Identification of Modified Histones as Binding Substrates of Human Spindlin Family Member 4 (SPIN4) by Peptide Arrays and Native Nucleosome Pulldown

Summary

Genetic variants in human Spindlin Family Member 4 (SPIN4) have recently been identified in patients with bone overgrowth, a novel developmental disorder. A protocol and new results describing the biochemical identification of modified histones in the context of nucleosomes as SPIN4 binding substrates are presented.

Abstract

Human Spindlin Family Member 4 (SPIN4) has just emerged as a novel regulator of bone growth. Loss-of-function variants in SPIN4 were identified in patients exhibiting a pre- and post-natal overgrowth syndrome, which is inherited in an X-linked dominant fashion. SPIN4 contains three tandem repeats of Tudor domains, each forming an individual aromatic cage. These Tudor domains are considered epigenetic reader domains for methylated arginine and lysine residues of histones based on sequence homology. However, the exact binding substrates of SPIN4 remain unclear. Human SPIN4, in both its wild-type and disease-causing mutant forms, was successfully generated via an insect cell expression system. By a histone modification peptide array, the differential binding affinities between wild-type and mutant SPIN4 were determined. Functional validations further demonstrated that SPIN4 is a bona fide reader of histone H3 Lysine 4 methylation (H3K4me), as evidenced by biochemical pulldown of native mononucleosomes. This integrated approach facilitates a rapid binding target identification of epigenetic reader proteins.

Introduction

Spindlin Family Member 4 (SPIN4) is an epigenetic reader and has been recently implicated in overgrowth syndrome, presenting with a constellation of pre- and postnatal overgrowth and tall stature^1,2. Spin4 knockout (KO) mice recapitulated the exact human phenotype, supporting the causality. To date, SPIN4 is the only identified exclusive epigenetic reader that negatively regulates bone growth. However, the biochemical properties of SPIN4, a critical epigenetic reader for growth, have not been characterized, leaving a critical knowledge gap in understanding the fundamental, disease-causing mechanisms in this novel developmental disorder. SPIN4 contains three tandem repeats of Tudor domains, which are generally considered methylated arginine or lysine histone reader protein domains³. A SPIN4^D82H missense variant has recently been identified in a patient with prenatal and postnatal bone overgrowth syndrome. This mutation affects the first Tudor domain of SPIN4; however, the relevant binding substrates of these Tudor domains remain unclear, limiting functional dissection of SPIN4-dependent regulation of chromatin and gene expression during bone development.

Identification of new histone-binding substrates of reader protein domains does not rely on a single, standardized approach. Post-translational modifications (PTMs) of histone mostly occur at histone tails that protrude outwards from the core nucleosome particles. The recognition of PTMs on histone tails by reader protein domains is highly variable and often requires multivalent, electrostatic interactions involving nucleosomal DNA, linker DNA, and/or other histone domains to collectively stabilize the reader-target interaction⁴. Currently, PTMs of histone tail peptides have been constructed in the form of microarrays and commercialized by several vendors⁵. While histone tail peptide-based PTM screening for binding targets of reader protein domains is high-throughput, it often yields false-positive hits in the absence of nucleosomes. Further, the experimental conditions for such screens are highly empirical and vary from one protein to another, making them technically difficult to address. Synthetic nucleosome-based PTM screening or pulldown provides a much more robust readout, but it is much more costly and low-throughput, since reconstituting nucleosomes carrying variable combinations of PTMs is inherently challenging.

This protocol describes a rapid approach for identifying histone PTMs in nucleosomes as binding targets of SPIN4. Recombinant SPIN4 was reconstituted and screened against a commercial histone PTM peptide array to identify potential binding targets. Native mono-nucleosomes containing diverse endogenous histone PTMs were subsequently generated from live cells. SPIN4 pulldown of mono-nucleosomes enabled antibody-based targeted validation of histone PTMs following the initial screening. This integrated biochemical strategy demonstrates a proof-of-principle approach for time- and cost-efficient target discovery and can be readily applied to identify binding substrates of other poorly characterized epigenetic reader proteins.

Protocol

All experimental procedures involving HEK293FT cells were conducted in compliance with the Institutional Biosafety Committee (IBC) guidelines of Virginia Commonwealth University.

