Liver hemojuvelin protein levels in mice deficient in matriptase-2 (Tmprss6)

Mutations of the TMPRSS6 gene, encoding the serine protease matriptase-2, lead to iron-refractory iron deficiency anemia. Matriptase-2 is a potent negative regulator of hepcidin. Based on in vitro data, it has recently been proposed that matriptase-2 decreases hepcidin synthesis by cleaving membrane hemojuvelin, a key protein of the hepcidin-regulatory pathway. However, in vivo evidence for this mechanism of action of matriptase-2 is lacking. To investigate the hemojuvelin-matriptase-2 interaction in vivo, an immunoblot assay for liver membrane hemojuvelin was optimized using hemojuvelin-mutant mice as a negative control. In wild-type mice, two hemojuvelin-specific bands of 35kDa and 20kDa were detected in mouse liver membrane fraction under reducing conditions; under non-reducing conditions, a single band of approximately 50kDa was seen. Phosphatidylinositol-specific phospholipase C treatment confirmed binding of the detected protein to the cell membrane by a glycosylphosphatidylinositol anchor, indicating that the major form of mouse liver membrane hemojuvelin is a glycosylphosphatidylinositol-bound heterodimer. Unexpectedly, comparison of liver homogenates from Tmprss6+/+ and Tmprss6-/- mice revealed significantly decreased, rather than increased, hemojuvelin heterodimer content in Tmprss6-/- mice. These data do not provide direct support for the concept that matriptase-2 cleaves membrane hemojuvelin and may indicate that, in vivo, the role of matriptase-2 in the regulation of hepcidin gene expression is more complex.

Effect of iron overload and iron deficiency on liver hemojuvelin protein

INTRODUCTION

Hemojuvelin (Hjv) is a key component of the signaling cascade that regulates liver hepcidin (Hamp) expression. The purpose of this study was to determine Hjv protein levels in mice and rats subjected to iron overload and iron deficiency.

METHODS

C57BL/6 mice were injected with iron (200 mg/kg); iron deficiency was induced by feeding of an iron-deficient diet, or by repeated phlebotomies. Erythropoietin (EPO)-treated mice were administered recombinant EPO at 50 U/mouse. Wistar rats were injected with iron (1200 mg/kg), or fed an iron-deficient diet. Hjv protein was determined by immunoblotting, liver samples from Hjv-/- mice were used as negative controls. Mouse plasma Hjv content was determined by a commercial ELISA kit.

RESULTS

Liver crude membrane fraction from both mice and rats displayed a major Hjv-specific band at 35 kDa, and a weaker band of 20 kDa. In mice, the intensity of these bands was not changed following iron injection, repeated bleeding, low iron diet or EPO administration. No change in liver crude membrane Hjv protein was observed in iron-treated or iron-deficient rats. ELISA assay for mouse plasma Hjv did not show significant difference between Hjv+/+ and Hjv-/- mice. Liver Hamp mRNA, Bmp6 mRNA and Id1 mRNA displayed the expected response to iron overload and iron deficiency. EPO treatment decreased Id1 mRNA, suggesting possible participation of the bone morphogenetic protein pathway in EPO-mediated downregulation of Hamp mRNA.

DISCUSSION

Since no differences between Hjv protein levels were found following various experimental manipulations of body iron status, the results indicate that, in vivo, substantial changes in Hamp mRNA can occur without noticeable changes of membrane hemojuvelin content. Therefore, modulation of hemojuvelin protein content apparently does not represent the limiting step in the control of Hamp gene expression.

