Impaired recognition memory and cognitive flexibility in the ratL5–L6 spinal nerve ligation model of neuropathic pain

2.4.1 von Frey test for mechanical allodynia

von Frey testing was carried out as described previously [49, 50,51, 52,54,55]. The arena used for von Frey testing consisted of six adjoining chambers with dimensions 25 cm (l) × 20 cm (w) × 14 cm (h). The sides were made of clear Perspex and the partitions between chambers from white melamine-coated chipboard. The arena was placed on a raised wire-mesh floor so that the experimenter could access the rats’ hind-paws from below. Rats received an initial habituation period of 30 min during which they were placed in individual chambers of the arena and no testing was performed. On subsequent test days, the rats were habituated to the arena for 20 min prior to assessment. von Frey filaments (Touch-Test® Sensory Evaluators, North Coast Medical, Inc., CA, USA) of different weights (0.16–180 g), were applied perpendicular to the plantar surface of the hind-paw, with sufficient force to cause slight buckling of the filament, up to a maximum of 6 s or until a positive result (flinching, licking or withdrawal of the paw) was observed. Filaments were applied to both left and right hind-paws five times (alternating between paws) in order of increasing weight until a 100% positive response (5 positive responses to 5 applications) was observed. The filament weight eliciting 50% response was calculated by plotting the % response versus filament weight for each rat. A mild detergent was used to clean the arena between each group of 6 rats.

2.4.2 Hargreaves test for thermal hyperalgesia

A commercial apparatus (IITC Life Science Inc., Woodland Hills, CA, USA) was used for Hargreaves testing as previously described [49,50,52,54]. The apparatus consisted of a three-chambered Perspex arena (chamber dimensions 22 cm × 20 cm × 15 cm, l×w×h) placed on top of a glass plate heated to 30 ± 1 °C. Rats were habituated to the arena for 30 min on the day before baseline testing began, and on test days were habituated to the arena for 20 min prior to testing. A moveable radiant heat source was positioned underneath the glass and could be focused on the rat’s hind-paw. The heat source was set to an active intensity of 30% and focused from below on the plantar surface of the rat’s hind-paw. The thermal stimulus was applied until a positive response (criteria similar to those used for von Frey testing) was recorded or until a cutoff time of 20 s was reached. Right and left hind-paws were tested 4 times, alternating between paws, and the average withdrawal latency for each paw was calculated. The arena was cleaned with mild detergent between testing each group of animals.

2.4.3 Acetone drop test for cold allodynia

The acetone-drop test was used to measure responding to an innocuous cold stimulus as previously described [49, 52, 53,54]. The arena used for this test was identical to that used for von Frey testing. A 1 ml syringe with a short length of polyethylene Portex® tubing (1 mm internal diameter) attached was used to apply approximately 0.2 ml of acetone (Sigma, Ireland) to the plantar surface of the hind-paw, without mechanically stimulating the paw. Each hind-paw was tested 3 times, alternating between paws, the number of positive responses (flinch, lick or withdrawal of the hind-paw) within 60 s of acetone application for each trial was recorded, and the total across trials calculated. The arena was cleaned with mild detergent between testing of each group of animals.

