Abstract
From 1990 to 1996, routine screening for whooping cough identified 399 patients with a calmodulin-dependent adenylate cyclase-positive test result and yielded 69 Bordetella pertussis isolates. None of the patients were fully vaccinated, and most were less than 6 months old. Analysis of total DNA by pulsed-field gel electrophoresis (PFGE) after XbaI, SpeI, or DraI macrorestriction yielded 19, 15, and 5 different patterns, respectively, whereas ribotyping failed to demonstrate any strain polymorphism. Discrimination among the isolates was improved by combining the PFGE profiles. Some patterns were more frequent, but the corresponding patients were not clearly epidemiologically related. The patterns for two strains obtained during a 3-month period from patients who were neighbors differed by the length of a single DNA fragment. These data strongly suggest that one type of isolate is widely spread throughout the world and is carried by individuals other than patients who develop a true illness.
Since the early 1990s, the incidence of whooping cough has increased notably in Europe including France (2, 17, 21) and had increased in the United States before that (3, 9), even though vaccination is still strongly recommended. Bacterial diagnosis of whooping cough is hindered by the fragility of Bordetella pertussis and the interference of the normal nasopharyngeal flora. Direct fluorescent-antibody testing and serological methods lack sensitivity and specificity, respectively (11). One rapid screening method is based on the detection of calmodulin-dependent adenylate cyclase (AC)-hemolysin by an assay with an alginate swab specimen from the nasopharynx (7). Multitarget PCR-based assays (13, 26) are sensitive and more efficient than culture, but nasopharyngeal sampling with Dacron swabs is required. No comparison of the results of the AC test and PCR-based diagnosis has been reported to date.
To distinguish among the clinical isolates responsible for pertussis outbreaks, existing epidemiological tools necessitate previous bacterial isolation (1, 14). Pulsed-field gel electrophoresis (PFGE) can reveal polymorphisms among clinical isolates, but isolates from a defined locality (1, 9, 17) had similar DNA profiles. A recent comparison of PFGE, random arbitrarily primed detection, and enterobacterial repetitive intergenic consensus-PCR showed that PFGE was the most discriminatory. PCR-based detection of B. pertussis repeat DNA sequences failed to reveal strain polymorphism (17).
The incidence of whooping cough diagnosed by both biochemical testing and bacterial isolation has increased in eastern France in the past 7 years. We used DNA fingerprinting to distinguish between endemic and epidemic isolates.
MATERIALS AND METHODS
Patients and samples.
Both hospitalized and ambulatory patients were included in this epidemiological study. They were eligible for participating in the study if they had signs of pertussis with intense, dry, and emetic coughing for at least 10 days. Apnea and cyanosis were frequent in infants younger than 3 months of age and sometimes necessitated intensive care in Strasbourg University Hospital. Samples were obtained by a hospital microbiologist or in an ambulatory consultation to limit the processing time. Nasopharyngeal swab specimens from both nostrils were obtained with calcium alginate-tipped applicators (Puritan Hardwood Products Co., Guilford, Maine). The swabs were immediately immersed in Stainer and Scholte medium and were agitated for several seconds (21, 22). Calmodulin-activated AC activity was assayed as described previously (7, 21). Samples were also inoculated onto fresh (≤7 days) Bordet-Gengou agar plates (Bacto Bordet-Gengou agar base; Difco Laboratories, Detroit, Mich.) supplemented with 15% (vol/vol) defibrinated sheep blood and 40 μg of cephalexin (Sigma) per ml. The plates were incubated at 37°C for at least 7 days in a humid atmosphere and were observed daily for the presence of “mercury droplet” colonies surrounded by a zone of hemolysis. B. pertussis was identified by classical methods, and the reference strain B. pertussis ATCC 18323 was used as a control. All isolates that were collected were stored at −80°C in a 9‰ (wt/vol) NaCl solution containing 20% (vol/vol) glycerol.
DNA preparations for PFGE analysis.
