Differential expression of PDE4 cAMP phosphodiesterase isoforms in inflammatory cells of smokers with COPD, smokers without COPD, and nonsmokers

Rachael Barber, George S. Baillie, Reinhard Bergmann, Malcolm C. Shepherd, Ruth Sepper, Miles D. Houslay, Gino Van Heeke


The expression profile of a panel of 15 cAMP phosphodiesterase isoforms was determined for inflammatory cell types of relevance to chronic obstructive pulmonary disease (COPD). In particular, the expression profiles for bronchoalveolar macrophages, peripheral blood monocytes, T lymphocytes, and neutrophils from smokers with and without COPD were compared. The phosphodiesterase expression profile was also analyzed for peripheral blood monocytes, T lymphocytes, and neutrophils from nonsmokers and compared with smokers. Qualitative RT-PCR identified transcripts for PDE4A10, PDE4A7, PDE4B1, PDE4B2, PDE4D1, and PDE4D2 isoforms as well as transcripts for both PDE3B and PDE7A in T cells, monocytes, and macrophages in all subjects. Transcripts for PDE4B3 and PDE4D4 were not observed in any of the cell types investigated. PDE4C was detected in all cells analyzed except for T cells. The long PDE4A4, PDE4D3, and PDE4D5 isoforms exhibited cell type-specific expression patterns. Semiquantitative and real-time quantitative RT-PCR were used to analyze differential expression between disease states and between cell types. PDE4A4 was found significantly upregulated in lung macrophages from smokers with COPD when compared with control smokers. Furthermore, PDE4A4 as well as PDE4B2 transcripts were detected in higher amounts in peripheral blood monocytes of smokers when compared with nonsmokers. Finally, PDE4D5 and PDE4C were differentially regulated in lung macrophages when compared with monocytes of the same subjects, irrespective of the disease state. The data obtained suggest that PDE4A4 may be relevant as a macrophage-specific anti-inflammatory target for COPD.

  • cAMP-specific phosphodiesterase family type 4
  • chronic obstructive pulmonary disease
  • inflammation

chronic obstructive pulmonary disease (COPD) is an inflammatory airway disease characterized by airflow obstruction and an irreversible decline in lung function. Smoking is the primary factor causing the disease (1). Inflammation in COPD is marked by the presence of neutrophils, macrophages, and CD8+ T cells, which are thought to play a major role in the pathogenesis of COPD (1, 20).

There is a paucity of effective therapies for COPD (10). In contrast to asthma, the inflammatory response in COPD is largely insensitive to corticosteroid treatment (1). There has been considerable interest in cAMP-specific phosphodiesterase family type 4 (PDE4) as a therapeutic target for COPD (5, 10, 18, 35). Promising clinical data have been generated using both Cilomilast (Ariflo) (8, 40) and Roflumilast (BY-217) (10). The development of PDE4 inhibitors has been based on a large body of evidence from many laboratories that show such compounds can exert potent anti-inflammatory actions on a wide range of cells in vitro and in a range of animal models in vivo (4, 13, 39). However, one factor constraining the therapeutic use of many PDE4 inhibitors is their ability to cause nausea, which has placed limitations on the dosing levels that can be employed and hence their anti-inflammatory efficacy (5, 10, 18, 35). Furthermore, previous reports demonstrated how nonselective PDE4 inhibition, using the prototypic inhibitor rolipram, has the potential to lead to a diminished host defense response in animal models (2).

Currently, 11 different PDE families have been described, representing more than 50 different cAMP and cGMP PDE variants (36). PDE4 activity alone is provided by a large family of isoforms that are encoded by four genes (PDE4A, -4B, -4C, and -4D) (6, 15, 16). Each isoform is characterized by a unique NH2-terminal region, which is often involved in their intracellular targeting and may also serve to regulate their catalytic activity (16). These isoforms fall into three separate categories: the so-called “long” isoforms, which have the upstream conserved region (UCR)1 and UCR2 regulatory modules linking the isoform-specific NH2-terminal regions to the catalytic unit; the “short” isoforms, which lack UCR1; and the “supershort” isoforms that both lack UCR1 and have a truncated UCR2. These UCR modules can interact with each other and also serve to influence the functioning of the PDE4 catalytic unit where they orchestrate the functional consequence of phosphorylation of the PDE4 catalytic unit by extracellular signal-regulated kinase and of phosphorylation of UCR1 by cAMP-dependent protein kinase (PKA) (7, 15, 16). These differentially targeted and regulated isoforms are expressed in a cell type-specific fashion (15, 16), implying that individual PDE4 isoforms play specific functional roles. Indeed, differences in the functional role of isoforms have been implied by antisense strategies (28) and by selective inhibition (27, 29), and, additionally, gene knockout studies have shown that the PDE4B and PDE4D gene families can perform specific roles (12, 21).

Tenor et al. (37) have described cAMP PDE enzymatic activities in human alveolar macrophages from normal and atopic asthmatic donors and concluded that there was no difference between disease states. The expression of PDE4A, PDE4B2, PDE4C, and PDE4D has previously been analyzed for peripheral blood leukocytes of normal and atopic subjects (9), whereas more recently, the identification of PDE4A, PDE4B, and PDE4D subtypes in CD4+ and CD8+ lymphocytes from healthy and asthmatic subjects has been noted (23).

To date, no studies have been reported describing the distribution of PDE4 isoforms in inflammatory cells of patients with COPD. An ability to define which PDE4 isoforms and splice variants are expressed in inflammatory cells and whether there is a differential regulation in COPD vs. non-COPD as well as in smokers vs. nonsmokers is of great interest as it may contribute to our understanding of the basic molecular pathology of this respiratory disease. Such analyses may also pinpoint particular PDE4 variants as possible new therapeutic targets with less potential for dose-limiting side effects or potential host defense impairment.

