Abstract
Nerve growth factor (NGF) exerts multiple functions on target neurons throughout development. The recent discovery of a point mutation leading to a change from arginine to tryptophan at residue 100 in the mature NGFβ sequence (NGFR100W) in patients with hereditary sensory and autonomic neuropathy type V (HSAN V) made it possible to distinguish the signaling mechanisms that lead to two functionally different outcomes of NGF: trophic versus nociceptive. We performed extensive biochemical, cellular, and live-imaging experiments to examine the binding and signaling properties of NGFR100W. Our results show that, similar to the wild-type NGF (wtNGF), the naturally occurring NGFR100W mutant was capable of binding to and activating the TrkA receptor and its downstream signaling pathways to support neuronal survival and differentiation. However, NGFR100W failed to bind and stimulate the 75 kDa neurotrophic factor receptor (p75NTR)-mediated signaling cascades (i.e., the RhoA-Cofilin pathway). Intraplantar injection of NGFR100W into adult rats induced neither TrkA-mediated thermal nor mechanical acute hyperalgesia, but retained the ability to induce chronic hyperalgesia based on agonism for TrkA signaling. Together, our studies provide evidence that NGFR100W retains trophic support capability through TrkA and one aspect of its nociceptive signaling, but fails to engage p75NTR signaling pathways. Our findings suggest that wtNGF acts via TrkA to regulate the delayed priming of nociceptive responses. The integration of both TrkA and p75NTR signaling thus appears to regulate neuroplastic effects of NGF in peripheral nociception.
SIGNIFICANCE STATEMENT In the present study, we characterized the naturally occurring nerve growth factor NGFR100W mutant that is associated with hereditary sensory and autonomic neuropathy type V. We have demonstrated for the first time that NGFR100W retains trophic support capability through TrkA, but fails to engage p75NTR signaling pathways. Furthermore, after intraplantar injection into adult rats, NGFR100W induced neither thermal nor mechanical acute hyperalgesia, but retained the ability to induce chronic hyperalgesia. We have also provided evidence that the integration of both TrkA- and p75NTR-mediated signaling appears to regulate neuroplastic effects of NGF in peripheral nociception. Our study with NGFR100W suggests that it is possible to uncouple trophic effect from nociceptive function, both induced by wild-type NGF.
Introduction
Nerve growth factor (NGF), discovered as a result of potent trophic actions on sensory and sympathetic neurons of the PNS in the 1950s (Levi-Montalcini and Hamburger, 1951), also regulates the trophic status of striatal and basal forebrain cholinergic neurons (BFCNs) of the CNS (Levi-Montalcini and Hamburger, 1951; Svendsen et al., 1994; Li and Jope, 1995; Kew et al., 1996; Conover and Yancopoulos, 1997; Lehmann et al., 1999). With the discovery of brain-derived neurotrophic factor (BDNF), Neurotrophin 3 (NT-3) and Neurotrophin 4 (NT-4), NGF is now known as a member of the neurotrophin family (Chao and Hempstead, 1995; Huang and Reichardt, 2001; Chao, 2003). NGF acts via two known receptors, the 140 kDa tyrosine receptor kinase A (TrkA) and the 75 kDa neurotrophin receptor (p75NTR), to transmit signals to the cytoplasm and nucleus of responsive neurons (Bothwell, 1995; Chao and Hempstead, 1995; Kaplan and Miller, 1997). NGF signaling through TrkA elicits many of the classical neurotrophic actions ascribed to NGF (Loeb and Greene, 1993). TrkB and TrkC mediate the signaling of NT-3 and NT-4, respectively (Huang and Reichardt, 2001; Chao, 2003). NGF and all members of the family also signal through p75NTR (Huang and Reichardt, 2001; Chao, 2003). p75NTR contributes to sphingomyelin-ceramide metabolism (Dobrowsky et al., 1994, 1995) and modulates RhoA activity to regulate axonal growth (Yamashita et al., 1999; Gehler et al., 2004). In addition, p75NTR has been shown to activate the NF-κB, Akt, and JNK pathways (Harrington et al., 2002; Roux and Barker, 2002) to either induce apoptosis or to promote cell survival and differentiation (Chao and Hempstead, 1995; Casaccia-Bonnefil et al., 1998; Salehi et al., 2000; Roux and Barker, 2002; Nykjaer et al., 2005).
Given its robust trophic effects, NGF has been investigated for therapeutic properties in neurodegenerative disorders (Apfel et al., 1994, 1998; Andreev et al., 1995; Apfel and Kessler, 1995, 1996; Blesch and Tuszynski, 1995; Anand et al., 1996; Apfel, 1999a,b, 2000, 2002; Aloe et al., 2012). In one example, NGF's robust trophic effects on BFCNs has suggested a role in treating Alzheimer's disease (AD) in which this population degenerates (Olson, 1993; Hefti, 1994; Scott and Crutcher, 1994; Blesch and Tuszynski, 1995; Knusel and Gao, 1996; Koliatsos, 1996; Eriksdotter Jönhagen et al., 1998; Williams et al., 2006; Mufson et al., 2008; Schindowski et al., 2008; Schulte-Herbrüggen et al., 2008; Cuello et al., 2010). Unfortunately, features of the biology of NGF have limited the extent to which it could be evaluated. Inability to cross the blood–brain barrier prevented systemic administration. Delivery via the ventricular system even at low doses resulted in pain and studies in primates demonstrated Schwann cell hyperplasia that served to compromise CSF flow (Winkler et al., 1997). A recent phase 2 trial in which NGF was delivered via virus to the basal forebrain demonstrated safety and was not associated with pain (Tuszynski et al., 2015). Marked sprouting of BFCN fibers was evidence of a potent trophic effect, but cognitive measures were unaffected (Rafii et al., 2014). Significant efforts were also invested to investigate NGF as a treatment for diabetic polyneuropathy (Apfel and Kessler, 1995, 1996; Anand et al., 1996; Tomlinson et al., 1996; Elias et al., 1998; Apfel, 1999b; Goss et al., 2002; Murakawa et al., 2002; Kanda, 2009). NGF treatment demonstrated some benefit in phase 2 trial at 0.1 and 0.3 μg/kg, but was associated with dose-dependent hyperalgesia at the injection site (Apfel, 1999a,b). A large-scale phase 3 trial with a dose of 0.1 μg/kg showed no beneficial effect (Apfel et al., 1998; Apfel, 2002).
NGF is not only a trophic factor, but also functions as one of the key molecules for mediating inflammatory pain and neuropathic pain in the PNS (Lewin and Mendell, 1993; Lewin et al., 1993; Chuang et al., 2001; Watanabe et al., 2008). Therefore, clinical trials in which large doses of NGF were infused in patients with AD had to be terminated due to the extreme side effects of pain (Aloe et al., 2012). Other clinical trials using NGF in treating diabetic neuropathies and peripheral neuropathies in HIV were also discontinued after reports of serious side effects such as back pain, injection site hyperalgesia, myalgia, and weight loss (Hellweg and Hartung, 1990; Lein, 1995; Apfel et al., 1998; Unger et al., 1998; Rask, 1999; McArthur et al., 2000; Quasthoff and Hartung, 2001; Schifitto et al., 2001; Apfel, 2002; Pradat, 2003; Walwyn et al., 2006). Therefore, the adverse effect of significant pain caused by NGF has severely limited its therapeutic use in treating neurodegenerative disorders. To overcome these pain-causing side effects of NGF, it is of paramount importance to elucidate the role of NGF and its receptor signaling by TrkA and p75NTR in nociception.
A large body of genetic and clinical evidence has pointed to both TrkA and p75NTR as contributing factors to sensitization of inflammatory pain mediated by NGF. For example, recessive mutations in TrkA cause hereditary sensory and autonomic neuropathy type IV [hereditary sensory autonomic neuropathy type V (HSAN V) (Online Mendelian Inheritance in Man (OMIM) #256800], also known as congenital insensitivity to pain with anhidrosis (Indo, 2001, 2002). Strong evidence supports a role of TrkA in mediating the sensitization effect of NGF: attenuation of TrkA expression (Malik-Hall et al., 2005; Alvarez and Levine, 2014) and pharmacological inhibition of TrkA-mediated signaling pathways extracellular signal-related kinase (ERK), phosphatidylinositol 3-kinase (PI3K), and phospholipase Cγ (PLCγ) all reduced NGF-induced hyperalgesia (Fang et al., 2005; Malik-Hall et al., 2005; Summer et al., 2006; Mantyh et al., 2011; Alvarez and Levine, 2014; Ashraf et al., 2016). Furthermore, NGF still evoked hyperalgesia in mice lacking p75NTR, pointing to the involvement of TrkA (Bergmann et al., 1998). Evidence that p75NTR has a role in pain signaling pathways is largely indirect. In one example, injecting a neutralizing antibody to p75NTR prevented NGF-induced pain behavior and NGF-mediated increases in action potentials in sensory neurons (Zhang and Nicol, 2004; Watanabe et al., 2008; Iwakura et al., 2010). Therefore, both TrkA and p75NTR signals contribute to pain induced by NGF. However, how these two receptors interact to mediate pain is poorly defined.
