Activation Mechanism and Cellular Localization of Membrane-Anchored Alginate Polymerase in Pseudomonas aeruginosa

Abstract

The exopolysaccharide alginate, produced by the opportunistic human pathogen Pseudomonas aeruginosa, confers a survival advantage to the bacterium by contributing to the formation of characteristic biofilms during infection. Membrane-anchored proteins Alg8 (catalytic subunit) and Alg44 (copolymerase) constitute the alginate polymerase that is being activated by the second messenger molecule bis-(3', 5')-cyclic dimeric GMP (c-di-GMP), but the mechanism of activation remains elusive. To shed light on the c-di-GMP-mediated activation of alginate polymerization in vivo, an in silico structural model of Alg8 fused to the c-di-GMP binding PilZ domain informed by the structure of cellulose synthase, BcsA, was developed. This structural model was probed by site-specific mutagenesis and different cellular levels of c-di-GMP. Results suggested that c-di-GMP-mediated activation of alginate polymerization involves amino acids residing at two loops, including H323 (loop A) and T457 and E460 (loop B), surrounding the catalytic site in the predicted model. The activities of the respective Alg8 variants suggested that c-di-GMP-mediated control of substrate access to the catalytic site of Alg8 is dissimilar to the known activation mechanism of BcsA. Alg8 variants responded differently to various c-di-GMP levels, while MucR imparted c-di-GMP for activation of alginate polymerase. Furthermore, we showed that Alg44 copolymerase constituted a stable dimer, with its periplasmic domains required for protein localization and alginate polymerization and modification. Superfolder green fluorescent protein (GFP) fusions of Alg8 and Alg44 showed a nonuniform, punctate, and patchy arrangement of both proteins surrounding the cell. Overall, this study provides insights into the c-di-GMP-mediated activation of alginate polymerization while assigning functional roles to Alg8 and Alg44, including their subcellular localization and distribution.IMPORTANCE The exopolysaccharide alginate is an important biofilm component of the opportunistic human pathogen P. aeruginosa and the principal cause of the mucoid phenotype that is the hallmark of chronic infections of cystic fibrosis patients. The production of alginate is mediated by interacting membrane proteins Alg8 and Alg44, while their activity is posttranslationally regulated by the second messenger c-di-GMP, a well-known regulator of the synthesis of a range of other exopolysaccharides in bacteria. This study provides new insights into the unknown activation mechanism of alginate polymerization by c-di-GMP. Experimental evidence that the activation of alginate polymerization requires the engagement of specific amino acid residues residing at the catalytic domain of Alg8 glycosyltransferase was obtained, and these residues are proposed to exert an allosteric effect on the PilZ_Alg44 domain upon c-di-GMP binding. This mechanism is dissimilar to the proposed mechanism of the autoinhibition of cellulose polymerization imposed by salt bridge formation between amino acid residues and released upon c-di-GMP binding, leading to activation of polymerization. On the other hand, conserved amino acid residues in the periplasmic domain of Alg44 were found to be involved in alginate polymerization as well as modification events, i.e., acetylation and epimerization. Due to the critical role of c-di-GMP in the regulation of many biological processes, particularly the motility-sessility switch and also the emergence of persisting mucoid phenotypes, these results aid to reach a better understanding of biofilm-associated regulatory networks and c-di-GMP signaling and might assist the development of inhibitory drugs.

Keywords: Pseudomonas aeruginosa; alginate; polymerases.

FIG 1

Highly conserved amino acids of Alg8 are involved in c-di-GMP-dependent regulation of alginate polymerization. (a) The in silico fusion of Alg8-PilZ_Alg44 was modeled using the Phyre2 server. Residues selected for site-specific mutagenesis were shown on loop A (magenta), loop B (blue), and the PilZ domain (green). Mutations of bolded and underlined residues were responsive to the absence and presence of RocR overproduction (i.e., reduced levels of c-di-GMP), while for other shown residues alginate production was abolished independent of RocR. (b) Alginate quantification of PDO300Δalg8 transformants harboring various plasmids containing respective site-specific mutants of alg8 with (+) and without (−) the rocR gene. (c) Immunoblot analysis of envelope fractions developed using an anti-Alg8 antibody showed that none of mutations affected the Alg8 localization to the envelope fraction (lanes 3 to 7 and lanes 10 to 14). Lanes 1 to 8 and 2 to 9 represent negative and positive controls, respectively. For estimating relative protein amounts, the protein band intensity was analyzed by the ImageJ software. The protein band based on genomic expression derived from 300Δ8+alg8 (lanes 2 and 9) was set as 1.0 in density, and relative densities of other bands were calculated as follows: lane 3 (1.14), lane 4 (0.937), lane 5 (1.06), lane 6 (1.03), lane 7 (1.48), lane 10 (0.9), lane 11 (0.848), lane 12 (0.817), lane 13 (0.985), and lane 14 (0.987). (d) Highly conserved amino acids of Alg8 whose replacement with alanine decoupled alginate polymerization from c-di-GMP-dependent and MucR-dependent regulation. Alginate quantification was performed for PDO300ΔmucRΔalg8 transformants with plasmids harboring respective site-specific mutants of alg8 with (+) and without (−) the rocR gene. The data in histograms in panels b and d represent the means ± SD for four independent repetitions, and treatments with different lowercase italic letters above the bars are significantly different (post hoc Tukey's HSD test, P < 0.05). CDM, cell dry mass; 300, PDO300; ND, not detectable; MCS5, pBBR1MCS-5.

