[Frontiers in Bioscience 4, d132-140, February 1, 1999]

Current Issue

Received: 8/26/98
Accepted: 11/8/98

Send correspondence to:

Dr Philip N. Rather,
Departments of Medicine and of Molecular Biology and Microbiology,
Case Western Reserve University,
10900 Euclid Ave,
Cleveland, OH 44106,

Tel: 216-368-0744,
Fax: 216-368-2034,

E-mail: pxr17@po.cwru.edu, David.macinga@spcorp.com


Providencia, Aminoglycoside Resistance, Gene Regulation


Copyright © Frontiers in Bioscience, 1995


David R. Macinga 1 and Philip N. Rather 1, 2

1 Division of Infectious Diseases, 2 Departments of Medicine and of Molecular Biology and Microbiology, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH 44106


1. Abstract
2. Introduction
2.1 Aminoglycoside resistance mechanisms.
2.2 Aminoglycoside resistance genes.
3. AAC(2’)-Ia in Providencia stuartii
3.1. Identification of the AAC(2’)-Ia enzyme
3.2. Isolation of the aac(2’)-Ia gene.
3.3. Distribution of aac(2’)-Ia.
3.4 Physiological functions for aac(2’)-Ia.
4. aac(2’)-Ia regulation
4.1. Strategies for isolation of regulatory genes.
4.2. Regulatory genes.
4.2.1. aarA
4.2.2. aarB
4.2.3. aarC
4.2.4. aarD
4.2.5. aarE
4.2.6. aarF
4.2.7. aarG
4.2.8. aarP
6. Role of quorum sensing in AAC(2’)-IA regulation
7. Perspective
8. Acknowledgements
9. Reference


Instrisic chromosomal acetyltransferases involved in aminoglycoside resistance have been identified in a number of bacteria. In Providencia stuartii, a chromosomal acetyltransferase (AAC(2’)-Ia) has been characterized in detail. In addition to the ability to acetylate aminoglycosides, the AAC(2’)-Ia enzyme has at least one physiological function, which is the acetylation of peptidoglycan. This modification is likely to influence the autolytic system in P. stuartii. The regulation of aac(2’)-Ia expression is extremely complex involving at least seven regulatory genes acting in at least two pathways. This complexity in regulation indicates that aac(2’)-Ia expression must be tightly controlled in response to different environmental conditions. This presumably reflects the importance of maintaining correct levels of peptidoglycan acetylation. In this review, a summary of data will be presented involving both the physiological and genetic aspects of aac(2’)-Ia in P. stuartii .


2.1. Aminoglycoside resistance mechanisms

The aminoglycosides are a large family of antibiotics which have been used extensively in the clinic since the introduction of Streptomycin over 50 years ago.

Aminoglycosides inhibit protein synthesis by binding to the ribosome and are bactericidal (1, 2, 3). Bacteria may acquire aminoglycoside resistance by decreased uptake/permeability, alterations of the ribosome, or through the acquisition of aminoglycoside modifying enzymes (4, 5, 6, 7). The first two mechanisms are rare and usually account for low-level resistance. High level resistance arising from the acquisition of aminoglycoside modifying enzymes is the most common and clinically important mechanism (6). Modification by aminoglycoside modifying enzymes results in inactivation of the aminoglycoside by decreasing it’s affinity for ribosomes (8). Three types of aminoglycoside modifying enzymes have been described: the aminoglycoside phosphotransferases (APH) which phosphorylate hydroxyl groups; the aminoglycoside nucleotidyltransferases (ANT) which adenylylate hydroxyl groups; and the aminoglycoside acetyltransferases (AAC) which acetylate amino groups. The aminoglycoside modifying enzymes are further subdivided based on the position of the aminoglycoside that is modified. The AAC(2')-Ia protein reviewed here acetylates amino groups present at the 2' carbon of the deoxystreptamine core (9, 10, 11). For a detailed description of the nomenclature of the aminoglycoside modifying enzymes see Shaw et al., 1993 (6).

2.2. Aminoglycoside resistance genes

The majority of genes encoding aminoglycoside modifying enzymes are carried on plasmids or other mobile genetic elements (6). This association has contributed to the rapid spread of aminoglycoside resistance throughout the bacterial kingdom. Recently, a number of chromosomally encoded N-acetyltransferases have been identified which are not associated with mobile genetic elements. These include AAC(2')-Ia from Providencia stuartii (10), AAC(2')-Ib from Mycobacterium fortuitum (12), AAC(2')-Ic from Mycobacterium tuberculosis, AAC(2')-Id from Mycobacterium smegmatis (13), AAC(6’)-Ic from Serratia marcescens (14), AAC(6’)-If from Acinetobacter sp. 13 (15), AAC(6’)-Ig from Acinetobacter haemolyticus (16), AAC(6’)-Ii from Enterococcus faecium (17), and AAC(6’)-Ik from Acinetobacter sp. 6 (18). In each case, the resistance gene has been found to be universally present in the species from which they were identified, regardless of the aminoglycoside resistance phenotype, suggesting that they may be involved in a housekeeping function separate from acetylating aminoglycosides.


3.1. Identification of the AAC(2’)-Ia enzyme

Early studies by both Chevereau et. al., and Yamaguchi et. al. led to the identification of the AAC(2’) enzyme in Providencia (9, 11), referred to hereafter as AAC(2’)-Ia. Studies using the purified enzyme from Providencia strain GN1544 demonstrated that AAC(2’)-Ia exhibited a pH optimum of 6.5 for lividomycin-A inactivation, and that the enzyme lost activity upon exposure to 65oC for 5 minutes (11). Interestingly, the pH optimum was dependent on the substrate used in the assays. The failure to transfer the AAC(2’) resistance phenotype by conjugation, together with the inability to eliminate the resistance phenotype from Providencia strains by curing with acridine orange or ethidium bromide, led to the first suggestion that the aac(2’)-Ia gene was chromosomally encoded (11). In a subsequent study it was shown that all aminoglycoside sensitive P. stuartii strains possessed low-level 2’-N-acetylating activity by the phosphocellulose binding assay (19). This data suggested that these P. stuartii strains all contained the aac(2’)-Ia gene and expressed it at low-levels. Furthermore, McHale et al. demonstrated no direct correlation between AAC(2’) resistance profiles and the presence of plasmid DNA in clinical isolates of P. stuartii (20).

