Artificial expression constructs for selectively regulating gene expression in cortical excitatory neurons

Artificial expression constructs with specific promoters and enhancers enable precise and cost-effective labeling of cortical excitatory neurons, overcoming the limitations of existing technologies by providing high specificity and suitability for human applications.

JP7590962B2Active Publication Date: 2024-11-27ALLEN INSTITUTE
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
JP2021523074
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2019-07-09
Filing Date
2019-11-05
Publication Date
2024-11-27
Estimated Expiration
2039-11-05

AI Technical Summary

Technical Problem

Existing technologies for selectively labeling and interfering with specific cell types in the brain, such as cortical excitatory neurons, are expensive, require multiple transgenic crosses, and often result in heterogeneous cell type labeling, making them unsuitable for human applications.

Method used

Artificial expression constructs, including specific promoters and enhancers like eHGT_058h, eHGT_078h, and mscRE4, are used to selectively drive gene expression in targeted central nervous system cell populations, such as cortical excitatory neurons, allowing precise labeling and interference.

Benefits of technology

These constructs provide high specificity and efficiency in labeling targeted cell types, reducing the need for multiple transgenic crosses and enabling applications in humans.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007590962000013
    Figure 0007590962000013
  • Figure 0007590962000014
    Figure 0007590962000014
  • Figure 0007590962000015
    Figure 0007590962000015
Patent Text Reader

Abstract

Artificial expression constructs are described for selectively modulating gene expression in selected central nervous system cell types. The artificial expression constructs can be used to selectively express synthetic genes or modulate gene expression, particularly in cortical excitatory neurons, including those primarily within cortical layers 2 / 3, 4, 5, and 6, and those with extracerebral (ET), intracerebral (IT), and pyramidal tract (PT) projections.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Patent Application No. 62 / 755,988, filed November 5, 2018, No. 62 / 806,600, filed February 15, 2019, No. 62 / 806,684, filed February 15, 2019, and No. 62 / 872,021, filed July 9, 2019, each of which is incorporated by reference in its entirety as if fully set forth herein.

[0002] STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT This invention was made with Government support under Grant 1R01-DA036909 awarded by the National Institutes of Health. The Government has certain rights in the invention.

[0003] Reference to sequence listing The sequence listing associated with this application is provided in text format in lieu of a hard copy and is incorporated herein by reference. The name of the text file containing the sequence listing is A166-0007PCT_ST25.txt. The text file is 597KB, was created on November 5, 2019, and has been submitted electronically via EFS-Web.

[0004] The present disclosure provides an artificial expression construct for selectively driving gene expression in cortical excitatory neurons. The artificial expression construct can be used to selectively express synthetic genes or modulate gene expression in cortical excitatory neurons, particularly those primarily in cortical layers 2 / 3, 4, 5, and 6, and those with extracerebral (ET) processes, intracerebral (IT) processes, and pyramidal tract (PT) processes. [Background technology]

[0005] To fully understand the biology of the brain, it is necessary to distinguish and define different cell types, and to identify vectors that can selectively label and disrupt them in order to study them further. In mice, recombinase driver strains have been used to great effect to label cell populations that share marker gene expression. However, the creation, maintenance, and use of such strains to label cell types with high specificity can be expensive and often requires triple transgenic crosses that result in low frequencies of experimental animals. Moreover, these tools are not applicable to humans, as they require germline transgenic animals.

[0006] Recent advances in single-cell profiling, such as single-cell RNA-seq and investigations of neuronal electrophysiology and morphology, have revealed that many recombinant driver lines label a heterogeneous mixture of cell types, often including cells from multiple subclasses. For example, the Rbp4-Cre mouse driver line, commonly used to label layer 5 (L5) neurons, labels cells with very distinct connectivity patterns: L5 intratelencephalic (IT, also called cortico-cortical) and pyramidal tract (PT, also called cortico-subcortical) neurons. Summary of the Invention

[0007] The present disclosure provides artificial expression constructs that selectively drive gene expression in targeted central nervous system cell populations. Targeted central nervous system cell populations include cortical excitatory neurons, e.g., those primarily within cortical layers (L) 2 / 3, 4, 5, and / or 6, as well as those with extracerebral (ET) projections, intracerebral (IT) projections, and / or pyramidal tract (PT) projections. Certain artificial expression constructs disclosed herein target specific excitatory cell types, while others selectively drive gene expression across multiple excitatory neuron types.

[0008] For example, an artificial expression construct comprising a promoter, the eHGT_075h enhancer, and a gene encoding an expression product can result in selective gene expression in L2 / 3 IT cortical excitatory neurons.

[0009] An artificial expression construct comprising the promoter, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_439m, and / or eHGT_254h enhancer, and a gene encoding an expression product can result in selective gene expression in L4 IT cortical excitatory neurons.

[0010] Specific examples of artificial expression constructs that include a promoter, a mscRE4 enhancer, a linked mscRE4, and / or a linked mscRE16 enhancer, and a gene encoding an expression product can result in selective gene expression in L5 PT cortical excitatory neurons. Examples of these expression constructs include T502-057 (vAi3.0), 981 (vAi5.0), 1052 (vAi10.0), CN1818 (vAi128.0), CN2014 (vAi129.0), and vAi130.0.

[0011] An artificial expression construct containing a promoter, a linked core of the mscRE4 enhancer, and a gene encoding an expression product can result in selective gene expression in L5 PT and L5 ET cortical excitatory neurons.

[0012] Artificial expression constructs comprising a promoter, mscRE1, mscRE11, and / or mscRE16 enhancer, and a gene encoding an expression product can result in selective gene expression in L5 PT and L5 IT cortical excitatory neurons.

[0013] An artificial expression construct comprising a promoter, the mscRE13 enhancer, and a gene encoding an expression product can result in selective gene expression in L6 IT cortical excitatory neurons.

[0014] Specific examples of artificial expression constructs containing a promoter, the mscRE10 enhancer, and a gene encoding an expression product can result in selective gene expression in L6 CT cortical excitatory neurons. Examples include 995 (vAi15.0).

[0015] An artificial expression construct comprising a promoter, the eHGT_440h enhancer, and a gene encoding an expression product can result in selective gene expression in a subtype of L6b cortical excitatory neurons.

[0016] An artificial expression construct comprising a promoter, the eHGT_078h enhancer, and a gene encoding an expression product can result in selective gene expression in L2 / 3 IT, L4 IT, L5 IT, L5 NP, and L5 PT cortical excitatory neurons.

[0017] Selective expression of a gene encoding an expression product can be achieved in L2 / 3 IT, L5 IT, and L6b neurons using the 1036(vAi16.0) artificial expression construct described herein, which contains the mscRE10 enhancer.

[0018] Selective expression of genes encoding expression products can be achieved in L2 / 3 IT, L5 PT, L6 CT, and L6b neurons using the 988 (vAi7.1), 1010 (vAi6.1), and / or 1011 (vAi7.2) artificial expression constructs described herein. These constructs contain the mscRE4 enhancer.

[0019] Pan-excitatory and / or widespread expression in cortical excitatory neurons can be selectively achieved using an artificial expression construct comprising a promoter, eHGT_073h, eHGT_073m, eHGT_077h, and / or eHGT_078m enhancer, and a gene encoding an expression product. In certain embodiments, pan-excitatory expression refers to expression in at least four types of cortical excitatory cells that is limited to non-expression in inhibitory cells and glial cells.

[0020] The artificial expression constructs described herein can further label other distinct cell types. For example, in addition to L5 PT cells, an artificial expression construct containing a promoter, an mscRE4 enhancer, and a gene encoding an expression product can result in gene expression in subcortical populations in CEAc, substantia nigra, pars compacta (or pars compacta, SNc), and (ProS). Similarly, in addition to L5 PT cells, an artificial expression construct containing a promoter, a linking core of the mscRE4 enhancer, and a gene encoding an expression product can result in gene expression in the subiculum, CA1 pyramidal neurons, a subset of dentate gyrus granule cells, scattered striated neurons, and rare cerebellar Purkinje cells.

[0021] As shown by the above discussion, certain artificial expression constructs disclosed herein include engineered enhancers, such as the concatenated cores of mscRE4, eHGT_078h, and eHGT_078m enhancers, as well as concatemers of mscRE4 and mscRE16 enhancers. In relation to MscRE4, the functional 155 base pair (bp) core of the mscRE4 enhancer (SEQ ID NO: 29) was concatenated (SEQ ID NO: 30) to minimize the size required to drive gene expression. Despite being a 3x concatemer, SEQ ID NO: 30 is shorter in length than the original mscRE4 enhancer (SEQ ID NO: 28, which contains 555 bp). When used to construct artificial expression constructs such as rAAV, such concatemers allow more space for enhancer-bound cargo genes, which is highly desirable, for example, in gene therapy vectors. For example, many therapeutic cargo genes are too large to fit into AAV vector designs, so space (length of sequence) is at a premium.

[0022] As described in more detail throughout this disclosure, certain artificial expression constructs disclosed herein include T502-050, T502-054, vAi34.0, vAi33.2, vAi45.0, vAi1.0, T502-057, T502-059, TG978, TG981, TG988, TG995, TG996, TG999, TG1002, TG1010, TG1011, TG1021, TG1 036, TG1037, TG1038, TG1046, TG1047, TG1048, TG1049, TG1050, TG1052, CN1402, CN1457, CN1818, CN1416, CN1452, CN1461, CN1454, CN1456, CN1772, CN1427, CN1466, CN1954, CN1955, CN2137, CN2139, and CN2014.

[0023] Many of the drawings submitted herein are better understood in color, and Applicants reserve the right to consider color versions of the drawings as part of the original submission and to present color images of the drawings at a later proceeding. [Brief description of the drawings]

