Maternal Tmprss6 is required for hepcidin suppression and fetal health Open Access

TO THE EDITOR:

Iron requirements are markedly increased during pregnancy to support maternal expanded erythropoiesis and fetal development.¹ Hepcidin, the master regulator of systemic iron homeostasis, is suppressed in mouse and human pregnancy, facilitating dietary iron absorption and transfer to the fetus.^2,3 The mechanism of hepcidin suppression in pregnancy remains unknown, although changes in maternal iron status are thought to contribute.^4,5 Tmprss6 is a key inhibitor of BMP6-hemojuvelin-BMP receptor interactions, particularly in iron deficiency. Tmprss6 has been shown to act on several downstream components of the hepcidin induction pathway, independently of its proteolytic activity.^6,7 Tmprss6 is a critical effector of hepcidin suppression. Tmprss6 knockout mice are unable to sustain pregnancy⁸; thus, it has not been possible to define the effects of disrupting this system during pregnancy on hepcidin expression and fetal outcomes. Using a small interfering RNA (siRNA) targeting maternal hepatic Tmprss6 in a mouse model of pregnancy, we sought to establish whether inhibition of maternal hepatic Tmprss6 abrogated hepcidin suppression in pregnancy and, furthermore, whether this has subsequent downstream impacts on fetal outcomes.

Mice were used under the approval of the Walter and Eliza Hall Institute of Medical Research's Animal Ethics Committee (approval 2020.034). Experimental details are summarized with each experiment and detailed in the supplemental Materials, available on the Blood website. Briefly, liver RNA was isolated using the ISOLATE II RNA Mini Kit (Bioline). RNA was reverse transcribed to complementary DNA (SensiFAST cDNA synthesis kit; Bioline), and gene expression levels were measured by reverse transcription (RT)-qPCR using SensiFAST SYBR No-ROX kit or SensiFAST Probe No-ROX kit (Bioline) on a LightCycler 480 II (Roche). Sequences of primers for RT-qPCR are provided in supplemental Materials (supplementary Table 1). Serum hepcidin was measured by enzyme-linked immunosorbent assay (Intrinsic LifeSciences). Tissue nonheme iron was measured as previously described.⁹ Serum iron was measured using the Pointe Iron/Total Binding Iron Capacity (TIBC) reagent set on serum separated on the day of collection and frozen at –80°C. Hemoglobin concentration and mean cell volume (MCV) were evaluated with the Advia2120i on EDTA anticoagulated blood.

First, we determined the time course of hepcidin suppression in mouse pregnancy. We measured hepatic hepcidin (Hamp) messenger RNA (mRNA) expression and serum hepcidin (HAMP) levels in C57Bl/6 mice receiving standard chow (∼180 mg/kg iron) on alternate days from time point embryonic day 6.5 (E6.5) onward. Hamp mRNA suppression occurred at a time point after E8.5 and before E10.5 (Figure 1A).

Next, we sought to establish whether hepcidin regulation in pregnancy is influenced by the inhibition of Tmprss6. TMPRSS6 is a membrane protease that likely cleaves hemojuvelin, a critical component of the BMP receptor complex required for canonical BMP6 and BMP2 signaling and for homeostatic hepcidin regulation in response to iron stores.^10,Tmprss6 knockout mice exhibit upregulated hepcidin expression and systemic iron depletion, including microcytic anemia, and are infertile. Thus, investigating the effects of Tmprss6 inactivation in pregnancy has not been possible using this knockout model.⁸ We used an in vivo RNA inhibition approach to overcome this constraint. We treated mice with a subcutaneous injection of 5 mg/kg of either a N-acetylgalactosamine (GalNAc)-siRNA conjugate targeting hepatic Tmprss6 (Tmprss6 siRNA; Silence Therapeutics GmbH, Berlin, Germany) or a nontargeting control (NTC) siRNA conjugate (complementary to luciferase RNA).¹¹ Treatment commenced at least 1 day before mating and was administered every 21 days, for up to a maximum of 3 doses, depending on the timing of successful mating (Figure 1B).

