Download Behavior of Bacterial Capsules Towards Different Proteins: A pH-Dependent Study and more Lecture notes Bacteriology in PDF only on Docsity!
TOMCSIK,J. & GUEX-HOLZER,S. (1954). J. gen. Microbiol. 10,97-
Demonstration of the Bacterial Capsule by means of a pH -dependent, Salt-like Combination with Proteins
BY J. TOMCSIK AND S. GUEX-HOLZER
Institute for Hygiene and Bacteriology, University of Basel, Switzerland
SUMMARY :^ The capsules of bacteria and of^ Cryptococcus^ neoforrnans^ are invisible under the phase-contrast microscope. They can be made visible through addition of a number of proteins at a certain, usually very narrow, pH range lying on the acid side of the isoelectric point of the proteins. The pH range a t which the reaction occurs shifts with the isoelectric point of the protein used. The optimal pH range depends, on the other hand, upon the capsular substance of the micro-organism. The non-specific capsular reaction is elicited through a salt-like combination of several proteins with the bacterial capsule, leading to precipitation at a pH value on the alkaline side of the isoelectric point of the bacterial surface and on the acid side of the isoelectric point of the proteins. The reaction is reversible; it disappears on changing the pH. The non-specific capsular reaction at an inter-isoelectric pH zone of the bacterial surface and of the proteins is not a ‘swelling’ reaction. A swelling of the capsule might occur with some bacteria as a secondary phenomenon on the alkaline side of the reaction zone. The reaction occurs in a broader zone in members of the genus Bacillus^ where it reveals a distinct shrinkage of the capsule at the acid side of the reacting zone. Nakamura (1923) studied the effect of acids and alkalis on the egg-white lysis of sarcina. He observed that HC1 at n-/2OO to ~ / 1 6 0 0prevented the lysozyme activity. When, however, an hour later the suspension was made slightly alkaline with 0.05 N-NaOH, an instantaneous lysis occurred. When studying the effect of lysozyme on the cell structure of capsulated lysozyme- sensitive members of the genus Bacillus (Tomcsik & Guex-Holzer, 1952),we re- peated the experiment of Nakamura and extended his observation by examining bacteria, after the addition of egg-white at different pH values, under the phase-contrast microscope. During this experiment, we made the unexpected observation that the capsule of these bacteria was beautifully demonstrated in the presence of a 1/50 dilution of egg-white a t about ~/320-HC1.The fractionation of egg-white revealed that the albumin, globulin and lysozyme fractions all were able to induce a non-specific capsular reaction; however, the optimal pH zone for eliciting this reaction was different with the three different proteins. The purpose of this paper is to describe this phenomenon and to study the behaviour of a representative member of capsulated micro- organisms towards several proteins, including proteins with different iso- electric points. MATERIALS AND METHODS Organisms. Bacillus anthracis : strain Vollum (NCTC 7200) ; an attenuated strain (A,) isolated by us from a Pasteur vaccine. A strain of B. megaterium, which produced a capsule on ordinary agar medium, obtained through the
kindness of J. P. Aubert (Service de Fermentations, Institut Pasteur, Paris).
G M X I 7
(^98) J. Torncsilc and S. Guez-Holxer
Bacillus M. isolated and described by us (Tomcsik, 1950; Tomcsik & Guex,
1951). A strain of B. subtilis which produced a capsule in 3 hr. agar cultures and abundant slime which consisted of D-glutamic acid polypeptide (strain studied by Tomcsik & Guex, 1951). A strain of Streptococcus haemolyticus,
group C (human), isolated from a case of otitis media (obtained from A. Grum-
bach, Hygiene Institute, University of Zurich). Strains of Klebsiella pneu- moniae, types A, B and C (from the Statens Seruminstitut, Copenhagen)
and another type B isolated in our Institute from a guinea-pig epidemic.
Diplococcus pneumoniae, type 111, a strain isolated in our laboratory from
a human case of pneumonia. Cryptococcus neoformans, strains D.U. and R.E.
