The TetR Family of Regulators

Abstract

SUMMARY

The most common prokaryotic signal transduction mechanisms are the one-component systems in which a single polypeptide contains both a sensory domain and a DNA-binding domain. Among the >20 classes of one-component systems, the TetR family of regulators (TFRs) are widely associated with antibiotic resistance and the regulation of genes encoding small-molecule exporters. However, TFRs play a much broader role, controlling genes involved in metabolism, antibiotic production, quorum sensing, and many other aspects of prokaryotic physiology. There are several well-established model systems for understanding these important proteins, and structural studies have begun to unveil the mechanisms by which they bind DNA and recognize small-molecule ligands. The sequences for more than 200,000 TFRs are available in the public databases, and genomics studies are identifying their target genes. Three-dimensional structures have been solved for close to 200 TFRs. Comparison of these structures reveals a common overall architecture of nine conserved α helices. The most important open question concerning TFR biology is the nature and diversity of their ligands and how these relate to the biochemical processes under their control.

INTRODUCTION

Prokaryotes use signal transduction systems to sense alterations in the environment and respond accordingly. These signal transduction systems can be broadly divided into two categories: one-component systems and two-component systems (1, 2). In one-component systems, the sensory and output functions are located on the same polypeptide, while in two-component systems, the sensory and output functions are located on separate polypeptides. While the term two-component system is better known, one-component systems are actually much more abundant in prokaryotes (2). There are at least 20 families of prokaryotic one-component systems that can be defined by amino acid conservation in their DNA-binding domains and are defined by different conserved motifs (e.g., pfam and Interpro) (Table 1). The majority of one-component systems employ a helix-turn-helix DNA-binding domain, the exception being transcription factors of the MetJ family, which instead contain a ribbon-helix-helix domain (3). The DNA-binding domains are typically located at either the N- or C-terminal end of the polypeptide, depending on the particular family, although a few instances where the DNA-binding domain has a more central location are apparent. It has been suggested that there is a correlation between the location of the DNA-binding domain and repressor and activator activity. The suggestion was that repressors generally contain an N-terminal DNA-binding domain, while activators generally contain a C-terminal DNA-binding domain (4, 5). While this may hold true for many transcription factors, we would advise caution because there are well-documented exceptions to this rule (6).

Table 1.

Major families of one-component signal transduction systems

One-component system	Defining features	Reference(s)
AraC/XlyS	Involved in regulating pathways for the catabolism of various sugars, primarily transcriptional activators, C-terminal DNA-binding domain	196
ArgR	Involved in regulating amino acid metabolism, typically function as transcriptional repressors, N-terminal DNA-binding domain	197
ArsR/SmtB	Involved in regulating metal homeostasis, primarily transcriptional repressors, DNA-binding domain located near the center of the protein	198
AsnC/Lrp	Involved in regulating amino acid metabolism, function as both transcriptional activators and repressors, N-terminal DNA-binding domain	199
Crp/Fnr	Involved in regulating many cellular processes, may function as activators and repressors, C-terminal DNA-binding domain	200
DeoR	Involved in regulating sugar metabolism, typically function as repressors, N-terminal DNA-binding domain	201
DtxR	Involved in regulating metal homeostasis, primarily transcriptional repressors, N-terminal DNA-binding domain	202
Fur	Involved in regulating metal homeostasis, primarily transcriptional repressors, N-terminal DNA-binding domain	202
GntR	Involved in regulating numerous cellular processes, typically function as transcriptional repressors, N-terminal DNA-binding domain	203
IclR	Involved in regulating carbon metabolism, function as both transcriptional activators and repressors, N-terminal DNA-binding domain	204
LacI	Involved in regulating carbon metabolism, typically function as transcriptional repressors, N-terminal DNA-binding domain	205
LuxR	Involved in regulating quorum sensing, typically function as activators, C-terminal DNA-binding domain	206
LysR	Involved in regulating many cellular processes, function as both activators and repressors, N-terminal DNA-binding domain	207
MarR	Involved in regulating antibiotic resistance, typically function as transcriptional repressors, DNA-binding domain located near the center of the protein	208
MerR	Involved in regulating metal homeostasis, typically function as transcriptional repressors, N-terminal DNA-binding domain	209
MetJ	Involved in regulating many cellular processes, typically function as transcriptional repressors, N-terminal DNA-binding domain	3
ModE	Involved in regulating metal homeostasis, function as both transcriptional activators and repressors, N-terminal DNA-binding domain	210
PadR	Poorly characterized family, N-terminal DNA-binding domain	211
TetR	Involved in regulating antibiotic resistance, typically function as repressors, N-terminal DNA-binding domain	14
Xre	Involved in regulating various cellular processes, typically function as transcriptional repressors, N-terminal DNA-binding domain	212, 213

The naming of protein families is characterized by a founder effect of sorts, where the family name is derived from the first characterized member. One-component systems are no exception. This can be misleading, however, as not every member of a particular family is likely to be involved in regulating the same basic process as the founder. For example, many regulators in the AraC family are known for their role in sugar metabolism as AraC itself regulates genes required for arabinose catabolism (7). However, some members of the family recognize small molecules other than sugars and play a role in the regulation of virulence, morphological development and antibiotic production (8–10). In fact, some AraC family regulators (e.g., MarA and SoxS) are believed to lack a ligand-binding domain and may not serve as one-component signaling systems at all. Similar to the case for AraC family regulators, not all ArsR or MerR homologs bind metals like the founding member of the family. ArsR homologs have been identified as part of toxin-antitoxin systems (11), and MerR homologs are now known to respond to various chemical stressors (12).

The TetR family of regulators (TFRs) is a large and important family of one-component signal transduction systems (13, 14). While members of this family are best known for their roles as regulators of antibiotic efflux pumps, this in fact describes a minority of their functional roles. Indeed, characterized members are known to regulate numerous aspects of bacterial physiology and to interact with a vast array of ligands (Fig. 1).

Fig 1.

TFRs are known to interact with an exceptionally diverse set of small molecules, including antibiotics, metabolites, and cell-cell signaling molecules.

TetR FAMILY REGULATORS

All TetR family regulators (TFRs) consist of an N-terminal DNA-binding domain and a larger C-terminal domain. The proteins are almost exclusively α helical and function as dimers. In most cases the C-terminal domains interact with one or more ligands, in turn altering the regulator's ability to bind DNA. The exceptional diversity of these ligands is a chief source of interest in these regulators and is a central focus in this review. The name “TFR” is derived from the TetR protein, which was the first family member to be discovered and characterized in detail. Like TetR, many TFRs are repressors; however, there are important exceptions that are activators or that have roles unrelated to transcription.

The inducible nature of tetracycline resistance in Escherichia coli was recognized in the mid-1960s (15). The protein factor responsible for the regulation and induction of tetracycline resistance, which we now know as TetR, was partially purified a decade later (16). The sequence of tetR and many of the molecular details surrounding the regulation of tetracycline resistance were unraveled in the 1980s (17–21). We now know that TetR is the repressor of the tetracycline efflux pump encoded by tetA (Fig. 2). In the absence of tetracycline, a pair of TetR dimers bind to overlapping operator sequences in the intergenic region between the divergently transcribed tetR and tetA genes. When tetracycline is present, it binds directly to TetR, trapping it in a conformation that is incompatible with DNA binding (22). This allows transcription of both tetR and tetA.

Fig 2.

TetR regulates the expression of the tetracycline resistance determinant encoded by tetA. (A) In the absence of tetracycline, a pair of TetR dimers bind to repeated palindromic sequences in the intergenic region between tetR and tetA. (B) When present, tetracycline is bound by TetR, causing a conformational change such that TetR can no longer bind DNA. This allows for expression of the tetracycline efflux pump encoded by tetA.

More than 240 TFRs have been at least partially characterized (Table 2), and while TetR remains one of the central models for the family, it is clear that TetR does not represent the enormous diversity seen in the family. Its well-documented role as a regulator of antibiotic efflux is shared by at most 25% of the TFR family members (23). We know that other TFRs function as both repressors and activators (e.g., DhaS), serve as local or global regulators (e.g., AmtR), and can interact with small-molecule or protein ligands (e.g., SlmA). TFRs can be autoregulatory, can be under the control of other transcription factors (e.g., AtrA), or may undergo posttranscriptional regulation (e.g., HapR). In spite of many years of investigation, central questions remain unanswered. For example, while the repressing (i.e., DNA-bound) and induced (i.e., ligand-bound) conformations of TetR have been described in detail, the manner in which the protein converts from one form to the other has not. Furthermore, it is unlikely that the conformational transitions of TetR describe those of all other TFRs, and indeed, the structure of TetR is atypical for the family as a whole (24). It is unclear whether there are distinct conformational subgroups within the family or whether each protein is in fact unique. More globally, in the vast majority of cases, the ligand(s) bound by TFRs have yet to be identified. In this review we discuss what we can learn about TFRs from genomics and structural studies and how this informs, and is informed by, the roles attributed to TFRs in bacterial physiology through more detail-oriented molecular genetic investigation. We incorporate phylogenomics as a new means of organizing TFRs.

Table 2.

