GB2048302A

GB2048302A - Identification of bacteria

Info

Publication number: GB2048302A
Application number: GB8014293A
Authority: GB
Original assignee: National Research Development Corp UK; National Research Development Corp of India
Current assignee: National Research Development Corp UK; National Research Development Corp of India
Priority date: 1979-05-02
Filing date: 1980-04-30
Publication date: 1980-12-10
Also published as: WO1980002433A1; DE3070143D1; JPH0321160B2; DK556780A; US4603108A; JPS56500399A; EP0018825B1; EP0018825A1; DK160323C; DK160323B; GB2048302B

Description

1

SPECIFICATION The identification of bacteria

GB 2 048 302 A 1 1 1 This invention relates to the identification of bacteria.

Clinical bacteriology laboratories are frequently called upon to test for the presence of pathogenic bacteria in clinical specimens and further to identify the type of bacteria concerned to help guide the 5 clinician in the choice of treatment, e.g. antibiotic, which is to be used to combat infection.

Customary bacterial identification procedures rely upon a series of characterisation tests on the basps of which the unknown organism is assigned to a defined group of bacteria. These tests include tests in which bacteria are classified by their ability to metabolise various substrates, metabolism being determined by resultant changes in the substrate media, e.g. pH changes, which maybe detected by 10 use of coloured indicators. For such metabolic tests it is necessary, however, to grow the bacteria, usually in complete growth media, and this requires considerable expenditure of time so that it is rarely possible to identify the bacteria on the same day as the specimen arrives in the laboratory. Often, if conventional bacteria identification procedures are used, definitive identification is not possible until 48-72 hours after receipt of the specimen. In the meantime, the treatment prescribed by the clinician 15 can at best be only a guess and may be initially incorrect with the consequence that the infection persists and the condition of the patient deteriorates. There is a pressing need, therefore, for increases in the speed of identification of bacteria which causes infections so that the correct treatment, e.g. antibiotic, may be prescribed without undesirable delay.

Very recently bacterial identification procedures have been proposed which rely only to a limited 20 extent upon bacterial growth for identification, but depend more upon determination of enzymes which are initially present in the bacteria or are induced after a relatively short period of time, e.g. a few hours, and these tests enable bacterial identification to be made earlier than previously, sometimes on the same day as receipt of the sample. Each of these bacterial identification procedures, however, -generally permit identification only of bacteria within certain limited groups and it is thus necessary to carry out a 25 preliminary identification, usually requiring bacterial growth, to select the particular identification procedure which should be used.

A new bacterial identification procedure has now been devised which relies purely upon determination of enzymes present within the organisms, and additionally advantgeously provides a single identification procedure by which a very wide range of commonly encountered bacteria may be 30 identified very rapidly.

Accordingly the present invention comprises a procedure for the identification of bacteria, in which bacteria are subjected to a combination of tests for determination of:

(a) lipase (b) a-glucosidase 35 (c) P-glucosidase (d) P-xylosidase (e) P-glucuronidase (f) P-galactosidase (g) whole cell acid phosphatase 40 (h) acid phosphatase in the presence of agents which disrupt the bacterial cell permeability barrier (i) D L-alanyi-p-naphthylamine specific pepticlase (j) L-arg inyi-p-naphthyla mine specific peptidase (k) N-p-L-glutamyi-p-naphthylamine specific pepticlase 45 (1) glycyi-p-na phthyla mine specific peptidase (m) L-4-hydroxyproiyi-p-naphthylamine specific pepticlase (n) L-leucyi-p-naphthylamine specific peptidase (0) L-leucyl4-methoxy-p-riaphthyla mine specific pepticlase (P) L-lysyl-p-naphthyla mine specific peptidase. 50 (q) L-proiyi-p-naphthylamine specific peptidase (r) L-pyrrolidonyi-p-naphthylamine specific peptidase (S) alanyl-p-nitroalanine specific peptidase (t) glutamyi-p-nitroaniline specific peptidase (U) diacetyl/acetoin producing enzymes 55 (v) p-nitrophenylalanine ammonia-lyase (W) tryptophanase (X) dioxyribonuclease (y) glutamate decarboxylase (Z) cytochrome oxidase 60 The procedure of the invention may be used for identification of a very wide range of bacteria, including most of those bacteria which are customarily encounted clinically. In particular, the procedure may be used to identify the commonly encountered bacterial qroups: Aeromonas, Acinetobacter. Alcaligenes, Bordatella, Citrobacter, Edwardsiella, Enterobacer, Escherichia, Flavobacterium, Hafnia, 2 G13_2 048 302 A 1 2 Klebsiella, Providencia, Proteus, Pseudomonas, Salmonella, Serratia, Shigella, Staphylococcus and Streptococcus. For example the procedure of the invention has been used to identify the following bacterial species: Aeromonas hydrophyla Aeromonas formicans Acinetobacter calcoaceticus var. anitratus Acinetobacter calcoaceticus var. lowoffi Alcallgenes faecalls Bordatella bronchiseptica Citrobacter freudi Citrobacter koseri Edwardsiella tarda Enterobacter aerogenes Enterobacter agglornerans Enterobacter cloacae Escherichia cofl Flavobacterium meningosepticum Hatnia alvei Klebsiella oxytoca Klebsiella pneumon/ae (sensu lato) Klebslella rhinoscleromatis Providencia alcalifaciens Providencia stuarN Pro te us mirabilis Proteus morganfl Proteus rettgeri Proteus vulgarls Pseudomonas aeruginosa Pseudomonas cepacia Pseudomonas fluorescens Serratia marescens Serratia rubidaea Serratia fiquefaciens Staphylococcus aureus Staphylococcus epidermidis Staphylococcus saprophyticus Streptococcus sp It will be appreciated, however, that the procedure of the invention may also be applied to the identification of other bacterial species besides those listed above.

