Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2014 Jan 22;34(4):1494-509.
doi: 10.1523/JNEUROSCI.4528-13.2014.

TREK2 expressed selectively in IB4-binding C-fiber nociceptors hyperpolarizes their membrane potentials and limits spontaneous pain

Cristian Acosta  1 , Laiche Djouhri, Roger Watkins, Carol Berry, Kirsty Bromage, Sally N Lawson
Affiliations

TREK2 expressed selectively in IB4-binding C-fiber nociceptors hyperpolarizes their membrane potentials and limits spontaneous pain

Cristian Acosta et al. J Neurosci. .

Abstract

Ongoing/spontaneous pain behavior is associated with ongoing/spontaneous firing (SF) in adult DRG C-fiber nociceptors (Djouhri et al., 2006). Causes of this SF are not understood. We show here that conducting (sometimes called uninjured) C-nociceptors in neuropathic pain models with more hyperpolarized resting membrane potentials (Ems) have lower SF rates. Understanding the control of their Ems may therefore be important for limiting pathological pain. We report that TREK2, a leak K(+) channel, is selectively expressed in IB4 binding rat C-nociceptors. These IB4(+) C-neurons are ∼10 mV more hyperpolarized than IB4(-) C-neurons in vivo (Fang et al., 2006). TREK2 knockdown by siRNA in these neurons in culture depolarized them by ∼10 mV, suggesting that TREK2 is responsible for this ∼10 mV difference. In vivo, more hyperpolarized C-nociceptor Ems were associated with higher cytoplasmic edge-TREK2 expression (edge-TREK2). Edge-TREK2 decreased in C-neurons 7 d after axotomy, and their Ems depolarized by ∼10 mV. This again supports a contribution of TREK2 to their Ems. These relationships between (1) Em and TREK2, (2) SF rate and Em, and (3) spontaneous pain behavior and C-nociceptor SF rate suggested that TREK2 knockdown might increase spontaneous pain. After CFA-induced inflammation, spontaneous foot lifting (a measure of spontaneous pain) was (1) greater in rats with naturally lower TREK2 in ipsilateral small DRG neurons and (2) increased by siRNA-induced TREK2 knockdown in vivo. We conclude that TREK2 hyperpolarizes IB4 binding C-nociceptors and limits pathological spontaneous pain. Similar TREK2 distributions in small DRG neurons of several species suggest that these role(s) of TREK2 may be widespread.

Keywords: DRG; K2P; TREK2; nociceptors; spontaneous firing; spontaneous pain.

