Method of stratifying a subject with polyneuropathy and in virto method for diagnosing polyneuropathy in a subject

Abstract

The present invention relates to an in vitro method for diagnosing polyneuropathy (PNP) in a subject as well as a method of stratifying a subject with polyneuropathy (PNP). Further, the present invention relates to a method for determining therapeutic treatment of PNP in a subject as well as to the use of any method of the present invention for determining a treatment regime for a subject suffering from PNP.

Description

METHOD OF STRATIFYING A SUBJECT WITH POLYNEUROPATHY AND IN VIRTO METHOD FOR DIAGNOSING POLYNEUROPATHY IN A SUBJECTThe present application claims the benefit of priority of EP Patent Application EP 24178562.5 filed on 28 May 2024, the content of which is hereby incorporated by reference in its entirety for all purposes.FIELD OF THE INVENTION

[0001] The present invention relates to an in vitro method for diagnosing polyneuropathy (PNP) in a subject as well as a method of stratifying a subject with polyneuropathy (PNP). Further, the present invention relates to a method for determining therapeutic treatment of PNP in a subject as well as to the use of any method of the present invention for determining a treatment regime for a subject suffering from PNP.BACKGROUND OF THE INVENTION

[0002] Polyneuropathies (PNPs) denote common diseases affecting peripheral nerves and have diverse underlying causes, which complicates their diagnosis. Single cell transcriptomic studies have defined the cellular architecture of rodent nerves. Whether this is replicated in humans and how human nerves transcriptionally respond to PNPs is unknown so far.

[0003] Specifically, polyneuropathies (PNPs) denote diseases of multiple peripheral nerves and range among the most common neurological diseases, affecting at least 4% of the middleage and elderly population1. Patients affected by PNPs suffer from progressive sensory and motor impairment and an often painful life-long aggravating disease with an enormous socioeconomic impact2 3. The underlying causes of PNPs are very diverse and comprise diabetic, toxic inflammatory / immune-mediated, hereditary, metabolic and vitamin deficiency4-6. Despite comprehensive diagnostic work-up, the underlying cause of PNPs is unclear in many patients (20-30 %)5. Biopsy of the sensory sural nerve at the lateral ankle is often the final diagnostic step to assess specific causes in human PNP patients, but even biopsy does not lead to a diagnosis in many patients7. This could be due to an insufficient depth of analysis of this tissue.Fully exploiting this precious biomaterial for mechanistic understanding and diagnostic potential is especially important to detect treatable causes such as immune mediated neuropathies, which account for up to 10% of the PNPs45and are treatable.

[0004] With the present invention, the inventors generated a single nuclei atlas of human peripheral nerves from 33 PNP patients and four controls (365,708 nuclei) combined with subcellular spatial transcriptomics in eight patients. The inventors identified and spatially validated novel and partially human-specific markers of perineurial fibroblasts (CXCL14), and myelinating (MLIP) and non-myelinating Schwann cells (GRIK3, PRIMA 1). Perineurial cells were more heterogeneous than expected from rodents. Nerve-associated immune cells contained diverse macrophages with an endoneurial to epineurial phenotypic gradient and previously undescribed classical dendritic cells (eDC) type 2. In PNPs, endoneurial cells including myelinating Schwann cells were lost, while Schwann cells with a repair phenotype increased, which correlated with disease severity and expansion of endoneurial lipid- associated macrophages. Transcriptional changes in non-endoneurial vascular cells and perineurial hyperplasia and disruption were shared across PNP subtypes. Additionally, the present invention enables that PNP patients were categorized into clusters identified by single cell transcriptomics with specific phenotypes beyond existing clinical classification.

[0005] The inventors and others previously performed single cell transcriptomic analyses of rodent peripheral nerves8-13. This identified novel markers of endoneurial cells, unexpected complexity of hematopoietic cell types in peripheral nerves, and a transcriptional response to autoimmunity that was shared between the peripheral and central nervous system. However, human peripheral nerves have not been characterized using similar technologies. Whether the cellular composition and cell type markers are replicated cross-species has not been investigated yet. Moreover, it remains unknown how human nerve cells respond to PNPs transcriptionally and compositionally on a single cell level and whether this is shared or distinct between PNP subtypes.

[0006] Here, by combining single nuclei and spatial transcriptomics of human sural nerve biopsies, the inventors provide the first large-scale single cell atlas of human peripheral nerves. They defined the cellular micro-milieu in human peripheral nerves and detected a unique cellular composition of the perineurium and abundant nerve-associated leukocytes. The inventors identified novel cell type markers of myelinating and non-myelinating Schwann cells (SC) and perineurial fibroblasts, which were validated spatially. Repair and damage SC, and occurrence of lipid-associated macrophages were shared features of PNPs. The inventors also found that PNPs affected cells beyond the endoneurium, causing perineurial cell proliferationand perineurial thickening preferentially in inflammatory PNPs. Finally, patients could be categorized into patient clusters identified solely by single cell transcriptomics which hinted towards mechanistic patterning of patients. The inventors thus provide a single cell atlas for the human peripheral nerve in health and PNPs and discover potential for an unbiased diagnostic classification of PNPs.

[0007] Thus the inventors were able to identify a unique cellular architecture of human nerves, to define PNPs as ‘pan-nerve diseases’, and to discover potential for unbiased diagnostic classification of PNP using single cell technologies.SUMMARY OF THE INVENTION

[0008] The present invention relates in a first aspect to an in vitro method for diagnosing polyneuropathy (PNP) in a subject, comprising a) determining by single cell RNA sequencing(i) a proportion of one or more cell type cluster characterized by one or more nerve- associated marker gene(s) and / or by one or more immunological marker gene(s) in a sample obtained from the subject;(ii) an expression level of one or more nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i);(b) diagnosing said subject as suffering from PNP if one or more of the following are fulfilled;(i) the proportion of one or more cell type cluster of (a)(i) characterized by the one or more nerve-associated marker gene(s) and / or by the one or more immunological marker gene(s) is / are altered compared to a control sample;(ii) the expression level of (a)(ii) of one or more of the nerve-associated differentiation gene(s) and / or of one or more of the immunological differentiation gene(s) within the one or more cell type cluster of (i) is altered compared to a control sample; wherein the altered proportion of one or more cell type cluster of (b)(i) and the altered expression levels of (b)(ii) are indicative for PNP.

[0009] In a second aspect, the present invention relates to a method of stratifying a subject with polyneuropathy (PNP), comprising(a) determining by single cell RNA sequencing(i) a proportion of one or more cell type cluster characterized by one or more nerve- associated marker gene(s) and / or one or more immunological marker gene(s) in a sample obtained from the subject;(ii) an expression level of one or more nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i);(b) stratifying said subject as suffering from a distinct PNP subtype if one or more of the following are fulfilled(i) the proportion of one or more cell type(s) cluster of (a)(i) characterized by the one or more nerve-associated marker gene(s) and / or by the one or more immunological marker gene(s) is altered compared to a control sample;(ii) the expression level(s) of (a)(ii) of the one or more gene(s) of the nerve-associated differentiation gene(s) and / or of the one or more immunological differentiation gene(s) within the one or more cell type cluster of (i) is altered compared to a control sample; wherein the altered proportion of one or more cell type(s) cluster of (b)(i) and the altered expression levels of (b)(ii) are indicative for the distinct PNP subtype.

[0010] For these methods of the present invention, it is preferred that the cell type cluster is characterized by one or more nerve-associated marker gene(s) selected from the group consisting of mySC, nmSC, repairSC, damageSC, Adipo, ArtEC, endoC, epiC, LEC, PC1 , PC2, periCi , periC2, perC3, ven_capEC1 , ven_capEC2, venEC, and VSMC.

[0011] Additionally or alternatively, the cell type cluster may be characterized by one or more immunological marker gene(s) which is / are selected from the group consisting of CD4, Treg, MAIT, CD4_CD8, CD8, NK_CD8, NK, B, Plasma, Mast, Macrol , Macro2, Macro3, Macro4, Macro5, Macro6, Macro?, Macro8, Macro9, MacrolO, Macro11 , Macro12, Macro13, Macro14, Macrol 5, Macro16, Macrol 7, Macrol 8, cDC2_1 , cDC2_2, Macro_cDC, Macro_Granulo, Granulol , Granulo2, and cDC-1_pDC.

[0012] It is preferred for the methods of the present invention that the nerve-associated marker gene(s) is / are one or more of the gene(s) as depicted in Table 1. It is preferred for the methods of the present invention that the immunological marker gene(s) is / are one or more of the gene(s) as depicted in Table 2. It is further preferred that the nerve-associated differentiation gene(s) and the immunological differentiation gene(s) is / are one or more of the gene(s) as depicted in Tables 4 to 6.

[0013] In the methods of the present invention, PNP may be characterized by a decreased proportion of the one or more cell type cluster characterized by one or more nerve-associated marker gene(s) mySC and / or endoC; and an increased proportion of the one or more cell type(s) cluster VSMC, damageSC, periC3, repairSC, venEC, Adipo, and LEC. It may also bethat PNP is characterized by a decreased proportion of cells type cluster characterized by one or more immunological associated marker gene(s) NK, MacrolO, Macro?, and cDC2_1 ; and an increased proportion of the one or more cell type cluster Macro18, Macro_cDC, Macro5, Macro_Granulo, Macro6, and Macro"! .

[0014] In the context of the present invention, PNP may comprise(i) an increased expression level of one or more of the nerve-associated differentiation gene(s) of cell type(s) cluster ven_capEC1 , PC2, artEC, endoEC, nmSC, PC1 , mySC, periC2, periCi , ven_capEC2, epiC, Adipo, VSMC, repair SC, venEC, periC3, and LEC;(ii) an increased expression level of one or more of the immunological differentiation gene(s) of cell type cluster Macro2, Mast, and T_NK.

[0015] Further, it may also be preferred that PNP comprises(i) in the cell type cluster mySC the increased expression level of the nerve-associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of the following: CD74, CD53, COL15A1 , DCN, TNXB, COL1A1 , and IL4R;(ii) in the cell type cluster nmSC the increased expression level of the nerve-associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of the following: CSF2RA, IL13RA, TGFBI, and IL10RA;(iii) in the cell type cluster repairSC the increased expression level of the nerve-associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of TMEM47 and GALR1 ;(iv) in the cell type cluster PC2 the increased expression level of the nerve-associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of the following: PCDH11Y, MFAP5, and NLGN4Y.

[0016] It is preferred for the in vitro method for diagnosing of the present invention that PNP subtypes are characterized by a decreased proportion of the cell type(s) cluster mySC, and an increased proportion of the cell type cluster repairSC and / or damageSC.

[0017] It is also preferred for the in vitro method for diagnosing of the present invention that distinct PNP subtypes are characterized according to the following:(i) a first PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) damageSC, VSMC, periC3, repairSC, venC, Adipo, and ven_capEC2; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) ven_capEC1 , nmSC, arEC, epiC, periC2, PC1 , mySC, and LEC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Plasma, Macro 17, CD4_CD8, Macro_cDC, Macro_Granulo, Macro5; CD4, NK_CD8, and Treg; and / or(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) NK, Mast, Macro13; Macro14; cDC2_1 , and Macro2;(ii) a second PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) periC3, damageSC, venEC, Adipo, and repairSC; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) mySC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Macro6, Macro5, Macro17, Macro_Granulo, Macro3, and Macro_cDC; and / or(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro2, cDCD2_1 , Macrol 0, Granulol , T reg, NK_CD8, and Granulo2;(iii) a third PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) VSMC, Adipo, repairSC, LEC, and epiC; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) endoC and mySC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macrol 8, Macro5, Macro_cDC, Macro6, Macrol , and Macro_Granulo; and / or(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro7, CD8, NK, and Macrol 0.

[0018] Additionally, or alternatively, distinct PNP subtypes are characterized according to the following:(i) a first PNP subtype comprises an increased expression level of the nerve-associated differentiation gene(s) of the PC2 cell type cluster;(ii) a second PNP subtype comprises(a) an increased expression level of one or more gene(s) of the nerve-associated differentiation gene(s) of the ven_capEC1 , artEC, and PC2 cell type cluster; and / or(b) an increased expression level of the immunological differentiation gene(s) of the Macro2 cell type cluster;(iii) a third PNP subtype comprises(a) an increased expression level of one or more gene(s) of the nerve-associated differentiation gene(s) of the nmSC, periC2, PC2, PC1 , ven_capEC1 , and artEC cell type cluster; and / or(b) an increased expression level of the immunological differentiation gene(s) of the Macro2 cell type cluster.

[0019] It may be encompassed in the methods of the present invention that step (a) additionally comprises one or more of the following: spatial transcriptomics, histology, determining electrical conductivity, determining axon characteristics, such as axon diameter, axon integrity, myelin characteristics, neurological disability of the subject, bulk tissue transcriptomics, epigenetic analysis, protein quantification by techniques, such as flow cytometry, mass spectrometry.

[0020] In the context of the present invention, the sample may be a neurological sample obtained from the subject, preferably the neurological sample may be a nerve biopsy, more preferably a sural nerve biopsy; a skin biopsy, or a biopsy from any other tissue containing peripheral nerves or liquids, such as blood, tear fluid, urine, cerebrospinal fluid, or cellular material comprising nerves, meningeal biopsies or cerebrospinal fluid.

[0021] In a further aspect, the present invention relates to a method for determining therapeutic treatment of PNP in a subject comprising conducting the method of stratifying or the in vitro method for diagnosing of the present invention.

