In the Seychelles' sample, we measured renal function using the accepted gold standard, inulin clearance, and found that ABCB1 genetic variants are associated with the haemodynamic profile, i.e. GFR, ERPF and RVR in these individuals of African descent with a positive family history of hypertension. We confirmed this association of the ABCB1 gene with renal function, estimated using the simplified MDRD equation, in a large population-based study of Caucasian subjects. This is, to our knowledge, the first time that the ABCB1 gene has been shown to be associated with renal function in the general population.
As this is not a true replication study (i.e. different ABCB1 marker variants were found to be associated with renal function in the two populations), these results are preliminary and need to be confirmed in other studies. This may result from the fact that the Seychelles and Swiss populations have different ethnic backgrounds and different linkage disequilibrium structures. In addition, the Seychelles sample was enriched in hypertensive individuals, whereas the Swiss one was not, which may further explain differences in findings between the two populations. We corrected our models for ascertainment (i.e. for the fact that recruitment was constrained to include two hypertensive siblings in each family). Such a correction aims to determine what the results would have been had the participants not been ascertained this way. This correction results in better generalizability of our observations.
In the African sample, (1) analyses restricted to subjects with at least 3 grand-parents of African descent led to very similar results; and (2) we confirmed the association with GFR estimated using MDRD in an analysis that included 119 additional African subjects. These results are thus unlikely to be due to population stratification. The confirmation of an association in the Caucasian sample with variants in the same gene adds further evidence against a spurious association.
The frequency of the 2677T allele typically varies between 0–10% in subjects of African descent, 39–46% in Caucasians and 36–44% in Asians , whereas the frequency of the 3435T allele typically varies between 16–27% in subjects of African descent, 48–57% in Caucasians, and 41–66% in Asians.  Of note, the frequencies of the 2677T (31%) and 3435T (54%) alleles in our sample are higher than those reported in other groups of African descent. The Seychelles' population is a relatively young population (the islands were first inhabited at the end of the 18th century) of small size (currently 84,000 inhabitants) that has undergone a population bottleneck during the middle of the 19th century followed by rapid growth (census data). Genetic drift may explain why the allele frequency differs in Seychelles versus other African populations (in Africa or elsewhere). It is not clear to what extent our observations may reflect the facts that homozygosity at the 3435T allele is a marker of African descent and African ethnicity is a risk factor for developing chronic kidney disease. Given the large interethnic differences in allele frequencies observed for ABCB1 variants, it is important that the associations of ABCB1 variants with renal function are explored further in other ethnic groups.
The TT genotype of the C3435T polymorphism is associated with a significant reduction in PGP expression in various cell types, including renal proximal tubular cells. [26–29] Although synonymous (i.e. not leading to an amino-acid change in PGP), the 3435 C>T variant influences mRNA expression by acting on its stability  and on the substrate specificity of PGP . The 2677 G>T variant is functional and leads to an Ala893Ser amino acid change in PGP. The 3435 C>T and 2677 G>T polymorphisms are in strong linkage disequilibrium. The fact that the 2677 G>T and 3435 C>T ABCB1 variants (and markers in strong linkage disequilibrium with these variants) were not associated with estimated GFR in the CoLaus study suggests that the true causal variant(s) is (are) in linkage disequilibrium with the variants genotyped in the Seychelles sample or that the two populations differ in either modifier genes or environmental components (e.g. the Balkanic toxin [32, 33]) affecting the impact of the causal variant. Further studies are needed to clarify how ABCB1 genetic variants are associated with renal function.
The association reported here between the ABCB1 gene and renal haemodynamics does not necessarily mean a causal relationship. However, several independent lines of evidence of an association with renal function strengthen our confidence that these results are not spurious. First, several experimental findings in vitro and in vivo [5, 8, 9] and findings in humans [11, 12] suggest that either PGP itself and/or ABCB1 genetic variants play a role in post-transplantation nephrotoxicity induced by calcineurin inhibitors, such as ciclosporine. Although inconsistent (i.e. the T allele is associated with nephrotoxicity in one study, but with nephroprotection in the other), these findings nevertheless suggest that PGP influences renal function via exogenous substances. Hence PGP may influence renal function via the control of intracellular concentration of a known toxin in the tubular epithelial cells. By analogy, we put could speculate that PGP influences renal function via the control of intracellular concentration of endogenous substances, by an as yet unknown mechanism. Second, Atanasova et al  found the ABCB1 haplotype 2677G/3435T to be protective against Balkan endemic nephropathy, a slowly progressive nephropathy associated with a high incidence of epithelial tumors of the upper urinary tract. Because of the very specific and stable geographical localization of this nephropathy, it has been proposed that an environmental toxin could play a role in its etiology.  However, the familial aggregation of this nephropathy suggests that genetic factors are also likely to be involved.  Third, endothelins are important regulators of renal blood flow and glomerular filtration rate  that can influence PGP-mediated transport [35, 36]. Bosentan, an endothelin receptor antagonist, activates the pregnane X receptor , which itself activates ABCB1 expression .