JOURNAL OF BACTERIOLOGY, Apr. 1977, p. 393-398 Copyright ©) 1977 American Society for Microbiology
Vol. 130, No. 1 Printed in U.S.A.
Homogeneity of Lipopolysaccharides from Acholeplasma PAUL F. SMITH
Department of Microbiology, School of Medicine, University of South Dakota, Vermillion, South Dakota 57069 Received for publication 19 January 1977
Five methods were employed to determine the heterogeneity or homogeneity of lipopolysaccharides from four acholeplasmal species, Acholeplasma axanthum, A. granularum, A. laidlawii, and A. modicum. A. axanthum lipopolysaccharide behaved as a single component in all tests. A. granularum exhibited two components of identical composition and antigenic specificity. A. modicum lipopolysaccharide behaved as three components in two tests, but all three were similar in composition and identical serologically. The separable components of lipopolysaccharides from A. granularum and A. modicum probably represent size differences only. A. laidlawii lipopolysaccharide contained two distinct components by all methods. One was identified as the previously reported amino sugar polymer, whereas the other was a lipopolysaccharide containing both neutral and amino sugars.
We have reported the isolation of a new class of lipopolysaccharides from various members of the class Mollicutes, namely, Acholeplasma and Anaeroplasma species, Thermoplasma acidophilum, and Mycoplasma neurolyticum (7, 10). These lipopolysaccharides are membrane associated and can be extracted into the aqueous phase of hot 45% phenol. Nucleic acidfree preparations contain covalently bonded fatty acids, glycerol, and neutral and amino sugars. Amino and neutral sugars vary both qualitatively and quantitatively among different species. These lipopolysaccharides also serve as specific antigens in complement fixation and Ouchterlony gel diffusion tests using antisera prepared against homologous membranes (11). These membrane-associated lipopolysaccharides possess some of the physical characteristics of lipopolysaccharides from gram-negative bacteria (8, 10), some of which have been shown to be heterogeneous either as to size or chemical composition or both (4). This communication reports the results of studies to determine the heterogeneity or homogeneity of the lipopolysaccharides obtained from four species of Acholeplasma.
a 24-h culture (10%, vol/vol) of a given organism. After 24 h of stationary incubation at 37°C, the organisms were harvested, washed, and freeze-dried as described previously (10). Isolation and purification of the lipopolysaccharides from lipid-free cellular residues by hot aqueous phenol extraction, nuclease digestion, and gel filtration were performed as already documented (10). The freeze-dried products appeared as cottony white materials essentially free of protein, phosphorus, and absorbance at 260 nm. Five methods were employed to assess heterogeneity or homogeneity: permeation chromatography of deacylated lipopolysaccharides, permeation chromatography of the acetylated derivatives, permeation chromatography of the methylated derivatives, polyacrylamide gel electrophoresis in sodium dodecyl sulfate-containing gels, and Ouchterlony gel diffusion. Approximately 5-mg amounts of lipopolysaccharide were used for each analysis. Mild deacylation was carried out by incubating lipopolysaccharides in 0.5 ml of 1 N NaOH at 37°C for 10 min, followed by immediatepapplication to the chromatographic column. Peracetylation was accomplished by heating lipopolysaccharides at 1000C for 12 h in 1.0 ml of dry pyridine acetic anhydride (3:1, vol/vol) under nitrogen in Teflon-lined screw-capped tubes. The pyridine and acetic anhydride were removed by aeration with a stream of nitrogen, and the residue was dissolved in chloroform-methanol (1:1, vol/vol) for application to the column. Permethylation was MATERIALS AND METHODS achieved by reacting lipopolysaccharide twice with The organisms used in this study were Achole- the methyl sulfinyl radical in dimethyl sulfoxide plasma axanthum S743, A. granularum BTS39, A. according to the method of Hakamori (5). Excess modicum PG49, and A. laidlawii B. One hundred- reagents were removed by dialysis against deionized liter batches of medium, composed of tryptose water, and the resulting product was freeze-dried. (Difco), 2%, yeast extract (Difco), 0.5%, glucose, Permethylated lipopolysaccharides were applied to 0.5%, and NaCl, 0.5%, pH 7.8, were distributed in 1- chromatographic columns in chloroform-methanol liter quantities in 3-liter flasks and inoculated with (1:1, vol/vol). Assessment of complete acetylation 393
