|Home | About | Journals | Submit | Contact Us | Français|
The specification of left-right asymmetry is an evolutionarily conserved developmental process in vertebrates. The interplay between two TGFβ ligands, Derrière/GDF1 and Xnr1/Nodal, together with inhibitors such as Lefty and Coco/Cerl2, have been shown to provide the signals that lead to the establishment of laterality. However, molecular events leading to and following these signals remain mostly unknown. We find that APOBEC2, a member of the cytidine deaminase family of DNA/RNA editing enzymes, is induced by TGFβ signaling, and that its activity is necessary to specify the left-right axis in Xenopus and zebrafish embryos. Surprisingly, we find that APOBEC2 selectively inhibits Derrière, but not Xnr1, signaling. The inhibitory effect is conserved, as APOBEC2 blocks TGFβ signaling, and promotes muscle differentiation, in a mammalian myoblastic cell line. This demonstrates for the first time that a putative RNA/DNA editing enzyme regulates TGFβ signaling, and plays a major role in development.
In vertebrates, left-right asymmetry is manifested anatomically by the selective folding and positioning of a subset of internal organs, like the heart and the gut. The emergence of laterality is also obvious during early development through genes that are expressed specifically in the left lateral plate mesoderm, such as the secreted factors Nodal and Lefty, both members of the TGFβ signaling molecules, and the transcription factor Pitx2. However, the key symmetry breaking event occurs before the left side specific expression of Nodal, Lefty and Pitx2, and involves signals emanating from the posterior mesoderm portion of the late organizer tissue in frogs at the end of gastrulation, and similar organizing centers in other vertebrates (Kupffer's vesicle in fish and the node in birds and mammals; Essner et al., 2002; Schweickert et al., 2007; Shook et al., 2004). These early signals from the posterior mesoderm are subsequently transferred to the lateral plate mesoderm. The key event in breaking symmetry is mediated by unidirectional flow mediated by cilia (Essner et al., 2005; Hirokawa et al., 2006; Schweickert et al., 2007). In the chick it involves cell migration (Gros et al., 2009), and in frogs long range diffusion of the ligand has been invoked (Tanaka et al., 2007). Regardless of the mechanism, however, initial signals form the first organizing center (posterior mesoderm) in turn generate a second signaling center (left lateral plate mesoderm) to ultimately convey right versus left positional information, guiding the appropriate folding and positioning of the internal organs (Bisgrove and Yost, 2006; Hirokawa et al., 2006; Lee and Anderson, 2008; Levin, 2005; Mercola, 2003; Raya and Belmonte, 2006; Shiratori and Hamada, 2006; Speder et al., 2007; Whitman, 2001).
The early signals emanating from the first organizing center in posterior mesoderm result from an interplay between TGFβ ligands and inhibitors. Two ligands: Xenopus-nodal-related-1 (Xnr1)/nodal and Derrière/GDF1, together with their inhibitor Coco/Dante/Cerl2, mediate signaling (Hashimoto et al., 2004; Marques et al., 2004; Pearce et al., 1999; Vonica and Brivanlou, 2007). Interestingly, Derrière and Xnr1 ligands exist as both homo- and heterodimer, share the same set of inhibitors, and are part of a positive feed back loop inducing each other's expression (Eimon and Harland, 2002; Tanaka et al., 2007). Moreover, in addition to Xnr1, both Derrière and Coco activity were required in the posterior mesoderm for correct left-right axis determination (Vonica and Brivanlou, 2007). Derrière controls the levels of Xnr1 expression and Coco acts as an inhibitor of both ligands. Ultimately, the balance between extracellular inhibitors and activators establishes differential thresholds of signaling on each side of the embryo. To trigger nodal-type TGFβ signaling, ligands bind and activate receptor complexes that consist of two type I receptors, ALK4 and ALK7, and two type II receptors, ActRIIA and ActRIIB. The receptor complex also requires participation of co-receptors of either EGF-CFC type such as XCR/Cripto, or TMMEF1 (Cheng et al., 2003; Dorey and Hill, 2006; Gritsman et al., 1999; Onuma et al., 2005; Yan et al., 2002; Yeo and Whitman, 2001). From the three XCR co-receptors that are expressed during early Xenopus embryogenesis, only XCR2 is present at the time when left-right asymmetry is specified. Activated receptors in turn activate the signal transducers Smad2/3. Consistent with their involvement, intracellular factors that interfere with ALK4 and Smad2/3 interaction, such as Ttrap, have shown to influence the right-left axis (Esguerra et al., 2007). Smad2/3 associate with Smad4, translocate to the nucleus, and regulate transcription of specific genes in cooperation with other transcription factors. In left-right axis determination, different thresholds of Xnr1 and Derrière signaling on each side lead to different transcriptional output segregating left cell fates from right by selective expression of genes such as Pitx2 on the left lateral plate mesoderm, setting up the second organizing center (Bisgrove and Yost, 2001; Brand, 2003; Levin, 2005; Wright, 2001).
