Effect of chronic lithium administration on endothelium-dependent relaxation of rat corpus cavernosum: the role of nitric oxide and cyclooxygenase pathways

Authors


Ahmad R. Dehpour, Department of Pharmacology, School of Medicine, Tehran University of Medical Sciences, Tehran, PO Box 13145–784, Iran.e-mail: dehpour@medscape.com

Abstract

OBJECTIVE

To verify the effect of chronic lithium administration on the endothelium-dependent relaxation of rat corpus cavernosum, as lithium is a major drug for treating bipolar disorder and some studies showed that lithium might cause erectile dysfunction in such patients, by a mechanism as yet unknown.

MATERIALS AND METHODS

LiCl (600 mg/L) was dissolved in drinking water and Sprague–Dawley rats received the solution for 30 days; control rats received tap water. After 30 days corporeal strips were prepared from both groups, mounted under tension in oxygenated organ baths, and pre-contracted with phenylephrine (7.5 µm). After equilibration, the strips were relaxed by acetylcholine (10 nm to 1 mm) in the presence or absence of indomethacin (a cyclooxygenase inhibitor; 20 µm). Furthermore, the relaxant responses to sodium nitroprusside (1 nm to 1 mm), a nitric oxide (NO) donor, were investigated in both groups. NADPH-diaphorase histochemistry was used to identify NO synthase within cavernosal tissue strips of both groups.

RESULTS

The acetylcholine-dependent relaxation was significantly lower in lithium-treated rats than in controls. Although indomethacin decreased significantly the relaxant responses to acetylcholine in controls, it increased the relaxant responses in lithium-treated rats. NADPH-diaphorase staining was greater in the chronic lithium-treated than in control preparations. Sodium nitroprusside produced similar relaxation in both groups.

CONCLUSION

Chronic lithium administration can impair the endothelium-dependent relaxation of rat corpus cavernosum; NO availability might decrease after lithium administration and the cyclooxygenase pathways might have a role in this effect.

Abbreviations
ED

erectile dysfunction

SNP

sodium nitroprusside

NO(S)

nitric oxide (synthase)

L-NAME

Nω-nitro-l-arginine methyl ester

EC50

effective concentration for a 50% response

PI

phosphoinositol.

INTRODUCTION

Lithium has largely met its initial promise as the first drug to be discovered in the modern era of psychopharmacology [1]. Although over 50 years have elapsed since its effects on mania were first described [2], lithium is still a mainstay in the treatment of mood disorders [3–6]. However, as with many other drugs, lithium has unwanted side-effects. Some studies showed that lithium can result in erectile dysfunction (ED) in patients who take it [7–10]. Aizenberg et al.[10] showed that 31.4% of bipolar or schizo-affective patients receiving lithium as the sole medical treatment reported sexual dysfunction. In another study, Ghadirian et al.[9] reported that a combination of lithium and other psychotherapeutic drugs, especially benzodiazepines, increased the frequency of sexual dysfunction to up to a half. Sexual dysfunction is also one of the most common side-effects that clinicians encounter in treating bipolar patients when using a combination of mood stabilizers (lithium, divalproex, or carbamazepine) and antipsychotics [11]. Moreover, it seems that more attention should be given to lithium-induced sexual dysfunction, as its presence can have important consequences for clinical management and compliance [9–11].

Many cases of ED might be due to an impairment of cavernosal smooth muscle physiology and functioning [12]. Relaxation of cavernosal smooth muscle is critical for inducing and maintaining penile erection. Nitric oxide (NO) catalysed by NO synthase (NOS) is now considered to be both the principal neurotransmitter and endothelium-derived relaxing factor controlling the relaxation of cavernosal smooth muscle [13–16]. It is postulated that endothelium-generated NO appears to have a major role in maintaining an erection [17]. NO acts by stimulating guanylyl cyclase to produce cGMP, which consequently causes cavernosal smooth muscle to relax [13]. Two products of the cyclooxygenase pathway, prostaglandins E1 and E2, are also implicated in the control of penile erection [18]. Prostaglandins interact with NO in many ways. Some studies showed that the basal release of NO and prostaglandins has a protective role in many physiopathological conditions such as ED [19–22].

Although previous studies reported ED in a considerable percentage of patients treated with lithium, little if anything is known about the mechanism of this sexual dysfunction. As the impairment of the mechanisms that cause relaxation of corpus cavernosum smooth muscle can lead to ED [12], in the present study we investigated the effect of chronic lithium treatment on the endothelium-mediated relaxation of rat isolated corpus cavernosal smooth muscle. To assess changes in the NO system in the corpus cavernosum of lithium-treated rats we further identified NOS by NADPH-diaphorase staining, and assessed the response of the smooth muscle to sodium nitroprusside (SNP), a NO donor.