Recently, patients in consanguineous Swedish families suffering from length-dependent loss of pain that often leads to bone fractures and joint destruction were shown to harbor a homozygous missense mutation in NGF (Einarsdottir et al., 2004; Carvalho et al., 2011). The disorder was labeled HSAN V, OMIM #608654. Genetic analysis of these HSAN V patients revealed a point mutation (661C>T) causing a substitution of tryptophan (W) for arginine (R) at position 211 in the proform of the NGF polypeptide (pro-NGFR221W); this residue corresponds to the position 100 in the mature protein (NGFR100W) (Einarsdottir et al., 2004). Unlike HSAN type IV, which results from mutations in TrkA, HSAN V patients appear to have normal cognitive function, suggesting that the mutant NGF may retain its trophic functions in the CNS (Einarsdottir et al., 2004). We and others reasoned that NGFR100W provides a tool with which to decipher possible differences in the trophic and nociceptive actions of NGF (Covaceuszach et al., 2010; Capsoni et al., 2011, 2014).
Initial characterization of NGFR100W revealed that the R100 mutation may disrupt the processing of pro-NGF to mature NGF in cultured cells, resulting in relatively higher percentage of NGF secreted as the pro-form (Larsson et al., 2009). Given the difficulties in expressing NGFR100W, Capsoni et al. (2011) examined a different series of residues at position 100, including NGFR100E and NGF double mutant (NGFP61S/R100E), using recombinant techniques. They discovered that these NGFR100 mutants bound normally to TrkA, but failed to bind to p75NTR (Covaceuszach et al., 2010). This finding suggested that failure to activate p75NTR signaling was sufficient to attenuate pain induced by NGF and that TrkA signaling had little or no effect in pain (Capsoni et al., 2011). The investigators speculated that NGFR100 mutant would allow for development of p75NTR antagonists such as NGFR100W as a “painless NGF” therapeutic agent (Malerba et al., 2015).
In the present study, we examined the binding and signaling properties of the mature form of the naturally occurring mutant NGF (NGFR100W) in HSAN V and compared the effects with those of wtNGF and NGF mutants that selectively engage and signal through either TrkA or p75NTR. We discovered that NGFR100W retains binding and signaling through TrkA to induce trophic effects, but not binding or activation of p75NTR. Our findings are evidence that NGFWT acts via both p75NTR and TrkA to cause pain. Our findings are consistent with a necessary role for TrkA in both acute sensitization and delayed priming of nociceptive responses. In contrast, signaling through p75 appears to be necessary for acute sensitization, but does contribute to priming. The integration of both TrkA and p75 signaling thus appears to regulate neuroplastic effects of NGF in peripheral nociception.
Materials and Methods
Ethics statement.
All experiments involving the use of animals were approved by the Institutional Animal Care and Use Committee of University of California–San Diego and University of California–San Francisco (UCSF). Surgical and animal procedures were performed strictly following the National Institutes of Health's Guide for the Care and Use of Laboratory Animals.
Chemicals, oligos, and reagents.
Streptavidin-QD605, Streptavidin-QD-655 (Q10103MP, Q10123MP) conjugates were from Invitrogen; all other chemicals were from Sigma-Aldrich unless noted otherwise. Recombinant extracellular domains of p75NTR were a generous gift from Dr. Sung Ok Yoon of Ohio State University. Prostaglandin E2 (PGE2) was from Sigma-Aldrich (catalog #82475). NGF (wtNGF, NGFR100W, KKE, Δ9/13) proteins were produced in our own laboratory (Sung et al., 2011).
Cloning.
Mouse pro-NGF was amplified by PCR from a NGF-GFP plasmid (a generous gift from Professor Lessmann, Mainz, Germany). The forward primer sequence was: 5′-acgaattccaccatgtccatgttgttctacactctgatcactgcg-3′ and the reverse primer sequence was: 5′-gatggatccttcgtgccattcgattttctgagcctcgaagatgtcgttcagaccgccaccgacctccacggcggtggc-3′. The reverse primer contains a sequence coding for the 17 aa AviTag: GGGLNDIFEAQKIEWHE. The sequence was based on the #85 AviTag peptide sequence described previously (Schatz, 1993). One glutamic acid residue was added to the C-terminal AviTag based on a finding by Avidity (www.avidity.com) that it greatly enhanced the biotinylation rate of the AviTag (Beckett et al., 1999). Platinum pfx DNA polymerase (Invitrogen, catalog #11708021) was used following the manufacture's instructions. The 50 μl reaction was denatured at 94°C for 4 min, followed by 25 cycles of amplification (30 s at 94°C; 30 s at 50°C; 90 s at 68°C). An additional extension was performed at 68°C for 4 min. The PCR product was purified and digested with EcoRI (Fermentas, catalog #FD0274) and BamHI (Fermentas, catalog #FD0054) and was ligated in-frame into the pcDNA3.1-myc-His vector that was predigested with EcoRI/BamHI. The resulting construct was designated as pcDNA3.1-NGFavi. BirA was amplified by PCR from pET21a-BirA (Addgene, plasmid #20857) (Howarth et al., 2005) using a forward primer (forward primer: 5′-gtgaac atg gctagcatgact-3′) and a reverse primer (5′-ggtgctcgagtcatgcggccgcaagct-3′ (containing an XhoI site). PCR was performed using Pfx as described above. The PCR product was digested with XhoI (Fermentas, catalog #FD0694) and subcloned into pcDNA3.1 myc.his (+) vector (Invitrogen) that was precut with EcoRV (Fermentas, catalog #FD0303) and XhoI. The resulting plasmid was designated as pcDNA3.1-BirA. All primers were from Elim Biopharmaceuticals. All constructs were verified by sequencing (Elim Biopharmaceuticals). NGFR100W was cloned using the same method, but with a point mutation at 661 (C>T). The KKE and Δ9/13 mutant constructs were obtained from Dr. K. Neet of Rosalind Franklin University (Hughes et al., 2001; Mahapatra et al., 2009) and subcloned to the pcDNA3.1-myc-His vector with an Avitag (Sung et al., 2011).
Protein purification.
HEK293FT cells were grown in 15 cm plates to 70% confluency. Cells were changed to 25 ml of DMEM-high glucose, serum-free medium that was supplemented with 50 μm d-biotin (Sigma-Aldrich, catalog #B4639). Then, 15–21 μg f pcDNA3.1-NGFavi, NGFR100Wavi, KKEavi, and Δ9/13avi plasmids DNA plus 15–21 μg of pcDNA3.1-BirA plasmid DNA were mixed with 1 ml of DMEM-high glucose medium and 60 μl of Turbofect (Fermentas, catalog #R0531). The mixture was incubated at room temperature for 15 min and then added into the medium by the dropwise method. Transfected HEK293FT cells were incubated at 37°C, 5% CO2. Seventy-two hours after transfection, media were collected for protein purification.
Media were harvested and adjusted to 30 mm phosphate buffer, pH 8.0, 500 mm NaCl, 20 mm imidazole, and a mixture of protease inhibitors (1 mm PMSF from Sigma-Aldrich, catalog #P7626, and 1 μl/ml aprotinin from Sigma-Aldrich, catalog #A6279). After incubation on ice for 15 min, media were cleared by centrifugation at 18,000 rpm for 30 min using a Beckman JA-20 rotor. Ni-NTA resins (Qiagen, catalog #30250) were rinsed with the washing buffer (30 mm phosphate buffer, pH 8.0, 500 mm NaCl, 20 mm imidazole, and a mixture of protease inhibitors from Sigma-Aldrich, catalog #S8820). Ni-NTA resins were added to the media at a concentration of 0.3 ml Ni-NTA/100 ml of media and incubated overnight with rotation at 4°C. The media/Ni-NTA slurry was loaded onto a column and the captured Ni-NTA resins were washed with 10 ml of wash buffer and eluted with 5 ml of elution buffer (30 mm phosphate buffer, pH 8.0, 500 mm NaCl, 300 mm imidazole, protease inhibitors). Every 500 μl volume of elution was collected. The purity and concentration of NGF was assessed by SDS-PAGE using a silver staining kit (Fast Silver, G-Biosciences, catalog #786–30). Known quantities of NGF purified from mouse submaxillary glands were used as standards. The first two eluted fraction normally contained most of purified proteins.
Cell culture and transfection.
PC12 cells or a subclone of PC12 cells, PC12M, PC12nnr5 cells were cultured as described previously (Wu et al., 2001, 2007). NIH3T3-TrkA, NIH3T3-p75NTR cells were as described previously (Huang et al., 1999). HEK293FT cells (Invitrogen, catalog #R70007) cells were cultured in DMEM-high glucose medium (4.5 g/L glucose, Mediatech, catalog #10-013-CV), 10% FBS, and 1% penicillin/streptomycin.
Administration of wtNGF, NGFR100W, prostaglandin E2 (PGE2), and inhibitors to adult rats.
