Reply to: GNPAT variant is associated with iron phenotype in healthy Taiwanese women: A population without the HFE C282Y mutation

To the Editor

We read with interest the report by Hsiao et al. that Taiwanese women not selected for iron phenotypes, who were heterozygous or homozygous for the glyceronephosphate O-acyltransferase (GNPAT) p.D519G (rs11558492) polymorphism, had higher serum iron levels than GNPAT wild-type subjects. Women with GNPAT p.D519G also had higher levels of serum iron and transferrin saturation (TS) after a test dose of oral iron. The response to oral iron was not significantly different in Taiwanese women with or without TMPRSS6 rs855791, a polymorphism associated with reduced activity of the enzyme matriptase-2, an inhibitor of hepcidin transcription.¹ Moreover, the association of GNPAT p.D519G with baseline serum iron levels was independent of TMPRSS6 rs855791, age, and amount of menstrual blood loss. In our exome sequencing study, GNPAT p.D519G was also associated with high-iron phenotypes in Caucasian men with HFE p.C282Y homozygosity.² The important results of Hsiao et al. provide further support for a role of GNPAT in the regulation of iron metabolism in a population in which HFE p.C282Y is rare.

The allele frequency of GNPAT p.D519G in the Taiwanese cohort was 12%, similar to that in the general population of the 1000 Genomes Project (14%). Thus, GNPAT p.D519G is a relatively common determinant of the higher baseline serum iron levels and response of serum iron/TS measures to oral iron among Taiwanese women. In non-anemic young Caucasian women with hypoferritinemia, test doses of oral iron ≥60 mg increased circulating hepcidin levels and decreased fractional iron absorption on the subsequent day.³ Taken together, these observations suggest that GNPAT p.D519G alters hepcidin regulation and iron absorption in response to oral iron, although this is unproven. At present, there are no published observations of the possible effects of GNPAT p.D519G on hepcidin expression or iron absorption in Caucasian women with or without p.C282Y.

In our paper,² we did not propose that GNPAT p.D519G is a “HFE modifier,” but rather that it is a modifier of iron status in HFE p.C282Y homozygotes. Hsiao et al. suggest that GNPAT p.D519G acts independently of HFE on hepcidin regulation, and their results indicated that the effect of the GNPAT p.D519G allele does not involve interaction with matriptase-2. However, their observations do not exclude a possible interaction of GNPAT p.D519G with wild-type HFE protein or any other upstream regulator of hepcidin transcription. We observed that siRNA-mediated knockdown of GNPAT in HepG2/C3A cells markedly reduced HAMP mRNA expression but had no demonstrable effect on the BMP/SMAD pathway.³ Thus, taken together, our results and those of Hsiao et al. are consistent with the possibilities that GNPAT p.D519G either affects a gene or genes upstream of HAMP, or alters hepcidin expression directly. Elucidating GNPAT’s putative role in regulating hepcidin requires further study.

Abbreviations

GNPAT

glyceronephosphate O-acyltransferase

TS

Transferrin saturation