Mispositioned Neurokinin-1 Receptor-Expressing Neurons Underlie Heat Hyperalgesia in Disabled-1 Mutant Mice

Visual Abstract


Introduction
Reelin (Reln) is a secreted glycoprotein that binds both the apolipoprotein E receptor 2 (Apoer2) and the verylow-density lipoprotein (Vldl) receptors. Reln binding, in turn, leads to receptor clustering, followed by recruitment and phosphorylation of the intracellular adaptor protein Disabled-1 (Dab1) by Src-family kinases (D'Arcangelo et al., 1999;Herz and Chen, 2006). Mice with mutations in Reln, or both Apoer2 and Vldlr, or Dab1 have similar patterns of mispositioned neurons generated during nervous system development Trommsdorff et al., 1999). The Reln-Dab1 signaling pathway regulates neuronal positioning in the spinal cord in a cell-specific manner. Neurons that express Dab1 sustain migratory errors in mutant mice, whereas other neurons are usually appropriately positioned (Phelps et al., 2002).
Reln-expressing and Dab1-expressing neurons are concentrated in the superficial dorsal horn (Laminae I-II) and in the lateral reticulated area of Lamina V, where nociceptive afferents terminate and contain projection neurons that integrate and convey nociceptive signals to supraspinal targets (Villeda et al., 2006;Akopians et al., 2008;Basbaum et al., 2009;Todd, 2010). The lateral spinal nucleus (LSN), a small area embedded in the dor-solateral funiculus that transmits noxious information to the brainstem and thalamus (Burstein et al., 1990;Olave and Maxwell, 2004), also contains Reln-labeled and Dab1-labeled neurons (Villeda et al., 2006). Importantly, functional abnormalities characterize Reln -/and dab1 -/mice: an increased sensitivity to noxious heat (hyperalgesia) and a pronounced reduction in mechanical and chemical sensitivity. Cold and visceral pain transmission, however, appeared normal (Villeda et al., 2006;Akopians et al., 2008;Wang et al., 2012). Taken together, these findings indicate that the biological basis of different pain modalities is segregated in mutants of the Reln-signaling pathway.
Despite these profound nociceptive abnormalities in the processing of pain messages, the alterations in the dorsal horn that underlie these opposite sensitivities to heat and mechanical stimuli are unclear. Although the termination pattern of nociceptive primary afferents is generally similar in wild-type and mutant mice, Reln -/-Lamina I is compacted and the area of Laminae I-II outer is smaller than in Reln ϩ/ϩ mice (Villeda et al., 2006;Yvone et al., 2017). In addition, the isolectin B4 (IB4) terminal field that defines Lamina II dorsal inner appears shifted dorsally in Reln -/yet contains greater numbers of Dab1-expressing neurons than found in Reln ϩ/ϩ mice (Yvone et al., 2017).
More relevant to the heat hypersensitivity is our report of misplaced Dab1 cells that express Neurokinin-1 receptors (NK1Rs) in Reln -/and Apoer2 -/-/Vdlr -/mice (Villeda et al., 2006;Akopians et al., 2008). The NK1R-expressing neurons are targeted by the peptidergic (substance P-expressing) primary afferents (C-fibers; Al Ghamdi et al., 2009), respond to noxious heat stimulation, and the great majority are projection neurons that transmit noxious thermal information rostrally (Littlewood et al., 1995;Marshall et al., 1996;Spike et al., 2003). After ablation of NK1R-expressing neurons with the cytotoxic substance P-saporin conjugate (SP-SAP), responses to acute pain were unchanged but the thermal hyperalgesia associated with chronic pain was reduced (Mantyh et al., 1997;Nichols et al., 1999). Here, we asked whether the noxious heat hypersensitivity detected in Reln -/and dab1 -/mice is related to the mispositioned neurons that express NK1Rs.
