Genetic evidence for an essential role of neuronally expressed IL-6 signal transducer gp130 in the induction and maintenance of experimentally induced mechanical hypersensitivity in vivo and in vitro

In gp130^fl/fl and SNS-gp130^-/- mice, experimental tumors as assessed from the tumor size at 10 days after tumor cell inoculation were similar in both groups (data not shown). Tumor growth was accompanied by increasing mechanical hypersensitivity and a decrease of mechanical withdrawal thresholds to 68.4 ± 10.9% of control values in gp130^fl/fl mice (n = 17). On the other hand, at the first day after inoculation mechanical sensitivity was not significantly changed in SNS-gp130^-/- mice (SNS-gp130^-/-, n = 11, 94.2 ± 15.9% in comparison to gp130^fl/fl, n = 17, 68.4 ± 10.9%, p < 0.05; two way RM ANOVA, Tukey post-test; genotype: F(1, 166) = 15.376; p < 0.001, time points: F(6, 166) = 7.739; p < 0.001, genotype × time points: F(6, 166) = 3.372; p < 0.01; Figure 1A). Mechanical thresholds dropped by 30% within the first three days in SNS-gp130^-/- mice but mechanical hypersensitivity significantly recovered from day 6 after tumor cell inoculation (SNS-gp130^-/-, n = 11, 79.9 ± 10.4% in comparison to gp130^fl/fl, n = 17, 35.4 ± 4.0%, p < 0.001; ANOVA). Differences between the two groups became even more evident at later time points.

As a possible mechanism two general possibilities are plausible. It is generally accepted that increased efficiency of spinal synaptic transmission is a major mechanism of mechanical hyperalgesia and allodynia. Alternatively, sensitization of primary nociceptive afferents could occur. To determine whether peripheral nociceptor sensitivity to mechanical stimuli was affected we performed standard teased fiber recordings from nociceptors at 7 to 10 days post inoculation, in vitro. Nociceptors with receptive fields within the tumor region had significantly lower mechanical von Frey thresholds than fibers innervating healthy skin in gp130^fl/fl mice (untreated: median: 32 mN, 17.65 mN and 83.5 mN as upper and lower quartile, n = 66 vs. tumor: median 16 mN, 11.4 mN and 22.6 mN as lower and upper quartile, n = 28). In healthy skin, 41% of mechanosensitive fibers responded to mechanical stimuli equal to 22.6 mN or lower whereas 59% were sensitive to 32 mN or higher (n = 66). In tumor associated skin, a significantly larger percentage of fibers (78%) responded to von Frey mechanical stimulation with less than 22.6 mN (n = 28, p < 0.01; χ²-test, Figure 1B). In contrast, mechanical sensitivity was similar of nociceptors projecting into healthy skin (n = 35) or tumor skin in SNS-gp130^-/- mice (n = 20; n.s.; χ²-test, Figure 1C, E). These results suggest that the signal transducer gp130 expressed is causally involved in tumor-associated mechanical hypersensitivity of nociceptors.

Although cancer pain appears to be unique and distinct from other chronic pain states [20] it seems to share at least some characteristics associated with inflammation [21] and also neuropathy [22]. Therefore, we aimed to address whether gp130 plays a role in regulating mechanical hypersensitivity in other persistent pain models. In the CCI model of neuropathic pain, both, gp130^fl/fl and SNS-gp130^-/- mice consistently developed a significant decrease of mechanical thresholds after nerve lesion. While control mice remained hypersensitive throughout the entire observation period, SNS-gp130^-/- mice partially recovered from the hypersensitive stated (day 28: percent decrease related to base line threshold gp130^fl/fl 23.9 ± 1.9%, n = 13, vs. SNS-gp130^-/- 67.2 ± 10.8%, n = 14; p < 0.05, p < 0.001; two way RM ANOVA, Tukey post-test; genotype: F(1, 181) = 4.820; p < 0.05, time points: F(6, 181) = 28.333; p = < 0.001, genotype × time points: F(6, 181) = 2.257; p < 0.05; Figure 2). This suggests that gp130 is also involved in the maintenance of neuropathic mechanical hypersensitivity.

Likewise, mechanical hypersensitivity is generally observed after subcutaneous injection of CFA. In gp130^fl/fl mice, subcutaneous injection (1 mg/ml) into the plantar site of the hind-paw resulted in paw swelling and a drop of mechanical von Frey thresholds to 31.3 ± 7.1% after 6 hours. The CFA-induced mechanical hypersensitivity was significantly attenuated in SNS-gp130^-/- mice (70.3 ± 6.8%, n = 8, p < 0.001; two way RM ANOVA, Tukey post-test; genotype: F(1, 104) = 21.439; p < 0.001, time points: F(6, 104) = 19.368; p < 0.001, genotype × time points: F(6, 104) = 5.165; p < 0.001; Figure 3A). Furthermore, control animals maintained the dramatic reduction of mechanical thresholds at 48 hours while SNS-gp130^-/- mice were largely resistant to mechanical hypersensitivity also in the maintenance phase of CFA induced inflammation (SNS-gp130^-/- 88.8 ± 7.1% vs. gp130^fl/fl 37.8 ± 8.7%, n = 8, p < 0.001; ANOVA). Six days after injection mechanical thresholds of SNS-gp130^-/- mice returned almost to baseline values whereas control mice were still mechanically hypersensitive (SNS-gp130^-/- 94.5 ± 5.2% vs. gp130^fl/fl 59.3 ± 9.7%, n = 8, p < 0.001; ANOVA). This difference could not be simply explained by a major reduction of inflammation, since at least the degree of swelling was similar in both mouse strains at 48 and 144 hours (Figure 3B, C).

