Mtmr8 is essential for vasculature development in zebrafish embryos

Background Embryonic morphogenesis of vascular and muscular systems is tightly coordinated, and a functional cooperation of Mtmr8 with PI3K in actin filament modeling and muscle development has been revealed in zebrafish. Here, we attempt to explore the function of Mtmr8 in vasculature development parallel to its function in muscle development. Results During early stage of somitogenesis, mtmr8 expression was detected in both somitic mesodem and ventral mesoderm. Knockdown of mtmr8 by morpholino impairs arterial endothelial marker expression, and results in endothelial cell reduction and vasculogenesis defects, such as retardation in intersegmental vessel development and interruption of trunk dorsal aorta. Moreover, mtmr8 morphants show loss of arterial endothelial cell identity in dorsal aorta, which is effectively rescued by low concentration of PI3K inhibitor, and by over-expression of dnPKA mRNA or vegf mRNA. Interestingly, mtmr8 expression is up-regulated when zebrafish embryos are treated with specific inhibitor of Hedgehog pathway that abolishes arterial marker expression. Conclusion These data indicate that Mtmr8 is essential for vasculature development in zebrafish embryos, and may play a role in arterial specification through repressing PI3K activity. It is suggested that Mtmr8 should represent a novel element of the Hedgehog/PI3K/VEGF signaling cascade that controls arterial specification.

Background MTM (myotubularin myopathy) family factors are members of the growing class of dual-specificity phosphatases (DSPs) including PTEN, which can dephosphorylate the products of phosphoinositide 3-kinases (PI3K), and are negative regulators of the PI3K/Akt signaling pathways [1]. A potential function for PI3K and PTEN has been suggested in both angiogenic signaling [2,3] and various models of muscle defects [4]. And, dual-specific phosphatase-5 (Dusp-5) has been identified to play a functional role in vascular development through counteracting the function of Snrk-1, a serine threonine kinase in angioblast development [5]. However, functions of MTM family members in these processes are not clear.
In vertebrates, vascular and muscular systems are tightly connected. And, the cells in dorsal aorta and myotomal muscle are both derived from mesoderm.
Endothelial cells (ECs) forming intersegmental vessel (SE) migrate following a path initially along the somite boundary and later between notochord and somitic tissue [6]. These data suggested that a dynamic connection between somitogenesis and vasculature development might exist, and the connection might be influenced by some signaling molecules. For example, semaphorin-plexin-signaling was proved to play significant roles in both somitogenesis and vasculature formation in zebrafish [6]. Furthermore, zebrafish perlecan was demonstrated to play a central function in skeletal muscle and cardiovascular development [7].
Several signaling pathways, such as vascular endothelial growth factor (VEGF) signaling, and PI3K/Akt and Hedgehog (Hh) signaling which genetically interacts with each other [8][9][10], had been shown to be involved in vascular formation. In vitro studies suggested a role for AKT/PKB as a downstream effector of VEGF signaling [11,12]. Moreover, over-expression of the downstream effector AKT/PKB rescued VEGF receptor blockade [13]. In addition, activation of protein kinase A (PKA), a negative regulator of hedgehog pathway, effectively inhibited Akt, which demonstrated a direct role for AKT in regulating Hedgehog signaling [13,14]. Myocardial Hh activation triggered by FGF signaling, is essential for Vegf expression [15]. However, these signaling interactions form complicated network and need further explanation.
Specific genes controlling artery/vein specification have been identified in different vertebrate species [16]. In zebrafish embryo, sonic hedgehog (Shh) and Vegf, expressed in the notochord and somite respectively, are required for arterial specification [9,17]. Conversely, activation of PI3K/AKT signaling in angioblasts promotes venous specification [18]. However, the precise mechanism of signaling interactions in vascular development remains to be elucidated. Interestingly, our recent study has shown a functional cooperation of the MTM family member Mtmr8 with PI3K in muscle development in zebrafish, and revealed a possible participation of Mtmr8 in the Hedgehog (Hh) pathway [19]. To explore the function of Mtmr8 in vasculature development, here, we further observed the vascular development defects of Mtmr8 loss-of-function. Subsequently, we found that the Mtmr8, as a negative regulator of PI3K, affected the Hedgehog and Vegf pathway in the blood vessel development in vivo. The data indicate that Mtmr8 is essential for vasculature development in zebrafish.

