Diana Miao ‘14, Yann Gibert1, Aileen Zhen1, Victoria Lattanzi1 and Paula Fraenkel1,2

1Division of Hematology/Oncology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA; 2Supervising Scientist

Hemojuvelin (Hjv) is a member of the repulsive-guidance molecule (RGM) family that upregulates transcription of the iron regulatory hormone hepcidin by activating the bone morphogenetic protein (BMP) signaling pathway in mammalian cells. Hepcidin has been called the master regulator for iron, and hepcidin expression and regulation deviations have been implicated in disease such as juvenile hemochromatosis and thalassemia. This paper evaluates the effects of hjv on hepcidin expression in embryonic zebrafish development. In contrast with previous studies in the adult mammalian model that suggests hjv regulates the expression of hepcidin, we found in the embryonic vertebrate model that the hjv transcript is only weakly expressed in the liver at the time of hepcidin expression. Furthermore, overexpression of hjv fails to increase expression of hepcidin and increases non-heme iron staining. A possible new alternate function for hjv may be in notochord and somite development, as hjv was strongly expressed in the notochord and somites of the zebrafish embryo and morpholino knockdown of hjv impaired the development of these structures. Overexpressing hjv also increases liver size. Overall, while hepcidin expression seems to be mediated by proteins other than hjv in the embryonic zebrafish model, hjv may have other necessary functions in somite and liver development.


Regulation of tissue iron levels is necessary for proper vertebrate metabolism, homeostasis, and growth.  Iron plays a significant role as a component of hemoglobin, which is required for oxygen transport. As the vertebrate organism has no means for actively excreting iron, iron uptake must be carefully controlled [1].

The body recycles the majority of iron stores by breaking down aged red blood cells. Iron lost from the skin and intestinal tract is replenished from dietary iron. Dietary iron can either be stored inside a cell as ferritin, in which case it will be disposed in feces when the cell dies, or it can be exported from of the cell by binding to the transmembrane protein ferroportin (fpn). Another important class of proteins, transferrins, are glycoproteins that bind to iron in the blood to control free iron levels in body fluids. This paper will focus on the regulation of hepcidin, a peptide hormone produced by the liver that acts as the “master regulator of iron metabolism” by inhibiting ferroportin. Hepcidin meanwhile, is hypothesized to be inhibited by the soluble form of the protein hemojuvelin (hjv or RGMc).

Hepcidin is a small, cysteine-rich peptide with antimicrobial properties and has been implicated in iron regulation.  Hepcidin is mostly found in the liver, circulates through blood vessels, and is excreted in urine.  Hepcidin binding to fpn causes fpn to be internalized and degraded by proteasome interactions [2, 3], thus promoting iron retention inside the cell. Hepcidin inhibits intestinal cells from exporting iron from the enterocyte into the adjacent capillaries, thereby suppressing iron absorption in the intestines. Hepcidin also promotes iron storage in the liver and inhibits the recycling of iron by macrophages [2-4]. To achieve a homeostatic balance of iron in the body, hepcidin is negatively regulated in relation to iron levels: high iron levels stimulate hepcidin release, whereas erythropoietic activity inhibits hepcidin. Overexpression of hepcidin causes hypoferremia, while hepcidin deficiency induces iron overload.

Hjv is a protein in the repulsive guidance molecule (RGM) family, which functions in iron regulation by signaling through the bone morphogenetic protein (BMP) pathway. In the mammalian model, hjv has been found to upregulate the iron regulatory hormone hepcidin [5].  Membrane-bound hjv binds to neogenin, increases intracellular iron storage, and increases hepcidin expression through the BMP pathway in vitro [9].

While iron concentration does not directly influence hjv expression, increased iron concentration promotes binding to the membrane as well as its activation [2, 3]. Conversely, low iron levels cause hjv to stay in a soluble state [10].  In mammalian models, furin-mediated cleavage of membrane-bound hjv produces soluble hjv [2, 11], which antagonizes membrane-bound hjv [2, 3] and reduces hepcidin expression.

