Vol. 59, No. 8
INFECTION AND IMMUNITY, Aug. 1991, p. 2781-2789 0019-9567/91/082781-09$02.00/0 Copyright C) 1991, American Society for Microbiology
Phospholipase Activity of Mycobacterium leprae Harvested from Experimentally Infected Armadillo Tissue PAUL R. WHEELER* AND COLIN RATLEDGE Department of Applied Biology, University of Hull, Hull HU6 7RX, United Kingdom Received 22 February 1991/Accepted 28 May 1991
Three types of phospholipase activity-phospholipase Al, A2, and lysophospholipase-were detected in Mycobacterium leprae harvested from armadillo tissue at about 25% of the specific activity found in a slowly growing mycobacterium, Mycobacterium microti, which was grown in medium to optimize its phospholipase activity. The highest activity found was lysophospholipase, which released fatty acid from 2-lyso-phosphatidylcholine. Phospholipase activity was detected by using phosphatidylcholine and phosphatidylethanolamine. Differences in relative activities with these three types of substrate distinguished phospholipase activity in M. leprae extracts from armadillo liver extracts. Furthermore, retention of activity in M. keprae after NaOH treatment showed that the activity associated with M. leprae was not host derived. The specific activity of phospholipase was 20 times higher in extracts of M. kprae than in intact M. leprae organisms. Diazotization, a treatment which abolishes activities of surface enzymes exposed to the environment by the formation of covalent azide bonds with exposed amino groups, did not affect M. leprae's phospholipase activity, with one exception: release of arachidonic acid from phosphatidylcholine, which was partially inhibited. Phenolic glycolipid I, the major excreted amphipathic lipid of M. leprae, inhibited phospholipase activity, including release of arachidonic acid, for both M. leprae- and armadillo-derived activity.
Generally, in diseases caused by Rickettsia, any damage is limited to the individual cells that have been parasitized, though the cumulative effect can result in serious disease. In a nonaggressive role, phospholipases of pathogens may be involved solely in nutrition, hydrolyzing phospholipids to release fatty acids (Fig. 1) which the pathogen can then use as a carbon source or a starting material for biosynthesis of more complex lipid or longer-chain fatty acids. M. leprae has the ability to use fatty acids for such metabolic activities: [14C]palmitate (a 16-carbon fatty acid) is incorporated into phenolic glycolipid I (PGL-I) (a characteristic capsular lipid [7]), is catabolized to 14CO2 (6, 31), and is elongated to longer-chain fatty acids (32). Therefore, in the present work, the first study of phospholipase activity in M. leprae, the phospholipases sought are those which hydrolyze the ester bond between the glycerol and fatty acyl moiety of phospholipids to release fatty acids (Fig. 1). There is a further way in which phospholipid metabolism in both the host and M. leprae could be affected. PGL-I, a major excreted amphipathic lipid from M. leprae must be able to insert in host membranes (24). This insertion may affect the activity of phospholipases on the phospholipids which are themselves major membrane constituents. An example of a microbial lipid modulating host cell function in this way is lipopolysaccharide from gram-negative bacteria, which increases the rate at which arachidonate is released from phospholipids containing arachidonoyl moieties as part of its effect on control of macrophage function and metabolism (1). Therefore, in the present work, the effects of PGL-I on phospholipase activities from host tissue and from M. leprae are investigated.
