A Fast and Economical Sample Preparation Protocol for Interaction Proteomics Analysis

A simple and fast immunoprecipitation (IP) protocol is designed with the sample preparation incorporated, applicable to both low and high throughput. This new protocol combines two procedures based on magnetic beads in 96‐well plate format. Protein complexes are captured by antibodies and magnetic beads conjugated with protein A. Proteins are washed and on‐bead digested by using Single‐Pot solid‐phase sample preparation (SP3). The whole IP‐SP3 approach can be completed in one day, which is considerably faster compared to the classical approach. No major quantitative differences are found between SP3 and FASP (filter‐aided sample preparation) or a longer incubation protocol. Taken together, the IP‐SP3 protocol is a fast and economical approach easily applicable for large‐scale protein interactome analysis.


DOI: 10.1002/pmic.201900027
method is easier and faster to perform without compromising the sensitivity.
Our new protocol employs 96-well plates for all steps and can be conveniently handled by a manual multi-channel pipette. We combined the use of two procedures based on magnetic beads, namely 1) the capture of antibody-bound protein complexes by magnetic beads conjugated with protein A, and 2) the washing of proteins and their trypsin digestion by Single-Pot solid-phase sample preparation (SP3 [5,6] ) protocol. SP3 is based on the use of magnetic beads coated with carboxylate hydrophobic groups that bound proteins at high organic solvent concentration. Proteins are digested on-bead by trypsin in an aqueous buffer. The absence of organic solvent concomitantly releases proteins/peptides from the beads into the digestion buffer. The consecutive IP followed by SP3 can be completed in 1 day, which is substantially faster than the classical approach that requires ࣙ3 days. [2] We applied the protocol to reveal synaptic protein-protein interactions. [7] Frozen mouse brain cortex was extracted in 0.5% n-dodecyl β-D-maltoside extraction buffer as previously described [1] (one cortex in 3 mL extraction buffer), and centrifuged twice at 16 000 × g for 20 min to remove insoluble materials. We recovered about 2 mg protein per mL of buffer. Four hundred and fifty µL of supernatant was dispensed to each well of the 96-well plate (Deepwell plate 96/500 µL from Eppendorf). Ten µg of antibody (rabbit polyclonal antibodies against Dmxl2, Gpr158, Gabra4, and Mpp2, custom-made by Genscript) and 60 µL of Protein A Magbeads (Genscript) were added to each well. This standard protocol works well for most IPs on typical synaptic proteins. [8] We expect that 5-10-fold less input material and antibody should be sufficient in case a highly sensitive LC-MS/MS platform is employed, such as the Orbitrap HF-X [9] or the ion mobility QTOF mass spectrometer (TimsTOF pro) equipped with PASEF technology. [10] The plate was covered with a water-tight silicone sealing mat (Eppendorf) and incubated at 4°C for 3 h in rotation at 20 rpm on an ELMI Intelli-Mixer. The plate was then placed on a 24-pins magnetic plate (MagnaBot 96 Magetic separation device from Promega), in which the beads were pulled rapidly to the side of the well closest to the magnetic bar. This facilitates the aspiration of solution from the bottom of the well with no interference from Gpr158, Gabra4, and Mpp2 IPs using SP3 or FASP). Each experiment was performed in triplicate. The median and number of proteins is indicated in each box. B) Overlap of proteins identified between SP3 and FASP protocol with each antibody and all combined. Only proteins identified by at least two peptides were included (N). C) LFQ intensity comparison between SP3 and FASP protocol for the bait protein of each antibody and two wellknown interactors. D) LFQ intensity comparison for Dmxl2 and two well-known interactors between overnight and 3 h incubation for antibody-antigen interaction using IP-SP3 protocol. E) Potential interactors most enriched in each IP-SP3 experiment. Each node represents a protein, color-coded by iBAQ intensity (log10 scale). the beads. The solution was discarded. The washing procedure was repeated four times with 350 µL 0.1% Triton-X 100 in 25 mm HEPES buffer (pH 7.4) at room temperature. Next, 50 µL 2% SDS in 75 mm Tris buffer (pH 7.5) containing 1 µL Tris(2carboxyethyl)phosphine hydrochloride was added to each well and shaken in a Thermoshaker (Eppendorf) at 800 rpm, 50°C, for 45 min. Afterward, 0.5 µL S-Methyl methanethiosulfonate was then added and shaken for 5 min at room temperature.
