Vol. 26, No. 1
INFECTION AND IMMUNITY, Oct. 1979, p. 262-269
0019-9567/79/10-0262/08$02.00/0
Modulation of the Immune Response to Sheep Erythrocytes by Lipid-Free Glycerol Teichoic Acid FRANK W. CHORPENNING,* JOHN J. LYNCH, JR., HAROLD R. COOPERt AND JOHN W. OLDFATHER The Ohio State University, Columbus, Ohio 43210 Received for publication 3 April 1979
The 4-day response of C3H/HeJ mice to sheep erythrocytes was suppressed by a lipid-free teichoic acid with an average molecular weight of 2,900 when it was administered by the intraperitoneal route. Enhancement was not observed at that time, and neither suppression nor enhancement could be demonstrated by the intravenous route. Either suppression or enhancement of background plaques could be induced, depending upon the timing. Dosage influenced the degree of suppression from 8 to 100 jtg, whereas suppression of background plaques required only 1 ,tg of lipid-free teichoic acid. The kinetics of the sheep erythrocyte response was altered by treatment of the mice with lipid-free teichoic acid, delaying the peak until day 5 and producing enhancement at that time. Although lipid-free teichoic acid was shown to be toxic for mouse splenocytes (50% lethal dose, ca. 200 Mug) in vitro, no effect at the levels employed was observed in vivo. The data presented indicate that modulatory activity is influenced by route, timing, dosage, and apparently the number of antibody-secreting cells.
Enhancement or suppression of unrelated immune responses by a variety of bacterial (as well as nonbacterial) substances is a well-established phenomenon (reviewed in reference 23); however, the mechanisms of action are essentially unknown. With several of these modulators, such as lipopolysaccharide (LPS) (13) and dextran (3), it has been shown that the direction of the effect (enhancement or suppression) is influenced by dosage, route, and time of administration. Speculation regarding mechanism has been directed toward various aspects of the modulatory molecule, including cellular binding site, toxicity, and the importance of a lipid component. Some of these substances, such as LPS and streptococcal exotoxin (17), are toxic. Others, such as dextran, are nontoxic. Most work toward identification of target cells has been done with LPS. Employing C3H/St mice, Skidmore et al. (25) have shown that both B lymphocytes and macrophages are involved in immunomodulation by LPS. A requirement for lipid (lipid A) has been demonstrated (6) for enhancement of immune responses by LPS, but obviously lipid is not required for action by other substances, such as dextran. Thus, it appears that there are differences in binding or target site between modulators in which lipid is important and those in which it is not a component. t Present address: Microbiological Associates, Walkersville, MD 21793.
A fatty acid requirement has been reported for suppression of sheep erythrocyte (SRBC) responses in C3H/HeJ mice by a glycerol teichoic acid (GTA) extracted from Streptococcus pyogenes (20, 21); however, under the conditions of those experiments, no enhancement was observed. We have recently shown (7) that purified GTA from a Bacillus sp. (ATCC 29726), even though lacking covalently bound fatty acids, can adsorb to cell membranes. Since this membranebinding activity may be involved in the mechanism of immunomodulation, we examined the effect of this purified lipid-free teichoic acid (LFTA) on responses to SRBC. (A preliminary report of portions of this work was presented at the 61st Annual Meeting of the Federation of American Societies for Experimental Biology, Chicago, Ill., 1 to 8 April 1977.) MATERUILS AND METHODS Animals. Female C3H/HeJ mice and Swiss mice were purchased from Jackson Laboratories, Bar Harbor, Maine, and were maintained in plastic cages on the usual food (Laboratory Animal Chow 5010, Ralston-Purina Co., St. Louis, Mo.). They were 10 to 12 weeks of age when examined. Teichoic acid. Lipid-containing teichoic acid (LCTA) was prepared by extraction with an equal volume of 95% aqueous phenol from Bacillus sp. ATCC 29726 and was purified as described by Decker et al. (9). Briefly, it was dialyzed, lyophilized, and dissolved in phosphate-buffered saline. This solution was treated with ribonuclease (type III-B, Sigma
262
VOL. 26, 1979
IMMUNOMODULATION BY LIPID-FREE TEICHOIC ACID
