Matriptase-2 regulates iron homeostasis primarily by setting the basal levels of hepatic hepcidin expression through a nonproteolytic mechanism

Matriptase-2 (MT2), encoded by TMPRSS6, is a membrane-anchored serine protease. It plays a key role in iron homeostasis by suppressing the iron-regulatory hormone, hepcidin. Lack of functional MT2 results in an inappropriately high hepcidin and iron-refractory iron-deficiency anemia. Mt2 cleaves multiple components of the hepcidin-induction pathway in vitro. It is inhibited by the membrane-anchored serine protease inhibitor, Hai-2. Earlier in vivo studies show that Mt2 can suppress hepcidin expression independently of its proteolytic activity. In this study, our data indicate that hepatic Mt2 was a limiting factor in suppressing hepcidin. Studies in Tmprss6−/− mice revealed that increases in dietary iron to ∼0.5% were sufficient to overcome the high hepcidin barrier and to correct iron-deficiency anemia. Interestingly, the increased iron in Tmprss6−/− mice was able to further upregulate hepcidin expression to a similar magnitude as in wild-type mice. These results suggest that a lack of Mt2 does not impact the iron induction of hepcidin. Additional studies of wild-type Mt2 and the proteolytic-dead form, fMt2S762A, indicated that the function of Mt2 is to lower the basal levels of hepcidin expression in a manner that primarily relies on its nonproteolytic role. This idea is supported by the studies in mice with the hepatocyte-specific ablation of Hai-2, which showed a marginal impact on iron homeostasis and no significant effects on iron regulation of hepcidin. Together, these observations suggest that the function of Mt2 is to set the basal levels of hepcidin expression and that this process is primarily accomplished through a nonproteolytic mechanism.

Matriptase-2 (MT2) is a type-II transmembrane serine protease.It is composed of a short N-terminal cytoplasmic domain, a transmembrane domain, and an extracellular domain of 735 amino acids, which contains a membraneproximal stem region, a predicted activation domain, and a C-terminal catalytic domain (Fig. 1A) (1).MT2 is encoded by the TMPRSS6 gene in humans and the Tmprss6 gene in mice and is predominantly expressed in the liver (1).It plays a key role in iron homeostasis by suppressing hepatic hepcidin gene transcription (1)(2)(3)(4).Hepcidin is an iron-regulatory hormone that is secreted mainly by hepatocytes (3,4).Hepcidin inhibits iron efflux from duodenal enterocytes that absorb dietary iron, from iron-recycling macrophages in the spleen and in the liver, and from iron-storing hepatocytes into the circulation by blocking the plasma-membrane iron exporter, ferroportin (FPN) (Fig. 1B) (5,6).Lack of hepcidin causes juvenile hemochromatosis, a particularly severe iron overload disorder (7,8).In humans, mutations in the TMPRSS6 gene result in an inappropriately high hepcidin expression and cause ironrefractory iron-deficiency anemia (IRIDA) (3,4,9).Similar phenotypes and alopecia indicating low iron levels in the body are reported in mouse models with a global Tmprss6 knockout (Tmprss6 −/− ), a truncated Tmprss6 that lacks the coding sequence for the catalytic domain (mask), or an Ile286Phe substitution in the CUB domain (Fig. 1A) (10)(11)(12)(13)(14).Our earlier studies suggest that this Ile286Phe substitution in Mt2 reduces the interaction with its binding partners (15).
MT2 is synthesized as a zymogen and undergoes activation by autocleavage on the cell surface (1,36,37).In vitro studies indicate that the activated MT2 cleaves multiple components of the hepcidin induction pathway, including HJV, Alk3, ActRIIA, Hfe, and Tfr2, and reduces their levels on the cell surface (Fig. 1C) (13,38).Co-expression of the hepatocyte growth factor activator inhibitor-2 (Hai-2) inhibits the Mt2 proteolytic activity (Fig. 1D) (39,40).Hai-2, encoded by the Spint2 gene, is a broad membrane-associated serine protease inhibitor (39,41).Initial in vivo studies suggest that MT2 suppresses hepcidin expression by inhibiting the function of HJV.The combined disruption of both Hjv and Tmprss6 genes in mice displays a phenotype that is indistinguishable from Hjv-null (Hjv −/− ) mice with a marked suppression of hepcidin expression and severe iron overload (12).Combined ablation of Tmprss6 and Hfe or Tfr2 genes shows a phenotype similar to the Tmprss6 −/− mice with an inappropriately high hepcidin expression and iron deficiency anemia (42,43).Further studies imply that Mt2 can suppress hepcidin expression independently of Hjv because increased expression of exogenous Mt2 in the liver of Hjv −/− mice is still able to significantly reduce hepcidin (13).MT2 can bind to HJV through its stem region (38).In mask mice that lack detectable Mt2 protein, the increased hepcidin expression is associated with a decrease, rather than an increase, in hepatic Hjv protein (44,45).These findings support the idea that Mt2 suppression of hepcidin is not mediated through its proteolytic activity to cleave and to inactivate its substrates.
Mechanistic studies reveal that Mt2 suppression of hepcidin expression requires not only its catalytic domain but also the stem region (Fig. 1A) and that it has to be associated with the plasma membrane (40).This is consistent with the observations that the IRIDA-causing mutations in MT2 are found throughout the entire extracellular domain (14).Interestingly, our recent studies show that Mt2 is able to suppress hepatic hepcidin expression independently of its proteolytic activity and suggest that this process is likely accomplished through substrate association (15).Hepatic hepcidin expression is positively regulated by the levels of bodily iron load through the BMP-signaling pathway.The liver possesses an elegant but incompletely defined machinery to modulate the transcription of hepcidin gene in response to the changes in bodily iron status (3,(46)(47)(48).Hepatic Tfr2 is stabilized by iron-saturated holo-Tf (49,50), and the transcription of Bmp6 gene in the liver sinusoidal endothelial cells is elevated by increased liver iron storage (16,51).Tfr2 and Bmp6 are, therefore, thought to be the sensors for increased iron in circulation and in storage, respectively, to enhance hepcidin expression (Fig. 1E).MT2 is the major negative regulator of hepcidin expression in the liver (1-4).However, the involvement of Mt2 in iron regulation of hepcidin and the extent to which the proteolytic activity is needed for the function of Mt2 in iron homeostasis is not clear.
In this study, we investigated the roles of Mt2 in iron regulation of hepcidin expression by using Tmprss6 −/− mice and hepatocyte-specific Spint2 knockout (Spint2 fl/fl ;Alb-Cre + ) mice.Results indicated that neither the proteolytic nor nonproteolytic function of Mt2 is involved in the iron regulation of hepcidin.Our data support the idea that Mt2 is a limiting factor in the suppression of hepcidin expression and that the major function of Mt2 is to set the basal levels of hepcidin expression.

