New Ways of Finding New Genes for Old Diseases

In this issue of JASN, Besse et al.¹ report the identification of loss of function variants in ALG9 as a novel cause of renal and hepatic cysts, ranging from few to a polycystic-like phenotype, in a cohort of genetically uncharacterized patients with autosomal dominant polycystic kidney disease (ADPKD) and autosomal dominant polycystic liver disease (ADPLD) genes. ALG9 encodes α-1,2-mannosyltransferase and was a strong candidate gene for these diseases, because other defects in the endoplasmic reticulum (ER) protein biogenesis pathway result in overlapping phenotypes.² Biallelic mutations in ALG9 are associated with congenital disorder of glycosylation type 1L, one of a group of diseases also associated with polycystic kidney disease (PKD).³ Its role in cystogenesis is confirmed using functional assays of polycystin-1 (PC1), the PKD1 gene product, and protein maturation and uniquely, by using a “genotype”-first approach of identifying loss of function ALG9 mutations in individuals with renal cysts from a large population-based cohort with linked exome and electronic health care record data. In addition to adding to our knowledge of the genetic architecture and molecular pathology of renal cystic disease, this study elegantly demonstrates how such a “genotype”-first approach can be used to unambiguously assign and refine a gene-disease association when the initial observation is made in a singleton or few individuals.

The majority of individuals with ADPKD harbor a mutation in either PKD1 or PKD2.⁴ These genes encode a large ion channel complex localized to the primary cilium of uncertain function. Studies on the basis of linkage analysis or direct sequencing in well characterized cohorts suggest that approximately 80% will have an identifiable pathogenic or likely pathogenic variant in PKD1 and that approximately 15% will have an identifiable pathogenic or likely pathogenic variant in PKD2, with the remainder having no identifiable mutation in either gene.⁵ In parallel with advances in and access to clinical and research molecular testing, renal imaging has defined the natural history of ADPKD through measurement of renal size, renal volume, and cyst burden. Mutations in PKD1 are associated with larger absolute kidney size compared with mutations in PKD2 and predict a faster rate of decline in renal function.⁶ This is reflected in ESKD occurring some 20 years earlier in individuals with a PKD1 mutation compared with PKD2.⁵ Loss of function mutations in PKD1 also predict a more rapid decline in renal function compared with missense mutations.⁵ However, there is considerable clinical variability in disease progression between and within families that is not accounted for simply by these genic and allelic effects.

Mutations in PKD1 and PKD2 are also associated with polycystic liver disease. In contrast, ADPLD is associated with minimal renal cystic disease and greater genetic heterogeneity. Mutations in PRKCSH, SEC63, LRP5, ALG8, SEC61B, and PKHD1 have been identified in ADPLD, revealing that disruption of the ER protein biogenesis pathway is a principle mechanism in hepatic cystogenesis.²

Imaging in ADPKD has also defined a group of patients with atypical renal cystic disease in whom the mutation detection rate is less than that in those with typical disease.⁷ In atypical disease (unilateral, asymmetric, segmental, or lopsided cystic disease or unilateral or bilateral reduction in renal size), the mutation detection rate may be up to 45%, suggesting that other genes are responsible for multiple renal cysts.⁷ This is well documented for rare conditions, such as von Hippel Lindau disease, tuberous sclerosis, COL4A1-related disease, orofacial digital syndrome type 1, HNF1B-associated renal disease, and autosomal dominant tubulointerstitial kidney disease, where other clinical features are typically present to suggest the diagnosis. However, clinical presentations, including multiple renal cysts in normal-sized kidneys with or without liver cysts, especially at the extremes of age with no clear family history can present a diagnostic dilemma with implications for monitoring, treatment, family screening, and reproductive counseling.