1. Purification of recombinant human SPIN4

1. Clone full-length wild-type SPIN4 or SPIN4^D82H with an N-terminal 1× FLAG epitope tag into a baculoviral vector^6,7.
2. Prepare baculovirus by transfecting Sf9 cells with baculoviral vectors, harvest viruses when cells burst, aliquot, and store the virus at −80 °C^6,7.
3. In a 500 mL Erlenmeyer flask, add 100 mL growth medium, 1.5 × 10⁸ Sf9 cells, and 1 mL P1 baculovirus.
4. Incubate the culture at 27.5 °C in an orbital shaker incubator set to 125 rpm for 72 h in the dark.
5. Purify FLAG-SPIN4 using FLAG affinity gel in BC150 lysis buffer containing protease inhibitors and elute the protein using 1× FLAG peptide at 0.1 mg/mL.
6. To completely remove the 1× FLAG peptide, dialyze eluted FLAG-SPIN4 against BC150 Protein Storage buffer at 4 °C in the following order:
   1. First dialysis: 1 L, 1 h.
   2. Second dialysis: 2 L, overnight.
   3. Third dialysis: 1 L, 1 h.
7. Concentrate the dialyzed protein to the desired volume (for SPIN4, from 600 µL to 150 µL).
8. Determine protein concentration using a Bicinchoninic Acid (BCA) assay, and calculate molar concentration accordingly.
9. The amount of crude protein recovered at each purification step and the corresponding estimated yields are summarized in the Supplementary Table 1. Purification was performed from 250 mL of Sf9 culture using 400 µL of anti-FLAG affinity gel slurry.
10. Dilute concentrated SPIN4 by 10× in BC150 Protein Storage Buffer (Supplementary Table 2). Load 1, 2, 5, 10, and 20 µL onto a 15% SDS–PAGE gel and visualize by Coomassie blue staining (Figure 1A).
11. Quantify band intensity using an image analysis software to determine relative signal intensity among lanes loaded with equal volumes.
12. Divide the intensity ratio by each protein’s molecular weight to obtain the molar ratio.
13. Use this molar ratio to correct the molar concentration estimated in Step 1.8.
14. Aliquot purified protein, snap-freeze in liquid nitrogen, and store at −80 °C. For short-term use, store at 4 °C overnight before proceeding to the next experiment.

2. Histone peptide array

1. Prepare peptide array blocking buffer by dissolving 5 g Bovine Serum Albumin (BSA) in BC150 Protein Storage Buffer to a final volume of 100 mL.
2. Optionally outline the reaction area on the peptide array using a hydrophobic marker pen to minimize reagent consumption.
3. Place several sheets of filter paper in a 15 cm petri dish and moisten them with water.
4. Position the peptide arrays on the moistened filter papers.
5. Carefully add 300 µL blocking buffer to cover each array area completely.
6. Cover the petri dish and incubate at room temperature for 1–3 h.
7. Prepare FLAG-SPIN4 in blocking buffer to a final concentration of 1 µM.
8. Prepare 1× FLAG peptide in blocking buffer to a final concentration of 1 µM.
9. To improve the signal-to-noise ratio and reduce undesired background, perform the following steps in a cold room. Move the blocked arrays (with the petri dish) to the cold room before adding proteins.
10. Remove excess blocking buffer from the arrays.
11. Carefully add the desired FLAG-SPIN4 protein or FLAG peptide as a control onto each block.
12. Cover the petri dish and incubate overnight in the cold room.
13. Add approximately 50 mL peptide array wash buffer to a Western blot wash box and bring it to the cold room.
14. Discard the protein solution and immediately transfer the arrays into the wash buffer, reaction side up.
15. Wash the arrays five times as follows: the first two washes for 1 min each, and the remaining three washes for 5 min each.
16. Rinse the arrays with peptide array blocking buffer to remove residual detergent, then return them to the Petri dish.
17. Prepare anti-FLAG antibody solution in 1:2000 dilution and add 300 µL per block.
18. Incubate at room temperature for 1 h.
19. Repeat Steps 2.14 to 2.16 to wash the arrays.
20. Prepare HRP-conjugated goat anti-mouse IgG (secondary antibody) solution in 1:6000 dilution and add 300 µL per block.
21. Incubate at room temperature for 1 h.
22. Repeat Steps 2.14 to 2.16 to wash the arrays.
23. Place the arrays inside a transparent sheet protector.
24. Add approximately 1 mL Enhanced Chemiluminescence (ECL) substrate, ensuring even coverage of the reaction surface.
25. Remove air bubbles and capture chemiluminescent and bright-field images using an appropriate imaging system (e.g., digital imagers or autoradiographic films) (Figure 1B).
26. Analyze signal intensity according to the peptide array software and associated online tools (Figure 2A–C).

