Differential effects of nordihydroguaiaretic acid (NDGA) on b-cell subsets: Reversal of NDGA-induced antibody suppression by cyclic GMP is subset specific

Int. J. lmmunopharmac., Vol. 6, No. 1, pp. 2 7 - 3 4 , 1984. Printed in Great Britain.

0192-0561/84 $3.00 + .00 © 1984 International Society for Immunopharmacology.

DIFFERENTIAL EFFECTS OF NORDIHYDROGUAIARETIC ACID (NDGA) ON B-CELL SUBSETS: REVERSAL OF NDGA-INDUCED ANTIBODY SUPPRESSION BY CYCLIC GMP IS SUBSET SPECIFIC JOANNA. WESS and DOUGLAS L. ARCHER* Division of Microbiology, Department of Health and Human Services, Public Health Service, Food and Drug Administration, 1090 Tusculum Avenue, Cincinnati, Ohio 45226, U.S.A. (Received 24 February 1983 and in final form 15 June 1983)

Abstract--Nordihydroguaiaretic acid (4,4'-[2,3-dimethyl tetramethylene]-dlpyrocatechol) (NDGA), a reportedly specific lipoxygenase inhibitor, suppressed the in vitro murine plaque-forming cell (PFC) response to a thymusdependent (TD) antigen, and the two subclasses of thymus-independent (TI) antigens, TI 1 and TI2, at a final concentration of 33/~M. Suppression kinetics were inconsistent with those observed in previous experiments for other lipoxygenase inhibitors, butylated hydroxyanisole (BHA) and butylated hydroxytoluene (BHT), in that BHA and BHT exert suppression early in the 5-day PFC culture system, whereas NDGA suppressed through the 96th h of the 120 h culture period. The sulfhydryl-protective agent 2-mercaptoethanol (2ME) protected both the TD and TI responses. Dibutyryl cyclic GMP (dbcGMP) failed to restore NDGA-suppressed TD and TI1 PFC responses but restored the TI2 response when added at 0-, 24-, 48-, 72-, and 96-h at concentrations of 1- 2 mM. NDGA inhibited lipopolysaccharide-(LPS-) induced increases in murine splenocyte cyclic GMP (cGMP) levels, and dbcGMP failed to accelerate the onset of the TI2 PFC response appreciably. The results of these and other laboratory studies indicated that NDGA may not be a specific inhibitor of lipoxygenase. Furthermore, the B-cell subset responding to TI2 antigens may be separated from the TD- and TIl-responding subsets because of the ability of dbcGMP to restore NDGA-suppressed TI2 responses but not the TD or TII response. The results suggest a fundamental difference in the biochemical pathways of B-cell subsets, and that cGMP metabolism in some cells may be linked to specific protein synthesis.

synthetase (Panganamala, Miller, Gwebu, Sharma & Cornwell, 1977). The result of lipoxygenase inhibition was a lowering of guanylate cyclase activity, resulting in an a b r o g a t i o n o f cyclic G M P ( c G M P ) accumulation. Lipoxygenase products of arachidonic acid have recently been implicated as playing a central role in tumor promotion (Valone, Obrist, Tarlin & Bast, 1983; Fischer, Mills & Slaga, 1982; Nakadate, Yamamoto, Ishii & Kato, 1982). Three other phenolic compounds were recently shown to inhibit the primary in vitro antibody response to a thymus-dependent (TD) antigen through a mechanism involving cGMP metabolism. These compounds were butylated hydroxyanisole (BHA) (Wess & Archer, 1981), butylated hydroxytoluene (BHT), and methyl paraben (MP) (Wess & Archer, 1982). The suppression of antibody synthesis caused by these three chemicals was freely reversed by the addition to cultures of exogenous cGMP or by elevating extracellular calcium levels in the medium. In the Wess & Archer study (1982), N D G A induced

N o r d i h y d r o g u a i a r e t i c acid ( 4 , 4 ' - [ 2 , 3 - d im e th y i tetramethylene]-dipyrocatechol) (NDGA) is a phenolic antioxidant derived from the creosote bush (Larrea divaricata, L). N D G A demonstrates antitumor properties in vivo and has been shown to inhibit malignant melanoma in man when applied topically (Smart, Hogle, Robins, Broom & Bartholomew, 1969). Blalock, Archer & Johnson (1981) in a study comparing the in vitro antitumor (anticellular) and immunosuppressive properties of various phenolic chemicals, demonstrated that N D G A inhibits clone formation of both malignant human and mouse cells. The anticellular effect and in vitro immunosuppressive effect of N D G A were manifest at similar doses of NDGA, thus indicating a link between these two biological systems. Recently N D G A has been used as a probe of the arachidonate system because of its reported abilities to inhibit lipoxygenase-catalyzed oxidation of arachidonic acid (Coffey, Hadden & Hadden, 1981; Kelley, Johnson & Parker, 1979) and prostaglandin

