- Email: [email protected]
Phagovar determination of Pseudomonas aeruginosa and a comparision of the results with mitomycin C induced pyocin production
Zbl. Bakt. Hyg., LAbt. Orig. A 249,225-234 (1981) Aus dem Institur fur Hygiene und Mikrobiologie der Universitat Wurzburg (Vorstand: Prof. Dr. H. P. R. Seeliger)
Phagovar Determination of Pseudomonas aeruginosa and a Comparision of the Results with Mitomycin C Induced Pyocin Production Phagovarbestimmung von Pseudomonas aeruginosa im Vergleich zur Mitomycin-C-induzierten Pyocinproduktion VERONICA STICHT-GROH
Received November 14, 1980
Summary One hundred Pseudomonas aeruginosa isolates recovered from intensive care areas of a university hospital and other clinical sites were characterized on the basis of their susceptibility to a set of twenty one phages. These strains were previously mitomycin C induced for their pyocin production and their serovars were also determined. All isolates identified on the basis of belonging to the same O-serovars and pyocin patterns, belonged to one phagovar group. Strains which were non-identifiable with commercial antisera, had to be typed by both phagovar assay and mitomycin C induced pyocin production. Neither phagovar assay nor pyocin production alone gave enough individual characteristics of anyone isolate, as to allow proper identification. Patterns with the same configurations in both phagovar and pyocin groups were detected among strains within entirely different O-serovar groups.
Zusammenfassung Hundert Pseudomonas-aeruginosa-isoiexe, gesammelt aus Bereichen der Intensivpflege und von anderen Krankenstationen, wurden auf der Grundlage ihrer Phagen-Empfindlichkeit bestimmt. In einer getrennten Studie waren die gleichen Sramrne zuvor bereits serologisch und hinsichtlich ihrer Pyocinproduktion untersucht worden. Isolate mit gleicher O-Serovar und gleichem Pyocinverhalten gehorten stets ein und derselben Phagovar an. Stamrne, die sich mit kauflichen Seren nicht identifizieren liefsen, mufsren sowohl der Phagentypisierung unterworfen, als auch im Hinblick auf ihre Mitomycin-C-induzierte Pyocinproduktion untersucht werden. Keine dieser beiden Methoden erlaubte fur sich allein eine ausreichende Charakterisierung der einzelnen Isolate. Vielmehr ergab sich, dag Stamme, die nur bei der Phagentypisierung oder nur in ihrer Pyocinproduktion iibereinsrimrnten, verschiedenen Serovaren angehorten.
226
V.Sticht-Groh
Introduction
It is known that especially in intensive care units (1), - neurosurgical, surgical and medical- Pseudomonas aeruginosa (P.a.) is a major source of superinfection in the critically ill. The importance of eradicating this organism from these settings, and of tracing them to the source of infection cannot be overemphasized. To achieve this, accurate, and standardized identifying methods are required. Presently, most clinical laboratories still recognize P. a. on the basis of its gross phenotypic characteristics, plus a few biochemical reactions and antibiotic susceptibility data. These characters are rarely specific enough to distinguish between individual isolates unless some unusual feature is present. Therefore, other, more reliable parameters should be looked for, such as pyocin production (2) or bacteriophage susceptibility (3). This latter for P.a. was first performed by Meitert in 1964 (4), followed by the different pyocin production methods developed by Farmer (5) and Gillies (6). None of these became universially accepted for routine laboratory use due to certain advantages and disadvantages in their methodology. Flagellar antigen determination although accurate and reliable (7) appears too cumbersome for every day use in its present form. In a previous study (8) mitomycin C induced pyocin production was tested in conjunction with an initial O-serovar determination of one hundred randomly selected P. isolates in a hospital laboratory setting. The originally described pyocin production method of Farmer (5) was simplified with the use of eighteen selected "indicator" strains. It was concluded that neither the use of commercially available antisera nor pyocin production alone were sufficient to identify anyone isolate. Mitomycin C induced pyocin production without the O-serovar determination proved faulty, as so called 'identical' pyocin 'types' occurred within the different serovars. In the present study phagovar ' determination was used in addition to O-serovar and pyocin determination on the above mentioned isolates. The obtained results were compared to mitomycin C induced pyocin production for their accuracy and routine applicability. Material and Methods In this study a bacteriophage set was used contairung twenty one phages. (Kindly supplied to us by the Cross-Infection Reference Laboratory, Colindale Avenue, London NW9 SHT, England). This kit contains the eighteen phages of Lindberg (9), plus three additional ones of Asheshov (1966). Phage assay test
The bacteriophages and their propagating strains were obtained in dried form. The dried suspensions of the propagating strains were subcultured on blood agar plates. The suspensions of the phages were diluted in nutrient broth (Oxoid) in tenfold dilutions till 10- 9 , and approx. 0.02 ml of each dilution was then applied to the surface of a nutrient agar plate. The plates were previously flooded with the propagating strains grown in a 1 This term is used throughout this presentationwith the recommendations put forward in the latest adition of the International Code of Nomenclature of Bacteria, Bacteriological Code, 1976 Revision.
