Expression and purification of the alpha subunit of the epithelial sodium channel, ENaC

Accepted Manuscript Expression and Purification of the Alpha Subunit of the Epithelial Sodium Channel, ENaC Bharat G. Reddy, Qun Dai, Carmel M. McNicholas, Catherine M. Fuller, John C. Kappes, Lawrence J. DeLucas PII: DOI: Reference:

S1046-5928(15)30064-4 http://dx.doi.org/10.1016/j.pep.2015.09.014 YPREP 4789

To appear in:

Protein Expression and Purification

Received Date: Revised Date: Accepted Date:

14 July 2015 5 September 2015 11 September 2015

Please cite this article as: B.G. Reddy, Q. Dai, C.M. McNicholas, C.M. Fuller, J.C. Kappes, L.J. DeLucas, Expression and Purification of the Alpha Subunit of the Epithelial Sodium Channel, ENaC, Protein Expression and Purification (2015), doi: http://dx.doi.org/10.1016/j.pep.2015.09.014

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Expression and Purification of the Alpha Subunit of the Epithelial Sodium Channel, ENaC Bharat G. Reddy
, Qun Dai*, Carmel M. McNicholas§, Catherine M. Fuller§
, John C. Kappes
 *, #, and Lawrence J. DeLucas∞
¶

Depts. of Biochemistry, Medicine*, Cell, Developmental and Integrative Biology§, and Optometry∞, and the Center for Biophysical Sciences and Engineering
University of Alabama at Birmingham, Birmingham, AL 35294 #

Birmingham Veterans Affairs Medical Center, Research Service, Birmingham, AL 35233

¶ To whom correspondence should be addressed: Dr. L. J. DeLucas [email protected]

1

Abstract The epithelial sodium channel (ENaC) plays a critical role in maintaining Na+ homeostasis in various tissues throughout the body. An understanding of the structure of the ENaC subunits has been developed from homology modeling based on the related acid sensing ion channel 1 (ASIC1) protein structure, as well as electrophysiological approaches. However, ENaC has several notable functional differences compared to ASIC1, thereby providing justification for determination of its three-dimensional structure. Unfortunately, this goal remains elusive due to several experimental challenges. Of the subunits that comprise a physiological hetero-trimeric αβγENaC, the α-subunit is unique in that it is capable of forming a homo-trimeric structure that conducts Na+ ions. Despite functional and structural interest in αENaC, a key factor complicating structural studies has been its interaction with multiple other proteins, disrupting its homogeneity. In order to address this issue, a novel protocol was used to reduce the number of proteins that associate and co-purify with αENaC. In this study, we describe a novel expression system coupled with a two-step affinity purification approach using NiNTA, followed by a GFP antibody column as a rapid procedure to improve the purity and yield of rat αENaC.

2

Introduction

The epithelial sodium channel (ENaC) is a member of the ENaC/Degenerin (DEG) superfamily[1]. ENaCs are expressed in multiple tissues throughout the body including the bladder, kidney, brain, and lung, among others. Physiologically, ENaC has been implicated in many different diseases and may be a therapeutic target in conditions such as acute respiratory distress syndrome (ARDS), deafness, cystic fibrosis, hypertension, inflammatory bowel disease, polycystic kidney disease, glioma, and many more [2]. Due to the many systems and diseases in which ENaC has been implicated, drug development would be significantly aided if a high-resolution structure of ENaC were available. However, to date no such structure has been reported. This is in marked contrast to the avian ASIC1 protein, an acid gated sodium channel, (and also a member of the Deg/ENaC superfamily), which has been successfully crystalized (cASIC1) [3-6]. Despite the overall low sequence identity of 17-20%, the extracellular domain of ENaC and cASIC1 exhibit 25% identity and 37% similarity. By using the ASIC1 structure as a model, ENaC’s finger, thumb, palm, knuckle, and β-ball domains have all been clearly defined [3], and a model of ENaC based on the structure of cASIC1 has been reported [7]. The function of ENaC and ASIC1 differ in several important ways. Alpha ENaC has a self-inhibition domain of 26 residues that is cleaved by proteases to become an active channel, a feature which is not found in ASIC1 [8, 9]. As the name suggests, ASIC1 is an acid sensing channel that is activated when the pH is reduced from 7.5 to 4-6 [10], while ENaC is generally considered constitutively active[1]. Another significant difference between ENaC and ASIC1 is their respective selectivity for Na+/K+, which in ASIC is ≤10, while in ENaC is >500 [1]. This difference is based in the selectivity filter where ENaC is mostly permeable to small monovalent cations (Na+, Li+, H+), which is close to the binding site for amiloride, a diuretic that inhibits a 3

physiological heterotrimeric αβγENaC with a Ki of ~ 100 nM [1, 5, 11, 12]. It is this sensitivity to the diuretic which characterizes the Deg/ENaC superfamily as “amiloride-sensitive sodium channels”. ENaC sensitivity to amiloride is approximately one hundred-fold greater than ASIC1 [1]. Additionally despite the availability of an ENaC model based on the ASIC1 structure, there are several regions in ENaC’s extracellular loop that have no sequence homology to ASIC1 yet are known to play a critical role in channel regulation [7]. These differences in activation and ion selectivity among others give credence to the pursuit of the ENaC structure. ENaC has been isolated from many cell types [2] and has been recombinantly expressed in different expression systems including Xenopus laevis oocytes [13], Saccharomyces cerevisiae [14], SF9 insect cells[15], rabbit reticulocyte cell extracts [16], MDCK [17], COS-7 [18], HEK293 [18], and CHO [19] cells. Members of the ENaC superfamily are glycoproteins generally between 530-740 amino acids in length, consisting of two transmembrane domains, two relatively small cytoplasmic domains predicted to be unstructured by the secondary structure predictor program Jpred[20], plus a large extracellular region that is largely ordered. Stoichiometrically, the arrangement of the ENaC superfamily has been under great debate. It is known that ENaC is a heteromeric channel consisting of at least three subunits designated α, β, and γ [21, 22]. There is another subunit, δ, which is functionally similar and interchangeable with the α subunit, however its physiological function is not clearly understood [23]. Initially the assembly of ENaC was thought to be a heterotetramer consisting of two α’s, with a single β and γ subunit [24-26], (other subunit arrangements consisting of eight to nine ENaC subunits have also been proposed [27-30]). However, resolution of the cASIC1 structure clearly reveals a trimeric channel [4]. To attain full activity of ENaC, co-expression of the three subunits α, β, and γ is required [22]. On the amino acid level, α, β, and γ ENaC are ~30% identical [22], whereas across 4