The isolates were grown at 37°C on Bordet-Gengou agar plates for at least 72 h, until colonies 1 mm in diameter appeared. The bacteria were harvested with a Pasteur pipette and were resuspended in 10 ml of 10 mM Tris-HCl–5 mM EDTA–1 M NaCl (pH 8.0) and then pelleted by centrifugation at 4°C and 5,000 × g, for 10 min and processed as described elsewhere (18). DNA fingerprinting was carried out after restriction with XbaI, SpeI, or DraI. The gels were run at 12°C with a Geneline Transverse Alternating Field Electrophoresis (TAFE) system (Beckman) at 140 mA in 0.6× TAFE running buffer (20× TAFE running buffer is 200 mM Tris, 0.5 mM free acidic EDTA, and 87 mM acetic acid [pH 8.2]). Separation of the restricted DNA fragments started with 4-s pulses for 2 h, followed by 12-s pulses for 8 h, 8-s pulses for 6 h, and 6-s pulses for 2 h. The gels were stained for 30 min with 2 μg of ethidium bromide per ml and were washed in water before being photographed under UV light at 300 nm. Pulsotypes were compared and classified in dendrograms by using the Dice coefficient and the unweighted pair group method with arithmetic mean clustering methods provided by Molecular Analyst (version 1.5) and Fingerprinting (version 1.12) software (Bio-Rad).
Ribotyping.
Ribotyping was performed as described previously (19) by using Immobilon P membranes (Millipore). Intact DNA embedded in agarose for PFGE analysis was restricted with AccI, PvuII, or SalI endonuclease, according to the manufacturer’s recommendations (New England Biolabs, Beverly, Mass.). Southern blots were made by using 32P-nick-translated DNA fragments (20) corresponding to the Pfu DNA polymerase-amplified whole 16S and 23S Escherichia coli rRNA genes (rDNAs) (15).
RESULTS
Disease epidemiology.
From 1990 to 1996, 867 patients were included in this study. The number of the patients sampled was stable until 1994, but the number increased during the last 2 years (almost 300 patients in 1995 and 1996). The AC test was positive for 399 patients (45.9%) and progressed as described above. Sixty-nine B. pertussis isolates (representing 17.3% of the positive AC tests) were obtained. The sex ratio for patients infected with B. pertussis isolates was 59% males and 41% females, and that for patients with AC-positive test results was 54% males and 46% females.
As shown in Fig. 1, 67 and 59% of AC test-positive and culture-positive patients, respectively, were less than 6 months old. The frequency of AC test positivity and culture positivity did not differ between children aged 6 months to 18 months (9 and 6%, respectively) and those aged 18 months to 6 years (13 and 12%, respectively). Older patients with a diagnosis of pertussis after sampling comprised 11% of the AC test-positive patients and 23% of the culture-positive patients. Figure 1 indicates the vaccination status of the patients. None of those with a positive biochemical test result or from whom bacteria were isolated had been fully vaccinated when pertussis occurred. The majority of patients were unvaccinated patients, but infants who were less than 3 months old accounted for 38% of the AC test-positive patients and 38% of the culture-positive patients. Only two of these patients had received a first vaccine injection. The patients in the 3- to 6-month and the 18-month to 5-year age groups with a diagnosis of pertussis were not completely vaccinated, and the patients in these groups combined corresponded to 19 and 26% of the subjects with positive AC test results and B. pertussis culture positivity, respectively. Overall, 65% of the patients had not been vaccinated at all, but 35% were less than 3 months old. Among the patients in the 6- to 10-year and ≥10-year age groups, no patient received a full vaccination because the last vaccination dose at age 6 years, at least, was omitted. The patients in these groups combined accounted for 11 and 23% of the subjects with positive AC test results and culture positivity, respectively.
FIG. 1.
Percentages, as compiled histograms, of the patients with different vaccination statuses and with a diagnosis of pertussis with an AC-positive test result (AC+; left columns, n = 331) and from whom B. pertussis bacteria were isolated (Bp+; right columns; n = 69), according to age (bottom). Vaccination statuses were as follows: no vaccination dose (
), one vaccination dose (□), two vaccination doses (
), three vaccination doses (▥), three vaccination doses without a booster dose at 18 months (
), four vaccination doses including that at 18 months (
), and five vaccination doses (■). The normal vaccination is three doses monthly from the 3rd to the 6th months of life and two booster doses at 18 months.
Fingerprinting by PFGE and ribotyping of B. pertussis DNA.