Phagovar Determination of Pseudomonas aeruginosa
227
Table 1. Coding of Serogroups, Sites, Phagovar and Pyocin Patterns I) Numbering of the Serogroups was according to the schema devised by the Subcommittee on Pseudomonadaceae and Related Organisms of the International Committee of Systematic Bacteriology S.A. = self-agglutinable N. T. = not typable Code Groups
I) Phagovar Patterns 7,1214, M4 7, 21, F8, 109, 1214 7, 21, 24, 44, F8, 109, 1214, M4, M6 7,21,24,44, F8, 109,352, 1214,M4,colll 7, 21, 24,44, FlO, 1214, M4, col 11 7,21, 24, 73, 109, 119X, 1214, M4, M6 7,24,44. 73,F8, 109, 1214 7,24,44,68,109, 119X, 1214, M4, 188/1 7,24, M4, col 11 7,44,F8, 109, 1214 7,73, 109, 1214, M4, col 11 21,24 21,24.109, M4, col 11 21,24, 68,F8,Fl0, 109, 1214 21,31,68, F8, FlO, 109, 119X, 352 21,44,F8, 109, 1214 21, 119X, 1214 24,44, F8, 1214 24,F8, 109,352, 1214 31, 188/1 31,44, F8, 119X 31,44, F8, 109, 1214, 188/1 44, 109, 1214 F8, 109, 352, 1214 FlO, 109 No Susceptibility V) Pyocin Patterns
111111 111112 111114 111131 111134 111411 111414 111421 111424 121411 141111 211131 211231 331112 341132 371112
I II III IV V VI VII VIII IX
X XI XII XIII XIV XV XVI XVII XVIII XIX
XX XXI XXII XXIII XXIV XXV XXVI
Code Groups
Pyocin Patterns 374152 383848 411241 1/421434 542841 611111 611211 611231 621844 81412/55 83124/66 883586 886874 8/6/266 55/38 8/6/281668 not inducible
1 2 3
4 5 6 7
8 9
10 11 12 13 14 15 16
II) Sites Neurosurgical Intensive Care Unit (ICU) Surgical Intensive Care Unit (ICU) Medical Intensive Care Unit (ICU) Other hospital wards Outside hospitals Community practices
Code Groups 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Code A B C D E F
nutrient agar broth in the log phase of the organism. Incubation followed at 37 DC overnight. The next day the bacteriophage titer was calculated and the routine test dilution (RTD) determined. The RTD was defined as the highest dilution just failing to give a confluent lysis. All twenty one phages were propagated once, using the agar layer method described by Adams (10). All batches of the twenty one phages so produced were tested
XlII XIV XV XVI
XII
Vill IX X XI
IV V VI VII
III
II
Code Groups
30/2
29/ 1
Total
P hag o var
1
4
Serogroups
4
2
3/ 1
6/ 1
27/ 2
5
3
24/1 23/1
13/1
4/1
23/1
17
4
31/1
1
4.10
7/2 1/1 19/1
6
5
11/ 1
11/1
9/117/ 1
14/1 2/1
14/ 1
16/1 14/1 26/1
14/1
16/1
15/1
14 · / 1 · · 16 / 1
15
6
10/1
1
7
6/ 1 1/1
1/1
8/1
27/1
13
8
18/1
20/ 1
6
10 2
10.13
T able 2. Distribu tion of Serogroups, Phagovar versus Py ocin Patte rns o f the T otal Number of Pseudomonas strains tested
1
22/ 1
12
11
16 4
16
1/2
25/1
4
NT
14/1
4
SA
10
1
3
1
8
2
3
4
2
5
100
T ot al
::r-
0
if b ..,
;:;.
~
:<
00
N N
6/ 1
22/1
9/ 1
6/ 1
3 2/2
7
Legend : D ifferent Pyocin C ode Gro ups sec Ta ble 1. Number of isolates.
• = •• =
7/ 1
T ot al = 100
11
1 22/ 1
22/ 1
2/ 1
22/1
17/1
2
3
2
1
2
18
16/1
13/1
14/1
22/1 4 22/ 1
XXV
3/1
XXVI
3/3
28/ 1 6/1
21/1
6/1
8
23/3 22/ 1
7/1
7/ 1
XXIV
XXIII
24/3
24/3
XXII
XX I
24/1
1311
6/2
5/1
XX
XIX
XVIII
XVII
'";:j
::r~
'"" -e
\0
N N
~
'"
0
5"
(Jq
l::
rD
,..,
~
'"
~
;:l
0
3
0
l:: 0-
rD '"
0
;:l
0"
~
5·
3
~ ,..,
'"
t::l
~ ,..,
-c
0
(Jq
230
V.Sticht-Groh
to determine their RTD. These were also compared to that of the standard phage suspensions. The tests were carried out on the same day using the same batches of medium and broth culture. Provided that the newly prepared phage suspensions showed a similar lytic spectrum to that of the standard phage, the batch was considered acceptable. As a rule, the appearance of grades 3, 4 or 5 reactions where none should be, or a complete absence was an indication for the rejection of the freshly prepared batch. Variations of grades 4 to 5 or 3 to 4 etc. were permissible. Results were recorded as: 5, maximum (homologous strain) till 1, where only a weak lysis was present. In general P. a. phages propagate easily and relativelyhigh titers were produced. Batches with a titer of 1000 or less were rejected. The clinical isolates to be tested were grown overnight on nutrient agar plates passed the following day to fresh nutrient broth, and incubated at 37°C for 2 hrs in a waterbath. The surface of the medium (2 large Petri dishes for each strain) was flooded with the inoculated broth and the residual fluid immediately drawn off. The plates were allowed to dry and approx. 0.02 ml of each of the twenty one phages with their proper RTD adjusted applied to the surface of the agar. The plates were incubated overnight at 37°C and the following day the reactions were read and recorded. Coding of the phagovar groups The different bacteriophage susceptibility patterns referred to as phagovar patterns or phagovarieties were obtained when each of the one hundred P.isolates were assayed in the twenty one phages listed as follows: 7, 16, 21, 24, 31, 44, 68, 73, F7, F8, FlO, 109, 119X, 352, 1214, M4, M6, col 11, col 18, col 21, and 188/1. The different phagovarieties were divided into twenty six phagovar code groups. These groups were constructed with the following criterium in mind that two or more isolates belonged to the same code, if they did not differ in their lytic spectra in more than two major reactions. The serovar and pyocin codes were the same as reported in from a previous study (8) (Table 1).