[0022] The present invention further relates to the use of the method of stratifying or of the in vitro method for diagnosing of the present invention for determining a treatment regime for a subject suffering from PNP.BRIEF DESCRIPTION OF THE FIGURES

[0023] Figure 1 shows the species-specific cellular landscape of human peripheral nerves. Figure 1A gives a schematic overview of the experimental procedure. Sural nerve biopsies were collected from 33 polyneuropathy (PNP) patients and four controls (CTRL) from three German centers (Essen, Munster, Wurzburg) and processed by single nuclei RNA- sequencing (snRNA-seq). Additionally, subcellular spatial transcriptomics (Xenium) was performed in a subgroup of eight patients. All 37 tissues were histologically characterized by quantifying myelin thickness (g-ratio), number of intactly myelinated axons and axon diameters. Figure 1 B shows LIMAP of 365,708 high-quality nuclei from 37 human sural nerves, showing 23 main clusters. Figure 1C gives a schematic depicting of novel marker genes of human myelinating Schwann cells (mySC), non-myelinating Schwann cells (nmSC), and perineurial cells (periC) found in this dataset were detected when reanalyzing published rodent dataset (Yim et al., Gerber et al., Wolbert et al.) and whether these markers were previously described in rodent or human literature. Figure 1D shows a representative section of an H&E staining overlaid with spatial-seq (Xenium) showing predicted clusters in spatial-seq of a sural nerve cross section of CTRL patient S24. Each dot represents one cell. Figure 1E to Figure 1G shows representative spatial-seq images of selected genes expressed in (E) perineurial cells, (F) in cells of the blood nerve barrier (BNB), and (G) in mySC (blue highlight) and nmSC (red highlight) in the sural nerve of CTRL patient S24. Each dot represents the expression of one transcript, a dotted line marks the perineurium, and a solid line surrounds individual vessels.

[0024] Figure 2 shows the heterogeneity of human nerve-associated immune cells. Figure 2A shows UMAP of 18,436 nuclei representing 35 immune cell (IC) subclusters of 37 human sural nerves. Figure 2B shows feature plots of known endoneurial and epineurial macrophage marker genes. Color encodes gene expression. Figure 2C shows gene ontology term enrichment analysis of marker genes (Iog2 fold change > 2, adjusted p value < 0.001) expressed by the Macro18 IC cluster in a one vs. all IC cluster comparison. Figure 2D shows representative spatial-seq images of the macrophage marker MS4A7 and the endoneurial macrophage markers CX3CR1 and TREM2 in the sural nerve of CTRL patient S24. Each dot represents the expression of one transcript, a dotted line marks the perineurium, and a solid line surrounds individual vessels. Figure 2E shows expression of immunoglobulin heavy (IGH) chain genes in the B cell (B) and plasma cell (plasma) cluster of the IC subclusters. Dot size encodes the percentage of cells expressing the transcript, and color shows the average gene expression.

[0025] Figure 3 shows identifying a pan-neuropathy cellular transcriptional response pattern. Figure 3A shows differences in cluster abundance in patients with polyneuropathy (PNP) vs. controls (CTRL), left: main clusters (see Fig. 1B), right: immune cell subclusters (see Fig. 2A). Only clusters with a Iog2 fold change > 0.5 are shown. Figure 3B shows correlations between individual single nuclei and PNP status (right), clinical Inflammatory Neuropathy Cause and Treatment (INCAT) disability score (middle), and g-ratio of myelinated axons (right) are shown. Red indicates a positive correlation and blue a negative correlation. Figure 3C shows the number of differentially expressed genes (DEG) per cell cluster that were determined in a pseudo bulk approach and are visualized. Figure 3D shows a neighborhood graph displaying the number of DEG between PNP and CTRL in a clustering-independent cellular neighborhood-based approach. Each dot represents a neighborhood, and the dot size represents the size of the neighborhood (Nhood), while edges depict the number of cells shared between neighborhoods. Color code shows the number of DEG. The right plot shows a magnification of the left plot, highlighting neighborhoods / clusters with the highest number of DEG. Figure 3E shows Volcano plots of DEG in myelinating Schwann cells (mySC), nonmyelinating SC (nmSC), repair SC, and pericytes 2 (PC2) between PNP and CTRL. The horizontal dashed line represents an adjusted p value of 0.1 and the vertical dashed lines display a Iog2 fold change of 2. Selected transcripts above those thresholds are labeled.

[0026] Figure 4 shows PNP subtypes that affect different cellular compartments of peripheral nerves. Figure 4A shows proportion of cells, split by disease group and colored by the main clusters. Figure 4B and Figure 4C show changes of cluster abundance in patients with vasculitic (VN), chronic inflammatory demyelinating (CIDP), and chronic idiopathic axonal (CIAP) polyneuropathy vs. controls (CTRL). The left plots show changes of the main clusters (as in Fig. 1B), the left plots show differences in the immune cell subclusters (see Fig. 2A). Only clusters with a Iog2 fold change > 0.5 are shown. Figure 4D shows representative sections of H&E staining of sural nerves overlaid with spatial seq (Xenium) showing predicted Schwann cell clusters in CTRL (S24, left plot), and VN (S30, right plot) patients. Each dot represents one cell. Figure 4E shows representative spatial transcriptomics images of perineurial marker genes of CTRL (S24), VN (S30), CIDP (S01), and CIAP (S14). Each dot represents the expression of one transcript. Figure 4F shows the density of PTPRC transcripts in the endoneurial (left plot) and epineurial (right plot) per disease group (n = 2 per disease group). Figure 4G shows neighborhood graphs displaying the number of DEG between PNP patients with VN (left), CIDP (middle) or CIAP (right) and CTRL in a clustering-independent cellular neighborhood-based approach. Dot size represents neighborhoods (Nhood), whileedges depict the number of cells shared between neighborhoods. Color code shows the number of DEG.

[0027] Figure 5: Descriptive statistic of the snRNa-seq cohort and experimental quality metrics. Figure 5A shows the average descriptive statistics of all patients: average age in years, gender, results of the motoric test of nerve conduction velocity (NCV) in meters per second, and INCAT disability score are shown. Figure 5B shows a schematic illustrating the technical study design and the simplified nuclei purification protocol (see methods-section given below). Figure 5C shows LIMAP of main clusters (see Fig. 1 B) split by center (left) or patients (right). Figure 5D shows the number of genes per nucleus for each individual sample. Figure 5E shows the proportion of cells, split by samples and colored by cluster.

[0028] Figure 6: Annotating nerve-associated cells in an unsupervised manner Figure 6A shows marker genes of main cell clusters (see Fig. 1 B). Figure 6B shows gene ontology term enrichment analysis of marker genes (Iog2 fold change > 2, adjusted p value < 0.001) expressed by the periC clusters in a one vs. all main cluster comparison. Figure 6C shows LIMAP of main clusters (see Fig. 1 B) with automatic annotation based on published rodent data (Yim et al., Gerber et al.). The rightmost plot shows vascular endothelial clusters only, automatically annotated with published human data (Mathy et al.). Figure 6D shows expression of novel cell marker genes in a published rodent dataset (Yim etal.) and this human study.

[0029] Figure 7: Spatially characterizing peripheral nerves and confirming novel transcripts. Figure 7A- Figure 7B show representative spatial transcriptomics images of (A) endothelial cell (Ec) markers and (B) Schwann cell (SC) markers in the sural nerve from a CTRL patient (S24). Each dot represents the expression of one transcript, a dotted line marks the perineurium, and a solid line surrounds the vasculature.

[0030] Figure 8: Defining endoneurial immune cells. Figure 8A shows marker genes of immune cell clusters (see Fig. 2A). Dot size encodes the percentage of cells expressing the transcript, and color code shows the average expression. Figure 8B shows gene ontology term enrichment analysis of differentially expressed (DE) genes upregulated (upper plots, Iog2 fold change > 2, adjusted p value < 0.1) or downregulated (lower plots, Iog2 fold change < 2, adjusted p value < 0.1) in PNP compared to CTRL in myelinating Schwann Cells (mySC) (left), non-myelinating SC (nmSC) (middle), and pericytes 2 (PC2) (right) clusters.

[0031] Figure 9: Histological characterization of sural nerve biopsies. Figure 9A shows representative overview images of one perineurium of a control patient (CTRL, S24) and patients with vasculitic (VN, S30), chronic inflammatory demyelination (CIDP, S01), and chronic idiopathic axonal (CIAP, S14) polyneuropathy stained with toluidine blue. Figure 9B shows g-ratios per sample. Each dot represents one g-ratio quantification. Figure 9C to Figure 9E shows histological measures (g-ratio, count of intactly myelinated axons, and diameter of axons) per disease group. Each dot represents the mean of the corresponding histological measure per patient. Figure 9F to Figure 9G shows the correlation between (F) g-ratio- (top), count- (middle), diameter of intactly myelinated axons (bottom) and the motoric nerve conduction velocity (NCV) of the tibial nerve (F), and between percentage of myelinating Schwann cells (mySC) (top), repair SC (bottom) and count of intactly myelinated axons. Each dot represents one patient. Figure 9H shows representative sections of H&E stainings of sural nerves overlaid with spatial-seq (Xenium) showing predicted Schwann cell clusters in CIDP (S01) and CIAP (S14) patients. Figure 9I shows quantification of predicted mySC, nmSC, and repairSC clusters in spatial-seq (Xenium) in the four disease-groups (n = 2 per group). Figure 9J to Figure 9K show the density of (J) CD3E and (K) MS4A1 transcripts in the endoneurial (left plot) and epineurial (right plot) per disease group (n = 2 per disease group). Figure 9L shows representative sections of H&E stainings of sural nerves overlaid with spatial-seq (Xenium) showing predicted T_NK cells in CTRL (S24), VN (S30), CIDP (S01) and CIAP (S14) patients. Figure 9M shows quantification of predicted T_NK cells in spatial-seq per disease group (n = 2 per group). Disease groups: CTRL (n = 4), VN (n = 5), CIDP (n = 9), CIAP (n = 11), cancer-associated / paraproteinemic (PPN, n = 2), diabetic (DPN, n = 2), other inflammatory (OIN, n = 2), other non-inflammatory (ONIN, n = 2).DETAILED DESCRIPTION OF THE INVENTION

[0032] The present invention relates in a first aspect to an in vitro method for diagnosing polyneuropathy (PNP) in a subject, comprising(a) determining by single cell RNA sequencing(i) a proportion of one or more cell type cluster characterized by one or more nerve- associated marker gene(s) and / or by one or more immunological marker gene(s) in a sample obtained from the subject;(ii) an expression level of one or more nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i);(b) diagnosing said subject as suffering from PNP if one or more of the following are fulfilled;(i) the proportion of one or more cell type cluster of (a)(i) characterized by the one or more nerve-associated marker gene(s) and / or by the one or more immunological marker gene(s) is / are altered compared to a control sample;(ii) the expression level of (a)(ii) of one or more of the nerve-associated differentiation gene(s) and / or of the one or more of the immunological differentiation gene(s) within the one or more cell type cluster of (i) is altered compared to a control sample; wherein the altered proportion of the one or more cell type cluster of (b)(i) and the altered expression levels of (b)(ii) are indicative for PNP.

[0033] “Polyneuropathy” as used herein means the simultaneous malfunction of many peripheral nerves throughout the body. The list of possible causes of PNP is long: impaired sugar metabolism, infections, malnutrition or alcoholism are just as much a part of it as chemotherapy, autoimmune diseases or genetic changes. The inventors of the present invention inter alia focus on the myelin sheath, i.e. the fat-rich protective sheath around the nerve fibers. The inventors of the present invention suspect that the fat metabolism in the nerves of PNP patients is disturbed, triggering a local immune reaction that sets further processes in motion that destroy the nerve sheath.

[0034] In the context of the present invention, “single cell RNA sequencing” examines the nucleic acid sequence information from individual cells with optimized next-generation sequencing technologies, providing a higher resolution of cellular differences and a better understanding of the function of an individual cell in the context of its microenvironment.

