Comparing the relative amount of testosterone required to restore sexual arousal, motivation, and performance in male rats

Abstract

It is well established that male rat reproductive behaviors including sexual arousal, motivation, and performance are dependent on circulating levels of testosterone (T). The present study was designed to (1) compare the relative amount of T required to restore these different aspects of behavior in castrated rats, and (2) create an animal model for clinical populations with sexual impairments. Twenty-nine male Long–Evans rats were tested before and after castration for sexual performance (copulation), motivation (partner preference), and arousal (50 kHz ultrasonic vocalizations; measured together with scent marking). Sexual arousal was also inferred from copulation data. Rats were then assigned to one of four groups, and T was re-introduced via Silastic capsule implants varying in length and content: No T (empty capsules), Low T (2 mm capsules), Medium T (5 mm capsules), or High T (two 10 mm capsules). The highest dose was intended to restore physiological levels. Results indicate that High T is required for 50 kHz vocalizations, while Medium T was sufficient for the restoration of copulation, partner preference, and scent marking. These data suggest that sexual arousal may be most sensitive to reductions in testosterone. The role of T levels in measures of generalized and specific (sexual) arousal is discussed in the context of other reproductive behaviors. Furthermore, because the Low T group showed impairments across all behaviors during post-implant tests, we propose that these animals may provide a good animal model for studying clinical conditions marked by reduced motivation and arousal, including Hypoactive Sexual Desire Disorder.

Introduction

Many decades ago, Frank Beach proposed that reproductive behaviors in male mammals include two separate, but interacting components: appetitive measures (often referred to as sexual motivation) and consummatory measures (often referred to as sexual performance) (Beach, 1956; reviewed by Meisel and Sachs, 1994, Hull et al., 2002). For males appetitive behavior is defined as the motivation to approach a sexually receptive female partner, usually expending energy to attain this goal (Sachs, 2007), while consummatory behaviors refer to the physical components of mating (Meisel and Sachs, 1994, Roselli et al., 2003, Nelson, 2005). Beach's original work also referenced a sexual arousal mechanism that was primarily responsible for bringing males to a threshold for initiating copulation. However, in recent years the term “sexual arousal” has evolved as an entity separate from (although less clearly defined than, and still linked very closely to) sexual motivation and performance (Pfaus et al., 2003, Sachs, 2007). In trying to identify how hormones and neurons give rise to these different aspects of reproductive behaviors, a variety of assays have been developed.

In male rats, sexual performance is typically measured with standard copulation tests, where subjects are given unrestricted access to sexually receptive females. Common measures of performance include mounts, intromissions and ejaculations (Meisel and Sachs, 1994, Harding and McGinnis, 2003, Harding and McGinnis, 2004, Harding and McGinnis, 2005). Standard copulation tests have also been used to measure sexual arousal and sexual motivation, specifically by examining latencies to initiate copulation (Saito and Moltz, 1986, Mendelson and Pfaus, 1989, Pfaus et al., 1990). However, because this test is so closely tied to performance, it has profound limitations in the measure of sexual motivation (Everitt, 1990) and although some argue it is a good measure of peripheral sexual arousal (Sachs, 2007), this seems limited in that it neglects “central” arousal in the brain (Bancoft, 2005). For these reasons, assays that differentiate between sexual performance and other components may be preferred.

The most common test used to assess sexual motivation in rodents is the partner preference test, which measures the orientation of a male towards a sexual partner. Orientation is quantified as the amount of time spent with either a sexually receptive female or a non-receptive female (Kindon et al., 1996, Vagell and McGinnis, 1997, Vagell and McGinnis, 1998, Kelliher and Baum, 2001, Harding and McGinnis, 2003, Harding and McGinnis, 2004, Harding and McGinnis, 2005). In some experiments, an intact male is used in the place of a non-receptive female (i.e., Paredes and Baum, 1995). Additionally, motivation can be quantified by (1) counting the anticipatory level changes in a bi-level apparatus previously paired with sexual stimuli (Pfaus et al., 1990, Pfaus and Phillips, 1991, Roselli et al., 2003), or (2) measuring time spent in an environment linked to sexual events using the conditioned place preference test (Oldenburger et al., 1992, Pfaus et al., 2001, Harding and McGinnis, 2004).