Mycobacterium leprae, the leprosy bacillus, infects and multiplies inside host cells, principally Schwann cells in nerve tissue (19) and macrophages (18). It does not appear to cause damage directly, but tissue damage during leprosy occurs as a result of immunopathological reactions (14, 15, 20). Nevertheless, M. leprae has an immuno-suppressing effect on its host and can go on to reach extremely high densities such as 109 bacteria (-0.1 mg of dry weight) per gram of tissue in human lepromatous leprosy (15) and over 1010 bacteria per gram of tissue in experimentally infected armadillos and mice (22). The search for immunomodulatory substances derived from M. leprae is therefore a search for virulence factors. Recently, lipomannan and lipoarabinomannan have been suggested as candidates for modifying host cell behavior, as they are extracellular and have a phosphatidylinositol anchor which could insert into the lipid bilayer in host cell membranes (10). Phospholipases, the subject of the present work, could also affect host cell membranes by deacylating membrane phospholipids if the phospholipases are extracellular or during periods when M. leprae organisms are intimately associated with host cell membranes. Phospholipases can be microbial virulence factors. In their most damaging forms, such as phospholipases C from Clostridium perfingens, they cause widespread tissue damage, are hemolytic and necrotizing (16, 17), and are the major cause of the pathology of gangrene. Phospholipase C hydrolyzes the choline-phosphoryl bond (see Fig. 1) and so is not like the acylhydrolases which are the subject of this present work. Nevertheless, acylhydrolases with damaging effects are known in nature. Examples are the phospholipases A in bee and cobra venom (5). Phospholipases may be needed by the bacteria to gain entry or force an escape through host cell membranes. The phospholipases A of Rickettsia species appear to play a part in both of these processes (25, 33). *
MATERIALS AND METHODS
Mycobacteria and harvesting. M. leprae was grown in nine-banded armadillos (Dasypus novemcinctus L.) and was harvested by a method including differential centrifugation, density gradient centrifugation, and partition in an aqueous
Corresponding author. 2781
2782
WHEELER AND RATLEDGE
INFECT. IMMUN.
---,o0-CH2 CH2.(0N(CH3)3
D
TABLE 1. Radioisotopically labelled substrates used in this study
(D P02
(DP,02
Water A2
OH2 A1
-----
-HO
0-00
------CO
----
OCH2
CH2
O2
Substrate
H2
CH2
L-lyso-3-PC, 1-[14C]palmitoyl L-3-PC, 1,2-di[1-14C]palmitoyI L-3-PC, 1-palmitoyl-2[1-14C] palmitoyl L-3-PC, 1-stearoyl-2[1-14C] arachidonyl L-3-phosphatidylethanolamine,
radioactivity
Specific (as supplied) (Ci/mol)
Supplier
56 114 54-57
Amersham NENa NEN
53 110
NEN Amersham
1,2-di[1-14C]palmitoyl " New England Nuclear, division of DuPont Biotechnology
Lipid
FIG. 1. Phospholid molecule and phospholipases. The lipid is PC with its acyl chains in the lipid part of a lipid-water system. The sites of actions of phospholipases A1, A2, C, and D are shown by arrows.
two-phase system (for details, see reference 29; 34). Mycobacterium microti was grown for 15 days at 37°C in a shaker-incubator, shaken at 200 rpm in Dubos medium with added liposomes prepared as described by Kondo et al. (13) to give a concentration of lipid equal to 150 ,ug ml-' (32). M. microti was harvested by washing three times with buffered Tween 80 (0.1% Tween 80 in 1.5 mM morpholineethanesulfonic acid [MES], pH 7.0). All studies with intact mycobacteria were done by adding portions as suspensions in buffered Tween 80 to incubation mixture detailed below. Surface treatments of intact bacteria. Some suspensions of M. leprae were treated with 0.5 M NaOH at 25°C for 30 min to abolish any host-derived activities adsorbed to the surface of the bacteria. The suspensions were neutralized with 1 M N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) base. It was confirmed that this treatment abolished phospholipase activity in extracts of armadillo tissue when the enzyme activity was sought, and enzyme activity was not detected in armadillo liver extracts treated with NaOH by the method routinely used for assaying phospholipase in this study (details of methods given below). Also, suspensions of M. leprae were diazotized with 7-diazonaphthalene-1,3-disulfonic acid (ANDS) in exactly the same way as described previously (30). During diazotization, covalent bonds form between amino groups on surface proteins and ANDS so that any enzyme activity on the surface and exposed to ANDS is irreversibly affected (21). The bacteria were washed free of ANDS. Preparation of cell extracts of mycobacteria. Cell extracts were prepared by ultrasonic disruption and centrifugation twice at 20,000 x g for 10 min as described previously (29,
Systems.
