Targeting ion channels in leukemias: a new challenge for treatment.

Leukemias, as other cancers, bear several genetic alterations of tumor-related genes, such as point mutations, translocations, epigenetic modifications, often accompanied by gene amplification or inactivation. The identification of tumor-related genes provides considerable insight into the biology of leukemias and opens the way to more specific pharmacological treatments. These genes comprise several ion channels and pumps, as the transport mechanisms associated with volume control, proliferation and apoptosis are often altered in cancers. In leukemic cells, such changes are observed as early as the stem cell stage. Ion channels can regulate other malignant features, such as lack of differentiation, increased migratory and invasive phenotype and chemoresistance. The role of certain voltage-gated K(+) channels, such as K(v)11.1 (also known as hERG1) can be largely attributed to modulation of cell adhesion to the extracellular matrix (ECM). K(v)11.1 exerts pleiotropic regulatory effects by forming multiprotein membrane complexes with integrin receptors in both acute myeloid leukemias (AML) and acute lymphoblastic leukemias (ALL). By recruiting growth factor and chemokine receptors, these complexes form signaling hubs that control neoplastic progression. Work in mice shows that blocking K(v)11.1 has a protective effect in acute leukemias. Ion channels are most promising targets for anti-leukemic therapy, because of their accessibility from the extracellular side and the thorough understanding of their pharmacology. In ALL cells, K(v)11.1 inhibitors abrogate the protective effect of bone marrow stromal cells and enhance the cytotoxicity of some common antileukemic drugs. Hence, ion channel modulators could overcome chemoresistance in acute leukemias, a major hindrance to therapeutic success.

Targeting ion channels in cancer: a novel frontier in antineoplastic therapy.

Targeted therapy is considerably changing the treatment and prognosis of cancer. Progressive understanding of the molecular mechanisms that regulate the establishment and progression of different tumors is leading to ever more specific and efficacious pharmacological approaches. In this picture, ion channels represent an unexpected, but very promising, player. The expression and activity of different channel types mark and regulate specific stages of cancer progression. Their contribution to the neoplastic phenotype ranges from control of cell proliferation and apoptosis, to regulation of invasiveness and metastatic spread. As is being increasingly recognized, some of these roles can be attributed to signaling mechanisms independent of ion flow. Evidence is particularly extensive for K(+) channels. Their expression is altered in many primary human cancers, especially in early stages, and they frequently exert pleiotropic effects on the neoplastic cell physiology. For instance, by regulating membrane potential they can control Ca(2+) fluxes and thus the cell cycle machinery. Their effects on mitosis can also depend on regulation of cell volume, usually in cooperation with chloride channels. However, ion channels are also implicated in late neoplastic stages, by stimulating angiogenesis, mediating the cell-matrix interaction and regulating cell motility. Not surprisingly, the mechanisms of these effects are manifold. For example, intracellular signaling cascades can be triggered when ion channels form protein complexes with other membrane proteins such as integrins or growth factor receptors. Altered channel expression can be exploited for diagnostic purposes or for addressing traceable or cytotoxic compounds to specific neoplastic tissue. What is more, recent evidence indicates that blocking channel activity impairs the growth of some tumors, both in vitro and in vivo. This opens a new field for medicinal chemistry studies, which can avail of the many available tools, such as blocking antibodies, antisense oligonucleotides, small interfering RNAs, peptide toxins and a large variety of small organic compounds. The major drawback of this approach is that some ion channel blockers produce serious side effects, such as cardiac arrhythmias. Therefore, drug developing efforts aimed at producing less harmful compounds are needed and we discuss possible approaches toward this goal. Finally, we propose that a novel therapeutic tactic could be developed by unlocking ion channels from multiprotein membrane signaling complexes.

hERG1 channels are overexpressed in glioblastoma multiforme and modulate VEGF secretion in glioblastoma cell lines.