The PFGE profiles obtained after XbaI digestion contained from 10 to 15 DNA fragments ranging from 50 kb to 550 to 600 kb (Fig. 2). Some DNA preparations were retested six times over a 1-year period and always gave reproducible profiles after digestion with XbaI or SpeI, as determined previously with the TAFE system (18). This accounts for the stability of the epidemiological marker and that of genomes of B. pertussis during this study. The best and most reproducible resolution was for fragments in the range of 100 to about 500 kb, and the pattern generally comprised 8 to 10 DNA fragments. For the pulsotypes obtained after digestion with XbaI (Fig. 2), smaller fragments were not separated and were not visible enough, whereas the size of the upper DNA fragment could not be determined precisely. Therefore, differentiation of the patterns was done with the variable bands, whose sizes ranged from 120 to 530 kb. Nevertheless, when the upper DNA fragment is considered with the one that is 600 kb in length, the different lengths of genomes varied from 3,200 to 3,800 kb, which are comparable to those reported previously (23). From the profiles obtained after digestion with XbaI, some DNA fragments appeared to be very well conserved, such as those at 135, 160, and about 200 kb (see Fig. 3). One DNA fragment of about 310 kb was always encountered except in a pulsotype 19 strain (from patient 6). Pulsotypes like those observed in Fig. 2, lanes 6, 7, and 1, were considered to be very similar on the basis of the separation of two DNA fragments of 270 to 275 kb in lane 1; these fragments comigrated in lanes 6 and 7. The corresponding patterns were designated 1 and 1′ and will be considered again in this work. The pulsotypes of the strains in lanes 2 and 5 (Fig. 2) were considered to be identical and were assigned to pattern (pulsotype) 15. The pulsotypes of the strains in lanes 9, 15, 16, 18, 19, and 21 were distinguished from those of the strains in lanes 4, 17, and 20 by the addition of a large DNA fragment of approximately 330 kb. This addition did not seem to affect significantly the length of any other visible fragment, even that of about 600 kb. The two groups of patterns were designated patterns 13 and 12, respectively. Pulsotype 15 (Fig. 3) differed from pattern 13 by the presence of a DNA fragment of 400 kb and a different distribution of DNA fragments ranging from 200 to 250 kb. Pulsotype 15 differed from patterns 6 and 17 (Fig. 3) by the presence of DNA fragments ranging from 150 to 250 kb and another one of 120 kb for pulsotype 17 (Fig. 3). Finally, 19 different XbaI digestion patterns (Fig. 3) were characterized from the series of 69 isolates, and these different patterns accounted for the pronounced polymorphism of the B. pertussis strains obtained from within a small geographic area (about 12,000 km2 with 1,700,000 individuals). Pulsotypes 13, 15, and 1 contained 22, 9, and 11 isolates (61% of the total), respectively.
FIG. 2.
PFGE analysis of XbaI-restricted intact DNAs from random B. pertussis isolates evidenced polymorphism and frequent pulsotypes. Lanes: T, polymers of bacteriophage λ DNA (New England Biolabs); 1, patient 43; 2, patient 60; 3, patient 55; 4, patient 52; 5, patient 61; 6, patient 11; 7, patient 23.
FIG. 3.
Computerized comparison according to the dendrogram and by schematic magnification of the 19 XbaI profiles and the number of corresponding fingerprints among the 70 B. pertussis isolates analyzed, including ATCC 18323. Nb, number.
The intact DNAs of the isolates were also analyzed after restriction with SpeI, and again, evidence of polymorphism and frequently occurring fingerprints was obtained (Fig. 4). The 15 patterns identified by restriction with SpeI contained 10 to 12 DNA fragments ranging from 50 to 550 kb (Fig. 4 and 5A). The sum of the estimated lengths of these fragments also approximately corresponded to the length of the chromosome determined as described above, indicating that numerous small DNA fragments were not visible in the PFGE profiles after restriction with either XbaI or SpeI. As for the profiles obtained after restriction with XbaI, several DNA fragments appeared to be ubiquitous in the 15 different profiles; they were 60, 80, 100, 130, and 170 kb in length. A great variability in DNA bands of 210 to 450 kb was observed, and these encompassed four to seven fragments and allowed the differentiation of PFGE profiles. Another 480-kb SpeI DNA fragment obtained after restriction with SpeI seemed to be almost constant in the genomes. As shown in Fig. 4, the patterns of many isolates corresponded to pattern 5 (Fig. 5A). Pattern 5 was distinguished from pattern 6 by a 200-kb DNA fragment, and isolates with this pattern were clustered with those with pattern 7 in the dendrogram analysis and supported relationships of clonality. Other pulsotypes, like those shown in Fig. 5A, patterns 3 and 14, differed from each other by the presence of DNA fragments of 275 and 125 kb, respectively. As for the profiles obtained after restriction with XbaI, two pulsotypes (pulsotypes 5 and 9) obtained after restriction with SpeI covered 56 and 10% of the isolates, respectively.
