Published ahead of print on August 21, 2003, doi:10.1165/rcmb.2002-0108OC
American Journal of Respiratory Cell and Molecular Biology. Vol. 30, pp. 242-248, 2004
© 2004 American Thoracic Society
DOI: 10.1165/rcmb.2002-0108OC
Intermolecular Interaction Between R Domains of Cystic Fibrosis Transmembrane Conductance Regulator
Sanhita Gupta,
Junxia Xie,
Jianjie Ma and
Pamela B. Davis
Departments of Pediatrics and Physiology and Biophysics, Case Western Reserve University School of Medicine, Cleveland, Ohio
Address correspondence to: Pamela B. Davis, Department of Pediatrics, Case Western Reserve University at Rainbow Babies and Children's Hospital, Biomedical Research Bldg, Rm 831, 2109 Adelbert Road, Cleveland, OH 44106-6006. E-mail: pbd{at}po.cwru.edu
 |
Abstract
|
|---|
The function of the R domain of cystic fibrosis transmembrane conductance regulator (CFTR) has not yet been fully established. The cis-trans proline isomerase cyclophilin A stimulates channel activity, and stimulation depends on the presence of highly conserved prolines at positions 740, 750, and 759. When the prolines at these positions, which normally exist in the cis conformation, are locked into the trans conformation by mutation to alanine (the P3A mutant), the open probability of P3A is high and is not further increased by cyclophilin A. We speculated that one mechanism by which this could occur was by promoting CFTR dimerization, which has been shown to increase open probability, and that the P3A-CFTR might favor dimerization more strongly than the native sequence. To test the hypothesis that R–R interaction occurs and is stronger in the P3A-R mutants, we investigated R–R interactions. GST-R and StrepII-R proteins expressed in Escherichia coli could interact with R domain protein translated in vitro as well as with full-length CFTR. In similar assays, the P3A mutant of R domain also interacts with R domain and P3A-R. The P3A-R–P3A-R interaction is stronger than the R–R interaction, which corroborates our data from the channel study and supports our hypothesis. Studies of deletion constructs of the isolated R domain and of full-length CFTR localize the region of interaction to the C-terminal portion of R (after amino acid 708). Particularly, the last 22 a.a. residues (838–859) of R are essential for this binding. R–R interaction possibly plays a role in channel gating.
Abbreviations: amino acid, a.a. • ATP-binding cassette, ABC • circular dichroism, CD • cystic fibrosis transmembrane conductance regulator, CFTR • glutathione sulfo-transferase, GST • phosphate-buffered saline, PBS • polymerase chain reaction, PCR • protein kinase A (cAMP-dependent), PKA • open probability, Po • single chain Fv, scFv
|
Introduction
|
|---|
Cystic fibrosis is an autosomal recessive disorder caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene that encodes a chloride-conducting channel which belongs to the ATP-binding cassette (ABC) transporter family. The CFTR protein consists of two similar motifs, each containing six membrane-spanning helices and a nucleotide-binding domain (NBD) common to all members of the ABC transporter family. The two motifs of CFTR are linked together by an intracellular regulatory (R) domain, which is unique to CFTR (1).
The R domain contains consensus sites for phosphorylation by cAMP-dependent protein kinase (PKA), which are the basis for regulation of the CFTR chloride channel (2–4). Circular dichroism (CD) on the R domain suggested that this peptide has little well-ordered secondary structure (helical content 5%). Also, the CD spectra of phosphorylated R did not show any significant change, indicating that phosphorylation probably introduced only a small change in conformation of R domain (5). In the middle of the R domain sequence, there are three proline residues, P740, 750, and 759, that are highly conserved among the CFTR molecules in different animal species (6). Because proline residues impart some structure to protein because they exist in the cis configuration, as opposed to the trans configuration typical of the other amino acids, it may be that the three conserved prolines contribute to what little structure exists in the R domain.