Developmental defects of hemato-vascular progenitors in mtmr8 morphants
During zebrafish somitogenesis, mtmr8 transcript was detected in early somitic mesoderm between 1-13 somites, and later in ventral mesoderm [19], where multipotential progenitors gives rise to at least two different tissues: the hematopoietic and endothelial lineages [20]. In order to reveal further functions of Mtmr8, we analyzed the expression pattern of some known early marker genes for hemato-vascular progenitors in lateral mesoderm of mtmr8 morphants by in situ hybridization. As shown in Figure 1, the early angioblast/hematopoietic mesoderm marker genes, scl and hbae1, are expressed in anterior lateral mesoderm (ALM) and/or posterior lateral mesoderm (PLM) in Cont-MO injected embryos ( Figure 1A, C), whereas at the same embryonic stage of 12 somites, significant expression reduction is observed in mtmr8 morphants ( Figure 1B, D). And, Mtmr8 knockdown simultaneously affects the formation of somitic mesoderm, because myod expression is significantly reduced in mtmr8 morphants in comparison to Cont-MO embryos ( Figure 1E, F). As Scl was demonstrated to be very critical for zebrafish endothelial and artery development [21], we further assayed expression of fli1, a marker for endothelial cells and also for a subset of the hematopoietic cells in early development stage [22]. Compared to Cont-MO embryos ( Figure 1G), fli1 is also significantly reduced in anterior lateral mesoderm (ALM) of the mtmr8 morphants, whereas its expression defect is mild in posterior lateral mesoderm (PLM) (Figure 1H). Moreover, the expression differences of these marker genes were confirmed by quantitative RT-PCR. As shown in Figure 1I, the expression levels of scl, hbae1, myod, and fli1 are greatly reduced to 45.8%, 24.8%, 9.0%, and 51.9% of the control embryos. The data indicate that Mtmr8 knockdown affects endothelial cell formation and results in developmental defects of hemato-vascular progenitors.

