Ano1 is a better marker than c-Kit for transcript analysis of single interstitial cells of Cajal in culture

The interstitial cells of Cajal (ICC) drive the slow wave-associated contractions in the small intestine. A commonly used marker for these cells is c-Kit, but another marker named Ano1 was recently described. This study uses single-cell RT-PCR, qPCR and immunohistochemistry to determine if Ano1 could be reliably used as a molecular marker for ICC in single-cell mRNA analysis. Here, we report on the relationship between the expression of c-Kit and Ano1 in single ICC in culture. We observed that Ano1 is expressed in more than 60% of the collected cells, whereas c-Kit is found only in 22% of the cells (n = 18). When we stained ICC primary cultures for c-KIT and ANO1 protein, we found complete co-localization in all the preparations. We propose that this difference is due to the regulation of c-Kit mRNA in culture. This regulation gives rise to low levels of its transcript, while Ano1 is expressed more prominently in culture on day 4. We also propose that Ano1 is more suitable for single-cell expression analysis as a marker for cell identity than c-Kit at the mRNA level. We hope this evidence will help to validate and increase the success of future studies characterizing single ICC expression patterns.


INTRODUCTION
Pacemaker cells called the interstitial cells of Cajal (ICC) drive the slow waveassociated contractions in the small intestine [1][2][3]. Research on the physiology and biochemistry of these cells through patch clamping and other electrophysiological techniques helped to gain insight into their function as the pacemakers of the gut [4][5][6][7]. However, other molecular techniques, such as RT-PCR and qPCR have only been used in a limited way because there is no pure culture of ICC or ICC cell line on which to perform gene expression analysis separately from the associated tissues, such as the enteric nerves and smooth muscle [8]. To date, only one transcriptomic analysis of ICC has been achieved after enrichment and purification of cell samples through fluorescence-activated cell sorting [9], but the implementation and validation of such a method requires specialized equipment and considerable economic investment. With the availability of transgenic mice with copGFP-expressing ICC [10], primary cultures have been used to improve identification in electrophysiological or immunohistochemical analyses [11,12] but not in single cell characterization. One alternative for the molecular analysis of ICC is the single-cell RT-PCR technique, which allows the collection of individual cells from a mixed culture [13]. The use of single-cell RT-PCR has increased in recent years thanks to the introduction of new technologies and the implementation of ready-to-use PCR products. However, the research on single cell expression profiling with ICC has been limited, with only few publications on the subject, mostly dedicated to the identification of the ICC through c-Kit expression, which is a broadly accepted ICC marker [1,3,14,15]. Some of the problems encountered performing singlecell RT-PCR with ICC are the small amount of genetic material obtained and the need to amplify the ICC marker c-Kit from every sample, since the most simple form of the protocol allows the amplification of only one target [16]. The large majority of experiments in ICC have been performed in culture, and previous reports have indicated that culture conditions may affect the biochemistry and function of ICC [17,18]. In particular, c-Kit is a gene that can be greatly influenced by the presence of serum factors like TGF-beta, which significantly decreases the half-life of c-Kit mRNA [19]. This phenomenon could affect the results of expression studies involving single-cell RT-PCR and using c-Kit as a molecular marker. Recent evidence has identified the calciumactivated chloride channel TMEM16A/anoctamin 1 (Ano1) in ICC [10]. Immunofluorescence studies reported 100% co-localization of this channel with c-Kit, and it is now accepted as an additional marker of ICC identity in both culture and tissue [20,21].

ICC primary cell culture
Short-term primary cultures of ICC were generated by enzymatic digestion of dissected small intestinal muscle tissue as previously described [22,23]. Small intestines were removed from 5-to 15-day old CD-1 mice (Charles River Laboratories) and dissected using blunt dissection. The gut wall was cut open at the mesenteric border, and then the mucosa was removed along with the mesentery. The muscle was cut into pieces and incubated for 15 min at 36ºC in HEPES-buffered saline (HS) with the addition of 1 mg/ml type F collagenase, 1 mg/ml bovine serum albumin, 0.5 mg/ml papain, 0.5 mg/ml soybean trypsin inhibitor and 0.2 mg/ml (−)-1,4-dithio-L-threitol (all from Sigma). After trituration of the smooth muscle, the cell suspension was settled on collagencoated cover slips and cultured for 3-4 days before use, using the Clonetics SmGM-2 system (Lomax, supplied by Cedarlane). All of the procedures were carried out in accordance with regulations from the Animal Research Ethics Board (AREB) of McMaster University in accordance with guidelines from the Canadian Council on Animal Care.

