Screening method of regeneration early conserved co-expression candidate gene set based on homologous gene mapping and application thereof

Abstract

The application discloses a screening method of a regeneration early conserved co-expression candidate gene set based on homologous gene mapping and application thereof, and belongs to the technical field of bioinformatics, single cell transcriptomics and regeneration biology. The application obtains the single cell transcriptome data of the tail fin regeneration early stage of Xenopus laevis and the DDC-induced mouse liver injury repair, and screens the specific up-regulated genes of the regeneration initiating cells RICs of the tail fin of Xenopus laevis and the second subpopulation LPLC_2 of mouse liver progenitor cell-like cells, respectively. c‑ Jun The regeneration early conserved co-expression candidate gene set including 1, 000 genes can be used for regeneration mechanism research and functional target screening.

Description

Technical Field

[0001] This invention relates to the fields of bioinformatics, single-cell transcriptomics, and regenerative biology, and more specifically to a method for screening a set of conserved co-expressed candidate genes in the early stages of regeneration based on homologous gene mapping and its application. Background Technology

[0002] The regenerative repair capacity after tissue injury varies significantly among different vertebrates. Lower vertebrates, such as amphibians, typically exhibit strong regenerative capabilities after damage to the tail fin, limbs, or other tissues. Mammals have relatively limited regenerative capacity, but retain strong damage repair and cellular remodeling capabilities in organs such as the liver. Comparing common molecular characteristics in the early stages of injury repair in different vertebrates and organs can help identify candidate regulatory nodes related to regeneration initiation, damage response, cell adhesion, extracellular matrix remodeling, and cell fate transition.

[0003] Single-cell transcriptome sequencing technology can resolve compositional and gene expression changes in different cell types or states during injury repair at single-cell resolution. Through single-cell data integration, dimensionality reduction clustering, cell type annotation, differential expression analysis, and functional enrichment analysis, key cell populations and their characteristic transcriptional programs that occur during injury repair can be identified.

[0004] However, cross-species comparative analysis of regeneration-related genes still faces technical challenges. Gene nomenclature rules, annotation completeness, and homology complexity vary across species. Existing conventional homology conversion tools primarily target direct homology between model species, failing to adequately adapt to the characteristics of the Xenopus allotetraploid genome. A large number of genes in the Xenopus genome simultaneously possess homologous copies of both L and S subgenomes. Directly comparing Xenopus gene symbols with mouse gene symbols can easily lead to problems such as homologous copy duplication counting, inconsistent gene symbols, misclassification of homology relationships, or omission of candidate genes. Currently, there is a lack of mature and universal analytical procedures for standardized mapping, deduplication, and comparability conversion between Xenopus L / S homologous copies and mammalian genes, thus affecting the accuracy and interpretability of cross-species candidate regulatory node screening.

[0005] In existing technologies, single-cell analysis of injury repair in a single species or organ is relatively common. However, there is still room for improvement in methods that unify the comparison of key cell populations in the early regeneration stage of lower vertebrates with those in the early regeneration stage of mammals. In particular, for the early regeneration model of the African clawed frog's caudal fin, how to handle the homologous copy problem of its L / S subgenome while preserving the characteristics of the RICs cell population, and how to establish a stable homology mapping pathway through intermediate species to further compare it with the early cell state of mouse liver injury repair, are key technical issues in the screening of candidate genes for early regeneration across species.

[0006] Therefore, establishing a technical process that combines single-cell target cell population identification, enriched gene screening, functional enrichment analysis, deduplication of allotetraploid L / S homologous copies, cross-species homology mapping, and intersection analysis of mammalian injury repair models is of great significance for discovering candidate genes co-expressed in the early stages of cross-species and cross-organ regeneration in vertebrates. Summary of the Invention

[0007] In view of this, the present invention provides a method for screening a set of conserved co-expressed candidate genes in the early stage of regeneration based on homologous gene mapping and its application.

[0008] The technical challenge of this invention lies in the fact that conserved genes cannot be directly screened between African clawed frogs and mice through gene symbol intersection. African clawed frogs are allotetraploids with homologous copies of L and S subgenomes, making direct comparison prone to duplicate counting, inconsistent nomenclature, misclassification of homology relationships, and omission of candidate genes. This invention addresses the problems of coexistence of homologous copies of L and S subgenomes in the African clawed frog allotetraploid genome, inconsistencies in cross-species gene nomenclature, and the instability of direct homology comparison between African clawed frogs and mice. It establishes a two-step homology gene mapping process: African clawed frog—tropical clawed frog—mouse. Candidate genes from the early regeneration target cell population of African clawed frogs are converted into comparable homologous genes in mice, and further intersection analysis is performed with upregulated genes from the early damage repair target cell population in mice to obtain a set of conserved co-expressed candidate genes in the early regeneration stage.

[0009] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0010] A method for screening a set of conserved co-expressed candidate genes in the early stages of regeneration based on homologous gene mapping includes the following steps: (1) Obtain early single-cell transcriptome data on tail fin regeneration in African clawed frogs and early single-cell transcriptome data on liver injury repair in mice with cholestasis; (2) Perform quality control, data integration, cell clustering, and cell type annotation on the single-cell transcriptome data respectively; (3) Screening specific upregulated genes of regeneration initiation cells from early data of African clawed frog tail fin regeneration to obtain the first candidate gene set; (4) Screening specific upregulated genes of the second subset of hepatic progenitor cells from early data of cholestatic liver injury repair in mice to obtain a second candidate gene set; (5) Perform two-step homologous gene mapping from African Xenopus to Tropical Xenopus to mouse on the first candidate gene set to obtain a set of comparable homologous genes for mice; (6) Perform an intersection analysis between the mouse comparable homologous gene set and the second candidate gene set to obtain a conserved co-expression candidate gene set in the early stage of regeneration.

