Biological constraints limit the use of rapamycin-inducible FKBP12-Inp54p for depleting PIP2 in dorsal root ganglia neurons

Background Rapamycin-induced translocation systems can be used to manipulate biological processes with precise temporal control. These systems are based on rapamycin-induced dimerization of FK506 Binding Protein 12 (FKBP12) with the FKBP Rapamycin Binding (FRB) domain of mammalian target of rapamycin (mTOR). Here, we sought to adapt a rapamycin-inducible phosphatidylinositol 4,5-bisphosphate (PIP2)-specific phosphatase (Inp54p) system to deplete PIP2 in nociceptive dorsal root ganglia (DRG) neurons. Results We genetically targeted membrane-tethered CFP-FRBPLF (a destabilized FRB mutant) to the ubiquitously expressed Rosa26 locus, generating a Rosa26-FRBPLF knockin mouse. In a second knockin mouse line, we targeted Venus-FKBP12-Inp54p to the Calcitonin gene-related peptide-alpha (CGRPα) locus. We hypothesized that after intercrossing these mice, rapamycin treatment would induce translocation of Venus-FKBP12-Inp54p to the plasma membrane in CGRP+ DRG neurons. In control experiments with cell lines, rapamycin induced translocation of Venus-FKBP12-Inp54p to the plasma membrane, and subsequent depletion of PIP2, as measured with a PIP2 biosensor. However, rapamycin did not induce translocation of Venus-FKBP12-Inp54p to the plasma membrane in FRBPLF-expressing DRG neurons (in vitro or in vivo). Moreover, rapamycin treatment did not alter PIP2-dependent thermosensation in vivo. Instead, rapamycin treatment stabilized FRBPLF in cultured DRG neurons, suggesting that rapamycin promoted dimerization of FRBPLF with endogenous FKBP12. Conclusions Taken together, our data indicate that these knockin mice cannot be used to inducibly deplete PIP2 in DRG neurons. Moreover, our data suggest that high levels of endogenous FKBP12 could compete for binding to FRBPLF, hence limiting the use of rapamycin-inducible systems to cells with low levels of endogenous FKBP12.


Background
The immunosuppressant macrolide, rapamycin, induces the dimerization of two naturally occurring protein domains: FK506 Binding Protein 12 (FKBP12) with the FKBP Rapamycin Binding (FRB) domain of mTOR [1]. These domains can be attached to other proteins to temporally and spatially control cell signaling with rapamycin or rapamycin analogs. For example, these domains were used to control cell growth and cell death [2], to translocate proteins to the plasma membrane or nucleus [3][4][5], and induce G protein-coupled receptor (GPCR) signaling [6].
