Molecular Typing of a Suspected Cluster of Nocardia farcinica Infections by Use of Randomly Amplified Polymorphic DNA, Pulsed-Field Gel Electrophoresis, and Amplified Fragment Length Polymorphism Analyses

Abstract

Randomly amplified polymorphic DNA (RAPD), pulsed-field gelelectrophoresis (PFGE), and amplified fragment length polymorphism (AFLP) analyses were used to investigate a possible outbreak of Nocardia farcinica. RAPD and PFGE analyses yielded irreproducible and unsatisfactory results, respectively. AFLP analysis seem to be a promising and welcome addition for molecular analysis of Nocardia isolates.

Health-care associated transmission or acquisition of Nocardia has only rarely been documented, and nocardial infections are not considered to be transmitted from person to person (3, 20). Between July and November 2003, four cases of disseminated infections with N. farcinica in renal transplant recipients in the Leiden University Medical Center (hospital A) were recognized. This small cluster prompted us to investigate the relatedness of these N. farcinica isolates in more detail by using two previously reported fingerprinting methods: randomly amplified polymorphic DNA (RAPD) analysis (13, 15, 19) and pulsed-field gel electrophoresis (PFGE) (2, 13). In addition, the use of amplified fragment length polymorphism (AFLP) (21), which has not yet been applied to Nocardia isolates, was explored.

The analysis included the 4 suspected outbreak isolates (isolate numbers A₁-A₄, Table 1) and 20 additional clinical isolates. These additional isolates were considered to be epidemiologically unrelated since they were collected from 20 individual patients out of 11 hospitals (B to L) over a 6-year period (1998 to 2003) in The Netherlands. Reference strains ATCC 3318, DSM43298, DSM43665, and DSM46004 were also included.

TABLE 1.

Clinical characteristics of a cluster of four solid organ transplant recipients with disseminated N. farcinica infections in hospital A^a

Patient (strain)	Age (yr)	Sex	Underlying condition	Diagnosis of N. farcinica infection		Main clinical presentation	Treatment	Outcome
				Date (day/mo/yr)	Biological sampleb
A1	47	F	Renal-pancreas Tx	23/10/03	BAL, pus from abscess right upper leg	Fever, pain right upper leg, pulmonary nodule left lung	SXT	Cured
A2	47	M	Renal Tx	14/10/03	BAL	Right sided pleuritic pain, right pulmonary upper-lobe infiltration	SXT	Cured
A3	52	M	Renal Tx	18/07/03	BAL, blood	Fever, subcutaneous nodules, cough, left pulmonary upper-lobe infiltration, three cerebral abscesses	SXT	Cured
A4	77	F	Renal Tx	26/11/03	BAL, sputum, blood	Fever, seizure, left pulmonary upper-lobe infiltration, multiple small cerebral abscess	SXT	Cured

^aAbbreviations: BAL, broncheoalveolar lavage; SXT, cotrimoxazole; Tx, transplantation.

^bOnly isolates from BAL samples were used for molecular genotyping.

N. farcinica isolates were cultured on sheep blood agar incubated aerobically at 35°C in an atmosphere enriched with 5% CO₂. DNA was isolated with established procedures using the QIAamp DNA minikit (Qiagen, Venlo, The Netherlands) and the MagNA Pure LC DNA extraction platform in combination with the MP LC DNA isolation kit III (Roche Diagnostics, Almere, The Netherlands). Molecular identification was performed by partial 16S rRNA gene sequencing (8). Primers for the RAPD analysis included ERIC1 (5′-ATGTAAGCTCCTGGGGATTCAC-3′), ERIC2 (5′-AAGTAAGTGACTGGGGTGAGCG-3′), M13 (5′-GAGGGTGGCGGTTCT-3′), DAF4 (5′-CGGCAGCGCC-3′), RTG2 (5′-GTTTCGCTCC-3′), RTG3 (5′-GTAGACCCGT-3′), RTG4 (5′-AAGAGCCCG-T-3′), and RTG6 (5′-CCCGTCAGCA-3′). All primers were purchased from Eurogentec (Seraing, Belgium) and evaluated on all isolates in duplicate. Only DAF4 and RTG6 gave more than four bands. The RAPD reaction mix consisted of 5 pmol of primer, 2 μl of DNA, and a RAPD bead (GE Healthcare, Diegem, Belgium) in a 25-μl reaction mixture. The PCR program for DAF4 consisted of 2 min at 94°C, followed by 45 cycles of 40 s at 94°C, 40 s at 45°C, and 40 s at 72°C, with a 5-min extension at 72°C. The PCR program for RTG6 consisted of 5 min at 95°C and 45 cycles of 60 s at 95°C, 60 s at 36°C, and 2 min at 72°C. PFGE analysis was performed according to the method of Blümel et al. (2). AFLP analysis was performed using established procedures with minor modifications (14). AFLP adapters were made by mixing equimolar amounts of complementary oligonucleotides (5′-CTCGTAGACTGCGTACAGGCC-3′ and 5′-TGTACGCAGTC-3′ for ApaI; 5′-GACGATGAGTCCTGAC-3′ and 5′-TAGTCAGGACTCAT-3′ for MseI; Eurogentec) and heating to 95°C, followed by slow cooling to ambient temperatures. Approximately 5 ng of genomic DNA was subjected to a combined restriction-ligation procedure containing 5 pmol of ApaI adapter, 50 pmol of MseI adapter, 2 U of ApaI (New England Biolabs, Beverly, MA), 2 U of MseI (New England Biolabs), and 1 U of T4 DNA ligase (Promega, Leiden, The Netherlands) in a total volume of 20 μl of 1× reaction buffer for 1 h at 20°C. Next, the mixture was diluted five times with 10 mM Tris-HCl (pH 8.3) buffer. One microliter of the diluted restriction-ligation mixture was used for amplification in a volume of 25 μl under the following conditions: 1 μM ApaI primer without selective residues (5′-Flu-GACTGCGTACAGGCCC-3′), 1 μM MseI primer with 1 selective C residue (underlined) (5′-GATGAGTCCTGACTAAC-3′), a 0.2 mM concentration of each deoxynucleoside triphosphate, and 1 U of Taq DNA polymerase (Roche Diagnostics) in 1× reaction buffer containing 1.5 mM MgCl₂. Amplification was done as follows. After an initial denaturation step for 4 min at 94°C in the first 20 cycles, a touchdown procedure was applied: 15 s of denaturation at 94°C, 15 s of annealing at 66°C, with the temperature for each successive cycle lowered by 0.5°C, and 1 min of extension at 72°C. Cycling was then continued for a further 30 cycles with an annealing temperature of 56°C. After completion of the cycles, incubation at 72°C for 10 min was performed before the reaction mixtures were cooled to room temperature. The amplicons were then combined with the ET550-R size standard (GE Healthcare) and analyzed on a MegaBACE 500 automated DNA platform (GE Healthcare) according to the manufacturer's instructions. The data was imported into Bionumerics software (Applied Maths, Sint-Martens-Latem, Belgium) and analyzed for percent similarity by the unweighted pair group method with arithmetic averages (UPGMA) using the Pearson correlation coefficient. Only DNA fragments from 60 to 550 bp were included in the analysis. The final interpretation of the dendrogram included a visual inspection of the fingerprints (6). Isolates were considered to be unrelated if their fingerprints differed by two or more bands.