[0011] This invention selects the tropical clawed frog as an intermediate species, which is closely related to the African clawed frog and can undertake homologous mapping of the L / S subgenome copy of the African clawed frog. At the same time, the homologous relationship of the tropical clawed frog genome is relatively simple, which facilitates further mapping to mouse genes, making it suitable as a bridging species between the African clawed frog and the mouse.

[0012] The technical challenge lies in the fact that the L and S copies of Xenopus laevis may correspond to the same Xenopus gene, which would still result in duplicate counting if left untreated. To address this, this invention extracts the correspondence between the L and S subgenomes and Xenopus genes, merges them into a unified mapping table, and removes missing and duplicate entries, thereby reducing the interference of allotetraploid structures on subsequent homology mapping and intersection analysis.

[0013] Furthermore, the early single-cell transcriptome data of African clawed frog tail fin regeneration were obtained from an early regeneration model of African clawed frog tail fin after amputation, and the early single-cell transcriptome data of mouse cholestatic liver injury repair were obtained from a cholestatic liver injury repair model induced by 3,5-diethoxycarbonyl-1,4-dihydrocortisone.

[0014] Furthermore, the first candidate gene set is a set of genes specifically upregulated by regeneration initiating cells (RICs) relative to basal epidermal cells (tp63+) and regenerating tissue cells (ROCs).

[0015] Furthermore, the second candidate gene set is a set of genes specifically upregulated by the second subpopulation of mouse hepatic progenitor-like cells, LPLC_2, relative to mature hepatocytes.

[0016] Furthermore, the two-step homologous gene mapping includes: mapping the African Xenopus gene to the tropical Xenopus homologous gene, and then mapping the tropical Xenopus homologous gene to the mouse homologous gene.

[0017] Furthermore, this also includes merging and deduplicating homologous copies of the L and S subgenomes of Xenopus laevis to reduce the impact of homologous copy duplication counts, inconsistent gene symbols, or misclassification on cross-species comparison results.

[0018] Furthermore, the set of conserved co-expressed candidate genes in the early regeneration stage is determined by the intersection of upregulated genes in Xenopus laevis RICs after homology mapping and upregulated genes in mouse LPLC_2, rather than by using a preset minimum number of genes as a screening criterion. The number of candidates in the set can vary in different datasets, screening thresholds, species models, or damage repair models.

[0019] The above method yields a set of conserved co-expressed candidate genes for early regeneration.

[0020] Furthermore, including Pmepa1 , Mmp11 , Tgif1 , Epcam , Rhob , Tubb2b , Tsc22d1 , Sox9 , Pdgfb , Mex3a , Jag1 , c-Jun , Mfge8 , Cd63 , Alcam , St14 , F2rl1 , Sestd1 , Tes , Sox4 and Arl4c .

[0021] The above candidate gene set can be applied in any of the following ways: Research on regeneration mechanisms; Conservative injury response analysis; Cross-species candidate regulatory node screening; Functional verification target selection; Research on tissue damage repair.

[0022] As can be seen from the above technical solution, compared with the prior art, the present invention has the following beneficial effects: (1) This invention converts the early regeneration genes of African clawed frog into comparable homologous genes of mouse through a two-step homologous gene mapping of African clawed frog-tropical clawed frog-mouse, thereby improving the standardization and interpretability of cross-species early regeneration gene comparison.

[0023] (2) In view of the characteristics of the Xenopus xenograft allotetraploid genome, the present invention merges and removes duplicates of homologous copies of the L and S subgenomes, thereby reducing the deviation caused by homologous copy duplication counts and gene naming differences.

[0024] (3) This invention uses African Xenopus RICs and mouse LPLC_2 as comparison objects to avoid comparison only at the whole tissue level, thereby improving the cellular state targeting of candidate gene screening.

[0025] (4) The conserved co-expressed candidate gene set in the early stage of regeneration obtained by the present invention can serve as a candidate gene pool for early stage of regeneration across species and organs, providing candidate targets for research on regeneration mechanisms, functional verification and tissue damage repair. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0027] Figure 1 This is a single-cell transcriptome UMAP analysis diagram of the early stage of tail fin regeneration in the African Xenopus laevis in Example 1 of the present invention, wherein... Figure 1 A is the unsupervised clustering UMAP diagram, used to show the spatial distribution of different clusters in the integrated single-cell data; Figure 1 B is a UMAP plot annotated with cell types, showing the distribution of cell types such as basal (tp63+), ROCs, RICs, goblet cells, SSCs, and notochord.

[0028] Figure 2 This is a heatmap of cell type marker gene expression in the early stage of tail fin regeneration in the African clawed frog in Embodiment 1 of the present invention, used to demonstrate the rationality of the annotation of each cell type.

[0029] Figure 3 This is a diagram showing the proportion of cell types at different time points during the early stage of tail fin regeneration in the African Xenopus laevis in Example 1 of the present invention, illustrating the changes in the proportion of major cell types in samples at 0 hpa, 1 hpa, 3 hpa, and 12 hpa.

[0030] Figure 4 This is a heatmap showing the correlation between different cell types in the early stage of tail fin regeneration in Xenopus laevis in Example 1 of the present invention. It is used to show the correlation between each cell type at the level of average expression profile and to help illustrate the transcriptional similarity between core cell states such as basal (tp63+), RICs and ROCs.

[0031] Figure 5This is a KEGG pathway enrichment analysis diagram of upregulated genes in Xenopus laevis RICs in Example 1 of the present invention. Specifically, three core cell types, basal (tp63+), RICs, and ROCs, were extracted from single-cell data after early integration of Xenopus laevis tail fin regeneration, and a three-cell subset was constructed. Differential expression analysis was performed on basal (tp63+), RICs, and ROCs, and differentially regulated genes in RICs were extracted for KEGG pathway enrichment analysis.

