Simultaneous Detection and Identification of Common Cell Culture Contaminant and Pathogenic Mollicutes Strains by Reverse Line Blot Hybridization

Abstract

We have developed a reverse line blot (RLB) hybridization assay to detect and identify the commonest mollicutes causing cell line contamination (Mycoplasma arginini, Mycoplasma fermentans, Mycoplasma hyorhinis, Mycoplasma orale, and Acholeplasma laidlawii) and human infection (Mycoplasma pneumoniae, Mycoplasma hominis, Mycoplasma genitalium, Ureaplasma parvum, and Ureaplasma urealyticum). We developed a nested PCR assay with “universal” primers targeting the mollicute 16S-23S rRNA intergenic spacer region. Amplified biotin-labeled PCR products were hybridized to membrane-bound species-specific oligonucleotide probes. The assay correctly identified reference strains of 10 mollicute species. Cell cultures submitted for detection of mollicute contamination, clinical specimens, and clinical isolates were initially tested by PCR assay targeting a presumed mollicute-specific sequence of the 16S rRNA gene. Any that were positive were assessed by the RLB assay, with species-specific PCR assay as the reference method. Initially, 100 clinical and 88 of 92 cell culture specimens gave concordant results, including 18 in which two or more mollicute species were detected by both methods. PCR and sequencing of the 16S-23S rRNA intergenic spacer region and subsequent retesting by species-specific PCR assay of the four cell culture specimens for which results were initially discrepant confirmed the original RLB results. Sequencing of amplicons from 12 cell culture specimens that were positive in the 16S rRNA PCR assay but negative by both the RLB and species-specific PCR assays failed to identify any mollicute species. The RLB hybridization assay is sensitive and specific and able to rapidly detect and identify mollicute species from clinical and cell line specimens.

Mollicute contamination is among the most frequently occurring problems associated with cell cultures and can render experimental results and biological products unreliable (6, 8). In different surveys, from 15 to 80% of cell cultures have been reported to be contaminated with mollicutes (19). Mollicute contamination is not always easy to detect because it usually produces neither turbid growth nor cytopathic effects (6). Of 20 mollicute species known to cause cell culture contamination, five (Mycoplasma arginini, M. fermentans, M. hyorhinis, M. orale, and Acholeplasma laidlawii) account for about 95% of episodes (13, 17).

M. pneumoniae, M. hominis, M. genitalium, Ureaplasma parvum, and U. urealyticum are the commonest pathogenic mollicute species found in clinical specimens. M. pneumoniae is a causative agent of primary atypical pneumonia in children and one of the most frequent causes of community-acquired pneumonia in adults (4). U. parvum and U. urealyticum are genitourinary commensal organisms that can cause urethritis and, in immunocompromised patients, abscesses and pyogenic arthritis (16). M. hominis may be involved in a variety of urogenital infections (11), and M. genitalium is an emerging cause of nongonococcal urethritis and is strongly associated with cervicitis and endometritis (2, 18).

Nonculture methods used for the detection and identification of mollicutes in clinical specimens and cell culture include DNA fluorochrome staining (14), immunological assays (1), nucleic acid hybridization (12), and PCR assay (8), which vary in speed, reliability, specificity, and sensitivity. PCR assay targeting “universal” mollicute sequences in the 16S rRNA gene have been described but are not entirely specific for mollicutes (8). The 16S-23S rRNA intergenic spacer region is a promising target for identification of mollicutes to species level because it harbors highly variable domains (5). We describe the development of a rapid, simple method, the reverse line blot (RLB) hybridization assay, based on this region for simultaneous detection and identification of the 10 commonest cell culture-contaminant and pathogenic mollicute species.

MATERIALS AND METHODS

Mollicute strains.

M. hyorhinis ATCC 17981, M. fermentans ATCC 19989, A. laidlawii ATCC 23206, M. pneumoniae ATCC 29342 (M129), M. genitalium ATCC 33530, and 14 serovars of U. parvum and U. urealyticum American Type Culture Collection (ATCC) reference strains were obtained directly from the ATCC. One strain each of M. arginini and M. orale, identified by sequencing of their 16S-23S rRNA intergenic spacer regions, was obtained from the Capital Pediatric Institute in Beijing, People's Republic of China. In addition, a well-characterized clinical isolate was used as our M. hominis reference strain.