[0024] [Figure 1A] TG978 (vAi4.1). Enhancer mscRE4 (eAi3.0). Representative epifluorescence images of mscre4-FlpO-WPRE virus induced expression in the brain of Ai65F reporter mice. [Figure 1B] TG978 (vAi4.1). Enhancer mscRE4 (eAi3.0). Representative epifluorescence images of mscre4-FlpO-WPRE virus induced expression in the brain of Ai65F reporter mice. [Figure 1C] TG978 (vAi4.1). Enhancer mscRE4 (eAi3.0). Single-cell RNA sequencing analysis of tdTomato-positive cells isolated from primary visual cortex (V1) of mscre4-FlpO-infected Ai65F mice. L2 / 3, layer 2 / 3; L5, layer 5; wm, white matter. [Diagram 2]TG981 (vAi5.0) Enhancer mscRE4 (eAi3.0). Representative epifluorescence images of mscre4-EGFP-WPRE viral expression in wild-type mouse brain. Brain sections were stained with anti-GFP antibody to visualize GFP fluorescence. [Diagram 3] TG988 (vAi7.1) enhancer mscRE4 (eAi3.0). (A) Representative epifluorescence images of mscre4-tTA2 virus induced expression in the brains of Ai63 reporter mice. Brain sections were stained with anti-dsred antibody to reveal tdTomato fluorescence. (B) mscre4-tTA2 virus was directly injected into the brains of Ai63 mice, and native tdTomato fluorescence was imaged within the primary visual cortex (V1 or VISp). Note that image parameters between the two images may differ. L2 / 3, layer 2 / 3; L5, layer 5; wm, white matter. [Figure 4] A, B: Representative epifluorescence images of mscre4-iCre virus induced expression in the brain of TG1010 (vAi6.1) enhancer mscRE4 (eAi3.0) Ai14 reporter mice. L5, layer 5; L6, layer 6; wm, white matter. [Diagram 5] A, B: Representative epifluorescence images of the mscre4-tTA2 virus induced expression in the brain of TG1011 (vAi7.2) enhancer mscRE4 (eAi3.0) Ai63 reporter mice. [Figure 6] TG1021 (vAi8.0Cre) enhancer mscRE4 (eAi3.0). Representative epifluorescence images of Mscre4-Cre-WPRE virus induced expression in the brain of Ai14 reporter mice. [Figure 7] TG1052 (vAi10.0) Enhancer 4XmscRE16 (eAi11.1). Representative epifluorescence image of 4Xmscre16-EGFP-WPRE virus expression in wild-type mouse brain. Virus was delivered directly into the brain by stereotaxic injection. [Figure 8]A, B: Representative epifluorescence images of mscre10-EGFP-WPRE viral expression in the brain of a wild-type mouse with TG995 (vAi15.0) and enhancer mscRE10 (eAi6.0). [Figure 9A] TG1036 (vAi16.0) enhancer mscRE10 (eAi6.0). Representative epifluorescence images of mscre10-FlpO-WPRE virus induced expression in the brain of Ai65F reporter mice. [Figure 9B] TG1036 (vAi16.0) enhancer mscRE10 (eAi6.0). Representative epifluorescence images of mscre10-FlpO-WPRE virus induced expression in the brain of Ai65F reporter mice. [Figure 9C] TG1036 (vAi16.0) enhancer mscRE10 (eAi6.0). Single-cell RNA sequencing analysis of tdTomato-positive cells isolated from primary visual cortex (V1) of mscre10-FlpO-WPRE-infected Ai65F mice. [Figure 10] A, B: Representative epifluorescence images of the mscre10-tTA2-WPRE virus induced expression in the brain of TG1048 (vAi18.0) enhancer mscRE10 (eAi6.0) Ai63 reporter mice. [Figure 11] TG996 (vAi19.0) enhancer mscRE11 (eAi7.0). Representative epifluorescence image of mscre11-EGFP-WPRE virus in wild-type mouse brain. Brain sections were stained with anti-GFP antibody to reveal GFP fluorescence. [Figure 12] A, B: Representative epifluorescence images of TG999 (vAi21.0) enhancer mscRE13 (eAi9.0) and mscre13-EGFP-WPRE viruses in wild-type mouse brains. Brain sections were stained with anti-GFP antibody to reveal GFP fluorescence. [Figure 13A] TG1037 (vAi22.0) enhancer mscRE13 (eAi9.0). Representative epifluorescence images of mscre13-FlpO-WPRE virus induced expression in the brain of Ai65F reporter mice. [Figure 13B] TG1037 (vAi22.0) enhancer mscRE13 (eAi9.0). (B) Single-cell RNA sequencing analysis of tdTomato-positive cells isolated from the primary visual cortex (V1) of mscre13-FlpO-WPRE-infected Ai65F mice. Cell types from top to bottom are Lamp5 Pich2 Dock5, Lamp5 Lsp1, Vip Chat Htr1f, Sst Tac1 Htr1d, Sst Calb2 Pdlm5, Sst Nr2f2 Necab, Pvalb Sema3e Kank4, Pvalb Rein Itm2a, L2 / 3 IT VISp Rred, L2 / 3 IT VISp Adamts2, L2 / 3 IT VISp Agmat, L2 / 3 IT ALM Sla, L6 IT VISp Penk Col27a1, L6 IT VISp Penk Fst, L6 IT VISp Col18a1, L5 IT VISp Hsd11b1 Endou, L5 IT VISp Whrn Tox2, L5 IT VISp Col27a1, L5 PT VISp C1qI2 Cdh13, L5 PT VISp Krt80, L6 IT These include VISp Car3, L4 IT VISp Rspo1, intronic-rich VISp L5 Endou, L6 CT VISp Gpr139, L6 CT VISp Ctxn3 Brinp3, L6 CT VISp Ctxn3 Sla, and L6b VISp Mup5. [Figure 14] TG1046 (vAi23.0) enhancer mscRE13 (eAi9.0). Representative epifluorescence images of mscre13-iCre-WPRE virus induced expression in the brain of Ai14 reporter mice. [Figure 15] TG1049 (vAi24.0) enhancer mscRE13 (eAi9.0). Representative epifluorescence images of mscre13-tTA2-WPRE virus induced expression in the brain of Ai63 reporter mice. [Figure 16]A, B: Representative epifluorescence images of TG1002 (vAi26.0) enhancer mscRE16 (eAi11.0) and mscre16-EGFP-WPRE viruses in wild-type mouse brains. Brain sections were stained with anti-GFP antibody to reveal GFP fluorescence. [Figure 17A] TG1038 (vAi27.0) enhancer mscRE16 (eAi11.0). Representative epifluorescence images of mscre16-FlpO-WPRE virus induced expression in the brain of Ai65F reporter mice. [Figure 17B] TG1038 (vAi27.0) enhancer mscRE16 (eAi11.0). Representative epifluorescence images of mscre16-FlpO-WPRE virus induced expression in the brain of Ai65F reporter mice. [Figure 17C] TG1038 (vAi27.0) enhancer mscRE16 (eAi11.0). Single-cell RNA sequencing analysis of tdTomato-positive cells isolated from primary visual cortex (V1) of mscre16-FlpO-WPRE-infected Ai65F mice. Cell types from top to bottom are Lamp5 Pich2 Dock5, Lamp5 Lsp1, Sst Mme Fam114a1, L2 / 3 IT VISp Agmat, L6 IT VISp Agmat, L6 IT VISp Penk Fst, L6 IT VISp Col23a1, Adamts2, L6 IT VISp Col18a1, L5 IT VISp Hsd11b1 Endou, L5 IT VISp Whrn Tox2, L5 IT VISp Batf3, L5 IT VISp Col6a1 Fezf2, L5 IT ALM Tmem163 Arhgap25, L5 IT ALM Cpa6 Gpr88, L5 PT VISp C1qqI2 Cdh13, L5 PT VISp Krt80, Intronic VISp L5 Endou, L6CT VISp Ctxn3 Brinp3, L6CT VISp Ctxn3 Sla, and LowAqp4. [Figure 18]TG1047 (vAi28.0) enhancer (mscRE16 (eAi11.0). Representative epifluorescence images of mscre16-iCre-WPRE virus induced expression in the brain of Ai14 reporter mice. [Figure 19A] TG1050 (vAi29.0) enhancer mscRE16 (eAi11.0). Representative epifluorescence images of mscre16-tTA2-WPRE virus induced expression in the brain of Ai63 reporter mice. [Figure 19B] TG1050 (vAi29.0) enhancer mscRE16 (eAi11.0). Representative epifluorescence images of mscre16-tTA2-WPRE virus induced expression in the brain of Ai63 reporter mice. [Figure 20] TG1149 / (T502-050;vAi33.0) EnhancerGrik1-enhScnn1a-2(eAi14.0). Representative confocal images of Hsp68-EGFP-WPRE-Grik1-enhScnn1a-2 virus induced expression in wild-type mouse brain. [Figure 21] A, B: Representative confocal images of TG1108 (vAi34.0) enhancer Scnn1a(Grik1)(eAi14.0)-FlpO-WPRE virus induced expression in the brain of Ai65 reporter mice. [Figure 22] A, B, TG1114 (vAi33.2) enhancer Scnn1a (Grik1) (eAi14.0). Representative epifluorescence images of Scnn1a (Grik1)-EGFP-WPRE virus in wild-type mouse brain. Brain sections were stained with anti-GFP antibody to reveal GFP fluorescence. [Figure 23] TG1109 (vAi45.0) enhancer mscRE12 (eAi8.0). Representative epifluorescence images of mscre12-FlpO-WPRE virus induced expression in the brain of Ai65F reporter mice. [Figure 24A] CN1402 (vAi106.0) Enhancer eHGT_058h (eAi106.0). (A) Fluorescent expression of CN1402 shown in whole mouse brain in sagittal sections. [Figure 24B] CN1402 (vAi106.0) enhancer eHGT_058h (eAi106.0). High-resolution image (left) showing non-overlapping of CN1402 SYFP fluorescence (red) and inhibitory marker Gad1 mRNA expression (white). The image on the right shows near-competitive overlap of CN1402 SYFP fluorescence (red) and cortical excitatory marker Slc17a7 mRNA expression (white). [Figure 24C] CN1402 (vAi106.0) Enhancer eHGT_058h (eAi106.0). Quantification of the specificity of CN1402 SYFP fluorescence in ALM and V1 mouse cortical regions based on multiplexed FISH data. Single cell transcriptomic characterization of SYFP fluorescent cells isolated from mouse V1. [Figure 24D]CN1402 (vAi106.0) Enhancer eHGT_058h (eAi106.0). After single cell gene expression analysis, cells were mapped to existing taxonomy of mouse cell types. The position of the blue circle reflects the extent of single cell mapping (towards the terminal lobe) and the size of the blue circle reflects the number of single cells mapped to that point in the hierarchy. The downward projecting bar reflects the number of cells mapping to that terminal branch of the cell type taxonomy. Cells listed from left to right include 169 L2 / 3 IT VISp Rrad, 168 L2 / 3 IT VISp Adamts2, 167 L2 / 3 IT VISp Agmat, 164 L4 IT VISp Rspo1, 163 L5 IT VISp Hsd11b1 Endou, 162 L5 IT VISp Whrn Tox2, 160 L5 IT VISp Batf3, 158 L5 IT VISp Col6a1 Fezf2, 157 L5 IT VISp Col27a1, 154 L6 IT VISp Penk Col27a1, 153 L6 IT VISp Penk Fst, 152 L6 IT VISp Col23a1 Adamts2, 149 L6 IT VISp Col18a1, 146 L6 IT VISp Car3, 144 L5 PT VISp Chrna6, 143 L5 PT VISp Lgr5, 142 L5 PT VISp C1qI2 PTgfr, 141 L5 PT VISp C1qI2 Cdh13, 140 L5 PT VISp Krt80, 134 L5 NP VISp Trhr Cpne7, 133 L5 NP VISp Trhr Met, 131 L6 CT Nxph2 Sla, 130 L6 CT VISp Krt80 Sla, 128 L6 CT VISp Nxph2 Wls, 127 L6 CT VISp Ctxn3 Brinp3, 126 L6 CT VISp Ctxn3 Sla, 122 L6 CT VISp Gpr139, 120 L6b Col8a1 Rprm, 119 L6b VISp Mup5, 118 L6b VISp Col8a1 Rxfp1, 115 L6b P2ry12, 114 L6b VISp Crh, 110 Lamp5 Krt73, 109 Lamp5 Fam19a1Pax6、108 Lamp5 Fam19a1 Tmem182、106 Lamp5 Ntn1 Npy2r、105 Lamp5 Plch2 Dock5、101 Lamp5 Lsp1、100 Lamp5 Lhx6 97 Sncg Slc17a8、96 Sncg Vip Nptx2、95 Sncg Gpr50、93 Vip Itih5、90 Serpinf1 Clrn1、89 Serpinf1 Aqp5 Vip、85 Vip Igfbp6 Car10、84 Vip Igfbp6 Pltp、82 Vip Lmo1 Fam159b、81 Vip Lmo1 Myl1、79 Vip Igfbp6 Mab21I1、78 Vip Arhgap36 Hmcn1、77 Vip Gpc3 Slc18a3、74 Vip Ptprt Pkp2、73 Vip Rspo4 Rxfp1 Chat、71 Vip Lect1 Oxtr、70 Vip Rspo1 Itga4、67 Vip Chat Htr1f、66 Vip Pygm C1qI1、61 Vip CrispId2 Htr2c、60 Vip CrispId2 Kcne4、58 Vip Col15a1 Pde1a、54 Sst Chodl、53 Sst Mme Fam114a1、52 Sst Tac1 Htr1d、50 Sst Tac1 Tacr3、49 Sst Calb2 Necab1、48 Sst Calb2 Pdlim5、46 Sst Nr2f2 Necab1、45 Sst Myh8 Etv1、44 Sst Chrna2 Glra3、42 Sst Myh8 Pibin、40 Sst Chrna2 Ptgdr、39 Sst Tac2 Myh4、37 Sst Hpse Sema3c、36 Sst Hpse Cbln4、34 Sst Crhr2 Efemp1、33 Sst Crh 4930553C11Rik、31 Sst Esrn1、29 Sst Tac2 Tacstd2、28 Sst Rxfp1 Eya1、27 Sst Rsfp1 Prdm8、23 Sst Nts、21 Pvalb Gabrg1、20 Pvalb Th Sst、18 Pvalb Calb1 Sst、17 Pvalb Akr1c18 Ntf3、16 Pvalb Sema3eKank4, 14 Pvalb Gpr149 IsIr, 11 Pvalb Reln Itm2a, 10 Pvalb Reln Tac1, 9 Pvalb Tpbg, 4 Pvalb Vipr2, 1 Meis2 Adamts19, 170 Astro Aqp4 171 OPC Pdgfra Grm5, 173 Oligo Serpinb1a, 174 Oligo Synpr, 175 VLMC Osr1 Cd74, 176 VLMC Osr1 Mc5r, 177 VLMC Spp1 Col15a1, 178 Peri Kcni8, 179 SMC Acta2, 180 Endo Ctla2a, and 181 Microglia Siglech. [Figure 25A] CN1457 (vAi107.0) Enhancer eHGT_078h (eAi107.0). Fluorescent expression of CN1457 shown in whole mouse brain in sagittal sections. [Figure 25B] CN1457 (vAi107.0) enhancer eHGT_078h (eAi107.0). High-resolution image (left) showing non-overlapping CN1457 SYFP fluorescence (red) and inhibitory marker Gad1 mRNA expression (white). The image on the right shows near-competitive overlap of CN1457 SYFP fluorescence (red) and cortical excitatory marker Slc17a7 mRNA expression (white). [Figure 25C] CN1457 (vAi107.0) Enhancer eHGT_078h (eAi107.0). Quantification of the specificity of CN1457 SYFP fluorescence in ALM and V1 mouse cortical regions based on multiplexed FISH data. [Figure 25D]CN1457 (vAi107.0) Enhancer eHGT_078h (eAi107.0). Single-cell transcriptional characterization of SYFP fluorescent cells isolated from mouse V1. Following single-cell gene expression analysis, cells were mapped to existing taxonomy of mouse cell types. The position of the blue circle reflects the extent of single-cell mapping (towards the terminal lobe) and the size of the blue circle reflects the number of single cells mapped to that point in the hierarchy. The downward projecting bar reflects the number of cells mapping to that terminal branch of the cell type taxonomy. Cells are the same as those listed in the figure brief in Figure 24D. [Figure 26] CN1416 (vAi108.0) enhancer eHGT_058m (eAi108.0). (A) Fluorescent expression of CN1416 shown in whole mouse brain in sagittal sections. (B) High resolution image (left) showing non-overlapping of CN1416 SYFP fluorescence (red) and inhibitory marker Gad1 mRNA expression (white). The image on the right shows near competitive overlap of CN1416 SYFP fluorescence (red) and cortical excitatory marker Slc17a7 mRNA expression (white). (C) Quantification of specificity of CN1416 SYFP fluorescence in ALM and V1 mouse cortical regions based on multiplexed FISH data. [Figure 27] CN1452 (vAi111.0) enhancer eHGT_073h (eAi111.0). (A) Fluorescent expression of CN1452 shown in whole mouse brain in sagittal sections. (B) Grayscale fluorescent image of DAPI and mFISH image of Gad1, Pvalb, Sst, SYFP (CN1452) and Vip mRNA in mouse visual cortex. (C) Co-staining of SYFP (CN1452) and Gad1 shows that only 7% of Gad1+ cells overlap with SYFP. N=43 cells from 1 animal. [Figure 28]CN1461 (vAi112.0) enhancer eHGT_073m (eAi112.0). (A) Fluorescent expression of CN1461 shown in whole mouse brain in sagittal sections. (B) Grayscale fluorescent image of DAPI and mFISH image of Gad1, Pvalb, Sst, SYFP (CN1461), and Vip mRNA in mouse visual cortex. (C) Co-staining of SYFP (CN1461) and Gad1 shows that only 1.5% of Gad1+ cells overlap with SYFP. N=130 cells from 1 animal. [Figure 29] CN1454 (vAi113.0) enhancer eHGT_075h (eAi113.0). (A) Fluorescent expression of CN1454 shown in whole mouse brain in sagittal sections. (B) High-resolution image (left) showing non-overlapping of CN1454 SYFP fluorescence (red) and inhibitory marker Gad1 mRNA expression (white). The image on the right shows near-competitive overlap of CN1454 SYFP fluorescence (red) and cortical excitatory marker Slc17a7 mRNA expression (white). (C) Quantification of specificity of CN1454 SYFP fluorescence in the V1 mouse cortical region based on multiplexed FISH data. [Diagram 30] CN1456 (vAi114.0) enhancer eHGT_077h (eAi114.0). (A) Fluorescent expression of CN1402 shown in whole mouse brain in sagittal sections. (B) High resolution image (left) showing non-overlapping of CN1402 SYFP fluorescence (red) and inhibitory marker Gad1 mRNA expression (white). The image on the right shows near-competitive overlap of CN1402 SYFP fluorescence (red) and cortical excitability marker Slc17a7 mRNA expression (white). (C) Quantification of specificity of CN1402 SYFP fluorescence in ALM and V1 mouse cortical regions based on multiplexed FISH data. Single-cell transcriptomic characterization of SYFP fluorescent cells isolated from mouse V1. (D) Following single-cell gene expression analysis, cells were mapped to existing classifications of mouse cell types. The position of the blue circle reflects the extent of single-cell mapping (towards the terminal lobe) and the size of the blue circle reflects the number of single cells mapped to that point in the hierarchy. The downward projecting bars reflect the number of cells mapping to that terminal branch of the cell type classification scheme. [Figure 31A] CN1818 (vAi128.0) Enhancer mscRE4 (3xCore) (eAi3.2). Expression of construct CN1818 examined by native fluorescence microscopy of cells labeled by retro-orbital injection. [Figure 31B] CN1818 (vAi128.0) enhancer mscRE4 (3xCore) (eAi3.2). Expression of construct CN1818 tested by hairpin chain reaction (HCR) RNA FISH targeting SYFP2 (derived from viral expression), Fam84b (expressed in L5 ET cells), and Rorb (expressed in L4 IT and L5 IT cells). FISH revealed a specificity rate of 78% in situ (62 Fam84b+ and SYFP2+ / 80 total SYFP2+). [Figure 32A] CN2014 (vAi129.0) enhancer mscRE4 (eAi3.0). Expression of construct CN2014 examined by native fluorescence microscopy of cells labeled by retro-orbital injection. [Figure 32B] CN2014 (vAi129.0) enhancer mscRE4 (eAi3.0). Expression of construct CN2014 tested by hairpin chain reaction (HCR) RNA FISH targeting SYFP2 (from viral expression), Fam84b (expressed in L5 ET cells), and Rorb (expressed in L4 IT and L5 IT cells). FISH revealed a specificity rate of 85% in situ (45 Fam84b+ and SYFP2+ / 53 total SYFP2+). [Diagram 33] CN1427 (vAi130.0) Enhancer mscRE4 (4x) (eAi3.1). Native tdTomato fluorescent expression in the V1 region of mouse brain slices. CN1427 serotype PHPeB virus was delivered by retro-orbital injection with analysis of reporter transgene expression 40 days after injection. [Diagram 34]CN1466 (vAi131.0) enhancer eHGT_078m (eAi128.0). (A) Expression of vector CN1466 (green) in mouse neocortical brain slice cultures after 25 days in vitro and 15 days post-infection. Mutually exclusive labeling CN1466 labeled neurons (green) and GABAergic neurons (red). (B) Expression of vector CN1466 in human neocortical brain slice cultures at 9 days in vitro and 9 days post-infection. Extensive pyramidal neuron labeling was observed. [Diagram 35] CN2139 (vAi134.0) Enhancer eHGT_439m (eAi131.0). Expression of vector CN2139 by retro-orbital delivery in mouse brain. Brain slices were subjected to fixed tissue immunohistochemistry with anti-GFP and anti-CTIP2 antibodies. Virus-labeled cells were observed in L4 of the neocortex. [Diagram 36] CN2137 (vAi135.0) Enhancer eHGT_440h (eAi132.0). Expression of vector CN2137 by retro-orbital delivery in mouse brain. Brain slices were subjected to fixed tissue immunohistochemistry with anti-GFP and anti-CTIP2 antibodies. Virus-labeled cells were observed in L6b of the neocortex. [Figure 37] (A) CN1954 (vAi132.0) Enhancer eHGT_078h (3x core) (eAi129.0). Expression of vector CN1954 in mouse neocortical brain slice cultures at 27 days in vitro and 20 days post-infection. (B) CN1955 (vAi133.0) Enhancer eHGT_078m (3x core) (eAi130.0). Expression of vector CN1955 in mouse neocortical brain slice cultures at 27 days in vitro and 20 days post-infection. [Figure 38] (A) vAi1.0 enhancer mscRE1 (eAi1.0). A) Expression of construct mscRE1-SYFP2 examined by natural fluorescence imaging of retro-orbital injection. (B) T502-059 (vAi2.0) enhancer mscRE1 (eAi2.0). A) Expression of construct mscRE3-SYFP2 examined by natural fluorescence imaging of retro-orbital injection. [Figure 39]A-D, T502-057 (vAi3.0) enhancer mscRE4 (eAi3.0). Expression of construct mscRE4-SYFP2 examined by native fluorescence imaging of retro-orbital injection. [Diagram 40] Cell source and quality control (QC statistics). Bar plots showing the number of cells flagged for each combination of QC criteria. N, number of cells collected. Unique fragments is the number of uniquely mapped fragments used in the analysis and was used for the first QC cutoff of >1e4 unique fragments. The fraction of fragments overlapping with the Encyclopedia of DNA Elements (Code) DNase-seq peak was calculated using uniquely mapped fragments and used for the second QC cutoff of >0.25. The fraction of fragments >250 bp in length was calculated using unique fragments providing a third QC cutoff of >0.1. [Diagram 41] Overview of enhancer discovery for viral tools. To build cell type-specific labeling tools, we isolated cells from adult mouse cortex and performed single-cell assays for transposase-accessible chromatin using sequencing (scATAC-seq). Samples were clustered and compared to single-cell RNA sequencing (scRNA-seq) datasets to identify clusters. Single cells matching the same transcriptome type were then pooled and the genomes searched for type-specific putative enhancers. These regions were cloned upstream of a minimal promoter in an adeno-associated virus (AAV) genome backbone, which was used to generate self-complementary adeno-associated virus vectors (scAAV) or recombinant adeno-associated virus vectors (rAAV). These viral tools were delivered retroorbitally or stereotaxically to label specific cortical populations. In cells with matching cell types, enhancers recruit their cognate transcription factors to drive cell type-specific expression. In other cells, the viral genome is present but the transcript is not expressed. However, not all enhancers work as expected, so enhancer constructs need to be tested for specificity. [Diagram 42]Fluorescence-activated cell sorting (FACS) gating examples. (4A) All FACS sorting followed a similar gating strategy: morphology and debris removal using forward scatter area (FSC-A) and side scatter area (SSC-A); removal of doublets / multiplets using forward scatter width (FSC-W) × forward scatter height (FSC-H) and side scatter width (SSC-W) × side scatter height (SSC-H) gating; and selection of live cells with or without fluorescent labeling using 4',6-diamidin-2-phenylindole (DAPI) and fluorophore signals. This panel shows an example of gating for direct fluorophore labeling of cells from injection of mscRE4-SYFP2. [Diagram 43]Comparison of Gm12878 platforms. Comparison of FACS-sorted scATAC-seq libraries with those previously generated using Fluidigm C1 (Buenrostro et al. Nature, 523 (2015)), sci-ATAC-seq (Cusanovich et al. Science 348 (2015) and Pliner et al., Mol. Cell 71 (2018)), or droplet-based indexing (10X Genomics) that allows data using a common cell line of human Gm12878 cells. For use in these comparisons, scATAC-seq data was generated using a FACS-based method for 60 Gm12878 cells. For each published dataset, raw data were obtained from GEO and aligned and analyzed using the same methods. For 10X Genome, aligned fragment positions and metadata were obtained from the 10X Genome website. Abbreviations used throughout plots: bu, Buenrostro et al., Nature 523 (2015) Fluidigm C1 ATAC-seq; cu, Cusanovich et al., Science 348 (2015) sci-ATAC-seq (2015); gr, Graybuck et al. (data presented herein) FACS scATAC-seq; pl, Pliner et al., Mol. Cell 71 (2018) sci-ATAC-seq (2018); and tx, 10X Genomics, 5k cells 10X ATAC-seq. Gray et al., Elife 1-30 (2017). Dual-axis QC criteria plot showing QC1 and QC2 cutoffs used for mouse cortical scATAC-seq data. [Diagram 44]Gm12878 platform comparison. Summary fragment length frequency plots. Fragment length is shown on the x-axis, and the fraction of reads with fragments of each bp size was calculated for each sample in each dataset. For this analysis, the median fraction at each fragment size is shown as a solid line, and the 25th and 75th percentiles are shown in the shaded areas. Abbreviations used throughout plots: bu, Buenrostro et al., Nature 523 (2015) Fluidigm C1 ATAC-seq; cu, Cusanovich et al., Science 348 (2015) sci-ATAC-seq (2015); gr, Graybuck et al. (data described herein) [Diagram 45] Samples were clustered in t-SNE space using a phenograph implementation of Louvain clustering. To identify cell types within these clusters, cells from each cluster were pooled and the number of fragments within 20 kb of each TSS was counted. Marker genes for transcriptome clusters were then selected from Tasic et al., Nature 563, 72-78 (2018) and correlation of TSS accessibility scores with log-transformed gene expression was performed. The scRNA-seq cluster with the highest correlation score was assigned as the identity of each phenotype cluster, and clusters with the same transcriptome mapping were combined for downstream analysis. The cluster with the highest correlation score was assigned as the identity of each cluster, and clusters with the same transcriptome mapping were combined for downstream analysis. [Figure 46A]scATAC-seq data. Dot plots show both the fraction of cells in each subclass expressing each gene (size of dot) and the median expression level within each subclass (color of dot). scATAC-seq data were grouped by subclass based on transcriptomic mapping, and aggregated fragment overlaps near genes of interest were plotted after fragment count normalization (track plot, right panel). Subclass-level gene expression profiles (dot plot, left panel) from Tasic et al. (2018, Nature) show highly specific expression of the Fam84b gene in the L5 PT subclass. Fam84b (family with sequence similarity 74, member B) is a transcription factor gene that was recently shown to be a highly selective marker gene for L5 PT neurons across two regions of mouse cortex (Tasic et al. (2018) Nature). We identified a peak of accessibility specific to L5 PT samples (mscRE4) 113 kb downstream from the Fam84b TSS. [Figure 46B]scATAC-seq data. Dot plots show both the fraction of cells in each subclass expressing each gene (size of dot) and the median expression level within each subclass (color of dot). scATAC-seq data were grouped by subclass based on transcriptomic mapping, and aggregated fragment overlaps near genes of interest were plotted after fragment count normalization (track plot, right panel). Subclass-level gene expression profile (dot plot, left panel) from Tasic et al. (2018) shows enrichment of Hsd11b1 expression in L5 IT and L5 PT cell types. Hsd11b1 (hydroxysteroid 11-beta dehydrogenase 1) is a gene involved in corticosteroid biosynthesis. It has been shown to be selectively expressed in L5 cells, with higher expression in some L5 IT types than in L5 PT cells {Tasic et al. (2018) Nature}. A peak of accessibility enriched in L5 IT cells but absent in L5 PT cells (mscRE16) was identified 34 kb upstream of the Hsd11b1 TSS. The cell types listed along the sides of Figures 46A and 46B are Lamp5, Sncg, Serpinf1, Vip, Sst Pvalb, L2 / 3 IT, L4, L5 IT, L6 IT, L5 PT, NP, L6 CT, L6b, Meis2, and CR. [Figure 46C] scATAC-seq data. Dot plots show both the fraction of cells in each subclass expressing each gene (dot size) and the median expression level within each subclass (dot color). scATAC-seq data were grouped by subclass based on transcriptomic mapping, and aggregated fragment overlaps near genes of interest were plotted after fragment count normalization (track plots, right panel). scATAC-seq data showed a peak of accessibility specific to mscRE10, located 34 kb upstream of Car3. [Figure 46D]scATAC-seq data. Dot plots show both the fraction of cells in each subclass expressing each gene (size of dot) and the median expression level within each subclass (color of dot). scATAC-seq data were grouped by subclass based on transcriptomic mapping, and aggregated fragment overlaps near genes of interest were plotted after fragment count normalization (track plot, right panel). (D) scATAC-seq data showed a peak of accessibility specific to mscRE13, located 86 kb upstream of Osr1. [Figure 47] mscRE location and cloning primers. [Figure 48] (A) Direct enhancer-driven expression of fluorophores was tested by cloning the putative mscRE4 or mscRE16 enhancers in scAAV constructs with a minimal promoter driving the fluorophore-WPRE3. After packaging, purified, and titrated scAAV was injected retro-orbitally into wild-type mice. (B) Two weeks after retro-orbital injection of rAAV with mscRE16 driving expression of EGFP (TG1002), cells were selectively labeled in L5 of the mouse cortex by EGFP expression, which is amplified here using antibody staining by immunohistochemistry (IHC). (C) Two weeks after retro-orbital injection of scAAV with mscRE4 driving expression of SYFP (T502-057), dim but distinct labeling of L5 PT cells was seen by native fluorescence without antibody amplification. [Figure 49A] Validation of cell type targeting of scAAV-mscRE4-SYFP2 virus by scRNA-seq. Enhancer-driven recombinase expression was tested using a scAAV construct with a minimal promoter driving EGFP-WPRE3. After packaging, mice were injected retro-orbitally. After 2 weeks, SYFP-expressing cells were visible in the cortex, which could be isolated by FACS and used for scRNA-seq. [Figure 49B]Validation of cell type targeting of scAAV-mscRE4-SYFP2 viruses by scRNA-seq. Centroid classifier mapping of labeled cells to data from Tasic et al. (2018, Nature) revealed 91.8% of cells mapped to the L5 PT transcriptome cell type. [Figure 50A] Electrophysiological characterization of mscRE4 labeled cells and demonstration of utility for electrophysiological recording of labeled neurons. Cortical slices from animals labeled with scAAV-mscRE4-SYPF2 (T502-057) virus were used for electrophysiological characterization. Example impedance amplitude profiles obtained from (yellow fluorescent protein) YFP+ and YFP- neurons in the VISp. For comparison, impedance amplitude profiles from unlabeled PT-like and IT-like neurons from somatosensory cortex are also shown. Resonance frequency is plotted as a function of input resistance. [Figure 50B] Electrophysiological characterization of mscRE4 labeled cells and demonstration of utility for electrophysiological recording of labeled neurons. Example voltage responses to a series of hyperpolarizing and depolarizing current injections of YFP+ and YFP- neurons. For reference, exemplary voltage responses obtained from unlabeled PT-like and IT-like neurons are also shown. [Figure 50C] Electrophysiological characterization of mscRE4 labeled cells and demonstration of the utility for electrophysiological recordings of labeled neurons. Input resistance, perpendicular ratio and resonant frequency for the three experimental conditions. [Figure 51] Additional electrophysiological characteristics of mscRE4-SYFP2 labeled cells. (A) Microscopy of example cells characterized by patch electrophysiology. On the left, a SYFP2 positive cell, on the right, a SFYP2 negative cell. (B) Input resistance, perpendicular ratio, and resonant frequency for the four experimental conditions: IT, YFP-, YFP+, and PT. [Figure 52]Stereotaxic labeling using enhancer-driven viruses. (A) Native fluorescence image of an animal with stereotaxic injection of mscRE4-EGFP in the primary visual cortex. The enhancer-driven virus was co-injected with a constitutive dTomato virus, rAAVDJ-EF1a-dTomato, at 0.1x the volume of the mscRE virus to provide location of the injection site (dotted outline). (B) Native fluorescence image of an animal stereotaxically injected with mscRE4-SYFP2 in the volumes indicated into the primary visual cortex. [Diagram 53] Several enhancer-driven recombinase viruses provide specific binary labeling. Three different recombinases and one transactivator were inserted downstream of the mscRE4 and promoter in the viral construct. After retro-orbital injection, labeling of L5 was found with various degrees of specificity using tTA2 (TG1011, SEQ ID NO: 88) in Ai63 reporter mice (rarest, most specific), FlpO (TG978, SEQ ID NO: 80) in Ai65F reporter mice (most complete and specific), iCre (TG1010, SEQ ID NO: 87) in Ai14 reporter mice (complete, but with background in L6), and dgCre in Ai14 reporter mice (least specific). Images show native fluorescence in the visual cortex 2 weeks after injection. For depictions, see Figure 68A-G. [Figure 54] Whole-brain images of retroorbitally delivered mscRE4-FlpO-WPRE3 (TG978, SEQ ID NO: 80) viral labeling reveal specific L5-restricted labeling throughout the cortex, and specific subcortical populations in the central amygdala nucleus, labeling of parts of the CeA, the pars saccularis (CEAc) that receives and processes pain signals; the substantia nigra compacta (or pars compacta, SNc), involved in movement control and affected in Parkinson's disease; as well as the prosbiculum (ProS). [Figure 55]Validation of cell type targeting of mscRE4-FlpO-WPRE3 virus by scRNA-seq. (A) Enhancer-driven recombinase expression was tested using a rAAV construct with a minimal promoter driving FlpO-WPRE3. After packaging, Ai65F mice were injected retro-orbitally. After 2 weeks, tdTomato-expressing cells were visible in the cortex, which could be isolated from L5 dissociations, sorted by FACS, and used for scRNA-seq. (B) Centroid classifier mapping of labeled cells to data from Tasic, et al. (2018, Nature) revealed 90.6% of cells mapped to L5 PT transcriptome cell types. The list of cell types along the right side from top to bottom is as follows: Sst Hpse Cbln4(3), L5 IT VISp Hsd11b1 Endou(2), L5 PT VISp Chrna6(2), L5 PT VISp Lgr5(2), L5 PT VISp C1ql2 Ptgfr(40), L5 PT VISp C1ql2 Cdh13(40), and L5 PT VISp Krt80(7). [Figure 56A] Double labeling and titration of viral copy number to achieve high titer and broad cross-sectional labeling. These experiments were performed by retro-orbital co-injection of mscRE4-FLPO (TG978, SEQ ID NO: 80) and mscRE16-EGFP (TG1002, SEQ ID NO: 86) viruses into FLP-dependent tdTOMATO reporter mice (Ai65F). See FIG. 68 for a depiction of this double labeling strategy. Corners of the fluorescent images identify fluorophores (anti-GFP, native tdTomato, and integration). [Figure 56B]Double labeling and titration of viral copy number to achieve high titer and broad cross-sectional labeling. These experiments were performed by retro-orbital co-injection of mscRE4-FLPO (TG978, SEQ ID NO: 80) and mscRE16-EGFP (TG1002, SEQ ID NO: 86) viruses into FLP-dependent tdTOMATO reporter mice (Ai65F). See FIG. 68 for a depiction of this double labeling strategy. Corners of the fluorescent images identify fluorophores (anti-GFP, native tdTomato, and integration). [Figure 56C] Double labeling and titration of viral copy number to achieve specific and exclusive labeling at low titers. These experiments were performed by retro-orbital co-injection of mscRE4-FLPO (TG978, SEQ ID NO: 80) and mscRE16-EGFP (TG1002, SEQ ID NO: 86) viruses into FLP-dependent tdTOMATO reporter mice (Ai65F). See FIG. 68 for a depiction of this double labeling strategy. Corners of the fluorescent images identify the fluorophores (anti-GFP, native tdTomato, and integrated). [Figure 57A] Enhancer-driven recombinase virus as a driver for cell labeling. Whole section image of mscRE4-FlpO injection shows labeling throughout L5 of the posterior cortex. The inset region on the right corresponds to the white box on the left. Layer overlay from the Allen Brain Reference Atlas shows that labeling is restricted to L5. tdTomato+ cells were dissected from all cortical depths and collected by FACS for scRNA-seq. Transcriptomic profiles were mapped to reference cell types from Tasic et al. (2018). 87.5% (28 of 32) of cells mapped to the L5 PT cell type. [Figure 57B]Enhancer-driven recombinase viruses as drivers for cell labeling. Full section images of mscRE10-FlpO injections show labeling in layer 6 (L6) of the cortex. The inset region on the right corresponds to the white box on the left. scRNA-seq of tdTomato+ cells shows that layer 6 corticothalamic (L6 CT) and L6b cell types were the most frequently labeled neuronal subclasses at 75% (n=36 of 48). [Figure 57C] Enhancer-driven recombinase viruses as drivers for cell labeling. Full section images of mscRE16-FlpO injections show labeling in L5 of the cortex. The inset region on the right corresponds to the white box on the left. scRNA-seq of tdTomato+ cells shows that the L5 IT cell type is the most frequently labeled neuronal subclass at 42% (n=20 of 48), although other subclasses are also labeled at this titer (Lamp5, 27%; L6 IT, 6%; L5 PT, 15%). [Figure 58] Retroorbital mscRE driver screening at multiple titers. Native fluorescence images of reporter mice retroorbitally (RO) injected with enhancer-driven recombinase virus at two titers: low RO, 1x10 genome copies, GC; high RO, 1x10 GC. Fluorescence is tdTomato. Scale bar size can be determined by the Scale Bar Key, where triangles indicate a scale of 100 μm, 7-point stars indicate a scale of 500 μm, and 5-point stars indicate a scale of 1000 μm. Arrows indicate where the layers are labeled, and the direction of the arrow indicates where the layers are labeled: L1, L2 / 3, L4, L5, L6, and L6b. [Figure 59A]Whole brain and cross-labeling of cell types. Whole brain imaging results using TissueCyte. Whole brain sections from Ai65F mice after retro-orbital injection of MscRE-FlpO were aligned to the Allen Institute Common Coordinate Framework (CCF) and mapped to the Allen Brain Atlas structural ontology. A high level overview of cell labeling across the structural ontology is presented by the classification plot. "Grey" is the root of the plot representing all grey matter regions, and each branch of a node indicates a child structure within each region. The size and color of the node represents the maximum signal found among all children of the node, allowing one to follow the tree to the source of high signal within each structure. The inserts display selected regions of high or specific signal. Region initials correspond to the Allen Brain Adult Mouse Atlas. [Figure 59B] Whole brain and cross-labeling of cell types. Further division of isocortical regions in the TissueCyte dataset down to the level of cortical layers allows for brain-wide quantification of layer-specific signal. A representative cortical section from the TissueCyte dataset is shown at the top along from most anterior to most posterior (left to right). Heatmaps show quantification of signal in each region and layer. Agranular regions lacking layer 4 have a hash in the L4 column. The front and back areas are FRP, ORBv1, ORBm, ORBI, PL, ILA, Aid, Mos, Alv, Mop, SSp-m, GU, ACAd, SSp-n, SSp-un, ACAv, SSp-uI, SSp-II, VISC , Aip, SSs, SSp-bfd, SSp-tr, AUDv, AUDd, AUDp, PTLp, RSPv, RSPd, PERI, VISam, TE, ECT, AUDpo, VISI, VISpm, and VISpI. [Figure 59C]Whole brain and cross-labeling of cell types. Diagram showing the use of coinjected recombinase viruses in a dual reporter system for co- or cross-labeling of cell types. In this experiment, one virus driving FlpO and a second virus driving iCre are co-injected into mice with genetically encoded Flp- and Cre-dependent reporters. In the target cell type, the enhancer drives the recombinase, permanently labeling that target cell type. If the enhancers selected are mutually exclusive, different populations are labeled. If they overlap, cross-labeling is possible. [Fig. 59D] Whole brain and cross-labeling of cell types. Native fluorescence images of Ai65F, Ai140 dual reporter mouse strains injected retro-orbitally with mscRE16-FlpO (red fluorescence) and mscRE4-iCre (green fluorescence). These enhancers are expected to label mutually exclusive cell types in L5 of the cortex. The white boxed area corresponds to the inset image, showing strong labeling of the L5 strain. [Figure 59E] Whole brain and cross-labeling of cell types. Number of cells in each layer for all cortical regions labeled with EGFP (mscRE4; L5 PT), tdTomato (mscre16; L5 IT), or both in the images in (Figure 59D). [Figure 60A] Whole-brain characterization of mscRE16-FlpO. TissueCyte images of mscRE16-FlpO, Ai65F mice 2 weeks after retro-orbital injection were registered to the Allen Institute Common Coordinate Framework (CCF) and each structure was scored on the adult mouse structures. For Figures 59A-E, these panels provide a high-level overview of cellular labeling across the entire structural ontology. The size and color of the node represent the maximum signal found among all children of the node, allowing one to follow the tree to the source of high signal within each structure. Insets display selected regions of high or specific signal. The bottom left inset shows projections of IT neurons across the corpus callosum. [Figure 60B]Whole-brain characterization of mscRE16-FlpO. Layer quantification on the same TissueCyte images registered to CCF for all isocortical regions. Agranular regions lacking L4 are indicated by white boxes in the L4 row. All initials correspond to the Allen Institute for Brain Science adult mouse 3D atlas. [Figure 61A] mscRE4 AAV vectors targeting rare L5 PT neurons in the human cortex. Human acute slice cultures dissected from the middle temporal gyrus (MTG) were infected with a quartet of viruses, i.e., two mscRE4-driven rAAVs expressing Cre or Flp recombinase, and two fluorescent reporter viruses, one expressing SYFP and the other expressing RFP. This strategy allows for high specificity by selecting only co-labeled neurons. Biocytin loading of co-labeled cells used for patch electrophysiology reveals a morphology consistent with human L5 PT neurons. [Figure 61B] mscRE4 AAV vectors targeting rare L5 PT neurons in the human cortex. Human acute slice cultures dissected from the middle temporal gyrus (MTG) were infected with a quartet of viruses, i.e., two mscRE4-driven rAAVs expressing Cre or Flp recombinase, and two fluorescent reporter viruses, one expressing SYFP and the other expressing RFP. This strategy allows high specificity by selecting only co-labeled neurons. Double fluorescent labeling of L5 PT neurons in the human cortex (Scale bar is 100 microns). [Figure 61C]mscRE4 AAV vectors targeting rare L5 PT neurons in the human cortex. Human acute slice cultures dissected from the middle temporal gyrus (MTG) were infected with a quartet of viruses, i.e., two mscRE4-driven rAAVs expressing Cre or Flp recombinase, and two fluorescent reporter viruses, one expressing SYFP and the other expressing RFP. This strategy allows high specificity by selecting only co-labeled neurons. Transcript validation was performed by mapping RNA extracted from labeled cells using patch sequences. RNA was reverse transcribed, amplified, sequenced, and mapped to a human MTG reference dataset, and a bootstrapped centroid classifier was used to match human L4 / 5 PT cell types in 100 out of 100 trials. [Fig.61D] mscRE4 AAV vectors targeting rare L5 PT neurons in the human cortex. Human acute slice cultures dissected from the middle temporal gyrus (MTG) were infected with a quartet of viruses, i.e., two mscRE4-driven rAAVs expressing Cre or Flp recombinase, and two fluorescent reporter viruses, one expressing SYFP and the other expressing RFP. This strategy allows for high specificity by selecting only co-labeled neurons. The electrophysiological properties of the co-labeled human L5 PT cells are consistent with previous studies of L5 PT cells, demonstrating the utility of this method for selective electrophysiological targeting. [Figure 62] Annotated sequence of CN1818 [Figure 63] 3xCore-mscRE4-SYFP2 virus (CN1818, SEQ ID NO: 109) was injected retro-orbitally into adult mice. Three weeks after injection, brains from injected mice were sectioned and imaged to assess targeted expression of SYFP2 fluorophore labeling. Strong expression of the SYFP2 reporter gene in adult mouse brain was observed after retro-orbital injection of CN1818. Labeled cells were primarily in layer 5 and had electrophysiological properties consistent with L5 PT neurons. [Figure 64](A) Nissl staining of the M1 region in a macaque brain slice showing neocortical layers and higher magnification of the boxed region showing numerous magnopyramidal Betz cells (white arrows). (B) Native YFP expression detected in Betz cells (white arrows) 4 days after infection with CN1818 and the corresponding Nissl staining of the same field. [Figure 65] (A) Expected viral labeling (green) and targeted patch clamp recording of a putative Betz cell in a cultured macaque M1 brain slice infected with CN1818, using Alexa dye filling (red) from a patch pipette. (B) Firing in response to a 1 sec, 3 nA current injection step showing a narrow action potential width. (C) Summary plot showing high firing rates in response to escalating current injection steps. [Figure 66A] Spike frequency acceleration and subthreshold membrane potential oscillations in the gamma band shown for CN1818 virally labeled macaque M1 putative Betz cells. [Figure 66B] Significant high speed sag, low input resistance (19MOhms). [Figure 66C] Subthreshold membrane resonance with peak resonant frequency of 5.3 Hz. [Figure 67] 3xCore-mscRE4-SYFP2 virus (CN1818, SEQ ID NO: 109) was applied to human surgical ex vivo cortical slice cultures. After incubation, the cortical slices were imaged under a microscope to assess target expression of the SYFP2 fluorophore label. CN1818 was found to label L5 PT neurons in human ex vivo neocortical brain slice cultures. The scale bar is 1 mm long. [Figure 68A] FIG. 1 shows a conventional Cre / lox system for generating cell type-specific labeling by breeding. [Figure 68B] FIG. 1 shows a conventional Flp / FRT system for generating cell type-specific labeling by breeding. [Figure 68C] FIG. 1 shows a conventional TET transactivator / TET responsive element (tTA2 / TRE) system used to generate cell type-specific labeling. [Figure 68D]Shows a mechanism for bypassing propagation by replacing the viral Cre driver. [Figure 68E] Shows a mechanism for bypassing propagation by replacing the viral Flp driver. [Fig. 68F] We demonstrate a mechanism for bypassing propagation by replacing the viral tTA driver and an additional layer of regulation via doxycycline treatment that can reduce or inactivate tTA2 activity. [Figure 68G] We show that direct labeling bypasses these systems entirely. A strong advantage of Cre- or Flp-dependent reporters is that they can be much brighter and are permanently turned on after recombination to remove the stop site. The tTA2 / TRE system is an additional mechanism for selective labeling that can be tunable by doxycycline treatment. [Figure 69] Diagrammatic overview of a multi-viral labeling system, where two different viruses, driven by the same or different enhancers, drive either a recombinase or a fluorophore. Upon injection into a reporter mouse, the enhancer-driven recombinase triggers excision of the arrest site in the target cell type, and the enhancer-driven fluorophore is directly expressed in another target cell type. If these cell types overlap in their use of viral enhancer elements, cross-reactive co-labeling can be observed. [Figure 70] Enhancer ID, labeled cell type, and validation method. [Figure 71] Summary of Vector Components. Sequence names, relative lengths, enhancers, promoters, product classes, primary products, and other components of the expression constructs described herein. [Figure 72] Classification and clustering of selected central nervous system cells. [Figure 73A] Enhancer targeting validation data. FM stands for Fluorescence Microscopy. [Figure 73B]Cell type specificity of the enhancers and vectors described herein. S = subset of types within a group; A = all types within a group; * = validated in mouse, RNA-seq, and a third modality; ~ = validated in mouse, RNA-seq, primate / human, and a fourth modality. [Figure 74] 1 is a schematic diagram of cortical layers with particular relevance to the primate visual cortex. This schematic is provided as an illustration of intracortical layers. [Fig. 75A] A database of human neocortical cell subclass-specific accessible chromatin elements. (A) Workflow for human neocortex epigenetic characterization. [Fig. 75B] Database of human neocortical cell subclass-specific accessible chromatin elements. High-quality nuclei (2858 from 14 specimens) visualized by tSNE and colored according to the transcriptome cell type (B), sorting strategy (C), or specimen (D) L, layer. (E) Transcriptomic abundance of 11 known cell subclass-specific marker genes across 75 cell types identified in the human middle temporal gyrus (Hodge et al., bioRxiv, 384826, 2018). [Figure 76] Mapping ATAC-seq clusters to RNA-seq cell types. The transcriptome cell types within a subclass were summed for cluster-wise mapping to obtain a cluster-wise mapping to subclasses. This plot represents the final mapped subclasses assigned as the most frequent mapping for each cluster, and these subclass identities are used for the pileup and calculations in Figures 75B, 77, and 78. [Figure 77] Characterization of human neocortical cell subclass-specific accessible genomic elements. Percent overlap between ATAC-seq peaks and previously identified DMRs (Lister et al., Science.341, 1237905, 2013; Luo et al., Science.357, 600-604, 2017), comparing actual peaks and randomized peak positions. Absolute numbers of detected peaks and peak-DMR overlaps are shown. [Figure 78] Accessible chromatin elements provide human genetic tools. Multiple enhancer-AAV vectors provide distinct subclass selectivity. Seven loci and ATAC-seq read pile-ups and expression patterns in mouse V1 for the seven AAV reporter vectors are shown. Scale 200 μm. [Figure 79]Sequences supporting the present disclosure. Enhancer Grik1-enhScnn1a-1 short form (SEQ ID NO: 188), Enhancer Grik1-enhScnn1a-1 (eAi14.0) (SEQ ID NO: 25), Enhancer mscRE1 (eAi1.0) (SEQ ID NO: 26), Enhancer mscRE3 (eAi2.0) (SEQ ID NO: 27), Enhancer mscRE4 (eAi3.0) (SEQ ID NO: 28), Enhancer mscRE4 core (SEQ ID NO: 29), Enhancer 3xmscRE4 core (eAi3.2) (SEQ ID NO: 30), enhancer mscRE4 (4x) (eAi3.1) (SEQ ID NO: 31), enhancer mscRE10 (eAi6.0) (SEQ ID NO: 32), enhancer mscRE11 (eAi7.0) (SEQ ID NO: 33), enhancer mscRE12 long form (SEQ ID NO: 34), enhancer mscre12 (eAi8.0) (SEQ ID NO: 35), enhancer mscRE13 (eAi9.0) (SEQ ID NO: 36), enhancer mscRE16 (eAi11.0) (SEQ ID NO: 37), enhancer 4XmscRE16 (eAi11.1) (SEQ ID NO: 38), enhancer eHGT_078h (eAi107.0) (SEQ ID NO: 39), enhancer eHGT_078h core (SEQ ID NO: 177), Enhancer eHGT_078h (3xCore) (eAi129.0) (SEQ ID NO: 40), enhancer eHGT_058h (eAi106.0) (SEQ ID NO: 41), enhancer eHGT_058m (eAi108.0) (SEQ ID NO: 42), enhancer eHGT_073h (eAi111.0) (SEQ ID NO: 43), enhancer eHGT_073m (eAi1 12.0) (SEQ ID NO: 44), enhancer eHGT_075h (eAi113.0) (SEQ ID NO: 45), enhancer HGT_077h (eAi114.0) (SEQ ID NO: 46), enhancer eHGT_254h (eAi127.0) (SEQ ID NO: 47), enhancer eHGT_078m (eAi128.0) (SEQ ID NO: 48), enhancer eHGT_078m Core (SEQ ID NO: 178), enhancer eHGT_078m (3xCore) (eAi130.0) (SEQ ID NO: 49), enhancer eHGT_439m (eAi131.0) (SEQ ID NO: 50), enhancer eHGT_440h (eAi132.0) (SEQ ID NO: 51),beta globin minimal promoter (SEQ ID NO: 52), minCMV (SEQ ID NO: 53), mutated minCMV promoter (SEQ ID NO: 54), Hsp68 minimal promoter (SEQ ID NO: 55), SYFP2 (SEQ ID NO: 56), EGFP (SEQ ID NO: 57), optimized Flp recombinase (SEQ ID NO: 58), improved Cre recombinase (SEQ ID NO: 59), WPRE3 (SEQ ID NO: 60), BGHpA (SEQ ID NO: 61), HA tag (SEQ ID NO: 62), HA tag (SEQ ID NO: 63), P2A (SEQ ID NO: 64), T2A (SEQ ID NO: 65), E2A (SEQ ID NO: 66), F2A (SEQ ID NO: 67), tet-transactivator (SEQ ID NO: 68), PHP.eB capsid (SEQ ID NO: 69), AAV9 VP1 capsid (SEQ ID NO: 70), plasmid backbone 1 (SEQ ID NO: 71), plasmid backbone 2 (SEQ ID NO: 72), T502-050 (vAi33.0) (SEQ ID NO: 73), T502-054 (vAi33.1) (SEQ ID NO: 179), T502-057 (vAi3.0) (SEQ ID NO: 74), T502-059 (vAi2.0) (SEQ ID NO: 75), vAi1.0 (SEQ ID NO: 76), vAi3 3.2 (TG1114) (SEQ ID NO: 77), vAi34.0 (TG1108) (SEQ ID NO: 78), vAi45.0 (TG1109) (SEQ ID NO: 79), TG975 (vAi4.0) (SEQ ID NO: 180), TG978 (vAi4.1) (SEQ ID NO: 80), TG979 (vAi4.2) (SEQ ID NO: 181), TG981 (vAi5.0) (SEQ ID NO: 81), TG982 (vAi6.0) ( SEQ ID NO: 182), TG987 (vAi7.0) (SEQ ID NO: 183), TG988 (vAi7.1) (SEQ ID NO: 82), TG995 (vAi15.0) (SEQ ID NO: 83), TG996 (vAi19.0) (SEQ ID NO: 84), TG997 (vAi20.0) (SEQ ID NO: 184), TG999 (vAi21.0) (SEQ ID NO: 85), TG1002 (vAi26.0) (SEQ ID NO: 86), T G1009 (vAi8.0dgCre) (SEQ ID NO: 185), TG1010 (vAi6.1) (SEQ ID NO: 87), TG1011 (vAi7.2) (SEQ ID NO: 88), TG1021 (vAi8.0Cre) (SEQ ID NO: 89), TG1022 (vAi9.0) (SEQ ID NO: 186), TG1036 (vAi16.0) (SEQ ID NO: 90), TG1037 (vAi22.0) (SEQ ID NO: 91),TG1038 (vAi27.0) (SEQ ID NO: 92), TG1045 (vAi17.0) (SEQ ID NO: 187), TG1046 (vAi23.0) (SEQ ID NO: 93), TG1047 (vAi28.0) (SEQ ID NO: 94), TG1048 (vAi18.0) (SEQ ID NO: 95), TG1049 (vAi24.0) (SEQ ID NO: 96), TG1050 (vAi29.0) (SEQ ID NO: 97), TG1052 (vAi10.0) (SEQ ID NO: 98), CN1402 (vAi106.0) (SEQ ID NO: 99), CN1416 (vAi108.0) (SEQ ID NO: 100), CN142 7 (vAi130.0) (SEQ ID NO: 101), CN1452 (vAi111.0) (SEQ ID NO: 102), CN1454 (vAi113.0) (SEQ ID NO: 103), CN1456 (vAi114.0) (SEQ ID NO: 104), CN1457 (vAi107.0) (SEQ ID NO: 105), CN1461 (vAi112.0) (SEQ ID NO: 106), CN1466 (vAi131.0) (SEQ ID NO: 107), CN1772 (vAi127.0) (SEQ ID NO: 108), CN1818 (vAi128.0) (SEQ ID NO: 109), CN1954 (vAi132.0) (SEQ ID NO: 110), CN1955 (vAi133.0) (SEQ ID NO: 111), CN2014 (vAi129.0) (SEQ ID NO: 112), CN2137 (vAi135.0) (SEQ ID NO: 113), CN2139 (vAi134.0) (SEQ ID NO: 114), myosin light chain kinase, green fluorescent protein, calmodulin chimera (SEQ ID NO: 115), genetically encoded green calcium indicator NTnC (SEQ ID NO: 116), calcium indicator TN-XXL (SEQ ID NO: 117), BRET type autoluminescent calcium indicator (SEQ ID NO: 118), Calcium indicator protein OeNL(Ca2+)-18u (SEQ ID NO: 119), GCaMP6m (SEQ ID NO: 120), GCaMP6s (SEQ ID NO: 121), GCaMP6f (SEQ ID NO: 122), channelrhodopsin-1 (SEQ ID NOs: 123 and 124), channelrhodopsin-2 (SEQ ID NOs: 125 and 126), CRISPR-associated protein (Cas) (SEQ ID NO: 127), Cas9 (SEQ ID NO: 128), CRISPR-associated endonuclease Cpf1 (SEQ ID NO: 129), ribonuclease 4 or ribonuclease L (SEQ ID NO: 130),Deoxyribonuclease II beta (SEQ ID NO: 131), sodium channel protein type 1 subunit alpha (SEQ ID NO: 132), voltage-gated potassium channel subfamily KQT member 2 (SEQ ID NO: 133), voltage-gated L-type calcium channel subunit alpha-1C (SEQ ID NO: 134), lactase (SEQ ID NO: 135), lipase (SEQ ID NO: 136), helicase (SEQ ID NO: 137), amylase (SEQ ID NO: 138), alpha glucosidase (SEQ ID NO: 139), transcription factor SP1 (SEQ ID NO: 140), Transcription factor AP-1 (SEQ ID NO: 141), heat shock protein 1 (SEQ ID NO: 142), CCAAT / enhancer binding protein (C / EBP) beta isoform a (SEQ ID NO: 143), octamer binding protein 1 (Oct-1) (SEQ ID NO: 144), transforming growth factor beta 1 (SEQ ID NO: 145), platelet-derived growth factor receptor (SEQ ID NO: 146), epidermal growth factor receptor (SEQ ID NO: 147), vascular endothelial growth factor (SEQ ID NO: 148), interleukin-8 receptor alpha (SEQ ID NO: 149), caveolin ( SEQ ID NO: 150), dynamin (SEQ ID NO: 151), clathrin heavy chain 1 isoform 1 (SEQ ID NO: 152), clathrin heavy chain 2 isoform 1 (SEQ ID NO: 153), clathrin light chain A isoform a (SEQ ID NO: 154), clathrin light chain B isoform a (SEQ ID NO: 155), Ras-related protein Rab-4A isoform 1 (SEQ ID NO: 156), Ras-related protein Rab-11A (SEQ ID NO: 157), platelet-derived growth factor (SEQ ID NO: 158), transforming growth factor beta 3 (SEQ ID NO: 159), neuroprotective agent Growth factor (SEQ ID NO:160), epidermal growth factor (EGF) (SEQ ID NO:161), GTPase HRas (SEQ ID NO:162), cocaine- and amphetamine-regulated transcript (chain A) (SEQ ID NO:163), protachykinin-1 (SEQ ID NO:164), substance P is positions 58-68 of protachykinin-1 (SEQ ID NO:165), oxytocin-neurophysin-1 (SEQ ID NO:166), oxytocin is positions 20-28 of oxytocin-neurophysin-1 (SEQ ID NO:167), and somatostatin (SEQ ID NO:168). The nucleic acid sequences described herein are intended to be representative of the sequences ofThe sequences are shown using standard letter abbreviations for the nucleotide bases. Only one strand of each nucleic acid sequence is shown, but the complementary strand is understood to be included in embodiments where appropriate. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0025] To fully understand the biology of the brain, it is necessary to distinguish and define different cell types, and to study them further, it is necessary to identify vectors that can selectively label and interfere with them. Tasic, Curr.Opin.Neurobiol.50,242-249(2018); Zeng&Sanes, Nat.Rev.Neurosci.18,530-546(2017). In mice, recombinase driver strains have been used to great effect to label cell populations that share marker gene expression. Daigle et al.,Cell 174,465-480.e22(2018); Taniguchi et al.,Neuron 71,995-1013(2011); Gong et al.,J.Neurosci.27,9817-9823(2007). However, the generation, maintenance, and use of such lines to label cell types with high specificity can be expensive and often requires triple transgenic crosses that result in low frequencies of experimental animals. Moreover, these tools are not applicable to humans because they require germline transgenic animals.

