International Journal of Impotence Research (2006) 18, 501–509 & 2006 Nature Publishing Group All rights reserved 0955-9930/06 $30.00 www.nature.com/ijir

REVIEW 11: a brief review of structure, expression and function

A Makhlouf, A Kshirsagar and C Niederberger

Department of Urology, University of Illinois at Chicago, Chicago, IL, USA

Phosphodiesterase 11 (PDE11) is the latest isoform of the phosphodiesterase family to be identified. Interest in PDE11 has increased recently because , an oral phosphodiesterase 5 inhibitor, cross reacts with PDE11. The function of PDE11 remains largely unknown, but growing evidence points to a possible role in male reproduction. The published literature on PDE11 structure, function and expression is reviewed. International Journal of Impotence Research (2006) 18, 501–509. doi:10.1038/sj.ijir.3901441; published online 5 January 2006

Keywords: phosphodiesterase; tadalafil; spermatogenesis;

Introduction possible role of PDE11 in spermatogenesis and potential effects, if any, of tadalafil on it. Introduction of orally administered phosphodiester- ase-5 (PDE5) inhibitors revolutionized the treatment of erectile dysfunction in the past decade.1 Of the 11 Overview of PDEs known (PDEs), PDE5 has been 1,6 the focus of much attention because it is the protein Mammalian PDEs are divided into 11 families. target of these inhibitors. was the first Some families (PDE1, PDE3, PDE4, PDE7 and PDE8) PDE5 inhibitor to be marketed followed by varde- are products of multiple genes (multigene), whereas nafil and tadalafil. All three compounds inhibit the others derive from a single gene (unigene) (Table 1). catalytic activity of PDE5, thereby preventing the Thus, the human genome contains 21 known genes degradation of intracellular cGMP. In turn, cGMP coding for PDEs. Variations in splicing and alternate activates cGMP-dependent protein kinase, which initiation sites create more variety in each family, with 6,7 phosphorylates a number of intracellular proteins a total number of 53 isoenzymes identified so far. resulting in decreased intracellular calcium concen- All PDEs share a highly conserved catalytic domain 1 tration and leading to penile smooth muscle relaxa- near the carboxyl-terminus, including 11 invariant 8 tion, vasodilatation, and subsequently, penile amino acids in the . Minor variations in the erection. Recently, there has been an increased catalytic domain are believed to be responsible for the interest in PDE11 because tadalafil, one of the newer different selectivities towards cAMP or cGMP sub- 8 PDE5 inhibitors approved for treatment of ED, cross- strate and various inhibitors (Table 1). More sub- reacts with PDE11.2–6 The expression and function stantial variation exists in the N-terminus part, where 1,9 of PDE11 are still poorly understood, and this most of the regulatory domains reside. present review will summarize the current state In the case of PDE11, there is a single gene in the of knowledge regarding PDE11 structure, function family (PDE11A according to standard nomencla- and distribution. Emphasis will be placed on the ture) with four splicing variants (PDE11A1, PDE11A2, PDE11A3 and PDE11A4). This appears to be the case in all mammalian species studied so far (human, rat and mouse).10–12 Correspondence: Dr A Makhlouf, Department of Urology, University of Illinois at Chicago, 840 S. Wood Street, MC 955 Chicago, IL 60612-7316, USA. E-mail: [email protected] Cloning and isolation of PDE11A Received 2 August 2005; revised 29 November 2005; accepted 29 November 2005; published online 5 January PDE11A was the latest member of the mammalian 2006 PDE family to be identified.13 As the entire human Phosphodiesterase 11 A Makhlouf et al 502 Table 1 Summary of PDE family characteristics1,7–9

PDE Genes (if Substrate Inhibitors, Ki or (IC50) Regulation Tissue expression family multigene)

2 þ 1 PDE1A cAMP4cGMP (14 mM) ( þ )Ca /CAM 1A: Thyroid, testis, brain PDE1B cAMP4cGMP (6 mM)(À)PKA, CamKII 1B: Brain, lymphocytes PDE1C cAMP ¼ cGMP Sildenafil (0.35 mM) 1C: Blood vessels, testis

2 cGMPXcAMP EHNA (1 mM) ( þ )cGMP, PKC Brain, heart, platelets, liver, Bay 60-7550 (0.047 mM)(þ /À)N-terminal-targeting thymocytes domain

3 PDE3A cAMP4cGMP (0.020 mM) ( þ )PKA, PKB 3A: Heart, blood vessels PDE3B (0.150 mM) (À)cGMP 3B: Adipocytes, (0.5–2 mM)(þ /À)N-terminal-targeting hepatocytes, lymphocytes domain

4 PDE4A cAMP (1 mM) ( þ )PKA, ERK, phosphatidic 4A: Lung, immune cells, acid brain, blood vessels PDE4B (0.8 nM)(À)ERK, caspases PDE4C (120 nM)(þ /À)N-terminal-targeting PDE4D Zardaverine (0.8–4 mM) domains