Mt2 is a limiting factor in the suppression of hepatic hepcidin expression
Earlier studies show that the heterozygous Tmprss6 +/− mice retain the ability to reduce hepcidin expression similar to the wild-type Tmprss6 Matriptase-2 sets the basal level of hepcidin expression iron deficiency when iron demands are high (12,52).To determine the lowest levels of Mt2 that is needed to suppress hepcidin expression, we expressed FLAG/MYC-tagged Mt2 (fMt2; Fig. 2A) in the liver of Tmprss6 −/− mice at 22% Tmprss6 mRNA of wild-type mice by intraperitoneal administration of AAV8 viral vector (Fig. 2, B and C).This AAV8 vector specifically expresses the gene of interest in hepatocytes, and the transduced gene is evenly expressed in hepatocytes throughout the liver (53), similar to the homogenous distribution of native Tmprss6 mRNA in hepatocytes (21).Our earlier studies indicate that the addition of a C-terminal FLAG/MYC epitope to Mt2 did not affect its ability to suppress hepcidin expression (13,15).In contrast to a full correction of the high hepcidin expression and iron deficient status by expressing a comparable Tmprss6 mRNA level as in wild-type mice (13,15), expression of 22% Tmprss6 mRNA in Tmprss6 −/− mice was only able to moderately elevate serum iron concentration, liver nonheme iron, hemoglobin (Hb), hematocrit (HCT), mean cell volume (MCV), and mean corpuscular hemoglobin (MCH) with no evident change of red blood cell (RBC) counts (Fig. 2, C-J).There was no significant reduction of hepatic hepcidin and Id1 mRNA levels and no evident improvement in alopecia (Fig. 2, K-M).The levels of liver nonheme iron are widely used as an indicator of iron storage in the body.Id1 (inhibitor of differentiation 1) is a direct downstream target of BMP signaling (Fig. 1C), and the levels of Id1 mRNA are widely used as a sensitive indicator for the status of BMP signaling.There is no known function of Id1 in iron homeostasis.In agreement with our earlier studies (54), the administration of the AAV8 viral vector did not increase Il-6 mRNA (Fig. 2N), suggesting that no inflammation was induced.Consistent with the earlier observations (12, 52), Tmprss6 +/− mice at 10-week-old displayed a moderately lower level of liver nonheme iron and HCT (Fig. 2, E and G).A trend of lower MCV when compared with the corresponding wild-type counterparts (Fig. 2H) was also noted.These results suggest that at least 50% of Tmprss6 mRNA seen in wild-type mice is required to efficiently suppress hepcidin expression.In conjunction with our earlier studies showing that increased Mt2 is able to further suppress hepcidin in wild-type and Tmprss6 −/− mice (13,15), these observations indicate that Mt2 is a limiting factor in maintaining iron homeostasis.

Lack of Mt2 does not significantly impact iron induction of hepatic hepcidin expression
To seek insights into the roles of Mt2 in the iron regulation of hepcidin, we first tested the extent to which the Tmprss6 −/− mice can elevate hepatic hepcidin expression in response to increased bodily iron load.Previous studies indicate that severe iron deficiency in mask mice can be corrected by feeding an iron-enriched diet with 2% carbonyl iron (11,45).To avoid gastrointestinal bleeding and other side effects of high dietary iron (55), we fed the Tmprss6 +/+ , Tmprss6 +/− , and Tmprss6 −/− mice a high iron diet containing 0.5% carbonyl iron (HID).The parallel control groups were fed an iron control diet (ICD) containing 48 ppm iron.Since feeding a 0.83% carbonyl iron diet for 10 days is insufficient to fully correct the low Hb and low HCT status in mask mice ( 56), all animals were analyzed at 3 weeks after feeding the HID or ICD (Fig. 3A).Matriptase-2 sets the basal level of hepcidin expression As expected, feeding a HID to wild-type mice (Tmprss6 +/+ /HID) markedly increased the bodily iron load as manifested by elevation of serum iron concentrations and liver nonheme iron levels (Fig. 3, B and C) and mild increases in Hb, HCT, and MCH (Fig. 3, D-F) when compared with the corresponding ICD group.Interestingly, feeding a HID to Tmprss6 −/− mice (Tmprss6 −/− /HID) also led to a rapid iron load to a comparable level of Tmprss6 +/+ /HID group within 3 weeks (Fig. 3, B and C).As a result, the anemia was fully corrected as indicated by marked increases in Hb, HCT, MCH, and MCV (Fig. 3, D-G).The RBC counts returned to the level of Tmprss6 +/+ /ICD group (Fig. 3H), and the alopecia disappeared (Fig. 3I).In comparison, mice in Tmprss6 −/− /ICD group remained severely iron deficient and anemic (Fig. 3,  B-I).In wild-type mice, increased iron load did not alter hepatic Tmprss6 mRNA levels (Fig. 3J).These observations indicate that the iron deficiency anemia in Tmprss6 −/− mice can be corrected by a diet containing 0.5% carbonyl iron.
Consistent with the earlier studies (3,29,57), increased bodily iron load significantly induced hepatic hepcidin expression in wild-type Tmprss6 +/+ mice by facilitating the Bmp signaling as displayed by a parallel elevation in Id1 mRNA (Fig. 3, K and L).Intriguingly, increased bodily iron load in Tmprss6 −/− mice was able to further upregulate both hepcidin and Id1 mRNA levels to a similar magnitude as those seen in wild-type Tmprss6 +/+ and heterozygous Tmprss6 +/− mice (Fig. 3, K and L).The relatively lower fold increases in Tmprss6 −/− /HID group were likely due to the higher basal levels in the control Tmprss6 −/− /ICD group (Fig. 3, K and L).The lack of significant change in IL-6 mRNA levels (Fig. 3M) ruled out the possibility of inflammation.Thus, these observations indicate that lack of Mt2 does not significantly impact the iron-mediated upregulation of hepatic hepcidin expression.Because of the severe iron deficiency in Tmprss6 −/− mice fed an ICD (Fig. 3, D-G), no iron-deprivation studies were conducted in these animals.
We next examined whether the iron-induced hepcidin expression in Tmprss6 −/− mice resulted from elevated Bmp6 mRNA and Tfr2 protein in the liver.Both Bmp6 and Tfr2 increase hepcidin expression (3,29,57).The transcription of Bmp6 gene is upregulated by increased iron storage (16,51), and hepatic Tfr2 is stabilized by increased iron-saturated holo-Tf (49,50).As shown in Figure 3, N and O, similar extents of elevation in Bmp6 mRNA and Tfr2 protein in the liver were detected between the Tmprss6 +/+ /HID and Tmprss6 −/− /HID groups when compared with their corresponding ICD controls.Additionally, a comparable degree of increase was also detected in Tmprss6 +/− mice.These results indicate that Tmprss6 −/− mice retain intact iron-sensing machinery in the liver and that Mt2 is not involved in the iron regulation of Tfr2 stability and Bmp6 gene transcription.Rather the abovementioned data suggest that the function of Mt2 is to set the basal level of hepcidin expression in the liver.

