Comparative study of commercially available and homemade anti-VAMP7 antibodies using CRISPR/Cas9-depleted HeLa cells and VAMP7 knockout mice

Cell culture

HeLa and Cos-7 cells (ATCC CCL-2 and CRL-1651, respectively) were maintained at 37°C and 5% CO₂ in a humidified incubator, and grown in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% fetal calf serum (FCS), 100 units/ml penicillin, and 100 μg/ml streptomycin (Gibco, Thermo Fisher Scientific). Cells were regularly split using Trypsin-EDTA to maintain exponential growth. Transfection of cells was performed using Lipofectamine 2000 according to the manufacturer's instructions. All culture media reagents were from Thermo Fisher Scientific.

Generation of VAMP7-depleted cells

Invalidation of VAMP7 was achieved using CRISPR RNA-CAS9 guide constructs based on a previously published protocol²¹. Briefly, sequences for sgRNA were preferentially chosen within the first exon region of VAMP7 genomic gene, with the help of the “RGEN Cas designer”²² web-based tool (http://rgenome.net/cas-designer/). To limit off-targets, an oligo sequence with ≤2 putative mismatches throughout the whole genome or an ‘out of frame’ score <66 were excluded. The sgRNA target sequences used are: 5’-caccgAACAGCAAAAAGAATCGCCA-3’ (forward) and 5’-aaacTGGCGATTCTTTTTGCTGTTc-3’ (reverse). Oligonucleotides (10 mM) were heated at 95°C for 5 min and annealed by ramping down the temperature from 95°C to 25°C at 5°C min^-1. Annealed primers were ligated into pSpCas9(BB)-2A-Puro (PX459) vector (Addgene) using the BbsI sites. After validation by sequencing, the targeting constructs were transfected into HeLa cells following a previously described protocol. An empty pSpCas9(BB)-2A-Puro (PX459) vector was used to generate “control” cells. At 24 h post-transfection, cells were diluted 1/10 and transfected ones were selected by 1 µg/ml puromycin addition for 72 h. The selected populations were then seeded into a 96-well plate at 1 cell per well. Clones derived from single cells were amplified and screened for deficiency by immunoblotting.

Plasmids

The human and rat GFP-VAMP7 constructs are the same as those that have been described previously²³. Plasmid containing mouse VAMP7 cDNA was a kind gift from Maurizio D’Esposito (IGCB, Naples, Italy). Mouse VAMP7 was amplified by PCR and cloned into pEGFP-C3 (Clontech) using HindIII / BamHI restriction sites.

Mouse colony husbandry and care

The wild-type (WT) and VAMP7 knockout (KO) littermate male cohort was established at the Mouse Clinical Institute animal facility as previously described¹⁰. They had a mixed 129/Sv-C57BL/6 genetic background. They were weaned at 4 weeks and housed two to six per M.I.C.E. cage by sex and litter regardless of the genotype under standard conditions and maintained in a room with controlled temperature (21−22°C) under a 12 h light/dark cycle (light on at 7:00 A.M.), with food (standard chow diet, Safe D04) and water available ad libitum. All experiments were performed in accordance with the European Communities Council Directive regarding the care and use of animals for experimental procedures (2010/63/UE) and were approved by the local ethical committee (CEEA40-Comité d’Ethique Buffon). Mice were euthanized by cervical dislocation. All efforts were made to ameliorate the suffering of animals and to reduce their number per experiment (1 animal cortex per condition for immunoprecipitation experiment).

Cortex isolation

The cortex from WT and VAMP7 KO 8 weeks-old C57bl/6 mice were isolated according to a previously published protocol¹⁰. Cortex were dissociated by pipette trituration in 500 µl TSE (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA) supplemented with 1% Triton X-100 and cOmplete protease inhibitor tablets (Roche Applied Science) and volume was adjusted in order to get ~2 mg/ml protein final concentration (considering 10 mg of tissue is equivalent to ~1 mg final protein). Lysis was performed by incubation under agitation at 4°C for 30min. After clarification by 16.000 x g centrifugation for 30 min, protein concentration of the supernatant was estimated using Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories) and immunoprecipitation was carried out (see below).

