A New Rubisco-like Protein Coexists with a Photosynthetic Rubisco in the Planktonic Cyanobacteria Microcystis

Abstract

Two genes encoding proteins related to large subunits of Rubisco were identified in the genome of the planktonic cyanobacterium Microcystis aeruginosa PCC 7806 that forms water blooms worldwide. The rbcL I gene belongs to the form I subfamily typically encountered in cyanobacteria, green algae, and land plants. The second and newly discovered gene is of the form IV subfamily and widespread in the Microcystis genus. In M. aeruginosa PCC 7806 cells, the expression of both rbcL I and rbcL IV is sulfur-dependent. The purified recombinant RbcL IV overexpressed in Escherichia coli cells did not display CO 2 fixation activity but catalyzed enolization of 2,3-diketo-5-methylthiopentyl-1-phosphate, and the rbcL IV gene rescued a Bacillus subtilis MtnW-deficient mutant. Therefore, the Microcystis RbcL IV protein functions both in vitro and in vivo and might be involved in a methionine salvage pathway. Despite variations in the amino acid sequences, RbcL IV shares structural similarities with all members of the Rubisco superfamily. Invariant amino acids within the catalytic site may thus represent the minimal set for enolization, whereas variations, especially located in loop 6, may account for the limitation of the catalytic reaction to enolization. Even at low protein concentrations in vitro, the recombinant RbcL IV assembles spontaneously into dimers, the minimal unit required for Rubisco forms I-III activity. The discovery of the coexistence of RbcL I and RbcL IV in cyanobacteria, the ancestors of chloroplasts, enlightens episodes of the chaotic evolutionary history of the Rubiscos, a protein family of major importance for life on Earth.

Function, Structure, and Evolution of the RubisCO-Like Proteins and Their RubisCO Homologs

“…9 does not explain the presence of RLPs in the picoeukaryote marine chlorophyte O. tauri, which encodes two distinct and highly divergent RLPs in its nuclear genome (17) in addition to a typical form I large The other unusual prokaryote-to-eukaryote lateral transfer in this scheme, which explains the presence of form II RubisCO in the Dinophyceae, has been previously analyzed in detail (44,51,52,58). The archaeal origin model of RubisCO/RLP evolution proposed here is substantially different from that reported previously by Ashida et al (7) and Carre-Mlouka et al (11), who speculated that bona fide RubisCOs arose in the YkrW lineage. However, those authors relied on much smaller sets of sequences and more limited numbers of phylogenetic reconstructions to reach their conclusions.…”

“…This conundrum was recently addressed, with two possibilities considered: (i) form III archaeal RubisCO preferentially uses an alternative substrate and does not require RuBP for catalysis, or (ii) alternative means to synthesize RuBP that are unique to archaea exist. The first possibility, if true, would also suggest that RuBP-dependent RubisCO activity might have evolved from a protein that possessed some alternative activity, a theory espoused by those that believe that RLP is an evolutionary precursor to RubisCO (7,11). However, exhaustive studies have thus far found no alternative to RuBP as a substrate or CO 2 acceptor for archaeal (or any other) RubisCO (26), suggesting that it is unlikely that form III RubisCO is being used for anything other than producing 3-phosphoglyceric acid (PGA) from RuBP and CO 2 .…”

“…Currently, detailed functional studies have been carried out for only four RLPs, C. tepidum RLP (30,31), the YkrW/MtnW proteins of Bacillus subtilis and Geobacillus kaustophilus (8,33,45,63), and the YkrW-like RLP from the cyanobacterium Microcystis aeruginosa (11). Thus far, the three-dimensional structures have been solved only for C. tepidum RLP (37), R. palustris RLP2 (this paper), and G. kaustophilus RLP (33,39).…”

“…The first group consisted of the RLPs from C. tepidum, R. palustris (RLP1 and RLP2), Archaeoglobus fulgidus, Mesorhizobium loti, and Sino- (8). In addition, a cyanobacterial (M. aeruginosa) RLP gene has also been shown to functionally complement the B. subtilis mutant (11). Detailed functional and structural relationships among bona fide RubisCO and RLP are extensively discussed below; clearly, bioinformatic analyses suggest discrete functions for at least some of the phylogenetically diverse RLPs discussed here.…”

Enzymatic Characterization of 5-Methylthioribose 1-Phosphate Isomerase fromBacillus subtilis

“…MTR-1-P isomerase was assayed in a coupling reaction with MtnB and MtnW, previously identified as MTRu-1-P dehydratase and 2,3-diketo-5-methylthiopentyl-1-phosphate enolase respectively. 2,7,15) In this assay, the reaction product MTRu-1-P is converted to 2-hydroxy-3-keto-5-methylthiopentenyl-1-phosphate (HK-MTPenyl-1-P) by sequential reactions of MtnB and MtnW. The reaction was started by adding 2 mM MTR-1-P to a mixture containing 50 mM Tris-HCl buffer (pH 8.1), 1 mM MgCl 2 , 0.5 mg of MTR-1-P isomerase, 5 mg of MtnB, and 15 mg of MtnW.…”