The procedure of the invention permits very rapid identification of bacteria, the enzyme determination tests used generally requiring only a relatively short period of incubation, e.g. from about minutes up to about 2 hours, usually from about 30 to about 90, minutes at about 401C, to give sufficient product for detection e.g. by spectroscopic measurements. It is believed that generally the enzymes which are determined during the procedure of the invention are constitutive enzymes for the bacteria concerned, though in some cases the enzymes may be induced enzymes the rates of bacterial 45 synthesis of which are very fast. Thus the tests used are generally suitably adapted for determination of constitutive erzymes. Without prejudice to the foregoing, however, determination of tryptophanase and deoxyribonuclease appears to require bacteria[ growth usually requiring incubation for a period of 2-2-1 hours.

2 Generally the tests which are used in the procedure of the invention for determination of bacterial 50 enzymes may comprise any suitable tests for the determination of the enzymes in question which are characteristically not dependent on growth of the organisms, and which may include those non-growth enzyme determination tests for the enzymes in question which are well known in the art.

These tests are usually of the kind in which the enzyme is determined by its ability to interact with a specific substrate. Interaction of the enzyme with the substrate, on incubation, usually gives rise to a 55 product which may be detected either directly or after further treatment which may include chemical synthesis from the initial enzyme product.

The product of the enzyme interaction may be detected by spectrometric measurements including fluorimetry or colorimetry. For example, the specific enzyme substrate may comprise an umbelliferyl derivative which on interaction with the enzyme gives rise to umbelliferone which is monitored fluorimetrically, or the substrate may comprise a nitrophenyl, nitroaniline or similar type of derivative which on interaction with the enzyme gives rise to a coloured product which is monitored colorimetrically.

in the simple case, the presence of the enzyme may be determined by simple visual observation, for instance, when interaction of the enzyme and substrate gives rise to a coloured product or a colour65 3 GB 2 048 302 A 3 change.

An example of an enzyme which may be determined by spectrometric measurement of the direct product of enzyme interaction with a substrate is cytochrome oxiclase, for instance, by interaction of sample with tetraphenyl tetramethyl-p-phenylene-diamine (TMPD) an indicator which is oxidised by cytochrome oxidase to give a purple colouration. Also acid phosphatases, beta ga I actosid ases and phenylalanine deaminases may be determined spectrometrically; for instance, using nitrophenyl derivatives, though it is normally necessary to treat the enzyme-substrate mixture with alkali subsequent to incubation so as to develop the nitrophenyl colouration by raising the pH above the pH optimum of the enzyme reaction.

Products which require further treatment after enzyme reaction with- the substrate before 10 monitoring, may be detected spectrometrically. For example, ammonia releasing enzymes, such as leucine deaminase may be determined by reacting the ammonia produced by the enzyme interaction _ with a colour producing reagent such as the Nessler reagent, and monitoring the resultant colour by colorimetric measurement. The ammonia released may be measured directly in the enzyme-substrate reaction mixture or may be removed, e.g. by dialysis, from the reaction mixture before assay.

Also, for example, directly producing enzymes may be determined by spectrometric monitoring of enzyme products after further treatment. In particular, diacetyl producing enzymes are determined by the Voges-Proskauer technique for testing for the presence of diacetyl/acetoin.

Furthermore, glutamate clecarboxylase activity may be determined by colorimetrically determing the carbon dioxide released as a result of interaction of the enzyme with glutamic acid. For instance, 20 carbon dioxide is determined colorimetrically by dialysis out of the reaction mixture after addition of acid e.g. H2SO1, through a hydrophobic membrane into a buffered indicator which changes colour as a result of the change in acidity due to the CO, Other enzymes may be determined by suitable techniques; for instance, substrate specific 25, pepticlases are determined by use of naphthylamine or nitroaniline derivatives, naphthylamine produced 25 being determined either directly by fluorimetry or by colorimetry after diazonium coupling.

The tests used for determination of the enzymes may be varied as desired, for instance to increase the organism specific selectivity of the tests. Thus, two tests for determination of phosphatase activity are included within the procedure of the invention, one for whole cell acid phosphatase activity and one for acid phosphatase activity in the presence of agents which disrupt the bacterial cell permeability 30 barrier. Any suitable agent may be used, though Cetrimide (Cetyl trimethyl-ammoniurn bromide) and lysozyme, especially in combination, are particularly preferred. The use of such an agent, for instance, has the effect of selectivity decreasing the acid phosphatase activity of Proteus bacteria and increasing that of Klebsiella bacteria.

The. invention includes kits of reagents for use in the procedure of the invention. Such kits typically 35 comprise separate specific substrates for each of the enzymes which it is desired to determine in the procedure of the invention. Thus, for example, a basic kit for use in the procedure of the present invention comprises separate specific substrates, for determination of a- z listed previously. Preferably these substrates are such as to give chromogenic products on interaction with corresponding enzymes to advantageously permit colorimetric monitoring. Additionally the kits may also comprise suitable 40 buffer solutions and other reagents, e.g. colour developing reagents, together with the enzyme substrates.

The procedure of the present invention is generally applicable to the identification of bacteria in clinical specimens including urine samples, throat swabs and sputum, wound swabs, stools and blood samples. Bacteria may be isolated from the specimens prior to identification. For example, bacterial 45 cultures are prepared from the specimens and colonies of the organisms to be identified are harvested -from the cultures after a sufficient period of growth e.g. normally about 18 hours, and made up into a suitable form e.g. suspension form, for determination by the procedure of the invention. In particularly preferred embodiments, however, it is envisaged that the procedure of the invention will be carried out on samples derived directly from clinical specimens, for instance, on samples derived directly from urine 50 samples, without need for growth of bacteria and isolation of single colonies.

Prior to assay of bacterial enzymes, however, the samples containing bacteria, whether derived directly from clinical specimens or derived as single colonies after bacterial growth, may in some cases be subjected to treatment to disrupt the permeability barrier of the bacteria and release the enzymes for assay. Any suitable treatment maybe used to disrupt the bacterial permeability barrier. Generally during 55 determination of cytoplasmic and periplasmic enzymes, such as P- galactosidases and acid phosphatases, such prior disruption of the bacterial permeability barrier may be desirable, though other enzymes, such as deaminases, which appear to be membrane associated may require the bacterial cells to be kept intact for enzyme activity to be maintained.