Results Table 2 shows the different distribution patterns of the isolates from the intensive care areas and other clinical sites as expressed by serogroups, phagovars and pyocin patterns. The majority of the strains (551100) fell into five main phagovar code groups: XVII (18/100), XXVI (111100), XVI (10/100), XII (81100), XVII (8/100). The rest of the groups contained seven strains or less. Isolates (11/100) which gave no bacteriophage susceptibility pattern fitting with the above mentioned twenty one phages assayed, fell into code group: XXVI. Phagovar code group XVII contained the greatest number of isolates (Table 3), i.e, 18/100. Fourteen of these originated in one intensive care area (A), all having the same O-serovar (0:11), and pyocin code: 22 (pattern: 611111). All these isolates were lysed by phages 21, 119X and 1214 or a combination of these, but not varying in their lytic spectra in more than two major reactions. Six of the fourteen strains came from repeat specimens. All these strains - previously assumed to be the 'same' on the basis of their identical O-serovars and pyocin patterns-fell into one phagovar code group. Such 'parallels' were also found with other isolates with the same O-serovars, and pyocin patterns. However, in ten isolates (10/100), (five instances) (Table 4), all originating from intensive care areas, where an 'identity' among strains was assumed on the basis of the same O-serovars and pyocin patterns, discrepancies were noted within their phagolytic spectra. In two strains an additional observation was made: identical O-serovars and phagovarities, but different pyocin patterns.
Phagovar Determination of Pseudomonas aeruginosa
231
Table 3. List of the Main Phagovar Code Groups
Phagovar Code
Sites
Groups
Pyocin Code Group~
o groups
Total No .
Repeats (R)
in Phngovur Code Groups
XII XII XII XII XII XII XII XII
F F F
A A D D D
XVI XVI XVI XVI XVI XVI XVI XVI XVI XVI
A A
XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII
D 0 F B
XXIV XXIV XXIV XXIV XXIV XXIV XXIV XXIV
A A 0 0 0
F D F F B E 0 E
A A A A A A
A A A A A A A A
D B 0
3 23 9 17 11 10 8 22
0 :3 0 :4 0:5 0:6 0:6 0 :7 0:8 0:11
24 23 7 7 1 19 3 1 31
0: 4 0:4 0:5 0 :5 0:5 0:5 0 :6 0:8 0:8 0 :4.10
5 24 6 3 22 22 22 22 22 22 22 22 22 22 22 22 22 22
0:2 0 :4 0:8 0 :10 0 :11 0:11 0:11 0:11 0:11 0:11 0: 11 0:11 0:11 0 :11 0: 11 0:11 0:11 0 :11
23 23 23 22 3 3 3 7
0:4 0:4 0:4 0:4 0 :8 0:8 0:8 0 :10
11
8
none
10
none
R R
R R
R
R 18
6
R
8
1
232
V. Sticht-Groh
Table 4. Discrepancies NuteJ Among Phagovar Patterns and Mltomydn C fnduced Pyoc;n Production Determinations Sites
Pyocin 0Serovar Codes
Phagovar Codes
Pattern
Phagovar Pattern
Discrepancies either in Pyocin or Lysovar Patterns
A
0 :4
23
XVI
611211
21,44, F8, 109, 1214
21,44,611211
A
0:4
23
XXIV
611211
F8, 109, 352, 1214
352
A
0: 4
24
XVI
61123 1
21, 44, F8, 109, 1214
21,1214, 61123 1
A
0:4
24
XXI
61123 1
31,44,F8, 109
31
Pyocin
B
0:2
6
XXIII
111411
44,109,1214
B
0:2
6
XXVI
111411
no susceptibility
B
0 :6
11
XII
141111
21,24
24
B
0:6
11
XVI
141111
21, 44, F8, 109, 1214
44, F8, 109, 1214
B
0:6
14
X
331112
7,44, F8, 109, 1214
44,F8
B
0:6
14
XI
331112
7,73,109, 1214, 73, M4, col 11 M4, col 11
Eleven of the one h undr ed isolates assayed, as already stated, were not susceptible to the above mentioned twenty one phages tested . Two out of eleven (2/100) were
previously (8) identified as P.-species. This result is to campare with mitomyc in C induced pyocin production, where only the non -P.a. strains were non-inducible. However, additional eight isolates gave a poo rly defined pyocin patterns. Strains which were previo usly considered no n-" typable" with commercial antisera, but mitomycin C induci ble, were characterized with the additional information derived from phagovar determination. On the basis of the same phagova r and pyocin codes, 'identity' of the isolates was assumed. Phage 68 lysed a relatively large number of the isolates (15%), by means of a weak up to 2 + reaction. This result was not considered, being thought nonspecific in nature. Phages F7, M6, and col 11 lysed none of the one hundred isolates tested. Phages 16, and col 21 lysed one stra in each.
