2822 A. Suarez et al. Eur. J. Immunol. 1997.27: 2822-2829

h a Suairez', Lourdes Mozo2, Abel Gayo', Josh Zam0rano3 and Carmen Gutierrez'? * ' Department of Immunology,

Hospital Central de Asturias, Centro Universitario, Oviedo, Spain Department of Functional Biology, Universidad de Oviedo, Spain American Red Cross, Immunology Department, Rockville, Maryland, USA

Requirement of a second signal via protein b a s e C or protein kinase A for maximal expression of CD40 ligand. Involvement of transcriptional and posttranscriptional mechanisms

High levels of CD40 ligand (CD40L) protein expression are induced on native T cells by increasing the intracellular Ca2+ concentration. In the present study we have shown that ionomycin induces CD40L gene transcription leading to mFWA accumulation which translates to high levels of protein expression. Con- versely, agents which increase the intracellular levels of cyclic AMP (CAMP), such as prostaglandin E2 (PGE2) or dibutyryl cyclic AMP (dbcAMP), were unable to induce CD40L expression on T lymphocytes. Cell activation by phor- bol 12-myristate 13-acetate (PMA) treatment had a slight effect on increasing CD40L mRNA and protein levels. However, PMA and dbcAMP synergized with ionomycin to significantly increase and to prolong the CD40L expression. Nuclear run-on assays revealed that PMA, but not dbcAMP, increased threefold the CD40L gene transcription rate induced by ionomycin. This effect was inde- pendent of de novo protein synthesis. In addition, at a posttranscriptional level, both reagents synergized with the Ca2+ ionophore to prolong the CD40L mRNA half-life by a mechanism which was also independent of de now protein synthe- sis. Moreover, when transcription was blocked with actinomycin D, an incre- ment of the CD40L transcript levels induced by PMA or dbcAMP on ionomycin- treated cells was observed in the presence of cycloheximide. This probably means that newly synthesized protein may contribute to the CD40L mRNA de- stabilization. In summary, these data show that PMA and dbcAMP synergized with ionomycin to increase the CD40L mRNA and protein levels. The up- regulatory effect of PMA was accomplished at a transcriptional and posttran- scriptional level, whereas dbcAMP exerted its synergistic effect exclusively at a posttranscriptional level.

1 Introduction

Upon activation, CD4' T cells transiently express on their surface a molecule which binds CD40. This molecule, called CD40 ligand (CD40L), is a 33-kDa type I1 trans- membrane protein with a significant structural homology to the extracellular domain of human TNF-a. It is prefer- entially expressed on activated CD4' cells, but it has also been described in activated mast cells, eosinophils, baso- phils and some CD8' T cells [l-51. CD40 is a 50-kDa type I cell surface protein present on B cells and other cell types such as dendritic cells, endothelial cells, thymic epi- thelial cells, monocytes and macrophages.

[I 167591

Received February 24, 1997; in final revised form July 30, 1997; accepted August 5 , 1997.

The interaction between CD40L and CD40 has a profound physiological significance, since it is necessary for B cells to acquire the capacity to proliferate and differentiate in response to cytokines. In vitro studies have shown that either mAb to CD40L or soluble forms of CD40 inhibit T cell-dependent B cell activation. Expression of CD40L is also essential for production of antibody responses in vivo [6, 71. Significantly, a mutation in the gene encoding CD40L is responsible for the X-linked hyper-IgM syn- drome characterized by the failure to produce antibodies of classes other than IgM [8, 91. A similar immunoregula- tory defect has been observed in CD40 or CD40L knock- out mice [lo, 111. Moreover, alterations in the expression of CD40L seem to play an important role in the develop- ment of autoimmune diseases. Thus, antibodies to CD40L prevent collagen-induced arthritis in mice, and a hyperex- pression of CD40L by B and T cells in human lupus has recently been described [ 12, 131.

Correspondence: Carmen GutiCrrez, Servicio de Immunologia, Interaction between CD40 and CD40L is also important in Hospital Central de Asturias, Julian Clavena s/n, E-33006 T cell-mediated responses as CD40L-defective mice are Oviedo, Spain unable to mount an optimal primary T cell response to Fax: +34-8-5106140 protein antigens [14-181. In addition, expression of

CD40L on Tcells is required to activate macrophages and Abbreviations: CD40L: CD40 ligand dbcAMP: Dibutyryl monocytes to produce pro-inflammatory cytokines cyclic AMP PKA: Protein kinase A CHX: Cycloheximide [19-211. Therefore, CD40L seems to play a master role in CSA: Cyclosporine A Act D: Actinomycin D

the mechanisms of immunoregulation by influencing both

00 14-2980/97/1111-2822$17.50 + .5 0/0 0 WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1997

Eur. J. Immunol. 1997.27: 2822-2829 Requirement of a second signal for maximal expression of CD40 ligand 2823

CD40L is rapidly induced by cross-linking of TCR by immobilized anti-CD3 mAb. The activation through the TCR can be bypassed by activation of PKC with PMA and by increasing the levels of intracellular Ca" by ionomycin, and in fact, expression of CD40L is achieved upon activa- tion of CD4' cells by these two agents. However, in a recent article [22] it has been demonstrated that activation of the Caz+/calmodulin pathway, but not that of the PKC pathway, was the essential signal for the expression of CD40L in T cells. In the present report we confirm this finding, showing that an increase in the intracellular Ca2+ is a sufficient signal to induce gene transcription and accu- mulation of CD40L mRNA. However, a co-stimulation with PMA augments and prolongs the CD40L gene expression by acting at the transcriptional and posttrans- criptional level. We have also shown, for the first time, that agents which increase the intracellular levels of CAMP, a physiological mediator in the regulation of the inflamma- tory responses [23], are capable of increasing the CD40L gene mRNA expression by posttranscriptional mecha- nisms .

