Fluorescence ImmunoPrecipitation (FLIP): a Novel Assay for High-Throughput IP
© The Author(s). 2016
Received: 21 May 2016
Accepted: 13 July 2016
Published: 15 August 2016
The immunoprecipitation (IP) assay is a valuable molecular biology tool applied across a breadth of fields. The standard assay couples IP to immunoblotting (IP/IB), a procedure severely limited as it is not easily scaled for high-throughput analysis.
Here we describe and characterize a new methodology for fast and reliable evaluation of an immunoprecipitation reaction. FLIP (FLuorescence IP) relies on the expression of the target protein as a chromophore-tagged protein and couples IP with the measurement of fluorescent signal coating agarose beads. We show here that FLIP displays similar sensitivity to the standard IP/IB procedure but is amenable to high-throughput analysis. We applied FLIP to the screening of mouse monoclonal antibodies of unknown behavior in IP procedures. The parallel analysis of the considered antibodies using FLIP and IP/western shows good correlation between the two procedures. We also show application of FLIP using unpurified antibodies (hybridoma supernatant) and we developed a publicly available tool for the easy analysis and quantification of FLIP signals.
Altogether, our characterizations of this new methodology show that FLIP is an appealing and reliable tool for any application of high-throughput IP.
KeywordsImmunoprecipitation Immunoblotting High-throughput HuEV-A vector Antibodies Screening
In the post-Human Genome Project era, increasing and growing interest is devoted towards proteomic approaches aimed at gaining new insight into the complex ways cells interpret and regulate genomic information. Consortia such as the Encyclopedia of DNA Elements consortium and the Human Proteome Organization (HPP) aim to fill the gap between the genomic and the proteomic world [1, 2]. Mass spectrometry and antibody capture analyses are the basic experimental strategies used towards this goal. An essential tool for the study of the function and interaction of proteins is the immunopurification of the protein of interest with or without its interacting partners and molecules [3, 4].
The immunoprecipitation assay (IP) is an important tool used in academic and industrial research labs for applications such as targeted protein purification, protein concentration, analysis of protein-protein interactions, identification/analysis of protein complexes and analysis of protein/DNA interactions using chromatin IP (ChIP). IP and ChIP are both inexpensive and highly informative, relying on a specific antibody to selectively bind to the target epitope with high affinity . By combining this binding reaction with a high molecular weight entity such as a bead, a bacterial cell or a meshwork of secondary antibodies, it is possible to isolate the protein of interest and its binding partner(s) from all the other cellular components in a microfuge tube . However, not all antibodies are well-suited for IP. For instance, for some IP applications the antibody should recognize the target protein in its three-dimensional folded state, which requires that the epitope be exposed on an accessible surface of the target protein . Therefore, there is a demand for high-throughput assays to screen antibodies capable of IP of target proteins [8, 9]. Moreover, because of the wide range of applications of the IP assay, quick and high-throughput ways to determine the success of an IP are necessary. Until now the standard procedure to validate antibodies for IP couples the IP assay to immunoblotting (IP/IB), a procedure that is challenging to scale to high-throughput analysis . Novel approaches for antibody validation combine immunoprecipitation with mass spectrometry analysis (IP/MS) but, despite the quantitative, highly sensitive and informative nature of this approach, IP/MS is still a more expensive and labor- and instrument-intensive procedure compared to the more accessible IP/IB [6, 11]. Moreover, success of IP/MS is highly dependent on both the cell type and the abundance of the protein of interest.
To our knowledge the only high-throughput assay developed for IP screening is LUMIER (luminescence-based mammalian interactome mapping) system [12, 13]. LUMIER relies on the quantification of the luminescence signal produced by the immunoprecipitation or co-immunoprecipitation of an over-expressed target protein tagged with Renilla luciferase. LUMIER was successfully applied to a range of applications but, to our knowledge, is still the only high-throughput option to classical IP/WB.
We have developed a novel assay for immunoprecipitation, termed FLuorescence IP (FLIP) that is quick, reliable, easy to perform and scalable because it does not rely on immunoblotting analysis. Additionally, due to the small amount of material required, the FLIP assay can be included in a standard IP/IB procedure without affecting the output of the analysis. Because of these features, the FLIP assay is ideal for high-throughput screening to identify IP-grade antibodies. Here we describe the FLIP concept and apply it to the high-throughput identification of IP-proficient mouse monoclonal antibodies.
HuEV-A Cloning and Cre Recombination Reaction
The HuEV-A vector was constructed from a one-step ligation reaction with two inserts and a vector fragment. The backbone for the HuEV-A vector was the previously described pCEP4-Puro plasmid . The first insert was a DNA fragment containing the chloramphenicol resistance gene, ccdB gene and att-R2 site, and was constructed by joining a NotI/KpnI segment of the Gateway Reading frame A cloning vector (Life Technologies, Cat #11828-029) containing the chloramphenicol resistance gene and the att-R2 site to a second synthetic DNA fragment (UEV) containing the complementary att-R1 site along with a series of epitope tags, recombination sites and the Venus YFP sequence with BsiWI (compatible with BsrGI) and EagI (compatible with NotI) termini. The pCEP Puro backbone vector was digested with KpnI and BsrGI and the inserts were ligated into the backbone using T4 DNA ligase in ligase buffer (Enzymatics, Beverly, MA; Prod. No L6030-HC). Vectors were transformed into E. coli and selected on ampicillin/chloramphenicol plates. The HuEV-A vector has been deposited at Addgene (plasmid number 68342).
