Vasoactive intestinal peptide inhibits cytokine production in T lymphocytes through cAMP-dependent and cAMP-independent mechanisms

Regulatory Peptides 84 (1999) 55–67 www.elsevier.com / locate / regpep

Vasoactive intestinal peptide inhibits cytokine production in T lymphocytes through cAMP-dependent and cAMP-independent mechanisms Hong-Ying Wang a , Xiaoming Jiang a , Illana Gozes b , Mati Fridkin c , Douglas E. Brenneman d , a, Doina Ganea * a

Department of Biological Sciences, Rutgers University, 101 Warren St., Newark, NJ 07102, USA Department of Clinical Biochemistry, Sackler School of Medicine, Tel Aviv University, Tel Aviv 69978, Israel c Department of Organic Chemistry, Weizmann Institute of Science, Rehovot 76100, Israel d Section on Development and Molecular Pharmacology, Laboratory of Developmental Neurobiology, National Institute of Child Health and Human Development, NIH, Bethesda, MD 20895, USA b

Received 25 March 1999; received in revised form 7 July 1999; accepted 13 July 1999

Abstract Previous reports indicate that VIP and the structurally related peptide PACAP, inhibit IL-2 and IL-10 production in antigen-stimulated T lymphocytes. Intracellular cAMP elevation appears to be the primary transduction pathway involved. However, in the lower concentration range, an additional, cAMP-independent transduction pathway appears to mediate the VIP inhibition of cytokine production. Here, we address this question by using VIP agonists and antagonists which act through cAMP-dependent and -independent pathways. The antagonists based on the neurotensin-VIP hybrid molecule did not affect the inhibitory effect of VIP/ PACAP on IL-2 and IL-10 production, confirming that astrocytes and T lymphocytes express different receptors. A lipophilic antagonist with increased membrane permeability, partially reversed the inhibitory effect of VIP/ PACAP, forskolin, prostaglandin E2, and 8-bromo-cAMP without significantly affecting cAMP levels, suggesting that it acts downstream of cAMP. Two VIP agonists inhibit IL-2 and IL-10 production. One of the agonists increases cAMP, whereas the second one does not induce cAMP/ cGMP. Our results indicate that VIP inhibits cytokine production in stimulated CD4 1 T cells through two separate mechanisms, which involve both cAMP-dependent and cAMP-independent transduction pathways.  1999 Elsevier Science B.V. All rights reserved. Keywords: PACAP; VIP-agonists; -antagonists; IL-2; IL-10; Transduction pathways.

1. Introduction Neuroimmunomodulation is partially mediated through soluble products of the nervous system such as neuropeptides released within the lymphoid microenvironment. The vasoactive intestinal peptide (VIP), a 28 aminoacid neuropeptide is widely distributed both in the central nervous system and in the lymphoid organs. In the nervous system VIP functions as a regulator of neuronal survival and differentiation, and of embryonic growth [1,2]. The *Corresponding author. Tel.: 1 1-973-353-1162; fax: 1 1-973-3531007. E-mail address: [email protected] (D. Ganea)

neurotrophic action of VIP in the CNS appears to be mediated through the release of survival-promoting substances from astrocytes [3]. In contrast to its neuronal survival and differentiation-promoting activity, VIP inhibits the proliferation of antigen-stimulated T lymphocytes through its inhibitory effect on the production of IL-2, a major T cell growth factor [4]. Previous studies from our laboratory indicated that VIP inhibits the transcription of IL-2 and IL-10, and reduces IL-4 production posttranscriptionally [5–8]. Three types of VIP receptors were recently cloned from various sources. VIP binds to the PACAP-specific receptor (PAC1) with a thousand fold less affinity than the specific ligand PACAP, whereas both VIP and PACAP bind with

0167-0115 / 99 / $ – see front matter  1999 Elsevier Science B.V. All rights reserved. PII: S0167-0115( 99 )00068-3

56

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

equal affinity to the VPAC1 and VPAC2 receptors [9]. Although the nature of the receptors present on astrocytes which mediate the neurotrophic activity of VIP is not entirely elucidated, they appear to be different from those present on lymphocytes. A specific VIP antagonist, the Met-hybrid, which antagonizes the neurotrophic activity of VIP [10], does not reverse the inhibitory effect of VIP on lymphocyte proliferation [11], and does not affect cAMP induction in cells transfected with either VPAC1 or VPAC2 [12]. Using Met-hybrid as a prototype, Gozes et al. [13,14] developed several VIP agonists and antagonists, based primarily on lipophilization to augment the ability of VIP to penetrate biological membranes, and on the substitution of methionine at position 17 with norleucine to increase peptide stability against oxidation. The lipophilic agonist affects neuronal survival more potently than VIP, without affecting cAMP, suggesting that the VIP neurotrophic activity is cAMP-independent. Both VPAC1 and VPAC2 are expressed in lymphocytes, particularly in CD4 1 T cells [15–17]. Recent experiments from our laboratory indicate that naive CD4 1 T cells do not express PAC1, and that both VPAC1 and VPAC2 agonists inhibit IL-2 production [18]. This suggests that, VPAC1 and VPAC2, but not PAC1, mediate the inhibitory effect of VIP on IL-2 production. In terms of transduction pathway, both PACAP/ VIP receptors are coupled to G proteins which activate adenylate cyclase [9]. Various cAMP-elevating agents were reported to inhibit IL-2 production [19–21]. These observations suggest that VIP inhibits IL-2 production through a cAMP-dependent pathway. However, at physiological concentrations, VIP is a better IL-2 inhibitor than FK (a known cAMP inducer), although it induces less cAMP than FK [6]. These results suggest that an additional secondary messenger is involved in the inhibitory effect of VIP. In the present study we address this question by using the VIP agonists and antagonists developed for the neurotrophic action of VIP [13]. The results suggest that an additional transduction pathway beside cAMP is involved in the VIP inhibition of IL-2, and that, as expected, the VIP lymphocytic receptors are different from the astrocytic ones.