The exact mechanism by which the R domain regulates the gating of CFTR chloride channel remains largely unknown. In a previous study, we showed that when wild-type CFTR captured in a lipid bilayer membrane is treated with cyclophilin A, the open probability (Po) of the channel is increased significantly (6). Cyclophilin A has enzymatic activity as a cis-trans peptidyl-prolyl isomerase. Because CFTR- R (708–835) was unaffected by cyclophilin A treatment, it was likely that proline residues in the deleted segment were isomerized by cyclophilin A and produced the increase in open probability. When three highly conserved proline residues in this segment (P740, P750, P759) were mutated to alanines, the resulting mutant of CFTR, which we named P3A, had Po significantly higher than that of the wild–type, and did not change on cyclophilin A treatment (6). These observations led to the hypothesis that the conversion of the proline residues from cis to trans configuration alters the conformation of the R domain to produce a more active channel. One possible mechanism for this increased activity is that the altered R conformation enhances the intermolecular interaction between CFTR molecules (e.g., formation of CFTR dimers), because evidence is accumulating that dimerization potentiates the activity of CFTR (7, 8). In this study, we test the hypothesis that the R domain interacts with itself and that this interaction is promoted by the P3A mutation.
|
Materials and Methods
|
|---|
Construction of Plasmids for Expression of R Domain, Mutants of R Domain, and Single-Chain Fv
c-DNA encoding amino acid (a.a.) 583–859 of CFTR (R domain) was amplified from wild-type CFTR c-DNA by PCR and cloned into BamHI and XhoI sites of the pGEX5X-2 (Amersham Pharmacia Biotech, Piscataway, NJ) downstream and in frame with glutathione sulfo-transferase (GST) sequence. R domain was also cloned with or without strepII sequence (NWSHPQFEK) at the N-terminal in pET17b (Novagen, Madison, WI). R domain and its mutants were cloned using standard recombinant DNA techniques in the BamHI and XhoI sites of the pET17b vector. The oligonucleotides used in construction of R domain and its mutants are as shown in Table 1. The 740–759 R mutant was made by megaprimer-directed PCR as described previously (9), and the oligonucleotide used for making the megaprimer is 5'-AGAAGGCTGTCCTTAGTAACGCTTCAGGCACGAAGG-3' and the reverse primer for R domain (Table 1) followed by a second round of PCR using the megaprimer and the forward primer for R domain (Table 1).
For single-chain Fv (scFv) to anti-human polymeric immunoglobulin receptor (pIgR) we used the same construct as described previously (10), subcloned into the KpnI site of pET17b (Novagen).
In Vitro Translation
The following proteins were translated using TNT system (Promega, Madison, WI) in the presence of 35S methionine according to manufacturer's instructions: R domain (a.a. 583–859 of CFTR), P3A-R, NEG2-R, 740–859 R, 817–859 R, 817–838 R, 583–625 R, 583–702 R, 740–759 R, 838–859 R, and scFv to pIgR cloned in pET17b vector.
Expression of Recombinant Proteins
Overnight cultures of Escherichia coli BL21 DE3 carrying pET17b-R /strepII-R construct or E. coli DH5 transformed with GST or GST-R construct was diluted 100-fold and incubated at 37°C until the OD600 reached 0.6–1, and then induced with isopropyl ß-D-thiogalactopyranoside (IPTG). After 3 h of induction, bacteria were collected, washed with phosphate-buffered saline (PBS), resuspended in PBS containing 1% Triton X-100 (vol/vol), and sonicated. Cellular debris was removed by centrifugation. The supernatant was used as a source of untagged R domain. The fusion protein and GST control protein were bound to glutathione sepharose (Amersham Pharmacia Biotech) and extensively washed with PBS containing 1% Triton X-100. The strepII-R protein was expressed and purified using procedures identical to that of GST-R except streptactin–agarose matrix (IBA, Gottingen, Germany) was used to purify the protein according to the manufacturer's recommendation. The integrity of the proteins bound to the matrix was confirmed by gel electrophoresis and Western blot using anti-R antibody (R&D, Minneapolis, MN). Matrix-bound GST and strepII proteins were used in the interaction studies.
