Light chain split luciferase (LSL) reporter system and assay

Abstract

The present disclosure provides compositions and methods for determining the folding of a protein of interest. The present disclosure further provides methods for determining the pathogenicity of mutations in proteins.

Description

30275 / 70546LIGHT CHAIN SPLIT LUCIFERASE (LSL) REPORTER SYSTEM AND ASSAYCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application 63 / 730,590, filed on December 11, 2024. The entire content of this application is incorporated herein by reference in their entirety.GOVERNMENT SUPPORT CLAUSE

[0002] This invention was made with government support under NS099160 awarded by the National Institutes of Health. The government has certain rights in the invention.INCORPORATION BY REFERENCE OF MATERIAL SUBMITTED ELECTRONICALLY

[0003] This application contains, as a separate part of the disclosure, a Sequence Listing in computer-readable form which is incorporated by reference in its entirety and identified as follows: 70546_SeqListing.xml; Size: 17,958 bytes; Created: November 19, 2025.FIELD OF THE INVENTION

[0004] The present disclosure provides a method of testing the conformation of a protein using a split reporter system in combination with an antibody fragment. As described herein, a method was designed for the identification of mutations, including pathogenic mutations associated with a disorder. In one embodiment, mutations in the NOTCH3 gene that are associated with CADASIL syndrome are investigated.BACKGROUND

[0005] Protein folding is a complex process by which a sequence of amino acids folds into a three-dimensional configuration. Genetic mutations can destabilize proteins, causing them to alter their properties or cellular location, which can lead to incorrect folding and improper functioning. In addition to underlying genetic mutations, protein misfolding can be caused by several factors, including translational errors, abnormal protein modifications, thermal or oxidative stress, and incomplete complex formations.30275 / 70546

[0006] Gel mobility assays have been used to identify potential suppressor residues (i.e., cysteine mutations paired with a disease allele) that reverse the gel mobility effects of single cysteine mutations (Lee et al., J Biol Chem, 2023: p. 104838.). However, there exists a need in the art for an improved assay to measure protein folding and to predict the impact of protein variants on disease.SUMMARY

[0007] The present disclosure provides methods for testing the conformation of a protein using a split reporter system in combination with an antibody fragment. In one embodiment, an expression construct is provided comprising a polynucleotide encoding a fusion protein, wherein the fusion protein comprises: a) an antibody fragment; b) a first detection moiety comprising a first portion of a split reporter; c) a protein sequence of interest; and d) a second detection moiety comprising a second portion of the split reporter; wherein the first and second detection moieties are capable of assembly when in proximity, and wherein the first and second detection moieties emit a detection signal when assembled. In one embodiment, the fusion protein comprises, from N-terminus to C-terminus: the antibody fragment, the first detection moiety, the protein sequence of interest, and the second detection moiety. In still another embodiment, the fusion protein comprises, from N-terminus to C-terminus: the protein sequence of interest, the first detection moiety, the antibody fragment, and the second detection moiety. In related embodiments, the order of the aforementioned components of the construct are provided in any order and, in related embodiments, the antibody fragment component may be replaced with a different protein that is useful for distinguishing wild-type versus pathological / misfolded, such as ApoE as described herein.

[0008] In still another embodiment of the present disclosure, an aforementioned method is provided further comprising one or more linkers. In one embodiment, the one or more linkers are located between the antibody fragment and the first detection moiety, the antibody fragment and the second detection moiety, the protein sequence of interest and the first detection moiety, and / or the protein sequence and the second detection moiety. In one embodiment, the one or more linker sequence is RS and / or VD.

[0009] In another embodiment of the present disclosure, an aforementioned method is provided wherein the antibody fragment comprises a light chain. In one embodiment, the light30275 / 70546chain is a rabbit light chain. In yet another embodiment, the rabbit light chain comprises a Cys80 residue. In still another embodiment, the rabbit light chain is 83G, 44F, 69B, 82A, 120B, 123D, or 145H. In one embodiment, the rabbit light chain is 83G.

[0010] The present disclosure also provides, in one embodiment, an aforementioned method wherein the first and second detection moieties comprise split β-lactamase, split GFP, or split luciferase. In one embodiment, the first and second detection moieties comprise split luciferase. In still another embodiment, the first and second detection moieties comprise SmallBiT and LargeBiT of Nanoluc.

[0011] In still another embodiment of the present disclosure, an aforementioned method is provided wherein the protein sequence of interest comprises one or more mutations relative to a reference protein sequence. In another embodiment, the protein sequence of interest contains one or more cysteine residues. In yet another embodiment, the protein sequence of interest contains multiple cysteine residues, and wherein two or more of the cysteine residues form a disulfide bridge. In still another embodiment, mutations to one or more cysteine residues disrupt formation of at least one disulfide bridge. In another embodiment, the fusion protein, when properly folded, assembles the first portion and the second portion of the split reporter. In yet another embodiment, the fusion protein, when misfolded, does not assemble the first portion and the second portion of the split reporter. In yet another embodiment, the light chain comprises a Cys80 residue, and wherein a cysteine of the protein sequence of interest forms a disulfide bridge with the Cys80 residue.

[0012] In another embodiment of the present disclosure, an aforementioned method is provided wherein the protein sequence of interest comprises a secreted protein, a cytoplasmic protein, a transmembrane protein, organelle-localized protein or a fragment thereof. In one embodiment, the protein of interest comprises NOTCH3 or FBN1.

[0013] The resent disclosure also provides compositions. In one embodiment, the present disclosure provides a composition comprising an aforementioned expression construct. In another embodiment, a composition comprising a collection of the aforementioned expression constructs is provided, thereby providing a library of protein sequence variants. In still another embodiment, a host cell comprising an aforementioned expression construct is provided by the present disclosure.30275 / 70546

[0014] In still another embodiment of the present disclosure, an aforementioned composition is provided comprising a collection of aforementioned host cells, thereby providing a library of protein sequence variants capable of being expressed and secreted from the collection of host cells.

[0015] In yet another embodiment, the present disclosure provides a method of determining whether a protein of interest is properly folded, the method comprising the steps of: (a) transfecting a host cell with an expression construct comprising a polynucleotide encoding a fusion protein, wherein the fusion protein comprises: i) an antibody fragment; ii) a first detection moiety comprising a first portion of a split reporter; iii) a protein sequence of interest; and iv) a second detection moiety comprising a second portion of the split reporter; wherein the first and second detection moieties are capable of assembly when in proximity, wherein the first and second moieties emit a detection signal when assembled, and wherein the protein sequence comprises one or more mutations relative to a reference protein sequence; (b) incubating the cell of (a) in a growth media under conditions that allow expression and secretion of the fusion protein; (c) collecting the growth media comprising the expressed and secreted fusion protein; (d) detecting a detection signal from the growth media; (e) determining if the fusion protein is properly folded by comparing the fusion protein detection signal to a signal from a reference protein.

[0016] In still another embodiment, a method of determining whether a protein of interest is properly folded is provided, the method comprising the steps of: (a) transfecting a host cell with an expression construct comprising a polynucleotide encoding a fusion protein, wherein the fusion protein comprises: i) an antibody fragment; ii) a first detection moiety comprising a first portion of a split reporter; iii) a protein sequence of interest; and iv) a second detection moiety comprising a second portion of the split reporter; wherein the first and second detection moieties are capable of assembly when in proximity, wherein the first and second moieties emit a detection signal when assembled, and wherein the protein sequence comprises one or more mutations relative to a reference protein sequence; (b) incubating the cell of (a) in a growth media under conditions that allow expression and of the fusion protein; (c) lysing the cell and collecting the cell lysate; (d) detecting a detection signal from the cell lysate; (e) determining if the fusion protein is properly folded by comparing the fusion protein detection signal to a signal from a reference protein.30275 / 70546

[0017] In still another embodiment, an aforementioned method is provided wherein the reference protein is a wild-type protein. In one embodiment, the cell is lysed by chemical lysis, enzymatic lysis, mechanical lysis, or physical lysis. In another embodiment, the chemical lysis comprises use of an alkali solution or detergent solution. In another embodiment, the detergent solution comprises 3-((3-cholamidopropyl) dimethylammonio)- 1 -propanesulfonate (CHAPS), cetyltrimethylammonium bromide (CTAB), nonyl phenoxypolyethoxylethanol (NP-40), polyethylene glycol sorbitan monolaurate (Tween™), sodium dodecyl sulphate (SDS), or t-Octylphenoxypolyethoxyethanol (Triton™ X-100). In yet another embodiment, the enzymatic lysis comprises use of lysozyme, lysostaphin, zymolase, cellulose, protease or glycanase. In still another embodiment, mechanical lysis comprises use of a high-pressure homogenizer or a bead mill. In one embodiment, physical lysis comprises use of ultrasonic waves, heat, osmotic shock, or cavitation.

[0018] In another embodiment, an aforementioned method is provided by the present disclosure, further comprising calculating a detection ratio by dividing the reference protein detection signal by the fusion protein detection signal, wherein a calculated detection ratio value greater than at least 2 corresponds to a properly-folded folding fusion protein and wherein a calculated detection ratio value less than at least 2 corresponds to a misfolded fusion protein. In yet another embodiment, the aforementioned method is provided further comprising: f) contacting the growth media with a reducing agent; and g) detecting a detection signal from the growth media. In one embodiment, the method is provided further comprising calculating a detection ratio by dividing the reference protein detection signal by the detection signal of (g), wherein a calculated detection ratio value greater than 2 corresponds to a properly-folded folding fusion protein and wherein a calculated detection ratio value less than 2 corresponds to a misfolded fusion protein. In one embodiment, the reducing agent is tris (2-carboxyethyl) phosphine hydrochloride (TCEP), p-mercaptoethanol (BME), or dithiothreitol (DTT).

[0019] The present disclosure further provides, in one embodiment, an aforementioned method further comprising: f) separating the host cell from the growth media; g) lysing the cell and collecting the cell lysate; h) contacting the host cell with a secondary antibody, wherein the secondary antibody is capable of binding to the antibody fragment, and wherein the secondary antibody comprises a detectable label; and i) adding the secondary antibody to the growth media; and j) detecting a signal from the secondary antibody; k) comparing the signal from the30275 / 70546secondary antibody with the signal from the growth media to measure secretion of the fusion protein. In still another embodiment a method further comprises: 1) treating cells with a thiol alkylator; and m) repeating steps a) through k). In some embodiments, a thiol alkylator includes iodoacetamide, PX-12, N-ethylmaleimide, disulfiram, ebselen, carmofur, auranofin, spebrutinib, osimertinib, and necrosulfonamide.

[0020] The present disclosure also provides, in one embodiment, a method of determining whether a protein of interest comprises a pathogenic protein sequence, the method comprising the steps of: (a) transfecting a host cell with an expression construct comprising a polynucleotide encoding a fusion protein, wherein the fusion protein comprises: i) an antibody fragment; ii) a first detection moiety comprising a first portion of a split reporter; iii) a protein sequence of interest; and iv) a second detection moiety comprising a second portion of the split reporter; wherein the first and second detection moieties are capable of assembly when in proximity, wherein the first and second moieties emit a detection signal when assembled, and wherein the protein sequence comprises one or more mutations relative to a reference protein sequence; (b) incubating the cell of (a) in a growth media under conditions that allow expression and secretion of the fusion protein; (c) collecting the growth media comprising the expressed and secreted fusion protein; (d) detecting a detection signal from the growth media; (e) determining if the fusion protein is properly folded by comparing the fusion protein detection signal to a signal from a reference protein; wherein a determination of proper folding corresponds to protein sequence comprising a benign mutation, and wherein a determination of improper folding corresponds to a protein sequence comprising a pathogenic mutation.

[0021] In still another embodiment of the present disclosure, a method of identifying a therapeutic compound or biomolecule is provided, the method comprising the steps of: (a) contacting a misfolded fusion protein according or a pathogenic protein according to an aforementioned method with a compound or biomolecule; (b) detecting a detection signal; (e) determining if the detection signal detected in (b) is improved relative to a detection signal in the absence of the compound or biomolecule; wherein an improved detection signal corresponds to a compound or biomolecule capable of rescuing a misfolded fusion protein or pathogenic protein.30275 / 70546BRIEF DESCRIPTION OF THE DRAWING

[0022] Fig. 1A-C shows a Light chain Split Luciferase (LSL) system to discriminate between wildtype and mutant NOTCH3 protein sequences. Fig. 1A shows representations of protein components are shown, with high activity wildtype nano luciferase represented at the top. In comparison, the split luciferase of LSL requires permissive conformations of the inserted test sequence that allows positioning of SmBiT in LgBiT to generate high activity (green insert of wildtype polypeptide is shown on the left). In contrast, if a variant causes large structural changes (red insert of pathogenic variants on the right), SmBiT is not able to interact with LgBiT to produce high activity. Fig. IB shows the design of the system includes a vector encoding components light chain variable domains of rabbit IgG (yellow), nano luciferase component SmBiT (mocha), test sequences of NOTCH3 or other genes (green), nano luciferase component LgBiT (latte). The LSL vector also includes a CMV promoter and polyadenylation signal sequence for expression by transfection into cells. A plasmid encoding iRFP is included so that transfection efficiency for normalization purposes. The effects of sequence variants cloned into the LSL vector are determined by referencing to wildtype LSL. Fig. 1C shows a comparison of LSL-NOTCH3 activity for wildtype and R90C canonical CADASIL mutation. Secreted nano luciferase activity from transfectants were compared after normalization to iRFP expression. The LSL-NOTCH3(l-3) wildtype construct generated high activity relative to the R90C mutant and to deletions of SmBiT and LgBiT. Activity generated by constructs containing wildtype NOTCH3 in LSL in which LgBiT was fused to SmBiT is also shown for comparison (bottom four). On the right, the ratio of iRFP normalized luciferase is quantified between WT and R90C mutant constructs. Significant differences are indicated (*) with p<0.05.

[0023] Fig. 2A-B shows the ability of LSL-NOTCH3 to discriminate between wildtype NOTCH3, CADASIL mutants, and non-pathogenic variants. Fig. 2A shows a schematic showing LSL-NOTCH3(l-3) vector used to analyze NOTCH3 EGF 1-3 variants. Cysteine (triangles) and non-cysteine (circles) residues were involved in pathogenic mutations. Amino acid changes are shown with pathogenic changes in red and non-pathogenic variants in cyan. Fig.2B shows the activity in media (iRFP normalized) of LSL-NOTCH3(l-3) transfected cells are shown, corresponding to variants in Fig.2A. All iRFP controlled values were referenced to WT (set at 1.0). Wildtype, pathogenic, and benign mutations are shown. All pathogenic mutans showed30275 / 70546significant differences from wildtype as indicated (*) with p<0.05. Benign variant R 113Q was not significantly different from wildtype.

[0024] Fig. 3A-D shows the effect of mutations at key NOTCH3 residues on LSL-NOTCH3 activity. Mutations were introduced into the NOTCH3 EGF 1-3 sequences of LSL-NOTCH3(1-3) in order to assess the effects of specific amino acid residues on LSL activity. Constructs are shown above each group of data which represents the iRFP normalized secreted luciferase activity for wildtype and mutants at NOTCH3 residues 49 (Fig. 3A), 90 (Fig. 3B), 75 (Fig.3C), and 146 (Fig. 3D). The wildtype residues are Cys49, Arg90, Arg75, and Cysl46, which were compared to mutations to all 19 other amino acids. Established pathological mutants are shown in red and WT is in cyan. Significant differences from wildtype are indicated (*) with p<0.05.

[0025] Fig. 4A -C shows a mapping of second cysteine mutations that suppress pathological NOTCH3 variant LSL function. Three mutations in NOTCH3 EGF 1-3 were analyzed and displayed in separate panels. Each panel shows, on the left, a schematic of mutations that were introduced into LSL-NOTCH3 and denotes the EGF repeat that harbors the mutation, depicted by a red circle (loss of cysteine) or by a red triangle (gain of cysteine). Black triangles represent the six cysteines of wildtype NOTCH3 in each repeat. The top construct (cyan dot) is the wildtype reference, the second construct is the established CADASIL mutant (in red), and all other constructs are second mutations in each of the remaining cysteine positions (changed to serine). Corresponding secreted LSL activity generated by constructs, normalized to iRFP expression, was determined for each construct, and all values were referenced to the WT construct (value of 1.0) shown in the chart to the right in each panel. The CADASIL mutants analyzed included C49Y (Fig. 4A; EGF 1), R90C (Fig. 4B; EGF 2), and C155Y (Fig.4C; EGF 3). Double mutants with significant differences from established CADASIL mutants are indicated (partial rescue of activity, *). Double mutants with significant differences from the ideal double mutant (yellow c and brown e double mutants in Fig.4A and Fig.4C) are indicated (weaker than ideal suppressor, #). Significance was considered as p<0.05.