Recent studies have led to considerable advancement in our understanding of the molecular mechanisms that underlie the relentless cell growth and invasiveness of human gliomas. Partial understanding of these mechanisms has (1) improved the classification for gliomas, by identifying prognostic subgroups, and (2) pointed to novel potential therapeutic targets. Some classes of ion channels have turned out to be involved in the pathogenesis and malignancy of gliomas. We studied the expression and properties of K(+) channels in primary cultures obtained from surgical specimens: human ether a gò-gò related (hERG)1 voltage-dependent K(+) channels, which have been found to be overexpressed in various human cancers, and human ether a gò-gò-like 2 channels, that share many of hERG1's biophysical features. The expression pattern of these two channels was compared to that of the classical inward rectifying K(+) channels, IRK, that are widely expressed in astrocytic cells and classically considered a marker of astrocytic differentiation. In our study, hERG1 was found to be specifically overexpressed in high-grade astrocytomas, that is, glioblastoma multiforme (GBM). In addition, we present evidence that, in GBM cell lines, hERG1 channel activity actively contributes to malignancy by promoting vascular endothelial growth factor secretion, thus stimulating the neoangiogenesis typical of high-grade gliomas. Our data provide important confirmation for studies proposing the hERG1 channel as a molecular marker of tumour progression and a possible target for novel anticancer therapies.

Interneurons transiently express the ERG K+ channels during development of mouse spinal networks in vitro.

During spinal cord maturation neuronal excitability gradually differentiates to meet different functional demands. Spontaneous activity, appearing early during spinal development, is regulated by the expression pattern of ion channels in individual neurons. While emerging excitability of embryonic motoneurons has been widely investigated, little is known about that of spinal interneurons. Voltage-dependent K+ channels are a heterogeneous class of ion channels that accomplish several functions. Recently voltage-dependent K+ channels encoded by erg subfamily genes (ERG channels) were shown to modulate excitability in immature neurons of mouse and quail. We investigated the expression of ERG channels in immature spinal interneurons, using organotypic embryonic cultures of mouse spinal cord after 1 and 2 weeks of development in vitro. We report here that all the genes of the erg family known so far (erg1a, erg1b, erg2, erg3) are expressed in embryonic spinal cultures. We demonstrate for the first time that three ERG proteins (ERG1A, ERG2 and ERG3) are co-expressed in the same neuronal population, and display a spatio-temporal distribution in the spinal slices. ERG immuno-positive cells, representing mainly GABAergic interneurons, were present in large numbers at early stages of development, while declining later, with a ventral to dorsal gradient. Patch clamp recordings confirmed these data, showing that ventral interneurons expressed functional ERG currents only transiently. Similar expression of the erg genes was observed at comparable ages in vivo. The role of ERG currents in regulating neuronal excitability during the earliest phases of spinal circuitry development will be examined in future studies.

Physical and functional interaction between integrins and hERG potassium channels.

Integrins are adhesion receptors capable of transmitting intracellular signals that regulate many different cellular functions. Among integrin-mediated signals, the activation of ion channels can be included. We demonstrated that a long-lasting activation of hERG (human ether-a-go-go-related gene) potassium channels occurs in both human neuroblastoma and leukaemia cells after the activation of the beta1 integrin subunit. This activation is apparently a determining factor inducing neurite extension and osteoclastic differentiation in both the cell types. More recently, we provided evidences that beta1 integrins and hERG channels co-precipitate in both the cell types. Preliminary results suggest that a macromolecular signalling complex indeed occurs between integrins and the hERG1 protein and that hERG channel activity can modulate integrin downstream signalling.

Megaesophagus in an asthmatic patient and beta2 stimulant treatment by inhalation.

Megaesophagus is a severe esophageal malformation. We report a case of megaesophagus in an asthmatic patient affected by congenital non-haemolytic anaemia and undergoing beta2 stimulant treatment by inhalation. Our case could be due to chronic beta2 receptor stimulation with imbalance of alpha and beta receptor, without any implication of favism.

Phosphatase inhibitors increase the open probability of ENaC in A6 cells.