Determination of abnormal vascular phenotypes in mtmr8 morphant embryos
Previous study has observed that mtmr8 is expressed predominantly in the vasculature around and after 24 hpf [19]. To determine whether Mtmr8 is required for vasculogenesis and angiogenesis, transgenic Tg(kdrl: GFP) la116 [23] embryos were injected with morpholino against mtmr8 as described previously [19], and their vasculature development was assayed. As shown in Figure 2, in comparison with normal vasculature phenotypes in Cont-MO embryos (Figure 2A), the primary intersegmental vessel (Se) [24] and dorsal longitudinal anastomotic vessel (DLAV) can sprout and form, but intensive defects are observed in mtmr8-MO morphants at 48 hpf. The vascular plexus in tail seems to be less complex and becomes narrow. Significantly, the mtmr8 morphants display a remarkable reduction in the size of dorsal aorta (DA), and the DA size reduction degree is correlated with severity of the impaired Se ( Figure 2B). The defected embryos are divided into three kinds according to relative normal Se number. The severely defected embryos with less than 5 normal Se and mild defected embryos with more than 5 normal Se in the trunk are 71.6% and 23.9% respectively, whereas normal phenotype is only 4.5% ( Figure 2J). To demonstrate the defected specificity, we co-injected 100 pg of zebrafish mtmr8 mRNA with the MO, and thereby led to an increase percentage of embryos in mild (69.7%) and normal (18.1%) phenotypes ( Figure 2C, J). And, we coinjected a mis-match mtmr8 mRNA which is unable to bind MO because of the existence of nucleotide substitutions but still encodes the same protein, with the MO, and could also rescue the morphant ( Figure 2J). Previous study observed that the expression of ptc1, a downstream target gene of the Hedgehog signaling pathway, was reduced in mtmr8 morphants, which could be rescued by co-injection with dnPKA mRNA [19]. And, the Hedgehog signaling pathway was shown to be essential for vasculture differentiation and to induce the expression of Vegf during the same process [9]. Consistently, the vascular defect of mtmr8 morphants could be rescued by co-injection of 100 pg dnPKA mRNA ( Figure  2D, J). And, scl overexpression could partially rescue the vascular defect of mtmr8 morphants, even though the defect reduction was very small ( Figure 2J). Moreover, HE stained longitudinal and transverse sections were analyzed. In comparison with that of Cont-MO embryos ( Figure 2E, G), the DA appears more narrow in the mtmr8 morphants ( Figure 2F, H, I). These data indicate that Mtmr8 is essential for the integrity of muscle and vasculature development, and that hedgehog pathway mediates these biological activities.
To clarify what stage of vasculogenesis is affected, we further observed blood circulation and vascular phenotypes in early morphant embryos. The movies at same stages of 28 hpf (Additional files 1) and 36 hpf (Additional files 2) reveal significant difference of blood cell circulation between normal control embryos and morphant embryos. In Cont-MO embryos, blood cells normally circulate in the trunk vessels, whereas almost no any blood cells flow in that of 28 hpf morphant embryos (Additional files 1), and only a few of blood cells move through arteries and veins in that of 36 hpf morphant embryos (Additional files 2). We also checked the cardiac system, however, no obvious cardiac defect was observed in the mtmr8 morphants (Additional files 3). Moreover, abnormal vasculogenesis was early observed in mtmr8 morphant embryos of Tg(kdrl:GFP) la116 transgenic zebrafish. As shown in Figure 3, when Se sprouts have produced at 28 hpf and complete vascular system has formed at 36 hpf in Cont-MO embryos ( Figure 3A, C), the corresponding vascular structures are absent or abnormal at the same stages in the mtmr8 morphant embryos ( Figure 3B, D). The above data further indicate that the vascular development defect begins from the early stage, because the down-regulation of multiple vascular marker genes for vasculogenesis has been observed in the mtmr8 morphant embryos.

Expression defects of molecular marker genes for artery/ vein and vascular endothelium in mtmr8 morphants
To further reveal molecular expression defects, we firstly used the endothelial cell marker fli1 to monitor the effect of mtmr8 knockdown on vascular development. As shown in Figure 4, in Cont-MO embryos, fli1 is abundantly expressed in axial vasculature and Se sprouts from the dorsal aorta at 26 hpf ( Figure 4A), whereas in mtmr8 morphants at the same stage, little fli1 transcript is observed in the corresponding axial vasculature and Se sprout regions ( Figure 4B). The data indicates that the endothelial-cell differentiation giving rise to the axial vasculature is defective in mtmr8 morphants. As the later-forming Se and new vessel growth were believed to form respectively by sprouting and from preexisting vessels [25], we next checked the formation of primary vasculature. The expression of ephrinB2a, an artery marker, was detected in presumptive DA region of Cont-MO embryos ( Figure 4C). However, its expression was severely reduced in the mtmr8 morphants ( Figure 4D), suggesting that the artery development should be impaired. The expression of vein marker flt4 was observed in PCV region of Cont-MO embryos ( Figure  4E), whereas the flt4-positive PCV regions were expanded, and its expression was partially reduced in the mtmr8 morphants ( Figure 4F). To confirm the role of Mtmr8 in artery formation, we further examined the expression changes of two Notch signaling markers notch5 and grl/hey2. As shown in Figure 4G-J, significant expression decrease of the two arterial marker genes was observed in the mtmr8 morphants.
Moreover, the effects of mtmr8 knockdown on vasculogenesis were judged by endogenous alkaline phosphatase activity. In Cont-MO embryos ( Figure 4K, L), the endogenous alkaline phosphatase activity labeled the major cerebral vessels and the subintestinal vessels (SIVs). However, all Mtmr8 morphants displayed an obvious reduction in the head signal and a total absence of SIV labeling ( Figure 4M, N). The obvious effects of mtmr8 knockdown on endothelial cell alkaline phosphatase activity further suggest that the vascular changes should be specific to endothelial cells, and the endothelial cell marker gene fli1 should be primary target for the mtmr8 functions.
Mtmr8 mediates PI3K/Akt signaling for artery specification In our previous study, mtmr8 knockdown was revealed to activate the PI3K/Akt pathway [19]. To determine whether the cooperation of Mtmr8 with PI3K plays a role during vasculogenesis, we firstly assayed the affect of PI3K inhibitor LY294002 on mtmr8 expression in zebrafish embryos. As shown in Figure 5, compared with control embryos ( Figure 5A, E), lower concentration of LY294002 (10 and 25 μM) has no obvious effect on mtmr8 expression in wild-type embryos ( Figure 5B, C, F, G), but significant increase expression of mtmr8 appears in the trunk vasculature when 50 μM LY294002 is used ( Figure 5D, H). Figure 5I shows the quantitative data from Q-RT-PCR.
In mtmr8 morphants, the expression of artery marker gene ephrinB2a was severely reduced ( Figure 4D) in the trunk region, but the reduced levels were variable, and might be classified into three kinds: normal, similar to consistent expression in wild-type ( Figure 5J), mild defect, patchy expression in some cells along the DA ( Figure 5K), severe defect, almost no any ephrinB2a expression in the trunk DA ( Figure 5L). In previous studies [19], when treated with 10 μM LY294002, about half of the mtmr8 morphants could be further deteriorated in muscle development, and the others were mildly or not obviously affected (Data not shown). However, 4 μM LY294002 had no obvious effect on the muscle phenotype of the morphants. Interestingly, when 4 μM LY294002 was used to treat the mtmr8 morphants, a significantly increasing proportion of normal embryos was found in the morphants ( Figure 5M). Subsequently, we checked phosphorylation status of another early arterial progenitor marker ERK (p42/44 MAP kinase) [18] in the mtmr8 morphants by Western blot analysis, and found that loss of Mtmr8 function significantly reduces ERK phosphorylation ( Figure 5N). The data indicate that Mtmr8 negatively mediates PI3K/Akt signaling for artery specification.