394
J. BACTERIOL.
SMITH
and methylation was made by determining the absence of free hydroxyl absorption in infrared spectra. These spectra were obtained on an instrument (Beckman model 18A) with films of appropriate lipopolysaccharides on NaCl crystals. Permeation chromatography was conducted exclusively on columns (1 by 57 cm) of controlled-poresize glass beads (Electro-Nucleonics, Inc., Fairfield, N.J.). Three pore sizes were used: CPG-10-75, with an operating range of 3 x 103 to 3 x 104; CPG-10-240, with an operating range of 2.2 x 104 to 1.2 x 106; and CPG-10-2000, with an operating range of 7 x 105 to 9 x 108. The use of different eluting solvents and the unknown absorption characteristics of the glass beads for the lipopolysaccharides and their derivatives precluded any accurate estimation of particle weights. Deionized water was used for elution of the deacylated lipopolysaccharides, and methanol served as eluting solvent for the peracetylated and permethylated derivatives. Void volumes were determined with blue dextran of molecular weight about 2 x 106, and bed volumes were determined with D-mannose or 2,3,4,6-tetramethylglucose. Two-milliliter fractions were collected; from these, known volumes were removed for neutral and amino sugar analyses. Prestained lipopolysaccharides were used for electrophoresis in sodium dodecyl sulfate-containing polyacrylamide gels (4). The lipopolysaccharides were dyed with Procion red MX-2B (ICI-United States, Inc., Wilmington, Del.) by the procedure of Anderson et al. (1). To 10 mg of lipopolysaccharide, dissolved in 1.0 ml of deionized water with the aid of sonication for 1 min in a sonic bath, was added 10 mg of dye contained in 1.0 ml of deionized water. After 5 min, 20 mg of solid NaCl was added, followed 30 min later by 2 mg of Na2CO3. After it had stood overnight, the clear solution was applied to a column (1 by 57 cm) of controlled-pore-size glass beads (CPG10-2000), and elution was carried out with water. The dyed lipopolysaccharide appeared in the void volume, whereas the unadsorbed dye eluted with the bed volume. The elutate comprising the void volume was freeze-dried. Gel electrophoresis was performed in gels (0.6 by 8.5 cm) of 5.6% polyacrylamide containing N,N'-methylenebisacrylamide (0.21%), ammonium persulfate (0.15%), N,NN,N'I"tetramethylethylenediamine (0.025%), and sodium dodecyl sulfate (1.0%) in tris(bhydroxymethyl) aminomethane (Tris) buffer (40 mM, pH 7.4), containing sodium acetate (20 mM) and ethylenediaminetetraacetic acid (2 mM) (2,4). The tracking dye was methylene blue (0.02%) in Tris buffer (10 mM, pH 8.0), containing ethylenediaminetetraacetic acid (1 mM). The disintegration reagent consisted of Tris buffer (10 mM, pH 8.0), containing ethylenediaminetetraacetic acid (1 mM), sodium dodecyl sulfate (4%), sucrose (30%), and dithiothreitol (2.5%). Before electrophoretic runs, 70 u.l of deionized water, 50 al of disintegration reagent, and 30 ,ul of tracking dye were added to 1 mg of dyed lipopolysaccharide, and the mixture was heated at 1000C for 5 min. After cooling, amounts varying from 20 to 50 ul were layered on top of the gels. Electrophoresis was conducted in Tris buffer (40 mM, pH 7.4) containing
sodium acetate (20 mM) and ethylenediaminetetraacetic acid (2 mM), employing a constant current of 5 mA per gel. Current was interrupted after the tracking dye migrated 7 cm into the gel. Gels were scanned in a microdensitometer (Canalco model E). Ouchterlony diffusion analyses were performed in petri plates containing Ionagar no. 2 (0.5%) and NaCl (0.85%). The lipopolysaccharides were dissolved in sodium phosphate buffer (100 mM, pH 8.0), at a concentration of 1 mg/ml with and without the addition of Triton X-100 at a final concentration of 1%. Diffusion was allowed to occur at 4 to 100C in a moist chamber for periods of up to 10 days. Antisera were prepared in rabbits against acholeplasmal membranes or pure lipopolysaccharides as detailed previously (11). All other hydrolytic and analytical procedures have been documented (6, 7, 10).
RESULTS Deacylated lipopolysaccharides and the peracetylated and permethylated derivatives were subjected to permeation chromatography on all three pore-sized columns, CPG-10-75, CPG-10240, and CPG-10-2000. The data presented in Fig. 1 to 3 are restricted to elution patterns from columns that gave the best examples of
0.5
j
4 N
0.2.5
-PGo -240 Ck
050
A.X loidb:ii
O.50 ° Ufi
%CPG-10-2000 0.25
0
0I25 i
1.01
20
30
40
50
Volume, ml
FIG. 1. Permeation chromatography ofdeacylated lipopolysaccharides on columns of controlled-poresize glass beads. Elution was with deionized water. Symbols: (0) neutral sugar, (0) amino sugar. Arrows indicate void (left) and pore (right) volumes. OD, Optical density.
LIPOPOLYSACCHARIDES FROM ACHOLEPLASMA
VOL. 130, 1977
separations; i.e., they
were
selected for the
greatest degree of apparent heterogeneity. Figure 1 presents the results for the deacylated
