Cholesterol oxidase (CO) and choline phosphohydrolase (CPH) exoenzymes were isolated from culture supernatants of Rhodococcus equi. ATCC 33701 and ...
Purification and Properties of Cholesterol Oxidase and Choline Phosphohydrolase from Rhodococcus equi R.S. Machang'u and J.F. Prescott
ABSTRACT
Cholesterol oxidase (CO) and choline phosphohydrolase (CPH) exoenzymes were isolated from culture supernatants of Rhodococcus equi ATCC 33701 and their hemolytic and cytotoxic activities examined. The purifications involved differential ammonium sulphate precipitation, ion exchange and gel filtration chromatography. A purification of 32.8-fold and a yield of 0.3% of CO were determined by synergistic hemolysis of sheep red blood cells (SRBC) presensitized with Staphylococcus aureus beta toxin. The enzymatic activity of CO was also demonstrated by oxidation of aqueous cholesterol suspensions. The activity of CO was reversibly inhibited by concentration. A purification of 412.4-fold and a yield of 1.7% of CPH were determined by hydrolysis of p-nitrophenyphosphorylcholine. Purity of both exoenzymes was confirmed by immunoblotting. On sodium dodecyl sulphate polyacrylamide gel electrophoresis, the CO had a molecular mass (Mr) of 60 kd and the CPH a Mr of 65 kd. Choline phosphohydrolase did not hydrolyse sphingomyelin. Sphingomyelinase C (SMC) activity was however demonstrated in concentrated culture supernatants. This dissociation of SMC from CPH activity indicates that R. equi produces two distinct phospholipase C exoenzymes, a CPH and a SMC. Both CO and CPH combined,
or individually, did not lyse native SRBC even with subsequent chilling of the cells at 4°C ("hot-cold" treatment). Purified CO lysed beta toxinsensitized SRBC. The CPH showed only minor hemolytic activity against such sensitized SRBC even at high concentrations. Combination of CO and CPH in lysis of beta toxin sensitized SRBC showed only minor additive rather than synergistic effects. Both exoenzymes were cytotoxic to a mouse macrophage cell line by the trypan blue dye exclusion assay, suggesting that these exoenzymes may contribute to the virulence of R. equi.
RESUME
Les proprietes hemolytique et cytotoxique des exoenzymes cholesterol oxydase (CO) et choline phosphohydrolase (CPH) provenant de surnageants de cultures de Rhodococcus equi ATCC 33701 ont ete etudiees. Pour la CO, une purification de 32,8 fois avec un rendement de 0,3 % a ete obtenue par hemolyse de globules rouges de mouton (GRM) presensibilies avec de la toxine beta de Staphylococcus aureus. L'activite enzymatique de la CO a aussi ete calculee par l'oxydation de suspensions aqueuses de cholesterol. Une purification de 412,4 fois avec un rendement de 1,7 No de la CPH a ete quantifiee par l'hydrolyse de la p-nitrophenyphosphorylcholine. La purete des deux enzymes a pu etre demontree par immuno-
empreinte. Des poids moleculaires de 60 kd et de 65 kd ont ete determines par electrophorese pour la CO et la CPH, respectivement. Bien que Ia CPH n'hydrolyse pas la sphingomyeline, une activite sphingomyelinase C etait presente dans les surnageants concentres des cultures demontrant que R. equi produit deux exoenzymes distinctes de phospholipase C. La CO et la CPH associees ou non n'ont pas produit de lyse des GRM meme lorsque refroidis a 4°C. Lorsque sensibilises par la toxine beta, la CO a hemolyse les GRM. La CPH n'avait qu'une faible activite hemolytique et ce, meme a hautes concentrations. Lorsque combinees ensemble, les deux exoenzymes n'ont produit qu'un effet additif de leur activite hemolytique. Ces deux exoenzymes ont demontre une activitg cytotoxique sur des macrophages de souris, laissant croire qu'elles peuvent etre associies A la virulence de cet agent. (Traduit par Dr Pascal Dubreuil)
INTRODUCTION Rhodococcus equi, a causal agent of foal pneumonia, is an intracellular parasite capable of surviving within and ultimately killing alveolar macrophages. Phagocytosed R. equi evade killing by equine alveolar macrophages in vitro, possibly by inhibiting phagosome-lysosome fusion
Department of Veterinary Microbiology and Immunology, Ontario Veterinary College, University of Guelph, Guelph, Ontario NIG 2W1. Reprint requests to Dr. J.F. Prescott. Supported by the Natural Sciences and Engineering Research Council and the Ontario Ministry of Agriculture and Food. Dr. R.S. Machang'u was supported by the Canadian Commonwealth Scholarship and Fellowship Plan. Submitted January 7, 1991.
332
Can J Vet Res 1991; 55: 332-340
and by inducing an early release of granular enzymes by the phagocytes (1-3). Rhodococcus equi produces membranolytic exoenzymes, originally called equi factors (4). These exoenzymes were shown to consist of a cholesterol oxidase (CO) (E.C. 1.1.3.6) and a sphingomyelin hydrolytic phospholipase C (PLC) (5). The CO causes synergistic hemolysis (SH) in vitro by inducing lysis of sheep red blood cells (SRBC) pretreated with Staphylococcus aureus beta toxin (sphingomyelinase C) or with Corynebacterium pseudotuberculosis phospholipase D (PLD) (E.C. 3.1.4.4) (5, 6). For hemolysis to occur, the phospholipids of the outer leaflet of the membrane bilayer must first be hydrolyzed by a phospholipase to make membrane cholesterol accessible for cleavage by the CO (7, 8). While the CO of R. equi involved in SH has been purified and characterized, the phospholipase C activity has been poorly characterized (5, 6). The interactive effects of these exoenzymes on cell membranes have not been fully described, although these effects may be relevant to the virulence of this organism. Diagnostic antibody tests based on inhibition of SH indicate the in vivo production of these exoenzymes, particularly the CO, and have been used in epidemiological studies of R. equi infections (9, 10). Antibody to these exoenzymes did not protect foals against the development of R. equi lung abscesses (11). Nevertheless, R. equi exoenzymes might be involved in the pathogenesis of R. equi infections through their membranolytic activity. Our approach to assessing the potential role for these exoenzymes in disease was to examine their activity in vitro, after isolating them in a purified state. We describe the purification from R. equi of CO and of a choline phosphohydrolase (CPH) (E.C. 3.1.4.3 ?) and then some properties of the purified enzymes, including activity against SRBC and a mouse peritoneal macrophage cell line.
