Modification of Dienes Mutual Inhibition Test for Epidemiological Characterization of Pseudomonas aeruginosa Isolates

Abstract

Pseudomonas aeruginosa is an important cause of community-associated and nosocomial infections related to exposure to aqueous environments. Such infections often occur in the setting of a common-source outbreak, in which case epidemiological characterization of isolates may be necessary. In this preliminary study, a modification of the Dienes mutual inhibition test, ordinarily used to assess the relatedness of swarming Proteus mirabilis strains, was used to study 15 P. aeruginosa isolates, with the results compared to those obtained by ribotype analysis. Complete concordance was noted between the results of the Dienes test and those of ribotyping. These observations suggest that further studies are warranted to assess the utility of the modified Dienes test as a simple, inexpensive, and reliable means for epidemiological typing of P. aeruginosa.

Pseudomonas aeruginosa is a ubiquitous gram-negative bacillus that is often found in moist environments (6). Community-associated P. aeruginosa infections may occur as a consequence of exposure to the organism in hot tubs, whirlpools, water slides, or swimming pools (8) or by use of contaminated sponges (2, 11). Keratitis due to P. aeruginosa may occur in individuals who store contact lenses in tap water and/or contaminated contact lens solutions (4). Intravenous drug users may develop endocarditis and osteomyelitis as a result of injection of contaminated injectables (9).

P. aeruginosa is also of nosocomial significance. The organism has been isolated from a variety of solutions including soaps, ointments, irrigation and dialysis fluids, eyedrops, and disinfectants, as well as from fomites such as showerheads, respiratory therapy equipment, sinks, and baths (12). Populations at risk for significant morbidity and mortality include intubated patients in intensive care units (17), patients on chronic ambulatory peritoneal dialysis (1), and burn patients (13).

Epidemiological investigations of P. aeruginosa infection have used molecular typing methods (10), as well as tools such as biotyping, phage typing, bacteriocin typing, and serotyping (7). The Dienes mutual inhibition test has been used as an epidemiological tool to characterize isolates of Proteus mirabilis (15, 18, 19, 20). In this test, if two different P. mirabilis colonies growing on an agar plate are genetically dissimilar, a clear line of demarcation forms at the point of intersection of swarming growth; conversely, if two isolates are identical or highly similar, the swarming edges merge without a conspicuous line of demarcation (3, 18, 19). Colonies of P. aeruginosa usually have a spreading morphology on agar media (8) and have the capability of adhering to central venous catheters (CVCs), including those impregnated with minocycline and rifampin (E. L. Munson et al., unpublished data). In this preliminary report, we propose a modified Dienes test for the epidemiological characterization of P. aeruginosa isolates.

Ten clinical isolates of P. aeruginosa, previously identified by the Vitek GNI+ card (bioMérieux, Marcy l'Etoile, France), and P. aeruginosa type strains ATCC 27853 and ATCC 35032 were selected for study. Three of the 12 strains were assayed in duplicate, for a total of 15 comparisons. The identities of the 15 isolates were coded and were not known by the personnel performing the experiments.

Triple-lumen polyurethane CVCs impregnated with minocycline and rifampin were obtained from Cook Critical Care (Bloomington, Ind.). CVCs were aseptically sectioned into 1-cm segments in a biological safety cabinet and placed on the first quadrant of 95% defibrinated sheep blood agar (SBA) plates (Remel, Lenexa, Kans.) previously inoculated by streaking the plate with an isolate of P. aeruginosa. After 18 to 24 h incubation in 35°C ambient air, CVC segments in the presence of confluent P. aeruginosa growth were transferred onto fresh SBA plates in close proximity in an orientation that resembled an equilateral triangle (Fig. 1). By using this format, bacterial growth emanating from each CVC segment could simultaneously be compared to the growth from the two other segments in the triangle. Each isolate, among the 15 examined in this study, was compared to all other isolates. The plates were incubated for 48 to 72 h at 35°C in ambient air. All assays were performed in duplicate. Intersection of colony growth from two isolates without a clearly visible line of demarcation was interpreted as indicating a high degree of relatedness (positive test result) (Fig. 1). Isolates with a readily observable line of demarcation were classified as being unrelated (negative test result) (Fig. 1).

