Composition for predicting drug responsiveness in patients with lupus nephritis, and use thereof

Abstract

The present invention relates to a composition for predicting drug responsiveness in patients with lupus nephritis, and a diagnostic kit and information provision method using same. More specifically, an IFN-I-responsive metagene signal is measured at the start of treatment and at 3 months in patients with lupus nephritis so that non-responsive patients whose IFN-I signaling is not sufficiently inhibited by standard remission induction therapy can be identified early, and thus a composition for predicting drug responsiveness, which enables the progression of kidney damage to be inhibited through preemptive treatment strategies such as avoiding unnecessary treatment and switching treatment to other drugs, can be provided. Therefore, the composition for predicting drug responsiveness, of the present invention, enables a customized treatment strategy to be established for each patient, and thus can contribute to overcoming treatment resistance and improving clinical prognosis.

Description

Composition for predicting drug responsiveness in patients with lupus nephritis and use thereof

[0001] The present invention relates to a composition for predicting drug responsiveness in patients with lupus nephritis, a diagnostic kit using the same, and a method for providing information.

[0002] Systemic lupus erythematosus (SLE) is a chronic autoimmune disease characterized by widespread inflammation and multiple organ involvement, and lupus nephritis (LN) is one of the most severe symptoms. LN occurs in approximately 35–60% of SLE patients and is a major cause of increased morbidity and mortality. Although current standard immunosuppressive therapy has improved LN treatment outcomes over the past few decades, a significant number of patients fail to achieve remission or experience frequent relapses.

[0003] Medical professionals often face a dilemma of balancing intensified immunosuppressive therapy with an increasing drug burden to preserve kidney function. To address this challenge, an in-depth understanding of the immunological mechanisms causing LN treatment resistance is essential for developing effective treatment strategies.

[0004] Type I interferons (IFN-I, i.e., IFN-α and IFN-β) are key amplifiers in the pathogenesis of SLE. IFN-I causes tissue damage by regulating various inflammatory processes, such as promoting autoreactive B cell survival, activating T cells, and stimulating cytokine production. Phase 3 clinical trials (TULIP-1 and TULIP-2) of the IFN-I receptor antagonist anifrolumab highlighted the importance of IFN-I signaling by showing that blocking the IFN-I signaling pathway significantly reduced the expression of the interferon-stimulating gene (ISG).

[0005] IFN-I signaling traditionally involves the phosphorylation of STAT1 and STAT2 proteins, which then complex with IRF9 to form the standard transcription factor ISGF3, activating its rapid and transient expression. However, there is evidence that other 'non-standard' IFN-I signaling pathways can sustain ISGF3 expression.

[0006] In particular, long-term IFN-I exposure induces the accumulation of STAT1, STAT2, and IRF9 proteins, which can form U-ISGF3. U-ISGF3 is an unphosphorylated ISGF3 complex that induces the expression of specific ISG subgroups. This excessive IFN-I response, characterized by the sustained expression of U-ISGF3-inducing genes, has been shown to contribute to hyperinflammation in high-risk COVID-19 patients. Interestingly, corticosteroid treatment significantly suppressed the expression of U-ISGF3-inducing genes by downregulating STAT1 expression in monocytes.

[0007] In renal biopsy tissues from LN patients, ISG is highly expressed across the immune cell population but is significantly absent in renal epithelial cells. The observation of a strong correlation in IFN-I signaling between blood and kidney samples suggests that IFN-I pathway activation is primarily triggered by extrarenal mechanisms. However, the mechanisms causing persistent intrarenal inflammation despite aggressive immunosuppressive therapy remain largely unknown. In this study, we performed a systematic single-cell analysis of temporal gene expression dynamics in circulating immune cells of SLE patients undergoing remission induction therapy to identify a distinct immunomodulatory network distinguishing between responders and non-responders.

[0008] The present invention aims to provide a composition for predicting drug responsiveness for early screening of patients with lupus nephritis who exhibit resistance to standard treatment.

[0009] The present invention provides a composition for predicting drug responsiveness in patients with lupus nephritis, comprising as an active ingredient a preparation capable of confirming the expression level of a protein selected from the group consisting of IRF7, LY6E, IFI44L, and IFI6, or a gene encoding the same.

[0010] The present invention provides a diagnostic kit for predicting drug responsiveness in patients with lupus nephritis comprising the above-mentioned composition for predicting drug responsiveness.

[0011] In addition, the present invention comprises the step of confirming the expression level of a protein or a gene encoding therein selected from the group consisting of IRF7, LY6E, IFI44L, and IFI6 in a sample isolated from a patient with lupus nephritis within 3 months of drug administration;

[0012] A step of comparing the expression levels of the above IRF7, LY6E, IFI44L, and IFI6 with the levels expressed in a sample isolated from a patient with lupus nephritis prior to drug administration; and

[0013] A method for providing useful information for predicting drug responsiveness in patients with lupus nephritis is provided, comprising the step of determining that sensitivity to the drug is indicated when the expression levels of IRF7, LY6E, IFI44L, and IFI6 are reduced compared to before drug administration.

[0014] The composition for predicting drug responsiveness according to the present invention can prevent the progression of kidney damage by avoiding unnecessary treatment and implementing preemptive treatment strategies, such as switching to other drugs, by early identifying non-responsive patients for whom IFN-I signal suppression is not sufficiently achieved by induction therapy, a standard treatment method, through the measurement of IFN-I responsive gene signals at the start of treatment and at 3 months of treatment in patients with lupus nephritis. Therefore, the composition for predicting drug responsiveness according to the present invention can demonstrate an effect that contributes to overcoming treatment resistance and improving clinical prognosis by enabling the establishment of personalized treatment strategies for each patient.

[0015] Figure 1a is a schematic diagram of the study design and data generation workflow. For single-cell RNA sequencing, peripheral blood samples were collected from 10 patients with lupus nephritis (5 complete responders and 5 non-responders) at the baseline time (before administration of corticosteroids and MMF) and at four time points at 3, 6, and 12 months after the start of treatment. Renal response was defined at 6 and 12 months according to the 2006 American College of Rheumatology response criteria and the 2012 Kidney Disease: Improving Global Outcomes guideline. CR is complete response, NR is non-response, and scRNA-seq is the single-cell RNA sequencing result.

