Detection of Cryptic Mosaicism in X-linked Alport Syndrome Prompts to Re-evaluate Living-donor Kidney Transplantation

Abstract

Background. 

Alport syndrome is a hereditary nephropathy caused by mutations in collagen IV genes and characterized by ultrastructural lesions of the glomerular basement membrane. Some patients have a negative family history with apparently de novo mutations. Although somatic mosaicism has been postulated, as cryptic mosaicism cannot be detected from mutational screening on peripheral blood samples, cases in kidney-confined mosaic form have been missed.

Methods. 

We report the case of a 24-year-old male patient with X-linked Alport syndrome diagnosis due to a COL4A5 pathogenic mutation (c.3334_3337dup [p.Gly1113Alafs^*25]). The same mutation had not been previously detected on a peripheral blood sample of maternal DNA. However, the mother, who was undertaking a clinical re-evaluation to take in consideration the possibility of a living-kidney transplantation, had experienced persistent microhematuria since the age of 10 years.

Results. 

A next-generation sequencing approach performed on maternal DNA from both peripheral blood sample and urine-derived podocyte-lineage cells unmasked the COL4A5 mutation only in the podocyte-lineage cells.

Conclusions. 

This finding unveils an early postzygotic event which can explain both the renal involvement and germline mosaicism. It changes the inheritance risk for each pregnancy raising it to 50% and underlines the need for different clinical management in the mother. This seems to indicate that a case-by-case more cautious approach is needed with mother-to-son kidney transplants.

Alport syndrome (AS) is a clinically heterogeneous nephropathy caused by mutations in collagen IV genes with a frequency of 1:5000 in X-linked form and 1:50 000 in autosomal disease.¹ It is characterized by hematuria, proteinuria, and progressive renal failure with ultrastructural lesions of the glomerular basement membrane (GBM) up to end-stage renal disease (ESRD).² AS is a genetically and clinically heterogeneous disorder to which all 3 main models of Mendelian inheritance, namely X-linked, autosomal recessive, and autosomal dominant, are applicable.³ X-linked AS (XLAS) is the most common form, accounting for about 80%–85% of cases of AS, with an estimated incidence of 1 in 5000 males.^2,4,5 The X-linked form is clinically characterized by severely affected males with persistent hematuria, proteinuria, and inevitable progression to ESRD (typically during the 2nd or 3rd decade). It also has a frequent association with extrarenal manifestations such as hearing loss and ocular anomalies. Female carriers of XLAS typically have good renal function, although about 15% develop end-stage renal failure by the age of 60 years. Variable clinical features are often related to a skewed X chromosome inactivation in individual tissues.^4-7 Thus, the carrier state should be viewed as at-risk rather than a benign condition. In patients with XLAS, approximately 10%–15% of COL4A5 mutations occur as spontaneous events.⁸ The existence of an autosomal dominant alport syndrome (ADAS) form is still debated. However, in the past decade, it has been shown that ADAS accounts for a significant fraction of cases.⁹ Furthermore, the existence of ADAS as a clinical entity distinct from thin basement membrane nephropathy has been supported by the employment of a next-generation sequencing (NGS) approach in patients presenting with GBM alterations overlapping with the abnormalities found in the XLAS and the autosomal recessive alport syndrome form. In these patients, a second causative mutation in COL4 genes has not been found. Recently, Pescucci et al¹⁰ contributed to delineating the overall clinical picture of ADAS which is characterized by hematuria and proteinuria evolving toward ESRD with a mean age of 55 years for ESRD onset.

