Abstract

Objective

   Disruption in the delicate symphony of genes, microRNA (miRNA), or
   protein expression can result in the dysregulation of the immune
   system, leading to the devastating consequences such as lupus nephritis
   (LN). The capacity of exosomes to transport miRNAs between cells and
   modify the phenotype of recipient cells implies their involvement in
   persistent kidney inflammation. This study unveils identifying two
   previously undiscovered exosomal miRNAs in the serum of LN patients,
   offering potential solutions to the current challenges in LN diagnosis
   and management.

Methods

   Initially, we used a reagent-based kit to isolate serum exosomes from
   patients with Systemic lupus erythematosus (SLE) and used Trizol method
   for total RNA extraction. Subsequently, we employed small RNA
   sequencing to screen for differential expression profiles of exosomal
   small RNAs. The RT-qPCR method was used to individually validate
   samples in both the screening and validation cohorts, enabling the
   identification of candidate small RNAs; specific to LN. We assessed the
   diagnostic potency using receiver operating characteristic (ROC) curve,
   and explored the biological roles of miRNAs using Gene Ontology (GO)
   and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses.

Results

   Compared to SLE patients without LN, SLE patients accompanied by LN
   exhibited significantly spiked levels of exosomal hsa-miR-4796-5p and
   hsa-miR-7974. The duo of miRNAs, hsa-miR-4796-5p and hsa-miR-7974,
   exhibited promising potential as biomarkers for diagnosing LN, with an
   AUC exceeding 0.8. Correlation analysis revealed a strong positive
   association between these miRNAs and proteinuria, as well as the SLE
   Disease Activity Index (SLEDAI) score. Moreover, the levels of two
   miRNAs in LN patients were significantly elevated in comparison to
   other autoimmune nephritis conditions, such as immunoglobulin A
   nephropathy (IgAN) and diabetic nephropathy (DN). Furthermore, the
   bioinformatics analysis indicated that this miRNAs duo can play a
   pivotal role in the regulation of immune processes by modulating signal
   pathways, such as the mTOR and PI3K-Akt signaling pathway.

Conclusion

   This study provides a new ground that serum exosomal miRNAs can
   effectively identify and predict LN in SLE patients.

   Keywords: systemic lupus erythematosus, lupus nephritis, exosomes,
   miRNA, biomarkers

Introduction

   Systemic lupus erythematosus (SLE) continues to pose significant
   challenges within the realm of medicine, causing characteristic blend
   of systemic and organ-specific clinical manifestations, coupled with
   extensive dysfunction of the immune system ([35]1, [36]2). LN stands
   out to be one of the most severe organic manifestations of SLE,
   affecting approximately 30-60% of adults and up to 70% of pediatric
   lupus patients ([37]3). Furthermore, it is noteworthy that LN
   contributes significantly to the elevated incidence of SLE, heightened
   mortality rates, and increased healthcare expenditures ([38]4).
   According to the guidelines, a reliable criterion for diagnosing LN is
   the histopathological confirmation obtained through renal biopsy
   ([39]5, [40]6). However, kidney biopsy is an invasive procedure
   associated with the risk of bleeding and is not easily repeatable.
   Consequently, it poses limitations to rheumatology and immunology
   physicians in their ability to dynamically monitor and manage the
   disease progression of SLE. Currently, commonly used laboratory markers
   for LN include urinary protein, serum creatinine, glomerular filtration
   rate, anti-dsDNA antibody, and serum complements ([41]7). However,
   these clinical parameters fall short of meeting the practical demands
   of clinical settings due to their insufficient sensitivity and
   specificity ([42]8, [43]9). Therefore, it is crucial to discover novel
   non-invasive markers capable of detecting LN activity, predicting
   relapses, and monitoring treatment responses.

