The Major Aminoglycoside-Modifying Enzyme AAC(3)-II Found in Escherichia coli Determines a Significant Disparity in Its Resistance to Gentamicin and Amikacin in China

Abstract

The aim of this study was to investigate the prevalence of aminoglycoside-modifying enzymes in Escherichia coli in different areas of China and to explore the relationship between pandemic enzyme type and bacterial resistance to antimicrobial agents in China. Gentamicin- or etimicin-resistant clinical isolates of E. coli were collected from different areas of China, and the in vitro antibacterial activity of 11 aminoglycoside agents was determined using standard (Clinical and Laboratory Standards Institute) agar dilution methods. Twelve aminoglycoside-modifying enzyme genes were detected by PCR and confirmed by DNA sequencing. A total of 205 E. coli strains were collected from nine hospitals in seven cities. All strains were highly resistant to gentamicin or etimicin, whereas resistance to tobramycin, netilmicin, and kanamycin was slightly lower. However, less than 15% of isolates were resistant to amikacin and isepamicin. Of the gentamicin-resistant strains, 88.2% and 86.7% were sensitive to isepamicin and amikacin, respectively. Five aminoglycoside-modifying enzyme genes were detected in 191 strains, whereas the remaining 14 strains were negative. The most common gene type was aac(3)-II (162 strains), followed by aac(6′)-I (50 strains), ant(3″)-I (28 strains), aph(3′)-II (20 strains), and ant(2″)-I (20 strains). Ninety-five strains yielded aac(3)-II only, whereas the others contained two or three genes. The three main gene combinations were aac(6′)-I/aac(3)-II (28 strains), aac(3)-II/ant(3″)-I (11 strains), and aac(3)-II/aac(6′)-I (10 strains). Regional bacterial resistance and enzyme distribution were roughly similar, although minor differences were found in Guangzhou, Jinan, and Dalian, which were the sources of most of the amikacin- or isepamicin-resistant strains. Chinese clinical isolates of E. coli remain highly resistant to gentamicin and etimicin, but are susceptible to amikacin and isepamicin. The dominant type of aminoglycoside-modifying enzyme, AAC(3)-II, might be the main source of the disparity in E. coli resistance to different aminoglycoside agents.

Introduction

Aminoglycosides are broad-spectrum antimicrobial agents with potent antibacterial activity and clear bactericidal effects. Combinations of aminiglycosides and β-lactams are used in the treatment of a variety of severe and resistant bacterial infections,^1,2 and thus the emergence of bacterial resistance to aminoglycosides is inevitable because of their widespread use. The rate of gentamicin resistance in Escherichia coli is over 40% in China and this seriously hinders the clinical use of this agent.^3,4

Bacterial mechanisms of resistance to aminoglycosides include decreased drug influx, antibacterial target mutation, drug modification, and active efflux. Aminoglycoside-modifying enzymes are the most important resistance factor, and these include acetyltransferase, nucleotidyltransferases, and phosphotransferases. Almost 100 of these enzymes are known, and the distribution and molecular types of the enzymes differs by region. This has resulted in a disparity in bacterial resistance to individual aminoglycoside agents, which determines the selection of aminoglycosides for use in clinics.^5,6 Bacterial resistance surveillance in China indicates that gentamicin or kanamycin resistance is severe in E. coli, but the rates of resistance to amikacin remain lower than 20%. Thus, the resistance mechanisms leading to regional differences are worthy of investigation.^3,4

Materials and Methods

Bacterial strains

A total of 205 nonduplicated clinical strains of gentamicin- or etimicin-resistant E. coli were collected from nine hospitals in southern (Guangzhou), eastern (Wenzhou and Hangzhou), and northern (Taiyuan, Beijing, Dalian, Jinan) China between January, 2008, and August, 2009. Strains were isolated from patients with infections, including urinary tract infections (103 strains, 50.2%), blood infections (32 strains, 15.6%), and respiratory infections (70 strains, 34.2%). All strains were screened using a Kirby–Bauer (KB) disk diffusion test to determine their sensitivity to gentamicin and etimicin (10 μg). Strains that produced inhibition zones with a diameter of less than 12 mm using either disk were retained for further investigation. Escherichia coli ATCC25922 was used as quality control strain in the study.