Discussion The different isolates defined on th e basis of their identical O-serovar and pyocin pattern also ha d the same phagovariety. With the determination of the O-serova r
Phagovar Determination of Pseudomonas aeruginosa
233
groups, the different lytic reactions either due to pyocins or phages, a total of fort y criteri a was established for each strain. If, therefore, the strains tested fell into one O-serov ar , pyocin and phagovar code th eir 'identity' could safely be assumed . Such observations should come as no suprise in isolat es derived from one outbreak.
However, no genetic correlation em be ascumed between pyocin producriofi Afid or ph agovar patt erns. These different characters of each strain depend solely on the fact, wh ich ' indicato r' strain or which ph ages were used to give the different lytic patterns. Without previous, i. e. pr imar y Ovscrovar determination either pyocin or phag ovar determinations alone will lead to erro neous results, as within entirely different 0 ant igenic gro ups the same p yocin or ph agovar patterns were noted. A combined determination of both pyocin and ph agovar code gro ups was essential to char acterize th e isolates when the O-serov ar agglutination was not possible (self-agglutinating stra ins). In these cases (Table 5), these additi onal characters were required to assume the 'sameness' among the isolates.
Table 5. Characterization of Isolates On the Basis of Pyoein and Phagovar Patterns Alone Sites
O-serovar
Pyocin Codes
Phagovar Codes
D (NeurologyInrensive) D (Neurology Intensive) D (Neurology Intensive)
0:8
O : Nd O:Nd
1 1 1
XIV XIV XIV
Nine out of the one hundred P.a. strains assayed (9%) gave no lytic activity with the twenty one ph ages assayed. Eight (81100) (8%) induced with mitomycin C gave po orly defined lytic zones. The lytic patterns due to pyocins were easier to read and evaluate than th ose given by ph ages. In five instances (10%) (Ta ble 4), a discrepancy was noted among patterns, where the 'sameness' of th e isolates was assumed by either one method plus scrova r typing. Therefore, choo sing a method for its routine applicability, the simpler, mor e easily performable mitomy cin C induct ion should be given preferenc e over phagovar typing . The two methods - mitomy cin C induction and phagovar typin g may be just as accurate ; both having a failur e quote of abou t 10% . The maintenance of the different phage suspensio ns, propagations and titrations, etc., requ ire time, material and a skilled staff. Furthermore, the lability in P.a. phage behaviour, the high innate lysogenicity occurring within this species (11), still cont ributes further to th e variability of th e results. On the other hand, ph agovar determinations, as already mentioned, proved essential in instances where the O-serovar determinat ion was unsuccessful. Consequently, phagovar identification shou ld be reserved for cases where additional characteristics are required to establish identity among the isolates. It is thu s recomm ended, that smaller hospital laboratories should carr y out a primary subdivision of all P. a. isolates recovered in an outbreak, on the basis of recognizing their different O-serovar. When a cluster of strains belonging to the same O-serova r is detected in one location, or when other reasons are to suspect
234
V.Sticht-Groh
hospital acquired infections due to P. a. attempts ought to be made to characterize the isolates by one step furth er. Regional or district laboratories should proceed to do this by subjecting the strains to mitomycin C induced pyocin production. At a reference laboratory level, isolates not sufficiently c aracterized either due to the lack of the O-serovar or pyocin determinations, should be subjected to phagovar assay. Such centers therefore, should have a complete set of both methods at hand. Finally, reference centres should untertake the standardization and improvement of already existing methods, and the development of more suitable ones for large scale epidemiological studies.
References 1. Praxis der Klinischen Hygiene in Anasthesie und Intensivpflege. Intensivmedizin, Notfallmedizin, Anasthesiologie Band 9. Georg Thieme Verlag, Stuttg art (1977) 2. Gov an, f. W.R.: Pyocin-ryping of Pseudomonas aeruginosa , in : Methods in Microbiology, vol. 10, p. 91. Academic Press, New York (1978) 3. Bergan, T.: Phage-t yping of Pseudom onas aerugin osa idem p. 61 4. M eitert, E.: Schema de lysorypie pour Pseudo m onas aeruginosa. Arch. roum . Path. expo 24 (1965) 439-458 5. Farm er Ill , j .]. and L. G. Herman: Epidemiological fingerprinting of Pseudomonas aeruginosa by production and sensitivity to pyocin and bacterioph age. Appl. Microb ial. 18 (1969) 760-7 65 6. Gillies, R. R. and j. W. R. Govan : T yping of Pseud om onas pyocyanea by pyocin pro ducti on. J. Path . Bact. 91 (1966) 339-3 45 7. Ansorg, R., W.Schmitt und V.Sch we rk : Antikorperbildung gegen somatische und flagellare Antigene von Pseudomonas aeruginosa, Med. Microbi ol. Immunol. 165 (1978) 181-1 89 8. Sticht-Groh, V .: Tracing of Pseudo mo nas aeruginos a Infection by the Use of Commercial Antisera and Pyocin Production and the Evaluation of the Results on the Basis of the ;(2 T est. Zbl. Bakt. H yg., I. Abr. Orig. A 244 (1979) 240-250 9. Lindberg, R. B., R. L. Latta, R. E. Brame, and f . A. M ontcrief : Defi.iit ive bacteriophage typing system for Pseud om onas aeruginosa. Bact. Pro c, 15 (1964) 81 10. Swanstrom, M. and M.H.Adams: Agar layer method for production of high titr e phage stocks. Proc. Soc. expo BioI. (N.Y.) 78 (1951) 372 11. Lindberg, R. B. and R. L. Latta: Phage Typing of Pseudom onas aeruginosa. J. infect. Dis. 130 (1974) 33-42 Dr. Veronica Sticht-Groh, Institut fiir Hygiene und Mikrobiologie, Josef-SchneiderStr. 2, Bau 17, D-7800 Wtirzburg
Phagovar Determination of Pseudomonas aeruginosa and a Comparision of the Results with Mitomycin C Induced Pyocin Production Phagovarbestimmung von Pseudomonas aeruginosa im Vergleich zur Mitomycin-C-induzierten Pyocinproduktion VERONICA STICHT-GROH
Received November 14, 1980
Summary One hundred Pseudomonas aeruginosa isolates recovered from intensive care areas of a university hospital and other clinical sites were characterized on the basis of their susceptibility to a set of twenty one phages. These strains were previously mitomycin C induced for their pyocin production and their serovars were also determined. All isolates identified on the basis of belonging to the same O-serovars and pyocin patterns, belonged to one phagovar group. Strains which were non-identifiable with commercial antisera, had to be typed by both phagovar assay and mitomycin C induced pyocin production. Neither phagovar assay nor pyocin production alone gave enough individual characteristics of anyone isolate, as to allow proper identification. Patterns with the same configurations in both phagovar and pyocin groups were detected among strains within entirely different O-serovar groups.