Mutations in hjv have been implicated in causing juvenile hemochromatosis [2, 5] and in thalassemia [6-8]. Juvenile (Type IIA) hemachromatosis displays abnormalities in pathways similar to those affected in hereditary (Type I) hemachromatosis, but symptoms start between ages 15-30 instead of in middle age and more advanced adulthood . The body overabsorbs iron, which then accumulates in major organs, especially the liver. Over time, this iron overload leads to chronic diseases such as diabetes and liver cirrhosis. Thalassemia causes the formation of irregular globin chains in hemoglobin, causing anemia. Exploring the hjv pathway  in the vertebrate model could illuminate new avenues for treatments of these diseases.

In this report, we overexpress and knock down hjv in the zebrafish, Danio rerio, and investigate the effects on hepcidin expression and iron distribution. Hepcidin has previously been explored in the postnatal mammalian mouse model, but little is known about the role of hepcidin in embryos. Previous studies in humans have found that hepcidin mRNA expression is much higher in iron-overloaded patients than in hereditary hemochromatosis patients and controls, suggesting that hepcidin and hjv mRNA are both upregulated in response to iron overload. Meanwhile, another study in mice found that hjv levels were not affected by parenteral iron loading, and this result was also found in humans [24]. The purpose of these experiments is two-fold: one, to develop the zebrafish embryo as a model to study embryonic regulation of hepcidin expression and iron levels in vertebrates; and two, to clarify the interactions between hepcidin and hjv.

Zebrafish are indispensable model organisms due to the ease with which they can be raised, observed, and manipulated.  Their small size, simple diet, and hardy nature make them easy to rear and breed. Their relatively short development time, egg to larvae in 72 hours, and clear embryos facilitate monitoring of embryonic development. Oocyte fertilization outside the mother also abrogates the need to sacrifice the mother organism [2], as is the case in the mouse model.

Already, the zebrafish has been established as a valuable model organism for studying anemia and iron metabolism disorders. Large-scale genetic screens have identified embryos with hypochromic anemia due to recessive mutations in various genes involved in iron regulation, including divalent metal transporter 1 (DMT1), transferrin receptor 1a (Tfr1a), and ferroportin1 (fpn1). Some studies have already been done on hepcidin in zebrafish. Hepcidin expression begins in wild-type zebrafish embryos at 36 hours postfertilization (hpf). It is highly conserved in the vertebrate lineage. Zebrafish hepcidin peptides are 52% identical to human hepcidin.  In human cells in vitro, zebrafish hepcidin mediates degradation of fpn and retention of cellular iron [2].

Zebrafish also offer potential for simple gene manipulation. We used microinjection of morpholino oligonucleotides [2, 3] and cRNA into fertilized zebrafish oocytes at the 1-cell stage to create knockdown and overexpression of hjv, respectively. Gene expression was analyzed through real-time quantitative polymerase chain reaction (qPCR) and whole-mount in situ hybridization (WISH).

In this experiment, we hypothesized that knocking down hjv would decrease hepcidin, as observed in the mammalian model. Conversely, we hypothesized that overexpressing hjv would increase hepcidin concentration. Contrary to expectations, our experiments revealed that neither knocking down nor overexpressing hjv had any effect on hepcidin concentration. However, we did find that hjv deficiency was associated with novel developmental defects in the notochord, somites, and liver. These new findings may suggest ancestral functions of hjv other than iron regulation.


The hjv transcript is weakly expressed in the liver at the time of hepcidin expression.

To assess hjv levels in the zebrafish embryo, we completed whole mount in situ hybridization on wild-type zebrafish embryos using hjv riboprobes at selected time points. Zebrafish at 11 hpf showed significant staining in the notochord (Figure 3A). Hjv staining extended to the somites by 18 hpf (Figure 3B). However, we were surprised to find that zebrafish failed to display hjv expression in the liver at 50 hpf and 72 hpf (Figures 3C and 3D), because hepcidin is strongly expressed in the liver at 55 hpf (Figure 2A).

Figure 3
Figure 3. The hjv transcript is not expressed in the zebrafish embryonic liver at the time of hepcidin expression. Whole mount in situ hybridization for hjv. Hjv is expressed in notochord at 11 hpf (A) (dorsal view), in the somites at 18 hpf (B) (lateral view with yolk removed), but not in the liver at 50 hpf (C) or 72 hpf (D).

Overexpression of zebrafish hjv fails to increase hepcidin expression in zebrafish embryo.