The proteolytic activity of Mt2 is not required in iron induction of hepatic hepcidin expression
Our earlier studies indicate that Mt2 can suppress hepcidin expression through a nonproteolytic mechanism in mice fed a 240-ppm iron diet (15).To determine the roles of Mt2 proteolytic activity in iron homeostasis in response to dietary changes in iron, we compared the abilities of wild-type fMt2 and the proteolytic dead form, fMt2 S762A , to suppress hepcidin expression under the control iron (48-ppm iron) and high iron (0.5% carbonyl iron) conditions.fMt2 and fMt2 S762A with a C-terminal FLAG/MYC epitope (Fig. 4A) were transduced into the liver of Tmprss6 −/− mice by AAV8 viral vectors as described in our earlier studies (15).AAV8-fMt2 and -fMt2 S762A viral vectors were intraperitoneally administered into the Tmprss6 −/− mice of both genders at 5 weeks of age.The Tmprss6 +/+ and Tmprss6 −/− controls were injected with a sterile PBS vehicle.Our earlier studies demonstrated that the intraperitoneal administration of empty AAV8 vector had no evident impacts on hepatic hepcidin expression and iron homeostasis in mice (58,59).At 2 weeks after viral administration to allow the expression of transduced Mt2 and Mt2 S762A , animals were randomly divided into two categories to feed either an ICD (48-ppm iron) or a HID (0.5% carbonyl iron) for 3 more weeks (Fig. 4B).All mice were euthanized at 10 weeks old for analysis.
As shown in Figure 4C, the mRNA levels of transduced fMt2 and fMt2 S762A were modestly higher than those of the wild-type control group.Both fMt2 and fMt2 S762A proteins in the liver membrane preparations were readily detected by an anti-FLAG antibody (Fig. 4D).Consistent with our earlier studies (15), the expressed fMt2 and fMt2 S762A in the liver of Tmprss6 −/− mice exhibited comparable abilities to reduce the high hepcidin and Id1 mRNA expression when fed an ICD.The hepcidin and Id1 mRNA were decreased to a similar level of wild-type Tmprss6 +/ + /ICD group (Fig. 4, E and F).Consistently, both fMt2 and fMt2 S762A displayed similar extents of increases in serum iron concentrations, liver nonheme iron levels, and blood parameters including Hb, MCV, and MCH (Fig. 4, G-K).The alopecia also disappeared as a result of the increased bodily iron load (Fig. 4L).These observations indicate that the proteolytic activity of Mt2 is not required for the suppression of hepcidin expression in mice with an adequate iron supply.
In the HID groups, the fMt2-transduced Tmprss6 −/− mice displayed comparable levels of increases in serum iron and liver nonheme iron levels as seen in the wild-type and Tmprss6 −/− controls (Fig. 4, G and H).Increased bodily iron load resulted in similar extents of hepcidin and Id1 mRNA increases between fMt2/HID and wild-type Tmprss6 +/+ /HID groups (Fig. 4, E and F).These results indicate that the transduced fMt2 acts similarly to endogenously expressed Mt2 in response to a high iron load.Intriguingly, increased bodily iron in the fMt2 S762A -transduced Tmprss6 −/− mice fed an HID (fMt2 S762A /HID) also led to analogous increases in hepcidin and Id1 mRNA as seen in fMt2/HID and wild-type Tmprss6 +/ + /HID groups (Fig. 4, E and F).These observations indicate that the proteolytic activity of Mt2 is not involved in the iron induction of hepcidin expression.In comparison, the Tmprss6 −/− /HID controls displayed a significantly higher level of hepcidin mRNA and a trend of increase in Id1 mRNA (Fig. 4, E and F).The lack of changes of IL-6 mRNA (Fig. S1) ruled out the possibility of inflammation as a cause of elevated Matriptase-2 sets the basal level of hepcidin expression hepcidin expression.These data suggest that under the high iron conditions, Mt2 also acts as a suppressor to downshift the hepcidin expression.Additional analysis revealed that expression of neither fMt2 nor fMt2 S762A significantly affected the iron-mediated increases in hepatic Tfr2 protein and Bmp6 mRNA (Fig. 4, D and M).These results imply that the Mt2 suppression of hepcidin is not accomplished by altering the hepatic Tfr2 levels or the Bmp6 gene transcription.Taken together, the data suggest that the proteolytic activity of Mt2 is not involved in the iron induction of hepatic hepcidin expression and that the primary function of Mt2 is to downshift the basal level of hepcidin expression.Hepatocyte-specific ablation of the serine protease inhibitor, Hai-2, displays a marginal effect on hepcidin expression in mice HAI-1 (encoded by the Spint1 gene) and HAI-2 (encoded by the Spint2 gene) are two closely related membrane-associated Kunitz-type serine protease inhibitors with a broad inhibitory spectrum (39,(60)(61)(62)(63)(64)(65)(66)(67).In transfected cells, Hai-2 blocks Mt2 cleavage of multiple key components in the hepcidin induction pathway (15,39,40,60).We hypothesized that the lack of difference between fMt2 and the protease-dead fMt2 S762A in suppressing hepcidin expression could result from a constant inhibition of Mt2 proteolytic activity by the endogenously expressed Hai-2 in the liver (Fig. 5A).To test this hypothesis, we examined the expression profiles of Hai-1 and Hai-2 in the liver.qRT-PCR analysis revealed that the liver predominantly expresses Spint-2 mRNA (Fig. 5B).In isolated cells from the mouse liver, Hai-2 mRNA was detected primarily in the hepatocyte population (Fig. 5C), which is similar to the expression profile of Tmprss6 mRNA (21).Thus, both Hai-2 and Mt2 are co-expressed in hepatocytes.
To determine the roles of hepatic Hai-2 in hepcidin expression, we generated a liver-specific Spint2 knockout (Spint2 fl/fl ;Alb-Cre + ) mouse model by crossing Spint2 fl/fl mice (68) with mice that express a Cre recombinase transgene driven by a hepatocyte-specific albumin (Alb) promoter.Global ablation of Spint2 gene in mice results in an embryonic lethality (69).qRT-PCR analysis revealed 99% depletion of Spint2 mRNA in the isolated hepatocytes from Spint2 fl/fl ;Alb-Cre + mice (Fig. 6A), indicating a hepatocyte-specific deletion.
Distinct from global Spint2 mutant mice, Spint2 fl/fl ;Alb-Cre + mice were born in the predicted Mendelian ratio.There was no evident difference in appearance when compared to wild-type Spint2 fl/fl ;Alb-Cre -littermates.Both Spint2 fl/fl ;Alb-Cre + and Spint2 fl/fl ;Alb-Cre -littermates had similar gender-dependent body weights at 5-and 8 weeks old (Fig. S2A).Both male and female Spint2 fl/fl ;Alb-Cre + mice were fertile.Depletion of Spint2 mRNA did not cause a compensatory increase of Spint1 mRNA (Fig. 6A) and did not alter the mRNA levels of other key iron regulatory genes, including Tmprss6, Hjv, Hfe, Tfr2, or Neo1, in the isolated hepatocytes (Fig. S2, B-F).
The roles of endogenous Hai-2 in iron homeostasis were examined by comparing 5-and 8-week-old Spint2 fl/fl ;Alb-Cre + mice with the gender and age-matched wild-type Spint2 fl/ fl ;Alb-Cre -littermates.All mice were fed a regular rodent diet containing 240-ppm iron after weaning.Consistent with the results in isolated hepatocytes (Fig. 6A), there were negligible amounts of Spint2 mRNA in the liver of Spint2 fl/fl ;Alb-Cre + mice (Fig. 6B).Compared with the corresponding wild-type Spint2 fl/fl ;Alb-Cre -littermates, we detected a trend increase in serum iron and liver nonheme iron levels only in 5-week-old male Spint2 fl/fl ;Alb-Cre + mice (Fig. 6, C and D).qRT-PCR analysis revealed a minor reduction in hepatic hepcidin and Id1 mRNA levels in this group of Spint2 fl/fl ;Alb-Cre + mice (Fig. 6, E and F).These results suggest that the moderate elevation of iron load in 5-week-old male Spint2 fl/fl ;Alb-Cre + mice likely results from the reduced Bmp signaling and hepcidin expression.No evident difference was observed in other groups (Fig. 6, C-F Matriptase-2 sets the basal level of hepcidin expression not significantly alter the blood parameters, including RBC counts, Hb, HCT, MCV, and MCH in Spint2 fl/fl ;Alb-Cre + mice (Table S1).Together, these results suggest that hepatic Hai-2, to a very limited extent, is involved in preventing the excessive suppression of hepatic hepcidin expression under steady state conditions.