[0035] The term “cell type cluster” as used herein means the grouping of such cells which show common characteristics on the basis of single cell RNA sequencing and on the basis of the expression level of predefined genes, which comprise one or more nerve-associated marker gene(s) and / or immunological marker gene(s). In a preferred embodiment of theinvention, (a) cell type cluster is / are (also) characterized based on rodent data. In a further preferred embodiment of the invention, (a) cell type cluster is / are characterized based on human data. Cell type cluster in the context of the invention may be endoneurial cell types including Schwann cells, characterized in particular by the nerve-associated marker genes SWOB and SOXW, of both myelinating type mySC, characterzied in particular by the nerve- associated marker genes MPZ, MBP, and / or PRX, and non-myelinating type nmSC, preferably characterized by the nerve-associated marker genes NCAM1, L1CAM, and / or CDH2, and endoneurial fibroblast cells endoC, preferably characterized by the nerve-associated marker genes S0X9, PLXDC1, and / or ABCA9. Further cell type cluster according to the invention may be Schwann cells expressing features of damage damageSC, preferably characterized by the nerve-associated marker genes EGR1, FOS, and / or JUN, and of repair15repairSC, preferably characterized by the nerve-associated marker genes NGFR, ATF3, GDNF, and / or RUNX2. Further cell type cluster according to the invention may be epineurial fibroblast cells epiC, preferably characterized by the nerve-associated marker genes CCBE1, and / or COMP, or perineurial fibroblast cells periCi -3, preferably characterized by the nerve-associated marker genes SLC2A 7 / GLUT1 , KRT19, and / or CLDN1. Further cell type cluster according to the invention may be vascular cells including vascular smooth muscle cells VSMC, preferably characterized by the nerve-associated marker genes ACTA2, and / or CARMN, or pericytes PC 1-2, preferably characterized by the nerve-associated marker genes PDGRFB, and / or RGS5, or endothelial cells EC, preferably characterized by the nerve-associated marker genes EGFL7, and / or PECAM1. According to the present invention, further cell type cluster may be endothelial cells EC separated into lymphatic endothelial cells LEC, preferably characterized by the nerve-associated marker genes PR0X1, LYVE1, and / or FLT4, or venous ven_EC, preferably characterized by the nerve-associated marker genes PLVAP, and / or ACKR1, or capillary capEC, preferably characterized by the nerve-associated marker genes ABCG2, and / or MFSD2A, or arterial artEC, preferably characterized by the nerve-associated marker genes SEMA3G, HEY1, and / or GJA5 continuum. Further cell type cluster according to the invention are venous / capillary EC ven_cap_EC2, which are preferably characterized by bloodnerve barrier (BNB) markers ABCB1 and SLC1A116, blood-brain barrier marker MFSD2A17, and tight junction transcript GJA 1 and therefore likely represented EC of the BNB. Further cell type cluster according to the invention may be nerve-associated leukocytes, such as myeloid lineage Macrol , which is preferably characterized by the immunological marker genes MS4A7, and / or CLECWA; such as Macro2, which is preferably characterized by the immunological marker genes MS4A7, and / or CX3CR1', such as Granulo, which is preferably characterized by the immunological marker genes SW0A8, and / or SW0A9; such as Mast, which is preferably characterized by the immunological marker genes CPA3 and / or MS4A2', such as T / NK cells, which are preferably characterized by the immunological marker genes CD3E, CD8A, NCR1,NKG7, and / or GZMA, or B cells, which are preferably characterized by the immunological marker genes CD79A, and / or IGHM.

[0036] Immunological cell type cluster according to the invention may be those as depicted in Tabel 2 as given herein. Those cell type cluster may comprise according to the invention myeloid lineage such as macrophage annotation Macro1-18, which is preferably characterized by the immunological marker genes LYZ, CD14, MRC1, CD163, and / or MS4A7; such as dendritic cells (DC) type 1 cDC1 , which are preferably characterized by the immunological marker genes CLEC9A, XCR1, and / or BATF3, such as plasmacytoid DC pDC, which is preferably characterized by the immunological marker genes CLEC4C, and / or IRF8, and such as DC type 2 cDC2, which is preferably characterized by the immunological marker genes FCER1A, CD1C, and / or CLEC10A such as T / NK cell clusters comprising CD4 T cell phenotypes, including naive CD4, which is preferably characterized by the immunological marker genes IL7R, CCR7, MAL, and / or SELL, such as Treg, which is preferably characterized by the immunological marker genes FOXP3, and / or IL2RA, such as MAIT, which is preferably characterized by the immunological marker genes KLRB1, and / or CXCR6, or such as CD8 T cells, which are preferably characterized by the immunological marker genes CD8A, and / or CCL5, or such as NK cells, which are preferably characterized by the immunological marker genes FCGR3A, NKG7, and / or NCR1. Further, cell type cluster in accordance with the present invention may be naive B cells B, which are preferably characterized by the immunological marker genes CD19, MS4A 1 / CD20, and / or CD79B, or plasma cells, which are preferably characterized by the immunological marker genes JCHAIN, and / or SDC7 / CD138, and may comprise IGHG1, IGHG3, IGHG4, and / or IGHA1 heavy chain gene expression. Further, cell type cluster may comprise granulocytes, which are preferably characterized by the immunological marker genes S100A8, S100A9; and / or mast cells, which are preferably characterized by the immunological marker genes CPA3, TPSAB1, and / or MS4A2.

[0037] The term “proportion of one or more cell type cluster” as used herein means the extent of the part of a distinct cell type cluster within the overall range of all cell type cluster.

[0038] The term “nerve-associated marker gene(s)” as used herein means one or more gene(s) which is / are expressed in a cell or a cluster of cells which is located in the nerve or in any tissue or surrounding texture in the direct or indirect neighbourhood of a nerve.

[0039] In the context of the present invention, the term “immunological marker gene(s)” means one or more gene(s) which is / are expressed in a cell or a cluster of cells which is assigned or associated with a cell of the immune system.

[0040] The term “control sample”, in the context of the present invention means any sample of a healthy subject or a subject not being suspected of having PNP.

[0041] In the context of the present invention, the term “altered” may mean a decrease or an increase of the respective value. For example, if the proportion of cells or of cell type cluster has altered, this may also mean that the relative abundance has changed.

[0042] The term “expression level of a gene” as used herein means the activity of a gene which comprises transcription of a gene. Accordingly, the expression level of a gene comprises determining the transcription of a gene on RNA level, and / or determining the transcriptional products on protein level.

[0043] In a second aspect, the present invention relates to a method of stratifying a subject with polyneuropathy (PNP), comprising(a) determining by single cell RNA sequencing(i) a proportion of one or more cell type cluster characterized by one or more nerve- associated marker gene(s) and / or one or more immunological marker gene(s) in a sample obtained from the subject;(ii) an expression level of one or more nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i);(b) stratifying said subject as suffering from a distinct PNP subtype if one or more of the following are fulfilled(i) the proportion of one or more cell type(s) cluster of (a) (i) characterized by the one or more nerve-associated marker gene(s) and / or by the one or more immunological marker gene(s) is altered compared to a control sample;(ii) the expression level(s) of (a)(ii) of one or more gene(s) of the nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i) is altered compared to a control sample; wherein the altered proportion of one or more cell type(s) cluster of (b)(i) and the altered expression levels of (b)(ii) are indicative for the distinct PNP subtype.

[0044] The expression “stratifying” as used in the context of the present invention may mean sorting individuals with a probable diagnosis of PNP or diagnosed with PNP into different classes or subtypes of PNP.

[0045] For the methods of the present invention, it is preferred that the cell type cluster ischaracterized by one or more nerve-associated marker gene(s) selected from the group consisting of mySC, nmSC, repairSC, damageSC, Adipo, ArtEC, endoC, epiC, LEC, PC1 , PC2, periCi , periC2, perC3, ven_capEC1 , ven_capEC2, venEC, and VSMC.

[0046] Additionally or alternatively, the cell type cluster may be characterized by one or more immunological marker gene(s) which is / are selected from the group consisting of CD4, Treg, MAIT, CD4_CD8, CD8, NK_CD8, NK, B, Plasma, Mast, Macrol , Macro2, Macro3, Macro4, Macro5, Macro6, Macro?, Macro8, Macro9, MacrolO, Macro11 , Macro12, Macro13, Macro14, Macrol 5, Macro16, Macrol 7, Macrol 8, cDC2_1 , cDC2_2, Macro_cDC, Macro_Granulo, Granulol , Granulo2, and cDC-1_pDC.

[0047] It is preferred for the methods of the present invention that the nerve-associated marker gene(s) is / are one or more of the gene(s) as depicted in Table 1 . It is preferred for the methods of the present invention that the immunological marker gene(s) is / are one or more of the gene(s) as depicted in Table 2. It is further preferred that the nerve-associated differentiation gene(s) and the immunological differentiation gene(s) is / are one or more of the gene(s) as depicted in Tables 4 to 6.

[0048] In the methods of the present invention, PNP may be characterized by a decreased proportion of the one or more cell type cluster characterized by one or more nerve-associated marker gene(s) mySC and / or endoC; and / or an increased proportion of the one or more cell type(s) cluster VSMC, damageSC, periC3, repairSC, venEC, Adipo, and / or LEC. It may also be that PNP is characterized by a decreased proportion of cells type cluster characterized by one or more immunological associated marker gene(s) NK, MacrolO, Macro7, and / or cDC2_1 ; and an increased proportion of the one or more cell type cluster Macrol 8, Macro_cDC, Macro5, Macro_Granulo, Macro6, and / or Macro!

[0049] In the context of the present invention, PNP may comprise(i) an increased expression level of one or more of the nerve-associated differentiation gene(s) of cell type cluster ven_capEC1 , PC2, artEC, endoEC, nmSC, PC1 , mySC, periC2, periCi , ven_capEC2, epiC, Adipo, VSMC, repair SC, venEC, periC3, and LEC;(ii) an increased expression level of one or more of the immunological differentiation gene(s) of cell type cluster Macro2, Mast, and T_NK.

[0050] In addition or alternatively, it may also be that PNP comprises(i) in the cell type cluster mySC the increased expression level of the nerve-associated differentiation gene(s) and of the immunological differentiation gene(s) of one or more of the following: CD74, CD53, COL15A1 , DCN, TNXB, COL1A1 , and IL4R; and / or(ii) in the cell type cluster nmSC the increased expression level of the nerve-associated differentiation gene(s) and of the immunological differentiation gene(s) of one or more of the following: CSF2RA, IL13RA, TGFBI, and IL10RA; and / or(iii) in the cell type cluster repairSC the increased expression level of the nerve-associated differentiation gene(s) and of the immunological differentiation gene(s) of one or more of TMEM47 and GALR1 ; and / or(iv) in the cell type cluster PC2 the increased expression level of the nerve-associated differentiation gene(s) and of the immunological differentiation gene(s) of one or more of the following: PCDH11Y, MFAP5, and NLGN4Y.

[0051] It is preferred for the in vitro method for diagnosing of the present invention that PNP subtypes are characterized by a decreased proportion of the cell type(s) cluster mySC, and an increased proportion of the cell type cluster repairSC and / or damageSC.

[0052] It is also preferred for the in vitro method for diagnosing of the present invention that distinct PNP subtypes are characterized according to the following:(i) a first PNP subtype may comprise(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) damageSC, VSMC, periC3, repairSC, venC, Adipo, and ven_capEC2; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) ven_capEC1 , nmSC, artEC, epiC, periC2, PC1 , mySC, and LEC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Plasma, Macro 17, CD4_CD8, Macro_cDC, Macro_Granulo, Macro5; CD4, NK_CD8, and Treg; and / or(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) NK, Mast, Macro13; Macro14; cDC2_1 , and Macro2;(ii) a second PNP subtype may comprise(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) periC3, damageSC, venEC, Adipo, andrepairSC; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) mySC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Macro6, Macro5, Macro17, Macro_Granulo, Macro3, and Macro_cDC;(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro2, cDCD2_1 , Macrol 0, Granulol , T reg, NK_CD8, and Granulo2;(iii) a third PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) VSMC, Adipo, repairSC, LEC, and epiC; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) endoC and mySC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macrol 8, Macro5, Macro_cDC, Macro6, Macrol , and Macro_Granulo;(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro?, CD8, NK, and Macrol 0.

[0053] In the context of the present invention, the first PNP subtype may be vasculitic polyneuropathy (VN). In the present invention, the second PNP subtype may be CIDP. CIPD as used herein may mean chronic inflammatory demyelinating polyneuropathy. It is also encompassed by the present invention that the third PNP subtype may be CIAP. CIAP as used herein may mean chronic idiopathic axonal polyneuropathy.

[0054] Additionally, or alternatively, distinct PNP subtypes may be characterized according to the following:(i) a first PNP subtype may comprise an increased expression level of the nerve-associated differentiation gene(s) of the PC2 cell type cluster;(ii) a second PNP subtype may comprise(a) an increased expression level of one or more gene(s) of the nerve-associated differentiation gene(s) of the ven_capEC1 , artEC, and PC2 cell type cluster; and / or(b) an increased expression level of the immunological differentiation gene(s) of the Macro2 cell type cluster;(ii) a third PNP subtype may comprise(a) an increased expression level of one or more gene(s) of the nerve-associated differentiation gene(s) of the nmSC, periC2, PC2, PC1 , ven_capEC1 , and artEC cell type cluster; and / or(b) an increased expression level of the immunological differentiation gene(s) of the Macro2 cell type cluster.

[0055] It may be encompassed in the methods of the present invention that step (a) additionally comprises one or more of the following: spatial transcriptomics, histology, determining electrical conductivity, determining axon characteristics, such as axon diameter, axon integrity, myelin characteristics, neurological disability of the subject, bulk tissue transcriptomics, epigenetic analysis, protein quantification by techniques such as flow cytometry, mass spectrometry.

[0056] In the context of the present invention, the sample may be a neurological sample obtained from the subject, preferably the neurological sample may be a nerve biopsy, more preferably a sural nerve biopsy, or skin biopsy, or a biopsy from any other tissue containing peripheral nerves or liquids, such as blood, tear fluid, urine, cerebrospinal fluid, or cellular material comprising nerves, or meningeal biopsies, or cerebrospinal fluid.

[0057] In a further aspect, the present invention relates to a method for determining therapeutic treatment of PNP in a subject comprising conducting the method of stratifying or the in vitro method of diagnosing of the present invention.

[0058] The present invention further relates to the use of the method of stratifying or the in vitro method of diagnosing of the present invention for determining a treatment regime for a subject suffering from PNP.

[0059] The following Tables are relevant for the present invention.

[0060] Table 1 : This Table gives an overview of examples of nerve-associated marker gene(s) of the present invention. Thus, this Table shows the top markers of the dataset of the main cluster of the present invention.

