Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Eosinophilia is a feature of allergic conditions, including asthma, and parasite infestation. The number of eosinophils has been correlated with the severity of disease, and epithelial damage caused by the release of eosinophil-derived products is thought to play a major role in the pathogenesis of bronchial hyperresponsiveness (BHR) (1). Human interleukin-5 (IL-5) is a cytokine that selectively induces the proliferation and differentiation of eosinophils (4, 5). IL-5 messenger RNA (mRNA) has been detected in the bronchial mucosa of asthmatic patients with eosinophilia but not in those patients without eosinophil infiltration or in normal controls (6). Recent work with double immunocytochemistry and in situ hybridization has revealed that the primary source of IL-5 in bronchial biopsies and bronchoalveolar lavage cells in patients with atopic asthma is activated T cells, representing more than 70% of positively staining cells (7). In animal models of pulmonary hyperresponsiveness, injection of monoclonal antibodies directed against IL-5 completely suppressed blood and tissue eosinophilia and BHR seen in response to allergen challenge (8, 9). In mice in which the IL-5 genes were inactivated, pulmonary eosinophilia in response to challenge with inhaled antigen was completely abolished (10).

Cyclosporin A (CsA) and FK506 are potent immunosuppressive agents that selectively inhibit calcium-associated signaling pathways regulating T-lymphocyte activation and cytokine gene expression (11). Treating guinea pigs with CsA and FK506 completely inhibited antigen- induced BHR and infiltration of eosinophils into the bronchial tissues (12, 13). CsA has also been shown to be effective in alleviating the clinical symptoms and lung function in cases of chronic severe asthma, as well as in glucocorticoid-resistant cases of the disease (14, 15). In a study using in situ hybridization, CsA and FK506 were found at the single-cell level to inhibit IL-5 mRNA expression (16). It has also been shown that CsA and FK506 inhibit IL-5 mRNA expression in peripheral blood mononuclear cells (PBMC) and CD4+ T-lymphocytes of patients with atopic asthma and normal controls (17).

Correlations between the induction of cytokine mRNA expression in activated T-lymphocytes and the secreted cytokine bioactivities have been observed, suggesting that mechanisms regulating mRNA accumulation are very important in the control of cytokine synthesis. The ultimate steady-state level of human IL-5 mRNA depends on the interplay of various factors influencing its rate of production and degradation. We investigated the effects of the immunosuppressive agents CsA and FK506 on the regulation of IL-5 mRNA expression in primary human T-lymphocytes and in the T-cell line HSB-2. A novel polymerase chain reaction (PCR)-based nuclear run-on assay was used to study the rate of IL-5 transcription, and a quantitative IL-5 reverse transcriptase-polymerase chain reaction (RT-PCR) was used to monitor the decay of IL-5 mRNA expression. In this present study, CsA and FK506 inhibited IL-5 transcription, providing a potential mechanism for their effectiveness in diseases with prominent eosinophilia.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Reagents and Cell Culture

Cells of the human T lymphoblastic leukemia line CCRF-HSB-2 (HSB-2) were obtained from the American Type Culture Collection (No. CCL 120.1; ATCC, Rockville, MD) and maintained in complete medium consisting of RPMI 1640 (Gibco-BRL, Gaithersburg, MD) supplemented with 10% (vol/vol) heat-inactivated fetal calf serum (FCS) (CSL, Melbourne, Australia), 2 mM glutamine (Gibco-BRL), 100 U/ml penicillin, 100 µg/ml streptomycin (Gibco-BRL), and 20 mM 4-(2-hydroxyethyl)-1-piperazine-N'-2-ethanesulfonic acid (HEPES) (CSL). PBMC were isolated from the blood of healthy volunteers through Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density-gradient centrifugation. Aliquots of PBMC (1 × 10⁶ cells/ml) were stimulated with PHA (Burroughs-Wellcome, Dartford, UK) (1 µg/ml) on day 0, diluted 5 times with complete medium supplemented with recombinant IL-2 (rIL-2) (Boehringer Mannheim, Germany) (5 U/ml) on day 3, and maintained in culture for an additional 4 days. A 10-fold increase in cell numbers was observed over the 7-day period. These cells will be referred to as 7-day cultured PBMC, and represent a cell population of which 99% consists of T cells as demonstrated by CD3 cell-surface phenotyping.

