MYPT1 reduction is a novel pathogenic factor of erectile dysfunction

knockout Y-Y.Z., Y-W.Z and L-S.W helped cell culture; X.C., L-S.W and T.T. helped plasmid construction; L-S.W. and C-J.L. contributed data calibration; Z.W.; J.S.; Y.B. and M-S.Z. analyzed the data; W.Z., X-N.Z., Y.B., and M-S.Z. Abstract: Erectile dysfunction (ED) is closely associated with smooth muscle dysfunction, but its underlying mechanisms remains incompletely understood. We here reported that the reduced expression of myosin phosphatase target subunit 1 (MYPT1), the main regulatory unit of myosin light chain phosphatase, was critical for the development of vasculogenic ED. Male MYPT1 knockout mice had reduced fertility and the penises displayed abolished erections as evidenced by reduced intracavernous pressure (ICP). The penile smooth muscles of the knockout mice displayed enhanced contractility and resistant relaxation. We further identified a natural compound lotusine that increased the MYPT1 expression by inhibiting SIAH1/2 E3 ligases-mediated protein degradation. This compound sufficiently restored the ICP and improved histological characters of the penile artery of Mypt1 haploinsufficiency mice. In diabetic ED mice (db/db), the decreased expression of MYPT1 was measured, and ICP was restored by lotusine treatment. We conclude that the reduction of MYPT1 is the major pathogenic factor of vasculogenic ED. The restoration of MYPT1 by lotusine improved the function of injured penile smooth muscles, and could be a novel strategy for ED therapy.


Introduction
Erectile dysfunction (ED) is the inability to achieve or maintain a penile erection sufficient for satisfactory sexual intercourse 1,2 . Nearly 52% of men between 40-70 years of age are reported to have ED 3,4 , and the worldwide prevalence is predicted to reach 322 million in 2025 5 . An interesting clinical feature of ED is the close association with cardiovascular diseases, e.g. more than half of all cases of diabetes, hypertension, and atherosclerosis are complicated with ED 5-7 . Therefore, ED phenotype is usually regarded as the end result of various cardiovascular diseases, and many investigators even consider ED as a vascular disease.
Erection is accomplished via multiple processes 1,5 . In brief, the pudendal artery provides the penile body with a blood supply and branches to dorsal and central arteries penetrating the corpus cavernosal tissue [8][9][10] . Upon sexual stimulation, the penile nerve and endothelium release the nitric oxide (NO) necessary for relaxation of the penile muscles, resulting in the filling of the sinuses with blood and the restriction of venous outflow 5,6,10,11 . After intercourse, the penis returns to a flaccid state through the initiation of penile smooth muscle contraction by neural and local factors 12, 13 .
Therefore, the balance of contraction/relaxation of penile smooth muscle plays a key role in the process of erection. We thus guess the impairment of the smooth muscles may be a critical pathogenic event in ED genesis, however, the molecular targets in the process remained unclear.
Although the penile artery and corpus cavernosum smooth muscle (CCSM) are highly specific, with unique properties, these muscles share several features common to smooth muscles 9 . Smooth muscle contraction is mediated by multiple signals, including depolarization and G protein-coupled receptor (GPCR) agonists.
MLCP is composed of three subunits: Myosin Phosphatase Target Subunit 1 (MYPT1), a PP1c catalytic core, and a 20 Kd kinase with unknown function. MYPT1 regulates the activity of the MLCP holoenzyme primarily via physical interaction with the PP1c catalytic subunit 16,18,20 . Ablation of MYPT1 expression leads to a significant alteration in contractile behaviors 18,21 , including hypercontractility and hyporelexation, whereas mutation of MYPT1 phosphorylation sites only slightly affects contraction 22 . This pattern implies that the protein level of intact MYPT1 rather than the phosphorylation status of MYPT1 is the primary factor regulating contractile behaviors. Here, we found that MYPT1 expression was reduced in the CC smooth muscle of ED patients. The ablation of MYPT1 altered the contractile properties of penile smooth muscle, resulting in ED genesis. Mutant smooth muscle was more sensitive to GPCR agonists and more resistant to nitric oxide mediated relaxation. Pharmacological intervention with MYPT1 restored penile erectile function. Our results thus revealed an essential role of MYPT1 in ED genesis and a novel therapeutic strategy for ED.

