THE JOURNAL OF BIOLOCKXL CHEMISTRY Vol. 259, No. 1, Issue of January 10, pp. 449-455,19&i 0 19&L by The American Society of Biological Chemists, Inc. Printed in U.S.A.

Distribution and Movement of Sterols with Different Side Chain Structures between the Two Leaflets of the Membrane Bilayer of Mycoplasma Cells*

(Received for publication, June 22, 1983)

Sanda Clejan and Robert BittmanS From the Department of Chemistry, Queens College of the City University of New York, Flushing, New York 11367

Mycoplasma gatiisepticum was adapted to grow with AS-sterols modified in the aliphatic side chain, and stopped-flow kinetic measurements of filipin associa- tion were made to estimate the sterol distribution be- tween the two leaflets of the membrane. Cholesterol derivatives with unsaturated side chains (desmosterol, cis- and trans-22-dehydrocholesterol, and cholesta- 5,22E,24-trien-3&ol) or an alkyl substituent (&sito- sterol) were predominantly (86-94%) localized in the outer leaflet of the bilayer. However, cholesterol, 20- isocholesterol, and sterols with side chains of varying lengths (in the 20(R)-n-alkylpregn-5-en-3B-ol series where the alkyl group ranged from ethyl to undecyl) were distributed nearly symmetrically between the two halves of the bilayer. Kinetic measurements of #I- [‘4C]sitosterol and [14C]desmosterol exchange between M. gallisepticum cells and an excess of sonicated sterol/ phosphatidylcholine vesicles confirmed the filipin- binding studies. More than 90% of these radiolabeled sterols underwent exchange at 37 “C with unlabeled sterols in vesicles over a period of 12-14 h in the presence of 2% (w/v) albumin. ,9-[‘4C]Sitosterol ex- change was characterized by biphasic exchange kinet- ics, indicative of two pools of sitosterol molecules in the cell membrane. Only a single kinetic pool was detected for [‘4C]desmosterol exchange. Stopped flow measurements of filipin binding to #I-sitosterol and stigmasterol also revealed an asymmetrical localiza- tion of these sterols in membranes of growing Myco- plasma. capricolum cells. When an early exponential culture of @-sitosterol- or stigmasterol-adapted M. ca- pricolum was transferred to a sterol-rich medium at 37 “C, approximately three-quarters of the &sitosterol or stigmasterol was localized in the outer leaflet after growth was continued for 6 h; in contrast, cholesterol was distributed symmetrically after about 1 h. The asymmetric localization of sterols with alkylated or unsaturated side chains suggests that growth-support- ing sterols need not be translocated extensively into the inner leaflet of the bilayers of M. gallisepticum and M. capricolum.

Mycoplasmas are the only organisms containing a single membrane and no cell wall that are capable of autonomous growth (1). As a result of the incorporation of high levels of sterols (without structural modification) into the cell mem-

* This research was supported in part by Grant HL 16660 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom requests for reprints should be addressed.

brane of the sterol-requiring mycoplasmas, the physical state of membrane lipids and some of the biochemical and physio- logical properties of the cell are altered (2). These features make mycoplasmas an attractive system with which to study the role of sterols in membranes. Other advantages of these prokaryotes are their lack of a cell wall, allowing direct contact between the membrane and exogenous lipids, and their ina- bility to synthesize or esterify sterols (which is a characteristic of eukaryotic systems). The inability of mycoplasmas to carry out pinocytosis is another feature that enables investigators to control the transfer of exogenous lipids to the membrane.

Sterols having a planar ring system, a free 3/Shydroxyl group, and an aliphatic side chain are taken up from the growth medium and support growth, apparently by regulating membrane fluidity (3); the ability of Mycoplasma capricolum to grow with sterols having other structural properties sug- gests that sterols may fulfill other roles as well (4). In previous work we measured the distribution of sterols between the two leaflets of the plasma membranes of two sterol-requiring mycoplasmas, M. cappricolum and Mycoplasma gallisepticum (5, 6). The distribution of sterols depends on the structure of the sterol molecule. Sterols bearing alkyl substituents at C- 24 or unsaturation at C-22 are distributed between the two halves of the bilayer of M. capricolum to a different extent than cholesterol (6). Derivatives of cholesterol with an unal- tered side chain but a modified steroid nucleus have the same distribution between the two halves of the M. capricolum membrane bilayer as cholesterol (6). In M. gullisepticum, cholesterol is distributed symmetrically in the two halves of the membrane bilayer, based on kinetic measurements of filipin binding to cells and membranes (7, 8) and of exchange of radiolabeled cholesterol between cells and acceptor mem- brane systems (9, 10). This report utilizes M. gullisepticum to describe the distribution of sterols bearing modifications in the acyclic side chain. Two types of structural changes were studied: 1) those bearing steric bulk, as exemplified by p- sitosterol (24a-ethylcholesterol) and desmosterol (24-dehy- drocholesterol), and 2) those containing shorter or longer side chains than cholesterol without the increase in steric bulk caused by introduction of alkyl substituents or double bonds. In addition, stereochemical features of the side chain were studied. The kinetics of sterol-filipin association and of p- [‘*C]sitosterol and [‘4C]desmosterol exchange from intact cells to lipid vesicles were analyzed to define some of the parameters that influence sterol movement and distribution in the M. gullisepticum membrane. We also studied the rates of incorporation and transbilayer movement of P-sitosterol, stigmasterol, and cholesterol in the membrane of M. cupri- colum ceils. The finding that the movement of the sterol molecule into the inner leaflet of the membrane is impeded by steric bulk and not by changes in the length of the side

449

by guest on March 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

450 Mycoplasma Sterol Distribution and Movement

chain raises questions about the roles of lipid-lipid and lipid- protein interactions in sterol translocation across myco- plasma membranes.