* Present address: FDA, 200 " C " ST., SW., Washington, DC, U.S.A. 27

28

JOANNA. WESSand DOUGLASL. ARCHER

suppression of antibody synthesis, but the suppression was not reversible by either cGMP or calcium. This result seemed inconsistent with N D G A being principally a lipoxygenase inhibitor. In light of these results, further studies, described herein, were initiated to determine the effects of N D G A on other aspects of humoral immunity, the thymus-independent (TI) immune responses. The latter have been further divided into the TI 1 and TI2 subclasses depending on the B-cell subset responding. The data suggest a differing role for cGMP in regulating various B-cell subset functions, or differential action of NDGA on the subsets. EXPERIMENTAL PROCEDURES

Animals Six- to 8-week-old female BDF1 (C57B1/6 × DBA/2) F~ mice (Harlan Labs., Indianapolis, IN) were used in this study. Animals were maintained on autoclavable mouse chow 5010C (Ralston Purina, St. Louis, MO) and acidified (pH 3) water, administered ad libitum. Materials NDGA (Sigma Chemical Co., St. Louis, MO) was administered by preparing the appropriate dilutions in absolute ethanol, applying the dose in 10/A volumes on 35 mm plastic culture dishes, and allowing the ethanol to evaporate before adding the spleen cells. N D G A was prepared fresh for each experiment. For kinetic studies N D G A was added to cultures in 10/al volumes of absolute ethanol. Appropriate ethanol controls were included for each time point. N2,O 2dibutyryl guanosine 3' :5'-cyclic monophosphoric acid (dbcGMP), N~,O z'-dibutyryl adenosine 3' :5'-cyclic monophosphoric acid (dbcAMP), guanosine 5 '-monophosphate (5' GMP), and 8-bromoguanosine 3' : 5 ' cyclic monophosphate (8 brcGMP) were added to 1 ml containing 1.5 × 107 BDF1 spleen cells. 2-mercaptoethanal (2ME), 99070 pure, was obtained from Matheson, Coleman, and Bell, Norwood, OH.

In vitro thymus-dependent PFC assays Dissociated mouse spleen cells were prepared and cultured as described by Mishell & Dutton (1967) at a concentration of 1.5 × 107 cells in 1 ml of culture. RPMI 1640 (Microbiological Associates, Bethesda, MD) to which were added sodium pyruvate (Microbiological Associates, Bethesda, MD) and fetal calf serum (FCS) (Gibco, Grand Island, New York, NY) to 10070, comprise the culture medium. Cultures were immunized with 3 × l06 sheep erythrocytes (SRBC) from a single sheep (Colorado Serum Co.,

Denver, CO). Numbers of plaque forming cells (PFC)/ culture were determined on day 5 of culture according to the method of Cunningham & Szenberg (1968). The number of cells/culture was counted in a hemacytometer, and viability was determined by trypan blue dye exclusion. Determination o f TI1 and TI2 PFC responses

Cultures prepared exactly as for the TD PFC response but not immunized with SRBC were immunized in vitro with 0.01 gg dinitrophenylaminoethyl-carboxylmethyl,o-Ficoll(DNP-FIC) (Biosearch, San Rafael, CA) for the TI2 response, and with 0.5 gg trinitrophenyl-lipopolysaccharide (TNP-LPS) (List Biological Labs, Inc., Campbell, CA) for the TIl response. Hapten-coupled sheep erythrocytes (SRBC) for the DNP-FIC and TNP-LPS hemolytic plaque assays were prepared by reacting SRBC and 2,4,6trinitrobenzenesulfonic acid (Eastman Kodak, Rochester, NY) under the conditions described by Rittenberg & Pratt (1969). PFC/culture was determined on day 5 of culture. The hemolytic plaque assays were carried out on microscope slides coated with 0.1070 agarose (Golub, Mishell, Weigle & Dutton, 1968). A 100 /al cell suspension was mixed with 0.5 ml of a 0.507o agarose solution and 100/al of hapten-coupled SRBC. Some plaque assays were performed according to the method of Cunningham & Szenberg (1968), but with slight modification; a final 1:40 dilution of guinea pig complement was used. Determination o f cGMP levels Mishell-Dutton cultures containing 1.5 × 107 BDF, spleen cells/ml were exposed to 33 gM NDGA (10/ag NDGA/1.5 x l07 cells) overnight at 37°C on a rocker platform. Cultures for each treatment group were treated in triplicate. Spleen cells were then resuspended in 0.85°7o NH4CI and incubated in cold for 5 min to lyse red cells. Spleen cells were then washed three times by successive centrifugations. LPS (25/ag/107 spleen cells) was added for 15 min (Watson, 1975) to the appropriate tubes. After the addition of an equal volume of 4°C, 0.15 M, phosphate-buffered saline (PBS), cells were centrifuged for 10 min in the cold at 400 × g, washed three times by successive centrifugations, and finally adjusted to a density of 5 × l07 cells/ml in PBS. Samples were collected for the determination of cGMP levels by the method of Goldberg, Haddox, Stephenson, Glass & Moser (1978). Commercial radioimmunoassay kits for cGMP determinations were purchased from Amersham, Arlington Heights, IL. Assays were carried out as described in the kit.