Mtmr8 is essential for normal vasculature development through regulating Hedgehog and Vegf signaling pathways
Hh signaling is multi-functional, and its role is temporal-spatially determined during embryogenesis [26]. Shh from the notochord promotes Vegf expression by the adjacent somite, which promotes expression of the artery-specific EphrinB2 in the dorsal aorta [27]. To determine the position of Mtmr8 in this regulatory   cascade, we tested its expression in embryos when these pathways were inhibited. As shown in Figure 6A and 6B, in 26 hpf wild-type embryos, ephrinB2a is absent in the dorsal aorta when exposed to 50 μM cyclopamine, and vegf expression is lost within the somites but persisted within the hypochord (Data not shown, the result same to Lawson et al [9]), which means that inhibition of Hedgehog signaling can partially disrupt Vegf expression. Then, we investigated whether cyclopamine could also inhibit the mtmr8 expression. As shown in Figure  6C-F, mtmr8 shows stronger expression in the posterior cardinal vein than in the dorsal aorta of wild-type control embryos ( Figure 6C-D), and its expression was obviously upregulated in the cyclopamine-treated embryos ( Figure 6E-F). Figure 6K shows the gradient up-regulation of mtmr8 under different doses of cyclopamine obtained from Q-RT-PCR.
Vascular function is tightly regulated by Vegf signaling pathway [28]. To assess whether Mtmr8 influenced the developing vasculature through Vegf pathway, we treated 10 hpf embryos with a Vegf receptor tyrosine kinase inhibitor, SU5416. In comparison with control embryos (Figure 6G), the embryos incubated with 5 μM of SU5416 failed to show vascular expression of mtmr8 in the trunk at 30 hpf ( Figure 6H), and the reduced transcript levels were detected under different doses of SU5416 by Q-RT-PCR ( Figure 6K), suggesting that Mtmr8 might act downstream of Vegf to determine arterial cell fate. Moreover, we checked the expression of vegf165 in Mtmr8 morphants and control embryos, and found a significant down-regulation of vegf165 expression within the somites at 26 hpf ( Figure 6I, J), and the quantitative down-regulation change was confirmed by Q-RT-PCR ( Figure 6L).
In zebrafish, over-expression of either vegf121 or vegf165 could rescue arterial differentiation blocked by a deficiency of Shh signaling [9]. Direct activation of PKA, an inhibitor of Hedgehog signaling, induces endothelial cell apoptosis and inhibits angiogenesis in vivo, and suppressing the PKA activity by expression of dnPKA, promotes endothelial cell survival and migration during angiogenesis [29,30]. Mtmr8 acts downstream of Vegf to determine arterial cell fate, and its knockdown can impair the Hedgehog signaling. On this basis, we further assessed whether exogenous Vegf and Hedgehog signaling were sufficient to rescue ephrinB2a expression in the absence of mtmr8 activity. As shown in Figure 6M, when injected with 100 pg dnPKA or vegf121+vegf 165 mRNA, the ephrinB2a expression in mtmr8 morphants can be rescued as shown by reducing percentage of embryos in the severe or mild defect group, which can not be rescued by control GFP mRNA (Data not shown). Moreover, we tested whether Vegf overexpression would lead to increase of mtmr8 expression. As shown in Figure 7, when 100 pg vegf121+vegf 165 mRNA was injected, ptc1 expression is not affected (Figure 7A, B), but the expression of mtmr8 was up-regulated ( Figure 7C, D) (n = 17/28) in the injected embryos. We therefore conclude that Mtmr8 expression in vasculature is dependent on Vegf, and Mtmr8 is essential for normal vasculature development through regulating Hedgehog and Vegf signaling pathways.