species human and rat ENaC orthologs are ~85% identical. Despite the requirement for all three α, β, and γ subunits to be expressed for ENaC to attain full activity, a homotrimeric αENaC channel was previously demonstrated to conduct sodium ions [16, 21]. Therefore, in an effort to reduce the complexity of producing purified, homogeneous protein to support the production of X-ray diffraction-quality crystals a homotrimeric α channel was pursued in these studies. Materials and Methods Cell culture and adaption to suspension Human embryonic kidney (HEK293) (Cat#11625-019, Life Technologies, Grand Island, NY, USA) and Chinese hamster ovary (CHO) (Cat#A11557-01, Life Technologies, Grand Island, NY, USA) cells were cultured as adherent monolayers at 37°C, 5% CO2 in DMEM-F12 supplemented with 10% fetal bovine serum (FBS), 100u/mL penicillin, 100u/mL streptomycin, 0.584 mg/mL glutamine, and 0.25 µg/mL Fungizone (Amphotericin B) (Life Technologies, Grand Island, NY, USA). For adaptation of ENaC transduced cells to suspension culture, subconfluent monolayer cultures were supplemented with an equal volume of CD4CHO/CD4HEK (Hyclone/GE, Logan, UT, USA)medium with 100u/mL penicillin, 100u/mL streptomycin, 0.584 mg/mL glutamine, 0.25 µg/mL Fungizone, and 2% FBS. After the monolayer culture reached confluency, the cells were dislodged by treatment with 0.25% trypsin-EDTA (Life Technologies, Grand Island, NY, USA), and placed in a flask comprising of the same medium, but with a reduced amount of FBS and incubated at 37°C. Every 4-5 days thereafter, for the next 6 weeks, 10% of the culture was placed into fresh medium with a gradual reduction of FBS to 0%. Suspension cells were maintained in 0% FBS media. If cells were transduced with lentivirus had its media supplemented with puromycin (5 ug/ml) to maintain gene insert selection. 5

GFP antibody production and immobilizing onto resin A monoclonal GFP antibody was generated by the University of Alabama at Birmingham’s Hybridoma Core Facility. BALB/c mice were immunized with purified GFP. The hybridoma was made with sensitized lymphocytes using standard protocols [31]. Hybridomas were grown in serum free media and the secreted antibodies were purified from the serum with protein G resin (GE Healthcare, Piscataway, NJ, USA). The purified GFP antibody yield was approximately 1mg/dL of medium. Purified GFP antibody was covalently bound to the free carbonyl group of Actigel ALD (Stereogene, Carlsbad, CA, USA). The GFP antibody resin was subsequently stored in 20% ethanol at 4°C. ENaC expression vector design Recombinant α rat ENaC (αrENaC) was expressed in HEK293 and CHO cells using a teton gene expression system [32] to obviate toxicity of overexpressing a constitutively active sodium channel. PCR amplified cDNA of the full-length αENaC open reading frame (orf) was ligated into a custom HIV-1-based lentiviral vector (designated K3843) comprising LTR-Psi-RREcts/ppt-TRE-multicloning site (BamHI and XhoI), tobacco etch virus (TEV) protease cleavage site (ENLYFQG) [33], enhanced green fluorescent protein (EGFP), [34] containing an A206K mutation to minimize self-dimerization [35], a 10x His-tag, a ubiquitin (Ubq) promoter, the puromycin resistance gene (puro), wpre [36], and ∆U3 LTR (Figure 1). Each of these specific vector genetic elements has been previously described[37]. This vector (designated K3849) expresses αENaC in-frame with eGFP, TEV and the 10x His coding sequences under transcriptional control of the inducible TRE promoter. The eGFP fusion facilitates both monitoring of protein expression in doxycycline induced cells and biophysical analysis of purified ENaC [37]. The downstream constitutively active Ubq promoter and puromycin gene enable selection of vector 6

transduced cells without expression of ENaC. The sequence of the ENaC expression vector (K3849) was verified by nucleotide sequence analysis. The αENaC expression vector was packaged and pseudotyped with the amphotropic VSV-G envelope glycoprotein as described previously [38, 39]. Full length αrENaC (78.9 kDa) with eGFP (26.9 kDa) and 6 glycosylation sites (~15kDa [40]) comprises a fusion protein of approximately 123 kDa. Transfection of cells with ENaC Monolayer cultures of HEK293 or CHO cells were grown in six-well plates and transfected using the calcium phosphate DNA precipitation method [41]. Each well of cell monolayer was transfected with 4 µg of lentiviral vector plasmid, 4 µg of the pCMV∆R8.2 packaging plasmid, and 2 µg of pMD.G (VSV-G) plasmid [42, 43]. Culture supernatants were collected 60 hours post-transfection, clarified by low-speed centrifugation (3,000 rpm for 10 min), and filtered through 0.22 µm pore-size sterile filters. The packaged vector particles were concentrated by ultracentrifugation at 125,000 g for 2 h at 4°C using an SW41 (Beckman Inc) rotor. The reverse tetracycline transactivator (rtTAS-M2) [44] expressing, doxycycline (dox) inducible, cell line was transduced with 50 ul lentiviral vectors at 105 cells/well into 96-well plates, infected at 37°C for over night, and then add 100 ul DMEM/F12 medium with 10% FBS. Exogenous expression of ENaC ENaC transduced CHO cells were continuously cultured in roller bottles to maintain a density 1 - 5x106 cells/mL and >95% viability as determined by trypan blue staining. Once suspension cell culture volume exceeded 2L with a total of at least 8 billion cells, the suspension culture was diluted with 2L of media to ~2x106 cells/mL and added to a pre-calibrated CelliGen 310 bioreactor (Eppendorf, New Brunswick, NJ, USA) with a 14L sterile glass vessel with a 7

temperature, pH, and dissolved oxygen probe fed with compressed O2, N2, CO2, and air as well as a 10%(w:v) sodium bicarbonate solution. During the entire time the culture was in the bioreactor, the temperature, pH, and dissolved oxygen (DO) was continuously monitored and maintained at 37°C, 7.2, and 50%, respectively. Both pH and DO concentration were independently verified daily with pH strips and a DO meter (Mettler Toledo, Columbus, OH, USA). Agitation was maintained between 30-50rpm depending on cell culture volume. Cell density, size, and viability were monitored with the Cellometer. Once the cells reached ~4x106 cells/mL, an additional 4L of medium was added. When cell density reached ~4x106 cells/mL at an 8 L volume, an additional 2L of medium was added. Finally, at a cell density of ~6x106 cells/mL the glucose concentration was measured with a GluCell glucose meter (CESCO Bioengineering, Taichung, Taiwan, R.O.C.) and 500mL of medium was added and supplemented with doxycycline and glucose to give final concentrations of 2ug/mL and 500mg/dL, respectively. The cells were induced for 2024 hours with a small aliquot taken to monitor eGFP expression. At the end of the protein induction time course, the cells were pelleted in a centrifuge at 1000 g. From a ~10.5L bioreactor run, yields were approximately ~6x107 cells or ~150g pellet. Solubilization of αrENaC from whole CHO cells Pelleted cells were resuspended in 5-6mL/g of solubization buffer (20 mM Tris (pH 8), 300 mM NaCl, 5% glycerol) supplemented with 0.5x Roche Complete Protease Inhibitor Cocktail (Roche Diagnostics, Indianapolis, IN, USA) and a final concentration of 1% (w:v) of nDodecyl β-D-maltoside (DDM) (Anatrace, Maumee, OH, USA) and placed on a stirrer for 1hr at 4°C. If the solution changed viscosity during the solubilization due to release of nuclear DNA, 20 µg/ml DNase-I (Sigma-Aldrich, St. Louis, MO, USA) was added and the solution was sonicated with a 550 Sonic Dismembrator (Thermo Fisher Scientific, Waltham, MA, USA) until 8