[0026] Recent advances in single-cell profiling such as single-cell RNA-seq (Tasic et al., Nature 563, 72-78 (2018); Tasic 2016, Nat Neurosci 19, 335-346), and investigations of neuronal electrophysiology and morphology (Gouwens 2019, Nat Neurosci 22, 1182-1195) have revealed that many recombinant driver lines label a heterogeneous mixture of cell types, often including cells from multiple subclasses. For example, the Rbp4-Cre mouse driver line, commonly used to label layer 5 (L5) neurons, also labels cells with highly distinct connectivity patterns: L5 intratelencephalic (IT, also referred to as cortico-cortical) and pyramidal tract (PT, also referred to as cortico-subcortical) neurons.

[0027] The present disclosure provides artificial expression constructs that selectively drive gene expression in targeted central nervous system cell populations, including: L2 / 3 IT cortical excitatory neurons, L4 IT cortical excitatory neurons, L5 PT cortical excitatory neurons, L5 PT and L5 ET cortical excitatory neurons, L5 PT and L5 IT cortical excitatory neurons, L6 IT cortical excitatory neurons, L6CT cortical excitatory neurons, L2 / 3 and 5 cortical excitatory neurons, L2 / 3 IT, L4 IT, L5 IT, L5 NP, L5 PT, and CR cortical excitatory neurons, pan-excitation and / or widespread expression in cortical excitatory neurons, L5 PT cortical excitatory neurons in combination with subcortical populations in the CEAc, substantia nigra, pars compacta (or pars compacta, SNc), and (ProS), as well as cells within the subiculum, CA1 pyramidal neurons, a subset of gingival granule cells, scattered stripper neurons, and cerebellar purge cells.

[0028] An artificial expression construct comprising the promoter, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_439m, and / or eHGT_254h enhancer, and a gene encoding an expression product can result in selective gene expression in L4 IT cortical excitatory neurons.

[0029] Specific examples of artificial expression constructs that include a promoter, a mscRE4 enhancer, a linked mscRE4, and / or a linked mscRE16 enhancer, and a gene encoding an expression product can result in selective gene expression in L5 PT cortical excitatory neurons. Examples of these expression constructs include T502-057 (vAi3.0), 981 (vAi5.0), 1052 (vAi10.0), CN1818 (vAi128.0), CN2014 (vAi129.0), and vAi130.0.

[0030] An artificial expression construct containing a promoter, a linked core of the mscRE4 enhancer, and a gene encoding an expression product can result in selective gene expression in L5 PT and L5 ET cortical excitatory neurons.

[0031] Artificial expression constructs comprising a promoter, mscRE1, mscRE11, and / or mscRE16 enhancer, and a gene encoding an expression product can result in selective gene expression in L5 PT and L5 IT cortical excitatory neurons.

[0032] An artificial expression construct comprising a promoter, the mscRE13 enhancer, and a gene encoding an expression product can result in selective gene expression in L6 IT cortical excitatory neurons.

[0033] Specific examples of artificial expression constructs that include a promoter, the mscRE10 enhancer, and a gene encoding an expression product can result in selective gene expression in L6 CT cortical excitatory neurons. Examples include 995 (vAi15.0).

[0034] An artificial expression construct comprising a promoter, the eHGT_440h enhancer, and a gene encoding an expression product can result in selective gene expression in a subtype of L6b cortical excitatory neurons.

[0035] An artificial expression construct comprising a promoter, the eHGT_078h enhancer, and a gene encoding an expression product can result in selective gene expression in L2 / 3 IT, L4 IT, L5 IT, L5 NP, and L5 PT cortical excitatory neurons.

[0036] Selective expression of a gene encoding an expression product can be achieved in L2 / 3 IT, L5 IT, and L6b neurons using the 1036(vAi16.0) artificial expression construct described herein, which contains the mscRE10 enhancer.

[0037] Selective expression of genes encoding expression products can be achieved in L2 / 3 IT, L5 PT, L6CT, and L6b neurons using the 988 (vAi7.1), 1010 (vAi6.1), and / or 1011 (vAi7.2) artificial expression constructs described herein. These constructs contain the mscRE4 enhancer.

[0038] Pan-excitatory and / or widespread expression in cortical excitatory neurons can be selectively achieved using an artificial expression construct comprising a promoter, eHGT_073h, eHGT_073m, eHGT_077h, and / or eHGT_078m enhancer, and a gene encoding an expression product. In certain embodiments, pan-excitatory expression refers to expression in at least four types of cortical excitatory cells that is limited to non-expression in inhibitory cells and glial cells.

[0039] The artificial expression constructs described herein can additionally label other distinct cell types. For example, in addition to L5 PT cells, an artificial expression construct comprising a promoter, an mscRE4 enhancer, and a gene encoding an expression product can result in gene expression in subcortical populations in CEAc, substantia nigra, pars compacta (or pars compacta, SNc), and (ProS). Similarly, in addition to L5 PT cells, an artificial expression construct comprising a promoter, a linking core of the mscRE4 enhancer, and a gene encoding an expression product can result in gene expression in the subiculum, CA1 pyramidal neurons, a subset of dentate gyrus granule cells, scattered striated neurons, and rare cerebellar Purkinje cells.

[0040] As shown by the proceeding discussion, certain artificial expression constructs disclosed herein contain engineered enhancers, such as the concatenated cores of mscRE4, eHGT_078h, and eHGT_078m enhancers, as well as a concatemer of mscRE4 and mscRE16 enhancers. In relation to MscRE4, the functional 155 base pair (bp) core of the mscRE4 enhancer (SEQ ID NO: 29) was concatenated (SEQ ID NO: 30) to minimize the size required to drive gene expression. Despite being a 3x concatemer, SEQ ID NO: 30 is shorter in length than the original mscRE4 enhancer (SEQ ID NO: 28, which contains 555 bp). When used to construct artificial expression constructs such as rAAV, such concatemers allow more space for enhancer-bound cargo genes, which is highly desirable, for example, in gene therapy vectors. For example, many therapeutic cargo genes are too large to fit into AAV vector designs, so space (length of sequence) is at a premium.

[0041] As described in more detail throughout this disclosure, certain artificial expression constructs disclosed herein include T502-050, T502-054, vAi34.0, vAi33.2, vAi45.0, vAi1.0, T502-057, T502-059, TG978, TG981, TG988, TG995, TG996, TG997, TG999, TG1002, TG1010, TG1011, TG1021, Examples include TG1036, TG1037, TG1038, TG1046, TG1047, TG1048, TG1049, TG1050, TG1052, CN1402, CN1457, CN1818, CN1416, CN1452, CN1461, CN1454, CN1456, CN1772, CN1427, CN1466, CN1954, CN1955, CN2137, CN2139, and CN2014.

[0042] Aspects of the disclosure are described in additional options and details below: (i) artificial expression constructs and vectors for selective expression of genes in selected cell types, (ii) compositions for administration, (iii) cell lines comprising the artificial expression constructs, (iv) transgenic animals, (v) methods of use, (vi) kits and commercial packages, (vii) exemplary embodiments, (viii) experimental examples, and (ix) concluding paragraphs.

[0043] (i) Artificial expression constructs and vectors for selective expression of genes in selected cell types. The artificial expression constructs disclosed herein include (i) an enhancer sequence that provides selective expression of a coding sequence in a targeted central nervous system cell type, (ii) the coding sequence to be expressed, and (iii) a promoter. The expression constructs can also include other regulatory elements if necessary or beneficial.

[0044] In certain embodiments, an "enhancer" or "enhancer element" is a cis-acting sequence that increases the level of transcription associated with a promoter and can function in either orientation relative to the promoter and coding sequence to be transcribed, and can be located upstream or downstream relative to the promoter or coding sequence to be transcribed. Art-recognized methods and techniques exist for measuring the function(s) of an enhancer element sequence. Specific examples of enhancer sequences utilized within the artificial expression constructs disclosed herein include mscRE1, mscRE3, mscRE4, concatemer of mscRE4 core, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, concatemer of mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073h, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, concatemer of eHGT_078h core, eHGT_078m, eHGT_078m core, eHGT_439m, eHGT_440h, and eHGT_254h.

[0045] In certain embodiments, the targeted central nervous system cell type enhancer is an enhancer that is uniquely or primarily utilized by the targeted central nervous system cell type. The targeted central nervous system cell type enhancer enhances expression of a gene in the targeted central nervous system cell type but does not substantially induce expression of a gene in other non-targeted cell types, and thus has neuro-specific transcriptional activity.

[0046] If the coding sequence is preferentially expressed in the selected neuronal cell and is not substantially expressed in other neuronal cell types, the product of the coding sequence is preferentially expressed in the selected neuronal cell type. In certain embodiments, the preferential expression is greater than 50% compared to the reference cell type, greater than 60% compared to the reference cell type, greater than 70% compared to the reference cell type, greater than 80% compared to the reference cell type, or greater than 90% compared to the reference cell type. In certain embodiments, the reference cell type refers to a non-targeted neuronal cell. The non-targeted neuronal cell may be in the same anatomical structure as the targeted cell and / or may project to a common anatomical region. In certain embodiments, the reference cell type is in an anatomical structure adjacent to the anatomical structure containing the targeted cell type. In certain embodiments, the reference cell type is a non-targeted neuronal cell that has a gene expression profile different from that of the targeted cell.

[0047] In certain embodiments, the product of the coding sequence can be expressed at low levels in non-selected cell types, for example, less than 1%, or 1%, 2%, 3%, 5%, 10%, 15% or 20% of the level at which the product is expressed in selected neural cells.In certain embodiments, the targeted central nervous system cell type is the only cell type that expresses the appropriate combination of transcription factors that bind to the enhancer disclosed herein to drive gene expression.Thus, in certain embodiments, expression occurs only in the targeted cell type.

[0048] In certain embodiments, target cell types (e.g., neural, neuronal, and / or non-neuronal) can be identified based on transcriptional profiles, such as those described in Tasic et al., 2018 Nature, and Hodge et al., Nature, 573, 61-68 (2019). Human cell types are further defined in an ontology framework defined at Bioontology.org. For reference, the following descriptions of neural cell types and distinguishing features are also provided:

[0049] Cortical Glutamatergic Neuron Classes. Glutamatergic neurons (also called excitatory neurons) produce the neurotransmitter glutamate, which is excitatory (promotes firing) when received by neurons with ionotropic receptors, and modulatory when received by neurons expressing metabotropic receptors. Most cortical glutamatergic neurons project outside of their resident zone (defined as the location of the primary cell body including the nucleus), and genetic markers correlate with these projection properties.

[0050] Cortical glutamatergic neuron subclasses. Subclasses of glutamatergic neurons are defined by both the layer in which the neuronal cell body (including the nucleus) resides and the primary projection pattern of these neurons. In mice, glutamatergic neurons are found in layers (L) 1, L2 / 3, L4, L5, L6, and the cortical subplate (also called L6b). In humans, glutamatergic neurons are found in L2, L3, L4, L5, L6, and L6b. In mice, L2 / 3 is often considered a single layer, while human cortical layers 2 and 3 are distinct. Intratelencephalic (IT, also called cortico-cortical) neurons project primarily from cortical cell bodies to other adjacent or distant cortical regions. Corticothalamic (CT) neurons project primarily from the cortex to the thalamus. Pyramidal tracts (PTs, also called cortical descending or extracortical neurons) project primarily from the cortex, usually from layer 5 of the cortex, to a variety of subcortical targets. Nearly projection (NP) neurons appear to have only localized projections within their cortical resident area.

[0051] In mice, projections and layer categories intersect in specific patterns that define glutamatergic neuron subclasses: L2 / 3 IT, L4 IT, L5 IT, L6 IT for IT neurons, L6CT for CT neurons, L5 PT for PT neurons, and L5 / 6 NP for NP neurons (some regions found in both layers). Projections of the L6b subclass of cells have not yet been clearly defined, but projections from L6b to local targets have been observed, as well as cortico-cortical projections to the anterior cingulate and subcortical projections to the thalamus. In mice, there are also very distinct types of neurons that arise alone. CR-Lhx5 cells correspond to Cajal-Retzius cells based on their location in L1 and expression of known Cajal-Retzius (CR) markers such as Trp73, Lhx5, and Reln.

[0052] In the human cortex, long-range cortical and subcortical projections are difficult to ascertain directly. However, a similar pattern of cell types is observed based on layer location and molecular correspondence to the projection classes seen in the mouse. Layer 4 cells tend to receive input from other cortical structures through the expression of specific genes such as RORB, by the lack of projection neurons, and through granular cytoarchitectures that are usually visualized by nuclear markers such as DAPI.

[0053] Overview of cortical glutamatergic subclasses: All: express the glutamate transmitters Slc17a6 and / or Slc17a7. They all express Snap25, lack expression of Gad1 / Gad2, and lack expression of Slc1A3. • L2 / 3 IT: Predominantly present in layer 2 / 3, with primarily intratelencephalic (cortico-cortical) projections. • L4 IT: Present mainly in layer 4, with projections that are either primarily regional or intratelencephalic (cortico-cortical). • L5 IT: Predominantly present in layer 5, with mainly intratelencephalic (cortico-cortical) projections. Also called L5a. L5 PT: Predominantly in layer 5, with predominantly cortico-subcortical (pyramidal or subcortical) projections. Also called L5b or L5 CF (corticofugal), or L5 ET (extratelencephalic). This subclass contains cells that are corticospinal projection neurons located in the primary motor cortex and adjacent regions and associated with motor neuron / movement disorders such as ALS. This subclass includes thick tufted pyramidal neurons, including distinctive subtypes found only in specialized areas, e.g., Betz, Meynert, and von Economo cells. • L5 NP: resident mainly in layer 5, with protrusions mainly nearby. • L6 CT: Present mainly in layer 6, with mainly corticothalamic projections. • L6 IT: Predominantly present in layer 6, with predominantly intracranial (cortico-cortical) processes. This subclass includes L6 IT Car3 cells, which are highly similar to the intracranial cortical processing cells. • L6b: Present mainly in the cortical subplate (L6b), with local (near the cell body) processes and some cortico-cortical processes from VISp to the anterior cyst, and cortico-subcortical processes to the thalamus. • Cajal-Retzius cells, a distinct subclass defined by a single type in CR:L1, express distinct molecular markers Lhx5 and Trp73.

[0054] Within each subclass, differentially expressed genes define multiple distinct experimentally targetable cell types. For example, within L2 / 3 IT cells in the primary visual cortex, three distinct cell types have been observed: L2 / 3 IT VISp Rrad, L2 / 3 IT VISp Adamts2, and L2 / 3 IT VISp Agmat, identified by expression of the Rrad, Adamts2, and Agmat genes, respectively. These genetic markers are primarily used to distinguish each cell type from related cell types within the cell subclass (in this case, L2 / 3 IT), and may not represent a single gene that distinguishes a cell type from all other cells within the cortex. Marker genes may need to be applied in combination to uniquely identify a given cell type.

[0055] Cortical GABAergic neuron classes. GABAergic neurons (also called inhibitory neurons) produce the neurotransmitter gamma-aminobutyric acid (GABA), which inhibits downstream neuronal firing. All cortical GABAergic neurons except one (called Meis2-Adamts19) share many gene expression markers, including Thy1 and Scn2b. The Meis2-Adamts19 type corresponds to a Meis2-expressing GABAergic neuron type that is mostly restricted to the white matter derived from the embryonic palmar border. Of the GABAergic types, it is the only type that robustly expresses the transcription factor Meis2 mRNA, transcribes a minimal number of genes, and does not express Thy1 and Scn2b.

[0056] Overview of cortical GABAergic subclasses: ●All: express the GABA synthesis genes Gad1 / GAD1 and Gad2 / GAD2. • Lamp5, Sncg, Serpinf1, and Vip: developmentally derived from neuronal precursors from the caudal ganglion luminal (CGE) or preoptic area (POA). • Sst and Pvalb: developmentally derived from neuronal precursors in the medial ganglionic emittance (MGE). ●Lamp5: Found in many cortical layers, especially the upper ones (L1-L2 / 3), and has mainly glial and single plexus morphology. • Sncg: Found in many cortical layers and has molecular overlap with Lamp5 and Vip cells, but expression of Lamp5 or Vip is inconsistent and expression of Sncg is more consistent. • Serpinf1: Found in many cortical layers and has molecular overlap with Sncg and Vip cells, but expression of Sncg or Vip is inconsistent and expression of Serpinf1 is more consistent. ●Vip: Found in many cortical layers, but especially in the upper layers (L1-L4), they highly express the neurotransmitter vasoactive intestinal peptide (Vip). ●Sst: Found in many cortical layers, but especially frequent in the lower layers (L5-L6). They highly express the neurotransmitter somatostatin (Sst) and frequently block dendritic input to postsynaptic neurons. This subclass includes sleep-active Sst Chodl neurons (which also express Nos1 and Tacr1), which are very distinct from other Sst neurons but express several shared marker genes with Sst. In humans, SST gene expression is often detected within layer 1 LAMP5+ cells. ●Pvalb: Found in many cortical layers, but especially frequent in the lower layers (L5-L6). They highly express the calcium-binding protein parvalbumin (Pvalb) and express the neuropeptide Tac1, which often attenuates the output of postsynaptic neurons. Most fast-spiking inhibitory cells strongly express Pvalb. This subclass includes chandelier cells, which have a distinct chandelier-like morphology and express the markers Cpne5 and Vipr2 in mice, and NOG and UNC5B in humans. Meis2: A distinct subclass defined only by the cortical GABAergic phenotype expressing the Meis2 gene and not expressing several other genes expressed by all other cortical GABAergic phenotypes (e.g., Thy1 and Scn2b). This type is found in L6b and subcortical white matter.

[0057] Cells located in the central nucleus of the amygdala (including CEA, CEAc) are involved in pain, anxiety, and fear processing. Cells in the substantia nigra compacta (SNc, also called pars compacta) are located in the midbrain, involved in motor control, and are adversely affected in Parkinson's disease. Cells in the prosbiculum (ProS) are located between the hippocampal CA1 region and the subiculum region.

[0058] The subiculum is the lowest component of the hippocampal formation. It lies between the entorhinal cortex and the CA1 subfield of the hippocampus proper. CA1 pyramidal neurons send their axons to the subiculum and deep layers of the entorhinal cortex. Granule cells in the dentate gyrus receive excitatory neuronal input from the entorhinal cortex and send excitatory output via mossy fibers to the hippocampal CA3 region. The cell bodies of striatal neurons are located within the subcortical basal ganglia of the forebrain. Purkinje cells send inhibitory projections to the deep cerebellar nuclei and constitute the predominant, if not the only, output of all motor coordination within the cerebellar cortex.

[0059] Non-neuronal subclasses: Astrocytes: Neuroectoderm-derived glial cells that express the marker Aqp4 and often express GFAP, but not the neuronal marker SNAP25. They can have a distinct star-shaped morphology and are involved in the metabolic support of other cells in the brain. Multiple astrocytic morphologies are observed in mice and humans. Oligodendrocytes: Glial cells derived from the neuroectoderm that express the marker Sox10. This category includes oligodendrocyte precursor cells (OPCs). Oligodendrocytes are the subclass primarily responsible for myelination of neurons. • VLMC: Vascular leptomeningeal cells (VLMC) are part of the meninges that surround the outer layer of the cortex and express the marker genes Lum and Col1a1. Pericytes: Blood vessel-associated cells, also known as mural cells, that express the marker genes Kcnj8 and Abcc9. Pericytes wrap around endothelial cells and are important in regulating capillary blood flow and are involved in blood-brain barrier permeability. SMCs: Specialized smooth muscle cells, also called mural cells, that are blood vessel-associated cells that express the marker gene Acta2. SMCs line the arteries of the brain and are involved in blood-brain barrier permeability. ●Endothelium: Cells that line the blood vessels in the brain. Endothelial cells express the markers Tek and PDGF-B. Microglia: Hematopoietic-derived immune cells that are brain-resident macrophages and perivascular macrophages (PVMs) that may be transiently associated with brain tissue or included as a by-product of brain dissection methods. Microglia are known to express Cx3cr1, Tmem119, and PTPRC (CD45).

[0060] In certain embodiments, the coding sequence is a heterologous coding sequence that encodes an effector element. An effector element is a sequence that is expressed to achieve and actually achieve an intended effect. Examples of effector elements include reporter genes / proteins and functional genes / proteins.

[0061] Exemplary reporter genes / proteins include those expressed by Addgene ID numbers 83894 (pAAV-hDlx-Flex-DTomato-Fishell_7), 83895 (pAAV-hDlx-Flex-GFP-Fishell_6), 83896 (pAAV-hDlx-GiDREADD-dTomato-Fishell-5), 83898 (pAAV-mDlx-ChR2-mCherry-Fishell-3), 83899 (pAAV-mDlx-GCaMP6f-Fishell-2), 83900 (pAAV-mDlx-GFP-Fishell-1), and 89897 (pcDNA3-FLAG-mTET2(N500)).Exemplary reporter genes include those encoding expressible fluorescent proteins or expressible biotin, blue fluorescent proteins (e.g., eBFP, eBFP2, Azurite, mKalama1, GFPuv, Sapphire, T-sapphire), cyan fluorescent proteins (e.g., eCFP, Cerulean, CyPet, Amcyanl, Midoriishi-Cyan, mTurquoise), green fluorescent proteins (e.g., GFP, GFP-2, tagGFP, turboGFP, EGFP, Emerald, Azami Green, Monomeric Azami Green (mAzamigreen), CopGFP, AceGFP, avGFP, ZsGreenl, Oregon Green™ (Thermo Fisher Scientific)), luciferase, orange fluorescent proteins (mOrange, mKO, Kusabira-Orange, Monomeric Examples of fluorescent proteins include, but are not limited to, Kusabira-Orange, mTangerine, tdTomato, and dTomato), red fluorescent proteins (e.g., mKate, mKate2, mPlum, DsRed monomer, mCherry, mRuby, mRFP1, DsRed-Express, DsRed2, DsRed-Monomer, HcRed-Tandem, HcRedl, AsRed2, eqFP611, mRaspberry, mStrawberry, Jred, and Texas Red™ (Thermo Fisher Scientific)), far-red fluorescent proteins (e.g., mPlum and mNeptune), yellow fluorescent proteins (e.g., YFP, eYFP, Citrine, SYFP2, Venus, YPet, PhiYFP, and ZsYellowl), and tandem conjugates.

[0062] GFP consists of 238 amino acids (26.9 kDa) and was originally isolated from Aequorea victoria / Aequorea aequorea / Aequorea forskalea, a jellyfish that fluoresces green when exposed to blue light. GFP from A. victoria has a major excitation peak at a wavelength of 395 nm and a minor excitation peak at 475 nm. Its emission peak is at 509 nm, which is in the lower green part of the visible spectrum. GFP from Renilla reniformis has a single major excitation peak at 498 nm. Due to its potential for widespread use and the evolving needs of researchers, many different variants of GFP have been engineered. The first major improvement was a single point mutation (S65T) reported in Nature in 1995 by Roger Tsien. This mutation dramatically improved the spectral properties of GFP, resulting in increased fluorescence, photostability, and a shift of the major excitation peak to 488 nm, with the peak emission maintained at 509 nm. By adding the 37°C folding efficiency (F64L) point mutant to this scaffold, enhanced GFP (EGFP) was obtained. EGFP has a folding efficiency of 9.13×10−21 m, also referred to as 55,000 L / (mol cm). 2 / has an extinction coefficient (denoted by ε), also known as the optical cross section of the molecule. Superfolder GFP, a series of mutations that allow GFP to rapidly fold and mature even when fused to poorly folding peptides, was reported in 2006.

[0063] "Yellow fluorescent protein" (YFP) is a genetic variant of the green fluorescent protein derived from the jellyfish Aequorea victoria. Its excitation peak is 514 nm and its emission peak is 527 nm.

[0064] Exemplary functional molecules include functional ion transporters, cellular transport proteins, enzymes, transcription factors, neurotransmitters, calcium reporters, channelrhodopsins, guide RNAs, nucleases, or designer receptors exclusively activated by designer drugs (DREADDs).

[0065] Ion transporters are transmembrane proteins that mediate the transport of ions across cell membranes. These ion transporters are widespread throughout most cell types and are important for regulating cellular excitability and homeostasis. Ion transporters participate in numerous cellular processes, such as action potentials, synaptic transmission, hormone secretion, and muscle contraction. Many important biological processes in living cells involve the movement of cations, such as calcium (Ca2+), potassium (K+), and sodium (Na+) ions, through such ion channels. In certain embodiments, ion transporters include voltage-gated sodium channels (e.g., SCN1A), potassium channels (e.g., KCNQ2), and calcium channels (e.g., CACNA1C).

[0066] Exemplary enzymes, transcription factors, receptors, membrane proteins, cellular trafficking proteins, signaling molecules, and neurotransmitters include enzymes such as lactase, lipase, helicase, α-glucosidase, amylase; transcription factors such as SP1, AP-1, heat shock factor protein 1, C / EBP (CCAA-T / enhancer binding protein), and Oct-1; receptors such as transforming growth factor receptor β1, platelet-derived growth factor receptor, epidermal growth factor receptor, vascular endothelial growth factor receptor, and interleukin-8 receptor α; cellular trafficking proteins such as membrane proteins, clathrin, dynamin, caveolin, Rab-4A, and Rab-11A; signaling molecules such as nerve growth factor (NGF), platelet-derived growth factor (PDGF), transforming growth factor β (TGFβ), epidermal growth factor (EGF), GTPases, and HRas; and neurotransmitters such as cocaine- and amphetamine-regulated transcription factors, substance P, oxytocin, and somatostatin.

[0067] In certain embodiments, the functional molecules include status such as reporters of neuronal function and calcium reporters. Intracellular calcium concentration is an important predictor of numerous cellular activities, including neuronal activation, muscle cell contraction, and second messenger signaling. A sensitive and convenient technique for monitoring intracellular calcium levels is via genetically encoded calcium indicators (GECIs). Among GECIs, a green fluorescent protein (GFP)-based calcium sensor named GCaMP is an efficient and widely used tool. GCaMP is formed by the fusion of M13 and calmodulin proteins to the N-terminus and C-terminus of a circular mutant GFP. Some GCaMPs produce distinct fluorescence emission spectra (Zhao et al., Science, 2011, 333(6051):1888-1891). Exemplary GECIs with green fluorescence include GCaMP3, GCaMP5G, GCaMP6s, GCaMP6m, GCaMP6f, jGCaMP7s, jGCaMP7c, jGCaMP7b, and jGCaMP7f. Additionally, GECIs with red fluorescence include jRGECO1a and jRGECO1b. AAV products containing GECIs are commercially available.For example, Vigene Biosciences offers AAV8-CAG-GCaMP3 (catalog number: BS4-CX3AAV8), AAV8-Syn-FLEX-GCaMP6s-WPRE (catalog number: BS1-NXSAAV8), AAV8-Syn-FLEX-GCaMP6s-WPRE (catalog number: BS1-NXSAAV8), AAV9-CAG-FLEX-GCaMP6m-WPRE (catalog number: BS2-CXMAAV9), AAV9-Syn-FLEX-jGCaMP7s-WPRE (catalog number: BS12-NXSAAV9), AAV9-CAG-FLEX -We provide AAV products including jGCaMP7f-WPRE (Cat. No.: BS12-CXFAAV9), AAV9-Syn-FLEX-jGCaMP7b-WPRE (Cat. No.: BS12-NXBAAV9), AAV9-Syn-FLEX-jGCaMP7c-WPRE (Cat. No.: BS12-NXCAAV9), AAV9-Syn-FLEX-NES-jRGECO1a-WPRE (Cat. No.: BS8-NXAAAV9), and AAV8-Syn-FLEX-NES-jRCaMP1b-WPRE (Cat. No.: BS7-NXBAAV8).