Chemical Co., St. Louis, Mo.), and the protein was precipitated with cold trichloroacetic acid, after which GTA was precipitated from the supernatant with ethyl alcohol. Contaminants were removed by separation on a Sephadex G-100 column, followed by passage through a Bio-Rad 50W X4 (H' form) cation-exchange resin, using 0.1 M acetate buffer. The eluate was dialyzed against water and lyophilized. LFTA was prepared from the LCTA by extraction (four times) with chloroform-ether (3:1) and drying under a stream of nitrogen gas. This was possible because LCTA prepared by the above method is stripped of any covalently bound fatty acids (7). Chemical analyses. Hexose was assayed by the anthrone method (27), hexosamine was assayed by the method of Boas (4), pentose was assayed by the method of Dische (10), and ribose (ribonucleic acid) was assayed by the orcinol-FeCl3 test. Sugars were also identified by gas-liquid chromatography (1). Phosphorus determinations followed the procedure of Chen et al. (5), and glycerol was determined by the glycerol dehydrogenase assay (16). Amino acids were identified and quantitated on an amino acid analyzer (Beckman Instruments, Inc., Fullerton, Calif.), and fatty acids were identified and quantitated by highpressure liquid chromatography, using the method of Durst et al. (12), as previously described in detail (7). Briefly, samples were saponified with methanol, neutralized, and evaporated to dryness under nitrogen gas. Residues were suspended in pyrogen-free water, acidified (pH 2), and extracted with chloroform-ether. The extracts were dried and acidified with 1 N alcoholic HCl, and potassium salts of the fatty acids were formed with 1 N alcoholic KOH. Extracts (samples) were then converted to their phenacyl ester derivatives with acetonitrile containing 60 nM a-p-dibromoacetophenone and 6 nM dicyclohexano-18-crown-6. The derivatized fatty acids were identified and quantitated on a Tracor 970 detector, using a Tracor 990 isochromatographic pump (Tracor Instruments, Austin, Tex.) and a ,u Bondapak/C18 column (Waters Associates, Milford, Mass.). C6 to C12 fatty acids were separated with a 90% methanol-10% water mobile phase, C12 to C20 fatty acids were separated with 95% methanol-5% water, and C20 to C24 fatty acids were separated with absolute methanol. LPS assays. Tests for LPS, designed to detect contamination by endotoxin, employed the Limulus amoebocyte lysate assay (Worthington Pyrostat Reagent, Worthington Biochemicals Corp., Freehold, N.J.). LPS used for a positive control and for modulator comparison was obtained from Difco Laboratories, Detroit, Mich. (extracted from Escherichia coli 0127:B8 by the Westphal method). Examination of modulatory effects. Each test or control group consisted of 10 mice. In initial experiments, the test animals were injected either intravenously or intraperitoneally (i.p.) with 8 jg of either LFTA or LCTA per mouse, divided into three subdoses 1 day apart. This total dosage was equal to 0.32 body weight and was calculated to be ,ig/g of average 1/1,000 of that used by Miller et al. (21), based on the phosphate content of each. On day 3, mice were injected by the same route with 0.5 ml of a 10% suspension of washed SRBC in phosphate-buffered saline. A control group received phosphate-buffered saline in
263
lieu of GTA, and a second control group received GTA but no SRBC. At 4 days after SRBC injection, all animals were sacrificed and their splenocytes were examined by the hemolytic plaque assay. Further experiments were designed to determine the effects of timing and dosage on modulation by LFTA. The former protocols were designed as described above, except the total dose of LFTA (8 jig) was given on either day 1, 2, or 3 of the experiment. The latter protocols were similar in design, except they included groups of mice given various doses of LFTA ranging from 0 to 100 l g and each dose was given i.p. 2 days before injection of SRBC. The effect of LFTA on the kinetics of the SRBC response was also studied. In these experiments, 10 ,ug of LFTA was given 2 days before injection of SRBC, and the hemolytic plaque response was assayed on days 1 to 9 postinjection of SRBC. Hemolytic plaque test. For the hemolytic plaque test, the technique of Jerne and Nordin (18) was modified by extending the incubation time (total, 2.5 h) and by spectrophotometric adjustment of the SRBC (10%) and complement concentrations. Spleen cells were separated in Hanks balanced salt solution, and tests were carried out in agarose in 60-mm petri plates. All tests were performed in triplicate. Tests for toxicity. Toxicity of LFTA for mouse splenocytes was determined by exposing 106 viable cells per culture suspended in RPMI 1640 medium (with 10% fetal calf serum and 100 Itg of gentamicin sulfate per ml) to various doses of LFTA. Cultures were incubated in 5% CO2 at 37°C for 4 days. Total viable cells (trypan blue exclusion) were counted and compared with control cultures. Twelve replicates were cultured for each LFTA dose and pooled for counting.
