Structural and biochemical studies of human membrane transporters require milligram quantities of stable, intact, and homogeneous protein. Here we describe scalable methods to screen, express, and purify human solute carrier transporters using codon-optimized genes.
Solute carriers (SLCs) are membrane transporters that import and export a range of endogenous and exogenous substrates, including ions, nutrients, metabolites, neurotransmitters, and pharmaceuticals. Despite having emerged as attractive therapeutic targets and markers of disease, this group of proteins is still relatively underdrugged by current pharmaceuticals. Drug discovery projects for these transporters are impeded by limited structural, functional, and physiological knowledge, ultimately due to the difficulties in the expression and purification of this class of membrane-embedded proteins. Here, we demonstrate methods to obtain high-purity, milligram quantities of human SLC transporter proteins using codon-optimized gene sequences. In conjunction with a systematic exploration of construct design and high-throughput expression, these protocols ensure the preservation of the structural integrity and biochemical activity of the target proteins. We also highlight critical steps in the eukaryotic cell expression, affinity purification, and size-exclusion chromatography of these proteins. Ultimately, this workflow yields pure, functionally active, and stable protein preparations suitable for high-resolution structure determination, transport studies, small-molecule engagement assays, and high-throughput in vitro screening.
Membrane proteins have long been targets for researchers and pharmaceutical industries alike. Of these, the solute carriers (SLCs) are a family of over 400 secondary transporter genes encoded within the human genome1. These transporters are involved in the import and export of numerous molecules, including ions2, neurotransmitters3, lipids4,5,6,7, amino acids8, nutrients9,10,11, and pharmaceuticals12. With such a breadth of substrates, these proteins are also implicated in a range of pathophysiologies through the transport of toxins13, transport of and inhibition by drugs of abuse14,15, or deleterious mutations16. Bacterial homologs have served as prototypes for the fundamental transport mechanism of several SLC families17,18,19,20,21,22,23,24,25. In contrast to human proteins, prokaryotic orthologs are often better expressed in the well-understood Escherichia coli expression system26,27 and are more stable in the smaller detergents which yield well-ordered crystals for X-ray crystallography28. However, sequence and functional differences complicate the use of these distantly-related proteins for drug discovery29,30. Consequently, direct study of the human protein is often needed to decipher the mechanism of action of drugs targeting SLCs31,32,33,34,35. While the recent advances in Cryo-electron Microscopy (Cryo-EM) have enabled structural characterization of SLCs in more native-like conditions36,37, difficulty in expressing and purifying these proteins remains a challenge for developing targeted therapeutics and diagnostics.
To alleviate this challenge, the RESOLUTE consortium (re-solute.eu) has developed resources and protocols for the large-scale expression and purification of human SLC-family proteins38. Starting with codon-optimized genes, we have developed methods for the high-throughput cloning and screening of SLC constructs. These methods were systematically applied to the whole family of SLCs, the genes were cloned into the BacMam viral expression system, and the protein expression was tested in human cell lines39 based on previously described methods for high-throughput cloning and expression testing40. In summary, the SLC gene is cloned from the pDONR221 plasmid into a pHTBV1.1 vector. This construct is subsequently used to transpose the gene of interest into a bacmid vector for transfecting insect cells, which includes a cytomegalovirus promoter and enhancer elements for expression in mammalian cells. The resulting baculovirus can be used to transduce mammalian cells for the expression of the target SLC protein.
We further developed standardized methods for large-scale expression and stable purification of selected SLCs (Figure 1). This protocol includes multiple checkpoints to facilitate effective troubleshooting and minimize variability between experiments. Notably, routine monitoring of protein expression and localization, as well as small-scale optimization of purification conditions for individual targets, were aided by Strep and Green Fluorescent Protein (GFP) tags41,42.
Ultimately, these chemically pure and structurally homogeneous protein samples can be used for structural determination by X-ray crystallography or Cryo-Electron Microscopy (Cryo-EM), biochemical target-engagement assays, immunization for binder generation, and cell-free functional studies via reconstitution into chemically defined liposomes.
