Cutaneous tactile allodynia associated with microvascular dysfunction in muscle

Cutaneous tactile allodynia (referred to henceforth as allodynia) is often found in patients with neuropathic pain, and is generally assumed to depend on the sensitization of the central nervous system in response to aberrant activity in damaged peripheral nerves [1]. However, allodynia is also caused by other injuries, such as that produced by ultraviolet radiation, and occurs in association with migraine headache [2] and fibromyalgia [3]. Allodynia is also prominent in complex regional pain syndrome (CRPS) [4], which can be initiated by either soft tissue (CRPS type-I) or nerve (CRPS type-II) injuries. Importantly, what both CRPS subtypes share with UV injury, migraine and fibromyalgia, besides allodynia, are significant vascular abnormalities caused by microvascular dysfunction [5‐9]. Also since CNS sensitization, which is critical for allodynia, is more pronounced following deep tissue injury than after cutaneous injury [10], it is possible that microvascular dysfunction in muscle may induce significant allodynia. However, few investigators have assessed vascular abnormalities in the etiology of chronic pain, and none have studied whether microvascular dysfunction in muscle contributes to allodynia. To address these questions, we investigated whether an ischemia-reperfusion (IR) injury produces allodynia in rats, and whether the allodynia is associated with microvascular dysfunction in muscle, and key mechanisms that underlie it. We show that microvascular dysfunction leads to persistent muscle ischemia, a reduction of intraepidermal nerve fibers, and allodynia correlated with muscle ischemia, but not with skin nerve loss. The affected hind paw muscle shows lipid peroxidation, an upregulation of nuclear factor kappa B, and enhanced pro-inflammatory cytokines, while allodynia is relieved by agents that inhibit oxidative stress, nuclear factor kappa B and cytokine activity. Allodynia is increased, along with hind paw muscle lactate, when these rats exercise, and is reduced by an acid sensing ion channel antagonist. Allodynia is also significantly correlated with muscle lactate before and after exercise.

We first tested whether allodynia is exhibited in rats with IR injury of the hind paw. A persistent significant reduction in mechanical paw-withdrawal threshold was observed following a 3 h IR injury induced using a tourniquet at the ankle (P = 0.0001) (Fig. 1a). This procedure produces a complete occlusion of blood flow to the hind paw, followed by prolonged reactive hyperemia (Fig. 2) and edema [11] on reperfusion. In addition to tactile allodynia, rats with what we have called chronic post-ischemia pain (CPIP) exhibit cold allodynia and mechanical hyperalgesia [11], as well as vascular abnormalities [12] that resemble symptoms in CRPS patients (Fig. 3).

To exclude the possibility that CPIP depends on a crush injury of afferent nerves, we examined whether allodynia was also induced after prolonged occlusion of the arteries supplying the hind paw. Mechanical paw-withdrawal thresholds were also persistently significantly reduced in rats whose hind paw blood vessels were occluded for 3 h (P = 0.011) (Fig. 1b), suggesting that allodynia was not caused by tourniquet-induced damage to underlying nerves.

If not nerve crush, it is likely that allodynia depends on IR injury and resultant microvascular dysfunction, which may include arterial vasospasms and capillary slow flow/no-reflow. Arterial vasospasms occur due to a reduction in nitric oxide-induced vasodilatation [13] and hyper-responsiveness of arterial smooth muscle cells to norepinephrine [14]. To determine whether CPIP rats exhibit arterial hyper-responsiveness, we used laser Doppler flowmetry to examine norepinephrine-induced reductions in blood flow in the hind paws of CPIP rats. We found that at 2 days post-perfusion, norepinephrine-induced reductions in CPIP hind paw blood flow were significantly enhanced (P = 0.0062) (Fig. 1c).

Slow flow/no-reflow is a condition where damage to the capillary endothelial cells causes them to swell, and to become clogged with platelets and white blood cells, thus obstructing the passage of red blood cells [15]. We used electron microscopy to assess the morphology of capillaries and endothelial cell thickness in CPIP hind paw muscles. Fig. 1d shows a normal capillary in the hind paw muscle of a sham-operated rat. In contrast, Fig. 1e shows a representative damaged capillary in CPIP hind paw muscle, with swollen endothelial cells and an occluded lumen. As shown in Fig. 1f, the endothelial cells in the hind paw muscle capillaries had significantly thicker (31.5%) cell walls in CPIP rats (P < 0.03971). These findings indicate that CPIP rats exhibit endothelial damage within hind paw muscle capillary beds.

