BURGERt - PNAS · 2005. 4. 22. · VOL. 56, 1966 MICROBIOLOGY: M. M. BURGER 913 TABLE 3 ANTIGENIC DETERMINANTS OFTEICHOIc ACIDS Maximal Source and repeating unit of _ Hapten Inhibitors

TEICHOIC ACIDS: ANTIGENIC DETERMINANTS, CHAINSEPARATION, AND THEIR LOCATION IN THE CELL WALL*

BY MAX M. BURGERtDEPARTMENT OF BIOLOGICAL CHEMISTRY, SCHOOL OF MEDICINE,

WASHINGTON UNIVERSITY, ST. LOUIS

Communicated by Maclyn McCarty, July 1, 1966

Teichoic acids are linear polymers of glycerophosphate or ribitolphosphate carry-ing sugar or D-alanine substituents.1 Their chain length varies between 8 and 50units and is dependent upon the isolation technique. Some occur in bacterial cellwalls where they can make up between 8 and 50 per cent of the wall's dry weight;some, the so-called "intracellular" ones, in an ill-defined area between cell wall andcell membranes. From the time of their discovery they were known to be anti-genic.2 Thus the group antigens of the streptococcal group D3 and the type anti-gens of Staphylococcus aureus were shown to be teichoic acids.4 Antisera, however,could be obtained only with whole cells, cell walls, or crude extracts; or they wereprobably a consequence of previous infections as was the case for human seraagainst Staphylococcus aureus.5 Pure teichoic acid preparations do not act as im-munogens.6' I

It will be shown that precipitation and the preparation of a particulate suspensionof teichoic acids render them immunogenic. Antisera thus obtained were used toelucidate the antigenic determinants of several teichoic acids. Furthermore, theyprovided the tool to answer the following two questions:

(1) Do teichoic acids from a single bacterial strain exist as a mixed population ofcompletely substituted and completely unsubstituted molecules, or is each moleculesubstituted in a quantitatively or qualitatively different manner?

(2) Can teichoic acids be located within the architecture of the bacterial cell wall?With regard to the first question, it has been almost tacitly assumed up to now

that different types of sugar substituents are present on the same polymer or, in thecase where only one type of sugar substituent occurs, that each molecule is onlypartially substituted. Polymers of the two organisms examined here (B. subtilis3610 and B. licheniformis) appeared to be mixtures of molecules which were eithercompletely devoid of substituents or carried a complete set of substituents. Nosingle macromolecule seemed to contain more than one type of sugar.

In answer to the second question, it was demonstrated that teichoic acids in theintact wall are not readily available to antibodies. It was also shown that treatmentof the cell walls with lysozyme increased the availability of antigenic sites, therebysuggesting that teichoic acids occur in a somewhat deeper layer of the wall.

Materials and Methods.-Antisera: Particulate suspensions of teichoic acids were obtained bythe slow addition at 40C of either 5 mg of cetylpyridinium chloride (10 mg/ml) or 5 mg methylatedbovine serum albumin (MBSA) (10 mg/ml) to 1 mg purified8 teichoic acid in 1.0 ml of 0.154 MNaCl. Various particle sizes were produced by altering both concentration of the reactants andtemperature of the reaction and selected with Millipore filters. After mixing the suspension withan equal volume of Freund's adjuvant, 1/3 of the resulting emulsion was injected intramuscularly,and the remainder in equal amounts into each footpad of a male New Zealand rabbit (7-9 lb).

910

Dow

nloa

ded

by g

uest

on

Janu

ary

6, 2

021

VOL. 56, 1966 MICROBIOLOGY: M. M. BURGER 911

This procedure was repeated on the 9th and 17th days following the first injection. Betweenthe 24th and 27th days the animals were bled from an ear vein and the sera were collected. Alarge sample of streptococcal group D antiserum was kindly made available by Marion Wood,Wellcome Laboratories, England.

In the experiments where whole walls or muramic acid-containing antigens were used, anti-bodies were purified by (NH4)2SO4 precipitation and absorption to a kieselguhr column previouslytreated with antigen. Elution with the hapten, however, gave only an 8% recovery.

Antigen and hapten preparations: Teichoic acids were extracted with a phenol procedure de-scribed earlier.8 Haptens were obtained by HF degradation" or alkali hydrolysis from polymerpreparations. Glucosyl glycerol and galactosyl glycerol are cross-contaminated to some degree(see Table 3).