We first examined the laminar distribution of NK1Rexpressing neurons in wild-type and mutant Reln and dab1 dorsal horns and then asked whether the NK1Rexpressing neurons were part of the Dab1 and Lmx1b populations. Next, we asked whether there were fewer NK1R-positive cells that expressed the immediate early gene, c-fos, after noxious heat stimulation in dab1 ϩ/ϩ compared to dab1 -/-Laminae I-II and LSN. Finally, to determine whether mispositioned NK1R-expressing neurons are responsible for the heat hypersensitivity in dab1 -/mice, we selectively ablated the superficially located NK1R-expressing neurons. We report that the ablation of NK1R-expressing neurons abolished the thermal hypersensitivity in dab1 -/mice, without altering their mechanical deficits.

Animals
All experiments were reviewed and approved by the Chancellor's Animal Research Committee at University of California, Los Angeles (UCLA) or the Institutional Care and Animal Use Committee at the University of California, San Francisco and conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory animals. Adult reeler (B6C3Fe -ala-Reln rl , The Jackson Laboratory) and dab1 KO mice (BALBcByJ dab1 -/-, gifts from Dr. J. Cooper, Fred Hutchinson Cancer Research Center, Seattle, WN, and Dr. B. Howell, SUNY Upstate Medical University, Syracuse, NY;Howell et al., 1997) were bred at UCLA and genotyped using polymerase chain reaction as described by D' Arcangelo et al. (1996) and Brich et al. (2003). For anatomic analyses, adult mice (4 -12 months) of either sex were deeply anesthetized with Pentobarbital (80 mg/kg), perfused with 4% paraformaldehyde, and postfixed for 3-4 h in the same fixative at 4°C. Lower lumbar spinal cord segments (L4-5) were dissected for all experiments and cervical enlargements (C5-8) were collected for studies in which we ablated NK1R-expressing neurons. Spinal cords were infiltrated with 30% sucrose, embedded and stored at -80°C.

Behavioral assays Thermal (heat)-induced Fos expression
We used six dab1 ϩ/ϩ and six dab1 -/age-matched male mice to compare the positions of the NK1R-expressing neurons activated by noxious heat. As the dab1 -/mice are larger, healthier, and had higher Fos expression than the Reln -/mice, we used dab1 mice for testing (Wang et al., 2012). Mice were lightly anesthetized with pentobarbital (50 -60 mg/kg) and 15 min later the left hindpaw was dipped into a 50°C water bath for 3 s/min for 10 min. One hour later, the mice were anesthetized (80 -100 mg/ kg) and perfused.

Hargreaves test of thermal (heat) sensitivity
Male dab1 mice, seven of each genotype, were tested with the Hargreaves' paw withdrawal test of heat sensitivity (Cao et al., 1998). Mice were placed in plastic cham-bers on a warm glass surface for 1 h, after which we focused a radiant heat source on the left hindpaw, and recorded the withdrawal latency.

von Frey test of mechanical sensitivity
Male dab1 mice (seven of each genotype) were placed in clear plastic chambers on an elevated wire platform and habituated for 1 h. We used the up-down method (Chaplan et al., 1994) with calibrated monofilaments (0.008 -4 g) to determine the mechanical withdrawal threshold. After the 0.4 filament was applied to the center of the left hindpaw, responses were recorded in series, and the 50% withdrawal threshold was determined.

Substance P analog conjugated saporin (SSP-SAP)
To determine whether mispositioned neurons that express NK1Rs in Laminae I-II and the LSN contribute to the heat hypersensitivity of dab1 -/mice, we selectively ablated these neurons by intrathecal injections of a peptidase-resistant analog of substance P [Sar 9 , Met (O 2 ) 11 , SSP] conjugated to the cytotoxic agent saporin Advanced Targeting Systems). Initially we tested wild-type mice with a range of concentrations (100 -400 ng in 1-l sterile PBS) for lumbar level intrathecal injections of SSP-SAP or control, unconjugated SAP. We found that 300 ng of SSP-SAP effectively killed NK1R-expressing neurons in lumbar superficial dorsal horn; the same concentration of unconjugated SAP had no effect.
