Smad1 and Smad5 are highly homologous, but can be individually targeted for loss of function studies using mutants and MO knockdown.(A) Comparison of zebrafish Smad1 and Smad5 protein structure. The MH1 and MH2 domains are highly similar, compared with the more divergent central linker region. The piggytail mutation (★) is located in the MH2 domain at amino acid 447 of Smad5. (B) In vitro transcription-translation of ³⁵S-methionine–labeled Smad proteins. Smad1 (S1) can be efficiently and specifically targeted for knockdown by the addition of the Smad1 translation-blocking MO (S1MO; lanes 1 and 2). The reaction is not inhibited by the addition of the Smad5 MO (S5MO; lane 3). Smad5 (S5) can be targeted for inhibition by the addition of the Smad5 MO, but not the Smad1 MO (lanes 4-6). (C) In control reactions, a distinct MO (G4MO, specific for the gata4 gene) was used in equivalent reactions against Smad1 or Smad5. In each case, the control MO had no affect on translation, confirming that inhibition seen in panel B is specific. (D) RT-PCR analysis comparing RNA from uninjected embryos and embryos injected with the splice site–blocking MO (S1MO1) collected at 32 hpf. βactin serves as a positive control and generates a band of 277 bp. Primers that span exon2-intron2 generate for smad1 a 232-bp fragment. S1MO1 injection efficiently decreases the amount of properly spliced smad1 message. The smaller fragment at 170 bp was shown by sequencing to encode a misspliced smad1 isoform.