The Substrate Anion Selectivity Filter in the Human Erythrocyte Cl - /HCO 3- Exchange Protein, AE1

Carbonic C-terminal HCO 3- C-terminal site. We of C-terminal region of human AE1, using a panel of individual introduced with and reagents last than a a small extracellular loop. we have assessed the sensitivity of AE1 introduced cysteine mutants to functional inhibition by small, hydrophilic sulfhydryl reagents to identify the residues in this region that form the anion selectivity filter and that line the anion translocation pore. A preliminary version of this work has the inhibitory effect of pCMBS on

Respiration and circulation of blood are the two essential physiological functions.
Removal of waste CO 2 from the body is equally important as the uptake of O 2 .High CO 2 levels cause tissue acidosis and depress central nervous system function (1).CO 2 is transported mainly as HCO 3  -in the blood (about 90%).Conversion of CO 2 to HCO 3 benefits the body in two ways: enhancement of blood capacity to carry CO 2 and pH buffering to stabilize the body's pH.The human anion exchanger 1 (AE1, also called SLC4A1 or Band 3) plays a critical role in the CO 2 transport system.In the systemic capillaries, the high CO 2 partial pressure drives CO 2 diffusion into the erythrocyte where it is converted to HCO 3 -.HCO 3 -is then transported out of the erythrocyte by AE1 in exchange for Cl -to prevent mass action from slowing the process.When the blood reaches the lung, the process is reversed and CO 2 is exhaled from the body.With a turnover rate of 5 x 10 4 anions S -1 , AE1 rapidly performs tightly coupled one to one electroneutral exchange of Cl -for HCO 3 -across the plasma membrane, placing AE1 among the fastest membrane transport proteins (2).Transport is especially rapid since AE1 is the most abundant integral membrane protein of the erythrocyte membrane (1.2 x 10 6 copies per cell and 50% of integral membrane protein) (3).
AE1 has additional roles in erythrocyte biology.AE1 interacts with the cytoskeleton via ankyrin to maintain the flexible biconcave disc shape of the erythrocyte (4).In aged erythrocytes, AE1 serves as the senescence antigen for clearance from circulation (5,6).AE1 also participates in the adhesion of malaria-infected erythrocytes to endothelial cells (7), as well as a host receptor for the Plasmodium falciparum invasion of erythrocytes (8).Mutation or deletion of the AE1 gene induces blood group antigens, variant erythrocyte morphologies, and human diseases, including the Diego a blood group antigen (9), Southeast Asian Ovalocytosis (10), and erythroid spherocytosis (11).
AE1 deficient mice display retarded growth, haemolytic anemia and a high rate of neonatal death (12).
Human AE1 belongs to a multi-gene family with three members.AE1 functions in erythrocytes and an N-terminally truncated form is in the kidney; AE2 is the housekeeping anion exchanger found in a variety of tissues; AE3 is expressed in heart, brain and retina.The three AE isoforms have high sequence conservation in their membrane domains and lower conservation in their cytoplasmic domains.Human AE1 is a 110 kDa glycosylated integral membrane protein composed of 911 amino acids.It exists as a homo-dimer in the lipid bilayer, but each monomer performs anion exchange activity independently (13).The 45 kDa cytoplasmic domain of AE1, whose structure has been determined by X-ray crystallography, forms the anchoring sites for cytosolic proteins, including haemologlobin, glycolytic enzymes and ankyrin, which anchors AE1 to the underlying cytoskeleton (14).The membrane domain performs anion transport alone (15).Erythrocytes from the AE1 knock out mouse showed continuous loss of the plasma membrane by forming rod-like blebs (12).Recently, AE1 was shown to be part of a large macromolecular complex in the erythrocyte membrane, with Rh protein and other membrane proteins (16).
AE1 functions by a ping-pong mechanism to exchange Cl -and HCO 3 -across the plasma membrane (17).The Ping-Pong model proposes that AE1 has a flexible transport site that undergoes conformation change to face extra and intracellular alternately (17).Extensive studies of the anion transport site demonstrate that it has distinct inward and outward facing conformations (18).Three-dimensional (19) and two-dimensional (20) cryoelectron microscopy structures are available for the membrane domain of AE1 at 20 Å resolution.
The C-terminal region of AE1 has been proposed to be involved in the anion translocation mechanism for many years.Two anion exchange inhibitors, pyridoxal phosphate and DIDS, both react with K851 in this region to inhibit the transport activity (21,22).The mutation, P868L, causes acanthocytosis and substantially increases anion exchange activity (23).Histidine 834 has been identified as undergoing conformational changes, close to the anion translocation site (24).Carbonic anhydrase binds to the Cterminal cytoplasmic region via the LDADD motif, forming a metabolon to facilitate HCO 3 -transport (25,26).These studies implicate the C-terminal region as the anion translocation site.We recently determined the topology of the C-terminal region of human AE1, using a panel of individual introduced cysteine mutants, probed with permeant and impermeant sulfhydryl reagents (27).The last two TMs were shorter than a standard TM, composed of only 16 amino acids each and were connected by a small extracellular loop.In the present paper, we have assessed the sensitivity of AE1 introduced cysteine mutants to functional inhibition by small, hydrophilic sulfhydryl reagents to identify the residues in this region that form the anion selectivity filter and that line the anion translocation pore.A preliminary version of this work has been published as an abstract (28).Cloning and expression of introduced cysteine mutants -The cloning strategy for the introduced cysteine mutations has been described previously (27).Mutant AE1 cDNAs were expressed by transient transfection of human embryonic kidney 293 cells (HEK), as previously described (29).HEK cells were plated onto 60 mm dishes in 4 ml DMEM medium, containing 5% (v/v) fetal bovine serum (FBS), 5% (v/v) calf serum (Life Technologies).Six to eight hours following seeding, cells were transfected with mutant plasmids, using the calcium phosphate precipitation method (29,30).Cells were then grown at 37 °C, in a 5% CO 2 atmosphere and harvested 48 hours post-transfection.

