Isolation of Two Isoforms of Phosphodiesterase 5 from Rat Penis

International Journal of Impotence Research (2003) 15, 129–136 & 2003 Nature Publishing Group All rights reserved 0955-9930/03 $25.00 www.nature.com/ijir Isolation of two isoforms of phosphodiesterase 5 from rat penis

C-S Lin1*, G Lin1 and TF Lue1

1Knuppe Molecular Urology Laboratory, Department of Urology, University of California San Francisco, California, USA

Inhibition of cGMP-specific phosphodiesterase type 5 (PDE5) has been shown to improve penile erection in patients with erectile dysfunction. We have reported previously the cloning of three PDE5 isoforms from human penile tissues. Here we report the cloning of two PDE5 isoforms from rat penile tissues. The similarity between rat and human PDE5A1-specific sequences were 68 and 88% at the nucleotide and amino-acid levels, respectively. Like the bovine and canine PDE5A1 sequences, the rat PDE5A1 sequence lacks the polyglutamine tract that appears to be unique to the human PDE5A1 sequence. The similarity between rat and human PDE5A2-specific sequences were 64 and 100% at the nucleotide and amino-acid levels, respectively. The equivalent of human PDE5A3-specific sequence was identified in the rat PDE5A gene; however, repeated efforts to clone the putative rat PDE5A3 isoform were not successful. Expression of PDE5A1 and A2 mRNA in various tissues was examined by Northern blotting and reverse transcription–polymerase chain reaction. Results from the two experimental procedures were largely in good agreement and indicated that PDE5A1 and A2 mRNA were expressed in a tissue-specific manner with PDE5A2 being the dominant isoform. International Journal of Impotence Research (2003) 15, 129–136. doi:10.1038/sj.ijir.3900983