3. Large-scale expression of FLAG-SPIN4 for pulldown assay

1. In a 500 mL Erlenmeyer flask, add 125 mL ESF921 medium, 1.8 × 10⁸ Sf9 cells, and 1.2 mL P1 baculovirus expressing FLAG-SPIN4 or FLAG-SPIN4^D82H.
2. Incubate at 27.5 °C in an orbital shaker incubator (125 rpm) for 72 h.
3. Aliquot the culture as follows:
   1. 4 mL of culture into a 15 mL tube for titration.
   2. 30 mL of culture into four 50 mL tubes.
4. Centrifuge the tubes at 600 × g for 15 min and discard the supernatant.
5. Snap-freeze the pellets in liquid nitrogen and store at −80 °C until use.

4. Titration of bead-to-lysate ratio for FLAG-SPIN4 and FLAG-SPIN4^D82H

1. Add 1 mL BC150 lysis buffer to each of the Sf9 pellets from 5 mL of Sf9 culture described in Section 3, Step 3.1.
2. Sonicate the lysates to ensure complete disruption: 30% amplitude, 5 s on/5 s off, for a total sonication time of 15 s.
3. Centrifuge at 20,000 × g, 4 °C, for 10 min, and transfer the supernatant to fresh 1.5 mL tubes.
4. Prepare 20 µL anti-FLAG affinity gel slurry for each pulldown reaction (60 µL for 3 titrations)
5. Wash the beads with BC150 lysis buffer:
   1. Add 1 mL BC150 lysis buffer and vortex briefly.
   2. Centrifuge at 3,000 × g for 1 min.
   3. Aspirate and discard the supernatant.
6. Resuspend the beads in 40 µL BC150 lysis buffer for every 1 pulldown reaction (120 µL for 3 titrations).
7. Combine the lysate in 1.5 mL tubes containing BC150 lysis buffer, and bead suspension (Supplementary Table 3)
8. Incubate the tubes on a tube rotator set to 15 rpm at 4 °C for 1.5 h.
9. Centrifuge at 3,000 × g for 1 min and aspirate the supernatant.
10. Wash the beads three times as follows:
    1. Resuspend beads in 500 µL BC150 lysis buffer and vortex briefly.
    2. Centrifuge at 3,000 × g for 1 min.
    3. Aspirate the supernatant.
11. Resuspend beads in 20 µL 1× Laemmli buffer and incubate at 95 °C for 5 min.
12. Centrifuge at 3,000 × g for 1 min.
13. Run the supernatant on a 15% SDS–PAGE gel and stain with Coomassie blue to estimate protein yield.
14. Based on the data, select a lysate-to-bead ratio that saturates the antibody binding sites on the beads. In the experiments, ≥300 µL of lysate per 20 µL of beads was sufficient to achieve full binding capacity.
    NOTE: “Saturation of antibody binding sites” is defined as the point at which increasing the lysate volume no longer results in greater recovery of the target protein.
    NOTE: At the same molarity, SPIN4^D82H exhibits a lower binding affinity to FLAG affinity gel than the wild-type counterpart.
    NOTE: In this experimental setting, 1 mL of lysate was derived from 6 million Sf9 cells (calculated as 1.5 × 10⁶ cells/mL × 4 mL), indicating that SPIN4 expressed from ~6 million Sf9 cells is sufficient to saturate 20 µL of anti-FLAG affinity gel slurry.

5. Large-scale preparation of anti-FLAG affinity gel–bound FLAG-SPIN4 as the bait

1. For each protein to be tested, collect two pellets from 30 mL cultures. Resuspend each pellet in 6 mL BC150 lysis buffer, transfer to a 15 mL centrifuge tube, and adjust the total volume to 13 mL with BC150 lysis buffer.
2. Sonicate each sample on ice at 100% amplitude, 15 s on/30 s off, for a total sonication time of 2 min.
3. Transfer the lysates to pre-chilled high-speed centrifuge tubes and centrifuge at 20,000 × g, 4 °C, for 15 min.
4. Add 200 × (number of samples + 1) × 1.1 µL of anti-FLAG affinity gel slurry to a 1.5 mL centrifuge tube.
5. Wash the beads with >5 column volumes (CV) of BC150 lysis buffer once.
6. Resuspend the beads in 400 × (number of samples + 1) × 1.1 µL of BC150 lysis buffer.
7. Transfer the supernatants obtained in Step 4 to 15 mL centrifuge tubes.
8. Add 500 µL of the Step 5.7 bead suspension to each tube.
9. In one tube, add 3 mL BC150 lysis buffer and 60 µL of 1× FLAG peptide (100 mM) as a negative control.
10. Incubate the samples on a tube rotator set to 120 rpm at 4 °C for 1.5 h.
11. Centrifuge at 2,000 × g for 5 min and discard the supernatant.
12. Resuspend the beads in an appropriate amount of supernatant and transfer to 1.5 mL tubes.
13. Centrifuge at 3,000 × g for 1 min and aspirate the supernatant.
14. Wash the beads five times with BC150 lysis buffer as follows:
    1. Resuspend the beads in 1 mL BC150 lysis buffer.
    2. For the 4th and 5th washes, place the samples on a tube rotator set to 20 rpm at 4 °C for 10 min.
    3. Centrifuge at 3,000 × g for 1 min and aspirate the supernatant.
15. Resuspend the beads in 200 µL BC150 lysis buffer.
16. Take 9 µL of the bead/buffer suspension and analyze protein content by 15% SDS–PAGE followed by Coomassie blue staining.
17. Because the protein input exceeds bead binding capacity, the amount of protein bound per bead volume should be similar across samples. Semi-quantify band intensity by digital image analysis and adjust bead amounts for subsequent reactions.