[0068] In certain embodiments, calcium reporters include genetically encoded calcium indicators GECI, NTnC, myosin light chain kinase, GFP, calmodulin chimera, calcium indicator TN-XXL, BRET-type autoluminescent calcium indicators, and / or calcium indicator protein OeNL(Ca2+)-18u).

[0069] In certain embodiments, the functional molecules include regulators of neuronal activity, such as channelrhodopsins (e.g., channelrhodopsin-1, channelrhodopsin-2, and variants thereof). Channelrhodopsins are a subfamily of retinylidene proteins (rhodopsins) that function as light-gated ion channels. In addition to channelrhodopsin 1 (ChR1) and channelrhodopsin 2 (ChR2), several channelrhodopsin variants have been developed. For example, Lin et al. (Biophys J, 2009, 96(5):1803-14) describe the creation of chimeras of the transmembrane domains of ChR1 and ChR2 in combination with site-directed mutagenesis. Zhang et al. (Nat Neurosci, 2008, 11(6):631-3) describe a red-shifted channelrhodopsin variant, VChR1. VChR1 has low light sensitivity and poor membrane trafficking and expression. Other known channelrhodopsin mutants include the ChR2 mutants described in Nagel et al., Proc Natl Acad Sci USA, 2003, 100(24):13940-5), ChR2 / H134R (Nagel, G. et al., Curr Biol, 2005, 15(24):2279-84), and ChD / ChEF / ChIEF (Lin, JY et al. Biophys J, 2009, 96(5):1803-14), which are activated by blue light (470 nm) but show no sensitivity to orange / red light. Additional mutants are described in Lin, Experimental Physiology, 2010, 96.1:19-25, and Knopfel et al., The Journal of Neuroscience, 2010, 30(45):14998-15004.

[0070] In certain embodiments, functional molecules include DNA and RNA editing tools such as CRISPR / CAS (e.g., guide RNA and nucleases such as Cas, Cas9 or cpf1). Functional molecules can also include engineered Cpf1 such as those described in US2018 / 0030425, US2016 / 0208243, WO / 2017 / 184768, and Zetsche et al. (2015) Cell 163:759-771, single gRNAs (see, e.g., Jinek et al. (2012) Science 337:816-821, Jinek et al. (2013) eLife 2:e00471, Segal (2013) eLife 2:e00563), or editases, guide RNA molecules, or homologous recombination donor cassettes.

[0071] Additional effector elements include Cre, iCre, dgCre, FlpO, and tTA2. iCre refers to codon-improved Cre. dgCre refers to an enhanced GFP / Cre recombinase fusion gene with an N-terminal fusion of the first 159 amino acids of the E. coli K-12 strain chromosomal dihydrofolate reductase gene (DHFR or folA) with a G67S mutation, modified to also contain an R12Y / Y100I instability domain mutation. FlpO refers to a codon-optimized form of FLPe, which greatly increases protein expression and FRT recombination efficiency in mouse cells. Similar to the Cre / LoxP system, the FLP / FRT system is widely used for gene expression (and the generation of conditional knockout mice mediated by the FLP / FRT system). tTA2 refers to tetracycline transactivator.

[0072] Exemplary expressible elements are expression products that do not contain effector elements, such as, for example, non-functional or defective proteins. In certain embodiments, expressible elements can provide a method for studying the effects of their functional counterparts. In certain embodiments, expressible elements are non-functional or defective based on engineered mutations that render them non-functional. In these aspects, non-expressible elements are as similar in structure as possible to their functional counterparts.

[0073] Exemplary self-cleaving peptides include 2A peptides, which lead to the production of two proteins from one mRNA. The 2A sequence is short (e.g., 20 amino acids), allowing more use in size-restricted constructs. Specific examples include P2A, T2A, E2A, and F2A. In certain embodiments, the expression construct includes an internal ribosome entry site (IRES) sequence. The IRES allows the ribosome to begin translation at a second internal site on the mRNA molecule, leading to the production of two proteins from one mRNA.

[0074] Coding sequences encoding the molecules (e.g., RNA, proteins) described herein can be obtained from publicly available databases and publications. The coding sequences can further include various sequence polymorphisms, mutations, and / or sequence variants, and such modifications do not affect the function of the encoded molecule. The terms "encode" or "encoding" refer to the property of a sequence of a nucleic acid, such as a vector, plasmid, gene, cDNA, mRNA, etc., that serves as a template for the synthesis of another molecule, such as a protein.

[0075] The term "gene" may include not only coding sequences, but also regulatory regions such as promoters, enhancers, and termination regions. The term may further include all introns and other DNA sequences spliced ​​from the mRNA transcript, as well as variants resulting from alternative splice sites. A sequence may also include degenerate codons of a reference sequence, or sequences that may be introduced to provide codon preferences in a particular organism or cell type.

[0076] The promoter may include a general promoter, a tissue-specific promoter, a cell-specific promoter, and / or a cytoplasm-specific promoter. The promoter may include a strong promoter, a weak promoter, a constitutive expression promoter, and / or an inducible promoter. An inducible promoter induces expression in response to a particular condition, signal, or cellular event. For example, the promoter may be an inducible promoter that requires a particular ligand, small molecule, transcription factor, or hormone protein to effect transcription from the promoter. Specific examples of promoters include minBglobin, CMV, minCMV, mutant minCMV, SV40, immediate early promoter, Hsp68 minimal promoter (proHSP68), and Rous sarcoma virus (RSV) long repeat (LTR) promoter. A minimal promoter does not have the activity to drive gene expression alone, but can be activated to drive gene expression when linked to a proximal enhancer element.

[0077] In certain embodiments, the expression construct is provided in a vector. The term vector refers to a nucleic acid molecule that can transfer or transport another nucleic acid molecule, such as an expression construct. The transferred nucleic acid is generally linked, e.g., inserted, into a vector nucleic acid molecule. The vector may include sequences that direct autonomous replication in a cell, or may include sequences that allow integration into host cell DNA. Useful vectors include, for example, plasmids (e.g., DNA or RNA plasmids), transposons, cosmids, bacterial artificial chromosomes, and viral vectors.

[0078] Viral vector is used broadly to refer to nucleic acid molecules that contain nucleic acid elements derived from viruses that facilitate the transfer and expression of non-natural nucleic acid molecules into cells. The term adeno-associated viral vector refers to a viral vector or plasmid that contains structural and functional genetic elements, or parts thereof, that are derived primarily from AAV. The term "retroviral vector" refers to a viral vector or plasmid that contains structural and functional genetic elements, or parts thereof, that are derived primarily from retroviruses. The term "lentiviral vector" refers to a viral vector or plasmid that contains structural and functional genetic elements, or parts thereof, that are derived primarily from lentiviruses, etc. The term "hybrid vector" refers to a vector that contains structural and / or functional genetic elements from more than one type of virus.

[0079] Adenovirus. "Adenovirus vector" refers to a construct that contains sufficient adenovirus sequences to (a) support packaging of an expression construct and (b) express coding sequences cloned therein in the sense or antisense orientation. Recombinant adenovirus vectors include genetically engineered forms of adenovirus. Knowledge of the genetic organization of adenovirus, a 36 kb linear double-stranded DNA virus, allows replacement of large pieces of adenoviral DNA with up to 7 kb of foreign sequences. In contrast to retroviruses, adenoviral infection of host cells does not result in chromosomal integration, because adenoviral DNA can replicate in an episomal manner without potential genotoxicity. Adenoviruses are also structurally stable, and no genome rearrangements have been detected after extensive amplification.

[0080] Adenoviruses are particularly suitable for use as gene transfer vectors due to their medium-sized genome, ease of manipulation, high titer, wide target cell range, and high infectivity. Both ends of the viral genome contain 100-200 base pair inverted repeats (ITRs), which are cis elements necessary for viral DNA replication and packaging. The early (E) and late (L) regions of the genome contain different transcription units that are divided by the initiation of viral DNA replication. The E1 region (E1A and E1B) encodes proteins involved in the regulation of transcription of the viral genome and several cellular genes. Expression of the E2 region (E2A and E2B) results in the synthesis of proteins for viral DNA replication. These proteins are involved in DNA replication, late gene expression, and host cell shutoff. The products of the late genes, including the majority of the viral capsid proteins, are expressed only after significant processing of a single primary transcript issued by the major late promoter (MLP). The MLP is particularly efficient during the late stages of infection, and all mRNAs issued from this promoter contain a 5'-tripartite leader (TPL) sequence making them the preferred mRNAs for translation.

[0081] Other than the requirement that the adenovirus vector be replication-defective, or at least conditionally defective, the nature of the adenovirus vector is not believed to be critical to the successful practice of certain embodiments disclosed herein. The adenovirus can be any of the 42 different known serotypes or subgroups A-F. In certain embodiments, adenovirus type 5 of subgroup C is the preferred starting material for obtaining a conditionally replication-defective adenovirus vector for use in certain embodiments. Adenovirus type 5 is a human adenovirus for which much biochemical and genetic information is known, and has been used historically in most constructs using adenovirus as a vector.

[0082] As shown, typical vectors are replication-defective and will not have the adenovirus E1 region.Therefore, it will be most convenient to introduce the polynucleotide encoding the gene of interest into the position where the E1 coding sequence has been removed.However, the insertion position of the construct within the adenovirus sequence is not critical.The polynucleotide encoding the gene of interest can also be inserted in place of the deleted E3 region in an E3 replacement vector or in the E4 region where a helper cell line or helper virus complements the E4 defect.

[0083] Adeno-associated virus (AAV) is a parvovirus discovered as a contamination of adenovirus strains. It is a ubiquitous virus that is not associated with any disease (antibodies are present in 85% of the human population in the United States). It is also classified as a dependent virus because its replication depends on the presence of a helper virus such as adenovirus. Various serotypes have been isolated, of which AAV-2 is the best characterized. AAV has single-stranded linear DNA that is encapsulated in the capsid proteins VP1, VP2, and VP3, forming icosahedral virions with a diameter of 20 to 24 nm.

[0084] AAV DNA is 4.7 kilobases long. It contains two open reading frames flanked by two ITRs. There are two major genes in the AAV genome: rep and cap. The rep gene codes for proteins involved in viral replication, while cap codes for the capsid proteins VP1-3. Each ITR forms a T-shaped hairpin structure. These terminal repeats are the only essential cis components of AAV for chromosomal integration. Thus, AAV can be used as a vector with all viral coding sequences removed and replaced with a cassette of genes for delivery. Three AAV viral promoters have been identified and named p5, p19, and p40 according to their map locations. Transcription from p5 and p19 leads to the production of rep proteins, while transcription from p40 produces capsid proteins.

[0085] AAV is prominent for use within the present disclosure due to its excellent safety profile and the fact that their capsids and genomes can be tailored to allow expression in selected cell populations. scAAV refers to self-complementary AAV. pAAV refers to plasmid adeno-associated virus. rAAV refers to recombinant adeno-associated virus.

[0086] Other viral vectors may also be employed, such as vectors derived from viruses such as vaccinia virus, poliovirus, and herpes virus, which offer several attractive features for a variety of mammalian cells.

[0087] Retroviruses. Retroviruses are common tools for gene delivery. "Retrovirus" refers to an RNA virus that reverse transcribes its genomic RNA into a linear double-stranded DNA copy and then covalently integrates its genomic DNA into the host genome. Once the virus is integrated into the host genome, it is called a "provirus." The provirus serves as a template for RNA polymerase II to induce the expression of RNA molecules that code for the structural proteins and enzymes required to produce new viral particles.

[0088] Exemplary retroviruses suitable for use in certain embodiments include Moloney murine leukemia virus (M-MuLV), Moloney murine sarcoma virus (MoMSV), Harvey murine sarcoma virus (HaMuSV), mouse mammary tumor virus (MuMTV), Gibbon monkey leukemia virus (GaLV), feline leukemia virus (FLV), spumavirus, Friend murine leukemia virus, murine stem cell virus (MSCV), and Ras sarcoma virus (RSV), and lentiviruses.

[0089] "Lentivirus" refers to a group (or genus) of complex retroviruses. Exemplary lentiviruses include HIV (including human immunodeficiency virus, HIV type 1 and HIV type 2), visnamesi virus (VMV), capric arthritis encephalitis virus (CAEV), equine infectious anemia virus (EIAV), feline immunodeficiency virus (FIV), bovine immunodeficiency virus (BIV), and simian immunodeficiency virus (SIV). In certain embodiments, an HIV-based vector backbone (i.e., HIV cis-acting sequence elements) may be used.

[0090] Improved safety for using some vectors can be provided by replacing the U3 region of the 5'LTR with a heterologous promoter to drive transcription of the viral genome during production of viral particles. Examples of heterologous promoters that can be used for this purpose include, for example, the viral Samia virus 40 (SV40) (e.g., early or late), cytomegalovirus (CMV) (e.g., immediate early), Moloney murine leukemia virus (MoMLV), Rous sarcoma virus (RSV), and herpes simplex virus (HSV) (thymidine kinase) promoters. Typical promoters can drive high levels of transcription in a Tat-independent manner. This replacement reduces the possibility of recombination to generate replication-competent virus, since the complete U3 sequence is not present in the virus production system. In certain embodiments, heterologous promoters have the added advantage of controlling how the viral genome is transcribed. For example, the heterologous promoter can be inducible, such that transcription of all or part of the viral genome occurs only when an inducer is present. Inducers include one or more chemical compounds or physiological conditions, such as temperature or pH, under which the host cells are cultured.

[0091] In certain embodiments, the viral vector comprises a TAR element. The term "TAR" refers to the "transactivation response" genetic element located in the R region of lentiviral LTR. This element interacts with the lentiviral transactivator (tat) genetic element to enhance viral replication. However, this element is not necessary in the embodiment in which the U3 region of 5'LTR is replaced by a heterologous promoter.

[0092] "R region" refers to the region within the retroviral LTR that begins at the beginning of the capping group (i.e., at the beginning of transcription) and ends just before the beginning of the poly(A) tract. The R region is also defined as flanking the U3 and U5 regions. The R region plays a role during reverse transcription in allowing the transfer of nascent DNA from one end of the genome to the other.

[0093] In certain embodiments, expression of heterologous sequences in viral vectors is increased by incorporating post-transcriptional regulatory elements, efficient polyadenylation sites, and optionally transcription termination signals into the vector. Various post-transcriptional regulatory elements can increase expression of heterologous nucleic acids. Examples include the woodchuck hepatitis virus post-transcriptional regulatory element (WPRE, Zufferey et al., 1999, J. Virol., 73:2886), the post-transcriptional regulatory element present in Hepatitis B virus (HPRE) (Smith et al., Nucleic Acids Res. 26(21):4818-4827, 1998), and the like (Liu et al., 1995, Genes Dev., 9:1766). In certain embodiments, the vector comprises a post-transcriptional regulatory element such as a WPRE or HPRE. In certain embodiments, the vector lacks or does not comprise a post-transcriptional regulatory element such as a WPRE or HPRE.

[0094] Elements that direct efficient termination and polyadenylation of heterologous nucleic acid transcripts can increase heterologous gene expression. Transcription termination signals are generally found downstream of polyadenylation signals. In certain embodiments, vectors include a polyadenylation sequence 3' of a polynucleotide encoding an expressed molecule (e.g., a protein). The term "poly(A) site" or "poly(A) sequence" refers to a DNA sequence that directs both the termination and polyadenylation of a nascent RNA transcript by RNA polymerase II. Polyadenylation sequences can promote mRNA stability by adding a poly(A) tail to the 3' end of the coding sequence, thus contributing to improved translation efficiency. Certain embodiments may utilize BGHpA or SV40pA. In certain embodiments, preferred embodiments of expression constructs include terminator elements. These elements can help to enhance transcription levels and minimize lead from the construct to other plasmid sequences.

[0095] In certain embodiments, the viral vector further comprises one or more insulator elements. The insulator elements may contribute to protecting viral vector expression sequences, such as effector elements or expressible elements, from integration position effects (i.e., position effects, see, for example, Burgess-Beusse et al., PNAS., USA, 99:16433, 2002, and Zhan et al., Hum. Genet., 109:471, 2001), which may be mediated by cis-acting elements present in genomic DNA and may result in deregulated expression of the imported sequence. In certain embodiments, the viral import vector comprises one or more insulator elements in the 3'LTR, and when the provirus is integrated into the host genome, the provirus comprises one or more insulators in both the 5'LTR and the 3'LTR by replicating the 3'LTR. Suitable insulators for use in certain embodiments include the avian β-globin insulator (Chung et al. Cell 74:505, 1993; Chung et al., PNAS USA 94:575, 1997; and Bell et al. Cell 98:387, 1999), the SP10 insulator (Abhyankar et al., JBC 282:36143, 2007), or other small CTCF recognition sequences that function as enhancer-blocking insulators (Liu et al. Nature, Biotechnology, 33:198, 2015).

[0096] Beyond the foregoing description, a wide range of suitable expression vector types will be known to those skilled in the art. These can include commercially available expression vectors designed for general recombinant procedures, such as plasmids containing one or more reporter genes and regulatory elements necessary for the expression of the reporter gene in cells. Numerous vectors are commercially available, for example from Invitrogen, Stratagene, Clontech, etc., and are described in many relevant guides. In certain embodiments, suitable expression vectors include any plasmid, cosmid, or phage construct capable of supporting the expression of a coding gene in mammalian cells, such as the pUC or Bluescript plasmid series. [Table 1] TIFF0007590962000002.tif254170TIFF0007590962000003.tif254170TIFF0007590962000004.tif30170

[0097] In certain embodiments, a vector (e.g., AAV) is selected that has a capsid that crosses the blood-brain barrier (BBB). In certain embodiments, the vector is modified to contain a capsid that crosses the BBB. Examples of AAVs with viral capsids that cross the blood-brain barrier include AAV9 (Gombash et al. Front Mol Neurosci. 2014; 7: 81), AAVrh.10 (Yang et al., Mol Ther. 2014; 22(7): 1299-1309), AAV1R6, AAV1R7 (Albright et al., Mol Ther. 2018; 26(2): 510), rAAVrh.8 (Yang et al., supra), AAV-BR1 (Marchio et al., EMBO Mol Med. 2016; 8(6): 592), AAV-PHP.S (Chan et al., Nat Neurosci. 2017; 20(8): 1172), and AAV-PHP.B (Deverman et al., Nat Biotechnol. 2016;34(2):204), AAV-PPS (Chen et al. Nat Med. 2009;15:1215), and PHP.eB capsid. The PHP.eB capsid differs from AAV9, such that, using AAV9 as a reference, the amino acids starting at residue 586 are changed from S-AQ-A (SEQ ID NO:169) to S-DGTLAVPFK-A (SEQ ID NO:170).

[0098] AAV9 is a naturally occurring AAV serotype that, unlike many other naturally occurring serotypes, can cross the BBB after intravenous injection. It transforms large parts of the central nervous system (CNS), thus allowing for minimally invasive treatment, as described, for example, in connection with clinical trials for the treatment of spinal muscular atrophy (SMA) syndrome with AveXis (AVXS-101, NCT03505099) and for the treatment of CLN3 gene-associated neuronal cerebrovascular disease (NCT03770572) (Naso et al., BioDrugs. 2017;31(4):317).

[0099] AAVrh.10 was originally isolated from rhesus macaques, exhibits low seropositivity in humans when compared to other common serotypes used in gene delivery applications (see Selot et al., Front Pharmacol. 2017;8:441), and is being evaluated in the LYS-SAF302, LYSOGENE, and NCT03612869 clinical trials.

[0100] Two variants isolated from a library of chimeric AAV vectors, AAV1R6 and AAV1R7 (in which the AAV1 capsid domain was swapped for AAVrh.10), retain the ability to cross the BBB and transduce the CNS while showing significantly reduced liver and vascular endothelial transduction.

[0101] rAAVrh.8, also isolated from rhesus macaques, demonstrates global transduction of glial and neuronal cell types in clinically important areas following peripheral administration and also displays reduced peripheral tissue tropism compared to other vectors.

[0102] AAV-BR1 is an AAV2 variant that displays the NRGTEWD ​​(SEQ ID NO: 171) epitope, isolated during in vivo screening of a random AAV display peptide library. It shows high specificity with high transgene expression in the brain with minimal off-target affinity (including the liver) (Korbelin et al., EMBO Mol Med. 2016;8(6):609).

[0103] AAV-PHP.S (Addgene, Watertown, Mass.) is a CREATE-generated variant of AAV9 that encodes the 7-mer sequence QAVRTSL (sequence number 172) and potently transduces neurons within the enteric nervous system and peripheral sensory afferents entering the spinal cord and brainstem.

[0104] AAV-PHP.B (Addgene, Watertown, Mass.) is a variant of AAV9 generated by the CREATE method that encodes the 7-mer sequence TLAVPFK (SEQ ID NO: 173). It transfers genes throughout the CNS more efficiently than AAV9 and transduces the majority of astrocytes and neurons across multiple CNS regions.

[0105] The AAV2 variant AAV-PPS, created by inserting the DSPAHPS (SEQ ID NO: 174) epitope into the capsid of AAV2, exhibits dramatically improved brain tropism compared to AAV2.

[0106] For additional information regarding capsids crossing the blood-brain barrier, see Chan et al., Nat. Neurosci, 2017 Aug:20(8):1172-1179.

[0107] (ii) Compositions for Administration. The artificial expression constructs and vectors of the present disclosure (referred to herein as physiologically active components) can be formulated with a carrier suitable for administration to cells, tissue slices, animals (e.g., mice, non-human primates), or humans. The physiologically active components in the compositions described herein can be prepared in neutral form, as a free base, or as a pharmacologically acceptable salt.

[0108] Pharmaceutically acceptable salts include the acid addition salts (formed with the free amino groups of the protein) and are formed with inorganic acids such as, for example, hydrochloric or phosphoric acids, or such organic acids as, for example, acetic, oxalic, tartaric, mandelic, etc. Salts formed with the free carboxyl groups are also derived from inorganic bases such as, for example, sodium, potassium, ammonium, calcium, or ferric hydroxides, as well as such organic bases as, for example, isopropylamine, trimethylamine, histidine, procaine, and the like.

[0109] Carriers for physiologically active ingredients can include solvents, dispersion media, vehicles, coatings, diluents, isotonic and absorption delaying agents, buffers, solutions, suspensions, colloids, etc. The use of such carriers for physiologically active ingredients is well known in the art. Any conventional media or agent can be used with the compositions described herein, unless it is incompatible with the physiologically active ingredient.

[0110] The phrase "pharmacologically acceptable carrier" refers to a carrier that does not produce an allergic or similar untoward reaction when administered to a human, and in certain embodiments, when administered intravenously (e.g., at the retroorbital plexus).

[0111] In certain embodiments, the compositions can be formulated for intravenous, intraparenchymal, intraocular, intravitreal, parenteral, subcutaneous, intraventricular, intramuscular, intrathecal, intraspinal, intraperitoneal, oral or nasal inhalation, or for direct injection or application to one or more cells, tissues, or organs.

[0112] The composition may comprise liposomes, lipids, lipid complexes, microparticles, microparticles, nanospheres, and / or nanoparticles.

[0113] The formation and use of liposomes are generally known to those skilled in the art. Liposomes have been developed with improved serum stability and circulation half-life (see, for example, U.S. Patent No. 5,741,516). In addition, various methods of liposomes and liposome-like preparations as potential drug carriers have been described (see, for example, U.S. Patent Nos. 5,567,434, 5,552,157, 5,565,213, 5,738,868, and 5,795,587).

[0114] The present disclosure also provides pharma- ceutically acceptable nanocapsule formulations of physiologically active ingredients. Nanocapsules can generally encapsulate compounds in a stable and reproducible manner (Quintanar-Guerrero et al. Drug Dev Ind Pharm 24(12):1113-1128, 1998; Quintanar-Guerrero et al. Pharm Res. 15(7):1056-1062, 1998; Quintanar-Guerrero et al., J. Microencapsul. 15(1):107-119, 1998; Douglas et al. Crit, Rev Ther Drug Carrier Syst 3(3):233-261, 1987). To avoid side effects due to intracellular polymer overload, such ultrafine particles can be designed using polymers that can be degraded in vivo. Biodegradable polyalkylcyanoacrylate nanoparticles that meet these requirements are contemplated for use in the present disclosure. Such particles are described in Couvreuret al., J Pharm Sci 69(2):199-202, 1980; Couvreuret al., Crit Rev Ther Drug Carrier Syst. 5(1)1-20, 1988; zur Muhlenet al., Eur J Pharm Biopharm, 45(2):149-155, 1998; Zambaux et al., J Control Release 50(1-3):31-40, 1998; and U.S. Pat. No. 5,145,684.

[0115] Injectable solutions may include sterile aqueous solutions, or dispersions and sterile powders for the extemporaneous preparation of sterile injectable solutions or dispersions (U.S. Pat. No. 5,466,468). For delivery by injection, the form is sterile and fluid to the extent that it can be delivered by syringe. In certain embodiments, the composition is stable under the conditions of manufacture and storage, and optionally contains one or more preservative compounds to protect against the contaminating action of microorganisms, such as bacteria and fungi. The carrier may be a solvent or dispersion medium containing water, ethanol, polyol (for example, glycerol, propylene glycol, and liquid polyethylene glycol, etc.), suitable mixtures thereof, and / or vegetable oils. The proper fluidity can be maintained, for example, by the use of a coating such as lecithin, by the maintenance of the required particle size in the case of dispersions, and / or by the use of surfactants. Prevention of microbial activity can be brought about by various antibacterial and / or antifungal agents, for example, parabens, chlorobutanol, phenol, sorbic acid, thimerosal, and the like. In various embodiments, the preparation includes isotonic agent(s), such as sugar(s) or sodium chloride. Prolonged absorption of the injectable composition can be achieved by including in the composition an agent that delays absorption, such as aluminum monostearate and gelatin. The injectable composition can be suitably buffered if necessary, and the liquid diluent is first rendered isotonic with sufficient saline or glucose.

[0116] Dispersions may also be prepared in glycerol, liquid polyethylene glycols, and mixtures thereof, as well as in oils. As indicated, under ordinary conditions of storage and use, these preparations may contain a preservative to prevent the growth of microorganisms.

[0117] Sterile compositions can be prepared by incorporating the physiologically active ingredient in an appropriate amount of a solvent containing other optional ingredients (e.g., as listed above), followed by filtration sterilization. Generally, dispersions are prepared by incorporating various sterilized physiologically active ingredients into a sterile vehicle containing a basic dispersion medium and other required ingredients (e.g., from those listed above). In the case of sterile powders for preparing sterile injectable solutions, the preferred preparation method can be vacuum drying and freeze-drying technology, which obtains a powder of the physiologically active ingredient and any additional desired ingredients from the solution previously sterile-filtered.

[0118] Oral compositions may be in liquid form, for example, as solution, syrup or suspension, or may be presented as a pharmaceutical product for reconstitution with water or other suitable vehicle before use.Such liquid preparations can be prepared by conventional means using pharma-ceutically acceptable additives such as suspending agents (e.g., sorbitol syrup, cellulose derivatives, or hydrogenated edible fats), emulsifying agents (e.g., lecithin or acacia), non-aqueous vehicles (e.g., almond oil, oily esters, or fractionated vegetable oils), and preservatives (e.g., methyl or propyl-p-hydroxybenzoate or sorbic acid). The compositions may take the form of, for example, tablets or capsules prepared with pharma- ceutically acceptable excipients such as binders (e.g., pregelatinized corn starch, polyvinylpyrrolidone, or hydroxypropylmethylcellulose), fillers (e.g., lactose, microcrystalline cellulose, or calcium hydrogen phosphate), lubricants (e.g., magnesium stearate, talc, or silica), disintegrants (e.g., potato starch or sodium starch glycolate), or wetting agents (e.g., sodium lauryl sulfate). Tablets may be coated by methods well known in the art.

[0119] Inhalable compositions can be delivered in the form of an aerosol spray presentation from a pressurized pack or nebulizer using a suitable propellant, for example, dichlorodifluoromethane, trichlorofluoromethane, dichlorotetrafluoroethane, carbon dioxide or other suitable gas. In the case of a pressurized aerosol, the dosage unit can be determined by providing a valve that delivers a metered amount. Capsules and cartridges, for example of gelatin, for use in an inhaler or filler can be formulated to contain a powder mix of the compound and a suitable powder base, such as lactose or starch.

[0120] Compositions also include microchip devices (U.S. Pat. No. 5,797,898), ophthalmic formulations (Bourlais et al., Prog Retin Eye Res, 17(1):33-58, 1998), transdermal matrices (U.S. Pat. Nos. 5,770,219 and 5,783,208), and feedback controlled delivery (U.S. Pat. No. 5,697,899).

[0121] Supplementary active ingredients can also be incorporated into the compositions.

[0122] Typically, the composition will contain at least 0.1% of the physiologically active ingredient, although the percentage of the physiologically active ingredient may of course vary, and may conveniently be 1 or 2% to 70% or 80% or more, or 0.5 to 99% by weight or volume of the total composition. Of course, the amount of the physiologically active ingredient in each physiologically useful composition may be prepared in such a way that a suitable dosage is obtained in any given unit dose of the compound. Factors such as solubility, bioavailability, biological half-life, route of administration, shelf life of the product, as well as other pharmacological considerations, are contemplated by those skilled in the art of preparing such pharmaceutical formulations, and therefore various compositions and dosages may be desirable.

[0123] In certain embodiments, for human administration, compositions should meet sterility, pyrogenicity, and general safety and purity standards as required by the U.S. Food and Drug Administration (FDA) or other applicable regulatory agencies in other countries.

[0124] (iii) Cell lines containing the artificial expression constructs. The present disclosure includes cells containing the artificial expression constructs described herein. Cells transformed with artificial expression constructs can be used for many purposes, including neuroanatomical studies, evaluation of functional and / or non-functional proteins, and drug screening to evaluate the regulatory properties of enhancers.

[0125] Although various host cell lines can be used, in certain embodiments, the cell is a mammalian neuronal cell. In certain embodiments, the enhancer sequence of the artificial expression construct is selected from the group consisting of mscRE1, mscRE3, mscRE4, a concatemer of mscRE4 core, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, a concatemer of mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, a concatemer of eHGT_078h, eHGT_078m, eHGT_078m Concatamers of core, eHGT_439m, eHGT_440h, and eHGT_254h, and / or artificial expression constructs are T502-050, T502-054, vAi34.0, vAi33.2, vAi45.0, vAi1,0, T502-057, T502-059, TG975, TG978, TG979, TG981, TG982, TG987, TG988, TG995, TG996, TG997, TG999, TG1002, TG1009, TG1010, TG1011, TG10 21, TG1022, TG1036, TG1037, TG1038, TG1045, TG1046, TG1047, TG1048, TG1049, TG1050, TG1052, CN1402, CN1457, CN1818, CN1416, CN1452, CN1461, CN1454, CN1456, CN1772, CN1427, CN1466, CN1954, CN1955, CN2137, CN2139, and / or CN2014, and the cell line is a human, primate, or murine neuronal cell. Cell lines that can be used for introduction in the present disclosure also include primary cell lines derived from living tissues such as rat or mouse brain, as well as organic cell cultures including brain slices from animals such as rats or mice. The PC12 cell line (available from the American Type Culture Collection, ATCC, Manassas, VA) has been shown to express several neuronal marker proteins in response to nerve growth factor (NGF). The PC12 cell line is considered to be a neuronal cell line and is applicable for use with the present disclosure.JAR cells (available from ATCC) are a platelet-derived cell line that expresses several neuronal genes, including the serotonin transporter gene, and may be used with the embodiments described herein.

[0126] WO91 / 13150 describes various cell lines, including neuronal cell lines, and methods for their production. Similarly, WO97 / 39117 describes neuronal cell lines and methods for the production of such cell lines. The neuronal cell lines disclosed in these patent applications are adapted for use in the present disclosure.

[0127] In certain embodiments, "neuronal cell" refers to a cell or cells located within the central nervous system, including neurons and glia, and cells derived from neurons and glia, including tumor cells and tumor cells derived from neurons or glia. "Cells derived from neural cells" refers to cells that are derived from, have originated from, or are differentiated from neural cells.

[0128] In certain embodiments, "neuron" describes something that is, relates to, or includes a neuronal cell. A nerve cell is defined by the presence of an axon and a dendrite. The term "neuron-specific" refers to activity that is found or occurs in a neuronal cell or a cell derived from a neuronal cell, but is not found or occurs or is substantially not found or occurs in a non-neuronal cell or a cell not derived from a neuronal cell, for example, a glial cell such as an astrocyte or an oligodendrocyte.