[0032] Figure 6 This is a UMAP diagram of the main cell types in the mouse DDC cholestatic liver injury repair model of the present invention, used to show the distribution of the main cell types after integrating single-cell data at three time points: D0, D17, and R2. After cell type annotation, the main cell types such as Hep, Chol, LVEC, and Ery were identified.

[0033] Figure 7 This is a heatmap of the expression of marker genes of major cell types in the mouse DDC cholestatic liver injury repair model of Example 1 of the present invention, used to demonstrate the rationality of the annotation of major cell types. Among them, the Hep cell population shows the expression characteristics of marker genes related to mature hepatocytes, the Chol cell population shows the expression characteristics of marker genes related to bile duct epithelial cells, and other immune cells, endothelial cells and interstitial cell populations also show the marker gene expression patterns consistent with their annotated identities.

[0034] Figure 8 This is a UMAP analysis diagram of mouse hepatobiliary epithelial cell subsets in Example 1 of the present invention. Specifically, Hep and Chol epithelial-related cells were extracted from the annotation results of the main cell types in the mouse DDC cholestatic liver injury repair model to construct a subset of hepatobiliary epithelial cells, and this subset was re-annotated with dimensionality reduction, clustering and cell state.

[0035] Figure 9 This is a diagram showing the proportion of mouse hepatobiliary epithelial cell subsets in Example 1 of the present invention. It is used to illustrate the changes in the proportion of cell types at three time points: D0, D17, and R2, and to help explain the dynamic changes in the state of hepatobiliary epithelial cells during cholestatic liver injury and early repair.

[0036] Figure 10This is a heatmap of marker gene expression in mouse hepatobiliary epithelial cell subsets in Example 1 of the present invention, used to demonstrate the rationality of the annotation of hepatobiliary epithelial cell subsets. Among them, Mature hepatocyte shows the expression characteristics of marker genes related to mature hepatocytes, Proliferating hep shows the expression characteristics related to proliferation, Cholangiocyte shows the expression characteristics related to bile duct epithelial cells, and LPLC_1 and LPLC_2 show the expression characteristics related to hepatic progenitor-like, bile duct-like, or epithelial state transitions during the damage repair process.

[0037] Figure 11 This is a KEGG pathway enrichment analysis diagram of upregulated genes in mouse LPLC_2 in Example 1 of the present invention. Specifically, based on the annotation results of hepatobiliary epithelial cell subsets, LPLC_2 was selected as the target cell population, and Mature hepatocyte was used as the comparison cell population for differential expression analysis. Upregulated differential genes of LPLC_2 relative to Mature hepatocyte were extracted for KEGG pathway enrichment analysis.

[0038] Figure 12 This is an intersection analysis diagram of candidate genes of Xenopus laevis RICs and upregulated genes of mouse LPLC_2 in Example 1 of the present invention. Specifically, the differentially upregulated genes of Xenopus laevis RICs are converted into a set of comparable homologous genes of mice through a two-step homologous gene mapping from Xenopus laevis to tropical Xenopus to mice, and then the intersection analysis is performed with the set of differentially upregulated genes of mouse LPLC_2 to screen out cross-species conserved candidate genes related to damage repair. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] In this invention, RICs refer to the regeneration initiation cells that appear in the early stages of tail fin regeneration in the African clawed frog. LPLC_2 refers to the second subset of hepatic progenitor-like cells in the early stage of repair of hepatobiliary epithelial cells in mouse DDC cholestatic liver injury. hpa refers to the number of hours following caudal fin amputation; DDC refers to 3,5-diethoxycarbonyl-1,4-dihydrocortisone.

[0041] This invention uses early regeneration RICs of African Xenopus tail fin and early repair of LPLC_2 in mouse DDC cholestatic liver injury as target cell states in the early stage of cross-species and cross-organ injury repair. Through single-cell transcriptome data integration, upregulated gene screening of target cell populations, two-step homologous gene mapping and intersection analysis, a set of conserved co-expressed candidate genes in the early stage of regeneration is obtained.

[0042] The method described in this invention can be implemented in the R language environment. The analysis software or packages used include Seurat, harmony, clusterProfiler, orthogene, ggvenn, ggplot2, and related species annotation databases, which are used to complete single-cell data quality control integration, target cell population upregulated gene screening, homologous gene mapping, intersection analysis, and result visualization.

[0043] Example 1 A set of conserved co-expression candidate genes in the early stage of regeneration based on homologous gene mapping and its screening method are described below: Step 1: Acquire and process single-cell transcriptome data from the early stage of tail fin regeneration in the African clawed frog.

[0044] The data were obtained from the GEO database, dataset number GSE245312, and included single-cell transcriptome sequencing data at four time points: 0 hPa, 1 hPa, 3 hPa, and 12 hPa after tail fin amputation in Xenopus laevis. The raw FASTQ files for each sample were downloaded, and upstream analysis was performed using CellRanger 7.2.0 software. Using Xenopus_laevis_v10.1 as the reference gene set, the CellRanger count program was run with default parameters to complete sequencing data preprocessing, reference genome alignment, cell barcode identification, UMI counting, and gene expression quantification, obtaining the single-cell gene-cell expression matrix for each sample.

[0045] The Seurat software package was used to read the 10x Genomics single-cell expression matrices obtained after CellRanger processing at each time point, and Seurat objects were constructed accordingly. During object construction, genes expressed in at least three cells were retained, and cells with at least 200 detectable genes were initially selected. Subsequently, the number of detectable genes (nFeature_RNA), the total number of UMIs (nCount_RNA), and the mitochondrial gene expression percentage (mt_percent) were calculated for each cell. Based on the distribution of quality control indicators, cells with nFeature_RNA > 2500, nCount_RNA > 2500, and mt_percent < 5 were retained for subsequent analysis.