Decreased MYPT1 protein expression in penile smooth muscles is associatied with ED.
To determine whether the level of MYPT1 protein is associatied with penile erection, we first assessed the penile function in a mouse line with smooth muscle-specific deletion of MYPT1. The pup production of Mypt1 +/+ , Mypt1 +/ΔSM and Mypt1 ΔSM /ΔSM mice strikingly differed. All of the matings of 8 female with 8 male C57BL/6 wild-type mice produced 288 pups in 52 litters, and all of the matings of 8 female wild-type C57BL/6 mice with 8 males of Mypt1 +/ΔSM mice produced 246 pups in 44 litters.
However, the mating of 6 female wild-type mice with Mypt1 ΔSM /ΔSM mice produced 10 pups in 2 litters (Table 1). These results show that male Mypt1 ΔSM /ΔSM mice presented subfertility, with small litter and pup numbers. To test whether subfertility was caused by an alteration in sexual desire, we evaluated sexual behaviors, including sniffing, mounting, grooming, and ejaculation within 45 min. The Mypt1 +/+ , Mypt1 +/ΔSM and Mypt1 ΔSM /ΔSM mice presented a comparable frequency of sexual behaviors (Table 2).
No vaginal plug was observed in wild-type females that mated with Mypt1 ΔSM /ΔSM mice.
We then detected the sperm quality and the level of sexual homoron. The sperm concentration and spern motility showed no difference among Mypt1 +/+ , Mypt1 +/ΔSM and Mypt1 ΔSM /ΔSM mice (Table 3), while the level testosterone and morphology of testis were also equal (Extended Figure 1, a-b). Thus, the subfertility of KO males was not likely caused by an alteration in sexual desire.
We then measured the penile ICP response to electrical stimulation to evaluate the penile erectile function in mice. Upon treatment with an electric stimulus (5 V, 12 Hz, 1 ms pulse width for 60 s), the ICP of Mypt1 +/+ mice increased rapidly and peaked at values of up to 55.49 ± 4.89 mmHg within 55 s; the ICP of Mypt1 +/ΔSM mice increased correspondingly, but the maximal pressure (42.30 ± 3.68 mmHg) was lower (p<0.05).
Strikingly, the ICP response to stimulation was abolished in Mypt1 ΔSM /ΔSM mice, and their maximal ICP value was significantly lower than that of the other groups (all p<0.001) (Figure 1, a-b).
We also collected corpus cavernosum (CC) bodies from 5 PED5i-invailed ED and 3 penile carcinoma patients. MYPT1 expression from ED patients was significantly decreased by western blot, 2 of the 5 ED biopsies were even showed no detectable MYPT1 expression (Extended Figure 2). Our collective observations showed that the ablation of MYPT1 led to ED.

Ablation of MYPT1 expression impairs the contractility of penile smooth muscle
Histological examination showed that all penises of Mypt1 +/+ , Mypt1 +/ΔSM and This observation suggests that MYPT1-deficient dorsal arteries showed a hypersensitive response to PE, particularly at a low dose. U46619 did not enhance the contractile response in mutant dorsal arteries, and the response sensitivity was unchanged ( Figure 2, e-f).
As CCSM is another important smooth muscle essential for penile erection, we then measured its contractile properties. Upon stimulation with 80 mM KCl, the peak force was not different among the Mypt1 +/+ , Mypt1 +/ΔSM , and Mypt1 ΔSM /ΔSM groups (Figure 2, g-h). When stimulated with 0.3 μM U46619, the force evoked on Mypt1 ΔSM /ΔSM muscle was significantly higher than that on Mypt1 +/ΔSM muscle (Figure 2, k-l), while the force produced by PE was comparable between these groups (p>0.05) (Figure 2, i-j). Thus, CC smooth muscle also displayed hyperresponsiveness to GPCR agonists.
RLC phosphorylation is tightly related to force development. To validate the hypothesis that the altered responses above were caused by natural force rather than by muscle remodeling, we measured RLC phosphorylation in the penile dorsal artery by urea/glycerol PAGE and Western blotting. As we expected, the induction of RLC monophosphorylation by PE was noticeably greater in mutant smooth muscle than in the control muscle ( Figure 3).