   MicroRNAs (miRNAs), tiny non-coding RNAs (18-25 nucleotides) that
   regulate gene expression by binding to messenger RNA (mRNA), play a
   crucial role in cell biology and disease ([44]10). This interaction
   prompts the restraint of mRNA translation and/or hastens its
   degradation, thus culminating in the curtailment of protein synthesis
   specific to certain target proteins ([45]11, [46]12). In 2008, Chen Xi
   et al. ([47]13) made the initial discovery of miRNA in human serum and
   provided evidence that it could serve as a new disease marker. In
   recent years, mounting evidence suggests the involvement of miRNA in
   the occurrence and development of various diseases, such as Alzheimer’s
   disease (AD), cancers, diabetes and autoimmune diseases
   ([48]14–[49]17). Extracellular vesicles(EVs) are membrane vesicles
   released by various cell types and can be categorized into three types:
   exosomes, microvesicles, and apoptotic bodies ([50]18). Numerous
   studies have demonstrated that EVs play a significant role in the
   development of autoimmune diseases through various mechanisms ([51]19).
   Microvesicles (MVs) are larger membrane vesicles derived from the cell
   plasma membrane surface. They can carry nuclear autoantigens and form
   immune complexes (ICs), which activate complements and cause damage to
   renal tissue ([52]20). Nielsen CT et al. utilized immune electron
   microscopy technology to provide evidence of co-localization between
   glomerular deposited immune complexes (ICs) and microvesicles, as well
   as galectin-3-binding protein (G3BP) in LN ([53]21). Exosomes, which
   are generated through the exocytosis of endosomal-derived intracellular
   membrane vesicles into the extracellular space ([54]22), were initially
   discovered by Johnstone and colleagues in 1983 during the culture of
   reticulocytes ([55]23). Previous studies have found that exosomes may
   contribute to a proinflammatory milieu and autoimmune inflammation in
   LN indirectly, either by directly interacting with their associated
   proinflammatory components or by triggering other cells to produce
   proinflammatory cytokines or materials ([56]19, [57]24). The
   dysregulation of circulating exosomal miRNAs in autoimmune diseases has
   been extensively studied, and there is increasing evidence confirming
   their involvement ([58]25, [59]26). However, the diagnostic potential
   of serum exosomal miRNAs in LN has not yet been fully explored.

   In this study, our objective was to investigate the differential
   expression of miRNAs in exosomes derived from the serum of patients
   with LN. We employed RNA sequencing to screen for differential
   expression profiles of exosomal small RNAs. Subsequently, we identified
   the top ten upregulated miRNAs as potential candidate miRNAs. The
   RT-qPCR method was employed in both the screening and validation
   cohorts, facilitating the identification of candidate miRNAs specific
   to LN. To evaluate their potential as biomarkers, we conducted
   receiver-operator characteristic (ROC) curve analysis. Furthermore, we
   utilized bioinformatics tools including Gene Ontology (GO) and Kyoto
   Encyclopedia of Genes and Genomes (KEGG) analyses to predict the target
   genes of these miRNAs and explore their potential functions and
   associated pathways. Our meticulous analysis revealed that serum
   exosomal miRNAs can be utilized as non-invasive biomarkers for the
   identification and prediction of LN in individuals with SLE.

Materials and methods

Subjects and study design

   This study enrolled 232 subjects from September 2022 to March 2023,
   including 116 patients with LN and 116 sex-age-matched patients with
   SLE without LN. All participants were hospitalized at Drum Tower of
   Nanjing Hospital and diagnosed by rheumatologists. The inclusion
   criteria for participants with SLE were based on the modified American
   College of Rheumatology 1997 revised criteria ([60]27, [61]28).
   Individuals with a history of malignant tumors, concurrent infections,
   metabolic abnormalities, or concurrent other autoimmune diseases were
   excluded from the study cohort. The subjects were divided into three
   phases ([62] Figure 1 ). During the discovery phase, we conducted miRNA
   sequencing by extracting exosomes from a 10 ml serum pool consisting of
   20 LN patients and 20 SLE without LN patients (GEO number:
   [63]GSE179950). In the training and validation phase, the expression
   levels of candidate miRNAs were validated in 192 samples using RT-qPCR
   assay. The activity of the disease was evaluated by the SLE Disease
   Activity Index (SLEDAI) ([64]29). Approval to conduct this study was
   provided by the Ethics Committee of the Affiliated Drum Tower Hospital
   of Nanjing University Medical School, with the assigned approval
   identification number 2020-327-01. Prior to their participation in this
   experiment, all individuals involved provided written informed consent.
   The clinical and demographic characteristics of all participants are
   presented in detail in [65]Table 1 .

Figure 1.

   [66]Figure 1
   [67]Open in a new tab

   Workflow of the study.

Table 1.

