Inflating bacterial cells by increased protein synthesis - PubMed (original) (raw)

Inflating bacterial cells by increased protein synthesis

Markus Basan et al. Mol Syst Biol. 2015.

Abstract

Understanding how the homeostasis of cellular size and composition is accomplished by different organisms is an outstanding challenge in biology. For exponentially growing Escherichia coli cells, it is long known that the size of cells exhibits a strong positive relation with their growth rates in different nutrient conditions. Here, we characterized cell sizes in a set of orthogonal growth limitations. We report that cell size and mass exhibit positive or negative dependences with growth rate depending on the growth limitation applied. In particular, synthesizing large amounts of "useless" proteins led to an inversion of the canonical, positive relation, with slow growing cells enlarged 7- to 8-fold compared to cells growing at similar rates under nutrient limitation. Strikingly, this increase in cell size was accompanied by a 3- to 4-fold increase in cellular DNA content at slow growth, reaching up to an amount equivalent to ~8 chromosomes per cell. Despite drastic changes in cell mass and macromolecular composition, cellular dry mass density remained constant. Our findings reveal an important role of protein synthesis in cell division control.

Keywords: cell division; cell size; cell volume; cellular DNA; growth rate.

© 2015 The Authors. Published under the terms of the CC BY 4.0 license.

PubMed Disclaimer

Figures

Figure 1

Figure 1

Cell size and content under different growth limitations

  1. Snapshots of bacteria from different culture conditions at similar OD600 (˜0.4) and the same magnification: I. glucose (λ ≈ 0.98/h); II. mannose (λ ≈ 0.41/h); III. glucose + 8 μM Cm (λ ≈ 0.32/h); IV. LacZ OE, glucose + 15 ng/ml cTc (λ ≈ 0.25/h). Cultures under different growth limitations (II–IV) exhibit large differences in cell size at comparable growth rates.
  2. Normalized cell size distributions, as quantified by automated image analysis, for cells taken from the conditions described in panel (A). Distributions for cells grown in mannose, Cm, and LacZ OE were taken at comparable growth rates. Inset, density distributions for cell volume normalized by average cell size. When normalized by mean cell size, the different distributions appear very similar.
  3. Mean cell volume obtained under the different growth limitations plotted against the corresponding growth rate of the culture (see Appendix Table S2 for standard deviations and Appendix Table S3 for the variation between repeats and different OD600).
  4. Cellular dry mass plotted against the corresponding growth rate of the culture, for each growth perturbation. The trends of cellular dry mass closely resemble the trends exhibited by cell volume (panel C).
  5. Cellular RNA plotted against the corresponding growth rate of the culture, for each growth perturbation.
  6. DNA content per cell. The trends in DNA content, as confirmed by DAPI staining (FigEV3), also closely follow the change in cell size shown in panel (C) (see FigEV1C for the correlation plot).
  7. Cellular dry mass plotted against cell volume. A tight correlation exists between these quantities under all growth limitations.

Figure 2

Figure 2. The threshold initiation model of cell size control

  1. Schematic of the initiation model. This model assumes that a threshold amount of the cell division protein X per cell, formula image, is required to trigger cell division: When the abundance of the protein X reaches a threshold level (represented by the dashed line), the cell divides at this size (left, cell I). In the LacZ OE strain, LacZ “compresses” the proteome fractions of X and other GR-dependent proteins G, while a certain fraction of the proteome Q remains constant and cannot be reduced (Hui et al, 2015). Hence, with LacZ OE, a cell at the size of cell I would contain a smaller amount of protein X, reducing it below the threshold level formula image, needed for cell division (middle, cell II). The cell would continue to grow and eventually divide when the cellular abundance of protein X reaches this threshold level (right, cell III). Due to the smaller proteome fraction of X under LacZ OE, a much larger cell is produced (see Box 1 for a quantitative analysis).
  2. Inverse of the average cellular protein content, 1/〈_P_〉, versus the growth rate for the different growth limitations; same symbols as in Fig1. According to the threshold initiation model, the plotted quantity reflects the growth rate dependence of the abundance of the cell division protein X under each mode of growth limitation (see Box 1).

Figure EV1

Figure EV1

Protein/RNA/DNA-volume correlation Symbols same as in Fig1.

  1. A, B Total cellular protein (A) and total cellular RNA (B) determined from Appendix Table S4, plotted against cell volume determined via microscopy, for each growth perturbation. The resulting direct proportionalities indicate that cell size closely corresponds to cellular protein under all growth limitations. This is related to the close correspondence between cell size and cellular dry weight (Fig2E), because cellular protein is the main constituent of cellular dry mass (Appendix Fig S6).
  2. C Cellular DNA plotted against cell volume determined via microscopy, for each growth perturbation. Remarkably, cellular DNA content closely follows the trends exhibited by cell size under all perturbations.