TFRs of known function

TFR	Organism	Descriptiona	Known ligand(s)	PDB ID	Reference(s)
AbyC	Verrucosispora maris AB-18-032	Located in the abyssomicin biosynthesis cluster; predicted to regulate abyD encoding a MFS export pump; mutation decreases abyssomicin synthesis			214
AcmG5	Streptomyces iakyrus	Located in the actinomycin G biosynthesis cluster			215
AcmP	Streptomyces chrysomallus ATCC 11523	Located in the actinomycin D biosynthesis cluster			94
AcmU	Streptomyces chrysomallus ATCC 11523	Located in the actinomycin D biosynthesis cluster			94
AcnR	Corynebacterium glutamicum	Regulates the aconitase (acn) gene		4AC6, 4ACI, 4AF5	145
AcrR	Escherichia coli	Regulator of the AcrAB multidrug efflux pump	Rhodamine 6G, ethidium, proflavine	3BCG, 3LHQ, 2QOP	112, 116
AcrR-like	Escherichia coli, Streptococcus uberis	Putative regulator of rdmC and mph(B) genes involved in spiramycin and tylosin resistance			103, 216
ActR (SCO5082)	Streptomyces coelicolor	Located in the actinorhodin biosynthesis cluster; regulates expression of the ActA and ActB efflux pumps	Actinorhodin, (S)-DNPA	2OPT, 3B6C, 3B6A	79
AcuR	Alcaligenes faecalis	Putative repressor for genes involved in dimethylsulfoniopropionate and acrylate catabolism			217
AcuR	Rhodobacter sphaeroides	Regulates expression of AcuI and DddL involved in dimethylsulfoniopropionate and acrylate catabolism	Acrylate		218
AdeN	Acinetobacter baumannii	Regulator of the AdeIJK efflux pump			219
AefR	Pseudomonas syringae	Regulates AHL production and is required for plant colonization		3CDL	40
AguR	Pseudomonas aeruginosa PAO1	Regulates AguBA required for agmatine utilization	Agmatine		181
AlnR2	Streptomyces sp. strain CM020	Located in the alnumycin biosynthesis cluster			220
AlpW	Streptomyces ambofaciens	Located in the alpomycin biosynthesis cluster and involved in the regulation of kinamycin biosynthesis; similar to gamma-butyrolactone receptors			221
AlpZ	Streptomyces ambofaciens	Located in the alpomycin biosynthesis cluster; similar to gamma-butyrolactone receptors			222
AmeR	Agrobacterium tumefaciens	Regulates the tripartite RND exporter AmeABC			223
AmiP	Streptomyces vinaceus-drappus	Located in the amicetin biosynthesis cluster			95
AmtR	Corynebacterium glutamicum	Global regulator of nitrogen control metabolism	GlnK		37
Ang8	Streptomyces sp. strain W007	Located in an angucyclinone biosynthesis cluster			224
ArpA	Streptomyces griseus	Involved in the regulation of antibiotic production and sporulation	A-factor (GBL)		225
ArpR	Pseudomonas putida S12	Regulates the ArpABC efflux pump involved in the export of multiple antibiotics			226
Asm2	Actinosynnema pretiosum	Located in the ansamitocin biosynthesis cluster and involved in the regulation of ansamitocin biosynthesis			227
Asm29	Actinosynnema pretiosum	Located in the ansamitocin biosynthesis cluster and involved in the regulation of ansamitocin biosynthesis			227
AtrA	Streptomyces coelicolor	Pleiotropic regulator of antibiotic biosynthesis			6
AtuR	Pseudomonas aeruginosa	Regulates genes required for acyclic terpene utilization			174
Aur1B	Streptomyces aureofaciens CCM 3239	Located in the auricin biosynthesis cluster			228
Aur1R	Streptomyces aureofaciens CCM 3239	Located in the auricin biosynthesis cluster; similar to gamma-butyrolactone receptors			229
AvaR1	Streptomyces avermitilis	Regulator of avermectin biosynthesis; similar to gamma-butyrolactone receptors	Avenolide		127
AvaR2	Streptomyces avermitilis	Similar to gamma-butyrolactone receptors			230
AvaR3	Streptomyces avermitilis	Pleiotropic regulator of antibiotic production; similar to gamma-butyrolactone receptors			230
AveI	Streptomyces avermitilis	Ortholog of AtrA; regulator of antibiotic production			231
Azi42	Streptomyces sahachiroi	Located adjacent to the azinomycin B biosynthetic gene cluster; thought to be beyond the boundaries of the cluster			232
BarA	Streptomyces virginiae	Involved in the regulation of virginiamycin; similar to gamma-butyrolactone-binding proteins	Virginiae butanolide (GBL)		233
BarB	Streptomyces virginiae	Involved in the regulation of virginiamycin; similar to gamma-butyrolactone-binding proteins			234
BarZ	Streptomyces virginiae	Located in the virginiamycin biosynthesis cluster; similar to gamma-butyrolactone-binding proteins			235
BdcR (YjgJ)	Escherichia coli	Regulator of BdcA expression			28
BecM	Streptomyces sp. strain DSM 21069	Located in the biosynthesis cluster for macrolactam BE-14106 biosynthesis			236
BepR	Brucella suis	Regulator of the BepDE efflux pump	Deoxycholate		237
BetI	Escherichia coli	Regulates expression of BetT, BetA, and BetB required for the synthesis of glycine betaine from choline	Choline		238
BioQ	Corynebacterium glutamicum ATCC 13032	Regulates biotin biosynthesis and import			189
BpeR	Burkholderia pseudomallei	Regulates the BpeAB-OprB multidrug efflux pump			239
BreR	Vibrio cholerae	Regulates the BreAB efflux pump in response to bile	Deoxycholate		39
Brp	Streptomyces clavuligerus	Gamma-butyrolactone receptor involved in the regulation of clavulanic acid and cephamycin C biosynthesis			240
BrtA	Listeria monocytogenes	Regulator of the MdrT efflux pump	Cholate		241
BspR	Burkholderia pseudomallei	Involved in regulating type III secretion systems			242
BtrR1	Bacillus circulans	Located in the butirosin biosynthesis cluster and involved in regulation			243
CalR1	Micromonospora echinospora	Located in the calicheamicin biosynthesis cluster			244
CampR	Rhodococcus sp. strain NCIMB 9784	Divergent to camK (6-oxocamphor hydrolase)			177
CamR	Pseudomonas putida	Regulator of camphor degradation			245
CasR	Rhizobium etli	Regulator of CasA required for colonization and infection of the host			246
CgmR (cg2894, Cgl2612)	Corynebacterium glutamicum	Multidrug resistance-related transcription factor	Ethidium bromide, malachite green	2ZOY, 2ZOZ, 2YVH, 2YVE	43, 247
ChlF1	Streptomyces antibioticus	Located in the chlorothricin biosynthetic gene cluster			248
ChryX5	Streptomyces albaduncus	Located in the chrysomycin biosynthesis cluster; a homolog is not present in the cluster for the related molecule ravidomycin			249
CifR	Pseudomonas aeruginosa	Regulator of the Cif toxin	Epibromohydrin		250
CmeR	Campylobacter jejuni	Regulator of the CmeABC efflux pump	Taurocholate, cholate, salicylate	2QCO, 3QPS, 3QQA	56
CmtI	Pseudomonas putida	Putative regulator of operons required for p-cymene/p-cumate degradation			175
CmtR	Pseudomonas putida	Putative regulator of operons required for p-cymene/p-cumate degradation			178
ComR	Escherichia coli	Regulator of ComC involved in copper permeability	Copper		72
CprA	Streptomyces coelicolor	Similar to gamma-butyrolactone receptors; involved in regulating sporulation and antibiotic production			134
CprB	Streptomyces coelicolor	Similar to gamma-butyrolactone receptors; involved in regulating sporulation and antibiotic production		1IU5, 1IU6	134
CprS	Streptomyces coelicolor	Similar to gamma-butyrolactone receptors			251
CymR	Pseudomonas putida	Regulator of the cym and cmt operons required for p-cymene and p-cumate degradation	p-Cumate		176
DarR (MSMEG_5346)	Mycobacterium smegmatis	First cyclic-di-AMP-responsive transcription factor to be identified in bacteria	Cyclic-di-AMP		142
DddH	Halomonas sp. strain HTNK1	Putative regulator of genes required for dimethylsulfoniopropionate and acrylate catabolism			252
DesT	Pseudomonas aeruginosa	Regulates the expression of the DesCB acyl-CoA desaturase operon	Oleate (corepressor), stearate (inducer)	3LSJ, 3LSR, 3LSP	166
DhaR	Rhodococcus rhodochrous	Regulator of haloalkane dehalogenase (DhaA)			143
DhaS	Lactococcus lactis	Regulator of the dha operon; functions as a transcriptional activator	DhaQ-dihydroxyacetone complex	2IU5	69
EbrR	Streptomyces lividans	Regulator of the EbrA efflux pump		3HTJ, 3HTI, 3HTH, 3HTA	253
EbrS	Streptomyces lividans	Regulator of the EbrC efflux pump			254
Ecm10	Streptomyces lasaliensis	Located in the echinomycin biosynthesis cluster			255
EmhR	Pseudomonas fluorescens	Regulates the EmhABC efflux pump that influences production of 2,4-diacetylphloroglucinol and is required for phenanthrene, anthracene, and fluoranthene efflux			256, 257
EncS	Streptomyces maritimus	Located in the enterocin biosynthesis gene cluster			258
EnvR (AcrS)	Escherichia coli	Divergent to the AcrEF efflux pump; may function as a switch for the alternative expression of AcrAB and AcrEF efflux pumps			259
EpeR	Streptomyces clavuligerus	Controls expression of the EpeA efflux pump			260
EsmT4	Streptomyces antibioticus Tu 2706	Located in the esmeraldin biosynthesis cluster			261
EthR	Mycobacterium tuberculosis	Regulator of ethA encoding a monooxygenase required for the activation of ethionamide	Hexadecyl octanoate	1T56	58
FabR	Escherichia coli	Regulator of genes required for unsaturated fatty acid synthesis	Unsaturated thioesters		165
Fad35R (Rv2506)	Mycobacterium tuberculosis	Regulator of Fad35 acyl-CoA synthetase	Palmitoyl-CoA		162
FadR (YsiA)	Bacillus subtilis	Regulator of fatty acid catabolism	Long-chain acyl-CoAs	1VIO	161
FadR	Pseudonocardia autotrophica	Regulates fad genes required for fatty acid degradation			158
FadR	Thermus thermophilus	Regulator of genes required for fatty acid degradation	Medium to long (C10 to C18) straight-chain fatty acyl-CoAs	3ANG, 3ANP	150
FarA	Streptomyces sp. strain FRI-5	Gamma-butyrolactone autoregulator that controls antibiotic production	IM-2 (GBL)		262
FasR	Corynebacterium glutamicum	Regulator of accD1 and fasA expression required for lipid synthesis			157
FrrA	Bradyrhizobium japonicum	Regulator of the FreABC efflux pump	Genistein, daidzein		263
HapR	Vibrio cholerae	Master quorum-sensing regulator		2PBX	264
HemR	Propionibacterium freudenreichii	Possible regulator of hem gene expression required for the conversion of glutamate to protoheme			190
HlyIIR	Bacillus cereus	Regulator of hemolysin II expression			265
HnoR (HdnoR)	Arthrobacter nicotinovorans	Repressor of 6-hydroxy-d-nicotine oxidase	6-Hydroxy-d- and 6-hydroxy-l-nicotine		266
HrtR	Lactococcus lactis	Regulator of the HrtB-HtrA transporter	Heme	3VP5, 3VP5, 3VOX	191, 46
IcaR	Staphylococcus epidermidis	Regulator of the ica operon required for biofilm formation		2ZCM, 2ZCN	267
IfeR	Agrobacterium tumefaciens	Regulator of the IfeAB efflux pump			268
JadR*	Streptomyces venezuelae	Located in the jadomycin biosynthesis cluster			269
JadR2	Streptomyces venezuelae	Similar to gamma-butyrolactone receptors; involved in the regulation of jadomycin biosynthesis	Jadomycin and chloramphenicol		133, 270
KanG	Streptomyces kanamyceticus	Located near the kanamycin biosynthesis cluster but probably beyond cluster boundaries			271
KijA8	Actinomadura kijaniata	Located in the kijanimicin biosynthesis cluster	Kijanimicin		272
KijC5	Actinomadura kijaniata	Located in the kijanimicin biosynthesis cluster			272
KijR	Streptomyces coelicolor	Regulator of KijX expression and kijanimicin resistance	Kijanimicin, saccharocarcins A and B		25
KinR	Streptomyces murayamaensis	Located in the kinamycin biosynthesis cluster			273
KirRII	Streptomyces collinus	Located in the kirromycin biosynthesis cluster			274
KsbA	Kitasatospora setae	Gamma-butyrolactone receptor protein; involved in regulating bafilomycin biosynthesis	GBLs		275
KstR	Mycobacterium tuberculosis	Regulator of lipid metabolism		3MNL	169
KstR2	Mycobacterium tuberculosis	Regulator of cholesterol metabolism			170
LanK	Streptomyces cyanogenus	Located in the landomycin biosynthetic pathway	Landomycin A and intermediates		78
Lct13	Streptomyces rishiriensis	Putative gamma-butyrolactone receptor protein; located in the lactonamycin biosynthesis cluster			276
Lct14	Streptomyces rishiriensis	Putative gamma-butyrolactone receptor protein; located in the lactonamycin biosynthesis cluster			276
LfrR	Mycobacterium smegmatis	Regulator of LfrA multidrug efflux pump	Proflavine	2WGB, 2V57	55
LitR	Vibrio fischeri	Involved in regulating luminescence and symbiotic light organ colonization			277
LiuQ (Bamb_4589)	Burkholderia ambifaria AMMD	Regulator of branched-chain amino acid degradation			183
LmrA	Bacillus subtilis	Regulator of the LmrB efflux pump	Flavonoids (quercetin, fisetin, galangin, catechin, coumestrol, genistein)		104
LplR	Rhodococcus erythropolis	Regulator of l-pantoyl lactone dehydrogenase gene expression			278
LuxR	Vibrio harveyi	Global regulator			279
LuxT	Vibrio harveyi	Global regulator			280
McbR	Corynebacterium glutamicum	Global regulator of l-methionine and l-cysteine biosynthesis	S-Adenosylhomocysteine		185
Mce3R	Mycobacterium tuberculosis	Putative regulator of lipid metabolism			281
MdoR	Mycobacterium sp. strain JC1	Regulator of genes required for methanol oxidation			147
MedORF28	Streptomyces sp. strain AM-7161	Located in the medermycin biosynthesis cluster			282
MepR	Pseudomonas putida	Regulates efflux pump involved in toluene resistance			283
MerO	Streptomyces sp. strain NRRL 30748	Located in the meridamycin biosynthesis cluster			284