[0026] Fig. 5A-D shows the effect of pathogenic versus benign NOTCH3 variants on additional parameters of LSL-NOTCH3 activity: Min / Max and Secretion Index. Fig. 5A shows the TCEP-responsiveness of LSL-NOTCH3 activity generated by wildtype NOTCH3 EGF 1-3 versus the CADASIL mutant R90C that was determined by addition of 2 pL TCEP (31.25 mM)30275 / 70546to the nano luciferase assay mixture. The amount of enzyme from the same assay was reduced in increments of 25%, and time courses of activity were displayed in the same graph. Each series of activities represents undiluted conditioned media. 75% media, 50% media and 25% media (in descending order of activity, with WT (green) and R90C mutant (red). Fig. 5B shows the minimum activity (Min; time 0 prior to TCEP addition) and the maximum activity (Max) after TCEP addition was used to calculate the Min / Max parameter. The paired Min / Max values are displayed for both the wildtype LSL-NOTCH3(l-3) WT versus R90C CADASIL mutant for all dilutions shown in (Fig. 5A). LSL activities produced by different concentrations of media did not affect the Min / Max parameter. Fig. 5C shows the Min / Max (parameter 2) of LSL-NOTCH3(l-3) for an expanded set of NOTCH3 mutants in EGF 1-3 were assessed by challenging media from transfectants with TCEP. The same mutants as in Fig. 2 were used, representing both pathological (red) and benign (cyan) NOTCH3 variants. The Min / Max values were all referenced to WT Min / Max. Fig. 5D shows the Secretion Index (parameter 3) for all mutants analyzed in Fig. 5C was determined by comparison of the maximum TCEP-induced activity secreted into the media to the maximum TCEP-induced activity produced by cell lysates. For analyses of Fig. 5C-D, significant differences from wildtype are indicated (*) with p<0.05.

[0027] Fig. 6A-E shows mapping locations within LSL important for pathogenic versus benign discriminatory capacity. Sequences in the light chain portion of LSL were altered to determine sequences that were important for high discriminatory function of LSL-NOTCH3. Fig.6A shows that Figs. 1-5 used the 83G light chain variable sequences in LSL-NOTCH3 constructs. Here, the light chain coding sequence was replaced by a series of rabbit monoclonal antibody light chain variable sequences listed on the x-axis. Secreted activity was normalized to iRFP for both WT and R90C versions of each light chain construct and displayed in the chart on the left which shows all constructs yield higher values for WT over mutant NOTCH3. On the right chart, the ratio of WT to mutant generated luciferase values is shown to demonstrate the magnitude of discrimination for each light chain. Fig. 6B shows constructs using 83G light chain in the LSL-NOTCH3 backbone were generated with small insertions or deletions 5’ to the SmBiT sequence. WT or R90C NOTCH3 EGF 1-3 reporter activities were determined and compared in the chart, which shows each value referenced to the respective WT construct (value set to 1.0). Fig. 6C shows the effect of TCEP-mediated reduction on activities from Fig. 6B were determined by adding TCEP to the reaction mixture and determining the Min / Max value. The30275 / 70546Min / Max values were referenced to WT (set to 1.0). Constructs with the largest differences between WT and R90C were considered the best at discrimination between WT and pathological mutants. Fig.6D assesses the role of the unpaired cysteine residue at position 80 of rabbit light chain; constructs were generated in LSL-NOTCH3(l-3) with mutation C80S. Analysis of the Min / Max parameter of these constructs that incorporated WT and NOTCH3 R90C was performed by measurement of luciferase activity over time after addition of TCEP. The timedependent activity after addition of TCEP for a representative set of R90C mutants with or without C80S is shown in Fig. 6E.

[0028] Fig. 7A-F shows LSL-NOTCH3 activity with insertion of additional regions of NOTCH3. Fig.7A shows the schematic of NOTCH3 and its 34 EGF repeats is shown above representations of subdomains of NOTCH3 selected for additional LSL-NOTCH3 analysis. Each subdomain is composed of three adjacent EGF repeats and included analysis of EGF (29-31) in the C-terminal region of NOTCH3. Domains were cloned in the LSL vector as before. Activity of the reporters were compared for WT, pathological mutants (red) and benign variants (cyan) by transfection and analysis of three parameters, iRFP normalized luciferase activity (parameter 1), TCEP-responsiveness (Min / Max; parameter 2), and Secretion Index (parameter 3). For all parameters in each chart, values were referenced to WT. EGF clusters analyzed included 2-4 (Fig. 7B), 3-5 (Fig. 7C), 4-6 (Fig. 7D), 5-7 (Fig. 7E), and 29-31 (Fig. 7F). For (Figs. 7B-F), significant differences from wildtype are indicated (*) with p<0.05.

[0029] Fig. 8A-C shows LSL-FBN1 activity corresponding to variants associated with Marfan’s disease and stiff skin syndrome. EGF domains of FBN1 were inserted into the LSL vector and transfected into cells. Mutations that are considered pathogenic and benign variants were compared to wildtype sequences. Secreted activity was analyzed, along with min / max and secretion indices. EGF domains 11-13 and 13-15 of FBN1 were used to examine variants linked to Marfan’s disease (Fig. 8A-8B); regions of FBN1 corresponding to TGFBP domain 4 with EGF domain 23 (residues 1528-1651) were used to test activity related to mutations causing stiff skin syndrome (Fig.8C).

[0030] Fig. 9A -C shows the effect of iodoacetamide on LSL-NOTCH3 variant activity. Cells were transfected with LSL-NOTCH3 EGF 1-3 constructs encoding wildtype, pathogenic CADASIL mutations (red), or benign variants (cyan). Cells were treated for 2 hours by addition30275 / 70546of iodoacetamide or water, and secreted luciferase activity determined. All values were referenced to water treatment and thus display the ratio of activity with to without iodoacetamide. Fig.9A shows secreted luciferase values were normalized to iRFP to determine parameter 1 increases. Fig. 9B shows that to determine changes in parameter 2, Min / Max increases after addition of iodoacetamide are displayed. Fig.9C shows that for parameter 3 assessment, the Secretion Index was determined for NOTCH3 variant reporters in cells treated with and without iodoacetamide. The ratio of treatment to water control is shown in the chart. At least three experiments were performed; significant differences in iodoacetamide stimulated activity for variants versus WT control (*) are indicated for p<0.05.

[0031] Fig. 10A-B shows the framework for interpretation of LSL assays for NOTCH3 variants. Fig. 10A shows a suggested workflow for analyzing the effects of gene variants. LSL reporters can be generated that correspond to genetic variants and compared by transfection, followed by analysis of luciferase activities after TCEP treatment of media and cell lysates. Fig.10B shows a framework to account for the results of LSL analysis of NOTCH3 benign and pathological mutants is shown. The cellular and biochemical effects of each variant result in differing patterns of parameter 1-3 alteration as discussed. In some embodiments, pathological variants of NOTCH3 alter disulfide arrangements and / or protein secretion.

[0032] Fig. 11 shows an experimental approach to evaluating potential mitigators of pathogenic NOTCH3 variants. To assess the potential beneficial effects of cysteine-binding small molecules on pathogenic variants of NOTCH3, a series of candidates were tested (Table 1) on a luciferase reporter (LSL-NOTCH3) that reflects normal folding of NOTCH3. In this system, pathogenic NOTCH3 mutants generate lower luciferase activity. Candidate agents were evaluated for their ability to increase luciferase activity of pathogenic NOTCH3 reporters. Fig.11 Top shows a schematic of the LSL vector in which a CMV reporter drives expression of a recombinant protein composed of an antibody light chain (L), small BiT of nanoLuciferase (S), EGF repeats of NOTCH3 (WT vs mutant), and large BiT of nanoLuciferase. Cotransfection of iRFP is used to normalize transfection efficiency. Cells transfected with LSL-NOTCH3 reporters are treated with vehicle or candidate mitigator agents (Cys binders) and the conditioned media is removed for analysis. Parameter 1 is the total normalized production of luciferase activity and is a composite reflection of proper disulfide bonding and secretion of the reporter protein. The ratio of Parameter 1 in control vs treated cells is the principle outcome reported in30275 / 70546this study. On some experiments, Parameter 2 is assessed, which reflects the unmasking of luciferase activity with chemical reduction of the conditioned media by TCEP. This is taken as an index of fraction of protein secreted that is favorably disulfide bonded. Chemicals which increased mutant reporter Parameter 1 and / or Parameter 2 are considered mitigators of pathogenic NOTCH3 conformation.

[0033] Fig. 12A-B shows a pathogenic NOTCH3 conformational mitigation across mutants by a series of candidate small molecules. Using the workflow outlined in Fig. 11 and in methods, the reporter output of cells transfected with mutants shown on the vertical axis was determined after individual treatment with candidate mitigator drugs shown on the horizonal axis. Three different clusters of EGF repeats of NOTCH3 were examined: 1 -3, 4-6, and 31 -33. The ratio of Parameter 1 (total secreted luciferase / iRFP) for drug vs control conditions was then normalized to the respective WT ratio and displayed in the heat map in (Fig. 12A). All experiments were performed at least three times for each mutant and drug combination. Statistically significant increases in drug stimulated Parameter 1 over control are shown in (Fig. 12B), with green indicating p<0.05.

[0034] Fig. 13A-C shows disulfiram, IAM, and auranofin as mitigators of pathogenic NOTCH3 conformational alterations. Fold changes in LSL-NOTCH3 activity (Parameter 1; total activity secreted to media) after treatment of transfected cells with disulfiram (Fig. 13A; 10μM), 1AM (Fig. 13B; 10μM), and auranofin (Fig. 13C; 1μM) are shown for reporters that include NOTCH3 EGF repeats 1-3 (left), repeats 4-6 (right), and repeats 31-33. Wildtype (WT) and benign variants are shown in blue, and pathogenic variants are shown in red. All values are shown with standard deviations. G209R was categorized as pathogenic because it induced gel mobility shifting and substantially suppressed LSL-NOTCH3 activity in Example 1. * p<0.05 compared to fold increase for WT reporter of each EGF repeat group.

[0035] Fig. 14A-C shows disulfiram, IAM, and auranofin increase the fraction of favorable disulfide bonding patterns NOTCH3 reporters. Fold change in TCEP-independent fraction of LSL-NOTCH3 activity (Parameter 2; min / max after TCEP addition normalized to value control) after treatment of transfected cells with disulfiram (Fig. 14A; 10μM), IAM (Fig. 14B; 10μM), and auranofin (Fig. 14C; 1μM) are shown for reporters that include NOTCH3 EGF repeats 1-3 (left), repeats 4-6 (right), and repeats 31-33. Wildtype (WT) and benign variants are shown in30275 / 70546blue, and pathogenic variants are shown in red. All values are shown with standard deviations. * p<0.05 compared to fold increase over WT reporter of each EGF repeat group.

[0036] Fig. 15A-J show small molecule mitigation of pathogenic NOTCH3 during cellular processing of target protein. (Fig. 15A) To assess cell-independent LSL reporter activity, conditioned media of cells transfected with LSL-NOTCH3 reporters corresponding to the EGF repeat 1-3 variants shown on the x-axis was treated with indicated agents. Luciferase activity was then measured; these were normalized to iRFP expressed in transfected cells (Parameter 1). There were no significant differences dependent on treatment conditions. (Fig. 15B-J) To assess if the effects of mitigators of mutant NOTCH3 affected LSL-NOTCH3 reporter activity in a time-dependent fashion, media (without or with mitigators) of cells transfected with WT or mutant reporters indicated were collected over time periods noted in the x-axis. Luciferase activities normalized to without drug control (at the same time point) are displayed on the y-axis.(Fig. 15B-G) were treated with disulfiram (5μM); (H-J) were treated with PX-12 (5μM). The NOTCH3 variants (from EGF repeats (1-3) (Fig. 15B-D) and from EGF repeats (4-6) (Fig. 15E-J)) used in reporter constructs are shown. All values are shown with standard deviations.

[0037] Fig. 16A-J shows dose-dependent effects of mitigators on pathogenic NOTCH3 conformational alterations. Media supplemented with doses of disulfiram (Fig. 16A-E) or auranofin (Fig. 16F-J) were added to cultures transfected with LSL-NOTCH3 (1-3) WT (Fig. 16A and Fig. 16F); LSL-NOTCH3 (1-3) R90C (Fig. 16B and Fig. 16G); LSL-NOTCH3 (1-3) C49Y (Fig. 16C and Fig. 16H); LSL-NOTCH3 (1-3) R75P (Fig. 16D and Fig. 161); and LSL-NOTCH3 (1-3) C108R (Fig. 16E and Fig. 16J). The media collected was assayed for luciferase activity which was normalized to iRFP levels and displayed on the y-axis (Parameter 1). All values are shown with standard deviations.

[0038] Fig. 17A-C shows additive effects of multiple mitigators on pathogenic NOTCH3. For selected LSL-NOTCH3 reporters shown to respond to two different mitigators, the level of luciferase production in media with single drug treatments vs dual treatments were compared (Parameter 1). All drugs were used at 10μM. NOTCH3 variants were tested from EGF repeats (1-3) (Fig. 17A), repeats (4-6) (Fig. 17B), and repeats (31-33) (Fig. 17C). Values shown are normalized to each reporter’s control expression level without drug treatment.30275 / 70546

[0039] Fig. 18A-B show the effect of amino acid residue at cysteine mutation position on mitigator function. A series of mutants which harbor changes at a single cysteine to all other non-cysteine residues were assayed for LSL-NOTCH3 activity. In Fig. 18A, NOTCH3 residue 49 of LSL-NOTCH3 (1-3) was mutated from cysteine to all other amino acids. The naturally occurring CADASIL mutation is C49Y. After transfection, cells were treated without or with disulfiram (10μM) and luciferase in the media quantified. Value of drug induced luciferase activity (Parameter 1) normalized to no drug controls are displayed. In Fig. 18B, the same procedure and analysis was used for analysis of NOTCH3 residue 146 of LSL-NOTCH3 (1-3). The naturally occurring CADASIL mutation is C146R. All values are shown with standard deviations. * p<0.05 compared to fold increase for drug treated WT reporter.

[0040] Fig. 19A-C shows mitigator effects on an alternative mutant NOTCH3 reporter system. Fig. 19A shows a schematic of an APOE-based split luciferase system for differentiating pathogenic NOTCH3 conformations. The system is similar to the LSL-NOTCH3 system except that NOTCH3 sequences of interest are cloned at the 5’ end and inverted split luciferase is flanked by APOE2. Cloned NOTCH3-ASL constructs bearing WT versus variant NOTCH3 fragments are transfected into cells and culture media assayed for activity. To test activity of mitigators of NOTCH3 conformational changes, media is supplemented with candidate small molecules before culture media assays. Transfection efficiency is controlled by measurement of iRFP expression. Fig. 19B shows variant NOTCH3 sequences from EGF repeats (1-3) shown were cloned into the NOTCH3-ASL vector. After transfection into 293 cells, normalized luciferase levels were determined. Conformational alterations in pathogenic mutants (red) were compared to WT and benign variants. Reduction of luciferase activity in pathogenic mutants is consistent with conformational alterations of CADASIL NOTCH3. All values are shown with standard deviations. * p<0.05 compared to WT reporter. Fig. 19C NOTCH3-ASL plasmids from Fig. 19B were transfected and treated without and with disulfiram (10μM). Conditioned media from drug treated groups were normalize to media without treatment on the y-axis. All values are shown with standard deviations. * p<0.05 compared to fold increase for drug treated WT reporter. The left panels of Fig. 19B and Fig. 19C show results for EGF repeats (1-3), the center panels correspond to EGF repeats (4-6), and the right panels are results for EGF repeats (31-33). WT and benign variants are in blue, and pathogenic variants are in red.30275 / 70546

[0041] Fig. 20A-B shows mitigating effects on FBN1 mutations linked to Marfan’s disease. Previously described LSL-FBN1 reporters (Example 1) with WT and benign variants (blue) or pathogenic mutations (red) were transfected as in LSL-NOTCH3 experiments. The L1038F variant is considered a variant of uncertain significance. Media with and without disulfiram (10μM; (Fig.20A)) or auranofin (1μM; (Fig. 20B)) were assayed for luciferase activity, and values normalized to media without drug for each reporter is shown on the y-axis. All values are shown with standard deviations. * p<0.05 compared to fold increase for drug treated WT reporter.DETAILED DESCRIPTION

[0042] As described herein, to predict the impact of protein variants on disease, a novel recombinant vector and custom experimental protocol were designed. While split reporters have been used to evaluate binding between separate polypeptides (US Application No. 13 / 455,521 and US Application No. 13 / 898,637) and to evaluate conformational changes of a single protein induced by ligand binding (US Application No. 11 / 805,460 and US Application No. 11 / 805,425, Sheahan et al., J Mol Med (Berl). 2016 Jul;94(7):799-808, Paulmurugan et al., Oncotarget. 2018 Apr 20:9(30):21495-21511. Paulmurugan et al.. Proc Natl Acad Sci U S A. 2006 Oct 24;103(43):15883-8, and Kang et al., Microbiol Spectr. 2023 Feb 23;11(2):e0233822), there remains a need for methods to study protein conformation of a single protein and / or a panel of mutations associated with a single protein.