Treatment with Tmprss6 siRNA (compared to NTC) resulted in effective knockdown of hepatic Tmprss6 mRNA expression in unmated mice and across pregnancy (Figure 1C), exceeding 90%. As expected, hepcidin was suppressed in NTC-treated mice at E14.5 compared with unmated NTC-treated controls, but in Tmprss6 siRNA–treated mice, hepcidin remained elevated, measured by both mRNA (Figure 1D) and protein (Figure 1E) readouts, across pregnancy. There was no change in serum hepcidin (and a markedly attenuated reduction in Hamp mRNA) at E14.5 in Tmprss6 siRNA–treated mice compared to their unmated controls. In this system, Tmprss6 inhibition did not change liver iron concentration (Figure 1F) compared to NTC-treated mice. Tmprss6 siRNA–treated mice still demonstrated lower liver iron in late pregnancy than unmated Tmprss6 siRNA–treated mice, indicating that some iron was still able to be mobilized (Figure 1F). Spleen iron was also increased in Tmprss6 siRNA–treated mice at all time points, reflecting reduced iron mobilization from the spleen after the destruction of senescent red cells (Figure 1G). In addition, Tmprss6 siRNA–treated mice showed evidence of profound systemic iron restriction in late pregnancy, with reductions in serum iron (Figure 1H), hemoglobin concentration (Figure 1I), and MCV (Figure 1J) compared with NTC-treated mice.

Further evaluation of the downstream targets of BMP/SMAD signaling demonstrated that although Id1 expression increased over pregnancy in the NTC-treated arm, no significant change in expression was seen after treatment with Tmprss6 siRNA (Figure 1K). Atoh8 and Smad7 were both increased in Tmprss6 siRNA–treated mice compared with NTC-treated mice; in treated mice, Atoh8 expression increased across pregnancy, whereas Smad7 decreased. Our data suggest that hepatic BMP target genes are regulated out of synchrony with Hamp in pregnancy.

Maternal Tmprss6 inhibition in mouse pregnancy produced adverse fetal impacts. Tmprss6 knockdown (compared to NTC) reduced placental iron concentration (Figure 2A), fetal liver iron content (Figure 2B; measured as previously described⁹), and fetal weight (Figure 2C) in samples collected at E14.5.

Next, we evaluated hepcidin regulation in the context of iron overload or iron deficiency. Female mice were fed with 1 of 4 experimental diets (low iron [2-4 parts per million (ppm) carbonyl iron], low control [35 ppm carbonyl iron], control [200 ppm carbonyl iron], or high iron [1000 ppm carbonyl iron]) for 4 weeks, before commencing mating. Hepcidin levels (measured by qPCR and serum enzyme-linked immunosorbent assay) were significantly reduced at E14.5 compared to dietary-matched unmated controls in all groups except the low iron group, in which hepcidin may have already been maximally suppressed (supplemental Figure 1A-B). Baseline liver iron corresponded to the diet and reduced across pregnancy in all groups except low iron (already iron deficient at baseline; supplemental Figure 1C). MCV was reduced in the low iron diet group at E8.5 and E14.5, reflecting iron deficiency (supplemental Figure 1D). Tmprss6 expression was largely consistent across pregnancy in all groups (supplemental Figure 1E); BMP6 expression reduced across pregnancy regardless of iron diet (supplemental Figure 1F). Id1, Smad7, and Atoh8 expression generally did not correlate with hepcidin suppression, regardless of iron status.

Previous studies using maternal or fetal hepcidin knockouts or exogenous maternal hepcidin administration demonstrated that maternal hepcidin is crucial for embryo and fetal iron endowment in a pregnancy.¹² The mechanism of hepcidin suppression in pregnancy remains uncertain, although it is not mediated via the erythroid-derived hepcidin–suppressing factor erythroferrone¹³ or via the pregnancy-related hormones estrogen and prolactin.¹⁴ Other proposed mechanisms include a placental-derived factor or simply a consequence of increased iron utilization. Our data indicate that the mechanism inhibiting hepcidin suppression likely occurs predominantly upstream of the BMP6-hemojuvelin-BMP receptor interaction (analogous to inhibition of BMP signaling by erythroferrone).¹⁵ 

Although regulation of Tmprss6 in pregnancy is not known, disruption of this system causes relative elevation of hepcidin levels across pregnancy, which drives maternal iron depletion and impairs placental iron transfer and fetal growth.