[0061] Table 2: This Table gives an overview of examples of immunological marker gene(s) of the present invention. Thus, this Table shows the top markers of the dataset of the immune cell cluster of the present invention.

[0062] Table 3: This Table shows the differentially expressed genes in polyneuropathy vs. control patients per cluster.

[0063] Table 4: This Table shows the differentially expressed genes in VN vs. control patients per cluster

[0064] Table 5: This Table shows the differentially expressed genes in Cl DP vs. control patients per cluster

[0065] Table 6: This Table shows the differentially expressed genes in CIAP vs. control patients per cluster0066] Table 7: Patient information623

[0067] The invention is further characterized by the following items:1 . An in vitro method for diagnosing polyneuropathy (PNP) in a subject, comprising(a) determining by single cell RNA sequencing(i) a proportion of one or more cell type cluster characterized by one or more nerve-associated marker gene(s) and / or by one or more immunological marker gene(s) in a sample obtained from the subject;(ii) an expression level of one or more nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i);(b) diagnosing said subject as suffering from PNP if one or more of the following are fulfilled(i) the proportion of one or more cell type cluster of (a)(i) characterized by the one or more nerve-associated marker gene(s) and / or by the one or more immunological marker gene(s) is / are altered compared to a control sample;(ii) the expression level of (a)(ii) of one or more of the nerve-associated differentiation gene(s) and / or of one or more of the immunological differentiation gene(s) within the one or more cell type cluster of (i) is altered compared to a control sample; wherein the altered proportion of one or more cell type cluster of (b)(i) and the altered expression levels of (b)(ii) are indicative for PNP.2. A method of stratifying a subject with polyneuropathy (PNP), comprising(a) determining by single cell RNA sequencing(i) a proportion of one or more cell type cluster characterized by one or more nerve-associated marker gene(s) and / or one or more immunological marker gene(s) in a sample obtained from the subject;(ii) an expression level of one or more nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i);(b) stratifying said subject as suffering from a distinct PNP subtype if one or more of the following are fulfilled(i) the proportion of one or more cell type cluster of (a)(i) characterized by the one or more nerve-associated marker gene(s) and / or by the one or more immunological marker gene(s) is altered compared to a control sample;(ii) the expression level(s) of (a)(ii) of one or more gene(s) of the nerve- associated differentiation gene(s) and / or of one or more immunologicaldifferentiation gene(s) within the one or more cell type cluster of (i) is altered compared to a control sample; wherein the altered proportion of one or more cell type cluster of (b)(i) and the altered expression levels of (b)(ii) are indicative for the distinct PNP subtype. The method of item 1 or 2, wherein the cell type cluster characterized by one or more nerve-associated marker gene(s) is selected from the group consisting of mySC, nmSC, repairSC, damageSC, Adipo, ArtEC, endoC, epiC, LEC, PC1 , PC2, periCi , periC2, perC3, ven_capEC1 , ven_capEC2, venEC, and VSMC. The method of any one of items 1 to 3, wherein the cell type cluster characterized by one or more immunological marker gene(s) is selected from the group consisting of CD4, Treg, MAIT, CD4_CD8, CD8, NK_CD8, NK, B, Plasma, Mast, Macrol , Macro2, Macro3, Macro4, Macro5, Macro6, Macro?, Macro8, Macro9, Macrol 0, Macrol 1 , Macrol 2, Macro13, Macro14, Macro15, Macro16, Macro17, Macro18, cDC2_1 , cDC2_2, Macro_cDC, Macro_Granulo, Granulol , Granulo2, and cDC-1_pDC. The method of any one of items 1 to 4, wherein the nerve-associated marker gene(s) is / are one or more of the gene(s) as depicted in Table 1. The method of any one of items 1 to 5, wherein the immunological marker gene(s) is / are one or more of the gene(s) as depicted in Table 2. The method of any one of items 1 to 6, wherein the nerve-associated differentiation gene(s) and the immunological differentiation gene(s) is / are one or more of the gene(s) as depicted in Tables 4 to 6. The method of any one of items 1 to 7, wherein PNP is characterized by a decreased proportion of the one or more cell type cluster characterized by one or more nerve- associated marker gene(s) mySC and / or endoC; and an increased proportion of the one or more cell type(s) cluster VSMC, damageSC, periC3, repairSC, venEC, Adipo, and LEC. The method of any one of items 1 to 8, wherein PNP is characterized by a decreased proportion of a cell type cluster characterized by one or more immunological associated marker gene(s) NK, MacrolO, Macro?, and cDC2_1 ; and an increased proportion of the one or more cell type cluster Macrol 8, Macro_cDC, Macro5, Macro_Granulo, Macro6, and Macrol .The method of any one of items 1 to 9, wherein PNP comprises(i) an increased expression level of one or more of the nerve-associated differentiation gene(s) of cell type cluster ven_capEC1, PC2, artEC, endoEC, nmSC, PC1, mySC, periC2, periCi, ven_capEC2, epiC, Adipo, VSMC, repair SC, venEC, periC3, and LEC;(ii) an increased expression level of one or more of the immunological differentiation gene(s) of cell type cluster Macro2, Mast, and T_NK. The method of any one of items 1 to 10, wherein PNP comprises(i) in the cell type cluster mySC the increased expression level of the nerve- associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of the following: CD74, CD53, COL15A1 , DCN, TNXB, COL1A1, and IL4R; and / or(ii) in the cell type cluster nmSC the increased expression level of the nerve- associated differentiation gene(s) and of the immunological differentiation gene(s) of one or more of the following: CSF2RA, IL13RA, TGFBI, and IL10RA; and / or(iii) in the cell type cluster repairSC the increased expression level of the nerve- associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of TMEM47 and GALR1; and / or(iv) in the cell type cluster PC2 the increased expression level of the nerve-associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of the following: PCDH11Y, MFAP5, and NLGN4Y. The method of any one of items 2 to 7, wherein PNP subtypes are characterized by a decreased proportion of the cell type cluster mySC, and an increased proportion of the cell type cluster repairSC and / or damageSC. The method of any one of items 2 to 7, and 12 wherein distinct PNP subtypes are characterized according to the following:(i) a first PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) damageSC, VSMC, periC3, repairSC, venC, Adipo, and ven_capEC2; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) ven_capEC1, nmSC, arEC, epiC, periC2, PC1 , mySC, and LEC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Plasma, Macro 17, CD4_CD8, Macro_cDC, Macro_Granulo, Macro5, CD4, NK_CD8, and Treg; and / or(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) NK, Mast, Macro13; Macro14; cDC2_1 , and Macro2;(ii) a second PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) periC3, damageSC, venEC, Adipo, and repairSC; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) mySC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Macro6, Macro5, Macro17, Macro_Granulo, Macro3, and Macro_cDC; and / or(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro2, cDCD2_1 , MacrolO, Granulol , Treg, NK_CD8, and Granulo2;(iii) a third PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) VSMC, Adipo, repairSC, LEC, and epiC; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) endoC, and mySC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Macro5, Macro_cDC, Macro6, Macrol , and Macro_Granulo;(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro7, CD8, NK, and MacrolO. The method of any one of items 2 to 7, and 12 or 13, wherein distinct PNP subtypes are characterized according to the following:(i) a first PNP subtype comprisesan increased expression level of the nerve-associated differentiation gene(s) of the PC2 cell type cluster;(ii) a second PNP subtype comprises(a) an increased expression level of one or more gene(s) of the nerve-associated differentiation gene(s) of the ven_capEC1 , artEC, and PC2 cell type cluster; and / or(b) an increased expression level of the immunological differentiation gene(s) of the Macro2 cell type cluster;(ii) a third PNP subtype comprises(a) an increased expression level of one or more gene(s) of the nerve-associated differentiation gene(s) of the nmSC, periC2, PC2, PC1 , ven_capEC1 , and artEC cell type cluster; and / or(b) an increased expression level of the immunological differentiation gene(s) of the Macro2 cell type cluster. The method of any one of items 1 to 14, wherein step (a) additionally comprises one or more of the following: spatial transcriptomics, histology, determining electrical conductivity, determining axon characteristics, such as axon diameter, axon integrity, myelin characteristics, neurological disability of the subject, bulk tissue transcriptomics, epigenetic analysis, protein quantification by techniques such as flow cytometry, mass spectrometry. The method of any one of items 1 to 15, wherein the sample is a neurological sample obtained from the subject, preferably wherein the neurological sample is a nerve biopsy, more preferably a sural nerve biopsy; skin biopsy, or a biopsy from any other tissue containing peripheral nerves or liquids; such as blood, tear fluid, urine, cerebrospinal fluid; cellular material comprising nerves, meningeal biopsies, or cerebrospinal fluid. A method for determining therapeutic treatment of PNP in a subject comprising conducting the method of any one of items 1 to 16. Use of the method of any one of items 1 to 16 for determining a treatment regime for a subject suffering from PNP.Further Definitions

[0068] It is noted that as used herein, the singular forms “a”, “an”, and “the”, include plural references unless the context clearly indicates otherwise. Thus, for example, reference to “a reagent” includes one or more of such different reagents and reference to “the method” includes reference to equivalent steps and methods known to those of ordinary skill in the art that could be modified or substituted for the methods described herein.

[0069] Unless otherwise indicated, the term “at least” preceding a series of elements is to be understood to refer to every element in the series. The term “at least one” refers, if not particularly defined differently, to one or more such as two, three, four, five, six, seven, eight, nine, ten or more. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the present invention.

[0070] The term “and / or” wherever used herein includes the meaning of “and”, “or” and “all or any other combination of the elements connected by said term”.

[0071] The term “less than” or in turn “more than” does not include the concrete number.

[0072] For example, less than 20 mean less than the number indicated. Similarly, more than or greater than means more than or greater than the indicated number, e.g. more than 80 % means more than or greater than the indicated number of 80 %.

[0073] Throughout this specification and the claims which follow, unless the context requires otherwise, the word “comprise”, and variations such as “comprises” and “comprising”, will be understood to imply the inclusion of a stated integer or step or group of integers or steps, but not the exclusion of any other integer or step or group of integer or step. When used herein, the term “comprising” can be substituted with the term “containing” or “including” or sometimes when used herein with the term “having”. When used herein “consisting of” excludes any element, step, or ingredient not specified.

[0074] The term “including” means “including but not limited to”. “Including” and “including but not limited to” are used interchangeably.

[0075] The term “about” preceding a respective value means plus or minus 10 % of that value, preferably plus or minus 5 %, more preferably plus or minus 2 %, most preferably plus or minus1 %, and also includes the value itself.

[0076] Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

[0077] It should be understood that this invention is not limited to the particular methodology, protocols, material, reagents, and substances, etc., described herein and as such can vary. The terminology used herein is for the purpose of describing particular embodiments only, and is not intended to limit the scope of the present invention, which is defined solely by the claims.

[0078] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, numerous equivalents to the specific procedures, embodiments, claims, and examples described herein. Such equivalents are considered to be within the scope of this invention and covered by the claims appended hereto.

[0079] A better understanding of the present invention and of its advantages will be gained from the following examples, offered for illustrative purposes only.EXAMPLES OF THE INVENTION

[0080] The following Examples illustrate the invention, but are not to be construed as limiting the scope of the invention.

[0081] MATERIALS & METHODS

[0082] Collection of patient samples

[0083] The inventors of the present invention collected sural nerves from 33 PNP patients from three centers: University Hospital of Munster (10), Essen (11), and Wurzburg (12). Additionally, the inventors obtained sural nerves from four control patients in Essen with traumatic nerve injuries, who received a sural nerve autograft as interposition, but were unaffected by polyneuropathy (see Table 7). Residual material from the sural nerve grafts were used for control samples in this study. In PNP patients, sural nerve biopsy was performed as part of the clinical diagnostic workup and a small portion of the nerve (approximately 0.2 cm) was used for this study. In Munster and Essen, the samples were immediately fresh-frozen after surgery in dry ice-cooled methylbutane and stored in liquid nitrogen (0 - 14 months, see Table 7) until nuclei extraction. In Wurzburg, the samples were embedded into OCT embedding matrix (Roth) until nuclei extraction (33 - 77 months, see Table 7). In Essen and Munster samples were collected from newly recruited patients. In Wurzburg, samples were collected from an existing biobank of OCT embedded cryopreserved sural nerve biopsies collected between 2016 and 2020. All experiments were carried out in accordance with the Declaration of Helsinki and were approved by the local ethical committees in Munster (2018-719-f-S), Essen (21-10376-BO), and Wurzburg (238 / 17; 15 / 19). All patients gave written informed consent to sample collection.

[0084] Patient characteristics

[0085] Patients were characterized clinically regarding diagnosis, disease activity, therapy response, disease duration, immunosuppressive therapy, relevant secondary diagnosis, electrophysiological studies and CSF analysis outlined in Table 7. Patients were diagnosed according to the EAN / PNS criteria 2021 for CIDP45and according to the PNS criteria for VN.

