Campylobacter Species - Journal of Clinical Microbiology - American ...

2 downloads 0 Views 594KB Size Report
Itoh, T., Y. Yanagawa, M. Shingaki, M. Takahashi, A. Kai, M. Ohashi, and G. Hamana. 1987. Isolation of Campylobacter pyloridis from human gastric mucosa and ...
JOURNAL OF CLINICAL MICROBIOLOGY, May 1991, p. 923-926 0095-1137/91/050923-04$02.00/0 Copyright C 1991, American Society for Microbiology

Vol. 29, No. 5

Effect of Inoculum Size on the Phenotypic Characterization of Campylobacter Species S. L. W. ON AND B. HOLMES* Identification Services Laboratory, National Collection of Type Cultures, Central Public

Health

Laboratory, London NW9 5HT, United Kingdom

Received 21 November 1990/Accepted 7 February 1991

The type strains of six Campylobacter species or subspecies were examined in eight tests used for the identification of such organisms. False-positive results were obtained in certain tests (growth on 1% bile, brilliant green, selenite, trimethylamine-N-oxide, 2,3,5-triphenyl tetrazolium chloride, and minimal media) when an inoculum yielding 107 to 108 CFU/ml was used. Each tolerance test was examined with blood, nutrient, and brucella agars as basal media. The type of basal medium used could also affect the test outcome. With the inoculum standardized to a density yielding approximately 106 CFU/ml, reproducible and pertinent results were obtained, provided an appropriate basal medium was used. However, 95% confidence limits for viable counts done on these basal media indicated that blood agar may yield more consistent results than the other agars examined.

Campylobacters are responsible for a number of clinical conditions and show considerable diversity of habitat (11). Campylobacter coli and C. jejuni subsp. jejuni are the most commonly isolated, and both cause enteritis in humans. Other, less frequently isolated campylobacters may be associated with various clinical conditions, including enteritis (6, 7, 10, 15-17). Certain species are largely of veterinary importance (5, 18), while one has so far been isolated only from the roots of salt marsh plants (8). The identification of campylobacters is difficult, since the strains have relatively fastidious growth requirements and are asaccharolytic and few biochemical tests have been found to provide discrimination. Alternative tests, such as resistance to chemical compounds, temperature tolerances, and growth conditions, are widely used. However, most workers perform the tests by methodologies unique to their own laboratories. Differences in methodologies may lead to differences in the outcomes of tests that are nominally the same. Certainly, there are discrepancies in phenotypic data published for campylobacters: for example, there are differences in the proportions of strains of C. fetus subsp. fetus showing tolerance of 1.5% NaCl (95% [13] to 0% [1]) and in the proportions of strains of Helicobacter pylori (formerly Campylobacter pylorn) reported to be tolerant of 1% glycine (100% [1] to 0% [6]). Less marked differences in campylobacter phenotypic data include the percentages of C. laridis strains exhibiting resistance to nalidixic acid (100% [1] to 69% [3]). Although differences in published data may reflect geographical variations, such extremes would not be expected. It is well established in antimicrobial susceptibility testing that the inadequate or inappropriate standardization of inocula can result in false-positive or -negative results (2, 4, 19). The aim of the present study was to determine the optimal inoculum size for certain tests commonly used in the differentiation of Campylobacter species.

*

MATERIALS AND METHODS Bacterial strains. The six strains used in this study are listed in Table 1 (all are type strains). Strains were designated as positive or negative controls for each test (Table 1) on the basis of the original species descriptions and also on the speed and ease of growth of each strain. Media. The basal media used were as follows: (i) nutrient broth no. 2 (Oxoid Ltd., Basingstoke, England) supplemented with 5% (all percentages are expressed as volume/ volume or weight/volume, when applicable) whole horse blood added after the autoclaved medium was cooled to approximately 50°C; (ii) nutrient broth no. 2, unsupplemented; and (iii) brucella broth (Difco Laboratories Ltd., East Molesey, England), unsupplemented. In all cases, 2% Japanese agar was used as the gelling agent. Various media were prepared by adding the following test substances to each of the three basal media before sterilization by autoclaving: 1% bile (desiccated ox bile; Difco), 1% glycine (BDH Ltd., Poole, England), 0.1% trimethylamine N-oxide (TMAO; Sigma Chemical Co. Ltd., Poole, England), and 1.5% NaCl (BDH; bringing the total NaCl concentration to 2%, as 0.5% NaCl is already present in each basal medium). Brilliant green medium was prepared by adding 1 ml of a 1% solution of brilliant green (BDH) to 1 liter of basal medium. Selenite (0.1%) and 2,3,5-triphenyl tetrazolium chloride (TTC; 0.04%) media were prepared by adding aseptically filter-sterilized solutions containing these test substances to each of the three autoclaved basal media. A minimal medium was prepared to the specifications of Roop et al. (12) but with a higher agar concentration (2%). Inoculum preparation. All bacterial suspensions were made in nutrient broth no. 2 (Oxoid). For each strain, an initial suspension containing approximately 108 CFU/ml was prepared by comparison with a calibration curve and adjustment of the optical density accordingly. The latter was determined with a Titertek Uniskan I (Flow Laboratories Ltd., Rickmansworth, Hertfordshire, England) fitted with a 405-nm filter. From each initial suspension, a range of suspensions containing 108 to 103 CFU/ml in 10-fold dilution steps was prepared. The mean number of CFU per milliliter

