Introduction

Top Abstract Introduction Methods Results Discussion References

PGE₂ is a product of the cyclooxygenation of arachidonic acid, that potently mediates diverse physiological responses and contributes to the pathogenesis of inflammatory, autoimmune and neoplastic diseases (Goetzl et al., 1995). PGE₂ is recognized and transduces cellular effects specifically by interacting with PGE₂ Rs of at least four distinct subtypes, designated the EP₁, EP₂, EP₃ and EP₄ Rs. All subtypes of PGE₂ Rs have recently been cloned and shown to be members of the G protein-coupled seven transmembrane-domain superfamily, but differ in structure, ligand-binding properties, tissue distribution, potency of effects, and signal transduction pathways (Coleman et al., 1990, 1994). EP₃ Rs and EP₄ Rs are expressed ubiquitously, whereas EP₁ and EP₂ have a more limited range of tissue expression. EP₁ Rs mediate increases in the intracellular concentration of calcium ([Ca⁺⁺]_i). EP₂ and EP₄ Rs activate AC and stimulate increases in the intracellular concentration of cAMP ([cAMP]_i) (Coleman et al., 1990, 1994; An et al., 1993; Yang et al., 1994; Regan et al., 1994; Honda et al., 1993; Watabe et al., 1993). EP₃ Rs have multiple isoforms that not only inhibit AC, resulting in a decrease in [cAMP]_i elevated by forskolin or other agonists, but also stimulate increases in [Ca⁺⁺]_i (Coleman et al., 1990, 1994; Regan et al., 1994; Honda et al., 1993; An et al., 1994; Namba et al., 1993; Irie et al., 1993).

PGE₂ potently and highly selectively modulates the development and specific functions of many immune cells. PGE₂ enhances elements of macrophage differentiation, but inhibits functional activation of macrophages (Keller et al., 1991; Schreiber et al., 1990). PGE₂ regulates apoptotic elimination of immature B cells, enhances mature B cell production of IgG1 and IgE, but inhibits production of IgM (Garrone et al., 1994; Brown et al., 1992; Roper and Phipps, 1992). PGE₂ inhibits T cell proliferation, differentiation, expression of membrane Rs, secretion of diverse cytokines, cytotoxicity and other specific effector functions in cellular immune reactions (Goetzl et al., 1995a). Some effects of PGE₂ on T cells appeared to differ among subsets of T cells, such as stimulation of the proliferative and cytokine responses of CD8⁺ T cells and concurrent suppression of the responses of CD4⁺ T cells (Goodwin and Ceuppens, 1983). Although PGE₂ suppresses virtually all effector functions of T cells, our recent results revealed that PGE₂ also stimulates some activities, including expression and activation of endogenous MMPs (Zeng et al., 1996a, 1996b), MMP-dependent basement membrane transmigration (Leppert et al., 1995), and protection of some T cells from activation-induced apoptosis (Goetzl et al., 1995b). The specificity and diversity of PGE₂ immunoregulatory effects can be best explained by each subset of T cells or other immune cells expressing a characteristic pattern of different EP Rs with distinctive signaling mechanisms.