The eluted proteins from Magbeads were transferred to a 96-well microplate with conical bottom (Microplate 96/V-pp from Eppendorf). Ten µg of Sera-Mag Speed beads [6] (Thermo Scientific, combination of GE Healthcare catalogue number www.advancedsciencenews.com www.proteomics-journal.com 45152101010250 and 65152105050250, washed twice in water) and 150 µL 100% acetonitrile were added to each well. After mixing by pipetting, the microplate was gently shaken for 20 min at room temperature. The microplate was transferred to the Epigentech magnetic raft and the solution was discarded. After washing four times in 75% ethanol and 4 times in 100% acetonitrile, the beads were incubated with 100 µL trypsin (0.6 µg trypsin) in 50 mm ammonium bicarbonate overnight at 37°C. The microplate was transferred to the 96 ring-shaped magnetic plate (96S Super Magnet Plate from Alpaqua). Sera-Mag Speed beads were captured around the perimeter of the well bottom, creating a bead free center for the easy collection of the peptides solution. The solution was transferred to a new 96-well plate for speedvac drying, and stored until LC-MS/MS.
For comparison, we also performed the sample preparation using the well-established FASP protocol. [4] In brief, eluted proteins were transferred to a Microcon-30 tube (Millipore), washed four times with 8 m Urea and then four times with 50 mm ammonium bicarbonate by centrifugation at 13 600 × g. Proteins were digested with trypsin overnight at 37°C. Peptides were collected by centrifugation, transferred to a new Eppendorf tube and dried in a speedvac.
Peptides were analyzed by a 2 h run on a nano-HPLC coupled to Orbitrap discovery mass spectrometer. [4] The top five most intense peptide peaks from MS1 spectrum were selected for MS/MS analysis. We use label-free quantification and enable normalization by MaxQuant (version 1.6.2.3, searched against the UniProt mouse proteome, 2015-02 release) with match between run to improve protein identification across the samples. Each experiment was performed in triplicate.
When all identified proteins were taken into account, coefficients of variation (CV) of about 20% were observed for all conditions ( Figure 1A). In some cases, FASP experiments had slightly lower CV (Dmxl2 and Gpr158), whereas in others SP3 experiments had slightly lower CV (Gabra4 and Mpp2). The overlap of identified proteins (with at least two peptides) between SP3 and FASP was high (overall 96%, Figure 1B). Equally, protein abundances (LFQ) recovered from SP3 and FASP experiments were similar, as exemplified by the bait proteins and their interactors in Figure 1C.
As previous studies often use overnight incubation for antibody-antigen interaction, [1,4,11] we compared the 3 h incubation used in our IP-SP3 approach to the overnight incubation protocols with the antibody against Dmxl2. There were no major quantitative differences in the recovery of bait proteins and their interactors between these protocols ( Figure 1D).
The four bait proteins examined in the present example study are expected to reside in distinct protein complexes. Therefore, each experiment could be used as negative control for each other to reveal the protein interactomes. We considered as potential interactors the proteins that showed at least 10-fold iBAQ intensity enrichment in all replicates of an IP set compared to the sets of other baits ( Figure 1E). The potential interactors with highest intensity sets are shown. However, the cross-reactivity of antibodies (common with polyclonal antibodies) resulting in false protein interactions cannot be excluded by performing small IP sets alone (see refs. 11,12), which emphasizes the value of highdensity, high-throughput approaches that are enabled by the protocol coined in this manuscript.
In short, we demonstrated that SP3 is a viable alternative to FASP [4] for IP-based interaction proteomics. Significantly, the cost of Sera-Mag Speed bead (<0.1 euros per experiment) is substantially lower than the Microcon-30 filter (about 4.5 euros per experiment), making the SP3 protocol more economical. With state-of-the-art mass spectrometers, such as the Orbitrap HF-X [9] or the ion mobility QTOP mass spectrometer (TimsTOF pro) equipped with PASEF technology, [10] it is expected that an IP sample could be analyzed in <15 minutes. The IPs and MS analyses are processed independently, each of which can be completed in a single day. Thus, the IPs and MS analysis of 96 samples from one plate could be synchronized in consecutive days. Taken together, this protocol opens up an avenue for a high-throughput, fast, and economical approach to large-scale protein interactome analysis (see ref. 13).