Cre recombination reactions were performed to generate HuEV-A expression plasmids lacking YFP in the tag. One unit of Cre recombinase (New England Biolabs, M0298), 0.25 μg of HuEV-A plasmid with intact tag, was incubated in 50 μl of 1X Cre buffer (New England Biolabs, Beverley, MA) at 37 °C for 30 min. Cre recombinase was then inactivated for 10 min at 70 °C and the DNA purified on column (Qiagen, 28106). The eluted DNA was then cut with PmlI enzyme for 1 h at 37 °C in CutSmart buffer (New England Biolabs) to cut the not recombined plasmids. PmlI has a unique site in the sequence between the LoxP sites. DNA was again column purified, transformed in ccdB resistant competent cells (Invitrogen, Carlsbad, CA; Prod. No A10460) and plated on ampicillin selection plates. Colonies were screened by colony PCR and the chosen clones verified by sequencing.
Cell Culture and Transient Transfection
HeLa-M2 cells were cultured in DMEM supplemented with 10 % FBS (Gemini, Prod. No 100–106) and 1 mM L-glutamine (ThermoFisher/Life Technologies, Prod. No 25030–081) . Cells were routinely split in fresh medium and new plates upon reaching 80–90 % confluence. During routine culture of the cells the medium was changed every 2–3 days.
The day before transfection 0.3 × 106 cells were plated in each well of a 6-well plate (Falcon). The day after plating, transfection was performed using Fugene-HD reagent according to the manufacturer’s instructions (Promega, Madison, WI; Prod. No E2311). For each well, 0.75 μg of DNA, 3 μl Fugene-HD and 50 μl of Opti-MEM (Life Technologies, Prod. No 31985–070) were incubated 15 min at room temperature. The transfection mixture was then dripped into each well containing 2 ml of complete media and the HeLa M2 cells plated the previous day. Transfection conditions (number of cells plated, DNA amount and Fugene-HD amount) used in Fig. 4 are reported in Additional file 1: Figure S4. We found that high transfection efficiency (>80 %) depended on (i) confluency of cells prior to plating (cell density >80 % confluence was detrimental); (ii) duration of incubation of plated cells prior to introduction of foreign DNA/Fugene (incubation after plating of >18 h was required); (iii) time after transfection before doxycycline treatments (doxycycline treatments immediately after transfection induced much lower expression compared to doxycycline treatments started 18–24 h after transfection); and purity/quality of DNA (all plasmids were prepared by Maxi or Midipreps [Invitrogen PureLink, plasmid filter kits, product number K210015]).
For the experiments presented here, doxycycline was added to the medium immediately after transfection at a final concentration of 1 μg/ml, which limited protein expression from the HuEV-A vector but shortened the timeline to cell harvest by a day. The expression of the desired protein under the indicated conditions was sufficient to guarantee a reliable FLIP and/or an IP/IB analysis in 96 % of the cases. Sporadic low expression was attributed to inaccurate cloning of the target protein into the expression vector.
Cell Lysis and Conventional Immunoprecipitation
For the parallel analysis of mouse antibodies (CDI Laboratories) using FLIP and IP/IB assays in a 96 well IP format (Fig. 5), an entire 6-well plate was used for each target protein. Cells were treated with 0 (one well), 50 ng/ml (2 wells) or 1 μg/ml (3 wells) doxycycline and lysates obtained from one well were used for one IP. After 24 h of doxycycline treatment, cells were lysed in Triton lysis buffer (20 mM HEPES, 1 % Triton X-100, 500 mM NaCl) supplemented at the time of use with protease inhibitor tablets (Roche, Prod. No 11697498001). 150 μl of lysis buffer were used to lyse the cells cultured in one well, and 120 μl of clarified cell lysate were used for each IP/ FLIP. IPs were performed with a normal mouse IgG Abs (Sigma, Prod. No sc-2025) as a negative control (using lysates from cells treated with 1 μg/ml doxycycline); FLAG-M2 antibody (Sigma, Prod. No F1804) as positive control (using lysates from cells treated with 1 μg/ml doxycycline); or the test antibody (using lysates from cells treated with 50 ng/ml and 1 μg/ml doxycycline). After SDS-PAGE and transfer of the proteins onto a PVDF-F membrane, proteins were probed with a rabbit polyclonal antibody against FLAG tag (Cell Signaling #2368) mixed with the test mouse antibody. Secondary goat antibodies conjugated to IRDye680 (anti-rabbit) or IRDye800 (anti-mouse) dyes (LiCOR), were used for the detection of the recognized bands on an Odyssey CLx scanner (LiCOR). The efficiency with which the tested mouse Ab was able to immunoprecipitate the over-expressed target protein was calculated by quantifying the intensity of the band of the protein immunoprecipitated with the tested mAb over the intensity of the band of the protein from the total lysate lane (IP INPUT) (Additional file 1: Figure S5). Correcting for the different volumes/amount loaded on the gel, the % of target protein immunoprecipitated (% tot. lysate) was calculated. FLIP was performed on the beads from FLAG IP, IgG IP and tested mAb IP and using only cell lysates from cells treated with 1 μg/ml doxycycline.