   Clinical and demographic characteristics of all participants.
   Characteristics SLE without LN (n=116) LN (n=116) P value
   Age, years 40(29-51) 35(27-47) 0.158
   Female, n (%) 97(89.81) 93(86.11) 0.365
   Proteinuria, n (%) 34(31.48) 98(90.74) <0.001***
   Hematuria, n (%) 19(17.59) 75(69.44) <0.001***
   Pyuria, n (%) 27(25.00) 50(46.30) 0.003**
   Cylinderuria, n (%) 0(0) 42(38.89) <0.001***
   24h proteinuria, median (IQR), mg/24h 137.0(67.0-201.0)
   2069.0(699.5-5489.7) <0.001***
   ACR, median (IQR), mg/g 9.2(5.6-24.3) 789.3(189.9-2308.1) <0.001***
   ESR, median (IQR), mm/h 22(10-42) 32(16-57) 0.003**
   Total protein, median (IQR), g/L 66.2(63.3-72.5) 55.0(48.8-64.3)
   <0.001***
   Blood albumin, median (IQR), g/L 37.9(35.4-40.6) 31.9(28.6-36.7)
   <0.001***
   Globulin, median (IQR), g/L 29.3(25.2-33.8) 22.7(18.4-28.6) <0.001***
   A/G, median (IQR) 1.29(1.12-1.54) 1.39(1.15-1.68) 0.056
   GLU, median (IQR), mmol/L 4.46(4.11-4.84) 4.37(3.90-4.86) 0.111
   Urea nitrogen, median (IQR), mmol/L 4.9(3.9-6.0) 7.4(5.2-11.5)
   <0.001***
   CREA, median (IQR), umol/L 49(41-56) 61(47-88) <0.001***
   Uric acid, median (IQR), umol/L 293(208-359) 391(296-452) <0.001***
   Total CO2, median (IQR), mmol/L 24.3(23.0-25.3) 23.6(21.1-25.2) 0.008**
   eGFR, median (IQR), ml/min/1.73m^2 132.7(110.6-164.9) 103.5(75.2-145.0)
   <0.001***
   C1q, median (IQR), mg/dL 15.9(13.7-18.4) 13.2(11.2-17.0) 0.007**
   C3, median (IQR), g/L 0.82(0.62-1.01) 0.76(0.47-0.96) 0.041*
   C4, median (IQR), g/L 0.12(0.07-0.19) 0.13(0.05-0.19) 0.702
   anti-dsDNA, median (IQR), IU/mL 87.43(18.40-253.71)
   112.31(19.74-457.58) 0.392
   ANA, n (%) 88(81.48) 86(79.63) 0.693
   25-(OH) D3, median (IQR), ng/mL 18.50(13.36-24.29) 14.06(8.76-28.19)
   0.051
   SLE-DAI, median (IQR) 4(2-5) 12(8-15) <0.001***
   [68]Open in a new tab

   Abbreviations: ACR, albumin-to-creatinine ratio; ESR, erythrocyte
   sedimentation rate; GLU, glucose; CREA, serum creatinine; eGFR,
   glomerular filtration rate; C1q, complement 1q; C3, complement C3; C4,
   complement C4; anti-dsDNA, anti-double stranded DNA antibody; ANA,
   antinuclear antibodies; SLE-DAI, systemic lupus erythematosus disease
   activity index. *P < 0.05, **P < 0.01, ***P < 0.001 (Mann-Whitney U
   test).

Library preparation and RNA sequencing

   Prior to sequencing, the integrity of RNA samples was validated using
   agarose gel electrophoresis. The quantification of the RNA samples was
   achieved using the NanoDrop ND-1000 instrument. To account for the
   heavy decoration of miRNAs with RNA modifications, several treatments
   were conducted before library construction. These treatments included
   3′-aminoacyl deacylation to 3′-OH for 3′-adaptor ligation, removal of
   3′-cP to 3′-OH for 3′-adaptor ligation, 5′-OH phosphorylation to 5′-P
   for 5′-adaptor ligation, and demethylation of m1A and m3C. The library
   construction and deep sequencing were carried out by BGI (Shenzhen,
   China) using the Illumina Next Seq instrument. The sequencing libraries
   were optimized for RNA biotypes and validated using the Agilent 2100
   Bioanalyzer. The sequencing process consisted of 50 cycles.

Exosomes isolation

   The serum samples were centrifuged at 2000 g for 30 min to separate
   cells and debris. Afterwards, they were stored at -80°C and thawed only
   when needed for use. Exosomes were isolated from serum using a
   commercial kit named as Total Exosome Isolation reagent (Thermo Fisher
   scientific, US) according to the manufacturer’s instructions. Briefly,
   a 100 μL serum sample was mixed with 20 μL of kit reagent and incubated
   at 4°C for 24 h. Then, the sample was centrifuged at 4°C at 10000 g for
   10 min. Discarded the supernatant while the exosome pellets were
   resuspended in 100 μL of 1×phosphate buffer saline (PBS) for further
   analysis.

