Effects of different protein supplements on mitogen responses of human peripheral blood lymphocytes

Journal o f l m m u n o l o g i c a l M e t h o d s , 75 (1984) 339-349

339

Elsevier JIM03320

Effects of Different Protein Supplements on Mitogen Responses of Human Peripheral Blood Lymphocytes 1 Richard L. Rumley, Stanley W. Chapman 2, Marie L. Hoover and Marvin A. Cuchens Veterans Administration Medical Center, and Departments of Medicine and Microbiology, University o f Mississippi Medical Center, Jackson, M S 39216, U.S.A.

(Received 2 July 1984, accepted 10 September 1984)

We evaluated the effects of 6 supplements often used in human lymphocyte cultures, including fetal calf serum, autologous human serum, pooled human AB serum, hypogammaglobulinemic human serum, bovine serum albumin and human serum albumin. Lymphocyte proliferation of unstimulated and mitogen activated peripheral blood mononuclear cells was measured by [ 3H]thymidine incorporation. The responses of cells stimulated with the T-cell mitogen phytohemagglutinin-P were significantly lower when cultured in bovine serum albumin supplemented media, but were otherwise not supplement dependent. In contrast, responses of cells stimulated with the B-cell mitogens Cowan I strain of S. aureus and antisera against the bt or 8 chain of human immunoglobulin were significantly effected by supplement. Cultures containing fetal calf serum and bovine serum albumin had high background responses without a proportional rise in cellular proliferation when B-lymphocyte-specificmitogens were utilized. Autologous human serum and pooled human AB serum contained immunoglobulin which interacted with each of the B cell mitogens, thus limiting their usefulness as in vitro supplements. Cells grown in human serum albumin supplemented media had minimal background and high stimulated responses to B-cell mitogens. These results indicate that human serum albumin is an optimal supplement for in vitro human lymphocyte proliferative assays since it supports high stimulated cell responses with low background activity, is devoid of immunoglobulin and had minimal variability among lots.

Key words: peripheral blood mononuclear cells - T lymphocytes - B lymphocytes - lymphocyte mitogens cellular proliferation - serum-free media - human serum albumin

This work was supported by research funds of the Veterans Administration. 2 Address correspondence to Dr. Stanley W. Chapman, Chief, Medical Service, VA Medical Center, Jackson, MS 39216, U.S.A. Abbreviations: PBM, peripheral blood mononuclear cells; PHA, phytohemagglutinin-P; CISA, Cowan I strain of S. aureus; HSA, human serum albumin; FCS, fetal calf serum; BSA, bovine serum albumin; ABS, pooled AB human serum; AHS, autologous human serum; HHS, hypogammaglobulinemic human serum; Ig, immunoglobulin; SI, stimulation index; cpm, counts per minute; A cpm, background subtracted counts. 0022-1759/84/$03.00 © 1984 Elsevier Science Publishers B.V.

