Preparation of retinol, RBP, and TTR

rRBP expression, refolding and purification

            All materials were purchased from Thermo Fisher Scientific unless stated otherwise. The human RBP IV gene without the signaling peptide (UniProt KB P02753) was codon-optimized for expression in E. coli using Genscript’s OptimumGene algorithm. The gene was synthesized using the NruI and BamHI restriction sites at the 5’ and 3’ positions, respectively, and ligated into the pTWIN1 plasmid (New England Biolabs). This plasmid allows for generation of a fusion construct containing an intein tag and a high affinity chitin binding domain that is governed under a T7 promoter system. The intein can undergo self-cleavage at room temperature and low pH to release the desired protein at high purity. BL21-DE3 cells (New England Biolabs) were transformed with ligated plasmid and cultured in 500 mL of LB media containing 100 mg/L ampicillin (Sigma-Aldrich).

     Protein expression was induced by addition of IPTG (IBI Scientific) to a final concentration of 750 µM during mid-log-phase growth as determined by OD₆₀₀ (0.6 – 0.8). After 4 hours at 37° C, the bacteria were harvested, pelleted, pooled and re-suspended in lysis buffer (25 mM Tris-HCl, 500 mM NaCl, 1mM EDTA (Sigma-Aldrich), 2 mM PMSF solubilized in isopropanol, pH 9.0). Cells were lysed by sonication for 45 minutes in an ice bath. The lysate was centrifuged at 4000 RPM in a swinging bucket rotor for 30 minutes at 4° C. rRBP was primarily expressed in the inclusion body pellet. The pellet was re-suspended in 25 mL of breaking buffer, prepared by dilution of 8 M guanidine hydrochloride (Sigma-Aldrich) into solubilizing buffer (25 mM Tris, 1 mM EDTA, pH 9.0, supplemented with 10 mM DTT and 2 mM PMSF) to a final guanidine hydrochloride (GdmHCl) concentration of 5 M; the protein solution was stirred vigorously overnight to ensure complete reduction and denaturation of the protein. 

     The next day, excess cell debris was pelleted by centrifugation at 4000 RPM for 45 minutes at 4° C, and the supernatant was collected and diluted 1:4 v/v with cold refolding buffer (25 mM Tris-HCl, 1 mM EDTA, 3 mM cysteine (Sigma-Aldrich), 0.3 mM cystine (Sigma-Aldrich), 0.5 mg fresh synthetic retinol (95%, Sigma-Aldrich) in ethanol (Koptec), pH 9.0) for a final GdmHCl concentration of 1 M. To ensure optimal folding, the retinol was added to the vigorously stirring refolding buffer immediately prior to addition of the reduced and denatured rRBP. Because retinol is highly sensitive to light, the column and solutions containing retinol were covered in aluminum foil. rRBP was allowed to refold while stirring vigorously at 4° C for 5 hours, upon which the solution was immediately applied to a chitin affinity column (New England Biolabs) equilibrated at 4° C with refolding buffer containing GdmHCl at an identical final concentration as the rRBP solution. The column was thoroughly washed with wash buffer (25 mM Tris, 500 mM NaCl, 1 mM EDTA, pH 9.0) to remove GdmHCl, adsorbed cellular protein contaminants and unbound retinol. The column was then equilibrated with elution buffer (25 mM Tris, 500 mM NaCl, 1mM EDTA, pH 6.5) and incubated at room temperature overnight to initiate cleavage of the intein. ROH-rRBP (rRBP bound to retinol) was eluted from the column and the retinol content checked by an absorbance scan from 260 nm – 380 nm. 

Anion exchange (AEX) chromatography

ROH-rRBP eluted from the chitin column was concentrated and buffer exchanged into AEX Buffer A (25 mM Tris, 1 mM EDTA, pH 8.0) in a 10 kDa Amicon-15 MWCO regenerated cellulose membrane filter (MilliporeSigma). The retentate was filtered through a 0.22 µm filter (Millipore) and slowly applied by syringe to a GE HiScreen diethylaminoethyl (DEAE) column pre-equilibrated with AEX Buffer A. The sample was allowed to adsorb for 10 minutes without flow. The column was then re-equilibrated for 10 minutes with AEX Buffer A at a flow rate of 1.0 mL/min. A step salt gradient was applied by mixing AEX Buffer A with high salt AEX Buffer B (25 mM Tris, 1 M NaCl, 1 mM EDTA, pH 8.0) at an 80:20 % v/v ratio. This solution was applied to the DEAE column at a flow rate of 1.0 mL/min to elute the protein. The column was then stripped for 20 minutes at 1.0 mL/min with 100 % AEX Buffer B before subsequent runs. 

