Resonance Assignments for the Tandem PWWP-ARID Domains of Human RBBP1

Biomolecular NMR Assignments (2019) 13:177–181 https://doi.org/10.1007/s12104-019-09873-2

ARTICLE

Resonance assignments for the tandem PWWP-ARID domains of human RBBP1

Weibin Gong1 · Xingzhe Yao2,3,5 · Qihui Liang1,5 · Yufeng Tong4 · Sarah Perrett1,5 · Yingang Feng2,3

Received: 5 December 2018 / Accepted: 16 January 2019 / Published online: 21 January 2019 © Springer Nature B.V. 2019

Abstract Retinoblastoma-binding protein 1 (RBBP1), also known as AT-rich interaction domain 4A (ARID4A), is a tumour suppressor involved in the regulation of the epigenetic programming in leukemia and Prader-Willi/Angelman syndromes. The ARID domain of RBBP1 binds to DNA non-specifically and has gene suppression activity. However, no structural data has been obtained for the human RBBP1 ARID domain so far. Here we report the near-complete 1H, 13C, 15N backbone and side-chain NMR assignment of a 27 kDa tandem PWWP-ARID domain construct that spans residues 171–414 with the removal of a short disordered region between the two domains. The predicted secondary structure based on the assigned chemical shifts is consistent with the structures of the isolated PWWP domain of human RBBP1 previously solved and the homologous ARID domains of other proteins.

Keywords RBBP1 · ARID domain · Tandem domains · NMR assignments · Secondary structure

Biological context regulation and epigenetic regulation functions that lead to cell cycle arrest and act as tumour suppressors (Binda et al. Retinoblastoma-binding protein 1 (RBBP1) and RBBP1-like 2006; Winter et al. 2012; Wu et al. 2008; Hurst et al. 2008). 1 (RBBP1L1) are components of the retinoblastoma (RB) RBBP1 and RBBP1L1 also regulate genomic imprinting transcriptional repressor complex and the Sin3-histone dea- epigenetically and are involved in genetic developmental cetylase (HDAC) complexes (Lai et al. 1999; Binda et al. disorders including Prader-Willi and Angelman syndromes 2006; Malovannaya et al. 2010). Both proteins show gene (Wu et al. 2006). Furthermore, RBBP1 and RBBP1L1 function as transcriptional coactivators for the androgen receptor (AR) and RB and play an integral part in the AR and RB * Sarah Perrett signalling pathways involved in the regulation of Sertoli cell [email protected] function and male fertility (Wu et al. 2013, 2017). * Yingang Feng RBBP1 specifically interacts with RB, exhibiting both [email protected] HDAC-dependent and -independent repression activities (Binda et al. 2006, 2008; Lai et al. 1999, 2001). Based on 1 National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules, Institute sequence homology to BRCA1, two regions, namely, R1 and of Biophysics, Chinese Academy of Sciences, R2 of RBBP1 were identified as responsible for the repres- Beijing 100101, China sion activities. The R1 region covers an AT-rich interaction 2 Shandong Provincial Key Laboratory of Synthetic Biology, domain (ARID), which binds to DNA nonspecifically (Pat- Qingdao Institute of Bioenergy and Bioprocess Technology, sialou et al. 2005) and is responsible for the HDAC-inde- Chinese Academy of Sciences, Qingdao 266101, China pendent repression activity. The R2 region (or R2 domain) 3 CAS Key Laboratory of Biofuels, Qingdao Institute located at the C-terminus of RBBP1 is responsible for the of Bioenergy and Bioprocess Technology, Chinese Academy HDAC-dependent repression activity of RBBP1 as R2 of Sciences, Qingdao 266101, China interacts with SAP30, a subunit of the mammalian Sin3A- 4 Department of Chemistry & Biochemistry, University HDAC co-repressor complex (Lai et al. 2001). RBBP1 and of Windsor, Windsor, ON N9B 3P4, Canada RBBP1L1 are also known as ARID4A and ARID4B, respec- 5 University of the Chinese Academy of Sciences, 19A Yuquan tively, due to the presence of an ARID domain. Besides the Road, Shijingshan District, Beijing 100049, China

Vol.:(0123456789)1 3 178 W. Gong et al.

