Laboratory Animals (1976) 10, 413-417.
413
DIENES TYPING OF PROTEUS STRAINS ISOLATED FROM BARRIER-MAINTAlNED
ANIMALS
by W. SIMPSON
and D. J. C. SIMMONS
Institute of Cancel' Research, Pol/ards Wood Research Station, Nigl1tingales Lane, Chalfont St Giles, Buckinghamshire,
HP84SP
SUMMARY
The possible use of Dienes' phenomenon for typing proteus strains as an aid to bacteriological monitoring of barrier-maintained animal units was investigated. It was rare for more than I Dienes' type to be isolated from an individual animal. A persistent relationship between I or 2 Diene's types and each strain of animal was demonstrated, although these same types were shown to be capable of causing clinical disease in their normal host. The findings suggest that Dienes' phenomenon may be of value in detecting outside contamination of barrier-maintained animals and also might usefully be applied to experimental epidemiological studies. Dienes' phenomenon occurs when 2 different types of swarming Proteus strains are inoculated at different points on the same agar plate. The 2 swarms never quite coalesce (Dienes, 1946), there being a degree of mutual inhibition between the strains. The swarms do however coalesce if the Proteus strains are of the same type. Dienes' phenomenon therefore forms the basis of a highly specific form of typing (Story, 1954). In epidemiologically unrelated strains coalescence occurs in 1 in 70 pairs tested (Skirrow, ]969). The technique is valuable when small numbers of isolates from a single outbreak of infection are being compared (Wilson & Miles, ]975). In this laboratory we are required to investigate the microbial flora of laboratory animals housed in barrier-maintained units, and any cross-infection and disease problems that may arise. Dienes' technique was considered as a possible aid to the fulfilment of this function. This entailed determining the pattern of distribution of proteus types isolated from various strains and species of animals bred at the Institute of Cancer Research. The technique has also been applied to isolates from sick animals, and to experimental epidemiological studies with germ-free mice. MATERTALS
AND METHODS
Animals The Proteus strains were isolated from the caeca of animals undergoing screening for pathogenic organisms. Specimens from 3 strains of rats, 5
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
414
W. SIMPSON
AND
D. J. C. SIMMONS
strains of mice and 1 strain of guinea-pig were cultured. These animals were obtained from 2 separate breeding units which are barrier maintained. There is interchange of staff between breeding rooms within each unit, but not between the units. A small survey of the Lister hooded inbred rat colony was undertaken to determine how many Proteus types were present in a single animal. Media The media used for isolation of Proteus was blood agar and selenite F enrichment broth. A streak from the swarm on a blood agar plate was inoculated on to a MacConkey agar plate. After 18 h incubation at 37°C single colonies were each seeded on to the centre of a diagnostic sensitivity test CDST) agar plate (Oxoid Ltd, Wade Road, Basingstoke, Hampshire, RG24 OPW). Individual colonies obtained by subculture from seleniJe F broth on to desoxycholate citrate agar plates or MacConkey agar plates were dealt with similarly. Specimens from the Lister hooded inbred rat survey were inoculated directly onto MacConkey agar plates which were incubated at 37°C for 18 h. 4 single Proteus colonies from each plate were then subcultured onto DST agar plates to obtain pure cultures. Dienes' typing was carried out on DST agar plates, which were freshly made and dried at 3rC for 90 min. When surface water was present on the plates aberrant results were obtained. In order to obtain clear lines of mutual inhibition a whole plate was used for each pair, in preference to the multiple inoculation method recommended by Story (1954) which frequently gave dubious results. Reference Types 4 different Proteus strains were selected as reference types from the first 15 isolates tested. All further isolates were tested against these strains. Length of survey The Proteus isolates typed were obtained over an 18 month period and, with the exception of the nude (nu Ilu) mice and guinea-pigs, not less than 50 animals of each strain were tested. Biochemistry All proteus isolates were biochemically typed (20E system; A.P.r. Laboratory Products Ltd, Philpot House, Rayleigh, Essex) in conjunction with the profile register. RESULTS
Number of Proteus types isolated The rate of Proteus isolations varied from 10-] 5 % of animals tested in the mouse strains to 95-100 % in the rat strains. 4 different Dienes' types were
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
DIENES
TYPING
OF PROTEUS
415
isolated initially and were designated CBT/1 to CBT/4. After testing further isolates against the reference strains of these types, 4 new types were found and numbered CBT/5 to CBT/8. All the Proteus strains isolated fell into 2 biochemical types, only CBT/4 was Proteus vulgaris, the rest being Proteus mirabilis. CBT/1 to CBTj4 comprised the majority of isolates. CBT/5 to CBT/8 were isolated infrequently. The few P. vulgaris isolates from rats and mice did not swarm. Table 1 shows the distribution of Dienes' types in the various animal strains. Table 2 summarizes the survey in Lister hooded inbred rats during which 2 Proteus types (CBT /1 and CBT /2) were isolated. Every animal except I was monocontaminated. of Proteus Dienes types in the animal strains Cancer Research.