EXPERIMENTAL PROCEDURES
Transfected HEK cells were washed twice with

Effect of introduced cysteine mutations on AE1 transport activity -To identify amino acids
involved in the AE1 anion translocation process, we constructed 80 individual cysteine mutations consecutively introduced into a cysteineless human AE1 cDNA, at all positions between F806 and C885 in the C-terminal region of human AE1.Cysteineless AE1 cDNA was constructed previously and mutant protein is fully functional (33).
Mutant AE1 proteins were expressed in HEK 293 cells by transient transfection.All proteins were well expressed to similar levels as shown on immunoblots probed with IVF12 antibody, which recognizes the extreme C-terminal tail (31) (Fig. 1).Transport activity of each mutant protein was assessed by monitoring intracellular pH changes associated with Cl -/HCO 3 -exchange in a whole cell assay (27).As shown in Fig. 2, introduced cysteine mutations impaired transport activity to varying degrees.Among 80 introduced cysteine mutant protein, 15 had less than 20% transport activity compared to the cysteineless AE1 protein, and were thus defined as functionally inactive (27).Interestingly, there are five functionally inactive mutants (M833C, H834C, F836C, T837C and Q840C) clustered at the N-terminal end of the second last transmembrane helix, which suggested that this region is critical for the proper folding of the last two transmembrane helices.Mutation of K851, which reacts with both pyridoxal phosphate and DIDS to inhibit anion transport, also inactivated transport function, highlighting the importance of this residue.
Eight AE1 mutant proteins were functionally inhibited by the membrane-impermeant reagent, pCMBS -Data from many studies has implicated the C-terminal portion of AE1 membrane domain as critical for ion translocation (34,35).To identify the residues that form the anion translocation pore in this region, we examined the sensitivity of all the introduced cysteine mutants to small hydrophilic, covalently-acting sulfhydryl-reagents (Fig. 3).If an introduced cysteine residue is in the anion translocation pore, covalent modification with any of these compounds will block the substrate pathway, and thus sterically inactive the transport activity.To determine the effect of sulfhydryl reagents, transport assays were performed before and after treatment with sulfhydryl reagents.
pCMBS is a negatively-charged, membrane-impermeant reagent that has been widely applied to probe membrane transport protein structure (36).Figure 4 provides an example of the effects of pCMBS on AE1 proteins.AE1C -was insensitive to pCMBS, as expected since it does not contain any cysteine residues.In contrast, L863C transport activity was greatly reduced by pCMBS, as seen by the reduction in the rate of pH change associated with HCO 3 -transport, following pCMBS treatment.We note that during the 10 min incubation period with pCMBS, the cytosol acidified to some degree (Fig. 4 A,B).The basis for this effect is unclear.However, upon pCMBS wash-out with Ringer's buffer, pH returned to a baseline very close to the starting level.The fact that pCMBS did not have any effect on AE1C -lends support to the idea that the pH drift during pCMBS incubation did not affect the ability to measure effects of pCMBS covalent modification on the activity of AE1 introduced cysteine mutants.
Among the 65 functionally active introduced cysteine mutants analysed, only eight mutants (V849C, V850C, T853C, L857C, A858C F861C, V862C, L863C) were inhibited by pCMBS (Fig. 5).V850C and L863C achieved the maximum effect with inhibition by pCMBS >70%, suggesting that these residues line the anion translocation pore.All the sensitive residues are hydrophobic, consistent with TM8 of AE1 in which the lining of the pore is hydrophobic (37).Interestingly, these residues lie at the extracellular ends of the last two transmembrane helices and the small extracellular connecting loop (Fig. 2).In addition to the pCMBS-inhibited mutants, some other mutants had anion exchange activity that was slightly increased by pCMBS.Small increases in activity following treatment with sulfhydryl reagents have been reported in studies of introduced cysteine mutants of both transporters and ion channels (38,39) and may reflect effects on cells rather than the protein itself.
The V849 and L863 region is sensitive to inhibition by both pCMBS and MTSEA -To explore further if any residues close to the intracellular side of the protein may line the translocation pore, we scanned all the functional cysteine mutants for sensitivity to the sulfhydryl reagent, MTSEA.The pKa of MTSEA is 8.5 (40), so that a significant fraction of the reagent is unprotonated and neutrally-charged in the pH range of transport assays (pH 7.2 to 7.8).Unprotonated MTSEA is thus membrane-permeant (41).As found for pCMBS, eight residues in the V849 to L863 region were inhibited by MTSEA (Fig. 6).Among these six were also inhibited by pCMBS (V849C, V850C, T853C, L857C, A858C and I863C).Two other mutants (S852C and A855C) were inhibited by MTSEA, but not pCMBS.Conversely, F861C and V862C were inhibited by pCMBS, but insensitive to MTSEA.Strikingly, in the entire region scanned for sensitivity to sulfhydryl reagents, F806 to C885, all pCMBS and MTSEA-sensitive mutants clustered narrowly between V849 and L863.The maximum transport inhibition by MTSEA was less than for pCMBS, only about 30%.This is may be explained by the covalent moiety of MTSEA, which is smaller than that of pCMBS.Alternatively, anionic pCMBS may have greater access to these sites than neutral/cationic MTSEA.