Interaction Assay
In vitro translated 35S labeled proteins (20 µl) were incubated with 20 µl of glutathione sepharose beads containing 1–2 µg of bound recombinant proteins in 500 µl of PBS containing 1% Triton X-100 and protease inhibitor cocktail (Sigma Chemical Co, St. Louis, MO). After extensive washing with PBS containing 1% Triton-X-100 and protease inhibitor, bound proteins were eluted in 25 µl of Laemmli sample buffer and resolved by SDS-polyacrylamide gel electrophoresis. Results were visualized by fluorography followed by autoradiography. For the competition experiments, in vitro translated cold R was added in addition to the 35S labeled proteins and matrix-bound GST fusion proteins. Both the cold and radioactive labeled R and P3A were translated using the same translation mix except 35S methionine was added to the latter instead of cold methionine. To quantitate unlabeled R or P3A used in the competition, 35S labeled R or P3A resolved from the same volume of reaction mix was extracted from the gel and counted in a Beckman LS-5801 liquid scintillation counter (Fullerton, CA). The quantity of the protein represented by these counts could then be calculated by taking into consideration the specific activity of 35S (1,000 Ci/mmol), the number of methionines (six) in the R domain, and the measured efficiency of the scintillation counter. After the competition reaction, the bound proteins were analyzed by SDS-polyacrylamide gel electrophoresis followed by autoradiography. The amount of labeled R or P3A bound to GST-R or GST-P3A was estimated by densitometric scanning of the radioactive bands in the gel. An IC50 was estimated for each individual experiment and the mean IC50 for the replicate experiment is presented in the text. The IC50 calculated for the competition reaction were compared by paired t test (SigmaStat 2.03).
Phosphorylation and Dephosphorylation Reactions
R domain translated in vitro (150 µl) was treated with 1 µl of Protein phosphatase 2A (Calbiochem, San Diego, CA) at 30°C for 30 min. An aliquot of the dephosphorylated mix was withdrawn and preserved as unphosphorylated R domain for the interaction assay. Okadaic acid was added to the remainder, and one tenth volume of solution containing 500 mM Tris, pH 7.5, 100 mM MgCl2, 1 mg/ml BSA was added, followed by PKA and ATP, and the reaction was incubated at 30°C for 1 h. The resulting translation mix was treated as phosphorylated R domain and used in subsequent interaction assays.
Cell Culture
The pCEP4 plasmids containing wild-type, NEG2, R, and P3A mutant CFTR were transfected into 293 HEK (293-EBNA; Invitrogen, San Diego, CA) cells using the LipofectAMINE reagent (11). The parent cells were passaged 1:6 2 d before transfection. One or two days after transfection, cells were used for isolation of membrane vesicles.
9HTEo-cells transfected with pCEP4-R (a.a. 583–859) (12) were used as a source of R domain expressed in mammalian cells. The cells were grown to near confluence and lysed with PBS containing 1% Triton X-100 and protease inhibitor cocktail, sonicated, centrifuged, and the cell debris discarded. The supernantant was used for interaction assay.
|
Results
|
|---|
R Domain Interacts with R Domain and P3A Translated In Vitro
To determine if R domain interacts with R domain, we constructed GST- and strepII-tagged R domains, which were used to pull down 35S-methionine–labeled R in an interaction assay. R-GST (60 kD) and StrepII-R (30 kD), expressed and purified from E. coli, were of the appropriate molecular weight and detected by anti-R antibody in a Western blot (Figure 1A, lanes 2 and 3). Translation reaction with no DNA added did not yield a labeled product (Figure 2, lane 1) and did not show a labeled product in a pull down assay with either GST or R-GST (data not shown). Both R-GST and strepII-R interacted with R domain translated in vitro and labeled with 35S methionine, whereas GST alone, as control, did not interact with R (Figure 2A). In addition, R-GST did not interact with anti-pIgR scFv, a protein of comparable size not directly related to CFTR (Figure 2A). R domain translated in vitro yields three bands on an SDS-PAGE (more prominent in Figure 2B), which are in close proximity to each other and are probably the result of multiple initiations of translation (from methionines 595 and 607, which are quite close to the translation start), not uncommon in in vitro translation reactions.