Somatic mosaicism is a well-known phenomenon observed in several inherited diseases, including AS. It is often responsible for variable expressivity and incomplete penetrance depending on the tissues involved and the grade of mosaicism in each tissue.¹¹ There have been few reports of somatic mosaicism in XLAS. However, in all the described cases, mosaicism was either detected in peripheral blood samples by high-depth NGS^8,11-13 or postulated, in the presence of >1 affected child, in parents with normal urinary findings and no collagen IV mutations on genomic DNA from peripheral leukocytes.^14,15 In this latter case, mosaicism in germ cells may be the result of either a mutation in the germ cell that thereafter undergoes mitotic divisions or an early postzygotic mutation in the mesoderm which gave rise to mosaicism in both the urinary tissues and the germline. Thus, the phenotype may or may not be expressed in the mosaic subject depending on the involved tissues and the proportion of mutated cells.¹⁶

To date, given the inaccessibility of podocytes, the grade of involvement of the disease-relevant tissue could not be ruled out. Several works have recently shown that podocyte excretion (podocyturia) is a silent phenomenon that precedes proteinuria^17,18 and may be a more sensitive indicator of glomerular disease activity and damage underlying the possibility of isolating podocytes in GBM damage-related disorders.¹⁹ Recently, we have proven that, from affected AS patients, it is possible to isolate a cellular population, defined as podocyte-lineage cells, which represents a promising tool for more precise treatments for AS.⁷

Going beyond these findings, employing this novel tool as source of good-quality DNA for an NGS approach, we detected a renal, otherwise cryptic mosaicism in the microhematuric mother of a patient with XLAS. This result completely changed the recurrence risk assessment and the clinical management of the family. In line with the concept of a personalized therapeutic intervention, it prompted us to re-evaluate the mother’s eligibility as a potential living donor for kidney transplantation.

MATERIALS AND METHODS

Samples Collection

Molecular diagnosis of AS was established at the Medical Genetics Unit in Siena (Azienda Ospedaliera Universitaria Senese). The family underwent genetic counseling, and blood samples were collected in EDTA-containing tubes for COL4A3, COL4A4, and COL4A5 mutational analysis (NGS and Sanger Sequencing). Proband and family members provided and signed a written informed consent form for the use of DNA samples for diagnostic purposes and for urine samples collection with the purpose of isolating podocytes-lineage cells. The study was approved by Azienda Ospedaliera Universitaria Senese Ethics Committee. Genomic DNA was isolated from EDTA peripheral blood samples and from urine-derived podocyte-lineage cells using a QIAamp DNA Blood Kit according to the manufacturer’s protocol (Qiagen, Hilden, Germany).

Urine-derived podocyte-lineage cells were isolated and cultured according to the previously published protocol.⁷

Podocytes-lineage Cells Isolation

Cells from the proband and his mother were isolated from the urine following a 2-step process as previously described.⁷ The urine was processed within 4 hours of collection. Urine samples were centrifuged into a 50-mL falcon tube at 2400 rpm of speed for 10 minutes. The pellet was washed with 10 mL of Washing Buffer, composed of Dulbecco’s phosphate-buffered saline without MgCl₂ and CaCl₂ and supplemented with 100 U/mL of penicillin, 100 μg/mL of streptomycin, and 500 ng/mL of amphotericin B, and again centrifuged at 2200 rpm for 10 minutes. After the removal of the supernatant, the pellet was resuspended in 250 µL of primary medium, composed of DMEM/high glucose and Ham’s F12 nutrient mix (1:1) and supplemented with 10% (vol/vol) fetal bovine serum (FBS), 100 of U/mL penicillin, 100 μg/mL streptomycin, the renal epithelial cell growth medium Single Quote kit supplements (recombinant human epidermal growth factor, insulin, gentamicin sulfate-amphotericin 1000, transferrin, hydrocortisone, triiodothyronine, epinephrine, and FBS), and 2.5 μg/mL amphotericin B. We than plated the pellet into a single well of a 12-well culture plate previously treated with gelatin 0.1% and incubated for 30 minutes at 37°C. Primary medium was added to the culture for the next 3 days (ie, 24, 48, and 72 h after plating). Approximately 96 hours after plating, one-third of the medium was removed (leaving 1 mL) and 1 mL of RE/MC expansion medium added (to select for podocyte-lineage cells growth). The medium composed of RE cells basal medium supplemented with renal epithelial cell growth medium Single Quote Kit, and MC medium is composed of DMEM/high glucose supplemented with 10% (vol/vol) FBS, 1% (vol/vol) GlutaMAX, 1% (vol/vol) nonessential amino acid, 100 U/mL penicillin, 100 μg/mL streptomycin, 5 ng/mL bFGF, 5 ng/mL PDGF-AB, and 5 ng/mL epidermal growth factor. The RE medium and MC medium are then combined together in a 1:1 ratio. The proliferation medium was changed daily, until 2 groups of small colonies were noticed, a first group of cells with a more regular appearance, smooth-edged contours, and cobblestone-like cells morphologies, and a second group more randomly arranged with a higher proliferation rate. Cells were split around 9–12 days after plating.