The enzymes assay tests of the procedure of the invention may be carried out by any suitable 60 method or means, including continuous flow and discrete sample analysis techniques, such as those which are well known in the art. In one embodiment a discrete analyser, such as the Kem-0-Mat system, is used. In another embodiment the enzymes may be assayed by automated continuous analysis techniques. In such continuous flow analysis methods it may be desirable to include a protein determination in view of the differing protein concentrations of various bacteria, so that the absolute 65 4 GB 2 048 302 A 4 relative enzyme activities of the bacteria may be determined. Also a protein assay may provide a measure of the blank in spectrometric assays for the absorbance due to the concentration of the organisms.

The conditions used during enzyme determinations may be varied as desired. For example, in continuous flow analysis techniques the relative organism to reagent concentration may be raised, e.g. a sample to reagent ratio in the range from about 1:3 up to about 3:1, to increase the levels of product formed on interaction of enzyme and substrate and thus permit detection of enzyme at lower bacterial suspension concentrations. Also, preferably relatively elevated temperatures e.g. temperatures of about 401C or more, may be used during incubation of sample and substrate to increase the rate of interaction. Preferably, using continuous flow techniques, it is possible to achieve bacterial identification 10 at a very early stage, especially within about 1 hour of the sample reaching the laboratory e.g. if apparatus comprising a single channel for each enzyme test is employed.

In further preferred embodiments, however, the procedure of the invention may be carried out using a test card or other suitable apparatus comprising a plurality of wells or compartments which separately contain specific enzyme substrates for each of the enzyme tests of the procedure and other 15.

reagents, as required, e.g. colour developing reagents. In use the sample, usually bacterial suspension, is added to each compartment and the development or absence of a detectible, e.g. coloured, product after a relatively short incubation preriod e.g. from about 20 minutes up to about 2 hours in preferred embodiments, indicates the presence or absence of the corresponding enzymes in the bacterial sample.

Such apparatus is included within the scope of the invention, and in particularly preferred embodiments 20 may be adapted to handling by automated techniques including automated, preferably computerised, spectro-metric scanning techniques which identify the bacterial species directly from the responses of the enzyme tests.

The procedure of the invention may incorporate additional tests for determination of bacterial enzymes besides those mentioned previously as tests (a)-(z); for instance to strengthen the 25 identification of certain groups of bacteria. Thus, for example, the procedure may include a test for determination of bacteria[ catalase to assist in identification of enterobacteria. Further the procedure may include a test for determination of bacterial urease activity.

In a particular embodiment the invention also includes a bacterial identification procedure for rapid differentiation of the commonly encountered bacterial groups Escherichia, Klebsiella spp, Proteus and 30 Pseudomonas spp, in which a sample comprising bacteria of one of these groups is subjected to a combination of tests for determination of bacterial acid phosphatase, beta-galactosidase, glutamate decarboxylase, phenylalanine deaminase, cytochrome oxidase, diacetyl producing enzymes and urease.

This limited combination of seven tests may be used to rapidly differentiate the bacterial groups Escherichia, Klebsiella spp, Proteus and Pseudomonas spp substantially as hereinbefore described with 35 reference to the full identification procedure incorporating the 26 test (a)-(z).

The invention also includes kits for use in this 7 test procedure; the basic kit typically comprising -separately specific substrates for acid phosphatase, diacetyl producing enzymes, beta-galactosidase, glutamate decarboxylase, phenylalanine deaminase, cytochrome oxidase and urease activity.

In general the procedures of the invention rely upon determination of the enzyme activity profiles 40 of the bacteria undergoing identification, and in accordance with the invention the combination of enzyme tests chosen, i.e. tests (a)-(z) or the limited combination of 7 tests, typically gives a unique "fingerprint" for the bacteria. The unique "fingerprint" for each species or group of bacteria may be determined with reference to the enzyme activity profiles of previously identified bacteria; for instance, from bacteria obtained from culture collections. Enzyme activity profiles may be determined qualitatively or quantitatively. The use of data-processing techniques may be desirable to facilitate the identification by comparison of enzyme profiles of unknown bacteria with those of previously identified bacteria. For example, processing of results obtained by discriminant function analysis, e.g. using the SPSS package (Statistical Package for Social Sciences) has been found to be particularly useful.

The procedures of the invention typically permit very rapid identification of bacteria, a single 50 procedure only being required for identification of a very wide range of bacteria, incuding most of those bacteria which are commonly encountered clinically. This provision of a single procedure advantageously obviates the need for a preliminary test and the undesirable delay that this causes.

The invention is further described by way of illustration only in the following description and examples which refer to the accompanying drawings:

Figure 1 which is a diagrammatic representation of a manifold used for continuous flow analysis of nitrophenol-releasing enzymes in the procedure of the invention; Figure 2 which is a diagrammatic representation of a similar manifold for analysis of ammonia releasing enzymes; W Figure 3 which is a diagrammatic representation of a similar manifold for analysis of diacetyl 60 producing enzymes; Figure 4 which is a diagrammatic representation of a similar manifold for analysis of glutamate decarboxylase; - Figure 5 which is a diagrammatic representation of a similar manifold for assay of cytochrome oxidase and protein; GB 2 048 302 A 5 Figure 6 which is a diagrammatic representation of a flow chart for a three-channel combination continuous flow analysis system for carrying out the procedure of the invention and comprising the manifolds of Figures 1 to 5, and Figure 7 which is a diagrammatic representation of a manifold used for continuous flow analysis of enzymes using substrates which release fluorescently active products.

Tests for determination of bacteria[ acid phosphatase, beta -ga lactosidase, glutamate decarboxylase, leucine deaminase, phenylalanine deaminase, cytochrome oxidase, urease and diacetyl producing enzymes are carried outby continuous flow analysis techniques using a three-channel combined system, a flow chart for which is given in Figure 6, comprising enzyme determination manifolds as shown diagrammatically in Figures 1 to 5.

The various enzyme determinations are carried out as follows:- Enzymes, utilising nitropheny substrates (beta-galactosidase, phenylalanine deaminase and acid phosphatase).