Next, we examined thermal and mechanical responses using the Hargreaves and von Frey tests, respectively, both before (baseline) and after the ablations. We injected seven male dab1 ϩ/ϩ and seven dab1 -/mice intrathecally with 300 ng of SSP-SAP. One month after SSP-SAP treatment we repeated the tests of heat and mechanical sensitivity with seven dab1 ϩ/ϩ and six dab1 -/mice. Mice were then perfused and their spinal cords immunostained for NK1Rs.

Hot plate test of thermal sensitivity
We also tested the mice in a model of heat pain sensitivity that is presumed to involve circuits in the brain and spinal cord. Unfortunately, because of profound motor instability or "reeling" of the mutants, these tests proved very difficult. By videotaping the mice and observing different behavioral responses on the hot plate, however, we were able to identify behavioral endpoints that distinguished wild-type and mutant mice. We conducted the hot plate test (1440 Analgesia Hot Plate, Columbus Instruments) using 22 dab1 female mice, evenly divided between genotypes. After a 30-min habituation period, mice received three trials each at 33°C baseline, 48°C, 52.5°C, and 55°C. Time to the first positive response was recorded, based on standard responses (shaking, licking, shielding hindpaw, or jumping), and alternate responses observed primarily, but not exclusively, in mutants, such as licking the trunk or leg, or intense grooming (anxious shaking or grooming that lasted Ն3 s or occurred three times). Unfortunately, we were not able to assess these mice after ablation of dorsal horn NK1R-expressing neurons, so we only report results from the untreated mice.

Experimental design and statistical analyses
The initial analysis of NK1R-immunoreactive neurons examined one dorsal horn from ϳ25 sections/mouse in 10 Reln and eight dab1 male or female mice, evenly divided by wild-type and mutant genotypes. Sections were photographed (Olympus AX70) and images were used to record the NK1R-expressing neurons identified with the microscope (40ϫ). Dorsal horns were divided into Laminae I-II, III-IV, and V-VI plus the LSN (as in Wang et al., 2012) and NK1R-expressing cells with one or more processes were counted. Values across sections from the same mouse and location were averaged to generate one value per mouse and location. Means between wild-type and mutant mice were compared at each location using the post hoc t comparisons under a two-way (genotype ϫ location) repeated measures (mixed-random effects) ANOVA model (JMP Software version 13, SAS Institute). The Fisher LSD criterion was used to control for multiple comparisons under this model.
To quantify subtypes of NK1R-expressing neurons that co-express Dab1 and/or Lmx1b in Laminae I-II, we analyzed six dorsal horn hemisections/mouse from six Reln ϩ/ϩ and six Reln -/mice by collecting confocal images at 40ϫ with a Zeiss LSM800 microscope. Confocal slices (3 m) were used to identify subtypes of NK1R cells: NK1R-only, NK1R-Dab1, NK1R-Lmx1b, and NK1R-Dab1-Lmx1b neurons. Means were compared between the two genotypes across four NK1R subsets using a 2 ϫ 4 repeated measure ANOVA model as above. The p values for comparing means of total NK1R neurons between the two genotypes were computed using a t test.
The Fos/NK1R analysis was conducted on three to five dorsal horns per mouse in six dab1 ϩ/ϩ and six dab1 -/male mice. Four to five 25ϫ confocal images were acquired (Zeiss LSM 510) for the Laminae I-II and LSN analyses. To image a total of 6-to 8-m/dorsal horn, we collected serial 3 m slices. Cell counts were conducted from 3-D reconstructions of each confocal slice with Neurolucida (v10; MicroBrightField RRID:nif-0000-10294). Double-labeled neurons per area per mouse were averaged and means between wild-type and mutant mice were compared using a two-way (genotype ϫ cell type) repeated measures ANOVA and post hoc t test.