Contaminant mollicute strains in cell cultures.

Cell cultures were received for mollicute screening from more than 30 clinical and research laboratories in Sydney, Australia, and tested by PCR assay with primers GPO-3 and MGSO as described previously (20). DNA samples from 92 cell cultures that gave positive results with this PCR assay were used in this study.

DNA extraction from cell culture fluid.

DNA extraction was carried out with the Roche Amplicor respiratory specimen preparation kit (Roche Diagnostics Co., Indianapolis, Ind.). Briefly, samples were centrifuged at 13,000 rpm for 10 min at room temperature. The supernatant was discarded, leaving 10 μl above the pellet, and 500 μl of wash solution was added. The sample was vortexed for 10 s and then centrifuged at 13,000 rpm for 10 min. The supernatant was removed, and the pellet was resuspended in 50 μl of lysis reagent, followed by 45 min of incubation at 60°C; 50 μl of neutralization solution was added, and the sample was vortexed and then stored at −20°C prior to PCR testing.

DNA extraction from clinical specimens and isolates.

We used a total of 100 human-pathogenic mollicute isolates or clinical specimens previously shown to contain a human-pathogenic mollicute in this study as follows: 10 M. hominis, 60 U. parvum, and 10 U. urealyticum clinical isolates plus 14 nasopharyngeal aspirate samples containing M. pneumoniae and 6 genital specimens containing M. genitalium from children (7, 8, 9). DNA was extracted as described previously (10) and stored at −20°C before testing.

Oligonucleotide design.

The mollicute 16S rRNA, 16S-23S rRNA intergenic spacer region, and 23S rRNA sequences in GenBank were compared with the Pileup and Pretty programs in the Multiple Sequence Analysis program group provided in WebANGIS (3rd version; Australia National Genomic Information Service). We designed “universal” mollicute primers and species-specific probes located in the 16S rRNA gene, 16S-23S rRNA intergenic spacer region, and 23S rRNA gene (Table 1).

TABLE 1.

Oligonucleotide primers and probes used in this study

Primer or probe namea	Tm (°C)b	GenBank accession no.	Sequencec	Optimal probe concentration (μM)
SPS1	78.9	AE002112	1642CCCTACGA(/G)GAACGTGGGGA(/G/C)TGGAT(/A)C(/T)ACCTCCT1673
SPS2	77.3	AE002112	1647CGA(/G)GAACGTGGGGA(/G/C)TGGAT(/A)C(/T)ACCTCCTTTC1676
SPA1	79.4	AE002112	2025CGTCCTTCATCGC(/A)CTG(/C/T/A)T(/C)T(/C)T(/A/G)GT(/A)C(/G)CCAA GGCATT(/C)CAC1991
SPA2	76.4	AE002112	2047CG(/C)T(/G/A)T(/A/G)GCT(/A)TA(/T)TCGC(/G)A(/T)GA(/G/T)TT(/A)A(/G) G(/A/T)C(/T)A(/G/C)CGTCCTTCATCG2014
A. laidlawii 1A	58.2	AF294996	30AAG TGT TAG TTA GCC TTT CTC CT8	0.078
A. laidlawii 2S	57.1	AF294996	152AAA TGA TGT CTG AAA AGA AAT AAG175	0.078
M. fermentans 1A	59.7	AF294992	102CCC ATA AAA AAG CCA CAT AAC82	0.078
M. fermentans 2S	58.2	AF294992	286CAT CAT AAC AAA CTA TAA CAA TAG GAA312	0.078
M. arginini 1A	58.4	AF294994	80AAA GAA CAA ATT GAG AGA TAG GTC57	0.078
M. arginini 2S	57.1	AF294994	169CAA TAG GTC TTA TAC TAC TAT TAA ACA AGA T199	0.078
M. orale 1A	58.7	AF294995	61GAA TAT TGG GCC ATT AAC TAT TT39	0.078
M. orale 2S	57.6	AF294995	201CAA TAG GTC AAA AAT ACT TAT ACG TAA227	0.156
M. hyorhinis 1A	57.8	AF294993	57CTA GAC ACG AAT CGA TTA TGT AAT34	0.078
M. hyorhinis 2S	57.9	AF294993	143AAC GAT CTT TTT TAT AAC CGA G164	0.156
M. hominis 1A	58	AF294991	77CAA AGA ACC GAG AGA TAA ATC T57	0.156
M. hominis 2S	57.3	AF294991	161CAA TAG GTC ATA CAA TTA ACA AAA C185	0.078
M. peneumoniae 1A	58.3	D14528	83ACC GAT AAA TAA ATG GAT TTT G62	0.156
M. pneumoniae 2S	59.8	D14528	158ACA TTT CCG CTT CTT TCA A176	0.312
M. genitalium 1A	59.2	D14526	54CAC CGA AAA AAT TAA TGG G36	0.312
M. genitalium 2S	58.8	D14526	120AAG AAT GTT TTT GAA CAG TTC TTT143	0.312
U. parvum 1A	57.3	AF059323	79GGC TTA TAT TCA TAT GGA TTT TAA TA54	0.078
U. parvum 2S	57.7	AF059323	91TAT TTT TAA AAA TTC ATA TGG TCG114	0.625
U. urealyticum 1A	59.9	X58561	223GGC TAA TAT TCA CAT GGA TTT TTA TA198	0.078
U. urealyticum 2S	58	X58561	236AAA TAT TTC AAA AGT TCA TAT GGT C260	0.625