MexL	Pseudomonas aeruginosa	Regulator of the MexJK efflux pump			285
MexZ (AmrR)	Pseudomonas aeruginosa	Regulates the MexXY (AmrAB) exporter involved in aminoglycoside resistance		2WUI	286
MlaM	Streptomyces sp. strain MP39-85	Located in the biosynthetic gene cluster for the macrocyclic lactam ML-449			92
MmfR	Streptomyces coelicolor	Gamma-butyrolactone-like receptor involved in regulating methylenomycin production			128, 287
MmyR	Streptomyces coelicolor	Gamma-butyrolactone-like receptor involved in regulating methylenomycin production			128, 287
MmyR	Streptomyces violaceoruber	Located in the methylenomycin biosynthesis cluster			288
MnbR	Comamonas sp. strain JS46	Putative regulator of mnb operon required for 3-nitrobenzoate oxidation			144
MonRII	Streptomyces cinnamonensis	Located in the monensin biosynthesis locus			289
MphR	Escherichia coli	Regulator of macrolide resistance	14-membered macrolides (erythromycin, oleandomycin)	3G56, 3FRQ	101
MSMEG_6564	Mycobacterium smegmatis	Global regulator of DNA repair genes			290
MtrR	Neisseria gonorrhoeae	Regulator of the mtr efflux pump		3VIB	291
NalC	Pseudomonas aeruginosa	Indirect regulator of the MexAB-OprM efflux pump through regulation of ArmR expression	Chlorinated phenols		292, 293, 294, 295
NalD	Pseudomonas aeruginosa	Regulator of the MexAB-OprM efflux pump			296
NapR3	Streptomyces aculeolatus	Located in the napyradiomycin biosynthesis cluster			297
NapR7	Streptomyces aculeolatus	Located in the napyradiomycin biosynthesis cluster			297
NcsR2	Streptomyces carzinostaticus	Gamma-butyrolactone receptor located in the neocarzinostatin biosynthesis cluster			298
NcsR3	Streptomyces carzinostaticus	Gamma-butyrolactone receptor located in the neocarzinostatin biosynthesis cluster			298
NcsR4	Streptomyces carzinostaticus	Located in the neocarzinostatin biosynthesis cluster			298
NemR (YdhM)	Escherichia coli	Regulator of N-ethylmaleimide reductase	N-Ethylmaleimide and other Cys modification reagents		299
NfxB	Pseudomonas aeruginosa	Regulator of the MexCD-OprJ efflux pump			300
NicS	Pseudomonas putida	Regulator of genes required for nicotinic acid degradation	Nicotinic acid and hydroxynicotinic acid		148
NonG	Streptomyces griseus	Located near the nonactin biosynthesis cluster but probably beyond cluster boundaries			301
OpaR	Vibrio parahaemolyticus	Global regulator			301
ORF20p	Streptomyces hygroscopicus	Located in the geldanamycin biosynthesis locus
OrfH2	Streptomyces griseoruber	Located in the hedamycin biosynthesis locus			302
OvmY	Streptomyces antibioticus	Located in the oviedomycin biosynthesis cluster			303
PaaR	Azoarcus evansii	Regulator of genes required for phenyl acetic acid degradation			304
PaaR	Thermus thermophilus	Regulator of genes required for phenyl acetic acid degradation	Phenylacetyl coenzyme A		150
PapR3	Streptomyces pristinaespiralis	Located in the pristinamycin biosynthesis cluster; similar to gamma-butyrolactone receptors			305
PapR5	Streptomyces pristinaespiralis	Located in the pristinamycin biosynthesis cluster; similar to gamma-butyrolactone receptors			305
PG1181	Porphyromonas gingivalis	Expressed in response to NO stress			306
PgaY	Streptomyces sp. strain PGA64	Located in the pga angucyclinone biosynthesis cluster			307
PhaD	Pseudomonas putida	Regulator of genes required for polyhydroxyalkanoate metabolism			167
PhlF	Pseudomonas fluorescens	Located in the 2,4-diacetylphloroglucinol biosynthesis cluster	2,4-Diacetylphloroglucinol (inducer), salicylate (corepressor)		75
PhlH	Pseudomonas fluorescens	Located in the 2,4-diacetylphloroglucinol biosynthesis cluster			308
PigZ	Serratia sp. strain ATCC 39006	Regulator of the ZrpADBC efflux pump			309
Pip (SCO4025)	Streptomyces coelicolor	Regulator of the Pep efflux pump	Pristinamycin I		100
PksA	Bacillus subtilis	Located in the bacillaene biosynthesis cluster			310
PlaR2	Streptomyces sp. strain Tü6071	Located in the phenalinolactone biosynthesis cluster			311
PltZ	Pseudomonas sp. strain M18	Located in the pyoluteorin biosynthesis cluster			312
PmeR (PSPTO_4302)	Pseudomonas syringae	Regulator of MexAB-OprM	Flavonoids		313
PqrA (SCO1568)	Streptomyces coelicolor	Regulator of the PqrB efflux pump			314
PsbI	Rhodopseudomonas palustris	Regulator of p-cumate catabolism	p-Cumate		179
PsrA	Pseudomonas aeruginosa	Regulator of the β-oxidation operon	Long-chain fatty acids	2FBQ	163
PydR	Pseudomonas putida KT2440	Regulator of pyrimidine reductive catabolic pathway			154
Pyr27	Actinosporangium vitaminophilum	Located in the pyrrolomycin biosynthesis cluster			315
Pyr3	Actinosporangium vitaminophilum	Located in the pyrrolomycin biosynthesis cluster			315
PyrO	Streptomyces pyridomyceticus	Located in the pyridomycin biosynthesis cluster; similar to gamma-butyrolactone receptors			316
QacR	Staphylococcus aureus	Regulator of the QacA efflux pump	Rhodamine 6G, dequalinium, crystal violet, berberine, DiOC3, methyl green, benzalkonium, tetraphenylarsonium, nitidine, palmatine	1JTX, 1JT6, 1JTY, 1JUM, 1JUP, 1JUS, 1JTO, 1QVT, 1QVU	60, 53
QdoR (YxaF)	Bacillus subtilis	Regulator of quercetin dioxygenase QdoI (YxaG)	Flavonoids (quercetin, fisetin, tamarixetin, galangin, genistein, coumestrol)		317
RamR (STM0580)	Salmonella enterica serovar Typhimurium	Regulator of the RamA efflux pump; mutations in the RamR binding site result in a multidrug resistance phenotype			318
RefZ (YttP)	Bacillus subtilis	Involved in the switch from medial to polar cell division			195
RegE	Actinoplanes friuliensis	Located in (or adjacent to) the friulimicin biosynthesis cluster			319
RemM	Streptomyces resistomycificus	Located in the resistomycin biosynthesis cluster			320
RemQ	Streptomyces resistomycificus	Located in the resistomycin biosynthesis cluster			320
RifQ	Amycolatopsis mediterranei	Located in the rifamycin biosynthesis cluster			91
RkI	Streptomyces strain sp. 88-682	Located in the RK-682 biosynthesis cluster			321
RmiR	Rhizobium etli	Regulator of NodTch			322
RmrR	Rhizobium etli	Regulator of the RmrAB efflux pump			323
RolR	Corynebacterium glutamicum	Regulator of resorcinol degradation	Resorcinol	3AQS, 3AQT	49
RphA3	Streptomyces griseoviridis	Located in the prodigiosin biosynthesis cluster			324
RrdA (SCO1104)	Streptomyces coelicolor	Regulator of antibiotic production			325
RutR (YcdC)	Escherichia coli	Regulator of pyrimidine synthesis	Uracil		326
Rv3066	Mycobacterium tuberculosis	Regulator of Mmr multidrug efflux pump	Ethidium	3V6G, 3V78	327
SaaR	Streptomyces ambofaciens	Gamma-butyrolactone receptor involved in regulating spiramycin production			328
SabR	Streptomyces ansochromogenes	Gamma-butyrolactone receptor involved in regulating nikkomycin production			329
SabR	Streptomyces acidiscabies	Gamma-butyrolactone receptor involved in regulating WS5995B production			330
SabS	Streptomyces acidiscabies	Gamma-butyrolactone receptor involved in regulating WS5995B production			330
SACE_7040	Saccharopolyspora erythraea	Regulator of morphological differentiation			331
SaqK	Micromonospora sp. strain Tu 6368	Located in the saquayamycin Z biosynthesis cluster			83
SAV3818	Streptomyces avermitilis	Global upregulator of antibiotic production in Streptomyces species			332
SbtR	Thermus thermophilus HB8	Contains an intermolecular disulfide bridge that may be involved in ligand affinity		3VUQ	333
SCAB1401	Streptomyces scabies	Located in the pyochelin biosynthesis cluster			334
ScbR	Streptomyces coelicolor	Gamma-butyrolactone-binding protein; pleiotropic regulator of antibiotic production	SCB1		251
ScbR2	Streptomyces coelicolor	Similar to gamma-butyrolactone-binding proteins; regulator of Cpk polyketide production and gamma-butyrolactone biosynthesis	Actinorhodin and undecylprodigiosin		131, 132, 133
SchA21	Streptomyces sp. strain SCC-2136	Located in the biosynthesis cluster for the angucyclinones Sch 47554 and Sch 47555			335
SchA4	Streptomyces sp. strain SCC-2136	Located in the biosynthesis cluster for the angucyclinones Sch 47554 and Sch 47555			335
SchR3	Streptomyces chartreusis	Located in the biosynthesis cluster for calcimycin (A23187)			93
SCO0253	Streptomyces coelicolor	Regulator of SCO0252	Tetracycline	3FIW	336
SCO0332	Streptomyces coelicolor	Regulator of SCO0330		2ZB9	337
SCO1712	Streptomyces coelicolor	Regulator of antibiotic production		3BNI	338, 160
SCO3201	Streptomyces coelicolor	Regulator of antibiotic production			339
SczA	Streptococcus pneumoniae	Regulator of metal ion homeostasis	Zn2+		71
SfmR1	Streptomyces lavendulae	Located in the saframycin A biosynthesis cluster			340
SimR	Streptomyces antibioticus	Located in the simocyclinone D8 biosynthesis cluster	Simocyclinones D8 and C4	2Y2Z, 2Y30, 2Y31	76
SlgR1	Streptomyces lydicus	Located in the streptolydigin biosynthesis cluster			341
SlmA	Escherichia coli	Nucleoid occlusion factor	FtsZ	3NXC	192, 193
SmcR	Vibrio vulnificus	Global regulator		3KZ9	342, 343
SmeT	Stenotrophomonas maltophilia	Regulator of the SmeDEF efflux pump	Triclosan	2W53	52, 61, 344
SMU_1349	Streptococcus mutans	Regulator of the TnSmu2 operon, which contains a secondary metabolite biosynthesis gene cluster			345, 346
SngR	Streptomyces natalensis	Gamma-butyrolactone receptor protein involved in regulating natamycin biosynthesis and sporulation			347
SocA3	Myxococcus xanthus	Involved in regulating morphological development			348
SpbR	Streptomyces pristinaespiralis	Gamma-butyrolactone receptor protein involved in regulating pristinamycin biosynthesis and sporulation			349, 305
SrpR	Pseudomonas putida	Regulator of the SrpABC efflux pump	SrpS		350, 351
SrrA	Streptomyces rochei	Gamma-butyrolactone receptor protein involved in regulating lankacidin and lankamycin biosynthesis and sporulation			352, 353
SrrB	Streptomyces rochei	Gamma-butyrolactone receptor protein involved in regulating lankacidin and lankamycin biosynthesis and sporulation			352
SrrC	Streptomyces rochei	Gamma-butyrolactone receptor protein involved in regulating lankacidin and lankamycin biosynthesis and sporulation			352
SscR	Streptomyces scabies	Gamma-butyrolactone receptor protein involved in regulating secondary metabolism	GBLs		354
SsfT2	Streptomyces sp. strain SF2575	Located in the biosynthesis cluster for the polyketide SF2575			99
Strop_2766	Salinispora tropica	Located in the salinilactam biosynthesis cluster			355
TamK	Streptomyces sp. strain 307-9	Located in the tirandamycin biosynthesis cluster			356
SwrT	Vibrio parahaemolyticus	Ortholog of V. harveyi LuxT; regulator of swarming motility			357
TarA	Streptomyces tendae	Gamma-butyrolactone receptor protein involved in regulating nikkomycin production			358
TcaR2	Micromonospora chalcea	Located in the tetrocarcin A biosynthesis cluster			359
TcmR	Streptomyces glaucescens	Located in the tetracenomycin C biosynthesis cluster			360
Tei8	Actinoplanes teichomyceticus	Located in the teicoplanin biosynthesis cluster			361
TetR	Escherichia coli	Regulator of tetracycline resistance	Tetracycline	2TCT, 1QPI	362
TetR	Arthrobacter oxydans	Putative regulator of genes required for phenyl acetic acid degradation			363
TetR	Streptomyces toxytricini	Putative regulator of the propionyl-CoA carboxylase complex			364
Tmn21	Streptomyces sp. strain NRRL 11266	Located in the tetronomycin biosynthesis cluster			365
TR	Mycobacterium peregrinum	Putative regulator of macrolide resistance			366
TrdK	Streptomyces sp. strain SCSIO1666	Located in the tirandamycin biosynthesis cluster			367
Tsn22	Streptomyces longisporoflavus	Located in the tetronasin biosynthesis cluster			GenBank accession no. FJ462704
TtgR	Pseudomonas putida	Regulator of the TtgABC efflux pump	Phloretin, naringenin, chloramphenicol, tetracycline, quercetin, luteolin	2UXP, 2UXI, 2UXH, 2UXU, 2UXO	51, 368, 369
TtgW	Pseudomonas putida	Divergent to the TtgGHI efflux pump but does not play a major role in regulation			370
TvrR	Pseudomonas syringae	Required for pathogenesis			371
TylP	Streptomyces fradiae	Gamma-butyrolactone receptor protein involved in regulating tylosin production and sporulation			372, 373
TylQ	Streptomyces fradiae	Gamma-butyrolactone receptor protein involved in regulating tylosin production			373
UidR	Escherichia coli	Regulator of the d-glucuronidase UidA			374
UrdK	Streptomyces fradiae	Located in the urdamycin biosynthesis cluster			84
VanT	Vibrio (Listonella) anguillarum	Global regulator			375
VarR	Streptomyces virginiae	Located in the virginiamycin biosynthesis cluster	Virginiamycin S		77
VceR	Vibrio cholerae	Regulator of VceCAB efflux pump	Carbonyl cyanide m-chlorophenyl hydrazone		376
VexR	Vibrio cholerae	Regulates the VexAB efflux pump which is expressed in response to bile, sodium dodecyl sulfate, or novobiocin			39
VlmE	Streptomyces viridifaciens	Located in the valanimycin biosynthesis cluster			377
VtpR	Vibrio tubiashii	Global regulator of virulence factors			378
XdhR (SCO1135)	Streptomyces coelicolor	Regulator of xanthine dehydrogenase			156