The dab1 (seven wild-type and seven mutant) mice were examined using the Hargreaves and von Frey tests of thermal heat and mechanical sensitivity, respectively, before and after SSP-SAP treatment (one mutant died before retest). Means from the Hargreaves test were com-pared with a two-way (genotype ϫ treatment) repeated measure ANOVA as above. The Mann-Whitney test was used to compare means of baseline and post SSP-SAP mechanical thresholds from the von Frey in the two genotypes, as these data do not have a normal distribution. The Wilcoxon matched pairs signed-rank test evaluated differences of treatments within genotypes (Sigma Plot 12, Systat Software Inc.).
As a final experiment, we tested dab1 mice on the hot plate test to extend our Hargreaves analyses and better interpret the heat responses of the mutant mice. Videos were analyzed blind to temperature, but the reeling behavior of mutants precludes blinding to genotype. Results of the three trials at each temperature were averaged for each of the 11 dab1 ϩ/ϩ and 11 dab1 -/female mice tested. Means were compared between dab1 ϩ/ϩ and dab1 -/mice with a two-way (genotype ϫ temperature) repeated measures ANOVA, followed by a post hoc t test.
To determine whether the numbers and subtypes of the adult NK1R-expressing neurons differ between wild-type and mutant mice, we quantified these neurons in Laminae I-II. Reln -/mice had greater numbers of total NK1R-expressing neurons in the superficial dorsal horn compared to Reln ϩ/ϩ mice (Reln ϩ/ϩ 13.8 Ϯ 0.7 neurons/  ϭ 100 m (A, B, E-G), 200 m (C, D).

The pattern of noxious heat-induced Fos expression in dab1 ؉/؉ and dab1 -/mice parallels the distribution of NK1R-expressing neurons
In the LSN, we determined that the 6 dab1 ϩ/ϩ mice had 1.3 Ϯ 0.2 Fos-immunoreactive NK1R neurons per section compared to only 0.5 Ϯ 0.1 double-labeled neurons in the 6 dab1 -/mice (p ϭ 0.0062; Fig. 5C). The percentage of total Fos-immunoreactive cells that bear NK1Rs in the LSN again is similar in both genotypes, with 49% of Fos-labeled cells being double-labeled. Of the total NK1R-immunoreactive neurons, 42% of dab1 ϩ/ϩ and 27% of dab1 -/-LSN were double-labeled after heat stimulation. These percentages reflect the overall loss of neurons in the mutant LSN (Villeda et al., 2006;Akopians et al., 2008;Wang et al., 2012).
We also observed that heat stimulation activated mispositioned NK1R-expressing cells in Reln -/superficial dorsal horn. For example, Figure 6A,C illustrates a Fos-immunoreactive NK1R-expressing cell near the border of Laminae II-III, in a similar position to the large NK1R-Dab1-Lmx1b neuron mispositioned in Reln -/mice ( Fig. 3A-A4). The dorsal white matter in the Reln -/mice also contained NK1R-immunoreactive processes as well as double-labeled Fos-NK1R-expressing neurons (Fig. 6A,B), and represents another area where ectopic NK1R-expressing neurons are found (Fig.  3C-C3). Together, these findings demonstrate that mispositioned NK1R-expressing neurons do indeed respond to noxious heat stimulation in both the Reln -/and dab1 -/dorsal horns.