2.5.1 Novel-object recognition test

The novel-object recognition test procedure used herein was based on a number of protocols described previously [56,57] with some modifications. Testing was carried out in a circular arena with a wooden base and metal sides (both painted black), with dimensions 75 cm (d) × 38 cm (h). The arena was illuminated by four 60 W bulbs which provided constant light intensity of 100lx. A video camera positioned above the arena was used to record behaviour during testing for subsequent analysis. The objects used were plastic Coca-Cola® bottles (filled with water) with a base diameter of 4.5 cm and 23.5 cm height and a plastic structure constructed from green and blue toy blocks with dimensions base area 5 cm² and height 16 cm. The objects had no apparent natural significance to the rats, and were secured to the base of the arena with white tack such that they were difficult to displace. Animals were habituated to the arena in the absence of objects for 20 min on the day before the test day. The test day comprised of three stages: (i) habituation, (ii) exposure 1 and (iii) exposure 2. Rats were introduced to the arena for a 3 min habituation period and then returned to their home cage for 7 min. During exposure 1, two identical objects (Coca-Cola® bottles) were placed in opposite quadrants of the arena, 16 cm from the perimeter. The rat was allowed to freely explore the arena and objects for a period of 3 min, after which the animal was removed from the arena and returned to its home cage for an interval of 5 min. Prior to exposure 2, one of the bottles was replaced with a novel object (plastic structure made from interlocking toy blocks). The animal was again allowed to freely explore the arena and objects for a period of 3 min and then returned to its home cage. The arena was cleaned with a mild detergent between rats to remove odours and olfactory cues, and faecal pellets were removed between exposures. Exploration of an object was defined as sniffing the object, rearing against the object or having the head directed towards the object within a 2 cm annulus of the object. Exploratory and general behaviours were manually rated from the DVD recordings of each of the three test stages, with the aid of Ethovision® behavioural tracking software (Noldus, The Netherlands). The software was also used to track the distance moved (in cm) by the animal during testing. The proportion of time spent exploring the object was assessed by calculating a discrimination ratio as follows:

2.5.2 Air-puff-induced passive avoidance

The procedure was similar to that described by Moriarty et al. [54]. Passive-avoidance testing was carried out in a specially constructed light/dark arena. The light compartment was made of white melamine-coated chipboard with dimensions of 30 cm (l) × 30 cm (w) × 40 cm (h) and was lit from above using a standard 60 W bulb, such that the compartment was maintained at a constant light intensity of 100l×. The dark compartment was made of dark grey Perspex with a black wooden lid, and its dimensions were identical to those of the light compartment. Light intensity in the dark compartment was negligible. The two compartments were separated by a manually controlled guillotine door. For delivery of air-puff directed at the rat’s face, an air-duster canister (Electrolube, GDP, UK) was mounted to the outside of the dark compartment, such that a nozzle could protrude up to 3.5 cm into the dark compartment through a 5 mm diameter hole in the compartment wall opposite the guillotine door. The nozzle was positioned 3 cm from the floor of the arena (approximately nose-height of the rat). Rats were habituated to the arena for 5 min 24 h prior to testing. During this period, rats were placed in the light compartment of the arena and allowed to freely explore both the light and dark compartments of the arena and then returned to their home cages. In the acquisition trial, the rat was placed in the light compartment of the arena, facing away from the guillotine door and the latency to enter the dark compartment was recorded up to a maximum of 300 s. Once the rat had entered the dark compartment, the guillotine door was lowered by the experimenter. Once the rat was facing the wall opposite the guillotine door, a single, brief puff of air (duration of ~1 s) was administered to the face. The rat was confined to the dark compartment for a further 90s post-air-puff, after which it was removed from the arena and returned to its home cage. If the rat did not enter the dark chamber within the 300-s period it was removed from the arena, returned to its home cage and excluded from further testing. For the retention trial (carried out 24 h post acquisition) the rat was again placed back into the light compartment of the arena, facing away from the guillotine door and the step-through latency to enter the dark compartment was recorded. If the animal did not enter the dark compartment within 300 s they were removed from the arena and returned to their home cage, and latency was recorded as 300 s. No air puff was administered during the retention trial and the arena was cleaned with mild detergent between rats to remove olfactory cues.

2.5.3 Morris water maze

Procedures for water maze cued test, acquisition training, forward probe trial, reversal training and reversal probe trials were similar to those described previously [58]. The Morris water maze (MWM) apparatus consisted of a circular white plastic pool, 2 m in diameter, and a platform made of clear Perspex (10 cm × 30 cm, d × h). The lighting of the maze was kept constant at 100lx (at water level) and the water temperature was maintained at 25 ± 3 °C throughout the testing. A video camera located above the apparatus was connected to a DVD recorder for subsequent behavioural tracking using Ethovision® software (Noldus, The Netherlands). The maze was surrounded by curtains, on which visual cues were hung during acquisition and reversal training and during probe trials. The cues consisted of geometric shapes (a filled circle, an open square and a series of wavy lines) printed in black on white A3 paper.