Keywords: rat; PDE5; isoforms; penis; cloning; tissue distribution

Introduction the treatment of a wide variety of ailments, includ- ing congestive heart failure, asthma, erectile dys- function (ED), etc.3 Phosphodiesterases (PDEs) are intracellular en- All PDE families except PDE6 have been detected zymes that catalyze the degradation of cAMP and/ at either the RNA or protein level in the human or cGMP. By counterbalancing adenyl and guanyl penile corpus cavernosum;4,5 however, only PDE5 cyclases that, respectively, catalyze the formation of has been unequivocally shown to regulate erectile cAMP and cGMP, PDEs serve to adjust the cellular function.5 While erection is induced by the accu- concentrations of cAMP and cGMP, thereby influen- mulation of cGMP after erotic stimuli trigger the 1 cing cellular functions. Despite having only two release of nitric oxide that activates guanyl cyclase, substrates, PDEs in the mammalian tissues are detumescence requires the hydrolysis of cGMP, comprised of 11 families (PDE1–PDE11) encoded which is catalyzed predominantly by PDE5. By from 21 distinct genes with each gene encoding selectively inhibiting the catalytic activity of PDE5, multiple isoforms. These 11 PDE families are 20– the widely prescribed sildenafil citrate allows the 45% homologous in their catalytic domain located concentration of cGMP to reach a level sufficient for near the C-terminal. Their N-terminals are divergent a satisfactory erection even in many ED patients. and may contain regulatory domains that bind to One of the side effects of sildenafil is a blue tinge to 1,2 calcium ions, calmodulin, proteins, and/or cGMP. vision, which is believed to be caused by cross- PDE1, 2, 3, 10, and 11 can hydrolyze both cAMP and reactivity with PDE6, a structurally PDE5-related cGMP, PDE4, 7, and 8 only cAMP, and PDE5, 6 and 9 molecule present in high concentration in the cone 1 only cGMP. Since their discovery in the late 1950s, and rod cells of the retina. Other side effects of both natural and synthetic compounds that inhibit sildenafil, such as headache, facial flushing, rhini- the catalytic activities of various PDEs have been tis, and dyspepsia, are believed to be because of developed and continue to be actively exploited for inhibition of PDE5 in the vascular and visceral smooth muscle in the relevant organs. We have reported the cloning of three PDE5 isoforms from human penile corpus cavernosum.6 *Correspondence: C-S Lin, Knuppe Molecular Urology Structurally, these isoforms differ only in the 50 end Laboratory, Department of Urology, University of Califor- nia San Francisco, CA 94143-1695, USA. of their messenger RNAs and the corresponding E-mail: [email protected] amino terminal of their proteins. By using isoform- Received 16 October 2002; revised 26 November 2002; specific oligonucleotide primers in reverse tran- accepted 20 December 2002 scription–polymerase chain reactions (RT–PCR), we PDE5 isoforms in rat penis C-S Lin et al 130 were able to show that PDE5A1 and A2 were I (Roche Molecular Biochemicals, Pleasanton, CA, expressed in a variety of tissues while PDE5A3 USA) at 371C for 30 min. The RNA was then purified was restricted to smooth muscle tissues. When by phenol/chloroform extraction and ethanol pre- assayed with lysates of transfected COS-7 cells, the cipitation. Quantity and purity of RNAs were three isoforms had similar cGMP Km of around 6 mM. measured by spectrophotometry with UV adsorption Sildenafil inhibited PDE5A1, A2, and A3 with IC50 at wavelengths 260 and 280. Integrity of RNAs was values of 28, 14, and 13 nM, respectively. assessed by the sharpness of the 28 s and 18 s In order to advance our understanding of the ribosomal RNA bands as visualized in agarose gels. function and tissue distribution of PDE5 isoforms, it is necessary to conduct experiments in an animal model such as the rat. Furthermore, the develop- Oligonucleotide primers ment of future PDE5 inhibitors also requires testing in an animal model such as the rat. As such, it is Several oligonucleotide primers used in polymerase important that we identify and isolate rat PDE5 chain reaction (PCR) experiments are listed in Table 1. isoforms. While rat PDE5A2 has been isolated,7 whether there are other PDE5 isoforms has not been reported. In the present study, we isolated PDE5A1 RACE–PCR and PDE5A2 isoforms from rat penis; we then compared their tissue distribution. Rapid amplification of cDNA ends–polymerase chain reaction (RACE–PCR)8 was used to identify and to Materials and methods isolate the 50 ends of PDE5 isoforms. Rat penile cavernosum RNAs were annealed to primer PDE5-RT (Table 1; primer 1, Figure 1) and reverse transcribed. RNA preparation The resulting single-stranded cDNA was then sub- jected to ‘tailing’ with dATP at the 30 end. The ‘‘A- Rat tissues were homogenized in Tri-Reagent RNA , extraction solution (Molecular Research Center, 5 Cincinnati, OH, USA). Following the recommended 1 procedure by the supplier, each tissue RNA (20– 50 mg) was treated with 10 units of RNase-free DNase RT , Table 1 Oligonucleotide primers used in RACE-PCR and RT- 3 PCR for PDE5 isoforms Tailing Name Sequence Use AAAAAAAAA RACE–PCR Annealing PDE-RT AATTCAGAGGCAGAGAT RT primer PDE560a CCTTCCTTGCACACAGGAAT Primer 1 (Figure 1) AAAAAAAAA PDE5-Nsi AACCATGCATTGACCATTTC Primer 2 (Figure 1) TTTTTTTTT 2 UAP GACTCGAGTCGACATCG Universal primer (T tail not shown) PCR

