Two-component systems (2CSs)

Bacteria must be able to sense and respond to their environment in order to survive. One major way microbes accomplish this is using signal transduction machinery such as two-component systems (2CSs). 2CSs are distributed widely throughout the prokaryotes and eukaryotes, with the exception of animals, rendering them as a target for the development of novel antimicrobial agents (85, 86).

As the most prevailing signal transduction mechanisms to control gene expression in bacteria, a typical 2CS consists of a membrane-associated sensor kinase and a cytoplasmic response regulator. In response to an external stimulus, the sensor protein undergoes auto-phosphorylation at a conserved histidine residue, and the signal is subsequently relayed to the cytosol via phosphorylation at the aspartic acid in the receiver domain of the response regulator (27). The signal transduction cascade may thus activate genes required for bacterial virulence during infections or survival in hostile environments (12). As is widely believed, 2CS proteins function as components of a signal transduction network, enabling bacteria to respond to complex environmental stimuli (5).

1.2.1 Classification of 2CSs

The 2CSs were classified into three major types by virtue of their functional domains: the classical system, the unorthodox system, and the hybrid system (204). Mostly, a 2CS sensor kinase carrying one input domain and one transmitter domain was called a classical (IT-type) sensors kinase. The others contain both the sensor kinase signature and a receiver domain of the response regulator, which were referred to as hybrid (ITR-type) sensors kinases. A small fraction of the hybrid sensors possesses an additional output domain at the carboxyl terminus and were referred to as unorthodox (ITRO-type) sensor kinases. For example, in P. aeruginosa PAO1, there are 42 IT-type sensor kinases, 12 ITR-type senor kinases, and 5 ITRO-type sensor kinases. Only three Hpt (histidine-containing phosphotransfer) modules had been observed (204). With the increasing number of bacterial 2CS identified, an emerging theme is the discovery of auxiliary factors, which are capable of influencing phosphotransfer but distinct from sensor kinases and response regulators. As the number of auxiliary factors increases, the members of a 2CS would surely increase, forming a three (and more) component system (27).

Auxiliary factors could either exert its function by targeting the response regulator, as in the case of the unorthodox 2CS response regulators RcsB and its auxiliary factor RcsA (122, 155, 251). Auxiliary factors could also target on the sensor protein. One of the examples in a multiple component system is the Rap and Spo0E family phosphatases that interact with, and stimulate the auto-phosphatase activity of, the Spo0F and Spo0A response regulators respectively (219).

1.2.2 The sensor kinase

Each sensor kinase contains a variable input domain that is adapted for the detection of a specific stimulus, typically a chemical ligand, and a conserved transmitter domain that transfers the signal to its cognate response regulator through a phosphorylation cascade (57).

A prototype sensor kinase is a homodimeric integral membrane domain in which the sensor domain is depicted as an extracellular loop within two membrane-spanning segments. The transmitter domain is localized within the cytoplasm. Recently the domains adapted by sensor proteins have been classified as extracellular sensor domains such as PDC (PhoQ-DcuS-CitA) (39, 40, 215), or membrane-embedded sensor domain similar to the phototaxis sensory rhodopsin II-transducer complex HtrII-SrII (84) and cytoplasmic sensor domains such as PAS (7, 123), GAF (102) or PHY (266). Although there is an apparent lack of uniformity between the specific conformational changes that take place within sensor domains upon changes in stimulus, the recent structural evidence suggests that the communication between

sensor and transmitter domains in sensor protein signaling is mediated by subtle structural changes along the dimmer interface and that related aspects of symmetry and asymmetry may also exist.

Signal termination in 2CSs usually occurs via the loss of the phosphoryl group from the response regulator. For example, the Spo0E family phosphatases (56, 88), which catalyze the dephosphorylation of the response regulator Spo0A involved in the initiation of endospore formation not only in Bacillus subtilis but also in other gram-positive bacteria, contain sequence and structural features similar to the chemotaxis phosphatases of CheZ (126, 275) and CheC/CheX/FliY families (188, 191, 230).

1.2.3 The response regulator

Consisted of a receiver domain and an output domain, the response regulator typically acts as the effector of 2CS. The input or receiver domain, which is usually conserved among regulators, received signals transferred from the sensor kinase and in some case from small phosphor donors such as phosphoramidates and acyl phosphates (258). Phosphorylation of the receiver domain is usually followed by a conformational change of the response regulator, which subsequently results in dimerization of the protein.

In contrast to the input domain, the output domain was rather diverse. Though most response regulators act as transcriptional regulators with various DNA-binding domains, the output domains of response regulators covered effector domains with a wide range of cellular functions, including RNA-binding, chemotaxis, two-component phosphorelay, cyclic di-GMP signaling, cAMP signaling and protein Ser/Thr phosphorylation (76). The properties that no restriction was applied on the variety of domains fused with the receiver domain and many 2CSs were able to interfere with the functioning of other signal transduction systems put 2CSs at the top of the bacterial signaling hierarchy.

1.2.4 The 2CS network

The increasing knowledge of 2CSs has revealed a diversity of designs controlling the flow of information within and between circuits. As a result, a 2CS network could be expected from two aspects, either the control of phosphorylation/dephosphorylation, or the signal integration and distribution among phosphotransfer pathways.

The control of phosphorylation/dephosphorylation covers single-step phosphotransfer, in which the sensor kinase exerts only its kinase activity, like the histidine kinase CheA in chemotaxis circuits (126), or exerts both kinase and phosphatase activity, as exemplified by

the E. coli 2CSs OmpR/EnvZ (11), PhoP/PhoQ (166) and CpxR/CpxA (217). In the latter case, a high expression level of the histidine kinase will increase the complex formation of the kinase/response regulator pair, with a commitment decrease in the response regulator phosphorylation.

The occurrence of phosphorelay in a 2CS could be found either in separate components, as in the sporulation phosphorelay of B. subtilis (180), or in hybrid proteins which usually contain reversible phosphorylation (4, 80) as also observed in RcsCDB phosphorelay regulating E. coli group I CPS biosynthesis (155). Possibly, the evolution of such complex phosphorelay machinery in 2CS signaling could provide additional points of control.

Finally, the autoregulation, which is resulted from a positive feedback from the 2CS itself, probably modulates the sensitivity to the applied stimulus, as explored in the BvgA/BvgS in Bordetella bronchiseptica (255). This may also create a short-term “learning”

behavior, as shown in 2CS PhoB/PhoR in E. coli (103). A negative autoregulation could even give rise to an oscillatory behavior in the expression of downstream genes, as observed in the 2CS CovS/CovR in Streptococcus pyogenes (93).

In view of signal integration and distribution among phosphotransfer pathways, the 2CS network contains branched pathways of either “one-to-many” in the chemotaxis regulatory system (126), “many-to one” in the quorum sensing network of Vibrio harveyi (151), or

“cross-regulation” via auxiliary proteins in S. enterica PmrD connector-mediated circuit (117, 119). Though much has been investigated in the complex 2CS network, a need remains for better understanding towards the properties of different circuit architectures to identify potential drug targets.