T. terrestris was purchased at a local market in Korea. An amount of 100 g of the T. terrestris fruit was ground and extracted with 1 L of 90% ethanol at 80℃ for 2 hours and then filtered through Whatman No. 2 filter paper. The filtrates were evaporated in a rotary evaporator (Eyela-NE, Rikakikai Co., Tokyo, Japan) under reduced pressure. The yield of T. terrestris fruit extract was 25% (wt/wt). The aqueous extracts were filtered and lyophilized (Thermo Fisher Scientific Inc., Rockford, IL, USA) to yield 55% (wt/wt) of the crude extract. These extracts were preserved in a refrigerator and used in the studies described below.

New Zealand white male rabbits weighing 2.5 to 3.0 kg were used (n=8/group). All the animals were cared for in accordance with the National Research Council publication Guide for the Care and Use of Laboratory Animals. The animal experiments were approved by the Institutional Animal Care and Use Committee of the research institute at the university. The rabbits were sacrificed with an overdose of ketamine hydrochloride (50 mg/kg) injected into the marginal vein of the ear, the penis was dissected immediately, and sections of the CC (2 mm×2 mm×10 mm) were prepared. The sections were then transferred to an organ bath; one end was connected to a muscle fixation ring and the other end was connected to an isometric tension transducer (SG-10). The sections were connected to a force displacement transducer (TSD 125C, Biopac Inc., Goleta, CA, USA), which is an isometric tension transformer, and the signals were recorded on a personal computer by using a four-channel data acquisition and analysis system for Windows (MP36R 4-Channel Systems, Biopac Inc.). The signals were relayed to a physiography instrument (Power-Lab, ADI Instruments, Sydney, Australia) and measured. Chart 5 software (ADI Instruments) was used for real-time monitoring of tension. Krebs-Henseleit (KH) solution was used for the organ bath at 37℃ and pH 7.4 and gassed continuously with 95% O₂/5% CO₂. With the initial tension of each section maintained at approximately 2 g, the KH solution was changed approximately every 30 minutes (total equilibration period, 2 hours) and allowed to reach a stable condition. Once a stable condition was reached, phenylephrine (PE) was added and the contraction level was observed. Each section was progressively stretched to the optimal point on its length-tension curve as determined by the active tension developed in response to PE. A stable condition was restored by washing the preparation with KH solution three times. These steps were repeated, and contracture within 100%±10% of the previous contraction was defined as the ideal resting optimal isometric tension. While maintaining resting optimal isometric tension, the tissues were contracted by PE (5×10^-6 M) pretreatment within the bath. A T. terrestris extract in five-fold dilutions (0.25 to 4 mg/mL) was added, and the relaxation level was assessed by observing the change in the tension curve.

The relaxation reactions were investigated in endothelium-intact or -denuded tissue groups to examine the mechanism by which the T. terrestris extract induced CC relaxation. Endothelium-intact or -denuded tissues from the CC were incubated for 30 minutes in KH solution containing various drugs. Next, the tissues were contracted with PE (5×10^-6 M), during which plateau state contraction was attained. The relaxation reaction was then observed following the addition of increasing concentrations of T. terrestris extract. The drug probes included N(G)-nitro-L-arginine methyl ester (L-NAME; 10^-5 M, nitric oxide [NO] synthesis inhibitor), propronolol (10^-5 M, β-receptor blocker), indomethacin (3×10^-5 M, cyclooxygenase inhibitor), glibenclamide (10^-5 M, K⁺_ATP channel inhibitor), 4-aminopyridine (10^-5 M, membrane potential-dependent K⁺ channel inhibitor), and methylene blue (10^-5 M, guanyl cyclase inhibitor) [5-7]. Eight sections were used for each group. The CC sections were pretreated with 3 mL of 0.3% 3-[(3-cholamide propyl)]-1-propane sulfonate (CHAPS) for 20 seconds to remove all vascular endothelial cells. The sections were rubbed lightly by using the thumb and index finger for 20 seconds and washed with KH solution, and the cells were removed by gently rolling the sections on dry paper. Removal of the vascular endothelial cells was confirmed by the failure of the loss of tissue contraction induced by acetylcholine (10^-5 M) in response to PE (5×10^-6 M).

The CC relaxation response shown in the organ bath was studied to determine whether it was expressed in vivo. Forty Sprague-Dawley male rats weighing 280 to 320 g were allocated to six groups according to the oral dosage received: control, 2.5, 5, 10, 50, and 100 mg/kg (n=8 each). Next, the groups were orally administered the same T. terrestris extract every day for 1 month before measurement of the ICP.