[0046] After quality control was completed, single-cell data from four time points (0 hPa, 1 hPa, 3 hPa, and 12 hPa) were merged, retaining the time point origin information for each cell. The merged data underwent NormalizeData standardization, FindVariableFeatures hypervariable gene screening, ScaleData data scaling, and RunPCA principal component analysis. In the ScaleData step, mt_percent was used as a covariate in regression to reduce the impact of mitochondrial proportion differences on subsequent analyses.

[0047] Based on the PCA results, the Harmony method was used to correct for potential batch differences between different time points. The top 30 Harmony principal components were selected to construct a cell nearest neighbor map, followed by unsupervised clustering and UMAP dimensionality reduction. The clustering resolution was set to resolution=0.1, and the UMAP parameters were set to min.dist=0.7 and n.neighbors=50. The unsupervised clustering results of cells in the UMAP space after integration are shown below. Figure 1 As shown in Figure A.

[0048] Based on clustering results, known cell type marker genes, and caudal fin regeneration-related cell characteristics, cell type annotation was performed on the integrated cell population. The annotated cell types included basal (tp63+), ROCs, RICs, goblet cells, SSCs, mesenchyme, muscle cells, erythrocytes, satellite cells, ionocytes, neuronal cells, myeloid cells, and notochord. The UMAP cell type annotation results are shown below. Figure 1 As shown in B.

[0049] Among them, the representative marker genes of basal(tp63+) include fgfbp2.L , tp63.L and nptx1.L Representative markers for ROCs include lef1 , tp63.L and fgf10.S Representative marker genes of RICs include nptx1.L , mmp1.S , junb , mmp8.S , tp63.L and jun The aforementioned marker gene expression characteristics support the molecular features of RICs, including damage response, basal-like epithelial features, extracellular matrix remodeling, and immediate early transcriptional responses.

[0050] To verify the rationality of cell type annotation, the average expression levels of representative marker genes in each cell type were calculated, and the average expression of each marker gene in different cell types was Z-score normalized to generate a heatmap of marker gene expression. The heatmap results show that each representative marker exhibits relatively high expression in its corresponding cell type, supporting the cell type annotation results. A heatmap of marker gene expression in cell types during the early stage of tail fin regeneration in the African Xenopus laevis is shown below. Figure 2 As shown.

[0051] To observe early changes in cell composition after caudal fin injury, the number and proportion of different cell types at time points of 0 hPa, 1 hPa, 3 hPa, and 12 hPa were statistically analyzed, and cell type composition ratio diagrams were plotted. These results illustrate the changes in the proportions of major cell types at different time points, particularly the dynamic changes in basal (tp63+), RICs, and ROCs, which are related to epithelial injury response and early regeneration. The cell type composition ratios at different time points in the early stage of caudal fin regeneration in Xenopus laevis are shown below. Figure 3 As shown.

[0052] To further observe transcriptional similarity among different cell types, genes with high variability were screened based on the average expression profile of each cell type, and Spearman correlation coefficients were calculated between cell types to create a correlation heatmap. These results are used to help illustrate a certain transcriptional correlation between core cell states such as basal (tp63+), RICs, and ROCs, but are not considered as sole evidence of pedigree causality. A heatmap showing the correlation between different cell types in the early stage of caudal fin regeneration in the African clawed frog is shown below. Figure 4 As shown.

[0053] Step 2: Screen upregulated genes of Xenopus laevis RICs and perform KEGG pathway enrichment analysis.

[0054] To obtain the first candidate gene set, three core cell types—basal (tp63+), RICs, and ROCs—were extracted from the integrated single-cell data of early tail fin regeneration in Xenopus laevis, constructing a three-cell subset. Using cell type as the cell identity, the FindAllMarkers function in Seurat was used to screen for upregulated genes in RICs.

[0055] The parameters for RICs upregulated gene selection are as follows: only.pos=TRUE, retaining only relatively upregulated genes in the target cell population; test.use="wilcox", using the Wilcoxon rank-sum test; slot="data", using the normalized expression matrix; min.pct=0.25, requiring the gene to be expressed in at least 25% of the cells in the target cell population; logfc.threshold=1, further retaining genes with a corrected p-value p_val_adj<0.01.

[0056] KEGG pathway enrichment analysis was performed on genes upregulated by RICs. First, the identifiable Xenopus gene symbols were converted to EntrezGeneID using the org.Xl.eg.db database. Then, the enrichKEGG function in the clusterProfiler package was used for enrichment analysis, with the species parameter set to organism="xla".

[0057] KEGG enrichment results showed that upregulated genes in RICs were enriched in pathways including the adipocytokine signaling pathway, arginine and proline metabolism, apoptosis, cytoskeleton in muscle cells, C-type lectin receptor signaling pathway, efflorescence, focal adhesion, necroptosis, ECM-receptor interaction, PPAR signaling pathway, Toll-like receptor signaling pathway, and Notch signaling pathway. These pathways collectively point to signaling processes related to extracellular matrix remodeling, cell adhesion, cytoskeleton regulation, damage response, cell death regulation, phagocytosis clearance, and regeneration. The KEGG pathway enrichment results for upregulated genes in Xenopus laevis RICs are shown below. Figure 5 As shown.

[0058] Among the genes upregulated by RICs and their enrichment pathways jun , jun.L , junbAP-1-related immediate early response genes showed upregulation. Among them, jun.L It appears in pathways such as apoptosis, C-typelectin receptor signaling pathway, and focal adhesion pathway, suggesting that AP-1-related transcriptional modules may be involved in damage response, cell adhesion, extracellular matrix signal sensing, and tissue remodeling in RICs.

