Introduction

Ca2+-activated Cl currents (CaCC) are abundant and are present in many different cell types, although with slight differences regarding their biophysical properties and pharmacology [30, 38]. CaCC mediates Ca2+-dependent Cl secretion in glands and flat epithelia and modifies cellular responses to adequate stimuli in muscle, nerve and receptors (Table 1). This is therefore why the druggable channels may be ideal pharmacological targets to control physiological functions or to correct defects in diseases such as cystic fibrosis [51, 93].

Table 1 Expression, function, and role in disease of anoctamins

There is hardly any tissue in which CaCC does not have a basic or at least a modulatory role (Table 1). Among these tissues are various sensory receptors, different types of smooth muscles, heart, endothelium, neuronal tissues, and epithelial organs (reviewed in [25, 34, 38, 48, 58, 68]). Anoctamins are expressed in all of these tissues (Table 1). As Ano 1 has been identified as a major component of the Ca2+-activated Cl channel [12, 100, 121], one might speculate that all anoctamins participate in formation of Ca2+-activated Cl currents in these organs. Cl currents generated by expression of Ano 1 have characteristics that come very close to those described for endogenous channels. Remarkably, silencing of Ano 1 by siRNA in a number of different cell types abolished CaCC, and mice lacking Ano 1 have defects in Ca2+-dependent Cl transport in a number of tissues, which supports the concept of Ano 1 forming an essential component of CaCC [82, 93, 113, 121]. As the basic molecular and physiological properties of anoctamins have been described already in previous reviews [30, 35, 51], we will focus in the present review on more recent findings and will speculate on additional functions of anoctamins, apart from their role as Ca2+-activated Cl channels.

Molecular aspects of anoctamins

Molecular aspects of Ano 1 have been already discussed in earlier reviews [30, 51]. The family of anoctamins consists of ten different proteins but only Ano 1 has been examined in more detail. Anoctamins do not show any obvious homology to other ion channels. The 986 amino acids of Ano 1 (Homo sapiens) form a tertiary structure with eight predicted transmembrane helices, intracellular NH2− and COOH ends and a pore, formed by the 5th and 6th transmembrane helices together with a p-loop dipping back into the membrane (Fig. 1a). This still preliminary structural information is based on mutagenesis of amino acids critical to pore formation (R647, K671, K694 in the a, c, d splicing isoform) (Fig. 1a), which changed the anion selectivity [121]. Moreover, Ano 1 comes as different splice variants, which has an impact on ion channel properties [31]. Thus, skipping of exon 6b may increase Ca2+ sensitivity while skipping of exon 13 abolishes the characteristic time-dependent activation observed for Ca2+-activated Cl channels. Interestingly, the pore loop contains three cysteines (at position 651, 656, 661 in the a, c, d isoform) and cysteine reagents were able to block Ca2+-activated Cl currents generated by Ano 1 [1, 51, 121]. All anoctamins carry at least one consensus site for N-glycosylation in their fifth extracellular loop (Fig. 1a).

Fig. 1
figure 1

Molecular properties of Ano 1: a Model for Ano 1 indicating eight transmembrane helices and a pore loop containing several cysteine and charged amino acid residues. At least one glycosylation site is present at the last extracellular loop. Ano 1 is differentially spliced (blue ovals a, b, c, d) and may contain non-canonical CAM binding sites in the N terminus. The C terminus contains two putative Erk1,2 phosphorylation sites. b Ano 1 forms stable dimers, with either two separate pores or one common pore. Whether anoctamins form heterodimers is currently unknown

The sequence homology within the putative pore region of anoctamins is considerable, while the overall homology is only moderate [51, 72]. Ano 1 and Ano 2 (TMEM16B) are close relatives with about 60% amino acid identity. In contrast to Ano 1 and Ano 2, the other anoctamins demonstrate an identity generally of 30% or below that may indicate a possible functional divergence of these proteins. Obviously anoctamin paralogs evolved from subsequent gene duplication events which were followed by a functional divergence of vertebrate anoctamins [72]. Interaction with auxiliary proteins would be suggested by the presence of protein interaction domains, which however are only present in Ano 2, 5, 7, and 9 (PDZ and coiled-coil domain) [72]. So far, Cl channel activity has been clearly identified for Ano 1 and Ano 2 [86, 96, 106, 108, 121]. We also found Ca2+-activated Cl currents generated by Ano 6 and Ano 7, while Ano 5, 8, 9, and 10 did not produce any measureable currents [99]. Ano 1 is clearly different from the other anoctamins since it is able to produce Ca2+-dependent Cl currents of much bigger amplitude when compared to Ano 2, 6, and 7 [99].

