Cyclosporin A depolarizes cytoplasmic membrane potential and interacts with Ca2+ ionophores

Biochimica et Biophysica Acta 886 (1986) 353-360 Elsevier

353

BBA 11708

Cyclosporin A depolarizes cytoplasmic m e m b r a n e potential and interacts with Ca 2 + i o n o p h o r e s

LS.szl6 M ~ t y u s a, M a r g i t Balfizs a, Adorjf.n A s z a l 6 s SS.ndor D a m j a n o v i c h a

b,

Sally M u l h e r n b a n d

Department of Biophysics, University Medical School of Debrecen, H-4012 Debrecen (Hungary), and b Division of Drug Biology and Nutrition, Food and Drug Administration, Washington, DC 20204 ¢U.S.A.) (Received August 9th, 1985) (Revised manuscript received January 2nd, 1986)

Key words: Cyclosporin; Membrane potential: Ca 2 + ionophore; Depolarization

Cytoplasmic membrane potential of mouse lymphocytes was determined with flow cytometry and fluorescence spectroscopy using 3,3'-dihexylcarbocyanine iodide (DiOC6(3)). The amount of this lipophilic cation incorporated into the cytoplasmic membrane is dependent upon the transmembrane potential, so the dye is suitable for continuous monitoring of this parameter, under controlled conditions. Membrane potential of the cells was decreased in the presence of cyclosporin A and cyclosporin G in a dose-dependent manner. However, the depolarization caused by Ca2+ ionophores, ionomycin and A23187, was reduced in the presence of cyclosporin A. Electron spin resonance spectroscopy with 5-doxylstearic acid as a probe indicated that cyclosporin A decreased the apparent motional freedom of membrane lipids. These data suggest incorporation of cyclosporin A into the cytoplasmic membrane, causing changes in ion fluxes. The membrane potential change induced by cyclosporin A may have selective biological consequences in certain subpopulations of iymphocytes.

Introduction

Cyclosporin [1] is a neutral cyclic peptide composed of 11 amino acid residues. Its stereospecific synthesis was given by Wenger [2,3]. Cyclosporins are widely used as immunsuppressive drugs in solid organ transplantation and in treatment of autoimmune disorders [4,5]. Unfortunately, the cytostatic effect of these antibiotics is combined with strong nephrotoxicity [6-8]. Hiestand and his co-workers [9] reported about new derivatives of cyclosporins with very similar properties to cyclosporin A without nephrotoxic side-effects. Abbreviations: DiOC6(3 ), 3,3'-dihexyloxacarbocyanine iodide; mRNA; messenger RNA; ESR, electron spin resonance; DMSO, dimethylsulfoxide.

The mechanism of action of cyclosporins is unclear. It appears from in vitro studies that the lymphokine, interleukin-2, production is inhibited by cyclosporin A at the level of mRNA synthesis [10,11]. Cyclosporin A does not block interleukin-2 in mouse lymphocyte cultures of Balb/C and C57BL/67 mouse spleen cells [12]. Handschumacher et al. [13] reported about cyclophylin, a specific cytosolic-binding 16 kDa protein that binds cyclosporin A. In order to understand the mechanism of the effect of cyclosporin A on lymphocyte activity, its influence of on the events triggering early cell activation must be investigated. Since cell activation is a very complex phenomenon, the selection of a sensitive early signal to monitor changes caused by cyclosporin A has great importance.

0167-4889/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)

