Are Cyanobacteria an Ancestor of Chloroplasts or Just One of the Gene Donors for Plants and Algae?

Abstract

Chloroplasts of plants and algae are currently believed to originate from a cyanobacterial endosymbiont, mainly based on the shared proteins involved in the oxygenic photosynthesis and gene expression system. The phylogenetic relationship between the chloroplast and cyanobacterial genomes was important evidence for the notion that chloroplasts originated from cyanobacterial endosymbiosis. However, studies in the post-genomic era revealed that various substances (glycolipids, peptidoglycan, etc.) shared by cyanobacteria and chloroplasts are synthesized by different pathways or phylogenetically unrelated enzymes. Membranes and genomes are essential components of a cell (or an organelle), but the origins of these turned out to be different. Besides, phylogenetic trees of chloroplast-encoded genes suggest an alternative possibility that chloroplast genes could be acquired from at least three different lineages of cyanobacteria. We have to seriously examine that the chloroplast genome might be chimeric due to various independent gene flows from cyanobacteria. Chloroplast formation could be more complex than a single event of cyanobacterial endosymbiosis. I present the "host-directed chloroplast formation" hypothesis, in which the eukaryotic host cell that had acquired glycolipid synthesis genes as an adaptation to phosphate limitation facilitated chloroplast formation by providing glycolipid-based membranes (pre-adaptation). The origins of the membranes and the genome could be different, and the origin of the genome could be complex.

Keywords: Paulinella chromatophore; chloroplast origin; cyanobacterial endosymbiosis; glycolipids; host-directed chloroplast formation; peptidoglycan; phylogenetic analysis.

Figure 1

Four major types of phylogenetic relationships between cyanobacterial and chloroplast proteins. Chloroplast-encoded proteins (or RNA) are marked with an asterisk and colored. [R] and [G] indicate red and green lineages, respectively. This is an extended, revised version of the original in [9,50].

Figure 2

Glycolipid biosynthesis pathways in cyanobacteria and chloroplasts. The distinction between galactose and glucose is shown by the 4’-OH group in red. Cyanobacterial genes are shown in cyan, whereas eukaryotic or non-cyanobacterial genes are shown in green. Note that the biochemical reactions are different in the two steps marked [1] in magenta. In other steps, biochemically identical reactions are catalyzed by structurally unrelated (isofunctional non-homologous) enzymes marked [2]. Eukaryotic paralogs and phylogenetically distant orthologs are marked [3] and [4], respectively.

Figure 3

Schematic phylogenetic trees showing different origins of chloroplast enzymes (or RNA) within the cyanobacterial diversity. (A) Canonical phylogenetic tree of cyanobacteria and chloroplasts. Various clades of cyanobacteria are shown according to [53,54]. Chloroplast-encoded proteins and RNA are shown in green. Nuclear-encoded chloroplast proteins are shown in black. The clade names are taken from [53]. According to [54], G. lithophora is sister to the main chloroplast clade represented by chloroplast rRNA. For simplicity, Clade H is used for G. lithophora. The positions of branching of RbcL and PsaA/B are shown with dotted lines because they do not match exactly with the canonical tree. (B) Different origins of chloroplast and cyanobacterial RbcL. (C) Probable origin of PsaA and PsaB. As explained in the text and Supplementary Materials, the phylogeny of PsaA and PsaB is very difficult. This diagram, which is consistent with previous results, is still a plausible hypothesis among many other alternatives. Different line colors are used to show different lineages such as cyan for cyanobacteria and green for chloroplast. Colored protein names are used to show chloroplast-encoded proteins.

Figure 4

Various scenarios on the origin of non-cyanobacterial proteins in the chloroplast. (A) The chromatophore of Paulinella is a reference, in which all components originate essentially from cyanobacteria. (B–D) Various possible versions of explanations for the non-cyanobacterial proteins in chloroplasts. (B) Cyanobacterial proteins introduced by the primary endosymbiosis were replaced by horizontal gene transfers during the long evolution of plants and algae. (C) Non-cyanobacterial proteins that are conserved in the three lineages must have been introduced before the divergence of the three lineages. (D) Non-cyanobacterial proteins might be introduced before chloroplast formation as a pre-adaptation. Note that the time ranges of chromatophore evolution and chloroplast diversification are about the same (100 million years). See text for details. The colors are used to indicate different lineages, such as cyan for cyanobacteria and green for chloroplast.

Figure 5

Unified schemes explaining the formation of membrane systems of (A) chromatophore and (B,C) chloroplast. The key components are phosphorus-free lipids. DGCC is a betaine lipid, and GL represents glycolipid trio (MGDG, DGDG, and SQDG). In scenario (B), glycolipids that were introduced as an adaptation to phosphate limitation form a membrane system, in which cyanobacterial photosynthesis and gene expression systems were incorporated. In scenario (C), the host cell acquired glycolipids before the cyanobacterial endosymbiosis, but then, the cyanobacterial glycolipid synthesis system was lost. Colors are used to indicate different lineages, such as cyan for cyanobacteria and green for chloroplast.