Abstract

In the present study, we assessed the involvement of the prefrontal cortex (PFC) in the ability of rats to perform crossmodal (tactile-to-visual) object recognition tasks. We tested rats with 3 different types of bilateral excitotoxic lesions: (1) Large PFC lesions, including the medial PFC (mPFC) and ventral and lateral regions of the orbitofrontal cortex (OFC); (2) selective mPFC lesions; and (3) selective OFC lesions. Rats were tested on 2 versions of crossmodal object recognition (CMOR): (1) The original CMOR task, which uses a tactile-only sample phase and a visual-only choice phase; and (2) a “multimodal pre-exposure” version (PE/CMOR), in which simultaneous pre-exposure to the tactile and visual features of an object facilitates CMOR performance over longer memory delays. Inclusive PFC lesions disrupted performance on both versions of CMOR, whereas selective mPFC damage had no effect. Lesions limited to the OFC caused delay-dependent deficits on the CMOR task, but failed to reverse the enhancement produced by multimodal object pre-exposure. This pattern of functional dissociations suggests complex, multidimensional contributions of the PFC and its subregions to crossmodal cognition.

Introduction

An important unresolved question in neuroscience concerns the mechanisms by which the brain stores and utilizes multisensory representations derived from the combination of unimodal sensory features of objects. The ability to use comprehensive multisensory object representations can clearly confer adaptive advantage by rendering environment- and object-appropriate behavior more efficient (Stein and Meredith 1993; Murray et al. 1998; Frassinetti et al. 2002). We have recently introduced a crossmodal object recognition (CMOR) paradigm to enable the systematic study of the neural bases of such processes in rodents (Winters and Reid 2010). In the CMOR task, rats are presented in a sample phase with only the tactile features of a to-be-remembered object. After a retention delay, 2 objects—a novel object and a copy of the sample object—are visually presented. Rats demonstrate crossmodal object recognition by preferentially exploring the novel object in the choice phase, much like in the common spontaneous object recognition (SOR) task for rodents (Ennaceur and Delacour 1988).

In an earlier study, we reported that 2 caudal cortical regions—the perirhinal (PRh) and posterior parietal (PPC) cortices—appear to interact in the service of crossmodal object recognition in rats (Winters and Reid 2010). This result is consistent with the notion that multisensory object representations rely on the contributions of a distributed network of brain regions, and we have subsequently addressed the question of what other areas might be involved. Specifically, we have posited that polymodal brain regions downstream of the PRh and PPC might be involved in aspects of crossmodal cognition required for performance on the CMOR task. One such brain area is the hippocampus, which receives myriad polysensory inputs (Amaral and Witter 1995; Lavenex and Amaral 2000) and has been previously implicated in various associative functions (Otto and Eichenbaum 1992; Rudy and Sutherland 1995; Vinogradova 2001; Manns et al. 2007). Our recent work, however, suggests that the hippocampus does not play an essential role in the tactile-to-visual CMOR task (Reid et al. 2012).

A second candidate region is the prefrontal cortex (PFC). Given the nature of executive functions attributed to the PFC (Robbins 1996; Dalley et al. 2004) and the wealth of bidirectional connections between PFC and higher-order sensory and association regions (Fuster 1997; Ongur and Price 2000; Uylings et al. 2003), it is possible that this area—or subregions within it—could play a valuable role in coordinating information processing between sensory modalities to aid in the formation and/or use of multisensory object representations. Indeed, the PFC has frequently been implicated in tasks requiring crossmodal cognition, particularly in human imaging studies (Banati et al. 2000; Laurienti et al. 2003; Weissman et al. 2004), but also in animal research with monkeys (Petrides and Iversen 1976; Aitken 1980; Gaffan and Harrison 1991; Tal and Amedi 2009) and rats (Whishaw et al. 1992; Lipton et al. 1999). In humans, anterior cingulate and surrounding dorsolateral prefrontal regions are commonly implicated in crossmodal cognition (Banati et al. 2000; Laurienti et al. 2003; Small 2004; Weissman et al. 2004; Ku et al. 2007). Consistent with such findings, anterior cingulate lesions have been shown to cause deficits in crossmodal matching in monkeys (Aitken 1980). Additionally, Fuster et al. (2000) demonstrated medial PFC (mPFC) neurons in monkeys that respond selectively to visual–auditory paired associations across time. Another prefrontal region implicated in crossmodal integration is the orbitofrontal cortex (OFC), which appears to contribute to crossmodal processing, especially when olfactory cues are involved (Lipton et al. 1999; Gottfried and Dolan 2003; Osterbauer et al. 2005; Small and Prescott 2005; Cerf-Ducastel and Murphy 2006). Evidence for an orbitofrontal role in crossmodal functions involving non-olfactory stimuli is less common (Diaconescu et al. 2011), but also not as frequently assessed and therefore remains an open question.

Thus, despite ample evidence for the involvement of prefrontal regions in crossmodal cognition, the specific contributions of the different areas remain unclear. The aim of the present study was to begin a systematic assessment of the contributions of rat PFC to crossmodal object recognition. For this initial study, we have chosen to focus on the mPFC, consisting of infralimbic, prelimbic, and rostral anterior cingulate cortices (Dalley et al. 2004), and the OFC. A few studies implicating the rat OFC in aspects of crossmodal information processing have highlighted the ventral and/or lateral OFC rather than the medial region (Whishaw et al. 1992; Lipton et al. 1999), and functional differences between the ventrolateral and medial OFC have previously been noted (Dalley et al. 2004); we therefore specifically targeted the ventrolateral OFC regions, also aiming to avoid inadvertent damage to the mPFC in our OFC groups.

We present data on the effects of inclusive bilateral PFC lesions, as well as separate mPFC and OFC lesions in the crossmodal and unimodal object recognition tasks from our previous study (Winters and Reid 2010). For comparison, the same rats were also tested in a novel variant of the CMOR task in which they are provided with a multimodal pre-exposure session prior to the tactile sample phase (Reid et al. 2012). In the original CMOR task, rats are limited to exploring the objects with only tactile or visual information in each behavioral phase; these separate sensory features of objects are never explored together. We have recently found, however, that simultaneous pre-exposure to both tactile and visual object information significantly affects the crossmodal recognition ability of rats, substantially increasing the length of the retention delay over which they can perform the CMOR task (Reid et al. 2012); such a mnemonic facilitation was not observed when the pre-exposure was unimodal (tactile- or visual-only) in nature. We have posited that the behavioral effect of multimodal object pre-exposure results from changes in the nature of the neural representations of the to-be-remembered objects, perhaps related to the formation and storage of an explicit multisensory object representation. It is therefore possible that different brain regions contribute to CMOR task performance with and without multimodal object pre-exposure. Accordingly, we also investigated the possible involvement of the PFC and its subregions in the behavioral facilitation produced by multimodal object pre-exposure. We predicted that PFC lesions would disrupt both aspects of crossmodal cognition. Specifically, we hypothesized that extensive PFC damage would disrupt processes related to the storage of multisensory object representations and comparison across sensory modalities. We posited that crossmodal comparison required for CMOR task performance might be specifically related to mPFC processing considering its common implication in multisensory tasks (Banati et al. 2000; Laurienti et al. 2003; Small 2004; Weissman et al. 2004; Ku et al. 2007) and its role in aspects of executive functions (Dalley et al. 2004). The OFC might be involved directly in associative multisensory representation within the PE/CMOR task, given its previous implication in predictive aspects of crossmodal cognition in rats (Lipton et al. 1999).

Materials and Methods

Subjects

Fifty-nine male Long-Evans rats were used in these experiments. The rats were kept on a 12-h on/12-h off reverse light/dark cycle with experiments carried out during the dark cycle. During experimentation, the rats were maintained on approximately 85% of their free-feeding body weight with 15–20 g of food/rat administered daily after trials. Rats were housed in pairs, and water was provided ad libitum. All procedures adhered to the guidelines of the Canadian Council on Animal Care and were approved by the Animal Care Committee at the University of Guelph.

Surgery

For each experiment, rats were randomly assigned either to the lesion group or to the control group. Before and during all surgeries, rats were deeply anaesthetized with isofluorane inhalation anaesthetic (Benson Medical Industries, Markham, ON, Canada). Animals also received a subcutaneous injection of the analgesic meloxicam (5 mg/mL; Boehringer Ingelheim, Burlington, ON, Canada) prior to surgery. They were then placed in a stereotaxic frame (Kopf Instruments, Tujunga, CA, USA) with the incisor bar set to –3.3 mm. The scalp was cut and retracted to expose the skull, and holes were drilled directly above the target region. Lesion groups received bilateral excitotoxic lesions, with infusions of 0.9 M N-methyl-d-aspartic acid (NMDA) dissolved in sterile phosphate buffer; and sham groups received identical infusions of phosphate buffer vehicle only.