PubMed Disclaimer

Figures

Figure 1.
Figure 1.
A, TREK2 antibody characterization. Top, GFP fluorescence or interference contrast (IC) shows HEK cells, transfected with plasmids for six rat K2P channels: TREK2, TWIK1, TREK1, TASK3, THIK2, and THIK1 (all subcloned in pCMV-SPORT vector). Bottom, Same fields show immunostaining for TREK2 that was intrinsically very low (GFP alone), increased substantially after transfection with the chimeric TREK2-GFP and with TREK2, but remained very low after transfection with all non-TREK2 K2P channels tested. This antibody therefore selectively recognizes TREK2 and not the other K2P channel tested, including TREK1, structurally the closest related to TREK2. Scale bar, 40 μm. B–D, TREK2 in DRG neurons is selectively expressed in IB4+ neurons. B, Triple fluorescence immunocytochemistry shows TREK2 (blue) to be highly colocalized with IB4 binding (red). RT97 (green) is a marker for neurofilament (against a highly phosphorylated epitope on NF200) that selectively stains DRG somata with A-fibers (i.e., myelinated). TREK2 was only visible in neurofilament-poor (NF-poor, thus C-fiber) neurons, as shown by lack of (green) NF staining in TREK2+ neurons. TREK2 staining is thus in IB4+ DRG neurons with C-fibers. Scale bar, 20 μm. C, D, %intensities refers to cytoplasmic immunointensities expressed as a percentage of maximum staining above background for that marker (see Materials and Methods); gray band(s) show(s) negative staining for TREK2 (<25% %intensity) and IB4 (<20% %intensity). C, TREK2+ neuronal profiles with nuclei were <400 μm2 in cross-sectional area, and mostly <300 μm2. D, TREK2 and IB4 %intensities were positively correlated in small neurons. (Spearman's correlation coefficient, rs, shown) and had a highly significant linear regression (r2 = 0.54, p < 0.0001). There was no significant correlation in medium to large neurons; these were negative for both (data not shown). E, F, TREK2 in DRG neurons of other species. Rat images included for comparison. E, Lumbar DRG cryostat sections, from adult rat and cat (L4) and 2 week female guinea pig (L6), immunostained by ABC against TREK2. As in rat, both cat and guinea pig showed small DRG neurons positively stained for TREK2. TREK2+ neurons in cat (although larger than rat) are consistent with them being small NF-poor neurons. Small NF-poor neurons in adult cat lumbar DRGs have a mean diameter ∼40 μm; maximum ∼50 μm (Perry and Lawson, 1993); in rats, mean diameter ∼18 μm, maximum ∼25 μm. Thus, in three species, a subpopulation of small neurons shows clear TREK2 immunostaining. Scale bar, 40 μm. F, Guinea pig DRG (as above) double-labeled for TREK2 immunocytochemistry and IB4 binding. As with rat, in guinea pig strong TREK2 staining (blue) is present in a subpopulation of small DRG neurons with clear IB4 binding; red dashed lines indicate outlines of large TREK2− neurons for comparison with small TREK2+. Scale bar, 20 μm. G, H, TREK2 and trkA are poorly coexpressed in rat DRG neurons. G, TREK2 (blue) and trkA (red) (double-immunofluorescence, pseudocolored) were not visibly colocalized. White dashed lines indicate outlines of large TREK2− neurons. Scale bar, 20 μm. H, Many neurons were negative for both trkA (<20% %intensity) and TREK2 (see below); others were TREK2+ or trkA+, and few were positive for both. %intensities of neurons of all sizes were weakly but negatively correlated (Spearman's). I, J, TREK2 and Nav1.9 are colocalized in small DRG neurons. Adjacent ABC-immunostained sections were examined because both primary antibodies are from the same species. I, TREK2 and Nav1.9 were strongly colocalized. Scale bar, 25 μm. J, Strong correlation (Spearman's) between their %intensities in all sizes of neuron.
Figure 2.
Figure 2.
