Nociceptive input after peripheral nerve injury results in cognitive impairment and alterations in primary afferent physiology in rats

Abstract

Pain alters cognitive performance through centrally mediated effects in the brain. In this study, we hypothesized that persistent activation of peripheral nociceptors after injury would lead to the development of a chronic pain state that impairs attention-related behavior and results in changes in peripheral neuron phenotypes. Attentional performance was measured in rats using the 5 choice serial reaction time titration variant to determine the initial impact of partial L5 spinal nerve ligation (pSNL) and the effect of persistent nociceptor activation on the resolution of injury. The changes in peripheral neuronal sensibilities and phenotypes were determined in sensory afferents using electrophysiologic signatures and receptive field properties from DRG recordings. Partial spinal nerve injury impaired attentional performance and this was further impaired in a graded fashion by nociceptive input via an engineered surface. Impairment in attention persisted for only up to 4 days initially, followed by a second phase 7 to 10 weeks after injury in animals exposed to nociceptive input. In animals with prolonged impairment in behavior the mechano-nociceptors displayed a persistent hypersensitivity marked by decreased threshold, increased activity to a given stimulus, and spontaneous activity. Nerve injury disrupts attentional performance acutely and is worsened with peripheral mechano-nociceptor activation. Acute impairment resolves but persistent nociceptive activation produces re-emergence of impairment in the attention-related task associated with electrophysiological abnormalities in peripheral nociceptors. This is consistent with the development of a chronic pain state marked by cognitive impairment and related to persistently abnormal peripheral input.

1.0. Introduction

Chronic pain after injury remains a major health problem that results in costs to patients and society manifest as reduced functional capacity and quality of life, medical costs, and time-expense from decreased work days and decreased productivity [22]. Additionally, persistence of pain contributes to increased opioid consumption and also may increase substance abuse disorder [35]. Despite the broad implications of chronic pain after injury, our understanding of the neurobiologic mechanisms of the transition from acute to chronic pain has been limited with few effective therapeutic developments. In many respects, this translation chasm can be attributed to the lack of effective animal models that permit mechanistic studies designed to both understand and prevent the transition to chronic pain from acute injury as chronic pain is not just a temporal continuum of acute nociception [29,33]. Appropriate animal models are critical for study and evaluation of not just neurobiology, but for testing pharmacologic interventions at defined time points in the evolution of the chronic pain state [1,28,29,33]. It is likely that only through understanding the neurobiology of these dynamic changes that effective treatments will emerge to prevent, halt or reverse the progression from acute to chronic pain.

It has been suggested that central rather than peripheral mechanisms are primarily responsible for chronic pain [24,30,43]. While central changes indisputably occur and chronic pain can result from central pathology, most chronic pain is initiated in the periphery and peripheral input is a critical component of its development [23,44]. Animal models of acute and neuropathic pain typically involve peripheral tissue injury or nerve damage. This results in the disruption of mechanosensory afferents (both sensitization and desensitization) and the emergence of unexcitable units [8–10]. These events contribute to the peripheral neuropathic state, with characteristic mechano-hyper- and hyposensitivity. Animal studies have largely focused on elicited sensory changes, particularly mechanosensitivity, thought to represent nociceptive input to the central nervous system [5]. Cognitive dysfunction following nerve injury is likely a measure of abnormal sensory input that is implicitly or explicitly perceived and likely surmised to be pain. One such non-reflexive measure of nociceptive input is place avoidance and we previously used a geometric surface (NOX) designed to activate sensitized fast-conducting nociceptive afferents (A-High Threshold MechanoReceptors: AHTMRs) after injury, which induced place avoidance reversed with morphine [10]. Higher level central nervous system dysfunction from noiciception has also been measured by reduced performance in a visual attention task, a variant of the five choice serial titration task [3,38,39,45]. Impaired attention is seen clinically as a result of acute and chronic pain [16,17,47,48]. This may occur with other cognitive impairments and can predispose to reduced function and quality of life [27,40,47,48]. The clinical impact of this suggests attentional impairment is a valuable measure for the relevance of pain in the animal model. In this study we hypothesized that persistent activation of peripheral nociceptors using the NOX surface after partial L5 spinal nerve ligation (pSNL) would delay recovery in cognitive function and this delayed recovery would manifest as continued abnormalities in mechanosensitive afferent nerves.

2.0. Methods

2.1. Animals.

Male Fisher 344 rats (N=32, 275–300g at beginning of experiments, Harlan/Envigo, Indianapolis IN) were used for all studies. Animals were maintained at 90% of their free-feeding weight accommodating for normal growth and weight gain throughout the study according to published growth curves. Animals were kept on a reversed light:dark cycle (dark 05:00–17:00) in a temperature and humidity controlled AAALAC approved vivarium immediately adjacent to the behavioral laboratory. Animals had ab lib access to water except during experimental sessions. All procedures were in accordance with guidelines provided by the National Institutes of Health and the International Association for the Study of Pain and were approved by the Institutional Animal Care and Use Committee of Wake Forest University Health Sciences.

2.2. Five choice serial reaction time titration variant (5CTV) procedure

Apparatus.

Commercially available operant chambers (Med Associates Inc., Fairfax, VT) equipped with one wall containing 5 apertures with LEDs and an opposite wall containing a food hopper and pellet dispenser were used for all studies as described previously [38]. Each chamber also contained a stimulus light with a red lens cap located in the middle of the wall with the food hopper, a stimulus light within the food hopper with a red lens cap, and a tone generator. Operant chambers were each separately enclosed within a sound- and light-attenuating expanded PVC-walled chamber (Med Associates). All behavioral sessions were controlled using custom software written in Med-PC IV using a PC-compatible computer and interface (Med Associates).

