Opposing modulation of Cx26 gap junctions and hemichannels by CO2.

KEY POINTS: A moderate increase in PCO2 (55 mmHg) closes Cx26 gap junctions. This effect of CO2 is independent of changes in intra- or extracellular pH. The CO2 -dependent closing effect depends on the same residues (K125 and R104) that are required for the CO2 -dependent opening of Cx26 hemichannels. Pathological mutations of Cx26 abolish the CO2 -dependent closing of the gap junction. Elastic network modelling suggests that the effect of CO2 on Cx26 hemichannels and gap junctions is mediated through changes in the lowest entropy state of the protein. ABSTRACT: Cx26 hemichannels open in response to moderate elevations of CO2 ( PCO2 55 mmHg) via a carbamylation reaction that depends on residues K125 and R104. Here we investigate the action of CO2 on Cx26 gap junctions. Using a dye transfer assay, we found that an elevated PCO2 of 55 mmHg greatly delayed the permeation of a fluorescent glucose analogue (NBDG) between HeLa cells coupled by Cx26 gap junctions. However, the mutations K125R or R104A abolished this effect of CO2 . Whole cell recordings demonstrated that elevated CO2 reduced the Cx26 gap junction conductance (median reduction 66.7%, 95% CI, 50.5-100.0%) but had no effect on Cx26K125R or Cx31 gap junctions. CO2 can cause intracellular acidification. Using 30 mm propionate, we found that acidification in the absence of a change in PCO2 caused a median reduction in the gap junction conductance of 41.7% (95% CI, 26.6-53.7%). This effect of propionate was unaffected by the K125R mutation (median reduction 48.1%, 95% CI, 28.0-86.3%). pH-dependent and CO2 -dependent closure of the gap junction are thus mechanistically independent. Mutations of Cx26 associated with the keratitis ichthyosis deafness syndrome (N14K, A40V and A88V), in combination with the mutation M151L, also abolished the CO2 -dependent gap junction closure. Elastic network modelling suggests that the lowest entropy state when CO2 is bound is the closed configuration for the gap junction but the open state for the hemichannel. The opposing actions of CO2 on Cx26 gap junctions and hemichannels thus depend on the same residues and presumed carbamylation reaction.

Cx26 keratitis ichthyosis deafness syndrome mutations trigger alternative splicing of Cx26 to prevent expression and cause toxicity in vitro.

The Cx26 mRNA has not been reported to undergo alternative splicing. In expressing a series of human keratitis ichthyosis deafness (KID) syndrome mutations of Cx26 (A88V, N14K and A40V), we found the production of a truncated mRNA product. These mutations, although not creating a cryptic splice site, appeared to activate a pre-existing cryptic splice site. The alternative splicing of the mutant Cx26 mRNA could be prevented by mutating the predicted 3', 5' splice sites and the branch point. The presence of a C-terminal fluorescent protein tag (mCherry or Clover) was necessary for this alternative splicing to occur. Strangely, Cx26A88V could cause the alternative splicing of co-expressed WT Cx26-suggesting a trans effect. The alternative splicing of Cx26A88V caused cell death, and this could be prevented by the 3', 5' and branch point mutations. Expression of the KID syndrome mutants could be rescued by combining them with removal of the 5' splice site. We used this strategy to enable expression of Cx26A40V-5' and demonstrate that this KID syndrome mutation removed CO2 sensitivity from the Cx26 hemichannel. This is the fourth KID syndrome mutation found to abolish the CO2-sensitivity of the Cx26 hemichannel, and suggests that the altered CO-2-sensitivity could contribute to the pathology of this mutation. Future research on KID syndrome mutations should take care to avoid using a C-terminal tag to track cellular localization and expression or if this is unavoidable, combine this mutation with removal of the 5' splice site.

Structural determinants of CO2-sensitivity in the ß connexin family suggested by evolutionary analysis.