The manifold shown in Figure 1 is used for determination of enzymes utilising nitrophenyl substrates i.e. beta-galactosidase, phenylalanine deaminase and acid phosphatase.

A stream of organism suspension 1 is mixed with an air segmented buffer stream 2 in the first 15 single mixing coil (SMC) 3, and is then mixed with the substrate stream 4 in the second SMC 5 and incubated for 18 minutes in a glass coil maintained in an oil bath 6 at 400C. The reaction is stopped by addition of a strong alkaline solution 7 (1.9M NH40H, 0.68M, NaOH, Triton- X-1 00 0.3 g/1) which also acts as a colour developer for the released p-nitro-phenyl molecule. The stream is then de-bubbled and passed through a 20mm flow-cell in a photometer 8 and the absorbance measured at 450nm for beta- 20 galactosidase and acid phosphatase and at 480nm for phenylalanine deaminase. The absorbance readings obtained are registered on a recorder 10. The figures given in the enclosed area 11 in Figue 1 are the flow rates used for the various reagent and reactant streams.

The various buffer solutions and substrates used are given below in Table 1.

TABLE 1: Enzyme systems tested with nitrophenol substrates Buffer Enzyme Substrate Ion pH Additions Acid phosphatase p -n I tropheny I O.1M Na 5.6 200g g/mI phosphate acetate of cetrimide 1.5mM and lysozyme P-galactosidase p-nitrophenyl- OAM K 7.4 200 ju g/ml P-D-galactophosphate of cetrimide pyranoside and lysozyme 0.5mM Phenylalanine DL-p4p-nitro- 0.1 M K 8.0 None deaminase pheny 0 - phosphate alanine 0.5mM It is believed that the interaction of phenylalanine deaminase with the IDL-P-(p-nitrophenyl)alanine substrate gives rise to production of pnitrophenylpyruvic acid which is the product which is measured by absorbance at 48Onm.

Ammonia-releasing enzymes (leucine deaminase and urease) With reference to Figure 2 the method used for assay of ammonia-releasing enzymes, i.e. leucine 30 deaminase and urease, is based on that of Bascomb and Grantham (1975 "Some Methods for Microbiological Assay" ed. Board & Lovelock pp 20-54, Academic Press, N.Y. ) but using different sized tubes. Ammonia released from the substrates is assayed by adding the Nessler reagent to ammonia which has been collected by dialysis into a 0.005 MHCI recipient stream. Attempts to measure the ammonia directly in the organisrn/substrate stream are not satisfactory due to troublesome 35_ base-line drift and non- reproducibility of the standards. Urease activity is determined in the absence of tris maleate or phosphate buffer as their presence caused a noisy base-line, the enzyme being measured in the presence of distilled water only (pH approximately 4.7-5.0). NH,CI solutions are used as standards in both assays. The substrates used are 5mM 1 -leucine in 0.5M phosphate borate, pH 8.0 and 1 OOmM urea in fresh distilled water. The Nessier reagent is prepared as described by Bascomb and. 40 6 GB 2 048 302 A Grantha m in the abovementioned publication and diluted 1:10 in fresh glass distilled water daily. A 20mm flow cell is used in the photometer and absorbance is measured at 420nm.

In Figure 2 the function of the various components and the relevance of the information given is similar to Figure 1. The broken line arrow 12 indicates the direction of flow across the dialyzer 13 5 without dialysis.

Diacetyl-producing enzymes The Voges-Proskauer reaction for the presence of acetoin/diacetyl is used, in the manifold illustrated in Figure 3, for determination of diacetyl- producing enzymes. The method used in developed from the described by Kamoun et al (Clin. Chem. 1972, 18, 355-357). The substrate solution contains 0.3M sodium pyruvate, 0.1 M acetate buffer (pH 4.5), O.1mM thiamine pyrophosphate and creatine 2g/l. 10 The 1 -naphthol colour-developing reagent is dissolved (25g/1) in 2M sodium hydroxide. Diacetyl solutions are used as standards. Absorbance is measured at 420nm using a 20mm flow call.

Glutamate decarboxylase Glutamate decarboxylase is determined, with reference to Figure 4, by interaction with substrate comprising glutamate to produce C02 which is dialysed from the reaction mixture into a buffered cresol 15 red solution where its presence causes a colour change in the indicator. The method used is based on those described by Leclerc (Annis. Inst. Pasteur, Paris (1967) 112, 713- 73 1), Moran and Witler (1976 J. Food Sci. 41, 165-167) and Technicon Methodology AA 11-08. Technicon sodium carbonate standards are used. The substrate used comprises 0.1 M acetate buffer (pH 3.8), 0.05M sodium glutamate and pyridoxal phosphate 20rng/1. After incubation of the sample and substrate 0.5M 20 sulphuric acid diluent containing Brij-35 (30% Technicon) 1 mi/1 is added to raise the ISH and assist removal of C02 produced by dialysis through a hydrophobic dialysis membrane. The colour reagent used contains 0.4M Tris, ammonia solution 28A111, Brij-35 20pl/l and cresol red 20,ug/1. Absorbance is measured at 420nm in a 1 Omm flow cell.

Cytochrome Oxidase and Protein Assays Cytochrome oxidase and protein are assayed using the manifold apparatus illustrated in Figure 5 which contains similar information and components having similar functions as in Figures 1 to 4.

The reagents used for assay of cytochrome oxidase activity are 0.05M Tris maleate.buffer pH 6.0 introduced prior to the first single mixing coil, 0.5mM NN N'N- tetramethyl-p-phenylene diamine dihydrochloride in 0.001 % (w/v) ascorbic acid introduced prior to the second single mixing coil. The 30 incubation period used is 17 minutes at room temperature and absorption is measured at 550nrn in a 1 Omm flow cell.

