Aldolases of the lactic acid bacteria

Arch. Microbiol. 110, 121-I28 (1976)

Archives of

Hicrnbiology �9 by Springer-Verlag 1976

Aldolases of the Lactic Acid Bacteria Demonstration of Immunological Relationships among Eight Genera of Gram Positive Bacteria Using an Anti-Pediococcal Aldolase Serum

JACK LONDON and NINA M. CHACE

Laboratory of Microbiology and Immunology, National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland 20014, U.S.A.

Abstract. Reciprocal qualitative and quantitative immunological experiments employing an anti-Pedio- coccus cerevisiae aldolase serum confirmed many of the interspecific relationships demonstrated previously among lactic acid bacteria with antisera prepared against the Streptococcusfaecalis fructose diphosphate aldolase. The extent of immunological relatedness observed between the Lactobacillus and Pediococcus aldolases was markedly greater than that noted between Pediococcus and Streptococcus aldolases indicating that the pediococci share closer phylogenetic ties with the rod-shaped lactobacilli than with their spherical counterparts in the streptococci. In addi- tion to confirming the existence of definitive, but distant, relationships between the lactic acid bacteria and certain gram positive nonsporeforming anaerobes, immunological cross-reactivity was also demonstrated between the pediococcal aldolases and those of Aerococcus viridans.

Key words: Fructose diphosphate aldolases - Phylo- genetic relationships - Lactic acid bacteria.

In two previous reports (London and Kline, 1973; London et al., 1975) certain phylogenetic relations between streptococci, lactobacilli, pediococci, eubacteria, propionibacteria and Butyribacterium rettgeri were inferred from an immunological study of their fructose diphosphate (FDP) aldolases using an anti- Streptococcus faecalis aldolase serum as a probe. A tentative phylogenetic map was constructed in which the positions of the various species were arranged according to the degree of immunological homology

* This paper is dedicated with deepest appreciation to Prof. Roger Y. Stanier on the occasion of his 60th birthday in token of what his friendship and guidance have meant to me. - J. L.

shared by their aldolase and the reference S. faecalis aldolase. Relying on a single protein marker in pre- paring this map posed several problems, not the least of which was the uncertainty in the positioning of the less-related species according to the antigenic similarities of the respective aldolases. Also, in this particular antigen-antibody system, the variations in the size of the immunologically reactive aldolases (London, 1974) could have conceivably distorted the immunological results of the immunodiffusion and microcomplement fixation experiments. It was imperative, therefore, that the results obtained with the relatively small S.faecalis aldolase (M. W. = 56000) be confirmed by using another, larger enzyme as a second reference.

To this end, the aldolase of Pediococcus cerevisiae NCDO 559 (M.W. = 176000) was purified to electrophoretic homogeneity (London, 1974) and used as an antigen to raise enzyme-specific antibodies in rabbits. The ensuing report describes the results of a sero- logical comparison ofaldolases from representatives of the genera Lactobacillus, Streptococcus, Pediococcus, Aerococcus, Eubacteriurn, Butyribacterium and Pro- pionibacteriurn using this anti-P.cerevisiae aldolase serum. The data from these experiments are integrated with those obtained with the S.faecalis aldolase as reference to provide a more accurate picture of the natural relationships among members of the above named genera.

MATERIALS AND METHODS Microorganisms. A list of the microorganisms used in these experiments is provided in Table 1. Streptococcus faecalis strain ATCC 27792 (formerly strain MR) was obtained from M. Rogosa (National Institute of Dental Research); Aerococcus viridans strains 779 and 784 were kindly provided by Dr. O. Mundt (Uni- versity of Tennessee); Pediococcus cerevisiae strains NCDO 559 and 990 were made available to us by Dr. E. Garvie (National Institute for Dairying Research). Strains ofEubacterium cylindroides were obtained from Dr. L. Holdeman (Virginia Polytechnic In-