Male C57Bl6 mice (> 8 weeks old) were used in all experiments. Mice were housed on a 12 h light/dark cycle with free access to chow and water and all animal use procedures were in accordance with ethical guidelines and animal welfare standards according to Austrian law. Mice were assigned to the following experimental groups: group I did not receive any treatment and was used as control; group II was inoculated with tumor cells; group III was injected intraplantarly with CFA; group IV underwent surgical procedure of chronic constriction injury (CCI). All groups contained gp130^fl/fl control and SNS-gp130^-/- mice, generated by gene targeting as described previously [10].

Standard testing procedures were used to quantify signs of pain-like behavior reflected in changes in mechanical sensitivity following inflammation, nerve lesion and tumor development. The area tested was the plantar side of the hind-paw where the tumor cells or CFA were inoculated. Baseline measurements were taken 2 times before treatment and daily thereafter up to 10 days post inoculation. Mice were placed in a plastic chamber with a wire mesh floor and allowed to habituate for 1 h before starting the test. Mechanical sensitivity at the site of tumor cells implantation was determined by measuring the paw withdrawal threshold in response to probing of the plantar surface of the hind-paw with calibrated von Frey monofilaments with bending forces between 2.8 and 45.3 mN. The withdrawal threshold was determined by increasing and decreasing stimulus intensity on the basis of the up-down method [64, 65], where a 11.4 mN stimulus was applied first. Behavioral test were performed in accordance with ethical guidelines and Austrian law. All measurements were done blindly.

An in vitro skin nerve preparation [10, 21, 66] was used to investigate the properties of unmyelinated afferent nerve fibers innervating the skin in the tumor area. In mice from group II electrophysiological recordings were performed 7 to 10 days post inoculation when a tumor mass had developed at the injection site covering the saphenous nerve territory. Animals were killed by CO₂ inhalation. The saphenous nerve was dissected with the skin of the dorsal hind-paw attached and mounted in an organ bath "inside-up" to expose the corium side. The preparation was superfused (15 ml/min) with an oxygen-saturated modified synthetic interstitial fluid solution containing (in mM) 108 NaCl, 3.48 KCl, 3.5 MgSO₄, 26 NaHCO₃, 1.7 NaH₂PO₄, 2.0 CaCl₂, 9.6 sodium gluconate, 5.5 glucose, 7.6 sucrose at temperature of 31 ± 1°C and pH 7.4 ± 0.05 [67]. The saphenous nerve was pulled into a separate chamber of the organ bath and placed on a small mirror. With the use of sharpened watchmakers' forceps fine filaments were teased from the desheathed nerve and placed on a gold wire recording electrode. Action potentials in single sensory neurons were recorded extracellularly, amplified (5000×), filtered (low pass 1 kHz, high pass 100 Hz), visualized on an oscilloscope and stored on a PC-type computer with Spike/Spidi software package for offline analysis employing a template-matching procedure [10, 12, 21]. The fibers were characterized as unmyelinated (C) according to their conduction velocity (< 2 m/s, calculated from the latency of unitary action potential to electrical stimulus at receptive field and distance of receptive field to recording electrode) and on the basis of the shape of the action potential. The receptive field of the primary afferent fiber was located by mechanical probing of the skin with a blunt glass rod. Only units with a signal-to-noise ratio > 2 were used for further analysis. Fibers were subject to a standard protocol of adequate mechanical and thermal stimuli. The mechanical threshold of each unit was determined with a set of calibrated von Frey monofilaments with uniform tip diameter 1.1 mm and bending forces ranging from 1 to 362 mN. The strength of the finest filament which evoked at least 3 action potentials was defined as activation threshold.

Lung carcinoma cells (ATCC clone 1642, American Type Cell Culture Collection) were cultivated on 25 cm² flasks in Dulbecco's modified Eagle's medium (DMEM, PAA, Vienna, Austria) with 4 mM L-glutamine and 10% fetal bovine serum (FBS), were grown to confluence, fed and passed once a week. Tumor cells were prepared for implantation by pouring off the media and rinsing with phosphate buffer saline (PBS). Trypsin-EDTA (0.5%, 1×, Gibco, Austria) was added for 2 min to detach cells from the flask. The enzymatic reaction was quenched by addition of modified DMEM. Just prior to implantation cells were counted, washed twice and then re-suspended in PBS for implantation. Mice were anaesthetized with isoflurane (Baxter, Vienna) and 7 × 10⁵ lung carcinoma cells in a volume of 25 μl PBS were injected subcutaneously in the plantar and dorsal site of the mouse hind-paw.

Complete Freud's Adjuvant (CFA, Sigma) was injected intracutaneously in a total volume of 25 μl and animals of group III were tested for mechanical sensitivity at 6, 24, 48, 72 and 144 h after injection. At 48 hours and the terminal day of the behavioral test, the vertical foot diameter of the injected and the non-injected paw was determined using a caliper.

The CCI model was obtained by three ligatures (7-0 prolene) on the right sciatic nerve with a distance of 1 mm each [68, 69]. The ligatures were tied around the nerve proximal to the trifurcation until a short flick of the hind-paw. Mice were anesthetized by intraperitoneal Phenobarbital injection before the procedure.

For statistical analysis the SigmaStat 3 software was used. Data are presented as mean ± SEM if not otherwise state. Two-way repeated measure ANOVA followed by Tukey post-hoc test or Mann-Whitney U-test for comparison between groups were used. For comparison of relative group sizes χ²-test was calculated. Differences were considered statistically significant at p < 0.05.