Abstract
Volume-regulated anion channels (VRACs) are key players in cell volume regulation and are involved in a variety of physiological processes. VRACs are composed of LRRC8 proteins, a protein family with five paralogs, termed LRRC8A, B, C, D and E. In a cellular environment, functional VRACs assemble as heteromeric channels containing LRRC8A and at least one other LRRC8 protein. Homomeric LRRC8A proteins assemble into hexameric channels but show reduced activity. The high-resolution structure of homomeric LRRC8A channels has revealed a modular assembly with a transmembrane spanning pore domain (PD) and intracellular leucin rich repeat (LRR) domains. The function of the different channel domains and the molecular activation mechanism of VRACs are not yet fully understood. Furthermore, the subunit stoichiometry and arrangement in heteromeric channels has remained elusive as no high-resolution structures of heteromeric channels were available at the beginning of this work. The resolution of these open question has thus become a major focus of this thesis. To provided tools for the study of the structure and function of VRACs, five synthetic nanobodies (sybodies) were generated that specifically bind to the intracellular LRR domains of LRRC8A. The sybodies act as modulators of VRAC activity, by either inhibiting or potentiating both, endogenous VRACs and homomeric LRRC8A channels. Homomeric LRRC8A channels were structurally analyzed in complex with these five sybodies by cryogenic electron microscopy (cryo-EM). Different binding epitopes were identified with the inhibitory sybodies binding to each of the six subunits on the accessible convex side of the LRR domains and the potentiating sybodies binding to every second LRR domain on the more buried concave side of the domains. While the inhibiting sybody Sb1 leads to a rigidification of the LRR domains in an arrangement that was originally observed in apo-LRRC8A channels, binding of the other two inhibiting and the two potentiating sybodies leads to a rearrangement of the LRR domains. Binding of the potentiating sybodies additionally increases the mobility of the LRR domains and the conformational changes in this region are transferred to the pore domain, probably resulting in increased channel activity. The effect of heteromerization on channel function was investigated in channels containing LRRC8A and LRRC8C subunits. To gain a better understanding of the building blocks of these channels, the structure of the non-functional homomeric LRRC8C channel was first determined by cryo-EM. The homomeric LRRC8C channels assemble into heptamers with reduced subunit contacts. To study the subunit stoichiometry in heteromeric LRRC8A/C channels, an absolute target quantification approach using LC-MS/MS was applied. The ratio of LRRC8A to LRRC8C subunits in overexpressed A/C channels remained stable at approximately 2:1, whereas endogenous A/C channels contain more A subunits with a ratio of 3.1:1. The 2:1 ratio 2 of overexpressed channels was confirmed in the first high-resolution structure of a heteromeric LRRC8A/C channel using Sb1 as a fiducial marker for LRRC8A subunits. The channel shows only one predominant arrangement with two adjacent pairs of A subunits and one pair of C subunits. While the LRR domains of the A-subunits arrange into the tightly interacting pairs already observed in the homomeric LRRC8A channels, LRR domains of the C subunits remain flexible and are therefore not resolved in the cryo-EM structure. In an LRRC8A/C channel without fiducial markers, the LRR domain of one C subunit is stabilized and shows a conformation comparable to its arrangement in homomeric LRRC8C channels. Incorporation of LRRC8C subunits into the tightly interacting LRRC8A subunits leads to the disturbance of subunit interactions and thus likely increase their mobility in the hexameric channel. As observed for the binding of potentiating sybodies, an increased mobility of the LRR domains and the transfer of this movement into the pore domain is presumably associated with channel activation. An increase in channel mobility was also observed in studies of two LRRC8C proteins with distinct mutations in the hinge region between the pore and the LRR domains, one leading to a conservative replacement of an amino acid, the second to the loss of the cytoplasmic domain due to a premature stop codon. These mutations were found in two patients, both suffering from a multisystemic disorder with a wide variety of different symptoms. I was able to show that both mutated LRRC8C proteins were biochemically stable and have retained their ability to assemble with LRRC8A, comparable to wild-type LRRC8C. Structural analysis of the homomeric assemblies of the two mutated LRRC8Cs revealed a similar conformation to the wild-type channels, although with strongly increased subunit mobility. Co-expression of mutated LRRC8C subunits with LRRC8A in HEK293 cells resulted in enhanced channel activity compared to co-expression of wild-type A and C, as demonstrated by functional studies using patch-clamp electrophysiology. This increase in channel function caused by the mutations was not strongly apparent in the functional measurements in fibroblast cells isolated from one of the patients, which can presumably be attributed to the expression of other LRRC8 paralogs in the primary cells, masking the gain-of-function effect of the mutated LRRC8C. The three different approaches used in this study to investigate VRAC activation all contributed to a common picture of relevant properties underlying this process. Manipulation of channel function by targeting the LRR domains, incorporation of other LRRC8 paralogs in heteromeric channels, and the introduction of activating mutations in LRRC8C all lead to an increase in subunit mobility, suggesting this property to be a critical part of the VRAC channel activation mechanism.