A commercial antibody to the human condensin II subunit NCAPH2 cross-reacts with a SWI/SNF complex component

Condensin complexes compact and disentangle chromosomes in preparation for cell division. Commercially available antibodies raised against condensin subunits have been widely used to characterise their cellular interactome. Here we have assessed the specificity of a polyclonal antibody (Bethyl A302-276A) that is commonly used as a probe for NCAPH2, the kleisin subunit of condensin II, in mammalian cells. We find that, in addition to its intended target, this antibody cross-reacts with one or more components of the SWI/SNF family of chromatin remodelling complexes in an NCAPH2-independent manner. This cross-reactivity, with an abundant chromatin-associated factor, is likely to affect the interpretation of protein and chromatin immunoprecipitation experiments that make use of this antibody probe.


Introduction
Antibody probes are critical tools for biomedical research. However, their utility depends on both affinity and specificity for the intended target. Cross-reactivity can occur when an antibody has affinity for more than one protein, and can be challenging to predict without empirical data. Cross-reactivity causes particular problems for immunoprecipitation-based applications that aim to identify proteins, or genomic regions, which interact with a protein of interest. In this case, interactions of the cross-reacting protein may be erroneously assigned to the intended antibody target. Control experiments are therefore essential to ensure that molecular interactions identified following immunoprecipitation are real, and not an artefact of cross-reactivity.
In this study, we perform immunoprecipitation mass spectrometry followed by control experiments, to show that a commercial antibody marketed for the detection of the condensin II subunit NCAPH2 cross-reacts with a subunit of SWI/SNF. Therefore, protein and chromatin immunoprecipitation experiments that rely on this antibody (Li et al., 2015;Wu et al., 2019), particularly those showing genome-wide co-localisation between condensin II and SWI/SNF (Wu et al., 2019), should be interpreted with caution.