Human leukemic T cells of the HSB.2 cultured line are early "double-negative" thymocytes bearing CD2 and CD7, but not CD3, CD4 or CD8 (Adams et al., 1970; Hara et al., 1988). HSB.2 T cells express a mean of 7300 Rs per cell for PGE₂, of which the EP₃ and EP₂ subtypes predominate and the EP₄ subtype is present at a lower level (Zeng et al., 1996a). The coexpression of EP₃, EP₂ and EP₄ Rs by HSB.2 T cells has permitted us to study distinctive effects of each subtype of PGE₂ R on regulation of T cell functions. We have previously demonstrated that PGE₂ stimulates increases in matrix metalloproteinase (MMP)-9 production and secretion by HSB.2 T cells, that are transduced by the EP₃ Rs through [Ca⁺⁺]_i-dependent enhancement of transcription of MMP-9 mRNA (Zeng et al., 1996a). Using HSB.2 T cells that coexpress multiple EP Rs as a model cell line, we now examine possible functional roles of EP₂ Rs and/or EP₄ Rs. We find now that PGE₂ stimulates production of IL-6 by HSB.2 T cells through EP₄ R and/or EP₂ R transduction of [cAMP]_i-dependent increases in IL-6 mRNA. This is the first report of separate functional effects of EP₃ Rs and EP₂ and/or EP₄ Rs in the same T cells that coexpress multiple EP Rs. We also have found that the mitogenic lectin concanavalin A alters EP R frequencies in T cells in a pattern that augments PGE₂ stimulation of HSB.2 T cell IL-6 mRNA and protein.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials. M&B 28767 (Rhone-Poulenc Rorer Research, Dagenham, Essex), PGE₂ and sulprostone (Schering Pharmaceuticals, Berlin), misoprostol (Searle, Skokie, IL), butaprost (Bayer U.K. Ltd.), IBMX, ovalbumin, concanavalin A (Con A), actinomycin D, forskolin and dibutyryl-cAMP (Sigma Chemical, St. Louis, MO) were obtained from the designated sources.

Culture of HSB.2 T cells. Human leukemic T cells of the HSB.2 line (Adams et al., 1970; Hara et al., 1988) were obtained from American Type Culture Collection and cultured in RPMI-1640 medium (UCSF Cell Culture Facility) with 25 mM HEPES, 10% (v:v) fetal bovine serum (FBS, Hyclone Laboratories, Logan, UT), 100 U/ml of penicillin and 100 µg/ml of streptomycin (complete RPMI medium). Cultures were maintained at 37°C in a humidified atmosphere of 5% CO₂/95% air and medium was changed every 1 to 3 days to maintain a density of 0.4-1.8 × 10⁶ cells/ml.

ELISA quantification of IL-6 secretion by HSB.2 T cells. Replicate suspensions of 5 × 10⁷ HSB.2 cells were washed three times with protein-free RPMI-1640 and incubated in 10 ml of protein-free Iscove's: RPMI-1640 medium (1:1, v:v) with 10⁹ M to 10⁶ M PGE₂, synthetic agonists, or other agents at 37°C in 5% CO₂ - 95% air for up to 24 hr. Under these conditions there were no significant changes in the total number of HSB.2 T cells or the total amounts of secreted proteins (Zeng et al., 1996a). HSB.2 T cells also were stimulated with 5 µg/ml of concanavalin A (Con A) alone or with 10⁹ M to 10⁶ M PGE₂ for up to 24 hr. The suspensions then were centrifuged at 2000 × g for 5 min. The supernatants from suspensions of unstimulated HSB.2 cells and of Con A-stimulated HSB.2 cells were collected and concentrated by centrifugation through Microcon-3 columns (Amicon, Inc. Beverly, MA). Less than 5% of IL-6 was lost when diluted standard IL-6 was concurrently concentrated by the same method. IL-6 in concentrated supernatants of unstimulated and Con A-stimulated cells was quantified by ELISA, according to the manufacturer's protocol (ENDOGEN, Woburn, MA).