^aMFS, major facilitator superfamily; AHL, acyl-homoserine lactone.

GENOMICS OF TFRs

A text-based search for TetR in the NCBI protein database gives well over 200,000 hits (as of 7 March 2013), and this number will continue to grow due to the explosion of whole-genome sequences available. The N-terminal DNA-binding domain of TFR family members is represented by conserved motifs or profiles in the public databases (e.g., IPR001647, PS50977, and pfam00440) and has been defined in previous reviews (14), aiding in the identification of TFRs from whole-genome sequences. While the vast majority of these TFRs have not been characterized, the availability of genome sequences allows us to examine different aspects of the genomics of TFRs.

Distribution of TFRs in Bacterial Genomes

Most sequenced bacterial genomes encode at least one TFR (14, 25). In the over 200 genomes that we examined, 23, from 8 genera, did not encode TFRs. TFRs were not found in at least some representatives from Borrelia, Chlamydia, Chlamydophila, Francisella, Helicobacter, Mycoplasma, Prosthecochloris, and Treponema. These are predominantly pathogens with genomes under 2 Mbp in size. In contrast, the Actinobacteria, along with other soil-dwelling isolates such as Burkholderia, Pseudomonas, and Rhizobium strains, encode the highest numbers of TFRs. Amycolatopsis (formerly Streptomyces) sp. strain AA4 encodes the greatest number of TFRs of the genomes we examined, at 212. Bacteria with large genomes tend to encode more TFRs (Fig. 3) (25). While in some instances this may be a function of the fact that bacteria with large genomes tend to encode a higher number of regulatory proteins, in other instances the situation may be more complex and indicate a preference for TFRs over other families of regulators. For example, Streptomyces coelicolor encodes 965 regulatory proteins in its approximately 8.7-Mbp genome (26). Of these regulators, 153 (15.8%) are TFRs, while only 34 (3.5%) are AraC family regulators and 40 (4.1%) are LysR family regulators (L. Cuthbertson and J. R. Nodwell, unpublished data). E. coli encodes 261 DNA-binding transcription factors in its 4.6-Mbp genome, of which 13 (5.0%) are TFRs, 28 (10.7%) are AraC family regulators, and 46 (17.6%) are LysR family regulators (27). Exceptions where bacteria with large genomes encode a relatively small number of TFRs include some deltaproteobacteria (e.g., Myxococcus and Stigmatella) and members of the phyla Planctomycetes and Verrucomicrobia. The evolutionary significance of this, if there is any, is not clear.

Fig 3.

Distribution of TFRs in sequenced genomes. Large genomes with a low number of TFRs are highlighted with a yellow box.

In some genera we observed a wide range in the number of TFRs in different species. For example, among the Mycobacterium spp., the pathogenic M. tuberculosis encodes 49 TFRs, M. leprae, known which is to have a reduced genome, encodes only 10, and the environmental isolates M. abscessus and M. smegmatis encode 138 and 137 TFRs, respectively. These data indicate a general trend that the number of TFRs encoded by an organism may reflect the diversity of environmental conditions that the organism encounters. Bacteria that grow in changeable niches, in particular the soil, are often enriched for TFRs while those that grow in close association with a host organism are not.

Conservation of TFRs

The availability of genome sequences allows us to examine the conservation of TFRs between strains and species. These comparisons may help to reveal TFRs associated with virulence traits or to distinguish newly acquired TFRs involved in specific adaptive responses from conserved TFRs more likely to be involved in regulating basic physiological processes. For example, a comparison of the TFRs in E. coli K-12 MG1655 and E. coli O157 EDL933 reveals that the two strains share 12 TFRs and that E. coli K-12 MG1655 encodes a single additional TFR not present in E. coli O157 EDL933. In E. coli O157 EDL933, one TFR, BdcR (formerly YjgJ), is truncated and lacks the DNA-binding domain. Further analysis indicates that this truncation is conserved in other O157 genomes as well as the genomes of some Shigella species. BdcR is a regulator of BdcA, a novel c-di-GMP-binding protein involved in biofilm dispersal (28). BdcR expression is thought to be regulated by NsrR, a protein that is involved in sensing nitric oxide (29) and that is also known to regulate other genes required for motility and biofilm development. While data on BdcR function are scant, the conserved deletion in E. coli O157 indicates that it may play a role in regulating an aspect of virulence.

A comparison of the TFRs in Pseudomonas aeruginosa PAO1 and the multidrug-resistant taxonomic outlier PA7 reveals that they have 36 TFRs in common and reveals TFRs unique to each strain that may play a role in the differences in virulence observed between strains. PAO1 encodes five TFRs absent in PA7 (PA1241, PA1290, PA2020, PA2766, and PA2931), while PA7 encodes two TFRs absent from PAO1 (PSPA7_2630 and PSPA7_4004). The PA7-specific TFRs are encoded within genomic islands of this isolate (30). PA2020, MexZ (also see TFRs and Antibiotic Resistance below), encodes a known regulator of the MexXY antibiotic resistance efflux pump (31). Mutations in MexZ are associated with isolates from chronic infections and small-colony variants (32, 33). In PA7, MexZ is truncated, lacking the DNA binding-domain, which leads to overexpression of MexXY and increased aminoglycoside resistance in this isolate (34).

Analyses of TFR conservation can be expanded to include many different species of the same genus. Conservation at the genus level may help to distinguish TFRs more likely to be involved in regulating basic cellular processes (e.g., fatty acid metabolism) as opposed to adaptive functions (e.g., resistance to specific antibiotics) and may point to more recently acquired traits. Our analysis of TFRs from members of the genus Streptomyces, the majority of which encode over 100 TFRs, reveals five TFRs that are conserved in all of the close to 70 strains sequenced as of 26 April 2013, with another seven TFRs highly conserved and missing in only one strain. One of these TFRs is more broadly conserved in Actinobacteria, while another two have been implicated in the regulation of antibiotic production in members of the genus (6, 35). We surmise that all 10 of these TFRs play an important role in regulating general processes important to antibiotic production and development in Streptomyces, while less conserved TFRs are more likely to play a role in regulating specific adaptive functions such as the catabolism of a specific carbon source or resistance to a specific antibiotic. It is interesting to note that four of the five conserved TFRs are type III TFRs (see “Predicting Target Genes” below) and that the regulatory targets cannot be predicted based on genomic orientation.

Predicting Operator Sites

Many TFRs bind palindromic, and often repeated, DNA operator sequences. Informatics approaches to identifying TFR operator sequences have been applied to small numbers of TFRs with success (24). In our experience, however, operator sites for TFRs of unknown function are often difficult to reliably predict. In many cases there is no obvious palindrome, and in others there are palindromes upstream of genes encoding TFRs or predicted targets that do not interact with the cognate TFR. In some cases, these may represent binding sites for other transcription factors. Ramos et al. (14) made use of protein-DNA crystals for QacR and TetR to identify amino acid positions that may generally be important in protein-DNA interactions and give specificity for a particular TFR for its operator sequence. It would be interesting to evaluate this approach to validate potential operator sequences identified through palindrome analysis or to perhaps predict the operator DNA sequence that is recognized by a TFR. Additional information such as DNase I footprinting can aid in the prediction of TFR operator sites from DNA sequence information (23).

Predicting Target Genes

TFRs can be classified into three types based on the orientation and proximity of their structural gene relative to adjacent genes on the chromosome (Fig. 4), and these relationships can be used to predict the regulatory target gene(s) of the TFR (23). The majority of TFRs are classified as type I: their genes show a divergent orientation to one of the adjacent genes, as is the case for tetR and tetA. This relationship is very predictive of a regulatory relationship in those cases where the intergenic region between the two genes is less than ∼200 bp. A longer intergenic region does not rule out a possible regulatory relationship; however, it is more rare in these cases. Type II TFRs are predicted to be cotranscribed with one or more adjacent genes based on orientation and a short distance (less than 35 bp) between genes. The majority of characterized TFRs are known or believed to be autoregulatory, and therefore type II TFRs would also be predicted to regulate the expression of cotranscribed genes. It should be noted, however, that an extensive investigation into autoregulation by TFRs is lacking, and certainly exceptions have been identified (e.g., AmtR [36–38]). In some cases, autoregulation is assumed based on other data (e.g., DNase I foot printing analysis for ActR [23]) but direct evidence is not available. The genes encoding type III TFRs show neither of these relationships with their neighboring genes. In these cases, putative regulatory relationships with neighbors, while they may exist, cannot be predicted by genomic orientation.

Fig 4.

Classification of TFRs based on the orientation and proximity of adjacent genes. (A) Type I TFRs are transcribed divergently from an adjacent gene. A regulatory relationship is predicted when this intergenic region is less than 200 bp. (B) Type II TFRs are predicted to be cotranscribed with and to regulate an adjacent gene based on a distance of less than 35 bp between genes. (C) Type III TFRs show neither of the above-described relationships with adjacent genes, and a regulatory relationship with the adjacent genes cannot be predicted.

Using this classification for TFRs, we can begin to take an inventory of the types of gene products regulated by TFRs (23). This inventory reveals that while the best-characterized TFRs do indeed regulate the expression of efflux pumps like the founding member of the family TetR, a large majority of TFRs actually regulate genes encoding cytoplasmic proteins. These proteins are almost exclusively predicted to be enzymes, and the diversity is extraordinary and includes all of the known functional classes (23). The biochemical functions of most of these enzymes are unknown.

Predicting Ligands

At this time, inducing ligands are known for 61 TFRs but remain unidentified for the vast majority of TFRs, including many of those that have been at least partly characterized. We have employed phylogenomics as a tool to predict ligands for TFRs of unknown function (25). Using this approach, we successfully identified the antibiotic kijanimicin as the inducing ligand for a previously uncharacterized TFR, KijR from Streptomyces coelicolor. Identifying the inducing ligand for KijR provided crucial insight into the function of its target gene, kijX, which acts as a kijanimicin deglycosylase. As discussed above, the majority of TFRs regulate enzymes of unknown function, and methods to identify the small-molecule ligands for TFRs will prove invaluable in determining the substrates and enzymatic functions carried out by the enzymes they regulate.

TFRs encoded in antibiotic biosynthesis clusters are known to interact with the products of those clusters (see TFRs and Antibiotic Resistance below) and can help us make predictions for ligands bound by TFRs of unknown function. For example, TFRs in the biosynthesis clusters for two structurally related polyether ionophores, calcimycin and monensin (TFRs SchR3 and MonRII, respectively), form a group in our phylogenetic analysis with the TFR of unknown function SSQG_00958 (Fig. 5A). Based on this clustering, we predict that SSQG_00958 binds a similar polyether ionophore and is involved in regulating resistance to the same molecule. SSQG_00958 is transcribed divergently from a putative exporter encoded by SSQG_00957. In another example, the gene encoding MlaM is located in the biosynthesis cluster for a macrolactam antibiotic and in our phylogenetic analysis falls into a larger group with two other TFRs, BecM and Strop_2766, located in the biosynthesis clusters for structurally related molecules (Fig. 5B). This cluster also contains numerous other TFRs of unknown function which we predict bind similar macrolactam antibiotics.

Fig 5.

Phylogenomics can be used to predict small-molecule ligands for TFRs of unknown function. (A) The TFR of unknown function SSQG_00958 is predicted to bind a polyether ionophore based on grouping with MonRII and SchR3. (B) TFRs encoded in the biosynthesis clusters for macrolactam antibiotics cluster together, leading to the prediction that all of the TFRs in this group interact with macrolactam antibiotics. (C) AefR may recognize a phytosterol based on clustering with BreR. (Adapted from reference 25.)

Ligand predictions based on phylogenomics are not limited to antibiotics. For example, BreR binds bile acids and is thought to be important to the survival of Vibrio cholerae in the intestinal tract (39). BreR and AefR share 30% identity (67% similarity) and grouped together in our analysis (Fig. 5C). AefR is involved in regulating quorum sensing and epiphytic fitness in the plant pathogen Pseudomonas syringae, but its inducing ligand is unknown (40). Given the similarities between BreR and AefR, we predict that the AefR-inducing ligand may be a phytosterol. Phytosterols share structural similarities with bile acids, and some (e.g., tomatidine) are known to have antimicrobial activity (41).