HSB-2 cells or 7-day cultured PBMC were stimulated with PMA (Sigma-Aldrich, St. Louis, MO) (4 ng/ml) and/ or ionomycin (Calbiochem, La Jolla, CA) (1 µg/ml). PBMC and 7-day cultured PBMC were also stimulated with PHA (1 µg/ml) or with rIL-2 (10 U/ml). Immediately prior to activation, aliquots of cells were exposed to CsA (a gift from Sandoz, Inc., East Hanover, NJ) or FK506 (a gift from Fujisawa Pharmaceutical Co. Ltd., Osaka, Japan). Unless otherwise stated, CsA was used at 1,000 ng/ ml and FK506 at 100 ng/ml. Ethanol was used to dissolve CsA and methanol to dissolve FK506. In control experiments, dilutions of ethanol or methanol equivalent to the concentrations in the specimens had no effect on IL-5 mRNA expression in PBMC or 7-day cultured PBMC (data not shown).

RNA Extraction and RT-PCR

RNA extraction, complementary DNA (cDNA) synthesis, PCR amplification, and detection of PCR products by hybridization were performed as previously described (18). Cells were lysed in 4 M guanidinium isothiocyanate, 25 mM sodium citrate (pH 7), 0.5% sarcosyl and 0.1 M 2-mercaptoethanol (solution D) and 20 µg of yeast transfer RNA (tRNA) was added, and total cellular RNA was extracted as described (19). Each PCR was performed on 4 × 10⁵ cell equivalents of cDNA. The primers were IL-5 sense: 5'-CGGATCCCACAGAAATTCCCACAA-3';  antisense: 5'-TGATATCCACTCGGTGTTCATTAC-3'; and -actin sense: 5'-TCACCAACTGGGACGACATG-3'; antisense: 5'-GTACAGGGATAGCACAGCCT-3'. PCR conditions were optimized to ensure that products were still accumulating exponentially when the reactions were terminated. Oligonucleotide primers were designed to bind to exons of IL-5 or -actin and to span an intron, so that any heteronuclear RNA, or genomic DNA copurified with the RNA, would yield a larger PCR product. The PCR products were size-fractionated by electrophoresis in 1.2% agarose, and detected either by staining with ethidium bromide or by Southern transfer to Hybond-N+ nylon membrane (Amersham, Amersham, UK), hybridization, and autoradiography at 70°C using Kodak X-ray film (Eastman Kodak, Rochester, NY). The hybridization primers were IL-5: 5'-CTTCAGTGCACAGTTGGTGAT-3'; and -actin: 5'-CAGCCATGTACGTTGCTATC-3', and were designed to bind to a region of cDNA between the two amplification primers. The intensities of the bands from the autoradiographs or negatives were quantitated with scanning densitometry (LKB densitometer; Uppsala, Sweden), or were scanned and analyzed with the NIH Image program (version 1.55; Wayne Rasband, National Institutes of Health, Bethesda, MD).