Antibacterial agents and reagents

Mueller–Hinton agar (MHA) and Mueller–Hinton broth (MHB) was purchased from Oxoid (Oxoid Limited, Hampshire, UK). Kanamycin, sisomicin, neomycin, ribostamycin, etimicin, netilmicin, and isepamicin were purchased from the National Institute for Quality Control of Pharmaceutical & Biological Products (Beijing, China). Amikacin was purchased from Qilu Pharmaceutical Co., Ltd. (Shangdong, China). Gentamicin was purchased from Ruixin Pharmaceutical Co., Ltd. (Zhejiang, China). Streptomycin was purchased from Dalian Merro Pharmaceuticals Co., Ltd. (Dalian, China). Tobramycin was a product of Alcon Laboratories, Inc. Genomic DNA extraction kits were purchased from Axygen Company (Hangzhou, China). PCR kits were purchased from Takara Company (Dalian, China).

Aminoglycoside minimum inhibitory concentration determination

Minimum inhibitory concentration (MIC) was determined using the standard agar dilution method of Clinical and Laboratory Standards Institute (CLSI; M02–A10).⁷ Briefly, 1–2 μl of a 10-fold diluted 0.5 McFarland unit fresh overnight culture of E. coli [approximately 1.0×10⁷ colony-forming units) CFU ml⁻¹] were inoculated onto MHA plates with serial double-diluted concentration of antibiotics, which resulted in 1.0–2.0×10⁴ CFU per inoculation point. The MIC was read after 18 hr incubation at 35°C. The MIC was defined as the lowest concentration capable of visual inhibition of bacterial growth.

Amplification and confirmation for genes of aminoglycoside-modifying enzymes

PCR amplification of following 12 common aminoglycoside-modifying enzyme genes was conducted: aph(3′)-II, aph(3′)-III, aph(3′)-VI, aac(6′)-I, aac (3)-I, aac (3)- II, aac(3)-III, aac(3)-IV, ant(2″)-I, ant(3″)-I, ant(4′)-Ia, and aac(6′)-aph(2″). Primers for PCR amplification were designed by reference to published methods.^8–10 The total PCR reaction volume was 50 μl, which included: 5 μl of 10× PCR buffer, 3 μl of MgCl₂, 1 μl of dNTP, 0.5 μl of DNA polymerase, 1 μl of primer, 5 μl of bacterial genomic DNA template, and sterile water to make the reaction volume up to 50 μl. PCR cycles were as follows: 94°C pre-denaturation for 5 min; followed by 94°C denaturation for 30 sec and annealing for 40 sec; 72°C extension for 40 sec, for 30 cycles; and finally 72°C extension for 5 min. PCR products were analyzed by electrophoresis using 1% agarose gel containing ethidium bromide, and visualized with ultraviolet (UV) light. Purified PCR products were sequenced and compared to GenBank for molecular type confirmation.

Results

MIC of aminoglycosides

A total of 205 strains of E. coli were collected from Guangzhou (22 strains), Wenzhou (36 strains), Hangzhou (55 strains), Taiyuan (19 strains), Beijing (22 strains), Jinan (42 strains), and Dalian (nine strains). All isolates were significantly resistant to gentamicin, streptomycin, and etimicin, followed by tobramycin, kanamycin, and netilmicin, while the resistance rates to amikacin and isepamicin were the lowest (Table 1). The amikacin-resistant bacterial strains came mainly from Jinan, Dalian, and Guangzhou (Group 2; the other cities made up Group 1). A total of 195 gentamicin-resistant strains were highly sensitive to isepamicin (172/195, 88.2%) and amikacin (169/195, 86.7%). Only one of 10 gentamicin-sensitive strains was resistant to amikacin.

Table 1.

