inhibition of nk cell-mediated cytotoxicity by oxysterols

CELLULAR IMMUNOLOGY 139, 541-549 (1992)

Inhibition of NK Cell-Mediated Cytotoxicity by Oxysterols’

OMERKUCUK,* JEANNETTESTONER-PICKING,* STANLEY~ACHNIN,~ LEO I. GORDON,# R. MICHAEL WILLIAMS,* LEONARD J. LB,*

ANDMAXWELLP.WESTERMAN*

*Section of Hematologv/Oncology, Chicago Medical-School and the Veterans Afairs Medical Center, North Chicago, Illinois 60064; iSection of Hematology/Oncology, University of Chicago,

Chicago, Illinois 60637; and $Yection of Hematology/Oncology, Northwestern University Medical School, Chicago, Illinois 6061 I

Received August 26, 1991; accepted October 16, 1991

Some of the oxidation products of cholesterol (oxysterols) have profound effects on plasma membrane structure and function. The present studies were undertaken to determine the effects of oxysterols on NK cell-mediated cytotoxicity. When mouse spleen cells were preincubated with certain oxysterols, NK cell cytotoxicity was inhibited without loss of effector cell viability. The strongest inhibition was observed with oxysterols that-are oxidized at the C-5, C-6, or C-7 positions of the sterol nucleus. Among these, 7&hydroxycholesterol caused more inhibition than 7cu-hydroxycholesterol suggesting that the spatial orientation of the hydroxyl group in the &position results in a greater perturbation in plasma membrane structure than that oriented in the (Y- position. In contrast, oxysterols that are oxidized at the C-20 and C-25 positions that are located on the C- 17 acyl chain had little or no inhibitory effect, suggesting that oxidation in the cholesterol nucleus which is situated closer to the phospholipid headgroups at the lipid bilayer-aqueous interlace results in a more profound effect on the plasma membrane physical structure. These results suggest that the lytic function of NK cell is sensitive to alterations in the physical state of its plasma membrane induced by oxysterols. 0 1992 Academic Press, IX.

INTRODUCTION

Oxysterols (OS) are derived from cholesterol oxidation under a variety of conditions associated with oxidant stress (l-4). The presence of OS in the plasma membrane leads to a number of changes in membrane structure, function, and morphology (5 7). Since NK cell cytotoxicity appears to be a membrane-mediated event (8, 9) we have studied the effects of various oxysterols on mouse NK cell function, We have found that certain OS inhibit mouse NK cell cytotoxicity in a dose-dependent fashion. This effect is not due to loss of NK cell viability and is abolished by serum lipoproteins which bind the OS and prevent their insertion into the NK cell plasma membrane. The experimental evidence suggests a mechanism that is independent of the capacity

’ Supported by grants from Veterans Affairs Regional Advisory Group, BRSG from UHS/The Chicago Medical School (SO7 RR05266-27 awarded by the Biomedical Research Support Grant Program, Division of Research Resources, NIH), Illinois Cancer Council (under Grant Nos. 2-S07-RR05893, Biomedical Research Support Grant and 5-P3O-CA27 142, Cancer Center Support Grant awarded by the National Cancer Institute), and Melvin Leichtling Memorial Fund.

541

0008-8749/92 $3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

542 SHORT COMMUNICATIONS

of OS to inhibit cellular sterol synthesis. The position of oxidation in the sterol molecule appears to determine whether an inhibitory effect of OS is observed. These results and recent observations on the effects of OS on membrane structure (7) support the hy- pothesis that membrane perturbations caused by OS insertion may result in altered cellular function including defective NK cell-mediated cytotoxicity.

MATERIALS AND METHODS

Animals. Normal 4- to 6-week-old male CBA/J mice were obtained from Harlon (Frederick, MD). The animals were aged-matched (6-15 weeks old) at the onset of each experiment.

Oxysterols. 6-Ketocholestanol, 7-ketocholesterol, and 20a-hydroxycholesterol were obtained from Sigma (St. Louis, MO), and 5a-hydroxy-6-ketocholestanol, 7a-hydroxycholesterol, 7@-hydroxycholesterol, and 25-hydroxycholesterol were obtained from Steraloids (Wilton, NH). The OS were added to lymphocyte suspensions as ethanolic solutions at 100 times the desired final concentrations.