NOTE: All codon-optimized RESOLUTE SLC genes have been deposited into AddGene43, the links to which are available on the list of RESOLUTE public reagents44. These genes have been cloned into the pDONR221 plasmid and allow direct cloning of the genes into the destination vector using recombination cloning45. To maximize parallelism, bacterial, insect, and mammalian cells are grown in block format for bacmid production (section 3), baculovirus amplification (section 5), and expression testing (section 6), respectively. For these steps, a micro-expression shaker is required to ensure sufficient mixing and aeration.
1. (High-throughput) cloning of SLCs into pHTBV1.1 bacmid
NOTE: The cloning step uses a recombination cloning protocol for efficient cloning and transformation into Escherichia coli (E. coli) using the heat-shock method46. The protocol is designed for high-throughput and parallel cloning of multiple targets or constructs but can be readily adapted to smaller scales.
2. Transposition
NOTE: The following steps are used to transpose the SLC genes from the pHTBV1.1 vector into a bacmid for BacMam baculovirus generation in Sf9 cells. Using the heat-shock method46, the pHTBV1.1 vector is transformed into DH10Bac competent E. coli cells, which contain a parent bacmid with a lacZ-mini-attTn7 fusion. Transposition occurs between the elements of the pHTBV1.1 vector and the parent bacmid in the presence of the transposition proteins provided by a helper plasmid48. See Table 2 for the composition of solutions used in this protocol.
3. High-throughput bacmid production
NOTE: The protocol describes the steps for extracting bacmids using a 96-well bacmid purification kit.
4. Transfection
NOTE: These steps are used to transfect Sf9 insect cells with the bacmid produced, which causes the insect cells to generate baculovirus particles (P0).
5. BacMam baculovirus amplification
NOTE: The following steps are used to amplify the initial P0 baculovirus to higher titer viral stocks; namely P1, P2, and P3. The final P3 titer is appropriate for transduction and protein expression. For efficiency and parallelism, this protocol uses fixed volumetric ratios for viral amplification, which have been empirically optimized. However, if the subsequently transduced cells do not show GFP fluorescence and increased cell diameter by microscopy or if protein expression fails (see sections 6 and 8), baculovirus amplification should be re-optimized for a low multiplicity of infection at each step after quantifying the baculovirus titer49,50,51,52, and infection monitored by GFP fluorescence microscopy and increased cell diameter53.
6. Transduction for expression testing
NOTE: The following section describes small-scale expression testing and can be modified for parallel testing of multiple constructs using deep well blocks.
7. High-throughput small-scale test purification
NOTE: The following steps describe a rapid test purification workflow in a 24-well block format for screening the expression levels of individual SLCs. See Table 2 for the composition of solutions used in this protocol.
8. Transduction for large-scale expression
NOTE: The following steps are the standard RESOLUTE protocol for SLC expression. Individual targets will require further optimization for the expression time, incubation temperature, and concentration of sodium butyrate. Further, we routinely optimize the baculovirus multiplicity of infection by testing various volumetric ratios of the P3 virus used to infect the suspension-adapted HEK293 cells in small-scale experiments. This is time efficient, uses techniques and equipment already at hand, and directly evaluates the desired experimental output. However, this empirical method requires re-optimization with each amplification of the P3 virus, and other methods are available to quantify the baculovirus particles49,50,51,52.
9. Protein purification
NOTE: The following is the standard RESOLUTE method for SLC purification for 5 L of cell culture. For each SLC target, the optimal detergent must be determined empirically. Prepare base buffer, detergent stock solution, wash, elution, and SEC buffers in advance (Table 2). For a list of the standard detergents tested, see Table 3. ATP and MgCl2 in the wash buffer reduce contamination by heat-shock proteins.
SLC genes can be cloned from RESOLUTE pDONR plasmids into BacMam vectors for mammalian expression
The described protocols for cloning, expression, and purification have proven successful for many SLC transporters across multiple protein folds. Nevertheless, the procedures include several checkpoints for monitoring progress, allowing for optimization to account for differences in expression, protein folding, lipid- and detergent-dependent stability, and sensitivity to buffer conditions.