We studied the cellular phosphatase inhibitors okadaic acid (OKA), calyculin A, and microcystin on the epithelial sodium channel (ENaC) in A6 renal cells. OKA increased the amiloride-sensitive current after approximately 30 min with maximal stimulation at 1-2 h. Fluctuation analysis of cell-attached patches containing a large number of ENaC yielded power spectra with corner frequencies in untreated cells almost two times as large as in cells pretreated for 30 min with OKA, implying an increase in single channel open probability (P(o)) that doubled after OKA. Single channel analysis showed that, in cells pretreated with OKA, P(o) and mean open time approximately doubled. Two other phosphatase inhibitors, calyculin A and microcystin, had similar effects on P(o) and mean open time. An analog of OKA, okadaone, that does not inhibit phosphatases had no effect. Pretreatment with 10 nM OKA, which blocks protein phosphatase 2A (PP2A) but not PP1 in mammalian cells, had no effect even though both phosphatases are present in A6 cells. Several proteins were differentially phosphorylated after OKA, but ENaC subunit phosphorylation did not increase. We conclude that, in A6 cells, there is an OKA-sensitive phosphatase that suppresses ENaC activity by altering the phosphorylation of a regulatory molecule associated with the channel.

HERG potassium channels are constitutively expressed in primary human acute myeloid leukemias and regulate cell proliferation of normal and leukemic hemopoietic progenitors.

An important target in the understanding of the pathogenesis of acute myeloid leukemias (AML) relies on deciphering the molecular features of normal and leukemic hemopoietic progenitors. In particular, the analysis of the mechanisms involved in the regulation of cell proliferation is decisive for the establishment of new targeted therapies. To gain further insight into this topic we report herein a novel approach by analyzing the role of HERG K(+) channels in the regulation of hemopoietic cell proliferation. These channels, encoded by the human ether-a-gò-gò-related gene (herg), belong to a family of K(+) channels, whose role in oncogenesis has been recently demonstrated. We report here that herg is switched off in normal peripheral blood mononuclear cells (PBMNC) as well as in circulating CD34(+) cells, however, it is rapidly turned on in the latter upon induction of the mitotic cycle. Moreover, hergappears to be constitutively activated in leukemic cell lines as well as in the majority of circulating blasts from primary AML. Evidence is also provided that HERG channel activity regulates cell proliferation in stimulated CD34(+) as well as in blast cells from AML patients. These results open new perspectives on the pathogenetic role of HERG K(+) channels in leukemias.

S-adenosyl-L-homocysteine hydrolase is necessary for aldosterone-induced activity of epithelial Na(+) channels.

The A6 cell line was used to study the role of S-adenosyl-L-homocysteine hydrolase (SAHHase) in the aldosterone-induced activation of the epithelial Na(+) channel (ENaC). Because aldosterone increases methylation of several different molecules, and because this methylation is associated with increased Na(+) reabsorption, we tested the hypothesis that aldosterone increases the expression and activity of SAHHase protein. The rationale for this work is that general methylation may be promoted by activation of SAHHase, the only enzyme known to metabolize SAH, a potent end-product inhibitor of methylation. Although aldosterone increased SAHHase activity, steroid did not affect SAHHase expression. Antisense SAHHase oligonucleotide decreased SAHHase expression and activity. Moreover, this oligonucleotide, as well as a pharmacological inhibitor of SAHHase, decreased aldosterone-induced activity of ENaC via a decrease in ENaC open probability. The kinetics of ENaC in cells treated with antisense plus aldosterone were similar to those reported previously for the channel in the absence of steroid. This is the first report showing that active SAHHase, in part, increases ENaC open probability by reducing the transition rate from open states in response to aldosterone. Thus aldosterone-induced SAHHase activity plays a critical role in shifting ENaC from a gating mode with short open and closed times to one with longer open and closed times.

Cyclic-nucleotide-gated channels: pore topology in desensitizing E19A mutants.