Protein assay The protein assay used is based on that of Lowry et al (195 1, J. Biol. Chem., 193, 265-275) and as described by Bascomb and Grantham in the abovementioned publication, though using smaller 35 tubes. The reagents used are alkaline copper solution prepared daily by mixing two mi sodium potassium tartrate (1 Og/1) with 2mi CUS04, 51-1,0 (5g/1) and 46mi 0.2M NaOH containing 0.37M NaCO, (anhydrous), in the order desribed. The alkaline copper reagent is introduced prior to the first single mixing coil. Folin-ciocalto reagent (BDH) is diluted 1:8 in distilled water daily and introduced prior to the second single mixing coil. The incubation period used is 17 minutes at room temperature and absorption is measured at 660nrn in a 1 Omm flow cell. Bovine serum albumin solutions are used as standards. A few drops of chloroform are added to the sodium tartrate, carbonate and protein standard solutions to prevent microbial contamination.

The enzyme assays described above are carried out in a combined continuous flow analysis system, the flowchart of which is illustrated in Figure 6, comprising three channels which are run 45 slmultaneously: one channel (A) for assaying enzymes utilising nitrophenyl derivatives, the second channel (B) for assaying protein and cytochrome oxidase, and the third channel (C) for assaying diacetyl producing, ammonia-producing and C02-proudicing (glutamate decarboxylase) enzymes. The NH3 and CO, products are dialysed from the reaction mixtures and the Nessler and 1 -naphtol colour-developing reagents added where appropriate between the incubation coil and the third SIVIC. The stream from the 50 single sample probe is divided into three, providing bacterial suspensions to each of thechannels A, B and C. The bacterial suspensions are maintained in cups on a revolving sample plate D, while substrates and buffer solutions are provided in continuous streams via the tubes.

All bacterial suspension are first tested for activity of three enzymes, one in each of the channels A, Band C. When the cycle is complete, the substrate, buffer and sample lines are transferred manually 55 to reagents for the next batch of three tests and the sampling of bacteria is restarted. This process is repeated four times including control runs with distilled water in the substrate lines to obtain the absorbance values of bacterial suspensions in the acid phosphatase, cytochrome oxidase, leucine deaminase and urease assays.

1 7 - GB 2 048 302 A 7 Example 1

A total of 199 suspensions of different bacteria are tested by the procedure outlined above using the continuous flow apparatus illustrated in the accompanying drawings. The bacteria are isolated from routine urine specimens cultured overnight at 370C on MacConkey agar plates (Difco Laboratories or Tissue Culture Services). Only plates that showed a homogeneous colony appearance are used. 5 Bacteria[ suspensions each comprising ten colonies in 5ml saline are prepared in a separate location and are brought to the automated system to ensure that automated identification is carried out without any knowledge of the colonial appearance of the samples. A 0.1 ml aliquot is removed from each suspension and added to 5ml nutrient broth which was then incubated at 37 OC for 2 hours on a Matburn rotary mixer. These suspensions are used for inoculation of a chosen set of conventional test 10 media and a nutrient agar slope to be kept for further testing. The remaining saline suspension is tested directly by the automated procedure. Cetrimide and lysozyme are included in the buffer solutions for acid phosphatase and A-galactosiclase assays and mixed with the saline organism suspensions for 2-y' minutes at room temperature prior to addition of the substrate to effect disruption of the bacterial permeability barrier. This treatment halved the absorbance at 340nm of all gram-negative bacteria 15 tested.

The results obtained using the automated procedure are given below in Table 2 and are compared with results obtained by conventional techniques from which the bacteria are previously identitied following the method of Cowan and Steel (11974 "Manual for the identification of Medical Bacteria" 2nd Edition, University Press, Cambridge).

TABLE 2

Distribution of positive results (%) in automated tests according to the invention.

E. coli Klebsiella Proteus spl) spp.

Pseudo- monas spl).

No. strains 125 39 16 19 Automated tests Acid phosphatase 1 95 100 0 Diacetyl (V-P) 0 51 0 0 j3-galactosidase 94 82 0 0 Glutamate decarboxylase 99 0 62 0 Leucine deaminase 0 8 94 0 Oxidase 0 0 0 95 Phenylalanine deaminase 2 0 100 0 Urease 0 62 75 10 Example 2

A total of 96 suspensions of different organisms is tested by the automated method described above in Example 1 and the preceding description except in this case each suspension comprises a single colony of organisms in 1 mi of saline. The automated test procedure adapted is the same as in Example 1 though with fewer control runs and similarly conventional tests are carried out on aliquots of each suspension for the sake of comparison. The results obtained, in terms of the success rate, are ' given below in Table 3 which also includes results for the success rate of the automated tests of Example 1.

8 GB 2 048 302 A 8 TABLE 3

Agreement between automated and conventional identification No. of strains tested E. col l Klebsiella Proteus Pseudo- Total SPI). SPP. monas spp.

Example 1 125 39 16 19 199 Example 2 61 -22 3 5 96 Correct. identification by automated assay (%) Example 1 Example 2 99 97 100 95 98 100 100 100 100 As can be seen the agreement rate achieved with the automated system in Example 1 is 98%. The three strains that are not identified include a strain of Klebslella spp. showing only acid phosphatase activity, a strain of E. colishowing only beta-galactosidase activity, and a strain of Pseudomonas spp. for which the cytochrome oxidase activity results are not available. Ail three organisms are originally classified as unidentified and on repeat testing are identified correctly. Thus with the choice of 8 enzyme activity tests used together with a protein assay test it is possible to correctly identify all four bacterial groups, namely Escherichia, Klebsiella spp., Proteus spp., Pseudomonas spp., by the automated testing procedure.

The success rate achieved in Example 2 was as good as that in Example 1. Thus it is possible to 10 identify some bacteria from a single colony by the use of 9 automated tests.

Example 3

A total of 304 suspensions of culture collection and freshly isolated strains were tested by the full 26 test procedure.of the invention as described below.

Culture collection strains were from the National Collection of Type Cultures, the Computer Trial 15 Laboratory at Colindale and the Bacteriology Department, St. Mary's Hospital Medical School; all had been characterised and identified by conventional methods. Fresh strains were isolated from routine urine specimens and identified using API strip 20E. All strains were cultured overnight at 370C on MacConkey agar, plates, without added sodium chloride (Difco Laboratory and Tissue Culture Services).