All 24 clinical isolates were identified as N. farcinica by partial 16S rRNA gene sequencing. With both primers, DAF4 and RTG6, different RAPD fingerprints were obtained in two consecutive experiments, indicating low reproducibility with these primers. Further analysis was considered irrelevant since the RAPD analysis was not reproducible. Despite several efforts, we were unable to obtain satisfactory results using PFGE analysis (results not shown). Because of these disappointing results, we decided to explore AFLP as an alternative typing method for Nocardia isolates. AFLP has been appreciated many times because of its high discriminatory power and reproducibility and easily withstands the comparison to PFGE analysis for many nosocomial pathogens (5, 12, 17, 18), which is generally considered to be the gold standard in epidemiological analyses. The result of the AFLP analysis is shown in Fig. 1. A cutoff similarity of 92% was arbitrarily chosen to discriminate between related and unrelated strains. As a result, fingerprints that differ by only 1 band are in a gray interpretation zone where it is arbitrary if they should be considered to be either related or unrelated. Three of four clinical isolates from the suspected outbreak in hospital A appeared to be related to each other: isolates A₁ and A₄ were indistinguishable, whereas only a minor difference (i.e., a one-band difference) was observed with isolate A₃ (92% similar). Isolate A₂ proved to be unique. Reference strains ATCC 3318 and DSM43665 yielded visually identical fingerprints. Since these are in fact the same strain, this nicely demonstrates the reproducibility of the AFLP method. All other isolates yielded unique patterns.

FIG. 1.

AFLP dendrogram created by UPGMA clustering using the Pearson correlation coefficient. AFLP fingerprints (not shown) were obtained from 24 clinical N. farcinica isolates and 4 N. farcinica reference strains. Strain numbers A₁ to A₄ represent N. farcinica isolates from four clustered patients in hospital A. All other clinical isolates were assumed to be unrelated (see the text). Reference strains ATCC 3318 and DSM43665 are identical. The 92% cutoff similarity level corresponds to isolates that have either visually identical fingerprints or a maximum of one band difference.

Reliable typing methods are a prerequisite for identifying the source and the mode of transmission of nosocomial infections. Previously, only a few reported presumptive clusters of Nocardia infections were studied using molecular techniques, and most of the earlier studies were based on the temporal relationship of contacts between the infected patients (1, 4, 9, 10, 15). RAPD analysis has been described before as a useful method for intraspecies discrimination in an epidemiological study of N. farcinica (7). However, using the same primers (DAF4 and RTG6) the RAPD analysis proved to be poorly reproducible under the conditions as described in the present study. In another report, three clustered cases of nocardiosis were shown to be caused by N. farcinica strains with identical RAPD fingerprints using primer P2 (13). Obviously, further optimization and standardization of the RAPD assay as a tool for typing of N. farcinica strains is necessary. A PFGE protocol has been described for the differentiation of N. farcinica isolates (2). However, problems with typing in PFGE procedures are well documented, and inconclusive fingerprinting results have also been reported in other genera (2, 11, 16). We were unable to successfully implement the Nocardia PFGE method as described in the literature (2). Despite being unable to verify our AFLP results to an accurate gold standard, AFLP analysis seems to be an affective alternative typing method for the differentiation of N. farcinica isolates.

Three of four N. farcinica isolates from the cluster of infections in hospital A were more closely related to each other than to the other strains, suggesting a common source of infection. However, hospitalization periods overlapped in case of patients A₁, A₂, and A₃, but not of patient A₄, suggesting another potential common source than the ward these patients were admitted to. Extensive rebuilding activities, which were taking place in hospital A in the same time period, might have been the source (2, 19). However, despite extensive environmental screening, Nocardia species were not isolated from environmental samples. Consequently, the source of this small cluster of N. farcinica infections could not be identified. Nevertheless, AFLP seems to be a promising and welcome addition to existing techniques for molecular typing of Nocardia isolates.