RT–PCR AAAAAAAAA TTTTTTTTT PDE5A-s TGATCACCGGGACTTTACCT PDE5A1 specific PDE5B-s TGCTATGTTGCCCTTTGGAG PDE5A2 specific rPDE5A3 AGCCTGGGTGTAGCAGCTT Rat PDE5A3 a Figure 1 Schematic representation of the RACE–PCR procedure. specific The top line represents a PDE5 mRNA with known sequence PDE797a GAGCACTGGTCCCCTTCAT Common (filled bar) and unknown sequence (open bar). A PDE5-specific downstream primer (arrow1; PDE5-RT, Table 1) was derived from the 50 end of primer the known sequence and used in RT. The resulting cDNA was Actin-s TCTACAATGAGCTGCGTGTG Upstream then ‘‘tailed’’ with a string of nucleotide A (from dATP) with primer terminal deoxynucleotide transferase. This artificial poly (A) tail Actin-a AATGTCACGCACGATTTCCC Downstream served as the annealing site for an upstream (universal) primer primer (Table 1) that contains the complementary poly (T) tail and additional sequences ( ) that include ‘‘rare-cutter’’ restric- aThe sequence of this primer (see Figure 4 also) is derived from rat tion sites such as SalI( ). The universal primer and a second PDE5A gene at a location equivalent to the human PDE5A3- PDE5-specific primer (arrow 2; PDE560a, Table 1) were used in specific exon. No rat mRNA has been detected by this primer PCR to amplify the cDNAs that correspond to the 50 end of the (with PDE797a as downstream primer). mRNA. The amplification products were then cloned and sequenced.

International Journal of Impotence Research PDE5 isoforms in rat penis C-S Lin et al 131 tailed’’ cDNA was then used as template in PCR, in reverse transcriptase. A total of 80 ml of TE buffer was which three primers were used, as described in the then added to make a 5 Â diluted library. A portion following. The PDE560a primer (Table 1; primer 2, of this library was further diluted to various Figure 1) served as a downstream primer. The dT-UAP concentrations (up to 100 Â dilution). In all, 1 mlof primer (Table 1; Figure 1) was used to anneal to the A- each dilution was then used in a 10 ml PCR to tailed cDNA (at an annealing temperature of 451C), and identify the optimal input within the linear ampli- the UAP primer (Table 1; Figure 1) was used to further fication range. In addition to the 1 ml diluted library, amplify the cDNAs at higher stringency (annealing the PCR mixture consisted of 10 ng of each of a temperature of 551C). The RACE–PCR products were primer pair and reagents supplied with the cloned into pMECA plasmid9 and sequenced by the Taq polymerase (Life Technologies, Inc., Gaithers- dideoxy chain termination method.10 burg, MD, USA). PCR was performed in the DNA Engine thermocycler (MJ Research, Inc., Watertown, MA, USA) under calculated temperature control. RT–PCR analysis The cycling program was set for 35 cycles of 941C, 5s; 551C, 5 s; 721C, 10 s, followed by one cycle of 721C, 5 min. The PCR products were electrophoresed RT–PCR was performed in an RT step and a PCR in 1.5% agarose gels in the presence of ethidium step. In the RT step, the cellular mRNAs were reverse bromide, visualized by UV fluorescence, and re- transcribed into a ‘‘library’’ of complementary DNAs corded by a digital camera connected to a computer. (cDNAs). This cDNA library was then used for the analysis of various genes in the PCR step. The RT procedure was performed with the SuperScript Northern blot analysis reverse transcriptase (Life Technologies, Inc., Gaithersburg, MD, USA) and its accompanying reagents. Briefly, 2.5 mg of each tissue RNA was Rat multiple tissue Northern blots were purchased annealed to 0.4 mg of oligo-dT primer in a volume of from Clontech Inc. (Palo Alto, CA, USA). Each of the 12 ml. A volume of 4 mlof5Â buffer, 2 ml of 0.1 M two membranes contains eight different rat tissue DTT, 1 ml of 10 mM dNTP, and 1 ml of Super- polyA+ RNAs at 2 mg/tissue/lane. Hybridization to Script reverse transcriptase was then added 32P-labeled cDNA probes was carried out with to bring the final reaction volume to 20 ml. After reagents supplied with the membranes at the 1 h of incubation at 421C, the RT mixture recommended conditions. Labeling of cDNA probes was incubated at 701C for 10 min to inactivate the was done by random priming using the Rediprime II

rPDE5A1 GAAAGACTGCAGGGCAAAGGGGGAGGGTCTCGAACACT-GTCCTGCTCCTCTGAGAGAGGGACCCTGC 66 | ||||| | ||||||||||||||||| | |||||| || |||||| || |||||| hPDE5A1 CGACTGGAGCAGGACGAAGGGGGAGGGTCTCGAGGCCGAGTCCTGTTCTTCTGAGGGACGGACCCCAG 68

