13 February 1974 S.A. MEDICAL JOURNAL 257 (Supplement-South African Journal of Laboratory and Clinical Medicine) LCM 9 The Enterotube System in the Identification of Enterohacteriaceae and Yersinia

M. H. FINLAYSON, B. GIBBS

SUMMARY Dulcitol agar for detection of acid production, absent in sp., most Enterobacters sp., Hafnia and Serratia, A trial of the Enterotube system for the identification of A rizona and Edwardsiella. and a comparison with the methods Urea agar for detection of urea cleavage, absent in at present in use in the Tygerberg Hospital Microbiology Escherichia, Shigella, Providencia, Enterobacter aerogenes, Laboratory, were carried out. One hunded cultures be­ Hafnia, Salmonella, A rizona and Edwardsiella. longing to the family Enterobacteriaceae including Phenylalanine agar for detection of phenylalanine , , Proteus vulgaris, deaminase by production of a green colouration after Proteus morgani, Proteus rettgeri, Providencia, Edward­ addition of ferric chloride solution. This reaction is posi­ siella, Shigella, Klebsiella, Enterobacter, Serratis, Sal­ tive only in the Proteus and Pro\'idencia groups. monella, Citrobacter and Arizona were tested with con­ ventional procedures and also with the Enterotube. The Hydrogen sulphide and indole agar for detection of conventional procedures, designed to identify the hydrogen sulphide and indole, the latter being detected organisms within 24 hours after isolation, were used. Three after the addition of Kovac's reagent to the compartment. of the cultures examined were not identified by the Hydrogen sulphide is produced by some Proteus strains Enterotube technique when their identity was established and by Salmonella, Citrobacter, Arizona and Edwardsie/la by routine conventional methods. These were non­ while indole is produced by Escherichia, .some Proteus -producing non-motile organisms which may have strains, Providencia and Edwardsiella. been members of the tribe Klebsiellae. Good agreement Dextrose agar showing acid production in all the was therefore observed in 97% of the cultures when the Enterobacteriaceae and also in Yersinia. It is claimed that Enterotube was used compared with the conventional the Enterotube provides a 'ready to use' system for the technique. Three cultures of and screening and differentiation of Enterobacteriaceae. It two of Yersinia pseudotuberculosis were identified by the avoids the preparation and sterilisation of media and Enterotube was used, compared with the conventional permits the simultaneous inoculation of the medium in the levels. 8 compartments with little risk of contamination. A redesigned Enterotube is now available. In this tube, S. Afr. Med. J., 48, 257 (1974). the lactose has been removed from the lysine decarboxy­ lase compartment, a test for ornithine decarboxylase has MATERIALS AND METHODS been added to one compartment, and the tests for pheny­ lalanine deaminase and dulcitol fermentation have been The original Enterotube system incorporates 9 biochemical combined in another compartment. An iron salt has been tests and comprises a moulded semicircular plastic tube added to this compartment, thus eliminating the need to divided into 8 compartments, each of which contains a add ferric chloride to the phenylalanine agar. The com­ slope of one of the following test media: partments containing tests for dextrose fermentation and the 2 carboxylase tests have been covered with sterile Citrate agar for detection of citrate utilisation; this wax, allowing for detection of gas from dextrose and would, if negative, exclude Escherichia, Shigella and improving the carboxylase reactions. In our opinion these Edwardsiella. changes greatly enhance the value of the Enterotube Lysine agar to detect lysine decarboxylase produced system, especially in the separation of Shigella strains, S. early by Hafnia and Serratia and possibly late by Kleb­ typhi and P. rettgeri, and also the more positive identifica­ siella and Enterobacter aerogenes. tion of . Lactose agar for detection of acid production, absent Only the original Enterotube is available in South in Shigella (except S. sonnei), Proteus sp., Providencia, Africa at present. This was tested by us and compared Salmonella, and Edwardsiella. with the conventional methods in use in our laboratory. These include the following biochemical reactions: lactose, Department of Medical Microbiology, University of SteIlen­ dextrose, sucrose, mannite, arabinose fermentation, with bosch and Tygerberg Hospital, Tiervlei, CP or without gas production. indole, hydrogen sulphide and M. H. FINLAYSON ornithine decarboxylase production, urea cleavage and B. GIBBS gluconate conversion. In addition, presence or absence of Date received: 25 September 1973. motility is recorded by a Craigie tube. 258 S.-A. MEDIESE TYDSKRIF 13 Februarie 1974 LKW 10 (Byvoegsel-Suid-Afrikaanse Tydskrif vir Laboratorium- en Kliniekwerk)

TABLE 11. ENTEROTUBE REACTIONS OF YERSINIA

Biochemical reactions

Phenyl Organism Dextrose alanine Urea Dulcitol Lactose Lysine Citrate Yersinia + +

One hundred strains of Enterobacteriaceae were tested Enterobacteriaceae, and their Enterotube reactions are by the original Enterotube and by our conventional shown in Table 11. Y. enterocolitica was not differentiated methods. These organisms are shown in Table I. from Y. pseudotuberculosis.

TABLE I. ORGANISMS TESTEI' BY ENTEROTUBE CONCLUSIONS No. Although our comparisons were limited because only 100 Escherichia coli 22 Enterotubes were available, we are of the opinion that the Shigella sp. 7 Enterotube system compares favourably with the conven­ Edwardsiella tarda 1 tional system in use in our laboratory. One of the great ... 2 advantages of the Enterotube system is that all tubes are Salmonella sp. " 9 inoculated in sequence, and rapidly, from a single isolated Arizona hinshawii 3 colony. Most of the cultures examined were correctly Klebsiella pneumoniae 26 identified, the only serious difficulties being with members Enterobacter aerogenes 6 of the Klebsiella-Enterobacter group. Thirty-two strains 8 of these organisms were tested and 3 strains (approximately Providencia sp. 4 10%) were possibly incorrectly identified. The redesigned Proteus mirabilis 3 Enterotube with an ornithine decarboxylase compartment P. vulgaris ... 3 should eliminate such problems. We are in agreement with P. morgani 3 Martin et al. ' in this regard, and also support their P. rettgeri ... 3 recommendations that lactose be removed from the lysine decarboxylase compartment (as has been done) and that arabinose be added in place of dulcitol. RESULTS In our opinion the Enterotube system, in particular the redesigned Enterotube, should give a valuable diagnostic Three of the organisms tested were not identified by the aid in small laboratories and also in field laboratories. Enterotube technique. They were non-motile organisms Superficially it would appear to be rather expensive, but which did not produce urease and which were possibly when the number and variety of tests is considered, it is Klebsiella or non-motile Enterobacter. They were, how­ doubtful whether the media could be produced at a lower ever, ornithine decarboxylase-negative and fermented cost in laboratories which are not equipped and staffed arabinose in the conventional tests and should therefore for large-scale media production. be classified as Klebsiella. Although not normally classified as Enterobacteriaceae, We wish to thank Roche Products (Pty) Ltd, Isando, Trans­ in view of their isolation from cases of gastro-enteritis, 5 vaal, for the supply of Enterotubes. strains of Yersinia, 3 of Y. enterocolitica, and 2 of Y. REFERENCE pseudotuberculosis were examined by the Enterotube 1. Martin, W. J., Wu, P. K. W. and Washington, J. A. (1971): Appl. technique. These strains were well-differentiated from the Microbial., 22, 96.