Zusammenfassung Hundert Pseudomonas-aeruginosa-isoiexe, gesammelt aus Bereichen der Intensivpflege und von anderen Krankenstationen, wurden auf der Grundlage ihrer Phagen-Empfindlichkeit bestimmt. In einer getrennten Studie waren die gleichen Sramrne zuvor bereits serologisch und hinsichtlich ihrer Pyocinproduktion untersucht worden. Isolate mit gleicher O-Serovar und gleichem Pyocinverhalten gehorten stets ein und derselben Phagovar an. Stamrne, die sich mit kauflichen Seren nicht identifizieren liefsen, mufsren sowohl der Phagentypisierung unterworfen, als auch im Hinblick auf ihre Mitomycin-C-induzierte Pyocinproduktion untersucht werden. Keine dieser beiden Methoden erlaubte fur sich allein eine ausreichende Charakterisierung der einzelnen Isolate. Vielmehr ergab sich, dag Stamme, die nur bei der Phagentypisierung oder nur in ihrer Pyocinproduktion iibereinsrimrnten, verschiedenen Serovaren angehorten.
226
V.Sticht-Groh
Introduction
It is known that especially in intensive care units (1), - neurosurgical, surgical and medical- Pseudomonas aeruginosa (P.a.) is a major source of superinfection in the critically ill. The importance of eradicating this organism from these settings, and of tracing them to the source of infection cannot be overemphasized. To achieve this, accurate, and standardized identifying methods are required. Presently, most clinical laboratories still recognize P. a. on the basis of its gross phenotypic characteristics, plus a few biochemical reactions and antibiotic susceptibility data. These characters are rarely specific enough to distinguish between individual isolates unless some unusual feature is present. Therefore, other, more reliable parameters should be looked for, such as pyocin production (2) or bacteriophage susceptibility (3). This latter for P.a. was first performed by Meitert in 1964 (4), followed by the different pyocin production methods developed by Farmer (5) and Gillies (6). None of these became universially accepted for routine laboratory use due to certain advantages and disadvantages in their methodology. Flagellar antigen determination although accurate and reliable (7) appears too cumbersome for every day use in its present form. In a previous study (8) mitomycin C induced pyocin production was tested in conjunction with an initial O-serovar determination of one hundred randomly selected P. isolates in a hospital laboratory setting. The originally described pyocin production method of Farmer (5) was simplified with the use of eighteen selected "indicator" strains. It was concluded that neither the use of commercially available antisera nor pyocin production alone were sufficient to identify anyone isolate. Mitomycin C induced pyocin production without the O-serovar determination proved faulty, as so called 'identical' pyocin 'types' occurred within the different serovars. In the present study phagovar ' determination was used in addition to O-serovar and pyocin determination on the above mentioned isolates. The obtained results were compared to mitomycin C induced pyocin production for their accuracy and routine applicability. Material and Methods In this study a bacteriophage set was used contairung twenty one phages. (Kindly supplied to us by the Cross-Infection Reference Laboratory, Colindale Avenue, London NW9 SHT, England). This kit contains the eighteen phages of Lindberg (9), plus three additional ones of Asheshov (1966). Phage assay test
The bacteriophages and their propagating strains were obtained in dried form. The dried suspensions of the propagating strains were subcultured on blood agar plates. The suspensions of the phages were diluted in nutrient broth (Oxoid) in tenfold dilutions till 10- 9 , and approx. 0.02 ml of each dilution was then applied to the surface of a nutrient agar plate. The plates were previously flooded with the propagating strains grown in a 1 This term is used throughout this presentationwith the recommendations put forward in the latest adition of the International Code of Nomenclature of Bacteria, Bacteriological Code, 1976 Revision.
