MOUSE MODELS FOR STUDYING PAIN IN SICKLE DISEASE: EFFECTS OF STRAIN, AGE, AND ACUTENESS

Abstract

Clinical management of severe pain associated with sickle cell disease (SCD) remains challenging. Development of optimal therapy would be facilitated by use of murine model(s) with varying degrees of sickling and pain tests that are most sensitive to vasoocclusion. We found that young (≤3 month old) NY1DD and S+S^Antilles mice (having modest and moderate sickle phenotype, respectively) exhibit evidence of deep tissue/musculoskeletal pain. Deep tissue pain and cold sensitivity in S+S^Antilles mice increased significantly with both age and incitement of hypoxia/reoxygenation (H/R). C57/BL6 mice (genetic background strain of NY1DD and S+S^Antilles) were hypersensitive to mechanical and heat stimuli, even without the sickle transgene. H/R treatment of HbSS-BERK mice with severe sickle phenotype resulted in significantly decreased withdrawal thresholds and enhanced mechanical, thermal and deep tissue hyperalgesia. Deep hyperalgesia incited by H/R in HbSS-BERK was ameliorated by CP 55940, a cannabinoid receptor agonist. Thus, assessment of deep tissue pain appears to be the most sensitive measure for studying pain mechanisms across mouse models of SCD, and HbSS-BERK mice may best model vaso-occlusive and chronic pain of SCD.

INTRODUCTION

Sickle cell disease (SCD) is associated with both unpredictable recurrent acute vasoocclusive pain episodes (“crises”) and chronic pain (NIH Publications, 2002; Dampier et al, 2002; Ballas et al, 2007; Smith et al, 2011). Opioid treatment remains a suboptimal approach due to side effects and altered clearance in SCD (Dampier et al, 1995; Darbari et al, 2011). Identification of underlying mechanisms and development of more effective analgesic control will be facilitated by animal models for the chronic and vaso-occlusive pain of human SCD.

We here sought to identify the most suitable mouse model(s) and appropriate pain measures that best represent characteristics of acute and chronic pain in human SCD. We previously found that HbSS-BERK mice exhibit tonic hyperalgesia (Kohli et al, 2010). We now characterize pain before and following incited vasoocclusion in mice having mild (NY1DD), medium (S+S^Antilles) or severe (HbSS-BERK) phenotypes (collectively referred to as ‘sickle’ mice henceforth) (Fabry et al, 1995; Paszty et al, 1997; Solovey et al, 2004). We also evaluated the effect of the models’ different genetic backgrounds on pain behaviors. To distinguish acute from chronic baseline pain, we incited vasoocclusion with hypoxia/reoxygenation (H/R) (Solovey et al, 2004). In sickle mice this induces acute sickling (Osarogiagbon et al, 2000), vasoocclusion (Kalambur et al, 2004) and an accompanying state simulating human ischemic/reperfusion (I/R) (Carden et al, 2000); endothelial activation (Solovey et al, 2004), increased leukocyte count, leukocyte/endothelial interaction, and emigration across endothelium (a hallmark of inflammation) (Kaul et al, 2000; Kaul et al, 2004); conversion of xanthine dehydrogenase to xanthine oxidase (Osarogiagbon et al, 2000); and activation of NFκB (Kaul et al, 2004; Kollander et al, 2010). None of these events develop in normal mice exposed to H/R.

Recently cannabinoids have been found protective against ischemia reperfusion injury and as analgesics to treat severe pain in experimental studies (Khasabova et al, 2008; Elikkottil et al, 2009; Tuma and Steffens, 2011). We observed that CP 55940, a cannabinoid receptor agonist that binds both CB1 and CB2 receptors, attenuated tonic and complete Freund adjuvant (CFA)-induced hyperalgesia in sickle mice (Kohli et al, 2010). It is therefore likely that cannabinoids may be effective in treating pain and also have a protective effect on ischemia reperfusion injury in SCD.

We observed variability in sensitivity to pain stimuli between different sickle mice as compared to respective control mice. Hypoxia/reoxygenation further exacerbated hyperalgesia in sickle mice, with variable sensitivity to different pain tests used. However, all sickle mice pre- and post-hypoxia/reoxygenation and irrespective of background strain uniformly exhibited sensitivity to grip force test that is an indicator of deep tissue/musculoskeletal pain. To confirm that grip force alterations are due to a nociceptive response, we examined the analgesic effect of cannabinoid, CP 55940 on H/R-induced changes in grip force measurement.

METHODS

All experiments were approved by the University of Minnesota Animal Care Committee.

Mice

Three types of transgenic mice expressing sickle hemoglobin (HbS) and their respective background controls were used.

