Dynamic expression of the TRPM subgroup of ion channels in developing mouse sensory neurons

Gene Expression Patterns 10 (2010) 65–74

Contents lists available at ScienceDirect

Gene Expression Patterns journal homepage: www.elsevier.com/locate/gep

Dynamic expression of the TRPM subgroup of ion channels in developing mouse sensory neurons Susanne Staaf a,b, Marina C.M. Franck b, Frédéric Marmigère b, Jan P. Mattsson c, Patrik Ernfors b,* a

Department of Bioscience, AstraZeneca R&D Mölndal, 431 83 Mölndal, Sweden Division of Molecular Neurobiology, Department of Medical Biochemistry and Biophysics, Karolinska Institutet, 171 77 Stockholm, Sweden c Albireo Pharmaceuticals, 430 33 Göteborg, Sweden b

a r t i c l e

i n f o

Article history: Received 2 February 2009 Received in revised form 10 September 2009 Accepted 12 October 2009 Available online 20 October 2009 Keywords: Nociceptive Non-nociceptive Peptidergic TRP channels Expression mRNA Neuron In situ hybridization Dorsal root ganglion Nodose ganglion Sensory nervous system TRPM8 Immunohistochemistry PCR

a b s t r a c t Despite the significance of transient receptor potential (TRP) channels in sensory physiology, little is known of the expression and developmental regulation of the TRPM (melastatin) subgroup in sensory neurons. In order to find out if the eight TRPM subgroup members (TRPM1–TRPM8) have a possible role in the sensory nervous system, we characterized the developmental regulation of their expression in mouse dorsal root ganglion (DRG) from embryonic (E) day 12 to adulthood. Transcripts for all channels except for TRPM1 were detected in lumbar and thoracic DRG and in nodose ganglion (NG) with distinguishable expression patterns from E12 until adult. For most channels, the expression increased from E14 to adult with the exception of TRPM5, which displayed transient high levels during embryonic and early postnatal stages. Cellular localization of TRPM8 mRNA was found only in a limited subset of very small diameter neurons distinct in size from other populations. These neurons did not bind isolectin B4 (IB4) and expressed neither the neuropeptide calcitonin gene-related peptide (CGRP) nor neurofilament (NF)200. This suggests that TRPM8+ thermoreceptive sensory neurons fall into a separate group of very small sized neurons distinct from peptidergic and IB4+ subtypes of sensory neurons. Our results, showing the expression and dynamic regulation of TRPM channels during development, indicate that many TRPM subfamily members could participate during nervous system development and in the adult by determining distinct physiological properties of sensory neurons. Ó 2009 Elsevier B.V. All rights reserved.

1. Results and discussion There are several types of sensory neurons in the DRG with responsiveness to different kinds of external and internal stimuli. These stimuli, i.e. nociceptive, thermal or mechanical, activate different receptors and ion channels that are present in the nerve terminals at the sensory receptive fields and their expression in selective subsets of DRG neurons determines the response profile of individual neurons to a given stimuli. TRP channels are a group of non-selective cation channels that play important functions in sensory neurons. These channels are divided into six subgroups called TRPM (melastatin), TRPC (classical), TRPV (vanilloid), TRPP (polycystin), TRPML (mucolipin) and TRPA (ankyrin). Despite the significance of TRP channels for sensory functions little is known about their expression in the DRG and in particular, about the

* Corresponding author. Address: Division of Molecular Neurobiology, Department of Medical Biochemistry and Biophysics, Karolinska Institutet, Scheelsv. 1, 171 77 Stockholm, Sweden. Tel.: +46 8 5248 7659; fax: +46 8 341960. E-mail address: [email protected] (P. Ernfors). 1567-133X/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.gep.2009.10.003

expression and regulation of the TRPM subfamily. TRPM1–M8 is functionally a very diverse group of ion channels with diverse functions and expression patterns (Fleig and Penner, 2004; Fonfria et al., 2006). TRPM8 is the only TRPM channel with a clearly assigned function in DRG neurons. It is activated by innocuous cool stimuli and responds to menthol and icilin with intracellular Ca2+ elevations (McKemy et al., 2002; Peier et al., 2002). Placing TRPM8 in any particular category of small size sensory neurons has been enigmatic and controversial. Originally, a lack of colocalization with known neuronal markers such as NF200, CGRP, IB4 and TRPV1 was reported (Peier et al., 2002). These findings were challenged by other groups showing coexistence of TRPM8 with TRPV1 (Okazawa et al., 2004) and NF200 (Kobayashi et al., 2005). In addition, two studies using tracers for TRPM8 in transgenic mice presented conflicting results where one showed presence of both CGRP and NF200 in a subpopulation of TRPM8+ neurons (Takashima et al., 2007) while the other reported lack of colocalization for TRPM8 with NF150, IB4 and CGRP (Dhaka et al., 2008). To give some insight on which of the TRPM subfamily channels may have importance in the sensory nervous system we analyzed

