How the brain senses CO2

Metabolism requires O2 and produces CO2 as a by-product. The amount of dissolved CO2 controls the acidity (pH) of our body fluids including blood. Many physiological processes are sensitive to pH, thus efficient control of the amount of dissolved CO2 in blood is a critical homoeostatic function. Special mechanisms exist to measure the amount of dissolved CO2 in blood ?if there is too much we breathe faster to drive off the excess, if there is too little we breathe less frequently. While there is evidence that CO2 is detected indirectly as pH, we have discovered a new molecular mechanism for the direct detection of CO2 ?CO2 binds to a type of channel called connexin 26 (Cx26), and in doing so causes it to open and release a chemical called ATP that activates neurons. Cx26 is sensitive to CO2 in exactly the physiological range required ?our normal levels of dissolved CO2 are at the mid-point of channel activation. Cx26 can thus respond to both increases and decreases in dissolved CO2. Cx26 is present in the correct areas of the brain to help control breathing. Our programme seeks to develop the genetic tools to analyze with great rigour the contributions of Cx26 to the measurement of CO2 and the control of breathing. Use of these tools will not only enable us to establish the causal link between Cx26 and behaviour, but also to determine the tissues, regions and cell types that are important for the detection of CO2. We also wish to understand the mechanism by which CO2 binds to the channel to change its conformation and cause it to open. We shall mix and match portions of CO2-sensitive and non-CO2-sensitive connexins to endow a previously non-sensitive connexin with CO2 sensitivity. This will tell us which part of the protein is important. We shall then identify the precise amino acid involved by mutating single amino acids in the critical region of the molecule. In parallel with this we shall use an analytical technique called NMR spectroscopy to directly and definitively test one possible way that CO2 could bind to the protein. Our proposal has the potential to transform understanding of how the brain senses CO2. This is likely to be important in understanding how CO2-sensing may be altered during pathologies such as congestive heart failure and chronic obstructive pulmonary disease.

We seek to understand the mechanisms by which brain senses the partial pressure of CO2 (PCO2) in arterial blood. Chemosensory reflexes regulate breathing to keep arterial PCO2 (and PO2) within tightly defined limits and thus maintain acid-base balance. Although disorders of chemosensory mechanisms are potentially life threatening, especially in the context of other pathologies such as congestive heart failure, knowledge of their underlying mechanisms remains surprisingly incomplete. We have recently discovered a new molecular transducer for CO2 chemoreception ?the connexin 26 (Cx26) hemichannel. During elevated arterial PCO2 (hypercapnia), increased CO2 binding to Cx26 causes it to open more and hence release more ATP. This ATP excites the respiratory network and causes the adaptive increase in breathing to hypercapnia. Cx26 is also sensitive to decreases in PCO2; closure of the channel, resulting in reduced ATP release, decreases the respiratory drive during hypocapnia and consequently helps to maintain PCO2. Cx26 is widely distributed throughout the brain in the subpial astrocytes and leptomeninges. This raises two important and major questions. Which are the key tissues and cell types for respiratory chemosensing ?the astrocytes or the leptomeninges? How can gating of a gap junction hemichannel be directly altered by CO2? We shall address these two problems in a coordinated fashion. Firstly, we shall develop the transgenic tools to delete Cx26 in selected cell types and regions. We shall then examine the effect on responses to CO2. This will provide definitive causal evidence linking Cx26 to the behavioural adaptive reflexes in responses to changes in PCO2 and will allow us to determine the most critical cellular components and regions for central CO2 chemosensing. Secondly, we shall make chimeric constructs by inserting structural motifs of Cx26 (CO2-sensitive) into a non-CO2-sensitive homologous connexin to identify the critical domains of the molecule. We shall complement the identification of key structural components by loss-of-function mutagenesis. We shall definitively test our favoured hypothesis that CO2 alters Cx26 hemichannel gating via carbamylation of a lysine residue by utilizing 15N-13C NMR spectroscopy to demonstrate directly the existence of the hypothesized carbamate group. The NMR studies will assist the design of the mutagenesis strategy and complement the functional evaluation of the mutant channels: we expect the existence of a carbamate group to be present in those mutant channels that are CO2-sensitive and absent in those that are not.