30) so that extracts contained membrane and cytosolic components. Extracts were in 2.5 mM MES adjusted to pH 7 with NaOH. If necessary, they could be concentrated by freeze drying and reconstituting in a small volume of water without loss of activity. Preparation of armadillo tissue extracts. Both infected and noninfected armadillo tissues were homogenized at pH 8.7 and centrifuged at 10,000 x g for 10 min; the supernatants were collected and recentrifuged at 10,000 x g for 10 min. The final supernatant was adjusted to pH 7.8 with citric acid and used as the armadillo tissue extract. Extracts of both mycobacteria and armadillo tissue could be stored at -70°C, frozen and rethawed up to three times, and lyophilized without discernible loss of phospholipase activity. Substrates and chemicals. Radioisotopically labelled substrates, all phospholipids, are given in Table 1. Other chemicals, including solvents, were Analar grade. Substrates used in enzyme assays were from Sigma and were of the highest purity available. The compositions of unlabelled phospholipids were as follows. L-a-Phosphatidylcholine (PC) (egg yolk, 99% pure) contained mainly palmitoyl and stearoyl acyl moieties. DL-a-Phosphatidylethanolamine (synthetic, 98% pure) was dipalmitoyl. L-a-2-lyso-Phosphatidylcholine (egg yolk, 99% pure) contained mainly palmitoyl and stearoyl acyl moieties. PGL-I is 3,6-di-O-methylglucosyl-
2,3-di-O-methylrhamnosyl-3-O-methylrhamnosyl-phenolphthiocerol (11); its diol and two mycocerosate moieties which make up the phthiocerol moiety are long chain (>24 carbon) length. PGL-I was provided by the Immunology of Leprosy (IMMLEP) (World Health Organization) bank. All aqueous solutions used in assays were filter sterilized. Enzyme assays. Enzyme assays were based on assays used in the very few previous studies on mycobacterial phospholipases but were modified by performing treatments such as ultrasonication and addition of a lysophospholipid with detergent properties, 2-lyso-phosphatidylcholine (lyso-PC) which should result in formation of micelles (5). Without these treatments, the phospholipids used in this study as enzyme substrates are in the form of liposomes (13). Thus, it was possible to present the phospholipase enzyme(s) with phospholipid substrates in different physical forms. The objective of testing many enzyme assays was to optimize detection of enzyme activity. This was done, for intact bacteria and their extracts, mainly with the relatively plentiful M. microti, and then selected assays were tried with M. Ieprae (Table 2). The assays were as follows. (i) Liposomes with dipalmitoylphosphatidic acid and PC (assay A). The assay with dipalmitoylphosphatidic acid and
VOL. 59, 1991
PHOSPHOLIPASE OF MYCOBACTERIUM LEPRAE
2783
TABLE 2. Optimizing the phospholipase assay with 1,2-di[ l4C 16:0]PC as substratea Controls: radioactiv-
Activity with the following:
ity corresponding to the following:
M. leprae
M. microti Assay system
Intact
bacteriab
Cell extractc
Intact
bacteriac
Cell
extractc
graycrol
fractionc
0_o 7
02.1
fractionb
A
Liposomes with DPPAd and PC 10 mM MES-Tween 80, pH 7 PBS, pH 7
0.2-0.5 (1-2) 0.2-0.4 (1-2)
13 (3)e ND
ND 1.5 (25)
ND ND
B
Micelles with DPPA, PC and lyso-PC 10 mM MES-Tween 80, pH 7 17-29 (74-128) 274-551 (70-141)NDD0-5 ND ND 0}2.5 PBS, pH 7 3.6-12.1 (15-20) 224 (50)e
09.