66

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74

the expression of the TRPM channels through embryonic and postnatal development in mouse thoracic and lumbar dorsal root ganglia as well as in the nodose ganglion. Except for TRPM1, we found transcripts of all channels with different expression patterns between E12 and adulthood. TRPM8 expression was investigated in more detail using in situ hybridization in combination with immunohistochemistry. We found by quantitative analysis that only a limited subset of very small diameter neurons, which did not label for IB4, CGRP or NF200, contained TRPM8. Consistently, size frequency histogram analysis placed the very small caliber TRPM8+ neurons in a separate category of neurons distinct from IB4+ and NF200+ neurons. 1.1. Developmental expression of TRPM channels in lumbar dorsal root ganglia Real-time PCR analysis using cDNA from lumbar DRG at E12, E14, E18, P0, P4, P12 and adult ages was performed to study the expression pattern of TRPM1–TRPM8 throughout development. By adding an equal amount of template in each assay, using an internal reference gene (mouse ribosomal protein 36B4) and normalizing against a common calibrator gene (adult lumbar TRPV1) that was set to 1, the relative expression levels could be compared between the different channels. All reactions produced one peak in the melting point curves. In adult tissue the mRNA expression levels were highest for TRPM3, 20% of TRPV1 mRNA levels, followed by TRPM2, TRPM4 and TRPM7 that expressed 10% of TRPV1 mRNA levels (Fig. 1). TRPM8 was expressed in lower amounts (3% of TRPV1 mRNA levels) and TRPM5 and TRPM6 levels were very low (<1% of TRPV1 mRNA levels). TRPM1 was not detected at all, neither in adult nor at any developmental stage analyzed. At embryonic stages, all channels showed increased expression from E12 until soon before or around birth. From that time point to P4 the expression decreased for all channels except for TRPM6, for which the mRNA levels remained unchanged. At the next stage analyzed, P12, the expression increased for all channels except for TRPM5, for which mRNA levels continued to decrease until adulthood at which time very little TRPM5 was detected. For TRPM2, M3, M4 and M6 the expression levels continued to increase between P12 and adulthood while TRPM7 and TRPM8 showed a small reduction in levels. This pattern suggests two expression waves. The second wave occurring after birth implies an importance of these channels in adult life either direct in sensory mechanisms or in more general ion homeostasis of the sensory neurons. 1.2. TRPM8 is expressed in neither peptidergic nor IB4+ small size neurons in mouse dorsal root ganglia To define the population of neurons in mouse lumbar DRG expressing TRPM8, colocalization studies were performed with common markers for neuronal subtypes. In situ hybridization with a digoxigenin labeled TRPM8 RNA probe showed very strong expression of the channel mRNA in a small subset of the neurons. Total number of neurons was determined by the nuclear neuronal marker Islet 1. Labeling was seen in 6.0% of all adult DRG neurons (Fig. 2), a finding that corresponds well with other studies reporting positive labeling in 5–10% of the neurons (Dhaka et al., 2008; Peier et al., 2002). To investigate if the decreased expression seen at P4 by quantitative PCR was caused by a loss of cells expressing TRPM8 or by lower expression levels in each cell, in situ hybridization was conducted on E18, P4 and P12 DRG (Figs. 3 and 4). The TRPM8 probe labeled 2.4%, 3.8%, 5.0% and 6.0% of the neurons in E18, P4, P12 and adult DRG, respectively (Table 2). Statistical analysis showed significantly less cells labeled in E18 than in P4 and in P4 than in P12 or adult, whereas there was no significant difference

between P12 and adult values. Hence, overall, the number of TRPM8+ neurons was found to steadily increase from E18 to adult stages. These findings suggest that increased expression levels in the DRG postnatally reflects an increase in the number of TRPM8 expressing neurons rather than marked alteration in expression levels within neurons already expressing TRPM8. However, our data suggest that the decline in mRNA levels detected by real-time PCR between E18 and P4 reflect reduced transcript levels in individual cells. To further investigate which population the TRPM8 labeled subset of cells belong to, immunohistochemical labelings with IB4, binding to small non-peptidergic neurons, NF200, which labels large proprioceptive and mechanoreceptive neurons, and CGRP, expressed by peptidergic nociceptive neurons, were performed on the same sections (Fig. 2). The neuronal nuclear marker Islet 1 was used to count the total number of neurons. IB4 stained 39%, NF200 was expressed by 38% and CGRP occurred in 40% of the total population of neurons in adult DRG. The same markers were also employed on P4 and P12 tissue (Table 2, Figs. 3 and 4). At these earlier stages, labeling frequencies for IB4 were only 16% at P4 and 33% at P12. Neuronal numbers expressing CGRP were also found to increase from 24% at P4 to 32% at P12, whereas similar numbers of NF200 expressing cells were counted at postnatal stages (27% at P4 and 29% at P12). TRPM8+ cells did not stain for IB4, NF200 or CGRP at any stage examined. The lack of coexpression in adults with either marker was consistent with the findings of Peier et al. (2002), where in situ hybridization combined with immunohistochemistry indicated no overlap between TRPM8 and NF200, TRPV1, CGRP or IB4 staining. Another study showed coexistence of TRPM8 with TRPV1 in the same cells in DRG sections, i.e. 29% of TRPM8+ cells also expressed TRPV1 (Okazawa et al., 2004). Yet another report presented a 23% TRPM8 labeling frequency and that 50% of the TRPM8+ cells colocalized with NF200 in rat DRG neurons (Kobayashi et al., 2005). Interestingly, contradicting the results of Okazawa et al. (2004), but confirming the findings of Peier et al. (2002), this study showed rare colocalization between TRPV1 and TRPM8. It might be of importance that both Kobayashi et al. (2005) and Okazawa et al. (2004) were using rat tissue which might stain differently for neuronal markers and TRP channels than mice do. Using a transgenic mouse expressing EGFP in the TRPM8 locus, TRPM8 expression was found in 7.8% of DRG neurons and also showed some colocalization with TRPV1 (Dhaka et al., 2008). Nerve endings from the TRPM8 population were localized in the outermost layers of epidermis close to, but not overlapping with CGRP+ fibers. These terminals correspond to the non-peptidergic/IB4-negative sensory free nerve endings previously identified in the epidermis in the intervibrissal fur on the mystacial pad of the rat (Fundin et al., 1997). Another study with a similar approach, using GFP as a tracer for TRPM8 neurons, showed a quite extensive overlap between GFP and NF200, CGRP and TRPV1 (Takashima et al., 2007). The lack of co-labeling of TRPM8 with IB4, NF200 or CGRP at any analyzed stage in the present study indicates that new TRPM8 neurons might not be recruited from any of these subtypes, in which case a transient co-expression at early developmental stages might be expected followed by a subsequent loss of the IB4, CGRP and NF200 identity. Differentiation of sensory neurons into specific subtypes is controlled by a dynamic interaction between extracellular signaling and control of gene expression exerted by neurotrophic growth factors and transcription factors (Marmigere and Ernfors, 2007). During embryonic development around 80% of all neurons express TrkA. The non-peptidergic population of neurons is generated from roughly half of the TrkA+ population that under influence of the runt homology domain transcription factor Runx1 starts to express Ret and extinguishes TrkA postnatally. These neurons colabel to 90% with IB4 and are dependent on GDNF for survival (Molliver