2. Material and methods

in DMSO at 100 mM. PGE2 was dissolved in 100% ethanol at a concentration of 10 mM, and 8-Br-cAMP and 8-Br-cGMP were dissolved in water at 100 mM. All reagents were stored at 2 208C. The VPAC1 agonist [R16]-chicken secretin was a generous gift from Dr. Patrick Robberecht (Universite Libre de Bruxelles, Bruxelles, Belgium) and the VPAC2 agonists RO 25-1553 and RO 25-1392 were generous gifts from Drs. Ann Welton and D.R. Bolin (Hoffmann-La Roche Inc., Nutley, NJ). Anti-murine CD3 monoclonal antibody (clone 1452C11) was purchased from Pharmingen (San Diego, CA) and stored at 48C. The monoclonal antibodies for the Elisa assays and the recombinant IL-2 and IL-10 standards were purchased from Pharmingen and stored at 48 and 2 808C, respectively.

2.2. VIP antagonists and agonists Antagonists. A hybrid molecule composed of neurotensin 6 – 11 -VIP7 – 28 called Met-hybrid (AVM) blocks VIP action on astroglia but not lymphocytes [10,11]. With AVM as a prototype, a further array of VIP antagonists were developed [13]. To stabilize the molecule against oxidation, methionine at position 17 in the VIP part of the AVM molecule was substituted with norleucine creating the Nle-hybrid (ANV). In addition, a fatty acyl moiety (stearic acid) was added at the N terminus to enhance membrane permeability, creating the stearyl-Nle-hybrid (SANV). Agonists. A strategy similar to the one described above was used to create VIP agonists. The substitution of methionine at position 17 with norleucine created the Nle-VIP (NV), significantly more stable than its natural counterpart, VIP [13]. The addition of a fatty acid moiety at the N terminus created the Stearyl-Nle-VIP (SNV) with higher membrane permeability [13]. The two synthetic peptides act as potent VIP agonists promoting neuronal survival in spinal cord cultures [13]. The nonlipophilic antagonists and agonist (AVM, ANV, and NV) were dissolved in 0.1 M acetic acid and the lipophilic antagonist and agonists (SANV and SNV) were dissolved in DMSO at a concentration of 10 23 M and stored at 2 808C.

2.1. Materials 2.3. Cells Peptides: VIP, PACAP-38, secretin, and the VPAC1 antagonist (Ac-Tyr 1 ,D-Phe 2 )GRF(1–29) (GRF-A) were purchased from American Peptide Co (Sunnyvale, CA), dissolved in phosphate-buffered saline at a stock concentration of 10 24 M and stored at 2 808C. Forskolin (FK), prostaglandin E2 (PGE2), 8-bromoadenosine-39:59cyclic monophosphate (8-Br-cAMP) and 8-bromoguanosine-39:59-cyclic monophosphate (8-Br-cGMP) were purchased from Sigma (St. Louis, MO). FK was dissolved

Female Balb / c mice (four to six weeks old purchased from Charles River Laboratories, Wilmington, MA) were sacrificed by cervical dislocation. The spleens were removed, and single cell suspensions were prepared in serum-free RPMI 1640 (Gibco, Grand Island, NY). After washing three times in serum-free RPMI-1640, the cells were resuspended in RPMI containing 10% fetal bovine serum (FBS, Atlanta Biologicals, Atlanta, GA).

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

2.4. Cell culture and stimulation Murine splenocytes (2 3 10 6 cells / ml) were cultured in 24-well tissue culture plates in RPMI medium (Gibco) with 10% fetal calf serum (Atlanta Biologicals), 2 mM L-glutamine, 100 U / ml penicillin, 10 mg / ml streptomycin (Gibco), and 5 3 10 25 M 2-mercaptoethanol (Sigma). The cells were preincubated with the VIP antagonists for 30 min at 378C, followed by the stimulation with 1 mg / ml anti-CD3 or 1 mg / ml ConA, and different concentrations of peptides or cAMP-elevating agents. Appropriate controls such as cells alone, cells with peptides, cells with stimulus were included. The cultures were maintained at 378C in a humidified incubator with 5% CO 2 .

2.5. Cytokine determination The amounts of IL-2 and IL-10 present in supernatants from stimulated cells were determined by Elisa assays developed by Pharmingen. Cell culture supernatants were harvested at 24 h for IL-2 and at 48 h for IL-10. The assays were performed according to the manufacturer’s recommended protocols. The Elisa assays were specific for murine IL-2 and IL-10 (did not cross-react with each other, with human IL-2 and IL-10, or with other murine cytokines such as GM-CSF, IFNg, IL-3, IL-4, IL-5, IL-6, and TNFa). The sensitivity limits were 5 pg / ml and 0.8 ng / ml for the IL-2 and IL-10 Elisa assays, respectively.