Inclusive Prefrontal Cortex Lesions

Nine rats received bilateral combined lesions to the mPFC (including anterior cingulate, infralimbic, and prelimbic cortices) and ventrolateral OFC. Coordinates were based on a combination of Pezze et al. (2009) for the mPFC lesion and Schiller and Weiner (2004) for the OFC lesion. For the mPFC lesion, 6 holes were drilled at the following locations relative to bregma: Anteroposterior (AP), +3.8, +3.2, and +2.6 mm, with all mediolateral (ML) coordinates at ±0.7 mm. Injections of NMDA were made at each of the 6 sites, administered 1.5 mm below dura, and a second injection was made 3.0 mm below dura at the AP +3.2 site in each hemisphere. For the OFC lesion, 2 holes were drilled at the following locations relative to bregma: AP, +3.2 mm; ML, ±2.4 mm, and injections were made at the 2 sites administered 5.5 mm below the skull at the site of injection. All injections were 0.2 µL and done over a period of 2 min with the needle left in place for an additional 2 min to allow the toxin to diffuse before retraction. One animal in the lesion group died following surgery, leaving 8 rats in the lesion group and 11 rats in the sham group.

Orbitofrontal Cortex Lesions

The lesion group (n = 10) received bilateral ventrolateral OFC lesions as described above, whereas the sham group (n = 10) received similar injections only with sterile phosphate buffer.

Medial Prefrontal Cortex Lesions

The lesion group (n = 10) received bilateral lesions to the mPFC, including anterior cingulate, infralimbic, and prelimbic cortices using the parameters described above. The sham group (n = 10) received equal volume injections of sterile phosphate buffer to the same locations.

For all surgeries, after the infusions were complete, the incision was sutured, and rats recovered in cages on heat pads until they were alert and ready to be returned to their home cages. After a week of recovery, habituation sessions began.

Histology

Following behavioral testing, all rats were anaesthetized by intraperitoneal injections of 2 mL of Euthansol (340 mg/mL; Schering Canada, Inc., Quebec, Canada) and perfused transcardially with 100 mL of phosphate-buffered saline (PBS; pH 7.4), followed by 250 mL of 4% neutral buffered paraformaldehyde (PFA; pH 7.4). Brains were removed and post fixed in 4% PFA and, following at least a 24-h delay, immersed in a 20% sucrose PBS solution until they sank. Coronal sections (60 μm) were sliced through the PFC using a cryostat, and every fifth section was mounted on a gelatin-coated glass slide. Sections were stained with cresyl violet and examined under a light microscope to assess the excitotoxic lesions. The extent of the damage was quantified with pixel volume measurements from drawings of the lesion within the PFC as specified by Paxinos and Watson (1998) compared with the pixel volume of the entire target region. Pixel volume measurements were made using ImageJ software (http://rsbweb.nih.gov/ij/). The percentage of lesioned area for each rat and the mean lesioned area for the lesion group were calculated with these values.

Apparatus

Object recognition was assessed in a Y-shaped apparatus (Winters et al. 2004; Forwood et al. 2005), which consisted of 3 identically constructed white plastic arms joined by a central triangular area. Each arm was 27 cm long, 10 cm wide, and 50 cm tall. One arm, the start arm, had a guillotine door placed 18 cm from the rear of the arm to allow the experimenter to control the entry of rats into the main exploratory area. The 2 choice arms in the exploratory area of the Y-shaped apparatus had 2 notches at 9 cm from the rear of the arm, where plastic barriers could be inserted to prevent physical access to the objects. A tripod-mounted JVC Everio digital camera was placed over the Y-shaped apparatus to record behavior.

Stimuli

Multiple object sets were used during the experiments. Object sets consisted of triplicates of 2 objects of approximately the same height and width, but which were otherwise visually and tactually distinct (see Fig. 1A for typical objects). All objects consisted of hard plastic, glass, ceramics, or a combination thereof. To control for odor cues, objects were wiped with a 50% ethanol solution before being placed at the end of the choice arms in each phase of each trial. A small amount of white tack was placed on the bottom of each object to affix it to the floor of the Y-shaped apparatus. Rats were never exposed to the same object set for more than one trial.

Figure 1.

The 4 versions of SOR used in this study. (A) Examples of typical object sets used for all versions. (B) Tactile-only SOR: The sample and choice phases are run in red light to prevent visual exploration of objects. (C) Visual-only SOR: The sample and choice phases are run in white light with transparent barriers separating the rat and the objects to prevent tactile exploration. (D) The original CMOR task: The sample phase is run in red light, and the choice phase is run in white light with the transparent barriers in place. Rats are therefore required to use the visual information available during the choice phase to recognize an object based on tactile information previously acquired in the sample phase. (E) The modified CMOR task with multimodal object pre-exposure (PE/CMOR): Format is the same as CMOR, except that rats can explore the eventual sample objects in white light and without the barriers in a pre-exposure session 24 h before the tactile sample phase; in the control (no pre-exposure) condition, rats remain in their home cages for this session. The multimodal pre-exposure extends CMOR task performance beyond the standard 1-h retention delay; we typically run PE/CMOR with a 3-h retention delay to compare performance in rats with and without pre-exposure.

Figure 1.

The 4 versions of SOR used in this study. (A) Examples of typical object sets used for all versions. (B) Tactile-only SOR: The sample and choice phases are run in red light to prevent visual exploration of objects. (C) Visual-only SOR: The sample and choice phases are run in white light with transparent barriers separating the rat and the objects to prevent tactile exploration. (D) The original CMOR task: The sample phase is run in red light, and the choice phase is run in white light with the transparent barriers in place. Rats are therefore required to use the visual information available during the choice phase to recognize an object based on tactile information previously acquired in the sample phase. (E) The modified CMOR task with multimodal object pre-exposure (PE/CMOR): Format is the same as CMOR, except that rats can explore the eventual sample objects in white light and without the barriers in a pre-exposure session 24 h before the tactile sample phase; in the control (no pre-exposure) condition, rats remain in their home cages for this session. The multimodal pre-exposure extends CMOR task performance beyond the standard 1-h retention delay; we typically run PE/CMOR with a 3-h retention delay to compare performance in rats with and without pre-exposure.

Behavioral Procedure

Habituation

Rats were habituated to the Y-shaped apparatus over 2 days prior to the beginning of testing. On both days, each rat spent 2 separate 5-min sessions in the Y-shaped apparatus, one session in red light and another in white light with the transparent plastic barriers in place; the order of exposure to these conditions was counterbalanced on both days. No objects were placed in the apparatus during habituation sessions.

Testing

SOR tasks exploit the innate tendency of rats to explore novel versus familiar objects preferentially. The CMOR task is an SOR task variation, which requires animals to compare visual object features with tactile features to recognize objects. For the basic CMOR experiments described here, all rats were run on the CMOR task, as well as unimodal (tactile- and visual-only) versions of SOR. The unimodal tasks are included to gauge the selectivity of any crossmodal impairments observed. See Figure 1 for schematic illustrations of the 4 SOR tasks used in the current study. All tasks were run with the same general format, each trial consisting of sample and choice phases separated by a retention delay (“immediate”, 1 h, or 3 h; see details below). During the sample phase, 2 identical objects were affixed at the rear of the choice arms. After the rat was placed into the apparatus, the guillotine door was raised, and object exploration behavior was scored from the moment the rat left the start box. Object exploration was recorded using a custom computer application that tracks the individual and total object exploration times, as well as the time spent in the apparatus. During the sample phase, rats were allowed to explore both objects for a total of 25 s of cumulative exploration or 3 min, if 25 s of object exploration was not accumulated. Object exploration was scored when the nose was directed toward a part of the object, or the area of the transparent barrier within the outline of the object (for visual-only conditions), at a distance of no more than 2 cm. At the end of the sample phase, the retention delay began. For the 1-h or 3-h delays, rats were returned to their home cages until the choice phase. For the “immediate” delay, rats were placed directly back into the start box of the Y-shaped apparatus, while the experimenter set up the choice phase. Thus, the immediate delay condition yielded a much shorter retention delay, the average of which was approximately 30–45 s. During the choice phase, an identical version of the object from the sample phase was presented along with a novel object. The side in which the novel object was placed was randomized. During the choice phase, object exploration was scored for 1 min.

