Antigen retrieval pre-treatment causes a different expression pattern of Cav3.2 in rat and mouse spinal dorsal horn.

Cav3 channels consist of three isoforms, Cav3.1 (α1G), Cav3.2 (α1H), and Cav3.3 (α1I), which produce low-threshold spikes that trigger burst firings in nociceptive neurons of the spinal dorsal horn (SDH) and dorsal root ganglion (DRG). Although Cav3.2 plays a crucial role in pathological pain, its distribution in SDH still remains controversial. One study showed that Cav3.2 is ubiquitously expressed in neurons, but another study implied that Cav3.2 is expressed restricted to astrocytes. To unravel these discrepancies, we used methods of immunohistochemistry either with or without antigen retrieval (AR) pre-treatment to detect Cav3 in SDH and DRG from both rats and mice. Moreover, Cav3.2 mRNA was detected in mice SDH using in situ hybridization. We found that the expression pattern of Cav3.2 but not Cav3.1 and Cav3.3 in SDH were largely different with or without AR pre-treatment, which showed a neuron-like and an astrocyte-like appearance, respectively. Double staining further demonstrated that Cav3.2 was mainly co-stained with the neuronal marker NeuN in the presence of AR but was with glial fibrillary acidic protein (GFAP, marker for astrocytes) in the absence of AR pre-treatment. Importantly, Cav3.2 mRNA was mainly co-localized with Cav3.2 but not GFAP. Together, our findings indicate that AR pre-treatment or not impacts the expression pattern of Cav3.2, which may make a significant contribution to the future study of Cav3.2 in SDH.

Introduction

T-type (Cav3) channels are low-voltageactivated (LVA) calcium channels that are encoded by three α1 subunit genes: α1G (Cav3.1), α1H (Cav3.2), and α1I (Cav3.3),[1] which are widely distributed in brain, spinal cord and dorsal root ganglion (DRG).[2] Besides, immunohistochemical (IHC) studies have demonstrated that the proteins of Cav3.1-3.3 are broadly expressed in the peripheral and central nervous system.[3-7] The role of Cav3.2 in pathological pain has been extensively studied by behavioral, electrophysiological, and molecular biological methods during the last decade. For example, the Western blot immunoassay found that the expression of Cav3.2 was increased in DRG and spinal dorsal horn (SDH) from rodents of cystitis-related bladder pain,[8] paclitaxel-induced peripheral neuropathy,[9] and loose ligatures of sciatic nerve induced neuropathic pain models,[10] etc. Moreover, T-type currents recorded in DRG were enhanced in various animal models of pathological pain.[11] Consistently, intrathecal injection of an antisense oligonucleotide (targeted to the α1-subunit of Cav3.2, but not Cav3.1 or Cav3.3) or Ttype channel blockers could alleviate pathological pain.[12,13] The above results suggest that Cav3.2 plays a crucial role in the modulation of pathological pain. Therefore, it is important to identify the cell-type specific expression of Cav3.2 channel in DRG and SDH. Up to now, the expression pattern of Cav3.2 in DRG is consensus, that is, Cav3.2 is mainly distributed in small- and mediumdiameter DRG neurons.[5,9,13-18] However, in SDH, the expression pattern of Cav3.2 is different. Chen et al. found that Cav3.2 was restricted to the neurons but not microglia or astrocytes.[5] In contrast, Li et al. found that Cav3.2 was co-localized to glial fibrillary acidic protein (GFAP)-positive cells (marker of astrocyte) but not NeuN- (marker of neuron) or OX42-positive (marker of microglia) cells.[9] One possible explanation for these differences might be that formaldehyde fixation compromises immunoreactivity of Cav3.2 antigen by preventing contact between the epitopes and the antibodies, which will affect the outcome of an IHC staining.[19,20] To test this hypothesis, we investigated the IHC staining of Cav3.2 in SDH and DRG sections with or without antigen retrieval (AR) pretreatment which can unmask the antigens in formaldehyde-fixed tissue sections. The IHC staining of other two subtypes (Cav3.1 and Cav3.3) were also tested as well. To further determine the nature of cell-type specific expression of Cav3.2 protein, we also investigated the distribution of Cav3.2 mRNA using in situ hybridization. Together, our results demonstrated that AR pre-treatment is essential for the IHC experiment of Cav3.2 in SDH. These results may provide a theoretical basis for such kinds of experiments in the future.