Ablation of NK1R-expressing neurons eliminates the heat hypersensitivity of dab1 -/mice
As first reported (Akopians et al., 2008), we again found that the dab1 -/mice have a reduced latency to withdraw from noxious heat in the Hargreaves test. For the present cohort of mice, the baseline withdrawal average latency for dab1 ϩ/ϩ mice was 12.2 Ϯ 0.9 s versus dab1 -/mice at 9.3 Ϯ 0.8 s (p ϭ 0.025, n ϭ 7 mice/genotype; Fig. 7A). After establishing baseline responsiveness, all mice received a single lumbar intrathecal injection of SSP-SAP and one month later we retested withdrawal latencies. Compared to baseline values, both dab1 ϩ/ϩ (15.7 Ϯ 0.82 s, n ϭ 7, p ϭ 0.0069) and dab1 -/mice (dab1 -/-15.83 Ϯ 0.88 s, n ϭ 6; p ϭ 0.0001; Fig. 7A) exhibited increases in their withdrawal latency, but most importantly, increases were significantly greater in the dab1 -/mice. In addition, the mean values of both genotypes did not differ after the SSP-SAP treatment (Fig. 7A). These results show that mispositioned NK1R-expressing neurons are required for the thermal (heat) hypersensitivity of dab1 -/mice. We presume that our finding of equivalent post treatment latencies reflects the fact that both correctly and incorrectly positioned NK1R-expressing neurons were ablated in the dab1 ϩ/ϩ and dab1 -/mice.
To determine the modality specificity of the behavioral changes observed, we also asked whether the loss of superficially located NK1Rs altered mechanical sensitivity. The baseline testing data replicated our previous results (Akopians et al., 2008) of increased thresholds in the mutant mice. The initial mean 50% threshold for the dab1 ϩ/ϩ mice was 0.69 Ϯ 0.02 g (n ϭ 7) compared to dab1 -/mice at 1.35 Ϯ 0.06 g (n ϭ 7; p ϭ 0.0159; Fig.7B). One month following SSP-SAP treatment, the withdrawal thresholds also differed between genotypes (dab1 ϩ/ϩ 0.80 Ϯ 0.02, n ϭ 7; dab1 -/-1.35 Ϯ 0.04; n ϭ 6; p ϭ 0.0034; Fig 7B). The mean 50% threshold for both dab1 ϩ/ϩ (0.80 Ϯ 0.02 g, n ϭ 7 p ϭ 0.063) and dab1 -/mice (1.35 Ϯ 0.04 g, p ϭ 0.999; n ϭ 6) after treatment, however, did not vary from their baseline values (Fig. 7B). We conclude that ablation of NK1R-expressing neurons does not alter the profound mechanical insensitivity seen in dab1 -/mice. It follows that the increased mechanical withdrawal thresholds are likely due to other groups of mispositioned neurons in the superficial dorsal horns of the dab1 -/mice. Importantly, to document the ablation of NK1R neurons after SSP-SAP, we compared the level of NK1R immunoreactivity in lumbar (directly targeted) and cervical sections from ablated mice to that in untreated spinal cord sections. Figure 8 shows representative levels of NK1R expression in cervical and lumbar spinal cord from untreated mice and after SSP-SAP treatment. NK1R expression was noticeably decreased in lumbar Lamina I and LSN after SSP-SAP (Fig. 8D,F) compared to the same Figure 5. Noxious heat induces Fos expression in dab1 ϩ/ϩ and dab1 -/dorsal horn. A, An example of the confocal images used to analyze NK1R-Fos expression in dab1 -/-Laminae I-II has a large misplaced NK1R-Fos-labeled cell body in Lamina I. B, After thermal stimulation dab1 ϩ/ϩ mice (black circles) have fewer Fos-expressing neurons in Lamina I-II than dab1 -/mice (red circles, ‫‪p‬ءء‬ ϭ 0.0028). Total NK1R-expressing neurons did not vary by genotype. NK1Rs and Fos-expressing neurons in Laminae I-II were less numerous in dab1 ϩ/ϩ than in dab1 -/mice ‫‪p‬ءء(‬ ϭ 0.0097). C, After heat stimulation, the total number of Fos-labeled neurons in the LSN was higher in dab1 ϩ/ϩ than dab1 -/mice ‫‪p‬ءءء(‬ ϭ 0.00064). The total NK1R-and NK1R-Fos-expressing neurons also were more numerous in dab1 ϩ/ϩ than dab1 -/-LSN (‫ء‬p ϭ 0.0113, ‫‪p‬ءء‬ ϭ 0.0062, respectively). Circles (dab1 ϩ/ϩ are black and dab1 -/are red) represent individual mice and diamonds show group mean Ϯ SEM values. Scale bar ϭ 50 m (A). areas in untreated mice (Fig. 8B). Somewhat unexpectedly, NK1R expression in cervical enlargement sections of treated mice was also reduced (Fig. 8C,E) compared to untreated mice (Fig. 8A), but to a lesser extent than in lumbar cord. This presumably resulted from rostral diffusion of the SSP-SAP reagent. Together, these results confirm the efficacy of the SSP-SAP treatment and suggest that the mispositioned NK1Rs in dab1 -/mice are responsible for the thermal hyperalgesia.