The cued test was used as a control procedure to assess rats’ vision, ability to swim and ability to identify the platform as an escape route from the maze. The pool was filled to 2 cm below the height of the platform and the top of the platform was covered with black plastic so that it was clearly visible. Rats underwent 4 trials each and both the start position and the platform position were quasi-randomised. Rats were given 120 s to locate the visible platform. Once the animal located the platform successfully, they were allowed to remain there for 10 s. If the animal did not locate the platform within 120 s, it was guided towards it by the experimenter and placed onto the platform for 10 s. The animal was then removed from the pool, dried off with a cotton towel and returned to a heated recovery cage for an inter-trial period (approximately 5–15 min). For acquisition training, the pool was filled to 2 cm above the level of the platform such that the platform was hidden below the surface of the water. The cues were hung on the curtains surrounding the pool and were visible at water level. The training consisted of four trials per day over five consecutive days, throughout which the platform was positioned in the southwest quadrant of the pool. For each trial, the animal was released from one of four release points in a quasi-random order, with its head facing towards the wall of the pool and away from the platform. Rats were given 120 s to locate the hidden platform. Once the animal located the platform successfully, they were allowed to remain there for 10 s. If the animal did not locate the platform within 120 s, it was guided towards it by the experimenter and placed onto the platform for 10 s. The animal was then removed from the pool, dried off with a cotton towel and returned to a heated recovery cage for an inter-trial period (approximately 5–15 min). The distance moved to get onto the platform (path length) was the primary outcome measure and was determined with the aid of Ethovision® software. The order in which the rats were tested was randomised every day to avoid any confounding effects of time of day. A probe trial was carried out the day after acquisition training was completed. The platform was removed from the maze, and rats were released from the northeast quadrant. The probe trial consisted of a single 120-s trail for each rat. The proportion of time spent and distance moved in each quadrant was determined with the aid of Ethovision® software. For reversal training, the protocol was the same as for acquisition training except that the platform was positioned in the opposite quadrant of the pool (northeast). Reversal training was performed over five consecutive days. The parameters tested were: proportion of time spent and proportion of distance moved in the “old” location or area where the platform had been located (southwest quadrant), and the proportion of time spent and proportion of distance moved in the “new” platform location (northeast quadrant). At the end of reversal training, a reversal probe trial was carried out. This was similar to the initial forward probe trial except the rats were all released from the southwest quadrant.

3.1.1 von Frey test

The 50% mechanical response threshold of the ipsilateral hind-paw was decreased in the SNL group compared with the sham control group on days 3–13 post-surgery and at the final time-point day 62/64 post surgery, suggesting persistent expression of mechanical allodynia in SNL rats (Fig. 2). The response threshold was also significantly decreased post-surgery in both sham and SNL groups compared with their respective baselines. Full details of the outcome of statistical analyses are described in the figure legends.

Fig. 2

Effect of SNL surgery on the 50% mechanical withdrawal threshold of the ipsilateral hind-paw. An overall Friedman’s ANOVA by ranks produced a significant result (χ² = 113.39, p<0.001). Friedman’s ANOVA by ranks on individual groups revealed significant effects of time in both sham (χ² =36.06, p<0.001) and SNL groups (χ² =43.15, p<0.001). SNL rats expressed mechanical allodynia from day 3 post surgery onwards compared with sham controls (Mann-Whitney Ö tests; *p<0.05, **p<0.01, ***p<0.001) and with pre-surgery baseline (Wilcoxon signed-ranks test; ⁺⁺p< 0.001). There was also a decrease in threshold in the sham controls post-surgery compared with baseline (Wilcoxon signed-ranks test; ^##p<0.01). n= 11–12

3.1.2 Hargreaves test

Paw withdrawal latency was decreased in the ipsilateral hind-paw of SNL rats compared with sham rats on days 8,16 and 63/65 post-surgery (Fig. 3). Latency was also decreased in both the sham (days 8 and 16) and SNL group (days 8, 16 and 63/65) compared with their respective baselines. These results suggest expression of thermal hyperalgesia in the SNL group which was still evident 63–65 days post-surgery.