2.6. Measurement of cyclic AMP and cyclic GMP Splenocytes (5 3 10 6 cells / ml) were pretreated for 5 min with antagonists, followed by the addition of VIP in the presence of the cAMP phosphodiesterase inhibitor, 3-isobutyl-1-methyl-xanthine (IBMX, Sigma). At different times (as indicated in text) the reaction was stopped by brief centrifugation, followed by addition of 1ml of 100% ethanol. The ethanol phase was collected by centrifugation and the pellets were extracted once more with ethanol. The ethanol from the two extractions was pooled and dried under vacuum. The intracellular cAMP/ cGMP contents were determined by using cAMP/ cGMP radioimmunoassay kits purchased from Amersham (Arlington Heights, IL). Samples were assayed in duplicate, and the amount of cAMP/ cGMP was determined by comparison with a standard curve obtained with predetermined concentrations of cAMP/ cGMP. The assay systems have a sensitivity of 320 pg cAMP/ ml and 140 pg cGMP/ ml, respectively.

2.7. Statistical analysis Results are presented as mean6SEM of triplicate cultures. Analysis of variance (ANOVA) followed by the Fisher PLSD was used to estimate the significance of the results. Differences of p , 0.05 were considered significant.

57

3. Results

3.1. The lipophilic antagonist SANV reverses the inhibitory effect of VIP on IL-2 and IL-10 production We showed previously that VIP and PACAP-38 inhibit IL-2 and IL-10 production in anti-CD3 stimulated T cells in a dose-dependent and specific manner [5,8,22]. The identification of the receptors and transduction pathways involved in the inhibitory effect of VIP/ PACAP on cytokine production has been hampered by the lack of specific VIP antagonists. We investigated the effect of the three antagonists, AVM, ANV, and SANV, on the inhibition of IL-2 and IL-10 production by VIP and PACAP-38. AVM was previously shown to antagonize the effect of VIP on astrocytes, but not on lymphocytes. As expected, AVM did not affect the inhibition of either IL-2 or IL-10 by VIP/ PACAP (Fig. 1). A similar lack of effect was observed for ANV, supporting the previous conclusion that astrocytes and lymphocytes have different VIP receptors [11]. SANV on the other hand reverses the inhibition of IL-2 and IL-10 for both VIP and PACAP (Fig. 1), and the effect is dose-dependent (Fig. 2). Since T lymphocytes express both VPAC1 and VPAC2, we investigated the effect of SANV on VPAC1 and VPAC2 agonists. Porcine secretin [12] and especially [R16 ]chicken secretin (Patrick Robberecht, personal communication) act as VPAC1 agonists. SANV reverses the inhibitory effect of porcine secretin, and to a lesser extent of [R16] -chicken secretin on both IL-2 and IL-10 production (Fig. 3A and B). Two cyclic VIP analogs (RO 25-1392 and RO 251553) act as VPAC2 agonists [23,24]. In our experimental system both RO analogs act as VIP agonists in the inhibition of IL-2 and IL-10 production, and SANV reverses the inhibitory effect (Fig. 3C and D). SANV may bind to the receptors, preventing the binding of the VPAC1 and VPAC2 agonists, or it may penetrate the membrane due to its lipophilic character, and block a second messenger. Since both VPAC1 and VPAC2 use cAMP as a major transduction pathway, and since other cAMP-elevating agents also inhibit IL-2 production, we investigated whether SANV reverses the inhibition of IL-2 production by FK, PGE2, and of 8-Br-cAMP. Similar to its effect on VIP, SANV reverses the inhibitory effect of FK, PGE2 and 8-Br-cAMP (Fig. 4). These results suggest that SANV acts on a secondary messenger shared by VPAC1, VPAC2, and several cAMP inducing agents. The reversal of the 8-BrcAMP activity indicates that SANV acts on a messenger localized downstream from cAMP. In spinal cord cultures both AVM and ANV, but not SANV, inhibit cAMP induction by VIP [13]. We reported previously that VIP induces cAMP in spleen cells, with a maximum induction at 5 min [22]. Spleen cells pretreated with SANV were cultured in the presence of VIP, and the

58

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

Fig. 1. Effect of antagonists on VIP and PACAP inhibition of IL-2 and IL-10 production. Spleen cells (2 3 10 6 cells / ml) were preincubated with the antagonists SANV, ANV, or AVM (10 26 M) for 30 min at 378C, followed by the addition of anti-CD3 (1 mg / ml) with or without VIP or PACAP-38 (10 27 or 10 28 M). IL-2 (A) and IL-10 (B) were quantitated by Elisa in supernatants harvested 24 h and 48 h later, respectively. Results are presented as mean6SEM of triplicate cultures. * indicates p , 0.05 compared to anti-CD3 plus VIP or anti-CD3 plus PACAP group.

intracellular cAMP was quantitated at different times. VIP induces a six-seven fold increase in intracellular cAMP levels at 5 min, followed by a decrease at later times; however, even at 60 min the cAMP level in VIP treated

cultures is approximately 4 times higher than in controls (Fig. 5). SANV does not reduce the cAMP levels at 5 min (when maximum cAMP induction occurs), but appears to affect cAMP accumulation at later timepoints (Fig. 5). At