FIG. 4.
PFGE analysis of SpeI-restricted DNAs from B. pertussis isolates also evidenced polymorphism and frequent pulsotypes. Lanes: T, polymers of bacteriophage λ DNA (New England Biolabs); 1, patient 21; 2, patient 7; 3, patient 2; 4, patient 4; 5, patient 28; 6, patient 42; 7, patient 58; 8, patient 32; 9, patient 29; 10, patient 6; 11, patient 35; 12, patient 38.
FIG. 5.
Computerized comparison according to the dendrograms and by schematic magnification of the 15 SpeI profiles (A) and the 5 DraI profiles (B) and the number (Nb) of corresponding fingerprints among the 70 B. pertussis isolates analyzed, including ATCC 18323.
Finally, PFGE analysis with DraI restriction revealed only five different pulsotypes (Fig. 5B) which were composed of approximately 10 DNA fragments also ranging from 50 to 500 kb. Most DNA fragments obtained after restriction with DraI were not constant within the five patterns except in the region of 50 to 110 kb. Pulsotypes 2 and 3 obtained after restriction with DraI were the most frequently encountered (30 and 36% of the isolates, respectively).
Ribotyping experiments revealed complete homogeneity in the profiles that were obtained (data not shown), despite the use of three restriction endonucleases.
Setting epidemiological data with combinations of pulsotypes obtained by PFGE.
The pulsotypes obtained after restriction with DraI, SpeI, and XbaI were combined to increase the level of discrimination (Table 1). This led to 35 combinations. The series of a given XbaI pulsotype were often differentiated by the SpeI profiles, as was the case for XbaI restriction patterns 13, 14, 15, 1, 1′, and 9. However, the three isolates harboring pulsotype 4 after restriction with XbaI were not further distinguished by restriction with either DraI or SpeI, although they appeared to be independent. The most frequently occurring pulsotype after XbaI restriction (pulsotype 13) was further distinguished only for isolates from four patients (those from patients 3, 47, 49, and 46) by restriction from DraI and SpeI, respectively. Also, XbaI pulsotypes 18, 14, 15, and 1 were distinguished by the profiles obtained after restriction with DraI and SpeI. Combinations of patterns obtained by PFGE after restriction with DraI-SpeI-XbaI, patterns 2-5-1, 3-5-13, and 5-9-15, appeared more frequently and represented 10, 26, and 9% of the clinical isolates, respectively (Table 1). Among these frequent combinations, patterns 2-5-1 and 3-5-13 were found for isolates which were obtained over a 5-year period. No epidemiological link (by age, nursery, family, or close circle) was easily evidenced for the other patients, and these isolates could be considered independent at the level of this study. The pulsotype combination 5-9-15 was found for six isolates collected in 1995 and 1996 from patients living in the city of Strasbourg, with two isolates found in two patients from the same family.
TABLE 1.
