Early diagnosis of clostridial gas gangrene using sialidase antibodies

Clinica Chimica A&, Elsevier

173 (1988) 251-262

251

CCA 04126

Early diagnosis of clostridial gas gangrene using sialidase antibodies Peter Roggentin

a, Gabriele

H.M. Gutsc~er-Gd~~ and Roland Schauer a

a, Raker

Hobrecht

b

* Biachemisches Znstitut, Christian-Albrechts-Uniuersiiiit, Kiel and b Untersuchungsinstitut Z des Sanittirsdienstes der Bundeswehr, Kiel-Kronshugen (FRG) (Received 10 July 1987; revision received 1 September 1987; accepted after revision 24 December 1987) Key words: Sialidase (neuraminidase); Sialidase antibody; Clostridia; Myonecrosis; ELISA; Diagnosis

In order to improve the diagnosis of gas gangrene, especially at an early stage of infection, new ways for the detection of the responsible C~ostrid~awere investigated. Sialidase, known to be excreted in large amounts by the most frequently occurring myonecrotizing clostridial species, Clostridium perfringens, Clostridium septicum, and Clostridium sordellii, was isolated. With polyclonal antibodies raised against these enzymes, two immunological assays were established, which are directed against the sialidase activity (sialidase inhibition test) and the enzyme protein (‘sandwich’ELISA), respectively. Using these assays, species-specific information about the presence of clostridial sialidase was obtained within 50 min or 6 h. Animal tests revealed that both assays are applicable 8-12 h after clostridiat infection, using resected tissues or wound fluids for estimations. The assays allow specific, sensitive, and qu~titative me~~ement of clostridial sialidases, and no si~fic~t interference by sialidases from other microbes or from host tissues occurred. The applicability of the new assays for an early diagnosis of gas gangrene in human patients is discussed.

In~oduction Suspected gas gangrene (myonecrosis) requires precise, early diagnosis and treatment, in order to prevent death from occurring within a few days. UnambiguCorrespondence to: R Schauer, Biochemisches Institut, Cbristian-Albrechts-U~v~~~t, senstrasse 40, D-2300 Kid, FRG. OOOP-8981/88/$03.506 1988 Elsevier Science Publishers B.V. (Biomedical Division)

Olshau-

252

ous diagnosis was thus far possible only by bacteriological investigations of tissue samples including isolation, identification, and assays of pathogenicity and sensitivity of the growing bacteria to antibiotics, especially with the clostridial strains [1,2]. Since these tests take a few days, further, more rapid methods of bacteriological diagnosis were applied to the clostridia and are summarized here. Microscopy only revealed the presence of gram-positive rods with round ends, ie the morphotype of the genus. Commercially available diagnostic sera against some clostridia predominantly showed cross reactions among the species. Gas chromatography of the volatile fermentation products were found to be typical but not specific for each species. Therapy of patients suspected for gas gangrene, using, eg antibiotics, hyperbaric O,-treatment and resection of infected tissue, had to start before the investigations were completed and the exact microbial diagnosis was known. If, however, information about the nature of wound pathogens is obtainable at an early stage of infection, a specific therapy can be rapidly applied, and a risky therapy like hyperbaric O,-treatment can be avoided in the case of non-clostridial infections. It is known [3] that those clostridial species (Clostridium perfringens, Clostridium sordellii, Clostridium septicum) most frequently causing myonecrosis produce and excrete the enzyme sialidase (EC 3.2.1.18). Earlier investigations [4,5] have shown that antibodies against C. perfringens sialidase inhibit the activity of this enzyme in a species-specific way; no inhibition was detected with sialidases of other Clostridia or other wound-infecting bacteria. Therefore, antibodies against sialidase from C. sordellii and C. septicurn were also investigated for species specificity. Based on these specificities, it was the aim of the present studies to develop immunological tests for the early diagnosis of clostridial gas gangrene. This appeared to be possible by the use of species-specific antibodies inhibiting the corresponding clostridial sialidase activity, which can be followed photometrically. A further immunological test (ELISA) was developed which is directed against the enzyme protein independent of its activity. Both assays were investigated for their applicability to specifically, quickly and quantitatively analyze sialidases isolated from bacterial cultures or present in wound fluids or tissue samples from guinea pigs infected with these Clostridia. Materials and methods Selection of bacterial strains and media for sialidase production