lipopolysaccharides. Although only the neutral shown for all lipopolysaccharides except A. laidlawii, the amino sugar patterns were identical to the neutral sugar patterns for A. axanthum, A. granularum, and A. modicum. The deacylated lipopolysaccharide from A. axanthum yielded only one peak. Since this lipopolysaccharide contains 90% of its fatty acids in N-acyl linkages, the alkaline treatment would not remove these fatty acids. Hence, aggregation due to lipid components on this particular lipopolysaccharide would not be averted. Nevertheless, results with derivatized lipopolysaccharide confirm that this single peak is indicative of homogeneity. Two peaks were observed for the lipopolysaccharide from A. granularum. These were collected and subjected to compositional analyses. Both appeared to be qualitatively and quantitatively similar. One slight difference was noted with respect to the quantity of glucose and galactose (Table 1). However, the small amount of fraction II available for analysis casts doubt on the significance of these differences. The deacylated lipopolysaccharide from A. laidlawii exhibited two distinct peaks. Peak I contained both neutral and amino sugar, whereas peak II was essentially devoid of neutral sugars. Analyses of the sugar composition of these two peaks (Table 1) corroborated these findings. Peak I contained predominately neutral sugars and the deoxyhexosamines together with some hexosamines, whereas peak II consisted entirely of glucosamine and galactosamine. The deacylated lipopolysaccharide from A. modicum yielded three sugar patterns are
395
peaks for neutral and amino sugars. These, too, were collected and subjected to sugar analyses. Qualitatively, all three were identical. Quantitatively, the neutral sugar content was the same, but some differences were noted in the deoxyhexosamine and hexosamine composition (Table 1). The results using peracetylated derivatives are exhibited in Fig. 2. Curves for both neutral and amino sugars are shown. The lipopolysaccharides from A. axanthum and A. granularum eluted as one peak. As with the deacylated lipopolysaccharide, the peracetylated derivative from A. laidlawii could be separated into two components: one contained neutral and amino sugars, and the smaller particle contained only amino sugars. The results for A. modicum are less clear, since the pattern obtained suggests distribution over a wide area of the columnn. Elution patterns for the permethylated derivatives are shown in Fig. 3. Permethylation interfered with the Elson-Morgan assay for amino sugars, reducing the sensitivity drastically. Therefore, the amino sugar pattern is shown only for the lipopolysaccharide from A. modicum since this material exhibited a rather diffuse pattern with the peracetylated derivative. All permethylated lipopolysaccharides exhibited a single peak by permeation chromatography. A second peak containing only amino sugars was noted with A. laidlawii, but is not apparent because only the neutral sugar pattern is drawn. Figure 4 shows the densitometric tracings of sodium dodecyl sulfate-polyacrylamide gel patterns of the predyed lipopolysaccharides. In all cases, the lipopolysaccharides entered the gel
TABLE 1. Neutral and amino sugar composition of fractions of deacylated lipopolysaccharides separated by permeation chromatography mol% of sugar in: Sugar
Neutral Glucose Galactose Mannose Amino Deoxyhexosamine Glucosamine Galactosamine
A. granularum a
II
I
54.2
39.9 60.1
97.8
45.8
A. modicum
A. laidlawii
II
I
II
III
5.8 94.2
6.2 93.8
1.0 99.0
32.3
46.8
68.5
67.7
51.9
30.4
2.2
53.0 47.0
2.1 Neutral: amino a See Fig. 1 for source of fractions.
48.3 41.7 1.9
17.3
6.9 76.5
33.4
39.8
60.2
1.98
1.91
1.86
396
SMITH
J. BACTERIOL. A. modlcun CPG-10-2000
I
0
0.10
~ ~ P2_9.00-0..
-I
SAS~~~~d 0.00_ 0
*
2 0 0
It)
0
20
30
40 Volume, ml
50
FIG. 2. Permeation chromatography of peracetylated lipopolysaccharides on columns of controlledpore-size glass beads. Elution was with methanol. Symbols: (0) neutral sugar, (0) amino sugar. Arrows indicate void (left) and pore (right) volumes. OD, Optical density.
completely. One band was observed with A. axanthum and A. modicum. Since only the neutral sugar-containing component of A. laidlawii lipopolysaccharide was subjected to electrophoresis, only one band appeared. The amino sugar polymer also exhibited one band that moved about 6 cm into the gel. The lipopolysaccharide from A. granularum produced two bands, a result similar to that obtained with the deacylated material on permeation chromatography. Ouchterlony gel diffusion analyses confirmed the results obtained by permeation chromatography and polyacrylamide gel electrophoresis. The results were identical whether or not intact or deacylated lipopolysaccharides, used as antigens in the diffusion analyses, were suspended in buffer or buffer containing 1% Triton X-100. Similarly, results were identical when antiserum was prepared against membranes or purified lipopolysaccharides. The lipopolysaccharides from A. axanthum and A. granularum each produced one sharp line of precipitation (Fig. 5). A. laidlawii exhibited two bands. When separated, component I (Fig. 1, Table 1) gave a line of identity with the more slowly diffusing component, whereas component II,
0 20
40 Volume, ml
30
50
FIG. 3. Permeation chromatography of permethylated lipopolysaccharides on columns of controlledpore-size glass beads. Elution with methanol. Symbols: (a) neutral sugar, (0) amino sugar. Arrows indicate void (left) and pore (right) volumes. OD, Optical density.
A-modicum
Ajaidlowii
A. gronulorum
A. oaxnhum _
I
0
I
2
_
3
_vM 6 7
_
4 cm
I
5
FIG. 4. Densitometric tracings of sodium dodecyl sulfate-polyacrylamide gel electrophoresis of lipopolysaccharides. Lipopolysaccharides were prestained with Procion red MX-2B.
VOL. 130, 1977
LIPOPOLYSACCHARIDES FROM ACHOLEPLASMA
~~~~~~~~~~~~~~~A
'Ia
397
1 3(-3WIcIS-
FIG. 5. Ouchterlony gel diffusion patterns of lipopolysaccharides or separated components or both versus specifwc antisera.