Target cells. YAC-I tumor cells were maintained in tissue culture flasks (Falcon) containing RPM1 1640 medium ,with Hepes (MA Bioproducts, Walkersville, MD), 5% heat-inactivated fetal calf serum (GIBCO, Grand Island, NY), 0.1 mg/ml genta- mycin, and 5 mM L-glutamine in a 5% COZ humidified incubator. Cultures were passaged twice weekly and periodically tested for mycoplasma contamination.

Preparation of lymphocytes. Spleens were removed immediately after cervical dis- location and single cell suspension of spleen cells were made in ice-cold RPM1 1640 with gentle teasing of cells out of the spleens by using a scalpel blade. After enrichment of the lymphocyte fraction using an Isolymph (Gallard-Schlesinger Chemical, Carle Place, NY) gradient, the cells were washed, and 2 X lo6 cells in 4 ml of RPM1 1640 were placed in tubes for incubation.

Incubation with oxysterols. One milliliter of lipoprotein depleted serum (LPDS) or normal human AB serum (ABS) was added to 4 ml of lymphocyte suspension, resulting in 20% lipoprotein-depleted medium (LPDM) or 20% lipoprotein replete medium (LPRM). LPDS was prepared as previously described (5) by ultracentrifugation flotation of lipoproteins of density < 1.25 g/ml. The residual bottom fraction is dialyzed ex- tensively against PBS, heated to 56°C for 45 min, and sterilized by membrane filtration prior to use. Ethanolic solution (50 pl) of OS was added resulting in final OS concentration of lop5 A4 (initially different concentrations of OS ranging from 2.5 X 10e6 to 2.5 X 10e5 M were used to determine dose-response). Since OS were dissolved in ethanol, all incubations contained a final concentration of 1% (v:v) ethanol. Incubations were carried out at 37°C for 45 min (initially different incubation periods were used to determine time dependence of OS inhibition) with gentle rocking in a water bath. Viability of lymphocytes was assessed by trypan blue dye exclusion both before and after incubation.

“Cr release assay. YAC-1 target cells ( 106) were labeled with 100 &i Na25’Cr04 for 90 min at 37°C and were washed three times. The desired concentration (100 ~1) of lymphocytes and 100 ~1 of 51Cr-labeled target cells (5,000) were added into triplicate wells of a 96-well, U-bottom microtiter plates (Corning) to obtain the desired effector: target (E:T) ratios. The plates were centrifuged at 300g for 2 min and then incubated for 4 hr at 37°C 5% CO2 in a humidified incubator. The plates were then centrifuged at 500g for 10 mitt, and 100 ~1 of supematant transferred to tubes for counting of

SHORT COMMUNICATIONS 543

radioactivity. Percentage of specific “Cr release was calculated by using the formula: (cpm experimental release-cpm spontaneous release)/(cpm maximal release-cpm spontaneous release) X 100, in which maximal release refers to the cpm obtained after adding Triton X-100 (0.5%) to the target cells just before centrifugation of the plates. Spontaneous release refers to the cpm released by target cells in the absence of effecters. All experimental and control values are the means of three replicates.

RESULTS

Cumulative results of all NK cell activity (NKCA) experiments are shown as a scattergram on Fig. 1. In each experiment, spleen from one mouse was used and two to four oxysterols were tested. Although percentage of SR values vary from mouse- to-mouse and day-to-day, at any given experiment OS-incubated cells never had a higher percentage of SR than control cells. The striking inhibition of NKCA by 5~ hydroxy-6-ketocholestanol and 7/3-hydroxycholesterol can be seen in Fig. 1. Less potent inhibitors, in decreasing order of potency, were 6-ketocholestanol, 7a-hydroxycholesterol, and 7-ketocholesterol. In contrast, the presence of 25hydroxycholesterol and 20a-hydroxycholesterol did not result in a significant inhibition compared to their controls in each experiment. Table 1 shows the percentage inhibition of NKCA caused by various OS, and the percentage of viability of lymphocytes after incubation with OS. There was no decreased viability in the effector cells after 45 min incubation with 1 o-5 M OS.