09.0
C
Liposomes with PC only 10 mM MES-Tween 80, pH 7 10 mM MES-Tween 80 and 5 mM CaCl2, pH 7 Glycerol-phosphate, pH 7
ND 8.1 (34)e
Micelles with PC only 10 mM MES-Tween 80, pH 7 Glycerol-phosphate, pH 7
1.2-3.7 (5-15) 2.6 (11)e
ND
32 (7)e 162-340 (36-76)
2.4-3.2 (3)
INFECTION AND IMMUNITY, Aug. 1991, p. 2781-2789 0019-9567/91/082781-09$02.00/0 Copyright C) 1991, American Society for Microbiology
Phospholipase Activity of Mycobacterium leprae Harvested from Experimentally Infected Armadillo Tissue PAUL R. WHEELER* AND COLIN RATLEDGE Department of Applied Biology, University of Hull, Hull HU6 7RX, United Kingdom Received 22 February 1991/Accepted 28 May 1991
Three types of phospholipase activity-phospholipase Al, A2, and lysophospholipase-were detected in Mycobacterium leprae harvested from armadillo tissue at about 25% of the specific activity found in a slowly growing mycobacterium, Mycobacterium microti, which was grown in medium to optimize its phospholipase activity. The highest activity found was lysophospholipase, which released fatty acid from 2-lyso-phosphatidylcholine. Phospholipase activity was detected by using phosphatidylcholine and phosphatidylethanolamine. Differences in relative activities with these three types of substrate distinguished phospholipase activity in M. leprae extracts from armadillo liver extracts. Furthermore, retention of activity in M. keprae after NaOH treatment showed that the activity associated with M. leprae was not host derived. The specific activity of phospholipase was 20 times higher in extracts of M. kprae than in intact M. leprae organisms. Diazotization, a treatment which abolishes activities of surface enzymes exposed to the environment by the formation of covalent azide bonds with exposed amino groups, did not affect M. leprae's phospholipase activity, with one exception: release of arachidonic acid from phosphatidylcholine, which was partially inhibited. Phenolic glycolipid I, the major excreted amphipathic lipid of M. leprae, inhibited phospholipase activity, including release of arachidonic acid, for both M. leprae- and armadillo-derived activity.
Generally, in diseases caused by Rickettsia, any damage is limited to the individual cells that have been parasitized, though the cumulative effect can result in serious disease. In a nonaggressive role, phospholipases of pathogens may be involved solely in nutrition, hydrolyzing phospholipids to release fatty acids (Fig. 1) which the pathogen can then use as a carbon source or a starting material for biosynthesis of more complex lipid or longer-chain fatty acids. M. leprae has the ability to use fatty acids for such metabolic activities: [14C]palmitate (a 16-carbon fatty acid) is incorporated into phenolic glycolipid I (PGL-I) (a characteristic capsular lipid [7]), is catabolized to 14CO2 (6, 31), and is elongated to longer-chain fatty acids (32). Therefore, in the present work, the first study of phospholipase activity in M. leprae, the phospholipases sought are those which hydrolyze the ester bond between the glycerol and fatty acyl moiety of phospholipids to release fatty acids (Fig. 1). There is a further way in which phospholipid metabolism in both the host and M. leprae could be affected. PGL-I, a major excreted amphipathic lipid from M. leprae must be able to insert in host membranes (24). This insertion may affect the activity of phospholipases on the phospholipids which are themselves major membrane constituents. An example of a microbial lipid modulating host cell function in this way is lipopolysaccharide from gram-negative bacteria, which increases the rate at which arachidonate is released from phospholipids containing arachidonoyl moieties as part of its effect on control of macrophage function and metabolism (1). Therefore, in the present work, the effects of PGL-I on phospholipase activities from host tissue and from M. leprae are investigated.
