Behavioral analyses

Experiments were performed following the guidelines of the International Association for the Study of Pain with approval of the University of California, San Francisco Institutional Animal Care and Use Committee or the Chancellor’s Animal Research Committee at UCLA. Mice were housed in a 12 h light/dark cycle at 21°C. Comparisons were made between reeler or dab1 mice and their wild type littermates.

1. Mechanical stimulation. 15 min after sodium pentobarbital (50 mg/kg) injection, we applied a modified alligator clip to the left hind paw at the base of the toes every 3 min for a 20 sec duration, for a total of 5 stimulus presentations.

2. Thermal (heat) stimulation. Mice were anesthetized using sodium pentobarbital (50 mg/kg) 15 min prior to the stimulation. The left hindpaw was dipped into a 50°C water bath for 3 sec per min, for a total of 10 min.

3. Formalin test. Mice were placed in Plexiglas testing chambers and allowed to acclimate to the environment for 30 min. Next 10 μl of 2% formaldehyde diluted in saline was injected subcutaneously into the plantar surface of one hindpaw. The duration of licking (nocifensive behavior) of the injected paw was recorded in 5 min intervals for one hour after the injection (Tjolsen et al., 1992).

4. Double cold plate test. In this test, the temperature of one aluminum plate is controlled with fluid circulation (from room temperature to 5°C), and the immediately adjacent plate set at room temperature (Bautista et al., 2007). Half the animals were placed initially on the room temperature plate and the others on the colder plate. Animals typically explore both plates. We recorded the time that the animal spent with both hindpaws in contact with the room temperature plate versus the colder plate.

5. Visceral pain test. Acetic acid was administrated intraperitoneally (100μl/10g body weight of 0.6% (v/v) in saline) and the evoked stretching responses were counted for the next 20 minutes.

Tissue preparation and immunohistochemistry

One hour after the heat or mechanical stimulus, mice were deeply anesthetized with sodium pentobarbital (100 mg/kg) and directly perfused through the heart with 4% paraformaldehyde in 0.1 M phosphate buffer (PB), ph 7.4. Alternately, after chemical and visceral stimuli, mice were anesthetized with 2.5% Avertin, then perfused transcardially with the fixative described above. The spinal cord was dissected out and post-fixed in 4% paraformaldehyde/PB for 3–4 hours at 4°C and then cryoprotected in 30% sucrose/PB overnight at 4°C. L4/L5 spinal cord sections (for mice in the heat, mechanical, and formalin studies) or thoracic spinal cord sections (for the visceral pain stimulus) were immunostained for Fos as previously described (Presley et al., 1990). Digitized images of 6–10 heavily Fos-labeled sections per animal were captured using both bright and darkfield illumination. Using the darkfield image as a guide, a line was drawn at the border between the relatively dark substantia gelatinosa (lamina II) and the more lucent or opaque lamina III. Fos immunoreactive profiles were counted in laminae I–II, III–IV, V–VI, and the lateral spinal nucleus as delineated in figure 1. Most Fos-positive cells found in the lateral reticulated part of the neck of the dorsal horn, i.e., the area which corresponds to the thin strands of lateral gray matter intermixed with longitudinal axon bundles (Menétrey et al., 1982), were included in laminae V–VI counts. Finally, to determine the distribution of Dab1-labeled cells in the reeler wild type and mutant lateral reticulated area and LSN of lumbar segments L4/5, we used a Dab1 antiserum (B3, 1:5000; gift of Dr. Brian Howell, Syracuse, NY) with standard avidin-biotin methods (Villeda et al., 2006, Akopians et al., 2008).

Open in a separate window

Fig. 1

Schematic of dorsal horn lamination and the area of the lateral spinal nucleus (LSN) used for Fos analysis. DLF, dorsolateral funiculus; DF, dorsal funiculus.

Enhanced noxious heat-evoked Fos expression in reeler and dab1 mutants

A second nociceptive consequence of deletion of the Reelin-signaling pathway is a greater sensitivity in the Hargreaves radiant heat test (Villeda et al., 2006, Akopians et al., 2008). Consistent with these findings, we now show that the reeler mutants have significantly greater Fos expression in laminae I–II of lumbar dorsal horn. The number of Fos-immunoreactive neurons in laminae III–IV, V–VI and in the LSN did not differ between genotypes (Fig. 2E, G, H), although there were more Fos-positive neurons in the lateralmost part of the reticulated area in wild type mice than in reeler mutants (Fig. 2G, H, arrows).