Studies by Swiatlo and Kocka suggested that aac(2’)-Ia expression in P. stuartii was inducible by exposure to aminoglycosides (21). In this study, two sensitive P. stuartii isolates were serially passed in media containing increasing amounts of gentamicin. By the fifth transfer, cells had gentamicin MICs that were at least 20-fold higher than the starting isolate and AAC(2’) enzyme activity showed a 10-fold increase. Passage of these highly resistant isolates for five transfers, in the absence of aminoglycoside, resulted in both the gentamicin MIC and enzyme activities returning to the normal levels seen in the starting isolates. This led the authors to conclude that the AAC(2’)-Ia expression is inducible by the presence of aminoglycosides. However, it was never shown if exposure to subinhibitory amounts of aminoglycoside increased AAC(2’)-Ia expression. This would have determined if a true induction event took place. Data from our lab does not support these results. We have demonstrated that AAC(2’)-Ia expression is not inducible by aminoglycosides as previously suggested (21). The conversion of sensitive P. stuartii isolates to those with high-level aac(2’)-Ia expression, primarily involves mutations in a number of regulatory genes which are discussed in detail below.

3.2. Isolation of the aac(2’)-Ia gene

The ability of the AAC(2’)-Ia enzyme to acetylate genatmicin was exploited to clone the aac(2)-Ia structural gene in the gentamicin sensitive E. coli strain DH5a (10). A library of partially digested Sau3AI fragments was prepared in pACYC184 from a clinical P. stuartii isolate (SCH75082831A), referred to hereafter as PR50 (10). Recombinants containing the aac(2’)-Ia gene were identified on plates containing 5 mg/ml gentamicin. E. coli recombinants containing the aac(2’)-Ia gene expressed a low level resistance to gentamicin, tobramycin, netilmicin and 6’-N-netilmicin, but not to 2’-N-netilmicin. This resistance profile was consistent with 2’-N-acetylating activity and phosphocellulose binding assays confirmed acetylation of 6-N-netilmicin, whereas 2’-N-netilmicin was not acetylated (10). DNA sequence analysis of the cloned insert and mutagenesis experiments allowed for identification of the aac(2’)-Ia open reading frame which is predicted to encode a protein of 20.1 kDa. The deduced AAC(2’)-Ia protein displays extensive similarity to a family of 2’-N-acetyltransferases identified in a number of Mycobacterial species and limited similarity to the AAC(6’)-Ic protein of Serratia marcescens .

3.3. Distribution of aac(2’)-Ia in Providencia

The prevelance of aac(2’)-Ia or related genes in other Providencia species has been investigated using Southern hybridization at low stringency. As expected, the aac(2’)-Ia gene was present in all isolates of P. stuartii tested (22). Sequences with homology to aac(2’)-Ia were also observed in Proteus penneri and Providencia rettgeri (22).

3.4. Physiological functions for AAC(2’)-Ia in P. stuartii

The universal presence of the aac(2’)-Ia gene in P. stuartii suggested a physiological role in metabolism. The similarity of aminoglycosides to cellular substrates containing amino sugars, such as N-acetylglucosamine and N-acetylmuramic acid led to the proposal that AAC(2’)-Ia may be involved in processes related to peptidoglycan or lipopolysaccharide (LPS) metabolism (10, 22, 23, 24). Reactions involving the acetylation of peptidoglycan have been demonstrated in P. stuartii and a relationship between AAC(2’)-Ia and peptidoglycan acetylation has been established by the work of K. Payie and A. Clarke (23, 24). Analysis of P. stuartii mutants with increased aac(2’)-Ia expression has revealed a corresponding increase in the levels of O-acetylation. Interestingly, biochemical analysis of the AAC(2’)-Ia enzyme has confirmed that it is also capable of acetylating aminoglycosides using acetate obtained from either acetyl CoA, soluble peptidoglycan fragments or N-acetylglucosamine (24). An additional phenotype associated with aac(2’)-Ia overexpression is altered cell morphology with cells appearing as shortened coccobaccili or as chains of cells (23).

To address the physiological function of AAC(2’)-Ia, a P. stuartii mutant (PR100) has been constructed which contains a frameshift mutation in the aac(2’)-Ia gene. This frameshift is predicted to result in a loss of function as it would result in a truncated polypeptide missing almost the entire COOH terminal portion of AAC(2’)-Ia. The aac(2’)-Ia frameshift results in a reduction in the intrinsic levels of aminoglycoside resistance from 8 mg/ml to 0.5 mg/ml and phosphocellulose binding assays have revealed a complete loss of gentamcin acetylation in extracts prepared from PR100 (24). Thus AAC(2’)-Ia is the sole acetyltransferase in the cell for aminoglycoside acetylation. The loss of the AAC(2’)-Ia enzyme resulted in a significant decrease in the levels of peptidoglycan O-acetylation (42%), relative to the 54% seen in the isogenic wild-type parent (23). The residual levels of O-acetylation are likely to result from a second O-acetyltransferase, which appears to be responsible for the majority of peptidoglycan O-acetylation. Furthermore, the loss of AAC(2’)-Ia results in significant changes in cell morphology with the formation of distorted rod-shaped cells which failed to constrict during cell division. These cells also displayed altered staining properties with uranyl acetate (23). It is known that O-acetylation blocks the activity of autolytic enzymes (muramidases) involved in peptidoglycan turnover. Moreover, O-acetylation also appears to be required for the activity of other autolysins (25, 26). These phentoypes observed in the aac(2’)-Ia frameshift mutant are consistent with altered levels of O-acetylation leading to significant changes in the autolytic system involved in cell wall turnover.