[0016] Figure 1b shows the results of the Uniform Manifold Approximation and Projection (UMAP) analysis on 238,366 single cells, color-coded by annotated immune cell types, where the heatmap in the lower right indicates the average expression levels of representative marker genes in each identified cell population. CM: classical monocyte; IM: intermediate monocyte; NM: non-classical monocyte; DC: conventional dendritic cell; pDC: plasmacytoid dendritic cell; B: B cell; PB: plasmablast; Tn: naive T cell; Tm: memory T cell; CTL: cytotoxic T cell; NK: natural killer cell.

[0017] Figure 1c shows the results of quantifying differentially expressed genes (DEGs) between complete responders and non-responders across the circulating immune cell compartment at 6 and 12 months after treatment, with the bar graph representing the number of DEGs by cell type (corrected P value < 0.05) and highlighting differences in transcriptional levels related to treatment response.

[0018] Figure 2 shows the results of distinguishing treatment responders and non-responders through monocyte transcription characteristics, Figure 2a shows the results of four distinct bone marrow cell subclusters (top right) identified through single-cell RNA sequencing analysis, and the heatmap shows differentially expressed genes (DEGs) between complete responders and non-responders across the entire subcluster at 6 and 12 months after treatment.

[0019] Figure 2b is a Venn diagram showing the number and proportion of upregulated and downregulated genes shared between classical monocytes and intermediate monocytes at 6 and 12 months after treatment.

[0020] Figure 2c is a Volcano plot showing the DEG between complete responders and non-responders in the integrated classical and intermediate monocyte profiles before and after remission induction therapy. CM: Classical monocytes; CR: Complete responders; DC: Dendritic cells; IM: Intermediate monocytes; NM: Non-classical monocytes; NR: Non-responders.

[0021] Figure 3 shows the results of identifying modules related to remission induction therapy through gene co-expression network analysis. Figure 3a is a hierarchical cluster dendrogram of 2,479 genes included in the weighted gene co-expression network analysis (WGCNA), where each module is assigned a random color and genes not clustered in any module are displayed in gray.

[0022] Figure 3b shows the heatmap results illustrating the correlation between module-specific genes and treatment outcomes, with the correlation coefficient and statistical significance indicated in each cell. *P < 0.05; **P < 0.01.

[0023] Figure 3c shows the temporal dynamics of unique gene expression for four clinically significant modules from the start of treatment to 3, 6, and 12 months after treatment for complete responders (CR, blue) and non responders (NR, red), where the line represents mean expression and the error bars represent standard error.

[0024] Figure 4 shows the results of the analysis of the biological pathway richness of gene co-expression modules. Functional richness analysis was performed on four clinically relevant WGCNA modules using Hallmark, Gene Ontology Biological Process (GOBP), Reactome, TRRUST v2, and literature-based gene sets indicated in italics. The bar graph shows the significance P-value and the ratio of duplicate genes for each module, and illustrates the dominant biological themes for each module.

[0025] Figure 5 shows the results of confirming the characteristics of early responsiveness through bulk RNA sequencing of individual patients. Figure 5a shows the isolation of monocytes from 13 patients with active proliferative lupus nephritis at baseline and 3 months after induction therapy, and clinical outcomes were evaluated after 1 year of follow-up. For bulk RNA sequencing, peripheral blood samples were collected before the administration of corticosteroids and MMF (baseline) and 3 months after the start of treatment. Renal response was defined at 6 and 12 months according to the 2006 American College of Rheumatology response criteria and the 2012 Kidney Disease: Improving Global Outcomes guideline.

[0026] Figure 5b shows the changes in expression of 18 representative genes of the 'CR-down / NR-up' module between baseline and 3 months after treatment, where red text indicates genes whose expression was significantly reduced in complete responders but not in non-responders. An asterisk (*) indicates a U-ISGF3-inducing gene. *P < 0.05; **P < 0.01.

[0027] Figure 6 shows the results of classifying treatment responses through various interferon-responsive metagene patterns, Figure 6a shows the temporal kinetics of the expression of interferon-responsive metagenes consisting of six genes (IRF7, ISG15, LY6E, IFI44, IFI44L, and IFI6) in monocytes during remission induction therapy, and the effect size including the 95% confidence interval calculated using a linear mixed-effects model shows different patterns between complete responders (CR, blue) and non-responders (NR, red) while considering repeated measurements within patients.

[0028] Figure 6b shows the results of evaluating the association between proteinuria levels, disease activity (SLEDAI-2K), and monocyte IFN-I responsive metagene expression through linear regression analysis.

[0029] The present invention will be described in more detail below.

[0030] This invention is a technology that identifies IFN-I responsive gene signals closely related to treatment responsiveness by analyzing transcriptome changes in immune cells at the single-cell level in a time-series manner during remission induction therapy, which is the standard treatment method for patients with lupus nephritis (LN). It was confirmed that the U-ISGF3-mediated IFN-I response signaling pathway identified in this invention was rapidly resolved in complete responders (CR), but remained continuously activated in non-responders (NR), thereby inducing treatment resistance. Accordingly, the inventors of this invention have completed this invention to utilize the sustained expression level of the U-ISGF3-induced gene as a key biomarker capable of predicting treatment non-responsiveness in lupus nephritis patients, thereby making it useful for evaluating treatment responsiveness and predicting prognosis.

[0031] Accordingly, the present invention provides a composition for predicting drug responsiveness in patients with lupus nephritis, comprising as an active ingredient a preparation capable of confirming the expression level of a protein selected from the group consisting of IRF7 (Interferon Regulatory Factor 7, Gene ID: 3665), LY6E (Lymphocyte Antigen 6 Family Member E, Gene ID: 4061), IFI44L (Interferon Induced Protein 44 Like, Gene ID: 10964) and IFI6 (Interferon Alpha Inducible Protein 6, Gene ID: 2537) or a gene encoding such protein.