5 cGMP Zaprinast (0.130 mM) ( þ )cGMP, PKG, PKA Smooth muscle, platelets, Sildenafil (0.10 mM) (À)Caspases Purkinje cells, gastro- (0.001 mM) intestinal epithelium, Tadalafil (0.010 mM) pulmonary endothelium

6 PDE6A cGMP Zaprinast (0.400 mM) ( þ ) Retinal photoreceptors PDE6B (0.125 mM) (À)cGMP, gamma and delta PDE6C Sildenafil (0.050 mM) subunits Vardenafil (0.011 mM) Tadalafil (2 mM)

7 PDE7A cAMP IBMX (4 mM) ( þ /À)PKA Brain, lymphocytes, kidney PDE7B Dipyridamole (0.6 mM)

8 PDE8A cAMP Dipyridamole (9 mM) PAS domain Thyroid PDE8B

9 cGMP Zaprinast (35 mM) Unknown Kidney

10 cGMP4cAMP Dipyridamole (1 mM) (À)cAMP Testis, brain Zaprinast (22 mM)

11 cAMP and Zaprinast (12 mM) Unknown Skeletal muscle, prostate, cGMP Dipyridamole (0.4 mM) testis, corpus cavernosum, heart (but see text and Table 3)

CAM ¼ calmodulin c; ERK ¼ extracellular signal-regulated protein kinase; PAS ¼ periodic acid-Schiff stain; PDE ¼ phosphodiesterase; PKA ¼ ; PKC ¼ protein kinase A; PKG ¼ protein kinase G. Only some of the most commonly reported inhibitors for each PDE are listed. Tissue expression is representative of most reports and is not necessarily exhaustive.

genome sequence is now available, it appears to be cloned the A3 and A4 isoforms of PDE11A by PCR the last. The first published report of PDE11A was with degenerate primers from the conserved cataly- by Fawcett et al.,13 who obtained a partial sequence tic domain of other PDEs, using a human multi- of PDE11A from a commercially available expressed tissue cDNA library template. No other variants of sequence tag (EST) database based on homology human PDE11A have since been reported. with other mammalian PDEs. These investigators then used the PDE11A partial sequence to identify the PDE11A gene from a human skeletal muscle Gene structure of PDE11A cDNA library. Soon after, the same group cloned PDE11A2 and PDE11A3 cDNAs from a human testis Among the mammalian superfamily of PDEs, PDE11 library.12 Almost simultaneously, Yuasa et al.10 is phylogenetically related to the other GAF-

International Journal of Impotence Research Phosphodiesterase 11 A Makhlouf et al 503 Stop codon (all forms) ATG ATG ATG ATG (PDE11A3) (PDE11A4) (PDE11A2) (PDE11A1) HSPDE11A gene ∗∗ >300 Kb Exon # 12 3 4567891011 12 13 14 15 16 17 18 19 20 21 22 23 GAF a GAF b Catalytic domain

490 aa (55.8 kDa)

AUG Stop PDE11A1 1784 bp GAF a domain

7 Exons 8-23 GAF b domain 576 aa (65.8 kDa) AUG Stop Catalytic domain PDE11A2 2141 bp ∗ Phosphorylation site ?56 Exons 8-23 684 aa (78.1 kDa) AUG Stop

PDE11A3 2205 bp

162 4 5 Exons 8-23 934 aa (104.8 kDa) AUG Stop

PDE11A4 ∗∗ 4476 bp

364 5 Exons 8-23 Figure 1 Gene structure of HSPDE11A (Homo sapiens PDE11A) and its four known variants. Gene structure is based on Yuasa et al.14 Exon numbers follow Yuasa et al.,14 in contrast to GeneBank that lists the 20 exons of PDE11A4 only. The 50-most exon(s) of PDE11A2 have not been studied.12 Sizes of the cDNA (in bp) and the predicted molecular weight (in kDa) of the open-reading frame protein are shown. containing PDEs (PDE2, PDE5, PDE6 and PDE10) for A1 and A3 have canonical TATA motifs, whereas based on sequence homology14 (GAF ¼ cGMP bind- that for A4 is TATA-less but has a GC-rich region ing PDE, Anabaena adenylyl cylcase and E. coli with a CCAAT box and Sp-1 . It is FhlA domain after the proteins where it was initially speculated that these differences in promoters are identified). GAFs are regulatory domains that in responsible for the different patterns of tissue some instances bind small molecules such as cGMP distribution exhibited by the various isoforms.14 or are involved in protein–protein interactions (see Sequence analysis of the upstream gene region Zoraghi et al.15 for a review). PDE11 most closely demonstrated the presence of SRY and Sox-5 motifs. resembles PDE5 by sequence, with 50% identity This hints at a potential role for PDE11A expression and 70% similarity in the catalytic domain.13 In in testicular development. However, no functional humans, the PDE11 subfamily consists of a single studies of the PDE11A promoter have been pub- gene, which has been localized to chromosome lished to date. 2q31 by fluorescent in situ hybridization.14 This Alternate splicing is responsible for further varia- was later confirmed by the Human Genome Project tion in the PDE11 family. PDE11A1, the shortest complete sequence. The gene spans a long stretch of form, does not include exons 1–6. Thus, it lacks the DNA (4300 kb) and is spread over more than 20 first GAF domain, and only includes a part of the exons (Figure 1). Alternate transcription initiation second GAF. PDE11A2 and PDE11A3 are slightly sites and alternate splicing are responsible for longer, including the entire second GAF and parts of the generation of the four isoforms of PDE11A the first GAF domain,10,12,14 whereas the longest (Figure 1).12,14 form, PDE11A4, includes both domains in addition The gene structure of PDE11A variants is detailed to two phosphorylation sites in the N-terminus in Figure 1. All four variants contain exons 8–23, part.10 which comprise the C-terminus of the protein with the catalytic domain. At the 50 end, however, the isoforms differ significantly. Omori’s group10,14 Biochemical characteristics of PDE11A identified three distinct transcription initiation sites for PDE11A1, PDE11A3 and PDE11A4, under the PDE11A is a dual-substrate PDE, acting on both control of three different promoters. The promoters cAMP and cGMP. Most studies report similar