Specific procedural manipulations were used for the various versions of object recognition. For the tactile-only SOR task (Fig. 1B), to prevent the rats from acquiring visual information about the objects, both the sample and choice phases were run in red light (Winters and Reid 2010). Specifically, during the tactile-only phases, a red-tinted lamp mounted on the video tripod was used with the overhead lights off. In the visual-only SOR task (Fig. 1C), to prevent the rats from acquiring tactile information, the objects were placed behind a transparent plastic barrier inserted at the 9-cm notch in each of the 2 choice arms of the Y-shaped apparatus for both the sample and choice phases. For the CMOR task (Fig. 1D), the sample phase was run in red light, and the choice phase was run in normal white light with the objects behind the transparent plastic barriers. Each group was tested in each of these 3 tasks with a 1-h retention delay in counterbalanced order. The OFC-lesioned rats and their sham groups were also tested in the CMOR task with the immediate retention delay after being tested on the PE/CMOR task (see below for further details).

To assess the role of PFC regions in the facilitative effects of multimodal object pre-exposure, we used a variant of the CMOR task in which rats were simultaneously pre-exposed to the tactile and visual features of the to-be-remembered objects 24 h prior to the tactile sample phase (Reid et al. 2012). For this multimodal pre-exposure CMOR (PE/CMOR) task (Fig. 1E), rats were tested in 2 conditions, counterbalanced. On one trial, rats received a 5-s multimodal pre-exposure to the sample object 24 h prior to the tactile sample phase, and on the other trial, rats remained in their home cages until the sample phase (“no pre-exposure” condition). Pre-exposure was run in normal white lighting without the barrier, so that the objects could be explored with tactile and visual features available simultaneously, and rats were allowed to explore until cumulative object exploration time reached 5 s. Twenty-four hours following the pre-exposure session, the typical CMOR trial proceeded as described above, but with a 3-h retention delay between the tactile sample and visual choice phases. We have previously shown that a multimodal—but not unimodal (tactile- or visual-only)—pre-exposure session reverses the memory impairment observed when CMOR is tested with a 3-h retention delay (Reid et al. 2012). All rats were tested in the PE/CMOR task after being tested on the original CMOR and unimodal SOR tasks.

Data Analysis

The discrimination ratio was used as the primary measure of object recognition and was calculated as: (Novel object exploration − familiar object exploration)/(novel object exploration + familiar object exploration) in the choice phase. For initial testing of each lesion type, 2 (group) × 3 (task: CMOR, tactile SOR, and visual SOR) mixed-factors analyses of variance (ANOVA) with repeated measures were performed to assess object recognition ability. For PE/CMOR testing, the discrimination ratio data were analyzed using 2 (group) × 2 (pre-exposure condition) ANOVAs with repeated measures. As an additional assessment of recognition performance, we used 1-sample t-tests to compare each group mean in each task to a discrimination ratio of 0, which indicates no novel or familiar object preference. Three additional behavioral measures were also scored for all tasks to ensure that general exploration of objects did not differ between groups: The total choice object exploration, total sample object exploration, and sample phase duration (i.e., the time required for rats to reach the 25-s sample exploration criterion). An additional control measure, duration of the pre-exposure session, was also analyzed for the PE/CMOR experiments. All control measures were analyzed using mixed-factors ANOVAs. All statistical analysis was conducted using SPSS 18.0 for windows with a significance level of α = 0.05.

As a general note, preventing rats from touching the objects resulted in significant reductions in overall object exploration times in the visual-only conditions, particularly in the choice phase (see Tables 1–3 for general exploratory data). Accordingly, relatively small differences in novel and familiar exploration times could produce strong object recognition (i.e., discrimination ratio) scores. Thus, in order to preclude the possibility that experimenter bias or error might have substantially influenced the data, 2 experimenters scored all of the choice phase data offline for all experiments. Both scorers were blind to the lesion group assignment for each subject, as well as the identity of the novel and familiar object for each individual trial. The inter-rater reliability for choice phase scoring was highly significant (Pearson r = 0.58–0.88, all P < 0.05) for all task types in all experiments. Thus, it is unlikely that observer bias influenced the results reported in this study.

Table 1

General exploratory data from inclusive PFC lesion experiments

Group Task/delay Pre-exposure phase duration Total sample exploration Sample phase duration Choice exploration 
PFC lesion CMOR/1 h — 22.97 ± 1.52 126.97 ± 11.4 2.46 ± 0.25 
Tactile/1 h — 21.74 ± 0.01 133.49 ± 13 12.43 ± 1.76 
Visual/1 h — 5.72 ± 0.45 180 ± 0 2.31 ± 0.44 
PE/CMOR/3 h 18.32 ± 2.13 21.05 ± 1.27 148.69 ± 11.82 2.02 ± 0.51 
CMOR/3 h — 23.05 ± 1.11 150.98 ± 10.79 2.18 ± 0.32 
Sham CMOR/1 h — 22.66 ± 1.12 120.71 ± 14.43 3.76 ± 0.45 
Tactile/1 h — 23.31 ± 0.71 125.83 ± 11.88 18.43 ± 1.31 
Visual/1 h — 5.3 ± 0.43 180 ± 0 3.21 ± 0.56 
PE/CMOR/3 h 19.67 ± 3.7 21.91 ± 0.91 120.16 ± 8.15 4.09 ± 0.63 
CMOR/3 h — 23.01 ± 0.56 168.05 ± 5.25 3.9 ± 0.43 
Group Task/delay Pre-exposure phase duration Total sample exploration Sample phase duration Choice exploration 
PFC lesion CMOR/1 h — 22.97 ± 1.52 126.97 ± 11.4 2.46 ± 0.25 
Tactile/1 h — 21.74 ± 0.01 133.49 ± 13 12.43 ± 1.76 
Visual/1 h — 5.72 ± 0.45 180 ± 0 2.31 ± 0.44 
PE/CMOR/3 h 18.32 ± 2.13 21.05 ± 1.27 148.69 ± 11.82 2.02 ± 0.51 
CMOR/3 h — 23.05 ± 1.11 150.98 ± 10.79 2.18 ± 0.32 
Sham CMOR/1 h — 22.66 ± 1.12 120.71 ± 14.43 3.76 ± 0.45 
Tactile/1 h — 23.31 ± 0.71 125.83 ± 11.88 18.43 ± 1.31 
Visual/1 h — 5.3 ± 0.43 180 ± 0 3.21 ± 0.56 
PE/CMOR/3 h 19.67 ± 3.7 21.91 ± 0.91 120.16 ± 8.15 4.09 ± 0.63 
CMOR/3 h — 23.01 ± 0.56 168.05 ± 5.25 3.9 ± 0.43 

Note: Data are expressed as mean (seconds ± SEM) of the total time spent in the pre-exposure phase (pre-exposure phase duration), the total time exploring the sample objects directly (total sample Exploration), the total time spent exploring the objects in the first minute of the choice phase (choice exploration), and the total time spent in the sample phase (sample phase duration).

CMOR: standard crossmodal object recognition task with no pre-exposure; PE/CMOR: CMOR task with multimodal pre-exposure 24-h prior to the sample phase; tactile: tactile-only SOR; visual: visual-only SOR; 1 h: 1-h retention delay; 3 h: 3-h retention delay.

Table 2

General exploratory data from mPFC lesion experiments

Group Task/delay Pre-exposure phase duration Total sample exploration Sample phase duration Choice exploration 
mPFC lesion CMOR/1 h — 25 ± 0 60.97 ± 8.42 5.29 ± 0.27 
Tactile/1 h — 25 ± 0 57.09 ± 4.64 13.06 ± 1.75 
Visual/1 h — 17.34 ± 1.27 166.54 ± 9.43 5.84 ± 1.08 
PE/CMOR/3 h 12.52 ± 1.31 23.85 ± 0.44 132.47 ± 9.54 3.69 ± 0.41 
CMOR/3 h — 23.75 ± 0.79 107.26 ± 10.29 3.17 ± 0.27 
Sham CMOR/1 h — 25 ± 0 65.68 ± 10.72 7.52 ± 0.68 
Tactile/1 h — 25 ± 0 62.76 ± 4 18.09 ± 1.86 
Visual/1 h — 19.15 ± 1.06 180 ± 0 5.18 ± 0.69 
PE/CMOR/3 h 12.58 ± 1.91 23.96 ± 0.61 107.65 ± 10.75 4.88 ± 0.28 
CMOR/3 h — 24.68 ± 0.33 94.83 ± 8.57 4.56 ± 0.23 
Group Task/delay Pre-exposure phase duration Total sample exploration Sample phase duration Choice exploration 
mPFC lesion CMOR/1 h — 25 ± 0 60.97 ± 8.42 5.29 ± 0.27 
Tactile/1 h — 25 ± 0 57.09 ± 4.64 13.06 ± 1.75 
Visual/1 h — 17.34 ± 1.27 166.54 ± 9.43 5.84 ± 1.08 
PE/CMOR/3 h 12.52 ± 1.31 23.85 ± 0.44 132.47 ± 9.54 3.69 ± 0.41 
CMOR/3 h — 23.75 ± 0.79 107.26 ± 10.29 3.17 ± 0.27 
Sham CMOR/1 h — 25 ± 0 65.68 ± 10.72 7.52 ± 0.68 
Tactile/1 h — 25 ± 0 62.76 ± 4 18.09 ± 1.86 
Visual/1 h — 19.15 ± 1.06 180 ± 0 5.18 ± 0.69 
PE/CMOR/3 h 12.58 ± 1.91 23.96 ± 0.61 107.65 ± 10.75 4.88 ± 0.28 
CMOR/3 h — 24.68 ± 0.33 94.83 ± 8.57 4.56 ± 0.23 

Note: Data are expressed as mean (seconds ± SEM) of the total time spent in the pre-exposure phase (pre-exposure phase duration), the total time exploring the sample objects directly (total sample exploration), the total time spent exploring the objects in the first minute of the choice phase (choice exploration), and the total time spent in the sample phase (sample phase duration).