Finally, we asked whether the heat hypersensitivity revealed using the Hargreaves test, which likely reflects increased excitability of circuits that underlie a spinallymediated withdrawal reflex response, is also manifest in the hot plate test, which we believe involves supraspinal as well as spinal cord processing of "pain" messages. Consistent with our hypothesis, Figure 9 shows that despite the coordination difficulties characteristic of these mutant mice, there was a significant decrease of the latency to respond to the 55°C hot plate stimulation (dab1 ϩ/ϩ 11.5 Ϯ 1.01 s, n ϭ 11; dab1 -/-8.6 Ϯ 0.87 s, n ϭ 11, p ϭ 0.0395).

Discussion
This study addresses the basis for the thermal hyperalgesia that occurs in Reln-signaling mutants. First, we found increased numbers of neurons that express NK1Rs in dorsal horn Laminae I-II in Reln -/and dab1 -/and fewer in mutant LSN compared to their respective wild-type mice. We then showed that some NK1R neurons co-express Dab1 and Lmx1b, confirming their excitatory phenotype (Cheng et al., 2004(Cheng et al., , 2005Dai et al., 2008;Szabo et al., 2015;Yvone et al., 2017). Next, we showed that noxious heat stimulation induced Fos-expression in significantly greater numbers of NK1R-immunoreactive neurons in dab1 -/-Laminae I-II and fewer in dab1 -/-LSN than in dab1 ϩ/ϩ mice. Importantly, based on their incorrect position in Lamina I, in the dorsal white matter, or near the Laminae II-III border, we conclude that at least some of the incorrectly positioned NK1R neurons in Reln -/and dab1 -/mice express Fos in response to noxious heat stimulation, which establishes that these neurons are functional. By ablating the superficial NK1Rbearing neurons with intrathecal SSP-SAP, we eliminated the thermal hyperalgesia of the dab1 -/mice, without altering their mechanical sensitivity. We conclude that misplaced NK1R-expressing neurons, specifically those that populate the superficial laminae of the dorsal horn, underlie the heat hypersensitivity that characterizes the Reln-signaling pathway mutants.
NK1R-expressing neurons constitute the majority of Lamina I projection neurons, are excitatory, and transmit moderate to intense pain-provoking information to supraspinal, brainstem and thalamic targets (Littlewood et al., 1995;Marshall et al., 1996;Todd et al., 2005;Todd, 2010). Another population of more ventrally located NK1R-expressing projection neurons has dorsally-directed dendrites that intermingle with the Lamina I projection neurons (Brown et al., 1995; Todd, 2010). Both populations of NK1R projection neurons are targeted by transient receptor potential vanilloid 1 (TRPV1)-expressing primary afferent fibers that respond to noxious heat stimulation. In fact, because intrathecal capsaicin-induced ablation of TRPV1 terminals completely abrogated both acute noxious heat-evoked behaviors, as well as the heat hyperalgesia that occurs in the setting of tissue injury, we previously concluded that the TRPV1 afferents provide the predominant, and likely the exclusive noxious heat input to the dorsal horn (Cavanaugh et al., 2009).