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74

67

Fig. 1. (A–H) Developmental expression of transient receptor potential channels TRPM1–TRPM8 in lumbar dorsal root ganglia (DRG) measured by real-time PCR. Each data point is calibrated against TRPV1 expression in adult lumbar DRG using the 2DDCT method. Error bars show standard deviation (n = 3). All transcripts but TRPM1 were detected. TRPM3 was the most abundant in adult tissue followed by TRPM4, TRPM2, TRPM7 and TRPM8 having lower amounts and finally TRPM5 and TRPM6 expressing very low levels.

and Snider, 1997; Molliver et al., 1997). Runx1 and TrkA has a mutually exclusive expression pattern in adults (Chen et al., 2006). Expression of TRPM8 is critically dependent on Runx1 since TRPM8 expression fails in Runx1/ mice (Chen et al., 2006). These results contradict the conclusions of Peier et al. (2002) suggesting that TRPM8 is expressed in TrkA+ neurons. The findings that expression of TRPM8 was abolished in TrkA/ mice seems to be a consequence of extinction earlier in development at a time before TrkA is down-regulated in this population. Interestingly, expression of TRPM8 is not affected in Ret/ mice indicating its indepen-

dence of Ret signaling. In contrast, a number of other genes regulated by Runx1 have been suggested to be under the control of Runx1-mediated Ret expression and subsequent signaling by Ret (i.e. TrpA1, MrgA1, MrgA3 and MrgB4), since their expression fails in both Runx1 and Ret null mutant mice (Chen et al., 2006; Luo et al., 2007). Our findings that IB4 staining steadily increased from P4 to adult stage correlate well with earlier observations, again showing that many non-peptidergic neurons are established after birth and that reshaping of neurons continues several weeks postnatally. The

68

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74

Fig. 2. Expression of TRPM8 in adult lumbar DRG. In situ hybridization with digoxigenin labeled RNA probe for TRPM8 (A, D and G) and immunohistochemical labeling of isolectin B4 (IB4) (B), neurofilament 200 (NF200) (E) and calcitonin gene-related peptide (CGRP) (H). Inverted in situ hybridization images merged with immunohistochemical photos (C, F and I) shows lack of colocalization of TRPM8 with any of the markers. Arrows point at TRPM8 cells not positive for IB4 (A–C), NF200 (D–F) or CGRP (G–I). Scale bar: 50 lm.

IB4-negative neurons with persistent TrkA expression develop into peptidergic neurons and are known to express CGRP in the adult (Averill et al., 1995). In accordance with this, we found that 40% of the total number of neurons was CGRP+ and many emerged after birth, since there was as steady increase in expression from P4 and onwards. Size is a classical way of classifying neuronal subtypes (Goldstein et al., 1991; Lawson, 1979; Molliver et al., 1995). In the absence of established markers for the small diameter cell type identified here, size frequency distribution was used to further characterize this population (Fig. 5). The strong correlation between different cell populations and cell area was clear in this study where large and medium sized cells were labeled with NF200 measuring from 400 to 1000 lm2 and the IB4+ cells were smaller with an area of typically 200–500 lm2. The size frequency histogram further enforces that the TRPM8+ neurons falls into an

exclusive population of particularly small diameter cells, mostly with areas comprised between 100 and 300 lm2. 1.3. Developmental expression of TRPM channels in thoracic dorsal root ganglia and nodose ganglia Nerves from the lumbar DRG project primarily to the lower limbs and skin whereas the thoracic ganglia innervate the gut more extensively. These different specificities suggest different neuronal subtypes to be present in varying ratios in the thoracic and lumbar ganglia. It is thus likely that the expression levels of TRPM mRNAs also differ if the channels are restricted to certain neuronal subtypes. This hypothesis was addressed by quantitative measurement of mRNA in thoracic DRG. In addition to this, mRNA levels were measured in NG to elucidate the possible involvement of TRPM channels in the vagal sensory system. In thoracic DRG

