ABSTRACT
Aims/Introduction
The impact of rare pathogenic variants on diabetic kidney disease (DKD) has not been investigated in detail. Previous studies have detected pathogenic variants in 22% of Caucasian patients with DKD; however, this proportion may vary depending on ethnicity and updates to the database. Therefore, we performed a whole‐genome analysis of patients with DKD in type 2 diabetes mellitus in Japan, utilizing a recent database to investigate the prevalence of kidney‐related pathogenic variants and describe the characteristics of these patients.
Materials and methods
Whole‐genome sequencing was performed, and variants were analyzed following the GATK Best Practices. We extracted data on 790 genes associated with Mendelian kidney and genitourinary diseases. Pathogenic variants were defined based on the American College of Medical Genetics criteria, including both heterozygous and homozygous variants classified as pathogenic or likely pathogenic.
Results
Among 79 participants, heterozygous pathogenic variants were identified in 27 (34.1%), a higher prevalence than previously reported. No homozygous pathogenic variants were detected. The identified heterozygous pathogenic variants were roughly divided into 23.7% related to glomerulopathy, 36.8% related to tubulointerstitial disease, 10.5% related to cystic disease/ciliopathy, and 28.9% related to others. Diagnostic variants were found in 10 patients (12.7%) in seven genes (ABCC6, ALPL, ASXL1, BMPR2, GCM2, PAX2, and WT1), all associated with autosomal dominant congenital disease.
Conclusions
This study identified a considerable number of patients with DKD in Japan who carried kidney‐related heterozygous pathogenic variants. These findings suggest potential ethnic differences and highlight the impact of database updates on variant detection.
Keywords: Diabetic kidney disease, Rare pathogenic variant, Whole‐genome sequence
Previous studies have identified pathogenic variants in 22% of Caucasian patients with diabetic kidney disease (DKD); however, this proportion may vary depending on ethnicity and updates to the database. Whole‐genome sequencing of 79 patients with DKD in Japan revealed that 34.1% had kidney‐related heterozygous pathogenic variants, and 12.7% had diagnostic variants for autosomal dominant congenital diseases, a higher prevalence than reported previously. These findings highlight ethnic differences and the importance of database updates in variant detection.
INTRODUCTION
The number of patients with diabetic kidney disease (DKD), a long‐term complication of type 2 diabetes mellitus (DM) that is associated with high cardiovascular complication and mortality rates, is increasing worldwide 1 , 2 . DKD can be clinically diagnosed based on prolonged diabetes duration and the presence of diabetic retinopathy, without the need for a kidney biopsy 3 . Since some patients with DM develop kidney dysfunction while others do not, despite similar glycemic control, genetic factors may underlie the development of DM and DKD 4 . In particular, genome‐wide association studies have identified several single‐nucleotide polymorphisms (SNPs) associated with the etiology of type 2 DM and DKD 5 , 6 , 7 . However, there is a problem known as “missing heritability,” wherein common variants such as SNPs can only explain a small fraction of the predicted disease risk 8 .
The impact of rare variants has gained attention as a potential explanation for missing heritability, and advances in next‐generation sequencing technology have enabled detailed studies of these variants. In 2019, Groopman et al. studied 370 patients with diabetic nephropathy in the United States and found that 6 (1.6%) carried diagnostic variants with an autosomal dominant (AD) inheritance pattern, which may directly influence disease development 9 . In 2021, a study on rare variants identified pathogenic variants in 22.0% of Caucasian patients with DKD in type 1 and 2 DM, including diagnostic variants found in 2.8% 10 . More recently, in 2023, another study in the United States reported that 18 (8.7%) of 206 patients with nephropathy associated with diabetes mellitus tested positive for diagnostic variants 11 . As such, the continuous evolution of genomic data may result in changes to the number of known relevant variants over time, requiring repeated updates to the information and diagnoses 12 .
In addition, the frequency of genetic variants considerably differs based on ethnicity 12 , 13 . For example, a rare variant of APOL1 is considered a genetic risk factor for CKD 13 ; however, this variant is extremely rare in East Asia and has almost never been observed 14 , 15 . In Japan, diagnostic variants have been identified in 11% of adult patients undergoing hemodialysis for unknown primary diseases 16 . Although genome‐wide association studies have identified genes linked to diabetic retinopathy and diabetic nephropathy, no studies in Japan have investigated rare pathogenic variants associated with the missing heritability of DKD 17 , 18 .
We performed whole‐genome sequencing (WGS) on samples from patients with DKD in type 2 DM in Japan, utilizing a recent database to determine the proportion of patients with pathogenic variants and assess their clinical characteristics.
MATERIALS AND METHODS
Study participants
Patients from the University of Tokyo Diabetic Kidney Disease (UT‐DKD) cohort who consented to provide samples for WGS were included in this study. The details of the UT‐DKD cohort have been described in previous reports 19 , 20 . Briefly, the UT‐DKD cohort consisted of patients with CKD stages G3a/G3b who have type 2 DM and no apparent kidney diseases other than DKD, regardless of the presence of albuminuria. All participants in this study were residents of Japan and regularly visited our hospital. We did not collect data on their ethnic origin.
Baseline data collected at the start of enrollment included age, sex, body mass index, history of hypertension, ischemic heart disease, diabetic retinopathy, family history of DM and kidney disease, duration of DM, and use of angiotensin‐converting enzyme inhibitors or angiotensin II receptor blockers. Additionally, hemoglobin A1c, uric acid levels, estimated glomerular filtration rate (eGFR), urine sediment findings, and urine albumin‐to‐creatinine ratio were assessed. A family history of DM and kidney disease was defined as having affected relatives within the second degree of kinship, based on information obtained through patient interviews. When albuminuria data were unavailable, proteinuria data were used instead, with the protein‐to‐creatinine ratio converted to an albumin‐to‐creatinine ratio 21 . The Japanese Modification of Diet in Renal Disease (MDRD) equation was used to calculate eGFR 22 . All eGFR data were collected for 30 months from the baseline visit, and the annual eGFR decline for each patient was calculated every 10 months (i.e., at four time points) using the least squares method 19 .