Materials and Methods

Animals

Sprague-Dawley (SD) rats (5-8 w) and wild type (WT) C57/BL6 mice (6-8 w) of either sex were obtained from the Animal Center of Nanchang University. Cav3.2 knock-out (KO) mice (C57/BL6) were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). Animals were housed in controlled room temperature (RT, 22-25°C) and maintained on a 12-h light/dark cycle with water and food ad libitum. All experimental procedures were approved by the Ethics Committee of Nanchang University.

Tissue collection and preparation

The rats and mice were deeply anesthetized by intraperitoneal injection of 1.5 g/kg urethane and perfused transcardially with saline followed by cold 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer. The L4-L5 spinal cord segments were removed and post-fixed in 4% PFA for 6 h, and then cryoprotected in a 30% sucrose solution for 3 days. The L4 and L5 DRGs were post-fixed in 4% PFA for 6 h and cryoprotected in a 20% sucrose solution for 12 h. After embedding in optimal cutting temperature medium, 30 μm (IHC staining) or 15 μm (in situ hybridization) transverse spinal cord sections and 15 μm DRG sections were prepared using a freezing microtome (CM1950, Leica, Nussloch, Germany).

Immunostaining

All the spinal cord and DRG sections for IHC staining were first rinsed with 0.01 M phosphate-buffered saline (PBS) three times. Then they were divided into no-AR (without AR pre-treatment) or AR (with AR pre-treatment) groups. Sections in AR groups were subjected to heat-induced AR in a water bath (HH-2, China) at 98°C for 10 min by using 0.01 M sodium citrate buffer (pH 6.0, Solarbio) and were cooled to RT. Next, the free-floating sections were incubated in a blocking solution (0.3% Triton X-100, 1% albumin from bovine serum, and 1% normal donkey serum in PBS) for 30 min at RT to prevent nonspecific staining and then were incubated in a solution containing primary antibodies for three nights at 4˚C. Primary antibodies used were rabbit anti-Cav3.1-3.3 (1:200, Cat# ACC-021, Cat# ACC-025, Cat# ACC-009, Alomone), goat anti-Cav3.2 (1:50, Cat# sc- 16261, Santa Cruz), guinea pig anti-NeuN (1:200, Cat# 266004, Synaptic Systems), goat anti-GFAP (1:2000, Cat# ab53554, Abcam). After washing, the sections were incubated with appropriate secondary antibodies: donkey anti-rabbit Alexa Fluor 488 (1:400, Cat# A-21206, Invitrogen), donkey anti-goat Alexa Fluor 555 (1:400, Cat# A- 21432, Invitrogen), donkey anti-guinea pig Cy5 (1:400, Cat# 706-175-148, Jackson ImmunoResearch) overnight at 4°C. After mounting the sections with Aqua- Poly/mount medium (Polysciences, Inc., Warrington, PA), they were observed under a Zeiss LSM700 confocal microscope (Germany) with ZEN 2010 software. The specificity of Cav3.1-3.3 antibodies (Alomone) was tested by either omission or pre-absorption of primary antibodies with peptide antigens using sections with AR pre-treatment. Moreover, the specificity of anti-Cav3.2 (Alomone) was also confirmed by the Cav3.2 KO mice. Co-localization images of Cav3.2 with other antibodies were taken with a 63x oil objective at a 0.5x zoom. The parameters for acquiring the images, such as the number of recording pixels, electrical shutter speed, gain, and pinhole were kept unchanged throughout the whole experiment process. The quantification of co-localization was assessed by the ZEN 2010 software as our previous study.[21]