MATERIALS AND METHODS PURIFICATION OF CHOLESTEROL OXIDASE
Cholesterol oxidase activity was monitored by SH assay of SRBC pretreated with S. aureus beta toxin (12). Rhodococcus equi ATCC 33701 was grown in 12 L of trypticase soy broth (Difco Laboratories, Detroit, Michigan) for 72 h at 37°C with shaking (100 rpm). Cell free supernatant obtained by filtration through a 0.45 um membrane (about 11.5 L) was concentrated to 800 mL by ultrafiltration through a 10,000 molecular weight cut off membrane (Pellicon IPTG, Millipore Corporation, Bedford, Massachusetts) and then filter sterilized (0.45 jsm, Millipore). Ammonium sulphate (560 g/L) was added to the concentrated culture supernatant and the mixture was shaken (60 rpm) at room temperature (RT) for 30 min until the salt was dissolved. Precipitation was continued over 24 h at 4°C. The precipitate was collected by centrifugation (8000 x g, 20 min, 4°C), dissolved in 65 mL of 20 mM Tris-HCl, pH 8.0, containing 4 mM cysteine-HCl and sodium azide 0.03% (w/v) (buffer 1), and dialyzed overnight against 1 L of the same buffer at 4°C. The dialysate (165 mL) was filter sterilized (0.45 Am) and purified further by batch ion exchange. Diethylaminoethyl (DEAE) Sephadex A-50 (30 gm) was swollen overnight in 2 L of buffer 1 and packed in two chromatographic columns (2.5 x 100 cm) connected in series. The columns were equilibrated overnight with the same buffer and monitored by an ultraviolet detector (Gilson UV 111 B, Middleton, Wisconsin) at 280 nm until a zero baseline was recorded (Fisher Recordall 5000, Bausch and Lomb, Austin, Texas). Two hundred mL of equilibrated beads were removed from the first column and mixed with the crude CO material by gentle shaking (60 rpm) for 1 h in a beaker. The mixture was then loaded back onto the column, the column washed with one column-volume (about 500 mL) of buffer 1, and then eluted with a NaCl gradient (0-2.0 M). The gradient was applied with a peristaltic pump (P-3, Pharmacia, Uppsala, Sweden) at a flow rate of 30 mL h-' while collecting 15 mL fractions (Ultrovac, LKB, Bromma, Sweden). Elution was stopped when a zero baseline was recorded in about 30 h.
Pooled fractions with SH activity from batch anion exchange (390 mL) were concentrated using polyvinylpyrrolidone (PVP) (Sigma Chemicals, St. Louis, Missouri) to 70 mL, then precipitated in a glass flask with two volumes of cold (- 70°C) acetone for 90 min at 4°C. The mixture was centrifuged (1,000 x g, 20 min, 4°C) and the precipitate dissolved in buffer 1. This solution was centrifuged (10,000 x g, 25 min, 4) to remove undissolved debris and the supernatant retained for further purification. Sephadex G-25 (Pharmacia) (30 gin) was swollen in 500 mL buffer 1 for 6 h, loaded onto a column (1.6 x 100 cm) and equilibrated overnight with the same buffer. The material (56 mL) obtained after acetone precipitation was applied onto the column and eluted with buffer 1, while collecting 15 mL fractions at a flow rate of 15 mL h- l. Fractions with SH activity were pooled, filter sterilized, PVP concentrated to 10 mL, and dialyzed for 12 h against 1 L of 50 mM morpholinethanosulphonic (MES) buffer, pH 5.7 (Aldrich Chemicals, Milwaukee, Wisconsin). Dialyzed CO material (26 mL) was applied in 5 mL aliquots to a Fast Protein Liquid Chromatography (FPLC)Mono S HR 5/5 cation exchange column (Pharmacia) equilibrated with 50 mM MES buffer. The column was eluted with 35 mL of a NaCl gradient (0-0.5 M), in the 50 mM MES buffer, while collecting fractions of 1 mL. Active fractions (11 mL) were pooled, supplemented with bovine serum albumin (BSA) at a final concentration of 0.1 mg/mL, and lyophilized. The lyophilate from cation exchange-purified CO material was reconstituted with 1 mL of 50 mM Tris-HCl, pH 8.0 (buffer 2) and applied in 200 uL aliquots to a FPLCSuperose 12-HR 10/30 (Pharmacia) gel filtration column equilibrated with buffer 2. Elution was with buffer 2 and active fractions (9 mL) were pooled and supplemented with BSA as described. This material, purified CO, was frozen (- 70°C) in small aliquots. Standardization of CO was based on SH activity. One hemolytic unit (HU) was defined as the amount of CO which caused 50%o lysis of an equal volume of 2% SRBC preincubated 333
with 2 HU/mL of S. aureus beta toxin (12). The enzymatic activity of CO was confirmed by oxidation of aqueous cholesterol suspension by CO (5), using Streptomyces sp. CO (Sigma) as the reference standard. The total protein content of purified material was determined by the micro-Lowry method using BSA as the standard (13). PURIFICATION OF CHOLINE PHOSPHOHYDROLASE
The CPH was prepared from 16 L of a 72 h culture of R. equi ATCC 33701. Bacteria were removed as described above for CO preparation and the cell free supernatant concentrated by ultrafiltration to 1 L. The concentrated supernatant was then precipitated by ammonium sulphate (560 g/L) and the precipitate reconstituted with buffer 2 to 200 mL and then fractionated by batch ion exchange as described above for CO isolation. Fractions with CPH activity from batch anion exchange were pooled, filter sterilized, concentrated to 210 mL by PVP, and desalted on a Sephadex G-25 (Pharmacia) column. Active fractions were pooled, PVP concentrated to 12 mL, and applied to a Sephadex G-100 S (Pharmacia) column (2.5 x 70 cm) equilibrated with buffer 2, and eluted with the same buffer. Active fractions obtained by gel filtration were pooled, PVP concentrated to 12 mL, and loaded onto a Mono Q-HR 5/5 FPLC column (Pharmacia) equilibrated with buffer 2, and eluted with a NaCl gradient (0-0.5 M) in the same buffer. Fractions with CPH activity were pooled (24 mL), desalted on a Econo-Pac DG 10 column (Bio-Rad), loaded onto a Mono Q-HR 5/5 FPLC column, and anion exchange purification was repeated as described above. Fractions with peak activity were pooled (9 mL), lyophilized, reconstituted with buffer 2 (600 AL), loaded onto a FPLC Superose 6-HR 10/30 column in 200 AL aliquots, and separated by gel filtration. Activity of the CPH material was determined by the p-NPPC hydrolysis assay using the PLC of Clostridium perfringens (Sigma) as the reference standard. One unit of CPH activity was defined as the amount of this exoenzyme that liberated 1 334
Amole
of water soluble organic phosphorus from egg yolk L-caphosphatidylcholine (14). The total protein content was determined by the micro-Lowry assay (13). Sphingomyelinase activity was also examined by the method of Colley et al (7). The purified CPH was stored frozen (- 70°C) in small aliquots.