[0129] In certain embodiments, non-neuronal cell lines including mouse embryonic stem cells may be used. Cultured mouse embryonic stem cells can be used to analyze the expression of gene constructs using transient transfection with plasmid constructs. Mouse embryonic stem cells are pluripotent and undifferentiated. These cells can be maintained in this undifferentiated state by leukemia inhibitory factor (LIF). Withdrawal of LIF induces differentiation of embryonic stem cells. In culture, the stem cells form a variety of differentiated cell types. Differentiation is triggered by the expression of tissue-specific transcription factors, and the function of enhancer sequences can be evaluated. (See, for example, Fiskerstrand et al., FEBS Lett 458:171-174, 1999.)

[0130] The method of differentiating stem cells into neuronal cells includes replacing the stem cell culture medium with a medium containing basic fibroblast growth factor (bFGF), heparin, N2 supplement (e.g., transferrin, insulin, progesterone, putrescine, and selenate), laminin, and polyornithine. The process of producing myelinating oligodendrocytes from stem cells is described in Hu et al., 2009, Nat. Protoc. 4:1614-22. Bibel et al., 2007, Nat. Protoc. 2:1034-43, describes a protocol for producing glutamatergic neurons from stem cells, while Chatzi et al., 2009, Exp. Neurol. 217:407-16, describes a procedure for producing GABAergic neurons. The procedure includes exposing stem cells to all-trans RA for 3 days. After subsequent culture in serum-free neuronal induction medium containing Neurobasal medium supplemented with B27, bFGF and EGF, 95% of GABA neurons develop.

[0131] US Publication No. 2012 / 0329714 describes the use of prolactin to increase neural stem cell numbers, and US Publication No. 2012 / 0308530 describes culture surfaces with amino groups that promote neuronal differentiation into neurons, astrocytes, and oligodendrocytes. Thus, the fate of neural stem cells can be controlled by a variety of extracellular factors. Commonly used factors include brain-derived growth factor (BDNF, Shetty and Turner, 1998, J. Neurobiol. 35:395-425), fibroblast growth factor (bFGF, U.S. Patent No. 5,766,948, FGF-1, FGF-2), neurotrophin-3 (NT-3) and neurotrophin-4 (NT-4), Caldwell et al. al., 2001, Nat. Biotechnol. 1, 19:475-9), ciliary neurotrophic factor (CNTF), BMP-2 (U.S. Pat. Nos. 5,948,428 and 6,001,654), isobutyl 3-methylxanthine, leukemia inhibitory growth factor (LIF, U.S. Pat. No. 6,103,530), somatostatin, amphiregululin, neurotrophins (e.g., cyclic adenosine monophosphate, epidermal growth factor (EGF), dexamethasone (a glucocorticoid hormone), forskolin, GDNF family ligands, potassium, retinoic acid (U.S. Pat. No. 6,395,546), and transforming factor alpha and TGF-beta (U.S. Pat. Nos. 5,851,832 and 5,753,506).

[0132] In certain embodiments, the yeast one-hybrid system may also be used to identify compounds that inhibit specific protein / DNA interactions, such as transcription factors of the mscRE1, mscRE3, mscRE4, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073h, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, concatemers of eHGT_078h core, eHGT_078m, concatemers of eHGT_078m core, eHGT_439m, eHGT_440h, and / or eHGT_254h enhancers.

[0133] Transgenic animals are described below. Cell lines can also be derived from such transgenic animals. For example, primary tissue cultures from transgenic mice (e.g., as described below) can provide cell lines that have expression constructs already integrated into the genome (see, e.g., MacKenzie and Quinn, Proc Natl Acad Sci USA 96:15251-15255, 1999).

[0134] (iv) Transgenic animals. Another aspect of the disclosure includes a transgenic animal, the genome or cells of which contain an artificial expression construct comprising mscRE1, mscRE3, mscRE4, a concatemer of mscRE4 core, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, a concatemer of mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073h, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, a concatemer of eHGT_078h core, eHGT_078m, a concatemer of eHGT_078m core, eHGT_439m, eHGT_440h, and / or eHGT_254h operably linked to a heterologous coding sequence. In certain embodiments, the genome or cells of the transgenic animal are selected from the group consisting of T502-050, T502-054, vAi34.0, vAi33.2, vAi45.0, vAi1.0, T502-057, T502-059, TG975, TG978, TG979, TG981, TG982, TG987, TG988, TG995, TG996, TG997, TG999, TG1002, TG1009, TG1010, TG1011, TG1021, and / or artificial expression constructs comprising TG1022, TG1036, TG1037, TG1038, TG1045, TG1046, TG1047, TG1048, TG1049, TG1050, TG1052, CN1402, CN1457, CN1818, CN1416, CN1452, CN1461, CN1454, CN1456, CN1772, CN1427, CN1466, CN1954, CN2137, CN2139, and / or CN2014.In certain embodiments, when a non-integrating vector is used, the transgenic animal contains in one or more of its cells mscRE1, mscRE3, mscRE4, a concatemer of mscRE4, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, a concatemer of mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073h, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, a concatemer of eHGT_078h core, eHGT_078m, eHGT_078m Concatamers of core, eHGT_439m, eHGT_440h, eHGT_254h and / or T502-050, T502-054, vAi34.0, vAi33.2, vAi45.0, vAi1.0, T502-057, T502-059, TG975, TG978, TG979, TG981, TG982, TG987, TG988, TG995, TG996, TG997, TG999, TG1002, TG1009, TG1010, TG1011, TG102 1, TG1022, TG1036, TG1037, TG1038, TG1045, TG1046, TG1047, TG1048, TG1049, TG1050, TG1052, CN1402, CN1457, CN1818, CN1416, CN1452, CN1461, CN1454, CN1456, CN1772, CN1427, CN1466, CN1954, CN1955, CN2137, CN2139, and / or CN2014.

[0135] Detailed methods for producing transgenic animals are described in U.S. Patent No. 4,736, 866. The transgenic animals can be any non-human species, but preferably include non-human primates (NHPs), sheep, horses, cows, pigs, goats, dogs, cats, rabbits, chickens, and rodents, such as guinea pigs, hamsters, monkeys, rats, mice, and ferrets.

[0136] In certain embodiments, the construction of transgenic animals results in organisms with engineered constructs present in all cells in the same genomic integration site.Therefore, cell lines derived from such transgenic animals will be as consistent as engineered constructs in all cells in the same genomic integration site, and therefore will undergo the same position effect mutations.In contrast, introducing genes into cell lines or primary cell cultures can result in heterologous expression of constructs.The drawback of this approach is that the expression of introduced DNA can be affected by the specific genetic background of the host animal.

[0137] As indicated above in relation to cell lines, the artificial expression constructs of the present disclosure can be used to genetically modify mouse embryonic stem cells using techniques known in the art. Typically, the artificial expression constructs are introduced into cultured mouse embryonic stem cells. The transformed ES cells are then injected into blastocysts from the host mother, and the host embryo is reimplanted into the mother. This results in chimeric mice whose tissues are composed of cells derived from both the embryonic stem cells present in the cultured cell line and the embryonic stem cells present in the host embryo. Usually, the mouse from which the cultured ES cells used for the transfer are derived is selected to have a coat color that is different from the host mouse whose embryo the transformed cells are injected into. The chimeric mice then have a variety of coat colors. The chimeric mice are bred with the appropriate strain to produce offspring that carry the transgene, as long as at least some of the germline tissue is derived from the genetically modified cells.

[0138] In addition to the delivery methods described above, the following techniques are also contemplated as alternative methods for delivering artificial expression constructs to target cells or selected tissues and organs of animals, particularly cells, organs, or tissues of vertebrates: ultrasound (e.g., ultrasound as described in U.S. Pat. No. 5,656,016), intraosseous injection (U.S. Pat. No. 5,779,708), microchip devices (U.S. Pat. No. 5,797,898), ophthalmic formulations (Bourlais et al., Prog Retin Eye Res, 17(1):33-58, 1998), transdermal matrices (U.S. Pat. Nos. 5,770,219 and 5,783,208), and feedback-controlled delivery (U.S. Pat. No. 5,697,899).

[0139] (v) Methods of Use. In certain embodiments, compositions containing physiologically active ingredients described herein are administered to a subject to produce a physiological effect.

[0140] In certain embodiments, the disclosure includes the use of the artificial expression constructs described herein to regulate the expression of a heterologous gene partially or completely encoded at a position downstream of an enhancer in an engineered sequence. Accordingly, methods of using the disclosed artificial expression constructs in the study, research, and potential development of drugs to prevent, treat, or alleviate symptoms of a disease, dysfunction, or disorder are provided herein.

[0141] Certain embodiments include methods of administering to a subject an artificial expression construct comprising SEQ ID NOs: 25-51, 177-178, and / or 188, and / or SEQ ID NOs: 73-114, and / or 179-187, as described herein, to drive selective expression of a gene in a selected neural cell type.

[0142] Certain embodiments include methods of administering to a subject, which may be an isolated cell, a cell network, a tissue slice, a laboratory animal, a veterinary animal, or a human, an artificial expression construct comprising SEQ ID NOs:25-51, 177-178, and / or 188, and / or SEQ ID NOs:73-114, and / or 179-187 described herein to drive selective expression of a gene in a selected neural cell type.

[0143] As is well known in the medical arts, the dosage of any one subject will depend on many factors, including the subject's size, surface area, age, the particular compound being administered, sex, time and route of administration, general health, and other drugs being administered concomitantly. Doses of compounds of the present disclosure will vary, but in certain embodiments, dosages are within the range of 10 to 150 mg of the artificial expression construct of the present disclosure. 5 ~10 100 In certain embodiments, patients receiving intravenous, intraparenchymal, intraspinal, retroorbital, or intrathecal administration may receive 10 6 ~10 22 A copy of the artificial expression construct can be injected.

[0144] "Effective amount" is the amount of composition required to produce a desired physiological change in a subject. Effective amounts are often administered for research purposes. The effective amounts disclosed herein may cause statistically significant effects in animal models or in vitro assays.

[0145] The amount of expression constructs and the time of administration of such compositions will be within the purview of one of ordinary skill in the art having the benefit of the present teachings. However, it is likely that administration of an effective amount of the disclosed compositions can be achieved by a single administration, such as, for example, a single injection of a sufficient number of infectious particles to produce an effect in the subject. Alternatively, in some circumstances, it may be desirable to provide multiple or successive administrations of the artificial expression construct composition or other genetic construct over either a relatively short or relatively long period of time, as may be determined by the individual overseeing the administration of such compositions. For example, the number of infectious particles administered to a mammal may be as little as 10 to 20 mg / kg, given either as a single dose or divided into two or more administrations, as may be required to achieve the intended effect. 7 , 10 8 , 10 9 , 10 10 , 10 11 , 10 12 , 10 13 , or even higher infectious particles / ml. Indeed, in certain embodiments, it may be desirable to administer two or more different expression constructs in combination to achieve a desired effect.

[0146] In certain circumstances, it may be desirable to deliver the artificial expression construct in an appropriately formulated composition as disclosed herein, either by pipette, retro-orbital injection, subcutaneously, intraocularly, intravitreally, parenterally, subcutaneously, intravenously, intraparenchymal, intracerebroventricularly, intramuscularly, intrathecally, intraspinally, orally, by oral or nasal inhalation, intraperitoneally, or by direct application or injection into one or more cells, tissues, or organs. Methods of administration may also include those forms as described in U.S. Patent No. 5,543,158, U.S. Patent No. 5,641,515, and U.S. Patent No. 5,399,363.

[0147] (vi) Kits and commercial packages. Kits and commercial packages contain the artificial expression constructs described herein. The expression constructs can be sequestered. In certain embodiments, the components of the expression product can be sequestered from one another. In certain embodiments, the expression product can be in a vector, in a viral vector, in a cell, in a tissue slice or sample, and / or in a transgenic animal. Such kits can further include one or more reagents, restriction enzymes, peptides, therapeutic agents, pharmaceutical compounds, or a means of delivery of the composition, such as a syringe, injectable, etc.

[0148] Kit or commercial package embodiments also include instructions for use of the included components, for example, in basic research, electrophysiological studies, neuroanatomical studies, and / or in the study and / or treatment of a disorder, disease or condition.

[0149] The following illustrative embodiments and experimental examples are included to demonstrate specific embodiments of the present disclosure. Those of skill in the art will appreciate in light of the present disclosure that many changes can be made to the specific embodiments disclosed herein and still obtain like or similar results without departing from the spirit and scope of the present disclosure.

[0150] (vii) Exemplary embodiments. 1. The linking core of the enhancer disclosed herein. 2. The linked core of embodiment 1, wherein the core is selected from SEQ ID NOs: 29, 177, and / or 178. 3. The linked core of embodiment 1 or 2, wherein the linked core comprises 2, 3, 4, 5, 6, 7, 8, 9, or 10 copies of SEQ ID NO: 29, 177, and / or 178. 4. The linked core of embodiment 3, comprising 3 copies of SEQ ID NO:29. 5. The linking core of embodiment 4, comprising SEQ ID NO: 30. 6. The linked core of embodiment 3, comprising 3 copies of SEQ ID NO: 177. 7. The linking core of embodiment 6, comprising SEQ ID NO: 40. 8. The linked core of embodiment 3, comprising 3 copies of sequence number 178. 9. The linking core of embodiment 8, comprising SEQ ID NO:49. 10. An artificial expression construct comprising: (i) an enhancer selected from mscRE1, mscRE3, mscRE4, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073h, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, eHGT_078m, eHGT 439m, eHGT_440h, eHGT 254h, and / or a concatemer of any of embodiments 1 to 8; (ii) a promoter; and (iii) a heterologous coding sequence. 11. The artificial expression construct of embodiment 10, wherein the heterologous coding sequence encodes an effector element or an expressible element. 12. The artificial expression construct of embodiment 11, wherein the effector element comprises a reporter protein or functional molecule. 13. The artificial expression construct of embodiment 12, wherein the reporter protein comprises a fluorescent protein. 14. The artificial expression construct of embodiment 12, wherein the functional molecule comprises a functional ion transporter, enzyme, transcription factor, receptor, membrane protein, cellular transport protein, signaling molecule, neurotransmitter, calcium reporter, channelrhodopsin, CRISPR / CAS molecule, editase, guide RNA molecule, homologous recombination donor cassette, or designer receptor that is exclusively activated by designer drugs (DREADD). 15. The artificial expression construct of embodiment 11, wherein the expressible element comprises a non-functional molecule. 16. The artificial expression construct of embodiment 15, wherein the non-functional molecule comprises a non-functional ion transporter, an enzyme, a transcription factor, a receptor, a membrane protein, a cellular transport protein, a signaling molecule, a neurotransmitter, a calcium reporter, a channelrhodopsin, a CRISPR / CAS molecule, an editase, a guide RNA molecule, a homologous recombination donor cassette, or a DREADD. 17. An artificial expression construct according to any of embodiments 10 to 16, comprising a concatemer of enhancers selected from mscRE1, mscRE3, mscRE4, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073h, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, eHGT_078m, eHGT_439m, eHGT_440h and eHGT_254h. 18. The artificial expression construct according to embodiment 17, wherein the concatemer comprises 2, 3, 4, 5, 6, 7, 8, 9, or 10 copies of the selected enhancer. 19. The artificial expression construct according to embodiment 18, wherein the concatemer comprises 3 or 4 copies of mscRE4, or 3 or 4 copies of mscRE16. 20. An artificial expression construct according to any of embodiments 10 to 19, wherein the artificial expression construct is associated with a capsid that crosses the blood-brain barrier. 21. The artificial expression construct of embodiment 20, wherein the capsid comprises PHP.eB, AAV-BR1, AAV-PHP.S, AAV-PHP.B, or AAV-PPS. 22. An artificial expression construct according to any of embodiments 10 to 21, wherein the expression construct comprises or encodes a skip element. 23. The artificial expression construct according to embodiment 22, wherein the skip element comprises a 2A peptide and / or an internal ribosome entry site (IRES). The artificial expression construct of embodiment 23, wherein the 24.2A peptide is selected from T2A, P2A, E2A, or F2A. 25. The artificial expression construct is mscRE1, mscRE3, mscRE4, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073h, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, eHGT_078m, eHGT_439m, eHGT_440h, or eHGT_254h 25. An artificial expression construct according to any of embodiments 10 to 24, comprising a set of features selected from: an enhancer selected from pBGmin or minBglobin, and / or a concatemer according to any of embodiments 1 to 9; a promoter selected from pBGmin or minBglobin; an expression product selected from EGFP, SYFP2, IRES2, FlpO, Cre, iCre, dgCre or tTA2; and a post-regulatory element selected from WPRE3 and / or BGHpA. 26. A vector comprising a linked core and / or an artificial expression construct according to any one of embodiments 1 to 25. 27.T502-050, T502-054, vAi34.0, vAi33.2, vAi45.0, vAi1.0, T502-057, T502-059, TG975, TG978, TG979, TG981, TG982, TG987, TG988, TG995, TG996, TG997, TG999, TG1002, TG1009, TG1010, TG1011, TG1021, TG1022, TG1036, TG10 37, TG1038, TG1045, TG1046, TG1047, TG1048, TG1049, TG1050, TG1052, CN1402, CN1457, CN1818, CN1416, CN1452, CN1461, CN1454, CN1456, CN1772, CN1427, CN1466, CN1954, CN1955, CN2137, CN2139, and CN2014. 28. The vector of embodiment 27, wherein the vector comprises a viral vector. 29. The vector of embodiment 28, wherein the viral vector comprises a recombinant adeno-associated viral (AAV) vector. 30. An adeno-associated virus (AAV) vector comprising at least one heterologous coding sequence, wherein the heterologous coding sequence is under the control of a promoter and an enhancer selected from mscRE1, mscRE3, mscRE4, mscRE10, mscRE11, mscRE12, mscRE13, mscRE16, Grik1_enhScnn1a-2, eHGT_058h, eHGT_058m, eHGT_073h, eHGT_073m, eHGT_075h, eHGT_077h, eHGT_078h, eHGT_078m, eHGT_439m, eHGT_440h, eHGT_254h, and / or a concatemer of any of embodiments 1-9. 31. The AAV vector of embodiment 30, wherein the AAV vector is replicable. 32. A transgenic cell comprising a ligated core, an artificial expression construct, and / or a vector according to any of the previous embodiments. 33. The transgenic cell of embodiment 32, wherein the transgenic cell is a cortical excitatory neuron. 34. The transgenic cell of embodiment 32 or 33, wherein the transgenic cell is a layer (L) 2, L3, L4, L5, or L6 cortical excitatory neuron. 35. The transgenic cell of any of embodiments 32 to 34, wherein the transgenic cell is an L4 IT cortical excitatory neuron, an L5 PT cortical excitatory neuron, an L5 ET cortical excitatory neuron, an L5 IT cortical excitatory neuron, an L5 NP cortical excitatory neuron, an L6 IT cortical excitatory neuron, an L6CT cortical excitatory neuron, or a CR cortical excitatory neuron. 36. The transgenic cell of embodiment 32, wherein the transgenic cell is derived from a subcortical population in the CEAc, substantia nigra, pars compacta, subiculum, or prosciutto (ProS). 37. The transgenic cell of embodiment 32, wherein the transgenic cell is a CA1 pyramidal neuron, a dentate granule cell, a striatal neuron, or a cerebellar Purkinje cell. 38. A non-human transgenic animal comprising the linked core enhancer, artificial expression construct, vector, and / or transgenic cell according to any of the preceding embodiments. 39. The non-human transgenic animal of embodiment 38, wherein the non-human transgenic animal is a mouse or a non-human primate. 40. An administrable composition comprising a linked core, an artificial expression construct, a vector, or a transgenic cell according to any of the preceding embodiments. 41. A kit comprising a linking core, an artificial expression construct, a vector, a transgenic cell, a transgenic animal, and / or an administrable composition according to any of the preceding embodiments. 42. A method for selectively expressing a heterologous gene in a neuronal population in vivo or in vitro, the method comprising providing to a sample or subject comprising a neuronal population an administrable composition according to embodiment 40 in a sufficient dosage and for a sufficient period of time, thereby selectively expressing the gene in the neuronal population. 43. The method of embodiment 42, wherein the heterologous gene encodes an effector element or an expressible element. 44. The method of embodiment 43, wherein the effector element comprises a reporter protein or functional molecule. 45. The method of embodiment 44, wherein the reporter protein comprises a fluorescent protein. 46. ​​The method of embodiment 44, wherein the functional molecule comprises a functional ion transporter, enzyme, transcription factor, receptor, membrane protein, cellular transport protein, signaling molecule, neurotransmitter, calcium reporter, channelrhodopsin, CRISPR / CAS molecule, editase, guide RNA molecule, homologous recombination donor cassette, or DREADD. 47. The method of embodiment 43, wherein the expressible element comprises a non-functional molecule. 48. The method of embodiment 47, wherein the non-functional molecule comprises a non-functional ion transporter, enzyme, transcription factor, receptor, membrane protein, cellular transport protein, signaling molecule, neurotransmitter, calcium reporter, channelrhodopsin, CRISPR / CAS molecule, editase, guide RNA molecule, homologous recombination donor cassette, or DREADD. 49. The method of any of embodiments 42-48, wherein providing comprises pipetting. 50. The method of embodiment 49, wherein the pipetting is to a brain slice. 51. The method of embodiment 50, wherein the brain slice comprises excitatory neurons. 52. The method of embodiment 50 or 51, wherein the brain slice comprises layer (L) 2, L3, L4, L5, and / or L6 cortical excitatory neurons. 53. The method of any of embodiments 50-52, wherein the brain slice comprises L4 IT cortical excitatory neurons, L5 PT cortical excitatory neurons, L5 ET cortical excitatory neurons, L5 IT cortical excitatory neurons, L5 NP cortical excitatory neurons, L6 IT cortical excitatory neurons, L6 CT cortical excitatory neurons, and / or CR cortical excitatory neurons. 54. The method of any of embodiments 50 to 53, wherein the brain slice comprises subcortical populations in the CEAc, substantia nigra, pars compacta, subiculum, and / or prosbiculum (ProS). 55. The method of any of embodiments 50-54, wherein the brain slice comprises CA1 pyramidal neurons, dentate gyrus granule cells, stripe neurons, and / or cerebellar Purkinje cells. 56. The method of any of embodiments 50 to 55, wherein the brain slice is mouse, human, or non-human primate. 57. The method of embodiment 48, wherein providing comprises administering to a living subject. 58. The method of embodiment 57, wherein the living subject is a human, a non-human primate, or a mouse. 59. The method of embodiment 56 or 57, wherein administration to a living subject is by injection. 60. The method of embodiment 59, wherein the injection comprises intravenous injection, intraparenchymal injection, intracerebroventricular (ICV) injection, intravesical magna (ICM) injection, or intrathecal injection. 61.T502-050, T502-054, vAi34.0, vAi33.2, vAi45.0, vAi1.0, T502-057, T502-059, TG975, TG978, TG979, TG981 , TG982, TG987, TG988, TG995, TG996, TG997, TG999, TG1002, TG1009, TG1010, TG1011, TG1021, TG1022, TG1036, T Artificial expression constructs including G1037, TG1038, TG1045, TG1046, TG1047, TG1048, TG1049, TG1050, TG1052, CN1402, CN1457, CN1818, CN1416, CN1452, CN1461, CN1454, CN1456, CN1772, CN1427, CN1466, CN1954, CN1955, CN2137, CN2139, and CN2014. EXAMPLES

[0151] Example 1 Individual neurons or non-neuronal cells were isolated from mouse cortex by FACS and examined using an assay for transposase-accessible chromatin with next-generation sequencing (ATAC-seq). This strategy allowed for the interrogation of both abundant and very rare cell types in the same way. 25 individual or combination Cre- or Flp-driver lines in combination with reporter lines were utilized, many of which were characterized using single-cell RNA-seq (Tasic et al., 2018, Nature 563:72-78), and selective retrograde labeling to sample cell populations in the adult mouse brain. Shared GABAergic cell types have been observed across the two distal poles of the mouse cortex, while dispersed glutamatergic cell types from different cortical regions have been observed (Tasic et al., 2018, Nature 563:72-78). Thus, dissections focused on the visual cortex for glutamatergic cell types, but allowed for more extensive cortical sampling for GABAergic cell types. Reverse phase labeled cells were collected only from the visual cortex. In total, we collected 3,381 single cells from 25 driver-reporter combinations in 60 mice, 126 retrograde labeled cells from injections into three targets across seven donors, and 96 samples labeled with one retro-orbital injection from the generated viral tool. After FACS, individual cells were subjected to ATAC-seq and sequenced in batches of 60-96 samples using MiSeq (Materials and Methods in Example 1). Quality control filtering was performed to select 2,416 samples with >10,000 uniquely mapped paired end fragments, of which >10% had fragment sizes greater than 250bp and >25% of the fragments overlapped with high-depth cortical DNAse-seq peaks generated by ENCODE (Yue et al., 2014, Nature, 515:355-364).

[0152] Previous studies have shown that most recombinase driver strains label two or more transcriptome cell types (Tasic et al., 2018, Nature 563:72-78, and Tasic et al., 2016, Nat Neurosci 19:335-346). To increase the resolution of the chromatin accessibility profile beyond that provided by the driver strains, the scATAC-seq data were clustered using a novel feature-free method for the calculation of pairwise Jaccard distances. These distances were used to cluster by t-stochastic neighborhood embedding (t-SNE), followed by phenographic clustering (Figure 45). Cluster identity was then assigned by comparing the accessibility near the transcription start site (TSS ±20kb) with the scRNA-seq dataset for VISp (Tasic et al., 2018, Nature 563:72 -78) using median correlation.

[0153] Layer 5 of the visual cortex contains L5 IT neurons that project to other cortical regions, proximal projecting (L5 NP) neurons that have only local projections, and L5 PT neurons that have long axonal projections to subcortical brain regions such as the thalamus (Tasic et al., 2018, Nature 563:72-78; Harris et al., 2018, biorXiv, 292961). The driver line Rbp4-Cre labels both L5 IT and L5 PT neurons in the cortex (Tasic et al., 2016, Nat Neurosci 19:335-346). To disaggregate these populations, L5 PT and L5 IT neurons were identified within the scATAC-seq dataset based on correlation with scRNA-seq cell types, labeling of these cells with Rbp4-Cre, and retrograde labeling from a known L5 PT target region, the lateral posterior nucleus (LP) of the thalamus. We pooled populations of L5 PT and L5 IT scATAC-seq samples into subclass-specific tracks and performed searches near transcriptional marker genes for 500 bp putative enhancer elements that were specific to L5 PT or L5 IT cells and had strong sequence conservation. These regions are referred to as mouse single-cell regulatory elements (mscREs).

[0154] Putative mscREs were cloned upstream of a minimal beta globin promoter driving SYFP2 or EGFP expression in viral constructs to generate AAV (Figure 48A). These constructs were packaged into PHP.eB-serotype viruses capable of crossing the blood-brain barrier for retro-orbital injection into wild-type mice (Chan et al., 2017, Nat. Neurosci 20:1172-1179). In total, 4 mscREs were screened for L5 PT cells and 2 mscREs for L5 IT. Two weeks after retro-orbital injection, brains of infected mice were collected and screened for expression by visual inspection of native fluorescence and by enhancing SYFP2 and EGFP signals with immunohistochemistry. Three of the enhancers provided labeling of cells in L5, while the others showed no off-target or detectable labeling.

[0155] To assess the specificity of cell type labeling, stereotaxic injections of these viruses were performed in VISp, labeled cells were sorted by FACS, and scRNA-seq was performed as previously described (Tasic et al., 2018, Nature 563:72-78). scRNA-seq expression profiles were compared to the VISp reference dataset using a central classification of cell types (Materials and Methods in Example 1). The MscRE4 element conferred specificity for L5 PT cells (Figure 48B), mscRE1 conferred specificity for L5 PT cells, and mscRE16 conferred specificity for L5 IT cells. scRNA-seq of FACS-sorted cells was also performed from retro-orbital labeling of mscRE4 and mscRE1 viruses, yielding similarly specific results (>92% for mscRE4). Direct labeling of cells by stereotaxic injection induced a similar innate immune response as anterior gradient labeling, whereas retrograde injection did not cause significant upregulation of immune-related pathways at the time of harvest. For MscRE4, labeling of L5 PT cells was confirmed by electrophysiological characterization of labeled versus unlabeled cells in the cortex. Cells labeled by MscRE4 had characteristics of L5 PT neurons, whereas cells that were negative for labeling did not (Figure 49A). This demonstrates the utility of these viral tools for electrophysiology experiments targeting specific subclasses for which driver strains are not available.

[0156] L5 PT cells are often difficult to isolate from single cell suspensions when in a heterogeneous mixture with other cell types due to differences in cell viability (Tasic et al., 2016, Nat Neurosci 19:335-346, and Tasic et al., 2018, Nature 563:72-78). We bootstrapped the scATAC-seq dataset by sorting cells labeled by mscRE4 for FACS using retro-orbital injection of mscRE4-driven virus (T502-057). As expected, based on the scRNA-seq analysis, 55 of 61 high-quality mscRE4 scATAC-seq profiles clustered together with other L5 PT samples (90.2%).

[0157] While direct fluorophore labeling provided sufficient signal to sort cells by FACS or perform patch clamp experiments, the use of enhancers to drive expression of the recombinase allowed expression of previously generated mouse reporter lines driving fluorophores, activity reporters, opsins, or genes too large to package into AAV (Daigle et al., 2018, Cell 174(2):465-480, and Madisen et al., 2015, Neuron 85(5):942-958). To test the specificity of enhancer-driven recombinase expression, we cloned mscRE4 into constructs containing minimal beta globin enhancers driving dgCre (TG1009), iCre (TG1010), FlpO (TG978), or tTA2 (TG1011) and packaged them into PHP.eB viruses. These viruses were delivered by retro-orbital injection into mice carrying genetically encoded reporters for each recombinase (Ai14 for dgCre and iCre, Ai65F for FlpO, and Ai63 for tTA2). Labeling was characterized by native fluorescence sectioning and microscopy (Materials and Methods in Example 1). FlpO, dgCre, and tTA2 resulted in highly specific labeling of cells in L5 of the mouse cortex. For FlpO viruses, whole-brain microscopy was also performed using the TissueCyte system, and strong specific labeling of L5 cells was found throughout the cortex with bright labeling of pyramidal projections to subcortical targets. Finally, whole brain co-labeling of both L5 IT and L5 PT populations by retro-orbital injection of mscRE4-FlpO (to label L5 PT cells, red, TG978) and mscRE16-EGFP (to label L5 IT cells, green, TG1002) was tested in the same Ai65F animals. Separate labeling of these two cell populations in L5 by microscopy was found, demonstrating that multiple enhancer-driven viruses can be used to simultaneously label transcriptionally defined populations of cell types in the same animal.

[0158] Materials and Methods for Example 1. Mouse breeding and housing. Mice were housed at the Allen Institute for Brain Science under Institutional Care and Use Committee protocols 1508 and 1802, with no more than five animals per cage, maintained on a 12-hour day-night cycle, and provided with food and water ad libitum. Animals with anophthalmia or microphthalmia were excluded from the study. Animals were maintained on a C57BL / 6J genetic background.

[0159] Retrograde labeling. Stereotaxic injections of CAV-Cre (Hnasko et al., 2006, Proc. Natl. Acad. Sci. USA 103:8858-8863) were performed using the Paxinos adult mouse brain atlas (Paxinos & Franklin, The Mouse Brain Stereotaxic Coordinates Compact 3 rd The brains of heterozygous or homozygous Ai14 mice were subjected to FACS using coordinates obtained from {Ed., Academic Press, NY, 2008}. TdT+ single cells were isolated from VISp by FACS.

[0160] Single-cell ATAC. Single-cell suspensions of cortical neurons were generated as previously described (Gray et al., 2017, eLife 6:e21883), but omitting papain instead of proteinase K for dissociation of some samples. Individual cells with high fluorophore labeling (tdTomato or SYFP2) for neuronal sorting and low fluorophore labeling for non-neuronal cell labeling and low DAPI were then sorted into 200 μL 8-well strip tubes containing 1.5 μL of tagging reaction mix (0.75 μL Nextera reaction buffer, 0.2 μL Nextera Tn5 enzyme, 0.55 μL water). After collection, cells were spun down briefly in a benchtop centrifuge and then immediately tagged for 30 min at 37 °C in a PCR machine. After tagmentation, 0.6 μL of proteinase K stop solution was added to each tube (5 mg / mL proteinase K solution (Qiagen), 50 mM EDTA, 5 mM NaCl, 1.25% SDS) followed by incubation at 40°C for 30 min in a PCR machine. The tagged DNA was then purified using AMPure XP beads (Beckman Coulter) at a ratio of 1.8:1 resuspended beads to reaction volume (3.8 μL added to 2.1 μL), with a final elution volume of 11 μL. Libraries were indexed and amplified by adding 15uL of 2X Kapa HiFi HotStart ReadyMix and 2uL of Nextera i5 and i7 index to each tube, followed by incubation at 72°C for 3 minutes and PCR (95°C for 1 minute, 22 cycles of 98°C for 20 seconds, 65°C for 15 seconds, and 72°C for 15 seconds, then a final extension at 72°C for 1 minute). After amplification, sample concentrations were measured in duplicate using the Quant-iT PicoGreen assay (Thermo Fisher). For each sample, the average concentration was calculated by comparison to a standard curve, and the mean and standard deviation of the concentrations were calculated for all samples. Samples with concentrations more than 2 standard deviations above the mean were not used for downstream steps as they were found in initial experiments to govern the sequencing run.All other samples were pooled by combining 5 μL of each sample in a 1.5 mL tube. The combined library was then purified by adding Ampur XP beads in a 1.8:1 ratio and finally eluted in 50 μL. The mixed library was then quantified using a Bioanalyzer High Sensitivity DNA Kit (Agilent).