View larger version (32K):
[in this window]
[in a new window]
|
Figure 1. Expression and purification of R domain peptides. (A) Detection of R-GST and StrepII-R. GST (lane 1) and R-GST (lane 2), purified using glutathione sepharose beads and StrepII-R (lane 3), purified using streptactin beads, were analyzed on 10% SDS-PAGE and subjected to Western blot analysis with anti-R antibody. (B) P3A-R-GST, purified using glutathione sepharose beads, was analyzed on 10% SDS-PAGE and subjected to Western blot analysis with anti-R antibody.
|
|
View larger version (32K):
[in this window]
[in a new window]
|
Figure 2. R domain translated in vitro interacts with R-GST, StrepII-R, and P3A-GST. (A) Twenty microliters of R domain translated in vitro, labeled with 35S methionine, was mixed with GST, R-GST, or StrepIIR bound to sepharose or agarose beads for 1 h in 1% Triton at 4°C. The beads were washed with PBS containing 1% Triton and analyzed by SDS-PAGE, fluorography, and autoradiography. Lane 1: translation without DNA; lane 2: translation with pET17b anti-pIgR scFv; lane 3: translation with pET17b R; lane 4: scFv pull down with strepII-R; lane 5: R pull down with strepII-R; lane 6: R pull down with GST; lane 7: R pull down with R-GST; lane 8: scFv pull down with GST; lane 9: scFv pull down with R-GST. (B) Same as A, except P3A was translated in vitro and GST (lane 1), R-GST (lane 2), and P3A-R-GST (lane 3) were used to pull down P3A-R.
|
|
Because prolines 740, 750, and 759 in the R domain have a critical role in increasing open probability of the channel (6), we cloned the R domain mutant with the three conserved prolines mutated from full-length P3A-CFTR. This mutant of R, expressed as P3A-R-GST or P3A-R translated in vitro, was used in the interaction assay to determine whether P3A-R interacts with P3A-R and/or R domain. P3A-R-GST expressed and purified from E. coli yielded a single band on Western blot analysis with anti-R domain antibody at 60 kD (Figure 1B). R-GST and P3A-R-GST, but not GST alone, pulled down P3A-R translated in vitro (Figure 2B). Significantly higher concentrations of cold R are required to disrupt the P3A-R–P3A-R interactions than are required to disrupt R–R interactions, indicating stronger interaction between the mutant Rs than the wild-type Rs (Figure 3A). The concentration of R required for half maximal displacement or half maximal inhibition (IC50) for R–R binding was 0.42 ± 0.02 nM, compared with 1.0 ± 0.00 nM (P < 0.001, n = 3), for P3A-R–P3A-R. The competitive binding experiment was repeated three times for each interaction, IC50 was calculated from each individual graph, and a mean value is presented. The other observation leading to the same inference (Figure 3B) is that two times more R than P3A is necessary to disrupt a heterodimer R–P3A-R interaction (IC50 [R] = 1.4 ± 0.03 nM, P < 0.001; IC50 [P3A-R] = 0.7 ± 0.03 nM, P < 0.001). These results were generated from four individual experiments. IC50 was calculated from each individual graph and a mean value is presented.
View larger version (13K):
[in this window]
[in a new window]
|
Figure 3. P3A-R–P3A-R interaction is stronger than R–R interaction. (A) R-GST was used to pull down 0.7 pmol of 35S labeled R (filled circles) and P3A-R-GST was used pull down 0.7 pmol of 35S labeled P3A-R (open circles). Cold R was added in increasing amounts to the reactions as indicated in the x-axis of the graph. The radio-activeproteins bound to the beads were separated in a SDS-polyacrylamide gel. After autoradiography, the intensity of interacted R or P3A was determined by densitometric scanning. This graph shows the mean of three independent experiments. (B) Same as A, except P3A-R–GST was used to pull down P3A in both, and either cold R (open circles) or cold P3A (filled circles) was used as competitor. This graph is the mean of four independent experiments.
