Bacterial Physiology and Virulence: The Cultures Converge
Close up view of the ppGpp binding site of the
RNAP/DksA/ppGpp complex. Source.
by Fred Neidhardt
Growth dominates the attention of many bacteriologists. It has done so for over a century, inspiring explorations into the complex biochemistry and physiology that produce new cells able to grow, survive harsh environments, and live to grow another day.
Likewise, since the earliest days of microbiology, virulence has been a central focus. In fact, studies of how bacteria cause disease have in sheer number dominated the field for the simple reason that more than intellectual curiosity has been involved: human health has demanded that one learn to cure infectious diseases and protect against them.
Until recently, researchers in these two arenas of microbial exploration shared precious little beyond basic technology and a knowledge of bacterial cell structure and function. Separate scientific cultures developed, as is so often the case in human endeavors. Not uncommonly, investigations of infectious disease proceeded largely in medical school departments of internal medicine or pediatrics, while explorations of the intricacies of microbial growth processes were pursued at the same schools in basic science microbiology departments. That situation has been changing in the past couple of decades, and finally the frontiers of bacterial physiology and of virulence (molecular pathogenesis) have virtually fused. An international symposium (Metabolism Meets Virulence) held in Höhenkammer (Germany) in April, 2009 helped signal this watershed. A scan of the topics covered [Ref 1] reveals some of the subjects in which these two areas of research have become intertwined. Today, each subject benefits from attention to the other.
Many examples could be used to illustrate this sea-change, but none is more fully documented than the intimate involvement of the bacterial stringent control system in Legionnaire’s disease of humans. The central player in the stringent control system is ppGpp (guanosine 5’diphosphate, 3’diphosphate), a nucleotide long known as a major governor of bacterial growth processes. A recent review [LINK] co-authored by Michael Cashel (who in Jonathan Gallant’s laboratory discovered this “magic spot” [LINK and LINK]) brings the ppGpp story up to date. This nucleotide alarmone (as it is frequently dubbed) and its helper protein, DskA, have now been shown to orchestrate the complex alternation between two differentiated intracellular forms of the bacterium Legionella pneumophila, a process required for its pathogenicity.
For those interested in the direction of current microbiological research, this tale of ppGpp in disease is enthralling. Here we shall concentrate on Legionella virulence, but readers are directed to a recent review [LINK] that presents in scholarly detail the known involvement of ppGpp in many infectious diseases—and suggests the likely extension to many others.
The following account is a brief summary by one who has been a student of growth and ppGpp regulatory effects for four decades. To ease this author’s conscience for not mentioning all the talented researchers who have explored the stringent system or pioneered studies of Legionella virulence, I express my regret at not paying due respect to these fellow adventurers. Their work is the basis of our story, and I trust that the two recent comprehensive reviews [LINK and LINK] of these areas will be consulted to learn their names and their contributions to this community effort.
a. The ppGpp Regulatory System
In what sense is ppGpp a master regulator of bacterial growth?
There are two answers to this question, and they concern two perhaps related physiological activities of bacteria: response to stress, and growth rate control. The first answer is that ppGpp at moderate to high concentration triggers some of the specific responses of bacteria to environmentally imposed stress (such as for carbon and energy, or for a required nutrient), as well as the general response bacteria make when they can no longer grow. ppGpp initiates the transformation from growing, log-phase cells to non-growing, stationary-phase cells. These are some of the most thoroughly understood roles for ppGpp. The second answer we’ll come to later.
Free-living bacterial cells possess scores of specific stress response systems that are called into play when particular toxic situations are encountered, or when the supply of required nutrients limits growth. Changes in pH, temperature, salt concentration, redox state, UV or ionizing radiation, and many other stresses are countered by synthesis or activation of a variety of protective proteins and processes. (A Gordon Research Conference on Microbial Stress Responses has met biannually for decades to discuss the latest news about these response networks.) Many (but not all) of these systems are affected by, or directly involve the production of ppGpp, which then facilitates or potentiates the cell’s response. Moreover, when growth is no longer possible and the ultimate stress response is elicited—the transition to stationary phase—ppGpp is absolutely required.
