Membrane proteins are a heterogeneous class of proteins that, as their name suggests, reside in or are associated with cellular membranes. These proteins often fulfill essential biological roles such as the transport of ions and metabolites, signal transduction, propagation of electrical impulses and controlling organellar and cellular shape (1). Consequently, membrane proteins make up a significant fraction of a typical eukaryotic proteome. Computational studies suggest that, for example, the membrane protein content of the human proteome is about 30%, while this is 31% Caenorhabditis elegans and 20% in Drosophila melanogaster (2,3). Membrane proteins are also medically relevant because 60% of these represent human drug targets (5). Given their biological importance it is therefore not surprising that in humans different progressive diseases are associated with dysfunctional membrane proteins (Table 1). In fact, many of these disease-linked membrane proteins contain (missense) mutations that affect their folding, stability or function (4). About 183000 disease-linked mutations in human membrane proteins are currently known, according to the MutHTP database (5).
Cystic fibrosis (CF) represents an established example of a disease related to genetic defects in a membrane protein. CF is a common life-limiting genetic disorder that affects about 70000 people globally with a high prevalence in those of Northern European descent (6,7). CF is observed in one out of ~ 2500 newborns in Europe and in one out of ~ 3500 newborns in the United States. CF is caused by mutations in the CFTR gene (Cystic Fibrosis Transmembrane Conductor), which encodes an anion channel that controls salt, fluid and pH balance in many organs (8). Dysfunction of CFTR strongly impairs the transport capacity of chloride and bicarbonate across the cell membrane of epithelial cells. Not surprisingly, CF is a multiorgan disorder characterized primarily by airway blockage by thickened mucus, which leads to chronic infections of the lung and local airway inflammation, which the major cause of morbidity and mortality, as well as blockage of pancreatic ducts, thereby resulting in pancreatic insufficiency (6,7). CFTR belongs to the superfamily of ABC (ATP-binding cassette) proteins that are found in all biological domains. Unlike other ABC proteins, that typically transport substrates against their concentration, CFTR conducts anions down their concentration gradient. Structurally, CFTR is a single polypeptide comprising two transmembrane domains (TMDs) with six membrane-spanning helices, which together make up the ion channel, and two cytosolic nucleotide binding domains (NBDs), that bind ATP (8). The overall architecture of CFTR is, therefore, similar to other ABC proteins. The domains are arranged into two pseudo-symmetrical halves connected by a so-called cytoplasmic regulatory domain (RD) of about 200 residues. The RD is unique to CFTR and controls the activity of CFTR. Gating of the CFTR channel is regulated through phosphorylation of the RD, which activates the channel, and ATP binding by the NBDs, that drives opening and closing of the channel pore (8). More than 2000 mutations in the CFTR gene have been found in CF patients and of these 200 are known to be pathologically relevant. Despite this large number of mutations, only one mutation (deletion of phenylalanine 508) is responsible for the majority of CF cases globally (6,7). About 82% of the CF population carry the ΔF508 variant. This mutation affects both the folding and thermal stability of the protein, resulting in its increased degradation, and thereby explaining the virtual absence of the mature protein within the cell membrane (6,7). Recently, small-molecule drugs have been developed that act as CFTR modulators through directly binding to the protein in order to target the primary molecular defect. Five classes of modulators have been described that act on CFTR via different mechanisms, namely: potentiators, correctors, stabilizers, read-through agents and amplifiers. Potentiators improve channel activity, while correctors increase the amount of CFTR in the plasma membrane (8). Recently, detailed structures of human CFTRΔF508 in complex with two correctors were reported (9). Here, I will discuss these structures as well as how these correctors rescue this CFTR mutant.
