Even though CFTR belongs to the ABC transporter superfamily, it has evolved to function as a channel, which makes it a challenge to compare the similarities and contrast the differences with the general ABC gating mechanism. The ATP Switch model proposed by Higgins and Linton describes how ABC transporters can couple ATP catalysis (ATP binding and hydrolysis) to gating and provides a good starting point for this investigation.[11
] As CFTR is a chloride channel, the switch model must be adapted to accommodate the structural and functional differences. Finally, in order to make a balanced analysis of the basic ABC transporter mechanism and how CFTR fits this, it would be useful to refer to other well-studied transporters, including HlyB and Pgp.
The ATP switch model
The ATP Switch model is based on the conformational changes between an open and closed dimer driving the ABC gating cycle, hence making the NBD pivotal to the mechanism. An energy-converting device, called an NBD, acts as an engine for driving biological processes by using ATP as a source of potential energy. Functional diversity is attributed to the varied TMDs, which are linked to the NBD engine to drive specific transport mechanisms. The general structure and function of NBDs are highly conserved; hence, the NBD dimer-centered ATP Switch mechanism is a good starting point for a general ABC gating mechanism. The most suitable approach to compare and contrast CFTR with the ATP switch model is perhaps by tackling it in a step-by-step fashion [Figures and ].
Initiation of the transport cycle. ATP Switch model requires ligand binding for the initiating step, which acts as an inbuilt latch mechanism. The cycle is permitted only if there is a ligand to transport, thereby preventing futile catalysis in the presence of high cellular levels of ATP.[11
In CFTR, ATP at the NBD2 site functions as both the ligand and the hydrolysable energy source.[8
] How can futile catalysis be prevented in CFTR despite the high cellular levels of ATP “hydrolysable ligand”? The answer lies in the preceding regulatory steps, unique to CFTR, that are outside of the gating cycle . This allows Cl- concentrations to be sensed indirectly and the information transmitted through complex signaling pathways ending in phosphorylation of the R domain and subsequent ATP binding at the NBD1 site.[8
In the ATP Switch model, the ligand binds to the TMDs, which transmits conformational changes to the NBDs that increase their affinity to ATP. In CFTR model, the ATP “hydrolysable ligand” binds directly to the NBD, therefore skipping the additional step that involves energy-dependent conformational changes. This may in part explain the reduced energy requirements of CFTR gating, which require only a single catalytically active NBD2 site.
The ABCC family of transporters, which include CFTR, have asymmetric NBDs with stably bound ATP at NBD1 and catalytically active NBD2 driving the gating cycle. Structural asymmetry is demonstrated by the 27% homology between NBD1 and NBD2 , compared with almost identical domains of bacterial transporters.[8
] Function asymmetry is demonstrated by photolabeled ATP and vanadate trapping experiments.[17
] ABC transporters need a change in affinity at the ligand-binding site to facilitate ligand binding and release, which involves energy-consuming conformational changes. CFTR has a stable ionic pore in the open state, which may explain in part the reason for a single catalytically active NBD2 site.
Main features of this step are formation of the closed NBD dimer, subsequent transmission of conformational changes to the TMDs, and finally ligand transport.[11
] These basic functions are shared by the ATP Switch model and CFTR.
Conformational changes at the TMDs are used for ligand transport in normal ABC transporters or the opening of the TMD anion pore in CFTR. This functional diversity of TMDs is reflected in their structural and mechanistic variations, although this does not detract from underlying similarity of all ABC proteins being driven by highly conserved biological engines, the NBDs.[1
The “power stroke” is a defining step in a process and for the ATP Switch model, it is the transport of the ligand across the lipid membrane. Mechanical energy for this step comes from the formation of a closed NBD dimer, which transmits the necessary conformational changes to the TMDs for ligand transport.[18
] It appears that CFTR cannot be viewed to have such a distinctive power stroke step as it functions as a channel to facilitate the flow of many chloride ions over a prolonged period. However, opening of the channel can be viewed as the major stroke in the gating cycle and therefore ATP binding, which powers this step, can then be interpreted as the energy input for this “power stroke.”
ABC transporters require the ligand-binding site to switch from inside-facing high affinity in the basal state to outside-facing low affinity after the translocation step in order to allow the release of the ligand on the other side of the membrane.[10
] CFTR on the other hand does not have a chloride-binding site as such, although positively charged residues throughout the pore attract chloride ions and facilitate rapid flow through the pore.[19
] This lack of a chloride-binding site, that would need energy-dependent conformational changes to modify its affinity, may be part of the reason behind CFTR requiring only a single catalytically active NBD2 site for the gating cycle.