Lack of hepatic Hai-2 does not alter iron regulation of hepcidin expression
We next examined whether hepatic Hai-2 is required for iron regulation of hepcidin expression.Both Spint2 fl/fl ;Alb-Cre + and the wild-type Spint2 fl/fl ;Alb-Cre -littermates of both genders were fed an iron-deficient diet (IDD; 2-6 ppm iron), an ICD (48 ppm iron), or a HID (0.5% carbonyl iron) for 3 weeks before analysis (Fig. 7A).As shown in Figure 7, B-D, mice fed a HID displayed marked increases in the liver and spleen nonheme iron levels as well as serum iron concentrations in both Spint2 fl/fl ;Alb-Cre + and Spint2 fl/fl ;Alb-Cre - mice to the similar extents.Consistently, comparable decreases in the liver and spleen nonheme iron were also detected between the IDD groups (Fig. 7, B and C).Although mice fed an IDD displayed no significant reduction of serum Matriptase-2 sets the basal level of hepcidin expression iron levels relative to their corresponding ICD controls (Fig. 7D), all these mice had a trend toward iron deficiency anemia as manifested by mild decreases in Hb, HCT, MCV, or MCH (Table S2).These observations indicate that depletion of hepatic Hai-2 does not significantly impact the bodily iron loading in mice.
As expected, hepatic hepcidin, Id1, and Bmp6 mRNA levels in wild-type Spint2 fl/fl ;Alb-Cre -mice of both genders were decreased by bodily iron load deprivation and elevated by increased body iron load (Fig. 7, E-G).These observations are consistent with the idea that hepcidin is induced via the Bmp signaling and that Bmp6 expression in the liver is enhanced by increased liver iron load.In comparison, increased iron load in Spint2 fl/fl ;Alb-Cre + /HID groups displayed similar extents of enhancement in hepatic hepcidin, Id1, and Bmp6 mRNA as seen in the corresponding Spint2 fl/fl ;Alb-Cre -/HID controls (Fig. 7, E-G).Consistently, similar extents of decrease in hepatic hepcidin, Id1, and Bmp6 mRNA were observed by iron depletion between the Hai-2-ablated Spint2 fl/fl ;Alb-Cre + mice (Spint2 fl/fl ;Alb-Cre + /IDD) and wild-type Spint2 fl/fl ;Alb-Cre - mice (Spint2 fl/fl ;Alb-Cre -/IDD) of both genders (Fig. 7, E-G).The lack of change of the Spint2 or Spint1 mRNA levels by different bodily iron load (Fig. 7, H and I) suggests that their expressions are not regulated by the Bmp signaling.Additionally, we also examined the induction of hepcidin expression by acute iron increase in Spint2 fl/fl ;Alb-Cre + mice.No significant difference was detected when compared with Spint2 fl/fl ;Alb-Cre -mice (data not shown).Together, these observations indicate that lack of hepatic Hai-2 does not impact the iron regulation of hepcidin expression, and suggest that hepatic Hai-2 does not play an important role in Mt2 suppression of hepcidin.