For patients with available abdominal ultrasound or CT scan data, these imaging modalities were used to assess the presence of kidney cysts, ureteral dilation, and kidney stones. Additionally, medical records were reviewed to confirm whether patients had undergone a renal biopsy, and if so, the pathology results were verified.
Analysis methods
WGS was performed using 150 bp paired‐end reads on an Illumina HiSeq X ten instrument (Illumina, San Diego, CA, USA), and variants were processed following GATK Best Practices (detailed in the Data S1: Supplementary Methods). The GRCh38 reference genome was used. Variant annotation was performed using ANNOVAR (ClinVar database, December 31, 2022) 23 . We extracted data on 790 genes associated with Mendelian kidney and genitourinary diseases, as previously reported worldwide (Table S1) 9 , 10 , 24 , 25 , 26 , 27 , 28 . Among these, variants classified as pathogenic or likely pathogenic based on the American College of Medical Genetics guidelines were defined as pathogenic variants 29 . For each gene, the analyzed region spanned from 1,000 bp upstream of the transcription start site, including the promoter region, to the end of the gene 30 , 31 . Sanger sequencing was performed to verify variants that met any of the following conditions, where WGS alone was considered insufficient to confirm their existence: FILTER not equal to PASS, QUAL < 100, coverage < 20×, and variant fraction < 20% 32 . Allele frequencies of the identified pathogenic variants were confirmed using the gnomAD 33 (https://gnomad.broadinstitute.org) and jMorp 34 (https://jmorp.megabank.tohoku.ac.jp) databases.
Statistical analysis
Categorical variables are presented as n (%) and were compared using Fisher's exact test. Continuous variables are expressed as mean ± SD or median (interquartile range), as appropriate. To compare continuous variables, Student's t‐test was used for normally distributed data, while the Mann–Whitney U‐test was applied for skewed distributions. In this study, the outcome was defined as the presence of a pathogenic variant to determine whether a patient harbored such variants based on clinical findings.
We also analyzed the number of pathogenic variants in each patient and compared the clinical features among three groups: 0, 1, and ≥2 variants. For continuous variables, anova was used for normally distributed data, while the Kruskal‐Wallis test was applied for skewed distributions. Categorical variables were compared using the chi‐squared test.
Multivariate logistic regression analysis was performed to assess the outcome, adjusting for age, sex, duration of DM, absence of diabetic retinopathy, and presence of a family history of kidney disease. Previous research has indicated that the absence of diabetic retinopathy and the presence of a family history of kidney disease are significant indicators of kidney diseases other than DKD 9 , 35 . P‐values < 0.05 were considered statistically significant. All statistical tests were performed using JMP 17.0.0 (SAS Institute Inc., Cary, North Carolina, USA).
Ethics
This study was approved by the Ethics Committee of the University of Tokyo Graduate School of Medicine (ethical approval number: G10132) and was conducted in accordance with the Declaration of Helsinki and institutional guidelines. Written informed consent was obtained from all participants.
RESULTS
Seventy‐nine patients who provided informed consent and underwent whole‐genome analysis were included in this study. Renal biopsies were performed on 2 out of the 79 patients, both of whom were diagnosed with diabetic nephropathy based on histological findings. The average age of the participants was 72 years [interquartile range: 66–76]. Of the total, 25 patients (31.6%) had diabetic retinopathy, and 9 (11.4%) had a family history of kidney disease (Table 1). The mean coverage of variants obtained using WGS was 16.0 ± 0.3×.
Table 1.
Comparison of clinical features between cases with and without heterozygous pathogenic variants
| Variable | All | Heterozygous pathogenic variant (−) | Heterozygous pathogenic variant (+) | P‐value |
|---|---|---|---|---|
| n = 79 | n = 52 | n = 27 | ||
| Age (year) | 72 (66–76) | 71 (66–76) | 73 (68–76) | 0.790 |
| Male, n (%) | 66 (83.5) | 43 (82.7) | 23 (85.2) | 1.000 |
| Body Mass Index (kg/m2) | 25.3 ± 3.3 | 25.3 ± 3.5 | 25.4 ± 2.8 | 0.940 |
| HbA1c (%) | 6.9 ± 0.6 | 6.8 ± 0.5 | 6.9 ± 0.7 | 0.475 |
| Uric acid (mg/dL) | 6.2 ± 1.0 | 6.3 ± 0.8 | 6.1 ± 1.3 | 0.490 |
| Duration of diabetes mellitus (year) | 19.6 ± 9.4 | 18.2 ± 9.1 | 22.3 ± 9.5 | 0.061 |
| Hypertension, n (%) | 67 (84.8) | 44 (84.6) | 23 (85.2) | 1.000 |
| Ischemic heart disease, n (%) | 17 (21.5) | 12 (23.1) | 5 (18.5) | 0.776 |
| Diabetic retinopathy, n (%) | 25 (31.6) | 15 (28.9) | 10 (37.0) | 0.459 |
| ACE inhibitor/ARB user, n (%) | 57 (72.2) | 39 (75.0) | 18 (66.7) | 0.441 |
| Family history, n (%) | ||||
| Diabetes mellitus | 39 (49.4) | 25 (48.1) | 14 (51.9) | 0.815 |
| Kidney disease | 9 (11.4) | 3 (5.8) | 6 (22.2) | 0.056 |
| eGFR (mL/min/1.73m2) | 47.2 ± 7.9 | 47.4 ± 7.7 | 47.0 ± 8.3 | 0.841 |
| Occult blood in urine | 18 (22.8) | 12 (23.1) | 6 (22.2) | 1.000 |
| Albumin‐to‐creatinine ratio (mg/gCr) | 41 (8–166) | 42 (7–225) | 38 (9–153) | 1.000 |
| eGFR slope (mL/min/1.73m2/year) | −0.3 (−1.8–1.2) | −0.1 (−1.8–1.5) | −0.8 (−1.8–0.3) | 0.267 |
Values are expressed as mean ± SD or median (interquartile range). Since 19 patients had proteinuria data but not albuminuria, the urine protein‐to‐creatinine ratio was converted to the urine albumin‐to‐creatinine ratio.