To evaluate the role of hjv expression in zebrafish embryos at 55 hpf, when it does not normally appear, we injected embryos with hjv cRNA and analyzed hepcidin expression. Whole mount in situ hybridization showed no noticeable increase in hepcidin expression (Figures 4A and 4B), suggesting that hjv has no role in regulating hepcidin during embryonic zebrafish development. However, there was a small decrease in liver size (Figures 4C and 4D).

Figure 4a-d
Figure 4A-D. Overexpression of zebrafish hjv fails to increase hepcidin expression in zebrafish embryos. A-D. Whole mount in situ hybridization at 55 hpf for hepcidin (A, B) or foxa3 (C, D) on embryos injected at the one-cell stage with hjv cRNA. Foxa3 was used as a marker for the liver (arrowhead) and intestine (arrow). WISH showed no increase in hepcidin expression.

We used quantitative real-time PCR to more finely assess possible small changes in hepcidin expression. qPCR results confirmed that injection of hjv cRNA failed to increase embryonic hepcidin levels (Figures 4E and 4F). In fact, compared to b-actin and LFABP, hepcidin levels seem to have slightly decreased, though this decrease did not meet the criteria for a significant change. This decrease may be due to slight toxicity from the side-effects of microinjection.

Figure 4e-f
Figure 4E-F. Overexpression of zebrafish hjv fails to increase hepcidin expression in zebrafish embryos. . E, F. Quantitative real-time RT-PCR at 72 hpf reiterated lack of change in hepcidin expression relative to β-actin (E) or to LFABP (F) following injection of hjv cRNA.

Overexpressing hjv increases non-heme iron staining at 55 hpf.

To study iron levels in zebrafish embryos, we used whole mount non-heme iron staining at 55 hpf. Wild-type and hjv MO2 embryos exhibited normal iron staining (Figures 5 A, B, G, and H). Interestingly, hjv cRNA displayed increased iron staining in the proctodeum and somites (Figures 5C and 5I), contrary to the expected iron-deficient phenotype seen in weissherbst (Figures 5F and 5L). Chianti mutants, which are deficient in transferrin receptor 1, also displayed increased staining in the proctodeum and somites (Figures 5D and 5J). Chianti showed increased staining in the dorsal spinal cord as well. In contrast, decreased iron staining was seen in transferrin-a deficient gavi (Figures 5E and 5K). Again, hjv seems to be independent of hepcidin; hjv MO would be expected to trigger decreased hepcidin and increased embryonic iron deposition, but iron levels in the embryonic tissues did not appear to have increased. This finding seems to contradict the hypothesis that hjv modulates iron transport in hepatocytes (Mal).
Figure 5a-f
Figure 5g-i

Figure 5j-l
Figure 5. Whole mount non-heme iron staining of zebrafish embryos at 55 hpf with 5x additional magnification of boxed regions. We observed normal iron staining in uninjected wild-type (A, J) and hjv MO injected fish (B, H), but increased iron staining (black arrows) in the somites and proctodeum (terminal gut) of hjv cRNA injected (C, I) fish. The increased staining resembles erythroid transferrin receptor deficient mutant chianti (cia) (D, J). Chianti also had increased iron staining in the dorsal spinal cord (blue arrows). As expected, decreased intraembryonic iron staining was observed in the transferrin-a deficient mutant gavi (gav) (E, K) and in the ferroportin deficient mutant weissherbst (weh) (F, L). Scale bar represents 200 microns.

Knockdown of hjv does not significantly impair hepcidin expression at 55 hpf, though it does affect liver size.

To assess the effects of hjv on hepcidin expression, we examined hjv MO knockdown embryos for expression of hepcidin using whole mount in situ hybridization. We stained embryos at 55 hpf using hepcidin riboprobes to check hepcidin transcript levels. We also ran foxa3 riboprobes in parallel to assess liver and intestine size.

Control embryos showed moderate hepcidin expression (Figures 2A and 2G). Heat-shocked tg(hsp70:bmp2b) embryos showed a  large increase in hepcidin expression (Figures 2B and 2H), while dorsomorphin-treated embryos showed no hepcidin expression (Figures 2C and 2I). Because hepcidin is regulated within the BMP pathway, it is expected that increasing BMP expression using heat-shock would increase hepcidin expression. Dorsomorphin, meanwhile, blocks BMP receptors and as expected, completely abrogates hepcidin expression.