Discussion
In this study, we explored the underlying mechanism by which hepatic Mt2 suppresses hepcidin expression.Our results indicate that hepatic Mt2 was a limiting factor in iron homeostasis, and that lack of Mt2 did not impact the magnitude of hepcidin induction by iron.Further studies suggest that the major function of Mt2 was to set the basal levels of hepatic hepcidin expression and that this process was accomplished primarily by its nonproteolytic function.Consistently, ablation of hepatocyte Hai-2 in mice only displayed a marginal effect on iron homeostasis.In conjunction with the earlier studies (15), our data support the idea that Mt2 regulates iron homeostasis primarily by setting the basal levels of hepatic hepcidin via a nonproteolytic mechanism.Earlier studies show that the heterozygous Tmprss6 +/− mice develop normally and are phenotypically indistinguishable from wild-type littermates (52).They have similar abilities to reduce hepcidin expression as the wild-type counterparts but are prone to iron deficiency when iron demands are high or when dietary iron is restricted (12,52).Here, we explored the minimal levels of Mt2 that are needed to suppress hepcidin expression.Results indicate that the expression of 22% Tmprss6 mRNA of wild-type mice in Tmprss6 −/− mice was unable to correct the high hepcidin expression and iron deficiency status.Interestingly, the siRNA and antisense oligonucleotide knockdown studies in wild-type mice reveal a significant increase in hepcidin expression when hepatic Tmprss6 mRNA levels were decreased by 60% (70,71).This increase in hepcidin expression was also augmented by further reduction of Tmprss6 mRNA (70,71).These observations suggest that an efficient suppression of hepatic hepcidin expression requires at least 50% of Tmprss6 mRNA seen in wild-type mice.Conversely, our earlier studies show that increased Mt2 in wild-type mice is able to further suppress hepcidin mRNA levels and to increase serum iron concentrations (13).Since MT2 expression is not transcriptionally regulated by iron (21,72), these findings indicate that hepcidin expression is negatively correlated with hepatic Tmprss6 mRNA levels, and support the idea that Mt2 is limiting factor in the suppression of hepcidin.
The liver possesses an elegant but incompletely defined machinery that can sense the levels of bodily iron load and positively regulate the transcription of the hepcidin gene (3,(46)(47)(48)57).This iron sensing machinery appears to be altered by the TMPRSS6 gene mutations in humans (3,4,9), global ablation of Tmprss6 gene (Tmprss6 −/− ) or a deletion of the Mt2 catalytic domain (mask) in mice (10)(11)(12)(13), because they all result in an inappropriately high hepcidin and iron-deficiency anemia.Here we investigated the roles of Mt2 in iron regulation of hepcidin expression.First, our data validated the earlier observations that iron-deficiency anemia in mask mice could be corrected by feeding a rodent diet with 0.83% or 2% carbonyl iron (11,45,56).We found that feeding a 0.5% carbonyl iron diet in Tmprss6 −/− mice was sufficient to overcome the high hepcidin barrier and to load iron into the liver to a similar extent as seen in Tmprss6 +/+ and Tmprss6 +/− mice within 3 weeks.The lower "high iron diet" minimized the deleterious side effects of damage to the intestines and kidneys.Interestingly, when the response of hepatic hepcidin expression to increased bodily iron load was compared, our data revealed that the Tmprss6 −/− mice retained the comparable responses of wild-type Tmprss6 +/+ and heterozygous Tmprss6 +/− mice to elevate hepatic hepcidin expression by enhancing the Bmp signaling.The absolute levels of hepcidin mRNA in the Tmprss6 −/− mice remained higher than those seen in Tmprss6 +/+ and Tmprss6 +/− mice, similar to the ICD groups.Further studies indicated that lack of Mt2 did not impact the increases in hepatic Tfr2 protein and Bmp6 mRNA by increased iron load in Tmprss6 −/− mice.Hepatic Tfr2 is stabilized by increased iron-saturated holo-Tf in the circulation (49,50), and the transcription of Bmp6 gene in the liver sinusoidal endothelial cells is upregulated by an increased intracellular iron (16,51,73).These results suggest that Tmprss6 −/− mice have intact iron-sensing and response machinery in the liver.Together, these observations support the idea that the function of Mt2 is to set the basal levels of hepcidin expression in the liver.
MT2 is a membrane-anchored serine protease.It cleaves multiple components of the hepcidin induction pathway including Tfr2 and Hjv, and reduces their levels on the cell surface in transfected cells (13,38).Our earlier in vivo studies indicate that Mt2 can suppress hepcidin expression independently of its proteolytic activity in mice under the steady state conditions (15).Here we investigated the role of proteolytic activity of Mt2 in iron regulation of hepcidin expression by taking advantage of fMt2 and the protease-dead fMt2 S762A .Our data indicate that the protease-dead fMt2 S762A behaved similarly to wild-type fMt2 in the suppression of hepatic hepcidin expression and the increase of bodily iron load under both adequate and high iron conditions.Consistently, expression of neither wild-type Mt2 nor fMt2 S762A impacted the iron-induced increases in hepatic Tfr2 protein.These observations suggested that the proteolytic activity of Mt2 is not involved in the iron induction of hepcidin expression.In support of this idea, the decreased, rather than increased, levels of Hjv protein are found in the liver of mask mice that lack the Mt2 catalytic domain (44).Our earlier studies also suggest that the levels of hepatic Hjv is not a limiting factor in the induction of hepcidin (54,58).Thus, Mt2 suppression of hepcidin expression was unlikely mediated through its proteolytic activity to cleave and to inactivate Hjv and other hepcidininducing components.Rather, these findings favor our earlier hypothesis that Mt2 suppress hepcidin expression through the interaction with its binding partners (15).