While the organizing centers and the molecular nature of the signals involved in the establishment of laterality have been characterized, events occurring downstream of signaling remain poorly understood. Performing a global analysis of genes that are regulated by Derriere in the posterior mesoderm, we find unexpectedly that APOBEC2 (Anant et al., 2001; Liao et al., 1999), an evolutionarily conserved member of the cytidine deaminases family of proteins, is induced by TGFβ signaling, and has a direct involvement in left-right specification. We show that APOBEC2 activity is necessary for left-right specification by selectively inhibiting Derrière but not Xnr1 signaling in Xenopus. In addition, APOBEC2 is required as a TGFβ inhibitor for the differentiation of a mammalian myoblastic cell line. These experiments highlight the role of an unexpected player in the context of TGFβ signaling pathway in general, and a previously unknown biological function for APOBEC2 in the context of left-right axis determination during vertebrate embryogenesis.
The activity of the TGFβ ligand Derrière is necessary for the proper development of laterality during early vertebrate embryogenesis (Vonica and Brivanlou, 2007). To identify the molecular network underlying right-left specification downstream of Derrière signaling, we performed microarray analysis comparing mRNA expression between wild type and Derrière-depleted embryonic explants (Figure 1A). This analysis led to the identification of a number of Derrière-regulated genes (Supplementary Data at http://www.ncbi.nlm.nih.gov/projects/geo, accession numbers GSM349689, GSM349690, GSE13882; and Table S1). Surprisingly, we find that Xenopus laevis APOBEC2 (xA2, Genbank accession #BC074467, Supplementary Figure 1A), the ortholog of the mammalian APOBEC2/ARCD-1 DNA/RNA editing enzyme (Anant et al., 2001), is down-regulated upon Derrière depletion. Because editing enzymes have not been previously associated with TGFβ signaling, we followed on the molecular and embryonic characterization of xA2 activity. Decreased xA2 expression upon Derrière depletion implied that signaling by this ligand induces xA2. RT-PCR analysis confirmed that xA2 was moderately decreased in Derrière-depleted embryos at both stage 12 and 18 (Supplementary Figure 1B, C), confirming the microarray result, and strongly decreased in embryos overexpressing a dominant-negative form of the nodal-specific type 1 receptor ALK4 (DN ALK4(KR), Figure 1B, Yeo and Whitman, 2001). In addition, overexpression of Derrière and Xnr1 induced and maintained xA2 expression in stage 10 and 18 ectodermal explants (Figure 1B). Together, these experiments indicate that xA2 is a target of TGFβ signaling.
To begin the characterization of xA2 function, we determined its spatio-temporal pattern of expression during embryogenesis. RT-PCR analysis revealed that expression of xA2 begins weakly at the onset of gastrulation, and then increases at neurulation (Figure 1C). The weak expression in gastrulation occurs mostly in dorsal marginal cells, including the late organizer (Figure 1D, E). xA2 expression is confined to the paraxial somitic mesoderm during neurulation (Figure 1G, and H) as well as the heart and cement gland at tailbud stages (Figure 1I). Expression pattern of zebrafish A2 (zA2) displayed a similar pattern with higher levels of zA2 RNA in the shield (the equivalent of the organizer, Figure 1N, O), and subsequent expression in the somites and heart, as reported (Figure 1Q, and R; Mikl et al., 2005; Thisse et al., 2004). In addition, and consistent with xA2 being a target for TGFβ signaling, we find that the expression pattern of xA2 overlaps with the expression domains of Derrière during gastrula and neurula stages, when the specification of laterality occurs (Figure 1J-M; Sun et al., 1999).