[0086] Nuclei extraction and purification

[0087] For each patient, approximately 10 mg pieces (median of 13.9 mg) were cut on dry ice as starting material from the original samples. Nerve pieces were cut into smaller pieces and processed using the Miltenyi Biotec nuclei extraction protocol. In summary, samples were transferred to a gentleMACS C-tube (Cat. no. 130-093-237) containing 2 ml lysis buffer (nuclei extraction buffer (Cat. no. 130-128-024) + 0.2 U / pl RNase inhibitor (EO0381) and processed on a gentleMACS dissociator with program 4C_nuclei_1. After dissociation, the nuclei suspensionwas filtered through a pluriStrainer Mini 70-pm cell strainer (pluriSelect®) and washed in 200-400 pl resuspension buffer (PBS with 0.1 % BSA and 0.2 LI / pL RNase inhibitor (EO0381) depending on sample size and expected nuclei count. Nuclei suspension was then filtered through a pluriStrainer Mini 40-pm cell strainer (pluriSelect®), placed on a 1.5 ml DNA-LoBind tube (Merck). The nuclei suspension was stained with Trypan Blue to manually assess nuclei viability and count using a Fuchs-Rosenthal chamber. Equal volumes between samples were used for downstream application single nuclei RNA seq.

[0088] Single nuclei RNA-sequencing and generation of count matrices

[0089] Single nuclei suspensions were loaded into a Chromium Next GEM Chip G and placed into the Chromium X Single Cell Controller and processed with Chromium Next GEM Single Cell 3' Kit v3.1 reagents (all 10X Genomics). Sequencing was performed on Illumina Nextseq 2000 and Novaseq 6000 with a 28-8-0-91 read setup. The inventors used cellranger v7.0.1 (10X Genomics) to generate count matrices with default parameters but an optimized transcriptome reference v1.146for GRCh38.

[0090] Single nuclei analysis

[0091] CellBender v0.3.047was used to remove background noise. In CellBender, the number of expected cells and total droplets was based on the UM I curve and the learning rate was reduced based on the automated output report. Further downstream analysis was performed with the R package Seurat v5.0.148. To increase computational efficiency and reduce memory usage of this large dataset, the inventors used BPCells vO.1.0. Low quality cells were filtered for each sample individually by inspecting quality control plots and removing cells with higher mitochondrial percentages (range: 1-5%), low (<200) or very high molecule counts per cell (range: 6000-9000). Doublets were removed using scDblFinder v1.16.049with default parameters. The inventors used Seurat for normalization (LogNormalize, default parameters), identification of highly variable genes (vst method, 2000 features), scaling, and performing PCA (default parameters). Next, the inventors used atomic sketch integration (method LeverageScore, 5000 cells) in Seurat to reduce memory usage. Batch effects were accounted for by integrating the samples with scVI v1.0.45°. After inspecting the effect of batch removal, the full dataset was integrated with Seurat. The scVI integrated full dataset was used to calculate the UMAP embeddings with Seurat (30 dimensions). The inventors identified clusters with the FindNeighbors and FindClusters (resolution 0.7) functions. The inventors calculated the top expressing genes of each cluster with FindMarkers (min. pct = 0.1 , logfc_threshold = 0.25, p_val_adj < 0.05, two-sided Wilcoxon rank-sum test). To subcluster the immune cells, the inventors selected the immune cell clusters of the main clusters. The inventors then reperformed identification of highly variable genes, scaling, and PCA as explained above. The data were integrated with reciprocal PCA using Seurat and after plausibility checks of the batch removal, UMAP was calculated (30 dimensions). Clusters were identified withFindNeighbors and FindClusters (resolution 2.3). The top expressing genes were identified as explained for the main clusters. Enrichment analysis of top markers (avg_log2FC > 1 , p_val_adj < 0.001) was carried out with enrichR 3.251with the GO Biological Process 2023 database. Differentially expressed genes between conditions were identified using a pseudobulk method with Libra v1 .0.052(edgeR 4.0.753, two-sided likelihood ratio test). Volcano plots of DE genes were created with Enhanced Volcano v1.20.0. Additionally, the inventors determined DE genes in a cluster-independent manner with miloDE v0.0.0.954following the tutorial. Briefly, neighborhoods were assigned based on the scvi integrated data (k = 30, prop = 0.1 , d = 30) and DE testing was conducted between conditions (min_count =10). To determine the cluster abundance, the inventors used propeller55(part of speckle v1.2.0). The inventors used the co-varying neighborhood analysis (CNA), implemented in the rcna package v0.0.9956, to compute correlations between clinical phenotypes or histological measures and our single cell data independent of clustering controlling for age and sex. To classify patients the inventors used the PCA of the cluster abundance and multi-resolution variational interference (MrVI)26. PCA of the cluster abundances and variable contributions were calculated and visualized with FactoMineR v2.957. To run MrVi v0.2.0, the Seurat object was converted to h5ad with sceasy v0.0.758and read as an AnnData object in Python. The MrVi model was trained following the tutorial. The samplesample distances were predicted with MrVi (get_local_sample_representation) for each cell and then averaged over all cells. Sample-samples distances were visualized in a heatmap using pheatmap v1.0.12 in R (ward.D2, euclidean distance measure).

[0092] Comparison with published datasets

[0093] The inventors downloaded the publicly available annotated data from Yim et al.9(GEO GSE182098, sciatic nerve), Gerber et al.10(https: / / snat.ethz.ch / seurat-objects.html, 10X Genomics P60), Mathys et al. (https: / / compbio.mit.edu / scBBB / ROSMAP vascular cells)14, and Wolbert et al.8(GEO GSE142541 , mouse). If necessary, the data were preprocessed with Seurat, i.e. normalized with LogNormalize, highly variable genes were identified, scaled, and PCA was performed with default parameters. To identify novel marker genes, cell markers in the published datasets were determined with the FindMarkers function in Seurat (min. pct = 0.1 , logfc.threshold = 0.25, p_val_ad < 0.05). To annotate the dataset of the inventors based on the published dataset, rodent gene names were converted to orthologues using homoIogene v.1.4.68 (homologeneData2 database). The inventors then classified their cells based on the annotated reference dataset using FindTransferAnchors and TransferData functions in Seurat with default parameters. Cells were labeled as unknown if the Seurat prediction score was below 0.3.

[0094] Histology: Semi-thin sections and analysis

[0095] After biopsy,1Zs of the sural nerve was embedded in Epoxy resin (see Fig. 1A, experimental setup). Semithin sections (1 pm) of sural nerves were cut using a EM UC7ultramicrotome (Leica Microsystems) and stained with toluidine blue. The sections were scanned with a slide scanner (Munster / Wurzburg: Grundium Ocus40; Essen: Zeiss Axio Scan Z.1 , Hitachi HV F203SCL camera). After blinding, the entire number of total myelinated axons per sural nerve were counted manually by one investigator using the CellCounter plug-in of Imaged v1.36. Physiologically unmyelinated axons (diameter <1 pm) and Remak-bundle fibers were not included.

[0096] The Imaged g-ratio plugin was used to determine myelin thickness and axonal diameter of 15% of the total myelinated axons per sural nerve. In randomly chosen myelinated axons the inner and outer rim of a myelin sheath were manually traced by one investigator. The g-ratio was calculated by dividing the axonal circumference (i.e. , inner rim of the myelin sheath) by the outer circumference of the respective myelin sheath, presuming circular axons. The g-ratio values of each patient were plotted against the calculated axon diameters.

[0097] Preparation of sural nerve cross sections and Xenium spatial transcriptomics

[0098] The inventors created a custom stand-alone gene panel for spatial transcriptomics using the commercial Xenium Analyzer platform. The panel was designed based on single nuclei analysis of the inventors, mostly focussing on cells that demonstrated a potential role in polyneuropathies, including Schwann cells (mySc, nmSC, repairSC), perineurial cells and vascular cells such as pericytes and endothelial cells. The inventors included a minimum of four top marker genes of each of the clusters and also defined twenty DE genes that were differentially expressed between the different PNP subtypes. The inventors additionally included well known markers of both epi- and endoneurial fibroblasts and different subtypes of leukocytes including B cells, T cells, and macrophages, adding up to a total of 99 genes.

[0099] In total, 8 sural nerve formalin-fixed paraffin-embedded (FFPE) samples were included for Xenium spatial transcriptomics: 2 CTRL (S22, S24), 2 CIDP (S01 , S11), 2 CIAP (S04, S14) and 2 VN (S29, S30) samples (see Table 7). Samples were primarily selected based on their optimal / intact morphology in the semithin sections and their center with the aim to limit center bias.

[0100] Sample preparation and processing was carried out by the CMCB Technology Platform Core Facility EM and Histology and the DRESDEN-concept Genome Center at the Technical University Dresden, in Germany. Samples were processed according to manufacturer protocols (10x protocols, CG000580, CG00582 and CG00584). Briefly, 4 pm cross sections were collected, floated in a 37°C water bath, and adhered to Xenium slides (10x Genomics, PN 1000460). Samples were deparaffinized in 2x Xylene and rehydrated in a descending series from 100% ethanol to MilliQ water. After inserting slides into xenium cassettes, samples were decross-linked and incubated overnight (22 hours) with padlock probes, followed by a post-hybridization wash. Subsequently, padlock probes were ligated, followed by rolling circle amplification, autofluorescence quenching, and nuclear staining (DAPI). Slides were loaded on a 10X XeniumAnalyzer (software v. 1.6.1.0), for region selection, with each region corresponding to one entire nerve bundle.

[0101] Xenium data analysis

[0102] Xenium data was loaded into Seurat. The inventors removed cells with less than 10 molecules per cell. The inventors performed normalization with SCTransform (default parameters) and computed PCA for each sample. To match our snRNA-seq data to our Xenium data, each cell was classified with the FlndTransferAnchors and TransferData function based on the previously annotated snRNA-seq data of the inventors (downsampled to 1000 cells per cluster). Cells with a prediction score below 0.3 were labeled as unknown. The 24 main clusters were then aggregated into 8 larger groups (SC, endoC, periC, epiC, VSMC, PC, EC, IC). For manual quantification of endo- vs. epineurial transcripts the Xenium images were loaded into Xenium Explorer 1.3.0 (10x Genomics) and endoneurial areas were manually outlined using the selection tool while using visualization of Schwann cell marker transcripts (EGR2, NGFR, SOX10, SWOB) to reliably identify such endoneurial areas. The area of each image visibly unoccupied by nerve- associated tissue was selected as epineurium. The area (in pm2), the number of total segmented cells, and the number of immune cell-associated transcripts (leukocytes: PTPRC / CD45, B cells: MS4A 1 / CD20, T cells: CD3E) was quantified and recorded in each selection area. The density of the respective transcript per total endorial and epineurial area were calculated for each sample. Selected transcripts were visualized in Xenium Explorer 1.3.0 (10x Genomics). Images were integrated with corresponding H&E staining to visualize the spatial orientation.

[0103] Example 1 : Comprehensive species-specific single nuclei transcriptional atlas of human sensory nerves.

[0104] The inventors here first created a large single cell transcriptomics atlas of human sural nerves of 37 donors (see Fig. 1A). Sural nerves were collected in three centers from 33 PNP patients and four controls (see Fig. 5; see Table 7). Control samples (CTRL) were residual material of surgical sural nerve autografts (‘interpositions’) from patients with traumatic nerve injuries, but unaffected by PNPs. The inventors established protocols to extract nuclei of sufficient number and quality and performed single nuclei RNA-sequencing (snRNA-seq) of all samples (see methods-section described above, see Fig. 5B). After analytical removal of doublets and low quality nuclei and batch correction, this resulted in 365,708 total high-quality nuclei (see Fig. 5C- 5D). The inventors then clustered the nuclei (henceforth termed ‘cells’ for simplicity) and annotated the clusters by using both predefined marker genes (see Fig. 6A) and automatic annotation based on rodent data9’10’14(see Fig. 6B). The inventors identified endoneurial cell types including Schwann cells (SWOB, SOXW), of both myelinating (mySC; MPZ, MBP, PRX) and non-myelinating (nmSC; NCAM1, L1CAM, CDH2) type and endoneurial fibroblast cells (endoC; S0X9, PLXDC1, ABCA9) (see Fig. 1B, see Fig. 6A). Schwann cells expressing featuresof damage (damageSC; EGR1, FOS, JUN) and of repair15(repairSC; NGFR, ATF3, GDNF, RUNX2) were also detected, which were not present in the reference rodent single cell data9 10. Moreover, the inventors found epineurial (epiC; CCBE1, COMP) and perineurial fibroblast cells (periCi -3; SLC2A 7 / GLUT1 , KRT19, CLDN1) (see Fig. 1 B and Fig. 6A). This perineurial cell heterogeneity was greater than previously described in rodent single cell datasets8-10’12. The periC3 cluster expressed genes marker-associated with extracellular matrix and collagen fibril organization, which distinguished it from the other two pericyte clusters (see Fig. 6B). Vascular cells included vascular smooth muscle cells (VSMC: ACTA2, CARMN), pericytes (PC1-2; PDGRFB, RGS5) and endothelial cells (EC; EGFL7, PECAM1) (see Fig. 6A). Based on known markers (see Fig. 6A) and published reference scRNA-seq data14(see Fig. 6C) endothelial cells (EC) separated into lymphatic endothelial cells (LEC; PR0X1, LYVE1, FLT4) and a venous (ven_EC: PLVAP, ACKR1) to capillary (capEC: ABCG2, MFSD2A) to arterial (artEC: SEMA3G, HEY1, GJA5) continuum. The inventors identified a cluster of venous / capillary EC (ven_cap_EC2), which expressed blood-nerve barrier (BNB) markers ABCB1 and SLC1A 116, blood-brain barrier marker MFSD2A17, and tight junction transcript GJA 1 and therefore likely represented EC of the BNB (see Fig. 1B and Fig. 6A). Nerve-associated leukocytes were mainly of myeloid lineage (Macrol : MS4A7, CLEC10A; Macro2: MS4A7, CX3CR1' Granulo: S100A8, S100A9; Mast: CPA3, MS4A2) outnumbering T / NK cells (T / NK: CD3E, CD8A, NCR1, NKG7, GZMA) and B cells (B: CD79A, IGHM) (see Fig. 1 B and Fig. 6A). The inventors thus created a transcriptional cellular atlas of human peripheral nerves replicating and extending cell populations identified in single cell rodent studies8-13.