In situ hybridization

The method for in situ hybridization was modified from a previous publication.[15] The Cav3.2 probe (5’-ACAAUGCCAUCAAAGAUGUUGUAGGGGUUCCGAAUG- 3’) was designed for mouse Cacna1h gene (NM_021415.4) and was labeled at 5’-end with digoxigenin. Briefly, spinal cord sections were treated with proteinase K (20 μg/mL) for 20 min and fixed in 4% PFA for 15 min at RT. After that, the sections were rinsed with PBS containing 0.1% Tween-20 (PBST). Next, they were treated with 0.25% acetic anhydride in 0.1 M triethanolamine (pH 8.0) and rinsed with PBST three times. Following this, sections were incubated in pre-hybridization solution (50% formamide, 5×SSC, 0.1% Tween-20, 0.3 mg/mL yeast tRNA, and 5×Denhardt’s Solution) for 4 h at 37°C, and then incubated in the hybridization solution (pre-hybridization solution plus digoxinlabeled probe) for 16 h at 37°C. After hybridization, the sections were washed three times with 2X saline-sodium citrate, containing 0.1% Tween-20 (SSCT) at 37°C (15 min each) and one time with 1X SSCT and 0.5x SSCT at RT, respectively. Afterward, the sections were blocked with 0.5% sheep serum (Solarbio) for 1.5 h and incubated with alkaline phosphatase-conjugated anti-digoxigenin antibody (Roche Diagnostics) overnight at 4°C. Then they were incubated with a mixture of nitroblue tetrazolium chloride (NBT, Sangon Biotech) and 5-bromo-4-chloro-3-indolylphosphate (BCIP, Sangon Biotech) for 1~3 h for color development. In addition, to compare the results of the in situ hybridization with that of the IHC, partial sections were further incubated with anti-Cav3.2 (Alomone) and GFAP as well as 4’,6- Diamidino-2-Phenylindole (DAPI, 1:1000, Cat# D9542, Sigma, sections were incubated with DAPI for 10 min at RT before mounting) and were visualized by an FSX100 microscope equipped with a digital camera system (Olympus, Japan) or the Zeiss LSM700 confocal microscope.

Statistical analysis

SPSS version 17.0 (SPSS Inc, Chicago, IL, USA) was used for statistical analysis. All data are expressed as mean ± SEM. Shapiro–Wilk test was used to assess the normality of data. Differences between groups were compared using unpaired Student’s t-test. P<0.05 was considered statistically significant.

Results

Immunolabeling of Cav3 isoforms in SDH and DRG of SD rats with or without AR pre-treatment

Although the distributions of Cav3.1, Cav3.2, and Cav3.3 mRNAs have been identified in both SDH and DRG,[2] IHC studies in these areas were mostly focused on Cav3.2. However, hyperalgesia could also be observed in Cav3.1 null mice,[22] suggesting Cav3.1 might contribute to the sensory perception of pain as well. Therefore, we first investigated the expression of Cav3.1, Cav3.2, and Cav3.3 isoforms in SDH and DRG with or without AR pre-treatment. As shown in Figure 1, regardless of whether sections were processed with (right) or without (left) AR, Cav3.1-immunoreactivity (IR) showed a punctate staining pattern in SDH (Figure 1 A,D) and Cav3.3-IR was primarily distributed around the neuron-like membranes (Figure 1 C,F). However, we found a significantly different expression pattern of Cav3.2-IR in sections processed with AR from those without AR pre-treatment. For sections lacking AR pre-treatment, Cav3.2-IR showed a glial cell-like appearance (Figure 1B), which is similar to the results of Li et al.[9] However, for those sections with AR pre-treatment, Cav3.2-IR showed a neuron-like appearance (Figure 1E), which was in line with the findings of Chen et al.[5] Next, we investigated whether AR-caused different IHC staining exists in other tissues, such as DRG, which is mostly studied for the mechanisms of pain. As shown in Figure 1 G-L, Cav3.1, Cav3.2, and Cav3.3-IR were highly expressed at neuronlike membranes and cytoplasm. In addition, unlike the observations of Cav3.2-IR in SDH, AR pre-treatment did not affect the expression pattern of Cav3.2 in DRG. Together, these findings demonstrate that AR pre-treatment altered the IHC staining results of Cav3.2 in rat SDH but not DRG, without any effect on Cav3.1 and Cav3.3 isoforms.

Figure 1.