CMOR: standard crossmodal object recognition task with no pre-exposure; PE/CMOR: CMOR task with multimodal pre-exposure 24-h prior to the sample phase; tactile: tactile-only SOR; visual: visual-only SOR; 1 h: 1-h retention delay; 3 h: 3-h retention delay.

Table 3

General exploratory data from OFC lesion experiments

Group Task/delay Pre-exposure phase duration Total sample exploration Sample phase duration Choice exploration 
OFC lesion CMOR/1 h — 23.55 ± 0.66 119.8 ± 13.06 3.45 ± 0.4 
Tactile/1 h — 23.66 ± 0.92 120.93 ± 6.62 12.4 ± 1.2 
Visual/1 h — 7.05 ± 0.59 180 ± 0 2.85 ± 0.31 
PE/CMOR/3 h 15.17 ± 3.25 25 ± 0 76.65 ± 11.82 4.91 ± 0.63 
CMOR/3 h — 24.71 ± 0.34 119.73 ± 12.49 3.05 ± 0.32 
CMOR/imm — 23.56 ± 0.66 119.8 ± 13.06 3.96 ± 0.28 
CMOR/1 hR — 24.58 ± 0.49 103.94 ± 11.37 4.38 ± 0.33 
Sham CMOR/1 h — 23.68 ± 0.63 111.71 ± 10.07 4.11 ± 0.59 
Tactile/1 h — 24.41 ± 0.4 114.88 ± 10.13 16.29 ± 1.8 
Visual/1 h — 7.58 ± 0.88 180 ± 0 3.15 ± 0.35 
PE/CMOR/3 h 20.93 ± 5.1 25 ± 0 65.77 ± 8.15 4.8 ± 0.6 
CMOR/3 h — 24.51 ± 0.38 114.22 ± 8.75 2.53 ± 0.28 
CMOR/imm — 23.68 ± 0.63 111.76 ± 11.06 4.21 ± 0.63 
CMOR/1 hR — 25 ± 0 78.12 ± 7.35 4.41 ± 0.36 
Group Task/delay Pre-exposure phase duration Total sample exploration Sample phase duration Choice exploration 
OFC lesion CMOR/1 h — 23.55 ± 0.66 119.8 ± 13.06 3.45 ± 0.4 
Tactile/1 h — 23.66 ± 0.92 120.93 ± 6.62 12.4 ± 1.2 
Visual/1 h — 7.05 ± 0.59 180 ± 0 2.85 ± 0.31 
PE/CMOR/3 h 15.17 ± 3.25 25 ± 0 76.65 ± 11.82 4.91 ± 0.63 
CMOR/3 h — 24.71 ± 0.34 119.73 ± 12.49 3.05 ± 0.32 
CMOR/imm — 23.56 ± 0.66 119.8 ± 13.06 3.96 ± 0.28 
CMOR/1 hR — 24.58 ± 0.49 103.94 ± 11.37 4.38 ± 0.33 
Sham CMOR/1 h — 23.68 ± 0.63 111.71 ± 10.07 4.11 ± 0.59 
Tactile/1 h — 24.41 ± 0.4 114.88 ± 10.13 16.29 ± 1.8 
Visual/1 h — 7.58 ± 0.88 180 ± 0 3.15 ± 0.35 
PE/CMOR/3 h 20.93 ± 5.1 25 ± 0 65.77 ± 8.15 4.8 ± 0.6 
CMOR/3 h — 24.51 ± 0.38 114.22 ± 8.75 2.53 ± 0.28 
CMOR/imm — 23.68 ± 0.63 111.76 ± 11.06 4.21 ± 0.63 
CMOR/1 hR — 25 ± 0 78.12 ± 7.35 4.41 ± 0.36 

Note: Data are expressed as mean (seconds ± SEM) of the total time spent in the pre-exposure phase (pre-exposure phase duration), the total time exploring the sample objects directly (total sample exploration), the total time spent exploring the objects in the first minute of the choice phase (choice exploration), and the total time spent in the sample phase (sample phase duration).

CMOR: standard crossmodal object recognition task with no pre-exposure; PE/CMOR: CMOR task with multimodal pre-exposure 24-h prior to the sample phase; tactile: tactile-only SOR; visual: visual-only SOR; 1 h: 1-h retention delay; 3 h: 3-h retention delay; imm: immediate retention delay; 1 hR: repeat of 1-h CMOR task.

Results

Inclusive PFC Lesions

Histology

Excitotoxin-induced brain damage was primarily restricted to the PFC, specifically the defined mPFC and ventrolateral OFC regions. The estimated average percent of total PFC loss was 61% (±5.41%); estimated damage within the 2 subregions was 49% (±7.1%) and 62% (±8.1%) for ventrolateral OFC and mPFC, respectively. Figure 2 illustrates the extent of the largest and smallest PFC lesions through several sections of the brain. In some cases, the lesions extended into other regions, including the primary motor cortex, secondary motor cortex, and claustrum, but this unintended damage was always unilateral.

Figure 2.

Extent of the inclusive excitotoxic prefrontal cortex lesions. The largest lesion is illustrated in gray, and the smallest lesion is in black. Anatomical figures are from Paxinos and Watson (1998). Coronal sections are 4.20 mm through 2.20 mm anterior to bregma.

Figure 2.

Extent of the inclusive excitotoxic prefrontal cortex lesions. The largest lesion is illustrated in gray, and the smallest lesion is in black. Anatomical figures are from Paxinos and Watson (1998). Coronal sections are 4.20 mm through 2.20 mm anterior to bregma.

Behavior

Rats were first tested on counterbalanced trials of the CMOR and unimodal (tactile- and visual-only) SOR tasks. An analysis of the discrimination ratio data with mixed-factors ANOVA revealed significant group (F1,15 = 6.32, P = 0.02) and task (F1,22 = 8.53, P = 0.004) effects; the interaction term was not significant (F1,22 = 1.11, P = 0.33). Inspection of Figure 3A, however, suggests that the predicted difference in the CMOR task was present, and planned comparison independent-samples t-tests supported this interpretation, as CMOR was the only task in which there was a significant group difference (CMOR: t17 = 2.39, P = 0.03; tactile SOR: t < 1; and visual SOR: t < 1). The analysis of the discrimination ratios compared with chance preference (0) with 1-sample t-tests for the lesion group indicated significant novel object preference in the tactile (t6 = 10.34, P < 0.001) and visual (t6 = 3.59, P = 0.02) SOR tasks, but not the CMOR task (t6 = 2.37, P = 0.06). Conversely, sham animals displayed significant novel object preference in all tasks (CMOR: t10 = 5.48, P < 0.001; tactile: t10 = 8.64, P < 0.001; and visual: t10 = 8.01, P < 0.001).

Figure 3.

Performance of rats with inclusive PFC lesions on (A) CMOR, as well as tactile-only and visual-only SOR, each with a 1-h retention delay between sample and choice phases and (B) 3-h CMOR with and without multimodal object pre-exposures 24 h prior to the tactile sample phase. Data are presented as average discrimination ratio (±SEM). *P < 0.05; **P < 0.01, lesion versus sham.

Figure 3.

Performance of rats with inclusive PFC lesions on (A) CMOR, as well as tactile-only and visual-only SOR, each with a 1-h retention delay between sample and choice phases and (B) 3-h CMOR with and without multimodal object pre-exposures 24 h prior to the tactile sample phase. Data are presented as average discrimination ratio (±SEM). *P < 0.05; **P < 0.01, lesion versus sham.