NY1DD sickle mice: NY1DD transgenic mice have linked human α and β^S globin (αHβS) genes on a mouse β^major D/D (D indicating natural deletion) background (Fabry et al, 1992). These mice, express ~26% HbS, and normal hemoglobin (Hb) and reticulocyte levels, and therefore, exhibit a mild SCD phenotype. These animals exhibit inflammation, and endothelial activation, which are further enhanced by hypoxia/reoxygenation (H/R) (Solovey et al, 2004; Belcher et al, 2003). These mice have not been characterized for pain behaviors. However, these mice were suggested to have enhanced nociception on the basis of facilitation of tail flick reflex (Lunzer et al, 2007).

S+S^Antilles mice: The S+S^Antilles mice are homozygous for the natural deletion of mouse β-major globin locus and transgenic expression of human α-, β^S-, and β^S-Antilles globin transgenes, with Antilles globin having a second mutation at β²³ position (valine to isoleucine) in addition to the β^S mutation on position β⁶ (Fabry et al, 1995). These mice express approximately 42% of human β^S and 36% of β^S-Antilles and show vascular pathology and inflammation which are further exacerbated by hypoxia (Belcher et al, 2003; Kalambur et al, 2004). Pain has not been analyzed in these mice.

HbSS-BERK mice: These mice are homozygous for knockout of both murine αand β globins and carry a single copy of the linked transgenes for human αand β^S globins. Therefore, HbSS-BERK mice express human α and β^S globin chains with ~99% human HbS, but no murine αor β globins (Paszty et al, 1997). These mice exhibit cutaneous and deep tissue hyperalgesia (Kohli et al, 2010).

Control mice: NY1DD and S+S^Antilles mice are on C57/BL6 background. Therefore C57/BL6 mice from the same colony were used as controls. HbSS-BERK mice are from a mixed genetic background. Thus, similar littermate mixed genetic strain background mice expressing normal human hemoglobin, HbAA-BERK were used as control for HbSS-BERK mice, from the same breeding.

All mice were bred, genotyped and phenotyped as described (Kohli et al, 2010; Solovey et al, 2004). Mean ages of young mice were: C57/BL6, 2.2±0.2 mo.; NY1DD, 2.6±0.2 mo.; S+S^Antilles 3.0±0 mo. Old mice were: C57/BL6, 6.5±0 mo.; NY1DD, 5.3±0 mo.; S+S^Antilles, 7.8±0.8 mo. HbSS-BERK were 5.7±0.7 mo., and HbAA-BERK, 5.5±0.5 mo. All experiments were performed on age- and sex-matched sickle and control (C57/BL6) mice after appropriate approvals from the Institutional Animal Care and Use Committee at the University of Minnesota.

Hypoxia/reoxygenation

Following baseline pain measurements, mice were exposed to hypoxia (H) with 8% O₂ and 92% N₂ for 3h in a chamber maintained at 24°C, followed by re-oxygenation (R) with room air, using a previously standardized procedure (Solovey et al, 2004). Pain measures were obtained during re-oxygenation, after 1–2h and 18h of re-oxygenation (HR1) as described in Table 1. This was repeated after 24h, followed by pain measures after 1–2h, 18h and 7d of reoxygenation (HR2). Hypoxia induces approximately 10-fold increase in sickle RBC in sickle mice but not in normal mice (Osarogiagbon et al, 2000).

Table 1.

Experimental Design for Hypoxia/Reoxygenation treatment

Group	Treatment	Reoxygenation at room air
H/R1	3 h of hypoxia on day 1	1–2 h pain measurement	18 h pain measurement
H/R2	3 h of hypoxia on day 2	1–2 h pain measurement	18 h pain measurement
Day 7*	none	pain measurement

^*Baseline pain measures were obtained before 1^st hypoxia treatment.

Pain behaviors

All behavioral tests were performed after mice were habituated to each testing apparatus in a controlled environment prior to testing as described by us earlier (Kohli et al, 2010). Order of analysis was critically followed post-hypoxia to enable us to allow sufficient acclimatization time before each test. Therefore, mice were tested after 1–2h of reoxygenation following hypoxia in this order: 1. Mechanical sensitivity. 2. Sensitivity to heat, 3. Grip force. 4. Cold sensitivity.

