Evidence for a specific role for muscarinic receptors in crossmodal object recognition in rats

Highlights

• Rats display crossmodal (tactile-to-visual) object recognition (CMOR).

• Acetylcholine may be important for multisensory binding.

• Scopolamine disrupts memory encoding in CMOR and unimodal object recognition.

• At retrieval, scopolamine selectively disrupts CMOR performance.

• Muscarinic receptors may help to bind multisensory object features.

Abstract

Acetylcholine (ACh) has been implicated in numerous cognitive functions, including multisensory feature binding. In the present study, we systematically assessed the involvement of cholinergic muscarinic receptors in several variations of an object recognition task for rats. In the standard spontaneous object recognition (SOR) task, tactile and visual properties of objects were freely available throughout the sample and choice phases. In the tactile- and visual-only unimodal SOR tasks, exploration in both phases was restricted to tactile and visual information, respectively. For the basic crossmodal object recognition (CMOR) task, sample object exploration was limited to tactile features, whereas choice objects were available only in the visual domain. In Experiment 1, pre-sample systemic administration of scopolamine (0.2 mg/kg) disrupted performance on standard SOR, both unimodal SOR tasks, and basic CMOR, consistent with a role for muscarinic receptors in memory encoding. Conversely, in Experiment 2, pre-choice systemic scopolamine selectively impaired object recognition on the CMOR task. For Experiment 3, the inclusion of multimodal, but not unimodal pre-exposure to the to-be-remembered objects prevented scopolamine from disrupting performance on the CMOR task when given prior to the choice phase. These results suggest that ACh is necessary during the choice phase of the CMOR task to facilitate the binding of object features across sensory modalities, a function that is not required for the other tasks assessed. Multimodal object pre-exposure might preclude the requisite contribution of ACh in the choice phase by allowing rats to bind important visual and tactile object information prior to testing.

Introduction

Despite acquiring sensory information through segregated systems, our brains integrate much of this material into unified, multisensory constructs. This binding of unimodal stimuli shapes our perception of the world and can dramatically impact our interaction with it (Freeman et al., 2013). Although there exists a great deal of interest in this process, the precise mechanisms by which we formulate cohesive, multisensory representations are still poorly understood.

We have developed a crossmodal object recognition (CMOR) task to permit detailed analysis of the underlying neurobiological processes responsible for the formation of multisensory object representations in the brain (Winters & Reid, 2010). With this task, we have shown that rats are capable of visually recognizing an object they have tactually explored one hour prior. Extending the retention delay to three hours necessitates a brief, multimodal pre-exposure session to facilitate recognition on this more mnemonically demanding variation of the CMOR task (PE/CMOR) (Reid, Jacklin, & Winters, 2012). We hypothesize that simultaneously experiencing the tactile and visual properties of the object help to bind the separate modalities together, establishing a multisensory object representation in the brain. Our previous work has focused on delineating the anatomical substrates mediating the integration of object features across modalities and comparing the differential contribution of critical structures in the two versions of this task. We have demonstrated that successful performance on the two CMOR tasks relies on distinct brain regions, suggesting that the memory enhancing effects of multimodal pre-exposure is likely a consequence of engaging different neural mechanisms to enable effective recognition of objects across sensory modalities (Reid, Jacklin, & Winters, 2013). To further characterize the neurobiology of multisensory integration, we have conducted a series of experiments in the current study to compare the role of the cholinergic system at different phases of the CMOR and PE/CMOR tasks.

Acetylcholine (ACh) has long been implicated in the acquisition and encoding of new memories (Hasselmo, 2006, Hasselmo and McGaughy, 2004). Consistent with this view, disrupting cholinergic signaling with non-selective antagonists of muscarinic receptors or permanent pre-training lesions of the cholinergic cell clusters in the basal forebrain is detrimental to certain mnemonic processes (Aigner et al., 1991, Berger-Sweeney et al., 2000, Ghoneim and Mewaldt, 1975, Hepler et al., 1985, Whishaw, 1985, Winters and Bussey, 2005), and specific temporal analyses indicate an essential role for acetylcholine in the encoding of new memories for long-term storage, but not for their subsequent retrieval (Aigner and Mishkin, 1986, Aigner et al., 1991, Ghoneim and Mewaldt, 1975, Hasselmo and McGaughy, 2004, Ogura and Aigner, 1993, Robbins et al., 1997, Tang et al., 1997, Warburton et al., 2003, Winters et al., 2006).

For the current series of experiments, we assessed the contribution of cholinergic muscarinic receptors to recognition tasks while systematically manipulating the perceived sensory modality of the objects during the sample (encoding) and choice (retrieval) phases. We predicted that, consistent with previous findings, blocking muscarinic receptors prior to the sample phase would impair performance on a standard spontaneous object recognition (SOR) task for which rats are able to utilize both tactile and visual information to recognize objects across a retention delay. Conversely, disrupting cholinergic signaling prior to the choice phase of the SOR task should not interfere with the recognition abilities of rats. Furthermore, isolating the tactile and visual modalities of the object properties into separate, unimodal components should not affect the nature of cholinergic involvement during the encoding and retrieval of an object representation. In our standard CMOR task, however, the animals’ experience with the tactile and visual properties of an object is separated by a delay, and proper recognition during the choice phase requires a comparison between the incoming visual information and a previously formed tactile representation. The rats never experience the two modalities together in this task and have therefore never been afforded the opportunity to bind the sensory stimuli into a cohesive, multisensory object representation. Because of this, we speculate that during the testing phase of the CMOR task, two mental processes must be engaged: recall of the tactile representation and a comparison of that information with the incoming visual properties of the object. While cholinergic signaling is not necessary for the recall of object memories (Winters et al., 2006), recent work in rats has indicated that ACh is essential for binding the separate features of stimuli both within and across modalities (Botly and De Rosa, 2007, Botly and De Rosa, 2008, Botly and De Rosa, 2009a, Botly and De Rosa, 2009b, Botly and De Rosa, 2012). We therefore predicted that administration of scopolamine prior to the choice phase would impair recognition on the standard CMOR task but would spare performance when rats are provided with multimodal pre-exposure to the to-be-remembered objects, a manipulation that presumably facilitates crossmodal feature binding and the formation of a multisensory object representation.