It is also entirely possible that anoctamins have diverse functions either as plasma membrane localized proteins or in intracellular compartments. In this respect, it is rather exiting that Ano 6 was found to induce scramblase activity in platelets [110]. Ano 6 induces Ca2+-dependent translocation of phosphatidylserine (PS) from the inner leaflet of the plasma membrane to the outer leaflet and is therefore essential for platelet aggregation. Patients suffering from the rare Scott syndrome, which is characterized by impaired blood coagulation, were shown to have a loss-of-function mutation in TMEM16F [15, 110]. Although Ano 1 does not have scramblase activity [110], it will be interesting to learn whether other anoctamins also facilitate PS scrambling.

Structural aspects of TMEM16 proteins were illuminated by two more recent publications, indicating that Ano 1 does not exist as a single protein [28, 102]. The data from both reports indicate that Ano 1 exists as obligate homodimer, similar to CLC-0 channels [62, 71]. As the quaternary structure of Ano 1 is not altered by changes in cytosolic Ca2+, dimerization appears to be permanent. Dimerization of Ano 1 seems to be stable, independent of Ca2+, and does not depend on the cytoskeleton [102]. Moreover, the present data do not indicate interactions with other proteins [28, 102]. It will now be interesting to study whether each Ano 1 subunit forms its own independent Cl channel pore as described for ClC channels [24], or whether the homodimer forms only one pore (Fig. 1b). Also, future work will show if heterodimeric channels are formed, similar to the ClC channels [27]. A heterodimeric architecture could help to explain why some anoctamins influence each other. We found earlier that coexpression of Ano 9 reduces the activity of Ano 1 [99].

Typically several anoctamins are expressed in parallel in a given cell. A detailed analysis of expression of all ten anoctamins indicated that every cell type in mouse and human tissues expresses at least two, but often several different anoctamins, with Ano 6 being the most abundant paralog [51, 99]. Ano 6 and 10 are broadly expressed in mouse and human tissues. Ano 8 is also broadly expressed although at lower levels. Ano 2, 3, and 4 are preferentially expressed in sensory receptor cells and neuronal tissues, while Ano 5 was found in skeletal muscle and thyroid gland [99]. In addition, splice variants of one given anoctamin are coexpressed in cells. Moreover, we found that expression of anoctamins is not stable, i.e., it largely varies with age, cell density, and polarization (unpublished data from the author's laboratory), which may explain inconsistent findings regarding expression levels of Ano 1 [1, 102].

Regulation of Ano 1

Although Ano 1 is activated by a rise in intracellular [Ca2+], a canonical Ca2+ binding domain has not yet been identified. The team led by Galietta proposed a cluster of four joined glutamic acid residues in the first intracellular loop, which may serve as Ca2+ binding site [30, 31] (Fig. 1a). We found recently that calmodulin (CAM) is required for activation of Ano 1 [113]. CAM was shown to bind to Ano 1, possibly within the N terminus. Moreover, we found a requirement for cytosolic ATP to fully activate Ano 1 [113]. When membrane patches were excised after stimulation of Ano 1 expressing HEK293 cells, the activated Cl currents rapidly inactivated [17, 113]. Similar results were found for HeLa cells overexpressing Ano 1. Purinergic stimulation (100 μM ATP) activated a whole cell current in Ano 1 expressing HeLa cells but not in mock transfected cells (Fig. 2a). Cl currents could also be activated in cell attached patches from Ano 1-transfected HeLa cells, but currents inactivated immediately after excision of the cell membrane (Fig. 2b). Thus, cellular components are required to keep the channel active.