Withdrawal responses to mechanical stimuli

Mice were placed in glass enclosures (10 × 6.5 × 6.5 cm) on an elevated wire mesh floor. The mid-plantar surface of the hind paws was stimulated using von Frey monofilaments (Stoelting Co, Wood Dale, IL) with calibrated bending forces until the mouse withdrew its paw in response to the stimulus. Paw withdrawal threshold was determined using the up-down paradigm described by Chaplan et al. (1994) and modified by Hamamoto et al. (2007) for mice. Testing was initiated with the 6.0 mN (0.61 g) von Frey filament. Subsequent stimuli were presented based on the response to the previous stimulus. In the absence of a paw withdrawal to a given stimulus, a stronger stimulus was presented. In the case of a paw withdrawal, a weaker stimulus was presented. Six responses, starting with the negative response immediately before the first paw withdrawal, were recorded. The resulting pattern was tabulated and the 50% paw withdrawal threshold was calculated using the formula:

$$50\%\mathrm{g}\text{threshold}={10}^{({\mathrm{X}}_{\mathrm{f}}+\mathrm{\kappa }\mathrm{\delta })}/10,000$$

Where X_f = value (in log units) of the final von Frey filament used; κ = tabular value for the pattern of positive/negative responses; and δ = mean difference (in log units) between stimuli (here, 0.223).

The frequency (%) of paw withdrawal was obtained in response to 10 applications of a 9.8 mN (1.0 gram) filament. For each trial, the filament was applied for 1–2 s with an inter-stimulus interval of approximately 5 s. Only vigorous withdrawals were used to calculate paw withdrawal frequency (PWF).

Withdrawal responses to heat

Heat hyperalgesia was examined by determining paw withdrawal latency to a radiant heat source (Hargreaves et al, 1988). Mice were placed on a temperature controlled (30°C) glass platform, covered with ventilated glass containers and allowed to habituate for at least 30 minutes. Radiant heat was applied to the middle of the plantar surface of the hind paw and paw withdrawal latency (PWL) was determined. Testing consisted of 4 trials for each hind paw with each trial separated by at least 5 min. Mean paw withdrawal latency was calculated from the last 3 trials. The intensity of the heat source was adjusted so that, at baseline, mice withdrew their hind paws at ~9 s. A cut off time of 16 s was chosen to avoid tissue damage.

Grip force

To evaluate deep tissue hyperalgesia, the tensile force of peak forelimb exertion was measured using a computerized grip force meter. During testing each mouse was held by its tail and gently passed over a wire mesh grid and allowed to grip the wires. The peak force exerted against the transducer was recorded in grams. Final data represent the mean from three trials.

Withdrawal responses to cold

Mice were placed on a cold plate (4°C) and the latency (s) to initial lifting of either forepaw (paw withdrawal latency); and the number of forepaw withdrawals (paw withdrawal frequency) during a period of 2 minutes on the cold plate.

Treatment with cannabinoid receptor agonist following H/R

After baseline grip force measurements, HbSS-BERK mice were exposed to 3h of hypoxia and 1h of reoxygenation as described above under ‘Hypoxia/reoxygenation’. Grip force was measured again after H/R treatment. Half of the mice were then injected intraperitoneally (i.p.) with vehicle (2% DMSO) and the other half with 0.3 mg/Kg (-)-cis-3-[2-hydroxy-4(1,1-dimethylheptyl)phenyl]-trans-4-(3-hydroxypropyl) cyclohexanol (CP 55940; Tocris Bioscience, Ellisville, MO) as described by us earlier (Kohli et al, 2010). We selected 0.3 mg/Kg dose of CP 55940 because this dose was effective in ameliorating hyperalgesia and in HbSS-BERK mice without producing catalepsy. Grip force was measured 30, 60, 90 and 180 min, and 24h after the injections.

Bar test

To evaluate the possibility of catalepsy due to CP 55940, the forepaws of mice were placed on an elevated bar of 1 cm diameter, while the hindpaws rested on the counter, as described by us earlier (Kohli et al, 2010). Catalepsy was determined as an increase in time spent by mice on the bar.

Statistical Analyses

The results are presented as mean with standard error of the mean. For age-related results (Figure 1) among S+S^Antilles, NY1DD, and C57/BL6 mice, all data were compared using parametric analyses including 1-way analysis of variance for within- or between- group comparisons and unpaired t-test. Multiple comparisons after analysis of variance were performed using a Tukey’s test. For hypoxia/reoxygenation results (Figure 2), all data were compared first by 2-way analysis of variance for between-group comparisons over time and multiple comparisons after analysis using Tukey’s post hoc testing. For within-group comparisons over time, all data were compared using 1-way analysis of variance and multiple comparisons after analysis were completed using Tukey’s post hoc testing. For all data, P < .05 was considered significant. Data were analyzed using Prism software (Version 5.04; GraphPad, San Diego, CA).

Figure 1. Pain behaviors in young and old NY1DD and S+S^Antilles and HbSS-BERK mice.