Positively-charged MTSET inhibits mutants only in the extracellular loop region -Human
AE1 will accept transports a wide range of anionic substrates, with strong selectivity against cations (35,42).In the ClC Cl -channel, whose structure may serve as a guide for an anion transporter like AE1, the anion selectivity filter is formed by four residues at the ends of different helices (43).To examine if the V849-L863 region similarly forms the anion selectivity filter, we tested the effect of a cationic sulfhydryl reagent, MTSET (Fig. 3), on the transport activity of these residues.MTSET has a quaternary amine group, with a fixed positive charge (40).Figure 7 shows that MTSET did not inhibit cysteineless AE1C -.In contrast, mutants S852C, T853C, A855C, L857C and A858C were inhibited by MTSET to varying degrees, suggesting that this reagent can reach and react with most of the residues in the loop region.However, MTSET did not inhibit mutants at the ends of the transmembrane segments: V849C, V850C, F861C, V862C and L863C.
Notably all four of these mutants were inhibited by pCMBS or MTSEA.
Cl -reduces the inhibitory effect of pCMBS on mutants V849C and L863C -Kinetic and 35 Cl - NMR studies revealed that Cl -competes with AE1 inhibitors, PLP, DNDS and DIDS, when binding to the anion transport site (44)(45)(46)(47)(48).To examine competition between pCMBS and Cl -, we measured the pCMBS sensitivity of mutant transport activity in the presence and absence of 140 mM NaCl in Ringer's buffer.All of the pCMBS-inhibited mutants in the F806-C885 region were examined.Some mutants (T853C, L857C, F861C and V862C) were not significantly influenced by Cl -(Fig.8A).However, Cl - significantly reduced the inhibitory efficacy of pCMBS for mutants V849C, V850C, A858C and L863C.The magnitude of the Cl -effect can be quantified from the difference in pCMBS inhibition in the absence and presence of Cl -(Fig.8B).A clear trend is evident, where the ends of the V849-L863 region are most affected by Cl -and the centre least.This pattern of Cl -sensitivity strongly suggests that the last two transmembrane segments of AE1 form the anion selectivity filter.
Analysis of the inhibition data in the context of AE1 structural models provides additional insight (Fig. 9).Measurements of accessibility of introduced cysteine mutants to biotin maleimide suggested that the last two transmembrane segments are short and connected by a very short extracellular loop (27).In the present study, the only introduced cysteine mutants that are susceptible to inhibition by sulfhydryl reagents cluster at the extracellular ends of the last two transmembrane segments of AE1.In spite of the capacity of MTSEA to permeate the membrane, mutants sensitive to MTSEA to extend only about half way across the membrane (L863C).The short loop connecting the last two transmembrane segments also featured mutants sensitive to both MTSEA and pCMBS.Interestingly, although one might expect the loop to be highly accessible to labeling by sulfhydryl compounds, S852C and A855C were inhibited by MTSEA, but not pCMBS.Helical wheel analysis reveals that, in the T837-K851 region, mutants functionally inactivated when mutated are organized on one face of the helix (Fig. 9B).Directly opposite this face is a hydrophobic surface on which two valine residues sensitive to both pCMBS and MTSEA are found.In the A858-L874 region a similar pattern is found (Fig. 9C).Residues intolerant to cysteine mutation are on one helical face; opposite is a hydrophobic face where the pCMBS and MTSEAinhibited sites lie.At odds with the pattern is L863C, which is sensitive to both pCMBS and MTSEA and is found on the same helical face as the mutation-sensitive residues.

