Method for RNA sequencing of PFA-fixed or FFPE tissue using fixative exchange

Abstract

To address low reverse transcription (RT) yield of single-nucleus RNA sequencing (snRNA-seq) of formalin-fixed, paraffin-embedded samples generated in clinical processes, the present inventors developed Fixative-eXchange (FX)-seq, a novel snRNA-seq method for samples strongly fixed with paraformaldehyde or formalin and / or for FFPE samples. FX-seq readily removes PFA crosslinks using an organocatalyst to increase reverse transcription yield, and prevents leakage through additional Pt (II)-based crosslinking specific to non-Watson-Crick positions of RNA molecules. FX-seq is also applicable to PFA perfusion-based animal research, large-scale human cohort studies, and personalized drug treatment through precision medicine.

Description

RNA sequencing method of PFA-fixed or FFPE tissues using fixative exchange

[0001] The present invention relates to a technique for treating formalin-fixed, formalin-fixed paraffin-embedded tissue (FFPE), or H&E-stained tissue so as to enable nucleic acid analysis.

[0002] In clinical settings, formalin-fixed and paraffin-embedded (FFPE) tissues are produced as a top priority for pathological diagnosis and are used as a primary resource for long-term storage. Consequently, the number of such FFPE tissues worldwide is estimated to exceed billions. If these can be analyzed at single-cell resolution, it presents a significant opportunity to obtain rich data on human single-cell transcriptomes and is estimated to open up possibilities for better understanding and treating human diseases.

[0003] However, when performing single-cell or single-nuclear RNA sequencing on heavily fixed tissues such as FFPE tissues, difficulties in cDNA synthesis exist due to low reverse transcription efficiency, even when using FFPE samples with good RNA quality. For this reason, there is a high unmet need for high-quality, scalable single-nuclear RNA sequencing (snRNA-seq) methods targeting tissues and FFPE samples heavily fixed with paraformaldehyde (PFA).

[0004] The snPATHO-seq, snFFPE-seq, and snRandom-seq technologies presented in recent studies have demonstrated the feasibility of snRNA-seq in FFPE tissue samples. However, snPATHO-seq utilizes a pre-designed probe panel, which carries the potential for bias in transcriptome information. snRandom-seq suffers from heterogeneity compared to existing 3'-terminal transcriptome analyses due to the use of random primers, and presents challenges in accurate measurement as multiple cDNA strands are synthesized from a single RNA strand. Furthermore, snFFPE-seq has been reported to have limited scalability due to the low quality of transcripts extracted from human FFPE samples. In addition, existing technologies rely on high-temperature heat treatment or tissue degradation processes using proteases to improve reverse transcription efficiency.

[0005] These approaches induce RNA degradation due to heating or result in incomplete removal of cross-links due to limited catalytic activity. There is significant room for improvement in these issues by better understanding the principles of reverse transcription inhibition by PFA immobilization at the molecular level and optimizing the process. This is expected to further enhance the quality and scalability of FFPE snRNA-seq data.

[0006] To overcome these limitations, the inventors began their research by exploring the molecular mechanisms affecting reverse transcription efficiency in order to improve cDNA synthesis from RNA in samples strongly immobilized with PFA or formalin. Previous studies have shown that PFA reacts with primary amines of nucleic acids to form cross-links or generate methylol adducts. Modification of nucleobases due to these covalent bonds interferes with the recognition of Watson-Crick base pairs, thereby reducing the efficiency of reverse transcription using mRNA molecules.

[0007] In particular, severely modified poly(A) tails reduced the binding efficiency of oligo(dT) reverse transcriptase primers and interfered with the progression of reverse transcriptase, increasing the likelihood of enzyme cessation or premature termination of cDNA synthesis. Therefore, the removal of cross-linking and methylol adducts using PFA or formalin plays a key role in improving the quality of single-cell or snRNA sequencing data from fixed samples. Furthermore, it is crucial to minimize RNA degradation caused by harsh reaction conditions, including the high temperatures required for this de-crosslinking process.

[0008] To this end, this study developed a fixative exchange (FX-seq) capable of achieving high-quality single-nuclear transcriptome analysis by integrating a chemical approach to fixed tissues.

[0009] To achieve the above objective, one aspect of the present invention provides a cell or tissue pretreatment method comprising the step of exchanging a fixative.

[0010] Another aspect of the present invention provides a method for analyzing nucleic acids in a cell or tissue, comprising the step of detecting nucleic acids in the pretreated cell or tissue.

[0011] Another aspect of the present invention provides a pretreatment kit for formalin-fixed, formalin-fixed paraffin-embedded (FFPE), or H&E-stained cell or tissue samples comprising an RNase inhibitor, an organic catalyst, and a cross-linker.

[0012] Another aspect of the present invention provides a novel crosslinker having the structural formula of the following chemical formula II:

[0013] [Chemical Formula II]

[0014] .

[0015] Another aspect of the present invention provides a composition for fixing cells or tissues comprising PVSA and formaldehyde.

[0016] Another aspect of the present invention provides a composition for fixative exchange comprising 2-aminophenylboronic acid and PVSA.

[0017] Another aspect of the present invention provides a composition for fixative exchange comprising an organic catalyst and PVSA.

[0018] Another aspect of the present invention provides a buffer composition comprising PVSA.

[0019] Fixative exchange sequencing according to the present invention (hereinafter referred to as FX-seq) improves the preservation of RNA within cells or tissues and can efficiently reverse and remove cross-linking or methylol adducts resulting from PFA reactions. Furthermore, it can suppress severe RNA degradation caused by RNase contamination, thereby facilitating the analysis of nucleic acids in cells and / or tissues fixed in PFA or formalin. FX-seq has demonstrated applicability to various biological and clinical samples. Additionally, it enables transcriptome analysis via snRNA-seq or scRNA-seq in PFA or FFPE-preserved tissues, facilitating the storage and transport of clinical samples and enabling the construction of cell atlases through snRNA-seq or scRNA-seq analysis of stored human FFPE samples containing clinical information. In particular, it is possible to analyze the principles of human diseases and pathological markers using FFPE clinical tissues that have been produced and stored for conventional clinical pathology diagnosis. Moreover, based on this, it can provide new diagnostic criteria for precision medicine.

[0020] Furthermore, by integrating with experimental tools for analyzing biomolecules that can be used as drug targets, new opportunities for drug treatment planning can be opened by investigating features found in various disease samples. Additionally, fixing animal tissues by perfusing them with PFA allows for the long-term preservation of transcripts, significantly improving animal studies through snRNA-seq or scRNA-seq analysis. This minimizes the effort required to sample fresh or frozen tissues. Moreover, by minimizing unwanted RNA degradation and stress-induced gene expression during the process of isolating single cells from tissues, the stability of transcriptomic information is enhanced, thereby maximizing the reproducibility of the study.

[0021] Cells and tissues containing large amounts of endogenous RNase make it difficult to obtain accurate transcriptomic information due to the loss of transcriptomic data in conventional microfluidic-based analyses. However, FX-seq, which utilizes PFA fixation and PVSA, is expected to enable accurate analysis by effectively inactivating endogenous RNase. Finally, pathogenic cells can also be analyzed via FX-seq after ensuring high safety through the inactivation process via PFA fixation.

[0022] Figures 1a to 1t show the concept of FX-seq and the results of verifying it.

[0023] Figure 1a shows a schematic diagram of the FX-seq strategy. Strong PFA immobilization inhibited subsequent enzymatic reactions by generating methylol adducts at the primary amines located at the Watson-Crick base pair sites. A crosslinking agent to replace this supported RNA molecules within the cell through selective reactions to atoms not involved in Watson and Crick base pair recognition, and an organic catalyst facilitated the removal of methylol adducts under mild conditions.

[0024] Figures 1b and 1c show the validation results of the FX components in an in vitro model. PFA immobilization immediately inhibited cDNA synthesis of RNA molecules (Figure 1b). Cat.2 demonstrated better catalytic activity than the previously reported Cat.1 in in vitro transcribed RNA and total RNA extracted from HeLa cells (Figure 1c). Data are expressed as mean ± SD. P-values: NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001.

[0025] Figure 1d confirms the additional crosslinking effect by the regioselective crosslinking agent. Data are expressed as mean ± SD. P-values: NS ≥ 0.05, *P≤0.05 **P≤0.01 ***P≤0.001.

[0026] Figure 1e shows the results of measuring the amount of leaked RNA per 1,000 cells using a Qubit fluorescence analyzer. Data are expressed as mean ± SD. P-values: NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001.

[0027] Figure 1f shows PVSA as an effective RNase inhibitor against FX-seq.

[0028] Figure 1g shows the results of automated electrophoresis analysis when various RNase inhibitors were used during the fixative exchange process in mouse brain tissue fixed with PFA. Data are expressed as normalized fluorescence units (FU) per unit of size (bp). "Lower" indicates a lower molecular weight standard.

[0029] Figure 1h illustrates the experimental sequence for the optimized nuclear separation strategy and FX-seq.

[0030] Figures 1i to 1l show the results of performing snRNA-seq on three PFA-heavily fixed Litermate mouse brains repeatedly under four conditions (nuclei isolated without treatment (No treat ctrl), heat treatment without catalyst ((-) Cat.), heat treatment with catalyst (Cat.2), and the complete FX-seq procedure with additional crosslinking and catalyst). The individual components of FX-seq led to an increase in the number of detected genes (Figure 1j), UMI (Figure 1k), and gene diversity (Figure 1l).

[0031] Figures 1m to 1t show the results of UMAP clustering and cell type annotation for PFA-fixed control samples (n = 7497) and FX-seq (n = 8654) after overlapping nucleation. Individual UMAP visualization and annotation (Figures 1m and 1n) and UMAP analysis of the merged gene expression matrix (Figures 1o and 1p) improved the analytical resolution of cell type classification after FX-seq. Cell distribution (Figures 1q and 1r) and marker gene specificity (Figures 1s and 1t) were visualized.

[0032] Abbreviations: CGN, cerebellar granulosa; CB Int, cerebellar neuron; OB Int, olfactory progenitor neuron; OB neuroblast, olfactory progenitor neuroblast; Pallial Glut, cortical glutamatergic neuron; Astro, astrocyte; Oligo, oligoglia; OPC, oligoglia precursor; EC, endothelial cell; VLMC, vascular leptomeningeal cell; Purkinje, Purkinje cell; MSN, intermediate spinous neuron; Int neuron, interneuron; Bergmann, Bergmann glial cell; OEC, olfactory sheath cell; Ependyna, mammary gland cell; Chor, choroidal plexus epithelial cell; VLMC(OB), olfactory progenitor vascular leptomeningeal cell; VLMC(Pia), vascular leptomeningeal cell of the pia membrane; VSMC, vascular smooth muscle cell.

[0033] Figures 2a and 2b show the molecular structures of the organic catalysts used in FX-seq. Specifically, the molecular structure of Cat.1 is shown in Figure 2a, and the molecular structure of Cat.2 is shown in Figure 2b.

[0034] Figure 3 shows the results confirming the treatment effect of the organic catalyst and the effect according to the type of catalyst. Each condition was evaluated using qPCR Ct values ​​of human GAPDH, CDKN1A, PRKACA, KAT5, UPF, and RXRB genes. Data are expressed as mean ± SD. P-values: NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001.

[0035] Figure 4 confirms the treatment effect of the organic catalyst and the amount of gene detected according to the catalyst concentration conditions. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh, Slc17a7, and Gad1 genes.

[0036] Figure 5 confirms the effects of the organic catalyst treatment and the effects according to the treatment temperature conditions. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh, Slc17a7, and Gad1 genes.

[0037] Figure 6 confirms the treatment effect of the organic catalyst and the results according to pH treatment conditions. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh and Slc17a7 genes.

[0038] Figure 7 confirms the treatment effect of the organic catalyst and the performance of the PEG additive. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh and Slc17a7 genes.

[0039] Figures 8a to 8c confirm the effects of organic catalyst treatment on mouse brains fixed with PFA, depending on the presence or absence of treatment and the duration. Each condition was evaluated using qPCR Ct values ​​for mouse Gapdh and Slc17a7 genes (Figure 8a). A larger amount of the gene was detected in the experimental group with added organic catalyst compared to the experimental group without it, and the detected UMI also showed the same pattern when sequencing cell nuclei under the same conditions (Figure 8b). In the histograms of the detected UMI and the corresponding cells (Figure 8c), it was also found that more genes were detected when there was organic catalyst treatment and when the organic catalyst treatment time was longer, at 30 minutes.

[0040] Figure 9 confirms the effect of the PEG-8000 additive during reverse transcription. The qPCR Ct values ​​of the mouse Gapdh gene were normalized based on the condition without PEG addition and compared with the Log2 Fold Change values.

[0041] Figures 10a to 10c show the results of verifying the effect of additional crosslinking agents on cultured cells confirmed by Barnyard sequencing experiments. No significant species mixing was observed in either the control group without additional crosslinking or the experimental group with additional crosslinking (Figure 10a). However, as a result of using additional crosslinking agents, significantly higher UMIs (Figure 10b) and gene numbers (Figure 10c) were detected in both human cells and mouse cells.

[0042] Figures 11a and 11b show the qPCR results to confirm the fixation effects of additional crosslinkers of different lengths. Crosslinkers with different intermediate PEG linker lengths were shown in mouse brains fixed with PFA (Figure 11a). More genes were detected in the additional crosslinker with a longer PEG chain length of 11 compared to tissues with 3 PEG chains or those without additional fixation, and it was confirmed that treatment at a concentration of 0.1X was more effective than treatment at a concentration of 0.01X (Figure 11b).

[0043] Figure 12 shows the difference in the amount of gene detected according to the treatment conditions of the additional crosslinking agent by time and temperature.

[0044] Figures 13a and 13b confirm the additional crosslinking agent effect according to the organic catalyst treatment time.

[0045] Figures 14a to 14d illustrate the synthesis and regioselectivity of the FX-seq crosslinker. Figure 14a illustrates the synthesis process of the crosslinker. Figures 14b and 14c illustrate the synthesized regioselective crosslinker 1 Figure 14 shows the H NMR spectrum (Fig. 14b) and its peak assignment (Fig. 14c). Fig. 14d shows the Guanine-N7 reaction of the regioselective crosslinker.

[0046] Figures 15a to 15j confirm that the Guanine-N7 regioselective crosslinker minimizes the inhibition of reverse transcription (RT) in an in vitro environment.

[0047] Figures 15a through 15g show automated electrophoresis data of immobilized RNA in an in vitro environment. Data are expressed in fluorescence units (FU) normalized per unit of size (bp). "Lower" indicates a lower molecular weight standard. Experimental conditions: Figure 15a is a non-immobilized RNA control, Figure 15b is 1% PFA, Figure 15c is 4% PFA, Figure 15d is 8% PFA, Figure 15e is 0.5X regioselective crosslinker (CL), Figure 15f is 2X CL, and Figure 15g is 4X CL.

[0048] Figures 15h through 15j show bar graphs illustrating the relative reversal inhibition by reaction with PFA or CL at the corresponding concentrations. Primers targeted the nucleotide sequences at the 3' end (Fig. 15h), intermediate region (Fig. 15i), and 5' end (Fig. 15j). Data are expressed as mean ± SD. P-values: NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001.

[0049] Figures 16a to 16g confirm the efficacy of the FX-seq single nuclear separation strategy in various rat organs.

[0050] Figure 16a represents the heart, Figure 16b the lung, Figure 16c the liver, Figure 16d the spleen, Figure 16e the pancreas, Figure 16f the kidney, and Figure 16g the brain tissue; these are Brightfield images of single nuclei isolated via FX treatment from mouse organs heavily fixed with perfusion and PFA. Black and white: phase contrast, blue: DAPI.

[0051] Figures 17a to 17d illustrate FX-seq pretreatment strategies for various samples. Figures 17a to 17d illustrate schematic diagrams of nucleation strategies for PFA tissue (Figure 17a), FFPE block (Figure 17b), FFPE section (Figure 17c), and H&E stained section (Figure 17d).

[0052] Figures 18a to 18c evaluate the optimization of deparaffinization and rehydration of stored human FFPE block sections. Three reactions and resuspension with xylene resulted in superior gene detection sensitivity compared to three reactions and resuspension with limonene (Figure 18a). Two reactions and resuspension with xylene yielded better results than three reactions and resuspension (Figure 18b). In conclusion, it was confirmed that two reactions and resuspension with xylene retained the largest amount of gene not only for qPCR detection but also when measuring the concentration after library preparation for single-nuclear sequencing (Figure 18c).

[0053] Figures 19a and 19b show the results of sequencing (FX-seq) after crosslinking agent exchange confirmed in FFPE blocks and FFPE sections of a mouse brain.

[0054] Figures 20a and 20b confirm that PVSA acts as an effective and heat-stable RNase inhibitor with enzymatic compatibility in both a HeLa cultured cell model and mouse brain tissue heavily fixed with PFA. Figure 20a shows the results of automated electrophoresis of total RNA directly extracted from highly fixed mouse brain tissue without heat treatment. "Lower" indicates a lower molecular weight standard. No inhibitor (orange), 1% DEPC (red), 0.1% v / v SUPERase-In (blue), and 3% PVSA (green) were used as RNase inhibitors. Figure 20b confirms the comparison of RNase inhibitors in HeLa cell lines by quantified cDNA synthesis yields: B2M (left), GAPDH (center), and CNOT (right). Data are expressed as mean ± SD. P-values: NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001.

[0055] Figures 21a through 21d show the results of in situ validation of the FX-seq catalyst in PFA-fixed mouse brain tissue. Genes commonly expressed in the mouse brain were targeted: Gapdh (left), Slc17a7 (center), and Gad1 (right).

[0056] Figure 21a shows the results of a comparison of the reactivity of Cat.1 and Cat.2, in which Cat.2 showed more efficient reactivity than Cat.1 even in an in situ environment.

[0057] Figure 21b shows the results of comparing heat treatment without a catalyst with heat treatment with Cat.2. Heat treatment in high-concentration Tris buffer under FX conditions increased the cDNA synthesis efficiency in mouse brain tissue fixed with PFA, but the addition of Cat.2 further increased the reverse transcription efficiency.

[0058] Figure 21c shows the time-dependent treatment effect of Cat.2. When treated with Cat.2 for a long period, the cDNA synthesis efficiency is observed to stagnate, which is likely due to leakage of RNA molecules, as previously confirmed in the HeLa cell line model.

[0059] Figure 21d shows the stepwise evaluation of the FX components. Heat treatment without catalyst ((-) Cat. 2), heat treatment with Cat. 2 added (Cat. 2), and heat treatment with Cat. 2 added after additional crosslinking (FX-seq) were evaluated under high-concentration Tris buffer conditions of FX. Data are expressed as mean ± SD. P-value: NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001.

[0060] Figures 22a to 22c confirm the auxiliary effect of PVSA in crosslinker exchange sequencing (FX-seq). Each condition was evaluated through qPCR Ct values ​​of mouse Gapdh, Slc17a7, and Gad1 genes.

[0061] Figures 23a to 23l show individual results obtained from triple experiments of FX-seq using 4% PFA-fixed mouse brains.

[0062] Figures 23a to 23d show individual data of QC indicators obtained from four conditions of snRNA-seq: nuclei isolated without treatment (No Treat Ctrl), heat treatment without catalyst under high concentration Tris buffer conditions of FX ((-) Cat. 2), heat treatment with high concentration Tris buffer and catalyst (Cat. 2), and a complete FX-seq procedure with additional crosslinking and high concentration Tris buffer and catalyst (FX-seq). Figures 23a, UMI per nucleus (Figure 23b), detection rate of mitochondrial genes (Figure 23c), and detection rate of rRNA (Figure 23d) are shown.

[0063] Figures 23e and 23f show the number of labeled nuclei (Figure 23e) and labeled cell counts in annotated clusters in each brain of individual replicates in the UMAP analysis for the untreated control group (Figure 23f).

[0064] Figures 23g and 23h show the number of labeled nuclei in each brain (Figure 23g) and the number of labeled cells in annotated clusters (Figure 23h) in the UMAP analysis for FX-seq.

[0065] Figures 23i and 23j show the distribution of UMI according to cell type identified in the untreated control group (Figure 23i) and FX-seq (Figure 23j).

[0066] Figures 23k and 23l show dot plots of cell type-specific marker genes identified in the untreated control group (Figure 23k) and FX-seq (Figure 23l).

[0067] Abbreviations: CGN, cerebellar granulosa; CB Int, cerebellar neuron; OB Int, olfactory progenitor neuron; OB neuroblast, olfactory progenitor neuroblast; Pallial Glut, cortical glutamatergic neuron; Astro, astrocyte; Oligo, oligoglia; OPC, oligoglia precursor; EC, endothelial cell; VLMC, vascular leptomeningeal cell; Purkinje, Purkinje cell; MSN, intermediate spinous neuron; Int neuron, interneuron; Bergmann, Bergmann glial cell; OEC, olfactory sheath cell; Ependyna, mammary gland cell; Chor, choroidal plexus epithelial cell; VLMC(OB), olfactory progenitor vascular leptomeningeal cell; VLMC(Pia), vascular leptomeningeal cell of the pia membrane; VSMC, vascular smooth muscle cell.

[0068] Figures 24a to 24k show the results obtained by performing FX-seq using strongly fixed mouse brain tissue after perfusion and the results obtained by performing snRNA-seq using freshly extracted and lightly fixed mouse brain.

[0069] Figures 24a through 24e show the snRNA-seq QC metrics of lightly fixed mouse brain cell nuclei (n = 11,807) extracted fresh and strongly fixed after PFA perfusion and mouse brain cell nuclei (n = 7,569). They show the number of detected genes (Figure 24a) and UMI (Figure 24b). Both were higher in the freshly extracted samples. Figure 24e shows the proportion of sequencing reads mapped to mitochondrial genes (Figure 24c), rRNA (Figure 24d), and exon regions (Figure 24e). The high proportion of mitochondrial genes, rRNA, and exon regions indirectly indicates contamination with cytoplasmic contents because the tissue degradation process using proteolytic enzymes was not included during the preparation of the fresh samples.

[0070] Figures 24f to 24i show the results of UMAP analysis for each sample after the QC cutoff (UMI 200–40,000 and mitochondrial readings less than 1%) (Figures 24f and 24g). Individual clusters were annotated according to molecular criteria. In the merged clusters, clusters with the same annotation in the individual clustering were co-clustered, but slight differences in distribution were observed in UMAP for each experimental group (Figures 24h and 24i).

[0071] Figure 24j shows a color map of the proportion mapped to exon regions in the UMAP analysis of the merged gene expression matrix. The change in distribution between fresh samples and FX samples in the merged matrix coincided with the distribution axis of the exon readout ratio.

[0072] Figure 24k shows the distribution of exon readout ratios of the top 10 genes identified through differential gene analysis between fresh and fixed samples of the same cell type.

[0073] Abbreviations: OB Int, olfactory precursor neuron; Pallial Glut, cortical glutamatergic neuron; Int neuron, interneuron; DG neuron, diencephalon neuron; MSN, intermediate spinous neuron; Astro, astrocyte; Oligo, oligoglia; OPC, oligoglia precursor; Ependyma, thalamic epithelial cell; Chor, choroidal plexus epithelial cell; EC, endothelial cell; VLMC, vascular leptomeningeal cell; VSMC, vascular smooth muscle cell; HIPP Pyr, hippocampal pyramidal neuron; TE Int, limbic system neuron; Cholinergic, habenula cholinergic neuron; Dopaminergic, dopaminergic neuron; Dien Glut, diencephalon glutamatergic neuron.

[0074] Figures 25a and 25b show microscopic images of freshly extracted mouse brain cell nuclei and cell nuclei extracted by FX-seq from mouse brains that were heavily fixed after perfusion. Specifically, they show a microscopic image of nuclei isolated from fresh mouse brain tissue (Figure 25a) and mouse brain tissue heavily fixed with PFA after perfusion (Figure 25b). Grayscale: Phase contrast, Blue: DAPI.

[0075] Figures 26a to 26l show the validation of FX-seq components in an FFPE mouse brain tissue model.

[0076] Figure 26a shows the brains of three littermates extracted immediately after euthanasia, fixed with PFA, and embedded in a paraffin block. After processing and separation using FX-seq in the FFPE block, snRNA-seq was performed.

[0077] Figures 26b to 26d show the QC metrics for FX-seq of FFPE mouse brain blocks. It was confirmed that applying individual components of FX-seq led to an increase in the number of detected genes (Figure 26b), UMI (Figure 26c), and gene diversity (Figure 26d).

[0078] Figures 26e to 26l show the results of UMAP clustering and cell type annotation for FFPE control samples (n = 4,914) and FX-seq (n = 5,852) after nucleation. Individual UMAP visualization and annotation (Figures 26e and 26f) and UMAP analysis of the merged gene expression matrix (Figures 26g and 26h) improved the analytical resolution of cell type classification after FX-seq. The distribution of each cell type (Figures 26i and 26j) and the specificity of marker genes (Figures 26k and 26l) are visualized.

[0079] Abbreviations: CGN, cerebellar granulosaneuron; CB Int, cerebellar interneuron; OB Int, olfactory progenitor interneuron; OB Neuron, olfactory progenitor neuron; Astro, astrocyte; Oligo, oligoglia; OPC, olfactory progenitor; EC, endothelial cell; VLMC, vascular leptomeningeal cell; Purkinje, Purkinje cell; Pallial Glut, cortical glutamatergic neuron; Int Neuron, interneuron; Bergmann, Bergmann glial cell; OEC, olfactory sheath cell; Ependyma, olfactory epithelial cell.

[0080] Figures 27a to 27l show individual data obtained from three repeated FX-seq experiments using an FFPE mouse brain block.