Quantitative precipitations: Equal volumes of antigen (in 0.154 M NaCl) and serum (0.1 or0.2 ml, at least 24 hr old) were mixed and kept at room temperature for 1 hr. After 1-5 days at40C, the resulting precipitate was washed 3 times with 2.0 ml 0.154 M NaCl, and then dissolvedin 0.4 ml 0.1 N NaOH, and the absorption at 280 mi determined.10 Most teichoic acid prepara-tions contained less than 0.5% nucleic acids. The teichoic acids neither quenched nor contributedto the absorption coefficient of E ll59 = 13.3 for teichoic acid antibodies. In two cases (B.licheniformis and "intracellular" B. subtilis W-23) the polymers could be identified in the anti-body-antigen precipitate and the amounts of antigen recovered from the immunoprecipitate wereproportional to the protein content of the precipitate. Yields varied between 78 and 91%.

Results.-Immunogenicity: Purified teichoic acid preparations did not elicit theformation of precipitating antibodies in rabbits. Three different types of teichoicacids, variation in chain length, and the use of Freund's adjuvant did not change thenegative outcome.

Since teichoic acids in whole-wall preparations are immunogenic, it was concludedthat formation of particles containing these teichoic acids might induce immuno-genicity. Plescia et al.11 have succeeded in rendering nucleic acids immunogenicby precipitating them with MBSA. It seemed probable that teichoic acids whichpossess a phosphodiester backbone, as do nucleic acids, might also be precipitatedwith MBSA. In fact, precipitation was quantitative and the particulate prepara-tion elicited antibody formation in this complex form. Antibody titers were vari-able (Table 1) and on an average 4-6 times greater than those found for nucleicacids. Although only the complexed form of teichoic acids was capable of initiatingantibody formation, pure teichoic acids could be used as efficient boosters after 4-6months.The particle size of the complex turned out to be critical insofar as only a certain

size range gave rise to optimal responses (Table 2). Whether the upper size limitreflects the upper limit of phagocytosis by macrophages, and the lower the minimalsize for immunogenicity of these compounds, is not known.More variability in antibody response was observed with another complexing

agent: cetylpyridinium chloride. This detergent is a small molecular compounand, unlike MBSA, not immunogenic by itself. It is a powerful precipitating agentfor all teichoic acids. Although the highest antibody titers could be obtained withthis complex, it was less reliable in terms of reproducibility (Table 1). Optimalparticle size for cetylpyridinium chloride complexes was in the range found forXIBSA complexes. Over-all size of the particle seems, however, not to be the solefactor determining antibody response. Thus, trichloracetic acid-extracted teichoicacid, being about '/2 to 1/3 as large as phenol-extracted teichoic acid, gave rise tomuch lower titers (0.2 to 0.4 mg/ml), although the over-all particle size of the com-plex was optimal.

Dow

nloa

ded

by g

uest

on

Janu

ary

6, 2

021

912 MICROBIOLOGY: M. M. BURGER PROC. N. A. S.

TABLE 1ANTIBODY LEVELS OBTAINED WITH DIFFERENT COMPLEXING AGENTS

Precipitation Antibody titerAntigen with (mg/ml serum)

1. B. subtilis 3610 wall teichoic acid MBSA 1.012. B. licheniformis wall polymer MBSA 1.27

CPCl 2.773. B. subtih8 W-23 "intracellular" teichoic acid MBSA 0.65

MBSA 0.90CPC1 1.29CPC1 0.25

4. S. faecalis "intracellular" teichoic acid Commercial 0.52MBSA 0.88CPC1 0.46

5. B. subtilis W-23 wall teichoic CPC1 1.276. DNA salmon sperm MBSA 0.32

CPCI 0.087. RNA yeast MBSA 0.12

CPC1 0.00Optimal antibody precipitation was reached over a wide range of antigen concentration (4 ;sg-100 pg per

ml serum). MBSA = methylated bovine serum albumin; CPC1 = cetylpyridinium chloride.

