Gene panel sequencing identifies a likely monogenic cause in 7% of 235 Pakistani families with nephrolithiasis

Abstract

Background:

Nephrolithiasis (NL) affects 1 in 11 individuals worldwide and causes significant patient morbidity. We previously demonstrated a genetic cause of NL can be identified in 11–29% of pre-dominantly American and European stone formers. Pakistan, which resides within the Afro-Asian stone belt, has a high prevalence of nephrolithiasis (12%) as well as high rate of consanguinity (>50%).

Methods:

We recruited 235 Pakistani subjects hospitalized for nephrolithiasis from 5 tertiary hospitals in the Punjab province of Pakistan. Subjects were surveyed for age of onset, NL recurrence, and family history. We conducted high-throughput exon sequencing of 30 NL disease genes and variant analysis to identify monogenic causative mutations in each subject.

Results:

We detected likely causative mutations in 4 of 30 disease genes, yielding a likely molecular diagnosis in 7% (17 of 235) of NL families. Only 1 of 17 causative mutations was identified in an autosomal recessive disease gene. 10 of the 12 detected mutations were novel mutations (83%). SLC34A1 was most frequently mutated (12 of 17 solved families). We observed a higher frequency of causative mutations in subjects with a positive NL family history (13/109, 12%) versus those with a negative family history (4/120, 3%). Five missense SLC34A1 variants identified through genetic analysis demonstrated defective phosphate transport.

Conclusions:

We examined the monogenic causes of NL in a novel geographic cohort and most frequently identified dominant mutations in the sodium-phosphate transporter SLC34A1 with functional validation.

Introduction

Nephrolithiasis (NL) affects 1 in 11 individuals during their lifetime(Scales et al. 2012; Tasian et al. 2016; 2015). NL is associated with significant patient morbidity, recurrence, and healthcare costs (Rule et al. 2009; 2015). The causes of NL are not well understood. Until recently, monogenic causes of NL were thought to be limited to rare tubulopathies and genetic syndromes. However, we have established that a single-gene cause of NL can be identified in 11% and 16.7–29.4% of selected adult and pediatric stone formers, respectively(Braun et al. 2016a; Daga et al. 2017; Halbritter et al. 2015). These studies were comprised of predominantly American and European stone formers, and therefore the broader relevance of these findings worldwide is not clear.

Pakistan resides within the “Afro-Asian stone belt”, a region with a high prevalence of nephrolithiasis (12%)(Rizvi S.A.H. et al. 2002). Environmental risk factors, including chronic dehydration and nutrition, contribute to these higher rates of nephrolithiasis(Rizvi S.A.H. et al. 2002). Pakistan also has a high rate of consanguineous marriages (>50%)(Romeo and Bittles, 2014), which by nature of Mendelian genetics is associated with an increased prevalence of autosomal recessive genetic diseases(Overall A. D. J. et al. 2003). We hypothesized that NL disease gene mutations cause nephrolithiasis in a significant fraction of Pakistani stone formers. However, the prevalence of monogenic forms of nephrolithiasis in this population is unknown.

We therefore performed high-throughput gene panel sequencing analysis of 30 known NL-causing genes in a cohort of 235 Pakistani subjects with nephrolithiasis recruited from one of 5 different tertiary care hospitals in Punjab, Pakistan. We show that likely causative mutations in known NL-causing genes are present in 7% of stone formers. We furthermore provide practical therapeutic and preventative measures based on each molecular diagnosis.

Materials and methods

Human subjects

Subjects were recruited during admission for nephrolithiasis, which was confirmed by abdominal ultrasound. The subjects were admitted to one of 5 different tertiary care hospitals in Punjab, Pakistan from July 2014 to December 2016 (Suppl. Figure 1). Subjects received informed consent and provided clinical information (confirmed by their urologist and/or medical records), family history for pedigree construction, and a blood sample for DNA extraction. Consanguinity was determined by history from adult subjects and/or guardians for pediatric cases by asking if parents of subjects were related. Ten subjects did not consent to the study, and 7 subjects were excluded as they had an evident secondary cause of NL (including urinary tract infections and secondary hyperoxaluria). Of the remaining 242 families, adequate DNA samples for genetic studies were obtained in 235 (Suppl. Figure 1). For 31 families, additional family members were recruited upon consent for clinical information and DNA submission, allowing for multi-generational pedigree construction. Serum chemistries, urine metabolites and stone analyses were requested and obtained when available. In total, the cohort consisted of 440 individuals (235 initial probands, 115 additional affected family members and 90 unaffected family members) in 235 families. The study was approved by the institutional review board of Boston Children’s Hospital and the Ethical Review Committee for Medical and Biomedical Research, University of Health Sciences, Lahore, Pakistan. It adheres to the Declaration of Helsinki.