the amino sugar polymer, gave a line identity with the faster-moving component. The lipopolysaccharide from A. modicum produced three lines of precipitation. When the separated components I, II, and III (Fig. 1, Table 1) were employed as antigens, each gave only a single line, with all three fusing as one. Antiserum was absorbed with each of these components singly. Precipitates formed overnight at 4°C were removed by centrifugation. The supernatant serum was reabsorbed twice. Each of these three absorbed antisera was run against each of the three separated components as well as against unabsorbed sera. No precipitin lines appeared with any antigen against absorbed sera, whereas each gave a single line against the unabsorbed serum. These results indicated identical antigenic determinants on all three separated components. DISCUSSION The results of gel permeation chromatography of deacylated lipopolysaccharide and the
peracetylated and permethylated derivatives, sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and Ouchterlony gel diffusion suggest that the lipopolysaccharide from each of the four species of Acholeplasma exists as a single homogeneous entity. Deacylation of the lipopolysaccharide from A. axanthum does not remove a significant amount of the fatty acids that exist in N-acyl linkages. Therefore, the data for permeation chromatography of this product may be in error, since residual fatty acids would allow aggregation. However, all of the other methods that ensured disaggregation gave results confirming homogeneity. The lipopolysaccharide from A. granularum behaved as a mixture of two types of molecules on permeation chromatography of the deacylation product and in polyacrylamide gels. However, we conclude that only a single type of molecule exists, since the derivatives behaved as single components, only one precipitin line was seen in immunodiffusion analyses, and the composition of the two components separating upon deacylation exhibited almost identical results. The li-
398
SMITH
popolysaccharide from A. laidlawii proved to be contaminated with the amino sugar polymer described by Gilliam and Morowitz (3). Separation of the much larger lipopolysaccharide molecule from the smaller amino sugar polymer yielded a component homogeneous by all the methods employed. The deacylated lipopolysaccharide from A. modicum could be separated into three components of very similar sugar composition. Although these three components migrated at different rates in immunodiffusion plates, they proved to be antigenically identical. Permeation chromatography of derivatives and polyacrylamide gel electrophoresis suggested homogeneity. It is possible that this lipopolysaccharide is homogeneous as to structure but heterogeneous as to size. Any serious attempt to discern size of these molecules from the present data alone is fraught with error. Little is known of the adsorption characteristics of such molecules on the controlled-pore-size glass beads, and the diffusability of lipopolysaccharides in polyacrylamide gels complicates the electrophoretic migration. Nevertheless, if these problems are kept in mind, some speculation as to size can be made. Except for the lipopolysaccharide of A. axanthum, one can assume one diacylglycerol per molecule and calculate the apparent molecular weight. In the case of A. granularum this value is approximately 19,000, in good agreement with the permeation chromatography results, which suggest a molecular weight of about 20,000. In this regard might be mentioned the inability of this lipopolysaccharide to elicit antibody response in rabbits, in contrast to the other lipopolysaccharides (R. J. Lynn, personal communication), which could be explained as a lack of sufficient size. Size of bacterial lipopolysaccharides is a determining factor in the immune response of ,8-lymphocytes. Repeated antigenic determinants are required for multiple bonding to reactive B-cell receptors (9). The size of the lipopolysaccharide from A. laidlawii appears to be in the range of 100,000 to 150,000; that of A. axanthum seems to be about 100,000. The lipopolysaccharide from A. modicum has an apparent size of 36,000 based upon calculation from the content of diacylglycerol. This result is compatible with the smallest particle seen in the deacylated product
J. BACTERIOL.
on the gel permeation column. The remaining two components eluting from this column have apparent sizes of 72,000 and 142,000 and could be construed as dimer and tetramer, respectively. It must be emphasized that these values are only speculative but probably are within the proper order of magnitude. ACKNOWLEDGMENTS We are indebted to Leigh Ann Smith for performing the polyacrylamide gel electrophoresis and to W. R. Mayberry for the gas chromatographic analyses. The dye, Procion red MX-2B, was a generous gift from D. P. Raper, ICI-United States, Inc., Wilmington, Del. LITERATURE CITED 1. Anderson, M. W., A. Hendric, J. R. A. Millar, and A. C. Munroe. 1971. Electrophoresis of dyed polysaccharides on "phoroslides." Analyst 96:870-874. 2. Fairbanks, G., T. L. Steck, and D. F. H. Wallach. 1971. Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry 10:2606-2617. 3. Gilliam, J. M., and H. J. Morowitz. 1972. Characterization of the plasma membrane of Mycoplasma laidlawii. IX. Isolation and characterization of the membrane polyhexosamine. Biochim. Biophys. Acta 274:353-363. 4. Jann, B., K. Reske, and K. Jann. 1975. Heterogeneity of lipopolysaccharides. Analysis of polysaccharide chain lengths by sodium dodecylsulfate-polyacrylamide gel electrophoresis. Eur. J. Biochem. 60:239-246. 5. Lindberg, B. 1972. Methylation analysis of polysaccharides, p. 178-195. In S. P. Colowick and N. 0. Kaplan (ed.), Methods in enzymology, vol. 28. Academic Press Inc., New York. 6. Mayberry, W. R., P. F. Smith, and T. A. Langworthy. 1974. A heptose-containing pentaglycosyl diglyceride among the lipids of Acholeplasma modicum. J. Bacteriol. 118:898-904. 7. Mayberry-Carson, K. J., T. A. Langworthy, W. R. Mayberry, and P. F. Smith. 1974. A new class of lipopolysaccharide from Thermoplasma acidophilum. Biochim. Biophys. Acta 360:217-229. 8. Mayberry-Carson, K. J., I. L. Roth, and P. F. Smith. 1975. Ultrastructure of lipopolysaccharide isolated from Thermoplasma acidophilum. J. Bacteriol. 121:700-703. 9. Moller, G., 0. Sjoberg, and J. Andersson. 1973. Immunogenicity, tolerogenicity and mitogenicity of lipopolysaccharides, p. 44-48. In E. H. Kass and S. M. Wolff (ed.), Bacterial lipopolysaccharides. University of Chicago Press, Chicago. 10. Smith, P. F., T. A. Langworthy, and W. R. Mayberry. 1976. Distribution and composition of lipopolysaccharides from mycoplasmas. J. Bacteriol. 125:916-922. 11. Sugiyama, T., P. F. Smith, T. A. Langworthy, and W. R. Mayberry. 1974. Immunological analysis of glycolipids and lipopolysaccharides derived from various mycoplasmas. Infect. Immun. 10:1273-1279.