Stationary phase cells differ greatly from growing ones. [Ref 2 and 3] They look different, and they act differently. Stationary cells are smaller and tougher, thus harder to break open by physical agents such as freeze-thaw cycles or grinding with alumina, or by treatment with penicillin-like antibiotics or chaotropic agents. Cylindrical cells, such as Escherichia coli, become more spherical in stationary phase, their cytosol and DNA nucleoid become condensed and their periplasm expanded. Their outer membrane, wall, and cell membrane take on altered chemical compositions, accounting for the increased toughness of the cells. Metabolism proceeds differently in stationary phase cells: protein and stable RNA are turned over more quickly; some ribosomes are degraded, others sequestered as inactive 100S dimers. The pathways of central metabolism are re-directed in ways that increase efficient use of the products of protein and RNA breakdown as well as residual metabolic byproducts (such as acetate) secreted during growth that now re-enter the cell.
These new products and activities require the activation of hundreds of genes that encode the proteins responsible. Note however, that while some of these proteins are the very same ones elicited by specific stresses, they are made here in cells that have not been exposed to those individual stresses. Clearly, a unique process produces the non-growing, stationary cell. A complex cascade of regulators led in large measure by the RNA polymerase sigma factor called σS (or sigma-S, an alternate to the major sigma factor functioning during growth, σ70) brings about the transformation to the refractory, non-growth state. Dozens of stationary phase genes are transcribed by RNA polymerase programmed by σS. Moreover, the products of some of these genes are themselves transcription factors that direct RNA polymerase to still other stationary phase genes. As a result, a total make-over of the bacterial cell occurs.
In a nutshell, rapid synthesis of σS leads to replacement of σ70, thereby initiating the differentiation leading to the stationary cell. And this is where ppGpp functions big time: ppGpp both promotes both rapid synthesis of σS [LINK and LINK] and its displacement of σ70 from core RNA polymerase [LINK]. As a result, ppGpp triggers the transformation to stationary phase. Cells lacking the ability to synthesize ppGpp are locked in growth mode and cannot transform into the stationary form [LINK].
The second reason for calling ppGpp a master regulator is the great likelihood that it governs the structure and composition of exponentially growing cells, tying properties such as cell size and ribosome content to the growth rate. The basal levels of ppGpp vary inversely with steady state growth rates [LINK] in accord with such a modulatory role. Nonetheless, using mutant strains that completely lack ppGpp to test this possibility turns out not to be easy. The difficulty arises from the rapid appearance of suppressor mutations in RNA polymerase in these cells. The suppressor mutations possibly affect the interaction of polymerase with the promoters of rRNA operons, because different laboratories [LINK and LINK] have obtained conflicting results with these ppGpp0 strains, and therefore the issue of growth rate control by ppGpp has remained unsettled for many years. Recent work, however, which successfully avoids the confusion caused by RNA polymerase mutants, provides strong evidence that growth rate control does not occur in the absence of ppGpp (Potrykus, K., H. Murphy, N. Philippe, and M. Cashel. 2010. ppGpp is the major source of growth rate control in E. coli. Environmental Microbiology. In press.).
How does metabolic stress generate a ppGpp signal?