Fig. 3

Effect of SNL surgery on ipsilateral hind-paw withdrawal latency in the Har-greaves test. The overall Friedman’s ANOVA by ranks was significant (χ² =37.10, p< 0.001). Friedman’s ANOVA by ranks for the SNL group indicated a significant effect of time in that group (χ² = 16.56, p = 0.001). Paw withdrawal latency was decreased in the SNL group compared with the sham control group at all post-surgery time points (Mann-Whitney U tests; **p<0.01, *p<0.05). Withdrawal latency decreased over time in the sham group (Wilcoxon signed-ranks tests; ^#p<0.05 days 8 and 16 vs. baseline) and in the SNLgroup (⁺p<0.01 days 8,16 and 63/65 vs. baseline). n= 10–12.

3.1.3 Acetone-drop test

The number of responses to acetone application was significantly greater in the SNL group than in the control group on days 12 and 61/63 post-surgery (Fig. 4). There was a strong trend for a similar increase on day 4, that just failed to reach the level of statistical significance (p = 0.059). The number of responses was also increased post-surgery compared with baseline in both sham and SNL groups compared with their respective baselines. These data suggest persistent expression of cold allodynia in the nerve-ligated rats.

Fig. 4

Effect of SNL surgery on the total number of hind-paw withdrawal responses following acetone application to the ipsilateral hind-paw. The overall Friedman’s ANOVA by ranks was significant (χ² =57.54, p<0.001) and Friedman’s ANOVA by ranks showed effects of time in both the sham (χ² =26.04, p< 0.001) and the SNL (χ² = 26.06, p< 0.001) groups. The number of ipsilateral hind-paw withdrawal responses to acetone was significantly greater in SNL rats compared with sham controls on days 12 and 61/63 post-surgery (Mann-Whitney U tests; **p<0.01) and a similar trend was observed on day 4 (p = 0.056) There was an increase in the number of responses post-surgery compared with baseline in both the sham (Wilcoxon signed-ranks tests; ^#p<0.05) and the SNL groups (⁺⁺p<0.01) compared with their respective baselines. n= 11–12.

3.2.1 Novel-object recognition

Discrimination ratios for the identical objects in exposure 1, and familiar and novel objects in exposure 2 were calculated according to the formula given in Section 2.5.1 There were no differences in object exploration (object 1 vs. object 2) in either group or between groups in exposure 1 (Fig. 5). In exposure 2, only the sham group displayed a significant preference for the novel object as shown by a greater discrimination ratio for the novel-object than for the familiar object. Familiar object exploration was significantly lower and novel-object exploration was significantly higher in the sham group than in the SNL group, suggesting there was a deficit in recognition memory in the nerve-injured rats (Fig. 5). SNL surgery was not associated with alterations in locomotor activity as measured by the total distance moved during the 5 min novel-object habituation trial (sham 1402.9 ±172.2 cm vs. SNL 1284.2 ± 132.7 cm, p = 0.62, Student’s unpaired two-tailed t-test), suggesting that the reduced novel-object discrimination was not due to surgery-related impairment of movement.

Fig. 5

Novel-object recognition in sham and SNL rats. Each exposure was analysed by one-way ANOVA (exposure 1: F(₃,₃₇) = 0.049, p = 0.985; exposure 2:F(3,₃₇) = 3.534, p = 0.025). There were no differences in object discrimination in exposure 1. In exposure 2, sham rats only expressed a preference for the novel object (Fisher’s LSD post hoc test; **p<0.01 vs. sham-familiar). The exploration of the familiar object was higher in the in the SNL group than in the sham group and exploration of the novel object was lower in the SNL group than in the sham group (^#p<0.05). n = 8–ll.