[0043] In one embodiment, the recombinant vector is composed of the light chain of a recombinant antibody, SmBiT (or “SBT”) of nano-luciferase, cDNA encoding protein variant of interest, and LgBiT (or “LBT”) of nano-luciferase (LSL construct). When wild type cDNA sequences of NOTCH3 are inserted into LSL, the construct generates high luciferase activity. When mutant cDNA sequences of NOTCH3 (that cause CADASIL) are inserted, the construct generates low luciferase activity. The custom experimental protocol includes a two-step process: 1) the conditioned media of cells transfected with LSL constructs are analyzed for luciferase activity and 2) TCEP is added to the reaction mixture, which blocks the inhibitory function of the mutant cDNA sequence, leading to maximal activity. The luciferase component of the LSL module is unaffected by reducing agents because it does not contain cysteine residues. The TCEP value is used as a normalization value and accounts for variations in transfection30275 / 70546efficiency and / or secretion efficiency. This two-step process performed in the same reaction mix enables reproducible assessments that are highly sensitive because normalization is performed in the same reaction well. The potential for LSL includes, but is not limited to, the ability to quantify the likelihood that a genomic sequence variant causes disease, a capability of importance in determining impact of variants of uncertain significance in genomic studies. A second impact of LSL is that it offers a direct route to screen drug libraries for agents with clinical activity. LSL constructs expressed in cell lines have been generated for specific mutations of NOTCH3. These cell lines can be used with drug libraries to identify chemicals that increase mutant LSL reporter activity by reversing the inhibition of LSL by mutant NOTCH3. The application of LSL to cDNA sequence variants beyond NOTCH3 has the potential to generally accelerate drug screening which may impact a wide range of genetic disorders.

[0044] Stereotyped mutations in NOTCH3 drive CADASIL, the leading inherited cause of stroke and vascular dementia. The vast majority of these mutations result in alterations in the number of cysteines in the gene product. However, non-cysteine altering pathogenic mutations have also been identified, making it challenging to discriminate pathogenic from benign NOTCH3 sequence variants. Experimental methods to discriminate between pathological and benign variant sequences are needed for clinical-genetic analysis of NOTCH3 and other polymorphic disease-related genes. Presented herein is a method for quantitative assessment of NOTCH3 mutants, the light chain split luciferase (LSL) assay. In LSL, NOTCH3 mutant fragments, cloned between a split luciferase open reading frame, are transfected into cells, producing secreted luciferase activity that is dependent on the normal structure of NOTCH3. Insertion of point mutants that cause CADASIL results in significantly lower activity. Using a panel of 47 sequences, the sensitivity and specificity was determined of LSL for pathogenic NOTCH3 mutation discrimination to be 100% and 93%. Application of LSL permitted ascertainment of prior work that cysteine alterations are the most important contributor to NOTCH3 protein pathology and that paired cysteine mutations have position-dependent suppressor effects on CADASIL NOTCH3 structural abnormalities. Two additional parameters from the LSL analysis (TCEP rescue of activity and secretion index), were also shown to be useful in characterizing NOTCH3 mutants, but these were less sensitive than total unreduced activity for pathogenic mutant determination. The spacing and primary sequence of the light chain module were shown to be an important component of the LSL assay, as a single light chain30275 / 70546cysteine is critical for pathogenic sequence discrimination. To assess the potential generalizability of the assay, the applicability of the assay was demonstrated for discrimination of mutations in FBN1 responsible for Marfan’s disease and Stiff Skin Syndrome (SSS).Furthermore, the activity of CADASIL mutant reporters is amplified by application of cysteinereactive iodoacetamide. As such, it is possible that LSL could be deployed to screen for novel compounds that suppress pathogenic conformations of NOTCH3.

[0045] CADASIL, the most common inherited cause of stroke and vascular dementia, results from a set of mutations in NOTCH3 (Wang, Handb Clin Neurol, 2018. 148: p. 733-743. and Chabriat et al., Lancet Neurol, 2009. 8(7): p. 643-53.). Stereotyped mutations in NOTCH3 drive CADASIL, the leading inherited cause of stroke and vascular dementia. The vast majority of these mutations result in alterations in the number of cysteines in the gene product. However, non-cysteine altering pathogenic mutations have also been identified, making it challenging to discriminate pathogenic from benign NOTCH3 sequence variants. Experimental methods to discriminate between pathological and benign variant sequences are needed for clinical-genetic analysis of NOTCH3 and other polymorphic disease-related genes. Molecular genetic analysis has suggested that the overwhelming majority of CADASIL mutations alter the number of cysteines of the gene product. The resulting mutant gene products contain an odd number of cysteines due to either loss or gain of a cysteine residue (Joutel et al., Lancet, 1997. 350(9090): p. 1511-5. And Rutten et al., Neurology, 2020. 95(13): p. el835-el843.). Notch proteins are highly conserved and composed of an array of EGF repeats, each consisting of three pairs of cysteines that form disulfide bonds (Wouters et al., Protein Sci, 2005. 14(4): p. 1091-103.). The changes in cysteine number in NOTCH3 linked to CADASIL has led to the hypothesis that mispairing of cysteine residues plays a key role in pathogenesis (Young et al., Commun Biol. 2022. 5(1): p. 331.).

[0046] A detailed analysis of the effects of CADASIL mutations on protein mobility was recently described for the first three EGF repeats of NOTCH3 (Lee et al., J Biol Chem, 2023: p.104838.). All CADASIL mutants tested showed retarded mobility in native gels under nonreducing conditions, indicating that gross structural alterations result from disease mutations. A comprehensive analysis demonstrated that 18 of 18 loss of cysteine mutants generated structural alterations; meanwhile, the vast majority of non-cysteine mutant residues did not affect gel30275 / 70546mobility. The residues to which lost cysteines were mutated did not affect the degree of mobility shift.

[0047] Furthermore, gel mobility assays enabled evaluation of potential suppressor residues (ie second cysteine mutations paired with a disease allele) that reverse the gel mobility effects of single cysteine mutations (Lee et al., J Biol Chem, 2023: p. 104838.). Notably, cysteine partners of the mutant cysteine residues were the most potent suppressors of mobility shifting. These studies validated the primacy of cysteine disulfides in maintaining gross structural integrity of the NOTCH3 protein and indicated that mispaired cysteines may participate in pathological structural alterations. Since gel mobility shifting of NOTCH3 can be performed in most basic labs and identifies pathogenic mutants with high sensitivity and specificity, it is predicted to be an accessible and specific method to query effects of individual NOTCH3 variants.

[0048] In the present disclosure, an alternate assay was developed to gel mobility shift experiments to assess the effects of variants in NOTCH3 protein. The new approach was motivated by the desire to 1) expand analysis beyond the first three EGF-like repeats that were the focus of gel shift assays; 2) develop methods to quantify the impact of disulfide bonding on NOTCH3 proteinopathy; 3) permit scale-up of an assay that could be used in high throughput screening.

[0049] Accordingly, herein a second accessible method is described for analyzing NOTCH3 protein variants that permits the quantification of pathological properties in NOTCH3: Light chain Split Luciferase (LSL). LSL is composed of a custom vector for expressing variant gene fragments and a measurement workflow; these components form a flexible experimental platform for variant gene product analysis. The applicability of LSL to differentiate CADASIL NOTCH3 mutants from wildtype was verified and then applied LSL to NOTCH3 variants across several EGF domains to quantify their pathological potential. The novel LSL workflow generates three parameters that distinguish the impact of NOTCH3 variants on disulfide bonding and protein secretion. Furthermore, LSL was adapted to variants of FBN1 that are linked to independent diseases and discuss potential uses of LSL for high throughput screening.Definitions

[0050] The plural herein shall equally denote the singular, and the singular shall equally denote the plural wherever reasonable. The words “for example” or “by way of example” or30275 / 70546similar phrases shall be interpreted as equivalent to “for example but not by way of limitation”, such that any example shall not limit the generality to which the example pertains.

[0051] The terms “approximately” and “about” refer to quantities that are within close range of a reference amount. With respect to polynucleotides, a sequence that is approximately / about a specified length is within 5% of the recited length.

[0052] The term “cysteine residue” refers to the amino acid cysteine either in isolation or as part of a polypeptide and / or protein.

[0053] The terms “detecting” or “determining” or “measuring” as used herein generally means identifying and / or quantifying the presence, absence, or folding status of a protein or polypeptide of interest. In various embodiments, detection signals are produced by the methods described herein, and such detection signals may be optical signals which may include but are not limited to, colorimetric changes, fluorescence, turbidity, mass-to-charge ratio of ions, and luminescence. Detecting or determining, in still other embodiments, also means quantifying a detection signal, and the quantifiable signal may include, but is not limited to, transcript number, amplicon number, protein number, and number of metabolic molecules. In this way, sequencing or bioanalyzers are employed in certain embodiments.

[0054] The term “disulfide bond” or “disulfide bridge” refers to covalent linkages formed between the sulfur atoms of two cysteine residues within a protein or polypeptide.

[0055] The term "expression vector" or “vector” refers to any molecule used to transfer coding information to a host cell. In various aspects, the expression vector is a nucleic acid, a plasmid, a cosmid. a virus, or an artificial chromosome.

[0056] The term "host cell" refers to a cell that has been transformed, transfected, or transduced by an expression vector described herein, which is then expressed by the cell. A host cell is, in various aspects, a prokaryotic or eukaryotic cell. In various aspects, the host cell is a bacteria cell, a protist cell, a fungal cell, a plant cell, or an animal cell. The term also refers to progeny of the parent host cell, regardless of whether the progeny is identical in genotype or phenotype to the parent, as long as the gene of interest is present.

[0057] In one embodiment, a reporter system such as NanoLuc is contemplated for use with the methods described in the present disclosure. The term “NanoLuc” refers to the NanoLuc30275 / 70546complementation protein system as used in the art. The system consists of two components termed LgBiT (18-kDa protein fragment) and SmBiT (11-amino-acid peptide fragment), which have been optimized for minimal self-association and stability. When LgBiT and SmBiT are optimally fused to two interacting proteins, or to different portions of a single protein, the formation of active luciferase is achieved when, for example, the protein of interest is properly folded.

[0058] The term “nucleic acid" refers to a polymer composed of a multiplicity of nucleotide units (ribonucleotide or deoxyribonucleotide or related structural variants) linked via phosphodiester bonds. Nucleic acids include RNA, cDNA, genomic DNA. In various embodiments, the RNA, cDNA, or genomic DNA is RNA, cDNA, or genomic DNA from a pathogen. In one embodiment, the DNA is cell-free DNA (cfDNA) which may be, for example, of tumor cell origin and therefore useful in detecting cancer.

[0059] The term “protein” refers to a polymer of amino acid residues, wherein a protein may be a single molecule or may be a multi-molecular complex. The term, as used herein, can refer to a subunit in a multi-molecular complex, polypeptides, peptides, oligopeptides, of any size, structure, or function. It is generally understood that a peptide can be 2 to 100 amino acids in length, whereas a polypeptide can be more than 100 amino acids in length. A protein may also be a fragment of a naturally occurring protein or peptide. The term protein may also apply to amino acid polymers in which one or more amino acid residues is an artificial chemical analogue of a corresponding naturally occurring amino acid. A protein can be wild-type, recombinant, naturally occurring, or synthetic and may constitute all or part of a naturally- occurring, or non-naturally occurring polypeptide. The subunits and the protein of the protein complex can be the same or different. A protein can also be functional or non-functional.

[0060] Generally, other nomenclature used herein and many of the laboratory procedures in cell culture, molecular genetics and nucleic acid chemistry and hybridization, which are described below, are those well-known and commonly employed in the art. (See generally Ausubel et al. (1996) supra; Sambrook et al, Molecular Cloning: A Laboratory Manual, Second Edition, Cold Spring Harbor Laboratory Press, New York (1989), which are incorporated by reference herein). Standard techniques are used for recombinant nucleic acid methods, polynucleotide synthesis, preparation of biological samples, preparation of cDNA fragments,30275 / 70546isolation of mRNA and the like. Generally enzymatic reactions and purification steps are performed according to the manufacturers' specifications.Expression constructs, vectors and host cells

[0061] A nucleic acid construct used to express the recombinant protein, such as NOTCH3 in one embodiment, can be one which is expressed extrachromosomally (episomally) in a transfected cell or one which integrates, either randomly or at a pre-selected targeted site through homologous recombination, into the host cell's genome. A construct which is expressed extrachromosomally comprises, in addition to recombinant protein-encoding sequences, sequences sufficient for expression of the protein in the cells and, optionally, for replication of the construct. It typically includes a promoter, recombinant protein-encoding DNA and a polyadenylation site. The DNA encoding the recombinant protein is positioned in the construct in such a manner that its expression is under the control of the promoter. Optionally, the construct may contain additional components such as one or more of the following: a splice site, an enhancer sequence, a selectable marker gene under the control of an appropriate promoter, an amplifiable marker gene under the control of an appropriate promoter, and a matrix attachment region (MAR) or other element known in the art that enhances expression of the region where it is inserted, and / or DNA homologous to genomic DNA in the recipient cell, to target integration of the DNA to a selected site in the genome (to target DNA or DNA sequences).

[0062] Any eukaryotic and prokaryotic vector is contemplated for use in the instant methods, including mammalian, yeast, fungal, insect, plant or viral vectors useful for a selected host cell. The term "vector" is used as recognized in the art to refer to any molecule (e.g., nucleic acid, plasmid, or virus) used to transfer coding information to a host cell. The term "host cell" is used to refer to a cell which has been transformed, or is capable of being transformed, by a vector bearing a selected gene of interest which is then expressed by the cell. The term includes mammalian, yeast, fungal, insect, plant and protozoan cells, and the progeny of the parent cell, regardless of whether the progeny is identical in morphology or in genetic make-up to the original parent, so long as the selected gene is present. In general, any vector can be used in methods of the invention and selection of an appropriate vector is, in one aspect, based on the host cell selected for expression of the expression construct.30275 / 70546

[0063] Examples include, but are not limited to, mammalian cells, such as Chinese hamster ovary cells (CHO) (ATCC No. CCL61); CHO DHFR-cells; serum-free, suspension-adapted CHO DHFR cell line was created at CMC ICOS (SFSA DG44 cells); human embryonic kidney (HEK) 293 or 293T cells (ATCC No. CRL1573); or 3T3 cells (ATCC No. CCL92). Other suitable mammalian cell lines, are the monkey COS-1 (ATCC No. CRL1650) and COS-7 (ATCC No. CRL1651) cell lines, and the CV-1 cell line (ATCC No. CCL70). Still other suitable mammalian cell lines include, but are not limited to, Sp2 / 0, NS1 and NSO mouse hybridoma cells, mouse neuroblastoma N2A cells. HeEa, mouse L-929 cells, 3T3 lines derived from Swiss, Balb-c or NIH mice, BHK or HaK hamster cell lines, which are also available from the ATCC.

[0064] Further exemplary mammalian host cells include primate cell lines and rodent cell lines, including transformed cell lines. Normal diploid cells, cell strains derived from in vitro culture of primary tissue, as well as primary explants, are also suitable.

[0065] Similarly useful as host cells include, for example, the various strains of E. coli (e.g., HB101, (ATCC No. 33694) DH5α, DH10, and MC1061 (ATCC No. 53338)), various strains of B. subtilis, Pseudomonas spp., Streptomyces spp., Salmonella typhimurium and the like.

[0066] Many strains of yeast cells known to those skilled in the art are also available as host cells for expression of a GOI and include, for example, Saccharomyces cerevisiae, Schizosaccharomyces pombe, Kluyveromyces strains, Candida, Pichia ciferrii and Pichia pastoris.

[0067] Additionally, where desired, insect cell systems may be utilized in the methods of the present invention. Such systems include for example and without limitation, Sf-9 and Hi5 (Invitrogen, Carlsbad, CA).

[0068] Exemplary fungal cells include, without limitation. Thermoascus aurantiacus, Aspergillus(filamentous fungus), including without limitation Aspergillus oryzaem, Aspergillus nidulans, Aspergillus terreus, and Aspergillus niger, Fusarium (filamentous fungus), including without limitation Fusarium venenatum, Penicillium chrysogenum, Penicillium citrinum, Acremonium chrysogenum., Trichoderma reesei, Mortierella alpina, and Chrysosporium lucknowense.30275 / 70546

[0069] Exemplary protozoan cells include without limitation Tetrahymena strains and Trypanosoma strains.

[0070] An expression plasmid according to the disclosure is further described in the following Example. The Example serves only to illustrate various embodiments of the present disclosure and is not intended to limit the scope of the invention in any way.Antibodies and fragments

[0071] As used herein, an "antibody" refers to a protein consisting of one or more polypeptides substantially encoded by immunoglobulin genes or fragments of immunoglobulin genes. The recognized immunoglobulin genes include the kappa, lambda, alpha, gamma, delta, epsilon and mu constant region genes, as well as myriad immunoglobulin variable region genes. Light chains are typically classified as either kappa or lambda. Heavy chains are typically classified as gamma, mu, alpha, delta, or epsilon, which in turn define the immunoglobulin classes, IgG, IgM, IgA, IgD and IgE, respectively.