Figure 1: Domain structures of bacterial enzymes that produce and
degrade ppGpp. N and C designate the amino (N) and carboxy (C)
ends of the proteins. Enzymatic domains synthesize (Synthetase)
and hydrolyze (Hydrolase) ppGpp. Regulatory domains (TGS and
ACT) interact with molecules that control the enzymatic activity of
these proteins. RSH and SpoT are bifunctional enzymes, but the
hydrolase region of RelA proteins is inactive. RelP, RelQ, and
RelV are small synthetase fragments found so far only in a few
A metabolic regulator of this significance must be made with alacrity in response to cell stress, and must also be removed quickly when it is appropriate for growth to resume. Both processes are handled throughout the bacterial world by a super family of enzymes collectively designated RSH (for RelA/SpoT Homologue). The name is derived from the fact that in E. coli, where the stringent system was discovered and subsequently elucidated, ppGpp is synthesized not only by the monofunctional enzyme RelA, but also by the bifunctional enzyme SpoT. Both enzymes use ATP to synthesize ppGpp from GDP (or pppGpp from GTP), but only SpoT can also hydrolyze the two alarmones to PP and GDP (or GTP). The domains of the enzymes that make and degrade ppGpp are shown in [Figure 1]. Many bacteria (e.g. some that are Gram-positive) contain one or more small gene fragments (relP, relQ, relV) that encode only an active ppGpp synthetase domain [LINK]; the stress conditions to which their protein products (RelP, RelQ, and RelV) respond are unknown.
This might be a good place to recognize that ppGpp is not solely a bacterial regulator. Genes for RSH enzymes are found throughout the plant world, encoded among the nuclear genes but with their product enzymes present within the chloroplasts. Studies indicate a necessary role for the ppGpp system in seed production and plant fertilization [LINK].
Figure 2: Diagram of the bacterial stringent response system. Two
parallel pathways synthesize pppGpp (which is subsequently con-
verted into ppGpp) from ATP and GTP in response to any of several
stress signals. Through interactions with RNA polymerase (RNAP),
ppGpp alters much of metabolism (– indicates inhibition; + indicates
Why should there be one bifunctional enzyme in some bacteria (and plants), and two enzymes (one monofunctional and one bifunctional) in E. coli and some other organisms? The answer is not known, but it may reflect how different organisms receive signals from the environment and sense metabolic stress. Being such a potent molecule, rapid and precise adjustment of the cellular level of ppGpp is absolutely critical. In E. coli, RelA molecules reside on a small fraction of ribosomes and synthesize ppGpp when uncharged tRNA accumulates during amino acid starvation; in contrast, SpoT is located elsewhere in the cytosol and responds to many stresses, including starvation for carbon and energy, iron, phosphate, and fatty acids [LINK] [Figure 2]. Little is known about the function of Rel P, Q, and V.
For our discussion of Legionnaire’s disease, to follow in a moment, we note that, like E. coli, L. pneumophila has both enzymes (RelA and SpoT).
How does ppGpp exert its manifold effects?
Figure 3. Models of how ppGpp might inhibit and activate RNA
polymerase (RNAP). (a) Direct inhibition by ppGpp and DksA
of transcription from ribosomal RNA promoters. (b) Direct
activation of transcription from amino acid biosynthesis
promoters. (c) Inhibition by ppGpp and possibly DksA of
transcription from ribosomal promoters, either (left path) by
favoring core RNAP programming by alternative sigma factors
(σS and σH ) or (right path) by σN perhaps involving some
factor yet to be discovered. Source.
Nature seems to have favored ppGpp with a chameleon-like character, allowing it to act in some instances as a direct inhibitor of transcription, in others as an activator, and in still other cases as an indirect agent that reprograms RNA polymerase by favoring the use of alternate sigma factors. The genetic evidence that ppGpp interacts directly with RNA polymerase has been confirmed by cross-linking ppGpp analogs to the polymerase [LINK and LINK]) and by co-crystallizing the enzyme with ppGpp [LINK]. All results indicate that ppGpp binds, perhaps at more than one site, where it interacts with the β and β' subunits of the polymerase. How this interaction brings about diverse and opposite effects on transcription initiation is related to the fact that promoters that are to be activated by ppGpp are differently structured (with AT-rich segments) than those that are to be inhibited (with correspondingly GC-rich segments). General re-programming of RNA polymerase with σS is achieved by fostering its synthesis over that of σ70 [LINK and LINK], and perhaps by a specific interaction at RNA polymerase itself that favors σS binding over that of σ70 [LINK].