Overview of CFTR modulators
In recent years novel CF therapies have emerged that are not focused on treating the symptoms of the disease but rather rescue CFTR’s channel function. These novel therapeutic approaches are based on small molecule drugs (termed CFTR modulators) that directly target the underlying dysfunctions caused by specific CF mutations (10). High-throughput screening of chemical libraries in combination with different cell models is typically used for the identification of CFTR modulators. Currently, four CFTR modulators are commercially available and are successfully used to improve the short and long-term clinical results of CF patients. These compounds enhance or restore the expression, function and stability of dysfunctional CFTR. Based on their mode of action, CFTR modulators are classified into five main groups – potentiators, correctors, stabilizers, read-through agents and amplifiers. Of these, three correctors (lumacaftor, tezacaftor, elexacaftor) and one potentiator (ivacaftor) are currently FDA-approved (10). The chemical structure of three FDA-approved CF modulators are shown in figure 1. Potentiators and correctors act on CFTR via different mechanisms. Potentiators restore or improve channel activity, thereby allowing CFTR-dependent conductance of anions. Ivacaftor, for example, improves CFTR function through stimulating decoupling between ATP hydrolysis and opening and closing of the channel. Recently, structural and biophysical evidence has been provided suggesting that ivacaftor binds at the interface between the two TMDs (10). Correctors are small molecule drugs that rescue CFTR folding, processing, and membrane transport. This increases the amount of membrane-embedded CFTR. Although these compounds act through different mechanisms, they typically improve the conformational stability of CFTR during folding within the ER lumen and represent therefore small molecule chaperones (10). CFTR correctors are classified into different groups based on their mode of action. Correctors from different groups act through unrelated mechanisms and can therefore be combined to synergistically support CFTR folding. Lumacaftor, for example, restores the folding and function of the ΔF508 variant in vitro but its binding site on CFTR is not clear yet. Despite these encouraging in vitro results, a clinical study showed that treatment of CF patients with lumacaftor alone did not improve lung function. Combining lumacaftor and ivacaftor proved more successful. This combination is now used in the clinic to improve CFTR function, slow lung deterioration and reduce other symptoms in ΔF508 patients. In addition to ΔF508, lumacaftor was also able to restore CFTR function of variants carrying other mutations in vitro. Tezacaftor is an optimized corrector which is based on the structure of lumacaftor and exhibits improved pharmacokinetic properties as well as reduced side effects. Lumacaftor and tezacaftor belong to the same group, type 1 correctors, and thus act through a similar mechanism. Like lumacaftor, tezacafor is clinically used in combination with ivacaftor for the treatment of ΔF508 patients. However, the dual combinations lumacaftor/ivacaftor and tezacaftor/ivacaftor only show a modest improvement in ΔF508 patients and, therefore, high-throughput screens are employed to identify novel, more promising correctors from massive chemical libraries (10).
Biological features of CFTR
Human CFTR is a membrane protein of 1480 residues and with a mass of about 186 kDa. Following translation by cytosolic ribosomes, CFTR is subjected to several posttranslational maturation steps in the endoplasmic reticulum as well as Golgi complex. After this, it is transported to the cell membrane of epithelial cells where it functions as a chloride channel. CFTR is a member of the superfamily of ATP binding cassette (ABC) proteins, which represents the largest family of membrane-embedded transporters (8). ABC proteins are ubiquitously present in all cells and they are involved in the transport of a variety of substrates into and out of cells. Substrate transport by ABC proteins is driven by ATP hydrolysis. Interestingly, CFTR is the only member of the ABC family that conducts anions down their concentration gradient in contrast to other ABC proteins that typically transport substrates against their concentration. Cloning of the CFTR gene in 1989 was the first step in understanding the molecular basis of CF (11,12,13). This allowed its sequencing, which revealed the domain organization of the protein. It is a single polypeptide, consisting of two pseudo-symmetrical halves that are made up of two TMDs and two cytosolic NBDs (figure 2 adapted from 8). The N and C-terminus are localized to the cytoplasm. The overall arrangement of CFTR is therefore similar to that of other ABC proteins. The TMDs are composed of six membrane-spanning helices that form the ion conduction pathway, while the NBDs are responsible for binding and hydrolysis of ATP. However, only one NBD is catalytically active. The transmembrane helices are closely intertwined with each other, forming two double six-helix bundles, while they are also connected with the NBDs. The TMDs are connected to the NBDs via small helices, so-called coupling helices. These helices are conserved at the structural level but not the level of the primary sequence. These helices transfer conformational changes within the NBDs to the TMDs and are therefore crucial for the activity of ABC proteins. It is therefore not surprising that many CF-causing mutations are located at the NBD/TMD interface, including ΔF508. The two halves of CFTR are connected by a regulatory domain of about 200 residues, that is only present in CFTR.