Procedure.

The 5CTV procedure was conducted as reported previously [38]. All behavioral sessions were conducted on weekdays during the dark phase of the light:dark cycle. Briefly, animals were trained in the task using 4 phases of training. The first phase consisted of training the animal to obtain a single 45 mg chocolate-flavored rat chow pellet (Formula F0299, Bio-Serv Inc., Flemington, NJ) by head entry into the food hopper for a maximum of 100 trials or 30 min. Food availability was signaled by illumination of the food hopper light. Once the rat obtained a minimum of 80 pellets within 30 min for 2 consecutive days, animals were trained to nosepoke into the middle aperture on the opposite wall to obtain 2 food pellets dispensed into the food hopper (phase 2). Sessions were initiated by delivery of two pellets into the food trough and illumination of the food trough light. Once the animal interrupted the head entry detector on the food trough, the trough lamp was turned off after 2 sec and the trials were initiated, signaled by illumination of the house light. Each trial consisted of the LED in the middle nose poke being illuminated for 30 sec (cue duration) during which time a nose poke resulted in the LED being turned off and the food trough light being illuminated and delivery of two food pellets. Head entry detection at the food trough initiated a 2 sec reward cycle timer, after which the food trough lamp was turned off and an inter-trial interval (ITI) timer of 5 sec was initiated. After the ITI the next trial began, signaled by illumination of the middle nose poke LED. If the animal responded in a nose poke other than the middle one (incorrect response) or did not respond within 30 sec (limited hold, omission of response), the LED was turned off and a 2 sec time-out period was initiated during which all lights were turned off. Responses in any of the nose pokes during this time-out period reset the 2 sec time-out timer. At the end of the time-out, the next trial was initiated, signaled by illumination of the house light and after the ITI, illumination of the middle nose poke LED. Responses during the ITI were recorded as premature responses and resulted in initiation of a time-out. Sessions consisted of 50 trials or 30 min whichever came first. Animals were required to complete all 50 trials with a minimum of 80% correct responses for 3 consecutive sessions before the third phase of training began. The third phase of training was similar to phase 2, except that one of the five nosepoke apertures was illuminated at random during successive trials and nosepoke into the illuminated aperture was considered the correct response. The criterion for advancement to the final phase of training was the same as for advancement from phase 2 to phase 3. The fourth and final phase of training was the titration phase. For this phase, the duration of aperture illumination (cue duration) was set to 30 sec in the first trial and correct responses resulted in a decrease in the cue duration in successive trials, while incorrect or omitted responses resulted in an increase in the cue duration in successive trials. The cue durations followed the array (sec): 30, 25, 20, 15, 10, 8, 6, 4, 2, 1, 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.2, 0.1. The trial duration was the same as the cue duration or 5 sec, whichever was greater, thus permitting the animal at least 5 sec to respond (limited hold, LH). Failure to respond within the LH was recorded as an omission. Incorrect responses or omissions resulted in a 5 sec time-out, signaled by turning off all lights. The end of the time-out period was signaled by illumination of the red stimulus light on the food hopper wall, followed by initiation of the next trial after the intertrial interval of 5 sec. Sessions consisted of a maximum of 100 trials or 30 min.

Behavioral endpoints.

The primary outcome measure for the 5CTV method was the log of the median cue duration (log MCD), determined from the individual cue durations for trials 16–100. The log of the MCD has been found to be more consistently normally distributed than MCD. The first 15 trials were omitted as the animal is titrating the cue down to maximal performance; as the cue is based on up down titration, there is only titration down during these trials and no information about maximal performance in the task in the up down methodology. The number of correct, incorrect, omissions, and total trials completed was also recorded, as was the latency to make correct or incorrect responses and the latency to retrieve the food reward. The % correct, %incorrect, and %omissions was calculated by dividing the number of correct, incorrect, or omission trials by the total trials completed, respectively. The %accuracy was calculated by dividing the number of correct trials by the sum of correct and incorrect trials. Latency to make a correct response and latency to retrieve the food reward were analyzed as secondary outcomes. The other measures were recorded and analyzed for comparison with the traditional 5 choice serial reaction time task method. The measures typically reported for the classical 5 choice serial reaction time task were also recorded: number of correct, incorrect, and omission trials, total trials completed, %accuracy (defined as number of correct trials divided by the sum of the number of correct and incorrect trials) and the %corrrect, %incorrect, and %omission trials (number of correct, incorrect, or omission trials divided by the number of total trials completed, respectively). The number of subjects was determined using a power calculation using G*Power ver. 3.1.9.2 (http://www.gpower.hhu.de/en.html). This calculation was based on the MCD, with the ability to detect an effect size η² of 0.15 at power of 0.9 and alpha level of 0.05, with repeated measures ANOVA design with 4 groups and 50 time point comparisons, yielding an estimated N of 8 rats/group.

Partial L5 ligation.

Partial L5 spinal nerve ligation (pSNL) was performed as described previously [8]. Briefly, under isoflurane anesthesia the right L5 nerve was visualized under a surgical microscope following a dorsal incision along the back of the animal and removal of the L6 transverse process with microrongeurs. One-third to one-half of the L5 nerve was ligated using 8–0 nylon suture and the muscle and skin were closed in separate layers with absorbable suture. Sham animals received anesthesia only for a similar duration as the pSNL rats (20 min). Animals received 75,000 U of penicillin G procaine s.c. after the procedure.

Effects of pSNL on 5CTV performance.