The analysis of the choice exploration control measure with mixed-factors ANOVA yielded an unexpected significant group × task interaction (F1,34 = 5.42, P = 0.02). Post hoc analysis indicated that there was an impact of lesion in the crossmodal (t17 = 2.26, P = 0.04) and tactile tasks (t17 = 2.80, P = 0.01), but not the visual (t17 = 1.21, P = 0.24) with lesion animals exploring less than shams in the choice phase. There was also a significant main effect of lesion (F1,17 = 8.91, P = 0.01) for choice object exploration. Despite these effects of PFC lesions on general exploration in the choice phase, both groups did explore within typically accepted values. Furthermore, and importantly, there were no significant differences in sample exploration measures, which suggest that PFC lesions did not affect exploration of the to-be-remembered objects during acquisition. There were also significant task effects for all 3 general exploration measures due to the different stimulus properties available to rats in the different task phases; rats explored significantly more in tactile conditions than in visual conditions in which physical access to the objects was restricted (Winters and Reid 2010). None of the other control analyses yielded significant effects (see Table 1 for descriptive statistics).

The same 2 groups of rats were next tested on 2 additional trials of CMOR with a 3-h retention delay: One trial with multimodal pre-exposure and another trial without pre-exposure. We have previously shown that rats fail on the original CMOR task with a 3-h retention delay (Reid et al. 2012); the no pre-exposure trials were therefore included as a control condition.

The analysis of the discrimination ratio with a mixed-factor ANOVA revealed a significant main effect of the lesion group (F1,16 = 8.34, P = 0.01), but the effects of pre-exposure condition (F1,16 = 1.17, P = 0.295) and lesion × pre-exposure condition interaction (F < 1) were not significant. However, Figure 3B suggests that the predicted impairment on PE/CMOR was present, and a planned comparison of the lesion group revealed a significant difference between shams and lesioned animals on pre-exposure trials (t16 = 2.85, P = 0.01), but not on the no pre-exposure trials (t16 = 1.25, P = 0.23). Additional analysis of the discrimination ratio data indicated that only the sham group with pre-exposure demonstrated significant novel object preference (lesion, no pre-exposure: t6 = 0.67, P = 0.54; lesion, pre-exposure: t6 = 0.11, P = 0.92; sham, no pre-exposure: t10 = 1.28, P = 0.23; sham, pre-exposure: t10 = 5.20, P < 0.001), further supporting the assertion that PFC lesions abolished the facilitative effect of multimodal object pre-exposure. Because neither separate mPFC nor OFC lesions disrupted PE/CMOR performance (see below), we assessed the possibility that the absolute quantity of PFC tissue damage was related to the PE/CMOR deficit by performing a correlation analysis between the percentage of PFC tissue lost in the lesion group and the discrimination ratio on PE/CMOR trials (with pre-exposure only). This analysis, however, revealed no significant correlation (Pearson r = −0.114).

The analysis of sample duration with mixed-factor ANOVA revealed an unexpected significant group × pre-exposure condition interaction (F1,16 = 8.76, P = 0.01) and a significant main effect of pre-exposure condition (F1,16 = 7.23, P = 0.02). Post hoc analysis on sample duration suggested that sham rats on pre-exposure trials were quicker to reach criteria during the sample phase; however, importantly, there was not a significant difference in total sample exploration time. Also, as in the previous CMOR experiment, there was a significant main effect of lesion on choice exploration (F1,16 = 10.33, P = 0.01), with lesioned animals spending less time exploring objects during the choice phase. However, choice exploration was within expected ranges, and overall sample exploration was not altered, suggesting that exploration of the object to be remembered was not seriously impacted (see Table 1 for descriptive statistics).

mPFC Lesions

Histology

Excitotoxin-induced brain damage was generally restricted to the mPFC. The estimated average mPFC loss was 68% (±3.29%). Figure 4 illustrates the extent of the largest and smallest mPFC lesions through several sections of the brain. In a few cases, the lesions extended into other regions, including the primary motor cortex, secondary motor cortex, and corpus callosum (forceps minor), but this damage was always unilateral.

Figure 4.

Extent of the excitotoxic medial prefrontal cortex lesions. The largest lesion is displayed in gray, and the smallest lesion is shown in black. Coronal sections are 4.20 mm through 2.20 mm anterior to bregma.

Figure 4.

Extent of the excitotoxic medial prefrontal cortex lesions. The largest lesion is displayed in gray, and the smallest lesion is shown in black. Coronal sections are 4.20 mm through 2.20 mm anterior to bregma.

Behavior

Rats were first tested on counterbalanced trials of the CMOR and unimodal (tactile- and visual-only) SOR tasks with a 1-h retention delay. Lesions of the mPFC did not appear to disrupt CMOR or unimodal SOR task performance (Fig. 5A). A mixed-factors ANOVA performed on the discrimination ratio did not find a significant lesion group × task interaction (F1,35 = 1.20, P = 0.31), nor were the main effects of group (F < 1) or task (F1,35 = 1.59) significant. One-sample t-tests against chance (0) indicated that both groups demonstrated significant novel object preference in all tasks (lesion, CMOR: t9 = 5.02, P = 0.001; lesion, tactile: t9 = 5.51, P < 0.001; lesion, visual: t9 = 3.64, P = 0.005; sham, CMOR: t9 = 3.29, P = 0.009; sham, tactile: t9 = 4.13, P = 0.003; sham, visual: t9 = 5.25, P = 0.001).

Figure 5.

Performance of rats with mPFC lesions on (A) CMOR, tactile-only, and visual-only SOR, each with a 1-h retention delay between sample and choice phases and (B) 3-h CMOR with and without multimodal object pre-exposures 24 h prior to the tactile sample phase. Data are presented as the average discrimination ratio (±SEM). Dashed line (0.0) represents equal preference for sample and novel objects.

Figure 5.

Performance of rats with mPFC lesions on (A) CMOR, tactile-only, and visual-only SOR, each with a 1-h retention delay between sample and choice phases and (B) 3-h CMOR with and without multimodal object pre-exposures 24 h prior to the tactile sample phase. Data are presented as the average discrimination ratio (±SEM). Dashed line (0.0) represents equal preference for sample and novel objects.

Mixed-factors ANOVA of the 3 general exploration measures found only anticipated significant difference between tasks (see Table 2 for descriptive statistics).

The same rats were next tested on 2 unique trials of CMOR with a 3-h retention delay: One trial with pre-exposure (PE/CMOR) and one without pre-exposure, counterbalanced. Mixed-factors ANOVA was performed on the discrimination ratio, revealing a significant main effect of pre-exposure condition (F1,18 = 18.13, P < 0.001) with pre-exposure trials having higher discrimination ratios compared with trials without pre-exposure (Fig. 5B); however, there was not a significant lesion group × pre-exposure condition interaction (F < 1) or main effect of lesion (F1,18 = 1.16, P = 0.30). Analysis of the discrimination ratios with 1-sample t-tests versus 0 found that only the sham and lesion groups with pre-exposure demonstrated significant novel object preference (lesion, no pre-exposure: t9 = 1.06, P = 0.32; lesion, pre-exposure: t9 = 3.99, P = 0.003; sham, no pre-exposure: t9 = 0.70, P = 0.50; sham, pre-exposure: t9 = 2.96, P = 0.02).

There was an unexpected significant main effect of pre-exposure condition for the sample duration (F1,18 = 6.23, P = 0.02), with rats on pre-exposure trials being slower to reach criterion than on trials without pre-exposure, but there was no difference in terms of overall sample object exploration. Additionally, there was a significant main effect of lesion on choice object exploration (F1,18 = 12.44, P = 0.002), with lesion animals exploring less than sham animals in the choice phase (see Table 2 for descriptive statistics).

OFC Lesions

Histology

Excitotoxin-induced brain damage was generally restricted to the ventral and lateral regions of OFC. The estimated average percent of ventrolateral OFC loss was 60% (±2.93%). Figure 6 illustrates the extent of the largest and smallest OFC lesions through several sections of the brain. In a few cases, the lesions extended into other regions, including the corpus callosum (forceps minor), claustrum, agranular insular cortex (ventral), and striatum, but this unintended damage was always unilateral.

Figure 6.

Extent of the excitotoxic ventral and lateral orbitofrontal cortex lesions. The largest lesion is shown in gray and the smallest lesion in black. Coronal sections are 4.20 mm through 2.20 mm anterior to bregma.

Figure 6.

Extent of the excitotoxic ventral and lateral orbitofrontal cortex lesions. The largest lesion is shown in gray and the smallest lesion in black. Coronal sections are 4.20 mm through 2.20 mm anterior to bregma.