To investigate a possible role of xA2 in left-right specification during early development, we performed loss-of-function experiments using translation-blocking morpholinos (xA2 MO1 and 2, Figure 2A and Supplementary Figure 2A). Both MOs had the same phenotype, and in subsequent experiments we used xA2 MO1 (henceforth named xA2 MO). xA2 MO, along with GFP RNA used as lineage tracer, was microinjected to target either the right or left paraxial mesoderm (PM, see Materials and Methods, and Vonica and Brivanlou, 2007). During normal development, the heart folds to the right, and the gut loops counterclockwise (Figure 2C). Random deviation from this pattern is called heterotaxia. Injected embryos were examined at stage 48 for heart rotation and intestinal looping as morphological readout for laterality. Heterotaxia was detected when xA2 protein was depleted on the left, but not the right side (Figure 2D, E, and Table 1). No effect was seen on axial development for the amount of MO used in this study (10 ng, Supplementary Figure 2C, D). Consistently, examination of molecular markers specific for the left side at stage 23-24 revealed that depletion of xA2 on the left blocked expression of Xnr1 and Lefty in the left LPM (Figure 2H-Q and Table 2). As left LPM markers are induced by early signals emanating during neurula from the posterior paraxial mesoderm (Ohi and Wright, 2007), we examined the effect of xA2 MO on the posterior PM. We find that reduction of xA2 activity leads to a decrease in Xnr1 and an anterior expansion of Derrière expression, as seen in in situ hybridization of posterior poles of stage 18 embryos (Figure 3C-H, internal views, anterior is up, left side of the embryo is on the right side of each picture), and confirmed in RT-PCR (Supplementary Figure 3J). Thus, disturbance in the balance of Derrière and Xnr1 signaling early in the posterior PM is the probable cause of the laterality phenotype. Finally, a left-right phenotype was observed in zebrafish embryos when MOs targeting zA2 were used (zA2 MO, Figure 2F,G, Supplementary Figure 2B, and Table 1). A slight developmental delay of 1 - 2 hours was noted in zA2MO-injected embryos during somitogenesis stages, whereas in situ hybridization with a cmlc2 probe revealed a lack of heart looping in zA2MO-injected embryos even 5 – 6 hours after obvious looping in uninjected embryos. Further, looping defects were also observed after 48 hpf (data not shown), indicating that the role of xA2 in the context of right-left specification is evolutionarily conserved.
To address the specificity of the xA2 MO phenotype and to position the input of xA2 in the pathway, we performed rescue experiments. xA2 MO was co-injected with DNA encoding xA2 lacking the MO binding site, or mRNA for an inducible Smad2 (GRVP16hSmad2Δ3; (Vonica and Brivanlou, 2007). GRVP16hSmad2Δ3 protein is retained in the cytoplasm until addition of dexamethasone (Dex) to the embryos. Coinjection of xA2 MO with the mutant xA2 DNA rescued the heterotaxia phenotype (Table 1), providing evidence for specificity. Additionally, coinjection of xA2 MO with GRVP16hSmad2Δ3 RNA, followed by Dex induction at neurula stages, rescued both the left-right asymmetry, and expression of molecular markers in the left LPM (Figure 2P; Table 1). The fact that induction of Smad2 was sufficient to rescue the phenotype when activated at neurula stages suggests that the rescue is not due to an indirect effect on TGFβ signaling. This experiment establishes that xA2 activity is required upstream of Smad2, in the context of left-right specification in vivo. Taken together, this evidence demonstrates that: (i) xA2 activity is required on the left side of the embryos for proper establishment of laterality and expression of left-side-specific markers; (ii) xA2 MO effect on LPM gene expression is due to interference with early posterior PM signals; (iii) the activity of xA2 in the context of left -right specification appears to be conserved between teleosts and amphibians.
To address the molecular mechanism of the effect of xA2 depletion, we tried to rescue the left-right phenotype with TGFβ ligands. Coinjection of xA2 MO and Xnr1 RNA directed at the left PM failed to rescue the xA2 depletion phenotype, including the left side expression of Lefty (Table 1, ,2;2; Supplementary Figure 3 B, C). Instead, embryos had short trunks and tails, and blastopore closure defects (Supplementary Figure 3A, and data not shown). However, when Xnr1 RNA was injected separately from paraxial xA2 MO, and directed at the left LPM, it did rescue the xA2 depletion phenotype (Table 2). These results indicate that the defect in TGFβ signaling causing the phenotype of xA2 depletion could be more complex than the absence of nodal/Xnr1 expression in posterior paraxial mesoderm, and that the lateral plate mesoderm retained the ability to respond to local nodal signals. To explain this unexpected result, we compared gene expression in posterior poles of neurula stage embryos with injections directed at the PM with xA2 MO alone, or coinjections of MO and Xnr1 RNA (Figure 3I, and Supplementary Figure 3 D-I). derrière and MyoD, a known target of TGFβ signaling in paraxial mesoderm (reviewed in (Dale, 1997; Green, 2002; Schier and Talbot, 2005; Smith, 1995), but not Brachyury, which is expressed in the notochord (Smith et al., 1991), were increased in xA2-depleted embryos, and had even higher levels when Xnr1 RNA and xA2 MO were associated. This apparent synergy between xA2 MO and an overexpressed TGFβ ligand suggests that depletion of xA2 could increase TGFβ signaling in posterior mesoderm.