Checkpoints during SLC cloning and small-scale expression
In the cloning steps, agarose gel electrophoresis should be used to ensure the correct size of the PCR and digestion products. Similarly, the Gateway and transposition reactions can be validated with a colony PCR reaction (Figure 2A,B). Baculovirus generation can be monitored using standard techniques as necessary49,50,51,52. The initial expression should be done at a small scale, evaluating protein yield by SDS-PAGE (Figure 2B). Similarly, the fraction of green fluorescent cells, total protein expression, and protein localization should be noted using fluorescence microscopy (Figure 2C,D). Protein expression should be optimized for cell type, temperature, time, and the necessity of co-expressing chaperones or complex partners. The expression can be further optimized by modifying the construct to truncate disordered N- and C-termini, based on secondary structure prediction56,57,58, and testing the types and placement of affinity tags. Protein stability should be evaluated at a small scale by FSEC (Figure 2E), SEC-based thermal shift assay (SEC-Ts), and DSF41,42,59,60,61,62. Small molecules, such as substrates and inhibitors, detergents, cholesterol hemisuccinate, lipids, and pH should be tested for improving protein stability considering the protein's function and native subcellular environment and subsequent purification buffers modified accordingly. In both small- and large-scale expression setups, cells should be monitored using microscopy for viability and contamination.
Optimization of transporter purification at large scale
Each step of large-scale protein purification should be evaluated by SDS-PAGE, including in-gel fluorescence to specifically monitor the GFP-tagged protein and enzymatic removal of that tag. In practice, the GFP-tagged SLC-expressing cells appear yellowish-green. After Twin-Strep-tag chromatographic elution, the eluent containing the purified protein appears fluorescent neon-green under white light. Chemically and structurally homogeneous protein should yield a single monodisperse A280 peak during size-exclusion chromatography (Figure 2F,G), and should show a single band on SDS-PAGE. The SDS-PAGE band corresponding to the expected SLC, and any unexpected bands, should be analyzed using tryptic digestion mass-spectrometry. Multiple bands on the SDS-PAGE gel indicate either proteolytic degradation, contaminating proteins, or SDS-resistant oligomers. Contaminating proteins may be removed by increasing the NaCl concentration of the solubilization buffer or changing the affinity tag. Proteolysis can be limited by improving the protein's purity, ensuring all steps are done at 4 °C or on ice, and optimizing the protocol to minimize the time of each step. If the SEC profile has a broad peak, multiple peaks, or large void peak (such as the purple trace of Figure 2F), the construct and purification conditions should be optimized at a small scale using FSEC, SEC-Ts, or DSF41,42,59,60,61,62.
Figure 1: Schematic of RESOLUTE workflow for SLC expression and purification. Step-by-step illustration of recombination cloning, BacMam baculovirus preparation, protein expression and purification, and downstream applications. Abbreviations: SLC = solute carrier; Cryo-EM = cryo-electron microscopy; SEC = size exclusion chromatography. Please click here to view a larger version of this figure.
Figure 2: Representative results for SLC expression and purification. (A) Colony PCR of high-throughput SLC cloning into pHTBV1.1-C3CGFP-SIII-10H-GTW. (B) Coomassie-stained SDS-PAGE of single small-scale, parallel expression test of 24 different full-length SLCs. (C) In-cell fluorescence of a GFP-tagged SLC localizing primarily to the plasma membrane. (D) In-cell fluorescence of a GFP-tagged SLC with significant intracellular localization. (E) Representative FSEC traces for four SLCs resolved on a hydrophilic, neutral silica-based UHPLC column. Representative SEC traces for six SLCs purified on either a (F) dextran-agarose or (G) agarose size exclusion chromatography columns. Molecular weights of the SLC complex and detergent used for purification are indicated where the oligomeric state has been experimentally determined. Abbreviations: SLC = solute carrier; GFP = green fluorescent protein; FSEC = fluorescence-detection size exclusion chromatography. Please click here to view a larger version of this figure.
Figure 3: Downstream applications of purified SLCs. (A) Micrograph of SLC1A1 in detergent. (B) 2D class averages of SLC1A1 in detergent. (C) Raw fluorescence of CPM thermal denaturation assay of SLC10A6 incubated with various concentrations of Taurolithocholic acid 3-sulfate. (D) First derivative of CPM thermal denaturation of SLC10A6. The SLC10A6's melting temperature increased by 10 °C in the presence of 120 µM Taurolithocholic acid 3-sulfate. Abbreviations: SLC = solute carrier; CPM = N-[4-(7-diethylamino-4-methyl-3-coumarinyl)phenyl]maleimide. Please click here to view a larger version of this figure.