Five each of the most pathogenic strains of C. perfringens, C. sordellii, and C. isolated from human gas gangrene wounds, as well as a strain of C. were taken from the strain collection of Untersuchungsinstitut I des Sanitiitsdienstes der Bundeswehr, Kronshagen, FRG. They were cultured in parallel in brain-heart infusion (Difco, Detroit, MI, USA), Todd-Hewitt broth (Difco), liver-liver broth, or chopped meat medium, the latter two home-made as described earlier [6], to find the conditions for optimal sialidase production. To determine the optimal time for harvesting, samples were taken from the cultures at different times

septicum, chauvoei,

253

after 1: 10 inoculations, and sialidase activity was measured with the substrate sialyl-cY(2-3)-lactose by the method of Warren [7]. Sialidase isolation One strain of each species producing highest sialidase activity (see ‘Results’) was cultivated in the optimal medium (chopped meat medium) by subsequent inoculation of 20 ml, 400 ml and 4 1 of broth. The cells were grown anaerobically at 35 o C in the dark and were harvested by centrifugation (10 000 X g, 20 min, 4” C) when sialidase activity in the medium was at a maximum. Sialidase activity in the medium was concentrated by ultrafiltration, (NH4)$04 precipitation, and gel filtration, as described in refs. [&lo]. Fractions containing sialidase activity were combined, concentrated using ultrafilter membranes (Sartorius, Gottingen, FRG) and dialyzed against 0.1 mol/l sodium acetate buffer, pH 5.5 (this solution was used for animal immunization), or against double distilled water followed by freeze-drying (basis of standard enzyme solutions). Protein concentrations were measured photometrically (Hitachi 220, Tokyo, Japan) at 280 nm [ll] or by the method of Peterson [12]. Animal experiments Production and enrichment of sialidase antibodies Antibodies were raised in sheep and rabbits by a method similar to that described in ref. [5]. Each 100 mU of partially purified sialidase in 100-250 ~1 of 0.1 mol/l sodium acetate buffer, pH 5.5, was subcutaneously injected at weekly intervals over a period of 2-3 mth. For the first injection, the sialidase solution was sonicated for 10 min (Braunsonic 220, Braun, Melsungen, FRG) with an equal volume of complete Freund’s adjuvant (Difco). The increase in antibody levels was detected using the enzyme inhibition test developed for sialidase detection (see below). Animals were bled and, after coagulation, the antiserum was separated by centrifugation (500 X g; 20 min; 20 o C). 50 ml of serum was filtered through Sephadex G-200 (130 x 7.5 cm; 50 mmol/l sodium phosphate buffer, pH 7) and fractions containing sialidase antibodies inhibiting this enzyme activity were combined, concentrated by pressure dialysis (Amicon cell), dialyzed against double-distilled water and freeze-dried. Sialidase occurrence in wouna!~after clostridial infection Clostridia (see above) were cultivated in chopped meat medium for 24 h at 35 o C under anaerobic conditions. The bacteria were washed and resuspended in 0.15 mol/l NaCl solution. Portions of 0.25~ml suspension containing 106-lo8 bacterial cells were mixed with equal amounts of aqueous, 0.9 mol/l CaCl, solution [13] and injected intramuscularly into the hind leg of guinea pigs as described earlier for rabbits [5]. Four, eight, and twelve hours after infection, samples of tissue and fluid were removed from the infected region after anaesthetising and killing the animals. To each 100-200 mg of wound tissue 600 ~1 of 0.15 mol/l PBS/Tween buffer, pH 6.8, consisting of 8 g/l NaCl, 0.2 g/l KCl, 1.15 g/l Na,HPO,, 0.2 g/l KH,PO,,