MATERIAL AND METHODS

Male Sprague–Dawley rats (Pasteur Institute; 200–250 g) were used throughout the study; they were housed in a light-controlled room with a 12-h day/night cycle, and were given free access to food and water. Experiments were conducted in accordance with the recommendations of the ethics committee of Tehran University. The rats were divided into two main experimental groups of control and chronic lithium-treated rats. The latter received 600 mg/L LiCl in tap water for 30 consecutive days; control rats received tap water with no lithium supplement. At the end of the treatment period, serum lithium levels were determined by an atomic absorption spectrophotometer (Perkin-Elmer, Norwalk, CT, USA).

The rats were then killed by cervical dislocation, their penises surgically removed at the level of the crural attachments to the pubo-ischial bones, and promptly placed in a Petri dish containing Krebs-bicarbonate solution (containing in mM: NaCl, 118.1; KCl, 4.7; KH2PO4, 1.0; MgSO4, 1.0; NaHCO3, 25.0; CaCL2, 2.5; and glucose, 11.1), bubbled with a mixture of 95% O2 and 5% CO2. The glans penis and urethra were excised and the corpus cavernosum tissue was then dissected free from the tunica albuginea. The two corpus cavernosa were separated by cutting the fibrous septum between them. They were mounted separately in 20-mL organ chambers with one end tied to an electrode holder and the other to a wire connected to a force transducer (Narco F-60, Narco Biosystems, Houston, TX, USA). The chambers contained Krebs-bicarbonate solution (PH 7.4) at 37 °C equilibrated with 95% O2 and 5% CO2. The strips were allowed to equilibrate under optimum resting tension of 0.2–2 g, and after equilibration for 60 min the contractile response to phenylephrine (7.5 µm) was measured. The optimum resting tension for corpus strips in this manner was 1.5 g and this tension was applied in all subsequent experiments. The presence of a functional endothelium was tested by determining the relaxation response to 1 µm acetylcholine in strips pre-contracted with 7.5 µm phenylephrine. In all the experiments, each strip was used only once.

The following drugs were used: phenylephrine hydrochloride, SNP, acetylcholine chloride, LiCl, Nω-nitro-l-arginine methyl ester (L-NAME) and indomethacin (Sigma, St. Louis, MO, USA). All drugs were freshly dissolved in distilled water, except indomethacin, which was dissolved in absolute ethanol.

In the control and lithium-treated groups, concentration-response curves for phenylephrine (10 nm to 1 mm) were obtained by the cumulative addition of phenylephrine to the chamber in half-log increments. The effective concentration for a 50% response (EC50) of phenylephrine in the two experimental groups was compared.

In the control and lithium-treated groups, when the contractile response to phenylephrine (7.5 µm) in corporeal tissue strips reached a steady state, concentration-response curves for relaxation were constructed by the cumulative addition of acetylcholine every 2 min for 20 min at half-log increments in concentrations of 10 nm to 1 mm.

In the next experiment, cumulative doses of acetylcholine were added to the organ bath containing tissues from control and lithium-treated rats: (a) after a 30-min exposure L-NAME (1 µm); and (b) after a 20-min incubation with indomethacin (20 µm).

For the histochemical study, the corpus cavernosum was immediately removed and fixed for 24 h in ice-cold 0.1 m PBS, PH 7.4, containing 0.1% glutaraldehyde and 4% paraformaldehyde, followed by cryoprotection in 15% sucrose for 4 h and 30% sucrose overnight at 4 °C. The fixed blocks were cut into 10 µm sections in a cryostat and mounted onto gelatine-chrome-alum-coated glass slides.

To identify NOS, NADPH-diaphorase staining in corpus cavernosum was examined histochemically using a modified method of Basar et al.[23], by incubating free-floating or slide-mounted tissue sections with 1 mm NADPH, 2 mm nitroblue tetrazolium, 0.1 m PBS, and 0.3% Triton X-100 at room temperature in the dark overnight. The reaction was stopped by rinsing the sections in PBS. Histochemical control experiments, in which NADPH was excluded from the reaction mixture, gave no positive staining. The tissue sections were covered with a polymeric mountant, examined using a light microscope with a filter, and photographed.

NADPH-diaphorase-positive nerves and endothelial cells are easily detected using this stain, and seen as a highly localized, densely blue region. The NADPH-positive nerves and endothelial cell staining intensity were evaluated in × 400 fields of each corpus cavernosum.

In the control and lithium-treated rats, when the contraction stabilized, concentration-response curves for SNP (1 nm to 1 mm) were obtained by the cumulative addition of SNP to the chamber in half-log increments. The EC50 of SNP in two groups was compared.

The data are expressed as the mean (sem), and assessed using one-way anova followed by Newman–Keuls post hoc test, with statistical significance considered at P < 0.05.