NGS and Variant Interpretation

The library preparation was performed using the Ion AmpliSeq Library Kit 2.0 (Thermo Fisher). This kit allowed us to obtain a barcoded library for all 151 exons of COL4A3/COL4A4/COL4A5 genes compatible with the Ion S5 platform, according to the ThermoFisher protocol. Libraries were purified using Agencourt AMPure XP system and quantified using the Qubit dsDNA HS Assay Kit (ThermoFisher). We pooled the library using an equimolar ratio and clonally amplified by emulsion polymerase chain reaction (PCR) using the Ion Chef system (Ion 520 kit chef, ThermoFisher). The DNA templates were loaded to Ion 520 chip and sequenced on the Ion Torrent S5, according to the ThermoFisher protocol. Postrun analysis was conducted using the latest version (v5.8) of the data analysis software Torrent Suite (ThermoFisher). Coverage assessment was performed using the “coverageAnalysis” plug-in (v5.8.0.8) that gave information about the amplicon read coverage and variants using the “variantCaller” plug-in (v5.8.0.19). The mean depth of the sample was 4811x (average base covered depth), and 97.44% of the base target had a coverage of ≥500x. In the region of interest, the coverage of the amplicon was 1352x and all the bases involved had a coverage of >500x. Identified variants were checked in the datasets of the following online databases: NCBI dbSNP Build 151 (https://www.ncbi.nlm.nih.gov/snp), ExAC Browser Beta (http://exac.broadinstitute.org/), and 1000 genomes (http://www.1000genomes.org). Variant validation was performed using the PE Big Dye Terminator Cycle Sequencing Kit on an ABI Prism 3130 analyzer (Applied Biosystem). The version 4.9 of Sequencher Software (Genes Code Corporation, Ann Arbor, MI) was used for sequence analysis.

Human Androgen Receptor Assay X-inactivation Assay

Enzymatic digestion with the HpaII enzyme was performed on mother peripheral blood DNA and on urine-derived podocyte-lineage DNA. Digestion was performed in a final volume of 10 µL with: HpaII enzyme (1 U), Buffer (1X), DNA (500 ng), H₂O up to proper volume. The sample was digested o/n at 37°C in a heating block. The day after that, 100 ng of digested DNA and 100 ng of nondigested DNA were amplified simultaneously by PCR in a final volume of 25 µL (Buffer 1X, MgCl₂ 3 mmol/L, dNTPs 200 µM, both Fw and Rv primers 12.5 pmol, Taq 1 U, and H₂O up to proper volume). The samples were amplified in a thermocycler following a proper PCR program: 95°C for 5′, 95°C for 1′, 60°C for 30′′ (×35 cycles), 72°C for 30′′, 72°C for 5′, and 15°C to infinite. Amplification was confirmed by agarose gel electrophoresis. PCR products were separated on an Applied Biosystems ABI3130 sequencer platform taking advantage of the fluorescent tail (6-carboxyfluorescein) linked to a primer and analyzed by GeneMapper analysis software (ThermoFisher Scientific, Waltham, MA) (Figure S1A and S1B, SDC, https://links.lww.com/TP/B863).

CASE REPORT

Here, we report the case of a 25-year-old male who came for a clinical re-evaluation during the process of obtaining a living-donor transplant from his mother. The proband was the only child. He had experienced persistent microhematuria since the age of 11 months. The occurrence of proteinuria was reported when he was 3 years old (Figure 1). His laboratory profile exhibited a serum creatinine of 2.6 mg/dL and an estimated glomerular filtration rate (estimated by chronic kidney disease epidemiology collaboration equation) of 32.3 mL/min. When he was 6 years old, a kidney biopsy revealed areas of thinning, thickening, and splitting of the GBM, consistent with a clinical diagnosis of AS. At the time of genetic counseling, audiometric evaluation displayed a neurosensorial hearing loss, while an ophthalmoscopic evaluation was reported to be normal. The proband was undergoing dialysis and waiting for kidney transplantation.