This study was designed to investigate in detail the expression profile of 12 known PDE4 subtypes and splice variants as well as of PDE7A, PDE3A, and PDE3B subtypes in inflammatory cell types of relevance to COPD. In particular, the expression profile was analyzed and compared for lung macrophages, peripheral blood monocytes, T lymphocytes, and neutrophils of smokers with and without COPD, as well as for peripheral blood monocytes, T lymphocytes, and neutrophils of nonsmokers.



CD4 and CD8 multisort kits, CD14 and CD16 microbeads, magnetic activated cell sorting (MACS) VS+ columns, MACS CS columns, VarioMACS glass stand, VarioMACS magnet, and the VarioMACS VS+ adaptor were all obtained from Miltenyi Biotech (Bergisch Gladbach, Germany). Ficoll-Paque research grade and first-strand cDNA synthesis kit were obtained from Amersham Pharmacia Biotech (Little Chalfont, Bucks, UK). Buffer RLT, QIAshredder, HotStarTaq 2× Master Mix kit, and RNeasy mini kit were obtained from Qiagen (Crawley, UK). Cell Strainer (100 μm) was obtained from Becton-Dickinson (Oxford, UK). RPMI 1640 with HEPES and PBS without Ca2+/Mg2+ were obtained from Invitrogen Life Technologies (Paisley, UK). Access RT-PCR kit, 100-bp DNA ladder, RQ1 RNase-free DNase, and Wizard PCR Preps DNA Purification System were purchased from Promega (Southampton, UK). MasterAmp PCR Optimization kit, MasterAmp 2× PCR PreMix H, and MasterAmp AmpliTherm DNA polymerase were from Epicentre Technologies (Cambio, Cambridge, UK). Primers were from Sigma-Genosys (Pampisford, Cambridgeshire, UK). Human whole brain total RNA was from Clontech (Basingstoke, UK). Square tissue culture plates (25 cm × 25 cm) and cell lifters/scrapers were from Fisher Scientific (Loughborough, UK). All other reagents were purchased from Sigma-Aldrich (St. Louis, MO).

Collection of clinical samples.

Clinical samples were obtained from volunteer subjects who had given their informed and written consent. Approval for the study was given by the local ethics committees (Guy’s Drug Research Unit and the Institute of Experimental and Clinical Medicine). The study conformed to the declaration of Helsinki.

Peripheral blood and bronchoalveolar lavage (BAL) samples were obtained from two groups (groups 1 and 2) of subjects at either the Guy's Drug Research Unit (London, UK) or the Institute of Experimental and Clinical Medicine (Tallinn, Estonia). Additional blood samples were taken from five nonsmoking subjects (group 3) from a volunteer donor panel at the Novartis Horsham Research Center. Group 1 consisted of 11 smokers diagnosed with mild to moderate COPD according to the Global Initiative for Chronic Obstructive Lung Disease criteria (11a). Group 2 was 10 smokers matched to the subjects in group 1 but diagnosed without COPD, as judged by a FEV1/FVC (forced expiratory volume in 1 s/forced vital capacity) ratio of >70%. The inclusion criteria for group 1 subjects were as follows: male or female COPD patients ages 45–60 years; nonatopic; current smokers (>5 pack-years, having smoked for >10 years and currently smoking >5 cigarettes per day); airflow limitation (FEV1 60–70% predicted for age and height, FEV1/ FVC ratio <70%) and with <10% reversibility to salbutamol; able to withhold bronchodilator therapy for at least 48 h; if female, using a medically acceptable form of contraception and having a negative result for pregnancy test before bronchoscopy, or be surgically sterile; and willing and able to give written informed consent.

The inclusion criteria for group 2 subjects were the same as for group 1 subjects, but they were diagnosed as not having any airflow limitation (FEV1/FVC >70% predicted for age and height) and with <10% reversibility to salbutamol.

The exclusion criteria for both groups were as follows: having taken regular inhaled or topical nasal corticosteroids in the preceding 3 mo; having received a course of prednisolone in the preceding 3 mo; having taken theophyllines within 14 days and other bronchodilators with 48 h of the scheduled bronchoscopy; having experienced an upper or lower respiratory tract infection within the preceding 4 wk; having chest X-ray appearances in previous month suggesting bilateral neoplasia, consolidation, or collapse; having an alcohol intake >28 units per week; participation in a clinical study within the last 3 mo; morbid obesity; and any clinically relevant history of serious cardiac or immunological disorder or other respiratory diseases as determined by the investigator. Table 1 summarizes the relevant clinical data for all subjects sampled for this study. Each subject from groups 1, 2, and 3 donated 100 ml of peripheral blood from which monocytes, T cells, and neutrophils were isolated. Subjects from groups 1 and 2 underwent a bronchoscopy with 150 ml of saline to provide BAL fluid (BALF) from which macrophages were isolated. Bronchoscopy was performed according to international guidelines (42), and BAL was obtained from the right middle lobe using three 50-ml aliquots of warmed saline. Patients fasted for 8 h before bronchoscopy. The procedure was performed under light sedation with 2.5–10 mg iv of midazolam together with 0.3–0.6 mg iv of atropine. Upper airway lignocaine topical anesthesia was applied. The dose did not exceed 6 mg/kg of lignocaine to avoid lignocaine toxicity from systemic absorption. Continuous administration of 2 l/min of supplemental oxygen was maintained throughout the procedure and for 60 min afterward. Nebulized salbutamol was administered as required after the procedure was completed to relieve bronchospasm. Lung function measurement was repeated once the patient had recovered from sedation and before discharge on the following day.