^aThe suffixes S and A indicate sense and antisense, respectively.

^bValues provided by the primer synthesiser (Sigma-Aldrich).

^cNumbers represent the base positions at which the primer sequences start and finish (starting at point 1 of the corresponding gene sequence in GenBank). Letters in parathenses and separated by a shill indicate alternative nucleotides in different species.

Identification of mollicute strains by species-specific PCR assay.

We tested 92 contaminated cell cultures, 100 clinical isolates or specimens, and 21 reference strains by mollicute species-specific PCR assay, based on the 16S-23S rRNA intergenic spacer region, as previously described (7, 8, 9).

Sequencing and sequence searching.

The identities of mollicute species in clinical and cell culture specimens for which the species-specific PCR assay and RLB hybridization results were discrepant were determined by PCR assay and sequencing of the 16S-23S rRNA intergenic spacer region with the SPS2 and SPA1 primer pair. In addition, we determined the 16S rRNA sequences for cell culture specimens in which no mollicute species was identified by either RLB hybridization or species-specific PCR assay, as previously described (8). Sequencing was performed by Applied Biosystems (Foster City, Calif.) BigDye terminator chemistry on an ABI Prism 373 DNA sequencer. The sequence search was performed with the FastA program group accessed through WebANGIS.

Mollicute 16S-23S rRNA intergenic spacer region PCR assay.

The PCR mixture was prepared as previously described (8). Positive and negative controls were processed in parallel with each sample tested to identify possible false-negative results and PCR contamination. Two universal primer pairs (SPS1 plus SPA2 and SPS2 plus SPA1) were designed from conserved nucleotide sequences of the 16S and 23S rRNA gene regions. The mollicute 16S-23S rRNA intergenic spacer region was amplified in a nested PCR assay with SPS1 and SPA2 as the outer primers and SPS2 and SPA1 as the inner primers. Inner primers were biotin labeled at the 5′ end (Sigma-Aldrich). The primer sequences are listed in Table 1. For the first-round amplification, denaturation, annealing, and elongation temperatures and times were 96°C for 10 s, 65°C for 10 s, and 74°C for 1 min, respectively, for 30 cycles, in a Perkin-Elmer 9600 thermal cycler (Eppendorf). Conditions were the same for the second-round amplification except that annealing was performed at 70°C for 10 s. We analyzed 10 μl of each PCR product by gel electrophoresis with 1.5% agarose. Gels were stained with 0.5 μg of ethidium bromide per ml, and bands were visualized with a UV transilluminator. Biotin-labeled amplicons from this PCR assay were used in the RLB assay.

RLB hybridization.

The RLB hybridization assay was based on a method described previously (21) except that we changed the hybridization temperature to 60°C, used streptavidin-β-peroxidase conjugate (Roche Diagnostics Co.) diluted 1:5,000 in 2× SSPE (1× SSPE is 0.18 M NaCl, 10 mM NaH2PO4, and 1 mM EDTA [pH 7.7])-0.5% sodium dodecyl sulfate and decreased the time of exposure to X-ray film (Hyperfilm; Amersham) to 5 min.