[0081] Figures 27a through 27d show individual data of QC indicators obtained from snRNA-seq under four conditions: nuclei isolated without fixative exchange (No Treat Ctrl), uncatalyzed heating under high concentration Tris buffer conditions of FX ((-) Cat.2), heating with Cat.2 added (Cat.2), and heating with Cat.2 added after additional crosslinking treatment (FX-seq). The number of detected genes (Figure 27a), the number of UMIs per nucleus (Figure 27b), the ratio of mapped reads for mitochondrial genes (Figure 27c), and the ratio of mapped reads for rRNA (Figure 27d) are shown in their respective figures. All metrics yielded similar values ​​in the biological replicates.

[0082] Figures 27e and 27f show the labeled nuclei of individual copies (Figure 27e) and the distribution of the number of labeled cells in annotated clusters (Figure 27f) in the UMAP analysis for the untreated control group.

[0083] Figures 27g and 27h show the labeled nuclei (Figure 27g) and the distribution of the number of labeled cells in annotated clusters (Figure 27h) in each brain of individual replicates in the UMAP analysis for FX-seq.

[0084] Figures 27i and 27j show the distribution of UMI according to cell type identified in the untreated control group (Figure 27i) and FX-seq (Figure 27j).

[0085] Figures 27k and 27l show dot plots of cell type-specific marker genes identified in the untreated control group (Figure 27k) and FX-seq (Figure 27l).

[0086] Abbreviations: CGN, cerebellar granulosa neuron; CB Int, cerebellar interneuron; OB Int, olfactory progenitor interneuron; OB neuron, olfactory progenitor neuron; Astro, astrocyte; Oligo, oligoglia; OPC, olfactory progenitor; EC, endothelial cell; VLMC, vascular leptomeningeal cell; Purkinje, Purkinje cell; Pallial Glut, palliative glutamate neuron; Int neuron, interneuron; Bergmann, Bergmann glial cell; OEC, olfactory sheath cell; Ependyma, olfactory epithelial cell.

[0087] Figures 28a to 28u show the results of FX-seq treatment on thin FFPE sections and H&E stained sections.

[0088] Figure 28a shows the analysis of 10 μm FFPE and H&E sections of a mouse brain FFPE block processed and separated by FX-seq.

[0089] Figures 28b to 28e show the QC metrics of FFPE sections (n ​​= 11,334) and H&E stained sections (n ​​= 8,648). They show the number of detected genes (Fig. 28b), the number of UMIs per nucleus (Fig. 28c), the ratio of mapped reads for mitochondrial genes (Fig. 28d), and the ratio of mapped reads for rRNA (Fig. 28e).

[0090] Figures 28f and 28g show the results of UMAP analysis of FFPE sections and H&E stained sections. In the UMAP analysis of the merged gene expression matrix, the same cell type was co-clustered even though they were separated under different conditions. Although each cluster was individually annotated, the analysis results showed that the gene expression patterns of the corresponding cell types were similar even under different conditions via FX-seq.

[0091] Figures 28h and 28i show the cell type annotations and cell number distributions of each cluster. Figure 28h shows the cell number distribution of the FFPE section, and Figure 28i shows the cell number distribution of the H&E stained section.

[0092] Figures 28j and 28k show the expression of representative marker genes of each cluster. Figure 28j shows the FFPE section, and Figure 28k shows the marker gene expression.

[0093] FIG. 28l applied FX-seq to the nuclei of human reticular metastatic cancer samples prepared from 4μm and 10μm FFPE and H&E sections (indicated as 4F, 10F, 4H, and 10H, respectively), and the analysis results are shown in FIG. 28m to FIG. 28u (n = 8,590 in 4F, n = 10,395 in 10F, n = 14,658 in 4H, and n = 11,206 in 10H).

[0094] Figure 28 shows the results of UMAP analysis for the combined gene expression matrix of four conditions.

[0095] Figure 28n shows that individual nuclei in UMAPs labeled according to experimental conditions (Figure 28m) are distributed biasedly depending on the presence or absence of H&E staining.

[0096] Figures 28o through 28r show the QC metrics after performing FX-seq in Figure 28l. They show the number of detected genes (Figure 28o), the number of UMIs per nucleus (Figure 28p), the ratio of mapped reads to mitochondrial genes (Figure 28q), and the number of described UMIs according to top genes aligned with the number of detected UMIs under each condition (Figure 28r). 4F had fewer genes and UMIs than 10F due to physical subsampling, and H&E staining had a greater impact than cross-sectional thickness.

[0097] Figure 28s shows the distribution of UMI in each cell type labeled according to experimental conditions.

[0098] Figure 28 shows the distribution of cell counts for each cell type.

[0099] Figure 28u shows a dot plot of marker genes specific to each cluster.

[0100] Abbreviations: CGN, cerebellar granulosa neuron; CB Int, cerebellar internal neuron; Purkinje, Purkinje cell; OB Int, olfactory progenitor internal neuron; OB Immature, olfactory progenitor immature neuron; OB neuroblast, olfactory progenitor neuroblast; Pallial Glut, palatal glutamate neuron; Int neuron, interneuron; DG neuron, dentate gyrus neuron; Cholinergic, habenura cholinergic neuron; MSN, intermediate spinous neuron; Dien glut, diene glutamate neuron; Astro, astrocyte; Bergmann, Bergmann glial cell; Oligo, oligodendrocyte; Ependyma, posterior epithelial cell; Chor, choroidal plexus epithelial cell; EC, endothelial cell; VLMC, vascular leptomeningeal cell; COP, oligodendrocyte precursor; OPC, oligodendrocyte precursor; SMC, smooth muscle cell; LEC, lymphoid endothelial cell.

[0101] Figures 29a to 29e show the results of comparing the effects of crosslinking agent exchange treatment and the type of reverse transcription primer on human cancer FFPE tissue. The gene body coverage of cDNA synthesized using oligo(dT) (Figure 29a) and the gene body coverage of cDNA synthesized using random primers (Figure 29b) are shown. Figures 29c and 29d show the number of UMIs for single-nuclear transcriptome results using oligo(dT) and random primers, respectively, before and after the application of FX-seq. It was clearly shown that UMIs increased in both cases. Among these, comparisons were made using barcoded random primers or dT-based primers during reverse transcription. Figure 29e is a UMAP diagram of cell types discovered using random primers and oligo(dT).

[0102] Figures 30a to 30j show the results of FX-seq analysis using mouse brain FFPE and H&E sections.

[0103] Figures 30a and 30b show microscopic images of cell nuclei extracted from mouse brain FFPE sections at 10x magnification (Fig. 30a) and 40x magnification (Fig. 30b). Black and white: phase contrast, blue: DAPI.

[0104] Figures 30c and 30d show microscopic images of cell nuclei extracted from mouse brain H&E sections at 10x magnification (Fig. 30c) and 40x magnification (Fig. 30d). Black and white: phase contrast, blue: DAPI.

[0105] Figures 30e and 30f show the results of individually clustered UMAP analysis of mouse brain FFPE sections (Figure 30e) and mouse brain H&E sections (Figure 30f).

[0106] Figures 30g and 30h show the number of UMIs by annotated cell type in mouse brain FFPE sections (Figure 30g) and mouse brain H&E sections (Figure 30h).

[0107] Figures 30i and 30j show the results of DEG analysis showing marker genes in mouse brain FFPE sections (Figure 30i) and mouse brain H&E sections (Figure 30j).

[0108] Abbreviations: CGN, cerebellar granulosa neuron; CB Int, cerebellar internal neuron; Purkinje, Purkinje cell; OB Int, olfactory progenitor internal neuron; OB Immature, olfactory progenitor immature neuron; OB neuroblast, olfactory progenitor neuroblast; Pallial Glut, palatal glutamate neuron; Int neuron, interneuron; DG neuron, dentate gyrus neuron; Cholinergic, habenura cholinergic neuron; MSN, intermediate spinous neuron; Dien glut, diene glutamate neuron; Astro, astrocyte; Bergmann, Bergmann glial cell; Oligo, oligodendrocyte; Ependyma, posterior epithelial cell; Chor, choroidal plexus epithelial cell; EC, endothelial cell; VLMC, vascular leptomeningeal cell; COP, oligodendrocyte precursor; OPC, oligodendrocyte precursor; SMC, smooth muscle cell; LEC, lymphoid endothelial cell.

[0109] Figures 31a and 31b show the morphological evaluation and RNA quality evaluation of reticular metastatic cancer tissue.

[0110] Figure 31a shows a microscopic image of a 4 μm H&E-stained section of a reticular metastatic cancer FFPE sample. Figure 31b shows an automated electrophoresis graph of total RNA extracted from a reticular metastatic cancer FFPE sample. Data are expressed in fluorescence units (FU) normalized per unit size (bp). "Lower" indicates a lower molecular weight standard.

[0111] Figures 32a to 32m show the scalability of FX-seq to human surgical specimens in which gastrointestinal stromal tumor (GIST) FFPE blocks were identified by FX-seq analysis.

[0112] Figure 32a shows surgically excised human GIST tissue immediately immersed in 4% PFA, fixed at 4°C for 48 hours, and then embedded in an FFPE block.

[0113] Figures 32b to 32e show the QC metrics after performing FX-seq. They show the number of detected genes (Fig. 32b), the number of UMIs per nucleus (Fig. 32c), the ratio of mapped reads for mitochondrial genes (Fig. 32d), and the ratio of mapped reads for rRNA (Fig. 32e). It was confirmed that the nuclei isolated from the FFPE block had low contamination by mitochondrial genes and rRNA.

[0114] Figure 32f shows the results of UMAP analysis on transcriptome distribution and clustering. Various cell types were identified based on molecular criteria.

[0115] Figure 32g shows a dot plot of marker genes specific to each cluster.

[0116] Figures 32h and 32i show the cluster network configured based on PAGA analysis (Figure 32h) and diffusion pseudotime heatmap (Figure 32i).

[0117] Figures 32j to 32m show transcriptome pathways and corresponding gene expression patterns identified by FX-seq. It shows the resolved tumor pathway 1 (Figure 32j) and the corresponding gene expression change per cluster (Figure 32k). It shows the resolved tumor pathway 2 (Figure 32l) and the corresponding gene expression change per cluster (Figure 32m). The blue frame image on the right shows the 20-fold magnified expression profile change of pathway 2 in tumor subclusters 9 through 11.

[0118] Abbreviations: LEC, Lymphoid Endothelial Cell; EC, Endothelial Cell; SMC, Smooth Muscle Cell; pDC, Plasma Dendritic Cell; cDC, Classical Dendritic Cell; NK Cell, Natural Killer Cell; DPT, Proliferating Time-like Cell.

[0119] Figures 33a to 33c confirm the morphological evaluation and RNA quality evaluation of gastrointestinal stromal tumors (GIST).

[0120] Figure 33a shows a microscopic image of a 4 μm H&E stained section of a GIST FFPE sample.

[0121] Figure 33b shows the automated electrophoresis graph of total RNA extracted from GIST FFPE samples. Data are expressed in fluorescence units (FU) normalized per unit of size (bp). "Lower" indicates a lower molecular weight standard.

[0122] Figure 33c shows the UMI of the annotated cell type shown in Figure 32.

[0123] Abbreviations: NK cell, Natural Killer cell; PC, Plasma cell; MC, Macrophage; cDC, Classical dendritic cell; pDC, Plasma dendritic cell; FB, Fibroblast; SMC, Smooth muscle cell; EC, Endothelial cell; LEC, Lymphoid endothelial cell.

[0124] Figures 34a to 34c show differential gene expression profiles of tumor populations in gastrointestinal tumors (GIST). Figure 34a shows the expression of the GIST marker gene (ANO1) and GIST oncogenes (KIT and PDGFRA). Figure 34b shows the expression of cell proliferation marker genes. Figure 34c shows the expression of genes related to the function of Cajal cells (ICC).

[0125] Figures 35a to 35c show the gene expression profiles associated with pathway 1 of a tumor population in GIST.

[0126] Figures 36a to 36c show the gene expression profiles associated with pathway 2 of a tumor population in GIST.

[0127] Figures 37a to 37j show the results of FX-seq analysis on stored colorectal cancer (CRC) FFPE tissue.

[0128] Figure 37a shows a full microscopic image of the H&E-stained portion of the CRC FFPE. Figures 37b through 37d show magnified images of various morphological regions. Figure 37b shows the boundary between the non-tumor (NT) and tumor (T) regions. Figure 37c shows the tumor region with collapsed mucosal tissue morphology, and Figure 37d shows the non-tumor region with intact tissue morphology. Figure 37e shows the results of automated electrophoresis of total RNA extracted from the FFPE section of the CRC specimen. Data are expressed in fluorescence units (FU) normalized per unit size (bp). "Lower" indicates a lower molecular weight standard.

[0129] Figures 37f to 37i show quality indicators of a CRC FX-seq experiment comparing spatially resolved tumor and non-tumor regions. They show the detected genes (Figure 37f), UMI (Figure 37g), mitochondrial readout ratio (h), and intron alignment ratio (Figure 37i).

[0130] Figure 37j shows the annotated cell types of UMI. Abbreviations: SMC, smooth muscle cells; EC, endothelial cells; LEC, lymphoid endothelial cells; Enteric glia, intestinal glial cells.

[0131] Figures 38a to 38m show the results of FX-seq analysis including spatial annotation of stored human colorectal cancer (CRC) FFPE specimens.

[0132] Figure 38a shows the schematic procedure of FX-seq for a stored CRC FFPE specimen. Tumor (T) and non-tumor (NT) regions were distinguished in adjacent FFPE sections based on cancer regions identified by a pathologist based on H&E section imaging.

[0133] Figure 38b shows the results of UMAP analysis of transcripts labeled by region origin in the FFPE section (n = 47,966 in T, n = 13,559 in NT).

[0134] Figure 38c shows the representative marker genes of each cluster.

[0135] Figure 38d shows the distribution of cell populations indicated by region in Figure 38b.

[0136] In Fig. 38e, the proportion of the number of cells belonging to each cluster was calculated from all cells identified in each region. The colored dots represent the same cell type annotations as shown in Fig. 38b.

[0137] Figure 38f shows the results of UMAP analysis to identify sub-clusters of the merged gene matrix of epithelial progenitor cells and tumor populations indicated by the green dotted line in Figure 38b.

[0138] Figure 38g shows the distribution of cell populations indicated in Figure 38f according to regional origin. It was confirmed that the progenitor cell population had a mixed distribution of nuclei in the T and NT regions, whereas the tumor population had a specific distribution of nuclei in the T region.

[0139] Figure 38h shows the representative marker genes of the cell types identified by the sub-cluster analysis of Figure 38f.

[0140] Figure 38i shows the results of cluster-level CNV inference in gene expression of tumor and epithelial progenitor cell populations using the inferCNV package.

[0141] Figure 38j shows the results of UMAP analysis to identify sub-clusters of the fibroblast cluster indicated by the orange dotted circle in Figure 38b.

[0142] Figure 38k shows the distribution of fibroblast subclusters indicated in Figure 38j according to regional origin.

[0143] Figure 38l shows the distribution of cell numbers in each cluster shown in Figures 38j and 38k.

[0144] Figure 38m shows the representative marker genes identified by sub-cluster analysis of Figure 38j.

[0145] Abbreviations: SMC, smooth muscle cell; EC, endothelial cell; LEC, lymphoid endothelial cell; Enteric glia, intestinal glial cell; EP, epithelial cell; Pro, progenitor cell; NMF, normal fibroblast; CAF, cancer-associated fibroblast.

[0146] Figures 39a and 39b show the differential gene expression profiles of tumor populations in colorectal cancer (CRC). Figure 39a shows the gene expression of subclusters including epithelial progenitor cells and tumor populations. Figure 39b shows the gene expression in all identified populations.

[0147] Figures 40a and 40b show the results of inferred CNV analysis for individual cells.

[0148] Figures 41a to 41f show the FFPE block QC and sequencing results of various colorectal cancer patient groups. It was confirmed that the UMI values ​​of each sequenced block had a high correlation with the Ct values ​​obtained by qPCR QC (Figure 41a). In addition, the correlation between the number of UMIs in the sequencing results and each RNA Integrity Number (RIN) (Figure 41b), DV200 (Figure 41c), and DV500 (Figure 41d) was confirmed.

[0149] These are the single-nuclear transcripts, identified cell types, and UMAP plots obtained from FFPE blocks of 24 patients after QC and sequencing (Fig. 41e). The UMAP plot shown is the UMAP plot drawn by dividing the patients into groups according to their microsatellite mutations (Fig. 41f).

[0150] Figures 42a to 42e show the results of crosslinker exchange sequencing (FX-seq) on sarcoma FFPE samples from various patient groups. Crosslinker exchange sequencing was performed on Ewing sarcoma FFPE tissues derived from various patients. The number of UMIs (Fig. 42a) and the number of detected genes (Fig. 42b) for Ewing sarcoma samples derived from each patient are shown in the figure. Various clusters were detected using the Leiden algorithm (Fig. 42c), and these were organized according to the patient (Fig. 42d) or the patient-derived sample number (Fig. 42e).

[0151] Figures 43a to 43c show the results of confirming the effect of additional PFA fixation before crosslinker exchange sequencing and the optimization of cell concentrations for organic catalyst treatment. Each condition was evaluated through qPCR Ct values ​​of human GAPDH, B2M, and CNOT genes.

[0152] Figure 44 shows the results of verifying the compatibility evaluation between cell hashing and a fixed sample.

[0153] Figure 45 confirms the compatibility of fixed nuclei processed by FX-seq with a microfluidics-based single-cell analysis platform. Data are expressed as fluorescence units (FU) normalized per unit of size (bp). "Lower" indicates a lower molecular weight standard.

[0154] Figure 46 shows the results of evaluating the feasibility of analyzing the accessibility of fixed chromatin.

[0155] Figure 47 illustrates the technical scalability of FX-seq for multi-omics analysis.

[0156] Fixed cell or tissue pretreatment method

[0157] To achieve the above objective, one aspect of the present invention provides a cell or tissue pretreatment method comprising the step of exchanging a fixative.

[0158] Fixed fluid exchange step

[0159] As used herein, the term “fixative” may be a solvent for fixing cells or tissues. One embodiment of the fixative may be formalin and / or paraformaldehyde.

[0160] The term "fixation exchange" as used herein may be expressed as FX-seq and means removing molecular bonds caused by the existing fixative and / or replacing it with another fixative for the analysis of cells or tissues. The solution used for fixing the cells or tissues may be formalin, paraformaldehyde (PFA), formaldehyde, etc., but is not limited thereto. The removal of molecular bonds caused by the fixative may be performed by adding an organic solvent. In this case, one embodiment of the organic solvent may be aminophenylboronic acid. Specifically, it may be 2-aminophenylboronic acid. Specifically, it may be 2-aminophenylboronic acid hydrochloride. The FX-seq can improve the preservation of RNA within cells or tissues by using a selective crosslinking agent on the molecular structure, and can efficiently reverse / remove crosslinks or methylol adducts resulting from the PFA reaction by using an organic catalyst.

[0161] At this time, the cell or tissue may be fixed with paraformaldehyde and / or formalin.

[0162] In addition, the fixative exchange can be performed by reacting an organic catalyst with a cell or tissue. Specifically, the fixative exchange can be performed by reacting a cell or tissue fixed with paraformaldehyde and / or formalin. In this case, the organic catalyst may be any one selected from the group consisting of aminophenylboronic acid, cyclic boronic acid ester, phosphonic acid ester, and bismuth salt.

[0163] At this time, the organic catalyst may be a compound represented by the following chemical formula I.

[0164] [Chemical Formula I]

[0165]

[0166] In the above chemical formula I, R A is a hydroxyl, C 1-6 Alkoxy, mercapto, amino, C 1-6 alkylamino or di(C 1-6 It may be an alkyl)amino. In some embodiments, R A is a hydroxyl, C 1-4 Alkoxy, mercapto, amino, C 1-4 alkylamino or di(C 1-4 It may be an alkyl)amino. Preferably, R A It can be an amino acid.

[0167] In some embodiments, R A can be substituted at the ortho, meta, or para position for the boronic acid group (-B(OH)2) of chemical formula I. For example, R A It can be substituted at the ortho or para position for the boronic acid group of chemical formula I.

[0168] In the above chemical formula I, R B is H; halo; hydroxy; C 1-6 Alkoxy; Amino; C 1-6Alkylamino; di(C 1-6 Alkyl)amino; nitro; cyano; carboxy; C 1-6 Alkyl carbonyl; carbamoyl; C 1-6 Alkylcarbamoyl; di(C 1-6 Alkyl)carbamoyl; or halo, hydroxy, C 1-6 Alkoxy, amino, C 1-6 alkylamino, di(C 1-6 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-6 It may be alkyl. In some embodiments, R B is halo; hydroxy; C 1-4 Alkoxy; Amino; C 1-4 Alkylamino; di(C 1-4 Alkyl)amino; nitro; cyano; carboxy; C 1-4 Alkyl carbonyl; carbamoyl; C 1-4 Alkylcarbamoyl; di(C 1-4 Alkyl)carbamoyl; or halo, hydroxy, C 1-4 Alkoxy, amino, C 1-4 alkylamino, di(C 1-4 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-4 It can be an alkyl.

[0169] In some embodiments, the organic catalyst may be a compound represented by the following chemical formula I-1.

[0170] [Chemical Formula I-1]

[0171]

[0172] In the above chemical formula I-1, R B is equal to that defined for chemical formula I.

[0173] Preferably, the organic catalyst may be aminophenylboronic acid. In this case, the aminophenylboronic acid may be 2-aminophenylboronic acid or 3-aminophenylboronic acid, and preferably 2-aminophenylboronic acid. Additionally, the 2-aminophenylboronic acid may be a 2-aminophenylboronic acid hydrochloride salt.

[0174] At this time, the organic catalyst may be treated at a concentration of about 0.1 mM to about 1 M. Additionally, the organic catalyst may be treated at a concentration of about 1 mM to about 500 mM, about 5 mM to about 300 mM, about 10 mM to about 200 mM, about 20 mM to about 100 mM, about 30 mM to about 80 mM, about 40 mM to about 70 mM, or about 50 mM to about 60 mM. Specifically, if the organic catalyst is 2-aminophenylboronic acid, it may be treated at about 1 mM, about 5 mM, about 10 mM, about 15 mM, about 20 mM, about 25 mM, about 30 mM, about 35 mM, about 40 mM, about 45 mM, about 50 mM, about 55 mM, or about 60 mM.

[0175] In addition, the reaction time of the above organic catalyst can be about 1 minute to about 24 hours, about 30 minutes to about 60 minutes, about 5 minutes to about 12 hours, or about 10 minutes to about 6 hours. Specifically, when the organic catalyst is 2-aminophenylboronic acid, the reaction can be carried out for about 1 minute, about 2 minutes, about 5 minutes, about 10 minutes, about 15 minutes, about 20 minutes, about 30 minutes, about 40 minutes, about 50 minutes, about 60 minutes, about 2 hours, about 3 hours, about 4 hours, about 5 hours, about 6 hours, about 7 hours, about 8 hours, about 9 hours, about 10 hours, about 11 hours, about 12 hours, about 13 hours, about 14 hours, about 15 hours, about 16 hours, about 17 hours, about 18 hours, about 19 hours, about 20 hours, about 21 hours, about 22 hours, about 23 hours, or about 24 hours.

[0176] In addition, the above organic catalyst reaction may be carried out at about 4°C to about 99°C, about 10°C to about 70°C, about 20°C to about 60°C, about 30°C to about 65°C, about 40°C to about 60°C, or about 50°C to about 55°C. At this time, if the organic catalyst is 2-aminophenylboronic acid, it may be carried out at about 30°C, about 35°C, about 40°C, about 45°C, about 50°C, about 55°C, or about 65°C.

[0177] Additionally, the organic catalytic reaction may be carried out at about pH 6.5 to about pH 8.8, about pH 7 to about pH 8.5, or about pH 7.8 to about pH 8.3. Additionally, the organic catalytic reaction may be carried out at about pH 7.0, about pH 7.1, about pH 7.2, about pH 7.3, about pH 7.4, about pH 7.5, about pH 7.6, about pH 7.7, about pH 7.8, about pH 7.9, about pH 8.0, about pH 8.1, about pH 8.2, about pH 8.3, about pH 8.4, about pH 8.5, about pH 8.6, about pH 8.7, or about pH 8.8.

[0178] In addition, the above organic catalytic reaction can be treated by adding Tris buffer at a concentration of 10 mM to 1 M. In addition, the above Tris buffer can be treated at a concentration of about 20 mM to about 500 mM, about 40 mM to about 400 mM, about 60 mM to about 300 mM, and about 80 mM to about 200 mM.

[0179] In addition, the above organic catalytic reaction can be carried out by adding Polyethylene glycol (PEG) to the reaction solution. At this time, the PEG may have a molecular weight of 100 to 8000. Preferably, the PEG may be PEG-8000. At this time, the concentration of PEG-8000 in the organic catalytic reaction solution may be about 1% to about 20%, about 4% to about 16%, or about 8% to about 12%.

[0180] Step of adding crosslinkers

[0181] A cell or tissue pretreatment method may further include a step of adding crosslinkers to the cells or tissues.

[0182] At this time, if the organic catalyst is contacted directly, DNA and / or RNA may not be preserved. Therefore, a crosslinker may be added to preserve DNA and / or RNA. In this case, the crosslinker may be added before reacting the organic catalyst.

[0183] In some embodiments, the crosslinker may be a compound represented by the following chemical formula II.

[0184] [Chemical Formula II]

[0185]

[0186] In the above chemical formula II, Z 1a , Z 2a , Z 1b and Z 2b Each is independently O, OH, S, SH, NH, NH2, N(C 1-6 alkyl) or NH(C 1-6 It may be an alkyl. In some embodiments, Z 1a , Z 2a , Z 1b and Z 2b Each can independently be OH, SH, or NH2.