Distribution of B. pertussis ATCC 18323 and isolates from patients with whooping cough according to PFGE profiles obtained after restriction with DraI, SpeI, and XbaI and the 35 combinations that were revealed
| Patient no. or strain | Pulsotype obtained after restriction with the following enzyme:
|
||
|---|---|---|---|
| DraI | SpeI | XbaI | |
| 1 | 1 | 8 | 1 |
| ATCC 18323 | 1 | 1 | 6 |
| 2 | 1 | 1 | 6 |
| 3 | 1 | 5 | 13 |
| 4 | 1 | 14 | 14 |
| 5 | 2 | 2 | 18 |
| 6 | 2 | 13 | 19 |
| 7 | 2 | 5 | 15 |
| 8 | 2 | 5 | 15 |
| 9 | 2 | 5 | 15 |
| 10 | 2 | 5 | 1 |
| 11 | 2 | 5 | 1 |
| 12 | 2 | 5 | 1 |
| 13 | 2 | 5 | 1 |
| 14 | 2 | 5 | 1 |
| 15 | 2 | 5 | 1 |
| 16 | 2 | 5 | 1 |
| 17 | 2 | 5 | 4 |
| 18 | 2 | 5 | 4 |
| 19 | 2 | 5 | 4 |
| 20 | 2 | 5 | 8 |
| 21 | 2 | 7 | 1 |
| 22 | 2 | 6 | 1 |
| 23 | 2 | 4 | 1 |
| 24 | 2 | 5 | 1′ |
| 25 | 3 | 5 | 13 |
| 26 | 3 | 5 | 13 |
| 27 | 3 | 5 | 13 |
| 28 | 3 | 5 | 13 |
| 29 | 3 | 5 | 13 |
| 30 | 3 | 5 | 13 |
| 31 | 3 | 5 | 13 |
| 32 | 3 | 5 | 13 |
| 33a | 3 | 5 | 13 |
| 33b | 3 | 5 | 13 |
| 33c | 3 | 5 | 13 |
| 34 | 3 | 5 | 13 |
| 35 | 3 | 5 | 13 |
| 36 | 3 | 5 | 13 |
| 37 | 3 | 5 | 13 |
| 38 | 3 | 5 | 13 |
| 39 | 3 | 5 | 13 |
| 40 | 3 | 5 | 13 |
| 41 | 3 | 5 | 1 |
| 42 | 3 | 6 | 1′ |
| 43 | 3 | 5 | 1′ |
| 44 | 3 | 5 | 9 |
| 45 | 3 | 9 | 16 |
| 46 | 3 | 6 | 13 |
| 47 | 3 | 4 | 13 |
| 48 | 3 | 12 | 10 |
| 49 | 3 | 15 | 13 |
| 50 | 4 | 1 | 17 |
| 51 | 4 | 5 | 18 |
| 52 | 4 | 5 | 12 |
| 53 | 4 | 3 | 12 |
| 54 | 4 | 11 | 12 |
| 55 | 4 | 3 | 7 |
| 56 | 4 | 10 | 14 |
| 57a | 4 | 10 | 14 |
| 57b | 4 | 10 | 14 |
| 58 | 5 | 9 | 15 |
| 59 | 5 | 9 | 15 |
| 60 | 5 | 9 | 15 |
| 61 | 5 | 9 | 15 |
| 62a | 5 | 9 | 15 |
| 62b | 5 | 9 | 15 |
| 63 | 5 | 6 | 5 |
| 64 | 5 | 10 | 3 |
| 65 | 5 | 10 | 9 |
PFGE analysis of family-related and geographically related cases of infection provided evidence that macrorestriction with only one enzyme may not be efficient enough to distinguish among the isolates. Results for three of these related isolates are presented in Fig. 6. The isolates were from two patients with illnesses in two distinct families (lanes 5 to 12) and two geographically related patients (lanes 1 to 4) from a single village where isolates were obtained after an interval of 3 months. In the first series of family-related cases of illness, no discrimination was obtained by macrorestriction with XbaI and SpeI (Fig. 6, lanes 5 to 8 and 10 to 12), even though the isolates were sampled over a period of 1 month in the second series (patients 35 and 38). However, in the two geographically related series, while the pulsotypes of the isolates from patients 24 and 21 obtained after restriction with XbaI appeared to be almost identical, with only a slight difference in the resolution of two DNA fragments of about 265 kb (Fig. 6, lanes 1 and 2), the corresponding pulsotypes for the isolates obtained after restriction with SpeI (Fig. 6, lanes 3 and 4) were different because of the different lengths of one DNA fragment (215 and 240 kb). After restriction with XbaI, pattern 1′ for the isolate from patient 24 was finally distinguished from pulsotype 1 for the isolates from patients 21, 23, and 22, which also showed a unique and more intense DNA band at 275 kb. In fact, it seemed that after restriction with SpeI a 240-kb fragment in the pulsotype of the isolate from patient 24 was reduced to a 215-kb DNA fragment in the pulsotype of the isolate from patient 21. Such a variation was not attributed to DNA separation since no variation in the other DNA fragments was observed. It must be emphasized that the isolate from patient 21 was obtained 3 months after the isolate from patient 24 was obtained, suggesting the loss of a small genetic element in the isolate from patient 21 that was observed in the profile obtained after restriction with SpeI. The two corresponding isolates may be assumed to be related, suggesting evolution within B. pertussis isolates.