[0032] The above composition may further include a preparation capable of confirming the expression level of a protein selected from the group consisting of ISG15 (Interferon-stimulated gene 15, Gene ID: 9636) and IFI44 (Interferon Induced Protein 44, Gene ID: 10561) or a gene encoding the same.

[0033] 상기 조성물은 BATF2(Basic Leucine Zipper ATF-Like Transcription Factor 2, Gene ID: 116071), XAF1(X-linked inhibitor of apoptosis-associated factor-1, Gene ID: 54739), BST2(Bone Marrow Stromal Cell Antigen 2, Gene ID: 684), DDX58(DEAD (Asp-Glu-Ala-Asp) box polypeptide 58, Gene ID: 23586), DDX60(DExD / H-box helicase 60, Gene ID: 55601), EPSTI1(Epithelial Stromal Interaction 1, Gene ID: 94240), HERC5(HECT And RLD Domain Containing E3 Ubiquitin Protein Ligase 5, Gene ID: 51191), HERC6(HECT and RLD domain containing E3 ubiquitin protein ligase 6, Gene ID: 55008), IFI27(Interferon Alpha Inducible Protein 27, Gene ID: 3429), IFI35(Interferon induced protein 35, Gene ID: 3430), IFIH1(Interferon Induced with Helicase C Domain 1, Gene ID: 64135), IFIT1(Interferon-induced protein with tetratricopeptide repeats 1, Gene ID: 3434), IFIT3(Interferon-induced protein with tetratricopeptide repeats 3, Gene ID: 3437), IFITM1(Interferon-induced transmembrane protein 1, Gene ID: 8519),MX1(MX dynamin like GTPase 1, Gene ID: 4599), MX2(MX dynamin like GTPase 2, Gene ID: 4600), OAS1(2'-5'-Oligoadenylate Synthetase 1, Gene ID: 4938), OAS2(2'-5'-Oligoadenylate Synthetase 2, Gene ID: 4939), OAS3 (2'-5'-oligoadenylate synthetase 3, Gene ID: 4940), OASL (2'-5'-Oligoadenylate Synthetase Like, Gene ID: 8638), PLSCR1 (phospholipid scramblase 1, Gene ID: 5359), RTP4 (Receptor Transporter Protein 4, Gene ID: 64108), STAT1 (Signal transducer and activator of transcription 1, Gene It may additionally include a preparation capable of confirming the expression level of a protein selected from the group consisting of ID: 6772), STAT2 (Signal transducer and activator of transcription 2, Gene ID: 6773), and TMEM140 (Transmembrane protein 140, Gene ID: 55281) or the gene encoding it.

[0034] In the present invention, the preparation capable of confirming the protein expression level may be an antibody, peptide, or nucleotide that specifically binds to the protein, but is not limited thereto.

[0035] The above antibodies may be polyclonal antibodies, monoclonal antibodies, human antibodies, humanized antibodies, or fragments thereof. Examples of the above antibody fragments include Fab, Fab', F(ab')2, Fv fragments, diabodies, linear antibodies (Zapata et al., Protein Eng. 8(10):1057-1062(1995)), single-strand antibody molecules, or multispecific antibodies formed from antibody fragments.

[0036] In addition, the above antibody is preferably used in an immunoassay method, and the immunoassay method includes, but is not limited to, Western blot, radioimmunoassay, radioimmunoprecipitation, immunoprecipitation, immunohistochemical staining, ELISA (enzyme-linked immunosorbent assay), flow cytometry, or immunofluorescence staining.

[0037] In the present invention, the agent capable of confirming the expression level of the gene may be an antisense oligonucleotide, a primer pair, or a probe that specifically binds to the mRNA of the gene, but is not limited thereto.

[0038] The method for determining the expression level of the above gene is specifically to measure the level of mRNA, and methods for measuring the level of mRNA include reverse transcription polymerase chain reaction (RT-PCR), real-time reverse transcription polymerase chain reaction, RNase protection assay, Northern blot and DNA chip, but are not limited thereto.

[0039] In the present invention, the drug may be administered in combination with a corticosteroid and an immunosuppressant, and the immunosuppressant may be selected from the group consisting of cyclophosphamide, mycophenolate mofetil, azathioprine, and tacrolimus.

[0040] The composition of the present invention may confirm changes in expression levels in samples isolated from patients within 3 months of drug administration.

[0041] More specifically, the present invention can identify high-risk patients who do not achieve sufficient IFN-I signal inhibition early by measuring the IFN-I responsive metagenetic signaling pathway at the time of treatment initiation and at 3 months, and this early prediction can provide the advantage of preventing non-response to combination therapy of corticosteroids and immunosuppressants, which is the standard treatment method for lupus nephritis, and inhibiting the progression of kidney damage due to treatment intensification.

[0042] In the present invention, the sample may be selected from the group consisting of blood, urine, cells, and tissues.

[0043] The present invention provides a diagnostic kit for predicting drug responsiveness in patients with lupus nephritis comprising the above composition.

[0044] The above kit may include primers, probes, or antibodies capable of measuring the expression level of a gene or the amount of protein.

[0045] When the above kit is applied to a PCR amplification process, it may optionally include reagents required for PCR amplification, such as a buffer, a DNA polymerase (e.g., a heat-stable DNA polymerase obtained from Thermus aquaticus (Taq), Thermus thermophilus (Tth), Thermus filiformis, Thermis flavus, Thermococcus literalis, or Pyrococcus furiosus (Pfu)), a DNA polymerase cofactor, and dNTPs; and when the above kit is applied to an immunoassay, the kit of the present invention may optionally include a substrate for a secondary antibody and a label. Furthermore, the kit according to the present invention may be manufactured into a plurality of separate packages or compartments containing the above-mentioned reagent components.

[0046] In addition, the present invention comprises the step of confirming the expression level of a protein or a gene encoding therein selected from the group consisting of IRF7, LY6E, IFI44L, and IFI6 in a sample isolated from a patient with lupus nephritis within 3 months of drug administration;

[0047] A step of comparing the expression levels of the above IRF7, LY6E, IFI44L, and IFI6 with the levels expressed in a sample isolated from a patient with lupus nephritis prior to drug administration; and

[0048] A method for providing useful information for predicting drug responsiveness in patients with lupus nephritis is provided, comprising the step of determining that sensitivity to the drug is indicated when the expression levels of IRF7, LY6E, IFI44L, and IFI6 are reduced compared to before drug administration.