NDGA Separation of B-cell Subsets

29

Table 1. Effect of addition of 33 laM NDGA added at various times to the anti-SRBC anti-DNP-FIC and anti-TNP-LPS PFC response anti_SRBC t Time added (h) 0 24 48 72 96 116

Direct PFC/culture* anti-DNP-FIC t

anti-TNP-LPS t

Ethanol control ~ 33/aM NDGA Ethanol control 33 laM NDGA Ethanol control 33 taM NDGA 567_+ 72 <25 9967+_ 321 633+_321 1273+_ 36 56+_35 958_+131 <25 6767+_ 645 300+_210 1077+_115 40+_27 625+-206 <25 5033+_1577 <100 673+_ 53 <10 608+- 45 16+_16 5433+_1069 466_+466 487+_ 57 <10 458+_197 133+_45 4200+_ 324 1267+_670 697_+193 247_+24 333+_121 467+_80 6033+_ 34 5233_+205 ND§ ND

* Means of triplicate cultures +_ SEM, background corrected. t PFC response determined by means of Cunningham & Szenberg (1968). Antigen-stimulated cultures to which 10/al of absolute ethanol was added at the time indicated. § ND = not done. RESULTS

Effect o f N D G A added at various times on the TD, TI1, and TI2 PFC responses Data presented in Table 1 summarize the effects of N D G A on the TD, TI1, and TI2 responses when N D G A was added at various times to culture dishes to a final concentration of 33/aM. Suppression of all three responses occurs when N D G A is added at 0 h (onset of culture), 24, 48, 72 and 96 h, but not when added 4 h before harvest (116 h data point not shown for TI2 response). These suppression kinetics differ from those of previously studied phenolic antioxidants ( B H A , B H T , and MP) which are also lipoxygenase inhibitors and exert inhibition only when added at 0, 24, and 48 h (Archer, Bukovic-Wess & Smith, 1977a,b; Kutz, Hindsdill & Weltman, 1980). The suppression caused by N D G A when added at 48 h of culture or later suggests that the B-cell is directly affected by N D G A , since macrophage processing and T-cell triggering (events definitely involved in some TI responses) should be completed before 48 h of culture (Dutton, 1976).

Effect o f 2 M E and dbcGMP on NDGA-suppressed TD, TI1, and TI2 PFC responses The NDGA-suppressed T D P F C response was restored by the addition of 2ME (a sulfhydrylprotective agent) at 0 and 24 h of culture, whereas only modest restoration was observed when 2ME was added at 48 h of culture (Table 2). Dithiothreitol (DTT) at 10-3 M could substitute for 2ME in P F C restoration (data not shown). Perhaps D T T operates via a mechanism different f r o m that of 2ME; however, Broome and Jeng (1973) demonstrated that D T T was far weaker than 2ME in terms of lymphoid cell growth promotion. The addition o f d b c G M P (1 - 2 mM) to NDGA-suppressed T D response cultures failed to

Table 2. The effect of 2ME, added at various times, on the NDGA-suppressed TD PFC response to SRBC Direct anti-SRBC PFC/culture* Time of 2ME addition (h) t Control (no NDGA) No 2ME 0 24 48 72 96

Experiment 1 8367_+875 <100 4133+584 3400+_ 58 233+_233 <100 <100

Experiment 2 15950+- 206 <100 16500!-_1681 8800+-1300 2766+_ 986 <100 ND *

* Means of triplicate cultures -+ SEM, background corrected. t 2ME added at time indicated to final concentration of 5 x 10-5 M. * ND = not determined.