Decreased hemojuvelin protein levels in mask mice lacking matriptase-2-dependent proteolytic activity

Matriptase-2, a membrane protein encoded by the Tmprss6 gene, is a negative regulator of hepcidin expression. Although matriptase-2 has been proposed to cleave membrane hemojuvelin, we have recently found decreased hemojuvelin protein levels in Tmprss6 -/- mice. The purpose of this study was to confirm this observation by determining hemojuvelin protein levels in another strain of mice with disrupted Tmprss6 gene, and to determine the effect of matriptase-2 deficiency on the expression of other membrane proteins participating in the bone morphogenetic protein signal transduction. Mask mice, which lack the proteolytic domain of matriptase-2, displayed decreased liver hemojuvelin protein content, while Id1 mRNA level, an indicator of hemojuvelin-dependent signal transduction, was increased. Protein levels of bone morphogenetic protein receptors Alk3 and Acvr2a were unchanged, and transferrin receptor 2 and neogenin protein levels were slightly decreased. The results confirm that the loss of matriptase-2 increases bone morphogenetic protein-dependent signaling, while paradoxically decreasing liver hemojuvelin protein content. The regulation of transferrin receptor 2 protein levels by transferrin saturation was not affected in mask mice. How the loss of matriptase-2 proteolytic activity leads to decreased hemojuvelin protein levels is at present unclear.

Liver and muscle hemojuvelin are differently glycosylated

Background

Hemojuvelin (HJV) is one of essential components for expression of hepcidin, a hormone which regulates iron transport. HJV is mainly expressed in muscle and liver, and processing of HJV in both tissues is similar. However, hepcidin is expressed in liver but not in muscle and the role of the muscle HJV is yet to be established. Our preliminary analyses of mouse tissue HJV showed that the apparent molecular masses of HJV peptides are different in liver (50 kDa monomer and 35 and 20 kDa heterodimer fragments) and in muscle (55 kDa monomer and a 34 kDa possible large fragment of heterodimer). One possible explanation is glycosylation which could lead to difference in molecular mass.

Results

We investigated glycosylation of HJV in both liver and muscle tissue from mice. PNGase F treatment revealed that the HJV large fragments of liver and muscle were digested to peptides with similar masses, 30 and 31 kDa, respectively, and the liver 20 kDa small fragment of heterodimer was digested to 16 kDa, while the 50 kDa liver and 55 kDa muscle monomers were reduced to 42 and 48 kDa, respectively. Endo H treatment produced distinct digestion profiles of the large fragment: a small fraction of the 35 kDa peptide was reduced to 33 kDa in liver, while the majority of the 34 kDa peptide was digested to 33 kDa and a very small fraction to 31 kDa in muscle. In addition, liver HJV was found to be neuraminidase-sensitive but its muscle counterpart was neuraminidase-resistant.

Conclusions

Our results indicate that different oligosaccharides are attached to liver and muscle HJV peptides, which may contribute to different functions of HJV in the two tissues.

Effect of stimulated erythropoiesis on liver SMAD signaling pathway in iron-overloaded and iron-deficient mice

Expression of hepcidin, the hormone regulating iron homeostasis, is increased by iron overload and decreased by accelerated erythropoiesis or iron deficiency. The purpose of the study was to examine the effect of these stimuli, either alone or in combination, on the main signaling pathway controlling hepcidin biosynthesis in the liver, and on the expression of splenic modulators of hepcidin biosynthesis. Liver phosphorylated SMAD 1 and 5 proteins were determined by immunoblotting in male mice treated with iron dextran, kept on an iron deficient diet, or administered recombinant erythropoietin for four consecutive days. Administration of iron increased liver phosphorylated SMAD protein content and hepcidin mRNA content; subsequent administration of erythropoietin significantly decreased both the iron-induced phosphorylated SMAD proteins and hepcidin mRNA. These results are in agreement with the recent observation that erythroferrone binds and inactivates the BMP6 protein. Administration of erythropoietin substantially increased the amount of erythroferrone and transferrin receptor 2 proteins in the spleen; pretreatment with iron did not influence the erythropoietin-induced content of these proteins. Erythropoietin-treated iron-deficient mice displayed smaller spleen size in comparison with erythropoietin-treated mice kept on a control diet. While the erythropoietin-induced increase in splenic erythroferrone protein content was not significantly affected by iron deficiency, the content of transferrin receptor 2 protein was lower in the spleens of erythropoietin-treated mice kept on iron-deficient diet, suggesting posttranscriptional regulation of transferrin receptor 2. Interestingly, iron deficiency and erythropoietin administration had additive effect on hepcidin gene downregulation in the liver. In mice subjected both to iron deficiency and erythropoietin administration, the decrease of hepcidin expression was much more pronounced than the decrease in phosphorylated SMAD protein content or the decrease in the expression of the SMAD target genes Id1 and Smad7. These results suggest the existence of another, SMAD-independent pathway of hepcidin gene downregulation.