To our surprise, zebrafish injected with hjv MO1 and hjv MO2 showed no significant change in hepcidin expression (Figures 2D and 2E), despite previous studies in the postnatal mouse model showing that decreasing hjv caused hepcidin deficiency and massive iron overload. There does seem to be a slight decrease in liver size in hjv MO embryos, however, as shown by decreased foxa3 staining (Figures 2J and 2K).

Figure 2a-k
Figure 2A-K. Whole mount in situ hybridization at 55 hpf for hepcidin (blue arrow). Foxa3 (G-K) was used as a marker for the liver (arrowhead) and intestine (black arrow). Compared to controls (A), bmp2b induction (B) increased hepcidin expression, while treatment with dorsomorphin (C) from 28-55 hpf abrogated hepcidin expression. Liver size remained unchanged (G, H, I). Knockdown of hjv by blocking translation (D, J), or by interfering with splicing (E, K), did not significantly change hepcidin expression, but slightly reduced liver size.

To ensure that hjv splicing had been altered in hjv MO2 injected zebrafish, we ran gel electrophoresis on the RT-PCR products of mRNA from wild-type and injected zebrafish. The gel showed a decrease in mRNA length of about 50 nucleotides (Figure 2L). This is consistent with aberrant splicing of the hjv transcript. Analysis by quantitative real-time RT-PCR revealed no significant change in hepcidin to b-actin and hepcidin to LFABP ratios following hjv MO1 and MO2 knockdown (Figures 2M and 2N), supporting  the idea that hjv expression does not influence hepcidin expression.

Figure 2l
Figure 2L. Gel electrophoresis on RT-PCR products of spliced mRNA confirmed a deletion of about 50 nucleotides, consistent with aberrant splicing of hjv transcripts.
Figure 2m-n
Figure 2M-N. Quantitative real-time RT-PCR at 72 hpf demonstrated no significant change in hepcidin to the ubiquitous gene β-actin (M) and hepcidin to the liver-specific gene LFABP (N) levels following hjv knockdown by hjv MO2 or a combination of hjv MO1 and MO2.

Morpholino knockdown of hjv results in notochord and somite abnormalities.

In mammalian studies, hemojuvelin (hjv) was associated with the iron regulatory pathway. However, we found that hjv plays a role in somite and notochord formation in zebrafish embryos. Light microscopy images of the zebrafish from the dorsal view show that embryos injected with hjv morpholino 1 (Figures 1B) or hjv morpholino 2 (Figure 1C) have notochord abnormalities. Hjv MO1 targets the 5’ UTR of hjv, inhibiting mRNA translation. Hjv MO2 is a non-overlapping morpholino that targets exon donor 2, which interferes with the proper splicing of mRNA transcripts. We used in situ hybridization with no tail riboprobes to emphasize the kinked zebrafish notochord phenotype (Figures 1F and 1G). The no tail gene is persistently expressed in the developing notochord [19].

Hjv mismatch morpholino 2 (MMO2) was designed from hjv MO2 with a five nucleotide substitution mismatch, rendering it unable to interfere with splicing. It acted as a negative control for the side-effects of morpholino injection. Wild-type embryos and embryos injected with MMO2 (Figures 1A, 1E, and 1F) showed no notochord abnormalities. Injection of both morpholinos aggravated the kinked notochord phenotype (Figure 1D).

Figure 1a-g
Figure 1A-G. Morpholino knockdown of hjv results in notochord and somite abnormalities. A-E. Light microscopy of zebrafish embryos at 15 hpf in dorsal view. Uninjected embryos (A) and embryos injected with a mismatch control morpholino (E) had straight notochords, while embryos injected with hjv MO1 (B) or MO2 (C) had distorted notochords (black arrows). Injecting both morpholinos (D) aggravated the phenotype. F,G. Whole mount in situ hybridization at 18 hpf with no tail, which stains the notochord, highlights the bent notochord in hjv MO1 injected morphants.