4.1. Strategies for isolation of regulatory genes

Studies on aac(2’)-Ia regulation have been conducted primarily at the transcriptional level. Primer extension analysis has been used to identify a promoter for aac(2’)-Ia which contains a -10 consensus sequence (TATAAT) for the 70 form of RNA polymerase and the sequence CTTTTT at the -35 region (10). This -35 sequence does not conform to known consensus sequences and predicts that aac(2’)-Ia transcription may require ancillary factors (see aarP below). The regulation of aac(2’)-Ia has been studied using transcriptional lacZ fusions to the aac(2’)-Ia promoter region. In a wild-type P. stuartii strain, such as PR50, the expression of an aac(2’)-lacZ fusion is low, resulting in a pale blue colony phenotype on X-gal plates. Using this fusion, we have determined that transcription of aac(2’)-Ia is not inducible by aminoglycosides (10). However, aminoglycoside resistant mutants arise a high frequency (10-6-10-7) when selected at 4X the MIC for gentamicin. In addition, the majority of these mutants display a dark blue phenotype on X-gal plates indicating increased transcription of the aac(2’)-Ia gene. This demonstrates that the mutations leading to aminoglycoside resistance are trans-acting since they simultaneously activated the chromosomal copy of aac(2’)-Ia and the aac(2’)-lacZ fusion present on a low copy plasmid. Selection of spontaneous and mini-Tn5Cm induced mutations which activate both copies of the aac(2’)-Ia promoter has been our strategy to identify negative regulators of aac(2’)-Ia. This approach has led to the identification of a complex regulatory network that involves genes designated aar (aminoglycoside acetyltransferase regulator) which are summarized in table 1.

Table 1. Regulatory genes of aac(2’)-Ia




Negative regulator. Probable integral membrane protein. Involved in response to AR-factor


Negative regulator. Identity is unknown.


Negative regulator at high cell density. Highly conserved in bacteria. Essential gene.


Negative regulator. CydD homolog


Positive effector. UbiA homolog required for ubiquinone biosynthesis.


Positive effector. Novel locus required for ubiquinone biosynthesis.


Negative regulator. Sensor kinase.


Transcriptional activator. Related to the MarA/SoxS family.

4.2. Regulatory genes

4.2.1. aarA

The aarA gene was identified as a mini-Tn5Cm insertion that increased expression of an aac(2’)-lacZ fusion 3-4 fold in liquid growth conditions (27). Loss of function mutations in aarA also resulted in a gentamicin resistance level that was increased 8-fold above wild-type. Null mutations in aarA are highly pleiotrophic and additional phenotypes include; loss of production of a diffusible yellow pigment and altered morphology with abberant cell separation after division. This results in a very distinctive cell chaining phenotype that is most prominent in cells at mid-log phase. Furthermore, the effects of the aarA mutation are much stronger when cells are grown on agar plates, where the expression of an aac(2’)-lacZ fusion is increased 8-10 fold. The AarA polypeptide is 31.1 kDa in size and very hydrophobic with at least two possible transmembrane domains. Homology searches of the databases with AarA resulted in no significant matches to other proteins. Genetic evidence implicates AarA in a pathway required for response to an extracellular pheromone signal, AR-factor, that acts to reduce aac(2’)-Ia expression (see below). Furthermore, an additional P. stuartii gene (cma37) which is regulated by quorum sensing is strongly dependent on a functional aarA gene for expression. Studies with this fusion again indicate that AarA is required for cells to sense the extracellular signal which activates the cma37 fusion. The stronger phenotypes associated with aarA deletions seen on solid media, relative to liquid growth are consistent with a role in response to an extracellular signal due to the increased accumulation of AR-factor in the surrounding agar vs. diffusion in liquid. Therefore, the pleiotrophic phenotypes, such as loss of pigmentation and defective cell division seen in cells with the aarA gene deleted, may result from a defect in sensing the quorum signal AR-factor.

4.2.2. aarB

The aarB3 mutation originally designated aar3 (10) results in a 10 to 12-fold increase in aac(2’)-Ia transcription. In the aarB3 background, the levels of aminoglycoside resistance are increased 128-fold above wild-type, suggesting that this mutation further increases aminoglycoside resistance in a manner independent of aac(2’)-Ia expression. The aarB3 mutation results in a small colony phenotype and cells which are dramatically shortened. The identity of the aarB gene remains to be determined. However, the use of a plasmid library to complement the aarB mutation has resulted in the identification of a high copy suppressor which contains the P. stuartii hemB homolog. This raises the possibility that the aarB mutation is in a component of the electron transport chain. This would be consistent with the levels of aminoglycoside resistance increased in a manner that is not proportional to aac(2’)-Ia expression.

4.2.3. aarC

The aarC gene was identified by a mutation (aarC1) which simultaneously activated the chromosomal aac(2’)-Ia gene and a plasmid encoded aac(2’)-lacZ transcriptional fusion (28). The deduced AarC protein is 40 kDa and is highly conserved to a family of proteins that is widespread in bacteria. The E. coli homolog is GcpE and studies in our lab have demonstrated that gcpE is essential for E. coli viability and have also shown that aarC is an essential gene in P. stuartii (28). Furthermore, complementation experiments have shown that aarC and gcpE are functionally equivalent. The missense allele, aarC1, results in a number of pleiotrophic phenotypes including; slow growth, altered cell morphology, and increased aac(2’)-Ia expression at high cell density. The biochemical function of AarC remains to be determined.