Determinant groups: If substituted by sugars, the polyglycerophosphate back-bone of teichoic acids seemed not to be involved in determining immunologic speci-ficity, or only to a minor extent. Rather, polyglycerophosphate appeared to be acarrier for the sugar constituents which turned out to be the chief antigenic deter-minants. However, in situations where the backbone is unsubstituted, McCarty hasshown that it can be antigenic.2 Whether purified polyglycerophosphate alone is animmunogen and not only an antigen, remains to be proved. In four cases examined(Table 3: 1, 3,4, and 5), D,IL-a-glycerophosphate did not act as a hapten inhibitor,nor could oligoglycerophosphate act as a precipitant or a hapten inhibitor. Thisfinding is probably a consequence of the complete sugar substitution of the poly-glycerophosphate backbone where the reactivity of the underlying glycerophosphatedeterminants is effectively shielded.12 McCarty has recently shown antigenicityof D-alanine substituents of teichoic acids.12 D-alanine, D,L-ethylalanine, andD,L-methylalanine at a concentration of 4 X 10-2 M did not inhibit any of ourantibody-precipitating systems. This finding may be a consequence of the highlability of alanine esters in teichoic acids, a type of substituent which might havebeen lost in our immunogen preparations.The most potent hapten inhibitors are shown in Table 3. They are indicative of

the antigenic-determinant sites. Unsubstituted sugars alone had in all cases verylow inhibitory capacities. Except for antibodies to B. subtilis W-23 wall teichoicacid, unphosphorylated compounds were more effective than phosphorylatedones. The structure of S. faecalis teichoic acid, i.e., the streptococcal group D

antigen, has been elucidated.'3 TheTABLE 2 structure of B. subtilis W-23 "intra-

DEPENDENcE OF ANTIBODY LEVEL OBTAINED cellular" teichoic acid is not completelyON THE PARTICLE SIZE OF IMMUNOGEN

Particle size Antibody titer) established. The occurrence of glu-(pA) (mg/ml serum cosaminyl substituents'4 is, however,0.45-0.8 0.6-0.90.8-1.2 1.6-1.9 supported by cross-reactivity of the1. -3.0 0.9O1.3 antiserum with hyaluronic acid and3.0-10.0 0.40.8Above 10 0. 0- 09 by strong inhibition of precipitation byAntibody titers are ranges obtained from three rabbits. chitobiose.

Dow

nloa

ded

by g

uest

on

Janu

ary

6, 2

021

VOL. 56, 1966 MICROBIOLOGY: M. M. BURGER 913

TABLE 3ANTIGENIC DETERMINANTS OF TEICHOIc ACIDS

MaximalSource and repeating unit of _ Hapten Inhibitors inhibition

teichoic acid Type Concentration (%)1. B. 8ubtilis 3610 2(a-D-glucosyl)

glycerophosphate 2(ct-D-glucosyl) glycerol 3 X 10-2 813(a-D-glucosyl) glycerol 3 X 10-2 8

2. B. lichniformis3(fl-D-galactosyl) glycerophosphate 3(fl-D-galactosyl) glycerol 1.3 X 10-2 39

together with3(c-D-glucosyl) glycerophosphate 3(a-D-glucosyl) glycerol 1.4 X 10-2 54

3. B. subtilis W-23 "intracellular"2(N-acetylglucosaminyl) glycero- 2(N-acetylglucosaminyl) 6.6 X 10-3 71phosphate glycerol

L-a-glycerophosphate, 6.6 X 10-3 0glucosamine, and N-acetylglucosamine

4. S. faecalis2(kojibiosyl) glycerophosphate 2(kojitriosyl) glycerol

and and 1 X 10-2 842(kojitriosyl) glycerophosphate 2(kojibiosyl) glycerol

5. B. subtilis W-23 wall4(D-glucosyl) ribitol-1(2)phos- 4(D-glucosyl) ribitol-1(2) 4.5 X 10-2 81phate phosphate

Inhibition, %, is given at optimal antibody/antigen ratios and at optimal hapten concentration. Haptens incases 1, 3, and 4 were obtained by HF degradation and purified by chromatography on charcoal celite columns.9Glucosaminyl glycerol contained about 15% glucosyl ribitol, galactosyl glycerol about 5% glucosyl glycerol, andglucosyl glycerol about 20% galactosyl glycerol. Partialls purified glycosylated glycerol derivatives from B.subtiliu 3610 and B. licheniformis were a gift from Dr. L. Glaser.

Within a series of possible haptenic inhibitors against B. subtilis W-23 wall tei-choic acids, glucosylribitol phosphate was found to be the most powerful one.Details will be reported later.15Chain separation: Using human antisera, Kabat's group has shown that chains of

Staphylococcus aureus teichoic acids carry either a-N-acetylglucosamine substituentsor the 13-anomeric form, but not both together on the same backbone.5