In vivo, TREK2 is in small C-fiber nociceptors; their Ems relate to TREK2 levels. A–G, DRG neurons injected in vivo with fluorescent dye after Em recording and sensory receptor property identification (see Materials and Methods). A, Examples (arrows) of fluorescent dye-labeled neurons before (top) and after (bottom) TREK2 ABC immunocytochemistry. TREK2%intensities (see Materials and Methods) were determined at edge-, mid-, and inner- (perinuclear) cytoplasm (B, inset). The two C-fiber nociceptors (C-Nocis) have (left) a −52 mV Em and strong cytoplasmic, including edge, TREK2 intensity; then (mid-left) a −42 mV Em, with weak edge- but stronger mid- and inner-TREK2. The Aδ nociceptor (mid-right) and Aα/β LTM (far right) were TREK2− (see Materials and Methods and below). Scale bar, 20 μm. B–G, Cytoplasmic edge- and mid-TREK2%intensities of ≥25% (of maximum cytoplasmic staining; see Materials and Methods) were deemed positive. The negative staining (<25%) is shown as gray bands. B, C, Most C-nociceptors, not A-fiber neurons, were TREK2+. This was the case for edge-TREK2 (B), mid-TREK2 (C), and inner-TREK2 (data not shown). Only C-nociceptors and a C-cooling unit were edge-TREK2+; all A-fiber neurons were edge-TREK2−. Most were also mid-TREK2−; only a single Aδ-LTM and an Aβ nociceptor were weakly mid-TREK2+. Median edge- and mid-TREK2 were highly significantly greater (p < 0.0001) in C-nociceptors than in all A-neurons (Mann–Whitney). No differences between edge- or mid-TREK2 in A-neuron subtypes were seen (Kruskall–Wallis tests). Similar patterns to mid-TREK2 were also seen for inner-TREK2 (data not shown). D, Em magnitude in C-nociceptors was inversely related to edge-TREK2 (linear regression) but not to mid-TREK2 (E) or inner-TREK2 (data not shown), although weak trends are seen (arrows indicate C-neurons illustrated in A). F, G, Em magnitude in A-fiber neurons was unrelated to edge- or mid-TREK2 immunointensity in all, or any subgroup, of A-neurons examined. Edge-TREK2%intensities (F) and mid-TREK2%intensities (G) and inner-TREK2%intensities (data not shown) were all negative (<25%). ****p < 0.0001.
Figure 3.
Figure 3.
siRNA knockdown of TREK2 in vitro caused depolarization of IB4+ DRG neurons. A, B, Western blots to show extent of TREK2 knockdown with siRNA in cultured DRG neurons. Western blots were of neuron-enriched cultures after 2 DIV. They were of adult rat DRG neurons (with NGF and GDNF) that had been transfected 6 h after plating with either 5 nm scrambled (scr) siRNA or with one of three siRNAs (A–C) from Ambion (see Materials and Methods) against rat TREK2. Naked siRNA was added 1 d after plating. Treatments and doses indicated below lanes in A. A, Top row, Single TREK2 band of ∼55 kDa. Bottom row, β-Tubulin III as loading control. B, Quantification of the three repeats of the experiment in A with tissue from three different rats relative to their loading controls. TREK2 protein levels were unaffected by scr; and compared with scr, the greatest TREK2 knockdown was with TREK2 siRNA C (Silencer Select preDesigned from Ambion) at either 5 or 10 nm (downward arrows); siRNA C at 10 nm was therefore chosen for later use in this study. C, D, TREK2 knockdown by siRNA C resulted in depolarization of IB4+ DRG neurons. C, Selection of cultured DRG neurons for Em measurement after transfection with scr-FAM (5 nm) or siRNA C (10 nm) plus scr-FAM (5 nm). Neurons selected had to show both FAM (green) fluorescence indicating likelihood of transfection plus IB4 binding (red; IB4+). White arrows indicate 2 neurons fulfilling both criteria. Scale bar, 20 μm. D, In small IB4+ FAM+ cultured DRG neurons (2 DIV), TREK2 knockdown by TREK2 siRNA C (plus scr-FAM) resulted in a highly significant median Em depolarization by ∼10 mV, compared with only scr-FAM treatment. *p < 0.05. **p < 0.01. ***p < 0.001. ****p < 0.0001. Blue dashed arrows indicate theoretical Ems calculated with the GHK equation, for normal DRG neurons and for those with no TREK2, using the percentage (69%) of Em K+ standing current that was due to TREK2 in TREK2-expressing small neonatal DRG neurons (Kang and Kim, 2006).
Figure 4.
Figure 4.
