Beruflich Dokumente
Kultur Dokumente
Elsevier
JFM 00012
The rapid automated conductance assay system has been used in experiments concerned with
estimating the bacteriological quality of fish in terms of total and selective counts. Results of the former
method were obtained in one-fifth of the time taken by conventional methods and the correlations
between the methods were good. Selective counts of Escherichia colt, coliforms, and vibrios were
completed in one-third of the normal time. The instruments were easy to use and the results easy to
interpret. The method holds out good promise for automating such routine fish bacteriology.
Introduction
Media
Nutrient broth (Oxoid CM 1), seawater broth (SWB) containing (g/l) beef extract
10, peptone 10, 750 ml sea water and 250 ml water, and the trimethylamine oxide
(TMAO) medium of Wood and Baird (1943) were used for total counts. MacConkey
broth (Oxoid) and modified Reasoner's medium (Silverman and Munoz. 1979)
containing (g/l) proteose peptone, 5; yeast extract, 3; lactose, 10; NaCI, 7.5: sodium
lauryl sulphate, 0.2; sodium deoxycholate, 0.1, pH 6.5; were used for counts of
coliform bacteria and E. cog respectively. Thiosulphate citrate bile salts broth
(TCBS, CM 333 recipe without agar, Oxoid Manual) was used for counts of vibrios.
When required the above media were solidified with agar (15 g/l).
Fish e x t r a c t s
Standard methods were used for the bacteriological examination of cod and
haddock obtained from commercial sources and stored in ice for various times.
Samples (usually 100 g) were homogenised in chilled 0.1% peptone-phosphate buffer
pH 7.0-7.2 (400 ml). Standard plate counts were determined on dilutions of the
homogenate using a modified Miles and Misra technique (Bousfield et al., 1973), the
plates were incubated at 2 0 ° C for 5 days. The selective counts were done similarly
on the appropriate medium with incubation at 3 7 ° C for 24 h.
Conductance measurements
The apparatus described by Richards et al. (1978) was used in the initial part of
this work. This consisted of modified glass test tubes bearing platinum electrodes
which were placed in a water bath controlled precisely to 20 °C for standard plate
counts or 3 7 ° C for selective counts and connected to the measuring and control
system. The data were presented in paper tape form to a computer, the results then
being obtainable in numerical or graphical form. In the main part of the study, a
commercial instrument ('Malthus') produced by Malthus Instruments Ltd., William
Clowes Street, Burslem, Stoke-on-Trent was used. It has been described by Baynes et
al. (1983) and Porter et al. (1983). Media and samples (up to 128 simultaneously)
were held in 12-ml glass ampoules, containing reusable platinum electrodes printed
on a ceramic base, which were placed in a water bath and connected to the control
box and a microcomputer. The data were available continuously in numerical form
and off line in graphical form. The microprocessor was programmed to display on a
video display unit when a given preset GB occurred with any particular sample. In
this case the computer display indicated when three consecutive changes of > 2/~S
had occurred. Readings were taken every 25 min, for 24 h.
129
Results
T o t a l counts
The relationship between the detection times and the plate counts for fish extracts
tested in nutrient broth, seawater broth and T M A O broth is shown in Fig. 2 and the
correlation coefficients together with their standard deviations are shown in Table I.
The T M A O medium gives the best correlation with counts, the least scatter about
the line of correlation and the shortest detection time. The detection times were
much shorter than the conventional incubation time of 5 days.
Selective counts
It was known from experience that the numbers of coliforms and vibrios would
be low or zero in the commercial fish used. Accordingly, the fish samples were
120 -
80
GB
40
-20 I I I
10 20 3O
Time (h)
Fig. 1. Typical plots of conductance changes with time for a fish sample assayed in triplicate in nutrient
broth at 20 o C.
130
8-
6--
o
o
4- °° o
0
0
2-
I I I I I
8 16 24 32 4O
Detection time (h)
7- b
4-
9
c~
0
3-- ~o Io
2--
1--
I I I I I I 1 °l°n~l"
4 12 20 28 36
Detection time (h)
7-
c
6-
5-
O
o 4-
O)
o
-- 3 -
2-
I i i , I I ?o%
4 12 20 2B 36
Detection time (In)
Fig. 2. Plots of detection times (h) and inoculum number ( N ) for assays in: (a) nutrient broth, (b)
seawater broth, (c) TMAO medium.