Combining information on TFRs from both phylogenomics and genomic context can also provide a powerful tool for predicting small-molecule ligands for TFRs. As the majority of TFRs are transcribed divergently from their target genes, in cases where the function of the target gene is known, this organization can lead to a prediction of a possible TFR ligand. For example SCO4099 from S. coelicolor is transcribed divergently from SCO4098, which encodes a putative streptogramin A acetyltransferase (vat) homolog. Our phylogenomics analyses coupled with additional database searches identify numerous TFRs sharing high similarity to SCO4099 in other actinomycetes; however, no ligands have been identified for any of them (Fig. 6) (25). These homologs are transcribed divergently from additional gene products implicated in resistance to streptogramin antibiotics (e.g., vgaA and vgbA) as well as gene products known to be involved in antibiotic resistance but not specifically in streptogramin resistance (e.g., mgtA/oleD and ereA). Using a combination of genomics approaches, we can predict that SCO4099 and related TFRs may bind a streptogramin antibiotic and that the genes regulated by these TFRs include both known and potentially novel streptogramin resistance genes.

Fig 6.

Combining information from genomic context with phylogenomics can also lead to ligand predictions for TFRs. (A and C) All of the TFRs in the group shown (A) (data are from reference 25) are type I TFRs predicted to regulate genes involved in streptogramin resistance (C). (B) Structure of the streptogramin antibiotic pristinamycin.

TFR STRUCTURAL BIOLOGY

General Structure of TFRs

X-ray crystal structures are currently available for close to 200 TFRs. Despite the vast sequence divergence seen in TFRs, structural data reveal that all family members share common structural features both in the DNA-binding domains (which are conserved in terms of primary sequence) and also in the ligand-binding domains (which are not) (24) (Fig. 7). The overall conserved structure of TFRs consists of nine α helices. The DNA-binding domain is composed of helices 1 to 3. Helices 2 and 3 form a helix-turn-helix motif, with helix 3 serving as the recognition helix that fits into the major groove upon DNA binding. The length of helix 1 is variable and can range from 12 to 23 residues (24). In many TFRs, helix 1 is preceded by a positively charged region responsible for making contacts with the DNA minor groove (see below) (42).

Fig 7.

TFRs share nine conserved α helices. In the front view, the DNA-binding domain is made up of helices 1 to 3. In the side view, helices 5 to 7 in the ligand-binding domain form a central triangle. In the top view, helices 8 and 9 from each monomer form a four-helical bundle that makes up the dimer interface. The structure of Rha06780 (PDB ID 2NX4) is shown, as it shows a structure typical of TFRs (24).

The ligand-binding domain is formed by conserved helices 4 to 9. Contacts between helix 1 of the DNA-binding domain and helices 4 and 6 of the ligand-binding domain link the two domains and are responsible for transmitting structural changes between the two domains upon ligand binding (see below). The ligand-binding domain can be divided into two structural subdomains. Helices 5 to 7 form a central triangle, while helices 8 and 9 make up the dimerization interface, forming a four-helix bundle with the same helices from the other monomer. In addition to the nine conserved helices, some TFRs, including TetR itself, contain a long insertion between helices 8 and 9 that may be involved in additional contacts to make up the dimer interface. It has been noted that while TetR serves as an important model for the family, its structure, along with that of another model TFR, QacR, is actually atypical compared to the majority of TFRs of known structure (24).

Interactions of TFRs with DNA

As of February 2013, structures have been solved for seven TFR-DNA complexes: CgmR, DesT, HrtR, QacR, SimR, TetR, and TM1030 (42–47). Based on the TFR-DNA structures currently available, it is clear that while TFRs share structurally similar DNA-binding domains, the mechanisms involved in DNA binding differ in significant ways between proteins. As discussed above, the DNA-binding domain is composed of helices 1 to 3, with helix 3 being responsible for the majority of DNA contacts. Helices 3 and 3′ recognize adjacent major grooves; thus, the spacing between these two helices in the TFR dimer is crucial for structural compatibility with stable DNA binding. In all cases investigated to date, this spacing is the target of conformational changes associated with ligand binding (see below). In general, TFR binding seems to induce a bend in the DNA, although at present there is no sequence or structural explanation for what determines either the direction of bending (toward or away from the TFR) or the degree of bending (43, 44, 47).

For some TFRs (e.g., TetR and QacR) the majority of TFR-DNA contacts are base specific, while for others (e.g., CgmR, DesT, HrtR, and SimR) the majority of TFR-DNA contacts are with the phosphate backbone. In the TetR-DNA complex, Lys48, located C-terminal to the DNA-binding domain, also makes an important DNA contact. The equivalent residue in SimR, Lys71, makes a similar contact, but this contact is absent from other TFR-DNA structures, including DesT and QacR. In SimR, additional DNA contacts are made between the N-terminal “arm” of SimR and the DNA minor groove. Positively charged arginine residues in the arm of SimR mediate these contacts. Sequence alignments and structural predictions reveal that a similar arm may be found in the majority of TFRs (42).

The QacR-DNA complex is distinct from that of other TFRs in that two QacR dimers bind cooperatively. Unlike many other transcription factors (e.g., the lambda phage repressor cI), where this cooperativity is due to protein-protein interactions between adjacent dimers (48), in QacR, cooperative binding is brought about by an alteration in the structure of DNA. Specifically, the interaction of QacR with DNA causes local underwinding that increases the distance between adjacent major grooves, and it is this conformation that most favorably forms the repressed complex with two QacR dimers. A slight widening of the major groove was also seen in the structure of DesT in complex with oleoyl coenzyme A (oleoyl-CoA) and DNA, indicating that this structural change is not limited to the QacR-DNA complex.

TFR-Ligand Interactions

At this time, ligands have been identified for 61 TFRs and X-ray crystal structures solved for 21 TFR-ligand complexes (Table 2). This information allows us to begin comparing the types of ligands recognized by TFRs and the mechanisms of ligand recognition. The known TFR ligands are extraordinarily diverse and include antibiotics, bile acids and other toxic molecules, cell-cell signaling molecules, carbon sources, proteins, fatty acids and fatty acid derivatives, and metal ions (Fig. 1, 5, and 6). This diversity supports a role for TFRs in regulating an equally diverse array of cellular processes from basic carbon and nitrogen metabolism to quorum sensing and antibiotic resistance. Structures are available for TFRs in complex with simple ligands such as citrate and resorcinol (49) to very complex molecules such as acyl-CoA derivatives (44) and antibiotics with multiple functional groups such as simocyclinone (50).

There are many ways that TFRs can interact with ligands. Structural data suggest that there are at least three different points at which ligands can enter a TFR ligand-binding site (Fig. 8). For example, ActR, QacR, SmeT, TetR, and TtgR all have a “side entry” opening distal to the dimerization interface that is believed to be the site of access for different ligands (22, 51–54). Ligands appear to enter CmeR, CgmR, HrtR, LfrR, and SimR via an entry point closer to the “front” of the protein (43, 46, 50, 55, 56). Finally, DesT, EthR, and FadR exhibit a relative “top entry” (44, 57, 58). It is unclear what, if anything, these differing mechanisms of ligand entry mean in terms of the type of ligand bound or the structural influence of ligand binding. For RolR and RutR, which bind resorcinol and uracil, respectively, there is no obvious entrance to the ligand-binding pocket (49). Rather, the ligand is trapped inside an otherwise inaccessible proteinaceous cage (Fig. 8).

Fig 8.

TFRs display different ligand entry points. Based on current TFR-ligand structures, the ligand-binding cavity may be accessible from the side (e.g., ActR), front (e.g., SimR), or top (e.g., DesT) of the TFR. In some structures (e.g., RolR), the ligand is not accessible to the external environment and the entry point cannot be determined. SlmA interacts with a protein rather than a small-molecule ligand. Residues involved in protein-protein interactions are colored in red.

Each tetracycline-binding pocket in TetR is composed primarily, but not exclusively, of residues from an individual monomer (22). This is also seen for the ligand-binding pockets of the majority of TFRs (e.g., ActR, CmeR, and QacR [53, 54, 56]). In contrast, the SimR ligand-binding cavity is composed of residues from both monomers such that each binds either the aminocoumarin or the angucyclinone moiety of the simocyclinone ligand (50).

Two molecules of Mg²⁺-tetracycline are bound by each dimer of TetR (22). This is also the case for many TFRs (e.g., SimR, CmeR, and MphR [50, 56, 59]), but different drug-binding stoichiometries are seen in some others. In the case of ActR, each ActR dimer is capable of binding either two molecules of actinorhodin or four molecules of (S)-2,4-dinitrophenyl acetate [(S)-DNPA] (54). In contrast, each dimer of LfrR binds only a single molecule of proflavine (55). Like LfrR, the majority of QacR-ligand structures show a single ligand within one monomer of each dimer. However, one structure of QacR in complex with two different ligands, ethidium and proflavine, within the same monomer has been solved (60). For CgmR, different binding stoichiometries are seen for different drugs, and the size of the drug is thought to play a role in the number of molecules required for CgmR derepression (43). TtgR also shows an interesting stoichiometry of binding to the plant antimicrobial phloretin, binding two molecules within one monomer and one in the other (51), while in SmeT, two molecules of triclosan were seen within a single monomer, while none were observed in the other (61).

Structures are available for four TFRs (CgmR, EbrR, QacR, and TtgR) in complex with different drugs, and analysis of these structures may shed light on how a single TFR may recognize a diverse set of ligands (43, 51, 53, 60). Based on the structures currently available, different drugs appear to be accommodated by different drug-binding sites within a single binding cavity. The structure of QacR has been solved in complex with six different cationic drugs. It shows a large binding pocket (1,100 ų) lined with negatively charged residues that form several separate drug-binding sites. In CgmR, ethidium bromide and methylene blue were found in the same inducer-binding pocket but were bound by different networks of hydrogen bonds. Structures are available for TtgR in complex with five different ligands, two antibiotics and three plant antimicrobials. TtgR contains a large, mainly hydrophobic, binding pocket with two distinct drug-binding sites: a high-affinity site and a general binding site. The high-affinity site is smaller and was occupied by only one of the five TtgR ligands. The general binding site is broader and was found to be occupied by all five ligands.

The first two reported structures of EthR were solved in conjunction with a fortuitous ligand, in once case hexadecyl octanoate (58) and in the other two cases uncharacterized molecules consisting of a six-membered ring (62). These structures have been critical in the design of synthetic EthR ligands (see below), and subsequent structures of EthR have been solved in complex with a number of these molecules (63–65). In one study, two related analogs were found to bind EthR with different orientations, supporting the extremely plastic nature of the EthR ligand-binding pocket (63). While EthR is known to recognize a large variety of ligands, only a small number of residues were found to be in contact with all ligands (66).

Similar to the case of EthR, the structure of CmeR shows how two structurally similar molecules can fit very differently into the same binding pocket (56). The structure of CmeR has been solved in complex with two structurally similar bile acids, taurocholate and cholate. Despite the similarity of these molecules, they were found within the same binding pocket but in opposite orientations, lying antiparallel to each other. Not only is DesT able to recognize different ligands, both saturated and unsaturated fatty acids, but its ability to do so is crucial to its function (44). Binding of the unsaturated oleoyl-CoA increases DNA-binding affinity, while the saturated palmitoyl-CoA serves as the inducer. DesT activity is regulated by the ratio of the two different ligands rather than through a single ligand. A Phe-rich cluster in DesT senses which ligand is bound. This changes the hydrophobic core to create a binding cavity tailored to each particular ligand.

Crystal structures are also available for seven uncharacterized TFRs in complex with bound ligands (3EGQ, 3CJD, 3KKD, 2QIB, 2D6Y, 4ICH, and 2IEK). It is difficult to determine if these TFR-ligand interactions are biologically relevant, but in at least one case, the genes adjacent to the TFR on the chromosome, and hence the predicted regulated genes (23) (see Genomics of TFRs above), indicate a potentially relevant relationship. The TFR Jann_2994 from the alphaproteobacterium Jannaschia sp. strain CCS1 was crystalized with stearic acid (PDB ID 3CJD). Jann_2994 is adjacent to a putative PspA homolog, which is potentially involved in regulating cytoplasmic membrane integrity as well as a putative fatty acid desaturase.

At least three TFRs, AmtR, DhaS, and SlmA, are known to interact with proteins rather than small-molecule ligands (GlnK, DhaQ, and FtsZ, respectively) (67–69). Residues of SlmA involved in interactions with the cell division protein FtsZ have recently been identified (70). These residues form an active site on the ligand-binding domain that in the crystal structure of SlmA is partially blocked by the DNA-binding domain (Fig. 8). The authors proposed that in the DNA-bound form of SlmA, the entire FtsZ interaction interface would be exposed, with implications for SlmA function (see TFRs and Cell Division below) (70).

Two TFRs, SczA and ComR, bind metals, but the molecular details of these interactions are not known (71, 72). Further structural studies will provide clues as to the mechanisms surrounding how TFRs specifically recognize metal ions.