Minimum Inhibitory Concentration (mg/L) of Aminoglycosides against Escherichia coli and Antimicrobial Resistance

	Percentage (%)			MIC (mg/L)			aac(3)-II(+)		aac(3)-II(–)		Group 1 ^c			Group 2
Agents	R ^a	I ^a	S ^a	50% ^b	90% ^b	Range ^b	50%	90%	50%	90%	50%	90%	Range	50%	90%	Range
SM	70.2	15.6	14.1	64	>256	2–>256	64	>256	64	>256	64	>256	2–>256	64	>256	2–>256
NM				4	128	<0.125–>256	4	128	4	>256	4	128	0.125–>256	4	128	1–>256
RB				8	>256	<0.125–>256	8	>256	64	>256	8	>256	2–>256	8	>256	2–>256
TB	44.9	20.0	35.1	8	256	≤0.5–>256	8	64	32	>256	8	64	≤0.5–>256	16	>256	≤0.5–>256
KN	45.4	6.3	48.3	32	>256	<0.125–>256	16	>256	64	>256	32	>256	0.125–>256	64	>256	4–>256
GT	95.1	1.5	3.4	64	>256	≤2–>256	64	256	16	>256	4	>256	≤2–>256	32	>256	≤2–>256
SS				64	>256	<0.125–>256	64	>256	32	>256	64	>256	<0.125–>256	64	>256	8–>256
ET	76.5	13.7	9.8	32	>256	<0.125–>256	32	>256	32	>256	32	>256	0.125–>256	32	>256	8–>256
NT	29.3	27.8	42.9	16	>256	≤0.5–>256	16	128	16	>256	16	128	≤0.5–>256	16	>256	≤0.5–>256
AK	11.7	1.5	86.8	2	>256	<0.125–>256	2	16	2	>256	2	8	0.125–>256	2	>256	<0.5–>256
IP	11.2	—	88.8	1	256	≤0.5–>256	1	2	1	>256	1	2	≤0.5–>256	1	256	≤0.5–>256

The breakpoints of etimicin and isepamicin refer to those of netilmicin and amikacin, respectively.

R, I, and S indicteresistance, intermediate, and susceptibility.

50%, 90%, and range meaning MIC₅₀, MIC₉₀, and MIC_range.

Group 1 included Hangzhou, Wenzhou, Beijing, and Taiyuan; Group 2 included Guangzhou, Jinan, and Dalian.

SM, streptomycin; NM, neomycin; RB, ribostamycin; TB, tobramycin; KN, kanamycin; GT, gentamicin; SS, sisomicin; ET, etimicin; NT, netilmicin; AK, amikacin; IP, isepamicin.

Distribution of aminoglycoside-modifying enzyme genes in E. coli

Five aminoglycoside-modifying genes were detected in 191 drug-resistant strains, whereas the remaining 14 strains were negative. The most common gene type was aac(3)-II (162 strains), followed by aac(6′)-I (50 strains), ant(3″)-I (28 strains), aph(3′)-II (20 strains), and ant(2″)-I (20 strains). A total of 18 different enzyme combinations were found (a maximum of three genes in a bacterial strain). aac(3)-II (95 strains), aac(6′)-I/aac(3)-II (28 strains), aac(3)-II/ant(3″)-I (11 strains), and aac(6′)-I (10 strains) were the four most common combinations. The regional gene distribution was similar, with the exception of Guangzhou, Jinan, and Dalian, which coincided with the results of the antibacterial sensitivity profile. The 23 isepamicin-resistant strains (22 strains were also amikacin resistant) had the following enzyme profiles: AAC(3)-II (seven strains), AAC(6′)-I/AAC(3)-II (two strains), AAC(6′)-I (two strains), APH(3′)-II (four strains), APH(3′)–II/AAC(3)-II (one strain), ANT(3″)-I (one strain), AAC(6) –I/APH(3)-II/ANT(3)-I (one strain), APH(3)–II/AAC(3)–II/ANT(3)-I (one strain), and negative (four strains). The bacterial resistance spectrum for aac(3)-II–positive strains was different from negative strains. The former was mainly resistant to gentamicin, kanamycin, and tobramycin, whereas the latter was resistant to almost all agents (Tables 1 and 2).

Table 2.