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

59

Fig. 2. Effect of different doses of antagonists on VIP inhibition of IL-2 production. Spleen cells (2 3 10 6 cells / ml) were preincubated with the antagonists SANV, ANV, or AVM (10 26 or 10 27 M) for 30 min at 378C, followed by the addition of anti-CD3 (1 mg / ml) with or without VIP (10 27 M). IL-2 was quantitated by Elisa in supernatants harvested 24 h later. Results are presented as mean6SEM of triplicate cultures. * indicates p , 0.05 compared to anti-CD3 plus VIP group.

the present time, the significance of this later reduction is not clear. In contrast, the VPAC1 antagonist GRF-A, used as control, reduces the cAMP levels at 5 min by approximately 50% (Fig. 5). The effect of AVM and ANV was assayed at the time of maximal cAMP induction. As expected, AVM and ANV had little if any effect on cAMP induction (Fig. 6), since these two hybrid molecules do not act as VIP antagonists on T lymphocytes.

3.2. The lipophilic agonist SNV inhibits IL-2 and IL-10 production in anti-CD3 stimulated T cells. The lack of correlation between cAMP levels and IL-2 inhibition for low concentrations of VIP and FK [22], suggests the existence of a second transduction pathway in the inhibitory action of VIP on IL-2 production. The two VIP agonists NV and SNV differ in their ability to induce cAMP in astrocytes [13]. Therefore, we investigated their effects on IL-2 and cAMP induction in lymphocyte cultures. Both NV and SNV inhibit IL-2 and IL-10 production, although slightly less efficient than VIP (Fig. 7). As expected, VIP and NV induce cAMP, whereas SNV does not (Fig. 8). These results suggest that, in addition to the well described cAMP-dependent pathway, a second cAMP-independent pathway may function in transducing the VIP signal in T lymphocytes. To further substantiate the possibility that two transduction pathways mediate the inhibitory effect of VIP on IL-2 production, we investigated whether SNV contributes to the inhibition of IL-2 exerted by either FK (a strict cAMP inducing agent) or VIP. SNV did not significantly affect the

VIP inhibitory effect, whereas it increases the inhibitory action of FK (Table 1). The effect was additive rather than synergistic, suggesting that the two pathways are not connected. Since in some cell types, VIP and PACAP induce cGMP in addition to cAMP [25,26], and since increases in intracellular cGMP inhibit cytokine production in activated T cells [27], we investigated the possibility that VIP/ PACAP inhibit IL-2 and IL-10 production via a cGMPdependent pathway. Stimulated and unstimulated spleen cells were treated with VIP, SNV, or 8-Br-cGMP, and the amounts of intracellular cGMP/ cAMP were determined. As expected, VIP, but not SNV, induced an increase in intracellular cAMP (Fig. 9A). However, neither VIP nor SNV induced detectable cGMP levels (Fig. 9B). The lack of cGMP induction in splenic lymphocytes by VIP or SNV argues against the involvement of a cGMP-dependent pathway in the inhibition of cytokine production.

3.3. Similar effect of SNV and SANV on anti-CD3 and ConA-stimulated T cells We reported previously that VIP inhibits IL-2 production in T cells stimulated through TCR (T cell receptor)-engagement, but not through a combined phorbol ester / calcium ionophore treatment [28]. Both anti-CD3 and ConA stimulation of T cells involve TCR-associated signals. Therefore, we compared the effects of the SNV and SANV on the IL-2 production in anti-CD3 and ConAstimulated cultures. In both cases, VIP, and to a lesser

60

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

Fig. 3. Effect of SANV on VPAC1 and VPAC2 agonists-induced inhibition of IL-2 and IL-10 production. Spleen cells (2x10 6 cells / ml) were preincubated with SANV (10 26 M) for 30 min at 378C, followed by the addition of anti-CD3 (1 mg / ml) with or without different concentrations of VPAC1 (A, B) or VPAC2 (C, D) agonists. IL-2 (A, C) and IL-10 (B, D) were quantitated by Elisa in supernatants harvested 24 h and 48 h later, respectively. Results are presented as mean6SEM of triplicate cultures. * indicates p , 0.05 compared to the corresponding no antagonist group.

degree SNV, inhibit IL-2 production, and SANV partially reverses the inhibitory effect (Fig. 10).

4. Discussion Previous reports indicated that VIP and the related neuropeptide PACAP affect the expression of certain immune cytokines, including IL-2 and IL-10 [5,8,22,29]. CD4 1 T cells, the source of many immune cytokines, express mRNA for at least two of the cloned VIP

receptors, i.e. VPAC1 and VPAC2 [16,28]. Both receptors appear to be involved in the inhibitory effect of VIP/ PACAP on IL-2 and IL-10 production [18]. Although there are some differences between the two receptors in terms of transduction pathways such as higher peptide concentrations required for cAMP induction through VPAC2 as compared to VPAC1 [12], and coupling of VPAC2 to a calcium-activated chloride channel [30], both receptors use cAMP as the major transduction pathway [30–32]. In lymphoid cells, similar to other cell types, VIP induces cAMP accumulation [5,22,29]. Furthermore, various

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

61

Fig. 4. Effect of SANV on the inhibition of IL-2 induced by cAMP-elevating agents. Spleen cells (2 3 10 6 cells / ml) were preincubated with SANV (10 26 M) for 30 min at 378C, followed by the addition of anti-CD3 (1 mg / ml) with or without different concentrations of VIP, FK, PGE2 or 8-Br-cAMP. IL-2 was quantitated by Elisa in supernatants harvested 24 h later. Results are presented as mean6SEM of triplicate cultures. * indicates p , 0.05 compared to the corresponding no antagonist group.