122

Streptococcus: S. faecalis ATCC 27792 S. lactis ATCC 19435

Lactobacillus: L. acidophilus ATCC 19992 L. bulgaricus ATCC 11842 L. casei ssp. casei 64H L. casei ssp. aMctosus OC45 L. coryniformis ssp. coryniformis M34 L. r ssp. torquens M30 L. deIbrueekii ATCC 9649 L. tactis ATCC 123'15 L. leichmannii ATCC 4797 L. salivarius ATCC 11741 L. xylosus ATCC 15577

Aerococcus: A. viridans ATCC 11563 A. viridans 779 A. viridans 784

Pediococeus: P. acidilactici ATCC 25740 P. acidilactici ATCC 25741 P. aeidilaetici ATCC 25742 P. acidilactici ATCC 25743 P. cerevisiae 559 P. cerevisiae 990 P. cerevisiae ATCC 8042 P. pentosaceus ATCC 25744 P. parvulus ATCC 13371

Propionibacterium : P. acnes ATCC 6919 P. arabinosum ATCC 4965 P. intermedium ATCC 14072 P. pentosaceum ATCC 4875 P. peterssonii ATCC 4870

Eubacterium : E. aerofaciens ATCC 25986 E. cylindroides 3594 E. limosum ATCC 8486

Others: Butyribacterium rettgeri ATCC 10852 Microbacterium thermosphactum ATCC 11609

Arch. Microbiol., Vol. 110 (1976)

Table 1 Strain designation of microorganisms

stitute) and Butyribacterium rettgeri was obtained from Dr. C. Wittenberger (National Institute of Dental Research). All other strains were obtained from the American Type Culture Collection.

Cultivation of the Microorganisms. With the exception of A. viridans the media used to cultivate the various bacterial strains and the conditions of growth have been cited elsewhere (London and Kline, 1973). The three strains of Aerococeus were maintained in trypticase soy broth (Difco Labs., personal communication, Dr. O. Mundt). Cultures were grown at 30~ in test tubes with loosely fitted screw caps for 24 h and stored at 8 ~ C. Batch cultures were grown in 1 1 Erlenmeyer flasks containing 500 ml of trypticase soy broth and harvested after 12 - 16 h of incubation at 30 ~ C.

Biochemical Determinations and Preparation of Cell Extracts. The preparation of cell-free extracts by ultrasonic disruption and the conditions of their storage have been described previously (London et al., t975). Aldolase activity in cell-free extracts was measured spectrophotometrically according to the procedure of Groves et al. (1966). Protein concentrations of cell-free extracts were determined by the biuret method (Gornall et al., 1949).

Immunological Procedures. The procedure for purifying the P. cerevisiae strain 559 aldolase to electrophoretic homogeneity and a biochemical characterization of the enzyme have been described (London, 1974). High titer anti-aldolase serum was raised in three 6-month old male Australian white rabbits by a series of four weekly intradermal injections of a suspension containing 0.353 mg P. cerevisiae aldolase (0.117 mg per rabbit); 20 p-g methylated bovine serum albumin; 0.6 ml of complete Freund's adjuvant and phos- phate buffered (0.02 M) normal saline to a final volume of 1.0 ml. Each rabbit received 0.5 ml of this suspension. Five days after the final injection, the rabbits were bled from the central ear artery and between 40 and 50 ml of blood were removed; the rabbits were bled again at day 10 and 15. After overnight storage at 5~ the clotted erythrocytes and fibrin were separated from the plasma by centrifugation and the clarified serum was stored as 5 ml afiquots at - 40 ~ C.

Immunodiffusion experiments were carried out according to the Stollar and Levine (1963) modification of the Ouchterlony technique. Details of these experiments are presented elsewhere (London and Kline, t973). The convention of Gasser and Gasser (1971) was used

to summarize immunodiffusion results; the interpretation of these results has been discussed previously (London and Kline, 1973).

