The mechanical conflict-avoidance assay is used as a non-reflexive readout of pain sensitivity in mice which can be used to better understand affective-motivational responses in a variety of mouse pain models.
Pain comprises of both sensory (nociceptive) and affective (unpleasant) dimensions. In preclinical models, pain has traditionally been assessed using reflexive tests that allow inferences regarding pain’s nociceptive component but provide little information about the affective or motivational component of pain. Developing tests that capture these components of pain are therefore translationally important. Hence, researchers need to use non-reflexive behavioral assays to study pain perception at that level. Mechanical conflict-avoidance (MCA) is an established voluntary non-reflexive behavior assay, for studying motivational responses to a noxious mechanical stimulus in a 3 chamber paradigm. A change in a mouse’s location preference, when faced with competing noxious stimuli, is used to infer the perceived unpleasantness of bright light versus tactile stimulation of the paws. This protocol outlines a modified version of the MCA assay which pain researchers can use to understand affective-motivational responses in a variety of mouse pain models. Though not specifically described here, our example MCA data use the intraplantar complete Freund’s adjuvant (CFA), spared nerve injury (SNI), and a fracture/casting model as pain models to illustrate the MCA procedure.
Pain is a complex experience with sensory and affective components. A reduction in the threshold of pain perception and hypersensitivity to thermal and/or mechanical stimuli are key features of this experience, which stimulus-evoked pain behavior tests can capture (like Hargreaves' test of heat sensitivity and the von Frey test of mechanical sensitivity)1,2. Although such tests give robust and reproducible results, they are limited by their reliance on reflexive withdrawal from a perceived noxious stimulus. This has called into question an ongoing reliance of pain research on these tests alone. To that end, pain researchers have for several years been exploring alternative/complementary behavioral tests for use in rodent pain models in an effort to capture more of the affective and/or motivational components of pain. These un-evoked, voluntary, or non-reflexive measures (e.g., wheel running, burrowing activity, conditioned place preference3,4,5) are being implemented in an attempt to improve the translatability of preclinical pain research.
The mechanical conflict avoidance (MCA) assay was originally described by Harte et al. in 20166, is used predominantly in rats7,8, and represents a modification of an earlier approach – the place escape-avoidance paradigm. In this approach, a noxious stimulus of the hind paw is performed in an otherwise desirable (dark) chamber to drive purposeful behavior of the animal to escape/avoid such stimulation9,10. Instead of relying on manual noxious stimulation of the hind paw by an observer, the MCA assay forces mice to negotiate a potentially noxious stimulus to escape an aversive environment and reach the dark chamber. The conflict/avoidance that gives the assay its name arises from these two competing motivations: escape brightly-lit areas and avoid noxious stimulation of the paws. The MCA assay also shares features with conditioned place-preference testing, where the pairing of pain relief with environmental cues drives changes in behavior that reflect a preference for the pain-relieving/rewarding context11.
Fundamentally speaking, all these assays share a similar approach: using a shift in an animal's preference for one aversive environment over another as an indicator of their affective/motivational state. The MCA assay is a 3 chamber paradigm consisting of a brightly lit chamber followed by a dark middle chamber with adjustable height probes and a dark third chamber without any aversive stimuli. An uninjured mouse is typically motivated to escape to a darkened chamber, given the innate aversion of rodents to bright light12. In this example, the natural motivation to escape a brightly-lit environment overcomes the disinclination to encounter hind paw stimulation (the adjustable height probes), which occurs exclusively in the darkened environment. In contrast, a mouse experiencing pain (due to inflammation or neuropathy, for example) may opt to spend more time in the brightly-lit environment, since there is motivation to avoid the unpleasant tactile experience of the mechanical probes in the setting of ongoing tactile hypersensitivity.
This article describes a modified version of the MCA assay. We have adapted the original method (which was performed in rats6) for use in mice. We have also reduced the number of probe heights tested from six to three (0, 2, and 5 mm above floor height) in order to streamline data acquisition. This approach has been tested across multiple pain models, and validated with known analgesics, indicating that pain hypersensitivity and/or the associated affective and motivational changes are driving these changes in behavior. This approach is relatively quick to conduct and adaptable when compared to other non-reflexive measures, which can take many days of habituation and training1,2. In concert with other measures of pain, MCA can generate valuable insights into the affective and motivational aspects of pain.
All experiments involving the use of mice and the procedures followed therein were approved by Institutional Animal Care and Use Committees of MD Anderson Cancer Center and Stanford University, in strict accordance with the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals.