Reduced MYPT1 expression impairs nitric oxide-mediated relaxation of penile smooth muscle
As activation of nitric oxide signaling is fundamental for erection, we then assessed the

Pharmacological intervention with MYPT1 restores penile function in ED mice
Given that a reduced MYPT1 level is essential for the ED phenotype, the upregulation of MYPT1 could be expected to restore penile function. We screened natural compounds from herbs in an in vitro cell model and obtained active compounds with potent activity towards inducing MYPT1 expression. Among these compounds, lotusine (Lot1) had relatively strong activity. In vitro, upon treatment with 0.5 µM L0t1, MYPT1 expression increased up to 2-fold without apparent cytotoxicity. MYPT1 expression increased to about 3-fold while being treated with 0.5 µM Lot1 ( Figure 5, a-b). This activity was also validated in vivo by i.p. injection. After injection with Lot1 (5 mg/kg body weight), MYPT1 expression was elevated 1.3-fold in C57/BL6 mice and 2-fold in Mypt1 +/ΔSM mice ( Figure 5c) but did not increase in Mypt1 ΔSM /ΔSM mice.
As Mypt1 +/ΔSM mice showed a moderate ED phenotype, we used these mice as an ED disease model, although the penile function was not impaired enough to affect fertility. We treated these mice and their littermates with Lot1 via consecutive i.p.
injections. The Lot1-treated mutant penises showed sizes comparable to those of control penises ( Figure 6c). Surprisingly, the lumen diameter of the mutant dorsal artery and CC was also restored significantly ( Figure 6

MYPT1 protein expression can regulate penile erection in db/db mice.
Diabetes is a risk factor in ED genesis, and diabetic (db/db) mice is usually considered to be an ED mice model. We first detected the expression of eNOS and MYPT1 in the dorsa artery and CC by immunofluorescence staining, the results showed eNOS has slightly decreased in db/db mice (Extended Figure 6), while the MYPT1 protein was significantly reduced (Extended Figure 6). This result showed that in diabetic ED mice model, the decreased MYPT1 existed.
40-60% of patients failed with PED5-Is in certain disease states, especially in diabetic ED 27 . We treated db/db mice with 5 mg/kg Lot1 for 28 days and found that These results showed that Lot1 improved the erectile function and penile artery diameter of ED mice with a dose-dependent effect. These results showed that Lot1 can rescued the ED phenotype in diabetic mice.

Lotusine induces an increase in MYPT1 expression by inhibiting the ubiquitin pathway
To investigate the regulatory mechanism of Lot1, we compared Mypt1 mRNA levels in A7r5 cells treated with and without Lot1. Treatment with 0.5 μm Lot1 did not affect the Mypt1 mRNA level (p>0.05) (Figure 8a), suggesting that a posttranslational pathway regulates the MYPT1 protein level. We then used an anti-MYPT1 antibody to pull down MYPT1 protein from lysates of A7r5 cells treated with 20 μM MG132 and 0.5 μM Lot1 for 24 h and subjected the precipitated mixture to immunoblotting to detect ubiquitin.
The level of total MYPT1 was elevated after respective treatment with MG132 and Lot1, while the level of ubiquitinated MYPT1 treated with Lot1 was decreased ( Figure   8b). In addition, both the SIAH1 and SIAH2 E3 ligases, which contain an MYPT1 binding motif, were detected in the immunoprecipitated (Figure 8b). To test if Lot1 directly binds to SIAH1/2 E3 ligases, we expressed recombinant GST-SIAH1 and GST-SIAH2 proteins in bacteria and subjected these proteins to binding affinity assay with a Biacore system. The result showed that Lot1 could directly bind the recombinant proteins with high affinities (SIAH1: KD=2.4e-7 and SIAH2: KD=5.03e-7) (Figure 8, cf). Together results showed that lotusine upregulated MYPT1 protein through directly binding SIAH1 and SIAH2.

