**Paradigms for Preclinical Investigations of Female Sexual Function and Dysfunction (HSDD and FSAD)**

Kelly A. Allers and Bernd Sommer *Boehringer Ingelheim Pharma GmbH & Co KG, Biberach an der Riss, Germany* 

## **1. Introduction**

A prerequisite for successful discovery of treatment options for female sexual dysfunction is a deeper understanding of sexual function per se, e.g. which physiological systems are involved, and what goes wrong in a dysfunctional state.. The availability of animal models, which capture the physiological underpinnings of a disorder and have been shown to respond to existing clinical treatments are a major success factor in this endeavor. For example, the translational ability of penile erection models are extremely good, since they have high predictive value for drugs used to treat erectile dysfunction. Predictive validity of models of female sexual dysfunction is less well established. However, results from recent efforts to back-translate effects from drugs that have shown efficacy in clinical trials into laboratory animals provide promising starting points for a better disease understanding and model validation.

This chapter will briefly outline those models that have potential in helping understand female sexual function and dysfunction, followed by three examples (flibanserin, bremolantide, apomorphine) of how clinically efficacious compounds contributed to elucidate the physiology and pharmacology underlying both the natural and pathological states (Bechara et al. 2004; Caruso et al. 2004b; Clayton et al. 2010; Safarinejad 2008) of arousal and desire disorders, the major indications in terms of prevalence (Johannes et al. 2009).

Modeling a complex human behavior in the laboratory imparts many challenges. Only a few fundamental issues faced in the laboratory are similar in the clinic – ie. the necessity to define a 'normal' range of responses for a given function before attempting to assess a dysfunctional range. In the clinic, questionnaire based tools have been developed, that can define 'healthy' and 'dysfunctional' states of sexual behavior within human populations. Furthermore, the two behavioral domains 'desire' and 'arousal' become distinguishable when dysfunction is present (DeRogatis et al. 2011). This can principally be achieved in the lab too, since each species has a clear set of behaviors used to solicit sexual contact. Although it is not always easy to separate desire from arousal in rodents as these two components of the sexual response are temporally connected in a natural setting, they can be isolated and studied as distinct physiological pathways. However, in order to have translational relevance, such laboratory studies must address measures aligned with those used clinically. With this request a number of issues arise that are unique to working with

Paradigms for Preclinical Investigations

treatments for sexual desire dysfunction.

**2.1 Behavioral models assessing sexual motivation** 

therefore decreases motivation (Paredes et al. 1997).

different questions related to motivated sexual behavior.

of Female Sexual Function and Dysfunction (HSDD and FSAD) 77

There is a long history in the behavioral psychology literature regarding assessment of motivation, incentives, and reward in the study of sexual function and the interpretation of this data (Agmo et al. 2004; Matthews et al. 1997; Pfaus 1996; Pfaus 2009b). This literature is ultimately relevant to the study of sexual function and interpretation of the studies described herein, but it is beyond the scope of this chapter. What will follow is a short description of current models with potential use in studying both normal and pathological sexual behaviors, what kind of information they provide, and their utility in development of

Behavioral models which are used to assess sexual motivation in general fall under two categories: those that measure directly the appetitive or proceptive behaviors used by females to show interest in a partner, and those that measure the reward value of a sexual

Pacing chambers: Pacing chambers are experimental boxes designed to allow a female rodent the ability to 'pace' the amount of contact she has with the male. The bi-level chamber is a commonly used design. In this design, the chamber itself consists of two levels with ramps at either end, allowing the female to escape or pursue her partner over both levels (Mendelson et al. 1987; Pfaus et al. 1992). Another type of pacing chamber consists of a plexiglass cage with a barrier in the middle. The barrier contains an opening big enough for the female to pass through, but not the male; alternatively another type of escape route is provided such as a chamber adjacent to a testing arena (Paredes et al. 1999; Peirce et al. 1961). In both of these designs, the female can approach or avoid the male according to her own level of motivation. In these studies behaviors that are meant to 'solicit' the attention of the male (eg ear wiggles, hops and darts, etc.) can be readily measured. In addition, the response to male mounting attempts can be assessed. Pacing chambers are crucial to assessing motivation for sexual contact in a female rodent, as a paradigm that does not allow the female to control the pacing of sexual contact does not induce reward, and

Partner Preference: Partner preference paradigms determine whether the test animal has more attraction to one animal over another. For example, a study examined whether a sexually receptive female rat prefers to spend time investigating another female rat, or a male rat (Clark et al. 2004). The goal of this type of study can be to determine what factors influence motivation for sexual contact as compared to social contact. In another study, female rats were allowed either paced copulation with almond-scented male rats or nonpaced copulations with non-scented males. When entered into a partner preference study, the females solicited almond-scented males more than non-scented, suggesting the animals associated the scent with more rewarding sexual contact (Coria-Avila et al. 2005). Clearly there are many variations on partner preference studies that can be utilised, answering

Primate models: Primate sexual behavior models provide an advantage over rodent models in that long-term primate pairs can be studied – mimicing the human situation. The hormone-dependency of primate sexual behavior is less strict than in rodents – again more like humans. Rhesus monkeys have commonly been used to assess proceptive behaviors.

encounter, providing a measure of the incentive to engage in sexual activity.

**2.1.1 Behavioral models assessing appetitive/proceptive behaviors** 

laboratory animals and are eloquently summarized by a quote from Professor Jim Pfaus, a well respected researcher in sexual function. 'Animals don't lie, but they don't talk, either'. There is no equivalent of a Female Sexual Function Index (Rosen et al. 2000 ) or clinical interview for laboratory rats. The interpretation of their behavior is the responsibility of a well-informed researcher who is devoted to understanding the motivations and actions of the animals. At first sight physiological studies appear easier to translate, a neuron fires or it doesn't, a neurotransmitter is released or it isn't, and usually standard statistical methods scan determine if a given effect is significant. However, whether these data are meaningful to the human situation again requires a careful interpretation of the study.