EXPERIMENTAL PROCEDURES

Materials Egg phosphatidylcholine, cholesterol, desmosterol, bovine serum

albumin (Fraction V, fatty acid poor), oleic and palmitic acids, and deoxyribonuclease were obtained from Sigma. &Sitosterol (referred to as sitosterol below) and stigmasterol were from Steraloids (Wilton, NH). Cholesterol, sitosterol, and stigmasterol were recrystallized several times from ethanol. Desmosterol was purified by preparative thin layer chromatography using Silica Gel H plates (Analtech, Newark, DE) developed with diethyl ether/petroleum ether (31, v/ v). The sterols and PC' migrated as single spots on Silica Gel G thin layer chromatography plates as previously described (6) when visu- alized by sulfuric acid spray and charring. The unsaturated side chain derivatives of cholesterol, cis- and trans-22-dehydrocholesterol and cholesta-5,22E,24-trien-3~-01, were generously supplied by Dr. Henry Kircher, University of Arizona, and characterized as described else- where (11-13). 20-Isocholesterol, which was also a gift from Dr. Kircher, was free of traces of the 20(R) stereoisomer, as judged by gas-liquid chromatography on a 5% OV-101 column and by argenta- tion thin layer chromatography of the steryl acetate precursor. The series of cholesterol analogs in which the length of the side chain varies and the terminal methyl branch is absent (2O(R)-n-alkylpregn- 5-en-3P-01s) was generously supplied by Dr. Toshiyuki Akiyama, University of Tokyo. These analogs, which are referred to by the total number of carbon atoms they contain (14, 15), were synthesized, characterized, and shown to be pure by Morisaki et al. (16). [6-methyl- 3H]Thymidine (30.6 Ci/mol) was obtained from New England Nu- clear. [4-"C]Sitosterol (58 Ci/mol) and [26(27)-'4C]desmosterol (53 Ci/mol) were purchased from Amersham (Arlington Heights, IL). Labeled sterols were analyzed for radiochemical purity by counting zones scraped from thin layer plates (silica gel, Analtech); the radio- chemical purity of all of the sterols was 398%.

Methods Growth of Mycoplasma Cells and Adaptation to Various Sterols-

M. gallisepticum strain A5969 and M. capricolum (California kid) were grown in a modified Edward medium (17) containing 10 Fg of the desired sterol/ml, and palmitic and oleic acids (10 pg of each/ml), and albumin (1% for growth of M. gallisepticum and 2% for growth of M. capricolum). The lipids were added as ethanolic solutions as described previously (6). The ethanol concentration did not exceed 0.5% (vfv). In the exchange experiments with ["C]sitosterol and ["Cc] desmosterol, 0.005 pCi of ['4C]sitosterol and 0.01 pCi of ['4C]desmos- terol were added per ml of medium. The leakiness and recovery of M. galiisepticum cells after separation from vesicles were monitored by supplementation of the medium with 0.25 gCi/ml of [6-methyL3H] thymidine. The organisms were adapted to grow with each sterol by making at least three passages with the desired sterol (10 pg/ml) prior to inoculation of a large volume of growth medium. The cells were adapted to grow in a sterol-poor medium by making serial passages into media containing decreasing sterol concentrations (5). The ex- change studies with ["C]sitosterol and ['4C]desmosterol were con- ducted using cells from logarithmic cultures (AG40 of about 0.20 for growth on sitosterol and about 0.15 for growth on desmosterol). The organisms were harvested and washed, and the cells were treated with deoxyribonuclease in the presence of 20 mM MgC12 (5, 18). The cells were resuspended in 0.40 M sucrose, 50 mM Tris buffer, pH 7.4, containing 20 mM MgCI, (STM buffer). Membranes were prepared by sonication of dilute cell suspensions at 0 "C and collected by centrifugation (5).

r'CJSitostero1 and r'C]Desmosterol Exchange Kinetics-The rates of [14C]sitosterol and ['4C]desmosterol exchange between M. gallisep- ticum cells and an excess of sterolfegg PC vesicles were measured at 37 "C with gentle shaking as described in the accompanying paper for the exchange of ["C]cholesterol (18). Vesicles (7.5 mM total lipid) were prepared in STM buffer from egg PC and sitosterol or desmos- terol by sonication under a stream of nitrogen as described previously for egg PC and cholesterol (10). The vesicles contained the same sterol/phospholipid molar ratio as that in the mycoplasma cell mem-

' The abbreviation used is: PC, phosphatidylcholine.

brane. The aqueous dispersions were centrifuged at 30,000 X g for 40 min at 4 "C to remove undispersed lipids, large multilamellar parti- cles, and titanium fragments shed from the tip. The exchange exper- iments were performed in STM buffer containing penicillin and deoxyribonuclease (18)- Bovine serum albumin (2% w/v, final con- centration) was also included, unless noted otherwise. Cells were incubated with a 40-100-fold excess of acceptors with respect to lipid concentration, which minimized saturation of the vesicles and back exchange of the radiolabeled sterol. About 80-90% of the 13H]thy- midine-labeled components was retained in the cell pellet after 12- 14 h of the incubation with the lipid vesicles, indicating that the cells containing sitosterol or desmosterol remained largely intact. Aliquots were withdrawn in duplicates at various times, cells were separated from vesicles, and the fraction of "C-sterol exchanged was calculated from the ratio of "Cf3H counts as described previously (18). The half-time for equilibration of [''C]sitosterol between the two kinetic pools we observed was calculated using the relationship tl,, = 0.693/