For the experiments described in Fig. 3, HeLa cells were plated and transfected in multiple 3-cm wells as described above, and all wells were treated with 1 μg/ml doxycycline for 24 h starting from the day after transfection. Cells were lysed in Triton buffer supplemented with protease inhibitors, and FLIP was performed as described below in 1.5 ml tubes (low throughput FLIP).
For the IP performed with hybridoma supernatant (Fig. 7), lysates from cells expressing UPF1 protein from the HuEV-A vector were used. Lysates from about two 3-cm wells were used to IP the tagged UPF1 protein using 200 μl (≈1:3 dilution) of hybridoma supernatant, (of unknown antibody concentration), or control mouse IgG antibodies (5 μg). The IP was conducted in 0.7 ml of total volume in 1.5 ml tubes using protein A/G beads [16, 17].
A Mini-FLIP assay (described below) used to evaluate the same hybridoma supernatants was performed using 40 μl (1:3 dilution) of hybridoma supernatant in a total volume of 120 μl.
Before performing FLIP assays, the YFP content of lysates from cells expressing the YFP-tagged protein of interest was measured to determine the amount of YFP present in each μl of lysate. The fluorescence of 10 μl of lysate was measured in clear-bottom black 384-well plates and compared to a standard curve obtained measuring the fluorescence of known amounts of recombinant YFP (MBL International Corporation; Prod. No JM-4998-100) (a typical standard curve ranged from 0 to 500 ng of recombinant YFP). If an autogain was set for the fluorescence reading, the samples and the standard curve dilutions were always read on the same plate. In this study, an excitation wavelength of 475 nm, and emission wavelength of 527 nm, were used for YFP detection using a Synergy H1 reader (BioTek). After quantifying the amount of YFP, an estimate of FLIP efficiency was determined from the curve in Fig. 3a. Considering that the curve in Fig. 3a was obtained using a well-behaved FLAG antibody, 700 μl of lysate containing at least 0.5 ng/μl of YFP (at least 350 ng of total YFP per IP) should be used for each FLIP to ensure a reliable measurement and quantification of fluorescence from the beads using FLIP performed in 1.5 ml tubes (large volume FLIP). Our characterizations showed that the sensitivity of the FLIP increased with an increase of YFP-protein in the lysate used for the assay. To ensure reliable FLIP results, especially for antibodies with weak affinity, the use of more than 350 ng of YFP-protein in the initial lysate is recommended.
A much lower amount of YFP and lysate is needed for mini-FLIP assays, in which 100–120 μl of lysate containing at least 0.3 ng/μl of YFP-protein can be used.
We have observed the YFP fluorescence on the beads to be highly stable in our FLIP assays. YFP signal coating the agarose beads was still observed up to one week after IP, with a negligible decrease in FLIP sensitivity, when the plate containing the FLIP beads was stored in the dark at 4 °C.
Large Volume FLIP
Lysates from cells expressing the YFP-protein of interest were aliquoted into 1.5 ml tubes (700 μl per tube). 5 μg of tested or control IgG antibodies was added directly into the lysates and incubated for 1 h at 4 °C on a nutating mixer (Fisher Scientific) to ensure continuous agitation of the sample. 50 μl of Protein A/G agarose beads (Santa Cruz, Prod. No sc-2003) slurry was then added to each tube and incubated for 30 min at 4 °C on a roller. Beads were washed 5 times in Triton buffer (800 μl of buffer per wash). During the last wash, 10 μl of beads/buffer solution were transferred into black 384 well plates for analysis with a microscope, while the remainder of the beads were collected and processed for immunoblotting analysis. For the latter analysis, the immune complexes were eluted from the beads by adding 1X NuPage LDS sample buffer (Life Technologies, Prod. No NP0007) supplemented with 350 mM β-mercaptoethanol. Samples were heated at 75 °C for 10 min and stored at −20 °C until immunoblotting analysis.
High Throughput (HiT) FLIP
For the parallel analysis of mouse monoclonal antibodies using FLIP and IP/IB assays presented in Fig. 5, a 96-well format was used for IP/FLIP. 700 μl of lysates prepared as described above were aliquoted to each well of a 96-deep-well plate (Nunc). The 96-well format allowed for manipulation of the samples using multichannel pipettes. Each YFP-protein expressed in HeLa cells was treated with 0, 50 or 1000 ng/ml doxycycline as described above. 5 μg of antibodies was added to the appropriate wells containing the target YFP-protein of interest. Every antibody required 4 wells for analysis: mouse IgG control Ab was added to a well containing lysates of cells treated with 1000 ng/ml doxycycline; anti FLAG-M2 Ab was added to a well containing lysates of cells treated with 1000 ng/ml doxycycline; mAb to be tested was added to a well containing lysates of cells treated with 1000 ng/ml doxycycline and to a well containing lysates of cells treated with 50 ng/ml doxycycline. The plate was sealed with aluminum sealing tape and incubated 1 h at 4 °C on a nutator to ensure continuous agitation of the sample. 50 μl of Protein A/G agarose bead slurry were added to each well and incubated for 30 min at 4 °C on a nutator. Beads were then washed 5 times with Triton buffer without disturbing the bead pellet when removing the supernatant. Approximately 50–100 μl of Triton buffer was left in each well during washes, and long loading tips (Fisher) were used to discard the washes to minimize bead loss. Multichannel tip adaptors attached to a vacuum pump were used for quick handling of the samples during washes. After the last wash (with 800 μl), 10 μl of beads from the mouse IgG IP and from the IP with the tested Ab was collected from each well and transferred to a 384 well plate. These samples were then analyzed for FLIP while the rest of the beads were processed for immunoblotting analysis. The left-over beads were collected by centrifugation and the last wash was discarded using loading tips. The bead pellet can be quickly touched to get rid of most of the buffer before eluting the immuno-complexes into 1X NuPage LDS sample buffer supplemented with 350 nM β-mercaptoethanol. The plate was then heated to 75 °C for 10 min and the sample buffer containing the immune complexes that were eluted from the beads was transferred to 96-well PCR plates, sealed and stored at −20 °C until immunoblotting analysis.