[0072] A typical full-length (intact) immunoglobulin (antibody) structural unit is known to comprise a tetramer. Each tetramer is composed of two identical pairs of polypeptide chains, each pair having one "light" (about 25 kDa) and one "heavy" chain (about 50-70 kDa). The N-terminus of each chain defines a variable region of about 100 to 110 or more amino acids primarily responsible for antigen recognition. The terms variable light chain (VL) and variable heavy chain (VH) refer to these light and heavy chains respectively.

[0073] Antibodies exist as intact immunoglobulins or as a number of well-characterized fragments that can be produced, inter alia, by digestion with various peptidases. Thus, for example, pepsin digests an antibody below the disulfide linkages in the hinge region to produce F(ab)'2; a dimer of Fab which itself is a light chain joined to VH-CHI by a disulfide bond. The F(ab)'2 may be reduced under mild conditions to break the disulfide linkage in the hinge region thereby converting the (Fab')2 dimer into a Fab' monomer. The Fab' monomer is essentially a Fab with part of the hinge region (see, Fundamental Immunology, W. E. Paul, ed., Raven Press. N. Y. (1993), for a more detailed description of other antibody fragments). While various antibody fragments are defined in terms of the digestion of an intact antibody, one of skill will appreciate that such Fab' fragments may be synthesized de novo either chemically or by utilizing recombinant DNA methodology. Thus, the term antibody, as used herein also includes whole30275 / 70546antibodies, antibody fragments either produced by the modification of whole antibodies or synthesized de novo using recombinant DNA methodologies.

[0074] The terms “antibody fragment,” “antigen-binding fragment,” or “antibody binding domain” refer to at least one portion of an antibody, or recombinant variants thereof, that contains the antigen-binding domain, i.e., an antigenic determining variable region of an intact antibody, that is sufficient to confer recognition and specific binding of the antibody fragment to a target, such as an antigen and its defined epitope. Examples of antigen-binding fragments include, but are not limited to, Fab, Fab’, F(ab’)2, and Fv fragments, single-chain (sc)Fv (“scFv”) antibody fragments, linear antibodies, single domain antibodies such as sdAb (either VL or VH), camelid VHH domains, and multi-specific antibodies formed from antibody fragments.

[0075] Antibody fragments can be made by various techniques, including but not limited to proteolytic digestion of an intact antibody as well as production by recombinant host cells (e.g., E. coli or phage), as described herein.Split reporters

[0076] The term “detection moiety” or “reporter” refers to a moiety capable of detecting as described above. In some embodiments, a detection moiety can comprise a selectable splitmarker gene and / or split-reporter gene to facilitate identification and selection of construct assembly that have been transduced with the vector. In some embodiments, a detection moiety comprises luciferase, beta-lactamase, and / or GFP. In some embodiments, a detection moiety is a polypeptide whose expression results in an easily detectable property, such as enzymatic activity. In some embodiments, a detection moiety can be detected by energy transfer-based methods like bioluminescence resonance energy transfer (BRET). In some embodiment, a detection moiety comprises a detectable enzyme or signal, such as. NanoLuc,

[0077] In some embodiments, the present disclosure can further comprise a selectable splitmarker gene and / or split-reporter gene to facilitate identification and selection of construct assembly that have been transduced with the vector. Either selectable split-marker or splitreporter gene may be flanked with appropriate regulator sequences to allow expression in host cells. Examples of selectable split-markers include, without limitation, luciferase, betalactamase, and GFP.30275 / 70546

[0078] In some embodiments, split-reporter genes may be used for identifying transduced cells and for evaluating the functionality of regulatory sequences. As disclosed herein, a split-reporter gene is a gene that is not present in or expressed by the recipient organism or tissue and that encodes a polypeptide whose expression results in an easily detectable property, such as enzymatic activity and when in proximity, the two split pieces recover detectable activity.Expression of the split-reporter gene can be assayed at a suitable time after the nucleic acid has been introduced into the recipient cells. Examples of split-reporter genes include, without limitation, genes encoding for luciferase, genes encoding for beta-lactamase, and genes encoding for green fluorescent protein. Suitable expression systems are well known in the art and may be prepared using known techniques or obtained commercially. In some embodiments, a construct with a minimal 5' flanking region showing the highest level of expression of the reporter gene is identified as the promoter. Such promoter regions may be linked to a reporter gene and used to evaluate agents for the ability to modulate promoter-driven transcription. While energy transferbased methods like bioluminescence resonance energy transfer (BRET) are one way to assess molecular interactions, split-reporters are adapted to complementation-based methods that rely on the interaction of subunits to reconstitute the detectable enzyme or signal. For example, with NanoLuc, interaction of LgBiT with a small complementary peptide, an active enzyme is formed that generates a bright luminescent signal in the presence of substrate. For the study of protein: protein interactions or, as used in the present disclosure, the study of protein folding as it relates to pathogenic mutations, the complementary peptide is in one embodiment Small BiT (SmBiT, 11 amino acid peptide), which has been optimized to have low affinity for LgBiT. The two subunits only form an enzyme when fused to target proteins and the two target proteins interact, or when fused to different portions of the same protein that is properly folded, bringing SmBiT and LgBiT together, enabling a sensitive method to assay protein association / disassociation kinetics and folding / misfolding using a simple bioluminescent signal. Proteins of interest

[0079] In various embodiments, engineered cells express a protein of interest. Polypeptides and proteins of interest can be of scientific or commercial interest, including protein-based therapeutics. Proteins of interest include, among other things, secreted proteins, non-secreted proteins, intracellular proteins or membrane-bound proteins. Polypeptides and proteins of interest may be referred to as “recombinant proteins.” The expressed protein(s), including fusion proteins30275 / 70546described herein, may be produced intracellularly or secreted into the culture medium from which it can be recovered and / or collected and / or assayed for folding properties as described herein. The term “isolated protein” or “isolated recombinant protein” refers to a polypeptide or protein of interest, that is purified away from proteins or polypeptides or other contaminants that would interfere with its therapeutic, diagnostic, prophylactic, research or other use.

[0080] By “purifying” is meant increasing the degree of purity of the protein in the composition by removing (partially or completely) at least one product-related impurity from the composition. Recovery and purification of proteins is accomplished by any downstream process, particularly the harvest operation, resulting in a more “homogeneous” protein composition that meets yield and product quality targets (such as reduced product-related impurities and increased product quality).

[0081] As used herein, the term “isolated” means (i) free of at least some other proteins or polynucleotides with which it would normally be found, (ii) is essentially free of other proteins or polynucleotides from the same source, e.g., from the same species, (iii) separated from at least about 50 percent of polypeptides, polynucleotides, lipids, carbohydrates, or other materials with which it is associated in nature, (iv) operably associated (by covalent or noncovalent interaction) with a polypeptide or polynucleotide with which it is not associated in nature, or (v) does not occur in nature.

[0082] Proteins of interest include, in one embodiment, Neurogenic locus notch homolog protein 3 (NOTCH3).

[0083] Proteins of interest include antigen-binding proteins. Antigen-binding protein refers to proteins or polypeptides that comprise an antigen-binding region or antigen-binding portion that has affinity for another molecule to which it binds (antigen). Antigen-binding proteins encompass antibodies, peptibodies, antibody fragments, antibody derivatives, antibody analogs, fusion proteins (including single-chain variable fragments (scFvs) and double-chain (divalent) scFvs), muteins, multispecific proteins, and bispecific proteins. An scFv is a single chain antibody fragment having the variable regions of the heavy and light chains of an antibody linked together. See U. S. Patent Nos. 7,741,465, and 6,319,494 as well as Eshhar et al., Cancer Immunol Immunotherapy (1997) 45: 131-136. An scFv retains the parent antibody's ability to specifically interact with target antigen.30275 / 70546

[0084] In some embodiments, proteins of interest may include proteins that bind specifically to one or more CD proteins, lipid-binding proteins such as ApoE, HER receptor family proteins, cell adhesion molecules, growth factors, nerve growth factors, fibroblast growth factors, transforming growth factors (TGF), insulin-like growth factors, osteoinductive factors, insulin and insulin-related proteins, coagulation and coagulation-related proteins, colony stimulating factors (CSFs). other blood and serum proteins blood group antigens: receptors, receptor-associated proteins, growth hormones, growth hormone receptors, T-cell receptors; neurotrophic factors, neurotrophins, relaxins, interferons, interleukins, viral antigens, lipoproteins, integrins, rheumatoid factors, immunotoxins, surface membrane proteins, transport proteins, homing receptors, addressins, regulatory proteins, and immunoadhesins.

[0085] In some embodiments, a reference protein is a wild-type protein or a mutant protein. In some embodiments, a mutant protein is associated with a known phenotype. In some embodiments, the known phenotype is associated with a pathogenic or benign pathology.

[0086] In some embodiments, a protein of interest comprises one or more disulfide bonds. Secreted and cell-surface proteins are often stabilized by disulfide bonds. Early in protein folding, disulfide formation is error-prone; the wrong cysteines are connected, or the correct cysteines are paired but in a temporal order that inhibits folding. Cells correct for this with a specialized redox environment in the endoplasmic reticulum (ER), equipped with catalysts of disulfide formation and isomerization (Wilkinson et al., Biochim Biophys Acta, 1699(1-2): 35-44, 2004). The presence of properly formed disulfide bonds in a polypeptide gene product is an indication that it is correctly folded and presumptively active.

[0087] In some embodiments, a construct (e.g.. a fusion protein as described herein) is treated with a reducing agent. In some embodiments, a reducing agent comprises tris(2-carboxyethyl)phosphine (TCEP). In some embodiments, the reducing agent is used to break a disulfide bond between the protein of interest and a rabbit light chain. In some embodiments, breaking the disulfide bonds between the protein of interest and the rabbit light chain using a reducing agent allows two split reporter portions as described herein to come together as a “rescue.” In some embodiments, a reducing agent comprises a control for proteins that comprise a negative signal when assembled.30275 / 70546

[0088] In some embodiments a fusion protein comprises, from N-terminus to C-terminus: the protein sequence of interest, the first detection moiety, the antibody fragment, and the second detection moiety. In other embodiments, a fusion protein comprises, from N-terminus to C-terminus: the antibody fragment, the first detection moiety, the protein sequence of interest, and the second detection moiety. In other embodiments, the four elements the antibody fragment, the first detection moiety, the protein sequence of interest, and the second detection moiety can be in any order. In still other embodiments, an antibody fragment in any of the aforementioned constructs is replaced with another protein sequence useful according to the present disclosure, such as ApoE.

[0089] In some embodiments, linkers may be included between one or more elements of a fusion protein. These linkers may improve assay sensitivity and / or specificity. In some embodiments, a linker may be short sequences that provide necessary flexibility for split luciferase (or other reporter system) components to assemble. Exemplary linker sequences are described herein. In various embodiments, the linker may be 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 or more residues.

[0090] In some embodiments, a method provided by the present disclosure comprises calculating a detection ratio by, for example, dividing a reference protein detection signal by a fusion protein detection signal, wherein a calculated detection ratio value above a threshold value corresponds to a properly-folded fusion protein and wherein a calculated detection ratio value below a threshold value corresponds to a misfolded fusion protein.

[0091] In some embodiments the threshold value is 0, 1, 2, 5, or 10. In some embodiments, the threshold value may be specific to the protein of interest. For example, with CADASIL, the WT / pathogenic ratio cutoff for pathogenic may be 2 in one embodiment. In other embodiments, a detection ratio depends on the specific protein of interest. In some embodiments, the detection ratios could be determined based on comparison of known benign vs known pathogenic variants for a protein of interest.

[0092] In some embodiments, the present disclosure provides an expression construct comprising one or more of the following components: a NOTCH protein or polypeptide comprising SEQ ID NO: 2 or a FBN1 protein or polypeptide sequence comprising SEQ ID NO: 10, SEQ ID NO: 12, or SEQ ID NO: 14; a rabbit light chain comprising any one of SEQ ID NOs:30275 / 705463-8; a split luciferase comprising SEQ ID NO: 15 and SEQ ID NO: 16, and optionally one or more linker sequences including, for example, VD and / or RS between an aforementioned component. In other embodiments, a protein of interest such as ApoE, optionally with a linker, is part of the construct. In various embodiments, the order of the components, from N-terminus to C-terminus, can be in any order. As described herein, the protein of interest, including, for example, a NOTCH3 protein or a FBN1 protein or another protein of interest, may comprise a wildtype sequence or may comprise a mutant sequence. In some embodiments, the protein of interest comprises a sequence comprising, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or more mutations relative to a wildtype sequence. In some embodiments, the sequences of the components are provided below.Thiol Alkylators

[0093] In some embodiments a method further comprises: 1) treating cells with one or more thiol alkylators; and m) repeating steps a) through k). In some embodiments, the one or more thiol alkylators include: 2-Iodoacetamide (IAM); 2-(butan-2-yldisulfanyl)-lH-imidazole (PX-12); N-acetylcysteine amide; N-ethylmaleimide (NEM); Disulfiram (DSF); 4-(2-Aminoethyl)benzenesulfonyl fluoride hydrochloride (AEBSF); Ebselen; Carmofur; Rabeprazole; Tenatoprazole; Omeprazole; Lansoprazole; Pantoprazole; Auranofin; Spebrutinib; Evobrutinib; Osimertinib; Ibrutinib; Zanubrutinib; Necrosulfonamide; Sortorasib; or a combination thereof. In preferred embodiments, the one or more thiol alkylators comprise iodoacetamide, PX-12, N-ethylmaleimide, disulfiram, ebselen, carmofur, auranofin, spebrutinib, osimertinib, necrosulfonamide, or a combination thereof.

[0094] In some preferred embodiments, treating cells with one or more thiol alkylators further comprises administering the one or more thiol alkylators at a concentration of about O. OOlpM, about O. OlpM, about O.lpM, about 1μM, about 2 pM, about 3 pM, about 4 pM, about 5 pM, about 6 pM, about 7 pM, about 8 pM, about 9 pM, about 10 pM, about 20 pM, about 100 pM, or about 1000 pM. In some preferred embodiments, treating cells with one or more thiol alkylators further comprises administering the one or more thiol alkylators at a concentration of about 1μM, about 2.5μM, about 5μM, or about 10μM.Exemplary Sequences

[0095] LSL-E3: (83G (bold)-SBT (underlined)-Notch3 (italicized)-LBT (bold underlined))30275 / 70546

[0096] MDTRAPTQLLGLLLLWLPGATFAQVLTQTPSSVSAAVGGTVTINCQSSQS VGNNNRLAWFQQKPGQPPKRLIYDASTLESGVPSRFKGSGSGTQFTLTISDLECDD AATYYCTGGYSGGIVAFGGGTEVVVKVTGYRLFEEILVDAPPCLDGSPCAAGGRCTQL PSREAACLCPPGWVGERCQLEDPCHSGPCAGRGVCQSSWAGTARFSCRCPRGFRGPDCSL PDPCLSSPCAHGARCSVGPDGRFLCSCPPGYQGRSCRSDVDERSNFTLEDFNGDyVEOTA AYNLDQVLEOGGVSSLLONLAVSVTPIORIVRSGENALKIDIHVIIPYEGLSADQMA QIEEVFKVVYPVDDHHFKVILPYGTLVIDGVTPNMLNYFGRPYEGIAVFDGKKITVT GTLWNGNKIIDERLITPDGSMLFRVTINS (SEQ ID NO: 1)

[0097] Exemplary Notch3 sequence

[0098] APPCLDGSPCANGGRCTQLPSREAACLCPPGWVGERCQLEDPCHSGPCAGRG VCQSSVVAGTARFSCRCPRGFRGPDCSLPDPCLSSPCAHGARCSVGPDGRFLCSCPPGYQ GRSCRSDVDE (SEQ ID NO: 2)

[0099] Light Chain 44F

[0100] MDTRAPTQLLGLLLLWLPGATFAQVLTQTASSVSAAVGGTVTINCQASQSVY DNNFLSWYQQKPGQPPKLLIYSASTLASGVPSRFKGSGSGTQFILTINDLECDDAATYYC AGGYSINSGELYVFGGGTEVVVK (SEQ ID NO: 3)

[0101] Light Chain 69B

[0102] MDTRAPTQLLGLLLLWLPGARCAYDMTQTPASVEVAVGGTVTINCQASQSIS TALSWYQQKPGQPPKLLIYLASTLASGVPSRFKGSGSGTQFTLTISDLECDDAATYYCQQ TYSYSNVDNTFGGGTEVVVK (SEQ ID NO: 4)

[0103] Light Chain 82A

[0104] MDTRAPTQLLGLLLLWLPGATFAQVLTQTASPVSAAVGSTVTINCQSSQIVY NNNRLSWFQQKPGQPPKQLIYDASILASGVSSRFKGSGSGTQFTLTISDVQCDDAATYYC LGSYDCSSADCRAFGGGTEVVVK (SEQ ID NO: 5)

[0105] Light Chain 120B

[0106] MDTRAPTQLLGLLLLWLPGATFAAVLTQTPSPVSAAVGGTVSISCQASQSVA NNNWLSWFQQKPGQPPKLLIYHASTLETGVPSRFKGSGSGTQFTLTISDVQCDDAATYY CLGAYSGGSDNAFGGGTEVVVK (SEQ ID NO: 6)30275 / 70546