6. Preparation of Mono-nucleosomes from HEK293FT cells

1. Grow HEK293FT cells in 15 cm tissue culture dishes to approximately 90% confluency using 10% Fetal Bovine Serum (FBS) in DMEM media at 37 °C in a 5% CO₂ humidified incubator.
2. Early passage HEK293FT cells are obtained commercially and authenticated by STR profiling.
   NOTE: One 15 cm dish yields ~27 million cells, which is enough for 1–2 pulldown reactions. For six reactions (including negative controls), prepare at least three 15 cm dishes. If the target histone post-translational modification (PTM) is rare or unstable, increase cell number 2–10× accordingly.
3. (Optional) Depending on the histone modification of interest, add appropriate PTM inhibitors (Supplementary Table 2) to the culture medium and incubate for 24 h.
   NOTE: This step may be omitted if abundant histone PTMs are present.
4. When cells reach near 100% confluency, collect them as follows:
   1. Aspirate the culture medium.
   2. Rinse cells gently with 10 mL ice-cold PBS, then aspirate.
   3. Add 5 mL ice-cold PBS.
   4. Scrape cells on ice and transfer to a 50 mL tube (all dishes can be combined).
   5. Centrifuge at 13,000 × g, 4 °C, for 10 min.
   6. Discard the supernatant.
5. (Optional) Either proceed immediately or freeze the cell pellet for later use:
   1. Estimate the pellet volume.
   2. Resuspend in 9× pellet volume of Buffer G to reach a final glycerol concentration of 20%.
   3. Snap-freeze in liquid nitrogen and store at −80 °C.
6. If proceeding with frozen pellet, thaw the pellet on ice, centrifuge at 1,300 × g, 4 °C, for 5 min, and discard supernatant.
7. Resuspend the (fresh or thawed) pellet in 10× pellet volume of TMSD buffer (40 mM Tris-HCl, pH 7.5, 5 mM MgCl₂, 250 mM sucrose, 1 mM DTT) containing protease inhibitors using a serological pipette.
8. Prepare 10% NP-40 in TMSD buffer freshly, and add 0.11× total reaction volume to achieve a final NP-40 concentration of ~1%.
9. Incubate the suspension on a tube rotator set to 120 rpm at 4 °C for 5 min.
10. Mix 10 µL of the suspension with 10 µL trypan blue, and observe under a microscope to monitor nuclear staining. Ensure that nuclei are completely isolated from cytosol (Figure 3A).
11. When approximately 95% of nuclei are trypan blue–positive, proceed to the next step. For HEK293FT cells, 1% NP-40 incubation for 5 min is typically sufficient. For other cell types, verify and adjust incubation time.
12. Centrifuge at 1,300 × g, 4 °C, for 5 min, and discard the supernatant.
13. Resuspend the pellet in 10× pellet volume of TMSD buffer.
14. Repeat Steps 6.12 to 6.13 five times to completely remove residual NP-40.
15. Resuspend the pellet in 5× pellet volume of Buffer B.
16. Add 1 Unit micrococcal nuclease (MNase) per 10,000 nuclei, and incubate in a 37 °C water bath. Occasionally agitate the tube during the incubation to keep the nuclei evenly resuspended
17. At the indicated time points (0.5, 1, and 2.5 h), withdraw 50 µL of the reaction into 1.5 mL tubes, immediately add 1 µL of 0.5 M EDTA-Na (pH 8.0), and vortex.
18. Extract DNA from each time point following the procedure below:
    1. Add 80 µL H₂O, 30 µL 10% SDS, and 20 µL 5 M NaCl; vortex.
    2. Add 200 µL phenol/chloroform; vortex thoroughly.
    3. Centrifuge at 20,000 × g, RT, 3 min.
    4. Carefully transfer the supernatant to a new 1.5 mL tube.
    5. Add 1 mL 100% ethanol and 20 µL 3 M sodium acetate (pH 5.2); vortex.
    6. Centrifuge at 20,000 × g, RT, 5 min.
    7. Remove supernatant, add 1 mL 70% ethanol.
    8. Centrifuge at 20,000 × g, RT, 1 min.
    9. Remove the supernatant and dry the pellet at 55 °C with the cap open for 10 min.
    10. Resuspend the pellet in 50 µL TE buffer.
19. Run samples on a 1% TBE agarose gel to confirm mono-nucleosome DNA size (147 bp) (Figure 3B).
20. (Optional) Micrococcal Nuclease (MNase) digestion efficiencies vary from cell types, additional MNase may be added mid-reaction if necessary to completely remove poly-nucleosomes. For instance, if 1 h and 2 h samples appear similar, add more MNase at 3 h.
21. When a predominant ~147 bp band is observed, or when additional MNase/extended incubation no longer alters fragment size, stop the reaction by adding 0.5 M EDTA-Na (pH8.0) to a final concentration of 2 mM.
22. Centrifuge at 1,300 × g, 4 °C, for 5 min. Collect the supernatant (S1).
23. Resuspend the pellet in 10× pellet volume of TE buffer + protease inhibitors and incubate the sample on a tube rotator set to 20 rpm at 4 °C for at least 30 min.
    NOTE: Mono-nucleosome extraction efficiencies vary from cell types. 0-300 mM NaCl may be used to efficiently extract nucleosomes.
24. Centrifuge at 1,300 × g, 4 °C, for 5 min. Collect the supernatant (S2).
25. Transfer 50 µL each of S1 and S2 to separate 1.5 mL tubes.
26. Analyze all fractions by agarose gel electrophoresis (Steps 6.17 to 6.18) to assess DNA size distribution and extraction efficiency. Most mono-nucleosomes are found in the S2 fraction (Figure 3C).
27. Transfer S2 to high-speed centrifuge tubes and spin at 20,000 × g, 4 °C, for 10 min to remove insoluble material.
28. Collect the supernatant and dialyze against BC150 lysis buffer (without inhibitors):
    1. First dialysis: 1 L, 1 h.
    2. Second dialysis: 2 L, overnight.
29. After dialysis, centrifuge again at 20,000 × g, 4 °C, for 10 min to remove any remaining insoluble debris.
30. Collect the final supernatant.
31. Mix 5 µL of the supernatant with 35 µL BC150 lysis buffer and 10 µL 5× Laemmli buffer, heat at 95 °C for 5 min, and load 1, 2, 5, 10, and 20 µL on a 15% SDS–PAGE gel.
32. Load 1, 2, and 5 µL of 0.1 mg/mL BSA in Laemmli buffer as a reference for yield estimation.
33. Stain with Coomassie blue, and estimate nucleosome yield semi-quantitatively using an image analysis software (Figure 3D).