Nothing is simple in biological regulation (life is too interesting for simplicity and biologists have come to expect complexity), and so one need not be surprised that ppGpp does not act alone. A protein called DksA is also an important player in the stringent regulatory system, one needed to stimulate the accumulation and function of σS during the early transition to stationary phase and to serve as a vital accessory factor in general for ppGpp. But this protein does even more. DksA also can substitute for ppGpp in some activities, and even have an opposite effect in others [LINK].
These various modes of ppGpp action are illustrated in Figure 3.
By these many mechanisms, ppGpp effects a vast change in the transcription pattern of E. coli [LINK].
b. Legionnaire’s Disease
What sort of disease would target veterans?
The answer is: the “opportunistic pathogen” called Legionella pneumophila. In 1976, at the Bellevue-Stratford Hotel in Philadelphia, PA, this organism caused an outbreak of a pneumonia-like illness that felled well over 200 people, 34 of whom died. Most of the dead were armed service veterans who as members of the American Legion had gathered at the hotel to celebrate the nation’s bicentennial. Long and intensive epidemiologic study finally indicted the culprit: L. pneumophila, a Gram-negative bacterium common in aquatic environments but hitherto unknown as a human pathogen. L. pneumophila has existed in nature for millions of years, growing slowly as surface biofilms within fresh water, and also as a parasite of freshwater amoebae (LINK). Unfortunately our modern cooling towers, humidifiers, and misters produce aerosols that can carry significant numbers of L. pneumophila cells into our lungs. Once having gained entrance, the bacteria multiply within macrophages and then travel from cell to cell using the same strategies as when infecting amoebae. People who are immunocompromised, even if only by virtue of their age, are at risk of contracting Legionnaire’s disease. (Presumably the age of the veterans celebrating at the Bellevue-Stratford that hot July contributed to their susceptibility to the aerosols delivered by the hotel's air conditioning system.)
How does the alternation of differentiated forms function during infection?
Figure 4. Model of how SpoT governs the life cycle of Legionella
pneumophila in macrophages. Beginning at the upper left, ppGpp
triggers the transformation of the replicative form into the
highly resilient, motile, transmissive form. After being phago-
cytized, the transmissive bacteria inhibit fusion with lysosomes
(small, dashed empty vacuole). Favorable growth conditions
within a vacuole formed from the endoplasmic reticulum (ER)
leads to hydrolysis of ppGpp by the bacterial SpoT enzyme, and
transformation into the replicative form ensues. Bacterial
replication eventually is restricted within the ER-derived vacuole,
causing SpoT to synthesize ppGpp, which triggers transformation
of the Legionella back to the transmissive form. The transmissive
cells resist lysosomal degradation and migrate to new macrophages.
This successful parasite employs a dramatic strategy: the organisms alternate between two differentiated forms, a thin-walled, non-motile replicative form and a thick-walled, motile, infectious transmissive form. By being more resilient and resistant to stress than the replicative form, the transmissive form resembles E. coli cells in stationary phase, but it has its own peculiarities [LINK]. Once engulfed by a macrophage, it secretes proteins (via the Dot/lcm type IV secretion system) that help establish a vacuole that protects it from lysosomal degradation. Here conditions are favorable for growth, calling for transition from the transmissive to the replicative form. Genes for the transmissive phase are repressed by the mRNA-binding protein Csr and by the sRNA chaperone Hfq. The entire flow of metabolism is shifted toward growth and cell replication [marvelously displayed by a total transcriptional analysis here]. Multiplication of L. pneumophila cells, now differentiated into their replicative form, proceeds until eventually nutrient supplies within the phagocyte are exhausted. At this point, the progeny of the original invading bacterium again alter their metabolism and transcription pattern drastically, now returning to the traits characteristic of the transmissive phase. Eventually the host macrophage is lysed, thereby releasing transmissive phase L. pneumophila able to infect new macrophages [Figure 4].