Lysates from cells expressing the YFP-protein of interest were aliquoted into PCR tubes (100–120 μl per tube). FLIP was then conducted using the large volume FLIP procedure as described above, except 10 μl of protein A/G slurry per tube and 100 μl of Triton buffer for each wash was used. 10 μl of beads was collected during the last wash, and the fluorescence coating the beads was analyzed with a fluorescence microscope.
After IP, a distinct and sharp fluorescent ring around the edges of the beads was indicative of a positive FLIP, whereas a uniform, unfocused fluorescence on the bead was indicative of background signal. A simple examination of the collected pictures by eye is sufficient for qualitative determination of FLIP results, and discrimination of FLIP-positive from FLIP-negative antibodies.
In this report, FLIP analysis was performed using pictures of YFP-coated beads analyzed with ImageJ (Micro-manager 1.4.18, University of California). For a simple quantification of the total mean fluorescence performed using ImageJ, it was important that the beads were allowed to settle and form a uniform layer on the bottom of the plate, and that a comparable number of beads were analyzed for different IPs from the same lysate. A low but comparable number of beads was preferable over an excess of beads distributed in different layers in the well. An EVOS FL auto (Life technologies) or a BD-pathway microscope (BD bioscience) was used to automatically collect images of the YFP coating the beads after IP, but any fluorescence microscope with an appropriate filter should be suitable for the acquisition of FLIP images. 384-well black plates with clear bottoms were used and pictures were taken with a 20X magnification, keeping the microscope parameters (exposure time, contrast and gain) constant for all the pictures taken for the same experiment. All the pictures were then opened with ImageJ software and the brightness/contrast set at equal values for all the pictures of the same experiment. The total mean fluorescence of the pictures was then measured and the value obtained from the corresponding IgG control IP subtracted from the mean fluorescence obtained from the IP using the tested mAb (or FLAG Ab) ([mean fluorescence from IP using FLAG antibodies] – [mean fluorescence from IP using IgG control antibodies]).
Software for FLIP Image Analysis
The FLIP image analysis software was developed in Python utilizing the Sci-Kit Image analysis package and a multitude of other supporting packages. The function of the software is to isolate and quantify the image signal from two regions of interest, the prominent bright ring at the perimeter of the agarose bead, and the body of the bead itself. Once the features are isolated, the average intensities of the respective regions are calculated and the FLIP signal is returned as the difference between the signals for the two regions. The true FLIP signal for the real-IP image was then calculated by subtracting the FLIP signal for its corresponding IgG control. Before using this software it is recommended that a uniform distribution of beads in the sample be visually confirmed in advance. A free and publicly available analysis tool for FLIP pictures can be found at http://openslice.fenyolab.org/flip.
FLIP Assay Using HuEV-A Expression Vector
In the FLIP assay, following cell lysis and incubation of lysate, beads, and antibody, the bead/immune-complex mixture is washed and then directly visualized with a fluorescence microscope or automated microscopy system. If the YFP-tagged protein of interest has been successfully immunoprecipitated the beads will be coated with a fluorescent signal detectable under light with an appropriate excitation wavelength. On the other hand, if the IP fails, the beads will not fluoresce and no signal will be detected by fluorescence microscopy. Quantification of the fluorescence signal of the beads against a control signal from a control IP performed using immunoglobulins from a non-immunized mouse (IgG) provides a reliable indication of the success of the IP (Fig. 1).
In contrast to the standard IP/IB analysis, the FLIP assay is a much faster procedure because it is based on the direct observation of fluorescent target proteins coating the beads. Moreover the FLIP assay can be integrated into any IP workflow since FLIP analysis requires only a small fraction of the bead/immune-complex product; the left-over beads can be processed for elution and follow-up analysis as required (Fig. 1).
FLIP is a Sensitive and Scalable Assay
Next we characterized the sensitivity of the FLIP assay compared to standard IP/IB procedures. For these analyses we used proteins overexpressed from the HuEV-A vector that, despite the still double band visible by immunoblotting analysis, is an easy to clone vector that enables Tet-regulation and high expression in mammalian cells.