FIGURE 1.: Family’s pedigree. Pedigree of the family is depicted and segregation of the mutated allele is reported. Results from the DNA from peripheral blood samples and urine-derived podocytes-lineage cells are schematically represented. AMI, acute myocardial infarction; M1, mutation 1; Wt, wild type.

Molecular investigation performed at 6 years of age, using single-strand conformation polymorphism scanning of the COL4A5 gene, had revealed a COL4A5 pathogenic mutation (c.3334_3337dup [p.Gly1113Alafs^*25]). The same mutation had not been detected on the maternal DNA from peripheral blood samples by single-strand conformation polymorphism analysis. This led us to a conclusion of “sporadic” XLAS due to a de novo mutation. However, the proband’s mother, who is currently 53 years old, had reported persistent microhematuria since the age of 10 years.

A recent assessment of the abnormalities in the shape of red cells in her urine led to a conclusion of glomerular hematuria. The mother’s laboratory profile reported a microalbuminuria level equal to 7.2 mg/L, with a serum creatinine of 0.80 mg/dL and an estimated glomerular filtration rate (estimated by chronic kidney disease epidemiology collaboration equation) of 90.8 mL/min. The urine albumin/creatinine ratio is 0.14 mg/g. No other laboratory tests concerning audiometric examinations or fundus oculi evaluation were possible because the patient did not consider it appropriate to carry out further laboratory tests due to the presence of only renal symptoms. Thus, to better establish mother’s eligibility as living donor, we decided to more closely investigate the molecular background to rule out either a digenic pattern of transmission with a second maternally inherited COL4 mutation or a maternal somatic mosaicism for the COL4A5 identified mutation. NGS, performed on genomic DNA from peripheral blood samples of both the mother and the son, confirmed that while the son exhibited the presence of the (p.[Gly1113Alafs*25]) COL4A5 pathogenic mutation, the mother did not (Table S1, SDC, https://links.lww.com/TP/B863). No other pathogenic variant or variant of uncertain significance was detected in both samples allowing us to exclude a digenic pattern of transmission. NGS was then performed on DNA extracted from proband’s mother urine-derived podocyte-lineage cells. Surprisingly, the pathogenic COL4A5 mutation was detected in 49% of analyzed fragments (Table S2, SDC, https://links.lww.com/TP/B863) and confirmed by Sanger sequencing analysis (Figure S2, SDC, https://links.lww.com/TP/B863). This allowed us to identify a cryptic mosaicism with a high percentage of disease-relevant cells harboring the mutation. Furthermore, X chromosome inactivation performed in both blood and urine samples of the mother using the Androgen Receptor Assay (HUMARA)²⁰ detected a ratio of 41%–58% and of 60%–40%, respectively (Figure S1C, SDC, https://links.lww.com/TP/B863). Finally, to define which of the 2 X chromosomes was inherited by the proband, we performed the GeneMapper Analysis by comparing the undigested peripheral blood sample DNA of the proband with the blood sample DNA of the mother.

The maternal mutated allele which was passed down to the son turned out to be 58% methylated (Figure S3, SDC, https://links.lww.com/TP/B863).

These data taken together allowed us to conclude that the mutated allele had a slightly unbalanced X chromosome inactivation.

DISCUSSION

In the past decade, high-depth NGS tools have allowed us to reach the conclusion that somatic mosaicism is more common than what we previously thought. This phenomenon explains the reduced penetrance and variable expressivity. If confined to inaccessible tissues and thus cryptic, mosaicism can impair an accurate risk assessment and proper clinical management.