A subclade of connexins comprising Cx26, Cx30, and Cx32 are directly sensitive to CO2. CO2 binds to a carbamylation motif present in these connexins and causes their hemichannels to open. Cx26 may contribute to CO2-dependent regulation of breathing in mammals. Here, we show that the carbamylation motif occurs in a wide range of non-mammalian vertebrates and was likely present in the ancestor of all gnathostomes. While the carbamylation motif is essential for connexin CO2-sensitivity, it is not sufficient. In Cx26 of amphibia and lungfish, an extended C-terminal tail prevents CO2-evoked hemichannel opening despite the presence of the motif. Although Cx32 has a long C-terminal tail, Cx32 hemichannels open to CO2 because the tail is conformationally restricted by the presence of proline residues. The loss of the C-terminal tail of Cx26 in amniotes was an evolutionary innovation that created a connexin hemichannel with CO2-sensing properties suitable for the regulation of breathing.

The poor design of clinical trials of statins in oncology may explain their failure - Lessons for drug repurposing.

Statins are widely used to treat hypercholesterolaemia. However, by inhibiting the production of mevalonate, they also reduce the production of several isoprenoids that are necessary for the function of small GTPase oncogenes such as Ras. As such, statins offer an attractive way to inhibit an "undruggable" target, suggesting that they may be usefully repurposed to treat cancer. However, despite numerous studies, there is still no consensus whether statins are useful in the oncology arena. Numerous preclinical studies have provided evidence justifying the evaluation of statins in cancer patients. Some retrospective studies of patients taking statins to control cholesterol have identified a reduced risk of cancer mortality. However, prospective clinical studies have mostly not been successful. We believe that this has occurred because many of the prospective clinical trials have been poorly designed. Many of these trials have failed to take into account some or all of the factors identified in preclinical studies that are likely to be necessary for statins to be efficacious. We suggest an improved trial design which takes these factors into account. Importantly, we suggest that the design of clinical trials of drugs which are being considered for repurposing should not assume it is appropriate to use them in the same way as they are used in their original indication. Rather, such trials deserve to be informed by preclinical studies that are comparable to those for any novel drug.

ABT-737 and pictilisib synergistically enhance pitavastatin-induced apoptosis in ovarian cancer cells.

There is considerable interest in redeploying drugs for use in combination with other oncology therapeutics. The single-agent activity of statins in ovarian cancer has been widely reported, however the drug concentration required to cause cell death is considerably higher than that achieved in patients receiving statin treatment for hypercholesterolemia. Unfortunately, statins can cause myopathy when administered in high doses. One solution to this is to identify drugs that could be used in combination with statins to reduce the dose required and those that may potentially reduce the incidence of adverse side effects. When the BH3 mimetic ABT-737, or the phosphatidylinositol 3-kinase inhibitor pictilisib, were combined with pitavastatin in cell growth assays using Ovcar-3 and Igrov-1 cells, the drug combinations were more effective than pitavastatin alone. In support of this, ABT-737 or pictilisib markedly increased cell death induced by pitavastatin in several ovarian cancer cell lines. The drugs were also synergistic in apoptosis assays. These observations suggested that either BH3 mimetics or pictilisib in combination with pitavastatin could be used in a subset of ovarian tumours, particularly those sensitive to BH3 mimetics, and phosphatase and tensin homolog inhibition, in the treatment of ovarian cancer.

Dietary geranylgeraniol can limit the activity of pitavastatin as a potential treatment for drug-resistant ovarian cancer.