(obtained from Dr E. W. Evans, Department of Bacteriology, University of
Michigan, Ann Arbor, U.S.A.). Culture media. Gladstone & Fildes (1940) CCY agar medium was used throughout the work with the following exceptions : the capsule production
of the B. anthracis strains was not optimal on CCY medium in presence of
20 yo (v/v) CO,; thus Difco peptone agar was used for this purpose. The streptococcus and pneumococcus were grown on blood agar and Cryptococcus
neoformans on Sabouraud medium. B. subtilis, B. rnegaterium and Bacillus M.
were incubated at 32O, all the other strains a t 37". Proteins. Albumin and globulin fractions of horse serum as well as those of egg-white were prepared by ammonium sulphate fractionation (Sahyun, 1944). Crystallized lysozyme was prepared according to the method of Alderton & Fevold (1946); this preparation was compared with a sample obtained from the Mann Research Laboratories, New York. The other prepara- tions were as follows : pepsin and trypsin (Kahlbaum); isoelectric casein (Difco); haemoglobin and protaminsulphate (Hoffmann La Roche, Basel) ;
bovine plasma albumin fraction V (Armour Laboratories, London).
Preparation of the reacting mixture of bacteria and proteins at different pH values. The bacteria were usually suspensions in 0.85 yo NaCl from 24 hr. agar cultures. The turbidity was less than in tube no. 3 of the McFarland BaSO, standard series (Kolmer, 1924), giving ten or more bacteria in the reacting mixture per field in microscope preparations with the oil-immersion objective no. 90 and ocular no. 10. Different dilutions of the proteins in physiological salt solution or in distilled water were added to equal volumes of bacterial suspension; 0.3 ml. of this mixture was added to 0.3 ml. of different buffer solutions which covered a pH range between 2 and 9 at intervals of 0.4 pH unit up to pH 4.4, and at larger intervals on the more alkaline side of the series. The buffer solutions of Teorell & Stenhagen (1938) were used throughout, if not otherwise mentioned. Two loopfuls of these mixtures placed on slides were examined under cover-glasses with the oil- immersion objective of the phase-contrast microscope of Wild and Co. (Heer- brugg, Switzerland). The microscopic examinations could be carried out immediately after setting up the series, but the results were hardly changed when the series was kept at room temperature for 20 hr. The microscopic examinations were usually carried out 1 hr. after setting up the series. The photographs were made with Leitz Ortholux microscope and Aristophot
larger at the optimal reaction than that found by the indian ink method. Only in preparations kept in a wet chamber for several hours could a secondary swelling occasionally be observed. The only pneumococcus strain studied gave the optimal reaction at a still higher pH value than did the klebsiellas of types A and C. A swelling of the capsule was observed regularly only with this organism (Pl. 2, fig. 18).
J. Torncsik and S. Guex-Holxer
Table 1. Non-specijic pH-dependent capsular reaction of several bacteria with bovine serum diluted 1/ pH of the buffer solution 2.0 2.4 2.8 3.2 3-6 4 4.4 4- Observed reaction
7-^ h^ --^7
Organism (^) r^ A -- v
Cryptococcus neoformans +
K. pneumoniae type A, C ( + I
Klebsiella pneumoniae type B -
Streptococcus haemolytincs group C -
Lliplococcus pneumoniae type III -
Bacillus subtilis -
B. anthracis, ‘Vollum’ ( 4 ) B. anthracis (vaccine) ( k ) B. megaterium ( k ) Bacillus M (k)
f
( + I
(+)
( + I
(+ +)
f
( + + + ) (+++) (+ +)
k -^ -^ -
+ + + + denotes the maximal reaction, the capsule appearing in the phase-contrast microscope as
+ denotes a very faint grey capsule.
+ + and + + + denote grades between + and + + + +.
k denotes the appearance of a faint grey capsule only on some bacteria. ( ) denotes shrinkage of the capsule and increased light scattering of the bacterial bodies.
a dark, grey body.