Additionally, two groups used these domains to directly and selectively deplete the lipid PIP 2 in cultured cells [3,4] and show that PIP 2 was important for GPCR signaling and ion channel function [7][8][9]. Both groups used 1) a plasma membrane-anchored FRB domain and 2) a cytosolic PIP 2 -specific phosphatase (yeast Inositol polyphosphate 5-phophatase (Inp54p) or mammalian type IV 5-phosphatase) fused to FKBP12. In cell lines transfected with both of these components, rapamycin promoted dimerization of the FRB domain with FKBP12, and induced rapid translocation of the phosphatase to the plasma membrane where it hydrolyzed PIP 2 . PIP 2 hydrolysis was visualized with a biosensor containing the pleckstrin homology (PH) domain of PLC∂1 (PLC∂1-PH) fused to a fluorescent protein [10][11][12]. This biosensor dissociates from the plasma membrane and enters the cytosol when PIP 2 is hydrolyzed to phosphatidylinositol 4-phosphate (PI(4)P) and inorganic phosphate.
To date, this rapamycin-inducible system has been used in cell lines. Given the widespread importance of PIP 2 in signaling and ion channel function [8,13,14], we hypothesized that this system, if adapted for use in animals, could also shed light on how alterations in PIP 2 affect animal physiology and behavior. For example, PIP 2 modulates Transient Receptor Potential (TRP) ion channels involved in heat and cold sensation, including TRPV1 and TRPM8 [15][16][17][18][19][20][21]. Moreover, we recently found that thermosensation and nociceptive sensitization could be reduced by indirectly decreasing PIP 2 concentration in DRG [22].
Here, we sought to directly and selectively reduce PIP 2 concentration in the plasma membrane of nociceptive DRG neurons to study the in vivo importance of PIP 2 in regulating thermal sensitivity and nociceptive sensitization. To accomplish this goal, we knocked FKBP12-Inp54p fused to a variant of yellow fluorescent protein (Venus) into the CGRPα locus. CGRPα is a marker of peptidergic sensory neurons, a subset of which expresses the thermosensor TRPV1 [23,24]. We generated a second mouse containing a CFP-tagged, membrane-tethered FRB domain knocked into the ubiquitously expressed Rosa26 locus. By crossing both of these mice together, we were able to express both components of the PIP 2 phosphatase system in peptidergic, small diameter DRG neurons and evaluate the performance of this system in vitro and in vivo. Unfortunately, we found that Venus-FKBP12-Inp54p did not translocate to the plasma membrane in DRG neurons following rapamycin treatment. Furthermore, our data suggests that a biological constraint-namely high levels of endogenous FKBP12limits translocation in murine DRG neurons.