View this table:
Table 1.

Relevant clinical data for study groups 1 and 2

Isolation of macrophages from BALF.

BALF was collected as described above, chilled, kept on ice at all times, and filtered through 100-μm cell strainers into 50-ml Falcon tubes to remove epithelial cells and excess mucus. The cells were separated from the filtrate by centrifugation at 4°C for 10 min at 400 g. The supernatant fraction was poured off, and the cell pellet was washed by gentle resuspension in 40 ml of ice-cold Ca2+/Mg2+-free PBS, followed by another centrifugation at 4°C for 10 min at 400 g. The final cell pellet was resuspended in a total of 10 ml of ice-cold Ca2+/Mg2+-free PBS, and the total cell count was determined using a hemocytometer. The cells were pelleted again as above and resuspended in cold serum-free RPMI 1640 medium to not more than 1 × 105 cells per ml. One hundred milliliters of cell suspension were then plated onto 25-cm × 25-cm square tissue culture plates and incubated for 2 h at 37°C. After this incubation period, the nonadherent population of cells was aspirated and discarded. The adherent cells, which consisted of >95% alveolar macrophages, were gently washed with 20 ml of RPMI 1640 by tilting the plate, and the washings were aspirated again. This method yielded >95% of the total alveolar macrophages attached to the plates. An estimate of the macrophage cell number was made, and for RNA extraction, cells were lysed in 100 μl of RLT buffer for every 106 cells present, but using not less than 350 μl. A cell lifter was used to distribute and collect the lysate, which was thoroughly disrupted by pipetting up and down and flash-frozen at −70°C. For enzyme activity analysis, cells were lysed as described below.

Isolation of monocytes, T cells, and neutrophils from peripheral blood.

Blood samples were collected using either 50-ml Falcon tubes containing 2 ml of filter-sterile 3.8% wt/vol trisodium citrate or using 10-ml heparin vacutainer tubes. Peripheral blood mononuclear cells (PBMCs) were separated from granulocytes and red blood cells using a Ficoll gradient essentially according to the manufacturer's instructions. Briefly, the blood was diluted 1:1 with Ca2+/Mg2+-free PBS. Up to 25 ml of this mixture was layered carefully onto 15 ml of Ficoll-Paque in a 50-ml Falcon tube at room temperature. A Ficoll gradient was established by centrifugation at 400 g for 30 min at room temperature. The layer containing PBMC was recovered and transferred to fresh ice-cold Falcon tubes. The remaining serum and Ficoll was aspirated to leave the red cell pellet. T cells and monocytes were purified sequentially from the PBMC layer, whereas neutrophils were purified from the red cell pellet, as described below. Blood cells in the recovered PBMC layer were washed twice with ice-cold MACS buffer (PBS + 0.4% bovine serum albumin + 2 mM EDTA), counted, and resuspended in 80 μl of MACS buffer for every 107 cells present. To this, 20 μl of each CD4 and CD8 multisort MACS beads were added, and T cells were separated using a MACS VS+ column as per the manufacturer's instructions. Monocytes were purified from the first flow-through fraction of this procedure using CD14 MACS microbeads. Neutrophils were isolated from the red blood cell pellet generated by the Ficoll gradient. The cell pellet was drained, and red blood cells were selectively lysed twice by resuspension in cold sterile H2O for exactly 30 s, immediately followed by the addition of an equal volume of 2× PBS, before centrifugation. The remaining white blood cells were washed in 20 ml of MACS buffer and resuspended in a final volume of 50 μl of MACS buffer for every 5 × 107 cells present. Neutrophils were recovered by adding an equal volume of CD16 MACS microbeads followed by separation on a type C MACS column as described by the manufacturer. In all cases, the recovered cells were counted and centrifuged at 400 g for 5 min at 4°C, and the supernatant fraction was discarded. The recovered pellets were lysed in 60 μl of RLT buffer for every 106 cells present, but using not less than 350 μl. Samples were kept chilled at all times to minimize degradation. The lysate was thoroughly disrupted by pipetting up and down and flash-frozen at −70°C.

Extraction of total RNA and cDNA synthesis.

For all cell types isolated, frozen lysates were warmed at 37°C for 10 min and homogenized using QIAshredder columns (Qiagen) according to the manufacturer's instructions. RNA was extracted from the homogenized lysates using QIAgen RNeasy mini or midi kits following the kit instructions. The RNA was treated with RQ1 DNase (1 U/μl) to remove genomic DNA contamination. RNA (75 μl) was mixed with 20 μl of RQ1 and 5 μl of 20× RQ1 buffer (0.6 mM Tris·HCl, 72 mM MgCl2, 120 mM NaCl) and incubated at 37°C for 30–45 min. After this, the RNA was repurified using the “clean-up” protocol of the RNeasy mini kit. To check for successful genomic DNA removal, each sample was tested in RT-PCR for β-actin with and without reverse transcriptase using Promega Access RT-PCR kit.

Once samples were confirmed as DNA free, first-strand cDNA was synthesized using the Pharmacia first-strand cDNA synthesis kit. Five micrograms of RNA were diluted to 20 μl using nuclease-free water and then heated to 65°C for 10 min. To this, 14 μl of a mix containing random primers, oligo(dT), buffer, and enzyme (supplied with kit) were added. This was incubated for 1 h at 37°C followed by a 90°C incubation for 5 min to inactivate the reverse transcriptase. The completed reaction was diluted with nuclease-free water to 1 ml to allow addition of 5 μl (equivalent to the addition of 25 ng of RNA) for subsequent PCR.

Specific amplification by PCR and primer design.