Table 1. Distribution
+ indicates
at the Institute
of
isolation of Proteus type.
Mice
Rats Lister August Wistar hooded inbred inbred inbred CBT/]
2 3 4 5 6 7 8
+ +
+ +
+
CBA/ CBA/ C57/ DBA/2 Nudes GuineaCa T6T6 BL inbred (nu nu) pigs inbred inbred inbred
+ +
+ +
+ + +
+ +
+
+
+
+ +
Table 2. Numbers and types of Proteus detected in a survey of the Lister hooded inbred colony.
Rats
No. caeca cultured
No. proteus strains isolated
No. CBT/! only
No. CBT/2 only
No. CBT/! & CBT/2
Babies* Adultst
41 42
28 33
12
1
43
9
0
Total
95
83
6]
2]
1
*8-20 days old.
52
17-8 months old.
Isolates from sick animals P. mirabilis was isolated in pure culture from head and neck abscesses of nude mice (nu nu). The type, CBT/3, was the same as that isolated from their caeca. It was also isolated from the caeca of healthy mice in the same group.
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
416
W. SIMPSON
AND
D. J. C. SIMMONS
P. mirabilis, CBTj2, was isolated in pure culture from the urine of a female August inbred rat with a urinary tract infection. This type was isolated from the rat's caecum and is a type normally associated with this strain of rat. A preputial gland abscess found in a CBAjCa mouse was cultured. P. mirabilis only was isolated and Dienes' typed as CBT!3. This type was isolated from the mouse's caecum and is a type found normally in our CBAjCa breeding colonies. Germ-free mice In an experiment which required the 'conventionalisation' of germ-free mice by placing them in a room with conventional mice, it was found that the germ-free mice acquired Proteus rapidly. The types isolated, CBTj2 and CBT!3 were the same as those present in the conventional
mice.
DISCUSSION
Taking account of the limited nature of this investigation into the application of Dienes' typing to barrier-maintained animal colonies, 2 findings of interest have emerged. Firstly, Proteus types associated with a particular strain or species of animal remained constant over a long period, irrespective of both the introduction of new or the interchange of existing staff within breeding rooms. Environmental changes also had no effect. Secondly, it is unusual to isolate more than 1 Proteus type from an individual animal. Story (1954) and de Louvois (1969) demonstrated a high incidence of autoinfection in patients with urinary tract infections. They isolated the same Dienes' type from the faeces and urine of affected patients. Different types were usually found in the faeces of other patients in the same hospital ward. We have shown that the Proteus type isolated from a clinical infection in an animal probably originated from the same animal. However, in view of the persistent association between particular Proteus types and epidemiologically related animals, the Proteus isolated from the clinical infection could have arisen as a result of cross-infection from another animal in the same colony. The persistent association of 1 or 2 Dienes' types with an animal colony suggests that this method of Proteus typing might be a valuable aid in detecting outside contamination of an animal unit or of cross-infection within the unit. However a new Proteus type might have difficulty in establishing itself because of the stable relationship between Proteus type and animal strain. An investigation of the source of Proteus types that contaminate the intestines of young animals, and the continued presence of these Proteus types into adult life is being carried out in a rat colony contaminated with 2 Proteus types. ACKNOWLEDGEMENTS
This work was supported
by grants to the Chester Beatty Research Institute
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
DIENES TYPING OF PROTEUS (I nstitute of Cancer Research: Royal Cancer Hospital) Research Council and the Cancer Research Campaign.
417 from
the Medical
REFERENCES de Louvois, J. (1969). Serotyping and the Dienes reaction on Protells mirabilis from hospital infections. JOllrnal of Clinical Pathology 22, 263-268. Dienes, L. (1946). Reproductive processes in Proteus cultures. Proceedings of the Society of Experimental Biology and Medicine 63, 265-276. Skirrow, M. B. (1969). The Dienes (mutual inhibition) tt'st in the investigation of proteus infections. JOllrnal of Medical Microbiology 2, 471-477. Story, P. (1954). Proteus infections in hospital. Journal of Pathology and Bacteriology 68, 55-62. Wilson, G. S. & Miles, A. (1975). Topley and Wilson's principles of bacteriology, virology and immunity, 6th ed, chap. 27. London: Arnold.