354 Changes in receptor distribution and mobility may play a decisive role in the mediation between external stimuli and the complex metabolic processes of a cell connected with transformation [14]. The proximity and distribution of cell surface receptors and ligand-binding sites can be measured by fluorescence resonance energy transfer also in a flow cytometer [15]. These changes may also be an early indicator of cell activation. One of the earliest effect in lymphocyte activation is thought to be an increase in the ion flux of Ca 2 + into the cells [16]. In the present work, we used flow cytometry and fluorescence spectroscopy [17] to measure another possible early event, changes in the membrane potential of mouse lymphocytes in the presence and absence of cyclosporins. We have used the potential-sensitive fluorescent probe DiOC6(3) monitoring the membrane potential [18]. Our results suggest that membrane depolarization of mouse lymphocytes is one of the earliest effect of cyclosporin A. The depolarization occurred with pharmacological concentrations of cyclosporin A [19 22] (i.e., with concentrations producing immunosuppression); no substantial differences were found between the effects of cyclosporins A and G. Electron spin resonance spectrometry was shown to be a useful tool for investigating plasma membrane and drug interaction [23,24]. We used this technique to study whether cyclosporin A would instantly bind to lymphocytes. Such binding could result in plasma membrane disturbances. M a t e r i a l s and M e t h o d s

Cells. Spleen or thymus cells from A / J , B a l b / C or C3H mice were suspended in phosphatebuffered saline containing low levels of potassium (1.5-4.2 mM KCI). Erythrocytes and lymphocytes were separated by gradient centrifugation on a Histopaque-1077 cushion. Viability of the cells was above 95%, as determined with the fluorescein diacetate test [25]. All experiments were carried out at room temperature. Flow ~ytometric determination of membrane potential. The cell number was adjusted to 2- 105 c e l l s / m l and the samples were stained with DiOC6(3 ), dissolved in DMSO. The dye c o n -

centration used in each experiment was determined to be non-toxic to the cells. The amount of the dye in the cell interior and membrane (and also the fluorescence intensity) depended upon the cytoplasmic membrane potential, as it was tested by changing the extracellular potassium concentration or applying ionophores with known effects on the cytoplasmic membrane potential [18]. Coulter EPICS V and FACS III (Becton-Dickinson) flow cytometers were used for data acquisition and analysis. The laser was tuned to 488 nm and the output was kept at 400 mW. The flow-rate of the cells was about 500 cell/s to obtain good resolution. The forward low-angle light scatter and right-angle green fluorescence signals were collected. The fluorescence was gated on the scatter signal to minimize artefacts. The coefficient of variation and the mean were calculated for each histogram. Spectrofluorimetric experiments. These experiments were carried out in a Perkin-Elmer MPF4 4 / A spectrofluorimeter. Samples were excited at 488 nm and the emission was monitored at 520 nm. The slits of the grating monochromators were 6 / 6 nm. The cell density was 106 c e l l / m l in rectangular 1 × 1 cm cuvettes. When the membrane potential was determined with 0.7/~M dye concentration, the fluorescence intensity was inversly proportional to the dye concentration inside of the cells, i.e., to the membrane potential. The reason for this phenomenon is the known aggregation of the dye above a certain concentration level in the intracellular space, resulting in a quenching of fluorescence [26]. Spin-labeling the cell membranes. The spin label, 5-doxylstearic acid, was dissolved in ethanol and the appropriate volume was dried into glass testtubes under vacuum. The usual amount of spin label was 8 . 1 0 s mol for 2.107 spleen lymphocytes. The cell suspension was added in about 30 /~l phosphate-buffered saline to the test-tubes containing the spin label. After the necessary contact time (at least 30 s), the cell suspension was transferred into a 50-/d micropipette capillary tube and sealed at the bottom with Critoseal ~. ESR spectrometry. ESR spectra were recorded at X-band with a Varian E-9 Century series spectrometer operated at 9.5 kHz, 100 field modulation, 4 G modulation amplitude, 100 G sweep

355

range and 20 mW microwave power. The temperature of the probe was set to 20+_0.5°C by a Varian variable temperature accessory, using N 2 gas flow. To evaluate the effect of cyclosporin A on lymphocyte membranes, 2- 107 spleen cells were labeled and an ESR spectrum was obtained. A similar procedure was followed for assessing the effect of 1 ~1 DMSO on the same volume of identical spleen cell suspensions. The amount of cyclosporin A used for the ESR measurements was 10-times the amount used in the membrane potential measurements. Evaluation of the E S R spectra. The equation used for these calculations is from Hubbel and McConnell [27] as used in experiments similar to ours by Bunerfield et al. [28]: S-