Results
Rapamycin induces translocation of Venus-FKBP12-Inp54p from the cytoplasm to plasma membrane in cell lines Before generating knockin mice, we set out to verify that the rapamycin-inducible phosphatase components functioned in our hands as described [3,4]. For these experiments, we modified the FRB-CFP construct described in Varnai et al. (2006) by replacing the native FRB domain with the destabilized FRB PLF mutant [25,26] to generate FRB PLF -CFP. This mutation confers greater sensitivity to rapamycin analogs, like C20-Marap, that can be used in vivo [25,26]. This construct also contains the palmitoylation sequence of human growth associated protein 43 (GAP43), a sequence that promotes plasma membrane localization in cell lines and DRG neurons [4,27]. Additionally, we replaced CFP in the yeast Inp54p construct described in Suh et al. (2006) with a yellow fluorescent protein (Venus) to permit simultaneous visualization of Venus-FKBP12-Inp54p and FRB PLF -CFP in live or fixed cells. The yeast phosphatase was chosen so that it could be immunologically distinguished from endogenous mouse 5-phosphatases.
When cotransfected into human embryonic kidney 293 (HEK293) cells [28], FRB PLF -CFP localized to the plasma membrane ( Figure 1A), Venus-FKBP12-Inp54p was localized to the cytoplasm ( Figure 1B), and PLC∂1-PH-RFP was bound to the PIP 2 -rich plasma membrane ( Figure 1C, quantified in Figure 1G), as expected. The apparent localization of the Venus-FKBP12-Inp54p construct to the plasma membrane at sites of cell-cell contact represents an artifact called "pseudolocalization" [29], and is not actual membrane localization. After treatment with 1 μM rapamycin, there was no change in membrane localization of the FRB PLF domain ( Figure 1D), but Venus-FKBP12-Inp54p translocated to the plasma membrane ( Figure 1E) and hydrolyzed PIP 2 , as evidenced by displacement of PLC∂1-PH-RFP to the cytoplasm ( Figure 1F). Furthermore, rapamycin reduced Gαq-coupled GPCR signaling (using 1 mM carbachol stimulation and calcium imaging as readout; data not shown). Lastly, rapamycin induced translocation of Venus-FKBP12-Inp54p to the plasma membrane in additional cell lines, including Rat1 fibroblasts, HeLa cells, and COS7 cells (data not shown).
The Rosa26 locus ubiquitously drives expression in all cell types, including DRG neurons [32,33]. Thus, we targeted FRB PLF -CFP to the plasma membrane, using a strong CMV early enhancer element and chicken betaactin (CAG) promoter in the Rosa26 locus, to drive higher gene expression than the endogenous Rosa26 promoter alone [34]. Rosa-FRB PLF heterozygous and homozygous mice were viable and fertile. Using immunohistochemistry, we found that FRB PLF -CFP was present on the plasma membrane of approximately 99% of all DRG neurons ( Figure 2C). For all conditions, n > 500 neurons were counted from two animals.
Rapamycin did not induce translocation of Venus-FKBP12 -Inp54p from the cytoplasm to the plasma membrane in DRG neurons We next crossed Rosa-FRB PLF mice with CGRP-Inp54p mice to generate compound heterozygotes that expressed both components in peptidergic DRG neurons ( Figure 3A). Intrathecal (i.t.) injection of rapamycin was previously used to dimerize mTOR-FKBP12 in DRG neurons in vivo [35,36]. In light of this information, we intrathecally injected CGRP-Inp54p +/− (controls) and Rosa-FRB PLF /CGRP-Inp54p compound heterozygous mice with rapamycin (3 injections, spread over 24 h). Animals were perfused shortly after the final rapamycin injection to fix proteins that may have translocated. DRG were then dissected, cryo-sectioned, and mounted for imaging. Translocation of Venus-FKBP12-Inp54p was assessed by quantifying the location of endogenous Venus fluorescence without immunoenhancement, as we were unable to use antibodies to GFP/Venus because they cross-react to CFP, a highly related protein.
Venus and CFP were present in our mice as singlecopy knockins, making it difficult to see endogenous fluorescence without increasing gain, and explaining why signals were weak. Unfortunately, this rapamycin treatment did not induce translocation of Venus-FKBP12-Inp54p from the cytoplasm to the plasma membrane in controls or in double heterozygous mice ( Figure 3A, quantified in Figure 3B), suggesting that this two-component system was non-functional in DRG neurons in vivo.