The preclinical testing of a drug candidate would ideally occur in a 'disease' model, i.e. a model that mimics some or all of the symptoms observed in the human disorder. Since a drug to treat sexual dysfunction should be expected to restore natural sexual behavior, NOT to promote increased sexual behavior in a healthy individual, the appropriate model should show some hyposexuality. However, manipulations that induce a state of hyposexuality in animals are not without disadvantage to research. On one hand, these models will allow the investigator to assess a drug's ability to restore natural behavior. On the other hand, whatever hormonal/pharmacological/behavioral manipulation that was used to induce hyposexuality becomes a confound that may bias the experiment. Careful interpretation of the data is required, and preferably a known predictive ability for clinical efficacy within the model can provide confidence in its usefulness.

With the abovementioned clinically efficacious compounds, substantial multidisciplinary investigation has ensued to elucidate the pharmacological mechanism of action of these drugs. This is just the boost that female sexual dysfunction research has needed, clinically efficacious drugs that can be used in back-translation to validate animal models, and the resources of the pharmaceutical industry to tenaciously investigate pharmacological mechanisms.

#### **2. Measuring 'desire' in animal models**

There is no single definition of desire. The DSM-IV-R defines the primary symptom of Hypoactive Sexual Desire Disorder as 'Deficient or absent sexual fantasies and desire for sexual activity'(American Psychiatric Association 2000), which suggests that sexual desire is the presence of fantasies and the desire for sexual activity….although this is clearly not a functional working definition. Perhaps Agmo's strategy in his recent review - to simply consult the dictionary - is more relevant (Agmo et al. 2004). The Merriam Webster dictionary suggests desire is: 1**:** conscious impulse toward something that promises enjoyment or satisfaction in its attainment, 2*a*: longing, craving *b* **:** sexual urge or appetite, 3**:** a usually formal request or petition for some action (http://www.merriam-webster.com/dictionary). With this definition paradigms that assess the different components can be obtained. We can assess whether an animal is motivated to seek a partner for sexual activity and whether this activity was enjoyed using paradigms that allow the female rat to choose place, and timing of copulation (and sometimes the partner as well), and studies that determine if there was reward value to the encounter. Studies such as the bilevel pacing chamber, in which the female rat solicits the male with species specific behaviors demonstrate a type of 'request' for action. Craving can be assessed in a similar way as it has been done in studies of drug addiction and craving, by determining with operant tasks whether an animal will 'work' toward achieving a goal (sexual activity or drug of addiction).

laboratory animals and are eloquently summarized by a quote from Professor Jim Pfaus, a well respected researcher in sexual function. 'Animals don't lie, but they don't talk, either'. There is no equivalent of a Female Sexual Function Index (Rosen et al. 2000 ) or clinical interview for laboratory rats. The interpretation of their behavior is the responsibility of a well-informed researcher who is devoted to understanding the motivations and actions of the animals. At first sight physiological studies appear easier to translate, a neuron fires or it doesn't, a neurotransmitter is released or it isn't, and usually standard statistical methods scan determine if a given effect is significant. However, whether these data are meaningful

The preclinical testing of a drug candidate would ideally occur in a 'disease' model, i.e. a model that mimics some or all of the symptoms observed in the human disorder. Since a drug to treat sexual dysfunction should be expected to restore natural sexual behavior, NOT to promote increased sexual behavior in a healthy individual, the appropriate model should show some hyposexuality. However, manipulations that induce a state of hyposexuality in animals are not without disadvantage to research. On one hand, these models will allow the investigator to assess a drug's ability to restore natural behavior. On the other hand, whatever hormonal/pharmacological/behavioral manipulation that was used to induce hyposexuality becomes a confound that may bias the experiment. Careful interpretation of the data is required, and preferably a known predictive ability for clinical efficacy within the

With the abovementioned clinically efficacious compounds, substantial multidisciplinary investigation has ensued to elucidate the pharmacological mechanism of action of these drugs. This is just the boost that female sexual dysfunction research has needed, clinically efficacious drugs that can be used in back-translation to validate animal models, and the resources of the

There is no single definition of desire. The DSM-IV-R defines the primary symptom of Hypoactive Sexual Desire Disorder as 'Deficient or absent sexual fantasies and desire for sexual activity'(American Psychiatric Association 2000), which suggests that sexual desire is the presence of fantasies and the desire for sexual activity….although this is clearly not a functional working definition. Perhaps Agmo's strategy in his recent review - to simply consult the dictionary - is more relevant (Agmo et al. 2004). The Merriam Webster dictionary suggests desire is: 1**:** conscious impulse toward something that promises enjoyment or satisfaction in its attainment, 2*a*: longing, craving *b* **:** sexual urge or appetite, 3**:** a usually formal request or petition for some action (http://www.merriam-webster.com/dictionary). With this definition paradigms that assess the different components can be obtained. We can assess whether an animal is motivated to seek a partner for sexual activity and whether this activity was enjoyed using paradigms that allow the female rat to choose place, and timing of copulation (and sometimes the partner as well), and studies that determine if there was reward value to the encounter. Studies such as the bilevel pacing chamber, in which the female rat solicits the male with species specific behaviors demonstrate a type of 'request' for action. Craving can be assessed in a similar way as it has been done in studies of drug addiction and craving, by determining with operant tasks whether an animal will 'work' toward achieving a

pharmaceutical industry to tenaciously investigate pharmacological mechanisms.

to the human situation again requires a careful interpretation of the study.

model can provide confidence in its usefulness.