The cells that will form the posterior PM are exposed to sequential TGFβ signals during Xenopus development, starting with the maternal Vg1, followed by multiple Xnr ligands and Derrière during blastula, and again Derrière during gastrula. To understand which TGFβ signal is affected by xA2 depletion, and its timing, we again used the dexamethasone-inducible Smad2 construct, GRVP16hSmad2Δ3, in coinjections with xA2 MO directed at the PM. Activation of this construct in A2-depleted embryos at neurula stage restored the left-right axis and expression of Xnr1 in the left LPM, indicating that the inducible Smad2 restored the posterior TGFβ signal decreased by the reduction in Xnr1 expression. In contrast, activation in gastrula did not rescue the left-right phenotype and produced the same trunk defects seen with Xnr1 RNA coinjection (Table 1 and data not shown). These results indicate that the phenotypic synergy between xA2 depletion and TGFβ overexpression described above probably results from molecular interactions that occur no later than the gastrula stage, when xA2 and derrière expressions overlap (Figure 1 F, K). Together, the synergy between xA2 depletion and TGFβ activity, the timing of the effect, and expression data suggest that xA2 depletion increases Derrière activity at gastrula stage. This leads to defects in posterior patterning at neurula stage, including the absence of Xnr1 expression and subsequent left-right specification defects.
The hypothesis that an increase in Derrière activity is the cause of the xA2 phenotype was tested directly by partially depleting the Derrière protein. An amount of Derrière MO that had no effect by itself rescued the left-right phenotype of xA2 depletion in coinjections targeted at the left PM (Table 1). In addition, we found that overexpression of the inducible Smad2 construct alone caused a left-right phenotype when activated in gastrula (stage 11), but not neurula (stage 16), (Table 1), confirming that increased TGFβ signaling in cells fated to become left PM can by itself produce left-right defects.
To begin the molecular and biochemical characterization of xA2 activity, we challenged the mesoderm-inducing activity of both Derrière and Xnr1 in animal cap explants. These explants, which give rise solely to epidermis when cultured alone (Figure 4A), can be induced by TGFβ ligands, including Derrière and Xnr1, to form mesoderm (reviewed in (Harland and Gerhart, 1997) and to elongate. We find that expression of xA2 alone did not change the morphology of animal caps (Figure 4B) or induce any germ layer-specific marker (not shown), suggesting that while xA2 is a target of TGFβ signaling, it has no ability to induce new cell types on its own. Interestingly, however, xA2 RNA overexpression selectively inhibited both the elongation and the induction of mesodermal markers induced by Derrière, but not by Xnr1 (Figures 4D, F, G, H). In agreement with the inhibitory effect of xA2 on Derrière, coinjection of xA2 MO with a low amount of Derrière RNA increased the expression of mesodermal fate genes (Figure 4I), and overexpression of xA2 RNA in the left PM produced a left-right phenotype by blocking Xnr1 expression in the left LPM (Table 2). Transcriptional assays confirmed these results independently. Derrière, but not Xnr1 induction of the Smad2 responsive promoter, ARE-lux (Liu et al., 1997), was strongly reduced when xA2 was co-expressed (Figure 4J). The transcriptional effect was specific for the TGFβ pathway, as xA2 did not inhibit activation of the siamois reporter gene (Brannon et al., 1997) by Wnt8 (Supplementary Figure 3L). We also generated a triple mutant of the putative enzymatic site (E102Q, C130S, and C132S, Supplementary Figure 1A). Although the mutant was as stably expressed as the wild type (Supplementary Figure 3K), it did not change the Derrière-mediated transcriptional activation of ARE-lux in a statistically significant way (Figure 4K), failed to rescue the XA2 MO left-right phenotype when expressed at low levels, and did not cause a left-right phenotype when expressed alone at high level on the left side (Table 3). Taken together, these experiments suggest that xA2 needs its deaminase domain for the left-right function and for inhibiting Derrière signaling upstream of transcriptional response. Examination of the state of Smad2 activation revealed that C-terminal phosphorylation of endogenous Smad2 induced by derrière, but not Xnr1, is strongly inhibited by co-expression of xA2 (Figure 4L). Thus, inhibition of Smad2-dependent mesoderm induction, transcriptional activation, and C-terminal phosphorylation, suggests that xA2 regulates the TGFβ pathway upstream of Smad2.