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74

69

Fig. 3. Expression of TRPM8 in P4 lumbar DRG. In situ hybridization with digoxigenin labeled RNA probe for TRPM8 (A, D and G) and immunohistochemical labeling of IB4 (B), NF200 (E) and CGRP (H). TRPM8 did not colabel with IB4, NF200 or CGRP illustrated by merged images (C, F and I). Arrows elucidate this lack of coexpression for IB4 (A–C), NF200 (D–F) and CGRP (G–I). Scale bar: 50 lm.

(squares in Fig. 6) the expression patterns were similar to the ones previously seen in lumbar DRG. The only consistent difference was that the levels were significantly higher in lumbar than in thoracic tissue in the adult for all channels apart from TRPM5, for which levels were very low in both tissues. All channels showed similar expression patterns from E12 and forward differing only at the adult stage for TRPM2, TRPM3 and TRPM4. Here, expression dropped in thoracic DRG opposed to continued increase in lumbar DRG. Similar to lumbar DRG, transcripts for TRPM1 could not be detected at any developmental stage in thoracic DRG. To further investigate the origin of the reduced mRNA levels in thoracic compared to lumbar DRG in situ hybridization of TRPM8 was performed on adult thoracic DRG sections. Results showed that 0.79 ± 0.18% of the total number of neurons were TRPM8+. This is significantly lower than in adult lumbar DRG where 6.0 ± 0.58% cells were TRPM8+ (P < 0.001). Hence, the lower transcript levels shown in the real-time PCR for thoracic DRG are caused by fewer

TRPM8 expressing cells rather than lower expression levels in each cell. TRPM8 axons originating from lumbar DRG are known to innervate skin to mediate cool sensations in the external environment. Also the gastrointestinal tract, to which thoracic afferents project, has been shown to display thermosensitivity (Villanova et al., 1997). In addition, cold stimuli applied to the stomach or rectum has been shown to lower the threshold for visceral perception in patients with irritable bowel syndrome (Li et al., 2004; Zuo et al., 2006). However, the mechanism for detection of cold in the gut has not been elucidated. Our results showing presence of the cold receptor TRPM8 in thoracic DRG suggests that the visceral thermoreception might be, at least partly, attributed TRPM8. The fact that very few neurons expressed TRPM8 could reflect a less substantial cold sensitivity of visceral afferents compared to skin afferents. Nodose ganglion has vagal afferents to various organs including the gastrointestinal tract, lungs and heart. Similar to DRG function, specialized cells convey sensory information from the target

70

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74

Fig. 4. Expression of TRPM8 in P12 lumbar DRG. In situ hybridization with digoxigenin labeled RNA probe for TRPM8 (A, D and G) and immunohistochemical labeling of IB4 (B), NF200 (E) and CGRP (H). Again, as shown for P4 and adult tissue, no overlap occurred between TRPM8 and IB4, NF200 or CGRP (C, F and I). Lack of costaining for individual cells is further illustrated with arrows. Scale bar: 50 lm.

Table 1 Primers used for real-time PCR. Gene

Accession Nos.

Forward primer

Reverse primer

TRPM1 TRPM2 TRPM3 TRPM4 TRPM5 TRPM6 TRPM7 TRPM8 TRPV1 36B4

AF047714 NM_138301 NM_177341.4 NM_175130 NM_020277 NM_153417 NM_021450 NM_134252 NM_001001445 X15267

50 -TGCGACGAAGGAGGAGTCAT-30 50 -CTTCCCCGTGCCTAACGA-30 50 -ATGGAGTAAGCATGCACCGTTT-30 50 -GCTCATCGCCATGTTCAGCTA-30 50 -ACCTGGCCACTGAAGCTGAT-30 50 -TCAAAAGACCCTCACAGATGCA-30 50 -TGCTGCTGGATATAGTGAATGTTGT-30 50 -AGGGAGGACTTGGATGTGGAA-30 50 -GATGACTTCCGGT GGTGCTT-30 50 -GAGGAATCAGATGAGGATATGGGA-30

50 -GCTGTAATTAAATGTTTTCTGAATGGTAAC-30 50 -CAGCGGTGTAAAAGGGAGGAT-30 50 -TGAGGGCCCATGTCTCGTAT-30 50 -GCGCTGTGCCTTCCAGTAG-30 50 -CCACCAGATCTTGGTCAGGAAT-30 50 -CGGCCACAGTAACACCTGACT-30 50 -GTCGACGCTCTTCTACTACAGCTTT-30 50 -CCCAGATGAAGAGAGCTTGCA-30 50 -GCCCACGTTGGTGTTCCA-30 50 -AAGCAGGCTGACTTGGTTGC-30

tissues to the CNS (Berthoud and Neuhuber, 2000). NG is placodederived in contrast to DRG which arises from the neural crest (Zhuo et al., 1997). To see if levels and developmental expression

patterns of TRPM channels were a consequence of differences in innervation or cellular origin, quantitative PCR analysis was conducted in material from NG (triangles in Fig. 6). TRPM2, M3, M4,