3.2.2 Air-puff passive avoidance

Exposure to air-puff produced a significant passive-avoidance response in both sham and SNL groups, as indicated by the increase in retention step-through latency compared with acquisition but there was no significant difference in the response between the sham and the SNL groups (Fig. 6). These results suggest that SNL surgery did not affect passive-avoidance responding in this paradigm.

Fig. 6

Passive-avoidance responding following air-puff exposure in sham and SNL rats. An overall Friedman’s ANOVA including both groups and both time points was significant (χ² = 18.52, p< 0.001). Both the sham (^##p < 0.01) and SNL (^#p< 0.05) groups showed an increase in step-through latency in the retention trial compared with the acquisition trial (Wilcoxon signed-ranks tests) but there was no difference in the retention step-through latency between sham and SNL groups. n = 8–10.

3.2.3 Morris water maze

3.2.3.1 Acquisition training and forward probe trial

In the acquisition training phase, the path length decreased over time in both the sham and the SNL groups, which suggests that rats successfully learned to locate the platform (Fig. 7A). There were no between-group differences, indicating that both groups learned equally well, and therefore SNL surgery did not affect the acquisition of spatial reference memory. A similar pattern of results was observed for the latency to get onto the platform (data not shown). In the forward probe trial, the percentage distance moved in each of the quadrant zones, and the platform and annulus zone, by sham and SNL groups, was compared by Student’s unpaired two-tailed r-tests. There were no differences in the percentage distance moved (Fig. 7B) or the percentage time spent (data not shown) in any of the zones, suggesting there was no effect of SNL surgery on spatial memory as assessed by the paradigm. As expected, the percentage distance moved in the SW quadrant, where the platform had been located, was above the level of chance (25%, Fig. 7B).

Fig. 7

(A) Path length to locate the platform in sham and SNL rats in the acquisition phase of the Morris water maze. Two-way repeated measures ANOVA of log transformed data revealed an overall effect of time (F(4,84) = 21.97, p<0.001). The path length decreased over days in both the sham (^##p<0.01 days 2–5 vs. day 1) and the SNL groups (⁺⁺⁺p< 0.001 days 3–5 vs. day 1; Fisher’s LSD post hoc tests). There were no between-group differences at anytime point. (B) Percentage distance moved in arena zones during the probe trial in sham and SNL groups. The dashed line represents the % distance animals would move in each quadrant by chance (25%). Distance moved by both sham and SNL groups in the SW quadrant, where the platform had been located, was above the level of chance. There were no between-group differences (Student’s unpaired r-tests). n= 11–12.

3.2.3.2 Reversal training and reversal probe trial

The percentage distance moved in the southwest quadrant decreased over time in both sham and SNL groups. The proportion of distance moved in the SW quadrant, i.e. the old platform location, was significantly greater in the SNL rats compared with the sham rats on day 2 (Fig. 8A). A similar result was observed in assessing the percentage time spent in the old quadrant location, with SNL rats spending a significantly greater proportion of time in the old quadrant location than the sham controls on day 2 of reversal training (sham 16.9± 1.6% vs. SNL 24.2 ±2.8%; p<0.05, data not shown). These results suggest that the SNL rats adapted more slowly to the change in platform location compared with sham controls and tended to revert to the old location more frequently. The percentage distance moved in the new platform location increased over time in both the sham and SNL groups. There were no significant differences between the groups at any of the time points (data not shown). In the reversal probe trial, the percentage distance moved by sham and SNL rats in each of the defined zones was calculated. The percentage distance moved in the NE quadrant was, as expected, greater than the level of chance as this is where the platform had most recently been located. Student’s unpaired two-tailed t-tests were used to compare the percentage distance moved in each of the four quadrant zones and the platform and annulus zone in sham and SNL groups. The sham group spent a significantly greater proportion of their total distance moved in the NE quadrant (new platform location) than the SNL group in the reversal probe trial (Fig. 8B). These data indicate that despite reversal training over 5 days, learning and memory of the new platform location was impaired in the SNL group compared with the sham controls. Taken together, the results of the reversal training and reversal probe tasks suggest a pattern of impaired cognitive flexibility, with SNL rats tending to return to the old platform location more than sham controls, and to spend less time in the new platform location. This was also evident from the tracking of the paths of individual rats in the reversal trials, which was done with the aid of Ethovision® software (see Fig. 8C).