2 Materials and methods

2.1 Reagents and mAb

Ionomycin (Io), PMA, cycloheximide (CHX), dibutyryl cyclic AMP (dbcAMP), actinomycin D (Act D), L-leucine methyl ester, prostaglandin (PG)Ez, o-phenylenediamine dihydrochloride and p-nitrophenyl phosphate disodium salt were obtained from Sigma Chemical Co. (St. Louis, MO). Cyclosporine A (CSA) was purchased from Sandoz (Basel, Switzerland). Anti-human CD40L mAb (M90) was a kind gift from Dr. Elaine K. Thomas from Immunex Corp. (Seattle, WA). R-PE-conjugated goat F(ab'), anti- mouse IgG, affinity-purified goat Ab to human y and p chains and peroxidase-conjugated goat-anti human IgG and IgM were obtained from Tagoimmunologicals (Cama- rillo, CA). Affinity-purified goat Ab to human a chain and alkaline phosphatase-conjugated goat-anti human IgA were obtained from Sigma Chemical Co. (St. Louis, MO).

2.2 Cell isolation and culture

Buffy coats from healthy blood donors were kindly pro- vided by the Centro Comunitario de Transfusiones (Oviedo). PBMC were obtained by Ficoll-Hypaque den- sity centrifugation (Lymphoprep, Nycomed, Oslo, Nor- way). Purified peripheral T cells were obtained by deplet- ing monocytes and NK cells by incubation with 1 mM L- leucine methyl ester [24]. To prevent T cell activation caused by rosetting treatment, the samples were enriched in T cells by centrifugation through a 52 YO Percoll density gradient as described [25]. Cells recovered from the pellet were CD3' >97%, CD19' <2%, CD14' e l % , CD16'56' < 1% as estimated by flow cytometry. For Ig production, B cells were isolated from PBMC by two cyc- les of rosetting with neuraminidase-treated SRBC. The B cell-enriched population obtained consisted of >90 'YO B cells (CD20+) and <1 YO T cells (CD3') as determined by flow cytometry.

Cell cultures of either PBMC and PBT cells were carried out at a concentration of 2 x 106 cells/ml at 37 "C in 5 % C02 in RPMI 1640 medium containing 2 mM L-glutamine and 25 mM Hepes (Bio Whittaker, Verviers, Belgium) and supplemented with 10 % heat-inactivated FCS (ICN Flow, Costa Mesa, CA) and antibiotics. Cells were stimulated with ionomycin, PMA, dbcAMP, PGE2 or a combination of these reagents. The final concentrations of the stimu- lants used in culture were: ionomycin 1 pg/ml, PMA 10 ng/ ml, dbcAMP 0.5 mM and PGEz 1 pM.

2.3 Flow cytometry

Surface expression of CD4OL was assessed by indirect immunofluorescence analysis. PBMC or PBT cells were first incubated with anti-CD4OL mAb (M90) or an irrele- vant antibody used as negative control for 30 min at 4 "C. Cells were washed twice with PBS and incubated with PE- conjugated goat anti-mouse IgG for 30 min at 4°C. After two final washes the staining intensity was analyzed using a FACScan flow cytometer and CellQuest software (Bec- ton Dickinson). Fluorescence of 10 OOO cells was accumu- lated for analysis and the specific fluorescence intensity was quantified as the median fluorescence intensity (MFI) and calculated by subtracting the background fluores- cence.

2.4 Northern blot analysis

Total cellular RNA was isolated by the method of Chom- czynski and Sacchi [26]. Ten micrograms of total RNA were subjected to electrophoresis in a denaturing 1.2% agarose gel containing 2 YO formaldehyde and transferred to nylon membranes as described [27]. Prehybridization, hybridization washes and autoradiography were per- formed by standard methods [27]. Briefly, blots were pre- hybridized for 8 h at 42°C in 5 x (SSPE), 1 % SDS, 5 x Denhart's solution, 5 x dextran sulfate, 50% formamide and 100 pgh l salmon sperm DNA. Hybridization was car- ried out at 42 "C for 48 h with the same solution containing the CD40L cDNA probe radioactively labeled with 32P- dCTP by using a random priming labeling kit from Strata- gene (La Jolla, CA). Later, blots were stripped and rehy- bridized with the GAPDH cDNA probe. The cDNA probe for CD40L was obtained by PCR amplification of first- strand cDNA prepared from RNA isolated from PBMC stimulated with 10 pg/ml Con A plus 10 ng/ml PMA for 8 h. Primer sequences used for CD4OL amplification were as follows: sense, 5 '-ATGATCGAAACATACAACCA- AA-3 ' and antisense, 5 ' -TCAGAGTITGAGTAAGCCA- AAG-3'. The size of the product was 846 base pairs and its identity was confirmed by its specific size by electrophore- sis in a 2 % agarose gel and by enzymatic digestions with Taq I and Pvu 11. The GADPH cDNA probe was obtained in a similar way. After hybridization, blots were stringently washed three times during 1 h with 0.2 x SSC and 0.5 YO SDS at 65 "C and exposed to X-rax film at -70 "C. Bands on the autoradiograms were integrated by scanning densi- tometry using the Bio Image Intelligent Quantifier pro- gram (Ann Arbor, MI) and values were normalized to the signal of the control GAPDH probe.