Once responding was stable in the 5CTV procedure, defined as a minimum of 5 successive sessions during which the MCD was 1 sec or less and did not vary by more than 15% from the mean across sessions, animals were assigned to one of four groups. One group received pSNL surgery and all behavioral sessions were conducted using the standard grid bar floor (Med Associates). Two additional groups had pSNL surgery with behavioral sessions conducted with engineered surfaces designed to stimulate sensitized AHTMRs in a graded fashion as reported previously [9,10]. One surface consisted of a stainless steel plate that contained contiguous square pyramidal shapes that were 5 mm at the base and 0.2 mm at the apex (NOX 0.2). Another was similar with an apex size of 1.0 mm (NOX 1.0). These two surfaces were chosen as on the NOX 0.2 maximally activates AHTMR or mechanical nociceptors and the NOX1.0 does so to a very limited degree as described and shown to create different place preferences after injury [10]. The plate was shaped to fit within the operant chamber with the pyramidal surfaces machined onto a 7”x9.5” rectangle in the middle of the plate. A fourth group of rats received sham surgery and behavioral sessions were conducted on NOX 0.2 plates. Animals received pSNL or sham surgery on a Monday with no 5CTV session conducted on that day. 5CTV sessions were conducted each weekday for 10 weeks beginning 24 hr after pSNL or sham surgery.

2.3. Electrophysiology

At the conclusion of behavioral studies, the animals were anesthetized for electrophysiology study with the investigator blinded to experimental treatment. Animals were deeply anesthetized with 3% isoflurane, the trachea was intubated and the lungs ventilated using pressure controlled ventilation (Inspira PCV, Harvard Apparatus, Holliston, MA) with isoflurane in humidified oxygen. Heart rate and noninvasive blood pressure were monitored throughout as a guide to depth of anesthesia. Anesthetized animals were immobilized with pancuronium bromide (2 mg/kg) and inspired and end tidal isoflurane concentration maintained at 2% throughout the study. A dorsal incision was made in the lumbar midline and L4 dorsal root ganglion (DRG) and adjacent spinal cord were exposed by laminectomy as previously described [6]. The tissue was continuously superfused with oxygenated artificial cerebrospinal fluid [aCSF (in mM): 127.0 NaCl, 1.9 KCl, 1.2 KH2PO4, 1.3 MgSO4, 2.4 CaCl2, 26.0 NaHCO3, and 10.0 D-glucose]. The spinal column was secured using custom clamps and the anesthetized animal was transferred to a preheated (32–34°C) recording chamber where the superfusate was slowly raised to 37°C (MPRE8, Cell MicroControls, Norfolk, VA) and its flow rate kept at 1 ml per minute (bath exchange time of ~5 s). Pool temperature adjacent to the DRG was monitored with a thermocouple (IT-23, Physitemp, Clifton, NJ). Rectal temperature (RET-3, Physitemp) was maintained at 36 ± 1°C with radiant heat.

The electrophysiological recordings from each animal were limited to a maximum duration of 50 minutes to diminish the likelihood that experimental manipulation would result in primary sensory afferent sensitization. DRG neuronal somata were impaled with quartz micropipettes (80–250 MΩ) containing 1 M potassium acetate. DC output from an Axoclamp 2B amplifier (Axon Instruments/Molecular Devices, Sunnyvale, CA) was digitized and analyzed off-line using Spike2 (CED, Cambridge, UK). Sampling rate for intracellular recordings was 21 kHz throughout (MicroPower1401, CED).

2.4. Cell inclusion criteria

Both cells with positive mechanically sensitive RF (P-RF) and no mechanically sensitive RF (N-RF) were included. Only cells capable of generating a somatic action potential (AP) (by current somatic injection, 25- and 500-ms pulses) and with impalements stable long enough to adequately explore the full extent of the skin at the L4 dermatome (>2 min) were included. Unexcitable cells (stable impalements of more than 5 min with steady Em of −40 mV or greater but unexcitable to peripheral mechanical stimuli and intrasomal injection of current) were noted for general statistical purposes only.

2.5. Cellular classification protocol

To identify the RF, the skin was searched, applying gentle pressure with a fine-tipped brush. For nonresponding afferents, subsequent searches used increasingly stiffer probes and finally sharp-tipped watchmaker forceps. Afferents with cutaneous RFs were distinguished from those with deep RFs by displacing skin to ensure that RFs tracked rather than remained stationary. Mechanical thresholds were characterized with calibrated von Frey filaments (Stoelting, Wood Dale, IL) with the mechanical threshold (MT) being the minimum von Frey hair producing neuronal activity. The cellular classification process was performed in a sequential manner and by the combination of multiple parameters to narrow down a given afferent identity as described [9]. The result of these procedures was combined with specific cellular properties (AP shape and somatic passive characteristics) to assign every cell into one of three simplified categories: low-threshold mechanoreceptor (LTMR), A-fiber high-threshold mechanoreceptor (AHTMR), or C-fiber high-threshold mechanoreceptor (CHTMR), based on the strongest defining characteristics and to compare afferents between sham and pSNL groups and NOX procedures, innervating either glabrous or hairy skin [9]. In addition, the application of vibratory stimuli was used to assess cellular response characteristics within the RF (tuning fork of 256 and 512 Hz; SKLAR Instruments, West Chester, PA) because response to vibration with high fidelity has been shown to be 100% predictive of the LTMR population [6].

2.6. P-RF and N-RF Neurons: Somatic Electrical Properties

Active membrane properties of all excitable neurons were analyzed, including the amplitude and duration of the AP and afterhyperpolarization (AHP) of the AP, along with the maximum rates of spike depolarization and repolarization; AP and AHP durations were measured at half-amplitude (D50 and AHP50, respectively) to minimize hyperpolarization-related artifacts. Passive properties were analyzed including Em, input resistance (Ri), time constant (τ), inward rectification, and, where possible, rheobase; all but the latter were determined by injecting incremental hyperpolarizing current pulses (≤0.1 nA, 500 ms) through balanced electrodes.