Additional studies in mice with hepatocyte-specific ablation of Hai-2 strengthen the idea that Mt2 suppresses hepcidin expression independent of its proteolytic activity.HAI-2 and HAI-1 are two closely related membrane-associated Kunitztype serine protease inhibitors with a broad inhibitory spectrum, including MT2, matriptase, hepsin (also known as Tmprss1), and prostasin (39, 60-67, 74, 75).Matriptase is a close family member of MT2 (1).Other studies indicate that the function of matriptase in vivo is controlled by HAI-2.Lack of Hai-2 suppression of the matriptase-mediated cell surface proteolysis causes neural developmental defects (76), prostate cancer invasion (77)(78)(79)(80), and severe intestinal epithelial damage Matriptase-2 sets the basal level of hepcidin expression (68,81).We found that the hepatocytes predominantly expressed Hai-2, but not Hai-1, which is similar to that of Mt2 (21).Our earlier studies showed that marked increases of Hai-2 in the liver of wild-type mice did not significantly affect hepcidin expression and Bmp signaling (15).This observation could be interpreted to mean that the endogenously expressed Hai-2 is sufficient to inhibit the proteolytic activity of Mt2.In this study, we detected a mild decrease in hepatic hepcidin mRNA levels and moderate increases in serum iron and liver nonheme iron only in 5-week-old male mice with hepatocytespecific Hai-2 ablation.This decrease in hepcidin expression could result from the lack of Hai-2 to inhibit the Mt2 proteolytic activity.Alternatively, this decrease in hepcidin expression could also be caused by the lack of Hai-2 to interfere with the function of Mt2.In vitro studies show that HAI-2 binds to and forms a complex with MT2 (39).Additionally, it is also possible that this decreased hepcidin expression is caused by altered functions of hepatic hepsin and prostasin.Both hepsin and prostasin are expressed in the liver, and they are essential for various liver metabolism and energy homeostasis (82)(83)(84).Further studies indicate that lack of hepatic Hai-2 did not significantly impact the iron regulation of hepcidin in mice.These data strengthened the idea that the proteolytic activity of Mt2 is not involved in iron regulation of hepcidin.Together, these observations suggest that hepatic Hai-2 might play a minor role in iron homeostasis by increasing hepcidin expression in young mice.We speculate that this process is likely accomplished by Hai-2-mediated inhibition of the proteolytic activity or non-proteolytic function of Mt2.
In summary, our data support the idea that Mt2 regulates iron homeostasis by setting the basal levels of hepatic hepcidin gene expression via a nonproteolytic mechanism.Based on the findings in this and previous studies showing that the increased hepcidin expression in mask mice is associated with a decrease, rather than increase, in hepatic Hjv protein (44,45) and that at least three disease-causing mutations in human MT2 have unaltered proteolytic activity (85)(86)(87), we propose a model in which Mt2 inhibits hepcidin expression primarily by binding to and interfering with the complex in hepatocytes that facilitates the Bmp signaling, which leads to a downshift in the Bmp signaling and hepcidin expression to a constant extent (Fig. 8).In this scenario, the iron induction of hepcidin is accomplished by hepatic Tfr2 that is stabilized by ironsaturated holo-Tf in the circulation (49,50) and Bmp6 in the liver sinusoidal endothelial cells whose transcription is upregulated by increased iron store (16,51,73).Mt2 is not involved in the iron regulation of hepcidin expression.The proteolytic activity of Mt2 may play a modest role in iron homeostasis at younger age when iron demands are high.
To study the underlying mechanism by which Mt2 suppresses hepcidin, five-week-old Tmprss6 −/− mice of both genders were intraperitoneally injected with AAV8-Mt2 or AAV8-Mt2 S762A viral vectors at the doses indicated in the figure legends.Since our earlier studies demonstrated that an intraperitoneal administration of an empty AAV8 vector had no evident impacts on hepatic hepcidin expression and iron homeostasis in mice (58,59), the control groups were injected with sterile PBS vehicle.At 2 weeks after the viral administrations, animals were randomly divided into two groups and started to be fed an ICD (48 ppm iron; TD.09488) or a HID (0.5% carbonyl iron; TD.140464) for an additional 3 weeks.Mice were euthanized at around 10 weeks old for analysis (Figs. 2B and 4B).Age, gender, and background-matched wild-type littermates were included as additional controls.
To study the roles of Mt2 in iron regulation of hepcidin expression, 7-week-old Tmprss6 +/+ , Tmprss6 +/− , and Tmprss6 −/− mice of both genders were randomly divided into two groups and fed either an iron control diet (ICD; 48 ppm iron) or a high iron diet (HID; 0.5% carbonyl 3 weeks.Mice were euthanized at around 10 weeks old for analysis (Fig. 3A).
The Spint2-floxed (Spint2 fl/fl ) mice on a C57BL/6J background were generated as previously described (68).Homozygous hepatocyte-specific conditional Spint2 knockout (Spint2 fl/fl ;Alb-Cre + ) mice and littermate Cre -controls were generated by crossing Spint2 flflt mice with mice expressing a Cre recombinase transgene driven by a hepatocyte-specific albumin (Alb) promoter on a C57BL/6J background (Jackson Laboratory).Genotyping was performed by using mouse tail snipping and PCR.The mice of both genders were euthanized at about 5 and 8 weeks of age for analysis.To study the roles of Hai-2 in iron regulation of hepcidin expression, six-week-old Spint2 fl/fl ;Alb-Cre + and Spint2 fl/fl ;Alb-Cre -mice of male or female were randomly divided into three groups and fed an irondeficient diet (IDD; 2-6 ppm iron; TD.110669), an iron control diet (ICD; 48 ppm iron) or a HID (0.5% carbonyl iron) for 3 weeks.Mice were euthanized at around 9 weeks old for analysis (Fig. 7A).
To study the induction of hepcidin expression by acute iron increase, six-week-old Spint2 fl/fl ;Alb-Cre + and Spint2 fl/fl ;Alb-Cre -mice were first fed an iron-deficient diet (IDD; 2-6 ppm iron; TD.110669) for 3 week to deplete bodily iron.Animals were then randomly divided into two categories and intraperitoneally injected with either Venofer (American Regant, Inc #NOC 0517-2340-10) at 2 mg Fe per 100 g body weight or 30% sucrose/saline vehicle.At 4 h post-administration, mice were euthanized for analysis.