reverse suppression in previous experiments (Wess & Archer, 1982). In those same experiments (Wess & Archer, 1982), the T D P F C suppressions caused by B H A , BHT, and M P were fully restored by dbcGMP. Data presented in Table 3 show that d b c G M P (1 - 2 mM) failed to restore the NDGA-suppressed TI1 P F C response, but 2ME, however, added at the onset of culture did protect the response from N D G A . In contrast, either 2ME or d b c G M P fully restores the NDGA-suppressed TI2 P F C response (Table 3); d b c A M P (0.4 mM) added to NDGA-suppressed TI2 cultures was without effect (data not shown). In other experiments with the TI2 response, 10 Mg of 8 brcGMP (21/aM) restored the NDGA-suppressed TI2 response by more than 8007o; 5' G M P , at concentrations ranging from 10/ag/ml to 1 m g / m l (0.028 - 2.8 mM), had no restorative capacity (data not shown). In summary, 2ME protected TD, TI 1, and TI2 P F C responses from N D G A . In contrast, dbcGMP restored only the NDGA-suppressed TI2 P F C response but was

30

JOANNA. WESS and DOUGLAS L. ARCHER Table 3. Effect of 2ME and dbc GMP on NDGA-suppressed anti-DNP-FIC and anti-TNPLPS PFC response Direct PFC/culture t anti-DNP-FIC ¢ anti-TNP-LPS Exp. 1 Exp. 2 Exp. 1 Exp. 2

Culture content* Control§ 33 taM NDGA 33 gM NDGA + lmM dbc GMP 33/aM NDGA + 2mM dbc GMP 33/aM NDGA + 2ME (10-SM) * t # § [I

575+_150 <25 950-+197 988_+188 808_+30

400± 76 <25 200_+64 467_+194 NDII

263_+ 13 <25 <25 <25 675_+175

517_+153 <25 <25 <25 ND

All materials added at time of antigen. Means of triplicate cultures +__SEM, background corrected. Modified Cunningham slide method. Antigen-stimulated cultures. Not done. Table 4. Effect of dbc GMP on onset of TI2 PFC response Direct anti-DNP-FIC PFC/culture* 24 h 48 h

Culture content Control t dbc GMP (1 mM) :~ dbc GMP (2 mM)

117+ 105 17_+ 17 67-+ 17

133__.133 92-+ 59 83_+ 51

* Means of triplicate cultures _+ SEM, background corrected. t Antigen-stimulated response. Added same time as antigen. Table 5. The effect of adding dbc GMP (to 1 mM) on the NDGA-suppressed TI2 (anti-DNP-FIC) PFC response Culture content and time of dbc GMP addition No NDGA 33 /aM NDGA 33 taM NDGA 33 /aM NDGA 33 /aM NDGA 33 /aM NDGA *

+ + + +

dbc dbc dbc dbc

GMP GMP GMP GMP

Time dbc GMP added (h)

Direct anti-DNP-FIC PFC/culture

NA* NA 0 48 96 116

1983+298 <100 492+113 900___ 73 1375-+290 213-+113

Not added,

w i t h o u t effect o n T D or TI1 responses. D b c G M P a l o n e did n o t elevate P F C responses o f u n s t i m u l a t e d cultures, n o r did it accelerate the onset o f the TI2 P F C response (Table 4). D a t a in T a b l e 5 show t h a t I m M o f d b c G M P c a n at least partially reverse the suppression o f TI2 P F C r e s p o n s e caused by N D G A (33 taM) w h e n it is a d d e d to cultures at 0, 48, or 96 h, b u t n o t w h e n it is a d d e d 4 h b e f o r e harvest (116 h). T h e m o d e s t r e s t o r a t i o n o b s e r v e d at the 0 a n d 48 h a d d i t i o n times is likely to be due to the c o n c e n t r a t i o n o f d b c G M P used (1 mM). This c o n c e n t r a t i o n yielded variable r e s t o r a t i o n at 0 h (Table 3), whereas 2 m M gave full r e s t o r a t i o n reproducibly at 0 h (Table 3). T h e reasons for this result are unclear at present.

Effect o f NDGA on cGMP accumulation in LPSstimulated splenocytes N D G A is r e p o r t e d to be a n effective lipoxygenase i n h i b i t o r (Tappel & M a r r , 1954) a n d as such prevents g u a n y l a t e cyclase triggering by the hydroxy- a n d h y d r o p e r o x y - d e r i v a t i v e s o f a r a c h i d o n a t e . Direct m e a s u r e m e n t s of c G M P levels in N D G A - t r e a t e d splenocytes ( b o t h resting a n d LPS-activated) were t h e r e f o r e made. Triplicate m e a s u r e m e n t o f c G M P levels in u n t r e a t e d (resting, n o N D G A ) , treated (resting, N D G A - e x p o s e d ) , activated (LPS-exposed), a n d NDGA-exposed, LPS-activated ceils indicated the following: (i) N D G A h a d n o effect o n c G M P levels in resting splenic lymphocytes (control resting value is 5.4 p m o l e s / 5 × 10 r ceils, whereas N D G A - t r e a t e d