Expression of Cav3.1-3.3 isoforms (Alomone) in SDH and DRG of SD rats with or without AR pre-treatment. Representative confocal images of Cav3.1 (A), Cav3.2 (B), and Cav3.3 (C) in SDH without AR pre-treatment (No AR). Representative confocal images of Cav3.1 (D), Cav3.2 (F), and Cav3.3 (F) in SDH with AR pre-treatment (AR). Representative images of Cav3.1 (G), Cav3.2 (H), and Cav3.3 (I) in DRG without AR pre-treatment (No AR). Representative images of Cav3.1 (J), Cav3.2 (K), and Cav3.3 (L) in DRG with AR pre-treatment (AR).

Immunolabeling of Cav3 isoforms in SDH and DRG of mice with or without AR pre-treatment

Since the Cav3.2-IR showed a different IHC staining pattern in rat SDH, we wonder whether this difference is due to the species of animal. Besides rat, mouse is the most commonly used animal species. Therefore, we next performed the same experiment on mice SDH. We found that Cav3.1- (Figure 2 A,D) and Cav3.3-IR (Figure 2 C,F) were similar in mice SDH sections with or without AR pre-treatment. However, the results of Cav3.2-IR (Figure 2 B,E) were affected by AR pre-treatment, which is highly similar to that of the rats. AR pre-treatment led to the Cav3.2-IR exhibited a neuron-like appearance in mice SDH, too. However, no differences of Cav3.1- and Cav3.3-IR in SDH from both rat and mouse were observed whether pre-treated with AR or not.

Figure 2.

Expression of Cav3.1-3.3 isoforms (Alomone) in mice SDH sections pre-treated with AR or not. Representative images of Cav3.1 (A), Cav3.2 (B), and Cav3.3 (C) in SDH without AR pre-treatment (no AR). Representative images of Cav3.1 (D), Cav3.2 (E), and Cav3.3 (F) in SDH with AR pre-treatment (AR).

Discrepancies of Cav3.2-IR from two different sources of antibodies with or without AR pre-treatment

Anti-Cav3.1-Cav3.3 antibodies used in the above study were all from the Alomone Company. We were curious whether the different results of Cav3.2-IR in SDH would have been the same if an alternative commercial anti-Cav3.2 antibody was used. Therefore, we next compared the Cav3.2-IR in SDH by using two different commercial sources of anti-Cav3.2 (Alomone vs Santa Cruz). As shown in Figure 3 A-C, without AR pre-treatment, the IR of the two sources of Cav3.2 antibodies were similar in rat SDH sections, which both showed a glial cell-like appearance and a substantial colocalization. However, for those sections with AR pre-treatment, the IR of anti- Cav3.2 antibody from Alomone was mostly expressed in neuron-like cells (Figure 3D), while there was no change for the IR of anti-Cav3.2 antibody from Santa Cruz which still shows a glial cell-like appearance (Figure 3E). Furthermore, after pretreating with AR, the extent of co-staining was significantly decreased (Figure 3F). These data suggested that AR pre-treatment has a different effect on Cav3.2-IR from various commercial sources of antibodies.

Figure 3.

Expression of anti- Cav3.2 from two companies in rat SDH with or without AR pre-treatment. A) Representative image of Cav3.2 (Alomone, green) in SDH without AR pre-treatment (No AR). B) Representative image of Cav3.2 (Santa Cruz, red) in SDH without AR pretreatment (No AR). C) Merged images of (A) and (B). D) Representative image of Cav3.2 (Alomone, green) in SDH with AR pre-treatment (AR). E) Representative image of Cav3.2 (Santa Cruz, red) in SDH with AR pre-treatment (AR). F) Merged images of (D) and (E).

Co-localization of Cav3.2 with GFAP and NeuN in sections of SDH pre-treated with AR or not