DISCUSSION
The human AE1 Cl -/HCO 3 -exchanger has an anion translocation rate about 10 fold lower than an ion channel (2,49).Although the substrate translocations rates of AE1 and chloride channels are similar, mechanistically AE1 differs because coupling of inward and outward ion translocation events is obligatory (17).As well, the rigid structures found for ion channels (50) differ from AE1, which has been shown to undergo large conformational changes during anion transport (51).The present study, focussed on the C-terminal portion of the AE1 membrane domain (F806-C885), analysed the sensitivity of AE1 introduced cysteine mutants to sulfhydryl reagents.The fidelity of the substituted cysteine accessibility method (SCAM) has recently been shown by comparison of SCAM results to the crystal structures for the bacterial K + channel and lactose permease (50,52).Here SCAM data has allowed identification of amino acids forming the AE1 anion selectivity filter.The last two transmembrane segments of AE1 are also revealed to form part of the transmembrane ion translocation pathway.
Like ion channels, AE1 is exquisitely selective toward the charge of its substrates, transporting only anions (35).The data presented here paint a picture of the substrate anion charge filtering in AE1.In the scan of the F806-C885 region, only introduced cysteine mutants in the V849-L863 sequence were susceptible to inhibition by three different sulfhydryl reagents.The basis for inhibition by these compounds is their ability to permeate aqueous regions, to form a covalent adduct and to sterically block the confined space of the ion translocation pathway.V849-L863 is therefore the only part of the C-terminal portion of the membrane domain, which meets these criteria.
The charge filter is cleanly delineated in these experiments (Fig. 9).The S852-L857 region is modelled as an extracellular loop on the basis of accessibility to the aqueous reagent, biotin maleimide, which has the greatest ability to label extramembraneous regions and which does not label transmembrane segments (27).In this loop four mutants (S852C, T853C, A855C and L857C) and A858C on the edge of the loop are inhibitable by MTSET, which has a fixed positive charge.MTSET has no effect on AE1 transport activity in regions deeper into the last two transmembrane segments.
The failure to inhibit deeper sites does not represent a lack of accessible or inhibitable residues, since five mutants in this deeper region were inhibited by pCMBS and or MTSEA.Exclusion of MTSET is consistent with charge filtering, to exclude cations.A second line of evidence demonstrating powerful substrate charge filtering was provided by experiments with pCMBS in the absence and presence of extracellular Cl - (Fig. 8).A clear trend in the effect of Cl -was evident across the region.Cl -maximally blocked pCMBS at the ends of the V849-L863 region and had the least effect in the middle.That is, the effects of MTSET and Cl -were reciprocal; Cl -was most potent in the region where MTSET had no effect and least effective at sites where MTSET inhibited transport.The ability of Cl -to reduce pCMBS efficacy at sites deeper in the membrane is consistent with charge filtering to allow anionic Cl -into the deeper sites, where Cl -competes with pCMBS for site occupancy.We conclude that MTSET has defined the boundaries of the cation exclusion charge filter in AE1.In the last two TMs cations cannot permeate deeper into the membrane than S852 and A858, respectively.
The cation exclusion filter is likely formed from two elements: the positive charge on K851 and the d + charge helical dipole at the N-terminus of the last TM (53).Support for a role of K851 in charge filtering also comes from studies of K851, modified by reaction with bisulfosuccinimydyl suberate (54,55).These studies led to the conclusion that loss of charge on K851 results in loss of AE1 transport function, but that this could be compensated at acid pH, where the net charge on the AE1 surface was increased.