+ k d , where kd and k b are the first order rate constants for ["C] sitosterol translocation from the inner to outer monolayers (slower exchanging pool) and from the outer to inner monolayers (faster exchanging pool), respectively. These rate constants were calculated from the slopes of the fast and slow phases, gl and g,, respectively, in semilogarithmic plots of the percentage of "C-sterol exchanged uersus time (19). The half-times for exchange of ["C]sitosterol and ["CC] desmosterol from the rapidly exchangeable pool, tS0, were calculated using the relationship tso = 0.693/g1.

Filipin Binding Kinetics-Initial rates of filipin binding to sterols in mycoplasma cells and membranes were measured at 360 nm as described previously (7, 8). The initial reaction rate was first order with respect to sterol and filipin. Second order rate constants for filipin association with sterol in intact cells and isolated membranes, designated as and kmembranes, were calculated as described else- where (7). All rate measurements were made in STM buffer at 10 "C. The air pressure used to initiate mixing of equal volumes of filipin solution with mycoplasma suspensions was reduced to 30 p.s.i. to minimize cell shearing. At least six measurements of the initial rates of filipin binding were made at each sterol concentration.

Analytical Procedures-Lipids were extracted from cells and iso- lated membranes and lipid phosphorus and sterol content were de- termined using procedures described before (5). The wavelengths used for estimating sterol concentration were as follows: 542 nm for sitosterol; 555 nm for desmosterol; and 550 nm for the other sterols. We used the standard curve obtained with cholesterol for assay of the concentrations of the cholesterol analogs bearing side chains of varying lengths (C22-C32).

RESULTS

Growth of M. gallisepticum with Sterok; Differing in Side Chain Structure-One of the attractive features about using mycoplasmas as a system with which to investigate the rela- tionship between sterol structure and transbilayer distribu- tion, in addition to their well known inability to synthesize sterols, is their ability to incorporate sterols from an exoge- nous supply without structural modification. We adapted M. gallisepticum to grow with various A5-sterols with identical steroid nuclear structures but different aliphatic side chains (see Fig. 1 for structures). At 10 pg/ml, sitosterol supported growth nearly as well as cholesterol; good growth was also obtained with cholesta-5,22E,24-trien-3/3-01, cis-22-dehydro- cholesterol, and desmosterol (Fig. 2 A ) . However, cells grew less well with trans-22-dehydrocholesterol, although it has an extended side chain resembling that of cholesterol, and with 20-isocholesterol, which differs from cholesterol only with respect to the configuration a t C-20. In addition, the cell yields were low with trans-22-dehydrocholesterol and 20- isocholesterol. Lower amounts of these sterols were incorpo- rated into the cell membrane compared with sitosterol and desmosterol (which differ from cholesterol in having a 24a- ethyl group and a AZ4-bond, respectively), cholesterol, and the sterols bearing unsaturation at C-22 and (2-24 (Table I). Sterols with short side chains also supported growth. For example, C26 was especially effective (Fig. 2B), although the cell membranes contained less sterol than cells grown on

by guest on March 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Mycoplasma Sterol Distribution and Movement 45 1

cholesterol (Table I). Substantial growth was found with C28; the cholesterol analogs C22 and C24 supported growth less well, and cell death was apparent after about 25 h. With C32, there was a marked reduction in the maximum absorbance of the culture and substantial cell death after 23 h (Fig. 2B); with C30, the cell yield was poor, although the cultures reached maximum absorbance values comparable to those grown with C22 and C24 and the sterol contents of the membranes were similar. Thus, various sterols can approach the effectiveness of cholesterol as a growth supporter of M. gallisepticum (sitosterol, desmosterol, cis-22-dehydrocholes- terol, and C26), and others promote substantial growth (cho- lesta-5,22E,24-trien-3@-01, C22, C24, C28). However, poor growth-promoting efficiencies were found with sterols having greatly extended side chains (as in C32 and trans-22-dehydro- cholesterol) and with 20-isocholesterol.

[14C]Sitosterol and ['4C]Desmosterol Exchange between M. gallisepticum Cells and Lipid Vesicles-On incubation with an excess of sonicated vesicles in STM buffer, cells remained intact (see "Methods") and the sterol/phospholipid molar ratio was maintained constant. More than 90% of the labeled sterol was removed from the cells during a 12-14-h period of

HO

I

rn 7

". 3

6 c

6

x-f 9

FIG. 1. Structures of the A5-3fl-hydroxysterols with which M. gallisepticum was grown. 1, cholesterol; 2, 20-isocholesterol; 3, 2O(R)-n-alkylpregn-5-en-3@-01: C22 (n = 01, C24 ( n = 2), C26 (n = 4), C28 ( n = 61, C30 (n = 8), C32 ( n = 10); 4, desmosterol (24- dehydrocholesterol); 5, sitosterol (24oc-ethylcholesterol); 6, stigmas- terol (24a-ethylcholesta-5,22E-dien-3~-01); 7, tram-22-dehydrocho- lesterol (cholesta-5,22E-dien-3P-o1); 8, cis-22-dehydrocholesterol (cholesta-5,22Z-dien-3@-01); and 9, cholesta-5,22E,24-trien-3P-ol.