2824 A. Su6rez et al. Eur. J. Immunol. 1997.27: 2822-2829

2.5 Nuclear run-on transcription assay

The transcription rate of the CD40L gene was measured by nuclear run-on transcription assay as described elsewhere [28]. Briefly, lo8 cells were washed twice with PBS, lysed for 5 min in 4 ml ice-cold lysis buffer (10 mM Tris-HC1 pH 7.5,3 mM MgCl,, 10 mM NaCl and 0.5 YO NP40), cen- trifuged at 500 x g for 5 min and the lysis repeated. The nuclei were resuspended in 100 pl glycerol buffer (50 mM Tris-HC1 pH 8, 40% glycerol, 5 mM MgC12, 0.1 mM EDTA) and incubated at 26°C for 20 rnin with 80 p1 of transcription buffer (12.5 mM Tris-HC1 pH 8.0, 6 mM MgCl,, 125 mM KCl, 2 mM DTT, 1 mM each of ATF', CTP and GTP) in the presence of 125 pCi [a-32P]UTP (3000 Ci/ mol; ICN). The reaction was terminated by incubation at 37°C for 20 rnin with 200 p1 stop buffer (10 mM Tris-HC1 pH 7.4, 0.5 M NaCl, 50 mM MgCl,, 2 mM CaC12), 40 U DNase and 150 U RNasin. After 45 rnin of proteinase K digestion (200 pglml) in 1 % SDS at 37"C, the labeled RNA was phenokhloroform extracted, precipitated twice in 2.5 M ammonium acetate and resuspended in 50% formamide. Equal numbers of cpm of 32P-RNA were hybridized for 24 h on nylon filters slot-blotted with 1.5 pg of the CD4OL and GAPDH cDNA probes and a non- relevant PCR product obtained from Bluescript (pBSK) as a negative control. Hybridization, washing, exposure and quantification of transcription rate levels were performed as described above.

2.6 Culture assay for Ig production

To assay T cell helper activity, freshly isolated T cells were cultured at 2 x lo6 cells/ml for 12 h in complete medium with the stimuli indicated. Cells were then fixed with 1 YO paraformaldehyde for 10 min and washed five times in RPMI [29]. After that, 1 x lo6 isolated B cells were cul- tured with l X lo6 fixed activated T cells for 12 days. Con- trol cultures were performed with B cells alone. Ig con- centrations in the culture supernatants were measured by ELISA.

2.7 Ig determination

Ig concentrations in the supernatants of cultured cells were measured by isotype-specific ELISA. For IgG, IgM and IgA determination, microtiter plates (Dynatech) were coated with affinity-purified goat Ab to human y , p or a chains at the appropriate dilution and kept overnight at 4°C. After washing, wells were blocked with PBS contain- ing 1 % BSA (w/v) for 2 h at 37°C. Supernatants and human Ig standards diluted in blocking solution were added to microtiter wells and incubated for 2 h. Bound Ig was detected with peroxidase-conjugated goat anti-human IgG and IgM or alkaline phosphatase-conjugated goat anti-human IgA using o-phenylenediamine dihydrochlo- ride or p-nitrophenyl phosphate disodium, respectively, as a substrate. Absorbance was determined at a wavelength of 492 nm for IgG and IgM and at 405 nm for IgA. Amounts of Ig were calculated according to the standard curves. The lower limit of detection was 12 ng/ml for IgM and IgG and 8 ng/ml for IgA.

3 Results

3.1 A second signal via PKC or PKA increments and prolongs the CD40L protein and mRNA expression obtained with ionomycin

In the present experiment we have studied the surface expression of CD40L on purified T cells after 5 h of cul- ture with various stimuli, as shown in Fig. l. We have also analyzed the CD40L mRNA steady-state levels in PBMC stimulated for 3 h with the same reagents (Fig. 2). All experiments at mRNA level were performed with total PBMC since in previous experiments no significant differ- ences of the mRNA accumulation were found in purified T cells or PBMC. We have found that cell activation with ionomycin alone induced significant levels of CD4OL mRNA which translated to a high level of protein expres- sion. However, a concomitant activation of PKC through the addition of PMA to cell culture exerted a significant synergistic effect with ionomycin increasing the transcript levels and protein expression considerably. PMA alone, although it did not induce accumulation of CD40L mRNA, slightly increased the percentage of CD40L- positive cells and the specific MFI. Activation of PKA by agents increasing the intracellular concentration of CAMP, such as PGE2 or the hydrophobic analogue of CAMP, dbcAMP, also synergized with ionomycin to increase the protein and, to a lesser extent, the CD40L mRNA levels. The signal delivered by these agents, in the absence of ionomycin, was unable to induce either the protein or mRNA expression of the CD4OL gene. Simultaneous acti- vation of cells with PMA, PGE2 and ionomycin did not result in an increase of the CD40L protein or mRNA expression compared with that obtained with ionomycin plus PMA.