The rats were anesthetized with ketamine (50 mg/kg) injected intraperitoneally. The animals were placed in the supine position on an operating table, a catheter (polyethylene-50 tube) was inserted into the carotid artery, and blood pressure (BP) was monitored. The prostate was exposed by a midline incision of the abdomen. Additionally, the pelvic ganglion, located on the posterolateral side of the prostate, was assessed, and the pertinent pelvic nerves and cavernosal nerves were assessed and dissected without injury. The cavernosal nerves were stimulated by using an electric stimulator and platinum electrodes. To assess an erection in response to electrical stimulation of the nerves, the penis was dissected up to the penile crura. The tunica albuginea of the penis was assessed by dissecting the ischial corpus cavernosal muscles surrounding the penile crura, and a 25-gauge needle was installed (polyethylene-50 tube pretreated with 250 U/mL heparin) in the CC [8].

BP and ICP were measured by using a BP manometer transducer and recorder (PowerLab, ADI Instruments). Chart 5 software (ADI Instruments) was used for real-time BP monitoring. Electrical stimulation was performed at a voltage of 5 V with a 60-second duration.

Pressure was measured before, during, and after stimulation to examine the hemodynamic reaction in response to autonomic nerve stimulation. Additionally, to examine the reactions induced by nerve stimulation, the percentage of maximum ICP/mean systemic arterial blood pressure (max ICP/MAP×100) was measured by stimulating the CC nerves, and differences among the groups were compared.

Contraction of the smooth muscle sections was equalized for 1 hour in KH solution and then induced with PE (5×10^-6 M), after which the tissues were exposed to the T. terrestris extract for 10 minutes. The tissues were frozen rapidly in liquid nitrogen and homogenized in ice-cold 6% trichloroacetic acid. The homogenized tissues were centrifuged for 15 minutes, and cellular proteins in the supernatant were separated by using conventional methods. The samples were divided into a normal CC group and experimental groups (0.25, 0.5, 1, 2, and 4 mg/mL T. terrestris extract), and the concentrations of released cGMP (n=8/group) and cAMP (n=8/group) were measured by immunoassay. The immunoassay was based on the competitive binding technique, in which the cGMP and cAMP in a sample compete with a fixed amount of horseradish peroxidase-labeled cGMP and cAMP for sites on rabbit polyclonal antibodies. During incubation, the polyclonal antibodies become bound to a microplate coated with goat antirabbit antibodies. Following a wash to remove excess conjugate and unbound sample, a substrate solution was added to determine the bound enzyme activity. Color development was stopped, and the absorbance was read at 450 nm. The color intensity was inversely proportional to the concentrations of cGMP and cAMP in the sample.

PE, L-NAME, glibenclamide, 4-aminopyridine, indomethacin, methylene blue, propranolol, and CHAPS were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA). The composition of the KH solution (pH 7.4) was (mM/L) as follows: NaCl, 118.1; NaHCO₃, 25; KCl, 4.6; KH₂PO₄, 1.2; CaCl₂, 2.5; MgSO₄, 1.21; and glucose, 11.0.

The statistical significance was analyzed by a t-test and one-way analysis of variance followed by Scheffe's F test by using the SPSS ver. 14.0 (SPSS Inc., Chicago, IL, USA). The results are expressed as the mean±standard deviation. A p-value <0.05 indicated statistical significance.

The T. terrestris extract produced a concentration-dependent relaxation response at 0.25 to 4 mg/mL (Fig. 1). Glibenclamide (10^-5 M, K_ATP blocker), 4-aminopyridine (10^-5 M, K⁺ channel blocker), methylene blue (10^-5 M, guanylate cyclase inhibitor), indomethacin (3×10^-5 M, prostaglandin pathway cyclooxygenase inhibitor), and L-NAME (10^-5 M, NO/NO synthase [NOS] inhibitor) were each administered respectively before extract treatment to investigate the mechanism of action of the relaxation response. The group of endothelium-denuded tissues from the CC was also investigated. Relaxation of the CC was inhibited in the endothelium-removed and L-NAME pretreatment groups (p<0.05) (Fig. 2).

The T. terrestris extract was orally administered every day at 2.5, 5, 10, 50, or 100 mg/kg for 1 month. Next, the ICP/MAP (%) was measured during neural stimulation of the CC. After 1 month of oral administration, a significant concentration-dependent increase in ICP/MAP was observed in accordance with the increasing dosage of the T. terrestris extract (p<0.05). The increase in ICP was observed at a dose of 2.5 mg/kg·d T. terrestris extract, and the maximum effect occurred at 100 mg/kg·d compared with that in the control group (p<0.05) (Fig. 3).

The cAMP concentration in the CC increased significantly after administration of the T. terrestris extract. However, the maximum concentration was observed at 0.5 mg; at 1, 2, and 4 mg, the values were lower than those at 0.5 mg. The cAMP concentrations at 0.5 and 4 mg were lower than those at 1 and 2 mg, revealing no concentration-dependent pattern (Fig. 4A). The cGMP concentration in the CC after administration of the T. terrestris extract was not statistically significant (Fig. 4B).