Discussion
Previous studies have showed that mutations in most myotubularin family genes are causative for human neuromuscular disorders [31]. And, knockdown of zebrafish myotubularin and mtmr8 also impairs the embryo muscle development [19,32]. Moreover, Mtmr8 has been shown to negatively regulate the PI3K/AKT pathway [19]. In the current study, we have further found that mtmr8 knockdown leads to activation of PI3K/Akt signaling, which impairs arterial endothelial marker gene expression, and results in endothelial cell impairment and vasculogenesis defects. Significantly, mtmr8 expression has been detected in somitic and lateral mesoderm, where the angioblasts are committed to endothelial lineage differentiation and restricted to arterial or venous fate [19,33]. And, significant expression reduction of endothelial progenitor marker genes in mtmr8 morphants indicates that the impaired arterial differentiation should be the consequence of defects in endothelial cell formation. These results suggest that mtmr8 should be essential for the endothelial cell differentiation and vasculature development through repressing the activity of PI3K in zebrafish.
Hedgehog signaling from notochord has been demonstrated to promote somitic expression of Vegf, and thereby to promote expression of the artery-specific EphrinB2 in dorsal aorta [27]. Mtmr8 expression has been detected in somitic and lateral mesoderm, where angioblasts are restricted to arterial or venous fate, suggesting that the impaired arterial differentiation in mtmr8 morphants might be the consequence of defects in Hedgehog/Vegf signaling pathways. And, the expression of mtmr8 is significantly repressed in the embryos treated with inhibitor of Vegf signaling pathway ( Figure  5G-H), suggesting that Mtmr8 might act downstream of Vegf to determine arterial cell fate. The concomitant down-regulation of ephrinB2a, ptc1 and vegf in Mtmr8 morphants raises the possibility that Mtmr8 acts upstream of Hedgehog pathway. We hypothesize that Mtmr8 regulates artery differentiation through two parallel ways (Figure 8). In one way, Mtmr8 may act downstream of Vegf. In other way, Mtmr8, Hedgehog and Vegf may control artery specification through a regulatory loop (Figure 8). In normal physiological condition, Mtmr8 is a downstream factor of Vegf signaling pathway, and is partially regulated by the Hedgehog signaling pathway that is affected by the expression of Mtmr8. Additionally, the down-expression of the hematovascular progenitor marker genes hbae1 and fli1 in mtmr8 morphants (Figure 1) suggests that mtmr8 might also regulate hematovascular progenitor specification and differentiation (Figure 8).