Table 2 shows the effect of preincubation with OS at different concentrations on NKCA. At 2.5 X 10M6 A4 no inhibition of NKCA was observed by 5a-hydroxy-6- ketocholestanol, 6-ketocholestanol, and 25-hydroxycholesterol. A twofold increase in the concentration of OS in the incubation medium resulted in 4 1 and 3 1% inhibition by 5a-hydroxy-6-ketocholestanol and 6-ketocholestanol while no inhibition was observed with 25-hydroxycholesterol. A fourfold increase in OS concentration to 10m5

. . . . . . . . . - , . . . ’ :. *

C 5o6K 78 6K 7 a 7K 25 20

osc

FIG. I A. scattergram showing the results of NK cell cytotoxicity expressed as 5’Cr specific release in the presence of various oxysterols (c, control; 5a6k, 5a-hydroxy-6-ketocholestanol; 7& 7@-hydroxycholesterol; 6k, 6-ketocholestanol; 7q 7a-hydroxycholesterol; 7k, 7-ketocholesterol; 25,25-hydroxycholesterol; 20,20~ hydroxycholesterol). Concentration of oxysterols in the incubation medium was 10m5 M, and the E:T ratio was 50. Control cells were incubated in 1% ethanol. OSC denotes oxygenated sterol compound.


TABLE 1

Inhibition of NK Cell Cytotoxicity after Incubation of Effector Cells with 10eJ M Oxysterol for 45 Min in Lipoprotein Depleted Medium

Oxysterol

Sd-Hydroxy-6-ketocholestanol 7j+Hydroxycholesterol 6-Ketocholestanol 7wHydroxycholesterol ‘I-Ketocholesterol 25-Hydroxycholesterol 20cY-Hydroxycholesterol

% NKCA inhibition”

(means + SE)

9lr 8 85+ 8 77+ 18 5lk 12 47r 17

6rt 4 4+ 4

‘% Effector cell viability

(means + SE)

902 4 92+ 9 99* 1 94 t 10 96-t 4 87~ 10 99+ 2

No. of experiments

18 13 12 10 9 I 4

Note. Effector cell to target cell ratio is 50: 1.

’ % Inhibition = % specific “Cr release in control medium - I specific JICr release in OS x loo

% specific “Cr release in control medium

M resulted in 84 and 86% inhibition by the first two OS, while the presence of 25- hydroxychoiestero1 resulted in a 23% inhibition. At a IO-fold increase in OS concentration to 2.5 X lop5 M 100 and 94% inhibition was observed with 5cu-hydroxyd- ketocholestanol and 6-ketocholestanol, respectively, whereas the inhibition with 25- hydroxycholesterol was at the 29% level in this set of experiments. The same changes in concentration of OS in the preincubation medium resulted in no loss of effector cell viability except when 5a-hydroxy-6-ketocholestanol was used which resulted in a loss of effector cell viability at the highest concentration level (data not shown).

Table 3 shows that the presence of lipoprotein replete 20% AB serum in the preincubation medium prevented the inhibitory effect of OS on NKCA. Control incubation media containing 20% ABS and 20% LPDS gave similar results and the presence of 1% ethanol in the media caused no inhibition (data not shown).

TABLE 2

Inhibition of Natural Killer Cell Cytotoxicity in the Presence of Different Concentrations of OS in the Incubation Medium

Concentration of OS in incubation medium (% inhibitionr

Oxysterol 2.5 x lO-‘j M 5.0 x lo-6 M 1.0 x 10-S M 2.5 X lO-J M

SLu-Hydroxy-6-ketocholestanol 0 41+ 4 84k 6 100 6-Ketocholestanol 0 31& 15 86k 10 94k 8 25-Hydroxycholesterol 0 0 23 -c 22 29 rt 20

Note. Results of three experiments are expressed as percentage of inhibition (means + SE). Effector cell to target cell ratio is 50: 1.

n % Inhibition = 96 SR in control medium - W SR in OS-containing medium % SR in control medium

x loo.