Reverse transcription-polymerase chain reaction. Replicate suspensions of HSB.2 T cells were washed and incubated at 5 × 10⁶ cells/ml in 5 ml protein-free Iscove's:RPMI-1640 medium (1:1, v:v) at 37°C for 24 hr. PGE₂ or other agents were added to the suspensions at 24 hr, 12 hr, 4 hr and 1 hr before harvesting suspensions of control and treated HSB.2 T cells for isolation of poly (A⁺) RNA by the Fast Track kit (Invitrogen, San Diego, CA) and total RNA by a TRIzol Reagent kit (GIBCO-BRL, Grand Island, NY). First-strand cDNAs were synthesized from HSB.2 T cell poly (A⁺) RNA or total RNA with oligo-dT primers and Superscript II reverse transcriptase (GibcoBRL), and were used as templates for PCR. PCR was carried out for 25-30 cycles of 95°C for 1 min, 58°C for 1 min, and 72°C for 1 min. The primers used are: IL-6 external primers of sequences: 5' (upstream)-CCAGGAGCCCAGCTATG and 5' (downstream)-CATTTGCCGAAGAGCC (Demuth et al., 1996). Five µl of the resultant RT-PCR products were amplified again in RT-PCR for 7-15 cycles under the same conditions as the first PCR, but with the IL-6 internal primers of sequences: 5' (upstream)-AACTCCTCCTCCACAAGCG and 5' (downstream)-TGGACTGCAGGAACTCCTT (Demuth et al., 1996). Both the external and internal primers for IL-6 included the intron-exon border to avoid amplification of genomic DNA. Analysis of EP R mRNA in control HSB.2 T cells and Con A-stimulated HSB.2 T cells was performed by RT-PCR using the same volume of cDNA for each subtype of PGE₂ R. The upstream primers were 5'-CTCGCCGCCCTGGTGTGCAACACGC for EP₁ R, 5'-TTCATCCGGCACGGGCGGACCGC for EP₂ R, 5'-TGTGTCGCGCAGTACCGGCG for EP₃ R, and 5'-CCTCCTGAGAAAGACAGTGCT for EP₄ R. The downstream primers were 5'-GGCCTCCCAGGCGCTCGGTGTTAGGCC for EP₁ R, 5'-GTCAGCCTGTTTACTGGC ATCTG for EP₂ R, 5'-CGGGCCACTGGACGGTGTACT for EP₃ R and 5'-AAGACACTCTCT GAGTCCT for EP₄ R (An et al., 1993, 1994; Regan et al., 1994; Funk et al., 1993). The constitutively-expressed gene encoding GAPDH was used as an internal control in RT-PCR to normalize the amounts of RNA in each sample. The cDNA of the GAPDH gene was amplified from the same volume of cDNAs as for IL-6 and the EP Rs, but only 10-12 cycles of PCR. The resultant RT-PCR products were resolved by electrophoresis in 2% agarose gels.

	Results

Top Abstract Introduction Methods Results Discussion References

Coexpression of the EP₃, EP₂, and EP₄ subtypes of PGE₂ R and Con A-induced alteration in the profile of EP Rs in HSB.2 T cells. The cultured line of human leukemic HSB.2 T cells coexpresses 7282 ± 1805 (mean ± S.E.) PGE₂ Rs/cell, that consist of the EP₂, EP₃ and EP₄ subtypes. Application of a series of synthetic agonists and antagonists, that bind preferentially to one or more subtypes of PGE₂ Rs to block binding of [³H]PGE₂ to HSB.2 T cells suggested predominant protein expression of EP₃ Rs, fewer EP₂ Rs and EP₄ Rs, and no EP₁ Rs (Zeng et al., 1996a). The presence of functional EP₂ Rs and EP₄ Rs was demonstrated by stimulation of increases in [cAMP]_i by PGE₂, misoprostol (EP₄/EP₂/EP₃ R-selective agonist) and butaprost (EP₂ R-specific agonist), whereas the presence of functional EP₃ Rs was indicated by increases in the [Ca⁺⁺]_i by PGE₂ and sulprostone (EP₃/EP₁ R-selective agonist) (Zeng et al., 1996a). Coexpression of mRNA encoding EP₃ Rs, EP₂ Rs and EP₄ Rs in unstimulated HSB.2 T cells was confirmed by RT-PCR analysis of poly (A⁺) RNA or total RNA from HSB.2 T cells (lane 1 in fig. 1). When HSB.2 T cells were stimulated with 5 µg/ml of the mitogenic lectin Con A for 1 to 24 hr, expression of mRNA encoding EP₃ Rs and EP₂ Rs were down-regulated by Con A in a time-dependent manner, whereas EP₄ R mRNA was maintained or up-regulated by Con A (fig. 1). mRNA encoding the EP₁ Rs was not detected in either unstimulated or Con A-stimulated HSB.2 T cells (data not shown).