In cyclic-nucleotide-gated "CNG" channels, the pore-loop "P-loop" is formed by the amino acid residues R345-S371 (here called R1-S27). Residue E19 determines the channel's interaction with extracellular divalent cations and contributes to ion conduction. Neutralization of this residue with alanine introduces channel desensitization. We have used serial cysteine mutagenesis to study P-loop topology in the alpha subunit of the mammalian rod CNG channels containing the E19A substitution. The pore topology was tested in the closed channel state and, when cGMP was present, during and after desensitization. With E19A substitution, the T15C, T16C, I17C and T20C mutants desensitized more slowly than controls. Moreover, the typical rundown produced by the I17C substitution in the wild-type "w.t." background was considerably reduced. Overall, with the E19A substitution, the accessibility pattern tested by applying the thiol-specific reagents Cd2+ and MTSET from the cytoplasmic side of the plasma membrane was similar to that observed with the w.t. Moreover, P22C channels were not inhibited by Cd2+ and MTSET (which do not cross the lipid bilayer) applied from the inside, but were blocked by MTSEA (which permeates the plasma membrane) also applied from the inside. This suggests that the residues following E19 remain accessible from the external side after E19A substitution. Thus, although the residues T15 to T20 seemed to participate in the structural rearrangements producing desensitization, no major P-loop remodelling occurs in desensitizing channels.

HERG potassium channels are more frequently expressed in human endometrial cancer as compared to non-cancerous endometrium.

HERG K(+)channels, besides contributing to regulate cardiac and neuronal excitability, are preferentially expressed in tumour cell lines of different histogenesis, where their role in the development and maintenance of the neoplastic phenotype is under study. We show here that both herg gene and HERG protein are expressed with high frequency in primary human endometrial cancers, as compared to normal and hyperplastic endometrium. RT-PCR and immunohistochemistry, using specific anti-HERG antibodies developed in our laboratory, were applied to tissue specimens obtained from 18 endometrial cancers and 11 non-cancerous endometrial tissues. herg RNA and HERG protein are expressed in 67% and 82%, respectively, of cancerous, while in only 18% of non-cancerous tissues. In particular, no expression was found in endometrial hyperplasia. Moreover, electrophysiological experiments confirmed the presence of functioning HERG channels on the plasma membrane of tumour cells. On the whole, these data are the first demonstration of the presence of HERG channels in primary human neoplasias, and could candidate HERG as a potential tool capable of marking cancerous versus hyperplastic endometrial growth.

The nicotinic receptor beta 2 subunit is mutant in nocturnal frontal lobe epilepsy.

Clustered attacks of epileptic episodes originating from the frontal lobe during sleep are the main symptoms of autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE, MIM 600513). Despite the clinical homogeneity, three forms of ADNFLE have been associated with chromosomes 20 (ENFL1; ref. 1), 15 (ENFL2; ref. 2) and 1 (ENFL3; ref. 3). Mutations of the gene encoding the neuronal nicotinic acetylcholine receptor alpha 4 subunit (CHRNA4 ) have been found in ADNFLE-ENFL1 families, but these mutations account for only a small proportion of ADNFLE cases. The newly identified locus associated with ENFL3 harbours several candidate genes, including CHRNB2 (ref. 8), whose gene product, the beta 2 nicotinic acetylcholine receptor (nAChR) subunit, co-assembles with the alpha 4 nAChR subunit to form the active receptor.

Cyclic nucleotide-gated channels: intra- and extracellular accessibility to Cd2+ of substituted cysteine residues within the P-loop.

