MyD88 and TRIF mediate the cyclic adenosine monophosphate (cAMP) induced corticotropin releasing hormone (CRH) expression in JEG3 choriocarcinoma cell line

JEG3 choriocarcinoma cell line was obtained from American Type Tissue Culture Collection (Manaaas, VA) and cultured in MEM (Invitrogen Life Technologies, Carlsbad, CA) supplemented with 10% fetal bovine serum, 10 mM HEPES, 1 mM sodium pyruvate and 100 nM of penicillin/streptomycin (Invitrogen Life Technologies). A cell permeable cyclic AMP (cAMP) analog, 8-bromoadenosine 3',5' cAMP was obtained from Sigma-Aldrich (St. Louis, MO).

The CRH-luciferase vector, pGL3-CRH 663, was characterized and described previously [8]. Dominant negative (DN) cDNA constructs of MyD88 and interleukin-1 receptor-associated kinase (IRAK)2 have been characterized and described previously [9]. The NH₂-terminally deleted DN-MyD88 coding for amino acids 152–296 inhibits IL1 induced NF-kB activation [9]. IRAK2 is downstream to MyD88 and DN-IRAK2 coding for aa 1–96 inhibits MyD88 induces signaling [9]. The pcDNA3 empty vector and pCMV-β-galactosidase vectors have been described previously [10]. Luciferase data obtained from the cells transfected with pcDNA3 empty vector was used to assess the specificity of dominant negative vectors (DN-MyD88, DN-TRIF) to suppress cAMP-induced luciferase activity. The viral cAMP response element (CRE)-luciferase vector was described by Giebler et al [11]. This vector contains CRE and GC-rich flanks form a critical DNA element (called the viral CRE) that is obligatory for human T cell leukemia virus type 1 (HTLV-1) protein Tax transactivation [11]. Dr. Shizuo Akira generated the dominant negative TIR domain-containing adapter inducing IFN-beta (TRIF) by truncating the full length 712 aa TRIF to 162 aa Toll/Interleukin-1 receptor domain (TIR)-only domain [12]. The AP-1 luciferase construct was kindly obtained from Dr. Phillip Koeffler (Cedars-Sinai Medical Center, UCLA School of Medicine).

JEG3 cells were plated at a concentration of 50,000 cells/well in 24-well plates and cultured in MCDB-131 with 10% fetal bovine serum overnight. Cells were co-transfected the following day with FuGene6 Transfection Reagent following the manufacturer's instructions. The Roche Fugene6 transfection system is routinely used by our laboratory and others, and does not affect the cell viability or induce cytokine production in the transfected cells [10]. The reporter genes CRH-Luciferase (0.5 μg) and either empty vector or dominant negative mutants of MyD88 and TRIF were transfected into the JEG3 cells. Reporter gene CRE-luciferase (0.5 μg) was transfected to assess the effect of TLR stimulation on CRE expression. pCMV-β-galactosidase cDNA (0.1 μg) was transfected to normalize the results for transfection efficiency as described earlier [10].

After overnight transfection, the cells were stimulated with various concentrations of cAMP. CRH-luciferase cDNA transfected JEG3 cells were stimulated with cAMP for 5 hrs or 20 hrs in separate sets of experiments. Cells were then lysed and luciferase activity was measured with a Promega kit (Promega, Madison, WI) and a luminometer. The adequacy of the transfection was assessed by co- transfecting the cells with β-galactosidase vector and performing calorimetric galactosidase assay. This is a very well established and accepted method of assessing the transfection efficiency. The cells transfected with β-galactosidase are also considered to be transfected with the vector in question, in our case CRH-luciferase vector. β-galactosidase activity was determined by calorimetric method as described earlier [10]. The luciferase data was corrected for transfection efficiency by dividing the luciferase measurement with the galactosidase measurement. This result was expressed as the relative light unit (RLU).

Each experiment was performed in triplicate or quadruplicate and repeated at least three independent times. To study the effect of the varying combinations and concentrations of cAMP and the DNA-vectors across each assay, one-way ANOVA was used to examine the difference in luciferase units across the different treatment groups. Tukey's test was then used to compare the average luciferase units between each of the treatment groups against the control while properly adjusting for multiple comparisons. The data presented are the results of a representative experiment.

cAMP is known to induce CRH expression in the trophoblasts. We first confirmed and extended those observations in JEG3 cells by transiently transfecting them with CRH-luciferase expression vector. We treated the cells with different concentrations of cAMP (0.1, 0.5, 1 μM) for 5 or20 hours and observed that cAMP treatment increased CRH promoter activation in a dose dependent manner and this was evident at 5 h (data not shown). Based on this data we stimulated the cells with 0.5 μM cAMP for 5 hr in the rest of the experiments. The transcription factor AP-1 has been shown to mediate cAMP induced CRH promoter activation in humans [13]. As anticipated cAMP stimulation induced luciferase activity significantly above media treated control in JEG3 cells transiently transfected with AP-1 luciferase vector (Figure 1). Pair-wise comparisons using Tukey's test found that this increase was maximal at 0.3 μM cAMP concentration.