1. MCA construction
2. Mouse MCA habituation and testing
The MCA assay has been used successfully with several mechanistically distinct mouse pain models. Figure 2 shows data where the outcome measure of choice was crossing the mid-point of chamber 2 (Figure 2A). The data obtained by using the halfway point versus escape into chamber 3 are very similar, ~40 s for halfway versus ~45 s for chamber 3 escape in the spared nerve injury (SNI) model of neuropathic pain with 5 mm probe height13.
In the CFA-induced inflammatory pain model, control hind paw (intraplantar) injection of saline does not change escape latency versus baseline. Those mice that were injected with CFA in one hind paw showed a significant increase in escape latency 4 days post-injection, but only when the probe height was raised to 5 mm. Crucially, this increased latency to escape at 5 mm was not seen in mice that received the NSAID carprofen (10 mg/kg, i.p.) 90 min before the beginning of testing (Figure 2B).
The spared nerve injury (SNI) model of traumatic neuropathic pain, is also associated with a significant increase in the latency to escape versus baseline when probe height was set to 5 mm. This effect was seen in SNI mice, but not their sham surgery controls. This increased escape latency was also prevented by systemic administration of the opioid analgesic buprenorphine (25 µg/kg, i.p.) 90 min prior to testing (Figure 2C). Increased escape latency was also observed in mice that did not undergo a baseline round of MCA testing prior to nerve injury (Figure 2D). In this case, the increased escape latency in SNI mice at 5 mm was prevented by gabapentin (30 mg/kg, i.p.) administered 90 min prior to testing. Collectively, this suggests that MCA can detect pain-related changes in stimulus aversion and avoidance in two widely-used models of inflammatory and neuropathic pain.
MCA was further tested in the fracture/casting model of the chronic pain condition complex regional pain syndrome (CRPS) which is established by a closed right distal tibia fracture followed by 3 weeks of casting14. This clinically informed model exhibits acute phase peripheral inflammation, as well as long-term immune activity in the central nervous system with persistent hindlimb allodynia. Similar to the CFA and SNI models, increased escape latencies were observed in the fracture/casting model (Figure 3A). Prior to the injury, the latency to escape from chamber 1 increased proportionally to the probe height. After the injury, the escape latency remained unchanged at 0 mm but significantly increased at the 2 mm and 5 mm probe height for males and the 5 mm probe height for females when compared to baseline (Figure 3B).
Figure 1: Schematic and images of the MCA device. (A) Potential outcome measures in the MCA assay, (marked by clockface icons): latency to exit chamber 1 (I), latency to cross more than 50% of chamber 2 (dotted line; II), the total amount of time spent in chamber 2 (III), latency to reach the escape chamber (IV) or percent time spent in each chamber (V). Animals experiencing pain on average show greater values for I, II, and IV, and reduced values for III. A reduced value for III necessarily increases the proportion of time spent in chamber 1 and/or chamber 3, which would be captured by outcome measure V. Created with Biorender.com. (B) Images illustrating the MCA device (and chambers numbered 1, 2, and 3) with the LEDs switched off (top left), the LEDs switched on (bottom left). (C) A view of the chambers from above with the doors opened. Please click here to view a larger version of this figure.
Figure 2: Inflammatory and neuropathic pain augment avoidance in the MCA assay. (A) Depiction of the specific outcome measure used here: latency to cross the chamber 2 midpoint. (B) Intraplantar injection of CFA significantly increased the latency to escape (red squares) versus saline controls (black circles) when the probe height was set to 5 mm. Intraperitoneal carprofen (10 mg/kg) attenuated the CFA-induced increase in escape latency (blue triangles). Data are plotted as mean escape latency ± standard error of the mean (SEM); n = 7 males/group. (C) Spared nerve injury (SNI) surgery significantly increased chamber 1 escape latency versus sham surgery controls (black circles), when probe height was set to 5 mm (red squares). Intraperitoneal buprenorphine (25 mg/kg) significantly attenuated this increase in escape latency (blue triangles). Data are plotted as mean escape latency ± SEM; n = 6-7 males per group. (D) SNI-induced increase in escape latency was reversed by use of the analgesic gabapentin (green triangles). Data are plotted as mean escape latency ± SEM; n = 8 males/group. ## = p < 0.01, ***/### = p < .001, for the indicated comparisons (two-way ANOVA, Bonferroni post-hoc). This figure has been modified from13. Please click here to view a larger version of this figure.