International Journal of Impotence Research Phosphodiesterase 11 A Makhlouf et al 504 Table 2 Studies of the biochemical activity of PDE11

Study Isoenzyme cGMP cAMP Inhibitors studied (selected)

13 7 7 Fawcett et al. HSPDE11A1 Km ¼ 0.52 0.34 mM Km ¼ 1.04 0.23 mM IBMX IC50 ¼ 49.8 mM (cGMP) Vmax ¼ 3.9 pmol/min/mg Vmax ¼ 3.6 pmol/min/mg Zaprinast IC50 ¼ 12.0 mM (cGMP) Vmax (A/G) ¼ 0.92 Dipyridamole IC50 ¼ 0.37 mM (cGMP)

12 Hetman et al. HSPDE11A2 Km ¼ 3.7 mM Km ¼ 3.3 mM Zaprinast IC50 ¼ 28 mM Vmax (A/G) ¼ 1.0 Dipyridamole IC50 ¼ 1.8 mM Sildenafil IC5040.5 mM HSPDE11A3 Km ¼ 4.2 mM Km ¼ 5.7 mM Zaprinast IC50 ¼ 5 mM Vmax (A/G) ¼ 1.0 Dipyridamole IC50 ¼ 0.82 mM

10 Yuasa et al. HSPDE11A3 Km ¼ 1.570.07 mM Km ¼ 3.070.28 mM Vmax (A/G) ¼ 2.4 HSPDE11A4 Km ¼ 1.470.06 mM Km ¼ 3.070.26 mM 7 7 Vmax ¼ 120 4.6 pmol/min/mg Vmax ¼ 270 28 pmol/min/mg Vmax (A/G) ¼ 2.2

11 Yuasa et al. RNPDE11A1t Km ¼ 1.670.15 mM Km ¼ 3.170.28 mM Zaprinast IC50 ¼ 1871.8 mM Vmax (A/G) ¼ 2.9 Dipyridamole IC50 ¼ 0.5370.03 mM RNPDE11A2 Km ¼ 1.370.08 mM Km ¼ 2.270.24 mM Zaprinast IC50 ¼ 1670.43 mM Vmax (A/G) ¼ 1.9 Dipyridamole IC50 ¼ 0.2570.02 mM 7 7 7 RNPDE11A3 Km ¼ 1.6 0.23 mM Km ¼ 4.0 0.98 mM Zaprinast IC50 ¼ 17 0.35 mM 7 Vmax (A/G) ¼ 3.2 Dipyridamole IC50 ¼ 0.20 0.02 mM RNPDE11A4 Km ¼ 1.670.08 mM Km ¼ 3.970.13 mM Zaprinast IC50 ¼ 2072.5 mM Vmax (A/G) ¼ 2.3 Dipyridamole IC50 ¼ 0.2670.02 mM

5 Weeks et al. HSPDE11A4 Km ¼ 0.97 mM Km ¼ 2.4 mM Dipyridamole IC50 ¼ 0.8470.22 mM 7 Vmax (A/G) ¼ 4 to 10 Tadalafil IC50 ¼ 0.073 0.003 mM 7 Sildenafil IC50 ¼ 3.8 0.75 mM 7 Vardenafil IC50 ¼ 0.84 0.16 mM

HSPDE11 ¼ Homo sapiens PDE11; RNPDE11 ¼ Rattus norvegicus PDE11. Studies are listed in chronological order of publication. Vmax values are not for pure and are therefore only meaningful when comparing hydrolytic rates for cGMP and cAMP. Vmax (A/G) is the ratio of Vmax for cAMP to Vmax for cGMP.