We next asked if the TGFβ inhibitory activity of APOBEC2 was conserved in mammals by testing its activity in the myoblastic C2C12 cell line. These cells can differentiate into myotube when shifted from high to low serum (McMahon et al., 1994). Smad2-mediated TGFβ signaling, however, inhibits this differentiation, and conversely, blocking TGFβ signaling accelerates formation of myotubes (Joulia et al., 2003; Massague et al., 1986; Olson et al., 1986; Rios et al., 2001; Rios et al., 2002). Interestingly, we find that while mouse APOBEC2 (mA2) is not expressed in undifferentiated C2C12 cells, its expression is induced during myotube differentiation (Figure 5A). As this pattern of expression suggests a potential role for A2 during myogenesis, we analyzed the consequences of reducing mA2 expression levels during C2C12 differentiation. Cells transfected with a specific siRNA targeting mA2 (siA2, Figure 5B, Supplementary Figure 2E) decreased expression of early (Myogenin) and late (Skeletal-α-Actin) muscle differentiation markers indicating that mA2 is required for myogenesis.
To address the role of mA2 in the process, plasmids encoding CMV-mA2 were co-transfected together with the TGFβ-responding plasmid p3TP-lux (Wrana et al., 1992), used as reporter for transcriptional activation in C2C12 cells. Consistent with our observations in Xenopus, mA2 inhibited TGFβ1-induced transcription of the p3TP-lux reporter in C2C12 cells (Figure 5C). Interestingly, mA2 also inhibited the endogenous levels of TGFβ signaling, suggesting a potential direct role in myoblast differentiation in vivo. The activity of mA2 was specific, as the closely related family member AID did not repress TGFβ-mediated induction in the same experimental conditions (Figure 5D). The TGFβ inhibitory activity of APOBEC2 thus seems to be conserved in mammals.
This study shows that expression of a putative RNA/DNA editing enzyme, APOBEC2, is regulated by TGFβ signaling, and in turn regulates the pathway by ligand-specific inhibition, generating a negative feedback loop. This work also assigns a previously unrecognized biological function for APOBEC2 in the context of axis determination during early embryogenesis. The contribution of APOBEC2 to the TGFβ pathway, its mechanism of action, and its role during left-right specification are all unexpected, and discussed below.
We show that APOBEC2 (A2) regulates TGFβ signaling. A2 is a member of the cytidine deaminase family of DNA/RNA editing enzymes that emerged at the beginning of vertebrate radiation (Conticello, 2008). These enzymes can mutate C to U in DNA or RNA (Conticello et al., 2007; Navaratnam and Sarwar, 2006; Smith, 2008). The other members of the family are APOBEC1, AID and APOBEC3s (with several variants in primates). AID and APOBEC2 are the only members conserved among chordates, while APOBEC1 and APOBEC3s are mammalian-specific (Conticello et al., 2007). APOBEC1 edits apolipoprotein B mRNA in a tissue specific manner by introducing a premature stop codon leading to a shorter protein. AID has been shown to edit the immunoglobulin locus directly on genomic DNA to generate antibody diversification. Different variants of APOBEC3 protein edit retroelements and retroviruses, such as HIV-1, to provide retroviral immunity. The fact that TGFβ signaling regulates expression of xA2 suggests that, just as in the context of immuno-diversification or retroviral editing, non-autonomous signals control editing activity in early embryogenesis. However, while editing functions described for previous members of the family affect cellular activities without changing their fate, A2 affects cell fate decisions with implications in axis development. Selective inhibition of Derrière, but not Xnr1, by A2 also provides a hint to how signaling by similar TGFβ ligands using the same receptor complex and signal transducers (Smad2 and 3 for nodal and derrière) can be segregated by autonomous factors. A2 provides a negative feedback loop only for derrière, but not for Xnr1, presumably when presented as homodimers. The fate of signaling by heterodimers of the two ligands has not been addressed by this study.