The dimer is at the center of the ATP Switch mechanism and it forms by the rigid-body rotation of the main nucleotide-binding site (Walker A and B motifs) on one NBD around the complimentary nucleotide-binding site (LSGGQ signature motif) on the second NBD.[20
] Once the two ATPs are bound, they provide the “molecular glue” at the dimer interface for maintaining the head-to-tail conformation.[21
] Even though the CFTR protein has an asymmetric dimer, the highly conserved nature of NBDs dictates that dimer formation is fundamental to gating function of all ABC proteins, with CFTR being no exception .
The main feature of this step is ATP hydrolysis driving the disassociation of the dimer, which initiates the formation of an open dimer state. The associated conformational changes are transmitted reciprocally from NBDs to TMDs and facilitate the return to the basal state.
The ABCC family, in which CFTR is a prominent member, have branched in the tree of ABC protein development to posses only a single catalytically active NBD2 site with the gating cycle driven by the catalysis of a single ATP molecule.[8
] Although this is a variation of the standard mechanism, it does not seem to conflict with the formation and dissociation of the dimer, which is essential to the ATP Switch model.
ATP hydrolysis initiates the process leading to the restoration of the protein to its basal state, but it is important to note that this site does not restore the channel directly. Studies with vanadate on Pgp and orthovandate on CFTR have shown that these molecules can replace Pi
after the hydrolytic step to form a non-reactive NBD dimer.[17
] The hydrolyzed ATP molecule is replaced with ADP-vanadate to leave the protein in a transitional active state. This active state has been shown by Basso and colleagues to be particularly apparent when CFTR is treated with orthovanadate as closure of the channel is significantly delayed and this may imply that a subsequent step such as ADP release or even ligand binding is required to fully restore the protein.
The essential aspect of this step, common to both models, is the restoration of the transporter to its resting state. This process is quite complex, involving many individual steps and is currently not well understood.
The sequential release of the catalysis products, Pi
followed by ADP, from the binding pocket facilitates the disassociation of the NBD dimer. This in turn transmits conformational changes to the TMDs that are restored to their resting state. The vanadate experiments mentioned above helped elucidate the order in which the products of ATP catalysis are released,[17
] and suggests that the release of ADP may be the final step in resetting the transporter.
Repulsion of the two products, ADP and Pi
, within the binding pocket is one explanation for the release of the catalysis products at the end of the cycle,[20
] although this may not be the case for all ABC proteins due to their diversity. Studies of MutS have shown that ADP only dissociates from the NBD and is replaced by ATP upon substrate binding.[23
] Further studies are required to elucidate the exact mechanistic steps needed to reset the CFTR channel to its basal state.
Final words on cystic fibrosis transmembrane conductance regulator and the ATP switch model
In comparing the ATP Switch and CFTR models, it is obvious that despite the differences, they are fundamentally quite similar. The ATP Switch model attempts to encompass the diversity of transporters by relating back to the highly conserved NBD characteristics of this superfamily, therefore allowing various transporters to be grouped together under its umbrella mechanism.
It seems to me that the ATP Switch model is comparable with the CFTR mechanism in many aspects. CFTR gating is based on ATP binding, driving the opening of the channel followed by hydrolysis to close it, which the associated formation and dissociation of the dimer, respectively.[8
] This dimer-based perspective is the foundation of the ATP Switch mechanism, proposed by Higgins and Linton[11
] , and naturally explains the naming of the model in terms of an on-off dimer “switch.”
Considering the view that “early in evolution an engine called a nucleotide - binding domain (NBD) was created
” and to this day has a highly conserved sequence makes it likely that the gating mechanism has also been conserved.[1
] The ability of the engine to power a diverse range of processes comes from its coupling to various TMDs, which show high sequence variation. In conclusion, it is likely that there is a general gating mechanism applicable to the ABC superfamily and the ATP Switch model fits many of the studies on CFTR and other ABC transporters.
Evaluating the ATP switch model
The ATP Switch model exemplifies much of the current accepted data on ABC transporters; however, it would be of use to examine a few other well-known transporter mechanisms. First, these further mechanisms should be beneficial for analyzing the points raised in comparing and contrasting the CFTR gating cycle to the ATP Switch model. And second, testing the ATP Switch model, designed to represent all the members of the ABC superfamily, against other well-known mechanisms should bring about constructive assessment of this general model.