RESULTS The purified phenol extracts contained approximately equimolar amounts of glycerol and phosphorus, the average values of different lots being 4.89 ,Lmol of glycerol per mg and 4.86 gmol of phosphorus per mg. They also contained small amounts of protein (18 amino acids detected), hexose, and fatty acids (Table 1). Since these compounds varied with the extract and individually yielded low molar ratios to glycerol, they did not appear to be associated with the teichoic acid molecule, but were probably impurities. These extracts are referred to as LCTA, although they actually may contain only free fatty acids. Teichoic acid prepared in the same way was extracted four times with chloroform-ether and used in the following experiments. The chemical analysis is shown in Table 2. Before extraction, this preparation contained 36.6 ,ug of fatty acids per mg, distributed as shown in the high-pressure liquid chromatography profile (Fig. 1). After chloroform-ether extraction, no fatty acids could be detected (Fig. 2), indicating that those demonstrated before extraction were not covalently bound in the purified material. This prep-
264
INFECT. IMMUN.
CHORPENNING ET AL.
TABLE 1. Chemical analysis ofpurified phenol extracts" % umol/ Molar ~ tg/mg Determination Dee g/mg (wt) mg ratio 450.0 45.0 4.89 Glycerol 301.2 30.1 4.86 1:1 Phosphate (calculated) (15.1) (151.0) (Phosphorus)Y 62.7 6.3 Hexose (Glucosamine) (15.7) (1.6) 0.09 1:54 14.8 1.5 Amino acids (total, 18) (Alanine) (0.22) (0.02) 0.025 1:196 81.7 8.2 Fatty acids a Average of three extracts before chloroform-ether extraction. b Molar ratio to glycerol. c Substances within parentheses are included in the measurements shown immediately above each. TABLE 2. Chemical analysis of the LFTA used in these experiments
Determination
pg/mg
Glycerol Phosphorus Hexoseb Glucosamine
494 170 54 14 10 0.04
mol/ 5.4 5.5
Mratior 1:1
0.08 1:68 Amino acids 0.004 Alanine 1:1,350 Fatty acids 0.0¢ aMolar ratio to glycerol. b There were trace amounts (0.05 >0.05 0.1 0.5 1.0
INFECTION AND IMMUNITY, Oct. 1979, p. 262-269
0019-9567/79/10-0262/08$02.00/0
Modulation of the Immune Response to Sheep Erythrocytes by Lipid-Free Glycerol Teichoic Acid FRANK W. CHORPENNING,* JOHN J. LYNCH, JR., HAROLD R. COOPERt AND JOHN W. OLDFATHER The Ohio State University, Columbus, Ohio 43210 Received for publication 3 April 1979
The 4-day response of C3H/HeJ mice to sheep erythrocytes was suppressed by a lipid-free teichoic acid with an average molecular weight of 2,900 when it was administered by the intraperitoneal route. Enhancement was not observed at that time, and neither suppression nor enhancement could be demonstrated by the intravenous route. Either suppression or enhancement of background plaques could be induced, depending upon the timing. Dosage influenced the degree of suppression from 8 to 100 jtg, whereas suppression of background plaques required only 1 ,tg of lipid-free teichoic acid. The kinetics of the sheep erythrocyte response was altered by treatment of the mice with lipid-free teichoic acid, delaying the peak until day 5 and producing enhancement at that time. Although lipid-free teichoic acid was shown to be toxic for mouse splenocytes (50% lethal dose, ca. 200 Mug) in vitro, no effect at the levels employed was observed in vivo. The data presented indicate that modulatory activity is influenced by route, timing, dosage, and apparently the number of antibody-secreting cells.
Enhancement or suppression of unrelated immune responses by a variety of bacterial (as well as nonbacterial) substances is a well-established phenomenon (reviewed in reference 23); however, the mechanisms of action are essentially unknown. With several of these modulators, such as lipopolysaccharide (LPS) (13) and dextran (3), it has been shown that the direction of the effect (enhancement or suppression) is influenced by dosage, route, and time of administration. Speculation regarding mechanism has been directed toward various aspects of the modulatory molecule, including cellular binding site, toxicity, and the importance of a lipid component. Some of these substances, such as LPS and streptococcal exotoxin (17), are toxic. Others, such as dextran, are nontoxic. Most work toward identification of target cells has been done with LPS. Employing C3H/St mice, Skidmore et al. (25) have shown that both B lymphocytes and macrophages are involved in immunomodulation by LPS. A requirement for lipid (lipid A) has been demonstrated (6) for enhancement of immune responses by LPS, but obviously lipid is not required for action by other substances, such as dextran. Thus, it appears that there are differences in binding or target site between modulators in which lipid is important and those in which it is not a component. t Present address: Microbiological Associates, Walkersville, MD 21793.