Finally, the AE1 transport site is thought to have inward and outward conformations, which can be recruited to face the cell surface with highest substrate concentration (56).
The effects of Cl -that we observed could be explained, in part, by such transport site recruitment.
Extracellular loop S852-L857 is dense with sites where sulfhydryl reagents will cause AE1 transport inhibition.The loop is thus localized to a region where addition of the covalent moiety of MTSET, PCMBS and MTSEA will cause inhibition.The pattern of inhibition with sulfhydryl reagents is consistent with the loop in a confined space, at the entry to the anion translocation channel.Further evidence for the localization of this loop in an inaccessible location comes from several sources.The classical anion exchange inhibitor, H 2 DIDS, covalently reacts with K851 ( 22).An anti-H 2 DIDS antibody was unable to bind H 2 DIDS, when covalently attached to native AE1, but able to bind to H 2 DIDS on denatured AE1, leading to the conclusion that H 2 DIDS is buried in the AE1 structure (57).Similarly, fluorescence energy transfer experiments indicated that the distance between stilbene disulfonates and the cytoplasm was less than the thickness of the bilayer, suggesting that stilbene disulfonates bind beneath the plane of the lipid bilayer (58).
The loop is likely to be in an extended b-conformation.In a b-conformation amino acid side-chains alternate in their orientation.Consistent with this, P854C and S856C could not be inhibited by any of the sulfhydryl reagents, suggesting that their side chains point away from the pore.In contrast, T853, A855 and L857 were inhibitable by sulfhydryl reagents, indicating that these side-chains are arranged to face into the anion translocation pathway.
The last two TMs form a portion of the AE1 anion translocation pore.In these TMs V849, V850, A858, F861, V862 and L863 were all positions where inhibition by sulfhydryl reagents occurred.V849 and T837 are found on one helical face, consistent with being on one wall of the translocation pore (Fig. 9B).Interestingly, other residues on this face are hydrophobic (isoleucine and valine).Pore linings formed of hydrophobic side chains have been a common feature of channels and have been proposed to reduce interaction of substrates with the pore linings (50,59).Interestingly, on the face opposite V849 and T837 are three residues whose mutation to cysteine results in loss of function.We propose that these residues localize to the face opposite the anion translocation pore.We previously speculated that T837 is required for proper insertion of the TM into the bilayer during biosynthesis (27).The role of Q840 may be in interactions between transmembrane segments.Importantly, K851 is on the helical face directly opposite the pore lining.As discussed above we have proposed an essential role of K851 in charge filtering.K851 is likely remote form the pore so that the fixed positive change on the residue does not impede ion flow, but its electric field can contribute to charge filtering.Consistent with this model, K851C is functionally inactive, suggesting an essential role for the residue.Also, covalent modification of K851 with DIDS or PLP did not block the ability of Cl -to bind AE1 (47,48), consistent with a location of K851 remote from the Cl -interacting pore.
On the last TM A858, F861 and V862 were sulfhydryl reagent inhibitable (Fig. 9C).As seen in the other TM, this helical face is formed of hydrophobic amino acids (phenylalanine, valine, leucine, alanine), again consistent with a hydrophobic pore lining.L863 stands apart from the other sulfhydryl-inhibitable mutants, on the opposite face of the helix.If A858, F861 and V862 form the pore lining of AE1, then L863 would be directly opposite the pore and it would be difficult to understand how sulfhydryl reagents could affect L863C activity.P860 could induce a kink in the TM, altering the helical phase so that A858, F861 and V862 are on the same helical face.Alternatively the presence of L863 on the opposite helical face from A858, F861 and V862, suggests the possibility that the last TM of AE1 is conformationally mobile such that L863 can alternately face the pore or not, perhaps as part of the transport cycle.Recent studies on the ability to label H834 with diethyl pyrocarbonate provided compelling evidence for large conformational flexibility in the last two TMs of AE1 (24).