Sensitivity to nociceptive stimuli is shown in different groups of mice comprising, young and old, NY1DD and S+S^Antilles mice (sickle mice) and their age-matched control C57/BL6 mice; and HbSS-BERK (severe sickle phenotype) and their age-matched control HbAA-BERK mice. Mean age in months ± SEM for, ‘Young’ mice: C57/BL6, 2.2 ± 0.2; NY1DD, 2.6 ± 0.2; S+S^Antilles, 3.0 ± 0.0, and ‘Old’ mice, C57/BL6, 6.5 ± 0.0; NY1DD, 5.3 ± 0.2; S+S^Antilles, 7.8 ± 0.8; HbAA-BERK (control), 5.5 + 0.5 and HbSS-BERK, 5.7 + 0.7. Comparisons were made between age-matched, NY1DD and C57/BL6; between S+S^Antilles and C57/BL6; and S+S^Antilles and NY1DD and differences in nociceptive sensitivity due to age were also compared between ‘young’ and ‘old’ mice (A–F). (A) Mechanical threshold using von Frey filaments, measure of cutaneous nociception is shown. (B) Paw withdrawal frequency (PWF) in response to 10 applications of a 9.8 mN (1.0 g) von Frey monofilament in the same mice. A higher PWF indicates increased nociception. No statistically significant difference was observed between age-matched sickle and C57/BL6 mice or between younger and older mice. (C) Grip force measurements indicating deep tissue/musculoskeletal pain are shown. Lower grip force suggests increased deep tissue hyperalgesia. Both types of sickle mice show decreased grip force compared to age-matched C57/BL6 ‘young’ NY1DD mice compared to ‘young’ S+S^Antilles mice and ‘old’ S+SAntilles show decreased grip force as compared to ‘young’ S+S^Antilles. (D) Paw withdrawal latency (PWL) on the cold plate in ‘seconds’ when the mouse lifted its paw after being placed on the cold plate maintained at 4°C, is shown. ‘Old’ S+S^Antilles mice show significantly decreased PWL (increased cold sensitivity) as compared to age-matched NY1DD and C57/BL6 and ‘young’ S+S^Antilles mice. (E) Paw withdrawal frequency (PWF) on a cold plate at 4°C in a 2 min period is shown. Old S+S^Antilles mice demonstrate significantly higher PWF (increased cold sensitivity) than young S+S^Antilles, and age-matched C67/BL6 and NY1DD. (F) PWL in seconds in response to a heat stimulus targeted to the intraplantar surface of the hind paw is shown. Of note, NY1DD show significantly less sensitivity to heat (higher PWL). (G–L) Sensitivity to different nociceptive stimuli in HbSS-BERK (sickle) as compared to HbAA-BERK. Please note that HbSS-BERK show significantly high sensitivity to all nociceptive stimuli as compared to HbAA-BERK. Significance (A–L), *P<.05, **P<.01 and ***P<.001, Each data point is the mean + SEM of measurements from 7–13 different mice. BW, body weight.

Figure 2. Comparisons of pain behaviors between different strains of control mice.

Columns in each graph represent the mean ± SEM of C57/BL6 (left) and HbAA-BERK (right) mice. (A) Mechanical withdrawal thresholds of C57/BL6 mice were significantly lower (*P = .0012), i.e., more sensitive, than HbAA-BERK mice. (B) No differences were observed between the two types of mice for PWF to von Frey filaments. (C) Weight-normalized grip force. (D) PWL on a cold plate maintained at 4°C. (E) PWF on a cold plate at 4°C over a 2 min period. (F) PWL to heat was lower (more sensitive) in C57/BL6 compared to HbAA-BERK mice (^$P = .044). Mouse ages (mean ± SEM): C57/BL6 = 6.0 months ± 0.6; HbAA-BERK mice = 5.5 months ± 0.5. n= 7–10 mice of each type. Statistical analyses were conducted using unpaired t-tests. BW, body weight; PWL, paw withdrawal latency; PWF, paw withdrawal frequency.

RESULTS

Response to painful stimuli in different sickle mice

HbSS-BERK mice, whose phenotype is the most severe for sickling and vaso-occulsion, exhibited the greatest degree of hyperalgesia on all nociceptive measures tested. The magnitude of hyperalgesia in these mice increased with age. Deep hyperalgesia (measured by decreased grip force) occurred early in life, even with modest sickle hemoglobin levels in NY1DD and S+S^Antilles mice (Figure 1C), which did not exhibit mechanical or heat sensitivity (Figures 1A, B and F). S+S^Antilles mice were particularly sensitive to cold and this increased with age (Figures 1D and E). NY1DD and S+S^Antilles were less sensitive to heat and mechanical stimuli than HbSS-BERK mice. Thus, different strains of mice exhibit different degrees of hyperalgesia to various sensory modalities. Grip force appears to be the most sensitive measure for detecting hyperalgesia in all of the genetic strains used, however, HbSS-BERK mice displayed the greatest degree of hyperalgesia across all pain tests.