[0187] In some embodiments, Z 1a and Z2a They may be identical or different from each other. In some embodiments, Z 1b and Z 2b They may be identical or different from each other. In some embodiments, Z 1a and Z 1b They may be identical or different from each other. In some embodiments, Z 2a and Z 2b They may be the same or different from each other.

[0188] In the above chemical formula II, R 1a and R 1b Each can independently be F, Cl, Br, or I. In some embodiments, R 1a and R 1b Each can independently be Cl or Br. In some embodiments, R 1a and R 1b They may be the same or different from each other.

[0189] In the above chemical formula II, Q a and Q b Each independently directly bonds, O, S, NH, NH2, N(C 1-6 alkyl), or N(C 1-6 It may be alkyl)2. In some embodiments, Q a and Q b Each can independently be directly bonded, O, S, NH, or NH2. In some embodiments, Q a and Q b They may be the same or different from each other.

[0190] In the above chemical formula II, L is C optionally interrupted by NH, O, or S. 2-40 It may be an alkylene. In some embodiments, L is C optionally interrupted by NH, O, or S. 10-40 It may be an alkylene. In some embodiments, L is C optionally interrupted by NH, O, or S. 20-30 It can be an alkylene. For example, L is C 10-40Alkylene; C 20-30 Alkylene; C interrupted by NH, O, or S 10-40 Alkylene; or C interrupted by NH, O, or S 20-30 It can be an alkylene.

[0191] In the above chemical formula II, R 4a and R 4b Each independently halo; hydroxy; C 1-6 Alkoxy; Amino; C 1-6 Alkylamino; di(C 1-6 Alkyl)amino; nitro; cyano; carboxy; C 1-6 Alkyl carbonyl; carbamoyl; C 1-6 Alkylcarbamoyl; di(C 1-6 Alkyl)carbamoyl; or halo, hydroxy, C 1-6 Alkoxy, amino, C 1-6 alkylamino, di(C 1-6 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-6 It may be alkyl. In some embodiments, R 4a and R 4b Each independently halo; hydroxy; C 1-4 Alkoxy; Amino; C 1-4 Alkylamino; di(C 1-4 Alkyl)amino; nitro; cyano; carboxy; C 1-4 Alkyl carbonyl; carbamoyl; C 1-4 Alkylcarbamoyl; di(C 1-4 Alkyl)carbamoyl; or halo, hydroxy, C 1-4 Alkoxy, amino, C 1-4 alkylamino, di(C 1-4 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-4 It may be alkyl. In some embodiments, R 4a and R 4b They may be the same or different from each other.

[0192] In the above formula II, l and m may each be an integer from 1 to 3 independently. In some embodiments, l and m may each be 2 or 3 independently. In some embodiments, l and m may be the same or different from each other.

[0193] In the above formula II, o and p may each be an integer from 0 to 2 independently. For example, o and p may each be 0 or 1 independently. In some embodiments, o and p may be the same or different from each other.

[0194] In some embodiments, the compound of Formula II may have a chemical structure symmetric with respect to L.

[0195] In some embodiments, the crosslinker may be a compound represented by the following chemical formula II-1.

[0196] [Chemical Formula II-1]

[0197]

[0198] In the above chemical formula II-1, Z 1a , Z 2a , Z 1b , Z 2b , R 1a , R 1b , Q a , Q b , L, R 4a , R 4b , o and p are as defined for Chemical Formula II.

[0199] In some embodiments, the crosslinker may be a compound represented by the following chemical formula IIA.

[0200] [Chemical Formula IIA]

[0201]

[0202] In the above chemical formula IIA, R 1a , R 1b , Q a , Q b , L, R4a , R 4b , o and p are as defined for Chemical Formula II.

[0203] In the above chemical formula IIA, R 2a , R 3a , R 2b and R 3b Each independently H or C 1-6 It may be alkyl. In some embodiments, R 2a , R 3a , R 2b and R 3b Each independently H or C 1-4 It can be an alkyl.

[0204] In some embodiments, R 2a and R 3a They may be identical or different from each other. In some embodiments, R 2b and R 3b They may be identical or different from each other. In some embodiments, R 2a and R 2b They may be identical or different from each other. In some embodiments, R 3a and R 3b They may be the same or different from each other.

[0205] In some embodiments, the crosslinker may be a compound represented by the following chemical formula IIA-1.

[0206] [Chemical Formula IIA-1]

[0207]

[0208] In the above chemical formula IIA-1, R 1a , R 1b , Q a , Q b , L, R 2a , R 3a , R 2b , R 3b , R 4a , R 4b , o and p are as defined for the chemical formula IIA.

[0209] In some embodiments, the crosslinker may be a compound represented by the following chemical formula IIB.

[0210] [Chemical Formula IIB]

[0211]

[0212] In the above chemical formula IIB, R 1a , R 1b , Q a , Q b , R 2a , R 3a , R 2b , R 3b , R 4a , R 4b , o and p are as defined for the chemical formula IIA.

[0213] In the above chemical formula IIB, R 5a and R 5b H, H2, and C are independently 1-6 alkyl or (C 1-6 It may be alkyl)2. In some embodiments, R 5a and R 5b H, H2, and C are independently 1-6 alkyl or (C 1-6 It may be alkyl)2. In some embodiments, R 5a and R 5b They may be the same or different from each other.

[0214] In the above chemical formula IIB, L a and L b C each independently 1-6 It may be an alkylene. In some embodiments, L a and L b C each independently 1-4 It may be an alkylene. In some embodiments, L a and L b They may be identical or different from each other. For example, L a and L bEach can be methylene, ethylene, propylene, or butylene.

[0215] In the above chemical formula IIB, Q 1 can be a direct bond, NH, O, or S. In some embodiments, Q 1 It can be NH, O, or S.

[0216] In the above formula IIB, n may be an integer from 1 to 20. In some embodiments, n may be an integer from 5 to 15.

[0217] In some embodiments, the crosslinker may be a compound represented by the following chemical formula IIB-1.

[0218] [Chemical Formula IIB-1]

[0219]

[0220] In the above chemical formula IIB-1, R 1a , R 1b , R 2a , R 3a , R 2b , R 3b R 4a , R 4b , R 5a , R 5b , o, p and n are as defined for the chemical formula IIB.

[0221] In one embodiment, the crosslinker may include a compound represented by the following chemical formula 1.

[0222] [Chemical Formula 1]

[0223] .

[0224] In addition, the reaction time of the crosslinker can be about 1 minute to about 24 hours, about 5 minutes to about 12 hours, about 10 minutes to about 6 hours, or about 30 minutes to about 3 hours. Specifically, when the organic catalyst is 2-aminophenylboronic acid, the reaction can be carried out for about 1 minute, about 2 minutes, about 5 minutes, about 10 minutes, about 15 minutes, about 20 minutes, about 30 minutes, about 1 hour, about 2 hours, about 3 hours, about 4 hours, about 5 hours, about 6 hours, about 7 hours, about 8 hours, about 9 hours, about 10 hours, about 11 hours, about 12 hours, about 13 hours, about 14 hours, about 15 hours, about 16 hours, about 17 hours, about 18 hours, about 19 hours, about 20 hours, about 21 hours, about 22 hours, about 23 hours, or about 24 hours.

[0225] In addition, the reaction of the crosslinker can be carried out at about 4°C to about 99°C, about 10°C to about 35°C, about 20°C to about 30°C, or about 23°C to about 25°C. At this time, if the crosslinker is the compound, it can be carried out at about 30°C, about 35°C, about 40°C, about 45°C, about 50°C, about 55°C, or about 65°C.

[0226] Step of adding an RNase inhibitor

[0227] The step of adding an RNase inhibitor may be further included. The RNase may continue to be included in the fixative substitution step. Additionally, the RNase inhibitor may be any one selected from the group consisting of polyvinyl sulfonic acid, polyvinyl sulfate, polyvinyl phosphonic acid, polyvinyl phosphate, polyvinyl nitric acid, polyethyleneimine, polyvinyl chloride, polyvinyl bromide, polystyrene, polyvinyl acetate, polystyrene sulfonate, and heparan sulfate.

[0228] Preferably, the RNase inhibitor may be polyvinyl sulfonic acid (PVSA). Additionally, the length of the vinyl chain of the RNase inhibitor may be at least 10 mer. Preferably, the length of the vinyl chain may be about 10 mer to about 1000 mer.

[0229] The above RNase inhibitor may be a recombinant protein-based RNase inhibitor, diethyl pyrocarbonate (DEPC).

[0230] The step of separating the nucleus

[0231] The method may further include a step of isolating the nucleus by treating with a protease. The protease may be any one selected from the group consisting of serine protease, cysteine ​​protease, and aspartic protease. Specifically, the protease may be one of collagenase, dispase, pepsin, papain, proteinase K, subtisilin, trypsin, or other specific proteases or other non-specific proteases. In this case, the protease may be cold-active, meaning it is capable of reacting at low temperatures. Preferably, the protease may be proteinase K.

[0232] Deparaffinization stage

[0233] The method may further include a deparaffinization step of fixed cells and / or tissues. The deparaffinization may be any one selected from the group consisting of xylene, toluene, limonene, and mineral oil. In this case, the step may be performed by adding xylene and / or limonene. In this case, preferably, deparaffinization may be performed using xylene.

[0234] [1 Specific Example]

[0235] FX Seq Analysis Method 1

[0236] The above cell or tissue pretreatment method can be specifically performed as follows.

[0237] The nucleic acids of paraformaldehyde (PFA) fixed cells and FFPE tissues were analyzed by performing the following steps:

[0238] A step of reacting paraformaldehyde (PFA) fixed cells and FFPE tissues with a crosslinker having the structure of Formula 1 and / or reacting the crosslinker with PVSA; a step of treating with 2-aminophenylboronic acid and / or treating with 2-aminophenylboronic acid and PVSA; a step of treating with a protease and / or treating with a protease and PVSA; a step of reverse transcribing mRNA into cDNA; and a step of analyzing the cDNA.

[0239] FX Seq Analysis Method 2

[0240] The above cell or tissue pretreatment method can be specifically performed as follows.

[0241] The nucleic acids of paraformaldehyde (PFA) fixed cells and FFPE tissues were analyzed by performing the following steps:

[0242] A step of reacting paraformaldehyde (PFA) fixed cells and FFPE tissues with a crosslinker having the structure of Formula 1 and PVSA; a step of treating with 2-aminophenylboronic acid and / or treating together with 2-aminophenylboronic acid and PVSA; a step of treating with a protease and / or treating together with a protease and PVSA; a step of reverse transcribing mRNA into cDNA; and a step of analyzing cDNA.

[0243] FX Seq Analysis Method 3

[0244] The above cell or tissue pretreatment method can be specifically performed as follows.

[0245] The nucleic acids of paraformaldehyde (PFA)-fixed cells or FFPE tissues were analyzed by performing the following steps:

[0246] A step of analyzing the tissue by staining the FFPE tissue with H&E; a step of reacting a crosslinker having the structure of Formula 1 and / or reacting the crosslinker with PVSA; a step of treating with 2-aminophenylboronic acid and / or treating with 2-aminophenylboronic acid and PVSA; a step of treating with a protease and / or treating with a protease and PVSA; a step of reverse transcribing mRNA into cDNA; and a step of analyzing the cDNA.

[0247] FX Seq Analysis Method 4

[0248] The above cell or tissue pretreatment method can be performed as follows.

[0249] The nucleic acids of paraformaldehyde (PFA)-fixed cells or FFPE tissues were analyzed by performing the following steps:

[0250] A step of reacting para-formaldehyde (PFA) fixed cells or FFPE tissues with a crosslinker having the structure of Formula 1 and / or reacting the crosslinker with PVSA; a step of treating with 2-aminophenylboronic acid and / or treating with 2-aminophenylboronic acid, protease, and PVSA; a step of reverse transcribing mRNA into cDNA; and a step of analyzing the cDNA.

[0251] FX Seq Analysis Method 5

[0252] The above cell or tissue pretreatment method can be specifically performed as follows.

[0253] The nucleic acids of paraformaldehyde (PFA)-fixed cells or FFPE tissues were analyzed by performing the following steps:

[0254] A step of treating paraformaldehyde (PFA) fixed cells or FFPE tissues with 2-aminophenylboronic acid and / or together with 2-aminophenylboronic acid and PVSA; a step of treating with protease and / or together with protease and PVSA; a step of reverse transcribing mRNA into cDNA; and a step of analyzing cDNA.

[0255] FX Seq Analysis Method 6

[0256] The above cell or tissue pretreatment method can be specifically performed as follows.

[0257] The nucleic acids of paraformaldehyde (PFA) fixed cells were analyzed by performing the following steps:

[0258] A step of treating paraformaldehyde (PFA) fixed cells with 2-aminophenylboronic acid and / or treating them together with 2-aminophenylboronic acid and PVSA; a step of reverse transcribing mRNA into cDNA; and a step of analyzing cDNA.

[0259] FX Seq Analysis Method 7

[0260] The above cell or tissue pretreatment method can be specifically performed as follows.

[0261] The nucleic acids of paraformaldehyde (PFA) fixed cells were analyzed by performing the following steps:

[0262] A step of reacting a paraformaldehyde (PFA) fixed cell with a crosslinker having the structure of Formula 1 and / or reacting the crosslinker with PVSA; a step of treating with 2-aminophenylboronic acid and / or treating with 2-aminophenylboronic acid and PVSA; a step of reverse transcribing mRNA into cDNA; and a step of analyzing the cDNA.

[0263] FX Seq Analysis Method 8

[0264] The above cell or tissue pretreatment method can be specifically performed as follows.

[0265] The nucleic acids of paraformaldehyde (PFA)-fixed cells or FFPE tissues were analyzed by performing the following steps:

[0266] A step of treating paraformaldehyde (PFA) fixed cells or FFPE tissues with 2-aminophenylboronic acid and / or together with 2-aminophenylboronic acid and PVSA; a step of treating with a protease and / or together with a protease and PVSA; a step of adding a fluorescent probe to nucleic acids; and a step of analyzing spatial information based on fluorescent imaging.

[0267] FX Seq Analysis Method 9

[0268] The above cell or tissue pretreatment method can be specifically performed as follows.

[0269] The nucleic acids of paraformaldehyde (PFA)-fixed cells or FFPE tissues were analyzed by performing the following steps:

[0270] A step of reacting paraformaldehyde (PFA) fixed cells and FFPE tissues with a crosslinker having the structure of Formula 1 and / or reacting the crosslinker with PVSA; a step of treating with 2-aminophenylboronic acid and / or treating with 2-aminophenylboronic acid and PVSA; a step of treating with a protease; a step of adding a fluorescent probe to the nucleic acid; and a step of analyzing spatial information based on fluorescent imaging.

[0271] 10 FX Seq Analysis Methods

[0272] The above cell or tissue pretreatment method can be performed as follows.

[0273] The nucleic acids of paraformaldehyde (PFA)-fixed cells or FFPE tissues were analyzed by performing the following steps:

[0274] Steps for treating paraformaldehyde (PFA) fixed cells and FFPE tissues with 2-aminophenylboronic acid and / or together with 2-aminophenylboronic acid and PVSA; step of treating with protease and / or together with protease and PVSA; step of adding a space barcode; step of reverse transcribing mRNA into cDNA; and step of analyzing cDNA.

[0275] FX Seq Analysis Method 11

[0276] The above cell or tissue pretreatment method can be performed as follows.

[0277] The nucleic acids of paraformaldehyde (PFA)-fixed cells or FFPE tissues were analyzed by performing the following steps:

[0278] Steps for reacting paraformaldehyde (PFA) fixed cells and FFPE tissues with a crosslinker having the structure of Formula 1 and / or reacting the crosslinker with PVSA; step of treating with 2-aminophenylboronic acid and / or treating with 2-aminophenylboronic acid and PVSA; step of treating with a protease and / or treating with a protease and PVSA; step of adding a space barcode; step of reverse transcribing mRNA into cDNA; and step of analyzing the cDNA.

[0279] FX Seq Analysis Method 12

[0280] The above cell or tissue pretreatment method can be performed as follows.

[0281] The nucleic acids of paraformaldehyde (PFA)-fixed cells or FFPE tissues were analyzed by performing the following steps:

[0282] Steps for treating paraformaldehyde (PFA) fixed cells and FFPE tissues with 2-aminophenylboronic acid and / or together with 2-aminophenylboronic acid and PVSA; step of adding a space barcode; step of treating with a protease and / or together with a protease and PVSA; step of reverse transcribing mRNA into cDNA; and step of analyzing the cDNA.

[0283] Method for analyzing nucleic acids in fixed cells or tissues

[0284] Another aspect of the present invention provides a method for analyzing nucleic acids in a cell or tissue, comprising the step of detecting nucleic acids in a cell or tissue pretreated by the above method.

[0285] At this time, the cell or tissue may be formalin-fixed or formalin-fixed paraffin-embedded (FFPE). In addition, the nucleic acid may be RNA, and in one embodiment may be mRNA.

[0286] Optimization of FX and nuclear separation processes

[0287] To address the issues related to PFA fixation, a method was first adopted to promote the removal of PFA adducts under mild conditions using an organic catalyst. For this purpose, an approach utilizing an organic catalyst was introduced based on the study by Karmakaret et al. It was confirmed that the reaction between PFA and RNA actually inhibits reverse transcription (Fig. 1b). As a result of further screening for catalyst candidates to secure better catalytic activity conditions, it was confirmed that Catalyst 2 (Cat.2) exhibited superior performance compared to previously reported catalysts in RNA synthesized under in vitro conditions and in RNA isolated after fixation of cultured HeLa cells (Figs. 1c, 2a, and 2b). To optimize the treatment of the organic catalyst, optimal conditions were found after testing the catalytic treatment effect (Fig. 3), catalyst concentration (Fig. 4), temperature (Fig. 5), pH conditions (Fig. 6), solution additives (Fig. 7), reaction time (Figs. 8a to 8c), and reverse transcription (Fig. 9).

[0288] Next, the FX-seq protocol was optimized for fixed cultured HeLa cells (Fig. 1d). During the process of adjusting the catalytic treatment time to maximize reverse transcription efficiency, it was discovered that as the catalytic treatment time increased, the amount of RNA leaking out of the cytoplasm increased, resulting in RNA leakage from the fixed cells (Fig. 1e). To prevent this loss, the cells were further fixed using a molecularly structured site-specific Pt-based chemical crosslinker synthesized by combining a guanine-N7 specific cisplatin derivative with a PEG (Polyethylene Glycol) linker (Figs. 14a to 14c). It was confirmed that this crosslinker can selectively bind to guanine-N7 atoms without interfering with Watson-Crick base pair formation, thereby forming additional crosslinks between RNA and surrounding materials (Fig. 14d).

[0289] Experimental results using isolated RNA confirmed that site-specific crosslinking on guanine-N7 minimized the effect on reverse transcription efficiency without affecting Watson-Crick base pair formation. This demonstrated that a cDNA synthesis yield similar to that of unfixed RNA could be maintained (Figs. 15a to 15j and Fig. 10). Subsequently, additional in situ crosslinking conditions were optimized to minimize RNA leakage and ultimately improve cDNA yield without affecting the removal of PFA reactants (Figs. 1d, 1e and Figs. 11a to 13b).

[0290] In addition, to extend the applicability of FX-seq to more complex biological contexts, the nuclear separation procedure was optimized for strongly fixed tissues and organs. The optimized protocol successfully extracted nuclei from various mouse organs (Figs. 16, 17, 18, and 19). However, it was observed that RNA quality was significantly degraded due to RNase activity or RNA autolysis within the tissues during processing steps, such as heating for nuclear separation and removal of PFA reactants (Fig. 1g). To address this, effective and cost-effective RNase inhibitors (RIs) were screened to replace expensive protein-based RNase inhibitors, and PVSA was discovered, which is heat-stable and does not inhibit enzymatic reactions for RNA sequencing.

[0291] PVSA is a non-reactive competitor that binds to the active site of RNase. It inhibits RNase activity while remaining undetectable to RNA, thereby maintaining the RNA in its initial state and improving in situ reverse transcription yield. In the unheated group, both PVSA and protein-based RIs showed similar results in effectively protecting against RNA degradation (Fig. 20a); however, under moderate heating conditions for catalytic treatment, PVSA was found to be the most effective among the tested RIs (Figs. 1f and 1g). Furthermore, it was confirmed that while the nucleic acid-reactive chemical DEPC (Diethyl pyrocarbonate) reacts with RNA to inhibit subsequent reverse transcriptase reactions, PVSA does not have a negative effect on subsequent enzymatic reactions, such as those involving proteases (Fig. 20b). Finally, each component of FX-seq was further verified by quantifying in situ synthesized cDNA from nuclei extracted from strongly immobilized mouse brain tissue using qPCR (Figs. 21 and 22).

[0292] FX-seq application to PFA-fixed and FFPE mouse brains

[0293] To perform single-nucleus sequencing on nuclei isolated after FX treatment, the existing sci-RNA-seq3 protocol was modified and combinatorial cell barcoding was applied (see Examples 20 and 21).

[0294] To verify the reproducibility and performance of FX-seq, mice were perfused with PFA, brain tissue was extracted, and fixed at 4°C for 48 hours. Subsequently, FX treatment was applied to the fixed brain, and sequencing for single-nuclear transcriptome analysis was performed on the isolated nuclei (n = 3, Fig. 1h, Fig. 1i). FX-seq restored the cDNA synthesis efficiency that was limited in PFA-fixed tissue, successfully obtaining transcriptome information that was previously undetectable in fixed biological samples.

[0295] In particular, by improving data quality based on unique molecular identifiers (UMI) and gene count, more single-nuclear transcriptome information was recovered and various cell types were identified compared to the PFA control group (PFA ctrl) in which nuclei were isolated without fixative exchange, the uncatalyzed heating experimental group ((-) Cat.2) under high-concentration Tris buffer conditions of FX, and the heating experimental group with added Cat.2 (Cat.2) (number of nuclei extracted from each brain: n = 7,645, 8,815, 8,815, 9,344; Figs. 1j to 1n).

[0296] Despite applying a nuclear isolation procedure optimized to preserve RNA quality, cell type identification in PFA-fixed samples was limited due to low reverse transcription efficiency (Figs. 23a to 23h). However, FX-seq significantly improved cell type identification performance by accurately classifying neurons, and in particular, succeeded in distinguishing medium spiny neurons (MSNs) from inhibitory neurons. In independent analyses and UMAP clustering results combined with PFA controls, FX-seq more accurately classified various cell types, such as Purkinje cells, olfactory ensheathing cells (OECs), and macrophages, according to existing molecular classification criteria (Figs. 10 to 1t and Figs. 23i to 23l).

[0297] To verify the accurate performance of FX-seq, we performed conventional snRNA-seq on weakly fixed nuclei (fixed with 0.1% PFA at 4°C for 10 minutes) using EasySci's latest protocol (see Methods). Although weakly fixed nuclei showed higher transcriptome coverage, applying FX-seq to strongly fixed nuclei resulted in transcriptome information with UMI and gene numbers similar to those of the weakly fixed nuclei (Figs. 24a and 24b). However, the process of isolating fresh nuclei did not involve protease treatment, resulting in incomplete nuclear separation and a higher proportion of cytoplasmic fraction (Figs. 25a and 25b).

[0298] This was indicated by higher mitochondrial and ribosomal RNA read rates and increased exon alignment rates in RNA sequences (Figs. 24c to 24e). This trend was also observed in differences within cell types distinguished by unsupervised clustering. Cell types were independently labeled for single-cell transcriptome information from two groups of weakly fixed fresh nuclei and fixation-post-FX-seq treated nuclei (Figs. 24f and 24g), and in the unsupervised clustering analyzed after merging the gene expression matrices of the two groups, the same cell types were closely located (Figs. 24h and 24i). However, it was confirmed that the bias in cell distribution within the clusters of fresh nuclei and FX-seq treated nuclei followed the axis of exon alignment rate distribution (Fig. 24j). This was further supported by differences in the distribution of exon alignment rates of differentially expressed genes between the two groups within the same cell type (Fig. 24k). Overall, FX treatment restored complex cell resolution in fixed tissues to be similar to fresh nuclei and was confirmed to have superior nuclear separation stability.

[0299] To extend the utility of FX-seq and simulate human FFPE samples used in clinical practice, we treated three mouse-derived mouse brain FFPE tissues identically with FX-seq (Fig. 26a). To mimic actual clinical conditions, the extracted brains were fixed by immersing them in a PFA solution without PFA perfusion, followed by paraffin embedding. Although the overall quantitative indicators of the acquired transcriptome data were degraded due to the paraffin embedding process, FX-seq increased both the total number of UMIs and the number of detected genes despite the quality degradation caused by FFPE treatment (Figs. 26b to 26d).

[0300] In particular, FX treatment improved cell type annotation in the merged UMAP clustering after independent cell type analysis and provided additional information that could clearly distinguish specific cell populations (e.g., palatal glial cells (Satb2 expression), non-palatal neurons (Zfhx3+ neurons), inhibitory interneurons, etc.) (Figs. 26e to 26l and Fig. 27). On the other hand, sequencing results of the untreated control sample did not provide sufficient information to distinguish these specific cell types and simply classified them as general neurons.

[0301] Application to FFPE and H&E sections derived from mouse brain and human clinical tissue

[0302] In the next step of the study, the compatibility of FX-seq with FFPE mouse brain sections and hematoxylin and eosin (H&E) stained sections was investigated. With slight protocol modifications, up to approximately 200,000 nuclei were successfully isolated from single 10 μm FFPE or H&E mouse brain sections (Figs. 28a and 28c). Although the total UMI and number of genes decreased due to the physical cleavage of nuclei during the sectioning process of the FFPE samples, similar clusters were observed compared to the FFPE blocks, confirming that transcriptomic information was preserved (Figs. 28b and 28c).