ARID and R2 domains, the two proteins contain three Royal deletion and alteration of the disordered region between the family domains: Tudor, PWWP, and chromo barrel (Gong two domains. The assignments serve as a starting point for et al. 2012). We previously showed that only the chromo bar- structure determination and continued functional studies of rel domain can bind to histone tails and is thus responsible the RBBP1 ARID domain. for the epigenetic regulation function of RBBP1 (Gong et al. 2012). The Tudor domain of RBBP1 contains two interdigitated subdomains and only shows DNA binding capability (Gong et al. 2014). Methods and experiments Despite many functional studies of RBBP1 and RBBP1L1, no structural data is available for the ARID Protein expression and purification domains, which is crucial to fully understand the structural mechanism of its repression activity. After extensive trials, The DNA encoding a truncation mutant of the tandem we were unable to obtain a soluble construct of the isolated PWWP-ARID domains (corresponding to amino acid resi- ARID domain of RBBP1, but we obtained a soluble con- dues 171–414 of the full-length protein with the deletion struct containing both the PWWP domain and the ARID of residues 281–303 and mutation to lysine of Asp279, domain (Fig. 1a). The NMR spectral quality of the tandem Glu304 and Glu306 in the acidic linker region between the PWWP-ARID domain construct was improved after the two domains) was cloned into the pET-30a expression vec- removal of a disordered region (residues 281–303) between tor. The plasmid was used to transform Escherichia coli the two domains (Fig. 1b) and further improved after a triple BL21(DE3) cells. The transformed cells were first grown in residue mutation (Asp279Lys, Glu304Lys and Glu306Lys) 20 mL LB medium overnight and were then transferred to 15 (Fig. 1c), making it suitable for 3D heteronuclear NMR 500 mL M9 minimal medium containing N-NH4Cl (1 g/L) experiments. We did not observe significant backbone amide and 13C-glucose (4 g/L) as the sole nitrogen and carbon chemical shift changes of the dispersed peaks between the sources, respectively. The cells were then grown at 37 °C wild type and mutants, and so presumably the conformation till the optical density at 600 nm (A600) reached 0.7–0.8. of the final mutant used for resonance assignment repre- After induction with 0.7 mM IPTG, 15N- or 15N/13C-labelled sents that of the wild type protein. Here, we report the near recombinant proteins were over-expressed at 18 °C for 48 h. complete 1H, 13C and 15N chemical shift assignments of the The expressed proteins were first purified by Ni–NTA tandem PWWP-ARID domain of human RBBP1, which affinity column (Chelating Sepharose Fast Flow, GE Health- contains residue 171–414 of the full-length protein with the care), followed by gel filtration chromatography using a

Fig. 1 The 1H-15N HSQC spectra of the tandem PWWP-ARID c The triple residue mutant (Asp279Lys, Glu304Lys and Glu306Lys) domain constructs. a The wild-type tandem PWWP-ARID domain on top of the loop deletion. The spectra were recorded on a Bruker (residues 171–414). b The loop-deletion (residues 281–303) mutant. 600 MHz NMR spectrometer

1 3 179 Resonance assignments for the tandem PWWP-ARID domains of human RBBP1

Superdex 75 column (GE Healthcare). Over 20 mg purified aromatic side chain resonance assignments were obtained protein was obtained from 500 mL of M9 medium. from 2D 1H–13C HSQC and 3D 1H–13C NOESY–HSQC spectra (NOE mixing time of 150 ms) recorded for the aro- NMR spectroscopy matic region. All spectra were processed with NMRPipe (Delaglio et al. 1995) and analyzed with NMRViewJ (John- NMR samples consisted of 0.7 mM protein in 20 mM son and Blevins 1994). NaH 2PO4-Na2HPO4, 500 mM NaCl, 1 mM DTT, 1 mM EDTA and 10% (v/v) D 2O at pH 6.0. All NMR experiments Assignment and data deposition were performed at 302 K on a Bruker 600 MHz spectrometer or a Bruker 950 MHz spectrometer equipped with a The tandem PWWP-ARID domain construct contains 229 triple resonance CryoProbe. 1H–15N and 1H–13C HSQC, amino acids, including a C-terminal 6 × histidine tag. For HNCO, HN(CA)CO, HNCA, HN(CO)CA, HNCACB and the assignment, residues are numbered according to the CBCA(CO)NH spectra were recorded for backbone assign- sequence in the construct. The 1H-15N HSQC spectrum ments. The side chain assignments were based on C(CCO) of the tandem PWWP-ARID domain and the assignment NH, HBHA(CO)NH, HCCH–TOCSY, CCH–TOCSY, of the amide signals are shown (Fig. 2). All non-proline and 1H–15N and 1H–13C NOESY-HSQC experiments. The backbone amide proton and nitrogen signals of the tandem

