It has been assumed that bacteria adapt to nutrient limitation by adjusting the number of ribosomes, no matter what they are being starved for. Instead, two recent studies show that Escherichia coli uses different approaches depending on whether its growth is limited by the availability of carbon, nitrogen or phosphate.
Fundamentals of bacterial growth are deliciously complex and yet important because a thorough understanding of microbial growth may someday enable the engineering of microorganisms that could provide our planet with food, fuel and fibre. A long-standing hypothesis postulates that bacterial growth rates in various media (ranging from minutes to several hours) are determined by changing the numbers of highly efficient ribosomes (factories that produce proteins) per cell, rather than by varying the efficiency of a fixed number of ribosomes. The constraint for this hypothesis is that all the nutrients must be present in excess of the cell’s ability to use them, so that the cell can adapt whether rich or lean cuisines are available1,2.
It is also well known that guanosine pentaphosphate, (p)ppGpp, a guanine nucleotide analogue produced by cells in response to starvation for sources of carbon, nitrogen or phosphate and many other cellular stresses, plays a role in this phenomenon of growth-rate control, as it inhibits rRNA transcription and thereby limits formation of ribosomes3 (Fig. 1). Bacterial mutants devoid of this alarmone (ppGpp0 cells) respond differently than wild-type cells when growth is slowed by the presence of an excess of poorly utilized nutrients. In wild-type cells the ribosome level varies greatly, with high levels at fast growth rates and low levels at slow growth. By contrast, in ppGpp0 cells, the ribosome levels are high at all growth rates (for example, ten-fold higher than wild-type cells at very slow growth) and the excess ribosomes are apparently normal, that is, they do not seem to be defective4.
Two recent articles in Nature Microbiology5,6 describe what happens when growth is not limited by availability of poorly utilized nutrients, but instead limited by starvation for each of three classes of nutrients: sources of carbon, nitrogen or phosphate. The two groups provide evidence that cells regulate translation differently depending on which nutrient is being limited. These different strategies are intimately involved with complexities related to translation regulation and to transcription–translation coupling, which recently gained much attention7,8. Although the approaches used in the two studies differ, the results from each manuscript seem to reinforce each other.
Iyer et al. allow cells to grow until a limiting amount of a carbon source (glycerol) or nitrogen source (ammonia) is exhausted and study what happens during the first hour after growth stops (Fig. 1). They use a dual fluorescent hybridization probe technique applied to one gene (lacZ) that can detect how much mRNA is present per cell due to RNA polymerase transcription of DNA at the beginning (head) and at the end of that gene (tail). This gives a measure of the speed of transcription and reveals effects due to the presence of possible pauses or termination events that can prevent transcription of the whole gene. The efficiency of translation and speed of protein synthesis on mRNA by ribosomes was also estimated using the catalytic activity of the lacZ product, β-galactosidase. The results indicate that in wild-type cells, carbon starvation mostly affects the rate of transcription that is coupled to translation and thus slows down both processes. By contrast, nitrogen starvation slows translation rates independently of transcription. Interestingly, during nitrogen starvation, translation seems to be uncoupled from transcription and instead (p)ppGpp becomes necessary to coordinate both processes by an unknown mechanism, which appears to be dispensable during carbon starvation.
On the other hand, Li et al. use pumps (chemostats) to maintain a constant low level of a limiting nutrient (glucose, ammonia or phosphate), so growth occurs at different slow steady-state rates; these cells are partially starved but never stop growing (Fig. 1). The authors employ mathematical modelling and perform classical protein and RNA measurements to document RNA/protein ratios, and carry out ribosome profiling on sucrose gradients to estimate the total number of ribosomes and the fraction judged as functional. The results indicate that during steady-state glucose limitation, there is only a small fraction of engaged ribosomes and they operate at fast elongation rates. By contrast, during steady-state nitrogen limitation, the number of working ribosomes is a bit higher but elongation rates are slow. In both instances, there is a large reservoir of excess ribosomes not engaged in protein synthesis, which the authors argue would allow the cells to quickly adapt to better conditions by growing faster. In addition, Li et al. find that although during phosphate starvation there is a notable decrease in the overall ribosomal pool, a high fraction of ribosomes is active and works with fast elongation rates. With respect to regulation exerted by (p)ppGpp, based on studies with a ΔrelA strain (partially devoid of (p)ppGpp) under nitrogen-starvation conditions, they find the alarmone to be a key mediator of ribosomal activity modulating the transition from translational initiation to elongation.