Fig. 2
figure 2

Ano 1 channels require cytosolic components to maintain activity: a Whole cell patch clamp experiments in mock-transfected and Ano 1-expressing HeLa cells. Stimulation of purinergic receptors by ATP (10 μM) activates a substantial whole cell current only in Ano 1-expressing cells as indicated by the current/voltage relationships. b Patch clamp experiments with Ano 1-expressing and mock-transfected HeLa cells. Activation of Ano 1 currents by ATP in cell-attached patches of Ano 1-expressing cells. After excision of the membrane patch, the current inactivates immediately. Patch clamp conditions were as described in [113]

Although Ano 1 carries multiple putative phosphorylation sites for protein kinase (PK) A and C, CAMK and CK2, none of these kinases seem to be relevant for gating of Ano 1 [51]. Thus, although the presence of calmodulin and ATP appears important for activation of Ano 1, calmodulin-dependent kinase (CAMKII) is not required [32, 113]. However, we identified two extracellular regulated kinase (Erk1,2) sites at the C terminus which are relevant for receptor-mediated activation of Ano 1 (Fig. 1a), while activation through direct increase in intracellular Ca2+ does not seem to require phosphorylation by Erk1,2 [113]. This finding could have important implications for gating of Ano 1 and suggests that Ano 1 can be activated through other Ca2+-independent pathways.

When expressed in Fisher rat thyroid cell (FRT), we found Ano 1, 2, and 6 well expressed in the plasma membrane, while other Ano 5, 7, 8, 9, and 10 were much less membrane localized, which probably has a substantial impact on the current size [99, 113]. We further noticed that overexpression of Ano 1 in FRT cells [99], HeLa cells (Fig. 3a), HEK293 cells [52], and oocytes from Xenopus laevis (Fig. 3b–f) induced large baseline Cl currents, which were due to spontaneously active Ano 1 channels. Enhanced baseline conductance that is found upon expression of Ano 1 is not due to enhanced intracellular Ca2+ concentrations; since they are at similar levels or even lower in cells expressing anoctamins (data not shown). Large baseline Cl conductance is not due to possible patch clamp artifacts since it was also found in fluorescence measurements with HeLa cells. In HeLa cells overexpressing Ano 1, I influx induced quenching of YFP-I152L, while additional stimulation of the cells with 100 μM ATP did not further increase influx of iodide and YFP-quenching. This suggests that no additional Cl conductance is activated by an increase in intracellular Ca2+ by ATP-stimulation (Fig. 3a). This could indicate a higher Ca2+ sensitivity of the overexpressed channel, which therefore may be already active at baseline Ca2+ levels. We may also suggest that additional regulatory subunits are missing, which keep the channel shut under control conditions.

Fig. 3
figure 3

Receptor activation of Ano 1: a Summary curves of fluorescence quenching experiments with HeLa cells stably expressing I sensitive YFP in the absence or presence of Ano 1. Application of extracellular I to Ano 1-expressing cells (right panel) leads to quenching of YFP-fluorescence without Ca2+-dependent stimulation. Additional increase of intracellular Ca2+ by extracellular ATP (10 μM) does not further increase quenching (not shown), indicating that overexpression of Ano 1 leads to a spontaneously active anion conductance in these cells. b Double electrode voltage clamp experiments with Xenopus laevis oocytes expressing P2Y2 receptors (left panel) or coexpressing P2Y2 receptors and mAno 1 (right panel). Oocytes expressing P2Y2-R only, activate endogenous xANO 1 upon stimulation with 100 μM ATP. Much larger currents are activated in mANO 1 expressing oocytes, which are slightly inwardly rectifying (insert). These oocytes also demonstrate a large baseline Cl conductance even in the absence of ATP. c Summary of the ATP-activated and ionomycin (1 μM)-activated conductances measured in P2Y2 (−Ano 1) and P2Y2 + Ano 1 (+Ano 1) expressing cells. d Summary of ATP-activated conductances in and P2Y2 + Ano 1 expressing cells and inhibition by DIDS (100 μM) and NFA (10 μM). e, f Effect of removal of extracellular Cl, Na+, and Ca2+ on ANO 1 currents (e) and conductance (f). Experimental conditions as described in [29, 99]