affinities (Km) for both substrates (Table 2). How- in Figure 1). In vitro, these residues were found to ever, there have been slight discrepancies in the be phosphorylated by both cAK and cGK. However, reported Vmax ratio for the breakdown of cAMP whether these sites are phosphorylated in intact vs cGMP. Beavo’s group found a Vmax ratio close to 1 tissues and the in vivo role of this phosphorylation in PDE11A1, PDE11A2 and PDE11A3,12,13 whereas have yet to be elucidated. others reported a higher velocity for cAMP,10 Several inhibitors of PDE11 have been identified especially for the PDE11A4 isoform.5 The consensus and tested. The non-selective inhibitors IBMX, remains, however, that at physiologic concentra- zaprinast and dipyridamole have all been documen- tions, PDE11 is capable of degrading both cGMP ted to inhibit PDE11 (Table 2). Furthermore, the and cAMP.6,16 PDE5-selective inhibitor tadalafil cross-reacts with The regulation of PDE11 function is very poorly PDE11 (see below for more discussion).2,3,5 This is understood. Like other GAF-containing PDEs, not surprising given the sequence similarity bet- PDE11A4 has two GAF domains located in tandem ween PDE11 and PDE5. In fact, it appears that as near the N-terminal.12,14 In PDE5, binding of cGMP PDE5 inhibitor variants are modified to react less to GAF a, and perhaps to GAF b, enhances the rate and less with PDE6, they gain in PDE11 reactivity.4 of phosphorylation of the by cAK (cAMP- Unfortunately, there are no PDE11-specific inhibi- dependent protein kinase) or cGK (cGMP-dependent tors available as of this writing,16 and this has protein kinase).17 Binding of cGMP to the GAF a site presented a hurdle in the study of the physiological has also been shown to increase the catalytic activity role of PDE11. of PDE5 directly, without requiring phosphorylation As tadalafil is widely used to treat erectile of the enzyme.18,19 However, similar allosteric dysfunction, its effects on PDE11 deserve special control of PDE11A4 could not be demonstrated to mention. Tadalafil is selective for PDE5, but it cross- 5,13,14 4 date. Nevertheless, it is believed that PDE11A4 reacts with PDE11 with reported IC50 values of 37 5 is subject to physiologic regulation. Two putative and 73 nM for PDE11A1 and PDE11A4, respec- phosphorylation sites have been identified by Yuasa tively. Tadalafil fold-selectivity for PDE5 over et al.14 at Ser-117 and Ser-162 (marked by asterisks recombinant PDE11A1 was initially reported as 53

International Journal of Impotence Research Phosphodiesterase 11 A Makhlouf et al 505 and later as 6.7.2 A greater fold-selectivity (B40- was localized to the glandular epithelium by fold) was reported for the longest isoform, immunohistochemistry studies.21,22 PDE11A4.5 The difference between these values for The expression of PDE11 in the testis is somewhat fold-selectivity could be owing to the isoform of more debatable. Seven studies from at least three PDE11 tested, variation in preparation of the different labs found PDE11 expression in the testis, proteins or to differences in assay conditions. Even whereas one recent study failed to detect it21 with a 40-fold selectivity, some authors argue that (Table 3). Most, but not all, evidence points to clinical effects of tadalafil on PDE11 cannot be PD11A3 as the major testicular isoform. This finding excluded.20 In contrast to tadalafil, fold-selectivity was demonstrated in the case of mRNA using A3- of PDE5 vs PDE11 isoforms for sildenafil or specific PCR in both rats and humans.10,11 This vardenafil have been reported as 1500- and 640-fold result agrees with Western blot findings from two for PDE5/PDE11A12 and 1000- and 9300-fold for groups who found an approximately 75 kDa band PDE5/PDE11A4.5 in testicular extracts, which is close to the predicted size of PDE11A3.16,22 Unfortunately, both groups used the same commercial antibody, and in one of the two studies, a heavier band (100 kDa) had Expression and tissue distribution a stronger signal.22 Furthermore, another group, of PDE11A using two monoclonal antibodies, failed to detect any PDE11 expression in the testis altogether.21 The tissue distribution of PDE11A remains a topic of These discrepancies could be owing to differences active study and some controversy.16,21,22 The initial in antibody and/or assay sensitivity and specificity. reports of PDE11A11–13 found evidence for PDE11 We should note that the group that detected expression in skeletal muscle, prostate, testis and testicular PDE11 used 100 mg per Western blot lane salivary glands. Several studies since then have and found PDE11 in a wide array of tissues.22 In supported and sometimes contradicted these find- contrast, the investigators who reported no testicu- ings. A summary of all studies published to date on lar expression of PDE11 loaded only 20 mg per the expression of PDE11 is presented in Table 3. lane and failed to detect it in the majority of tissues Unfortunately, interpretation of these studies is they examined.21 Finally, studies using immuno- rendered difficult because of differences in the histochemistry have localized the PDE11 signal species (human, rat or mouse) and isoforms to the germinal epithelium in the seminiferous (PDE11A1, PDE11A2, PDE11A3 or PDE11A4) stu- tubules and to Leydig interstitial cells as well as died. Furthermore, some studies investigated mRNA spermatozoa.22,23,25 presence, whereas others reported protein expres- The expression of PDE11 in tissues other than sion, making direct comparisons difficult. prostate and testis has been documented but not The presence of PDE11 in the prostate is the best- studied extensively.10,13,22–24 Immunohistochemis- documented aspect of all the summarized studies. In try detected PDE11 in the smooth muscle of corpora fact, prostatic expression of PDE11 mRNA and/or cavernosa,24 in vascular endothelium and smooth protein was detected in every published study that muscle,22 in acidophils and lactotrophs of the included prostate specimens.10,11,13,21,22,25 How- anterior pituitary23 and in parasympathetic ganglia ever, some discrepancy exists in the literature as to of the heart.21 Whether the presence of PDE11 in which PDE11 isoform is present in the prostate. The these tissues is of any physiological significance two initial reports identifying PDE11 reported the remains to be established. presence of the longest transcript (10 kb) on North- ern blots of prostatic RNA, whereas other tissues had shorter transcripts.10,13 Using specific primers, Potential role of PDE11 in male Yuasa et al.10 identified this transcript as PDE11A4, reproduction and the view that PDE11A4 is expressed in the prostate, whereas PDE11A3 was restricted to the Although PDE11 was identified almost 6 years ago, testis gained acceptance.20 However, the same group very little is known about its physiologic function. later found that, in the rat, PDE11A4 is not Advances in this regard have been hampered by the expressed in the prostate.11 Two studies found the absence of a specific inhibitor and by the lack of a major protein band detected in the prostate clear pattern of PDE11 tissue expression. The recent (56 kDa13 and 70 kDa22) to be too small for the creation of a mouse PDE11 À/À knockout strain,25 expected weight of PDE11A4 (104 kDa). Still, a band however, holds some promise to identify the role of of the right size (100–105 kDa) was seen in studies PDE11 in body functions. In addition, the wide- from at least two different labs with three different spread use of the PDE5 inhibitor tadalafil, which antibodies.21,22 Thus, the weight of the evidence cross-reacts with PDE11, has hinted at a potential clearly argues for the presence of PDE11A4 in role for PDE11 in spermatogenesis and sperm human prostate, with other isoforms probably capacitation in select patients with pre-existing expressed as well. The location of the PDE11 signal fertility problems.26

