Chloride reported missing from brain and extracellular matrix is a suspect

It is exciting when a new physiological technique emerges and leads to results that challenge established concepts. In this issue of The Journal of Physiology, Normoyle et al. (2024) apply their recently developed technique for measuring chloride concentration to brain extracellular space (ECS) and reveal results that are strikingly different from those obtained with classical ion-selective microelectrodes (ISMs). The concentration of extracellular chloride ([Cl−]o) in the ECS has been long understood to be about 120 mM, close to that in the cerebrospinal fluid (CSF). This is because Cl o − ${\mathrm{Cl}}_{\mathrm{o}}^-$ is in contact with CSF at ventricular and pial surfaces, and at the perivascular spaces. This understanding is supported by studies using ISMs. It was surprising therefore when Normoyle and co-workers used an optical imaging technique to show that resting [Cl−]o was less than half that in the CSF or bathing medium. The authors attributed the dramatic reduction in [Cl−]o to the presence of about 50 mM of anionic charge groups attached to the extracellular matrix that extends throughout the ECS. The charge value was inferred from the measured [Cl−]o value and corroborated by analysis of the literature on the matrix composition. The explanation was that the fixed negative charges would enforce local electroneutrality by displacing mobile Cl o − ${\mathrm{Cl}}_{\mathrm{o}}^-$ . This type of Donnan exclusion is seen in cartilage and elsewhere but not so far in ECS. As the authors discuss, this finding may change our understanding of GABA-mediated neuronal inhibition that is based on the transmembrane chloride equilibrium potential. This is relevant also to seizure and oedema mechanisms. Another important consequence of the study is that it estimates the concentration of fixed negative charges in the ECS. With any new finding with conceptual implications, it is necessary to examine the methods. Here the investigators have prepared a chloride-sensitive dye, N(4-AminoButyl)Phenantridinium (ABP), attached it to a 10 kDa dextran molecule (to keep the dye in the ECS) and used fluorescence lifetime imaging (FLIM) to measure [Cl−]o. The technical details are described in prior publications. The validation presented here included formulation of sulfated agarose blocks with fixed negative charge where Donnan exclusion was observed using the ABP–dextran technique. Physiological measurements were made in organotypic hippocampal slice cultures and in vivo. The slice studies included enzymatic digestion to remove some of the negatively charged chondroitin sulfate component of the extracellular matrix. The methods are technically complex, but they appear carefully validated and the results convincing. Nevertheless, the findings raise some issues. A major question is why none of the prior studies with ISMs revealed the drop in [Cl−]o. Since their introduction in the early 1970s, ISMs have been used in hundreds of studies. While relatively few have measured [Cl−]o, three studies are mentioned in the article and there are several more both in brain tissue and in the retina, all made by independent groups, and usually involving ISM measurements of additional extracellular ions. In all cases the resting [Cl−]o measurements were close to those in CSF or bathing medium, in contrast with the results of Normoyle et al. It is useful to focus on one set of ISM-based [Cl−]o measurements, not discussed in the paper, namely those involving spreading depression (SD) and the related condition of ischaemia. During SD, both Kraig & Nicholson (1978) and Hansen & Zeuthen (1981) recorded substantial falls in [Cl−]o from normal baselines along with changes in [Na+]o, [K+]o and [Ca2+]o. The fall in [Cl−]o was consistent with the fall in [Na+]o expected to maintain electroneutrality. There were some ‘missing anions’ that might be attributed to fixed charges but equally the missing anions could have been HCO3− or anionic substances released during SD or ischaemia. A critical fact here is that the ECS volume is diminished during SD and ischemia from about 20% to as small as 5% (Syková & Nicholson 2008, Table 7). Mobile ions can cross cell membranes, but fixed extracellular anions cannot; thus, from a naïve perspective, the concentration of 50 mM of anionic charge postulated by Normoyle et al. could effectively become 200 mM – more than enough to exclude all mobile anions and even exceed the anionic content of CSF. This assumes the anionic sites are uniformly distributed. Normoyle et al. suggest the Cl o − ${\mathrm{Cl}}_{\mathrm{o}}^-$ distribution is heterogeneous. However micro regions where [Cl−]o is similar to the [Cl−] of the CSF do not appear in the in vivo data presented by Normoyle et al. As noted in the article, a future study combining both ABP–dextran and ISMs in both agarose blocks and brain tissue seems important. Probably during peak SD and certainly during the prolonged condition of ischaemia, the ABP–dextran method should not record any Cl o − ${\mathrm{Cl}}_{\mathrm{o}}^-$ if the fixed anions are uniformly distributed. A valuable aspect of combined studies is that they could reveal unsuspected limitations of Cl-ISMs leading to a re-interpretation of some previous studies. The prevailing concept (Nicholson & Phillips 1981) is that an ISM records the ionic concentration in diffusion equilibrium with the ECS in a volume a few micrometres in extent around the tip of the ISM. Is this true in a charged matrix? Another avenue of research suggested by these studies would be a biophysical model incorporating multi-ion Donnan equilibrium accounting for other extracellular ions as well as Cl−. In the modelling context, note that both small monovalent cations and anions (similar in size to Cl−) appear to diffuse through the ECS equally well (Nicholson & Phillips 1981). Is this consistent with a 50 mM fixed anionic charge? The elegant studies of Normoyle et al. may change our concepts of the role of Cl o − ${\mathrm{Cl}}_{\mathrm{o}}^-$ in brain physiology and our ideas about the ECS matrix, and even the interpretation of some ISM data. Yet there are still mysteries to be resolved. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. None declared. Sole author. None.