A translational approach to vocalization deficits and neural recovery after behavioral treatment in Parkinson disease

Abstract

Parkinson disease is characterized by a complex neuropathological profile that primarily affects dopaminergic neural pathways in the basal ganglia, including pathways that modulate cranial sensorimotor functions such as swallowing, voice and speech. Prior work from our lab has shown that the rat model of unilateral 6-hydroxydopamine infusion to the medial forebrain bundle that is useful for studying limb sensorimotor deficits also yields vocalization deficits that may be amenable to treatment with intensive exercise. This affords us an opportunity to explore the potential mechanisms underlying behavioral and neural recovery as a result of intervention for cranial sensorimotor deficits associated with Parkinson disease. Our methods include recording and acoustic analysis of male rat ultrasonic vocalizations in a control condition, after neurotoxin infusion (Parkinson disease model), and after targeted vocalization training. We also use well-established behavioral and immunohistochemical methods to assess the level of neurochemical recovery in the striatum of the basal ganglia after our interventions. Our findings, although preliminary, prompt us to look in other brain regions extraneous to the striatum for potential underlying mechanisms of recovery. Thus, our future work will focus on the underlying mechanisms of behavioral recovery in a Parkinson disease model in the hope that this will lead to improved understanding of brain function and improved treatment for voice and swallowing disorders.

Learning outcomes: Readers will gain an understanding of how a rat model of Parkinson disease is used to study vocalization deficits and interventions.

Introduction

Parkinson disease (PD) is characterized by a complex neuropathological profile that primarily affects dopaminergic neural pathways in the basal ganglia (Bergman and Deuschl, 2002, Braak et al., 2004), including pathways that modulate cranial sensorimotor functions such as swallowing, voice and speech. Thus, PD leads to dysphagia, dysphonia, and dysarthria that compromise health and quality of life (Fox et al., 2002, Ho et al., 1998, Plowmann-Prine et al., 2009). Common voice deficits associated with PD include breathiness, hoarse vocal quality, decreased vocal loudness, and decreased frequency variability (Darley et al., 1969a, Darley et al., 1969b, Fox et al., 2002, Ho et al., 1998, Logemann et al., 1978). Speech and voice therapy in the form of intensive exercise has been shown to improve these deficits and quality of life in patients with PD (Ramig et al., 2001a, Ramig et al., 2001b, Sapir et al., 2006). However, the underlying mechanisms of these behavioral interventions are not well understood. Data from a human imaging study after intensive intervention for parkinsonian dysarthria (Lee Silverman Voice Therapy) showed a more ‘normalized’ pattern of activation in the cortical motor and premotor areas of the brain and also additional recruitment of right anterior insula, dorsolateral prefrontal cortex, and basal ganglia (caudate head, putamen) (Liotti et al., 2003). Therefore, although data are limited to one study, it appears that behavioral intervention can alter neuronal patterns of activity.

In contrast to limited clinical data regarding the effects of behavioral therapy on neural pathways underlying cranial functions, there is a body of animal research demonstrating that intensive limb exercise leads to sparing of striatal dopamine neurons if started early in the disease process (Anstrom et al., 2007, Mabandla et al., 2004, Tillerson et al., 2001, Woodlee and Schallert, 2004). However, it is not known if behavioral interventions for vocalization deficits have these same effects. Thus, there is limited knowledge of the underlying mechanisms for improvement in cranial sensorimotor control with treatment for PD.

It is tempting to use knowledge gained from behavioral studies of animal limb sensorimotor systems as explanations for the positive effects observed after voice and speech treatment in PD. However, common drug and surgical interventions that clearly benefit limb function do not appear to benefit cranial motor function to the same extent (Ciucci et al., 2008a, Fuh et al., 1997, Hunter et al., 1997, Narayana et al., 2009, Potulska et al., 2003). There is not yet an adequate explanation or empirical data to explain the differential effects of pharmacologic and surgical treatments for PD on limb versus cranial systems. Perhaps cranial sensorimotor systems are modulated by dopamine in a different manner or by other non-dopaminergic neurotransmitters than those employed for limb actions. Putative mechanisms such as these should be investigated in future research.

Exploring potential mechanisms for cranial deficits in persons with PD requires invasive procedures and control of age, disease severity, medications, and environment. To overcome these methodological limitations, animal models can be used as a translational approach to studying interventions for PD. It is essential that we investigate vocalization deficits in awake animals with appropriate behavioral assays. While it is intuitively appealing to apply well-used measurements made in humans to the study of animals, our measures must be relevant to the animal in terms of vocalization, and sensitive to changes with PD and intervention. This concept is summarized by Cenci, Whishaw, and Schallert (2002):

…the modeling of human-like symptoms in animals should be made primarily on the basis of an expectation of functional similarity, rather than on one of physical identity. The first question to ask is not whether a rat would show a given neurological symptom, but rather, how that neurological symptom would manifest itself in a rat. (p. 574)

Although we formulate different measures than what we use in human evaluation, these measures can provide important insights into sensorimotor control processes relevant to humans when interpreted in the appropriate context.