First-strand cDNA was used as a template for PCR analysis of the following PDE isoforms and splice variants: PDE4A (generic) and splice variants PDE4A4, PDE4A10, and PDE4A7 (2EL); PDE4B (generic) and splice variants PDE4B1, PDE4B2, and PDE4B3; PDE4C (generic); PDE4D (generic) and splice variants PDE4D1, PDE4D2, PDE4D3, PDE4D4, and PDE4D5; PDE3A (generic); PDE3B (generic); and PDE7A (generic). PCR primers were designed to cover either a conserved region within a PDE isoform (generic) or the unique 5′ end of the mRNA to target a specific splice variant. The sequences used were obtained from the GenBank database, and GenBank accession numbers are shown in Table 2. Additional primers were designed to amplify β-actin and transferrin as amplification controls and for normalization purposes. Primers were designed using the Applied Biosystems software Primer Express and selected for their position within the sequence and calculated annealing temperature. Amplified PCR fragments were fractionated and visualized on a 2% agarose gel in 1× TAE buffer (40 mM Tris-acetate and 1 mM EDTA) containing 1 μg/ml of ethidium bromide. Sequences of PCR primers used are listed in Table 1. Amplification conditions and primer sequences were tested and optimized on first-strand cDNA synthesized from total brain RNA (Clontech). Reaction conditions were optimized to enable detection in as little as 25 ng of total RNA. Brain RNA similarly converted to cDNA served as positive control for all PCR reactions as all tested PDE variants were found to be present in brain tissue. PCR reactions were performed on each sample in duplicate or triplicate using the following conditions: 18.75 pmol of each forward and reverse primer; 12.5 μl of 2× PCR mix; 5 μl of the cDNA synthesis reaction equivalent to 25 ng of RNA; and nuclease-free water to a total volume of 25 μl. Amplifications were performed using either Epicentre 2× PCR mix H and 0.625 units of Epicentre MasterAmp Amplitherm DNA polymerase for amplification of PDE4A4, PDE4B3, and β-actin or Qiagen HotStarTaq Mastermix for all other reactions and subjected to the following program on a Biometra T3 Thermocycler: 95°C for 5 min; 40 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 1 min; a final extension of 72°C for 10 min; followed by pause at 4°C. Semiquantitative RT-PCR was performed in duplicate by halting the PCR amplification during the exponential phase as determined for each variant in preliminary experiments. Preliminary experiments had shown that β-actin was a better and more constant standard for comparisons. PCR products were separated on a 2% agarose gel. DNA bands were visualized by ethidium bromide staining, and their intensities were normalized vs. β-actin. Fluorescent intensities were measured and analyzed using ImageQuant software (Molecular Dynamics). Real-time quantitative PCR was performed using an ABI 7900 instrument. The PDE4A4 primer and probe sequences were as follows [forward: 5-CGCACCGGCCCATAGAG-3′; reverse: 5′-TGCCAGTGCCATGGAAGGA-3′; 6-carboxyfluorescein dye (FAM)-probe: 5′-ACCCGCATGTCCTG-3′] and were used at the concentrations of 800 and 250 nM, respectively. Control reagents for β-actin, including primers and VIC-reporter probe, were purchased from ABI. Primers were used at 900 nM and probe at 200 nM.

View this table:
Table 2.

Data for the oligonucleotide forward and reverse primers used to amplify 15 different PDE variants, transferrin, and β-actin

Assay of PDE4 phosphodiesterase activity.

The PDE4 phosphodiesterase activity was measured essentially as described previously (27). Macrophages were harvested by centrifugation for 3 min at 1,000 g at 4°C. The medium was removed, and the cell pellet was rinsed with PBS (137 mM NaCl, 3 mM KCl, 1 mM KH2PO4, and 6 mM Na2HPO4, pH 7.4) before its removal and subsequent addition of lysis buffer (25 mM HEPES, 2.5 mM EDTA, 50 mM NaCl, 50 mM NaF, 30 mM sodium pyrophosphate, 10% glycerol, and 1% Triton X-100, pH 7.5). Protease inhibitors were added to the lysis buffer, namely “complete EDTA-free protease inhibitor cocktail” (Roche Diagnostics, Mannheim, Germany), to give final concentrations of 40 μg/ml PMSF, 150 μg/ml benzamine, 1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin A, and 1 μg/ml antipain. Lysis was allowed to progress on ice for 1 h with gentle agitation during which time cell disruption occurred, allowing PDE activity to be ascertained. PDE activity was then determined using a modification of the Thompson and Appleman (38) two-step procedure as described previously by us (31). Briefly, a 2 μM mixture of [3H]cAMP (Amersham Biosciences) and unlabeled cAMP in 20 mM Tris, pH 7.4, and 10 mM MgCl2 assay buffer was mixed with complete cell lysates to a final concentration of 1 μM cAMP. To determine the total PDE4 activity, the selective inhibitor rolipram was included in the reaction mixture to a final concentration of 10 μM (16). Reactions were incubated at 30°C for 20 min with frequent agitation. cAMP hydrolysis was stopped by plunging the tubes into boiling water for 3 min and then placing them on ice for 10 min. Twenty-five microliters of venom from Crotalus attrox (Sigma) was added, and the samples were incubated at 30°C for a further 10 min. Dowex resin (Sigma 1X8–400), pH 3.0, stored as a 50:50 Dowex/water mixture, was prepared immediately before the assay by the addition of one volume of 100% ethanol to two volumes of the Dowex/water mixture to create a slurry. Four hundred microliters of the Dowex slurry were added to each reaction, vortexed, and placed on ice for 15 min. During this period, samples were frequently mixed by repeated inversion. Samples were then vortexed before being centrifuged at 13,000 g at 4°C for 3 min. One hundred and fifty microliters of the supernatant fraction from each tube were then placed in a corresponding scintillation vial containing 1 ml of scintillant and taken for counting on a beta counter.