Pre-clinical and retrospective studies of patients using statins to reduce plasma cholesterol have suggested that statins may be useful to treat cancer. However, prospective clinical trials have yet to demonstrate significant efficacy. We have previously shown that this is in part because a hydrophobic statin with a long half-life is necessary. Pitavastatin, the only statin with this profile, has not undergone clinical evaluation in oncology. The target of pitavastatin, hydroxymethylglutarate coenzyme-A reductase (HMGCR), was found to be over-expressed in all ovarian cancer cell lines examined and upregulated by mutated TP53, a gene commonly altered in ovarian cancer. Pitavastatin-induced apoptosis was blocked by geranylgeraniol and mevalonate, products of the HMGCR pathway, confirming that pitavastatin causes cell death through inhibition of HMGCR. Solvent extracts of human and mouse food were also able to block pitavastatin-induced apoptosis, suggesting diet might influence the outcome of clinical trials. When nude mice were maintained on a diet lacking geranylgeraniol, oral pitavastatin caused regression of Ovcar-4 tumour xenografts. However, when the animal diet was supplemented with geranylgeraniol, pitavastatin failed to prevent tumour growth. This suggests that a diet containing geranylgeraniol can limit the anti-tumour activity of pitavastatin and diet should be controlled in clinical trials of statins.

Evolutionary adaptation of the sensitivity of connexin26 hemichannels to CO2.

CO2 readily combines with H2O to form [Formula: see text] and H+ Because an increase of only 100 nM in the concentration of H+ (a decrease of 0.1 unit of pH) in blood can prove fatal, the regulated excretion of CO2 during breathing is an essential life-preserving process. In rodents and humans, this vital process is mediated in part via the direct sensing of CO2 via connexin26 (Cx26). CO2 binds to hemichannels of Cx26 causing them to open and allow release of the neurotransmitter ATP. If Cx26 were to be a universal and important CO2 sensor across all homeothermic animals, then a simple hypothesis would posit that it should exhibit evolutionary adaptation in animals with different homeostatic set points for the regulation of partial pressure of arterial CO2 (PaCO2). In humans and rats, PaCO2 is regulated around a set point of 40 mmHg. By contrast, birds are able to maintain cerebral blood flow and breathing at much lower levels of PaCO2 Fossorial mammals, such as the mole rat, live exclusively underground in burrows that are both hypoxic and hypercapnic and can thrive under very hypercapnic conditions. We have therefore compared the CO2 sensitivity of Cx26 from human, chicken, rat and mole rat (Heterocephalus glaber). We find that both the affinity and cooperativity of CO2 binding to Cx26 have been subjected to evolutionary adaption in a manner consistent with the homeostatic requirements of these four species. This is analogous to the evolutionary adaptation of haemoglobin to the needs of O2 transport across the animal kingdom and supports the hypothesis that Cx26 is an important and universal CO2 sensor in homeotherms.

Altered CO2 sensitivity of connexin26 mutant hemichannels in vitro.

Connexin26 (Cx26) mutations underlie human pathologies ranging from hearing loss to keratitis ichthyosis deafness (KID) syndrome. Cx26 hemichannels are directly gated by CO2 and contribute to the chemosensory regulation of breathing. The KID syndrome mutation A88V is insensitive to CO2, and has a dominant negative action on the CO2 sensitivity of Cx26WT hemichannels, and reduces respiratory drive in humans. We have now examined the effect of further human mutations of Cx26 on its sensitivity to CO2 : Mutated Cx26 subunits, carrying one of A88S, N14K, N14Y, M34T, or V84L, were transiently expressed in HeLa cells. The CO2-dependence of hemichannel activity, and their ability to exert dominant negative actions on cells stably expressing Cx26WT, was quantified by a dye-loading assay. The KID syndrome mutation, N14K, abolished the sensitivity of Cx26 to CO2 Both N14Y and N14K exerted a powerful dominant negative action on the CO2 sensitivity of Cx26WT None of the other mutations (all recessive) had a dominant negative action. A88S shifted the affinity of Cx26 to slightly higher levels without reducing its ability to fully open to CO2 M34T did not change the affinity of Cx26 for CO2 but reduced its ability to open in response to CO2 V84L had no effect on the CO2-sensitivity of Cx26. Some pathological mutations of Cx26 can therefore alter the CO2 sensitivity of Cx26 hemichannels. The loss of CO2 sensitivity could contribute to pathology and consequent reduced respiratory drive could be an unrecognized comorbidity of these pathologies.