To study the effects of PD and behavioral treatment in animals, the first step is to model the effects of the disease. These models are often based on a systemic pharmacologic administration or brain lesion. A model, in the best of circumstances, can only serve as an approximation to the human situation it is designed to reflect. As such, findings are interpreted with the appropriate caution that should always be applied to all experimentation in the medical and social sciences. Albeit imperfect, an appropriate animal model can provide critical insights into putative mechanisms underlying health and disease in humans.

One widely used technique for creating a model of PD in the rat is depleting dopamine unilaterally with micro-infusion of the neurotoxin 6-hydroxydopamine (6-OHDA) to the medial forebrain bundle, a major dopaminergic pathway affected by PD. The medial forebrain bundle consists of neurons that originate in the substantia nigra and synapse in the striatum. The neurotoxin 6-OHDA leads to the degeneration of dopamine cells and quantifiable deficits that mimic those found in humans in the early stages of PD (Cenci et al., 2002, Fulceri et al., 2006, Marshall, 1979, Meredith and Kang, 2006, Ungerstedt and Arbuthnott, 1970). This model has been used in prior examinations of PD-related sensorimotor deficits and neural modulation associated with intervention (Cenci et al., 2002, Fleming et al., 2005, Meredith and Kang, 2006). Therefore the 6-OHDA model can be useful in examining some aspects of the deficits associated with PD.

Because of its extensive prior use, the 6-OHDA model of PD is associated with a wealth of behavioral tests to estimate lesion severity and recovery with intervention. The “forelimb asymmetry test,” also called the “cylinder test,” capitalizes on the unilateral lesion model because the affected limb can be compared with the intact limb during spontaneous movement. To perform this test, the awake rat is placed into a tall, clear plexiglass cylinder and observed while engaging in natural behaviors. Specifically, the use of the each forelimb is counted while the rat rears and explores. Rats with unilateral dopamine depletion preferentially use the unimpaired forelimb for support during exploration and show little or no use of the impaired forelimb (Schallert and Woodlee, 2005, Tillerson et al., 2001, Tillerson et al., 2002). Using a simple formula, the counted data are converted to a score that correlates highly with the amount of dopamine depletion found in the striatum (Woodlee & Schallert, 2004). Thus, this behavioral measure provides an accurate estimate of the extent of brain lesion and the associated sensorimotor deficit. Behavioral tests of this kind have been very useful in the study of brain regions and pathways underlying limb motor impairments in models of PD. The extent to which these tests and resultant measures apply to brain mechanisms underlying cranial sensorimotor disruption is unknown.

Most of the work in the 6-OHDA model of PD has been done in the limbs, but there are preliminary data concerning cranial motor systems from our laboratories. Specifically, we have found that a unilateral 6-OHDA lesion also leads to vocalization deficits (Ciucci et al., 2007, Ciucci et al., 2008b, Ciucci et al., 2009). Under normal conditions, rats produce calls in the ultrasonic frequency range. A subset of these ultrasonic calls have frequencies that center around 50-kHz and are used to locate other rats, during play, in mother–pup interactions, and during sexual encounters (Bialy et al., 2000, Brudzynski, 2005, Brudzynski and Ociepa, 1992, Brudzynski and Pniak, 2002, McGinnis and Vakulenko, 2003). These calls are thought to be semiotic, or convey meaning (Brudzynski, 2005). In our work, which utilizes sexual encounters to elicit calls, we determined that three different types of calls are produced while a male rat calls to a female rat in estrus: simple, frequency modulated, and harmonic (Fig. 1) (Ciucci et al., 2009). A rat in the control (unlesioned) condition primarily produces frequency modulated calls, but produces mostly simple calls after a unilateral 6-OHDA lesion (Ciucci et al., 2009, Ciucci et al., 2008b). Within each call type, we also analyzed duration, bandwidth, and intensity and found that bandwidth and intensity are also vulnerable to striatal dopamine depletion (Ciucci et al., 2007, Ciucci et al., 2009). Our data suggest that similar to humans, striatal dopamine loss is associated with vocalization deficits and that the 6-OHDA model is useful for studying appropriately formulated paradigms regarding brain changes associated with PD and recovery with intervention. Thus, with the unilateral 6-OHDA model of PD, we study the nature of recovery for vocalization deficits with and without behavioral interventions. Specifically, we are addressing whether there is a behavioral improvement of vocalization following intensive vocal exercise and whether this recovery involves rescue or regeneration of striatal dopamine neurons.