A previously described immunoelectrophoresis procedure (London et al., 1971) was used to confirm the homogeneity of the pediococcal aldolase preparation and to determine the specificity of the antiserum. Figure 1 shows an example of a typical plate; 5 p-1 of purified pediococcal aldolase (6 p-g) were loaded into the upper well while the lower well received 5 p-1 of crude cell-free extract (30 ~tg of protein); following electrophoresis, 30 gl of anti- pediococcal aldolase serum placed into the center trough and the plate was incubated at 4 ~ C overnight. The appearance of a single precipitin line in both the upper and lower portion of the slide indicates that the antigen was flee of contaminating protein and that the antiserum did not contain extraneous cross reacting material. Sera taken from the rabbits prior to immunization produced no precipitate when substituted for the antisera in immunoelectrophoresis experiments.

Microcomplement fixation experiments were carried out according to the procedure of Wasserman and Levine (1961); minor modifications of this technique have been published elsewhere (London et al., 1971). The method of Hook and Muschel (1964) was used to titrate guinea pig complement. Anti-pediococcal aldolase sera was used at a dilution of 1:25000 and 1:35000 with the homologous antigen. Cell-free extracts of P. cerevisiae NCDO 559 containing the homologous antigen were used in the range of 1 to 10 p-g of protein while heterologous antigens were used at concentrations ranging from 2 to 40 gg of protein. Since the comparative amounts of complement fixed by both homologous and heterologous antigens were linear and parallel functions of the logarithm of the antibody concentration (Fig.2), results from these experiments were directly comparable (Sarich and Wilson, 1966). The index of dissimilarity (I.D.) for each of the heterologous aldolases was calculated from these data using the equation of Champion et al. (1974). For the pediococcal aldolase system, the term m of their equation equaled 137. To permit a comparison of results from the aldolase system with immunological data from other protein sys- tems, results of the microcomplement fixation experiments were also expressed as immunological distance units where the immunological distance = log I.D. x 100 (Prager and Wilson, 1971; Champion et al., 1975).

J. London and N. M. Chace: Aldolases of Lactic Acid Bacteria J23

Fig. 1. Amido black protein stain of immunoelectrophoresis plate of aldose : anti-aldolase complex. See text for details

100

'=- 80

60

40

20

100

L. _ bu!ga

A. v#idans

F 500 1000 I 101000 5000

ANTISERUM DILUTION

i 50000

Fig. 2. Relation of percent complement fixed to log of the antibody concentration

RESULTS

Differentiation of the Aldolases by Immunodiffusion Experiments

The aldolases of all available strains of Pediococcus were compared with each other by multiple cross- matches. Figure 3 shows several examples of such cross-matching. The three strains of P. cerevisiae and the single strain of Pediocoecus pentosaceus produced confluent precipitin lines with each other (Fig. 3A) indicating antigenic identity. The extract of P. cerevisiae NCDO 559 which contained the homologous antigen, produced faint spurs against the wells containing extracts ofPediococcus acidilactici (Fig. 3 B). This pattern of partial identity established that the aldolases of the latter do not share complete immuno- logic homology with the former. In a cross-match between the P. acidilactici and P. parvulus aldolases (not shown) the former produced large, distinct spurs against the latter. Hence, the P. parvulus aldolase is

the most immunologically distant enzyme in this set; these results agree with those of the anti-Strepto- coccus faecalis aldolase study (London and Kline, 1973; London et al., 1975). The aldolases of the three strains of Aerococcus viridans examined failed to react with the anti-S, faecalis aldolase serum but did react with the anti-pediococcal aldolase serum. Since the Aerococcus aldolases were of the same molecular weight as the P. parvulus enzyme (approximately 60000, unpublished data) they were cross-matched with one another. The A. viridans aldolases were found to be less related to the reference P. cerevisiae aldolase than the P. parvulus enzyme. Results of these cross-matches are summarized in Figure 4 and es- tablish the following order of relatedness; P. eere- visiae ~ P. pentosaceus > P. acidilactici > P. parvulus > Aerococcus viridans.