Behavior

Rats were first tested on counterbalanced trials of the CMOR and unimodal (tactile- and visual-only) SOR tasks with a 1-h retention delay. Analysis of the discrimination ratio data with mixed-factors ANOVA revealed group (F1,17 = 4.10, P = 0.06) and interaction terms (F1,24 = 3.20, P = 0.07) that approached significance, whereas the effect of task was clearly not significant (F1,24 = 2.48, P = 0.12). Examination of Figure 7A suggests that the predicted difference in the CMOR task was present, and planned comparison independent samples t-tests supported this interpretation, as CMOR was the only task in which there was a significant group difference (CMOR: t17 = 3.63, P = 0.002; tactile SOR: t18 = 0.68, P = 0.51; visual SOR: t18 = 0.51, P = 0.62). One-sample t-tests comparing the discrimination ratio scores to 0 also indicated that rats demonstrated significant novel object preference in all cases, except when lesioned animals were tested on the CMOR task (lesion, CMOR: t9 = 0.74, P = 0.48; lesion, tactile: t9 = 6.91, P < 0.001; lesion, visual: t9 = 3.90, P = 0.004; sham, CMOR: t9 = 8.30, P < 0.001; sham, tactile: t8 = 3.75, P = 0.005; sham, visual: t9 = 3.85, P = 0.004).

Figure 7.

Performance of rats with OFC lesions on (A) CMOR, tactile-only, and visual-only SOR, each with a 1-h retention delay between sample and choice phases; (B) 3-h CMOR with and without multimodal object pre-exposures 24 h prior to the tactile sample phase; (C) CMOR with the “immediate” retention delay and no pre-exposure; and (D) Retest with the 1-h delay CMOR task (no pre-exposure). Data are presented as the average discrimination ratio (±SEM). Dashed line (0.0) represents equal preference for sample and novel objects. **P < 0.01, lesion versus sham.

Figure 7.

Performance of rats with OFC lesions on (A) CMOR, tactile-only, and visual-only SOR, each with a 1-h retention delay between sample and choice phases; (B) 3-h CMOR with and without multimodal object pre-exposures 24 h prior to the tactile sample phase; (C) CMOR with the “immediate” retention delay and no pre-exposure; and (D) Retest with the 1-h delay CMOR task (no pre-exposure). Data are presented as the average discrimination ratio (±SEM). Dashed line (0.0) represents equal preference for sample and novel objects. **P < 0.01, lesion versus sham.

In analyses of the 3 general exploration measures, only the typical task effects were found in a mixed-factor ANOVA (see Table 3 for descriptive statistics).

The same rats were next tested on 2 unique trials of CMOR with a 3-h retention delay: One trial with pre-exposure (PE/CMOR) and one without pre-exposure, counterbalanced. The discrimination ratio scores were analyzed with mixed-factors ANOVA that revealed only a significant main effect of pre-exposure condition (F1,18 = 15.19, P < 0.001). The main effects of lesion (F < 1) and lesion group × pre-exposure condition interactions (F < 1) were not significant (Fig. 7B). Further analysis with 1-sample t-tests comparing discrimination ratios to 0 revealed that both the sham and lesion groups demonstrated significant novel object preference on pre-exposure trials, but not on trials without pre-exposure (lesion, no pre-exposure: t9 = 1.14, P = 0.28; lesion, pre-exposure: t9 = 3.89, P = 0.004; sham, no pre-exposure: t9 = 0.17, P = 0.87; sham, pre-exposure: t9 = 5.34, P < 0.001).

In terms of general exploration, there was an unexpected main effect of pre-exposure condition on choice exploration (F1,18 = 19.27, P < 0.001) and sample duration (F1,18 = 32.84, P < 0.001), with rats on pre-exposure trials being quicker to reach criterion and exploring more in the choice phase (see Table 3 for descriptive statistics).

We next tested the delay-dependence of the OFC lesion-induced impairment in the CMOR task. Our results with mPFC lesions did not support our initial hypothesis regarding a role for the mPFC in crossmodal object feature comparison. The selective CMOR deficit in OFC-lesioned rats, however, suggests that the ventrolateral OFC might contribute to this function. We therefore hypothesized that, if the ventrolateral OFC is a comparator for crossmodal object features, a function required for correct choice phase performance, CMOR task impairment should be observed in OFC-lesioned animals regardless of the length of the retention interval. Accordingly, the same groups of rats were each tested on one additional trial in the original CMOR task (i.e., without pre-exposure), but with the immediate delay condition to minimize mnemonic demand. Analysis of the discrimination ratio scores, however, with an independent-samples t-test revealed no significant difference between lesion and sham groups (t18 = 0.31, P = 0.76; Fig. 7C), and both groups demonstrated significant novel object preference (sham: t9 = 2.48, P = 0.035; lesion: t9 = 2.96, P = 0.016). Furthermore, no significant differences were found between the lesion and sham groups in terms of choice exploration, total sample exploration, or sample duration (Table 3).

Finally, to ensure that there was no recovery of function, the lesion and sham rats were again run on the CMOR task with a 1-h retention delay. An independent-samples t-test analysis of the discrimination ratio scores indicated that OFC-lesioned rats were still significantly impaired on the 1-h CMOR task compared with shams (t10 = 2.95, P = 0.01; Fig. 7D). No significant differences were found between the lesion and sham groups in terms of choice exploration, total sample exploration, or sample duration (Table 3).

Discussion

The neural bases of multisensory integration remain poorly understood. The CMOR paradigm has enabled us recently to embark on a systematic analysis of the brain circuitry involved in rodent tactile-to-visual crossmodal object recognition. The current study extends our previous findings (Winters and Reid 2010; Reid et al. 2012) by demonstrating that prefrontal circuitry is integrally involved in particular facets of CMOR performance. Specifically, the current results indicate that large PFC lesions, including the anterior cingulate, mPFC, and ventrolateral OFC, disrupt spontaneous CMOR task performance and reverse the facilitative mnemonic effect of multimodal object pre-exposure. More selective lesions of the ventrolateral OFC impaired the basic CMOR task in a delay-dependent manner, but spared the mnemonic enhancement produced by multimodal pre-exposure. Finally, selective mPFC lesions failed to affect performance in either version of CMOR. This pattern of findings provides important insights into the role of the PFC and its subareas in tactile-to-visual crossmodal object recognition.

The “inclusive” PFC lesions in the current study encompassed the ventrolateral OFC and a large proportion of the mPFC, including the insular, prelimbic, and rostral components of the anterior cingulate cortex. Homologous PFC regions in humans and nonhuman primates have previously been implicated in aspects of multisensory integration or crossmodal object recognition (Petrides and Iversen 1976; Aitken 1980; Gaffan and Harrison 1991; Banati et al. 2000; Laurienti et al. 2003; Weissman et al. 2004; Tal and Amedi 2009). Moreover, these areas are all polysensory in the sense that they are reciprocally connected with various association cortices (Fuster 1997; Ongur and Price 2000; Uylings et al. 2003). We therefore predicted that large PFC lesions would affect performance in both CMOR tasks used in the current study. Indeed, this is what we found. Specifically, inclusive PFC lesions produced a selective impairment in the original CMOR task when compared with the unimodal (tactile- and visual-only) SOR tasks. This type of selectivity has been seen previously following “disconnection” lesions of the PRh and PPC, and we have posited that these caudal cortical regions cooperate as part of a distributed network underlying CMOR task performance (Winters and Reid 2010). The PFC might represent another node in this putative network, interacting with the PRh and PPC to coordinate crossmodal object recognition; indeed, PFC regions are known to be connected with both the PRh and PPC (Reep et al. 1996; Ongur and Price 2000). Interestingly, bilateral PRh and PPC lesions disrupted visual and tactile SOR performance, respectively, in our previous study (Winters and Reid 2010). We took these results to indicate relatively unimodal contributions of these cortical regions to CMOR task performance. Conversely, the selective CMOR deficit produced by bilateral PFC lesions in the current study suggests contributions specific to the crossmodal nature of this task.

Given the established role of the PFC in higher-order and executive cognitive functions (Dalley et al. 2004), we initially hypothesized that mPFC lesions and, by extension, inclusive PFC damage would disrupt the comparison between tactile and visual representations in the CMOR task. The absence of effects with mPFC lesions, however, argues against such a role for this brain area. Conversely, ventrolateral OFC lesions selectively impaired CMOR task performance. The delay dependence of the CMOR deficit induced by OFC lesions, however, argues against a simple crossmodal comparison function. Were the OFC merely coordinating crossmodal evaluation to facilitate a match-mismatch decision, there would be no obvious reason to expect this function to be limited to more mnemonically taxing conditions.