Fig. 5. Lack of effect of SANV on cAMP induction by VIP. Spleen cells (5 3 10 6 cells / ml) were pretreated with SANV or GRF-A (10 26 M) for 5 min, followed by the addition of 10 27 M VIP. Intracellular levels of cAMP were determined at 5, 15, 25 and 60 min as described in Methods. Results are presented as mean6SEM of triplicate cultures.

62

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

Fig. 6. Effect of VIP antagonists on cAMP induction by VIP. Spleen cells (5 3 10 6 cells / ml) were pretreated with SANV, ANV, or AVM (10 26 M) for 5 min, followed by the addition of 10 27 M VIP. Intracellular levels of cAMP were determined at 5 min as described in Methods. Results are presented as mean6SEM of triplicate cultures.

cAMP elevating agents inhibit IL-2 expression in T cells and thymocytes [19–21]. These observations suggest that cAMP functions as the second messenger for the inhibitory effect of VIP on IL-2 expression. The role of cAMP in IL-10 expression is more controversial. Although the transduction pathways involved in the modulation of IL-10 expression in macrophages and T lymphocytes are largely unknown, cAMP appears to have an opposite effect in the two cell types. cAMP inducing agents promote IL-10 production in macrophages, and inhibit IL-10 production in stimulated T cells [8]. For both IL-2 and IL-10 the question remains whether the inhibitory effect of VIP/ PACAP is exerted solely through cAMP. Indeed, when we previously compared the effects of VIP and FK in terms of cAMP induction and IL-2 inhibition, we observed that at concentrations which are physiologically relevant for VIP (10 29 M), the peptide induces less cAMP production than FK, while acting as a more potent IL-2 inhibitor [22]. This observation suggests that a second pathway, in addition to the one involving cAMP, may be involved in transducing the VIP signal. Here we address this question by using VIP agonists / antagonists which do not affect cAMP levels. Gozes et al. [10] developed a specific VIP antagonist, AVM, that interacts with the VIP receptors on astrocytes and spinal cord cells with an affinity 10 fold higher than VIP and antagonizes the neurotrophic activity of VIP. In contrast, the antagonist does not reverse the inhibitory effect of VIP on lymphocyte proliferation, and 1000 fold higher concentrations of AVM are required to displace VIP from lymphocytes [11]. These results suggest different functional VIP receptors on lymphocytes and astrocytes. Using AVM as a prototype, several other VIP antagonists were developed, including the nonlipophilic ANV and the

lipophilic SANV [13,14]. Both ANV and SANV inhibit the VIP neuronal survival activity, whereas only ANV inhibits cAMP induction in astrocytes [13]. When tested in lymphocytes, the AVM and ANV were ineffective in terms of reversing the inhibitory activity of VIP on IL-2 and IL-10 production, with SANV inducing a partial reversal. Compared to its effect on neuronal survival, SANV is much less potent in lymphocytes (the biologically active antagonist / VIP concentration ratio is 10–100 in lymphocyte cultures, compared to 0.001 in the spinal cord cultures). These results confirm the previous conclusion that lymphocytes and astrocytes express different functional VIP receptors [11]. The effect of SANV could be explained by two different mechanisms, a receptor-independent uptake due to its enhanced lipophilicity, or the existence of an astrocytic-type receptor on T lymphocytes. We favor the first mechanism, based on the lack of effect of AVM and ANV. However, a very low abundance of the astrocytictype receptor on T cells could also explain the lack of effect of AVM and ANV, and the reduced potency of SANV. The functional VIP receptors on astrocytes and spinal cord cells appear to be different from VPAC1 or 2 [12,33]. On the other hand, CD4 1 T lymphocytes express VPAC1 and VPAC2 mRNA [15,16,18,27], and both receptors appear to be involved in the inhibition of IL-2 and IL-10 production [18]. Both receptors use cAMP as a transduction pathway. Similar to astrocytes, SANV does not affect significantly the VIP-induced cAMP increases in lymphocytes. However SANV partially reverses the inhibitory effects of both VPAC1 and VPAC2 specific agonists and of various cAMP-elevating agents on IL-2 production. Since SANV also reverses the inhibitory effect of 8-Br-cAMP,

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

63

Fig. 7. Effect of VIP agonists on IL-2 and IL-10 production. Spleen cells (2 3 10 6 cells / ml) were stimulated with anti-CD3 (1 mg / ml) in the presence or absence of different concentrations of VIP, NV (Nle-VIP), or SNV (Stearyl-Nle-VIP). IL-2 (A) and IL-10 (B) were quantitated by Elisa in supernatants harvested 24 h and 48 h later, respectively. Results are presented as mean6SEM of triplicate cultures. * indicates p , 0.05 compared to the corresponding controls without peptides.

we propose that this antagonist interferes not with the induction of cAMP, but rather with a secondary message downstream of cAMP. Possible targets for SANV include several cAMP controlled molecules which function in the transcriptional activation of the IL-2 gene, such as the

p50 / RelA family [34–36], the fos / jun family [37,38], the ICER family [39], and NFAT [35]. Experiments from our laboratory suggest that VIP affects the expression of members of both ICER and jun family in anti-CD3 stimulated spleen T cells (unpublished observations).