4.2.4. AarD

The aarD gene is a trans-acting negative regulator of aac(2')-Ia which was identified as a mini-Tn5Cm insertion resulting in the activation of an aac(2’)-IacZ transcriptional fusion (29). The mini-Tn5Cm insertion (designated aarD1) results in a 5-fold activation of the aac(2’)-lacZ fusion, a 3-fold increase in the levels of aac(2’)-Ia mRNA accumulation, and a 32-fold increase in aminoglycoside resistance over that of wild-type P. stuartii.

The aarD locus has been cloned by complementation and encodes two polypeptides, AarD and OrfX, which exhibit extensive homology to the Escherichia coli CydD and CydC proteins respectively (29, 31, 32). The CydD and CydC proteins comprise a heterodimeric ABC transporter complex which is involved in formation of a functional cytochrome d oxidase complex (32, 33, 34). Mutations in cydD and cydC result in the loss of the cytochrome d oxidase both spectroscopically and functionally (32, 34, 35). P. stuartii aarD mutants exhibit phenotypic characteristics consistent with a defect in the cytochrome d oxidase including hyper-susceptibility to the respiratory inhibitors Zn2+ and toluidine blue (29, 30, 34). Introduction of the E. coli cydDC genes into the aarD1 background leads to complementation of all mutant phenotypes suggesting that the two loci are functional homologues (29).

The increased aac(2')-Ia expression observed in the aarD1 background contributes minimally to the overall increase in gentamicin resistance since introduction of the aarD1 mutation into an aac(2')-Ia mutant strain also results in a 32-fold increase in gentamicin resistance. Since previous studies have demonstrated that uptake of aminoglycosides is dependent on the presence of a functional electron transport system (7, 36, 37), and since electron transport is defective in the aarD1 background (29), it is probable that a decrease in aminoglycoside uptake accounts for the high level of resistance observed in aarD mutants. It seems unlikely that aarD plays a direct role in the regulation of aac(2')-Ia since ABC transporters are not known to function as transcriptional regulators (38). It has recently been proposed that aac(2')-Ia expression is influenced by an uncharacterized regulatory pathway that responds to changes in the redox state of the membrane (see below) (39). Mutations in aarD are predicted to alter the redox state of the membrane and thus indirectly affect aac(2')-Ia expression.

An interesting phenotype of the aarD1 mutation is sensitivity to a self-produced extracellular factor (29). This phenotype is not unique to P. stuartii, as E. coli mutants (cydD, cydAB) lacking cytochrome d oxidase are also sensitive to a self-produced extracellular factor. The identity of this factor has not been established.

4.2.5. AarE

The aarE gene was identified by selecting for P. stuartii mutants resistant to gentamicin (40). The aarE1 allele resulted in a level of gentamicin resistance that is increased to 256 mg/ml, relative to the 4 mg/ml observed in the isogenic parent. Surprisingly, the accumulation of aac(2’)-Ia mRNA was significantly reduced in the aarE1 background. Analysis of the aarE gene has shown it to be the ubiA homolog, which encodes an octaprenyltransferase required for the second step of ubiquinone biosynthesis. The loss of ubiquinone function is predicted to decrease the uptake of aminoglycosides, which is likely to explain the high-level aminoglycoside resistance. The decreased aac(2’)-Ia mRNA accumation may reflect a requirement for uniquinone, either directly or indirectly in a regulatory process involved in aac(2’)-Ia mRNA stability.

4.2.6. AarF

The aarF locus of P. stuartii is a positive regulator of aac(2')-Ia expression with the level of aac(2')-Ia mRNA being dramatically decreased in an aarF null mutant (39). Despite the lack of aac(2')-Ia expression, aarF null mutants exhibit a 256-fold increase in gentamicin resistance over the wild-type strain. P. stuartii aarF null mutants also exhibit severe growth defects under aerobic growth conditions and have been found to lack detectable quantities of the respiratory cofactor ubiquinone.

The wild-type aarF gene has been cloned and encodes a 62.5 kDa polypeptide which exhibits extensive amino acid identity to two putative adjacent open reading frames from Escherichia coli designated yigQ and yigR (39, 41). Disruption of the yigR gene has confirmed that this locus is required for ubiquinone production in E. coli (39). Heterologous complementation studies demonstrate that aarF and the E. coli yigQR loci are functionally equivalent. Three ubiquinone biosynthesis genes, ubiB, ubiD, and ubiE, have been mapped near yigQR at minute 86 on the E. coli chromosome (41, 42, 43, 44). Complementation experiments with known ubi mutants have demonstrated that the yigQR locus is genetically distinct from ubiB, ubiD and ubiE suggesting that yigQR (aarF) represents a novel locus required for ubiquinone production.

Previous studies have shown that ubiquinone deficient E. coli mutants accumulate gentamicin poorly and as a result exhibit increased gentamicin resistance (36, 37). Therefore, the high-level gentamicin resistance observed in the aarF and yigR mutants is likely associated with decreased accumulation of the drug resulting from the absence of aerobic electron transport. It seems unlikely that aarF is directly involved in the regulation of aac(2')-Ia. It has been proposed that a reduced form of ubiquinone acts as an effector molecule in an uncharacterized regulatory pathway that activates the expression of aac(2')-Ia (39). In ubiquinone deficient aarF mutant strains, this regulatory cascade would be disrupted resulting in decreased aac(2')-Ia expression (see below).

4.2.7. AarG

The aarG gene was identified in a genetic screen for negative regulators of aac(2’)-Ia. A recessive mutation (aarG1) results in an 18-fold increase in the expression of b-galactosidase from an aac(2’)-lacZ fusion (45). Direct measurments of RNA from the chromosomal copy of aac(2’)-Ia have confirmed this increase at the level of RNA accumulation. Taken together, these results demonstrate that loss of aarG results in increased aac(2’)-Ia transcription. The aarG1 allele also results in enhanced expression of aarP, encoding a transcriptional activator of aac(2’)-Ia (45). Genetic experiments have shown that in an aarG1, aarP double mutant, the expression of aac(2’)-Ia is significantly reduced over that seen in the aarG1 background. However, the levels of aac(2’)-Ia in this double mutant are still significantly higher than in a strain with only an aarP mutation. Therefore, the aarG1 mutation increases aac(2’)-Ia expression by both aarP dependent and independent mechanisms.