(Tfl -- T I )aN,zL ( T , , - T ± ) ~ I a N,

where x L refers to single-crystal parameters obtained from the data of Hubbel and McConnell [27]. The definitions of 2T~, and 2T~ are shown also in their paper, and a N, = 1/3(TI, + 2 T I ) . The correction factor of 1.6 G was used in the estimation of 2T~ values. ESR spectra obtained from membranes of intact cells with spin labels of 5doxylstearic acid show contribution mostly from spin labels of restricted motion with negligible contribution from free-moving spin labels. The spin-label concentration was kept low enough to avoid spin-spin interaction. Standard deviation of the measurements of 0.012T' were between 0.012 and 0.007 G for the three sets of experiments shown in Table I. Materials. Valinomycin, ionomycin and ionophore A23187 were obtained from Calbiochem; DiOC6(3 ) was from Kodak; 5-doxylstearic acid was from Aldrich-Chemie, GmbH. All other reagents were of analytical grade. Results and Discussion

The effect of cyclosporins A and G on the O'tOplasmic membrane potential Fig. 1 shows the effect of cyclosporin A on the membrane potential of mouse spleen cells measured in a flow cytometer. Since in these experiments 250 nM DIO6(3 ) was applied as a potential

t9

~6

~

\/

~6°t ~41 3)

4

/2=

o

/" /-

o/"

Ib ' i b - ~

D

Ftuorescence IntensiN (arbitrary units) Fig. 1. Effect of cyclosporin A on the membrane potential of mouse spleen cells measured in a flow cytometer. 2.105 cells/ml were stained with 250 n m DiOC6(3 ). Curve 1 indicates the fluorescence intensity distribution of DiOC6(3 ) dye without cyclosporin A treatment. Curves 2, 3 and 4 were taken after treatment with 0.5, 1 and 10 ~ g / m l of cyclosporin A, respectively. The X-axis shows the fluorescence intensity (proportional with the membrane potential), the Y-axis indicates the number of cells. The inset shows the spectrofluorimetric titration of the depolarizing effect of cyclosporin A on mouse spleen cells. The change of the fluorescence intensity is plotted against the cyclosporin A concentration. Here, the increase of the fluorescence intensity is proportional with the depolarization. | n these types of experiments, 1 • 106 cells/ml were stained with 0.7 ~M DiOC6(3 ).

probe, the fluorescence intensity was proportional to the dye concentration in the cell membrane and cytoplasm. The amount of the dye changes with the transmembrane potential in these cellular compartments. Pharmacological doses of cyclosporin A decreased the cytoplasmic membrane potential, i.e., the histograms were shifted to the left side. The data shown in Fig. 1 clearly demonstrate the dose dependence of the fluorescence distribution of spleen cells pretreated with 0.5, 1 and 10 t~g/ml cyclosporin A. The inset shows the spectrofluorimetric titration of the depolarizing effect of cyclosporin A on mouse spleen cells. The effect of cyclosporin A on the membrane potential of cells was similar as determined by flow cytometry. The cyclosporin A has the same effect on the membrane potential of mouse thymus cells (data not shown). Relevant controls, treated with the same amount of DMSO, the solvent for cyclosporins A and G, did not cause changes in the fluorescence intensity of the cells. The cyclosporin G had essentially the same effects on the membrane potential as did the cyclosporin A.

356 Thymus cells had similar sensitivity and direction of changes in membrane potential as did spleen cells. In these experiments, the cells were suspended in phosphate-buffered saline containing 10% fetal calf serum to stabilize cells.

z!

t.} o

The protective effect of insulin pretreatment Insulin is known to counteract the immunosuppressive effect of cyclosporin A [28]. The insulin provided a significant protection against the decrease of membrane potential upon addition of cyclosporin A.

