by
|
by |
|
| This article will appear in a forthcoming issue of Trends in Microbiology. | |
Abstract
Bacterial geneticists dream of being able to unravel the regulatory networks by which a bacterial cell controls its growth. For any cell, optimal growth requires the coordination of metabolic processes with DNA replication and cell division. For bacteria with complex developmental transformations, growth must also be coordinated with these transformations [1,2]. Until recently, understanding bacterial growth and development required the painstaking isolation of regulatory mutants, followed by gene cloning and molecular characterization. Now, however, new post-genomic methodologies are accelerating the rate we can gain information about such complex processes. Detailed DNA-array and proteomic analyses of the cell cycle of Caulobacter crescentus now hint at the power of using such techniques in combination, and suggest that the long-held dream of readily mapping genetic networks could soon be fulfilled.
The choice of a tractable experimental system is as crucial to successful analysis of cell cycle regulation as the methods used. C. crescentus arguably provides the best experimental system to study the cell cycle because large quantities of cells can be synchronized by a simple method [3,4]. Caulobacter reproduce by asymmetric cell division, producing a motile swarmer cell and a sessile stalked cell (figure 1). Cell cycle progression in these cells is accompanied by an ordered series of developmental events that give rise to an asymmetric pre-divisional cell. Until recently, the expression pattern of cell cycle and developmental genes had to be analyzed one gene at a time. Work over the past 25 years has generated data on the cell cycle expression pattern of approximately 70 genes and/or proteins. These studies have led to the important conclusion that most of the genes required for the progression of the cell cycle and for development, and for the regulation of these events, are differentially expressed during the cell cycle. Furthermore, the expression of cell-cycle-regulated genes often occurs just before their time of action. After all this work, two recent articles describing the cell cycle expression patterns of 2,966 genes and 979 proteins provide a stunning proof of principle for the application of post-genomic analysis methods to the study of the bacterial cell cycle [5,6].
Microarray Analysis of Cell Cycle Gene Expression
In the first paper, Laub et al. used DNA microarrays to analyze the expression patterns of 90% of Caulobacter genes. Swarmer cells were isolated and allowed to proceed synchronously through the 150-minute cell cycle. RNA was isolated from cell aliquots taken every 15 minutes, and the level of each mRNA was measured using microarrays. The expression pattern of 72 genes previously shown to be cell cycle regulated was found to be essentially the same by microarray analysis, providing an impressive confirmation of the ability of microarrays to measure cell cycle expression of bacterial mRNAs. In total, 553 genes, representing 19% of the genome, produced RNA whose level changed substantially during the cell cycle. Among these were 247 genes of unknown function; their differential expression suggests that they might be involved in cell cycle and/or developmental events.
Peaks of gene expression were observed throughout the cell cycle. When genes with similar expression patterns were grouped, it became clear that genes involved in a specific cell-cycle-regulated event were often expressed just before the execution of this event. This was already known for the well-studied flagellar genes [7] and for genes required for pili synthesis [8]. The general rule is that genes encoding proteins involved in the formation of complex multiprotein structures are coexpressed in an order that reflects the order of assembly of these structures. Knowing the order of expression of genes involved in a complex cellular function can provide important hints about the order of action of the proteins involved. For example, the genes encoding proteins involved in cell division are expressed in a precisely ordered fashion, with the cell division initiation gene ftsZ being expressed first, followed by ftsI, then ftsW, and finally by ftsQ and ftsA. The ordered expression of ftsZ and ftsQA was recently shown to be important as a checkpoint coupling DNA replication and cell division [9]. If the order of cell division gene expression reflects the order of assembly of their protein products in the cell division apparatus, the order of assembly of cell division proteins in Caulobacter might be different from that suggested by studies in Escherichia coli [10]. An intriguing alternative is that the order of action of cell division proteins in E. coli is different from the current model.
The most exciting aspect of this study is that it identified potential regulators of cell cycle progression. Twenty-seven two-component regulatory proteins and sigma factors are amongst the newly identified genes with a differential pattern of expression. The study of these genes should allow the mapping of the regulatory networks that control the Caulobacter cell cycle. An impressive preview of what is possible was provided when it was determined that the global cell cycle response regulator CtrA is either directly or indirectly involved in the regulation of 26% of cell-cycle-regulated genes.