**2. Measuring 'desire' in animal models** 

goal (sexual activity or drug of addiction).

There is a long history in the behavioral psychology literature regarding assessment of motivation, incentives, and reward in the study of sexual function and the interpretation of this data (Agmo et al. 2004; Matthews et al. 1997; Pfaus 1996; Pfaus 2009b). This literature is ultimately relevant to the study of sexual function and interpretation of the studies described herein, but it is beyond the scope of this chapter. What will follow is a short description of current models with potential use in studying both normal and pathological sexual behaviors, what kind of information they provide, and their utility in development of treatments for sexual desire dysfunction.

### **2.1 Behavioral models assessing sexual motivation**

Behavioral models which are used to assess sexual motivation in general fall under two categories: those that measure directly the appetitive or proceptive behaviors used by females to show interest in a partner, and those that measure the reward value of a sexual encounter, providing a measure of the incentive to engage in sexual activity.

## **2.1.1 Behavioral models assessing appetitive/proceptive behaviors**

Pacing chambers: Pacing chambers are experimental boxes designed to allow a female rodent the ability to 'pace' the amount of contact she has with the male. The bi-level chamber is a commonly used design. In this design, the chamber itself consists of two levels with ramps at either end, allowing the female to escape or pursue her partner over both levels (Mendelson et al. 1987; Pfaus et al. 1992). Another type of pacing chamber consists of a plexiglass cage with a barrier in the middle. The barrier contains an opening big enough for the female to pass through, but not the male; alternatively another type of escape route is provided such as a chamber adjacent to a testing arena (Paredes et al. 1999; Peirce et al. 1961). In both of these designs, the female can approach or avoid the male according to her own level of motivation. In these studies behaviors that are meant to 'solicit' the attention of the male (eg ear wiggles, hops and darts, etc.) can be readily measured. In addition, the response to male mounting attempts can be assessed. Pacing chambers are crucial to assessing motivation for sexual contact in a female rodent, as a paradigm that does not allow the female to control the pacing of sexual contact does not induce reward, and therefore decreases motivation (Paredes et al. 1997).

Partner Preference: Partner preference paradigms determine whether the test animal has more attraction to one animal over another. For example, a study examined whether a sexually receptive female rat prefers to spend time investigating another female rat, or a male rat (Clark et al. 2004). The goal of this type of study can be to determine what factors influence motivation for sexual contact as compared to social contact. In another study, female rats were allowed either paced copulation with almond-scented male rats or nonpaced copulations with non-scented males. When entered into a partner preference study, the females solicited almond-scented males more than non-scented, suggesting the animals associated the scent with more rewarding sexual contact (Coria-Avila et al. 2005). Clearly there are many variations on partner preference studies that can be utilised, answering different questions related to motivated sexual behavior.

Primate models: Primate sexual behavior models provide an advantage over rodent models in that long-term primate pairs can be studied – mimicing the human situation. The hormone-dependency of primate sexual behavior is less strict than in rodents – again more like humans. Rhesus monkeys have commonly been used to assess proceptive behaviors.

Paradigms for Preclinical Investigations

are overactive, or both.

primes is certainly warranted.

of Female Sexual Function and Dysfunction (HSDD and FSAD) 79

means of assessing thoughts or fantasies in animals. Furthermore, the underlying neuropathology of HSDD is essentially unknown. However, neuroimaging studies indicate that there are clearly differences in brain activation patterns of women suffering from HSDD when compared to healthy volunteers, confirming a neurological basis for symptoms (Arnow, 2009). With the substantial literature from preclinical studies, we are beginning to understand how the neurotransmitter pathways involved in sexual motivation not only function, but also how they can be dysfunctional. Michael Perleman introduced the concept of the 'Sexual Tipping Point'®, which incorporates the myriad of influences on a person when calculating their sexual desire (Perelman 2009). This concept describes the 'excitatory' and 'inhibitory' factors that are imparted by physiological, psychosocial, and cultural influences. Certainly from a neurobiological perspective, this concept holds true (Bancroft et al. 2009 ). Scientific evidence for this kind of balance in the neurobiology of desire has recently been reviewed, and an evidence based hypothesis generated for the underlying pathology of HSDD (Bancroft et al. 2009; Pfaus 2009b). Neurotransmitters such as dopamine and norepinephrine have a role in stimulating sexual activity, while serotonin plays an inhibitory role. Steroid and stress hormones also play a role in inducing sexual excitation or inhibition. These neurochemical pathways provide the 'balance' which can be tipped towards or away from motivation for sexual contact depending on intrinsic and environmental factors. The basic hypothesis for the neuropathology of HSDD is that either the systems regulating sexual excitation are inefficient, or the systems regulating inhibition

Preclinical studies have demonstrated multiple means of making a female animal 'hyposexual' by manipulating the excitatory and inhibitory pathways regulating sexual function. Following are descriptions and suggestions of what might be useful in drug

Hormone manipulations: The simplest method of achieving low desire in animals is to remove hormone cycling by ovariectomy. Without further manipulation, rodents do not engage in sexual contact and this could potentially be used to assess compounds meant to restore sexual health (Lopez et al. 2007). However, ovariectomised animals are not an ideal model for HSDD since a large proportion of patients are premenopausal and do not show altered hormone status. They may, however, serve to model decreased libido in postmenopausal women. A further option is to exploit a rat's intrinsic hormone cycling. As a female rat moves from estrus into metestrus her physiology adapts by sending a 'stop' signal in the brain. This is hormonally mediated, but with correlated neurophysiological changes (Hawcock et al. 2010; Lopez et al. 2007; Richards et al. 2010). Attempting laboratory studies with only animals in metestrus is not easy, as this stage of their cycle lasts only a few hours, but subhormone priming can induce a steady-state that resembles metestrous behaviorally and in histopathology of vaginal smears (Hawcock et al. 2010). Furthermore, this type of priming has been used to show increases in sexual functioning with both apomorphine and melanotan II (the precursor to bremelanotide)(Allers et al. 2010d; Hawcock et al. 2010). In keeping with this idea, multiple studies using the behavioral paradigms mentioned above have been utilised with sub-estrus hormone primes which induce hyposexual behavior. Further investigation of the potential of sub-estrus hormone

Pharmacological manipulations: Two models have potential utility that use pharmacological alterations. First is SSRI-induced hyposexuality. Chronic or sub-chronic treatment with SSRIs in rats reduces sexual activity, although the effects on proceptivity compared to

development, for testing of potential compounds to treat HSDD.