|
|
R-GST Interacts with R Domain in CFTR
To determine whether R domain in the context of CFTR can interact with the R domain, expressed separately, we performed the following experiment. Wild-type CFTR vesicles were diluted with the interaction assay buffer and tested for interaction with GST or R-GST. As shown (Figure 4A), R-GST, but not GST alone, interacts with CFTR. However, neither GST nor R-GST was able to pull down CFTR-R (708–835) (Figure 4B), suggesting that R in R-GST interacts specifically with R domain in CFTR. However, when only a.a. 817–838 are deleted (CFTRNEG2), interaction is restored. Interestingly, although in wild-type CFTR, R-GST pulls down predominantly Band C, the fully glycosylated form of CFTR, in CFTRNEG2, predominantly Band B is pulled down. This may be because more Band B is available in the vesicles prepared from the CFTRNEG2 transfections (this observation was consistent in all three trials) or it may indicate that the putative site of interaction is influenced by glycosylation status, particularly if the stabilization of the helical NEG2 region is removed.
Interaction Occurs Whether or Not R Is Phosphorylated
We used two independent methods to determine whether phosphorylation of R domain is significant in the interaction. First, we expressed R domain from pET17b-R transformed in E. coli DE3 and used this in the pull down assay assuming that bacterially expressed R domain remains unphosphorylated in the absence of a source of PKA in E. coli. We used R domain expressed in 9HTEo- cells transfected with pCEP-R as a source of phosphorylated R. This R domain, by its size and by Western blot with antibodies to phosphoserine, is largely phosphorylated (12). Figure 5A shows that R domain from both prokaryotic (lane 4) and mammalian (lane 5) cells is recognized by anti-R antibody in a Western blot analysis and interacts with R-GST in the interaction assay (Figure 5).
View larger version (43K):
[in this window]
[in a new window]
|
Figure 5. R–R interaction occurs with both phosphorylated and dephosphorylated R. (A) E. coli DE3 was transformed with pET17b+R plasmid, grown to log phase, and induced with 0.5 mM IPTG for 3 h. The E. coli extract was fractionated into soluble and membrane fractions. The uninduced bacterial lysate (lane 1), induced bacterial lysate (lane 2), and the fractions (lane 3: insoluble, lane 4: soluble) were analyzed by Western blot using anti-R antibody. Lane 5: 9HTEo-cells transformed with R domain were lysed and the cell lysate was analyzed by Western blot using anti-R antibody. (B) Soluble fraction from E. coli was allowed to interact with GST (lane 1) or R-GST (lane 2) and the associated protein(s) was analyzed by Western blot using anti-R antibody. 9HTEo-cell extract was mixed with GST (lane 3) or R-GST (lane 4) and the associated protein(s) was analyzed using Western blot using anti-R antibody. (C) In vitro 35S-labeled R domain (lane 2) was dephosphorylated using protein phosphatase (lane 3) and rephosphorylated using PKA in the presence of phosphatase inhibitor (lane 4). Apparent molecular weight shifts slightly due to addition of highly charged phosphate groups. Lane 1 shows translation without DNA (control). (D) The translated R domain pulled down with GST (lane 1) and R-GST (lane 2), dephosphorylated R pulled down with GST (lane 3) and R-GST (lane 4), and rephosphorylated R pulled down with GST (lane 5) and R-GST (lane 6) are shown.
|
|
In the second approach, R domain translated in rabbit reticulocyte lysate was dephosphorylated using protein phosphatase 2A and rephosphorylated in the presence of okadaic acid by PKA and ATP. Figure 5D shows that both dephosphorylated and rephosphorylated R interacts with RGST. Therefore, using two independent approaches, it appears that the R domain is capable of interaction with R whether or not it is phosphorylated.