7. Modified nucleosome pulldown

1. Prepare 2× bead blocking buffer by dissolving 1 g BSA in BC150 lysis buffer to a final volume of 10 mL, then vortex until fully dissolved.
2. Transfer FLAG-SPIN4-WT and FLAG-SPIN4^D82H beads containing equivalent molar amounts of protein (from Section 5, Steps 5.16 to 5.17) into separate 1.5 mL tubes.
3. Add an equal volume of negative control beads (from Section 5, Step 10) to the FLAG-SPIN4-WT sample.
4. Adjust the total volume of all bead samples to be identical using BC150 lysis buffer.
5. Add an equal volume of 2× bead blocking buffer to each tube containing beads and BC150 lysis buffer.
6. Incubate the samples on a tube rotator set to 20 rpm at 4 °C overnight (O/N).
7. Thaw an appropriate amount of 1× nucleosome extracted from HEK293FT cells on ice. The required volume depends on the desired nucleosome-to-SPIN4 molar ratio. For initial optimization, use a 1:1 molar ratio of nucleosome to SPIN4.
8. Mix 1× nucleosome with an equal volume of 2× bead blocking buffer thoroughly.
9. Combine the mixture from Step 7.8 with the pre-blocked beads in a 15 mL tube.
10. Incubate the mixture on a tube rotator set to 20 rpm at 4 °C for 4 h or overnight.
11. Centrifuge at 3,000 × g, 4 °C, for 3 min.
12. Carefully discard the supernatant.
13. Resuspend the beads in 1 mL BC150 lysis buffer and transfer to 1.5 mL tubes.
14. Wash the beads three times as follows:
    1. Centrifuge at 3,000 × g, 1 min, and aspirate the supernatant.
    2. Resuspend the beads in 1 mL BC150 lysis buffer.
    3. Occasionally agitate the tube to keep the beads evenly resuspended during washing.
15. Wash the beads two additional times as follows:
    1. Centrifuge at 3,000 × g, 1 min, and aspirate the supernatant.
    2. Resuspend beads in 1 mL BC150 lysis buffer.
    3. Rotate at 20 rpm, 4 °C, for 10 min.
16. After washing, resuspend the beads in 9 column volumes (CV) of BC150 lysis buffer.
17. Take 160 µL from each sample, mix with 40 µL 5× Laemmli buffer, and heat at 95 °C for 5 min.
    ​NOTE: A FLAG peptide elution of the protein-nucleosome complex is recommended, yielding cleaner results with less background.
18. Centrifuge at 3,000 × g, room temperature, for 5 min.
19. Load 5 µL of the supernatant on a 15% SDS–PAGE gel and perform silver staining (Figure 4A).
20. Load another 5 µL of the supernatant on a 15% SDS–PAGE gel, transfer to a PVDF membrane, and probe with the appropriate primary and secondary antibodies specific for the target histone PTMs (Figure 4B).