Antisera against B. subtilis 3610 wall teichoic acid precipitated only completelyglucosylated polymers (Table 4) from a teichoic acid preparation containing a ratioof glucose to glycerophosphate of less than 1.9 All unsubstituted glycerophosphatepolymers were left in solution. Precipitated material was completely substitutedwith glucose as demonstrated by its absolute alkali resistance.9 HF degradation9gave rise to 2(a-D-glucosyl) glycerol. B. subtilis 3610 teichoic acid from membranes,or so-called "intracellular" teichoic acid, is therefore a mixture of unsubstitutedglycerophosphate polymers and completely glucosylated polyglycerophosphate ashas been suggested earlier.9

Recently it had been shown that B. subtilis W-23 wall teichoic acid consists of amixture of polyribitolphosphate and completely glucosylated polyribitolphos-phate.'5 On the grounds of both of these findings it might be predicted that some of

TABLE 4CHAIN SEPARATION OF B. subtilis 3610 "INTRACELLULAR" TEICHOIC ACID

Glycosylation ofPolyglycerophosphate

Polymer examined A BTeichoic acid before precipitation 35 47Precipitated with antiserum 96 101Remaining in supernatant 5 2

Glucose and glycerophosphate were enzymatically determined after acid hydrolysis. Yields variedfrom 78 to 95%. A and B are two different teichoic acid preparations.

Dow

nloa

ded

by g

uest

on

Janu

ary

6, 2

021

914 MICROBIOLOGY: M. M. BURGER PROC. N. A. S.

the so-called partially substituted teichoic acids are mixtures of nonsubstituted andcompletely substituted polymers.Yet another case was found where a mixture of two types of related cell wall

polymers occurred in the same organism. Polymers can be extracted from B.licheniformis walls which resemble teichoic acids, but where the sugars are built intothe backbone itself.'6 They contain the following constituents: glycerophosphate/glucose/galactose/amino sugars (1:0.46:0.36:0.01). If such a preparation is in-jected into rabbits and the antibodies obtained are used to precipitate a polymerpreparation with a ratio of 1:0.49:0.47:0.31 (same order), the material precipitatedhad aratioof 1:0.52:0.48:0.19. Inthepresenceof 1.36 X 10-2Mgalactosylglycerolas a hapten inhibitor, galactose precipitation was inhibited 83 per cent, glucose 16per cent, while the presence of 1.46 X 10-2M glucosyl glycerol inhibited the precipita-tion of 76 per cent glucose polymer and 28 per cent galactose polymer. Galactoseand glucose appear thus to be located on different glycerophosphate polymers.Each bacterium carries both polymers since a wall extract from a cloned strain con-tained both types.The same two polymers were enzymatically synthesized in vitro and could be

separated in a similar fashion (Table 5). The occurrence of galactose and glucose ontwo different polymer chains in the biosynthetic material was suggested earlier,based on independence of the two enzymatic processes during a stepwise synthesis ofthe two polymers.'6In a control experiment, the presence of both hapten inhibitors did not suppress

precipitation of each sugar beyond the inhibition found with each hapten alone(Table 5, expt. 3). This is meant to indicate that even the residual amount of theglucose polymer precipitated in the presence of its glucosylglycerol inhibitor is notbound to a galactose polymer and vice versa, since the amount of glucose polymerprecipitated in the presence of galactosyl glycerol and glucosyl glycerol, i.e., bothinhibitors, would have to be lower.

Location of teichoic acids in cell walls: Of all possible antigenic sites of the wallteichoic acids in situ, only 19 per cent are reached by antibodies in B. subtilis 3610,while B. licheniformis walls are capable of binding 71 per cent of the theoreticalvalue (Fig. 1). Thus, most of the wall's teichoic acid of B. subtilis 3610 seems not tobe readily available on the surface and might be buried in a deeper layer or might besterically in an unfavorable situation for the incoming antibodies. This is furtherillustrated in Figure 1 by a gradual increase in antibody-antigen interaction duringlysozyme treatment.' B. licheniformis polymers, however, which are not teichoic

TABLE 5CHAIN SEPARATION OF ENZYMATICALLY PREPARED B. licheniformis WALL POLYMERS

Polymer Mixture Polymer RecoveredInhibitor (Hapten) Added as Antigen in Precipitate

Expt. Type Concentration (M) Type Cpm Cpm Per cent

1. 3(G-D-galactosyl) glycerol 1.35 X 10-2 Galactose 1350 599 37Glucose 1610 1485 92

2. 3(ox-D-glucosyl) glycerol 1.2 X 10-2 Galactose 1350 1065 81Glucose 1610 510 31

3. 3(j3-D-galactosyl) glycerol 1.7 X 10-2 Galactose 1350 607 45plus plus

3(a-D-glucosyl) glycerol 1.3 X 10-2 Glucose 1610 632 39The polymers were enzymatically synthesized from tritiated and C14-labeled precursors and isolated together

by phenolextraction before immunoprecipitation. Hapten concentrations chosen are at, or beyond, the opti-mally inhibitory ones.