Experiments were performed on adult male Sprague Dawley rats (220–240 g, Charles River Laboratories; RRID:RGD_737891). All experiments were performed following protocols that have been approved by the University of California–San Francisco Committee on Animal Research and conformed to the National Institutes of Health's Guidelines for the Care and Use of Laboratory Animals. Either 200 ng of wtNGF or NGFR100W was injected intradermally on the dorsum of one hindpaw of adult rats. Nociceptive thresholds in the injected paws were then tested over time.
For studies using K252a (Sigma-Aldrich catalog #K1639) and GW4869 (Sigma-Aldrich catalog #D1692), both inhibitors were dissolved in DMSO (2 μg/μl) and then diluted to a concentration of 0.2 μg/μl in saline at the time of the experiments. Five minutes before injection of NGF, 5 μl (1 μg) of inhibitors was administered intradermally on the dorsum of the hindpaw at the same site where NGF was injected. The experimental design is illustrated in Figure 7A.
Randall–Selitto mechanical test and Hargreaves thermal test.
Mechanical nociceptive threshold was measured using the Randall–Selitto paw pressure test (Randall and Selitto, 1957) with an Ugo Basile Algesymeter (Ferrari et al., 2010, 2013, 2015a,b). This device exerts a linear increase in force to the dorsum of the hindpaw of the rats. Before the test, rats were kept in individual restrainers for 20 min to acclimatize them to the experimental environment. The restrainers had openings that allowed rats to extend hindpaws during the test. Mechanical thresholds were calculated as the average of three readings.
Thermal threshold was measured using the Hargreaves test, which applies heat stimuli by an adjustable high-intensity movable halogen projector lamp (Malmberg and Yaksh, 1993). Baseline responses were first measured and then averaged from at least two readings. Then, 600 ng of wtNGF or NGFR100W was injected in the plantar surface of one hindpaw under isofluorane anesthesia. After recovery from the anesthesia, in less than 1 min, rats were placed on a glass plate and acclimatized for 20 min. Thermal nociceptive thresholds were evaluated within 1 h after the injections.
Binding and internalization assays of NGF-QD.
At 50% confluence of NIH3T3-TrkA and NIH3T3-p75NTR, cells were starved in serum free DMEM-high glucose medium for 4 h at 37°C. NGF or NGF mutants were conjugated with QD 655-streptavidin (Invitrogen, catalog #Q10121MP) on ice 4°C for 30 min. Then, 0.2 nm conjugate was applied to cells. Cells were incubated at 20°C for 20–30 min, washed with serum free DMEM-high glucose medium, and then surface binding was quantified. For the internalization assay, cells were incubated with 1 nm conjugates for 30 min or 2 h, for 3T3-TrkA and 3T3-p75NTR, respectively, at 37°C. Cells were washed and then subjected to imaging.
Pull-down assay.
The recombinant protein, Fc-75NTR extracellular domain (ECD) was a kind gift from Dr. Sung Ok Yoon, the Ohio State University. Five microliters of supernant from insect lysate expressing Fc-p75NTR was incubated with a range of either wtNGF or NGFR100W (∼0–10 ng) overnight at 4°C. Avidin–agarose beads (20 μl) were added/incubated for 2 h at 4°C. Beads were washed, boiled with protein sample buffer, and subjected to SDS-PAGE. Western blotting was performed using the function blocking antibodies against the extracellular domain of p75NTR (REX) (Mischel et al., 2001).
Dorsal root ranglion (DRG) culture, live-cell imaging, and data analysis.
Embryonic DRGs at embryonic day 15 (E15) to E16 were isolated from Sprague Dawley rats as described previously (Cui et al., 2007; Wu et al., 2007; Sung et al., 2011) with some minor modifications. Cells were maintained with alternation between growth media (MEM media containing 10% heat inactivated FBS and 100 ng/ml of NGF) and selection media (MEM media containing 0.5–1 μm cytosine β-d-arabinofuranoside (Sigma-Aldrich catalog #C1768 and 100 ng/ml NGF) every 2 d.
For survival analysis, only the cells with round and transparent cell bodies were counted as DRG neurons to exclude possible fibroblast populations. In phase contrast images, the cell bodies of DRGs look whitish and transparent, whereas fibroblasts have black and flattened cell bodies.
For live cell imaging, dissociated DRGs were cultured in microfluidic chambers for 7–10 d (Cui et al., 2007). The microfluidic chambers, manufactured in-house, were plated onto 24 mm × 48 mm glass coverslips that were precoated with poly-L-lysine (Sigma-Aldrich catalog P8920) as described previously (Taylor et al., 2006). Dissociated DRG neurons were plated into the cell body chamber. The growth/selection scheme outlined above was repeated. Axons from the DRG neurons started to cross the microgrooves after 3 d and reached the axonal chamber in another 7–8 d. Before live imaging of axonal transport of NGF, all compartments (cell body and axonal chambers) of the DRG neurons were thoroughly rinsed and depleted of NGF in NGF-free, serum-free MEM for 3 h. NGF-QD605 was prepared following the protocol described above. NGF-QD605 was added to a final concentration of 0.2 nm to the axonal chambers for 2 h at 37°C. Live cell imaging of NGF-QD605 transport with the axons was performed using a modified inverted microscope (Nikon TE300) for pseudo-TIRF illumination (Zhang et al., 2010). The microscope stage was equipped to maintain a constant temperature (37°C). CO2 level (5%) was maintained using CO2-independent medium (Invitrogen catalog #18045-088) during live imaging. The laser beam of 532 nm was used and penetrated ∼1 μm into aqueous solution at an incident angle. Fluorescence emission was filtered with QD605/20 emission filter (Chroma Technology). Time-lapse images were acquired at the speed of 10 frames/s and were captured using an EMCCD camera (Cascade 512B, Photometric). All data were processed and analyzed using a MATLAB software pipeline (RRID:SCR_001622).
Single-cell patch-clamp recording.
All recordings were obtained from small- to medium-diameter cells from cultured DRGs at room temperature. One-electrode whole-cell voltage-clamp recording was performed using Axopatch-1D amplifiers (Molecular Devices). A 2–5 MW sized patch electrode and puffing pipette were used. All cells were clamped at −60 mV holding potential to measure low pH-evoked current.
Standard external solution contained the following (in mm): 145 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 10 HEPES, and 10 glucose, adjusted to pH 7.4 with NaOH, and was used to incubate DRG neurons. To block possible synaptic transmission via activation of ionotropic receptors, 1 mm TTX, 10 mm CNQX, 50 mm AP5, and 50 mm picrotoxin were added to the bath. Internal recording solution contained the following (in mm): 130 K-gluconate, 10 HEPES, 0.6 EGTA, 5 KCl, and 2.5 Mg-ATP, adjusted to pH 7.3 with KOH. Low pH puffing solution was made from external solution and was adjusted to pH 5.5 by adding 1 n HCl.
Cells were starved for 2 h with MEM before recording. To evoke a low pH response, pH 5.5 external solution was applied briefly through an additional glass pipette placed 50 mm away from the recorded cell and by a Picospritzer (5–10 psi, for 50–100 ms, Picospritzer II, General Valve). The responses seen at this configuration were not evoked by mechanical stimulation due to air puffing as when neutral pH solution was used, so no response was observed. wtNGF or NGFR100W at a concentration of 50 ng/ml was applied directly to the bath and incubated for 10 min, followed by low pH puffing. Data analysis was done in pClamp software (Molecular Devices, RRID:SCR_011323) by calculating the charge transfer for 100 ms and normalized by computing the ratio of response after NGF/response before NGF.
Antibodies and SDS-PAGE/blotting.
Standard protocols were followed for SDS-PAGE and blotting. Rabbit anti-pro-NGF IgGs were a generous gift of Dr. B.L. Hempstead of Cornell University. Rabbit anti-NGF IgGs were from Santa Cruz Biotechnology (catalog #sc-549; RRID:AB_632012). Rabbit IgGs against Trk-pTyr490, pErk1/2, total Erk1/2, and p-Cofilin were from Cell Signaling Technology (catalog #9141; RRID:AB_2298805, 9101; RRID:AB_331646, 9102; RRID:AB_330744, 3311; and RRID:AB_330239 respectively). Mouse IgGs against pJNK and mouse IgGs against total PLC-γ were from Santa Cruz Biotechnology (sc-6254; RRID:AB_628232, sc-7290; RRID:AB_628119, respectively) and rabbit IgGs against pPLC-γ were from GeneTex (GTX61714; RRID:AB_10621534). Rabbit functional blocking antibodies against the extracellular domain of p75NTR (REX) (Mischel et al., 2001) was a generous gift of Dr. L. Reichardt of UCSF. Rabbit monoclonal antibodies against pAkt were from Epitomics (catalog #2214-1; RRID:AB_1266979). Rabbit anti-AviTag IgGs were from Genescript (catalog #A00674; RRID:AB_915553).
Immunostaining.