POLYACRYLAMIDE GEL ELECTROPHORESIS AND IMMUNOBLOTTING
buffer was used for the second antibody. The nitrocellulose membranes with immune complexes were developed with a reagent mixture containing 10 mg each of naphthol phosphate and Fast Red (Sigma) dissolved in 10 mL of 0.1 M Tris buffer, pH 9.2. INTERACTION OF PURIFIED CO AND CPH WITH SRBC
(a) Effects ofpurified CO and CPH on SRBC: Twofold dilutions of 100 tL Discontinuous sodium dodecyl of purified CO (starting at sulphate polyacrylamide gel electro- 5120 HU/mL, final concentration phoresis (SDS-PAGE) was done under range 256 to 0.13 HU/well) and reducing conditions (15) on 4% stack- 100 AL CPH (starting at 860 U/mL, ing gel and 120o separating gels of final concentration range 43 to 0.75 mm thickness in a Protean II 0.02 U/well) were prepared in two Mini-gel apparatus (Bio-Rad, rows of 12 wells each of a "U"-bottom Richmond, California). Samples from microtiter plate (Linbro, Hamden, each purification step were mixed with Connecticut) containing 100 tL of reducing buffer (15), boiled for 4 min diluent buffer (PBS, 1 mM MgSO4, at 90°C, and aliquots (10 AL) loaded pH 7.4). A third row contained 100 AL onto individual lanes. One lane was of serial twofold dilutions of preparaloaded with 5 yL of denatured low tions of S. aureus beta toxin (starting range molecular weight standards (Bio- at 1280 HU/mL, final concentration Rad). Separation was for 40 min at range 64 to 0.02 HU/well) standard200 V and gels were stained with ized by "hot-cold lysis" assay (17). Coomassie blue (0.1 7o) (15). Gels were A fourth row containing diluent alone also stained by silver reagent was included as a control. One hundred
(Bio-Rad). Immunoblotting was done as described by Towbin and Gordon (16). After SDS-PAGE, proteins were electrophoretically transferred to a nitrocellulose paper for 1 h at 100 volts in a Mini-blot transfer apparatus (Bio-Rad). For CO, transferred protein were analyzed against homologous rabbit hyperimmune serum and also against heterologous rabbit serum against the CO of Streptomyces sp. (Sigma). For CPH, a homologous hyperimmune rabbit serum was used. Rabbits were immunized by intramuscular injection of about 50-100 jig of purified enzyme followed by a booster vaccination of the same quantity in an equal volume (400 AL) of Freund's incomplete adjuvant three weeks later; rabbits were bled for serum two weeks later. Skimmed milk (10%o) in Tris buffered saline (TBS) (0.5 M NaCl, 0.02 M Tris, pH 7.4) was used as the sample buffer and sera were prepared in 1/25 dilutions in this buffer. Alkaline phosphatase-conjugated goat antirabbit IgG (1/1000 w/v) in the TBS
yL of native SRBC (207) (Woodlyn Laboratories, Guelph, Ontario) were added to each well and the plates examined for "hot-cold" lysis (17) by incubating at 37°C for 1 h with shaking (100 rpm) and subsequently at 4°C for 30 min. Hemolysis was determined spectrophotometrically by measuring hemoglobin release at 540 nm (12). To examine effects of CO and CPH on sensitized SRBC, serial doubling dilutions of the exoenzymes were prepared in two rows of a "U" microtiter plate as described above. A third row (control) contained diluent only. Sensitized SRBC (100 AL) were added to all wells and the plate incubated for 2 h at room temperature (RT) then examined for hemolysis spectrophotometrically. Sensitization was by incubating SRBC (2%), washed and suspended in PBS, with S. aureus beta toxin at a final concentration of 2 HU/mL of SRBC for 1 h at RT with shaking. One hemolytic unit of S. aureus beta toxin was defined as the amount that gave 50%o hemolysis of an equal volume of SRBC (27o) after "hot-cold" treatment (17). The effects on sensitized
SRBC of more concentrated CO (starting at 1024 HU/well), and CPH (starting at 497.5 U/well) were examined in a similar manner. (b) Effects on SH of a mixture of CO and CPH: Serial twofold dilutions of a mixture containing 50 AL of CO (starting at 5120 HU/mL) and 50 AL (starting at 860 U/mL) of CPH were prepared across a row of 12 wells of a "U"-bottom microtiter plate (Linbro) containing 100 AL of diluent buffer. In a second row of wells, containing 50 AL of diluent buffer, serial twofold dilutions of CO (starting at 5120 U/mL) were prepared and then 50 L of 50 mM Tris-HCl, pH 7.4 containing 1 mM MgSO4, was added to each well to obtain a final volume of 100 AL/ well of the mixture. A third row contained 50 AL of serial twofold dilutions of CPH (starting at 860 U/mL), also with equal volumes of the Tris-HCl buffer. At the end, row 1 contained dilutions of CO and CPH mixture (range 128 to 0.03 HU/well of CO, and 21.5 to 0.01 U of CPH/well), row 2 contained CO only (range 128 to 0.03 HU/well) and row 3 contained CPH only (21.5 to 0.01 U/well). The control row (Row 4) contained 100 ItL of the Tris-HCl buffer, without exoenzymes. Equal volumes (100 AL) of sensitized 2% SRBC in phosphatebuffered saline (PBS) were then added to all wells and the plate incubated (2 h at RT) with shaking. Hemolysis was determined spectrophotometrically. EFFECT OF CO AND CPH ON A MOUSE PERITONEAL MACROPHAGE TUMOR CELL LINE