Relative expression RT-PCR
To assess the relative expression of c-Kit and Ano1 compared to GAPDH, primary cultures of ICC were prepared as described above. Day 0 corresponded to a stabilized culture in serum-free solution, while days 2 and 4 correspond to that many days of culture in the normal culture medium. The Ambion Cells to cDNA II Kit for cDNA extraction was used according to the manufacturer's instructions. The internal primers designed for c-Kit and Ano1 are also qPCR compatible. We detected GAPDH expression with the primers GAPDHF 5'-CCATGGAGAAGGCCGGGG and GAPDHR 5'-CAAAGTTGTCATGGATGACC (PCR product: 198 bp). The program included an initial denaturing of 95ºC for 5 min, followed by 35 cycles of 10 s of denaturation at 95ºC and annealing/extension at 60ºC for 5 s. A melting curve was applied to ensure the specificity of the PCR products (65 to 95ºC with 0.5ºC steps every 5 s).

Single cell isolation and RNA extraction
In order to isolate the ICC in primary culture for RT-PCR, we utilized a patch clamp rig as described elsewhere [23]. ICC-MP were identified by their roughly triangular shape with a process at each apex, found singly. Protease (0.1 mg/ml) was used to detach ICC from the collagen-coated coverslips. Unpolished, lowresistance pipettes were used to remove cells from the coverslips. Cells were removed by applying negative pressure to the pipette. A no cell (NC) control was included. For it, we simulated the collection of a cell by lowering the pipette into the bath solution. The pipettes for single cell extraction contained 0.6 µl of RNase-free 10x RT Buffer with RNase inhibitor (20 units per sample) to a final volume of 6 µl. The contents of the pipette were expelled with positive pressure into a PCR tube containing 12.5 µl of RNase-free RT mixture consisting of 2.3 µM oligo (dT), 150 µM dNTPs, 1.2 mM dTT, 3.6 mM MgCl 2 and 1.4 µl of 10x RT Buffer (Life Technologies) along with 0.5 µl of 1% NP40 detergent to cause cell membrane disruption.
The reaction was incubated at 65ºC for 2 min. After the addition of 1 µl reverse transcriptase (Superscript III, Invitrogen), the sample was placed at 50ºC for 90 min. For positive controls, tissue extracted from adult CD-1 murine brains was triturated in a mortar with a pestle in liquid nitrogen. Afterwards, we weighed 10-20 mg of tissue and collected it in Eppendorf tubes with 500 µl of lysis solution from an RNeasy RNA isolation kit (Qiagen). The RNA was obtained from the lysis solution using an affinity column and was collected for cDNA synthesis using the instructions of the Superscript III First Strand Synthesis Kit (Invitrogen).

Single-cell RT-PCR
The single cells obtained from primary cultures were tested for Ano1 and c-Kit expression using a nested approach. The external primers for pre-amplification were: Ano1F 5'-TGTACTTTGCCTGGCTTGGAGC and Ano1R 5'-CACCTGGC AATGCAGCCGTA (PCR product: 700 bp); and c-kitF 5'-GCTCAT TGGCTTTGTGGTTGCAG and c-kitR 5'-ATGCGCCAAGCAGGTTCACAA (PCR product: 404 bp). For nested PCR we used the internal primers: Ano1intF 5'-CAACTACCGATGGGACCTCAC and Ano1intR 5'-AATAGG CTGGGAATCGGTCC (PCR product: 170 bp); and c-kitintF 5'-ATA GACCCGACGCAACTTCCT and c-kitintR 5'-AACTGTCATGGCAGCATC CGAC (PCR product: 150 bp). Pre-amplification of the targets was carried out on half of the single cell cDNA, or 200 ng of tissue cDNA, by cycling 30 times at 50ºC and extending for 1 min at 72ºC. Then, nested PCR was carried out using internal specific primers. The PCR protocol was performed on a CFX96 thermal cycler (Bio-Rad Laboratories Canada Ltd.): initial denaturation for 3 min at 94ºC, then 35 amplification rounds of denaturation for 15 s at 94ºC, alignment for 15 s at 55-58ºC, and extension for 30 s at 72ºC. The final extension was 5 min at 72ºC. For both amplifications, recombinant Taq Polymerase was used according to the manufacturer's instructions (Life Technologies). Negative controls were performed without a template; no false amplifications were obtained. The resulting products were analyzed via agarose electrophoresis in 1.5% agarose gels (Invitrogen) stained with 1 µg/ml ethidium bromide (Sigma-Aldrich). Images were obtained with a Gel-Doc 2000 documentation system (Bio-Rad Laboratories Canada Ltd.). The identities of all of the amplicons produced were confirmed by sequencing (MOBIX Laboratories, McMaster University).