[0105] The inventors systematically compared cell type markers in humans with their published rodent counterparts8-10. As expected, the majority of cell marker transcripts were shared between species. However, several genes expressed in mySC (MLIP), nmSC (GRIK3, PRIMA 1), and periC (CXCL14) have not been described in the literature. When the inventors re-analyzed published rodent datasets8-10, they could detect Mlip in mySC in two of those9 10, although they were not mentioned in the text (see Fig. 1 C). In contrast, GRIK3, PRIMA 1 and CXCL14, were not detectable at relevant levels (see Fig. 6D), which indicates that they could be specific for humans. The inventors thus identified undescribed and species-specific transcripts in perineurial and Schwann cells in humans.

[0106] Example 2: Spatially dissecting the human nerve ultrastructure confirms novel transcriptional cell type markers.

[0107] For morphological confirmation and characterization, the inventors employed an in situ amplification-based spatial transcriptomics approach (Xenium; ‘spatial-seq’) to localize predefined transcripts at subcellular-resolution18. The inventors of the present application designed a custom panel of 99 RNA transcripts combining known (e.g., PRX) and novel (e.g., GRIK3) markers of nerve-associated cell types. Notably, highly transcribed genes such as myelin protein-encoding genes (e.g. MPZ) had to be excluded to prevent optical overcrowding (see methods- section described above). The inventors then visualized these RNA targets in a formalin-fixed paraffin-embedded (FFPE) cross-section of a human sural nerve graft (CTRL) with optimally maintained morphology.

[0108] First, the inventors analyzed the overall cellular organization. By integrating spatial and snRNA-seq, each cell in spatial-seq was matched to a snRNA-seq cluster, which were aggregated into 8 larger groups (SC, endoC, periC, epiC, VSMC, PC, EC, IC), and then visualized spatially (see Fig. 1 D). As expected, each nerve consisted of multiple bundles (fascicles) formed by endoneurial cells (SC, endoC) and surrounded by a clearly demarcated layer of perineurial cells (periC). Vascular cell types (EC, VSMC, PC) formed vessels of either arterial, capillary, or venous morphology with EC forming the thin innermost layer (see Fig. 1 D). Epineurial cells (epiC) were located in between fascicles and were interspersed with immune cells that were rare in the endoneurial areas (see Fig. 1D). The inventors thus confirmed the snRNA-seq cluster annotation using spatial features. This also provides a histological annotation of peripheral nerve cells in an automated manner.

[0109] Second, when comparing periC and epiC transcripts identified in the snRNA-seq clusters, the previously undescribed perineurial markers (CXCL14) indeed co-localized with known perineurial fibroblast transcripts (CLDN1, SLC2A1, KRT19) and outlined the perineurium (see Fig. 1 E). Among vascular cells, as predicted from snRNA-seq data, transcripts associated with epiEC (PLVAP) were located in vessels in the epineurium (see Fig. 1F). In contrast, the BNB markers ABCB1 and SLC1A1, expressed by the ven_capEC2 cluster in snRNA-seq, were enriched in endoneurial vessels together with known EC markers (PECAM1, EGFL7) -'i9(see Fig. 1F). This supports the interpretation of the inventors that the ven_capEC2 cluster represents cells of the BNB. Other vascular cell markers (e.g., SEMA3G in artEC) were preferentially associated with their respective vessel type (see Fig. 7A).

[0110] The inventors then focussed on the human Schwann cell (SC) transcriptome. Known pan-SC markers (SOXW, SWOB, EGR2Y'iwere abundantly expressed across the endoneurium (see Fig. 7B). Known transcriptional markers of mySC (PRX, CDH1, SLC36A2)9 nco-localized with transcripts newly identified in the mySC cluster (MLIP) (see Fig. 1G) and known transcriptional markers of nmSC (L1CAM, CDH2, CHLiy2-''9-20co-localized with novel transcripts of the nmSC cluster (PRIMA 1, GRIK3) (see Fig. 1G). The inventors thus spatially validated known and novel transcriptional markers of perineurial fibroblasts, the BNB, and human myelinating and non-myelinating Schwann cells.

[0111] Example 3: Human peripheral nerves contain complex immune cells and replicate rodent macrophage heterogeneity.

[0112] The inventors next sought to deeply characterize human nerve-associated immune cells. To capture the immune cell (IC) heterogeneity, the inventors sub-clustered all leukocyte nuclei (n= 18,436) at high resolution (see Fig. 2A). Clusters representing megakaryocytes / platelets (PF4, GP9, PPBP) and red blood cells (HBA, HBB) were undetected arguing against relevant blood contamination (see Fig. 2A). The majority of IC clusters (79% of all IC nuclei) were of myeloid lineage with mainly macrophage annotation (Macro1-18: LYZ, CD14, MRC1, CD163, MS4A7) (see Fig. 2A and Fig. 8A, see Table 2). The transcriptional phenotype and ontogeny of nerve- associated macrophages in rodents differ depending on their endo- vs. epineurial location8’21 22. When directly testing this in humans, macrophage clusters (MS4A7) exhibited a gradient of expression ranging from markers with known endoneurial (CX3CR1, TREM2) to epineurial (LYVE1, F0LR2, TIMD4) macrophage location in rodents (see Fig. 2B). When analyzing expression of a representative and tentatively endoneurial macrophage cluster (Macro18) its gene signature was enriched in lipid metabolism-related pathways (see Fig. 2C). The inventors next analyzed markers of epi- (F0LR2) vs. endoneurial (CX3CR1) macrophages ( / WS4A7)21in spatial transcriptomics. CX3CR1 was preferentially detected endoneurial, although it was also present within epineurial vessels, and FOLR2 was mostly located epineurial (see Fig. 2D). The inventors thus replicated the location-specificity of rodent nerve-associated macrophage phenotypes in humans. The inventors additionally identified smaller myeloid clusters that could be assigned to classical dendritic cells (DC) type 1 (cDC1 : CLEC9A, XCR1, BATF3), plasmacytoid DC (pDC: CLEC4C, IRF8), and classical DC type 2 (cDC2: FCER1A, CD1C, CLEC10A) (see Fig. 2A and Fig. 8A). cDC1 , but not cDC2, have been previously described in rodent nerves13.The less abundant T / NK cell clusters formed a continuum ranging from CD4 T cell phenotypes, including naive CD4 (JL7R, CCR7, MAL, SELL), Treg (F0XP3, IL2RA), MAIT (KLRB1, CXCR6) to CD8 T cells CD8A, CCL5) and NK cells (FCGR3A, NKG7, NCRT) (see Fig. 2A and Fig. 8A). Smaller clusters of nuclei represented naive B cells CD19, MS4A 1 / CD20, CD79B,) and plasma cells (JCHAIN, SDC7 / CD138) (see Fig. 2A and Fig. 8A) with mainly IGHG1, IGHG3, IGHG4, and IGHA 1 heavy chain gene expression (see Fig. 2E). Granulocytes (S100A8, S100A9) and mast cells (CPA3, TPSAB1, MS4A2) (see Fig. 2A and Fig. 8A) were also detected. The inventors thus delineated known and unknown nerve-associated leukocytes in humans in unprecedented detail.

[0113] Example 4: Diverse nerve-associated cell types respond to polyneuropathy.

[0114] Having characterized human nerve-associated cells in health, the inventors next aimed to understand how PNP affected the transcriptome and the cellular organization of peripheral nerves. The inventors found that most cell types were present in all human donors, although some patients showed a grossly skewed composition (see Fig. 1E). When comparing PNP and CTRL, the most apparent change in PNP was a loss of mySC and the occurrence of damageSC and repairSC (see Fig. 3A), known to be induced by nerve damage in rodents1523-25. Additionally, PNP induced an increase of leukocyte clusters; especially Macrol and Granulo (see Fig. 3A). Among IC subclusters, an increase of the endoneurial lipid-associated Macrol 8 subcluster was the most abundant change in PNP patients (see Fig. 3A). Overall, several cell clusters normally predictedto be located within the endoneurium (mySC, endoC) were less abundant, while periC3 increased in PNP (see Fig. 3A). This indicated loss of cells located within the endoneurium as a shared feature of human PNPs.

[0115] When testing which individual single nuclei were associated with disease, the PNP status positively correlated with nuclei in the leukocyte and repairSC clusters and negatively correlated with the mySC cluster after correcting for confounders (see methods-section above) (see Fig. 3B, left panel). This means that mySC are less likely to be present in PNP patients on a per nucleus level. In addition, the periC3 cluster positively correlated with PNP disease status (see Fig. 3B, left panel). These clusters similarly correlated with clinical disease severity as quantified by the clinical INCAT score (see Fig. 3B, middle panel) and myelin thickness (inversely measured by g- ratio) (see Fig. 3B, right panel). PNPs thus induced strong compositional changes that also reflected disease severity and affected perineurial cell types.

[0116] The inventors then characterized which and how nerve cells transcriptionally responded to disease. The number of differentially expressed genes (DEG) was highest in vascular clusters (ven_capEC1 , PC2, artEC) followed by endoneurial fibroblasts (endoC) (see Fig. 3C). Using a clustering-independent cellular neighborhood-based approach, the inventors accordingly found that cellular neighborhoods with the highest number of DEG were mostly located in vascular clusters (PC2, ven_capEC1 , artEC) and in the Macro2 cluster (see Fig. 3D). Next, the inventors more specifically tested how PNP influenced gene expression of selected nerve-associated cells (mySC, nmSC, repairSC, PC2) (see Fig. 3E). The inventors found that many of the DEG in the mySC cluster were associated with fibrotic tissue remodeling (e.g., DON, TNXB), extracellular matrix formation (e.g., COL1A1, COL15A 1), and immune regulation (e.g., CD53, IL4R, CD74). The nmSC cluster exhibited transcripts associated with immunosuppression and -modulation (IL10RA, IL13RA1, CSF2RA, TGFBI), and the repairSC cluster expressed genes for nerve repair and neuronal differentiation (GALR1, TMEM47) (see Fig. 3E; see Table 3). GO term enrichment analysis showed that upregulated genes were associated with cell differentiation in the mySC cluster and with cell migration in the nmSC cluster (see Fig 8B). In the PC2 cell cluster (pericytes), DEG indicated upregulation of molecules related to vessel maintenance (MFAP5) and cell adhesion (NLGN4Y, PCDH11Y), while genes indicating an interferon-induced immune response (e.g., IFIT3, OASL, MX1) were downregulated (see Fig. 3E). Consequently, downregulated genes were enriched for antiviral response GO terms (see Fig. 8B). In summary, PNPs induced transcriptional changes in Schwann cells but also widely outside of the endoneurium, especially in non-endoneurial vascular cells. This suggests that PNPs widely affect the cellular micro-milieu of peripheral nerves and may thus constitute ‘pan-nerve diseases’.

[0117] Example 5: PNP subtypes preferentially affect different cellular compartments of peripheral nerves.

[0118] The etiology of PNP is complex, multiple underlying conditions can lead to PNPs, andthe heterogeneity of PNP mechanisms is poorly defined. The inventors therefore next sought to identify cellular or transcriptional patterns specific to individual PNP subtypes. The inventors first classified the available PNP patients by integrating all diagnostic information (see Table 7) into seven subtypes of PNPs: vasculitic (VN, n = 5), chronic inflammatory demyelinating (CIDP, n = 9), chronic idiopathic axonal (CIAP, n = 11), cancer-associated paraproteinemic (PPN, n = 2), diabetic (DPN, n = 2), other inflammatory (OIN, n = 2), and other non-inflammatory (ONIN, n = 2) (see Fig. 4A). The inventors characterized nerves histologically (see methods-section; see Fig. 9A). Myelin thickness (inversely quantified by the g-ratio) (see Fig. 9B and Fig. 9C) and the average number of intactly myelinated axons (see Fig. 9D) decreased in PNPs, while axon diameter was less affected (see Fig. 9E). Histological findings in PNP subtypes were in accordance with expectations and also correlated with electrophysiological measures (see Fig. 9F).

[0119] When analyzing the cellular composition determined by snRNA-seq, CTRL samples had the highest relative proportions of the mySC cluster and a low proportion of the repairSC cluster (see Fig. 4A). Across conditions, the number of intactly myelinated axons positively correlated with cluster proportions of mySC and negatively correlated with repairSC (see Fig. 9G). The inventors next compared cluster proportions between groups with at least four samples per group (VN, CIDP, CIAP, CTRL), while disregarding other diagnostic groups due to their small size. In direct comparison with controls, a relative increase of damageSC and repairSC, and loss of mySC was detected in VN, CIDP, CIAP (see Fig. 4B). A gain of the periC3 cluster (perineurial cells) and the venEC cluster was shared between VN and CIDP. An increase of the VSMC cluster was found in both VN and CIAP. Loss of the LEC cluster was specific to VN (see Fig. 4B). The inventors next compared immune cell sub-cluster proportions between the groups. The lipid-associated Macro18 cluster was increased in all three PNP groups compared to CTRL (see Fig. 4C). In contrast, only VN showed an expansion of the plasma cluster. Each PNP subtype thus induced a unique pattern of compositional cellular changes in peripheral nerves.