TABLE 3

Presence of Lipoproteins (AB Serum) in the Incubation Medium Prevent the Inhibition of NK Cell Cytotoxicity by OS (lo-’ M)

Percentage specific release (means f SE)

Oxysterols 20% LPDS + ethanol 20% LPDS + OS 20% ABS + OS

SwHydroxy-6-ketocholestanol 11 + 2.0” 0.6 f 0.1 11 f 0.5 7B-Hydroxycholesterol 33 t- 11.0 3.0 f 1.0 35 f 9.0 6-Ketocholestanol 15 t- 3.0 0.5 f 0.2 17 + 4.0 7wHydroxycholesterol 33 ?I 11.0 12.0 + 5.0 33 It 12.0 25-Hydroxycholesterol 29 4 0.9 30 f 2.0 33 2 3.3

a Results of three experiments are expressed as mean percentage of specific release + SE. Effector cell to target cell ratio is 50: 1.

DISCUSSION

The results described above indicate that certain oxysterols are potent inhibitors of mouse NK cell cytotoxicity. The position of oxygen in the OS molecule appears to determine the extent of this inhibitory effect. While the OS oxidized at the C-5, C-6 and C-7 carbons are inhibitory, the OS oxidized at the C-20 or C-25 position have minimal inhibitory effect on NKCA. Figure 2 shows a schematic representation of OS used in this study. According to the currently accepted model of the plasma membrane lipid bilayer (10) shown in Fig. 3, the hydroxyl or keto groups of various OS will be in juxtaposition with different areas of the adjacent phospholipid molecules depending on the site of oxidation. The OS molecule is expected to orient itself similar to the cholesterol molecule with the hydrophilic 3@-OH group being closer to the membrane surface and the C-17 acyl chain extending into the hydrophobic interior of the membrane (10). This orientation results in the interactions of C-5, C-6, or C-7 OS with the surrounding phospholipids in regions that are closer to the hydrophilic surface which contain their head groups and acyl chain ester bonds. In contrast, the interactions of C-20 or C-25 OS are with the tail portions of phospholipid acyl chains in the hydrophobic interior of the bilayer membrane. These are relative positions since molecules in the membranes are mobile. Specifically, the lipid acyl chains are disordered with numerous gauche and cis bonds resulting in a “kink” structure which is less ordered than the all-truns lipid acyl chain structure pictured in Fig. 3. Thermal motion allows for the lipid molecules to translate laterally within the membrane as well as to allow for motion perpendicular to the membrane surface resulting in small perturb- antes. Finally, thermomechanical fluctuations can be induced within a membrane allowing for ripples of membrane segments to occur. Since the interaction of cholesterol with neighboring phospholipids has a profound effect on the physical properties of the membrane (11, 12) the presence of OS in the plasma membrane, may alter the structure and organization of the lipid bilayer and thereby affect important cellular functions. The specific interactions between sterols and phospholipids can be either affected by oxysterol induced changes in the phospholipid hydrogen bonding network as well as packing changes caused by steric hindrances due to the presence of oxide groups at different positions on the cholesterol molecule.

546 SHORT COMMUNICATIONS p3 ‘lIi5?-

R4

HO \ \ Rl

932

2b

nci. 2. Schematic representation of sterols: (a) cholesterol, R,=H, R*=H, R,=H, %=H; 7whydroxy- cholesterol, R,=H, R2=OH, R,=H, R.+=H; 7P-hydroxycholesterol, R,=OH, R*=H, R3=H, &=H; 20whydroxycholestero1, R,=H, R2=H, R3=OH, %=H; 25-hydroxycholesterol, RI=H, R,=H, R3=H, %=OH; (b) 6-ketocholestanol, RS=H; ScY-hydroxy-6-ketocholestanol, RS=OH; (c) 7-ketocholesterol.