International Journal of Impotence Research nentoa ora fIptneResearch Impotence of Journal International 506

Table 3 Studies of the expression and tissue distribution of PDE11

Study Species mRNA detection Protein detection (year) Probe/primers Technique Result Antibody Technique Result

Fawcett Human 428-bp probe of Northern blot Skeletal muscle (8.5 kb A3?), Three polyclonal Western blot Prostate (56 kDa A1), skeletal et al.13 N-terminus of prostate (10.5 kb A4?46.0 kb anti-sera (EPH2, muscle (78 kDa A3465 kDa catalytic domain A1?) EPH3, EPH4) to three A2456 kDa A1) Dot blot Prostate4testis, skeletal muscle, different PDE11 kidney, liver, pituitary, salivary peptides (amino glandbadrenal, pancreas, spinal acids 115–129, 410– cord, trachea and mammary, and 424 and 454–468) thyroid glands

Yuasa Human Probe common to Northern blot Prostate (10 kb A4, 6 kb A3?, 2 kb et al.10 A3 and A4 ?)4Testis (3 kb A3?) Dot blot Prostate4testis ¼

pituitary4salivary, thyroid, 11 Phosphodiesterase pancreas, liver A3- and A4- RT-PCR Prostate A4 only Makhlouf A specific primers Testis A3 only

Yuasa Rat Primers specific for RT-PCR A2 in the brain, lung, skeletal tal et et al.11 A2, A3 and A4 muscle, spleen, testis and prostate A3 in the testis only A4 in the brain, heart, kidney and liver (not prostate)

Baxendale Human and Polyclonal anti- Immunohisto- Testis (seminiferous epithelium et al.23,24 mouse PDE11 catalytic chemistry and Leydig cells), pituitary domain peptide (anterior acidophils and (EPH3) to amino lactotrophs), corpus cavernosum acids 410–424a and vascular sm. muscle

Loughney Human Two monoclonal Western blot 417L: Prostate et al.21 antibodies (417L and (105 kDa)4pituitary 479F) to two (105 kDa)bheart (105 kDa). Not different epitopes in the penis, liver, testis or sk. near carboxyl- muscle terminal (mapped to 479F: Prostate amino acids 410–424 (105 kDa)4pituitary (105 kDa); and 454–467, sk. muscle and heart (175 kDa, respectively) artifact?). Not in the testis, liver, penis. No A1–A3 bands Immunohisto- Prostatic glandular epithelium chemistry and cardiac parasympathetic ganglia