71

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74 Table 2 Quantification of neuronal subtypes. Age

P4 P12 Adult Adult

Tissue

DRG (lumbar) DRG (lumbar) DRG (lumbar) NG

% of neurons TrpM8+

IB4+

NF200+

CGRP+/TrkA+ ***

3.8 ± 0.44** (n = 25) 5.0 ± 0.37* (n = 22) 6.0 ± 0.58 (n = 32) 0.51 ± 0.16 (n = 23)

16 ± 1.9 (n = 8) 33 ± 2.3** (n = 12) 39 ± 2.1 (n = 10) 33 ± 3.7 (n = 8)

28 ± 3.1 (n = 8) 29 ± 1.9 (n = 8) 38 ± 2.1** (n = 10) 8.2 ± 2.1 (n = 4)

24 ± 1.4 (n = 6) 32 ± 1.8** (n = 8) 40 ± 2.0* (n = 11) 9.4 ± 1.1 (n = 5)

Data is presented as means ± SEM, n = number of sections counted. Total number of neurons was determined by Islet 1 immunohistochemistry on top of the in situ hybridized sections. *Represents P < 0.05 and **represents P < 0.01. Significance levels show comparisons to previous time points. TrpM8 at P4 compares against E18 which displayed 2.4% ± 0.25 expressing cells. ***Numbers are for CGRP in DRG, TrkA in NG.

Fig. 5. Size frequency histogram showing the area of cells staining for TRPM8, IB4 and NF200, respectively. TRPM8 cells falls into a separate group of very small neurons comparing to IB4 cells, which are small and medium sized, and NF200 cells being medium and large sized neurons.

M6 and M8 transcript levels were very low from E12 to E18 whereafter a marked increase was seen to P0. TRPM7 differed to this pattern by having a distinct decrease from E12 to E18 whereafter the expression increased to P0. For TRPM2, M3, M4 and M7 channels there were small decreases or unchanged mRNA levels between P0 and P4 followed by increasing levels at P12 and adult. This early postnatal reduction was not seen for TRPM6 and M8 for which the levels increased until adulthood. Expression levels of TRPM5 were very similar to its expression in both lumbar and thoracic DRG, with the only major difference that it peaked at E14 instead of E18 before declining to very low levels at E18. TRPM5 increased transiently at P0 whereafter it declined to very low levels in the adult. TRPM1 was not detected in material from any stage. Hence, TRPM2, M3, M4, M6 and M8 had a later onset in NG than in DRG, with increased expression from E18 and forward, as compared to the DRG where the channels were detected already at E12–E14. Apart from this, the expression levels and regulation of expression were similar between DRG and NG. In a previous study we also observed a resembling regulation of TRPC channel expression in placode and neural crest derived tissues (Elg et al., 2007). TRPM8 has previously been identified in NG neurons projecting to the gut and cool stimuli elicited a Ca2+ response in isolated NG neurons (Zhang et al., 2004). To get further insight if the differences in innervation and function between NG and DRG were reflected in the cellular subtype distribution of TRPM8 colocalization studies were performed using in situ hybridization and immunohistochemistry. TRPM8 was expressed by 0.51% of NG neurons and localized to very small size neurons. This shows that the NG contains more than 10 times lower numbers of TRPM8+ neurons than the lumbar DRG (Table 2). The relative abundance of mRNA in the NG as detected by real-time PCR suggests that the mRNA levels in each cell were substantially higher in NG than in lumbar DRG, but also that fewer afferents might be sensi-

tive to cold stimuli. The NG innervate a range of different organs and only a few of the target organs may be thermosensitive, which might be reflected by the low number of TRPM8+ cells in the NG as compared to the DRG. Furthermore, it cannot be excluded that that other cold receptors are also present on vagal afferent neurons. Similarly to what was found in DRG, no colocalization was identified between TRPM8 and IB4 (0 of 310 cells), NF200 (0 of 57 cells) or TrkA (0 of 70 cells) in NG, indicating that the TRPM8+ cells in NG represent a unique small size neuronal population. 1.4. Conclusions The overall increasing expression levels of TRPM2, M3, M4, M6, M7 and M8 in sensory ganglia throughout development open for undiscovered sensory functions for these channels. All channels but TRPM1 had detectable amounts of mRNA in varying degrees throughout development in both DRG and NG. The TRPM8+ population, which has a known sensory function in thermosensation, was further characterized, both quantitatively and qualitatively using in situ hybridization and immunohistochemistry in P4, P12, adult lumbar and thoracic DRG and in adult NG. Results showed expression in a limited subpopulation of neurons of both DRG and NG that did not colabel with IB4, NF200 or CGRP/TrkA at any of the developmental stages tested. Size frequency histograms clearly pointed out the TRPM8 population as a unique collection of neurons with very small cell soma area with the majority of cells measuring between 100 and 300 lm2 in the DRG. 2. Experimental procedures All procedures involving animal handling were approved by the local Ethical Committee for Animal Experiments.

72

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74

Fig. 6. (A–H) Developmental expression of TRPM1–TRPM8 in thoracic DRG (squares) and NG (triangles) measured by real-time PCR. Each data point is calibrated against TRPV1 expression in adult lumbar DRG using the 2DDCT method. Error bars show standard deviation (n = 3). Transcript expression in thoracic DRG was similar to that observed in lumbar DRG (Fig. 1). mRNA levels in NG were also in the same range for each channel as in lumbar DRG but exhibit different patterns of regulation during embryonic development.