View larger version (50K): [in this window] [in a new window]   Fig. 1.   Coexpression of the EP3, EP2, and EP4 Rs and time-course of Con A-induced alteration in the profile of EP Rs in HSB.2 T cells. HSB.2 T cells were incubated for 24 hr in medium (lane 1) and 5 µg/ml of Con A was added to the incubation at 1, 2, 4, 12, or 24 hr (in lanes 2-6, respectively) before harvesting the cells. RT-PCR analysis of the same poly (A+) RNA or total RNA with a probe specific for each EP R-subtype of human PGE2 Rs or GAPDH, a constitutively expressed mRNA were performed as described in Methods. The result presented was representative of those from three experiments.

PGE₂ stimulation of IL-6 secretion by unstimulated- and Con A-stimulated-HSB.2 T cells. PGE₂ enhanced secretion of immunoreactive IL-6 by unstimulated HSB.2 T cells in a time-dependent manner (fig. 2A). Since the basal levels of IL-6 secreted by unstimulated HSB.2 T cells were low, we chose to measure IL-6 cumulative secretion by HSB.2 T cells during 24 hr of incubation. At 24 hr, the secretion of IL-6 by HSB.2 T cells was increased significantly by 10⁷ M PGE₂ to a mean maximal level 8.8-fold higher than that of PGE₂-untreated controls. At 24 hr, IL-6 secretion was increased by PGE₂ in a concentration-dependent manner (fig. 2B), with significant mean enhancement to levels 5.0-, 9.3- and 13.2-fold higher than that of unstimulated HSB.2 cells by 10⁸ M, 10⁷ M, and 10⁶ M PGE₂, respectively (fig. 2B). Incubation of HSB.2 T cells with 5 µg/ml of Con A for 24 hr enhanced secretion of IL-6 to a mean 12-fold higher than that of HSB.2 T cells incubated in medium alone (fig. 2C). When HSB.2 T cells were concurrently stimulated with Con A and a range of concentrations of PGE₂ for 24 hour, IL-6 secretion was further enhanced 2.5-, 6.7-, 8.8-, and 10.2-fold by 10⁹, 10⁸, 10⁷, and 10⁶ M PGE₂, respectively, when compared with Con A-evoked IL-6 secretion (0.25 pg/ml culture medium) (fig. 2C). After 24 hr of Con A stimulation without and with PGE₂, more than 70% of HSB.2 T cells were still viable, as assessed by trypan blue exclusion.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Time-course and PGE2 concentration-dependence of stimulation of secretion of IL-6 by HSB.2 T cells without and with Con A-stimulation. A: Time-course of secretion of IL-6 in HSB.2 T cells by PGE2 alone. HSB.2 T cells were incubated with medium for 24 hr and 107 M PGE2 was added to the incubation at 1, 2, 4, 12, or 24 hr before harvesting the cell supernants. B: Dependence of IL-6 secretion on PGE2 concentration with a 24 hour incubation. Fold-increases in A and B were calculated relative to the level of IL-6 secretion with medium alone (fold-increase = 1.0). In the absence of PGE2, the IL-6 secretion by HSB.2 T cells was .02 pg/ml culture medium. C: PGE2 concentration-dependence of stimulation of secretion of IL-6 by Con A-stimulated HSB.2 T cells. HSB.2 T cells were exposed to 5 µg/ml of Con A for 24 hours and 107 M PGE2 was added to the incubation as described in A. "+"=with and "-"=without. The results presented in A, B and C represent the mean ± S.E. of the results from three or four experiments performed in duplicate.

PGE₂ enhancement of IL-6 mRNA in unstimulated- and Con A-stimulated-HSB.2 T cells. We then examined whether PGE₂ regulates IL-6 at a transcriptional level. Pretreatment of HSB.2 cells with 1 µg/ml of the cellular RNA synthesis inhibitor actinomycin D for 2 hr and continued incubation for 24 hr together with 10⁶ M PGE₂, resulted in a mean of 42% (n = 4) reduction in the PGE₂-evoked increases in IL-6 secretion compared to that elicited by PGE₂ without actinomycin D. Actinomycin D alone failed to change the basal level of IL-6 secretion by HSB.2 T cells (data not shown). Therefore, HSB.2 T cell IL-6 response to PGE₂ required de novo RNA synthesis.

