The interplay between the polar growth determinant DivIVA, the segregation protein ParA, and their novel interaction partner PapM controls the Mycobacterium smegmatis cell cycle by modulation of DivIVA subcellular distribution

Abstract

The genus of Mycobacterium includes important clinical pathogens (M. tuberculosis). Bacteria of this genus share the unusual features of their cell cycle such as asymmetric polar cell elongation and long generation time. Markedly, control of the mycobacterial cell cycle still remains not fully understood. The main cell growth determinant in mycobacteria is the essential protein DivIVA, which is also involved in cell division. DivIVA activity is controlled by phosphorylation, but the mechanism and significance of this process are unknown. Here, we show how the previously established protein interaction partner of DivIVA in mycobacteria, the segregation protein ParA, affects the DivIVA subcellular distribution. We also demonstrate the role of a newly identified M. smegmatis DivIVA and ParA interaction partner, a protein named PapM, and we establish how their interactions are modulated by phosphorylation. Demonstrating that the tripartite interplay affects the mycobacterial cell cycle contributes to the general understanding of mycobacterial growth regulation.

Keywords: ParA; chromosome segregation; mycobacteria; polar growth.

Fig 1

Identification of PapM as a novel M. smegmatis ParA interaction partner. (A) BTH analysis shows T18-PapM interactions. The blue color of the colony indicates interaction. (B) Affinity chromatography confirmation of PapM-ParA interactions. The SDS-PAGE showing the His-PapM binding to GST-ParA but not to the GST column. First, the lysate of the E. coli BL21 producing GST-ParA (lane 2) or GST (lane 3) was loaded on GSH-Sepharose; next, after the washing of unbound proteins, lysate of the E. coli BL21 producing His-PapM (lane 1) was loaded on both columns. The proteins retained at resin were eluted with glutathione (lanes 4 and 5). The bottom panel shows the Western blotting confirmation of His-PapM presence in eluate in the presence of GST-ParA but not in control GST eluate. Anti-His antibody was used for His-PapM detection. Presented result is the representative example of independent four experimental repeats.

Fig 2

Phosphorylation enhances PapM binding to DivIVA while diminishing ParA recruitment to DivIVA in E. coli. (A) Colocalization of PapM-mTurquoise2 (MT2) with Ics-mCherry. (B) Colocalization of PapM-mT2 with mCherry-DivIVA. (C) Colocalization of mCherry-DivIVA and EGFP-ParA. (D) Colocalization of mCherry-DivIVA, EGFP-ParA and PapM-mT2. (E) Colocalization EGFP-ParA with mCherry-DivIVA in presence of His-PknB_KD. (F) Colocalization of mCherry-DivIVA, EGFP-ParA, and PapM-mT2 in the presence of His-PknB_KD. (G) Colocalization of phosphoablative mCherry-DivIVAT74A variant and EGFP-ParA in the presence of His-PknB_KD. (H) Colocalization of PapM-mT2 with mCherry-DivIVA in the presence of His-PknB_KD. Left panels show the profiles of fluorescence measured along the E. coli BL21 cells, and the right panels show the representative images of E. coli cells producing EGFP-ParA (green), mCherry-DivIVA (red), and PapM-mT2 (blue). The number of the cells used for analysis (n) is indicated in the plots; scale bar, 5 µm. (I) The ratio of EGFP-ParA (left panel) and PapM-mT2 (right panel) fluorescence at the pole to the fluorescence at the mid-cell. The ratio higher than 1 (dashed line) means more fluorescence signal at the poles; the ratio lower than 1 illustrates the fluorescence in the mid-cell. The data come from at least two independent biological replicates. The statistical significance between strains determined by Wilcoxon test (two-sided) with Holm method used for multiple comparisons is marked with asterisks: P-values ≤0.05 (*), ≤0.01 (**), ≤0.001 (***), and ≤0.0001 (****).

Fig 3

papM deletion or its overexpression in wild-type and ΔparA background affects M. smegmatis culture growth rate and cell length. (A) The growth curves of ΔpapM, ΔparA, and ΔparAΔpapM strain compared to the wild-type (WT) strain; inset, example image of branched ΔpapM, ΔparA M. smegmatis cells. The image shows the overlay of DAPI fluorescence (nucleoid stain, blue) and brightfield. Scale bar, 5 µm. (B) The growth curve of papM overexpressing strain in wild-type (papM↑) and ΔparA background (ΔparA papM↑) as compared to control strains with empty pMVp_ami plasmid (WT + pMVp_amiØ and ΔparA + pMVp_amiØ). (C) Cell length distribution in ΔpapM (478 cells), ΔparA (346 cells), and double mutant ΔparAΔpapM (350 cells) strain, compared to wild-type strain (WT) (482 cells). (D) Cell length distribution in papM overexpressing strain in wild-type (433 cells) or ΔparA background (392 cells) as compared to controls with empty pMVp_ami vector in the wild-type (488 cells) or ΔparA background (351 cells). The data come from at least two independent biological replicates. The statistical significance between strains determined by Student’s t-test (two-sided) with Holm method used for multiple comparisons is marked with asterisks: P-values ≤0.05 (*), ≤0.01 (**), ≤0.001 (***), and ≤0.0001 (****).