FIG. 6.
PFGE analysis of geographically related and family-related B. pertussis isolates. Lanes: L, polymers of bacteriophage λ DNA (New England Biolabs); 1, 2, 3, 4, 9, and 10, XbaI-restricted intact DNAs of isolates from patients 24, 21, 62a, 62b, 38, and 35, respectively; 3, 4, 7, 8, 11, and 12, SpeI restriction of the same DNAs, respectively.
DISCUSSION
Although the sampling procedure and the diagnostic techniques have been identical during the past 7 years, the numbers of cases of whooping cough and the rates of isolation of B. pertussis have increased in Eastern France, to a ratio of two cases of bacterial isolation per 100,000 individuals in 1994, but the AC tests revealed a sixfold greater incidence of pertussis among people. This incidence of whooping cough is greater than the evaluation of 3.5 cases per 100,000 individuals in France (2). As reported previously (6) and according to the data obtained in this study, newborns and young infants constitute the population at major risk, but all the patients included in the study presented with an incomplete vaccination status. In fact, normally vaccinated children in France represent 95% of the population less than 6 months old, and 90% of the population from 2 to 7 years of age has been vaccinated, whereas the booster dose given at age 6 years remains facultative (4). It must be noted that for about 10 years (1980 to 1990) the last immunization was rarely administered in eastern France. This lack of receipt of the last immunization may have favored the asymptomatic or nonsymptomatic carriage of B. pertussis by previously immunized people without the development of true whooping cough. Noncompliance with immunization due to the fear of well-publicized neurologic side effects can also play a role in the upsurge in the incidence of pertussis. While a diagnosis of whooping cough remained poorly evoked in adults (3, 10), such persons who are transitory infectious carriers may be vectors for the spread of a B. pertussis strain of a given pulsotype (isolates with patterns 13, 14, 15, 1, and 4 after restriction with XbaI were more frequently detected in this study). This would explain the absence of an epidemiological relatedness of the patients included in the study. This hypothesis is strengthened by the observation that incompletely vaccinated people with pertussis were found in populations positive for pertussis by both a positive AC test result and bacterial isolation.
Again, the study with the 69 B. pertussis isolates confirmed that the results of PFGE analysis provided epidemiological markers of choice (9, 17). Even though different data were obtained for Bordetella bronchiseptica isolates (19), variations in B. pertussis were not found by three ribotyping procedures. The use of XbaI and SpeI restriction endonucleases provided consistent assessments of the polymorphisms of B. pertussis strains. Restriction with XbaI, SpeI, and DraI identified 19, 15, and 5 different pulsotypes, respectively, and accounted for the polymorphisms of the bacteria collected over a 7-year period in a small geographic area, although 76% of the isolates were collected during the last 3 years. Use of combinations of PFGE patterns greatly improved the ability to distinguish the isolates since 35 patterns were recorded when combinations were used. Nevertheless, it appeared that each XbaI, SpeI, and, to a lesser extent, DraI restriction site may have concerned regions susceptible to genetic rearrangements. In this study, for isolates from patients with whooping cough in a small village detected over a 3-month interval, two clonal XbaI profiles were observed, and further examination of the SpeI fingerprints improved the ability to discriminate between the isolates.
The frequent pulsotype combinations did not necessarily account for a direct epidemiological link between isolates and the corresponding patients. After comparison, pattern 13 for isolates from patients 31, 33, and 40 obtained after restriction with XbaI (Table 1 and Fig. 3) seemed to be identical to the pattern most encountered for isolates involved in an outbreak at Fort Smith, Alberta, Canada (9), as well as the pattern for isolates involved in another recent French study of isolates from Paris and its suburbs (17). This observation suggests that B. pertussis is spread not only by people who have developed a true illness. Adults and health care workers may be affected by unrecognized B. pertussis infections (10, 12, 16). This bacterial carriage may be endemic and may be unknown for most patients. The efficacy of the common pertussis vaccine, which is estimated to decrease significantly during the 5th year after immunization (5, 8, 11, 16), strengthened this hypothesis. In a recent work, 25% of the adult population with a cough for 2 or 3 weeks had nonsymptomatic pertussis (24), in accordance with the results of this study. In conclusion, the rapid processing of samples from patients with whooping cough allowed significant detection of B. pertussis, although to a lesser extent than the AC test did. PFGE typing with two or three endonucleases, e.g., XbaI, SpeI, and DraI, confirmed the high degree of polymorphism among B. pertussis strains and strengthened the ability to distinguish isolates, but the predominant strains suggested the long-term spread of the bacteria. These routes of spread might be reduced by more efficient vaccination programs.