Results and discussion
In order to identify proteins that physically interact with the condensin II complex, we performed affinity purification mass spectrometry using a polyclonal antibody raised against the C-terminus of the NCAPH2 subunit (Bethyl Labs A302-276A, referred to hereafter as B276). Immunoprecipitation was performed from mouse embryonic stem cell (mESC) whole cell extract, without crosslinking, using either B276 or a rabbit IgG control. The abundance of co-purified proteins was then assessed by Label Free Quantification. As expected, all 5 subunits of the pentameric condensin II complex were significantly enriched in the Ncaph2 immunoprecipitate compared to the IgG control ( Figure 1A). In addition, numerous components of the SWI/ SNF complex family were also enriched to a similar or greater degree than known condensin II subunits. Indeed, the most significantly enriched protein identified in these experiments was the Smarcc1/Baf155 subunit.
In an independent experiment performed in a different laboratory, immunoprecipitations were performed using nuclear lysates from human HEK-293T cells, either after crosslinking with formaldehyde or with non-crosslinked material. Co-purified proteins were identified using MS/MS and analysed with the SAINTexpress algorithm (Teo et al., 2014) to determine fold Scatter plot showing the enrichment of proteins in immunoprecipitations conducted using the Bethyl Labs anti-NCAPH2 (B276) antibody versus rabbit IgG control (n = 4, from cultures grown and processed in parallel), as determined by mass spectrometry and label-free quantification. The x-axis depicts mean Log2 fold-enrichment. The y-axis shows the Log10 probability of enrichment from Bayes-moderated t-tests. Condensin II and SWI/SNF subunits are labelled in blue and red, respectively.  (Table 1 and Table 2). For both crosslinked and non-crosslinked cells, all components of the condensin II pentamer were identified with high enrichment scores, as well as numerous components of the SWI/SNF complex, confirming the independent results observed in mESC.
These results were unexpected, because previous biochemical characterisation of SWI/SNF complexes did not identify condensin subunits as robust interactors (Ho et al., 2009;Kadoch et al., 2013). We therefore considered two possibilities: a direct physiological interaction between condensin II and SWI/SNF, leading to bona fide co-purification, or, on the contrary, a direct recognition of one or more SWI/SNF components by the B276 anti-NCAPH2 antibody, resulting in an experimental artefact.
To distinguish these possibilities, we first performed reciprocal immunoprecipitations using B276 ( Figure 2A) and an antibody against the SWI/SNF subunit Smarca4 ( Figure 2B). Confirming the mass spectrometric analysis, B276 immunoprecipitated both the Smc4 subunit of condensin and the SWI/SNF subunit Smarca4 ( Figure 2A). However in the reciprocal experiment, the anti-Smarca4 antibody robustly immunoprecipitated the SWI/SNF component Arid1a, as expected, but Ncaph2 was not detected by western blots using the B276 probe ( Figure 2B).
In a second independent experimental setup, CRISPR-mediated homology-directed repair was used to homozygously integrate a cassette encoding a double FLAG tag at the 3' terminus of the endogenous Ncaph2 open reading frame in mESCs (Ncaph2 FLAG/FLAG , Figure 2C). Seamless integration of the epitope tag was confirmed by Sanger sequencing. Anti-FLAG antibodies were then used to perform affinity purification from whole cell extract of Ncaph2 FLAG/FLAG or control Ncaph2 +/+ mESCs, without crosslinking, and peptides were identified by mass spectrometry and label free quantification as described above.
All 5 condensin II subunits were significantly enriched in the Ncaph2 FLAG/FLAG immunoprecipitate, whereas SWI/SNF subunits were not ( Figure 2D). Thus, although several SWI/SNF subunits were immunoprecipitated by the B276 anti-NCAPH2 antibody in both human and mouse cells, we were unable to validate these interactions using two independent approaches. This suggested that the immunoprecipitation of SWI/SNF by B276 occurred independently of the NCAPH2 target in both mouse and human cells.
To directly test this hypothesis, we made use of an HCT-116 human cell line in which C-terminally tagged NCAPH2 protein can be rapidly depleted by the addition of indole-3-acetic acid (auxin) to the culture medium ( Figure 3A) (Takagi et al., 2018). Whole cell extracts were prepared from cells cultured either in the presence or absence of auxin treatment (5 hours), and immunoprecipitations were conducted using B276, or using a custom anti-NCAPH2 antibody raised in our laboratory (Methods).
In the absence of auxin treatment, the custom antibody efficiently immunoprecipitated both NCAPH2 and the condensin subunit SMC4, but this was prevented by the auxin-mediated depletion of NCAPH2 ( Figure 3B). The SMARCA4 subunit of SWI/SNF was not detectable by immunoblot in either condition. In contrast, the B276 antibody immunoprecipitated SMARCA4 but not NCAPH2 or SMC4 in this cell line ( Figure 3B). Why B276 can immunoprecipitate Ncaph2 from mESC and HEK-293T cell extracts but not from this HCT-116 cell line is unclear, but could be attributable to the C-terminal AID: mCherry tag in the latter. Importantly, the immunoprecipitation of SMARCA4 by B276 was unaffected by auxin-mediated depletion of NCAPH2. These results confirm that the immunoprecipitation of SWI/SNF complexes by B276 occurs independently of the intended NCAPH2 bait, and therefore do not reflect bona fide physical interactions between condensin II and SWI/SNF.
To further investigate details of SWI/SNF components binding to the 276 antibody, a western blot was performed on HEK 293T cell nuclear lysate as well as a recombinant human condensin II complex (Kong et al., 2020) ( Figure 4A). While recombinant protein resulted in one band which ran in the position of NCAPH2 at around ~80 kDa, the blot of nuclear lysate produced two bands, one at around 80 kDa, likely corresponding to NCAPH2, and an additional band around 135 kDa. We then performed an IP using B276 followed by western blotting with either B276 or an antibody recognising SMARCC1 whose result suggested that both NCAPH2 and SMARCC1 were pulled down. However, the additional 135 kDa band was observed at the same respective height in both blots developed either with B276 or with an anti-SMARCC1 antibody ( Figure 4B).
This finding leads us to suspect that B276 binds with considerable affinity to SMARCC1, in addition to NCAPH2, accounting for the considerable enrichment of SMARCC1 in the B276 IP mass-spectrometry data ( Figure 1A, Table 1 and Table 2).
The cross-reactivity of the B276 antibody with another protein, specifically an abundant chromatin-associated factor, may affect the interpretation of protein and chromatin immunoprecipitation experiments performed using this antibody (Li et al., 2015;Wu et al., 2019). In particular the notion that Condensin II and components of the SWI/SNF chromatin remodeller complexes colocalize on a genome-wide scale (Wu et al., 2019) should be revisited in light of the presented results.
Genome editing. To generate the NCAPH2 FLAG/FLAG mESC line used in Figure 2, the following sgRNA protospacer sequences were cloned into pX335-U6-Chimeric_BB-CBh-hSpCas9n(D10A): guide 1 = GGTGGAAAGTAGTATATACC guide 4 = ACTCAAGGCTGGGCCATGGA The homology-directed repair template was ordered as a double stranded DNA fragment (gBlock, IDT) and cloned using MluI sites into pEGFP-C1 (Clontech). Immunoblot (IB) of nuclear lysate from formaldehyde crosslinked 293T cells probed with B276 yields two bands, a specific band at ~80kDa, that is also present when B276 is used to probe a blot prepared from recombinant condensin II complex (Kong et al., 2020), and a non-specific band (*) at ~135kDa. B. Immunoprecipitation (IP) performed with B276 using 293T cell nuclear lysate, followed by immunoblot probed with either B276 or mouse Anti-SMARCC1 (sc-32763), demonstrating that the non-specific band in B276 (*) runs in the same position as SMARCC1. Uncropped blots are available via Figshare (Wood & Cutts, 2020).