However, primate models also introduce further complexities of social relationships and status, different social structures which can regulate sexual activity. In rhesus monkeys, it is well established that the laboratory setting can substantially alter 'normal' sexual motivation when these relationships cannot be controlled for (Wallen 1990). Marmosets are small primates that form long term pair-bonds and have been successfully utilised for studies of sexual motivation in a laboratory setting. This design has allowed detailed study of hormonal influences and pharmacological manipulations to be assessed for their effects on sexual behavior within the pair (Barnett et al. 2006; Kendrick et al. 1985).

#### **2.1.2 Behavioral models assessing reward value of sexual activity**

Place Preference: Place preference is a test in which an animal has a choice between two chambers, and learns to associate one with a rewarding experience (Tzschentke 2007). For example, at the beginning of an experiment, the animal is placed into a cage which is separated in half by a plexiglass divider, but with an opening through which the animal can pass freely. With no conditioned stimulus, the animal will have no preference between the two sides and spends equal time in both. However if an animal is taught to associate a positive experience in one chamber, then the animal will choose to spend time in the chamber in which the positive experience occurred. Alternatively, animals show a preference for a given chamber at the start of the test, and the experimental manipulation is used to change the preference to the alternate chamber. This test has been used to demonstrate a rewarding experience for paced sexual contact (Paredes et al. 1997).

Operant behavior designs: In operant behavior studies, the animal is taught that a sequence of behaviors will lead to a given outcome. For example, a female rat learns that pressing a lever will lead to access to a sexually active male rat (Bermant et al. 1966). In this situation, if the female rat is motivated to engage in sexual behavior, pressing the lever and achieving this access to a male serves as a reinforcer. In one such study, the male was removed following a mount, an intromission, or an ejaculation. The latency for the female to press the level for further contact was recorded. The response latencies were shortest when the females were only mounted, intermediate when intromissions occurred, and longest following ejaculation. These data indicate that the animal was more motivated to continue sexual contact before the male ejaculated than after (Bermant 1961).

#### **2.2 Disorders of desire and potential animal models**

Hypoactive sexual desire disorder (HSDD) is defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) as persistent or recurrent deficiency or absence of sexual fantasies and thoughts, and/or desire for, or receptivity to, sexual activity, which causes personal distress or interpersonal difficulties and is not caused by a medical condition or drug (American Psychiatric Association 2000). This disorder has a prevalence of between 6-10% in females assessed from the US population (Johannes et al. 2009; Simon 2010).

In designing preclinical experiments to investigate the biology of HSDD, the ideal would be a population of animals which actually exhibit the symptoms of this disorder with the same underlying neuropathology as patients. Clearly that is not possible. We can measure sexual motivation with the multiple behavioral models mentioned above, and in some situations distress and interpersonal difficulties may even be inferred from animal behavior such as aggressive rejection in rodents or fighting between primate pair-mates, there is no objective

However, primate models also introduce further complexities of social relationships and status, different social structures which can regulate sexual activity. In rhesus monkeys, it is well established that the laboratory setting can substantially alter 'normal' sexual motivation when these relationships cannot be controlled for (Wallen 1990). Marmosets are small primates that form long term pair-bonds and have been successfully utilised for studies of sexual motivation in a laboratory setting. This design has allowed detailed study of hormonal influences and pharmacological manipulations to be assessed for their effects on

Place Preference: Place preference is a test in which an animal has a choice between two chambers, and learns to associate one with a rewarding experience (Tzschentke 2007). For example, at the beginning of an experiment, the animal is placed into a cage which is separated in half by a plexiglass divider, but with an opening through which the animal can pass freely. With no conditioned stimulus, the animal will have no preference between the two sides and spends equal time in both. However if an animal is taught to associate a positive experience in one chamber, then the animal will choose to spend time in the chamber in which the positive experience occurred. Alternatively, animals show a preference for a given chamber at the start of the test, and the experimental manipulation is used to change the preference to the alternate chamber. This test has been used to

demonstrate a rewarding experience for paced sexual contact (Paredes et al. 1997).

sexual contact before the male ejaculated than after (Bermant 1961).

**2.2 Disorders of desire and potential animal models** 

population (Johannes et al. 2009; Simon 2010).

Operant behavior designs: In operant behavior studies, the animal is taught that a sequence of behaviors will lead to a given outcome. For example, a female rat learns that pressing a lever will lead to access to a sexually active male rat (Bermant et al. 1966). In this situation, if the female rat is motivated to engage in sexual behavior, pressing the lever and achieving this access to a male serves as a reinforcer. In one such study, the male was removed following a mount, an intromission, or an ejaculation. The latency for the female to press the level for further contact was recorded. The response latencies were shortest when the females were only mounted, intermediate when intromissions occurred, and longest following ejaculation. These data indicate that the animal was more motivated to continue

Hypoactive sexual desire disorder (HSDD) is defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) as persistent or recurrent deficiency or absence of sexual fantasies and thoughts, and/or desire for, or receptivity to, sexual activity, which causes personal distress or interpersonal difficulties and is not caused by a medical condition or drug (American Psychiatric Association 2000). This disorder has a prevalence of between 6-10% in females assessed from the US

In designing preclinical experiments to investigate the biology of HSDD, the ideal would be a population of animals which actually exhibit the symptoms of this disorder with the same underlying neuropathology as patients. Clearly that is not possible. We can measure sexual motivation with the multiple behavioral models mentioned above, and in some situations distress and interpersonal difficulties may even be inferred from animal behavior such as aggressive rejection in rodents or fighting between primate pair-mates, there is no objective

sexual behavior within the pair (Barnett et al. 2006; Kendrick et al. 1985).