The anti-pediococcal aldolase sera reacted weakly with the aldolases of Lancefield group D and group N streptococci (not shown) and not at all with extracts of representatives of the other Lancefield groups; the two reacting groups could not be distinguished from each other by the antisera. In marked contrast t o these results, a number of Lactobacillus aldolases reacted strongly with the anti-pediococcal aldolase sera. Aldolases of Lactobacillus coryniformis ssp. coryniformis and ssp. torquens produced minor spurs against the aldolases of L. casei ssp. casei and L. casei ssp. rhamnosus (Fig. 5A); no infraspecific differences were observed. With the exception of the P. acidi- lactiei aldolases, extracts of the above mentioned Lactobacitlus species produced the strongest interspecific cross-reactions observed in this series of experiments. Aldolases of the L. leichmannii group (which include L. lactis, L. delbrueckii and L. bul-

124 Arch. Microbiol., Vol. 110 (1976)

Fig. 3A and B. Results of immunodiffusion experiments comparing pediococcal atdolases. (A) Sample wells contain extracts (approximately 0.8 units enzyme activity) of the following Pediococcus species: (1) P. cerevisiae 559; (2) P. cerevisiae 990; (3) P. pentosaceus 25744; (4) P. cerevisiae 8042. The center well contains 15 gl anti-P, cerevisiae 559 aldolase serum. (B) Sample wells contain extract of the following Pediococcus species: (1) P. acidilactici 25740; (2) P. cerevisiae 559; (3) P. acidilactici 25741 ; (4) P. acidilactici 25742; (5) P. acidilactici 25743. The center well contains, 215 gl anti-P, cerevisiae aldolase serum

P. cerevisiae

t

t

t

Fig. 4~

P, pentosaceus

P. acidilactic/

l P. parvulus

t ......... viridans

Hierarchical ordering of pediococcal and aerococcal aldolases as determined with anti-P, cerevisiae aldolase serum. Arrows indicate the dominant antigen in the paired cross-match; = sign indicates confluent precipitates (reaction of identity)

garicus) form a group of apparent identical specificity (Fig. 5B) but are less immunologically related to the reference enzymethan is the L. salivarius aldolase. The L. acidophilus aldolase produced verY faint precipitates when reacted with the anti-pediococcal aldolase serum (not shown) similar to those produced by the streptococci. A summation of the data is shown in Figure 6. With one minor difference to be discussed later, the ordering of 'the species is the same as that observed with the anti-streptococcal aldolase serum (London and Kline, 1973).

The structural homologies between the aldolases of the lactic acid bacteria and those of the anaerobic eubacteria and propionibacteria as demonstrated with

the anti-streptococcal aldolase sera (London et al., 1975) were confirmed with anti-pediococcal aldolase sera. Crossmatches with extracts of Propionibacterium petersonii, P. pentosaceum, P. intermedium, P. acnes and P. arabinosum gave faint but discernable reactions of apparent identity with the anti-pediococcal aldolase serum. A reaction of apparent identity was also observed between Eubacterium cylindroides and Eu- bacterium limosum; however, both produced spurs against the aldolase of Eubacterium aerofaciens.

Quantitative Immunological Studies

While the titers of the antipediococcal aldolase sera are significantly lower than those of the antistreptococcal aldolase sera, their strength was sufficient to react with a number of key heterologous cell-free extracts to provide another spatial reference point for mapping. The results of the microcomplement fixation studies are summarized in Table 2 and can be compared with the antistreptococcal aldolase results which are also shown. Unfortunately, a direct comparison of the numerical values cannot be made because no at tempt was made to demonstrate reciprocity among the different antisera (Champion et al., 1975). However, the relative distances between related species approximate the degree of relatedness. A comparison of the immunological distance values tend to confirm the immunodiffusion results. With the exception of the P. cerevisiae 8042 I.D. value which is not significantly different from that of P. acidilactici, the pediococci can be divided into three distinct immunological groups. The minor antigenic differences observed between strains of L. coryniformis and L. casei with