254

and 0.5 ml/l Tween 20, was added m an Eppendorf tube. After shaking for 10 min, the suspension was centrifuged for 10 min (Centrifuge 5414,lOOOO X g, Eppendorf, Hamburg, FRG), and the supernatant containing sialidase was used for the estimations. Cotton wool plugs soaked with wound fluids were left for 10 min in 1 ml of the PBS/Tween buffer, the resulting suspension was centrifuged and also investigated for sialidase in the supernatant. Sialidase assays Calorimetric

tests

Clostridial cultures and enriched sialidase preparations were investigated for sialidase activity by the method of Warren [7]. Purified sialidase and extracts from infected tissues were measured for enzyme activity by the fluorimetric test of Potier et al [14] modified according to Berg et al [15]. Assays were performed at 37O C in the presence of 0.1 mol/l sodium acetate buffer, pH 5.5, and 0.1 mmol/l4-methylumbelliferyl-a-D-Nacetylneuraminic acid (MU-NeuSAc) in a total volume of 0.1 ml. The reaction was stopped after 10 min of incubation by the addition of 0.9 ml of 0.13 mol/l glycine buffer, pH 10, containing 0.06 mol/l NaCl and 0.04 mol/l Na,CO,. Liberated 4-methylumbelliferone was measured with a filter fluorimeter (Perkin-Elmer, Uberlingen, FRG) at 365 nm for excitation and 450 nm for emission. 10 ~1 of sample solution were usually tested. Immunological-photometric

tests

The test for the inhibition of sialidase activity was developed with antibodies against sialidases from C. perfringens, C. septicum, and C. sordellii derived from sheep, modified according to ref. [5]. Each freeze-dried antibody portion was dissolved (20 mg/ml) in 0.1 mol/l Tris-HCl buffer, pH 7.4. Antibody solution (20 ~1) was added to 20 ~1 of sample solution and incubated for 30 min at 20 o C. The 100% value of sample sialidase activity was obtained by using 20 ~1 of Tris-HCl buffer instead of antibody solution. After incubation, 50 ~1 of 0.1 mol/l sodium acetate buffer, pH 5.5, and 10 ~1 of MU-NeuSAc were added, and sialidase activity was measured as described above. For the measurement of sialidase occurrence independent of enzyme activity a sandwich enzyme-linked immuno-sorbent assay (ELISA) was developed [16]. Microtiter plates (96 well, PVC, Titertek FLOW, Meckenheim, FRG) were used as solid phase. The first sialidase antibody required for this assay was raised in sheep, dissolved in 50 mmol/l sodium carbonate buffer, pH 9.6, and used for coating the wells (100 pi/well) by a 30 min incubation at 20°C. After this as well as after the following absorption steps, the wells were washed three times for 5 min with PBS/Tween. For the second step standard sialidase solution or test samples (100 pi/well) were routinely incubated for 80 min at 20 QC. The incubation conditions of this step can be modified to 40 min at 37 o C. The second antibody, raised in rabbits, was diluted in PBS/Tween and incubated (100 pi/well) for 60 min at 20°C with the antigen (sialidase) already absorbed. The third antibody (peroxidase-coupled goat anti-rabbit IgG) was diluted 1: 1000 with PBS/Tween as suggested by the