Bacterial suspensions of 8 colonies in 4ml of sterile saline were used for all enzyme assays.

Enzyme activities were measured using 3 different analytical systems. The continuous flow system, as described in Example 1, was used to determine activity of cytochrome oxidase, glutamate decarboxylase and protein content. A semi-automated continuous flow method was used for detection of enzymes listed in Table 4, by fluorimetry using the continuous flow manifold shown in Figure 7. A discrete analyser (Kem-0-Mat) was used to perform the, tests listed in Table 5.

In the semi-automated method buffered substrate solution (200YO and bacterial suspension (200jul) were added manually to Auto Analyzer cups. The bacterial and substrate mixture was incubated for 90 min. at 401C and the fluorescence measured in a Lochbrt fluorimeter attached to a Newton sampler using a continuous flow manifold as detailed in Figure 7. This system allowed measurement of 100 samples per hour. The enzymes tested and the composition of the reagents are given in Tables 4(a) 30 and 4(b).

k m 1 GB 2 048 302 A 10 composition are given in Table 5. For all tests 350M1 of diluent and 70A1 of reagent 1 were dispensed into each euvette. Bacterial suspensions were dispensed in 50p] aliquots for the DNAase test and in 25jul aliquots for all other tests. Minimum incubation periods for tryptophanase and DNAase were 2h, for VP and PNPA 1 1h, and for the remaining tests one hour. At the end of the incubation period reagent 2 2 was added to the cuvettes containing substrate and bacterial suspension in 50pi aliquots for the VP, PNPA, PNPP, Pl\1PP+C+L and PNPG+C+L tests and in 3501A1 aliquots for the INDOLE test. Final absorbance was read 2 min. after addition of the second reagent, except for VP and INDOLE tests which were read after 20 min. Absorbance measurements from the 3 different analytical systems were fed into a computer for calculation of enzyme activity and specific enzyme activity of each organism 10 suspension in each enzyme test.

R 1; 9 GB 2 048 302 A 9 TABLE 4 (a) - Hydrolases Enzyme Buffered substrate Manifold reagent Buffer Substrate Concn. mm Lipase Tris -HC1 4 -methylumbel if eryl - 0.05 HCI,2OmM pH 9.0, 0-05M nonanoate a-Glucosidase Tris 4-methylumbel iferyl - 0.2 HCl,20mM phosphate a-D-glucopyranoside pH 8.0, O.1M P-Glucosidase Tris 4-methylumbeliferyl- 0.5 HCI, 20mM phosphate P-D -glucopyranoside pH 8.0, OAM P -Xylosidase Tris 4-methylumbel iferyl 0.2 HCI, 20mM phosphate P-D-xylopyranoside pH 8.0, O-1M P'-Glucuronidase sodium 4-methylumbeliferyl- 0.5 glycine acetate P-D-glucuronide buf fer pH5.6, OAM pH 8.8, imm All substrates were supplied by Koch Light. They were dissolved in methoxyethanol and diluted with appropriate buffer to the final concentrations listed.

TABLE 4(b) - Substrates used for peptidase assays i Substrate DL -Alan ine -P -naphthy lamide hydrochloride L-Arginine-p- naphthylamide hydrochloride (arg NAP) N -y-L-Gutamyl -p-naphthylamide Glycyl-p-naphthylamide hydrochloride L-4-Hydroxyprolyi-p-naphthylamide (OH pro-NAP) L-Leucyl-p-naphthylamide hydrochloride (leu-NAP) L-Leucyl-4mothoxy-)S-naphthylamide hydrochloride (leu-4-m-NAP) L-Lysyl-.8 '-naphthylamide carbonate L-Prolyl -p Inaphthylamide hydrochloride (pro-NAP) L-Pyrrolidonyl-p-naphthylamide (pyr-NAP) Suppl ier Sigma Sigma Sigma Koch Light Sigma Sigma Sigma Koch Light Sigma Sigma All substrates were dissolved in dimethyl sulphide and diluted to a final substrate concentration of 5 O.lmM with 0.1M Tris phosphate buffer,pH 8.0, containing cobaltous nitrate (2M). The reagent used in the continuous flow manifold was 0. 1 M Tris phosphate buffer, pH 8.0.

The discrete analyser, Kem-0-Mat, (Coulter Electronics) was used to perform the tests listed in Table 5. The bacterial suspensions were placed in the sample cups, buffers were dispensed by the diluent syringe, substrates and reagents (for developing the colour of the reactant end-products) were 10 dispensed by the reagent syringe. To increase sample throughout and lengthen incubation pd'riods cuvettes containing bacterial suspensions, buffer and substrate were incubated outside the Kern-OMat. The cuvette changer was used for simultaneous insertion of all 32 cuvettes and for their removal. Two program cartridges were used to perform each test. Program cartridge 1 was used for distribution of bacterial suspensions, buffer, cofactors and substrate to the cuvette and for reading the initial absorbance of each cuvette. The cuvettes were then removed and placed in a 4WC incubator. The diluent and reagent syringes and probes were rinsed and charged with the reagents for the next enzyme test. After incubation the cuvettes of each enzyme test were returned to the analyser. Program cartridge 2 was used for adding the second reagent and for reading the final absorbance. The details of reagent N TABLE 5 (2)

Test Diluent Reagentl Reagent 2 Acid phosphatase 0.03M sodium acetate 5mM p-nitrophenyl- 1.9M ammonium hydroxide (PN1PIP+C+Q buffer pH 5-6, containing disodium phospate containing 0.68M NaOH and cetrimide 0.007g/1 Triton-X-100, 0.3g/1 lysozyme 0.0079/[and NaCI 2.8g/1 P-Galactosidase 0.03M potassium phosphate 8mM p-nitrophenyl - 1.9M ammonium hydroxide (PN1PG+C+Q buffer, pH 7.5, containing P-D-galactopyranoside containing 0. 68M NaOH and cetrimide, 0.007g/1 Triton-X-100, 0.3!g.l lysozyme 0.00Tg/1 and NaCI 2.8 g/I Tryptophanase Peptone (Oxoid L37) 51911 2mM L-tryptophan 78% (vlv) ethanol containing (INDOLE) and NaCC 6.75911 in 1116MHCl and p-dimethyl - distilled water arninocirmarnaldehyde 2.37g/1 Deoxyri bonuc lease 0.07M -!."ris-IHCI buffer, distilled water - - - - - - - - - - - - - - (DNA) pH 9-0, containing 1M9012 ^ ethidium bromide 23mg111 and deoxyribonucleic acid (Sigma calf thymus) 227,mg/1 All diluent solutions contain Triton-X-100, 0.lml/i,, to decrease air bubbles on cuvette walls.