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

64

Fig. 8. Effect of VIP agonists on cAMP induction. Spleen cells (5 3 10 6 cells / ml) were treated with VIP, NV (Nle-VIP), or SNV (Stearyl-Nle-VIP) (10 27 and 10 28 M). Intracellular levels of cAMP were determined at 5, 20, and 40 min as described in Methods. Results are presented as mean6SEM of triplicate cultures.

Table 1 Stearyl-Nle-VIP (SNV) potentiates the inhibitory effect of FK, but not of VIP a % IL-2 Inhibition Treatment

No SNV

No VIP or FK

0

1 SNV (10 28 M)

1 SNV (10 29 M)

5364.8

3662.4

VIP (10 29 M)

5265.5

6461.3 ( . 100)

5463.4 (88)

FK (10 28 M)

3562.0

7962.8 (88)

7561.5 (71)

1963.8

6560.9 (72)

4864.0 (55)

FK (10 29 M) a

6

Spleen cells (2 3 10 cells / ml) were stimulated with anti-CD3 (1 mg / ml) in the presence or absence of different concentrations of SNV plus VIP or FK. IL-2 contents in the 24 h cell culture supernatants were determined by Elisa. Numbers in parantheses indicate the result of an expected additive effect.

Based on the strategy used to create the VIP antagonists, Gozes et al. [13,14] synthesized two specific VIP agonists, the nonlipophilic NV and the lipophilic SNV. SNV is 100 fold more potent than VIP in promoting neuronal survival, and does not induce cAMP in astrocytes [13]. Thus the neurotrophic effect of VIP is probably cAMP-independent. In our experimental system, both NV and SNV inhibit IL-2 and IL-10 production, although 10 fold less efficiently than VIP. This is a surprising finding, since the agonists are more stable than VIP. However, this is reminiscent of our previous observations regarding the effect of another class of stable VIP agonists, the cyclic RO 25 analogs, which also inhibit cytokine production 10 fold less efficiently than VIP [22]. Since the RO 25 analogs bind solely to VPAC2, the concomitant binding of VIP to both VPAC1 and VPAC2 may explain the higher efficiency of the VIP.

A similar situation could exist for NV; at the present time we do not know whether NV binds to both PACAP/ VIP receptors. In contrast to VIP and NV which induce cAMP in lymphocytes, the SNV is inactive in this respect. This suggests, that a second pathway, in addition to cAMP, functions in the inhibitory effect of VIP on IL-2 gene expression in T lymphocytes. The nature of the second transduction pathway(s) involved in the transmission of the inhibitory signal of VIP on cytokine production remains to be determined. Although cAMP induction appears to be the major intracellular event associated with the three PACAP/ VIP receptors, in muscle cells and pinealocytes signaling through the PACAP/ VIP receptors has been also associated with increases in intracellular cGMP [24,25]. However, since neither VIP nor SNV induce detectable intracellular cGMP

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

65

Fig. 9. Effect of VIP and SNV on cAMP and cGMP induction. Spleen cells (5 3 10 6 cells / ml) were treated with 10 27 M VIP, SNV or 8-Br-cGMP in the presence or absence of anti-CD3 (1 mg / ml) for 30 min. Intracellular cAMP (A) and cGMP (B) levels were determined as described in Methods. Results are presented as mean6SEM of triplicate cultures. * indicates p , 0.05 compared to the corresponding anti-CD3 or medium controls.

in splenic lymphocytes, cGMP does not appear to be the second messenger involved. Other possible transduction pathways have been described. In spleen cells, VIP also activates nuclear protein kinase C [40]. Also, the high affinity astrocyte VIP receptors are associated with the

nuclear translocation of alpha and delta protein kinase C and Ca 11 mobilization [41], and in lung membranes VIP interacts directly with a membrane-bound calmodulin [42]. In conclusion, a novel non-cAMP-dependent pathway is suggested for the VIP inhibition of cytokine production in

66

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

Fig. 10. VIP, SNV and SANV affect in a similar manner anti-CD3 and ConA-stimulated T cells. Spleen cells (2 3 10 6 cells / ml) were stimulated with anti-CD3 (1 mg / ml) or ConA (1 mg / ml) in the presence or absence of VIP or SNV. When testing the effect of SANV, the cells were preincubated with SANV (10 26 M) for 30 min at 378C, followed by the addition of anti-CD3 (1 mg / ml) or ConA (1 mg / ml) and VIP (10 27 and 10 28 M). IL-2 was quantitated by Elisa. Results are presented as mean6SEM of triplicate cultures. For the SANV effect * indicates p , 0.05 compared to the corresponding controls without SANV.

T cells. The elucidation of the transduction pathways which mediate the effect of VIP/ PACAP in lymphocytes could contribute to the development of stable VIP/ PACAP analogs with specifically designed immunomodulatory functions.