Statistical analysis.

Statistical analysis of the semiquantitative RT-PCR results was performed using Mann-Whitney's or Wilcoxon's test and Student's t-test or paired t-test as indicated.


This study set out to determine the expression profile of 12 PDE4 splice variants by qualitative RT-PCR in inflammatory cells of smokers diagnosed with or without COPD. We also analyzed the profile of the cAMP-specific PDE3A, PDE3B, and PDE7A enzymes because PDE3 activity, together with that of PDE4, provides the major cAMP PDE activities in cells associated with immune responses (11). Analysis of PDE7A was performed since this enzyme has been suggested, by Li and coworkers (25), to be required for T cell activation. The qualitative analysis was followed by a semiquantitative RT-PCR or real-time quantitative analysis to evaluate differences in expression patterns of these PDE forms highlighted by qualitative RT-PCR. Finally, we also analyzed the PDE4 enzyme activity in BAL macrophages from COPD subjects and controls.

Three groups of subjects were recruited. Group 1 consisted of 11 smokers diagnosed with COPD, whereas group 2 consisted of 10 smokers matched for age, gender, and smoking history (see materials and methods) but without COPD. Five samples in each group were used for RNA profiling work, and the remaining samples in each group were used for enzyme activity determinations. Group 3 was made up of five nonsmokers who volunteered to donate peripheral blood samples for PCR analysis. Because elevated cAMP levels have been shown to affect the expression of certain PDE4 variants (24, 30), care was taken to avoid exposure to cAMP-elevating agents (e.g., β2-adrenoceptor bronchodilators) before sampling.

Evaluating primer pairs.

To undertake this study, gene-specific amplification conditions and primer pairs needed to be optimized and validated for each PDE4 splice variant. Human total brain RNA was used for this purpose and was included as a positive control for all the RT-PCR experiments. Table 2 shows the primer sequences used for RT-PCR and amplicon sizes, and Fig. 1 illustrates schematically where the PDE subtype generic and isoform-specific RT-PCR primers are located. Figure 2 shows a typical agarose gel with the RT-PCR expression profile for the 15 PDE variants in whole human brain. All amplicons migrated at the expected rate. As expected from the primer design, the PDE4D1/2 pattern produced two bands, as a fragment from both PDE4D1 and from PDE4D2 was amplified. All amplicons were further subjected to a diagnostic restriction enzyme analysis to confirm the identity of the fragments produced (data not shown). Bands on the agarose gel were scored as positive only when the amplification was reproducible in each PCR replicate. Splice variant-specific amplicons for PDE4 were scored as positive only when the expected fragment for the corresponding generic PDE4 subtype was also amplified in the same sample.

Fig. 1.

Schematic presentation of the cAMP phosphodiesterases (PDE) 3, 4, and 7 gene structure and location of the RT-PCR primers used. Grey boxed areas indicate sequence common to the different splice variants within the subtype. Single lines represent unique nucleotide sequence specific to a particular splice variant. Lightly hatched boxed areas denote the conserved upstream conserved region (UCR)1 and UCR2 in the PDE4 family. Thickly hatched boxed areas denote conserved catalytic domains across the PDE families. PDE4A7 is a truncated PDE4A form lacking catalytic activity. A Graphic represents a PDE subtype generic RT-PCR primer, able to amplify all splice variants of a specific PDE subtype; Graphic denotes a PDE splice variant-specific RT-PCR primer.

Fig. 2.

A 2% agarose gel showing the PDE expression profile pattern obtained for whole human brain using optimized PCR primer and reaction conditions (as shown in Table 2). The outer lanes contain the 100-bp DNA size ladder (Promega). Lane 1, PDE4A generic (338 bp); lane 2, PDE4A4 (257 bp); lane 3, PDE4A10 (281 bp); lane 4, PDE4A7 (301 bp); lane 5, PDE4B generic (546 bp); lane 6, PDE4B1 (236 bp); lane 7, PDE4B2 (264 bp); lane 8, PDE4B3 (214 bp); lane 9, PDE4C (95 bp); lane 10, PDE4D generic (262 bp); lane 11, PDE4D1 (163 bp); lane 12, PDE4D1/2 (221 and 307 bp); lane 13, PDE4D3 (292 bp); lane 14, PDE4D4 (347 bp); lane 15, PDE4D5 (212 bp); lane 16, PDE3A (353 bp); lane 17, PDE3B (305 bp); lane 18, PDE7A (254 bp); lane 19, transferrin (401 bp); lane 20, β-actin (466 bp); lane 21, water control. The amplified fragments migrated as expected. GenBank accession numbers for all PDE isoforms studied here are shown in Table 2.

Assessing transcripts in cells of the immune system by qualitative RT-PCR.

Figure 3 shows an agarose gel with a representative, qualitative RT-PCR expression profile for the indicated PDE isoforms for BAL macrophages and peripheral blood monocytes, T cells, and neutrophils from a smoker with COPD. The results for each cell type and comparison between smokers with and without COPD and nonsmokers are summarized in Table 3.

Fig. 3.

Representative agarose gels showing the qualitative RT-PCR PDE expression profile for human bronchoalveolar macrophages (A), human peripheral blood monocytes (B), T lymphocytes (C), and neutrophils (D) from a smoker with chronic obstructive pulmonary disease (COPD). Gel loading was identical as in Fig. 2.

View this table:
Table 3.

