Further characterization of a rat model of varicella zoster virus-associated pain: Relationship between mechanical hypersensitivity and anxiety-related behavior, and the influence of analgesic drugs

Abstract

Persistent herpes zoster-associated pain is a significant clinical problem and an area of largely unmet therapeutic need. Progress in elucidating the underlying pathophysiology of zoster-associated pain and related co-morbidity behavior, in addition to appropriately targeted drug development has been hindered by the lack of an appropriate animal model. This study further characterizes a recently developed rat model of zoster-associated hypersensitivity and investigates (a) response to different viral strains; (b) relationship between viral inoculum concentration ('dose') and mechanical hypersensitivity ('response'); (c) attenuation of virus-associated mechanical hypersensitivity by clinically useful analgesic drugs; and (d) measurement of pain co-morbidity (anxiety-like behavior) and pharmacological intervention in the open field paradigm (in parallel with models of traumatic peripheral nerve injury). Varicella zoster virus was propagated on fibroblast cells before s.c. injection into the glabrous footpad of the left hind limb of adult male Wistar rats. Control animals received injection of uninfected fibroblast cells. Hind-limb reflex withdrawal thresholds to mechanical, noxious thermal and cooling stimuli were recorded at specified intervals post-infection. Infection with all viral strains was associated with a dose-dependent mechanical hypersensitivity but not a thermal or cool hypersensitivity. Systemic treatment with i.p. morphine (2.5 mg/kg), amitriptyline (10 mg/kg), gabapentin (30 mg/kg), (S)-(+)-ibuprofen (20 mg/kg) and the cannabinoid WIN55,212-2 (2 mg/kg) but not the antiviral, acyclovir (50 mg/kg), was associated with a reversal of mechanical paw withdrawal thresholds. In the open field paradigm, virus-infected and nerve-injured animals demonstrated an anxiety-like pattern of ambulation (reduced entry into the central area of the open arena) which was positively correlated with mechanical hypersensitivity. This may reflect pain-related co-morbidity. Further, anxiety-like behavior was attenuated by acute i.p. administration of gabapentin (30 mg/kg) in nerve-injured, but not virus-infected animals. This model will prove useful in elucidating the pathophysiology of zoster-associated pain and provide a tool for pre-clinical screening of analgesic drugs.

Figure 1

Nature and duration of VZV-induced hypersensitivity to A) punctuate mechanical stimulation (electronic von Frey device) and B) dynamic mechanical stimulation (cotton bud). Animals (n = 3) were infected with VZV (strain Dumas) on day 0 and behavioural testing was performed at intervals until sensory thresholds returned to baseline values. The dotted line in B) represents the threshold for hypersensitivity i.e. paw withdrawal latency (PWL) <8 secs is consistent with dynamic mechanical hypersensitivity as described in the method by Field et al., (1999). +p<0.05 statistical difference in threshold response between ipsilateral and contralateral hind paws (one way ANOVA followed by Tukey test); *p<0.05 significant difference between response threshold at this time point compared to mean pre-infection baseline threshold (one way ANOVA followed by Dunnett’s test).

Figure 2

Ipsilateral paw withdrawal thresholds (PWT) in response to punctate mechanical stimulation following infection with different viral strains. A) Dumas (n = 9) -∎- and Ellen (n = 6) -▴- infected animals compared to uninfected fibroblast-injected control animals (n = 9) -∘- (+p<0.05 statistical difference between virus-infected and control animals, one way ANOVA followed by Tukey test; *p<0.05 significant difference between response threshold at this time point compared to mean pre-infection baseline threshold, one way ANOVA followed by Dunnett’s test); B) PWT responses expressed as the mean percentage decrease from baseline (*p<0.05 statistical difference between viral strains, one way ANOVA followed by Tukey test); and C) comparison of VZV-induced mechanical hypersensitivity following injection with a viral strain known to be associated with development of PHN following herpes zoster infection (n = 6), a viral strain not associated with the development of PHN following herpes zoster infection (n = 6) and the Dumas strain (n = 9) (*p<0.05 statistical difference between PHN-associated and not associated strains; +p<0.05 statistical difference between the PHN associated strain and Dumas strain; #p<0.05 statistical difference between viral strain not associated with development of PHN and Dumas strain - one way ANOVAs followed by Tukey test). Time at which VZV infection was performed (↑).

Figure 3

Viral dose-response relationship. Dose was defined as percentage cytopathic effect (cpe) (Dumas strain, 15%, 35% and 100% cpe), whilst response was defined as ‘the mean (± sem) percentage decrease from baseline ipsilateral paw withdrawal thresholds to punctate mechanical stimulation on day 14 post-infection’.

Figure 4

Comparison of the effect of vehicle (n = 13) versus drug administration on ipsilateral paw withdrawal response thresholds to punctate mechanical stimulation in VZV (Dumas)-infected animals. Arrow represents period of drug/vehicle administration. Each drug was given by the intraperitoneal route in a twice daily paradigm (08:00 and 18:00) for 4 consecutive days (18 - 21 post-infection inclusive), and each animal received only one drug. Sensory thresholds were examined 2 hours after the morning dose. Time course of the effect of A) morphine 2.5mg/kg (n = 6); B) amitriptyline 10mg/kg (n = 7); C) gabapentin 30mg/kg (n = 5); D) (S)-(+)-ibuprofen 20mg/kg (n = 6); E) WIN55,212-2 2mg/kg (n = 6); and F) acyclovir 50mg/kg (n = 7). Morphine, amitriptyline and gabapentin were dissolved in saline/sterile water, while (S)-(+)-ibuprofen, WIN55,212-2 and acyclovir were prepared in 40% DMSO solution. (+p<0.05 statistical difference between vehicle-administered and active drug-administered animals, one way ANOVA followed by Tukey test; *p<0.05 significant difference between response threshold at this time point compared to mean pre-infection baseline threshold, one way ANOVA followed by Dunnett’s test).