Distribution of Aminoglycoside-Modifying Enzyme Genes in 205 Strains of Escherichia coli

	Number of modifying enzyme gene (%)
Regions (strains)	aph(3′)-II	aac(6′)-I	aac(3)-II	ant(2″)-I	ant(3″)-I
G1^a (132)	9	35	106	8	22
	(5)	(19.4)	(58.9)	(4.4)	(12.2)
G2^a (73)	11	15	56	12	6
	(11)	(15)	(56)	(12)	(6)

G1, G2 are group 1 and group 2 in Table 1.

Discussion

Antibiotic-modifying enzymes are a major mechanism for bacterial resistance to aminoglycosides. Three groups and nearly 100 aminoglycoside-modifying enzymes have been reported. Variation in the chemical structure of aminoglycosides means that a single enzyme can inactivate several aminoglycoside agents and the same agent can also be inactivated by different enzymes, which results in partial cross-resistance and regional differences in resistance to aminoglycosides.^5,6,11 The results of a national bacterial resistance survey in China showed that Enterobacteriaceae were highly resistant to gentamicin with resistant rates of 59.7% in E. coli (the resistant rates in the hospitals collecting the isolates were 37.6–68.4%), 36.8% in Klebsiella pneumoniae, 40.8% in Enterobacter cloacae, 38% in Enterobacter aerogenes (38%), 28.3% in Serratia spp, and 38.8% in Proteus mirabilis. However, resistance to amikacin was found to be less common. These results suggested that different mechanisms were responsible for the disparity in bacterial resistance to aminoglycoside agents.^3,4,12

The results of the current study show that E. coli was highly resistant to gentamicin, streptomycin, and etimicin, whereas resistance to amikacin and isepamicin was low. Over 80% of isolates could be inhibited with 2 mg L⁻¹ or less of isepamicin, and 88.2% of gentamicin-resistant strains were sensitive to this agent. These results are consistent with a report from Taiwan, where the extended-spectrum β-lactamase-producing E. coli had resistance rates to gentamicin, amikacin, and isepamicin of 75%, 18.7%, 18.7%, respectively, whereas K. pneumoniae had resistance rates of 70.3%, 26.7%, and 26.7%, respectively.¹³ The current study further found that most strains possessed aminoglycoside-modifying enzyme genes. aac(3)-II was the dominant type, while others included aac(6′)-I, ant(3″)-I, aph(3′)-II, and ant(2″)-I. Of these strains, 61.3% (117/191) contained only one gene, in particular 95 strains that contained aac(3)-II, whereas the remaining strains had two to three genes. The dominant enzyme, AAC(3)-II, can modify gentamicin, tobramycin, netilmicin, sisomicin, and dibekacin. AAC(6′)-I is a broad-spectrum enzyme that can inactivate amikacin, tobramycin, netilmicin, kanamycin, isepamicin, sisomicin, and dibekacin. ANT(3″)-I mainly modifies the 3″-hydroxyl moiety of streptomycin and the 9-hydroxy moiety of spectinomycin. ANT(2″)-I uses gentamicin, tobramycin, sisomicin, kanamycin, and duazomycin as substrates. APH(3′)-II contributes to the resistance to kanamycin, neomycin, and ribostamycin. The composition and distribution of enzymes coincided with the characteristics of resistance to gentamicin, kanamycin, and netilmicin, rather than amikacin or isepamicin. The resistance phenotype has remained stable for a long time in China, and the sustained predominance of AAC(3)-II could explain the disparity in resistance to aminoglycoside agents by E. coli, although aac(6′)-1 has shown a slight increase in recent years.^8–10,14

The aminoglycoside resistance spectrum and the modifying enzyme type differ greatly among areas. AAC(6′)-I determines amikacin resistance by Enterobacter and Klebsiella in Japan. The main gentamicin-resistant enzymes were ANT(2″)-I and AAC(3)-1 in the United States. AAC (6′)-I had a higher amikacin resistance detection rate in France, Belgium, and Greece, whereas gentamicin resistance in Germany was correlated with ANT(2″)-I and AAC(3)–IV.^15–17 Although AAC (6′)-I has increased in recent years, AAC(3)-II has long been the major enzyme in China and showed no significant regional variation. However, the different resistance pattern detected in isolates from Guangzhou, Jinan, and Dalian might not result from the minor enzyme type differences, and this change requires further elucidation, such as the mutant aminoglycoside-modifying enzymes as reported by Toth et al. In that report, a APH(2″)-IIa mutant with increased catalytic efficiency to amikacin and isepamicin was derived by PCR mutagenesis.^{8–10,18–20}