PCR-based Nuclear Run-on Assay

Cells were harvested by centrifugation at 400 × g for 5 min at 4°C, washed three times with ice-cold phosphate-buffered saline (PBS), and resuspended in 5 ml of ice-cold lysis buffer (3 mM MgCl₂, 1 mM KCl, 10 mM Tris-Cl, pH 7.4) containing 0.3% NP-40 (Sigma). Cells were incubated at room temperature until lysed, and were then layered on 10 ml of sucrose cushion (30% sucrose in lysis buffer containing no NP-40). Following centrifugation at 600 × g for 10 min at 4°C, the nuclear pellet was resuspended in 100 µl of nuclear storage buffer (50 mM Tris-Cl, pH 8.3; 40% glycerol; 5 mM MgCl₂; 0.1 mM ethylene diamine tetraacetic acid [EDTA]). Nuclei were either stored in liquid nitrogen or used immediately. Fresh or thawed nuclei (200 µl) were split into two aliquots and incubated for 30 min at 30°C in 20% glycerol; 30 mM Tris-Cl, pH 8.0; 2.5 mM MgCl₂; 150 mM KCl; 1 mM dithiothreitol (DTT); and 40 U of ribonuclease inhibitor (RNasin) (Promega, Madison, WI), with or without 0.5 mM each of rATP, rCTP, rGTP and rUTP. No ribonucleotide triphosphates (rNTPs) were added to the second aliquot, preventing any transcription from occurring. Transcription was terminated by lysing the nuclei with 200 µl of Solution D (19). Samples were then subjected to RT-PCR as described earlier.

Quantitative RT-PCR and Studies of the Rate of IL-5 mRNA Degradation

Quantitative RT-PCR and studies of the rate of IL-5 mRNA degradation were performed as described (20, 21). Plasmid pSPAG2 (kindly provided by A. Guiffre, Haematology Dept., St. Vincent's Hospital, Sydney, Australia) was derived from pSP64 (poly A), and contains two 81-bp oligonucleotides with the binding sequences for the human IL-5 amplification primers. The competitor RNA (cRNA) was generated by performing a transcription reaction as described. The resultant cRNA was separated from the template DNA through band interception (20) and quantitated by absorbance at 260 nm.

cDNA synthesis was done in a final volume of 48 µl, which contained 10 µl of RNA consisting of 5 µl of total cellular RNA (2 × 10⁵ cell equivalents) and an equal volume of known numbers of cRNA molecules. cDNA was then subjected to PCR for 35 cycles. Amplification of IL-5 cRNA and IL-5 mRNA yields PCR products of 103 bp and 320 bp, respectively. PCR products were separated by agarose-gel electrophoresis, transferred to Hybond-N+ nylon membrane, and hybridized as described (18). The sequence of the cRNA hybridization primer was as described (20).

Laser densitometry was used to calculate the signal strength of both IL-5 mRNA and IL-5 cRNA bands. The ratios of IL-5 cRNA to IL-5 mRNA were calculated for each point. A graph of the number of cRNA molecules against the cRNA:mRNA ratio was generated, the line of best fit was determined by linear regression analysis, and the point at which the ratio was equal to 1 was calculated. Because the number of cRNA molecules is known, the number of IL-5 mRNA molecules in the samples can be obtained. To determine RNA stability, the number of IL-5 RNA molecules was plotted against the time in hours, and mRNA half-life was calculated through linear regression analysis. Values were expressed as the mean and SE of triplicate experiments. All statistical analyses were done with Student's paired t test. A P value < 0.05 was considered statistically significant.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

CsA and FK506 Reduce IL-5 mRNA Abundance

The effect of CsA on IL-5 mRNA accumulation was investigated with RT-PCR on total cellular RNA extracted from human PBMC and 7-day cultured PBMC. In a trial of the clinical efficacy of CsA in patients with chronic glucocorticoid-dependent asthma, whole-blood concentrations of 151 ng/ml were obtained (22). In the present experiments, cells were stimulated with PHA in the absence or presence of increasing concentrations of CsA ranging from 150 to 1,000 ng/ml. As shown in Figure 1, no IL-5 expression was detectable in unstimulated PBMC or 7-day cultured PBMC (Lanes a). IL-5 was strongly induced in the PHA-stimulated cells (Lanes b). CsA at 1,000 ng/ml completely inhibited IL-5 expression, with substantial inhibition at 150 to 600 ng/ml (Figure 1A, Lanes c through f). There was no appreciable difference in the amount of -actin mRNA in these samples, demonstrating comparable RNA content in each sample. Similar experiments were performed with FK506. In various clinical trials, FK506 has been reported to be approximately 20- to 100-fold more potent than CsA. PBMC or 7-day cultured PBMC were stimulated with PHA with or without various concentrations of FK506. In the presence of 100 ng/ml FK506, IL-5 mRNA expression was completely inhibited (Figure 1B, Lane f). PBMC were more sensitive than 7-day cultured PBMC to FK506, with complete inhibition of IL-5 mRNA accumulation at 5 ng/ml in PBMC and 10 ng/ ml in 7-day cultured PBMC.