Wayman Mouse Expressed Lac-Z Cytochemistry Prostate (glandular epithelium), Polyclonal antibody Immunohisto- Testis (germinal epithelium and et al.25 insert in PDE11 for b- testis (seminiferous tubules). (EPH3) to amino chemistry in Leydig cells) and spermatozoa knockout cassette galacotsidase Not aorta. Other tissues not acids 410–424 PDE11 in À/À mice reported þ / þ mice Phosphodiesterase 11 A Makhlouf et al 507 Spermatogenesis refers to the differentiation of sk. 4 testis spermatogonia into spermatids through the first and

4 second meiotic divisions. There is ample evidence that this process is regulated through a cAMP- 70 kDa)

100 kDa dependent mechanism. First, b 4 expression has been documented in spermatogonia, and it peaks during early meiosis.27 Second, CREM isoform sequence. 70 kDa) and bladder (cAMP-responsive element) is a transcription factor b that binds cAMP and promotes the transcription of CC (100 kDa tected signal with PDE11A1, 4 several genes involved in spermatogenesis. Knock- A4) muscle (100 kDa, 70 kDa) tissues not tested Spermatozoa (acrosomal cap and flagellum). Other tissues not studied (100 kDa (100 kDa only) Wide variety of tissues including the adrenal, spleen, prostate, sm. muscle, testis out mice deficient in CREM are infertile because of an arrest in spermatogenesis.28 This can be reversed which was raised against a peptide

13 by implantation of wild-type spermatogonia, indi- ., cating that the defect is specific to the spermato- 28

et al gonia and not a systemic hormonal problem. Given that PDEs regulate cAMP levels through its degrada- tion, it stands to reason that they play a role in Western blot Prostate (70 kDa WesternImmunohisto- chemistry Spermatozoa (75 kDa A3?). Other Immunohisto- chemistry regulating spermatogenesis. However, it should be stressed that many PDEs are expressed in cells of the spermatogenic pathway,16 and assigning a role to PDE11 in particular, at least in humans, is not warranted by the evidence so far. Another area where PDE11 is suspected of playing a physiological role is in sperm capacitation. Sperm exiting the epididymis are incapable of fertilization Commercial (FabGennix Inc.) polyclonal anti-body (PD11A-112A) to residues 454–468 Commercial (FabGennix Inc.) polyclonal anti-body (PD11A-112A) to residues 454–468 until they undergo capacitation in the female genital tract.29 Several factors influence capacita- tion, including factors secreted by the prostate.30 The mechanism of capacitation is also cAMP depen- dent.25 Several cytokines cause release of cAMP, which in turn causes an influx of calcium into the spermatozoon triggering capacitation.25 By keeping cAMP levels low, PDEs are believed to prevent premature capacitation.25 Two lines of evidence suggest the involvement of PDE11 in spermatogenesis and capacitation: animal 26,31 but not epididymal sperm studies and clinical trials with tadalafil, and the PDE11 knockout mouse model.25 Daily administration of high doses of tadalafil (410 mg/kg/day) to beagle dogs was shown to cause atrophy of the germinal epithelium in 20–100% of animals with subsequent oligospermia in 40–75%.31 RT-PCR Present in whole testis extract, This effect was non-reversible upon drug cessation, suggesting damage to the germinal stem cell popula- tion. Whether this effect is owing to PDE11 inhibi- tion or to another non-specific effect of tadalafil remains unknown. Interestingly, this negative effect on spermatogenesis was not observed when mice were given similarly high doses of the medica- 31,32 human sequence of PDE11A tion. It can be speculated that these differences are owing to species-specific patterns of PDE11 expression. Unfortunately, there are no published

refer to the antibody by name only (EPH3). We identified this with the similarly named anti-serum in Fawcett studies on PDE11 expression in canine testis. Clinical studies of the effect of tadalafil on 23,24 spermatogenesis in humans have yielded mixed Human Mouse Primers from results. In a double-blind cross-over study, Pomara et al.26 found that acute tadalafil administration was

16 associated with a statistically significant decrease in 22 . sperm straight line velocity in a group of 16 males Both abstracts Studies are sorted byPDE11A2, chronological PDE11A3 order and of PDE11A4, publication. respectively. In See ‘result’ text column, for the further notation discussion. A1, Amino-acid A2, locations A3 in or the A4 antibody refers column correspond to to putative the identification PDE11A1 of the de D’Andrea et al a Baxendale and Fraser spanning residues 410–424. who were less than 40 years old with a history of

