These two cases of COL4A5-related nephropathy were remarkable for various reasons. For instance, we found that immunolabeling for COL4A5 could be decreased or lost in Bowman’s capsule while preserved in the GBM, a finding that has never been reported in this condition before. Additionally, the ultrastructural abnormalities seen in both cases were consistent with Alport syndrome but also came with variable degrees of IgA deposition. Lastly, the p.Gly1170Ser mutation was associated with a unique yet similar histopathological portrait in both mother and son.

It is intriguing that in contrast to Bowman’s capsule, the GBM showed preserved expression of COL4A5 in the two family members and did so even if it was ultrastructurally abnormal. One explanation could be that the COL4A3-4-5 trimers were more efficiently processed than the COL4A5-5-6 trimers by harboring only one mutation per trimer instead of two. This explanation would be in keeping with the observation that G-changing missense mutations in several of the COL4A5 G-X-Y repeats have been associated with more favorable outcomes than truncating mutations but still found to alter the structure of the GBM [5, 6].

As mentioned, Bowman’s capsule in the mother showed a segmental decrease in COL4A5 expression on the second renal biopsy. This other finding was not unexpected in that the gene defect affected chromosome X and was thus bound to be inactivated in only a subgroup of cells within the kidney. It could also be for the same reason that COL4A5 expression in Bowman’s capsule appeared normal in the mother on the first biopsy, i.e., that it was preserved because the glomeruli sampled were from a region where X-inactivation was favorable.

In view of the gene defect uncovered, the other interesting aspect of our report was the presence of IgA deposition in both cases and mesangiopathy in the mother. While a recent exome-wide analysis of familial IgAN from 46 pedigrees has already led to the identification of gene defects in COL4A5 [7], it has only been seldom reported afterwards that the two conditions could coexist. Our findings should thus be seen as a timely contribution to this emerging context. They also suggest that IgAN in Alport syndrome can take a while to manifest in its typical form as it was indeed more manifest in the mother.

As for the reason that could explain why a gene defect in COL4A5 can lead to IgAN, some authors have argued that GBM thinning could simply allow IgA to translocate more readily from blood to mesangium and elicit a pathological response at this new location [3]. Such a scenario would be congruent with the observation that IgAN has also been described with gene defects in COL4A3 and COL4A4. However, one would then wonder why mesangial deposition of the other immunoglobulins or of C3 is not a very common finding in Alport syndrome.

In our view, a more plausible explanation would be that the susceptibility to develop a COL4A3-, COL4A4- or COL4A5-related form of IgAN is driven by coinherited gene polymorphisms. For instance, this renal lesion was seen in our blood-related cases while it has not previously been associated with the COL4A5 p.Gly1170Ser mutation. Along the same line, many individuals who were found to carry pathogenic mutations in COL4A3, COL4A4 or COL4A5 were also blood-related [7].

As it stands, there is evidence to suggest the involvement of IgAN-associated polymorphisms in Alport syndrome. In particular, a genome-wide association study of sporadic IgAN has led to the identification of a susceptibility locus (2q36) that encompasses the COL4A3 and COL4A4 genes [8] and that could thus harbor a polygene through which IgA1 processing is coordinated. In the case of COL4A5, VSIG1 could correspond to another polygene of interest given that it is involved in gastrointestinal immune regulation and is also localized in locus Xq22.3 [9].

A question that will need to be addressed in the course of future investigations is whether COL4-related IgAN could be treated as idiopathic IgAN, especially if galactose-deficient IgA1 were to be involved in disease progression. This question will become particularly relevant if complement pathway inhibitors or B-cell-directed strategies in idiopathic IgA were eventually found to be more effective and better tolerated than the previously described approaches.

In conclusion, we have reported for the first time two cases of X-linked Alport syndrome where the only structure affected by a decrease in COL4A5 immunolabeling was Bowman’s capsule and where neither COL4A3 nor COL4A4 could have thus been at cause. The cases described were also striking in being blood-related and both showing glomerular IgA deposition. These observations suggest that the molecular and clinical impacts of the mutation identified were conditioned in mother and son by a coinherited gene defect in or near the Xq22.3 locus.