Altered Erythro-Myeloid Progenitor Cells Are Highly Expanded in Intensively Regenerating Hematopoiesis

Regeneration of severely damaged adult tissues is currently only partially understood. Hematopoietic tissue provides a unique opportunity to study tissue regeneration due to its well established steady-state structure and function, easy accessibility, well established research methods, and the well-defined embryonic, fetal, and adult stages of development. Embryonic/fetal liver hematopoiesis and adult hematopoiesis recovering from damage share the need to expand populations of progenitors and stem cells in parallel with increasing production of mature blood cells. In the present study, we analyzed adult hematopoiesis in mice subjected to a submyeloablative dose (6 Gy) of gamma radiation and targeted the period of regeneration characterized by massive production of mature blood cells along with ongoing expansion of immature hematopoietic cells. We uncovered significantly expanded populations of developmentally advanced erythroid and myeloid progenitors with significantly altered immunophenotype. Their population expansion does not require erythropoietin stimulation but requires the SCF/c-Kit receptor signaling. Regenerating hematopoiesis significantly differs from the expanding hematopoiesis in the fetal liver but we find some similarities between the regenerating hematopoiesis and the early embryonic definitive hematopoiesis. These are in (1) the concomitant population expansion of myeloid progenitors and increasing production of myeloid blood cells (2) performing these tasks despite the severely reduced transplantation capacity of the hematopoietic tissues, and (3) the expression of CD16/32 in most progenitors. Our data thus provide a novel insight into tissue regeneration by suggesting that cells other than stem cells and multipotent progenitors can be of fundamental importance for the rapid recovery of tissue function.

B-lymphopoiesis gains sensitivity to subsequent inhibition by estrogens during final phase of fetal development

Adult B-lymphopoiesis is suppressed by the inhibitory effects of elevated estrogens during pregnancy. At the same time, hematopoietic cells in the fetal liver are resistant to this suppression by estrogens and ensure active production of B-cells. We investigated whether this unresponsiveness to estrogens of fetal cells also applies to cells obtained from a newborn liver and projects into the adult hematopoiesis when fetal liver cells are transplanted to adult mice. Mixtures of fetal liver (E14.5), neonatal liver (P0.5) and adult bone marrow (BM) cells were co-transplanted into adult primary and secondary recipients treated with high doses of estrogen in the Ly5.1/Ly5.2 congenic mouse model. Total chimerism as a proportion of all nucleated blood cells, chimerism as a proportion of B220+ B-cells, and of other blood cell lineages as well, were determined by flow cytometry. B-lymphopoiesis derived from fetal liver (E14.5) stem cells remained resistant to estrogen after transplantation into both primary and secondary adult recipients, for up to 280 days. In contrast, B-lymphopoiesis derived from neonatal liver (P0.5) stem cells was resistant to estrogen only for approximately 50 days after the primary transplantation to the adult BM microenvironment. These results provide further evidence for a critical developmental period of B-lymphopoiesis during its fetal liver stage. In the mouse, critical developmental events that allow for the subsequent expressed sensitivity of B-lymphopoiesis for suppression by estrogens after sexual maturation appear to occur during the period of late-stage fetal liver hematopoiesis before its migration to the bone marrow.