Acknowledgements This work was supported in part by the Charles & Johanna Busch Biomedical Award (DG), and by the PHS grants MH49079-01A3, AI41786-01R21, and AI4178602(DG). We thank Dr. Patrick Robberecht (Universite Libre de Bruxelles, Belgium) for his generous gift of [R16 ]-chicken secretin, and Drs. Ann Welton and D.R. Bolin (Hoffmann-La Roche Inc., Nutley, NJ) for their gift of RO 25-1553 and RO 25-1392.

References [1] Gozes I, Brenneman DE. Neuropeptides as growth and differentiation factors in general and VIP in particular. Mol Neurosci 1993;4:1–9.

[2] Gressens P, Hill JM, Gozes I, Fridkin M. Brenneman, D.E. Growth factor function of vasoactive intestinal peptide in whole cultured mouse embryos. Nature 1993;362:155–8. [3] Brenneman DE, Phillips TM, Festoff BW, Gozes I. Identity of neurotrophic molecules released from astroglia by vasoactive intestinal peptide. Ann NY Acad Sci 1997;814:167–73. [4] Ganea D. Regulatory effects of vasoactive intestinal peptide on cytokine production in central and peripheral lymphoid organs. Adv Neuroimmunol 1996;6:61–74. [5] Sun L, Ganea D. VIP inhibits IL-2 and IL-4 production through different mechanisms in T cells activated via the TCR / CD3 complex. J Neuroimmunol 1993;48:59–70. [6] Tang H, Sun L, Xin Z, Ganea D. Downregulation of cytokine expression in murine lymphocytes by PACAP and VIP. Ann NY Acad Sci 1996;805:768–78. [7] Wang H-Y, Xin Z, Tang H, Ganea D. Vasoactive intestinal peptide inhibits IL-4 production in murine T cells by a post-transcriptional mechanism. J Immunol 1996;156:3243–53. [8] Martinez C, Delgado M, Gomariz RP, Ganea D. Vasoactive intestinal peptide (VIP) and pituitary adenylate cyclase activating polypeptide (PACAP38) inhibit IL10 production in murine T lymphocytes. J Immunol 1996;156:4128–36. [9] Rawlings SR, Hezareh M. Pituitary adenylate cyclase activating polypeptide (PACAP) and PACAP/ vasoactive intestinal polypeptide receptors: actions on the anterior pituitary gland. Endocrine Reviews 1996;17:4–29. [10] Gozes I, McCune SK, Jacobson L, Warren D, Moody TW, Fridkin M, Brenneman DE. An antagonist to vasoactive intestinal peptide

H.-Y. Wang et al. / Regulatory Peptides 84 (1999) 55 – 67

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19] [20]

[21]

[22]

[23]

[24]

[25]

[26]

affects cellular functions in the central nervous system. J Pharmacol Exp Therapeut 1991;257:959–66. Gozes Y, Brenneman DE, Fridkin M, Asofsky R, Gozes I. A VIP antagonist distinguishes VIP receptors on spinal cord cells and lymphocytes. Brain Res 1991;540:319–21. Usdin TR, Bonner TI, Mrzey E. Two receptors for vasoactive intestinal peptide with similar specificity and complementary distributions. Endocrinology 1994;135:2662–80. Gozes I, Lilling G, Glazer R, Ticher A, Ashkenazi IE, Davidson A, Rubinraut S, Fridkin M, Brenneman DE. Superactive lipophilic peptides discriminate multiple vasoactive intestinal peptide receptors. J Pharmacol Exp Therapeut 1995;273:161–6. Gozes I, Lilling G, Davidson A, Bardea A, Reshef A, Glazer R, Zamostiano R, Ashur-Fabian O, Ticher A, Ashkenazi IE, Moody TW, Rubinraut S, Fridkin M, Brenneman DE. Development of V.I.P. agonists and antagonists with tissue and receptor specificity: effects on behavioral maturation, sexual function, and the biological clock. Ann NY Acad Sci 1996;805:59–171. Gomariz RP, Garrido E, Leceta J, Martinez C, Abalo R, Delgado M. Gene expression of VIP receptor in rat lymphocytes. Biochem Biophys Res Commun 1994;203:599–1604. Delgado M, Martinez C, Johnson M, Gomariz RP, Ganea D. Differential expression of vasoactive intestinal peptide receptors 1 and 2 (VIP-R1 and VIP-R2) in murine lymphocytes. J Neuroimmunol 1996;68:27–38. Johnson M, McCormack R, Delgado M, Martinez C, Ganea D. Murine T lymphocytes express vasoactive intestinal peptide receptor 1 (VIP-R1) mRNA. J Neuroimmunol 1996;68:109–19. Jiang X, Wang H-Y, Yu J, Ganea D. VIP1 and VIP2 receptors but not PVR1 mediate the effect of VIP/ PACAP on cytokine production in T lymphocytes. Ann NY Acad Sci 1998;865:397–407. Novak TJ, Rothenberg EV. cAMP inhibits induction of IL-2 but not IL-4 in T cells. Proc Natl Acad Sci USA 1990;87:9353–8. Anastassiou ED, Paliogianni F, Balow JP, Yamada H, Boumpas DT. PGE2 and other cAMP elevating agents modulate IL-2 and IL-2Ra gene expression at multiple levels. J Immunol 1992;148:2845–52. Katamura K, Shintaku N, Yamauchi Y, Fukui T, Ohshima Y, Mayumi M, Furusho K. Prostaglandin E2 at priming of naive CD4 1 T cells inhibits acquisition of ability to produce IFNg and IL-2, but not IL-4 and IL-5. J Immunol 1995;155:4604–12. Tang H, Welton A, Ganea D. Neuropeptide regulation of cytokine expression. Effects of VIP and Ro 25-1553. J Interferon Cytokine Res 1995;15:993–1003. Gourlet P, Vertongen P, Vandermeers A, Vandermeers-Piret MC, Rathe J, De Neef P, Waelbroeck M, Robberecht P. The long acting vasoactive intestinal polypeptide agonist RO 25-1553 is highly selective of the VIP2 receptor subclass. Peptides 1997;18:403–8. Xia M, Sreedharan SP, Bolin DR, Gaufo GO, Goetzl EJ. Novel cyclic peptide agonist of high potency and selectivity for the type II vasoactive intestinal peptide receptor. J Pharmacol Exp Ther 1997;281:629–33. Schaad NC, Vanecek J, Rodriguez IR, Klein DC, Holtzclaw L, Russell JT. Vasoactive intestinal peptide elevates pinealocyte intracellular calcium concentrations by enhancing influx: evidence for involvement of a cGMP-dependent mechanism. Molec Pharmacol 1995;47:923–33. Murphy KS, Jin JG, Grider JR, Makhlouf GM. Characterization of PACAP receptors and signaling pathways in rabbit gastric muscle cells. Am J Physiol 1997;272:G1391–9.