131
TABLE 1
C o r r e l a t i o n coefficient ( r ) b e t w e e n c o n d u c t a n c e d e t e c t i o n times f o u n d in 3 m e d i a a n d c o n v e n t i o n a l c o u n t
with residual s t a n d a r d d e v i a t i o n (S.D.) of detection time. a n d regression e q u a t i o n by analysis of variance.
spiked with mixtures of 10 recent isolates from fish grown in nutrient broth for 48 h
at 20°C, mixed, diluted, and 1 ml added to each of the three media. A typical
relationship between detection times and numbers for vibrios is shown in Fig. 3 and
the results are summarised in Table II. In the medium most appropriate for the
organisms, i.e. MacConkey's and modified Reasoner's for E. coil and coliforms and
TCBS broth for vibrios, there was a fast response and a good relationship between
numbers and detection times. Long detection times resulted from large numbers of
other organisms and were therefore not acceptable as a positive result. With the
tolerance levels as shown in the table such confusion did not occur. The relationships
appear to be independent of the total count of the sample (10°-105/g).
Discussion
I I 1 I I
2 4 6 8 10
Detection time (h)
T A B L E II
D e t e c t i o n time (h)
between GB and change in bacterial numbers they can be used to measure initial
numbers in a sample. The detection routine in effect measures the time taken for the
population of bacteria to adapt to the medium and to grow. Such results are
available long before those of conventional plate count methods in which the colony
forming unit (cfu) has to replicate until the colony is visible. Experience in our own
and other laboratories indicates that the correlation coefficient between detection
times and counts rarely exceeds 0.9. One reason for this is that each cell in the
inoculum whether present in the free state or as part of a clump contributes to the
overall change, but both single cells and clumps gi+e rise to single cfus in the plate
count. Also, injured cells recover better in liquid than on solid media. The conduc-
tance measurements are made with a precision of 1 part in 105 but it is well known
that the error in routine bacteriological counts is high (see Hall, 1982; Postgate,
1969). It is difficult, as Richards et al. (1978) pointed out, to calibrate a precise
method against an inaccurate method however well established. Thus, higher correla-
tion coefficients would not necessarily be expected.
It is assumed that conductance assays depend upon the increased number of
charged ions resulting from the metabolism of the medium by the microbes in the
inoculum. Accordingly, the choice of culture media even for total counts can
influence the results. It has been shown recently (Easter et al., 1983) that some fish
spoilage bacteria can use T M A O as an electron acceptor during microaerophilic
growth. As T M A O is a neutral uncharged molecule, and the product of its reduction
trimethylamine is basic and highly charged, it is an excellent choice of substrate in a
medium for fish spoilage as it is crucial to the metabolism of the culture and yields a
good conductance change. The microaerobic growth conditions in a conductance
assay system may be more relevant to the changes in quality of fish as they are more
akin to the redox conditions existing during spoilage. The results may therefore
correlate more highly with other spoilage parameters such as taste panel score and
chemical indices than the routine conventional bacteriological count. It is clear,
however, from the data presented that the results for total counts are available far
133
References
Baynes, N.C., J. Comrie and J.H. Prain, 1983. Detection of bacterial growth by the Malthus conductance
meter. Med. Lab. Sci. 40, 149-158.
Bousfield, l.J., G.L. Smith and R.W, Trueman, 1973. The use of semi-automatic pipettes in the viable
counting of bacteria. J. Appl. Bacteriol. 36, 297-299.
Easter, M.C., D.M. Gibson and F.B. Ward, 1983. A conductance method for the assay and study of
bacterial trimethylamine oxide reduction. J. Appl. Bacteriol. 52, 357-365.
Hall, L.P., 1982. A manual of methods for the bacteriological examination of frozen foods. 3rd edn.,
Campden Food Preservation Research Association, Chipping Campden.
Jason, A.C., 1983. A deterministic model for monophasic growth of batch cultures of bacteria. Antonie
van Leeuwenhoek. 49, 513-516.
Porter, I.A., T.M.S. Reid, W.J. Wood, D.M. Gibson and G. Hobbs, 1983. Conductance measurements for
determining antibiotic sensitivity. In: Antibiotics: assessment of antimicrobial activity and resistance,
edited by A.D. Russell and L.B. Quensel, Academic Press, London, pp. 49-60.
134
Postgate, J.R., 1969. Viable counts and viability. In: Methods in Microbiology, Vol. 1, edited by J.R.
Norris and D.W. Ribbons, Academic Press, London,pp. 611-628.
Richards, J.C.S,, A.C. Jason, G. Hobbs. D.M, Gibson and R.H. Christie. 1978. Electronic measurement of
bacterial growth. J. Phys. E: Sci. lnstrum. 11, 560-568.
Silverman, M.P, and E.F. Munoz, 1979. Automated electrical impedance technique for rapid enumeration
of faecal coliforms in effluents from sewage treatment plants. Appl. Envir. Microbiol. 37. 521-526.
Wood. A.J. and E.A. Baird, 1943. Reduction of trimethylamine oxide by bacteria. I. The Entero-
bacteriaceae. J. Fish. Res. Bd. Can. 6, 194-201.