ANO1 and c-KIT immunohistochemistry
For immunohistochemistry, both musculature whole-mount tissue and cultured cells were made from the proximal jejunum of CD1 mice processed according to the following protocol. Tissues were fixed in ice-cold acetone for 10 min. After incubation with 5% normal goat serum for 1 h to block non-specific staining, tissues were incubated with monoclonal rat anti-c-Kit (ACK4, 1:200, Cedarlane) overnight, followed by Cy3 conjugated goat anti-rat IgG (1:600, Jackson ImmunoResearch) incubation for 1 h at room temperature. After c-Kit staining, the tissues were fixed again with 4% (w/w) paraformaldehyde in phosphatebuffered saline (PBS) for 1 h. The tissues were incubated with rabbit anti-ANO1 (1:100, AbCam Inc.) and then with Alexa 488-conjugated goat anti-rabbit IgG (1:200, Jackson ImmunoResearch). All of the antibodies were diluted in 0.3% Triton X-100 in PBS (pH 7.4). Control tissues were prepared by omitting primary antibodies. Pictures were taken using a confocal microscope (Zeiss LSM 510) with excitation wavelengths (543 nm and 488 nm) appropriate for Cy3 and Alexa 488.

Ano1 and c-Kit relative expression
In our relative expression analysis, we observed that c-Kit levels remain constant during primary culture, but they were low on day 4 compared to Ano1, which increased around fourfold (p < 0.05, Fig. 1A). This suggests that Ano1 is a better candidate for ICC identification in single cell expression analyses. The effect of serum on c-Kit mRNA regulation has been shown in other cell types [19], but the precise dynamics of c-Kit transcription, translation and cys-acting mechanisms in ICC require further investigation.

Standard single-cell Ano1 and c-Kit RT-PCR
Since Ano1 levels were highest on day 4, we looked for Ano1 and c-Kit expression in single ICC at this point during culture. We were able to amplify two genes from single interstitial cells of Cajal: Ano1 and c-Kit. We found that of the eight Ano1-positive cells, only one exhibited c-Kit expression (Fig. 1B). Of 11 cells tested, 3 cells did not exhibit Ano1 or c-Kit amplification. These cells were not taken into account in the percentage reported because we cannot be sure that any PCR product could be amplified from their cDNA.

Multiplexed RT-PCR and immunohistochemistry
The possibility of a multiplexed approach to increase the number of targets amplified has been reported elsewhere [24]. We applied this methodology to single ICC for this study. We revisited the ratio of Ano1 and/or c-Kit expression in these multiplexed experiments with similar results (n = 18 cells; 12 Ano1-positive cells; 4 c-Kit-positive cells). Additional genes were amplified from the Ano1-positive cells (voltage-gated porins and potassium channels, data not shown), confirming that cDNA had been synthesized correctly and the lack of c-Kit expression was not an artifact of the technique used. At the protein level we found 100% co-localization between Ano1 and c-Kit in both tissue and cultured cells (Fig. 2), which is consistent with previous reports. This suggests that our negative c-Kit results could be the product of the low levels of c-Kit mRNA present in the cells at the moment of the extraction, a diminished half-life of its mRNA, or a combination of the two factors, whereas the Ano1 transcript levels increased. NC denotes a no cell control, while the negative control was performed without template. The identity of the products was confirmed by sequencing. Fig. 2. c-Kit (red) and ANO1 (green) immunoreactivities in mouse jejunum musculature. A1 through A3 -Wholemount preparations show a dense ICC-MP network. B1 through B3 -Cultured preparations show triangle or multipolar-shaped ICC. Co-localization of c-Kit and ANO1 was 100% in ICC-MP of both tissue (A3) and cultured cells (B3).
While Ano1 transcript levels in cultures seem to increase by day 4, it is likely that they reach a steady-state level. This mainly because Ano1 is required for the generation of slow waves [25] and also because the recorded slow waves maintain their characteristics stably after several days in culture [26]. ANO1 has been previously related to cell division and regulatory volume decrease (RVD) [27][28][29]. Therefore, the expression of ANO1 in our system could be related to compensation of osmotic homeostasis after stress produced by the tissue disruption process needed to generate ICC primary cultures. In prostate cancer, evidence suggests that ANO1 could regulate swellingactivated Ca 2+ entry through BCL2 activation and could thus regulate calcium homeostasis in these cells [28]. ANO1 has also been found overexpressed in gastrointestinal stromal tumors [30,31], which originate from the ICC and present high BCL2 levels. However, the role of ANO1 expression regulation over BCL2 and over calcium oscillations remains to be investigated. ANO1 Ca 2+ -activated Clchannels play an important role in slow wave generation. A recent study with Tmem16a -/mice showed the complete loss of pacemaker activity in the muscle of the Tmem16a -/mouse antrum and small intestine, whereas the ICC network and c-Kit immunoreactivity appeared normal [24]. Loss of pacemaker activity was found in both W/W v and Tmem16a -/mice, suggesting that ANO1 shares the same functional significance as traditional ICC marker c-Kit at the mRNA level. Our study showed the higher success of single-cell PCR from ICC using Ano1, so Ano1 is a better marker than c-Kit for transcript analysis of single ICC. We expect that the use of Ano1 as a marker for single cell identification at the mRNA level will increase the success rate of further single ICC PCR experiments in the field, allowing validation and faster up-scaling to medium and high-throughput platforms.