Figure 3: Tibial fracture/casting induced chronic pain augment avoidance in the MCA assay. Fracture/casting significantly increased escape latency at 3 weeks post-injury (W3) versus baseline (BL) in males at the 2 mm and 5 mm probe heights and in females at the 5 mm probe height (n = 5/sex). Data from each mouse are depicted in faded black (males) or cayenne (females) with mean represented by dark lines. **/*** = p < 0.01/< 0.001 versus sex- and probe height-matched baseline value by two-way ANOVA, Tukey post-hoc. Please click here to view a larger version of this figure.
Supplementary File 1: 3D printer file SpikeBed-MCA. When printed in a suitably biocompatible and washable material, such as nylon 12, SpikeBed-MCA.stl produces the platform of tactile probes which protrude through the floor of chamber 2. Please click here to download this File.
Supplementary File 2: 3D Printer file MCA_baseplate. When printed in a suitably biocompatible and washable material, such as nylon 12, MCA_baseplate.stl produces the floor of chamber 2, through which the tactile probes protrude. Please click here to download this File.
As with all behavioral tests, proper handling, randomization, and blinding to the treatment of animals is essential throughout. Given the multifactorial inputs into complex behaviors and decision-making, it is imperative that animals are handled, habituated, and tested as consistently as possible while minimizing distress. Care should also be taken to reproduce the timing of mouse placement in chamber 1, switching on the LED lights, and removing the barrier, since differences here could influence subsequent behavior.
It should be noted that the different outcome measures depicted in Figure 1A are inter-related. For example, a mouse entering chamber 2 usually crosses the halfway point of chamber 2 and then almost always completes escape into chamber 3. This means that outcome measures I, II and IV are inter-related. Outcome measures III and V measure the total dwell time in chamber 2, and the proportion of dwell time in all 3 chambers, respectively. Therefore, these measures are closely related to one another. However, a mouse can theoretically accrue substantial dwell time in chamber 2 whether the halfway crossing or escape into chamber 3 had a low latency, high latency, or did not occur at all.
Several variations or modifications of this method have been reported. In addition to the different outcome measures listed here (Figure 1), investigators could vary the progression of probe height in an effort to highlight differences in sensitivity. Since there were no statistically significant differences with the intermediate 2 mm probe height, it may be more efficient to run mice only at 0 mm and 5 mm. Alternatively, a probe height between 2 and 5 mm, or repeated runs at a 5 mm probe height may begin to unmask differences that were not otherwise apparent. In addition, evaluation of dwell time in each chamber can be used as a readout of motivation and activity. This can be useful in instances where some mice run quickly through to chamber 3 but then return to chamber 1 to further explore. In these situations, latency to entry into chamber 3 alone would not capture this subtlety. Raising the testing cutoff time beyond the 2 min limit set here may also prove worthwhile for some investigators. Finally, we cannot exclude the possibility that repeat testing of the same animals (more than the three times described here), or testing with greater frequency (fewer than 4-7 days between tests) may introduce habituation or learning effects. For these reasons, the inclusion of naïve, unmanipulated control groups at every time point is encouraged. Ultimately, variations in behavior are highly likely to be pain model specific and warrant further investigation in these and other pain models.
The pain-inducing models used here (CFA, SNI, fracture/casting) are typically associated with hypersensitivity in other pain behavior tests, which corresponds to an increase in avoidance/escape latency. The MCA assay may also be able to detect a loss of sensory acuity (via increased time spent in chamber 2, for example), though this has not been formally tested. MCA has some limitations which warrant consideration. Aversion to bright light is a key means of motivating entry into chamber 2 and therefore a driver of subsequent conflict. Any pathological feature associated with a particular mouse model that might alter aversion to bright light (e.g., visual impairment) should be carefully considered before employing this test. The contribution of anxiety to escape latency has also not been systematically tested, though chronic inflammatory and neuropathic pain models have been reported to show signs of anxiety-like behavior in mice in other tests, there continues to be a debate regarding this15,16. That said, a contribution of pain-related anxiety to these behavioral outcomes cannot be confirmed or ruled out at this time. Since MCA has multiple inputs into the outcome measures, this comes with more potential confounds to consider.