Unlike other ion channels, CFTR function is not regulated by changes in the concentration of the translocated substrate ion. Rather, the R domain controls the activity of CFTR but how does it regulate CFTR activity? This domain contains 19 potential phosphorylation sites and phosphorylation of the R domain by protein kinase A is required for channel opening. Following phosphorylation, ATP binding opens the channel, whereas ATP hydrolysis closes it. CFTR activity is mainly controlled by phosphorylation of the R domain owing to the high and relatively constant cellular concentration of ATP (8). Specifically, it has been shown that dimerization of the NBDs is required for channel opening and this only occurs after phosphorylation of the regulatory domain. The dephosphorylated regulatory domain is located between the two NBDs, thereby preventing their dimerization. Upon phosphorylation, this domain is released from its inhibitory position, allowing dimerization of the NBDs and subsequent opening of the ion channel. For ABC proteins, two different structural conformations for the TMDs are typically observed, namely: the outward-facing (OF) and the inward-facing conformation (IF). In the OF, the external gate at the extracytoplasmic region is open, while in the IF, the internal gate at the cytosolic side is open. Thus, in ABC proteins access to the substrate translocation pathway is regulated at both ends. Unlike other ABC proteins, the ion conduction pathway of CFTR forms a continuous channel permeable for anions primarily chloride and bicarbonate (8). Although larger anions such as SCN- and nitrate can also be accommodated in the CFTR channel, they typically bind with high affinity within the channel. These ions are therefore translocated at a relatively low rate.
Structural features of human CFTR bound to folding correctors
Over the past years, detailed structures of human and zebrafish CFTR in different conformations were obtained by cryo-EM. These established its overall three-dimensional architecture, representing an inverted “V”, which is similar to the structure of other ABC transporters (14,15,16,17). Moreover, cryo-EM structures of human CFTR bound to either lumacaftor or tezacaftor at atomic resolution were reported recently (9). To obtain these structures, human CFTR was recombinantly expressed in HEK cells. Subsequently, cells were detergent solubilized and CFTR was affinity purified followed by size exclusion chromatography. Purified CFTR was incubated with lumacaftor or tezacaftor, frozen and subjected to cryo-EM. The structure of unphosphorylated, ATP-free CFTR and phosphorylated, ATP-bound CFTR in complex with lumacaftor were obtained at 3.9 and 2.7 Å, respectively (9). These are shown in figure 3 in ribbon and surface representation with lumacaftor as purple spheres, tezacaftor as blue spheres and ATP as green spheres, respectively. In the unphosphorylated, ATP-free form (top panels), CFTR is represents an inverted V, a structure that is typical for an ABC transporter in its inward-facing conformation. Each TMD is made up of a six-helix bundle with the helices segregating in the cytoplasm. The two NBDs are separated with the R-domain (in blue) wedged in between and are connected by a cytosolic coupling helix to the TMDs. The NBDs consist of two subdomains, namely: the head and the tail. The head comprises the nucleotide binding core and is made up β-sheets and a central helix, while the tail is an α-helical subdomain that contains the ABC-specific signature motif. Lumacaftor is observed at TMD1. In the phosphorylated, ATP-bound form (left bottom panel), the NBDs dimerized and contain two ATP molecules. Moreover, they are located closely to the TMDs, forming an elongated molecule. In contrast to other ABC transporters, the transmembrane helices of CFTR do not split into a wing-like structure but remain tightly packed to shield the ion channel. In this conformation, lumacaftor is also observed at TMD1, indicating that lumacaftor binds to CFTR in both conformations within the binding site located in TMD1 (9). The hydrophobic headgroup of lumacaftor sticks into the hydrophobic pocket that is formed by transmembrane helix (TMH) 1, 2, 3 and 6, while its polar moiety extends out of the binding pocket (right top and left bottom panel). Thus, lumacaftor inserts deep into a hydrophobic pocket of TMD1. Additionally, the structure of phosphorylated, ATP-bound CFTR in complex with tezacaftor (a structural analog of lumacaftor) was determined by cryo-EM at 3.8 Å resolution (9). This structure is shown in the right bottom panel of figure 3 and reveals CFTR as an elongated molecule with dimerized NBDs. Tezacaftor is, moreover, bound to TMD1 at the same location as lumacaftor. Like lumacaftor, the hydrophobic head group of tezacaftor is located inside the hydrophobic pocket with its hydrophilic part outside of the pocket.