incubation in the presence of 2% albumin (Fig. 3). Since the kinetic data for ['4C]sitosterol exchange (Fig. 3A) can be fitted by the sum of two exponentials, it is concluded that ["C] sitosterol is distributed between two pools in the cell mem- brane. Thus semilogarithmic plots of the percentage of ["C] sitosterol remaining in M. gallisepticum cells uersus time are biphasic (Fig. 3A). By extending the straight line of the slower process of the semiexponential plot to time zero, we found that the faster phase represents approximately 76% of the total [ '*C]sitosterol undergoing exchange with nonlabeled si- tosterol in vesicles. ['4C]Cholesterol also gave biphasic kinet- ics of exchange between cells and vesicles (10, 18), and it was presumed that the rapidly and slowly exchangeable pools represent sterol molecules initially in the outer monolayer and inner monolayer of the bilayer, respectively. The size of the faster exchangeable pool of ['4C]sitosterol molecules, r,,,,, and the half-time of equilibration of [14C]sitosterol between the two pools, tl/,, were calculated from the fast and slow exponential constants, g, and g,, respectively (19). Table I1 lists these parameters and also shows the corresponding val- ues obtained for ['4C]cholesterol exchange (18). The fraction of ['4C]sitosterol molecules in the rapidly exchangeable pool is considerably greater than that of [14C]cholesterol (0.76 and 0.58, respectively). A comparison of the slopes of the two phases obtained for exchange with sitosterol and cholesterol (Fig. 3A compared with Fig. 2 of Ref. 18, sterol/phospholipid molar ratio of 0.9) shows that the more accessible pool under- went exchange faster with cholesterol than sitosterol (see g, and ts0 values in Table 11). However, the slower exchange process, which is thought to represent sterol movement from the inner to outer leaflet, occurred at the same rate. Table I1 also shows that tllz is lower for the equilibration of sitosterol between the two pools than for equilibration of cholesterol. This unexpected result arises because the half-time is a func- tion of the sum of the rate constants (kab + kba), and the rate constants are calculated from g,, g,, and the pool size (19). (The pool size is determined from the extension of the straight line of the slower process to the y axis of Fig. 3A.) Fig. 3B shows that only a single kinetic pool was detected for ["C] desmosterol exchange. This is in contrast to the semilogarith- mic plots of ['4C]sitosterol and ['4C]cholesterol exchange be- tween cells and vesicles, which are characterized by a rapid and a slow rate. Bovine serum albumin (2%, w/v) stimulated the rate of ['4C]desmosterol exchange by a factor of 1.9, as found in our previous study of ['4C]cholesterol exchange be- tween M. gallisepticum cells and lipid vesicles (10).

Rates of Filipin Binding to Sterols in Cells and Membranes- In several recent studies we have used stopped flow measure-

FIG. 2. Growth of M. gallisepti- cum A5969 in Edward medium sup- plemented with bovine serum albu- min (1% w/v), palmitic and oleic acids (each 10 rg/ml), and sterol (10 pg/ml). A , the sterols used were: a, cho- lesterol; w, sitosterol; 0, cis-22-dehydro- cholesterol; V, cholesta-5,22E,24-trien- 3@-01; V, desmosterol; and 0, trans-22- dehydrocholesterol or 20-isocholesterol. Inset, plot of growth after 20 h uersu concentration of cholesterol (a) and si- tosterol (=). B, the sterols were: ., (226, V, C28; 0, C30; 0, C22 or C24; Y, C32.

0 . 2 0

0.10

0.0 I

j 0.oa

0.04

0 0 2

I C

I -

i -

L

I

0.20 -

0 10-

0.01-

3 0 0 0 - t 0 04-

0 . 0 2 -

I

by guest on March 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

452 Mycoplasma Sterol Distribution and Movement

ments of filipin-sterol association to probe the transbilayer distribution of cholesterol (8, 20) and other free sterols (6) in membranes that can be prepared in both sealed and unsealed states. An example of the initial, linear increase in transmit- tance arising from filipin binding is shown for sitosterol- containing cells (Fig. 4A). The initial rate of absorbance change, dA/dt , with cells and membranes is first order in sterol (Fig. 4B) and filipin (7) concentrations. The extent of cell membrane disruption induced by filipin binding is mini- mized by the use of initial velocity measurements, high sterol/ filipin molar ratios, and low temperature. Fig. 4B shows that the rates of binding of filipin to isolated mycoplasma mem-

TABLE I Growth and sterol content of M. gallisepticum cells obtained from

media supplmented with different A5-sterok Cultures were grown at 37 "C in a modified Edward medium

supplemented with sterol, palmitic acid and oleic acid (10 wg/ml of each), and 1% fatty acid-poor bovine serum albumin.

Mem- Sterol used for brane Sterol content

growth pro- of membrane' tein"'

mg pgUs/ mg Sitosterol 0.135 6.2 95 Desmosterol 0.130 5.8 85 Cholesta-5,22E,24- 0.115 6.0 107

cis-22-Dehydro- 0.138 6.0 107

trans-22-Dehydro- 0.065 3.9 70

Cholesterol 0.155 6.4 124 20-Isocholesterol 0.065 4.0 45 c22 0.103 5.0 73 C24 0.105 5.4 C26 0.160 6.4 67 C28 0.115 5.8 79 C30 0.100 3.2 68 C32 0.047 NDd NDd

trien-36-01

cholesterol

cholesterol

a The absorbance of the culture at 640 nm was measured after 20 h of growth. The values are representative examples from one culture with each sterol.