67

[27] Bauer H, Jung T, Tsikas D, Stichtenoth DO, Frolich JC, Neumann C. Nitric oxide inhibits the secretion of T helper 1 and T helper 2 associated cytokines in activated human T cells. Immunology 1997;90:205–11. [28] Xin Z, Jiang X, Wang H-Y, Denny TN, Dittel BN, Ganea D. Effect of vasoactive intestinal peptide (VIP) on cytokine production and expression of VIP receptors in thymocyte subsets. Regul Pept 1997;72:41–54. [29] Boudard F, Bastide M. Inhibition of mouse T cell proliferation by CGRP and VIP. J Neurosci Res 1991;29:29–41. [30] Inagaki N, Yoshida H, Mizuta M, Mizuno N, Fujii Y, Gonoi T, Miyazaki J, Susumu S. Cloning and functional characterization of a third PACAP receptor subtype expressed in insulin-secreting cells. Proc Natl Acad Sci USA 1994;91:2679–83. [31] Sreedharan SP, Patel DR, Huang JX, Goetzl EJ. Cloning and functional expression of a human VIP receptor. Biochem Biophys Res Commun 1993;193:546–53. [32] Lutz EM, Sheward WJ, West KM, Morrow JA, Fink G, Harmar AJ. The VIP2 receptor: molecular characterization of a cDNA encoding a novel receptor for vasoactive intestinal peptide. FEBS Lett 1993;334:3–8. [33] Ashur-Fabian O, Giladi E, Brenneman DE, Gozes I. Identification of VIP/ PACAP receptors on rat astrocytes using antisense oligodeoxynucleotides. J Mol Neurosci 1997;9:211–22. [34] Chen D, Rothenberg EV. IL-2 transcription factors as molecular targets of cAMP inhibition: delayed inhibition kinetics and combinatorial transcription roles. J Exp Med 1994;179:931–42. [35] Tsuruta L, Lee H-J, Masuda ES, Koyano-Nakagawa N, Arai N, Arai K, Yokota T. cAMP inhibits expression of the IL-2 gene through the NFAT site, and transfection of NFAT cDNAs abrogates the sensitivity of EL-4 cells to cAMP. J Immunol 1995;154:5255–64. [36] Ho H-Y, Lee H-H, Lai M-Z. Overexpression of mitogen-activated protein kinase kinase reversed cAMP inhibition of NFkB in T cells. Eur J Immunol 1997;27:222–6. [37] Hsueh Y-P, Lai M-Z. C-jun N-terminal kinase but not mitogenactivated protein kinase is sensitive to cAMP inhibition in T lymphocytes. J Biol Chem 1995;270:18094–8. [38] Tamir A, Granot Y, Isakov N. Inhibition of T lymphocyte activation by cAMP is associated with downregulation of two parallel mitogen-activated protein kinase pathways, the extraacellular signalrelated kinase and c-Jun N-terminal kinase. J Immunol 1996;157:1514–22. [39] Bodor J, Spetz A-L, Strominger JL, Habener JF. cAMP inducibility of transcriptional repressor ICER in developing and mature human T lymphocytes. Proc Natl Acad Sci USA 1998;93:3536–41. [40] Zorn N, Russell DH. Vasoactive intestinal peptide activation of nuclear protein kinase C in purified nuclei of rat splenocytes. Biochem Pharmacol 1990;40:2689–94. [41] Fatatis A, Holtzclaw LA, Avidor R, Brenneman DE, Russell JT. Vasoactive intestinal peptide increases intracellular calcium in astroglia: synergism with alpha-adrenergic receptors. Proc Natl Acad Sci USA 1994;91:2036–40. [42] Stallwood D, Brugger CH, Baggenstoss BA, Stemmer PM, Shiraga H, Landers DF, Paul S. Identity of a membrane-bound vasoactive intestinal peptide-binding protein with calmodulin. J Biol Chem 1992;267:19617–21.