Discussion
In this report, we found that downregulation of MYPT1 significantly altered the contractile properties of penile smooth muscles, and was sufficient to cause serious ED phenotypes and histological changes. On the other side, the up-regulation of MYPT1 not only improved penile erection but also restored tissue structures. As MYPT1 expression has been proved to be efficiently regulated by pathological stimuli 28 , our result thus revealed a novel pathogenic mechanism for vasculogenic ED. Based on this result, the up-regulation of MYPT1 by a drug such as lotusine would reverse the affected smooth muscle to physiological status and hence cure erectile dysfunction.
Based on our observations, besides the imbalance in NO/cGMP/PKG pathway 29 , the ED genesis may also occur via the following scenario: pathological stimulation induces downregulation of MYPT1 in penile smooth muscle; the reduction in MYPT1 leads to enhanced contractile responses to agonists and resistance to nitric oxide; the resultant hypercontractility inhibits the blood filling process, and the penis adopts a persistent flaccid state, and resistance to nitric oxide inhibits the vessel dilation process.
As blood filling is required not only for erection but also for the penile nutrient supply, the arterial narrowing caused by the persistent flaccid state further limits blood filling and hence an erection. The interplay of these regulatory processes appears to contribute heavily to ED genesis. MYPT1 is encoded by the 4.6 kb Ppp1r12a gene and is ubiquitously and constantly expressed as a housekeeping gene 30 . The MYPT1 protein contains SIAH1/2 E3 ligase binding motifs, implying the involvement of a ubiquitin-E3 ligase-proteasomal degradation pathway in MYPT1 regulation 31 . Our recent report validated this pathway in colonic smooth muscle 32 . In this report, we demonstrated that this pathway also exists in penile smooth muscle. Interestingly, miRNA and several pathological factors (e.g. LPS and high glucose) can regulate MYPT1 expression 33-37 . Decreased MYPT1 expression was detected in gastric fundus and gastric antrum smooth muscle from diabetes patients 38 and ob/ob mice 39 , moreover the activity of SIAH1 was increased in high glucose environment 40 . This implies the involvement of MYPT1 in the ED phenotype complicated with other disorders such as diabetes 5-7 or bowel inflammation 41 .
Nitric oxide in penile tissue is released from nerves or synthesized by eNOS and mediates smooth muscle relaxation by stimulating guanylyl cyclase (GC) activity towards cGMP production, thereby mediating relaxation through cGMP-dependent kinase (PKGIα) 11,42 . PDE5 inhibition amplifies nitric oxide-cGMP signaling by overproducing cGMP, resulting in enhanced relaxation 11 . PDE5 inhibitors are currently used for first-line treatment of ED, and several new selective inhibitors are in development. However, according to current clinical results, at least 30% of ED patients do not respond to PDE5 inhibitors, in diabetic ED patients, the failed-treated patients were up to 40%-60% 5 . Meanwhile, the eNOS in the diabetic animal model did not significantly decrease in CC 43,44 . We here found that the ablation of MYPT1 expression led to reduced sensitivity to nitric oxide for penile smooth muscle relaxation. This effect may result from the abolition of PKGI-MYPT1-Ser695 signaling during relaxation 18,21 and, interestingly, might partially explain the therapeutic failure of PDE5 inhibitors.

Male sexual behavior test
Sexual behavior was tested as described in a previous report 23 . A male (8-10 weeks) and a female (6-8 weeks) mouse were manually placed in the same chamber, and their sexual behaviors were visually assessed at 8:00 pm over a 45-min measurement period.
Before behavior recording, the female mice were hormonally primed for 48 hours via subcutaneous injection of 10 μg of estradiol benzoate (Sigma-Aldrich) and were then injected with 500 μg of progesterone (Sigma-Aldrich) at 3-5 hours before recording.
Both reagents were dissolved in sesame oil. The plug in the vaginal orifice was observed at 9:00 am on the second day.