In cyclic nucleotide-gated (CNG) channels from the bovine rod, the pore loop "P-loop", connecting the S5 and S6 transmembrane segments, is formed by the residues R345-S371 (here named R1-S27). It determines channel selectivity and contributes to gating. We have studied its topology, by testing the accessibility to Cd2+ of serially substituted cysteine residues. Channels were expressed in Xenopus oocytes. The accessibility of V4C, S6C, T16C, 117C, T20C, P22C and S27C from the cytoplasmic side of the plasma membrane was tested by applying 1-100 microM Cd2+ to the inner face of inside-out patches, at negative membrane potentials. Under these conditions, the effect of Cd2+ on wild-type channels was negligible. The accessibility of the same residues from the external side of the membrane was tested by measuring CNG current inhibition persisting after wash-out of Cd2+ applied to outside-out patches. T16C and I17C channels were strongly inhibited by Cd2+ from the inside, in the presence of cGMP. The Kd for T16C block was 16 microM. Thus the T16 and I17 residues participate directly in channel function and are accessible from the cytoplasmic side when the channels are open. In contrast, V4C, T20C and P22C residues were only inhibited when 100 microM Cd2+ was applied externally, suggesting that V4C, T20C and P22C face the outer side of the P-loop.

Methylation increases the open probability of the epithelial sodium channel in A6 epithelia.

We used single channel methods on A6 renal cells to study the regulation by methylation reactions of epithelial sodium channels. 3-Deazaadenosine (3-DZA), a methyltransferase blocker, produced a 5-fold decrease in sodium transport and a 6-fold decrease in apical sodium channel activity by decreasing channel open probability (P(o)). 3-Deazaadenosine also blocked the increase in channel open probability associated with addition of aldosterone. Sodium channel activity in excised "inside-out" patches usually decreased within 1-2 min; in the presence of S-adenosyl-l-methionine (AdoMet), activity persisted for 5-8 min. Sodium channel mean time open (t(open)) before and after patch excision was higher in the presence of AdoMet than in untreated excised patches but less than t(open) in cell-attached patches. Sodium channel activity in excised patches exposed to both AdoMet and GTP usually remained stable for more than 10 min, and P(o) and the number of active channels per patch were close to values in cell-attached patches from untreated cells. These findings suggest that a methylation reaction contributes to the activity of epithelial sodium channels in A6 cells and is directed to some regulatory element closely connected with the channel, whose activity also depends on the presence of intracellular GTP.

The properties of cysteine mutants in the pore region of cyclic-nucleotide-gated channels.

The properties of the pore of cyclic-nucleotide-gated channels are thought to depend mostly on the structure of the "P loop" which connects the S5 and S6 transmembrane segments. We applied scanning cysteine mutagenesis to the P loop of the alpha subunit of cGMP-gated channels from bovine rod, i.e. to the amino acid residues K346 to S371 (here named K2-S27). Mutant channels were expressed in Xenopus laevis oocytes and currents recorded from inside-out patches, in symmetrical sodium and in the absence of divalent cations. Cysteine mutations had minor effects on the affinity for cGMP and the selectivity to monovalent alkali cations. However, in mutants within the segment W9-T16, open probability was strongly reduced. It was less than 0.2 in saturating cGMP compared to 0.8 in wild-type channels. Moreover, W9C and L12C mutants were outward rectifiers, while T16C was an inward rectifier. These mutant channels showed a time dependence in the development of the steady-state current. Currents from I17C channels progressively decayed in inside-out patches. This rundown was prevented by 1 mM dithiothreitol on the cytoplasmic side of the plasma membrane, suggesting that I17 is an intracellular residue. On the other hand, mutants within the segment T20-S27 had normal open probabilities, around 0.8 in saturating cGMP, but altered single-channel conductance.

Cyclic nucleotide-gated channels. Pore topology studied through the accessibility of reporter cysteines.