ACE, angiotensin‐converting enzyme; ARB, angiotensin II receptor blocker; eGFR, estimated glomerular filtration rate; HbA1c, hemoglobin A1c.
Pathogenic variants located in 24 genes (29 positions) were identified in 27 patients (34.1%; Figure 1; Table 2). All variants were heterozygous single‐nucleotide variants (SNVs), and no homozygous or autosomal recessive compound heterozygous SNVs were detected. As a result, in this study, the term “pathogenic variants” specifically refers to heterozygous pathogenic variants. Of the 29 SNVs, 17 had low coverage, eight were confirmed by Sanger sequencing, and nine could not be confirmed due to insufficient DNA samples or technical difficulties. Therefore, these SNVs are listed as unverified in Table 2. Among the 29 SNVs, 26 (89.7%) exhibited higher minor allele frequencies in the Japanese population (ToMMo 54kJPN) 34 than in international databases (gnomAD) 33 .
Figure 1.
Flow chart of study design and the results of whole‐genome sequence analysis. This figure presents a flow chart illustrating the participant selection process in the UT‐DKD cohort of 79 patients with diabetes mellitus and a pipeline of whole‐genome sequence analysis methods. Whole‐genome sequences were narrowed down using a gene list including 790 genes associated with Mendelian forms of kidney and genitourinary diseases, and the characteristics of participants with and without heterozygous pathogenic variants were compared. DKD, diabetic kidney disease; UT‐DKD, University of Tokyo Diabetic Kidney Disease.
Table 2.
The characteristics of all heterozygous pathogenic variants
| rs number | Chr | Postion (Hg38) | Gene name | Ref/Alt | Impact | Code | Verification | ACMG criteria | Allele frequency | |
|---|---|---|---|---|---|---|---|---|---|---|
| ToMMo † (54kJPN) | gnomAD ‡ (Total) | |||||||||
| rs121918010 | 1 | 21,573,781 | ALPL | T/C | Exonic | c.748T>C:p.F250L, c.814T>C:p.F272L, c.979T>C:p.F327L | Yes | P/LP | 0.00277 | 0.00005 |
| rs753886165 | 1 | 215,758,592 | USH2A | T/A | Intronic | c.11389+3A>T | Yes | P/LP | 0.00001 | <0.00001 |
| rs397518039 | 1 | 215,877,882 | USH2A | T/C | Splicing | c.8559‐2A>G | No | P | 0.00011 | 0.00001 |
| rs140049204 | 2 | 202,552,887 | BMPR2 | C/T | Exonic, Splicing | c.1585C>T:p.R529C | No | LP | NA | 0.00005 |
| rs727503968 | 3 | 121,790,112 | IQCB1 | G/A | Exonic | c.691C>T:p.R231*, c.1090C>T:p.R364* | Yes | P | 0.00015 | 0.00003 |
| rs146395801 | 6 | 10,874,407 | GCM2 | G/A | Exonic | c.1109C>T:p.T370M | Yes | LP | 0.00166 | 0.00009 |
| rs1583446897 | 6 | 51,934,296 | PKHD1 | C/T | Exonic | c.5935G>A:p.G1979R | Yes | P/LP | 0.00021 | <0.00001 |
| rs199568593 | 6 | 52,046,089 | PKHD1 | A/G | Exonic | c.2507T>C:p.V836A | Yes | P/LP | 0.00621 | 0.00011 |
| rs113993993 | 7 | 66,994,210 | SBDS | A/G | Splicing | c.258+2T>C | Yes | P/LP | 0.00390 | 0.00363 |
| rs120074160 | 7 | 66,994,286 | SBDS | T/A | Exonic | c.184A>T:p.K62* | Yes | P/LP | 0.00123 | 0.00053 |
| rs111033220 | 7 | 107,690,203 | SLC26A4 | C/T | Exonic | c.1229C>T:p.T410M | Yes | P | 0.00042 | 0.00015 |
| rs375415632 | 9 | 133,424,479 | ADAMTS13 | G/A | Splicing | c.330+1G>A | Yes | LP | 0.00004 | <0.00001 |
| rs201021899 | 10 | 100,827,045 | PAX2 | A/C | Exonic | c.1058A>C:p.Q353P, c.1151A>C:p.Q384P, c.1127A>C:p.Q376P, c.1141A>C:p.S381R | No | LP | 0.00581 | 0.00012 |
| rs767846762 | 11 | 103,176,241 | DYNC2H1 | AA/− | Exonic | c.5681_5682del:p.H1896Yfs*9 | Yes | P | 0.00039 | 0.00001 |
| rs373745258 | 11 | 128,839,972 | KCNJ1 | G/A | Exonic | c.329C>T:p.P110L, c.272C>T:p.P91L, c.323C>T:p.P108L | Yes | LP | NA | 0.00003 |
| rs121907892 | 11 | 64,593,747 | SLC22A12 | G/A | Exonic | c.672G>A:p.W224*, c.774G>A:p.W258*, c.111G>A:p.W37* | Yes | P | 0.02224 | 0.00038 |