Phagovar Determination of Pseudomonas aeruginosa
227
Table 1. Coding of Serogroups, Sites, Phagovar and Pyocin Patterns I) Numbering of the Serogroups was according to the schema devised by the Subcommittee on Pseudomonadaceae and Related Organisms of the International Committee of Systematic Bacteriology S.A. = self-agglutinable N. T. = not typable Code Groups
I) Phagovar Patterns 7,1214, M4 7, 21, F8, 109, 1214 7, 21, 24, 44, F8, 109, 1214, M4, M6 7,21,24,44, F8, 109,352, 1214,M4,colll 7, 21, 24,44, FlO, 1214, M4, col 11 7,21, 24, 73, 109, 119X, 1214, M4, M6 7,24,44. 73,F8, 109, 1214 7,24,44,68,109, 119X, 1214, M4, 188/1 7,24, M4, col 11 7,44,F8, 109, 1214 7,73, 109, 1214, M4, col 11 21,24 21,24.109, M4, col 11 21,24, 68,F8,Fl0, 109, 1214 21,31,68, F8, FlO, 109, 119X, 352 21,44,F8, 109, 1214 21, 119X, 1214 24,44, F8, 1214 24,F8, 109,352, 1214 31, 188/1 31,44, F8, 119X 31,44, F8, 109, 1214, 188/1 44, 109, 1214 F8, 109, 352, 1214 FlO, 109 No Susceptibility V) Pyocin Patterns
111111 111112 111114 111131 111134 111411 111414 111421 111424 121411 141111 211131 211231 331112 341132 371112
I II III IV V VI VII VIII IX
X XI XII XIII XIV XV XVI XVII XVIII XIX
XX XXI XXII XXIII XXIV XXV XXVI
Code Groups
Pyocin Patterns 374152 383848 411241 1/421434 542841 611111 611211 611231 621844 81412/55 83124/66 883586 886874 8/6/266 55/38 8/6/281668 not inducible
1 2 3
4 5 6 7
8 9
10 11 12 13 14 15 16
II) Sites Neurosurgical Intensive Care Unit (ICU) Surgical Intensive Care Unit (ICU) Medical Intensive Care Unit (ICU) Other hospital wards Outside hospitals Community practices
Code Groups 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Code A B C D E F
nutrient agar broth in the log phase of the organism. Incubation followed at 37 DC overnight. The next day the bacteriophage titer was calculated and the routine test dilution (RTD) determined. The RTD was defined as the highest dilution just failing to give a confluent lysis. All twenty one phages were propagated once, using the agar layer method described by Adams (10). All batches of the twenty one phages so produced were tested
XlII XIV XV XVI
XII
Vill IX X XI
IV V VI VII
III
II
Code Groups
30/2
29/ 1
Total
P hag o var
1
4
Serogroups
4
2
3/ 1
6/ 1
27/ 2
5
3
24/1 23/1
13/1
4/1
23/1
17
4
31/1
1
4.10
7/2 1/1 19/1
6
5
11/ 1
11/1
9/117/ 1
14/1 2/1
14/ 1
16/1 14/1 26/1
14/1
16/1
15/1
14 · / 1 · · 16 / 1
15
6
10/1
1
7
6/ 1 1/1
1/1
8/1
27/1
13
8
18/1
20/ 1
6
10 2
10.13
T able 2. Distribu tion of Serogroups, Phagovar versus Py ocin Patte rns o f the T otal Number of Pseudomonas strains tested
1
22/ 1
12
11
16 4
16
1/2
25/1
4
NT
14/1
4
SA
10
1
3
1
8
2
3
4
2
5
100
T ot al
::r-
0
if b ..,
;:;.
~
:<
00
N N
6/ 1
22/1
9/ 1
6/ 1
3 2/2
7
Legend : D ifferent Pyocin C ode Gro ups sec Ta ble 1. Number of isolates.
• = •• =
7/ 1
T ot al = 100
11
1 22/ 1
22/ 1
2/ 1
22/1
17/1
2
3
2
1
2
18
16/1
13/1
14/1
22/1 4 22/ 1
XXV
3/1
XXVI
3/3
28/ 1 6/1
21/1
6/1
8
23/3 22/ 1
7/1
7/ 1
XXIV
XXIII
24/3
24/3
XXII
XX I
24/1
1311
6/2
5/1
XX
XIX
XVIII
XVII
'";:j
::r~
'"" -e
\0
N N
~
'"
0
5"
(Jq
l::
rD
,..,
~
'"
~
;:l
0
3
0
l:: 0-
rD '"
0
;:l
0"
~
5·
3
~ ,..,
'"
t::l
~ ,..,
-c
0
(Jq
230
V.Sticht-Groh
to determine their RTD. These were also compared to that of the standard phage suspensions. The tests were carried out on the same day using the same batches of medium and broth culture. Provided that the newly prepared phage suspensions showed a similar lytic spectrum to that of the standard phage, the batch was considered acceptable. As a rule, the appearance of grades 3, 4 or 5 reactions where none should be, or a complete absence was an indication for the rejection of the freshly prepared batch. Variations of grades 4 to 5 or 3 to 4 etc. were permissible. Results were recorded as: 5, maximum (homologous strain) till 1, where only a weak lysis was present. In general P. a. phages propagate easily and relativelyhigh titers were produced. Batches with a titer of 1000 or less were rejected. The clinical isolates to be tested were grown overnight on nutrient agar plates passed the following day to fresh nutrient broth, and incubated at 37°C for 2 hrs in a waterbath. The surface of the medium (2 large Petri dishes for each strain) was flooded with the inoculated broth and the residual fluid immediately drawn off. The plates were allowed to dry and approx. 0.02 ml of each of the twenty one phages with their proper RTD adjusted applied to the surface of the agar. The plates were incubated overnight at 37°C and the following day the reactions were read and recorded. Coding of the phagovar groups The different bacteriophage susceptibility patterns referred to as phagovar patterns or phagovarieties were obtained when each of the one hundred P.isolates were assayed in the twenty one phages listed as follows: 7, 16, 21, 24, 31, 44, 68, 73, F7, F8, FlO, 109, 119X, 352, 1214, M4, M6, col 11, col 18, col 21, and 188/1. The different phagovarieties were divided into twenty six phagovar code groups. These groups were constructed with the following criterium in mind that two or more isolates belonged to the same code, if they did not differ in their lytic spectra in more than two major reactions. The serovar and pyocin codes were the same as reported in from a previous study (8) (Table 1).