The N-RF cells were separated in two different populations based on the shape of the AP [9]: neurons with inflection in the repolarizing phase (S-type neurons) and neurons without this inflection (F-type neurons). To more clearly determine the presence of this inflection, the first differential of the AP were used (presence or absence of a second additional negative component in the time course of the AP differential). Since RF properties, especially response characteristics, were used to define differences in the fast-conducting afferents (F-type neurons), the ability to more accurately define and categorize these two populations without RFs was not possible.

2.7. P-RF Neurons Only: CV

Because intact lumbar DRGs serve multiple nerves, spike latency was obtained by stimulating the RF at the skin surface using a bipolar electrode (0.5 Hz); this was performed following natural stimulation to prevent potential alterations in RF properties by electrical stimulation. All measurements were obtained using the absolute minimum intensity required to excite neurons consistently without jitter; this variability (jitter) in the AP generation latency (particularly at significantly shorter latencies), seen at traditional (i.e., two- to threefold threshold) intensity has been presumed to reflect spread to more proximal sites along axons. Any neuron with jitter was rejected (3 cells). Stimuli ranged in duration from 50 to 100 μs; utilization time was not taken into account. Conduction distances were measured for each afferent on termination of the experiment by inserting a pin through the RF (marked with ink at the time of recording) and carefully measuring the distance to the DRG along the closest nerve.

2.8. Statistical analysis

Prior to analysis, parametric assumptions were evaluated for all variables using histograms, identification of outliers with boxplots, descriptive statistics, and the Shapiro–Wilk test for normality. Data are reported as medians (range) if not normally distributed or means (standard deviation) if normally distributed. Student’s t-test and repeated measures analysis of variance (ANOVA) were used for normally distributed data and Friedman test and Mann Whitney U-test were used for not normally distributed data. Post-hoc analyses were conducted using Holm-Sidak correction for multiple comparisons across time after surgery for MCD for the 5CTV data using baseline data prior to surgery as control. For between group post-hoc analyses, comparisons were made for each time point, correcting the p-value by dividing 0.05 by the number of time points (50) for a final p-value of 0.001 required for significance. Analyses were carried out using OriginPro 9.5 (OriginLab, Northampton, MA) or Prism 6.0g (GraphPad Software, San Diego, CA). A p-value of 0.05 was considered significant unless otherwise noted above. There are no missing data or outliers to report and no animals were excluded from the study.

3.0. Results

3.1. Effects of nerve injury on performance in the 5CTV visual attention task

Median Cue Duration.

Overall a disruption of performance in the 5CTV occurred that was dependent upon time after surgery [F(49,1372)=21.7, p<0.0001] and group [F(3,28)=6.8, p=0.001], with a significant interaction between time and group [F(147,1372)=4.7, p<0.001]. Performance was significantly different between pSNL/NOX 0.2 and all other groups (Fig. 1). Using a within-subject design comparing effects of surgery to baseline values, performance in the 5CTV task for pSNL/NOX 0.2 was disrupted on days 1–4 after surgery, returning to baseline values from days 7–32, and then disrupted in a second phase from days 51–67 (p<0.05, Fig. 1). There was also an increase between at days 31–35 (p<0.05). This is likely the beginning of manifestations in the attentional disruption and was not considered a separate phase of transition, but a part of the secondary phase of persistent dysfunction. In pSNL/NOX 0.2, MCD increased 50-fold (from baseline of 0.5 ± 0.05 sec to 25 ± 2.4 sec, mean±SD) 24 hr after surgery, and this effect had partially resolved to 4.6 ± 3.6 sec by day 4 (Fig. 1) and was no different from baseline by day 5. The average MCD across days 51–67 after injury was increased 9-fold (4.5±1.3 sec) compared to presurgery baseline as well in this group (Fig. 1). Behavior was disrupted in a similar acute manner in pSNL/NOX 1.0 with the initial disruption from baseline occurring on days 1–3 (51-fold at 24 hr, 22.6±3.5 versus 0.4±0.04 sec). However, no chronic secondary phase emerged in pSNL/NOX 1.0. Behavior on the grid bar floor, pSNL/bars, was disrupted for a shorter time and to a lesser degree with impaired performance occurring for only the first 2 days after surgery (maximum 8-fold increase at 48 hr, 4.2±1.8 sec versus 0.5±0.05 sec at baseline), returning to normal thereafter and for the remainder of the study. In Sham/NOX 0.2, the rats performed the 5CTV task daily on the NOX 0.2 surface, but did not show any significant disruption of performance at any time compared to baseline values. Comparing groups within individual days and correcting for multiple comparisons, the pSNL/NOX 0.2 group was significantly different from the pSNL/bars group on days 1 and 2 and days 51–67. The pSNL/NOX 1.0 group was significantly different than the pSNL/bars group on the bar floor for days 1 and 2 only.

Figure 1.

Performance in the 5CTV attention task following partial L5 ligation (pSNL). Performance was assessed prior to and after pSNL for 5 days a week over a ten week period. Mean (SEM) are shown prior to injury (baseline, BSL) and after pSNL or sham surgery (N=8/group). *, significantly different from BSL within group, p<0.05. #, significantly different from pSNL group that performed the task on the grid bar floor, p<0.0001.

Correct response and reward latency.