Exosomes sizing

   For measuring the size, exosomes were first diluted in 1 ml of PBS. The
   mixture was gently inverted to ensure even distribution and then slowly
   added to the NTA sample cell using a 1 ml disposable syringe. Three
   sample videos, each lasting 30 to 60 seconds, were recorded. The
   ZetaView Nanoparticle Tracking Analysis (NTA) system was utilized to
   measure the average diameter of the exosomes. This system has the
   capability to characterize the size distribution of small particles in
   liquid samples and can detect diameters ranging from 20 to 1000 nm
   ([69]30). The results obtained from processing the Software ZetaView
   (Zeta View 8.04.02) were expressed as the mean standard deviation (SD)
   of the three video recordings.

Transmission electron microscopy

   A total of 20 μL of exosomes were applied onto the copper grid of the
   electron microscope and allowed to sit at room temperature for 10 min.
   Subsequently, 20 μL of 2% phosphotungstic acid was added to the copper
   grid for negative staining, which was carried out for 10 min. Excess
   staining was removed using filter paper. Once the copper grid was
   dried, transmission electron microscopy (TEM) was performed to examine
   the exosomes. The test was conducted under the at 120KV. A bilayer
   membrane structure was selected, and particles with diameters ranging
   from 100 to 200 nm were captured in the photographs.

Western blotting

   The exosomes obtained from serum were treated with a lysis solution
   containing RIPA lysis buffer. After 30 min of lysis on ice, the mixture
   was centrifuged at 4°C at 12000 g for 10 min. The resulting supernatant
   was collected, and the protein concentration in exosomes was measured
   using the Micro BCA protein detection kit (Thermo Fisher Scientific,
   California, USA). The remaining proteins were added to 5 × SDS loading
   buffer and heated at 99°C for 5 min. Subsequently, 20 μg of protein was
   loaded onto a 0.2 µm PVDF membrane following the manufacturer’s
   protocol. The loading process involved applying a voltage of 80 V for
   30 min, followed by 120 V for 1 h. The membranes were then blocked with
   5% skim milk at room temperature for 1 h and incubated overnight at 4°C
   with primary antibodies targeting CD63, TSG101, and Calnexin (Abcam,
   UK) ([70]21). They were washed four times with 1× TBST for 15 minutes
   each and incubated with the appropriate secondary antibody for 1 h at
   room temperature. Finally, the membranes were washed three times with
   TBST for 10 min each, the ECL exposure solution was applied to the
   film, and pictures were taken.

RNA extraction and RT-qPCR

   Total RNAs were extracted from serum exosomes using the Trizol reagent
   (Invitrogen, USA) and were dissolved in water treated by
   diethylpyrocarbonate (DEPC). Briefly, Trizol is a phenol-guanidine
   isothiocyanate solution that effectively lyses biological material and
   denatures proteins ([71]31). It is designed to preserve RNA integrity.
   After adding chloroform and separating the phases, proteins are
   extracted in the organic phase, DNA is separated at the interface, and
   RNA is left in the aqueous phase. The aqueous phase can be carefully
   aspirated and isopropanol can be added to precipitate the RNA. The
   content and purity of acquired RNA were detected by OneDrop-2000 (Nano
   Drop Technologies) miRNAs were reverse transcribed into cDNA using
   miRNA 1st Strand cDNA Synthesis Kit (Vazyme Biotech). The quantitative
   RT-qPCR reaction was performed in 96-well plates with the miRNA
   Universal SYBR qPCR Master Mix (Vazyme Biotech). The primer details for
   the 10 candidate miRNAs can be found in [72]Supplementary Table 1 .

Pathway analysis

   For a more in-depth insight into the function of the target genes of
   miRNAs, we conducted gene ontology (GO) and annotations Kyoto
   Encyclopedia of Genes and Genomes (KEGG). Predictions of miRNAs for LN
   were performed using the DAVID database
   ([73]http://david.ncifcrf.gov/), which integrates biological data and
   analysis tools to offer detailed information on the functional
   annotations of genes/proteins. To be significant the P-value (P < 0.05)
   was considered.

Statistical analysis

   Continuous variables were represented as the median (interquartile
   range [IQR]) based on the normality test. The Spearman rank test was
   used to compare miRNA expression and clinical variables between two
   groups. Receiver operator characteristic (ROC) analysis was performed
   to calculate the area under the ROC curves (AUC), which evaluated the
   diagnostic efficiency of the candidate miRNAs. Statistical analyses
   were conducted using SPSS software, version 20.0 (SPSS Inc., Chicago,
   IL, USA), and GraphPad Prism version 9 (GraphPad Software Inc., La
   Jolla, CA, USA). The Mann-Whitney U test (*P < 0.05) was used to
   determine statistical significance. SPSS binary logistic regression was
   employed to predict the probability of joint diagnosis. The correlation
   analysis data were analyzed using Spearman rank correlation analysis.

Data and materials availability

   miRNA sequencing data, deposited in GEO under [74]GSE179950. The raw
   data is available upon request from the corresponding author, providing
   a valuable resource for further research.