Affinity purification

The TG50 (for “Thierry Galli #50”) serum raised against VAMP7 was generated by Covalab (Villeurbanne, France; animal house registration number C21 464 04 EA) using immunization of New Zealand white rabbit with GST-VAMP7 (1-188 aa, as previously described for TG11²⁴ and TG18²⁵) and then purified by affinity chromatography. Briefly, serum was clarified and de-lipidated by high-speed centrifugation (70,000 rpm) and applied on a 6xHis-VAMP7 (coiled-coil 1-188 aa) covalently cross-linked HITrap-NHS column (GE Healthcare) overnight at 4°C using a peristaltic pump (0.3 ml/min). We washed the column with filtered and degassed PBS (20 ml, 1 ml/min). Specific antibodies were eluted using 200 mM Glycine/HCl pH 2.2, and collected in tubes containing neutralizing buffer (TBS 10X, 1X final). For each fraction, protein concentration was quantified by optical density (280 nm) measurement.

Immunoblotting

References for all tested antibodies and reagents used for immunoblotting are listed in Table 1 and Table 2, respectively. Cells were washed in cold phosphate buffered saline (PBS) and lysed 20 min in TSE (50mM Tris-HCl, pH 8.0, 150mM NaCl, 1 mM EDTA) supplemented with 1% Triton X-100 and cOmplete protease inhibitor tablets (Roche Applied Science). Lysates were clarified by centrifugation 30 min at 16.000 x g, and protein concentration was estimated using Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories). Following heat denaturation at 95°C during 5 min, proteins were separated by 15% SDS-PAGE and transferred to a 0.45-µm nitrocellulose filter (Amersham, GE Healthcare) at 40 mA overnight. Membrane was blocked with 5% low-fat milk in TBS buffer for 20 min at room temperature and probed with indicated primary antibodies diluted in 5% skimmed milk in TBS-T, overnight at 4°C. Following primary antibody incubation, the filter was washed 3× 5 min in TBS-T at room temperature, and probed with HRP- or fluorescently-labeled secondary antibodies diluted in TBS-T for 1 h at room temperature. The filter was then washed 3× 5 min in TBS-T at room temperature. All incubation and washes were performed with gentle rocking. Proteins were detected by enhanced chemiluminescence (ThermoFisher Scientific) and imaged using ImageQuant LAS-4000 (Fujitsu Life Sciences), or scanned in an Odyssey infrared imaging system (Li-Cor).

Immunofluorescence staining

The following immunofluorescence staining protocol was performed, with all steps carried out at room temperature unless stated otherwise (see Table 3 for reagent details). HeLa cells were washed once in PBS, fixed with 4% paraformaldehyde in PBS for 20 min, quenched for 20 min with 50 mM NH₄Cl in PBS and permeabilized by treatment with 0.3% Triton X-100 in PBS for 4 min. After blocking with 10% FCS + 0.3% Triton in PBS for 30 min, cells were incubated overnight with the primary antibodies diluted in 3% FCS + 0.3% Triton in PBS at 4°C. After several washes with 3% FCS, 0.3% Triton in PBS, cells were then incubated with the secondary antibodies in 3% FCS, 0.3% Triton in PBS for 1 h, then washed several times in 0.3% Triton in PBS. Coverslips were partially dried and mounted in Prolong medium (Invitrogen), and then left to set overnight. Fluorescence microscopy and imaging were performed using an upright epifluorescence microscope (DMRA2, Leica Microsystem) equipped with a CMOS camera (Orca Flash 4.0 LT, Hamamatsu) and a HCX PL APO 100x/1.40-0.70 oil CS oil-immersion Leica objective. Dilution of primary antibodies (Table 3) was formerly optimized to get relatively equivalent signal intensity in the WT cells for the same microscope settings (time of exposure, binning, objective, etc.), allowing direct comparison of signal between antibodies.