2.1. Real-time PCR For expression analysis of TRPM channels in embryonic and postnatal development, DRG and NG were excised from female C57/Bl6JOlaHsd (Harlan Netherlands, Horst, The Netherlands) mice at different ages and quickly frozen on dry ice. From embryos all lumbar and thoracic DRGs were used in contrast to postnatal and adult tissues where only L4–L6 and T9–T12 were dissected. RNA was extracted with Trizol (Invitrogen, Paisley, Scotland), DNase treated with DNAfree (Ambion, Cambridgeshire, UK) and used for cDNA synthesis with iScript (Bio-Rad, Hercules,

CA). Real-time PCR was performed using 20 ng cDNA per 25 ll reaction, mixed with primers (Eurogentec, Seraing, Belgium) (Table 1) and SybrGreen PCR master mix (Applied Biosystems, Foster City, CA) in a 7500 real-time PCR system (Applied Biosystems). The resulting PCR products created a single band in agarose gel electrophoresis, produced one specific signal in melting curves and their identity was confirmed with sequencing. For normalization of the PCR data the 2DDCT method was used (Livak and Schmittgen, 2001) where DDCT = ((CT (target gene)  CT (reference gene))  (CT (calibrator gene)  CT (reference gene)). In this study, target gene stands for TRPM1–TRPM8,

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74

reference gene for mouse acidic ribosomal protein 36B4 and calibrator gene for TRPV1 in adult lumbar DRG. 2.2. In situ hybridization Adult female C57/Bl6JOlaHsd mice were perfusion fixed with 4% phosphate buffered paraformaldehyde pH 7.4. DRG (L4–L6 and T11–T12) or NG were excised, postfixed for 1–2 h and incubated overnight at 4 °C in 20% sucrose. After snap freezing in TissueTek (Sakura, Zoeterwoude, The Netherlands) the ganglia were cut in 14 lm sections and stored at 80 °C until use. The procedure for P4 and P12 animals were similar except for not being perfusion fixed before dissection. For E18 tissue, the embryo was fixed in its entirety over night and cryoprotected in 20% sucrose at 4 °C for an additional 24 h before sectioning at 14 lm. For preparation of digoxigenin labeled probe, RNA from adult L4–L6 DRG was extracted with Trizol, DNase treated with DNAfree and used as template for cDNA synthesis with Superscript II reagent (Invitrogen). A 748 bp long TRPM8 fragment was amplified from the cDNA using PfuUltra DNA polymerase (Stratagene, La Jolla, CA) and primers (50 -TCTGAAGACCCTGGCCAAAGT-30 and 50 AATGTGGATGAGCCTTAGCGTG-30 ) in an iCycler (Bio-Rad). After purification on agarose gel the fragment was cloned into pCR4Blunt-TOPO vector (Invitrogen) and sequenced in a 3730 DNA Analyzer (Applied Biosystems). Digoxigenin labeled RNA was produced by incubating 50 -digested vector DNA containing the TRPM8 fragment with Dig labeling mix, RNase inhibitor, transcription buffer (Roche, Mannheim, Germany), DTT and T3 RNA polymerase (Promega, Madison, USA). Incubation occurred at 37 °C for 3 h whereafter the RNA was precipitated with LiCl and dissolved in RNase free H2O. For hybridization, the RNA probe (5 ng/ll) was mixed in hybridization solution (0.19 M NaCl, 10 mM Tris (pH 7.2), 5 mM NaH2PO42H2O/Na2HPO4 (pH 6.8), 50 mM EDTA (Merck, Darmstadt, Germany), 10% Dextran Sulphate, 1 mg/ml Yeast tRNA (Sigma–Aldrich, St. Louis, MO), 1 Denhardt’s solution and 50% Formamide (JT Baker, Deventer, The Netherlands). After denaturation for 2 min at 95 °C, 200 ll of the probe was applied to each section and incubated overnight in a humid chamber at 69 °C. The sections were washed 4  20 min in washing buffer (2 SSC (Sigma–Aldrich), 0.1% Tween 20 (Research Organics, Cleveland, MO) and 50% Formamide), followed by 3  20 min in MABT buffer (0.1 M Maleic acid, 0.15 M NaCl (Merck) and 0.1% Tween 20 (pH 7.5)). Blocking solution (2% blocking agent (Roche) and 20% normal goat serum (Jackson Immunoresearch, West Grove, USA) dissolved in MABT) was applied for 1 h followed by overnight incubation in room temperature with anti-digoxigenin antibody (Roche) (1:2000 in blocking solution). After washing 6  20 min with MABT buffer and 3  10 min with Buffer B3 (100 mM Tris, 100 mM NaCl, 50 mM MgCl2 and 0.1% Tween 20 (pH 9.5)) detection was achieved by incubation for 0.5–2 days in NBT/BCIP solution (100 mg/ml/50 mg/ml (Roche) in buffer B3). A final wash in PBS stopped the reaction and the sections were mounted in glycerol:water 9:1. 2.3. Immunohistochemistry In situ hybridized sections were washed 3  10 min with PBS, blocked with 4% normal goat serum, 1% bovine serum albumin (Sigma–Aldrich) and 0.1% Triton X-100 (Sigma–Aldrich) in PBS and incubated overnight at 4 °C with mouse-anti-NF200 clone N52 (1:200), rabbit-anti-CGRP (1:200), rabbit-anti-TrkA (1:5000) or mouse-anti-Islet-1 (1:100) diluted in blocking solution. The monoclonal NF200 antibody (Sigma–Aldrich, clone N52) was raised against the carboxyterminal tail segment of enzymatically dephosphorylated pig neurofilament H-subunit and detects both