no longer viscous. The solubilized αrENaC solution was centrifuged at 100,000g for 1hr at 4°C. A sample was collected for future analysis. Isolation of CHO cell membranes and solubilization of αrENaC Pelleted cells were resuspended in 5-6mL/g of hypotonic buffer (10mM HEPES (pH 7.2), 1mM EDTA), supplemented with 0.5x Roche Complete Protease Inhibitor Cocktail and placed in a pre-chilled N2 cell disrupter (Parr Instruments, Moline, Illinois, USA). N2 gas was added to a final pressure of 1000psi and allowed to sit at 4°C for 20 minutes on a stirrer before releasing the pressure. The resulting solution was supplemented with sucrose to a final concentration of 250mM and placed back in the N2 cell disrupter on a stirrer for 5 minutes before releasing the pressure. To ensure proper breakage, the lysed cells were stained with trypan blue and observed under a microscope. The subsequent lysate was centrifuged at 1000g for 10 minutes to pellet unbroken cells and organelles. In order to pellet the membrane, the supernatant was centrifuged at 100,000g for 1hr at 4°C. The resulting membrane pellets were resuspended with a Dounce homogenizer in a high salt wash buffer (20 mM Tris (pH 8), 500 mM NaCl) containing 0.5x Roche Complete Protease Inhibitor Cocktail and was centrifuged at 100,000g for 1hr at 4°C. The pellets were collected and stored in a -80°C freezer. The membrane pellets were resuspended with a Dounce homogenizer at 10mL/g in solubization buffer and a final concentration of 1% (w:v) of DDM and placed on a stirrer for 1hr at 4°C. The solubilized membranes were centrifuged at 100,000g for 1hr at 4°C. Samples were collected for future analysis. Purification of αrENaC After centrifugation, the lysate was mixed with NiNTA resin (Qiagen, Venlo, Limburg, Germany) pre-equilibrated with solubilization buffer and supplemented with imidazole to a final concentration of 60mM and batch bound for 4-16hrs. Next, the resin was loaded onto a low pres9

sure chromatography column and attached to an AKTA FPLC system (GE Healthcare, Piscataway, NJ, USA). The beads were thoroughly washed with 10 column volumes of wash buffer (20 mM Tris (pH 8), 150 mM NaCl, 75mM imidazole, and 1mM DDM), and eluted with 4 column volumes of elution buffer (20 mM Tris (pH 8), 150 mM NaCl, 1M imidazole, and 1mM DDM). Samples that were positive for αrENaC via SDS PAGE under GFP fluorescence were collected and batch bound onto pre-equilibrated GFP antibody (ab) resin with cleavage buffer (20 mM Tris (pH 8), 150mM NaCl, and 1mM DDM), overnight. Subsequently, the GFP ab column was washed twice with a high salt wash buffer (20 mM Tris (pH 8), 1M NaCl, and 1mM DDM). To elute αrENaC from the GFP ab column, an on-column cleavage was conducted in cleavage buffer, supplemented with 1:10 (w:w) TEV to αrENaC (estimated via UV extinction coefficient), and rocked overnight at 4°C. Additionally PNGase F (NEB, Ipswich, MA, USA) can be added to the cleavage buffer with TEV according to manufacturer recommendations to cleave the six N-linked glycosylations. The elution from the GFP ab column was run over a 1mL NiNTA resin column pre-equilibrated with cleavage buffer to remove TEV and any nonspecifically bound His containing proteins, collected, concentrated with Amicon Ultra-15 100,000 NMWL (Merck KGaA, Darmstadt, Germany) spin concentrators to a final volume ~0.5mL and run over a Yarra S-3000 size exclusion column (Phenomenex, Torrance, CA, USA) on an AKTA. Samples were collected at each step for future analysis. SDS PAGE with coomassie staining and Western blot analysis Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS PAGE) gels were either 7.5% hand poured gels or 4-20% gradient Mini-PROTEAN TGX gels (BioRad, Hercules, CA, USA). EZ-run protein markers (Thermo Fisher Scientific, Waltham, MA, USA) were used for calibration and gels were stained with GelCode Coomassie (Thermo Fisher Scientific, Waltham, 10

MA, USA). For western blots, a GFP monoclonal mouse antibody (Abgent, San Diego, CA, USA) or a penta-His HRP conjugated antibody (Qiagen, Venlo, Limburg, Germany) were used. An αrENaC polyclonal antibody directed against a peptide sequence corresponding to residues 79-101 (sequence NLMKGNREEQGLGPEPAAPQQPTC; generated in collaboration with Dr. M. Knepper, NIH) of human αENaC centering around the N-terminal cytoplasmic edge of the first transmembrane domain, was also used. Secondary anti-mouse and anti-rabbit antibodies were HRP-conjugated (Southern Biotech, Birmingham, AL, USA). Enhanced chemiluminescence (ECL) was detected either using X-ray film, or on a G:BOX Chemi XT4 (Syngene, Frederick, MD, USA) imager. Additionally eGFP was detected in SDS PAGE gels using the G:BOX fitted with eGFP filters. TEV protease expression and purification A plasmid containing TEV fused to MBP and a 6x His-Tag (pRK793 Addgene plasmid #8827) was expressed in E. Coli and purified on a NiNTA column as previously described [45] followed by a Superdex 75 column. TEV protease was concentrated to ~0.7mg/mL and stored as single use aliquots at -80°C. Screening with fluorescence size exclusion chromatography The fluorescence size exclusion chromatography (FSEC) setup was a modified version of what was previously reported by Gouaux’s group [46]. A Superose 6 or Superdex 200 10/30 SEC column (GE Healthcare, Piscataway, NJ, USA) was attached to a 1260 Infinity fluorescence detector FPLC (Agilent Technologies, Santa Clara, CA, USA) and equilibrated with 50mM Tris (pH 7.5), 200mM NaCl, 0.1% NaN3, and 0.03% DDM. Samples were filtered with a 0.1µM Anotop syringe filter (GE Healthcare, Piscataway, NJ, USA) before loading onto the FSEC. The SEC column was calibrated with various fluorescent conjugated proteins and a high molecular 11