Results

To investigate the binding substrates of human SPIN4, wild-type SPIN4 and a patient-derived mutant associated with overgrowth syndrome were reconstituted using a Sf9 insect cell expression system. Specifically, this is an aspartic acid (D) to histidine (H) missense mutation at residue 82 (D82H), which resides within the first Tudor domain of SPIN4, implying its role in regulating the substrate binding of SPIN4. A baculoviral vector encoding a single FLAG-tagged SPIN4 was used for protein expression in Sf9 cells, followed by a one-step FLAG affinity purification and 1× FLAG peptide competitive elution (Figure 1A). Equal molarities of purified SPIN4 proteins were then subjected to incubation with a Histone Peptide Array consisting of 384 unique histone PTM combinations. SPIN4 proteins bound to specific histone PTMs were detected by FLAG antibody and HRP-conjugated secondary antibody (Figure 1B). The chemiluminescence signals were quantified digitally. SPIN4 binds to H3K4me1/2/3 peptides, and such binding affinity is negatively impacted by other PTMs in-cis, such as H3R2me2a, H3R2me2s, and H3K9me3 (Figure 2A-C). In contrast to SPIN4, SPIN4^D82H exhibited a general reduction of binding to all SPIN4 targets, indicating that Tudor domain 1 is required for substrate binding and SPIN4^D82H is a loss-of-function mutant (Figure 2A-C).

As peptide-based binding assays often yield false-positive or non-specific binding results, SPIN4 binding to histone PTMs was further validated in their native state as mononucleosomes. Nuclei were extracted from HEK293FT cells (Figure 3A), and native chromatin was digested with a time course of MNase (Figure 3B). A short digestion yields mono-nucleosomes with some di- and tri-nucleosomes, while prolonged digestions generate mostly mono-nucleosomes despite a lower molecular weight, indicating the elimination of linker DNA and over-digestion (Figure 3B). The digested nucleosomes were not effectively released to the supernatant (S1) (Figure 3C). Therefore, mono-nucleosomes were further extracted from the digested pellet using Tris-EDTA (TE) buffer from the supernatant (S2) (Figure 3C). After dialysis and concentration, these native mono-nucleosomes were quantified by a serial titration on SDS-PAGE gels followed by Coomassie blue staining (Figure 3D).

Next, the molarities of reconstituted SPIN4 and extracted native mono-nucleosomes were calculated at a 1:1 molar ratio for pulldown experiments. FLAG-SPIN4 proteins were bound to FLAG agarose beads as bait. Following the pulldown, samples were run on SDS-PAGE and silver-stained. While histone proteins from SPIN4-bound mono-nucleosomes were detected, SPIN4^D82H exhibited an overall reduction of nucleosome binding, consistent with the observations using the peptide arrays (Figure 4A). H3K4me1/2/3 were further detected by western blot, comparing SPIN4 and SPIN4^D82H pulldown samples. Consistently, SPIN4^D82H exhibited no quantifiable binding to H3K4me1/2/3, validating these dual approaches for the rapid identification of histone PTMs as functionally relevant binding substrates of epigenetic readers (Figure 4B).

Figure 1: Reconstitution of human SPIN4 and SPIN4^D82H for histone peptide-based PTM screening. (A) Coomassie staining of reconstituted SPIN4 proteins. (B) Screening of binding substrates comparing SPIN4 and SPIN4^D82H using a Histone Peptide Array. Peptide PTM candidates exhibiting differential binding are indicated by the array serial numbers. Please click here to view a larger version of this figure.