Secondary measures included the latency to make a correct response and the latency to retrieve the food reward following a correct response. There was no significant main effect of group on latency to make a correct response [F(3,28)=1.6, p=0.2], however there was a significant main effect of time after surgery [F(49,1372)=11.8, p<0.0001] and a significant interaction between group and time [F(147,1372)=3.0, p<0.0001]. Post-hoc comparisons within group demonstrated a significant increase in the latency to make a correct response for pSNL/NOX 0.2 on days 1–4 (maximum at 24 hr, 12.8±1.8 sec) compared to baseline (1.06±0.05 sec), with no differences thereafter. There was a significant increase in this measure for pSNL/NOX 1.0 on days 1 (14.3±3.1 sec) and 2 (9.2±3.9 sec) compared to baseline (1.01±0.06 sec) after surgery. No other significant differences were found for the other two groups. Comparing across groups on individual days and correcting for multiple days demonstrated that the latency to make a correct response was significantly increased for both pSNL/NOX 0.2 and pSNL/NOX 1.0 groups on days 1 and 2 compared to pSNL/Bars. No other significant differences were found. There was a significant main effect of time after surgery on latency to retrieve the food reward [F(49,1372)=1.8, p=0.0008] and a significant interaction between time after surgery and group [F(147,1372=1.3, p=0.006] but no main effect of surgical group [F(3,28)=1.5, p=0.2]. Post-hoc analysis demonstrated that the latency to retrieve the food reward was increased on the first day after surgery (9.6±3.1 sec) for pSNL/NOX 0.2 compared to baseline (1.6±0.07 sec), with no other differences for this group and no differences after surgery for any other group. Comparing within each day, the latency to retrieve the food reward was significantly increased only for the pSNL/NOX 0.2 group on the first day after surgery relative to pSNL/Bars with no other differences at any time point.

Classical 5 choice serial reaction time task trial outcome measures.

The 5CTV method titrates task difficulty to the threshold of performance capability in real time based on trial outcome, and therefore the more traditional measures of performance in the classical 5 choice serial reaction time task cannot be directly compared across paradigms. However, these classical measures were collected and analyzed for comparison with data typically reported using the classical method.

Total trials completed was significantly reduced after surgery, with a main effect of surgical group [F(3,28)=6.9, p=0.01], time after surgery [F(49,1372)=11.4, p<0.0001] and a significant interaction [F(147,1372)=3.6, p<0.0001]. The total trials completed was reduced on days 1 (55.8±4.6), 2 (86.8±7.7), 35 (90.3±5.2), and 52 (89.8±5.3) after surgery for the pSNL/NOX 0.2 group compared to baseline (100±0) and on days 1 (59.9±7.1), 2 (85.3±8.7), and 65 (81.1±12.7) compared to baseline (100±0) for the pSNL/NOX 1.0 group, with no differences at any time point for the other two groups. The number of correct trials was also reduced after surgery with a significant main effect of group [F(3,28)=5.0, p=0.007], time after surgery [F(49,1372)=19.0, p<0.0001] and a significant interaction [F(147,1372)=4.0, p<0.0001]. Correct trials decreased compared to baseline (57.2±0.3) in the pSNL/NOX 0.2 group on days 1–4 (14.5±4.6 to 47.5±6.8) and days 51–57, 60, 64, and 67 (mean±S.D. across days 47.2±1.2) after surgery (p<0.05). For the pSNL/NOX 1.0 group, the number of correct trials was decreased compared to baseline (58.2±0.7) on days 1 (16.8±6.9), 2 (37.6±8.1), and 65 (46.0±7.0) after surgery. Correct trials was decreased on day 2 (49.3±4) compared to baseline (57.8±0.7) after surgery in the pSNL/Bar group and at no time after surgery in the Sham/NOX 0.2 group. Incorrect trials had no group effect [F(3,28)=2.7, p=0.06] but there was an effect of time after surgery [F(49,1372)=7.7, p<0.0001] and a significant interaction between group and time after surgery [F(147,1372)=1.7, p<0.0001]. The number of incorrect trials was decreased compared to baseline (13.0±2.5) in the pSNL/NOX 0.2 group only on days 1 (2.5±1.5) and 2 (2.0±0.6) after surgery, and in the pSNL/NOX 1.0 group only on days 1–4 (baseline 16.3±1.8, 24 hr 0.8±0.4, 4 d 6.9±1.9) after surgery. There were no differences for the pSNL/bars or sham/NOX 0.2 groups for incorrect trials compared to baseline. The number of omissions was affected over time in a similar fashion as incorrect trials [group: F(3,28)=2.9, p=0.051; time: F(49,1372)=9.1, p<0.0001; interaction: F(147,1372)=1.8, p<0.0001]. The number of omissions was increased only on day 2 after surgery for the pSNL/NOX 0.2 (baseline 29.9±2.5, 2d 44.8±0.8) and pSNL/NOX 1.0 groups (baseline 25.6±2, 24hr 42.4±1.8, 4d 39±1.7) with no differences from baseline for the other two groups. There was no significant main effect on %accuracy (baseline: 82±3% pSNL/NOX 0.2, 79±2% pSNL/NOX 1.0, 81±3% pSNL/Bars, 81±3% Sham/NOX 0.2) after surgery with respect to group [F3,28)=2.5, p=0.08], time after surgery [F(147,1372)=1.3, p=0.1] and no interaction [F(147,1372)=0.5, p=1.0].