In small DRG neurons, peripheral axotomy decreased edge-TREK2 and depolarized Ems. A, Seven days after L5 spinal nerve axotomy, Ems of ipsilateral L5 C-fiber neurons in vivo were ∼10 mV more depolarized than normal (unpaired t test with Welch's correction, p = 0.034). *p < 0.05. B, Edge, mid, and inner (Fig. 2B, inset) cytoplasmic TREK2%intensities in normal (N) versus axotomized ipsilateral (Ax) L5 small (cross-sectional area <400 μm2) DRG neurons. Medians plus interquartile ranges are shown. For each region (edge, mid, inner), medians for N and Ax groups were compared (Mann–Whitney tests). After axotomy, compared with normal, edge-TREK2 decreased (p < 0.001) ipsilaterally and mid-TREK2 showed a small decrease ipsilaterally. In contrast, inner TREK2 increased ipsilaterally. *p < 0.05. ***p < 0.001. C, Edge-TREK2 is normally high. D, Ipsilaterally, edge-TREK2 was reduced substantially 7 d after axotomy. E, Higher-power (100× objective) enlargement of regions in white rectangles in D. Arrows indicate pale edge-TREK2 staining. Scale bars, 20 μm.
Figure 5.
Figure 5.
Axotomy 7 d earlier decreased Nav1.8 and Nav1.7 as well as TREK2 expression. A, Blinded subjective scoring (0 = no visible staining; 7 = maximum staining intensity) of whole cytoplasmic Nav1.8 immunointensities in L5 DRG sections from 4 normal (N) and 4 axotomized rats ipsilaterally (Ax). Medians plus interquartile ranges are shown. The median decreased significantly after axotomy (Mann–Whitney test). **p < 0.01. B, Nav1.8 (red) in TREK2+ neurons (blue): double immunofluorescence of L5 DRG sections. Left, Normal untreated. Right, After L5 spinal nerve axotomy 7 d earlier. Images are illustrative of results in the 4 rats per group. In merged images (bottom), coexpression shows as purple/mauve. Nav1.8 and TREK2 are normally colocalized, but Nav1.8 was markedly decreased in TREK2+ neurons as well as medium- and large-size neurons after axotomy. White dashed lines indicate neuronal outlines after axotomy because these are hard to detect due to decreased edge-TREK2 and decreased Nav1.8 in TREK2+ neurons. Scale bar, 20 μm. C, Nav1.7 (red) in TREK2+ (blue) neurons. Images are illustrative of similar results from 4 normal and 4 axotomized rats. Left, Normal untreated. Right, One week after ipsilateral L5 spinal nerve axotomy. Both antibodies were raised in rabbit. Immunostainings for TREK2 and Nav1.7 were thus in adjacent sections. Both sections were double-labeled with IB4 (green), and numbers are to aid alignment of neurons between sections. Normally, Nav1.7 expression was high in TREK2+ neurons (e.g., 1, 2, and 4) but was not limited to these neurons (e.g., 3). After axotomy, compared with normal, Nav1.7 immunostaining was substantially reduced in small TREK2+ neurons (4–6); TREK2 is retracted from the neuron edge; 1 is a large neuron, negative for all these stains. As in B, neuronal outlines are indicated with white dashed lines. Scale bar, 20 μm.
Figure 6.
Figure 6.
A, SF rate was related to Em in C-neurons in neuropathic pain models. SF rate in intact L4 C-nociceptors was significantly linearly correlated with Em 7 d after SNA and mSNA surgery (see Materials and Methods and Results). B, C, SFL 1 d after CFA was related to ipsilateral TREK2 levels. In ipsilateral L5 DRG small neurons in 5 rats 1 d after CFA-induced cutaneous inflammation, there were significant linear correlations of (B) SFLd and (C) number of SFLn with median ipsilateral L5 small DRG neuron TREK2 pixel densities for each rat, whether in entire cytoplasm, or edge, mid, or inner cytoplasmic regions. Thus, SFL after CFA in individual rats was related to ipsilateral TREK2 levels in small DRG neurons.
Figure 7.
Figure 7.