Mouse peritoneal tumor cell macrophages (ATCC TIB 71 RAW 264.7) were obtained from Dr. S. Rosendal, University of Guelph. Cells were grown in cell culture bottles containing 10 mL of RPMI medium from a stock preparation (500 mL of RPMI, penicillin 10,000 I.U. + streptomycin 40 mg, 5 mL of 200 mM L-glutamine, 5 mL bovine calf serum, and 0.25 g (1:250) trypsin) (Flow Laboratories, Mclean, Virginia). Culture bottles were incubated in 5 % CO2 and cells were harvested after a monolayer was formed (about 72 h). For cell harvest, the culture medium was removed and 3 mL of antibiotic-trypsin-versene (ATV) was added for 1 min. The ATV solu-
tion was then removed, and cells detached by a light tap on the bottle. Cells were re-suspended in 10 mL of RPMI and adjusted to 5 x 103 cells/mL RPMI. One hundred AL of the cell suspension were placed in two rows of three wells each in a 96-well cell culture plate and placed in a CO2 incubator for 1 h at 37°C to allow the cells to attach. After attachment, the first row was supplemented with 100 AL of CO (512 U/well) and the second row (control) with 100 ttL of RPMI. To examine CPH effects cells were prepared in another cell culture plate as described above and rows 1 and 2 supplemented with 100 AL CPH (1 U/well) and with 100 AL RPMI respectively. The CO and CPH were prepared in 50 mM Tris-HCl, pH 7.4 and 2 mL aliquots were dialyzed overnight against 200 mL of RPMI before use. The mixtures were incubated for a further 4 h, then centrifuged for 5 min at 1000 rpm, and 100 AL of the supernatants in all wells replaced with 100 lsL of 0.47o trypan blue dye solution in PBS. Plates were then incubated at RT for 5 min and again centrifuged for 5 min. Aliquots (150 AL) of the supernatant were removed from wells and the relative proportions of live and dead cells (dead cells retained dye) were recorded in a "blinded" fashion by two observers, among at least 300 cells attached to the wells. Counts of dead cells were compared statistically against controls by the analysis of variance procedure using a statistical analysis system (SAS PROC. ANOVA) for personal computers (SAS Institute Inc., Cary, North Carolina). Five independent studies were done with CO
and three studies with CPH. Cells incubated with boiled CO were also compared to cells incubated with RPMI for proportions of dead cell counts in two independent studies. RESULTS PURIFICATION OF CHOLESTEROL OXIDASE
The R. equi culture supernatant was initially ultrafiltered to give a concentrated retentate enriched with CO and CPH. The progressive purification of CO, as determined by SH, is summarized in Table I. Figure IA shows a representative elution profile of CO fractions following desalting of acetone precipitated material on Sephadex G-25 column. Synergistic hemolysis was present only within the small peak. Desalting on Sephadex G-25 removed residual acetone and markedly concentrated the CO (Fig. 2, lane 5). Cation exchange (Fig. iB) isolated CO to near homogeneity. Figure 1 C shows a characteristic elution profile of this cation exchange fractionated CO on Superose 12-HR gel filtration chromatography, which purified the exoenxyme. Purified CO (SH activity 2560 HU/mL) induced total lysis of sensitized SRBC. It also oxidized aqueous suspensions of cholesterol. This material was free of CPH. A recovery of 0.3%o and a purification of 32.8-fold was recorded by the SH hemolytic activity of the CO material (Table I). Stabilization of the CO with BSA was required for storage of material at room temperature or at - 70°C, since the preparation otherwise soon lost activity (data not shown).
TABLE I. Purification of cholesterol oxidase from Rhodococcus equi ATCC 33701 with activity determined by synergistic hemolysis of sheep red blood cells Specific
Total
Purification
(mg)
activity (HU/mg of protein)
Yield (%)
8.2 x 106
5.7 x 103
1.4 x 103
100
3.4 x 106
1.2 x 103
2.8 x 103
41
4.0 x 106
2.9 x 102
1.4 x 104
48.7
10
7.2 x 105
1.3 x 101
5.5 x 104
8.8
39.3
1.1 x 105
1.3
8.5 x 104
1.3
60.7
2.3 x 104 L
0.5
4.6 x 104
0.3
32.8
step
Vola (mL)
activity (HU)
Supernatant (> 10,000 MW)
800 165
(NH4)2S04
precipitation
Total protein
Fold
puriflcation
1 2.0
DEAE ion 390 exchange Sephadex G-25 70 Cation exchange 11 (FPLC) Gel filtration 9 (FPLC) aStarting volume of culture, 12
335
A
Figure 5 shows a Coomassie blue stained SDS-PAGE analysis of CPH fractions taken at different steps of purification. Lane 8 shows a single band of Mr 65 kd which was the purified CPH. A silver stained gel of purified CPH (lane 9) showed, in addition to the CPH band, a faint band at about 66 kd. A positive immunoblot reaction of purified CPH with homologous antiserum showed the apparent purity of this exoenzyme material (Fig. 5, lane 10).
B
1.o
0.9
FI
z 0.7 a Q
0.6.
0
0.5-
0.3
0.4
.