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
413
DIENES TYPING OF PROTEUS STRAINS ISOLATED FROM BARRIER-MAINTAlNED
ANIMALS
by W. SIMPSON
and D. J. C. SIMMONS
Institute of Cancel' Research, Pol/ards Wood Research Station, Nigl1tingales Lane, Chalfont St Giles, Buckinghamshire,
HP84SP
SUMMARY
The possible use of Dienes' phenomenon for typing proteus strains as an aid to bacteriological monitoring of barrier-maintained animal units was investigated. It was rare for more than I Dienes' type to be isolated from an individual animal. A persistent relationship between I or 2 Diene's types and each strain of animal was demonstrated, although these same types were shown to be capable of causing clinical disease in their normal host. The findings suggest that Dienes' phenomenon may be of value in detecting outside contamination of barrier-maintained animals and also might usefully be applied to experimental epidemiological studies. Dienes' phenomenon occurs when 2 different types of swarming Proteus strains are inoculated at different points on the same agar plate. The 2 swarms never quite coalesce (Dienes, 1946), there being a degree of mutual inhibition between the strains. The swarms do however coalesce if the Proteus strains are of the same type. Dienes' phenomenon therefore forms the basis of a highly specific form of typing (Story, 1954). In epidemiologically unrelated strains coalescence occurs in 1 in 70 pairs tested (Skirrow, ]969). The technique is valuable when small numbers of isolates from a single outbreak of infection are being compared (Wilson & Miles, ]975). In this laboratory we are required to investigate the microbial flora of laboratory animals housed in barrier-maintained units, and any cross-infection and disease problems that may arise. Dienes' technique was considered as a possible aid to the fulfilment of this function. This entailed determining the pattern of distribution of proteus types isolated from various strains and species of animals bred at the Institute of Cancer Research. The technique has also been applied to isolates from sick animals, and to experimental epidemiological studies with germ-free mice. MATERTALS
AND METHODS
Animals The Proteus strains were isolated from the caeca of animals undergoing screening for pathogenic organisms. Specimens from 3 strains of rats, 5
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
414
W. SIMPSON
AND
D. J. C. SIMMONS
strains of mice and 1 strain of guinea-pig were cultured. These animals were obtained from 2 separate breeding units which are barrier maintained. There is interchange of staff between breeding rooms within each unit, but not between the units. A small survey of the Lister hooded inbred rat colony was undertaken to determine how many Proteus types were present in a single animal. Media The media used for isolation of Proteus was blood agar and selenite F enrichment broth. A streak from the swarm on a blood agar plate was inoculated on to a MacConkey agar plate. After 18 h incubation at 37°C single colonies were each seeded on to the centre of a diagnostic sensitivity test CDST) agar plate (Oxoid Ltd, Wade Road, Basingstoke, Hampshire, RG24 OPW). Individual colonies obtained by subculture from seleniJe F broth on to desoxycholate citrate agar plates or MacConkey agar plates were dealt with similarly. Specimens from the Lister hooded inbred rat survey were inoculated directly onto MacConkey agar plates which were incubated at 37°C for 18 h. 4 single Proteus colonies from each plate were then subcultured onto DST agar plates to obtain pure cultures. Dienes' typing was carried out on DST agar plates, which were freshly made and dried at 3rC for 90 min. When surface water was present on the plates aberrant results were obtained. In order to obtain clear lines of mutual inhibition a whole plate was used for each pair, in preference to the multiple inoculation method recommended by Story (1954) which frequently gave dubious results. Reference Types 4 different Proteus strains were selected as reference types from the first 15 isolates tested. All further isolates were tested against these strains. Length of survey The Proteus isolates typed were obtained over an 18 month period and, with the exception of the nude (nu Ilu) mice and guinea-pigs, not less than 50 animals of each strain were tested. Biochemistry All proteus isolates were biochemically typed (20E system; A.P.r. Laboratory Products Ltd, Philpot House, Rayleigh, Essex) in conjunction with the profile register. RESULTS
Number of Proteus types isolated The rate of Proteus isolations varied from 10-] 5 % of animals tested in the mouse strains to 95-100 % in the rat strains. 4 different Dienes' types were
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
DIENES
TYPING
OF PROTEUS
415
isolated initially and were designated CBT/1 to CBT/4. After testing further isolates against the reference strains of these types, 4 new types were found and numbered CBT/5 to CBT/8. All the Proteus strains isolated fell into 2 biochemical types, only CBT/4 was Proteus vulgaris, the rest being Proteus mirabilis. CBT/1 to CBTj4 comprised the majority of isolates. CBT/5 to CBT/8 were isolated infrequently. The few P. vulgaris isolates from rats and mice did not swarm. Table 1 shows the distribution of Dienes' types in the various animal strains. Table 2 summarizes the survey in Lister hooded inbred rats during which 2 Proteus types (CBT /1 and CBT /2) were isolated. Every animal except I was monocontaminated. of Proteus Dienes types in the animal strains Cancer Research.