Results

Identification and characterization of serum exosomes

   The morphology of serum exosomes was verified using TEM and the
   particle size using NTA. The results showed the presence of elliptical
   or bowl-shaped particles with an average diameter of 127 nm ([75]
   Figures 2A, B ). Western blot analysis also revealed the presence of
   typical EV proteins such as CD63 and TSG101 and absence of calnexin
   ([76] Figure 2C ). These findings demonstrate the existence of exosomes
   in serum, providing a basis for further investigation.

Figure 2.

   [77]Figure 2
   [78]Open in a new tab

   Characteristics of serum exosomes. (A) Morphology of vesicles under
   transmission electron microscopy. (B) Particle size distribution of
   vesicles measured by Nanoparticle Tracking Analysis. (C) Western
   blotting revealed the presence of exosomal biomarkers such as CD63, and
   TSG101.

miRNA profiles of serum exosomes from LN patients

   To determine the miRNA profile in LN patients, all participants were
   divided into two groups according to the clinical standard, including
   SLE with LN and SLE without LN. Exosomes were extracted from the mixed
   serum of 20 LN patients and 20 SLE patients without LN, and miRNA
   sequencing was performed. The distribution characteristics and Venn
   analyses of miRNA revealed a difference between LN and SLE without LN
   ([79] Figures 3A, B ). In comparison to SLE patients without LN, 382
   upregulated and 350 downregulated miRNAs were observed in LN, which met
   the criteria for sequencing detection of log2 fold-change > 2 in
   scatter plot analysis ([80] Figure 3C ). As shown in [81]Figure 3D ,
   the top 10 upregulated miRNAs visualized via hierarchical clustering
   were investigated as candidate markers of LN.

Figure 3.

   [82]Figure 3
   [83]Open in a new tab

   Analysis of differentially expressed miRNAs in serum exosomes of LN
   patients. (A) Venn diagram of serum exosomes derived miRNAs in LNs and
   SLE without LN patients. (B) The profiles of various length of miRNAs
   in serum exosomes between the two cohorts. (C) Scatter plots of
   differentially expressed miRNAs. Red and green dots indicated
   upregulated and downregulated miRNAs (log2 fold change > 2 between the
   two compared cohorts), and black dots indicated non-differentially
   expressed miRNAs. (D) Hierarchical clustering indicated the profiles of
   top 10 upregulated and downregulated miRNAs between two cohorts.

Validation of RNA sequencing by RT-qPCR in the training phase

   To verify the results of sequencing, a total of 24 pairs of subjects
   (24 LN and 24 SLE without LN patients) were recruited in the training
   phase, employing the RT-qPCR method. To assess the reliability and
   repeatability of this method, a standard curve of different
   concentrations of synthetic miRNA was constructed to determine whether
   there was a linear relationship between 1 pmol/L and 10 nmol/L ([84]
   Figure 4A ). Out of the top ten upregulated miRNAs, five miRNAs were
   successfully amplified using the RT-qPCR method ([85] Figures 4B–F ).
   According to [86]Figure 4 , the absolute quantitative results indicate
   a significant increase in three serum exosomal miRNAs (hsa-miR-638,
   hsa-miR-4796-5p, and hsa-miR-7974) in LN patients compared to SLE
   patients without LN.

Figure 4.

   [87]Figure 4
   [88]Open in a new tab

   Identification of differentially expressed serum exosomes derived
   miRNAs in LNs and SLE without LN patients. (A) Linear standard curve of
   serum exosomes derived miRNAs concentration. (B–F) Differential
   expression of 5 miRNAs verified by RT-qPCR in LN and SLE without LN
   patients. hsa-miR-638, hsa-miR-4796-5p and hsa-miR-7974 were
   significantly upregulated in LNs compared with SLE without LN patients.
   P value of the Mann-Whitney U test: (*P < 0.05, **P < 0.01, ****P <
   0.0001; ns, no significant difference).