The aarG1 allele also confers a multiple antibiotic resistance phenotype (Mar) to P. stuartii resulting in increased resistance to tetracycline, chloramphenicol and fluoroquinolones. This Mar phenotype in the aarG1 background is partially due to overexpression of aarP, which is know to confer a Mar phenotype in both P. stuartii and E. coli (see below). However, a mechanism independent of aarP overexpression also accounts for increased levels of intrinsic resistance in the aarG1 background. This mechanism could involved increased expression of a second activator with a target specificity similar to that of AarP.

The aarG gene encodes a protein with similarity to sensor kinases of the two-component family with the strongest identity to PhoQ (57%). Immediately upstream of AarG is an open reading frame designated aarR which encoded a protein with 75% amino acid identity to PhoP, a response regulator (46). The regulatory phenotypes associated with the aarG1 mutation may result from a failure to phosphorylate the putative response regulator AarR, which functions as a repressor of aarP, and possibly aac(2’)-Ia.

4.2.8. aarP

A central component in the activation of aac(2')-Ia expression is a small transcriptional regulator designated AarP. The aarP gene was originally isolated from a multicopy library of P. stuartii chromosomal DNA based on the ablility to activate aac(2')-Ia expression in trans (47). The presence of aarP in multiple copies led to an 8-fold increase in aac(2')-Ia mRNA accumulation. Studies utilizing an aac(2')-lacZ transcriptional fusion demonstrate that this increase results from an activation of aac(2')-Ia transcription. Chromosomal disruption of the aarP locus results in a fivefold reduction in aac(2')-Ia mRNA levels and eliminates the induction of aac(2')-Ia expression normally observed during logarithmic growth (48). These studies indicate that aarP is required for the normal expression pattern of aac(2')-Ia observed in wild-type P. stuartii. Furthermore, expression of aarP has been shown to be increased in the aarB, aarC and aarG mutants, demonstrating that aarP contributes to the overexpression of aac(2')-Ia in these mutant backgrounds (10, 28, 45).

The aarP gene encodes a 16 kDa protein which contains a putative DNA binding helix-turn-helix motif and belongs to the AraC/XylS family of transcriptional activators (47, 49). Several lines of evidence indicate that the AarP protein directly interacts with the aac(2’)-Ia promoter region to activate transcription (48). In vivo transcriptional activation studies utilizing a series of 5’ deletion derivatives of the aac(2’)-Ia promoter demonstrate that sequences extending to -67 relative to the transcriptional start are required for activation by AarP. A 4-base insertion at -47 in the context of the full length promoter abolishes activation of the aac(2’)-Ia promoter by AarP. Purified AarP protein binds to a wild-type aac(2’)-Ia promoter fragment in electrophoretic mobility shift assays, but does not to bind a derivative containing the 4 base insertion at -47. Finally, DNaseI footprint analysis indicates that AarP protects a region of the aac(2')-Ia promoter extending from -29 to -46. This protected region partially overlaps the -35 region of the aac(2’)-Ia promoter (10) suggesting that AarP functions as a Class II activator and mediates activation of aac(2’)-Ia by interaction with the sigma subunit of RNA polymerase (50).

The AarP protein exhibits extensive homology with the E. coli MarA and SoxS proteins which are activators involved in the multiple-antibiotic-resistance phenotype and the oxidative stress response (47, 51- 55). Previous research has demonstrated that there is overlap in the in vivo targets for MarA and SoxS (52, 56- 60). This overlap is thought to be the result of the high degree of similarity in the helix-turn-helix domain of these proteins (54). AarP, which exhibits high homology to MarA and SoxS in the helix-turn-helix domain, was found to activate targets of both MarA and SoxS in vivo (47). The putative AarP binding site displays some similarities with the proposed binding sites for MarA and SoxS (48).

Expression of aarP appears to be governed by a mechanism which differs from those controlling MarA and SoxS expression. Unlike the MarA and SoxS proteins, which are located in operons containing a gene which regulates their expression, the aarP message appears to be monocystronic. Expression of aarP was found to be slightly elevated in the presence of tetracycline but was not elevated in the presence of a potent inducer of MarA, salicylate (47). Recent studies of aarP expression have revealed that the AarP message accumulates as cell density increases (48). Furthermore, the addition of spent culture media leads to increased aarP message accumulation suggesting that aarP is subject to regulation by a density dependent cell signaling mechanism.


The regulation of aac(2’)-Ia expression is also subject to control by cell to cell signaling or quorum sensing (61). The accumulation of aac(2’)-Ia mRNA exhibits two levels of growth phase dependent expression. First, as cells approach mid-log phase, a significant increase is observed relative to cells at early-log phase. This increase at mid-log phase is the result of increased aarP expression. Second, as cells approach stationary phase, the levels of aac(2’)-Ia mRNA are decreased to levels that are at least 20-fold lower than those at mid-log phase. This decrease at high density is mediated by the accumulation of an extracellular factor (AR-factor) (61). The growth of P. stuartii cells in spent (conditioned) media from stationary phase cultures resulted in the premature repression of aac(2’)-Ia in cells at mid-log phase. The analysis of AR-factor has shown it to be between 500 and 1000Da in size, heat stable and sensitive to proteases. These characteristics are consistent with a small peptide, although the exact structure has not been determined.