Study of HlyB and Pgp has provided the valuable structural and functional insight for CFTR research. This may be exemplified by the generalized schematic diagrams of an ABC protein being based on Pgp  as well as the substrate-assisted catalysis (SAC) models of HlyB emphasizing the importance of a latch mechanisms, respectively. Outlined below are some of the mechanistic advances linked to CFTR research and suggestions of future areas of interest.
Models of HlyB
The ABC transporter HlyB is an essential Escherichia coli protein that exports hemolysin A, a virulence factor that influences metabolism of host cells, across the cell envelope of the bacterium.
The previous model of NBD function in the HlyB transporter was general base catalysis that hypothesized the nucleotide-binding pocket to contain all the necessary amino acids orientated favorably to facilitate ATP hydrolysis, thereby lacking an intrinsic latch mechanism within the gating cycle. The new SAC model, on the other hand, requires ligand binding to form a head-to-tail NBD dimer that creates a microenvironment favorable for ATP hydrolysis. These modifications of physical properties, such as H-bonding and salt bridge formation, occur in specific residues (Asp630/Glu630/H662) that are involved in hydrolysis.[21
] So, the SAC provides a theory for how ligand binding and subsequent dimerization can be coupled to ATP hydrolysis without futile catalysis, which emphasizes the importance of the latch mechanism in the ATP Switch model and the need for a further regulatory step in the CFTR mechanism.
The D-loop is a highly conserved and characteristic sequence of ABC proteins located in the major catalytic region of an NBD and it has had little scientific attention over the years . It has been proposed by Hanekop et al
. that the D-Loop is part of the molecular machinery allowing the two distant nucleotide-binding sites to interact with each other, which is essential to the cooperativity required in function of the NBD dimer.[21
] Further investigation is required to shed light on this motif, especially as the process of dimer formation and dissociation is fundamental to the ATP Switch model.
The mechanochemistry energy model of HlyB looks at two sources of potential energy within the cycle: The first is mechanical energy from dimerization allowing rigid body motion of the helical domains (33 kj/mol) and the second is the chemical energy released from ATP hydrolysis (36 kj/mol). The two potential energies are of comparable magnitude, thereby further supporting the mechanochemical model.[21
] As a side remark, this model does raise the question of where in the cycle the power stroke is located or even if there are two such strokes, one for mechanical and one for chemical energy release.
Models of P-glycoprotein
Pgp plays an important part in chemotherapeutic resistance of cancer cells and is capable of transporting numerous drugs of variable structures. Due to its pathophysiological significance, it has been widely researched and should help in understanding the general ABC gating mechanism.
Isolated NBD experiments may be inconclusive for studying the gating mechanism as they lack important structural restrictions imposed by TMDs, which apply under physiological conditions. TMDs may be necessary for the formation of important intermediates, such as the occluded Pgp state, and their absence has been proposed to be the reason for symmetrical dimer formation in isolated NBD experiments.[25
] The CFTR protein forms an asymmetrical dimer, similar in structure to the occluded Pgp state, although the basis of its formation may be more heavily based on differences in the primary structure of the nucleotide-binding sites than the dynamics of the gating mechanism. However, even if in isolated nucleotide experiments, an asymmetrical CFTR dimer still forms, this does not retract from the importance of TMD interactions with the NBDs throughout gating mechanism and the need for more advanced experimental techniques in the future to study complete proteins.
ATP catalysis powers the gating mechanism; however, the number of ATP molecules required for a single cycle is debatable for different ABC proteins. Most ABC proteins are thought to require the catalysis of two ATP molecules for a complete cycle, with studies on Pgp showing that the energy from a single ATP catalysis and subsequent electrostatic repulsion between the products is insufficient to restore the transporter to its basal state.[25
] This contrasts to the ABCC family of transporters that require only a single ATP to drive the complete cycle.[8
] However, it seems that the ATP Switch model is sufficiently versatile to accommodate these differences as the model is centered around the dimer that is universal to all ABCs, and so, such differences appear as specifics within a general mechanism.
An interesting way of looking at gating is from a thermodynamic perspective, which analyzes the energy-driven conformational changes between various Pgp intermediates.[26
] This kind of model is depicted in terms of numerous transition states and pathways with thermodynamic pressures dictating the actual mechanism, which allows for many potential pathways but a single most likely route that perhaps may also be reflected in the CFTR mechanism.