Increased susceptibility to iron deficiency of Tmprss6-haploinsufficient mice

To the editor:

We read with interest the paper by Finberg et al¹  that describes the Hjv^−/−Tmprss6^−/− double mutant mouse and characterizes the hematologic phenotype of Tmprss6^+/− mouse. TMPRSS6 encodes a liver-expressed type II transmembrane serine-protease, that, cleaving the membrane Bone Morphogenetic Protein (BMP) coreceptor hemojuvelin (HJV),²  down-regulates the Sons of Mothers Against Decapentaplegic (SMAD)–dependent hepcidin expression. This function is essential for erythropoiesis, since loss of TMPRSS6 activity in both humans³  and mice^4,5  causes iron-deficiency anemia.

Our recent work on Tmprss6^+/− mice complements the results of Finberg et al.¹ Tmprss6^+/− mice develop normally and are phenotypically indistinguishable from wild-type littermates. At 3 weeks of age they have lower hemoglobin (Hb), red blood cell (RBC) counts, and mean corpuscular volume (MCV; Figure 1A) and higher hepcidin and Id1 levels compared with wild-type (Figure 1B). Adult Tmprss6^+/− mice have Hb and RBC similar to wild-type but reduced MCV, as reported.¹  Transferrin saturation, which shows a trend toward reduction in 8-week-old animals in the previous study,¹  is significantly reduced in our study. Tmprss6^+/− mice of this age show mildly elevated Id1 and hepcidin production, inappropriately high for their low levels of transferrin saturation (Figure 1B-C). C-reactive protein mRNA levels are similar in Tmprss6^+/− and wild-type mice (data not shown), ruling out a potential contribution of inflammation to hepcidin expression. In agreement with the observation of a more severe anemia in young versus adult patients in a large iron-refractory iron deficiency anemia (IRIDA) pedigree,⁶  our results show that young Tmprss6^+/− mice, which have increased iron demands, are indeed more iron-deficient than adult mice.

Female mice at 18.5 gestational days that have high iron demands, have lower liver iron concentration (LIC) than wild-type, even if hematologic parameters are similar.¹  To assess the effect of iron restriction, we fed 4-week-old animals of both genotypes with normal or iron-deficient diet for 4 weeks. The iron-deficient diet induced a more pronounced decrease of transferrin saturation in Tmprss6^+/− than in wild-type mice (Figure 1C) and a trend toward a more severe microcytic anemia, although the difference reached statistical significance only for MCV (Figure 1A). However, as in pregnant mice,¹ Hamp was strongly down-regulated in both genotypes (Figure 1B), suggesting that iron deficiency overrides the SMAD pathway activation observed in Tmprss6-haploinsufficient mice.

We have not measured LIC that was found decreased in 8-week-old Tmprss6^+/− mice.¹  However, transferrin receptor 1 (TfR1) mRNA, a LIC indirect measure, is significantly increased only in adult Tmprss6^+/− mice fed a standard diet (Figure 1B). Likely high basal TfR1 mRNA levels in young wild-type mice that have high iron requests and in adult wild-type animals on an iron-deficient diet blunt the difference with Tmprss6^+/−.

In conclusion, systemic iron homeostasis is mildly compromised in Tmprss6-haploinsufficient mice, which are prone to iron deficiency when iron demands are high for body growth and erythropoiesis expansion or when dietary iron is restricted. In humans, genomewide association studies show that common TMPRSS6 variants influence iron parameters, Hb, and erythrocyte traits.^7–10  All these findings suggest that susceptibility to iron deficiency may be modulated by TMPRSS6 mutations even at the heterozygous state.