[0059] It should be noted that the above results indicate that AP-1 family-related transcriptional modules are associated with the early regeneration state of RICs. The infection or disease-related pathway names appearing in KEGG enrichment do not indicate that the sample has been infected with the corresponding pathogen or disease process, but should be understood as the enrichment of molecular modules related to damage response, inflammation-like response, cell adhesion, cytoskeleton regulation, and cell state transition shared in the relevant pathways.

[0060] Step 3: Obtain and process single-cell transcriptome data from the mouse DDC cholestatic liver injury repair model.

[0061] The data were derived from publicly published studies on the spatiotemporal mapping of cholestatic liver injury and repair in mice and were indexed in the CIRSTA database (Cholestatic Injury and Repair Spatio-Temporal Atlas) on the STOmicsDB platform. This database contains single-cell transcriptome data at multiple time points during DDC-induced cholestatic liver injury and the repair phase after DDC withdrawal. This embodiment selected three representative time points—D0, D17, and R2—for analysis. D0 represents the normal liver state without DDC treatment, D17 represents the cholestatic liver injury stage 17 days after DDC treatment, and R2 represents the early liver injury repair stage on day 2 after DDC withdrawal.

[0062] First, cells from the periportal vein region and bile duct-related areas were extracted based on Hepzone information, using the selection criterion Hepzone%in%c(“PV”,“”). Here, PV represents portalvein, i.e., the periportal vein region. This selection strategy was used to improve the targeting of subsequent analyses for key areas of cholestatic liver injury and the status of bile duct-related repair.

[0063] The mitochondrial gene ratio was calculated for single-cell samples at time points D0, D17, and R2. Based on the cell mass distribution at different time points, quality control thresholds were set as follows: for D0 samples, cells with nFeature_RNA > 200 and nFeature_RNA < 5000 were retained; for D17 samples, cells with nFeature_RNA > 200 and nFeature_RNA < 6000 were retained; and for R2 samples, cells with nFeature_RNA > 200 and nFeature_RNA < 8000 were retained. For all three time points, mt_percent < 5 was required.

[0064] After quality control was completed, cell objects from the three time points D0, D17, and R2 were merged, and NormalizeData standardization, FindVariableFeatures high-variability gene screening, ScaleData data scaling, and RunPCA dimensionality reduction were performed sequentially. In the ScaleData step, mt_percent was used as a covariate for regression. Subsequently, the Harmony method was used to correct for potential batch differences between different time points. Based on the Harmony integration results, the top 30 principal components were selected for unsupervised clustering and UMAP dimensionality reduction. The clustering resolution was set to resolution=0.1, and the UMAP parameters were set to n.neighbors=50 and min.dist=0.6.

[0065] Based on known liver cell marker genes, preliminary annotation of major cell types was performed on single-cell data from the integrated mouse DDC cholestatic liver injury repair model. The annotation revealed the major cell types as Hep, Chol, LVEC, Mac, B, cDC1, LAM, pDC, Neutrophil, T, NK, HSC, and Ery. The UMAP results for the major cell types in the mouse DDC model are shown below. Figure 6 As shown.

[0066] To verify the rationality of the annotation of major cell types, the average expression level of marker genes in each cell type was calculated, and the average expression of each marker gene across different cell types was normalized using Z-scores, resulting in a heatmap of marker gene expression. The results showed that in Hep... Alb , Hnf4a , Cat , Fah , Hmgcs2 Genes with relatively high expression in Chol Krt19 , Sox9 , Krt7 , Muc1In the mouse DDC model, marker genes were relatively highly expressed, and other immune cells, endothelial cells, and stromal cell populations also exhibited marker gene expression patterns consistent with their annotated identities. A heatmap of marker gene expression in major cell types is shown below. Figure 7 As shown.

[0067] To further focus on the relationship between hepatocytes, hepatic progenitor-like cells, and bile duct-like cells during the DDC injury repair process, two types of epithelial-associated cells, Hep and Chol, were extracted from the initial annotation objects to construct a subset of hepatobiliary epithelial cells. This subset was then re-standardized, screened for highly variable genes, scaled, PCA dimensionality reduction, Harmony integration, unsupervised clustering, and UMAP dimensionality reduction.

[0068] Sub-clustering analysis of hepatobiliary epithelial cells used the first 30 harmony principal components, with a clustering resolution of 0.6 and UMAP parameters of n.neighbors=50 and min.dist=0.6. After sub-clustering and marker gene annotation, hepatobiliary epithelial cells were classified into five cell states: Mature hepatocyte, Proliferating hep, LPLC_1, LPLC_2, and Cholangiocyte. The UMAP results for mouse hepatobiliary epithelial cell subsets are shown below. Figure 8 As shown.

[0069] The proportions of five hepatobiliary epithelial cell states at three time points (D0, D17, and R2) were statistically analyzed, and subpopulation composition ratios were plotted. These results illustrate the dynamic changes in the states of epithelial-related cells such as Mature hepatocyte, Proliferating hep, LPLC_1, LPLC_2, and Cholangiocyte during cholestatic injury and early repair. The mouse hepatobiliary epithelial cell subpopulation composition ratios are shown below. Figure 9 As shown.

[0070] To validate the rationality of sub-annotation in hepatobiliary epithelial cells, the average expression levels of marker genes in different subpopulations were calculated, and heatmaps were generated after Z-score normalization. The results showed that in Mature hepatocytes... Apoc1 , Alb and Ttr It shows relatively high expression; Proliferating hep Dhfr , Gsta4 , Gstm1 , Adh1 , Gstm3 and Aldh1a7 It shows relatively high expression; in LPLC_1 Cfb , Gm20547 , Apoh, Cfi , Fgb and Gc Genes such as LPLC_2 are relatively highly expressed; Anxa5 , Bicc1 , Spint2 , Dsg2 , Myl12a , Cldn3 and App It is expressed at a relatively high level in Cholangiocytes. Spp1 and Krt19 They exhibit relatively high expression. A heatmap of expression of marker genes in mouse hepatobiliary epithelial cell subsets is shown below. Figure 10 As shown.

[0071] The above results support the view that LPLC_2 is an important cellular state with bile duct-like, progenitor-like, and epithelial junction remodeling characteristics in the repair process of DDC cholestatic liver injury, and can serve as a target cell population in the early stage of mammalian injury repair.

[0072] Step 4: Screen for upregulated genes in mouse LPLC_2 and perform KEGG pathway enrichment analysis.

[0073] To obtain a second set of candidate genes, LPLC_2 was used as the target cell population and Mature hepatocytes as the comparison cell population. The FindMarkers function in Seurat was used to screen for upregulated genes of LPLC_2 relative to mature hepatocytes.

[0074] The differential analysis parameters are as follows: ident.1="LPLC_2", ident.2="Mature hepatocyte"; only.pos=TRUE; test.use="wilcox"; slot="data"; min.pct=0.25; logfc.threshold=3; genes with p_val_adj<0.05 were further retained. Based on the above screening criteria, 426 LPLC_2 genes with high fold-wise upregulation relative to Mature hepatocyte were obtained in this embodiment. This gene set represents the molecular programs related to bile duct-like, progenitor-like, epithelial junction remodeling, and injury response acquired by LPLC_2 relative to mature hepatocytes during the repair process of DDC cholestatic liver injury.

[0075] KEGG pathway enrichment analysis was performed on upregulated genes in LPLC_2. First, the differentially expressed genes were converted to mouse EntrezGeneID, and then enrichment analysis was performed using the enrichKEGG function in the clusterProfiler software package, with the species parameter set to organism="mmu".

[0076] KEGG enrichment results showed that LPLC2 upregulated genes were mainly enriched in the cadherin signaling pathway, focal adhesion pathway, MAPK signaling pathway, ECM-receptor interaction pathway, integrin signaling pathway, PI3K-Akt signaling pathway, Hippo signaling pathway, cornified envelope formation pathway, and cell adhesion molecule interaction pathway. These results suggest that LPLC2 involves transcriptional regulation programs related to intercellular adhesion, cell-extracellular matrix junctions, integrin-mediated signal transduction, focal adhesion formation, epithelial remodeling, and tissue repair. The pathway enrichment analysis results of mouse LPLC2 upregulated genes are shown below. Figure 11 As shown.

[0077] It should be noted that the appearance of tumor, infection, or disease-related pathway names in KEGG enrichment does not indicate that the sample has a corresponding tumor or pathogen infection. These pathways contain a large number of shared gene modules related to cell adhesion, MAPK signaling, PI3K-Akt signaling, cytoskeleton regulation, cell state transition, and damage repair. In this embodiment, they should be interpreted as enrichment of damage repair-related molecular modules.

[0078] Among the genes upregulated by LPLC_2 c-Jun It showed a significant upward adjustment, and c-Jun It appeared in enriched pathway-related genes such as the focal adhesion pathway and the MAPK signaling pathway. This result suggests... c-Jun / AP-1-related immediate early damage response transcriptional modules may be involved in the acquisition of bile duct-like state, cell junction remodeling, cell adhesion regulation, and damage repair response processes in LPLC_2.

[0079] Step 5: Perform two-step homologous gene mapping on the upregulated genes of Xenopus laevis RICs.

[0080] Using the upregulated genes of Xenopus laevis RICs obtained in step two as input, a cross-species homologous gene mapping process from Xenopus laevis to mice was established. Since Xenopus laevis is an allotetraploid, a large number of genes in its genome have homologous copies in both the L and S subgenomes. Directly comparing Xenopus laevis gene symbols with mouse gene symbols can easily lead to homologous copy duplication counting, inconsistent gene nomenclature, and misclassification of homologous relationships.

[0081] This embodiment employs a two-step homologous gene mapping strategy: African Xenopus-Tropical Xenopus-Mouse.

[0082] First, the XenbaseGenepageToGeneIdMapping.txt file provided by the Xenbase database is read. This file includes fields such as GenePageID, GenePageSymbol, Xt_GeneID, Xt_Symbol, Xl_L_GeneID, Xl_L_Symbol, Xl_S_GeneID, and Xl_S_Symbol. The Xenbase mapping file can be updated along with the database; therefore, the file download date is recorded during implementation. In this embodiment, the download date of the Xenbase mapping file used is June 3, 2026.

[0083] The first step involved extracting the mapping relationships between copies of the L subgenome of Xenopus africanus and those of Xenopus laevis, as well as between copies of the S subgenome of Xenopus africanus and those of Xenopus laevis. The L and S copy mapping results were then merged into a unified Xenopus africanus-to-Xenopus laevis mapping table, with duplicate rows and missing entries removed.

[0084] The second step involves left-joining the list of RICs-upregulated genes obtained in step two with the aforementioned Xenopus laevis-to-Xenopus mapping table to obtain the Xenopus laevis homologous genes corresponding to the RICs-upregulated genes. Xenopus laevis genes that fail to match Xenopus laevis homologous genes are marked as unmapped entries. For cases where L / S homologous copies map to the same Xenopus laevis gene, merging and deduplication are performed to reduce the impact of duplication counts caused by Xenopus laevis allotetraploid genomes.

[0085] The third step involves using the orthogene software package to map the deduplicated Xenopus genes to mouse homologous genes. The input species is set to Xenopus (Silurana) tropicalis, and the output species is set to Mus musculus. The mapping method can be homogene or other available homology databases, preserving complex homology relationships. The final result is a set of comparable mouse homologous genes derived from the upregulated genes of Xenopus RICs.