The OFC may contribute to the CMOR task by performing top-down object priming during the sample phase. Evidence from humans indicates that the OFC may be involved in top-down facilitation of visual recognition (Fenske et al. 2006). Bar et al. (2006) have suggested that partially complete object information is sent from early visual stages to the PFC, specifically the OFC, which then generates expectations about the identity of the image and projects these “initial guesses” back to the temporal lobe; such a mechanism could reduce the number of object representations under consideration, substantially facilitating object recognition (Bar 2003; Fenske et al. 2006).

We propose that a similar mechanism might be at work in the CMOR task. When rats have not previously had multimodal object pre-exposure, they have no explicit association between the tactile and visual features of the to-be-remembered objects. Thus, the tactile sample phase of the CMOR task provides only an incomplete representation of the sample object. From this limited information, rats with a functioning OFC may be able to extrapolate candidate visual representations through a top-down process similar to that proposed by Bar (2003). Such a process may not be necessary when the retention delay between sample and choice phases is relatively short, as a simple inferential crossmodal matching strategy might suffice under limited mnemonic load. However, as the delay increases, the incomplete object representation acquired in the tactile sample phase may be particularly vulnerable to trace decay and/or interference from intervening items (Winters et al. 2007; McTighe et al. 2010). One important extension of this study for future analysis could involve assessing the possibility that ventrolateral OFC lesions increase vulnerability to interference in the CMOR task as retention delays are increased, and whether manipulations aimed at reducing such interference, such as those performed by McTighe et al. (2010), would help to reduce or prevent the impairing effects of OFC lesions. Such a finding would provide further support for the putative top-down contribution of the OFC to CMOR task performance.

Clearly, further research is needed to test this notion, but the findings from the present study are intriguing in their implication of the OFC in a specific facet of crossmodal object recognition. Although we did not assess inclusive PFC-lesioned animals for delay-dependent CMOR task impairment, this remains an important question for future analysis. It seems likely that the larger PFC lesions produced in this study would also cause delay-dependent CMOR deficits if the proposed OFC-based mechanism is the underlying cause of the CMOR impairment observed in PFC-lesioned rats. Nevertheless, this remains an empirical question, and it is possible that the more extensive damage produced by inclusive PFC lesions could cause delay-independent effects just as these lesions impaired PE/CMOR performance in the absence of similar effects with more selective PFC damage.

Inclusive damage to the PFC blocked the normally facilitative effects of multimodal object pre-exposure on CMOR task performance with retention delays longer than 1 h (Reid et al. 2012). Thus, regions of the PFC may also be more directly involved in the multisensory associative process required for the development of multisensory object representations. We originally suggested that such a role may be related to OFC function. Unlike inclusive PFC damage, however, OFC lesions did not disrupt PE/CMOR task performance, suggesting that the PFC contribution to this form of crossmodal memory is relatively independent of its role in the more inferential version of CMOR. Indeed, assuming the beneficial mnemonic effects of multimodal object pre-exposure stem from the formation of an explicit association between the tactile and visual features of the to-be-remembered objects, a top-down process such as that proposed above for the OFC would be superfluous in the PE/CMOR task. The putative multisensory object representation formed in the pre-exposure phase would likely be reactivated on tactile sample exploration, thereby providing the animals with a comprehensive object trace in the sample phase. The pattern of results reported here suggest that the PFC may be directly involved in the associative process of multisensory object representation, either in enabling the linkage between tactile and visual object representational areas, or perhaps as the final repository of a multisensory representation (Lacey et al. 2007).

Whatever the specific role of the PFC in PE/CMOR performance, it appears to remain intact in the absence of a fully functioning OFC. Moreover, this contribution does not appear to require an intact mPFC region, as mPFC lesions also did not reverse the CMOR enhancing effect of multimodal object pre-exposure. Indeed, damage to the mPFC failed to disrupt either version of crossmodal object recognition in the current study. This finding is somewhat surprising given the numerous studies implicating homologous regions in multisensory processing. The anterior cingulate cortex, in particular, is often implicated in multisensory integration (Aitken 1980; Banati et al. 2000; Laurienti et al. 2003). It is notable that the mPFC lesions produced in the present study did not extensively damage the anterior cingulate cortex, being restricted to only the rostral portion. Accordingly, the possible effect of more extensive anterior cingulate lesions in CMOR task performance remains a question for future research.

The overall pattern of spared and impaired behavior in this study is intriguing, as it suggests compensatory mechanisms within the PFC for its contribution to the PE/CMOR task. Whereas inclusive PFC lesions disrupted PE/CMOR task performance, separate damage to either of the 2 components comprising these inclusive lesions—the mPFC and ventrolateral OFC—failed to cause impairment. This outcome suggests that, rather than a specific region, what matters most to PE/CMOR task performance is that a minimally sufficient level of PFC functionality remains to mediate the proposed associative process underlying the effect of the multimodal pre-exposure phase. Both mPFC and ventrolateral OFC appear to be able to compensate in this capacity for loss of functionality in the other. Our correlational analysis suggests that this may not be a simple mass action effect related to the total amount of PFC tissue damage, but rather may reflect the importance of retaining a sufficient threshold level of overall PFC functionality. Thus, although these ideas remain largely speculative, and the exact nature of PFC involvement in the CMOR and PE/CMOR tasks cannot yet be clearly articulated, the functional dissociations observed in the present study indicate potentially complex roles of prefrontal regions in crossmodal cognition. Future research with complementary techniques, such as immediate early gene imaging and reversible lesions, should help to further elucidate the specific contributions of prefrontal areas to crossmodal object recognition.

The roles of PFC regions in crossmodal object recognition should also be assessed for different sensory modalities to determine the generality of these effects. The current study, however, provides important new insights into the contributions of the PFC to multisensory integration. Moreover, as the current findings suggest, targeted modifications of the CMOR task can help to elucidate the neural bases of crossmodal cognition, for instance, by illustrating the functional dissociations relating to differential experience with to-be-remembered objects. Indeed, it is apparent from the present results that very brief multimodal exposure to an object can dramatically alter the crossmodal mnemonic capabilities of an organism, and these effects may be the manifestation of underlying reorganization of the neural circuitry representing that object. Future work with variations on the CMOR task should continue to aid in the elucidation of neural circuitry dynamics involved in crossmodal cognition.

Funding

This work was supported by National Sciences and Engineering Research Council (NSERC) of Canada Discovery Grant (400176) to B.D.W., an Alexander Graham Bell Canada Graduate Scholarship to J.M.R., and an Ontario Graduate Scholarship to D.L.J.

Notes

Conflict of Interest: None declared.

References

Aitken
PG
Lesion effects on tactual to visual cross-modal matching in the rhesus monkey
Neuropsychologia
 , 
1980
, vol. 
18
 (pg. 
575
-
578
)
Amaral
DG
Witter
MP
Paxinos
G
Hippocampal formation
The rat nervous system
 , 
1995
San Diego
Academic Press
(pg. 
443
-
493
)
Banati
RB
Goerres
GW
Tjoa
C
Aggleton
JP
Grasby
P
The functional anatomy of visual-tactile integration in man: a study using positron emission tomography
Neuropsychologia
 , 
2000
, vol. 
38
 (pg. 
115
-
124
)
Bar
M
A cortical mechanism for triggering top-down facilitation in visual object recognition
J Cogn Neurosci
 , 
2003
, vol. 
15
 (pg. 
600
-
609
)
Bar
M
Kassam
KS
Ghuman
AS
Boshyan
J
Schmid
AM
Dale
AM
Hamalainen
MS
Marinkovic
K
Schacter
DL
Rosen
BR
, et al.  . 
Top-down facilitation of visual recognition
Proc Natl Acad Sci USA
 , 
2006
, vol. 
103
 (pg. 
449
-
454
)
Cerf-Ducastel
B
Murphy
C
Neural substrates of cross-modal olfactory recognition memory: an fMRI study
Neuroimage
 , 
2006
, vol. 
31
 (pg. 
386
-
396
)
Dalley
JW
Cardinal
RN
Robbins
TW
Prefrontal executive and cognitive functions in rodents: neural and neurochemical substrates
Neurosci Biobehav Rev
 , 
2004
, vol. 
28
 (pg. 
771
-
784
)
Diaconescu
AO
Alain
C
McIntosh
AR
The co-occurrence of multisensory facilitation and cross-modal conflict in the human brain
J Neurophysiol
 , 
2011
, vol. 
106
 (pg. 
2896
-
2909
)
Ennaceur
A
Delacour
J
A new one-trial test for neurobiological studies of memory in rats. 1: behavioral data
Behav Brain Res
 , 
1988
, vol. 
31
 (pg. 
47
-
59
)
Fenske
MJ
Aminoff
E
Gronau
N
Bar
M
Top-down facilitation of visual object recognition: object-based and context-based contributions
Prog Brain Res
 , 
2006
, vol. 
155
 (pg. 
3
-
21
)
Forwood
SE
Winters
BD
Bussey
TJ
Hippocampal lesions that abolish spatial maze performance spare object recognition memory at delays of up to 48 h
Hippocampus
 , 
2005
, vol. 
15
 (pg. 
347
-
355
)
Frassinetti
F
Bolognini
N
Ladavas
E
Enhancement of visual perception by crossmodal visuo-auditory interaction
Exp Brain Res
 , 
2002
, vol. 
147
 (pg. 
332
-
343
)
Fuster
JM
The prefrontal cortex—anatomy, physiology, and neuropsychology of the frontal lobe
 , 
1997
Philadelphia
Lippincott-Raven
Fuster
JM
Bodner
M
Kroger
JK
Cross-modal and cross-temporal association in neurons of frontal cortex
Nature
 , 
2000
, vol. 
405
 (pg. 
347
-
351
)
Gaffan
D
Harrison
S
Auditory-visual associations, hemispheric specialization and temporal-frontal interaction in the rhesus monkey
Brain
 , 
1991
, vol. 
114
 