Immunoprecipitation

Transfected Cos7 cells were washed once in PBS 1X then lysed as described in the “immunoblotting” section. Immunoprecipitation experiments were carried out as followed (see Table 4 for reagent details). Briefly, 1 mg of protein extract was submitted to immunoprecipitation by incubation overnight at 4°C with 2.5 µg of antibodies that were pre-coupled with 25 μl magnetic beads (Dynabeads M-280, Invitrogen). Beads were then extensively washed with TSE-1% Triton and beads resuspended in 2X Laemmli buffer. Samples were loaded on 4–12% Bis-Tris NuPAGE (ThermoFisher Scientific) or RunBlue SDS (Expedeon) gels with manufacturer-recommended electrophoresis buffer, processed for western blotting using HRP-coupled secondary antibodies and enhanced chemiluminescence (ThermoFisher Scientific).

For immunoprecipitation of endogenous VAMP7, 1 mg of mouse cortex extracts (see “Cortex isolation” section) were submitted to immunoprecipitation as for cell extracts, excepted that 5 µg of antibodies, 40 μl magnetic beads (Dynabeads M-280, Invitrogen), fluorescent secondary antibodies and an Odyssey infrared imaging system (LI-COR, Lincoln, Nebraska, USA) were used.

Quantitative analysis

All quantification analyses was performed using ImageJ (1.49n) and data were computed in Microsoft Excel.

For western blotting signal quantification, a 20-pixel-wide straight line vertically crossing each lane was drawn to generate an intensity profile (see Figure 1A and B, left panel). Local background correction was performed by manually drawing a segmented line under the peaks representing the bands detected by western blotting (Figure B, right panel). The VAMP7 band was defined as the ~25 kDa band that would disappear or diminish in intensity in the KO extract compared with the WT extract. Areas under all peaks and the VAMP7 one, shown in grey and blue, respectively (Figure 1B, right panel), were measured. In order to estimate the signal-to-noise ratio of each antibody, taking into account the intensity of the band of interest over the intra-lane non-specific ones and the specificity of the signal in the control condition versus the KO ones, a so-called “WB specificity index” was calculated for each lane of the western blot, using the following equation:

Figure 1. Comparative immunoblot analysis of endogenous VAMP7 level of expression in control or CRISPR/Cas9-modified VAMP7-depleted HeLa cells.

(A) Immunoblot performed on lysates from control (Ctrl) and VAMP7 knockout (KO) HeLa cells. An equal amount of total protein extracts from each condition was run in replicates. Following transfer on nitrocellulose and blocking, membrane was sliced and each piece was probed with indicated anti-VAMP7 antibodies (expected size: ~25 kDa). Time of exposure for each condition is provided. For loading control, membrane was washed and incubated with anti-α-tubulin antibody (expected size: ~50 kDa). (B) Example of quantification of western blotting signal from (A) (see dotted red line). Intensity profile (right panel) was generated from a 20-pixel-wide straight line (yellow, left panel) across each lane. On the intensity profile (right panel), areas corresponding to the VAMP7 signal and the non-specific bands are shown in blue and gray, respectively (see Methods section for details). (C) Quantification of each antibody tested by western blotting shown in (A). The “specificity index” represents the signal-to-noise ratio of the antibodies, reflecting the intensity of the VAMP7 band amongst the overall signal per lane (including non-specific bands) and its specificity in the control condition compared to KO (see Methods section for details). AP, affinity purified.

For immunocytochemistry, rather than a raw intensity comparison between control and KO conditions, we chose to rely on the subcellular distribution of VAMP7 to reflect the correct staining pattern of the tested antibodies better. This was estimated using a previously described method with slight modifications²⁶, for each antibody, in control and VAMP7 KO cells. A 20-pixel-wide band (Figure 2B) was drawn from the nucleus (defined from the edge of DAPI staining) to the cell leading edge and going through the perinuclear region pointing to the most elongated part of the cell. Length of this band was normalized by binning in order to obtain a 10-pixel-long line from which a fluorescence intensity profile was generated. Profile from control and KO cells were plotted in function of the distance from nucleus to periphery (i.e. 0–100% on the x axis, Figure 2B, lower panel). A so-called “IF specificity index” was calculated by measuring the area between the control and the KO intensity profiles, which can be computed by the following equation:

Figure 2. Comparative immunostaining of VAMP7 endogenous level of expression in control or depleted HeLa cells.