T-lymphopoiesis is Severely Compromised in Ubiquitin-Green Fluorescent Protein Transgenic Mice

Tagging cells of experimental organisms with genetic markers is commonly used in biomedical research. Insertion of artificial gene constructs can be highly beneficial for research as long as this tagging is functionally neutral and does not alter the tissue function. The transgenic UBC-GFP mouse has been recently found to be questionable in this respect, due to a latent stem cell defect compromising its lymphopoiesis and significantly influencing the results of competitive transplantation assays. In this study, we show that the stem cell defect present in UBC-GFP mice negatively affects T-lymphopoiesis significantly more than B-lymphopoiesis. The production of granulocytes is not negatively affected. The defect in T-lymphopoiesis causes a low total number of white blood cells in the peripheral blood of UBC-GFP mice which, together with the lower lymphoid/myeloid ratio in nucleated blood cells, is the only abnormal phenotype in untreated UBCGFP mice to have been found to date. The defective lymphopoiesis in UBC-GFP mice can be repaired by transplantation of congenic wild-type bone marrow cells, which then compensate for the insufficient production of T cells. Interestingly, the wild-type branch of haematopoiesis in chimaeric UBC-GFP/wild-type mice was more active in lymphopoiesis, and particularly towards production of T cells, compared to the lymphopoiesis in normal wild-type donors.

Effect of Erythropoietin on the Expression of Murine Transferrin Receptor 2

Erythropoietin (EPO) downregulates hepcidin expression to increase the availability of iron; the downregulation of hepcidin is mediated by erythroferrone (ERFE) secreted by erythroblasts. Erythroblasts also express transferrin receptor 2 (TFR2); however, the possible role of TFR2 in hepcidin downregulation is unclear. The purpose of the study was to correlate liver expression of hepcidin with the expression of ERFE and TFR2 in murine bone marrow and spleen at 4, 16, 24, 48, 72 and 96 h following administration of a single dose of EPO. Splenic Fam132b expression increased 4 h after EPO injection; liver hepcidin mRNA was decreased at 16 h. In the spleen, expression of TFR2 and transferrin receptor (TFR1) proteins increased by an order of magnitude at 48 and 72 h after EPO treatment. The EPO-induced increase in splenic TFR2 and TFR1 was associated with an increase in the number of Tfr2- and Tfr1-expressing erythroblasts. Plasma exosomes prepared from EPO-treated mice displayed increased amount of TFR1 protein; however, no exosomal TFR2 was detected. Overall, the results confirm the importance of ERFE in stress erythropoiesis, support the role of TFR2 in erythroid cell development, and highlight possible differences in the removal of TFR2 and TFR1 from erythroid cell membranes.

Cell cycle and differentiation of Sca-1+ and Sca-1- hematopoietic stem and progenitor cells

Hematopoietic stem and progenitor cells (HSPCs) are crucial for lifelong blood cell production. We analyzed the cell cycle and cell production rate in HSPCs in murine hematopoiesis. The labeling of DNA-synthesizing cells by two thymidine analogues, optimized for in-vivo use, enabled determination of the cell cycle flow rate into G2-phase, the duration of S-phase and the average cell cycle time in Sca-1⁺ and Sca-1^- HSPCs. Determination of cells with 2n DNA content labeled in preceding S-phase was then used to establish the cell flow rates in G1-phase. Our measurements revealed a significant difference in how Sca-1⁺ and Sca-1^- myeloid progenitors self-renew and differentiate. Division of the Sca-1⁺ progenitors led to loss of the Sca-1 marker in about half of newly produced cells, corresponding to asymmetric cell division. Sca-1^- cells arising from cell division entered a new round of the cell cycle, corresponding to symmetric self-renewing cell division. The novel data also enabled the estimation of the cell production rates in Sca-1⁺ and in three subtypes of Sca-1^- HSPCs and revealed Sca-1 negative cells as the major amplification stage in the blood cell development.