8

6

7

View larger version (28K): [in this window] [in a new window]   Fig. 3.   PGE2-induced increases in IL-6 mRNA in HSB.2 T cells. A: Time-course of PGE2-induced increases in mRNA encoding IL-6. The samples are: medium control (lane 1) and 107 M PGE2 after incubation for 1 hr, 4 hr and 24 hr (lane 2-4, respectively). B: Actinomycin D inhibition of increases in IL-6 mRNA elicited by PGE2. HSB.2 T cells were pretreated without and with 1 µg/ml actinomycin D for 24 hr, and then stimulated with or without 107 M PGE2 for 1 additional hour. Lane 1-4 are medium control, PGE2, actinomycin D, PGE2 plus actinomycin D, respectively. C: Effect of Con A-induced changes in mRNA encoding IL-6. HSB.2 cells were incubated in medium (lane 1) or treated with 5 µg/ml of Con A for 1, 4, 12, and 24 hr in lanes 2-5, respectively. D: Effects of a combination of Con A and PGE2 on mRNA encoding IL-6. HSB.2 T cells were pretreated without and with 5 µg/ml Con A for 24 hour and then treated with or without 107 M PGE2 for 1 additional hour. Lane 1-4 are control, PGE2 plus Con A, PGE2, and Con A alone, respectively. In A-D, RT-PCR analysis were performed as described in Method. The results presented in A-D were representative of those from three experiments.

7

EP₄/EP₂ R mediation of PGE₂ stimulation of IL-6 secretion. We next determined which EP Rs mediate the effect of PGE₂ on IL-6 by employing agonists selective for one or more subtypes of EP R. PGE₂ and the EP₄/EP₂/EP₃ R-agonist misoprostol (fig. 4C) stimulated the secretion of IL-6 by HSB.2 T cells to a similar extent (fig. 4A), suggesting a dependence on EP₄, EP₂ and/or EP₃ Rs. The fact that two EP₃ R-directed agonists sulprostone and M&B 28767 (fig. 4C) did not increase the secretion of IL-6 by HSB.2 T cells (fig. 4A) ruled out the involvement of EP₃ Rs. PGE₂ stimulation of HSB.2 T cell secretion of IL-6 thus is mediated by EP₄ and/or EP₂ Rs. The fact that the EP₂ R-selective agent butaprost (fig. 4C) is a weak agonist, as assessed in its ability to elevate [cAMP]_i in HSB.2 T cells (Zeng et al., 1996a), may explain its failure to stimulate IL-6 secretion (fig. 4A). With a similar specificity for EP₄/EP₂ Rs, the increases in IL-6 mRNA evoked by PGE₂ were mimicked by the EP₄/EP₂/EP₃ R-agonist misoprostol, but not by the EP₃ R-directed agonists sulprostone and M&B 28767 (fig. 4B). That butaprost is a weak agonist for the EP₂ R may explain its failure in increasing IL-6 mRNA (fig. 4B). Thus increases in both IL-6 message expression and protein secretion by PGE₂ are dependent on EP₄ Rs and/or EP₂ Rs.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   PGE2 R subtype-specificity of transduction of stimulation of IL-6 secretion (A) and mRNA (B) by HSB.2 T cells. A: HSB.2 were incubated for 24 hr with medium, 107 M PGE2, sulprostone or M&B 28767 (EP3 R-directed), and misoprostol (EP4/EP2/EP3-selective) or 106 M butaprost (EP2 R-selective). Fold-increase calculated and basal IL-6 secretion were similar to that described in Fig. 2B. Each bar and bracket represents the mean ± S.E. of the results of three experiments performed in duplicate. The levels of statistical significance were determined by a Student's t test for each value relative to control, *= P < .01. B: HSB.2 T cells were incubated in serum-free medium for 24 hours and stimulated with 107 M PGE2 or the subtype-selective agonists for 1 hr. RT-PCR analysis of poly (A+) RNA with probes specific for IL-6 or GAPDH were carried out as described in Method. The samples from lane 1-6 were: control, 107 M PGE2, sulprostone, M&B 28767, misoprostol, and 106 M butaprost. The results were representative of those from three experiments. C. Structures of PGE2 analogues selective for each EP R used in the study.