255

manufacturer (Sigma, Munich, FRG) and incubated (100 $/well) for 60 min at 20 o C to react with the second sialidase antibody. After final washing, each well was filled with 150 ~1 of a substrate solution composed of 3.4 mg o-phenylenediamine (Sigma), 4 ~1 of an 8.8 mol/l H,O, solution, and 10 ml of 0.15 mol/l sodium citrate-sodium phosphate buffer, pH 5.0. Colour development was stopped by adding 50 ~1 of 2.2 mol/l H,SO,/well after 30 min of peroxidase action at 20 o C in the dark. Colour intensity was measured with a microtiter reader (Multiscan MC 3402, Titertek) at 492 nm. The final value was calculated by subtraction of the value measured for unspecific absorption (without the first antibody) from that for specific absorption. In order to keep performance time to a minimum, antibody concentrations (1 pg-10 mg/ml), incubation times (15-120 min), and temperature (20 or 37” C) were varied resulting in the method described. Optimized antibody concentrations for each sialidase measured are given in results. Both immunological sialidase assays were carried out with some other sialidases to exclude a cross reaction with the homologous system which would result in a false-positive reaction. Fowlplaque virus sialidase (virus particles) was a gift of Prof. H.-D. Klenk (Marburg, FRG); sialidase of Trypanosoma rangeli was donated by Prof. M.E.A. Pereira (Boston, MA, USA). Enzymes of Vibrio cholerae and Arthrobatter ureafaciens were bought from Behringwerke (Frankfurt, FRG) and Calbiochem (Frankfurt, FRG), respectively. All sialidases used for the cross reaction studies were applied in concentrations of ca. 50 mu/ml similar to the test samples. ReSUltS

Selection oj bacterial strains, broth, and time of harvest for sialidase production In all cases, the Clostridia excreted sialidase of highest activity in chopped meat

medium, followed by Todd-Hewitt broth, brain-heart infusion, and liver-liver broth in a decreasing order. The C. perfringens strain A99 was chosen from the five strains of this species tested and harvested 12 h after inoculation into chopped meat medium, reaching a final sialidase activity of 8.6 mU/mg of medium protein. Strain G5 of C. sordellii was grown for 24 h in the optimum medium resulting in 17 mU of- sialidase activity per mg protein. All strains of C. septicum exhibited increasing sialidase activity after this time, even five days after inoculation. Nevertheless, strain h18 was harvested after 72 h of growth, producing 2.8 mU sialidase/mg protein.

Sialidase isolation

The enzyme of C. perfringens A99, enriched by the method described, reached specific activity of 8 U/mg protein with a yield of 26%. This corresponds to purification factor of 930. Sialidase of C. sordellii G5 was enriched 412-fold up to U/mg with 31% recovery. A specific activity of 1 U/mg with a yield of 11% and purification factor of 357 were obtained for the enzyme of C. septicum H18.

a a 7 a

256

\_

\ 1

2

3

4

5

6

drlutmn

7

8

9 10

step

Fig. 1. Inhibitory potency of antibody preparations used against sialidases from C. perfringens A99 (A), C. sordeliii G5 (M) and C. septicurn H18 (@), showing different behaviour during stepwise 1: 2 dilutions.

Production of sialidase antibodies Increasing antibody levels in sheep and rabbits were followed by enzyme inactivation tests during immunization. A typical curve has been published earlier (see Fig. 2 in [5]). During the preparation of antisera, the inhibiting fractions mainly included the IgG-fraction of serum proteins (Fig. 3 in [5]). Isolated antibody pools after G-200 filtration showed differences in their potency of maximum enzyme inhibition, as shown in Fig. 1. Antibodies against C. perfringens sialidase completely inhibited this enzyme, and the inhibitory potency decreased only after many dilution steps. The antibodies against C. sordellii and C. septicum sialidase reached an enzyme inhibition level of only 79% and 508, respectively, which was further diminished after a few dilution steps. Concentration of the antibody solution had no effect on this inhibition level. Methods of sialidase detection The MU-Neu5Ac test [14,15] was found to be a rapid and exact method for the measurement of even small sialidase activities (minimum 10 &J/ml) from various sources. All sialidases used were detected by this test. The adverse effects of colour quenching by high concentrations of protein and fluorescent substances were largely eliminated by a 1: 100 dilution of the samples in this test. The sialidase inhibition test was established with antibody concentrations able to inhibit up to 50 mu/ml of sialidase activity of samples, this being the highest activity measurable with the subsequent MU-Neu5Ac test without further dilutions of the sample. As described above, sialidase activity was inhibited up to a maximum degree, which was found to be species-dependent (Fig. 1). To avoid a marked decrease of sialidase activity even without antibodies during incubation at 37 o C and pH 7.4 [5], the incubation temperature was lowered to 20 o C. A sandwich-ELISA measuring the concentrations of sialidase protein independent of its activity was developed for the enzymes from C. perfringens and C. septicum. However, corresponding experiments with the enzyme from C. sordellii