In vivo growth of mantle cell lymphoma xenografts in immunodeficient mice is positively regulated by VEGF and associated with significant up-regulation of CD31/PECAM1

Mantle cell lymphoma (MCL) is an aggressive lymphoma subtype with dismal prognosis. New treatments are needed to improve outcome of relapsed/ refractory disease. Recently, several drugs targeting at least partially the process of angiogenesis have been successfully tested in the therapy of MCL. Molecular mechanisms that regulate MCL-induced angiogenesis and that might represent potential new druggable targets remain, however, incompletely understood. We established two mouse models of human MCL by subcutaneous xenotransplantation of JEKO-1 and HBL-2 cell lines into immunodeficient mice. Histological analyses of xenografts confirmed their neovascularization. The growth of xenografts was significantly suppressed by single-agent therapy with bevacizumab, monoclonal antibody targeting vascular endothelial growth factor (VEGF). Subsequently, we analysed expression of 94 angiogenesis related genes in ex vivo isolated JEKO-1 and HBL-2 cells compared to in vitro growing cells using TaqMan low-density arrays. The most up-regulated genes in both JEKO-1 and HBL-2 xenografts were genes encoding platelet/endothelial cell-adhesion molecule (CD31/PECAM1), VEGF receptor 1 (FLT1), hepatocyte growth factor (HGF), angiogenin (ANG) and transcription factor PROX1. The most downregulated genes in both JEKO-1 and HBL-2 xenografts were midkine (MDK) and ephrine B2 (EPHB2). In summary, our results demonstrate an important role of angiogenesis in the biology of MCL and provide preclinical evidence of potent anti-MCL activity of bevacizumab. In addition, gene expression profiling of 94 angiogenesis-related targets revealed several in vivo up-regulated and down-regulated transcripts. The most differentially expressed target in both MCL tumours was CD31/PECAM1. Whether any of these molecules might represent a potential druggable target in MCL patients remains to be elucidated.

Iron Overload Causes Alterations of E-Cadherin in the Liver

Iron overload causes tissue damage in the liver, but its initial effects at the molecular and cellular level are not well understood. Epithelial cadherin (E-cad) is a major adhesion protein in adherens junctions and is associated with several signal transduction pathways. Dysfunction of E-cad causes instability of adherens junctions, which leads to cell invasion, cell migration, and carcinogenesis. We found in liver samples from iron-overloaded mice that the apparent molecular mass of E-cad was reduced from 125 to 115 kDa in sodium dodecyl sulphate polyacrylamide gel electrophoresis under reducing conditions and immunoblotting, and that the cellular expression of E-cad was decreased in immunohistochemistry. The mRNA level of E-cad, however, did not change significantly, suggesting that the alterations are posttranslational. Interestingly, incubation of control liver extracts with Fe2+ alone also produced the same mobility shift. Neither an oxidant nor an antioxidant influenced this shift in vitro, suggesting that reactive oxygen species, which are generated by iron and known to cause damage to macromolecules, are not involved. Treatment of the 115 kDa E-cad with deferoxamine, an iron chelator, thus removing Fe2+, shifted the molecular mass back to 125 kDa, demonstrating that the shift is reversible. The observation also implies that the alteration that causes the mobility shift is not due to transcriptional control, deglycosylation, and proteolysis. This reversible mobility shift of E-cad has not been previously known. The alteration of E-cad that causes the mobility shift might be an initial step to liver diseases by iron overload.