Proteomic Analysis of the Cell Cycle
One limitation of microarray studies is that not all changes in mRNA level reflect a change in protein concentration. For example, even though cckA, parA, parB and flbD mRNAs were found to vary during the cell cycle in the microarray study, the concentration of their protein products is constant [11-13]. Ideally, microarray studies should be complemented by analysis of the pattern of protein expression as proteins carry out most cellular activities. Changes in protein activity resulting from protein modification or degradation also elude DNA-microarray analysis. The second paper, reporting a proteomic study of the Caulobacter cell cycle [6], nicely complements and extends the microarray study. Proteins from synchronized cells were pulse-labeled with radioactive methionine to measure their rates of synthesis at different stages of the cell cycle, and were then separated by 2D gel electrophoresis. This procedure reproducibly resolved 979 proteins, of which 144 (15%) were found to be differentially synthesized during the cell cycle. One caveat of this analysis is that the information provided by 2D gels is mostly about abundant soluble proteins.
Cluster analysis of the expression patterns revealed that the regulated proteins fall into 23 groups, each with distinct cell cycle expression profiles. The identity of 91 of the 144 cell-cycle-regulated proteins was determined by peptide-mass fingerprinting using a matrix-assisted laser desorption time-of-flight (MALDI-TOF) spectrophotometer. A crucial finding of the proteomic study was that, of the 63 proteins expressed during the cell cycle whose genes were analyzed in the microarray study, 49 (78%) had identical or very similar cell cycle expression patterns. Although this supports the validity of using DNA microarray data to make inferences about protein expression, ten of the genes encoding cell-cycle-regulated proteins showed no variation of mRNA level and four had inverse expression patterns. These results suggest that a significant number of proteins are subject to post-transcriptional regulation, underscoring the importance of proteomic analyses of gene expression.
The most striking finding of this study is the identification of a set of 48 proteins that are degraded in the course of a single cell cycle, providing the first evidence of a global role for proteolysis in control of the bacterial cell cycle. Highly unstable proteins are often regulatory proteins required for a short period of time [14]. Their presence at the wrong time can be detrimental to the cell. For example, proteins crucial for cell cycle progression such as CtrA (reference 15), FtsZ (reference 16), and CcrM (which encodes an essential DNA methyltransferase) [17] were part of the set of unstable cell-cycle-regulated proteins, confirming previous studies. The chemotaxis proteins CheYI and CheD were also part of this set, indicating that most of the chemotaxis machinery is degraded at every cell cycle, probably during swarmer cell differentiation like the previously studied McpA (reference 18). Approximately half of the unstable proteins were differentially synthesized during the cell cycle, strongly suggesting that these proteins are only present during a short time-period. These proteins are prime candidates for important regulators of cell cycle progression.
Metabolic Requirements of the Cell Cycle
The results from these studies provide a global view of the types of proteins whose cell cycle synthesis might provide an advantage for Caulobacter cells. Many such proteins were known from classical studies, most of them involved in developmental control or in organelle synthesis. The surprising finding from this study is that some proteins involved in metabolic functions are also cell cycle regulated. This opens the door to a true understanding of cell cycle control and of the cell's metabolic requirements as it proceeds through the developmental cell cycle. For example, the expression of genes and/or proteins involved in energy metabolism and macromolecular synthesis was upregulated during swarmer cell differentiation, probably because the energy requirement of the cell increases as the growth rate increases during differentiation. A potential regulator of amino acid biosynthesis was also upregulated during swarmer cell differentiation, and enzymes involved in amino acid biosynthesis were increased predominantly after swarmer differentiation, reflecting an increased demand for amino acids. The timed expression of enzymes required for the synthesis of redox cofactors and proteins involved in the oxidative stress response suggest that the cell is subject to more changes (or more likely to be affected by changes) in redox potential at specific stages of its life cycle.
From the Old Days to the Future
In the 'old days', a time-consuming genetic screen could identify important regulators or participants in a pathway and set the stage for years of work in many laboratories. The post-genomic analyses described in these articles have accomplished a very similar task by pointing the way to genes and proteins that are likely to be important for cell cycle progression. The expression patterns obtained in these studies are like the phenotype of genetic screens, giving a strong indication of the event involving a specific protein. However, the timing information can sometimes be misleading and post-genomic methods will never replace well-designed genetic screens. Further studies of the cell cycle genes and proteins identified are required, especially as it is now clear that timing is not everything and that the subcellular location of proteins can be as important as its time of expression. The exciting prospect for the future is that as the identified genes are mutated, the complete network of genes whose normal expression is dependent - either directly or indirectly - on the mutated gene, can be easily identified using post-genomic methods.