Retinol stripping 

ROH-rRBP was chilled to 4° C then stripped of retinol by 1:1 v/v dilution in chilled diethyl ether with gentle rocking overnight at 4° C. The organic phase was carefully removed without disturbing the organic/aqueous interface and replaced with fresh, cold diethyl ether prior to rocking for an additional 3 hours. The aqueous phase, containing apo-rRBP, was then slowly removed with a syringe and placed under nitrogen at room temperature to evaporate trace diethyl ether. Apo-rRBP was concentrated and buffer exchanged to PBS (10 mM phosphate, 150 mM NaCl, pH 7.4) by ultrafiltration at 4° C with a 10 kDa Amicon membrane filter centrifuged at 4000 rpm for 10 minutes. To reduce aggregation during the centrifugation process, the apo-rRBP concentrations were kept near 0.25 mg/mL. The retentate was then filtered through a 0.45 µm filter (MilliporeSigma) to remove any large aggregates and quantified via absorbance at 280 nm on an Eppendorf BioSpectrometer using an extinction coefficient (ε₂₈₀) of 40,400 M^-1 cm^-1. 

SDS-PAGE gel electrophoresis

            Samples were diluted by addition of 2X Novex Tricine SDS Sample Buffer and boiled at 85° C for two minutes. The denatured protein was loaded alongside an EZ-Run Protein Ladder (Fisher BioReagents, Fair Lawn, NJ) on a Novex 10 – 20 % Tricine gradient gel and electrophoresed using Novex Tricine SDS Running Buffer for 90 minutes at 125 V. The gel was then stained with Coomassie Brilliant Blue R-250 and analyzed with a Bio Rad ChemiDoc imaging system.

Preparation of holo-rRBP from apo-rRBP

Holo-rRBP was prepared by dilution of fresh, concentrated retinol (ROH) or retinoic acid (RA) at a 3-fold molar excess in ice cold ethanol into 10 µM apo-rRBP. The concentration of ligand in ethanol was calculated by using an extinction coefficient (ε₃₂₅) of 52,480 M^-1 cm^-1 and (ε₃₅₀) of 44,300 M^-1 cm^-1 for ROH and RA, respectively. The final ethanol concentration was kept < 1 % v/v in the protein sample. The mixture was equilibrated for 24 hours at 4° C to ensure ligand binding. Since serum holo-RBP demonstrates ligand to protein absorbance ratios of ~ 1.0, confirmation of binding was determined by an absorbance scan, with an A_330/280 ratio of 0.95 or higher considered a satisfactory preparation of ROH-rRBP, and an A_338/280 ratio of 0.95 or higher considered a satisfactory preparation of RA-rRBP. 

Western Blot

Apo-rRBP, serum-derived RBP (Athens Research and Technology, Cat # 16-16-180216) and recombinant human albumin (Sigma, Cat # A9731) were diluted into LDS 4X Sample buffer (NuPAGE), boiled and loaded into 4-12% Bis-Tris SDS-PAGE gel (NuPage) at a total protein concentration of 75 ng per well. SDS-PAGE was run in MES Running Buffer (50 mM MES, 50 mM Tris Base, 0.1% w/v SDS, 1 mM EDTA, pH 7.3) at 200 V for 35 minutes. Samples were transferred to 0.45 micron polyvinylidene difluoride membranes at 30 V for one hour before being collected and washed 3X at 5 minutes each in TBST (20 mM Tris, 135 mM NaCl, 0.1 % v/v Tween-20, pH 7.6). Membranes were blocked for 1 hour at room temperature in blocking buffer (5% w/v non-fat dry milk (LabScientific) dissolved in TBST) before incubation at 4 ℃ overnight with either mouse monoclonal anti-human RBP4 (Abnova, 1:500 dilution, Cat # MAB3211) in blocking buffer or blocking buffer alone (for no-primary controls). After overnight staining, all membranes were washed 3X at 5 minutes each in TBST before secondary antibody incubation at room temperature for 1 hour in the dark with goat anti-mouse IgG IRDye 800CW (Li-Cor, 1:5000 dilution, Cat # 926-32210) antibody in blocking buffer. Probed membranes were washed 3X at 5 minutes each in TBST and subsequently imaged using a Li-Cor Odyssey Imager.