340

Introduction

In vitro assays of mitogen stimulated lymphocytes require certain macromolecules to maintain viability and support mitogen directed functions (Yachnin and Raymond, 1975). A variety of different sera or serum components have been utilized in human and other mammalian lymphocyte cultures (Cooperband et al., 1967; Alford, 1970; H~yry and Defendi, 1970). Many of the supplements utilized contain biologically active substances that can activate (Melchers and Andersson, 1974) or suppress (Shiigi et al., 1975) cellular function. Furthermore, some sera may also introduce heterologous components that interfere with the assay of the biological products (Smethurst and Williams, 1977), modulate mitogen activity (Banck, 1982; Weiler et al., 1982), or potentially expose laboratory workers to infectious agents, (CDC, 1983). Finally, variability among serum sources, suppliers, or even lots (Cooperband et al., 1967; Shiigi and Mishell, 1975) result in unpredictable effects of sera on lymphocyte in vitro assays. Such sera or media influences may produce major problems in in vitro studies of lymphocytes from humans with their heterogenous environments and genetic constitution (Holt et al., 1966; Ivanyi et al., 1973). These problems are magnified when mitogens specific for B-lymphocytes are employed in serum-supplemented cultures of human peripheral blood mononuclear cells (PBM) (Schuurman et al., 1980; Banck, 1982). Several studies have addressed these potential problems by employing serum-free or component defined media. For example, Polet and Spieker-Polet (1975) demonstrated that albumin (human or bovine) is necessary for in vitro mitogen activated human lymphocyte proliferation. Others have successfully used bovine serum albumin with essential salts, hormones, and nutrients to support human lymphocyte proliferation (Arai et al., 1977), antibody formation (Tanno et al., 1982), and mixed lymphocyte culture assays (Hsia et al., 1979). Bovine serum albumin is frequently employed as a supplement because of its availability. However, it contains tightly bound fatty acids, that are expensive to remove, and have been shown to modulate lymphocyte responses in vitro (Spieker-Polet and Polet, 1981). Only minor differences have been noted in the background and stimulated responses of cultures supplemented with human serum or human albumin using the T-cell mitogens concanavalin-A and phytohemagglutinin-P (PHA) (Polet and Spieker-Polet, 1975; Banck, 1982). However, B-cell mitogen induced responses have not been systematically evaluated employing different sera or protein supplements. In the light of the problems noted, our laboratory compared the effects of 6 culture supplements on the in vitro proliferative responses of mitogen activated human PBM. Lymphocyte proliferative responses were measured after stimulation with the T-cell mitogen PHA and the B-cell mitogens Cowan I strain of S. a u r e u s (CISA) and antisera against human 8 or # chain of immunoglobulin. Data is presented demonstrating that human serum albumin is an optimal supplement to support human PBM proliferation when stimulated with both T-cell and B-cell mitogens.

341

Materials and Methods

Blood was drawn from each donor after informed consent. The donors were normal, healthy laboratory workers free of underlying illness or medication ingestion for a week or more. Ages ranged from 26 to 37 years. This study was reviewed by the Internal Review Board of The University of Mississippi Medical Center.

Protein supplements Six protein supplements were used: human serum albumin (HSA; 25%, Alpha Therapeutic Corporation, Los Angeles, CA and Cutter Biological Laboratory, Berkley, CA): fetal calf serum (FCS; Flow Laboratories, McLean, VA); bovine serum albumin (BSA, Fraction V, not fatty acid free; Sigma Chemical Co., St. Louis, MO); pooled human AB serum (ABS; Flow Laboratories); autologous human serum (AHS drawn daily from each lymphocyte donor) and hypogammaglobulinemic human serum (HHS) obtained from a patient with common variable hypogammaglobulinemia. AHS was used the same day it was drawn while aliquots of HHS, ABS, and FCS were frozen to - 7 0 ° C until utilized. All sera were heated at 56°C for 30 min prior to use. Employing standard radial immunodiffusion techniques with commercial reagents (Helena Laboratories, Beaumont, TX) FCS, HSA, and BSA did not contain detectable human immunoglobulin (Ig). By the same method HHS had 10/xg/ml of IgD and 40/Lg/ml of IgM with no IgG or IgA detectable. Ig levels in AHS and ABS were not measured but assumed to be normal. Mitogens PHA (Difco Laboratories, Detroit, MI) was utilized as a T-cell mitogen. CISA (a gift of Arthur White, Indiana University, Indianapolis, IN) was prepared by formalin and heat treatment according to the method described by Schuurman et al. (1980), and was utilized as a B-cell mitogen. The antiserum preparations used were made from pooled sera from goats immunized with purified human myeloma IgM or IgD proteins prepared as previously described (Cuchens et al., 1978). The IgG fractions of the antisera were isolated using a protein A-Sepharose affinity column (Pharmacia Fine Chemicals, Piscataway, N J). The cross-reactive antibodies were removed by passage over appropriate immunoabsorbent columns conjugated with normal human serum, human y-globulin, IgG, IgM, IgD, or IgA, and the /~ or 8 chain specificity of the antisera was confirmed by immunodiffusion. The antisera were dialyzed versus phosphate-buffered saline and filter sterilized with a 0.45 ~m HATF filter (Millipore Corporation, Bedford, MA). The protein content of the anti-8 and anti-~ reagents was determined by ultraviolet spectrophotography (280 X wavelength) and stored in aliquots at - 7 0 ° C until utilized. Culture conditions The PBM were harvested from heparinized blood by density gradient centrifugation using Mono-Poly Resolving Medium (Flow Laboratories). Cells were washed 3 times and resuspended to 1 × 10 6 cells/ml in RPMI-1640 supplemented with