FIG. 1.

Illustration of the modified Dienes test for three P. aeruginosa isolates growing on 5% defibrinated blood agar. Isolate 1 is on top, isolate 9 is to the right, and isolate 10 is to the left. A positive result (P. aeruginosa isolates 1 and 9) and two negative results (P. aeruginosa isolates 10 and 1 and P. aeruginosa isolates 10 and 9) were obtained.

Ribotyping was performed with the RiboPrinter microbial characterization system (Qualicon, Wilmington, Del.), as described previously (5, 16). Briefly, isolates were inoculated into tubes containing lysis buffer, placed in a heating block at 80°C for 30 min, and then transferred to the RiboPrinter instrument. Within the RiboPrinter, the remaining steps were entirely automated, including cleavage of DNA with the restriction enzyme PvuII, fragment separation by gel electrophoresis, and modified Southern blotting. The DNA fragments were then hybridized with a labeled DNA probe derived from the Escherichia coli rrnB rRNA operon. The bands were detected with a chemiluminescent substrate. An image was captured by using a customized charge-coupled device camera and electronically transferred to the system's computer. Similarity coefficients were calculated on the basis of both relative banding intensity and band position.

P. aeruginosa isolates yielding both a negative Dienes test and a ribotyping similarity coefficient of greater than 0.850 were evaluated further by pulsed-field gel electrophoresis (PFGE), performed as described previously (14). Genomic DNA in agarose was digested with the restriction enzyme PvuII, and the resulting fragments were separated by electrophoresis in 1% agarose on a CHEF-DRII apparatus (Bio-Rad, Richmond, Calif.) with the following conditions: 200 V for 23 h at switch times ramped from 5 to 40 s. Strains were considered different by PFGE if more than three bands were different (5, 21). Strains with one to three different bands were considered subtypes.

Growth was observed to radiate from the respective CVC segments for all 15 isolates examined in this study (Fig. 1). Approximately 72 h of incubation was necessary to achieve a convergence of growth patterns. The results of the modified Dienes test revealed two groups of P. aeruginosa isolates that appeared to be related: isolates 5 and 13 and isolates 1, 9, 11, and 12 (Table 1). All other combinations yielded modified Dienes test results indicative of unrelatedness.

TABLE 1.

Results of modified Dienes testing for 15 isolates of P. aeruginosa^a

Isolate	Result for isolate:
	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1	−	−	−	−	−	−	−	+	−	+	+	−	−	−
2		−	−	−	−	−	−	−	−	−	−	−	−	−
3			−	−	−	−	−	−	−	−	−	−	−	−
4				−	−	−	−	−	−	−	−	−	−	−
5					−	−	−	−	−	−	−	+	−	−
6						−	−	−	−	−	−	−	−	−
7							−	−	−	−	−	−	−	−
8								−	−	−	−	−	−	−
9									−	+	+	−	−	−
10										−	−	−	−	−
11											+	−	−	−
12												−	−	−
13													−	−
14														−

^aEach isolate was tested against all 14 other isolates. +, seamless merger of colony growth; A −, presence of a line of demarcation between colony growths.

Ribotyping delineated the same two clusters (Fig. 2 and Table 2). A similarity coefficient of ≥0.9897 was obtained with P. aeruginosa isolates 1, 9, 11, and 12. The similarity coefficient obtained with isolates 5 and 13 was 0.9680. Two other clusters of P. aeruginosa strains (isolates 3 and 4 and isolates 2, 6, and 15) yielded similarity coefficients of 0.9334 to 0.9461 by ribotyping. No relatedness was observed in comparisons of these strains by the modified Dienes test.

FIG. 2.

Dendrogram (A) and ribotype profile (B) characterization of 15 P. aeruginosa test isolates.

TABLE 2.