**2.1.2 Behavioral models assessing reward value of sexual activity** 

means of assessing thoughts or fantasies in animals. Furthermore, the underlying neuropathology of HSDD is essentially unknown. However, neuroimaging studies indicate that there are clearly differences in brain activation patterns of women suffering from HSDD when compared to healthy volunteers, confirming a neurological basis for symptoms (Arnow, 2009). With the substantial literature from preclinical studies, we are beginning to understand how the neurotransmitter pathways involved in sexual motivation not only function, but also how they can be dysfunctional. Michael Perleman introduced the concept of the 'Sexual Tipping Point'®, which incorporates the myriad of influences on a person when calculating their sexual desire (Perelman 2009). This concept describes the 'excitatory' and 'inhibitory' factors that are imparted by physiological, psychosocial, and cultural influences. Certainly from a neurobiological perspective, this concept holds true (Bancroft et al. 2009 ). Scientific evidence for this kind of balance in the neurobiology of desire has recently been reviewed, and an evidence based hypothesis generated for the underlying pathology of HSDD (Bancroft et al. 2009; Pfaus 2009b). Neurotransmitters such as dopamine and norepinephrine have a role in stimulating sexual activity, while serotonin plays an inhibitory role. Steroid and stress hormones also play a role in inducing sexual excitation or inhibition. These neurochemical pathways provide the 'balance' which can be tipped towards or away from motivation for sexual contact depending on intrinsic and environmental factors. The basic hypothesis for the neuropathology of HSDD is that either the systems regulating sexual excitation are inefficient, or the systems regulating inhibition are overactive, or both.

Preclinical studies have demonstrated multiple means of making a female animal 'hyposexual' by manipulating the excitatory and inhibitory pathways regulating sexual function. Following are descriptions and suggestions of what might be useful in drug development, for testing of potential compounds to treat HSDD.

Hormone manipulations: The simplest method of achieving low desire in animals is to remove hormone cycling by ovariectomy. Without further manipulation, rodents do not engage in sexual contact and this could potentially be used to assess compounds meant to restore sexual health (Lopez et al. 2007). However, ovariectomised animals are not an ideal model for HSDD since a large proportion of patients are premenopausal and do not show altered hormone status. They may, however, serve to model decreased libido in postmenopausal women. A further option is to exploit a rat's intrinsic hormone cycling. As a female rat moves from estrus into metestrus her physiology adapts by sending a 'stop' signal in the brain. This is hormonally mediated, but with correlated neurophysiological changes (Hawcock et al. 2010; Lopez et al. 2007; Richards et al. 2010). Attempting laboratory studies with only animals in metestrus is not easy, as this stage of their cycle lasts only a few hours, but subhormone priming can induce a steady-state that resembles metestrous behaviorally and in histopathology of vaginal smears (Hawcock et al. 2010). Furthermore, this type of priming has been used to show increases in sexual functioning with both apomorphine and melanotan II (the precursor to bremelanotide)(Allers et al. 2010d; Hawcock et al. 2010). In keeping with this idea, multiple studies using the behavioral paradigms mentioned above have been utilised with sub-estrus hormone primes which induce hyposexual behavior. Further investigation of the potential of sub-estrus hormone primes is certainly warranted.

Pharmacological manipulations: Two models have potential utility that use pharmacological alterations. First is SSRI-induced hyposexuality. Chronic or sub-chronic treatment with SSRIs in rats reduces sexual activity, although the effects on proceptivity compared to

Paradigms for Preclinical Investigations

**3.1 Models of peripheral arousal** 

tool for early drug screening (Booth et al. 2010).

**3.2 Disorders of arousal and potential animal models** 

years.

copulation.

of Female Sexual Function and Dysfunction (HSDD and FSAD) 81

of the physiological mechanisms of arousal are endless. The circuitry from brainstem autonomic centers to the genitalia has been well studied, as have been the peripheral nerve innervation and structure and function of the genitalia. The understanding of healthy functioning arousal pathways, tissues, and biochemistry has grown substantially in recent

Below is a brief description of methods used to investigate peripheral arousal. Rather than review in detail the numerous techniques available, we will give a short description of the

Lordosis behavior: It is arguable whether a lordosis reflex posture can be used as a measure of arousal. This is a posture adopted by female rats which allows the male access for intromission and ejaculation. The reason for including this here is twofold: female rats will exhibit lordosis even in non-paced mating situations where a 'rewarding' sexual experience does not occur and is therefore distinct from motivation or desire, and female rats will typically resist intromission if they are in a non-receptive hormone state or in pain. For these reasons, possibly lordosis may be a surrogate marker for arousal, in that physiologically, her body may be prepared for sexual contact and it is not uncomfortable to engage in