Overall, both reports indicate that not only do different starvation regimens affect ribosomal content and activities differently, but surprisingly (p)ppGpp also plays different roles under each starvation condition. Both of these conclusions present a dilemma: the sum of individual C-, N- and P-starved states (that is, growth in an excess of poorly utilizable nutrients) is apparently not equal to its parts. At least two types of future experiments can be imagined that might help to resolve these issues. One type requires finding ways that ppGpp0 cells can be starved for single-nutrient sources to clearly define which processes truly require (p)ppGpp and which do not. This is not currently possible because these cells cannot survive under such conditions. The other type involves limiting growth in the presence of an excess of a combination of poorly utilizable sources of either carbon, nitrogen or phosphate and letting the cells decide how they wish to adapt, so it could be established whether or to which extent any adaptation mechanism overrides the other strategies.
How might an understanding of starvation and growth be useful in the future? The DNA–RNA–protein central dogma is based on standard information transfer. One path of current research is to focus on constructing alternative DNA–RNA–product information systems (the so-called orthogonal systems) that are relying on synthetic biology9. Special ribosomes already have been engineered to exclusively recognize modified mRNA codes and synthesize proteins with the use of modified transfer RNAs. For these systems to co-exist within normal cells, many alterations will be necessary. For one, the orthogonal ribosomes will have to possess special regulatory properties that allow them to accumulate and function independently of normal ribosomes. The papers discussed here clearly reveal that different starvation conditions may allow manipulation of very different ribosome populations with respect to both numbers and functions. Perhaps the two ribosome populations co-exist, one functioning according to normal information transfer and another according to an alternative code. The ability to selectively manipulate heterogeneous ribosome populations within a single cell is an exciting way to separately optimize sustaining functions from those needed to engineer efficient synthesis of novel products.
Schaechter, M., Maaloe, O. & Kjeldgaard, N. O. J. Gen. Microbiol. 19, 592–606 (1958).
Bremer, H. & Dennis, P. P. in Escherichia coli and Salmonella: Cellular and Molecular Biology (ed. Neidhardt, F. C) 1553–1569 (ASM Press, Washington, DC, 1996).
Potrykus, K. & Cashel, M. Annu. Rev. Microbiol. 62, 35–51 (2008).
Potrykus, K., Murphy, H., Philippe, N. & Cashel, M. Environ. Microbiol. 13, 563–575 (2011).
Iyer, S., Le, D., Park, B. R. & Kim, M. Nat. Microbiol. 3, 741–748 (2018).
Li, S. H.-J. et al. Nat. Microbiol. https://doi.org/10.1038/s41564-018-0199-2 (2018).
McGary, K. & Nudler, E. Curr. Opin. Microbiol. 16, 112–117 (2013).
Artsimovitch, I. Mol. Microbiol. 108, 467–472 (2018).
Liu, C. C., Jewett, M. C., Chin, J. W. & Voigt, C. A. Nat. Chem. Biol. 14, 103–106 (2018).
The authors declare no competing interests.
About this article
Cite this article
Potrykus, K., Cashel, M. Growth at best and worst of times. Nat Microbiol 3, 862–863 (2018). https://doi.org/10.1038/s41564-018-0207-6
Linking bacterial growth, survival, and multicellularity – small signaling molecules as triggers and drivers
Current Opinion in Microbiology (2020)
Frontiers in Microbiology (2020)
The Absence of (p)ppGpp Renders Initiation of Escherichia coli Chromosomal DNA Synthesis Independent of Growth Rates