[0049] The above method may further confirm the expression level of one or more proteins selected from the group consisting of ISG15 and IFI44 or genes encoding them in samples isolated from lupus patients within 3 months of drug administration.

[0050] In addition, the above method may further confirm the expression level of one or more proteins selected from the group consisting of BATF2, XAF1, BST2, DDX58, DDX60, EPSTI1, HERC5, HERC6, IFI27, IFI35, IFIH1, IFIT1, IFIT3, IFITM1, MX1, MX2, OAS1, OAS2, OAS3, OASL, PLSCR1, RTP4, STAT1, STAT2, and TMEM140 in samples isolated from lupus patients within 3 months of drug administration.

[0051] According to the present invention, when the expression levels of the protein or the gene encoding it were examined in samples from patients with lupus nephritis prior to drug administration and within 3 months after drug administration, complete responders showed a gradual inhibition of IFN-I signaling, whereas non-responders maintained IFN-I-driven gene expression exhibiting persistent inflammatory characteristics. Bulk RNA sequencing results using monocytes from an independent group of LN patients showed that the expression of six IFN-I responsive genes (IRF7, ISG15, LY6E, IFI44, IFI44L, IFI6) significantly decreased at the 3-month mark in complete responders, but not in non-responders. These gene signaling pathways showed a correlation with proteinuria and disease activity scores.

[0052] Accordingly, the composition for predicting drug responsiveness in patients with lupus nephritis according to the present invention can demonstrate substantial effects that contribute to overcoming treatment resistance and improving clinical prognosis by accurately predicting the treatment responsiveness of patients with lupus nephritis and enabling the establishment of personalized treatment strategies for each patient.

[0053] Hereinafter, the present invention will be described in detail with reference to examples to aid in understanding. However, the following examples are merely illustrative of the content of the present invention and the scope of the present invention is not limited to the following examples. The examples of the present invention are provided to more completely explain the present invention to those with average knowledge in the art.

[0054] <Experimental Example>

[0055] The following experimental examples are intended to provide experimental examples that are commonly applied to each embodiment according to the present invention.

[0056]

[0057] 1. Research Design and Participants

[0058] This study was conducted based on data and biological samples obtained from the Korean SLE (KUDOS) cohort, which was prospectively enrolled from patients with biopsy-confirmed lupus nephritis (LN) at Hanyang University Rheumatology Hospital in Korea. Patients were recruited between March 2018 and July 2020. All patients were 18 years of age or older and met the 2012 SLICC criteria or the 2019 European / American Rheumatology Association (EULAR / ACR) SLE classification criteria.

[0059] From the KUDOS cohort, a homogeneous group of 10 patients was retrospectively selected for single-cell RNA sequencing. This group included only patients who demonstrated a complete response (CR, n = 5) or no response (NR, n = 5) at 12 months, while partial responders were excluded (Group 1). To validate the study results, an additional 13 patients (CR, n = 7; NR, n = 6) from the same cohort were selected to perform bulk RNA sequencing of monocytes (Group 2). Eligible participants exhibited active proliferative LNs confirmed by renal biopsy, persistent proteinuria (urinary protein to creatinine ratio [UPCR] ≥ 500 mg / g or 24-hour urine protein ≥ 500 mg / day), and / or active urinary sedimentation (more than 5 erythrocytes or leukocytes per high-magnification field or presence of cell casts).

[0060] All patients received standard induction therapy using high-dose corticosteroids and mycophenolate mofetil (MMF, 2 g / day) in accordance with LN management guidelines following kidney biopsy and baseline blood collection. Biological agents such as anifrolumab, belimumab, and rituximab were not administered. This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of Hanyang University Hospital (IRB No.: HYUH2017-08-035). All study participants submitted written informed consent.

[0061]

[0062] 2. Sample Collection and Clinical Evaluation

[0063] Blood samples were prospectively collected from all participants of the KUDOS cohort at predetermined time points. Peripheral blood mononuclear cells (PBMCs) were isolated via Ficoll-Paque density gradient centrifugation (GE Healthcare), resuspended in frozen medium (RPMI 1640, 20% fetal bovine serum, 10% DMSO), and stored in liquid nitrogen. For Group 1 (single-cell RNA sequencing), PBMC samples were analyzed at the baseline time point (before administration of high-dose corticosteroids and MMF) and at four time points: 3, 6, and 12 months after the start of treatment. For Group 2 (bulk RNA sequencing), monocytes were isolated from PBMCs collected at the baseline time point and at 3 months after treatment.

[0064] Baseline and follow-up clinical data on renal function, including serum creatinine, estimated glomerular filtration rate, blood urea nitrogen, and UPCR obtained from single urine samples or 24-hour urine collections, were collected. In addition, SLE-related autoantibodies (e.g., anti-dsDNA antibodies) and complement levels (C3, C4, CH50) were evaluated.

[0065] Disease activity was measured using the Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K), and the histopathological classification of the LN, including activity and chronicity indices, was performed according to the revised 2018 International Society of Nephrology / Rapid Physiology (ISN / RPS) guidelines. Treatment response was classified according to the ACR and Renal Diseases: Global Outcome Improvement Guidelines.

[0066]

[0067] 3. Single-cell RNA sequencing and data processing

[0068] Single-cell RNA sequencing was performed using the 10X Genomics Chromium platform and sequenced on an Illumina sequencer. Raw data was processed with Cell Ranger (v6.1.2), aligned to the human genome reference sequence GRCh38-2020-A, and then imported into Seurat (v4.3).

[0069] Through quality control, cells expressing fewer than 200 or more than 6,000 genes, cells with a mitochondrial content exceeding 10%, or cells identified as doublets by scDblFinder (v1.13.10) were removed. Data were normalized using Seurat's SCTransform, batch correction was performed using Batchelor (v1.19.0) and Harmony (v1.2.0), and visualized using UMAP with the top 20 mutual nearest neighbor components for PBMCs and the top 16 principal components for myeloid cells.

[0070]

[0071] 4. Differential Gene Expression and Functional Analysis

[0072] Differential expression analysis was performed on pseudobulk profiles using DESeq2 (v1.30.1). Transcripts with an adjusted P-value of less than 0.05 and an absolute log₂ fold change of greater than 0.25 after adaptive contraction were considered significant. Gene set module scores were calculated using Seurat's AddModuleScore. Overexpression analysis was performed using the hypergeometric test (fgsea v1.16.0) with gene sets from Hallmark, Gene Ontology, Reactome, TRRUST (version 2), and published IFN signaling pathways.

[0073]

[0074] 5. Weighted Co-emergence Network Analysis (WGCNA)

[0075] Weighted Gene Co-expression Network Analysis (WGCNA) was performed on genes filtered by variance. After excluding low-quality genes and samples through standard quality filtering, a weighted correlation matrix encoded using pairwise Pearson correlations was generated (soft-threshold β = 6). Subsequently, modules were identified using hierarchical clustering of topological overlap matrices and the dynamic hybrid tree-cut algorithm (minimum module size 50, threshold 0.3).

[0076]

[0077] 6. Monocyte RNA Sequencing and Analysis

[0078] Monocytes were isolated from peripheral blood mononuclear cells (PBMCs) using magnetic-activated cell sorting (Pan Monocyte Isolation Kit, Miltenyi Biotec). RNA was extracted using TRIzol (Thermo Fisher Scientific) and rRNA and globin mRNA were removed (QIAseq FastSelect-rRNA HMR Kit, Qiagen). The library prepared with the QIAseq FX Single Cell RNA Kit was sequenced (Illumina Novaseq X, 150 bp paired-end reads). Reads were aligned using BWA (v0.7.17) (GRCh38-2020-A) and counted using SAMtools (v1.17).

[0079]

[0080] 7. Statistics and Reproducibility

[0081] All analyses were performed in R, and results were visualized using ggplot2 (v3.4.2) and GraphPad Prism (v9.0.0). Group differences were evaluated using Student's t-test (with Welch's correction), the Mann-Whitney test, or Fisher's exact test. For metagenetic analysis, normalized expression values ​​for selected ISGs were extracted from bone marrow cells. Z-scores were calculated for each gene in all cells, and metagenetic scores were generated by calculating the 10% truncated mean of the total gene Z-scores per cell. A linear mixed-effects model was fitted to the metagenetic Z-scores using bounded maximum likelihood, with patient identity used as the random effect and sampling time as the fixed effect. Predictors of continuous variables were identified using linear regression. A P-value < 0.05 was considered statistically significant, and the Benjamini-Hochberg method (FDR < 0.05) was used for multiple test correction.

[0082]

[0083] <Example 1> Confirmation of Patient Characteristics

[0084] The present invention included a total of 23 patients from the KUDOS cohort. Group 1 (n = 10) underwent single-cell transcriptome analysis of PBMCs, and Group 2 (n = 13) underwent large-scale transcriptome analysis of isolated monocytes. The initial clinical characteristics of the two groups are shown in Table 1.

[0085] The two groups exhibited similar underlying clinical characteristics. The median age was 33 years in Group 1 and 38 years in Group 2, and women predominated in both groups. Clinical markers were similar between the two groups: median SLEDAI-2K score (17 vs. 16), serum creatinine (0.6 mg / dL in both), and UPCR (1,616 vs. 1,218 mg / g). All patients had biopsy-confirmed proliferative lupus nephritis (LN); class IV was predominant in Group 1, while group 2 showed a more heterogeneous form. Histological activity and chronic markers were also similar. All patients received mycophenolate mofetil (MMF) and oral glucocorticoids for remission induction therapy.

[0086] Treatment response was identical between the two groups in Group 1 (50% CR, 50% NR), while Group 2 showed a complete response (CR) of 53.8% and a non-response (NR) of 46.2%. Within each patient group, there were no significant baseline differences between the complete response and non-response groups in demographics, lupus nephritis activity parameters, renal histology, or prior immunosuppressive drug exposure (Table 2).

[0087] Group 1:scRNA-seq analysisGroup 2:bulk RNA-seq analysisNumber1013DemographicsAge, years33 (21, 52)38 (23, 58)Female10 (100)11 (84.6)Lupus nephritis activity parametersUrine protein-to-creatinine, mg / g1616 (846, 5308)1218 (534, 2939)Serum creatinine, mg / dl0.6 (0.4, 1.6)0.6 (0.5, 1.4)C3, mg / dl48 (24, 61)50 (39, 106)C4, mg / dl9 (2, 15)9 (4, 23)anti-dsDNA antibody positive6 (60)11 (84.6)SLEDAI-2K17 (10, 22)16 (8, 21)Renal histologyISN / RPS classIII06 (46.2)III + V02 (15.4)IV9 (90)4 (30.8)IV + V1 (10)1 (7.7)LN activity index8 (5, 14)6 (2, 11)LN chronicity index1 (0, 5)1 (0, 5)TreatmentPrior immunosuppressive therapy(ever exposed)Hydroxychloroquine10 (100)13 (100)Mycophenolate mofetil7 (70)7 (54)Cyclophosphamide2 (20)7 (54)Azathioprine3 (30)6 (46)Cyclosporine2 (20)1 (8)Tacrolimus1 (10)3 (23)Methotrexate4 (40)4 (31)Mizoribine1 (10)1 (8)Induction treatment regimenMycophenolate mofetil based10 (100)13 (100)Glucocorticoids10 (100)13 (100)Renal responseComplete response5 (50)7 (53.8)Non-response5 (50)6 (46.2).