In vivo measurements of intracavernous pressure (ICP) and mean arterial pressure (MAP)
To evaluate the penile erectile function of mice, the ICP in response to electrostimulation of cavernous nerves was measured according to a previous report 24 .
Mice were anesthetized by intraperitoneal injection of pentobarbital (40 mg/kg), and the bladder and prostate were exposed via a midline suprapubic incision. The testes and epididymides were repositioned into the abdomen to expose cavernous nerves. The penis was denuded of skin and fascia, and a heparinized (100 IE/mL) 25-gauge needle was inserted into the CC and connected to a pressure transducer via polyethylene (PE)-50 tubing. The catheter was connected to a pressure transducer connected to a PowerLab 8/SP data acquisition system (Chart 5.0 software; ADInstruments, Colorado Springs, CO). The parameters for electric stimulation of cavernous nerves were as follows: 5 V, 12 Hz, 1 ms pulse width, and 60 s duration. MAP was measured as previously reported (ALC Non-invasive Blood Pressure System, Shanghai Alcott Biotech, China) 21 .

Immunofluorescence
Penises were fixed with 4% formaldehyde supplemented with 0.002% picric acid for 4 hours at room temperature and were then dehydrated in 30% sucrose PBS overnight at 4 º C. Fixed tissues were embedded in optimal cutting temperature (OCT) compound

Histological examination of mouse penises
Mouse penises were fixed in 4% formaldehyde at 4°C for 2 hours, dehydrated in butyl alcohol at 4°C overnight, embedded in paraffin, and cut into 5-μm sections. Sections were stained with hematoxylin/eosin (H-E), the morphology of penile arteries was examined via microscopic imaging (DotSlide, Olympus).
The protein concentrations were measured with bicinchoninic acid protein assay reagent (Thermo Scientific Pierce). Equal amounts of protein were separated by 8-12% SDS-PAGE, and the proteins were then transferred to a PVDF membrane. The membrane was sequentially probed with a primary antibody and a secondary antibody.
To visualize the immunoreaction signal, the membrane was incubated in Subpico Western Solution (Sudgen) and exposed to film. The following primary antibodies were

Measurement of myosin light chain phosphorylation 26
Myosin light chain phosphorylation in the penile artery was measured via urea/glycerol PAGE. Briefly, after stimulation with 10 μM PE, the artery was immediately frozen in liquid nitrogen at the indicated time point and stored in a solution of 10% TCA/10 mM DTT in acetone at -80 º C. The tissues were then ground to a slurry with 10% TCA/10 mM DTT in H2O. The mixtures were centrifuged at 3,000 × g for 3 min, and then the supernatants were discarded. The precipitates were washed 2 times with acetone and one time with diethyl ether, dried at room temperature after removal of the diethyl ether and dissolved in 8 M urea sample buffer. An anti-RLC antibody was used as the primary antibody for Western blotting. The relative ratio of phosphorylated RLC to total RLC was calculated.

Preparation of recombinant SIAH1 and SIAH2 proteins
Mouse Siah1 and Siah2 cDNAs were amplified from smooth muscle tissue by RT-PCR kit (Invitrogen) and confirmed by DNA sequencing. The coding fragments were respectively inserted into a pGEX-5x-1 vector via BamH I/EcoRI polyclonal restriction sites, and then transformed the recombinant plasmids to BL21 E.Coli. After induction with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG), the bacteria successfully expressed recombinant GST-SIAH1 and GST-SIAH2 proteins. We purified these proteins with a glutathione sepharose resin (Cat. 17-0756-01, GE Healthcare) and the purity of these proteins was higher than 95%.

Binging assay for lotusine and recombinant SIAH1 and SIAH2 proteins
The respective binding affinities of lotusine with SIAH1 and SIAH2 were determined by a Biacore T200 instrument. The SIAH1-GST and SIAH2-GST protein (with the concentration of 20 μg/ml) were respectively immobilized on a CM5 chip at pH 4.0 by using the amine coupling kit (GE Healthcare). A dilution series of lotusine (0 to 5 μM) was passed on SIAH1-GST and SIAH2-GST at 20 μl/min and the association phases (120 seconds) were recorded. The binding kinetics were analyzed with a 1:1 binding model of Biocore evaluation software. The kinetic rate constants, Ka, and Kd were derived for each reaction. KD values were calculated by Kd/Ka ratio.

Statistical analysis
Data are presented as means ± SEM. Differences between two groups were evaluated by paired or unpaired t-tests. Multiple group comparisons were performed by using one-way ANOVA followed by Tukey's test. P ≤ 0.05 was considered statistically significant. All statistical analyses were performed in GraphPad software.