251

* 0.1

10

1

mu/ml

Fig. 2. Correlation between the absorbance and the amount of sialidase from C. perfringens (A) and C. technique. For experimental details see ‘Materials and Methods.’

sepziaun (0) in the sandwich-ELISA

were not successful. Figure 2 shows that excellent correlations between the photometric absorbance and sialidase amount, as well as a low zero value, can be obtained using freeze-dried antibody fractions in concentrations of 10 pg powder/ml buffer for the first, and 100 pg/ml for the second sialidase antibody These concentrations were used for the assay of both enzymes. For a further characterization of the four solutions, protein content was measured giving values of 0.22 pg/ml for the first, and 6.2 pg/ml for the second antibody solution directed against the sialidase from C. perfringens. The protein contents of the antibody solutions against C. septicum sialidase were 0.92 and 7.5 pg/ml, respectively. For calibration, freshly prepared sialidases were used giving the standardization curve shown iu Fig. 2. Using a semi-logarithmic presentation, a linear region between 0.2 and 10 mU si~dase/~ sample was obtained. The curve resembled a substrate saturation curve resulting in high accuracy for small sialidase amounts. This accuracy, however, was diminished at higher enzyme concentrations, which are also detectable by this test. All components of the ELISA method, except the final peroxidase substrate solution, can be stored frozen for longer than one year without loss of reactivity. To compensate for variations during individual handling of the assays, each ELISA series included sialidase standard solutions. Immunological cross reactions

To ensure that the immunological tests were not affected by siahdases from other Clostridia or from other sources, resulting in false-positive values, enzymes (each 50

258 TABLE

I

Cross-reactions between sialidases (50 mU/ml) of different origins and the antibodies of C. perfringens, C. septic-urn, and C. sordellii with two immunological test systems early diagnosis of clostridial infections Source of sialidase

Sialidase

inhibition

C. perfringens Clostridium perfringens (culture medium) Clostridium septicurn (culture medium) Clostridium sordellii (culture medium) Clostridium chauvoei (culture medium) Vibrio cholerae (purified) Arthrobacter ureafaciens (purified) Fowl plague virus (suspension) Ttypanosoma rangeli (culture medium)

test (W inhibition) C. septicum

C. sordellii

against sialidases developed for the

Sandwich-ELISA C. perfringens

(mu/ml) C. septicum

100

0

0

50

0

0

50

0

0

50

0

0

79

0

0

35

12

29

0

2

0

0

0

0

0

0

0

10

0

3

21

17

4

0

0.1

17

12

10

0.3

0.1

mu/ml) of different origin were introduced into the inhibition test and the sandwich-ELISA instead of the homologous sialidase. From the results listed in Table I it is clear that the three clostridial species studied produce immunologically distinct sialidases, which react specifically in both assays with their respective antibodies. This behaviour was observed with crude and enriched enzymes. Cross reactions were, however, measured in the enzyme inhibition tests between antibodies against the three clostridial sialidases studied and crude sialidase solutions of C. chauvoei, Arthrobacter ureafaciens and Tvpanosoma rangeli or, the particle suspension of fowl plaque virus. In contrast, such influences on the sandwich-ELISA method were very low (maximum 6% in the case of sialidase from A. ureafaciens), which enables a much more specific determination of the sialidases from C. perfringens and C. septicum. The reason for the observed cross reactions was not investigated, as pure sialidases were not available to exclude influences of other components of the sialidase preparations used. No interference of guinea pig sialidase from muscle with the immunological assays was detected. This was also the case for sialidase from non-infected human tissues (unpubl. obs.). Application of the tests to gas gangrene samples

Earlier investigations [5] have shown that measurable sialidase activities in the sera of animals occur lo-12 h after infection. However, at the site of infection sialidase should be detectable earlier. As patient samples from this stage of infection were not available, animals were infected and tissue samples as well as cotton plugs

259 TABLE II Measurement of sialidase activity in tissue samples and wound fluids of guinea pigs infected with C. septicurn, or C. sorakifii dgrending on time after infection

pwfrivtgens,C.