NDGA Separation of B-cell Subsets value is 4.7 pmoles/5 x 10' cells) and (ii) N D G A completely abolished LPS-induced, increased levels of cGMP (LPS-stimulated value is 11 pmoles/5 x 10' cells, and the NDGA-treated, LPS-stimulated value is 4.5 pmoles/5 x 107. cells). The correlation coefficient of the standard curve was 0.962; standard error of the mean of individual replicates varied from 1 to 14O7o. DISCUSSION

Classification of antibody responses into TD and TI has provided a convenient system by which to explore mechanism of antigen triggering of B lymphoctyes. But recent evidence indicated that such a strict separation may not be possible and that a continuum of antigens, based on their degree of Tcell dependence, likely exists (Letvin, Benacerraf & Germain, 1981). Studies of the antigen TNP-Ficoll have focused attention on this classification problem. TNP-Ficoll can elicit an antibody response in congenitally athymic mice, but the presence of T cells dramatically increases such responses (Archer, Smith & Wess, 1978; Mond, Mongini, Sieckmann & Paul, 1980; Letvin et al., 1981). Likewise, the macrophage has been shown to be a participant accessory cell in the antibody response to TNP-Ficoll (Boswell, Sharrow & Singer, 1980; Chused, Kassan & Mosier, 1976). TNP-LPS has been shown to be essentially Tcell independent (Jacobs and Morrison, 1975) and much less macrophage-dependent than TNP-Ficoll (Chused et al., 1976). Further complicating such studies is the fact that B-ceU subsets responding to TNP-LPS (prototype TI1 antigen) and DNP-Ficoll (prototype TI2 antigen) are demonstrably different by several criteria. TI1 antigens such as TNP-LPS and Brucella abortus cell wall stimulate B-cells from neonatal and adult mice (Mosier, Mond & Goldings, 1972; Mosier, Mond, Zitron, Scher & Paul, 1977) as well as from C B A / N defective mice (Mond, Scher, Mosier, Blaese & Paul, 1978; Tittle & Rittenberg, 1980). The antibody response to TI1 antigens is characterized by no T-cell and little macrophage involvement (Mosier et al., 1972; Motta, Portnoi & T r u f f a - B a c h i , 1981), and responding B-cells demonstrate high density I-A and low surface IgD (Zitron, Mosier & Paul, 1977; Greenstein, Lord, Horan, Kappler & Marrack, 1981). TI2 responsive Bcells are activated by antigens such as TNP- (or DNP-) Ficoll (FIC), which stimulate a more mature population of B cells (Mond et al., 1980). TI2 responsive B cells differ further from TI1 responsive B cells in that their response to antigen is macrophagedependent (Chused et al., 1976; Boswell et al., 1980),

31

somewhat T-cell-dependent (Mond et al., 1980; Archer et al., 1978; Levin et al., 1981), and sensitive to cyclosporin A (Kunkel & Klaus, 1980). They also exhibit low density I-A and high surface IgD (Greenstein et al., 1981; Zitron et al., 1977). In discussing the effects of chemicals such as NDGA on the antibody response, the investigator must therefore be cognizant of the lack of rigid definitions for TDand especially TI-antibody responses. Data presented in Table 1 demonstrate that 10/ag N D G A (33/aM final concentration) added to cultures of BDF1 splenocytes immunized with SRBC, TNPLPS, or DNP-Ficoll, inhibits the resulting day-5 immune response when added at 0-, 24-, 48-, 72-, or 96-h of culture, but not when added 4 h before harvest (116 h) (not shown for DNP-Ficoll). Inhibition caused by the addition of N D G A at 48 h and later would suggest that the B-ceU is directly affected by NDGA, since macrophage processing and T-cell triggering (if involved in the TI responses) should be completed before 48 h of culture. Thus clonal expansion of specific antigen-triggered B-cells and/or antibody production by antigen-specific plasma cells were considered as possible targets of NDGA. The sulfyhdryl-protective agent 2ME protected the TD, TI1, and TI2 responses from NDGA, but only when added early in the culture period. The TD response was restored when 2ME was added at 0 and 24 h of culture (Table 2); the same 2ME rescue kinetics were observed for the NDGA-suppressed TI2 response (data not shown). In contrast, the NDGA-suppressed TI1 PFC response was restored by 2ME only when added at culture onset (0 h). Data concerning the effects of cGMP on NDGAsuppressed cultures point to other major differences in the TD, TI 1, and TI2 PFC responses. We previously established that NDGA-suppressed TD response could not be restored by exogenous dbcGMP (Wess & Archer, 1982). Similarly, dbcGMP failed to restore the NDGA-suppressed TI1 PFC response (Table 3). In contrast, the TI2 PFC response was fully restored by 1 - 2 mM dbcGMP (Table 3); in related experiments, 8 brcGMP (21/aM) also restored the TI2 response, but 5 ' G M P administered over a wide concentration range failed to reverse N D G A suppressions. The reasons for the high dosage ( 1 - 2 mM) requirement for dbcGMP relative to 8 brcGMP are unclear but consistent with previous findings (Wess & Archer, 1981). The dbcGMP appears to be able to reverse NDGA suppressions (NDGA added at 0 h) when added any time through 96 h of culture. In fact, the data suggest that the later it is added, the more complete the restoration. The dbcGMP does not simply act by accelerating the TI2