International Journal of Impotence Research Phosphodiesterase 11 A Makhlouf et al 508 infertility. It is unknown whether the inhibition no studies to date have investigated PDE11 activity of PDE11 is the mechanism behind the observation in infertile males, and the possibility of PDE11 or whether the changes in the parameters will malfunction underlying cases of infertility remains translate to impact clinical fertility rates. In a larger purely speculative. 6-month, double-blind, placebo-controlled study, In conclusion, we have summarized what is Hellstrom et al.33 looked at the effects of 10 or known about PDE11 structure, expression and func- 20 mg of tadalafil on semen parameters and repro- tion so far. Clearly new studies are needed to further ductive hormones in 421 men older than 45 without our understanding of this PDE family. a history of infertility. Inclusion criteria included a sperm concentration of 420 million, 50% sperm motility and 50% normal sperm morphology. Those with a history of infertility or abnormal hormone References profile were excluded. The study found that in men older than 44, tadalafil had no significant affect 1 Lin CS, Xin ZC, Lin G, Lue TF. Phosphodiesterases as on sperm concentration, morphology or motility therapeutic targets. Urology 2003; 61: 685–691. nor had any affect on investigated hormone levels of 2 Saenz de Tejada I, Angulo Frutos J, Gaudo M, Florio V. testosterone, LH or FSH. It can be surmised that the Comparative selectivity profiles of Tadalafil, Sildenafil and Vardenafil using an in vitro phosphodiesterase activity assay. affect of PDE11 inhibition, if any, by this medication Int J Impot Res 2002; 14(Suppl 4): S20–S32. was not detected or significant. Any potential effect 3 Gbekor E, Bethel S, Fawcett L, Mount N, Phillips S. Selec- of long-term tadalafil use on fertility parameters in tivity of sildenafil and other phosphodiesterase type 5 (PDE5) men with a history of infertility or younger than age inhibitors against all human phosphodiesterase families. Eur Urol Suppl 2002; 1(Suppl 1): 63. 45 years, however, remains unproven. 4 Maw GN, Allerton CM, Gbekor E, Million WA. Design, PDE11 À/À knockout mice were recently described synthesis and biological activity of beta-carboline-based 25 by Wayman et al. Knockout mice were found to be type-5 phosphodiesterase inhibitors. Bioorg Med Chem Lett overall phenotypically normal without any gross or 2003; 13: 1425–1428. lethal abnormalities. However, a small but measurable 5 Weeks JL, Zoraghi R, Beasley A, Sekhar KR, Francis SH, Corbin JD. High biochemical selectivity of tadalafil, sildenafil decrease in sperm concentration of ejaculated sperm 3 and vardenafil for human phosphodiesterase 5A1 (PDE5) over from 29.3 to 23.8 million/cm was detected. This PDE11A4 suggests the absence of PDE11A4 cross-reaction in difference was accentuated by a slight increase in patients. Int J Impot Res 2005; 17: 5–9. necrospermia in the PDE À/À knockout (45% viable 6 Bischoff E. Potency, selectivity, and consequences of nonselec- tivity of PDE inhibition. Int J Impot Res 2004; 16(Suppl 1): sperm in PDE À/À vs 56% in wild type). In addition, S11–S14. pre-ejaculated sperm from knockout mice showed a 7 Pyne NJ, Furman BL. phosphodiesterases in 53% increase in the capacitated fraction, suggesting pancreatic islets. Diabetologia 2003; 46: 1179–1189. premature capacitation. However, it should be 8 Manallack DT, Hughes RA, Thompson PE. The next generation stressed that the knockout mice had fertility rates of phosphodiesterase inhibitors: structural clues to ligand and substrate selectivity of phosphodiesterases. J Med Chem 2005; similar to those of wild-type mice, and were able to 48: 3449–3462. reproduce without difficulty. This suggests that 9 Maurice DH, Palmer D, Tilley DG, Dunkerley HA, Netherton redundant mechanisms, such as other PDE isoforms, SJ, Raymond DR et al. Cyclic nucleotide phosphodiesterase may have mitigated the absence of PDE11. activity, expression, and targeting in cells of the cardiovas- cular system. Mol Pharmacol 2003; 64: 533–546. 10 Yuasa K, Kotera J, Fujishige K, Michibata H, Sasaki T, Omori K. Isolation and characterization of two novel phosphodies- Future directions terase PDE11A variants showing unique structure and tissue- specific expression. J Biol Chem 2000; 275: 31469–31479. Most of the work on PDE11’s function has focused 11 Yuasa K, Ohgaru T, Asahina M, Omori K. Identification of rat cyclic nucleotide phosphodiesterase 11A (PDE11A): compar- on spermatogenesis. However, recall that the pros- ison of rat and human PDE11A splicing variants. Eur J tate is the site most consistently associated with Biochem 2001; 268: 4440–4448. PDE11 expression. It is possible that changes in 12 Hetman JM, Robas N, Baxendale R, Fidock M, Phillips SC, sperm function seen with inhibitors of PDE11 are Soderling SH et al. Cloning and characterization of two splice variants of human phosphodiesterase 11A. Proc Natl Acad Sci owing to changes in the contribution of the prostate USA 2000; 97: 12891–12895. to the seminal fluid, a contribution known to 13 Fawcett L, Baxendale R, Stacey P, McGrouther C, Harrow I, 30 contain factors required for sperm capacitation. Soderling S et al. Molecular cloning and characterization of a Another area of research is idiopathic infertility distinct human phosphodiesterase gene family: PDE11A. Proc owing to asthenospermia. One notable difference Natl Acad Sci USA 2000; 97: 3702–3707. 14 Yuasa K, Kanoh Y, Okumura K, Omori K. Genomic organiza- between studies showing no deleterious effect of tion of the human phosphodiesterase PDE11A gene. Evolu- tadalafil vs those showing changes in sperm para- tionary relatedness with other PDEs containing GAF domains. meters is that the patients studied in the former were Eur J Biochem 2001; 268: 168–178. normal fertile volunteers, whereas the latter con- 15 Zoraghi R, Corbin JD, Francis SH. Properties and functions of 26,33 GAF domains in cyclic nucleotide phosphodiesterases and sisted of infertility clinic patients. This raises other proteins. Mol Pharmacol 2004; 65: 267–278. the possibility that infertile men may have a 16 Baxendale RW, Fraser LR. Mammalian sperm phospho- different response to PDE11 inhibition. However, diesterases and their involvement in receptor-mediated cell