Interestingly, it was recently shown that mA2 is required for normal muscle development. mA2-deficient mice exhibit slightly reduced muscle mass, increased slow:fast fiber ratio, and developed myopathy at 8-10 months of age (Sato et al., 2009). We show here that A2 is required for terminal differentiation of C2C12 myoblasts in vitro and that it can inhibit TGFβ signaling. It is tempting to speculate that the defects observed in vivo in mutant mice might be mediated by increased TGFβ signaling due to GDF-type ligands expressed in muscle cells, such as Myostatin/GDF8 and GDF11 (Artaza et al., 2002; Lee et al., 2005), in a A2-deficient background. This is also supported by the decrease in slow:fast fiber ratio seen in myostatin-deficient mice (Amthor et al., 2007; Girgenrath et al., 2005), indicating that APOBEC2 and myostatin have opposite effects on muscle in vivo.
Regarding the mechanism of the A2 effect in the context of left-right specification, it can act either as a DNA/RNA editing enzyme or by an independent, currently unknown, mechanism. While our experiments do not address this directly, the fact that conserved amino acids in the enzymatic domain required for the function of well established cytidine deaminases are also required for A2 biological activity suggests that A2 could at an editing level. Alternatively, A2 could, directly or indirectly, affect the translation or stability of proteins involved in TGFβ signaling.
If A2 acts as an editing enzyme, does it edit DNA, RNA, or both? The answer to this question is currently unknown. At the genomic level, APOBEC2 has been suggested to provide global changes in the methylation state of DNA in zebrafish embryos (Rai et al., 2008). While our study does not address this directly, a number of observations suggest that amphibian A2 is not involved in global changes of genomic DNA. First, the ligand-specific nature of the TGFβ inhibitory effect, directed at Derrière but not nodal, is not consistent with global methylation changes. Second, we note that overexpression of zAPOBEC2a (one of the two variants in the fish), unlike zAID and zAPOBEC2b, had no effect on genomic demethylation (Rai et al., 2008). This also suggests that the amphibian A2 is more similar to APOBEC2a, rather than the APOBEC2b ortholog in the fish. Finally, while in the fish both AID and APOBEC2 have been implicated in demethylation, we find that only A2 and not AID displays inhibitory activity in C2C12 cells, thus arguing against demethylation being causal to the activity.
We show that A2 activity is necessary for the specification of the left-right axis downstream of signaling. In agreement with this role, xA2 expression overlaps that of derrière in cells of the posterior mesoderm. Evidence for the appropriate timing of A2 activity in the context of left-right specification is provided by our rescue experiments where the laterality defects generated by xA2 depletion were rescued by induction of Smad2 activity at early neurula stages. This is the same developmental stage when the posterior mesoderm acts as a signaling center to pattern the left lateral late mesoderm, through Derrière and Xnr1 signaling. We therefore propose that selective inhibition of Derrière signaling in posterior mesoderm by APOBEC2 is required for the process underlying specification of laterality in the left lateral plate mesoderm. While this explains why A2 is necessary for proper left-right axis specification, and links the phenotypic outcome to molecular events, it also raises interesting questions. First, neither MO-mediated reduction of A2 function in zebrafish (Rai et al., 2008), nor knockout of APOBEC2 function by insertional disruption in the mouse (Mikl et al., 2005) have been reported to affect laterality. In the mouse, as stated above, other cytidine deaminases, (for example APOBEC3s), are present, that can theoretically compensate redundantly for lack of APOBEC2. In the zebrafish study, depletion was achieved with the same morpholino oligonucleotide as the one used in our work, and a possible explanation would be that a discreet phenotype like heart orientation could have escaped observation. Second, the mechanism of the selective effect of A2 on Derrière/GDF, as opposed to nodal signaling, remains unexplained. It remains to be seen if other differences reported on the molecular pathways activated by Nodal versus Derrière (Dorey and Hill, 2006; Howell et al., 2002) are connected to the role of A2.
Regardless of the outcome, A2, and perhaps other members of the DNA/RNA editing family of enzymes, provides yet another unexpected level of regulation for the TGFβ signaling pathway.