Rapamycin treatment did not affect behavior in Rosa-FRB PLF /CGRP-Inp54p heterozygous mice in vivo
We recently found that PIP 2 levels in DRG at the time of inflammation enduringly affected thermal hypersensitivity, with preemptively reduced PIP 2 levels correlating with an enduring reduction in thermal hypersensitivity [22]. Thus, we hypothesized that complete Freunds adjuvant (CFA)-induced thermal hypersensitivity would be blunted in Rosa-FRB PLF /CGRP-Inp54p mice if rapamycin caused a significant depletion of PIP 2 . Although we did not observe translocation in rapamycininjected animals, we hypothesized that behavior might be a more sensitive readout of system performance than translocation. First, we performed control experiments with CGRP-Inp54 +/− mice to ensure that expression of this phosphatase alone had no effect on behavioral responses. We tested noxious thermal and mechanical sensitivity in wild-type (WT) and CGRP-Inp54p heterozygous littermates before and after inflaming one hindpaw with CFA, to model chronic inflammatory pain. We found that WT and CGRP-Inp54p +/− mice showed no significant differences in paw diameter, thermal sensitivity or mechanical sensitivity before or after CFA injection ( Figure 4A-C). Thus, expression of Venus-FKBP12 -Inp54p in CGRPα-positive neurons did not alter nociceptive behavioral responses.
Next, we tested the effects of preemptive rapamycin i.t. injections on thermal and mechanical hypersensitivity in CGRP-Inp54p +/− (control) mice and Rosa-FRB PLF / CGRP-Inp54p compound heterozygous mice. After baseline testing, we injected these mice twice with rapamycin (i.t.) at 6 hour intervals, then injected CFA into one hindpaw immediately after the second injection. CFA injections were successful, as evidenced by edema ( Figure 4D), thermal hypersensitivity ( Figure 4E) and mechanical allodynia ( Figure 4F) in the inflamed paw, but not the control paws. However, there were no significant differences between control and experimental groups, further suggesting that this two-component system was non-functional in DRG neurons. Rapamycin treatment stabilized FRB PLF -CFP but did not induce translocation of Venus-FKBP12-Inp54p in cultured DRG neurons We next evaluated whether rapamycin could induce translocation of Venus-FKBP12-Inp54p to the plasma membrane in cultured DRG neurons. To test this, we cultured DRG neurons from Rosa-FRB PLF /CGRP-Inp54p double heterozygous mice, treated cells for 10 min with vehicle or 1 μM rapamycin, then imaged both components (CFP and Venus fluorescence) using confocal microscopy. As with our in vivo studies above, rapamycin treatment did not induce translocation of Venus-FKBP12-Inp54p to the plasma membrane ( Figure 5). We then treated cultured DRG neurons from Rosa-FRB PLF / CGRP-Inp54p double heterozygous for longer periods of time. Unfortunately, we still were unable to detect translocation even after 24 hours ( Figure 6A-B) or 48 hours (data not shown). Notably however, prolonged treatment with rapamycin stabilized FRB PLF -CFP, as evidenced by increased fluorescence signal after 24 hours ( Figure 6A-B, quantified in Figure 6C; all gain settings the same). The FRB PLF domain can be stabilized within hours after dimerizing with endogenous FKBP12 [25,37].
Our data suggested that DRG neurons might contain high levels of endogenous FKBP12 that compete with Venus-FKBP12-Inp54p for binding to FRB PLF -CFP. Moreover, we hypothesized that HEK293 cells might express lower levels of endogenous FKBP12 than DRG neurons, given that Venus-FKBP12-Inp54p did translocate to the membrane in HEK293 cells expressing FRB PLF -CFP ( Figure 1). Indeed, we found that endogenous FKBP12 levels were significantly higher in DRG when compared to HEK293 cells ( Figure 7A-B). Although the level of FKBP12 is only 1.5 higher in total DRG lysate ( Figure 7B), this is likely an underestimation of FKBP12 in DRG neurons due to dilution by nonneuronal DRG cells, as FKBP12 is expressed more highly in neurons than non-neuronal surrounding cells of the DRG ( Figure 7C-D). COS7 cells also contained low levels of FKBP12 (data not shown), possibly explaining why Venus-FKBP12-Inp54p translocated to the plasma membrane in this cell line as well (see above). To delineate the localization of FKBP12, we immunostained DRG sections from WT animals with antibodies to FKBP12. FKBP12 was found throughout the cytoplasm in all neurons, and was often concentrated at the membrane in large diameter DRG neurons ( Figure 7C). Notably, the satellite cells that surround DRG neurons (marked by DRAQ5-positive nuclei) contained lower levels of FKBP12 ( Figure 7C). Likewise, in cultures of dissociated DRG, high levels of FKBP12 were detected in βIII Tubulin + neurons (a neuronal-specific marker), while βIII Tubulin -, DRAQ5 + cells had lower levels of FKBP12 ( Figure 7D; quantified by image intensity analysis; p < 0.0001, data not shown). Thus, FKBP12 was present at high levels in DRG neurons, and at low levels in non-neuronal cells in the DRG.