Knockdown of hjv also resulted in somite irregularities. Light microscopy from the anterior-posterior dimension showed clearly delineated V-shaped somites in wild-type and hjv MMO2-injected embryos (Figures 1H and 1I). Embryos injected with hjv MO1 and MO2 displayed fewer U-shaped somites (Figure 1J).

Figure 1h-j
Figure 1H-J. Light microscopy of the tail at 24 hpf lateral view (top) with additional 3.5x enlargement of area labeled in red (below). Uninjected (H) and MMO2 injected (I) embryos had clearly delineated chevron-shaped somites (arrows), while hjv morphants (J) had U-shaped, flattened somites.


Hjv does not function in regulating hepcidin in the embryonic zebrafish model.

Hjv knockdown by morpholinos and hjv overexpression with cRNA had no significant effect on hepcidin transcript levels in the embryonic zebrafish model. A temporal separation between hepcidin and hjv expression also supported the hypothesis that hjv does not regulate hepcidin. This discovery came as a surprise, since postnatal hjv knockout mice exhibit extreme hepcidin deficiency and iron overload [5]. Thus, while mutations in hjv are correlated with low hepcidin and hjv gene expression is found in the same areas of the skeletal and cardiac muscles and liver as hepcidin, our results suggest that embryonic iron homeostasis in vertebrates does not require hjv.

Alternately, differences in species development may account for the different uses for embryonic hepcidin. Fish utilize stored iron in the yolk regulated by ferroportin and transferrin to obtain iron for erythropoiesis. All the iron the fish embryo needs for development is deposited in the yolk prior to fertilization. Placental mammals, in contrast, have access to a dynamic store of iron from the mother’s blood circulation via the placenta. Future studies will be needed to test whether mammalian embryos need hepcidin.

Though hjv does not seem to influence iron homeostasis in the embryonic zebrafish, it may play a role in the adult zebrafish. Unfortunately, we have no model thus far to test this hypothesis, as the effects of MO injection dissipate after four days. One possible method for future experimentation is developing a transgenic zebrafish strain with an inducible construct that regulates hjv expression and  specifically targets the liver. This strain would allow us to study the effects of hjv specifically in the adult zebrafish liver independent of deleterious side effects of generalized hjv deficiency, like the observed notochord and somite mutations.

Hjv plays a role in notochord signaling during zebrafish development.

The notochord irregularities in hjv MO injected embryos suggest an original role for hjv in regulating somite and notochord development, possibly via the BMP pathway. This role may not have been previously discovered in the mouse model because other RGMs could compensate for hjv.

The effects of hjv on notochord and somite formation are  significant because the notochord provides structural support prior to vertebral column formation and influences somite formation [20]. Additionally, chemical signals from the notochord induce thoracic mesoderm signaling, so problems with the notochord cause large-scale malformations in the zebrafish. Fish injected with hjv MO do not survive to maturity due to fatal developmental defects.

Future experiments need to be conducted to determine which proteins hjv interacts with in notochord signaling. Continued research is also needed to determine whether the somite phenotype is a sideeffect of notochord defects or an autonomous effect of hjv knockdown. Hjv has already been shown to interact with neogenin, and the neogenin knockdown fish shows a similar muscle phenotype [21]. Future experiments could include simultaneously injecting reduced doses of both hjv MO and neogenin MO. If these morpholinos act synergistically, we could deduce that hjv and neogenin interact genetically to regulate somite formation. Another interesting interaction to explore would be between hjv and MyoD, a protein known to interact with the BMP pathway [22].

The robust expression of hjv in skeletal and cardiac muscle has been noted previously in the adult murine model [23], but little research has been done to explore the function of hjv in these tissues. Our results suggest that hjv may have a role in embryonic vertebrate muscle development antecedent to its role as a regulator of hepcidin.

Understanding the interactions of hepcidin within the BMP pathway has practical applications in curing disease.

Mutations in hepcidin have been implicated in causing juvenile hemochromatosis and thalassemia, while inadequate hepcidin levels are thought to exacerbate iron overload. In these diseases, hepcidin deficiency causes elevated absorption of iron from the diet and increased recycling of iron by macrophages, contributing to severe iron overload. Understanding the role of hepcidin in the zebrafish creates a new model for researching iron-regulatory defects and illuminates new clinical pathways for future medicines.  The new finding of hjv involvement in notochord signaling and somite development also suggests a new role for hjv in the embryonic vertebrate model.