Results Table 2 shows the different distribution patterns of the isolates from the intensive care areas and other clinical sites as expressed by serogroups, phagovars and pyocin patterns. The majority of the strains (551100) fell into five main phagovar code groups: XVII (18/100), XXVI (111100), XVI (10/100), XII (81100), XVII (8/100). The rest of the groups contained seven strains or less. Isolates (11/100) which gave no bacteriophage susceptibility pattern fitting with the above mentioned twenty one phages assayed, fell into code group: XXVI. Phagovar code group XVII contained the greatest number of isolates (Table 3), i.e, 18/100. Fourteen of these originated in one intensive care area (A), all having the same O-serovar (0:11), and pyocin code: 22 (pattern: 611111). All these isolates were lysed by phages 21, 119X and 1214 or a combination of these, but not varying in their lytic spectra in more than two major reactions. Six of the fourteen strains came from repeat specimens. All these strains - previously assumed to be the 'same' on the basis of their identical O-serovars and pyocin patterns-fell into one phagovar code group. Such 'parallels' were also found with other isolates with the same O-serovars, and pyocin patterns. However, in ten isolates (10/100), (five instances) (Table 4), all originating from intensive care areas, where an 'identity' among strains was assumed on the basis of the same O-serovars and pyocin patterns, discrepancies were noted within their phagolytic spectra. In two strains an additional observation was made: identical O-serovars and phagovarities, but different pyocin patterns.
Phagovar Determination of Pseudomonas aeruginosa
231
Table 3. List of the Main Phagovar Code Groups
Phagovar Code
Sites
Groups
Pyocin Code Group~
o groups
Total No .
Repeats (R)
in Phngovur Code Groups
XII XII XII XII XII XII XII XII
F F F
A A D D D
XVI XVI XVI XVI XVI XVI XVI XVI XVI XVI
A A
XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII XVII
D 0 F B
XXIV XXIV XXIV XXIV XXIV XXIV XXIV XXIV
A A 0 0 0
F D F F B E 0 E
A A A A A A
A A A A A A A A
D B 0
3 23 9 17 11 10 8 22
0 :3 0 :4 0:5 0:6 0:6 0 :7 0:8 0:11
24 23 7 7 1 19 3 1 31
0: 4 0:4 0:5 0 :5 0:5 0:5 0 :6 0:8 0:8 0 :4.10
5 24 6 3 22 22 22 22 22 22 22 22 22 22 22 22 22 22
0:2 0 :4 0:8 0 :10 0 :11 0:11 0:11 0:11 0:11 0:11 0: 11 0:11 0:11 0 :11 0: 11 0:11 0:11 0 :11
23 23 23 22 3 3 3 7
0:4 0:4 0:4 0:4 0 :8 0:8 0:8 0 :10
11
8
none
10
none
R R
R R
R
R 18
6
R
8
1
232
V. Sticht-Groh
Table 4. Discrepancies NuteJ Among Phagovar Patterns and Mltomydn C fnduced Pyoc;n Production Determinations Sites
Pyocin 0Serovar Codes
Phagovar Codes
Pattern
Phagovar Pattern
Discrepancies either in Pyocin or Lysovar Patterns
A
0 :4
23
XVI
611211
21,44, F8, 109, 1214
21,44,611211
A
0:4
23
XXIV
611211
F8, 109, 352, 1214
352
A
0: 4
24
XVI
61123 1
21, 44, F8, 109, 1214
21,1214, 61123 1
A
0:4
24
XXI
61123 1
31,44,F8, 109
31
Pyocin
B
0:2
6
XXIII
111411
44,109,1214
B
0:2
6
XXVI
111411
no susceptibility
B
0 :6
11
XII
141111
21,24
24
B
0:6
11
XVI
141111
21, 44, F8, 109, 1214
44, F8, 109, 1214
B
0:6
14
X
331112
7,44, F8, 109, 1214
44,F8
B
0:6
14
XI
331112
7,73,109, 1214, 73, M4, col 11 M4, col 11
Eleven of the one h undr ed isolates assayed, as already stated, were not susceptible to the above mentioned twenty one phages tested . Two out of eleven (2/100) were
previously (8) identified as P.-species. This result is to campare with mitomyc in C induced pyocin production, where only the non -P.a. strains were non-inducible. However, additional eight isolates gave a poo rly defined pyocin patterns. Strains which were previo usly considered no n-" typable" with commercial antisera, but mitomycin C induci ble, were characterized with the additional information derived from phagovar determination. On the basis of the same phagova r and pyocin codes, 'identity' of the isolates was assumed. Phage 68 lysed a relatively large number of the isolates (15%), by means of a weak up to 2 + reaction. This result was not considered, being thought nonspecific in nature. Phages F7, M6, and col 11 lysed none of the one hundred isolates tested. Phages 16, and col 21 lysed one stra in each.
Discussion The different isolates defined on th e basis of their identical O-serovar and pyocin pattern also ha d the same phagovariety. With the determination of the O-serova r
Phagovar Determination of Pseudomonas aeruginosa
233
groups, the different lytic reactions either due to pyocins or phages, a total of fort y criteri a was established for each strain. If, therefore, the strains tested fell into one O-serov ar , pyocin and phagovar code th eir 'identity' could safely be assumed . Such observations should come as no suprise in isolat es derived from one outbreak.