What is the role of the stringent control system in differentiation?
The transition from the replicative to the transmissive form is initiated when growth of L. pneumophila in macrophages (or in broth cultures) is restricted, thus triggering the massive transcriptional response needed to develop the new suite of structural features and activities. This cascade of regulatory events is orchestrated in large measure by the ppGpp stringent control system. In fact, the correspondence between the triggering of L. pneumophila differentiation within the phagocyte and the triggering of E. coil to enter stationary phase in broth cultures is uncanny. All the familiar players are in action. Both RelA and SpoT stand guard to monitor the environment within the vacuoles where the replicative form of the bacteria are growing. At the first hints of nutrient restriction, ppGpp is synthesized. If the deficit concerns amino acid availability, presumably RelA provides the alarm signal. But good evidence in broth cultures shows that SpoT also plays a role; for example, when an appropriate supply of fatty acids is not forthcoming, ACP (Acyl-Carrier Protein) interacts with SpoT, which then synthesizes ppGpp, just as happens in E. coli; this might also happen inside the vacuole, since SpoT is essential for formation of the transmissive form in macrophages. In addition, under these conditions a DksA homologue (72% identical to that of E. coli) is critical for L. pneumophila growing in broth to differentiate to the transmissive form including flagellar gene activation, evasion of lysosomes, and cytotoxicity toward macrophages. The relationship between DksA and ppGpp is not simple; rather, the two can act cooperatively as well as independently, depending on the context. For example, ppGpp seems to be essential for differentiation of the bacteria within the macrophage to become transmissive to new cells, while DskA is dispensable. In broth cultures, however, DksA is essential for flagellar morphogenesis. And in many situations DksA cooperates with ppGpp responding to the level of ppGpp to stimulate expression of particular genes involved in transmissive cell traits. This situation mirrors the complex interactions of DksA and ppGpp in E. coli [LINK].
How ppGpp and DskA exert their effect is still under study. This pair seem to have both direct and indirect effects on many activators and repressors of genes related to differentiation to the transmissive form, but the exact molecular biology of their interaction is still under study. At any rate, a cascade of regulatory events is initiated when growth in macrophages is restricted, and a massive transcriptional response produces enzymes that develop the structural features and activities characteristic of the transmissive form of L. pneumophila. The return to the replicative form in the newly invaded cell would presumably resemble the scenario described for the original invasion of fresh macrophages upon infection, commencing with a lowered level of ppGpp. The regulatory proteins involved in this cascade are numerous; a review [LINK] summarizes current information about these L. pneumophila-specific virulence factors.
Our interest lies in the fact that the cascade is initiated by the same master system that governs growth of most—perhaps all—bacterial cells.
Retrospection enjoys advantages over other thought processes. Looking backward produces, if not simpler, at least more obvious explanations and rationales for what at one time were puzzles. And so it appears with ppGpp (“magic spot”) and virulence. How could it be otherwise—a highly developed mechanism that not only enables fast growing bacteria to hunker down and protect themselves during adverse times but also permits a reversal to their growing state—would surely be a prime starting point for evolution to craft a successful intracellular parasite.
Or, do we have the story backwards?
- Nova Acta Leopoldina. (July, 2010) Metabolism meets virulence. 111 (378).
- Henge-Aronis, R. (1996) Chp 93 in E. coli and Salmonella, ASM Press.
- Huisman et al. (1996) Chp 106 in E. coli and Salmonella, ASM Press.
Fred Neidhardt is F.G. Novy Distinguished University Professor, Emeritus, Department of Microbiology and Immunology, University of Michigan Medical School at Ann Arbor .
Dalebroux ZD, Svensson SL, Gaynor EC, & Swanson MS (2010). ppGpp conjures bacterial virulence. Microbiology and molecular biology reviews : MMBR, 74 (2), 171-99 PMID: 20508246