A fatty acid requirement has been reported for suppression of sheep erythrocyte (SRBC) responses in C3H/HeJ mice by a glycerol teichoic acid (GTA) extracted from Streptococcus pyogenes (20, 21); however, under the conditions of those experiments, no enhancement was observed. We have recently shown (7) that purified GTA from a Bacillus sp. (ATCC 29726), even though lacking covalently bound fatty acids, can adsorb to cell membranes. Since this membranebinding activity may be involved in the mechanism of immunomodulation, we examined the effect of this purified lipid-free teichoic acid (LFTA) on responses to SRBC. (A preliminary report of portions of this work was presented at the 61st Annual Meeting of the Federation of American Societies for Experimental Biology, Chicago, Ill., 1 to 8 April 1977.) MATERUILS AND METHODS Animals. Female C3H/HeJ mice and Swiss mice were purchased from Jackson Laboratories, Bar Harbor, Maine, and were maintained in plastic cages on the usual food (Laboratory Animal Chow 5010, Ralston-Purina Co., St. Louis, Mo.). They were 10 to 12 weeks of age when examined. Teichoic acid. Lipid-containing teichoic acid (LCTA) was prepared by extraction with an equal volume of 95% aqueous phenol from Bacillus sp. ATCC 29726 and was purified as described by Decker et al. (9). Briefly, it was dialyzed, lyophilized, and dissolved in phosphate-buffered saline. This solution was treated with ribonuclease (type III-B, Sigma
262
VOL. 26, 1979
IMMUNOMODULATION BY LIPID-FREE TEICHOIC ACID
Chemical Co., St. Louis, Mo.), and the protein was precipitated with cold trichloroacetic acid, after which GTA was precipitated from the supernatant with ethyl alcohol. Contaminants were removed by separation on a Sephadex G-100 column, followed by passage through a Bio-Rad 50W X4 (H' form) cation-exchange resin, using 0.1 M acetate buffer. The eluate was dialyzed against water and lyophilized. LFTA was prepared from the LCTA by extraction (four times) with chloroform-ether (3:1) and drying under a stream of nitrogen gas. This was possible because LCTA prepared by the above method is stripped of any covalently bound fatty acids (7). Chemical analyses. Hexose was assayed by the anthrone method (27), hexosamine was assayed by the method of Boas (4), pentose was assayed by the method of Dische (10), and ribose (ribonucleic acid) was assayed by the orcinol-FeCl3 test. Sugars were also identified by gas-liquid chromatography (1). Phosphorus determinations followed the procedure of Chen et al. (5), and glycerol was determined by the glycerol dehydrogenase assay (16). Amino acids were identified and quantitated on an amino acid analyzer (Beckman Instruments, Inc., Fullerton, Calif.), and fatty acids were identified and quantitated by highpressure liquid chromatography, using the method of Durst et al. (12), as previously described in detail (7). Briefly, samples were saponified with methanol, neutralized, and evaporated to dryness under nitrogen gas. Residues were suspended in pyrogen-free water, acidified (pH 2), and extracted with chloroform-ether. The extracts were dried and acidified with 1 N alcoholic HCl, and potassium salts of the fatty acids were formed with 1 N alcoholic KOH. Extracts (samples) were then converted to their phenacyl ester derivatives with acetonitrile containing 60 nM a-p-dibromoacetophenone and 6 nM dicyclohexano-18-crown-6. The derivatized fatty acids were identified and quantitated on a Tracor 970 detector, using a Tracor 990 isochromatographic pump (Tracor Instruments, Austin, Tex.) and a ,u Bondapak/C18 column (Waters Associates, Milford, Mass.). C6 to C12 fatty acids were separated with a 90% methanol-10% water mobile phase, C12 to C20 fatty acids were separated with 95% methanol-5% water, and C20 to C24 fatty acids were separated with absolute methanol. LPS assays. Tests for LPS, designed to detect contamination by endotoxin, employed the Limulus amoebocyte lysate assay (Worthington Pyrostat Reagent, Worthington Biochemicals Corp., Freehold, N.J.). LPS used for a positive control and for modulator comparison was obtained from Difco Laboratories, Detroit, Mich. (extracted from Escherichia coli 0127:B8 by the Westphal method). Examination of modulatory effects. Each test or control group consisted of 10 mice. In initial experiments, the test animals were injected either intravenously or intraperitoneally (i.p.) with 8 jg of either LFTA or LCTA per mouse, divided into three subdoses 1 day apart. This total dosage was equal to 0.32 body weight and was calculated to be ,ig/g of average 1/1,000 of that used by Miller et al. (21), based on the phosphate content of each. On day 3, mice were injected by the same route with 0.5 ml of a 10% suspension of washed SRBC in phosphate-buffered saline. A control group received phosphate-buffered saline in