The chromosomal aac(2’)-Ia gene was originally identified as an aminoglycoside resistance gene. The universal presence of this gene in P. stuartii led to the identification of a role for AAC(2’)-Ia in the O-acetylation of peptidoglycan. The possibility of additional roles for the AAC(2’)-Ia enzyme in cellular metabolism, such as a role in LPS acetylation, have been suggested (10, 23, 24). However, the additional roles of AAC(2’)-Ia in P. stuartii, if any, remain to be identified.

The complex regulatory networks controlling aac(2’)-Ia expression presumably reflect the importance of maintaining correct levels of AAC(2’)-Ia expression and subsequent O-acetylation of peptidoglycan. One pathway of aac(2’)-Ia regulation involves a quorum sensing signal AR-factor. The response to AR-factor involves the putative integral membrane protein AarA, which may serve in a transport complex for AR-factor. In addition, the aarC gene may be involved in the AR-factor regulatory pathway, as aarC mutations only display a phenotype in cells at high density. A second pathway for aac(2’)-Ia appears to involve an unidentified component of the electron transport chain, possibly a ubiquinone derivative. A model has been presented in which loss of the aarD gene may be predicted to lead to an accumulation of the reduced form of ubiquinone (ubiquinol) by preventing the formation of the cytochrome bd oxidase complex (39). Consistent with this model, aarD mutations lead to increased aac(2’)-Ia expression and two mutations which block ubiquinone production (aarE, aarF) result in reduced aac(2’)-Ia expresssion.

A third pathway of regulation involves the transcriptional activator aarP, which is a central activator of aac(2’)-Ia. The aarP gene is subject to negative regulation by the aarB, aarC and aarG genes. The AarG sensor kinase acts to negatively regulate aarP and aac(2’)-Ia by possibly phosphorylating the response regulator AarR. The signal which alters the kinase/phosphatase activity of AarG is unknown. However, the identification of this signal will be crucial for our understanding of regulation of aac(2’)-Ia and of the multiple antibiotic resistance phenotype accompanied by AarP overexpression. In addition to the mutations described, we have identified additional regulatory mutations that remain to be characterized. The analysis of these genes may provide important clues regarding the physiological processes which are coupled to aac(2’)-Ia regulation.


Work in my lab on the aac(2’)-Ia gene has been supported by the National Institutes of Health, the National Science Foundation and the Department of Veterans Affairs.


1. Anderson, P., Davies, J. & B. D. Davis: Effect of Spectinomycin on polypeptide synthesis in extracts of Escherichia coli. J. Mol. Biol. 29, 203-215 (1967)

2. Davis, B. D., Tai, P.-C. & B. J. Wallace: Complex interactions of antibiotics with the ribosome. In Ribosomes, pp. 771-789. Edited by M. Nomura, A. Tissieres & P. Lengyel. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory (1974)

3. Flaks, J. G., Cox, E. C. & J. R. White: Inhibition of polypeptide synthesis by streptomycin. Biochem. Biophys. Res. Com. 7, 385-389 (1962)

4. Benveniste, R. & J. Davies: Mechanisms of antibiotic resistance in bacteria. Ann. Rev. Biochem. 42, 471-506 (1973)

5. Foster, T. J.: Plasmid-derived resistance to antimicrobial agents. Microbiol. Rev. 47, 361-409 (1983)

6. Shaw, K. J., Rather, P. N., Hare, R. S. & G. H. Miller: Molecular genetics of aminoglycoside resistance genes and familial relationships of the aminoglycoside-modifying enzymes. Microbiol. Rev. 57, 138-163 (1993)

7. Taber, H. W., Mueller, J. P., Miller, P. F. & A. S. Arrow: Bacterial uptake of aminoglycoside antibiotics. Microbiol. Rev. 51, 439-457 (1987)

8. Dickie, P., Bryan, L. E. & M. A. Pickard: Effect of enzymatic adenylylation on dihydrostreptomycin accumulation in Escherichia coli carrying an R-factor: model explaining aminoglycoside resistance by inactivating mechanisms. Antimicrob. Agents Chemother. 14, 569-580 (1978)

9. Chevereau, M., Daniel, P. J. L., Davies, J. & F. LeGoffic: Aminoglycoside resistance in bacteria mediated by gentamicin acetyltransferase II, an enzyme modifying the 2'-amino group of aminoglycoside antibiotics. Biochemistry 13, 1141-1156 (1974)

10. Rather, P. N., Orosz, E., Shaw, K. J., Hare, R. & G. H. Miller: Characterization and transcriptional regulation of the 2'-N-acetyltransferase gene from Providencia stuartii. J. Bacteriol. 175, 6492-6498 (1993)

11. Yamaguchi, M Mitsuhashi, S. Kobayashi, F. & H. Zenda: A 2'-N-acetylating enzyme of aminoglycosides. J. Antibiotics 27, 507-515 (1974)

12. Aínsa, J. A., Martin, C., Gicquel, B. & R. Gomez-Lus: Characterization of the chromosomal aminoglycoside 2'-N-acetyltransferase gene from Mycobacterium fortuitum. Antimicrob. Agents Chemother. 40, 2350-2355 (1996)

13. Aínsa, J. A., Pérez, E., Pelicic, V., Berthet, F.-X., Gicquel, B. & C. Martin: Aminoglycoside 2'-N-acetyltransferase genes are universally present in mycobacteria: characterization of the aac(2')-Ic from Mycobacterium tuberculosis and the aac(2')-Id gene from Mycobacterium smegmatis. Mol. Microbiol. 24, 431-441 (1997)

14. Shaw, K. J., Rather, P. N., Sabatelli, F., Mann, P., Munayyer, H., Mierzwa, R., Petrikkos, G., Hare, R. S., Miller, G. H., Bennett, P. & P. Downey: Characterization of the chromosomal aac(6')-Ic gene from Serratia marcescens. Antimicrob. Agents Chemother. 36, 1447-1455 (1992)

15. Lambert, T., Germand, G. & P. Courvalin: Characterization of the chromosomal aac(6')-If gene of Acinetobacter sp.13 and the aac(6')-Ih plasmid gene of Acinetobacter baumannii. Antimicrob. Agents Chemother. 38, 1003-1011 (1994)

16. Lambert, T., Germand, G., Galimand, M. & P. Courvalin: Characterization of an Acinetobacter haemolyticus aac(6')-Ig gene encoding an aminoglycoside 6'-N-acetyltransferase which modifies amikacin. Antimicrob. Agents Chemother. 37, 2093 (1993).