* Cells were grown in 500 ml of medium. Values are the average of two different cell preparations. Not determined because of poor growth and extensive cell lysis.

I I 1 I I I 2 4 6 8 IO 12 14

TIME (hours)

branes (which are open; see Ref. 10) are faster than those with the corresponding intact cells. Table I11 presents the second order rate constants for filipin binding to various A5-sterols in intact cells (kcell,) and isolated membranes (kmembmnes). High ratios of 12ce&membranes (0.86-0.94) indicate that sitosterol and the cholesterol derivatives with unsatu- rated side chains, desmosterol, cholesta-5,22E,24-trien-3P-o1, and cis- and truns-22-dehydrocholestero1, accumulate in the outer half of the bilayer. However, cholesterol, 20-isocholes- terol, and the cholesterol analogs with short and long side chains are nearly symmetrically distributed between the two leaflets, with 52-60% of the sterol in the outer half of the bilayer.

In a previous report we used second order rate constants of filipin-cholesterol association in cells and membranes to show that cholesterol moves rapidly from the outer to inner half of the lipid bilayer of growing M. eapricotum (5) . The results of stopped flow kinetic measurements of the filipin-sitosterol and filipin-stigmasterol association in M. capricolum cells and membranes are given in Table IV. To test the relative abilities of cholesterol, sitosterol, and stigmasterol to undergo rapid translocation across the M. capricolum cell membrane, we adapted this organism to grow with 1.25 pg of each sterol/ml. When a culture of the adapted strain was transferred to a sterol-rich medium (10 pg/ml), the growth rate was enhanced; the sterol content of the membranes was increased by more than 3-fold for sitosterol and stigmasterol within 6 h of growth, and more than 4-fold for cholesterol within 4 h (Table IV). When samples were withdrawn at various time intervals after the cultures had been supplemented with sterol, stopped flow studies with the intact cells and isolated membranes revealed that sitosterol and stigmasterol were predominantly localized in the outer leaflet of the membrane. For example, after 4 h of growth stimulation and sterol incorporation, approximately three-fourths of the sitosterol or stigmasterol was estimated to be in the outer half of the bilayer. Choles- terol, however, was distributed about equally between the two halves of the bilayer after only 1 h of incubation. Furthermore, the amount of sitosterol or stigmasterol incorporated was lower than that of cholesterol. These experiments show that the rates of sitosterol and stigmasterol transbilayer movement in growing M. capricolurn cells are inhibited relative to the

2 4 6 8 IO 12

TIME (hours)

FIG. 3. First order kinetic plots of ['4C]sitosterol or ['4C]desmosterol exchange fromM. gallisepticum cells to sitosterol/PC or desmosterol/PC vesicles at 37 "C. The molar ratio of sterol/phospholipid was 0.90. A , sitosterol exchange in STM buffer in the presence of albumin (2%, w/v). The values of percentage of ["C] sitosterol remaining in the cells are plotted k S.E. uersus time of incubation with vesicles; cells were separated at intervals as described under "Methods," The arrow represents the size of the rapidly exchanging kinetic pool (see text). B, desmosterol exchange in STM buffer in the presence (0) and absence (0) of albumin (2%, w/v).

by guest on March 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Mycoplasma Sterol Distribution and Movement 453

TABLE I1 Kinetic parameters for exchange of [14C]sterols between M.

gallisepticum cells and sterolleg PC vesicles The kinetic parameters were determined as described under "Meth-

ods" using the derivation of Bloj and Zilversmit (19). For the exchange of ["CJsitosterol (Fig. 3A) or of ['4C]cholesterol (18) from M. galli- septicum cells, g, and g2 are the rapid and slow exponential rate constants. The size of the outer pool, router, is obtained from ro = kJ ( k d + k b ) . tm is the time required for the "C-sterol in the more accessible pool to be reduced by 50%. For ["C]desmosterol exchange only one kinetic pool was found (Fig. 38). Values are the average of two different cell preparations for sitosterol and desmosterol and four different preparations for cholesterol. The sterol/phospholipid molar ratio was 0.9. The exchange experiments were conducted a t 37 "C in STM buffer containing 2% albumin. For the exchange of ["Cldes- mosterol in the absence of albumin (Fig. 38) , the values of g, and tW were -0.105 h" and 6.5 h, respectively. The slopes of the rapid and slow phases ( R I and g2) in the semilogarithmic plots (Fig. 3) were calculated by a least squares regression analysis of the data points. The correlation coefficients for g1 and g2 of the 14C-sterols presented in this table ranged from 0.93 to 0.96.

"C-Sterol -&!I -& rmbr t v z tm

h-l h" h h -

Cholesterol 0.277 0.167 0.58 6.2 2.5 Sitosterol 0.204 0.165 0.76 4.9 3.4 Desmosterol 0.196 1.0 3.5

TABLE I11 Second order rate constants for association of filipin with sterols in

M. gallisepticum cells and membranes The second order rate constants were calculated from stopped flow

traces of the initial rates of filipin-sterol association at 10 "C as described previously (7). Initial rates were measured using a t least four sterol concentrations in cells and membranes from each culture. The filipin concentration was maintained constant (7 to 9 p~ final concentration). The second order rate constants presented are the averages from two cell cultures of M. gallisepticum adapted to grow with each sterol with the exception of sitosterol and cholesterol, with which three cultures were grown, and desmosterol, with which four cultures were grown. The error limits are standard errors of the mean.