Receptor activation of Ano 1

Xenopus laevis oocytes are known for their endogenous Ca2+ sensitive Cl currents (Fig. 3b). It is this endogenous Xenopus Ano 1 that made expression cloning of Ano 1 in oocytes difficult and that prompted Schroeder and colleagues to move to oocytes from Axolotl, which are free of CaCC [100]. Ca2+-activated Cl currents typically show more or less outward rectification, depending on the increase in intracellular Ca2+ [38]. An example for receptor (P2Y2) mediated activation of endogenous CaCC is shown in the left panel in Fig. 3b. Extracellular ATP (100 μM) activates a transient outwardly rectifying Cl current. Expression of mouse Ano 1 in oocytes largely augmented Cl currents activated by ATP (or ionomycin; Fig. 3b, c). Interestingly, these Cl currents induced by overexpression of Ano 1 were linear or even slightly inwardly rectifying (Fig. 3b, right panel), when compared to endogenous currents, which show slight outward rectification (Fig. 3b, left panel). A similar current behavior was noticeable in the cloning papers by Schroeder and Yang [100, 121]. Thus, there seem clear differences in current properties between endogenous CaCC and overexpressed anoctamins. Notably, a differential time and voltage dependence was reported recently, depending on whether anoctamin was examined as whole cell currents or in cell excised inside/out membrane patches [17]. When we determined the ion selectivity of overexpressed mAno 1 under bi-ionic conditions, i.e., after ion replacements in the bath, we found a substantial ATP (Ca2+)-activated Na+ conductance in addition to the Cl conductance (Fig. 3e, f). This pronounced Na+ permeability is not explained by the only moderate selectivity of Ca2+-activated Cl channels for Cl over Na+ by only 0.1 [88]. It is possible that expression of Ano 1 translocates additional Ca2+/Na+ influx pathways to the cell membrane that may shape the current voltage relationship of whole cell currents. Also, depending on the number of Cl channels activated by increase in intracellular Ca2+, the cell membrane potential will be more or less depolarized which will activate voltage gated Ca2+ influx pathways.

Ano 1 and more

Not much is known currently about the interaction of TMEM16A with other proteins. Current data indicate dimerization of Ano 1 and interaction with components of the intracellular signaling pathways such as CAM [28, 52, 102, 113]. An interesting finding concerns the interference of Ca2+-activated Cl channels and cystic fibrosis transmembrane conductance regulator (CFTR). CFTR is a cAMP/PKA/ATP-regulated Cl channel that also controls the function of other membrane proteins [56]. It has been reported that CFTR inhibits endogenous Ca2+-activated Cl currents (CaCC) in Xenopus oocytes, bovine pulmonary artery endothelium, and isolated parotid acinar cells, by an unknown mechanism [54, 85, 118, 119]. Moreover, also volume-regulated anion channels were inhibited by expression of CFTR, which shares some properties with Ca2+-activated Cl currents [1, 116]. Additional studies demonstrated that upregulation or downregulation of CFTR resulted in a parallel up- and downregulation of cAMP-, Ca2+-, and volume-regulated Cl conductance [50]. Although these data clearly suggested a relationship of these different anion conductances, they have been regarded as separate molecular entities [2]. Recently, we could recapitulate the CFTR-dependent inhibition of CaCC by demonstrating inhibition of Ano 1 by coexpressed CFTR after cAMP-dependent activation. However, we found no evidence for a direct molecular interaction, as both proteins did not coimmunoprecipitate (data not shown). A recent paper examined a possible molecular interaction of both channels by Förster resonance energy transfer (FRET), but detected only minimal FRET between CFTR and Ano 1 [102]. Thus, the authors suggested that regulation of Ano 1 by CFTR is not due to a direct interaction.

Developmental aspects of anoctamins

Members of the anoctamin family are widely expressed during vertebrate embryogenesis [37, 91]. This is perhaps not surprising given that functions typically associated with CaCC, including secretion and smooth muscle contraction, are active during development [79, 97]. Despite their abundant expression, little is known about the functions of anoctamin family members during embryogenesis. This is, in part, because null mice have been reported only for Ano 1 [90]. Further, only Ano 5 has been associated with a congenital defect in humans so far [9, 41].

Because of its widespread expression and the fundamental importance of CaCC, Ano 1 has been identified by a number of investigators examining diverse questions in a number of tissues. For example, Ano 1 was identified in a screen for genes expressed in the zone of polarizing activity (ZPA) of the embryonic mouse limb bud [92]. Subsequent studies have found that this gene is expressed in a number of tissues of all three germ layers in the developing embryo [37, 91]. Mice homozygous for a null mutation of Ano 1 die within the first month of life and exhibit decreased calcium-activated chloride currents in a number of tissues [44, 82, 93, 94]. These mice demonstrated a severe failure to thrive and therefore it has been complicated to identify the precise cause of death. As early as embryonic day (E)14.5, all homozygous mutant mice exhibit abnormal tracheal morphology and develop gaps in the cartilage rings that normally surround the ventrolateral trachea [90, 93]. Intriguingly, expression of Ano 1 was never detected in the embryonic chondrogenic mesenchyme and so it was hypothesized that the cartilage defects are secondary to defects of the respiratory epithelium or smooth muscle where Ano 1 is abundantly expressed. A similar tracheal patterning defect has been observed in CFTR knockout mice, CFTR deficient pigs, and perhaps in CF patients [70, 93]. It remains to be seen if this defect is attributable to decreased chloride secretion in epithelial or mesenchymal cells. The generation of a conditional null allele of Ano 1 will facilitate the characterization of this and other developmental defects observed in Ano 1 null mice. Preliminary data suggest that deletion of Ano 1 specifically from the embryonic airway epithelium does not lead to tracheal malformation (JRR unpublished data).