The effect of surgery on %correct, %incorrect, and %omissions follows from the measures above and is consistent with the relative lack of effect on total trials completed other than the first week after surgery. There was a significant main effect on %correct trials for surgical group [F(3,28)=4.4, p=0.01], time after surgery [F(49,1372)=19.5, p<0.0001], and their interaction [F(147,1372)=3.5, p<0.0001]. The %correct trials was decreased on days 1–4, 35, 51, 57, and 60 after surgery for the pSNL/NOX 0.2 group and on days 1 and 2 for the pSNL/NOX 1.0 group. There were no differences at any other time for the pSNL/bars or sham/NOX 0.2 groups. There was no difference between groups for %incorrect trials [F(3,28)=2.6, p=0.07] after surgery but there was a main effect of time [F(49,1372)=7.6, p<0.0001] and a significant interaction between group and time after surgery [F(147,1372)=1.8, p<0.0001]. Individual differences from baseline were significantly different for the pSNL/NOX 0.2 and pSNL/NOX 1.0 groups as listed above for total incorrect trials. The effect on %omissions followed from the effects listed above for total omissions, with a main effect of group [F(3,28)=5.9, p=0.003], time after surgery [F(49,1372)=18.9, p<0.0001] and their interaction [F(147,1372)=2.8, p<0.0001]. The same individual differences from baseline for %omissions were the same as listed above for total omission trials for each surgical group and time point.

3.2. Electrophysiology

Twelve weeks after the pSNL or sham surgery, intracellular recordings were obtained from as many of the animals as technically possible; this included 128 sensory neurons innervating the L4 dermatome from 25 animals (Fig. 2A). Recordings were obtained in 39/128 neurons from Sham/NOX 0.2 animals (n=7) and classified as (Fig 2B): P-RF (36/39): LTMR (19/36) (6/19 hairs and 13/19 LTMR-RA), HTMR (10/36) (AHTMR: 9/10 and CHTMR: 1/10), MS (7/36) and N-RF (3/39): F-type (1/3), S-Type (2/3), UNEX (0/3). The remaining 89/128 neuronal recordings were obtained from pSNL animals (n=18) exposed to different NOX apices or the standard bar floor (A. NOX 1.0: 7/18, B. NOX 0.2: 8/18 and C. bars: 3/18 animals). The neurons from the animals who underwent these treatments were classified: (A) NOX 1.0 (26/89): P-RF (13/26): LTMR (9/13) (4/9 hairs and 5/6 LTMR-RA), HTMR (1/13) (AHTMR: 1/1 and CHTMR: 0/1), MS (3/13) and N-RF (13/26): F-type (5/13), S-type (2/13), UNEX (6/13). (B) NOX 0.2 (47/89): P-RF (25/47): LTMR (12/25) (8/12 hairs and 4/12 LTMR-RA), HTMR (7/25) (AHTMR: 6/7 and CHTMR: 1/7), MS (6/25) and N-RF (22/47): F-type (7/22), S-type (0/22), UNEX (15/47). (C) Bars (16/89): P-RF (16/16): LTMR (7/16) (3/7 hairs and 4/7 LTMR-RA), HTMR (6/16) (AHTMR: 5/6 and CHTMR: 1/6), MS (3/16) and N-FR (0/16). Due to the small number of HTMR afferents (1/13) recorded from animals pSNL/NOX 1.0, this group of afferents (A) was excluded from further analysis. Percentile representation of the above distributions is shown in Fig. 3A.

Figure 2.

A. Schematic diagram of the in vivo ePhys rat L4 preparation (lateral view) after pSNL in L5. Diagram of lateral flank illustrates the areas where the neuronal receptive field (RF) of the studied afferents were found in hairy skin and glabrous skin (lower left) (Blank circles: LTMRs; HTMR: red triangles). B. Flowchart and classification of the neurons included in the study: With mechanical RF (P-RF): low threshold mechanoreceptors (LTMR, black), high threshold mechanoreceptor (HTMR, red), muscular spindle (MS, dark gray) and without RF (N-RF): fast AP dynamic mechanically unresponsive (F-type, gray), slow AP dynamics mechanically unresponsive (S-type, light gray), unresponsive (mechanically and electrically) (UNEX, white).

Figure 3.

A. Effect of L5-pSNL and exposure to NOX surface on the L4 afferent recording population distribution (pie charts). B. Percental distribution of mechanically sensitive and insensitive afferents after L5-pSNL and exposure to NOX surface (left). Effects of the L5-pSNL and exposure to NOX surface on the mechanical sensitivity of L4 afferents (right) (*=p<0.05; **=p<0.01; ***=p<0.001). C. Representative of the nociceptive response (AHTMR) to mechanical threshold (VFH) and suprathreshold (pinch) stimulation after exposure to NOX 0.2 mm² apex in sham (Sham/NOX 0.2; upper) and L5-pSNL (pSNL/NOX 0.2; lower) animals. SA: spontaneous activity. Scale bars: 2 s, 20 mV.

3.3. Effect of the NOX surface on mechano-excitability after injury

Acute pain physiology was used to evaluate the overall effects of chronic exposure to the NOX surface on the excitability of peripheral sensory neurons, comparing the effects of exposure to NOX 0.2 between sham and pSNL animals. As shown in Fig. 3B (left panel) recorded neurons from Sham/NOX 0.2 animals show a significantly (p<0.001) higher proportion of P-RF afferents (92.3%) than pSNL/NOX 0.2 animals (53.2%). In contrast, the proportion of mechano-excitable afferents did not differ in neurons from Sham/NOX 0.2 and pSNL/Bar animals (P-RF Sham/NOX 0.2: 92.3% vs pSNL/Bar: 100%) or between pSNL/NOX 0.2 and pSNL/NOX 1.0 (P-RF pSNL/NOX 0.2: 53.2% vs pSNL/NOX 1.0: 50 %).