[0088] Group 1: scRNA-seq analysisGroup 2: bulk RNA-seq analysisTotal(n = 10)CR(n = 5)NR(n = 5)Total(n = 13)CR(n = 7)NR(n = 6)Age, years33 (21, 52)37 (21, 52)25 (21, 45)38 (23, 58)48 (23, 58)36 (25, 44)Female10 (100)5 (100)5 (100)11 (84.6)6 (86)5 (83)Urine protein-to-creatinine, mg / g1616(846, 5308)1685(846, 1822)1547(1069, 5308)1218(534, 2939)1052(534, 1607)1827(730, 2939)Serum creatinine, mg / dl0.6 (0.4, 1.6)0.6 (0.4, 0.8)0.8 (0.5, 1.6)0.6 (0.5, 1.4)0.6 (0.5, 0.9)0.6 (0.5, 1.4)anti-dsDNA antibody positive48 (24, 61)3 (60)3 (60)50 (39, 106)6 (86)4 (67)C3, mg / dl9 (2, 15)51 (44, 54)45 (24, 61)9 (4, 23)46 (41, 72)52 (39, 106)C4, mg / dl6 (60)5 (2, 10)11 (4, 15)11 (84.6)9 (6, 19)8 (4, 23)SLEDAI-2K17 (10, 22)16 (10, 20)18 (14, 22)16 (8, 21)16 (8, 21)14 (10, 20)ISN / RPS classIII0006 (46.2)6 (86)0III + V0002 (15.4)03 (50)IV9 (90)5 (100)4 (80)4 (30.8)1 (14)3 (50)IV + V1 (10)01 (20)1 (7.7)LN activity index8 (5, 14)6 (5, 12)8 (6, 14)6 (2, 11)8 (4, 10)4 (2, 11)LN chronicity index1 (0, 5)0 (0, 4)1 (0, 5)1 (0, 5)1 (0, 4)2 (0, 5)Ever immunosuppressive therapyHydroxychloroquine10(100)5 (100)5 (100)13(100)7 (100)6 (100)Mycophenolate mofetil7 (70)2 (40)5 (100)7 (54)2 (29)5 (83)Cyclophosphamide2 (20)2 (40)07 (54)2 (43)4 (67)Azathioprine3 (30)3 (60)06 (46)3 (43)3 (50)Cyclosporine2 (20)02 (40)1 (8)01 (17)Tacrolimus1 (10)01 (20)3 (23)1 (14)2 (33)Methotrexate4 (40)3 (60)1 (20)4 (31)3 (43)1 (17)Mizoribine1 (10)1 (20)01 (8)1 (14)0Induction treatmentMycophenolate mofetil based10(100)5 (100)5 (100)13(100)7 (100)6 (100).

[0089] Data were expressed as n(%) or median (minimum, maximum). No significant difference was observed between Group 1 and Group 2, and individual P-values ​​were not reported.

[0090] * C3: Complement 3, C4: Complement 4, ISN / RPS: International Society of Nephrology / Renal Pathology Society, LN: lupus nephritis, scRNA-seq: single-cell RNA sequencing, SLEDAI-2K: Systemic Lupus Erythematosus Disease Activity Index 2000.

[0091]

[0092] <Example 2> Confirmation of changes in immune cell transcripts after induction therapy

[0093] To identify the transcriptional kinetics underlying the treatment response to lupus nephritis, single-cell transcriptome profiling of peripheral blood proteins (PBMCs) was performed on 10 female patients in Group 1. Blood samples were collected at baseline (before administration of high-dose corticosteroids and MMF) and at 3, 6, and 12 months after standard remission induction therapy (Fig. 1a). All patients demonstrated a renal response (CR or NR) by 6 months, which was maintained stably for 12 months.

[0094] Analysis of all 40 samples revealed a total of 238,366 normal single cells. Through unsupervised hierarchical clustering, five major immune cell compartments were identified: NK and T cells, monocytes and dendritic cells (DCs), B cells, plasmablasts, and plasmacytoid dendritic cells (plasmacytoid DCs), and a total of 13 distinct cell types were identified (Fig. 1b). After immunosuppressant treatment, a slight increase in circulating B cells was observed in the non-responder group at 6 and 12 months compared to baseline.

[0095] Next, analog expression was analyzed in 6- and 12-month samples to identify differentially expressed genes (DEGs) between complete responders and non-responders. Among all immune cell compartments, monocytes and dendritic cells showed the highest levels of DEGs (Fig. 1c). Enrichment analysis revealed high expression of genes involved in inflammatory pathways, IFN-I / II signaling, and IL-2 and IL-6-mediated responses in myeloid cells. Notably, these transcriptional characteristics showed significant differences between CR and NR after standard remission induction therapy.

[0096] In summary, it can be seen that monocytes and dendritic cells are major determinants of the treatment response in lupus nephritis.

[0097]

[0098] <Example 3> Confirmation of Convergent Transcriptional Changes in Classical Monocytes and Intermediate Monocytes

[0099] The inventors of the present invention analyzed 31,536 single-cell transcripts focusing on bone marrow cell-specific profiles and identified four major cell types consistently present throughout the samples (Fig. 2a). These four major cell types are classical monocytes (CM), intermediate monocytes (IM), non-classical monocytes (NM), and dendritic cells (DC). Differential expression analysis between complete responders and non-responders was performed at 6 and 12 months, revealing distinct patterns of gene expression in each subgroup, identifying 146 DEGs in CM, 110 DEGs in IM, 71 DEGs in NM, and 68 DEGs in DC. Among these cell groups, CM and IM not only exhibited the highest number of DEGs (both upregulation and downregulation) but also shared more than one-third of these DEGs (Fig. 2b).

[0100] The above results suggest the existence of common biological mechanisms mediating treatment-related immune modulation. Based on these results, we hypothesized that specifically integrating the gene expression profiles of CM and IM could yield unique characteristics reflecting the treatment response of LN patients. In particular, no significant DEGs were observed between complete responders and non-responders of CM and IM in the baseline comparison (Fig. 2c).

[0101]

[0102] <Example 4> Confirmation of Treatment-Related Gene Co-expression Module

[0103] To identify gene co-expression patterns associated with immune transcriptional changes during treatment, WGCNA was performed on CM and IM transcriptome data. Based on variance-stabilizing transformations, 2,479 genes above the 75th percentile (default) were selected, and the baseline genetic states of patients who achieved a complete response (CR) or were non-responders (NR) were compared. Eight modules of the co-expression gene network were identified using a dynamic hybrid tree cut algorithm for module clustering (Fig. 3a). Subsequently, Pearson correlation coefficients were calculated between the module eigengenes—values ​​representing the summarized principal components of gene expression—and the associated treatment responses. Through this module-based analytical framework, clinically relevant co-expression transcriptional networks could be systematically identified.

