Restriction fragment length polymorphism typing demonstrates substantial diversity among Pneumocystis jirovecii isolates

Abstract

Better understanding of the epidemiology and transmission patterns of human Pneumocystis should lead to improved strategies for preventing Pneumocystis pneumonia (PCP). We have developed a typing method for Pneumocystis jirovecii that is based on restriction fragment length polymorphism (RFLP) analysis after polymerase chain reaction amplification of an approximately 1300 base-pair region of the msg gene family, which comprises an estimated 50-100 genes/genome. The RFLP pattern was reproducible in samples containing >1000 msg copies/reaction and was stable over time, based on analysis of serial samples from the same patient. In our initial analysis of 48 samples, we found that samples obtained from different individuals showed distinct banding patterns; only samples obtained from the same patient showed an identical RFLP pattern. Despite this substantial diversity, samples tended to cluster on the basis of country of origin. In an evaluation of samples obtained from an outbreak of PCP in kidney transplant recipients in Germany, RFLP analysis demonstrated identical patterns in samples that were from 12 patients previously linked to this outbreak, as well as from 2 additional patients. Our results highlight the presence of a remarkable diversity in human Pneumocystis strains. RFLP may be very useful for studying clusters of PCP in immunosuppressed patients, to determine whether there is a common source of infection.

Figure 1. Stability of the RFLP banding pattern in samples obtained over time

Representative gels (left) and corresponding blots (right) show the RFLP pattern of 2 samples (A and B) collected from 5 individuals at different time points. The samples from patients 1 and 2 are BAL samples that were obtained 15 and 111 days apart, respectively; from patient 3 are sputum samples that were obtained 10 days apart; and from patients 4 and 5, are oral wash samples that were both obtained 1 day apart. The samples from patient 2 were obtained during 2 separate episodes of PCP. Although the samples from patients 4 and 5 had msg copy numbers below the cut-off of reproducibility of our assay (patient 4 B, 260 msg copies/reaction; patient 5 A and B, both 900 msg copies), the RFLP pattern is identical (patient 5) or highly similar (patient 4). All samples were digested with Dra1.

Figure 2. Similarity analysis of RFLP patterns

Dendrogram derived by BioNumerics software from RFLP analysis of 37 samples following agarose gel electrophoresis. Eleven samples with less than 1,000 msg copies per assay were excluded from this analysis. The Dice coefficient was used to calculate similarities, and UPGMA for cluster analysis. The position tolerance was 1.9%. The percent similarity scale is shown above the dendrogram, and indicated by the numbers at the individual nodes. The samples grouped with the same colour come from the same country: orange, Italy; green, Netherlands; blue, USA. Lambda/HindIII molecular weight markers were used as internal standards in the gel analysis to normalize the banding pattern from 6 different gels. BAL samples that were collected from the same patient are indicated by a red box around the number. The green box indicates a sputum pair collected from the same patient. Five major clusters that were identified by the BioNumerics software are indicated by a blue bars on the right and are numbered 1 to 5.

Figure 3. RFLP analysis of isolates from a cluster of patients with PCP

A. RFLP analysis of 10 isolates from German renal transplant patients with previously reported evidence of nosocomial Pneumocystis transmission. Each lane corresponds to one patient. Identical banding patterns are evident in all samples. There is an additional band in sample 8a, but of note, msg copies of this sample are 230/assay, below the cut off of reproducibility. Sample 2a also had <1,000 msg copies/assay. B. RFLP analysis of 9 additional BAL samples that were unrelated to the outbreak. Samples 12, 13, and 14 show a pattern identical to the outbreak pattern. These 3 samples are from 2 patients transplanted in the same Germany hospital but who were not previously linked to the outbreak. The last 6 samples are from patients unrelated to the outbreak, and all show a unique pattern. The first sample, 6a, is the same as in the top panel and is included to allow comparison with the outbreak pattern. All samples (top and bottom) were digested with Dra1 and separated by gel electrophoresis.