To date, few asymptomatic or barely symptomatic females have been promptly recognized as XLAS patients harboring germline mosaicism after 1 or more affected children were diagnosed with XLAS.^13-15 In 50% of cases, germline mosaicism is associated with somatic mosaicism. Therefore, simultaneous examination of several different tissues, such as buccal smear, hair, fibroblasts, and urinary cells, is necessary to detect it. Sometimes, in the presence of informative families, even in the absence of pathogenic variants in saliva and peripheral leukocytes, somatic and/or germline mosaicism has been stated in the parents. Indeed, lack of good-quality DNA from urinary cell sediment has so far prevented the execution of an NGS approach on disease-relevant cell types such as kidney-derived cells. Doing so would allow us to detect even low-grade somatic mosaicism and reach a clear molecular diagnosis. Furthermore, urinary cell sediment is enriched for kidney transitional cells which, given their different embryonic origin, are not representative of the kidney glomerular structure involved in AS pathogenesis.

Here, we applied our recently established cell system⁷ and high-depth NGS approach. We unmasked a cryptic mosaicism, which involves ovarian cells and kidney and explains the persistent microhematuria reported in the proband’s mother. Surprisingly, the pathogenic COL4A5 variant was detected in 49% of analyzed fragments, allowing us to conclude that likely a sizeable proportion of cells at the kidney level harbor the mutation. However, we cannot ignore the possibility that our estimation of mosaic variant percentage using cultured cells could be biased toward the mutated allele. This is due to either a preferential clonal expansion of COL4A5-mutated cells in culture or because a very low proportion of wild-type unaffected cells could fall into urine over time.

According to the X chromosome inactivation results,²⁰ the level of kidney involvement could likely be related to a skewed X-inactivation rather than to the degree of somatic mosaicism. This result allows us to conclude that an early postzygotic event occurred in the intermediate mesoderm which differentiating in the nephrotomes develops in the excretory units of the urinary system and in the gonads, which then bring the causative mutations in germ cells (Figure 2).

FIGURE 2.: Mosaic event occurring in the intermediate mesoderm. Figure is modified from the Netter Atlas of anatomy. The mutational event in COL4A5 is depicted in the box on the right. The occurrence of an early mutational event in the intermediate mesoderm leads to a full mutational load in kidneys and gonads.

In a fertile female, this finding changes the risk of having a child with XLAS. The inheritance risk for each pregnancy can indeed be as high as 50%. The proband’s mother’s genotype at the podocyte level is comparable to the genotype of a carrier female with a germline heterozygote COL4A5 pathogenic mutation. This evidence should urge us to reconsider her clinical management given that 15% of females with XLAS develop end-stage renal failure by the age of 60 years. Notably, when we considered kidney transplantation, this careful molecular characterization prompted us to re-evaluate the mother’s eligibility as kidney living donor. We were obliged to more thoroughly investigate her kidney function to ensure a successful procedure for the son and avoid ESRD in the mother. In our view, the mother was eligible for kidney donation due to her normal kidney function likely in line with the type of mutation and its mosaic state, absence of proteinuria, and negligible risk of future kidney failure at her age. Kidney transplant has not yet been performed since the proband’s creatinine clearance remains at 45 mL/min.

Our evidence suggests that a more cautious genetic evaluation is necessary in the presence of apparently sporadic AS. This should be taken into account, alongside the other assessments required for living-donor kidney transplant. It also suggests the importance of a new sequencing approach on DNA from podocyte-lineage cells in patients with clinical and histologic findings highly suggestive of AS for whom molecular diagnosis has not been reached on DNA from peripheral blood samples.

ACKNOWLEDGMENTS

We would like first to thank the Alport syndrome patients and their families. The "Cell lines and DNA bank of Rett Syndrome, X-linked mental retardation and other genetic diseases", members of the Telethon Network of Genetic Biobanks (project no. GTB18001), funded by Telethon Italy, and the EuroBioBank network and the A.S.A.L. Onlus association, provided us with specimens. We are also thankful for the donation from "Graziano and Marco Laurini" to A.R. We also say thank you to the research funding program by the Alport Syndrome Foundation (ASF), the Pedersen Family, and the Kidney Foundation of Canada (KFOC). This work is generated by the European Reference Network for Rare Kidney Diseases (ERKnet).