[0303] The difference in cell types between the FFPE block and the section samples is attributed to the fact that only a portion of the brain was collected in the case of the sections, resulting in a change in the cell population distribution. When FX-seq was applied to H&E stained sections, the read rates for mitochondrial and ribosomal RNA increased, while the number of UMIs and genes decreased (Figs. 28d and 28e). This is thought to be the result of increased accumulation of cell debris due to reduced proteolytic efficiency and harsh staining conditions (Figs. 30a to 30d). Nevertheless, the results of merging and analyzing the transcriptome information of the FFPE and H&E sections showed that individually labeled cell types appeared in the same cluster, suggesting that reliable and consistent transcriptome analysis is possible on both platforms (Figs. 28f to 28k and Figs. 30e to 30j).

[0304] Scalability of the sample

[0305] The effects of section thickness and H&E staining on the sequencing data quality of human clinical samples were investigated. FX-seq was applied to surgical specimens of metastatic cancer with omental metastasis from primary bladder cancer (approx. 2.5 cm x 1.2 cm per slide) (Figs. 31a and 31b). The specimens were fixed with 4% PFA at 4°C for 48 hours and then treated with FFPE. Subsequently, the FFPE blocks were sectioned to thicknesses of 4 μm and 10 μm, and H&E staining was performed (Fig. 28). Similar to the mouse experiments, a decrease in UMI and an increase in mitochondrial RNA ratio were observed due to H&E staining and thin sections; however, single-nuclear transcripts obtained from the FFPE and H&E sections were consistently co-identified by cell type (Figs. 28m and 28n). In addition, the recovered UMI and the number of cells per cell population in each experimental group showed similar trends overall (Figs. 28s and 28t). This suggests that FX-seq can recover high-quality single-nuclear level transcripts in FFPE and H&E sections of human surgical specimens and can be extended to various tissues (Fig. 28u). Furthermore, it indicates that by using different reverse transcription primers, it is possible to extract information on the full-length RNA as well as transcriptome information at the 3' end (Fig. 29).

[0306] Large-scale single-nuclear transcriptome analysis using FX-seq and transcriptome pathway analysis associated with TKI resistance

[0307] Large-scale transcriptome analysis at the single-cell level identifies a more diverse range of cell types, deepening the understanding of cellular heterogeneity and biological phenomena, and enabling the identification of rare cell types. To evaluate the scalability of FX-seq, FFPE blocks were analyzed from gastrointestinal stromal tumors (GISTs) that underwent palliative surgery due to tumor recurrence despite responding to tyrosine kinase inhibitor (imatinib) treatment for 5 years. This tumor originated in the small intestine and measured approximately 1.5 cm x 0.8 cm x 1 cm (Figs. 32a, 33a, and 33b).

[0308] In a single experiment, approximately 4 million nuclei were isolated from 1 / 10 of a homogenized GIST FFPE block, and 199,293 nuclei were analyzed using an average of about 4,600 sequencing reads per nucleus. The sequencing data were processed through data quality and replicator analysis, generating a total of 171,992 single-nuclear transcripts with an average UMI of 1,543.7 (Figs. 32b to 32f). Transcriptome analysis identified various cell types, including rare immune cell populations such as common dendritic cells (cDCs) and plasma cell dendritic cells (pDCs) (Fig. 32g).

[0309] The largest cell cluster was identified as tumor cells strongly expressing ANO1, a specific genetic marker of interstitial cells of Cajal (ICC), the progenitor cells of GIST tumors (Fig. 34a). It was confirmed that these tumor cells exhibited exclusive expression of the oncogenes KIT and PDGFRA (Fig. 34a), suggesting that the expression levels of these genes change independently during the development from ICC to GIST tumors. In particular, PDGFRA with high PDGFRA expression highOnly the population expressed cell proliferation markers such as CENPF, TOP2A, and MKI67 (Fig. 34b). In the results of trajectory analysis reconstructed using pseudotime trajectory or Partition-based graph abstraction (PAGA) analysis for the analysis of connectivity between single-cell transcripts (Figs. 32h and 32i), non-proliferative KIT from normal ICC [expressing MYH11, MYLK, and TPM1 (Fig. 33c)] high PDGFRA that proliferates through cells high Two pathways for conversion into cells were discovered.

[0310] We hypothesized that these two pathways represent transcriptome information capable of explaining the resistance mechanism caused by the KIT mutation (p.N822K), a representative secondary mutation in imatinib-resistant cancer cells, and presumed that cells proliferate along these two pathways. In pathway 1, the transcriptome showed transient increases in the expression of DNA methylation-related genes, including PCNA, DNMT1, and UHRF1, and sustained upregulation of the EZH2 histone methylase (Fig. 32j). Additionally, transient expression of several genes, including ALK, ATAD2, and MYO3A, was observed (Figs. 32k and 35).

[0311] These transient gene expression phenomena contribute to imatinib resistance along with epigenetic modifications, and proliferating PDGFRA represented by the overexpression of DNMT3B, CIT, and DIAPH3 highA hypothesis was proposed that this leads to altered transcriptome patterns in the cells. In the case of Pathway 2, similar overexpression of EZH2 was observed without changes in DNA methylation-related genes (Fig. 32l). In particular, in the non-proliferative tumor population (Cluster 9 in Fig. 32l), gene expression of GDF15, PURPL, PLEKHG1, GFRA1, and SYNDIG1 was observed to increase transiently before decreasing at the end of Pathway 2. In proliferative cells, abnormal gene expression was observed, including germline-specific gene expression such as SPAG1, SPTG4, ADCY10, and CCNG2, along with cell proliferation markers (Figs. 32m and 36).

[0312] In summary, two pathways involving abnormal gene expression patterns and transient changes in gene expression provide important insights into cellular responses to imatinib and the acquisition of resistance. Pathway 1 involves changes in gene expression related to DNA methylation, which is consistent with previous findings regarding abnormal DNA methylation associated with imatinib resistance in chronic myeloid leukemia (CML). Pathway 2 is characterized by the expression of germline-specific genes that are not typically expressed in other tumors and normal cells, and is considered a rare cell population with novel pathways associated with imatinib resistance in large-scale snRNA-seq analysis. Overall, we demonstrated the high scalability and data quality of FX-seq analysis on clinical FFPE samples and confirmed that it offers opportunities to understand the dynamic characteristics of integrated cellular processes associated with human diseases and to discover novel molecular markers.

[0313] Possibility of Integrating Pathological Diagnosis and FX-seq

[0314] Finally, FX-seq was used to analyze single-nuclear transcripts in human colorectal cancer (CRC) FFPE slides stored for approximately two months. Based on pathological annotations, tumor regions (T) and non-tumor regions (NT) were physically separated (Figs. 37a to 37j), and a sequencing library was constructed by separating nuclei through FX processing and barcoding using region-specific reverse transcription primers to distinguish pathological regions (see Fig. 38a, Examples 20 and 21). This allowed for the application of FX-seq to track the spatial origin of pathologically labeled regions.

[0315] Sequencing results identified cell types typically present in CRC, such as immune cells, endothelial cells, intestinal glial cells, and epithelial cells (Figs. 38b and 38c). Additionally, differences in the spatial distribution of cell types could be confirmed through barcodes labeling tumor / non-tumor regions within each cluster (Figs. 38d and 38e), and large epithelial lineage clusters distinct from normal epithelial cells could be classified into epithelial progenitor cells and tumor cells (Figs. 38f and 38g). This was cross-validated using molecular criteria of loss of PIGR expression and increase in CDK8 expression (Figs. 38c and 39), and distinct transcriptomic differences between progenitor cells and tumor cells were confirmed through further detailed cluster analysis (Fig. 38h).

[0316] Additionally, the inferCNV package was applied to the analyzed transcriptome information to analyze copy number variation (CNV) in epithelial progenitor cells and tumor cells relative to normal epithelial cells. In tumors, amplification of Chr7p, Chr8q, Chr13q, and Chr20p and a decrease in Chr1p were observed, whereas changes in CNV were minimal in epithelial progenitor cells (Figs. 38i and 40). This is consistent with previous CRC scRNA-seq studies and demonstrates that FX-seq can generate high-quality data from clinical samples.

[0317] FX-seq provides insights into the complex tumor microenvironment (TME) of CRC tissue. Further sub-clustering analysis of the fibroblast population revealed two distinct subgroups of activated cancer-associated fibroblasts (CAFs) acting as key regulators in the TME (Fig. 38j). Analysis of fibroblast subtypes based on the spatial distribution separating tumor and surrounding tissues distinguished CAF cells biased toward the tumor region, as well as fibroblasts (NMFs) and their progenitor cells evenly distributed across both tumor and non-tumor regions (Figs. 38k and 38l). CAF cells were further subdivided into CAF-A and CAF-B based on transcriptome status, and genes specific to each group were identified (Fig. 38m). CAF-A expresses genes such as FAP, PDPN, COL1A2, and MMP2, while CAF-B expresses genes such as ACTA2, TAGLN, and PDGFRA; cluster-specific gene analysis showed a strong correlation with previous studies.

[0318] Furthermore, the analysis revealed additional marker genes more specific to each cluster, providing further insight into the heterogeneity of the CAF population. In conclusion, FX-seq demonstrated that it is a useful tool for linking surgical pathological diagnoses with relevant transcriptomes and for investigating cellular heterogeneity in the TME in greater depth.

[0319] Additionally, based on transcriptome information and CNVs of tumor cells observed in colorectal cancer tissue, the patient's tumor cells were further classified as intrinsic-consensus molecular subtypes 2 (iCMS2), as reported in previous single-cell transcriptome analysis studies on colorectal cancer tissue. Simultaneously, increased expression of LGR5 was observed in the tumor population, supporting the possibility that this occurred through expansion from LGR5+ crypt-base stem cells (Fig. 39). iCMS2 is reported to be primarily microsatellite stable (MSS) and does not possess mutations in KRAS, which is consistent with the MSS and KRAS wild-type diagnostic results confirmed during the pathological diagnosis of the sample. Furthermore, the detection of a CAF population in the tumor microenvironment (TME) suggests the presence of a fibrotic environment within the tumor. In summary, the tumors in this study were classified as iCMS2_MSS_F according to the proposed IMF classification using intrinsic epithelial subtype (I), microsatellite instability (M), and fibrosis (F).

[0320] These analysis results demonstrated the potential of FX-seq as a diagnostic tool by analyzing single-nuclear transcriptomes in FFPE sections. This method can identify cancer subtypes requiring information at the single-cell or single-nuclear resolution level, thereby eliminating the need for additional surgical procedures to obtain fresh specimens. FX-seq can provide new opportunities for clinical diagnosis by utilizing FFPE sections generated from routine histology. In particular, it is possible to analyze desired hypotheses by establishing cohorts with FFPE blocks from various patient groups (Fig. 41).

[0321] Cell or tissue pretreatment kit

[0322] Another aspect of the present invention provides a cell or tissue sample pretreatment kit comprising an RNase inhibitor, an organic catalyst, and a crosslinker.

[0323] At this time, the kit may be for the pretreatment of formalin-fixed or formalin-fixed paraffin-embedded (FFPE) cells or tissues. Additionally, the kit may further include a cell separation kit. Additionally, the RNase inhibitor may be PVSA.

[0324] In addition, the organic catalyst may be 2-aminophenylboronic acid hydrochloride. Also, the crosslinker is as described above.

[0325] New Crosslinker

[0326] Another aspect of the present invention provides a compound having the structural formula of the following chemical formula II.

[0327] [Chemical Formula II]

[0328]

[0329] In the above chemical formula II, Z 1a , Z 2a , Z 1b and Z 2b Each is independently O, OH, S, SH, NH, NH2, N(C 1-6 alkyl) or NH(C 1-6 It is an alkyl). In some embodiments, Z 1a , Z 2a , Z 1b and Z 2b Each can independently be OH, SH, or NH2.

[0330] In some embodiments, Z 1a and Z 2a They may be identical or different from each other. In some embodiments, Z 1b and Z 2b They may be identical or different from each other. In some embodiments, Z 1a and Z 1b They may be identical or different from each other. In some embodiments, Z 2a and Z 2b They may be the same or different from each other.

[0331] In the above chemical formula II, R 1a and R 1b is independently F, Cl, Br, or I. In some embodiments, R 1a and R 1b Each can independently be Cl or Br. In some embodiments, R 1a and R 1b They may be the same or different from each other.

[0332] In the above chemical formula II, Q a and Q b Each independently directly bonds, O, S, NH, NH2, N(C 1-6 alkyl) or N(C 1-6 It is alkyl)2. In some embodiments, Q a and Q b Each can independently be directly bonded, O, S, NH, or NH2. In some embodiments, Q a and Q b They may be the same or different from each other.

[0333] In the above chemical formula II, L is C optionally interrupted by NH, O, or S. 2-40 It is an alkylene. In some embodiments, L is C optionally interrupted by NH, O, or S. 10-40 It may be an alkylene. In some embodiments, L is C optionally interrupted by NH, O, or S. 20-30 It can be an alkylene. For example, L is C 10-40 Alkylene; C 20-30 Alkylene; C interrupted by NH, O, or S 10-40 Alkylene; or C interrupted by NH, O, or S 20-30 It can be an alkylene.

[0334] In the above chemical formula II, R 4a and R 4b Each independently halo; hydroxy; C 1-6 Alkoxy; Amino; C 1-6 Alkylamino; di(C 1-6Alkyl)amino; nitro; cyano; carboxy; C 1-6 Alkyl carbonyl; carbamoyl; C 1-6 Alkylcarbamoyl; di(C 1-6 Alkyl)carbamoyl; or halo, hydroxy, C 1-6 Alkoxy, amino, C 1-6 alkylamino, di(C 1-6 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-6 It is an alkyl. In some embodiments, R 4a and R 4b Each independently halo; hydroxy; C 1-4 Alkoxy; Amino; C 1-4 Alkylamino; di(C 1-4 Alkyl)amino; nitro; cyano; carboxy; C 1-4 Alkyl carbonyl; carbamoyl; C 1-4 Alkylcarbamoyl; di(C 1-4 Alkyl)carbamoyl; or halo, hydroxy, C 1-4 Alkoxy, amino, C 1-4 alkylamino, di(C 1-4 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-4 It may be alkyl. In some embodiments, R 4a and R 4b They may be the same or different from each other.

[0335] In the above formula II, l and m are each independently integers from 1 to 3. In some embodiments, l and m may each independently be 2 or 3. In some embodiments, l and m may be the same or different from each other.

[0336] In the above formula II, o and p are each independently integers from 0 to 2. For example, o and p may each independently be 0 or 1. In some embodiments, o and p may be the same or different from each other.

[0337] In some embodiments, the compound of Formula II may have a chemical structure symmetric with respect to L.

[0338] In some embodiments, the crosslinker may be a compound represented by the following chemical formula II-1.

[0339] [Chemical Formula II-1]

[0340]

[0341] In the above chemical formula II-1, Z 1a , Z 2a , Z 1b , Z 2b , R 1a , R 1b , Q a , Q b , L, R 4a , R 4b , o and p are as defined for Chemical Formula II.

[0342] In some embodiments, the crosslinker may be a compound represented by the following chemical formula IIA.

[0343] [Chemical Formula IIA]

[0344]

[0345] In the above chemical formula IIA, R 1a , R 1b , Q a , Q b , L, R 4a , R 4b , o and p are as defined for Chemical Formula II.

[0346] In the above chemical formula IIA, R 2a , R 3a , R 2b and R 3b Each independently H or C 1-6 It may be alkyl. In some embodiments, R 2a , R 3a , R 2b and R 3b Each independently H or C1-4 It can be an alkyl.

[0347] In some embodiments, R 2a and R 3a They may be identical or different from each other. In some embodiments, R 2b and R 3b They may be identical or different from each other. In some embodiments, R 2a and R 2b They may be identical or different from each other. In some embodiments, R 3a and R 3b They may be the same or different from each other.

[0348] In some embodiments, the crosslinker may be a compound represented by the following chemical formula IIA-1.

[0349] [Chemical Formula IIA-1]

[0350]

[0351] In the above chemical formula IIA-1, R 1a , R 1b , Q a , Q b , L, R 2a , R 3a , R 2b , R 3b , R 4a , R 4b , o and p are as defined for the chemical formula IIA.

[0352] In some embodiments, the crosslinker may be a compound represented by the following chemical formula IIB.

[0353] [Chemical Formula IIB]

[0354]

[0355] In the above chemical formula IIB, R 1a , R 1b , Q a , Q b , R 2a , R 3a , R 2b , R3b , R 4a , R 4b , o and p are as defined for the chemical formula IIA.

[0356] In the above chemical formula IIB, R 5a and R 5b H, H2, and C are independently 1-6 alkyl or (C 1-6 It may be alkyl)2. In some embodiments, R 5a and R 5b H, H2, and C are independently 1-6 alkyl or (C 1-6 It may be alkyl)2. In some embodiments, R 5a and R 5b They may be the same or different from each other.

[0357] In the above chemical formula IIB, L a and L b C each independently 1-6 It may be an alkylene. In some embodiments, L a and L b C each independently 1-4 It may be an alkylene. In some embodiments, L a and L b They may be identical or different from each other. For example, L a and L b Each can be methylene, ethylene, propylene, or butylene.

[0358] In the above chemical formula IIB, Q 1 can be a direct bond, NH, O, or S. In some embodiments, Q 1 It can be NH, O, or S.

[0359] In the above formula IIB, n may be an integer from 1 to 20. In some embodiments, n may be an integer from 5 to 15.

[0360] In some embodiments, the crosslinker may be a compound represented by the following chemical formula IIB-1.

[0361] [Chemical Formula IIB-1]

[0362]

[0363] In the above chemical formula IIB-1, R 1a , R 1b , R 2a , R 3a , R 2b , R 3b R 4a , R 4b , R 5a , R 5b , o, p and n are as defined for the chemical formula IIB.

[0364] In some embodiments, the crosslinker may include a compound represented by the following chemical formula 1.

[0365] [Chemical Formula 1]

[0366]

[0367] In some embodiments, the crosslinker can be manufactured according to the following reaction formula I.

[0368] [Reaction Equation I]

[0369]

[0370] In the above reaction equation I, Z 1a , Z 2a , Z 1b , Z 2b , R 1a , R 1b , Q a , Q b , L, R 4a , R 4b , o, p, l and m are as defined for the above chemical formula II.

[0371] In the above reaction scheme I, X is a halo. For example, X can be F, Cl, Br, or I. In some embodiments, X is R1a and R 1b It may be the same or different from each other.

[0372] In the above reaction equation I, Q c and Q d Depending on the reaction, each Q a and Q b It can be a reactive functional group that forms. For example, Q c and Q d Each can independently be an amino, hydroxy, or mercapto. In some embodiments, Q c and Q d They may be the same or different from each other.

[0373] In some embodiments, a crosslinker having a left-right symmetric chemical structure can be prepared according to the following reaction scheme II.

[0374] [Reaction Equation II]

[0375]

[0376] In the above reaction equation I, Z 1a , Z 2a , R 1a , Q a , L, R 4a , o, l, X and Q c is as defined for the above reaction equation II.

[0377] definition

[0378] In this specification, the term "halogen" or "halogen atom" refers to an atom belonging to Group 17 of the periodic table. Halogen atoms include F, Cl, Br, I, etc. The term "halo" refers to a halogen substituent.

[0379] The term "alkyl" refers to a fully saturated branched or unbranched (or straight-chain or linear) hydrocarbon. The alkyl may be a substituted or unsubstituted alkyl. The C 1-20 Alkyl groups are, for example, C 1-15 , C 1-10 , or C 1-6It may be an alkyl. The above C 1-6 Alkyl is C 1-5 , C 1-4 , C 1-3 , or C 1-2 The alkyl may be an alkyl. The alkyl may be, for example, methyl, ethyl, n-propyl, isopropyl, n-butyl, isobutyl, sec-butyl, n-pentyl, isopentyl, neopentyl, iso-amyl, or n-hexyl.

[0380] The term "haloalkyl" refers to an alkyl group substituted with one or more halogens, and may include a perhaloalkyl group in which all hydrogens of the alkyl group are substituted with halogens.

[0381] The term "hydroxy" refers to the -OH functional group (hydroxyl group).

[0382] The term "mercapto" refers to an -SH functional group.

[0383] The term "carbonyl" refers to -C(=O)-.

[0384] The term "alkoxy" refers to an alkyl group bonded to an oxygen atom. The above C 1-20 Alkoxy, for example, C 1-15 , C 1-10 , or C 1-6 It may be an alkoxy. The above C 1-6 Alkoxy is C 1-5 , C 1-4 , C 1-3 , or C 1-2 It may be an alkoxy. The alkoxy may be methoxy, ethoxy, propoxy, butoxy, etc.

[0385] The term "amino" refers to -NH2.

[0386] The term "alkylamine" refers to an amine in which one H of an amino (-NH2) is substituted with an alkyl group.

[0387] The term "di(alkyl)amine" refers to an amine in which both H groups of an amino (-NH2) are substituted with alkyl groups. The two alkyl groups in a di(alkyl)amine may be the same or different.

[0388] The term "nitro" refers to -NO2.

[0389] The term "cyano" refers to a functional group formed by a triple bond between a carbon atom and a nitrogen atom, denoted as -CN.

[0390] The term "carboxy" refers to -COOH. A carboxyl salt refers to the conjugate base of a carboxylic acid.

[0391] The term "alkoxycarbonyl" refers to a monovalent substituent in which the -OH group of a carboxyl group is replaced by an alkoxy group. For example, C 1-6 Alkoxycarbonyl is C 1-6 It refers to -C(=O)- substituted with an alkoxy.

[0392] The term "carbamoyl" refers to -CONH2.

[0393] The term "alkylcarbamoyl" refers to a substituent in which one hydrogen atom of -NH2 in a carbamoyl is substituted with an alkyl group.

[0394] The term "dialkylcarbamoyl" refers to a substituent in which two hydrogen atoms of the -NH2 group in a carbamoyl are each substituted with an alkyl group. In a dialkylcarbamoyl, the two alkyl groups may be the same or different.

[0395] At this time, the crosslinker can selectively bind to the guanine-N7 atom without interfering with Watson-Crick base pair formation.

[0396] Tissue fixation composition

[0397] Another aspect of the present invention provides a composition for fixing cells or tissues comprising PVSA and formaldehyde. The PVSA is as described above.

[0398] Composition for exchanging fixative

[0399] Another aspect of the present invention provides a composition for fixative exchange comprising an organic catalyst and PVSA. In this case, the organic catalyst and PVSA are as described above.

[0400] Composition for re-tablet

[0401] Another aspect of the present invention provides a resuspended solution comprising a buffer for resuspending and buffering action and PVSA. The buffer may be one of Tris, citrate, phosphate, HEPES, MOPS, or carbonate buffer.

[0402] In addition, the present invention may be compatible with other commercial single-cell / nuclear library production platforms such as 10X Chromium (Fig. 45). Therefore, FX-seq is expected to enable efficient single-nuclear transcriptome analysis on a large scale in PFA-fixed or human clinical FFPE samples and various animal models, and has potential applications not only in single-nuclear transcriptome analysis but also in proteomics, epigenetics, chromatin accessibility analysis (ATAC-seq), spatial transcriptomics, and spatial proteomics.

[0403] The present invention will be explained in more detail below through examples. However, these examples are intended to illustrate one or more specific embodiments, and the scope of the present invention is not limited to these examples.

[0404] I. Preparation

[0405] Preparation Example 1. HeLa cell culture

[0406] HeLa cells were cultured in Dulbecco's modified Eagle's medium (DMEM, Gibco) supplemented with 10% heat-inactivated fetal bovine serum (HI-FBS, Gibco) and 1% penicillin-streptomycin (Gibco). Cells were cultured at 37°C under 5% CO₂ conditions and subcultured three times per week.

[0407] Preparation Example 2. Animal Experiment

[0408] All animal handling and experiments were conducted in accordance with the guidelines of the Yonsei University Institutional Animal Care and Use Committee (IACUC). Eight-week-old C57BL / 6N wild-type mice were purchased and housed in an animal facility under strict SPF conditions. The mice were sacrificed between 8 and 10 weeks of age.

[0409] Preparation Example 3. Research on Human Subjects

[0410] All human patient samples used in this study were approved by the Institutional Review Board (IRB) of Yonsei University Severance Hospital and were obtained through pre-operative written consent.

[0411] Preparation Example 1. Expression and purification of Tn5

[0412] The pTXB1-Tn5 plasmid was donated by Rickard Sandberg (addgene_60240). The expression and purification of Tn5 were performed using a slightly modified version of Sandberg's method. The expression plasmid pTXB1-Tn5 was transformed into chemically competent cells (C3013, NEB) with an ampicillin selection marker. To prepare 500 mL of cell culture, 250 mL of LB medium (MP Biomedicals) was added to each of two 1 L flasks, and 5 mL of cells cultured overnight were inoculated into each medium.

[0413] Cells were cultured at 37°C with shaking at 250 rpm in LB medium (MP Biomedicals) and ampicillin until the A600 (optical density) value reached the range of 0.5–0.9. Subsequently, the culture medium was cooled in a refrigerator for 30 minutes, and IPTG (SG bio) at a final concentration of 0.25 mM was added. After additional overnight incubation at 25°C at 250 rpm, the culture medium reached an A600 value of 3.0. The culture medium was stored at -80°C until the next step after harvest.

[0414] For protein extraction and purification, the pellet obtained from 500 mL of cultured cells was resuspended in 80 mL of HEGX buffer (20 mM HEPES-NaOH, pH 7.2, 800 mM NaCl, 1 mM EDTA, 10% glycerol, 0.1% Triton X-100) with the addition of complete protease inhibitors (Roche). Subsequently, sonication and polyethyleneimine (PEI) precipitation were performed according to Sandberg's method. The purified lysate was filtered through a 0.45 μm filter (Millex), and 10 mL of chitin resin slurry (NEB) was added to the filtrate.