Influence of genetic strain on pain measures in sickle mice

It is well known that mice exhibit strain-dependent sensitivities in behavioral measures of nociception (Mogil et al, 1999). Since, the transgenic sickle mice used in this study were obtained from two different strains of mice (C57/BL6 for NY1DD and S+S^Antilles and HbAA-BERK for HbSS-BERK), we compared behavioral measures of pain in these different strains of control mice to determine if enhanced sensitivity to any of the pain measures observed in the sickle mice could be attributed to differences in sensitivity of the background controls or to the sickle phenotype. As shown in Fig. 2, C57/BL6 mice exhibited a decrease in paw withdrawal threshold to mechanical stimuli and a decrease in paw withdrawal latency to heat as compared to the HbAA-BERK control mice. Although we do not believe that the increased sensitivity of sickle mice is due to strain differences (due to the reasons described below), these results underscore the need to compare background strains of mice so that behavioral measures of pain can be carefully interpreted.

Influence of hypoxia/reoxygenation on mechanical and deep tissue pain

Mechanical withdrawal threshold obtained using von Frey monofilaments decreased after 1h of H/R1 and remained decreased until 18h after reoxygenation in HbSS-BERK as compared to their baseline and to HbAA-BERK mice, but not in S+S^Antilles mice (Figures 3A and B). H/R2 caused a further decrease in threshold in HbSS-BERK. PWF evoked by the standard von Frey filament did not change after H/R1 but increased following H/R2, returning back to baseline after 18h in HbS-BERK (Figure 3C). A decrease in withdrawal threshold occurred in HbAA-BERK following 18h of H/R2 (Figure 3A) but not after H/R1.

Figure 3. Hypoxia/reoxygenation-induced pain behaviors in sickle mice.

Following baseline pain measurements, mice were exposed to first hypoxia treatment of 3h, followed by pain measures after 1–2h and 18h of re-oxygenation (HR1). This was repeated after 24h, followed by pain measures after 1–2h, 18h and 7d of reoxygenation (HR2). Different measures of pain are shown for sickle, S+S^Antilles and HbSS-BERK and their control C57/BL6 and HbAA-BERK, respectively. Data are shown as mean ± SEM. Comparisons were made between control and sickle mice following each treatment and significance is indicated with [*] sign. Within each mouse changes in nociception following treatments were compared to baseline (BL) and significance is shown with the # sign. (A and B) Paw withdrawal threshold to von Frey monofilaments are shown. Lower thresholds are indicative of increased sensitivity to cutaneous nociception. HbSS-BERK compared to HbAA-BERK: *P<.05, **P<.01, ***P<.001, ****P<.0001; significant differences from initial baseline value: ^#P<.05. (C and D). PWF with 9.8 mN (1.0 g) von Frey monofilament is shown. Higher PWF indicates increased nociception. In (C), *P<.01, **P<.001, ***P<.0001, HbSS-BERK compared to HbAA-BERK; ^#P<.05, compared to baseline (D) ^#P<.05 and ^##P<.001, compared to baseline in C57/BL6 and S+S^Antilles. (E and F) Grip force measurements indicative of deep tissue/musculoskeletal pain are shown. Lower grip force is suggestive of increased deep pain. In (E) *P<.05, **P<.001, ***P<.0001 as compared to HbAA-BERK and #P<.05, compared to baseline. In (F) *P<.05, compared to C57/BL6 and ^#P<.05 and ^##P<.001 compared to baseline. (G and H) Paw withdrawal latency (PWL) on the cold plate in ‘seconds’ when the mouse lifted its paw after being placed on the cold plate maintained at 4°C, is shown. In (G) *P<.05, **P<.01, ***P<.0001, compared to HbAA-BERK and ^#P<.05, as compared to baseline. In (H) *P<.05, as compared to C57/BL6 and ^#P<.05, ^##P <.01 compared to baseline. (I and J) Paw withdrawal frequency (PWF) on a cold plate at 4°C in a 2 min period is shown. In (I) *P<.01, **P<.001, compared to HbAA-BERK and ^#P<.01, ^##P<.001, compared to baseline. In (J) *P<.05, **P<.01, compared to C57/BL6; ^#P<.001 compared to baseline. (K and L) PWL in seconds in response to a heat stimulus targeted to the intraplantar surface of the hind paw is shown. Shorter latency indicates increased sensitivity to heat. In (K) *P<.0001, compared to HbAA-BERK and ^#P<.05, ^##P<.01, ^###P<.001 compared to baseline. In (L) *P<.01, **P<.0001 compared to C57/BL6. All data were obtained from 5.7 ± 0.7 mo old HbSS-BERK (open squares), 5.5 ± 0.5 mo old HbAA-BERK (solid squares), 5.6 ± 0.1 mo old S+S^Antilles (open triangles) and 6.0 ± 0.6 mo old C57/BL6 (solid triangles) mice. N = 7–10 per type of mice. BW, body weight.