[0104] Four of these gene modules showed a significant correlation with renal outcomes after remission induction therapy (Fig. 3b). The green module (r = 0.72, P = 9e-07), which showed a strong correlation with early treatment response at 3 months in patients with complete response, was defined as the Early Signature. The yellow module showed a positive correlation with CR (r = 0.61, P = 8e-05) and was named 'CR-up'. The brown module showed a negative correlation with both CR (r = -0.55, P = 6e-04) and NR (r = -0.76, P = 1e-07), and was designated as 'Common-down'. The blue module, which shows a negative association with CR (r = -0.58, P = 2e-04) but a positive association with NR (r = 0.54, P = 7e-04), was labeled 'CR-down / NR-up'. To clearly show the temporal changes in these transcription profiles, the mean unique gene values ​​of each module were plotted for both the CR and NR groups at baseline and at 3, 6, and 12 months after treatment (Fig. 3c).

[0105]

[0106] <Example 5> Confirmation of sustained atypical IFN-I signal activation in non-responders

[0107] Next, to evaluate the functional importance of each module, an overexpression analysis was performed to identify key biological pathways mediating monocyte-specific immune regulation.

[0108] The 'early signaling' module was found to be significantly abundant in previously reported glucocorticoid-related gene signaling pathways in peripheral blood leukocytes of SLE patients and features key genes such as FKBP5, IL1R2, PHC2, and IRAK3 (Fig. 4a). This module also overlaps with genes (e.g., ZBTB16, SLA, SAP30, and SLC1A3) that are primarily regulated by glucocorticoids rather than IFN-I, as demonstrated by Northcott and colleagues.

[0109] The 'CR-up' module showed abundance in Reactome pathways associated with eukaryotic translation and significantly overlapped with the COVID-19_EARLY_DOWN signal, which included genes such as CLEC10A, EEF1A1, HLA-DRA, and RPL15 among the top 500 genes downregulated in female blood after SARS-CoV-2 infection (Fig. 4b). Unlike these modules, the 'common down' module consisted of genes that were consistently downregulated regardless of elongation results (Fig. 4c). Within these features, typical inflammatory cytokines (e.g., TNF, IL1B, CXCL chemokines) and transcription regulators (e.g., CD83, JUNB, KLF6, NR4A3) were co-expressed, and the Hallmark TNF signaling pathway was remarkably abundant.

[0110] In the 'CR-down / NR-up' module, genes of the Hallmark IFN-α response pathway were found to be high. To analyze the components of IFN signaling in more detail, a set of genes was selected from published literature focusing on genes experimentally verified in human cells (Table 3). Subsequently, the ISGF3-inducing gene group and the U-ISGF3-inducing gene group were analyzed separately.

[0111] As a result, all 29 U-ISGF3-inducing genes, including IFI35, IFIT3, IRF7, and EPSTI1, were present in the 'CR-down / NR-up' module, which showed a higher level of expression abundance than the ISGF3-inducing genes (Fig. 4d). In contrast, 30 of the 48 ISGF3-inducing genes, including many typical ISGs such as CXCL16, MYD88, HLA-F, and UBA7, were not identified in the module.

[0112] ISGF3-inducible genesU-ISGF3-inducible genesSTAT2 / IRF9-inducible genesIFN-γ-specific genesIFN-α-sensitive NF-κB target genes1ARHGAP27BATF2CLDN4CXCL9CCL42B3GALNT1XAF1MSR1CASP7CCL33BLZF1BST2JUNDGBP3CXCL84AIDADDX58DESTLR3TRAF15CUTADDX60DGKASTAT3ICAM16CXCL16EPSTI1CFBSOCS1TNF7CXorf38ISG15APOBEC3GHLA-ATNFAIP88DLL1HERC5EIF2AK2HLA-GNFKB19EHD4HERC6ISG15SERPING1IL610EIF2AK2IFI27MX1HLA-DRAIRF811TENT5AIFI35IFIT3HLA-DMAIL2712SHFLIFI44CCL8HLA-DRB1CTNNB113HES4IFI44LSAA2CD74KLF614HLA-FIFIH1DPYSL4HLA-DMBCCL2015IFI16IFIT1VSIG8HLA-DPA1STAT5A16INSIG1IFIT3CH25HIL32IL1R117MASTLIFITM1CD74GBP5CXCL918SLFN5IRF7TXNIPIDO1TNFSF919MT1MMX1CCL7GABBR1IL1A20MYD88MX2GBP1PTGS221NCOA7OAS1PLAAT4CXCL322NFE2L3OAS2IRF1IRAK223NUDCD1OAS3WARSINHBA24OGFROASLGBP2KDM6B25PANX1PLSCR1CXCL10HES126PARP10RTP4IL18BPFOSL127PCGF5STAT128PHACTR4STAT229PI4K2BTMEM14030PLEKHA431PNPT132PPM1K33PRKD234TMEM17135RAB2036RASGRP337TRIM6938FAM122C39SAMD940SRA141STARD542TDRD743TGM244TNFAIP645TRIM546UBA747UNC93B148ZFYVE26

[0113]

[0114] <Example 6> Confirmation of IFN-I responsive signal in monocytes in an independent patient group

[0115] The abundant selective expression of U-ISGF3-inducing genes in the 'CR-down / NR-up' module suggests that this atypical IFN-I pathway plays a specific role in treatment resistance. Based on these module-based insights, 18 representative genes from the 'CR-down / NR-up' module were selected to construct a concentrated gene signaling pathway that explains more than 95% of the module variance. Notably, 11 of these genes were U-ISGF3-inducing genes (Table 4). As expected, the normalized expression of these 18 genes showed significant differences between the renal response groups at 6 and 12 months after remission induction therapy. This feature was tested in an independent group of 13 patients with active proliferative lymph nodes (Group 2), including 7 complete responders and 6 non-responders (Fig. 5a, Table 1). To identify early genetic characteristics associated with LN treatment outcomes, high-throughput RNA sequencing was performed on monocytes isolated from PBMCs collected at baseline (before remission induction therapy) and 3 months after treatment, and renal response was assessed at 12 months.