Vol. 130, No. 1 Printed in U.S.A.
Homogeneity of Lipopolysaccharides from Acholeplasma PAUL F. SMITH
Department of Microbiology, School of Medicine, University of South Dakota, Vermillion, South Dakota 57069 Received for publication 19 January 1977
Five methods were employed to determine the heterogeneity or homogeneity of lipopolysaccharides from four acholeplasmal species, Acholeplasma axanthum, A. granularum, A. laidlawii, and A. modicum. A. axanthum lipopolysaccharide behaved as a single component in all tests. A. granularum exhibited two components of identical composition and antigenic specificity. A. modicum lipopolysaccharide behaved as three components in two tests, but all three were similar in composition and identical serologically. The separable components of lipopolysaccharides from A. granularum and A. modicum probably represent size differences only. A. laidlawii lipopolysaccharide contained two distinct components by all methods. One was identified as the previously reported amino sugar polymer, whereas the other was a lipopolysaccharide containing both neutral and amino sugars.
We have reported the isolation of a new class of lipopolysaccharides from various members of the class Mollicutes, namely, Acholeplasma and Anaeroplasma species, Thermoplasma acidophilum, and Mycoplasma neurolyticum (7, 10). These lipopolysaccharides are membrane associated and can be extracted into the aqueous phase of hot 45% phenol. Nucleic acidfree preparations contain covalently bonded fatty acids, glycerol, and neutral and amino sugars. Amino and neutral sugars vary both qualitatively and quantitatively among different species. These lipopolysaccharides also serve as specific antigens in complement fixation and Ouchterlony gel diffusion tests using antisera prepared against homologous membranes (11). These membrane-associated lipopolysaccharides possess some of the physical characteristics of lipopolysaccharides from gram-negative bacteria (8, 10), some of which have been shown to be heterogeneous either as to size or chemical composition or both (4). This communication reports the results of studies to determine the heterogeneity or homogeneity of the lipopolysaccharides obtained from four species of Acholeplasma.
a 24-h culture (10%, vol/vol) of a given organism. After 24 h of stationary incubation at 37°C, the organisms were harvested, washed, and freeze-dried as described previously (10). Isolation and purification of the lipopolysaccharides from lipid-free cellular residues by hot aqueous phenol extraction, nuclease digestion, and gel filtration were performed as already documented (10). The freeze-dried products appeared as cottony white materials essentially free of protein, phosphorus, and absorbance at 260 nm. Five methods were employed to assess heterogeneity or homogeneity: permeation chromatography of deacylated lipopolysaccharides, permeation chromatography of the acetylated derivatives, permeation chromatography of the methylated derivatives, polyacrylamide gel electrophoresis in sodium dodecyl sulfate-containing gels, and Ouchterlony gel diffusion. Approximately 5-mg amounts of lipopolysaccharide were used for each analysis. Mild deacylation was carried out by incubating lipopolysaccharides in 0.5 ml of 1 N NaOH at 37°C for 10 min, followed by immediatepapplication to the chromatographic column. Peracetylation was accomplished by heating lipopolysaccharides at 1000C for 12 h in 1.0 ml of dry pyridine acetic anhydride (3:1, vol/vol) under nitrogen in Teflon-lined screw-capped tubes. The pyridine and acetic anhydride were removed by aeration with a stream of nitrogen, and the residue was dissolved in chloroform-methanol (1:1, vol/vol) for application to the column. Permethylation was MATERIALS AND METHODS achieved by reacting lipopolysaccharide twice with The organisms used in this study were Achole- the methyl sulfinyl radical in dimethyl sulfoxide plasma axanthum S743, A. granularum BTS39, A. according to the method of Hakamori (5). Excess modicum PG49, and A. laidlawii B. One hundred- reagents were removed by dialysis against deionized liter batches of medium, composed of tryptose water, and the resulting product was freeze-dried. (Difco), 2%, yeast extract (Difco), 0.5%, glucose, Permethylated lipopolysaccharides were applied to 0.5%, and NaCl, 0.5%, pH 7.8, were distributed in 1- chromatographic columns in chloroform-methanol liter quantities in 3-liter flasks and inoculated with (1:1, vol/vol). Assessment of complete acetylation 393