| c.672G>A:p.W224*, c.774G>A:p.W258*, c.111G>A:p.W37* | Yes | P | 0.02224 | 0.00038 | ||||||
| c.672G>A:p.W224*, c.774G>A:p.W258*, c.111G>A:p.W37* | Yes | P | 0.02224 | 0.00038 | ||||||
| c.672G>A:p.W224*, c.774G>A:p.W258*, c.111G>A:p.W37* | Yes | P | 0.02224 | 0.00038 | ||||||
| c.672G>A:p.W224*, c.774G>A:p.W258*, c.111G>A:p.W37* | Yes | P | 0.02224 | 0.00038 | ||||||
| rs1250438314 | 11 | 32,391,999 | WT1 | G/A | Exonic | c.232C>T:p.H78Y, c.1369C>T:p.H457Y, c.718C>T:p.H240Y, c.769C>T:p.H257Y, c.1420C>T:p.H474Y | No | LP | 0.00001 | 0.00012 |
| rs770829226 | 13 | 51,965,038 | ATP7B | A/C | Intronic | c.1708‐5T>G | Yes | P/LP | 0.00018 | 0.00001 |
| rs191759494 | 15 | 45,108,159 | DUOX2 | C/T | Exonic | c.1462G>A:p.G488R | Yes | P/LP | 0.00309 | 0.00011 |
| rs67867306 | 16 | 16,177,500 | ABCC6 | T/− | Exonic | c.2542delA:p.M848Cfs*83, c.2200delA:p.V734Cfs*83 | Yes | P/LP | 0.01408 | 0.00020 |
| c.2542delA:p.M848Cfs*83, c.2200delA:p.V734Cfs*83 | No | P/LP | 0.01408 | 0.00020 | ||||||
| c.2542delA:p.M848Cfs*83, c.2200delA:p.V734Cfs*83 | No | P/LP | 0.01408 | 0.00020 | ||||||
| rs72650699 | 16 | 16,202,045 | ABCC6 | G/A | Exonic | c.1132C>T:p.Q378*, c.790C>T:p.Q264* | No | P | 0.00214 | 0.00006 |
| rs145583876 | 16 | 89,102,626 | ACSF3 | G/A | Exonic | c.689G>A:p.W230* | Yes | P | 0.00582 | 0.00010 |
| rs28999113 | 16 | 88,809,834 | APRT | A/G | Exonic | c.407T>C:p.M136T | Yes | P | 0.00173 | 0.00004 |
| rs104894507 | 16 | 88,810,450 | APRT | C/T | Exonic | c.294G>A:p.W98* | Yes | P | 0.00020 | 0.00001 |
| rs756128712 | 17 | 2,033,837 | DPH1 | G/− | Exonic | c.288delG:p.E97Kfs*8, c.273delG:p.E92Kfs*8 | Yes | P | 0.00505 | 0.00006 |
| rs80356484 | 17 | 42,911,000 | G6PC1 | G/T | Exonic | c.648G>T:p.L216L | Yes | P/LP | 0.00156 | 0.00004 |
| rs387906978 | 17 | 8,121,693 | HES7 | C/A | Exonic | c.571G>T:p.D191Y, c.556G>T:p.D186Y | Yes | P | 0.00031 | 0.00004 |
| rs766433101 | 20 | 32,434,600 | ASXL1 | CACCACTGCCATAGAGAGGCGGC/− | Exonic | c.1705_1727del:p.E574Rfs*15, c.1888_1910del:p.E635Rfs*15 | Yes | P/LP | 0.00010 | 0.00008 |
| rs202119339 | 20 | 62,314,320 | LAMA5 | C/T | Exonic | c.8488G>A:p.A2830T | Yes | LP | 0.01548 | 0.00022 |
| c.8488G>A:p.A2830T | Yes | LP | 0.01548 | 0.00022 | ||||||
| c.8488G>A:p.A2830T | No | LP | 0.01548 | 0.00022 | ||||||
| c.8488G>A:p.A2830T | No | LP | 0.01548 | 0.00022 | ||||||
ACMG, American College of Medical Genetics and Genomics; Alt, alternative; Chr, chromosome; LP, likely pathogenic; NA, not available; P, pathogenic; Ref, reference; ToMMo, Tohoku Medical Megabank Organization.
Allele frequency panel of 69,000 Japanese individuals from the Tohoku Medical Megabank Organization 34 .
Genome Aggregation Database v4.1.0 33 .
The gene names of the detected heterozygous pathogenic variants are shown in Figure 2 and classified into the following four disease categories based on OMIM (https://www.omim.org) and the Protein Atlas (https://www.proteinatlas.org): glomerulopathy (23.7%), tubulointerstitial (36.8%), cystic or ciliopathy (10.5%), and others (28.9%) 36 . Among these, frameshift variants were found in ABCC6 (N = 3), ASXL1 (N = 1), DPH1 (N = 1), and DYNC2H1 (N = 1). Stop‐gain variants were detected in ABCC6 (N = 1), ACSF3 (N = 1), APRT (N = 1), IQCB1 (N = 1), SBDS (N = 1), and SLC22A12 (N = 5) (Table S2).
Figure 2.
Characteristics of heterozygous pathogenic variants. The gene names of the detected heterozygous pathogenic variants are shown, along with their classification into four categories (glomerulopathy, tubulointerstitial, cystic or ciliopathy, and others). The number of corresponding variants detected is indicated in parentheses. Stripes highlight diagnostic variants known to cause congenital diseases with autosomal dominant inheritance patterns.