Vaginal blood flow models: Typically these types of studies are carried out in aneasthetised animals, usually rats or rabbits (Beharry et al. 2003; Giuliano et al. 2010; Hale et al. 2003). A basic paradigm uses peripheral nerve stimulation (ie pudendal or pelvic) to stimulate blood flow to the clitoris and vagina. These studies have been critical in elucidating the mechanisms by which blood flow to the genitals is regulated, and the neurotransmitters and hormones involved in such regulation. The blood flow itself can be measured by techniques, such as plethysmography or laser doppler flowmetry. In addition to the standard rate or amplitude of flow, further characterisation of blood flow as measured by laser doppler can elicit surrogate measures of autonomic activation to the vagina (Allers et al. 2010d; Allers et al. 2010c). These measures are obtained by fast Fourier transform analysis of the oscillations within blood flow and are similar to those used to assess heart rate variability. Using this type of analysis responses to naturally arousing stimuli (a male rat) and experimental manipulations meant to induce arousal (drugs, nerve section, hormone status) are observed. In vitro studies: Studies on gential tissues also help to define the intricate biochemical and physiological mechanisms for healthy sexual functioning (Aughton et al. 2008; Wilson et al. 2009). These types of studies may investigate the mechanisms involved in vaginal smooth muscle contraction and relaxation, lubrication, or neurotransmitter release from innervating nerves. In addition, one unique model utilises an ex vivo brain tissue assay which shows responses that are highly predictive of lordosis behavior. In a drug development project where increased lordosis could be predictive of clinical efficacy, this assay provides an ideal

Female sexual arousal disorder (FSAD) is defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) as the inability to attain or maintain until completion of sexual activity adequate lubrication in response to sexual excitement (American Psychiatric Association 2000). Unfortunately, this definition does not truly seem to describe many FSAD patients since several studies have now shown that a

methods which can be utilised and the types of data generated.

receptivity are unclear (Frye et al. 2010; Guptarak et al. 2010; Matuszczyk et al. 1998). While SSRI-induced sexual dysfunction would not be diagnosed as HSDD in humans, it is still a disorder which requires treatment, and it provides a model of hyposexuality in animals with a well characterised clinical correlate. A further model that was recently presented uses a disruption of early sexual reward which causes animals to 'learn' not to enjoy sexual contact. During her first sexual experiences a female rat is treated with naloxone, which diminishes the reward signal obtained. In further testing this results in a long lasting decrease in sexual contact (Pfaus et al. 2008).

Natural variation in sexual motivation: One study has demonstrated that within a normal population of female laboratory rats, a subgroup can be found which show more avoidance of sexual contact when presented with a male. Hence, when compared to the population as a whole, these animals are hyposexual. Translational relevance remains to be determined however, since apomorphine – a drug that has demonstrated clinical efficacy – does not induce increased sexual behavior in this model (Snoeren et al. 2011)

Marmoset model: A marmoset monkey model of hyposexual activity has been developed that utilises a combination of sub-optimal hormone priming and separation of the male and female pair mates to induce low levels of copulation on reunion. In addition, because the marmoset forms long-term pairs, a study design can include more than one hormone prime in a cross-over design to assess hormone dependence of compounds. This model has recently been used to assess the efficacy of flibanserin, a drug that has demonstrated efficacy in HSDD women in the clinic (Aubert et al. 2009; Tannenbaum et al. 2007).

#### **3. Measuring arousal in animal models**

There are two types of arousal that are relevant to sexual function, generalised arousal and peripheral arousal. Generalised arousal is a state of awareness or attention that is given to an organism's surroundings. An animal with a high state of generalised arousal has greater awareness of sensory cues, more motor activity, and a high degree of emotional reactivity. Peripheral arousal during sexual activity is the physiological preparation for sexual contact, such as vaginal and clitoral engorgement and lubrication.

Generalised arousal is relevant to sexual function as varying levels of sensory awareness or attention may be linked to motivation (Schober et al. 2011). In this sense, generalised arousal may be more closely linked to desire, than to peripheral arousal. Indications of decreased general arousal in sexual dysfunction, in particular women with HSDD support this link to desire. Women with HSDD show different brain activation patterns compared to healthy controls when viewing erotic videos, women with HSDD demonstrate differences in attending to sexual cues compared to healthy controls, and also have different electroencphalographic excitability as measured by the P300 wave when compared to both healthy controls and to women with arousal disorder (Arnow et al. 2009; McCall et al. 2006; Vardi et al. 2009). Consequently, it is possible that HSDD results from decreased generalised arousal or is a cause of it. Either way, it is clearly an important part of the sexual cycle. Generalised arousal is thought to be a key feature of an animal preparing to engage in any motivated behavior, not just sexual contact. If generalised arousal is necessary for motivated behavior, and motivated behavior is best measured in the aforementioned models, perhaps these behaviors are a good surrogate for generalised arousal.

Peripheral arousal related to sexual function typically refers to the autonomic response of the body as it prepares for sexual contact. In theory, the possibilities for preclinical research of the physiological mechanisms of arousal are endless. The circuitry from brainstem autonomic centers to the genitalia has been well studied, as have been the peripheral nerve innervation and structure and function of the genitalia. The understanding of healthy functioning arousal pathways, tissues, and biochemistry has grown substantially in recent years.

Below is a brief description of methods used to investigate peripheral arousal. Rather than review in detail the numerous techniques available, we will give a short description of the methods which can be utilised and the types of data generated.