[0120] The inventors then analyzed spatial transcriptomics of sural nerves from VN, CIDP, and CIAP patients in comparison to control (n = 2 per group). Predicted repairSC were increased in the endoneurium in VN compared to CTRL (see Fig. 4D). Moreover, mySC were considerably less abundant in VN compared to CTRL (see Fig. 4D). Predicted mySC were also less abundant in CIDP and CIAP compared to CTRL although to a lesser extent than in VN (see Fig. 9H to Fig. 9I). Quantification of the density of selected leukocyte-associated transcripts (PTPRC / CD45, MS4A 1 / CD20, CD3E) showed that endo- and epineurial B and T cells increased in VN, CIDP and in CIAP compared to CTRL (see Fig. 4F and Fig. 9J). Predicted T_NK cells were more abundant in VN and, to a lesser extent, in CIDP and CIAP than in CTRL (see Fig. 9L and Fig. 9M). The inventors thus spatially validated loss of mySC and gain of repairSC and leukocytes in PNPs, especially in VN, as observed in snRNA-seq.

[0121] The inventors next aimed to spatially validate the increase of the periC3 (perineurial cell)cluster in snRNA-seq. Plotting known (e.g. CLDN1, SLC2A 1, KRT19) and novel (e.g. CXCL14) perineurial markers, the inventors found that the perineurium consisted of a single cell layer in CTRL samples, but formed 3-5 cell layers in PNP samples which was most pronounced in CIDP and appeared regionalized (see Fig. 4E). The perineurium was also structurally less well defined with a more dispersed and heterogeneous appearance in PNP samples (see Fig. 4E). Perineurial hyperplasia and fibrotic dispersion thus occur across PNP subtypes, but preferentially in CIDP.

[0122] Next, the inventors tested for subtype-specific transcriptional alterations using a clustering independent approach. In VN, most DEG were located in neighborhoods in the PC2 (pericyte) cluster (see Fig. 4G) in accordance with vascular-focused inflammation. In CIDP, DEG were also identified in neighborhoods of vascular clusters (ven_capEC1 , artEC, PC2) and the Macro2 cluster. In CIAP, DEG were distributed in a more widespread pattern and also involved the nmSC and the periC2 (perineurial cell) clusters (see Fig. 4G). In summary, the inventors defined loss of mySC and non-endoneurial transcriptional alterations as shared across PNPs. Each PNP subtype additionally induced unique compositional and transcriptional responses, for example immune-associated changes in VN and perineurial dispersion in CIDP.

[0123] Example 6: Conclusion of results

[0124] The inventors aimed to exploit the full potential of human nerve biopsies using state-of- the-art techniques. Here, the inventors present a single nuclei transcriptomics atlas of human sural nerves including 365,708 nuclei of 33 PNP patients and four controls integrated with spatial transcriptomics at subcellular resolution. Capitalizing on this first large-scale technology application to human peripheral nerves and a dataset 10-fold larger than previous rodent studies8-13the inventors dissected the cellular heterogeneity of human peripheral nerves in detail and found unexpected heterogeneity in perineurial cells. Additionally, the inventors disentangled nerve-associated immune cells and found a multitude of macrophage clusters with an endoneurial to epineurial phenotypic gradient and previously undescribed eDC type 2. The inventors also discovered and validated previously undescribed transcriptional markers of perineurial fibroblasts (CXCL14) and myelinating (MLIP) and non-myelinating SC (GRIK3, PRIMA 1), which were partially human-specific. In human PNP, the inventors identified a loss of myelinating SC, the occurrence of damage and repair SC, and a gain of a lipid-associated macrophage population, as shared features of PNPs. In general, PNP exhibited a loss of cells located in the endoneurium. However, transcriptional changes also affected multiple non-endoneurial cell populations. Thus, PNPs may constitute ‘pan-nerve diseases’. Notably, multiple PNPs, especially CIDP, showed perineurial hyperplasia and dispersion. Using single cell transcriptomics to classify PNP patients, the inventors defined patient clusters with specific clinical, histological and cellular phenotypes beyond existing clinical classification.

[0125] Among the novel Schwann cell markers, the MLIP gene encodes a muscular laminin- interacting protein, which is required to maintain muscular integrity27and could serve similarfunctions in SC. The PRIM A 1 -encoded protein organizes esterases at the neuromuscular junction in terminal SC28, but has not been described in nmSC. GRIK3 encodes a glutamate ionotropic receptor29and variants in GRIK3 have been associated with CMT type C30. These genes may thus be involved in structural organization of SC in human peripheral nerves. The inventors additionally identified novel markers for perineurial cells. Human CXCL14 (synonymously BRAK) is constitutively expressed by epithelial tissues and regulates microglial development in the brain31and neurovascular patterning in the eye32. It is thus conceivable that the perineurium, regarded as an epithelial barrier33, expresses CXCL14. The species specificity of some of the markers (GRIK3, PRIMA1, CXCL14) could be due to i) size or phylogeny of the mammal, ii) diseased samples, which dominated our dataset, or iii) the purely sensory sural nerve analyzed in humans. The inventors find differences unlikely to be caused by discrepancies between single nuclei vs. single cell methods because one rodent study analyzed nuclei9and markers were also not identified.

[0126] As a common denominator between the different entities of neuropathies, the inventors observed a relative loss of cells located within the endoneurium. Whether this reflects true cellular loss (i.e. shrinkage) of the endoneurium or cellular transdifferentiation or epi- / perineurial cell proliferation remains to be determined. It is conceivable that loss of myelin-capable axons could be associated with a reduction of SC and other cells located within the endoneurium. Alternatively, a quantitative increase of perineurial cells (e.g. proliferation of the periC3 cluster) could reduce the relative ratio of peri- to endoneurial cells and thereby simulate a relative loss of cells with endoneurial location. The perineurium is a lamellated structure made up of concentric cell layers bordered on each side by a basement membrane34. A linear relationship has been described between fascicle diameter and perineurium thickness35, i.e. larger fascicles have a thicker perineurium. Additionally, distal sections have a greater perineurium thickness than proximal segments35. Thickening of the perineurial basement membrane is a characteristic of diabetic PNP36. Focal or generalized perineurial thickening has also been reported in the rare condition ‘perineuritis’ that can occur either primary or secondary to inflammatory diseases37. Perineurial accumulation of immune cells has also been demonstrated in the early stages of inflammation in a rodent model of autoimmune neuropathy38. The transcription factor F0XD1, which was expressed in PNP samples, fate-labels myofibroblast precursors, which contribute to rodent perineurial fibroblasts39and might play a role in tissue fibrosis40. Overall, perineurial thickening per se has not been studied systematically in PNPs and - to our knowledge - the perineurial distortion described here has not been reported previously. This could reflect a maladaptive response of the nerve microstructure to chronic damage in diverse PNPs.

[0127] Diagnosing polyneuropathies can be challenging due to the diverse range of possible causes and the many cases, in which the etiology remains unclear (20-30 %)5. Even histology of the sural nerves often fails to establish the etiology of PNP7. Here, the inventors used data-driven approaches to classify patients solely based on snRNA-seq into pathophysiological groups withdistinct clinical phenotypes. Therefore, the inventors hypothesize that snRNA-seq could represent a PNP classification tool towards personalized single cell-based neurology / neuro-pathology41-43. This underlines the known heterogeneity in the course and treatment response of neuropathies and suggests that the cellular architecture of the nerve should be taken into account in future studies.REFERENCES: Hanewinckel, R. et al. Prevalence of polyneuropathy in the general middle-aged and elderly population. Neurology87, 1892-1898 (2016). Hoffman, E. M. et al. Impairments and comorbidities of polyneuropathy revealed by population-based analyses. Neurology 84, 1644-1651 (2015). Jennum, P., Ibsen, R. & Kjellberg, J. Health, social and economic consequences of polyneuropathy: a controlled national study evaluating societal effects on patients and their partners. Eur. Neurol. 73, 81-88 (2015). Visser, N. A., Notermans, N. C., Linssen, R. S. N., van den Berg, L. H. & Vrancken, A. F. J. E. Incidence of polyneuropathy in Utrecht, the Netherlands. Neurology 84, 259-264 (2015). Hanewinckel, R., van Oijen, M., Ikram, M. A. & van Doorn, P. A. The epidemiology and risk factors of chronic polyneuropathy. Eur. J. Epidemiol. 31 , 5-20 (2016). Verghese, J., Bieri, P. L., Gellido, C., Schaumburg, H. H. & Herskovitz, S. Peripheral neuropathy in young-old and old-old patients. Muscle Nerve 24, 1476-1481 (2001). Prada, V. et al. Diagnostic value of sural nerve biopsy: retrospective analysis of clinical cases from 1981 to 2017. Front. Neurol. 10, 1218 (2019). Wolbert, J. et al. Redefining the heterogeneity of peripheral nerve cells in health and autoimmunity. Proc Natl Acad Sci USA 117, 9466-9476 (2020). Yim, A. K. Y. et al. Disentangling glial diversity in peripheral nerves at single-nuclei resolution. Nat. Neurosci. 25, 238-251 (2022). Gerber, D. et al. Transcriptional profiling of mouse peripheral nerves to the single-cell level to build a sciatic nerve ATIas (SNAT). eLife 10, (2021). Carr, M. J. et al. Mesenchymal precursor cells in adult nerves contribute to mammalian tissue repair and regeneration. Cell Stem Cell 24, 240-256. e9 (2019). Toma, J. S. et al. Peripheral Nerve Single-Cell Analysis Identifies Mesenchymal Ligands that Promote Axonal Growth. eNeuro 7, (2020). Lovatt, D. et al. scRNA-seq generates a molecular map of emerging cell subtypes after sciatic nerve injury in rats. Commun. Biol. 5, 1105 (2022). Garcia, F. J. et al. Single-cell dissection of the human brain vasculature. Nature 603, 893- 899 (2022). Arthur-Farraj, P. J. et al. Changes in the Coding and Non-coding Transcriptome and DNA Methylome that Define the Schwann Cell Repair Phenotype after Nerve Injury. Cell Rep. 20, 2719-2734 (2017). Palladino, S. P. et al. The Human Blood-Nerve Barrier Transcriptome. Sci. Rep. 7, 17477 (2017). Wood, C. A. P. et al. Structure and mechanism of blood-brain-barrier lipid transporter MFSD2A. Nature 596, 444-448 (2021).Janesick, A. et al. High resolution mapping of the tumor microenvironment using integrated single-cell, spatial and in situ analysis. Nat. Commun. 14, 8353 (2023). Chen, B., Banton, M. C., Singh, L., Parkinson, D. B. & Dun, X.-P. Single cell transcriptome data analysis defines the heterogeneity of peripheral nerve cells in homeostasis and regeneration. Front. Cell. Neurosci. 15, 624826 (2021). Kalinski, A. L. et al. Analysis of the immune response to sciatic nerve injury identifies efferocytosis as a key mechanism of nerve debridement. eLife 9, (2020). Ydens, E. et al. Profiling peripheral nerve macrophages reveals two macrophage subsets with distinct localization, transcriptome and response to injury. Nat. Neurosci. 23, 676-689 (2020). Wang, P. L. et al. Peripheral nerve resident macrophages share tissue-specific programming and features of activated microglia. Nat. Commun. 11, 2552 (2020). Gomez-Sanchez, J. A. et al. After nerve injury, lineage tracing shows that myelin and remak schwann cells elongate extensively and branch to form repair schwann cells, which shorten radically on remyelination. J. Neurosci. 37, 9086-9099 (2017). Norrmen, C. et al. mTORCI Is Transiently Reactivated in Injured Nerves to Promote c-Jun Elevation and Schwann Cell Dedifferentiation. J. Neurosci. 38, 4811-4828 (2018). Szepanowski, F. et al. LPA1 signaling drives Schwann cell dedifferentiation in experimental autoimmune neuritis. J. Neuroinflammation 18, 293 (2021). Boyeau, P. et al. Deep generative modeling for quantifying sample-level heterogeneity in single-cell omics. BioRxiv (2022) doi:10.1101 / 2022.10.04.510898. Cattin, M.-E. et al. Expression of murine muscle-enriched A-type lamin-interacting protein (MLIP) is regulated by tissue-specific alternative transcription start sites. J. Biol. Chem. 293, 19761-19770 (2018). Petrov, K. A. et al. Schwann cells sense and control acetylcholine spillover at the neuromuscular junction by a 7 nicotinic receptors and butyrylcholinesterase. J. Neurosci. 34, 11870-11883 (2014). Strutz, N. et al. Identification of domains and amino acids involved in GLuR7 ion channel function. J. Neurosci. 21, 401-411 (2001). Vedi, M. et al. 2022 updates to the Rat Genome Database: a Findable, Accessible, Interoperable, and Reusable (FAIR) resource. Genetics 224, (2023). Li, Z., Li, Y. & Jiao, J. Neural progenitor cells mediated by H2A.Z.2 regulate microglial development via Cxcl14 in the embryonic brain. Proc Natl Acad Sci USA 116, 24122-24132 (2019). Ojeda, A. F., Munjaal, R. P. & Lwigale, P. Y. Knockdown of CXCL14 disrupts neurovascular patterning during ocular development. Dev. Biol. 423, 77-91 (2017). Ubogu, E. E. Biology of the human blood-nerve barrier in health and disease. Exp. Neurol. 328, 113272 (2020).Hill, R. E. & Williams, R. E. A quantitative analysis of perineurial cell basement membrane collagen IV, laminin and fibronectin in diabetic and non-diabetic human sural nerve. J. Anat. 201 , 185-192 (2002). Andreis, F. R. et al. On the relationship between fascicle diameter and perineurium thickness in the ulnar nerve of pigs, in 2023 11th International IEEE / EMBS Conference on Neural Engineering (NER) 1-4 (IEEE, 2023). doi:10.1109 / NER52421.2023.10123782. Hill, R. E. & Williams, P. E. Perineurial cell basement membrane thickening and myelinated nerve fibre loss in diabetic and nondiabetic peripheral nerve. J. Neurol. Sci. 217, 157-163 (2004). Collins, M. P, Hadden, R. D. M. & Shahnoor, N. Primary perineuritis, a rare but treatable neuropathy: Review of perineurial anatomy, clinicopathological features, and differential diagnosis. Muscle Nerve 68, 696-713 (2023). Mausberg, A. K. et al. Trapped in the epineurium: early entry into the endoneurium is restricted to neuritogenic T cells in experimental autoimmune neuritis. J. Neuroinflammation 15, 217 (2018). Nikoloudaki, G. & Hamilton, D. W. Assessing the fate and contribution of Foxdl -expressing embryonic precursors and their progeny in palatal development, homeostasis and excisional repair. Sci. Rep. 14, 4969 (2024). Ajay, A. K. et al. Deletion of STAT3 from Foxdl cell population protects mice from kidney fibrosis by inhibiting pericytes trans-differentiation and migration. Cell Rep. 38, 110473 (2022). Lim, J. et al. Transitioning single-cell genomics into the clinic. Nat. Rev. Genet. 24, 573-584 (2023). Heming, M., Borsch, A.-L., Wiendl, H. & Meyer Zu Hdrste, G. High-dimensional investigation of the cerebrospinal fluid to explore and monitor CNS immune responses. Genome Med.14, 94 (2022). Meyer Zu Hdrste, G., Gross, C. C., Klotz, L., Schwab, N. & Wiendl, H. Next-Generation Neuroimmunology: New Technologies to Understand Central Nervous System Autoimmunity. Trends Immunol. 41 , 341-354 (2020). Clark, I. C. et al. Identification of astrocyte regulators by nucleic acid cytometry. Nature 614, 326-333 (2023). Van den Bergh, P. Y. K. et al. European Academy of Neurology / Peripheral Nerve Society guideline on diagnosis and treatment of chronic inflammatory demyelinating polyradiculoneuropathy: Report of a joint Task Force-Second revision. Eur. J. Neurol. 28, 3556-3583 (2021). Pool, A.-H., Poldsam, H., Chen, S., Thomson, M. & Oka, Y. Recovery of missing single-cell RNA-sequencing data with optimized transcriptomic references. Nat. Methods 20, 1506- 1515 (2023).Fleming, S. J. et al. Unsupervised removal of systematic background noise from dropletbased single-cell experiments using CellBender. Nat. Methods 20, 1323-1335 (2023). Hao, Y. et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat. Biotechnol. 42, 293-304 (2024). Germain, P.-L., Lun, A., Garcia Meixide, C., Macnair, W. & Robinson, M. D. Doublet identification in single-cell sequencing data using scDblFinder. FWOORes. 10, 979 (2021). Lopez, R., Regier, J., Cole, M. B., Jordan, M. I. & Yosef, N. Deep generative modeling for single-cell transcriptomics. Nat. Methods 15, 1053-1058 (2018). Kuleshov, M. V. et al. Enrichr: a comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 44, W90-7 (2016). Squair, J. W. et al. Confronting false discoveries in single-cell differential expression. Nat. Common. 12, 5692 (2021). Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139-140 (2010). Missarova, A., Dann, E., Rosen, L., Satija, R. & Marioni, J. Sensitive cluster-free differential expression testing. BioRxiv (2023) doi:10.1101 / 2023.03.08.531744. Phipson, B. et al. propeller: testing for differences in cell type proportions in single cell data. Bioinformatics 38, 4720-4726 (2022). Reshef, Y. A. et al. Co-varying neighborhood analysis identifies cell populations associated with phenotypes of interest from single-cell transcriptomics. Nat. Biotechnol. 40, 355-363 (2022). Husson, F, Josse, J. & Le, S. FactoMineR : An R package for multivariate analysis. J. Stat. Softw. 25, (2008). Cakir, B. et al. Comparison of visualization tools for single-cell RNAseq data. NAR Genom. Bioinform. 2, lqaa052 (2020).