The chemical nature of the capsular substance of C. neoformans was deter- mined by Drouhet, Segretain & Aubert (1950), by Evans & Mehl (1951) and by Evans & Theriault (1953); xylose, mannose and glucuronic acid were found in the purified capsular polysaccharide. C. neoformans showed in our studies the lowest pH zone as compared with any other capsulated micro- organisms, with each of the proteins tested. No swelling of the capsule was observed at the optimal pH. Neill, Castillo, Smith & Kapros (1949) also did not observe swelling, when studying the specific capsular reaction of this micro-organism by the Neufeld test. On the contrary, a definite contraction of the capsule was observed, with our method, on the acid side of the reacting pH zone. In the second group of bacteria, those which produce polypeptide capsules, B. anthracis and B. subtilis could be separated. Capsular substance of
B. anthracis was separated from the bacterial polysaccharide by Tomcsik &
Szongott (1933); Bodon & Tomcsik (1934) demonstrated that only the poly- peptide and not the polysaccharide antibody gives a specific reaction with the capsule of B. anthracis. The chemical nature of this capsular substance was cleared up by Ivanovics & Bruckner (1937) who showed that it is a D-glutamic
acid polypeptide ; these observations were confirmed and extended by Hanby
Reaction of bacterial capsules with proteins 101 & Rydon (1946), Hanby, Waley & Watson (1948), Ambrose & Hanby (1949), Housewright & Thorne (1950). Ivanovics & Erdos (1937) isolated the same substance from B. subtilis. In the case of B. anthracis grown on agar medium with 20 yo (v/v) atmospheric CO, concentration, the non-specific capsular reaction i s easily demonstrable with all sera and proteins over a broad zone of pH values (Pl. 1, figs. 2, 3, 5). On the more acid side of this zone (see Table 1) a definite contraction of the capsule was observed (Pl. 1, fig. 6). The bacteria and the contracted capsules showed a peculiar light-scattering which made photography difficult. Our B. subtilis strain had to be incubated only a few hours at 32" in order to show a considerable proportion of capsulated bacilli (Pl. 1, fig. lo). In older cultures the polypeptide is given off as a slimy substance. The pH zone of the non-specific capsular reaction was narrower than with B. anthracis and the contracted capsules at acid reaction were hardly visible. A slime reaction, analogous to the non-specific capsular reaction, can be demonstrated occasionally by increasing the concentration of protein. In these cases the pH range at which the capsular and the slime reactions can be demonstrated is the same.
We listed in the third group of capsulated bacteria B. megaterium and
Bacillus M which have a complex capsular composition. As shown in recent
studies (Tomcsik, 1951 ; Tomcsik & Guex-Holzer, 1951, 1952), the polypeptide antibody produced by intravenous injections of killed capsulated B. anthracis organisms into rabbits reveals a homogeneously distributed capsular sub- stance in these bacteria (Pl. 2, fig. 14). On the other hand, the homologous specific antibodies, which can be specifically absorbed with the serologically distinct polysaccharides of these two bacteria, induce the visibility of a charac- teristic capsular structure (Pl. 2, fig. 15), as well as of the cell wall (Tomcsik & Guex-Holzer, 1952). The analysis of these polysaecharides will be reported in
another communication. It is, however, interesting in view of the present work
that these polysaccharides do not contain uronic acids. The reaction of these bacilli with proteins, that is to say, the non-specific capsular reaction, occurred practically at the same pH range as with the 23. anthracis strains. The shrinking of the capsule and the light scattering of these bacilli is very pro-
nounced on the more acid side of the reacting zone. In comparing micro-
photographs, figs. 14 and 15 of P1. 2, showing the specific capsular reactions with the polypeptide and with the polysaccharide substances of the capsules, with the microphotographs, figs. 11 to 13 of P1.2, on the non-specific capsular reactions with proteins, the similarity with the polypeptide reaction is apparent. Later in this paper it will be shown that a purified polysaccharide isolated from these bacteria does not give a precipitation reaction with pro- teins a t an appropriate pH range as the polypeptide does.
