Claxton DP and Gouaux E (2018), Expression and purification of a functional het...

XB-ART-55124

PLoS One 2018 Jan 01;137:e0201210. doi: 10.1371/journal.pone.0201210.

Show Gene links Show Anatomy links

Expression and purification of a functional heteromeric GABAA receptor for structural studies.

Claxton DP , Gouaux E .

???displayArticle.abstract???
The GABA-gated chloride channels of the Cys-loop receptor family, known as GABAA receptors, function as the primary gatekeepers of fast inhibitory neurotransmission in the central nervous system. Formed by the pentameric arrangement of five identical or homologous subunits, GABAA receptor subtypes are defined by the subunit composition that shape ion channel properties. An understanding of the structural basis of distinct receptor properties has been hindered by the absence of high resolution structural information for heteromeric assemblies. Robust heterologous expression and purification protocols of high expressing receptor constructs are vital for structural studies. Here, we describe a unique approach to screen for well-behaving and functional GABAA receptor subunit assemblies by using the Xenopus oocyte as an expression host in combination with fluorescence detection size exclusion chromatography (FSEC). To detect receptor expression, GFP fusions were introduced into the α1 subunit isoform. In contrast to expression of α1 alone, co-expression with the β subunit promoted formation of monodisperse assemblies. Mutagenesis experiments suggest that the α and β subunits can tolerate large truncations in the non-conserved M3/M4 cytoplasmic loop without compromising oligomeric assembly or GABA-gated channel activity, although removal of N-linked glycosylation sites is negatively correlated with expression level. Additionally, we report methods to improve GABAA receptor expression in mammalian cell culture that employ recombinant baculovirus transduction. From these methods we have identified a well-behaving minimal functional construct for the α1/β1 GABAA receptor subtype that can be purified in milligram quantities while retaining high affinity agonist binding activity.

???displayArticle.pubmedLink??? 30028870
???displayArticle.pmcLink??? PMC6054424
???displayArticle.link??? PLoS One
???displayArticle.grants??? [+]

Species referenced: Xenopus
Genes referenced: gabarap

???attribute.lit??? ???displayArticles.show???

