Computational and genetic reduction of a cell cycle to its simplest, primordial components

Abstract

What are the minimal requirements to sustain an asymmetric cell cycle? Here we use mathematical modelling and forward genetics to reduce an asymmetric cell cycle to its simplest, primordial components. In the Alphaproteobacterium Caulobacter crescentus, cell cycle progression is believed to be controlled by a cyclical genetic circuit comprising four essential master regulators. Unexpectedly, our in silico modelling predicted that one of these regulators, GcrA, is in fact dispensable. We confirmed this experimentally, finding that ΔgcrA cells are viable, but slow-growing and elongated, with the latter mostly due to an insufficiency of a key cell division protein. Furthermore, suppressor analysis showed that another cell cycle regulator, the methyltransferase CcrM, is similarly dispensable with simultaneous gcrA/ccrM disruption ameliorating the cytokinetic and growth defect of ΔgcrA cells. Within the Alphaproteobacteria, gcrA and ccrM are consistently present or absent together, rather than either gene being present alone, suggesting that gcrA/ccrM constitutes an independent, dispensable genetic module. Together our approaches unveil the essential elements of a primordial asymmetric cell cycle that should help illuminate more complex cell cycles.

Figure 1. Minimal model of Caulobacter crescentus cell cycle.

(A) Schematic of the cell cycle. Localization of (total) CtrA and phosphorylated CckA proteins indicated. (B) Circuit diagram of the mathematical model, reduced from the biological model shown in Figure 5B. Methylation and compartmentalisation are discrete events affecting ctrA transcription and activity of the CckA phosphorelay, respectively. (C) Mathematical description of model: ordinary differential equations and discrete events. See Text S1 for parameter values and justification.

Figure 2. ΔgcrA::Ω mutant cells are viable but have growth and morphological defects.

(A) Immunoblots showing GcrA, CtrA, and PilA (reporter for compartmentalization) steady-state levels for WT cells grown in M2G. (B) Simulated protein levels (solid lines) of GcrA and CtrA in WT cells, averaged over two compartments/cells where appropriate and incorporating imperfect cell cycle synchrony (see Text S1). Times of simulated events are indicated by an arrow (DNA replication initiation), an arrowhead (SW to ST differentiation of the SW daughter cell), a dotted line (ctrA P₁ hemi-methylation), and a dashed line (compartmentalisation). CtrA and GcrA relative protein quantifications from immunoblots are the mean of three biological replicates; error bars are data ranges. Both datasets normalized to maximum ST/PD value. See Figure S1 for simulated profiles in ST and SW compartments and for ctrA promoter expression. (C) Time course of colony appearance following ΦCr30-mediated generalized transduction with ΔgcrA::Ω lysates from a ΔgcrA::Ω; xylX::P_xyl-gcrA donor strain. Error bars are standard deviation from three biological replicates. Transductants scored on PYE (left panel, blue curves) or M2G (right panel, red curves) media supplemented with spectinomycin (30 µg/ml) and streptomycin (5 µg/ml) to select for ΔgcrA::Ω transduction. Table (right) shows conditions or strains used and number key for corresponding curves. Column (far right) shows whether ΔgcrA::Ω transducing lysate was added to cells or media. Total number of Spc^R clones obtained after transduction of cells containing the pMT335gcrA plasmid (7 and 10) reflects the efficiency of transduction of the ΔgcrA::Ω marker. Relative to this value, 33% in PYE and 96% in M2G of NA1000 transduced cells give Spc^R clones, confirming that NA1000 ΔgcrA::Ω colonies are not due to suppressors. (D, E) Differential interference contrast (DIC) micrographs of cells grown in PYE (D) or M2G (E). Scale bar represents 2 µm. Value above micrograph shows doubling time (with standard deviation from at least three biological replicates) for each strain. (F) Relative abundance of replication origin versus terminus (Cori/ter ratio) in WT and mutant cells grown in PYE (blue bars) or M2G (red bars). Ratios normalized to WT value. Triplicate measurements from two independent DNA extractions. Error bars are standard deviation. (G) Immunoblots showing steady-state levels of various proteins in WT and mutant cells in M2G. Lowest row shows levels of acid-extracted RsaA protein in M2G detected by Coomassie Brilliant Blue staining.

Figure 3. ΔgcrA::Ω cells suffer from insufficiency in FtsN.