Immunostaining was performed according to published protocols (Weissmiller et al., 2015). Briefly, PC12nnr5 cells (RRID:CVCL_C128) were cultured on coverslips that were precoated with Matrix gel (BD Biosciences). Aafter serum starvation for 2 h, cells were treated with either 50 ng/ml wtNGF or 50 ng/ml NGFR100W. Cells were then fixed for 10 min with 4% paraformaldehyde at 37°C and permeabilized for 15°C at room temperature with 0.1% Triton X-100; 3% BSA and 5% goat serum in PBS were used for blocking. The cells were then incubated overnight at 4°C with blocking solution containing 1/100 diluted primary antibody, active RhoA from New East Biosciences (catalog #26904; RRID:AB_1961799). After washing 3× primary antibody with PBS and rocking fir 5 min, the cells were incubated with a 1/800 dilution of Alexa Fluor 488-goat anti-mouse IgG (Invitrogen catalog #A1100) for 1 h at room temperature with rocking covered with foil. After 3 washes with PBS, nuclei were labeled with 1 μg/ml Hoechst 33342 (bisBenzimide H 33342 trihydrochloride, Sigma-Aldrich catalog #B2261) for 5 min at room temperature. Cells were rinsed, air-dried, mounted, and examined with a Leica microscopy using a 100× oil objective lens.
Statistical analysis.
All experiments were repeated at least three times independently. Statistical analyses of results and calculation of p-values were performed using Prism 5 software (GraphPad Software; RRID:SCR_015807). For un-pairwise comparisons, the Student's t test was used. For multiple comparisons, the Tukey one-way ANOVA (RRID:SCR_002427) test was used. Data are presented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001.
Results
NGFR100W does not elicit acute thermal or mechanical hyperalgesia in vivo
Previous studies suggested that the NGF mutation associated with HSAN V disrupted either the processing of the proform (i.e., NGFR221W) or secretion of the mature form (NGFR100W), resulting in preferential secretion of the NGFR221W (Larsson et al., 2009; Covaceuszach et al., 2010; Carvalho et al., 2011; Capsoni, 2014). However, these findings and interpretations are perplexing because HSAN V patients report no autonomic symptoms (Capsoni, 2014), which would argues against increased levels of pro-NGF, a p75 ligand that induces death of sympathetic neurons (Roux and Barker, 2002; Nykjaer et al., 2005; Khodorova et al., 2013). Furthermore, these patients have intact mental abilities, arguing against clinically meaningful CNS neuronal loss and dysfunction. To further explore the binding, signaling, and actions of NGFR100W, we produced and characterized the mature form of the protein in HEK293 cells.
Although superficial sensation in patients homozygous for NGFR100W is normal, it is unknown whether this mutant NGF can induce nociceptor sensitization or transition to chronic pain. Therefore, we measured the behavioral response of adult rats to mechanical stimuli or noxious thermal stimulus after a single injection of either wtNGF or NGFR100W. Proteins were administered intradermally to the hindpaws of adult male Sprague Dawley rats following published protocols (Taiwo et al., 1991).
We used the Randall–Sellito method to measure mechanical threshold (Randall and Selitto, 1957). For all experiments, baseline values were measured before the injection of NGF. Previous reports showed that 10 ng to 1 μg of injected wtNGF induced significant mechanical hyperalgesia (Andreev et al., 1995; Malik-Hall et al., 2005). When 200 ng of wtNGF (Fig. 1A–C, blue) was injected to the dorsum of the rat's hindpaw, the mechanical nociceptive threshold was reduced by ∼25% compared with baseline (p < 0.0001, unpaired t test, 0 m vs 60 m on the first day; Fig. 1B,C, blue). Mechanical hyperalgesia was evident as soon as 15 min and reached the maximum by 1 h, an effect that lasted for ≥24 h. Hyperalgesia was diminished to ∼15% by 5 d (Fig. 1C). In marked contrast to wtNGF, injection of 200 ng of NGFR100W had no effect on mechanical nociceptive threshold (Fig. 1B,C, red).
NGFR100W does not induce acute sensitization in adult rats. A–C, Mechanical hyperalgesia. The Randall–Selitto method was used to measure mechanical hyperalgesia in adult rats. After intradermal injection of 200 ng of wtNGF (n = 6) or NGFR100W (n = 6) into the rat's hindpaws, mechanical threshold was measured at the indicated times. A significant decrease in mechanical nociceptive threshold in rats injected with wtNGF was seen within 15 min, reached a maximum by 1 h, and lasted at least 5 d. In contrast, NGFR100W did not produce a significant decrease in the threshold during the 5 d period. D–F, Thermal hyperalgesia. The Hargreaves test was used to measure thermal hyperalgesia. Baseline response was first measured for each test animals before NGF injection. Hindpaws were injected with 600 ng of either wtNGF or NGFR100W. Thermal threshold was measured at the indicated times. Animals that were injected with wtNGF showed a significant decrease at 30 min. The decrease was dissipated after 1 h. NGFR100W failed to reduce thermal nociceptive threshold during the entire 1 h test period. Six rats for each group (n = 6) were used in the test. Unpaired t tests were performed against the baselines within each NGF injection group to produce p-values. Comparions were done between the threshold before injection with either wtNGF or NGFR100W to produce p-values that were noted in the figure. *p < 0.05, **p < 0.01, ***p < 0.001. G–I, PGE2-hyperalgesic priming effect. On d 7 after NGF administration, PGE2 was injected intradermaly (100 ng/5 μl) to test hyperalgesic priming. Naive rats (nontreated with NGF) are shown as a control. Naive animals showed acute hyperalgesia within 30 min, but not at 4 h, after PGE2 injection. Rats pretreated with intradermal injection of wtNGF showed a decrease in the mechanical nociceptive threshold after PGE2 injection at 30 min, like the controls, but unlike controls, it lasted longer than 4 h. Rats pretreated with NGFR100W showed significant decrease in mechanical nociception after PGE2 injection, which lasted at least 4 h and was comparable to wtNGF. Data are presented as mean ± SEM. Unpaired t tests were performed versus values before PGE2 injection. ***p < 0.001 Six paws (n = 6) were used for wtNGF or NGFR100W.
We then measured thermal hyperalgesia using the Hargreaves method (Hargreaves et al., 1988). Intraplantar injection of 600 ng of wtNGF to the hindpaw of rats induced thermal hyperalgesia, as demonstrated by as much as a 34.8% decrease in nociceptive threshold (p = 0.0186, 0 vs 45 m, unpaired t test; Fig. 1D–F, blue). Acute thermal hypergelsia was first observed 20–30 min after injection, with the maximal effect at 45 min, followed by a return to baseline (Fig. 1E,F, blue). In contrast, injection of 600 ng of NGFR100W did not induce acute thermal hyperalgesia (Fig. 1E,F, red). We conclude that, unlike wtNGF, NGFR100W does not induce acute thermal or mechanical hyperalgesia.
NGFR100W induces hyperalgesic priming
To test whether NGFR100W could contribute to chronic pain, we took advantage of the hyperalegesic priming paradigm with the injection of prostaglandin E2 (PGE2) (Reichling and Levine, 2009). Because wtNGF was shown to induce hyperalgesic priming in rats (Ferrari et al., 2010), we used this model to examine responses to wtNGF and NGFR100W. Seven days after injection of 200 ng of either wtNGF or NGFR100W, PGE2 (100 ng/5 μl) was injected into the same injection site into adult rats (Fig. 1G). If a priming effect for chronic pain is absent, then injection of PGE2 only induces acute hyperalgesia that disappears by 4 h (Ferrari et al., 2010, 2013, 2015b). However, in the presence of priming, the same dose of PGE2 induces a much prolonged hyperalgesia to mechanical stimuli (Aley and Levine, 1999; Reichling and Levine, 2009).
As reported previously (Ferrari et al., 2010), intradermal injection of wtNGF (200 ng) induced acute mechanical hyperalgesia that lasted through day 7 (Fig 1H,I). As before, NGFR100W failed to induce acute mechanical hyperalgesia (Fig. 1H,I). On the seventh day after measuring the threshold baseline, PGE2 was injected into the same site as for NGF and the mechanical threshold was measured at 30 min and 4 h thereafter. Consistent with previous findings (Aley and Levine, 1999; Ferrari et al., 2010), in animals treated with wtNGF, PGE2 induced prolonged hyperalgesia unattenuated at least for 4 h (Fig. 1H,I, blue), whereas in naive subjects, PGE2 induced only acute hyperalgesia at 30 min and the hyperalgesic effect was largely dissipated at 4 h (Fig. 1H,I, black, naive). Remarkably, NGFR100W was also as potent as wtNGF in inducing priming with prolonged PGE2 hyperalgesia, lasting at least 4 h (Fig. 1H,I, red, NGFR100W). We conclude that NGFR100W retains the binding and signaling necessary to induce hyperalgesic priming (Fig. 1H,I) despite its inability to induce acute hyperalgesia (Fig. 1B,C,E,F). These findings suggest that studies of NGFR100W may provide insights into mechanisms underlying the nociceptive and trophic functions of NGF in sensory neurons.
