BOSTON – The recent identification of two gene mutations in a cohort of African Americans with nondiabetic kidney disease helps explain the disproportionately higher rates of kidney disease in this population and represents a disease-mechanism pathway that could lead to new treatments and possibly a cure, Dr. David J. Friedman said at the meeting.
Dr. Friedman of Beth-Israel Deaconess Medical Center, Boston, and his colleagues recently reported the association between two independent variants in the apolipoprotein L1 (APOL1) gene on chromosome 22 and focal segmental glomerulosclerosis (FSGS) and hypertension-attributed end-stage kidney disease in blacks (Science 2010;329:841-5). Not only do the investigators believe that APOL1 is very important to the understanding of nondiabetic renal disease in blacks, “we think the variants in the gene are among the most powerful that have been discovered to date,” Dr. Friedman stressed.
The disparity between the rates of end-stage renal disease (ESRD) in blacks and whites in the United States is “incredible,” Dr. Friedman stated, noting that the incidence rate is four to five times higher in blacks, according to the 2010 United States Renal Data System annual report. “People have been debating for decades whether the major cause of this disparity is genes or environment. No doubt both are important, but given how strongly this phenotype travels in families, I think we can say with certainty that genes play an important role.”
The APOL1 discovery came on the heels of an earlier association linking FSGS, nondiabetic ESRD, and HIV nephropathy in blacks with the MYH9 gene located on the same chromosome, Dr. Friedman explained. “This was quite striking, because we used to think of the three conditions as entirely different diseases, yet each one had exactly the same locus.”
Despite the strong association and several years spent looking for causal mutations using fine mapping sequences, the causal variants remained elusive until Dr. Friedman and his colleagues approached the problem from a different perspective. “We asked, 'How could any disease gene that's this deleterious become so common in a population?' We assumed there was something in this [genetic region] that was beneficial once upon a time to human evolution in Africa,” he said. Using mathematical techniques, “we realized that because of the effects of natural selection, the disease gene interval was much larger than anyone thought and probably contained at least five genes.” Consequently, the investigators tested new variants in other genes for association with renal disease in African Americans, looking specifically for variants that had not yet been documented, he said.
In a cohort of 205 African Americans with biopsy-proven FSGS and no family history of the disease and 180 African American control subjects, “we saw that variants in the neighboring APOL1 gene were much more strongly associated with renal disease, and unlike the MYH9 variants, which were located in regions of the gene that did not encode for protein, the APOL1 variants were protein-coding sequences.” The investigators determined that the top two variants almost always co-occurred on the same chromosome and each changed an amino acid somewhere on the protein. “We called this the g1 risk allele, and when we controlled for it, a new variant popped up, which we called the g2 allele,” he said. Controlling for both the g1 and g2 alleles, “the entire association of this region disappeared and there was no signal left for MYH9.”
The investigators also tested the genetic variants in hypertension-associated ESRD in a larger cohort of 1,030 African Americans with the disease and 1,025 geographically matched control subjects and found that the same two variants had a tremendous impact on the development of the disease. “When combined together, the P value was on the order of 10 to the minus 60, or 35 orders of magnitude greater than the very best MYH9 [result],” Dr. Friedman said. Surprisingly, he noted, we found that these disease variants follow a recessive pattern and together the odds ratio was on the order of 7-10, while the very largest effect sizes of the common variants that affect hypertension or diabetes will confer odds ratios of about 1.4-1.5.”
The APOL1 gene and these variants “tend to fall into a different category that we've all been familiar with in the past,” Dr. Friedman explained. “Most disease variants are either very rare with powerful effects or common with relatively modest effects. The APOL1 variants have a surprising combination of effect size and frequency such that 50%-60% of African Americans carry g1 and/or g2 risk alleles, and 50% are risk homozygous, meaning they are in the highest risk for kidney disease: That translates into about 3.5 million individuals.” Further, while the odds ratios for the more common forms of nondiabetic kidney disease in this population range from 7 to 10, “we're starting to see odds ratios in the range of 30 for diseases like HIV nephropathy.”