[0415] This mixture was transferred to an Econo-Pack chromatography column (Bio-Rad) and washed by gravity with 200 mL of HEGX buffer (20 times the resin volume). The mixture was divided into two 50 mL Falcon tubes and spun at 10 rpm for 1 hour in a cold environment to allow intein-tagged Tn5 to bind to the chitin resin. Sandberg's method was performed to remove the intein tag. 20 mL of elution buffer (100 mM DTT in HEGX buffer) was added to the top of the column, and 5 mL of buffer was passed through the column.

[0416] Subsequently, the column was sealed and incubated at 4°C for 48–72 hours to release Tn5 through ligation. Tn5 elution was performed in 3 mL aliquots, and each fraction was verified by SDS-PAGE. The fraction containing Tn5 was concentrated to 5 mL using a Mohr Amicon Ultra-15 centrifuge filter (Millipore). The samples underwent two buffer exchanges with 1 L of 2X concentration Tn5 dialysis buffer (100 mM HEPES-KOH, pH 7.2, 800 mM NaCl, 0.2 mM EDTA, 2 mM DTT, 10% glycerol). For long-term storage at -80°C, Triton X-100 was removed from the solution, and the NaCl concentration was increased to 800 mM. The final protein concentration was measured using Pierce 660 nm protein assay reagent (Thermo Fisher). Tn5 transposomes were assembled according to Sandberg's method using Tn5-MEDS B / B oligonucleotides. Approximately 1 mg of purified Tn5 transposase was obtained from a single 500 mL culture, and it could be stored for a long period at -80°C and maintained its activity even after 2 years of storage.

[0417] Analysis Method 1. Sequencing Read Preprocessing and Gene Expression Matrix Generation

[0418] Fastq files generated by sequencing were processed with minor modifications based on the sci-RNA-seq3 analysis pipeline. Briefly, generated reads were annotated to the corresponding cell type if they met the criterion of Levenshtein edit distance (ED) < 2 based on reverse transcription and ligation indices. Individual reads were clipped to the AAAAAAAA sequence and default settings using trim_galore v0.6.4. The clipped reads were mapped to the GRCm39 reference genome for mice and the GRCh38.p13 reference genome for humans using the default settings of STAR v2.7.9a123.

[0419] Gene annotations were performed using GENCODE V42 for humans and GENCODE VM27 for mice. Duplicate reads were filtered using the unique molecular identifier (UMI) sequence, reverse transcription index, ligation adapter index, and the last coordinate of the read 2 mapping. Gene expression matrices for individual cells were generated using the HTSeq package by assigning mapped regions to the intron or exon regions of each gene. In the case of multiple mapped reads, they were assigned to the gene closest to the end of the mapped read; however, if another gene existed within 100 bp of the closest gene, that read was deleted.

[0420] Analysis Method 2. Processing and Visualization of Single Nuclear RNA Sequencing Data

[0421] The gene expression matrix was generated using the Python package scanpy v1.9.0 based on the number of genes identified during the preprocessing stage. Subsequent analyses, excluding details on analysis metrics, are available in Table 1. Cells that did not meet specific quality control criteria, such as the number of UMIs, the number of detected genes, and the mitochondrial read ratio, were excluded from subsequent analyses. After merging the initial gene matrices, only genes containing protein-coding sequences, pseudogenes, and long non-coding RNAs were used for further analysis. Genes with no counts in the entire dataset were excluded from the analysis. Doublets, formed by the merging of two or more cell transcripts, were identified using Scrublets with various parameters and filtered using manual thresholds. The cell expression matrix was normalized by the average UMI of all cells, a fictitious count of 1 was added, and the matrix was log-transformed.

[0422] The data was scaled to have a mean of 0 and unit variance. The dimensionality-reduced dataset, obtained using Principal Component Analysis (PCA), was applied to Naver Graph computation, cells were clustered using the Leiden algorithm, and embedded into UMAP with optimal parameters for each experiment. Genes specific to each cluster were identified by using the scanpy.tl.rank_genes_groups function and filtering using parameters set for each experiment. Clusters were assigned to specific cells based on marker genes identified according to molecular criteria from previous studies.

[0423] For mouse brain cell type annotation, single-cell and single-nuclear transcriptome atlases and marker genes from previous studies were referenced. For human clinical sample cell type annotation, the Human Protein Atlas was referenced to identify cell types specific to marker genes in each tissue. Clusters containing marker genes from different clusters simultaneously were further filtered as potential dichotomous populations. For subcluster analysis, the original gene expression matrices corresponding to cells in each subcluster were re-analyzed starting from the gene filtering step. The Partition-based Graph Abstraction (PAGA) algorithm was applied to construct pathways between the identified clusters, and cells were aligned according to diffusion pseudotime.

[0424]

[0425] Analysis Method 3. CNV Inference

[0426] To infer CNVs from the gene expression matrix, inferCNV v1.3.3 was used in Figure 35 to infer CNV status from epithelial lineage transcripts. Original gene expression matrices from cells with a UMI of 200 or higher were utilized in inferCNV. Considering the low number of reads per cell, cells with an average read count of less than 0.001 per gene were filtered out. CNVs were inferred as three states—neutral, lost, and gained—using the HMM3 method, and were analyzed at the sample level in Figure 35 and at the individual cell level in Figure 40.

[0427] II. Synthesis of Novel Crosslinking Agents

[0428] Example 1. Synthesis of crosslinking agent

[0429]

[0430] All reactions were carried out under dry argon or nitrogen air conditions, and the reaction temperature was controlled using a heating mantle (Misung Scientific).

[0431] C18 flash chromatography was performed using a CombiFlash Rf 150 purification system (Teledyne) and a C18 Purichem flash column (pore size: 120 Å, particle size: 50 μm, cartridge mass: 26 g, Purichem). Protons ( 1 H) Nuclear magnetic resonance (NMR) spectra were recorded in MeOD-d4 solution (Sigma Aldrich) using a Bruker Avance III HD 300 MHz instrument.

[0432] 1 1H NMR data is expressed as chemical shift (δ ppm) [Multiplicity, coupling constant (Hz), integration], and mass spectra were acquired using InfinityLab LC / MSD and 1260 series LC.

[0433] Dichloro(ethylenediamine)platinum(II) [Pt(en)Cl2, 100 mg, 0.31 mmol, Sigma-Aldrich] and dimethylformamide (DMF, Sigma-Aldrich) were added to a round-bottom flask containing a magnetic stirring bar. Silver nitric acid (AgNO3, 58 mg, 0.31 mmol, Sigma-Aldrich) was dissolved in 6 mL of dimethylformaldehyde (DMF) and added dropwise, and the mixture was stirred for 16 hours under dark room temperature conditions. The reaction mixture was filtered through a Millex-GV membrane filter (pore size: 0.22 μm, Millipore) to remove silver chloride (AgCl).

[0434] A portion of NH2-PEG11-NH2 (65 mg, 0.12 mmol, Broadpharm) was added and stirred overnight at 50°C in the dark. After the reaction was complete, the mixture was concentrated under reduced pressure. The residue was dissolved in 3 mL of Milli-Q water and stored overnight at 4°C in the dark. The mixture was filtered through a Millex-GV membrane filter to remove insoluble yellow particles [Pt(en)Cl(NO3)]. The filtered solution was concentrated under reduced pressure and purified by C18 flash column chromatography (a method in which the concentration of 0.1% formic acid and acetonitrile was gradually increased from 0% to 100% over 10 minutes in water and acetonitrile). The solvent of the separated fraction was removed under reduced pressure, and the residue was dissolved in 3 mL of methanol. This solution was filtered again through a Millex-GV membrane filter to remove the remaining [Pt(en)Cl(NO3)], and finally, yellow oil (119 mg, yield: 88.2%, Figure 15a) was obtained.

[0435] 1 H-NMR (300 MHz, MeOD-d4): δ = 3.97-3.68 (m, 52H), 2.98-2.56 (m, 14H). LC-MS(ESI+) m / z [M-2H] 2+ : 562.4

[0436] III. Nucleic Acid Analysis of PFA-Fixed or FFPE Tissues Using Fixative Exchange

[0437] Example 2. Confirmation of RNA reverse transcription inhibition by PFA

[0438] Template DNA was obtained by PCR reaction in which the sequences of the PHF11 or SET7 / 9 genes were inserted into the pcDNA3 vector. Subsequently, the process was performed using the Hi-Scribe T7 High Yield RNA Synthesis Kit (NEB) according to the manufacturer's protocol. In vitro transcribed RNA was purified and eluted using a Monarch RNA column according to the manufacturer's protocol.

[0439] Purified single PHF11 and SET7 / 9 RNA (50 ng / μL) was fixed by shaking at 37°C over time using 4% PFA dissolved in 1X PBS. The fixed RNA was purified via column purification and quantified using the Qubit RNA BR kit (Invitrogen). The reverse transcription inhibitory effect of PFA was evaluated through reverse transcription and qPCR quantification.

[0440]

[0441] PCR primer sequence Sequence number pcDNA FCCA CTG CTT ACT GGC TTA TCG3pcNDA RGCT GGC AAC TAG AAG GCA CA4

[0442] Reverse transcription of immobilized RNA was performed as follows: 2 μL of 0.1 ng / μL immobilized RNA, 2 μL of 5X Maxima H Minus reaction buffer, 0.25 μL of Maxima H Minus reverse transcriptase, 0.25 μL of SUPERase-In, 1 μL of 10 μM reverse primer, 0.5 μL of 10 mM dNTP, and 2 μL of nuclease-free water. The concentration of reverse-transcribed cDNA was quantified according to the qPCR protocol described below.

[0443] Primer sequence SEQ ID NO: PHF11 3' FAGA CTA TGA AGA AAT CGG GAG TGC5PHF11 3' RCTC TTC CAA CTG CTG CC6SET7 / 9 3' FAGT TCT CCA GGG CAC GTA TG7SET7 / 9 3' RGGC TTC CTC CAT CTG GGT AAT8

[0444] Example 3. Removal of RNA cross-linking under in vitro conditions

[0445] PHF11 RNA transcribed in vitro was fixed at room temperature for 16 hours using 4% PFA dissolved in 1X PBS at a concentration of 200 ng / μL. The fixed RNA was purified by column purification according to the manufacturer's protocol and quantified using the Qubit RNA BR kit (Invitrogen). The degree of cross-link removal was evaluated after treatment with an organic catalyst solution (10 ng / μL fixed RNA, 10 mM Tris pH 7.0, 10 mM Cat.1 or Cat.2, 0.1% v / v SUPERase-In) at 55°C for 30 minutes.

[0446] The treated RNA was purified by adding an equal volume of SPRIselect (Beckman Coulter) and eluted according to the manufacturer's protocol. After quantification, the purified RNA was reverse transcribed as follows: 2 μL of 0.1 ng / μL PHF11 RNA, 2 μL of 5X Maxima H Minus reaction buffer, 0.25 μL of Maxima H Minus reverse transcriptase, 0.25 μL of SUPERase-In, 1 μL of 10 μM reverse primer, 0.5 μL of 10 mM dNTP, and 2 μL of nuclease-free water. The reverse-transcribed cDNA was quantified using the qPCR protocol.

[0447] Example 4. Removal of PFA cross-linking from HeLa cell total RNA

[0448] Total RNA was extracted from freshly cultured HeLa cells using the column purification method according to the manufacturer's protocol. The extracted fresh total RNA was fixed at room temperature for 16 hours using 200 ng / μL of purified total RNA and 4% PFA dissolved in 1X PBS. The fixed RNA was purified via column purification according to the manufacturer's protocol and quantified using the Qubit RNA BR kit (Invitrogen). To check the degree of cross-link removal, the fixed RNA was treated with an organic catalyst solution (10 ng / μL fixed RNA, 10 mM Tris pH 7.0, 10 mM Cat.1 or Cat.2, 0.1% v / v SUPERase-In) at 55°C for 30 minutes.

[0449] The treated RNA was purified by adding an equal volume of SPRIselect (Beckman Coulter) and eluted according to the manufacturer's protocol. After quantification, the purified RNA was reverse transcribed as follows: 2 μL of 0.1 ng / μL total RNA, 2 μL of 5X Maxima H Minus reaction buffer, 0.25 μL of Maxima H Minus reverse transcriptase, 0.25 μL of SUPERase-In, 1 μL of 10 μM oligo dT primer, 0.5 μL of 10 mM dNTP, and 2 μL of nuclease-free water. The reverse-transcribed cDNA was quantified using the qPCR protocol described below.

[0450] Primer sequence Sequence number Oligo dTTTT TTT TAA GCA GTG GTA TCA ACG CAG AGT ACG TTT TTT TTT TTT TTT TTT TTV9Oligo dT RAGC AGT GGT ATC AAC GCA GA10GAPDH FGAG AAG GCT GGG GCT CAT TT11KAT5 FTGT AAA GTA GAA GTT GGG GGT GG12UPF FCGC ACT GTA CCA AGG CAA TG13CDKN1A FCCA AAC ACC TTC CAG CTC CT14PRKACA FGCG ATT CAA CCT GTG TGC TG15RXRB FGGC TGC ATG ATT TTT GCC CT16

[0451] Example 5. Additional RNA cross-linking

[0452] RNA transcribed in vitro was shaken at 37°C for 15 minutes using 50 ng / μL of purified PHF11 RNA. The reaction solution contained 1X MOPS (1X PBS for PFA crosslinking) and 1-8% PFA or 0.5X to 4X of the Pt crosslinker synthesized in Example 1. Here, the 1X crosslinker corresponded to 58.7 μM in all experiments. After the reaction was complete, each sample was immediately placed on ice to dilute, and the degree of crosslinking was evaluated using an Agilent 4200 Tapestation. The reverse transcription reaction was prepared with the following solution: 2 μL of 0.1 ng / μL crosslinked PHF11 RNA, 4 μL of 5X SuperScript buffer, 0.5 μL of SuperScript reverse transcriptase, 0.5 μL of SUPERase-In, 1 μL of 0.1 M DTT, 2 μL of 10 μM reverse primer, 1 μL of 10 mM dNTP, and 9 μL of nuclease-free water. The reaction solution was incubated at 55°C for 30 minutes, and subsequently, the degree of inhibition of each crosslinker was quantified by qPCR.

[0453] Primer sequence SEQ ID NO: PHF11 3' FAGA CTA TGA AGA AAT CGG GAG TGC TGC17PHF11 3' RCTC TTC CAA CTG CTG CC6PHF11 mid FTTT CGG GTG GCT GTA CAT GA18PHF11 mid RCCA AGA AGG ACG ACG CAG T19PHF11 5' FCTC CTT CCC ACC GGT GTC TT20PHF11 5' RTCC TCA CAT TCC ACA AGT CCT G21

[0454] Example 6. HeLa cell line experiment

[0455] For HeLa cell line experiments, 1.4 to 2 million cells were collected, centrifuged at 200 g for 2 minutes at room temperature, washed once with 1X PBS (Gibco), and fixed with 4% PFA dissolved in 1X PBS at 4°C for 15 minutes. Subsequently, the cells were centrifuged at 200 g for 3 minutes at 4°C and washed with washing buffer (1X PBS, 0.025% Tween-20). The washed cells were permeabilized with PBST solution (1X PBS, 0.05% TritonX-100, 0.4 mg / mL BSA) at room temperature for 15 minutes. Afterward, the cells were washed with washing buffer.

[0456] Fixed cells were filtered over ice using a 40 μm cell filter (pluriStrainer) with a wide-bore pipette tip, and after counting the cells, they were diluted to a final concentration of 5 k cells / μL. The filtered cells were incubated twice in a CL(-) solution (1X MOPS, 0.025% Tween-20) for 5 minutes in a refrigerator. Subsequently, a CL(+) solution (1X MOPS, 0.025% Tween-20, 0.2X Cross-linker) was added and the mixture was shaken at 25°C for 15 minutes.

[0457] The cross-linker concentration was optimized according to the sample type, as reaction rates differ significantly between cell lines lacking ECM and tissues rich in ECM. For example, a cross-linker concentration of 0.2X was effective in cell line experiments, whereas treatment with a cross-linker concentration of 2X was effective in mouse brain FFPE sections (the reference for a 1X concentration is approximately 58.7 μM). The effective reaction concentration also varies depending on the batch size of the synthesized Pt cross-linker and the temperature and time of the additional cross-linking reaction; typically, concentrations ranging from 0.01X to 10X can be used. The effective cross-linker concentration for each tissue is listed in each example. Cross-linker-treated cells were centrifuged at 200 g, resuspended in wash buffer, and counted. Cells were dispensed according to experimental conditions to achieve a final concentration of 500 cells / μL. Subsequently, the cells were washed with CT(-) solution (200 mM Tris pH 8.3) and shaken at 55°C at various time intervals with CT(+) solution (200 mM Tris pH 8.3, 50 mM Cat.2). Afterward, the samples were immediately placed on ice and centrifuged at 200 g.

[0458] A portion of the supernatant was taken, and RNA concentration was measured using the Qubit RNA BR assay (Invitrogen) according to the manufacturer's manual. The remaining supernatant was discarded, and the cell pellet was resuspended in PBSRI solution (1X PBS, 0.4 mg / mL BSA, 0.1% v / v SUPERase-In) and counted. Cells were reverse transcribed at 55°C for 30 minutes as follows: 11 μL of 4,300 cells / μL, 4 μL of 5X SuperScript buffer, 1 μL of SuperScript reverse transcriptase, 1 μL of SUPERase-In, 1 μL of 10 μM reverse primer, 1 μL of 10 mM dNTP, and 1 μL of 100 mM DTT. The concentration of the reverse-transcribed cDNA was quantified according to the qPCR protocol described below.

[0459] Primer sequence Sequence number Oligo dTTTT TTT TAA GCA GTG GTA TCA ACG CAG AGT ACG TTT TTT TTT TTT TTT TTT TTV9Oligo dT RAGC AGT GGT ATC AAC GCA GA10GAPDH FGAG AAG GCT GGG GCT CAT TT11B2M FTGA GTA TGC CTG CCG TGT GA22B2M RATC TTC AAA CCT CCA TGA TGC T23CNOT FGTC CAA AAC CTG ACT GCA TGT ATC24CNOT RGGT GTT TAC CCG CCT GCA T25

[0460] Example 7. qPCR Screening

[0461] In this study, experimental and sample conditions were optimized through qPCR screening. Cell lines, whole tissues, and tissue sections underwent cross-linking, reverse transcription, nuclear separation, and washing as described above. For in situ experiments, an equal volume of 2X lysis mixture (20 mM Tris pH 8.0, 400 mM NaCl, 100 mM EDTA pH 8.0, 4.4% SDS, 3.34 mg / mL Proteinase K) was added to reverse-transcribed nuclei, followed by gentle suspension. The lysis reaction was performed at 55°C for 30 minutes and at 85°C for 15 minutes, and the reverse-transcribed cDNA was purified and eluted using SPRIselect. The lysis step was omitted in in vitro experiments.

[0462] For the qPCR reaction mixture, 10 μL of TOPreal SYBR Green qPCR UDG PreMIX, 1 μL of 10 μM reverse primer, 1 μL of 10 μM forward primer, 2 μL of eluted cDNA, and 6 μL of nuclease-free water were mixed, and a total of 18 μL was dispensed into each well.

[0463] For technical accuracy, all conditions in the plate were repeated three times. The qPCR reaction was performed using the ViiA 7 Real-Time PCR System with the following cycle settings: an initial denaturation step of 15 minutes at 95°C, an amplification step of 40 cycles of 10 seconds at 95°C, 15 seconds at 60°C, and 30 seconds at 72°C (15 seconds at 95°C and 1 minute at 60°C for PowerUP SYBR Green Master Mix), a final extension step of 10 minutes at 72°C, and a melting curve measurement step of 15 seconds at 95°C, 1 minute at 60°C, and 30 seconds at 95°C.

[0464] Primer sequence Sequence number CNOT_FGTC CAA AAC CTG ACT GCA TGT ATC24CNOT_RGGT GTT TAC CCG CCT GCA T25B2M_FTGA GTA TGC CTG CCG TGT GA22B2M_RATC TTC AAA CCT CCA TGA TGC T23GAPDH_FGAG AAG GCT GGG GCT CAT TT11Gapdh_FACT GAG CAA GAG AGG CCC TA26Slc17a7_FGGG AGG AGA GGG TTG TTC CT27Gad1_FAGA CCT CCG ATA CAC TGA CCA28oligodT_RAGC AGT GGT ATC AAC GCA GA10

[0465] Example 8. PFA fixation of mouse brain

[0466] To obtain PFA-fixed mouse brain tissue, 8- to 10-week-old mice were anesthetized (oxygen 0.8 L / min, isoflurane 3%). Subsequently, 0.8 L / min of oxygen and 2% isoflurane were used to maintain anesthesia. The mice were perfused cardiacly with cold 1X PBS (Gibco), followed by additional perfusion with a solution of cold 4% PFA dissolved in 1X PBS. After perfusion was complete, the heads of the mice were decapitated to collect the whole brains. The detached brains were placed in a 1X PBS, 4% PFA solution and fixed by shaking in a refrigerator for 48 hours.

[0467] Example 9. FFPE embedding of mouse brain

[0468] Mice aged 8-10 weeks were euthanized in a CO2 chamber, immediately decapitated to extract the brains, and placed in a 4% PFA (EMS) solution (1X PBS) for 48 hours in a cold room with shaking fixation. The solution was replaced after 2-3 hours to maintain uniform fixation quality. After 48 hours, the sufficiently fixed brains were placed in an embedding cassette, and FFPE block fabrication was carried out according to the following protocol: 70% ethanol for 1 hour (1 time), 80% ethanol for 1 hour (1 time), 95% ethanol for 1 hour (2 times), 100% ethanol for 1 hour (3 times), xylene for 1 hour (3 times), and heated (56-58°C) paraffin wax for 1 hour (3 times), after which they were finally embedded in a paraffin block.

[0469] Leica ASP 300S (Leica) was used for tissue processing, and Leica Arcadia C / H (Leica) was used for paraffin embedding.

[0470] Example 10. FFPE embedding of human surgical samples

[0471] Freshly obtained human tumor samples were placed in 4% PFA (EMS) solution (1X PBS) as soon as possible after acquisition and fixed in a cold room for 48 hours. The solution was replaced after 2-3 hours of fixation to maintain uniform tissue quality. FFPE blocks were fabricated by processing the sufficiently fixed human clinical samples in the same manner as the mouse brain tissue FFPE embedding protocol.

[0472] Example 11. FFPE block segmentation

[0473] In the present invention, FFPE blocks were sectioned to thicknesses of 4 μm, 10 μm, or 30 μm. The blocks were trimmed horizontally using a Leica RM2255 microtome (Leica), and the flattened blocks were sectioned to the desired thickness onto silane-coated microslides (Muto) or stored directly in laboratory tubes. The sectioned slides were refrigerated with silica gel to maintain sample quality.

[0474] Example 12. H&E dyeing

[0475] 4 μm and 10 μm FFPE sections were used for H&E staining. Staining was performed using an Autostainer XL (Leica) or CV5030 (Leica) according to the following protocol: xylene 7 min 3 times, 100% ethanol 1 min 2 times, 95% ethanol 1 min 2 times, 1 min water wash, staining in Harris's hematoxylin solution for 5 min, 1 min water wash, differentiation in 1% HCl for 3 seconds, 30 second water wash, 3 second bluing in 0.3% ammonia solution, counterstaining in eosin Y for 20–60 seconds, 95% alcohol 30 second washes 2 times, dehydration in 100% ethanol for 1 min 3 times, and xylene clearing for 2 min 3 times.

[0476] Example 13. Homogenization, deparaffinization, and rehydration of FFPE samples

[0477] The FFPE block was removed from the embedding cassette, placed on a 150 x 20 mm Petri dish, and the paraffin wax outside the tissue was removed with a razor blade, taking care not to remove the tissue. The shaved block was finely ground with a razor blade, and the cut tissue was placed in a 2 mL tube along with 5 mm stainless steel beads (Qiagen) and 1.8 mL of xylene (Duksan) or limonene (Sigma-Aldrich). The mixture was homogenized in a chemical hood at 30 Hz for 4 minutes using a Tissue Lyser II (Qiagen) instrument.

[0478] The homogenized tissue solution was pelleted at 1,000 g for 3 minutes at room temperature, the supernatant was removed, and the pellet was resuspended in xylene and reacted at room temperature for 5 minutes, after which the same process was repeated once more. Deparaffinization and rehydration were performed by rotating at 15 rpm at room temperature. Subsequently, the tissue was resuspended in pre-cooled 100% ethanol and reacted at room temperature for 5 minutes, and the same process was repeated once, followed by one reaction each in 90% and 70% ethanol. From the 90% ethanol stage onwards, the solution was pelleted at 1,000 g for 3 minutes in a bucket centrifuge maintained at 4°C. The rehydrated tissue was pelleted again by centrifugation at 4°C (1,000 g, 3 min), suspended in BCL solution (3X SSC, 3% PVSA, 0.025% Tween-20), and passed through a 100 μm cell strainer followed by a 20 μm cell strainer if necessary.

[0479] Filtered tissues were washed twice with BCL solution and used as a preliminary step before further cross-linking. For 4 or 10 μm FFPE sections, the tissue attached to the slide was scraped off with a razor blade to maintain a “scroll” shape, and the collected FFPE section scrolls were deparaffinized and rehydrated in the same manner as above. H&E stained sections were also manually scraped off the slides. Since they were already deparaffinized during the staining process, only rehydration was performed, and no further homogenization was carried out. Prior to the experiment, when using FFPE or H&E sections, 1-2 slides were processed with FX-seq to preliminarily evaluate nuclear yield and RNA quality.