Sexual arousal has also been measured with a variety of tasks (for review, see Sachs, 2007), including, but certainly not limited to, penile reflexes and non-contact erections in appropriate contexts (Sachs et al., 1994, Pfaus, 1996, Sachs, 2007), sniffing and pursuit behaviors (Giantonio et al., 1970), and more recently ultrasonic vocalizations in mice (James et al., 2006). It should also be noted that the concept of sexual arousal is often equated with sexual motivation (Stewart, 1995), adding to the confusion in the literature (Sachs, 2007). Male rats emit ultrasonic calls at two different frequencies: 20 and 50 kHz. While 20 kHz vocalizations are emitted in times of distress, 50 kHz vocalizations are more likely to reflect positive experiences (Portfors, 2007). These higher-frequency vocalizations may play an important role in coordinating reproductive activity, including attracting and soliciting sexual partners and regulating sexual interactions during copulation (McIntosh et al., 1979, Barfield and Thomas, 1986, Matochik and Barfield, 1991, Harding and McGinnis, 2005). Additionally, vocalizations are reported to reflect high levels of arousal, while being positively correlated with appetitive behaviors (Burgdorf et al., 2008). Our laboratory and others have shown that 50 kHz vocalizations can be elicited from sexually active males in response to salient sexual cues that are associated with reproduction, including soiled estrous bedding (Harding and McGinnis, 2003, McGinnis and Vakulenko, 2003, Harding and McGinnis, 2004, Harding and McGinnis, 2005) or sexually receptive females appearing briefly in a copulation arena (McGinnis and Vakulenko, 2003, Harding and McGinnis, 2005). Based on earlier research, we considered intromission and mount latencies as indirect measures of peripheral arousal (Saito and Moltz, 1986) and ultrasonic vocalizations in response to sexually specific stimuli as a measure of central arousal (James et al., 2006). Unlike partner preference, this task requires little energy expenditure to attain a goal (Sachs, 2007), and therefore we considered it to be separate from sexual motivation.

Arousal, motivational, and performance aspects of male rat reproductive behavior depend on circulating levels of testicular androgens, as evidenced by their reduction in occurrence in the weeks following castration (Meisel and Sachs, 1994, Hull et al., 2002, Harding and McGinnis, 2003, Bancoft, 2005) and their ability to be reinstated with physiological levels of testosterone (T) (Merkx, 1984, Matochik and Barfield, 1991, Harding and McGinnis, 2003, Bancoft, 2005) by either subcutaneous T injection (Hart, 1974), or implantation of T filled Silastic capsules under the skin (McGinnis and Dreifuss, 1989, McGinnis et al., 1989, McGinnis and Mirth, 1989, Krey and McGinnis, 1990). Replacement of T at sub-physiological levels has been shown to restore some components of sexual behavior in castrated male rodents including erectile function (Armagon et al., 2006) and ejaculatory behavior (Clark et al., 1995), as well as scent marking (Matochik and Barfield, 1991); while 50 kHz vocalizations appear to require larger amounts of T (Matochik and Barfield, 1991). The lower limit of T needed to restore these behaviors however, is not well clarified, and to our knowledge, no studies have examined the levels of T required to restore sexual motivation as measured by partner preference.

This study was also designed with the intent to create a new animal model for clinical conditions with symptoms of low sexual desire or arousal, including hypoactive sexual desire disorder, a condition that is marked by lack of sexual fantasies and desires for some period of time (Meuleman and Van Lankveld, 2005). The model developed here may closely mimic the human condition: producing males that have previously been sexually active that see a decline in reproductive behavior. By using a restoration paradigm, the proposed model allows for baseline measures of sexual arousal, motivation, and performance to be taken, for behaviors to be significantly reduced below “normal” levels, and for the effects of different drug treatments to be measured and compared to original levels of behavior.

The present study was therefore designed with two specific goals: (1) to compare the amount of T required to reinstate different components of reproductive behaviors in castrated male rats, and (2) to create an animal model for reduced—but not absent—reproductive behavior, the purpose of which is to aid future studies of impairments in male sexual behavior. Standard copulation tests were performed to measure male sexual performance, and provided some measure of peripheral sexual arousal. Partner preference was used to determine sexual motivation. We elected to measure 50 kHz vocalizations under two different conditions as a marker of (central) sexual arousal, and scent marking in response to sexual cues as another indicator of general ambulatory arousal that is androgen-dependent and closely tied to vocalizations (Matochik and Barfield, 1991). Rats were tested for all behaviors before and after castration as well as after implantation of Silastic capsules containing varying amounts of T. Based on previous findings, we expected vocalizations to require larger amounts of T to be restored compared to other sexual behaviors.