Gene panel sequencing

We performed mutation analysis using barcoded multiplex PCR (Fluidigm 48.48-Access Arrays™) based approach that we developed previously(Braun et al. 2016a; Halbritter et al. 2012, 2013, 2015). We designed 518 target specific primers for 381 coding exons and the adjacent splice sites of 30 genes that are known monogenic causes of nephrolithiasis or nephrocalcinosis (defined by an OMIM phenotype). The genes screened included: AGXT, APRT, ATP6V0A4, ATP6V1B1, CA2, CASR, CLCN5, CLCNKB, CLDN16, CLDN19, CYP24A1, FAM20A, GRHPR, HNF4A, HOGA1, HPRT1, KCNJ1, OCRL, SLC12A1, SLC22A12, SLC2A9, SLC26A1, SLC34A1, SLC34A3, SLC3A1, SLC4A1, SLC7A9, SLC9A3R1, VDR, and XDH (Suppl. Table 1). Amplicon sizes were chosen to range from 250 to 300 bp. Primer sequences will be provided upon request. The use of barcoded multiplex PCR (Fluidigm 48.48-Access Arrays™ system) allowed parallel amplification of all 518 amplicons in 48 patients at a time. Subsequently, the pooled libraries were sequenced on an Illumina MiSeq® instrument. Sequence reads were aligned to the human reference sequence using CLC Genomics Workbench™ (CLC-bio, Aarhus, Denmark)(Braun et al. 2016a; Halbritter et al. 2012, 2013, 2015) to identify variants that were distinct from the reference human genome (Suppl. Figure 2 and Suppl. Figure 3). We aimed for minimum 20× coverage of 80% of exons in each sample and repeated sequencing of samples that did not meet this initial standard.

Mutational analysis

We excluded synonymous variants and variants that occur with minor allele frequencies >1% in the dbSNP (version 147) database (Suppl. Figure 2 and Suppl. Figure 3). Variant frequency was further assessed in exome and genome databases EVS, ExAC, and gnomAD of healthy control individuals. Remaining variants were then evaluated as in Suppl. Figure 2 and Suppl. Figure 3. All remaining variants were confirmed in original subject DNA by Sanger sequencing with segregation in family members when DNA available.

Statistical testing

Odds ratios were measured using the online resource at MedCalc (see web resource below).

Expression of SLC34A1 wildtype and mutant RNA in Xenopus laevis oocytes

Animal studies adhere to the NIH guide for the Care and Use of Laboratory Animals or the equivalent. Xenopus laevis oocytes were purchased from EcoCyte and transferred to the standard extracellular solution for oocyte experiments containing 100 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 10 mM HEPES, pH 7.4 adjusted with Tris. Inorganic phosphate was added from a 1M K₂HPO₄/ KH₂PO₄ stock premixed to give pH 7.4. Modified Barth’s solution for storing oocytes contained (in mM): 88 NaCl, 1 KCl, 0.41 CaCl₂, 0.82 MgSO₄, 2.5 NaHCO₃, 2 Ca(NO₃)₂, 7.5 HEPES, pH 7.5 adjusted with Tris and supplemented with 5 mg/l doxycyclin and 5 mg/l gentamicin. All standard reagents were obtained from either Sigma-Aldrich or Fluka (Buchs, Switzerland). Oocytes were injected with 50 nl of cRNA (200 ηg/μl for single injection, 100 ng/μl total RNA for coinjection) of wildtype or mutant SLC34A1 constructs or water as negative control (NI). Constructs were designed as discussed in Supplementary Methods 1. Experiments were performed 3 days after injection.

³²Phosphate uptake

Water injected control oocytes (NI), oocytes expressing wildtype and mutant SLC34A1 (6–8 oocytes/group) were first allowed to equilibrate in ND96 solution (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 5 mM Hepes, pH 7.4 adjusted with Tris) without tracer. After aspiration of this solution, oocytes were incubated in ND96 solution containing 1mM cold P_i and ³²P_i (specific activity 10 mCi mmol-1 Pi, Perkin Elmer). Uptake proceeded for 10 min, then oocytes were washed 4 times with ice-cold ND96 and lysed individually in 2% SDS for 30 min before addition of the scintillation cocktail (Emulsifier-Safe, PerkinElmer). The amount of radioactivity in each oocyte was measured by scintillation counting (Tri-Carb 29000TR, Packard).