[0116] As a result, among the 18 genes representing the 'CR-down / NR-up' module, the expression of 6 ISGs (IRF7, ISG15, LY6E, IFI44, IFI44L, and IFI6) decreased significantly up to 3 months in patients who achieved CR, whereas no similar changes were observed in patients who achieved NR (Fig. 5b).

[0117] MEIFI35 * 0.984023IFIT3 * 0.980404IRF7 * 0.979748ISG15 * 0.97929EPSTI1 *0.973166IFITM30.969677LY6E0.965348OAS1 * 0.964869RSAD20.962528PARP120.961854PARP90.961659IFIT1 * 0.957489SP1100.957333IFITM1 * 0.955741XAF1 * 0.954553IFI44 * 0.951816IFI44L * 0.951034IFI60.950487

[0118]

[0119] <Example 7> Confirmation of treatment results based on changes in IFN-I gene expression patterns

[0120] To determine whether the aforementioned six ISGs could serve as early signals to track treatment response in LNs, a linear mixed-effects model was adopted to analyze gene expression dynamics over time and to examine both inter-donor and intra-donor variability. In this approach, individual cells from single-cell RNA sequencing data were modeled as a function of six IFN-I responsive metagenes (IRF7, ISG15, LY6E, IFI44, IFI44L, and IFI6) (Fig. 6a). Analysis of complete responders revealed that metagene expression was significantly downregulated at all post-treatment time points. These were at 3 months (β = -0.753), 6 months (β = -0.534), and 12 months (β = -0.485) compared to baseline (all P < 0.001). Surprisingly, non-responders showed a gradual increase in metagene expression over time, with the effect size increasing from 3 months (β = 0.146) to 6 months (β = 0.695) and 12 months (β = 0.769), all of which were significant when compared to baseline (P < 0.001). There was a distinct difference in expression patterns between the CR and NR groups, indicating that the trajectories of the IFN-I response in monocytes differ during remission induction therapy.

[0121] Linear regression analysis further confirmed a significant association between monocyte metagene expression after remission induction therapy and proteinuria (R²= 0.435, P < 0.0001) and SLEDAI-2K score (R²= 0.352, P = 0.0005) (Fig. 6b). Similar trends were observed in the NK / T cell and B cell compartments, but to a lesser degree.

[0122] In summary, the above results suggest that the sustained expression of selected IFN-I responsive genes in circulating immune cells is a significant signal indicating non-response to treatment in lupus nephritis (LN).

[0123]

[0124] Foregoing, specific parts of the present invention have been described in detail. It will be apparent to those skilled in the art that such specific descriptions are merely preferred embodiments and do not limit the scope of the invention. Accordingly, the actual scope of the invention is defined by the appended claims and their equivalents.

Claims

A composition for predicting drug responsiveness in patients with lupus nephritis, comprising as an active ingredient a preparation capable of confirming the expression level of a protein selected from the group consisting of IRF7, LY6E, IFI44L, and IFI6, or a gene encoding such protein. In paragraph 1, The above composition is a composition for predicting drug responsiveness in patients with lupus nephritis, further comprising a preparation capable of confirming the expression level of a protein selected from the group consisting of ISG15 and IFI44 or a gene encoding the same. In paragraph 1 or 2, The above composition is a composition for predicting drug responsiveness in patients with lupus nephritis, further comprising a preparation capable of confirming the expression level of a protein selected from the group consisting of BATF2, XAF1, BST2, DDX58, DDX60, EPSTI1, HERC5, HERC6, IFI27, IFI35, IFIH1, IFIT1, IFIT3, IFITM1, MX1, MX2, OAS1, OAS2, OAS3, OASL, PLSCR1, RTP4, STAT1, STAT2, and TMEM140 or a gene encoding the same. In paragraph 1, The above-mentioned drug is a composition for predicting drug responsiveness in patients with lupus nephritis, wherein a corticosteroid and an immunosuppressant are administered in combination. In paragraph 4, A composition for predicting drug responsiveness in patients with lupus nephritis, wherein the above-mentioned immunosuppressant is selected from the group consisting of cyclophosphamide, mycophenolate mofetil, azathioprine, and tacrolimus. In paragraph 1, The above composition is a composition for predicting drug responsiveness in patients with lupus nephritis, which confirms changes in expression levels in samples isolated from patients within 3 months of drug administration. In paragraph 6, The above sample is a composition for predicting drug responsiveness in patients with lupus nephritis, selected from the group consisting of blood, urine, cells, and tissues. A diagnostic kit for predicting drug responsiveness in patients with lupus nephritis comprising a composition of any one of claims 1 to 7. A step of determining the expression level of one or more proteins selected from the group consisting of IRF7, LY6E, IFI44L, and IFI6, or genes encoding the same, in samples isolated from patients with lupus nephritis within 3 months of drug administration; A step of comparing the expression levels of the above IRF7, LY6E, IFI44L, and IFI6 with the levels expressed in a sample isolated from a patient with lupus nephritis prior to drug administration; and A method for providing useful information for predicting drug responsiveness in patients with lupus nephritis, comprising the step of determining that sensitivity to the drug is indicated when the expression levels of IRF7, LY6E, IFI44L, and IFI6 are reduced compared to before drug administration. In Paragraph 9, The above method is a method that provides useful information for predicting drug responsiveness in patients with lupus nephritis, further confirming the expression level of one or more proteins selected from the group consisting of ISG15 and IFI44 or genes encoding them in samples isolated from patients with lupus within 3 months of drug administration. In Article 9 or Article 10, The above method is a method for providing useful information for predicting drug responsiveness in patients with lupus nephritis, which further confirms the expression level of one or more proteins selected from the group consisting of BATF2, XAF1, BST2, DDX58, DDX60, EPSTI1, HERC5, HERC6, IFI27, IFI35, IFIH1, IFIT1, IFIT3, IFITM1, MX1, MX2, OAS1, OAS2, OAS3, OASL, PLSCR1, RTP4, STAT1, STAT2, and TMEM140 in samples isolated from patients with lupus within 3 months of drug administration.

Images