\ i

"1n

z

i Ptuorescence

Scaring the membrane potential changes The range for observing membrane potential changes was established by adding 1-5 ~ g / m l valinomycin to determine the upper limit, and 1.0 /zg/ml of gramicidin for the lower limit, both dissolved in DMSO. Changing the extracellular potassium concentration between 1.5 and 160 mM resulted in a dramatic change of fluorescence signals. The fluorescence signals were high at the low, and very low at the highest potassium concentration. However, the lowest fluorescence signals were measured after treating the cells with the channel-forming ionophore, gramicidin. These effects provided firm evidence that changes in the fluorescence signals were parallel with those of the membrane potential. The selection of the flow-cytometer settings and the dye concentration made it possible that the mean zero potential was near to the left side of the coordinate system for each histogram, and the mean fluorescence values of the controls stained with DiOC6(3 ) were very close to those of the cells having the maximum potential values (i.e., the fluorescence intensity at 1.5 mM extracellular potassium concentration). Low amounts of valinomycin (1-5 /~g/ml) at this low potassium concentration could increase the mean value of the fluorescence intensity distribution of the cell population by 5-10% (Fig. 2). Thus, the fluorescence intensity displayed in Figs. 1 and 2 can be taken as directly proportional to the membrane potential. The cytoplasmic membrane potential of resting lymphocytes is, for all practical purposes, a Donnan potential of the unequal distribution of the potassium ions between the extra- and intracellular space, so that linear potassium con-

Intensity

Fig. 2. Effect of ionophores on the membrane potential of mouse thymus cells. - - , Fluorescence intensity distribution of thymus cells (3-105 cells/ml) stained with DiOC6(3) (50 nM) as it was measured in a flow cytometer: . . . . . of addition of 2 #g/ml valinomycin; .-. , after addition of 1 ~tg/ml gramicidin. centration changes are expected to alter the membrane potential in a logarithmic fashion. The cationic dyes have directly been exchanged with inorganic cations, such as potassium or sodium, across the cell membrane. This fact makes the fluorescence intensity scales of the flow cytometric histogram directly and logarithmically proportional with cation transport.

Effect of mobile ionophores on the membrane potential in the presence of cyclosporin A The potassium-specific valinomycin decreased the effect of cyclosporin A on the lymphocyte membrane potential (data not shown). However, since certain dose-range valinomycin had a membrane-potential-changing effect into the opposite direction as cyclosporin A, tow other mobile carriers were also tested in combination with cyclosporin A. Time-course of changes in fluorescence of DiOC6(3)-stained thymic cells was followed in this case in a flow cytometer. Fig. 3A shows the effect of 1 /~g/ml ionomycin on membrane potential of mouse thymus cells. The solid line indicates the fluorescence intensity distribution of control population, while the dotted line shows the depolarizing effect of ionomycin. In Fig. 3B we can see the result of a similar experiment with 1.5 /xg/ml cyclosporin A-pretreated cells. The cyclosporin A itself caused a small

357

I r

l

10d

A

A I

I

u

A

8C (1/ t.)

I

~CD

60

u

0

•

..Q

E

B

J

Z

0

1

I

I

1

I

1

2

3

4

5

°100

B

f' [

. I

I

gtuorescence Intesity Fig. 3. Effect of cyclosporin A on the depolarization caused by ionomycin of mouse thymus cells. (A) Fluorescence intensity distribution of thymus cells (solid line). The curve indicated with dashed line was taken after addition of 1 / t g / m l ionomycin. The extracellular Ca 2÷ concentration was 1 mM. The staining was identical to that described in the legend to Fig. 2. (B) Effect of ionomycin after treating the cells with 1.5/~g/ml cyclosporin A. The solid line shows the fluorescence intensity distribution of cyclosporin A-treated cells. The curve indicated with dashed line was obtained after addition of 1 /~g/ml ionomycin.

decrease in the membrane potential of cells (solid line). Addition of 1 /~g/ml ionomycin (dashed line) caused a further depolarization, but this effect was significantly smaller compared to the control sample. This phenomenon was time-dependent, as shown in Fig. 4A. The full circles indicate the fluorescence intensity of thymus cells treated with 1/~g/ml ionomycin. The open circles show the protective effect of cyclosporin A on the depolarizing effect of ionomycin. Pretreatment of cells with cyclosporin A (1.5 /tg/ml cyclosporin A) significantly decreased the effect of the ionophore A23187, despite the fact that it itself causes depolarization. Data in Fig. 4B show an identical counteraction of cyclosporin A, except ionophore