To study the CD40L expression over time, peripheral T cells were activated with ionomycin, ionomycin plus PMA or ionomycin plus dbcAMP and the protein expres- sion analyzed at different times. Peak levels were obtained after 6 h of culture in all three conditions analyzed, but an up-regulated expression was found when a second stimulus with either PMA or dbcAMP was employed together with ionomycin (Fig. 3). The MFI and the percentage of CD40L-positive cells decreased after 24 h in cultures with ionomycin alone, whereas it remained for at least 48 h in cultures with ionomycin plus PMA or ionomycin plus dbcAMP. It is of interest that at 20 h we observed a second peak of CD40L surface protein which was considerably more noticeable when the kinetic expression was studied with total PBMC instead of purified T cells (data not shown). This biphasic kinetics has also been observed by other authors when CD4' T cells were activated in the presence of endothelial cells [30].

These results suggest that an increase of the intracellular Ca2+ is a sufficient signal to obtain a high level of CD40L expression both at the protein and the mRNA level, but additional stimuli result in an up-regulation and stabiliza- tion of its expression. These stimuli may act by increasing the transcription rate or, at the posttranscriptional level, by stabilizing the accumulated mRNA induced by the Ca2+ signal.

Eur. J . Immunol. 1997.27: 2822-2829 Requirement of a second signal for maximal expression of CD40 ligand 2825

Medium PMA

C - lo lo+PMA

kjhTJ loo 101 10 103 104 loo 10' 10 lo3 lo4

lwPGE2 IoidbcAMP

CD40L FLUORESCENCE INTENSITY

3.2 PMA, but not dbcAMP, synergizes with ionomycin to increase the CD40L transcription rate

A nuclear run-on transcription assay was performed to determine whether the positive effect of PMA or dbcAMP on the CD40L expression was due to an increased tran- scription rate. Nuclei were isolated from PBMC after 90 min of culture in medium alone or with ionomycin, PMA, ionomycin plus PMA and ionomycin plus dbcAMP. Results from a representative experiment are shown in Fig. 4. Without stimulation, a low rate of CD40L tran- scription was observed, which was highly increased by

CD40L

GAPDH

'i I I

0

Figure 2. Expression of CD4OL mRNA with different stimuli. Total cellular mRNA was extracted from PBMC cultured for 3 h with the stimuli indicated and at the same concentration as in Fig. 1. Northern blot was performed and the specific CD40L mRNA quantified by densitometric scanning of autoradiographs. Relative values obtained after normalization to the GAPDH sig- nal are shown. Data are representative of four independent exper- iments with similar results.

Figure 1. Surface expression of CD40L with different stimuli. Enriched peripheral T cells (97 % CD3') were cultured for 5 h in medium alone or with PMA, dbcAMP, PGEz, ionomy- cin (Io), 10 + PMA, 10 + dbcAMP and 10 + PGEz and analyzed using flow cytometry. Each reagent was employed at the following concentration: 10 1 pg/ml, PMA 10 ng/ml, dbcAMP 0.5 mM and PGEz, 1 KM. CD40L expression (solid line) was determined by indi- rect staining with the mAb M90. The percent- age of CD40L-expressing cells (minus control staining) and the MFI on a logarithmic scale of the positively staining population calculated by subtracting the background fluorescence signal (dotted lines) is represented in the top right cor- ner of each histogram. Profiles are from a single experiment and are representative of at least four experiments for each stimulant.

ionomycin, but not by PMA. However, PMA, but not dbcAMP, synergized with ionomycin to increase the rate of transcription threefold compared with ionomycin alone. These results suggest that activation of the PKC, but not of the PKA metabolic pathway, has a positive regulatory effect on the CD40L gene expression by increasing the transcription rate in ionomycin-activated cells.

3.3 PMA and dbcAMP increase CD40L mRNA stability

We studied whether the effect of PMA and dbcAMP on CD40L mRNA accumulation was exerted at a posttran- scriptional level. PBMC were activated with ionomycin for 3 h to induce a significant mRNA expression. Then, PMA or dbcAMP, with or without 1 pM CSA to block selec- tively CD40L gene transcription, was added to the cell cul- tures. I t has been previously shown that this agent totally inhibits the transcription of the CD40L gene [31]. The amount of mRNA was studied by Northern blot after 90 min of culture (Fig. 5) . In the presence of CSA no detectable amount of CD40L mRNA in control cultures was evident, indicating that in the absence of gene tran- scription, the ionomycin-induced mRNA disappeared within a short period of time. However, under the same experimental conditions, a quantifiable amount of mRNA was observed in cultures with PMA or dbcAMP, suggesting that these agents stabilize the ionomycin-induced CD40L transcript levels. PMA was more efficient than dbcAMP since a higher amount of mRNA was obtained with the for- mer reagent. Control cultures (without CSA) produced the expected results. Ionomycin promoted the CD40L mRNA expression which was increased by adding PMA or dbcAMP.