(A) Position of eight himar1 transposon (Tn) insertions (yellow arrowheads) that ameliorate slow growth of ΔgcrA::Ω mutant cells on PYE. (B) Trace of N6-methyladenosine (m6A) marked DNA at ftsN locus of WT cells from ChIP-Seq experiment performed with antibodies to m6A. Figure reports number of times given nucleotide position sequenced from chromatin precipitates. Stars show predicted positions of GAnTC methylation sites at ftsN locus. (C) Occupancy of GcrA across ftsN locus as determined by ChIP-Seq using polyclonal antibodies to GcrA. Occupancy expressed as percentage of total reads. Chromosome coordinates below chart in (C) match scheme shown in (A). In (A–C) grey-shaded region denotes ftsN promoter fragment (P_ftsN) used in promoter probe experiments shown in (G, H). (D) DIC image of ΔgcrB ΔgcrA::Ω cells harbouring P_ftsN::Tn2 insertion [depicted in (A)] grown in PYE. Scale bar represents 2 µm. Values above micrograph indicate doubling time (with standard deviation from three biological replicates) and, in brackets, the number of days after which the majority of colonies are visible. (E) Relative Cori/ter ratio in ΔgcrB ΔgcrA::Ω cells harbouring either P_ftsN::Tn1 or P_ftsN::Tn2 insertion. Cells grown in PYE (blue bars) or M2G (red bars). Ratios normalized to WT value. Triplicate measurements from two independent DNA extractions. Error bars are standard deviation. (F) Immunoblots showing steady-state levels of FtsN in WT, gcrA mutants, and derivatives harbouring different Tn insertions at ftsN locus in PYE. (G) P_ftsN- or P_{ftsN *}-lacZ (GAnTC mutant) activity after depletion of GcrA for 5 h (white) or 24 h (red) in ΔgcrB ΔgcrA::Ω xylX::P_xyl-gcrA cells grown in M2GX (M2G containing 0.3% xylose) or depleted by washing and cultivation in M2G. At 24 h of GcrA depletion, the absolute levels of β-galactosidase are 637±8 Miller units for P_ftsN and 608±6 Miller units for P_ftsN*. (H) β-galactosidase activity of P_ftsN- or P_{ftsN *}-lacZ fusions in NA1000 cells grown in PYE (blue) or M2G (red), values below histograms. In (G, H) data are from four independent experiments; error bars are standard deviation.

Figure 4. Dispensability and genetic interactions of gcrA-ccrM regulatory module.

(A) Immunoblots showing abundance of various proteins in WT and mutant strains in PYE. FtsN relative protein quantifications were normalized to the NA1000 sample value and represent the average and standard deviation of three independent experiments. Schematic shows Tn position in ccrM::Tn allele. (B) Representative DIC micrographs of gcrA and ccrM mutant cells grown in PYE. Leftmost figure reproduced from Figure 2D. Scale bar represents 2 µm. Arrowheads indicate stalks. Values above micrograph show doubling time for each strain (with standard deviation from at least three biological replicates) and, in brackets, the number of days after which the majority of colonies are visible. (C) Relative Cori/ter ratio in mutant strains grown in PYE (blue bars) or M2G (red bars). Ratios normalized to WT value. Triplicate measurements from two independent DNA extractions. Error bars are standard deviation. (D) Tn insertion bias in coding sequences (CDS) of ΔgcrA::Ω cells relative to WT cells as determined by Tn-seq. Abscissa shows position as function of genome position, and ordinate gives insertion ratio. Peaks show CDSs with the highest number of Tn insertions. Noncoding sequences are not included (see Figure S7). (E) Inverse ratio shown compared to ratio in (D), with peaks indicating CDSs receiving fewest insertions in ΔgcrA::Ω cells relative to WT cells. (F) Simulated protein levels of CtrA in ΔgcrA (solid line) and ΔgcrA ΔccrM (dashed line) averaged as in Figure 2B. Both curves normalised to maximum PD value.

Figure 5. Coconservation of gcrA/ccrM suggests a primordial cell cycle regulatory circuit.

(A) gcrA and ccrM are generally coconserved in the Alphaproteobacteria (adapted from Figure 2 and Table S2 of [43]). One species of each genus analysed is shown (37 out of a total of 65 genomes). (B) Current biological model of cell cycle regulation in Caulobacter crescentus, with the dispensable GcrA/CcrM module highlighted. Question marks indicate elements of the circuit that have yet to be fully elucidated.