0.3
-
0.2
-
0.1
-
EFFECTS OF CO AND CPH ON SRBC AND MOUSE MACROPHAGES
C
-
-260 a
IGO
0.2
0.1 -
, K_
.... ............
5
10
1S
10 NU MOBER 5
F R IACT ION
Is
20
5
.40
2S
Fig. L.A. Representative elution profile from Sephadex G-25 column of materials containing CO obtained after batch (DEAE) anion exchange. Fractions with CO activity (+ +) were within the first of the two peaks recorded. B. Representative edution profile of Mono S-HR cation exchange chromatography of acetone precipitated and desalted (Sephadex G-25) material containing CO. Approximately 1 mg protein was applied to the column and eluted with a 0-0.5 M NaCl gradient in 50 mM morpholinethanosulfonic add (pH 5.7) at a flow rate of 0.5 mL/min. Fractions of 1 mL were coflected. Boxed areas show fractions with SH activity. C. Representative edution profile from Superose 12 HR gel filtration of materials containing CO obtained after cation exchange. Approximately 0.5 mg protein was applied to the column and eluted with 50 mM Tris-HCI buffer, pH 8.0, at a flow rate of 0.5 mL/min. One mL fractions were coilected and fractions (boxed) with peak hemolytic activity (pooled) represented purified CO.
Analysis by SDS-PAGE of samples taken at different steps of purification is shown in Fig. 2. The CO band was estimated to be 60 kd. The only other band present in purified CO sample (lane 7) was BSA (Mr 67 kd) which was added to stabilize the CO material. The immunoblot reaction of purified CO material with homologous antiserum (lane 9), and with hyperimmune rabbit serum to Streptomyces sp. CO (lane 10), demonstrated the purity of the material and confirmed the identity of the 60 kd band as CO. Figure 3 shows the inhibition of SH activity by concentration of CO. Such inhibition was evident throughout the stages of purification of the
is summarized in Table II. Figure 4A shows the elution profile following Mono Q-HR anion exchange of CPH material and Fig. 4B shows the elution profile following a second anion exchange under similar conditions. Figure 4C shows the elution profile after Superose 6-HR gel filtration, which purified the CPH to near homogeneity. A 1.7% yield and a purification of 412.4-fold was achieved (Table II). Purified CPH did not oxidize aqueous cholesterol suspensions. The purified CPH material hydrolyzed p-NPPC to release p-nitrophenol but did not hydrolyze sphingomyelin (Table III). Crude ultrafiltered culture supernatant materials from two difexoenzyme. ferent batches revealed the presence of SMC (364 and 544 U/mL, respecPURIFICATION OF CHOLINE tively). The isoelectric point (pI) of PHOSPHOHYDROLASE CPH was estimated at 4.5 by chromaThe progressive purification of CPH tofocusing (data not shown).
336
No hemolysis was recorded after incubation (2 h at RT) of native SRBC with relatively large amounts of CO (256 HU/100 AL 2% SRBC) or CPH (43 U/100 yL 20o SRBC). "Hot-cold" treatment did not facilitate lysis by CO or CPH of unsensitized cells, but gave total lysis of SRBC incubated with beta toxin of S. aureus (32 to 64 HU beta toxin/100 AL 2% SRBC). Total lysis of sensitized SRBC was recorded with CO (128 to 32 HU CO/100 AL 2% SRBC), whereas CPH alone (21.5 U/ 100 AL 2% SRBC) caused only slight (about 15Gb) lysis. Absorbance readings of the wells containing concentrated CO (256 to 1024 HU/100 AL 2% SRBC) recorded values below baseline (control wells) levels, Fig. 3. Synergistic hemolytic activity was however progressively restored with serial doubling dilutions of the CO material to a peak (80% hemolysis) at 80 HU/ 100 AL 2% SRBC of the CO. Only slight SH (about 15%) was recorded with the concentrated CPH (497.5 U/ 100 IAL 2% SRBC). This slight hemolyis was equivalent to that seen with lower concentrations (21.5 U/ 100 ML 2% SRBC) of the CPH. Figure 6 shows the SH effect caused by serial dilutions of a mixture containing 128 HU of CO and 21.5 U of CPH. Hemolysis (50%) recorded with the mixture was similar to that caused by CO alone, although a slightly more intense lysis was recorded with the mixture than with the concentrations of CO alone. Table IV shows the mortality of mouse peritoneal macrophage tumor cells exposed to 512 HU of CO in five studies or to 1 U of CPH in three studies. The effect of larger quantities
Effect of CO c n
n on lye
02 O.1S 2 3
4 5
6 7
8 9 10 11 12 13 14 15 16 17 1s1 s 20 21 22 42) ofpur0Id CO
Dlu St
a
(HU/ws):; 129: (9 1024
Fig. 3. Variations of synergistic hemolysis with CO concentration. Serial doubling dilutions of concentrated CO, starting with 1024 HU/well (*); serial doubling dilutions of nonconcentrated CO, starting with 128 HU/well (U); sensitized SRBC without CO (A).
significant difference in the number of dead macrophages between samples treated with RPMI or with boiled CO. DISCUSSION Fig. 2. SDS-PAGE of samples of Rhodococcus equi culture supernatants coilected at different steps of CO purification. Lane 1: low range molecular weight standards. Lane 2: 50 jig concentrated culture supernatant. Lane 3: 30 ytg ammonium sulphate precipitated material. Lane 4: 5 tg DEAE (batch) separated material. Lane 5: 10 itg acetone precipitated and desalted (Sephadex G-25) material. Lane 6: 1 j4g Mono S-HR cation exchange purified material. Lane 7: 10 ,g Superose 12-HR gel filtration purified CO. Lane 8: 5 ug silver stained SDS-PAGE of the purified CO. Lanes 9 and 10 show immunoblot reactions of purified CO material with homologous rabbit serum and with antiserum to Streptomyces sp. CO, respectively. The CO band (arrow) is shown at 60 kd. The band at 66 kd (lanes 6 to 10) is BSA.