However, no genetic correlation em be ascumed between pyocin producriofi Afid or ph agovar patt erns. These different characters of each strain depend solely on the fact, wh ich ' indicato r' strain or which ph ages were used to give the different lytic patterns. Without previous, i. e. pr imar y Ovscrovar determination either pyocin or phag ovar determinations alone will lead to erro neous results, as within entirely different 0 ant igenic gro ups the same p yocin or ph agovar patterns were noted. A combined determination of both pyocin and ph agovar code gro ups was essential to char acterize th e isolates when the O-serov ar agglutination was not possible (self-agglutinating stra ins). In these cases (Table 5), these additi onal characters were required to assume the 'sameness' among the isolates.
Table 5. Characterization of Isolates On the Basis of Pyoein and Phagovar Patterns Alone Sites
O-serovar
Pyocin Codes
Phagovar Codes
D (NeurologyInrensive) D (Neurology Intensive) D (Neurology Intensive)
0:8
O : Nd O:Nd
1 1 1
XIV XIV XIV
Nine out of the one hundred P.a. strains assayed (9%) gave no lytic activity with the twenty one ph ages assayed. Eight (81100) (8%) induced with mitomycin C gave po orly defined lytic zones. The lytic patterns due to pyocins were easier to read and evaluate than th ose given by ph ages. In five instances (10%) (Ta ble 4), a discrepancy was noted among patterns, where the 'sameness' of th e isolates was assumed by either one method plus scrova r typing. Therefore, choo sing a method for its routine applicability, the simpler, mor e easily performable mitomy cin C induct ion should be given preferenc e over phagovar typing . The two methods - mitomy cin C induction and phagovar typin g may be just as accurate ; both having a failur e quote of abou t 10% . The maintenance of the different phage suspensio ns, propagations and titrations, etc., requ ire time, material and a skilled staff. Furthermore, the lability in P.a. phage behaviour, the high innate lysogenicity occurring within this species (11), still cont ributes further to th e variability of th e results. On the other hand, ph agovar determinations, as already mentioned, proved essential in instances where the O-serovar determinat ion was unsuccessful. Consequently, phagovar identification shou ld be reserved for cases where additional characteristics are required to establish identity among the isolates. It is thu s recomm ended, that smaller hospital laboratories should carr y out a primary subdivision of all P. a. isolates recovered in an outbreak, on the basis of recognizing their different O-serovar. When a cluster of strains belonging to the same O-serova r is detected in one location, or when other reasons are to suspect
234
V.Sticht-Groh
hospital acquired infections due to P. a. attempts ought to be made to characterize the isolates by one step furth er. Regional or district laboratories should proceed to do this by subjecting the strains to mitomycin C induced pyocin production. At a reference laboratory level, isolates not sufficiently c aracterized either due to the lack of the O-serovar or pyocin determinations, should be subjected to phagovar assay. Such centers therefore, should have a complete set of both methods at hand. Finally, reference centres should untertake the standardization and improvement of already existing methods, and the development of more suitable ones for large scale epidemiological studies.
References 1. Praxis der Klinischen Hygiene in Anasthesie und Intensivpflege. Intensivmedizin, Notfallmedizin, Anasthesiologie Band 9. Georg Thieme Verlag, Stuttg art (1977) 2. Gov an, f. W.R.: Pyocin-ryping of Pseudomonas aeruginosa , in : Methods in Microbiology, vol. 10, p. 91. Academic Press, New York (1978) 3. Bergan, T.: Phage-t yping of Pseudom onas aerugin osa idem p. 61 4. M eitert, E.: Schema de lysorypie pour Pseudo m onas aeruginosa. Arch. roum . Path. expo 24 (1965) 439-458 5. Farm er Ill , j .]. and L. G. Herman: Epidemiological fingerprinting of Pseudomonas aeruginosa by production and sensitivity to pyocin and bacterioph age. Appl. Microb ial. 18 (1969) 760-7 65 6. Gillies, R. R. and j. W. R. Govan : T yping of Pseud om onas pyocyanea by pyocin pro ducti on. J. Path . Bact. 91 (1966) 339-3 45 7. Ansorg, R., W.Schmitt und V.Sch we rk : Antikorperbildung gegen somatische und flagellare Antigene von Pseudomonas aeruginosa, Med. Microbi ol. Immunol. 165 (1978) 181-1 89 8. Sticht-Groh, V .: Tracing of Pseudo mo nas aeruginos a Infection by the Use of Commercial Antisera and Pyocin Production and the Evaluation of the Results on the Basis of the ;(2 T est. Zbl. Bakt. H yg., I. Abr. Orig. A 244 (1979) 240-250 9. Lindberg, R. B., R. L. Latta, R. E. Brame, and f . A. M ontcrief : Defi.iit ive bacteriophage typing system for Pseud om onas aeruginosa. Bact. Pro c, 15 (1964) 81 10. Swanstrom, M. and M.H.Adams: Agar layer method for production of high titr e phage stocks. Proc. Soc. expo BioI. (N.Y.) 78 (1951) 372 11. Lindberg, R. B. and R. L. Latta: Phage Typing of Pseudom onas aeruginosa. J. infect. Dis. 130 (1974) 33-42 Dr. Veronica Sticht-Groh, Institut fiir Hygiene und Mikrobiologie, Josef-SchneiderStr. 2, Bau 17, D-7800 Wtirzburg