Claims

CLAIMS1 . An in vitro method for diagnosing polyneuropathy (PNP) in a subject, comprising(a) determining by single cell RNA sequencing(i) a proportion of one or more cell type cluster characterized by one or more nerve-associated marker gene(s) and / or by one or more immunological marker gene(s) in a sample obtained from the subject;(ii) an expression level of one or more nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i);(b) diagnosing said subject as suffering from PNP if one or more of the following are fulfilled(i) the proportion of one or more cell type cluster of (a)(i) characterized by the one or more nerve-associated marker gene(s) and / or by the one or more immunological marker gene(s) is / are altered compared to a control sample;(ii) the expression level(s) of (a)(ii) of one or more of the nerve-associated differentiation gene(s) and / or of one or more of the immunological differentiation gene(s) within the one or more cell type cluster of (i) is altered compared to a control sample; wherein the altered proportion of one or more cell type cluster of (b)(i) and the altered expression levels of (b)(ii) are indicative for PNP.

2. A method of stratifying a subject with polyneuropathy (PNP), comprising(a) determining by single cell RNA sequencing(i) a proportion of one or more cell type cluster characterized by one or more nerve-associated marker gene(s) and / or one or more immunological marker gene(s) in a sample obtained from the subject;(ii) an expression level of one or more nerve-associated differentiation gene(s) and / or of one or more immunological differentiation gene(s) within the one or more cell type cluster of (i);(b) stratifying said subject as suffering from a distinct PNP subtype if one or more of the following are fulfilled(i) the proportion of one or more cell type cluster of (a)(i) characterized by the one or more nerve-associated marker gene(s) and / or by the one or more immunological marker gene(s) is altered compared to a control sample;(ii) the expression level(s) of (a)(ii) of one or more gene(s) of the nerve- associated differentiation gene(s) and / or of one or more immunologicaldifferentiation gene(s) within the one or more cell type cluster of (i) is altered compared to a control sample; wherein the altered proportion of one or more cell type(s) cluster of (b)(i) and the altered expression levels of (b)(ii) are indicative for the distinct PNP subtype.

3. The method of claim 1 or 2, wherein the cell type cluster characterized by one or more nerve-associated marker gene(s) is selected from the group consisting of mySC, nmSC, repairSC, damageSC, Adipo, ArtEC, endoC, epiC, LEC, PC1 , PC2, periCi , periC2, perC3, ven_capEC1 , ven_capEC2, venEC, and VSMC.

4. The method of any one of claims 1 to 3, wherein the cell type cluster characterized by one or more immunological marker gene(s) is selected from the group consisting of CD4, Treg, MAIT, CD4_CD8, CD8, NK_CD8, NK, B, Plasma, Mast, Macrol , Macro2, Macro3, Macro4, Macro5, Macro6, Macro?, Macro8, Macro9, Macrol 0, Macrol 1 , Macrol 2, Macro13, Macro14, Macro15, Macro16, Macro17, Macro18, cDC2_1 , cDC2_2, Macro_cDC, Macro_Granulo, Granulol , Granulo2, and cDC-1_pDC.

5. The method of any one of claims 1 to 4, wherein the nerve-associated marker gene(s) is / are one or more of the gene(s) as depicted in Table 1.

6. The method of any one of claims 1 to 5, wherein the immunological marker gene(s) is / are one or more of the gene(s) as depicted in Table 2.

7. The method of any one of claims 1 to 6, wherein the nerve-associated differentiation gene(s) and the immunological differentiation gene(s) is / are one or more of the gene(s) as depicted in Tables 4 to 6.

8. The method of any one of claims 1 to 7, wherein PNP is characterized by a decreased proportion of the one or more cell type cluster characterized by one or more nerve- associated marker gene(s) mySC and / or endoC; and an increased proportion of the one or more cell type(s) cluster VSMC, damageSC, periC3, repairSC, venEC, Adipo, and LEC.

9. The method of any one of claims 1 to 8, wherein PNP is characterized by decreased proportion of cells type cluster characterized by one or more immunological associated marker gene(s) NK, MacrolO, Macro?, and cDC2_1 ; and an increased proportion of the one or more cell type cluster Macrol 8, Macro_cDC, Macro5, Macro_Granulo, Macro6, and Macrol .

10. The method of any one of claims 1 to 9, wherein PNP comprises(i) an increased expression level of one or more of the nerve-associated differentiation gene(s) of cell type(s) cluster ven_capEC1 , PC2, artEC, endoEC, nmSC, PC1 , mySC, periC2, periCi , ven_capEC2, epiC, Adipo, VSMC, repair SC, venEC, periC3, and LEC;(ii) an increased expression level of one or more of the immunological differentiation gene(s) of cell type cluster Macro2, Mast, and T_NK.

11. The method of any one of claims 1 to 10, wherein PNP comprises(i) in the cell type cluster mySC the increased expression level of the nerve- associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of the following: CD74, CD53, COL15A1 , DCN, TNXB, COL1 A1 , and IL4R; and / or(ii) in the cell type cluster nmSC the increased expression level of the nerve- associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of the following: CSF2RA, IL13RA, TGFBI, and IL10RA; and / or(iii) in the cell type cluster repairSC the increased expression level of the nerve- associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of TMEM47 and GALR1 ; and / or(iv) in the cell type cluster PC2 the increased expression level of the nerve-associated differentiation gene(s) and of immunological differentiation gene(s) of one or more of the following: PCDH11Y, MFAP5, and NLGN4Y.

12. The method of any one of claims 2 to 7, wherein PNP subtypes are characterized by a decreased proportion of the cell type cluster mySC, and an increased proportion of the cell type cluster repairSC and / or damageSC.

13. The method of any one of claims 2 to 7, and 12, wherein distinct PNP subtypes are characterized according to the following:(i) a first PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) damageSC, VSMC, periC3, repairSC, venC, Adipo, and ven_capEC2; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) ven_capEC1 , nmSC, arEC, epiC, periC2, PC1 , mySC, and LEC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Plasma, Macro 17,CD4_CD8, Macro_cDC, Macro_Granulo, Macro5; CD4, NK_CD8, and Treg; and / or(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) NK, Mast, Macro13; Macro14; cDC2_1 , and Macro2;(ii) a second PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) periC3, damageSC, venEC, Adipo, and repairSC; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) mySC; and / or(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Macro6, Macro5, Macro17, Macro_Granulo, Macro3, and Macro_cDC; and / or(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro2, cDCD2_1 , MacrolO, Granulol , Treg, NK_CD8, and Granulo2;(iii) a third PNP subtype comprises(a) an increased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) VSMC, Adipo, repairSC, LEC, and epiC; and / or(b) a decreased proportion of the one or more cell type cluster characterized by one or more of the nerve-associated marker gene(s) endoC, and mySC;(c) an increased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro18, Macro5, Macro_cDC, Macro6, Macro"! , and Macro_Granulo;(d) a decreased proportion of the one or more cell type cluster characterized by one or more of the immunological marker gene(s) Macro?, CD8, NK, and MacrolO.

14. The method of any one of claims 2 to 7, and 12 or 13, wherein distinct PNP subtypes are characterized according to the following:(i) a first PNP subtype comprises an increased expression level of the nerve-associated differentiation gene(s) of the PC2 cell type cluster;(ii) a second PNP subtype comprises(a) an increased expression level of one or more gene(s) of the nerve-associated differentiation gene(s) of the ven_capEC1 , artEC, and PC2 cell type cluster; and / or(b) an increased expression level of the immunological differentiation gene(s) of the Macro2 cell type cluster;(ii) a third PNP subtype comprises(a) an increased expression level of one or more gene(s) of the nerve-associated differentiation gene(s) of the nmSC, periC2, PC2, PC1 , ven_capEC1 , and artEC cell type cluster; and / or(b) an increased expression level of the immunological differentiation gene(s) of the Macro2 cell type cluster.

15. The method of any one of claims 1 to 14, wherein step (a) additionally comprises one or more of the following: spatial transcriptomics, histology, determining electrical conductivity, determining axon characteristics, such as axon diameter, axon integrity, myelin characteristics, neurological disability of the subject, bulk tissue transcriptomics, epigenetic analysis, protein quantification by techniques such as flow cytometry, mass spectrometry.

16. The method of any one of claims 1 to 15, wherein the sample is a neurological sample obtained from the subject, preferably wherein the neurological sample is a nerve biopsy, more preferably a sural nerve biopsy, a skin biopsy, or a biopsy from any other tissue containing peripheral nerves or liquids, such as blood, tear fluid, urine, cerebrospinal fluid, or cellular material comprising nerves, or meningeal biopsies, or cerebrospinal fluid.

17. A method for determining therapeutic treatment of PNP in a subject comprising conducting the method of any one of claims 1 to 16.

18. Use of the method of any one of claims 1 to 16 for determining a treatment regime for a subject suffering from PNP.

Images