394
J. BACTERIOL.
SMITH
and methylation was made by determining the absence of free hydroxyl absorption in infrared spectra. These spectra were obtained on an instrument (Beckman model 18A) with films of appropriate lipopolysaccharides on NaCl crystals. Permeation chromatography was conducted exclusively on columns (1 by 57 cm) of controlled-poresize glass beads (Electro-Nucleonics, Inc., Fairfield, N.J.). Three pore sizes were used: CPG-10-75, with an operating range of 3 x 103 to 3 x 104; CPG-10-240, with an operating range of 2.2 x 104 to 1.2 x 106; and CPG-10-2000, with an operating range of 7 x 105 to 9 x 108. The use of different eluting solvents and the unknown absorption characteristics of the glass beads for the lipopolysaccharides and their derivatives precluded any accurate estimation of particle weights. Deionized water was used for elution of the deacylated lipopolysaccharides, and methanol served as eluting solvent for the peracetylated and permethylated derivatives. Void volumes were determined with blue dextran of molecular weight about 2 x 106, and bed volumes were determined with D-mannose or 2,3,4,6-tetramethylglucose. Two-milliliter fractions were collected; from these, known volumes were removed for neutral and amino sugar analyses. Prestained lipopolysaccharides were used for electrophoresis in sodium dodecyl sulfate-containing polyacrylamide gels (4). The lipopolysaccharides were dyed with Procion red MX-2B (ICI-United States, Inc., Wilmington, Del.) by the procedure of Anderson et al. (1). To 10 mg of lipopolysaccharide, dissolved in 1.0 ml of deionized water with the aid of sonication for 1 min in a sonic bath, was added 10 mg of dye contained in 1.0 ml of deionized water. After 5 min, 20 mg of solid NaCl was added, followed 30 min later by 2 mg of Na2CO3. After it had stood overnight, the clear solution was applied to a column (1 by 57 cm) of controlled-pore-size glass beads (CPG10-2000), and elution was carried out with water. The dyed lipopolysaccharide appeared in the void volume, whereas the unadsorbed dye eluted with the bed volume. The elutate comprising the void volume was freeze-dried. Gel electrophoresis was performed in gels (0.6 by 8.5 cm) of 5.6% polyacrylamide containing N,N'-methylenebisacrylamide (0.21%), ammonium persulfate (0.15%), N,NN,N'I"tetramethylethylenediamine (0.025%), and sodium dodecyl sulfate (1.0%) in tris(bhydroxymethyl) aminomethane (Tris) buffer (40 mM, pH 7.4), containing sodium acetate (20 mM) and ethylenediaminetetraacetic acid (2 mM) (2,4). The tracking dye was methylene blue (0.02%) in Tris buffer (10 mM, pH 8.0), containing ethylenediaminetetraacetic acid (1 mM). The disintegration reagent consisted of Tris buffer (10 mM, pH 8.0), containing ethylenediaminetetraacetic acid (1 mM), sodium dodecyl sulfate (4%), sucrose (30%), and dithiothreitol (2.5%). Before electrophoretic runs, 70 u.l of deionized water, 50 al of disintegration reagent, and 30 ,ul of tracking dye were added to 1 mg of dyed lipopolysaccharide, and the mixture was heated at 1000C for 5 min. After cooling, amounts varying from 20 to 50 ul were layered on top of the gels. Electrophoresis was conducted in Tris buffer (40 mM, pH 7.4) containing
sodium acetate (20 mM) and ethylenediaminetetraacetic acid (2 mM), employing a constant current of 5 mA per gel. Current was interrupted after the tracking dye migrated 7 cm into the gel. Gels were scanned in a microdensitometer (Canalco model E). Ouchterlony diffusion analyses were performed in petri plates containing Ionagar no. 2 (0.5%) and NaCl (0.85%). The lipopolysaccharides were dissolved in sodium phosphate buffer (100 mM, pH 8.0), at a concentration of 1 mg/ml with and without the addition of Triton X-100 at a final concentration of 1%. Diffusion was allowed to occur at 4 to 100C in a moist chamber for periods of up to 10 days. Antisera were prepared in rabbits against acholeplasmal membranes or pure lipopolysaccharides as detailed previously (11). All other hydrolytic and analytical procedures have been documented (6, 7, 10).
RESULTS Deacylated lipopolysaccharides and the peracetylated and permethylated derivatives were subjected to permeation chromatography on all three pore-sized columns, CPG-10-75, CPG-10240, and CPG-10-2000. The data presented in Fig. 1 to 3 are restricted to elution patterns from columns that gave the best examples of
0.5
j
4 N
0.2.5
-PGo -240 Ck
050
A.X loidb:ii
O.50 ° Ufi
%CPG-10-2000 0.25
0
0I25 i
1.01
20
30
40
50
Volume, ml
FIG. 1. Permeation chromatography ofdeacylated lipopolysaccharides on columns of controlled-poresize glass beads. Elution was with deionized water. Symbols: (0) neutral sugar, (0) amino sugar. Arrows indicate void (left) and pore (right) volumes. OD, Optical density.
LIPOPOLYSACCHARIDES FROM ACHOLEPLASMA
VOL. 130, 1977
separations; i.e., they
were
selected for the
greatest degree of apparent heterogeneity. Figure 1 presents the results for the deacylated