Discussion
We successfully generated two knockin mice that each expressed components of the rapamycin-inducible PIP 2 depletion system. FRB PLF -CFP and Venus-FKBP12 -Inp54p were expressed in the appropriate cell types and each of these proteins was targeted to the correct subcellular location (membrane and cytoplasm, respectively). While Venus-FKBP12-Inp54p translocated to the membrane in cell lines expressing FRB PLF -CFP, we were unable to detect rapamycin-induced translocation of these components in DRG neurons in vitro or in vivo. Furthermore, rapamycin treatment of double heterozygous mice did not alter thermal sensitivity as we would have expected if the system had worked in vivo. While this chemically-induced translocation tool is widely used for manipulation in cell lines, our data collectively suggest that high levels of endogenous FKBP12 limit its functionality in DRG neurons (Figure 8).
While rapamycin did not induce translocation in DRG neurons, it did enhance CFP-FRB PLF protein fluorescence intensity, suggesting that rapamycin interacted with FRB PLF and promoted dimerization to endogenous FKBP12. The FRB domain mutation used in our Rosa-FRB mouse consists of three point mutations: K2095P, T2098L, and W2101F [26]. These mutations allow for the use of rapamycin analogs that do not cross-react with the wildtype, endogenous FRB domain of mTOR. One of these mutations (T2098L) is responsible for protein destabilization, and this destabilized FRB mutant has been previously used for successful dimerization in vivo [25,26,37]. Interestingly, this destabilization is extended to proteins fused to the FRB PLF mutant, such as fluorescent tags, and is reversed upon FKBP12rapamycin-FRB PLF complex formation, with a half- maximal result approximately 8 hours after rapamycin treatment [25,37]. The fact that FRB PLF -CFP protein levels were increased demonstrates that rapamycin was in fact reaching its intended target in DRG neurons. However, Venus-FKBP12-Inp54p failed to translocate, suggesting that endogenous FKBP12 could be responsible for this stabilization. Indeed, DRG neurons contain higher levels of endogenous FKBP12 than HEK293 cells ( Figure 7). Endogenous FKBP12 in DRG neurons could potentially out-compete Venus-FKBP12-Inp54p for binding to FRB PLF , and thus prevent Venus-FKBP12 -Inp54p from translocating to the membrane. Alternatively, the levels of mTOR, the protein that contains the endogenous FRB domain, could have a similar effect on sequestering the transgenic Venus-FKBP12-Inp54p protein away from the membrane-tagged FRB PLF domain. The rapamycin analog C20-Marap binds to FRB PLF without interacting with endogenous mTOR [26]; therefore, this compound could potentially be used to rule out the role of endogenous mTOR as a source of translocation inhibition. However, it should be emphasized that use of rapalogs would not overcome the problem we identified, namely interaction of FRB with endogenous FKBP12.
Overexpression of Inp54p in cell lines can lead to loss of cell adhesion, induction of membrane blebbing, and ultimately cell death [38,39]. Expression after   transfection in cultured cells tends to be on the timescale of a few days, whereas these mice express Venus-FKBP12-Inp54p throughout the life of the animal. It is possible that some compensation occurs when Inp54p is expressed over longer time scales.
In addition to elevated levels of endogenous FKBP12, other factors might limit rapamycin-induced translocation in vivo. Based on our experiments with FRB PLF , FKBP12-Inp54p, and PH constructs in cell lines, we noticed that the ratio between the three proteins varied highly between individual cells. Indeed, others similarly noted that the ratio of each component was critical for experimental success [29,40]. Therefore, the ratio of FRB PLF -CFP to Venus-FKBP12-Inp54p in our mice might be suboptimal for translocation.
Given the widespread use of rapamycin-induced dimerization to study biological processes in cell lines, it is perhaps remarkable to note that there is only one publication describing the use of rapamycin-induced heterodimerization in vivo [25]. Notably, in this study Stankunas et al. use rapamycin analogs to dimerize and stabilize a cytoplasmically localized FRB PLF fusion protein with endogenous FKBP12 (also cytoplasmic) in cultured mouse embryonic fibroblasts (MEFs) and embryonic forelimb tissue. They did not attempt to dimerize FRB PLF with an engineered FKBP-domain to translocate protein, nor did their study utilize this system in neurons. Karpova and colleagues used a FKBP homodimerization system, consisting of two mutated FKBP domains (F36V) to regulate neurotransmission in transgenic mice [41]. Thus, it appears that no lab has successfully found a way to inducibly heterodimerize engineered FRB with engineered FKBP in vivo.