3.4. Neurons with receptive fields

The mechanical threshold of LTMR and HTMR neurons were significantly altered over time by pSNL and the magnitude of these effects correlated with the animals’ exposure to the NOX surface after the procedure. For LTMRs mechanical thresholds in pSNL/NOX 0.2 group significantly increased (e.g. desensitized) compared to Sham/NOX 0.2 (MT: 1.6 mN [0.2–39.2] vs 0.7 mN [0.08 −3.9], respectively). In contrast, AHTMRs showed the opposite effect, with mechanical threshold significantly decreased (e.g., sensitized) in pSNL/NOX 0.2 animals compared to pSNL/Bar animals (260 mN [100–260] vs 588 mN [260–588]; p<0.05) or Sham/NOX 0.2 (588 mN [100–1000]; P<0.01; Fig. 2B, right panel). Furthermore, 6 of 7 AHTMR afferents from pSNL/NOX 0.2 animals developed spontaneous activity (SA) after activation. This SA was not observed in any other group (Fig. 3C).

3.5. Effect of the NOX surface on cellular electrical excitability

The proportion of cells without RF (N-RF) was also altered by injury and by exposure to the NOX surface. There were very few N-RF cells in the Sham/NOX 0.2 (8%) or pSNL/Bar (0%) animals, but these were commonly found using an unbiased search strategy in pSNL/NOX 1.0 (50%) and pSNL/Bar (47%) groups (p<0.001). Given that N-RF cells have no receptive field, Em was the only parameter that could be compared with mechano-sensitivity units. Em of these cells without receptive field (F and S-type) were significantly depolarized (−51±3 mV) compared to those with receptive fields (LTMR and HTMR, −61±2 mV; p<0.01). In contrast, unexcitable cells without RF (UNEX) were significantly hyperpolarized (Em: −68±1 mV; p<0.01)). No statistical difference was found in other electrical properties of mechanosensitive afferents. A summary of their characteristics per group appears in Table 1.

Table 1.

Effect of L5-pSNL and exposure to NOX surface on somatic cellular electrical properties

								Cellular Electrical Properties
								Active
					Passive			Spike			AHP
				CV	Em	Ri	T	Amplitude	D50	MDR	MRR	Amplitude	AHP50
	NOX	Type	N	m/sec	mV	MΩ	Ms	mV	ms	dV/s	dV/s	mV	ms
pSNL	0.2	LTMR	12	24.7±4.4	−60±3	138±26	2.6±0.5	39±2	0.7±0.09	116±17	−64±8	8±1.1	4±0.8
		HTMR	7	4.9±1.5	−63±4	166±16	3.5±1.2	57±6	1.2±0.2	127±19	−69±12	12±2.7	11± 4.8
pSNL	Bars	LTMR	7	26.2±5.7	−63±3	95±3	1.8±0.2	47±4	0.6±0.06	159±23	−92±21	9±1.7	4±0.7
		HTMR	6	4.8±2.5	−55±2	145±4	1.8±1.2	48±5	1.7±0.5	95±25	−66±18	12±1	4±0.9
Sham	0.2	LTMR	19	22.6±2	−59±2	99±8	1.8±0.3	36±2	0.7±0.03	104±8	−61±3	9±0.8	3±0.4
		HTMR	10	4.9±1.7	−59±3	200±50	4±0.5	48±6	1.1±0.2	122±19	−81±8	8±1.1	5±1.1

Data are presented as ± standard error.

4.0. Discussion

The main findings in this study are that pSNL alone produces little effect on a task requiring sustained visual attention, repeated stimulation of peripheral nociceptive afferents using an engineered surface greatly increases the acute disruptive effect of pSNL on visual attention, and re-emergence of disrupted behavior is associated with electrophysiologic changes in peripheral neurons. Disrupted behavior occurs in a graded fashion, with the 0.2 mm² apex producing disruption longer than the 1.0 mm² apex in the first week, presumably due to increased nociceptive input from sensitized AHTMR afferents during performance of the operant task. The re-emergence of disrupted behavior after approximately 7 weeks (when tactile hypersensitivity is likely still present) occurs selectively in the presence of NOX 0.2 and persists for over a month. This suggests repeated and prolonged peripheral nociceptive input during the behavioral paradigm prevents normal recovery in this cognitive task and induces the development of a chronic disability state. The neuronal signature of primary afferents in animals from the pSNL/NOX 0.2 group after 12 weeks of injury with chronic nociceptive input is similar to the effects of the same surgical injury at early time points [9,11]. The alterations in the electrical signature of primary afferents after injury largely resolves over time, although the resolution is incomplete even 10–12 weeks after injury [11]. The absence of changes in peripheral afferent physiology and attentional performance in the other groups long after injury suggests that the persistent altered peripheral input to the central nervous system induced by injury coupled with the NOX 0.2 surface has important chronic behavioral consequences from an injury which otherwise has only acute effects.

Many peripheral nerve injury models demonstrate chronic mechanical hypersensitivity manifest as decreased paw withdrawal thresholds [13,28,32]. The behavioral relevance of these reflexive measures has been questioned, however there are few alternatives, particularly for repeatedly assessing the same animal over time. The latter feature is important to determine mechanisms that contribute to the transition from acute to chronic pain states and develop therapies to enhance pain resolution after surgery or injury. Operant techniques utilizing reinforced behavior are theoretically ideal for assessing development of behavioral disruption longitudinally, however the typical nerve injury models in rodents have not reliably produced robust effects in these. One limitation is that most operant behavior techniques must be conducted in isolated chambers, and experimenter interference is disruptive. Therefore, providing external nociceptive stimulation, such as application of von Frey filaments, during the operant task is problematic. Here we utilized engineered surfaces to provide graded levels of AHTMR activation, and presumably mechanical nociceptive input, by virtue of the normal body weight of the animal during performance of the operant task and in the absence of participation by the investigator. We previously demonstrated these surfaces produce avoidance selectively in animals following pSNL [9]. Thus, failure to establish acute or chronic effects of peripheral nerve injury in some operant behaviors may be from lack of sufficient nociceptive input from the injury alone. While we have previously shown that ligation of L5/L6 produces changes in intravenous opioid and intrathecal clonidine self-administration long after injury in rats, this injury does not affect operant responding for food reinforcement or intracranial self-stimulation when behavior is performed on standard bar floor surfaces [19,20,36,37].