[0161] scATAC sequencing, alignment, and filtering. Mixed libraries containing 60-96 samples each were sequenced on an Illumina MiSeq at a final concentration of 20-30 pM. After sequencing, raw FASTQ files were aligned to the GRCm38(mm10) mouse genome using Bowtie v1.1.0 as previously described (Gray et al., 2017, eLife 6:e21883). After alignment, duplicate reads were removed using samtools rmdup, which resulted in only single copies of uniquely mapped paired reads in BAM format. For analysis, samples were filtered to remove those with fewer than 10,000 paired end fragments (20,000 reads) and at least 10% of sequenced fragments greater than 250 bp. An additional filter was created using ENCODE whole cortex DNase-seq HotSpot peaks (sample ID ENCFF651EAU from experiment ID ENCSR00COF). Samples with less than 25% of paired-end fragments that overlapped DNase-seq peaks were removed from downstream analysis. Cells that passed these criteria both had sufficient unique reads for downstream analysis and had high-quality chromatin accessibility profiles assessed by fragment size analysis. As an additional QC check, aggregated scATAC-seq data was compared to bulk ATAC-seq data from matching Cre-driver strains, when available. The aggregated single-cell dataset was found to be in good agreement with previously published bulk datasets.

[0162] Jaccard distance calculation, PCA and tSNE embedding, and density-based clustering. To compare ScATAC-seq samples, all cells were downsampled to an equal number of uniquely aligned fragments (10,000 per sample). These fragments were extended to a length of 10 kb, and then any overlapping fragments within each sample were collapsed into regions based on the outer bounds of the overlapping fragments. The number of overlapping regions between each pair of samples was then counted and divided by the total number of regions in both samples to obtain a Jaccard similarity score. These scores were converted to Jaccard distances (1-Jaccard similarity), and the resulting matrix was used as input for t-SNE. After t-SNE, samples were clustered in t-SNE space using the RPhenograph package with settings that yielded >100 clusters to obtain small groups of similar neighborhoods (Levine et al., 2015, Cell 162:184-197).

[0163] Correlation with single-cell transcription factors. Phenomenologically defined neighborhoods were assigned to cell subclasses and clusters by comparing the accessibility scores of regions within 20 kb of each transcription start site (TSS) to the median expression values ​​of scRNA-seq clusters from mouse primary visual cortex (Tasic et al., 2018, Nature 563:72-78) (Materials and Methods in Example 1). This strategy of neighborhood assignment and correlation allowed the resolution of cell types in the scATAC-seq data to approach that of the scRNA-seq data, because types that are separated too far apart would resolve into the same transcriptome type by correlation. To assess the robustness of these assignments, we used a bootstrap clustering method in which we randomly discarded 20% of the scATAC-seq samples, performed t-SNE, assigned clusters, and performed 100 comparisons with the scRNA-seq clusters. As an alternative to phenomenon clustering, these analyses were also performed by selecting the five nearest neighbors of each sample in t-SNE space and performing the same count and correlation analyses as above.

[0164] Integrating cell classes and peak calling. Aligned reads from single cell subclasses / clusters were used to create tag directories and call chromatin accessible peaks using HOMER (findPeaks-region-o auto). The resulting peaks were converted to BED format and used as input for DiffBind / differential enrichment analysis.

[0165] Viral genome cloning. Enhancers were cloned from C57Bl / 6J genomic DNA using enhancer-specific primers and Phusion high-fidelity polymerase (M0530S, NEB). Standard molecular cloning approaches were then used to insert individual enhancers into rAAV or self-complementary adeno-associated virus (scAAV) backbones containing the minimal β-globin promoter, gene, and bovine growth hormone polyA. Plasmid integrity was verified by Sanger sequencing and restriction digestion to confirm intact inverted terminal repeat (ITR) sites.

[0166] Virus packaging and titration. Prior to transfection, 10 51 μg of AAV viral plasmid, 190 μg of pHelper, and 105 μg of AAV-PHP.eB were mixed with 5 mL of Opti-MEM I medium (reduced serum, GlutaMAX; ThermoFisher Scientific) and 1.1 mL of 1 mg / mL 25 kDa linear polyethylenimine (Polysciences) in PBS at pH 4–5. This cotransfection mixture was incubated at room temperature for 10 min. Recombinant AAV of the PHP.eB serotype was generated by adding 0.61 mL of this cotransfection mixture to each of ten 15 cm dishes of HEK293T cells (ATCC) at 70–80% confluence. 24 h after transfection, the cell medium was replaced with DMEM (high glucose, L-glutamine, and sodium pyruvate; ThermoFisher Scientific) with 4% FBS (Hyclone) and 1% antibiotic-antimycotic solution. 72 hours after transfection, cells were harvested by scraping in 5 mL of medium and pelleted at 1500 rpm for 15 min at 4 °C. The pellet was suspended in a buffer containing 150 mM NaCl, 10 mM Tris, and 10 mM MgCl2 (pH 7.6) and frozen in dry ice. The cell pellet was quickly thawed in a 37 °C water bath and then passed through a syringe five times with a 21-23 G needle, followed by three more rounds of freeze / thaw and incubation with 50 U / ml benzonase (Sigma-Aldrich) for 30 min at 37 °C. The suspension was then centrifuged at 3,000 × g and purified using a layered iodixanol step gradient (15%, 25%, 40%, and 60%) by extraction of a volume below the 40-60% gradient layer interface by centrifugation at 58,000 rpm for 90 min at 18 °C in a Beckman 70Ti rotor. Virus was concentrated using Amicon Ultra-15 centrifugal filter units by centrifugation at 3,000 rpm at 4°C, reconstituted in PBS containing 5% glycerol and 35 mM NaCl, and then stored at -80°C.

[0167] Retro-orbital injection. To introduce AAV virus into the bloodstream, C57Bl / 6J, Ai14, Ai65F, or Ai63 mice (Madisen et al., 2015, Neuron 85(5):942-958) aged 21 days or older were briefly anesthetized with isoflurane and injected with 1 × 10 10 ~1×10 11 Viral genome copies (gc) were delivered to the retro-orbital sinus in a maximum volume of 50 μL or less. This approach has been previously utilized to cross the blood-brain barrier and deliver AAV virus to the mouse brain with high efficiency (Chan et al. 2017, Nat Neurosci 20(8):1172-1179). For delivery of multiple AAVs, viruses were premixed and then delivered simultaneously to the retro-orbital sinus. To analyze virally delivered transgenes in the brain, animals were allowed to recover and then sacrificed 1–3 weeks post-infection.

[0168] Stereotactic injections and tissue processing. Viral DNA was packaged into the PHP.eB serotype to produce recombinant adeno-associated viruses (rAAV) for mscRE4-minBGprom-EGFP-WPRE3 (TG981), mscRE4-minBGprom-IRES2-tTA2-WPRE3 (TG1011), and mscRE4-minBGprom-FlpO-WPRE3 (TG978) viruses (titers: 1.64x1014, 5.11x1013, and 6.00x1013, respectively), or self-complementary AAV (scAAV) for mscRE4-minBGprom-SYFP2-WPRE3-BGHpA (T502-057) virus (titer 1.34x1013) (Chan et al., Nat. Neurosci 20:1172-1170, 2017). Each virus was injected at 250 nL and 50 nL using a pressure injection system (Nanoject II, Drummond Scientific The rAAV-mscRE4-minBGprom-EGFP-WPRE3 and scAAV-mscRE4-minBGprom-SYFP2-WPRE3 viruses were delivered to both the primary visual cortex (VISp; coordinates: A / P:-3.8, ML:-2.5, DV:0.6) of male and female C57Bl6 / J and wild-type transgenic mice (Htr2a-Cre(-), SST-IRES-Cre; Ai67(-), Cck-IRES-Cre(-)) for rAAV-mscRE4-minBGprom-EGFP-WPRE3 and scAAV-mscRE4-minBGprom-SYFP2-WPRE3 viruses, respectively, or heterozygous Ai65F and Ai63 mice for rAAV-mscRE4-minBGprom-FlpO-WPRE3 and rAAV-mscRE4-minBGprom-tTa2-WPRE3 viruses, respectively. To mark the injection site, rAAV-EF1a-tdTomato or rAAV-EF1a-EGFP was co-injected with the experimental virus at a dilution of 1:10. Expression of all viruses was analyzed 14 days after injection. For tissue processing, mice were perfused transcardially with 4% paraformaldehyde (PFA) and fixed in 30% sucrose for 1–2 days. 50 μm sections were prepared using a freezing microtome, and fluorescent images of the injections were captured from the mounted sections using a Nikon Eclipse TI epifluorescence microscope.

[0169] Immunohistochemistry. Mice were perfused transcardially with 0.1 M phosphate-buffered saline (PBS) followed by 4% paraformaldehyde (PFA). Brains were removed and fixed overnight in PFA followed by an additional incubation in 30% sucrose overnight. Coronal sections (50 μm) were cut using a freezing microtome and native fluorescence or antibody enhancement was analyzed in mounted sections. Free-floating brain sections were stained using rabbit anti-GFP antibody to enhance enhanced green fluorescent protein (EGFP) fluorescence. Briefly, sections were rinsed three times in PBS, blocked for 1 h in phosphate-buffered saline (PBS) containing 5% donor donkey serum, 2% bovine serum albumin (BSA) and 0.2% Triton X-100 and incubated overnight at 4 °C in anti-GFP primary antibody (1:2000, Abcam ab6556). The next day, sections were washed three times in PBS, incubated in blocking solution containing Alexa® 488-conjugated secondary antibody (1:1500, Invitrogen), washed in PBS, and mounted in Vectashield containing DAPI (H-1500, Vector Labs). Epifluorescence images of native or antibody-enhanced fluorescence were acquired on a Nikon Eclipse Ti microscope or on a TissueCyte 1000 (Tissue Vision) system.

[0170] Viral titers were measured using quantitative PCR (qPCR) with a primer pair that recognizes a 117 bp region in the AAV2 ITR (forward: GGAACCCCTAGTGATGGAGTT (SEQ ID NO: 175), reverse: CGGCCTCAGTGAGCGA (SEQ ID NO: 176)). QPCR reactions were performed using QuantiTect SYBR Green PCR Master Mix (Qiagen) and 500 nM of primers. To determine viral titers, positive control AAV with known titers and freshly produced virus with unknown titers were treated with DNAse I. Serial dilutions (1 / 10, 1 / 100, 1 / 500, 1 / 2500, 1 / 12500, and 1 / 62500) of both the positive control and the freshly produced virus were loaded onto the same qPCR plate. A standard curve of viral particle concentration versus Cq value was generated based on the positive control virus, and the titer of the fresh virus was calculated based on the standard curve.

[0171] Single-cell RNA sequencing and cell type mapping. scRNA-seq was performed using the SMART-Seq v4 kit (Takara catalog number 634894) as previously described (Tasic et al., 2018, Nature 563:72-78). Briefly, single cells were sorted into 8-well strips containing SMART-Seq lysis buffer with RNase inhibitor (0.17 U / μL) and immediately frozen on dry ice for storage at -80°C. SMART-Seq reagents were used for reverse transcription and cDNA amplification. Samples were tagged and indexed using the NexteraXT DNA Library Preparation kit (Illumina FC-131-1096) with the NexteraXT Index Kit V2 Set A (FC-131-2001) according to the manufacturer's instructions, except that the volumes of all reagents, including cDNA, were reduced to 0.4x the recommended volumes. Full documentation of scRNA-seq procedures can be found in the 'Documentation' section of the Allen Institute data portal (http: / / celltypes.brain-map.org / ). Samples were sequenced on an Illumina HiSeq 2500 or Illumina MiSeq to a median depth of XX reads per cell as 50 bp paired-end reads. Reads were aligned to GRCm38(mm10) using STAR v2.5.3 (Dobin et al., 2013, Bioinformatics 29:15-21) in towpassMode and exonic read counts were quantified using the GenomicRanges package for R as described in Tasic et al. (2018, Nature 563:72-78). To determine the corresponding cell type for each scRNA-seq dataset, we utilized the scratch.hicat package in R (Tasic et al., 2018, Nature 563:72-78). Marker genes that distinguished each cluster were selected, and this gene panel was then used in a bootstrap central classifier that performed 100 rounds of correlation using 80% of the marker panel randomly selected in each round.

[0172] Physiology. Coronal mouse brain slices were prepared using the NMDG protection recovery method (Ting et al., 2014, Methods Mol. Biol. 1183:221-242). Mice were deeply anesthetized by intraperitoneal administration of Advertin (20 mg / kg) and perfused transcardially with an artificial cerebrospinal fluid (ACSF) solution containing 92 NMDG, 2.5 KCl, 1.25 NaH2PO4, 30 NaHCO3, 20 HEPES, 25 glucose, 2 thiourea, 5 Na-ascorbate, 3 Na-pyruvate, 0.5 CaCl2·4H2O, and 10 MgSO4·7H2O. Slices (300 μm) were sectioned on a Compresstome VF-200 (Precisionary Instruments) using zirconium ceramic blades (EF-INZ10, Cadence). After sectioning, slices were transferred to a heated (32–34°C) collection chamber filled with NMDG ACSF under constant carbonization. After 12 min, slices were transferred to a holding chamber containing ACSF consisting of (in mM): 92 NaCl, 2.5 KCl, 1.25 NaH2PO4, 30 NaHCO3, 20 HEPES, 25 glucose, 2 thiourea, 5 Na-ascorbate, 3 Na-pyruvate, 128 CaCl2·4H2O, and 2 MgSO4·7H2O, continuously bubbled with 95 / 5 O2 / CO2.

[0173] For patch-clamp recordings, slices were placed in a submerged heated (32–34 °C) recording chamber continuously perfused with ACSF under constant carbonization containing 119 NaCl, 2.5 KCl, 1.25 NaH2PO4, 24 NaHCO3, 12.5 glucose, 2 CaCl2·4H2O, and 2 MgSO4·7H2O (pH 7.3–7.4). Neurons were observed with an Olympus BX51WI microscope and infrared differential contrast optics, and a 40x water-immersion objective. Patch pipettes (3–6 MΩ) were pulled from borosilicate glass using a horizontal pipette puller (P1000, Sutter Instruments). Electrical signals were acquired using a Multiclamp700B amplifier and PClamp10 data acquisition software (Molecular Devices). Signals were digitized at 10–50 kHz (Axon Digidata1550b) and filtered at 2–10 kHz. Pipette capacitance was compensated and bridges were balanced throughout whole-cell current-clamp recordings. Access resistance ranged from 8 to 25 MΩ).

[0174] Data were analyzed using custom scripts written in Igor Pro (Wavemetrics). All measurements were performed at resting membrane potential. Input resistance (R N ) were calculated using the maximum and steady-state voltage deflections, respectively. N The maximum and steady state of the voltage sag was determined. N The resonant frequency (f R ) were determined from the voltage response to constant amplitude sinusoidal current injections, either linearly increased from 1–15 Hz for 15 s or logarithmically increased from 0.2–40 Hz for 20 s. Impedance amplitude profiles were constructed from the ratio of the fast Fourier transform of the voltage response to the fast Fourier transform of the current injection. Rcorresponds to the frequency at which the maximum impedance was measured. The majority of neurons included in the examples described here were located in the primary visual cortex (n=10 YFP+, 10 YFP-), but recordings were also made from the motor cortex (n=1 YFP+) and primary somatosensory cortex (n=4 YFP). For illustrative purposes, we also compared the properties of YFP+ and YFP- neurons to 32 L5 pyramidal neurons located in the somatosensory cortex from uninfected mice. To classify these neurons as IT-like or PT-like, we used the Partition Analysis of Clustering (diana) from the cluster package in R (Maechler and Rousseeuw, 2012, R package version 1(2),56). I h Ih-associated membrane properties are known to distinguish IT and PT neurons across many brain regions (Baker et al., 2018, J. Neurosci. 38:5441-5455. Thus, features included in the clustering were Ih-associated membrane property-sag ratio, R N and f R To assess the statistical significance of the clustering, we used the sigclust package in R (Huang et al., 2015, J Comput Graph Stat 24(4):975-993).

[0175] Example 2 Predicted brain-wide labeling of neuronal subclasses with enhancer-driven adeno-associated viruses (AAVs). Individual neurons and non-neuronal cells from transgenically labeled mouse cortex were isolated by fluorescence-activated cell sorting (FACS) and examined using an assay for transposase-accessible chromatin with next-generation sequencing (scATAC-seq). Buenrostro et al., 2015, Nature 523:486-90), Cusanovich et al., 2015, Science(80):348,910-914. This strategy allows the investigation of both abundant (e.g., intratelencephalic L4 IT neurons in layer 4, 17% of primary visual areas of the cortex, VISP, neurons) and very rare cell types (e.g., Sst Chodl neurons, 0.1% of VISp neurons) in the same way. To broadly and specifically sample cells in the mouse brain, we used 25 different Cre or Flp-driver lines, or combinations of them crossed to appropriate reporter lines (Figure 42). Many of the same lineages were previously characterized by single-cell RNA-seq. Tasic et al., 2018, Nature 563, 72-78. In addition, we employed retrograde labeling with recombinase-expressing viruses to selectively sample cells with specific protrusions (Retro-ATAC-seq). This method yielded scATAC-seq libraries of comparable quality to previously published scATAC-seq studies (Figure 43, Figure 44). Buenrostro et al., 2015, Nature 523, 486-90; Pliner et al., 2018, Mol. Cell 71, 858-871.e8; Cusanovich et al., 2015, Science 348, 910-4.

[0176] To generate scATAC-seq data that are directly comparable to scRNA-seq datasets (Tasic et al., 2018, Nature 563:72-78), dissections focused on the visual cortex for glutamatergic cell types, but allowed for more extensive cortical sampling for GABAergic cell types. This strategy is rooted in the observation that GABAergic cell types are shared across the two distal poles of the mouse cortex, while glutamatergic cell types differ between different cortical regions (Tasic et al., 2018, Nature 563:72-78). Retro-ATAC-seq cells were collected only from the visual cortex. In total, we collected 3,381 single cells from 25 driver-reporter combinations in 60 mice, 126 retrogradely labeled cells from injections into three targets across seven donors, and 96 samples labeled by retro-orbital injection of one of the viral tools generated according to this disclosure. After FACS, individual cells were processed using ATAC-seq and sequenced in batches of 60-96 samples using MiSeq (Materials and Methods in Example 2). Quality control (QC) was performed by filtering 2,416 samples with >10,000 uniquely mapped paired end fragments. >10% of the paired end fragments had a fragment size greater than 250 bp, and >25% of the fragments overlapped with high-depth cortical DNAse-seq peaks generated by Encyclopedia of DNA Elements (ENCODE) (Figure 42). Yue et al., 2014, Nature 515:355-64.

[0177] Previous studies have shown that most recombinase driver strains label more than one transcriptome cell type. Tasic et al., 2018, Nature 563:72-78; Tasic et al., 2016, Nat. Neurosci. 19, 335-346. To increase the cell type resolution of chromatin accessibility profiles beyond that provided by the driver strains, scATAC-seq data were clustered using a novel, featureless method for calculating pairwise Jaccard distances. These distances were used for principal component analysis (PCA) and t-stochastic neighbor embedding (t-SNE), followed by phenographic clustering (Figure 45, Materials and Methods in Example 2). Levine et al., 2015, Cell 162:184-197. This clustering method clearly grouped cells from class-specific driver strains, separating them into multiple clusters as expected based on the transcriptional analysis. Cluster identity was then assigned by comparison of accessibility near transcription start sites (TSS ±20 kb) to the scRNA-seq dataset generated for VISp using median correlation (Figure 45, Materials and Methods in Example 2). Tasic et al., 2018, Nature 563:72-78. Subclass-level assignments for each driver line were found to closely match those observed for the same driver line by scRNA-seq. Once assigned, clusters from the same subclass (e.g., Vip or layer 5, L5, IT) or different cell types (e.g., Pvalb Vipr2) were aggregated for peak calling and examination of accessibility patterns (Figure 46A-D). Comparison of these scATAC-seq aggregate profiles with previously published ATAC-seq from cortical populations shows strong correspondence between aggregate profiles and populations, and comparison with previously published cortical scATAC-seq data demonstrates increased cell type resolution using the current dataset generated by this lab. Cusanovich et al.,2018,Cell 174,1309-1324.e18, Preissl et al.,2018,Nat.Neurosci.21:432-439.

[0178] L5 in the mouse cortex contains three major subclasses of excitatory neurons: intertelencephalic (IT) neurons that project to other cortical regions, proximal projection (L5 NP) neurons with mostly localized projections, and corticosubgrade (a subset of which is called pyramidal tract, L5 PT) neurons that project to subcortical brain regions such as the thalamus. Tasic et al., 2018, Nature 563:72-78; Harris et al., bioRxiv, 2018 doi:10.1101 / 292961. The driver line Rbp4-Cre labels both L5 IT and L5 PT neurons in the cortex, but not L5 NP. Tasic et al., 2018, Nature 563:72-78. ScATAC-seq clustering identified L5 PT and L5 IT neurons in the generated dataset based on correlation with scRNA-seq cell types (Figure 45). Labeling of these cells with Rbp4-Cre and retrograde labeling from a known L5PT target region, the lateral thalamic nucleus (LP), verified that these cells were likely L5 IT (Rbp4-Cre+ only) and L5 PT neurons (Rbp4 and LP Retro-ATAC-seq). A search was performed near transcriptional structural marker genes for 500bp putative enhancer regions that were specific to L5 PT or L5 IT cells and had strong sequence conservation (Figure 46A-D). These regions are referred to as mouse single cell regulatory elements (mscRE, Figure 47).

[0179] To functionally test the mscREs, their genomic sequences were cloned upstream of a minimal β-globin promoter driving the fluorescent proteins SYFP2 or EGFP in a recombinant adeno-associated virus (rAAV) genome (Figure 48A). These constructs were packaged using the PHP.eB serotype, which can cross the blood-brain barrier, allowing delivery by retro-orbital injection. Four mscREs were screened for L5 PT cells and two for L5 IT (Figure 47). Chan et al., 2017, Nat. Neurosci. 20:1172-1179. Two weeks after retro-orbital injection, brains of infected mice were collected and screened for expression by visual inspection of native fluorescence and by enhancing SYFP2 and EGFP signals with immunohistochemistry. Two of these enhancers provided specific labeling of cells in L5 (Figure 48C, right) and were selected for further validation.

[0180] To evaluate the utility of enhancer-driven fluorophores as viral tools, retro-orbital injections of mscRE4-SYFP2 virus were performed in additional animals. From two of these, L5 of VISp was dissected, labeled cells were sorted by FACS, and scRNA-seq was performed as previously described. Tasic et al., 2018, Nature 563:72-78. scRNA-seq expression profiles were compared to the VISp reference dataset using a central classification of cell types (Materials and Methods in Example 2). Tasic et al., 2018, Nature 563:72-78. MscRE4-SYFP2 virus was found to provide >91% specificity for L5 PT cells within L5 (Figure 49B). Labeling of L5 PT cells was confirmed by electrophysiological characterization of labeled vs. unlabeled cells in the cortex (Figures 49B, 50A, 51). Cells labeled by mscRE4 had characteristics of L5 PT neurons, whereas cells that were negative for labeling more closely matched L5 IT neurons. Baker et al., J.Neurosci.38,5441-5455,2018. This experiment demonstrates the utility of these viral tools for electrophysiology experiments targeting specific neuronal subclasses for which driver lines are not available. Finally, we tested direct stereotaxic injection of mscRE4 fluorophore virus into the VISp. Although specificity was dependent on the injection volume, we found that extremely bright and specific labeling could be achieved by using stereotaxic injections, likely reflecting a loss of specificity when the viral genome copy number per cell was high (Figure 52A, Figure 52B).

[0181] L5 PT cells are often difficult to isolate from single cell suspensions when in a heterogeneous mixture with other cell types due to differences in cell viability, and there are currently no reliable driver lines to selectively label L5 PT cells. Tasic et al., 2018, Nature 563:72-78; Tasic et al., Nat. Neurosci. 19, 335-346, 2016. We enriched our scATAC-seq-generated datasets by sorting mscRE4-labeled cells for FACS using retro-orbital injection of mscRE4-SYFP2 virus. As predicted based on scRNA-seq analysis, 55 of 61 high-quality mscRE4 scATAC-seq profiles clustered together with other L5 PT samples (90.2%).

[0182] Although fluorophore expression provided sufficient signal to sort cells by FACS or perform patch clamp experiments, expression of recombinase from specific enhancer viruses would expand the utility of these tools as drivers for reporter lines expressing fluorophores, activity reporters, opsins, or other genes that are too large to package into AAV. See Daigle et al., 2018, Cell 174(2):465-480; Madisen et al., Neuron, 85, 942-958, 2015. To test the specificity of enhancer-driven recombinase expression, mscRE4 was cloned into constructs containing a minimal β-globin promoter driving destabilized Cre (dgCre), iCre, FlpO, or tTA2, and the constructs were packaged into PHP.eB virus (Figure 53). These viruses were delivered by retro-orbital injection into mice carrying genetically encoded reporters for each recombinase (Ai14 for dgCre and iCre, Ai65F for FlpO, and Ai63 for tTA2). Madisen et al., Nat. Neurosci. 13, 133-40, 2010; Madisen et al., Neuron 85, 942-958, 2015; Daigle et al., Cell 174, 465-480.e22, 2018. Labeling was assessed by sectioning and microscopy of native fluorescence (Figure 53). FlpO, iCre, and tTA2 viral constructs led to labeling of cells in L5 of the mouse cortex with various levels of specificity, whereas dgCre showed non-specific labeling of cortical layers. The same strategy was applied to inject two different titers (1 × 10 10 and 1×10 11We screened both mscRE4 and mscRE16 drivers of FlpO, iCre, and / or tTA2 by retro-orbital injection at 100 ng / ml (total genome copies, GC). The specificity and completeness of labeling was found to be highly dependent on both the injection titer and the combination of recombinase reporters used in these experiments (Figure 58). Based on these experiments, a single titer of each FlpO virus was selected for deep characterization and additional animals were injected for scRNA-seq and whole-brain two-photon tomography by TissueCyte (Figures 57A-C, 59A, 59B, and 60). Each of these viruses was found to have a high level of layer and subclass specificity in the cortex, with 87.5% of cells labeled by mscRE4-FlpO corresponding to L5 PT cells (Figure 57A) and 42% of cells labeled by mscRE16-FlpO corresponding to L5 IT cells (Figure 57C), with little overlap. TissueCyte images revealed that the two viruses labeled additional subcortical populations (mscRE4 in APr, CEa, and hip, Figure 59A, and mscRE16 in pon, BLA, and hip, Figure 60).

[0183] Viruses can also be co-administered to label multiple cell populations, either exclusively or crosswise (Figure 59C). This strategy reduces the need for triple or quadruple crossing to obtain co-labeled populations of cells. Whole brain co-labeling of both L5 IT and L5 PT populations was tested by retro-orbital injection of mscRE4-iCre (to label L5 PT cells in green) and mscRE16-FlpO (to label L5 IT cells in red) in the same Ai65F;Ai140 animals (Figure 59D). Separate labeling of these two cell populations was found at L5 by microscopy (Figure 59E), demonstrating that multiple enhancer-driven viruses can be used to simultaneously label or disorder prospectively defined subclass populations in the same animal.

[0184] Materials and Methods for Example 2. Mouse breeding and housing and retrograde labeling were performed as described in the Materials and Methods section of Example 1.

[0185] Single cell ATAC. Single cell suspensions of cortical neurons were generated as previously described, except for using papain instead of pronase for some samples and adding trehalose to the dissociation and sorting medium for some samples. Gray et al.,Elife 1-30,2017 doi:10.7554 / eLife.21883. FACS was then used to individually select live neuronal cells by gating on negative DAPI (and positive fluorophore label (tdTomato, EGFP, or SYFP2)) or live non-neuronal cells by gating on negative DAPI and negative fluorophore label.

[0186] For GM12878 scATAC, cells were obtained from the Coriell Institute and grown in T25 culture flasks in RPMI1640 medium (Gibco, Thermo Fisher Cat catalog number 11875093) supplemented with 10% fetal bovine serum (FBS) and Penn / Strep. At 80% confluence, cells were transferred to a 15 mL conical tube, centrifuged, and washed with PBS containing 1% FBS. Cells were then resuspended in 1% FBS and 2 ng / mL DAPI (DAPI * The cells were resuspended in PBS containing 20 mM NaCl (Life Technologies catalog number D1306).

[0187] Single cells were sorted into 200 μL 8-well strip tubes containing 1.5 μL of tagmentation reaction mixture (0.75 μL Nextera reaction buffer, 0.2 μL Nextera Tn5 enzyme, 0.55 μL water). After collection, cells were spun down briefly in a benchtop centrifuge and then immediately tagged for 30 min at 37 °C in a thermocycler. After tagmentation, 0.6 μL of proteinase K stop solution was added to each tube (5 mg / mL proteinase K solution (Qiagen), 50 mM EDTA, 5 mM NaCl, 1.25% SDS) followed by incubation for 30 min at 40 °C in a thermocycler. The tagged DNA was then purified using AMPure XP beads (Beckman Coulter) at a ratio of 1.8:1 resuspended beads to reaction volume (3.8 μL added to 2.1 μL) for a final elution volume of 11 μL. Libraries were indexed and amplified by adding 15uL of 2X Kapa HiFi HotStart ReadyMix and 2uL of Nextera i5 and i7 index to each tube, followed by incubation at 72°C for 3 minutes and PCR (95°C for 1 minute, 22 cycles of 98°C for 20 seconds, 65°C for 15 seconds, and 72°C for 15 seconds, then a final extension at 72°C for 1 minute). After amplification, sample concentrations were measured in duplicate using the Quant-iT PicoGreen assay (Thermo Fisher). For each sample, the average concentration was calculated by comparing to a standard curve, and the mean and standard deviation of the concentrations were calculated for each sample batch. Samples with concentrations more than 2 standard deviations above the mean were not used for downstream steps, as they were found in initial experiments to govern the sequencing run. All other samples were pooled by combining 5μL of each sample in a 1.5mL tube. The combined library was then purified by adding Ampure XP beads at a ratio of 1.8:1, with a final elution of 50 μL. The mixed library was then quantified using a Bioanalyzer High Sensitivity DNA Kit (Agilent).

[0188] scATAC sequencing, alignment, and filtering were performed as described in the Materials and Methods section of Example 1. Jaccard distance calculations, PCA and tSNE embedding, and density-based clustering were also performed as described in the Materials and Methods section of Example 1, except that when comparing ScATAC-seq samples, fragments were extended to 1 kb in length and samples were clustered in t-SNE space using the RPhenogh package with k=6.

[0189] Correlation with single-cell transcription factors. Phenomenologically defined neighborhoods were assigned to cell subclasses and clusters by comparison of accessibility of transcription start site (TSS) neighborhoods to the median expression values ​​of scRNA-seq clusters at cell type (e.g., L5 PT Chrna6) and subclass (e.g., Sst) levels from mouse primary visual cortex. Tasic et al., Nature 563, 72-78, 2018. To score each transcription start site (TSS), the TSS position was obtained from the RefSeq gene annotation provided by the UCSC Genome Browser database and windows from the TSS + / - 20kb were generated. The number of fragments of all samples in each cluster that overlapped with these windows was then counted. For comparison, differentially expressed marker genes were selected from the Tasic et al., Nature 563, 72-78, 2018 scRNA-seq dataset using the scratch.hicat package for R. We then correlated the expression cluster scores with the log-transformed median exon read count values ​​of this marker gene set for each scRNA-seq cluster from primary visual cortex, and assigned the transcriptomic cell type with the highest score correlation. This strategy of neighbor assignment and correlation allowed the resolution of cell types in the scATAC-seq data to approach that of the scRNA-seq data, because types that are split too far apart would be resolved into the same transcriptional subclass or type by correlation.

[0190] scATAC-seq grouping and peak calling. For downstream analysis, cell type assignments were grouped to subclass level except for highly distinct cell types (Lamp5 Lhx6, Sst Chodl, Pvalb Vipr2, L6 IT Car3, CR, and Meis2). Unique fragments of all cells within these subclasses / distinct type groups were aggregated into BAM files for analysis. Aligned reads from single cell subclasses / clusters were used to create tag directories and chromatin accessibility peaks were called using HOMER with the settings "findPeaks-region-o auto". Resulting peaks were converted to BED format. Heinz et al.,Mol.Cell 38,576-589,2010.

