NephMadness 2015: Genetic Nephrology Region

Conall O'Seaghdha, MB MRCPI; Paul Phelan, MD

Disclosures

March 02, 2015

Editorial Collaboration

Medscape &

In This Article

GWAS in Nephrology vs Next-Generation Sequencing

These conference rivals each have a loyal following who will relish this first-round contest. GWAS promised much when it exploded onto the scene some years back, and many predicted several national championships which have failed to materialize. There is similar enthusiasm for Next-Generation Sequencing at present, a team that has huge aspirations and expects silverware this season.

GWAS in Nephrology

The human genome consists of approximately 3 billion nucleotides of DNA sequence. Areas of variance at an individual nucleotide, termed SNPs, occur across the genome at intervals of about one per 300 base pairs of DNA. SNPs in close physical proximity are more likely to be inherited together as part of a group (haplotype). This phenomenon, referred to as linkage disequilibrium, allows for one SNP to serve as a proxy for the presence of other SNPs in that haplotype.

This is the concept of a "tag SNP" and obviates the need for individual genotyping of every variant. This is the principle behind GWAS. We have witnessed an explosion of GWAS for complex traits including renal function (eGFR), CKD, and hypertension. GWAS are usually designed to detect relatively common SNPs (minor allele frequency > a pre-determined level, for example 5%).

GWAS in Nephrology have not proven to be as clinically useful as initially hoped. Like GWAS in other complex diseases, many variants with tiny effect sizes have been uncovered, but these variants only explain a small proportion of total heritability of the disorders.

Large consortiums have been created to try to provide the necessary power to detect more variants but overall, the effect sizes remain small. Examples of these GWAS meta-analyses include the CHARGE consortium (n=29,163), the Global BPgen Consortium (n=34,433) and ICBP-GWAS (n=69,395 with validation in a further 133,661 individuals) for hypertension. ICBP-GWAS reported 29 SNPs independently associated with blood pressure, but together they explained only 0.9% of the BP variation in the cohort. This reflects the genetic complexity of the trait. Large renal function GWAS collaborations have also been formed and have demonstrated similar findings of numerous SNPs with tiny effect sizes. These findings may one day lead to useful risk scores for CKD or hypertension but demonstrate the limited clinical relevance of individual or a small group of SNPs.

Despite these limitations, genome-wide approaches have proven useful in our field. The discovery of the APOL1 variants came from initial identification of variants in the nearby gene MYH9 gene on chromosome 22 (see Team APOL1). We have also learned much about UMOD nephropathy and the function of Tamm-Horsfall protein (Uromodulin) from a renal function GWAS in which UMOD variants popped up as being genome-wide significant for eGFR (see Team AD Tubulointerstitial Nephritis).

It was mentioned above that this locus appears to confer shared risk for both hypertension and kidney disease (and also cardiovascular events). Another example of shared risk loci comes from the ICBP-GWAS study where a variant in the phospholipase C epsilon gene (PLCE1) was associated with blood pressure variance. A coding missense mutation in PLCE1 has been described causing steroid-resistant FSGS.

Variants in SHROOM3 have been consistently associated with CKD/eGFR in large GWAS, but any potential mechanism remained unclear. Due to this finding, a group from Mount Sinai performed a set of experiments in a transplant cohort which was hugely insightful. They genotyped transplant donors for the SHROOM3 variant which correlated with increased SHROOM3 protein expression and allograft fibrosis on protocol biopsies. It also associated with eGFR in the recipient. They identified the risk allele to be located in a sequence for the transcription factor TCF7–L2, which enhanced SHROOM3 expression that in turn regulated TGF-B induced renal fibrosis. In a mouse model, SHROOM3 knockdown strongly abrogated interstitial fibrosis. This exciting research with real translational potential was made possible by SNP associations in GWAS cohorts of the general population and firmly demonstrates the power of this approach.

Another demonstration of the power of GWAS comes from a study in >20,000 individuals of European and Asian ancestry which shed light on the complex genetic architecture of IgA nephropathy. Several genome-wide significant variants were identified which were predominantly located in pathways of immunity and inflammation, including variants with overlapping susceptibility to autoimmune disorders such as inflammatory bowel disease. The study demonstrated that disease onset was related to the number of risk loci present, although they still only explained a small proportion of the variance in disease onset. A striking observation was the association of the genetic risk score with pathogen diversity, particularly helminth diversity. Helminths are common in the soil in Asia and may explain the increased incidence of IgA nephropathy in some geographical areas and the known association of mucosal infections as a trigger for IgA nephropathy.

Many SNPs from GWAS will have very small p values but will not reach genome-wide significance when corrected for multiple testing. This may be due to overly stringent criteria for genome-wide significance or underpowered studies.

One method of using these variants in a clinical useful way is to perform pathway analysis not limited to only genome-wide significant or replicated variants. A recent paper in JASN employed pathway analysis on GWAS data examining the development of new-onset diabetes mellitus after transplant (NODAT) post renal transplantation. This study implicated β-cell dysfunction in the pathophysiology of NODAT, contrary to traditional thinking that the etiology was merely insulin resistance.

Another example is from the Wellcome Trust Case Control Consortium where no genome-wide significant associations were observed for hypertension in the original study. However, pathway analysis revealed interconnected networks in dopamine signaling including genes coding for the AMPA, NMDA, and GABA-A receptors. This suggests that the regulation of vascular smooth muscle tone is important. There is inherent bias in pathway analysis, however, as it is reliant on accuracy and depth of input from the pathway databases, but it does provide an additional use of GWAS data including non-genome-wide significant SNPs.