	Fig 1. Expression of GluClα-EGFP in oocytes.(A) Cartoon design of EGFP fusion to the M3/M4 loop of GluClα. (B) Four days after injecting 50 ng of synthetic mRNA, receptor expression is visualized by EGFP fluorescence at the oocyte surface by confocal microscopy. (C) Detergent (C12M) solubilization of oocytes and FSEC analysis on a Superose6 size exclusion column captures a monodisperse elution profile consistent with a pentameric assembly. Arrows indicate peak elution times for ferritin (440 kDa) and EGFP (27 kDa) standards. Receptor expression levels are sensitive to the total concentration of mRNA injected. FSEC traces were acquired from the same batch of oocytes. Absolute fluorescence intensities are shown.
	Fig 2. Expression of GABAA receptors in oocytes.(A) The α1/β2 receptor demonstrates increased expression levels and homogeneity relative to the α1 subunit alone. The arrow indicates a population reduction of smaller oligomers in the presence of the β2 subunit. (B) Co-injection of γ2S mRNA with the α1 subunit does not improve receptor homogeneity. (C) Injection of all three subunits results in attenuated expression levels relative to α1/β2. All FSEC traces were acquired from the same batch of oocytes. Absolute fluorescence intensities are plotted on the same scale.
	Fig 3. Investigation of α1 and β2 subunit association in oocytes through FRET.Photobleaching of EGFP fused to the α1 subunit results in increased β2-mKalama fluorescence at the oocyte surface relative to a control region, indicating that the subunits are in close enough proximity for subunit fluorescent proteins to undergo FRET.
	Fig 4. Truncation of the M3/M4 loop does not perturb receptor assembly or function in oocytes.(A) Cartoon of subunit construct design for α1 and β2 subunits shows locations for GFPuv fusion and loop truncation. A polypeptide linker (red line) connects GFPuv to the N-terminus of the α1 subunit. (B) Expression of GFPuv-α1/β2 increases receptor homogeneity relative to α1-EGFP/β2. However, an increase in cleaved GFPuv fluorescence signal is observed (arrow). (C) Replacement of the native M3/M4 loop in both subunits with a tri-peptide induces a right shift in the elution profile, consistent with a smaller hydrodynamic radius. Absolute fluorescence intensities are shown on the same scale. The FSEC traces shown in (B) and (C) were obtained from the same batch of oocytes. (D) Two-electrode voltage clamp of GFPuv-α1-LT/β2-LT demonstrating that the receptor retains gating activity and sensitivity to zinc.
	Fig 5. Removal of glycosylation sites on the N-terminus reduces α1/β2 expression in oocytes.(A) Sequence alignment of the N-terminus for GABAA and GluClα subunits identifying residues for deletion (red line) and predicted sites for glycosylation (*). The red bar represents the first predicted α-helix in the extracellular domain. A blue box outlines the predicted signal peptide. GluClαcryst is the sequence used to obtain the crystal structure. (B) Sequence alignment of the C-terminus for GABAA and GluClα subunits identifying residues for deletion (red line). (C) Removal of an 11-residue tail from the α1 subunit in the GFPuv-α1-LT/β2-LT construct (Fig 4) does not change receptor behavior. (D) Deletion of the N-terminus (ΔN) in either α1 or β2 subunit reduces receptor expression in the α1-EGFP/β2 construct. (E) Site directed mutagenesis of predicted glycosylation sites reduces α1-EGFP/β2 receptor expression. Absolute fluorescence intensities are shown on the same scale. FSEC traces shown in (D) and (E) were obtained from the same batch of oocytes.
	Fig 6. Mutation of other predicted glycosylation sites in the β2 subunit alters receptor expression and assembly in oocytes.(A) Site directed mutation of consensus glycosylation sites in the β2 subunit alters receptor expression and monodispersity, which is in contrast to mutation of a second predicted glycosylation site in the α1subunit (B). Absolute fluorescence intensities are shown on the same scale. FSEC traces were obtained from the same batch of oocytes.
	Fig 7. The α1-EGFP/β2 receptor cannot be solubilized by detergent from Sf9 cells.(A) Diagram of insect cell expression vectors harboring one (pFastBac1) or both (pFastBac Dual) full-length GABAA subunits. EGFP has been placed in the M3/M4 loop of the α1 subunit. (B) EGFP fluorescence is observed 48 hours post infection for pFastBac1 or pFastBac Dual constructs, suggesting receptor targeting to the cell membrane. (C) FSEC analysis at the indicated time points to measure expression suggests that most of the receptor breaks down upon whole cell solubilization with C12M detergent. (D) FSEC profiles do not improve using the bicistronic pFastBac Dual vector, which ensures each cell possesses a copy of both subunits. Absolute fluorescence intensities for (C) and (D) are shown on the same scale.
	Fig 8. Identification of α1-EGFP/β2 receptor expression parameters in mammalian cells.(A) Diagram of pEG BacMam vector used for generating recombinant baculovirus to infect mammalian cells. (B) Addition of 10 mM sodium butyrate to the virus-transduced culture increases receptor yield but induces aggregation. Absolute fluorescence intensities are shown. (C) Infecting cells with a relative MOI ≥ 1 (β:α) increases receptor monodispersity. Area-normalized traces are shown. Cells were grown at 37°C for FSEC traces in (B) and (C). (D) Dropping the temperature of infected cells increases receptor yield and reduces aggregation observed at higher temperatures. Absolute fluorescence intensities are shown in (D).
	Fig 9. Optimized mammalian expression parameters apply to other receptor GABAA receptor subtypes.(A) A boost in ρ1-EGFP expression levels is observed at 30°C with addition of 10 mM sodium butyrate. (B) Similar to oocytes, the β subunit is required to form a homogeneous receptor in mammalian cells. In addition, the α1-EGFP/β1-LT receptor demonstrates higher expression levels than the α1-EGFP/β2-LT receptor. Expression for each experiment was performed at 30°C. For comparison, the fluorescence scale in (B) is approximately 2-fold greater than in (A). (C) Time course of α1-EGFP/β1-LT expression at 30°C (relative MOI = 0.6, α1:β1) followed by EGFP-FSEC suggests that the receptor demonstrates peak expression levels 60–72 hours post infection. (D) The α1-LT/β1-LT receptor displays GABA-gated current in oocytes by TEVC measurement.
	Fig 10. Purification and characterization of the α1/β1 receptor from mammalian cells.(A) Preparative size exclusion chromatography of IMAC-purified material isolates homogeneous receptor as suggested by FSEC analysis of four representative elution fractions collected across the SEC peak (B). (C) SDS-PAGE analysis of SEC fractions across the elution peak shown in (A) indicates diffuse subunit bands, consistent with glycosylation. (D) Purified receptor shows specific binding in C12M or L-MNG detergent. Ligand binding experiments by SPA reveal high affinity binding sites for 3H-muscimol (KD = 24 nM, Bmax = 1690 CPM or 0.76 pmol ligand, E) or GABA (Ki = 98 nM, F) as measured by a direct binding isotherm or competition assay, respectively.

References [+] :

Allen, De novo mutations in epileptic encephalopathies. 2013, Pubmed