Portal de Periódicos da CAPES

Gene Regulation: Hacking the Network on a Sugar High

2008; Elsevier BV; Volume: 30; Issue: 1 Linguagem: Inglês

10.1016/j.molcel.2008.03.005

1097-4164

Tom Ellis, Xiao Wang, James J. Collins,

Bioinformatics and Genomic Networks

In a recent issue of Molecular Cell, Kaplan et al., 2008Kaplan S. Bren A. Zaslaver A. Dekel E. Alon U. Mol. Cell. 2008; 29: 786-792Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar determine the input functions for 19 E. coli sugar-utilization genes by using a two-dimensional high-throughput approach. The resulting input-function map reveals that gene network regulation follows non-Boolean, and often nonmonotonic, logic. In a recent issue of Molecular Cell, Kaplan et al., 2008Kaplan S. Bren A. Zaslaver A. Dekel E. Alon U. Mol. Cell. 2008; 29: 786-792Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar determine the input functions for 19 E. coli sugar-utilization genes by using a two-dimensional high-throughput approach. The resulting input-function map reveals that gene network regulation follows non-Boolean, and often nonmonotonic, logic. If you had just spent $500 and a day standing in line to get your hands on the latest gadget, would you immediately take a hammer to it? It's probably the last thing a normal person would do, yet last June this is precisely what a handful of curious individuals did to Apple's iPhone (BBC News Online, 2007BBC News Online (2007). Hackers lift the bonnet on iPhone (http://news.bbc.co.uk/1/hi/technology/6269014.stm).Google Scholar). Their motivation? To uncover the technology hidden within and, in doing so, to help a community of hackers and bootleggers get started on modifications and imitations. The approach is called reverse engineering—where detailed analyses of the workings of an existing system are undertaken to help build independent copies, new devices, or simply just to understand the system's logic without knowing its blueprint beforehand. Reverse engineering is a method that is now being employed by systems biologists to understand the workings of biological systems. Much like the curious individuals dissecting their latest purchases, systems biologists use approaches that effectively open up a system they are keen to explore. Perturbations are applied to the system, and its functional architecture can be inferred from how its individual parts behave in response to different conditions. In the case of gene networks, this top-down approach is used to parse out the logical interactions between known parts, and much like in electrical engineering, a circuit diagram of Boolean gates like AND, OR, and NOR can typically be put together to describe the system. The top-down method is relatively new to biology and has had some significant successes in the past few years (Gardner et al., 2003Gardner T.S. di Bernardo D. Lorenz D. Collins J.J. Science. 2003; 301: 102-105Crossref PubMed Scopus (894) Google Scholar, Ma'ayan et al., 2005Ma'ayan A. Jenkins S.L. Neves S. Hasseldine A. Grace E. Dubin-Thaler B. Eungdamrong N.J. Weng G. Ram P.T. Rice J.J. et al.Science. 2005; 309: 1078-1083Crossref PubMed Scopus (266) Google Scholar, Yeger-Lotem et al., 2004Yeger-Lotem E. Sattath S. Kashtan N. Itzkovitz S. Milo R. Pinter R.Y. Alon U. Margalit H. Proc. Natl. Acad. Sci. USA. 2004; 101: 5934-5939Crossref PubMed Scopus (407) Google Scholar). However, this approach often suffers from the low resolution of indirect experimental measurements due to technological restrictions. Now, Kaplan et al., 2008Kaplan S. Bren A. Zaslaver A. Dekel E. Alon U. Mol. Cell. 2008; 29: 786-792Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar avoid this limitation by exploring a relatively complex regulatory system in two dimensions with high-resolution, high-throughput measurements (Figure 1). Their equivalent of the iPhone is E. coli, and the parts of interest within their model organism are 19 sugar-utilization genes. These genes are known to be regulated by a network motif described as dense overlapping regulons (DOR), where a set of regulators combinatorially control a set of output genes (Shen-Orr et al., 2002Shen-Orr S.S. Milo R. Mangan S. Alon U. Nat. Genet. 