Fig. 2 2D 1H–15N HSQC spectrum of the RBBP1 tandem PWWP- by a cross. The side chain NH peaks are labelled with a subscript ARID domain at 302 K. Assignments are indicated by the one-letter “sc”. The spectrum was recorded on a Bruker 950 MHz NMR spec- amino acid code and the sequence number. The backbone NH signal trometer of Y158 is under the plot contour level, and the position is indicated

1 3 180 W. Gong et al.

Fig. 3 a The secondary structure of the RBBP1 tandem PWWP-ARID domain predicted by the software TALOS-N. The predicted α-helix and β-sheet probabilities of each residue are plotted in black and red, respectively. The secondary structure elements are indicated on the top of the figure.b Sequence alignment of the three ARID domains of RBBP1, JARID1 and DRI. The sequence of RBBP1 ARID is coloured as magenta, together with predicted secondary structure elements. The secondary structure elements of DRI ARID are labelled underneath the aligned sequences

PWWP-ARID domain were assigned, except for those of shifts for aliphatic hydrogen atoms, such as H α of Arg119, Lys112, Arg155, Leu156, Ser186, Phe214–His220 and Hβ of Val11, Leu126, Trp172, and Lys193, and Hγ of His224–His229 (His6 tag). According to previous predic- Lys193. This is possibly due to the shielding effects of sur- tion, the missing of backbone signals of Phe214-His220 are rounding residues, which will be clarified by the structure possibly due to their disordered conformation (Gong et al. determination in future. 2012), while other missing backbone signals are possibly The secondary structure of the tandem PWWP-ARID due to the local conformation and dynamic properties of domain was predicted from the assigned chemical shifts these residues. Ignoring the C-terminal invisible segment using the software TALOS-N (Shen and Bax 2013). The and the non-native His tag, the assignment of the backbone results indicate that PWWP-ARID contains eight β-strands resonances of H, H N, C α and C β of residues Asn1–Gln213 and nine α helices (Fig. 3a). The predicted secondary struc- was completed to 98.2%. Assignment of the aliphatic and ture (β1–β6 and α1–α3) of the PWWP domain in the tan- aromatic side chains was achieved to 87.9% (side chain dem domains is consistent with the structure of the RBBP1 amino and guanidine group atoms of lysine and arginine PWWP domain (PDB ID: 2YRV) previously deposited residues, OH, SH, side chain 13C′, 13Cξ, and quaternary 13C by the RIKEN Structural Genomics/Proteomics Initia- were excluded). In detail, 98.0%, 98.6%, 96.7%, 98.0%, tive (RSGI). The ARID domain of RBBP1 has only ~ 30% 98.2%, 98.0%, and 96.1% of assignments were obtained for sequence identity with the ARID domains of JARID1 and backbone N, Cα, C′, HN, Hα, side chain C β, and H β, respec- Dead ringer protein (DRI), which recognize the CCGCCC tively. There are some unusual upfield-shifted chemical motif and AT-rich DNA sequence, respectively (Tu et al.