NGFR100W does not potentiate low H+-evoked response in sensory neurons in vitro
Among the known mechanisms of pain transduction, NGF produces acute hypersensitivity by potentiating nociceptive ion channels such as the capsaicin receptor (also known as TRPV1) and acid-sensing ion channels (ASICs) (Szallasi and Blumberg, 1999; Qiu et al., 2012; McCleskey and Gold, 1999; Chuang et al., 2001; Julius and Basbaum, 2001; Mamet et al., 2003; Yen et al., 2009). ASICs and TRPV1 are proton-gated nonselective cation channels that mediate acid-evoked pain in peripheral sensory neurons (Caterina et al., 2000; Davis et al., 2000; Chen et al., 2002; Hellwig et al., 2004). We next performed patch-clamp recordings to test whether NGFR100W no longer elicited nociceptive response at the cellular level in cultured rat E15.5 DRG sensory neurons. At 8 d in culture (DIC8), neurons were starved for NGF for 2 h. We used a moderate acidic solution, pH 5.5, to activate ASICs and TRPV1 by puffing the patch-clamped cell body (Fig. 2A). Proton-evoked currents were measured before and after a 10 min application of either wtNGF or NGFR100W (Fig. 2B). The responses were normalized by calculating the ratio of the “after NGF” value to the “before NGF” value. Consistent with previous studies (Koplas et al., 1997; Shu and Mendell, 1999a,b), 10 min of wtNGF treatment produced an acute sensitization, as displayed by a 1.5-fold increase in proton evoked current (Fig. 2C). However, NGFR100W did not induce hypersensitization (Fig. 2C; wtNGF vs NGFR100W; p < 0.01; paired t test).
Low H+-evoked response by single-cell patch clamping. Rat E15.5 DRG neurons were cultured as described in the Materials and Methods. At DIC5, DRG neurons were deprived of NGF for 2 h. A, Phase contrast image of DRG neurons (under 60× magnification) showing the experimental setup. The patch pipette approached from the left and another glass pipette that applied brief puffs of pH 5.5 solution to the nearby cell body was placed at a distance. This pipette did not induce mechanical responses. B, Whole-cell patch-clamp recording was performed at a holding potential of −60 mV in DRG. The proton-evoked response was measured after a brief application of moderate acidic solution of pH 5.5 (blue arrow) onto the cell body to induce inward current (designated as “before NGF response”). Then, 50 ng/ml of wtNGF or NGFR100W (green arrow) was applied to the bath solution for 10 min and three additional puffs were applied to record the after NGF response. C, The data were normalized by calculating the ratio of the after NGF response to the before NGF response. Bar graphs represent mean ± SEM. wtNGF sensitized the inward current by 1.45-fold, but not NGFR100W (p = 0.0096; wtNGF vs NGFR100W; unpaired t test). **p < 0.01.
NGFR100W induces differentiation and supports survival of rat E15.5 DRG neurons
HSAN V patients show reduced responses to painful stimuli but retain normal cognitive function (Einarsdottir et al., 2004). We thus speculated that NGFR100W would sustain trophic signaling, resulting in the differentiation and survival of DRG sensory neurons (Winter et al., 1988; Wu et al., 2007). To compare the bioactivity of NGFR100W with wtNGF, we performed a dose–response survival assay of DRG neurons for both wtNGF and NGFR100W (0, 10, 50, and 100 ng/ml). We used E15.5 DRG and measured the number of healthy DRGs at DIC8 after treating with either wtNGF or NGFR100W (Fig. 3A). At DIC8, phase contrast images of DRG cultures were taken and survival was quantitated (Fig. 3B). Unpaired t test was performed to compare the number of healthy DRGs treated with wtNGF or NGFR100W. At any concentration of treatment, NGFR100W exerted a trophic effect to support the survival of DRGs comparably to wtNGF. As a control, cultures with no addition of NGF showed significant death (Fig. 3A). We thus conclude that NGFR100W supports the survival of rat E15.5 DRG neurons and that it does so as effectively as wtNGF.
DRG survival assays. Rat E15.5 DRG neurons were cultured as described in the Materials and Methods. Parallel cultures were supplied with either wtNGF or NGFR100W at a range from 0 to 100 ng/ml. A, Phase contrast images of DRG neurons at 8 d in vitro were captured and representative images are shown. Negative control (no NGF, 0 ng/ml) failed to maintain the survival of DRGs. NGFR100W maintained the survival as potently as wtNGF. B, The survival rate of DRG neurons (i.e., cell counts) by wtNGF or NGFR100W was not significantly different. E, Unpaired t test was performed on wtNGF versus NGFR100W (p > 0.9999 at 0, 10, and 50 ng/ml, p = 0.4000 at 100 ng/ml). Data are presented as mean ± SEM.
NGFR100W binds to and is internalized through TrkA, but not p75NTR
NGF binds and signals through TrkA and/or p75NTR receptors to effect neuronal function (Frade and Barde, 1998; Yoon et al., 1998; Sofroniew et al., 2001; Chao, 2003). We then explored whether NGFR100W differed from wtNGF in binding and internalization through TrkA and p75NTR. We used a NIH3T3 cell line that stably expresses either TrkA or p75NTR (Hempstead et al., 1991; Kaplan et al., 1991; Zhou et al., 1994; Huang et al., 1999) to perform in-cell binding assays. NGFR100W and wtNGF were each labeled with Quantum Dots 655 (QD655) before incubating with either NIH3T3-TrkA-, or NIH3T3-p75NTR cells. Saturable binding was demonstrated by using a range of NGF concentrations (Fig. 4A,B). Binding data were fit to a classical hyperbolic binding curve (one-site binding) and nonlinear regression analysis using Prism 6 software. The results showed that NGFR100W binding to TrkA (Kd = 2.79 nm) was essentially indistinguishable from that for wtNGF (Kd = 2.27 nm; Fig. 4A). In contrast, NGFR100W showed minimal binding to p75NTR even at concentrations at which wtNGF binding was essentially saturated (Fig. 4B).
Classical hyperbolic saturation curves of wtNGF and NGFR100W bound to TrkA or p75NTR-expressing NIH3T3 cells. A, B, Either NIH3T3 TrkA or p75NTR cells were exposed to wtNGF-QD655/or NGFR100W-QD655 for 20–30 min at 20°C with a range of different NGF-QD655 concentrations. Nonlinear regression analysis using observed data were performed to examine the binding of NGF to the TrkA receptor (A) or p75NTR (B). Blue squares and red circles represent surface-bound wtNGF and NGFR100W, respectively. C, D, Biotinylated wtNGF (C) or NGFR100W (D) was incubated with recombinant extracellular domain of p75NTR at 4°C for 2 h. The complex that contained biotinylated NGF was pulled down using streptavidin–agarose (SA) beads at 4°C for 2 h. The beads were washed and boiled in SDS sample buffer. Both an aliquot of the supernatant (control) and the bead-bound samples were analyzed by Western blotting using a specific antibody against p75NTR (REX). Representative blot shows p75NTR bound to wtNGF-SA beads (C) and p75NTR bound to NGFR100W-SA beads (D). Relative REX signal from pulled-down beads was significantly increased with increased amount of wtNGF, as shown in C. E, ANOVA analysis was performed with different amounts of wtNGF, suggesting that pulled-down p75 ECDs were significantly increased according to the increased amount of wtNGF (F(4,5) = 44.41, p = 0.0004). D, In contrast, pulled-down p75 ECD was not detectable even with 10 ngf of NGFR100W and only reached the level of baseline Rex signal intensity. Data are presented as mean ± SEM.
To further confirm that NGFR100W binding to p75NTR was markedly reduced, we performed in vitro binding assays for p75NTR using the extracellular domain (p75NTR-ECD). Either wtNGF or NGFR100W in their biotinylated forms at a final concentration ranging from 0 to 0.25 nm was incubated with recombinant p75NTR-ECD; streptavidin–agarose beads were used to pull down p75NTR-ECD that bound to biotinylated wtNGF or NGFR100W. The levels of p75NTR-ECD were assayed by immunoblotting using the anti-p75NTR antibody, REX. Figure 4, C and E, shows that increasing concentrations of wtNGF resulted in increasing amounts of p75NTR-ECD in the pull-down complex. In contrast, NGFR100W failed to pull down detectable p75-ECD even at the highest concentration (Fig. 4D). We conclude that the binding affinity of NGFR100W for p75NTR is markedly reduced with respect to that for wtNGF; therefore, NGFR100W fails to bind p75NTR.