TABLE II. Purification of choline phosphohydrolase from Rhodococcus equi ATCC 33701 with activity determined by hydrolysis of p-nitrophenylphosphorylcholine
Purification step
Supernatant (> 10,000 MW)
(NH4)2S04
precipitation
Specific activity
Vola (mL)
Total activity (U)
Total protein (mg)
(U/mg of protein)
Yield
(qOl)
Fold purification
1000
6.9 x 105
7.1 x 103
9.7 x 101
100
I
200
3.9 x 105
1.5
103
2.6 x 102
56.5
2.7
210
2.8 x 105
7.8 x 102
3.6 x 102
40.6
3.7
3.0 x 104
7.5
4.0 x 103
4.3
41.2
2.4 x 104
7.2 x 10-1
3.3 x 104
3.5
343.8
2.2 x 104
5.6 x 10-1
3.9 x 104
3.2
402.0
1.2 x 104
3.0 x 10-1
4.0 x 104
1.7
412.4
x
DEAE ion
exchange Gel filtration
12 Sephadex G-100 S Anion exchange I 24 (FPLC) Anion exchange II 9 (FPLC) Gel filtration 9 (FPLC) aStarting volume of culture 16 L
of CPH was not examined because of macrophages was significantly higher the small quantity of purified enzyme (p < 0.05) in samples treated with CO available. The mean number of dead or CPH than in controls. There was no
In this study CO was purified to homogeneity and CPH to near homogeneity from the culture supernatant of R. equi ATCC 33701. The molecular mass of CO was estimated by SDSPAGE analyses to be 60 kd, which is in good agreement with the 61 kd reported by Linder and Bernheimer (5). Batch DEAE anion exchange (pH 8.0) was used because preliminary studies had shown CO to be positively charged at pH 8.0 (i.e. pI > 8.0). The pI of CO was earlier reported to be between 9 and 10 (5). The unexpected increase in total SH activity (Table I) after batch DEAE anion exchange may reflect removal by DEAE of components inhibitory or competitive to CO. The CO eluted from cation exchange readily lost activity, possibly due to the MES buffer or the low pH (5.7). Additions of BSA (0.1 mg/mL) (18) stabilized the CO material when stored frozen - 70°C. One interesting feature was the failure of concentrated CO material to induce SH (Fig. 3). Dilution of CO, however, restored SH. This failure of concentrated CO to induce SH was possibly due to aggregation of the exoenzyme material following concentration. It was also interesting that the supernatants of SRBC samples treated with concentrated CO gave hemoglobin absor-
337
A ,O.S
*0.4
0.3
02
F01
z4 I
B
6.
la
0. ml
i
C W_t~w 1I
as
&4 60 is
Is
is
30
49
AS
IC
0.4
L
0.2
2000
F A
0 S 10
1S
20
2S 300
15
20
2S
3
01.1
0
5
10
FRACTION
30
NUMBER
Fig. 4.A.Representative elution profile of Mono Q-HR anion exchange chromatography of post Sephadex G-100 S material containing CPH. Approximately 1 mg protein of sample was applied to the column and eluted over a NaCI gradient (0-0.5 M) in 20 mM Tris-HCI buffer at a flow rate of 0.5 mL/min. One mL fractions were collected. Boxed areas show fractions containing CPH activity. B. Representative elution profile following a second Mono Q-HR anion exchange of materials containing CPH. Fractions obtained after the first anion exchange were pooled, desalted and 200 ug protein were applied to the column. Elution was with a 0-0.5 M NaCI gradient in 20 mM Tris buffer at a flow rate of 0.5 mL/min while collecting 1 mL fractions. Fractions with peak activity (boxed areas) were pooled. C. Representative elution profile of CPH material following Superose 6 HR gel filtration of materials containing CPH. Fractions obtained after the second anion exchange were pooled, lyophilized, and reconstituted with Tris-HCI, pH 8.0 containing sodium azide (0.03'!) (w/v). About 200 sg protein was applied to the column and eluted in 1 mL fractions with the same buffer. Fractions with peak activity (boxed) represented purified CPH.
bance readings lower than the controls, suggesting absorption of the hemoglobin released by sensitized SRBC or its cleavage to nonabsorbing components.
p-NPPC hydrolysisb (U)
Hemolysisa (HU)
Exoenzyme
L)
338
*§':t,.0;2M\"q_
perfringens PLC, Rhodococcus equi CPH and Staphylococcus aureus beta toxin
I..
0z
on increasing activity of concentrated CO (Fig. 3) coupled with the need to dilute the exoenzyme to reach the linear range. However, since R. equi supernatant also has SMC which might contribute to hemolysis (5), the use of SH to determine CO activity was suitable only for purified fractions. For this reason, estimates of purifica-
TABLE II. Hemolytic, p-NPPC and TPNAL-sphingomyelin hydrolytic activity/mL of Clostriduwm
.0.
3
Synergistic hemolysis was a useful measure of CO activity and was used rather than a direct cholesterol oxidation assay (5), since a standardization curve of the latter was linear over only a narrow range. In addition, results obtained on a direct cholesterol oxidation assay were difficult to interpret probably because of the dilution effect
Sphingomyelin hydrolysisC (HU)
C. perfringens PLC 2560 100 2043 R. equi purified CPH NDd 860 ND S. aureus beta toxin 2560 ND 2019 aHemolysis was assessed after incubation of mixtures of the exoenzymes with native SRBC for 1 h at 37°C and subsequently at 4°C for 30 min bp-NPPC hydrolysis was determined by calibration against a standard C. perfringens PLC COne unit of sphingomyelinase activity gave an OD reading of 0.023 at 330 nm following hydrolysis of TPNAL-sphingomyelin dNot detected 1
.3;4
2
5S;
..
:;....
'V.
-:7 M. ':1:: ~:-
z: :.
6.
*31. *I: :...... '..:'!. ...
,
.:
,-.
...
.:
.:
..