weight calibration kit (GE Healthcare, Piscataway, NJ, USA) to develop a standard molecular weight curve. Patch clamp recordings of αrENaC Whole-cell current recordings were obtained using standard Amphotericin B (SigmaAldrich, St. Louis, MO, USA) perforated patch methodology at room temperature from cells grown on coverslips and mounted on a flow through chamber on the stage of a Leica DM IRB inverted microscope (Leica Microsystems, Heidelberg, Germany). Bath solution exchange was achieved using a pinch valve control system converging on an 8-1 manifold. Tips of borosilicate recording pipettes (5-7 mΩ) were back filled with pipette solution (150 mM KCl, 10mM HEPES, pH 7.2 (Tris/HCl)) then with the same solution containing ~0.2mg/ml Amphotericin B [47]. Currents were obtained using an Axopatch 200B patch clamp amplifier (Axon Instruments, Molecular Devices, USA) with voltage commands and data acquisition was controlled by Clampex software (pClamp 10, Axon Instruments) and digitized (Digidata 1440A interface, Axon Instruments) at a sampling frequency of 2 kHZ. Current-voltage relationships were obtained using a pulse protocol in which cells were stepped from a -40 mV holding potential from -100 to +80 mV in 20 mV increments for 250 ms. Mean currents were obtained from the average of 3 sweeps during the 200-250 ms period of each sweep using Clampfit software (Axon Instruments). Bath solutions contained 140mM NaCl, 4.0mM KCl, 1.8mM CaCl2, 1.0mM MgCl2, 10mM glucose, 10mM HEPES pH 7.4 (NaCl/HCl). Amiloride concentration was 10 µM. Appropriate vehicle controls were performed including a paired t-test control vs amiloride.

12

Results Expression of αrENaC Several constructs were designed and tested to optimize αrENaC expression and purification. A common problem with many of the earlier constructs was weak binding affinity of the His tag (Supplemental Figure 1a). We suspected that the high detergent concentration occluded/buried the His-tag in the detergent micelle preventing it from interacting with the NiNTA resin in the protein-detergent complex [48]. Therefore, constructs were designed to sequester the His tag from the detergent micelle. For example, constructs were made using eGFP as the epitope tag, or with a 10x His-tag placed on either the N or C-terminus of αrENaC. However, constructs with eGFP (in SF9 cells) or a SUMO-tag on the N-terminal were not properly expressed and solubilized with DDM. In contrast, placing eGFP at the C-terminal was compatible with αrENaC expression and subsequent solubilization with DDM (Supplemental Figure 1b). Additionally, it was noted that during the initial NiNTA purification step, when a sufficient amount of NiNTA resin was used, nearly all solubilized αrENaC would bind to the NiNTA with an insignificant amount of ENaC detected in the NiNTA flow through (Supplemental Figure 1a). This is in contrast with previous construct designs where the majority of the solubilized αrENaC remained in the NiNTA flow through. After evaluating many construct variations, the final construct used is shown in Figure 1. Although much of our early αrENaC expression studies were performed in HEK293 cells, the final construct was expressed in CHO cells and demonstrated to be amiloride-sensitive in patch clamp experiments (Supplemental Figure 2). As shown in Figure S2b, the observation of amiloride sensitive current is evidence for αrENaC activity in the plasma membrane of the induced CHO cells. Initially, when not induced to express αrENaC, the HEK293 cells grew well as 13

adherent cultures. However, two problems occurred as a result of attempts to scale-up expression via adaption from adherent cells to suspension cultures. First, when adapted as a suspension culture, much of the expressed αrENaC could not be solubilized from the plasma membrane (Supplemental Figure 3b). Additionally, the small amount that was solubilized consisted predominantly of monomeric protein (αrENaC 108kDa plus ~15kDa resulting from glycosylation and ~70kDa contributed by the DDM detergent[49] thus resulting in a total molecular weight of ~173kDa). These results suggest that most of the αrENaC produced was a DDM-insoluble aggregate. The second issue was that the adapted HEK293 suspension cells were not healthy. Viability was approximately 50-70% at the start of the adaptation process, but the majority of these cells would die in subsequent weeks. As seen in Supplemental Figure 3a, HEK293 suspension cells induced to express αrENaC are surrounded by cellular debris and not viable as determined by Trypan Blue staining (results not shown). After testing several other expression systems (including SF9 insect cells), it was determined that CHO cells were better suited for αrENaC expression. Figure 1 shows αrENaC expression in induced suspension CHO cells. The viability of these cells was >98% as determined with trypan blue staining. The differences between expression of αrENaC in HEK293 and CHO cells can be seen in the FSEC traces between the two expression systems (Supplemental Figure 3c). αrENaC Detergent Screen Detergents are typically used to extract membrane proteins from their lipid bilayer environment as well as throughout the purification process. . However, detergent solubilization of the membrane protein can cause partial protein unfolding or even complete denaturation. Utilizing a list of candidate detergents developed by Wiener [50], 70 different detergents were evaluated for their ability to extract αrENaC from adherent HEK293 cells. SDS PAGE gels were run on the 14

pellet and supernatant of the solubilized cells followed by immunoblot analysis (data not shown). Of the 70 detergents, 9 were determined to extract αrENaC from HEK293 adherent cells (Supplemental Table 1). These detergents were subsequently analyzed via FSEC to determine the molecular weights for the protein-detergent complexes. Of the 9 detergents tested, several maintained αrENaC as a trimer and/or a monomer on the FSEC based on a previously determined calibration curve for the column. αrENaC with GFP fusion and glycosylations is approximately a 130kD monomer (as determined with markers on a SDS-paged gel) or 390kD trimer plus a 70kD detergent micelle. Of the 70 detergents evaluated, one of the more promising detergents was DDM based on FSEC relative yield and homogeniety. The FSEC chromatogram of αrENaC is shown in Figure 2a. It should be noted that only peaks around the 28, 31, and 36 minutes have fluorescent eGFP proteins show in the eGFP fluorescence of Figure 2b, while all the other peaks expressed fluorescence that did not correspond to eGFP, but instead corresponded to the intrinsic fluorescence of solubilized CHO cells as shown in the control. Due to this artifact, future FSEC chromatograms focus on the portion of the FSEC run that contains the region of eGFP fluorescence. Different Cell Lysis Techniques Can Affect Protein Stability One point that should be considered when dealing with membrane proteins is the effects of the lysis process on protein stability. There exist many different techniques to lyse mammalian cells including sonication, exposure of cells to high detergent concentrations, and nitrogen decompression. To varying degrees all of these techniques worked in lysing αrENaC expressing CHO cells, but more importantly, the resulting condition of αrENaC is different. This is keenly be observed in in the FSEC chromatograms of Figure 3. In the whole cell sonication and solubilization (blue) versus solubilized (red) curve, the appearance of a peak around 36 minutes 15