To extend these analyses, we then assayed whether wtNGF and NGFR100W differed with respect to TrkA- or p75NTR-receptor-mediated internalization. Because NGFR100W showed reduced or absent binding to p75NTR, NGFR100W would also fail to be internalized via p75NTR. We performed live cell imaging using NIH3T3-TrkA- or NIH3T3-p75NTR-expressing cells (Huang et al., 1999). In addition to wtNGF and NGFR100W, we also took advantage of two well characterized NGF mutants: the KKE mutant, which shows a significant decrease in binding affinity for p75NTR but unaltered binding affinity for TrkA (Ibáñez et al., 1992; Mahapatra et al., 2009), and the Δ9/13 mutant, which poorly binds to TrkA while maintaining normal binding affinity for p75NTR (Hughes et al., 2001). Accordingly, KKE and Δ9/13 NGF served as positive controls for binding to TrkA and p75NTR, respectively. We produced monobiotinylated forms of KKE and Δ9/13 NGF along with wtNGF and NGFR100W to facilitate conjugation to QD655. These QD-655 labeled forms of NGF were incubated with cells at a final concentration of 0.2 nm at 37°C (30 min for NIH3T3-TrkA cells and 2 h for NIH3T3-p75NTR cells) to allow for internalization. Incubations were followed by extensive washing at 4°C in PBS (3×). QD655 signals were captured by live imaging and defined as internal if the QD signal was found within the perimeter of the cell and at the same focal level as the nucleus. As with wtNGF and the KKE mutant, NGFR100W was internalized by NIH3T3-TrkA cells (Fig. 5A). The process was receptor mediated because premixing with 100× wtNGF (i.e., not conjugated to QD 655) eliminated NGFR100W-QD655 internalization. The Δ9/13 mutant was not internalized in NIH3T3-TrkA cells (Fig. 5A,C). As predicted, NGFR100W and the KKE mutant failed to be internalized into NIH3T3-p75NTR cells; with both wtNGF and the Δ9/13 mutant, bright QD signals were detected inside NIH3T3-p75NTR cells (Fig. 5B,C). As a control, nonconjugated QDs were not detected in either cell line. These results are further evidence that NGFR100W is similar to wtNGF in that it binds to TrkA, but differs in not binding to p75NTR.
Imaging analysis of binding and internalization into NIH3T3-TrkA- or -p75NTR-expressing cells. A, NIH3T3-TrkA cells were cultured on coverslips that were precoated with poly-L-lysine. Cells were rinsed and serum-starved for 2 h. Cells were then incubated with 0.2 nm of wtNGF, or NGFR100W, KKE, or delta9/13-QD655 conjugates or with QD655 alone for 30 min at 37°C. After extensive rinses, QD655 signals were captured by live cell imaging. Representative images are shown. wtNGF, NGFR100W, and the KKE mutant that were used as positive controls for the TrkA receptor showed bright QD655 signals inside the cells. To ensure that internalization of NGFR100W was receptor mediated, we premixed NGFR100W with 100-fold unlabeled NGF before live imaging. The results show little internalization of NGFR100W, indicating the internalization of NGFR100W into NIH3T3-TrkA cells was TrkA specific. B, Similarly, NIH3T3-p75NTR cells were used to investigate internalization of the different forms of NGF proteins and representative images are shown after incubating with NGF-QD655 for 2 h at 37°C. The results show that both wtNGF and Δ9/13, which are known to bind to the p75NTR receptor, were internalized into NIH3T3-p75NTR cells and the QD655 signals were mostly concentrated around the peripheries of the cell. No signals were observed in the NIH3T3-p75NTR cells when treated with either the KKE mutant or NGFR100W. C, Internalized QD655 within the cells were quantitated. Data are presented as mean ± SEM.
NGFR100W activates TrkA-mediated signaling pathways, but fails to stimulate a p75NTR downstream effector
Based on the binding and internalization results, we predicted that NGFR100W activated the TrkA-mediated, but not the p75NTR-mediated signaling, pathway. We next determined whether wtNGF and NGFR100W activated two main effectors of TrkA downstream signaling cascades: Erk1/2 and Akt using PC12 cells (Fig. 6A). For this purpose, PC12 was stimulated with 50 ng/ml of either wtNGF or NGFR100W. Cells then were lysed and analyzed using immunoblotting. Figure 6, H, shows that NGFR100W induced phosphorylation of TrkA, Erk1/2, and Akt to the extent comparable to wtNGF. We also performed semiquantitative measurement of pTrkA activated by wtNGF or NGFR100W. The signals for pTrkA were normalized against GAPDH (Fig. 6-1). We did not detect a significant difference in pTrkA (Fig. 6-1). These findings suggest that NGFR100W maintains its ability to activate TrkA, Erk 1/2, and Akt signaling pathways, which are primarily involved in providing trophic support for neuronal survival and differentiation.
Analysis of TrkA- and p75NTR-mediated signaling pathways. A–C, PC12 cells were serum-starved and treated with 50 ng/ml of either wtNGF or NGFR100W for the indicated time intervals. Cell lysates were analyzed by SDS-PAGE and immunoblotting with specific antibodies as indicated. Treatment with either wtNGF or NGFR100W induced phosphorylation; i.e., activation of TrkA, Erk 1/2, and Akt. The blots were reprobed for total Akt or Erk1/2 as loading controls. We also measured the signals for pTrkA activated by wtNGF or NGFR100W. The signals for pTrkA were normalized against GAPDH (Fig. 6-1). D, E, Analysis of downstream signaling of p75NTR in PC12 nnr5 by immunoblotting. The cells were prepared, treated, and cell lysates analyzed by SDS-PAGE/immunoblotting with specific antibodies as in A. There was a significant reduction in phosphorylation of Cofilin by NGFR100W compared with wtNGF. F, Immunostaining of RhoA-GTP in PC12 nnr5. PC12 nnr5 were plated on the coverslip coated by poly-L-lys, starved for 2 h, treated with either wtNGF or NGFR100W, and the preparations were fixed and permeabilized, followed by the protocol. Differential interference contrast imaged cells show single staining for active form of RhoA (RhoA-GTP) or double staining for active RhoA and nucleus. RhoA-GTP staining revealed that RhoA activation was stronger in the cell treated with wtNGF than in the cell treated with NGFR100W. G, H, Analysis of PLC-γ signaling in PC12 cells by immunoblotting. G shows that PLC-γ stimulated by NGFR100W differs significantly from the one by wtNGF. In contrast, the same lysate showed similar amount of activation of Erk1/2 and Akt. PC12 cells were pretreated with the p75NTR inhibitor Pep15, followed by treatment with 50 ng/ml NGF. In parallel samples, cells were treated with vehicle, 50 ng/ml NGF, or 50 ng/ml NGFR100W. Cell lysates were analyzed by SDS-PAGE/immunoblotting with specific antibodies as indicated. The data show thatactivation of PLC-γ was markedly suppressed by NGFR100W compared with wtNGF. wtNGF failed to fully activate PLC-γ when p75NTR was functionally inhibited, similar to partially activated PLC-γ when treated by NGFR100W. Data are presented as mean ± SEM. p values were calculated using student unpaired t-test. *p < 0.05, **p < 0.01, ***p < 0.001.
Figure 6-1
We then assessed whether both wtNGF and NGFR100W stimulated two signaling pathways downstream of p75NTR receptors. We investigated whether RhoA and Cofilin, two effectors downstream to p75NTR (Yamashita and Tohyama, 2003; Vardouli et al., 2005), were phosphorylated or activated by wtNGF or NGFR100W. Using PC12nnr5 cells that express p75NTR with little or no TrkA (Loeb and Greene, 1993), we tested whether treatment of these cells with either wtNGF or NGFR100W induced phosphorylation of Cofilin (i.e., p-Cofilin) by immunoblotting using an antibody that specifically recognizes the phosphorylated form of Cofilin. Our results show that Cofilin was activated when cells were treated by wtNGF. However, the level of phosphorylated Cofilin was significantly less in cells treated by NGFR100W (Fig. 6D,E). We then assayed for activation of RhoA, a signaling molecule upstream of Cofilin in the p75NTR signaling cascades (Vardouli et al., 2005). As demonstrated by immunostaining with specific antibody to activated RhoA (i.e., RhoAGTP), RhoAGTP exhibited marked activation by wtNGF, as evident by strong cytosolic staining (Fig. 6F). In contrast, NGFR100W resulted in much less activation of RhoAGTP, with only sparse speckles of signals in the cytoplasm (Fig. 6F). These data further confirm that NGFR100W is ineffective in activating signaling cascades downstream of p75NTR.
Interestingly, we observed that NGFR100W was unable to fully induce phosphorylation of PLC-γ (Fig. 6G). Therefore, we investigated whether failure of activation of the p75NTR signaling by NGFR100W was responsible for the inability to fully phosphorylate PLC-γ. Indeed, there has been evidence suggesting that activated p75NTR downstream effectors positively affect downstream of TrkA signaling (Ruiz-Argüello et al., 1996; Basáñez et al., 1997; Ruiz-Argüello et al., 1998; Cremesti et al., 2002). For example, others have shown that ceramide-induced changes in membrane microenvironments facilitate PLC signaling. We therefore speculated that treating cells with wtNGF under conditions in which p75NTR was inhibited would cause a reduction in PLC-γ activation, as was the case for NGFR100W (Fig. 6H). We used PC12 cells to determine whether p75NTR inhibition affected PLC-γ by possible crosstalk between p75NTR and TrkA. Our results showed that phosphorylation of PLC-γ was decreased when PC12 was pretreated with a p75NTR inhibitor, TAT-pep5 (Head et al., 2009), followed by stimulation with wtNGF (Hasegawa et al., 2004), as shown in Figure 6F. These results suggest that failure of p75NTR signaling by NGFR100W leads to failure of PLC-γ signaling by as yet unknown mechanisms, even though NGFR100W stimulates other downstream signaling such as Erk1/2 and Akt under TrkA. NGFR100W activates most TrkA downstream signaling events, but not those mediated by p75.
