SUMMARYIn angiosperms, shoot branching greatly determines overall plant architecture and affects fundamental aspects of plant life. Branching patterns are determined by genetic pathways conserved widely across angiosperms. In Arabidopsis thaliana (Brassicaceae, Rosidae) BRANCHED1 (BRC1) plays a central role in this process, acting locally to arrest axillary bud growth. In tomato (Solanum lycopersicum, Solanaceae, Asteridae) we have identified two BRC1-like paralogues, SlBRC1a and SlBRC1b. These genes are expressed in arrested axillary buds and both are down-regulated upon bud activation, although SlBRC1a is transcribed at much lower levels than SlBRC1b. Alternative splicing of SlBRC1a renders two transcripts that encode two BRC1-like proteins with different C-t domains due to a 3¢-terminal frameshift. The phenotype of loss-of-function lines suggests that SlBRC1b has retained the ancestral role of BRC1 in shoot branch suppression. We have isolated the BRC1a and BRC1b genes of other Solanum species and have studied their evolution rates across the lineages. These studies indicate that, after duplication of an ancestral BRC1-like gene, BRC1b genes continued to evolve under a strong purifying selection that was consistent with the conserved function of SlBRC1b in shoot branching control. In contrast, the coding sequences of Solanum BRC1a genes have evolved at a higher evolution rate. Branch-site tests indicate that this difference does not reflect relaxation but rather positive selective pressure for adaptation.

Amplification and diversification of transcriptional regulators that control development is a driving force of morphological evolution. A major source of protein diversity is alternative splicing, which leads to the generation of different isoforms from a single gene. The mechanisms and timing of intron evolution nonetheless remain unclear, and the functions of alternative splicing-generated protein isoforms are rarely studied. In Solanum tuberosum, the BRANCHED1a (BRC1a) gene encodes a TCP transcription factor that controls lateral shoot outgrowth. Here, we report the recent evolution in Solanum of an alternative splice site in BRC1a that leads to the generation of two BRC1a protein isoforms with distinct C-terminal regions, BRC1a(Long) and BRC1a(Short), encoded by unspliced and spliced mRNA, respectively. The BRC1a(Long) C-terminal region has a strong activation domain, whereas that of BRC1a(S) lacks an activation domain and is predicted to form an amphipathic helix, the H domain, which prevents protein nuclear targeting. BRC1a(Short) is thus mainly cytoplasmic, while BRC1a(Long) is mainly nuclear. BRC1a(Long) functions as a transcriptional activator, whereas BRC1a(Short) appears to have no transcriptional activity. Moreover, BRC1a(Short) can heterodimerize with BRC1a(Long) and act as a dominant-negative factor; it increases BRC1a(Long) concentration in cytoplasm and reduces its transcriptional activity. This alternative splicing mechanism is regulated by hormones and external stimuli that control branching. The evolution of a new alternative splicing site and a novel protein domain in Solanum BRC1a led to a multi-level mechanism of post-transcriptional and post-translational BRC1a regulation that effectively modulates its branch suppressing activity in response to environmental and endogenous cues.

The carboxysomal polypeptides of Thiobacillus neapolitanus with apparent molecular masses of 85 and 130 kDa were isolated and subjected to N-terminal sequencing. The first 17 amino acids of the two peptides were identical. The sequence perfectly matched the deduced amino acid sequence of an open reading frame in the carboxysome operon. The gene was subsequently named csoS2. Expression of the gene in Escherichia coli resulted in the production of two peptides with apparent molecular masses of 85 and 130 kDa. Immunospecific antibodies generated against the smaller peptide recognized both peptides; the peptides were named CsoS2A and CsoS2B, respectively. A digoxigenin-hydrazide glycosylation assay revealed that both CsoS2A and CsoS2B are post-translationally modified by glycosylation. CsoS2 was localized to the edges of purified carboxysomes by immunogold electron microscopy using the monospecific CsoS2A antibodies. The molecular mass of CsoS2A calculated from the nucleotide sequence was 92.3 kDa.