[0086] In the aforementioned homology mapping process, the number of successfully mapped Xenopus genes, the number of unmapped genes, the number of uniquely mapped Xenopus genes, the number of uniquely mapped Tropical Xenopus genes after deduplication, the number of genes reduced due to L / S homology copy merging, and the number of Tropical Xenopus genes derived from the merging of multiple Xenopus L / S copies were statistically analyzed. These statistical results are used to evaluate mapping efficiency and the effectiveness of L / S copy merging, and can be output as a homology mapping results table.

[0087] During homology mapping, Jun-related genes are verified at the gene ID level to avoid errors caused by... jun , jun.L , junb , junb.L Other similar nomenclature can lead to duplicate counting, incorrect merging, or omission of candidate genes. After the above processing, a set of mouse comparable homologous genes is obtained for comparison with the set of mouse LPLC_2 upregulated genes.

[0088] Step 6: Perform intersection analysis on candidate homologous genes of Xenopus laevis RICs and upregulated genes of mouse LPLC_2.

[0089] The set of comparable homologous genes in mice obtained through a two-step mapping process from Xenopus laevis to Xenopus tropicae to mice was subjected to an intersection analysis with the set of mouse LPLC_2 upregulated genes obtained in step four. It should be noted that in the intersection analysis, Xenopus laevis RIC candidate genes were not directly compared using their original Xenopus laevis gene symbols. Instead, they were first converted to comparable mouse gene symbols through homologous gene mapping before being compared with mouse LPLC_2 upregulated genes.

[0090] In this embodiment, the mouse comparable homologous gene symbols obtained from African Xenopus RICs after two-step homology mapping are organized into a first gene set, and the mouse LPLC_2 upregulated differential gene symbols are organized into a second gene set. The two gene sets can be compiled into a CSV format input file, where the first column can be named "African Xenopus," containing the mouse comparable homologous gene symbols obtained from the two-step homology mapping of upregulated genes from African Xenopus RICs; the second column can be named "Mouse," containing the mouse LPLC_2 upregulated differential genes relative to Mature hepatocytes.

[0091] The two gene symbols listed above are then standardized and cleaned. This standardization and cleaning includes: converting the input gene symbols to character type, removing leading and trailing spaces, deleting true missing values, deleting empty strings, and deleting invalid entries such as NA, NaN, NULL, or null. Duplicate entries are also removed from each gene list. The first cleaned gene set is denoted as xenopus_fin, and the second cleaned gene set is denoted as mouse_liver.

[0092] An intersection operation was performed on the cleaned xenopus_fin and mouse_liver sets to obtain common candidate genes that coexist in both the early RICs homologous candidate gene set of Xenopus tail fin regeneration and the early LPLC_2 upregulated gene set of mouse DDC cholestatic liver injury repair. This step can be implemented using the intersect function in R, and the number of common genes can be further counted and a common gene list file can be exported. To facilitate the visualization of the overlap between the two candidate gene sets, a Venn diagram can be drawn using the ggvenn package and output in PDF format. The intersection analysis results are shown below. Figure 12 As shown.

[0093] In this embodiment, intersection analysis revealed 21 common candidate genes between the mouse comparable candidate gene set of African Xenopus RICs after homology mapping and the mouse LPLC_2 upregulated gene set. These 21 common candidate genes are... Pmepa1 , Mmp11 , Tgif1 , Epcam , Rhob , Tubb2b , Tsc22d1 , Sox9 , Pdgfb , Mex3a , Jag1 , c-Jun , Mfge8 , Cd63 , Alcam , St14 , F2rl1 , Sestd1 , Tes , Sox4 and Arl4c .

[0094] Step 7: Identify the set of conserved co-expressed candidate genes in the early stages of regeneration.

[0095] Based on the intersection analysis results in step six, a set of conserved co-expressed candidate genes in the early stage of regeneration was identified, supported by the set of homologous candidate genes of early RICs in the tail fin regeneration of Xenopus laevis and the set of upregulated genes of LPLC_2 in the early stage of repair of cholestatic liver injury in mouse DDC.

[0096] in, c-Jun The simultaneous presence of these genes in the candidate gene set of Xenopus laevis RICs after homology mapping and in the mouse LPLC_2 upregulated gene set suggests... c-Jun The / AP-1-related transcriptional module exhibits activation characteristics in the early cellular state of caudal fin regeneration in lower vertebrates and in the early cellular state of liver injury repair in mammals. Except c-Jun In addition, in the intersection of genes Epcam , Sox9 , Jag1 , Alcam and Pdgfb These genes are also associated with epithelial state transition, cell adhesion, extracellular matrix remodeling, cell communication, phagocytic clearance, or damage repair. Therefore, this common candidate gene set can serve as an early candidate gene pool for cross-species and cross-organ regeneration.

[0097] In this embodiment, c-Jun / AP-1-related genes are considered key candidate regulatory modules because: c-Jun It possesses transcriptional regulatory properties; c-Jun In Xenopus laevis RICs, they are upregulated and enter pathway modules related to damage response, cell adhesion, or cell death. c-Jun Jun was upregulated in the mouse LPLC_2 upregulated gene set; Jun also appeared in the common candidate gene set of both models. These results support... c-Jun / AP-1 serves as a candidate conserved regulatory node for early cross-species and cross-organ regeneration.

[0098] The conserved co-expression candidate gene set obtained in the early regeneration stage of this invention is not limited to a preset minimum number of genes, but is obtained through a process of screening upregulated genes in Xenopus laevis RICs, two-step homologous gene mapping, screening upregulated genes in mouse LPLC_2, and intersection analysis. The number of candidate genes can vary in different datasets, screening thresholds, species models, or damage repair models.