Materials and methods

Zebrafish strains, maintenance and determination of genotype

Zebrafish were maintained as described. Ethical approval was obtained from the Institutional Animal Care and Use Committees of Children’s Hospital and Beth Israel Deaconess Medical Center.  Tg (hsp70:bmp2b) zebrafish are described elsewhere [12].  Heterozygote carriers of tg(hsp70:bmp2b) were identified by crossing with WT zebrafish, subjecting the progeny embryos at the shield-stage to heat shock at 37o C for 40 min, and assessing the percentage of ventralized or dorsalized embryos produced [12].  Hypochromic anemia mutants used included chianti (ciaTu25f), gavi (gavIT029), and weissherbst (wehTp85c) [10, 13, 14]. Chianti mutants suffer from hypochromic anemia due to a mutation in transferrin receptor 1(tfr1), which is necessary for iron uptake into erythroid precursors. Gavi are also hypochromic anemia mutants; they have a mutation in transferrin-a, which transports iron to developing red blood cells. Weissherbst fish have a mutation in ferroportin1 (fpn1) that renders them unable to transport iron from the yolk sac to the rest of the embryo. Affected gavi and weissherbst fish do not survive to maturity due to fatal anemia, so instead, heterozygous fish were raised to maturity and in-crossed to yield anemic mutants. Mutants were scored at 48 hpf for anemia by looking for the absence of red color in hemoglobin in circulating blood under a light microscope.

Morpholino and cRNA injection

Antisense morpholino oligonucleotides (MOs)[15], obtained from Gene Tools, Inc. (Philomath, OR), were designed either to interfere with translation or to impair appropriate splicing of transcripts. Morpholinos for hjv (sequences available upon request) were injected at the one-cell stage at 0.2 mM concentration, or 3 nL in 1x Danieau medium, supplemented with phenol red to improve visibility during injection and fluorescein. Embryos were scored at 24 hpf for fluorescence to ensure successful injection, as the fluorescein should have diffused into all cells of injected embryos. Non-fluorescent embryos were not evaluated.

Full-length zebrafish hjv was cloned into the pCS2+ vector. The vector was digested with NotI and sense hjv cRNA was synthesized using the SP6 mMachine Kit (Ambion, Austin, TX). The cRNA was injected at a concentration of 1000 ng/microliter in the same way as hjv MO injections.

Heat shock

For assessment of bmp2b expression, embryos at 48 hpf were incubated at 37oC (heat shock) for 40 min and then returned to 28.5 oC for 6 hours of incubation.  The embryos were then fixed in RNAlater (Ambion) at selected time points.

Chemical treatment

Embryos were treated with 40 µM dorsomorphin [2, 4, 6] dissolved in dimethyl sulfoxide (DMSO) from 28-55 hpf.  Control embryos were treated with DMSO only.

Whole mount in situ hybridization

Whole mount in situ hybridizations were performed as previously described [16]. Development of endogenous pigments in live fish was inhibited by supplementing the embryo medium with 1-phenyl-2-thiourea (PTU) at 8 hpf at a final concentration of 0.2 mM. Embryos were fixed at selected time points in 4 % paraformaldehyde-PBS (PFA) overnight at 4ºC and then transferred to methanol for storage at -80ºC. Day one of in situ hybridization involved protein digestion by Proteinase K for 30 minutes to expose cellular RNA. Embryos were then incubated at 69ºC with 20 µL probe in formamide solution. The following antisense riboprobes were generated for use in the in situ hybridizations: hemojuvelin, hepcidin [10], foxa3 [12], and no tail (gift of G. Begemann). On day two, embryos were exposed to antibodies, and on day three, embryos were stained using nitro blue tetrazolium (NBT) and 5-Bromo-4-chloro-3-indolyl phosphate (BCIP). Embryos were left in 1 mL of staining solution in the dark for about 3 hours.

Photomicrographs of in situ hybridizations were obtained with a BX51 compound microscope (Olympus, Center Valley, PA) at 100x magnification using Nomarski optics and a Q-capture 5 digital camera (QImaging, Surrey, BC, Canada). Images were processed using Adobe Photoshop software.  Scale bars represent 100 microns, unless otherwise indicated.