Yachnin et al. (5) have shown that oxysterols are inserted into peripheral blood cell membranes after a brief incubation in lipoprotein-depleted serum. The effects of oxysterols on various cellular functions and morphology can be explained by two mechanisms: (a) some OS are potent inhibitors of 3-hydroxy-3-methyl-glutaryl-CoA (HMG- CoA) reductase, the rate limiting enzyme in cellular cholesterol synthesis; or (b) OS inserted in the plasma membrane as cholesterol analogues alter membrane physical properties (5). Preincubation of cytolytic T lymphocytes (CTL) with 25-hydroxycholesterol for 24 hr causes inhibition of CTL activity, which is due to inhibition of HMG- CoA reductase (13). Exposure to various OS in culture for l-5 days inhibits CTL activity ( 14) or plaque-forming cell (PFC) response ( 15). The inhibition of CTL and PFC by 25-hydroxycholesterol and 20a-hydroxycholesterol after a prolonged exposure of the effector cells to these compounds is due to inhibition of cholesterol synthesis ( 14, 15). In contrast, even prolonged incubation with 78- and 7a-hydroxycholesterol does not result in significant depression of sterol biosynthesis while they inhibit the CTL activity, suggesting that the inhibition of cytolytic function by certain OS is not due to inhibition of HMG-CoA reductase. In our experiments the short preincubation


PHOSPHOLIPID HEAD GROUP

- ,

PHOSPHOLlPlD AWL CHAINS

FIG. 3. Schematic representation of bilayer plasma membrane composed of phospholipids and cholesterol. (R indicates serine, ethanolamine, choline, etc.).

time is not sufficient to induce significant changes in cellular sterol synthesis; however, the cell-membrane is rapidly enriched in OS due to the lack of lipoproteins in the incubation medium. When incubation is carried out in lipoprotein containing medium the insertion of OS into effector cells is inhibited due to the binding of OS by lipoproteins (5). The inhibition of NK cell lytic activity by 5cu-hydroxy-6-keto- and 6-ketocholestanol and by 7@-hydroxy-, 7a-hydroxy-, and 7-ketocholesterol in our experiments is most likely due to insertion of these OS into membrane and not due to inhibition of HMG- CoA reductase. Even though 25- and 20a-hydroxycholesterol are significantly more potent inhibitors of HMG-CoA reductase compared to the other OS, they have minimal inhibitory effect on NKCA. Similarly, Streuli et al. (16) found that the capacity of oxygenated sterol compounds to inhibit E-rosette formation was independent of their inhibitory effect on sterol synthesis. Our results support the membrane insertion hy- pothesis of Yachnin et al. (5).

It is now widely accepted that the dynamics of membrane proteins play an important role in cellular functions. The fluidity of the lipid bilayer membrane matrix is the


main mechanism controlling the dynamic features of membrane proteins ( 17- 19). Thus, the function of membrane carriers, receptors, and enzymes can be markedly influenced by changes in the lipid microviscosity (20-23). Roozemond and Bonavida observed that fluidization or rigidification of the plasma membrane of either effector or target cells affect different stages of the NK cell-mediated cytolytic events (9).

Cholesterol is a particularly important component of plasma membranes of mam- malian cells ( 11). The cholesterol content of the membrane helps determine its fluidity ( 12,24). Oxysterols which are oxidation products of cholesterol result from lipid per- oxidation caused by oxygen radicals or other free radicals (l-4). These molecules may reach sufficient quantities in plasma membranes of human blood cells to cause changes of the membrane properties (2). We have shown that OS insertion results in changes of membrane fluidity and protein structure in normal human red blood cells (7). Insertion of 20a-hydroxycholesterol ( 10% of total membrane cholesterol) immobilized the lipid acyl chains to a degree equivalent to enriching total membrane cholesterol by 50% while red cell membrane protein helical structure was not altered. The insertion of 7a-hydroxycholesterol into erythrocyte membranes, however, resulted in an increase in protein helical structure, with no change in lipid acyl chain mobility. The mechanism of cytotoxicity by NK cells, although not completely elucidated, involve membrane- related events such as recognition, adhesion, activation, and signal transduction which may be influenced by the physical characteristics of the plasma membrane (9, 13, 17- 23, 25, 26). Changes in membrane fluidity could alter the exposure of receptors, in- tramembrane mobility or conformation of proteins which would result in impairment of cell functions dependent on membrane proteins. Our results suggest that inhibition of NK cell activity by certain OS may be due to changes in membrane protein conformation. Preliminary results show that OS do not inhibit the conjugation of effector and target cells (unpublished data). Other possible mechanisms include the effects of OS on the release of lytic proteins from NK cells, mobility of NK cells, and the activities of enzymes and other proteins in NK cell membrane which play a potential role in the lytic mechanism.