Pt 5
(pg. 
2133
-
2144
)
Gottfried
JA
Dolan
RJ
The nose smells what the eye sees: crossmodal visual facilitation of human olfactory perception
Neuron
 , 
2003
, vol. 
39
 (pg. 
375
-
386
)
Ku
Y
Ohara
S
Wang
L
Lenz
FA
Hsiao
SS
Bodner
M
Hong
B
Zhou
YD
Prefrontal cortex and somatosensory cortex in tactile crossmodal association: an independent component analysis of ERP recordings
PLoS One
 , 
2007
, vol. 
2
 pg. 
e771
 
Lacey
S
Campbell
C
Sathian
K
Vision and touch: multiple or multisensory representations of objects?
Perception
 , 
2007
, vol. 
36
 (pg. 
1513
-
1521
)
Laurienti
PJ
Wallace
MT
Maldjian
JA
Susi
CM
Stein
BE
Burdette
JH
Cross-modal sensory processing in the anterior cingulate and medial prefrontal cortices
Hum Brain Mapp
 , 
2003
, vol. 
19
 (pg. 
213
-
223
)
Lavenex
P
Amaral
DG
Hippocampal-neocortical interaction: a hierarchy of associativity
Hippocampus
 , 
2000
, vol. 
10
 (pg. 
420
-
430
)
Lipton
PA
Alvarez
P
Eichenbaum
H
Crossmodal associative memory representations in rodent orbitofrontal cortex
Neuron
 , 
1999
, vol. 
22
 (pg. 
349
-
359
)
Manns
JR
Zilli
EA
Ong
KC
Hasselmo
ME
Eichenbaum
H
Hippocampal CA1 spiking during encoding and retrieval: relation to theta phase
Neurobiol Learn Mem
 , 
2007
, vol. 
87
 (pg. 
9
-
20
)
McTighe
SM
Cowell
RA
Winters
BD
Bussey
TJ
Saksida
LM
Paradoxical false memory for objects after brain damage
Science
 , 
2010
, vol. 
330
 (pg. 
1408
-
1410
)
Murray
EA
Malkova
L
Goulet
S
Milner
AD
Cross-modal associations, intramodal associations, and object identification in macaque monkeys
Comparative neuropsychology
 , 
1998
Oxford
Oxford University Press
(pg. 
51
-
69
)
Ongur
D
Price
JL
The organization of networks within the orbital and medial prefrontal cortex of rats, monkeys and humans
Cereb Cortex
 , 
2000
, vol. 
10
 (pg. 
206
-
219
)
Osterbauer
RA
Matthews
PM
Jenkinson
M
Beckmann
CF
Hansen
PC
Calvert
GA
Color of scents: chromatic stimuli modulate odor responses in the human brain
J Neurophysiol
 , 
2005
, vol. 
93
 (pg. 
3434
-
3441
)
Otto
T
Eichenbaum
H
Neuronal activity in the hippocampus during delayed non-match to sample performance in rats: evidence for hippocampal processing in recognition memory
Hippocampus
 , 
1992
, vol. 
2
 (pg. 
323
-
334
)
Paxinos
G
Watson
C
The rat brain in stereotaxic coordinates
 , 
1998
London
Academic Press
Petrides
M
Iversen
SD
Cross-modal matching and the primate frontal cortex
Science
 , 
1976
, vol. 
192
 (pg. 
1023
-
1024
)
Pezze
MA
Dalley
JW
Robbins
TW
Remediation of attentional dysfunction in rats with lesions of the medial prefrontal cortex by intra-accumbens administration of the dopamine D(2/3) receptor antagonist sulpiride
Psychopharmacology (Berl)
 , 
2009
, vol. 
202
 (pg. 
307
-
313
)
Reep
RL
Corwin
JV
King
V
Neuronal connections of orbital cortex in rats: topography of cortical and thalamic afferents
Exp Brain Res
 , 
1996
, vol. 
111
 (pg. 
215
-
232
)
Reid
JM
Jacklin
DL
Winters
BD
Crossmodal object recognition in rats with and without multimodal object pre-exposure: no effect of hippocampal lesions
Neurobiol Learn Mem
 , 
2012
, vol. 
98
 (pg. 
311
-
319
)
Robbins
TW
Dissociating executive functions of the prefrontal cortex
Philos Trans R Soc Lond B Biol Sci
 , 
1996
, vol. 
351
 (pg. 
1463
-
1470
discussion 1470–1461
Rudy
JW
Sutherland
RJ
Configural association theory and the hippocampal formation: an appraisal and reconfiguration
Hippocampus
 , 
1995
, vol. 
5
 (pg. 
375
-
389
)
Schiller
D
Weiner
I
Lesions to the basolateral amygdala and the orbitofrontal cortex but not to the medial prefrontal cortex produce an abnormally persistent latent inhibition in rats
Neuroscience
 , 
2004
, vol. 
128
 (pg. 
15
-
25
)
Small
DA
Crossmodal integration–insights from the chemical senses
Trends Neurosci
 , 
2004
, vol. 
27
 (pg. 
120
-
123
; discussion 123–124.
Small
DM
Prescott
J
Odor/taste integration and the perception of flavor
Exp Brain Res
 , 
2005
, vol. 
166
 (pg. 
345
-
357
)
Stein
BE
Meredith
MA
The merging of the senses
 , 
1993
Cambridge, MA
MIT Press
Tal
N
Amedi
A
Multisensory visual-tactile object related network in humans: insights gained using a novel crossmodal adaptation approach
Exp Brain Res
 , 
2009
, vol. 
198
 (pg. 
165
-
182
)
Uylings
HB
Groenewegen
HJ
Kolb
B
Do rats have a prefrontal cortex?
Behav Brain Res
 , 
2003
, vol. 
146
 (pg. 
3
-
17
)
Vinogradova
OS
Hippocampus as comparator: role of the two input and two output systems of the hippocampus in selection and registration of information
Hippocampus
 , 
2001
, vol. 
11
 (pg. 
578
-
598
)
Weissman
DH
Warner
LM
Woldorff
MG
The neural mechanisms for minimizing cross-modal distraction
J Neurosci
 , 
2004
, vol. 
24
 (pg. 
10941
-
10949
)
Whishaw
IQ
Tomie
J
Kolb
B
Ventrolateral prefrontal cortex lesions in rats impair the acquisition and retention of a tactile-olfactory configural task
Behav Neurosci
 , 
1992
, vol. 
106
 (pg. 
597
-
603
)
Winters
BD
Bartko
SJ
Saksida
LM
Bussey
TJ
Scopolamine infused into perirhinal cortex improves object recognition memory by blocking the acquisition of interfering object information
Learn Mem
 , 
2007
, vol. 
14
 (pg. 
590
-
596
)
Winters
BD
Forwood
SE
Cowell
RA
Saksida
LM
Bussey
TJ
Double dissociation between the effects of peri-postrhinal cortex and hippocampal lesions on tests of object recognition and spatial memory: heterogeneity of function within the temporal lobe
J Neurosci
 , 
2004
, vol. 
24
 (pg. 
5901
-
5908
)
Winters
BD
Reid
JM
A distributed cortical representation underlies crossmodal object recognition in rats
J Neurosci
 , 
2010
, vol. 
30
 (pg. 
6253
-
6261
)