ACKNOWLEDGMENTS
We are very grateful to the pediatricians and pediatric care units, especially those at the Strasbourg University Hospital, for assistance and information concerning the patients. We thank D. Young for help with the English.
This work was supported by funds from the Institut de Bactériologie de la Faculté de Médecine de Strasbourg.
REFERENCES
- 1.Beall B, Cassiday P K, Sanden G N. Analysis of Bordetella pertussis isolates from an epidemic by pulsed-field gel electrophoresis. J Clin Microbiol. 1995;33:3083–3086. doi: 10.1128/jcm.33.12.3083-3086.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Baron S, Bégué P, Desenclos J C, Drucker J, Grimpel E, Guiso N, Njamkepo E. Evaluation épidémiologique et microbiologique de la coqueluche en France en 1993–1994. Bull Epidemiol Hebd. 1995;19:83–85. [Google Scholar]
- 3.Bass J W, Wittler R R. Return of epidemic pertussis in the United States. Pediatr Infect Dis J. 1994;13:343–345. doi: 10.1097/00006454-199405000-00002. [DOI] [PubMed] [Google Scholar]
- 4.Bégué P, Baron S, Grimpel E. Epidémiologie de la coqueluche en Europe en 1995. Med Mal Infect. 1995;25(Suppl. 1):1263–1267. [Google Scholar]
- 5.Bortolussi R, Miller B, Ledwith M, Halperin S. Clinical course of pertussis in immunized children. Pediatr Infect Dis J. 1995;14:870–878. doi: 10.1097/00006454-199510000-00010. [DOI] [PubMed] [Google Scholar]
- 6.Christie C D C, Marx M L, Marchant C D, Reising S F. The 1993 epidemic of pertussis in Cincinnati. N Engl J Med. 1994;331:16–21. doi: 10.1056/NEJM199407073310104. [DOI] [PubMed] [Google Scholar]
- 7.Confer D L, Zackrisson G, Lagergard T, Eaton J W. Rapid diagnosis of pertussis. Scand J Infect Dis. 1990;22:175–177. doi: 10.3109/00365549009037899. [DOI] [PubMed] [Google Scholar]
- 8.Davis S F, Sutter R W, Strebel P M, Orton C, Alexander V, Sanden G N, Cassell G H, Thacker L, Cochi S L. Current outbreaks of pertussis and Mycoplasma pneumoniae infection: clinical and epidemiological characteristics of illnesses manifested by cough. Clin Infect Dis. 1995;20:621–628. doi: 10.1093/clinids/20.3.621. [DOI] [PubMed] [Google Scholar]
- 9.de Moissac Y R, Ronald S L, Peppler M S. Use of pulsed-field gel electrophoresis for epidemiological study of Bordetella pertussis in a whooping cough outbreak. J Clin Microbiol. 1994;32:398–402. doi: 10.1128/jcm.32.2.398-402.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Deville J G, Cherry J D, Christensen P D, Pineda E, Leach C T, Kuhls T L, Viker S. Frequency of unrecognized Bordetella pertussis infections in adults. Clin Infect Dis. 1995;21:639–642. doi: 10.1093/clinids/21.3.639. [DOI] [PubMed] [Google Scholar]
- 11.Halperin S A, Bortolussi R, Wort A J. Evaluation of culture immunofluorescence and serology for the diagnosis of pertussis. J Clin Microbiol. 1989;27:752–757. doi: 10.1128/jcm.27.4.752-757.1989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.He Q, Viljanen M K, Nikkari S, Lyytikainen R, Mertsola J. Outcomes of Bordetella pertussis infection in different age groups of an immunized population. J Infect Dis. 1994;170:873–877. doi: 10.1093/infdis/170.4.873. [DOI] [PubMed] [Google Scholar]
- 13.Houard S, Hackel C, Herzog A, Bollen A. Specific identification of Bordetella pertussis by the polymerase chain reaction. Res Microbiol. 1989;140:477–487. doi: 10.1016/0923-2508(89)90069-7. [DOI] [PubMed] [Google Scholar]