Next, we determined the CD40L mRNA half-life in cells activated with ionomycin, ionomycin plus PMA or iono- mycin plus dbcAMP (Fig. 6A, B, C). After 3 h of culture, mRNA decay measurements in the presence of Act D, a general transcription inhibitor, showed that whereas the CD40L mRNA half-life induced by ionomycin alone was

2826 A. SuArez et al. Eur. J. Immunol. 1997.27: 2822-2829

A. a+,

I

0 4 8 1 2 1 6 2 0 2 1 2 8 3 2 3 6 4 0 4 4 4 8

TIME (hours)

B.

Y -I w 0 w 2 t (I)

B a -1 0

0 ae

0 4 8 1 2 1 6 2 0 2 4 2 8 3 2 3 6 4 0 U 4 8

TIME (hours) -.- Medium lo * lo+PMA * lo+dbcAMP

50 min, addition of PMA or dbcAMP prolonged the mRNA half-life up to 120 and 70 min, respectively. These results clearly show that activation of the PKC or PKA metabolic pathway by PMA or dbcAMP, respectively, exert a posttranscriptional function stabilizing the ionomycin-induced CD40L mRNA.

3.4 The posttranscriptional effect mediated by PMA or dbcAMP is not dependent on de ~ O Y O protein synthesis

We analyzed whether the increase of the CD40L mRNA levels induced by a second signal delivered by PMA or dbcAMP on ionomycin-activated cells was dependent on de novo protein synthesis. Either CHX, Act D or both agents together were added to cells preactivated with iono- mycin for 3 h and subsequently cultured with PMA or dbcAMP for a further 90 min. Following this, total RNA was extracted and the amount of CD40L transcripts ana-

Figure 3. Kinetics of surface expression of CD40L. Enriched peripheral T cells were cul- tured at 2 x lo6 celldml in medium alone or with ionomycin (Io), 10 + PMA and 10 + dbcAMP at the same concentration as in Fig. 1. The cells were harvested at the times shown and stained with anti-CD40L mAb (M90) for flow cytometric analysis as described in Sect. 2.3. (A) MFI was calculated after sub- tracting the background staining from the total fluorescence. (B) Percentage of CD4OL- positive cells.

lyzed by Northern blot. In Fig. 7 it may be observed that the CD40L mRNA levels induced by PMA or dbcAMP were not reduced by blocking protein synthesis with CHX, but rather that a slight increase of the accumulated mRNA was observed, which indicates that the positive effect of PMA or dbcAMP on the CD40L gene expression was not dependent on de novo protein synthesis. When the tran- scription was inhibited by Act D , the presence of CHX also up-regulated the CD40L mRNA induced by PMA or dbcAMP, suggesting that CHX exerted a positive post- transcriptional effect probably by blocking the synthesis of destabilizing factors and resulting in an increment of the accumulated CD40L mRNA.

3.5 CHX did not increase CD40L transcription rate

We also considered, as has been described for other genes [32], the possibility that the up-regulatory effect of CHX on the CD40L mRNA levels was due to an increased tran- scription rate in addition to the posttranscriptional effect described above. Nuclear run-on assays were performed in

CD40L

GAPDH

pBSK

- 0

Figure 4. Transcriptional effects of ionomycin (Io), PMA and dbcAMP on CD40L expression. After 90 min of culture with the stimuli indicated at the same concentrations as in Fig. 1, PBMC were harvested and lysed to obtain the nuclei for the run-on tran- scription assay. The histogram represents relative transcription rates, which were calculated after normalization to the respective GAPDH signals. The experiment shown is one of three experi- ments performed.

CD4OL

GAPDH

CSA - - - + + + PMA - + - - + -

dbcAMP - - + - - +

Figure 5 . Posttranscriptional effects of PMA and dbcAMP on the CD40L gene expression. PBMC were cultured for 3 h with 1 pg/ ml 10 to obtain CD40L mRNA, then 1 pM CSA was added to inhibit CD40L gene transcription and cells were cultured for a fur- ther 90 min in the presence of 10 nglml PMA or 0.5 mM dbcAMP. Control cultures without CSA were performed under the same conditions. CD40L mRNA levels were analyzed by Northern blot as described in Sect. 2.4.

Eur. J. Immunol. 1997.27: 2822-2829 Requirement of a second signal for maximal expression of CD40 ligand 2827

A B C CD4OL

GAPDH

TIME Imin) TIME (minuter) TIME (rnin)

Figure 6 . PMA and dbcAMP increase the stability of CD40L mRNA. PBMC were stimulated with 10, 10 + PMA and 10 + dbcAMP. After 3 h of culture the transcription was arrested by adding 5 pg/ml Act D and Northern blot was performed at different times. The half-life was calculated by the measurement of mRNA decay calculated after normalization to the respective GAPDH signals. In the presence of 10, the CD40L mRNA half-life was 50 min (A), with 10 + PMA 120 min (B) and in the presence of 10 + dbcAMP 75 min (C). Representative results from two independent experiments are shown.

PBMC after 90 min of incubation with ionomycin or iono- mycin plus PMA in the presence or absence of CHX. The results show that CHX did not increase the transcription rate of the CD4OL gene in either case, which indicates that the positive effect of CHX was exclusively posttranscrip- tional (Fig. 8).

3.6 T helper activity by differentially activated T cells

Lastly, we analyzed whether the increased expression of CD40L on differentially activated T cells was functional and able to induce Ig secretion. To this purpose we studied the production of IgM, IgG and IgA by B cells cultured with T cells previously activated with 10, 10 + PMA and 10 + dbcAMP and fixed to exclude cytokine production.