6O 0

1

2

3 4 Time [min]

5

Fig. 4. Protective effect of cyclosporin A on the depolarization caused by ionomycin and ionophore A23187 of mouse thymus cells. (A) The time dependence of the depolarization caused by 1 /~g/ml ionomycin without cyclosporin A pretreatment (e) and with 1.5 /tg/ml cyclosporin A pretreatment (©). The fluorescence intensities were obtained from the mean values of flow cytometric histograms. The experimental conditions were identical to those in Fig. 3. (B) The results of a similar experiment are shown, except that the thymus cells have been treated with 1 / t g / m l ionophore A23187 instead of ionomycin.

A23187 (1 /tg/ml) ws used instead of ionomycin. These data were confirmed also with spectrofluorimetric investigations. Specificity of the action of cyclosporin A on the membrane potential

Since most of the experiments were carried out with spleen cells, the question of selectivity arises, as the spleen cells consist of approximately equal amounts of T and B lymphocytes. On the other hand, the immunosuppressive effect of cyclosporin A is very specifically limited to cytolytic and helper subclasses of T lymphocytes [20,29]. The

358

ESR spectrometry of spleen lympho~ytes pretreated with ~Tclosporin A

"

,1 Fig. 5. ESR spectra of 5-doxylstearic acid probe in mouse spleen cells before (top) and after (bottom) addition of 10 p,g cyclosporin A.

spectrofluorimetric titration of thymus or spleen cells with cyclosporin A for changes in m e m b r a n e potential did not show any selectivity. The decrease in m e m b r a n e potential of T and B lymphocytes occurred with equal probability. Data are available that cyclosporin A influences the biological function of B lymphocytes as well [30]. Thus, it is quite acceptable that the cyclosporin A decreased the m e m b r a n e potential of both T and B lymphocytes. The specific response of the cellular immunity and the relatively unharmed humoral immunity come from the different degree of response of the two cell types to the same depolarizing effect of cyclosporin A.

Results of the ESR experiments are summarized in Table I and Fig. 5. This probe is known to provide information on the mobility of phospholipids in the plasma membrane [28]. Parallel and perpendicular components of the motionally averaged nitrogen hyperfine tensor, order parameter and polarity factor, the usual spectral parameters of this type of measurement, were compared with cyclosporin A-pretreated and untreated spleen cells. The results provide evidence that cyclosporin A pretreatment significantly decreased the mobility of the phospholipids in the plasma membrane, as expressed by the change in the order parameter, S, value. Since these measurements were done immediately after the addition of cyclosporin A to the spin-labeled cells, a rapid equilibrium between cyclosporin A and cellular membrane c o m p o n e n t s can be assumed. Although for obvious technical reasons, i.e., the limited sensitivity of the measurement, the concentration of the cyclosporin A had to be increased, which, together with the cell ratio, was not higher than in the case of the highest cyclosporin A concentration in the flow cytometric experiments. A recent publication on the effect of altered m e m b r a n e fluidity on natural killer cell-mediated cytotoxicity emphasizes the possible importance of our finding, regarding the slight but still significant increase in m e m b r a n e rigidity upon addition of cyclosporin A [31]. In summary, results from these experiments support the conclusion that cyclosporin A binds to

TABLE 1 CHARACTERISTIC ESR PARAMETERS OF SPLEEN CELLS Characteristic ESR parameters of spleen cells before and after DMSO or cyclosporin A treatment. Significant differences can be found in the order parameter, S, after cyelosporin A treatment, indicating a decreased mobility of the lipid probe in the membrane. Cell treatment

T~

T~

T'= T~ - Ti

a x,

S

n

Spleen cells, 2.10 v (control) Spleen cells, 2.107 + DMSO (1 /~1) Spleen cells, 2. 107+ 100/~g of cyclosporin A in 1 /zl DMSO