Xenopus embryos were fertilized in vitro and cultured in 0.1X MMR. For left-right axis experiments, embryos were injected, 10 nl per cell in the lateral subequatorial sector of dorsal blastomeres at the 4 cell stage for paraxial localization, or in the lateral subequatorial sector of ventral blastomeres for lateral plate mesoderm localization. These localizations correspond approximately to the 32 cell stage C2 and C3 blastomeres, which we have previously described to have paraxial and lateral plate localization (Vonica and Brivanlou, 2007). Coinjections of GFP RNA (1 ng, for phenotype) and LacZ RNA (1 ng, for double in situ hybridization staining) were used to select correctly injected embryos. Staging of embryos was done according to Nieuwkoop and Faber. For RT-PCR of posterior paraxial mesoderm, MO oligonucleotides were injected in both dorsal blastomeres and posterior dorsal fragments containing paraxial mesoderm were cut at stage 18 with a gastromaster (Xenotek Engineering). For RT-PCR and biochemistry of animal caps, all four blastomeres were injected in the animal pole, caps were cut at stage 9 and cultured in 0.5X MMR. For transcription assays, each ventral bastomere of 4 cell stage embryos was injected in the animal pole and embryos were collected at stage 11. For specific localization of different injection sites, 10 ng control fluorescent MO (GeneTools, LLC) was injected at 4 cell stage in the marginal lateral side of the left dorsal blastomere, and Cherry H2b RNA was injected in the marginal lateral side of the left ventral blastomere. Embryos were allowed to develop until stage 20, fixed in formaldehide and exposed to UV of the appropriate wavelength for imaging. Zebrafish embryos were injected at the 1 to 4 cell stage and reared at 28.5°C to appropriate developmental stages. Embryos were staged according to established morphological criteria (Kimmel et al., 1995).
Xenopus APOBEC2 (XA2) wild-type (CD327544, in pCMV-SPORT6 from ATCC) was used for in vitro translation and antisense probe transcription. For in vivo overexpression and rescue experiments we used CS2+XA2 FLAG, constructed as follows: XA2 open reading frame was amplified by PCR from the original vector with the oligonucleotides (S) 5’ CCG GAA TTC ACC ACC ATG GCg CAa cGa CAg AAc AAT TCA CAG TCT TCC AAG GAT 3’ where small case indicates conservative mutations in the MO-homology region, and (AS) 5’ CGG CTC TAG ATT ACT TGT CAT CGT CGT CCT TGT AGT CTC TTA GAA TCT CTG CCA GCT T 3’. The vector contains a C-terminal Flag tag. CS2+ MutXA2 was derived by mutagenizing E102 to Q (GAG to cAG), C130 to S (TGC to TcC) and C134 to S (TGT to TcT). CS2+XA2 HA was constructed by cloning the 5’ end of CS2+XA2 FLAG restricted with HinDIII into CS2+FLAG-PSP-XA2 HA construct.
Zebrafish APOBEC2 (zA2) (BC0907580, from Open Biosystems) was cloned into CS2+ as EcoRI/XbaI fragment from the original pDNR-LIB vector, and used to generate sense RNA for expression, and sense and antisense RNA probes.
The complete coding sequence of mouse APOBEC2 was PCR-amplified from C2C12 cDNA and cloned in the EcoRI site of the pcDNA3.1 vector (Invitrogen), giving rise to the CMV-Apobec2 plasmid.
MO oligonucleotides for XA2 and control fluorescent MO were obtained from Gene Tools, LLC. XA2 MO: 5’ GTG AAT TAT TTT GCC TCT GTG CCA T 3’; XA2 MO2: 5’ CTG TGC CAT TGT GAA TAA ACA GAG A 3’. The negative control siRNA (DS Scrambled Negative) and the siRNA targeting Apobec2 (NM_009694 duplex 2) were purchased from IDT. Derrière and Xnr1 MO have been described (Vonica and Brivanlou, 2007).
Wild-type and 5' mutated xA2 RNA (1 μg) were in vitro translated and labeled with [S35]Met and rabbit reticulocyte lysate (Promega, Inc.) in the absence or presence of xA2 MO (10 μg, (Taylor et al., 1996). Wild-type zA2 RNA was similarly translated in vitro in the presence or absence of 10 μg zA2 MO or zA2 Mut MO. Translation reactions were run on 4-14% Invitrogen precast gels, dried and exposed to film. RNA synthesis was performed as previously described (Vonica and Brivanlou, 2007). RNAs for overexpression were in vitro transcribed from: CS2++CocoFLAG (Bell et al., 2003), CS2 Xnr1 and CS2 Xnr2 (from C. Wright), pSP64T Xnr4 (from D. Melton), pBS Xnr5 (from M. Asashima), pCS2+Vg1(Ser)HA (from J. Heasman), CS2++derrière (Cheng et al., 2003), CS2+GRVP16hSmad2Δ3 (Vonica and Brivanlou, 2007).
For transcription assays, the nodal/activin specific reporter gene A3Luc (Liu et al., 1997) was coinjected with the inducible Smad2 construct GRVP16hSmad2Δ3 RNA in the animal pole at 4 cell stage. Embryos were cultured until stage 11, when they were processed for luciferase assays (Luciferase Assay System, Promega) as described (Vonica and Gumbiner, 2002). Results are expressed as arbitrary luciferase units. All assays were done in triplicate. Significance was calculated with Student's t test for unpaired data.