Dow

nloa

ded

by g

uest

on

Janu

ary

6, 2

021

VOL. 56, 1966 MICROBIOLOGY: M. M. BURGER 915

FIG. 1.-Availability of antigenic sites in Bacillus cell walls.B. licheniformis 1.6-mg walls (containing 15% polymers) wereincubated with 13.5 mg purified antibodies against B. licheni-

2Q0r O formis wall polymers, and 2-mg B. subtilis 3610 walls (contain-

B subtilis 3610 ing 33% teichoic acid) with 8.8-mg purified antibodies againstED75 ~ 75/ B. subtilis 3610 wall teichoic acid. Optimal antibody precipi-

tation was reached with 25 Ag B. licheniformis polymer in the-6 500 / 50 SB licheniformis presence of 1.39-mg antibodies in 1 ml and with 75 jog B.- [; subtilis 3610 teichoic acid in a solution of 1.0-mg antibodies°25 25r in 1 ml. After incubation for 2 hr and removal of the walls,

the unabsorbed antibodies remaining in the supernatant were0L 3 6 9 0 2 4 determined by titration with antigen. The percentage of

theoretical binding sites occupied is defined as 100 times theMinutes after addition of Lysozyme ratio of antibodies actually bound by the walls versus the

amount the teichoic acid contained by this wall would havebound if it were in solution at the equivalence point.

acids, as was mentioned earlier, have most of their antigenic sites on the wall sur-face, i.e., easily available to their antibodies. A second case of a true teichoic acidwas examined, therefore, to find out whether this different location within the wallmight fall into a more general pattern. B. subtilis W-23 wall teichoic acid turnedout to be mostly buried underneath the wall surface and unavailable to its anti-bodies as was the case for B. subtilis 3610 teichoic acid. A similar conclusion wasreached independently in a study on the glucosylation of B. subtilis W-23 walls.Only 20 per cent of the teichoic acid backbone content of B. subtilis W-23 walls wasable to act as acceptor for glucose when the intact wall was used as a substrate. 18In a wall preparation of B. subtilis 3610 which still contained large amounts of

plasma membranes and probably parts of cytoplasm on the inside of the walls(verified with electron micrographs), the amount of antibodies adsorbed to thewalls was identical to that adsorbed to pure walls after trypsinization and repeatedwashings. It is thus conceivable that teichoic acids occur only on the external sideof the cell and not on the side adjoining the cytoplasmic membrane. Support forthis notion comes from the fact that whole B. subtilis 3610 cells absorb only 10 percent less teichoic acid antibodies than corresponding amounts of pure wall prepara-tions.

Discussion.-Teichoic acids can be rendered immunogenic by complexing themwith a cationic precipitating agent. MBSA and cetylpyridinium chloride werefound to be two such promoting agents. Although the size of the particles in sus-pension could be shown to be critical for eliciting antibodies, proper size appearsnot to be a sufficient requirement since polyvinylamine and spermidine, despitebeing complete precipitating agents for teichoic acids, did not support immunogeni-city, regardless of the size of the complex achieved.

It was suggested that MBSA conveys immunogenicity to nucleic acids by protect-ing them from degradation by nucleases in the challenged animal." In fact, thy-mus DNA which was only partially complexed with MBSA could be degraded to acertain extent with DNase in vitro. Such a partially DNase-sensitive DNA turnedout to be a weaker immunogen than DNA completely covered up with MBSA. 19

In the case of teichoic acids, however, no enzyme is known which would be cap-able of degrading teichoic acids in animal tissues. The problem of the mechanismof action of these promoters of immunogenicity therefore remains an open one.During an attempt to separate glucose-containing from galactose-containing

chains in B. licheniformis wall polymers, amino sugars were coprecipitated, although

Dow

nloa

ded

by g

uest

on

Janu

ary

6, 2

021

916 MICROBIOLOGY: M. M. BURGER PROC. N. A. S.

the original polymer preparation used as an immunogen was practically devoid ofamino sugars. Amino sugars are neither part of the backbone, nor are they sub-stituents of this teichoic acidlike polymer. Together with muramic acid they arepart, however, of the mucopeptide structure of the wall. M\uramic acid was pre-viously found in preparations of these teichoic acidlike polymers in B. lichenifor-mis.8 Such a coprecipitation of amino sugars might thus be interpreted to meanthat parts of the cell wall mucopeptide are tightly-possibly covalently-bound tothe teichoic acidlike wall polymers.