Ano 5 is one of the members of the family that has been associated with a disease other than cancer in humans. During embryogenesis, Ano 5 is expressed in developing mesenchymal cells including the somites, myotome-derived skeletal muscle precursors, cardiac myofibers, and chondrocytes [73]. Mutations in Ano 5 have been linked to gnathodiaphyseal dysplasia, a disease of the bones, and two forms of muscular dystrophy [9, 115]. The function of Ano 5 in musculoskeletal cells and the mechanisms by which these mutations contribute to disease remain to be identified. As described, only Ano 1 and Ano2 have been reported to encode channels with CaCC activity and endogenous and overexpressed Ano 5 have been localized primarily to intracellular vesicles [73, 115]. The development of animal models and the identification of other human mutations in anoctamin family proteins will facilitate the identification of functions for these proteins during normal development.

Also postnatal expression of anoctamins may vary considerably with age. This is of particular interest for Cl secretion in the intestinal epithelium. A Ca2+-dependent Cl secretion has been detected in the colon and small intestine of neonatal mice, which disappeared in older mice [74]. In both mouse proximal and distal colon, Ca2+-dependent Cl secretion through muscarinic stimulation fades with increase in animal age (Fig. 4a). This Ca2+-dependent Cl secretion was shown to be activated during infection with rotavirus, which is probably the reason for infantile gastroenteritis in more than 600,000 patients every year worldwide [5, 61]. We found recently activation of Ca2+-dependent Cl secretion in the distal colon of young mouse pups, which disappeared completely in older animals [83]. We further demonstrated activation of Ano 1 by an active peptide of the rotavirus toxin NSP4. Interestingly, expression of Ano 1 is found in the neonatal colon throughout epithelial cells, but appears to be expressed only in basolateral membranes in epithelial cells of adult mouse colon (Fig. 4b, c). Similarly expression of Ano 1 was found in the basolateral membrane of colonic epithelial cells of adult guinea pigs where it may contribute to production of electrogenic K+ secretion [39]. In contrast, clear luminal staining is seen in the distal colon of younger animals (Fig. 4c). Taken together, Ano 1 appears as a major player in rotavirus-induced diarrhea in the infant/juvenile intestine and may become a major pharmacological target for the treatment of the worldwide common rotavirus diarrhea.

Fig. 4
figure 4

Role of Ano 1 for rotavirus toxin induced Cl - secretion: a Age dependence of Ca2+-activated Cl secretion (muscarinic basolateral stimulation with 100 μM carbachol) in mouse proximal and distal colon. Cl secretion was detected as CCH-activated short circuit currents (I sc). b, c Immunohistochemistry of Ano 1 in colon of neonatal, younger, and older mice. Ano 1 is found in the apical membrane only in colon mucosa of younger animals but is located basoalerally in adult animals. d Model for rotavirus toxin NSP4 induced Cl secretion, which occurs through Ca2+-activated Ano 1 channels. The receptor for NSP4 and the ATP-release mechanism are unknown

An intracellular function of anoctamins?

Regulated secretion in exocrine cells occurs through exocytosis of secretory granules and the subsequent release of vesicular proteins. Secretory granules of pancreatic acinar cells express ClC Cl channels which are believed to facilitate acidification and exocytosis of these granules [112]. A number of transport proteins have been identified in secretory granules of exocrine glands and it thus could be possible that other Cl channels, apart from ClC channels, contribute to granular acidification and exocytosis. Ano 1 is clearly plasma membrane localized in overexpressing HEK293 and Cos-7 cells, according to fluorescence images and electron microscopy [113]. However, in some fluorescence images, a portion of Ano 1 appears to be localized also intracellularly, particularly those from mouse pancreatic and submandibular acinar cells [121] (Fig. 5a). In immunofluorescence, Ano 1 was detected in apical as well as basolateral membranes and also in the cytosol of exocrine cells from mouse pancreas and salivary glands. When analyzing proteins in whole cell lysates and the isolated granular fraction of pancreatic and submandibular glands, we detected Ano 1 in both fractions, supporting a possible role of Ano 1 in intracellular granules (Fig. 5b).