Details of the 27 patients with heterozygous pathogenic variants are summarized in Table S3. Among these variants, those associated with an AD inheritance pattern and potentially influencing disease onset were defined as “diagnostic variants”. Ten patients (12.7%) had diagnostic variants in seven genes: ABCC6, ALPL, ASXL1, BMPR2, GCM2, PAX2, and WT1, all of which are associated with AD congenital diseases (Tables 3 and S3). None of the 10 patients underwent a renal biopsy. Of these 10 patients, three (30.0%) had diabetic retinopathy. Two patients (20.0%) had a family history of kidney disease, similar to the proportion of patients with heterozygous pathogenic variants overall (N = 6, 22.2%). Additionally, none of the patients had more than five kidney cysts, dilated ureters, or ureteral stones. Two patients tested positive for occult blood in their urine, but their urine sediment contained fewer than five red blood cells per high‐power field, and they were not diagnosed with hematuria. A comparison of clinical features between patients with and without diagnostic variants revealed no significant differences, except for uric acid levels (Table S4).
Table 3.
The characteristics of patients with diagnostic variants causing congenital diseases exhibit an autosomal dominant inheritance pattern
| Patient No. | Age | Sex | rs number | Gene name | Code | Category | Manner of inheritance | Diabetic retinopathy | Family history of kidney disease | Kidney cyst (≥5) | Ureteral dilation | Kidney stone | Occult blood in urine |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 77 | M | rs121918010 | ALPL | c.748T>C:p.F250L, c.814T>C:p.F272L, c.979T>C:p.F327L | Tubulointerstitial | AD, AR | Yes | No | No data | No data | No data | No |
| rs145583876 | ACSF3 | c.689G>A:p.W230*, c.689G>A:p.W230* | Other | AR | |||||||||
| 2 | 68 | M | rs140049204 | BMPR2 | c.1585C>T:p.R529C | Tubulointerstitial | AD | No | No | No | No | No | No |
| rs1583446897 | PKHD1 | c.5935G>A:p.G1979R | Cystic / Cilliopathy | AR | |||||||||
| rs120074160 | SBDS | c.184A>T:p.K62* | Other | AR | |||||||||
| 3 | 77 | M | rs146395801 | GCM2 | c.1109C>T:p.T370M | Other | AD, AR | No | Yes | No | No | No | No |
| 4 | 76 | M | rs201021899 | PAX2 | c.1058A>C:p.Q353P, c.1151A>C:p.Q384P, c.1127A>C:p.Q376P, c.1141A>C:p.S381R | Glomerulopathy | AD | No | No | No | No | No | No |
| 5 | 74 | M | rs1250438314 | WT1 |
c.232C>T:p.H78Y, c.1369C>T:p.H457Y, c.718C>T:p.H240Y, c.769C>T:p.H257Y, c.1420C>T:p.H474Y |
Glomerulopathy | AD, SMu | No | No | No | No | No |
Yes |
| rs202119339 | LAMA5 | c.8488G>A:p.A2830T | Glomerulopathy | AR | |||||||||
| 6 | 64 | M | rs767846762 | DYNC2H1 | c.5681_5682del:p.H1896Yfs*9 | Cystic / Cilliopathy | AR, DR | No | No | No | No | No | No |
| rs67867306 | ABCC6 | c.2542delA:p.M848Cfs*83, c.2200delA:p.V734Cfs*83 | Other | AD, AR | |||||||||
| rs202119339 | LAMA5 | c.8488G>A:p.A2830T | Glomerulopathy | AR | |||||||||
| 7 | 58 | M | rs67867306 | ABCC6 | c.2542delA:p.M848Cfs*83, c.2200delA:p.V734Cfs*83 | Other | AD, AR | Yes | No | No data | No data | No data | Yes |
| 8 | 73 | M | rs199568593 | PKHD1 | c.2507T>C:p.V836A | Cystic / Cilliopathy | AR | No | Yes | No data | No data | No data | No |
| rs67867306 | ABCC6 | c.2542delA:p.M848Cfs*83, c.2200delA:p.V734Cfs*83 | Other | AD, AR | |||||||||
| 9 | 77 | F | rs72650699 | ABCC6 | c.1132C>T:p.Q378*, c.790C>T:p.Q264* | Other | AD, AR | No | No | No data | No data | No data | No |
| rs202119339 | LAMA5 | c.8488G>A:p.A2830T | Glomerulopathy | AR | |||||||||
| 10 | 72 | M | rs766433101 | ASXL1 | c.1705_1727del:p.E574Rfs*15, c.1888_1910del:p.E635Rfs*15 | Glomerulopathy | AD | Yes | No | No | No | No | No |
Variants reported to cause congenital diseases with an autosomal dominant inheritance pattern are shown in boldface. AD, autosomal dominant; AR, autosomal recessive; DR, digenic recessive; F, female; M, male; sMu, somatic mutation.
We compared the clinical features of patients with and without heterozygous pathogenic variants. Univariate analysis showed no significant differences in clinical parameters, including albuminuria levels, baseline eGFR, and duration of DM, between patients with and without heterozygous pathogenic variants, although patients with variants tended to have a family history of kidney disease (with variants, N = 6 (22.2%); without variants, N = 3 (5.8%); P = 0.056; Table 1). In the multivariate analysis, the duration of DM, family history of kidney disease, and the absence of diabetic retinopathy were not significantly associated with the presence of heterozygous pathogenic variants (adjusted odds ratio [95% confidence interval]: 1.04 [0.98–1.10], 4.28 [0.89–20.44], and 0.76 [0.26–2.19], respectively; Table S5).
Participants were categorized into three groups based on the number of heterozygous pathogenic variants: 0 (N = 52), 1 (N = 18), and ≥2 (N = 9), and their clinical characteristics were compared (Table S6). The duration of DM and a family history of kidney disease were significantly associated with the number of heterozygous pathogenic variants (Table S6, P = 0.009 and P = 0.040, respectively). However, no clear trend was found linking an increasing number of variants to longer DM duration or a higher proportion of individuals with a family history of kidney disease. There was also no statistically significant association between the number of heterozygous pathogenic variants and the urine albumin‐to‐creatinine ratio or eGFR slope.