View larger version (39K): [in this window] [in a new window]   Figure 1.   Effect of CsA and FK506 on induction of IL-5 mRNA accumulation. Cells (2 × 106/ml) were exposed to medium alone (Lanes a) or to PHA (1 µg/ml; Lanes b through f) in the absence (Lanes b) or presence of CsA (A: Lanes c through f) or FK506 (B: Lanes c through f) for 20 h. CsA concentrations were: Lane C, 150 ng/ml; Lane d, 300 ng/ml; Lane e, 600 ng/ml; Lane f, 1,000 ng/ml. FK506 concentrations were: Lane c, 1 ng/ml; Lane d, 5 ng/ ml; Lane e, 10 ng/ml; Lane f, 100 ng/ml. Total cellular RNA was extracted and IL-5 and -actin mRNA accumulation analyzed with RT-PCR, gel electrophoresis, and Southern hybridization. Results shown are representative of two independent experiments.

To examine the signal requirements for inhibition of IL-5 mRNA expression by CsA and FK506, the effects of stimulating T-lymphocytes with different activators were investigated. IL-5 mRNA accumulation was induced in 7-day cultured PBMC with PHA; PMA and ionomycin; PMA alone; ionomycin alone; or rIL-2. No IL-5 mRNA was detectable in the unstimulated 7-day cultured PBMC or in the cells cultured with CsA or FK506 alone. CsA inhibited IL-5 mRNA expression in response to PHA; PMA and ionomycin; PMA; or ionomycin (Figure 2A). In contrast, IL-5 expression induced by rIL-2 was resistant to inhibition by CsA, as previously observed (44). Similar findings were obtained with FK506 (Figure 2B). As with CsA, induction of IL-5 mRNA by rIL-2 was not inhibited by FK506. The effects of CsA and FK506 on IL-5 mRNA were confirmed with quantitative RT-PCR. In both 7-day cultured PBMC and in the T-cell line HSB-2, CsA or FK506 decreased the amount of IL-5 mRNA induced by PMA and ionomycin stimulation by 72 to 99% (Table 1).

View larger version (28K): [in this window] [in a new window]   Figure 2.   Effect of CsA and FK506 on induction of IL-5 mRNA via different signaling pathways. Seven-day cultured PBMC (2 × 106 cells/ml) were exposed to medium alone or stimuli as shown in the presence (+) or absence () of CsA (A) or FK506 (B) for 20 h. IL-5 mRNA was detected as described in Figure 1, except that Southern hybridization was not performed. Results shown are representative of two independent experiments.

CsA and FK506 Inhibit the Rate of IL-5 Transcription in 7-day Cultured PBMC and HSB-2 Cells

Nuclear run-on assays were performed to determine whether the decrease in total cellular levels of IL-5 mRNA in response to CsA and FK506 was attributable to a decrease in the rate of IL-5 transcription. Conventional run-on assays were done, but IL-5 transcription was not detected. Therefore, a novel, more sensitive assay was developed, in which nuclei are isolated from activated cells and samples are incubated with or without rNTPs, described as positive (+) and negative () transcription reactions, respectively. RNA is then extracted and subjected to RT- PCR. The difference between the two samples is a measure of gene transcription (21). For this assay, 7-day cultured PBMC or HSB-2 were stimulated for 3 h in the presence of PMA and ionomycin. IL-5 transcripts were readily detected in both cell populations (Figure 3, P + I, Lanes marked "+ transcription reaction"). No signals were detected in the negative control reactions lacking rNTPs (Figure 3, Lanes marked " transcription reaction"). CsA and FK506 inhibited IL-5 transcription in both cell types; transcription was abolished by CsA and markedly reduced by FK506 (Figure 3). In addition to the transcription reactions, cellular RNA was extracted from aliquots of cells removed at the time the nuclei were isolated, in order to determine total cellular IL-5 mRNA levels. The effects of CsA or FK506 on transcription were similar to their effects on total cellular IL-5 mRNA levels (Figure 3). CsA and FK506 did not inhibit transcription or total cellular RNA levels for -actin (Figure 3).