## **3.1 Models of peripheral arousal**

80 Sexual Dysfunctions – Special Issue

receptivity are unclear (Frye et al. 2010; Guptarak et al. 2010; Matuszczyk et al. 1998). While SSRI-induced sexual dysfunction would not be diagnosed as HSDD in humans, it is still a disorder which requires treatment, and it provides a model of hyposexuality in animals with a well characterised clinical correlate. A further model that was recently presented uses a disruption of early sexual reward which causes animals to 'learn' not to enjoy sexual contact. During her first sexual experiences a female rat is treated with naloxone, which diminishes the reward signal obtained. In further testing this results in a long lasting

Natural variation in sexual motivation: One study has demonstrated that within a normal population of female laboratory rats, a subgroup can be found which show more avoidance of sexual contact when presented with a male. Hence, when compared to the population as a whole, these animals are hyposexual. Translational relevance remains to be determined however, since apomorphine – a drug that has demonstrated clinical efficacy – does not

Marmoset model: A marmoset monkey model of hyposexual activity has been developed that utilises a combination of sub-optimal hormone priming and separation of the male and female pair mates to induce low levels of copulation on reunion. In addition, because the marmoset forms long-term pairs, a study design can include more than one hormone prime in a cross-over design to assess hormone dependence of compounds. This model has recently been used to assess the efficacy of flibanserin, a drug that has demonstrated efficacy

There are two types of arousal that are relevant to sexual function, generalised arousal and peripheral arousal. Generalised arousal is a state of awareness or attention that is given to an organism's surroundings. An animal with a high state of generalised arousal has greater awareness of sensory cues, more motor activity, and a high degree of emotional reactivity. Peripheral arousal during sexual activity is the physiological preparation for sexual contact,

Generalised arousal is relevant to sexual function as varying levels of sensory awareness or attention may be linked to motivation (Schober et al. 2011). In this sense, generalised arousal may be more closely linked to desire, than to peripheral arousal. Indications of decreased general arousal in sexual dysfunction, in particular women with HSDD support this link to desire. Women with HSDD show different brain activation patterns compared to healthy controls when viewing erotic videos, women with HSDD demonstrate differences in attending to sexual cues compared to healthy controls, and also have different electroencphalographic excitability as measured by the P300 wave when compared to both healthy controls and to women with arousal disorder (Arnow et al. 2009; McCall et al. 2006; Vardi et al. 2009). Consequently, it is possible that HSDD results from decreased generalised arousal or is a cause of it. Either way, it is clearly an important part of the sexual cycle. Generalised arousal is thought to be a key feature of an animal preparing to engage in any motivated behavior, not just sexual contact. If generalised arousal is necessary for motivated behavior, and motivated behavior is best measured in the aforementioned models, perhaps

Peripheral arousal related to sexual function typically refers to the autonomic response of the body as it prepares for sexual contact. In theory, the possibilities for preclinical research

decrease in sexual contact (Pfaus et al. 2008).

**3. Measuring arousal in animal models** 

such as vaginal and clitoral engorgement and lubrication.

these behaviors are a good surrogate for generalised arousal.

induce increased sexual behavior in this model (Snoeren et al. 2011)

in HSDD women in the clinic (Aubert et al. 2009; Tannenbaum et al. 2007).

Lordosis behavior: It is arguable whether a lordosis reflex posture can be used as a measure of arousal. This is a posture adopted by female rats which allows the male access for intromission and ejaculation. The reason for including this here is twofold: female rats will exhibit lordosis even in non-paced mating situations where a 'rewarding' sexual experience does not occur and is therefore distinct from motivation or desire, and female rats will typically resist intromission if they are in a non-receptive hormone state or in pain. For these reasons, possibly lordosis may be a surrogate marker for arousal, in that physiologically, her body may be prepared for sexual contact and it is not uncomfortable to engage in copulation.

Vaginal blood flow models: Typically these types of studies are carried out in aneasthetised animals, usually rats or rabbits (Beharry et al. 2003; Giuliano et al. 2010; Hale et al. 2003). A basic paradigm uses peripheral nerve stimulation (ie pudendal or pelvic) to stimulate blood flow to the clitoris and vagina. These studies have been critical in elucidating the mechanisms by which blood flow to the genitals is regulated, and the neurotransmitters and hormones involved in such regulation. The blood flow itself can be measured by techniques, such as plethysmography or laser doppler flowmetry. In addition to the standard rate or amplitude of flow, further characterisation of blood flow as measured by laser doppler can elicit surrogate measures of autonomic activation to the vagina (Allers et al. 2010d; Allers et al. 2010c). These measures are obtained by fast Fourier transform analysis of the oscillations within blood flow and are similar to those used to assess heart rate variability. Using this type of analysis responses to naturally arousing stimuli (a male rat) and experimental manipulations meant to induce arousal (drugs, nerve section, hormone status) are observed. In vitro studies: Studies on gential tissues also help to define the intricate biochemical and physiological mechanisms for healthy sexual functioning (Aughton et al. 2008; Wilson et al. 2009). These types of studies may investigate the mechanisms involved in vaginal smooth muscle contraction and relaxation, lubrication, or neurotransmitter release from innervating nerves. In addition, one unique model utilises an ex vivo brain tissue assay which shows responses that are highly predictive of lordosis behavior. In a drug development project where increased lordosis could be predictive of clinical efficacy, this assay provides an ideal tool for early drug screening (Booth et al. 2010).

#### **3.2 Disorders of arousal and potential animal models**

Female sexual arousal disorder (FSAD) is defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) as the inability to attain or maintain until completion of sexual activity adequate lubrication in response to sexual excitement (American Psychiatric Association 2000). Unfortunately, this definition does not truly seem to describe many FSAD patients since several studies have now shown that a

Paradigms for Preclinical Investigations

function and dysfunction.