A close up of the corrector binding site is provided in figure 4, showing lumacaftor (as purple spheres in the top left panel) and tezacaftor (as blue spheres in the right top panel) in the hydrophobic pocket of TMD1. Lumacaftor mainly interacts through van der Waals interactions with CFTR. The residues that facilitate molecular recognition of lumacaftor are shown in the bottom left panel with lumacaftor as purple ball and sticks. Lumacaftor interacts with CFTR through a salt bridge with Lys68. The polar moiety that extends outside of the binding pocket interacts with Ile70, Asn71 and Arg74 of TMH1 and Leu365 and Ile368 of TMH6, thereby linking these TMHs together. The pocket-lining residues that facilitate recognition of tezacaftor are shown in the bottom right panel with tezacaftor as blue ball and sticks. Tezacaftor interacts with the same set of pocket-lining residues as lumacaftor, although it does not form a salt bridge with Lys68, but a hydrogen bond with Arg74 instead. Furthermore, tezacaftor interacts with fewer residues of TMH1 in contrast to lumacaftor. Mutagenesis of these pocket-lining residues reduced the binding of either lumacaftor and tezacaftor, thereby reducing their ability to rescue CFTRΔF508. Collectively, these results emphasize the functional significance of the lumacaftor/tezacaftor binding site. Both correctors thus bind to the same site in TMD1 and thereby promote the folding of TMD1 but not of other domains. Consequently, corrector binding does not affect the overall structure of CFTR. The structural data presented in figure 4 suggest that binding of lumacaftor or tezacaftor link TMHs 1, 2, 3 and 6, which stabilizes the partially folded TMD1. Hence, CFTRΔF508 is rescued by these correctors from the increased stability of TMD1. This, in turn, shows that the corrector binding site is the site of action to restore CFTR folding.
Ion channels, aquaporins, G-protein-coupled receptors and drug resistance proteins are examples of membrane proteins that fulfill crucial cellular roles, ranging from the transport of ions to cell signaling and the efflux of toxic compounds (1). Owing to their essential function, it is not surprising that membrane proteins make up 20-30% of a typical eukaryotic proteome (2). Moreover, membrane proteins are also medically relevant because 60% of them act as drug targets and dysfunctional membrane proteins are the underlying cause of different disorders and diseases, including cancer and neurodegenerative diseases. The aberrant activity of membrane proteins is often caused by mutations that affect their folding, stability or function (4). About 183000 disease-linked mutations in human membrane proteins are currently known! (5). CF is an established example of a genetic disease caused by mutations in the gene encoding the CFTR membrane protein. CF is a common life-limiting heritable disorder with a high prevalence in those of Northern European descent (6,7). CFTR is a chloride channel that resides the plasma membrane of epithelial cells and regulates salt, fluid and pH balance. More than 2000 mutations in the CFTR gene have been found and 200 of these are known to be pathologically relevant. Despite this large number of mutations, only one mutation (deletion of phenylalanine 508) is responsible for the majority of CF cases globally (6,7). Recently, novel CF therapies have emerged that are not focused on treating the symptoms of the disease but rather rescue CFTR’s channel function. These novel therapeutic approaches are based on small molecule drugs (termed CFTR modulators) that directly target the underlying dysfunctions caused by specific CF mutations (10). Based on their mode of action, CFTR modulators are classified into five main groups – potentiators, correctors, stabilizers, read-through agents and amplifiers. Currently, four CFTR modulators are FDA-approved and are successfully used to improve the short and long-term clinical results of CF patients. These include three correctors (lumacaftor, tezacaftor, elexacaftor) and one potentiator (ivacaftor) (10). CFTR is a so-called ABC protein, which comprise a large group of membrane-embedded transporters that are present in domains of life (8). These proteins are involved in the transport of a variety of substrates into and out of cells, which is catalyzed by ATP hydrolysis. CFTR is unique amongst ABC transporters in that conducts anions down their concentration gradient in contrast to other ABC proteins. CFTR is a single polypeptide with a domain arrangement similar to other ABC transporters, namely: two TMDs and two cytosolic NBDs (8). The TMDs are composed of six membrane-spanning helices that form the ion conduction pathway, while the NBDs are responsible for binding and hydrolysis of ATHowever, only one NBD is catalytically active. In addition, a regulatory domain of about 200 residues that connects the two halves of the protein and is unique to CFTR. This domain controls the activity of CFTR through its phosphorylation by protein kinase A that is required for channel opening, while, following phosphorylation, ATP binding opens the channel and ATP hydrolysis closes it (8). Structurally. CFTR exhibits an overall three-dimensional architecture that resembles an inverted “V”, similar to the structure of other ABC transporters (14,15,16,17). Moreover, cryo-EM structures of human CFTR bound to either lumacaftor or tezacaftor at atomic resolution were reported recently (9). These structures are presented in figure 3 and 4 and show that lumacaftor and tezacaftor bind at the same site within TMD1, which stabilizes only the partially folded TMD1 through linking TMHs that make up the corrector binding site. Based on these results a mechanistic model for lumacaftor and tezacaftor was proposed (figure 5 adapted from 9).