In summary, the MCA test provides a non-reflexive readout of pain sensitivity in mouse models. The outcome measures are influenced by factors other than reflexive sensitivity and provide a composite measure of pain sensitivity and affective/motivational state. The amount of time needed to run each test, and the level of skill and specialized equipment required compare favorably with other non-reflexive measures of pain, such as gait analysis or conditioned place preference5,13. Though still somewhat novel, the approach has been adopted and independently verified by multiple teams of investigators, predominantly in rats. Partial sciatic nerve ligation increased exit latency17 and morphine-dependent withdrawal in rats7. Another study in rats proposed that counting the number of crossings, using spinal cord injury and chronic constriction injury models in rats, may serve as a useful outcome measure8. Crucially, this study also identified an increase in probe avoidance in sham surgery controls, indicating that inclusion of a naïve group alongside sham/vehicle controls is warranted. Future applications of MCA could focus on the variation between mouse strains and/or pain models, the impact of anxiety on assay performance, and integration of posture analysis or gait kinematics to better understand differences in behavioral adaptations to noxious stimuli.
The translational gap between preclinical mouse studies and the development of novel therapeutics continues to present a cause for concern. With this in mind, the MCA assay complements existing tools in pain research and helps give a more complete picture of the many sensory and affective dimensions of pain.
The authors have nothing to disclose.
GM is supported by an NDSEG Graduate Fellowship. VLT is supported by NIH NIGMS grant #GM137906 and the Rita Allen Foundation. AJS is supported by Department of Defense grants W81XWH-20-1-0277, W81XWH-21-1-0197, and the Rita Allen Foundation. We are grateful to Dr. Alexxai Kravitz at Washington University School of Medicine for designing and making freely available the 3D printer files for the chamber 2 floor and probe plate.
32.8ft 3000K-6000K Tunable White LED Strip Lights, Dimmable Super Bright LED Tape Lights with 600 SMD 2835 LEDs | Lepro | SKU: 410087-DWW-US | For lighting chamber 1. https://www.lepro.com/32ft-dimmable-tunable-white-led-strip-lights.html |
3D printed 'spike bed' and 'chamber 2 floor' | Shapeways | N/A | Optional, for mechanical probes as an alternative to blunted map pins. |
70% ethanol | Various | N/A | To clean MCA between mice. |
Acryl-Hinge 2 | TAP Plastics | N/A | for attaching chamber lids to rear walls. https://www.tapplastics.com/product/plastics/handles_hinges_latches/acryl_hinge_2/122 |
Chemcast Cast Acrylic Sheet, Clear | TAP Plastics | N/A | 3mm thick. For front wall of chamber 1. https://www.tapplastics.com/product/plastics/cut_to_size_plastic/acrylic_sheets_cast_clear/510 |
Chemcast Cast Transparent Colored Acrylic, Transparent Dark Red – 50% | TAP Plastics | N/A | 3mm thick. 50% light transmission. For walls and lids of chambers 2 and 3. https://www.tapplastics.com/product/plastics/cut_to_size_plastic/acrylic_sheets_transparent_colors/519 |
Chemcast Translucent & Opaque Colored Cast Acrylic, Sign Opaque White – 0.1% | TAP Plastics | N/A | 3mm thick. For side walls and lid of chamber 1. https://www.tapplastics.com/product/plastics/cut_to_size_plastic/acrylic_sheets_color/341 |
Disinfectant (e.g. Quatricide) | Pharmacal Research Laboratories, Inc. | 65020F | To disinfect MCA at the end of a testing session. |
Dry-erase markers and board | Various | N/A | To add experimental info to the beginning of video footage. |
Map pins | Various | N/A | Optional, for mechanical probes. Use sandpaper to blunt sharp points before use. Can be used in place of 3D-printed parts. |
Paper towels | Various | N/A | To clean/disinfect MCA. |
SCIGRIP Weld-On #3 Acrylic Cement | TAP Plastics | N/A | For assembling acrylic sheets into chambers and affixing hinges. https://www.tapplastics.com/product/repair_products/plastic_adhesives/weld_on_3_cement/131 |
Stopwatch | Various | N/A | To record escape latencies/dwell times in real-time or from recorded video. |
Timer | Various | N/A | To ensure LED turn-on, barrier removal and test completion are timed consistently. |
Video camera | Various | HDRCX405 Handycam Camcorder | To record mouse behavior in the MCA device. Can be substituted with any consumer-grade video camera capable of 1080p resolution. |
Tripod | Famall | N/A | Any tripod that can hold the camera at bench height for recording MCA footage is acceptable. |