Effect of dzyerent proteins on the non-speciJic capsular reaction The pH zones in which the non-specific capsular reaction occurs with bovine serum remain essentially the same when horse or human serum is used. Human serum has an advantage inasmuch as precipitation and agglutination, which
Reaction of bacterial capsules mith proteins 108 The concentration of proteins used in the Tables 2 and 3 was 0.05y0 (w/v). With some of the proteins, as for example casein and haemoglobin, 0 - 0 0 5 ~ 0 (w/v) may elicit a good capsular reaction, whereas with others, such as albumin, less than 0-05y0 (w/v) gave an indistinct reaction. Of the proteins studied, only pepsin and trypsin gave a completely negative reaction with each of the capsulated organisms. Gelatin, bovine plasma albumin fraction no. V and egg albumin gave only a weak reaction with the Bacillus species. All the other non-basic proteins used reacted well with each of the capsulated micro-organisms. A correlation between the isoelectric point of the proteins and the shifting of the reacting pH zone was observed with every micro-organism when using horse serum albumin, globulin or haemoglobin. The higher the isoelectric point of the protein, the higher the pH zone a t which the optimal non- specific capsular reaction occurred. The behaviour of some of the capsulated organisms towards lysozyme and protamine confirmed this observation. When a reaction occurred at all with these basic proteins, the reacting zone was extended up to weak alkaline reaction. A visible capsular reaction with basic proteins was found only with the Bacillus species. A peculiar feature of this reaction was the contraction of the bacterial capsule and the increased refractivity of the bacilli (Pl. 1, fig. 6). Only with the reaction of B. anthracis with lysozyme could the capsule be demonstrated with almost normal appear- ance at pH 6 and 7, when the mixture was incubated several hours a t room
temperature (P1.c 1, fig. 7). The microscopic examination of B. megateriurn
and Bacillus M had to be carried out at this pH very quickly with lysozyme,
since these micro-organisms were dissolved very rapidly even a t room tem- perature by a 0.05 yo (wlv) solution of purified lysozyme. Another characteristic feature of the reaction with basic proteins and some of the bacteria was in the deformation of the bacterial cell. The strongest deformation was observed in the reaction of protamine on Bacillus M. A visibly turbid suspension of these bacilli was killed completely when left for 3 hr. at room temperature, pH 9, with 0.1-0-025y0 (w/v) protamine solutions. A marked deformation of the bacterial bodies was observed under these circumstances down to 0-006 yo (wlv) concentration of protamine. The deformation was very much less distinct with micro-organisms not belonging to the genus Bacillus. The correlation of the pH zone of the non-specific reaction with the iso- electric point of the protein is best seen when the upper pH limit giving the strongest reaction is taken into consideration, as can be seen in Table 4. The pH values given in this table were established with electrometric measure- ments. The data given in Table 4 suggest that these proteins might form electrostatic compounds with the bacterial capsule a t an inter-isoelectric pH value. The possibility of the occurrence of other binding forces cannot be excluded since some proteins, e.g. gelatin and ovalbumin, react differently with different capsules. Furthermore, casein, a protein which gives strong
capsular reactions, reacts in a pH zone which corresponds rather to the
globulin than to the albumin zone, despite the fact that its isoelectric point
104 J. Tomcsik and S. Guex-Holxer is identical with that of the albumin. Differences in the reaction towards different micro-organisms are most distinct when using basic proteins such as lysozyme and protamine.
Table 4. Upper pH limit of the capsular reaction with proteins of diflerent isoelectric points Isoelectric Cryptococcus Bacillus Protein point neoformans anthracis Albumin (horse serum) 4.9 3.0 4. Haemoglobin 6-8 4.4 5. Lysozyme 9.5 c. 6 c. 7
Protamine 10-12^ -
Globulin (horse serum) 5.4 3.2 4.