740–859, 817–859, and 838–859 Mutants of R Domain Do Not Interact with R-GST
Our previous study shows that P740, P750 and P759 in the R-domain play an important role in CFTR channel function. It has also been reported that a short segment of R domain (a.a. 817–838), rich in negatively-charged amino acids when added exogenously as a peptide, stimulates the wild-type CFTR channel activity (11). The CFTR–NEG2 channel is independent of PKA regulation, and its activity remains unaltered by cyclophilin A treatment. However, an R domain mutant with an NEG2 deletion (a.a. 817–838) interacts with R, whereas two other deletion mutants (a.a. 740–859) and (a.a. 817–859) do not interact with R-GST (Figure 6). In addition, the NEG2 mutant of full-length CFTR interacts with R-GST, so it is unlikely that the NEG2-R results are spurious (Figure 4C). Interestingly, although 740–759 R, from which the critical prolines are deleted, interacts with R, so these proline residues are not essential for the interaction. However, a mutant lacking a.a. 838–859 does not interact. This suggests that the amino acid residues immediately following NEG2 in the R domain are important for the interaction. We made deletion mutants 583–625 and 583–702, in which the same number of amino acids were deleted from the N-terminus of the R domain as 740–859 and 817–859 had deleted from the C-terminus. Both of these mutants interacted with the R domain. Therefore, alteration in size of the R domain alone probably did not cause the lack of interaction in the C-terminus mutants.
View larger version (24K):
[in this window]
[in a new window]
|
Figure 6. Interaction of R with mutants of R domain. The deletion mutants of R shown in this figure were constructed using standard recombinant DNA techniques and were translated in vitro using TNT coupled transcription translation system, labeled with 35S methionine. These labeled proteins were used in the interaction assay with R-GST, analyzed by SDS-PAGE and autoradiography.
|
|
|
Discussion
|
|---|
We pulled down R domain with R-GST but not GST alone, indicating that R interacts with R. However, because the size of GST is comparable to that of the R domain, we were concerned that it might cause a change in R domain conformation that promotes interaction. To address this issue, we constructed a StrepII-R fusion protein where an octapeptide was fused to R instead of the much larger GST. Like R-GST, StrepII-R interacted with R. Moreover, the R–R interaction in vitro seems to be specific, because R does not interact with anti-pIgR scFv, a protein of comparable size that is not related to CFTR, and cold R competes with labeled R in a dose-dependent manner. However, portions of large proteins expressed in vitro may not assume the same conformation as they do in the context of the whole molecule. To test whether R domain interacts with R domain in the context of CFTR, we used GST-R to attempt to pull down wild-type CFTR and CFTR–R. Wild-type CFTR, but not the mutant form missing a.a. 708–835, could be pulled down in this assay, further confirming the specificity of the R–R interaction, demonstrating that exogenous R can interact with R in its native conformation in the context of the intact channel, and focusing attention on the C-terminal portion of R as a potential site for interaction.
When the three critical proline residues in the R domain are mutated to alanine, the resulting mutant, P3A-R, interacts with both R and P3A-R. If our hypothesis, that the all-trans conformation about the prolines favors R–R interaction, is correct, then P3A-R–P3A-R interaction should be stronger than the R–R interaction. Results from competition experiments indicate that significantly larger concentration of cold R is necessary to disrupt the interaction of P3A-R with P3A-R than is required to disrupt the interaction of R with R. In addition, the interaction of the heterodimer R–P3A-R is disrupted by a lower concentration of P3A than R. Hence the change in conformation of R domain around the proline residues does appear to favor the R–R interaction. We speculated that this interaction favors dimer formation, which is reflected functionally in the increase in open probability of the CFTR channel (6).
Although the phosphorylation status of the R domain plays a major role in determining CFTR channel activity, the R–R interaction occurs whether the R domain is phosphorylated or not. This observation suggests that the conformational change induced by isomerization about the proline residues results in a better presentation of the site for R–R interaction, whatever the phosphorylation status. However, the assays we performed on phosphorylated and dephosphorylated R domain are qualitative, and it is possible that modest differences in the affinity of R for R are entrained by phosphorylation.