Mycobacterium leprae, the leprosy bacillus, infects and multiplies inside host cells, principally Schwann cells in nerve tissue (19) and macrophages (18). It does not appear to cause damage directly, but tissue damage during leprosy occurs as a result of immunopathological reactions (14, 15, 20). Nevertheless, M. leprae has an immuno-suppressing effect on its host and can go on to reach extremely high densities such as 109 bacteria (-0.1 mg of dry weight) per gram of tissue in human lepromatous leprosy (15) and over 1010 bacteria per gram of tissue in experimentally infected armadillos and mice (22). The search for immunomodulatory substances derived from M. leprae is therefore a search for virulence factors. Recently, lipomannan and lipoarabinomannan have been suggested as candidates for modifying host cell behavior, as they are extracellular and have a phosphatidylinositol anchor which could insert into the lipid bilayer in host cell membranes (10). Phospholipases, the subject of the present work, could also affect host cell membranes by deacylating membrane phospholipids if the phospholipases are extracellular or during periods when M. leprae organisms are intimately associated with host cell membranes. Phospholipases can be microbial virulence factors. In their most damaging forms, such as phospholipases C from Clostridium perfingens, they cause widespread tissue damage, are hemolytic and necrotizing (16, 17), and are the major cause of the pathology of gangrene. Phospholipase C hydrolyzes the choline-phosphoryl bond (see Fig. 1) and so is not like the acylhydrolases which are the subject of this present work. Nevertheless, acylhydrolases with damaging effects are known in nature. Examples are the phospholipases A in bee and cobra venom (5). Phospholipases may be needed by the bacteria to gain entry or force an escape through host cell membranes. The phospholipases A of Rickettsia species appear to play a part in both of these processes (25, 33). *
MATERIALS AND METHODS
Mycobacteria and harvesting. M. leprae was grown in nine-banded armadillos (Dasypus novemcinctus L.) and was harvested by a method including differential centrifugation, density gradient centrifugation, and partition in an aqueous
Corresponding author. 2781
2782
WHEELER AND RATLEDGE
INFECT. IMMUN.
---,o0-CH2 CH2.(0N(CH3)3
D
TABLE 1. Radioisotopically labelled substrates used in this study
(D P02
(DP,02
Water A2
OH2 A1
-----
-HO
0-00
------CO
----
OCH2
CH2
O2
Substrate
H2
CH2
L-lyso-3-PC, 1-[14C]palmitoyl L-3-PC, 1,2-di[1-14C]palmitoyI L-3-PC, 1-palmitoyl-2[1-14C] palmitoyl L-3-PC, 1-stearoyl-2[1-14C] arachidonyl L-3-phosphatidylethanolamine,
radioactivity
Specific (as supplied) (Ci/mol)
Supplier
56 114 54-57
Amersham NENa NEN
53 110
NEN Amersham
1,2-di[1-14C]palmitoyl " New England Nuclear, division of DuPont Biotechnology
Lipid
FIG. 1. Phospholid molecule and phospholipases. The lipid is PC with its acyl chains in the lipid part of a lipid-water system. The sites of actions of phospholipases A1, A2, C, and D are shown by arrows.
two-phase system (for details, see reference 29; 34). Mycobacterium microti was grown for 15 days at 37°C in a shaker-incubator, shaken at 200 rpm in Dubos medium with added liposomes prepared as described by Kondo et al. (13) to give a concentration of lipid equal to 150 ,ug ml-' (32). M. microti was harvested by washing three times with buffered Tween 80 (0.1% Tween 80 in 1.5 mM morpholineethanesulfonic acid [MES], pH 7.0). All studies with intact mycobacteria were done by adding portions as suspensions in buffered Tween 80 to incubation mixture detailed below. Surface treatments of intact bacteria. Some suspensions of M. leprae were treated with 0.5 M NaOH at 25°C for 30 min to abolish any host-derived activities adsorbed to the surface of the bacteria. The suspensions were neutralized with 1 M N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) base. It was confirmed that this treatment abolished phospholipase activity in extracts of armadillo tissue when the enzyme activity was sought, and enzyme activity was not detected in armadillo liver extracts treated with NaOH by the method routinely used for assaying phospholipase in this study (details of methods given below). Also, suspensions of M. leprae were diazotized with 7-diazonaphthalene-1,3-disulfonic acid (ANDS) in exactly the same way as described previously (30). During diazotization, covalent bonds form between amino groups on surface proteins and ANDS so that any enzyme activity on the surface and exposed to ANDS is irreversibly affected (21). The bacteria were washed free of ANDS. Preparation of cell extracts of mycobacteria. Cell extracts were prepared by ultrasonic disruption and centrifugation twice at 20,000 x g for 10 min as described previously (29,
Systems.