CFTR is synthesized cotranslationally by cytosolic ribosomes and the nascent polypeptide is inserted into the endoplasmic membrane. The individual CFTR domains fold individually once they are synthesized, which is followed by assembly of its mature structure. The N-terminal TMHs of TMD1 are produced early during the biogenesis of CFTR and adopt a structure that is thermodynamically unstable. Binding of lumacaftor or tezacaftor (yellow sticks) links the TMHs that form the corrector binding site, thereby stabilizing TMD1 in the endoplasmic membrane. This makes it less susceptible to cellular degradation and increases the lifetime of TMD1, allowing proper folding of the other domains. Conceivably, the increased lifetime of TMD1 partially rescues folding defects in other parts of CFTR such as CFTRΔF508. It is expected that the mechanistic insight offered by these novel CFTR structures in complex with lumacaftor and tezacaftor will facilitate the design of improved type 1 correctors.
1. von Heijne G. (2007). The membrane protein universe: what's out there and why bother? J Intern Med. 261: 543-57.
2. Almén MS, Nordström KJV, Fredriksson R, Schiöth HB. (2009). Mapping the human membrane proteome: a majority of the human membrane proteins can be classified according to function and evolutionary origin. BMC Biol. 7:50.
3. Krogh A, Larsson B, von Heijne G, Sonnhammer EL. (2001). Predicting transmembrane protein topology with a hidden Markovmodel: application to complete genomes. J Mol Biol. 305:567-580.
4. Sanders CR, Myers JK. (2004). Disease-related misassembly of membrane proteins. Annu Rev Biophys Biomol Struct. 33: 25-51.
5. Kulandaisamy A, Binny Priya S, Sakthivel R, Tarnovskaya S, Bizin I, Hönigschmid P, Frishman D, Gromiha MM. (2018). MutHTP: mutations in human transmembrane proteins. Bioinformatics. 34: 2325-2326.
6. Cutting GR. (2015). Cystic fibrosis genetics: from molecular understanding to clinical application. Nat Rev Genet. 16: 45-56.
7. Elborn JS. (2016). Cystic fibrosis. Lancet. 388: 2519-2531
8. Csanády L, Vergani P, Gadsby DG. (2019). Structure, Gating, and Regulation of the CFTR Anion Channel. Physiol Rev. 99: 707-738.
9. Fiedorczuk K, Chen J. (2022). Mechanism of CFTR correction by type I folding correctors. Cell. 185: 158-168.
10. Lopes-Pacheco M. (2020). CFTR Modulators: The Changing Face of Cystic Fibrosis in the Era of Precision Medicine. Front Pharmacol. 10: 1662.
11. Kerem B, Rommens JM, Buchanan JA, Markiewicz D, Cox TK, Chakravarti A, Buchwald M, Tsui LC (1989). Identification of the cystic fibrosis gene: Genetic analysis. Science 245: 1073–1080.
12. Riordan JR, Rommens JM, Kerem B, Alon N, Rozmahel R, Grzelczak Z, Zielenski J, Lok S, Plavsic N, Chou JL, et al. (1989). Identification of the cystic fibrosis gene: Cloning and characterization of complementary DNA. Science 245: 1066–1073.
13. Rommens JM, Iannuzzi MC, Kerem B, Drumm ML, Melmer G, Dean M, Rozmahel R, Cole JL, Kennedy D, Hidaka N. et al. (1989). Identification of the cystic fibrosis gene: Chromosome walking and jumping. Science 245: 1059–1065.
14. Liu F, Zhang Z, Csanády L, Gadsby DC, Chen J. (2017). Molecular Structure of the Human CFTR Ion Channel. Cell. 169: 85-95.e8.
15. Zhang Z, Liu F 2, Chen J. (2018). Molecular structure of the ATP-bound, phosphorylated human CFTR. Proc Natl Acad Sci U S A. 115: 12757-12762.
16. Zhang Z, Chen J. (2016). Atomic Structure of the Cystic Fibrosis Transmembrane Conductance Regulator. Cell. 167: 1586-1597.e9.
17. Zhang 1, Liu F, Chen J. (2017). Conformational Changes of CFTR upon Phosphorylation and ATP Binding. Cell. 170: 483-491.e8.