[0480] Example 14. Homogenization of strongly fixed tissue

[0481] Similar to FFPE block homogenization, tissues strongly fixed with PFA without paraffin embedding were homogenized using a Tissue Lyser II (Qiagen) instrument with a cold HM solution (10 mM Tris pH 7.0, 3% PVSA, 21 mM MgCl2, 1 mM CaCl2) instead of xylene or limonene. The HM solution was a modified version of the conventional tissue degradation method with added PVSA; the homogenized tissue was filtered through a 100 μm cell strainer and then subjected to secondary filtration through a 20 μm cell strainer. The filtered tissue was suspended in the HM solution and pelleted at 1,000 g for 3 minutes in a bucket centrifuge maintained at 4°C, repeating this process a total of three times. The tissues were then used immediately for subsequent protocols or flash-frozen at -80°C for storage until further use.

[0482] Example 15. Fixative Exchange

[0483] Homogenized and washed tissue samples were permeated by centrifuging at 3 rpm for 15 minutes at room temperature in PBST3 solution (1X PBS, 0.2% TritonX-100, 3% PVSA). Subsequently, the samples were pelleted at 1,000 g for 3 minutes in a bucket centrifuge maintained at 4°C, the pellet was recovered, and treated in a CL(-) solution (1X MOPS, 3% PVSA, 0.025% Tween-20) in a cold room for 5 minutes, after which the same process was repeated once more. A CL(+) solution (1X MOPS, 3% PVSA, 0.025% Tween-20, Cross-linker) was added, and the RNA was cross-linked by shaking at 25°C for 15 minutes. Afterward, the samples were pelleted at 1,000 g for 3 minutes in a bucket centrifuge maintained at 4°C, and the supernatant was removed. The recovered pellet was washed once with CT(-) solution (200 mM Tris pH 8.3, 3% PVSA), then pelleted again at 1,000 g for 3 minutes (4°C), carefully resuspended in CT(+) solution (200 mM Tris pH 8.3, 50 mM Cat.2, 3% PVSA, 8% w / v PEG-8000), and shaken at 55°C for 30 minutes to remove PFA crosslinks. After removing PFA, the tissue was pelleted again at 1,000 g for 3 minutes in a bucket centrifuge maintained at 4°C, the supernatant was removed, and the process proceeded to the subsequent step for nucleation.

[0484] Example 16. Nuclear separation

[0485] The pelleted FX-treated tissue was washed once with DG(-) solution (50 mM Tris pH 7.0, 10 mM EDTA pH 8.0, 10 mM NaCl, 1% PVSA, 0.2% Tween-20). Then, the tissue was pelleted at 1,000 g for 3 minutes in a bucket centrifuge maintained at 4°C, the supernatant was removed, and the tissue was resuspended in DG(+) solution (50 mM Tris pH 7.0, 10 mM EDTA pH 8.0, 10 mM NaCl, 1% PVSA, 0.2% Tween-20, 0.1 mg / mL Proteinase K) and shaken at 37°C for 20 minutes.

[0486] Digestion conditions were optimized as the strength of the ECM varied depending on the tissue type. For example, all human tissues analyzed in this study were digested in DG(++) solution (50 mM Tris pH 7.0, 10 mM EDTA pH 8.0, 10 mM NaCl, 1% PVSA, 1% Tween-20, 0.1 mg / mL Proteinase K) at 37°C with shaking for 20 to 40 minutes. After digestion, the tissues were pelleted at 1,000 g for 3 minutes in a bucket centrifuge at 4°C, the supernatant was removed, and the digested pellets were resuspended in a sucrose solution (1X PBS, 1% PVSA, 1.5 M sucrose) and then pelleted by centrifuging at 3,000 g for 20 minutes in a bucket centrifuge at 4°C.

[0487] For fragment samples, the samples were briefly filtered through a 40 μm cell strainer prior to sucrose concentration separation. After removing the supernatant while taking care not to disturb the pellet, the pelleted nuclei were resuspended in PBST1 solution (1X PBS, 0.2% TritonX-100, 1% PVSA) and subjected to secondary permeabilization by rotating at 3 rpm for 15 minutes at room temperature. However, for fragment samples, this secondary permeabilization step was omitted as excessive TritonX-100 treatment can lead to severe RNA leakage from physically severed nuclei. The secondary permeabilization step was also omitted for pelleted nuclei from FFPE blocks.

[0488] The recovered nuclei were suspended in BR solution (5X SSC, 0.025% Tween-20, 0.1% v / v SUPERase-In), and the washing process was repeated a total of three times by pelleting at 1,000 g for 3 minutes in a bucket centrifuge at 4°C. After the final wash, the nuclei were resuspended in PBSRI (1X PBS, 0.04% w / v BSA, 1% v / v SUPERase-In); PVSA was not used in this step as it inhibits the reverse transcription reaction. The prepared nuclei were stained with DAPI (Invitrogen) and counted using a hematocytometer and an EVOS imaging system (Life Technologies).

[0489] Example 17. Preparation of biological 3-replicate experiments from PFA-fixed or FFPE mouse brains

[0490] To verify the efficacy of FX-seq treatment in PFA-fixed or FFPE mouse brains, brain tissues were collected from three litterm mice. Each brain tissue was treated individually in parallel up to the reverse transcription step.

[0491] The control group did not undergo additional cross-linking or organic catalyst treatment prior to enzymatic nuclear separation. The (-) Cat.2 control group was not additionally cross-linked but was treated in (-) Cat.2 solution (200 mM Tris pH 8.3, 3% PVSA, 8% w / v PEG-8000) at 55°C for 30 minutes. The Cat.2 group was treated with an organic catalyst without additional cross-linking.

[0492] The FX group was sequentially treated with additional cross-linking and organic catalysts prior to enzymatic digestion. Each brain of the biological triplicate was divided into four groups (untreated control, (-) Cat.2, Cat.2, and FX), labeled with separate barcodes during sci-RNA-seq3 reverse transcription, and subsequently processed together. Each experimental group and brain was identified based on the reverse transcription barcode during single-cell transcriptome analysis.

[0493] Example 18. Fresh extraction using FX-seq and comparative experiment of weakly fixed brains and strongly fixed brains

[0494] Tissues were prepared according to the Easy-sci RNA-seq protocol to compare the performance of FX-seq with fresh samples. Fresh mouse brains were collected after euthanasia, the cerebellum was removed, and the brains were cut into small pieces in chilled buffer (1X PBS, 1% DEPC). After centrifugation at 200 g for 5 minutes, 1 mL of EZ lysis buffer (Sigma-Aldrich) and 10 μL of DEPC were added to the pellet and gently mixed 10 times. The mixture was placed on ice for 5 minutes and then processed through a 40 μm cell filter (Pluristrainer).

[0495] The tissue remaining on the filter was treated by adding 500 μL of EZ lysis buffer and 0.1% SUPERase-In (Invitrogen). The treated tissue was centrifuged at 500 g for 5 minutes, then resuspended in 1 mL of EZ lysis buffer and 0.1% SUPERase-In and pipetted three times. After centrifugation at 500 g for 5 minutes, the tissue was fixed on ice with 1X PBS and 0.1% PFA solution for 10 minutes. The nuclei were pelleted at 500 g for 3 minutes, resuspended in EZ lysis buffer, and the washing and pipetting process was repeated. Finally, the tissue was filtered using a 20 μm cell filter and washed with a washing solution (10 mM Tris pH 7.5, 10 mM NaCl, 3 mM MgCl2); the nuclei were then resuspended in PBSRI to perform reverse transcription.

[0496] Example 19. FX-seq of human colorectal cancer FFPE sections considering spatial context

[0497] For FX-seq analysis including spatial labeling of human colorectal cancer (CRC) FFPE sections, a pathologist first analyzed the H&E-stained sections to annotate the tumor and non-tumor regions. Based on these annotations, the cancerous region (T) and the peri-cancer region (NT) of adjacent 10 μm FFPE sections were physically scraped and separated using a blade. The separated sections underwent simultaneous deparaffinization and rehydration processes, followed by FX-seq processing. The detailed methods and subsequent protocols were referenced above.

[0498] Example 20. Reverse transcription and ligation

[0499] Reverse transcription of the combinatorial barcoding process for single-cell sequencing experiments was performed according to the protocol described below. In each well, 2 μL of 5X Maxima H Minus reaction buffer (Thermo Fisher Scientific), 0.5 μL of 200 U / μL Maxima H Minus Reverse Transcriptase, 0.5 μL of 20 U / μL SUPERase-In (Invitrogen), 2 μL of a mixture of 25 μM barcode oligo-dT primers and 2.5 mM dNTPs, 3 μL of 40% w / v PEG-8000 (Sigma-Aldrich), and 2 μL of nuclei prepared at 7.5k–10k nuclei / μL were mixed and briefly pulsed using a plate centrifuge maintained at 4°C. The plates were incubated at 55°C for 30 minutes to perform reverse transcription.

[0500] After reverse transcription, 30 μL of 1X PBS (0.025% Tween-20) was added to each well, collected in a 5 mL tube, and centrifuged at 1,000 g for 3 minutes in a bucket centrifuge maintained at 4°C. The pelleted nuclei were suspended in 270 μL of 1X PBS (0.04% w / v BSA, 1% v / v SUPERase-In) per 96-well plate to prepare a nuclear suspension solution. The ligation reaction was performed on the suspended nuclei according to the following steps. 0.5 μL of Quick Ligase Enzyme (Quick Ligase), 5 μL of 5X Quick Ligation Buffer, 2 μL of 10 μM ligation oligonucleotide, and 2.5 μL of nuclear suspension solution were mixed in each well and reacted at 25°C for 10 minutes after pulse centrifugation.

[0501] After ligation, 30 μL of 1X PBS (0.025% Tween-20) was added to each well, collected in a 5 mL tube, and centrifuged at 1,000 g for 3 minutes. The pelleted nuclei were filtered using a Flowmi cell filter (SP Bel-Art), and after an additional washing step, the nuclei were counted and resuspended in 1X PBS (0.025% Tween-20) at a concentration of 200–500 nuclei / μL. The remaining nuclei were stored at -80°C. All oligonucleotide barcode sequences used in this study were taken from sci-RNA-seq3 (see Table S11 in Junyu Caoet.al., Nature, 566(7745):496-502, 2019).

[0502] Example 21. Library preparation and sequencing

[0503] 5 μL of the nuclear solution suspended after ligation was mixed with 3 μL of elution buffer (Qiagen), 1.33 μL of second strand synthesis buffer (NEB), and 0.66 μL of second strand synthesis enzyme mix (NEB). The mixture was reacted at 16°C for 1 hour. For the tagmentation step, 4 μL of 5X tagmentation buffer (50 mM TAPS-NaOH pH 8.5, 25 mM MgCl2, 50% DMF), 5 μL of nuclease-free water, and 1 μL of Tn5 enzyme were mixed over ice and reacted at 55°C for 7 minutes. Subsequently, 5 μL of 0.1% SDS (Sigma-Aldrich) was added, and the mixture was reacted at 55°C for 15 minutes to inactivate the enzyme.

[0504] 1x SPRIselect was added to each well, and washed and eluted to 16.5 μL according to the manufacturer's protocol. PCR amplification was performed by mixing 20 μL of 2X NEB Next High-Fidelity PCR Master Mix (NEB), 2 μL of 10 μM barcode Illumina P5 adapter, 2 μL of 10 μM barcode Illumina P7 adapter, and 16 μL of the eluted library. The amplification conditions consisted of 9 to 14 cycles of amplification steps: 5 minutes at 72°C, 30 seconds at 98°C, followed by 10 seconds at 98°C, 30 seconds at 66°C, and 1 minute at 72°C, followed by a final extension reaction for an additional 5 minutes at 72°C.

[0505] The amplified libraries were eluted using 0.8 times the volume of SPRIselect according to the manufacturer's protocol. The final library concentration was measured using the 1X dsDNA HS assay with a Qubit 4.0 fluorescence detector (Invitrogen). Library quality was verified using an Agilent 4200 Tapestation (Agilent) and HSD1000 screen tape (Agilent). The libraries were sequenced using a NextSeq 2000 instrument (Illumina) with P2 (2x100 bp) or P3 (2x150 bp) paired-end kits (Manufacturer specifications: P2 approximately 400 million reads, P3 approximately 1.2 billion reads). The number of detected cells and reads per cell can be found in Table 9.

[0506]

[0507] Experimental Result 1. Verification of FX Seq Efficacy

[0508] The components required for FX-seq were validated in an in vitro model (Figs. 1b and 1c). cDNA synthesis yields were evaluated using qPCR. PFA immobilization immediately inhibited the cDNA synthesis yield of RNA molecules (Fig. 1b). Cat.2 demonstrated better catalytic activity than previously reported Cat.1 in in vitro transcribed RNA and extracted total RNA from HeLa cells (Fig. 1c). The ΔCt values ​​for each experimental group were normalized to the unimmobilized RNA control. Data are expressed as mean ± SD. P-values ​​(NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001) were determined by Bonferroni's multiple comparison test following unpaired one-way ANOVA analysis.

[0509] Figure 1d measures the additional crosslinking effect by a regioselective crosslinking agent in HeLa cells. It was confirmed that treatment with Cat.2 under optimized conditions improved cDNA synthesis yield, but the prolonged reaction time caused RNA leakage. The degree of RNA leakage during product removal due to PFA fixation was measured by Qubit RNA concentration in the supernatant, and cDNA synthesis yield was measured by qPCR to determine ΔCt values. Data are expressed as mean ± SD. P values ​​(NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001) were determined by Bonferroni's multiple comparison test following unpaired one-way ANOVA analysis.

[0510] Figure 1e measured the amount of RNA leaked per 1,000 cells using Qubit. Additional crosslinking with a new crosslinking agent minimized RNA leakage, thereby improving RNA preservation. Data are expressed as mean ± SD. P values ​​(NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001) were determined by Bonferroni's multiple comparison test following unpaired one-way ANOVA analysis.

[0511] Figure 1f shows PVSA as an effective RNase inhibitor against FX-seq. PVSA is negatively charged and mimics the phosphate backbone of RNA, inhibiting RNase activity in situ through competitive inhibition.

[0512] Figure 1g shows the results of automated electrophoresis analysis when various RNase inhibitors were used during the process of extracting RNA from mouse brain tissue fixed with PFA. It was confirmed that PVSA has the highest compatibility with the FX-seq protocol, which includes appropriate heat treatment along with enzymatic degradation. It was confirmed that the optimized nuclear separation protocol using PVSA minimizes RNA degradation compared to other inhibitors. Data are expressed as normalized fluorescence units (FU) per unit of size (bp).

[0513] Figure 1h illustrates the optimized nuclear separation strategy and the experimental sequence for FX-seq. Physically homogenized tissues were degraded by proteases following treatment with a regioselective crosslinking agent and an organic catalyst to separate single-nucleus suspensions, which were then filtered by sucrose cushion centrifugation. The single nuclei were labeled with combinatorial barcoding using a slightly modified sci-RNA-seq3 protocol. The resulting libraries were sequenced on an NGS sequencer and analyzed.

[0514] Figures 1i to 1l show that snRNA-seq was performed on three Litermate mouse brains heavily fixed with PFA using four conditions (nuclei isolated without fixative exchange (No Treat Ctrl), catalytic heating in high concentration Tris buffer of FX ((-) Cat.2), heating with Cat.2 added (Cat.2), and heating with Cat.2 added after additional cross-linking treatment (FX-seq)). The individual components of FX-seq led to an increase in the number of detected genes (Figure 1j), UMI (Figure 1k), and gene diversity (Figure 1l).

[0515] Figures 1m through 1t show the results of UMAP clustering and cell type annotation for PFA-fixed control samples (n = 7497) and FX-seq (n = 8654) after overlapping nucleation. Visualization and annotation of individual UMAP (uniform manifold approximation and projection) (Figures 1m and 1n) and UMAP analysis of the merged gene expression matrix (Figures 1o and 1p) improved the analytical resolution of cell type classification after FX-seq. Cell distribution (Figures 1q and 1r) and marker gene specificity (Figures 1s and 1t) were visualized.

[0516] Experimental Result 2. Confirmation of Treatment Effects of Organic Catalysts and Effects According to Catalyst Type

[0517] Total RNA from cultured HeLa cells was extracted and fixed with PFA. During the treatment of the fixed RNA with an organic catalyst, the effects of the catalyst itself and the effects of the type of organic catalyst applied were verified through triplicate experiments. It was confirmed that the presence or absence of the catalyst during high-temperature treatment was directly related to the amount of gene detected. It was confirmed that the effect of the organic catalyst of the present invention (Cat.2) was superior to that of the previously reported organic catalyst (Cat.1). The qPCR Ct values ​​for each treatment condition were normalized to the control group that did not undergo treatment after fixation and converted into ΔCt values. The same experimental conditions as those in Fig. 1c were evaluated for a wider variety of genes.

[0518] Refer to the sections on HeLa cell culture, removal of PFA cross-linking from HeLa cell total RNA, and qPCR screening in the above examples. Among these, the addition of a catalyst and the type of catalyst were varied under the organic catalyst treatment conditions. Each condition was evaluated using qPCR Ct values ​​for human GAPDH, CDKN1A, PRKACA, KAT5, UPF, and RXRB genes. Data are expressed as mean ± SD. P-values ​​(NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001) were determined by Bonferroni's multiple comparison test following unpaired one-way ANOVA analysis. The results are shown in Figure 3.

[0519] Experimental Result 3. Confirmation of Treatment Effect of Organic Catalyst and Detected Gene Amounts According to Catalyst Concentration Conditions

[0520] The effects of the catalyst itself and the effects of the catalyst concentration were confirmed during organic catalyst treatment on mouse brains fixed with PFA. It was confirmed that the amount of gene detected was enhanced through high-temperature treatment compared to the control group that did not undergo high-temperature treatment. Furthermore, it was confirmed that the presence or absence of the catalyst during high-temperature treatment enhanced the amount of gene detected, and that the amount of detected genes also increased as the catalyst concentration increased. The PFA fixation of mouse brains, homogenization of strongly fixed tissues, fixative exchange, nucleation, and qPCR screening sections of the above examples were referenced. Among these, under the organic catalyst treatment conditions, the addition of the catalyst and the concentration of the added catalyst were varied. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh, Slc17a7, and Gad1 genes. The results are shown in Figure 4.

[0521] Experimental Result 4. Verification of the treatment effect of the organic catalyst and the effect according to treatment temperature conditions

[0522] The effects of the catalyst itself and the effects of the catalyst treatment temperature were confirmed during organic catalyst treatment on mouse brains fixed with PFA. It was confirmed that more RNA was retained in the cell nuclei reacted with the addition of 10 mM Tris buffer and PVSA RNA degradase inhibitor compared to the control group without crosslinking agent exchange treatment. Furthermore, it was confirmed that the presence or absence of the catalyst during high-temperature treatment enhanced the amount of gene detected, and the amount of detected genes increased with increasing treatment temperature. The PFA fixation of mouse brains, homogenization of strongly fixed tissues, fixative exchange, nuclear separation, and qPCR screening sections of the above examples were referenced. Among these, the addition of the catalyst and the treatment temperature conditions were varied under the organic catalyst treatment conditions. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh, Slc17a7, and Gad1 genes. The results are shown in Figure 5.

[0523] Experimental Result 5. Verification of Treatment Effect of Organic Catalyst and According to pH Treatment Conditions

[0524] The effects of the catalyst itself and pH were verified during organic catalyst treatment on mouse brains fixed with PFA. It was confirmed that the presence or absence of the catalyst during high-temperature treatment is directly related to the amount of gene detected, and that the catalyst exhibits similar performance under pH conditions commonly used in biological experiments. The PFA fixation of mouse brains, homogenization of strongly fixed tissue, fixative exchange, nuclear separation, and qPCR screening sections of the above examples were referenced. Among these, the addition of the catalyst and the pH of the Tris buffer were varied under the organic catalyst treatment conditions. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh and Slc17a7 genes. The results are shown in Figure 6.

[0525] Experimental Result 6. Verification of Treatment Effect of Organic Catalyst and Performance of PEG Additive

[0526] The effects of the catalyst itself and the PEG additive were confirmed during organic catalyst treatment on mouse brains fixed with PFA. It was confirmed that the presence or absence of the catalyst during high-temperature treatment was directly related to the amount of gene detected, and the addition of PEG-8000 during treatment also enhanced the amount of gene detected. The PFA fixation of mouse brains, homogenization of strongly fixed tissue, fixative exchange, nucleation, and qPCR screening sections of the above examples were referenced. Among these, the organic catalyst treatment conditions were varied by the addition of the catalyst and the presence or absence of PEG. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh and Slc17a7 genes. The results are shown in Figure 7.

[0527] Experimental Result 7. Confirmation of treatment effects according to the presence or absence of organic catalyst treatment and time in PFA-fixed mouse brains

[0528] Each condition was evaluated using qPCR Ct values ​​for mouse Gapdh and Slc17a7 genes (Fig. 8a). As a result, a larger amount of the gene was detected in the experimental group with added organic catalyst compared to the group without, and the detected UMI also showed the same pattern when sequencing cell nuclei under the same conditions (Fig. 8b). This suggests that the detection sensitivity of qPCR corresponds to the sensitivity of the sequencing results, indicating that the sequencing results can be predicted from the qPCR experiment results. In the histograms of the detected UMI and the corresponding cells (Fig. 8c), it was also observed that more genes were detected when organic catalyst treatment was applied, and when the organic catalyst treatment time was longer at 30 minutes. The sections regarding PFA fixation of mouse brains, homogenization of strongly fixed tissues, fixative exchange, nuclear separation, qPCR screening, reverse transcription and ligation, library preparation, and sequencing in the above examples were referenced. Among these, the treatment with the organic catalyst and the treatment time were varied. The results are shown in Figure 8.

[0529] Experimental Result 8. Confirmation of the effect of PEG-8000 additive during reverse transcription

[0530] Single nuclei were isolated from mouse brains fixed with PFA after undergoing a crosslinking agent exchange process. The efficiency of reverse transcription was evaluated by adding a PEG additive to the isolated single nuclei during reverse transcription. PEG is generally known to enhance the efficiency of biological reactions, and this is known to be an effect caused by molecular clustering. Similarly, in fixed cell nuclei, the addition of PEG maximized the efficiency of reverse transcription, thereby increasing the sensitivity of the detected genes. In particular, unlike fresh or weakly fixed nuclei, fixed nuclei do not undergo rupture of the cell or nuclear membrane even at high concentrations of PEG, allowing for the specific use of high concentrations of PEG. The PFA fixation of mouse brains, homogenization of strongly fixed tissues, fixative exchange, nuclear isolation, and qPCR screening sections of the above example were referenced. Among these, optimization was performed by varying the concentration of the PEG-8000 substance used for reverse transcription. The qPCR Ct values ​​of the mouse Gapdh gene were normalized based on the condition without PEG addition and compared using the Log2 Fold Change value. The results are shown in Figure 9.

[0531] Experimental Result 9. Verification of the effect of additional crosslinking agents in cultured cells confirmed by Barnard sequencing experiment

[0532] The fixation effect of additional crosslinking agents was verified through Barnyard tests on human cells (HEK293T) and mouse cells (NIH / 3T3) weakly fixed with PFA (4°C, 15 min). Barnyard tests are typically used to evaluate the accuracy of sequencing techniques. In Barnyard tests, human and mouse cells are placed in different wells for reverse transcription, and then for the subsequent ligation reaction, they are reacted in the same well using the split-pool technique. When the experiment is conducted with normal cells, it is normal for only human or only mouse genes to be detected in the sequencing analysis, which identifies them as single nuclei.

[0533] Although no significant species mixing was observed in either the control group without additional cross-linking or the experimental group with additional cross-linking (Fig. 10a), significantly higher UMIs (Fig. 10b) and gene counts (Fig. 10c) were detected in both human and mouse cells through additional cross-linking. Refer to the sections on HeLa cell culture, HeLa cell line experiments, qPCR screening, reverse transcription and ligation, library preparation, and sequencing in the above examples. The results are shown in Fig. 10.

[0534] Experimental Result 10. Confirmation of qPCR results according to the PEG linker length of the additional crosslinking agent

[0535] The fixation effects of crosslinkers with different intermediate PEG linker lengths were compared in mouse brains fixed with PFA (Fig. 11a). It was confirmed that the additional crosslinker with a longer PEG chain length of 11 resulted in the detection of more genes than tissues with 3 PEG chains or those without additional fixation, and that treatment at a concentration of 0.1X was more effective than treatment at a concentration of 0.01X (Fig. 11b). It was determined that the longer the intermediate PEG linker, the better the efficiency of the crosslinking reaction and the better the retention of RNA molecules, which is a reasonable reason. The parts regarding PFA fixation of mouse brains, homogenization of strongly fixed tissues, fixative exchange, nuclear separation, qPCR screening, and crosslinker synthesis in the above examples were referenced. Among these, the type and concentration of the additional crosslinker were varied. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh gene. The results are shown in Fig. 11.

[0536] Experimental Result 11. Confirmation of differences in gene detection amounts according to treatment conditions of additional crosslinking agent by time and temperature.

[0537] The fixation effect of the additional crosslinker at high (55°C) or low (10°C) temperatures was verified in mouse brains fixed with PFA. The qPCR Ct values ​​were converted to ΔCt values ​​by normalizing them to the control group without each crosslinker exchange. In both short high-temperature treatments and relatively long low-temperature treatments, a larger amount of the gene was detected compared to the control group without additional crosslinker fixation, suggesting versatility in temperature selection for the additional crosslinking reaction. The PFA fixation of mouse brains, homogenization of strongly fixed tissues, fixative exchange, nuclear separation, and qPCR screening sections of the above examples were referenced. Among these, the concentration and duration of the additional crosslinker fixation were varied. Each condition was evaluated using the qPCR Ct values ​​of the mouse Gapdh gene. The results are shown in Figure 12.