DISCUSSION
For the first time, we report the prevalence of rare heterozygous pathogenic variants in patients with DKD in type 2 DM in Japan based on whole‐genome analysis. Among the 79 patients, 34.1% were found to have kidney‐related heterozygous pathogenic variants associated with Mendelian kidney or genitourinary diseases, suggesting a higher‐than‐expected frequency of such variants in patients with DKD in Japan. Additionally, 12.7% of the patients (10 out of 27) carried diagnostic variants known to cause congenital diseases with an AD inheritance pattern. A significant proportion of the identified heterozygous pathogenic variants were associated with tubulointerstitial disease. However, no clear clinical associations with the presence of these variants were identified in this study.
We compared the prevalence of pathogenic and diagnostic variants in our study with those reported by Lazaro‐Guevara et al. (2021), who analyzed these variants in a DKD population 10 . We found that the pathogenic variant rate in our study was 34.1%, while the diagnostic variant rate was 12.7%. In contrast, Lazaro‐Guevara et al. 10 reported rates of 22.0% and 2.8%, respectively. Thus, our findings reflect a higher prevalence of both pathogenic and diagnostic variants in the DKD population. We also compared the diagnostic variant rate with the reports by Groopman et al. and Dahl et al., which identified diagnostic variants in 1.6% and 8.7% of patients with DKD, respectively 9 , 11 . Therefore, our study demonstrates a higher prevalence of pathogenic and diagnostic variants in the DKD population compared to previous reports.
The higher variant detection rate in our study, compared to previous reports, can be attributed to several factors. The first reason is the difference in the analyzed population 9 , 10 , 11 . Our cohort consisted entirely of an East Asian population, while previous studies primarily focused on Caucasian populations 9 , 10 , 11 . This ethnic difference may have contributed to the higher prevalence of heterozygous pathogenic variants in our study. Ethnic variations have been shown to influence the prevalence and development of various kidney diseases, including IgA nephropathy and focal segmental glomerulosclerosis, suggesting that ethnic background plays a significant role in the development of DKD 37 , 38 . Indeed, the majority of the variants identified in our study (N = 26, 89.7%) were more frequent in the Japanese population than in the global population (Table 2), indicating that the Japanese population has unique genetic characteristics. Second, we utilized recent databases for annotation, and differences in the databases may have influenced our results. The number of submitted records in ClinVar was approximately 0.5 million in September 2017; by May 2024, it had increased to over 4.3 million 23 , 39 . When investigating the frequency of rare variants in previous reports, it is crucial to consider not only the results but also the version of the database on which the data are based.
The most common heterozygous pathogenic variants identified in this study were associated with tubulointerstitial‐related genes (Figure 2). Additionally, variants in genes linked to congenital nephrotic syndrome (LAMA5, PAX2, and WT1) and nephronophthisis due to ciliopathy (DYNC2H1 and IQCB1) were detected (Table S2) 40 , 41 , 42 . Notably, 14 of the 27 patients (51.9%) with heterozygous pathogenic variants carried frameshift or stop‐gain variants.
Of the 27 patients with heterozygous pathogenic variants, 10 (12.7%) carried diagnostic variants in genes associated with AD congenital diseases, suggesting that these variants may potentially influence disease development. However, the significance of these diagnostic variants remains unclear, as no imaging or clinical findings strongly indicate hereditary conditions other than DKD. Recent studies have also identified patients with heterozygous pathogenic variants linked to autosomal recessive diseases, such as phenylketonuria and classical homocystinuria, who exhibit mild symptoms 43 , 44 . These findings highlight the need for further research into the clinical implications of these variants.
We investigated whether clinical characteristics could identify patients with heterozygous pathogenic variants, but no statistically significant associations were found in either univariate or multivariate analyses, likely due to the small sample size (Tables 1 and S5). Although DM duration and a family history of kidney disease were associated with the number of heterozygous pathogenic variants, no clear trend was observed linking the number of variants to longer DM duration or a higher prevalence of a family history of kidney disease (Table S6). Additionally, diabetic retinopathy, a key clinical feature of DKD, was not significantly associated with a higher risk of heterozygous pathogenic variants. Among the 10 patients with diagnostic variants, three had diabetic retinopathy, supporting the clinical diagnosis of DKD but highlighting the challenge of identifying DKD patients with diagnostic variants based solely on clinical data. Uric acid levels were higher in the group with diagnostic variants than in those without (Table S4). However, the variants identified in the diagnostic variant group were not associated with elevated uric acid levels. In contrast, five patients in the non‐diagnostic variant group, carrying pathogenic stop‐gain variants in SLC22A12 (linked to lower uric acid), had mean uric acid levels of 4.6 ± 1.6 mg/dL, compared to 6.3 ± 0.9 mg/dL in those without the variant 45 . These findings suggest that the variant may influence the observed differences. As few studies have examined the genetic background of DKD, further research with a larger sample size is needed.
This study has several limitations. First, the small sample size and single‐center cohort in Japan may limit the generalizability of the findings. The limited sample size also prevented a multivariate analysis that could account for all potential confounders related to heterozygous pathogenic variants. Future studies with larger numbers of patients are required. Second, nine SNV positions could not be confirmed using the Sanger method due to a lack of remaining DNA samples, difficulties with primer design, or interference from contiguous background sequences in the forward primer. Third, since DKD in this cohort was not diagnosed through renal biopsy, the possibility of other underlying diseases, which are clinically difficult to detect, cannot be ruled out. Fourth, the association between DKD pathology (including disease development and progression) and heterozygous pathogenic variants remains unclear. Finally, this study was descriptive, as WGS analysis was performed solely on the DKD population, without comparisons to healthy individuals or DM patients without CKD.