Above results suggested that immunostaining of Cav3.2 (Alomone) has neuron-like and glial-cell like appearances with or without AR pre-treatment, respectively. To further confirm the cell-type specific expression of Cav3.2, we next co-stained Cav3.2 with GFAP and NeuN. As shown in Figure 4A-E, it is clear that without AR pre-treatment, Cav3.2 was largely co-stained with GFAP but not NeuN. In contrast, with AR pre-treatment, Cav3.2 was major co-stained with NeuN but not GFAP (Figure 4F-J). Quantitative analysis (Figure 4K) demonstrated that the co-localized ratio of Cav3.2 with GFAP in sections without AR pre-treatment was higher (30.4±4.4% vs 9.4±0.4%, n=7, P<0.01) than that of the sections pretreated with AR. The results for the co-localization of Cav3.2 with NeuN were vice versa (12.4±0.7% vs 28.2±1.3%, n=7, P<0.001). In control experiments by omitting the primary antibodies or pre-absorbing the primary antibodies with peptide antigens, no IHC signal was detected (Supplementary Figures 1 and 2). Moreover, we tested the specificity of Cav3.2 antibody (Alomone) by using the SDH sections with AR pre-treatment from WT and Cav3.2 KO mice. Strong Cav3.2-IR was observed in SDH of WT mice (Figure 4L) but not Cav3.2 KO mice (Figure 4M), which demonstrated the specificity of Cav3.2 antibody (Alomone). Furthermore, western blotting using SDH tissues from mice implied that Cav3.2 (Alomone) is expressed in WT mouse rather than KO mouse, suggesting its fine specificity for western blot (Supplementary Figure 3). Together, these results suggested that AR pre-treatment caused the co-localization of Cav3.2 shifted from GFAP to NeuN.

Figure 4.

Expression pattern of Cav3.2 (Alomone) in rat SDH with or without AR pre-treatment. Representative images of Cav3.2 (A, green), GFAP (B, red), NeuN (C, magenta), merged images of Cav3.2 and GFAP (D), and merged images of Cav3.2 and NeuN (E) in SDH without AR pre-treatment (No AR). Representative images of Cav3.2 (F, green), GFAP (G, red), NeuN (H, magenta), merged images of Cav3.2 and GFAP (I), and merged images of Cav3.2 and NeuN (J) in SDH with AR pre-treatment (AR). K) Quantitative colocalization results of Cav3.2-IR with NeuN- and GFAP-IR. Representative images of Cav3.2 (Alomone) in SDH from wild-type mice (L) and Cav3.2 knock-out (KO) mice (M). **P<0.01, ***P<0.001.

To identify the nature distribution pattern of Cav3.2, we next performed the in situ hybridization experiment to study the expression of Cav3.2 mRNA, which showed that the Cav3.2 mRNA positive stain cells have a neuron- like appearance (Figure 5A). The sections of in situ hybridization were further processed to IHC co-stain for Cav3.2, GFAP, and DAPI. Figure 5 B-G showed that most of the Cav3.2 mRNA positive cells are also anti- Cav3.2-positive but not GFAP-positive cells. For conclusion, these results demonstrated that the expression pattern of Cav3.2-IR with AR pre-treatment is closer to that of Cav3.2 mRNA, suggesting that the Cav3.2 positive cells should be neurons but not astrocytes.

Figure 5.

Expression pattern of Cav3.2 mRNA in mouse SDH. A) Microscopic images of a DIG-labeled Cav3.2 probe for in situ hybridization in mice SDH. Triple-labeling of in situ hybridization sections for Cav3.2 mRNA (B, purple) and the following IHC staining of Cav3.2 (C, green, Alomone), GFAP (D, red), DAPI (E, magenta), merged images of Cav3.2 and GFAP (F), and merged images of Cav3.2 and DAPI (G). Note that Cav3.2 mRNA is expressed in Cav3.2-positive cells (arrows) but not GFAP-positive astrocytes.

Discussion

The immunohistochemistry technique is used to detect cell or tissue antigens.[23] Principal factors affecting the outcome of IHC studies include the following: i) tissue fixation and processing, ii) unmasking of epitopes, and iii) sensitivity of the detection system.[24] Formaldehyde is one of the most popular fixatives due to its low cost, ease of preparation, and because of its well morphologic detail with few artifacts. However, formaldehyde fixation could mask or damage some binding sites of antibodies, which might lead to the loss of IR.[25,26] AR pretreatment, which is mainly based on hightemperature heating of tissues, is an effective method to overcome these disadvantages. [27,28] Since 2007, genetic linkage studies have implicated Ca gene coded protein as a pain modulation related protein. However, there were disagreements as to the expression pattern of Cav3.2 in SDH of SD rat.[5,9]We therefore raised the hypothesis that AR pre-treatment might account for this difference. To test this hypothesis, we observed the effect of AR pre-treatment on IHC staining of Cav3.1-3.3 in rat SDH and DRG as well as mouse SDH.