The dab1 mutants also show a heightened response to heat stimulation in the Hargreaves test (Akopians et al., 2008). Now we show that dab1 mutants also have increased heat-induced Fos expression compared to their control counterparts in the superficial (laminae I–II) and deep dorsal horns (laminae III–VI, Fig. 2F). By contrast, the number of Fos-labeled neurons in the LSN of dab1 mutants is reduced compared to wild type mice. These somewhat paradoxical results may be explained by our earlier findings that ~50% of the LSN neurons at lumbar levels are mispositioned, most likely to other areas of the dab1 mutant dorsal horn (Akopians et al., 2008).

Lateral reticulated area: Fos expression in Dab1-containing neurons of reeler wild type, but not the mutant

As mispositioned neurons in reeler mutants contain high levels of Dab1 protein, we examined the distribution of Dab1 cells in reeler wild type and mutant lateral reticulated area at L4/L5 segmental levels. Figure 3C illustrates several prominent Dab1-immunoreactive neurons (arrow) wedged between the lateralmost white matter bundles in a wild type mouse; other Dab1-positive neurons are located further laterally and dorsally, in the LSN. We detected fewer Dab1 cells in the lateral reticulated area of reeler mutants (Fig. 3D, arrow). On average, wild type mice had almost twice the number of Dab1 neurons in the lateral reticulated area (mean + SEM; wild type 4.2 + 0.25 cells/hemisection; mutant 2.1 + 0.28; p<0.0001) and LSN (wild type 2.2 +0.25; mutant 1.3 + 0.28; p<0.04) compared to reeler mutants. On the other hand, reeler mutants have ectopically located Dab1-positive cells along the outer (dorsal) border of lamina I and at the medial border adjacent to the dorsal columns (Fig. 3F, arrowhead). These findings suggest that some projection neurons in the lateral reticulated area express Dab1 and require the presence of the Reelin-signaling pathway for correct positioning in the deep dorsal horn.

We next asked if the prominent Dab1-labeled neurons in the lateral reticulated area expressed Fos following noxious mechanical and heat stimulation. Interestingly, we observed Fos-immunoreactive Dab1 cells in the wild type lateral reticulated area (Fig. 3C, E, arrowheads on insets) following both noxious stimuli but they were rarely found in reeler mutants (Fig. 3D, F), consistent with our hypothesis that some neurons in the lateral reticulated area are mispositioned. Several ectopically positioned Dab1 neurons expressed Fos (Fig. 3F at arrowhead, enlarged in inset) in reeler mutants following noxious heat stimulation.

Both reeler and dab1 mutants show reduced nocifensive behavior and Fos expression after hindpaw formalin

Intraplantar formalin induced the typical biphasic pattern of licking (Tjolsen et al., 1992) in both reeler and dab1 wild type control mice (Fig. 4A, E). However, in reeler mutants both the first (0–10 min) and second phases (11–60 min) of behavior were significantly decreased (Fig. 4A). The response to formalin in the dab1 mutants was also reduced, but interestingly, only during the second phase of behavior (Fig. 4E). Consistent with the behavior observed, formalin-induced Fos expression was significantly reduced in laminae I through VI of the spinal cord, in both reeler and dab1 mutants (Fig. 4C, D). The magnitude of the reduction was comparable in all regions of the dorsal horn (Fig. 4B, F) but there appears to be an especially striking reduction of Fos expression in the lateral reticulated area of the reeler mutant mice (Fig. 4C, D; arrows).

Open in a separate window

Fig. 4

Formalin-evoked nocifensive behavior and Fos expression in reeler (A–D) and dab1 (E–H) mice. Formalin was injected into the hindpaw and the duration of licking/biting the injected paw was measured in 5-min intervals for 1 hr. A, E: Compared to wild type mice, reeler mutants had reduced nocifensive behavior in phases I and II (A), whereas in dab1 mutants, a decrease was seen only in phase II (E). B, F: Fewer Fos immunoreactive cells were found in reeler (B) and dab1 (F) mutant dorsal horn laminae than in wild type mice. Fos expressing neurons were decreased in the reeler lateral spinal nucleus (LSN), but not dab1 mutants. C, D, G, H: Representative sections illustrate distribution of Fos immunoreactive neurons in wild type (C, G), reeler (D) and dab1 mutant (H) mice. Significant differences between wild type and mutant mice, *p<0.05; **p<0.01; ***p<0.001. Scale bar C, D, G, H = 200 μm.