ACKNOWLEDGMENT

We thank Dr. Ronald Herberman (Pittsburgh Cancer Institute, Pittsburgh, PA) for his critical review of the manuscript.

REFERENCES

1. Smith, L. L., and Johnson, B. H., Free Radicals Med. Biol. 7, 285, 1989. 2. Smith, L. L., “Cholesterol Autoxidation.” Plenum Press, New York, 198 1. 3. Bachowski, G. J., Thomas, J. P., and Girotti, A. W., Lipids 23, 580, 1988. 4. Morin, R. J., and Peng, S. K., Ann. Clin. Lab. Sci. 19, 225, 1989. 5. Yachnin, S., Streuli, R., Gordon, L. I., and Hsu, R., Curr. Top. Hematol. 2, 245, 1979. 6. Gordon, L. I., Bass, J., and Yachnin, S., Proc. Natl. Acad. Sci. USA 17,4313, 1980. 7. Rooney, M. W., Yachnin, S., Kucuk, O., Lis, L. J., and Kauffman, J. W., Biochim. Biophys. Acta 820,

33, 1985. 8. Herberman, R. B., Reynolds, C. W., and Ortaldo, J. R., Annu. Rev. Zmmunol. 4,65 1, 1986. 9. Roozemond, R. C., and Bonavida, B., .T. Zmmunol. 134,2209, 1985.

10. Yeagle, P., “The Membranes of Cells” pp. 120-138. Academic Press, San Diego, CA, 1987. 11. Nes, W. R., Lipids 9, 596, 1974. 12. Demel, R. A., and DeKryuff, B., Biochim. Biophys. Acta 457, 109, 1976. 13. Heiniger, H.-J., Brunner, K. T., and Cerottini, J.-C., Proc. Natl. Acad. Sci. USA 75, 5683, 1978.


14. Sprangrude, G. J., Sherris, D., and Daynes, R. A., Transplantation 33, 482, 1982. 15. Humphries, G. M. K., and McConnell, H. M., J. Immunol. 122, 121, 1979. 16. Streuli, R. A., Chung, J., Scanu, A. M., and Yachnin, S., J. Immunol. 123, 2897, 1979. 17. Borochov, H., and Shinitzky, M., Proc. Natl. Acad. Sci. USA 73,4526, 1976. 18. Edelman, G. M., Yahara, I., and Wang, J. L., Proc. Natl. Acad. Sci. USA 70, 1442, 1973. 19. Edelman, G. M., Science 192, 2 18, 1976. 20. Papahadjopoulos, D., Cowden, M., and Kimelberg, H., Biochim. Biophys. Acta 330,8, 1973. 21. Wiley, J. S., and Cooper, R. A., Biochim. Biophys. Acta 413, 425, 1975. 22. Inbar, M., and Shinitzky, M., Eur. J. Immunol. 5, 166, 1975. 23. Farias, R. N., Bloj, B., Morero, R. D., Sineniz, F., and Trucco, R. E., Biochim. Biophys. Acta 415, 23 1,

1975. 24. Cooper, R. A., Leslie, M. H., FischkofI, S., Shinitzky, M., and Shattil, S. J., Biochemistry 17, 327, 1978. 25. Henkart, P. A., Annu. Rev. Immunol. 3, 31, 1985. 26. Allison, A. C., and Ferluga, J., N. Engl. J. Med. 295, 165, 1976.

inhibition of nk cell-mediated cytotoxicity by oxysterols

Documents