Corresponding author. 923

924

J. CLIN. MICROBIOL.

ON AND HOLMES TABLE 1. Strains used as positive and negative controls for the characters tested Test

Positive control

Negative control

1% Bile Brilliant green 1% Glycine 2% NaCl Selenite TMAO TTC Minimal medium

C. jejuni NCTC 11351 C. coli NCTC 11366 C. jejuni NCTC 11351 C. sputorum subsp. bubulus NCTC 11367 C. jejuni NCTC 11351 C. laridis NCTC 11352 C. jejuni NCTC 11351 C. coli NCTC 11366

C. sputorum subsp. bubulus NCTC 11367 C. jejuni NCTC 11351 C. fetus subsp. venerealis NCTC 10354 C. jejuni NCTC 11351 C. fetus subsp. venerealis NCTC 10354 C. jejuni NCTC 11351 C. hyointestinalis NCTC 11608 C. jejuni NCTC 11351

control media (blood, nutrient, and brucella agars). Inoculation of test plates. Air-dried test plates were inoculated with 20-,ul dilutions of the six strains. Only the appropriate positive and negative control strains were inoculated on each medium. The spots were allowed to dry (for not more than 15 min), and the plates were incubated in a Variable Atmosphere Incubator (VAIN; Don Whitley Scientific Ltd., Shipley, Yorkshire, England; calibrated to provide an atmosphere of 2% H2-5% C02-5% 02-88% N2 at 37°C) and examined for up to 3 days. All tests were performed three times, on separate occasions and with different batches of media, to confirm that they were reproducible. Any growth on the test plates was compared with that on the accompanying unsupplemented basal medium control plates and recorded as normal or as only faintly discernible.

brucella agar base. The negative controls often exhibited growth at some dilutions, albeit generally at high viable cell concentrations, ranging from 107 to 108 CFU/ml. In some cases, growth of the negative controls could be disregarded, since it was only faintly discernible (indicated by a "w" in Table 3) and easily distinguished from normal growth. The negative control for brilliant green, with blood and nutrient agars, yielded growth with each of the dilutions used. The same patterns of results were obtained on each occasion tested. Use of a standardized inoculum of 106 CFU/ml. All strains were retested by the same methods as before but with a standardized inoculum of 106 CFU/ml (Table 4). All positive and negative controls behaved as expected, regardless of the basal medium used, with the exception of the positive control for growth on 2% NaCl, when brucella agar was used as the basal medium.

RESULTS

DISCUSSION

Growth on unsupplemented basal medium control plates. Differences in the estimates for viable CFU were seen among the three basal media (Table 2). In addition, 95% confidence limits for counts made on blood agar showed less variance than did those made on either nutrient or brucella agar. The quality of growth of all Campylobacter strains examined was not as good on brucella agar as on nutrient agar and blood agar, with the latter giving the most luxuriant growth of all. Growth on test plates. Positive results for selenite and TTC were accompanied by the development of orange and crimson colors, respectively. The patterns of growth obtained with each of the dilutions of the appropriate control strains on each test medium and on the minimal medium are summarized in Table 3. The positive controls generally performed as expected, except for those for brilliant green and 2% NaCl tolerance, for which growth could only be obtained at a high inoculum density, 2107 CFU/ml, with