J. London and N. M. Chace: Aldolases of Lactic Acid Bacteria 125

Fig. 5 A and B. Results of immunodiffusion experiments comparing Lactobacillus aldolases. (A) Sample wells contain extracts (between 0.3-0.5 units of aldolase activity) of the following Lactobacillus species: (1) L. casei 64H; (2) L. coryniformis ssp. coryniformis M34; (3) L. casei ssp. rhamnosus OC-45. Center well contains 25 gl anti-P, cerevisiae aldolase serum. (B) Sample wells contain extracts of the following Lactobacillus species (between 0.3-0.5 units aldolase activity): (1) L. salivarius 11741 ; (2) L. delbrueck ii 9649; (3) L. bulgaricus 11842; (4) L. lactis 12315; (5) L. leichmannii 4797. Center well contains 35 gl anti-P, cerevisiae aldolase serum

immunodiffusion are not reflected in their I.D. values and the two species cluster in a single immunological group. Of greater significance is the fact that the general ordering of the Lactobacillus aldolases is the same with both immunological procedures. But the two most intriguing observations are (1) that the pediococci are more closely related to certain lactobacilli than to the streptococci, and (2) that they are also more closely related to the propionibacteria than are the streptococci.

D I S C U S S I O N

The original phylogenetic map (London and Kline, 1973) which at tempted to describe the natural relationships among the homofermentat ive lactic acid bacteria was based on immunological studies carried out with a single reference protein, the Strepto-

coccus faecalis F D P aldolase. Although evolutionary divergence between the streptococci, pediococci and lactobacilli was apparent f rom the reactions of non- identity (double-spurred precipitin reactions) observed in multiple immunodiffusion cross-matches, an accurate placement of the diverging lines of lactic acid bacteria with respect to one another was impossible with only a single immunological distance value. Accord- ingly, the various lines were arranged in an arbitrary fashion. The microcomplement fixation studies with anti-Pedicoccus cerevisiae aldolase serum reported here have provided a second set of immunological distance values which were used to prepare a two dimensional representation of the antigenic relationships of the lactic acid bacterial F D P aldolases (Fig. 7). On this map, the species were aligned to an

L, coryniform~

Z. case/ssp. COSO[

t Fig. 6.

L. casei ssp. rhamnosus

l L. sahvarius

t L. delbr

l l L. bu/gaficu,~

Hierarchical ordering of Lactobacillus species as determined with anti-pediococcal aldolase serum. Arrows denote the dominant antigen in the paired cross-match; = sign indicates confluent precipitates (reaction of identity)

optimal two-point fit using a scale of 1 m m per I.D. unit (for the original drawing). With a few exceptions, the aldolases of most species could be comfortably positioned by the pair of I.D. values; only the aldolases of the Lactobacillus leichmannii and L. acidophilus are shown out of alignment. In accordance with the anti-pediococcal aldolase data, these species should fall slightly below and to the right of the positions shown on the map. The three most distal streptococcal clusters and the L . j u g u r t i - L . helveticus group are shown in their original positions (London and Kline, 1973) because the cross-reactions between their aldolases and the anti-pediococcal aldolase sera were too


Table 2. Quantitative estimation of antigenic homology of certain aldolases by microcomplement fixation

Organism Antigenic similarity determined with

anti-P, cerevisiae aldolase anti-S, faecalis aldolase

Index of dissimilarity Immunological distance Index of dissimilarity Immunological distance

Pediococcus P. cerevisiae 559 1 P. cerevisiae 990 1.13 P. cerevisiae 8042 1.49 P. pentosaceus 25744 1.21 P. acidilactici 25740 1.64 P. acidilactici 25742 1.65 P. parvulus 19371 16.3 A. viridans 11563 33.2 A. viridans 783 45.5 A. viridans 779 44.4 S. faecal& MR 51.5 S. lactis 19435 57.2 L. casei 64H 5.4 L. casei v. rhamnosus OC91 6.0 L. coryniformis M34 6.9 L. salivarius 12315 28 L. bulgaricus 11842 28.8 L. xylosus 15572 41 L. acidophilus 19992 58 M. thermosphactum 51 B. rettgeri 42

