Cx43-20 kDa isoform is generated from Internal Ribosome Entry Site (IRES). a
twist ISH lcompared with Cx43 immunostaining in stage 24 embryos; IgG control antibody; arrow: neural crest; asterisk: eye. Drawing: Cx43 and twist expression summary. Scale bar = 80 μm. b Cx43-20k protein levels (analyzed by WB) and N-cadherin mRNA (analyzed by qPCR) in st23 NC cells (n = 61 embryos, N = 3). c–h Cx43 putative IRES and activity of Cx43 isoforms. c Predicted methionine-initiated polypeptide. Green: constructs used in e. Red arrows: abundant peptides found in Xenopus; gray arrows: less abundant or absent peptides in Xenopus. d WB of embryos at NC migratory stages with antibody against the Cx43 C-terminus. Major bands correspond to Cx43FL and Cx43-20k (arrows). e WB of neural crest expressing the Cx43-20k-HA with an antibody against HA. The only band generated corresponds to the M276-initiated peptide (Cx43-20k). f Injection of Cx43-20k, but not Cx43-11k, leads to induction of N-cadherin expression in blastula embryos; percentage of embryos displaying N-cadherin expression, (nCx43-11k = 20 embryos nCx43-20k = 20 embryos, N = 3). Scale bar = 70 μm. g WB against Cx43 from embryos at st21 injected with Cx43-HA (lane 1) and with Cx43 mutated in methionine 213 (M213L Cx43, lane 2). h WB’s quantification (nlane1 = 20 embryos nlane2 = 20 embryos, N = 4). i WB of st22 neural crest cells with antibody against the Cx43 C terminus. Lane1: control neural crest; lane 2: neural crest injected with Hif-1α morpholinos; lane 3: neural crest expressing a dominant active form of Hif-1α. j Quantification of Cx43 isoforms (ncontrol = 50; nHif-1αMO = 50; nHif-1αΔ = 50 neural crests, N = 6 independent experiments). Hif-1αMO and Hif-1αΔ were previously validated25. In f, h, and j histograms represent mean and bars show s.e.m. (in j one-way ANOVA p < 0.001, two-tailed t test, p** < 0.01, p*** < 0.001, p**** < 0.0001). N number of independent experiments; n sample size. Spread of data in bar charts is shown as overlying dots