International Journal of Food Microbiology, 16 (1t392) 13I- 139 © 1992 Elsevier Science Publishc~ B.V. All rights reserved 11168-16115/92/$1)5.00

131

FOOD 005119

Variation in growth kinetics and phenotype of Aeromonas spp. from clinical, meat processing and fleshfood sources J. Andrew Hudson Meat bulustry Research Institute of New Zealand. tlamihon. N~,n.Zealand

(Received 17 Octuber 1991;accepted 17 April 1992)

Sixty-four strains of motile aerormmads (A. hydrophila. A. sohna and A. ctu'iae), isolated from clinical meal processing and ready-to-eat fleshfood sources, and the A. hydrophila type strain were tested with respect to their growth kinetics at 4°C and 37°C. and the reportcd indicators of pathogenicity: autoagglutinatinn and haemolysis(tested using a CAMP reaction}. Between the species, A. cm'iae grew the fastest at 37~C and had the highest percentage of strains nut able to grow al 4°C {after 200 h incubation). Within the species, food-derivedstrains of A, hydrophila were better adapted to growth at lower temperatures than Ihose from clinical or meat proccssiog sources. Clinical strains of A. hydrophila autoagglutinated more frequently than those from other ~mrces, but no differences in CAMP reactions were noted. Acrommms ravine and A. sohrin isolates appeared to b¢ homogeneous with respect to growth kinetics at the temperatures tested. A comparison of the growth kinetics of the A. hydrophila type strain and a food-derived A. hydrophila strain clearly reflected the latter's enhanced ability to grow at low temperatures. Key words: At,rmuonas; Growth kinetics; Food

Introduction Motile a e r o m o n a d s are r e g a r d e d as being capable of growth at refrigeration t e m p e r a t u r e s . Reports in the literature give m i n i m u m growth t e m p e r a t u r e s of - 0 . 1 - 1 . 6 ° C (Walker and Stringer, 1987), and 0-10*C ( R o u f a n d Rigney, 1971). Many experiments on these organisms have b e e n carried out at 4 or 5°C (e.g., Knlachel, 1990). Toxigenic isolates from chickens, however, were shown to be mesophiles (Kirov et al,. 1990), with m i n i m u m theoretical growth t e m p e r a t u r e s , predicted by extrapolation from the curve produced by plotting the square root of the growth rate versus t e m p e r a t u r e (Ratkowsky et al., 1982), of 4A-6.50C. Kirov ¢t al. (1990) showed that these isolates would c o m p e t e poorly against P s e u d o m o n a s fragi at

Correspondence address: J.A. Hudson, Meat Industry Research Institute of New Zealand (Inc.), PO Box 617, Hamilton, New Zealand. Phone, (07) 8556159.

132 refrigeration temperatures, even though the opposite has been reported for one strain of A. hydrophila in experiments on fish mince and surimi (Ingham and Potter, 1988). It was observed by the author, that of nine Aeromonas hydrophila cultures grown at 4°C, the type strain and one other grew poorly, whereas the remaining seven food-derived strains grew well (unpublished observations), indicating the existence of two distinct groups of strains. Similar observations had been noted by Rouf and Rigney (1971), who classified A. hydrophila strains as either mesophiles or psychrophiles. Further to this, Knochel (1990) showed an association between the source of strains and their growth kinetics. Palumbo et al. (1989) have shown that food-derived A. hydrophila strains are essentially the same as clinical isolates with respect to their biochemistry and virulence-associated characteristics. However, no investigation of the strains' growth kinetics was attempted. In contrast, environmental strains and clinical strains are generally associated with different DNA homology groups (Altwegg et al. 1990) Attempts to correlate biochemical characteristics of motile Aeromonas isolates with pathogenicity have generally found that such characteristics vary with the species. For example, strains of the species A. caciae generally do not show haemolysis, but those of A. hydrophila and A. sobria do (Ljungh, 1987). Kn0ehel (1989) did, however, find that Aeromonas strains from low temperature sources ( < 15°C) produced lower titres of haemolysin at 37°C than strains from warm ( > 24°C) sources. A characteristic reported to correlate with mouse pathogenicity is autoagglutination, in this test Aeromonas strains arc incubated in brain heart infusion broth for 17-18 h at 35°C. At the end of this period spontaneously pelleted (SP +) cultures are noted. The cultures are then vortex mixed and divided in two. One aliquot is boiled for one hour and then examined for pelleting after boiling (PAB +) compared with the unboiled aliquot. The SP- PAB + phenotype was found to be more frequent in strains associated with invasive infections (Janda et al., 1987). Kn0chel (1989) found that only 2% of Aeromonas isolates from low temperature sources were PAB ÷. The work described here investigated the existence of two growth kinetic phenotypes and used two simple tests, haemolysis using a CAMP test and autoagglutination, to attempt to determine which group(s) is/are potentially pathogenic.