Mechanism of Induction by Ligands

Structures are available for six TFRs (TetR, QacR, HrtR, DesT, SimR, and CgmR) in both DNA-bound and ligand-bound conformations, providing insight into the structural mechanism of derepression (22, 42–47, 50, 53). It has been noted that in most apo-protein structures, the TFR most closely resembles the ligand-bound or induced form rather than the DNA-bound form. It is therefore unlikely that the comparison of apo and ligand-bound structures provides a meaningful insight into induction and that it is necessary to compare the ligand-bound and DNA-bound forms.

In all cases, ligand binding is associated with increased separation of the DNA-binding domains of the two TFR monomers relative to the DNA-bound form. This may be accompanied by further conformational changes involving helix 4 and helix 6, which are in direct contact with the DNA-binding domain. In TetR, ligand binding causes a shift in helix 6 resulting in the pendulum-like motion of helix 4. Using equilibrium protein-unfolding experiments, Reichheld et al. (73) provided evidence that TetR does not occupy two distinct folded states (i.e., DNA bound and tetracycline bound) but rather that ligand binding increases the folding cooperativity between the N- and C-terminal domains. It was suggested by Reichheld et al. (73) that this increases the stability of a conformation in which the DNA-binding domains are too far apart to support binding to adjacent major grooves in the DNA.

Similar to the case for TetR, a pendulum-like motion was noted in helix 4 of both QacR and CgmR, but in these cases, ligand binding caused a coil-to-helix transition in helix 5 and a relocation of helix 6 in QacR. In CgmR, a widening of the inducer-binding pocket and shift in helix 6 was observed. In HrtR, heme binding was shown to cause a coil-to-helix transition in helix 4, resulting in a rigid-body motion of the DNA-binding domain to an orientation not compatible with DNA binding. The case of DesT is perhaps not as simple, as DNA-binding and induced forms of the protein are both bound to ligands, albeit different ones. In the DNA-free form (bound to the inducing ligand palmitoyl-CoA as opposed to oleoyl-CoA), a helix-to-coil transition of helix 4 is seen along with an ordering of the L8-L9 loop and movement of helix 6 and helix 7. These changes in DesT again result in a widening of the distance between the DNA-binding domains. SimR represents yet another variation, where there is no reorientation between the DNA-binding and ligand-binding domains but rather a rigid-body motion of the two SimR monomers relative to each other that results in a widening between the two DNA-binding domains.

It is difficult to posit a universal structural model for the transition between the repressing and induced conformations for TFRs, and indeed, it is unclear whether there are true commonalities throughout the family. Certainly key structural elements, including the conserved helices of the DNA-binding domain and the conserved helix 5 to 7 triangle and four-helix dimerization interface, are relevant. While at first glance it may be difficult to directly apply the Reichheld model for allosteric regulation of TetR (73) to SimR given that there is no reorientation between the DNA-binding and ligand-binding domains in the case of SimR, structural flexibility along the monomer interface may be important in this case. The ligand-binding cavity of SimR is composed of residues from both monomers, and as a result ligand binding will undoubtedly decrease the flexibility between them. Recent work has challenged the Reichheld model (74); however, this work was based primarily on X-ray crystallographic analysis of the protein bound to artificial peptide inducers and therefore should be interpreted with caution. Our view is that nuclear magnetic resonance (NMR) analysis of one or more TFRs, preferably for those where there are X-ray data on both the ligand- and DNA-bound forms (e.g., CgmR, DesT, HrtR, QacR, SimR, or TetR), in which the structural transitions that occur upon ligand binding are monitored would be an ideal means of settling debate in this area.

TFRs AND ANTIBIOTIC RESISTANCE

There are numerous TFRs involved in regulating resistance to antibiotics and other toxic compounds. These TFRs can be divided into three categories: (i) TFRs regulating self-resistance in antibiotic-producing organisms, (ii) TFRs regulating specific antibiotic resistance in nonproducing organisms, and (iii) TFRs regulating multidrug resistance.

TFRs Regulating Self-Resistance in Antibiotic-Producing Organisms

Numerous TFRs have been identified in the biosynthesis clusters for antibiotics and other secondary metabolites in species of Streptomyces and related actinobacteria. Of these, six TFRs, i.e., ActR, KijA8, LanK, PhlF, SimR, and VarR, have been shown to bind the products of the biosynthetic pathways in which they are encoded (25, 75–79). These TFRs primarily regulate the expression of efflux pumps required for antibiotic export but may also regulate the expression of late-stage biosynthetic genes.

Actinorhodin is a benzoisochromanequinone antibiotic produced by S. coelicolor. The biosynthetic pathway for this compound is encoded in a 22-kb region that includes the actR gene and its target operon actAB, which encodes two efflux pumps believed to export actinorhodin from the cell. The biosynthesis of actinorhodin involves a typical type II polyketide synthase that first generates an 18-carbon octaketide (80). This molecule is tailored into a 3-ring intermediate, and, late in the pathway, two of these intermediates are covalently joined to generate the mature six-ring compound. ActR binds both the final biosynthetic product actinorhodin and three-ring biosynthetic intermediates, including (S)-DNPA (79). Genetic evidence suggests that in actinorhodin-producing cells (S)-DNPA and/or other 3-ring intermediates serve to activate the expression of efflux genes, the only known self-resistance mechanism, before the final product is synthesized (81). Furthermore, there are now several reports that the export proteins are required for efficient, high-yield biosynthesis of actinorhodin (81, 82). The biochemical basis for reduced actinorhodin biosynthesis in cells defective in the actAB operon is not well understood, but it has been interpreted as evidence that initial activation of the actinorhodin export genes is primarily dependent on intermediates. However, it is also clear that sustained expression of the actinorhodin efflux pumps throughout the culture (i.e., including cells that produce actinorhodin and those that do not) requires the actinorhodin final product (81). Thus, actinorhodin is believed to act as a cell-cell signal to trigger export and resistance in nonproducing cells.

Like ActR, LanK and SimR are also able to bind both the final products of the biosynthetic pathways in which they are encoded and biosynthetic intermediates. LanK from Streptomyces cyanogenus S136 is located in the biosynthesis cluster for the glycosylated angucyclic polyketide antibiotic landomycin A. LanK regulates both the landomycin A efflux pump encoded by lanJ and the downstream gene lanZ1 (78). LanZ1 is an epimerase required for synthesis of sugar residues required for later-stage landomycin biosynthesis. Thus, as is the case for ActR, at least one step in the induction of the LanK target operon involves the interaction of the repressor with an immature landomycin intermediate. TFRs are also located in the biosynthesis clusters for the related angucyclinone antibiotics urdamycin and saquayamycin, but the role of these TFRs in regulating antibiotic biosynthesis and export has not been investigated (83, 84).

SimR is located in the biosynthesis cluster for simocyclinone D8 in Streptomyces antibioticus Tü 6040 (85, 86). Simocyclinone D8 is a structurally complex inhibitor of DNA gyrase (87, 88). The final molecule is composed of four parts: an angucyclic polyketide, a d-olivose sugar, a tetraene linker, and an aminocoumarin moiety. SimR regulates expression of the simocyclinone efflux pump encoded by simX and is induced by both simocyclinone D8 and the intermediate simocyclinone C4, which lacks the aminocoumarin functional group (76). It is not clear, however, that the interaction of SimR with the C4 intermediate is biologically relevant. Unlike the ActR case, where intermediates are bound more tightly than the finished product, or the LanK case, where induction is required for the completion of biosynthesis, the C4 intermediate binds more weakly than the mature D8 molecule, and there are no known biosynthetic steps that depend on the SimX export protein.

TFRs are encoded in many of the antibiotic biosynthesis gene clusters found in actinomycetes; however, they are also associated with the biosynthesis of other classes of compounds in a great many organisms. For example, PhlF is located in the 2,4-diacetylphloroglucinol biosynthesis cluster of Pseudomonas fluorescens (89). Biosynthesis of 2,4-diacetylphloroglucinol is of interest, as it occurs via a type III polyketide synthase (PKS) thought to be rare in bacteria (90). PhlF binds to the intergenic region between phlF and phlA, repressing expression of the phlABCD operon (75). DNA binding is enhanced in the presence of salicylate and disrupted by the biosynthetic product of the cluster 2,4-diacetylphloroglucinol.

TFRs are also present in the biosynthesis clusters for diverse polyketides, including ansamycins (e.g., rifQ in the rifamycin cluster [91]), macrolactams (e.g., mlaM in the ML-449 cluster [92]), and polyether ionophores (e.g., schR3 in the calcimycin cluster [93]). TFRs are not limited to polyketide biosynthesis clusters but are found in biosynthesis clusters for nonribosomal peptides (e.g., acmP and acmU in the actinomycin cluster [94]) and nucleoside antibiotics (e.g., amiP in the amicetin cluster [95]).

KijR and Pip from S. coelicolor are involved in regulating antibiotic resistance in a nonproducing organism (see below) and are closely related to KijA8 and VarR, respectively (25) (Fig. 9), raising the possibility that KijR and Pip were acquired by horizontal gene transfer from a producing organism. KijA8 regulates expression of the putative kijanimicin efflux pump encoded by kijA5 in response to kijanimicin in the producing organism Actinomadura kijaniata (25). Similarly, VarR regulates expression of the virginiamycin efflux pump in Streptomyces virginiae in response to virginiamycin S (77). Environmental microbes are highly resistant to antibiotics (96) and provide a reservoir for resistance elements in other environmental microbes as well as in clinically relevant pathogens (97, 98). This raises questions as to the possible origins of other TFRs, for example, TetR and MphR of E. coli, in producing organisms. No TFR has been identified in the biosynthesis clusters for oxytetracycline or chlorotetracycline, but a TFR is present in the cluster for the glycosylated anticancer tetracycline SF2575, although our analysis does not show a close relationship to TetR from E. coli (25, 99). The closest homolog of MphR is, however, found in the environmental microbe Myxococcus xanthus (25, 59), revealing possible origins for MphR in the environment.

Fig 9.

Grouping of TFRs involved in antibiotic resistance. (A and B) KijA8 and KijR (A) and VarR and Pip (B) group together in phylogenomics analysis, indicating that KijR and Pip may have been horizontally acquired from an antibiotic-producing organism. (C) Many TFRs controlling the expression of multidrug efflux pumps cluster together in phylogenomics analysis. (Adapted from reference 25.)

TFRs Regulating Specific Antibiotic Resistance in Nonproducing Organisms

The first characterized TFR, TetR, the founding member of the family, is the regulator of tetracycline resistance. Despite this, only a limited number of TFRs have been implicated in specific antibiotic resistance in nonproducing organisms: TetR, KijR, MphR, and Pip and the paralogous TFRs LmrA and QdoR.

Like for TetR, the target of Pip in S. coelicolor (SCO4025) is efflux pump gene of the major facilitator superfamily, pep (SCO4024) (100). Unlike tetR and tetA, pip and pep are cotranscribed. As discussed above, Pip shares a high degree of similarity with VarR in the viginiamycin biosynthesis cluster.

KijR and MphR regulate enzymes involved in antibiotic inactivation (25, 101). KijR regulates the expression of kijX, which encodes a novel antibiotic deglycosylase, and shares similarity with kijA8 in the kijanimicin biosynthesis cluster (see below) (25). MphR regulates expression of mphA, encoding a macrolide phosphotransferase, and mrx, encoding a membrane protein required for high-level resistance (101, 102). Another, unnamed TFR is found upstream of genes encoding a macrolide phosphotransferase (mphB) and a putative methyl esterase (rdmC-like) required for high-level macrolide resistance in some strains of E. coli as well as Streptococcus uberis (103). Despite the fact that they both regulate macrolide resistance genes, this unnamed TFR and MphR were found in separate groups in our analysis (25).

LmrA and QdoR are paralogous TFRs in Bacillus subtilis that bind plant flavonoids (104). LmrA and QdoR regulate expression of their own genes as well as those for LmrB, QdoI, and YxaH. The LmrA/QdoR regulon is organized into two operons: lmrA-lmrB and qdoR-qdoI-yxaH. LmrB is an efflux pump of the major facilitator superfamily. QdoI is a quercetin dioxygenase, responsible for flavonoid inactivation. YxaH is a membrane protein of unknown function.

Rather than regulating a specific antibiotic resistance mechanism, EthR from M. tuberculosis regulates the expression of EthA, an enzyme required for activation of the antibiotic ethionamide (105–108). While EthA is active against a broad range of substrates, including two other tuberculosis prodrugs, isoxyl and thiacetazone (106, 109, 110), the natural substrate for EthA is an unknown molecule believed to be distinct from ethionamide which is not an inducer of ethA expression. Due to its toxicity, ethionamide is currently used as a second-line drug primarily in the treatment of drug-resistant strains of tuberculosis. Activators of EthR are of interest for use in conjunction with ethionamide, as they would increase EthA expression, and therefore activation of ethionamide, allowing for lower ethionamide concentrations to be used (64).