We do not believe, however, that an increased afferent input to the altered dorsal horn circuits underlies the heat hypersensitivity of these mutant mice. In fact, we observed neither a noticeable change in the number of TRPV1-expressing dorsal root ganglion cell bodies nor in the pattern of TRPV1 afferent terminal expression in the lumbar superficial dorsal horn between the mutant and wild-type mice. We appreciate, however, that we have not ruled out possible sprouting of individual TRPV1-positive afferents in the mutant mice. Sprouting could provide a more distributed heat input to the mispositioned NK1R neurons in Lamina I-II, the consequence of which would be greater heat pain information transfer to the brain and to spinal reflex withdrawal circuits. Taken together, we conclude that the increased heat pain sensitivity in the mutant mice reflects the increased number of NK1Rexpressing neurons, many of which are mispositioned, but nevertheless receive a TRPV1 input comparable to that of the wild-type mice. Of course, an increase in the number of NK1R-expressing projection neurons would also enhance the feedforward spino-bulbo-spinal facilitatory loops that increase the excitability of nociresponsive neurons in deep dorsal horn (Rahman et al., 2008), many of which engage heat-induced withdrawal reflex circuitry. As both the hot plate and Hargreaves tests revealed increased sensitivity to heat in the mutant mice, we conclude that mispositioned NK1R-expressing projection and interneurons respond to the stimulation.
Unresolved is the question concerning the consequence of mispositioning of NK1R-expressing neurons of the LSN to Laminae I-II. Earlier studies reported that some LSN neurons, which unlike Lamina I neurons do not receive a direct input from primary sensory neurons, express NK1Rs, project to the ventrolateral medulla, and induce Fos expression in response to noxious thermal stimulation (Olave and Maxwell, 2004). Taken together with the present results we suggest that the heat hypersensitivity phenotype manifest in the mutant mice results from mispositioned NK1R neurons, some of which originated in the LSN and would therefore receive substance P Figure 7. SSP-SAP eliminates the thermal hypersensitivity of dab1 -/mice. A, Hargreaves test results illustrate that baseline values (green circles) for the dab1 ϩ/ϩ and dab1 -/withdrawal responses differ by genotype (‫ء‬p ϭ 0.0245). After SSP-SAP treatment (black circles), the dab1 ϩ/ϩ ‫‪p‬ءء(‬ ϭ 0.0069), and Figure 7. continued dab1 -/-‫‪p‬ءءء(‬ Ͻ 0.0001) responses were reduced compared to their respective baseline values. B, The von Frey responses in dab1 ϩ/ϩ and dab1 -/mice do not change with SSP-SAP treatment. Compared to the dab1 ϩ/ϩ , dab1 -/mice have much slower responses both before (‫ء‬p ϭ 0.0159) and after SSP-SAP treatment (‫ء‬p ϭ 0.0034). Circles represent individual mice and diamonds show group mean Ϯ SEM. New Research inputs from dorsally located interneurons rather than primary sensory afferents (Gutierrez-Mecinas et al., 2018). Interestingly, Sikandar et al. (2017) found that LSN neurons predominantly process input from deep, rather than cutaneous tissues. In future studies, it will be of interest to assess the responsiveness of the mutant mice to deep tissue, including muscle, injury.

Mispositioned NK1R-expressing neurons are functional
Importantly, our recording of heat-provoked Fos induction in mispositioned NK1R neurons in the superficial dorsal horn illustrates that the synaptic circuits targeted by the complement of TRPV1 afferents is maintained. This conclusion is consistent with other reports that misposi- tioned neurons in Reln -/mice receive substantial functional synaptic input from the same sources as in wildtype mice (Caviness, 1976;Terashima et al., 1987;Yip et al., 2003), although the dendritic arbor of the mispositioned Dab1-labeled neurons is often stunted, as seen in the hippocampus (Niu et al., 2004). For example, retrograde and anterograde tracing studies comparing Reln ϩ/ϩ and the essentially inverted Reln -/cerebral cortex found evidence of similar corticospinal (Terashima et al., 1983) and corticocortical (Caviness and Yorke, 1976;Terashima et al., 1985) connections, plus reciprocal connections of motor, somatosensory, and visual cortices with their appropriate thalamic targets (Caviness and Frost, 1983;Terashima et al., 1987). In the spinal cord, the supraspinal and primary afferent innervation of mispositioned Reln -/sympathetic preganglionic neurons is also organized correctly (Yip et al., 2003). Thus, if migrating NK1R projection neurons from lateral Lamina V in a Reln -/mouse failed to stop, and instead continued circumferentially to settle in Laminae I-II, these projection neurons would likely continue to receive substantial synaptic input from the rather large population of substance P-expressing interneurons (Liu et al., 1994;Xu et al., 2008;Dickie et al., 2019) .