Web resources

UCSC Genome Browser, https://genome.ucsc.edu

Exome Variant Server, http://evs.gs.washington.edu/EVS

Exome Aggregation Consortium, http://exac.broadinstitute.org

Genome Aggregation Database, http://gnomad.broadinstitute.org/

Online Mendelian Inheritance in Man (OMIM), http://www.omim.org

Polyphen2, http://genetics.bwh.harvard.edu/pph2

Sorting Intolerant from Tolerant (SIFT), http://sift.jcvi.org

MutationTaster, http://www.mutationtaster.org

MedCalc Odds Ratio Calculator, https://www.medcalc.org/calc/odds_ratio.php

RESULTS

Cohort characteristics

We recruited a diverse cohort of stone formers across sex, age, nephrolithiasis history, and familial inheritance, as evident from demographic data of probands from these families. 143 probands were male, and 92 were female (M:F ratio 1.6:1). Further clinical data was available for 229 of 235 probands. The age of first NL episode spanned pediatric and adult age groups: 1–20 years, 74/229 (32%); 21–40 years, 108/229 (47%); 41–60 years, 42/229 (18%); 61–80 years, 5/229 (2%). The median age of onset was 28 years (range 1–76 years). 113 (49%) of 229 probands reported a history of recurrent stones. 47% (109/229) of probands had a positive family history of NL. 53% (122/229) were born from consanguineous unions.

Monogenic causative mutation detection

Using high-throughput exon sequencing of 30 NL disease genes in 235 subjects with renal stones, we identified a molecular diagnosis in 7% (17 of 235 NL families) in 4 different known NL genes SLC34A1, SLC4A1, SLC7A9, and APRT (Figure 1). Panel sequencing results were confirmed by Sanger sequencing (Suppl. Figure 4 and 5). Heterozygous mutations were detected in the 3 dominant genes SLC34A1 (12 families), SLC4A1 (3 families), SLC7A9 (1 family), while biallelic mutations were identified only in 1 recessive gene APRT (1 family) as described in Table 1 and Figure 1. 10 of 12 mutations (83%) were novel. Mutations in SLC34A1 were identified in two families UF26 and UF30 where additional family members with stones were recruited, and segregation was confirmed (Suppl. Figure 4).

Figure 1.

Fraction (%) of 235 Pakistani NL families in whom a causative mutation was detected in one of 30 known NL genes (“solve rate”). The flow diagram depicts the distribution of NL families by solve rate and by mode of inheritance (recessive versus dominant). NL nephrolithiasis. “Asterisk” Mutations in SLC34A1 determined based on standardized genetics analysis and functional analysis (in case of missense mutations)

Table 1.

Molecular genetic diagnoses established in 3 dominant genes in 20 individuals from 16 NL families (6.8%) and in 1 recessive gene in 1 family (0.4%) out of 235 families with NL from Punjab province in Pakistan