Vector Name | Antibiotic markers | Tags for Purification | Screening Primers | bp added during PCR | ||
pHTBV1.1-C3CGFP-SIII-10H-GTW | AmpR | C-terminal | pHTBV-F (CTATAGACTCTATAGGCACACC) | ~350 | ||
3C protease site GFP | ||||||
Twin-Strep | GFP-R (CTGTCGTACAGATGAACTTCAAGGTC) | |||||
His10 | ||||||
pDONR221 | KanR | none | M13 Forward | ~190 | ||
M13 Reverse |
Table 1: Plasmids used for RESOLUTE cloning and BacMam generation.
Solution | Composition |
DH10Bac selection plates | LB-agar plates containing 50 µg/mL kanamycin, 10 µg/mL tetracycline, 7µg/mL gentamycin, 40 µg/mL IPTG, and 100 µg/mL bluo-gal |
2x LB medium (with antibiotics) | 2x LB broth containing 50 µg/mL kanamycin, 10 µg/mL tetracycline, and 7 µg/mL gentamycin |
HT Lysis buffer | 50 mM HEPES-NaOH, pH 7.5, 250 mM NaCl, 5% glycerol, EDTA-free protease inhibitor |
HT Wash buffer | 50 mM HEPES-NaOH, pH 7.5, 250 mM NaCl, 5% glycerol, 0.03% DDM/0.003% CHS |
HT Elution buffer | 50 mM HEPES-NaOH, pH 7.5, 250 mM NaCl, 5% glycerol, 0.03% DDM/0.003% CHS, 100 mM D-Biotin |
Base buffer | 50 mM HEPES-NaOH, pH 7.5, 150 mM NaCl |
Detergent stock solution | 10% (w/v) detergent, with or without 1% CHS as appropriate. Mix at 4 °C overnight and store at -20 °C. |
Strep wash buffer | 50 mM HEPES-NaOH, pH 7.5, 150 mM NaCl, detergent at 3-fold CMC, 10 mM MgCl2, 1 mM ATP |
Strep elution buffer | 50 mM HEPES-NaOH, pH 7.5, 150 mM NaCl, 100 mM D-Biotin, detergent at 3-fold CMC |
Size exclusion chromatography (SEC) buffer | 20 mM HEPES-NaOH, pH 7.5, 150 mM NaCl, detergent at 2-fold CMC. Filter through a 0.22 µM membrane. |
Sodium butyrate | 1 M solution in DPBS, store at -20 °C for long-term use. |
Table 2: A list of solutions used in this protocol and their composition.
Detergent system | Extraction concentration | Purification concentration (% w/v) |
DDM | 1% | 0.03% |
DDM + CHS | 1% + 0.1% | 0.03% + 0.003% |
DM | 1% | 0.25% |
DM+CHS | 1% + 0.1% | 0.25% + 0.025% |
NG | 1% | 0.60% |
OG | 1.5% | 1.50% |
LDAO | 1% | 0.07% |
LDAO + CHS | 1% + 0.1% | 0.07% + 0.007% |
C12E8 | 1% | 0.015% |
C12E8 + CHS | 1% + 0.1% | 0.015% + 0.0015% |
C12E9 + CHS | 1% + 0.1% | 0.01 + 0.001% |
CYMAL-5 | 1% | 0.40% |
LMNG | 1% | 0.005% |
LMNG + CHS | 1% + 0.1% | 0.005% + 0.0005% |
GDN | 1% | 0.02% |
Digitonin | 0.05% | |
OGNG + CHS | 1% + 0.1% | 0.18% + 0.018% |
C12E10 + CHS | 1% + 0.1% | 0.04% + 0.004% |
Fos Choline-12 | 1% | 0.14% |
Table 3: Standard detergents used to test membrane solubilization and SLC monodispersity and stability.