[0191] ATAC population of Sst neurons. ATAC-seq of populations of neurons derived from Sst-IRES2-Cre;Ai14 mice was performed as previously described. Gray et al., Elife 1-30, 2017. doi:10.7554 / eLife.21883. Briefly, cells from adult mouse visual cortex were microdissected and FACS sorted into 8-well strips as described above, but at 500 cells per well instead of single cells as in scATAC-seq. Cell membranes were lysed and nuclei pelleted before resuspending in the same tagging buffer as described above in a higher volume (25 μL). Staining was carried out for 1 hour at 37° C., followed by the addition of 5 μL cleanup buffer (900 mM NaCl, 300 mM EDTA), 2 μL 5% SDS, and 2 μL proteinase K, incubation at 40° C. for 30 minutes, and cleanup with AMPure XP beads (Beckman Coulter) at a ratio of 1.8:1 beads to reaction volume. Samples were amplified using KAPA HotStart Ready Mix (Kapa Biosystems, Cat. No. KK2602) and 2 uL each of Nextera i5 and i7 primers (Illumina), quantified using a Bioanalyzer, and sequenced on an Illumina MiSeq.

[0192] Comparison with bulk ATAC-seq data. For comparison with previously published studies, data were used from Mo et al., Neuron 86, 1369-1384, 2015, GEO accession number GSE63137 for Camk2a, Pvalb, and Vip neuronal populations, and from Gray et al. (Elife 1-30, 2017), GEO accession number GSE87548 for Cux2, Scnn1a-Tg3, Rbp4, Ntsr1, Gad2, mES, and genomic controls. Mo et al., Neuron 86, 1369-1384, 2015, Gray et al., Elife 1-30, 2017 doi:10.7554 / eLife.21883. For these comparisons, the population ATAC-seq of Sst neurons described above was also included. For each population, reads from all replicates were merged and each region was downsampled to 6.4 million reads. Peaks were then called using HOMER as described above for the aggregated scATAC-seq. BED-formatted peaks of scATAC-seq aggregates with and without the bulk ATAC-seq dataset were used as input for comparison using the DiffBind package in R as previously described. Gray et al.,Elife 1-30,2017 doi:10.7554 / eLife.21883.

[0193] Identification of mouse single-cell regulatory elements. To identify genes that were uniquely expressed in the L5 PT and L5 IT subclasses across all glutamatergic subclasses, a targeted search for mouse single-cell regulatory elements (mscREs) was performed by performing pairwise differential expression analysis of scRNA-seq clusters. Unique peaks were then searched for within 1 Mbp of each marker gene, and these peaks were manually inspected for low or no accessibility in off-target cell types and for conservation. In cases where highly conserved regions overlapped the peak region but the peak was not centered on the highly conserved region, peak selection was adjusted to include adjacent highly conserved sequences. For cloning, primer searches were centered on a 500 bp region centered in the middle of the selected peak region and included up to 100 bp on either side. Final region selection and PCR primers are shown in Figure 47.

[0194] As described in the Materials and Methods section of Example 1, the following techniques were performed: viral genome cloning; viral packaging, titration, and titer determination; retro-orbital injection; stereotactic injection (except that each virus was delivered bilaterally at 250 nL, 50 nL, and 25 nL); immunohistochemistry; and single-cell RNA sequencing and cell type mapping.

[0195] Comparison with previous scATAC-seq studies. For comparison with the GM12878 dataset, raw data from Cusanovich et al. (Science 348, 910-4, 2015) were downloaded from GEO accession number GSE67446, GEO accession number GSE65360 from Salav et al. (2015), and GEO accession number GSE109828 from Pliner et al. (Mol. Cell 71, 858-871.e8, 2018). See Buenrostro et al., Nature 523, 486-90, 2015. Processed 10x Genomics data were obtained from the 10x Genomics website. Samples from Buenrostro, Cusanovich, Pliner, and Gm12878 from this lab were aligned to the hg38 human genome using the same bowtie pipeline described above for mouse samples to obtain fragment positions per cell. 10x Genomics samples were analyzed using the fragment positions provided by 10x Genomics. For comparison to TSS regions, we used the RefSeq gene tables provided by the UCSC Genome Browser database for hg19 (for the 10x data) and hg38 (for the other datasets). For comparison to ENCODE peaks, we used ENCODE Gm12878 DNA-seq HotSpot results from ENCODE experiment ID ENCSR000EJD aligned to hg19 (ENCODE file ID ENCFF206HYT) or hg38 (ENCODE file ID ENCFF773SPT).

[0196] For comparison with previously published mouse cortex datasets, raw FASTQ files were downloaded from GEO accession number GSE111586 from Cusanovich, et al. (Cell 174, 1309-1324.e18, 2018) and GEO accession number GSE100033 from Preissl et al. Nat. Neurosci. 21, 1, 2018. Multiplexed files were aligned to the mm10 genome using Bowtie v1.1.0 and demultiplexed using an R script before removing duplicate position alignments. Only barcodes with >1,000 mapped reads were retained for analysis. Per-barcode statistics were processed using the same algorithms used for per-cell statistics from datasets generated by this lab, and Cusanovich et al., Cell 174, 2018 datasets that passed established QC criteria were subjected to the same analysis pipeline as data generated by this lab after demultiplexing and duplicate read removal. Metadata from Cusanovich et al. (Cell 174, 2018) was obtained from the Mouse sci-ATAC-seq Atlas website http: / / atlas.gs.washington.edu / mouse-atac / .

[0197] Physiological, patch clamp recordings, and data analysis were performed as described in the Materials and Methods section of Example 1.

[0198] TissueCyte Imaging and Analysis. TissueCyte images were collected, registered, and segmented as previously described. Oh et al. (Nature 508,207-214,2014). After registration, 3D arrays of signal binned to 25um voxels were analyzed in R by background subtraction and averaging the signal in the best structure in the Allen Brain Atlas structural ontology. To propagate signal from fine to coarse structures in the ontology, a hierarchical calculation was performed that assigned each parent the maximum value of its child nodes in the ontology from the bottom to the top of the ontology. The ontology was then filtered to remove very fine structures, and the resulting ontology relationships and structure scores were displayed using the classification and metacoder R packages for R. Foster et al.,bioRxiv 071019,2016 doi:10.1101 / 071019.

[0199] Software for analysis and visualization. Analysis and visualization of ScATAC-seq and transcript structure datasets were performed using R v. 3.5.0 or higher in Rstudio IDE (integrated development environment for R) or Rstudio Server Open Source Edition, and the following packages: data.table, dplyr, Matrix, matrixStats, purrr, and reshape2 for general data analysis and manipulation; GenomicAlignments, GenomicRanges, and rtracklayer for analysis of genomic data; cowplot, ggbeeswarm, ggExtra, ggplot2, and rgl for plotting and visualization; Rphenograph and Rtsne for clustering and size reduction; scratch.hicat and scratch.io for analysis of transcript structure datasets; metacodeR and taxa for taxonomic analysis and visualization; and plater for management of plate-based experimental results and metadata.

[0200] Example 3 Human single neuron epigenetic assessment of neocortical cell classes. Primates, and especially the human neocortex, are greatly expanded and complex compared to other mammals such as rodents (Zeng et al., Cell. 149, 483-496, 2012; Rakic, Nat Rev Neurosci. 10, 724-735, 2009). The expansion of the neocortex enables human-centric abilities such as language and reasoning that are disrupted in human diseases such as schizophrenia and autism (King et al., JAMA Netw Open. 1, e184777-e184777, 2018; van den Heuvel et al., JAMA Psychiatry. 70, 783-792, 2013). This structure contains billions of cells, grouped into tens if not hundreds of molecularly defined cell types (Zeisel et al., Science. 347, 1138-1142, 2015; Tasic et al. Nat. Neurosci. 19, 335-346, 2016; Tasic et al., Nature. 563, 72, 2018; Hodge et al., bioRxiv, 384826, 2018).

[0201] To understand these cells and their regulation, we generated high-quality datasets of accessible chromatin using both bulk and single human brain nuclei via ATAC-seq from multiple fresh neurosurgical specimens (bulk n=5, single n=14) (Buenrostro et al., Nature. 523, 486-490, 2015; Graybuck et al., bioRxiv, 525014, 2019; Gray et al., eLife Sciences. 6, e21883, 2017). We prepared 3660 single-nucleus ATAC-seq libraries (median of 48542 unique mapping reads) and used 2858 quality-filtered nuclei for clustering and mapping (Figure 75A, and Materials and Methods in Example 3). 27 ATAC-seq clusters were identified that mapped to 18 human brain temporal lobe transcriptionally defined cell types (Hodge et al., bioRxiv, 384826, 2018) (Figure 75B). These cell types span three major classes of brain cell types: excitatory, inhibitory, and non-neuronal; and 11 cell type subclasses: excitatory layer 2 / 3 (L23), layer 4 (L4), layer 5 / 6 intratelencephalic (L56IT), and deep non-intratelencephalic neurons (DL); inhibitory LAMP5, VIP, SST, and PVALB neurons, as well as non-neuronal astrocytes, microglia, and oligodendrocytes / OPCs. Identified cell types were typically identified in a predictive sorting strategy (Figure 75B), and all cell types were populated by multiple specimens.

[0202] To identify putative regulatory elements within each subclass, data were aggregated for all nuclei within each subclass and subclass-specific peaks were called with Homer (Heinz et al., Molecular Cell. 38, 576-589, 2010), revealing peaks in close proximity to recently identified transcriptional subclass-specific marker genes (Hodge et al., bioRxiv, 384826, 2018), validating the clustering and mapping strategy. Furthermore, within peaks, Chroma VAR (Schep et al. Nature Methods. 14, 975-978, 2017) identified predicted cell type characteristic transcription factor motifs, including DLX1 in inhibitory neurons and NEUROD6 in lower layer excitatory neurons, and these accessibility correlated with their transcript abundance across subclasses (Hodge et al., bioRxiv, 384826, 2018) (paired t-test for correlation, DLX1 t=3.0 p<0.01, NEUROD6 t=5.4 p<0.001). These observations indicate strong concordance between RNA-seq and ATAC-seq data modalities.

[0203] To assess accessibility and correspondence between epigenetic modifications and primary sequences, we calculated overlap between subclass snATAC-seq peaks and previously identified differentially methylated regions (DMRs) (Lister et al., Science.341, 1237905, 2013; Luo et al., Science.357, 600-604, 2017) and aggregated by subclass. For all cell subclasses, we observed greater overlap of ATAC-seq peaks and DMRs than would be expected by chance alone (Figure 77E), providing the epigenetic components of thousands of independently validated human neocortical cell subclasses.

[0204] To explore the relationship of these elements to genes, the cell subclass peaks were subset into sets of all peaks, subclass-specific peaks, transcription start site (TSS)-distal peaks (>20 kb from any RefSeq TSS), and intersections of subclass-specific peaks with TSS-distal peaks. This analysis revealed particularly strong DMR overlap in the TSS-distal peaks (ANOVA F = 3.6; all peaks vs. TSS-distal p < 0.05; all peaks vs. TSS-distal and subclass-specific; p < 0.01 [Sidak post hoc corrected probability]). To further characterize the ATAC-seq peaks, their primary sequence conservation was then calculated by phyloP score (Pollard et al., Genome Res. 20, 110-121, 2010). All cell subclass peak sets were on average more conserved than random DNA stretches. In particular, the TSS-distal peak had a higher conservation score than all peaks (paired t-test, p<.001, t=5.4, df=10), and the inhibitory neuron subclass peaks had significantly greater conservation than those of the excitatory neuron subclass (heterosedastic t-test, p<0.05 for all peak sets, t=2.6, df=5.6; p<0.05 for TSS-distal peak set, t=2.5, df=5.9), consistent with previous observations by Luo et al. (Science. 357, 600-604, 2017).

[0205] Overall, the high conservation and validation via molecularly independent techniques together suggest that ATAC-seq identifies bona fide functional genomic elements that confer human neocortical cell type identity.

[0206] To count human-accessible chromatin regions shared with mouse ("conserved") and human-specific regions ("divergent"), Jaccard similarity coefficients between human peaks and mouse peaks mapped to the human genome were calculated for all cell subclasses. All mouse subclasses display the highest Jaccard similarity enrichment with their orthologous human subclasses and all but one human subclass map, as expected reciprocally. In addition, non-neurons show the strongest inter-species epigenetic similarity, followed by inhibitory and excitatory neurons, which show the weakest but still greater than random similarity. Quantifying the conserved and divergent peaks for each species revealed thousands of conserved peaks in each category, many more than would be expected by chance alone. Furthermore, we observe a much greater primary sequence conservation in the conserved peaks than in the divergent peaks in both species (heterosedastic t-test; human t=10.3, p<0.001; mouse t=6.6, p<0.001), suggesting that these elements carry out important evolutionarily shared functions. We observed that across 11 cortical subclasses, 34±10% (mean±sd) of all human-accessible chromatin elements are conservedly detected in mouse. In conclusion, many functional genomic elements are conserved between human and mouse and across all major neocortical cell subclasses.

[0207] These data, which established a high-quality, high-resolution catalog of human neocortex accessible genomic elements, were used as a tool to link cell subclasses with brain diseases and traits. Linkage disequilibrium score regression (LDSC, see Bulik-Sullivan et al. Nature Genetics. 47, 291-295, 2015; Finucane et al., Nat Genet. 47, 1228-1235, 2015) was used to find significant associations between human brain cell subclass ATAC-seq peaks and SNPs identified in 15 well-powered genome-wide association studies of brain diseases or traits (see Materials and Methods in Example 3). Overall similar association patterns were observed using either ATAC-seq peaks or DMRs (Lister et al., Science. 341, 1237905, 2013; Luo et al., Science. 357, 600-604, 2017), as well as generally weak associations of outgroup traits (Crohn's disease) and outgroup peak sets (The ENCODE Project Consortium, Nature. 489, 57-74, 2012), suggesting that these analyses are robust to experimental techniques.

[0208] We divided the subclass peak sets into conserved and divergent subsets and generally found stronger associations between brain diseases / traits and conserved peaks. We observed significant associations (passing Bonferroni-corrected p-value significance cutoffs) between multiple neuronal (but not non-neuronal) subclass peak sets and educational achievement and schizophrenia, similar to previous analyses of RNA-seq data (Skene et al. Nature Genetics. 50, 825, 2018; Girdhar et al. Nature, Neuroscience. 21, 1126-1136, 2018; Cusanovich et al., Cell. 174, 1309-1324.e18, 2018), and found these relationships to be stronger in conserved regions than in divergent regions. Similar to previous reports, the strongest association was also observed between microglial peaks and Alzheimer's disease (Skene et al. Nature Genetics. 50, 825, 2018; Girdhar et al. Nature, Neuroscience. 21, 1126-1136, 2018; Cusanovich et al., Cell. 174, 1309-1324.e18, 2018), although these results did not pass the cutoff for significance, likely due to low overall heritability and therefore power dependence in Alzheimer's disease studies. Interestingly, this microglia-Alzheimer's association was stronger in divergent peaks than in conserved peaks, suggesting that human-specific modes of microglial gene expression contribute to Alzheimer's pathology.

[0209] Because the number of divergent peaks in humans outnumbered the conserved peaks, we inferred that the overall heritability of neuronal-associated traits (educational achievement and schizophrenia) was largely conserved or divergent. Summing the subclass-associated heritability revealed that conserved peaks contained the majority of heritability and were significantly more numerous than divergent peaks. Overall, these analyses suggest that cross-species epigenetic analyses may enable the discovery of conserved functional genomic elements that illuminate human health and disease.

[0210] To determine whether these functional genomic elements could provide useful genetic tools, several subclass-specific peaks were cloned into adeno-associated virus (AAV) reporter expression vectors to test subclass-specific enhancer activity (Dimidschstein et al. Nature, Neuroscience. 19, 1743-1749, 2016). Peaks were selected to be near known subclass-specific marker genes from RNA-seq (Hodge et al., bioRxiv, 384826, 2018) and to exhibit subclass-specific accessibility. Several enhancers were discovered (Zerucha et al., J. Neurosci. 20, 709-721, 2000) that drove distinct reporter expression patterns in mice consistent with expected subclass-specific accessibility profiles (Figure 78), suggesting that the ATAC-seq enhancer discovery described herein is a generalizable strategy for identifying cell class / type-specific genetic tools.

[0211] Because these tools are non-species restricted, the study focused on eHGT_022 near the LAMP5 / VIP cell marker CXCL14, which is conservedly accessible in LAMP5 and VIP neuronal subclasses in humans and mice. AAV vectors driving either the human or mouse orthologue of eHGT_022 were both sufficient to drive expression in upper layer enriched interneurons in both mice and humans, and these reporter-positive cells were found to correspond specifically to LAMP5 and VIP neurons in both mice and humans. These observations, together with those in a companion manuscript (Graybuck et al., bioRxiv, 525014, 2019), suggest that ATAC-seq can identify specific cell type and subclass enhancers enabling a useful genetic tool in humans and other species.

[0212] Human brain function and disease are often difficult to study because model organisms do not recapitulate human brain circuits or display clear clinically relevant phenotypes. In particular, functionally relevant cell types are unknown in many conditions, leading to undertreatment of many debilitating brain disorders. It is therefore important to understand human brain-specific circuit components and their regulatory devices to provide pathways for therapeutic intervention. In this study, we cataloged human neocortical functional genomic elements with cell type precision, providing the most high-resolution dataset of human brain chromatin accessibility to date. This advances our knowledge of human brain chromatin structure and reveals cell type-specific logic in gene regulation. This knowledge is expected to guide models of human cognitive circuits as well as aid in gene therapy for unmet clinical needs.

[0213] Materials and Methods for Example 3. Neurosurgical Tissue Acquisition. A pipeline was established for the regular delivery of fresh neurosurgical tissue to the Allen Institute for processing from a network of surgeons in Seattle WA. These samples are naturally resected to access epilepsy foci or tumors. Experiments are limited to the temporal cortex, most often the middle temporal gyrus. These samples are immersed in ACSF.7 (recipe in Table 3) prior to carbonization, rapidly transported to the laboratory with carbonization, sliced ​​into 350 μm slices on a vibratome, and continued carbonization in ACSF.7 until dissociation.

[0214] Bulk tissue ATAC-seq. MTG tissue slices were harvested after bubbling in ACSF.7 for up to 16 h and processed with NeuroTrace 500 / 525 (catalog number N21480 from ThermoFisher Scientific, 1 / 100 in ACSF.7) to highlight laminar cortical structures. Fine forceps were used to trim white matter and meningeal tissue, then layers 1-6 were dissected into six different low-binding Eppendorf 1.5 mL tubes (MilliporeSigma catalog number Z666548) under a fluorescent microscope as in Hodge et al. (bioRxiv, 384826, 2018). The supernatant was discarded and replaced with 50-100 µL of Nextera DNA Library Reaction (numbers FC-121-1031 from Illumina) containing 0.1% IGEPAL-630 (substitute for NP-40) and then transposed by vigorously pipetting up and down 25-50 times using a P200 pipette, followed by incubation at 37 °C for 1 h. The reaction was then quenched by adding 1 mL of ice-cold nuclei isolation medium, the sample was pelleted at 1000 g for 5 min at 4 °C, resuspended in 1 mL of fresh homogenization buffer (recipe in Table 3), the nuclei were released from the sample using 10-15 strokes of a loose Dance pestle followed by 10-15 strokes of a tight Dance pestle, the nuclei were then filtered through a 70 µm nylon mesh strainer, and the nuclei were pelleted at 10000 g for 10 min at 4 °C. For staining, nuclei were resuspended in 500 μL ice-cold blocking buffer (recipe in Table 3) containing 1 / 500 PE-NeuN antibody (MilliporeSigma catalog no. FCMAB317PE) and 1 μg / mL 4'-diaminophenylindazole (DAPI, MilliporeSigma catalog no. D9542), samples were rocked at 4°C for 30 min, then pelleted at 1000g for 5 min at 4°C, and finally samples were resuspended in 500 μL fresh ice-cold blocking buffer before sorting cells on a FacsARIA III.

[0215] Using scattering profiles to remove debris and doublets, bulk samples were sorted at 5000-10000 cells per sample into 200 μL blocking buffer in low-binding Eppendorf 1.5 mL tubes as DAPI+NeuN+ from layers 1-6 or DAPI+NeuN- from layers 1 and 5. Sorted nuclei were pelleted at 1000g for 10 min at 4°C, subsequently resuspended in 50 μL proteinase K cleanup buffer (recipe in Table 3), incubated at 37°C for 30 min, and then frozen at -20°C until library preparation and sequencing.

[0216] For library preparation, tagged DNA was purified with 1.8x volume / volume Ampure XP beads (Beckman-Coulter catalog number A63881), eluted in 11 μL water, and then PCR amplified with Nextera Index Kit primers (number FC-121-1012 from Illumina) using KAPA HiFi HotStart ReadyMix (KAPA Biosystems number KK2602) in a 30 μL reaction (72°3:00, 95°1:00, cycle 17x [98°:20, 65°:15, 72°:15], 72°1:00). PCR products were purified using 1.8x Ampure XP beads, and libraries were quantified using Agilent BioAnalyzer High Sensitivity DNA Chips (catalog number 5067-4626). Sample libraries were then pooled equally and sequenced with paired-end 50 bp reads on either an Illumina MiSeq (Allen Institute) or NextSeq machine (SeqMatic, Fremont CA USA). Fastq files were processed as described below.

[0217] Single-cell ATAC-seq. We modified the single-cell ATAC-seq workflow from the bulk sample workflow in several ways, most notably by performing the transposition reaction after sorting rather than before sorting, and by omitting DAPI except in non-neuronal samples (due to uncertainties that DAPI may interfere with transposition).

[0218] Specific MTG tissue layers were collected and dissected as for bulk samples, but these layers were immediately denuded to release nuclei, which were then stained in blocking buffer containing PE-NeuN antibody rather than DAPI. Single NeuN+ nuclei from each layer were sorted into individual wells of a 96-well plate, using scatter profiles to exclude debris and doublets. Single nuclei-to-event correspondence was confirmed by test sorting single NeuN+ events into flat-bottom 96-well plates with 40 μL blocking buffer containing DAPI, followed by pelleting at 3000 g for 1 min and microscopy. These tests typically yielded >95% single nuclei filled wells and undetectable doublets. When glial cells were sorted, neurons were first sorted from the sample using PE-NeuN+ staining, then glial cells were sorted as DAPI+ NeuN- events after treatment with DAPI (1 μg / μL) for 1–2 min.

[0219] Single NeuN+ cells were sorted into 1.5 μL of Nextera Tn5 transposition reaction (0.6 μL Tn5 enzyme, 0.75 μL tagging buffer, 0.15 μL 1% IGEPAL CA-630) in an Eppendorf semi-skirted 96-well plate (MilliporeSigma catalog no. EP0030129504). Immediately after sorting, the plate was spun down briefly, vortexed briefly, spun down again, and then incubated at 37 °C for 30 min for transposition. After transposition, 0.6 μL of proteinase K cleanup buffer was added, the sample was vortexed briefly, spun down, and incubated at 40 °C for an additional 30 min, and then the plate was frozen until library preparation. Library preparation of single cell samples was the same as that of bulk samples, except that the number of amplification cycles was increased from 17 to 22 cycles due to the lower input DNA content.

[0220] Bulk ATAC-seq sample clustering. Peaks were called for all 39 bulk samples from five independent specimens using MACS2 (Zhang et al., Genome Biology. 9, R137, 2008), then DiffBind (Ross-Innes et al., Nature. 481, 389-393, 2012) to identify 73742 differential peaks for all contrasts between sample types (classification strategies and specimens). Of these, 1524 discriminated experimental samples and were discarded for clustering. With the remaining 72218 peaks that were found to be specific to discriminate pairwise combinations of any sorting strategy, correlations between bulk samples were reanalyzed using the reads of these peaks. The correlation matrix revealed groupings of non-neuronal, upper layer neuronal, and lower layer neuronal samples. One sample (H17.03.009 L1 NeuN+) was omitted from this analysis. This is because this sample appeared intermediate between NeuN+ and NeuN cells, likely resulting from a sorting error.

[0221] ATAC-seq data preprocessing and quality control. Sample-specific Fastq files were obtained using standard built-in Illumina dendexing protocols. Each fastq file was mapped to the human genome reference hg38 patch 7 using bowtie2 and flag--no-mixed--no-discordant-X 2000 to generate sample-specific bam files, which were then filtered for low-quality mapping, secondary mapping, and unmapped reads using samtools view-q 10-F 256-F 4, and then filtered for duplicate reads using samtools rmdup. These filtered read bam files were then converted to BED files using bedTools bamToBed, and quality control calculations of average ENCODE overlap and TSS enrichment scores were performed. The average ENCODE overlap bed file was converted to fragment format so that the percentage of unique fragments overlapping with ENCODE Project DNaseI hypersensitive peaks from adult human frontal cortex (studies ENCSR000EIK and ENCSR000EIY; The ENCODE Project Consortium, Nature. 489, 57-74, 2012; Sloan et al., Nucleic Acids Res. 44, D726-D732, 2016) was assessed using bedTools intersectBed (Quinlan and Hall, Bioinformatics. 26, 841-842, 2010) and the average of these two numbers was taken. For TSS enrichment scores, the published technique of Chen et al. (Chen et al., Nat Meth. 13, 1013-1020, 2016) was used. This technique sums the overlap of reads within a 2 kb window surrounding all human TSSs, segments this 2 kb window into 40 50 bp bins, then normalizes the total read counts to the outer 4 bins (the first and last 2 bins), and finally reports a TSS enrichment score as the maximum height of that normalized read count graph.It was noticed that while this technique performed well for all bulk samples, it gave spuriously abnormally high scores for some single cells with low read counts. As a result, it was modified to set the TSS enrichment score to 1 (no enrichment) for single cells with less than 500 reads or TSSs calculated to be greater than 20 (likely spurious events).

[0222] These quality control metrics were used to filter out low-quality cells (ENCODE overlap <15% and TSS score <4). Additionally, cells with less than 10,000 unique read pairs were filtered out, as these many reads are required for the clustering approach. Of the 3,660 initial cells, the analysis was limited to 2,858 high-quality nuclei for clustering.

[0223] Single-cell clustering: Bootstrap clustering. Single cells were clustered using the extended fragment Jaccard distance calculation between cells as implemented by the Lowcat package (Graybuck et al., bioRxiv, 525014, 2019). To achieve this, we first excluded read data on chromosomes X, Y, and M to prevent differential chromosome-biased clustering. It was then randomly downsampled with fragments extended to a normalized length of 1000 bp with the same center as described in Materials and Methods of Example 1. Jaccard distance was then calculated as described in Materials and Methods of Example 1.

[0224] Finally, this 2858x2858 Jaccard distance matrix was reduced in dimension to a 2858x29 matrix with principal component scores between 2 and 30 using princomp in R. Principal component 1 was omitted as it was highly correlated with quality control metrics, suggesting that this axis primarily reflects the quality of the cell library. Principal components above 30 contain little cell type information, so excluding them represents a noise removal step. The resulting 29 PCs (principal components) are used to call cell clusters and visualize them using tSNE.

[0225] To call cell clusters on this 2858x29 principal component matrix, an iterated Jaccard-Louvain clustering technique was bootstrapped using k=15 nearest neighbors. Each bootstrapping round was repeated 200 times, each time including only 80% (2286) of the cells, and the frequency with which each cell co-clustered with all other cells was tabulated. This co-clustering frequency matrix was then hierarchically clustered by Euclidean distance, and 27 cell type clusters were called by cutting the tree to present visually obvious co-clustered cell blocks. Repeating this process with a stricter variable 50-90% cell inclusion yielded a similar cluster structure with similar cluster membership, whereas randomizing the Jaccard distance matrix prior to principal component analysis and bootstrap clustering did not result in clusters in the dataset. Together, these analyses suggest that the identified clusters represent real and reproducible cell groups.

[0226] Single-cell clustering: comparing feature set choices. We also attempted to cluster cells using feature sets other than Jaccard distance between cells. These feature sets included: 1) a list of all peaks detected from the entire aggregated dataset (236588 peaks with region flags, called using Homer findPeaks (Heinz et al., Molecular Cell. 38, 576-589, 2010)), 2) a list of all RefSeq gene TSS regions, extended + / - 10kb (27021 regions), 3) all 321184 non-overlapping 10kb bins across the entire human genome, and 4) a list of "GeneBins", defined as the genomic region of each gene between the midpoint boundaries between each RefSeq gene transcription region. For each feature set, we calculated the counts in the region of each cell to observe cell groups, then identified principal components and visualized the cell groups by tSNE of principal components 2:50. Jaccard distance revealed the cleanest qualitative separation between cells and cell clusters, and a wide range of tSNE turbidity values ​​maintained these separations.

[0227] Mapping clusters to transcriptomic cell types: assimilation of epigenetic and transcriptional information. The goal was to map 2858 high-quality ATAC-seq profile cells discovered by a large-scale RNA-seq study to human brain cell types (Hodge et al., bioRxiv, 384826, 2018). To do this, we first sought the best technique to produce gene-level information from ATAC-seq data to correlate with RNA-seq transcript counts. Four techniques were tried: 1) read counts in RefSeq gene bins, 2) read counts in RefSeq gene bodies, 3) read count extensions + / - 10kb in RefSeg gene TSS regions, and 4) Cicero gene activity scores (Cusanovich et al., Cell. 174, 1309-1324.e18, 2018; Pliner et al., Molecular Cell. 71, 858-871.e8, 2018). With these four sets of gene-level information calculated for each cell, single cells were mapped to RNA-seq cell types using each epigenetic feature set (using a subset of 831 marker genes) and the RNA-seq cluster gene count median per million (CPM) as the best correlation, resulting in four different mappings for each cell.

[0228] The 831 marker genes were selected to be informative marker genes for RNA-seq clustering and to contain rich epigenetic information. This was achieved by using the select_markers function with default parameters from the scratch.hicat R package (Tasic et al., Nature.563,72,2018). This gave 2791 transcriptome marker genes, which were further filtered by intersecting with the top 10 percent of genes with the highest total Cicero gene activity score for all 2858 cells to give 831 combined transcriptome and epigenetic marker genes for mapping.

[0229] The four sets of cell-type mapping yielded four tables of cell type abundance in the dataset. We then compared the four cell type abundance tables to the "expected" cell type abundance, calculated by multiplying the sum of the number of cells sorted with each sorting strategy by the expected cell type frequency with each sorting strategy, with the RNA-seq dataset as the true gold standard. By correlating the four cell type abundance tables with the expected abundance tables (Pearson correlation of log-transformed abundance values ​​+1), it became clear that the Cicero gene activity score supplies the most reliable gene-level information for epigenetic to transcriptomic mapping purposes.

[0230] Mapping clusters to transcriptomic cell types: Bootstrapping mapping for final mapping calls. Using Cicero gene activity scores, the cell-wise mapping procedure was bootstrapped 100 times while retaining a variable 50-90% of genes from each round, and the most frequently mapped transcriptomic cell types were applied to each single ATAC-seq cell. We then reported the percentage of constituent cells of each cluster that map to each cell type, summed by cell type subclass.

[0231] Cluster-wise mapping was also performed for each of the 27 ATAC-seq clusters using the same bootstrap mapping procedure, except that Cicero gene activity scores were aggregated by average across cells within each cluster before mapping. We report the number of 100 times each cluster mapped to each cell type and summed by transcriptomic subclass in Figure 76.

[0232] Cluster-wise mapping was observed to be largely consistent with, but cleaner than, cell-wise mapping (Figure 76). Therefore, cluster-wise mapping was selected as the final mapping procedure. Thus, each cell is assigned a final mapped transcriptome cell type and cell type subclass (shown in Figure 76) as a result of its ATAC-seq cluster membership.

[0233] Peak calling. Peaks were called in both bulk and aggregated single-cell data using Homer findPeaks with region flags (Heinz et al., Molecular Cell. 38, 576-589, 2010). This program was found to outperform Hotspot, MACS2, and SICER for identifying small regions corresponding to likely enhancers while still capturing peak boundaries. Peak sizes are a median of 400-500 bp across subclasses.

[0234] Identifying transcription factor motifs using ChromVAR. ChromVAR (Schep et al., Nature Methods. 14, 975-978, 2017) was used to identify the accessibility of transcription factor motifs in cells. Homer findPeaks was used to call peaks in the aggregate of all single cells and the bulk library (236588 peaks), which were then resized to a standard 150bp size with the same center. We downloaded 452 transcription factor motifs from JASPAR (using the JASPAR2018 R package; Tan, JASPAR2018: Data Package for JASPAR2018, 2017) and 1764 from cisBP (as included in the R package chromVARmotifs, Schep et al., Nature Methods. 14, 975-978, 2017) and used chromVAR to aggregate and quantify motif accessibility in all 2858 single cells. Cell type subclass variance motifs among subclasses were found by ranking the subclass average motif accessibility by the standard deviation among subclasses (including DLX1 and NEUROD6).