Animal no.

Infected with:

Time (h> $e tion

1 2 3 4 5 6 7 8 9 10 1L 12 I3 14 15

C. C. C. C. C. C. C. C. C. C. C. C. C. C. C.

~e~~~~g~ ~~jRge~ Perfringens Perfiingens Perfringens septicurn septicum septicum septicum septicum so&&i sordellii so&e& sorde& swrfeffii

4 4 8 g 12 4 8 8 12 12 4 4 8 g I2

-

Tissue samples ’ Activity (mIJ/

Inhibitidn

mljb

test

I.0 0.5 (3) 21.5 73.3 (367) 23.2 0 0.4 (2) 0.6 2.9 (8) 1.8 1.1 0.5 (3) 11.0 4*8(13) 21.4

nd. a n.d. + n.d. n.d. n.d. + nd. n.d. nd. f nd.

Wound fluids (in cotton plugs) ELISA (A4921

Activity (mu/ Plug)

Inhibitian test

ELISA

0.2f 0.03 0.59 > 0.8 > 0.8 0.0s 0.17 0 0.46 0.20

5.4 0.7 20.6 33.8 156 0 0.3 0.4 1.2 1.4 0.2 0.4 3.8 3.8 16.7

+ + -t + + + + + f ic

I.25 (53) 0.10 (0) 1.39 (64) 1.64 (64) 1.63 (64) n.d. 0.03 (0.3) 0.32 (1.3) 0.12 (8) 0.42 (2)

CA4921 ’

’ Enzyme activity was determined by the calorimetric assay, the inhibition test, and the sandwich-ELISA. b Values in brackets = mu/g tissue. ’ Values in brackets = mIJ/phtg. ’ nd. = not detected.

soaked with wound fltids were collected 4, 8, and 12 h after infection. Different amounts of sialidase were measured after these times, according to the clostridial species used for infection, the individual animals and the type of sample, as shown in Table II. As expected, significant differences were found among the individual animals. IIowever_ the eqxriments demonstrate that it is possible to detect qxxifrtally the clostridial infection a few hours after its onset, independent of whether tissues or fluids are used for estimation. Eight or 12 h after infection high sialidase activities from all three clostridial species could be detected by inhibition with antibodies in a species-specific marmer. After the same amount of time, positive results were obtained with the ELI,SA, which were even visible by eye if the absorption was greater than 0,2. After 4 h, low but significantly elevated sialidase activities were present in both tissue samples and ,wound fluids infected with C. perfringens and C. sordellii,as determined by the inhibition test and, in the case of C. perfringem,by the ELISA method (Table II>. Thus, the methods described allowed the detection of clostridial sialidases within the following times: The MU-Neu5Ac test requires I5 min, the inhibition test needs 50 min, and with the ELISA results are obtained within 6 h. The first of these assays is unspecific with regard to the origin of sialidase. The second one is specific9 but

not quantitative. The third assay is both specific and quantitative. As the latter assay is directed to the presence of enzyme protein, it is independent of enzyme activity.