Fig 4

papM deletion alters ParA dynamics. (A) Kymographs showing cumulative EGFP-ParA fluorescence intensity during the cell cycle in the control strain (ΔparA + pMVp_nat egfp-parA, 30 cells analyzed) and in the ΔpapM background (ΔparAΔpapM + pMV306p_nat egfp-parA, 25 cells analyzed). The time zero is the beginning of the new cell cycle detected as the visible separation of daughter cells accompanied by EGFP fluorescence at the new poles. Mid-cell is marked with a black line, and the ParA assembly at the mid-cell in the control strain is marked with red ellipse. The data come from two independent biological replicates. (B) The bubble plot showing the percentages of mobile and immobile PAmCherry-ParA molecules in control (ΔparA + pMVp_nat PAmcherry-parA) and ΔpapM (ΔparAΔpapM + pMV306p_nat PAmchery-parA) strain as determined by PALM. The number of tracks analyzed is 1,166 for control (49 cells) and 590 tracks (42 cells) for ΔpapM. The diffusion coefficients are indicated.

Fig 5

Polar accumulation of mCherry-DivIVA is altered in ΔparA but not ΔpapM M. smegmatis strain as compared to the wild-type control. The analyses were performed in WT control strain (400 cells), ΔpapM (398 cells) and ΔparA (400 cells) strains producing mCherry-DivIVA (apart of the wild-type DivIVA). (A) The ratio of mCherry-DivIVA fluorescence intensity at the old pole to average fluorescence intensity in the mid-cell. (B) The ratio of fluorescence intensity at the new pole to average fluorescence intensity in the mid-cell. (C) The ratio of mCherry-DivIVA fluorescence intensity at the old cell pole to its fluorescence at the new cell pole. (D) Representative images of the cells of each studied M. smegmatis strain. The mCherry-DivIVA fluorescence (red) is merged with the brightfield image. Scale bar, 5 µm. The data come from at least two independent biological replicates. The statistical significance between strains determined by Wilcoxon test (two-sided) with Holm method used for multiple comparisons is marked with asterisks: P-values ≤0.05 (*), ≤0.01 (**), and ≤0.001 (***).

Fig 6

The growth rate and mCherry-DivIVA redistribution during the cell cycle are affected by parA and papM deletion. (A) The cell growth rate measured as the cell length increment (the difference of the cell length of the newly born mother cell and the daughter cell at the time preceding cell division - the mid-cell mCherry-DivIVA signal appearance) divided by cell cycle length shown in B (103, 90, and 89 cells analyzed for WT, ΔpapM, and ΔparA, respectively). (B) The cell cycle length - the time elapsed between the appearance of the mid-cell mCherry-DivIVA signal in the mother cell and the appearance of mid-cell mCherry-DivIVA signal in its daughter cell (the number of cells analyzed as in A). (C) The time between the detection of the mid-cell mCherry-DivIVA signal and detection of the separated cell poles (84, 78, and 65 cells analyzed for WT ΔpapM and ΔparA, respectively). The analyses were performed in the WT control strain and ΔpapM and ΔparA strains producing mCherry-DivIVA (apart of the wild-type DivIVA). (D) The percentage of the cells with the visible mCherry-DivIVA signal in the snapshot images (487, 573, and 241 cells analyzed for WT, ΔpapM, and ΔparA, respectively). The data come from two independent biological replicates. The statistical significance between strains determined by Student’s t-test (two-sided) with Holm method used for multiple comparisons is marked with asterisks: P-values ≤0.05 (*), ≤0.01 (**), ≤0.001 (***), and ≤0.0001 (****).

Fig 7

The model of ParA-PapM-DivIVA interplay. (A) The scheme of M. smegmatis cell cycle showing ParA and DivIVA dynamics. (B) Model of ParM-ParA-DivIVA interactions. Black arrows are established relations, gray arrows indicate postulated relations, blue arrows indicate the relations suggested by the mutant M. smegmatis strain phenotypic analyses presented here, and yellow arrows illustrate conclusions based on in E. coli colocalization experiment.