To further characterize slow flow/no-reflow in CPIP rats, we used intra-arterial perfusion of India ink to stain patent blood vessels in the hind paw digital muscle of CPIP rats [16]. As shown in Fig. 1g, most vessels in the muscle capillary bed of the contralateral CPIP hind paw are stained with India ink, indicating that they are patent. In contrast, many vessels in the ipsilateral CPIP hind paw are poorly stained (Fig. 1h). Quantifying this, we found that the ipsilateral CPIP hind paw had significantly fewer patent blood vessels than the contralateral hind paw up to 7 days post-reperfusion (P = 0.0002) (Fig. 1i). This indicates that CPIP rats have persistent capillary slow flow/no-reflow in their hind paw muscles.

Muscle is particularly vulnerable to ischemia associated with microvascular dysfunction following IR injury [17]. To assess muscle ischemia, we examined the reduction of triphenyltetrazolium chloride (TTC) to triphenylformazan (formazan red) in CPIP hind paw digital muscle. TTC reduction is indicative of mitochondrial respiration and serves as an indicator of cellular oxygenation [17, 18]. Contralateral CPIP hind paw muscle exhibits normal TTC staining, indicated by uniform dark red colour due to effective reduction of TTC to formazan red (Fig. 4a). However, TTC staining is decreased in ipsilateral CPIP hind paw muscle, indicated by the patchy pink-white staining (Fig. 4b). A spectrophotometric assay was used to quantify formazan red in supernatant from homogenized CPIP hind paw muscle tissue. There was significantly less formazan red derived from the ipsilateral, as compared to the contralateral, CPIP hind paw muscle up to 7 days post-reperfusion (P = 0.0314) (Fig. 4c). This indicates that CPIP rats exhibit persistent muscle ischemia associated with reduced mitochondrial respiration.

It is possible that microvascular dysfunction in CPIP rats could induce secondary injury of the distal nerve endings, as previously reported in CRPS-I patients [19]. Thus, we examined intraepidermal nerve fiber (IENF) density in the CPIP hind paw skin using immunocytochemical staining with antibodies to protein gene product 9.5 (PGP9.5), a pan-neuronal marker [20]. Sham rats exhibits normal PGP9.5 skin staining with many IENFs observed in the epidermis (Fig. 4d). However, epidermal PGP9.5 staining is decreased in ipsilateral CPIP hind paw skin, indicating there is a reduction in IENFs (Fig. 4e). Quantifying PGP9.5 staining over the first week after IR injury, we found a significant reduction in IENF density in the epidermis of CPIP hind paw skin (P = 0.0133) (Fig. 4f). However, when we plotted IENF density against mechanical paw-withdrawal thresholds, we found no significant correlation between IENF density and allodynia in CPIP rats at either 2 or 7 days post-reperfusion (P > 0.05) (Fig. 4g). Also, CPIP rats did not exhibit any significant alteration of conduction velocity in the sural nerve (P > 0.05) (Fig. 4h). This is consistent with the absence of a tourniquet-evoked injury to peripheral nerve at the ankle, and similar to findings in CRPS-I patients, who by definition do not have detectable injury at the level of the peripheral nerve.

Microvascular dysfunction after IR injury is initiated by oxygen free radicals generated when oxidases that accumulate in ischemic tissue reduce molecular oxygen arriving on reperfusion [21]. We assessed free radical-induced lipid peroxidation in hind paw muscle of CPIP rats, and whether CPIP allodynia is attenuated by the antioxidant N-acetyl-L-cysteine (NAC) or the free radical scavenger 4-hydroxy-2,2,6,6-tetramethylpiperydine-1-oxyl (TEMPOL) given 2 days post-reperfusion when allodynia peaks. Malondialdehyde (MDA), a product of lipid peroxidation, was significantly elevated in ipsilateral, compared to contralateral CPIP hind paws (P = 0.0296) (Fig. 5a), and CPIP allodynia was significantly attenuated by NAC (Fig. 5b) (P = 0.0002) and TEMPOL (P < 0.00001) (Fig. 5c), suggesting a key role of free radicals.

After IR injury, free radicals stimulate the production of pro-inflammatory cytokines [22], following upregulation of nuclear factor kappa B (NFκB) [23]. We used ELISAs to determine the levels of pro-inflammatory cytokines and NFκB in CPIP hind paw muscle, as well as examining the effects of interleukin-1 receptor antagonist (IL-1RA), and an inhibitor of NFκB (pyrrolidine ditiocarbamate, PDTC), on CPIP allodynia 2 days post-reperfusion. The cytokines tumor necrosis factor-α (P = 0.0085) (Fig. 5d), IL-6 (P = 0.0014) (Fig. 5e) and IL-1β (P = 0.0021) (Fig. 5f), and the transcription factor NFκB (P = 0.0073) (Fig. 5g), were all significantly elevated in the CPIP hind paws early after reperfusion, and both IL-1RA (P = 0.0302) (Fig. 5h) and PDTC (P = 0.0009) (Fig. 5i) significantly elevated paw-withdrawal thresholds of CPIP rats, suggesting that NFκB and pro-inflammatory cytokines play a role in CPIP allodynia.