Family_Individual	Nucleotide change	Amino acid change	Zygosity state	PPh2, Evolutionary Conservation	gnomAD Allele frequencya	Ref.	Sex	Age of onset (years)	Stone recurrence	Family history	Parental consanguinity	Phenotype prior to mutational analysis
Dominant genes
SLC34A1 (Type 2 sodium/phosphate cotransporter)
UF26–2	c.511G>A	p.Val171Ile	het	0.881, D r.	1/37/276942	Novel	F	35	Yes		Y	NL
UF26–3	c.511G>A	p.Val171Ile	het	0.881, D r.	1/37/276942	Novel	M	30	No	Y		NL
UF26–4	c.511G>A	p.Val171Ile	het	0.881, D r.	1/37/276942	Novel	M	20	Yes			NL
UF30–3	c.653C>T	p.Ala218Val	het	0.899, C i.	0/23/276526	Novel	M	22	No	Y	Y	CaOx NL
UF30–4	c.653C>T	p.Ala218Val	het	0.899, C i.	0/23/276526	Novel	F	30	No			NL
URO-11	c.652G>A	p.Ala218Thr	het	0.784, C i.	2/100/27657 4	Novel	M	25	Yes	Y	Y	NL
URO-12	c.398C>T	p.Alal33Val	het	0.991, C.i.	3/1003/2771 88	Y	F	55	Yes	N	N	NL
URO-13	c.1225G>A	p.Gly409Ser	het	0.999, C.e.	0/27/246234	Novel	M	17	No	Y	Y	NL
URO-17	c.293dup	p.Ala99Argfs* 37	het	NA	NP	Novel	M	18	No	N	Y	NL
URO-29	c.398C>T	p.Alal33Val	het	0.991, C.i.	3/1003/2771 88	Y	M	40	Yes	Y	Y	NL
URO-88	c.293dup	p.Ala99Argfs* 37	het	NA	NP	Novel	M	40	Yes	N	N	NL
URO-117	c.1006+1G>A	Splice site	het	NA	0/16/246190	Y	F	17	No	Y	Y	NL
URO-133	c.398C>T	p.Alal33Val	het	0.991, C.i.	3/1003/277188	Y	M	8	Yes	N	Y	CaOx NL
URO-140	c.1006+1G>A	Splice site	het	NA	0/16/246190	Y	M	36	Yes	Y	N	NL
URO-169	c.398C>T	p.Alal33Val	het	0.991, C.i.	3/1003/277188	Y	M	52	Yes	Y	N	NL
SLC4A1 [Anion exchanger (Diego blood group)]
URO-35	c.286C>T	p.Arg96Cys	het	0.838, M.m.	0/73/246164	Novel	F	20	No	Y	N	NL
URO-178	c.911G>A	p.Arg304Gln	het	0.752, X.t.	0/23/237418	Novel	M	1	No	N	Y	NL
URO-196	c.1541G>C	p.Arg514Pro	het	0.948, C.i.	0/1/30782	Novel	F	40	No	Y	N	NL
SLC7A9 [Solute carrier family 7 (glycoprotein associated amino acid transporter light chain, bo,+ system), member 9]
URO-152	c.1076_1077dup	p.Ile360Valfs* 3	het	NA	NP	Novel	M	24	No	Y	N	NL
Recessive gene
APRT (Adenine phosphoribosyltransferase)
URO-77	c.292_293del	p.Trp98Glyfs* 11	Hom	NA	NP	Novel	M	3	No	Y	Y	NL

CaOx calcium oxalate; C.e. Caenorhabditis elegans; D.r. Danio rerio; F female; het heterozygous; Hom homozygous; M male; M.m. Mus musculus; N no; NA not applicable; ND No data; NL nephrolithiasis; NP not present; PPh2 Polyphen2-HumVar (http://genetics.bwh.harvard.edu/pph2/); X.t. Xenopus tropicalis; Y yes

^aFrequency reported as “number of homozygotes/number of variant alleles/total alleles” in Genome Aggregation Database (gnomAD)

Genotype relationships

We next evaluated whether molecular diagnoses correlated with geographic location (i.e. founder effects) and specific clinical features. First to assess for founder mutations, genetic mutations were mapped according to the specific town or city where each subject lived (Figure 2). Mutations were diffusely distributed through Punjab and surrounding provinces in Pakistan. No SLC34A1 mutations that were detected in multiple families clustered within specific towns or cities, suggesting no founder effects in our cohort.

Figure 2.

Regional distribution of 12 different mutations in 4 known NL genes (SLC34A1, SLC4A1, SLC7A9, APRT) detected in NL families in Punjab and 2 neighboring provinces of Pakistan. The SLC34A1 mutations p.Ala99Argfs*37, c.1006+1G>A, p.Ala218Val and p.Ala133Val were observed in multiple families (2, 2, 2, and 4, respectively) but did not aggregate within sub-regions of Pakistan, suggesting these are not founder alleles

We next assessed for relationships between the age of stone onset and molecular diagnosis. To begin, we plotted frequency of mutation detection versus age of onset, observing that mutation detection frequency increases with earlier age of onset (Figure 3): 1–20 years, 8 of 74 (11%); 21–40 years, 7 of 108 (7%); 41–60 years, 2 of 42 (5%); 61–80 years, 0 of 5 (0%).

Figure 3.

Distribution of age of onset in NL families in whom a causative mutation was detected in 3 dominant and 1 recessive gene. Frequency of subjects solved for causative mutations in one of the known NL genes are binned across ages of onset. Note that genetic diagnoses are most frequent during first two decades of life. Relative frequencies of “solved” versus “unsolved” cases were: 1–20 years, 8 of 74 (11%); 21–40 years, 7 of 108 (7%); 41–60 years, 2 of 42 (5%); 61–80 years, 0 of 5 (0%)

Finally, we compared mutation detection frequency to history of stone recurrence, family history of stone disease, and parental consanguinity (Suppl. Table 2). We observed a higher frequency of mutation detection in subjects with a positive NL family history (13/109, 12%) versus those with a negative family history (4/120, 3%) that was statistically significant (Suppl. Table 2). There were no correlations between frequency of mutation detection and stone recurrence or consanguinity status. These analyses indicate that certain clinical factors such as earlier age of onset and family history were associated with an increased likelihood of identifying a molecular diagnosis.