is larger the whole cell sonicated and solubilized sample (blue). In turn, the isolated membranesolubilized sample has the smallest 36 minute peak (orange). This 36 minute peak corresponds to free eGFP (Figure 2b). By comparison, sonication is a harsh technique for cell lysis and for ENaC, appears to increase cleavage of the peptide bond between αrENaC and eGFP. However, the process of solubilizing whole cells on a stirrer is not as gentle as expected. Of the three techniques investigated, nitrogen decompression proved to be the best technique in that it was the least likely to destabilize ENaC protein. Total cells αrENaC vs Isolated Membrane αrENaC After lysing cells to purify the membrane protein, there exists the option of isolating membrane followed by solubilization in detergent or alternatively, directly adding detergent to the lysed cells resulting in solubilization of the entire cell contents. Although the latter protocol is much simpler and quicker (albeit, at the expense of requiring more detergent), the process of isolating membranes proved to be a valuable step in the purification of αrENaC. Figure 3 compares nitrogen decompression followed by solubilization of αrENaC from isolated membranes (orange) versus solubilized αrENaC from whole cell lysate (blue and red). There are two distinct populations of αrENaC in the whole cell lysate solubilization solution (regardless of the use of sonication) centering on the 27.7 minute and 31.1 minute peaks, while in the nitrogen decompression/solubilized isolated membranes, there is a large single peak around 29.7 minutes. According to our calibrated curve, the 29.7 minute peak elutes at approximately 300kD. Also, it must be noted that with the nitrogen decompression followed by solubilization of isolated membrane, there is a smaller peak before the 29.7 peak centering around 27.5 minutes. The end result is αrENaC extracted from purified membranes are more homogenous that αrENaC extracted from whole cells. 16

Salt Anions Can Affect the Stoichiometric Arrangement of αrENaC It has long been known[51] that the choice of salt anions and cations have an effect on protein stability. As shown in Figure 4, FSEC runs were conducted with different sodium based salts. Following a Hofmeister anion series Cl-> Br->ClO3->I->NO3-, different anions were used with Na+ where the progression of the anions increases a protein’s propensity to denature and therefore reduces protein stability [52]. Only NaBr is shown as the FSEC chromatographic traces of NaClO3, NaI, and NaNO3 were almost identical. The type anion used affected the stoichiometry for αrENaC in that with NaCl there were two species of protein; one centering on the 27.7 minute peak and one on the 31.1 minute peak. However, with Br- as the anion, the 27.7 minute peak begins to disappear while the 31.1 minute peak becomes sharper. This 31.1 minute peak corresponds to an approximate monomer (based on the definition of an active ENaC channel that only a trimer should be capable of conducting sodium ions, this peak is likely to be nonfunctional). Additionally, the appearance of a 36 minute peak corresponds to eGFP. Trading Cl- for other anions such as Br- allows αrENaC to become more homogenous at the expense of the potential formation of a functional channel. Another possible explanation for the presumed loss of function is based on the initial ASIC1 structure that showed a Cl- ion positioned at the binding interface of the subunits [3]. It is possible that replacement of Cl- with any other anion (especially larger anions) disrupts the inter-subunit packing. It has been shown that the activity of αβγENaC is affected by the use of different anions [53]. Purification of αrENaC A fundamental problem associated with αrENaC purification is its propensity to bind to many different proteins [54-56]. When purifying αrENaC in NaCl, there were many other proteins co-eluting with αrENaC in the NiNTA elution after an extensive wash with 75mM imidaz17

ole (Figure 5a). The 10x His-Tag located on the C-terminal domain of eGFP provides tight binding between the His-Tag and NiNTA. Imidazole concentrations as high as 100 mM can be used in the binding buffer, yet a majority of αrENaC remains bound to the NiNTA (data not shown). Likewise, in the elution process, 500 mM imidazole was not sufficient to completely remove αrENaC from NiNTA but instead required a 1M imidazole to elute most of the bound αrENaC. A key change in the NiNTA elution purity occurred when replacing NaCl with NaBr in the buffer. When using NaBr, αrENaC eluted from the NiNTA column was significantly purer, yielding only two major bands on the Coomassie gel as opposed to the numerous bands that can be seen with NaCl (Figure 5a). In the NaBr NiNTA elution profile, the top band is αrENaC, while the bottom band is actin (results confirmed via tandem mass spectrometry). Figure 5b utilizes the eGFP of the αrENaC fusion protein to monitor αrENaC in an SDS PAGE gel on a fluorescence imager. This figure shows that the TEV cleavage site was accessible to the TEV protease. The data also suggest that αrENaC has progressed through the trans-Golgi network, allowing it to be properly glycosylated (glycosylation is known to occur at six different sites) [57]. As shown in Figure 5B, this glycosylation can be cleaved by PNGase F. In addition to actin, there were many other proteins albeit in smaller quantities that necessitated additional purification steps. After developing a GFP antibody resin, the NiNTA elution was loaded on the GFP antibody resin column and washed with high salt. The salt wash helped remove many of the contaminating proteins including actin. However, eluting αrENaC off the column was problematic. Both antibody specific peptides and low pH 3 washes did not elute αrENaC from the column. As a result, on-column cleavage with TEV was used to successfully eluted αrENaC. Figure 5c shows a Coomassie gel and western blot using an alpha ENaC antibody after NiNTA to clean up the TEV protease followed by a final SEC (Yarra S-3000) purification step. Our purification of αrENaC ultimately

18

produced ~0.5mg of ~90% pure αrENaC as estimated by extinction coefficient and ImageJ respectively from a 150g pellet of CHO cells. Discussion The expression of ENaC subunits and fully active ENaC is well characterized in several species. To date, there is a lack of published protocols for the expression of milligram quantities of ENaC necessary to support crystallographic, NMR or EM structural analysis. Traditional high protein yield expression systems such as SF9 insect cells and E. coli, failed to express folded full-length soluble αrENaC in our experience (data not shown; SF9 cells did express αrENaC in our laboratory and as reported by another group [15], but it resulted in a high molecular weight aggregate as determined by FSEC). Utilization of the unique expression system described here resulted in production of αrENaC in both HEK293 and CHO cell lines. However, CHO cells proved superior for the desired scale up required to produce milligram quantities for subsequent structural studies. . One of the characteristics of a fully active ENaC channel is the proteolytic processing that occurs in the trans-Golgi network by furin and on the surface of the membrane by the serine protease family of channel activating proteases (CAPs)[58]. It is with these cleavages that αrENaC is able to attain full activity. In αrENaC, there are two cleavage sites for furin, (a protease in the Golgi), which releases an inhibitory tract of amino acids (Asp-206 - Arg-231). Adding back the cleaved peptide inhibits an active αrENaC channel in an amiloride independent manner[9]. One particular feature of the αrENaC purified here is the distinct lack of proteolytically cleaved channels (Figure 5b and 5c). With that being said, there is a small species of αrENaC in the eGFP fluorescence gel (Figure 5b) that is the result of some low occurrence cleavage events. However, we do not know if this is a result of degradation, furin, or a CAP. 19