One dramatic feature of the sulfhydryl reagent transport inhibition data is the marked clustering of sensitive mutants at the extracellular ends of the last two TMs (Fig. 9).Surprisingly, pCMBS and MTSEA greatly inhibited transport activity of mutants localized close to the extracellular end of these helices, but had no effect on the remainder of the two TMs.Inhibition by sulfhydryl reagents extended only 3 and 6 residues from the ends of the last two TMs, respectively.In a helical conformation each residue translates 1.5 Å, so 3 and 6 residues corresponds to a depth of only 4.5 and 9 Å respectively from the end of the helix, which is only a fraction of the distance across the bilayer.The observation is also unique in comparison to related studies of ion channels in which clear "stripes" of inhibitable mutants were observed down the length of the pore lining TM (37).The data is also distinguished from the results in a study of TM8 of AE1, in which a clear "stripe" of pore lining residues running the length of the TM were identified (37).In that study pCMBS was able to inhibit introduced cysteine mutants until a point just before E681, which has been defined as the permeability barrier.
MTSEA was effective beyond E681, as a result of its membrane permeability.The last two TMs therefore differ significantly from the structure of ion channel pores and the pore structure contributed by TM8.It may be that the last two TMs are kinked in such a way that before V849 and after L863 the TM does not form the pore lining.There are two other possible explanations for the observation: 1.The uninhibited mutation sites are inaccessible to MTSEA and pCMBS, a possibility, which seems unlikely and 2. These sites are in a large open structure, which is sufficiently open that addition of the sulfhydryl reagent's covalent adduct does not occlude the translocation pore.
The AE1 plasma membrane Cl -/HCO 3 -exchanger has a remarkably high ion translocation rate.However, the basis for the anion translocation event has not been established.Here we have presented evidence that the last two transmembrane segments of AE1 are intimately involved in the translocation process.These last two short (16 residues each) helices are connected by a small extracellular loop.The loop region is required for selection of substrate ions on the basis of charge.We also presented evidence that at least the outer portions of these last two transmembrane segments form a portion of the transmembrane anion translocation pathway.
Identification of this significant portion of the anion translocation pathway paves the way to identify the site of anion exchange, which is likely nearby this region, close to the centre of the bilayer.alternately with Cl -free (white), Cl -containing (gray) Ringer's buffer.Cells were then perfused with Cl -free Ringer's buffer, containing 500 µM pCMBS (black bar) and incubated for 10 min, followed by wash-out with Cl -free Ringer's buffer for 10 min.
Curves are broken because the light path was blocked to prevent possible decomposition of pCMBS under high intensity light (black).Cells were then perfused with Cl -containing, Cl -free and Cl -containing Ringer's buffer to re-assess transport rates.Inhibition of anion exchange was determined by comparison of the initial 100 second rates of alkalinization and acidification before and after pCMBS incubation.
Fluorescence ratio (503 nm/440 nm) is a measure of intracellular pH.Typical assays span the pH range 7.2-7.8.