Hepcidin expression in the liver of mice with implanted tumour reacts to iron deficiency, inflammation and erythropoietin administration

Cancer is known to be an important cause of anaemia due to several factors including iron deficiency and inflammation. Hepcidin, a key regulator of iron metabolism, is up-regulated by iron and inflammatory stimuli such as interleukin 6, and decreased by iron deficiency, enhanced erythropoiesis and hypoxia. It is supposed to play a crucial role in changes of iron metabolism in anaemia of chronic disease, which is characterized by sequestering iron in macrophages and decreasing its availability for red blood cell production. To study the effect of tumour growth on hepcidin expression, we implanted human melanoma cells into mice and studied the changes of the amount of liver hepcidin mRNA by real-time PCR. We observed development of anaemia, which correlated with the size of the tumour. Hepcidin expression significantly decreased with the anaemia development, but in late stages we observed an increase of its expression together with an increase of mRNA for interleukin 6. However, the increase of hepcidin expression could be inhibited by exogenous erythropoietin administration. In our model of tumour growth, hepcidin expression reflected anaemia development and iron deficiency, erythropoietin administration and inflammation, and we suppose that it could therefore serve as a useful marker of these clinical situations common in cancer patients and play a role in the pathogenesis of cancer-associated anaemia.

Second bone marrow transplantation into regenerating hematopoiesis enhances reconstitution of immune system

In bone marrow transplantation (BMT), hematopoiesis-reconstituting cells are introduced following myeloablative treatment, which eradicates existing hematopoietic cells and disrupts stroma within the hematopoietic tissue. Both hematopoietic cells and stroma then undergo regeneration. Our study compares the outcomes of a second BMT administered to mice shortly after myeloablative treatment and the first BMT, with those of a second BMT administered to mice experiencing robust hematopoietic regeneration after the initial transplant. We evaluated the efficacy of the second BMT in terms of engraftment efficiency, types of generated blood cells, and longevity of function. Our findings show that regenerating hematopoiesis readily accommodates newly transplanted stem cells, including those endowed with a robust capacity for generating B and T cells. Importantly, our investigation uncovered a window for preferential engraftment of transplanted stem cells coinciding with the resumption of blood cell production. Repeated BMT could intensify hematopoiesis reconstitution and enable therapeutic administration of genetically modified autologous stem cells.

Hematopoiesis Remains Permissive to Bone Marrow Transplantation After Expansion of Progenitors and Resumption of Blood Cell Production

The immense regenerative power of hematopoietic tissue stems from the activation of the immature stem cells and the progenitor cells. After partial damage, hematopoiesis is reconstituted through a period of intense regeneration when blood cell production originates from erythro-myeloid progenitors in the virtual absence of stem cells. Since the damaged hematopoiesis can also be reconstituted from transplanted hematopoietic cells, we asked whether this also leads to the transient state when activated progenitors initially execute blood cell production. We first showed that the early reconstitution of hematopoiesis from transplanted cells gives rise to extended populations of developmentally advanced but altered progenitor cells, similar to those previously identified in the bone marrow regenerating from endogenous cells. We then identified the cells that give rise to these progenitors after transplantation as LSK CD48^– cells. In the submyeloablative irradiated host mice, the transplanted LSK CD48^– cells preferably colonized the spleen. Unlike the endogenous hematopoiesis reconstituting cells, the transplanted whole bone marrow cells and sorted LSK CD48^– cells had greater potential to differentiate to B-lymphopoiesis. Separate transplantation of the CD150^– and CD150⁺ subsets of LSK CD48^– cells suggested that CD150^– cells had a greater preference to B-lymphopoiesis than CD150⁺ cells. In the intensively regenerating hematopoiesis, the CD71/Sca-1 plot of immature murine hematopoietic cells revealed that the expanded populations of altered myeloid progenitors were highly variable in the different places of hematopoietic tissues. This high variability is likely caused by the heterogeneity of the hematopoiesis supporting stroma. Lastly, we demonstrate that during the period when active hematopoiesis resumes from transplanted cells, the hematopoietic tissues still remain highly permissive for further engraftment of transplanted cells, particularly the stem cells. Thus, these results provide a rationale for the transplantation of the hematopoietic stem cells in successive doses that could be used to boost the transplantation outcome.