23.75

8.87

14.88

13.83

0.601

4

23.80

8.90

14.90

13.87

0.600

2

24.25

8.70

15.55

13.88

0.626

6

359

the cytoplasmic membrane and either directly or indirectly decreases the transmembrane potential. Both the flow cytometric and spectrofluorimetric measurements confirmed that it had an effect on various mobile carrier ionophores but did not alter the effect of the channel-forming agent, gramicidin, or of high extracellular potassium (160 mM). In addition, the ESR experiments indicated a decreased motional freedom of lipids in the cytoplasmic membrane immediately upon addition of cyclosporin A to the cells. The principal target of the biological action of cyclosporin A is thought to be a subpopulation of T cells responsible for the production of interleukin-2 [32,33]. Since the splenic lymphocytes are made up of approximately equal numbers of T and B cells, one would expect the cyclosporin A to affect less than half of the splenic cell preparation. However, the results indicate that all lymphocytes are influenced by cyclosporin A. This is consistent with the observation that cyclosporin A also interacts with B cells [30]. Thymus lymphocytes showed essentially the same type of interaction with cyclosporin A as the spleen cells. This would suggest that cyclosporin A nonspecifically binds to the lipid membrane a n d / o r perhaps to a protein in the membrane of the cell. The selectivity of the action of cyclosporin A may be thus due to the particular requirements of different cells for ion fluxes directing regulatory processes during their metabolism. It is also possible, as was suggested earlier [33], that membranes of different cells 'react' differently with agents such as cyclosporin A, thereby allowing selectivity in cellular responses among the different types of cells. Recently, it was observed that cyclosporin A binds to calmodulin [34]. Since cyclosporin A interfers with the normal ion transport of the cells, it is possible that Ca 2+ transport changes precede binding of cyclosporin A to calmodulin. Also interleukin-2 production can be inhibited at the mRNA level when the ion concentration and distribution among the different compartments, such as intracellular and extracellular space, are changed. It is known that the D N A - d e p e n d e n t R N A polymerase of both bacterial and eukaryotic cells are un~,:~ the control of the ion concentration of their microenvironment [35]. Since the doses used in this study were in the

range where immunosuppression occurs [20-22, 24], it is reasonable to conclude that the cyclosporin A-caused cation transport changes in the cellular membrane may represent the earliest and principal mode of action of cyclosporin A as an immunosuppressive agent. The changes in ion fluxes of the cytoplasmic membrane may have far-fetching consequences in the cellular metabolism [36,37]. However, to verify our above view, further experiments are necessary. Experiments are currently in progress in our laboratory to identify the subpopulations of mouse and human lymphocytes with different responsiveness to cyclosporin A and to identify biophysical and other parameters of this responsiveness. References 1 Borel, J.F., Feurer, C., Gubler, U. and Stahelin, H. (1976) Agents Actions 6, 468-475 2 Wenger, M.R. (1983) Helv. Chim. Acta 66, 2308 2321 3 Wenger, M.R. (1983) Helv. Chim. Acta 66, 2672-2702 4 Borel, J.F. (1983) Transplant. Proc. 15, Suppl. 1, 2219 5 Cohen, D.J., Loertscher, R., Rubin, M.F., Tilney, N.L., Carpenter, C.B, and Strom, T.B. (1984) Ann. Int. Med. 101, 667-682 6 Whiting, P.H., Thompson, A.W., Blair, T.T. and Simpson, J.G. (1982) Br. J. Exp. Pathol. 63, 88-94 7 Kennedy, M.S., Dug, H.J., Siegel, M., Crowley, J.J., Storb, R. and Thomas, E.D. (1983) Transplantation 35, 211-215 8 Baxter, C.R., Duggin, G.G., Horvfith, J.S., Hail, B.M. and Tiller, D.J. (1984) Res. Commun. Chem. Pathol. Pharmacol. 45, 69-80 9 Hiestand, P.C., Gunn, H.C., Gale, J.M., Ruffel, B. and Borel, J.F. (1985) Immunology 55, 249-255 10 Kr6nke, M., Leonard, W.J., Depper, J.M., Arya, S.K., Wong-Staal, F., Gallo, R.C., Waldmann, T.A, and Greene, W.C. (1984) Proc. Natl. Acad. Sci. USA 82, 5214-5218 11 Elliott, J.F., Lin, Y., Mizel, S.B., Bleackley, R.Ch., Harnish, D.G. and Paetkau, V. (1984) Science 226, 1439-1441 12 Kauffmann, Y., Chang, A.E., Robb, R.J. and Rosenberg, S.A. (1984) J. immunol. 133, 3107 3111 13 Handschumacher, R.E., Harding, M.W., Rice, J., Drugge, R.J. and Speicher, R.W. (1984) Science 226, 544-546 14 Zidovetzki, R., Yarden, Y., Schlessinger, J. and Jovin, T.M. (1981) Proc. Natl. Acad. Sci. USA 78, 6981-6985 15 Damjanovich, s., Tr6n, L., Sz6116si, J., Zidovetzki, R., Vaz, W.L.C., Regateiro, F., Arndt-Jovin, D.J. and Jovin, T.M. (1983) Proc. Natl. Acad. Sci. USA 80, 5985-5989 16 Tsien, R.Y., Pozzan, T. and Rink, T.J. (1982) Nature 295, 68-71 17 Shapiro, H.M. (1981) Cytometry 1, 301-312 18 Waggoner, A.S. (1979) Annu. Rev. Biophys. Bioeng. 8, 47-68