Among patients with clinically diagnosed DKD in Japan, a considerable number had kidney‐related heterozygous pathogenic and diagnostic variants, with a prevalence higher than previously reported. However, the clinical significance of these variants remains unclear, highlighting the need for larger cohort studies to further investigate their implications.
DISCLOSURE
The authors declare no conflict of interest.
Approval of the research protocol: This study was approved by the Ethics Committee of the University of Tokyo Graduate School of Medicine (ethical approval number: G10132) and was performed in accordance with the Declaration of Helsinki and the institutional guidelines.
Informed Consent: Written informed consent was obtained from all the participants.
Approval date of Registry and the registration no. of the study/trial: N/A.
Animal studies: N/A.
Supporting information
Table S1. List of the 790 genes associated with Mendelian kidney and genitourinary diseases.
Table S2. Category and phenotype of heterozygous pathogenic variants.
Table S3. Characteristics of patients with all heterozygous pathogenic variants.
Table S4. Comparison of clinical features between cases with and without diagnostic variants.
Table S5. Predictors for having heterozygous pathogenic variants (multivariate logistic regression analysis).
Table S6. Comparison of clinical features by the number of heterozygous pathogenic variants.
Data S1. Supplementary Methods and Supplementary References.
ACKNOWLEDGMENTS
The authors greatly appreciate all of those who participated in the survey. We would also like to thank Editage (www.editage.com) for editing the English language. This research was supported by the Center of Innovation Science and Technology‐based Radical Innovation and Entrepreneurship Program (COI STREAM), which was aimed at promoting industry‐academia collaboration, and the Japan Agency for Medical Research and Development under Grant Number (20 K17275).
DATA AVAILABILITY STATEMENT
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.
REFERENCES
- 1. Gonzalez‐Perez A, Saez M, Vizcaya D, et al. Incidence and risk factors for mortality and end‐stage renal disease in people with type 2 diabetes and diabetic kidney disease: A population‐based cohort study in the UK. BMJ Open Diabetes Res Care 2021; 9: e002146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Hoogeveen EK. The epidemiology of diabetic kidney disease. Kidney Dial 2022; 2: 433–442. [Google Scholar]
- 3. KDOQI . KDOQI clinical practice guidelines and clinical practice recommendations for diabetes and chronic kidney disease. Am J Kidney Dis 2007; 49(2 Suppl 2): S12–S154. [DOI] [PubMed] [Google Scholar]
- 4. Ahlqvist E, van Zuydam NR, Groop LC, et al. The genetics of diabetic complications. Nat Rev Nephrol 2015; 11: 277–287. [DOI] [PubMed] [Google Scholar]
- 5. Suzuki K, Akiyama M, Ishigaki K, et al. Identification of 28 new susceptibility loci for type 2 diabetes in the Japanese population. Nat Genet 2019; 51: 379–386. [DOI] [PubMed] [Google Scholar]
- 6. Sandholm N, Cole JB, Nair V, et al. Genome‐wide meta‐analysis and omics integration identifies novel genes associated with diabetic kidney disease. Diabetologia 2022; 65: 1495–1509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Hishida A, Nakatochi M, Sutoh Y, et al. GWAS meta‐analysis of kidney function traits in Japanese populations. J Epidemiol 2024; 34: 526–534. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Manolio TA, Collins FS, Cox NJ, et al. Finding the missing heritability of complex diseases. Nature 2009; 461: 747–753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Groopman EE, Marasa M, Cameron‐Christie S, et al. Diagnostic utility of exome sequencing for kidney disease. N Engl J Med 2019; 380: 142–151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Lazaro‐Guevara J, Fierro‐Morales J, Wright AH, et al. Targeted next‐generation sequencing identifies pathogenic variants in diabetic kidney disease. Am J Nephrol 2021; 52: 239–249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Dahl NK, Bloom MS, Chebib FT, et al. The clinical utility of genetic testing in the diagnosis and Management of Adults with chronic kidney disease. J Am Soc Nephrol 2023; 34: 2039–2050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Garcia SA, Costa M, Leon A, et al. The challenge of managing the evolution of genomics data over time: A conceptual model‐based approach. BMC Bioinformatics 2022; 23(Suppl 11): 472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Parsa A, Kao WH, Xie D, et al. APOL1 risk variants, race, and progression of chronic kidney disease. N Engl J Med 2013; 369: 2183–2196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Wojcik MH, Lemire G, Berger E, et al. Genome sequencing for diagnosing rare diseases. N Engl J Med 2024; 390: 1985–1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Ideozu JE, Liu M, Riley‐Gillis BM, et al. Diversity of CFTR variants across ancestries characterized using 454,727 UK biobank whole exome sequences. Genome Med 2024; 16: 43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Fujimaru T, Mori T, Chiga M, et al. Genetic diagnosis of adult hemodialysis patients with unknown etiology. Kidney Int Rep 2024; 9: 994–1004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Imamura M, Takahashi A, Matsunami M, et al. Genome‐wide association studies identify two novel loci conferring susceptibility to diabetic retinopathy in Japanese patients with type 2 diabetes. Hum Mol Genet 2021; 30: 716–726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Taira M, Imamura M, Takahashi A, et al. A variant within the FTO confers susceptibility to diabetic nephropathy in Japanese patients with type 2 diabetes. PLoS One 2018; 13: e0208654. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Hirakawa Y, Yoshioka K, Kojima K, et al. Potential progression biomarkers of diabetic kidney disease determined using comprehensive machine learning analysis of non‐targeted metabolomics. Sci Rep 2022; 12: 16287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Yoshioka K, Hirakawa Y, Kurano M, et al. Lysophosphatidylcholine mediates fast decline in kidney function in diabetic kidney disease. Kidney Int 2022; 101: 510–526. [DOI] [PubMed] [Google Scholar]