1 3 181 Resonance assignments for the tandem PWWP-ARID domains of human RBBP1

2008; Iwahara et al. 2002). The predicted secondary struc- AT-rich interaction domains (ARIDs) recognize DNA. EMBO ture elements (α4–α9 and β7–β8) of the RBBP1 ARID J 21:1197–1209 Johnson BA, Blevins RA (1994) NMRView: a computer program domain in the tandem domain construct are similar to those for visualization and analysis of NMR data. J Biomol NMR in the ARID domain of DRI, except that it lacks the first 4:603–614 and last α-helices (α1 and α8) in the DRI ARID domain, Lai A et al (1999a) RBP1 recruits both histone deacetylase-dependent and that the boundaries of α6 and α7 helices of the two and -independent repression activities to retinoblastoma family proteins. Mol Cell Biol 19:6632–6641 ARID domains are also significantly different (Fig. 3b). Like Lai A, Marcellus RC, Corbeil HB, Branton PE (1999b) RBP1 induces the RBBP1 ARID domain, the JARID1 ARID domain also growth arrest by repression of E2F-dependent transcription. lacks the two helices. The structure determination of the Oncogene 18:2091–2100 RBBP1 PWWP-ARID double domain based on the reported Lai A et al (2001) RBP1 recruits the mSIN3-histone deacetylase complex to the pocket of retinoblastoma tumor suppressor family pro- chemical shift assignments is underway, which will provide teins found in limited discrete regions of the nucleus at growth a more detailed structural basis for the differences in DNA arrest. Mol Cell Biol 21:2918–2932 recognition of these different ARID domains. All assigned Malovannaya A et al (2010) Streamlined analysis schema for high- chemical shifts were deposited in the BioMagResBank at throughput identification of endogenous protein complexes. Proc Natl Acad Sci USA 107:2431–2436 the University of Wisconsin-Madison under the Accession Patsialou A, Wilsker D, Moran E (2005) DNA-binding properties of Number 27677. ARID family proteins. Nucleic Acids Res 33:66–80 Shen Y, Bax A (2013) Protein backbone and sidechain torsion angles Acknowledgements This work was support by a National Key R&D predicted from NMR chemical shifts using artificial neural net- Program of China Grant 2017YFA0504000 to SP; National Natural works. J Biomol NMR 56:227–241 Science Foundation of China Grants 31470747 to WG, 31670735 Tu S et al (2008) The ARID domain of the H3K4 demethylase RBP2 and 31661143023 to YF; a Beijing Natural Science Founda- binds to a DNA CCGCCC motif. Nat Struct Mol Biol 15:419–421 tion Grant 5172026 to WG; and the CAS Center for Excellence in Winter SF, Lukes L, Walker RC, Welch DR, Hunter KW (2012) Allelic Biomacromolecules. variation and differential expression of the mSIN3A histone deacetylase complex gene Arid4b promote mammary tumor growth and metastasis. PLoS Genet 8:e1002735 Wu MY, Tsai TF, Beaudet AL (2006) Deficiency of Rbbp1/Arid4a References and Rbbp1l1/Arid4b alters epigenetic modifications and sup- presses an imprinting defect in the PWS/AS domain. Genes Dev Binda O, Roy JS, Branton PE (2006) RBP1 family proteins exhibit 20:2859–2870 SUMOylation-dependent transcriptional repression and induce Wu MY, Eldin KW, Beaudet AL (2008) Identification of chromatin cell growth inhibition reminiscent of senescence. Mol Cell Biol remodeling genes Arid4a and Arid4b as leukemia suppressor 26:1917–1931 genes. J Natl Cancer Inst 100:1247–1259 Binda O, Nassif C, Branton PE (2008) SIRT1 negatively regulates Wu RC, Jiang M, Beaudet AL, Wu MY (2013) ARID4A and ARID4B HDAC1-dependent transcriptional repression by the RBP1 family regulate male fertility, a functional link to the AR and RB path- of proteins. Oncogene 27:3384–3392 ways. Proc Natl Acad Sci USA 110:4616–4621 Delaglio F et al (1995) NMRPipe: a multidimensional spectral pro- Wu RC, Zeng Y, Chen YF, Lanz RB, Wu MY (2017) Temporal-spatial cessing system based on UNIX pipes. J Biomol NMR 6:277–293 establishment of initial niche for the primary spermatogonial stem Gong W et al (2012) Structural insight into recognition of methylated cell formation is determined by an ARID4B regulatory network. histone tails by retinoblastoma-binding protein 1. J Biol Chem Stem Cells 35:1554–1565 287:8531–8540 Gong W, Wang J, Perrett S, Feng Y (2014) Retinoblastoma-binding Publisher’s Note Springer Nature remains neutral with regard to protein 1 has an interdigitated double Tudor domain with DNA jurisdictional claims in published maps and institutional affiliations. binding activity. J Biol Chem 289:4882–4895 Hurst DR et al (2008) Alterations of BRMS1-ARID4A interaction modify gene expression but still suppress metastasis in human breast cancer cells. J Biol Chem 283:7438–7444 Iwahara J, Iwahara M, Daughdrill GW, Ford J, Clubb RT (2002) The structure of the Dead ringer-DNA complex reveals how

1 3