_
:4 e .:
' ,,'_
.' ,:
__Ft.
f
Fig. 5. SDS-PAGE of samples of Rhodococcus equi culture supernatants collected at different steps of CPH purification. Lane 1: Low molecular weight standards; Lane 2: 70 jig concentrated culture supernatant; Lane 3: 75 tg ammonium sulphate precipitated material. Lane 4: 30 Ag DEAE (batch) anion exchange material. Lane 5: 5 Ag Sephadex G-100 S purified CPH. Lane 6: 3 Ug Mono Q-HR anion exchange purified CPH. Lane 7: 2 ug Mono Q-HR anion exchange (repeated) purified CPH. Lane 8: 1 1tg Superose 6-HR gel filtration purified CPH. Lane 9: 1 Ug silver stained SDS PAGE of Superose 6-HR gel filtration purified CPH. The band at 65 kd (arrow) is CPH. Lane 10 shows an immunoblot reaction of purified CPH with homologous rabbit antiserum.
Interation of a CO + CPH mixture with sensited SRBC
0
a
16
32
64
126 S12 1024 206 DU6kinos do00 + CP6 mnibus
4066
3192 16364
Staing _tn _sw Maw : CO 126 HU; CPH 21.S U
Fig. 6. Interactions of R. equi CO and CPH with sensitized SRBC. One hundred 1tL aliquots of a 2% suspension of SRBC sensized with beta toxin (2 HU/mL) were incubated (2 h, RT) with equal volumes of serial doubling dilutions of a mixture starting with 128 HU of CO and 21.5 U of CPH/well) (U) and then examined for hemolysis. Dilutions of CO (starting with 128 HU/well) with Tns-HCI (*); dilutions of CPH (starting with 21.5 U/well) with Tns-HCI (4). Beta toxin sensitized SRBC without exoenzymes ( v ). Y axis, absorbance (540 nm).
Linder and Bernheimer (5) purified CO to near homogeneity but did not report percentage yields or fold purification of this exoenzyme. The present purification procedure for R. equi CO has extended previous studies (5, 6) by describing the comprehensive purification, yield, and recovery of CO material, by showing the apparent instability of CO which could be overcome by the addition of BSA, by showing failure of concentrated CO to lyse sensitized SRBC, and by demonstrating a hemoglobin binding or absorption effect of concentrated CO. Production of CO has also been demonstrated among organisms of the genus Streptomyces and in other genera such as Brevibacterium, Corynebacterium, Nocardia and Mycobacterium (20). Most of these organisms are inhabitants of soil and have membranes rich in sterols. Cholesterol oxidase may be used by these organisms to acquire sterols from cholesterol in the environment for synthesis of cellular components, or as a source of carbon for internal metabolism (20, 22). Cross-reaction of antibody to the CO of Streptomyces sp. with the CO of R. equi (Fig. 2) suggests that the CO of these soil organisms, genetically related to R. equi, may be closely related. The initial steps in CPH purification were similar to those of CO purification. For the final steps, repeated anion exchange was done to obtain a nearly purified CPH preparation
(Fig. 4B). An additional, barely visible, band (Mr 66 kd) was possibly BSA from the filtration systems (Fig. 5, lane 8). Purified CPH material remained stable when frozen (- 70°C). This is the first report describing the isolation from R. equi of a CPH in a highly purified state. This CPH exoenzyme (Mr 65 kd, pI 4.5) is clearly distinguishable from the broad substrate-specific PLC (Mr 74 kd, pl about 6.5) with apparent SMC activity described earlier (5, 6). The CPH isolated in the present study did not hydrolyze sphingomyelin (Table III). Since sphingomyelinase C activity was demonstrated in culture supernatants, R. equi therefore produces at least two distinct exoenzymes with PLC activity: a SMC (5, 6) and a CPH.
No lysis of native SRBC was recorded even with relatively large amounts of CO or CPH. Chilling did not facilitate lysis of native SRBC by CO or CPH. This showed that CO and tion and percentage yield of CO based CPH were not themselves hemolytic to on SH may be erroneous. Neverintact SRBC and were not "hot-cold" theless, the approximately 30-fold conhemolysins. With sensitized SRBC, centration required to purify CO suptotal lysis was achieved with CO alone ports earlier reports (5,19) that this (32 to 64 HU). Choline phosphoexoenzyme is produced in large hydrolase (43 U) gave only slight amounts in culture supernatant. The (15%) lysis, which was not enhanced CO could be readily identified in culby concentration. Hemolysis of beta ture supernatants (Fig. 2, lane 2). The toxin sensitized SRBC by mixtures of discrepancy for loss of purification CO and CPH did not appreciably diffrom cation exchange (60.7) to final gel fer from hemolysis by CO alone filtration (32.8) may be explained by (Fig. 6) showing that the interactive minor loss of activity of the final effects were at best additive but not purified enzyme. synergistic. The 15% lysis of SRBC by CPH presumably occurred after cleavage of TABLE IV. Percentage of mortality of mouse peritoneal macrophage tumor cell line (RAW 264.7) phosphorylcholine headgroups from a exposed to R. equi cholesterol oxidase (CO), choline phosphohydrolase (CPH), or RPMI (control) CPH-susceptible membrane phosphoas assessed by the trypan blue dye exclusion assay lipid present in small amounts in the Treatment SRBC membrane. This phospholipid CO Study CPH RPMI Heated CO might be phosphatidylcholine (23) because R. equi CPH did not preferen1 29.6a 16.1 tially cleave sphingomyelin, as demon2 28.1 8.0 3 22.1 13.5 strated by the absence of "hot-cold" 4 17.7 12.3 and the inability to hydrolyze lysis 5 12.7 1.9 sphingomyelin (Table III). Although 6 33.3 16.0 CPH did not cause intense hemolysis, 7 26.8 9.9 8 its potent p-NPPC hydrolytic effect 29.3 9.9 9 8.2 6.6 suggest that CPH will attack mem10 5.2 6.3 branes containing choline phosphoSignificanceb p < 0.05 p < 0.05 NS lipids. The production of this exoaEach value represents the percentage of dead cells in not less than 300 cells counted 'The significance of the principal with respect to the control samples was assessed by analysis of enzyme in minor amounts compared to CO suggests that CO is the major variance exoenzyme of R. equi involved in SH. NS, not significant
339
It would be interesting, however, to isolate and characterize the R. equi SMC enzyme and examine its effect on SRBC and interaction with CO and CPH. Cholesterol enrichment of SRBC has been shown to facilitate lysis by CO without participation of other agents such as phospholipases C (24). Since lysis of membranes by CO and PLC depends on their constitution, we examined a macrophage cell line as another target of these exoenzymes. The CO caused significant killing of the mouse macrophage tumor cell line, when compared to controls without CO or with heated exoenzyme (Table IV). This toxic effect indicated that failure of CO to lyse native SRBC may not be representative of the exoenzyme's interaction with all biological membranes, although the SRBC were not examined for other less degradative membrane alterations. It is possible, therefore, that the presence of larger amounts of cholesterol in RAW 264.7 cells may have facilitated cytotoxicity by CO of the macrophages. One other possible cytotoxic mechanism might have been indirectly through hydrogen peroxide production following CO action on serum present in the cell culture medium. The cytotoxicity of CO on the mouse macrophage tumor cell line suggests that CO alone or in conjunction with other factors may have a role in the pathogenesis of R. equi pneumonia through cell damaging mechanisms. Cytotoxicity by CPH suggests that the macrophage membrane was also affected by this exoenzyme, thus also implicating CPH in the pathogenesis of R. equi through macrophage damage. Synergistic action on the mouse tumor cell line by CO and CPH was not examined. We were also unable to conclude whether the CO or the CPH was more cytotoxic. By demonstrating cytotoxic effects of CO and CPH on mouse macrophages, this
340
work suggests that these R. equi exoenzymes may be involved in the virulence of this organism by their action on host macrophages or other cells, and justifies further studies of the pathogenic interactions of R. equi exoenzymes with foal alveolar macrophages.