Cleavage of αrENaC by furin would result intwo specific bands on a reduced SDS-PAGE, a large C-terminal domain of about 52.1kD and a smaller N-terminal domain about 23.9kD. There are several disulfide bonds that can maintain the connection between the N- and C-terminal domains[59]. In Figure 5c, the western blot to the N-terminal domain of alpha ENaC only shows a single band at the full-length 78.9kD (no 23.9kD band is visible). This result confirms what Kleyman et al. observed when only expressing αrENaC in CHO cells [19]. Without the β and γ subunits, α subunit is not subjected to proteolytic processing and as a result, exists as a single species on a gel. To observe physiologically active ENaC channels, αβγENaC subunits need to be expressed together, but due to the differential processing of the channels by proteases, it was previously shown that the sample is not homogenous [19]. Additionally, by using buffers containing NaCl, we show a large species of homotrimeric channels, but we also observe a significant amount of monomeric protein. In our purification, we decided to pursue a homogenous sample at the expense of physiological activity. Instead of expressing αβγENaC for a physiologically relevant channel, we expressed only αENaC with the goal being to produce uniformly processed channels. In addition, instead of using NaCl-based buffers to maintain a trimeric channel, we substituted a NaBr-based buffer to disassociate the ENaC timer thereby providing pure monomeric ENaC subunits. Furthermore, instead of isolating all expressed ENaC, we selectively isolated the membrane expressed fraction in order to obtain a single stoichiometric species. At each step, decisions were made to select for a uniformly processed monomeric subunit in an attempt to purify as homogenous a protein sample as possible at the expense of function. None-the-less the source of the αrENaC protein from the membranes did exhibit amiloride sensitive currents which suggests function (experiments performed using patch clamp technique). To support dif20

ferent studies such as crystallography, differential scanning calorimetry and dynamic light scattering, a homogenous sample is a necessity [60]. The question is, if we have pure channel subunits, can we reconstitute a channel? Once an homogenous channel is formed, it might be possible to restore partial function with the use of trypsin[58]. Another possible future strategy as a compromise between the ability to produce protein amenable to crystallization versus production of fully functional protein, would be pursuit of an αβγENaC channel with the furin sites and inhibitory tracts removed from the alpha[61] and gamma[62] subunits. However, as a necessary first step towards our overall goal of obtaining a high resolution structure, we produced a useful protocol for expression of milligram quantities of ~90% homogenous αrENaC. Acknowledgments This research was supported by the following NIH Grants NIGMS R01GM095639 and NIDDK R01DK037206, and the Virology, Genetic Sequencing and Flow Cytometry Cores of the UAB Center for AIDS Research (P30 AI27767), and the UAB CF Research & Translational Core Center (P30 DK072482). They antibodies were made by RDCC Analytical Imaging and Immunoreagent Core with supported from NIH P30 AR048311. We would also Casey Weaver for the use of his hybridoma and Suman Bharara and Kevin Macon for their help in culturing the hybridoma to produce the GFP antibody. Finally we would like to thank the late Dale Benos, without whom, none of this would have been done. References 1.

2. 3.

Kellenberger, S. and L. Schild, Epithelial Sodium Channel/Degenerin Family of Ion Channels: A Variety of Functions for a Shared Structure, ed. S. Kellenberger and L. Schild. Vol. 82. 2002. 735767. Qadri, Y.J., A.K. Rooj, and C.M. Fuller, ENaCs and ASICs as therapeutic targets. American journal of physiology. Cell physiology, 2012. 302(7): p. C943-65. Jasti, J., et al., Structure of acid-sensing ion channel 1 at 1.9 A resolution and low pH. Nature, 2007. 449(7160): p. 316-23.

21

4. 5. 6. 7.

8.

9.

10. 11. 12. 13.

14. 15.

16. 17. 18. 19. 20. 21. 22. 23.

24.

Gonzales, E.B., T. Kawate, and E. Gouaux, Pore architecture and ion sites in acid-sensing ion channels and P2X receptors. Nature, 2009. 460(7255): p. 599-604. Baconguis, I., et al., X-ray structure of acid-sensing ion channel 1-snake toxin complex reveals open state of a Na(+)-selective channel. Cell, 2014. 156(4): p. 717-29. Baconguis, I. and E. Gouaux, Structural plasticity and dynamic selectivity of acid-sensing ion channel-spider toxin complexes. Nature, 2012. 489(7416): p. 400-5. Kashlan, O.B., et al., Constraint-based, homology model of the extracellular domain of the epithelial Na+ channel alpha subunit reveals a mechanism of channel activation by proteases. J Biol Chem, 2011. 286(1): p. 649-60. Kashlan, O.B., et al., Allosteric Inhibition of the Epithelial Na+ Channel through Peptide Binding at Peripheral Finger and Thumb Domains. Journal of Biological Chemistry, 2010. 285(45): p. 3521635223. Carattino, M.D., et al., The Epithelial Na+ Channel Is Inhibited by a Peptide Derived from Proteolytic Processing of Its α Subunit. Journal of Biological Chemistry, 2006. 281(27): p. 1890118907. Krishtal, O.A. and V.I. Pidoplichko, A receptor for protons in the nerve cell membrane. Neuroscience, 1980. 5(12): p. 2325-2327. Benos, D.J., Amiloride: a molecular probe of sodium transport in tissues and cells. American Journal of Physiology - Cell Physiology, 1982. 242(3): p. C131-C145. Kashlan, O.B. and T.R. Kleyman, ENaC structure and function in the wake of a resolved structure of a family member. Am J Physiol Renal Physiol, 2011. 301(4): p. F684-96. Firsov, D., et al., Cell surface expression of the epithelial Na channel and a mutant causing Liddle syndrome: A quantitative approach. Proceedings of the National Academy of Sciences, 1996. 93(26): p. 15370-15375. Gupta, S.S. and C.M. Canessa, Heterologous expression of a mammalian epithelial sodium channel in yeast. FEBS Letters, 2000. 481(1): p. 77-80. Rao, U.S., A. Mehdi, and R.E. Steimle, Expression of amiloride-sensitive sodium channel: a strategy for the coexpression of multimeric membrane protein in Sf9 insect cells. Anal Biochem, 2000. 286(2): p. 206-13. Ismailov, I.I., et al., Triple-barrel Organization of ENaC, a Cloned Epithelial Na+ Channel. Journal of Biological Chemistry, 1996. 271(2): p. 807-816. Staub, O., et al., WW domains of Nedd4 bind to the proline-rich PY motifs in the epithelial Na+ channel deleted in Liddle's syndrome. Embo j, 1996. 15(10): p. 2371-80. Prince, L.S. and M.J. Welsh, Cell surface expression and biosynthesis of epithelial Na+ channels. Biochem J, 1998. 336 ( Pt 3): p. 705-10. Hughey, R.P., et al., Maturation of the epithelial Na+ channel involves proteolytic processing of the alpha- and gamma-subunits. J Biol Chem, 2003. 278(39): p. 37073-82. Cole, C., J.D. Barber, and G.J. Barton, The Jpred 3 secondary structure prediction server. Nucleic Acids Research, 2008. 36(suppl 2): p. W197-W201. Canessa, C.M., J.-D. Horisberger, and B.C. Rossier, Epithelial sodium channel related to proteins involved in neurodegeneration. Nature, 1993. 361(6411): p. 467-470. Canessa, C.M., et al., Amiloride-sensitive epithelial Na+ channel is made of three homologous subunits. Nature, 1994. 367(6462): p. 463-467. Haerteis, S., et al., The δ-Subunit of the Epithelial Sodium Channel (ENaC) Enhances Channel Activity and Alters Proteolytic ENaC Activation. Journal of Biological Chemistry, 2009. 284(42): p. 29024-29040. Firsov, D., et al., The heterotetrameric architecture of the epithelial sodium channel (ENaC). Vol. 17. 1998. 344-352. 22

25. 26. 27.