To interrogate the efficiency and sensitivity of the FLIP assay compared to standard IP/IB, a randomly chosen protein (HES-1) expressed from the HuEV-A vector was used together with a well-known commercial antibody against FLAG-tag (Anti-FLAG M2; Biorad). Different amounts of the HuEV-A expressed protein/cell lysate were used to perform FLIP and IP/IB in parallel. The tagged-HES-1 was expressed in previously transfected Tet-on HeLa cells (see experimental procedures). The fluorescence of the cell lysate was measured and the amount of YFP per microliter of solution deduced using a standard curve of purified YFP (Additional file 1: Figure S3). Different amounts of cell lysate derived from cells expressing the tagged protein were used in both FLIP assays or standard IP/IB using an anti-FLAG commercial antibody (Fig. 3a, b). The FLIP signal calculated using ImageJ was plotted versus the corresponding amount of YFP-protein used for the assay (Fig. 3a). We observed a linear increase of fluorescent signal over background by FLIP similar to the increase of the Immunoblot signal following IP, indicating a very similar sensitivity between the FLIP assay and the IP/IB assay.
To identify the minimum amount of cells necessary to obtain a reliable FLIP signal, the lysates from HeLa cells transfected and cultured in different size plates (96, 48, 24, 12, 6-well plates or 6-cm plates) were used to perform FLIP and IP/IB analysis in parallel. Even with minimal starting material (0.02×106 cells plated in one well of a 96-well plate) we observed a low but positive FLIP signal. For all the utilized conditions, a good correlation between the FLIP and the IP/IB analyses was observed (Fig. 4). A reliable FLIP signal consistently above background signal was obtained from 48-well plates.
Overall these observations show an excellent correlation between standard IP/IB analysis as compared to FLIP analysis. Therefore, the FLIP assay presents sensitivity comparable to the classical IP/IB analysis while minimizing assay time and sample handling. These features confer to the assay a much higher potential for high-throughput. These advantages of FLIP make our new technique a relevant and appealing tool for several molecular biology applications including the high-throughput screening of antibodies of unknown performance in immunoprecipitation.
High-Throughout FLIP Screening for IP Competent Mouse Antibodies
Overall these observations demonstrate the high-throughput capability of the FLIP assay applied for the screening of antibodies able to efficiently immunoprecipitate their target protein.
FLIP can be Used to Screen Unpurified Antibodies
This result demonstrated that unpurified antibodies can be efficiently used for FLIP analysis to predict the IP performance of the tested immunoglobulins.
Development of Software for Automated Analysis of FLIP Images
The quantification of the FLIP signals presented in Figs. 3, 4, 5, 6 and 7 was done using ImageJ by measuring the total mean fluorescence of the acquired images (see Experimental Procedures). This analysis, although extremely simple and accessible to any laboratory, does not account for the number of beads as well as their area, and is extremely sensitive to possible fluorescent contaminants that aberrantly increase the measured FLIP signals. Moreover we observed that in control IPs using normal mouse IgG certain agarose beads exhibited high fluorescence. Although high, this background fluorescence was easily distinguishable from the fluorescence caused by specific IP because the background fluorescence was always uniformly distributed on the beads, whereas the IP signal was clearly characterized by a ring of strong fluorescence around the beads. We therefore developed software that measures and compares the fluorescence on the rims of the beads with the fluorescence present on the body of the beads. This procedure was applied to every bead in the picture taken as a single entity. Quantifying FLIP signals in this way therefore is independent of the bead number and excludes signals from non-spherical shapes (likely contaminants and/or overlapping beads). We applied the improved analysis method to the FLIP analysis presented in Fig. 6 (mini-FLIP) and Fig. 7 (FLIP using unpurified antibodies) and obtained a remarkable improvement in FLIP performance, especially when used for more challenging applications such as using unpurified antibodies (Fig. 8). The linear correlation (R2) between the amount of target protein used for IP and mini-FLIP signal was slightly improved using the new analysis (from 0.917 to 0.947) (Fig. 6c versus b). More importantly, we observed a great increase in separation between FLIP signal from FLAG IP or tested antibody IP and background signal (IgG IP) (respectively, Fig. 6b versus Fig. 8a and Fig. 7 versus Fig. 8c and d). The quantified FLIP signal from control IPs was always negative using the new methodology of analysis, reflecting the fact that the edges of the control beads were actually darker than the body of the beads (Fig. 8a).
Moreover, the improved analysis clearly outperformed the ImageJ analysis in FLIP done using unpurified antibodies (Figs. 7 and 8c, d). Analysis of FLIP results with the improved method produced FLIP values that almost perfectly correlated with the IP/IB in our screening of hybridoma supernatants produced after immunization of mice with UPF1 protein (Fig. 8c, d and previous paragraph).
The principles used by the new quantification method are described in further details in the “experimental procedures” section. A publicly available analysis tool for FLIP pictures can be found at http://openslice.fenyolab.org/flip.
This new tool improves the sensitivity of the FLIP, especially for the low performing antibodies and moreover increases the throughput of the FLIP assay, rendering this method an extremely easy, fast, and reliable way to screen antibodies or even hybridoma supernatants for applications in immunoprecipitation.