Summary of PDE expression profiling results

Table 3 illustrates that multiple PDE4 variants were detected in each cell type, and at least one splice variant of each subtype was observed in each cell type. The notable exception was the PDE4C subtype, which was not detected in peripheral blood T cells, whereas present in all other cell types, although very weakly. Qualitative RT-PCR revealed the presence of mRNA for the long PDE4A10 and PDE4B1 as well as the short PDE4B2, PDE4D1, and PDE4D2 isoforms in all the cell types evaluated here (Table 3). Additionally, transcripts for the catalytically inactive PDE4A7 (2EL) splice variant (14) were consistently and ubiquitously expressed in all cell types analyzed, although amplification was weak compared with expression in brain tissue. Transcripts for both PDE3B and PDE7A were abundantly expressed in all inflammatory cell types investigated in this study. In contrast to this, the long PDE4B3 and PDE4D4 isoforms were not detected in any cell types analyzed in this study (Fig. 3, lanes 8 and 14, and Table 3), whereas transcripts for both of these isoforms in human brain were easily detected (Fig. 2). Additionally, the long PDE4A4, PDE4D3, and PDE4D5 isoforms clearly exhibited (Table 3) cell-specific expression patterns by qualitative RT-PCR. This was also shown for the minor PDE4C subfamily using a generic probe. PDE4A4, PDE4D3, and PDE4D5 are intriguing isoforms in that all have been shown to interact with other proteins in cells, indicating that they may form specific regulatory complexes (16). Although transcripts for PDE4D5 were not detected in BAL macrophages of both COPD and control subjects by qualitative RT-PCR, they were clearly present in blood monocytes of the same subjects as well as in their T cells and neutrophils (Fig. 3 and Table 3). PDE4D3 transcripts were not observed in neutrophils of either smokers with and without COPD or nonsmokers (Table 3). For PDE4A4, transcripts amplified very weakly in the qualitative RT-PCR experiments, resulting in barely detectable bands on the agarose gels. However, in semiquantitative mode, PDE4A4 was consistently detected in all samples analyzed here apart from lung macrophages from smokers without COPD (Table 3 and Fig. 4).

Fig. 4.

Agarose gels showing the semiquantitative RT-PCR data and box plots for PDE4A4 expression. A: semiquantitative RT-PCR data comparing PDE4A4 expression in macrophages of smokers with COPD with smokers without COPD. B: transformed data as relative expression normalized to β-actin in the form of a box plot. C: PDE4A4 expression in blood monocytes for smokers with COPD, without COPD, and for nonsmokers. D: shows again the transformed data from C as relative expression normalized to β-actin in the form of a box plot and compares the data for smokers (both with and without COPD) with nonsmokers. Lanes 1–5, samples from smokers with COPD; lanes 6–10, samples from smokers without COPD; lanes 11–15, samples from nonsmokers; lane w, water control. A 100-bp DNA size ladder brackets the patient samples. BA, β-actin; 4A4, PDE4A4. P values were calculated using the Student's t-test.

Although a qualitative RT-PCR approach did not allow for comparisons of amplification intensities and thus the identification of any differences in transcript levels among PDE variants, it did allow for a first approximation of relative transcript levels of a particular PDE variant across different cell types. For example, it was particularly clear that the ratio of the signal for PDE4B1 vs. PDE4B2 transcript amplification was consistently higher in monocytes compared with T cells and neutrophils (Fig. 3). In addition to this, we noted that amplification of PDE4A7 and PDE3A transcripts from all cell types analyzed here was consistently weaker than we found using brain as a source.

Comparing PDE4 expression patterns among smokers with and without COPD by semiquantitative and quantitative RT-PCR.

First, a semiquantitative RT-PCR approach was used to address putative disease-specific or cell type-specific differences observed in the qualitative RT-PCR analysis of PDE4 transcript levels.

When doing this, the transcripts of the PDE4A4 long isoform, when normalized to β-actin control, were found to be significantly upregulated in BAL macrophages from COPD patients compared with non-COPD subjects (P < 0.01, Fig. 4). In contrast, PDE4A4 expression was not seemingly different in blood monocytes of the same COPD subjects compared with non-COPD controls (Fig. 4). Interestingly, this was the only significant difference among the 15 PDE variants analyzed in inflammatory cells of smokers with COPD vs. smokers without COPD. It was, therefore, decided to further confirm the differential expression of PDE4A4, specifically in COPD vs. healthy smokers, by real-time quantitative PCR. The data in Fig. 5 confirm a significant (P < 0.02) upregulated expression for PDE4A4 in BAL macrophages of smokers with COPD.

Fig. 5.

Box plot comparing the expression of PDE4A4 between COPD subjects and controls, as measured by real-time quantitative PCR. Data are normalized vs. β-actin. P value (P = 0.0159) was calculated using the exact Mann-Whitney's test.

Further semiquantitative analyses were performed for peripheral blood cells to compare PDE expression levels between smokers and nonsmokers. It was noted that both PDE4A4 and PDE4B2 transcripts were significantly higher (P < 0.005 for both) in blood monocytes of smokers compared with nonsmokers (Fig. 4D and Fig. 6, respectively). Significant differences were not observed for the other PDE variants analyzed (data not shown). This semiquantitative RT-PCR approach did not reveal any other consistent and significant differences for PDE4B2 in monocytes or macrophages when comparing among smokers with and without COPD, and it did not reveal any differences between monocytes and macrophages from either smokers with COPD or smokers without COPD.

Fig. 6.

Agarose gel (A) and box plot (B) comparing the expression of PDE4B2 in monocytes among smokers (both with and without COPD) and nonsmokers. Box plot shows the data as normalized expression vs. β-actin. Lanes 1–5, samples from smokers with COPD; lanes 6–10, samples from smokers without COPD; lanes 11–15, samples from nonsmokers; lane w, water control. A 100-bp DNA size ladder brackets the patient samples. 4B2, PDE4B2. P values were calculated using the Student's t-test.