Sterol in cell membrane %lh k m s m b n n a kcslhlkmsmbnnr

Sitosterol Desmosterol Cholesta-5-22E,24-

trien-38-01 cis-22-Dehydro-

cholesterol trans-22-Dehydro-

cholesterol Cholesterol 20-Isocholesterol c22 C24 C26 C28 C30

x 1 0 4 M-* s-I

3.6 f 0.3 4.2 f 0.4 8.1 f 0.6 8.6 f 0.8 6.6 f 0.4 7.0 f 0.3

6.5 f 0.5 7.1 f 0.5

6.7 f 0.3 7.2 f 0.3

6.1 f 0.2 11.7 f 0.3 14.4 f 0.7 24.4 f 0.8 7.6 f 0.8 12.9 f 1.0 6.7 f 0.9 12.2 f 1.0 7.2 f 0.8 13.1 f 0.9 7.0 f 0.4 12.3 f 0.5 8.8 f 1.0 14.7 f 1.1

~

0.86 & 0.06 0.94 f 0.06 0.94 f 0.05

0.91 f 0.07

0.93 f 0.04

0.52 f 0.05 0.59 f 0.04 0.59 f 0.09 0.55 f 0.10 0.57 k 0.09 0.60 f 0.05 0.59 f 0.09

0.8 n 0 4) 03 > 0.6 ? Y

c

0

0 0.4 a 0

0.2

4 0 12 16

SITOSTEROL (yM)

FIG. 4. Stopped flow kinetics of association of filipin with C22 and sitosterol in M. gallisepticum cells and membranes at 10 "C. A, photograph of multiple traces of the initial transmittance changes a t 360 nm as a function of time (50 ms/large division) obtained on mixing filipin solutions with intact cells containing C22. The vertical axis is 50 mV/division; initial signal, 1.6 V. The initial 100-150-ms disturbance period was ignored, and initial rates were calculated from the linear transmittance increases (see Ref. 7). The final C22 concentration in the cell suspension, after mixing with filipin, was 11.2 pM. The final filipin concentration was 9 pM. 8, plot of the initial rate of absorbance change/s, dA/dt (which was calculated from the increase in the transmittance of filipin on binding to sitosterol) uersw sitosterol concentration in intact cells (0) and open isolated membranes (0). A t least six measurements of the initial rate were made a t each concentration. The final filipin concentration was 7.5 pM.

rate of cholesterol translocation into the internal membrane surface.

DISCUSSION

In this paper we have assessed how sterol transbilayer distribution in M. gallisepticurn is influenced by shortening and extending the length of the aliphatic side chain and by enlarging it in bulk by introducing double bonds or alkyl substituents. In a previous study we reported that sitosterol, stigmasterol, and ergosterol are localized predominantly in the outer leaflet of the M. capricolurn cell membrane (6). I t has been pointed out by Razin that conclusions about sterol

structural specificity drawn from studies with M. capricolurn may not be valid for other sterol-requiring mycoplasmas (21), since a very broad range of sterols can support growth of this organism (e.g. Refs. 6, 22-24). We have now analyzed sterol transbilayer distribution in M. gallisepticurn membranes. Sterols bearing alkyl substituents or unsaturation in the side chain are localized predominantly in the outer leaflet (Table H I ) , indicating that M. gallisepticurn and M. capricolurn share the same characteristic of accumulating sterols with bulky or rigid side chains in the outer half of the bilayer. Alkylation a t C-24 (as in sitosterol, stigmasterol, and ergosterol) increases the steric bulk of the side chain, and (because of branching) decreases the conformational freedom. Unsaturation (as in

by guest on March 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

454 Mycoplasma Sterol Distribution and Movement

TABLE IV Growth of M. capricolurn cells, sterol content, and sterol transbiluyer distribution in the membrane of adapted cells

transferred to sterol-rich medium M. cupricolurn was adapted to grow on low sterol (1.25 pglml), and cultures were grown to AG4,, = 0.10. At time

zero, 10 Mg of cholesterol (5), sitosterol, or stigmasterol were added per ml of medium supplemented with 2% fatty acid-poor bovine serum albumin and 10 Gg/ml of oleic and palmitic acids. Growth was continued at 37 "C for periods up to 6 h, and cells were harvested at the times indicated. Second order rate constants are the averages calculated from initial rates of stopped flow traces at 10 "C using at least two cultures with sitosterol and stigmasterol and at least nine different cultures with cholesterol.

Time after transfer to 10 pg

of sterol/ml Cholesterol" Sitosterol Stigmasterol Cholesterol" Sitosterol Stigmasterol Cholesteroln Sitosterol Stigmasterol

A610 Sterol content Kce!l.lkm.mb,a

h pgfmg membrane protein 0 0.08 0.10 0.10 33 25 20 0.73 0.83 0.89 1 0.16 0.11 0.11 44 33 32 0.53 0.80 2 0.24 0.13 0.13 92 42 41 0.45 0.74 4

0.85 0.26 0.20 0.20 127 62 59 0.45 0.75 0.79

6 0.23 0.22 77 70 0.70 0.78 Data with cholesterol represent results after transfer to 10% fetal calf serum (5).