- 21. Sumida K, Nadkarni GN, Grams ME, et al. Conversion of urine protein‐creatinine ratio or urine dipstick protein to urine albumin‐creatinine ratio for use in chronic kidney disease screening and prognosis: An individual participant‐based meta‐analysis. Ann Intern Med 2020; 173: 426–435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Matsuo S, Imai E, Horio M, et al. Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 2009; 53: 982–992. [DOI] [PubMed] [Google Scholar]
- 23. Landrum MJ, Lee JM, Benson M, et al. ClinVar: Improving access to variant interpretations and supporting evidence. Nucleic Acids Res 2018; 46(D1): D1062–D1067. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Mori T, Hosomichi K, Chiga M, et al. Comprehensive genetic testing approach for major inherited kidney diseases, using next‐generation sequencing with a custom panel. Clin Exp Nephrol 2017; 21: 63–75. [DOI] [PubMed] [Google Scholar]
- 25. Fujimaru T, Mori T, Sekine A, et al. Kidney enlargement and multiple liver cyst formation implicate mutations in PKD1/2 in adult sporadic polycystic kidney disease. Clin Genet 2018; 94: 125–131. [DOI] [PubMed] [Google Scholar]
- 26. Bleyer AJ, Westemeyer M, Xie J, et al. Genetic etiologies for chronic kidney disease revealed through next‐generation renal gene panel. Am J Nephrol 2022; 53: 297–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Knoers N, Antignac C, Bergmann C, et al. Genetic testing in the diagnosis of chronic kidney disease: Recommendations for clinical practice. Nephrol Dial Transplant 2022; 37: 239–254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Fang Y, Shi H, Xiang T, et al. Genetic architecture of childhood kidney and urological diseases in China. Phenomics 2021; 1: 91–104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: A joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med 2015; 17: 405–424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Le NQK, Yapp EKY, Nagasundaram N, et al. Classifying promoters by interpreting the hidden information of DNA sequences via deep learning and combination of continuous FastText N‐Grams. Front Bioeng Biotechnol 2019; 7: 305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Carvalho SG, Guerra‐Sá R, de C Merschmann LH. The impact of sequence length and number of sequences on promoter prediction performance. BMC Bioinformatics 2015; 16(Suppl 19): S5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Arteche‐Lopez A, Avila‐Fernandez A, Romero R, et al. Sanger sequencing is no longer always necessary based on a single‐center validation of 1109 NGS variants in 825 clinical exomes. Sci Rep 2021; 11: 5697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Karczewski KJ, Francioli LC, Tiao G, et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature 2020; 581: 434–443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Tadaka S, Hishinuma E, Komaki S, et al. jMorp updates in 2020: Large enhancement of multi‐omics data resources on the general Japanese population. Nucleic Acids Res 2021; 49(D1): D536–D544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. He F, Xia X, Wu XF, et al. Diabetic retinopathy in predicting diabetic nephropathy in patients with type 2 diabetes and renal disease: A meta‐analysis. Diabetologia 2013; 56: 457–466. [DOI] [PubMed] [Google Scholar]
- 36. Hays T, Groopman EE, Gharavi AG. Genetic testing for kidney disease of unknown etiology. Kidney Int 2020; 98: 590–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Xu LL, Zhou XJ, Zhang H. An update on the genetics of IgA nephropathy. J Clin Med 2023; 13: 123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Wang M, Chun J, Genovese G, et al. Contributions of rare gene variants to familial and sporadic FSGS. J Am Soc Nephrol 2019; 30: 1625–1640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. ClinVar submissions [Internet]. [Accessed 2024 May 15]. Available from https://www.ncbi.nlm.nih.gov/clinvar/submitters/.
- 40. Zhu V, Huang T, Wang D, et al. Ocular manifestations of the genetic causes of focal and segmental glomerulosclerosis. Pediatr Nephrol 2024; 39: 655–679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Braun DA, Warejko JK, Ashraf S, et al. Genetic variants in the LAMA5 gene in pediatric nephrotic syndrome. Nephrol Dial Transplant 2019; 34: 485–493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Sakakibara N, Nozu K, Yamamura T, et al. Comprehensive genetic analysis using next‐generation sequencing for the diagnosis of nephronophthisis‐related ciliopathies in the Japanese population. J Hum Genet 2022; 67: 427–440. [DOI] [PubMed] [Google Scholar]
- 43. Kalyta K, Stelmaszczyk W, Szczesniak D, et al. The Spectrum of the heterozygous effect in biallelic mendelian diseases‐the symptomatic heterozygote issue. Genes 2023; 14: 1562. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Hames A, Khan S, Gilliland C, et al. Carriers of autosomal recessive conditions: Are they really “unaffected?”. J Med Genet 2023; 61: 1–7. [DOI] [PubMed] [Google Scholar]
- 45. Misawa K, Hasegawa T, Mishima E, et al. Contribution of rare variants of the SLC22A12 gene to the missing heritability of serum urate levels. Genetics 2020; 214: 1079–1090. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. List of the 790 genes associated with Mendelian kidney and genitourinary diseases.
Table S2. Category and phenotype of heterozygous pathogenic variants.
Table S3. Characteristics of patients with all heterozygous pathogenic variants.
Table S4. Comparison of clinical features between cases with and without diagnostic variants.
Table S5. Predictors for having heterozygous pathogenic variants (multivariate logistic regression analysis).
Table S6. Comparison of clinical features by the number of heterozygous pathogenic variants.
Data S1. Supplementary Methods and Supplementary References.
Data Availability Statement
The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.