lipopolysaccharides. Although only the neutral shown for all lipopolysaccharides except A. laidlawii, the amino sugar patterns were identical to the neutral sugar patterns for A. axanthum, A. granularum, and A. modicum. The deacylated lipopolysaccharide from A. axanthum yielded only one peak. Since this lipopolysaccharide contains 90% of its fatty acids in N-acyl linkages, the alkaline treatment would not remove these fatty acids. Hence, aggregation due to lipid components on this particular lipopolysaccharide would not be averted. Nevertheless, results with derivatized lipopolysaccharide confirm that this single peak is indicative of homogeneity. Two peaks were observed for the lipopolysaccharide from A. granularum. These were collected and subjected to compositional analyses. Both appeared to be qualitatively and quantitatively similar. One slight difference was noted with respect to the quantity of glucose and galactose (Table 1). However, the small amount of fraction II available for analysis casts doubt on the significance of these differences. The deacylated lipopolysaccharide from A. laidlawii exhibited two distinct peaks. Peak I contained both neutral and amino sugar, whereas peak II was essentially devoid of neutral sugars. Analyses of the sugar composition of these two peaks (Table 1) corroborated these findings. Peak I contained predominately neutral sugars and the deoxyhexosamines together with some hexosamines, whereas peak II consisted entirely of glucosamine and galactosamine. The deacylated lipopolysaccharide from A. modicum yielded three sugar patterns are
395
peaks for neutral and amino sugars. These, too, were collected and subjected to sugar analyses. Qualitatively, all three were identical. Quantitatively, the neutral sugar content was the same, but some differences were noted in the deoxyhexosamine and hexosamine composition (Table 1). The results using peracetylated derivatives are exhibited in Fig. 2. Curves for both neutral and amino sugars are shown. The lipopolysaccharides from A. axanthum and A. granularum eluted as one peak. As with the deacylated lipopolysaccharide, the peracetylated derivative from A. laidlawii could be separated into two components: one contained neutral and amino sugars, and the smaller particle contained only amino sugars. The results for A. modicum are less clear, since the pattern obtained suggests distribution over a wide area of the columnn. Elution patterns for the permethylated derivatives are shown in Fig. 3. Permethylation interfered with the Elson-Morgan assay for amino sugars, reducing the sensitivity drastically. Therefore, the amino sugar pattern is shown only for the lipopolysaccharide from A. modicum since this material exhibited a rather diffuse pattern with the peracetylated derivative. All permethylated lipopolysaccharides exhibited a single peak by permeation chromatography. A second peak containing only amino sugars was noted with A. laidlawii, but is not apparent because only the neutral sugar pattern is drawn. Figure 4 shows the densitometric tracings of sodium dodecyl sulfate-polyacrylamide gel patterns of the predyed lipopolysaccharides. In all cases, the lipopolysaccharides entered the gel
TABLE 1. Neutral and amino sugar composition of fractions of deacylated lipopolysaccharides separated by permeation chromatography mol% of sugar in: Sugar
Neutral Glucose Galactose Mannose Amino Deoxyhexosamine Glucosamine Galactosamine
A. granularum a
II
I
54.2
39.9 60.1
97.8
45.8
A. modicum
A. laidlawii
II
I
II
III
5.8 94.2
6.2 93.8
1.0 99.0
32.3
46.8
68.5
67.7
51.9
30.4
2.2
53.0 47.0
2.1 Neutral: amino a See Fig. 1 for source of fractions.
48.3 41.7 1.9
17.3
6.9 76.5
33.4
39.8
60.2
1.98
1.91
1.86
396
SMITH
J. BACTERIOL. A. modlcun CPG-10-2000
I
0
0.10
~ ~ P2_9.00-0..
-I
SAS~~~~d 0.00_ 0
*
2 0 0
It)
0
20
30
40 Volume, ml
50
FIG. 2. Permeation chromatography of peracetylated lipopolysaccharides on columns of controlledpore-size glass beads. Elution was with methanol. Symbols: (0) neutral sugar, (0) amino sugar. Arrows indicate void (left) and pore (right) volumes. OD, Optical density.
completely. One band was observed with A. axanthum and A. modicum. Since only the neutral sugar-containing component of A. laidlawii lipopolysaccharide was subjected to electrophoresis, only one band appeared. The amino sugar polymer also exhibited one band that moved about 6 cm into the gel. The lipopolysaccharide from A. granularum produced two bands, a result similar to that obtained with the deacylated material on permeation chromatography. Ouchterlony gel diffusion analyses confirmed the results obtained by permeation chromatography and polyacrylamide gel electrophoresis. The results were identical whether or not intact or deacylated lipopolysaccharides, used as antigens in the diffusion analyses, were suspended in buffer or buffer containing 1% Triton X-100. Similarly, results were identical when antiserum was prepared against membranes or purified lipopolysaccharides. The lipopolysaccharides from A. axanthum and A. granularum each produced one sharp line of precipitation (Fig. 5). A. laidlawii exhibited two bands. When separated, component I (Fig. 1, Table 1) gave a line of identity with the more slowly diffusing component, whereas component II,
0 20
40 Volume, ml
30
50
FIG. 3. Permeation chromatography of permethylated lipopolysaccharides on columns of controlledpore-size glass beads. Elution with methanol. Symbols: (a) neutral sugar, (0) amino sugar. Arrows indicate void (left) and pore (right) volumes. OD, Optical density.
A-modicum
Ajaidlowii
A. gronulorum
A. oaxnhum _
I
0
I
2
_
3
_vM 6 7
_
4 cm
I
5
FIG. 4. Densitometric tracings of sodium dodecyl sulfate-polyacrylamide gel electrophoresis of lipopolysaccharides. Lipopolysaccharides were prestained with Procion red MX-2B.
VOL. 130, 1977
LIPOPOLYSACCHARIDES FROM ACHOLEPLASMA
~~~~~~~~~~~~~~~A
'Ia
397
1 3(-3WIcIS-
FIG. 5. Ouchterlony gel diffusion patterns of lipopolysaccharides or separated components or both versus specifwc antisera.