[0107] Light Chain 123D

[0108] MDTRAPTQLLGLLLLWLPGATFTIVMTQTPSSRSVPVGDTVTINCQATRSVYN NNHLAWYQQKPGQPPKLLIYSASTLASGVPSRFKGSGSGTQFTLTLSDVVCADAATYYC AGYNNGTDGFAFGGGTEVVVK (SEQ ID NO: 7)

[0109] Light Chain 145H

[0110] MDTRAPTQLLGLLLLWLPGATFAQVMTQTPSSTSAVVGGTVTINCQSSQSVY NNNRLSWYQQKPGQPPKLLIYYASKLASGVPSRFKGSGSGTQFTLTISDLECDDAATYY CLGSYDWSNSDCIAFGGGTEVVVK (SEQ ID NO: 8)

[0111] Exemplary linker sequence between SBT and NOTCH3

[0112] VD

[0113] Exemplary linker sequence between NOTCH3 and LBT

[0114] RS

[0115] LSL-FBNL Marfan (83G (bold)-SBT(underlined) -Marfan EGF11-13 (italicized)-LBT (bold and underlined))

[0116] MDTRAPTQLLGLLLLWLPGATFAQVLTQTPSSVSAA VGGTVTINCQSSQS VGNNNRLAWFQQKPGQPPKRLIYDASTLESGVPSRFKGSGSGTQFTLTISDLECDD AATYYCTGGYSGGIVAFGGGTEVVVKVTGYRLFEEILVD / NEC^M / PSLCTHG^CWT / GSFKCRCDSGFALDSEERNCTDIDECRISPDLCGRGQCVNEPGDFECKCDEGYESGFMMMK NCMDIDECQRDPLLCRGGVCHNTEGSYRCECPPGHQLSPNISACIDINERSVFTLEDFVGD WEOTAAYNLDOVLEOGGVSSLLONLAVSVTPIQRIVRSGENALKIDIHVIIPYEGLS ADQMAQIEEVFKVVYPVDDHHFKVILPYGTLVIDGVTPNMLNYFGRPYEGIAVFDGKKITVTGTLWNGNKIIDERLITPDGSMLFRVTINS (SEQ ID NO: 9)

[0117] Exemplary FBNL Marfan EGF 11-13 sequence

[0118] INECKMIPSLCTHGKCRNTIGSFKCRCDSGFALDSEERNCTDIDECRISPDLCG RGQCVNTPGDFECKCDEGYESGFMMMKNCMDIDECQRDPLLCRGGVCHNTEGSYRCE CPPGHQLSPNISACIDINE (SEQ ID NO: 10)

[0119] LSL-FBNL Marfan (83G (bold)-SBT (underlined)-Marfan EGF13-15 (italicized)-LBT (bold and underlined))30275 / 70546

[0120] MDTRAPTQLLGLLLLWLPGATFAQVLTQTPSSVSAAVGGTVTINCQSSQS VGNNNRLAWFQQKPGQPPKRLIYDASTLESGVPSRFKGSGSGTQFTLTISDLECDD AATYYCTGGYSGGIVAFGGGTEVVVKVTGYRLFEEILVDCQRDPLLCRGGVCfflVrEGS YRCECPPGHQLSPNISACIDINECELSAHLCPNGRCVNLIGKYQCACNPGYHSTPDRLFCVDI DECSIMNGGCETFCTNSEGSYECSCOPGFALMPDQRSCTDIDE S FTLEBFVGDWEQTA AYNLDQVLEOGGVSSLLONLAVSVTPIORIVRSGENALKIDIHVIIPYEGLSADQMA QIEEVFKVVYPVDDHHFKVILPYGTLVIDGVTPNMLNYFGRPYEGIAVFDGKKITVT GTLWNGNKIIDERLITPDGSMLFRVTINS (SEP ID NO: 11)

[0121] Exemplary FBN1, Marfan EGF 13-15 sequence

[0122] CQRDPLLCRGGVCHNTEGSYRCECPPGHQLSPNISACIDINECELSAHLCPNG RCVNLIGKYQCACNPGYHSTPDRLFCVDIDECSIMNGGCETFCTNSEGSYECSCQPGFAL MPDQRSCTDIDE (SEQ ID NO: 12)

[0123] LSL-FBN1, Stiff Skin (83G (bold)-SBT (underlined)-SSS (italicized)-LBT (bold and underlined))

[0124] MDTRAPTQLLGLLLLWLPGATFAQVLTQTPSSVSAAVGGTVTINCQSSQS VGNNNRLAWFQQKPGQPPKRLIYDASTLESGVPSRFKGSGSGTQFTLTISDLECDD AATYYCTGGYSGGIVAFGGGTEVVVKyTGYRLFEEILVDDTR GACFLDZRPRGDAGD TACSNEIGVGVSKASCCCSI. GKAWGTPCEMCPAVNTSEYKn. CPGGEGFRPNPITVn. EDIDEC QELPGLCQGGKCINTFGSFQCRCPTGYYLNEDTRVCDDVNERSVFTLEDFVGDWEQTAAY NLDQVLEQGGVSSLLQNLAVSVTPIQRIVRSGENALKIDIHVIIPYEGLSADQMAQIEEVF KVVYPVDDHHFKVILPYGTLVIDGVTPNMLNYFGRPYEGIAVFDGKKITVTGTLWNGNK IIDERLITPDGSMLFRVTINS (SEQ ID NO: 13)

[0125] Exemplary FBN1, Stiff Skin SSS sequence

[0126] DTRSGNCYLDIRPRGDNGDTACSNEIGVGVSKASCCCSLGKAWGTPCEMCPA VNTSEYKILCPGGEGFRPNPITVILEDIDECQELPGLCQGGKCINTFGSFQCRCPTGYYLN EDTRVCDDVNE (SEQ ID NO: 14)

[0127] Exemplary linker sequence between SBT and FBN1

[0128] VD

[0129] Exemplary linker sequence between FBN1 and LBT30275 / 70546

[0130] RS

[0131] Exemplary SBT sequence

[0132] VTGYRLFEEIL (SEQ ID NO: 15)

[0133] Exemplary LBT sequence

[0134] VFTLEDFVGDWEQTAAYNLDQVLEQGGVSSLLQNLAVSVTPIQRIVRSGENA LKIDIHVIIPYEGLSADQMAQIEEVFKVVYPVDDHHFKVILPYGTLVIDGVTPNMLNYFG RPYEGIAVFDGKKITVTGTLWNGNKIIDERLITPDGSMLFRVTINS (SEQ ID NO: 16)EXAMPLESExample 1 - Light chain Split Luciferase assay implicates pathological NOTCH3 thiol reactivity in inherited cerebral small vessel diseaseMaterials and Methods

[0135] DNA constructs. Expression constructs were built in the pCMV-Sport6 vector. The basic LSL construct was cloned in sequential steps that combined synthetic genes, PCR, and standard ligation. Antibody light chains were produced by either PCR or gene synthesis of rabbit variable domains of IgG light chain sequences that were originally generated against a cleavage fragment of NOTCH3 (Zhang et al., Sci Rep, 2021. 11 (1 ): p. 17246.). The gene synthesis or PCR approach incorporated SmBiT followed by a cloning site that enabled ligation of NOTCH3 fragments into the C-terminus of the SmBiT, whose stop codon was removed. LgBiT was synthesized by PCR and cloned by ligation; a stop codon was inserted at the C-terminus of this cassette.

[0136] Point mutations in NOTCH3 were derived by PCR of templates from mutants described (Lee et al., J Biol Chem, 2023: p. 104838.) or by nested PCR using oligos harboring specific mutations. DNA fragments were digested with appropriate restriction enzymes and cloned by ligations using T4 ligase. All clones were validated by sequencing of the inserted mutant genes.

[0137] Light chain mutation clones were generated by nested PCR onto the 83G backbone. Small insertions and mutations were created by incorporating corresponding alterations in oligonucleotides used for DNA amplification. Addition of the constant domain of the 83G light chain was accomplished by PCR.30275 / 70546

[0138] Cell culture. HEK293 cells were propagated in DMEM with 10% fetal bovine serum in 5% carbon dioxide chambers. Transfection was performed using PolyJet as recommended by the manufacturer. Briefly. 400ng of LSL DNA and 100ng of iRFP expression plasmid were mixed in 25μL serum free DMEM. DMEM (25μL) with 1.5μL PolyJet was mixed with diluted DNA and then added to media of cells in 24 well plates. After 6-12 hours, the media was changed to OptiMEM. After a further 36-48 hours, the conditioned media was analyzed for nanoLuc activity.

[0139] LSL analysis workflow. In all experiments, reference groups were included with the appropriate wildtype NOTCH3 sequence cloned into the LSL vector. Parameter 1 (secreted nanoLuc activity normalized to iRFP) was determined by mixing 25pL conditioned media with 6.25 μL reaction mixture in a white 96 well plate. Luciferase activity was determined on a plate reading luminometer (BioTek Synergy LX multi-mode reader). In noted experiments, the Luc / iRFP ratio was normalized to wildtype that was set to 1.0. For some experiments, Parameters 2 and 3 were determined as follows. For Parameter 2, 2 μL of TCEP (31.25mM) was added to the reaction mixture and repeated measurements of luciferase activity were determined over 30 minutes. A reaction curve was fitted to the equation: Y=Y0+(Plateau-Y0)*(1-exp(-K*x)) and the plateau value was determined. This was considered the Max level, which the initial value before adding TCEP was deemed the Min level. The Min / Max ratio was determined, with the working model that it was the percentage of protein in normal conformation relative to total protein. In noted experiments, the Min / Max was normalized to wildtype that was set to 1.0. For Parameter 3, cell lysates were prepared in 40μL Lumit Immunoassay Lysis Detection Kit buffer and 0.5 to 5 μL of lysate was assayed for nanoLuc activity. The assay was then treated with TCEP and the plateau level of activity was determined as in Parameter 2. The ratio of TCEP treated secreted activity to lysate activity was considered the secretion index for the LSL protein. In noted experiments, the Secretion Index was normalized to wildtype that was set to 1.0.

[0140] Statistics. Normality was determined by Shapiro-Wilk test. Significant differences were determined using one-way ANOVA with Dunn’s multiple comparisons test using GraphPad Prism 8.0.0 (224). p value < 0.05 was considered statistically significant.30275 / 70546Results

[0141] Light chain split luciferase assay (LSL) vector design. The overall objective was to generate a quantification tool to differentiate benign from pathogenic NOTCH3 protein. EGF-like domains, the site of all mutations that cause CADASIL, are conserved modules that each contain six cysteines that form three disulfide bonds (Wouters et al., Protein Sci, 2005. 14(4): p.1091-103. and Luca et al.. Science, 2017. 355(6331): p. 1320-1324.). The bonds are predicted to scaffold the peptide chain into a compact structure. A vast majority of pathogenic mutations in NOTCH3 disrupt cysteine number (Joutel et al., Lancet, 1997. 350(9090): p. 1511-5. and Joutel et al., Nature, 1996. 383(6602): p. 707-10.), which is predicted to distort disulfide bonding and to disrupt EGF-like domain structure. We reasoned that a split luciferase readout, incorporating NOTCH3 between luciferase sequences, could differentiate wildtype NOTCH3 protein sequences and mutant NOTCH3 sequences by virtue of differences in protein structure imparted by sequence variants (Fig. 1A).

[0142] The new split luciferase assay for investigation of NOTCH3 variants includes a genetic fusion of (1) light chain variable domain (L) of a rabbit antibody, (2) the SmBiT domain (S) of split nanoluciferase (NanoLuc) (Dixon et al., ACS Chem Biol, 2016. 11(2): p. 400-8.); (3) NOTCH3 sequence variants; and (4) the LgBiT domain (L) of NanoLuc (Dixon et al., ACS Chem Biol, 2016. 11(2): p. 400-8.) (Fig. IB; left side). Initial characterization of the LSL system was performed on NOTCH3 EGF domains 1-3, the site of both pathological and benign mutations in human NOTCH3. The system for evaluating NOTCH3 will be referred to as LSL-NOTCH3, and constructs featuring insertion of coding sequences NOTCH3 EGF domains 1-3 are referred to as LSL-NOTCH3-E3. After transfection of the LSL-NOTCH3-E3, secreted protein activity can be quantified for luciferase activity. The efficiency of transfection is normalized by co-transfection with iRFP (Filonov et al., Nat Biotechnol, 2011. 29(8): p. 757-61. and Hock et al., Cell Cycle, 2014. 13(2): p. 220-6.), which can be quantified by plate scanning without cell destruction (Fig. IB).

[0143] The light chain antibody fragment was originally conceived to promote secretion of the recombinant protein to the media and to provide a target for commonly available antibodies that enable quantification of recombinant protein. The choice of the split NanoLuc was motivated by its high level of activity (Boute et al., Front Pharmacol, 2016. 7: p. 27. and Hall et30275 / 70546al., ACS Chem Biol, 2012. 7(11): p. 1848-57.) and prior work showing that spatial approximation of SmBiT and LgBiT sequences reconstitutes activity (Dixon et al., ACS Chem Biol, 2016. 11(2): p. 400-8.). Both NanoLuc polypeptides are devoid of cysteines and predicted to be insensitive to reducing agents. The two sequences were cyclically permuted so that SmBiT is produced N-terminal to LgBiT, which has been effective in other split protein systems (Guo et al., Nat Commun, 2022. 13(1): p. 789).

[0144] The genetic fusion LSL-NOTCH3-E3 generates protein with high enzymatic activity detectable in secreted media (Fig. 1C; construct 6). No activity was generated above background when SmBiT or LgBiT was deleted (Fig. 1C; constructs 1-4), but the LSL-NOTCH3-E3 activity was lower than that generated by fusion of NOTCH3 sequences to intact nanoLuc (Fig. 1C; constructs 8-11). These studies are consistent with wildtype NOTCH3 EGF domains 1-3 folding in a manner which would permit spatial alignment of SmBiT and LgBiT, reflected by reconstitution of luciferase activity.

[0145] Light chain split luciferase assay (LSL) for quantification of abnormalities in NOTCH3. The effects of pathogenic mutations on LSL-NOTCH3 activity were compared the activity generated by a prototypical CADASIL-causing R90C gain of cysteine mutation (Fig. 1C). Mutant R90C LSL-NOTCH3 transfected cells produced markedly suppressed levels of luciferase activity (normalized to iRFP) that was consistent with the blocking of secreted LSL activity by conformational changes of the split luciferase module (Fig 1C; construct 7). High level activity was generated by mutant R90C sequences fused to intact nanoLuc, demonstrating that the suppressive action of the mutant largely depended on the split luciferase domain (Fig. 1C; constructs 9 and 11). There was a 47-fold decrease in LSL-NOTCH3 activity when the CADASIL mutant was compared to wildtype (Fig. 1C; right side).

[0146] To test whether the LSL-NOTCH3 system was capable of broadly differentiating pathogenically altered protein from wildtype and benign variants of NOTCH3 were generated a series of pathogenic and benign sequence variants that were transfected into cells (Fig. 2A). The conditioned media were tested for secreted luciferase activity (normalized to transfected iRFP levels), which was compared to wildtype activity. As shown in Fig. 2B (showing wildtype value set to 1), all pathogenic LSL-NOTCH3 variants generated significantly decreased levels of luciferase activity. In contrast, benign variants generated levels that were similar to wildtype.30275 / 70546

[0147] Secreted LSL-NOTCH3 activity was a parameter that readily discriminates between benign and pathogenic mutants (referred to as parameter 1). Two additional parameters generated by the LSL system will be described later.

[0148] Effect of specific amino acid changes on LSL-NOTCH3 activity. Previous gel shift studies highlighted the importance of cysteines in mutant NOTCH3 protein structure (Lee et al., J Biol Chem, 2023: p. 104838.). To test the concordance between LSL and structural (gel shift) assays, were generated LSL constructs that corresponded with gel shift experiments from (Lee et al., J Biol Chem, 2023: p. 104838.) to test 1) the effect of different residues in positions 49, 75, 90 and 146 (Fig. 3); and 2) the potential suppressor effects of second cysteine mutations within EGF repeats (Fig. 4).

[0149] Fig. 3 showed results of single mutant analysis of LSL-NOTCH3-E3 activity. The effects of mutations at residues 49 and 146, which are cysteine in the wildtype protein and are the sites of CADASIL mutations C49Y and C146R, were analyzed. All non-cysteine replacements at residue 49 yielded decreased secreted luciferase activity that was significantly changed from wildtype (Fig. 3A). All non-cysteine replacements at residue 146 resulted in a large drop-in activity compared to wildtype, with several, but not all, of these significant (Fig.3D).

[0150] Although most CADASIL mutations cause a gain or loss of cysteine, there are notable exceptions which include the R75P mutation that has been described in Asian populations.NOTCH3 migrates abnormally in gels when position 75 was mutated to proline, cysteine, or glycine, indicating that selective changes trigger protein structural alterations that may relate to pathology, all twenty amino acids into position 75 in LSL-NOTCH3-E3 were introduced and assessed secreted luciferase activity (Fig. 3C). Only proline, cysteine, and glycine mutants induced reductions in activity, confirming the selective nature of changes in this position that result in structural alterations of NOTCH3.