ergosterol, stigmasterol, desmosterol, cis- and trans- 22-de- hydrocholesterol, and cholesta-5,22E,24-trien-3P-o1) also de- creases the flexibility and increases the bulk of the side chain. The differences in the distribution of cholesterol and deriva- tives containing alkyl groups or double bonds suggest that the latter experience steric interference during translocation across M . gallisepticum and M. capricolum membranes. The kinetics of sitosterol and stigmasterol translocation from the outer leaflet of adapted M. capricolum cells (which is in contact with the exogenous supply of sterol) into the inner leaflet confirm this conclusion (Table IV); about one-half of the cholesterol moved into the inner half of the bilayer rapidly in these growing cells, whereas about three-fourths of the newly acquired sitosterol and stigmasterol was present in the outer half. In M . gallisepticum, the high ratios of kcells/kmembranes

in the filipin-binding experiments (Table 11) indicate that 86 k 6% of the sitosterol and 94 + 6% of the desmosterol are localized in the outer leaflet. The [14C]sitosterol and ["C] desmosterol exchange data are in good agreement with these results, since 76% and 100% of these sterols, respectively, were found in the more accessible pool (Fig. 3, Table 11). In contrast to the effects of side chain bulk and rigidity on sterol localization in the membrane, side chain length does not appear to impose an important constraint on sterol distribu- tion between the inner and outer halves of the M . gallisepticum membrane. Cholesterol and its analogs having shorter (C22, C24, C26) and longer (C28 and C30) side chains than isooctyl were found to have the same localization (Table 111). Thus the absence of branching at C-25 in the 20(R)-n-alkylpregn- 5-en-36-01s does not influence sterol distribution. Further- more, 20-isocholesterol and cholesterol have an identical dis- tribution between the two leaflets of the membrane, suggest- ing that the orientation of the side chain at C22 with respect to the C/D ring system does not influence sterol localization in M. gallisepticum. It therefore appears that the translocation process is particularly sensitive to inhibition by the insertion of double bonds at C-22 or C-24 or alkyl groups at C-24 of the sterol side chain.

All of the sterols we used are taken up into membranes spontaneously because of their lipophilicity, and become in- tercalated between fatty acyl chains of phospholipid mole- cules. Sterols with alkyl substituents or double bonds in the side chain may fit into spaces between adjacent phospholipids less well than the unalkylated, saturated analogs (Refs. 4 and 25 and references cited therein). The looser packing of these sterols with phospholipids in the membrane is expected to decrease the ability of the rigid sterol ring to restrict the

trans-gauche isomerizations of the fatty acyl chains. Since the sterol side chain extends deeply into the interior of the bilayer (26, 27), van der Waals interactions in this region would be weakened when sterols with bulky side chains are present. In fact, sterols with very long aliphatic side chains have been postulated to penetrate into the adjacent leaflet of the bilayer (28). This insertion would disrupt the membrane bilayer and increase the fluidity of the membrane in the liquid-crystalline state; it may explain why C30 and, in particular, C32 failed to function as effective growth supporters of M. gallisepticum. It is not known why cis-22-dehydrocholesterol was more ef- fective than the trans isomer with respect to growth support and cell yield. The kinked conformation of the side chain has been postulated to reduce the ability of this sterol to inter- digitate between phospholipid fatty acyl chains, producing a lethal effect on mouse fibroblast cells (29). It should be noted, however, that the cis isomer was not toxic to many species of Drosophila, whereas trans-22-dehydrocholesterol was toxic (13).

Several other factors, besides looser packing of the bulk lipids, may be responsible for the differences in growth and cell yield we found with the various sterols (Table I). The abilities of sterols with alkylated, unsaturated, or very long side chains to act as spacer and filler molecules may be diminished relative to cholesterol. In its role as a spacer molecule, cholesterol is considered to separate phospholipid head groups and disrupt head group interactions (30-33). Since acidic phospholipids (mainly phosphatidylglycerol and diphosphatidylglycerol) are synthesized by mycoplasmas, the sterol-induced decrease in the density of surface charges would be expected to modify intermolecular interactions be- tween polar head groups, leading to changes in conformations of membrane proteins. In its role as a filler molecule, choles- terol at low concentrations (<5 mol %) can occupy potential voids in the hydrocarbon region of the membrane, making the straightening of the fatty acyl chains energetically favorable (34). Small amounts of cholesterol may also exert a regulatory role on phospholipid biosynthesis by interacting with specific membrane-bound enzymes (35). It remains to be determined if low concentrations of other sterols can also interact speeif- ically with protein(s) in the membrane to enhance enzymatic activity, resulting in modification of phospholipid synthesis.

The physiological role of extensive sterol localization in each leaflet of the membrane bilayer remains to be estab- lished. The results we obtained in M. gallisepticum (Tables I1 and 111) and in M. capricolum (6) indicate that growth- supporting sterols need not be translocated extensively into

by guest on March 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Mycoplasma Sterol Distribution and Movement 455

the inner half of the bilayer. For example, sitosterol, cholesta- 5,22E,24-trien-3&01, and cis-22-dehydrocholesterol supported growth of M. gallisepticum and were incorporated into the membrane in high amounts, but with different distributions between the two leaflets compared with cholesterol. On the other hand, 20-isocholesterol and C30 gave the same trans- bilayer distribution as cholesterol, but were poor growth sup- porters and were taken up at lower levels. Similarly, in M. cupricolum (6), sitosterol, ergosterol, and stigmasterol sup- ported growth nearly as well as cholesterol but were localized differently than cholesterol in the two leaflets of the bilayer; cholesterol analogs with modifications in the steroid nucleus were distributed identically as cholesterol but were poorer growth supporters. No information is available concerning the possibility that other membrane components may be redistributed differently between the two halves of the bilayer of cholesterol-containing membranes compared with, for ex- ample, sitosterol-containing membranes. It is also not known how much sterol is needed in the inner leaflet of mycoplasmas for optimal membrane function or whether some specific membrane functions may be altered when the sterol distri- bution is changed.