- 14.Khattak M N, Matthews R C. Genetic relatedness of Bordetella species as determined by macrorestriction digests resolved by pulsed-field gel electrophoresis. Int J Syst Bacteriol. 1993;43:659–664. doi: 10.1099/00207713-43-4-659. [DOI] [PubMed] [Google Scholar]
- 15.Lane D J. 16S/23S rRNA sequencing. In: Stackebrandt E, Goodfellow M, editors. Nucleic acids techniques in bacterial systematics. Chichester, United Kingdom: John Wiley & Sons Press Ltd.; 1991. pp. 115–176. [Google Scholar]
- 16.Mink C A, Sirota N M, Nugent S. Outbreak of pertussis in a fully immunized adolescent and adult population. Arch Pediatr Adolesc Med. 1994;148:153–157. doi: 10.1001/archpedi.1994.02170020039006. [DOI] [PubMed] [Google Scholar]
- 17.Moissenet D, Valcin M, Marchand V, Grimpel E, Bégué P, Garbarg-Chenon A, Vu-Thien H. Comparative DNA analysis of Bordetella pertussis by pulsed-field gel electrophoresis, randomly amplified polymorphism DNA, and ERIC-polymerase chain reaction. FEMS Microbiol Lett. 1996;143:127–132. doi: 10.1111/j.1574-6968.1996.tb08471.x. [DOI] [PubMed] [Google Scholar]
- 18.Prévost G, Pottecher B, Dahlet M, Bientz M, Mantz J M, Piémont Y. Pulsed field gel electrophoresis as a new epidemiological tool for monitoring methicillin-resistant Staphylococcus aureus in an intensive care unit. J Hosp Infect. 1991;17:255–269. doi: 10.1016/0195-6701(91)90270-i. [DOI] [PubMed] [Google Scholar]
- 19.Register K B, Boisvert A, Ackermann M R. Use of ribotyping to distinguish Bordetella bronchiseptica isolates. Int J Syst Bacteriol. 1997;47:678–683. doi: 10.1099/00207713-47-3-678. [DOI] [PubMed] [Google Scholar]
- 20.Sambrook J, Fritsch E F, Maniatis T. Molecular cloning: a laboratory manual. Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory Press; 1989. [Google Scholar]
- 21.Scheftel J M, Matis J, Simeoni U, Lutz P, Haubensack M, Geisert J, Minck R. Use of Bordetella pertussis adenylate cyclase in the diagnosis of whooping cough in infants and young children. Med Microbiol Lett. 1992;1:78–84. [Google Scholar]
- 22.Stainer D W, Scholte M J. A simple chemically defined medium for the production of phase I Bordetella pertussis. J Gen Microbiol. 1971;63:211–220. doi: 10.1099/00221287-63-2-211. [DOI] [PubMed] [Google Scholar]
- 23.Stibitz S, Garletts T L. Derivation of a physical map of the chromosome of Bordetella pertussis Tohama I. J Bacteriol. 1992;174:7770–7777. doi: 10.1128/jb.174.23.7770-7777.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Strebel P, Nordin J, Edwards K, Hunt J, Besser-Wiek J, Burns J, Amundsen J. Program and abstract of the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington, D.C: American Society for Microbiology; 1997. Estimated incidence of pertussis among persons aged 10–49 years, Minnesota, 1995–1996, abstr. K-165; pp. 357–358. [Google Scholar]
- 25.Syedabubakar S N, Matthews R C, Preston N W, Owen D, Hillier V. Application of pulsed-field gel electrophoresis to the 1993 epidemic of whooping cough in the UK. Epidemiol Infect. 1995;115:101–113. doi: 10.1017/s0950268800058167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wadowski R M, Michaels R H, Libert T, Kingsley L A, Ehrlich G D. Multiplex-PCR-based assay for detection of Bordetella pertussis in nasopharyngeal swab specimens. J Clin Microbiol. 1996;34:2645–2649. doi: 10.1128/jcm.34.11.2645-2649.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]