[0151] In Figs. 1-2, a pathological gain of a cysteine at position 90 corresponding to the canonical R90C CADASIL mutation resulted in marked suppression of secreted luciferase activity. Gel shift abnormalities resulted only from mutations to cysteine at this location. In LSL-NOTCH3 experiment where position 90 was mutated to all possible amino acids (Fig. 3B), significantly lowered activity was found with mutations to cysteine and to tryptophan; all other30275 / 70546residues resulted in activities that were not significantly different from wildtype. This largely matches gel shifting data and again emphasizes the importance of cysteine changes in driving NOTCH3 structural alterations.

[0152] Fig. 4 shows experiments in which we used LSL to potentially confirm and extend prior work that led to discovery of suppressor mutations that reverse CADASIL abnormalities. Cysteine mutations may drive CADASIL pathology by inducing a cysteine imbalance in which free thiol groups are free to react with cysteines within NOTCH3 or within other proteins via new, dysfunctional disulfide bonds (Young et al., Commun Biol, 2022. 5(1): p. 331.). If that is the case, CADASIL mutant proteins that harbor second mutations of specific cysteines are predicted to suppress dysfunctional conformations. EGF repeat-specific second mutations in cysteine residues were identified that reversed gel shifts caused by CADASIL mutations C49Y, R90C, and C155Y (Lee et al., J Biol Chem, 2023: p. 104838.). These mutants were cloned and all potential suppressor cysteine mutations into LSL-NOTCH3-E3 to quantify the impact of double mutations (relative to single CADASIL mutations). As shown in Fig. 4A, in the context of the C49Y mutant (which results in the loss of the second cysteine of EGF1), the most effective second mutations that reversed suppressed LSL activity was a cysteine to serine alteration in the fourth cysteine residue of EGF1. This residue is normally paired with the second cysteine. Other loss of cysteine second mutations also resulted in significant increases in suppressed luciferase activity but suppressed the mutant defect to a much lower degree. For the gain of cysteine mutation R90C mutant (Fig. 4B), all second cysteine mutants modestly suppressed mutation-related effects on the LSL reporter, with none exhibiting significant superiority over others. For the loss of cysteine mutant C155Y, which causes a loss of the sixth cysteine of EGF3, all cysteine mutations in EGF3 counteracted the loss of LSL activity caused by the CADASIL mutations (Fig. 4C).However the mutation in the fifth cysteine, which normally pairs with the Cl 55 residue, was the only one that completely restored reporter activity. Overall, results from the LSL assay demonstrate that mutation of the cognate cysteine that is altered in CADASIL had a marked effect on activity of the reporter, which correlates with structural investigations using gel shifts that support the same model. In addition to identifying the most potent suppressor mutations, use of LSL appears to support that other residues have partial suppressor effects, and suggests that the assay may be more sensitive than gel shift assays.30275 / 70546

[0153] Role of disulfides in LSL inhibition by mutant NOTCH3. Gel mobility shifting abnormalities of mutant NOTCH3 are eliminated by reducing agents, an indication that abnormal disulfide bonds of mutant NOTCH3 participate in protein conformational changes (Lee et al.. J Biol Chem, 2023: p. 104838.). One might expect that if NOTCH3 disulfide status was related to suppression of luciferase activity, treatment with the reducing agent TCEP would rescue activity. To test how secreted LSL-NOTCH3 proteins respond to reducing agents, conditioned media including LSL proteins was challenged with TCEP and activity was followed over time (Fig. 5A). The wildtype LSL-NOTCH3(l-3) modestly increased in activity (-30% increase), while the R90C LSL-NOTCH3(l-3) sharply increased activity (-1000% increase); increases were seen at all concentrations of protein tested (multiple curves in Fig. 5A). When displayed as the ratio of the minimum activity to the plateau activity (Min / Max), the TCEP-inducible LSL levels showed much greater levels for mutant protein at all enzyme dilutions. The TCEP-inducible LSL levels, corresponding to low Min / Max values, was independent of protein concentration for both wildtype and R90C protein (Fig. 5B). Conceptually, the activity after TCEP was considered to be the total potential activity, whereas the activity before TCEP is considered activity that is actively blocked by abnormal disulfide bonding involving mutant NOTCH3.

[0154] To determine if elevated TCEP-inducibility of LSL-NOTCH3 luciferase activity is a feature of other pathogenic mutants of NOTCH3 and if it is specific for pathogenic vs benign variants, the analysis of LSL-NOTCH3 variants were expanded in Fig. 5C. After transfection of LSL-NOTCH3 constructs, the Min / Max levels of activity were determined after TCEP treatments. All experiments included corresponding wildtype constructs as a reference, which was expected to generate the highest Min / Max values (least TCEP-inducible). For EGF-like repeats 1-3 of NOTCH3, high Min / Max values were observed for wildtype and benign NOTCH3 variants. The Min / Max values were consistently suppressed for all pathological mutants, indicating that their activities were increased by reduction of the reporter proteins. Overall, there was excellent concordance between pathogenicity of mutations and TCEP-inducibility, indicating that the Min / Max value is a second parameter that reflects disease related changes in NOTCH3 structure (referred to as parameter 2).

[0155] Protein secretion abnormalities in subsets of mutant NOTCH3. For the pathological mutant R75P, the overall Min / Max levels with TCEP challenge was only moderately reduced compared to wildtype, indicating that some pathological mutants generate a surprisingly high30275 / 70546fraction of properly folded protein. Because all pathological mutants generated lower total secreted LSL activity, the low secreted level of LSL activity for these mutants may result from poor secretion. The ratio of LSL was measured in the media to that remaining in cells.

[0156] Transfections with wildtype and mutant LSL-NOTCH3 were analyzed by harvesting media and cells. Media was isolated from the mixture by centrifugation, and the cell pellet was separately analyzed by lysis in luciferase assay buffer. All luciferase values were obtained after treatment with TCEP to capture the total potential luciferase activity, irrespective of disulfidedependent conformational pathology. In all experiments, the ratio of media to cell associated activity was compared to that of the corresponding wildtype LSL-NOTCH3 (Fig. 5D).

[0157] For benign variant EGF-like repeat clusters, secretion ratios were the same as that of wildtype. However, some pathogenic mutations resulted in decreased secretion ratio (R90C, R75P, and C49Y). However, not all pathogenic mutations resulted in decreased secretion. Based on these data, the secretion ratio of NOTCH3 variants in the LSL system serves as a specific parameter for pathogenicity. The sensitivity of this third LSL parameter (referred to as parameter 3) was lower than that of the first two parameters in this limited mutation set.

[0158] In total, the LSL-NOTCH3 system generated three parameters that are useful for discrimination of pathogenic and benign NOTCH3 variants: a) secreted LSL activity; b) Min / Max value (TCEP-inducible activity); and c) secretion ratio.

[0159] Importance of the light chain module of LSL in pathological discrimination. Since there are no cysteines in SmBiT and LgBiT, the alteration of activity of LSL-NOTCH3 in the presence of TCEP is dependent on disulfide bonds in NOTCH3 and / or in the antibody light chain. To further characterize the basis of NOTCH3 variant-mediated regulation of LSL activity, a series of alterations in the light chain module of LSL were generated and tested their impact on the magnitude of discrimination between wildtype and R90C NOTCH3 sequences.

[0160] First, the light chain variable sequence of LSL was substituted. The original clone used the rabbit variable sequence from 83G, a monoclonal antibody generated against NOTCH3. Other variable sequences derived from the same animal after immunization were tested with wildtype and R90C NOTCH3 sequences. Compared to constructs with other light chain sequences, 83G generated the greatest luciferase activity in secreted media (left panel. Fig. 6A). Furthermore, 83G derived LSL clones produced the magnitude of discrimination between30275 / 70546wildtype and R90C sequences (ratio of wildtype to R90C shown in right panel of Fig. 6A). As such, among a series of light chain variants, 83G appeared to the best characteristics for studying NOTCH3 among those testes.

[0161] Second, the light chain variable portion of LSL were replaced with the complete light chain (variable plus constant domains [LC]; blue bar of Fig. 6B). In the presence of complete light chain, there was no difference in secreted protein activity between wildtype and mutant NOTCH3 fusions. Moreover, the activation of pathological mutant LSL-NOTCH3 in this context by TCEP was no longer present (blue bar of Fig. 6C).

[0162] Third, small insertions and deletions were made N-terminal to SmBiT in the LSL construct (green and black bars of Figs. 6B and 6C). Addition of two amino acids to the light chain of the LSL reporter reduced the discriminatory capacity of the system (green bar of Figs.6B and 6C). The mutant dependent reporter expression levels (Fig. 6B) and TCEP responsiveness (Fig. 6C) of the mutants were no longer present with deletions of 1-3 amino acids of the light chain module. Overall, these finding indicated that the light chain variable domain spacing relative to NOTCH3 may influence the discriminatory capabilities of the system.

[0163] A unique feature of rabbit IgG light chains is the presence of a disulfide bond that bridges the variable and constant domains; this cysteine bridge is absent from human IgG. In LSL, the absence of the constant domain removes the bonding partner for a conserved cysteine at residue 80. Cys80 of LSL could participate in disulfide bonding with pathological mutant NOTCH3, which is known to harbor free thiols and likely engages in abnormal disulfide bridges (Young et al., Commun Biol, 2022. 5(1): p. 331. and Lee et al., J Biol Chem, 2023: p. 104838.). The formation of Cys80 to mutant NOTCH3 disulfides could dramatically strain the SmBiT domain of the reporter protein and to strongly inhibit split luciferase activity; the suppression of activity should be relieved in the presence of TCEP. Indeed, when the Cys80Ser mutations was introduced into LSL, the mutant NOTCH3 reporter did not respond to TCEP (Figs. 6D and 6E). In sum, 1) spacing of NOTCH3 in relation to the light chain and 2) a single residue (Cys80) in LSL that may form a disulfide bridge with mutant NOTCH3 are key factors that make the LSL system useful for NOTCH3 sequence evaluation.

[0164] Application of LSL to other sets of NOTCH3 EGF repeats. The EGF repeats of NOTCH3 have conserved features but are heterogeneous, with differences in pathogenicity of30275 / 70546mutations in NOTCH3, which depends on the location within the protein sequence (Rutten et al., Genet Med, 2019. 21(3): p. 676-682.; Hack et al., Brain, 2022.; and Cho et al., Neurology, 2022.). To determine potential differences in sequence characteristics, the basal features of WT NOTCH3 EGF triplets were examined across the entire ectodomain for any of the three LSL parameters. The ectodomain was segmented into clusters of three EGF-repeats, and constructs were expressed in cells (Fig. 7A). Secreted luciferase activity (normalized to iRFP activity) was determined for wildtype and mutant NOTCH3 ectodomain fragments (Fig. 7B). The Min / Max values and the secretion parameters were also determined across domains for different mutants, relative to wildtype.

[0165] Without exception, decreased secreted luciferase was noted for constructs that included pathogenic mutations in EGF-like repeats 2-4, 3-5, 4-6, 5-7 and 29-31; with benign variants, there was only a single construct that generated a decrease in activity (G209R in EGF3-5; Fig. 7C). Decreases in Min / Max and secretion parameters were noted in pathogenic mutants but was less sensitive to the presence of pathogenic mutations. All of the constructs were aggregated together and determined that the overall sensitivity and specificity in discriminating pathogenic from non-pathogenic NOTCH3 mutations to be 100% and 93%.

[0166] Application of LSL to other disorders. Other degenerative disorders are caused by mutations in EGF-repeats encoded by other genes, including Marfan disease and Stiff Skin Syndrome (SSS), which are caused by mutations in FBN 1. To determine whether LSL could be applied to these disorders, a sample of mutations in FBN1 were cloned into the LSL framework (LSL-FBN1) and compared activity parameters generated from disease causing mutants to those of the corresponding wildtype sequence.

[0167] Mutations in FBN1 EGF-repeats 11-13 and 13-15 were analyzed by LSL-FBN1 and three parameters of LSL analysis are shown (Fig. 8A-B). This revealed that some, but not all, of the FBN1 pathogenic mutations correspond with alterations in iRFP-normalized luciferase activity. Similarly, some, but not all, of the pathogenic mutations yield differences in Min / Max (parameter 2) and none produced differences from wildtype in secretion ratios (parameter 3). All benign variants yielded three parameters that were not different from wildtype. Mutations in FBN1 that are linked to SSS (TFGBD4) showed alterations in all three parameters tested in an LSL-FBN1 reporter that included TGFBD4 and the adjacent EGF23 domain (Fig. 8C).30275 / 70546

[0168] Extrinsic regulators of mutant NOTCH3 structural transformation. To investigate potential modulators of mutant NOTCH3 structure in cells, LSL-NOTCH3(l-3) constructs were transfected into cells and then challenged cells with iodoacetamide, a general thiol alkylator. If free thiols of NOTCH3 underlie repression of pathological LSL reporter activity, capping thiols may increase luciferase activity. Further, this experiment could test the feasibility of the LSL system to screen for compounds that attenuate the impact of mutant NOTCH3. Fig. 9 shows that treatment of cells with iodoacetamide for 2 hours resulted luciferase activity increases that were greater for CADASIL mutants than wildtype NOTCH3 and benign variant LSL constructs. In addition, the Min / Max values (parameter 2) and secretion indices (parameter 3) were significantly increased for multiple mutants; there were no changes in these parameters for wildtype and benign LSL-NOTCH3 reporters. These studies suggest that alkylation of cysteine thiols can rescue LSL activity of pathological NOTCH3 mutants.Discussion

[0169] Although the initial molecular genetics of inherited small vessel disease implicated cysteine residues of NOTCH3 as disease-initiating factors, altered biochemical features of mutant NOTCH3 protein have only recently emerged. This study provides a new approach to characterizing differences between wildtype and pathogenic NOTCH3 features using a simple, quantifiable transfection assay.

[0170] LSL: a new assay for differentiating pathogenic polypeptide sequences. For NOTCH3 variants, LSL-NOTCH3 provides a high level of discriminatory capacity (wildtype and benign variants vs. pathogenic variants). There was also strong alignment between investigation of NOTCH3 mutant sequences that do not correspond to variants found in the population and prior work using a gel shift assay to assess protein configurations. For example, single mutant changes at residues 49, 75, 90, and 146 showed strong preference for cysteines in both this work and earlier gel shift studies. Specific, unexpectedly significant changes, such as R75G, demonstrated alterations in both protein conformation in gel shifts and by LSL. Suppressor studies in the LSL matched gel shift suppressor studies; both approaches indicate that the strongest suppressor mutants are found when both cysteines that are normally paired are mutated. In case studies, it appears that LSL may have increased sensitivity compared to gel shifting.30275 / 70546

[0171] Three parameters are obtained in each LSL assay: 1 ) the total activity of oxidized reporters (min / iRFP); 2) the fraction of reporter that is active under oxidized versus reduced conditions (Min / Max); and 3) the fraction of reporter that is secreted versus cell associated (secretion index).

[0172] One working model for what the assay is reading out is shown in Fig. 10. The normalized activity of LSL-NOTCH3 (parameter 1) is the overall level of active LSL enzyme, which is modulated by a) the fraction of LSL that achieves conformations that produce LSL activity and b) the secretion efficiency of the protein. Meanwhile, the Min / Max ratios of LSL-NOTCH3 activity (parameter 2) represents the fraction of LSL that achieves conformations that produce LSL activity; the max levels after TCEP treatment, as shown above, likely represent amount of LSL that is in a flexible state, unhindered by disulfide bonding, while the min level is the level of LSL that is active when the NOTCH3 module is disulfide bonded. Mutagenesis of C80 of the light chain suggests that when a pathological mutant of NOTCH3 (with free thiols) is part of LSL-NOTCH3, it is capable of bonding with C80, constraining the LSL-NOTCH3 protein in a manner that prevents SmBiT from LgBiT interaction. Finally, the secretion index (parameter 3) is simply the amount of protein secreted versus withheld in cells.

[0173] Utilizing all three LSL parameters in the context of NOTCH3, there are several conclusions. First, parameter 1 offers the most sensitive screen for pathogenicity of NOTCH3, since it integrates disulfide pathology and secretion pathology into a single assessment (sensitivity was 100% over all 5 EGF combinations). Second, pathological NOTCH3 mutants demonstrate different degrees of alteration of parameter 2 and parameter 3 (sensitivities of 93% and 29% over all 5 EGF combinations). The specificity of the three parameters for discrimination of pathological mutations were 93%, 96%, and 89%. This suggests that mutants may have dominant secretion problems or thiol reactivity issues depending on location of mutations; this interpretation could be affected by interrelationships between secretion and thiol reactivity. Third, non-cysteine pathological NOTCH3 mutations appear to have abnormalities in parameter 1 and parameter 2. The last finding indicates that even non-cysteine mutants alter thiol reactivity of NOTCH3. This implicates thiol reactivity as a CADASIL-associated abnormality that may link all NOTCH3 mutants, regardless of amino acid change or location.30275 / 70546

[0174] LSL for other genetic disorders. LSL was applied to FBN1 to investigate the ability to discriminate pathological from benign mutants related to Marfan’s disease and Stiff Skin Syndrome. In Marfan’s, the ability to distinguish pathological from benign variants of FBN1 was incomplete. The specificity of parameters 1-3 for pathogenic FBN1 mutations was 100%; however, the sensitivity for discrimination of pathogenic mutations was only 44%, 41%, and 15%. It should be noted that Marfan’s disease mechanisms are thought to be related to decreased expression of FBN1; in addition, unlike CADASIL, Marfan’s disease is caused by a large number of non-cysteine mutations. Thus, our LSL assessments of NOTCH3 and FBN1 suggest that the fundamental biochemical mechanisms that drive CADASIL and Marfan’s disease, though both are degenerative vascular diseases, are not identical.