Immunoprecipitation (IP) procedures are applied to a large variety of molecular biology assays, including protein purification, concentration, co-immunoprecipitation and chromatin immunoprecipitation [2–8]. The classical IP procedure is often coupled to immunoblotting analysis, a technique suited to low throughput analysis of only a few samples [10, 20, 21]. Here we describe and characterize FLIP (Fluorescence IP), which couples conventional IP to the direct observation and/or quantification of a fluorescently tagged target protein on the surface of beads. This approach is faster and cheaper than using IB to measure the success of an IP, and is thus amenable to high-throughput screens. A major limitation of the IP assay is often the availability of antibodies able to efficiently immunoprecipitate the target protein . Here we applied FLIP to the screening of mouse monoclonal antibodies in order to quickly screen and identify those able to immuneprecipitate their corresponding target protein.
Here we show that FLIP has a similar sensitivity to the IP/IB assay and that FLIP can also be miniaturized to increase throughput by minimizing both volume of cell lysate and amount of antibody used for IP. Another advantage of FLIP is the use of fluorescently labeled target protein, which ensures a degree of specificity for the antibodies that pass FLIP. Indeed, given that the target protein is the only fluorescently tagged protein in the cell lysate used for IP, the presence of fluorescence coating the agarose beads after immunoprecipitation indicates selective binding of the tested antibody to the target protein. On the other hand, the FLIP assay is “blind” to any possible non-specific binding of the tested antibody to proteins different from the fluorescently labeled target protein. Another possible limitation of FLIP is that it does not currently give any information about the ability of the tested antibody to recognize the target protein in its denatured state and therefore FLIP cannot predict the behavior of antibodies in immunoblotting assays. To this end, FLIP could theoretically be used using protein solutions/lysates pre-treated with denaturing conditions such as high SDS concentrations and/or reducing agents . This approach may be possible and somehow informative because of the high stability of GFP and GFP derived fluorophores that can withstand denaturing conditions undergoing partial denaturation and still remaining fluorescent .
Also, FLIP cannot assay the performance of an antibody in IP of endogenous proteins because it relies on the overexpression of the target protein. Nevertheless, we show here that FLIP can easily be integrated into classical IP/IB procedures with no disruption and no diminished sensitivity of the latter. This allows for further characterization of the antibodies after FLIP screening for the IB analysis of bands of the endogenous protein. Moreover, FLIP relies on the overexpression of the target protein as a fluorophore-tagged fusion and therefore requires the cloning of the gene of interest into a suitable expression vector. To this end we developed a flexible mammalian expression vector named HuEV-A that combines the simplicity of Gateway cloning with the flexibility of our 3xFLAG-V5-YFP tag [18, 19, 24]. We used the Life Technologies Ultimate ORF collection, compatible with Gateway cloning (Gateway entry clone format) to construct a HuEV-A library of proteins. Also, the HuEV-A tag can be dramatically reduced in size with simple Cre recombination to a 3xFLAG-V5 tag, or the entire tag can be eliminated by employing FLP recombination.
The FLIP assay is very similar in basic concept to the previously described LUMIER (luminescence-based mammalian interactome mapping) assay [12, 13] which has been applied to a wide range of applications, including identification of new protein-protein interactions [12, 25–27], validation of yeast two hybrid screening hits [28, 29] and identification of particular antibodies in patients’ blood (Luciferase Immunoprecipitation Systems or LIP) . LUMIER exploits the overexpression of a target protein fused to Renilla luciferase and utilizes quantification of luminescence signal after immunoprecipitation. As for FLIP, the success of the immunoprecipitation of the target fusion protein can be efficiently and easily determined by high-throughput measurements: measurement of luminescence after incubation with a luciferase substrate in the case of LUMIER and LIP, and measurement of the fluorescence coating the immunoprecipitated beads in the case of FLIP. Despite the great success and the many applications of the LUMIER system no alternative assays with high-throughput capability has been developed for IP analysis until now.
Our FLIP assay is not meant to substitute for the LUMIER assay, which is probably more sensitive compared to IP/WB and FLIP analysis itself. Our FLIP assay is meant to complement the previously known techniques (IP/WB and LUMIER assays) in high-throughput procedures challenging for luciferase reactions or in contexts in which microscopy observations are more amenable than luciferase signal quantification or in which fluorophore tagging is preferable to Renilla tagging. The FLIP assay is an additional tool for the fast screening of IP reactions. FLIP can also be envisioned as an orthogonal assay to the LUMIER system because “preys” tagged with FLAG and a fluorophore of choice can be used with Renilla tagged “baits” (see ) to verify and quantify efficiency of IP before luminescence quantification. In this context our HuEV-A expression vector that includes a FLAG, V5 and YFP tag would be optimal.
Compared to IP-Western blotting, some information such as the nonspecific reactivity of the antibody towards non-target proteins, and/or the preferential reactivity towards specific isoforms or splice variants of the target protein, are lost. Nevertheless, we foresee an extremely useful application of FLIP in screening for antibodies that function efficiently in IP assays. For instance, many unpurified antibodies can be tested by FLIP directly using hybridoma supernatants. After selection of the FLIP-positive clones, a reduced number of the antibodies would need to be purified and characterized further by IP/IB or other assays. Moreover, because of the minimal material used in the FLIP analysis, a new IP does not need to be performed again for the FLIP-positive antibodies, as the left-over beads from FLIP analysis can be used for the follow-up characterizations.