In voltage- and cyclic nucleotide-gated ion channels, the amino-acid loop that connects the S5 and S6 transmembrane domains, is a major component of the channel pore. It determines ion selectivity and participates in gating. In the alpha subunit of cyclic nucleotide-gated channels from bovine rod, the pore loop is formed by the residues R345-S371, here called R1-S27. These 24 residues were mutated one by one into a cysteine. Mutant channels were expressed in Xenopus laevis oocytes and currents were recorded from excised membrane patches. The accessibility of the substituted cysteines from both sides of the plasma membrane was tested with the thiol-specific reagents 2-aminoethyl methanethiosulfonate (MTSEA) and [2-(trimethylammonium)ethyl]methanethiosulfonate (MTSET). Residues V4C, T20C, and P22C were accessible to MTSET only from the external side of the plasma membrane, and to MTSEA from both sides of the plasma membrane. The effect of MTSEA applied to the inner side of T20C and P22C was prevented by adding 10 mM cysteine to the external side of the plasma membrane. W9C was accessible to MTSET from the internal side only. L7C residue was accessible to internal MTSET, but the inhibition was partial, approximately 50% when the MTS compound was applied in the absence of cGMP and 25% when it was applied in the presence of cGMP, suggesting that this residue is not located inside the pore lumen and that it changes its position during gating. Currents from T15C and T16C mutants were rapidly potentiated by intracellular MTSET. In T16C, a slower partial inhibition took place after the initial potentiation. Current from I17C progressively decayed in inside-out patches. The rundown was accelerated by inwardly applied MTSET. The accessibility results of MTSET indicate a well-defined topology of the channel pore in which residues between L7 and I17 are inwardly accessible, residue G18 and E19 form the narrowest section of the pore, and T20, P21, P22 and V4 are outwardly accessible.

Low expression of the ClC-2 chloride channel during postnatal development: a mechanism for the paradoxical depolarizing action of GABA and glycine in the hippocampus.

In early postnatal development, during the period of synapse formation, gamma-aminobutyric acid (GABA) and glycine, the main inhibitory transmitters in the adult brain, paradoxically excite and depolarize neuronal membranes by an outward flux of chloride. The mechanisms of chloride homeostasis are not fully understood. It is known that in adult neurons intracellular chloride accumulation is prevented by a particular type of chloride channel, the ClC-2. This channel strongly rectifies in the inward direction at potentials negative to ECl thus ensuring chloride efflux. We have tested the hypothesis that in the developing hippocampus, a differential expression or regulation of ClC-2 channels may contribute to the depolarizing action of GABA and glycine. We have cloned a truncated form of ClC-2 (ClC-2nh) from the neonatal hippocampus which lacks the 157 bp corresponding to exon 2. In situ hybridization experiments show that ClC-2nh is the predominant form of ClC-2 mRNA in the neonatal brain. ClC-2nh mRNA is unable to encode a full-length protein due to a frameshift, consequently it does not induce any currents upon injection into Xenopus oocytes. Low expression of the full-length ClC-2 channel, could alter chloride homeostasis, lead to accumulation of [Cl-]i and thereby contribute to the depolarizing action of GABA and glycine during early development.

Lithium blocks cell cycle transitions in the first cell cycles of sea urchin embryos, an effect rescued by myo-inositol.

Lithium is a classical inhibitor of the phosphoinositide pathway and is teratogenic. We report the effects of lithium on the first cell cycles of sea urchin (Lytechinus pictus) embryos. Embryos cultured in 400 mM lithium chloride sea water showed marked delay to the cell cycle and a tendency to arrest prior to nuclear envelope breakdown, at metaphase and at cytokinesis. After removal of lithium, the block was reversed and embryos developed to form normal late blastulae. The lithium-induced block was also reversed by myo- but not epi-inositol, indicating that lithium was acting via the phosphoinositide pathway. Lithium microinjection before fertilization caused arrest prior to nuclear envelope breakdown at much lower concentrations (3-5 mM). Co-injection of myo-inositol prevented the block. Microinjection of 1-2 mM lithium led to block at the cleavage stage. This was also reversed by coinjection of myo-inositol. Embryos blocked by lithium microinjection proceeded rapidly into mitosis after photolysis of caged inositol 1,4,5-trisphosphate. These data demonstrate that a patent phosphoinositide signalling pathway is essential for the proper timing of cell cycle transitions and offer a possible explanation for lithium's teratogenic effects.