Fig. 8

(A) Percentage distance moved in the SW quadrant (old platform location) by sham and SNL rats (data averaged over 4 trials per day). Two-way repeated measures ANOVA revealed a significant effect of time (F(₄,₈₄) – 20.43, p< 0.001) and a significant interaction of time and surgery (F(₄,₈₄) – 2.76, p – 0.033). The % distance moved in the old location decreased overtime in both groups (Fisher’s LSD post hoc tests; ^##p<0.01 sham days 2–5 vs. day 1, ⁺⁺p<0.01, SNL days 3–5 vs. day 1). The % distance moved in the old quadrant was significantly greater in the SNL group than the sham control group on day 2 (Fisher’s LSD post hoc test; *p<0.05). (B) Percentage distance moved in arena zones during the reversal probe trial in sham and SNL groups. The dashed line represents the % distance animals would move in each quadrant by chance (25%). The % distance moved in the NE quadrant was above the level of chance for both sham and SNL groups. The sham rats’ distance moved in the NE quadrant as a % of their total distance moved during the trial was greater than that of the SNL rats (Student’s unpaired t-test; *p<0.05). n= 11–12. (C) Representative images of individual Ethovision® tracks for an individual (i) sham or (ii) SNL rat. Images indicate that the proportion of time and movement in the old location was greater in the SNL rats than in the sham controls.

3.3 Immunohistochemistry

The synaptophysin density, or the area of positive synapto-physin staining as a percentage of the total tissue area, was calculated for all CA1 and mPFC (prelimbic area) images. No significant hemispheric differences were observed and therefore images from the right and left sides of the brain were pooled. The results from each rat were used to calculate an average density per rat and the data presented shows the average per group, taking each rat as an experimental unit (n = 4–6, see Figs. 9D and 10D). There were no significant differences in synaptophysin density between sham and SNL groups in either region of interest. Representative 60 × images from sham and SNL rats from both regions are presented in Figs. 9 and 10 (B and C). To investigate whether SNL surgery was associated with differences in the distribution of synaptophysin, double-labelling immunohistochemistry for synaptophysin and the vesicular GABA transporter vGAT, and synaptophysin and the vesicular glutamate transporter vGLUT was performed. Pearson’s coefficient was used as a measure of colocalisation. Similar to single-labelled images, an average Pearson’s value per rat was calculated and then used to calculate an average per group. Results are presented in Table 1 and representative 60× double stained images are presented in Fig. 11. There were no between-group differences in the colocalistaion of synaptophysin and either vGLUT or vGAT in the regions of interest, the mPFC and the CA1.

Fig. 9

(A) CA1 region of the hippocampus adapted from the Rat Brain Atlas [59]. The black squares indicate the area from which the images were obtained. (B) 60x representative image of synaptophysin immunostained section from the CA1 of a sham-operated and a (C) SNL-operated rat. Scale bar–30 μm. (D) Quantification of synaptophysin staining density in the CA1 of sham and SNL rats. There was no difference in the average % area of synaptophysin staining between groups (Student’s unpaired t–test, p > 0.05).

Fig. 10

(A) Medial prefrontal cortex (mPFC, prelimbic area) adapted from the Rat Brain Atlas [59]. The black squares indicate the area from which the images were obtained. (B) 60 × representative image of synaptophysin immunostained section from the mPFC of a sham-operated and a(C) SNL-operated rat. Scale bar–30μm. (D) Quantification of synaptophysin staining density in the mPFC of sham and SNL rats. There was no difference in the average % area of synaptophysin staining between groups (Student’s unpaired t–test, p > 0.05).

Fig. 11

Representative images of double immunostaining for synaptophysin and vGAT (A-D) and synaptophysin and vGLUT (E-H) in the CA1 region of the hippocampus and the mPFC (prelimbic area). Scale bar–30 μm.