32

JOANNA. WESSand DOUGLASL. ARCHER

PFC response (Table 4). These data also show clearly that very little if any clonal expansion of TI2-responding B cells occurs during the first 48 h of culture. Taken together, the data suggest that in the TI2-antigen-respondingB-lymphocyte subpopulation, NDGA has little effect on clonal expansion but may exert its suppressive effect on protein (antibody) synthesis. The fact that dbcGMP added at 96 h of culture (Table 5) to NDGA-suppressed splenocytes effects full reversal lends strong support to this contention, since full clonal expansion cannot occur in the remaining 24-h period. Nor did dbcGMP demonstrate the ability to accelerate clonal expansion (Table 4). It is clear, however, that NDGA does prevent cGMP formation induced by LPS (this manuscript) and by the tumor promoter PMA (Coffey & Hadden, 1983). Recent studies on cGMP support the hypothesis that NDGA exerts its immunosuppressive effect by inhibiting antibody (protein) synthesis by TI2-antigenresponding B-lymphocyte population. Johnson, Archer & Torres (1982) recently showed that cGMP or its inducers could replace helper cells in the production of gamma interferon. Without helper cells (Lyt 1 +), the interferon-producing cells (Lyt 2 + ) failed to respond to mitogen by producing gamma interferon. If cultures were DNA-synthesis-arrested by mitomycin C, dbcGMP restored full production of protein (gamma interferon) but failed to restore DNA synthesis (Johnson et al., 1982). Ananthakrishnan, Coffey & Hadden (1981) have shown that cGMP and calcium, when added to isolated, resting human lymphocyte nuclei, cause an increase in RNA synthesis

through a modification of initiation a n d / o r elongation. Johnson et al. (1982) concluded from their study that cGMP could act as a second messenger in gamma interferon production. Ananthakrishnan et al. (1981) likewise present compelling evidence that cGMP (with calcium) is the second messenger in mitogen stimulation of lymphocytes, whereby it is produced at or near the cell membrane and ultimately delivers an intranuclear signal that induces RNA synthesis. From our data, and that of Johnson et al., (1982), and Ananthakrishnan et al., (1981), a picture emerges in which the major role of cGMP is related to RNA and ultimately protein synthesis; proliferation of cells in these systems seems to be an unrelated issue. Data presented herein indicate that NDGA has a relatively pure effect on cGMP metabolism in the TI2-antigen-responsiveB-cell subset; NDGA may have multiple targets in the TD- and TIl-responsive B-cell subsets. Numerous studies in immunology (and more recently tumor promotion) use NDGA as a specific lipoxygenase inhibitor, but in fact it is probably not. Other known targets of NDGA include mitochondrial electron transport (Bhuvaneswaran & Dakshinamurti, 1972) adenylate cyclase (Coffey & Hadden, 1983), and prostaglandin synthetase (Panganamala et al., 1977). Great care should thus be taken, then, when interpreting experimental results based on NDGA related inhibitions of various biological effects, with special attention given to dose. Further studies on the differential effects of NDGA on TD, TI1, and TI2 in vitro PFC responses should help to determine differences in the biochemical pathways of the various B-cell subsets.