28. 29.

30. 31. 32. 33.

34. 35. 36. 37.

38. 39.

40. 41. 42.

43. 44.

Dijkink, L., et al., The epithelial sodium channel (ENaC) is intracellularly located as a tetramer. Pflügers Archiv, 2002. 444(4): p. 549-555. Anantharam, A. and L.G. Palmer, Determination of Epithelial Na+ Channel Subunit Stoichiometry from Single-Channel Conductances. The Journal of General Physiology, 2007. 130(1): p. 55-70. Snyder, P.M., et al., Electrophysiological and Biochemical Evidence That DEG/ENaC Cation Channels Are Composed of Nine Subunits. Journal of Biological Chemistry, 1998. 273(2): p. 681684. Eskandari, S., et al., Number of Subunits Comprising the Epithelial Sodium Channel. Journal of Biological Chemistry, 1999. 274(38): p. 27281-27286. Staruschenko, A., et al., Fluorescence Resonance Energy Transfer Analysis of Subunit Stoichiometry of the Epithelial Na+ Channel. Journal of Biological Chemistry, 2004. 279(26): p. 27729-27734. Staruschenko, A., et al., Epithelial Na+ channel subunit stoichiometry. Biophys J, 2005. 88(6): p. 3966-75. in Antibodies: A Laboratory Manual, second edition, E.A. Greenfield, Editor. 2014, Cold Spring Harbor Laboratory Press. Welman, A., J. Barraclough, and C. Dive, Tetracycline regulated systems in functional oncogenomics. Transl Oncogenomics, 2007. 2: p. 17-33. Kapust, R.B., et al., Tobacco etch virus protease: mechanism of autolysis and rational design of stable mutants with wild-type catalytic proficiency. Protein Engineering, 2001. 14(12): p. 9931000. Heim, R., A.B. Cubitt, and R.Y. Tsien, Improved green fluorescence. Nature, 1995. 373(6516): p. 663-664. Zacharias, D.A., et al., Partitioning of Lipid-Modified Monomeric GFPs into Membrane Microdomains of Live Cells. Science, 2002. 296(5569): p. 913-916. Zufferey, R., et al., Woodchuck hepatitis virus posttranscriptional regulatory element enhances expression of transgenes delivered by retroviral vectors. J Virol, 1999. 73(4): p. 2886-92. Hildebrandt, E., et al., Erratum to: A Stable Human-Cell System Overexpressing Cystic Fibrosis Transmembrane Conductance Regulator Recombinant Protein at the Cell Surface. Mol Biotechnol, 2015. 57(5): p. 406. Zufferey, R., et al., Self-Inactivating Lentivirus Vector for Safe and Efficient In Vivo Gene Delivery. Journal of Virology, 1998. 72(12): p. 9873-9880. Burns, J.C., et al., Vesicular stomatitis virus G glycoprotein pseudotyped retroviral vectors: concentration to very high titer and efficient gene transfer into mammalian and nonmammalian cells. Proceedings of the National Academy of Sciences, 1993. 90(17): p. 8033-8037. Kornfeld, R. and S. Kornfeld, Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem, 1985. 54: p. 631-64. Graham, F.L. and A.J. van der Eb, A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology, 1973. 52(2): p. 456-67. Naldini, L., et al., Efficient transfer, integration, and sustained long-term expression of the transgene in adult rat brains injected with a lentiviral vector. Proc Natl Acad Sci U S A, 1996. 93(21): p. 11382-8. Wu, X., et al., Development of a novel trans-lentiviral vector that affords predictable safety. Mol Ther, 2000. 2(1): p. 47-55. Urlinger, S., et al., Exploring the sequence space for tetracycline-dependent transcriptional activators: novel mutations yield expanded range and sensitivity. Proc Natl Acad Sci U S A, 2000. 97(14): p. 7963-8.

23

45. 46. 47. 48.

49. 50.

51. 52. 53. 54. 55. 56. 57. 58. 59.

60. 61. 62.

Tropea, J.E., S. Cherry, and D.S. Waugh, Expression and purification of soluble His(6)-tagged TEV protease. Methods Mol Biol, 2009. 498: p. 297-307. Kawate, T. and E. Gouaux, Fluorescence-Detection Size-Exclusion Chromatography for Precrystallization Screening of Integral Membrane Proteins. Structure, 2006. 14(4): p. 673-681. Akaike, N. and N. Harata, Nystatin perforated patch recording and its applications to analyses of intracellular mechanisms. Jpn J Physiol, 1994. 44(5): p. 433-73. Vergis, J.M. and M.C. Wiener, The variable detergent sensitivity of proteases that are utilized for recombinant protein affinity tag removal. Protein Expression and Purification, 2011. 78(2): p. 139-142. VanAken, T., et al., Alkyl glycoside detergents: synthesis and applications to the study of membrane proteins. Methods Enzymol, 1986. 125: p. 27-35. Vergis, J.M., M.D. Purdy, and M.C. Wiener, A high-throughput differential filtration assay to screen and select detergents for membrane proteins. Analytical biochemistry, 2010. 407(1): p. 111. Hofmeister, F., Zur Lehre von der Wirkung der Salze. Archiv für experimentelle Pathologie und Pharmakologie, 1888. 24(4-5): p. 247-260. Collins, K.D., G.W. Neilson, and J.E. Enderby, Ions in water: Characterizing the forces that control chemical processes and biological structure. Biophysical Chemistry, 2007. 128(2–3): p. 95-104. Collier, D.M. and P.M. Snyder, Extracellular Chloride Regulates the Epithelial Sodium Channel. Journal of Biological Chemistry, 2009. 284(43): p. 29320-29325. Meltzer, R.H., et al., Heteromeric assembly of acid-sensitive ion channel and epithelial sodium channel subunits. The Journal of biological chemistry, 2007. 282(35): p. 25548-59. Berdiev, B.K., Y.J. Qadri, and D.J. Benos, Assessment of the CFTR and ENaC association. Mol Biosyst, 2009. 5(2): p. 123-7. Cantiello, H.F., et al., Actin filaments regulate epithelial Na+ channel activity. Am J Physiol, 1991. 261(5 Pt 1): p. C882-8. Snyder, P.M., et al., Membrane topology of the amiloride-sensitive epithelial sodium channel. J Biol Chem, 1994. 269(39): p. 24379-83. Rossier, B.C. and M.J. Stutts, Activation of the Epithelial Sodium Channel (ENaC) by Serine Proteases. Annual Review of Physiology, 2009. 71(1): p. 361-379. Firsov, D., et al., Mutational Analysis of Cysteine-rich Domains of the Epithelium Sodium Channel (ENaC): IDENTIFICATION OF CYSTEINES ESSENTIAL FOR CHANNEL EXPRESSION AT THE CELL SURFACE. Journal of Biological Chemistry, 1999. 274(5): p. 2743-2749. McPherson, A., Introduction to protein crystallization. Methods, 2004. 34(3): p. 254-265. Sheng, S., et al., Furin cleavage activates the epithelial Na+ channel by relieving Na+ selfinhibition. Am J Physiol Renal Physiol, 2006. 290(6): p. F1488-96. Passero, C.J., et al., Defining an inhibitory domain in the gamma subunit of the epithelial sodium channel. American Journal of Physiology - Renal Physiology, 2010. 299(4): p. F854-F861.