Repair template sequence:
A

C G C G T A A G T G G C C T G G G A C A C A A G G G A T -GGGGCAGCGGCCCTGGACTCTACTGACACCTTGTT-TCCACAGGCTAATGACTACACAGTGGAGATCACTCA G C A G C C A G G A C T G G A G G C A G C T G T G G A C A C A A TGTCTCTGAGACTGCTCACACACCAGCGAGCCCACA C C C G C T T C C A G A C C T A T G C T G C A C C A T C C A T G G C C C A G C C T G A C T A C A A G G A C G A C G A T -G AC A AG G AC TAC A AG G AC G AC G AT G AC A AG T G AG TGGACAGCACTGAGGCAGGGGTGGAAAGTAGTATA T A C C T G G A G G T C T T T G C C C C T A A T G T G C T A T G G G G C C AT T C A C T C C A G T G C T G C C T C C T G G C T G G C C TAG C C TA ATA A AG T G T T G C TAC C C C AC C TG T T C AC C G G AC AG AC TAT T TA A AT G AG C T G C T G G TA C AG AG C AC AT G C AC AG AT TA AG TAC AT C C AT T TA ATGACAGGGCCTAGGCAATAGGTATAGTTCAGACGCGT
The pX335 derivative plasmids encoding sgRNAs and Cas9 nickase were transfected in combination with the repair template into E14 mESC using Lipofectamine 2000 (Thermo Fisher Scientific), according to manufacturer instructions. Transfected cells were FACS-purified based on GFP fluorescence and seeded at limiting dilution, then individual colonies were picked, expanded, screened by PCR band-shift assays, and then homozygous integration was verified by Sanger sequencing.