Involvement of [cAMP]_i and protein kinase A-dependent mechanisms in the effects of PGE₂ on IL-6 secretion. We next examined the signal transduction from the EP Rs that mediate the PGE₂ effects of IL-6 secretion by HSB.2 T cells. Dibutyryl-cAMP (a permeant cAMP analogue), forskolin (a direct activator of adenylyl cyclase) and IBMX (a cAMP phosphodiesterase inhibitor) are all capable of increasing cellular levels of cAMP. Optimal concentrations of dibutyryl-cAMP, forskolin and IBMX each increased the secretion of IL-6 by HSB.2 T cells to levels similar in magnitude to that elicited by 10⁷ M PGE₂ at 24 hr (fig. 5A). Simultaneous addition of PGE₂ and either dibutyryl-cAMP, forskolin or IBMX resulted in further enhancement of IL-6 secretion when compared with that elicited by PGE₂ alone (fig. 5A). Thus an increase in [cAMP]i is one second messenger mediating IL-6 responses of HSB.2 T cells. Again, this supports the dependence of PGE₂-evoked increases in IL-6 secretion on EP₄ and/or EP₂ Rs that are functionally coupled to increases in [cAMP]_i.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Involvement of [cAMP]i and protein kinase A on IL-6 secretion evoked by PGE2. A: [cAMP]i mediation of PGE2 effect on IL-6 secretion. HSB.2 T cells were incubated for 24 hours without (Frame Left) and with (Frame Right) 107 M PGE2 alone or with IBMX, dibutyryl-cAMP ((Bu)2cAMP) and forskolin (FSK) at respective final concentrations of 1 µM, 10 mM, and 10 µM. B: Protein kinase A requirement for PGE2 to increase IL-6 secretion. HSB.2 T cells were cotreated with 107 M PGE2 and a PKA inhibitor KT5720 or PKC inhibitor calphostin C (Cal C) at 106 M for 24 hr. In s A and B, fold-increases calculated and basal IL-6 secretion with buffer alone were similar to that described in Fig. 2. The levels of statistical significance were determined by a Student's t test for each value relative to control, + = P < .05, * = P < .01. However, **= P < .01 when compared to the PGE2 -stimulated level of IL-6 secretion.

6

6

	Discussion

Top Abstract Introduction Methods Results Discussion References

Prostaglandin E₂ has potent effects on T cell differentiation, proliferation, survival, adhesion, migration, expression of membrane Rs and diverse synthetic responses to antigens and mitogens (Goetzl et al., 1995a). Expression of a characteristic pattern of different subtypes of EP Rs by each subset of T cells is the principal determinant of specificity of the response of those T cells to PGE₂. Aspects of ligand specificity and transductional biochemical pathways have been partially defined for each of the subtypes of EP Rs expressed by T cells. The HSB.2 T cell is a "double-negative" thymocyte (Adams et al., 1970; Hara et al., 1988), that coexpresses a total of 7000 EP₃, EP₂ and EP₄ Rs per cell with a K_d of 3.7 nM (Zeng et al., 1996a). The results of ligand binding studies (Zeng et al., 1996a) and RT-PCR analyses (fig. 1) suggested that the levels of expression of EP R mRNA and protein by unstimulated HSB.2 T cells are in the order of prevalence of EP₃ Rs = EP₂ Rs >EP₄ Rs, without EP₁ Rs. The presence of functional EP₃ Rs or EP₂/EP₄ Rs was demonstrated by respective increases in [Ca⁺⁺]_i and [cAMP]_i by PGE₂ and EP R-selective agonists (Zeng et al., 1996a). The EP₃ Rs of HSB.2 T cells have been shown to mediate PGE₂ stimulation of MMP-9 production and secretion by a Ca⁺⁺-dependent mechanism (Zeng et al., 1996a). HSB.2 T cells do not produce IL-2 constitutively (Kasahara et al., 1985), which was also confirmed in our study. IL-6 is a multifunctional cytokine with many immunological and inflammatory activities, including regulation of differentiation and functions of B and T lymphocytes (Van Snick 1990). Although PGE₂ has been shown to modify IL-6 release from T lymphocytes (Della Bella et al., 1997), the EP R-specificity and transductional mechanisms have not been elucidated for defined sets of T cells. Coexpression of EP₃ Rs, EP₂ Rs and EP₄ Rs by HSB.2 early T cells allowed us now to examine their potentially important roles in regulating IL-6 secretion.