Figure 2: Quantification of SPIN4 binding affinity to peptide PTMs. Chemiluminescent signals were quantified digitally. Comparisons of SPIN4 and SPIN4^D82H for binding to every (A) H3K4me1 (B) H3K4me2 and (C) H3K4me3 containing peptides are presented as bar graphs. Please click here to view a larger version of this figure.

Figure 3: Generation of mono-nucleosomes from native chromatin. (A) Single nuclei suspension extracted from HEK293FT cells. Scale bar, 250 µm (B) Time-resolved serial digestions of native chromatin from HEK293FT nuclei. DNA purified from digested chromatin indicates a ladder of mono-, di-, and tri-nucleosomes. (C) Post-MNase digestion fraction (S1),TE-buffer containing 300 mM NaCl extracted mono-nucleosomes (S2). (D) Coomassie blue staining of extracted mono-nucleosomes from the S2 fraction. Please click here to view a larger version of this figure.

Figure 4: Mono-nucleosome pulldown by SPIN4 and SPIN4^D82H and targeted validation of H3K4me1/2/3 containing mono-nucleosomes. (A) Silver stain, and (B) Western blot of H3K4me1/2/3, using SPIN4 pulldown, FLAG-eluted mono-nucleosome samples. Please click here to view a larger version of this figure.

Supplementary Table 1: Calculation of protein yield. SPIN4 protein yield was calculated at each purification step. Please click here to download this file.

Supplementary Table 2: Buffer Compositions. The composition for all the buffers used in this study is described.Please click here to download this file.

Supplementary Table 3: Protein pulldown calculation. SPIN4 lysate and FLAG beads ratio were calculated.Please click here to download this file.

Discussion

A time- and cost-effective method for biochemically pinpointing specific histone PTMs as targets of a poorly characterized protein, and for assessing how its disease-associated mutation affects the target's binding affinity, was presented. The advantage of this method is that it requires no expensive or time-consuming reconstitution of synthetic nucleosomes. This approach leveraged the convenience of commercialized histone PTM peptide arrays while ensuring that SPIN4 is a bona fide H3K4me reader by using native mononucleosomes as physiologically relevant binding substrates. A critical step of this method is the successful reconstitution of a protein of interest. It is empirical and can be challenging. In some cases, truncated versions of candidate proteins could be used to overcome technical hurdles. A limitation of this method is the use of native nucleosomes, which are not biochemically defined. However, following this method, mono-nucleosomes after pulldown can be further examined by mass spectrometry to gain a global view of binding specificity and dissect combinations of histone PTMs in cis, thereby overcoming this limitation⁸.

An interesting observation from the results is that H3R2 citrullination (H3R2citr) appears to permit SPIN4 binding to H3K4me1/2/3, whereas H3R2me2a/s blocks this interaction. H3R2me2a/s have been well documented to antagonize recognition of H3K4me by reader proteins, as either asymmetrical or symmetrical di-methylation at H3R2 structurally hinders the proximal H3K4me recognition⁹. On the other hand, H3R2me1 appears to be permissive for H3K4me recognition⁹. However, the exact function of H3R2citr has not been well characterized. As H3R2citr and H3R2 methylation are mutually exclusive, this result suggests H3R2citr may function to block H3R2 methylation and facilitate H3K4me recognition. Intriguingly, while SPIN4^D82H loses binding to most H3K4me1/2/3 containing peptides, it retains some interactions with H3R2citr-H3K4me2/3 peptides, implying that H3R2citr might alleviate the defects of SPIN4^D82H.

It was also found that H3K9me1/2 permits H3K4me1/2/3 recognition by SPIN4 in-cis, while H3K9me3 blocks such interaction. Previously, another Spindlin family member SPIN1, has been shown to bind the H3K4me3-H3K9me3 peptide¹⁰. H3K9me3 is a stringent marker for constitutive heterochromatin present at telomeric, centromeric, and retrotransposons across the genome, while H3K9me1/2 are much more pervasive¹¹. While SPIN1 knockout (KO) mice exhibit muscle defects and are perinatal lethal, SPIN4-KO mice are viable but exhibit bone overgrowth resembling human patients^1,12. Together, these results suggest a molecular distinction in substrate specificity underlying Spindlin family members may contribute to distinct physiological functions.