Figure Legends Figure 1. αrENaC Construct Design and Expression (top) The overall genetic design of the lentiviral vector is depicted. The abbreviations are as follow: LTR, long terminal repeats; ψ, Psi packaging element; RRE, Rev response element; cts, central termination sequence; ppt, polypurine tract; TRE, Tet response element; TEV, tocabo etch virus; GFP, enhanced green fluorescent protein; His, 10x histidine tag; Ubq, ubiquitin promoter; Puro, puromycin resistant gene; wpre, woodchuck hepatitis virus posttranscriptional regu24

latory element. (bottom) Image of doxycycline induced αrENaC expression in suspension CHO cells obtained with a Zeiss Laser Scanning 710 Confocal Microscope showing a bright field image overlaid with an eGFP image. Figure 2. αrENaC solubilization with DDM (a) Whole suspension CHO cell sonicated lysates were solubilized with DDM and were injected on a FSEC running a Superose 6 10/30 column (blue). Additionally, a control was run with whole suspension non-transduced CHO cell sonicated lysate (grey). Fractions corresponding to the color bars below the FSEC peaks were run on SDS PAGE and analyzed for eGFP fluorescence with the Syngene G:BOX Chemi XT4(b). The size markers on the FSEC chromatogram correspond to running the following purified protein samples to observe retention time on the column: Blue Dextran (2000kDa), Thyroglobulin (669kDa), Ferritin (440kDa), R-phycoerytherin (240kDa), Fluoroscene Fab (100kDa), and Free-GFP (28kDa). The S on the SDS PAGE represents the solubilized lysates injected into the FSEC. Figure 3. FSEC of Differential Cell Lysis and Solubilization of αrENaC with DDM FSEC running a Superose 6 10/30 column injected with the whole suspension CHO cell sonicated and solubilized αrENaC with DDM (blue), whole suspension CHO cell solubilized αrENaC with DDM on a stirrer (red), and isolated CHO cell membrane using nitrogen decompression and solubilized αrENaC with DDM (orange). Figure 4. Salt Anions and αrENaC Stoichiometry FSEC running a Superose 6 10/30 column injected with whole suspension αrENaC CHO cell sonicated and DDM solubilized lysate with DDM using a NaCl based buffer (blue) and a NaBr based buffer (purple). Figure 5. αrENaC Purification (a) 4-20% gradient SDS PAGE stained with Coomassie Blue after concentrating the NiNTA elutions with NaCl based buffers on the left and a 7.5% SDS PAGE stained with Coomassie Blue after concentrating the NiNTA elutions with NaBr based buffers on the right. On the NaBr gel, both αrENaC and actin were verified via mass spectrometry on the cut gel bands. The gels were imaged on a desktop scanner. (b) Concentrated NiNTA elutions were digested with PNGase, TEV, or both were run on a 7.5% SDS PAGE stained with Coomassie Blue on the left. The Coomassie stained gel was imaged with an Li-Cor Odyssey imaging system which reads the infrared signal from Coomassie and the eGFP fluorescence was imaged with a Syngene G:BOX Chemi XT4 with eGFP filters. (c) αrENaC after all purification steps including the GFP antibody column and SEC was run on a 4-20% gradient SDS PAGE and stained with Coomassie (left) or run on a western blot probed with an anti-αENaC ab using ECL on the Syngene G:BOX Chemi XT4 (right).

25

Supplemental Figure 1. αrENaC Construct Iterations (a) Elution of solubilized αrENaC expressing adherent HEK293 cells lysed with DDM following batch binding to NiNTA resin on a stirrer (S). Ft refers to the flow through fraction. Eluates were analyzed using a 7.5% SDS PAGE and immunoblot probing with the anti-His ab for the initial constructs without eGFP (left) and with the anti-GFP ab for the C-terminal GFP construct (right) using ECL on X-ray film. (b) Testing the solubility αrENaC in DDM with whole adherent HEK293 cells on a stirrer with eGFP tags on the C- or N- terminal of the construct. Lysates were analysed using a 7.5% SDS PAGE and immunoblot probing with anti-GFP using ECL on film. Supplemental Figure 2. αrENaC Sensitive Current in CHO Cells Current voltage (IV) relationships from whole cell recordings from alpha ENaC infected CHO cells. Uninduced (a) and following 24 hour doxycycline induction (b) in the absence (square) and presence (circle) of 10 uM amiloride. * represents p<0.05. (c) Amiloride sensitive currents expressed as percent inhibition of basal current at -80mV. Data derived from IV relationships in (a) and (b). *** represents p<0.0005. Supplemental Figure 3. αrENaC Construct in Mammalian Cells (a) Image of doxycycline induced αrENaC expression in suspension HEK293 cells using a Zeiss Laser Scanning 710 Confocal Microscope with a bright field image overlaid with an eGFP image. (b) FSEC running a Superdex 200 10/30 column injected with whole HEK293 cell solubilized membrane with DDM on a stirrer of either adherent (light green), adapted to suspension (dark green), or uninfected suspension HEK293 cells (gray). Fractions were collected at the FSEC peaks corresponding to 20.6 and 25 minutes. Those fractions were run on a 7.5% SDS PAGE gel and immunoblotted with an anti-His ab. The blot was visualized with ECL on X-ray film. (c) FSEC running on a Superdex 200 10/30 column injected with suspension CHO (blue) or HEK293 (dark green) whole cell solubilized membranes with DDM on a stirrer, or uninfected adherent HEK293 cells with DDM on a stirrer (gray).

Supplemental Table 1. List of Detergents that Solubilize αrENaC

26

27

28

30

31

32

We have expressed milligram quantities of the membrane protein Epithelial Sodium Channel, ENaC, in mammalian cells. We have purified milligram quantities of ENaC. We utilized the GFP fusion as a purification tag on a GFP antibody column.

33