Support for the concept that teichoic acids are localized deeper in the wall thanB. licheniformis wall polymers seems to come from comparisons of the immunogenicresponse of these polymers when whole walls are used for immunization. The rela-tive antibody yield based on the polymer content of the walls injected and correctedfor different yields of the soluble antigens was 2-3 times higher for the B. licheni-formis wall polymer than for B. subtilis 3610 teichoic acid. This result is not neces-sarily a predicted one since differences in availability of these polymers are expectedto disappear as soon as the walls get into contact with the tissue lysozyme. Ourcurrent knowledge of the process of antigen uptake and recognition by the immuno-genic system is, however, so poor that this type of reasoning still belongs to the realmof speculation.The fact that antigenic teichoic acids can occur deep in a multilayered wall struc-

ture, as is suggested here, and at the same time display reactivity with antibodiesshould not be too much of a concern. Thus, it is well established that Streptococcusfaecalis teichoic acid-the group D specific antigen-does not occur within the cellwall at all but in an ill-defined area beneath the wall, and yet this location does notprevent it from reacting with its antibodies. The mucopeptide architecture hastherefore to be thought of as being a highly porous network or else as a sheet withlarge-sized discontinuities.Summary. Purified teichoic acids were rendered immunogenic and their anti-

genically active groups were determined and found to be the sugar substituents ofthe polymer backbone. Antisera with known specificities made it possible to estab-lish the fact that teichoic acids are probably mixtures of completely glycosylatedand completely unglycosylated chains. Partially glycosylated polymers could notbe detected. Polymers similar to teichoic acids containing more than one type ofsugar are also mixtures of homogeneous polymers, each molecule carrying only onetype of sugar.Polymers in the cell wall were found to be available to their antibodies to different

degrees. This is thought to reflect their relative depth of location in the cell wall.

The author is grateful for discussions with Drs. L. Glaser, H. N. Eisen, C. F. Cori, and A. B.Pardee.

* This work was supported by the St. Louis Heart Association and the National Science Foun-dation (grant GB 1033).

t Fellow of the Helen Hay Whitney Foundation. Present address: Program in BiochemicalSciences, Princeton University, Princeton, New Jersey 08540.

1 Baddiley, J., Federation Proc., 21, 1084 (1962).2 McCarty, M., J. Exptl. Med., 109, 361 (1959).3 Wicken, A. J., S. D. Elliot, and J. Baddiley, J. Gen. Microbiol., 31, 231 (1963).4 Juergens, W. G., A. R. Sanderson, and J. L. Strominger, J. Exptl. Med., 117, 925 (1963).

Dow

nloa

ded

by g

uest

on

Janu

ary

6, 2

021

VOL. 56, 1966 MICROBIOLOGY: M. M. BURGER 017

6 Torii, M., E. A. Kabat, and A. E. Bezer, J. Exptl. Med., 120, 13 (1964).6 Julianelle, L. A., and C. M. Wieghard, J. Exptl. Med., 62, 31 (1935).7 Haukenes, G., Aca Pathol. Microbiol. Scand., 55, 450 (1962).8 Burger, M. M., and L. Glaser, J. Biol. Chem., 239, 3168 (1964).9 Glaser, L., and M. M. Burger, J. Biol. Chem., 239, 3187 (1964).10Eisen, H. N., J. Immunol., 60, 77 (1948).11 Plescia, O., W. Braun, and N. C. Palczuk, these PROCEEDINGS, 52, 279 (1964).12 McCarty, M., these PROCEEDINGS, 52, 259 (1964).3Wicken, A. J., and J. Baddiley, Biochem. J., 87, 54 (1963).

14 Glaser, L., and M. M. Burger, unpublished observation.16 Chin, T., M. M. Burger, and L. Glaser, Arch. Biochem. Biophys., in press.16 Burger, M. M., and L. Glaser, J. Biol. Chem., 241, 494 (1966).17 This increase is not due to a release of teichoic acid into soluble form, since even after 4 min

lysozyme treatment, not more than 5% of the total teichoic-acid content of the wall was found insoluble form. The action of lysozyme within the first 4 min consists presumably in loosening therigid mucopeptide network.

18 Chin, T., and L. Glaser, in preparation.19 Burger, M. M., unpublished observation

Dow

nloa

ded

by g

uest

on

Janu

ary

6, 2

021