View larger version (29K): [in this window] [in a new window]   Figure 3.   CsA and FK506 inhibit the rate of IL-5 transcription. Nuclei were isolated from 7-day cultured PBMC (A) or HSB-2 cells (B) that had been stimulated for 3 h with PMA and ionomycin alone (P + I), with CsA (P + I + CsA), or with FK506 (P + I + FK506). Transcription reactions were run in the presence (+) or absence () of rNTPs, and IL-5 transcripts were detected with RT-PCR and Southern hybridization. Total cellular RNA was isolated from aliquots of these cells at the time the nuclei were isolated, and was assessed with RT-PCR and Southern hybridization. Results shown are representative of three independent experiments.

HSB-2 cells were stimulated with PMA and ionomycin for 0, 2, and 4 h, and nuclei were harvested and transcription reactions performed (Figure 4, left six tracks). No IL-5 transcripts were detected at 0 h. At 2 h, transcripts were detected in the + transcription reaction tracks, but little or no signal was present in the  transcription reaction tracks, indicating active transcription at this time. At 4 h, the signals in the + transcription reaction tracks were not greatly stronger than in the corresponding  transcription reaction tracks, indicating the presence of complete transcripts when the nuclei were harvested, and reduced generation of new transcripts. CsA (Figure 4A, center six tracks) virtually abolished IL-5 transcription at the 2- and 4-h time points. CsA also completely inhibited the signal for total IL-5 mRNA (Figure 4A, right six tracks). Likewise, FK506 inhibited the generation of new IL-5 transcripts at 2 and 4 h (Figure 4B, center six tracks), and it had a similar effect on total cellular RNA (Figure 4B, right six tracks).

View larger version (42K): [in this window] [in a new window]   Figure 4.   Time course of the effect of CsA and FK506 on the rate of IL-5 transcription. Nuclei were isolated from HSB-2 cells stimulated with PMA and ionomycin in the presence of CsA (A) or FK506 (B). Cells were cultured for 0, 2, and 4 h without inhibitors (left six tracks) or with inhibitors (center six tracks), and transcription reactions were run in the presence (+) or absence () of nucleotides. Total cellular IL-5 mRNA was also assessed (right six tracks). IL-5 transcripts and mRNA were detected with RT-PCR and Southern hybridization. Results shown are representative of two experiments.

CsA and FK506 Do Not Affect IL-5 mRNA Stability

The total abundance of any mRNA species is determined by both its rate of production and its rate of decay. In studies of mRNA stability, inhibitors of transcription are commonly used to abolish the production of new mRNA, so that the subsequent reduction in total mRNA abundance is a measure of the rate of mRNA decay. However, transcription inhibitors may themselves interfere with the regulation of mRNA degradation (23). In the case of human IL-5, transcription occurs only in the first few hours after activation. By 16 h, when IL-5 mRNA is still readily detectable, transcription has ceased (21). Therefore, studies of IL-5 mRNA degradation that commence at this time may be performed without the use of transcription inhibitors. In the present study, HSB-2 cells were cultured for 16 h with PMA and ionomycin in the presence or absence of either CsA or FK506, and RNA was then extracted 0 (equivalent to 16 h poststimulation), 4, 8, and 24 h later. -actin mRNA was assessed as an internal control to check the integrity of the RNA for each time point. IL-5 expression was analyzed with quantitative RT-PCR. The half-life of IL-5 mRNA in the cells was 2.60 ± 0.33 (mean ± SE) h, which was not significantly altered by CsA or FK506 (Figure 5; Table 2).