c. 8
Precipitation of the capsular substance with proteins at acid reaction As pointed out previously, several proteins react at pH values on the acid side of their isoelectric points with the capsule of various micro-organisms,. thereby making the capsule visible. Depending on the pH value in the reaction zone, the capsule appears either contracted, normal, or in some cases swollen. The reaction is sometimes homogeneous and extends to the whole depth of the capsule, as in the case of albumin (Pl. 1, fig. 2). In other cases, most frequently with casein and with haemoglobin and in lesser degree with globulin, the reaction appears usually as a precipitation at the surface of the capsule (Pl. 1, fig. 5; P1. 2, fig. 18). Thus microscopic examjnation gives the impression that the essential feature of the reaction consists of a precipitation. For the study of the precipitation of capsular substances with proteins
Bacillus M was selected, because this organism, like B. megaterium, contains
polypeptide as well as polysaccharide in its capsule (Tomcsik & Guex-Holzer, 1951, 1952). Three fractions were obtained from this organism: a D-glutamic acid polypeptide, a polysaccharide and a mucoprotein. The separation of the
D-glutamic acid polypeptide from the polysaccharide was good. It was found
that by mixing one volume each of polypeptide and albumin solution with eight volumes of acetate buffer O-2y0 polypeptide had to be added to 1-2y albumin to obtain the largest amount of precipitate in a pH zone where the buffer solutions had values from pH 3.3 to 4.2. On increasing the pH of the buffer to 4.4, a turbidity but no precipitation occurred. Less concentrated polypeptide required also 5 to 10 times the quantity of albumin to produce the greatest turbidity ; the precipitation was, however, not complete. The albumin used for the acid precipitation of the capsular substances was bovine plasma albumin fraction V. Protamine precipitated the polypeptide even at pH 7. The purified polysaccharide was isolated by hot-water extraction from the capsulated bacilli ; after hydrolysis it showed glucosamine and galactosamine, a trace of phosphorus, but no glutamic or uronic acids. This polysaccharide absorbed the antibodies from the homologous serum, revealing the charac- teristic capsular structure (Pl. 2, fig. 15) completely, but it gave no precipita- tion whatsoever with albumin a t acid reactions.
(^106) J. Tomcsik and S. Guex-Holxer
a pneumococcus C-polysaccharideprecipitation discovered by Tillett &z Francis (1930). A non-specific ‘quellung’ of several types of pneumococci at pH 4.0 by a variety of proteins was recognized for the first time by Jacox (1947). We assume that the principle underlying the ‘capsular swelling reaction ’ of pneumococci described by Jacox is identical with the principle of the non- specific capsular reaction described with various micro-organismsin this work ; nevertheless, the following differences should be pointed out. Jacox found that the reaction was inhibited by even as low a concentration of sodium chloride as 0.02 M. We generally used 0.85 yo saline, corresponding to 0.145 M-sodium chloride, for the solution or dilution of the proteins and for the suspension of bacteria. We observed only in the case of Cryptococcw neoformans and the pneumococci that the capsular reaction with albumin was favoured when the
NaCl concentration was decreased to 0. 0 7 ~. No such effect was observed,
however, when using globulin, haemoglobin or other proteins. According to Jacox the non-specific capsular reaction occurs only with pneumococci. He found no reaction with Klebsiella pneumoniae, a mucoid Escherichia coli, or a mucoid Streptococcus haemolyticus, although acid agglutination was observed under these conditions. We could demonstrate the non-specificcapsular reaction with all species of capsulated organisms studied, when using the appopriate pH range and with a number of different proteins. Jacox described the non-specific capsular swelling of pneumococci a t pH 4.0. He did not find a difference in the reaction pH zone with albumin or haerno-
globin. By using pH intervals of at least 0.4 unit, as we did, it is easy to
demonstrate that the reaction pH zones of albumin and of haemoglobin are different and also vary according to the organism tested. To demonstrate the difference between the reaction zones of such proteins as albumin and globulin, whose isoelectric points lay nearer to each other, pH differences of 0.2 should be used. The only protein which, in the work of Jacox, showed a higher pH
than 4.0 for eliciting the capsular swelling of pneumococci, was a 3-phospho-
glyceraldehyde dehydrogenase preparation. This protein has an isoelectric point a t pH 6.55 according to Cori, Slein & Cori (1948). This finding of Jacox fits well with our conception according to which the higher the isoelectric point of the protein, the higher the pH zone at which the non-specific capsular
reaction occurs. It would be interesting to see whether the phenomenon
observed by Lofstrom might not also be correlated with a high isoelectric point of the acute phase protein. Jacox assumes that in the absence of ionized salts, the pneumococcus polysaccharide-protein gel becomes increasingly hydrophilic and therefore capable of enlarging to produce the ‘quellung’ reaction. We have shown that the non-specific capsular reaction at different pH zones depends as much on the isoelectric point of the proteins used as on the species of organism which
produces polysaccharide or polypeptide capsular substance. We assume that
the reacting portions of the molecules of the capsular polysaccharides or of the polypeptide are, for instance, uronic acids or D-glutamic acid, leading in an inter-isoelectric zone to the formation of salt-like compounds with proteins ;
Reaction of bacterial capsules with proteins^107
the reaction is essentially a precipitation. The capsule itself may be contracted, of unchanged volume or, occasionally, swollen, depending on the pH value. Unfortunately we were not able to carry out electrophoretic determinations on the bacteria used, to determine their isoelectric points. It may be assumed that the isoelectric point of the bacterial surface is lower when the optimal capsular reaction occurs at lower pH. In consequence it should be lower with Cryptococcus neoformans and with Klebsiella pneurnoniae than with Bacillus anthracis. We found no data in the literature about the isoelectric point of capsulated Cryptococcus rheoformans or Bacillus anthracis. Harden & Harris (1953) point out that misleading conclusions can result from attempts to interpret information regarding cellular isoelectric points from data obtained by the dye method. They found by micro-electrophoretic measurements that the isoelectric point of Klebsiella pneumoniae is 2-48 and of Bacillus anthracis 3.1, i.e. the reverse of the values found by dye methods.