Neither reeler nor dab1 mutants differ from wild type mice in their cold pain responsiveness

To determine whether the altered heat responses generalized to a cold thermal stimulus, we also examined responsiveness on the two-platform cold plate preference assay (Bautista et al., 2007). Figure 5A and B show that reeler and dab1 mutants do not differ from their respective controls in their responsiveness to noxious cold stimulation (15°C, 10°C and 5°C).

Open in a separate window

Fig. 5

Cold and visceral pain responsiveness is unchanged in reeler and dab1 mutants. A, B: Regardless of temperature, wild type, reeler and dab1 mutants spent equivalent time on the coldest plate (compared to the adjacent plate). C: Number of abdominal stretches induced by i.p. acetic acid did not differ between wild type, reeler, or dab1 mutant mice. D: The number of Fos immunoreactive cells induced by acetic acid in the thoracic superficial dorsal horn did not differ between wild type and dab1 mutant mice. p>0.05 for all comparisons; error bars reflect SEM.

Positioning errors in the LSN and lateral reticulated area contribute to nociceptive abnormalities in reeler and dab1 mutants

Using Fos expression as a marker of activity, we found that intense heat and mechanical stimulation activated fewer LSN neurons in reeler and dab1 mutants than in their respective controls. This result is consistent with our present findings of fewer Dab1-positive LSN neurons in reeler mutants and with previous reports that reeler and dab1 mutants have fewer total neurons in the LSN (Villeda et al., 2006, Akopians et al., 2008). The “missing” neurons are likely displaced to other areas, including laminae I–II. The LSN neurons (in wild type mice) almost certainly correspond to those originally described in the dorsal lateral funiculus by Gwyn and Waldron (1968) and which subsequent studies demonstrated project widely to the brainstem, hypothalamus and thalamus (Menétrey et al., 1982, Burstein et al., 1990a, Burstein et al., 1990b). Taken together with more recent studies reporting that LSN neurons both receive and integrate nociceptive information (Olave and Maxwell, 2004, Pinto et al., 2010), we conclude that the LSN positioning errors in Reelin signaling pathway mutants likely contribute to the abnormalities in pain processing.

We also recorded a striking reduction of noxious stimulus (heat, mechanical and formalin)-evoked Fos expression in cells located in the lateral reticulated area of the deep dorsal horn, adjacent to the dorsolateral funiculus. And in a remarkable correspondence, we found that prominent Dab1-immunoreactive neurons are common in the lateral reticulated area of wild type mice, but they are reduced in number and presumably mispositioned in reeler mutants. Furthermore, mechanical and noxious heat stimulation induced Fos in a subset of these large Dab1 neurons, but only in wild type mice. Although the physiology of these Dab1-positive neurons is unknown, based on their size and location they likely correspond to the nociresponsive neurons described by Menétrey et al. (1980). As many neurons in this region project to thalamic targets, as do the majority of lateral lamina V neurons (Burstein et al., 1990b, Willis et al., 2001, Craig, 2003, Basbaum et al., 2009), their incorrect positioning in mutants probably contributes to reduced pain responsiveness. Because the Reelin-signaling pathway is required for postnatal dendritic development (Niu et al., 2004, Matsuki et al., 2008), mispositioned Dab1-labeled lamina V neurons would likely have stunted dendrites and, in addition, altered synaptic function in adults (Herz and Chen, 2006; Forster et al., 2010).

Large lateral lamina V neurons respond to both thermal and mechanical noxious stimulation (Menétrey et al., 1980, Craig, 2003), as do the Dab1 lamina V neurons that co-express Fos. It was, therefore, surprising that the loss of Reelin signaling only reduced mechanical pain responsiveness. An interesting possibility is that heightened thermal (heat) pain responsiveness reflects additional positioning errors, for example, one that places nociresponsive neurons of the LSN in a circuit that is more directly, and perhaps selectively, engaged by noxious heat responsive, TRPV1-expressing primary afferent nociceptors. To date we have documented a polysynaptic connection from nociceptors to neurons of the LSN, but we did not specifically address the connectivity of the TRPV1 subsets of nociceptors (Braz and Basbaum, 2009). The fact that heat and mechanical (and chemical) pain processing can be influenced in opposite directions, of course, supports the view that despite the presence of polymodal neurons in the dorsal horn, there may be considerable behaviorally relevant modality segregation of heat and mechanical pain processing beyond the primary afferent nociceptors. That cold and visceral pain responsiveness are unchanged in the mutant mice strongly supports that view.

This work was supported by the National Science Foundation (IOB-0518714 and IOB-0924143 to PEP) and the National Institute of Health (NS14627 and DA29204 to AIB). We thank Autumn Abadesco and Marianne Cilluffo for excellent technical assistance.