Acknowledgments-We are grateful to Drs. Toshiyuki Akiyama and Henry Kircher for kindly donating the cholesterol analogs. We thank Sylvia Schaffel for typing the manuscript and Bruce Robinson for preparing the art work.

REFERENCES 1. Razin, S. (1973) Adu. Microb. Physiol. 10, 1-80 2. Rottem, S. (1980) Biochim. Biophys. Acta 604,65-90 3. Razin, S. (1978) Microbiol. Rev. 42 , 414-470 4. Bloch, K. E. (1983) Crit. Rev. Biochem. 14, 47-92 5. Clejan, S., Bittman, R., and Rottem, S. (1978) Biochemistry 17 ,

6. Clejan, S., Bittman, R., and Rottem, S. (1981) Biochemistry 2 0 ,

7. Bittman, R., and Rottem, S. (1976) Biochem. Biophys. Res. Com-

8. Bittman, R., Blau, L., Clejan, S., and Rottem, S. (1981) Biochem-

4579-4583

2200-2204

mun. 71,318-324

istry 20, 2425-2432 9. Rottem, S., Slutzky, G. M., and Bittman, R. (1978) Biochemistry

17,2723-2726

10. Rottem, S., Shinar, D., and Bittman, R. (1981) Biochim. Biophys.

11. Kircher, H. W., and Rosenstein, F. U. (1982) J. Org. Chem. 47,

12. Rosenstein, F. U., Caruso, R., and Kircher, H. W. (1977) Lipids

13. Kircher, H. W., Phariss, R. L., Rosenstein, F. U., Baldwin, D.,

14. Suckling, K. E., and Boyd, G. S. (1976) Biochim. Biophys. Acta

15. Nakamura, T., Nishikawa, M., Inoue, K., Nojima, S., Akiyama, T., and Sankawa, U. (1980) Chem. Phys. Lipids 26, 101-110

16. Morisaki, M., Shibata, M., Duque, C., Imamura, N., and Ikekawa, N. (1980) Chem. Pharm. Bull. (Tokyo) 28, 606-611

17. Razin, S., and Rottem, S. (1976) in Biomedical Analysis of Mem- branes (Maddy, A. H., ed) pp. 3-26, Chapman and Hall, London

18. Clejan, S., and Bittman, R. (1984) J. Biol. Chem. 259,441-448 19. Bloj, B., and Zilversmit, D. B. (1976) Biochemistry 15, 1277-

20. Blau, L., and Bittman, R. (1978) J. Biol. Chem. 253,8366-8368 21. Razin, S. (1982) Curr. Top. Membr. Transp. 17 , 183-205 22. Odriozola, J. M., Waitzkin, E., Smith, T. L., and Bloch, K. (1978)

23. Lala, A. K., Buttke, T. M., and Bloch, K. (1979) J. Biol. Chem.

24. Efrati, H., Shinitzky, M., and Razin, S. (1980) FEBS Lett. 122,

25. Bittman, R., Leventhal, A. M., Karp, S., Blau, L., Tremblay, P.- A., and Kates, M. (1981) Chem. Phys. Lipids 28, 323-335

26. Stockton, G. W., and Smith, I. C. P. (1976) Chem. Phys. Lipids

27. Franks, N. P. (1976) J. Mol. Biol. 100,345-358 28. Suckling, K. E., Blair, H. A. F., Boyd, G. S., Craig, I. F., and

Malcolm, B. R. (1979) Biophys. Biochim. Acta 551,lO-21 29. Kircher, H. W., Rosenstein, F. U., and Kandutsch, A. A. (1981)

Lipids 16 , 943-945 30. Yeagle, P. L., Martin, R. B., Lala, A. K., Lin, H. K., and Bloch,

K. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 4924-4926 31. Yeagle, P. L. (1978) Accts. Chem. Res. 11, 321-327 32. Bloch, K. E. (1979) CRC Crit. Rev. Biochem. 7, 1-5 33. Wieslander, A., Christiansson, A., Rilfors, L., and Lindblom, G.

34. Presti, F. T., Pace, R. J., and Chan, S. I. (1982) Biochemistry 2 1 ,

35. Dahl, J. S., Dahl, C. E., and Bloch, K. (1981) J. Biol. Chem. 256,

Acta 649,572-580

1722-1724

12,297-300

and Fogleman, J. C. (1982) Lipids 17,209-214

436,295-300

1283

Proc. Natl. Acad. Sci. U. S. A. 75 , 4107-4109

254,10582-10585

59-63

17,251-263

(1980) Biochemistry 19, 3650-3655

3831-3835

87-91

by guest on March 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from

S Clejan and R Bittmanthe two leaflets of the membrane bilayer of mycoplasma cells.

Distribution and movement of sterols with different side chain structures between

1984, 259:449-455.J. Biol. Chem. 

  http://www.jbc.org/content/259/1/449Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/259/1/449.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 30, 2020

http://ww

w.jbc.org/

Dow

nloaded from