Whole mount embryo iron staining

PTU-treated embryos for iron-staining were fixed in 4% PFA at selected time points. Diaminobenzidine (DAB) enhanced-staining for ferric iron was performed as described [2, 4]  following fixation.  Embryos were incubated in 2.5% ferrocynanide/0.25 M HCl mix for 30 minutes, rinsed in PBSTween, incubated in 0.3% hydrogen peroxide in 100% methanol for 20 minutes, rinsed in PBSTween again, and then incubated at room temperature in (DAB) substrate. Wild-type embryos incubated in HCl without potassium ferrocyanide served as a negative control. After iron-staining, embryos were transferred to 70% glycerol and stored at 4ºC. Photomicrographs were taken soon after to avoid losing staining solution into the glycerol using an SZX51 zoom stereomicroscope (Olympus) at 40x magnification with a DP-71 camera (Olympus).

Quantitative analysis of gene expression

At specified time points, embryos were pooled in groups of 20, anesthetized with Tricaine, and fixed in RNAlater (Ambion). Embryo pools were frozen with liquid nitrogen and ground with mortar and pestle. RNA extraction and purification was performed using the RNeasy lysis kit (QIAGEN). RNA yield was quantified by absorption at 260 nm using a light spectrophotometer. RNA was stored at -20ºC before reverse transcriptase reaction.

To generate cDNA, reverse transcription was performed on RNA from 1 pool of embryos using Superscript II reverse transcriptase in a 20 µL reaction. Resulting cRNA was diluted to 100 µL and stored at -20ºC.

Quantitative real-time polymerase chain reaction was performed using 10 µL of cDNA template. For reactions normalized to b-actin, the 50 µL reaction cocktail also included 0.4 µM forward primer, 0.4 µM reverse primer, and 0.2 µM  probe, along with 1x Quantitect Probe PCR Master Mix (QIAGEN). Probes had a VIC reporter and TMARA quencher. Quantitative real-time RT-PCR for LFABP expression was performed by amplifying with LFABP primers [17] with 1x SYBR Green Master Mix (Applied Biosystems, Inc., Foster City, CA), according to the manufacturer’s instructions. Detection and analysis were performed on an ABI 7000 (Applied Biosystems). Reactions were performed in duplicate. Cycling parameters were 2 minutes at 50ºC and 10 min at 95 ºC, followed by 40 cycles of 15 seconds at 95 ºC to 1 minute at 60 ºC. Detection and analysis were performed on an ABI 7000 (Applied Biosystems).

Water replaced cDNA template in the non-template controls, which acted as a negative control for contamination. We also ran two reverse trancriptase negative (RT-) wells as negative controls against possible chromosomal DNA contamination remaining after RNA purification.

RT-PCR assay for hepcidin was normalized to β-actin expression as previously described [14]. Transcript abundance was expressed as a fold-increase over calibrator, according to the method described [18]. The calibrator group is in the first column of each graph. Data presented are the means and standard errors. N = 2-8 pools per time point or condition.

Biostatistical analysis

Heterogeneity among cohorts was analyzed by ANOVA using Prism 5 (GraphPad Software, Inc., San Diego, CA). Tests for heterogeneity used the natural log for assessment of transcript levels.  All estimates and standard errors presented have been converted back to the original units.  When the global P-value obtained from the ANOVA analysis was statistically significant, pairwise comparisons between the cohorts were performed using two-tailed Student’s t-tests with a Bonferroni correction for multiple comparisons.  P-values less than 0.05 were deemed statistically significant and are indicated by an asterisk.


Hentze MW, M.M., Andrews NC, Balancing acts: molecular control of mammalian iron metabolism. Cell, 2004. 117(3): p. 285-297.

Papanikolaou F, S.M., Ludwig EH, et al., Mutations in HFE2 cause iron overloadin chromosome 1q-linked juvenile hemochromatosis. Nat Genet, 2004. 36: p. 77-82.

Niederkofler C, S.R., Arber S, Hemojuvelin is essential for dietary iron sensing, and its mutation leads to severe iron overload. J Clin Invest, 2005. 115(8): p. 2180-6.