342 glutamine, penicillin 50 U / m l , streptomycin 50 ffg/ml and the appropriate protein supplements. All sera were used at a final concentration of 5%, a concentration empirically determined to be optimum. BSA or HSA were utilized at a final concentration of 1%. Higher concentrations of either albumin source inhibited lymphocyte proliferation. All supplemented media were prepared on the day of each experiment and filtered using a 0.22/~m GS filter (Millipore Corporation) prior to use in culture. PBM were cultured at 2 x 105 cells/well/0.2 ml of supplemented RPMI-1640 in round bottom microtiter plates (Flow Laboratories). Mitogen stimulated cultures received 10/~1 of the appropriate mitogen dilution while unstimulated cells received only medium. Triplicate or quadruplicate cultures were incubated at 37°C in a humidified 5% CO 2 atmosphere for 3, 5 or 7 days. The cultures were pulsed with 0.5 ffCi/well of [3H]thymidine (77.2 Ci/mmol; New England Nuclear, Boston, MA) 18 h prior to harvesting with an automated cell harvester (Brandel Biomedical, Gaithersburg, MA), and samples counted by liquid scintillation. Calculations The mean and standard error of the mean were calculated for each set of replicate wells and for each group of donor responses to added mitogen and supplement. A stimulation index (SI) was calculated using the following formula:

SI=

counts per minute (cpm) mitogen stimulated PBM cpm unstimulated PBM

Where applicable, the 2-tailed t-test was used to determine statistical significance.

Results Effects of serum supplement on unstimulated P B M cells As shown in Fig. 1, the mean background (unstimulated) [3H]thymidine incorporation of cells cultured in medium supplemented with HSA, HHS or AHS were not significantly different over time. Background responses of cells harvested on day 3 were not significantly different from each other regardless of the supplement used. However, when cultured for longer periods the mean background responses of PBM cultured in FCS, BSA and ABS supplemented media progressively increased. The responses of cells grown in these sera on days 5 and 7 were significantly greater ( P < 0.05) than those of cells cultured in HSA, HHS or AHS. Furthermore, cells cultured in ABS and FCS had greater variability in background cpm as indicated by the wide standard error bars. Although not depicted, variation in background stimulation was seen when different lots of FCS were utilized, not only among different donors, but also with PBM cells of the same donor performed on the same day with equivalent experimental conditions. In contrast, when different commercial sources or lots of HSA were compared in similar experiments, no differences in background responses or experimental variation were demonstrated.

343

30

20

CPM x I0 -s}

t

,o

357 HSA1%

357 357 357 357 357 HHS5% AHS5% BSAI% ABS5% FCS5% Serum Supplements Over Time (Days)

Fig. 1. Proliferative response of unstimulated human PBM cultured in media supplemented with serum or albumin. The results presented are derived from 7 donors. The mean cpm and standard error of the mean of D N A synthesis as measured by [3H]thymidine uptake at 3, 5 and 7 days of culture is shown.