263
lieu of GTA, and a second control group received GTA but no SRBC. At 4 days after SRBC injection, all animals were sacrificed and their splenocytes were examined by the hemolytic plaque assay. Further experiments were designed to determine the effects of timing and dosage on modulation by LFTA. The former protocols were designed as described above, except the total dose of LFTA (8 jig) was given on either day 1, 2, or 3 of the experiment. The latter protocols were similar in design, except they included groups of mice given various doses of LFTA ranging from 0 to 100 l g and each dose was given i.p. 2 days before injection of SRBC. The effect of LFTA on the kinetics of the SRBC response was also studied. In these experiments, 10 ,ug of LFTA was given 2 days before injection of SRBC, and the hemolytic plaque response was assayed on days 1 to 9 postinjection of SRBC. Hemolytic plaque test. For the hemolytic plaque test, the technique of Jerne and Nordin (18) was modified by extending the incubation time (total, 2.5 h) and by spectrophotometric adjustment of the SRBC (10%) and complement concentrations. Spleen cells were separated in Hanks balanced salt solution, and tests were carried out in agarose in 60-mm petri plates. All tests were performed in triplicate. Tests for toxicity. Toxicity of LFTA for mouse splenocytes was determined by exposing 106 viable cells per culture suspended in RPMI 1640 medium (with 10% fetal calf serum and 100 Itg of gentamicin sulfate per ml) to various doses of LFTA. Cultures were incubated in 5% CO2 at 37°C for 4 days. Total viable cells (trypan blue exclusion) were counted and compared with control cultures. Twelve replicates were cultured for each LFTA dose and pooled for counting.
RESULTS The purified phenol extracts contained approximately equimolar amounts of glycerol and phosphorus, the average values of different lots being 4.89 ,Lmol of glycerol per mg and 4.86 gmol of phosphorus per mg. They also contained small amounts of protein (18 amino acids detected), hexose, and fatty acids (Table 1). Since these compounds varied with the extract and individually yielded low molar ratios to glycerol, they did not appear to be associated with the teichoic acid molecule, but were probably impurities. These extracts are referred to as LCTA, although they actually may contain only free fatty acids. Teichoic acid prepared in the same way was extracted four times with chloroform-ether and used in the following experiments. The chemical analysis is shown in Table 2. Before extraction, this preparation contained 36.6 ,ug of fatty acids per mg, distributed as shown in the high-pressure liquid chromatography profile (Fig. 1). After chloroform-ether extraction, no fatty acids could be detected (Fig. 2), indicating that those demonstrated before extraction were not covalently bound in the purified material. This prep-
264
INFECT. IMMUN.
CHORPENNING ET AL.
TABLE 1. Chemical analysis ofpurified phenol extracts" % umol/ Molar ~ tg/mg Determination Dee g/mg (wt) mg ratio 450.0 45.0 4.89 Glycerol 301.2 30.1 4.86 1:1 Phosphate (calculated) (15.1) (151.0) (Phosphorus)Y 62.7 6.3 Hexose (Glucosamine) (15.7) (1.6) 0.09 1:54 14.8 1.5 Amino acids (total, 18) (Alanine) (0.22) (0.02) 0.025 1:196 81.7 8.2 Fatty acids a Average of three extracts before chloroform-ether extraction. b Molar ratio to glycerol. c Substances within parentheses are included in the measurements shown immediately above each. TABLE 2. Chemical analysis of the LFTA used in these experiments
Determination
pg/mg
Glycerol Phosphorus Hexoseb Glucosamine
494 170 54 14 10 0.04
mol/ 5.4 5.5
Mratior 1:1
0.08 1:68 Amino acids 0.004 Alanine 1:1,350 Fatty acids 0.0¢ aMolar ratio to glycerol. b There were trace amounts (0.05 >0.05 0.1 0.5 1.0