2002; 31: 64-68Crossref PubMed Scopus (2050) Google Scholar). The output of each gene within the network is measured with a library of fluorescent reporter strains, where in each strain a copy of the promoter of one gene of interest is used to drive expression of green fluorescent protein (GFP). The GFP levels provide an accurate readout of transcriptional activity for that gene in vivo, and the logic within the network can be revealed by measuring each gene's activity under different conditions. The process is closely analogous to an electrical engineer testing each part of a circuit with a voltmeter to see which components respond to different inputs and by how much. From this, the authors are able to assign a two-dimensional input function to each gene that describes how its output is controlled by the multiple input signals. Once determined, the input functions of the system can be compared and used to map the logic underlying the network. It is generally believed that bacterial sugar-utilization genes are regulated in a qualitatively similar manner. But the input functions inferred by Kaplan et al., 2008Kaplan S. Bren A. Zaslaver A. Dekel E. Alon U. Mol. Cell. 2008; 29: 786-792Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar show great diversity and cannot be explained by simple combinations of Boolean gates, challenging the usual analogy made between electronic circuits and gene regulatory networks. The most surprising results come from the input functions for genes involved in galactose metabolism. For the genes galP and galETKM, the input functions are nonmonotonic with respect to the input signals; as an input signal increases, transcription rates rise to a peak and then decrease. The authors hypothesize that this might be due to the fact that, unlike other sugars in this study, galactose has a dual use—it is used as both a carbon source and a cell-wall component. This functional duality and its relationship to the unexpected input functions may warrant further studies to provide a molecular-level understanding of nonmonotonic behavior. The authors also show that many of the two-dimensional input functions can be decomposed into the products of two one-dimensional functions. This finding could greatly reduce the experimental workload needed for high-resolution data, and the decomposition of input functions could give us some hints about the existence of general regulatory principles. However, there are exceptions such as the regulator of the fucose operon, fucR, the input function of which displayed two peaks and could not be decomposed into the products of two one-dimensional functions. Molecular reasoning will be needed to understand such anomalies; for example, unlike other genes in this study, fucR is controlled by an internal promoter in addition to an upstream promoter. The high-resolution input functions described by Kaplan et al., 2008Kaplan S. Bren A. Zaslaver A. Dekel E. Alon U. Mol. Cell. 2008; 29: 786-792Abstract Full Text Full Text PDF PubMed Scopus (103) Google Scholar provide a great tool for reverse-engineering gene regulation. A natural extension would be the addition of more dimensions. Other possible external inducers that could be studied include growth inhibition, heat shock, and DNA damage. The authors further raise a question of whether the input function diversity originates from the various promoter structures or from differences in upper-stream circuitry. An intriguing way to address this question would be to engineer equivalent gene regulatory networks through synthetic biology (Guido et al., 2006Guido N.J. Wang X. Adalsteinsson D. McMillen D. Hasty J. Cantor C.R. Elston T.C. Collins J.J. Nature. 2006; 439: 856-860Crossref PubMed Scopus (269) Google Scholar) or to modify gene regulation within the existing network by replacing the natural promoters with synthetic promoters generated from different libraries (Alper et al., 2005Alper H. Fischer C. Nevoigt E. Stephanopoulos G. Proc. Natl. Acad. Sci. USA. 2005; 102: 12678-12683Crossref PubMed Scopus (639) Google Scholar, Cox et al., 2007Cox 3rd, R.S. Surette M.G. Elowitz M.B. Mol. Syst. Biol. (Stevenage). 2007; 3: 145PubMed Google Scholar, Murphy et al., 2007Murphy K.F. Balazsi G. Collins J.J. Proc. Natl. Acad. Sci. USA. 2007; 104: 12726-12731Crossref PubMed Scopus (155) Google Scholar, Solem and Jensen, 2002Solem C. Jensen P.R. Appl. Environ. Microbiol. 2002; 68: 2397-2403Crossref PubMed Scopus (104) Google Scholar). Bottom-up approaches like these would no doubt complement the successes of the top-down approach detailed here. Diverse Two-Dimensional Input Functions Control Bacterial Sugar GenesKaplan et al.Molecular CellMarch 28, 2008In BriefCells respond to signals by regulating gene expression. The relation between the level of input signals and the transcription rate of the gene is called the gene's input function. Because most genes are regulated by more than one signal, the input functions are usually multidimensional. To understand cellular responses, it is essential to know the shapes of these functions. Here, we map the two-dimensional input functions of 19 sugar-utilization genes at high resolution in living E. coli cells. We find diverse, intricately shaped input functions, despite the similarity in the regulatory circuitry of these genes. Full-Text PDF Open Archive