HSB.2 T cell secretion of IL-6 was stimulated in a time-dependent and concentration-dependent manner and increased up to 13-fold by 10⁸ M - 10⁶ M PGE₂ (fig. 2). Dibutyryl-cAMP, forskolin and IBMX that increase the [cAMP]i in HSB.2 T cells by different mechanisms, mimicked PGE₂ in enhancing secretion of IL-6 from HSB.2 T cells (fig. 5), which supported a cAMP-dependent mechanism. The combination of PGE₂ and dibutyryl-cAMP, forskolin or IBMX resulted in a greater stimulation of IL-6 secretion than either type of agonist alone, suggesting mechanisms of additive mediation. These observations are consistent with previous findings that PGE₂ effects on IL-6 production by macrophages or astrocytoma cells are mediated by cAMP (Hinson et al., 1996; Fiebich et al., 1997). The cAMP-dependence of PGE₂ enhancement of IL-6 secretion by HSB.2 T cells suggests involvement of EP₄ Rs and/or EP₂ Rs. The predominant role of EP₄ Rs and/or EP₂ Rs was confirmed by finding that the EP₄/EP₂/EP₃ R-selective agonist misoprostol (IC₅₀=44 ± 1.3 nM) but not the EP₃ R-preferential agonists sulprostone (IC₅₀=2.3 ± .6 nM) and M&B 28767 (IC₅₀=4.0 ± .4 nM) (Zeng et al., 1996a), reproduced the effects of PGE₂ on IL-6 secretion by HSB.2 T cells. Thus, stimulation of HSB.2 T cell secretion of IL-6 by PGE₂ is mediated by the adenylyl cyclase-coupled EP₄ Rs and/or EP₂ Rs. Butaprost (EP₂ R) (IC₅₀>1 µM) (Zeng et al., 1996a), failed to increase IL-6 secretion (fig. 4) presumably because it is a weak agonist for EP₂ R-mediated adenylyl cyclase activity. Furthermore, PGE₂-evoked increases in IL-6 secretion were suppressed by a PKA inhibitor but not by a PKC inhibitor, indicating the central role for cAMP-dependent PKA activation and ruling out the involvement of PKC. The PK inhibitor data supported the possibility that IL-6 responses to PGE₂ are dependent on EP₄/EP₂ Rs coupled to cAMP-PKA pathway, but not on EP₃ Rs and the PKC pathway.