360 19 Bendtzen, K. and Danarello, Y. (1984) J. Immunol. 20, 43-51 20 Hess, A.D., Tutschka, P.J., Pu, Z. and Santos, G.W. (1982) J. Immunol. 128, 360-367 21 Miyawaki, T., Yachie, A., Ohzeki, S., Nagaoki, T. and Taniguchi, N. (1983) J. Immunol. 130, 2737-2742 22 Koponen, M,, Grieder, A. and Loor, F. (1984) Immunology 53, 55-62 23 Gr6f, P. and Belhgyi, J. (1983) Biochim. Biopbys. Acta 734, 319-328 24 Bendtzen, K. and Thomsen, M. (1984) IRCS Med. Sci. 12, 280-281 25 Rotman, B. and Papermaster, B.W. (1966) Proc. Natl. Acad. Sci. USA 55, 134-141 26 Sims, P.J., Waggoner, A.S., wang, C.H. and Hoffman, J.F. (1974) Biochemistry 13, 3315-3330 27 Hubbell, W.L. and McConnell, H.M. (1971) J. Am. Chem. Soc. 93, 314-326 28 Butterfield, D.A., Roses, A.D., Cooper, M.L., Appel, S.H. and Chesnut, D. (1979) Biochemistry 13, 5078 5082

29 Hess, A.D., Tutschka, P.J. and Santos, G.W. (1982) J. Immunol. 128, 355-359 30 Muraguchi, A., Butler, J.L., Kehnl, K.H., Falkoff, R.J. and Fauci, A.S. (1983) Am. J. Exp. Med. 158, 690-702 31 Roozemond, R.C. and Bonavida, B. (1985) J. Immunol. 134, 2209-2214 32 Bunjes, D., Hradt, M., Rollinghof, M, and Worgnes, H. (1981) Eur. J. Immunol. 11,657-661 33 Aszal6s, A. (1975) Antimicrob. Agents Chemother. 7, 754-757 34 Colombani, P.M., Robb, A. and Hess, A.D. (1985) Science 228, 337-339 35 Hippel, Von P.H. and McGhee, F.D. (1972) Annu. Rev. Biochem. 41, 231 300 36 L'Allemain, G., Franchi, A., Gragoe, E., Jr. and Pouyssegur, J. (1984) J. Biol. Chem. 259, 413-4319 37 L'Allemain, G., Paris, S. and Pouyssegur, J. (1984) J. Immunol. 132, 3064 3070