CD40L

GAPDH

CHX . - - + + + - - . + + + ACTD . - . - - - + + + + + + WA . + . . + - - + - . + .

&AMP . . + . - + . . + . - +

Figure 7. The posttranscriptional effect of PMA and the dbcAMP on the CD40L mRNA accumulation is independent of protein synthesis. PBMC were cultured for 3 h with 1 pg/ml 10 to promote gene transcription. Further protein synthesis, gene transcription or both were inhibited by addition of 10 &ml CHX, 5 &ml Act D or both reagents, respectively. Then the cells were cultured for a further 90 min with 10 ng/ml PMA or 0.5 mM dbcAMP. CD40L mRNA levels were analyzed by Northern blot and nor- malized as described in Sect. 2.4. Representative results from three independent experiments are shown.

As shown in Fig. 9, ionomycin activated T cells induced a significant secretion of all Ig isotypes, which was increased when T cells were activated by PMA or dbcAMP in addi- tion to ionomycin.

4 Discussion

In vivo expression of CD40L on the T cell surface is induced by cross-linking of antigen receptors by MHC- peptide complexes on APC. Although a significant expres- sion of CD40L may be obtained upon T cell activation with anti-CD3, maximal levels of surface protein expres-

CD40L

GAPDH

pBSK

+ X I 0

Figure 8. Addition of Cycloheximide did not increase the CD4OL transcription rate. Nuclei were prepared from PBMC stimulated with 1 pg/ml 10 or 10 plus 10 ng/ml PMA in the presence or absence of 10 &nl CHX for 90 min. Run-on transcription assay was performed as described in Sect. 2.5 and the relative transcrip- tion rate calculated with respect to the GAPDH gene transcrip- tion.

2828 A. Suhrez et al. Eur. J. Immunol. 1997.27: 2822-2829

Figure 9. T helper activity by differentially activated T cells. Puri- fied peripheral Tcells were stimulated for 12 h with 10, lo + PMA or 10 + dbcAMP at the same concentrations employed in Fig. 1. Then cells were fixed with 1 % paraformalde- hyde and cultured at 1 x lo6 cellslml with 1 x lo6 cells/ml of isol- ated B cells. After 12 days, culture supernatants were collected and assayed for Ig production by isotype-specific ELISA, as described in Sect. 2.7. Representative results from two independ- ent experiments are shown.

sion are achieved in vitro by the single or associated action of various polyclonal stimulators such as PHA, Con A, ionomycin and PMA [9,22]. In agreement with a previous report, we have observed that activation of the Caz+/cal- modulin pathway by ionomycin leads to high levels of CD40L protein expression [22]. Additionally, we have shown that the Ca2+ signal, in the absence of other stimuli, induces CD40L gene transcription leading to a significant CD40L mRNA accumulation. We have also shown that agents which increase the intracellular levels of cAMP and activate the CAMP-dependent kinase PKA, such as PGEz or dbcAMP, are, in spite of being unable to initiate CD40L gene transcription, capable of synergizing with ionomycin to enhance the levels of CD40L protein and to prolong its detection on the cell surface up to 48 h. Similarly, activa- tion of the PKC metabolic pathway by PMA also exerted a synergistic effect, up-regulating and prolonging the CD4OL expression in ionomycin-treated cells. However, whereas the increased levels of cAMP synergized with the Ca2+ signal by acting exclusively at a posttranscriptional level, activation of PKC by PMA exerted its synergistic effect at both, transcriptional and posttranscriptional level.

Using nuclear run-on assays, a spontaneous low rate of CD40L gene transcription was observed on PBMC which was highly increased by ionomycin treatment. Therefore, triggering of the Ca*+/calmodulin metabolic pathway must activate transcription factors implicated in the regulation of the CD40L gene expression. A putative candidate to promote the CD40L gene transcription may be the DNA- binding protein NF-ATp (preexisting nuclear factor of activated T cells) which binds to the NF-AT site in the pro- moter region of various cytokine genes, including TNF-a, a molecule which belongs to the same family of CD40L [33-351. NF-ATp appears in nuclear extracts of T cells within 10 min of stimulation, and this occurs via a calcium- dependent signaling process that is blocked by CSA and

FK-506, similarly to CD4OL gene expression [36-381. Fur- thermore, six potential NF-AT-binding sequences have been identified in the 5'-region of the mouse CD40L gene [39]. Nevertheless, and since PMA increases threefold the transcription rate over that obtained with ionomycin alone, it might be that transcription factors activated by PMA, such as AP-1, AP-2 or NF-xB may also play a role in the regulation of the CD40L gene expression.