30) so that extracts contained membrane and cytosolic components. Extracts were in 2.5 mM MES adjusted to pH 7 with NaOH. If necessary, they could be concentrated by freeze drying and reconstituting in a small volume of water without loss of activity. Preparation of armadillo tissue extracts. Both infected and noninfected armadillo tissues were homogenized at pH 8.7 and centrifuged at 10,000 x g for 10 min; the supernatants were collected and recentrifuged at 10,000 x g for 10 min. The final supernatant was adjusted to pH 7.8 with citric acid and used as the armadillo tissue extract. Extracts of both mycobacteria and armadillo tissue could be stored at -70°C, frozen and rethawed up to three times, and lyophilized without discernible loss of phospholipase activity. Substrates and chemicals. Radioisotopically labelled substrates, all phospholipids, are given in Table 1. Other chemicals, including solvents, were Analar grade. Substrates used in enzyme assays were from Sigma and were of the highest purity available. The compositions of unlabelled phospholipids were as follows. L-a-Phosphatidylcholine (PC) (egg yolk, 99% pure) contained mainly palmitoyl and stearoyl acyl moieties. DL-a-Phosphatidylethanolamine (synthetic, 98% pure) was dipalmitoyl. L-a-2-lyso-Phosphatidylcholine (egg yolk, 99% pure) contained mainly palmitoyl and stearoyl acyl moieties. PGL-I is 3,6-di-O-methylglucosyl-
2,3-di-O-methylrhamnosyl-3-O-methylrhamnosyl-phenolphthiocerol (11); its diol and two mycocerosate moieties which make up the phthiocerol moiety are long chain (>24 carbon) length. PGL-I was provided by the Immunology of Leprosy (IMMLEP) (World Health Organization) bank. All aqueous solutions used in assays were filter sterilized. Enzyme assays. Enzyme assays were based on assays used in the very few previous studies on mycobacterial phospholipases but were modified by performing treatments such as ultrasonication and addition of a lysophospholipid with detergent properties, 2-lyso-phosphatidylcholine (lyso-PC) which should result in formation of micelles (5). Without these treatments, the phospholipids used in this study as enzyme substrates are in the form of liposomes (13). Thus, it was possible to present the phospholipase enzyme(s) with phospholipid substrates in different physical forms. The objective of testing many enzyme assays was to optimize detection of enzyme activity. This was done, for intact bacteria and their extracts, mainly with the relatively plentiful M. microti, and then selected assays were tried with M. Ieprae (Table 2). The assays were as follows. (i) Liposomes with dipalmitoylphosphatidic acid and PC (assay A). The assay with dipalmitoylphosphatidic acid and
VOL. 59, 1991
PHOSPHOLIPASE OF MYCOBACTERIUM LEPRAE
2783
TABLE 2. Optimizing the phospholipase assay with 1,2-di[ l4C 16:0]PC as substratea Controls: radioactiv-
Activity with the following:
ity corresponding to the following:
M. leprae
M. microti Assay system
Intact
bacteriab
Cell extractc
Intact
bacteriac
Cell
extractc
graycrol
fractionc
0_o 7
02.1
fractionb
A
Liposomes with DPPAd and PC 10 mM MES-Tween 80, pH 7 PBS, pH 7
0.2-0.5 (1-2) 0.2-0.4 (1-2)
13 (3)e ND
ND 1.5 (25)
ND ND
B
Micelles with DPPA, PC and lyso-PC 10 mM MES-Tween 80, pH 7 17-29 (74-128) 274-551 (70-141)NDD0-5 ND ND 0}2.5 PBS, pH 7 3.6-12.1 (15-20) 224 (50)e
09.
09.0
C
Liposomes with PC only 10 mM MES-Tween 80, pH 7 10 mM MES-Tween 80 and 5 mM CaCl2, pH 7 Glycerol-phosphate, pH 7
ND 8.1 (34)e
Micelles with PC only 10 mM MES-Tween 80, pH 7 Glycerol-phosphate, pH 7
1.2-3.7 (5-15) 2.6 (11)e
ND
32 (7)e 162-340 (36-76)
2.4-3.2 (3)