In addition to methylated histones, Tudor domains have also been shown to recognize methylated non-histone proteins³. This work exclusively focuses on histone PTMs and does not explore potential non-histone targets of SPIN4. Future investigations are warranted to examine the full scope of SPIN4-binding targets to better understand the underlying molecular mechanisms of the X-linked bone overgrowth syndrome caused by SPIN4 mutations.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
1.5 mL Microcentrifuge Tubes	Sarstedt	72.690.001	General experiments
15 cm Tissue Culture Dish	Sarstedt	83.3903.300	General experiments
15 mL Conical Centrifuge Tubes	Sarstedt	62.554.101	General experiments
50 mL Conical Centrifuge Tubes	Sarstedt	62.548.004	General experiments
500 mL Erlenmeyer Flask	Stellar Scientific	CT229809	Sf9 cell culture
Amersham ImageQuant 800 CCD Imager	Cytiva	IMQ800	CCD imaging system
Anti-FLAG M2 Affinity Gel	Sigma-Aldrich	A2220	Biochemical pulldown
Anti-FLAG M2 antibody	Sigma-Aldrich	F1804	Western blot
Aprotinin	MedChemExpress	HY-P0017	Protease inhibitor
Array Analyze Sofrware	Active Motif	https://www.activemotif.com/documents/Array_Analyze_Software_v16.1.zip	MODified peptide array data analysis
ASI Agarose LE	Alkali Scientific	A7700	DNA gel
Bac-to-Bac Baculovirus Expression System	Thermo Scientific	10359016	Baculoviral expression
Benzamidine hydrochloride	MedChemExpress	HY-W018781	Protease inhibitor
Bovine Serum Albumin (BSA)	Sigma-Aldrich	A9647	Protein quantification standard
Bromophenol Blue	Thermo Scientific	A18469.08	Lammeli buffer
Coomassie Brilliant Blue R-250	Thermo Scientific	20278	Protien gel staining
Dithiothreitol (DTT)	Sigma-Aldrich	D5545	Lysis buffer
EDTA	RPI research products	E58100	Lysis buffer
Enhanced Chemiluminescence (ECL) substrate	Cytiva	RPN2232	Western blot
ESF 921 Insect Cell Culture Medium	Expression Systems	96-001	Sf9 cell culture
Ethanol (200 proof)	Spectrum Chemical	ET107	DNA precipitation
EVOS M5000 Imaging System	Invitrogen	M5000	General imaging
FLAG peptide	MedChemExpress	HY-P7006	Pulldown elution
Glycerol	Sigma-Aldrich	G77893	Storage buffer
H3K4me1 antibody	Active Motif	39300	Western blot
H3K4me2 antibody	Cell Signaling	9725T	Western blot
H3K4me3 antibody	Cell Signaling	9751T	Western blot
HEK293FT cells	Thermo Scientific	R70007	Human embryonic kidney fibroblastcell line
HEPES	Sigma-Aldrich	H3375	Lysis buffer
ImageJ	National Institutes of Health	https://imagej.net/ij/download.html	Java-based image processing program
Leupeptin hemisulfate	MedChemExpress	HY-18234A	Protease inhibitor
Micrococcal Nuclease (MNase)	NEB	M0247S	Endonuclease for chromatin digestion
Model 120 Sonic Dismembrator	Fisher Scientific	FB120110	Sonication
MODified Histone Peptide Array	Active Motif	13005	Protein binding assay
Pepstatin A	MedChemExpress	HY-P0018	Protease inhibitor
Peroxidase IgG Fraction Monoclonal Goat Anti-Mouse IgG, light chain specific	Jackson ImmunoResearch	115-035-174	Western blot
pFastBac1-Flag-SPIN4	Thermo Scientific	10360014	Baculoviral expression vector
pFastBac1-Flag-SPIN4-D82H	Thermo Scientific	10360014	Baculoviral expression vector
Phenol:Chloroform:Isoamyl Alcohol (25:24:1)	Sigma-Aldrich	P2069	DNA purification
Pierce Rapid Gold BCA Protein Assay Kit	Thermo Scientific	PIA53226	Protein quantificaiton
Pierce Silver Stain Kit	Thermo Scientific	24612	Protein detection
PMSF Protease Inhibitor	Thermo Scientific	36978	Protease inhibitor
Polyacrylamide Gel	Made in-house	N/A	Protein gel
Sf9 insect cells	ATCC	CRL-1711	Insect cell line for baculovirus expression
Slide-A-Lyzer MINI Dialysis Device, 10K MWCO	Thermo Scientific	69572	Protein dialysis
Sodium acetate	Sigma-Aldrich	S8750	DNA precipitation
Sodium Chloride	Fischer	BP358-1	General buffers
Sodium Dodecyl Sulfate (SDS)	Sigma-Aldrich	L3771	General buffers
Trypan Blue Solution (0.4%)	Thermo Scientific	15250061	Nuclei staining