Figure 1 .
Figure 1.Expression of introduced cysteine mutants in HEK 293 cells.Representative AE1 introduced cysteine mutants were expressed in HEK 293 cells.Mutant proteins on the PVDF membrane were probed on immunoblots with IVF 12 antibody, which recognizes the C-terminal tail of AE1 protein.Note that C885 is an endogenous cysteine residue in an otherwise cysteineless background.

Figure 2 .
Figure 2. Position of introduced cysteine mutations in the topology model of human AE1 and their relative anion exchange activity.A, the model was recently proposed on the basis of cysteine scanning mutagenesis (27).Dashed box highlights the region shown at larger scale in B. Transport activity of each mutant protein was assayed by measurement of pH changes associated with bicarbonate transport.Transport activity is expressed as activity relative to the human AE1C -mutant.Black, <20% of AE1C - activity; Dark gray, 20-60% of AE1C -activity; Light gray, >60% of AE1C -activity (27).Branched structure at amino acid 642 represents the single site of N-linked glycosylation.

Figure 4 .
Figure 4. Representative Cl -/HCO 3 -exchange assays with the cysteine-reactive reagent, pCMBS.HEK293 cells grown on coverslips were transiently transfected with cDNA encoding A, AE1C -and B, L863C.Two days post-transfection, cells were loaded with the pH sensitive dye, BCECF-AM, and placed in a fluorescence cuvette in a fluorimeter to monitor fluorescence.Transport rates were measured by perfusion of cells

Figure 5 .
Figure 5.Effect of pCMBS on AE1 transport activity.Individual AE1 introduced cysteine mutants were expressed in HEK293 cells grown on coverslips.Transport rates were measured before and after incubation with 500 µM pCMBS.Background transport activity of mock-transfected cells was subtracted from each assay.Arrowheads mark mutants not analyzed due to low transport activity.Error bars represent standard error.Uninhibited mutants were examined with n=1.Mutants sensitive to pCMBS had n=3-5.Asterisks mark mutants whose transport rate was significantly (p<0.01)inhibited by pCMBS.

Figure 6 .
Figure 6.Effect of MTSEA on AE1 transport activity.Transport rates were measured before and after incubation with 5 mM MTSEA.Background transport activity of mock-transfected cells was subtracted from each assay.Arrowheads mark mutants not analyzed due to low transport activity.Asterisks mark mutants whose transport rate was significantly (p<0.01)inhibited by MTSEA.Error bars represent standard error.Uninhibited mutants were examined with n=1.Mutants sensitive to MTSEA had n=3-5.

Figure 7 .
Figure 7. Effect of MTSET on AE1 transport activity.Introduced cysteine mutants between V849 and L863 were incubated with 5 mM MTSET.Transport rates were measured before and after MTSET treatment.Arrowheads mark mutants not analyzed due to low transport activity.The dashed line marks activity of AE1C -after pCMBS treatment.Black bars with asterisk represent mutants whose transport activity was significantly (p<0.01)inhibited by MTSET.Error bars represent standard error (n=3-5).

Figure 8 .
Figure 8. Charge filtering in the last two transmembrane segments.Transport activity was measured for AE1 introduced cysteine mutants before and after incubation with 500 mM pCMBS.Incubation with pCMBS was performed either in the absence (gray bars) or presence (hatched bars) of 140 mM NaCl.A, the inhibitory effect of pCMBS was measured as the % difference in activity in the absence and presence of pCMBS.B, the effect of Cl -on pCMBS sensitivity was measured from the difference in pCMBS sensitivity when pCMBS was incubated in the absence or presence of Cl -.Error bars represent standard error (n=3~5).Asterisks show statistical difference (p<0.01).

Figure 9 .
Figure 9. Location of sulfhydryl reagent-sensitive mutants.A, topology of the Cterminal portion of AE1.B, C helical wheel models with 3.6 residues/turn for B, by guest on October 5, 2017 http://www.jbc.org/Downloaded from Fig.2 Fig.3 Fig.7 Materials -Restriction endonucleases were from New England Biolabs.Pwo DNA polymerase was from Roche Diagnostics.Plasmid preparation kits were from Qiagen.