Changes in sensory-discriminative aspects of behavior following peripheral nerve injury have been widely studied in rodents, with recent increases in the study of the affective-motivational dimension of pain-related behavior. Investigations of cognitive aspects of behavioral changes induced by peripheral nerve injury have been relatively few by comparison, however pain induced by spared nerve injury disrupts working memory or sustained attention [14,26]. Spared nerve injury disrupted performance in the classical 5 choice serial reaction time task weeks after injury, decreasing the percentage of correct responses from approximately 90% at baseline to a minimum of 80–85% at the peak of effect 2–3 months after injury [26]. This effect size is similar to the number of correct responses at 7–10 weeks after injury in the pSNL/NOX 0.2 group. The latency to make correct responses in the spared nerve injury study increased from 0.75 sec to 0.85 sec across weeks as well. These changes are similar in magnitude to many studies using the classical 5 choice method to study disruption, and rarely does one find greater behavioral effects than this by experimental manipulations.

The present study used a cue titration paradigm rather than fixed duration to ask a fundamentally different question. The 5CTV method does not ask how well a subject can perform at a given level of difficulty, but rather at what level of difficulty the subject is able perform the task. The titrating method is similar to threshold determination for mechanical paw withdrawal and determining a median nociceptive level for footshock intensity [2,15]. The dynamic range is greater than methodology using predetermined levels of input or task difficulty by adjusting the input stimulus in real time according to the behavioral responses of individual subjects. Using such a method, we can capture a wide range of disruption in performance or cognitive ability across groups or individuals within groups in real time, and can do so longitudinally in the same subject. Although the dynamic advantages of this methodology provide a benefit, the overall pattern of effect was similar to conventional methodology, disruption in the first few days after surgery that subsided and re-emerged [26]. Effects on food maintained responding using a progressive-ratio schedule were also found using conventional methodology, only in the first 1–2 days as well [26]. This is similar to the present data, in that the latency to retrieve the food reward, typically a measure in the 5 choice serial reaction time task considered to be related to reinforcement strength, was only affected in the first 1–2 days after surgery, and only in the presence of the NOX surfaces. Unlike the study by Higgins et al., however, we did not find a prolonged effect of nerve injury on attention performance in the absence of enhanced nociceptive input using NOX [26]. This may be due to a lesser injury (pSNL versus spared nerve injury) that rapidly resolves in the absence of repeated nociceptive input. Interestingly, the dynamic range of effect size in the chronic phase of disruption is significantly greater using the 5CTV method, with the MCD being increased 9-fold over baseline compared to the decrease of 5–10% in correct or %correct responses using the classical 5 choice method [26]. A larger dynamic range and effect size should theoretically pose an advantage in future mechanistic or pharmacological studies.

Nociceptive input from surgical tissue injury, inflammation, and nerve injury produce changes in attentional performance and produce profound changes in the sensibility and electrical excitability of different subtypes of mechano-sensory neurons [4,7,9,10,12,18,21,26,29,32,34,39,41,42,45,49]. Some specific neuronal characteristics develop after nerve damage: (a) sensitization of nociceptive afferents [7,9,18,34] (b) desensitization of tactile units [4,6,9,41,49] and (c) increased number of non-responsive (F and S-types) and unexcitable cells (UNEX) [9]. The present data indicate that exposure to the NOX 0.2 surface greatly sustains this neuropathic state long after the nerve injury and somehow blocks the normal recovery process and its resolution with subsequent re-emergence of behavioral dysfunction. This behavioral dysfunction is correlated with persistent abnormalities in the peripheral neuronal circuitry that was detected electrophysiologically in a blinded fashion. Interestingly, limited changes in electrophysiologic parameters occur from this NOX 0.2 surface activating AHTMR alone in the absence of pSNL. These observations indicate that activation of fast conducting peripheral afferent nociceptors and neuronal changes after injury contribute to a sustained peripheral hypersensitivity state [10]. Together, our results indicate that the NOX 0.2 surface is ideal to induce a peripheral neuropathic state after nerve injury during experimental situations where the maintenance of this condition is desired.

4.1. Potential clinical implications

Current thinking regarding rehabilitation from surgery and recovery from post-operative or post-traumatic pain is that engagement in physical activity is critical soon into the recovery process, even if such activity induces mild to moderate pain [31]. These recommendations also provide the option for administration of analgesics and multimodal analgesia during this recovery period to minimize pain and stress [25,46]. The present data suggest repeated stimulation of primary sensory nociceptive afferents after injury may produce changes that exacerbate the development of behaviorally-relevant levels of increased nociceptive input. The lack of this development, either behaviorally or with regard to altered sensory afferent physiology, in the presence of the standard grid bar floor or the NOX 1.0 surface, suggests that the amount of nociceptive drive is a critical determinant of whether these changes occur after injury. Further studies are needed, including those that incorporate standard pharmacological interventions in the perioperative period after surgical trauma, however the present experimental model may be useful for studying molecular changes that occur in primary afferents after injury with repeated nociceptive stimulation, and their behavioral consequences. The present experimental model would appear to be a unique resource to understand the molecular mechanisms involved in the transition from acute to chronic pain state.