TFRs Involved in Regulating Multidrug Resistance

TFRs are also involved in regulating a number of multidrug resistance pumps, including AcrAB in E. coli, which is regulated by AcrR, and MexXY from Pseudomonas aeruginosa, which is regulated by MexZ (111, 112). The AcrAB efflux pump in E. coli is under the control of several global regulators, including MarA, Rob, SoxS, and SdiA (113, 114). AcrR is thought to play a role in fine-tuning the expression of acrAB rather than serving as an on-off switch (112). Nevertheless, mutations in acrR alone result in increased expression of acrAB and are associated with antibiotic-resistant clinical isolates (115). AcrR has been shown to interact with various synthetic compounds, including ethidium, proflavine, and rhodamine 6G (116); however, the physiological relevance of these ligands for AcrR and other TFRs regulating multidrug resistance pumps such as QacR may be questionable. Clinically, the so-called multidrug resistance pumps, particularly those of the RND family, are a major source of antibiotic resistance in Gram-negative bacteria (117). However, multidrug resistance is typically the result of mutations in the regulators (118) of these pumps, indicating that multidrug resistance is not the native function of these pumps and that they serve other natural functions (119). Identifying bona fide interacting partners for the regulators of these pumps, whether they are small-molecule or protein ligands, will help to elucidate their roles under physiological conditions. A role for AcrAB in removing toxic metabolites has been suggested (120), and it would be interesting to test these putative acrAB inducers as ligands for AcrR.

The MexXY transporter of P. aeruginosa is expressed under conditions of ribosome stress, including the presence of antibiotics that target the ribosome (121). Expression of mexXY is controlled by the TFR MexZ and requires ArmZ (PA5471) (122), which interacts with MexZ (31, 123). armZ encodes a homolog of RtcB, an RNA ligase involved in recovery from stress-induced RNA damage (124), and is cotranscribed with PA5470, which encodes a homolog of PrfH. PfrH is thought to function as a peptide release factor that recognizes mRNA signals other than normal stop codons, possibly signals that result from RNA damage (125). While antibiotics that target the ribosome induce MexZ expression, MexZ does not appear to interact directly with these antibiotics but rather responds to effects downstream of ribosome disruption. Similarly, while the MexXY efflux pump functions as a multidrug efflux pump, its native function is not antibiotic efflux per se but rather its increased expression is a response to ribosome stress. Our phylogenetic analysis reveals a group containing many TFRs regulating putative multidrug efflux pumps (Fig. 9). This group includes, for example, AcrR and EnvR of E. coli, MexZ and NalD of P. aeruginosa, and MtrR of Neisseria gonorrhoeae. Further studies will be required to determine whether this shared grouping is indicative of a common interacting partner (small molecule or protein) for these TFRs and a common function for the efflux pumps that they regulate.

TFRs AND CELL-CELL SIGNALING

GBL Signaling

Gamma-butyrolactone (GBL) signaling molecules are involved in the regulation of antibiotic production and morphological development in Streptomyces and other actinomycetes and are the most well characterized signaling molecules in these species. A-factor from Streptomyces griseus was the first GBL to be characterized, and its identification predates that of the acyl-homoserine lactone quorum-sensing molecules of Gram-negative bacteria (126). The TFR ArpA is the A-factor receptor in S. griseus and is part of a large group of closely related TFRs (Fig. 10) that bind GBLs and related signaling molecules such as avenolide from Streptomyces avermitilis (127) and the methylenomycin furans from S. coelicolor (128). In some cases, such as that of ArpA in S. griseus, GBL signaling plays a major role in both antibiotic production and morphological development (9). In other cases, such as that of ScbR and the GBL SCB1 in S. coelicolor, some global effects have been noted; however, the predominant role of GBL signaling is in the regulation of a single antibiotic gene cluster (129, 130).

Fig 10.

All known TFRs involved in gamma-butyrolactone (GBL) signaling form a single group (data are from reference 25). Within the GBL group, a subclade of TFRs known as the “pseudo”-GBL receptors are highlighted with a yellow box.

The clustering of all known and predicted GBL receptors in our analysis shows the separate clustering of the so called “pseudo”-GBL receptors and helps to identify putative receptors not associated with GBL biosynthetic enzymes (Fig. 10). Pseudo-GBL receptors such as JadR2 from S. venezuelae and ScbR2 from S. coelicolor are reported to play a role in the GBL signaling alongside their cognate GBL receptor (i.e., JadR3 [SVEN_5968] and ScbR) by regulating expression of GBL biosynthesis enzymes (131, 132). GBL signaling systems regulate antibiotic biosynthesis, and one report suggests that pseudo-GBL receptors may interact with the final antibiotic product being regulated (133). While the biological relevance of these data is questionable due to the high concentration of antibiotic used in these studies, the idea that pseudo-GBL receptors play a role in GBL signaling pathways is an interesting one.

While the majority of GBL receptors and pseudoreceptors are associated with GBL biosynthetic enzymes, a number of orphan receptors, not associated with biosynthetic gene clusters or resistance genes, have also been identified. Our data provide support for previous reports concerning the role of some of these proteins, namely, CprA and CprB from S. coelicolor, in regulating secondary metabolite biosynthesis and morphological differentiation in Streptomyces (130, 134). Some bacteria are known to recognize and even metabolize the quorum-sensing signals produced by other bacteria (135). For example, E. coli and Salmonella enterica do not produce acyl-homoserine lactones but are able to sense them through the receptor SdiA (136). It is tempting to speculate that the role of orphan GBL receptors (e.g., CprA and CprB) and GBL receptors in bacteria not known to produce GBLs (e.g., MSMEG_2193 and MSMEG_2195) may be to recognize GBLs produced by other microbes.

Quorum Sensing

In Vibrio cholerae, the TFR HapR plays a major role in quorum-sensing regulation at high cell density (137). HapR orthologs in other species of Vibrio include LuxR of V. harveyi (not to be confused with transcription factors of the LuxR family such as LuxR of V. fischeri) and LitR of V. fischeri. The crystal structures of HapR as well as the orthologous SmcR are available (138, 139), and while they both show a putative ligand-binding cavity, none of the known quorum-sensing molecules have been reported to bind these proteins. Instead, expression of HapR is regulated at the posttranscriptional level through interactions of hapR mRNA with a number of small RNAs (sRNAs) (140). Recent data indicate an integration of cell density with nutrient availability in quorum sensing (141), raising the possibility that another type of small-molecule signal may serve as a ligand for HapR. The TFR DarR from M. smegmatis has been shown to interact with cyclic-di-AMP (142), providing a precedent for interactions between TFRs and second messengers.

TFRs AND CARBON METABOLISM

TFRs have been implicated in both central pathways for carbon metabolism as well as peripheral pathways for the catabolism of specific carbon sources, including the degradation of pollutants and other waste products (e.g., DhaR and MnbR [143, 144]). AcnR in Corynebacterium glutamicum controls the expression of the aconitase gene, acn. Aconitase is a tricarboxylic acid cycle enzyme that converts citrate to isocitrate and is thought to be an important control point in tricarboxylic acid cycle activity in Corynebacterium (145). Structures are available for AcnR (Protein Data Bank [PDB] ID 4AC6, 4ACI, and 4AF5) and show a bound molecule of citrate with evidence for another putative ligand-binding pocket (146). Further studies will be required to determine if AcnR indeed binds multiple small-molecule ligands.

Numerous TFRs have been identified as regulators of the expression of catabolic pathways for different carbon sources. For example, in Lactococcus lactis, DhaS regulates the expression of dihydroxyacetone kinase, which is required for glycerol catabolism (69). Unlike the majority of TFRs, DhaS interacts with a protein rather than a small molecule and acts as a transcriptional activator. MdoR from Mycobacterium sp. strain JC1 also acts as a transcriptional activator (147). MdoR regulates expression of the mdo gene, which is required for oxidation of methanol. The TFRs NicS, PaaR, and RolR are involved in the regulation of metabolism pathways for nicotinic acid, phenyl acetic acid, and resorcinol, respectively (148–150). In each case, the TFR has been shown to interact with the molecule being degraded or a catabolic intermediate.

TFRs AND NITROGEN METABOLISM

AmtR is a master regulator of nitrogen metabolism in Corynebacterium (37). The AmtR regulon is composed of at least 33 genes. These encode proteins that import and metabolize different nitrogen sources as well as other regulators of nitrogen metabolism. Unlike most of the characterized TFRs, AmtR interacts with a protein rather than a small-molecule ligand. Consistent with its role in controlling nitrogen assimilation, AmtR interacts with the adenylylated form of GlnK, which accumulates under conditions of nitrogen limitation (67). To date, residues important for this interaction have not been characterized. AmtR homologs are found in Actinobacteria, including some species of Mycobacterium, Nocardia, Rhodococcus, and Streptomyces (25, 151). In Streptomyces, two OmpR-like regulators, GlnR and GlnRII, are the master regulators of nitrogen metabolism (152). Although not present in Corynebacterium, GlnR homologs are more conserved in actinobacteria than AmtR homologs; however, there are a few species that encode both (151). Mycobacterium abscessus, Nocardia farcinica, Rhodococcus jostii, Streptomyces avermitilis, and Streptomyces scabies all encode both AmtR and GlnR homologs. In these species, the genes for the AmtR homologs are divergent to genes involved in the import and degradation of urea (Fig. 11A), suggesting that like AmtR, they may regulate nitrogen metabolism. The gene for another AmtR homolog in Mycobacterium smegmatis is located in an operon with genes for a putative enoyl-CoA hydratase and a putative fatty acid-CoA ligase, which do not play an obvious role in nitrogen metabolism. Bioinformatic analysis suggests that mycobacteria contain putative GlnR-binding sites throughout their chromosomes, while AmtR-binding sites in strains encoding AmtR homologs were not identified (151). It is tempting to speculate that in strains encoding both GlnR and AmtR homologs, AmtR may function as a local rather than a global repressor.

Fig 11.

TFRs involved in nitrogen metabolism. (A) Homologs of AmtR, a global regulator of nitrogen metabolism in Corynebacterium, may act as local regulators in related organisms. (B) RutR and PydR homologs from separate clades within a larger group of TFRs predicted to be involved in nucleotide metabolism. (C) Homologs of XdhR may be involved in purine metabolism. (Adapted from reference 25.)

In E. coli, RutR is the master regulator of pyrimidine metabolism. RutR regulates transcription of the divergently transcribed rut operon, encoding gene products involved in the degradation of pyrimidines for use as a nitrogen source. RutR also regulates a number of other targets located elsewhere on the chromosome, including the carAB, gadAX, gadBC, ygiF-glnE, and gcl-hyi-glxR operons, which are involved in various aspects of pyrimidine and glutamate metabolism (153). A homolog of RutR, PydR, has been identified as the regulator of genes required for pyrimidine degradation via the reductive pathway in Pseudomonas putida (154). A crystal structure is available for RutR in complex with its inducing ligand uracil (155). Residues involved in uracil binding are conserved in PydR, indicating that PydR may also be a uracil-responsive transcription factor (154). Our analysis shows separate RutR and PydR subclades within a larger group of TFRs probably involved in the metabolism of pyrimidines as well as possibly purines (Fig. 11B). P. aeruginosa encodes four RutR-PydR homologs, and based on genomic context, all four homologs are likely to play a role in nucleotide metabolism. This RutR-PydR group serves as an example of how very similar TFRs may regulate different pathways involved in the same overall physiological process, in this case nucleotide metabolism. Another example of this is the involvement of FabR and DesT in regulating fatty acid saturation in E. coli and P. aeruginosa, respectively (see TFRs and Lipid Metabolism below).

XdhR from S. coelicolor regulates the expression of the divergently transcribed four-gene operon encoding the subunits of, and a maturation factor for, xanthine dehydrogenase. Xanthine dehydrogenase activity is responsible for the conversion of xanthine to uric acid, which can be broken down and used as a nitrogen source. XdhR may therefore provide a link between primary metabolism, morphological development, and antibiotic production in Streptomyces (156), and our analysis shows that XdhR is indeed well conserved in this genus. The potato pathogen Streptomyces scabies carries two xdhR homologs (Fig. 11C). SCAB82081 encodes an ortholog of XdhR, while the role of SCAB83171 is unclear. We have also identified XdhR homologs in a number of Gram-negative bacteria (e.g., Pseudomonas putida and Rahnella). These XdhR homologs are predicted to regulate a putative short-chain dehydrogenase of unknown function. Given the phylogenetic grouping, these XdhR homologs and the short-chain dehydrogenases they regulate should be investigated for a role in purine metabolism.