GWAS have revolutionized the search for genetic influences on complex diseases but it is far from a panacea. As a technique, GWAS are not designed to fully uncover the interplay of multiple genetic and environmental factors which cause CKD and hypertension. Genetic variant discovered by GWAS mostly have tiny independent effect sizes and none are likely to be obligatory for the phenotype to occur.

Team GWAS is a hot and cold side who can beat anyone on their day but also may succumb to unheralded opposition. This unpredictability makes them a fascinating team to follow. Their recent success will give them confidence. Will this be enough against the next-generation sequencing new kids?

Next-Generation Sequencing

Next-Generation Sequencing, including whole-exome sequencing (WES) and whole-genome sequencing (WGS), promises in-depth coverage of the exome/genome with improved coverage of rare variants and copy number variants (CNVs; large insertions and deletions). WES involves sequencing all exons, the coding proportion of the genes, which make up about 1% of the genome and where presumably most disease-causing variants lie. Deep sequencing projects, such as 1000 Genomes, demonstrate that rare variants, which are usually not covered in GWAS, constitute the majority of polymorphic sites in human populations.

An early example of the use of WES is the identification of a potassium channel mutation in the development of primary hyperaldosteronism, one of the more frequent causes of secondary hypertension. WES has facilitated gene discovery for several kidney diseases including FSGS and VUR (see Team VUR and Familial FSGS). It has also helped identify novel phenotypes for known genes in the case of COL4A mutations, which cause hereditary nephritis, presenting as familial FSGS. In nephronophthisis, an AR ciliopathy which causes tubulointerstitial nephritis, Next-Generation Sequencing has expanded the breadth of known causative genes. There are now 17 NPHP genes described, but despite this, many remain undiscovered.

The idea of WES is to remove much of the redundancy of the genome and maximize efficacy and cost-effectiveness. However, the sphere of epigenetics has taught us that non-coding portions of the genome can be vitally important, potentially heritable, and can influence expression of the genes and therefore phenotypes. Moreover, WES is not a perfect method for new gene discovery in familial disease, and exonic regions may still prove difficult to sequence with the potential to miss causative variants. This is highlighted by the problems in sequencing MUC1 as a cause of AD interstitial nephritis due to multiple repetitive regions in and around the gene. Successful sequencing relies on multiple reads of the regions of interest, with depths of <10X showing inconsistent call rates.

The major challenge with Next-Generation Sequencing lies in identifying the specific disease-causing variants from all the benign variants we carry. WES in one individual will typically reveal approximately 20,000 variants, and even when sequencing >1 individual in a family, multiple potentially pathogenic variants will be shared between family members. Filtering methods and in silico techniques may predict if a variant is likely to be damaging but have the potential also to inadvertently disregard pathogenic mutations.

The issue is compounded in African American families which have known increased genetic diversity. African American families may have many more variants uncovered by WES, compared to non-African ancestry families. Therefore, even bigger pedigrees are needed to identify disease causing mutations.

Aside from research settings, the utility of WES in clinical practice for precise molecular diagnosis is unknown. The spectrum of childhood nephrotic syndrome is an area where it may be useful, as several genes forming a large proportion of cases have been identified.

Data on likely responsiveness to certain treatments is a potential indication for testing. Other indications include transplantation, both for assessment of potential recurrence and for disease in living related donors.

The use of exon sequencing of 27 genes known to cause steroid resistant nephrotic syndrome (SRNS) that manifested before 25 years of age has been reported. A single-gene cause was detected in 29.5% (526 of 1783) of families, with younger presentations more likely to have a monogenic cause identified. A UK cohort of 36 patients (all <16 years at onset) with SRNS detected a pathogenic variant in 70% of familial causes and 15% of sporadic cases.

WGS is becoming more affordable and it is likely that very soon it may replace WES in the investigation of genetic diseases. So will WGS solve these issues that we have with WES? As sequencing is not confined to the exome, gene regulatory regions, enhancers, and promoters will be covered. It will certainly add more complexity by sequencing the entire genome and will uncover millions of variants in each individual sequenced. Sophisticated filtering and bioinformatic methods will need to be employed to identify likely pathogenic variants.

Current issues with WGS include a lower detection rates for CNVs versus single-nucleotide variants and incomplete concordance between different sequencing technologies. WGS will potentially provide information about countless medical conditions, many of undetermined significance. The huge volumes of data generated by such technologies will potentially greatly aid genetic interrogation of kidney diseases but will provide logistical and ethical challenges which must be overcome.

With a strong preseason behind it, Next-Generation Sequencing expects to win this contest with some to spare. GWAS has gone under the radar so far this season but has earned some notable recent victories and has a lot to prove to its doubters. This one will go down to the wire.

Dr. O'Seaghdha earned his medical degree from University College Dublin in Ireland and did his nephrology fellowship training in Ireland and subsequently in Sydney, Australia. He also completed the Harvard fellowship in nephrology. During his fellowship, he worked as a clinical researcher for three years in the Framingham Heart Study where his interests were in the epidemiology of CKD in the general population, novel biomarkers of CKD, and the genetic epidemiology of kidney disease. He was also editor of the nephrology blog Renal Fellow Network during this time. After his fellowship he was an attending physician and Transplant Nephrologist at Massachusetts General Hospital and Instructor in Medicine in Harvard Medical School. He returned to Ireland in 2013 to take up his current position as Consultant Nephrologist and Transplant Physician at Beaumont Hospital in Dublin and as Honorary Senior Lecturer in Medicine in the Royal College of Surgeons and in Trinity College Dublin. He is also the National Specialty Director for Nephrology higher training in Ireland.

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as:

processing....