SLC34A1 mutations

As SLC34A1 mutations were identified in 12 families, we compared age of onset with severity of mutation (Suppl. Figure 6) but did not detect a clear genotype-phenotype relationship.

5 of 7 SLC34A1 mutations that we identified are novel, including 4 of 5 missense mutations (Table 1). SLC34A1 encodes the sodium-phosphate co-transporter NaPiIIa. We evaluated the impact of these mutations on NaPiIIa cell surface localization and phosphate transport. When co-expressed with wildtype SLC34A1, all missense mutant constructs localized appropriately on the cell surface by confocal microscopy (Suppl. Figure 7). When mutated SLC34A1 RNA was expressed in Xenopus oocytes, the 5 novel missense mutations showed defective phosphate uptake in isolation (Figure 4). An additional variant of unknown significance (c.365G>C; p.Ser122Thr) was identified in family URO131 (Suppl. Table 3) but did not impair transporter localization or phosphate uptake (Figure 4 and Suppl. Figure 7). To assess for dominant negative effects of our mutations, mutant SLC34A1 RNA was co-injected in oocytes with wildtype RNA at equimolar ratios. No mutant constructs suppressed phosphate transport below the activity of oocytes expressing 50% of wildtype SLC34A1 RNA, suggesting these mutations do not have dominant negative effects on wildtype NaPiIIa.

Figure 4.

Effect of Pakistani stone former SLC34A1 mutations on PO4 transport. Uptake of radioactively labeled phosphate was measured in Xenopus laevis oocytes upon injection with SLC34A1 wild-type and mutant RNA or water as a control. Level of uptake was normalized as a percent of wild-type SLC34A1. The above histogram represents the mean of 3 experiments with standard error of the mean shown by error bars. “Asterisk” Statistically significant difference (p<0.05) from Wildtype condition

Discussion

Nephrolithiasis is a medical condition that afflicts patients worldwide, requiring invasive interventions and hospitalizations, precipitating kidney injury, and leading to growing healthcare costs(Rule et al. 2009; Scales et al. 2012). We have previously demonstrated that a single-gene cause can be identified in 11–29% of cases depending on age of onset(Braun et al. 2016a; Daga et al. 2017; Halbritter et al. 2015). However, these previous findings were based on cohorts primarily recruited from the United States and Europe.

Here, we performed mutational analysis in a large cohort of Pakistani stone formers with nephrolithiasis at presentation and provide the first large scale evaluation of monogenic stone disease in the Afro-Asian stone belt. We sequenced the coding regions of 30 established NL disease genes and identified the likely causative mutation in 17 of 235 (7%). This study confirms our previous findings (Braun et al. 2016a; Daga et al. 2017; Halbritter et al. 2015) that mutations in NL disease genes can explain a substantial fraction of disease in stone forming individuals.

We detected a causative mutation more frequently in those with a positive family history of NL than those lacking a family history of NL (Suppl. Table 2), which was reported in renal stone disease previously (Halbritter et al. 2015; STECHMAN MICHAEL J. et al. 2007). This finding is consistent with a vertical pattern of inheritance, which we would expect from the mainly dominant mutations identified in our cohort. We also identified a causative mutation more frequently in those with early age of onset relative to those with later age of onset, consistent with previous studies (Braun et al. 2016a; Daga et al. 2017; Halbritter et al. 2015). In future studies, it will be important to consider these genetic diagnoses in correlation with additional clinical presenting factors and biomarkers that impact patient management including stone analysis, serum chemistries and urine metabolite studies.

The frequency of causative mutations in Pakistani stone formers overall was lower than previously observed in Halbritter et al. 2015 (7.2% here versus 15%). Both studies had similar percentages of pediatric cases (32% in this study versus 36%), which should not, therefore, bias mutation detection. Differences in the genetic landscape of Pakistani stone formers versus other populations may also contribute. Moreover, “unsolved” Pakistani subjects may have mutations in novel NL genes not yet identified from other populations or copy number variants (deletion or insertions) not detected by gene panel sequencing. Alternatively, environmental factors such as chronic dehydration could have a more prevalent role in NL pathogenesis in Pakistan, increasing stone risk sufficiently in those lacking a monogenic cause(Rizvi S.A.H. et al. 2002).