NGFR100W is retrogradely transported in a fashion similar to wtNGF
Axons feature prominently in facilitating trafficking and signaling of NGF. A possibility existed that the loss of pain signaling in NGFR100W was due to its inability to be effectively transported retrogradely to the cell soma in sensory neurons. The ability of R100W to induce priming suggested that it was effectively transported retrograde in axons to engage the cell body responses that support priming. To test this possibility, we established microfluidic cultures of dissociated rat E15.5 DRG neurons. To visualize trafficking of NGF, we conjugated biotinylated wtNGF or NGFR100W with streptavidin-QD605 at a 1 NGF dimer to 1 QD605 ratio. This experimental paradigm was used to track axonal movement of a single NGF dimer by live imaging to produce highly quantitative results (Sung et al., 2011). wtNGF was retrogradely transported with an average speed of 1.5 μm/s (Fig. 7A), which is consistent with previous results (Cui et al., 2007; Sung et al., 2011). Based on the analysis of >100 endosomes/condition, the average moving speed of NGFR100W revealed no marked difference from that of wtNGF (Fig. 7B).
Analysis of retrograde axonal transport by live imaging. A, B, Rat E15.5 DRG neurons were cultured in microfluidic chamber. Kymographs of axonal movement of wtNGF and NGFR100W based on real-time imaging series of axonal transport assays are shown. The graphs represent spatial position of QD605 signals (in micrometers) over time (seconds). The results for both wtNGF and NGFR100W showed similar slopes, suggesting that NGFR100W moves at a speed within the axon similar to that of wtNGF. C, Overlayed kymographs of displacement of axonal QD605 signals for wtNGF and NGFR100W. The 100–150 QD605 signals for either wtNGF (blue) or NGFR100W (red) were analyzed and superimposed. The results demonstrate that axonal movement of wtNGF and NGFR100W behaves in a strikingly similar fashion. D, Total average transport speeds including pausing for wtNGF and NGFR100W were calculated to be 1.5 and 1.4 mm/s, respectively. The moving velocity without pausing, that is, during the “go” motion period, was 1.7 mm/s for both wtNGF and NGFR100W.
NGF is known to exhibit a “go and stop” behavior during transport (Cui et al., 2007; Sung et al., 2011). Therefore, we tested whether the moving speed during the “go” period differed significantly between wtNGF and NGFR100W. Our results demonstrated that this was not the case (Fig. 7D). By superimposing the kymographs of NGFR100W onto those of wtNGF (Fig. 7C), our results confirmed that NGFR100W behaved exactly like wtNGF during retrograde transit from the axonal terminal to the cell body. We thus conclude that the critical function in retrograde axonal transport of NGF in sensory neurons is preserved in NGFR100W.
Study of contribution of TrkA or p75NTR to NGF-induced sensitization effect in vivo
Our findings suggest that NGFR100W retained its ability to bind to and activate TrkA while failing to engage p75NTR. However, when injected into adult rats, NGFR100W still induced hyperalgesic priming without causing acute sensitization to mechanical stimuli. These results raise the possibility that TrkA and p75NTR may play a distinct role(s) in NGF-induced hyperalgesia and hyperalgesic priming. We used pharmacological reagents to selectively inhibit signaling downstream of TrkA and/or p75NTR: K252a for blocking TrkA activation and GW4869 for inhibiting neutral sphingomyelinase downstream of p75NTR.
Vehicles or inhibitors (K252a, GW4869, K252a + GW4869) were injected only once 5 min before NGF injection (Fig. 8A). We then measured the mechanical nociceptive threshold at 1 h and 7 d after NGF injection, as outlined in the experimental design (Fig. 8A). K252a produced a robust reduction (∼30%) in NGF hyperalgesia (p = 0.0001); GW4869 had a small but insignificant reduction (∼7%, p = 0.0557); combination of K252a with GW4869 caused a ∼25% reduction (p < 0.0001). Consistent with results presented in Figure 1, NGF-induced hyperalgesia had returned to baseline nociceptive threshold at day 7 after NGF treatment regardless of inhibitors (Fig. 8A).
Role of TrkA and p75NTR in nociceptive response and hyperalgesic priming. A, Five microliters (1 μg) of K252a, GW4869, or vehicle (10% DMSO) was administered intradermally on the dorsum of the hindpaw of adult rats at the same site where NGF was going to be injected. After NGF injection (200 ng of wtNGF), the Randall–Selitto method was used to measure mechanical hyperalgesia as in A. n = 4 for vehicle, K252a, GW4869; n = 6 for K252a + GW4869. Data are presented as mean ± SEM. p-values were calculated using unpaired t test. ***p < 0.001. B, Higher dosages of GW4869, 10 μg (n = 4) and 1000 μg (n = 4), were injected intradermally on the dorsum of the hindpaw of adult rats following by NGF injection as in A. The same method was applied to measure mechanical hyperalgesia used as in A. Data are presented as mean ± SEM. p-values were calculated using unpaired t test. *p < 0.05, **p < 0.01.
We then measured the impact of these inhibitor treatments on priming after injection of PEG2, as in Figure 1. Treatment with K252a, GW4869, or both did not change acute sensitization significantly 30 min after PEG2 injection (Fig. 8A). However, at 4 h after PEG2 injection, K252a and GW4869 each alone produced a moderate reduction in PGE2 hyperalgesia (p = 0.0076 for K252a; p0.0359 for GW4869); Strikingly, combination of K252a with GW4869 induced complete elimination of PGE2 hyperalgesia (p = 0.0006; Fig. 8A). These results further support an interaction between TrkA and p75NTR in mediating the priming effect by NGF.
Because there is discrepancy between NGFR100W injection studies (Fig. 1B,C) versus GW4869 treatment in inducing acute hypersensitivity (Fig. 8A), we repeated the Randall–Selitto experiment with higher dosages of GW 4869. We chose two different dosages, 10 μg and 1000 μg. Previous studies have demonstrated that 11.55 μg/injection of GW 4869 was effective in blocking acute sensitization by NGF (Khodorova et al., 2013, 2017). The 1000 μg dose was to ensure that a maximal effect was achieved. As demonstrated in Figure 8B, 10 μg/injection appeared to yield a maximal effect; both dosages, 10 μg and 1000 μg, successfully prevented NGF-induced acute hyperalgesia at 1 h after NGF injection, which is consistent with published studies (Khodorova et al., 2013, 2017), but also from our own studies presented in Figure 1, B and C. At 1 h after NGF injection, vehicle injection decreased the mechanical pain threshold ∼34.57% and 10 μg and 1000 μg of GW4869 decreased the threshold ∼19.14% (p = 0.0043) and ∼17.64% (p = 0.0303), respectively. (Fig. 8B). We then measured priming after injection of PGE2 with these two high dosages of GW4869. Both dosages also blocked priming significantly compared with vehicle (p = 0.0087 and p = 0.0043; Fig. 8B). These data further confirmed our result pointing to the loss of p75NTR downstream signaling in NGFR100W as a mechanism underlying loss of pain perception. Together, these data suggest that both TrkA and p75NTR are responsible for mediating acute hypersensitivity and priming.
Discussion
Studies on a naturally occurring mutation in HSAN V patients, NGFR100W, have provided new insights into NGF-mediation of nociceptor function. Using molecular, cellular, and biochemical techniques, we demonstrated that NGFR100W retained the ability to bind and signal through TrkA receptors and robustly support the trophic status of DRGs. In contrast, NGFR100W failed to bind or activate of p75NTR. Our in vivo studies have demonstrated that inhibition of TrkA and p75NTR signaling, as well as delivery of antisense reagents targeting these receptors, provided evidence for a role for both receptors in nociceptor regulation. Whereas TrkA activation mediated acute sensitization and priming, p75NTR signaling appeared to contribute only to priming, thereby augmenting the TrkA response. Remarkably, and in contrast to its ability to induce trophic responses via TrkA, NGFR100W induced nociceptor priming, but differed from wtNGF by not causing acute sensitization. These observations are evidence for distinct TrkA signaling pathways mediating trophic effects and nociceptor function and provide insights into the biology of HSAN V.
Consistent with previous results (Covaceuszach et al., 2010; Capsoni et al., 2011), we confirmed that, whereas NGFR100W binding and activation of TrkA is robust, binding and signaling through p75NTR is essentially absent. Important to its ability to prime nociceptors, NGFR100W was internalized at axonal tips of DRG sensory neurons and traveled toward the cell body at the speed and with the velocity characteristic of wtNGF. Given properties in common with wtNGF, it is unclear why NGFR100W failed to induce acute sensitization. Possibilities include differences with respect to wtNGF in the structure of the NGFR100W/TrkA complex, the structure of the complex in signaling endosomes, or the downstream partners recruited to these complexes. Furthermore, it is hard to exclude the possibility that TrkA downstream signaling events implicated in pain could be positively affected by crosstalk between p75NTR and TrkA, as suggested by only partial PLC activation by NGFR100W. This suggests the possibility that TrkA downstream events that support trophic functions are retained in NGFR100W, whereas p75NTR downstream and some TrkA downstream events that affect pain threshold are reduced in NGFR100W. In any case, it is apparent that differences must exist for the signaling events that subserve TrkA trophic functions and its acute effects on nociceptor sensitization. Further studies to decipher the basis for these differences will benefit from the ability to compare the signaling properties of NGFR100W/TrkA with wtNGF/TrkA.