1 J1 1, 1 4 A TABLE 5 (1)

Test Diacetyllacetoin producing enzymes (V-P reaction) (VP) p-Nitrophenylalanine ammonia-lyase (PNPA) Alanyl -p-nitroalanine specific peptidase (PNAA) Glutamyl -p -nitro aniline specific peptidase (PNAG) Acid phosphatase (PNIPP) 1.

Diluent 0.05M acetate buffer, pH 4.5, containing 0.15M sodium pyruvate, creatine lg/1, and NaCI 4.25g11 0.1M potassium phosphate buffer, pH 8.0 O.1M potassium phosphate buffer PH 8.0 0.07M Tris phosphate buffer PH 8.0 containing cobaltous nitrate 2pM 0.03M sodium acetate buffer, PH 5.6 containing NaCl ?.8g/l Reagent 1 0.4mM cocarboxylase 1OrnK DL-P-(p-nitrophenyl) alanine 3.5mM L-alanine4-nitroanilide hydrochloride (BDH) 3.5mM y-L-glutamyi 4- nitroanilide 5mM p-nitrophenyldisodium phosphate Reagent 2 2M NaOH containing a-naphthol 50g11 1.9M ammonium hydroxide containing 0.68M NaOH and Triton-X-100, 0.39/1 1.9M ammonium hydroxide containing 0.68M NaOH and Triton-X-100, 0.3 g/[ G) cc] N) 0 -p.

CO W 0 N 13 GB 2 048 302 A 13 Analysis of results Three hundred and four cultures failing, into 35 species were tested by both Kem-0-Mat and continuous flow systems on a total of 26 tests and a protein assay. Thirty one strains were excluded from the calculations because of missing values, mixed cultures and doubtful identity by conventional testing. The remaining 273 cultures were divided into a known set (Training Set) comprising of between 5 210 and 221 cultures of mainly culture collection strains; and an unknown set (Test Set) of 52-73 cultures, mainly of strains freshly isolated from cultured urine specimens and tested the day after the urine specimen was received.

The discriminant function analysis (DFA) using SPSS software was used to identify both training 10 and test sets.

The 35 species were divided either into 17 genera or sub-divided within each genus to species or groups of species as shown in Table 6.

14 GB 2 048 302 A 14 TABLE 6: Division and coding of bacterial taxa in discriminant function analysis (DFA) Coding used in DFA Number in groups Bacterial taxon 17 22A 22B 25A 25B 35 Escherichia 1 1 1,2 1,2 1,2 1 Klebsiella pneumoniae 2 2 3 3 3 2 sensu lato Klebsiella rhinoschlero- 2 2 3 4 4 3 matis Klebsielia oxytocum 2 2 3 - 3 3 4 Proteus mirabilis 3 3 4 5 5 5 Proteus morganii 3 3 4 6 6 6 Proteus rettgeri 3 0 0 0 0 7 Proteus vulgaris 3 3 4 7 7 8 Pseudomonas aeroginosa 4 4 5 8 8 9 Pseudomonas cepacia 4 5 6 9 9 10 Pseudomonas fluorescens 4 6 7 8 8 11 Staphylococcus epidermedis 5 7 8 11 11 12 Staphylococcus aureus 5 7 8 10 10 13 Staphylococcus sapro- 5 7 8 11 11 14 phyticus Streptococcus 6 8 9 12 12 1,5 Citrobacter freundii 7 9 10 113 13 16 Citrobacter kossori 7 10 11 14 14 17 Providencia alkali- 8 11 12 15 1,5 18 faciens Providencia-stuarti 8 11 12 15 15 19 Serratia marcescens 9 12 13 16 16 20 Serratia rubidaea 9 12 13 16 16.21 Serratia liquefaciens 9 12 0 16 0 22 Acinetobacter calco13 14 17 17 23 aceticus va. anitratus Acinetobacter calco- 10 13 14 17 17 24 aceticus va. lowfM Alcaligones faecalis 11 14 15 18 18 25 Alcaligenes bronchi- 11 15 16 19 19 26 septicus Flavobacterium 12 16 17 20 20 27 meningosepticum Enterobacter aerogenes 13 17 18 21 21 28 GB 2 048 302 A 15 TABLE 6 continued Coding used in DFA Number of groups Bactenial taxon 17 22A 22B 25A 25B 35 Enterobacter cloacae 13 18 ig, 22 2 q Enterobacter aglomerans 13 17 18 21 21 30 Hafnia alvei 14 19 20 23 23 31 Salmonella 15 20 0 0 0 32 Edwardsiella tarda 16 21 21 24 24 33 Aeromonas hydrophyla 17 22 22 25 25 34 Aeromonas formicans 17 22 22 25 25 35 Each grouping was subjected to DFA using all 26 tests (variables) and a number of restricted sets of variables, to determine the best mathematical treatment of the data.

The percentage agreements with previous identification obtained with the Training and Test Sets 5 in the various combinations are shown in Table 7 when all 26 variables were included.

TABLE 7 Effect of grouping of taxa in the DFA on identificationperformance using all 26 variables Training Set No. of No, of Agreement groups strains Test set No. of Agreement strains 17 215 74.5 64 45.2 22A 215 78.9 64 45.9 22B 215 80.9 68 47.2 25A 221 81.4 52 50.0 25B 215 82.3 58 48.3 215 70.3 64 56.9 Highest percent of agreement was obtained in the training set with the 25B grouping and with the test set with 25A grouping. The effect on agreement percentage of exclusion of selected variables from the DFA is shown in Tables Sand 9, using 22B and 25B groupings.