73

phosphorylated and nonphosphorylated NF200 with broad species crossreactivity (Shaw et al., 1986). The CGRP antibody (Peninsula Laboratories, San Carlos, CA, #T-4239) was raised against a synthetic human a-CGRP peptide and crossreacts with human, chicken and rat a-CGRP as well as human and rat b-CGRP in radioimmunoassays (manufacturer’s technical information). Staining of DRG sections exhibited a pattern undistinguishable to what has been demonstrated in earlier studies (Dong et al., 2001; Peier et al., 2002). The TrkA antibody (gift from LF Reichardt, UCSF) was raised against the extracellular domain of the rat TrkA receptor (Clary et al., 1994) and crossreacts with mouse (Molliver et al., 1995). Islet-1 (Developmental Studies Hybridoma Bank, Iowa City, IA, clone 39.4D5 developed by Thomas M. Jessell) was raised against the C-terminal portion of rat Islet 1 and has broad crossreactivity (manufacturer’s technical information). The antibody labeled all neurons in the DRG in accordance with previous studies (Thor et al., 1991). After washing 3  10 min with PBS incubation occurred for 2–4 h with secondary IgG coupled to AF488 or AF555 (Invitrogen). After repeated washing with PBS the slides were mounted in Vectashield (Vector Laboratories, Burlingame, CA). 2.4. Isolectin B4 (IB4) staining Slides were washed 3  10 min in PBST (PBS with 0.2% Tween 20), incubated for 2–3 h with IB4-biotin (Invitrogen, #I21414) diluted to 2 mg/ml in PBST, washed again 3  10 min in PBST and incubated 1 h with streptavidin-AF555 (Invitrogen). Finally, washing 3  10 min in PBST was performed before mounting in Vectashield. Labeling of small size DRG neurons was detected as previously described (Molliver et al., 1997; Snider and McMahon, 1998). Sections from all procedures were evaluated in a Nikon Eclipse 90i microscope with a Nikon DS-5MC digital camera. The images were imported into Adobe Photoshop 7.0 (Adobe Systems Inc., San José, CA) where brightness and contrast were adjusted (equal to the whole image) to increase visual clarity. 2.5. Statistics Comparison of cell numbers at different developmental stages were made with two-tailed Student’s t-test, assuming equal variance and approximated normal distribution for both populations. Acknowledgement We thank Johnny S. for technical support. This work was supported by Swedish Research Council, Swedish Brain Foundation, Bertil Hållsten Foundation, Linné Foundation (DBRM), ERC advanced Grant 232675 and the Karolinska Foundation. References Averill, S., McMahon, S.B., Clary, D.O., Reichardt, L.F., Priestley, J.V., 1995. Immunocytochemical localization of trkA receptors in chemically identified subgroups of adult rat sensory neurons. Eur. J. Neurosci. 7, 1484–1494. Berthoud, H.R., Neuhuber, W.L., 2000. Functional and chemical anatomy of the afferent vagal system. Auton. Neurosci. 85, 1–17. Chen, C.-L., Broom, D.C., Liu, Y., de Nooij, J.C., Li, Z., Cen, C., Samad, O.A., Jessell, T.M., Woolf, C.J., Ma, Q., 2006. Runx1 determines nociceptive sensory neuron phenotype and is required for thermal and neuropathic pain. Neuron 49, 365–377. Clary, D.O., Weskamp, G., Austin, L.R., Reichardt, L.F., 1994. TrkA cross-linking mimics neuronal responses to nerve growth factor. Mol. Biol. Cell 5, 549–563. Dhaka, A., Earley, T.J., Watson, J., Patapoutian, A., 2008. Visualizing cold spots: TRPM8-expressing sensory neurons and their projections. J. Neurosci. 28, 566– 575. Dong, X., Han, S.-k., Zylka, M.J., Simon, M.I., Anderson, D.J., 2001. A diverse family of GPCRs expressed in specific subsets of nociceptive sensory neurons. Cell 106, 619–632.

