Nt from conventional HMGA. This conclusion is further supported by bioinformatic prediction of disordered sequences with the Predictor Of Naturally Disordered Regions (PONDR) software [33], showing that XHMG-AT-hook1 has only 75.5 of disordered structure, while human HMGA2 and Xenopus XLHMGA2 are prototypes of intrinsically disordered proteins having 100 of disordered structure (data not shown).DiscussionIn this paper we report the cloning and developmental expression of a new gene, distantly related to HMGA1 and HMGA2, that we named Xhmg-at-hook. We have analyzed Xhmg-athook pattern of expression during Xenopus development and found that its main domains of expression are in the developing CNS, NCC and eye.The deduced XHMG-AT-hook protein shares with typical HMGA the AT-hook DNA binding domain, but, differently from HMGA1 and HMGA2, has 6 or 8 of such motifs. In our case, the comparison of XHMG-AT-hook1 DNA binding activity with that of typical HMGA shows a clear difference: XHMG-AT-hook1 protein does not bind the typical sequence targets recognized by both human and Xenopus HMGA proteins. Besides, also the protein-protein interaction activity of XHMG-AT-hook1 is different from those of typical HMGA. These results suggest that XHMG-AT-hook factors do not share the typical characteristics of the HMGA family and therefore should not be included in this family. Multi-AT-hook proteins have been described in other organisms: some highly divergent HMGA proteins (like D1 in Drosophila) and other proteins containing several AT-hooks (in plants) have been shown to behave as canonical HMGA. On the contrary, other AT-hook-containing proteins have been reported to exert different functions from HMGA proteins and classified as non-canonical HMGA proteins [34]. XHMG-AT-hook1? should therefore be included in this last category.Figure 5. XLHMGA2ba, but not XHMG-AT-hook1, interacts with the same molecular partners of mammalian HMGA. (A) Blue coomassie stained analysis of different HMGA molecular partners produced as GST-fused protein and of GST alone. PR: pocket region; CT: C-terminal region; ZnF: Zinc finger MedChemExpress CB-5083 region. (B) GST-pull down assays performed with the GST-fused HMGA molecular partners shown in panel A and IVT and [35S]methionine radiolabeled XHMG-AT-hook1 (HA ATH1), hHMGA2 (HA A2), and XLHMGA2ba (HA LA2ba). For each IVT protein used input is shown in lanes 1, 6, and 10 (10 of the amount used in GST-pull down experiments). GST alone is used as a negative control. doi:10.1371/58-49-1 biological activity journal.pone.0069866.gMulti-AT-Hook Factors in XenopusBy MO injection experiments, we have shown that they play a possibly redundant role in Xenopus laevis development. In fact, consistent with the pattern of mRNA expression, combined injection of MOs against the two mRNA forms expressed in early embryogenesis, Xhmg-at-hook1 and Xhmg-at-hook3, leads to reductions in the eye and parts of the pharyngeal skeleton. These effects are at least in part consistent with the reduced expression of the rostral brain marker Xotx2, of the neural marker nrp-1, and of the NCC marker Twist observed in MO injected embryos. In fact, in all embryos injected with both 23977191 MOs, Xotx2 and nrp1 expression was reduced in the developing eye, though some effects were also seen in other parts of the CNS. It is interesting to note that while injection of single MOs only produced weak phenotypic effects in a minority of embryos, upon combined injections of MoXat1 and MoXat3 there is a definite shift tow.Nt from conventional HMGA. This conclusion is further supported by bioinformatic prediction of disordered sequences with the Predictor Of Naturally Disordered Regions (PONDR) software [33], showing that XHMG-AT-hook1 has only 75.5 of disordered structure, while human HMGA2 and Xenopus XLHMGA2 are prototypes of intrinsically disordered proteins having 100 of disordered structure (data not shown).DiscussionIn this paper we report the cloning and developmental expression of a new gene, distantly related to HMGA1 and HMGA2, that we named Xhmg-at-hook. We have analyzed Xhmg-athook pattern of expression during Xenopus development and found that its main domains of expression are in the developing CNS, NCC and eye.The deduced XHMG-AT-hook protein shares with typical HMGA the AT-hook DNA binding domain, but, differently from HMGA1 and HMGA2, has 6 or 8 of such motifs. In our case, the comparison of XHMG-AT-hook1 DNA binding activity with that of typical HMGA shows a clear difference: XHMG-AT-hook1 protein does not bind the typical sequence targets recognized by both human and Xenopus HMGA proteins. Besides, also the protein-protein interaction activity of XHMG-AT-hook1 is different from those of typical HMGA. These results suggest that XHMG-AT-hook factors do not share the typical characteristics of the HMGA family and therefore should not be included in this family. Multi-AT-hook proteins have been described in other organisms: some highly divergent HMGA proteins (like D1 in Drosophila) and other proteins containing several AT-hooks (in plants) have been shown to behave as canonical HMGA. On the contrary, other AT-hook-containing proteins have been reported to exert different functions from HMGA proteins and classified as non-canonical HMGA proteins [34]. XHMG-AT-hook1? should therefore be included in this last category.Figure 5. XLHMGA2ba, but not XHMG-AT-hook1, interacts with the same molecular partners of mammalian HMGA. (A) Blue coomassie stained analysis of different HMGA molecular partners produced as GST-fused protein and of GST alone. PR: pocket region; CT: C-terminal region; ZnF: Zinc finger region. (B) GST-pull down assays performed with the GST-fused HMGA molecular partners shown in panel A and IVT and [35S]methionine radiolabeled XHMG-AT-hook1 (HA ATH1), hHMGA2 (HA A2), and XLHMGA2ba (HA LA2ba). For each IVT protein used input is shown in lanes 1, 6, and 10 (10 of the amount used in GST-pull down experiments). GST alone is used as a negative control. doi:10.1371/journal.pone.0069866.gMulti-AT-Hook Factors in XenopusBy MO injection experiments, we have shown that they play a possibly redundant role in Xenopus laevis development. In fact, consistent with the pattern of mRNA expression, combined injection of MOs against the two mRNA forms expressed in early embryogenesis, Xhmg-at-hook1 and Xhmg-at-hook3, leads to reductions in the eye and parts of the pharyngeal skeleton. These effects are at least in part consistent with the reduced expression of the rostral brain marker Xotx2, of the neural marker nrp-1, and of the NCC marker Twist observed in MO injected embryos. In fact, in all embryos injected with both 23977191 MOs, Xotx2 and nrp1 expression was reduced in the developing eye, though some effects were also seen in other parts of the CNS. It is interesting to note that while injection of single MOs only produced weak phenotypic effects in a minority of embryos, upon combined injections of MoXat1 and MoXat3 there is a definite shift tow.