Both ischemia [24] and exercise [25] produce lactate acidosis and muscle pain. We measured lactate levels in the CPIP hind paw muscle for 2 weeks post-reperfusion, and compared lactate levels with paw-withdrawal thresholds. We also assessed whether muscle lactate levels increased, and paw-withdrawal thresholds decreased, in CPIP rats that exercised (running on a circular treadmill). Basal lactate levels were significantly elevated in CPIP rats, particularly early after reperfusion (P = 0.00036) (Fig. 6a). Early post-reperfusion, when lactate levels peaked, CPIP rats showed a strong reluctance to exercise, indicated by significantly increased running stoppages (P = 0.0083) (Fig. 6b). Later post-reperfusion, when basal lactate levels were lower, running resulted in significantly higher lactate levels in the hind paw muscle (P = 0.0215) (Fig. 6c), and significantly lower paw-withdrawal thresholds (P = 0.00135) (Fig. 6d). There was also a significant inverse correlation between lactate level and paw-withdrawal threshold in unexercised and exercised CPIP rats (Fig. 6e). Finally, allodynia in CPIP rats is significantly reduced in rats treated with low doses of the acid sensing ion channel (ASIC) antagonist amiloride (P = 0.02472) (Fig. 6f). These results suggest that muscle ischemia results in the generation of lactate in CPIP rats contributing to pain and allodynia at rest, as well as to exercise-induced pain and increases in allodynia. Importantly, while allodynia was not correlated with cutaneous nerve abnormalities, it was correlated with a measure of muscle ischemia (i.e, lactate), and attenuated by blocking the actions of lactate at ASICs. Parallel to these observations in rats, human CRPS is also marked by increased pain and allodynia after exercise [26]. Although the anti-allodynic effects of amiloride argue for a role of ASICs in CPIP, it is important to note that the effects of amiloride are non-selective as it also affects both epithelial sodium channels and Na+/H+ exchangers [27]. Thus, additional experiments with more selective ASIC antagonists are warranted.

We describe here that allodynia after hind paw IR injury coincides with arterial hypersensitivity to norepinephrine, capillary slow flow/no-reflow and ischemia, and follows increased lipid peroxidation, NFκB, and pro-inflammatory cytokines in muscle tissue. Allodynia is also alleviated by agents that reduce these consequences of microvascular dysfunction, and is directly related to muscle tissue lactate. All these findings indicate that microvascular dysfunction and ischemia in muscle underlies persistent allodynia after IR injury. Our results are consistent with the demonstration that activity in muscle nociceptors is more potent than activity in cutaneous nociceptors in evoking central sensitization [10]. Also, while injecting capsaicin or carrageenan into skin produces allodynia that lasts at most hours, the same injections into muscle induce prolonged central sensitization and cutaneous allodynia lasting up to several weeks [28, 29]. Here, IR injury induced alterations in skin nerves, however, unlike the persistent muscle ischemia, these changes were not correlated with allodynia. The finding that allodynia is correlated with muscle lactate (but not skin IENF loss) argues for a more critical role of muscle changes in allodynia. However, we can not dismiss the possibility that the skin IENF loss has some role in the production of allodynia.

The relationship between microvascular dysfunction in muscle and pathological pain is strengthened by our evidence that allodynia is attenuated by reducing oxygen free radicals, NFκB and pro-inflammatory cytokines, all mediators of IR injury. This is also the first evidence that increased lactate associated with muscle ischemia induces allodynia, at rest or after exercise, in the rat hind paw, and that allodynia after IR injury is reduced with an ASIC antagonist. Importantly, ASIC channels on afferent nerves are tuned to respond to increased lactate [30], and ASIC-3 knock-out mice do not develop the prolonged allodynia or CNS sensitization seen in wild-type mice after muscle injections of acidic saline [31]. Microvascular dysfunction and ischemia in muscle may be critical, but largely ignored, mechanisms of chronic pathological pain, playing a significant role in the etiology of allodynia. See Fig. 7 for a summary schematic of the interpretations of our findings.

Muscle ischemia and painful symptoms after ischemia-reperfusion injury highlight the importance of microvascular dysfunction in muscle to cutaneous tactile allodynia, or painful hypersensitivity to touch, in chronic pain. Arterial vasospasms, endothelial cell injury and capillary slow flow/no-reflow after ischemia-reperfusion injury induce persistent allodynia dependent on oxygen free-radicals, nuclear factor kappa B, pro-inflammatory cytokines and lactate in muscle. Microvascular dysfunction leads to abnormalities in cutaneous nerves, as well as persistent ischemia in muscle, and that allodynia is significantly correlated with the muscle ischemia, but not with skin nerve changes. Besides nerve cells, pathology in muscle microvasculature may be an important new target for development of therapies for chronic pain.

All methods were approved by the McGill Animal Care and Ethics committees.