[0235] Global peak characterization by conservation. Once peaks were called for each subclass, peaks were subset into four sets: 1) all peaks (no subsets); 2) subclass-specific peaks detected only in that subclass and not in outgroup subsets of human keratinocytes or mouse E16.5 kidney ATAC-seq data downloaded from ENCODE (The ENCODE Project Consortium, Nature. 489, 57-74, 2012); 3) TSS-distal peaks that were not located <20 kb from any of the 27021 RefSeq gene TSS sites downloaded from the UCSC table browser (Karolchik et al., Nucleic Acids Res. 32:D493-D496, 2004); 4) subclass-specific and TSS-distal peaks. Overlap was calculated using bedtools intersectBed. In analyses that shuffled peak positions, for TSS-distal peaks, randomly generated comparison peak positions were restricted to the same TSS-distal genomic region.

[0236] For peak phyloP scores, bigWigSummary was used to retrieve phyloP values ​​from hg38.phyloP4way.bw or mm10.phyloP4way.bw. These files quantify base pair conservation across four mammals; Homo sapiens, Mus meculus, Galeoptera balegatus (Malayan flying lemur), and Tupaia chinensis (Chinese tree shrew). Ten values ​​distributed across each peak were returned and the maximum average of eight sets of three consecutive values ​​was calculated. This was done to find small regions of highly conserved regions of approximately 100 bp within each peak, which would result in a larger deviation between actual and random phyloP scores than a single peak-wise average. Peak-wise phyloP scores were compared to scores of peak regions randomly distributed across the genome by subtracting the random peak phyloP average from the actual peak phyloP average.

[0237] Transcriptomic cell type identification is consistent with methylation data. Using the dataset from Luo et al. (Science.357,600-604,2017 (Supplementary Table 3, including 1012 human and 1016 mouse methylation marker genes), published mCH gene body marker genes were correlated with the cluster-wise median FPKM of transcriptomic human cell types identified by Hodge et al. (bioRxiv,384826,2018) and mouse cell types by Tasic et al. (Nature.563,72,2018). Pearson correlation coefficients between normalized gene body mCH and RNA-seq cluster-wise median FPKM were calculated, and the best-correlated transcriptomic cell type was assigned to each methylation cell type. The specificity of the match was calculated as the difference between the best and second-best correlations. Importantly, all transcriptomic cell type assignments are consistent with the predicted subclasses by the original authors.

[0238] Quantification of overlap between ATAC-seq peaks and DMRs. First, we aggregated human DMRs from Luo et al. (Science.341,1237905,2013) and Lister et al. (Science.357,600-604,2017). For neuronal types, we downloaded the DMRs as calculated by the authors and then merged these DMRs using bedtools mergeBed. For non-neuronal types, we downloaded raw fastq files from Lister et al.'s GEO submission corresponding to bulk NeuN-negative cells from two human replicates (GSM1173774 and GSM1173777) and converted these to allc files using the pipeline analysis method from Luo et al. (Science.357,600-604,2017). These allc files were aggregated and used to find DMRs with methylpy DMRfind against allc files for all human subclasses from Luo et al. and the outgroup of human H1 cells from ENCODE (The ENCODE Project Consortium, Nature. 489, 57-74, 2012). The same set of bulk non-neuronal DMRs was used as one for comparison with astrocyte, oligodendrocyte / OPC, and microglia ATAC-seq classes (Figure 77).

[0239] Using the bed files corresponding to each subclass ATAC-seq peak set and each subclass DMR set, overlap between peaks and DMRs was quantified using bedtools intersectbed. 100 calculations of actual peak overlap were bootstrapped by removing 20 percent of the peaks each time and calculating the percentage overlap, and the average of these 100 measurements is reported.

[0240] Similarly, we randomized peak locations across the genome using bedtools shuffleBed 100 times, calculated the percentage overlap each time, and report the average of these 100 measurements. By definition, the disarticulation range of overlap percentages between real and randomized peaks established a false discovery rate <0.01. We also calculated the enrichment of DMR overlap for the ATAC-seq peak set, defined as the ratio of real peak-DMR overlap percentage to the overlap percentage of randomized peak locations.

[0241] Mouse to human interspecies comparison. A set of subclass-specific peaks was used to map between human and mouse subclasses, which are uniquely identified in that subclass only. The first subclass-specific mouse peak was mapped to hg38 using liftOver. The human peak overlap calculation was then bootstrapped 100 times against all mouse peaks randomly retaining 80% of the human peaks each time, and the Jaccard similarity coefficient (intersection on union) was averaged over the 100 runs. In addition, the genomic peak positions were shuffled 100 times, and the average Jaccard similarity coefficient was calculated each time. The enrichment of the Jaccard similarity coefficient was determined as the ratio of real to random.

[0242] Characterization of human conserved and divergent peaks began with all human peaks and subsets identified within the same homologous subclass that either intersected ("Conserved") or did not intersect ("Divergent") with mouse peaks mapped to hg38 by liftover. To characterize mouse conserved and divergent peaks, all mouse peaks were intersected with reciprocal mm10-mapped human peaks. For these four sets of peaks, phyloP scores were then calculated as above.

[0243] Cloning enhancers. Enhancers were manually selected from the ATAC-seq and RNA-seq data for cloning by the following criteria: 1) adjacent to known subclass marker genes, 2) peaks that are specifically accessible only in the subclass of interest, and 3) containing regions of high primary sequence conservation by phyloP score.

[0244] The selected enhancers were cloned into AAV expression vectors, derivatives of either pscAAV-MCS (Cell Biolabs catalog no. VPK-430) containing eHGT_019h, eHGT_017h, eHGT_022h, eHGT_022m, and eHGT_023h, or pAAV-GFP (Cell Biolabs catalog no. VPK-410) containing eHGT_078h, eHGT_058h, eHGT_060h, and hDLXI56i (Dimidschstein et al., Nature, Neuroscience. 19, 1743-1749, 2016; Zerucha et al., J. Neurosci. 20, 709-721, 2000). The enhancer was inserted upstream of the minimal beta globin promoter and SYFP2, a brighter EGFP alternative that is well tolerated in neurons, by standard Gibson assembly approaches (Kremers et al., Biochemistry. 45, 6570-6580, 2006). NEB stable cells (New England Biolabs no. C3040I) were used for transformation. scAAV plasmids were monitored by restriction analysis and Sanger sequencing for occasional (10%) recombination of the remaining ITRs.

[0245] Virus production. Enhancer AAV plasmid was maxiprepped and transfected together with helper plasmid and PHP.eB rep / cap packaging vector into one plate of AAV-293 cells (Cell Biolabs catalogue no. AAV-100) in polyethylimine max. The next day the medium was changed to 1% FBS and after 5 days the cells and supernatant were harvested and AAV particles were released by 3 freeze-thaw cycles. After freeze-thawing the lysate was treated with benzonase to degrade free DNA (2 μL benzonase, 37 degrees for 30 min, MilliporeSigma catalogue no. E8263-25KU), then cell debris was pre-removed with a low speed spin (1500g for 10 min) and finally the crude virus was concentrated via a 100 kDa molecular weight cut-off Centricon column (MilliporeSigma catalogue no. Z648043) to a final volume of 150 μL. This crude virus preparation was useful in both mouse and human virus studies.

[0246] Mouse virus testing. Mice were injected retro-orbitally at P42-P49 with 10 μL of crude virus preparation (1E11 genome copies) diluted in 100 μL of PBS and then sacrificed 18-28 days post-infection. For live epifluorescence, mice were perfused with ACSF.7 and live 350 μm physiological sections were cut in compressed sections of one hemisphere to analyze reporter expression. For antibody staining, the other hemisphere was drop-fixed in 4% PFA in PBS for 4-6 h at 4 degrees, then cryoprotected in 30% sucrose in PBS for 48-72 h, then embedded in OCT for 3 h at room temperature, then frozen on dry ice and sectioned at 10 μm thickness, followed by antibody staining using standard practice. Single-cell RNA-seq was accomplished as previously described (Tasic et al., Nat Neurosci. 19, 335-346, 2016; Tasic et al., Nature. 563, 72, 2018).

[0247] Human virus testing. Temporal cortical neurosurgical specimens were bubbled in cold ACSF.7 and kept sterile throughout the entire process. Tissue blocks were sliced ​​at a thickness of 350 μm, then the white matter and pial membrane were dissected out and removed. Typically, all layers are represented in good cortical slices. Slices were then allowed to recover at high temperature (bubbled ACSF.7 for 15 min at 30 degrees) before sodium reintroduction (bubbled ACSF.8 for 30 min at room temperature, recipe in Table 2; Ting et al., Scientific Reports.8, 8407, 2018). Slices were then seeded at the gas interface on Millicell PTFE cell culture inserts (MilliporeSigma no. PICM03050) in 6-well dishes on 1 mL of Slice Culture Medium (recipe in Table 3). After 30 min, slices were infected by applying high titer AAV2 / PHP.eB virus aliquots directly to the surface of the slices, 1 μL per slice. Slice culture medium was replenished every 2 days and reporter expression was monitored.

[0248] Single-cell RNA-seq was achieved on human virus-infected neurons by digestion in carbonized ACSF.1 / trehalose+blocker+papain (recipe in Table 3) for 1 h at 30°C, followed by gentle trituration in Low-BSA Quench buffer, shallow spin gradient centrifugation into High-BSA Quench buffer (100g, 10 min at room temperature), and resuspension in cell resuspension buffer. Myelin debris was also removed using 1 / 20 Myelin Bead Removal Kit II (Miltenyi catalog no. 130-096-733), and contaminating glial cells were sorted out employing 1 / 40 PE-anti-CD9 clone eBioSN4 (Thermo Fisher catalog no. 12-0098-42). Single SYFP2+ labeled human neurons were then sorted for sequencing using SMARTer V4 as previously described (Tasic et al., Nat Neurosci. 19, 335-346, 2016; Tasic et al. Nature. 563, 72, 2018).

[0249] Inference of GWAS cell subclass associations. Linkage disequilibrium score regression (LDSC, Bulik-Sullivan et al. Nature Genetics. 47, 291-295, 2015; Finucane et al. Nat Genet. 47, 1228-1235, 2015) was used to partition the heritability of various brain conditions to regions associated with accessible chromatin in 11 human cortical cell subclasses. The peaks are divided into conserved and divergent subsets. As an outgroup comparator, we also investigated heritability associated with the outgroup population of human keratinocytes downloaded from ENCODE.

[0250] Summary statistics from 21 genome-wide association studies (GWAS) including predicted brain-related disorders (schizophrenia, major depressive disorder, autism spectrum disorder, ADHD, Alzheimer's disease, Tourette's syndrome, bipolar disorder, eating disorders, obsessive-compulsive disorder, loneliness, BMI, PTSD) and predicted non-brain-related disorders (Crohn's disease and asthma) were downloaded from the PGC and EMBL / EBI GWAS repositories (see Table 2). 10 (N * h 2 ) < 3.6, where N is the number of patients in the study and h 2 represents the sum of heritability across all SNPs in the study, the power of the study (Finucane et al., Nat Genet. 47, 1228-1235, 2015). This exclusion was due to the lack of evidence for asthma (Demenais et al., Nat. Genet. 50, 42-53, 2018, log 10 (N*h 2 )=3.5, PTSD(Duncan et al.,Mol.Psychiatry.23,666-673,2018,log 10 (N*h 2 )=2.9), eating disorders (Duncan et al.,Am J Psychiatry.174,850-858,2017,log 10 N*h 2)=3.5), loneliness (Gao et al., Neuropsychopharmacology.42, 811-821, 2017, log 10 (N*h 2 )=3.3), obsessive-compulsive disorder (IOCDF-GC&OCGAS, Mol. Psychiatry. 23, 1181-1188, 2018, log 10 (N*h 2 )=3.5), and one major depressive disorder study (Major Depressive Disorders Working Group of the Psychiatry GWAS Consortium et al. (Mol. Psychiatry. 18, 497-511, 2013, log 10 (N * h 2 )=3.3) were excluded. All 15 included studies were performed on populations of European descent. Within these datasets, the analysis was limited to 1389227 high-confidence SNPs present in the HapMap3 list, and linkage disequilibrium maps from 1000 genomic European descent individuals were used to analyze cell subclass-associated chromatin trait and disease enrichment along with the LDSC baseline model LDv2.0 with 75 enumerated genomic feature categories. For statistical testing to identify significant Bonferroni enrichment, multiple hypothesis testing for correction of LDSC's block jackknife estimated p-value was used as previously suggested (Skene et al. Nature Genetics. 50, 825, 2018). This correction was calculated as 0.05 / 345 disease / subclass combinations=1.45e -4 Significance cutoffs of 180 and 150 tests were used as well. [Table 2] TIFF0007590962000006.tif254164TIFF0007590962000007.tif141166 [Table 3] TIFF0007590962000009.tif254168TIFF0007590962000010.tif254169TIFF0007590962000011.tif254169TIFF0007590962000012.tif157170

[0251] closing paragraph Also included are variants of the sequences disclosed and referenced herein. Guidance can be found using computer programs well known in the art, such as DNASTAR™ (Madison, Wisconsin) software, to determine which amino acid residues can be substituted, inserted, or deleted without abolishing biological activity.Preferably, the amino acid changes in the protein variants disclosed herein are conservative amino acid changes, i.e., substitutions of similarly charged or uncharged amino acids.Conservative amino acid changes involve the substitution of one of a family of amino acids that are related in their side chains.

[0252] Suitable conservative substitutions of amino acids in a peptide or protein are known to those of skill in the art and can generally be made without altering the biological activity of the resulting molecule. Those skilled in the art will generally recognize that single amino acid substitutions in non-essential regions of a polypeptide do not substantially alter biological activity (see, e.g., Watson et al. Molecular Biology of the Gene, 4th Edition, 1987, The Benjamin / Cummings Pub. Co., p. 224). Naturally occurring amino acids are generally classified into conservative substitution families as follows: Group 1: alanine (Ala), glycine (Gly), serine (Ser), and threonine (Thr); Group 2: (acidic): aspartic acid (Asp) and glutamic acid (Glu); Group 3: (acidic; also classified into polar, negatively charged residues and their amides): asparagine (Asn), glutamine (Gln), Asp, and Glu; Group 4: Gln and Asn; Group 5: (basic; also classified into polar, positively charged residues): arginine (Arg), lysine (Lys), and histidine (His); Group 6 (large aliphatic, nonpolar residues): isoleucine (I Group 7 (uncharged polar): tyrosine (Tyr), Gly, Asn, Gln, Cys, Ser, and Thr; Group 8 (large aromatic residues): phenylalanine (Phe), tryptophan (Trp), and Tyr; Group 9 (non-polar): proline (Pro), Ala, Val, Leu, Ile, Phe, Met, and Trp; Group 11 (aliphatic): Gly, Ala, Val, Leu, and Ile; Group 10: (small aliphatic, non-polar or slightly polar residues): Ala, Ser, Thr, Pro, and Gly; and Group 12 (sulfur-containing): Met and Cys. Additional information can be found in Creighton (1984) Proteins, W.H. Freeman and Company.

[0253] In making such changes, the hydrophilicity index of amino acids may be considered. The importance of the hydrophilic amino acid index in conferring interactive biological function on a protein is generally understood in the art (Kyte and Doolittle, 1982, J. Mol. Biol. 157(1), 105-32). Each amino acid is assigned a hydropathic index on the basis of its hydrophobicity and charge characteristics (Kyte and Doolittle, 1982). These values ​​are as follows: Ile (+4.5), Val (+4.2), Leu (+3.8), Phe (+2.8), Cys (+2.5), Met (+1.9), Ala (+1.8), Gly (-0.4), Thr (-0.7), Ser (-0.8), Trp (-0.9), Tyr (-1.3), Pro (-1.6), His (-3.2), Glutamate (-3.5), Gln (-3.5), Aspartate (-3.5), Asn (-3.5), Lys (-3.9), and Arg (-4.5).

[0254] It is known in the art that certain amino acids may be substituted by other amino acids with similar hydrophilicity index or score, and still result in proteins with similar biological activity, i.e. still obtain biologically functional equivalent proteins. When making such changes, substitution of amino acids with hydrophilicity index within ±2 is preferred, within ±1 is particularly preferred, and within ±0.5 is even more particularly preferred. It is also understood in the art that substitution of similar amino acids can be effectively made based on hydrophilicity.

[0255] As described in detail in U.S. Pat. No. 4,554,101, the following hydrophilicity values ​​have been assigned to amino acid residues: Arg (+3.0), Lys (+3.0), Aspartic acid (+3.0±1), Glutamic acid (+3.0±1), Ser (+0.3), Asn (+0.2), Gln (+0.2), Gly (0), Thr (-0.4), Pro (-0.5±1), Ala (-0.5), His (-0.5), Cys (-1.0), Met (-1.3), Val (-1.5), Leu (-1.8), Ile (-1.8), Tyr (-2.3), Phe (-2.5), Trp (-3.4). It is understood that an amino acid can be substituted for another amino acid having a similar hydrophilicity value and still obtain a biologically equivalent, and in particular an immunologically equivalent, protein. In such changes, substitutions of amino acids whose hydrophilicity values ​​are within ±2 are preferred, those within ±1 are particularly preferred, and those within ±0.5 are even more particularly preferred.

[0256] As discussed above, amino acid substitutions may be based on the relative similarity of the amino acid side-chain substituents, for example, their hydrophobicity, hydrophilicity, charge, size, and the like.

[0257] As noted elsewhere, variants of a gene sequence can include codon-optimized variants, sequence polymorphisms, splice variants, and / or variants that do not affect the function of the encoded product to a statistically significant extent.

[0258] Variants of the protein, nucleic acid, and gene sequences disclosed herein also include sequences having at least 70% sequence identity, 80% sequence identity, 85% sequence identity, 90% sequence identity, 95% sequence identity, 96% sequence identity, 97% sequence identity, 98% sequence identity, or 99% sequence identity to the protein, nucleic acid, or gene sequences disclosed herein.

[0259] "Percent sequence identity" refers to the relationship between two or more sequences, as determined by comparing the sequences. In the art, "identity" also means the degree of sequence relatedness between protein, nucleic acid, or gene sequences, as determined by the match between strings of such sequences. "Identity" (often referred to as "similarity") can be readily calculated by known methods, see Computational Molecular Biology (Lesk, A. M. ed.) Oxford University Press, NY (1988); Biocomputing: Informatics and Genome Projects (Smith, D. W. ed.) Academic Press, NY (1994); Computer Analysis of Sequence Data, Part I (Griffin, A. M., and Griffin, H. G. ed.) Humana Press, NJ (1994); Sequence Analysis in Molecular Biology (Von Heijne, G. ed.) Academic Press (1987); and Sequence Analysis Primer (Gribskov, M. and Devereux, J. ed.) Oxford University Press, NY (1992). Preferred methods for determining identity are designed to give the best match between the sequences tested. Methods to determine identity and similarity are codified in publicly available computer programs. Sequence alignment and percent identity calculations may be performed using the Megalign program of the LASERGENE bioinformatics computing suite (DNASTAR, Inc., Madison, Wisconsin). Multiple alignment of sequences can also be performed using the Clustal method of alignment (Higgins and Sharp CABIOS, 5, 151-153 (1989) with default parameters (GAP PENALTY=10, GAP LENGTH PENALTY=10).Related programs also include the GCG suite of programs (Wisconsin Package Version 9.0, Genetics Computer Group (GCG), Madison, Wisconsin); BLASTP, BLASTN, BLASTX (Altschul et al., J. Mol. Biol. 215:403-410 (1990); DNASTAR (DNASTAR, Inc., Madison, Wisconsin), and the FASTA program incorporating the Smith-Waterman algorithm (Pearson, Comput. Methods Genome Res., [Proc. Int. Symp.] (1994), Meeting Date 1992, 111-20. Editor(s): Suhai, Sandor. Publisher: Plenum, New York, 1997. Within the context of this disclosure, when sequence analysis software is used for the analysis, it will be understood that the results of the analysis will be based on the "default values" of the program referenced. As used herein, "default values" refers to any set of values ​​or parameters that are initially loaded into the software when it is first initialized.

[0260] Variants also include nucleic acid molecules that hybridize to sequences disclosed herein under stringent hybridization conditions and provide the same function as the reference sequences. Exemplary stringent hybridization conditions include overnight incubation at 43°C in a solution containing 50% formamide, 5X SSC (750mM NaCl, 75mM trisodium citrate), 50mM sodium phosphate (pH 7.6), 5X Denhard's solution, 10% dextran sulfate, and 20μg / ml denatured sheared salmon sperm DNA, followed by washing the filter in 0.1X SSC at 50°C. Changes in the stringency of hybridization and signal detection are primarily achieved by manipulating formamide concentration (lower percentages of formamide result in lower stringency), salt conditions, or temperature. For example, moderately high stringency conditions include overnight incubation at 37°C in a solution containing 6XSSPE (20XSSPE = 3M NaCl; 0.2M NaH2PO4; 0.02M EDTA, pH 7.4), 0.5% SDS, 30% formamide, 100 μg / ml salmon sperm blocking DNA, followed by washing at 50°C with 1XSSPE, 0.1% SDS. In addition, to achieve even lower stringency, washes performed after stringent hybridization can be performed at higher salt concentrations (e.g., 5X SSC). Modifications in the above conditions can be achieved through the inclusion and / or substitution of alternative blocking reagents used to suppress background in hybridization experiments. Exemplary blocking reagents include Denhardt's reagent, BLOTTO, heparin, denatured salmon sperm DNA, and commercially available proprietary formulations. The inclusion of certain blocking reagents may require modification of the hybridization conditions described above due to compatibility issues.

[0261] The term link is used broadly to describe joining together in a chain or series. It is used to describe the joining of nucleotide or amino acid sequences into a single nucleotide or amino acid sequence, respectively. The term "concatamerize" should be construed to describe "concatenate."

[0262] As will be understood by one of skill in the art, each embodiment disclosed herein can comprise, consist essentially of, or consist of its particular described element, step, ingredient, or component. Thus, the term "include" or "including" should be interpreted as describing "comprise, consist, or consist essentially of." The transitional term "comprise" or "comprises" means to anticipate the inclusion, even in major amounts, of an unspecified element, step, ingredient, or component, without limitation. The transitional phrase "consisting of" excludes any element, step, ingredient, or component that is not specified. The transitional phrase "consisting essentially of" limits the scope of the embodiment to the specified element, step, ingredient, or component, and those that do not have a substantial effect on the embodiment. A substantial effect would cause a statistically significant decrease in selective expression in the target cell population as determined by scRNA-Seq and the selected artificial expression construct / target cell population pairing.

[0263] Unless otherwise indicated, all numbers expressing properties, such as amounts of ingredients, molecular weights, reaction conditions, and the like, used in the specification and claims should be understood in all instances to be modified by the term "about." Accordingly, unless indicated to the contrary, the numerical parameters set forth in the specification and appended claims are approximations that may vary depending upon the desired properties to be obtained by the present invention. At the very least, and not as an attempt to limit the application of the doctrine of equivalents to the scope of the claims, each numerical parameter should be construed in light of the number of reported significant digits and by applying ordinary rounding techniques. Where further clarification is needed, the term "about" has the meaning that would reasonably be assigned by a person of ordinary skill in the art to a stated numerical value or range, i.e., ±20% of the stated value, ±19% of the stated value, ±18% of the stated value, ±17% of the stated value, ±16% of the stated value, ±15% of the stated value, ±14% of the stated value, ±13% of the stated value, ±12% of the stated value, ±11% of the stated value, ±10% of the stated value, ±9% of the stated value, ±8% of the stated value, ±7% of the stated value, ±6% of the stated value, ±5% of the stated value, ±4% of the stated value, ±3% of the stated value, ±2% of the stated value, or ±1% of the stated value.

[0264] Notwithstanding that the numerical ranges and parameters setting forth the broad scope of the invention are approximations, the numerical values ​​set forth in the specific examples are reported as precisely as possible, however, any numerical values ​​inherently contain certain errors necessarily resulting from the standard deviation found in their respective testing measurements.

[0265] As used in the context of describing the present invention (particularly in the context of the claims below), "a," "an," "the," and similar referents should be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context. The recitation of ranges of values ​​herein is merely intended to serve as a shorthand method of individually referring to each separate value falling within the range. Unless otherwise indicated herein, each separate value is incorporated herein as if it were individually recited herein. All methods described herein can be performed in any suitable order, unless otherwise indicated herein or clearly contradicted by context. The use of any and all examples or exemplary language (e.g., "etc.") provided herein is intended merely to better illuminate the invention and does not pose a limitation on the scope of the invention as otherwise claimed. No language in this specification should be construed as indicating any non-claimed element essential to the practice of the invention.

[0266] Grouping of alternative elements or embodiments of the invention disclosed herein is not to be construed as limiting. Each group member may be referenced and claimed individually or in any combination with other members of the group or other elements found herein. It is anticipated that one or more members of a group may be included in, or deleted from, a group for reasons of convenience and / or patentability. When such inclusion or deletion occurs, the specification is deemed to contain the modified group, thereby satisfying the description of all described Markush groups used in the appended claims.

[0267] Certain embodiments of the invention are described herein, including the best mode known to the inventors for carrying out the invention. Of course, variations on these described embodiments will become apparent to those skilled in the art upon reading the foregoing description. The inventors expect those skilled in the art to adopt such variations as necessary, and the inventors intend the invention to be practiced other than as specifically described herein. Accordingly, this invention includes all modifications and equivalents of the subject matter recited in the claims appended hereto as permitted by applicable law. Moreover, any combination of the above-described elements in all possible variations thereof is encompassed by the invention unless otherwise indicated herein or clearly contradicted by context.

[0268] Additionally, throughout this specification, numerous references are made to patents, printed publications, journal articles, and other written texts, the materials of which are referenced herein. Each of the referenced materials is individually incorporated herein by reference in its entirety for the teachings referenced.

[0269] Finally, it is to be understood that the embodiments of the invention disclosed herein are illustrative of the principles of the invention. Other modifications that may be employed are within the scope of the invention. Thus, by way of example, but not of limitation, alternative configurations of the invention may be utilized in accordance with the teachings herein. Thus, the invention is not limited to that precisely as shown and described.

[0270] The details set forth herein are presented by way of example and for the sole purpose of illustrative discussion of preferred embodiments of the present invention, and to provide what is believed to be the most useful and readily understood description of the principles and conceptual aspects of various embodiments of the present invention. In this regard, no attempt is made to show structural details of the present invention, but rather an attempt is required to provide a basic understanding of the present invention, and to make clear to those skilled in the art how some forms of the present invention may be embodied in practice, through the description taken in conjunction with the drawings and / or examples.

[0271] The definitions and explanations used in this disclosure are meant and intended to control in any future interpretations, unless clearly and unambiguously amended in the following examples, or unless the application of the meaning makes any interpretation meaningless or essentially meaningless. If the interpretation of a term makes it meaningless or essentially meaningless, the definition should be taken from a dictionary known to those skilled in the art, such as Webster's Dictionary, 3rd Edition, or Oxford Dictionary of Biochemistry and Molecular Biology (Ed. Anthony Smith, Oxford University Press, Oxford, 2004).

Claims

1. The sequence shown in SEQ ID NO: 177, The sequence shown in SEQ ID NO: 178, A sequence having at least 95% sequence identity to SEQ ID NO: 177, said sequence having the same enhancer activity as the sequence shown in SEQ ID NO: 177, or A sequence having at least 95% sequence identity to SEQ ID NO: 178, and having the same enhancer activity as the sequence shown in SEQ ID NO:

178. A nucleic acid comprising an artificial enhancer sequence comprising:

2. Two copies of the sequence set forth in SEQ ID NO:177, Two copies of the sequence set forth in SEQ ID NO: 178, two copies of a sequence having at least 95% sequence identity to SEQ ID NO: 177, said sequence having the same enhancer activity as the sequence shown in SEQ ID NO: 177; or Two copies of a sequence having at least 95% sequence identity to SEQ ID NO: 178, said sequence having the same enhancer activity as the sequence shown in SEQ ID NO:

178. The nucleic acid of claim 1 ,

3. Three copies of the sequence set forth in SEQ ID NO: 177, 3 copies of the sequence set forth in SEQ ID NO: 178; Three copies of a sequence having at least 95% sequence identity to SEQ ID NO: 177, said sequence having the same enhancer activity as the sequence shown in SEQ ID NO: 177; or The nucleic acid of claim 1, having three copies of a sequence having at least 95% sequence identity to SEQ ID NO: 178 and having the same enhancer activity as the sequence shown in SEQ ID NO:

178.

4. The sequence shown in SEQ ID NO: 40, The sequence shown in SEQ ID NO: 49, A sequence having at least 95% sequence identity to SEQ ID NO: 40, which has the same enhancer activity as the sequence shown in SEQ ID NO: 40, or A sequence having at least 95% sequence identity to SEQ ID NO:49, and having the same enhancer activity as the sequence shown in SEQ ID NO:

49. The nucleic acid of claim 1 ,

5. (i) the nucleic acid according to claim 1 or The sequence shown in SEQ ID NO: 39, The sequence shown in SEQ ID NO: 48, A sequence having at least 95% identity to the sequence shown in SEQ ID NO: 39, and having the same enhancer activity as the sequence shown in SEQ ID NO: 39, or an enhancer having a sequence having at least 95% identity to the sequence shown in SEQ ID NO: 48, said sequence having the same enhancer activity as the sequence shown in SEQ ID NO: 48; (ii) a promoter, and (iii) heterologous coding sequences 1. An artificial expression construct comprising:

6. The artificial enhancer sequence is One copy of the sequence set forth in SEQ ID NO: 177, or one copy of a sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 177, which has the same enhancer activity as the sequence set forth in SEQ ID NO: 177; Two copies of the sequence set forth in SEQ ID NO: 177, or two copies of a sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 177 and having the same enhancer activity as the sequence set forth in SEQ ID NO: 177; Three copies of the sequence set forth in SEQ ID NO: 177, or three copies of a sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 177, which has the same enhancer activity as the sequence set forth in SEQ ID NO: 177; One copy of the sequence set forth in SEQ ID NO: 178, or one copy of a sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 178, which has the same enhancer activity as the sequence set forth in SEQ ID NO: 178; Two copies of the sequence set forth in SEQ ID NO: 178, or two copies of a sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 178, which has the same enhancer activity as the sequence set forth in SEQ ID NO: 178, or Three copies of the sequence set forth in SEQ ID NO: 178, or three copies of a sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 178, and having the same enhancer activity as the sequence set forth in SEQ ID NO:

178.

6. The artificial expression construct of claim 5, comprising:

7. The artificial enhancer sequence is The sequence set forth in SEQ ID NO: 177, The sequence set forth in SEQ ID NO: 178, The sequence set forth in SEQ ID NO: 40, The sequence set forth in SEQ ID NO:49, A sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 177, and having the same enhancer activity as the sequence set forth in SEQ ID NO: 177; A sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 178, and having the same enhancer activity as the sequence set forth in SEQ ID NO: 178; A sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO: 40, and having the same enhancer activity as the sequence set forth in SEQ ID NO: 40; or A sequence having at least 95% sequence identity to the sequence set forth in SEQ ID NO:49, and having the same enhancer activity as the sequence set forth in SEQ ID NO:

49.

6. The artificial expression construct of claim 5, comprising:

8. 6. The artificial expression construct of claim 5, which is associated with a capsid comprising AAV-PHP.eB, AAV-BR1, AAV-PHP.S, AAV-PHP.B, or AAV-PPS.

9. 6. The artificial expression construct of claim 5, which is in a viral vector.

10. 10. The artificial expression construct of claim 9, wherein the viral vector comprises a recombinant adeno-associated viral (AAV) vector.

11. 6. The artificial expression construct of claim 5, comprising or encoding a skip element comprising T2A, P2A, E2A, F2A, or an internal ribosome entry site (IRES).

12. A transgenic cell comprising the artificial expression construct of claim 5.

13. The transgenic cell of claim 12, wherein the transgenic cell is a cortical excitatory neuron.

14. 14. The transgenic cell of claim 13, which is a layer (L) 2, L3, L4, L5, or L6 cortical excitatory neuron.

15. 14. The transgenic cell of claim 13, which is an L4 intratelencephalic (IT) cortical excitatory neuron, an L5 pyramidal tract (PT) cortical excitatory neuron, an L5 extratelencephalic (ET) cortical excitatory neuron, an L5 IT cortical excitatory neuron, an L5 proximal projection (NP) cortical excitatory neuron, an L6 IT cortical excitatory neuron, an L6 corticothalamic (CT) cortical excitatory neuron, or a Cajal-Retzius (CR) cortical excitatory neuron.

16. 13. The transgenic cell of claim 12, derived from a subcortical population in the central amygdala, pars external capsule (CEAc), substantia nigra, pars compacta, subiculum, or presubiculum (ProS).

17. The transgenic cell of claim 12, which is a CA1 pyramidal neuron, a dentate gyrus granule cell, a fiber neuron, or a cerebellar Purkinje cell.