[0538] Experimental Result 12. Confirmation of the effect of additional crosslinking agent according to organic catalyst treatment time

[0539] The fixation effect of the additional crosslinking agent was compared according to the organic catalyst treatment time in human cells (HEK293T, Fig. 13a) and mouse cells (NIH / 3T3, Fig. 13b) weakly immobilized with PFA (4°C, 15 min). It was confirmed that the amount of gene detected was enhanced depending on the presence or absence of the additional crosslinking agent, even at different organic catalyst treatment times. This suggests that the additional crosslinking agent can effectively prevent RNA leakage during high-temperature treatment. The procedure was conducted on HEK293T and NIH / 3T3 cells by referring to the sections on HeLa cell culture, HeLa cell line experiments, and qPCR screening in the above examples. Among these, the catalyst treatment time was varied under the organic catalyst treatment conditions. Each condition was evaluated through the qPCR Ct values ​​of human GAPDH, B2M, and CNOT genes, and mouse Gapdh, Slc17a7, and Gad1 genes. The results are shown in Fig. 13.

[0540] Experimental Result 13. Synthesis and Regioselectivity Confirmation of FX-seq Crosslinking Agent

[0541] Figure 14a shows the synthesis process of the crosslinking agent. Figures 14b and 14c show the synthesized regioselective crosslinking agent. 1 The H NMR spectrum (Fig. 14b) and its peak assignment (Fig. 14c) are shown. Fig. 14d shows the reaction of the Guanine-N7 regioselective crosslinker. The results are shown in Fig. 14.

[0542] Experimental Result 14. Confirmation that Guanine-N7 regioselective crosslinker minimizes inhibition of reverse transcription (RT) in vitro

[0543] Figures 15a through 15g show automated electrophoresis data of fixed RNA in an in vitro environment. Data are expressed in fluorescence units (FU) normalized per unit of size (bp). The experimental conditions are as follows: Figure 15a is a non-fixed RNA control, Figure 15b is 1% PFA, Figure 15c is 4% PFA, Figure 15d is 8% PFA, Figure 15e is 0.5X regioselective crosslinker (CL), Figure 15f is 2X CL, and Figure 15g is 4X CL. The 4% PFA and 2X CL concentrations are standard concentrations used for tissue fixation.

[0544] Figures 15h through 15j show bar graphs illustrating the relative inhibition of reverse transcription by reaction with PFA or CL at the corresponding concentrations. Due to the difficulty of RNA purification, the cross-linked RNA for each condition was appropriately diluted to avoid inhibition of RT and qPCR reactions. To measure the RT reaction without immobilization bias, the RNA ctrl (PFA) and RNA ctrl (CL) groups each contained an amount of diluted cross-linking agent equivalent to that of the experimental groups. The ΔCt values ​​for each experimental group were calculated through normalization relative to the control group. Three primer sets were designed to estimate the synthesis efficiency of the RT enzyme across the entire RNA region.

[0545] Specifically, the primers targeted the nucleotide sequences at the 3' end (Fig. 15h), intermediate region (Fig. 15i), and 5' end (Fig. 15j). While the cross-linking effect of CL showed a negligible reverse transcription inhibitory effect, PFA immobilization demonstrated a significant reverse transcription inhibitory effect. Although the inhibitory effect of the CL reaction accumulates as it reaches the 5' end, which requires a longer reaction by the RT enzyme, the maintenance effect on the RNA molecule is greater than the inhibitory effect of CL because FX-seq focuses on analyzing the nucleotide sequence at the 3' end. P-values ​​(NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001) were determined by Bonferroni's multiple comparison test following unpaired one-way ANOVA analysis. The results are shown in Fig. 15.

[0546] Experimental Result 15. Confirmation of the Efficacy of the FX-seq Single Nucleus Splitting Strategy in Rat Organs

[0547] Figure 16a shows a Brightfield image of a single nucleus isolated via FX treatment from mouse organs with high perfusion and PFA fixation, such as the heart, lung, liver, spleen, pancreas, kidney, and brain tissue. The experiment was performed by referring to the sections on PFA fixation of the mouse brain, homogenization of strongly fixed tissue, fixative exchange, and nucleus isolation in the above examples. Black and white: phase contrast, blue: DAPI. The results are shown in Figure 16.

[0548] Experimental Result 16. Optimization Evaluation of Deparaffinization and Rehydration of Stored Human FFPE Block Sections

[0549] Long-term stored human colorectal cancer FFPE blocks received from Severance Hospital, Yonsei University were sectioned to a thickness of 10 μm, and the conditions for deparaffinization and rehydration were optimized. The deparaffinization process typically uses limonene (Li.) or xylene (Xy.), both of which act as organic solvents to dissolve and remove the paraffin wax in which the tissue is embedded. At this time, the triple reaction and resuspension with xylene resulted in superior gene detection sensitivity compared to the triple reaction and resuspension with limonene (Fig. 18a).

[0550] In addition, the long reaction (5 min) with limonene yielded better results than the short reaction (2 min), and the two-reaction and resuspension with xylene yielded better results than the three-reaction and resuspension (Fig. 18b). In conclusion, it was confirmed that the two-reaction and resuspension with xylene retained the largest amount of gene not only for qPCR detection but also when measuring the concentration after library preparation for single-nuclear sequencing (Fig. 18c). Among the above examples, the FFPE embedding of mouse brain, FFPE block fragmentation, homogenization, deparaffinization, and rehydration of FFPE samples, fixative exchange, nuclear separation, qPCR screening, reverse transcription and ligation, library preparation, and sequencing were referenced. Among these, the deparaffinization and rehydration of FFPE samples were optimized. Each condition was evaluated using the qPCR Ct value of the human GAPDH gene. The results are shown in Fig. 18.

[0551] Experimental Result 17. Sequencing (FX-seq) results after crosslinking agent exchange confirmed in FFPE blocks and FFPE sections of mouse brains

[0552] The experiment was conducted by dividing the subjects into a control group, separated into single nuclei without crosslinking agent exchange, and an experimental group, separated into single nuclei after undergoing the crosslinking agent exchange step. Paraffin-embedded tissue of the whole mouse brain was used for the FFPE blocks, and FFPE sections were 10 μm thick. It was confirmed that more genes were detected through crosslinking agent exchange compared to the control group, whether the entire sample or thinly sectioned sections were used. The FFPE embedding of the mouse brain, FFPE block sectioning, homogenization, deparaffinization, and rehydration of the FFPE samples, fixative exchange, nuclear separation, and qPCR screening of the mouse brain in the above examples were referenced. Each condition was evaluated using the qPCR Ct values ​​of the mouse Gapdh and Slc17a7 genes. The results are shown in Fig. 19.

[0553] Experimental Result 18. Confirmation of the Efficacy of PVSA as an RNase Inhibitor

[0554] Figure 20a shows the results of automated electrophoresis of total RNA directly extracted from highly fixed mouse brain tissue without heat treatment. Data are expressed in fluorescence units (FU) normalized per unit of size (bp). No inhibitor (orange), 1% DEPC (red), 0.1% v / v SUPERase-In (blue), and 3% PVSA (green) were used as RNase inhibitors. As can be seen in Figure 1g, heat treatment enhanced RNase activity, resulting in severe RNA degradation in all groups except 3% PVSA. Therefore, PVSA was included in FX-seq to serve as an RNase inhibitor that is inexpensive, stable in high-temperature environments, and compatible with enzymatic reactions.

[0555] Figure 20b confirms the comparison of RNase inhibitors in HeLa cell lines using quantified cDNA synthesis yield. The ΔCt values ​​for each experimental group were calculated through normalization relative to the control group. Three gene sets were investigated: B2M (left), GAPDH (center), and CNOT (right). The results for all three genes confirmed the efficacy of PVSA during FX-seq treatment. DEPC is an effective reagent that blocks RNase enzymes via a reduction reaction, but it also reduced all other enzymes to an inactive state. Since FX-seq degrades tissues using enzymes, DEPC was concluded to be an incompatible RNase inhibitor in this context. P-values ​​(NS ≥ 0.05, *P≤0.05, **P≤0.01, ***P≤0.001) were determined by Bonferroni multiple comparison tests following unpaired one-way ANOVA analysis. The results are shown in Figure 20.

[0556] Experimental Result 19. In situ verification results of FX-seq catalyst in PFA-fixed mouse brain tissue

[0557] These are the experimental results measured by qPCR following in situ cDNA synthesis using nuclei extracted from perfusion-fixed mouse brain tissue. The ΔCt values ​​for each experimental group were calculated through normalization relative to the control group. Genes commonly expressed in mouse brains were targeted: Gapdh (left), Slc17a7 (center), and Gad1 (right).

[0558] In addition, a comparison of the reactivity of Cat.1 and Cat.2 confirmed that Cat.2 exhibited more efficient reactivity than Cat.1 even in an in situ environment (Fig. 21a). Furthermore, a comparison was made between heat treatment without a catalyst and heat treatment with Cat.2. While heat treatment in high-concentration Tris buffer increased cDNA synthesis efficiency in PFA-fixed mouse brain tissue, the addition of Cat.2 further increased reverse transcription efficiency (Fig. 21b). The time-dependent effects of Cat.2 treatment were examined (Fig. 21c). Long-term treatment with Cat.2 resulted in a plateau in cDNA synthesis efficiency, which is likely due to RNA molecule leakage, as previously observed in the HeLa cell line model.

[0559] Stepwise evaluation of the FX components was verified (Fig. 21d). Heating conditions of FX under high concentration Tris buffer conditions ((-) Cat. 2), heating with the addition of Cat. 2 (Cat. 2), and heating conditions with the addition of Cat. 2 after additional crosslinking treatment (FX-seq) were evaluated.

[0560] Data were expressed as mean ± SD. P-values ​​(NS ≥ 0.05, *P≤0.05 **P≤0.01 ***P≤0.001) were determined through Bonferroni's multiple comparison test following unpaired one-way ANOVA analysis.

[0561] Experimental Result 20. Confirmation of the auxiliary effect of PVSA in FX-seq

[0562] FFPE blocks of mouse brains stored for over one year were sectioned to a thickness of 30 μm, and experiments were conducted by dividing the subjects into groups based on the presence or absence of crosslinking agent exchange sequencing treatment and the use of PVSA, an RNA degradase inhibitor. Generally, concerns regarding transcript quality degradation caused by RNA degradases are relatively lower in fixed tissues compared to transcript analysis in fresh tissues. However, high-temperature treatment used to reverse RNA fixation is accompanied by the activation of RNA degradases within the tissue, which poses a problem regarding transcript quality degradation. To address this issue, FX-seq utilized PVSA throughout the entire process, including additional crosslinking reactions, the reversal of transcript fixation effects using organic catalysts, and proteolytic reactions for single nucleus separation. Consistent with verifications in other previous experiments, it was confirmed that the additional crosslinking reaction and the organic catalyst reaction enhanced the reverse transcription efficiency of transcripts compared to the control group.

[0563] Furthermore, in the experimental group in which PVSA was not added only during the additional crosslinking reaction and the organic catalytic reaction, the tissue was rehydrated and homogenized in a solution containing PVSA prior to the additional crosslinking reaction, leaving residual PVSA in the tissue. However, lower transcript preservation was confirmed with higher Ct values. In the control group, where only single nucleus isolation was performed without undergoing other reactions after tissue homogenization, it was confirmed that the presence or absence of PVSA significantly affects transcript preservation. Therefore, it was confirmed that each reaction in crosslinker exchange sequencing plays a critical role for high-sensitivity analysis, and that the utilization of PVSA in each reaction also plays a significant role in transcript preservation. Among the above examples, the FFPE embedding, FFPE block fragmentation, homogenization, deparaffinization, and rehydration of FFPE samples, fixative exchange, nucleus isolation, and qPCR screening of mouse brains were referenced. Each condition was evaluated through the qPCR Ct values ​​of the mouse Gapdh and Slc17a7 genes. The results are shown in Fig. 22.

[0564] Experimental Result 21. Confirmation of the effects of FX-seq using PFA-fixed mouse brains

[0565] Individual data of QC indices obtained from snRNA-seq under four conditions are shown (Figs. 23a to 23d): nuclei isolated without fixative exchange (No Treat Ctrl), uncatalyzed heating under high concentration Tris buffer conditions of FX ((-) Cat.2), heating with Cat.2 added (Cat.2), and heating with Cat.2 added after additional crosslinking treatment (FX-seq). Number of detected genes (Fig. 23a), number of UMIs per nucleus (Fig. 23b), detection rate of mitochondrial genes (Fig. 23c), and detection rate of rRNA (Fig. 23d). All analysis results yielded similar values ​​in biological copies.

[0566] In the UMAP analysis for the untreated control group, the labeled nuclei of individual replicates (Fig. 23e) and the distribution of the number of labeled cells in annotated clusters (Fig. 23f) were shown. Additionally, in the UMAP analysis for FX-seq, the labeled nuclei in each brain (Fig. 23g) and the distribution of the number of labeled cells in annotated clusters (Fig. 23h) were shown.

[0567] The distribution of UMI according to cell type identified in the untreated control group (Fig. 23i) and FX-seq (Fig. 23j) is shown in Figs. 23i and 23j. In addition, dot plots of marker genes specific to cell type identified in the untreated control group (Fig. 23k) and FX-seq (Fig. 23l) are shown in Figs. 23k and 23l.

[0568] Experimental Result 22. Comparison of snRNA-seq results between freshly extracted mouse brains and strongly fixed mouse brains treated with FX-seq.

[0569] This shows the snRNA-seq QC metrics of lightly fixed mouse brain cell nuclei (n = 11,807) extracted fresh and strongly fixed after perfusion (n = 7,569). These are the number of detected genes (Fig. 24a) and UMI (Fig. 24b). Both were found to be high in the freshly extracted samples. This represents the proportion of sequencing reads mapped to mitochondrial genes (Fig. 24c), rRNA (Fig. 24d), and exon regions (Fig. 24e). The high proportion of mitochondrial genes, rRNA, and exon regions indirectly indicates contamination of cytoplasmic contents due to limited tissue degradation in the fresh samples.

[0570] The results of UMAP analysis of each sample after the QC cutoff (UMI 200–40,000 and mitochondrial readings less than 1%) are shown (Figs. 24f and 24g). Individual clusters were annotated according to molecular criteria. In the merged clusters, clusters with the same annotation in the individual clustering were co-clustered, but slight differences in distribution were observed in UMAP for each experimental group (Figs. 24h and 24i).

[0571] Figure 24j shows a color map of the proportion mapped to exon regions in the UMAP analysis of the merged gene expression matrix. The change in distribution between fresh samples and FX samples in the merged matrix coincided with the distribution axis of the exon readout ratio.

[0572] Figure 24k shows the distribution of exon read rates of the top 10 genes identified by differential expression gene analysis between fresh and fixed samples of the same cell type. Genes abundant in fresh samples had high exon read rates, while genes abundant in fixed samples had low exon read rates.

[0573] This was confirmed by microscopic images of freshly extracted mouse brain cell nuclei and cell nuclei extracted by FX-seq from mouse brains that were strongly fixed after perfusion (Fig. 25). Additionally, microscopic images of nuclei isolated from fresh mouse brain tissue (Fig. 25a) and perfused, PFA-heavy fixed mouse brain tissue (Fig. 25b) are shown. Grayscale: Phase contrast, Blue: DAPI.

[0574] Experimental Result 23. Validation of FX-seq Components in an FFPE Mouse Brain Tissue Model

[0575] This shows that the brains of three littermates were extracted immediately after euthanasia, fixed with PFA, and embedded in a paraffin block (Fig. 26a). SnRNA-seq was performed after processing and separation using FX-seq in the FFPE block. The samples were tested under four conditions: separated nuclei without fixative exchange (No Treat Ctrl), uncatalyzed heating under high concentration FX Tris buffer conditions ((-) Cat.2), heating with Cat.2 added (Cat.2), and heating with Cat.2 added after additional cross-linking treatment (FX-seq).

[0576] This shows the QC metrics for FX-seq of FFPE mouse brain blocks. It was confirmed that individual components of FX-seq led to an increase in the number of detected genes (Fig. 26b), UMI (Fig. 26c), and gene diversity (Fig. 26d). Although various clusters could be identified in FFPE control samples through the optimized nuclear separation protocol, processing with FX-seq significantly improved the resolution of cell type classification.

[0577] The results of UMAP clustering and cell type annotation for FFPE control samples (n = 4,914) and FX-seq (n = 5,852) after overlapping nuclear removal are shown. Individual UMAP visualization and annotation (Figs. 26e and 26f) and UMAP analysis of the merged gene expression matrix (Figs. 26g and 26h) showed improved analytical resolution of cell type classification after FX-seq. The distribution of each cell type (Figs. 26i and 26j) and the specificity of marker genes (Figs. 26k and 26l) were visualized.

[0578] Experimental Result 24. Verification of individual data obtained from triple experiments of FX-seq using FFPE mouse brain blocks

[0579] Individual data of QC indices obtained from snRNA-seq under four conditions: nuclei isolated without fixative exchange (No Treat Ctrl), uncatalyzed heating under high concentration Tris buffer conditions of FX ((-) Cat. 2), heating with Cat. 2 added (Cat. 2), and heating with Cat. 2 added after additional cross-linking treatment (FX-seq). The number of detected genes (Fig. 27a), number of UMIs per nucleus (Fig. 27b), ratio of mapped reads for mitochondrial genes (Fig. 27c), and ratio of mapped reads for rRNA (Fig. 27d) were determined. All metrics yielded similar values ​​in the biological replicates.

[0580] In the UMAP analysis for the untreated control group, the labeled nuclei of individual copies (Fig. 27e) and the distribution of the number of labeled cells in annotated clusters (Fig. 27f) were examined. Additionally, in the UMAP analysis for FX-seq, the labeled nuclei of individual copies (Fig. 27g) and the distribution of the number of labeled cells in annotated clusters (Fig. 27h) were examined.

[0581] This shows the distribution of UMI according to cell type identified in the untreated control group (Fig. 27i) and FX-seq (Fig. 27j). In addition, dot plots were used to identify cell type-specific marker genes identified in the untreated control group (Fig. 27k) and FX-seq (Fig. 27l).

[0582] Experimental Result 25. Confirmation of the applicability of FX-seq to thin FFPE sections and H&E-stained sections

[0583] Compatibility testing between FX-seq and a clinical environment was performed. SnRNA-seq was conducted after isolating nuclei from mouse brain sections and human retinal metastatic cancer tissues, both H&E-stained and unstained sections. The results are shown in Figure 28.

[0584] This shows the analysis of FX-seq-treated and isolated 10 μm FFPE and H&E-stained sections of a mouse brain FFPE block (Fig. 28a). Fig. 28a shows the QC metrics of the FFPE sections (n ​​= 11,334) and H&E-stained sections (n ​​= 8,648). The number of detected genes (Fig. 28b), the number of UMIs per nucleus (Fig. 28c), the ratio of mapped reads for mitochondrial genes (Fig. 28d), and the ratio of mapped reads for rRNA were confirmed (Fig. 28e).

[0585] The results of UMAP analysis of FFPE sections and H&E stained sections are shown (Figs. 28f and 28g). In the UMAP analysis of the merged gene expression matrix, the same cell type was co-clustered even though they were separated under different conditions. Although each cluster was individually annotated, the analysis results showed that the gene expression patterns of the corresponding cell types were similar even under different conditions via FX-seq.

[0586] The cell type annotations and cell number distributions of each cluster were examined (Figs. 28h and 28i). In addition, the specific expression of representative marker genes of each cluster was examined (Figs. 28j and 28k).

[0587] FX-seq was applied to the nuclei of human-derived reticular metastatic cancer samples prepared from 4μm and 10μm FFPE and H&E stained sections (indicated as 4F, 10F, 4H, and 10H, respectively), and the analysis results are shown in Figures 28m to 28u (n = 8,590 in 4F, n = 10,395 in 10F, n = 14,658 in 4H, and n = 11,206 in 10H).

[0588] The results of UMAP analysis for the combined gene expression matrix under four conditions are shown (Fig. 28m). Additionally, in the UMAP (m) labeled according to experimental conditions, individual nuclei were distributed in a biased manner depending on the presence or absence of H&E staining (Fig. 28n).

[0589] Figure 28l shows the QC metrics of FX-seq. The number of detected genes (Figure 28o), the number of UMIs per nucleus (Figure 28p), the ratio of mapped reads to mitochondrial genes (Figure 28q), and the number of described UMIs according to top genes aligned with the number of detected UMIs in each condition (Figure 28r) were examined. 4F had fewer genes and UMIs than 10F due to physical subsampling, and H&E staining had a greater impact than cross-sectional thickness.

[0590] In addition, the distribution of UMI in each cell type labeled according to experimental conditions was examined (Fig. 28s). In addition, the distribution of cell numbers for each cell type was examined (Fig. 28t). Furthermore, marker genes specific to each cluster were plotted as a dot plot (Fig. 28u).

[0591] Experimental Result 26. Comparison of Effects of Crosslinking Agent Exchange Treatment and Reverse Transcription Primers on Metastatic Cancer Samples

[0592] Specifically, bladder cancer samples metastasized to the retina were compared before and after the application of FX-seq using either oligo(dT) primers or random primers. Transcriptome information was obtained by adding a reverse transcription primer with a random primer sequence (NNNNNN) to the existing oligo(dT)-based reverse transcription primers. Figure 29a shows the gene body coverage of cDNA synthesized using oligo(dT) and Figure 29b shows the gene body coverage of cDNA synthesized using random primers. It can be confirmed that when using random primers, the 5' end is read more than with oligo(dT), suggesting that the full length of the transcript can be analyzed even in fixed FFPE samples.

[0593] The number of UMIs in single-nuclear transcriptome results using oligo(dT) before and after the application of FX-seq (Fig. 29C) and the number of UMIs using random primers (Fig. 29D) were examined. In both cases, it was clearly evident that the UMIs increased (Fig. 29e). The cell types identified using additional random primers and the UMAP plots were presented. The sections regarding human subject studies, FFPE embedding of human surgical samples, FFPE block fragmentation, homogenization, deparaffinization, and rehydration of FFPE samples, fixative exchange, nuclear separation, reverse transcription and ligation, library preparation, and sequencing in the above examples were referenced. Among these, comparisons were made using barcoded random primers or dT-based primers during reverse transcription.

[0594] Experimental Result 27. Results of FX-seq analysis using mouse brain FFPE and H&E sections

[0595] The results are shown in Fig. 30. Microscopic images of cell nuclei extracted from mouse brain FFPE sections at 10x magnification (Fig. 30a) and 40x magnification (Fig. 30b) are shown. Black and white: phase contrast, blue: DAPI. Also shown are microscopic images of cell nuclei extracted from mouse brain H&E sections at 10x magnification (Fig. 30c) and 40x magnification (Fig. 30d). Black and white: phase contrast, blue: DAPI. Additionally, the results of individually clustered UMAP analysis of mouse brain FFPE sections (Fig. 30e) and mouse brain H&E sections (Fig. 30f) were examined. Furthermore, the number of UMIs by annotated cell type in mouse brain FFPE sections (Fig. 30g) and mouse brain H&E sections (Fig. 30h) in Fig. 28 are shown. In addition, the results of DEG analysis showing marker genes in the mouse brain FFPE section (Fig. 30i) and mouse brain H&E section (Fig. 30j) of Fig. 28 are shown.

[0596] Experimental Result 28. Confirmation of Morphological and RNA Quality Assessment of Oreticular Metastatic Cancer

[0597] The results are shown in Fig. 31. Specifically, microscopic images of 4 μm H&E-stained sections of reticular metastatic cancer FFPE samples were examined (Fig. 31a). Automated electrophoresis results of total RNA extracted from reticular metastatic cancer FFPE samples were graphed (Fig. 31b). Data are expressed as normalized fluorescence units (FU) per unit of size (bp).

[0598] Experimental Result 29. Verification of the Scalability of FX-seq to Human Surgical Specimens Confirmed by FX-seq Analysis of Gastrointestinal Stromal Tumor (GIST) FFPE Blocks

[0599] Figure 32a shows that surgically resected human GIST tissue was immediately immersed in 4% PFA, fixed at 4°C for 48 hours, and then embedded in an FFPE block. Nuclei were isolated from the FFPE block and combinatorially labeled using sci-RNA-seq3 for the preparation of an snRNA-seq library. Transcripts of 171,992 nuclei were analyzed after QC and removal of overlapping nuclei.

[0600] Specific marker genes for each cluster were plotted as a dot plot (Fig. 32b). Additionally, the QC metrics of FX-seq were shown in Fig. 32a. The number of detected genes (Fig. 32c), the number of UMIs per nucleus (Fig. 32d), the ratio of mapped reads for mitochondrial genes (Fig. 32e), and the ratio of mapped reads for rRNA (Fig. 32f) were examined. It was confirmed that the nuclei isolated from the FFPE block had low contamination by mitochondrial genes and rRNA.

[0601] The results of UMAP analysis on transcriptome distribution and clustering were examined. Various cell types were identified based on molecular criteria (Fig. 32g). A cluster network was constructed based on PAGA analysis (Fig. 32h) and a diffusion pseudotime heatmap (Fig. 32i).

[0602] The resolved transcriptome pathways and their corresponding gene expression patterns were shown. The resolved tumor pathway 1 (Fig. 32j) and the corresponding gene expression changes per cluster (Fig. 32k) were plotted. The resolved tumor pathway 2 (Fig. 32l) and the corresponding gene expression changes per cluster (Fig. 32m) were shown, and the blue frame image on the right confirmed the 20-fold magnified expression profile changes of pathway 2 in tumor sub-clusters 9 through 11.