The ionomycin induced CD4OL mRNA is rather unstable since in the absence of other activating factors (PMA or dbcAMP) is is rapidly degraded with a half-life of 50 min. These reagents prolong significantly the CD40L mRNA half-life by a mechanism which does not require de novo protein synthesis. A superinduction of CD40L mRNA could even be observed in the presence of the protein inhibitor, CHX. Rapid degradation of mammalian mRNA encoding oncoproteins and lymphokines has been related to the preence of AU-rich sequences in the 3' untranslated region, probably, although not yet established, by mecha- nisms involving a rapid deadenylation, which favors nucle- olytic digestion [40]. Tandem repeats of AUUUA motifs have been identified in the 3' untranslated region of the CD40L mRNA [41] which may be the target of nucleases or nuclease-competing factors. Since PMA or dbcAMP confer stability to the CD40L transcripts by mechanisms which do not require the synthesis of new factors, these stimulating agents may activate preexisting factors that bind at, or near, the nuclease target site, inhibiting the nucleolytic digestion. Alternatively, they may activate fac- tors which inhibit the RNase activity. Several studies sug- gest that phosphorylation could play a role in the regula- tion of the activity of AU-binding proteins, leading to sta- bilization of different mRNA [42]. Regions other than AU-rich sequences can also confer instability to a particu- lar mRNA. Inhibition of the synthesis of labile proteins, such as the RNase L system, may account for the CHX- dependent superinduction of the CD40L transcript levels [32, 431.

Although it has not been investigated in the present report, in addition to the effect on the stability of the CD40L mRNA, PMA or dbcAMP may increase the CD40L protein levels by acting at other posttranscriptional stages, such as the rate of translation, the proteolytic pro- cessing or the intracellular protein trafficking. This may be particularly true in the case of the dbcAMP, since it does not increase the CD40L gene transcription rate and has only a limited effect in prolonging the mRNA half-life, increasing considerably, however, the protein cell surface expression.

Expression of CD40L by ionomycin or ionomycin plus dbcAMP seems to be of physiological relevance since it was able to induce secretion of all isotypes of Ig. In vivo, it may be that a combination of inflammatory products, leading to an increase in intracellular Ca2+ and CAMP, may bypass the usually stringent control of T cell activation by the TCWCD3 complex and thereby recruit non-Ag- specific T cell responses, with possible pathological conse- quences. A number of physiological and pharmacological agents share the ability to elevate intracellular cAMP in T lymphocytes such as the E series of prostaglandins, hista- mine and derivatives as well as P2-adrenergic receptor agonists. Evaluation of human disease corroborates that

Eur. J. Immunol. 1997.27: 2822-2829

overproduction of these agents correlates with elevated Th2 and IgE responses [44]. More recently, it has been shown that cAMP inhibits anti-CD3 or IL-2 proliferation of Thl cells but not of Th2 cells [45, 461. It may be that cAMP selectively activates Th2 cells which in turn down- modulate inflammatory Thl-mediated functions. It should be of interest to study possible differential regulatory mechanisms contributing to CD40L expression in different T cell subsets. A s B cell responses are mediated by Th2 cells, a CAMP-dependent enhanced CD4OL expression may contribute to the autoantibody production which often is associated with inflammatory processes and to IgE production in allergic reactions. We are presently investi- gating the helper capacity in vitro of CAMP-stimulated T cell subsets to induce the switch to IgE production.

The authors are grateful to Dr. Elaine K. Thomas for providing the anti-human CD4OL mAb that allowed us to carry out this investiga- tion. We also thank David H. Wallace for the English correction of this manuscript. This work was supported by grant FISS 131195 from the Fondo de Investigaciones Sanitarias de la Seguridad Social.

Requirement

5 References

1 Banchereau, J., Bazan, F., Blanchard, D., Brikre, F., Galizzi, J. P., van Kooten C., Liu, Y. J., Rousset, E and Saeland, S., Annu. Rev. Immunol. 1994. 12: 881.

2 Hollenbaugh, D., Ochs, H. D., Noelle, R. J., Ledbetter, J. A. and Aruffo, A., Immunol. Rev. 1994.138: 23.

3 Hollenbaugh, D., Grosmarie, L. S., Kullas, C. D., Chalupny, N. J., Braesch-Andersen, S., Noelle, R. J., Stamenkovic, I., Ledbetter, J. A. and Aruffo, A., EMBO J. 1992.11: 4313.

4 Noelle, R. J., Roy, M., Shepherd, D. M., Stamenkovic, I., Ledbetter, J. A. and Aruffo, A., Proc. Natl. Acad. Sci. USA 1992. 89: 6550.

5 Peitsch, M. C. and Jongeneel, C. V., Int. Immunol. 1993. 5: 233.

6 Van den Eertwegh, A. J. M., Noelle, R. J., Roy, M., Shep- herd, D. M., Aruffo, A., Ledbetter, A. J., Boersma, W. J. A. and Claassen, E., J. Exp. Med. 1993.178: 1555.

7 Foy, T. M., Shepherd, D. M., Durie, E H., Aruffo, A., Led- better, J. A. and Noelle, R. J., J . Exp. Med. 1993. 178: 1567.

8 Aruffo, A., Farrington, M., Hollenbaugh, D., Li, X., Milato- vich, A., Nonoyama, S., Bajorath, J., Grosmarie, L. S., Sten- kamp, R., Neubauer, M., Roberts, R. L., Noelle, R. J., Led- better, J. A., Francke, U. and Ochs, H. D., Cell 1992. 72: 291.

9 Kroczek, R. A., Graf, D., Brugnoni, D., Giliant, S., Kort- hauer, U., Ugazio, A., Senger, G., Mages, H. W.,Villa, A. and Notarangelo, L. D., Immunol. Rev. 1994.138: 39.

10 Borrow, P., Xshon, A., Lee, S., Xu, J., Grewal,, I. S., Old- stone, M. B. A. and Flavell, R. A., J. Exp. Med. 1996. 183: 2129.