TREK2 knockdown in vivo decreased SFL 1 d after CFA. Ipsilateral SFL in rats 1 d after an intradermal plantar CFA injection in one hind foot in the same region as an intradermal plantar injection of either TREK2 siRNA or scrambled siRNA given 3 d earlier. Numbers in B and C (superscripts) and D and E (beside data points) relate to rats from which tissue in the lanes of A was obtained; these numbers are included to enable comparison of TREK2 protein expression with other data from the same rats. A, TREK2 Western blot of skin tissue, at the site of scr (1, 2) or siRNA (3, 4) injection 4 d previously, and 1 d after either saline (1, 3) or CFA injection (2, 4) at the same site. Top, The lower TREK2 band (MW ∼50 kDa, the band that we and others find in DRG tissues) was almost eliminated by siRNA, not scr, treatment, regardless of whether rats were injected with CFA or saline. The upper band was not reduced after siRNA. The MWs of these bands are consistent with them being two of the three known rat TREK2 isoforms (see Results and Discussion). Bottom, α-Tubulin loading controls. B, TREK2 knockdown by TREK2 siRNA in nerve fibers and epidermis (Ep) in plantar skin. Left, after scr siRNA injection. Right, after TREK2 siRNA injection. Both treatments were followed by CFA injection. Tissue was taken 4 d after scr/siRNA treatment (1 d after CFA). Triple fluorescence staining for β-tubulin III (blue, to stain nerve fibers), IB4 binding (green), and TREK2 (red). White dashed lines indicate junctions between epidermis (above) and dermis (below). In the merged images, colocalization of all three markers shows as white or very pale (bottom left image). After siRNA, red TREK2 staining in epidermis and in groups of nerve fibers (shown as blue and/or green) was reduced, showing substantial knockdown in vivo of TREK2 near the siRNA injection site. Scale bar, 25 μm. C, TREK2 knockdown by local TREK2 siRNA in the nerve innervating plantar skin. This knockdown, seen 5–10 mm from the siRNA injection site, occurred whether rats were subsequently treated with saline or CFA. There was decreased fiber TREK2 immunostaining (red) in TREK2 siRNA, not scr siRNA, treated rats, so that IB4+ fibers (green) showed little/no TREK2 (red). Scale bar, 50 μm. D, E, SFL after CFA increased after TREK2 siRNA compared with scr control. Ipsilateral SFLd (D) or SFLn (E) after siRNA C compared with scr treatment did not increase before saline injection (day 3, columns 1 and 2) or after saline injection (day 4, columns 3 and 4), or before CFA injection (day 3, columns 5 and 6). However, after CFA, it caused increased SFLd (D) and SFLn (E) in both scr (p < 0.01, columns 7 vs 5) and TREK2 siRNA (p < 0.001, column 8 vs 6) rats. This increase was ∼2-fold greater (for both SFLd and SFLn) after siRNA than after scr treatment (p < 0.05, column 8 vs 7). Statistics refer to both duration and events throughout. That is, the effect on SFL (duration and number of footlifts) of TREK2 knockdown was evident only after CFA treatment, when it doubled the CFA-induced SFL. *p < 0.05.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Acosta C, Davies A. Bacterial lipopolysaccharide regulates nociceptin expression in sensory neurons. J Neurosci Res. 2008;86:1077–1086. doi: 10.1002/jnr.21565. - DOI - PubMed
    1. Acosta C, McMullan S, Djouhri L, Gao L, Watkins R, Berry C, Dempsey K, Lawson SN. HCN1 and HCN2 in rat DRG neurons: levels in nociceptors and non nociceptors, NT3 dependence and influence of CFA-induced skin inflammation on HCN2 and NT3 expression. PLoS One. 2012;7:e50442. doi: 10.1371/journal.pone.0050442. - DOI - PMC - PubMed
    1. Akopian AN, Sivilotti L, Wood JN. A tetrodotoxin-resistant voltage-gated sodium channel expressed by sensory neurons. Nature. 1996;379:257–262. doi: 10.1038/379257a0. - DOI - PubMed
    1. Alvarez P, Gear RW, Green PG, Levine JD. IB4-saporin attenuates acute and eliminates chronic muscle pain in the rat. Exp Neurol. 2012;233:859–865. doi: 10.1016/j.expneurol.2011.12.019. - DOI - PMC - PubMed
    1. Alvarez-Leefmans FJ, Gamiño SM, Giraldez F, Noguerón I. Intracellular chloride regulation in amphibian dorsal root ganglion neurones studied with ion-selective microelectrodes. J Physiol. 1988;406:225–246. - PMC - PubMed

Publication types

Substances