Here we show that FLIP is a fast and reliable method that can partially substitute and easily complement the conventional IP/IB procedure. This new technique can be directly applicable to the high-throughput screening for the identification of IP-grade antibodies.
Immunoprecipitation (IP) is a powerful and informative procedure often coupled to immuno-blotting analysis (IP/IB). IP/IB has limited scalability to high-throughput pipelines. Here we describe a novel, fast and easily scalable method called FLIP (FLuorescence IP) that can partially substitute and complement the conventional IP/IB procedure. Beside the description and application of the new method, we also built and characterized a flexible plasmid (HuEV-A) for YFP protein tagging and we developed a publicly available analysis tool for FLIP pictures that improves the sensitivity of FLIP.
Ab, antibody; FLIP, fluorescence-IP; IB, immunoblot; IgG, immunoglobulin; IP, immunoprecipitation; IP/IB, immunoprecipitation followed by immunoblot analysis; mAb, mouse monoclonal antibody; S.D., standard deviation
Special thanks to Dr. Jessica Chapman-Lim and Dr. Beatrix Ueberheide at the NYULMC Proteomics Resource Center who performed the mass spectrometry analysis of the HuEV-A expressed proteins. We thank Dr. Shane Miersch, University of Toronto, for helpful discussions. Work reported here was supported by NIH grant U54 HG006434 to JDB.
Work reported here was supported by NIH grant U54 HG006434 to JDB. CDI Laboratories, Inc. produced antibodies described in this publication; JDB is a founder of CDI, owns stock, and serves on its Board of Directors. The terms of this arrangement have been reviewed and are managed by the NYU Langone Conflict of Interest Management Unit (CIMU).
Availability of data and materials
The publicly available analysis tool for FLIP pictures can be found at http://openslice.fenyolab.org/flip.
The HuEV-A vector has been deposited at Addgene (plasmid number 68342).
PM performed, analyzed and interpreted the experiments presented in the manuscript and wrote the manuscript. TL developed and implemented the publicly available software for FLIP analysis. DL performed the IP/Western blotting analysis using unpurified antibodies. DJE is the CDI laboratories representative involved in the production of all monoclonal mouse antibodies used in this manuscript. DF supervised the development of the software for the FLIP analysis and helped to draft the manuscript. JDB conceived of the study, supervised the study and helped to draft the manuscript. All authors read and approved the final manuscript.
CDI Laboratories, Inc. produced antibodies described in this publication; JDB is a founder of CDI, owns stock, and serves on its Board of Directors. The terms of this arrangement have been reviewed and are managed by the NYU LangoneConflict of Interest Management Unit (CIMU)
Consent for publication
Ethics approval and consent to participate
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Consortium EP. A user’s guide to the encyclopedia of DNA elements (ENCODE). PLoS Biol. 2011;9:e1001046.View ArticleGoogle Scholar
- Legrain P, Aebersold R, Archakov A, Bairoch A, Bala K, Beretta L, Bergeron J, Borchers CH, Corthals GL, Costello CE, et al. The human proteome project: current state and future direction. Mol Cell Proteomics. 2011;10:M111 009993.View ArticlePubMedPubMed CentralGoogle Scholar
- Huttlin EL, Ting L, Bruckner RJ, Gebreab F, Gygi MP, Szpyt J, Tam S, Zarraga G, Colby G, Baltier K, et al. The BioPlex Network: A Systematic Exploration of the Human Interactome. Cell. 2015;162:425–40.View ArticlePubMedPubMed CentralGoogle Scholar
- Malovannaya A, Lanz RB, Jung SY, Bulynko Y, Le NT, Chan DW, Ding C, Shi Y, Yucer N, Krenciute G, et al. Analysis of the human endogenous coregulator complexome. Cell. 2011;145:787–99.View ArticlePubMedPubMed CentralGoogle Scholar
- Harlow E, Lane D. Using antibodies : a laboratory manual. New York: Cold Spring Harbor Laboratory Press; 1998.Google Scholar
- Harlow E, Lane D. Immunoprecipitation: purifying the immune complexes. CSH Protoc 2006, 2006Google Scholar
- Tansey WP. Nondenaturing protein immunoprecipitation from Mammalian cells. CSH Protoc. 2007;2007:pdb prot4620.PubMedGoogle Scholar
- Landt SG, Marinov GK, Kundaje A, Kheradpour P, Pauli F, Batzoglou S, Bernstein BE, Bickel P, Brown JB, Cayting P, et al. ChIP-seq guidelines and practices of the ENCODE and modENCODE consortia. Genome Res. 2012;22:1813–31.View ArticlePubMedPubMed CentralGoogle Scholar