Our study further allowed for a comparison of PDE expression levels between peripheral blood monocytes and lung macrophages from the same subjects. This analysis showed that PDE4C transcript levels were significantly higher when monocytes differentiate to macrophages in the lung, irrespective of disease state (P = 0.005; Fig. 7). In contrast, PDE4D5 transcript levels decrease as the monocytes mature into lung macrophages (P < 0.01; Fig. 8).

Fig. 7.

Agarose gels showing a semiquantitative RT-PCR comparison of PDE4C expression in monocytes (A, lanes 1–10) vs. macrophages of the same subjects (B, lanes 1–10). C: box plot shows the data as normalized expression of PDE4C vs. β-actin. Lanes 1–5, samples from smokers with COPD; lanes 6–10, samples from smokers without COPD; lanes 11–13, samples from nonsmokers; lane w, water control. A 100-bp DNA size ladder brackets the patient samples. 4C, PDE4C. P values were calculated using Wilcoxon's test.

Fig. 8.

Agarose gel (A) showing a semiquantitative RT-PCR comparison of PDE4D5 expression in monocytes (lanes 1–13) vs. macrophages of the same subjects (lanes 14–23). B: bar graph shows the data as normalized expression of PDE4D5 vs. β-actin. Lanes 1–4 and 14–18, samples from smokers with COPD; lanes 5–9 and 19–23, samples from smokers without COPD; lanes 10–13, samples from nonsmokers; lane w, water control. A 100-bp DNA size ladder brackets the patient samples. 4D5, PDE4D5. P values were calculated using the paired t-test.

Because this study revealed a disease-specific difference in PDE4A4 expression in BAL macrophages, we decided to determine the PDE4 enzyme activity in BAL macrophages from COPD subjects and controls. As shown in Fig. 9, the total PDE4 activity was markedly and significantly higher in BAL macrophages from subjects with COPD compared with subjects without COPD.

Fig. 9.

Box plot comparing PDE4 enzyme activity in bronchoalveolar lavage macrophages from smokers with COPD and smokers without COPD. The P value (P = 0.02) was calculated using Mann-Whitney's test.


PDE4 cAMP phosphodiesterases are thought to hold great promise as therapeutic targets for inflammatory diseases, and, currently, a number of PDE4 inhibitors are in late-stage clinical trials for COPD (5, 10). However, use of most PDE4 inhibitors evaluated to date has resulted in side effects, including nausea and emesis (10). A large family of isoforms provides PDE4 activity in cells, and thus the identification of isoforms associated with inflammatory cell types of relevance to COPD may provide a new focus for the development of third generation PDE4 inhibitors that are targeted at specific PDE4 isoforms. When using any such approach, one would have to be mindful that side effects may also be related to specific PDE4 isoforms as well. Indeed, it has been suggested, on the basis of gene knockout studies done on mice and using a paradigm for emesis, that inhibition of PDE4D activity may be specifically linked to emesis (34).

Four genes encode at least 12 different PDE4 splice variants, and for those analyzed here, we show clear evidence for specific tissue distribution (6, 7, 15). Gene knockout studies (12, 21), selective inhibition (27, 29), antisense (28), and dominant negative strategies (33) imply that specific PDE4 isoforms may have particular functional roles. This is considered most likely achieved (7, 16) by their specific intracellular targeting, leading to the control of anchored PKA (36). A number of studies have been published describing the expression of particular PDE4 variants in leukocytes (9, 23). However, a detailed and comprehensive investigation of the expression and distribution of PDE4 splice variants in inflammatory cells of relevance to COPD has not been reported. In addition, no information is available on the PDE4 isoform expression profile in lung bronchoalveolar macrophages, and there is also no information on the comparison of expression in inflammatory cells obtained from smokers with COPD vs. smokers without COPD and vs. nonsmokers.

PDE4 enzymes have a very high specific activity and are expressed at very low levels endogenously (4, 7, 16). This makes it extremely challenging to detect native levels of expression, especially under conditions where cell numbers and samples are very limited, as in clinical studies. Furthermore, there is a paucity of effective isoform-specific antisera, and certain major isoforms have proved to be remarkably difficult to generate antisera against (see e.g., Ref. 15). RT-PCR provides an extremely powerful and sensitive means of assessing transcripts for proteins. Using qualitative RT-PCR analysis, we show that all cell types tested here have transcripts for at least one PDE4 variant representing each PDE4 subfamily. The only exception was T cells, which did not express any PDE4C, as measured by RT-PCR. In many cases, several splice variants representing a PDE4 subtype were present. Furthermore, none of the cells studied here expressed just a single PDE4 subtype. Additionally, we noted that PDE4D5 was absent in BAL macrophages of both COPD and non-COPD subjects, whereas this long isoform was clearly present in the monocytes from the same individuals. PDE4D3 was not observed in blood neutrophils but was detected in all other cell types.

Semiquantitative analyses revealed disease and cell type-specific expression patterns for certain PDE4s. Thus PDE4A4 was specifically and uniquely upregulated in lung macrophages of COPD subjects, whereas PDE4A4 and PDE4B2 were both upregulated in peripheral blood monocytes from smokers compared with nonsmokers (Figs. 4C and 6). PDE4D5 was downregulated in lung macrophages compared with blood monocytes from the same subjects, irrespective of the disease state (Fig. 8). Furthermore, PDE4C was very poorly expressed in blood monocytes, yet clearly upregulated in the macrophages of the same individuals, again irrespective of disease state. Real-time quantitative PCR further confirmed the upregulation of PDE4A4 in COPD subjects vs. healthy controls (Fig. 5).