S+S^Antilles and C57/BL6 showed a parallel increase in response to the 1.0 g von Frey filament following both H/R treatments (Figure 3D). Of note, C57/BL6 exhibited a lower withdrawal threshold than HbAA-BERK (Figure 2). Thus, responsiveness of rodents to hypoxia and its effects on mechanical sensitivity in sickle mice may in part be dependent on genetic strain.

Grip force decreased after H/R2 in HbSS-BERK and S+S^Antilles mice, but not in their respective controls (Figures 3E and F). These data demonstrate that sickling, vasoocclusion and ischemia-reperfusion injury induced by H/R possibly lead to deep tissue/musculoskeletal pain in sickle mice, and, therefore, control mice lacking sickle hemoglobin do not show increased sensitivity to grip force following H/R. Thus, grip force, which is a measure of deep pain, appears to be a sensitive measure of pain due to recurrent vasoocclusion in SCD.

Thermal sensitivity in response to hypoxia/reoxygenation

HbSS-BERK, S+S^Antilles and their respective controls HbAA-BERK and C57/BL6, respectively, exhibited a significant increase in cold sensitivity following H/R treatment (Figures 3G–J). However, neither the HbAA-BERK or C57/BL6 or S+S^Antilles showed increased sensitivity to heat, whereas, PWL to heat decreased significantly in HbSS-BERK following both H/R treatments (Figures 3 K and L). It is noteworthy, that C57/BL6 mice demonstrated greater hyperalgesia to heat than HbAA-BERK (Figure 2) and NY1DD (Figure 1F) mice.

Thus, HbSS-BERK exhibited increased thermal sensitivity following H/R, but inclusion of strain-specific control is essential to distinguish between the effects of H/R specifically versus strain-related characteristics.

The cannabinoid rceptor agonist, CP 55940 ameliorates deep hyperalgesia following hypoxia reoxygenation

HbSS-BERK mice injected with CP 55940 following H/R, showed a significant increase in grip force after 60 min of injection as compared to the grip force immediately after H/R (Figure 4). Grip force continued to increase in CP 55940 treated mice up to 24h (last period of observation) as compared to H/R and returned to baseline at this time point. In contrast, grip force did not increase in vehicle treated mice following H/R at any time point. Grip force remained significantly decreased in vehicle treated mice up to 24h as compared to baseline (before H/R), suggestive of sustained deep tissue pain incited by H/R. Therefore, a significant sustained increase in grip force from 60 min to 24 h after CP 55940 treatment suggests that this cannabinoid receptor agonist ameliorates deep tissue pain incited by H/R in HbSS-BERK mice.

Figure 4.

The cannabinoid receptor agonist CP 55940 reduces deep hyperalgesia following hypoxia reoxygenation. Following baseline grip force measurements, HbSS-BERK mice were exposed to 3h of hypoxia treatment and 1h of oxygenation at room air (H/R), and grip force was measured again. Mice were then injected with vehicle or CP 55940 and grip force was measured at indicated time on the graph. Hollow and solid circles represent vehicle and CP 55940 treatment, respectively. Data are shown as mean + SEM from 5 mice per treatment. Statistical significance was calculated by comparing each value with BL ($) or with H/R (*); and between vehicle and CP 55940 treatment (#). ^$P<0.05 and ^$$P<0.005, as compared to BL; *P<0.05 and **P<0.005 as compared to H/R; and ^#P<0.05 and ^##P<0.005, as compared to vehicle for that time point. Mean age of mice + SEM in months were, Vehicle, 12.25 + 0.8 and CP 55940, 11.7 + 0.7. Abbreviations, BL, baseline; H/R, hypoxia reoxygenation; Veh, vehicle; CP, CP 55940.

Importantly, mice injected with CP 55940 did not exhibit catalepsy. There was no significant difference in the time spent on the bar between CP 55940 and vehicle injected mice at any time points (data not shown), similar to our earlier observations (Kohli et al, 2010). Thus, the antinociceptive effects of CP 55940 were not due to impaired motor function.