14

J

12 10

.I. "-

X

L

8

+ I I

6

i ;

I 4

I

2

35 HSA 1%

35 lISA 1%

5 FCS 5%

3 5 HHS 5%

H 5

AHS 5%

3 5 ABS 5%

Serum Supplement with Time (days) Fig. 2. Effects of different sera or albumin sources on the maximum proliferative response of PHA stimulated PBM. Mean and standard error of 5 normal donors are presented. The shaded areas represent the mean background responses for these experiments. Results of 7 day cultures were always less than day 3 or 5 and are not shown.

344

Supplement effects on in vitro P H A stimulated cultures Peak responses of albumin supplemented cultures occurred at day 5 of culture (Fig. 2). FCS and AHS cultured cells had maximal peak responses occurring at day 3 while HHS and ABS cultured cells had essentially equivalent responses at day 3 and day 5. Proliferation of all cultures fell significantly at day 7 (data not shown). With the exception of BSA supplemented cultures, peak cellular responses with PHA stimulation were similar. Peak PHA responses of cells cultured in BSA were significantly lower compared to peak responses for FCS, HHS, AHS and ABS ( P < 0.05). PHA dose response studies were not supplement dependent and optimum stimulation was noted with 10 /~g per PHA per well. Similar to the results obtained with unstimulated cultures, less variability in the proliferative response was seen in HSA supplemented cell cultures while the greatest variability was noted in FCS cultured cells. In vitro effects of serum supplements on CISA stimulated cultures The peak response of CISA activated PBM occurred on day 7 of culture (Fig. 3). As noted in Fig. 4, the CISA stimulated responses of cells supplemented with HSA, BSA and FCS were dose related over the range of concentrations employed (106-108 bacteria/well). In contrast, when immunoglobulin containing AHS, ABS and HHS were utilized as culture supplements, this dose response was not evident with the same concentrations of bacteria. This modification of dose response implies mitogen supplement or cellular supplement interactions which may be due to immunoglobulin present in these sera. In support of this, utilizing higher concentrations of HHS, ABS and AHS resulted in a progressive decrease in cellular proliferation for all doses of CISA utilized (data not shown). 6 5 0

4

Z

2 1 57 HSA 1%

57 BSA 1%

57 FCS 5%

57 HHS 5%

57 AHS 5%

IN

7 ABS5%

Serum Supplement with Time (days) Fig. 3. The effects of protein supplements on time response of cultures stimulated with CISA. The mean and standard errors of 5 normal donors are presented at 5 and 7 days in culture. The responses noted at days 3 and 9 were always lower than day 5 or 7 and are not presented.

345

An increase in proliferative response over time was noted in unstimulated cultures utilizing FCS and ABS as supplements (Figs. 3 and 4). This background activity reduces background subtracted counts ( a cpm) a n d / o r SI for cells cultured in these 2 media. For example, A cpm and SI were significantly different when comparing cells cultured in HSA to those in FCS (A cpm: HSA = 30,600; FCS = 14,000; and SI: HSA = 18; FCS = 1.8). Although cells cultured in ABS had equivalent A cpm as HSA cultured cells (HSA = 30,600; ABS = 32,930), the higher background noted in ABS cultured cells resulted in a lower SI (HSA = 18; ABS = 3.9). Supplement activity on anti-I~ and anti-8 stimulated human cultures Due to the presence of high concentrations of immunoglobulin, ABS and AHS were not considered suitable for use in cultures stimulated with anti-immunoglobulins. In contrast, HSA, FCS and HHS were employed as supplements in anti-immunoglobulin stimulated cultures because they contain little or no Ig. The proliferative response of anti-immunoglobulin stimulated cultures increase over time with a peak response noted on day 7 (Fig. 5). The proliferative response of cells cultured for 9 days always fell below that on day 7 (data not shown). Although increased proliferative responses over time were observed in FCS supplemented cultures, most of this response is due to mitogenic effects of FCS and background subtracted responses are minimal or absent. The proliferation of PBM cultured in FCS containing media seemed to be suppressed when cultured with anti-& In contrast to the results observed for FCS, cells cultured in HSA had lower background stimulation resulting in increasing A cpm and SI. Responses in HHS cultured cells were intermediate to HSA and FCS. Although not depicted in this figure, cells cultured in