View larger version (14K): [in this window] [in a new window]   Figure 5.   Effect of CsA and FK506 on IL-5 mRNA stability in HSB-2 cells. Cells (2 × 106 cells/ml) were cultured with PMA and ionomycin alone (P + I), with CsA (P + I + CsA), or with FK506 (P + I + FK506) for 16 h. Cells were then harvested 0, 4, 8, and 24 h later. IL-5 mRNA abundance was analyzed with quantitative IL-5 RT-PCR. Data are presented as the mean ± SE of triplicate experiments.

View larger version (21K): [in this window] [in a new window]   Figure 6.   Effect of CsA and FK506 on IL-5 mRNA stability in 7-day cultured PBMC. Cells were stimulated with PMA and ionomycin (P + I) (A) or PHA (B) in the presence or absence of CsA or FK506 as described in Figure 5. Results are presented as the mean ± SE of triplicate experiments.

Similar experiments were performed on primary human T cells. In these experiments, 7-day cultured PBMC were activated with PMA and ionomycin for 16 h in the presence or absence of CsA or FK506. In the absence of inhibitors, the half-life of IL-5 mRNA was 2.24 ± 0.07 h, which was not significantly altered by CsA or FK506 (Figure 6A; Table 2). Experiments were also performed on 7-day cultured PBMC activated with PHA instead of PMA and ionomycin. After activation with PHA, the IL-5 mRNA half-life was 5.03 ± 0.47 h, and again CsA or FK506 did not significantly alter the stability of IL-5 mRNA (Figure 6B; Table 2).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The principal findings in this study were that CsA and FK506 markedly inhibit IL-5 gene transcription (Figures 3 and 4), but do not detectably decrease the stability of IL-5 mRNA (Figures 5 and 6; Table 2). The effect of CsA and FK506 on gene transcription was similar to their effect on total mRNA abundance (Figures 1 and 2; Table 1). The findings for total mRNA abundance are consistent with previous reports that CsA inhibits human IL-5 mRNA expression (16, 17). The results indicate that the reduction in total IL-5 mRNA levels induced by CsA and FK506 can be explained by inhibition of gene transcription. These experiments were performed on cells freshly derived from human peripheral blood and on the human T-leukemia cell line HSB-2. Because both cell populations gave similar results, it is unlikely that the findings with the primary cells were dependent on the presence of cells other than T cells.

Because IL-5 transcription could not be detected with conventional run-on assays, we developed a novel assay in which transcription in isolated nuclei was distinguished from RNA formed before the nuclei were isolated by running each sample with and without rNTPs (+ and  transcription reaction, respectively) (21). The amount of ongoing transcription at the time of nuclear harvest is represented by the difference between the two conditions. PCRs were terminated when products in the + transcription reaction samples were still rapidly accumulating, providing a clearer distinction between the + and  tracks. However, when the PCR was extended for several more cycles, -actin was always found in the  transcription reaction samples, as was IL-5 except in unstimulated cells. In isolated nuclei, the splicing machinery is inefficient, resulting in the predominance of unspliced RNA. However, with the novel method, the size of the PCR products was consistent with derivation from spliced mRNA. In isolated nuclei, any transcription in + transcription reaction samples may provide fragments that act as primers on preformed mature transcripts. Alternatively, splicing may not be completely absent in isolated nuclei. In either case, the PCR conditions would favor amplification of completely spliced transcripts.