It is important to perform phenotypic tests under standardized conditions to avoid misidentifications, particularly in a group, such as the campylobacters, with a complex taxonomic structure. Several factors should be considered in the performance and assessment of phenotypic tests useful in the identification of these organisms. Differences in inoculum size may explain discrepancies in, for example, the rates of tolerance of 1% glycine reported for H. pylori (100% [1] to 0% [6]). Tolerance was seen when the inocula were adjusted to a McFarland standard of 1, which we have found to be approximately equivalent to 1.0 x 108 CFU/ml. In this study, false-positive results were obtained in several key tests, generally when the inoculum size exceeded 107 to 108 CFU/ml. Our results indicate that, for some commonly used tolerance tests, an inoculum size of approximately 106 CFU/ml is the most appropriate (as indeed it is for determining susceptibility to antimicrobial agents [4]) when blood agar, nutrient agar, or brucella agar is

in each initial suspension

was

estimated by

use

of the plate

count technique of Miles and Misra (9) performed on three

TABLE 2. Numbers of CFU of the initial suspension on unsupplemented basal medium controls CFU/ml on the following agar:

Mean

11366 10354 11608 11351 11352 11367

Nutrient 95% Confidence limit (minimum-maximum)

Blood

NCTC

4.16 x 3.05 x 7.6x 3.24 x 3.23 x 3.75 x

108 108 107 108 108 108

95% Confidence limit (minimum-maximum) (3.47-4.84) x 108

(1.70-4.4)x 108 (0.18-1.3)x 108 (2.26-4.2)x 108 (1.27-5.18) x 108 (3.75-3.75) x 108

Mean

7.04 x 7.45 x 2.51 x 2.8x 3.25 x 2.51 x

107 106 107 107

(0.22-1.18) x108 (0.35-1.14)x 107 (0.38-4.6)X 107 (1.8-3.79)x 107

108

(0.39-6.1) x 108 (0.49-4.53)x 108

108

Brucella 95% Confidence limit (minimum-maximum)

Mean

2.07 x 1.9X 4.09x 2.3x 1.47X

108

107 105 108 107 3.6x 106

(0.59-3.5) (0.34-3.46) (0.33-7.85) (0.55-4.04) (1.08-1.86) (0.314.9)

x

108

x 107 x 105 x 108 x 107 x 106

VOL. 29, 1991

PHENOTYPIC CHARACTERIZATION OF CAMPYLOBACTERS

925

TABLE 3. Patterns of growth obtained with 10-fold dilutions of positive and negative control organism suspensions on a minimal medium and on test media with blood agar, nutrient agar, or brucella agar as a base Growth of the following controls with the indicated test substance': Agar base

Dilu-

Positive

tion fco

%

Bilr

Blood

108 107 106

S05

104

103 Nutrient

108

107 106

i05

104 103 Brucella

BiBi 1%

liant

green

1

Glycine

2%

NaCl

Negative

Sele-Mima TMAO nite

medium

1

cine

g

+++ +++ +++ +++ +++ +++ +++

+++ ++

+++ ++ (w)

+++ +++ +++ ++ +

+++ +++ +++ ++ +

+++ ++ +++ ++ +

+++ +++ +++ +++ ++

+++ +++ ++ ++ +

+++ +++ +++ ++ ++

+++ +++ +++ ++ +

+++ +++ +++ ++ ++

+++ +++ +++ ++ +

+

_

-

N2% SeleNaC Ie

-

-

-

-

_ _

-

Minimal

TMAOmmedium +++ (w) +

+++

+++ (w) + +++ (W) ++ (w) ++ (w) -

+++ ++ (W) ++ (w)

+ (w)

-

+++ +++ +++ +++ +++ +++ +++

-

++ +++ (w) -

+++

-

+++ +++ (w) +++-

-

+++ +++ +++ +++ +++

-

+

-

-

+++(w)

+++ +++ +++ +++ +++ +++ +++

-

-

++(w)

++ +

-

++

+

+++ +++ +++ +++ +++ ++ +++ +++ ++ + ++ ++

+++ +++ ++

+++ +

-

+++

-

++

-

+

+

-

+

-

108

+++ +++ +++ +++ +++ +++ +++

107

+++ ++

106

+++ ++

105 104 103

% Bile

Bl-

++ +

-

+++ +++ +++ +++ +++ _ ++ ++ ++ +++ ++

-

-

++

-

+

-

-

-

-

-

+++ +++ ++ (w) ++ +

-

-

-

-

-

-

-

-

-

-

++

+

++

-

+

+

++

-

-

-

+

-

+

-

-

-

(w) +_ (w)

-

-

-

-

-

++

_

+++ ++ -

-

-

-

-

_ _

_ _

_ _

_ _

-

_

a+ + +, Confluent or semiconfluent normal growth; + +, uncountable (.40 colonies, normal growth); +,