Our data help to explain the clinical manifestations of NGFR100W mutation. Studies of affected families point to considerable interpatient variability. Both homozygous and heterozygous individuals demonstrate orthopedic manifestations, the most frequent of which is multiple painless fractures, typically in the legs and feet (Einarsdottir et al., 2004; Minde et al., 2004; Larsson et al., 2009; Capsoni, 2014). Homozygous patients show decreased pain sensation, mainly at the forearms and legs. In contrast, they respond to truncal pain and register visceral pain. Distal testing of temperature thresholds showed increases in some homozygous and heterozygous patients. Sensitivity to soft touch, joint position, vibration sensation, and visceral pain are normal. Sural nerve biopsy reveals loss of C and A delta sensory fibers in both homozygotes and heterozygotes (Minde et al., 2004, 2009; Minde, 2006; Sagafos et al., 2016). All patients show reduced sensory innervation of skin and reduced sympathetic innervation of sweat glands, with more marked changes in homozygotes (Axelsson et al., 2009). Although a decrease in pain is consistent with the inability of NGFR100W to induce acute nociceptor sensitization, the clinical picture is that of a length-dependent sensory and sympathetic neuropathy. Indeed, this explains best the painless fractures and decreased pain sensation in distal lower limbs, increased distal thresholds for cold and heat perception, loss of sensory and sympathetic innervation of skin, and preservation of truncal pain. In view of the preservation of TrkA signaling by NGFR100W, the question arises as to how NGFR100W causes the syndrome. The likely cause is failure to secrete the protein in sufficient amounts to support the distal axons of sensory and sympathetic neurons. NGF is critical for the survival and maintenance of sympathetic and sensory neurons (Ibáñez et al., 1992; Kew et al., 1996; Casaccia-Bonnefil et al., 1998; Sofroniew et al., 2001; Chao, 2003). Decreased secretion of NGFR100W, as demonstrated in other studies, must be confirmed with human cells expressing the mutant protein (Larsson et al., 2009; Covaceuszach et al., 2010).
A causal link between NGF deficiency in innervated targets and neuronal dysfunction and degeneration was established during the earliest studies on NGF (Levi-Montalcini and Hamburger, 1951; Cohen and Levi-Montalcini, 1956; Levi-Montalcini and Angeletti, 1961, 1963; Levi-Montalcini, 1964) and has been shown in studies of the developing and mature nervous system. Attempts to increase NGF availability in the hopes of reducing neurodegeneration (Wilcox and Johnson, 1988; Apfel et al., 1994; Apfel and Kessler, 1995, 1996; Apfel, 1999a,b, 2002; McArthur et al., 2000; Cattaneo et al., 2008) have failed in part due to NGF dose-limiting pain (Eriksdotter Jönhagen et al., 1998; Apfel, 2000, 2002). That NGFR100W maintains trophic functions without acutely sensitizing nociceptors has suggested that this isoform of NGF might be used to provide trophic support without causing pain (Capsoni et al., 2011, 2014). That NGFR100W maintains the ability of wt NGF to prime nociceptors raises the caution that NGFR100W treatment may not fully avoid the pain induced by wild-type NGF.
Increasing evidence supports that both TrkA- and p75NTR-mediated signaling pathways are intimately involved in NGF-induced hyperalgesia (Nicol and Vasko, 2007). Extensive human genetic studies strongly support an essential role played by TrkA in pain sensation; mutations in TrkA that result in loss or reduced TrkA activity are associated with congenital insensitivity to pain with anhidrosis (Indo, 2001, 2002). In addition, inhibiting of TrkA-mediated signaling pathways such as Erk1/2 (Aley et al., 2001; Dai et al., 2002), PI3K/Akt (Zhuang et al., 2004), and PLC-γ (Chuang et al., 2001) has been shown to block NGF-induced sensitization both in vivo and in vitro. Consistent with these findings, we used specific antisense oligos to attenuate expression of TrkA via intrathecal administration or administered K252a to inhibit TrkA activation via acute intraplantar injection. Both approaches affirmed that TrkA is required for both the acute and chronic phase of sensitization induced by NGF.
Unlike TrkA, a role for p75NTR in NGF-induced hyperalgesia has been implicated largely by indirect evidence: (1) intrathecal administration of anti-p75NTR into animals reduced temperature hyperalgesia and mechanical allodynia after nerve injury (Obata et al., 2006); (2) direct application of an p75NTR antibody to a crushed sciatic nerve suppressed mechanical allodynia (Fukui et al., 2010); (3) pretreatment with a p75NTR antibody prevented the increase in the number of action potentials induced by NGF (Zhang and Nicol, 2004); (4) intraplantar injection of proNGF, which selectively activates p75NTR, and not TrkA, induced hyperalgesia (Khodorova et al., 2013); and (5) NGF-induced sensitization was attenuated by inhibiting p75NTR-mediated activation of the sphingomyelin-ceramide-sphingosine 1 phosphate and the c-JUN kinase pathway (Zhang et al., 2002, 2006; Doya et al., 2005; Obata et al., 2006; Khodorova et al., 2013).
We confirmed a role for p75NTR in acute sensitization of mechanical nociceptors using GW4869. However, a discrepancy exists between priming studies of NGFR100W in Figure 1 versus GW4869 in Figure 8A; NGFR100W induced a priming effect comparably to wtNGF, whereas GW4869 failed to block the priming effect by wtNGF even at higher dosages (Fig. 8B). This could be explained by the lack of specificity of GW4869 in blocking neutral SMase2 (Canals et al., 2011). The GW4869 specificity issue was also raised in studies showing that GW4869 not only blocks SMase, but also PLC and PP2A (Luberto et al., 2002). Moreover, previous studies have shown that GW4869 abrogated NGF-mediated TrkA trophic effects on cell viability (Candalija et al., 2014), suggesting that block of nSMase2 could lead to partial block of TrkA-mediated trophic signaling such as pAkt (Gills et al., 2012). Therefore, the GW4869 effect could not be the same as loss of p75 downstream by NGFR100W. This may explain the inconsistency in our data showing incapability of priming by GW4869 versus the priming capability by NGFR100W.
We are also aware of the specificity issues associated with the use of K252a inhibitor to block TrkA. K252a selectivity has been studied in many studies. For example, K252A is known to inhibit PKCs, which is also implicated in p75 downstream (Mizuno et al., 1993). Not only inhibiting TrkA, at certain range of concentration, K252a was also found to act as partial inhibitor of the PDGF receptor (Nye et al., 1992).
Given all the caveats associated with the use of inhibitors in our studies, our data suggest that p75NTR plays an important role in NGF-induced pain function (Zhang and Nicol, 2004; Watanabe et al., 2008; Iwakura et al., 2010; Khodorova et al., 2013, 2017). When p75 inhibition is combined with TrkA suppression, both acute response and priming resulted in a greater reduction in NGF-mediated hyperalgesic priming. These results support a role for p75NTR that is most evident in its ability to synergize with TrkA to mediate NGF-induced hyperalgesic priming. Therefore, the nociceptive functions of NGF include contributions from both TrkA and/or p75NTR.
To avoid the pitfalls associated with the use of inhibitors for TrkA- or p75NTR-signaling pathways, future studies using novel genetic models are needed to further study the contributions of TrkA and p75NTR to NGF-induced sensitization. For example, TrkA activity can be specifically inhibited by injection of nanomolar concentrations of derivatives of the general kinase inhibitor PP1 (1NMPP1 or 1NaPP1) to block NGF signaling in TrkA (F592A)-knock-in mice (Chen et al., 2005). Ablation of p75NTR can be induced in conditional p75NTR knock-out mice (Bogenmann et al., 2011; Wehner et al., 2016). Use of these models is expected to further clarify the contributions of TrkA and p75NTR and suggest the possibility that their signaling pathways interact to effect NGF-induced sensitization.
Footnotes
This work was supported by the National Institutes of Health [Grant PN2 EY016525 and University of California–San Diego (UCSD) Alzheimer's Disease Research Center P50 Pilot Grant to W.C.M.; Grant R01 NS084545 to J.D.L.; and UCSD T32 Neuroplasticity of Aging Training Grant to K.S.], the Down Syndrome Research and Treatment Foundation (W.C.M.), the Larry L. Hillblom Foundation (W.C.M.), the Alzheimer's Association (W.C.M., the Thrasher Research Fund (W.C.M.), and the Larry L. Hillblom Foundation (Tauopathy Foundation startup grant to C.W.). We thank Pauline Hu for assistance in DRG cultures, and R. Shibata and N.D. Storslett for assistance with protein purification.
The authors declare no competing financial interests.
- Correspondence should be addressed to Dr. Chengbiao Wu, Department of Neurosciences, University of California San Diego, Medical Teaching Facility, Room 312 MC-0624, 9500 Gilman Drive, La Jolla, CA 92093. chw049{at}ucsd.edu
References
In this issue