M E R A G P rPDE5A1 C--GGGTGGAAATCCGGCATCAGCGAGCCGCAGAAACCCGCGGCAAACACCATGGAACGAGCGGGCCC 133 | ||||||||| | | | ||| ||||| | || | ||||| |||||||| || || ||||| hPDE5A1 CTGGGGTGGAAAAGCAGTACCAGAGAGCCTCCGAGGCGCGCGGTGCCAACCATGGAGCGGGCCGGCCC 136 M E R A G P

S S V Q S ------Q Q Q R D Q D rPDE5A1 CAGCTCCGTGCAGTCG------CAGCAGCAGCGGGACCAGGACT 171 ||||| || |||| | ||||||||| |||| ||||||| hPDE5A1 CAGCTTCGGGCAGCAGCGACAGCAGCAGCAGCCCCAGCAGCAGAAGCAGCAGCAGAGGGATCAGGACT 204 S F G Q Q R Q Q Q Q P Q Q Q K Q Q Q R D Q D

W V E A W L D D H R D F T F S Y F V R K A T R rPDE5A1 GGGTGGAAGCGTGGCTGGATGATCACCGGGACTTTACCTTCTCTTACTTTGTTAGAAAGGCCACCAGA 239 ||| ||||| |||||||| |||||| |||||||||||||||| |||||||||||||| ||||||||| hPDE5A1 CGGTCGAAGCATGGCTGGACGATCACTGGGACTTTACCTTCTCATACTTTGTTAGAAAAGCCACCAGA 272 S V E A W L D D H W D F T F S Y F V R K A T R Figure 2 Comparison of the isoform-specific regions of rat and human PDE5A1 cDNA and amino-acid sequences. Identical nucleotides between the rat and human sequences are indicated by vertical lines. Gaps within each sequence are introduced for maximal homology alignment. The deduced amino-acid (single-letter code) sequences are shown above (for rat) and below (for human) the respective nucleotide sequences. GenBank accession numbers for the rat and human PDE5A1 sequences are AY155460 and AF155192, respectively.

International Journal of Impotence Research PDE5 isoforms in rat penis C-S Lin et al 132 kit (Amersham Life Sciences Inc., Arlington Heights, PDE5 isoforms by RACE–PCR. As reported pre- IL, USA). The membrane set was first hybridized to a viously,6 alignment of the published bovine PDE5A1 rat PDE5A1 isoform-specific probe (see text). After and rat PDE5A2 sequences allowed us to design an development of the autoradiograph, the membranes RT primer (primer 1, Figure 1; PDE-RT, Table 1) and were stripped of the probe and then hybridized to a a PCR primer (primer 2, Figure 1; PDE560a, Table 1). rat PDE5A2 isoform-specific probe (see text). After These primers were used in three separate RACE– development of the second autoradiograph, the PCR experiments with rat penile RNA samples as membranes were stripped of the probe, hybridized templates and resulted in the isolation of 22 PDE5 to a 32P-labeled b-actin cDNA probe (supplied by cDNA clones. Nine of these clones were PDE5A1 Clontech), and autoradiographed. and the others were PDE5A2, as determined by comparing to human PDE5A1 and PDE5A2 se- Results quences (Figures 2 and 3). The similarity between rat and human PDE5A1-specific sequences were 68 and 88% at the nucleotide and amino-acid levels, Identification and cloning of rat PDE5 isoforms respectively. Like bovine and canine PDE5A1 sequences,11,12 the rat PDE5A1 sequence lacks the polyglutamine (polyQ) tract that appears to be Since PDE5 isoforms apparently differ only at the 50 unique to the human sequence.6,13–15 The similarity ends of their mRNAs, we sought to clone possible rat between rat and human PDE5A2-specific sequences