Materials and methods

Cultures Sixty-five Aeromonas strains, as described in the following, were used in this study. The type strain of the species A. hydrophila (ATCC 7966) was obtained from the New Zealand Communicable Disease Centre, Porirua, New Zealana, under accession number 804. This strain was originally isolated from tinned milk with a fishy odour.

133 Sixteen clinical strains of Aeromonas spp. (five A. hydrophila, six A. caciae and five A. sobria) were used. These were a gift from Diann¢ Fraser and Des Till of the New Zealand Communicable Disease Centre. Twenty-two strains from a variety of retail ready-to-eat fleshfoods, including meat and seafoods, (20 A. hydrophila and two A. sobria) were obtained by the method previously described (Hudson and DeLacy, 1991). In this method a 20 g sample of food is homogcnised in alkaline peptone water and incubated for 24 h at 28°C. A dilution series is then prepared and plated onto Starch Ampicillin Agar (Palumbo et al., 1985). Pure cultures arc obtained from these plates after 24 h incubation at 28°C. An additional test, el~stin hydrolysis, was used to aid in the distinction between A. hydrophila and A. cm'iae, since 99% of A. cm'iae, strains are negative for this test and 86% of A. hydrophila strains are positive (Lee, 1987). The method was as described by Hudson et al. (1986) except that nutrient agar was used as the basal medium. Samples regarded as being derived from meat processing came from two sources: (i) freshly slaughtered lamb carcasses sampled before chilling at a beef and lamb export meat plant and (ii) products taken from a pork processing plant at various stages in smallgoods production up to the point of cooking, In the case of sheep carcass samples a 10 × 10 cm square sterile gauze pad was moistened with peptone water and the whole carcass surface (including the inside of the body cavity) swabbed firmly. Swabs were transferred to 225 ml of alkaline peptone water and attached bacteria resuspended using a Colworth Stomacher (A.J. Seward and Co. Ltd., London, England) for 2 min. The suspensions were subsequently treated in the same manner as food samples. The samples taken from a pork processing plant were also treated as described for food samples. In total. 12 A. hydrophila, five A. sobria and nine A. caviae isolates were obtained from the combined sources of lamb carcass swabs and pork processing samples. Identification of motile aeromonads was carried out as described by Hudson and DeLacy (1991). Cultures were stored on Nutrient Agar slopes under refrigeration.

Measurement of growth curl'es For all curves, a i-ml volume of culture grown overnight at 28°C in Nutrient Broth (Difco), was added to 25 ml of the same medium in 100- or 150-ml conical flasks. These were incubated at the temperatures indicated. Samples were aseptically withdrawn after various periods of incubation and the optical density was measured at 650 nm. Before sampling the conical flasks were thoroughly shaken by hand to mix the cultures. Growth parameters were calculated by performing a linear regression on the straight line portion of the logarithmic growth curve to give the specific growth rate (p.). The intercept of this line with the initial optical density was determined to give the lag time. These parameters were obtained for all strains at both 4°C and 37°C. Experiments providing data for construction of curves of the square root of the growth rate versus temperature were all carried out in triplicate, and the averaged

13,1

ta were used for curve construction. Such curves, described by Ratkowsky ct al. ~.1982), show straight line portions from the optimum temperature for growth io lower temperatures, so allowing the prediction of a theoretical minimum growth temperature. In these experiments two organisms were used, the type strain of A. hydrophila and a food-derived isolate.

Indicators of pathogenicity All isolates were tested for their ability to pellet (autoagglutinate) during incubation (SP + phenotypc) or after boiling for one hour (PAB + phenotype) by the method of Janda et al. (1987). This phcnotypic feature is reported to be correlated to mouse pathogenicity. As an indicator of haemolytic ability CAMP tests were performed. Test plates comprised a 10 ml basal layer of Columbia Blood Agar (Difco) overlayed with 5 ml of this medium containing 5% defibrinated sheep blood. The CAMP reaction organisms Staphylococcus aareus and Rhodococcas equi were streaked parallel to one another and well separated. Test organisms werc streaked at right angles to the CAMP reaction organisms so that the streak lines were no closer than I - 2 ram. Three test organisms were streaked per plate and plates incubated aerobically at 37°C. Haemolysis has been reported to correlate with mouse pathogenicity (Burke et al., 1984), but species-specific CAMP reactions have not been shown (Ben-Ruwin et al., 1990).

Statistical Analysis Comparisons were made between species from any of the three sources and within species for subgroups from different sources. Data for growth kineti~ were analysed by computer using one-way analysis of variance. Strains scored as no growth (after 2(Xl h incubation), were omitted from the calculations. Data for pathogenicity indicators were analysed by computer using a ehi-square test with the expected values calculated from the incidence of the phenotype in question across either all of the strains tested or the appropriate species.