More recently, additional genes have been identified in individuals with renal and/or hepatic cysts. These include GANAB and DNAJB11.^8,9 Both studies used genetically unresolved individuals with ADPKD and ADPLD and either whole-exome or targeted next generation sequencing to identify pathogenic mutations in multiple families, but these still comprised a small proportion of those in the discovery cohorts. GANAB encodes glucosidase II subunit-α, a functional partner of glucosidase II subunit-β (the PRKCSH gene product), and was shown to regulate PC1 maturation. DNAJB11 encodes an ER protein involved in protein folding and assembly and was also shown to regulate PC1 maturation in vitro. This further confirms that disruption of the ER protein biogenesis pathway is a common mechanism in renal as well as hepatic cystogenesis.

Besse et al.¹ identified two patients with loss of function mutations in ALG9 from a discovery cohort of 122 patients with genetically unresolved ADPKD or ADPLD using in vitro assays to confirm disruption of PC1 processing. To obtain further support for ALG9 as a cause of ADPKD or ADPLD, individuals with a predicted loss of function ALG9 mutation, no other known ADPKD or ADPLD gene mutation, preserved native kidney function, and accessible abdominal imaging were identified from the MyCode Community Health Initiative.¹⁰ This is a large unselected population–based cohort of >190,000 individuals with combined longitudinal electronic health record and whole-exome sequencing data. From this study, an additional 21 patients were identified from >90,000 exomes, 14 of which had at least one imaging study. Compared with controls, the majority of patients had four or more renal cysts over the age of 50 years. There was extreme clinical variability as has been seen previously in ADPLD and in the rare causes of ADPKD with a low risk of developing ESKD. Therefore, ALG9 can be added to GANAB and DNAJB11 as a cause of renal and hepatic cystic disease, often with atypical and mild features. Although accounting for a very small proportion of individuals suspected of having ADPKD or ADPLD, testing these genes in parallel with PKD1 and PKD2 will provide an opportunity for those who present with atypical disease to be given more precise information about clinical and family risk.

The use of large integrated datasets of genomic and health information has proved extremely valuable in supporting genome-wide and phenome-wide association studies, with the resultant benefit in understanding disease mechanisms, refining disease classification, and identifying drug targets to improve health outcomes. This has been particularly important for identifying genetic determinants of kidney function and CKD.¹¹ In addition to MyCode, many other similar studies have been established, including electronic medical records and genomics, the Million Veteran Program, the National Institutes of Health All of Us Research Program, and the United Kingdom Biobank. The ability to test rare and private mutations in addition to common variation for association with the numerous and specific phenotypes, including those derived from imaging studies, contained within an individual’s electronic health record is now possible with the rapid reduction in the cost of whole-exome and whole-genome sequencing, and it has the promise of fundamentally altering the way that medicine is delivered.

As this study reveals, these integrated datasets can also be simply and elegantly used for single-gene rare variant–single-phenotype association studies to rapidly accelerate gene discovery as well as genome-wide and phenome-wide association studies. However, there are some limitations to the former. There may be a lack of family history information, and patients may not be recallable for additional clinical and research studies. Therefore, we currently do not know the penetrance and variability of heterozygous ALG9 mutations in additional family members or why the patients identified initially underwent abdominal imaging, for example.

However, the identification of ALG9 as a cause of renal and hepatic cysts adds to our knowledge of the mechanisms underlying cystogenesis and provides opportunities to extend genetic testing to individuals with atypical disease in whom current testing is negative. It also identifies ALG9 as a potential modifier of disease expression due to PKD1 or PKD2 mutations. Further studies are required to address this directly.

The ability to confirm gene disease associations and refine genotype-phenotype correlations using resources, such as MyCode, will significantly enhance the delivery of precision medicine in the general nephrology clinic.

Disclosures

Dr. Sandford received research funding from Intercept Pharmaceuticals and has received advisory board and educational honoraria from Otsuka Pharmaceuticals Europe Ltd.

Funding

The University of Cambridge has received salary support in respect of Dr. Sandford from the National Health Service in the East of England through the Clinical Academic Reserve.