Validation of RNA sequencing by RT-qPCR in the validation phase

   To further verify the above results, 144 subjects (72 LN and 72 SLE
   without LN patients) were recruited in the validation phase. According
   to [89]Figures 5A-C , hsa-miR-4796-5p and hsa-miR-7974 are
   significantly elevated in LN compared with SLE without LN, while there
   is no difference in the level of hsa-miR-638 between the two cohorts.
   The diagnostic performance of two miRNAs, hsa-miR-4796-5p and
   hsa-miR-7974, was evaluated using ROC analysis. The area under the
   curve (AUC) values for hsa-miR-4796-5p and hsa-miR-7974 were 0.703 (95%
   CI: 0.6287-0.7767) and 0.757 (95% CI: 0.6766-0.8373), respectively.
   These AUC values were obtained for distinguishing between LN and SLE
   patients without LN, as shown in [90]Figures 5D, E . The AUC for the
   panel of these two miRNAs combined was 0.806 (95% CI: 0.7348 to
   0.8775), as demonstrated in [91]Figure 5F . Serological indicators,
   such as C1q and CREA, have been found to be closely associated with
   renal damage caused by SLE. However, our study suggests that their
   diagnostic value is limited. In order to improve the accuracy of
   predictions, we evaluated the combined use of these clinical indicators
   with a panel of two miRNAs. Encouragingly, the AUCs for the combined
   approach were 0.837 and 0.844, respectively ([92] Figures 5G–J ).
   Furthermore, our findings revealed a positive correlation between
   levels of hsa-miR-4796-5p and hsa-miR-7974 with markers such as 24-hour
   proteinuria, ACR, creatinine, urea nitrogen, and SLEDAI. Conversely, we
   observed a negative correlation with albumin (Alb), C1q, and vitamin D3
   levels ([93] Figure 5K , [94]Supplementary Table 2 ).

Figure 5.

   [95]Figure 5
   [96]Open in a new tab

   Diagnostic value of serum exosomes derived miRNAs in the validation
   phase. (A–C) Expression of 3 miRNAs in LNs compared with SLE without LN
   patients in large Samples. P value of the Mann-Whitney U test: (*P <
   0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; ns, no significant
   difference) (D) Receiver operator characteristic (ROC) curve of
   hsa-miR-4796-5p in distinguishing LNs from SLE without LN patients. (E)
   Receiver operator characteristic (ROC) curve of hsa-miR-7974 in
   distinguishing LNs from SLE without LN patients. (F) ROC combined
   diagnostic analyses of hsa-miR-4796-5p and hsa-miR-7974 in
   discriminating LNs from SLE without LN patients. (G, H) ROC curve
   analysis of a single C1q marker and the combination of C1q with
   hsa-miR-4796-5p and hsa-miR-7974 markers for distinguishing LNs from
   SLE without LN patients. (I, J) ROC curve analysis of a single CREA
   marker and the combination of CREA with hsa-miR-4796-5p and
   hsa-miR-7974 markers for distinguishing LNs from SLE without LN
   patients. (K) Correlation analyses between miRNAs and clinical
   variables. Red and blue color represent the positive correlation and
   negative correlation, and the depth of the color represents the degree
   of correlation. The presence of * indicates an absolute value of the
   correlation coefficient r >0.3 (*P < 0.05, **P < 0.01 Spearman rank
   correlation analysis).

Significance of hsa-miR-4796-5p and hsa-miR-7974 in clinical practice

   Autoimmune nephritis encompasses several types of kidney diseases,
   including immunoglobulin A nephropathy (IgAN), diabetic nephropathy
   (DN), anti-neutrophil cytoplasmic antibody (ANCA)-associated
   glomerulonephritis (GN), Henoch-Schonlein purpura nephritis (HSPN), and
   LN. In our hospital, the most prevalent types of autoimmune nephritis
   among patients are IgAN, DN, and LN. To investigate the specificity of
   miRNAs in diagnosing LN, we conducted experiments to evaluate the
   effectiveness of candidate exosomal miRNAs. A total of 20 patients with
   LN, 20 with IgAN, and 20 with DN were randomly enrolled in our study.
   Our findings revealed significantly higher levels of hsa-miR-4796-5p
   and hsa-miR-7974 in LN patients compared to those with IgAN and DN.
   ([97] Figures 6A, B ). Our findings reveal a noteworthy correlation
   between the levels of two specific miRNAs, hsa-miR-4796-5p and
   hsa-miR-7974, and the SLEDAI score. The correlation coefficients (r)
   and p-values for hsa-miR-4796-5p and hsa-miR-7974 are 0.423 (p=0.003)
   and 0.398 (p=0.005), respectively ([98] Supplementary Table S2 ).
   Notably, the high scoring group exhibited a substantial disparity in
   miRNA levels when compared to the low scoring group. ([99] Figures 6C,
   D ).

Figure 6.

   [100]Figure 6
   [101]Open in a new tab

   Significance of hsa-miR-4796-5p and hsa-miR-7974 in clinical practice.
   (A, B) Differential expression analysis of two miRNAs in different
   autoimmune nephritis diseases. Statistical significance was determined
   by the Mann-Whitney U test (*P < 0.05, **P < 0.01; ns, no significant
   difference). (C, D) Differential expression analysis of 2miRNAs in
   different LN severity groups. (0-4, no activity; 5-9, mild activity;
   ≥10, moderate to severe activity).