74

S. Staaf et al. / Gene Expression Patterns 10 (2010) 65–74

Elg, S., Marmigere, F., Mattsson, J.P., Ernfors, P., 2007. Cellular subtype distribution and developmental regulation of TRPC channel members in the mouse dorsal root ganglion. J. Comp. Neurol. 503, 35–46. Fleig, A., Penner, R., 2004. The TRPM ion channel subfamily: molecular, biophysical and functional features. Trends Pharmacol. Sci. 25, 633–639. Fonfria, E., Murdock, P.R., Cusdin, F.S., Benham, C.D., Kelsell, R.E., McNulty, S., 2006. Tissue distribution drofiles of the human TRPM cation channel family. J. Recept. Signal Transduct. Res. 26, 159–178. Fundin, B.T., Arvidsson, J., Aldskogius, H., Johansson, O., Rice, S.N., Rice, F.L., 1997. Comprehensive immunofluorescence and lectin binding analysis of intervibrissal fur innervation in the mystacial pad of the rat. J. Comp. Neurol. 385, 185–206. Goldstein, M.E., House, S.B., Gainer, H., 1991. NF-L and peripherin immunoreactivities define distinct classes of rat sensory ganglion cells. J. Neurosci. Res. 30, 92–104. Kobayashi, K., Fukuoka, T., Obata, K., Yamanaka, H., Dai, Y., Tokunaga, A., Noguchi, K., 2005. Distinct expression of TRPM8, TRPA1, and TRPV1 mRNAs in rat primary afferent neurons with adelta/c-fibers and colocalization with trk receptors. J. Comp. Neurol. 493, 596–606. Lawson, S.N., 1979. The postnatal development of large light and small dark neurons in mouse dorsal root ganglia: a statistical analysis of cell numbers and size. J. Neurocytol. 8, 275–294. Li, Y.Q., Wang, Y.M., Zuo, X.L., Guo, Y.T., Zhang, H.Y., Lu, X.F., Li, J.M., Desmond, P.V., 2004. Visceral perception thresholds after rectal thermal and pressure stimuli in irritable bowel syndrome patients. J. Gastroenterol. Hepatol. 19, 187–191. Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2-[Delta][Delta]CT method. Methods 25, 402–408. Luo, W., Wickramasinghe, S.R., Savitt, J.M., Griffin, J.W., Dawson, T.M., Ginty, D.D., 2007. A hierarchical NGF signaling cascade controls Ret-dependent and Retindependent events during development of nonpeptidergic DRG neurons. Neuron 54, 739–754. Marmigere, F., Ernfors, P., 2007. Specification and connectivity of neuronal subtypes in the sensory lineage. Nat. Rev. Neurosci. 8, 114–127. McKemy, D.D., Neuhausser, W.M., Julius, D., 2002. Identification of a cold receptor reveals a general role for TRP channels in thermosensation. Nature 416, 52–58. Molliver, D.C., Radeke, M.J., Feinstein, S.C., Snider, W.D., 1995. Presence or absence of TrKA protein distinguishes subsets of small sensory neurons with unique

cytochemical characteristics and dorsal horn projections. J. Comp. Neurol. 361, 404–416. Molliver, D.C., Snider, W.D., 1997. Nerve growth factor receptor trkA is downregulated during postnatal development by a subset of dorsal root ganglion neurons. J. Comp. Neurol. 381, 428–438. Molliver, D.C., Wright, D.E., Leitner, M.L., Parsadanian, A.S., Doster, K., Wen, D., Yan, Q., Snider, W.D., 1997. IB4-binding DRG neurons switch from NGF to GDNF dependence in early postnatal life. Neuron 19, 849–861. Okazawa, M., Inoue, W., Hori, A., Hosokawa, H., Matsumura, K., Kobayashi, S., 2004. Noxious heat receptors present in cold-sensory cells in rats. Neurosci. Lett. 359, 33–36. Peier, A.M., Moqrich, A., Hergarden, A.C., Reeve, A.J., Andersson, D.A., Story, G.M., Earley, T.J., Dragoni, I., McIntyre, P., Bevan, S., et al., 2002. A TRP channel that senses cold stimuli and menthol. Cell 108, 705–715. Shaw, G., Osborn, M., Weber, K., 1986. Reactivity of a panel of neurofilament antibodies on phosphorylated and dephosphorylated neurofilaments. Eur. J. Cell Biol. 42, 1–9. Snider, W.D., McMahon, S.B., 1998. Tackling pain at the source. New ideas about nociceptors. Neuron 20, 629–632. Takashima, Y., Daniels, R.L., Knowlton, W., Teng, J., Liman, E.R., McKemy, D.D., 2007. Diversity in the neural circuitry of cold sensing revealed by genetic axonal labeling of transient receptor potential melastatin 8 neurons. J. Neurosci. 27, 14147–14157. Thor, S., Ericson, J., Brännström, T., Edlund, T., 1991. The homeodomain LIM Protein Isl-1 is expressed in subsets of neurons and endocrine cells in the adult rat. Neuron 7, 881–889. Villanova, N., Azpiroz, F., Malagelada, J.-R., 1997. Perception and gut reflexes induced by stimulation of gastrointestinal thermoreceptors in humans. J. Physiol. 502, 215–222. Zhang, L., Jones, S., Brody, K., Costa, M., Brookes, S.J.H., 2004. Thermosensitive transient receptor potential channels in vagal afferent neurons of the mouse. Am. J. Physiol. Gastrointest. Liver Physiol. 286, G983–G991. doi:10.1152/ ajpgi.00441.2003. Zhuo, H., Ichikawa, H., Helke, C.J., 1997. Neurochemistry of the nodose ganglion. Prog. Neurobiol. 52, 79–107. Zuo, X.L., Li, Y.Q., Shi, L., Guo Ping, G., Kuang, R.G., Lu, X.F., Li, J.M., Desmond, P.V., 2006. Visceral hypersensitivity following cold water intake in subjects with irritable bowel syndrome. J. Gastroenterol. 41, 311–317.