Infection induced innate immune responses play a significant role in the pathogenesis of preterm delivery. Innate immune receptors, Toll like receptors, are expressed in the female reproductive system and placenta, and mediate the immune responses to intrauterine infections [14, 15]. MyD88 is the common adaptor molecule that transmits signals from all TLRs and the IL1 receptor to induce NF-κB activation [3, 16].

We have previously shown that MyD88 mediates the LPS-induced, TLR4 mediated, CRH promoter activation in the placenta [7]. In order to assess whether MyD88 plays a direct role in cAMP-induced CRH promoter activation, we cotransfected JEG3 cells with CRH-luciferase and β-galactosidase expression vectors and either nonsignaling dominant negative (DN)-MyD88 [9] or empty vector and stimulated the cells with cAMP. There was no increase in cell death after transfection or cAMP treatment as assessed by microscope.

We observed that expression of DN-MyD88 significantly blocked the cAMP-induced CRH expression in a dose dependent manner (Figure 2A). DN-MyD88 effect was maximal at 0.5 μM concentration, as the luciferase activity at this concentration was significantly lower than that seen with 0.3 μM DN-MyD88 (p = 0.03) (Figure 2A). These data suggest that inhibition of MyD88 signaling blocks the cAMP-induced CRH promoter activation.

MyD88-independent signaling is mediated via adaptor molecule TRIF, which transduces signals induced by TLR4 and TLR3 [3, 17]. MyD88-independent pathways are involved in interferon (IFN) regulatory factor (IRF)-3 activation and subsequent induction of IFN-β and IFN-inducible genes and delayed NF-kB activation by TLR4. To test if TRIF plays a role in cAMP-induced CRH promoter activation, we cotransfected the JEG3 cells transiently expressing CRH-luciferase and β-galactosidase expression vectors with either DN-TRIF or empty vector and stimulated them with cAMP. Pair-wise comparisons using Tukey's test found that expression of both 0.3 and 0.5 μM DN-TRIF significantly inhibited the cAMP-induced CRH-luciferase expression (p < 0.01 for each concentration) (Figure 2B). This decrease was maximal at the 0.3 μM concentration of DN-TRIF (Figure 2B). These data suggest that inhibition of TRIF signaling blocks the cAMP induced CRH promoter activation.

IRAK2 is downstream to MyD88 and mediates the MyD88 induced NF-κB activation [3, 9]. We transiently transfected the JEG3 cells expressing CRH-luciferase and β-galactosidase expression vectors with either empty vector or various concentrations of DN-IRAK2 cDNA. We stimulated the cells with cAMP for 5 hours and examined the CRH promoter activation by performing luciferase assays. We observed that expression of DN-IRAK2 cDNA did not block the cAMP-induced CRH promoter activation (data not shown). These data suggest that MyD88 regulation of cAMP-induced CRH expression is proximal to IRAK2.

The cAMP response element (CRE) is downstream of PKA and mediates the cAMP induced CRH promoter activation in trophoblasts [1, 8, 18]. We questioned whether MyD88 and TRIF regulated cAMP-induced CRE promoter expression. We first confirmed that 5 hr stimulation with cAMP induces CRE activation in a dose dependent manner in JEG3 cells transiently expressing the CRE-luciferase expression vector (Figure 3). This increase was maximal at 0.5 μM cAMP as CRE-luciferase levels were significantly higher at this concentration over those seen at the 0.3 umol concentration (p < 0.01) (Figure 3).

Next, we tested whether Myd88 and TRIF influences cAMP-induced CRE expression. In JEG3 cells transiently cotransfected with CRE-luciferase vector, expression of DN-MyD88 cDNA significantly decreased the cAMP-induced CRE-luciferase levels in a dose dependent manner (p < 0.01 for all concentrations). This decrease was maximal at 0.5 μM concentration of DN-MyD88; at this concentration of DN-MyD88 cAMP-induced luciferase activity was significantly lower than that observed with the 0.3 μM concentration (p = 0.04) (Figure 4A). In contrast, DN-TRIF did not inhibit cAMP-induced CRE promoter activation (Figure 4B).