Figure 5

Open field parameters across all groups: naïve (n = 15); sham partial sciatic nerve injury (PSNI) (n = 11); PSNI (n = 10); sham spinal nerve transection (SNT) (n = 8); SNT (n = 11); uninfected fibroblast-injected control (n = 12); VZV-infected (n = 12). Open field parameters measured during a 15 minute testing session under low level lighting conditions were: A) number of entries (frequency) into the inner zone (40 × 40 cm zone); B) duration in the inner zone; and C) total distance moved throughout the arena (100 × 100 cm). *p<0.05 compared to naïve animals (t-test, followed by Mann-Whitney rank sum test where normality test failed); #p<0.05 (t-test, followed by Mann-Whitney rank sum test where normality test failed) for sham versus experimental.

Figure 6

A) Open field arena (100cm²) and inner zone (40cm²); B-F) Examples of characteristic track patterns as monitored using an infrared camera.

Figure 7

Relationship between open field inner zone entry and mechanical hypersensitivity. Paw withdrawal threshold (PWT) was measured using the electronic von Frey device. A) VZV (Dumas)-induced mechanical hypersensitivity is positively correlated with anxiety-like behaviour (as reflected by reduced entry into the open field inner zone). Pearson correlation coefficient, R = 0.63 (p = 0.02). ▴ represents individual VZV-infected animals on day 14 p.i. (n = 12) while the dotted line represents the trend between variables. However, there was no correlation in uninfected fibroblast-injected control animals. Mean ± sem percentage decrease in mechanical hypersensitivity for VZV-infected animals on day 14 p.i. was 26.2% ± 3.7, while mean frequency into the inner zone was 7.1 ± 1.6. For uninfected fibroblast-injected animals, there was a 1.1% ± 3.8 increase in mechanical PWTs overall, while mean frequency into the inner zone was 12.6 ± 1.9. B) Mechanical hypersensitivity is positively correlated with anxiety-like behaviour in spinal nerve transected (SNT) animals (n = 11). Pearson correlation coefficient, R = 0.71 (p = 0.01). ∎ represents individual animals 14 days post-surgery while the dotted line represents the trend between variables. However, there was no such correlation in sham animals. Mean ± sem percentage decrease in mechanical hypersensitivity for SNT animals on day 14 was 50.1% ± 2.6, while mean frequency into the inner zone was 2.9 ± 0.7. For sham animals, there was a 0.8% ± 1.6 decrease in mechanical PWTs overall, while mean frequency into the inner zone was 8.6 ± 1.5.

Figure 8

Open field pharmacological validation: effect of acute administration of gabapentin (GP) (30 mg/kg) or vehicle (saline) in A) spinal nerve transected (SNT) and sham-operated animals; and B) VZV (Dumas)-infected and uninfected-fibroblast (Hel) control animals 14 days post-surgery/infection. Drug/vehicle (n = 5 - 8 per group) was administered by the intraperitoneal (i.p.) route 20 minutes before testing in the open field. The following parameters were measured during a 15 minute testing session: (i) inner zone frequency; (ii) inner zone duration; and (iii) total distance traveled in the arena. * p<0.05 (t-test) compared to SNT or VZV-infected animals; # p<0.05 (t-test) compared to sham or Hel-injected control animals.

Figure 9

Immunohistochemical demonstration of VZV IE62 protein in ipsilateral L4/5 DRG following infection with A) viral strain associated with development of PHN following acute herpes zoster infection at (i) x20 magnification (arrow indicates a VZV positive cell), (ii) x40 magnification demonstrating diffuse cytoplasmic staining, (iii) x40 magnification demonstrating intra-nuclear staining; B) VZV strain not associated with development of PHN following acute herpes zoster infection; C) Dumas strain/high cpe; D) medium cpe; E) low cpe; F) Ellen strain; and G) uninfected fibroblast control. Images were photographed at x20 magnification, scale bar 50 μm. The arrow indicates cells positive for VZV IE62 protein. These cells generally exhibited diffuse cytoplasmic staining.

Figure 10

Demonstration of viral protein in sensory neurons following infection with different viral strains. A) Percentage distribution of VZV positive DRG neurons across groups (n = 3 animals per group). Using an immunofluorescence technique, ipsilateral L4 and L5 DRG from VZV-infected animals were pooled and labeled using a polyclonal antibody directed against VZV IE62 protein (1:250 dilution). Cells were co-labeled with the neuron-specific nuclear protein, NeuN and only those cells demonstrating a nucleus were included in the analysis. The number of NeuN labeled cells is ≥100 per group, taken from a series of random sections 100 μm apart (number of sections per group therefore varies until at least 100 NeuN labeled cells had been sampled). The percentage of NeuN labeled cells positive for VZV IE62 protein was then calculated and is displayed above the relevant bar for each group. Analysis of cell size distribution of VZV positive NeuN labeled cells from animals infected with B) viral strain associated with development of PHN; C) Dumas strain and D) Ellen strain, is shown.