The most frequently mutated gene in our cohort was SLC34A1, which encodes the proximal tubule sodium-phosphate co-transporter NaPiIIa(Schlingmann et al. 2016). Recessive SLC34A1 mutations in humans cause infantile hypercalcemia with hypercalciuria and hypophosphatemia. Although more controversial, dominant mutations have been identified in subjects with NL or nephrocalcinosis and features of hypercalciuria and/or hyperphosphaturia (Braun et al. 2016a; Daga et al. 2017; Fearn et al. 2018; Halbritter et al. 2015; Prié et al. 2002; Rajagopal et al. 2014; Schlingmann et al. 2016). Our current study adds 5 novel SLC34A1 mutations and 12 novel families with dominant SLC34A1 NL disease to the previous 9 mutations and 10 families in the literature (Braun et al. 2016a; Daga et al. 2017; Fearn et al. 2018; Halbritter et al. 2015; Prié et al. 2002; Rajagopal et al. 2014; Schlingmann et al. 2016). The age of onset did not correlate with mutation severity and, in fact, varied within families with segregating heterozygous SLC34A1 mutations (Suppl. Figure 6, Table 1). This suggests that other factors (e.g. environmental triggers including diet) likely influence the age-related penetrance of these mutations.

Importantly, the 5 missense mutations in our cohort were functionally validated and harbor defective phosphate uptake activity (Figure 4) although not through dominant-negative effects on the wild-type protein (Suppl. Figure 8). Studies of Slc34a1 knockout mice suggest that dominant SLC34A1 mutations may rather cause stone disease through haploinsufficiency (Beck et al. 1998). Heterozygous null Slc34a1 knockout mice display abnormalities in calcium/phosphate homeostasis that increase NL risk including increased phosphate excretion and elevated 1,25-(OH)₂ Vitamin D₃ levels (Prié et al. 2001).

In our previous studies of consanguineous families with NL or other renal disease, we have identified autosomal recessive mutations in known disease genes in a substantial fraction of cases(Braun et al. 2016a, 2016b; Daga et al. 2017; Halbritter et al. 2015; Sadowski et al. 2015; Warejko et al. 2018). We were, therefore, surprised to identify a recessive cause of NL only once (0.4%) in this cohort despite the high frequency of Pakistani subject consanguinity (Suppl. Table 2). Because of the high degree of consanguinity, NL disease in Pakistani families may exhibit more complex genetic inheritance patterns, as was observed previously with limb-girdle muscular dystrophy on Réunion Island(Beckmann, 1996; Richard et al. 1995): single families may have multiple distinct recessive and/or dominant causative mutations, hampering monogenic mutation detection.

Another explanation is that the current cohort and our previous NL studies ascertained different forms of NL disease. The Pakistani NL subjects were uniformly recruited during hospitalization under the care of an expert urologist for a nephrolithiasis episode. In our previous reports, subjects were also recruited by nephrologists in specialty clinics and, in some cases, presented with broader and more severe clinical features in addition to NL such as failure to thrive, chronic kidney disease, acid-base disturbances, and electrolyte abnormalities(Braun et al. 2016a; Daga et al. 2017; Halbritter et al. 2015). The latter severe cases may therefore have distinct genetic etiologies (i.e. mutations in known recessive NL genes) from our Pakistani subjects with isolated NL.

Identifying the monogenic etiology in patients with nephrolithiasis can provide additional preventative, diagnostic, and therapeutic implications in concert with standard clinical diagnostics such as stone analysis and urine metabolite studies that currently guide patient care. For instance, in subject URO-77 bearing a recessive truncation mutation in APRT (Table 1, Supplementary Table 4), allopurinol therapy should be initiated to stabilize and/or improve renal function. Despite the declining costs of genetic testing, genetic diagnostics may not be economically feasible for all patients in resource limited settings such as Pakistan. Until genetic testing becomes more financially viable in these settings, it may be most relevant to cases with severe onset (i.e. early age, renal dysfunction), with strong family history, and where clinical testing does not reveal a precise etiology.

Conclusion

We examined the monogenic causes of NL in a novel geographic cohort and most frequently identified dominant mutations in the sodium-phosphate transporter SLC34A1 with functional validation.