TFRs AND LIPID METABOLISM

There are numerous parallels between the biosyntheses of polyketide antibiotics and fatty acids. We have noted above the involvement of TFRs in regulating resistance to numerous polyketide antibiotics, including tetracycline. TFRs also play a major role in regulating fatty acid metabolism as well as the metabolism of other lipid compounds, including sterols.

Fatty Acid Biosynthesis and Degradation

FasR from C. glutamicum is a regulator of lipid biosynthesis. In a fasR mutant, 17 genes were differentially expressed, including fasA, fasB, accB, accC, and accD1 (157). In addition, two other TFRs were found to be differentially expressed in the fasR mutant, one of which, Clg1640, may also play a role in fatty acid metabolism. Clg1640 is found in a group with FadR from Pseudonocardia autotropica (Fig. 12A). In P. autotropica, FadR controls an operon involved in fatty acid degradation (158). Also located in this group are the AtrA homologs from Streptomyces. AtrA is a pleiotropic regulator of antibiotic production in Streptomyces (6). The inducing ligand for AtrA is unknown, but given its grouping with two TFRs involved in fatty acid metabolism, a fatty acid derivative should be investigated. Transcriptomics studies support a role for AveI, the AtrA ortholog of Streptomyces avermitilis, in the regulation of carbon flux toward antibiotic production. Indeed, genes involved in fatty acid metabolism were downregulated in an aveI mutant (159). FasR itself is located in a group with another known regulator of antibiotic production in S. coelicolor, SCO1712 (160) (Fig. 12B), further highlighting a connection between fatty acid metabolism and antibiotic production. It is worth highlighting that AtrA acts as a transcriptional activator, in contrast to the role of most TFRs as repressors (6).

Fig 12.

TFRs involved in lipid metabolism. TFRs involved in lipid metabolism are found in many groups. (A to C) TFRs involved in fatty acid biosynthesis and degradation. (D) TFRs regulating fatty acid saturation. (E) TFRs involved in the synthesis and degradation of storage polymers. (F and G) TFRs involved in terpene utilization. (Adapted from reference 25.)

While FasR is currently the only known TFR involved in fatty acid biosynthesis, numerous TFRs are known to play a role in fatty acid degradation. In addition to FadR from P. autotropica (see above), two other TFRs have been called FadR. FadR from B. subtilis regulates five operons required for fatty acid degradation and recognizes long-chain fatty acyl-CoAs (161). FadR from Thermus thermophilus controls the expression of numerous genes implicated in fatty acid degradation (57). Although there are currently three TFRs bearing the name FadR, none grouped together in our analysis (25). While the available data support a role for all three TFRs in fatty acid degradation, the above proteins are clearly not orthologous.

In M. tuberculosis, at least two TFRs are known to play a role in lipid metabolism, Fad35R and Mce3R. Fad35R controls the expression of an acyl-CoA synthetase encoded by Fad35 in response to fatty acid derivatives (162). Mce3R represses the transcription of the virulence-related mce3 locus as well as other genes required for fatty acid degradation. Mce3R is among a group of TFRs containing duplicated TFR domains within a single peptide. Our analysis has shown that the N- and C-terminal TFR domains form separate clusters within the same group (Fig. 12C). This indicates that the N-terminal domains are more similar to each other than they are to the C-terminal domains, and this group of TFRs may be the result of a single duplication and gene fusion event. Although the inducing ligand for Mce3R has not been identified, it is located within a larger group with Fad35R (Fig. 12C), indicating that it may likewise be induced by a fatty acid derivative.

PsrA from P. aeruginosa responds to long-chain fatty acids to control expression of fad genes (163). In addition, PsrA plays a role in resistance to cationic antimicrobial peptides, antibiotic production, quorum sensing, and virulence, indicating that PsrA and long-chain fatty acids play an important role in the physiology of this important opportunistic pathogen.

Lipid Saturation

As mentioned above, FabR from E. coli and DesT from P. aeruginosa regulate different pathways involved in the same overall physiological process, in this case fatty acid saturation. Our phylogenetic analysis shows that they are located in the same group (25) (Fig. 12D). FabR regulates the expression of fabA and fabB, which are required for the synthesis of unsaturated fatty acids (164). The genes encoding FabA, FabB, and FabR are all located in different areas of the chromosome, and unlike most TFRs, FabR is not autoregulatory (165). While not essential for DNA binding by FabR, unsaturated thioesters (i.e., acyl-ACP or acyl-CoA) were found to enhance binding, while the FabR-DNA interaction was disrupted in the presence of saturated thioesters (165). DesT shows a similar pattern of ligand binding, where DNA binding is enhanced by unsaturated acyl-CoAs and disrupted by saturated acyl-CoAs (166). DesT regulates the expression of desC and desB, which are divergently transcribed from desT (166). The desC and desB genes encode a reductase and an acyl-CoA desaturase, respectively. Whereas FabR regulates the biosynthesis of unsaturated fatty acids, DesT regulates gene products required for the desaturation of preformed acyl chains.

Synthesis and Degradation of Storage Polymers

In Pseudomonas putida, PhaD controls the expression of genes involved in polyhydroxyalkanoate (PHA) metabolism (167). PHAs are produced as carbon storage granules and are being investigated for their potential as alternative plastics (168). PHA polymers are synthesized from (R)-3-hydroxyacyl-CoA, which can be produced from various intermediates of fatty acid degradation. Although not experimentally demonstrated, PhaD is thought to bind a fatty acyl-CoA intermediate of β oxidation (167). Interestingly, PhaD is located in a larger group with NGO0393 and NMB0810 from Neisseria gonorrhoeae and Neisseria meningitidis, respectively (Fig. 12E). The NGO0393 and NMB0810 orthologs are one of only two TFRs encoded by each species. The roles of NGO0393 and NMB0810 in the metabolism of storage polymers have not been investigated.

Terpene Utilization

Terpenes, including cholesterol, are an important class of natural product built from isoprene units. Two TFRs, KstR and KstR2, control cholesterol degradation in M. tuberculosis (169–172). The genes in the KstR and KstR2 regulons are known to be upregulated in vivo and are important for virulence of M. tuberculosis (173). Specific inducing ligands for KstR and KstR2 have not been identified. Although KstR and KstR2 are located in separate groups (Fig. 12F and G), other TFRs in both groups further suggest a role in terpene metabolism. For example, Rv0767c is in the same group as KstR and is located in an operon with Rv0764c, encoding a putative steroid demethylase. KstR2 is in a group with AtuR from P. aeruginosa. AtuR controls the expression of genes required for acyclic terpene utilization (174). A number of TFRs (CampR, CmtI, CmtR, CymR, and PsbI) that control the utilization of cyclic terpenes such as camphor and p-cymene have also been identified (175–179).

TFRs AND AMINO ACID METABOLISM

Three TFRs, AguR, LiuQ, and McbR, are involved in regulating amino acid metabolism. While AguR and LiuQ act as local regulators controlling the expression of adjacent genes involved in amino acid degradation, McbR acts as a global regulator for sulfur metabolism.

Agmatine is an intermediate in the arginine decarboxylase (ADC) pathway for arginine degradation (180). In P. aeruginosa PAO1, the TFR AguR controls expression of the aguAB operon, involved in agmatine utilization, and is induced by agmatine (181). AguR is conserved in many members of the genus Pseudomonas but is absent from P. aeruginosa PA14, where an alternative operon for agmatine metabolism also plays a role in biofilm formation (182).

The TFR LiuQ has been identified using comparative genomics as the regulator of the liuABCD genes, required for the degradation of branched-chain amino acids in Burkholderiales (183). Although DNA binding was not experimentally tested, a putative LiuQ binding site was identified.

McbR from C. glutamicum is a global regulator of sulfur metabolism, including the genes required for the biosynthesis of sulfur-containing amino acids (184, 185). DNA binding by McbR is modulated by the small molecule S-adenosylhomocysteine, a by-product of methylation reactions (185). Deletion of mcbR causes numerous pleiotropic effects on additional aspects of growth and metabolism, indicating that McbR plays a central role in regulating numerous physiological processes in C. glutamicum (186).

TFRs AND COFACTOR METABOLISM

Biotin

In many bacteria, BirA controls the expression of biotin biosynthesis genes (187). BirA is a bifunctional protein acting as the biotin-protein ligase as well as a transcriptional regulator. In some organisms, BirA lacks the transcriptional regulatory domain and the biotin biosynthesis genes are controlled by other transcription factors. In alphaproteobacteria, the GntR family regulator BioR controls the expression of biotin biosynthesis genes (188), while in Corynebacterium and certain related actinobacteria, the TFR BioQ is the regulator of biotin biosynthesis (189). BioQ interacts with a 13-bp palindromic region upstream of a number of biotin biosynthesis genes. Biotin itself was not found to disrupt the BioQ-DNA interaction in vitro, but this does not rule out the possibility of a biotin intermediate serving as a ligand for BioQ.

Heme

Two TFRs, HrtR from Lactococcus lactis and HemR from Propionibacterium freudenreichii, have been implicated in heme homeostasis. HemR is a putative regulator of genes required for the conversion of glutamate to protoheme in P. freudenreichii. It is transcribed divergently from hemX, encoding a putative heme transporter. The details of HemR DNA binding and ligand binding have not been investigated (190). L. lactis does not synthesize heme. HrtR senses intracellular heme and regulates expression of a heme exporter encoded by HrtBA (191).

TFRs AND CELL DIVISION

SlmA from E. coli binds and antagonizes polymerization of the bacterial tubulin homolog FtsZ, preventing cell division from occurring over the chromosome, in a process known as nucleoid occlusion (68, 192, 193). SlmA is not known to interact with a small molecule but rather interacts directly with the tubulin-like cell division protein FtsZ. This interaction is believed to be important for preventing the formation of cell division septa around unsegregated chromosomes. In one model for SlmA function (70), the FtsZ interaction interface on SlmA is completely exposed only when SlmA is bound to DNA. Hence SlmA affect FtsZ polymerization only in areas where DNA is present.

Based on our analysis (25), SlmA homologs are found in most members of the gamma- and betaproteobacteria. Exceptions in which this mechanism appears to be absent include Acinetobacter, Francisella, Legionella, Pseudomonas, Stenotrophomonas, Xanthomonas, and Neisseria. A possible SlmA ortholog in Bordetella parapertussis was found to be the product of a pseudogene. In E. coli, the absence of both the Min system, also involved in regulating FtsZ function, and SlmA results in a synthetic lethal phenotype (68). How FtsZ ring placement is regulated in Gram-negative bacteria such as Myxococcus, Campylobacter, and Bacteroides where both the Min system and SlmA are absent is unknown, but evidence suggests that there are as yet-unidentified factors involved in nucleoid occlusion (194).

In Bacillus, Noc, a protein unrelated to SlmA, controls nucleoid occlusion. Noc is a homolog of ParB chromosome-partitioning proteins. The TFR RefZ, however, is involved in regulating the transition from medial to polar cell division during sporulation, possibly as a direct effector of FtsZ polymerization (195). Our analysis does not indicate a close relationship between SlmA and RefZ.

FUTURE DIRECTIONS AND CHALLENGES

TFRs play an important role in regulating numerous aspects of bacterial physiology. Through genomics and structural studies, we have learned a great deal regarding the types of gene products regulated by TFRs and the mechanisms by which TFRs interact with both DNA and small molecules. Although genomics allows us to predict the target genes for the majority of TFRs, this cannot be done for type III TFRs, and other methodologies must be employed. Different models still exist as to the structural changes that TFRs undergo upon ligand binding and the precise molecular mechanisms behind derepression, and future NMR studies may help to resolve discrepancies in the current data. Determination of the identities of the small-molecule ligands, or other interacting partners, bound by the more than 200,000 TFRs in the public databases probably represents the most understudied and challenging area of TFR biology, and future work will be required to identify these ligands.

Biographies

Leslie Cuthbertson received her B.Sc. degree in molecular biology and genetics from the University of Guelph. She continued on to Ph.D. studies in microbiology with Chris Whitfield, where her studies were focused on the export of polysaccharides in Gram-negative bacteria. Upon completing her Ph.D., Dr. Cuthbertson began postdoctoral studies at McMaster University with Justin Nodwell, where she became interested in microbial chemical biology and more specifically how bacteria use transcription factors to recognize small-molecule signals. She believes that understanding how bacteria “see” their environment will be crucial in combatting pathogens while encouraging the growth of beneficial microflora.

Justin Nodwell is a Professor and Chair in the Department of Biochemistry at the University of Toronto. He trained in molecular biology and bacterial genetics with Jack Greenblatt and Richard Losick and worked from 1998 to 2013 at McMaster University, Department of Biochemistry and Biomedical Sciences and Michael DeGroote Institute for Infectious Diseases Research. His research involves the responses of bacterial cells to biologically active small molecules and the regulatory mechanisms that control secondary metabolism. Goals of this work include understanding antibiotic resistance and identifying new antimicrobial compounds.