17. Costa, Y., Galimand, M., Leclercq, R., Duval, J. & P. Courvalin: Characterization of the chromosomal aac(6')-Ii gene specific for Enterococcus faecium. Antimicrob. Agents Chemother. 37, 1896-1903 (1993)

18. Rudant, E., Bourlioux, P. & P. Courvalin: Characterization of the aac(6')-Ik gene of acinetobacter sp.6. FEMS Microbiol. Lett. 124, 49-54 (1994)

19. Miller, G., Sabatelli, F. Hare, R. & J. Waitz: Survey of aminoglycoside resistance patterns. Developments Industrial Microbiol. Ch.8 91-104 (1980)

20. McHale, P. J., Keane, C. T., & G. Dougan: Antibiotic resistance in Providencia stuartii isolated in hospitals. J. Clin. Microbiol. 13, 1099-1104 (1981)

21. Swiatlo, E. & F. Kocka: Inducible expression of an aminoglycoside-acetylating enzyme in Providencia stuartii. J. Antimicrob. Chemother. 19:27-30 (1987)

22. Clarke, A. J. Francis, D. & W. J. Keenleyside: The prevelance of gentamicin 2’-N-acetyltransferase in the Proteeae and its role in the O-acetylation of peptidoglycan. FEMS Microbiol. Letters 145, 201-207 (1996)

23. Payie K. G. Rather P. N. & A. J. Clarke: Contribution of gentamicin 2’-N-acetyltransferase to the O-acetylation of peptidoglycan in Providencia stuartii. J. Bacteriol. 177, 4303-4310 (1995)

24. Payie K. G. & A. J. Clarke: Characterization of gentamicin 2’-N-acetyltransferase from Providencia stuartii: its use of peptidoglycan metyabolites for acetylation of both aminoglycosides and peptidoglycan. J. Bacteriol 179, 4106-4114 (1997)

25. Clarke, A. J. & C. Dupont : O-acetylated peptidoglycan: its occurrence, pathobiological significance, and biosynthesis. Can. J. Microbiol 38, 85-91 (1992)

26. Clark, A. J. Extent of peptidoglycan O-acetylation in the tribe Proteeae. J. Bacteriol. 175, 4550-4553 (1993)

27. Rather P. N., & E. Orosz: Characterization of aarA, a pleiotrophic negative regulator of the 2’-N-acetyltransferase in Providencia stuartii. J. Bacteriol. 176, 5140-5144 (1994)

28. Rather P. N. Solinsky, K., Paradise, M. R. & M. M. Parojcic: aarC, an essential gene involved in density dependent regulation of the 2’-N-acetyltransferase in Providencia stuartii. J. Bacteriol. 179, 2267-2273 (1996)

29. Macinga, D. R. & P. N. Rather: aarD, a Providencia stuartii homologue of cydD: role in 2'-N-acetyltransferase expression, cell morphology and growth in the presence of an extracellular factor. Molec. Microbiol. 19, 511-520 (1996)

30. Delaney, J. M., Wall, D. & C. Georgopoulos: Molecular characterization of the Escherichia coli htrD gene: cloning, sequence, regulation and involvement with cytochrome d oxidase. J. Bacteriol. 175, 166-175 (1993)

31. Poole, R. K., Hatch, L., Cleeter, M. W. J., Gibson, F., Cox, G. B. & G. Wu: Cytochrome bd biosynthesis in Escherichia coli: the sequences of the cydC and cydD genes suggest that they encode the components of an ABC membrane transporter. Molec. Microbiol. 10, 421-430 (1993)

32. Bebbington, K. J. & H. D. Williams: Investigation of the role of the cydD gene product in the production of a functional cytochrome d oxidase. FEMS Microbiol. Lett. 112, 19-24 (1993)

33. Poole, R. K., Gibson, F. & G. Wu: The cydD gene product, component of a heterodimeric ABC transporter, is required for assembly of periplasmic cytochrome c and of cytochrome bd in Escherichia coli. FEMS Microbiol. Lett. 117, 217-224 (1994)

34. Poole, R. K., Williams, H. D., Downie, J. A. & F. Gibson: Mutations affecting the cytochrome d-containing oxidase complex of Escherichia coli K-12: identification and mapping of a fourth locus, cydD. J. Gen. Microbiol. 135, 1865-1874 (1989)

35. Georgiou, C. D., Fang, H. & R. B.Gennis: Identification of the cydC locus required for expression of the functional form of the cytochrome d terminal oxidase complex in Escherichia coli. J. Bacteriol. 169, 2107-2112 (1987)

36. Bryan, L. E. & S. Kwan: Roles of ribosomal binding, membrane potential, and electron transport in bacterial uptake of streptomycin and gentamicin. Antimicrob. Agents Chemother. 23, 835-845 (1983)

37. Bryan, L. E. & H. M.van den Elzen: Effects of membrane energy mutations and cations on streptomycin and gentamicin accumulation by bacteria: a model for entry of streptomycin and gentamicin in susceptible and resistant bacteria. Antimicrob. Agents Chemother. 12, 163-177 (1977)

38. Faith, M. J. & R. Kolter: ABC transporters: bacterial exporters. Microbiol. Rev. 57, 995-1017 (1993)

39. Macinga, D. R., Cook, G. M., Poole, R. K. & P. N. Rather: Identification and characterization of aarF, a locus required for production of ubiquinone in Providencia stuartii and Escherichia coli and for expression of 2'-N-acetyltransferase in P. stuartii. J. Bacteriol. 180, 128-135 (1998)