Specificity of dorsal horn circuits that process noxious heat and mechanical information
Of particular interest are the divergent effects of the Reln-signaling pathway mutations on heat and mechanical pain processing. Rats treated with SP-SAP and tested one month later had marked attenuation of both thermal and mechanical hyperalgesia in chronic pain models, but no changes detected in acute pain tests (Mantyh et al., 1997;Nichols et al., 1999). In contrast, we detected a small, but significant increase of heat pain thresholds after SSP-SAP treatment in the dab1 ϩ/ϩ mice, possibly because SSP-SAP is a more effective toxin than SP-SAP (Wiley, 2008) or that mice differ in their response com-pared to rats (Mantyh et al., 1997;Nichols et al., 1999;Suzuki et al., 2002). Based on the previous SP-SAP studies, one would conclude that the NK1R cells contribute to and may be required for both heat and mechanical hyperalgesia. That may be the case in wild-type, but the results in dab1 -/mice demonstrate that activity of NK1R cells, even when their number is increased, does not lead to increased mechanical hypersensitivity. In fact, the dab1 -/mice have significantly decreased mechanical sensitivity.
In this regard, Reln -/and dab1 -/mice bear considerable similarity to mice with a deletion of the testicular orphan nuclear receptor (TR4), which have increased mechanical thresholds and a profound loss of excitatory interneurons in the superficial dorsal horn, some of which express Reln (Wang et al., 2013). More recently, Szabo et al. (2015) reported decreased mechanical sensitivity in the conditional Lmx1b mutant mouse, which also is missing excitatory interneurons, including Reln-labeled cells. In addition, both of these mutants, as well as Reln -/mice, have a reduction in the size of Laminae I-II outer and altered lamination patterns (Wang et al., 2013;Szabo et al., 2015;Yvone et al., 2017). Conceivably, there are comparable interneuronal abnormalities in Reln -/mice that cause a loss of excitatory noxious mechanical input from mechanically responsive interneurons in the superficial dorsal horn. We previously demonstrated that ablation of the IB4 population of sensory neurons is associated with a remarkably selective increase in mechanical thresholds (Cavanaugh et al., 2009). Taken together, these results suggest that mispositioning of NK1R neurons does not underlie the increased mechanical thresholds in the Reln -/mice. Rather, this deficit likely reflects circuit abnormalities among superficial dorsal horn interneurons that are relatively selective for the processing of mechanical pain information (Peirs et al., 2015;Petitjean et al., 2015;Peirs and Seal, 2016;Cheng et al., 2017). That conclusion is consistent with many recent studies emphasizing that, at least at the level of dorsal horn circuitry, there is considerable specificity in the processing of nociceptive heat, nociceptive mechanical and pruritoceptive information (Kardon et al., 2014;Petitjean et al., 2015;Cheng et al., 2017). Figure 9. dab1 -/mice respond to 55°C stimulation on a hot plate faster than dab1 ϩ/ϩ mice. The hot plate test examined latency responses to 48°C, 52.5°C, and 55°C. Differences between genotypes were limited to 55°C, with dab1 -/responding more quickly than dab1 ϩ/ϩ mice (‫ء‬p ϭ 0.0395). Circles (black for dab1 ϩ/ϩ and red for dab1 -/-) represent individual mice and diamonds mark group mean Ϯ SEM.