To further investigate the site within the R domain responsible for the R–R interaction, based on the CFTR and CFTR-R pull down data, we focused on a.a. 708–835, because when this sequence was deleted, interaction was abolished. The critical prolines at 740–759 fall within this region. We also tested the NEG2 region, because we found that deletion of this 22 a.a. stretch (a.a. 817–838), which contains many negatively charged residues, results in a channel that is active without PKA phosphorylation. Moreover, the NEG2 peptide, added exogenously, can stimulate channel function (11), suggesting that it is capable of interaction at some site within CFTR. Both R domain mutants 740–859 and 817–859 were incapable of interacting with R-GST. This focused attention on a.a. 817–859 as critical for the R-R interaction. However, R domain with a.a. 817–838 deleted, as well as CFTR with a.a. 817–838 deleted, interacted with RGST. Thus, the NEG2 region is not required for interaction. This observation suggests that the R–R interaction possibly requires a.a. 838–859. Indeed the mutant R domain from which a.a. 838–859 were deleted did not interact with R. Attempts to construct and test the CFTR838–859 molcule did not meet with success, so this idea could not be tested in the context of the intact molecule. However, GST-R pulls down CFTR but not CFTR-R (708–835), which retains the residues 838–859. In CFTR-R mutant, the conformation of this sequence (838–859) may be altered by the large adjacent deletion, more so than by the small, but still adjacent, deletion in CFTR–NEG2. The N-terminal deletion mutants, 583–625 R and 583–702 R, both interacted with R-GST, and CFTR R, which retains these residues does not. Therefore, the site for R–R interaction does not reside in the region of a.a. 583–702.
Although our data strongly suggest that R–R interactions occur, the R domain cannot be the exclusive site for inter-domain interaction or be required for CFTR to assume the functionally active form, because CFTR-R generates an active channel, albeit one with low open probability. Moreover, several reports suggest that accessory molecules that link the C-termini of CFTR also promote CFTR activation, but this linkage also cannot be required for activation because C-terminal truncations of CFTR can form functional channels (13). It may be that the interactions that cause association between CFTR molecules are actually strongest at other sites, such as the hydrophobic membrane spanning domains, but in order for these interactions to occur, monomers must be in close proximity to one another. Such proximity can be achieved by covalent linkage, as in the tandem CFTR molecules made by Zerhusen and coworkers (14) that form channels with properties similar to native CFTR, or by connecting the C-termini of CFTR molecules via PDZ domains (7, 8). Our results clearly suggest that interaction between R domains is another way for such interaction to occur, which favors activation of CFTR channels especially when key proline residues assume the trans conformation. It is also possible that the monomeric form of CFTR is active (15), but the dimeric form is more active, and R–R interactions promote conversion from less active monomeric form to the more active dimer. If this is the case, then drugs such as cyclophilin A, a cis-trans peptidyl prolyl isomerase, may help shift the equilibrium from the less active monomer to the more active dimer. Such drugs might enhance the activity of CFTR that is present in low quantity because of splice mutations or transcriptional regulation, or mutants with partial activity that reach the plasma membrane. Cis-trans isomerization around the critical proline residues either by cyclophilin A treatment or mutation of prolines 740, 750, and 759 to alanine, activates the channel by increasing the number of openings and not the open time (6). Thus one could speculate that an alternative explanation for the enhanced channel activity with cyclophilin A treatment or the P3A mutant is that the P3A mutant or cyclophilin A treatment somehow enhances the ability of phosphorylated R to permit channel openings. Though the mechanism of how phosphorylation translates into channel openings or how P3A might enhance it is not clear at the present time, this remains an alternative possibility.
|
Acknowledgments
|
|---|
This work was supported by the grants T32 HL07415, R01 DKHL49003, and P50 HL60293, and by grants from the Cystic Fibrosis Foundation.