40. Paradise M. R., Cook G, Poole R. K. & P. N. Rather: aarE, a homolog of ubiA is required for transcription of the aac(2’)-Ia gene in Providencia stuartii. Antimicrob. Agents. Chemother 42:959-962 (1998)

41. Daniels, D. L., Plunkett, G., Burland, V. & F. R. Blattner: Analysis of the Escherichia coli genome: DNA sequence of the region from 84.5 to 86.5 minutes. Science 257, 771-777(1992)

42. Cox, G. B., Young, I. G., McCann, L. M. & F. Gibson, F: Biosynthesis of ubiquinone in Escherichia coli K-12: Location of genes affecting the metabolism of 3-octaprenyl-4-hydroxybenzoic acid and 2-octaprenylphenol. J. Bacteriol. 99, 450-458(1969)

43. Lee, P. T., Hsu, A. Y., Ha, H. T. & C. F. Clarke: A C-methyltransferase involved in both ubiquinone and menaquinone biosynthesis: isolation and identification of the Escherichia coli ubiE gene. J. Bacteriol. 179, 1748-1754 (1997)

44. Young, I. G., McCann, L. M., Stroobant, P. & F. Gibson: Characterization and genetic analysis of mutant strains of Escherichia coli K-12 accumulating the ubiquinone precursors 2-octaprenyl-6-methoxy-1,4-benzoquinone and 2-octaprenyl-3-methyl-6-methoxy-1,4-benzoquinone. J. Bacteriol. 105, 769-778 (1971)

45. Rather, P. N., Paradise, M. R., Parojcic, M. M. & S. Patel: A regulatory cascade involving aarG, a putative sensor kinase, controls the expression of the 2'-N-acetyltransferase and an intrinsic multiple antibiotic resistance (Mar) response in Providencia stuartii. Molec. Microbiol. 28, 1345-1353 (1998)

46. Groisman E. A., Heffron F. & F. Solomon. Molecular genetic analysis of the Escherichia coli phoP locus. J. Bacteriol. 174, 486-491 (1992)

47. Macinga, D. R., Parjocic, M. M. & P. N. Rather: Identification and analysis of aarP, a transcriptional activator of the 2'-N-acetyltransferase in Providencia stuartii. J. Bacteriol. 177, 3407-3413 (1995)

48. Macinga, D. R: Regulation of the 2'-N-acetyltransferase of Providencia stuartii by a small transcriptional acivator AarP and by genes involved in aerobic electron transport. Ph.D. Dissertation. Department of Molecular Biology and Microbiology. Case Western Reserve University (1997)

49. Ramos, J. L., Rojo, F., Zhou, L. & K. N Timmis: A family of positive regulators related to the Pseudomonas putida TOL plasmid XylS and the Escherichia coli AraC activators. Nucl. Acids Res. 18, 2149-2152 (1990)

50. Ishihama, A: Protein-protein communication within the transcription apparatus. J. Bacteriol. 175, 2483-2489 (1993)

51. Amabile-Cuevas, C. F. & B. Demple: Molecular characterization of the soxRS genes of Escherichia coli: two genes control a superoxide stress regulon. Nucl. Acids Res. 19, 4479-4484 (1991).

52. Cohen, S. P., McMurry, L. M. & S. B. Levy: marA locus causes decreased expression of OmpF porin in multiple-antibiotic-resistant (Mar) mutants of Escherichia coli. J. Bacteriol. 170, 5416-5422 (1988)

53. Cohen, S. P., Yan, W. & S. B. Levy: A multidrug resistance regulatory locus is widespread among enteric bacteria. J. Infect. Dis. 168, 484-488 (1993)

54. Gambino, L., Gracheck, S. J. & P. F. Miller: Overexpression of the MarA positive activator is sufficient to confer multiple antibiotic resistance in Escherichia coli. J. Bacteriol 175, 2888-2894 (1993)

55. Wu, J. & B. Weiss: Two divergently transcribed genes, soxR and soxS, control a superoxide response regulon of Escherichia coli. J. Bacteriol. 173, 2864-2871 (1991)

56. Ariza, R. R., Cohen, S. P., Bachhawat, N., Levy, S. B. & B. Demple: Repressor mutations in the marRAB operon that activate oxidative stress genes and multiple antibiotic resistance in Escherichia coli. J. Bacteriol. 176, 143-148 (1994)

57. Chou, J. H., Greenberg, J. T. & B. Demple: Posttranscriptional repression of Escherichia coli OmpF protein in response to redox stress: positive control of the micF antisense RNA by the soxRS socus. J. Bacteriol. 175, 1026-1031 (1993)

58. Liochev, S. I. & I. Fridovich: Fumarase C, the stable fumarase of Escherichia coli, is controlled by the soxRS regulon. Proc. Natl. Acad. Sci. USA 89, 5892-5896 (1992)

59. Miller, P. F., Gambino, L. F., Sulavik, M. C. & S. J. Gracheck: Genetic relationship between soxRS and mar loci in promoting multiple antibiotic resistance in Escherichia coli. Antimicrob. Agents Chemother. 38, 1773-1779 (1994)

60. Rosner, J. L. & J. L. Slonczewski: Dual regulation of inaA by the multiple antibiotic resistance (Mar) and superoxide (SoxRS) stress response systems of Escherichia coli. J. Bacteriol. 176, 6262-6269 (1994)

61. Rather P. N., Paradise, M. R. & M. M. Parojcic: An extracellular factor regulating expression of the chromosomal aminoglycoside 2’-N-acetyltransferase in Providencia stuartii. Antimicrob. Agents Chemother.; 41, 1749-1754 (1997)