DISCUSSION

In the present study, we studied three different transgenic mouse models expressing sickle hemoglobin to determine which model exhibits the greatest changes in nocifensive behaviors compared to control mice, and which behavioral measures are most sensitive for detecting hyperalgesia under normal conditions and following hypoxia/reperfusion incitement. First, we found that HbSS-BERK mice developed the greatest degree of hyperalgesia across all pain tests. This is consistent with our previous study (Kohli et al, 2010) showing that these mice developed age-dependent hyperalgesia to mechanical, heat and cold stimuli, and deep tissue hyperalgesia as indicated by a decrease in grip force. Of the models used in the present study, HbSS-BERK mice exhibit the greatest severity of hematologic disease. Second, grip force was the most sensitive measure for detecting hyperalgesia across different sickle mice and following hypoxia/reoxygenation. Thus, the occurrence of deep pain may be an indicator of ongoing sickling events, vasoocclusion and ischemic injury.

NY1DD and S+S^Antilles appear to be less sensitive to heat and mechanical stimuli and this may be due to their C57/BL6 background. C57/BL6 mice normally exhibit increased sensitivity to these stimuli (Mogil et al, 1999). Although C57/BL6 mice are innately hypersensitive, introducing sickle hemoglobin, as in the S+S^Antilles mice, on the C57/BL6 background increases cold sensitivity with age, perhaps due to recurrent H/R injury. Indeed, of 12 strains of mice tested, C57/BL6 was most sensitive to several nociceptive tests and therefore suggested to be least appropriate for studying nociception in genetically modified disease models (Mogil et al, 1999).

HbSS-BERK mice exhibited the greatest changes in pain sensitivity to different stimuli following hypoxia/reoxygenation over a period of time. Our results are consistent with a recent study that measured mechanical sensitivity sensitivity in HbSS-BERK mice following 3 h of hypoxia (Hillery et al, 2011). In our study recurrent exposure to H/R further decreased mechanical sensitivity, suggesting that each episode of H/R exposure results in progression of pain. In addition, H/R also increased deep tissue hyperalgesia and heat sensitivity in HbSS-BERK. The increase in hyperalgesia after recurrent hypoxia/reperfusion, particularly mechanical hyperalgesia, may be due to sensitization of nociceptors since hypoxia was shown to lower mechanical thresholds of most Cnociceptors (Fuchs et al, 2010). Grip force, on the other hand appears to demonstrate pain specific to sickling and therefore reperfusion injury following hypoxia (discussed in detail below).

We also observed increased sensitivity to heat in HbSS-BERK in response to H/R which persisted for 24h following both, the first and repeat exposure. The enhancement of increased pain sensitivity in HbSS-BERK mice beyond the existent chronic pain, suggest that repeated episodes of H/R injury, perhaps lead to ongoing chronic pain and difficult to treat episodes of “acute” pain following “crises.” H/R is known to result in increased inflammatory response, vasocclusion and ischemic injury (Solovey et al, 2004; Osarogiagbon et al, 2000; Kalambur et al, 2004; Carden et al, 2000; Kaul et al, 2000; Kaul et al, 2004), and both of them occurring collectively and repeatedly, may underlie severe pain in SCD.

The mechanisms underlying enhanced pain sensitivity in SCD are unclear. We have shown that BERK mice exhibited a variety of structural and neurochemical changes in the periphery and spinal cord (Kohli et al, 2010). For example, the skin of these mice was less innervated, suggesting peripheral neuropathy and neuropathic pain perhaps due to repeated vasoocclusive injury. The presence of nerve injury is supported by the thinner epidermis, which occurs following degeneration of motor and sensory nerves (Li et al, 1997). Also, lymphatic vessels exhibited abnormal structure in HbSS-BERK mice. Since lymphatic vessels are involved in immune cell trafficking, disruption of these vessels may contribute to persistent inflammation. The neuropeptides, CGRP and SP are increased in persistent and inflammatory pain states (Honore et al, 2000) and were also increased in the skin of HbSS-BERK mice (Kohli et al, 2010). Increased expression of CGRP and SP in the skin is suggestive of neurogenic inflammation and sensitization of nociceptors. Indeed, it has recently been shown that cutaneous nociceptors in HbSS-BERK mice exhibited sensitization to mechanical stimuli and this was mediated, at least in part, by TRPV1 receptors (Hillery et al, 2011). Furthermore, hypoxia itself activates TRPV1 activity via HIF-1α in sensory neurons of native rats (Ristoiu et al, 2011), and this may lead to additional nociceptor sensitization and sustained ongoing pain. It is unknown whether sensitization of nociceptors contributes to heat and cold hyperalgesia, and deep tissue hyperalgesia, exhibited by these mice.

