Antimicrobial Resistance in Clinical Isolates of Ureaplasma spp. from Samples in Germany

Ureaplasma urealyticum and Ureaplasma parvum are mollicutes species that colonize the urogenital tract of many asymptomatic people but are also thought to be associated with symptomatic infections. Using 170 strains isolated between 2016 and 2019 in a German university hospital, resistance was tested by a combination of commercial tests, molecular methods, and determination of MICs.

ABSTRACT

Ureaplasma urealyticum and Ureaplasma parvum are mollicutes species that colonize the urogenital tract of many asymptomatic people but are also thought to be associated with symptomatic infections. Using 170 strains isolated between 2016 and 2019 in a German university hospital, resistance was tested by a combination of commercial tests, molecular methods, and determination of MICs. Rates of resistance to macrolides, tetracyclines, and fluoroquinolones were 0%, 4.1%, and 7.1%, respectively.

INTRODUCTION

Species of the genera Mycoplasma and Ureaplasma belong to the mollicutes class, which is characterized by a remarkable reduction of their genomes. As a result, these bacteria possess limited metabolic capabilities and rely strongly on hosts for in vivo growth. The most striking characteristic of all mollicutes is the lack of a bacterial cell wall, resulting in the intrinsic resistance of clinically relevant species to beta-lactams. Ureaplasma parvum and Ureaplasma urealyticum are common bacteria of the human urogenital tract present in up to 80% of asymptomatic patients and can be the causative agents of severe infections in newborns (1–5). Furthermore, results of studies over the last few years have suggested an association between these microorganisms and symptomatic urogenital tract infections and fatal hyperammonemia in a limited number of adults, as well adverse pregnancy outcomes (1, 2, 5). Despite a high rate of therapeutic failure (6), antibiotic treatment may be necessary in some of these patients and includes macrolides, tetracyclines, and fluoroquinolones. As strains resistant to these antibiotic classes have been described internationally, knowledge about the resistance pattern of currently circulating strains is important for successful and, in many cases, empirical treatment. However, actual data about the resistance rates among Ureaplasma spp. are lacking in Germany and rare in Europe.

A total of 170 strains isolated from vaginal and cervical swabs (n = 69), urethral swabs (n = 55), urine (n = 10), ejaculates (n = 9), and rectal swabs (n = 27) from the southeast part of Germany were collected between December 2016 and January 2019. The study was approved by the Institutional Review Board of TU Dresden (no. EK 473122017). Ureaplasma strains in clinical material were analyzed using Mycoplasma IST2 (bioMérieux). In parallel, isolates were cultivated in mycoplasma/ureaplasma broth (Oxoid) and stored at −80°C. Isolates were diluted with broth and retested by the Mycofast RevolutioN (MR) (Elitech) test according to the recommendations of the manufacturer. The DNA of strains was isolated with the QIAamp DNA minikit (Qiagen). All isolates were divided into U. urealyticum and U. parvum using PCR (7). Based on pretesting, strains with suspected quinolone resistance were further characterized by PCR and partial sequencing of genes coding for ParC and ParE by methods reported previously (8, 9). The presence of the tetM element in the isolated strains as a precondition for tetracycline resistance in Ureaplasma was detected by PCR as described (10). To investigate the clonality of the confirmed elements, PCR products were sequenced, and the sequences were aligned. MICs of strains with suspected resistance to quinolones were determined according to standard procedure using levofloxacin (Sigma) (11).

Among the 170 strains (55.9% from men) included in the study, 68 (40.0%) were U. urealyticum. Whereas U. parvum was isolated with nearly the same frequency from female and male patients (57% versus 43%, respectively), most U. urealyticum isolates (75%) were cultivated from men. The occurrence of the tetM element was confirmed in 73 isolates (42.9%). Sequencing of PCR products was successful in 70 out of 73 isolates, and derived amino acid sequences showed seven different types (see Table S1 in the supplemental material). However, most elements can be assigned to the tetM sequence of Ureaplasma spp. reference strains Vancouver (36%) and Seattle (29%), respectively. Tetracycline-resistant strains were found in three of the seven tetM types (Table S1).

Testing of Ureaplasma isolates by commercial assays to determine resistance is a common method of characterizing strain collections (12–16). However, differences in test procedures in comparison with the CLSI microdilution assay (11) as the gold standard are the reason for discrepant results between methods (17). Data from different reports confirmed that determination of macrolide and tetracycline resistance by the IST2 and MR kit is largely comparable with the microdilution test but show that quinolone resistance is greatly overestimated by the IST2 test. To a much lesser extent, the MR test was also found to overestimate quinolone resistance (13). Using the IST2 and the MR kit, the following rates of macrolide, tetracycline, and quinolone resistance were found in the present study: 0%, 4.1%, and 65.9% with IST2 and 0%, 4.1%, and 7.1% with MR. Except for quinolone resistance measured by IST2, these rates are in accordance with the results of previous studies using microdilution assays with or without commercial kits for determining resistance from the United States, Switzerland, France, and Great Britain, which reported rates of 0% to 0.1% for macrolides, 0% to 8% for tetracyclines, and 1% to 6% for quinolones (12, 13, 18–21). Of note, significantly higher rates are described in particular populations and regions, e.g., China and Tunisia (15, 22–24). Isolates with quinolone (33% U. urealyticum) and tetracycline (75% U. urealyticum) resistance were retested with the recently available Mycoplasma IST3 kit, which offers an optimized panel of antibiotics in comparison with the IST2 kit, and results comparable with the MR test were demonstrated. Furthermore, 20 strains testing quinolone resistant using the IST2 but susceptible by the IST3 and MR kits were retested by MIC determination according to the CLSI method (11). All of these strains are levofloxacin susceptible (MIC, 0.25 to 2 μg/ml). Thus, the IST2 resulted in a remarkable percentage of false-positive quinolone-resistant strains (89%; n = 100). Despite a relatively high rate of tetM-carrying strains, only 7 isolates (4.1%) showed resistance to tetracycline with both the IST2 and MR tests. To confirm this result, 25 tetM-positive but tetracycline-susceptible Ureaplasma strains were investigated using the CLSI procedure (11) and demonstrated MICs to tetracycline between ≤0.25 and 1 μg/ml.