Results and Discussion

Initially analyses were carried out between the species without regard to their source, to discern inter-species differences. Analysis of variance showed that there were no statistical differences between species for either the generation or lag times at 4°C or for the lag times at 37°C. However, at 37°C a difference was found between strains of A. cat,iae and A. sobria for their generation times (P = 0.039). Data for these parameters are shown in Table 1. It should be noted that strains that did not grow at 4°C were omitted from these calculations. Analysis of the data for non-growing strains by the chi-square method showed that A. cal'iae strains deviated from the expected distribution to a significant extent (P = 0.~112), in that eight of the 15 strains tested were unable to grow at 4°(:.

TABLE I Mean growth kinetic data for Aermnonas spp. at 37 and 4°C Species

A. hydrophilu A. sobria A. cariae

n

38 12

7

37~C lag

gen

4°C lag

gen

(h)

(h)

(h)

(hi

q),.'~} 0.28 0.17

0.81]

27,41 26.31 23.19

27.84 28.57 26.32

I.I

0.59

gen., generation time. Slrains not growing at 4°C wcrc omincd from the calculations, For statistical analysis sec Results and Di~ussion. Aeromonas cariae deviated f r o m the e x p e c t e d values in the test for haemolysis o n s h e e p b l o o d a g a r ( p < 0.001) as only two o f 15 strains w e r e haemolytic (Table Ill, while the A. hydrophila strains w e r e significantly m o r e h a e m o l y t i c t h a n a r a n d o m distribution w o u l d predict ( P = 0.0454). This finding a g r e e s with t h a t of L j u n g h (1987). W i t h r e g a r d to o t h e r d a t a from the C A M P test, no significant differences w e r e f o u n d b e t w e e n the species with respect to r e a c t i o n s with S, aureus (Table il). However, r e a c t i o n s with R. equi d i f f e r e d b e t w e e n the species. O n l y o n e o f 15 strains o f A. cariae ( P < 0.001) showed e n h a n c e d haemolysis with R. equi (not a surprising result since this species is g e n e r a l l y non-haemolytic), b u t A. hydrophila was f r e q u e n t l y f o u n d to have a n e n h a n c e d haemolysis zone with R. equi ( P = 0.007). in r e g a r d to the a u t o a g g l u t i n a t i o n p r o p e r t i e s of the species, n o significant differences w e r e o b s e r v e d for precipitation a f t e r boiling, but A. sobria a u t o a g g l u t i n a t e d m o r e f r e q u e n t l y t h a n e x p e c t e d ( P < 0.001) a f t e r incubation, while A. hydrophila s h o w e d the o p p o s i t e t r e n d ( P = 0.0168). In this study a n u m b e r o f strains w e r e f o u n d to have a n SP + P A B - p h c n o t y p e ( d a t a not shown), a result not r e c o r d e d by J a n d a ct al. (1987). All o f these strains w e r e e i t h e r A. sobria o r A. caviae.

TABLE II Summary tff phenolypic characters of species of motile aen~mom,d Characteristic

Growth at 4°C Hacmolysis CAMP reaction with S. ram,us CAMP reaction with R. eqni "Precipitationbeforeboiling(SP') Precipitation after boiling(PAB + )

Numher (f.~) positive fi~reach characteristic A. h)~Irophila

A. ,~obria

A, cm'iue

( n = 38)

( n = 12)

( n = 15)

33 (87) 33 (87) 6 (16) 32 (84) 3 (8) 4(Ill

12 (Ilk'l) 8 (75) 2 (17) 8 (75) 8 (75) 2 (17)

7 (47) 2 (13) 5 (33) I (7) 5 (33) 3 (20)

For statistical analyses see Results and Discussion.

136 T A B L E Ill Mean growth kinelic data fi~r A. hydrophila from different sources Source Fk)od

370(` Generation time(h) Lag(h) 4~C

1.00 (I.41

Generation time (h) Lug(h)

19.98 14.81

Clinical

(I.611 (I.32 ND ND

Meat prtxdessing

11.52 0.09 38.28 41.52

ND, not determined. For statistical analyses see Results and Discussion.

In within-species analyses the various kinetic and phenotypie data were compared separately for all three species with respeet to the sources of the strains: food, clinical or meat processing. For A. hydrophila analysis of kinetic data for growth (omitting the type strain) (Table l i d showed a difference in the lag times at 37°C for strains of the species, with the food-derived isolates having significantly longer lag times than those from meat processing (P < 0.0(l i). Similarly, food-derived isolates had significantly longer generation times at 37°C, than isolates from either clinical sources (P = 0.03151 or meat processing ( P < (1.01}1). For the analysis of variance of kinetic values for organisms grown at 4°C, clinical strains were not included since only two were able to grow at this temperature. Food-derived strains had a shorter lag phase and grew at a faster rate (P = 0.(RII I) than those from meat processing (Table liD. Clinical isolates showed an inability to grow at 4°C, with three of the five strains unable to grow at this temperature. This inability to grow at 4°(:, was unlike other members of the species ( P < 0.001). No differences were noted for CAMP reactions between isolates from different sources (Table IV). Clinical isolates showed ,~ greater propensity for agglutination before (P=0.008) and after (P