Cav3 channels can be activated around resting membrane potential and are inactivated within a few tens of milliseconds.[29] Hence, they are also known as transiently open calcium channels. In the nervous system, Cav3 channels are involved in the regulation of neuronal excitability, thus they are linked to the pathogenesis of various neurological disorders, including chronic pain. In multiple animal models of neuropathic pain, Cav3 was upregulated in DRG and SDH.[30-32] Among the three Cav3 isoforms, Cav3.2 is the most concerned subtype with neuropathic pain.[12] Consistent with previous studies showing that Cav3.2 was mainly distributed in small- to medium- sized DRG neurons,[14,15] our data found that Cav3.2, as well as Cav3.1 and Cav3.3, had similar expression pattern in DRG with or without AR pre-treatment. Interestingly, previous studies showed that the distribution pattern of Cav3.2 in SDH is controversial. Chen et al. found that, in L5/6 spinal nerve ligation pain model, the expression of Cav3.2 (Santa Cruz, rabbit) in SDH of SD rats was upregulated seven days after nerve ligation.[5] Moreover, they found that Cav3.2 is expressed in neurons (co-localize with NeuN) but not astrocytes (co-localize with GFAP). However, in SDH of SD rats suffered from paclitaxel-induced peripheral neuropathy, the Cav3.2 (Alomone, rabbit) expression was increased and co-localized to GFAP positive but not NeuN positive cells.[9] These distinct results may be due to the antibodies they used were from different companies. Thus, we investigated the Cav3.2 expression in SDH of SD rats with the antibodies from the above two companies. Differently from Chen et al.’s study (Santa Cruz - rabbit),[5] our results found that Cav3.2-IR from Santa Cruz (goat) showed a glia-like appearance and was irrelevant to AR pre-treatment. This difference might be due to the different sourced anti-Cav3.2. However, Santa Cruz does not provide rabbit- sourced Cav3.2 which is out of the market. Currently, only a mouse-sourced Cav3.2 is available. Therefore, we performed the IHC staining of Cav3.2 using a rabbitsourced anti-Cav3.2 (C1868) from Sigma- Aldrich (data not shown). Whereas, the Cav3.2-IR (Sigma-Aldrich) displayed a punctate staining with or without AR pretreatment, suggesting the source of Cav3.2 has little effect on Cav3.2-IR. In contrast, AR pre-treatment influenced the results of Cav3.2-IR from Alomone (rabbit). Our triple-labeling experiments found that Cav3.2 (Alomone) was co-localized with NeuN but not GFAP after AR pre-treatment, further demonstrating that AR pre-treatment impacts the results of Cav3.2-IR in SDH.

Given that mice are also the commonly used rodents for pain animal models, we next observed the effect of AR on Cav3.2 (Alomone) on SDH sections from mouse and got the same results. A recent study demonstrated that Cav3.2 was co-stained with PKCγ-containing interneurons in the Cav3.2-GFP knock-in mice, suggesting Cav3.2 is expressed in neurons.[6] In line with this study, our IHC results of Cav3.2-IR in mice showed that under the pre-treatment of AR, Cav3.2-IR displayed the neuron-like appearance.

Although one previous report described Cav3.2 (α1H) mRNA was mostly expressed in the outermost layers (layers 1-2) of SDH,[2] our present findings indicate that Cav3.2 mRNA is present throughout the gray matter of spinal cord. Since the sections for in situ hybridization included sodium citrate and heat pre-treatment, post-IHC of Cav3.2 protein (Alomone) from in situ hybridization sections largely matched the distribution of Cav3.2 mRNA. Furthermore, our and previous electrophysiological studies have revealed the presence of T-type channels in SDH neurons,[33-35] which highly suggested that Cav3.2 is expressed in neurons. These data suggested that AR pretreatment might be necessary for specific antigens, including Cav3.2.

Taken together, we have demonstrated that AR pre-treatment affect the Cav3.2-IR in rat and mouse SDH and we believe that our findings make a significant contribution to the future research of Cav3.2 distribution in SDH from both rats and mice.