This work was carried out with the help of the Rockefeller Foundation.
REFERENCES ALDERTON,G. & FEVOLD,H. L. (1946). Direct crystallization of lysozyme from egg white and some crystalline salts of lysozyme. J. biol. Chem. 164, 1. AMBROSE, E. J. & HANBY,W. E. (1949). Evidence of chain folding in a synthetic polypeptide and in keratin. Nature, Lond. 163, 483. BODON,G. & TOMCSIK,J. (1934). Effect of specific antibody on the capsule of anthrax bacilli. Proc. SOC. exp. Biol., N. Y. 32, 122. CORI, G. T., SLEIN,M. W. & CORI,C. F. (1948). Crystalline ~-glyceraldehyde-3- phosphate dehydrogenase from rabbit muscle. J. biol. Chem. 173, 605. DROUHET, E., SEGRETAIN,G. & AUBERT,J. P. (1950). Polyoside capsulaire d'un champignon pathoghe Torulopsis neoformans ;relation avec la virulence. Ann. Inst. Pasteur, 79, 891. EVANS,E. E. & MEHL,J. W. (1951). A qualitative analysis of capsular polysac- charides from Cryptococcus neoformans by filter paper chromatography. Science, 114, 10. EVANS, E. E. & THERIAULT,R. J. (1953). The antigenic composition of Cryptococcus neoformans. J. Bact. 65, 571. GLADSTONE, G. P. & FILDES,P. (1940). A simple culture medium for general use without meat extract or peptone. Brit. J. exp. Path. 21, 161. HANBY,W. E. & RYDON,H. N. (1946). The capsular substance of Bacillus anthracis. Biochem. J. 40, 297. HANBY,W. E., WALEY,S. G. & WATSON,J. (1948). Synthetic polyglutamk acid. Nature, Lond. 161, 132. HARDEN, V. P. & HARRIS,J. 0. (1953). The isoelectric point of bacterial cells. J. B a t. 65, 198. HOUSEWRIGHT, R. D. & THORNE, C. B. (1950). Synthesis of glutamic acid and glutamyl polypeptide by Bacillus anthracis. J. Bact. 60, 89. IVANOVICS,G. & BRUCKNER,G. (1937). Chemische und immunologische Studien uber den Mechanismus der Milzbrandinfektion und Immunitat. 2. Irnmun- Forsch. 90, 304. IVANOVICS,G. & ERDOS,L. (1937). Ein Beitrag zum Wesen der Kapselsubstanz des Milzbrandbazillus. 2. ImmunForsch. 90, 5. JACOX,R. F. (1947). A new method for the production of non-specific capsular swelling of the pneumococcus. Proc. SOC. exp. B i d , N. Y. 66, 635.
- J. TOMCSIK & s. GUEX-HOLZER-REACTIONO F RACTERIi\L CAPSULES WITH PROTEINS. PLATE
- Journal of General Microbiology, Vol. 10, No.
- J. TOMCSIK & s. G U E X - H o I , z ~ R - - K E A C T I O N O F BACTERIAL CAPSULES WITH PROTISINS. PLATE