REFERENCES 1. HIETALA SK, ARDANS AA. Interaction of Rhodococcus equi with phagocytic cells from R. equi-exposed and non-exposed foals. Vet Microbiol 1987; 14: 307-320. 2. ZINK CA, YAGER JA, PRESCOTT JF, FERNANDO MA. Electron microscopic investigation of intracellular events after ingestion of Rhodococcus equi by foal alveolar macrophages. Vet Microbiol 1987; 14: 295-305. 3. YAGER JA. The pathogenesis of Rhodococcus equi pneumonia in foals. Vet Microbiol 1987; 14: 225-232. 4. FRASER G. The effect on animal erythrocytes of combinations of different diffusible substances produced by bacteria. J Pathol Bacteriol 1964; 88: 49-53. 5. LINDER R, BERNHEIMER AW. Enzymatic oxidation of membrane cholesterol in relation to lysis of sheep erythrocytes by corynebacterial enzymes. Arch Biochem Biophys 1982; 213: 395 404. 6. BERNHEIMER AW, LINDER R, AVIGAD L. Stepwise degradation of membrane sphingomyelin by corynebacterial phospholipases. Infect Immun 1980; 29: 123-131. 7. COLLEY CM, ZWAAL RFA, ROLEOFSEN B, VAN DEENEN LM. Lytic and nonlytic degradation of phospholipids in mammalian erythrocytes by pure phospholipases. Biochim Biophys Acta 1973; 307: 74-82. 8. LINDER R. Alteration of mammalian membranes by the cooperative and antagonistic actions of bacterial proteins. Biochim Biophys Acta 1984; 779: 432435.
9. PRESCOTT JF, COSHAN-GAUTHIER R, BARKSDALE L. Antibody to equi factor in the diagnosis of Corynebacterium equi pneumonia of foals. Can J Comp Med 1984; 43: 370-373. 10. SKALKA B. Dynamics of equi factor antibodies in sera of foals kept on a farm with differing histories of Rhodococcus equi pneumonia. Vet Microbiol 1987; 14: 269-276.
11. MACHANG'U RS, PRESCOTT JF. Role of antibody to extracellular proteins of Rhodococcus equi in protection against pneumonia of foals. Vet Microbiol 1991; 26: 323-333. 12. PRESCOTT JF, MACHANG'U RS, KWIECIEN J, DELANEY K. Prevention of foal mortality due to Rhodococcus equi pneumonia on an endemically affected farm. Can Vet J 1989; 30: 871-875. 13. PETERSON GL. A simplification of the protein assay method of Lowry et al. which is more applicable. Anal Biochem 1977; 83: 346-356. 14. KURIOKA S, MATSUDA M. Phospholipase C assay using p-nitrophenylphosphorylcholine together with sorbitol and its application to studying the metal and detergent requirement of the enzyme. Anal Biochem 1976; 75: 281-289. 15. LAEMMLI E. Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 1970; 227: 680-685. 16. TOWBIN H, GORDON JN. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets. Procedure and some applications. Proc Natl Acad Sci 1979; 76: 4350-4354. 17. HAQUE R, BALDWIN JN. Purification and properties of staphylococcal beta hemolysis. J Bacteriol 1964; 88: 1304-1309. 18. SCOPES RK. Protein Purification. New York: Springer-Verlag, 1982. 19. AIHARA H, WATANABE K, NAKAMURA R. Characterization of production of cholesterol oxidase in three Rhodococcus strains. J Appl Bacteriol 1986; 61: 269-274. 20. ARIMA K, NAGASAWA M, BAE M, TAMURA G. Microbial transformation of sterols. Part I. Decomposition of cholesterol by microorganisms. Agric Biol Chem 1969; 33: 1636-1643. 21. FERREIRA NP, TRACEY RP. Numerical taxonomy of cholesterol degrading soil bacteria. J Appl Bacteriol 1984; 57: 429446. 22. WATANABE K, SHIMIZU H, AIHARA H, NAKAMURA R, SUZUKI K-I, KOMAGATA K. Isolation and identification of cholesterol degrading Rhodococcus strains from food of animal origin and their cholesterol oxidase activities. J Gen Appl Microbiol 1986; 32: 137-147. 23. NELSON G. Lipid composition of erythrocytes in various mammalian species. Biochim Biophys Acta 1967; 144: 221-232. 24. LANGE Y, MATTHIES H, STECK TL. Cholesterol oxidase susceptibility of the red cell membrane. Biochim Biophys Acta 1984; 769: 551-562.