Whereas the molecular mechanism causing tetracycline resistance in ureaplasmas (transcriptionally active tetM element) and its influence on MICs to tetracycline (increased from ≤1 μg/ml in susceptible strains to 2 to >64 μg/ml in resistant isolates) are well characterized, many mutations associated with quinolone resistance of Ureaplasma spp. have been found in the genes coding for topoisomerases ParC and ParE. To correlate phenotypic quinolone resistance (≥4 μg/ml for moxifloxacin and/or levofloxacin according to CLSI) with corresponding mutations, the quinolone resistance-determining regions (QRDR) of parC (nucleotides 211 to 444 according to the sequence of the reference strain U. urealyticum SV10; GenBank accession no. NC_011374.1) and parE (1093 to 1752) of all 170 strains were sequenced. In the 12 strains consistently detected as quinolone resistant by the MR and IST2/3 kits, the following amino acid changes were confirmed: S83L (9×), D82N (1×), and E87K (1×) in ParC and R448K (1×) in ParE (Table 1). All changes have been described in different reports as associated with quinolone resistance, and measured MICs to levofloxacin were consistent with those found in other studies (8, 12–14, 18–26). Further changes in ParE were detected (9xR437I+G480S, 4xT392I, 2xE466K, 1xK420E+I437R+S480G, 1xL511I) but did not result in phenotypic resistance (data not shown). Among the quinolone-resistant strains in this study, 58% carried a tetM element. Interestingly, two U. urealyticum isolates (no. 993 and 21097) from rectal swabs of HIV-positive patients were quinolone and tetracycline resistant, confirming the occurrence of double resistance in particular cases.

TABLE 1.

Characterization of fluoroquinolone-resistant Ureaplasma isolates of the study

Strain no.	Sex/age (yr) of patienta	Clinical manifestation/coinfectionb	Samplec	Speciesd	tetMg	Antimicrobial susceptibilitye according to:		QRDR analysisf		LEV MIC (μg/ml)
						IST3	RevolutioN	ParC	ParE
3531	F/38	PPR	CS	Up	−	MOXs, LEVr	MOXs, LEVr	WT	R448K	8
14580	M/28	STI	UA	Up	−	MOXr, LEVr	MOXs, LEVr	S83L	WT	4
23008	M/43	HIV	RS	Uu	+	MOXs, LEVr	MOXs, LEVr	S83L	WT	4
993	M/25	HIV	RS	Uu	+ (R)	MOXr, LEVr	MOXs, LEVr	S83L	WT	8
8538	F/43	Unknown	CS	Up	−	MOXs, LEVr	MOXs, LEVr	S83L	WT	8
22724	M/34	Unknown	UA	Up	+	MOXs, LEVr	MOXs, LEVr	S83L	WT	4
24524	F/20	Unknown	CS	Up	+	MOXs, LEVr	MOXs, LEVr	D82N	WT	8
74454	M/69	Unknown	E	Up	−	MOXr, LEVr	MOXr, LEVr	S83L	WT	4
6062	M/30	Unknown	UA	Up	+	MOXr, LEVr	MOXr, LEVr	S83L	WT	4
18396	M/49	HIV	RS	Uu	+	MOXs, LEVr	MOXs, LEVr	S83L	WT	4
19356	F/35	Unknown	CS	Up	−	MOXs, LEVr	MOXs, LEVr	E87K	WT	4
21097	M/44	HIV	RS	Uu	+ (R)	MOXr, LEVr	MOXs, LEVr	S83L	WT	8

^aF, female; M, male.

^bPPR, preterm premature rupture; STI, supposed sexually transmitted infection; HIV, regular screening.

^cCS, cervical swab; UA, urethral swab; RS, rectal swab; E, ejaculate.

^dUp, U. parvum; Uu, U. urealyticum.

^eSusceptibility to fluoroquinolones; MOX, moxifloxacin; LEV, levofloxacin; r, resistant; s, susceptible.

^fAmino acid changes. WT, wild type.

^g(R), tetracycline resistance.

In conclusion, 153 (90%) of the 170 Ureaplasma spp. isolates investigated in the present study were susceptible to therapy with a standard antibiotic (macrolide, tetracycline, or fluoroquinolone) recommended for treatment of Ureaplasma infections and 168 isolates to a member of two of these classes of antibiotics.