HbSS-BERK demonstrated tonic deep tissue hyperalgesia measured using grip force test (Kohli et al, 2010). No other study has measured grip force in these mice. The observations herein that grip force was uniformly significantly lower in all three types of young as well as old sickle transgenics tested as compared to their respective age-matched controls, is suggestive of the specificity of grip force test to deep tissue injury and pain similar to musculoskeletal pain experienced by patients with SCD. It is likely due to ongoing episodes of sickling and the resultant reperfusion injury, which may start early in life and may be associated with avascular necrosis of bone and infarction of the bone marrow observed in SCD (Ballas et al, 2007; Ganguly et al, 2011). A decrease in grip force/increased deep tissue pain upon repeated H/R in HbSS-BERK and S+S^Antilles in this study suggest that this may be true. Notably, C57/BL6 mice do not show tonic sensitivity to grip force as compared to HbAA-BERK control mice. Repeated H/R incitement did not alter grip force response in C57/BL6, but altered their sensitivity to mechanical and cold hyperalgesia, suggesting that grip force sensitivity is a response specific to sickle animals. Recent elegant studies on HbSS-BERK sickle mice have demonstrated that inflammatory response mediated by E-selectin promotes leukocyte-sRBC interaction and vasoocclusion (Hidalgo et al, 2009). Therefore, pain in response to vasooclusion is unique to SCD, requiring robust and specific measures of pain. Together, these data suggest that changes in grip force may be more specific as a change in pain due to sickling and associated pathology in sickle mice.

Our earlier studies (Kohli et al, 2010) also suggested that central sensitization contributes to hyperalgesia in SCD. For example, we observed higher expression of IL-6, COX-2, and TLR4 in the spinal cord of HbSS-BERK mice as compared to controls. Moreover, activation of several mediators of central sensitization, including MAPK/ERK, p38 MAPK, and STAT3 were also increased in the spinal cord of HbSS-BERK mice. Electrophysiological studies of nociceptive spinal neurons are needed to determine whether their response properties and encoding of natural stimuli are altered in these mice.

Decreased grip force following the incitement of H/R in HbSS-BERK mice increases in response to cannabinoid treatment, suggesting that decreased grip force may be due to deep tissue pain and that CP 55940 has an analgesic effect on deep tissue pain in sickle mice. This is in agreement with our earlier studies demonstrating that CP 55940 reduced tonic and CFA-induced deep tissue hyperalgesia in HbSS-BERK mice (Kohli et al, 2010). In addition to having an analgesic effect, cannabinoids may provide a therapeutic benefit following ischemia-reperfusion injury. Experimental studies suggest a protective effect of cannabinoids and endocannabinoid system on vascular insult following ischemia-reperfusion injury (Tuma and Steffens, 2011). Cannabinoids, particularly CB2 receptor agonists have been shown to reduce inflammation, oxidative stress and membrane damage and enhance blood flow, which may provide a therapeutic advantage in preventing vascular injury due to vasoocclusion in SCD. Increased grip force up to 24h following CP 55940 injection may be due to the prevention of H/R-induced injury, in addition to the analgesic effect. However, the initial increase in grip force following the first 60 min after CP 55940 injection appears to be primarily due to the analgesic effect and/or reduction in H/R-induced injury. It is likely that cannabinoids may be effective in producing analgesia in severe pain due to vasooccluion during “crises.” Opioids are currently the mainstay to treat severe pain in SCD and may not be effective all the time (NIH Publications, 2002; Ballas et al, 2007). Moreover, high doses of opioids are required to treat pain in SCD as compared to treating analogous pain in other conditions. Taken together with the analgesic effect of CP 55940 in attenuating H/R-induced nociception, cannabinoids may be potentially effective in treating severe pain in SCD. Furthermore, the role of cannabinoids in ameliorating H/R-induced pathobiology, inflammation and oxidative stress in SCD appears promising and requires further investigation.

We conclude that murine sickle models vary in their usefulness for examining mechanisms and targets to treat pain in human SCD. HbSS-BERK appear to be the most appropriate mouse model to study pain mechanisms because of the severity of hematologic disease and lower nociceptive sensitivity of their genetic background (HbAA-BERK) as compared to S+S^Antilles with moderate hematologic disease on C57/BL6 strain with nociceptive hypersensitivity. The grip force test, which measures deep tissue/musculoskeletal pain, appears to be more sensitive and specific to sickle state in mice. CP 55940 ameliorated deep pain incited by H/R suggesting that cannabinoids may be useful as analgesics to treat severe pain in SCD. This study provides the rationale to consider clinical trials to evaluate deep pain and the analgesic potential of cannabinoids to treat pain in SCD.