Propionibacterium P. pentosaceum 4875 25 P. intermedium 10472 28 P. arabinosum 4965 28 P. aenes 6919 44.7

0 /9.6 129 5.3 20 130

17 19.8 129 8.3 23 136

21 9.2 96 21 11.4 105

121 39 159 152 n.r. n.r. 165 n.r. n.r. 164 n.r. n.r. 171 1.0 0 175 8.3 91 73 4.8 68 78.4 5.4 73 84 3.6 56

144 30 147 145 38 158 161 20 130 176 176 224 170 11.7 106 162 7.75 89

139 212 232 143 190 229 144 N.D. N.D. 165 191 230

n.r. = no reaction

weak to obtain reliable I.D. values. Similarly, the position of the Aerococcus aldolases are single point assign- ments since the anti-streptococcal sera failed to react with them.

Immunological reactions between the pediococcal aldolases and the anti-streptococcal and anti-pediococcal aldolase sera, respectively, demonstrate that this group of enzymes can be segregated into three distinct antigenic groups. Moreover, the relative extent of relatedness between the three antigenic clusters is roughly the same with both sets of antisera in immunodiffusion experiments.

The compact clustering of the three phenospecies, P. cerevisiae, P. pentosaceus and P. acidilactici indicate that these taxa have not diverged significantly from one another. In fact, our results indicate the species P. pentosaceus may be identical to P. cerevisiae; DNA hybridization studies should eventually resolve this question. However, if the P. parvulus aldolase is aligned on the map according to its I.D. values this species is pulled out of the pediococcal line to a position slightly closer to the streptococci (Fig. 7).

Whether this particular placement is due to an ex- perimental artifact resulting from the substantial size differences in the respective antigens (all other pediococcal aldolases have a M.W. of 176000 as contrasted with the 56-60000 M.W. streptococcal and P. parvulus enzymes) or actually reflects an evolutionary event independent and separate from the onto- geny of the main pediococcal cluster cannot be as- certained yet. The aerococci have tentatively been placed in line with P. parvulus simply because their enzymes belong to the same molecular weight group as th e latter.

In a previous report (London and Kline,/973) we speculated that the Lancefield group D streptococci might be more closely related to some species of lactobacilli than to certain species of streptococci. These speculations were based on the comparatively high degree of immunological homology observed between the aldolases of the S.faecalis and L. casei. The results obtained with anti-pediococcal aldolase sera reveal that the majority of the pediococci are more closely related to members of the subgenus Strepto-

J. London and N. M. Chace: Aldolases of Lactic Acid Bacteria 127

Fig. 7 A composite relatedness map based on immunological data obtained with anti-Streptococcusfaecalis and anti- P. cerevisiae aldolase sera

q A. viridans

J GroupN Streptococci

~ S . bovis-S.sanguis group e1"~S. #ai~varius group

_S.pyoqenes group

I Pr~176 rium

cerevisiae 990

ocidiloctlci L. ocidophilus

L. leichmannii group

L.cose__.A~ \\ \ \ \ \ \ \ \ \L._. salivorius

L_.. xylosus

M. thermosphoct urn

L. helvet

E__. cylindr oide s

E~ oerof o ciens

bacterium of the lactobacilli than to their spherical counterparts in the streptococci.

If these data are subsequently confirmed by in- vestigations with other sets of enzymes, a strong argu- ment can then be made for the genesis of a group of spherical bacteria from a rod-shaped progenitor instead of a more distantly related coccal form.

A comparison of the ordering of the species be- longing to the genus Lactobacillus by the two sets of anti-aldolase sera reveal that they are essentially identical (see Table 2). Therefore, it appears that the size differences among the five quaternary forms of the enzyme did not appreciably affect the immunological procedures used to measure the structural similarities between the aldolases. The single exception, namely, the P. parvulus aldolase, has been discussed above. Moreover, the ordering of this genus is roughly similar to that observed by Gasser and Gasser (1971) using D- and L-lactate dehydrogenases as evolutionary markers.