Fig. 5
figure 5

An intracellular function of Ano 1? a Immunohistochemistry of Ano 1 in mouse submandibular acinar cells indicates predominant basolateral, but also luminal and intracellular granular expression. b Western blot analysis of Ano 1 in total lysates or intracellular granules only of mouse pancreas and submandibular gland cells. In all samples, the expected 110 kDa band is detected. Only pancreatic acinar cells show a 150 kDa band which may reflect glycosylated Ano 1. c Cell swelling and light scattering by isolated granules in suspension. Because the granular membrane contains Cl channels, incorporation of the K+ ionophore valinomycin leads to KCl-induced osmotic swelling of the granules and change in light scattering

The Cl transport in isolated granules can be examined using light scattering (LS). LS is changed during application of the K+ ionophore valinomycin, when Cl channels are activated and vesicles are swelling (Fig. 5c, upper panel). An example for a change in LS in isolated mouse pancreatic granules is shown in the lower panel of Fig. 5. Using this technique, it will be interesting to compare Cl conductances and LS of granules isolated from pancreatic acini of Ano 1 null mice and from wild-type animals. A possible intracellular role of anoctamins has also been postulated for Ano 8, 9, and 10, as these anoctamins have been found exclusively in intracellular compartments when expressed in Fisher rat thyroid (FRT) and HEK293 cells [99]. Interestingly Ano 1, which is upregulated in human gastrointestinal stromal tumors (GIST, c.f. below), has been detected to a significant level in intracellular compartments [3, 26, 120].

Anoctamins in health and disease

Table 1 summarizes the role of Ca2+-activated Cl channels in epithelial and non-epithelial tissues. It is evident that CaCC has a role in most tissues, and thus it is not surprising to find anoctamins broadly expressed in all cell types. However, we clearly need to distinguish between data from mouse and human tissue. So far, most expression data are from mouse [37, 42, 82, 90, 93, 94, 99, 100, 121] and rather little is known about expression in native human tissues [3, 7, 65]. While mouse native tissue and cultured human cells [10, 31, 32, 51, 76] express significant amounts of Ano 1, expression levels in native human tissues are currently not well determined. In preliminary RT-PCR analysis of Ano 1-expression in human nasal and colonic epithelium, we were unable to detected transcripts for Ano 1, while other anoctamins, such as Ano 6, were readily expressed (unpublished data). These results parallel earlier observations showing that Ca2+-dependent Cl transport is small in human when compared to mouse airways [51] and that Ca2+-dependent Cl secretion in adult human colon relies entirely on the presence of CFTR [53, 64]. A recent study by the Verkman team even questioned the role of Ano 1 a major component of CaCC in airway and intestinal epithelial cells [76]. Their conclusions are based on the fact that broad inhibitors of CaCC such as tannic acid and the arylaminothiophene CaCCinh-A01, fully inhibited CaCC current in human bronchial and intestinal cells, while an Ano 1—specific inhibitor had little effects on Ca2+-activated Cl secretion in these cells [76]. In contrast, siRNA-knockdown of Ano 1 in HT29 colon epithelial cells and human airway epithelial cell lines inhibited CaCC [1, 32, 66]. At any rate, these results obtained from cultured cells may not necessarily reflect the true in vivo situation, particularly since expression of Ano 1 appears largely polarization/differentiation dependent (unpublished data).