Isolation of hepatocytes
Hepatocytes were isolated from 8-week-old Spint2 fl/fl ;Alb-Cre + and Spint2 fl/fl ;Alb-Cre -mice as previously described (51).Briefly, the livers were perfused using collagenase Type 2 (Worthington Biochemical Corporation) in Earle's Balanced Salt Solution (Sigma-Aldrich).Hepatocytes were pelleted by centrifugation (500 rpm) (Beckman Centrifuge, Allegra 6R) for 5 min at 4 C. Cell pellets were immediately lysed in the RA1 buffer of NucleoSpin RNAII kit for RNA preparation and qRT-PCR analysis of the genes of interest.

Blood parameters and serum iron assay
Blood parameters were analyzed by using Hemavet 950 (Drew Scientific).Serum iron concentration was detected by using a Pointe Iron/TIBC Reagent Set (Pointe Scientific).

Tissue nonheme iron assays
Tissue nonheme iron levels were determined as previously described (88) with the following modifications.Briefly, 50 to 150 mg wet tissues were digested in 250 to 750 μl of acid buffer at 65 C for 72 h.The supernatant was collected by centrifugation at 10,000g for 5 min, followed by the addition of chromogen (1.86 mM bathophenanthroline sulfonate, 143 mM thioglycolic acid in water) and OD measurement at 535 nm.Each sample was measured twice in triplicate.Iron concentration is expressed as micrograms of iron per gram of wet tissue.

qRT-PCR
Total RNA from mouse liver tissues and isolated hepatocytes was extracted using a NucleoSpin RNA kit (Macherey-Nagel).cDNA was synthesized using Oligo deoxythymidine primers (Invitrogen) and M-MLV reverse transcriptase (Invitrogen).The cDNA preparations from isolated mouse liver hepatocytes, Kupffer cells, sinusoidal endothelial cells, and hepatic stellate cells (HSC) are same as in our previous studies (89).qRT-PCR analysis was carried out in triplicate on each sample using the Power SYBR Green PCR master mix in a QuantStudio 12K Flex qPCR System (ThermoFisher).All primer sets used in these studies (Tables S3) were validated against the reference primers (β-actin) to ensure approximately equal efficiencies of amplification.The results are expressed as the amount relative to that of β-actin for each sample.

Immunodetection
Liver membrane fractions were prepared as previously described (44).Protein extracts from the liver membrane preparations were separated by using SDS-PAGE under reducing conditions.The transduced exogenous fMt2 and fMt2 S762A in the liver were probed directly by using an HRPcoupled mouse anti-FLAG M2 IgG (Sigma) and chemiluminescence.The endogenous Tfr2 and β-actin were detected by using purified rabbit anti-Tfr2 (90), mouse anti-β-actin (Sigma), and the corresponding secondary antibodies.

Statistical analysis
Two-tailed student t test was used to compare two sets of data.One-way ANOVA and Tukey's post-test were used for multiple comparisons.
Matriptase-2 sets the basal level of hepcidin expression

Figure 2 .
Figure 2. Mt2 is a limiting factor in the suppression of hepatic hepcidin expression.A, diagrams of Mt2 and fMt2 constructs with C-terminal FLAG/MYC tag.m: MYC.f: FLAG.B, experimental design to determine the suppression of hepcidin expression relative to hepatic Tmprss6 mRNA levels.5-week-old homozygous Tmprss6 −/− mice were intraperitoneally injected with AAV8-fMt2 at 2 × 10 11 viral genome particles per mouse or sterile PBS vehicle (−).Wildtype Tmprss6 +/+ and heterozygous Tmprss6 +/− mice at the same age were also injected with vehicle.Standard diet (PicoLab Laboratory Rodent Diet-5L0D; LabDiet) contains 240-ppm iron.The iron control diet (ICD) contains 48 ppm iron.All mice were euthanized at 10 weeks of age for analysis.Each group consists of at least four mice with similar numbers of males and females.C-L, hepatic Tmprss6 mRNA, serum iron, liver nonheme iron, hemoglobin (Hb), hematocrit (HCT), mean cell volume (MCV), mean corpuscular hemoglobin (MCH), RBC counts, hepatic hepcidin mRNA, and hepatic Id1 mRNA levels in wildtype Tmprss6 +/+ mice, heterozygous Tmprss6 +/− mice, as well as homozygous Tmprss6 −/− mice injected with AAV8-fMt2.M, a representative image of mice at the time of euthanasia.N, IL-6 mRNA levels in the liver.All qRT-PCR results are expressed as the amount relative to that of β-actin for each sample.The mean values and the standard deviation (SD) are presented.One-way ANOVA was used to analyze the data relative to wild-type Tmprss6 +/+ mice.Two-tailed student t test was also used to analyze the data between Tmprss6 +/+ and Tmprss6 +/− mice as well as Tmprss6 −/− mice with and without AAV8-fMt2 administration.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 3 .
Figure3.Mt2 regulates iron homeostasis by setting the basal levels of hepatic hepcidin expression.A, experimental design to assess the iron induction of hepatic hepcidin expression.Seven-weeks old wild-type Tmprss6 +/+ , heterozygous Tmprss6 +/− , and homozygous Tmprss6 −/− mice were randomly divided into two subgroups and fed either an iron control diet (ICD; 48 ppm iron) or a high iron diet (HID; 0.5% carbonyl iron).All mice were euthanized at 10 weeks of age for analysis.Each group consists of at least seven mice with similar numbers of male and female.B, serum iron assay.C, liver nonheme iron assay.D-H, blood parameters: Hb, HCT, MCH, MCV, and RBC counts.I, a representative image of mice at the time of euthanasia.Since the data for the ICD (48-ppm Fe) group were generated from the same sets of animal studies as in Figure2, we used the same images for the ICD group as in Figure2Min order to compare with the mice in the parallel HID (5000-ppm Fe) group.J-N, qRT-PCR analysis of hepatic Tmprss6, hepcidin, Id1, IL-6, and Bmp6 mRNA levels.All qRT-PCR results are expressed as the amount relative to that of β-actin for each sample.O, representative images of Western blot analysis for endogenous Tfr2 and β-actin in the liver membrane extracts (250 μg protein) from the mice as described above.Each panel was cropped from the same image.All data are expressed as mean ± SD.One-way ANOVA and Tukey's post-test were used to analyze the data relative to the Tmprss6 +/+ /ICD group.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 4 .
Figure 4. Mt2 suppression of hepcidin expression largely depends on its nonproteolytic function.A, diagrams of wild-type fMt2 and the proteolytic dead fMt2 S762A .B, experimental design to assess the role of Mt2 proteolytic activity in iron induction of hepcidin expression.Five-week-old Tmprss6 −/− mice of both sexes were intraperitoneally injected with AAV8-fMt2 or fMt2 S762A at 8 × 10 11 viral genome-particles per mouse, or the sterile PBS vehicle as control (−).At 2 weeks post the viral administrations, mice in each group were randomly divided into two subgroups and fed either an iron control diet (ICD; 48 ppm iron) or a high iron diet (HID; 0.5% carbonyl iron) for 3 weeks.Animals were euthanized at 10 weeks of age.Age-matched wild-type Tmprss6 +/+ (WT) littermates on the same background were injected with a sterile PBS vehicle as additional controls.Each group consists of at least five animals with similar numbers of male and female.C, qRT-PCR analysis of hepatic Tmprss6 mRNA.D, representative images of Western blot analysis for transduced fMt2 and fMt2 S762A , the endogenous Tfr2 and β-actin in the liver membrane extracts (250 μg protein) from the mice as described above.E and F, qRT-PCR analysis of hepatic hepcidin and Id1 mRNA.G, serum iron assay.H, liver nonheme iron assay.I-K, Blood parameters: Hb, MCV, and MCH.L, a representative image of mice at the time of euthanasia.M, qRT-PCR analysis of hepatic Bmp6 mRNA.All qRT-PCR results are expressed as the amount relative to that of β-actin for each sample.Data are expressed as mean ± SD.One-way ANOVA and Tukey's post-test were used to analyze the data relative to the WT-Tmprss6 +/+ /ICD group.*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 5 .
Figure 5. Spint2, not Spint1, is predominantly expressed in hepatocyte population of the liver.A, the predicted role of hepatic Hai-2 in Mt2 suppression of hepcidin expression and the predicted consequence of hepatic Spint2 gene knockout in mice.B, qRT-PCR analysis of Spint1 (Sp1) and Spint2 (Sp2) mRNA levels in the liver of 8-week-old wild-type mice on a C57BL/6J background.Each group consists of at least five animals.C, qRT-PCR analysis of Spint1 (Sp1) and Spint2 (Sp2) mRNA levels in isolated hepatocytes (HC), Kupffer cells (KC), sinusoidal endothelial cells (SEC), and hepatic stellate cells (HSC) from the liver of wild-type 129S mice (n = 4).All qRT-PCR results are expressed as the amount relative to that of β-actin for each sample.