Enrichment analysis of hsa-miR-4796-5p and hsa-miR-7974

   To explore the potential functions of hsa-miR-4796-5p and hsa-miR-7974,
   we performed KEGG pathway enrichment analysis and Gene Ontology
   analysis. The KEGG results indicated that hsa-miR-4796-5p was
   associated with HSV-1 infection, the MAPK signaling pathway, and the
   mTOR signaling pathway ([102] Figure 7A ). The GO project of
   hsa-miR-4796-5p involves intracellular signal transduction,
   insulin-like growth factor receptor signaling pathway, protein
   phosphorylation, and other related processes ([103] Figure 7B ). KEGG
   pathway enrichment analysis of hsa-miR-7974 revealed the involvement of
   miRNAs in the MAPK pathway, PI3K-Akt signaling pathway, and endocytosis
   ([104] Figure 7C ). The analysis also identified enrichment of GO terms
   targeted by hsa-miR-7974, such as cellular responses to signal
   transduction, endocytosis, protein trafficking, and other GO strains
   ([105] Figure 7D ). These results serve as a reminder that miRNA may
   play an important role in the development of LN by impacting these
   functions.

Figure 7.

   [106]Figure 7
   [107]Open in a new tab

   Enrichment analysis of hsa-miR-4796-5p and hsa-miR-7974. (A, B) KEGG
   pathway enrichment analysis and GO terms analysis of hsa-miR-4796-5p.
   (C, D) KEGG pathway enrichment analysis and GO terms analysis of
   hsa-miR-7974. The size of bubbles represents the number of genes
   enriched in this pathway, and the color of bubbles represents
   significance. P value represents the significance degree of enrichment.

Discussion

   LN is a severe manifestation of SLE, with approximately 40% of patients
   developing chronic kidney disease and 5-20% progressing to end-stage
   kidney disease (ESKD) within 10 years of their initial SLE diagnosis
   ([108]32). Additionally, patients undergoing immunosuppressive
   treatment for LN may experience various complications such as
   infection, osteoporosis, cardiovascular issues, and reproductive system
   problems ([109]4). Clearly, timely diagnosis and accurate evaluation of
   LN are essential for enhancing outcomes in SLE patients. The commonly
   used diagnostic methods for LN in clinical practice include 24-hour
   proteinuria quantification and kidney biopsy. However, these methods
   have certain limitations. Urine samples may have inaccurate retention
   time, and there may be poor patient compliance for urine protein
   testing. Additionally, kidney biopsy, although considered a valuable
   diagnostic tool, is an invasive procedure that carries the risk of
   bleeding and is challenging to replicate. Consequently, there is an
   immediate priority to investigate new non-invasive biomarkers that can
   effectively differentiate between LN and SLE.

   Liquid biopsy is an innovative diagnostic method for analyzing
   biological material in blood and other bodily fluids to identify
   disease status ([110]33). In recent years, the detection of miRNAs has
   been increasingly utilized for studying various autoimmune diseases.
   MiRNA, a small single-stranded endogenous non-coding RNA, has the
   remarkable capability to efficiently and precisely suppress the
   expression of its targeted transcripts, leading to alterations in
   cellular epigenetics and playing a vital regulatory role in both the
   innate and adaptive immune systems ([111]34, [112]35). Current research
   on miRNAs in SLE primarily focuses on serum miRNAs and PBMC miRNAs,
   with limited studies on exosomal miRNAs ([113]36, [114]37). Exosomes, a
   specific type of extracellular vesicles (EVs), have been identified in
   majority of body fluids ([115]38–[116]42). Composed of a lipid bilayer,
   the extracellular surface of exosomes serves as a protective barrier,
   safeguarding their contents such as proteins, mRNA, miRNAs, and other
   non-coding RNAs (ncRNAs) from degradation ([117]18). When exosomes
   circulate, the RNA molecules they contain, especially miRNA, play a
   vital role in facilitating communication between different tissues
   through paracrine and endocrine pathways ([118]43). Due to their
   excellent stability and accessibility, exosomal miRNAs hold potential
   as non-invasive biomarkers ([119]44).