Similarity coefficients obtained by ribotyping comparison of 15 P. aeruginosa isolates

Isolate	Similarity coefficient for isolate:
	12	11	1	9	8	10	14	4	3	7	2	15	6	5	13
12	100.00
11	99.77	100.00
1	98.98	99.29	100.00
9	98.93	99.17	98.97	100.00
8	74.93	77.07	75.27	77.59	100.00
10	67.47	67.80	67.19	68.02	64.81	100.00
14	58.96	58.98	58.06	59.47	73.32	53.73	100.00
4	30.31	30.42	31.37	33.09	51.18	49.58	40.99	100.00
3	35.18	35.45	36.76	38.45	56.27	41.35	42.32	94.61	100.00
7	31.71	31.91	32.64	35.16	52.16	41.19	40.85	82.07	84.43	100.00
2	63.66	63.77	64.31	64.70	37.02	41.70	40.95	73.29	75.63	70.99	100.00
15	59.21	59.16	59.91	60.55	35.51	41.84	37.89	78.59	80.14	70.11	94.27	100.00
6	67.27	67.17	68.27	68.95	40.12	47.39	43.44	72.92	74.35	77.70	94.04	93.34	100.00
5	39.06	38.83	40.57	39.48	21.07	20.78	23.97	52.75	55.50	50.61	80.70	67.61	65.20	100.00
13	45.90	45.92	46.77	46.50	27.01	27.38	31.25	56.31	58.78	50.71	82.63	73.86	68.69	96.80	100.00

The results of PFGE analysis of selected strains are presented in Fig. 3. Strains 5 and 13 yielded identical PFGE profiles. Differences of one to three bands were noted among comparisons of strains 1 and 9, strains 2 and 5, and strains 2 and 13. These pairs were considered subtypes of one another.

FIG. 3.

PFGE banding patterns for 10 P. aeruginosa test isolates. The numbers represent isolate numbers. S, molecular weight standards.

The three duplicate isolates were isolates 1 and 11, isolates 5 and 13, and isolates 9 and 12. All three pairs yielded positive modified Dienes test results and yielded similarity coefficients of greater than ≥0.9680 by ribotyping. Strains 5 and 13 were identical by PFGE; strains 1 and 9 differed by two bands and were considered subtypes by PFGE. Isolate 1 was a clinical isolate. It was recovered from a culture of an endotracheal suction specimen from a 64-year-old man with nosocomial pneumonia. It had been chosen randomly for inclusion in this study. Isolate 9 was P. aeruginosa type strain ATCC 27853. Interestingly, relatedness between strains 1 and 9 was demonstrated both by the modified Dienes test and by ribotyping.

Potential variables with the modified Dienes test include differential growth rates among different P. aeruginosa strains and the spacing between the CVC segments on test plates. These two factors could influence the length of incubation necessary to achieve sufficient organism growth to permit test interpretation. Mueller-Hinton agar, a less nutritive medium than SBA, resulted in diminished spread of organism growth around CVC segments (data not shown) and substantially increased the incubation times necessary to achieve a result. Use of CVC segments provided a convenient means of confining the assay to a relatively small surface area of an agar plate. This permitted greater reproducibility in experimental setup, easier delineation of lines of demarcation, and a faster turnaround time. The reason for using minocycline- and rifampin-impregnated CVC segments was the observation that P. aeruginosa adhered tightly to such segments, even following prolonged incubation and serial subculture of segments (E. L. Munson, S. O. Heard, and G. V. Doern, unpublished data). It is likely that other catheter types could also be used successfully in the modified Dienes test.

In summary, in this preliminary study, in which ribotyping and PFGE were used as comparison methods, a modification of the previously described Dienes mutual inhibition test performed well in the epidemiological characterization of clinical isolates of P. aeruginosa. The test was simple to perform and cheap and may have utility in the initial screening of P. aeruginosa isolates in suspected common-source outbreaks. This test is particularly appealing to smaller clinical microbiology laboratories with financial and technical constraints that may preclude use of molecular biology-based methods in epidemiological investigations.