rPDE5A2 GTGGAGCGGAG--TGCGTGGCGGGCGGGTCCCCCGAAGCTCGAGGAGTCGCCGCGAAGCCCGCGCGCG 66 ||||||| ||| ||| |||||||||||| ||||||||| ||||| | | || || | | | ||| hPDE5A2 GTGGAGCAGAGAGTGCTTGGCGGGCGGGTGCCCCGAAGCCTGAGGAATTGATGCATAGTCGGGGTGCG 68

rPDE5A2 CGCTTGCCTCCGGCGCGCCGATCTGGGCTGAACTAACAAGCTCTGCTTTGGAGAAGGGGCTGGAGCCC 134 | | |||| | ||| | ||||| |||| | | | |||| hPDE5A2 TG---GACTCCAGGACGCG------CCCCTTTGCAGAAAGAGTCCGTGCCC 110

rPDE5A2 CGCCACCGGCGCCTTGGGGCAACGGCCGTCGAGCAGCTGGTGCTGGCCGCCCGAGCCGGGAGGTCGAC 202 ||| ||| | || ||||| ||||||||| ||||| ||||||||||| | ||||||||| ||| hPDE5A2 CGCTGCCGCCTGCTCTGGGCAGCGGCCGTCGGGCAGCGGGTGCTGGCCGTCTTGGCCGGGAGGGTGAC 178

rPDE5A2 TCTCGCCGGGCGGGGATGGTGGACGAGCCGGCGGGG------ACTCGCCCGCGCGC 252 | |||| ||| | |||||| || ||||| || || ||||||| |||||| hPDE5A2 ACACGCCCGGCAGTGATGGTCGAGGAGCCCGCCCGGGTTCGACGCCCGGGTCCCACTCGCCGGCGCGC 246

rPDE5A2 GGCCGCCACAGGGCCCGGCGCTGCCCGCCAGGGGG-CGACGCCACGCGGGGTCCGCAGCCGGCGACCC 319 |||||| | ||| | |||| |||||||||||| || ||| ||||||| || | |||||| hPDE5A2 GGCCGCGGCTCTGCCGGTCGCTCCCCGCCAGGGGGGCGGCGCAGCGCGGGGCGAGCGGGAGGCGACGG 314

rPDE5A2 GCAGCGAAGAGGCGGCGGGGGCCATAAGGGAGCCGGTCGGGCACCCACGG------AACT 373 || ||| | || ||| |||||| | ||||||| ||| | | || | || hPDE5A2 GCGGCGGCGCGGAGGCCGGGGCCGTGAGGGAGCTGGTTCTGAGTCGCCGACCTCACCTCACCTCACCT 382

rPDE5A2 CCGCGGG------CCCCCAGAGTCTCCCC-AGTCCACCATCCGAGGAGTTCGGAAACGAGCGCCCC 432 |||| | || | ||| |||| |||||| ||| | |||||||| |||||| | ||| hPDE5A2 GCGCGCGTTGTTGCCCCTACGGAGCCTCCTGGAGTCCAGGATCGGCGGAGTTCGAAAACGAACTTCCC 450

M L P F G D K T R rPDE5A2 ACCTCTGCTATGTTGCCCTTTGGAGACAAAACGAGA 468 || | ||||||||||||||||||||||||||| ||| hPDE5A2 ACGTTTGCTATGTTGCCCTTTGGAGACAAAACAAGA 486 M L P F G D K T R Figure 3 Comparison of the isoform-specific regions of rat and human PDE5A2 cDNA and amino-acid sequences. Identical nucleotides between the rat and human sequences are indicated by vertical lines. Gaps within each sequence are introduced for maximal homology alignment. The deduced amino-acid (single-letter code) sequences are shown above (for rat) and below (for human) the respective nucleotide sequences. GenBank accession numbers for the rat and human PDE5A2 sequences are AY155461 and AF155193, respectively.

