The phylogeny of C/S1 bZIP transcription factors reveals a shared algal ancestry and the pre-angiosperm translational regulation of S1 transcripts

Abstract

Basic leucine zippers (bZIPs) form a large plant transcription factor family. C and S1 bZIP groups can heterodimerize, fulfilling crucial roles in seed development and stress response. S1 sequences also harbor a unique regulatory mechanism, termed Sucrose-Induced Repression of Translation (SIRT). The conservation of both C/S1 bZIP interactions and SIRT remains poorly characterized in non-model species, leaving their evolutionary origin uncertain and limiting crop research. In this work, we explored recently published plant sequencing data to establish a detailed phylogeny of C and S1 bZIPs, investigating their intertwined role in plant evolution, and the origin of SIRT. Our analyses clarified C and S1 bZIP orthology relationships in angiosperms, and identified S1 sequences in gymnosperms. We experimentally showed that the gymnosperm orthologs are regulated by SIRT, tracing back the origin of this unique regulatory mechanism to the ancestor of seed plants. Additionally, we discovered an earlier S ortholog in the charophyte algae Klebsormidium flaccidum, together with a C ortholog. This suggests that C and S groups originated by duplication from a single algal proto-C/S ancestor. Based on our observations, we propose a model wherein the C/S1 bZIP dimer network evolved in seed plants from pre-existing C/S bZIP interactions.

Figure 1. Conservation of C and S1 bZIP orthologs across angiosperm species.

The figure combines consistent topologies from independent lineage-specific Maximum Likelihood phylogenetic trees of angiosperm C and S1 bZIP subfamilies. For each species, values indicate the number of orthologs directly related to arabidopsis (A. thaliana) or rice (O. sativa) as reference bZIPs (in bold) for dicots or monocots, respectively. Notice that no direct correspondence exists between individual dicot and monocot orthologs (individual columns), but only between groups of orthologs (white blocks). Dashes indicate missing orthologs. See Suppl. Figs S1 and S2 for original C and S1 bZIP trees, respectively.

Figure 2. A phylogeny of C and S bZIP subfamilies in green plants.

Simplified phylogenetic trees of the C and S bZIP subfamilies showing the relationship between the two groups of orthologs in green plants. Known C and S1 bZIP orthologs from arabidopsis and rice are shown in bold. The topology represents the consensus of independent phylogenetic reconstructions shown in Suppl. Fig. S3. S1 bZIP orthologs are indicated based on the presence of conserved S1 5′uORFs.

Figure 3. Sequence alignment of 5′uORFs and experimental confirmation of SIRT from gymnosperms putative S1 bZIPs.

(A) Alignment of 5′uORFs from angiosperms and gymnosperms S1 bZIPs showing conservation across each lineage (indicated in violet and light blue respectively, according to the color code of Fig. 2). Residues known to be necessary for SIRT in arabidopsis are marked with a red star. Gymnosperms sequences tested for SIRT are indicated with a red arrow. (B) Schematic representation of the constructs used to test SIRT in the transient LUC expression assay. Rectangles represent 5′uORFs shown in panel A, with proportions reflecting length and distance from the main ORF. A red line in the mutant AtbZIP11 (Y39A) indicates the substitution point (negative control). An empty 35S:rLUC construct was used to normalize LUC activity data. (C) Results of relative LUC activity assays. Normalized LUC activity is presented relative to arabidopsis WT construct results in sorbitol. Values represent the average of at least four biological replicates. Error bars indicate SD from the mean. Stars indicate the significance of a two-tailed distribution t-test with unequal variance (**p < 0.01, ***p < 0.005). For the original results see Suppl. Table S2.

Figure 4. Presence pattern of C and S bZIP orthologs in different plant lineages and a model for the evolution of the C/S1 dimerization network.

(A) Schematic summary of C and S or S1 bZIP orthologs coexistence in different plant lineages according to our results. Ticks and dashes indicate presence and absence, respectively. Based on the presence pattern, the putative formation of C/S or C/S1 heterodimers can be postulated (lowest row, yellow and green background respectively). (B) The gene tree summarizes the findings presented in this paper, and illustrates a possible model for the emergence of the C/S1 bZIP dimerization network as described in the text. Abbreviations refer to the plant lineage names used in panel A. Dot pairs and dimer cartoons provide information on approximate time of appearance and ortholog group of the bZIP sequences involved in hypothetical interactions, respectively. (Photographic references: the chlorophyta picture (https://commons.wikimedia.org/wiki/File:Micrasterias_.jpg) is public. The charophyta picture (https://commons.wikimedia.org/wiki/File:Klebsormidium_bilatum_Belgium_%2814759117646%29.jpg) is licensed under the Creative Commons Attribution 2.0 Generic license. The license terms can be found on the following link: https://creativecommons.org/licenses/by/2.0/deed.en. The bryophyta picture (https://commons.wikimedia.org/wiki/File:Bryi1004.JPG) is licensed under the Creative Commons Attribution-Share Alike 4.0 International license. The license terms can be found on the following link: https://creativecommons.org/licenses/by-sa/4.0/deed.en. The lycopodiophyta, pteridophyta, gymnosperms, and angiosperms pictures (https://commons.wikimedia.org/wiki/File:Selaginella_canaliculata.jpeg, https://commons.wikimedia.org/wiki/File:DidzialapisSakys.JPG, https://commons.wikimedia.org/wiki/File:Abies_homolepis_cones.jpg, https://commons.wikimedia.org/wiki/File:Rosa_Red_Chateau01.jpg, respectively) are licensed under the Creative Commons Attribution-Share Alike 3.0 Unported license. The license terms can be found on the following link: https://creativecommons.org/licenses/by-sa/3.0/deed.en).