The immunological assays developed for a species-specific detection (inhibition test) and qu~tifi~ation (ELISA assay) of sialidases from C. ~e~~~~g~~~ C. sepi~&~~or C. ~~~d~~~i~ (with the exception of ELBA) were repressible, sensitive, and not substan~ally effected by beterologous sialidases. Both tests never failed in the presence of their homologous enzyme, if the activity was higher than 0.1 mU,/ml (inhibition test} or if the amount of sialidase protein was higher than the corresponding activity of 0.2 mU/ml (ELISA assay). This was independent of the purity of the enzymes, and they were, therefore, detectable in broth or wound fluids. The results indicate that the purity of the sialidases from the three Ckwtridia isolated for immunization and the quality of the corresponding antibodies was suitable for estab~s~ng the assays. For the sialidase of C. sordellii the ELISA could not be developed, as only a poor reaction of the second antibody with the enzyme occurred, leading to poor quantitative dete~a~ons. This behaviour may be explained by the relatively small molecular weight of this sialidase (~~ Da; Ro~entin et al, in prep.), possibly giving rise to smaller number of antigen& sites If?] that can easily be saturated by the first antibody. The reason for the lower inhibitory level of antibodies against C. sordellii and C. septicurn sialidase, when compared with that of C. perfringens sialidase (Fig. l), is also not clear. The region of the active centre of both enzymes may not be antigenic and therefore cannot be blocked by antibodies. The measurable inhibitory effect may be caused by a deformation of the sialidase protein molecule and its active centre after antibody binding to a peripheral part of the protein. The sialidase inhibition test would obviously fail to detect inactivated enzyme [lS]. However, with the ELISA method, inactivated sialidase can be estimated, as long as a suffi~e~t number of antigenic sites remain available. The latter assay, therefore, has a broader app~~ation. A further advantage of the ELBA is a lack of cross reactivity with heterologous sialidases. A disadvantage of the ELISA, however, when compared with the inhibition test, is the longer time required for its performance. These i~unolo~cal tests have been successfully applied for the analysis of clostridial sialidases in animal samples. Animals treated with vital bacterial cells yielded, even a few hours after infection, sufficient amounts of C~~~~ridjurn-sp~ific sialidase to be detectable by both assays. It can be expected under non-experimental conditions, for instance after injuries, that this time is longer, since in the contaminated wound, the clostridial spores are only able to germinate in the absence of oxygen, At this early stage of ~f~tion, when the vital cells begin to produce detectable boots of sialidase, the ~unol~~~ assays would prove to be far less time-consuming (1-S h) in comparison to the physiolo~~al tests still necessary for a

261

guaranteed determination of the bacterial species (l-2 days). Further rapid methods used in bacteriological diagnosis were also applied to clostridia, but were not found to be reliable with regard to species-specificity, with one exception: a diagnostic serum against C. novyi (Pasteur) showed no cross reactions with related species in immunofluorescence (unpubl. results). Both sialidase-dependent tests may be applicable to samples from other sources, in which clostridial sialidase production has been induced by, eg sialoglycoconjugates [19] and anaerobic conditions. Interest may exist in the estimation of clostridial infection of spoiled food stuffs and, more importantly, in gas gangrene patients. The application of the assays to human samples derived from patients suffering from clostridial myonecrosis has shown that clostridial sialidase, occurring in infected wounds in high concentrations [5], can be specifically and accurately be detected in this way (Roggentin et al, in prep.). Thus, it appears possible, especially with the ELISA, to diagnose a clostridial infection at an early stage, so enabling a specific and effective therapy and the exclusion of non-clostridial infections. By applying a similar immunological technique as described for these clostridial sialidases, it would be possible to specifically detect other microorganism species’ or groups by their excreted proteins if these show species- or genus-specific antigenicity. Such proteins may be enzymes or toxins known to be produced and excreted by pathogenic microorganisms including fungi, the detected or unambiguous diagnosis of which is often difficult. Acknowledgements