Circular dichroism (CD) spectroscopy

Apo-rRBP was dialyzed into fluoride containing buffer (10 mM phosphate, 150 mM NaF, pH 7.4) and filtered through a 0.22 µm filter. The protein (0.1 mg/mL for near UV measurements and 0.05 mg/mL for far UV measurements) was then loaded into a 1.0 mm cuvette and placed in an Aviv model 420 CD spectrometer operating at 25° C. All samples and buffers were scanned three times with a step size of 1 nm and a 5 s integration time, and then averaged to determine a mean CD signal. The mean CD signal from 0.22 µm filtered fluoride buffer was subtracted from the protein samples and the corrected data were converted to molar ellipticity using the exact concentration of apo-rRBP determined by absorbance. Data for near and far-UV were then combined to cover the entire range and the spectra smoothed by a 10 point Savitzky-Golay function.

Guanidine hydrochloride denaturation

Guanidine hydrochloride (GdmHCl) solutions (0 – 6 M) were prepared by volumetric dilution of 8 M GdmHCl stock (Sigma-Aldrich, 8 M, pH 8.5 in 50 mM bicine) into bicine-NaOH Buffer (50 mM bicine (Sigma-Aldrich), 150 mM NaCl, pH 8.5). Apo-rRBP, RA-rRBP and ROH-rRBP, prepared as described above, were diluted into the GdmHCl solutions to a final concentration of 500 nM rRBP. Samples were equilibrated for 7 days at room temperature in the dark. Fluorescence spectra were measured on a PTI QuantaMaster 40 fluorometer with excitation at 295 nm and emission from 315 nm to 500 nm. Spectra for three independent samples at each GdmHCl concentration were collected and averaged, with background subtracted. Unfolding was monitored by the change of intensity at the point of maximal signal difference between the folded and unfolded states (apo-rRBP, 320 nm; RA-rRBP, 360 nm; ROH-rRBP, 460 nm). 

ROH-rRBP equilibrium binding constant

All measurements were collected with a PTI QuantaMaster 40 fluorometer at room temperature. Synthetic retinol was used immediately upon opening. Briefly, ROH was dissolved in ice cold ethanol and vortexed. The stock was diluted 100X into ice cold ethanol, and the concentration measured by absorbance at 325 nm. All samples were prepared in the dark and wrapped in aluminum foil. The concentration of ethanol was 2 % v/v when applied to the rRBP samples. ROH was added to apo-rRBP in PBS (10 mM phosphate, 150 mM NaCl, pH 7.4) to produce a final rRBP concentration of 500 nM and incubated overnight at room temperature. rRBP tryptophan residues were excited at 295 nm and the tryptophan and retinol emission collected from 315 – 500 nm. All data points were corrected by background subtraction of the buffer, as ROH was not fluorescent in buffer under these conditions (data not shown). 

ROH-rRBP-TTR equilibrium binding constant

All measurements were collected with a PTI QuantaMaster 40 fluorometer using polarizers at room temperature. ROH-rRBP solutions were prepared as described above. Recombinant human TTR (rTTR) was produced in PBS (10 mM phosphate, 150 mM NaCl, pH 7.4) as described elsewhere. ROH-rRBP solutions were diluted into the rTTR stocks to a final ROH-rRBP concentration of 1 µM and variable rTTR concentrations between 0 and 4 µM. The fluorescence emission of bound retinol in ROH-rRBP was recorded using excitation at 330 nm and emission at 460 nm. The G-factor value was calculated for an rTTR-free sample of ROH-rRBP at 1 µM, and this value was held constant throughout the experiment [22]. Each sample was scanned 5 times over a 20 second period and the resultant values averaged to generate a single anisotropy value. This process was repeated with three independent samples in order to obtain a mean anisotropy and standard deviation.