[0603] Experimental Result 30. Confirmation of Morphological and RNA Quality Assessment of Gastrointestinal Stromal Tumors (GIST)

[0604] Microscopic images of 4 μm H&E stained sections of the GIST FFPE sample were examined (Fig. 33a). Automated electrophoresis results of total RNA extracted from the GIST FFPE sample were graphed (Fig. 33b). Data were expressed in normalized fluorescence units (FU) per unit of size (bp). The UMI of the annotated cell type indicated in Fig. 32 was shown (Fig. 33c).

[0605] Experimental Result 31. Confirmation of Differential Gene Expression Profiles of Tumor Populations in Gastrointestinal Tumors (GIST)

[0606] The expression of the GIST marker gene (ANO1) and GIST oncogenes (KIT and PDGFRA) was confirmed (Fig. 34a). In addition, the expression of cell proliferation marker genes was confirmed (Fig. 34b). The expression of genes related to the function of Cajal cells (ICC) was confirmed (Fig. 34c). Furthermore, the gene expression profile related to pathway 1 of the tumor population in GIST was confirmed (Fig. 35). The gene expression profile related to pathway 2 of the tumor population in GIST was confirmed (Fig. 36).

[0607] Experimental Result 32. FX-seq Analysis Results of Stored Colorectal Cancer (CRC) FFPE Tissue

[0608] A complete microscopic image of the H&E-stained portion of the CRC FFPE was examined (Fig. 37a). Additionally, magnified images of various morphological regions were presented. Fig. 37b shows the boundary between the non-tumor (NT) and tumor (T) regions. It showed a tumor region with collapsed mucosal tissue morphology (Fig. 37c) and a non-tumor region with intact tissue morphology (Fig. 37d).

[0609] The results of automated electrophoresis of total RNA extracted from FFPE sections of CRC specimens were examined (Fig. 37e). Data were expressed in fluorescence units (FU) normalized per unit of size (bp). Quality indicators of the CRC FX-seq experiment were examined by comparing spatially resolved tumor and non-tumor regions. These include detected genes (Fig. 37f), UMI (Fig. 37g), mitochondrial readout ratio (Fig. 37h), and intron alignment ratio (Fig. 37i). The UMI of the annotated cell type in Fig. 38 is shown (Fig. 37j).

[0610] Experimental Result 33. Verification of FX-seq analysis results including spatial annotation of archived human colorectal cancer (CRC) FFPE specimens

[0611] A schematic procedure for FX-seq on stored CRC FFPE specimens is shown (Fig. 37j). Tumor (T) and non-tumor (NT) regions were distinguished in adjacent FFPE sections based on cancer regions identified by a pathologist using H&E section imaging. Adjacent 10 μm FFPE sections were scraped, nuclei were separated via FX-seq, and regional origin (T vs. NT) was distinguished through the barcode of the first reverse transcription primer of the combination labeling.

[0612] The results of UMAP analysis of transcripts labeled by region origin in the FFPE section are shown (n = 47,966 in T, n = 13,559 in NT) (Fig. 38b). Additionally, representative marker genes for each cluster are shown in Fig. 38c. The distribution of cell populations indicated by region in Fig. 38b is shown in Fig. 38d.

[0613] The proportion of cells belonging to a cluster was calculated from all cells identified in each region (Fig. 38e). Colored dots represent the same cell type annotations as shown in Fig. 38b. The results of UMAP analysis to identify sub-clusters of the merged gene matrix of epithelial progenitor cells and tumor populations, indicated by the green dashed lines in Fig. 38b, were verified (Fig. 38f).

[0614] Figure 38f shows the distribution of cell populations according to regional origin (Figure 38g). It was confirmed that the progenitor cell population had a mixed distribution of nuclei in the T and NT regions, whereas the tumor cell population had a specific distribution of nuclei in the T region.

[0615] Representative marker genes of the cell types identified by the sub-cluster analysis in Fig. 38f were shown (Fig. 38h). Cluster-level CNV inference results in gene expression of tumor and epithelial progenitor cell populations were shown using the inferCNV package (Fig. 38i).

[0616] The results of UMAP analysis to identify sub-clusters of the fibroblast cluster indicated by the orange dashed circle in Fig. 38b are shown (Fig. 38j). Additionally, the distribution of fibroblast sub-clusters indicated by regional origin in Fig. 38j is shown (Fig. 38k).

[0617] The cell number distribution of each cluster shown in Figs. 38j and 38k is shown (Fig. 38l). In addition, representative marker genes identified by sub-cluster analysis in Fig. 38j are shown (Fig. 38m).

[0618] Experimental Result 34. Confirmation of Differential Gene Expression Profiles of Tumor Populations in Colorectal Cancer (CRC)

[0619] Gene expression in subclusters including epithelial progenitor cells and tumor populations was shown (Fig. 39a). Gene expression in all identified populations is shown in Fig. 39b.

[0620] Experimental Result 35. Confirmation of FFPE Block QC and Sequencing Results in Various Colorectal Cancer Patient Groups

[0621] FFPE blocks stored at the Department of Pathology, Severance Hospital, Yonsei University were received for the analysis of various colorectal cancer (CRC) samples. All blocks were sectioned to a thickness of 10–30 μm, followed by deparaffinization and rehydration, after which the tissues were lysed. The blocks that underwent QC were sectioned again, separated into single nuclei after crosslinking agent exchange, and analyzed. It was confirmed that the UMI values ​​of each sequenced block had a high correlation with the Ct values ​​obtained by qPCR QC (Fig. 41a).

[0622] This allowed the qPCR analysis performed in the present invention to preview the quality of sequenced single nuclei, thereby preventing poor-quality blocks from occupying sequencing reads and preventing the waste of reagents used in library construction. RNA of the tissue lysed from FFPE sections was analyzed by automated electrophoresis. Subsequently, the correlation between the UMI number of the sequencing results and each RNA Integrity Number (RIN) (Fig. 41b), DV200 (Fig. 41c), and DV500 (Fig. 41d) was confirmed.

[0623] R of the linear regression trend line measured in each correlation analysis 2When comparing the values, the values ​​measured using actual qPCR showed the strongest correlation compared to the RIN, DV200, and DV500 values ​​generally used for RNA QC. These are the single-nuclear transcripts, identified cell types, and UMAP plots obtained from FFPE blocks of 24 patients after completing QC and proceeding with sequencing (Fig. 41e). In the UMAP plot shown, the UMAP plots were plotted by dividing the patients into groups according to their microsatellite mutations (Fig. 41f). The human subject study, FFPE block fragmentation, homogenization, deparaffinization, and rehydration of FFPE samples, fixative exchange, nuclear separation, qPCR screening, reverse transcription and ligation, library preparation, and sequencing sections of the above example were referenced.

[0624] Experimental Result 36. Confirmation of Crosslinker Exchange Sequencing Results in Sarcoma FFPE Samples from Various Patient Groups

[0625] Crosslinking agent exchange sequencing was performed on Ewing sarcoma FFPE tissues derived from various patients. The number of UMIs (Fig. 42a) and the number of detected genes (Fig. 42b) for each patient's Ewing sarcoma sample were confirmed. Results of single-nucleus sequencing performed on Ewing sarcoma. Various clusters were detected using the Leiden algorithm (Fig. 42c), and these were organized according to the patient (Fig. 42d) or the patient-derived sample number (Fig. 42e). The sections regarding human subject studies, FFPE block fragmentation, homogenization, deparaffinization, and rehydration of FFPE samples, fixative exchange, nuclear separation, qPCR screening, reverse transcription and ligation, library preparation, and sequencing in the above examples were referenced.

[0626] Experimental Result 37. Confirmation of the effect of additional PFA fixation prior to crosslinker exchange sequencing and the optimization of cell concentration for organic catalyst treatment

[0627] Tissues extracted in clinical settings are typically fixed in formalin (4% PFA, 1X PBS) and embedded in paraffin wax for long-term storage. However, the exact fixation time varies by hospital and by pathology department staff. Among FFPE tissues produced in these varying environments, those with insufficient fixation, in particular, exhibit lower detectable gene levels because intracellular RNA is degraded by RNA-degrading enzymes during high-temperature processing. Additionally, tissues that are large or insufficiently fixed for other reasons, where RNA-degrading enzymes remain active, can be effectively processed.

[0628] Therefore, additional fixation with PFA prior to tissue processing provides additional structural support and can further immobilize and neutralize RNA-degrading enzymes. To verify this, human colorectal cancer tissue extracted from Severance Hospital, Yonsei University was embedded in FFPE blocks. Subsequently, the tissues were sectioned to a thickness of 30 μm, deparaffinized, and rehydrated. Afterward, the sections were incubated in an additional fixation solution (1X PBS, 4% PFA) for 48 hours or 30 minutes in a low-temperature (4°C) environment, followed by the application of the crosslinker exchange method. Sections fixed with additional PFA for 48 hours exhibited higher gene detection sensitivity, while additional fixation for a short duration (30 minutes) had no significant effect. Furthermore, crosslinker exchange sequencing involves high-temperature reactions; typically, in mouse brains, the organic catalytic reaction was carried out at a concentration of 1 to 4 million nuclei / mL.

[0629] Mouse brains fixed with PFA consistently undergo a fixation reaction for 48 hours after perfusion with 1X PBS and 4% PFA solution, resulting in a consistent and robust degree of fixation. Consequently, the activation of RNA degrading enzymes is relatively lower than that of human tissues that were not consistently fixed. Therefore, when the reaction was performed using sections of the same human colorectal cancer FFPE tissue with an organic catalyst at a lower cell concentration than that used in the mouse brain experiment, the RNA degrading enzyme inhibitor in the organic catalyst solution reacted sufficiently to effectively stop RNA degradation. Furthermore, it was confirmed via qPCR that the detection sensitivity of genes was significantly enhanced. Refer to the sections on human subject studies, PFA fixation of mouse brains, FFPE embedding of human surgical samples, FFPE block fragmentation, homogenization, deparaffinization, and rehydration of FFPE samples, fixative exchange, nuclear separation, and qPCR screening in the experimental methods mentioned above. Each condition was evaluated using qPCR Ct values ​​for human GAPDH, B2M, and CNOT genes.

[0630] Experimental Result 38. Verification of Compatibility Evaluation between Cell Hashing and Fixed Samples

[0631] Cell hashing is a method commonly used in multiplex analysis to distinguish individual cells or experimental groups; typically, it involves using a tag in which an oligonucleotide is linked to an antibody, or the oligonucleotide sequence itself serves as a tag. Compatibility between crosslinker exchange sequencing (FX-seq) technology and cell hashing technology was evaluated. To maintain hashing oligonucleotides within the cells, immobilized PFA mouse brains were homogenized and permeated, then resuspended in a solution containing hashing oligonucleotides (1X MOPS, hashing oligo). The mixture was reacted at 1000 rpm for 10 minutes at 25°C, after which 4% PFA was immediately added and reacted for 15 minutes in a low-temperature environment (4°C) (1X MOPS, hashing oligo, 4% PFA).

[0632] Subsequently, after the crosslinking agent exchange and single-nucleus separation processes, qPCR was used to evaluate whether hashing oligonucleotides were still present in the cells. As the concentration of hashing oligonucleotides introduced into the cells increased, the detected values ​​also increased. This suggests that the hashing oligonucleotides were well maintained within the cells. This implies that crosslinking agent exchange sequencing technology can analyze various samples in a single experiment by utilizing the principle of cell hashing. The PFA fixation of mouse brains, homogenization of strongly fixed tissues, fixative exchange, nuclear separation, and qPCR screening sections of the above example were referenced. The results are shown in Fig. 44.

[0633] Experimental Result 39. Confirmation of compatibility of FX-seq processed fixed nuclei with a microfluidics-based single-cell analysis platform

[0634] The feasibility of analyzing FX-seq-treated cells and nuclei was tested not only with combinatorial barcoding (sci-RNA-seq) analysis techniques but also with other single-cell analysis techniques and platforms. After homogenization of fixed mouse brain tissue, a crosslinking agent exchange process was performed, and single nuclei were isolated. The isolated nuclei were captured within microfluidic microdroplets using a 10X Chromium 3' Reagent kit and Chromium X instrument. The aqueous and oil phases of the captured microdroplet solution were separated, and the microdroplets remaining in the oil phase were isolated. After lysing the isolated microdroplets, synthesized cDNA was isolated using SPRI beads. Second strand synthesis (SSS) was performed on the isolated cDNA to synthesize complementary base sequences.

[0635] After the reaction was complete, libraries were prepared following the library preparation process described in the Chromium Single Cell 3' Reagent Kits User Guide (v3.1 Chemistry) and measured via automated electrophoresis. Data were expressed as fluorescence units (FU) normalized per unit size (bp). "Lower" indicates a lower molecular weight standard. It was confirmed that high-quality libraries were generated using a microfluidic-based single-cell analysis platform, suggesting that the fixed tissue processing of crosslinker exchange sequencing is easily compatible with and usable with other single-cell analysis techniques and platforms.

[0636] Refer to the parts of the above examples regarding PFA fixation of mouse brain, homogenization of strongly fixed tissue, fixative exchange, nuclear isolation, reverse transcription and ligation, library preparation, and sequencing. After isolating a single nucleus, a library was prepared using the 10X Chromium kit with the slight modifications mentioned above. The results are shown in Fig. 45.

[0637] Experimental Result 40. Results of evaluating the feasibility of analyzing chromatin accessibility of fixed chromatin

[0638] Chromatin accessibility analysis is a technique that typically analyzes euchromatin, where the chromatin is unwound and transcription is facilitated, and heterochromatin, where the chromatin is condensed and transcription is inhibited. Euchromatin allows easy access for enzymes such as Tn5, enabling the insertion and cleavage of specific oligonucleotides into the open chromatin; conversely, heterochromatin does not allow easy enzymatic access, resulting in poor insertion and cleavage of oligonucleotides. Therefore, chromatin accessibility analysis can analyze the transcriptional potential of individual genes by analyzing the sequences of euchromatin and comparing them to the entire human genome.

[0639] However, it is known that chromatin accessibility analysis is also difficult in samples strongly immobilized with PFA. We evaluated the feasibility of analyzing chromatin accessibility in immobilized samples treated with organic catalysts using the crosslinker exchange sequencing (FX-seq) technique. We confirmed the effects of organic catalyst treatment time and temperature on the chromatin of PFA-immobilized cells. Single nuclei isolated from HEK293T cells were incubated in 4% PFA at a low temperature (4°C) for 20 hours. Subsequently, euchromatin was extracted via a Tn5 enzyme reaction following the organic catalyst treatment and single nucleus isolation processes. It was determined by performing qPCR with primers complementary to the oligonucleotides attached to the enzyme that the accessibility of the Tn5 enzyme improved with a long reaction time (3 hours) at a relatively low temperature (37°C).

[0640] This suggests that the difficulty of analyzing chromatin accessibility due to PFA fixation can be overcome through crosslinker exchange sequencing (FX-seq) technology. The procedure was performed on HEK293T cells by referring to the Tagmentation step among the steps of HeLa cell culture, fixative exchange, qPCR screening, library preparation, and sequencing in the above examples. Among these, RNA degradase inhibitors were excluded under the organic catalyst treatment conditions, and the organic catalyst treatment time and temperature conditions were varied. Each condition was evaluated using the qPCR Ct values ​​of Illumina's P5-P7 adapter sequence. The results are shown in Fig. 46.

Claims

1. A cell or tissue pretreatment method comprising a step of exchanging the fixative.

2. In Paragraph 1, A cell or tissue pretreatment method in which the cell or tissue is fixed with paraformaldehyde and / or formalin.

3. In Paragraph 1, A cell or tissue pretreatment method in which the above fixative exchange is performed by reacting with an organic catalyst.

4. In Paragraph 1, A cell or tissue pretreatment method in which the above organic catalyst is a compound represented by the following chemical formula I: [Chemical Formula I] In the above chemical formula I, R A is a hydroxyl, C 1-6 Alkoxy, mercapto, amino, C 1-6 alkylamino or di(C 1-6 It is an alkyl)amino and; R B is H; halo; hydroxy; C 1-6 Alkoxy; Amino; C 1-6 Alkylamino; di(C 1-6 Alkyl)amino; nitro; cyano; carboxy; C 1-6 Alkyl carbonyl; carbamoyl; C 1-6 Alkylcarbamoyl; di(C 1-6 Alkyl)carbamoyl; or halo, hydroxy, C 1-6 Alkoxy, amino, C 1-6 alkylamino, di(C 1-6 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-6 It is an alkyl.

5. In Paragraph 4, The above organic catalyst is a cell or tissue pretreatment method represented by the following chemical formula I-1: [Chemical Formula I-1] In the above chemical formula I-1, R B is the same as defined in Paragraph 2.

6. In Paragraph 5, A cell or tissue pretreatment method comprising the above organic catalyst 2-aminophenylboronic acid.

7. In Paragraph 1, A cell or tissue pretreatment method wherein the above fixative exchange further includes the step of reacting the cell or tissue with a crosslinker.

8. In Paragraph 7, A cell or tissue pretreatment method in which the above crosslinker is a compound represented by the following chemical formula II: [Chemical Formula II] In the above chemical formula II, Z 1a , Z 2a , Z 1b and Z 2b Each is independently O, OH, S, SH, NH, NH2, N(C 1-6 alkyl) or NH(C 1-6 alkyl) and; R 1a and R 1b are each independently F, Cl, Br, or I; Q a and Q b Each independently directly bonds, O, S, NH, NH2, N(C 1-6 alkyl) or N(C 1-6 Alkyl)2 and; L is C arbitrarily interrupted by NH, O, or S 2-40 It is an alkylene; R 4a and R 4b Each independently halo; hydroxy; C 1-6 Alkoxy; Amino; C 1-6 Alkylamino; di(C 1-6 Alkyl)amino; nitro; cyano; carboxy; C 1-6 Alkyl carbonyl; carbamoyl; C 1-6 Alkylcarbamoyl; di(C 1-6 Alkyl)carbamoyl; or halo, hydroxy, C 1-6 Alkoxy, amino, C 1-6 alkylamino, di(C 1-6 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-6 It is alkyl; l and m are each independently integers from 1 to 3; o and p are each independently integers from 0 to 2.

9. In Paragraph 8, A cell or tissue pretreatment method in which the above crosslinker is a compound represented by the following chemical formula IIA: [Chemical Formula IIA] In the above chemical formula IIA, R 1a , R 1b , Q a , Q b , L, R 4a , R 4b , o and p are as defined in Paragraph 6, and R 2a , R 3a , R 2b and R 3b Each independently H or C 1-6 It is an alkyl.

10. In Paragraph 9, A cell or tissue pretreatment method in which the above crosslinker is a compound represented by the following chemical formula IIB: [Chemical Formula IIB] In the above chemical formula IIB, R 1a , R 1b , Q a , Q b , R 2a , R 3a , R 2b , R 3b , R 4a , R 4b , o and p are as defined in Paragraph 7, and R 5a and R 5b H, H2, and C are independently 1-6 alkyl or (C 1-6 Alkyl)2 and, L a and L b C each independently 1-6 It is an alkylene, and Q 1 is a direct bond, NH, O, or S, and n is an integer from 1 to 20.

11. In Paragraph 7, A cell or tissue pretreatment method in which the above crosslinker is reacted prior to an organic catalytic reaction.

12. In Paragraph 1, A cell or tissue pretreatment method comprising the additional step of adding an RNase inhibitor.

13. In Paragraph 12, A cell or tissue pretreatment method in which the above RNase inhibitor is PVSA (polyvinyl sulfonic acid).

14. In Paragraph 1, A cell or tissue pretreatment method that further includes the step of separating the nucleus by treating with protease.

15. A step of detecting nucleic acids in a cell or tissue pretreated by any one of the methods of claims 1 to 14; A method for analyzing nucleic acids in cells or tissues.

16. In Paragraph 15, A method for analyzing nucleic acids in cells or tissues, wherein the above nucleic acid is RNA.

17. In Paragraph 15, A method for analyzing nucleic acids in cells or tissues, comprising a step of detecting the nucleic acid and a reverse transcription step of converting mRNA into cDNA.

18. A pretreatment kit for formalin-fixed or formalin-fixed paraffin-embedded (FFPE) cell or tissue samples containing an RNase inhibitor, an organic catalyst, and a cross-linker.

19. In Paragraph 18, The above kit is a formalin-fixed or formalin-fixed paraffin-embedded (FFPE) cell or tissue sample pretreatment kit that further comprises a cell separation kit.

20. In Paragraph 18, A formalin-fixed or formalin-fixed paraffin-embedded (FFPE) cell or tissue sample pretreatment kit in which the above RNase inhibitor is PVSA.

21. In Paragraph 18, The above organic catalyst is a formalin-fixed or formalin-fixed paraffin-embedded (FFPE) cell or tissue sample pretreatment kit, which is a compound represented by the following chemical formula I: [Chemical Formula I] In the above chemical formula I, R A is a hydroxyl, C 1-6 Alkoxy, mercapto, amino, C 1-6 alkylamino or di(C 1-6 It is an alkyl)amino and; R B is H; halo; hydroxy; C 1-6 Alkoxy; Amino; C 1-6 Alkylamino; di(C 1-6 Alkyl)amino; nitro; cyano; carboxy; C 1-6 Alkyl carbonyl; carbamoyl; C 1-6 Alkylcarbamoyl; di(C 1-6 Alkyl)carbamoyl; or halo, hydroxy, C 1-6 Alkoxy, amino, C 1-6 alkylamino, di(C 1-6 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-6 It is an alkyl.

22. In Paragraph 18, The above crosslinker is a formalin-fixed or formalin-fixed paraffin-embedded (FFPE) cell or tissue sample pretreatment kit, which is a compound represented by the following chemical formula II: [Chemical Formula II] In the above chemical formula II, Z 1a , Z 2a , Z 1b and Z 2b Each is independently O, OH, S, SH, NH, NH2, N(C 1-6 alkyl) or NH(C 1-6 alkyl) and; R 1a and R 1b are each independently F, Cl, Br, or I; Q a and Q b Each independently directly bonds, O, S, NH₂, N(C 1-6 alkyl) or N(C 1-6 Alkyl)2 and; L is C arbitrarily interrupted by NH, O, or S 2-40 It is an alkylene; R 4a and R 4b Each independently halo; hydroxy; C 1-6 Alkoxy; Amino; C 1-6 Alkylamino; di(C 1-6 Alkyl)amino; nitro; cyano; carboxy; C 1-6 Alkyl carbonyl; carbamoyl; C 1-6 Alkylcarbamoyl; di(C 1-6 Alkyl)carbamoyl; or halo, hydroxy, C 1-6 Alkoxy, amino, C 1-6 alkylamino, di(C 1-6 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-6 It is alkyl; l and m are each independently integers from 1 to 3; o and p are each independently integers from 0 to 2.

23. A novel crosslinker represented by the following chemical formula II: [Chemical Formula II] In the above chemical formula II, Z 1a , Z 2a , Z 1b and Z 2b Each is independently O, OH, S, SH, NH, NH2, N(C 1-6 alkyl) or NH(C 1-6 alkyl) and; R 1a and R 1b are each independently F, Cl, Br, or I; Q a and Q b Each independently directly bonds, O, S, NH, NH2, N(C 1-6 alkyl) or N(C 1-6 Alkyl)2 and; L is C arbitrarily interrupted by NH, O, or S 2-40 It is an alkylene; R 4a and R 4b Each independently halo; hydroxy; C 1-6 Alkoxy; Amino; C 1-6 Alkylamino; di(C 1-6 Alkyl)amino; nitro; cyano; carboxy; C 1-6 Alkyl carbonyl; carbamoyl; C 1-6 Alkylcarbamoyl; di(C 1-6 Alkyl)carbamoyl; or halo, hydroxy, C 1-6 Alkoxy, amino, C 1-6 alkylamino, di(C 1-6 C optionally substituted with at least one of alkyl)amino, nitro, and cyano 1-6 It is alkyl; l and m are each independently integers from 1 to 3; o and p are each independently integers from 0 to 2.

24. In Paragraph 23, A novel crosslinker wherein the above crosslinker is a compound represented by the following chemical formula IIA: [Chemical Formula IIA] In the above chemical formula IIA, R 1a , R 1b , Q a , Q b , L, R 4a , R 4b , o and p are as defined in Paragraph 6, and R 2a , R 3a , R 2b and R 3b Each independently H or C 1-6 It is an alkyl.

25. In Paragraph 23, A novel crosslinker wherein the above crosslinker is a compound represented by the following chemical formula IIB: [Chemical Formula IIB] In the above chemical formula IIB, R 1a , R 1b , Q a , Q b , R 2a , R 3a , R 2b , R 3b , R 4a , R 4b , o and p are as defined in Paragraph 7, and R 5a and R 5b H, H2, and C are independently 1-6 alkyl or (C 1-6 Alkyl)2 and, L a and L b C each independently 1-6 It is an alkylene, and Q 1 is a direct bond, NH, O, or S, and n is an integer from 1 to 20.

26. In Paragraph 23, A novel crosslinker wherein the crosslinker is a compound represented by the following chemical formula 1: [Chemical Formula 1] .

27. In Paragraph 1, The novel crosslinker described above is characterized by not interfering with Watson-Crick base pair formation and selectively binding to a guanine-N7 atom.

28. A composition for fixing cells or tissues containing PVSA and formaldehyde.

29. Composition for fixative exchange comprising an organic catalyst and PVSA.

30. Buffer composition containing PVSA.

31. Regarding Paragraph 30, A buffer composition wherein the buffer is any one selected from the group consisting of Tris, citrate, phosphate, HEPES, MOPS, and carbonate.

Images