I

I ,-45-

3 IE

2

z//~

7//

!0 e 10 7 10 6

10 8 107 10 6

10 8 107 IO s

I0 ° 10 7 10 6

HSA1%

BSA 1%

FCS 5%

HHS 5%

10 ° 10 7 10 6

10 e 10' 106

AHS 5*

ABS 50

Serum Supplementand Mitogen Concentration[number/well) Fig. 4. Effects of protein supplements on dose response of CISA stimulated PBM. The mean and standard error of five normal donors are presented. Cells were cultured 7 days with the protein supplements noted and stimulated with multiple concentrations of CISA. Responses utilizing 109 CISA/well were always less than 108/well and are not shown.

346

anti-delta

antl-mu

30-

20

I0

357 lISA I%

357 FCS 5%

357 HHS 5%

357 HSA I%

357 FC$ 5%

357 HH$ 5%

Serum Supplements Over Time (Ooy$)

Fig. 5. Effects of protein supplements on the proliferative responses of human PBM stimulated with anti-human/~ and 8 chain reagents. Five donors were used for these experiments and the dose of anti-Ig used was over a 1000-fold range. The mean and standard error of the maximum proliferative response are presented. Time in culture and serum supplement are noted on the ordinate. H S A h a d p r e d i c t a b l e dose response curves with m a x i m a l s t i m u l a t i o n n o t e d at 90 / ~ g / m l of anti-8 a n d 55 / ~ g / m l of anti-/~. In contrast, dose response d a t a for cells s u p p l e m e n t e d with H H S a n d F C S were inconsistent.

Discussion L y m p h o c y t e cultures are s u p p l e m e n t e d with serum or other d e f i n e d p r o t e i n s to m a i n t a i n cellular viability. This s t u d y c o m p a r e s the effects of each of the serum s u p p l e m e n t s F C S , ABS, A H S , H H S , a n d the a l b u m i n s u p p l e m e n t s BSA a n d H S A on the in vitro proliferative responses of h u m a n P B M from several donors. T h e d a t a from these e x p e r i m e n t s d e m o n s t r a t e that the s e r u m or a l b u m i n source can influence cellular function of u n s t i m u l a t e d a n d s t i m u l a t e d cultures. T h e effects n o t e d can p r o d u c e conflicting conclusions when c o m p a r i n g the results of l y m p h o c y t e s cultured with these supplements. S u p p l e m e n t related differences in cell responses were most n o t a b l e when the B-cell mitogens C I S A a n d a n t i - i m m u n o g l o b u l i n reagents were utilized. Some supplem e n t s alone seemed to induce cellular p r o l i f e r a t i o n that could p o t e n t i a l l y m a s k B-cell proliferative responses. F u r t h e r , the i m m u n o g l o b u l i n c o n t a i n i n g s u p p l e m e n t s ABS, A H S and, to some degree, H H S were also n o t e d to m o d i f y responses to b o t h