**4.1 Apomorphine** 

(Subramony 2006).

al. 2011a).

sexual behavior in rodents.

the animal into different stages of the estrous cycle.

of Female Sexual Function and Dysfunction (HSDD and FSAD) 83

following examples demonstrate how the process of drug discovery and new drugs available for testing has contributed substantially to the scientific understanding of sexual

Drug development teams have adopted multiple strategies for assessing dopamine receptors as a drug target, including assessing the roles of individual dopamine receptor subtypes, dopamine reuptake, or using a classical dopamine receptor agonist (apomorphine) in different formulations. Apomorphine is a non-selective dopamine agonist was first discovered in 1869 and over time has been used as an emetic, a treatment for alcohol and morphine addictions, and to improve symptoms of Parkinson's Disease

Dopamine has long been known to be a modulator of sexual function. There are decades of literature reporting pro-sexual effects of dopamine agonists but also reports of the same paradigms producing reduced sexual behavior. Much research has focussed on elucidating the hormone dependence of effects, the regions of the brain where increased or decreased dopamine occurs during sexual contact in efforts to investigate the real role of dopamine in sexual function and where discrepancies in data may come from. Many review articles have highlighted ongoing questions related to dopamine in sexual function (Meisel et al. 2006; Paredes et al. 2004; Peeters et al. 2008; Pfaus 2009a; Stolzenberg et al. 2011b; Stolzenberg et

Regardless of the scientific debates over how dopamine regulates sexual function, clinical evidence has indicated that stimulating dopamine receptors may provide help for women with sexual dysfunction (Bechara et al. 2004; Caruso et al. 2004a). For this reason, studies were undertaken to investigate further why apomorphine can have opposing effects on

Using a partner preference paradigm in which a rat chooses to actively investigate either a sexually vigorous male or a castrated male, a measure of active investigation can be utilised to determine sexual interest compared to social interest. In this paradigm, female rats that are ovariectomised and given a sub-hormone prime to resemble metestrus do not show a preference for either male rat. Upon apomorphine treatment, a dose dependent increase in preference for the sexually vigorous male is observed. When this study is run with animals that have been fully hormonally primed to resemble behavioral estrous, they show a clear preference for the sexually vigorous male, which is decreased by apomorphine treatment (Hawcock et al. 2010). These data indicate that when ALL other conditions are equal dopamine receptor agonism can have the exact opposite effect simply by artificially placing

Further investigations suggest that the opposing effects of apomorphine occur in the naturally cycling animal in metestrus compared to estrous. Using the FFT analysis of laser doppler flowmetry in rats, these differential effects of apomorphine are also present. Metestrus animals had a significant increase in slow oscillatory activity of blood flow, which in this model is indicative of sexual arousal (Allers et al. 2010d). Furthermore, this increase could not be blocked by the peripheral antagonist domperidone, but could be attenuated with the centrally acting antagonist haloperidol, indicating apomorphine's actions did originate in the brain (Allers et al. 2010c). In estrous animals, apomorphine elicited a decrease in slow oscillatory activity indicating decreased arousal. The dose ranges in both

patient will rate her subjective feelings of arousal quite low, but tends to show a normal genital response (Laan et al. 2008; Rellini et al. 2006). Classifying patients into subgroups of mainly genital arousal, or subjective arousal dysfunction in studies using VPA begins to show some differences, but subjective assessments remain the best diagnostic tools (Both et al. 2010; Salonia et al. 2010). In healthy women, there is not always agreement between subjective and physiologically measured responses either, perhaps due to interoceptive differences in attending to one's own genital physiology.

Even though there is a wealth of information on healthy arousal responses, there is no evidence based hypothesis for the neuropathology of FSAD. Clear differences in sympathetic responsiveness has been demonstrated in patients as compared to controls using procedures such as hyperventilation or extreme exercise (Brotto et al. 2009; Meston 2000). This points to a pathology in the systemic autonomic nervous system, or centralautonomic interface, rather than in the genitalia.

The majority of models above cannot provide translational disease models for FSAD, since vaginal vasocongestion is not affected in a large number of women with FSAD, and the likely pathology is within the circuitry of the autonomic or central nervous system. Only one technique above has demonstrated that it can actually distinguish between healthy controls and women with FSAD, and has been demonstrated to have predictive utility.

Slow oscillations in vaginal blood flow: In this model, laser Doppler flowmetry or vaginal plethysmograph can be used to record blood flow within the vaginal wall (Allers et al. 2010d; Allers et al. 2010c). Rather than assessing peak amplitudes, rate or other typical measures, the trace is analysed by fast Fourier transform (FFT) to elucidate the oscillatory charactistics of flow. This technique is well established in vascular research and these oscillations are known to reflect autonomic nervous system input to the tissue being studied. In this way, these measures are a surrogate of autonomic input. When measured in the vagina, the changes seen during sexual arousal paradigms are specific to the vagina, meaning they do not occur in other tissues simultaneously. This method has been used to differentiate human patients, correlates highly with subjective arousal in both healthy and FSAD women, and has been used in animals to show arousal inducing effects of apomorphine and melanotan II, in addition to natural arousal (exposure to sexually active male). Upon apomorphine administration animals in metestrus respond with increased slow oscillations in this model, at the same doses that induce restoration of sexual behavior in a partner-preference test (Hawcock et al. 2010). This model has not yet been used clinically to demonstrate restoration of natural arousal in FSAD women with a drug candidate, but has great potential as a translational model for investigation of drugs to treat FSAD.

#### **4. Lessons from drug discovery**

The process of drug discovery requires that a target (eg receptor, enzyme, etc.) be identified with a reasonable hypothesis for why it is engaged in a given disease state. This hypothesis is then rigorously tested within the laboratory with tool compounds and drug candidates. Alternatively, and typical of the sexual health field, a prosexual side effect is noticed for a given drug candidate, and this is quickly followed up to assess its potential for therapeutic use in sexual medicine. This process leads to a substantial amount of research into pathways, disease states, and pharmacological mechanisms regarding a particular target being generated. Within sexual health groups in the pharmaceutical industry recent targets have included dopamine receptors, serotonin receptors, and melanocortin receptors. The following examples demonstrate how the process of drug discovery and new drugs available for testing has contributed substantially to the scientific understanding of sexual function and dysfunction.