the amino sugar polymer, gave a line identity with the faster-moving component. The lipopolysaccharide from A. modicum produced three lines of precipitation. When the separated components I, II, and III (Fig. 1, Table 1) were employed as antigens, each gave only a single line, with all three fusing as one. Antiserum was absorbed with each of these components singly. Precipitates formed overnight at 4°C were removed by centrifugation. The supernatant serum was reabsorbed twice. Each of these three absorbed antisera was run against each of the three separated components as well as against unabsorbed sera. No precipitin lines appeared with any antigen against absorbed sera, whereas each gave a single line against the unabsorbed serum. These results indicated identical antigenic determinants on all three separated components. DISCUSSION The results of gel permeation chromatography of deacylated lipopolysaccharide and the
peracetylated and permethylated derivatives, sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and Ouchterlony gel diffusion suggest that the lipopolysaccharide from each of the four species of Acholeplasma exists as a single homogeneous entity. Deacylation of the lipopolysaccharide from A. axanthum does not remove a significant amount of the fatty acids that exist in N-acyl linkages. Therefore, the data for permeation chromatography of this product may be in error, since residual fatty acids would allow aggregation. However, all of the other methods that ensured disaggregation gave results confirming homogeneity. The lipopolysaccharide from A. granularum behaved as a mixture of two types of molecules on permeation chromatography of the deacylation product and in polyacrylamide gels. However, we conclude that only a single type of molecule exists, since the derivatives behaved as single components, only one precipitin line was seen in immunodiffusion analyses, and the composition of the two components separating upon deacylation exhibited almost identical results. The li-
398
SMITH
popolysaccharide from A. laidlawii proved to be contaminated with the amino sugar polymer described by Gilliam and Morowitz (3). Separation of the much larger lipopolysaccharide molecule from the smaller amino sugar polymer yielded a component homogeneous by all the methods employed. The deacylated lipopolysaccharide from A. modicum could be separated into three components of very similar sugar composition. Although these three components migrated at different rates in immunodiffusion plates, they proved to be antigenically identical. Permeation chromatography of derivatives and polyacrylamide gel electrophoresis suggested homogeneity. It is possible that this lipopolysaccharide is homogeneous as to structure but heterogeneous as to size. Any serious attempt to discern size of these molecules from the present data alone is fraught with error. Little is known of the adsorption characteristics of such molecules on the controlled-pore-size glass beads, and the diffusability of lipopolysaccharides in polyacrylamide gels complicates the electrophoretic migration. Nevertheless, if these problems are kept in mind, some speculation as to size can be made. Except for the lipopolysaccharide of A. axanthum, one can assume one diacylglycerol per molecule and calculate the apparent molecular weight. In the case of A. granularum this value is approximately 19,000, in good agreement with the permeation chromatography results, which suggest a molecular weight of about 20,000. In this regard might be mentioned the inability of this lipopolysaccharide to elicit antibody response in rabbits, in contrast to the other lipopolysaccharides (R. J. Lynn, personal communication), which could be explained as a lack of sufficient size. Size of bacterial lipopolysaccharides is a determining factor in the immune response of ,8-lymphocytes. Repeated antigenic determinants are required for multiple bonding to reactive B-cell receptors (9). The size of the lipopolysaccharide from A. laidlawii appears to be in the range of 100,000 to 150,000; that of A. axanthum seems to be about 100,000. The lipopolysaccharide from A. modicum has an apparent size of 36,000 based upon calculation from the content of diacylglycerol. This result is compatible with the smallest particle seen in the deacylated product
J. BACTERIOL.
on the gel permeation column. The remaining two components eluting from this column have apparent sizes of 72,000 and 142,000 and could be construed as dimer and tetramer, respectively. It must be emphasized that these values are only speculative but probably are within the proper order of magnitude. ACKNOWLEDGMENTS We are indebted to Leigh Ann Smith for performing the polyacrylamide gel electrophoresis and to W. R. Mayberry for the gas chromatographic analyses. The dye, Procion red MX-2B, was a generous gift from D. P. Raper, ICI-United States, Inc., Wilmington, Del. LITERATURE CITED 1. Anderson, M. W., A. Hendric, J. R. A. Millar, and A. C. Munroe. 1971. Electrophoresis of dyed polysaccharides on "phoroslides." Analyst 96:870-874. 2. Fairbanks, G., T. L. Steck, and D. F. H. Wallach. 1971. Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane. Biochemistry 10:2606-2617. 3. Gilliam, J. M., and H. J. Morowitz. 1972. Characterization of the plasma membrane of Mycoplasma laidlawii. IX. Isolation and characterization of the membrane polyhexosamine. Biochim. Biophys. Acta 274:353-363. 4. Jann, B., K. Reske, and K. Jann. 1975. Heterogeneity of lipopolysaccharides. Analysis of polysaccharide chain lengths by sodium dodecylsulfate-polyacrylamide gel electrophoresis. Eur. J. Biochem. 60:239-246. 5. Lindberg, B. 1972. Methylation analysis of polysaccharides, p. 178-195. In S. P. Colowick and N. 0. Kaplan (ed.), Methods in enzymology, vol. 28. Academic Press Inc., New York. 6. Mayberry, W. R., P. F. Smith, and T. A. Langworthy. 1974. A heptose-containing pentaglycosyl diglyceride among the lipids of Acholeplasma modicum. J. Bacteriol. 118:898-904. 7. Mayberry-Carson, K. J., T. A. Langworthy, W. R. Mayberry, and P. F. Smith. 1974. A new class of lipopolysaccharide from Thermoplasma acidophilum. Biochim. Biophys. Acta 360:217-229. 8. Mayberry-Carson, K. J., I. L. Roth, and P. F. Smith. 1975. Ultrastructure of lipopolysaccharide isolated from Thermoplasma acidophilum. J. Bacteriol. 121:700-703. 9. Moller, G., 0. Sjoberg, and J. Andersson. 1973. Immunogenicity, tolerogenicity and mitogenicity of lipopolysaccharides, p. 44-48. In E. H. Kass and S. M. Wolff (ed.), Bacterial lipopolysaccharides. University of Chicago Press, Chicago. 10. Smith, P. F., T. A. Langworthy, and W. R. Mayberry. 1976. Distribution and composition of lipopolysaccharides from mycoplasmas. J. Bacteriol. 125:916-922. 11. Sugiyama, T., P. F. Smith, T. A. Langworthy, and W. R. Mayberry. 1974. Immunological analysis of glycolipids and lipopolysaccharides derived from various mycoplasmas. Infect. Immun. 10:1273-1279.