Figure EV2

Figure EV2. Cell counts from Coulter counter and plating

Cell counts for the three types of growth limitations were obtained by using a Coulter counter on the one hand and plating on the other hand. In general, there is a very good agreement between the results of the Coulter counter and plating counts. For LacZ overexpression (OE) in glucose minimal medium, plating (red asterisks) resulted in only slightly lower values (10%) than the Coulter counter method (filled red diamonds), demonstrating that almost all cells in the culture are viable even at large cell sizes. For the nutrient limitation series, the results of the two methods give good agreement except for the very slow growth conditions (GR < 0.7 h−1), where the Coulter counter method (green filled circles) resulted in a lower cell number as compared to plating (green asterisks). This results from the small cell sizes under these conditions, which are close to the detection limit of the Coulter counter. For translation inhibition with Cm, the values of plating (blue asterisks) are lower compared to those obtained from the Coulter counter method (filled blue triangles) by 20%. This indicates that a fraction of cells in these conditions is not viable when subjected to plating. Throughout this manuscript, we use the cell count in order to determine per cell quantities. In these calculations, due to the small deviations discussed above, for the actual cell count, we employ the values of plating for nutrient limitation at slow growth rates, but use the values of the Coulter counter methods for both translation inhibition and LacZ overexpression.

Figure EV3

Figure EV3

Cellular DNA content from DAPI staining Symbols same as in Fig1.

  1. Cellular DNA content measured by DAPI staining and microscopy plotted against growth rate. The results confirm the data for cellular DNA content (Fig2F). Error bars represent the standard deviations of the measured distributions over the cell population (n = 20).
  2. Cellular DNA content determined from DAPI staining versus DNA content determined from biochemical assay (see Materials and Methods section). The two methods of determining cellular DNA content show good agreement.
  3. Microscopy images of DAPI-stained DNA and cells. Cellular DNA is localized at single nucleoid foci per cell. This suggests that even in large cells with high cellular DNA content resulting from LacZ OE, only one completely replicated chromosome per cell is present. Therefore, the high cellular DNA content likely results from multiple incomplete replication forks. Moreover, these images demonstrate that cell division proceeds normally in the large cells resulting from LacZ OE (e.g., right image, upper left corner) and the large cell size is not caused by inhibition of cell division as reported under other conditions, for example, by Zaritsky et al (2006).

Figure EV4

Figure EV4

Relative abundances of exemplary protein levels under the three modes of growth perturbations Symbols same as in Fig1. In the threshold initiator model of cell division (Fig2A), changes in cell size are reflected by changes in the abundance of the regulatory protein X, ϕ X. The growth rate dependence of ϕ X required by our model to produce the observed cell sizes are presented in Fig2B. Here, we present the relative abundance of some exemplary proteins, as measured in a recent proteomic study (Hui et al, 2015), that bear some resemblance to ϕ X for each mode of growth perturbation. The abundances are normalized to that in the standard glucose condition.

  1. FtsZ, an essential protein that forms the Z-ring at the site of cell division.
  2. The DNA-associated protein, IHF, has a matching profile (IHF was not detected in the LacZ OE series, but we know that most proteins exhibit a linear decrease under this perturbation; see Appendix Fig S4). The use of a chromosome-associated protein for cell division control would be an effective strategy to coordinate cell division with the state of chromosome, as has been discussed in the context of eukaryotic cell division (Skotheim, in submission).
  3. HU, a DNA-binding protein.
  4. SSB, an essential DNA-binding protein required for DNA replication.

Figure EV5

Figure EV5. Chemiosmosis model of biomass–water homeostasis

Schematic illustration of the proposed chemiosmosis model of biomass–water homeostasis. The observed tight correlation between cellular dry mass (or protein) and cell size (Figs 1E and EV1A) under the different limitations holds even under growth perturbations like LacZ OE, which severely affects the canonical dependence of cell size on growth rate. Since cell size is defined by the sum of cell dry mass and water content, this means the cell keeps nearly a constant ratio of the cellular dry mass and water content. This remarkable coordination of water and dry mass content could be mediated by a homeostatic feedback loop illustrated schematically in this panel: A higher density of charged macromolecules (RNA and proteins) leads to the accumulation of counter ions (e.g., potassium ions and glutamate) via the Gibbs–Donnan effect (Chang, 2005) and thereby results in an increase in intracellular osmolarity (indicated by the red ⊕, left panel). This increase in osmolarity drives water into the cytoplasm, which leads to cell size expansion (dashed line, middle panel), lowering the density of macromolecules back toward the homeostatic level (right panel). Such a mechanism could account for the coordination of cytoplasmic volume with macromolecular content, even under unexpected perturbations like LacZ OE, and is consistent with the recently reported effects of rapid changes of external osmolarity on cell growth (Rojas et al, 2014).

References

    1. Amir A. Cell size regulation in bacteria. Phys Rev Lett. 2014;112:208102.
    1. Bernander R, Stokke T, Boye E. Flow cytometry of bacterial cells: comparison between different flow cytometers and different DNA stains. Cytometry. 1998;31:29–36. - PubMed
    1. Bipatnath M, Dennis PP, Bremer H. Initiation and velocity of chromosome replication in Escherichia coli B/r and K-12. J Bacteriol. 1998;180:265–273. - PMC - PubMed
    1. Boye E, Nordström K. Coupling the cell cycle to cell growth. EMBO Rep. 2003;4:757–760. - PMC - PubMed
    1. Campos M, Surovtsev IV, Kato S, Paintdakhi A, Beltran B, Ebmeier SE, Jacobs-Wagner C. A constant size extension drives bacterial cell size homeostasis. Cell. 2014;159:1433–1446. - PMC - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources