Previous Article | Next Article ![]()
Journal of Clinical Microbiology, April 2003, p. 1664-1672, Vol. 41, No. 4
0095-1137/03/$08.00+0 DOI: 10.1128/JCM.41.4.1664-1672.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Laboratoire de Microbiologie, Centre National de Référence des Haemophilus influenzae, Hôpital Purpan,1 Faculté de Médecine Toulouse Purpan, 31059 Toulouse Cédex 9,2 Hôpital de l'Archet, 06202 Nice,3 CHRU, Hôpital Calmette, 59037 Lille Cédex,4 CHU Jean Minjoz, 25030 Besançon Cédex,5 France INSERM 99-22, CHU Bichat-Claude Bernard, 75877 Paris Cédex 18, France6
Received 5 August 2002/ Returned for modification 30 August 2002/ Accepted 9 December 2002
|
|
|---|
|
|
|---|
In the study of Faden et al. (9), 200 children were monitored from birth through to 2 years of age to determine the nasopharyngeal colonization pattern of nontypeable H. influenzae. Forty-four percent of the children were colonized on one or more occasions, and the acquisition rate was greatest in the first year. Colonization mainly involves a dominant strain which can be followed by a series of other strains (1, 9, 15, 24). The level of carriage varies according to different factors, i.e., age, siblings, and living conditions. All of these factors can also have repercussions on the carriage of strains resistant to antibiotics or encapsulated type b strains. Anti-Hib vaccination has led to a sharp drop in the carriage of encapsulated type b strains (33).
The turnover of noncapsulated H. influenzae strains in the nasopharynx of healthy children and of otitis-prone children has been the object of several studies (9, 24, 27, 29, 35). These results were obtained in children attending or not attending day care centers (DCCs) or living in orphanages.
The present study was designed to monitor nasopharyngeal carriage of H. influenzae in healthy children attending DCCs in three different geographic areas of France. The technique of pulsed-field gel electrophoresis (PFGE) was used to characterize the epidemiology of H. influenzae colonization of children in the DCC and in the whole geographic area. Our goals were to determine whether the children were colonized with specific strains in a given geographic area, whether the same strains were observed over a period of time, what their turnover was between the seasons, and what was the profile of the ß-lactamase-producing strains.
|
|
|---|
Approval for the study was obtained from the Ethical Committee (Comité Consultatif pour la Protection des Personnes dans la Recherche Biomédicale, Aix en Provence, France, 1 December 1998), and in each area, the study was initiated with the agreement of the local health authority and of the DCC manager. Parental consent was obtained and a questionnaire was completed for each participant providing information regarding a subject's age, sex, family size, history of recent respiratory infection or antibiotic therapy and vaccine status (in particular, the anti-Hib status). In each area, the DCCs were randomized for the two study periods. In each DCC, and for each study period, the children were randomly selected among those for whom written parental consent had been given. Randomization was done by using a two-stage sampling method choosing the primary units (DCC) with probability proportional to size.
Nasopharyngeal samples were obtained by using a silicon cannula with a flexible tube fixed onto a syringe (Vygon, Paris, France). The tubes were transported in transport medium (Portagerm; bioMérieux, Marcy l'Etoile, France) to a designated hospital bacteriology laboratory in each area.
Specimens were plated on chocolate agar and sheep blood agar plates (bioMérieux) and incubated overnight at 37°C (under a CO2-enriched atmosphere for chocolate agar plates). H. influenzae isolates were identified according to standard microbiological procedures (25). Isolates were kept frozen locally at -70°C before being sent to the H. influenzae National Reference Center in Toulouse, France.
Bacteriological methods. Capsular typing was done by slide agglutination with type-specific antiserums a to f (Difco, BD, Le Pont de Claix, France). The capsular type was confirmed by molecular typing with the primers and methods reported by Falla et al. (10).The biotype was determined as described by Kilian (19).The production of ß-lactamase was assessed by a chromogenic cephalosporin test (Nitrocefine; bioMérieux). The ß-lactamase type, TEM or ROB, was determined by PCR by using the primers for blaTEM and blaROB proposed by Tenover et al. (34). Antibiotic susceptibility was determined by the disk diffusion test on Haemophilus test medium (18) with purchased disks (Bio-Rad, Marnes la Coquette, France) and by MIC determination by an agar dilution method which used Haemophilus test medium agar with an inoculum of 104 cells per spot (5).
PFGE. PFGE was performed as described previously (7) with a contour-clamped homogeneous electric field system (CHEF DR III apparatus; Bio-Rad Laboratories). PFGE fingerprints were analyzed with Bio-Profil 99 software, Bio-1D software was used for the gel analysis, and Biogene software was used for pattern comparison (Vilber Lourmat, Marne La Vallée, France). The similarity of the PFGE banding patterns was estimated with the Dice coefficient and the unweighted pair group method using arithmetic average algorithm.
Various dendrograms were obtained: one with the genotypes of all strains, one with the genotypes of the strains isolated in May to June, one with the genotypes of the strains isolated in November to December, and those with the genotypes of the strains from each area period by period. Each dendrogram is composed of restriction patterns for single isolates or for two or more isolates. The diversity index (DI) is defined as the ratio of the number of restriction patterns to the number of strains. An index close to 1 represents maximum diversity or heterogeneity; an index of 0.01 indicates maximum clonality or identity. The diversity is also assessed by the percentage of patterns for a single isolate (percentage of the number of patterns for single isolates over the number of patterns). A percentage close to 100 represents maximum heterogeneity.
Statistical analysis. Statistical analysis was performed by using EPI-INFO 6.4.fr software.
|
|
|---|
The characteristics of the strains according to the area and the period are shown in Table 1. Overall, the carriage rate of H. influenzae was 40.9% (688 out of 1,683 children), with a carriage peak between 21 and 24 months (56%). The prevalence of encapsulated H. influenzae carriers in children under 3 years was 1% (17 of 1,683), the prevalence for type f was 0.4% (7 of 1,683), and the prevalence for type e was 0.6% (10 of 1,683). None of the isolates were of type b.
|
View this table: [in a new window] |
TABLE 1. Carriage and characteristics of H. influenzae strains isolated during the two periods (year 1999) from children attending DCCs in three different geographic areas of France
|
H. influenzae carriage during the May-to-June 1999 period. During the May-to-June 1999 period, nasopharyngeal samples were taken from a total of 879 children. Overall, the carriage rate of H. influenzae was 41.8% (368 of 879). From area to area, H. influenzae carriage was 34.8% in Alpes Maritimes, 36.2% in Doubs, 58.8% in Nord (P < 0.001) (Table 1). Among the strains, 7 were capsulated, 3 were of type f (0.8%) and 4 were of type e (0.9%), and isolated in the Doubs and Nord areas. For the whole set (160 of 368), 43.4% of the strains were TEM-type ß-lactamase producers. The percentage of strains producing ß-lactamase was 40.3, 47.4, and 42.5% in Alpes Maritimes, Doubs, and Nord, respectively (P = 0.54). The carriage rate of ß-lactamase-producing strains by children was 14.05% in the Alpes Maritimes, 17.2% in Doubs, and 25% in Nord (Table 1) (P = 0.003).
H. influenzae carriage during the November-to-December period. During this period, 804 children were included for the three areas. Among them, 320 (39.8%) were H. influenzae carriers, i.e., 35.9% in Alpes Maritimes, 38.3% in Doubs, and 47.3% in Nord (P = 0.03). Ten of the corresponding strains were encapsulated (3.1%) (10 of 320), 4 (1.2%) were type f in the Alpes Maritimes and 6 (1.8%) were type e in the Doubs and Nord areas. TEM-type ß-lactamase-producing strains accounted for 46.5% (149 of 320). The proportion of strains producing ß-lactamase was 39.2, 54.3, and 42.4% in the Alpes Maritimes, Doubs, and Nord, respectively (P = 0.035). The carriage rate of ß-lactamase-producing strains was 14.1% in Alpes Maritimes, 21.2% in Doubs, and 19.6% in Nord (P = 0.065).
PFGE. From the 688 strains, 25 repeatedly yielded unsatisfactory weak pulsed-field gel electropherograms or were lost and were omitted from further investigations.
An example of PFGE fingerprints obtained after SmaI digestion is shown in Fig. 1. Analysis was carried out by using various dendrograms: that obtained from the PFGE fingerprints of the whole set of strains (663 strains), those obtained with the strains from each period (May to June, 353 strains; November to December, 310 strains) (Fig. 2), and those obtained with the strains from each area including just one or both of the periods (Fig. 3).
![]() View larger version (89K): [in a new window] |
FIG. 1. PFGE profiles of SmaI-restricted H. influenzae DNA of strains isolated from children attending DCCs in the Doubs area during the November-to-December period. Lanes (left to right): 1, H. influenzae ATCC 10211; 2, child from DCC 04; 3 to 8, children from DCC 05; 9, child from DCC 06; 10, lambda ladder; 11, child from DCC 07; 12 to 17, children from DCC 08; 18 and 19, children from DCC 09; 20, lambda ladder; 21 to 25, children from DCC 09; 26, child from DCC 08; 27 and 28, children from DCC 06; 29, uninterpretable; 30, H. influenzae ATCC 10211.
|
![]() ![]() ![]() ![]() View larger version (77K): [in a new window] |
FIG. 2. PFGE analysis dendrogram showing the genetic relationship among strains of H. influenzae isolated from children attending DCCs in three different geographic areas of France during the period from November to December 1999. For each area the number indicates the number of strains in the pattern; -, no strain in the pattern. The dendrogram is shown in four portions, in order from left to right, with the bottom of one portion corresponding to the top of the next.
|
![]() ![]() ![]() View larger version (50K): [in a new window] |
FIG. 3. PFGE analysis dendrogram showing the genotypic relationship among strains of H. influenzae isolated from children attending DCCs in the Alpes Maritimes area during the two study periods. For each period the number indicates the number of strains in the pattern; -, no strain in the pattern. The dendrogram is shown in three portions, in order from left to right, with the bottom of one portion corresponding to the top of the next.
|
|
View this table: [in a new window] |
TABLE 2. Comparison of restriction pattern distribution of H. influenzae strains for the two study periods
|
The children who were carriers of H. influenzae during the May-to-June period and also in November to December had strains with different restriction patterns during the two periods.
According to the area and the period, the number of restriction patterns was between 81 and 115 for the May-to-June period and between 71 and 77 for the November-to-December period. The DI was always higher during the May-to-June period (Table 3).
|
View this table: [in a new window] |
TABLE 3. Comparison of restriction pattern distribution of H. influenzae strains according to the area and study period
|
Among the 306 TEM-type ß-lactamase-producing strains, 297 were studied by PFGE and 194 patterns were found, including 103 patterns for single isolates. Within the same pattern, ß-lactamase-producing strains can coexist alongside non-ß-lactamase-producing strains.
Depending on the area, 24 to 28 new patterns were observed among the strains isolated in November to December. The percentage of strains in new genotypes during this period, compared to the situation in May to June, was 66 to 73% (Table 4).
|
View this table: [in a new window] |
TABLE 4. Comparison of restriction pattern distribution of ß-lactamase-producing H. influenzae strains according to the area and study period
|
|
|
|---|
Carriage of encapsulated strains other than type b was low: 0.4% carriers of type f and 0.6% carriers of type e. These two capsular types are usually found in children of preschool age (11, 23) and school age (3) or living in orphanages (27).
High genetic heterogeneity of nontypeable H. influenzae isolates has been reported in previous studies whatever the epidemiological markers used (4, 14, 21, 29, 30). Among 178 epidemiologically unrelated isolates, Saito et al. found 165 genotypes (28). On the other hand, among the 111 strains isolated from children living in the closed community of an orphanage, only 13 PFGE profiles were detected, with 83% of the isolates belonging to 5 patterns (27). In our study concerning the strains of H. influenzae colonizing children attending DCCs, 366 genotypes were found for 663 strains.
This study did not focus on the fate of the colonization in one child, like in the investigations of Trottier et al. (35) and Raymond et al. (27). It did not, therefore, assess the dynamics and the diversity of colonization in individual children. It does, however, provide a picture of the dynamics and the diversity of colonization in a community made up of children attending a DCC in a geographic area and finally in three areas during two periods 5 to 6 months apart.
A comparison of strains isolated during the same period in the three areas, providing an image on the national scale, shows that a third of the strains were simultaneously present in the three areas. This corresponds to the dispersion of strains (genotypes) in a population of children who are geographically remote and who have no contact. The same dispersion of the strains is observed on the scale of the area; children attending different DCCs are carriers of strains with the same genotype.
During a prospective study carried out with three children attending a DCC, it was shown that colonization of the nasopharynx by H. influenzae was a dynamic process corresponding to the carriage of a single strain for several months, the loss of this strain followed by the acquisition of a new strain (31). This same dynamic colonization process is also shown in the study by Trottier et al. (35) where, in the same children, there are resident strains and transient strains; after acquisition and carriage of a strain, there is loss followed or not by the acquisition of a new strain.
Among the children participating in the present study, colonization is also seen to be a dynamic process with persistencein the whole population, not in individual childrenof only 30% of the strains at a 6-month interval. This renewal of almost 70% of the strains was observed independently in the three geographic areas. It cannot be totally attributed to different children being recruited for the two study periods and stresses the diversity and the genetic heterogeneity of the nontypeable strains of H. influenzae.
Such a change of genotypes could be due to the introduction of new strains or to strains already present but undetected. Horizontal transfer of strains has been observed in DCCs for pneumococcus, with a turnover of 64% of the strains in the same children in the absence of selection pressure by antibiotics (2).
These high rates of clonal replacement have also been observed for Streptococcus mitis, which undergoes constant change in its natural habitat, the mouth and pharynx, among the members of a family. The strains of S. mitis colonizing the adults and children of one family belong to a multitude of distinct genotypes and change continually. This type of dynamics in the S. mitis population can result from exchanges within the family unit (but in fact few have been detected) or from the passage from one habitat to another (mouth and pharynx) or it could be due to in situ recombination (16).
For H. influenzae, this type of situation has not been reported. However, the fact that strains with reduced sensitivity to ampicillin and having the same genotype are found in adults and children supports the notion of the exchange of strains between the two populations and can partly explain the renewal of the strains colonizing children (7). So, the diversity is not thought to be simply due to exchanges of strains within DCCs leading to increased heterogeneity or genomic diversity but also to exchanges with adult subjects.
The carriage of strains resistant to antibiotics and in particular to ß-lactam is a source of concern in numerous countries and involves both Streptococcus pneumoniae and H. influenzae. In H. influenzae, it is the production of TEM-type ß-lactamase that is the most frequent. Unlike in other countries, where the prevalence of ß-lactamase is low (lower than 10%) (17, 26), in France the frequency of ß-lactamase-producing strains is high, over 35%, in particular for strains isolated from the nasopharynx or those causing acute otitis media (6, 13, 27).
In the present study, 44.5% of the strains produced TEM-type ß-lactamase. These strains showed a genomic diversity comparable to that of the whole set of strains. In addition, the dynamics of colonization by ß-lactamase-producing strains is also comparable to that of the sensitive strains with, depending on the area, a renewal rate of 66 to 73% of the strains over the 6-month study period. The spread of strains producing ß-lactamase does not correspond to clonal expansion under selection pressure (or not) from ß-lactam antibiotics. This result does not support the implantation of one or several strains in an area followed by their expansion and stability over time. It rather suggests acquisition of resistance by the transfer of genes, as of plasmids or transposons. The molecular epidemiology of these genetic elements should be able to confirm this hypothesis.
We thank all those who participated in the OTIFLOR study: Bruno Grandbastien, Alain Martinot, Nathalie Bernard-Rémy, and Sandrine Piechel, CHRU Lille; Brigitte Dunais, CHU Nice; Jean-Marie Estavoyer, CHU Besançon; and Didier Guillemot, Institut Pasteur, Paris, France.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»