(A) Control and VAMP7 knockout (KO) HeLa cells were grown on glass coverslips, fixed in paraformaldehyde, blocked and immunostained with the indicated mouse or rabbit anti-VAMP7 antibodies (green) and DAPI (blue). KO condition allows the estimation of background signal. Samples were imaged with an epifluorescence microscope using a 100X objective. Bars, 25 µm. (B) Example of quantification of immunofluorescence signal from (A). Distribution of VAMP7 fluorescence was measured across a 20-pixel-wide straight line (yellow) drawn from the border of the nucleus towards the plasma membrane in control and VAMP7 KO cells (left panel). Intensity profiles were plotted on the same graph (right panel) and area between control and VAMP7 KO curve (blue) was measured in order to generate a “specificity index” (see Methods section). (C) Quantification of each antibody tested by IF shown in (A). The “specificity index” reflects both pattern of distribution of VAMP7 across the cell and signal-to-noise ratio given by the tested antibodies (see Methods section for details). AP, affinity purified.

Antibodies with a high score exhibit both a more intense signal in the perinuclear region than the peripheral one in the control and a low signal distribution for VAMP7 KO condition compared to control. Poor score corresponds to either a correct distribution of VAMP7 signal in the control but a very high background in the KO condition, or an unspecific/random distribution of VAMP7 signal in the control, or a combination of both cases.

Tools development and description of antibodies tested in this study

In order to characterize and compare a set of commercially available and homemade (i.e from “Thierry Galli’s lab”, hereafter referred as “TG lab”) anti-VAMP7 antibodies (Table 1), we first generated VAMP7 knockout cells, using CRISPR/Cas9 engineering²¹, as described in the methods section. HeLa cells were chosen as they express VAMP7 endogenously in a detectable amount by western blotting and immunocytochemistry¹⁵.

Several of the tested antibodies were described in the provider’s datasheet to only work for immunofluorescence detection (e.g. Cell Signaling, catalogue number D81Y1R) or western blotting (e.g. Cell Signaling, catalogue number D4D5J) and were used accordingly. Our lab generated two antibodies, the mouse monoclonal “158.2”²⁷ and the rabbit polyclonal “TG50”, which are commercially available from Synaptic Systems and Covalab (TG50 as protein A purified serum), respectively. Only the in-house affinity-purified (AP) version of the TG50 antibody was included in this study because we wanted an affinity-purified serum as best possible positive control.

Characterization of antibodies by western blotting

We compared four monoclonal and four polyclonal rabbit antibodies by western blotting using control or VAMP7-KO HeLa cell extracts (Figure 1). We used an anti-tubulin antibody as loading control. All antibodies tested in the described conditions (Table 2) were sensitive enough to detect a prominent band at the expected molecular weight (~25 kDa) in the control condition. This band was absent in the VAMP7-KO cell lines in all cases. However, some non-specific bands were visible for all the tested antibodies in both control and VAMP7-KO cell lysates, particularly with polyclonal Synaptic Systems (catalogue number 232 003) and Sigma-Aldrich (catalogue number T6074) antibodies. Therefore, all the tested antibodies showed a signal that was specific for VAMP7, but they also showed variable background bands. As assessed by intensity profile analysis (Figure 1B) and our western blotting specificity index (Figure 1C), the TG lab (TG50) antibody showed the best signal-to-noise ratio using this Western blotting conditions, which may not be surprising, because it had been affinity-purified.