Western blots. For western blots in
Primary antibodies (detailed in Table 3) were added to the block solution at the dilution shown in the antibody Table. Membranes were incubated in the antibody dilutions with constant agitation, at 4°C overnight. Membranes were washed in TBS-Tween 20 solution (0.1% Tween20; 3 washes × 10 minutes). HRP-conjugated secondary antibodies were also diluted in the block solution (with 0.1% Tween20), and membranes were incubated with secondary antibody dilutions under constant agitation at room temperature for 1 hour. Membranes were then washed in TBS-Tween20 solutions (0.1% Tween20, 3 washes × 10 minutes). Membranes were stained with SuperSignal West Pico PLUS HRP substrate (Thermo Scientific) and then imaged with an ImageQuant LAS 4000 (GE Healthcare).
Immunoprecipitation using crosslinked cells was performed twice, in two independent experiments. Crosslinked cells nuclear lysate was prepared by lysing nuclei using RIPA buffer (50 mM Tris pH 8, 150 mM NaCl, 1% NP-40, 0.1% SDS, 0.5% DOC, protease and phosphatase inhibitors, 1mM DTT) and a Dounce homogeniser. DNA was fragmented with sonication, and DNA fragment size was confirmed to be less than 500 bp by running a sample on 1.5% agarose, stained with SYBR safe (Thermo). DNA was further digested by incubating with Benzonase with 2 mM MgCl 2 for 60 minutes, before being spun at 12,000 g. Nuclear lysate was incubated with dyna beads for 2 hrs, and beads were washed 5 times with RIPA buffer.
Un-crosslinked cell immunoprecipitation was performed once, as described in Bode et al. (2016). Briefly, nuclear lysate was prepared by solubilising nuclear pellet with nuclear lysis buffer (50 mM HEPES (pH7.9), 5mM MgCl 2 , 0.2% TritonX-100, 20% glycerol, 300 mM NaCl, and protease and phosphatase inhibitor) using a Dounce homogeniser with a tight pestle and incubating with Benzonase (Sigma) for 60 minutes at 4°C before being spun at 12,000 g. Nuclear lysate was diluted 1 in 2 with nuclei lysis buffer without NaCl before incubating with Dyna beads for 2 hrs at 4°C. Beads were washed 5 times with nuclear lysis buffer with 150mM NaCl.
After IP for both crosslinked and un-crosslinked samples, on bead digestion was performed as described in Hillier et al. (2019) Multiply charged ions (z = 2 -7) with intensity above 10,000 counts were fragmented in HCD (higher collision dissociation) cell at 30% collision energy, and the isolation window at 1.6 Th. The fragment ions were detected in ion trap with AGC at 10,000 and 35 ms or 50 ms maximum injection time. The dynamic exclusion time was set at 45 s and 30 s with ±10 ppm.
Mass spectrometry data analysis. Raw mass spectrometry data files were analysed with Proteome Discoverer 1.4 (Thermo). Database searches were carried out using Mascot (version 2.2) against the Uniprot human or mouse reference databases (January 2018) appended with the cRAP database (www.thegpm.org/ crap/) with the following parameters: Trypsin was set as digestion mode with a maximum of two missed cleavages allowed. Precursor mass tolerance was set to 10 ppm, and fragment mass tolerance set to 0.5 Da. Acetylation at the N-terminus, oxidation of methionine, carbamidomethylation of cysteine, and deamidation of asparagine and glutamine were set as variable modifications. Peptide identifications were set at 1% FDR using Mascot Percolator. Protein identification required at least one peptide with a minimum score of 20. Protein lists from bait and control experiments were analysed with the SAINTexpress (Significance Analysis of INTeractome) algorithm (Teo et al., 2014) to discriminate true interacting proteins from background binders. Using this method, we identified 693 and 1815 unique protein hits from the immunoprecipitation using non-crosslinked and crosslinked cells, respectively. Hits with the highest fold enrichment scores and SAINT probability score cut-off of 1 were members of either the condensin II or SWI/ SNF complexes, and are shown in Table 1 and Table 2 2 for crosslinked and non-crosslinked cells, respectively.
Identified peptides were analysed using the Contaminant Repository for Affinity Purification at crapome.org to calculate SAINT probability and fold enrichment scores (Mellacheruvu et al., 2013). Top hits were members of either the condensin II or SWI/SNF complex, and are shown in Table 1 and  Table 2, for crosslinked and non-crosslinked samples, respectively.

Western blotting.
To prepare western blots shown in Figure 4, samples of nuclear lysate and IP were mixed with 4x NuPAGE LDS Sample Buffer (Invitrogen) and run on a 4%-12% Bis-Tris NuPAGE gel (Invitrogen) against a Color Protein Standard broad range ladder (NEB) in NuPAGE MES SDS buffer. Protein was transferred on to a nitrocellulose membrane (Amersham), blocked with 5% milk powder in TBS-T, probed with either primary mouse anti-Smarcc1 (sc-32763) or rabbit anti-NcapH2 (A302-276A, lot 1) and secondary ECL anti-mouse or rabbit IgG HRP-linked (Cytiva). Western blot was developed on Hyperfilm ECL (Amersham) using ECL substrate (Pierce).