11 van Essen, D., Kikutani, H. and Gray, D., Nature 1995. 378: 620.

12 Buhlmann, J. E. and Noelle, R. J., J . Clin. Immunol. 1996. 16: 83.

13 Desai-Mehta, A., Lu, L., Ramsey-Goldman, R. and Datta, S. K., J. Clin. Invest. 1996. 97: 2063.

14 Grewal, I. S. and Flavell, R. A., Immunol. Today 1996. 17: 410.

of a second signal for maximal expression of CD40 ligand 2829

15 Guo, Y., Wu, Y., Shinde, S., Sy, M., Aruffo, A. and Liu, Y., J .

16 Shinde, S., Wu, Y., Guo, Y., Niu, Q., Xu, J., Grewal, I. S.,

17 Grewal, I. S., Xu, J. and Flavell, R., Nature 1995. 378: 617. 18 Stiiber, E., Strober, W. and Neurath, M., J. Exp. Med. 1996.

183: 693. 19 Cella, M., Scheidegger, D., Palmer-Lehmann, K., Lane, P.,

Lanzavecchia, A. and Alber, G., J. Exp. Med. 1996.184: 747. 20 Wagner, D. H., Stout, R. D. and Suttles, J., Eur. J . Immunol.

1994.24: 3148. 21 Kennedy, M. K., Picha, K. S., Fanslow, W. C., Grabstein, K.

H., Alderson, M. R., Clifford, K. N., Chin, W. A. and Mohler, K. M., Eur. J . Immunol. 1996.26: 370.

22 Niisslein, H. G., Frosch, K., Woith, W., Lane, P., Kalden, J. R. and Manger, B., Eur. J. Immunol. 1996. 26: 846.

23 Moore, A. R. and Willoughby, D. A., Clin. Exp. Immunol. 1995.101: 387.

24 Kanof, M. E. and Smith, P. D., in Coligan, J. E., Kruisbeck, A. M., Margulies, D. H., Shevach, E. M. and Strober, W. (Eds.), Current Protocols in Immunology, Greene Publishing and Wiley-Interscience, New York 1995. p. 7.1.1.

25 Romano, M. F., Lamberti, A., Bisogni, R., Tassone, P., For- misano, S., Venuta, S. andTurco, M. C., J. Immunol. 1996. 156: 2119.

26 Chomczynski, P. and Sacchi, N., Anal. Biochem. 1987. 162: 156.

27 Sambrook, J., Fritsch, E. E and Maniatis, T., Molecular Clon- ing: A laboratory manual, Cold Spring Harbor Laboratory Press, New York 1989.

28 Mpez-Boado, Y. S., Tolivia, J. and L6pez-Otin, C., J. Biol. Chem. 1994.269: 26871.

29 Nonoyama, S., Penix, L. A., Edwards, C. P., Lewis, D. B., Ito, S., Aruffo, A., Wilson, C. B. and Ochs, H. D., J. Clin. Invest. 1995. 95: 66.

30 Karrnann, K., Hughes, C. C. W., Fanslow, W. C. and Pober, J. S., Eur. J. Immunol. 1996. 26: 610.

31 Fuleihan, R., Ramesh, N., Horner, A., Ahern, D., Belshaw, P. J., Alberg, D. G., Stamenkovig, I., Harmon, W. and Geha, R. S., J. Clin. Invest. 1994. 93: 1315.

32 Zubiaga, A., Muiioz, E. and Huber, B. T., J. Immunol. 1991. 146: 3857.

33 Rao, A., Immunol. Today 1994.15: 274. 34 Hettmann,T. and Cohen, A., Eur. J. Immunol. 1995.25: 3356. 35 Yaseen, N. R., Maizel, A. L., Wang, F. and Sharma, S., J.

36 Schreiber, S. L. and Crabtree, G. R., Immunol. Today 1992.

37 Liu, J., Immunol. Today 1993.14: 290. 38 Fruman, D. A., Bierer, B. E., Benes, J. E., Burakoff, S. J.,

39 Tsitsikov, E. N., Ramesh, N. and Geha, R. S., Mol. Immunol.

40 Decker, C. J. and Parker, R., Trends Biochem. Sci. 1994. 19:

41 Graf, D., Korthauer, U., Mages, H. W., Senger, G. and Kro-

42 Xia, H., Kannapell, C. C., Fu, S. M. and Sung, S. J., Blood

43 Lindsten, T., June, C. H., Ledbetter, J. A., Stella, G. and

44 Fedyk, E. R., Adawi, A., Looney, R. J. and Phipps, R. P.,

45 Haraguchi, S., Good, R. A. and Day, N. K., Immunol. Today

46 Tamir, A., Granot, Y. and Isakov, N., J. Immunol. 1996. 1.57:

Exp. Med. 1996. 184: 955.

Flavell, R. and Liu, Y., J. Imrnunol. 1996. 1.57: 2764.

Biol. Chem. 1993.268: 14285.

13: 136.

Austen, K. E and Katz, H. R., J. Immunol. 1995. 154: 1846.

1994. 31: 895.

336.

czek, R. A., Eur. J. Immunol. 1992. 22: 3191.

1993. 82: 2806.

Thompson, C. B., Science 1989.244: 339.

Clin. Immunol. Immunopathol. 1996. 81: 101.

1995.16: 595.

1514.