   Our previous research has identified upregulated tsRNAs in the urine
   exosomes of patients with LN, which have shown promise in
   distinguishing LN from SLE ([120]45). In this study, we aimed to
   analyze the levels of circulating exosomal miRNAs in serum to
   investigate their potential significance in LN. Previous reports have
   suggested that there may be differences in the miRNA content between
   blood plasma and serum ([121]46, [122]47). Liu et al. ([123]48)
   recommended the use of blood plasma in exosome research, as platelets
   also contain significant amounts of RNA that could be released into the
   serum during the coagulation process. However, a recent study ([124]49)
   discovered that plasma prepared by centrifugation contains platelets
   and ery-ghosts, which co-isolate with EVs. In this study, we initially
   conducted RNA sequencing using serum exosomes from patients with LN. We
   compared the results to SLE patients without LN and identified 382
   upregulated miRNAs in LN. Then, we selected the top 10 upregulated
   miRNAs as potential markers of LN. To validate the expression levels of
   these candidate exosomal-miRNAs, we used an RT-qPCR assay, which is
   more sensitive and not limited by sequence-abundance bias compared to
   the microarray profiling assay ([125]50). Our data demonstrated that
   the levels of exosomal hsa-miR-4796-5p and hsa-miR-7974 were
   significantly elevated in patients with LN compared to SLE patients
   without LN in both the training and validation phases. These two miRNAs
   demonstrated a significant ability in diagnosing LN in patients with
   SLE, with an Area Under the Curve (AUC) above 0.8. In order to
   demonstrate higher diagnostic value, we formed a comprehensive team to
   integrate these two miRNAs with clinical parameters (such as C1q and
   CREA). The area under curve (AUC) values for these miRNAs were found to
   be 0.837 and 0.844, respectively. The SLE Disease Activity Index
   (SLEDAI) is an important tool for evaluating SLE activity based on
   clinical symptoms and auxiliary examinations. After scoring the 192
   enrolled patients, we observed a significant upregulation of these two
   miRNAs in moderate to severe cases. Therefore, the exosomal
   hsa-miR-4796-5p and hsa-miR-7974 in serum have the potential to serve
   as biomarkers for evaluating disease activity.

   In order to assess the specificity of these miRNAs duo in LN, we
   conducted separate experiments in immunoglobulin A nephropathy (IgAN)
   and diabetic nephropathy (DN). Our findings revealed that the levels of
   these two miRNAs were significantly elevated in LN compared to IgAN and
   DN. Therefore, these miRNAs not only serve as potential biomarkers to
   distinguish LN in SLE, but also exhibit good specificity in other
   autoimmune kidney diseases. The KEGG results revealed that
   hsa-miR-4796-5p and hsa-miR-7974 were found to be associated with
   various signaling pathways, such as the mTOR signaling pathway and
   PI3K-Akt signaling pathway which is implicated in the pathogenesis of
   LN ([126]51–[127]53). GO analysis showed that hsa-miR-4796-5p is
   enriched in insulin-like growth factor binding, which influences
   autoimmunity by modulating signaling pathways relevant to Th17/Treg
   balance ([128]54). These results suggests that miRNAs investigated in
   this study may play a regulatory role in the progression of LN by
   engaging in these signaling pathways.

   To recap, we have identified two serum exosomal miRNAs signatures in
   patients with LN. Specifically, we have demonstrated that
   hsa-miR-4796-5p and hsa-miR-7974, which are derived from serum
   exosomes, have the potential to be valuable biomarkers for
   differentiating between LN and SLE patients, as well as the autoimmune
   nephritis group. Furthermore, our study has shed light on the potential
   biological functions of these novel serum exosomal miRNAs. These
   findings provide a foundation for future research to explore the
   clinical applications and deeper understanding of serum miRNAs.
   However, it is important to note that our study was limited by the
   small sample size and single-center experience. Therefore, randomized
   clinical trials are the next frontier in evaluating these two miRNA
   signatures for LN diagnosis and prognosis.

Data availability statement

   The data presented in the study are deposited in the NCBI repository,
   accession number PRJNA745976.

Ethics statement

   The studies involving humans were approved by The Ethics Committee of
   the Affiliated Drum Tower Hospital of Nanjing University Medical
   School. The studies were conducted in accordance with the local
   legislation and institutional requirements. The participants provided
   their written informed consent to participate in this study.

Author contributions

   FC: Data curation, Formal Analysis, Investigation, Methodology,
   Software, Writing – original draft. BS: Conceptualization, Data
   curation, Investigation, Methodology, Writing – original draft. WL:
   Conceptualization, Data curation, Investigation, Methodology, Software,
   Writing – original draft. JGo: Data curation, Methodology, Writing –
   original draft. JGa: Data curation, Investigation, Methodology, Writing
   – original draft. YS: Methodology, Writing – original draft. PY:
   Funding acquisition, Project administration, Supervision, Writing –
   review & editing. ZL: Funding acquisition, Resources, Supervision,
   Conceptualization, Writing – review & editing.

Acknowledgments