REFERENCES

ANANTHAKRISHNAN,R., COFFEY, R. G. & HADDEN, J. W. (1981). Cyclic GMP and calcium in lymphocyte activation by phytohaemagglutinin. Lymph. Differ., 1, 183-196. ARCHER, D. L., Burovtc-WEss, J. A. & SMITH,B. G. (1977a). Inhibitory effect of an antioxidant, butylated hydroxyanisole, on the primary in vitro immune response. Proc. Soc. Exp. Biol. Med., 154, 289- 294. ARCHER, D. L., BUKOVIC-WEss,J. A. & SMITH, B. G. (1977b). Suppression of macrophage-dependent T-lymphocyte function(s) by gallic acid, a food additive metaholite. Proc. Soc. Exp. BioL Med., 156, 465- 469. ARCHER, D. L., SMITH, B. G. & WESS, J. A. (1978). Evidence of helper T-cell participation in the antibody response to thymus-dependent antigens. IRCS Med. Sci., 6, 76. BHUVANESWARAN,C. & DAKSHINAMURTI,K. (1972). Inhibition of electron and energy transfer in rat liver mitochondria by nordihydroguaiaretic acid. Biochem., 11, 85-91. BLALOCK,J. E., ARCHER,D. L. & JOHNSON, H. M. (1981). Anticellular and immunosuppressive activities of foodborne phenolic compounds. Proc. Soc. Exp. BioL Med., 167, 391 -393. BOSWELL, H. S., SHARROW, S. D. • SINGER, A. (1980). Role of accessory cells in B cell activation. I. Macrophage presentation of TNP-Ficoll: Evidence for macrophage-B cell interaction. J. Immunol., 124, 989-996. BROOME,J. D. & JENG,M. W. (1973). Promotion of replication in lymphoid cells by specific thiols and disulfides in vitro. J. Exp. Med., 138, 574- 590. CHUSED, T. M., KASSAN,A.S. & MOSIER, D. E. (1976). Macrophage requirement for the in vitro response to TNP-Ficoll: A thymic independent antigen. J. Immunol., 116, 1579- 1581.

NDGA Separation of B-cell Subsets

33

COFFEY, R. G. & HADDEN, J. W. (1983). Phorbol myristate acetate stimulation of lymphocyte guanylate cyclase and cyclic guanosine 3 ':5 '-monophosphate phosphodiesterase and reduction of adenylate cyclase. Cancer Res., 43, 150 - 158. COFFEY, R. G., HADDEN, E. M. & HADDEN, J. W. (1981). Phytohemagglutinin stimulation of guanylate cyclase in human lymphocytes. J. Biol. Chem., 256, 4418-4424. CUNNINGHAM, k . J. 8~ SZENBERG, A. (1968). Further improvements in the plaque technique for detecting single antibodyforming cells. Immunology, 14, 599-600. DUTTON, R. W. (1976). In Mitogens in Immunobiology. (Edited by J. J. Oppenheim and D. L. Rosenstreich) pp. 237 - 244. Academic Press, New York. FISCHER, S. M., MILLS, G. D. & SLAGA,T. J. (1982). Inhibition of mouse skin tumor promotion by several inhibitors of arachidonic acid metabolism. Carcinogenesis, 3, 1243 - 1245. GOLOBERG, N. D., HADDOX, M. K., STEPHENSON, J. H., GLASS, D. B. & MOSER, M. E. (1978). Redox modulation of splenic cell soluble guanylate cyclase activity; activation by hydrophilic and hydrophobic oxidants represented by ascorbic and dehydoascorbic acids, fatty acids, hydroperoxides, and prostaglandin endoperoxides. In Advances in Cyclic Nucleotide Research (Edited by George, W. J. and Ignarro, L. J.) pp. 101 - 129. Raven Press, New York. GOLUB, E. S., MlSHELL, R. E., WEIGLE,W. O. & DUTTON, R. W. (1968). A modification of the hemolytic plaque assay for use with protein antigens. J. Immunol., 100, 133- 137. GREENSTEIN, J. L., LORD, E. M., HOGAN, P., KAPPLER, J. W. & MARRACK, P. (1981). Functional subsets of B cells defined by quantitative differences in surface I-A. J. Immunol., 126, 2419- 2423. JACOBS, D. M. & MORRISON, D. C. (1975). Stimulation of a T-independent primary anti-hapten response in vitro by TNPlipopolysaccharide (TNP-LPS). J. Immunol., 114, 360-364. JOHNSON, H. M., ARCHER, D. L. & TORRES, B. A. (1982). Cyclic GMP as the second messenger in helper cell requirement for ).-interferon production. J. Immunol., 129, 2570-2572. KELLEY, J. P., JOHNSON, M. C. & PARKER, C. W. (1979). Effect of inhibitors of arachidonic acid metabolism on mitogenesis in human lymphocytes: possible role of thromboxanes and products of the lipoxygenase pathway. J. lmmunol., 122, 1563 - 1571. KUNKEL, A. & KLAUS, G. G. B. (1980). Selective effects of cyclosporin A on functional B cell subsets in mouse. J. Immunol., 125, 2526- 2531. KUTZ, S. A. H1NSDILL, R. D. & WELTMAN, D. J. (1980). Evaluation of chemicals for immunomodulatory effects using an in vitro antibody-producing system. Environ. Res., 22, 3 6 8 - 376. LETVlN, N. L., BENACERRAF, B. & GERMAIN, R. N. (1981). B-lymphocyte responses to trinitrophenyl-conjugated Ficoll: requirement for T-lymphocytes to trinitrophenyl-conjugated Ficoll: requirement for T-lymphocytes and IA-bearing adherent ceils. Proc. Natn. Acad. Sci., 78, 5113-5117. MISHELL, R. I. & DUTTON., R. W. (1967). Immunization of mouse spleen cell cultures from normal mice. J. Exp. Med., 126, 423 - 442. MOND, J. J., SCHER, I., MOSIER, D. E., BLAESE, M. & PAUL, W. E. (1978). T-independent responses in B-cell defective DBA/N mice to Brucella abortus and to trinitrophenyl (TNP) conjugates of Brucella abortus. Eur. J. Immunol,, 8, 4 5 9 - 463. MOND, J. J., MONGINI, P. K. A., SIECKMANN,D. & PAUL, W. E. (1980). Role of T-lymphocytes in response to TNP-AECMFicoll. J. Immunol., 128, 1066-1070. MOSIER, D. E., MOND, J. J. & GOLDINGS, E. A. (1972). The ontogeny of thymic independent antibody responses in vitro in normal mice and mice with an X-linked B cell-defect. J. Immunol., 119, 1874- 1878. MOSlER, D. E., MOND, J. J., ZITRON, I., SCrtER, I., & PAUL, W. E. (1977). Functional correlates of surface Ig expressions for T-independent antigen triggering of B cells. In The Immune System: Genes and the Cells (eds. Sercay, E. E., Herzenberg, L. A. and Fox, C. F.) pp. 699-706. Academic Press Inc., New York. MOTTA, I., PORTNO1, D. & TRUFFA-BACHI, P. (1981). Induction and differentiation of B memory cells by a thymusindependent antigen, trinitrophenylated lipopolysaccharide. Cell. Immunol., 57, 3 2 7 - 338. NAKADATE, T., YAMAMOTO, S., ISHII, M. & KATO, R. (1982). Inhibition of 12-0-tetradecanoylphorbol-13-acetate-induced epidermal ornithine decarboxylase activity by lipoxygenase inhibitors: possible role of product(s) of lipoxygenase pathway. Carcinogenesis, 3, 1411- 1414. PANGANAMALA,R. V., MILLER, J. S., GWEBU, E. T., SHARMA, H. M. & CORNWELL, D. G. (1977). Differential inhibitory effects of vitamin E and other antioxidants on prostaglandin synthetase plastelet aggregation and lipoxides. Prostaglandins, 14, 261 - 271. Rrr-mNBERG, M. B. & PRATt, K. L. (1969). Antinitrophenyl fiN-P) plaque assay. Primary response of Balb/c mice to soluble and particulate immunogen. Proc. Sue. Exp. Biol. Med., 132, 5 7 5 - 581. SMART, D. R., HOGLE, H. H., ROBINS, P. K., BROOM, A. D. & BARTHOLOMEW, D (1969). An interesting observation on nordihydroguaiaretic acid and a patient with malignant melanoma - - a preliminary report. Cancer Chemother. Rep., Part L $3, 147- 151. TAPPEL, A. L. & MARR, A. G. (1954). Antioxidants and enzymes. Effect of tocopherol, propylgallate, and nordihydroguaiaretic acid on enzymatic reactions. J. Agr. Food Chem., 2, 554-558.