Table 1. Distribution
+ indicates
at the Institute
of
isolation of Proteus type.
Mice
Rats Lister August Wistar hooded inbred inbred inbred CBT/]
2 3 4 5 6 7 8
+ +
+ +
+
CBA/ CBA/ C57/ DBA/2 Nudes GuineaCa T6T6 BL inbred (nu nu) pigs inbred inbred inbred
+ +
+ +
+ + +
+ +
+
+
+
+ +
Table 2. Numbers and types of Proteus detected in a survey of the Lister hooded inbred colony.
Rats
No. caeca cultured
No. proteus strains isolated
No. CBT/! only
No. CBT/2 only
No. CBT/! & CBT/2
Babies* Adultst
41 42
28 33
12
1
43
9
0
Total
95
83
6]
2]
1
*8-20 days old.
52
17-8 months old.
Isolates from sick animals P. mirabilis was isolated in pure culture from head and neck abscesses of nude mice (nu nu). The type, CBT/3, was the same as that isolated from their caeca. It was also isolated from the caeca of healthy mice in the same group.
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
416
W. SIMPSON
AND
D. J. C. SIMMONS
P. mirabilis, CBTj2, was isolated in pure culture from the urine of a female August inbred rat with a urinary tract infection. This type was isolated from the rat's caecum and is a type normally associated with this strain of rat. A preputial gland abscess found in a CBAjCa mouse was cultured. P. mirabilis only was isolated and Dienes' typed as CBT!3. This type was isolated from the mouse's caecum and is a type found normally in our CBAjCa breeding colonies. Germ-free mice In an experiment which required the 'conventionalisation' of germ-free mice by placing them in a room with conventional mice, it was found that the germ-free mice acquired Proteus rapidly. The types isolated, CBTj2 and CBT!3 were the same as those present in the conventional
mice.
DISCUSSION
Taking account of the limited nature of this investigation into the application of Dienes' typing to barrier-maintained animal colonies, 2 findings of interest have emerged. Firstly, Proteus types associated with a particular strain or species of animal remained constant over a long period, irrespective of both the introduction of new or the interchange of existing staff within breeding rooms. Environmental changes also had no effect. Secondly, it is unusual to isolate more than 1 Proteus type from an individual animal. Story (1954) and de Louvois (1969) demonstrated a high incidence of autoinfection in patients with urinary tract infections. They isolated the same Dienes' type from the faeces and urine of affected patients. Different types were usually found in the faeces of other patients in the same hospital ward. We have shown that the Proteus type isolated from a clinical infection in an animal probably originated from the same animal. However, in view of the persistent association between particular Proteus types and epidemiologically related animals, the Proteus isolated from the clinical infection could have arisen as a result of cross-infection from another animal in the same colony. The persistent association of 1 or 2 Dienes' types with an animal colony suggests that this method of Proteus typing might be a valuable aid in detecting outside contamination of an animal unit or of cross-infection within the unit. However a new Proteus type might have difficulty in establishing itself because of the stable relationship between Proteus type and animal strain. An investigation of the source of Proteus types that contaminate the intestines of young animals, and the continued presence of these Proteus types into adult life is being carried out in a rat colony contaminated with 2 Proteus types. ACKNOWLEDGEMENTS
This work was supported
by grants to the Chester Beatty Research Institute
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
DIENES TYPING OF PROTEUS (I nstitute of Cancer Research: Royal Cancer Hospital) Research Council and the Cancer Research Campaign.
417 from
the Medical
REFERENCES de Louvois, J. (1969). Serotyping and the Dienes reaction on Protells mirabilis from hospital infections. JOllrnal of Clinical Pathology 22, 263-268. Dienes, L. (1946). Reproductive processes in Proteus cultures. Proceedings of the Society of Experimental Biology and Medicine 63, 265-276. Skirrow, M. B. (1969). The Dienes (mutual inhibition) tt'st in the investigation of proteus infections. JOllrnal of Medical Microbiology 2, 471-477. Story, P. (1954). Proteus infections in hospital. Journal of Pathology and Bacteriology 68, 55-62. Wilson, G. S. & Miles, A. (1975). Topley and Wilson's principles of bacteriology, virology and immunity, 6th ed, chap. 27. London: Arnold.
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
Downloaded from lan.sagepub.com at UNIV OF GEORGIA on May 30, 2015