The effects of CsA and FK506 on IL-5 gene transcription (Figures 3 and 4) are similar to their effects on transcription of the genes for IL-2, IL-3, IL-4 and granulocyte- macrophage colony-stimulating factor (GM-CSF) (24). In T-cell activation, the calcium- and calmodulin-dependent phosphatase calcineurin dephosphorylates the cytoplasmic component of nuclear factor of activated T cells (NF-AT), which translocates to the nucleus, associates with the nuclear component of NF-AT, and binds to regulatory regions of cytokine genes to stimulate their transcription (27, 28). CsA and FK506 inhibit this process by forming complexes with intracellular proteins and inhibiting calcineurin (11, 29, 30). Thus, CsA and FK506 inhibit the effects of increased intracellular calcium ion concentration. NF-AT is involved in the transcriptional induction of several cytokines produced by activated T cells, including IL-2, IL-3, IL-4, and GM-CSF (24, 26, 31, 32). Characterization of the DNA-binding proteins interacting with the IL-5P motif in the murine IL-5 gene promoter has revealed the presence of NF-AT-related proteins (33). The human IL-5 gene promoter, although not as extensively characterized as its murine counterpart, contains an IL-5P motif that is highly homologous to the murine sequence. The results of the present study provide further evidence that transcriptional regulation of the human IL-5 gene may involve an NF-AT component.

In murine cells, CsA has been reported to inhibit IL-5 mRNA (34) or to have no effect (35). Some of the cells in the latter study were maintained in IL-2. It is therefore possible that the IL-2 pathway, which is resistant to the effects of CsA (Figure 2), contributed to the IL-5 expression observed in the present study. IL-2 induces several signaling pathways, including phosphorylation of the Janus kinase (JAK) signal transduction and activation of transcription (STAT) signaling pathway, which is not activated by the T-cell receptor (36). Cytokine expression induced by IL-2 is not associated with the increases in intracellular calcium levels that follow activation of the T-cell receptor (37). The lack of calcium dependence of the IL-2 signaling pathway is likely to explain the IL-2-induced resistance of IL-5 mRNA accumulation to CsA and FK506 (Figure 2).

IL-5 mRNA expression in 7-day cultured PBMC, a population of activated T-lymphocytes, is relatively resistant to the inhibitory effects of FK506 as compared with PBMC (Figure 1). In another study, the activation state of T-lymphocytes affected their susceptibility to inhibition by FK506. Membrane-bound and soluble IL-2 receptor expression was inhibited on resting but not activated human T-lymphocytes (38). Resistance to FK506 has been associated with a deficiency in expression of the FK506 binding protein FKBP12 in murine bone marrow-derived mast cells (39). Activation of T-lymphocytes may be associated with decreased expression of FKBP12 or an altered cellular ratio between the different FK506 binding proteins, resulting in reduced susceptibility of cytokine gene expression to inhibition by FK506 in the 7-day cultured PBMC.

Apart from effects on transcription, acceleration of mRNA degradation is another possible mechanism whereby CsA and FK506 might decrease cytokine mRNA abundance. However, we found that neither CsA nor FK506 accelerated the degradation of IL-5 mRNA (Figures 5 and 6). By contrast, FK506 decreased the stability of IL-2 and GM-CSF mRNA in human PBMC and in Jurkat cells (40), and in a mast cell line transfected with genomic regulatory sequences for IL-3, both endogenous and exogenous IL-3 mRNA are destabilized by CsA (41).

Stability of an mRNA species is governed by several factors, including AU rich motifs in the 3' untranslated region, as yet unidentified sequences elsewhere in the gene, and the cellular expression of regulatory proteins. IL-5 mRNA contains fewer AU rich motifs in its 3' untranslated region than the mRNAs of other cytokines, which may contribute to the lack of a destabilizing effect of CsA and FK506 on IL-5 mRNA. Studies of the mRNA for GM-CSF have revealed that sequences other than the AU rich motifs are also involved in regulating mRNA stability (42, 43). The IL-5 gene may lack these response elements, thus preventing CsA or FK506 from interacting with stabilizing/destabilizing proteins. Alternatively, the expression of particular regulatory proteins may be cell specific, and this may explain the different reported effects of CsA and FK506 on the various cytokines. Further studies of the effects of CsA and FK506 on the stability of other cytokine mRNAs in various cell types will provide useful information about the regulation of cytokine mRNA degradation by these agents.