347

CISA and anti-Ig stimulation. In contrast to the other supplements, lymphocytes cultured with HSA had low background stimulation and reproducible responses to the B-cell mitogens utilized. Additionally, the lot to lot and manufacturer variability noted for some of the supplements was not seen with HSA. The mitogenic properties of FCS have been previously described (Johnson and Russell, 1965). Potential mitogenic compounds in FCS include microbiological contamination (Shiigi and Mishell, 1975), prostaglandins (Okazaki et al., 1978), fatty acids (Kelly and Parker, 1979) and various other bioactive components (Leffert, 1974). Utilizing murine cells, other investigators (Hoffmann and Dutton, 1971; Braun and Unanue, 1980) have noted that certain albumin-like proteins in FCS may activate different lymphocyte populations in vitro, and that this effect is probably mediated by macrophages. In this regard, in vitro modulation of lymphocytes by monocytes has been noted for a variety of stimuli (Unanue, 1981). Although Spieker-Polet and Polet (1976) demonstrated that BSA may be utilized for in vitro culture of human PBM, they noted high background responses similar to our findings. These same authors subsequently noted that tightly bound lipids present in BSA produce these high background responses (Spieker-Polet and Polet, 1981). Removing the lipids from BSA and reconstituting with specific fatty acids can reverse the high backgrounds and promote stimulation in response to mitogens. However, BSA delipidation is expensive and may introduce other variables into the culture conditions. The elevated background counts seen in our lymphocyte cultures supplemented with ABS may be due to a variety of factors including the presence of immune complexes (Sinclair, 1978), complement components (Weiler et al., 1982) or the presence of blood group antigens (Mollison, 1983). In addition, immunoglobulin present in ABS and AHS may block or alter mitogen induced lymphocyte responses. Employing CISA as a B-cell mitogen Schuurman et al. (1980) demonstrated that lymphocyte responses were diminished in serum containing immunoglobulin. Similar results were noted by us in CISA stimulated cultures, as a predictable dose response was not seen with AHS and ABS. Also, utilizing higher concentrations of these immunoglobulin containing sera resulted in a proportional decrease in lymphocyte responses to optimal mitogen concentrations. In contrast, ceils cultured in HSA and HHS had a reproducible dose response to CISA. The presence of immunoglobulin in the culture media has been shown to suppress anti-lg stimulated lymphocyte proliferation (Kermani-Arab et al., 1977). In addition, immune complexes resulting from anti-Ig-immunoglobulin interactions may also modify lymphocyte responses (Sinclair, 1978). As such, AHS and ABS were not considered suitable for use in cultures stimulated by anti-Ig reagents. Although adequate for supplementation of CISA stimulated cultures, HHS, as employed in our studies, is not totally immunoglobulin free and the small amounts of IgM and IgD present may potentially interact with the appropriate antisera. This interaction may explain why HHS supplemented cell cultures had a reduction in proliferative response to the anti-immunoglobulin reagents when compared to cells cultured in HSA. The lack of readily available HHS from hypogammaglobulinemic patients and the variability of the serum immunoglobulin present hinders the general use of this

348 supplement. Further, depletion of immunoglobulin from normal human serum is both expensive and introduces further variables into the culture system. The results of our studies and those of other authors (Spieker-Polet and Polet, 1976; Banck, 1982) indicate that HSA is a good protein source for supporting in vitro human lymphocyte mitogen assays. HSA seems to support lymphocyte viability, is not mitogenic for resting lymphocytes and allows the expression of proliferative responses to both B- and T-cell mitogens. Furthermore, commercial HSA for in vivo use is free of pyrogen, immunoglobulin, insulin, hemoglobulin or transferrin. Since HSA is immunoglobulin free, mitogen supplement interactions with the B-cell mitogens CISA, anti-# and anti-8 are avoided. In addition, the low background responses over time of lymphocytes supplemented in HSA allow the reproducible measurement of the lower proliferative responses in cultures activated with anti-/~ or anti-6. Variability in cellular proliferation of anti-immunoglobulin stimulated cells, especially anti-6, has been noted (Chiorazzi, 1982). Some of this variability may be due to the antisera utilized or to differences in donor lymphocyte reactivity to the anti-immunoglobulin reagent. However, the supplement most often employed in these studies is FCS, which has been shown by our studies to significantly modulate the responses to anti-immunoglobulin reagents. Thus, some of the variability noted in prior studies may also result from the serum or protein supplements employed. In conclusion, although several supplements used in human lymphocyte proliferative assays seem to be adequate for the T-cell mitogen PHA, many of these supplements introduce factors that produce unacceptable experimental variation, mitogen incompatibility and cytological modulation, especially with the B-cell mitogens used in this study. Many of these problems can be avoided when human PBM are cultured in a defined media. Although further confirmation is required, HSA appears to be an optimum protein supplement for these cultures as it is devoid of immunoglobulin and is not itself mitogenic. In addition, HSA is widely available, relatively inexpensive, has a long shelf life, is species specific and has minimal lot or supplier variation.