Figure 6 .
Figure 6.Ablation of hepatic Spint2 mildly increases bodily iron load in male mice.A, qRT-PCR analysis of Spint1 and Spint2 mRNA levels in isolated hepatocytes from 8-week-old Spint2 fl/fl ;Alb-Cre -and Spint2 fl/fl ;Alb-Cre + mice.B, qRT-PCR analysis of Spint2 mRNA levels in the liver of 5 and 8-week-old Spint2 fl/fl ;Alb-Cre -and Spint2 fl/fl ;Alb-Cre + mice of both genders.C and D, serum iron and liver nonheme iron assays in 5 and 8-week-old Spint2 fl/fl ;Alb-Cre -and Spint2 fl/fl ;Alb-Cre + mice of both genders.E and F, qRT-PCR analysis of hepcidin and Id1 mRNA levels in the liver of 5 and 8-week-old Spint2 fl/fl ;Alb-Cre -and Spint2 fl/fl ;Alb-Cre + mice of both genders.All qRT-PCR results are expressed as the amount relative to that of β-actin for each sample.Each group consists of at least four animals.The means ± SD are presented.Two-tailed Student t test was used to analyze the data for each age and gender-matched group.The p values were presented.

Figure 7 .
Figure 7. Ablation of hepatic Spint2 does not affect iron regulation of hepcidin expression.A, experimental design to examine the effects of hepatocyte-specific Spint2 knockout on iron regulation of hepcidin expression.Five-week-old Spint2 fl/fl ;Alb-Cre -and Spint2 fl/fl ;Alb-Cre + mice of both genders were fed an iron deficient diet (IDD; 2-6 ppm iron), an iron control died (ICD; 48 ppm iron), or a high iron diet (HID; 0.5% carbonyl iron; TD.140464) for 4 weeks before euthanasia for analysis.B-D, liver nonheme iron, spleen nonheme, and serum iron assays.E-I, qRT-PCR analysis of hepcidin, Id1, Bmp6, Spint2, and Spint1 mRNA levels.All qRT-PCR results are expressed as the amount relative to that of β-actin for each sample.Each group consists of at least six animals.The means ± SD are presented.Two-tailed Student t test was used to analyze the data for each gender and diet-matched group.No statistical significance was noticed.

Figure 8 .
Figure8.A model for Mt2 suppression of hepcidin expression through its nonproteolytic activity.In wild-type mice, Mt2 inhibits hepcidin expression primarily by binding to and interfering with the hepcidin-inducing complex that contains BMP receptors (BMPRs), Hjv, Tfr2, Hfe, and Neo1 on the plasma membrane of hepatocytes.The iron induction of hepcidin expression is accomplished by holo-Tf-mediated stabilization of Tfr2 and the upregulated transcription of Bmp6 gene in the liver sinusoidal endothelial cells by increased iron store (A and B).Lack of Mt2 will upper-shift the hepcidin expression with no impact on iron-mediated induction (C and D).
Figure 1.Hepcidin is an iron regulatory hormone that is secreted mainly by hepatocytes.A, diagram of MT2 protein.Cyto, cytoplasmic domain; TM, +/+ counterparts but they are more prone to transmembrane domain; SEA, sea urchin sperm protein, enteropeptidase agrin; CUB, complement protein subcomponents C1r/C1s, urchin embryonic growth factor and bone morphogenetic protein 1 domain; L, low-density lipoprotein receptor class-A domain; Catalytic, serine protease (S/P) catalytic domain.The arrow indicates the predicted autocleavage activation site.B, hepcidin inhibits iron efflux from duodenum, spleen, and liver into the circulation by blocking the plasma-membrane iron exporter, FPN.C, diagram of the key components that are involved in the induction and suppression of hepcidin expression in the liver.The inhibitor of the differentiation-1 gene (Id1) is a direct downstream target of BMP signaling.There is no known function of Id1 in iron homeostasis.D, diagram for HAI-2 inhibition of the proteolytic activity of MT2.E, iron induction of hepcidin expression by increases in Bmp6 gene transcription and TfR2 protein stabilization.