The authors are indebted to Dr. Elke A. Randow and Dr. Norbert Langfeld, Untersuchungsinstitut I des SanitBtsdienstes der Bundeswehr, Kronshagen, for the animal experiments and preparation of tissue extracts. Thanks are also due to Claudia Szeiki, Roswitha Brede and Marzog El Madani (Biochemisches Institut, Universitat Kiel) for technical assistance. This work was supported by Fraunhofer Gesellschaft, Munich, Grant InSan 1-2880-V-4383. References 1 Hentschel M. Gasbrand. In: Hombostel H, Kaufmamr W, Siegenthaler W, eds. Innere Medizin in Praxis and Klinik, 2nd ed. Stuttgart: Georg Thieme, 1977;13202-13207. 2 Jancik JM. Praktische Hinweise zur Differenrentiahliagnose clostridialer histotoxischer Infektionen. Wehrmed Monatsschr 1980;3:79-83. 3 MiiIIer HE. Neuraminidases of bacteria and protozoa and their pathogenic role. Behring Inst Mitt 1974;55:34-56. 4 Sander-Wewer M, Gutschker-Gdaniec G, Kietzmann E, Roggentin P, Hobrecht R, Schauer R. Specific sialidase-antibodies - a possible tool for early diagnosis of clostridial myonecrosis. In: Chester MA, Heine&d D, Lundblad A, Svensson S, eds. Glycoconjugates Proc 7th Int Symposium. Lund: Rahms, 1983;448-449. 5 Schauer R, Sander-Wewer M, Gutschker-Gdaniec G, Roggentin P, Randow PA, Hobrecht R. Siahdase activity in the sera of patients and rabbits with clostridial myonecrosis. Clin Chim Acta 1985;146:119-127.

262 6 Roggentin P, ~~~er-G~~ G, Schauer R, Hobrecht R. Correlative properties for a differentiation of two ~~~trt~~ ~o~~Ii~i phenotypes and their distinction from Ci~r~~~ ~ifg~~~. Zbl Bakt Hyg A 1985;260:319-328. 7 Warren L. The thiobarbituric acid assay of sialic acids. J Biol Chem 1959;234:1971-1975. 8 Gutschker-Gdaniec G, Sander-Wewer M, Roggentin P, Hobrecht R, Schauer R. An efficient method for the isolation of sialidase from Clasrridhn perftingens. Hoppe-Seyler’s 2 Physiol Chem 1982;363:1028. 9 Gutmhker-Gdauiec G, Sander-Wewer M, Roggentin P, Hobrecht R, Schauer R. Comparison of molecular and enzymic properties of sialidases from different clostridial species. In: Chester MA, Heine&d D, Lundblad A, Svensson S, eds. Glycoconjugates. Proc 7th Int Symposium Lund: Rahms, 1983;450-451. 10 Neea S, Veh RW, Schauer R. Purification and characterization of neuraminidase from Clostrtdium perftingem Hoppe-Seyler’s Z Physiol 1975;356:1027-1042. 11 Warburg 0, Christian W. Isolierung und Rristallisation des ~~gsfe~en~ Enolase. Biochem Z 1942;310:384-421. 12 Peterson CL. A s~p~fi~tion of the protein assay method of Lowry et al which is more generally applicable. Anal B&hem 1977;83:346-356. 13 Liieier G. Experimentelle Infektionen durch anaerobe Sporenbildner (Clostridien) der Gasbrandgruppe. In: Eichler 0, Farah A, Herken H, Welch AD, eds. Handbuch der Experimentellen Pharmakologie, Vol. 16. Berlin: Springer, 1966;244-291. 14 Potier M, Mameli L, B6lisle M, Dallaire L, Melancon SB. Fluoromettic assay of neuraminidase with a sodium (4-methyl~belliferyl-a-D-N-acetylneuramina substrate. Anal B&hem 1979;94:287-296. 15 Berg W, Gutschker-Gdaniec G, Schauer R. Fluorescent staining of sialidases in polyacrylamide gel electrophoresis and ultrathin-layer isoelectric focusing. Anal B&hem 1985;145:339-342. 16 Oellerich M. Principles of enzyme-immunoassays. In: Berllmeyer H-U, ed. Methods of enzymatic analysis, 3rd ed., Vol I. Weinheim: Verlag Chemie, 1983;233-260. 17 Rabat EA. Kabat and Mayer’s Experimental Immunochemistry, 2nd ed. Springfield, IL.: Thomas, 1961. 18 Schauer R, Corfield AP. Sialidases and their ~bitors. Med Chem Adv 1981;5:423-434. 19 Schauer R, ed. Sialic acids - chemistry, metabolism and function. New York: Springer, 1982.