34

JOANNA. WESS and DOUGLAS L. ARCHER

TIITLE, T. V. & RITTENBERG, M. R. (1980). IgG B memory cell subpopulations: Differences in susceptibility to stimulations by TI1 and TI2 antigens. J. Immunol., 124, 202-206. VALONE, F. H., OBRIST, R., TARLIN, N. & BAST, R. C., Jr. (1983). Enhanced arachidonic acid lipoxygenation by K562 cells stimulated with 12-0-tetradecanoylphorbol-13-acetate. Cancer Res., 43, 197-201. WATSON, J. (1975). The influence of intracellular levels of cyclic nucleotides on cell proliferation and the induction of antibody synthesis. J. Exp. Med., 145, 9 7 - 111. WESS, J. A. • ARCHER, D. L. (1981). Restoration by cyclic guanosine monosphosphate and extracellular calcium of butylated hydroxyanis01e-suppressed primary murine thymus-dependent antibody response. Immunopharmacology, 3, 361 - 366. WESS, J. A. & ARCHER, D. L. (1982). Evidence from in vitro routine immunologic assays that some phenolic food additives may function as antipromoters by lowering intracellular cyclic GMP levels. Proc. Soc. Exp. Biol. Med., 170, 427 - 430. ZITRON, I. M., MOStER, D. E. & PAUL, W. E. (1977). The role of surface lgD in the response to thymic independent antigens. J. Exp. Med., 146, 1707-1718.