RESULTS

There was no significant difference in the weight increase of control and chronic lithium-treated rats; the serum level of lithium was 0.33 (0.05) mmol/L in chronic lithium-treated rats and it was undetectable in control rats.

In both groups phenylephrine caused concentration-dependent contraction in strips of corpus cavernosum (Fig. 1). There was no significant difference in the maximum contractile responses to phenylephrine between the control and treated groups, or in the contractile responses to 7.5 µm phenylephrine, at 75.81 (7.08) and 75.05 (9.01)% of maximum contraction, respectively. The EC50 was not significantly different between the groups, at 3.08 (0.28) and 2.96 (0.23) µm, respectively.

Figure 1.


Concentration-dependent contraction in response to phenylephrine in isolated cavernosal smooth muscles from control and chronic lithium-treated groups. Each group consisted of six rats.

There was a significantly lower (P < 0.001) maximum response to acetylcholine in the lithium-treated than in the control group, at 42.08 (2.20) and 112.45 (2.86)%, respectively (Fig. 2). In the lithium-treated group there was a significant difference (P < 0.01) with 1 µm acetylcholine.

Figure 2.


Concentration-dependent relaxation in response to acetylcholine in pre-contracted (phenylephrine 7.5 µm) rat corpus cavernosum in control and chronic lithium-treated groups. Each group consisted of six rats. (**P < 0.01; ***P < 0.001 vs control group).

Acute inhibition of NOS with 1 µm L-NAME significantly inhibited the acetylcholine-induced relaxation in the corporeal tissue of both control and lithium-treated rats (Fig. 3).

Figure 3.


Effect of 1 µm L-NAME on acetylcholine-induced relaxation of rat corpus cavernosum, showing the control and lithium-treated groups with and with no L-NAME. Each group consisted of six rats.

Pre-incubation with indomethacin (20 µm) attenuated (P < 0.001) the maximum response to acetylcholine in control rats, at 112.4 (2.8) vs 72.3 (6.5)%, in the absence or presence of indomethacin (Fig. 3). The notable result was that in vitro indomethacin significantly (P < 0.001) potentiated the maximum response to acetylcholine in lithium-treated rats, at 42.1 (2.2) vs 99.7 (4.9)% in the absence or presence of indomethacin, respectively (Fig. 4). There was no significant difference between th maximum responses to acetylcholine in the control group with no in vitro indomethacin and the lithium-treated group with in vitro indomethacin, at 112.4 (2.8) and 99.7 (4.9)%, respectively.

Figure 4.


The effect of 20 µm indomethacin on acetylcholine-induced relaxation of rat corpus cavernosum, showing control and lithium-treated groups with and with no indomethacin. Each group consisted of six rats.

There was no staining in control sections from which NADPH was omitted. There were NADPH-diaphorase-stained nerves and endothelial cells in corpus cavernosum of both control and lithium-treated rats, but there was obviously more NADPH-diaphorase staining in lithium-treated rats (Fig. 5).

Figure 5.

Histochemical localization of NOS with NADPH-diaphorase staining in the cavernosal tissue from control (A) and lithium-treated (B and C) rats. Each group consisted of six rats.

SNP produced concentration-dependent relaxation of corpus cavernosum pre-contracted by 7.5 µm phenylephrine (Fig. 6). Neither the maximum relaxation nor the EC50 of SNP-induced relaxation was different in the control or lithium-treated rats.

Figure 6.


Concentration-dependent relaxation in response to SNP in pre-contracted cavernosal muscles from control and chronic lithium-treated groups. Each group consisted of six rats.

DISCUSSION

In the present study the endothelium-mediated relaxation of corpus cavernosum was impaired in chronic lithium-treated rats. Relaxation of the smooth muscle was attenuated in the presence of L-NAME, a nonselective NOS inhibitor, showing this relaxation to be mainly mediated via NO. As contraction-response curves for phenylephrine were indistinguishable between control and chronic lithium-treated rats, and NO-mediated relaxation was decreased in the lithium-treated group, it seems that the nitrergic-mediated relaxation was decreased in the corpus cavernosum of chronic lithium-treated rats.

Many studies reported that chronic lithium treatment can result in sexual dysfunction in about a third of patients receiving it [7–10] and even combined therapy with other psychotropic drugs could increase the frequency of sexual dysfunction up to a half [9,11]. The most common problems in this group of patients were a reduction in the frequency of sexual thoughts, difficulties in achieving and maintaining erections, and loss of erection during sex [10]. However, the mechanism of this problem is as yet unknown. The results of the present study suggest that one of the mechanisms that leads to sexual dysfunction in chronic lithium-treated patients might be the local effect of the drug on the corpus cavernosum and its adverse effect on the endothelium-mediated relaxation of corporeal tissue.