Conclusion
In conclusion, our findings indicate that Mtmr8 functions as a key regulator of somitic muscle development and vasculature development in zebrafish embryos. The signaling pathways that function during vascular development are highly conserved. Further analysis for Mtmr8 function will be essential to understand the complex signaling hierarchy for the differentiation of endothelial cells and the development of vasculature in vertebrates.

Zebrafish husbandry
Danio rerio wild-type (WT) AB strain, and Tg(kdrl:GFP) la116 transgenic zebrafish (previous Tg(flk1:GFP) la116 ) [23] were used. Zebrafish were maintained and staged as described [34]. When necessary, embryos were anesthetized with 0.003% tricaine (Sigma). All of the experimental researches on zebrafish were performed with the approval of the animal ethics committee in the Institute of Hydrobiology, Chinese Academy of Sciences.

Antisense morpholino and mRNA microinjection
The sequences and doses of the injected morpholinos were described previously [19]. mtmr8, dnPKA, scl, vegf121 and vegf165 cDNAs were subcloned into PCS2 + plasmid for in vitro synthesis of capped mRNA, using SP6 RNA polymerase (Ambion), following the manufacturer's instructions. A mis-matched mtmr8 mRNA (mis-mtmr8) was constructed by using mis-match forward prime 5'-ATGGAGCACATCATAACGCCCAAAGTCG-3' (underlines indicate the changed nucleotide). All mRNA were re-suspended in water and co-injected at a concentration of 50-100 ng/μL.

Whole-mount in situ hybridization, endogenous alkaline phosphatase-based vessel staining and histological analysis
Embryos at different stages were collected and fixed as described [35]. Purified plasmids were linearized by selected restriction enzymes and used as templates for in vitro transcription using T7 or Sp6 RNA polymerase to generate DIG-labeled (Roche) sense and anti-sense probes. Whole mount in situ hybridization was performed as described previously [36,37].
At 72 hpf, embryos were fixed in 4% paraformaldehyde and dehydrated in methanol. Later, they were rehydrated to PBST and then used for staining, as described by Serbedzija et al [38]. Histological analysis was performed according to methods described by Lawson et al [17].

Real-time quantitative PCR (Q-RT-PCR) and Western blot detection
Real-time quantitative PCR was performed by DNA Engine Chromo 4 Real-Time System (MJ Research) with SYBR Green I Dye according to a previous report [39]. The ratio of these genes to b-actin in control embryos was set to 1 (100%), and the expression of all morpholino-injected or drug-treated embryos were normalized relative to this value. All samples were analyzed in triplicates and the results were expressed relative to the expression of b-actin using the 2 (-delta delta C(T)) method [40]. Data are presented as mean ± SD of three separate experiments.  Western blot detection for the ratio between phospho-p44/42 MAP kinase (Thr202/Tyr204) and p44/42 MAP kinase in developing zebrafish larvae was performed essentially as described [18]. The membranes were washed and incubated with Alkaline Phosphatase (AP)conjugated secondary antibodies. Images of blots were captured with a scanner, and quantitative densitometric analysis was performed using Scion Image.

Drug treatments
PI3K inhibitor, LY294002 (Promega) was dissolved in DMSO at stock concentration of 50 mM. For experiments, embryos were incubated in embryo media containing 0-50 μM of the drug from 10 hpf to 24 hpf. Control embryos were treated with the equivalent amount of DMSO solution (1%≤).
Cyclopamine and SU5416 were obtained from Calbiochem and were used at a concentration of 0-50 μm and 0-5 μM respectively. Zebrafish embryos were soaked in 6-well plates and were treated with the drugs from 10 hpf, until they were processed for in situ experiments. Control embryos were treated in embryo medium containing equivalent amount of DMSO solution (1% < ).

Statistical analysis
All the experiments were performed in triplicates (about 30 embryos in each experiment group). Data were presented as mean ± SD. The data in morphant embryos compared to the control (set to 1) were assessed by one-way analysis of variance (ANOVA), followed by the Tukey's post hoc test for multiple comparisons. A probability (*P) of <0.05 was considered statistically significant.