Conclusions
While rapamycin-induced translocation is highly effective for studying signaling events in a temporally-controlled manner in cell lines, our results-taken together with the lack of published reports of rapamycin-induced translocation in vivo-suggest that there are limitations that prevent the adaptation of this system for use in neurons in vitro and in vivo. Supporting this hypothesis, we found that brain lysates and DRG lysates had equally high levels of FKBP12, and it has previously been noted that high levels of FKBP12 mRNA are found throughout nervous tissue, including cerebral cortex and hippocampus, compared to non-neuronal tissue [42]. Therefore, elevated levels of endogenous FKBP12 could restrict the utilization of rapamycin-induced translocation in neuronal cells in general. To our knowledge, this issue has not been previously identified or raised. Our study could thus spur the development of new reagents, like novel rapalogs that interact with engineered versions of FKBP12 but not endogenous FKBP12. Such reagents, when combined with FRB mutants that do not interact with endogenous mTOR, could allow greater adoption of this dimerization system in vitro and in vivo.

Methods
All procedures and behavioral experiments involving vertebrate animals were approved by the Institutional Animal Care and Use Committee at the University of North Carolina at Chapel Hill.

DNA plasmid constructs
Constructs for rapamycin-induced PIP 2 depletion in HEK293 cells were obtained from Ken Mackie (University of Indiana), Tamas Balla (NICHD, Bethesda, MD USA) and Tobias Meyer (Stanford, Stanford, CA USA). The RFP-tagged PH domain of rat PLC∂1 was a kind gift from Ken Mackie. The CFP-tagged FRB domain was tethered to the plasma membrane using the first 20 amino acids of the human GAP43, as described in Várnai et al. (2006), was obtained from Tamas Balla, and cloned into pcDNA3.1(+). The FKBP-Inp54p yeast 5phosphatase construct was a gift from Tobias Meyer, cloned into pcDNA3.1(+), and modified with a Venus fluorescent protein tag.
Cell culture and live-imaging HEK293 cells were grown on glass bottom cell culture dishes (MatTek; P35G-0-10-C) in Dulbecco's Modified Eagle Medium (DMEM, Sigma) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells were transfected with 4 μl Lipofectamine 2000 (Invitrogen) and 1 μg total DNA per culture dish in Opti-MEM (Gibco) for 2 hours, at which point media was replaced with the supplemented DMEM. After 16-24 hours, supplemented DMEM was replaced with Hank's Balanced Salt Solution (HBSS Gibco 14025, supplemented with 9 mM HEPES, 11 mM D-glucose, 0.1% fatty-acid free BSA, pH 7.3) warmed to 37°C. After baseline imaging, HBSS was replaced with HBSS containing 1 μM rapamycin (Calbiochem). Each plate was imaged on a Leica TCS confocal microscope using a 40x objective, and maintained at 37°C throughout the imaging session using a heated stage attachment. Cells were treated for 10 minutes, at which point a final post-rapamycin image was taken. Membrane to cytoplasm ratio in pre-and post-rapamycin-treated cells was measured in cells expressing all three constructs, using NIH ImageJ software.

Generation of FRB PLF -CFP and Venus-FKBP12-Inp54p knockin mice
The GAP43-FRB PLF -CFP construct containing three point mutations of the FRB domain (K2095P, T2098L, and W2101F) was cloned into the Rosa26 targeting construct. This insert was placed under the control of the CAG promoter, and the entire CAG-GAP43-FRB PLF -CFP insert was followed by a self-excising neomycin resistance cassette (ACN) [43]. CGRP targeting was accomplished by recombineering of Calca targeting arms from a C57BL/6-derived bacterial artificial chromosome (BAC; RP24-136021). The start codon, located in exon 2, is common to CGRPα and calcitonin and was replaced with an AscI site to facilitate cloning. The Venus-FKBP12-Inp54p construct described above was cloned into this CGRP targeting construct, without an external promoter, but with the ACN cassette. Successful targeting of embryonic stem cells by homologous recombination was identified with Southern blot hybridization, using probes that flanked the 5' and 3' arms of the targeting constructs, as well as an internal neomycin probe. Chimeric mice were produced by blastocyst injection, and mated to C57BL/6 mice to establish the line.