In this study, only five strains of isepamicin-resistant E. coli were found to be positive for aac(6′)-I, which encodes an enzyme with isepamicin-modifying activity. The remaining strains might possess some alternative resistance mechanism, such as 16S rRNA methylase. Previous reports of investigations in Shanghai and Wenzhou (a city in this study, too) found amr(A) and rmt(B) methylase genes in amikacin-resistant E. coli, and rmt(B) detection rates were as high as 84.1–100%. These results indicate that gentamicin, kanamycin, and tobramycin resistance by E. coli in China is mainly related to production of aminoglycoside-modifying enzymes, whereas amikacin and isepamicin resistance results from methylation of 16S rRNA. However, the prevalence of methylase genes remains at a low level.^21–23

In conclusion, the clinical isolates of E. coli from China were mainly resistant to gentamicin, kanamycin, tobramycin, netilmicin, and etimicin, but they retained a high sensitivity to amikacin and isepamicin. The main resistance mechanism was the aminoglycoside-modifying enzyme AAC(3)-II. Thus, amikacin and isepamicin could be better choices for the treatment of severe infections in the clinic.

Footnotes

Disclosure Statement

There are no conflicts of interest.

References

Kumar

, Zarychanski

, Light

, Parrillo

, Maki

, Simon

, Lapotta

, Lapinsky

, Ellis

, Mirzanejad

, Martinka

, Keenan

, Wood

, Arabi

, Feistein

, Kumar

, Dodek

, Kravetsky

, Doucetle

Cooperative Antimicrob Therapy of Septic Shock (CATSS) Database Research Group. 2010. Early combination antibiotic therapy yields improved survival compared with monotherapy in septic shock: A propensity-matched analysis. Crit. Care Med., 38:1773–1785.

Hirsch

E.B.

, Tam

V.H.

2010. Detection and treatment options for Klebsiella pneumoniae carbapenemases (KPCs): An emerging cause of multidrug-resistant infection. J. Antimicrob. Chemother., 65:1119–1125.

Xiao

Y.H.

, Wang

, Zhao

C.Y.

, Gau

, Zheng

, Xue

, Li

X.Y.

, Hao

F.L.

2008. National bacterial resistant surveillance of Mohnarin 2006–2007. Chin. J. Nosocomiol., 8:1051–1056.

Wang

2008. Bacterial resistant surveillance of CHINET 2006. Chin. J. Infect. Chemother., 8:1–9.

Wright

G.D.

1999. Aminoglycoside-modifying enzymes. Curr. Opin. Microbiol., 2:499–503.

Jana

, Deb

J.K.

2006. Molecular understanding of aminoglycoside action and resistance. Appl. Microbiol. Biotechnol., 70:140–150.

Clinical Laboratory Standard Institute (CLSI). 2009. Performance Standards for Antimicrobial Susceptibility Testing. Nineteenth Informational Supplement, Jan 28,1 CLSI: Wayne, PA.

Yao

, Chen

, Hu

C.Q.

, Jin

S.H.

2005. Analysis of antibiotic resistance profile and gene patterns for aminoglycoside modifying enzymes in 64 strains of Escherichia coli. Chin. J. Microbiol. Immunol., 25:727–732.

Duan

J.C.

, Lv

X.J.

, Zhao

, Feng

, Yu

R.J.

, Zong

Z.Y.

, Gao

Y.Y.

, Xiong

Y.L.

2005. Aminoglycosides resistance mechanisms of extended spectrum β-lactamases-producing Enterobacteriaceae. Chin. J. Antibiot., 30:555–558.

10.

Z.S.