Characterization of antibodies by immunofluorescence

In order to better characterize these antibodies (Table 1), we performed immunostaining in control and VAMP7 KO HeLa cells. For this assay, the rabbit antibody from Cell Signalling Technology clone D8Y1R was used instead of the D4D5J clone, according to the manufacturer’s recommendations. To compare the specificity of the antibodies, we adjusted their dilution (Table 3) to get relatively equivalent signal intensity in the WT cells with the same acquisition time on the microscope. According to this assay, the mouse antibodies from Creative Diagnostics (CABT-37960MH), Synaptic Systems (158.2 –232 011), TG lab (158.2) and the rabbit antibodies Synaptic Systems (232 003) and TG lab (TG50) stained perinuclear membrane structures and vesicles dispersed in the cytoplasm, a typical and already described localization pattern for VAMP7 in HeLa cells^10,15,28. However, in the WT HeLa cells, the R&D Systems (MAB6117) antibody gave a homogenous signal, which spread into the nucleus, the Cell Signaling (14811) antibody, seemed to also stain perinuclear ER-like structures and the Sigma-Aldrich (T6074) antibody exhibited a diffuse cytoplasmic pattern with an absence of vesicular staining. All the tested antibodies seemed to show an overall lower-intensity signal in the VAMP7 KO cells compared to control. According to the “secondary-only” condition that reveals the internal background signal of the experiment (Figure 2A, right panel) and the distribution analysis we conducted (Figure 2B, see Methods for details) to compute the IF specificity index (Figure 2C), the most convincing signal-to-noise ratio was observed with the Creative Diagnostics (CABT-37960MH), Synaptic Systems (232 003) and TG lab (158.2 and TG50) antibodies.

Characterization of antibodies by immunoprecipitation

Taken together, immunoblot and immunofluorescence assays suggest that the homemade anti-VAMP7 antibodies (158.2 and TG50) showed the best endogenous signal-to-noise ratio in HeLa cells. To check the inter-species specificity of these antibodies, we carried out immunoprecipitation assays in Cos cells overexpressing mouse, rat or human GFP-tagged VAMP7 constructs (Figure 3A). We used 158.2 and TG50 for immunoprecipitation and immunoblot and species-specific IgG and anti-GFP antibodies as negative and positive controls for immunoprecipitation, respectively (Table 4). Both 158.2 and TG50 antibodies exhibited a sharp ~50 kDa band in all immunoprecipitation lanes, demonstrating a relatively equivalent ability to precipitate either mouse, rat or human VAMP7, while a very faint signal was observed in the negative control IgG IP lanes.

Figure 3. Comparison of available anti-VAMP7 antibodies for exogenous and endogenous immunoprecipitation.

(A) Cos cells overexpressing GFP-tagged mouse, rat or human VAMP7 constructs were lysed and VAMP7 was immunoprecipitated using indicated specific antibodies (“IP Antibody”). Normal isotype IgG (“IgGM” or “IgGR”) and GFP antibody were used as negative and positive control, respectively. Supernatants after immunoprecipitation (SN) and immunoprecipitates (IP) were probed with indicated VAMP7 antibodies (WB primary Ab). *VAMP7 band of interest (expected size of GFP-VAMP7 constructs: ~50 kDa). °Absence of band at VAMP7 corresponding size. (B) VAMP7 was immunoprecipitated from wild-type (“WT”) and VAMP7 knock-out (KO) mouse cortex lysates using rabbit anti-VAMP7 TG50 antibody and detected with mouse anti-VAMP7 158.2. Normal isotype IgG (IgM or IgGR) were used as negative control. Inputs (IN), supernatants after immunoprecipitation (SN) and immunoprecipitates (IP) were probed with mouse anti-VAMP7 158.2 antibody. *VAMP7 band of interest (expected size of GFP-VAMP7 constructs: ~50 kDa). °Absence of band at expected size. ~heavy and light chains of the antibody used for immunoprecipitation. AP = affinity purified.

We next wondered whether or not these antibodies could immunoprecipitate endogenous VAMP7 from tissue and test background signal in KO tissue. To this aim, we chose to perform immunoprecipitation on lysates from WT or VAMP7 KO mouse cortex using the rabbit TG50 antibody because it performed the best in previous assays and to observe immunoprecipitated VAMP7 with the mouse 158.2 antibody (Figure 3B). Although 158.2 antibody showed stronger background and multiple bands compared to the pattern seen in HeLa cells (Figure 1), a clear band at the expected molecular weight (~25 kDa) was present for the WT immunoprecipitation condition but not VAMP7 KO, demonstrating the specificity and low background noise of the TG50 antibody for immunoprecipitation of endogenous VAMP7 in mouse tissue extracts. Altogether, we showed here that both 158.2 and TG50 antibodies were able to immunoprecipitate VAMP7 from different species and that immunoprecipitation of endogenous VAMP7 could be performed in mouse tissue extracts using TG50 for IP and 158.2 for subsequent western blotting.