The partial suppression of IL-6 response to PGE₂ in HSB.2 T cells by the RNA synthesis inhibitor actinomycin D suggested a transcriptional mechanism with a requirement for IL-6 mRNA synthesis rather than mRNA stability. Northern analysis of HSB.2 T cell poly (A⁺) RNA, by hybridization with a cDNA probe specific for human IL-6 showed that one predominant transcript of 2.4 kb, corresponding in size to mRNA encoding IL-6, was increased after 1 hr of exposure to PGE₂ (data not shown). Since IL-6 mRNA levels were low in HSB.2 T cells, we employed semi-quantitative RT-PCR with higher sensitivity than Northern blot to monitor the changes of mRNA encoding IL-6. The transcriptional regulatory mechanism was confirmed by finding that PGE₂ increased IL-6 mRNA level in HSB.2 T cells. RT-PCR analysis revealed higher levels of IL-6 mRNA after 1 hr of exposure to PGE₂ (fig. 3). The PGE₂ transcriptional regulation of IL-6 production had been demonstrated in macrophages and astrocytoma cells (Hinson et al., 1996; Fiebich et al., 1997). The EP₄/EP₂ R specificity of the PGE₂ stimulatory effect on IL-6 secretion was confirmed by demonstrating the capacity of both PGE₂ and EP₄/EP₂/EP₃ R-agonist misoprostol, but not the EP₃ R-agonists sulprostone or M&B 28767, to elevate the level of IL-6 mRNA (fig. 4). Thus, HSB.2 T cell IL-6 responses to PGE₂ are dependent on EP₄/EP₂ Rs. Again, butaprost failed to increase IL-6 mRNA expression (fig. 4) presumably because it is a weak agonist for EP₂ Rs.

As T cells can be activated by exposure to antigen and multiple immune cytokines, the effects of PGE₂ on IL-6 production were investigated after pretreatment of HSB.2 T cells with Con A as a mitogenic stimulus of early T cells. A mitogenic concentration of Con A alone increased the level of IL-6 secretion, and conditioned greater sensitivity of HSB.2 T cells to PGE₂ and a higher magnitude of IL-6 response to the same concentration of PGE₂ (fig. 2). Con A increased IL-6 secretion to nearly 12-fold higher than the basal level (Fig. 2). After Con A pretreatment, HSB.2 T cells responded with increased IL-6 secretion to 10⁹ M PGE₂ that had no effect on unactivated HSB.2 T cells (fig. 2). In addition, the magnitude of IL-6 secretion evoked by 10⁸ M-10⁶ M PGE₂ was increased 7-10-fold by exposure of HSB.2 T cells to Con A, when compared with Con A treatment alone. We have also found that Con A suppressed mRNA encoding EP₃ Rs and EP₂ Rs and concurrently increased EP₄ R mRNA (fig. 1), without detectably altering the level of IL-6 mRNA (fig. 3C). Con A-activation of HSB.2 T cells consequently increased the representation of EP₄ Rs in relation to EP₃ Rs even though Con A also suppressed EP₂ R mRNA, which served to augment the role of EP₄ Rs as the transducers of PGE₂ stimulation of IL-6 secretion by Con A-stimulated HSB.2 T cells. This may explain the effect of Con A in enhancing both basal and PGE₂-stimulated levels of IL-6 secretion but with similar fold-increases in IL-6 secretion evoked by a range of concentrations of PGE₂ for unstimulated and Con A-stimulated HSB.2 cells, when compared with their respective basal IL-6 secretion without PGE₂ (fig. 2). Con A also enhanced the potency of PGE₂ as a stimulus of HSB.2 T cell expression of IL-6 mRNA. This was evidenced by our finding of a greater increase in expression of IL-6 mRNA when HSB.2 T cells were incubated with Con A and PGE₂, than when HSB.2 T cells were incubated with PGE₂ alone (fig. 3). Con A activation thus both up-regulates EP₄ R expression selectively and achieves sensitization of T cells to the effects of PGE₂. Con A stimulated HSB.2 T cells to greater PGE₂-evoked increases in IL-6 mRNA and secretion, at least in part by changing the profile of expression of EP Rs.

Therefore, PGE₂-evoked increases in IL-6 production and secretion are attributable to EP₄ Rs and/or EP₂ Rs in HSB.2 T cells, that are mediated through a transcriptional mechanism and cAMP-PKA-dependent events. Con A-induced alterations in the relative ratio of EP₄/EP₃ Rs directly contributed to the specificity of changes in IL-6 responses to PGE₂ in HSB.2 T cells.