[0099] The results of this invention can be output in the form of tables, graphs, or data files. The output results include, but are not limited to: a list of genes upregulated by Xenopus laevis RICs, a mapping table from Xenopus laevis to Tropical Xenopus, a mapping table from Tropical Xenopus to mice, a list of genes upregulated by LPLC_2 in mice, a table of KEGG enrichment results for genes upregulated by RICs, a table of KEGG enrichment results for genes upregulated by LPLC_2, a set of comparable genes in mice after homology mapping, intersection analysis results, Venn diagrams, and a set of conserved co-expression candidate genes in the early regeneration stage.

[0100] In this embodiment, the set of conserved co-expressed candidate genes in the early regeneration stage includes Pmepa1 , Mmp11 , Tgif1 , Epcam , Rhob , Tubb2b , Tsc22d1 , Sox9 , Pdgfb , Mex3a , Jag1 , c-Jun , Mfge8 , Cd63 , Alcam , St14 , F2rl1 , Sestd1 , Tes , Sox4 and Arl4c This candidate gene set can be used for research on regeneration mechanisms, analysis of conserved damage responses, screening of cross-species candidate regulatory nodes, prioritization of candidate genes, screening of functional validation targets, or research related to tissue damage repair.

[0101] Through the above embodiments, this invention establishes a method for screening a set of conserved co-expressed candidate genes in the early stage of regeneration based on homologous gene mapping. This method uses early regeneration RICs of the African Xenopus tail fin and early LPLC_2 cells in the mouse DDC cholestatic liver injury repair stage as target cell populations. Through single-cell transcriptome integration analysis, screening of upregulated genes in the target cell population, merging and deduplication of African Xenopus L / S homologous copies, two-step homologous gene mapping from African Xenopus to Tropical Xenopus to mouse, and cross-model intersection analysis, a set of conserved co-expressed candidate genes in the early stage of regeneration is obtained.

[0102] This invention can reduce the impact of duplicate counting and naming inconsistencies caused by the allotetraploid L / S homologous copies of Xenopus laevis on cross-species comparisons, enabling the upregulated genes of RICs in the Xenopus laevis tail fin regeneration model to be converted into a mouse comparable homologous gene set, thereby achieving candidate gene comparison between lower vertebrate regeneration models and mammalian injury repair models.

[0103] This invention employs intersection analysis between the set of upregulated homologous genes from Xenopus laevis RICs and the set of upregulated genes from mouse LPLC_2 to screen for candidate genes co-occurring in two different species, organs, and injury repair models. This screening strategy enhances the conservation and priority of candidate genes in the early stages of regeneration, providing a foundation for subsequent functional validation, mechanistic studies, and the screening of regeneration regulatory targets.

[0104] The conserved co-expression candidate gene set obtained in the early regeneration stage of this invention includes c-Jun Genes related to transcriptional regulation, and Epcam , Sox9 , Jag1 Genes related to epithelial state transition, cell adhesion, extracellular matrix remodeling, cell communication, phagocytosis and clearance, and damage repair. These candidate genes collectively constitute a cross-species, cross-organ early-stage candidate gene pool for damage repair.

[0105] The candidate gene set obtained in this invention is based on differential expression, functional enrichment, homology mapping, and intersection analysis of single-cell transcriptomes. The core of this invention lies in providing a reproducible and scalable method for screening conserved candidate genes in the early stages of cross-species regeneration, and the candidate gene set obtained by this method.

[0106] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0107] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for screening a set of conserved co-expressed candidate genes in the early stages of regeneration based on homologous gene mapping, characterized in that, Includes the following steps: (1) Obtain early single-cell transcriptome data on tail fin regeneration in African clawed frogs and early single-cell transcriptome data on liver injury repair in mice with cholestasis; (2) Perform quality control, data integration, cell clustering, and cell type annotation on the single-cell transcriptome data respectively; (3) Screening specific upregulated genes of regeneration initiation cells from early data of African clawed frog tail fin regeneration to obtain the first candidate gene set; The first candidate gene set is a set of genes specifically upregulated by regeneration initiating cells (RICs) relative to basal epidermal cells and regenerating tissue cells. (4) Screening specific upregulated genes of the second subset of hepatic progenitor cells from early data of cholestatic liver injury repair in mice to obtain a second candidate gene set; The second candidate gene set is a set of genes specifically upregulated by mouse hepatic progenitor cell-like cell subpopulation LPLC_2 relative to mature hepatocytes; (5) Perform two-step homologous gene mapping from African Xenopus to Tropical Xenopus to mouse on the first candidate gene set to obtain a set of comparable homologous genes for mice; (6) Perform an intersection analysis between the mouse comparable homologous gene set and the second candidate gene set to obtain a conserved co-expression candidate gene set in the early stage of regeneration.

2. The method as described in claim 1, characterized in that, The early single-cell transcriptome data of Xenopus laevis tail fin regeneration were obtained from an early regeneration model of Xenopus laevis tail fin after amputation, and the early single-cell transcriptome data of mouse cholestatic liver injury repair were obtained from a cholestatic liver injury repair model induced by 3,5-diethoxycarbonyl-1,4-dihydrocortisone.

3. The method as described in claim 1, characterized in that, The two-step homologous gene mapping includes: mapping the African Xenopus gene to the tropical Xenopus homologous gene, and then mapping the tropical Xenopus homologous gene to the mouse homologous gene.

4. The method as described in claim 3, characterized in that, It also includes merging and deduplicating homologous copies of the L and S subgenomes of Xenopus laevis.

5. The set of conserved co-expressed candidate genes in the early regeneration stage obtained by the method of any one of claims 1 to 4.

6. The candidate gene set as described in claim 5, characterized in that, Depend on Pmepa1 , Mmp11 , Tgif1 , Epcam , Rhob , Tubb2b , Tsc22d1 , Sox9 , Pdgfb , Mex3a , Jag1 , c-Jun , Mfge8 , Cd63 , Alcam , St14 , F2rl1 , Sestd1 , Tes , Sox4 and Arl4c composition.

Images