FIG 2

Role of conserved amino acid residues of Alg44 proposed to be localized to the periplasmic domain in alginate biosynthesis and purification of the Alg44 dimer produced by recombinant P. aeruginosa. (a) Uniprot analysis of the periplasmic domain of Alg44 shows the alignment of highly conserved regions among different alginate-producing species P. aeruginosa (PA), Azotobacter vinelandii (AZ), Pseudomonas fluorescens (PF), Pseudomonas putida (PP), and Pseudomonas syringae (PS). Dashed lines show the positions of these highly conserved regions in the Alg44 model predicted by the Phyre2 server. Alginate quantification showed that P266A, C267A, and/or C269A completely abolished alginate production and that for other residues it was significantly reduced. The data represent the means ± SD for four repetitions, and treatments with different lowercase italic letters above the bars are significantly different (post hoc Tukey's HSD test, P < 0.05). (b) Production of free uronic acids in liquid culture mediated by variants of Alg44 (flowthrough samples were obtained using filters with a 10-kDa cutoff), indicating that alginate polymerization was impaired by site-specific mutagenesis of highly conserved periplasmic amino acid residues of Alg44. The data represent the means ± the SD for four repetitions, and asterisks indicate pairs of significantly different values (post hoc Tukey's HSD test: *, P < 0.05; ***, P < 0.001). (c) Immunoblot analysis of envelope fractions developed using anti-His tag antibodies showed that mutations M259A, P266A, C267A, D268A, and C269A completely disrupted Alg44 localization to the envelope fraction (lane 2 and lanes 6 to 10). The intensity of other protein bands was consistent with the amount of alginate produced by complemented mutants (lanes 1, 3 to 5, and 12). For estimating relative protein amounts, the protein band intensity was analyzed by the ImageJ software. The band of genomic expression of alg44-6His (lane 11) was set as 1.0 in density, and the relative densities of other bands were calculated as follows: lane 1 (0.76), lane 3 (0.76), lane 4 (0.29), lane 5 (0.13), and lane 12 (2.1). Lanes 13 and 14 represent negative controls. (d) Composition of alginates impacted by various Alg44 variants (see also Tables S1 and S2 in the supplemental material). Ac.%, percentage of acetylation; F_G, molar fraction of guluronate (G) residue. (e) Purification of the Alg44 dimer. SDS-PAGE gel (stained with Coomassie brilliant blue) (lanes 1 to 3) and an immunoblot (lanes 4 to 6) of peak I (Fig. S7) showed a very stable dimer plus the monomer bands of Alg44. TM, transmembrane domain; CDM, cell dry mass; 300, PDO300; MCS5, pBBR1MCS-5; ND, not detectable.

FIG 3

Alg8 and Alg44 proteins appear localized and distributed in a nonuniform, punctate, and patchy arrangement in the cell envelope of P. aeruginosa. Schematic membrane models of the C-terminal fusions of Alg8 and Alg44 with sfGFP protein are presented on the left. In addition to the C-terminal fusion of PilN with sfGFP protein, all respective images of bacteria that appeared with coplanar orientation based on fluorescent distribution were visualized using CLSM and analyzed using IMARIS image analysis software (Bitplane), each at three independent times. Analysis of various numbers of optical sections (z-stack series), which were separated by 0.13 μm, showed fluorescent foci of PDO300Δalg8(pBBR1MCS-5:alg8:sfgfp) (columns a to c) and PDO300Δalg44(pBBR1MCS-5:alg44:sfgfp) (columns d to f) are arrayed in particular puncta and patchy patterns around the cells, while fluorescent foci in PDO300 (pBBR1MCS-5:pilN:sfgfp) (columns g to i) were uniformly distributed and not in any particular pattern, which ruled out possible technical artifacts. CM, cytoplasmic membrane; FLUOR, fluorescent; DIC, differential interference contrast.