Neuronal dissociation and imaging
Male WT and Rosa-FRB PLF /CGRP-Inp54p mice (3-4 weeks old) were decapitated without anesthesia, and the DRG were dissected into ice-cold Hank's Balanced Salt Solution (HBSS; Gibco,14175-095), and dissociated using collagenase (1 mg/mL; Worthington, CLS1) and dispase (5 mg/mL; Gibco, 17105-041) dissolved in HBSS. Neurons were plated onto coverslips coated with 0.1 mg/mL poly-D-lysine (Sigma P0899) and 5 μg/mL laminin (Sigma, L2020), and cultured in Neurobasal-A medium (Invitrogen, 10888022), supplemented with B-27 Supplement (Gibco, 17504-044), L-glutamine (Gibco, 25030-081), and penicillin-streptomycin (Gibco, 15140-122), and 5% fetal bovine serum. WT neurons were grown with no fetal bovine serum, but with the addition of 0.25 ng/mL nerve growth factor, and 0.5 ng/mL glial derived neurotrophic factor. WT neurons were fixed with 4% paraformaldeyde (warmed to 37°C) at 24 hours in vitro and then immunostained. For Rosa-FRB PLF +/− /CGRP-Inp54p +/− neurons, rapamycin was added to neuronal culture medium at a final concentration of 1 μM in half of the wells, and cultured for 24 and 48 hours. This allowed for comparison of treated neurons to vehicle controls. Dissociated neuronal cultures were prepared as described above, and fixed at 24 hours in vitro with 4% PFA warmed to 37°C for 30 min. Neurons were washed with PBS to remove fixative, mounted on slides, and imaged on a Leica TCS confocal microscope. CFP was excited with a 458 nm laser and detected with emission settings of 465-505 nm, and Venus was excited using a 514 nm laser and detected with emission settings of 528-587 nm.

Immunohistochemistry
For DRG tissue sections, male mice 4-6 weeks were injected intraperitoneally with pentobarbital, and perfused with 4% PFA in 0.1 M phosphate buffer, pH 7.4. Lumbar DRG (L2-L6) were dissected and post-fixed for 2 hours in 4% PFA. The DRG were subsequently cryoprotected in 30% sucrose, 0.1 M phosphate buffer, pH 7.3 at 4°C for 24 h, frozen in OCT TissueTek, cryosectioned at 20 μm, mounted on Superfrost Plus slides, and stored at −20°C until use. Tissue was rehydrated and washed in PBS to remove OCT embedding compound, and either coverslipped immediately, or prepared for immunohistochemistry (IHC). DRG sections and dissociated neurons were permeablized and blocked in TBS-Tx (0.05 M Tris, 2.7% NaCl, 0.3% Triton-X 100, pH 7.6) containing 10% normal goat serum (NGS) for 1 hr at room temperature. Sections were incubated overnight at 4°C with primary antibodies in TBS-Tx/10%NGS, washed, incubated at room temperature for 2 hours with secondary antibodies in TBS-Tx/10%NGS, washed, and mounted with Fluorogel

Drug administration for tissue extraction
Mice that were heterozygous for CGRP-Inp54p, or heterozygous for both Rosa-FRB PLF and CGRP-Inp54p, received three intrathecal injections (each containing 1 nmol rapamycin, dissolved in saline to a total volume of 5 uL) over the course of 24 hours. Mice were then perfused and dissected, as described above, two hours after the second round of injections.

Behavior
Male 3-to 4-month-old CGRP-Inp54p +/− and WT littermates (n = 12 per genotype), or Rosa-FRB PLF /CGRP-Inp54p and CGRP-Inp54p heterozygous littermates (n = 10 per genotype), were acclimated to the testing apparatuses and experimenter for 2 days prior to behavioral testing. The experimenter was blind to genotype throughout the experiment. For CGRP-Inp54p +/− and WT controls, baseline thermal and mechanical responses were monitored using Hargreaves and Von Frey apparatuses (as previously described in [22]) prior to intraplantar injection with complete Freund's adjuvant (CFA) into the left hindpaw. Behavior testing for both thermal and mechanical sensitization was carried out on subsequent days, as described previously [22]. Rosa-FRB PLF /CGRP-Inp54p double heterozygous and CGRP-Inp54p +/− littermate controls were used to determine the extent of rapamycin-induced depletion of PIP 2 in the reduction of pain sensitivity in vivo. For this, each mouse received two intrathecal injections of rapamycin (1 nmol in 5 μl), one just after baseline measurement, and one just prior to CFA injection into the hindpaw 6 hours later, and behavioral testing was carried out on subsequent days.