The development of SLC-targeting therapies has remained hampered due to the absence of systematic characterization of transporter function. This has led to disproportionally fewer drugs targeting this protein class relative to GPCRs and ion channels63, despite their numerous roles in normal and pathophysiological processes. RESOLUTE is an international consortium aimed at developing cutting-edge research techniques and tools to accelerate and improve current SLC research. As a part of RESOLUTE, we have developed these protocols for efficient cloning, construct screening, and large-scale expression and purification of human SLCs.
Here we describe scalable SLC cloning and expression methods that were successfully used to systematically explore the human SLC transporters, including putative and orphan SLCs. Notably, SLCs purified in this manner have been successfully used in subsequent studies of transporter structure, biochemistry, function, binder generation, and small-molecule binding. We regularly employ this method to purify milligram quantities of various SLCs and under optimal conditions, the entire protocol, including the cloning and tissue culture steps, can be completed in 4-5 weeks.
Our method is optimized for economy and parallelism to systematically evaluate multiple targets. However, this high-throughput method is also readily adapted to the parallel generation of constructs for a single target with various truncations or tags by using distinct cloning primers or vectors. This is similar to methods that also optimize multiple constructs for a target64, though our protocol offers further efficiencies with parallel cloning, baculovirus generation, and expression testing. Transfection offers a shorter time between construct cloning and expression by forgoing the baculovirus generation65 but is significantly more expensive and laborious for large-scale expression. In contrast, stable cell lines are likely less expensive for large-scale expression66, but generating highly-expressing clonal cell lines can require more time and specialized resources. Finally, while this protocol uses human cell lines for protein production, insect cells line such as from Spodoptera frugiperda and Trichoplusia ni have also been successfully used for large-scale SLC expression5,31,64. Expression in human cell lines increases media costs but offers more native-like post-translation modifications and lipid environment39,67.
While the protocol can be adapted for different membrane transporters and experimental needs, several factors influence the quality and yield of the purified protein samples. Though it is ideal to study full-length proteins, some degree of sequence truncation may be required to achieve better expression, purification, and reconstitution yields. All RESOLUTE SLC constructs have been tagged with a cleavable GFP, which is valuable in monitoring SLC expression, cellular localization, and purification. The suspension-adapted HEK293 cell expression system used in these experiments has led to superior yields and is recommended, although we routinely also produce proteins without complex glycosylation via the suspension-adapted HEK293 GnTl- cell line. The incubation temperature and length for protein expression by transduced cells should be optimized for each target, though we have found 72 h at 30 °C to be a good default.
All protein purification steps should be carried out on ice or at 4 °C and once the purified protein has been snap-frozen, freeze-thaw cycles should be avoided. The type and amount of the detergents used in membrane solubilization and purification buffers are critical and should be determined for each SLC empirically.
The SLCs purified with this method yield homogeneous and structurally and functionally intact samples, which can be used for a variety of biochemical and biophysical studies. Observing single, discrete particles of the solubilized and purified SLC protein by negative stain and Cryo-EM (Figure 3A,B) can be promising for subsequent structure determination37. The purified SLC in detergent can be used for biophysical assays such as thermal stability assay (Figure 3C,D) to investigate the protein interactions with small molecules such as substrates, inhibitors, or lipids59,62. Finally, the SLCs purified using this protocol in biochemical assays can be reconstituted into liposomes or nanodiscs for functional assays68 and used for antibody and nanobody generation and selection69,70. While it remains a challenge to adapt these methods to the throughput necessary for the discovery of new SLC-targeted small molecules 1, promising advances have been made in the field of in vitro high-throughput screening technologies71,72,73,74.
The authors have nothing to disclose.
This work was performed within the RESOLUTE project. RESOLUTE has received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement No 777372. This Joint Undertaking receives support from the European Union's Horizon 2020 research and innovation programme and EFPIA. This article reflects only the authors' views and neither IMI nor the European Union and EFPIA are responsible for any use that may be made of the information contained therein. The pHTBV plasmid was kindly provided by Prof. Frederick Boyce (Harvard).