Several final comments about the phylogenetic map are in order here. First, the map shown in Figure 7 is the product of studies carrier out with a single set of enzymes, the FDP aldolases. As such, the natural relationships depicted in the map reflect the evolutionary history of the enzyme which is a function of its amino acid substitution rate and while the map re-

quires verification with other sets of enzymes, the proteins selected for this purpose should exhibit the same rate of evolutionary change. Second, the point on the map which serves as a juncture for the various divergent lines of evolution and represents the common ancestor, only stations the lines or clusters in two dimensions. However, none of the lines are actually fixed because they can be rotated in a third dimension about the junction point and still maintain their correct alignment to each other. Additional reference enzymes will be needed to con- struct maps in which the clusters or evolutionary lines have been locked into a multidimensional configura- tion.

REFERENCES

Champion, A. B., Prager, E. M., Wachter, D., Wilson, A. C.: Microcomplement fixation. In: Biochemical and immunological taxonomy of animals and plants (C. A. Wright, ed.), pp. 397-416. London: Academic Press 1974

Champion, A. B., Soderberg, K. L., Wilson, A. C., Ambler, R. P. : Immunological comparison of Azurins of known amino acid sequence. Dependence of cross-reactivity upon sequence resemblance. J. Mol. Evol. 5, 291-305 (1975)

Gasser, F., Gasser, C. : Immunological relationships among lactic dehydrogenases in the genera Lactobacillus and Leuconostoc. J. Bact. 106, 113-125 (1971)


Gornall, A. G., Bardawill, C. J., David, M. IVI. : Determination of serum proteins by means of the biuret reaction. J. biol. Chem. 177, 751 --766 (1949)

Groves, W. E., Calder, J., Rutter, W. J. : Fructose diphosphate. II. Clostridium perfringens. In: Methods in enzymology, Vol. 9 (S. P. Colowick, N. O. Kaplan, eds.), pp. 486-491. New York: Academic Press 1966

Hook, W. A., Muschel, L. H. : Anticomplementary effect and complement activity of human sera. Proc. Soc. exp. Biol. (N.Y.) 117, 292 - 297 (1964)

London, J. : Variations in the quaternary structure of three lactic acid bacteria aldolases. Evidence for the existence of a Class I and Class II aldolase in Lactobacillus casei. J. biol. Chem. 249, 7977- 7983 (1974)

London, J., Chace, N. M., Kline, K. : Aldolases of lactic acid bacteria: immunological relationships among aldolases of streptococci and gram-positive nonsporeforming anaerobes. Inter. J. syst. Bacteriol. 25, 114-123 (1975)

London, J., Kline, K. : Aldolase of lactic acid bacteria: a case history in the use of an enzyme as an evolutionary marker. Bacteriol. Rev. 37, 453-478 (1973)

London, J., Meyer, E. Y., Kulczyk, S. : Comparative biochemical and immunological study of malic enzyme from two species of lactic acid bacteria: Evolutionary implications. J. Bact. 106, 126- 137 (1971)

Prager, E. M., Wilson, A. C. : The dependence of immunological crossreactivity upon sequence resemblance among lysozymes. II. Comparison of precipitin and microcomplement fixation results. J. biol. Chem. 246, 7010-7017 (1971)

Sarich, V. M., Wilson, A. C.: Quantitative immunochemistry and the evolution of primate albumins: microcomplement fixation. Science 154, 1563-1566 (1966)

Stollar, D., Levine, L.: Two-dimensional immunodiffusion. In: Methods in enzymology (S. P. Colowick, N. O. Kaplan, eds.), Vol. VI, pp. 848-854. New York: Academic Press 1963

Wasserman, E., Levine, L. : Quantitative micro-complement fixation and its use in the study of antigenic structure by specific antigen-antibody inhibition. J. Immunol. 87, 290-295 (1961)

Received March 1, 1976

Documents

Aldolases of the lactic acid bacteria