Irrespective of the uncertainty of the role of Ano 1 in human tissues, the numbers of reports is continuously increasing, indicating an essential role of Ano 1 in various epithelial tissues, visceral and vascular smooth muscles, neuronal cells, and receptors (summarized in Table 1). Loss of expression of Ano 1 in knockout animals leads to multiple defects in epithelial organs such as airways, salivary glands, pancreatic glands, hepatobiliary tract, and large intestine [23, 33, 59, 82, 93, 94, 114, 121]. More recent studies show the importance of Ano 1 in Cajal pacemaker and smooth muscle cells of airways, [20, 36, 44, 65, 122]. Lack of Ano 1 in mouse Cajal cells leads to reduced slow wave activity in gastrointestinal muscles [44], while expression alternatively transcribed Ano 1 was found in human interstitial cells of Cajal of patients with diabetic gastroparesis [67]. Meanwhile, anoctamins have also been identified in neuronal tissues and in olfactory and taste receptors, the inner ear and the retina [11, 16, 46, 69, 84, 89, 106, 108, 117]. Defects in expression or function of these anoctamins lead to ataxia, panic disorder, defects in pain perception, and nonsyndromic hearing defects (Table 1). However, recent results obtained in a mouse loxP/cre knockout model for Ano 2 provided the rather surprising result that although Ano 2 clearly is the ciliary Ca2+-activated Cl channel in the main olfactory epithelium that is virtually abolished in Ano 2−/− mice, it seems entirely dispensable for olfaction [8]. Moreover, anoctamins have been linked to muscular dystrophy, myopathy, gnathodiaphyseal dysplasia, hemolysis, and platelet adhesion, indicating a role of anoctamins in skeletal muscle, endothelium, erythrocytes, and cartilage/bone (Table 1).

Before Ano 1 was identified as Ca2+-activated Cl channel, it was actually known as cancer-associated protein in gastrointestinal stromal tumors (GISTs) and head and neck squamous cell carcinomas [13, 14, 120]. Ano 1 was therefore called DOG1 (discovered on gastrointestinal stromal tumor 1), TAOS2 (tumor amplified and overexpressed sequence 2), and ORAOV2 (oral cancer overexpressed 2). Ano 1 may promote cancer development, a concept that is supported by the fact that it is embedded within the 11q13 amplicon, which is associated with a poor prognosis in squamous cell carcinoma of the head and neck [43]. It could work in concert with cancer-relevant proteins such as cyclin D1 or Fas-associated death domain. In preliminary experiments, however, we could not find a positive correlation between expression of Ano 1 and cell proliferation (unpublished data). A previous report pointed out that amplification of Ano 1 in head and neck squamous cell carcinoma (HNSC) cells stimulates attachment, spreading, detachment, and invasion, which could account for its effects on migration [4]. Amplification of ANO 1 may therefore be a marker for distant metastasis in HNSC, and overexpression of ANO 1 may affect cell properties linked to metastasis.

The ability of cancer cells to metastasize is linked to their ability to migrate. Migration of cells depends on their intracellular proton concentration and their ability to increase their volume at the leading edge with extension of the lamellipodium and to shrink at retract their rear part [101, 107] (Fig. 6b). A central role in this cell shrinkage has the charybdotoxin-inhibited and Ca2+-sensitive IK1 (or Gardos) K+ channels. It is suggested that activation of IK1 requires parallel activation of a Cl exit pathway in order to maintain electroneutrality (Fig. 6b). Accordingly, anoctamins may serve as volume regulated Cl channels (Fig. 6a). In fact, we found earlier that Ano 1 produces volume-regulated chloride currents that are reduced in mice lacking expression of Ano 1 [1]. We showed that Ano 1 together with other anoctamins are activated by cell swelling through an autocrine mechanism that involves ATP release and binding to purinergic P2Y2-receptors. Ano 1 channels are activated by ATP through increase in intracellular Ca2+ and by a Ca2+- independent mechanism engaging extracellular regulated protein kinases (Erk1/2) [1]. Moreover, in erythrocytes, Ca2+-activated Gardos channels are also essential for cell shrinkage, and we also found that in addition, Ano 1 is also important for volume regulation in mouse red blood cells, as cell shrinkage after exposure to bacterial hemolysin was significantly attenuated and lysis was enhanced in erythrocytes isolated from Ano 1 null mice compared with wild-type littermates [103]. Other anoctamins are also related to cancer. Thus, expression of Ano 7, also known as NGEP, is enhanced in prostate cancers [6, 19, 47]. Moreover, splicing variants of Ano 6 (TMEM16F) are associated with metastatic capability of mammary cancers in mouse and is also associated with poor prognosis of patients with breast cancer [22]. It will be very interesting to learn more about the role of these anoctamins in cancer development and metastasis. The fact that these proteins are related to malignancy provides another convincing example for the role of ion channels in cancer [49].

Fig. 6
figure 6

Role of Ano 1 in cancer: a Potential role of anoctamins and Gardos K+ channels in volume regulation and cell shrinkage after hypotonic cell swelling or increase in intracellular Ca2+. ATP release leads to activation of puringeric signaling along with activation of Erk1,2. b Anoctamin and Gardos-mediated cell shrinkage at the rear end of migrating cells