- Gasper WC, Marinov GK, Pauli-Behn F, Scott MT, Newberry K, DeSalvo G, Ou S, Myers RM, Vielmetter J, Wold BJ. Fully automated high-throughput chromatin immunoprecipitation for ChIP-seq: identifying ChIP-quality p300 monoclonal antibodies. Sci Rep. 2014;4:5152.View ArticlePubMedPubMed CentralGoogle Scholar
- Burnette WN. “Western blotting”: electrophoretic transfer of proteins from sodium dodecyl sulfate--polyacrylamide gels to unmodified nitrocellulose and radiographic detection with antibody and radioiodinated protein A. Anal Biochem. 1981;112:195–203.View ArticlePubMedGoogle Scholar
- Marcon E, Jain H, Bhattacharya A, Guo H, Phanse S, Pu S, Byram G, Collins BC, Dowdell E, Fenner M, et al. Assessment of a method to characterize antibody selectivity and specificity for use in immunoprecipitation. Nat Methods. 2015;12:725–31.View ArticlePubMedGoogle Scholar
- Barrios-Rodiles M, Brown KR, Ozdamar B, Bose R, Liu Z, Donovan RS, Shinjo F, Liu Y, Dembowy J, Taylor IW, et al. High-throughput mapping of a dynamic signaling network in mammalian cells. Science. 2005;307:1621–5.View ArticlePubMedGoogle Scholar
- Blasche S, Koegl M. Analysis of protein-protein interactions using LUMIER assays. Methods Mol Biol. 2013;1064:17–27.View ArticlePubMedGoogle Scholar
- Symer DE, Connelly C, Szak ST, Caputo EM, Cost GJ, Parmigiani G, Boeke JD. Human l1 retrotransposition is associated with genetic instability in vivo. Cell. 2002;110:327–38.View ArticlePubMedGoogle Scholar
- Weidenfeld I, Gossen M, Low R, Kentner D, Berger S, Gorlich D, Bartsch D, Bujard H, Schonig K. Inducible expression of coding and inhibitory RNAs from retargetable genomic loci. Nucleic Acids Res. 2009;37:e50.View ArticlePubMedPubMed CentralGoogle Scholar
- Harlow E, Lane D. Immunoaffinity purification: coupling antibodies to protein a or g bead columns. CSH Protoc 2006, 2006Google Scholar
- Bjorck L, Kronvall G. Purification and some properties of streptococcal protein G, a novel IgG-binding reagent. J Immunol. 1984;133:969–74.PubMedGoogle Scholar
- Katzen F. Gateway((R)) recombinational cloning: a biological operating system. Expert Opin Drug Discov. 2007;2:571–89.View ArticlePubMedGoogle Scholar
- Nagai T, Ibata K, Park ES, Kubota M, Mikoshiba K, Miyawaki A. A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications. Nat Biotechnol. 2002;20:87–90.View ArticlePubMedGoogle Scholar
- Harlow E, Lane D. Antibodies: a laboratory manual. Cold Spring Harbor: Cold Spring Harbor Laboratory; 1988.Google Scholar
- Greenfield EA. Antibodies: a laboratory manual. Second edition. edn.CHS. Cold Spring Harbor, New York: Laboratory Press; 2014.Google Scholar
- Bradbury A, Velappan N, Verzillo V, Ovecka M, Chasteen L, Sblattero D, Marzari R, Lou J, Siegel R, Pavlik P. Antibodies in proteomics II: screening, high-throughput characterization and downstream applications. Trends Biotechnol. 2003;21:312–7.View ArticlePubMedGoogle Scholar
- Saeed IA, Ashraf SS. Denaturation studies reveal significant differences between GFP and blue fluorescent protein. Int J Biol Macromol. 2009;45:236–41.View ArticlePubMedGoogle Scholar
- Einhauer A, Jungbauer A. The FLAG peptide, a versatile fusion tag for the purification of recombinant proteins. J Biochem Biophys Methods. 2001;49:455–65.View ArticlePubMedGoogle Scholar
- Ellis JD, Barrios-Rodiles M, Colak R, Irimia M, Kim T, Calarco JA, Wang X, Pan Q, O’Hanlon D, Kim PM, et al. Tissue-specific alternative splicing remodels protein-protein interaction networks. Mol Cell. 2012;46:884–92.View ArticlePubMedGoogle Scholar
- Ryzhakov G, Teixeira A, Saliba D, Blazek K, Muta T, Ragoussis J, Udalova IA. Cross-species analysis reveals evolving and conserved features of the nuclear factor kappaB (NF-kappaB) proteins. J Biol Chem. 2013;288:11546–54.View ArticlePubMedPubMed CentralGoogle Scholar
- Taipale M, Tucker G, Peng J, Krykbaeva I, Lin ZY, Larsen B, Choi H, Berger B, Gingras AC, Lindquist S. A quantitative chaperone interaction network reveals the architecture of cellular protein homeostasis pathways. Cell. 2014;158:434–48.View ArticlePubMedPubMed CentralGoogle Scholar
- Caufield JH, Sakhawalkar N, Uetz P. A comparison and optimization of yeast two-hybrid systems. Methods. 2012;58:317–24.View ArticlePubMedGoogle Scholar
- Blasche S, Mortl M, Steuber H, Siszler G, Nisa S, Schwarz F, Lavrik I, Gronewold TM, Maskos K, Donnenberg MS, et al. The E. coli effector protein NleF is a caspase inhibitor. PLoS One. 2013;8:e58937.View ArticlePubMedPubMed CentralGoogle Scholar
- Burbelo PD, Ching KH, Klimavicz CM, Iadarola MJ. Antibody profiling by Luciferase Immunoprecipitation Systems (LIPS). J Vis Exp. 2009.Google Scholar