16 GB 2 048 302 A 16 TABLE 8 Effect of excluding selected variables in the DFA on identification performance, using 22B grouping No. of Agreement variables % Variables excluded, in addition to those excluded in the preceding line 26 80.9 - - - - - - - 81.4 PNAG 24 80.5 VP 23 75.8 INDOLE 22 76.3 PNAA 21 74.4 Glutamate decarboxylase 73.0 PNPA 67.4 Arg-NAP, leu-NAP, pro-NAP, p-glucuroni dase, DNAase 51.2 OH-pro-NAP, leu-4-m-NAP, pyr-NAP, a-glucosidase, P-glucosidase TABLE 9(a) Effect of excluding selected variables in the DFA on identification performance, using 25B grouping Percentage agreement No. of variables 26 25 24 23 23 No. of Bacterial strains genus 23 Escherichia 95.6 91.3 91.3 91.3 91.3 24 Klebsiella 87.5 87.5 87.5 87.5 83.3 26 Proteus 69.2 65.4 65.4 65.4 65.4 17 Pseudomonas 94.1 94.1 94.1 94.1 94.1 19 Staphylococcus 84.2 84.2 78.9 78.9 78.9 7 Streptococcus 85.7 85.7 85.7 85.7 85.7 13 Citrobacter 84.6 84.6 84.6 84.6 84.6 17 Providencia 94.2 94.2 94.2 82.3 94.2 21 Serratia 71.4 71.4 71.4 66.6 61.9 Acinetobacter 70.0 80.0 ' 70.0 70.0 60.0 7 Alcal igenes 100 100 100 100 100 2 Flavobacterium 100 100 100 100 100 22 Enterobacter 68.2 68.2 63.6 63.6 63.6 Hafnia 100 100 100 100 100 6 Edwardsiella 100 100 100 100 100 9 Aeromonas 77.8 77.8 77.8 77.8 77.8 1 All genera 83.8 84.8 82.0 81.4 81.0 228 Tests excluded.... PNAG PNAG PNAG PNAG VP VP a-glucosidase PNAG j6-xylosidase 1 17 GB 2 048 302 A 17 TABLE 9(b) Number of genera in each catagory of percentage agreement shown in Table 6(a) Percentage Number of variables 26 25 24 23 23 agreement 90-100 7 7 7 6 7 80-89.9 4 5 3 4 3 70- 79.9 3 2 4 3 2 -69.9 2 2 2 3 4 In both cases exclusion of the PNAG test results in a slight increase in the total percentage agreement as well as in the number of genera showing 80% agreement. It is believed that this was due to experimental inaccuries in the results arising from the PNAG test. Exclusion of further tests whether by computer selection or by inspection causes further decrease in the % of agreement which is unacceptable.

The taxa which proved difficult to separate are Proteus rettgeri, Acinetobacter anitratus, and the EnterbacterlSerratia species. This is probably a result of use of strains kept for a long time in culture.

Some of these strains when tested by both the conventional API 20 and R/B tube kits did not show 'good identification' and in a number of strains the results of one system did not agree with those 10 of the second.

More strains preferably freshly isolated ones, are needed for solving the difficulties in these taxa.

The results obtained show, however, that the full procedure of the invention gives greater than 80% accuracy of identification over a very wide range of different bacteria. It is believed that this accuracy can be increased if enzyme profile information from a large and more representative group of 15 previously identified bacteria is available for comparison.

Claims

CLAIMS 1. A procedure for the identification of bacterial in which

bacteria are subjected to a combination of tests of determination of bacterial enzymes (a)-(z) as hereinbefore defined. 20
2. A procedure for rapid differentiation of bacteria of the groups Escherichia, Klebsiella spp., Proteus and Pseudomonas spp., in which sample comprising bacteria of one of these groups is subjected to a combination of tests for determination of bacterial acid phosphatase, beta-galactosidase, glutamate decarboxylase, phenylalanine deaminase, cytochrome oxidase, diacetyl producing enzymes and urease.
3. A procedure according to Claim 1 or 2, in which the tests used, besides those for determination 25 of tryptophanase and cleoxyribonuclease, are suitably adapted for determination of constitutive enzymes.
4. A procedure according to Claim 3, in which the period of incubation for the constitutive enzyme test is from about 10 minutes up to about 2 hours.
5. A procedure according to Claim 3 or 4 in which the agent which is used to disrupt the bacteria, 30 cell permeability barrier is cetrimide and/or lysozyme.
6. A procedure according to any of the- preceding claims which is carried out on a sample derived directly from a clinical specimen.
7. A procedure according to any of the preceding claims comprising the use of continuous flow analysis techniques.
8. A procedure according to Claim 7, in which the procedure includes a protein determination.
9. A procedure according to Claim 7 or 8, in which the organism to reagent concentrations used in the tests are in the range of about 1:3 up to about 3:1.
10. A procedure according to any of the preceding claims in which incubation is carried out at a temperature of about 401C or more.
11. A procedure according to any of Claims 2-6 or 10 which is carried out using apparatus comprising a plurality of compartments which separately contain specific enzyme substrates and other reagents as required, and in which sample is added to each compartment and the bacteria is determined by the development or absence of detectable products in the compartments.
12. A kit comprising reagents for use in a procedure according to Claim 1 or 2 comprising 45 separate specific substrates for each of the enzymes which it is desired to determine.

18 GB 2 048 302 A 18
13. A kit according to Claim 12, in which the enzyme substrates are such as to give chromogenic products on interaction with corresponding enzymes.
14. A kit according to Claim 12 or 13, comprising suitable buffers and other reagents together with the enzyme substrates.
15. A kit according to any of Claims 12-14 comprising a test card or other suitable apparatus 5 comprising a plurality of wells or compartments which separately contain specific enzyme substrates for each of the enzyme tests of the procedure and other reagents as required.

Printed for Her Majesty's Stationery Office by the Courier Press, Leamington Spa, 1980. Published by the Patent Office, Southampton Buildings, London, WC2A 1 AY, from which copies may be obtained.

j, ky 1