[0175] For Stiff Skin Syndrome, though there are limited number of mutations, LSL-FBN 1 provided a clear difference between mutant and wildtype FBN1 sequences. All of the parameters for SSS LSL-FBN1 were abnormal, indicating that like CADASIL, protein biochemical changes in SSS may involve aberrant thiol reactivity. Overall, the utility of LSL to discriminate between pathogenic and benign variants is high for CADASIL.

[0176] Future studies. As a quantitative test, LSL has the potential to determine the degree of protein dysfunction. This may be useful in CADASIL, as it has recently become clear that different mutations result in higher risk of severe disease (Rutten et al., Genet Med, 2019. 21(3): p. 676-682.; Hack et al., Brain, 2022.; and Cho et al., Neurology, 2022.). The potential to biochemically verify this is attractive but will require careful phenotyping in large groups of patients. As a first step, several EGF repeats can be analyzed by LSL. In addition, there appears to be baseline differences in parameter 2 for different regions of NOTCH3, an indication that thiol reactivity of the protein differs along the expanse of the gene product. Since the LSL assay permits quantification of the effects of mutations with expanded dynamic range compared to gel shifting, future analysis across the NOTCH3 protein using LSL may permit a refined comparison of thiol reactivity of NOTCH3 mutants across independent EGF domains that map to different clinical severity.

[0177] Cysteine alkylation increased LSL-NOTCH3 activity in all pathological mutants but not in benign variants of NOTCH3 EGF 1-3. The non-cysteine mutants R75P and R61W demonstrated significant enhancement with iodoacetamide treatment, with the former exhibiting30275 / 70546the highest magnitude of response; this is consistent with a role of thiol reactivity in both cysteine and non-cysteine CADASIL mutations. Further, these results indicate that scaling up the LSL system may be a feasible approach to identifying cysteine reactive compounds that bind to NOTCH3 mutants, an objective that may contribute to screens for therapies. More generally, adaptation of LSL to other situations in which pathological variants are distinguished from wildtype could be useful in identification of small molecule modifiers that alter protein structure related to inherited disease.Example 2 - Cysteine-reactive mitigators of small vessel disease-related NOTCH3 mutants Materials and Methods

[0178] Candidate mitigators. Cysteine-reactive small molecules were purchased in purest form available from Sigma-Millipore, MolPort, TOCRIS Bioscience, Med-Chem Express, APExBIO, AdipoGen, and Thermo Fisher Scientific. All reagents were dissolved in DMSO and diluted in the same solvent concentrations that corresponded to 1000-fold of amounts used in cell culture experiments.

[0179] DNA constructs. Expression constructs for conducting the LSL assay have been previously described (Cartee et al. J Biol Chem. 2025 Mar;301(3): 108224. and Example 1), and new constructs for the current study were built on the same platform DNA vectors. Point mutations in NOTCH3 were incorporated by PCR of templates from mutants described (Lee et al. J Biol Chem. 2023 Jun;299(6): 104838.) or by nested PCR using oligos that included desired mutations. DNA fragments were digested with restriction enzymes and cloned into vectors with T4 DNA ligase. Clones were sequenced prior to use in expression assays. Point mutations of FBN1 have been previously described (Cartee et al. J Biol Chem. 2025 Mar;301(3): 108224. and Example 1). For NOTCH3-ASL constructs in Fig. 19, NOTCH3 EGF repeats 1-3 replaced the light chain sequence of the LSL vector by PCR; a secretion peptide sequence at the 5’ end was incorporated into all clones. Subsequently, the full length human APOE2 open reading frame (with its stop codon mutated, incorporating flanking SalPAgel and Sph I / Bgl 11 sites at the 5’ and 3’ ends of the ORF) replaced the NOTCH3 sequences of LSL clones to produce the reporter system shown in Fig. 19A.

[0180] Cell culture and small molecule treatments. HEK293 cells (293A; Qbiogene) were grown in DMEM with 10% fetal bovine serum in 5% carbon dioxide chambers. Gene30275 / 70546transfection was conducted with PolyJet as recommended by the manufacturer. As before (Cartee et al. J Biol Chem. 2025 Mar;301(3): 108224. and Example 1), 400ng of LSL vector and 100ng of iRFP plasmid were mixed in 25pL DMEM. DMEM (25pL) and 1.5pL PolyJet were added to diluted DNA and then dropped onto media of cells in 24 well plates. After 18-24 hours, the media was exchanged with OptiMEM supplemented with 0.1% DMSO or test drug at the concentrations described; stocks of drug were in DMSO and diluted 1:1000 into OptiMEM. After 2 hours, conditioned media was removed for analysis of nanoLuc activity.

[0181] Split luciferase analysis of small molecule activity on pathogenic mutants. LSL-NOTCH3 experiments were performed according to protocols described in prior work (Cartee et al. J Biol Chem. 2025 Mar;301 (3): 108224. and Example 1), with all experiments on mutant sequences compared to appropriate wildtype NOTCH3 sequence cloned into the LSL vector. The effects of small molecules were determined by comparing two luciferase parameters with and without candidate mitigators of interest. Parameter 1 (secreted nanoLuc activity normalized to iRFP) was determined as before (Cartee et al. J Biol Chem. 2025 Mar;301(3); 108224. and Example 1) (25pL conditioned media combined with 6.25 μL reaction mixture). Luciferase activity was determined in which plastic plates in a plate reading luminometer (BioTek Synergy LX multi-mode reader). In all experiments, unless noted, the candidate mitigator response was Luc / iRFP with small molecule referenced to without. For determination of Parameter 2 (reduction-unmasked Luc activity), 2 μL of TCEP (31.25mM) was mixed with the reaction mixture and luciferase activity was followed over 30 minutes. Time course values were fitted to the equation: Y=Y0+(Plateau-Y0)*(1-exp(-K*x)) to derive the plateau (max) and the initial value before adding TCEP was deemed the Min level. The Min / Max ratio was determined, corresponding most likely to the percentage of protein in normal conformation relative to total protein. Unless noted, the Min / Max with candidate mitigators was normalized to samples without candidate exposure. Parameter 3 from (Cartee et al. J Biol Chem. 2025 Mar;301(3): 108224. and Example 1) was not determined in this study. The LSL-NOTCH3 protocol was used to analyze NOTCH3-ASL reporters from Fig. 19, except that TCEP challenges were not performed (only Parameter 1 was assessed). The LSL-NOTCH3 methods described above were used for FBN1 analysis in Fig. 20.30275 / 70546

[0182] Statistics. Normality was assessed by the Shapiro-Wilk test. Significant differences were determined using one-way ANOVA with Dunn’s multiple comparisons test usingGraphPad Prism 10.4.1. A p value < 0.05 was considered statistically significant.Results

[0183] Small molecule screening. The 21 compounds selected for study were molecules with capacity to react with protein thiols. Two well-established cysteine alkylation agents, iodoacetamide (IAM) and N-ethylmaleimide (NEM), were selected and expected to have broad target range. Other reagents included FDA-approved or investigational drags with ability to alkylate cysteines. The largest groups of compounds were oncological medications (that bind pro-proliferative proteins via cysteines) and proton pump inhibitors that covalently modifygastric H+ / K+ ATPase (via a sulfenic acid group). The compounds and their chemical features are displayed in Table 1.Table 1: Twenty one cysteine-reactive candidates tested for ability to mitigate NOTCH3 conformational alterations. Compounds are listed in order of chemical complexity, which was obtained from the world wide web at pubchem.ncbi.nlm.nih.gov / compound. Lipophilicity data was calculated using the world wide web at swissadme.ch. Electrophilicity data was determined using the world wide web at esnuel.org. H = exceeds calculation limits.Chemical name Chemical reactivity FDA approval Complexity Lipophilicity Electrophilicity 2-lodoacetamide (IAM) Nucelophilic substitution (SN2) reaction No 44.9 0.16 24819 2-(butan-2-yl disulfa nyl )-l H- Nucelophilic substitution (SN2) reaction No 111 2.14 167 imidazole (PX-12)N-acetylcysteine amide Thiol disulfide exchange reaction No 149 -0.44 125 N-ethylmaleimide (NEM) No 165 0.48 253 Disulfiram (DSF) Thiol disulfide exchange reaction Yes 201 3.25 192 4 (2- Nucelophilic substitution(SN2) reaction No 239 1.62 56851 Aminoethyl)benzenesulfonylfluoride hydrochloride(AEBSF)Ebselen Covalent bonding, forming No 275 1.75 185 a thioselenide linkageCarmofur Covalent adduct formation, carbamoylation No 382 1.93 246 Rabeprazole Sulfenamide intermediate, then Covalent Yes 440 2.26 184 disulfide bondTenatoprazole Sulfenamide intermediate, then Covalent No 455 2.05 250 disulfide bondOmeprazole Sulfenamide intermediate, then Covalent Yes 459 2.31 234 disulfide bondLansoprazole Sulfenamide intermediate, then Covalent Yes 480 3.13 241disulfide bond30275 / 70546Chemical name Chemical reactivity FDA approval Complexity Lipophilicity Electrophilicity Pantoprazole Sulfenamide intermediate, then Covalent Yes 490 2.3 230 disulfide bondAuranofin Ligand exchange reaction Yes 532 1.12 H Spebrutinib No 561 3.41 H Evobrutinib No 595 3.48 H Osimertinib Yes 725 3.24 H Ibrutinib Yes 726 3.25 H Zanubrutinib Yes 728 3.17 H Necrosulfonamide No 760 1.53 H Sotorasib Yes 1030 4.1 H

[0184] NOTCH3 variant evaluation platform. As demonstrated in Example 1, the LSL-NOTCH3 platform enables differentiation of benign variants and pathogenic variants of NOTCH3 (Cartee et al. J Biol Chem. 2025 Mar;301(3): 108224. and Example 1). The platform employed human NOTCH3 variants (three EGF-like repeats at a time) cloned into the LSL backbone vector and then transfected into 293 cells along with an iRFP expression vector to normalize for transfection efficiency. High activity in the conditioned media was generated by WT and benign variants of NOTCH3; substantially lower activity was found after transfection with pathological NOTCH3 mutants. Impaired activity of mutants was a result of a combination of deficient expression levels, secretion efficiency, and / or conformational alterations that block luciferase activity reconstitution; total production of protein with retained luciferase activity is reflected by composite Parameter 1, the total luciferase secreted, normalized to iRFP levels.Conformational alterations that depend on disulfide bonding are reflected by Parameter 2, the ratio of unreduced luciferase activity normalized to total activity after unlocking aberrant disulfide bonds using the reductant TCEP.

[0185] Fig. 11 diagrams how LSL-NOTCH3 was used to evaluate the degree to which candidate drugs attenuate the impact of pathological NOTCH3 mutants. LSL-NOTCH3 was first transfected into cells with iRFP. After an overnight incubation, candidate drugs in fresh media were overlayed onto cells. The expression and activity of the LSL-NOTCH3 constructs were determined 2 hours after addition of fresh media (with DMSO or drug; Parameter 1). The ratio of expression with drug versus DMSO was calculated to measure the impact of drug. To determine the ability of the drug to block disulfide dependent pathological changes, Parameter 230275 / 70546was determined for DMSO or drug-treated groups by adding TCEP to the luciferase assay and following the increase in activity over time.

[0186] Attenuation of pathogenic properties ofNOTCH3 by pharmacological agents. The effect of 21 cysteine acting agents were evaluated on CADASIL mutants of NOTCH3 in EGF1-3, EGF4-6, and EGF31-33 (Fig. 12), regions that correspond to both high and low risk alleles (Rutten et al. Genet Med. 2019 Mar;21(3):676-682; Hack et al. Brain. 2023 Jul 3; 146(7):2913-2927; and Cho et al. Neurology. 2022 Aug l;99(5):e430-e439.). Significant increases in Parameter 1 values for at least one pathogenic reporter were identified for 10 of 21 compounds, including: iodoacetamide, PX-12, N-ethylmaleimide, disulfiram, ebselen, carmofur, auranofin, spebrutinib, osimertinib, and necrosulfonamide.

[0187] Significant increases in Parameter 1 values across half or more of the mutant reporters were identified for 5 of 21 compounds, including: IAM (13 of 16), N-ethylmaleimide (10 of 16), PX-12 (10 of 16), disulfiram (14 of 16), and auranofin (12 of 16). Overall, these test compounds had broad effects on pathogenic NOTCH3 variants, though the magnitude of the effects across mutants was heterogeneous.

[0188] The strongest increases in Parameter 1 values occurred for the following compounds: disulfiram (up to 7.7 times relative to wildtype responses), iodoacetamide (up to 12 times relative to wildtype responses), and auranofin (up to 7.2 times relative to wildtype responses). The primary data for these compounds are presented in Fig. 13. The most potent effects for disulfiram were seen against the R75P mutation; disulfiram increased absolute reporter levels to 10% of wildtype. For iodacetamide, the most potent effects were seen against R75P mutation; iodacetamide increased absolute reporter levels to 10% of wildtype. The most potent effects for auranofin were seen against the R75P mutation; auranofin increased reporter levels to 10% of wildtype. None of the agents fully reversed the effects of CADASIL mutations on Parameter 1. However, mutants that had higher baseline activity, C206R, C222Y, and C224Y, disulfiram increased the reporter value to 17%, 20% and 40% of wildtype, respectively.

[0189] Increases in Parameter 2 values were also identified for 16 / 21 compounds. Thus, a significant majority of compounds elevated both Parameter 1 and 2 for at least one mutant reporter. None of the 21 compounds had any effects on reporters corresponding to non-30275 / 70546pathogenic variants of NOTCH3. Fig. 14 shows the effects of iodoacetamide, disulfiram, and auranofin on Parameter 2 across all NOTCH3 variants tested.

[0190] Cell-free versus cell-dependent small molecule effects. For compounds that demonstrated activity on CADASIL mutants, Experiments on protein after secretion were performed by mixing conditioned media with selected drugs (Fig. 15 A). None of these studies showed increases in activity, indicating that the drugs do not act directly on the secreted LSL-NOTCH3 proteins. In time course studies of disulfiram and PX-12 treated cells, there was an increase in Parameter 1 which progressively rose with longer periods of incubation (Fig. 15B), which is consistent with an action of these drugs on cellular production and / or processing of the NOTCH3 protein.

[0191] Dose characterization. Dose / response correlations were defined in a series of studies on disulfiram, PX-12, and auranofin (tested against the most highly responsive mutant reporters; Fig. 16). These targeted studies indicate that the strongest effects on LSL-NOTCH3 production occur with 10μM disulfiram and PX-12 and 1μM auranofin.

[0192] Compound effects of cysteine reactive molecules. It was tested whether combining cysteine reactive drugs resulted in synergistic or additive or antagonistic effects (Fig. 17).Combining PX12 and disulfiram resulted in higher values of LSL-NOTCH3 reporter activity that with each drug used alone; the effects appeared additive.

[0193] Free thiol effects on small molecule potency. None of the drugs that affected pathogenic NOTCH3 reporters were active on wildtype reporters. Since pathogenic variants are largely cysteine altering mutations that result in a potentially unpaired thiol group, it was tested whether drug effectiveness was dependent on the loss of cysteine or, alternatively, on the gain of a non-cysteine amino acid. Accordingly, a mutations to the cysteine at residue 49. which was shown to respond well to disulfiram in the C49Y mutant, to all 18 other amino acids (Fig. 18A). Nearly all of the mutants responded to disulfiram, with increased reporter levels. At residue 146. which responds to PX-12 in the pathogenic C146R mutant, alteration to all 18 other amino acids resulted in similar responses to PX-12, except for one mutant (Fig. 18B). These two cases are consistent with the notion that small cysteine reactive compounds act on NOTCH3 because of a loss of cysteine rather than because of a gain of another amino acid.30275 / 70546

[0194] Effects of cysteine-reactive drugs on alternative NOTCH 3 conformational reporters. To confirm whether drugs are capable of attenuating conformational changes of pathological NOTCH3 variants under other circumstances, an alternative to the LSL-NOTCH3 assays was tested. This assay, NOTCH3-ASL (APOE split luciferase), is illustrated in Fig. 19 A. The test variant of NOTCH3 is cloned to the N-terminus of an inverted split nano-luciferase that is separated by an APOE2 open reading frame with a linker sequence. As shown in Fig. 19B. wildtype and benign NOTCH3 variants generated significantly higher secreted luciferase activity than all pathological mutants, which is consistent with the pathological NOTCH3 generation of abnormal protein that is poorly secreted or synthesized. Cells transfected with a series of pathogenic and non-pathogenic mutants were treated with disulfiram which resulted in increased levels of reporter expression in all pathogenic variants (Fig. 19C). WT NOTCH3-ASL expression was not affected by drug treatment. Because similar drug responses were observed using two reporter systems, it is likely that the drug effects reflect action upon mutant NOTCH3.