Nemeth E, T.M., Powelson J, et al., Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization. Science, 2004. 306: p. 2090–2093.

Huang FW, P.J., Pinkus GS, Fleming MD, Andrews NC A mouse model of juvenile hemochromatosis. J. Clin. Invest, 2005. 115(8): p. 2187–91.

Papanikolaou G, T.M., Christakis JI, et al., Hepcidin in iron overload disorders. Blood, 2005. 2005: p. 4103-4105.

Kattamis A. Papassotiriou I. Palaiologou D, e.a., The effects of erythropoetic activity and iron burden on hepcidin expression in patients with thalassemia major. Haemotologica, 2006. 91: p. 809-812.

Camberlain E. Zanninelli G, D.L., et al., Anemia in beta-thalassemia patients targets hepatic hepcidin transcript levels independently of iron metabolism genes controlling hepcidin expression. . Haematologica, 2008. 93: p. 111-115.

Zhang, A., et al., Evidence that inhibition of hemojuvelin shedding in response to iron is mediated through neogenin. J Biol Chem, 2007. 282(17): p. 12547-56.

Fraenkel PG, Gibert Y., Holzheimer JL, et al., Transferrin-a modulates hepcidin expression in zebrafish embryos. Blood, 2009. 113: p. 2843-2850.

Lin, L., et al., Soluble hemojuvelin is released by proprotein convertase-mediated cleavage at a conserved polybasic RNRR site. Blood Cells Mol Dis, 2008. 40(1): p. 122-31.

Chocron S, V.M., Rentzsch F, Hammerschmidt M, Bakkers J, Zebrafish Bmp4 regulates left-right asymmetry at two distinct developmental time points. Dev Biol, 2007. 305: p. 577-588.

Donovan A, B.A., Zhou Y, et al., Positional cloning of zebrafish ferroportin1 identifies a conserved vertebrate iron exporter. Nature, 2000. 403: p. 776-781.

Fraenkel PG, T.D., Donovan A, Zahrieh D, Zon LI, Ferroportin1 is required for normal iron cycling in zebrafish. J Clin Invest, 2005. 115: p. 1532-1541.

Nasevicius A, L.J., Ekker SC, Distinct requirements for zebrafish angiogenesis revealed by a VEGF-A morphant. Yeast, 2000. 17: p. 294-301.

Thisse B, H.V., Lux A, et al., Spatial and temporal expression of the zebrafish genome by large-scale in situ hybridization screening. Methods Cell Biol, 2004. 77: p. 505-519.

Goessling W, N.T., Lord AM, et al., APC mutant zebrafish uncover a changing temporal requirement for wnt signaling in liver development. Dev Biol, 2008. 320: p. 161-174.

Trant JM, G.S., Ackers J, Chung BC, Place AR, Developmental expression of cytochrome P450 aromatase genes (CYP19a and CYP19b) in zebrafish fry (Danio rerio). J Exp Zool, 2001. 290: p. 475-483.

Ransom DG, H.P., Odenthal J, et al., Characterization of zebrafish mutants with defects in embryonic hematopoiesis. Development, 1996. 123: p. 311-319.

Nusslein-Volhard C, D.R., Zebrafish: A Practical Approach. 2005, Oxford, UK: Oxford University Press.

Mawdsley, D., Cooper, HM, Hogan, BM, Cody, SH, Lieschke, GJ, and Heath, JK, The Netrin receptor Neogenin is required for neural tube formation and somitogenesis in zebrafish Dev. Biol, 2004 269(1): p. 302-15.

Lumsden AL, H.T., Dayan S, Lardelli MT, Richards RI, Huntingtin-deficient zebrafish exhibit defects in iron utilization and development. Hum Mol Genet, 2007. 16: p. 1905-1920.

Babitt JL, Huang FW, Wrighting DM, Xia Y, Sidis Y, Samad TA, Campagna JA, Chung RT, Schneyer AL, Woolf CJ, Andrews NC, Lin HY: Bone morphogenetic protein signaling by hemojuvelin regulates hepcidin expression. Nat Genet 2006. 38: p. 531–539.

Malyszko, J. Hemojuvelin: The Hepcidin Story Continues. Kidney Blood Press Res 2009. 32: p. 71-76.