Previous RT-PCR data (41) have indicated that PDE4B2 transcripts predominate in peripheral blood monocytes and neutrophils from healthy volunteers. Indeed, it was suggested that PDE4B2 constituted 80.2% of the PDE4 population in unstimulated monocytes and 99.7% in neutrophils, with PDE4D forming 0.3%, and PDE4A and PDE4C together <0.1% of the total PDE4 transcripts in neutrophils. Our data are clearly at variance with this, showing that transcripts for all four PDE4 subfamilies were readily and abundantly detected in both monocytes and neutrophils (Fig. 3). Moreover, transcripts for PDE4B1, in addition to PDE4B2, were clearly expressed in both monocytes and neutrophils as well as in T cells and BAL macrophages (Fig. 3). Transcripts for multiple splice variants within the PDE4A and PDE4D subfamilies were also clearly evident. The reasons for these differences are not clear but could include the use of suboptimal PCR conditions or primer sequences, as we found it was essential to thoroughly validate the primers and conditions used not only against plasmids encoding specific isoforms but especially on whole brain RNA, which we show here provides an excellent source of the full range of known PDE4 isoforms (Fig. 2). Furthermore, whereas LPS activation of monocytes has been shown to increase PDE4B2 transcript levels (26), we show here that, using peripheral blood monocytes and lung macrophages obtained from smokers with and without COPD, a difference in expression levels for PDE4B2 could not be observed between disease states and cell types using semiquantitative RT-PCR. This may indicate a difference from in vivo activation compared with that seen in vitro with LPS.

PDE4C is often ignored as contributing to inflammatory cell function based on its low-level expression in these cells. We confirm that PDE4C transcript levels are lower compared with those of other PDE4 subfamilies in blood monocytes and neutrophils and confirm that PDE4C transcripts are not found in blood T cells (Fig. 3) (11). However, PDE4C mRNA is clearly present (Fig. 6) in elevated amounts in BAL macrophages compared with monocytes from the same individuals. Although no differential expression was observed between COPD and healthy subjects, PDE4C cannot be excluded as a possible macrophage therapeutic target for COPD.

Transcripts for PDE3A, PDE3B, and PDE7A were detected in all four cell types analyzed, extending observations of their expression in primary CD4+ and CD8+ T lymphocytes (11). PDE3A was very weakly present, especially in lung macrophages and blood monocytes. PDE4A7 is a curious isoform that is NH2- and COOH-terminally truncated to such an extent that it is catalytically inactive (14). We clearly identified transcripts for it in a wide variety of blood cells, suggesting that further investigation of this isoform is warranted.

A key finding of this study is the specific upregulation of PDE4A4 in lung macrophages from COPD subjects. A key property of PDE4A4 is its ability to interact with the SH3 domains found in various Src family tyrosyl kinases involved in activation of immune system cells (32). This interaction serves to both target PDE4A4 to specific intracellular locations as well as affect the conformation of the catalytic unit indicated by a heightened sensitivity to inhibition by the archetypal PDE4-selective inhibitor rolipram (32). Importantly, the activity of PDE4A4 increases upon activation of U-937 monocytic cells by proinflammatory agents such as LPS and interferon-γ (27). This suggests a key role for this isoform in lowering localized cAMP levels to facilitate inflammatory cell activation. Additionally, overexpression of PDE4A5, the rodent homolog of PDE4A5, has been shown to protect fibroblasts against apoptosis (19), which might again suggest that inhibition of this isoform is of potential therapeutic importance in attenuating inflammatory lung disease. Intriguingly, PDE4A5 has also very recently been shown to interact specifically with the immunophilin XAP2 (3). Although the functional consequences of this are still unknown, we note that XAP2 is also able to interact with the aryl hydrocarbon (dioxin) receptor, which can be activated by pollutants (22).

Given the significant upregulation of PDE4A4 in BAL macrophages of smokers with COPD, we also measured the PDE4 enzyme activity in BAL macrophages of smokers with and without COPD. The data show that the total PDE4 activity is clearly upregulated in BAL macrophages of smokers with COPD. Unfortunately, we were unable to measure activity specifically associated with PDE4A. This may be due to low absolute amounts of PDE4A protein or due to poor immunoprecipitating capacity of the antiserum used. Although transcripts for PDE4A4 were markedly upregulated in these samples and this possibly contributed to the increased enzyme activity, we cannot rule out contributions from other PDE4 variants and/or increases due to posttranslational activation of all or some of the PDE4 variants. Nonetheless, the observation that the total PDE4 enzyme activity is markedly increased in macrophages of COPD subjects compared with non-COPD subjects is of interest.

Because macrophages are a hallmark of inflammation in COPD (1, 20), the finding of a highly specific and significant differential expression of PDE4A4 in macrophages of COPD subjects may well be of importance. Indeed, the results of this study lead us to suggest that PDE4A4 may be a PDE4 isoform-specific therapeutic target for COPD. Whereas PDE4A4 is clearly also expressed in other cell types, including immune cells, its disease-specific upregulation in lung macrophages could possibly provide a new route into a PDE4-targeted anti-inflammatory treatment for COPD, with less potential for dose-limiting side effects. Testing the new hypothesis that the PDE4A4 isoform is a plausible new therapeutic target for COPD will require a new emphasis on either making specific inhibitors for PDE4A4 or devising new strategies to disrupt PDE4A4-specific intracellular interactions (19, 27, 32) to help elucidate its role in COPD.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 11a.
  13. 12.
  14. 13.
  15. 14.
  16. 15.
  17. 16.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
View Abstract