International Journal of Impotence Research PDE5 isoforms in rat penis C-S Lin et al 133 rPDE5A3 GCTTGCTCTGACCCTGGAGGGGCGGGGAAGCCGGGGGTCTGGACCAGGCCACCGGGAAGCCTGGGTGT 68 || || ||| |||||| |||||||||||| |||| |||| || |||| || |||| hPDE5A3 CCTCGCCCTGGCCCTGGCTGGGCGGGGAAGCTGGGGTGAGGGACGAGTCCACGGGACAGCCCAAA--- 65

rPDE5A3 AGCAGCTTGCCCCGGTCGCCCCCA 92 ||| | || | | || | hPDE5A3 GGCAACATGACGGAACCTTGCCAAA 90 Figure 4 Comparison of the putative rat PDE5A3 isoform-specific sequence and the human PDE5A3 isoform-specific sequence. The putative rPDE5A3 sequence is derived from rat PDE5A gene; its actual cDNA has not been identified. Identical nucleotides between the rat and human sequences are indicated by vertical lines. Gaps within each sequence are introduced for maximal homology alignment. GenBank accession numbers for the rat gene and human PDE5A3 sequences are AB017578 and AF155193, respectively. The underlined sequence was used for the design of the rPDE5A3-specific primer (Table 1). were 64 and 100% at the nucleotide and amino-acid levels, respectively.

Lack of PDE5A3 isoform expression

Since all the above-isolated cDNA clones were either PDE5A1 or PDE5A2, we attempted to isolate the putative rat PDE5A3 isoform by performing one additional RACE–PCR experiment with rat prostate tissue. This resulted in the isolation of several additional PDE5A1 and PDE5A2 clones but not any PDE5A3 (data not shown). By aligning the human PDE5A3-specific sequence with the rat PDE5A gene sequence (GenBank accession number AB017578), an apparent rat PDE5A3-specific exon Figure 5 RT–PCR analysis for PDE5A1 and PDE5A2 expression was identified (Figure 4). The human and rat in rat tissues. Primer pair PDE5A-s and PDE-797a was used for sequences are not only highly similar (61%) but PDE5A1, primer pair PDE5B-s and PDE797a for PDE5A2, and also located in a comparable region between the A1- primer pair Actin-s and Actin-a for b-actin (Table 1). The reaction and A2-specific exons in their respective genes (data products were electrophoresed in a 1.5% agarose gel and stained with ethidium bromide. The RT–PCR product in each lane was not shown). Based on the sequence homology, we derived from 25 ng of tissue RNA. Size marker is a 100-bp ladder. designed a rat ‘‘PDE5A3-specific’’ primer (Figure 4; Table 1) and used it in several RT–PCR experiments with rat penis, prostate, and aorta tissues. The showed that both PDE5A1 and A2 mRNAs are results were all negative, further suggesting the lack approximately 7.5 kb in length and their expression of PDE5A3 expression in the rat. levels in various tissues were generally in agreement with the RT–PCR results. One exception is that whereas RT–PCR detected low isoform expression Tissue distribution of PDE5 isoforms levels in the stomach, Northern blotting detected high expression levels, especially for the A2 iso- form. This is probably because of differences in the Based on the rat PDE5A1- and A2-specific se- selection of tissue sites for RNA isolation. For ease quences, we designed oligonucleotide primers (Fig- of comparison, the RT–PCR and Northern blot ures 2, 3; Table 1) for RT–PCR analysis on various rat results are summarized in Table 2. tissues. The results (Figure 5) showed that PDE5A1 mRNA was expressed in fewer tissues and at lower levels than PDE5A2 mRNA. Particularly, heart, liver, Discussion and skeletal muscle appear to express small amounts of PDE5A2 but no PDE5A1. Generally, organs that contain substantial amounts of smooth We previously reported the cloning of three distinct muscle (aorta, intestines, prostate, urethra, urinary PDE5 isoforms from humans. Since most of our in bladder, and uterus) tend to express more PDE5 vivo research was conducted with rat models, we mRNA, particularly the A2 isoform. To assess the felt it was necessary to isolate the rat PDE5 isoforms. expression level more precisely, Northern blot By using an identical approach, that is, RACE–PCR analysis was also carried out. The results (Figure 6) cloning, we succeeded in the isolation of rat

International Journal of Impotence Research PDE5 isoforms in rat penis C-S Lin et al 134