[0195] Effects of cysteine-reactive drugs on FBNl-related mutant protein. Cysteine imbalanced proteins are also implicated in other genetic disorders. In Marfan’s disease, mutations found in the EGF-like repeats cause vascular degenerative changes that are, to date, untreatable (Dietz et al. Nature. 1991 Jul 25;352(6333):337-9.). Many of the mutations are similar to those found in CADASIL: they alter cysteine number that results in thiol imbalanced proteins. Accordingly, it was tested if LSL-FBN1 reporters that correspond to Marfan’s mutations could be enhanced by disulfiram or auranofin. Disulfiram increased select pathogenic LSL-FBN1 reporters as shown in Fig. 20A. In contrast, auranofin failed to increase activity of any of the LSL-FBN1 reporters (Fig. 20B). None of the benign mutants in FBN1 displayed beneficial responses to either drug in reporter assays.Discussion

[0196] Because CADASIL results from mutations that change cysteine number in NOTCH3, thiol imbalance of NOTCH3 and subsequent conformational changes in the protein are suspected to be the initiating steps in disease. This study was motivated by the need to identify approaches that could attenuate pathogenic NOTCH3 conformations. The results demonstrate that 1) multiple cysteine-reactive small molecules have the capacity to mitigate the aberrant properties30275 / 70546of C AD ASIL-related NOTCH3 mutants; and 2) select cysteine-reactive small molecules act broadly against subsets of NOTCH3 mutants.

[0197] Cysteine-reactive small molecules act on mutant NOTCH3. The present example demonstrates the feasibility of exploiting cysteine reactivity to counter effects of NOTCH3 conformational changes in CADASIL. In view of their shared thiol reactivity, it is likely that the small molecules highlighted here target NOTCH3 via covalent interactions with cysteine residues. One mechanism that is consistent with the results is that mutant NOTCH3 harbors unpaired thiols that react with other thiols of NOTCH3, rendering an aberrant protein secondary structure; however, in the presence of thiol reactive mitigator compounds these unpaired thiols are capped, thereby preventing abnormal intramolecular disulfide bonding. Another possibility is that unpaired thiols of mutant NOTCH3 react with other cellular molecules that bind to cysteine; thiol reactive mitigators may therefore also react with unpaired thiols to prevent pathogenic intermolecular interactions. In support of these mechanisms, improvement in Parameter 2 (fraction of properly disulfide bonded protein) was found in many of the drug / variant combinations (Fig. 13 and Fig. 14) and documented that effects of mitigators on multiple NOTCH3 mutants were independent of non-cysteine residue identity (Fig. 18).

[0198] Treatment of protein outside of cells did not change reporter activity, indicating that mitigators act on mutant NOTCH3 inside of cells, potentially when unpaired thiols are initially generated during protein synthesis and maturation. The unavailability of thiols to drugs after synthesis and secretion is consistent with the unexpectedly low amount of thiol in purified mutant NOTCH3 preparations that we previously described for a series of purified NOTCH3 mutants (Young et al. Commun Biol. 2022 Apr 7;5(1):331.).

[0199] Mitigation of both cysteine and non-cysteine NOTCH3 mutants. We note that non-cysteine mutants (R61W, R75P, and G209R) respond to mitigators, with higher magnitude than several cysteine mutants. Thus, cysteine targeting appears to beneficially affect both cysteine and non-cysteine pathogenic NOTCH3 mutants. In prior work, we note that the non-cysteine mutants have similar redox-dependent gel shift properties as cysteine mutants (Lee et al. J Biol Chem. 2023 Jun;299(6): 104838.); furthermore, Parameter 2 non-cysteine and cysteine mutants in LSL-NOTCH3 assays are also uniformly depressed (Cartee et al. J Biol Chem. 2025 Mar;301(3): 108224. and Example 1). Thus, experimental evidence points to the possibility that30275 / 70546the molecular basis of both cysteine and non-cysteine NOTCH3 mutant impairment may be aberrant disulfide bond formation.

[0200] The lower degree of benefit across a broad spectrum of Marfan’s disease reporters underscores differences between NOTCH3 and FBN1 disease pathophysiology and indirectly supports the role of cysteine imbalance in CADASIL. Whereas NOTCH3 disease is likely a cysteine-imbalance problem, FBN1 disorders such as Marfan’s have been proposed to result from independent mechanisms (Zeigler et al. Adv Exp Med Biol. 2021;1348:185-206. and Du et al. Int J Med Sci. 2021 May 27; 18(13):2752-2766.).

[0201] Addressing the challenge of molecular heterogeneity ofNOTCH3 mutants in CADASIL. One challenge in targeting NOTCH3 in CADASIL is molecular heterogeneity, as hundreds of different mutations have been described. In principle, heterogeneity could restrict the utility of highly specific NOTCH3 conformation targeting drugs, which may not work broadly across the patient population. The present study suggests an approach to neutralize issues surrounding molecular heterogeneity: using lower specificity cysteine-targeting compounds against mutant NOTCH3.

[0202] The broadest NOTCH3 mitigators, disulfiram and auranofin, altered NOTCH3 reporter activity a majority of mutations across nine EGF repeats and were also effective on gain and loss of cysteine mutations and against non-cysteine pathogenic mutants. The chemical structures of these agents make them unlikely to be highly specific for a narrow range of cysteines (Quadros Barse et al. J Biol Chem. 2025 Mar;301(3):108159.; Roder et al. Drugs R D.2015 Mar;15(l):13-20.; Yakisich et al. Biochem Biophys Res Commun. 2001 Nov 30;289(2):586-90.: Malka et al. FEBS J. 2009 Sep;276(17):4900-8.; Paranjpe et al.Carcinogenesis. 2014 Mar;35(3):692-702.; Xu et al. Microbiol Spectr. 2025 Nov4; 13(1 l):e0175225.; and Loo et al. Mol Pharm. 2004 Nov-Dec;l(6):426-33.). Tian et al recently showed that a multiple cysteine-targeting drugs in fact possess broader than initially realized thiol reactivity ranges (Tian et al. Nat Commun. 2025 May 26; 16(1):4863.). As such, further development of CADASIL targeting drugs could benefit from de-prioritizing complexity and specificity of small molecule drugs and, rather, emphasizing chemical reactivity.

[0203] In summary, the present work validates a discovery strategy that leverages thiol reactivity to impart therapeutic effects on pathological NOTCH3. The application of the LSL30275 / 70546assay system study provides a simple and rapid tool to facilitate derivative compound development and for exploration of expanded libraries of cysteine-reactive candidates. The results support the potential for repurposed drugs such as disulfiram and auranofin in NOTCH3 disease which is not restricted to specific mutations. Expansion of the approach provides a foundation for future potential optimization, testing, and deployment of therapeutics against currently untreatable conditions.

Claims

30275 / 70546CLAIMS1. An expression construct comprising a polynucleotide encoding a fusion protein, wherein the fusion protein comprises:a) an antibody fragment;b) a first detection moiety comprising a first portion of a split reporter;c) a protein sequence of interest; andd) a second detection moiety comprising a second portion of the split reporter;wherein the first and second detection moieties are capable of assembly when in proximity, andwherein the first and second detection moieties emit a detection signal when assembled.

2. The expression construct of claim 1, wherein the fusion protein comprises, from N- terminus to C-terminus: the antibody fragment, the first detection moiety, the protein sequence of interest, and the second detection moiety.

3. The expression construct of claim 1, wherein the fusion protein comprises, from N- terminus to C-terminus: the protein sequence of interest, the first detection moiety, the antibody fragment, and the second detection moiety.

4. The expression construct of any one of claims 1-3, further comprising one or more linkers.

5. The expression construct of 4, wherein the one or more linkers are located between the antibody fragment and the first detection moiety, the antibody fragment and the second detection moiety, the protein sequence of interest and the first detection moiety, and / or the protein sequence and the second detection moiety.

6. The expression construct of claim 4 or 5, wherein the one or more linker sequence is RS and / or VD.

7. The expression construct of any one of claims 1-6, wherein the antibody fragment comprises a light chain.

8. The expression construct of claim 1, wherein the light chain is a rabbit light chain.30275 / 705469. The expression construct of claim 8, wherein the rabbit light chain comprises a Cys80 residue.

10. The expression construct of claim 9, wherein the rabbit light chain is 83G, 44F, 69B, 82A, 120B, 123D, or 145H.

11. The expression construct of claim 10, wherein the rabbit light chain is 83G.

12. the expression construct of claim 1, wherein the first and second detection moieties comprise split β-lactamase, split GFP, or split luciferase.

13. The expression construct of claim 12 wherein the first and second detection moieties comprise split luciferase.

14. The expression construct of claim 13, wherein the first and second detection moieties comprise SmallBiT and LargeBiT of Nanoluc.

15. The expression construct of claim 1, wherein the protein sequence of interest comprises one or more mutations relative to a reference protein sequence.

16. The expression construct of claim 15, wherein the protein sequence of interest contains one or more cysteine residues.

17. The expression construct of claim 16, wherein the protein sequence of interest contains multiple cysteine residues, and wherein two or more of the cysteine residues form a disulfide bridge.

18. The expression construct of claim 17, wherein mutations to one or more cysteine residues disrupt formation of at least one disulfide bridge.

19. The expression construct of claim 15, wherein the fusion protein, when properly folded, assembles the first portion and the second portion of the split reporter.

20. The expression construct of claim 15, wherein the fusion protein, when misfolded, does not assemble the first portion and the second portion of the split reporter.

21. The expression construct of claim 16, wherein the light chain comprises a Cys80 residue, and wherein a cysteine of the protein sequence of interest forms a disulfide bridge with the Cys80 residue.30275 / 7054622. The expression construct of claim 1, wherein the protein sequence of interest comprises a secreted protein, a cytoplasmic protein, a transmembrane protein, organelle-localized protein or a fragment thereof.

23. The expression construct of claim 1, wherein the protein of interest comprises NOTCH3 orFBNl.

24. A composition comprising the expression construct of any one of claims 1-23.

25. A composition comprising a collection of expression constructs of any one of claims 1- 23, thereby providing a library of protein sequence variants.

26. A host cell comprising the expression construct of any one of claims 1-23.

27. A composition comprising a collection of host cells of claim 26, thereby providing a library of protein sequence variants capable of being expressed and secreted from the collection of host cells.

28. A method of determining whether a protein of interest is properly folded, the method comprising the steps of:(a) transfecting a host cell with an expression construct comprising a polynucleotide encoding a fusion protein, wherein the fusion protein comprises:i) an antibody fragment;ii) a first detection moiety comprising a first portion of a split reporter;iii) a protein sequence of interest; andiv) a second detection moiety comprising a second portion of the split reporter; wherein the first and second detection moieties are capable of assembly when in proximity,wherein the first and second moieties emit a detection signal when assembled, andwherein the protein sequence comprises one or more mutations relative to a reference protein sequence:(b) incubating the cell of (a) in a growth media under conditions that allow expression and secretion of the fusion protein;(c) collecting the growth media comprising the expressed and secreted fusion protein;30275 / 70546(d) detecting a detection signal from the growth media;(e) determining if the fusion protein is properly folded by comparing the fusion protein detection signal to a signal from a reference protein.

29. A method of determining whether a protein of interest is properly folded, the method comprising the steps of:(a) transfecting a host cell with an expression construct comprising a polynucleotide encoding a fusion protein, wherein the fusion protein comprises:i) an antibody fragment;ii) a first detection moiety comprising a first portion of a split reporter;iii) a protein sequence of interest; andiv) a second detection moiety comprising a second portion of the split reporter; wherein the first and second detection moieties are capable of assembly when in proximity,wherein the first and second moieties emit a detection signal when assembled, and wherein the protein sequence comprises one or more mutations relative to a reference protein sequence;(b) incubating the cell of (a) in a growth media under conditions that allow expression and of the fusion protein;(c) lysing the cell and collecting the cell lysate;(d) detecting a detection signal from the cell lysate;(e) determining if the fusion protein is properly folded by comparing the fusion protein detection signal to a signal from a reference protein.

30. The method of claim 28 or 29, wherein the reference protein is a wild-type protein.

31. The method of claim 29, wherein the cell is lysed by chemical lysis, enzymatic lysis, mechanical lysis, or physical lysis.

32. The method of claim 31, wherein chemical lysis comprises use of an alkali solution or detergent solution.

33. The method of claim 32, wherein the detergent solution comprises 3-((3- cholamidopropyl) dimethylammonio)-! -propanesulfonate (CHAPS),30275 / 70546cetyltrimethylammonium bromide (CTAB), nonyl phenoxypolyethoxylethanol (NP-40), polyethylene glycol sorbitan monolaurate (Tween™), sodium dodecyl sulphate (SDS), or t-Octylphenoxypolyethoxyethanol (Triton™ X-100).

34. The method of claim 31, wherein the enzymatic lysis comprises use of lysozyme, lysostaphin, zymolase, cellulose, protease or glycanase.

35. The method of claim 31, wherein mechanical lysis comprises use of a high-pressure homogenizer or a bead mill.

36. The method of claim 31, wherein physical lysis comprises use of ultrasonic waves, heat, osmotic shock, or cavitation.

37. The method of any one of claims 28-31, further comprising calculating a detection ratio by dividing the reference protein detection signal by the fusion protein detection signal, wherein a calculated detection ratio value greater than at least 2 corresponds to a properly-folded folding fusion protein and wherein a calculated detection ratio value less than at least 2 corresponds to a misfolded fusion protein.

38. The method of claim 28 or 29, further comprising:(f) contacting the growth media with a reducing agent; and(g) detecting a detection signal from the growth media.

39. The method of claim 38, further comprising calculating a detection ratio by dividing the reference protein detection signal by the detection signal of (g),wherein a calculated detection ratio value greater than 2 corresponds to a properly-folded folding fusion protein and wherein a calculated detection ratio value less than 2 corresponds to a misfolded fusion protein.

40. The method of claim 38, wherein the reducing agent is tris (2-carboxyethyl) phosphine hydrochloride (TCEP), P-mercaptoethanol (BME), or dithiothreitol (DTT).

41. The method of claim 28 or 29, further comprising:(f) separating the host cell from the growth media;(g) lysing the cell and collecting the cell lysate;(h) contacting the host cell with a secondary antibody,30275 / 70546wherein the secondary antibody is capable of binding to the antibody fragment, andwherein the secondary antibody comprises a detectable label; and (i) adding the secondary antibody to the growth media; and(j) detecting a signal from the secondary antibody;(k) comparing the signal from the secondary antibody with the signal from the growth media to measure secretion of the fusion protein.

42. The method of claim 28 or 29, further comprising:(l) treating cells with one or more thiol alkylators; and(m) repeating steps a) through k).

43. The method of claim 42, wherein the one or more thiol alkylators is iodoacetamide, PX- 12, N-ethylmaleimide, disulfiram, ebselen, carmofur, auranofin, spebrutinib, osimertinib, necrosulfonamide, or a combination thereof.

44. A method of determining whether a protein of interest comprises a pathogenic protein sequence, the method comprising the steps of:(a) transfecting a host cell with an expression construct comprising a polynucleotide encoding a fusion protein, wherein the fusion protein comprises:i) an antibody fragment;ii) a first detection moiety comprising a first portion of a split reporter;iii) a protein sequence of interest; andiv) a second detection moiety comprising a second portion of the split reporter; wherein the first and second detection moieties are capable of assembly when in proximity,wherein the first and second moieties emit a detection signal when assembled, and wherein the protein sequence comprises one or more mutations relative to a reference protein sequence;(b) incubating the cell of (a) in a growth media under conditions that allow expression and secretion of the fusion protein;(c) collecting the growth media comprising the expressed and secreted fusion protein; (d) detecting a detection signal from the growth media;30275 / 70546(e) determining if the fusion protein is properly folded by comparing the fusion protein detection signal to a signal from a reference protein;wherein a determination of proper folding corresponds to protein sequence comprising a benign mutation, andwherein a determination of improper folding corresponds to a protein sequence comprising a pathogenic mutation.

45. A method of identifying a therapeutic compound or biomolecule, the method comprising the steps of:(a) contacting a misfolded fusion protein according to claim 28 or a pathogenic protein according to claim 44 with a compound or biomolecule;(b) detecting a detection signal;(c) determining if the detection signal detected in (b) is improved relative to a detection signal in the absence of the compound or biomolecule;wherein an improved detection signal corresponds to a compound or biomolecule capable of rescuing a misfolded fusion protein or pathogenic protein.

Images