Animal caps cultured in 0.5X MMR, posterior poles or whole embryos, were collected at the indicated stages and RNA was purified in batches of 10 and processed for semiquantitative RT-PCR as described (Wilson and Melton, 1994). Each experiment was repeated three times. For real-time RT-PCR analysis, RNA was treated with DNA-free kit (Ambion) to remove contaminant genomic DNA and reverse transcribed with the SuperScriptIII kit (Invitrogen). cDNA was then analyzed with LightCycler 480 SYBR Green I kit (Roche) in the LightCycler 480 machine (Roche) and results were analyzed with the REST software (Pfaffl et al., 2002). Sequences of oligonucleotides used are described in Supplementary Methods.
C2C12 myoblasts were maintained in high-glucose DMEM with 15% FBS growth medium (GM). For induction of differentiation, GM was replaced with differentiation medium (DM), containing 2% horse serum. Where indicated, cells were cultured in presence of 5ng/ml recombinant human TGF-β1 (R&D Systems, Inc.) . siRNAs were transfected at a final 10nM concentration with Lipofectamine 2000 (Invitrogen). After six hours, the medium was changed to DM and the cells were analyzed after two days.
For luciferase assays, 300 ng of CMV-APOBEC2 plasmid were cotransfected in C2C12 cells along with 300 ng of the Firefly luciferase encoding plasmid p3TP-Lux and 30 ng of the Renilla luciferase-encoding plasmid pRL-TK (Promega, Inc.) in a 12-well plate. The day after, cells were analyzed for luciferase activity with the Dual Luciferase Assay System (Promega, Inc.), and results were expressed as the ratio of firefly vs. control renilla activity.
Whole mount in situ hybridization in Xenopus was as described (Harland, 1991), with full-length probes transcribed for the following vectors: pBS Xnr1 and pBS Xlefty-a (Branford et al., 2000), CS2++ Coco (Bell et al., 2003), pCMV-SPORT6 derrière (CA789921, from ATCC), pBS MyoD (from A. Salic, Harvard Med. Sch.). For double in situ hybridization, the marker was digoxigenin-labeled and stained with BMPurple (Roche), and the LacZ probe from the pSP6nucβGal (from R. Harland, Berkeley U.) was FITC-labeled and stained with Fast Red (Roche). All in situs were performed in a BioLane HT1 in situ robot (Holle & Huttner AG). The CS2+ Cherry H2b construct was a kind gift from A. Warmflash.
Zebrafish embryos were fixed in 4% paraformaldehyde overnight at 4°C. In situ hybridizations were performed essentially as previously described (Thisse et al., 1993). Zcmlc2 construct was obtained from P.D. Henion (Yelon et al., 1999). ZA2 sense and antisense probes were generated from CS2+ ZA2 (see above).
Imaging was done with a Zeiss HBO 100 microscope and AxioVision imaging program and minimally enhanced in Adobe Photoshop 7.0 imaging program.
For immunohistochemistry of heart muscle, stage 46 Xenopus embryos were fixed in 4% PFA and incubated with MF20 monoclonal antibody for light meromyosin (1:20, Developmental Studies Hybridoma Bank, University of Iowa) and stained with immunoperoxidase as described (Chen and Fishman, 1996).
Samples were run on Invitrogen 4-12% or 10% precast gels, using LDS sample buffer. For Smad 2 phosphorylation, ten animal caps from control or injected embryos were lysed in 1% NP40 buffer. The blots were blocked with 1% serum and 5% milk, and incubated with 1:2500 anti-total Smad2/3 (rabbit polyclonal), 1:500 anti-Phospho C-terminal Samd2/3 (rabbit monoclonal), both from Cell Signaling. The FLAG tag was visualized with the rabbit anti-FLAG (1:5000, Sigma).
We thank Drs. Ann C. Foley, Nina Papavasiliou, and Harold C. Smith for useful comments on the manuscript, Drs. G.T. Williams for the antibody against mouse A2 and K. McBride for the anti-AID antibody. We also thank Drs. M. Asashima, M. Blum, A. Salic, J. Heasman, P. D. Henion, D. Melton, C. Wright, A. Warmflash and H. J. Yost for plasmids. Blaine Cooper offered expert assistance for the manuscript. This work was supported by NIH grants R01 HD032105 to A. H. B., R03 HD057334 to A. V., and A. R. was supported by a postdoctoral fellowship from the Human Frontier Science Program Organization.