International Journal of Impotence Research Phosphodiesterase 11 A Makhlouf et al 509 signaling important for capacitation. Mol Reprod Dev 2005; 71: corpus cavernosum and the contribution of PDE11 inhibition 495–508. on the nerve stimulated relaxation. J Urol 2001; 165(Suppl): 17 Turko IV, Francis SH, Corbin JD. Binding of cGMP to both 340. allosteric sites of cGMP-binding cGMPspecific phosphodies- 25 Wayman C, Phillips S, Lunny C, Webb T, Fawcett L, Baxendale terase (PDE5) is required for its phosphorylation. Biochem J R et al. Phosphodiesterase 11 (PDE11) regulation of sperma- 1998; 329: 505–510. tozoa physiology. Int J Impot Res 2005; 17: 216–223. 18 Rybalkin SD, Rybalkina IG, Shimizu-Albergine M, Tang XB, 26 Pomara G, Morelli G, Turchi P, Caglieresi C, MOschini C, Beavo JA. PDE5 is converted to an activated state upon Canale D et al. Effect of acute in vivo sildenafil or tadalafil cGMP binding to the GAF A domain. EMBO J 2003; 22: treatments on semen parameters in patients with fertility 469–478. problem, a randomized, double-blind, crossover study. J Sex 19 Mullershausen F, Friebe A, Feil R, Thompson WJ, Hofmann F, Med 2004; 2(Suppl 2): 23 (PS-5-9). Koesling D. Direct activation of PDE5 by cGMP: long-term 27 Sinclair ML, Wang XY, Mattia M, Conti M, Buck J, Wolgemuth effects within NO/cGMP signaling. J Cell Biol 2003; 160: DJ et al. Specific expression of soluble adenylyl cyclase in 719–727. male germ cells. Mol Reprod Dev 2000; 56: 6–11. 20 Pomara G, Morelli G. Re: Montorsi et al. Vardenafil for the 28 Nantel F, Monaco L, Foulkes NS, Masquilier D, LeMeur M, treatment of erectile dysfunction: a critical review of the Henriksen K et al. Spermiogenesis deficiency and germ-cell literature based on personal clinical experience. Eur urol in CREM-mutant mice. Nature 1996; 380: 159–162. 2005; 47: 612–621. 29 de Lamirande E, Leclerc P, Gagnon C. Capacitation as a 21 Loughney K, Taylor J, Florio VA. 30,50-cyclic nucleotide regulatory event that primes spermatozoa for the acrosome phosphodiesterase 11A: localization in human tissues. Int J reaction and fertilization. Mol Hum Reprod 1997; 3: 175–194. Impot Res 2005; 17: 320–325. 30 Kshirsagar A, Niederberger C. Are prostates and male 22 D’Andrea MR, Qiu Y, Haynes-Johnson D, Bhattacharjee S, infertility related? AUANEWS 2005; 10: 9–10. Kraft P, Lundeen S. Expression of PDE11A in normal and 31 Cialiss (Tadalafil) label insert. http://www.fda.gov/cder/foi/ malignant human tissues. J Histochem Cytochem 2005; 53: label/2005/021368s006lbl.pdf, p 19. 895–903. 32 Sausen PJ, Reams RY, Morford LL, Dietsch GN. Tadalafil 23 Baxendale RW, Burslem F, Phillips SC. Phosphodiesterase has no effect on testes in rodents. Int J Impot Res 2002; type 11 (PDE11) cellular localization: progress towards 14(Suppl 4): S20–S32. defining a physiological role in testis and/or reproduction. 33 Hellstrom WJ, Overstreet JW, Yu A, Saikali K, Shen W, Beasley J Urol 2001; 165(Suppl): 340. Jr CM et al. Tadalafil has no detrimental effect on human 24 Baxendale RW, Wayman CP, Turner L, Phillips SC. Cellular spermatogenesis or reproductive hormones. J Urol 2003; 170: localization of phosphodiesterase type 11 (PDE11) in human 887–891.

International Journal of Impotence Research