Two species-specific oligonucleotide probes were designed for each mollicute species, with sequences in the highly variable domain of the 16S-23S rRNA intergenic spacer region used to maximize the specificity of hybridization. RLB results were regarded as positive when both probes gave positive results. The melting temperatures of the probes were between 57°C and 60°C (Table 1). To optimize hybridization conditions, duplicate samples were tested at two different hybridization temperatures (50 and 60°C), and each probe was tested at several twofold dilutions, starting at a concentration of 1.25 μM. In our laboratory, membranes labeled with mollicute species-specific probes have been stripped at least 15 times without any detectable loss of signal.

Sensitivity of the RLB assay.

Suspensions of M. arginini, M. fermentans, M. hyorhinis, M. orale, and Acholeplasma laidlawii reference strain DNA were serially diluted from 10⁻¹ to 10⁻⁸ with molecular biology-grade water and tested in parallel by the species-specific PCR assay and RLB hybridization. Comparison of the detection limits of the two methods was used to determine the assay sensitivity of the RLB assay.

RESULTS AND DISCUSSION

Mollicute strains are widespread human, animal, and plant parasites (15). Although many methods have been developed to detect and identify them, PCR is currently the most commonly used. Previously, we used a species-specific PCR assay to detect and identify mollicute species, particularly those encountered as cell culture contaminants. However, the complexity and expense of this technique limit its application for routine diagnosis. Thus, the aim of our study was to develop a more convenient assay capable of accurately and rapidly detecting and simultaneously identifying cell line-contaminant and clinically important mollicute species. The technique described here is based on the detection of mollicute species with amplification products from the 16S-23S rRNA intergenic spacer region with species-specific oligonucleotide probes in an RLB hybridization assay.

PCR amplification of mollicute 16S-23S rRNA intergenic spacer with universal primer pairs.

The universal primers (SPS1 plus SPA2 and SPS2 plus SPA1) designed from the mollicute 16S and 23S rRNA genes amplified all reference strains. As previously described (8), the PCR assay based on the 16S and 23S rRNA genes was not completely specific for mollicutes; 12 cell culture specimens which contained nonmollicute bacteria (as determined by 16S rRNA sequences) were also amplified (data not shown) (8). However, with the use of species-specific oligonucleotide probes in the RLB hybridization assay (see below), these false-positive results were avoided.

Specificity of RLB hybridization.

There was no nonspecific hybridization at either 50 or 60°C and no difference in hybridization results between the two temperatures. The hybridization temperature chosen for testing all strains was 60°C. The optimal probe concentrations varied with different probes from 0.078 to 0.625 μM (Table 1).

Most mollicute species-specific sequences in the 16S-23S rRNA intergenic spacer region variable domains are very different from each other except for U. parvum and U. urealyticum, which showed very little heterogeneity. The probes we designed for these species, U. parvum 1A and U. urealyticum 1A, differ by only three bases. Nevertheless, all 4 U. parvum and 10 U. urealyticum ATCC reference strains and 70 clinical isolates (60 U. parvum and 10 U. urealyticum) were successfully identified to species level with these species-specific probes. Theoretically, the use of two probes for each species should increase the specificity of the RLB assay. For each mollicute species tested, the results were identical for both species-specific probes.

Reference strains of 10 mollicute species, including one each of U. parvum and U. urealyticum, were correctly identified by the RLB assay with no nonspecific hybridization (Fig. 1). The results of the RLB hybridization assay and the species-specific PCR assay were compared for all 192 specimens or isolates that were positive in the mollicute group-specific PCR assay. Both methods gave the expected results for all 100 clinical isolates and specimens. In 12 of 92 16S rRNA PCR-positive cell culture specimens, no mollicute species was detected by the RLB assay or species-specific PCR assay, and sequencing of PCR amplicons failed to identify mollicute-specific sequences in any.

FIG. 1.