Figure 6 Northern blot analysis for PDE5A1 and PDE5A2 expression in rat tissues. A set of two Northern blot membranes was purchased from Clontech Inc. Each membrane contains eight different rat tissue mRNAs at 2 mg per tissue per lane. The membranes were first hybridized to a 32P-labeled rat PDE5A1 isoform-specific probe (Figure 2). After development of the autoradiograph (top 2 panels), the membranes were stripped of the probe and then hybridized to a 32P-labeled rat PDE5A2 isoform-specific probe (Figure 3). After development of the autoradiograph (bottom 2 panels), the membranes were stripped of the probe and then hybridized to a 32P-labeled b- actin cDNA probe. The final autoradiograph is shown here as the two middle strips. Size markers shown on the left of each panel are in kilobases (kb).

International Journal of Impotence Research PDE5 isoforms in rat penis C-S Lin et al 135 Table 2 Summary of expression levels of PDE5A1 and PDE5A2 PDE5A1 and PDE5A2 in tissue distribution. Since mRNA in various rat tissues RT–PCR is less quantitative than Northern blot analysis, we decided to compare tissue distribution RT–PCR Northern of PDE5 isoforms with both methods in the present study. As expected, RT–PCR was more sensitive and Tissue PDE5A1 PDE5A2 PDE5A1 PDE5A2 detected PDE5 expression (in particular, the A1 Adrenal gland ND ND ÀÀisoform) in tissues that were negative with the Aorta + +++ ND ND Northern blot analysis (eg, liver and skeletal mus- Brain + ++ + +++ cle). As summarized in Table 2 for ease to compare, Eye ND ND + +++ results from the two methods were largely in good Esophagus + + ND ND agreement, indicating that the RT–PCR method, Heart À +++ Kidney + ++ + ++ despite its ease of use, is reliable. It should be Large intestine + ++++ + ++++ pointed out that it remains unclear why the North- Liver À + ÀÀern method detected multiple RNA species in the Lung + ++ ++ +++ stomach with the A1-specific probe. Since the Ovary + +++ ND ND Penis + ++ ND ND stomach was also where the RT–PCR and Northern Prostate ++ +++ ++ ++++ results differed most, whether there was something Skeletal muscle À ++ ÀÀpeculiar to the preparation of the stomach RNA for Small intestine ++ +++++ ND ND the Northern blotting remains an open question. Spinal cord + ++ À + In summary, the present study demonstrated for Spleen ND ND ÀÀ Stomach + + ++ +++++ the first time the existence of PDE5A1 in the rat. It Testis ND ND ÀÀalso points out that, despite the existence of Thyroid ND ND + ++ PDE5A3-like sequence in the rat genome, PDE5A3 Urethra + +++ ND ND mRNA is not likely expressed. Finally, we for the Urinary bladder + +++ ND ND Uterus ++ +++ ND ND first time compared PDE5 isoform expressions with both the RT–PCR and Northern methods. The results ND: not determined; À: not detected; +: lowest level; +++++: indicated that PDE5A1 and A2 were expressed in a highest level. tissue-specific manner with PDE5A2 being the dominant isoform. Whether these isoforms are regulated by diseases (eg, ED) or drugs (eg, sildena- PDE5A1 and PDE5A2 but not PDE5A3. Rat PDE5A1 fil) is currently under investigation. is highly similar to bovine and canine PDE5A1 and together they differ from human PDE5A1 in not having the polyglutamine tract. Since elongated polyglutamine tract causes protein misfolding and Acknowledgements aggregation and is associated with a number of neurodegenerative disorders in humans,16 it is tempting to imagine the existence of a ‘‘mutant’’ This work was supported in part by grants from the PDE5A1 that may also be associated with neurode- California Urology Foundation (to C-S Lin) and from generative diseases. A recent discovery that andro- NIH (2R01-DK-45370, to TF Lue). gen receptor with elongated polyglutamine tract forms aggregates that alter axonal trafficking and mitochondrial distribution in motor neuronal pro- cesses17 further suggests the need to look for this References possibility. Despite repeated attempts to isolate or identify the 1 Francis SH, Turko IV, Corbin JD. Cyclic nucleotide phospho- putative rat PDE5A3, no such molecule was found. diesterases: relating structure and function. Prog Nucleic Acid By comparing the available rat and human PDE5A Res Mol Biol 2001; 65: 1 – 52. 2 Soderling SH, Beavo JA. Regulation of cAMP and cGMP gene sequences, we found them to be highly similar signaling: new phosphodiesterases and new functions. Curr in all three alternative first exons, including the A3- Opin Cell Biol 2000; 12: 174 – 179. specific exon. The homology includes not only the 3 Lin CS, Xin ZC, Lin G, Lue TF. Phosphodiesterases as transcribed regions but also their adjacent untran- therapeutic targets. Urology 2003; in press. scribed intronic sequences (data not shown). It is 4Ku¨the A et al. Expression of different phosphodiesterase genes in human cavernous smooth muscle. J Urol 2001; 165: 280 – therefore rather puzzling why the rat PDE5 pre- 283. mRNA could not be spliced into a PDE5A3-coding 5 Corbin JD, Francis SH, Webb DJ. Phosphodiesterase type 5 as a mRNA. More importantly, without the rat equiva- pharmacologic target in erectile dysfunction. Urology 2002; lent, elucidation of the function of human PDE5A3 60: 4 – 11. 6 Lin CS, Lau A, Tu R, Lue TF. Expression of three isoforms of would be more difficult. cGMP-binding cGMP-specific phosphodiesterase (PDE5) in RT–PCR is the only method that has been used human penile cavernosum. Biochem Biophys Res Commun previously to analyze the difference between 2000; 268: 628 – 635.