, Zhou

L.X.

, Zhao

J.Z.

, Yan

, Deng

S.J.

, Zhou

J.M.

2007. Detection of novel resistance genes encoded aminoglycoside modifying enzymes from clinical isolates of Escherichia coli. Chin. J. Nosocomiol., 17:914–916.

11.

Shaw

K.J.

, Rather

P.N.

, Hare

R.S.

, Miller

G.J.

1993. Molecular genetics of aminoglycoside resistance genes and familial relationships of the aminoglycoside-modifying enzymes. Microbiol Rev., 57:138–163.

12.

Xiao

Y.H.

, Wang

, Li

2008. Bacterial resistance surveillance in China: A report from Mohnarin 2004–2005. Eur. J. Clin. Microbiol. Infect. Dis., 27:697–708.

13.

Liao

C.H.

, Sheng

W.H.

, Wang

J.T.

, Sun

H.Y.

, Wang

H.K.

, Hsueh

P.R.

, Chen

Y.C.

, Chang

S.C.

2006. In vitro activities of 16 antimicrobial agents against clinical isolates of extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiellia pneumoniae in two regional hospitals in Taiwan. J. Microbiol. Immunol. Infect., 39:59–66.

14.

Zhang

Y.L.

, Li

J.T.

1998. Studies on the resistance of clinica lisolates and the resistant mechanism in Beijing region. Chin. J. Antibiot., 23:371–376.

15.

Neonakis

, Gikas

, Scoulica

, Manios

, Georgiladakis

, Tselentis

2003. Evolution of aminoglycoside resistance phenotypes of four Gram negative bacteria: An 8 year survey in a university hospital in Greece. Int. J. Antimicrob. Agents, 22:526–531.

16.

Miller

G.H.

, Sabatalli

F.J.

, Naples

, Hare

R.S.

, Shaw

K.J.

1995. The changing nature of aminoglycoside resistance mechanisms and the role of isepamicin- a new broad spectrum aminoglycoside. J. Chemother., 7,Suppl 2:31–44.

17.

Vakulenko

S.B.

, Mobashery

2003. Versatility of aminoglycosides and prospects for their future. Clin. Microbiol. Rev., 16:430–450.

18.

L.H.

, Pan

C.Q.

, Hu

D.K.

, Liu

C.B.

, Zhang

2008. New genotypes for aminoglycoside-modifying enzymes in Esherichia coli. Chin. J. Nosocomoil, 18:757–759.

19.

Zhang

X.Z.

, Huang

Y.M.

, Cheng

, Zhang

, Chen

, Fang

Y.C.

, Deng

2010. Drug resistant analysis of levofloxacin-resistant Escherichia coli to aminoglycosides. Chin. J. Nosocomiol., 20:1358–1361.

20.

Toth

, Frase

, Chow

J.W.

, Smith

, Vakulenko

S.B

. 2010. Mutant APH(2″)-IIa enzyme with increased activity against amikacin and isepamicin. Antimicrob. Agents Chemother., 54:1590–1595.

21.

, Zhang

, Han

, Sun

, Ni

2009. Plasmid-mediated 16S rRNA methylases in aminoglycoside-resistant Enterobacteriaceae isolates in Shanghai, China. Antimicrob. Agents Chemother., 53:271–272.

22.

X.D.

, Wu

C.M.

, Liu

H.B.

, Li

X.S.

, Beier

R.C.

, Xiao

, Qin

S.S.

, Huang

S.Y.

, Shen

J.Z.

2009. Plasmid-mediated ArmA and RmtB 16S rRNA methylases in Escherichia coli isolated from chickens. J. Antimicrob. Chemother., 64:1328–1330.

23.

F.Y.

, Yao

, Pan

J.Y.

, Chen

, Qin

Z-q.

, Parsons

, Yang

L-h.

, Li

Q-q.

, Zhang

X-q.

, Qu

, Wang

L-x.

2010. High prevalence of plasmid-mediated 16S rRNA methylase gene rmtB among Escherichia coli clinical isolates from a Chinese teaching hospital. BMC Infect. Dis., 10:184–194.