Received in original form July 9, 2002
Received in final form July 24, 2003
|
References
|
|---|
- Riordan, J. R., J. M. Rommens, B. Kerem, N. Alon, R. Rozmahel, Z. Grzelczak, J. Zielenski, S. Lok, N. Plavsic, J. L. Chou, M. L. Drumm, M. C. Ianuzzi, F. S. Collins, and L. C. Tsui. 1989. Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science 245:1066–1073.[Abstract/Free Full Text]
- Ostedgaard, L. S., O. Baldursson, and M. J. Welsh. 2001. Regulation of the cystic fibrosis transmembrane conductance regulator Cl- channel by its R domain. J. Biol. Chem. 276:7689–7692.[Free Full Text]
- Sheppard, D. N., and M. J. Welsh. 1999. Structure and function of the CFTR chloride channel. Physiol. Rev. 79:S23–S45.
- Gadsby, D. C., and A. C. Nairn. 1999. Control of CFTR channel gating by phosphorylation and nucleotide hydrolysis. Physiol. Rev. 79:S77–S107.
- Ostedgaard, L. S., O. Baldursson, D. W. Vermeer, M. J. Welsh, and A. D. Robertson. 2000. A functional R domain from cystic fibrosis transmembrane conductance regulator is predominantly unstructured in solution. Proc. Natl. Acad. Sci. USA 97:5657–5662.[Abstract/Free Full Text]
- Xie, J., J. Zhao, P. B. Davis, and J. Ma. 2000. Conformation, independent of charge, in the R domain affects cystic fibrosis transmembrane conductance regulator channel openings. Biophys. J. 78:1293–1305.[Medline]
- Wang, S., H. Yue, R. B. Derin, W. B. Guggino, and M. Li. 2000. Accessory protein facilitated CFTR-CFTR interaction, a molecular mechanism to potentiate the chloride channel activity. Cell 103:169–179.[CrossRef][Medline]
- Raghuram, V., D. D. Mak, and J. K. Foskett. 2001. Regulation of cystic fibrosis transmembrane conductance regulator single-channel gating by bivalent PDZ-domain-mediated interaction. Proc. Natl. Acad. Sci. USA 98:1300–1305.[Abstract/Free Full Text]
- Barik, S. 1995. Site-directed mutagenesis by double polymerase chain reaction. Mol. Biotechnol. 3:1–7.[Medline]
- Gupta, S., J. Eastman, C. Silski, T. Ferkol, and P. B. Davis. 2001. Single chain Fv: a ligand in receptor-mediated gene delivery. Genes Ther. 8:586–592.[CrossRef]
- Xie, J., L. M. Adams, J. Zhao, T. A. Gerken, P. B. Davis, and J. Ma. 2002. A short segment of the R domain of cystic fibrosis transmembrane conductance regulator contains channel stimulatory and inhibitory activities that are separable by sequence modification. J. Biol. Chem. 277:23019–23027.[Abstract/Free Full Text]
- Perez, A., K. A. Risma, E. A. Eckman, and P. B. Davis. 1996. Overexpression of R domain eliminates cAMP-stimulated Cl- secretion in 9/HTEo- cells in culture. Am. J. Physiol. 271:L85–L92.
- Rich, D. P., R. J. Gregory, S. H. Cheng, A. E. Smith, and M. J. Welsh. 1993. Effect of deletion mutations on the function of CFTR chloride channels. Receptors Channels 1:221–232.[Medline]
- Zerhusen, B., J. Zhao, J. Xie, P. B. Davis, and J. Ma. 1999. A single conductance pore for chloride ions formed by two cystic fibrosis transmembrane conductance regulator molecules. J. Biol. Chem. 274:7627–7630.[Abstract/Free Full Text]
- Ramjeesingh, M., C. Li, I. Kogan, Y. Wang, L. J. Huan, and C. E. Bear. 2001. A monomer is the minimum functional unit required for channel and ATPase activity of the cystic fibrosis transmembrane conductance regulator. Biochemistry 40:10700–10706.[CrossRef][Medline]
Copyright © 2004 American Thoracic Society.
|
Top
Abstract
Introduction