International Journal of Impotence Research PDE5 isoforms in rat penis C-S Lin et al 136 7 Kotera J et al. Expression of rat cGMP-binding cGMP-specific 13 Stacey P, Rulten S, Dapling A, Phillips SC. Molecular cloning phosphodiesterase mRNA in Purkinje cell layers during and expression of human cGMP-binding cGMP-specific postnatal neuronal development. Eur J Biochem 1997; 249: phosphodiesterase (PDE5). Biochem Biophys Res Commun 434 – 442. 1998; 247: 249 – 254. 8 Frohman MA, Dush MK, Martin GR. Rapid production of full- 14 Loughney K et al. Isolation and characterization of cDNAs length cDNAs from rare transcripts: amplification using a encoding PDE5A, a human cGMP-binding, cGMP-specific single gene-specific oligonucleotide primer. Proc Natl Acad 30,50-cyclic nucleotide phosphodiesterase. Gene 1998; 216: Sci USA 1988; 85: 8998 – 9002. 139 – 147. 9 Thomson JM, Parrott WA. pMECA: a cloning plasmid with 44 15 Yanaka N et al. Expression, structure and chromosomal unique restriction sites that allows selection of recombinants localization of the human cGMP-binding cGMP-specific based on colony size. BioTechniques 1998; 24: 922 – 928. phosphodiesterase PDE5A gene. Eur J Biochem 1998; 255: 10 Sanger F, Nicklen S, Coulson AR. DNA sequencing with 391 – 399. chain-terminating inhibitors. Proc Natl Acad Sci USA 1977; 16 Freiman RN, Tjian R. Neurodegeneration. A glutamine-rich 74: 5463 – 5467. trail leads to transcription factors. Science 2002; 296: 2149 – 11 McAllister-Lucas LM et al. The structure of a bovine lung 2150. cGMP-binding, cGMP-specific phosphodiesterase deduced 17 Piccioni F et al. Androgen receptor with elongated polyglu- from a cDNA clone. J Biol Chem 1993; 268: 22 863 – 22 873. tamine tract forms aggregates that alter axonal trafficking and 12 Kotera J et al. Novel alternative splice variants of cGMP- mitochondrial distribution in motor neuronal processes. binding cGMP-specific phosphodiesterase. J Biol Chem 1998; FASEB J 2002; 16: 1418 – 1420. 273: 26 982 – 26 990.

International Journal of Impotence Research