Acknowledgements The excellent technical assistance of Brenda Chapman and the secretarial assistance of Jan Orr and Cynthia Pace are gratefully acknowledged.

References Alford, R.H., 1970, Proc. Soc. Exp. Biol. Med. 133, 1443. Arai, S., I. Yamane, Y. Tanno and T. Takishima, 1977, Proc. Soc. Exp. Biol. Med. 154, 444. Banck, G., 1982, J. Immunol. Methods 51,279. Braun, J. and E.R. Unanue, 1980, Immunol. Rev. 52, 3. CDC-United States, 1983, Morbidity and Mortality WeeklyReports 32, 101. Chiorazzi, N., 1982, in: Human B-LymphocyteFunction, Activation and Immunoregulation,eds. A.S. Fauci and R.E. Ballieux(Raven Press, New York) p. 47.

349 Cooperband, S.R., J.A. Green, M.A. Kennedy and M.M. Grant, 1967, Nature (London) 214, 1240. Cuchens, M.A., L.N. Martin and G.A. Leslie, 1978, J. Immunol. 121, 2257. H~yry, P. and V. Defendi, 1970, Clin. Exp. Immunol. 6, 345. Hoffmann, M. and R.W. Dutton, 1971, Science 172, 1047. Holt, P.J.L., N.R. Ling, G.C.B. Winter, C.F. McCarthy, A.E. Read and J.M. Yoffey, 1966, Lancet i, 980. Hsia, S., R.S. Wilkinson and D.B. Amos, 1979, J. Immunol. Methods 27, 383. Ivanyi, L., S.J. Challacombe and T. Lehner, 1973, Clin. Exp. Immunol. 14, 491. Johnson. G.J. and P.S. Russell, 1965, Nature (London) 208, 343. Kelly, J.P. and C.W. Parker, 1979, J. Immunol. 122, 1556. Kermani-Arab, Z., G.A. Leslie and D.R. Burger, 1977, Clin. Exp. Immunol. 30, 450. Leffert, H.L., 1974, J. Cell Biol. 62, 767. Melchers, F. and J. Andersson, 1974, Eur. J. Immunol. 4, 687. Mollison, P.L., 1983, Blood Transfusions in Clinical Medicine (Blackwell Scientific Publications, Oxford, England) p. 278. Okazaki, T., M. Shimizu, C.E. Arbesman and E. Middleton, 1978, Prostaglandins 15, 423. Polet, H. and H. Spieker-Polet, 1975, J. Exp. Med. 142, 949. Schuurman, R.K.B., E.W. Gelfand and H.M. Dosch, 1980, J. Immunol, 125, 820. Shiigi, S.M. and R.I. Mishell, 1975, J. Immunol. 115, 741. Shiigi, S.M., L.S. Fujikawa and R.I. Mishell, 1975, J. Immunol. 115, 745. Sinclair, N.R. StC., 1978, Transplant. Proc. 10, 349. Smethurst, M. and D.C. Williams, 1977, Prostaglandins 13, 719. Spieker-Polet, H. and H. Polet, 1976, J. Biol. Chem. 251,987. Spieker-Polet, H. and H. Polet, 1981, J. Immunol. 126, 949. Tanno, Y., S. Arai and T. Takishima, 1982, J. Immunol. Methods 52, 255. Unanue, E.R., 1981, Adv. Immunol. 31, 1. Weiler, J.M., Z.K. Ballas, T.L. Feldbush and B.W. Needleman, 1982, Immunology 45, 247. Yachnin, S. and J. Raymond, 1975, Clin. Exp. Immunol. 22, 153.