A widespread hypothesis explaining the effects of lithium on neurotransmitter responses is that lithium inhibits uncompetitively inositol monophosphatase [24–29], an effect suggested by diminished signalling through phosphoinositol (PI)-linked second messengers and, in turn, Ca2+. Messenger molecules, such as acetylcholine, bind to the G-protein-coupled receptor on an endothelial cell and share the ability to induce large and transient increases followed by small and sustained increases in Ca2+ through activation of phospholipase C, which liberate inositol 1,4,5-triphosphate (IP3) and 1,2-diacylglycerol (DAG) [30–34]. Therefore, the diminished relaxation of corpus cavernosum in the present study can be explained by the probable effect of lithium on the PI cycle, as the activity of PI cycle in the corporeal endothelium is a crucial step for generating NO. Also, the results of the present study suggest that this effect of lithium occurs at concentrations less than the therapeutic levels, which are usually 0.8–1.0 mm[1,35]. This is in line with previous studies, which showed that the inhibitory effect of lithium on the PI metabolic cascade at the monophosphatase level is evident in intact cells at extracellular concentrations of > 0.1 mm[28].

As all known NOS isoforms have NADPH-diaphorase activity [36–38], as confirmed by the co-location of NOS with NADPH-diaphorase activity [39–42], the NADPH-diaphorase histochemical reaction is commonly used to visualize NOS protein. In the present study we used NADPH-diaphorase histochemistry to identify NOS in rat corpus cavernosum; NADPH-diaphorase histochemical staining of the corpus cavernosum strips in control and chronic lithium-treated rats showed that the NOS of endothelium and nerves was increased in lithium-treated rats. This result is in line with some studies indicating that chronic treatment with lithium increased the induction of NOS gene expression in the hypothalamus and astrocytes [43,44]. However, other studies reported different data, which showed that lithium had an inhibitory effect on the guanylyl cyclase pathway as it inhibited cGMP formation in rat and mouse brain [45]. In the present study the relaxant responses to SNP, which activates cGMP synthesis directly, was unaffected by chronic lithium administration. Thus, it is possible that the smooth muscle responsiveness to NO is not affected by chronic lithium treatment.

Endothelium-dependent relaxation in response to acetylcholine results in the production of prostaglandins (e.g. E1 and E2) and NO [18,46,47]. Prostaglandin E1 promotes relaxation of penile arterial smooth muscle via prostacyclin receptors, and of trabecular smooth muscle through interaction with prostaglandin E receptors, with some preliminary data supporting a key role for the EP2 receptor subtypes [20,48]. Prostacyclin and prostaglandin E receptors, coupled to G proteins, increase intracellular cAMP levels through adenylyl cyclase stimulation, and consequently cause relaxation of cavernosal tissue [20]. Therefore, the inhibition of the cyclooxygenase pathway by indomethacin would be expected to decrease acetylcholine-induced relaxation of corpus cavernosum. In the present study, indomethacin attenuated the endothelium-mediated relaxation of corpus cavernosum from control rats, suggesting that the cyclooxygenase pathway might be important in the endothelial relaxation of corporal strips. However, the effect of cyclooxygenase inhibition on acetylcholine-mediated relaxant responses of the corpus cavernosum is variable. Some authors reported no effect of indomethacin on the endothelium-dependent relaxation induced by acetylcholine in rabbit corpus cavernosum in vitro[49], while others reported an enhancement of the response in both human and rabbit corporeal tissues [50]. In agreement with our previous study [51], in this experiment with penile erectile tissue from rats, in vitro administration of indomethacin impaired the relaxant response to acetylcholine. However, it is not yet clear whether these disparities are due to differences between species or in the methods used. An interesting finding of the present study was that although in vitro administration of indomethacin decreased the endothelium-dependent relaxation of corporeal strips from control rats, it prevented the attenuation of relaxant responses to acetylcholine in lithium-treated rats. This result suggests a role for the cyclooxygenase pathway in the inhibitory effect of chronic lithium administration on endothelium-dependent relaxation of rat corpus cavernosum, such that the blockade by indomethacin of this pathway might lead to an improvement of the impaired endothelium-mediated relaxation of corpus cavernosum from chronic lithium-treated rats. However, more detailed studies are needed to verify the exact mechanism of action of lithium on cyclooxygenase pathway in this tissue.

In conclusion, the present study indicated that chronic lithium treatment impaired the endothelium-mediated relaxation of rat corpus cavernosum. NADPH-diaphorase staining intensity was greater in the lithium-treated group than the control, and the relaxant response to SNP was unaffected by chronic lithium treatment. Inhibition by indomethacin of the cyclooxygenase pathway improved the impaired endothelium-dependent relaxations in lithium-treated rats. Therefore, it could be concluded that chronic lithium treatment might alter both the NO and cyclooxygenase pathways, and lead to impairment of endothelial relaxation of rat corporeal tissue.

CONFLICT OF INTEREST

None declared.

Ancillary