3C protease | Produced in-house | ||
50 or 100 kDa cut-off centrifugal concentrators | Sartorius | VS0242 | |
5-Cyclohexyl-1-Pentyl-β-D-Maltoside | Anatrace | C325 | CYMAL-5 |
96-well bacmid purification kit | Millipore | LSKP09604 | Montage Plasmid Miniprep |
96-well block (2 mL) | Greiner Bio-One | 780271 | |
Adhesive plastic seals | Qiagen | 19570 | Tape Pads |
Agarose size exclusion chromatography column | Cytiva | 29091596 | Superose 6 Increase 10/300 GL |
Benzonase DNAse | Produced in-house | ||
BisTris | Sigma Aldrich | B9754 | |
Cholesteryl Hemisuccinate Tris salt | Anatrace | CH210 | CHS |
Cobalt metal affinity resin | Takara Bio | 635653 | TALON Metal Affinity Resin |
D(+)-Biotin | Sigma Aldrich | 851209 | |
Dextran-agarose size exclusion chromatography column | Cytiva | 28990944 | Superdex 200 Increase 10/300 GL |
Digitonin | Apollo Scientific | BID3301 | |
Dounce tissue grinder (40 mL) | DWK Life Sciences | 357546 | |
EDTA-free protease inhibitor cocktail | Sigma Aldrich | 4693132001 | cOmplete, EDTA-free Protease Inhibitor Cocktail |
Fetal Bovine Serum | Thermo Fisher | 10500064 | |
Fos-Choline-12 | Anatrace | F308S | FS-12 |
Glycerol | Sigma Aldrich | G5516 | |
Glyco-diosgenin | Anatrace | GDN101 | GDN |
Gravity flow columns | Cole-Parmer | WZ-06479-25 | |
HEK293 medium | Thermo Fisher | 12338018 | FreeStyle 293 medium |
HEPES | Apollo Scientific | BI8181 | |
Hydrophilic, neutral silica UHPLC column | Sepax | 231300-4615 | Unix-C SEC-300 4.6 x 150 |
Imidazole | Sigma Aldrich | 56750 | |
Insect transfection reagent | Sigma Aldrich | 71259 | Reagent |
Lauryl Maltose Neopentyl Glycol | Anatrace | NG310 | LMNG |
Magnesium Chloride Hexahydrate | Sigma Aldrich | M2670 | |
Micro-expression shaker | Glas-Col | 107A DPMINC24CE | |
NaCl | Sigma Aldrich | S9888 | |
n-Decyl-β-D-Maltoside | Anatrace | D322 | DM |
n-Dodecyl-b-D-Maltopyranoside | Anatrace | D310 | DDM |
n-Dodecyl-N,N-Dimethylamine-N-Oxide | Anatrace | D360 | LDAO |
n-Nonyl-β-D-Glucopyranoside | Anatrace | N324S | NG |
n-Octyl-d17-β-D-Glucopyranoside | Anatrace | O311D | OGNG |
Octaethylene Glycol Monododecyl Ether |
Anatrace | O330 | C12E8 |
Octyl Glucose Neopentyl Glycol | Anatrace | NG311 | OGNG |
Phosphate Buffered Saline | Sigma Aldrich | D8537 | DPBS |
Polyoxyethylene(10)dodecyl Ether | Anatrace | AP1210 | C12E10 |
Polyoxyethylene(9)dodecyl Ether | Anatrace | APO129 | C12E9 |
Porous seal for tissue culture plates | VWR | 60941-084 | Rayon Films for Biological Cultures |
Proteinase K | New England Biolabs | P8107S | |
Recombination enzyme mix | Thermo Fisher | 11791020 | Gateway LR Clonase II |
Serum-free insect media | Gibco | 10902088 | Sf-900 II serum-free media |
Sodium Butyrate | Sigma Aldrich | 303410 | |
Sonicator 24-head probe | Sonics | 630-0579 | |
Sonicator power unit | Sonics | VCX 750 | |
Strep-Tactin resin | IBA Life Sciences | 2-5030-025 | Strep-TactinXT 4Flow high- capacity resin |
Sucrose | Sigma Aldrich | S7903 | |
Sucrose Monododecanoate | Anatrace | S350 | DDS |
Suspension-adapted HEK293 cells | Thermo Fisher | A14527 | Expi293F |
Transfection reagent | Sigma Aldrich | 70967 | GeneJuice Transfection Reagent |