Ten mollicute species reference strains identified by the RLB hybridization method. The positions of 20 probes specific for mollicute species are shown on the left-hand side. The 16S-23S rRNA intergenic spacer region PCR amplicons were from the following species (top to bottom): A. laidlawii ATCC 23206; M. fermentans ATCC 19989; M. arginini from the Capital Pediatric Institute in Beijing, People's Republic of China; M. orale from the Capital Pediatric Institute; M. hyorhinis ATCC 17981; M. hominis, a well-characterized clinical isolate; M. genitalium ATCC 33530; M. pneumoniae ATCC 29342 (M129); U. urealyticum serovar 8, ATCC 27618; and U. parvum serovar 3, ATCC 27618.

Positive results were concordant in both assays for 76 cell culture specimens, including 18 in which two or more different mollicute species were detected; four specimens were positive only in the RLB assay (Table 2). After DNA was reextracted from these four original cell culture specimens, the amplicons generated by the nested 16S-23S rRNA intergenic spacer region PCR assay were sequenced, and the species-specific PCR assay was repeated. This confirmed the original RLB results for three cell culture specimens which contained M. orale. For the fourth, repeating the species-specific PCR assay confirmed the original RLB assay result, which identified the strain as M. hyorhinis. However, the sequencing result was unsatisfactory, suggesting that the cell culture specimen was possibly contaminated with other organisms as well.

TABLE 2.

Mollicutes strains identified in 92 cell line specimens by two methods

Species	No. of strains identified by:
	RLB hybridization assay	Species-specific PCR assay
M. fermentans	13	13
M. arginini	24	24
M. orale	9	9a
M. hyorhinis	16	16a
M. fermentans and M. arginini	5	5
M. fermentans, M. arginini, and M. orale	2	2
M. arginini and M. orale	4	4
M. orale and M. hyorhinis	7	7
No mollicute species detectedb	12	12
Total	92	92

^aM. orale and M. hyorthinis were identified in three and one cell culture specimens, respectively, by the species-specific PCR assay after reextraction of DNA.

^bTwelve specimens contained nonmollicute bacteria, as determined by 16S rRNA sequencing.

RLB sensitivity.

In the DNA serial dilution test, the RLB assay was as sensitive as the species-specific PCR assay for the detection of M. arginine and M. orale (detection limit, 10⁻³ and 10⁻⁴, respectively). The RLB assay was 100-fold more sensitive than species-specific PCR assay for detection of Acholeplasma laidlawii (the RLB assay detection limit was 10⁻⁶) and 10-fold more sensitive for detection of M. arginini and M. hyorhinis (the RLB assay detection limit was 10⁻⁴). Our initial results for cell culture specimens apparently confirmed the greater sensitivity of the RLB assay compared with the species-specific PCR assay, since it detected a greater number of positive results. If so, this could be explained by the fact that the RLB assay used the product of a nested rather than a single-amplification PCR assay (3). However, the sensitivity of the species-specific PCR assay was improved after DNA was extracted again from the original specimens. We concluded that the RLB assay is at least as sensitive as the species-specific nested PCR assay.

While our manuscript was in preparation, a PCR-microtiter plate hybridization assay for rapid detection of four human genital mycoplasmas in genitourinary samples was described (22). Although the principle is similar to that of the RLB assay, the RLB assay has a number of advantages. It is inexpensive; membranes and probes can be reused at least 15 times without loss of sensitivity, and up to 45 species can be detected simultaneously or 45 specimens can be tested for a single species on a single membrane. In future, more mollicute species-specific probes can be added to increase the range of detectable species of contaminants and further improve cell line quality control. The results are read objectively as the presence or absence of a well-defined dot rather than by detection of a color change, which may be more subjective. The RLB assay target, the 16S-23S rRNA intergenic spacer region, shows more species-specific variability than the 16S rRNA genes, as used in the plate hybridization assay.

The RLB assay can successfully identify more than one mollicute species in cell culture. Only the pathogenic species M. pneumoniae and M. genitalium were identified in clinical specimens. Further evaluation of a greater range of genital and respiratory specimens is needed before this method can be used routinely for identification of pathogenic mollicutes in clinical specimens, but these preliminary results are very promising.

In summary, although there is no “gold standard” for the detection and identification of mollicute species, our RLB hybridization assay is at least as sensitive and specific and more practicable than our previously validated species-specific PCR assay for identification of mollicute species in cell culture and clinical specimens.