The homeodomain regulates stable DNA binding of prostate cancer target ONECUT2

Abstract

The CUT and homeodomain are ubiquitous DNA binding elements often tandemly arranged in multiple transcription factor families. However, how the CUT and homeodomain work concertedly to bind DNA remains unknown. Using ONECUT2, a driver and therapeutic target of advanced prostate cancer, we show that while the CUT initiates DNA binding, the homeodomain thermodynamically stabilizes the ONECUT2-DNA complex through allosteric modulation of CUT. We identify an arginine pair in the ONECUT family homeodomain that can adapt to DNA sequence variations. Base interactions by this ONECUT family-specific arginine pair as well as the evolutionarily conserved residues are critical for optimal DNA binding and ONECUT2 transcriptional activity in a prostate cancer model. The evolutionarily conserved base interactions additionally determine the ONECUT2-DNA binding energetics. These findings provide insights into the cooperative DNA binding by CUT-homeodomain proteins.

Fig. 1. Structure of the OC2-PEG10 complex.

a Overall structure of the OC2-PEG10 DNA complex. The position of CUT and HOX domains on DNA and their respective helices are labeled (α1- α8) while the unmodelled loop is depicted as a dashed line. CUT and HOX domains are shown in green and blue, respectively. b, c Arrangement of DNA interacting residues in α3 helix of CUT domain and α8 helix of HOX domain of OC2 are shown. d Schematic representation of the protein-DNA contacts in the complex. Hydrogen bonds are shown as dashed lines and water molecules are depicted as cyan spheres. DNA interacting residues of CUT and HOX domains are shown in green and blue, respectively.

Fig. 2. The RR motif in the HOX domain of OC2 shows unique DNA interactions.

a, b PEG10 and TTR DNA sequences and corresponding conserved base-specific interactions with OC2 and OC1, respectively. The core sequence is shown in blue except the bases at which PEG10 and TTR vary, that are in red. Black triangles depict interaction sites; the respective base-interacting residues are indicated. The difference in interaction of the first arginine (OC2 R479 or OC1 R438) of RR pair is shown with a black rectangular outline. c–f Interaction of OC2 S364, equivalent OC1 S323, and OC2 R479 and equivalent OC1 R438, to DNA. Hydrogen bonds in panels (c–f) are shown as yellow dashed lines. g Interaction of OC2 R480 and equivalent OC1 R439 (yellow and orange dashed lines, respectively). h Interactions of the POU residues corresponding to the OC arginine pair [PDB 1E3O [10.2210/pdb1E3O/pdb] (OCT1) and 1AU7 [10.2210/pdb1AU7/pdb] (PIT1)]. Hydrogen bonds are shown in the same color as respective proteins. i The relative orientations of the OC arginine pair and corresponding OCT1 and PIT1 residues. j, k Structure-based sequence alignment of CUT (j) and HOX (k) domains of OC2, OC1, OCT1, PIT1 and SATB1. The amino acid ranges are indicated. The conserved serine (S364) in CUT is highlighted (in green) while the arginine pair (RR motif; R479/R480) in HOX are shown with a red border. The conserved glutamine (Q365) in CUT and asparagine (N476) in HOX are also highlighted (in green). The color of highlighted residues is based on Clustal scheme.

Fig. 3. The HOX domain drives stable association of OC2 to DNA.

a–c ITC binding analysis of intact OC2, CUT domain and CUT + HOX, to PEG10 DNA. The raw heats (differential power, DP; top) and binding isotherms (bottom) are shown and representative of three independent experiments (n = 3; technical replicates). Source data are provided as a Source Data file. d–g Amide hydrogen-deuterium exchange mass spectrometry (HDX-MS) characterization of DNA binding by OC2. Structure of OC2 bound to PEG10 DNA showing the helix α3 (purple) which appears to become much more structured upon PEG10 binding (d). The CUT and HOX are in green and gold, respectively, while the PEG10 is in wheat. The protein N- and C-termini are indicated. Relative deuterium uptake into OC2 with and without bound PEG10 is rendered on the structure using a blue-white-red gradient scale (e); regions of the protein not covered by the HDX-MS experiment are in black. No regions of increased uptake were observed but many parts of the protein experienced decreased (blue) deuterium uptake upon PEG10 binding with the maximum decrease being 0.28. Relative deuterium uptake into the CUT domain alone with and without bound PEG10 is plotted on the structure according to the same scale and color scheme as in panel e (f). There were no significant differences in deuterium uptake upon PEG10 binding of the CUT domain alone (consequently, mostly white except the black regions that showed no coverage). Note that the CUT(-DNA) structure is extracted from the OC2-PEG10 structure for demonstration only. Deuterium uptake plot for the helix α3 containing peptide (residues 358-371) in the four different conditions as indicated (g). The y-axis corresponds to the total number of amides in the peptide. The HDX-MS data shown is representative of three independent experiments (n = 3; technical replicates); P-values based on One-way ANOVA as provided by the software were 0.001 for OC2-PEG10 vs apo-OC2; 0.9 for CUT-PEG10 vs apo-CUT.

Fig. 4. DNA binding thermodynamics of OC2 base-specific mutants.

a–c ITC binding analysis of OC2SQ, OC2N and OC2RR, to PEG10 DNA. The raw heats (differential power, DP) for each injection are shown on top and binding isotherms are shown in the bottom. The data shown is representative of three independent experiments (n = 3; technical replicates). Source data are provided as a Source Data file.

Fig. 5. Base-specific interactions by OC2 are needed for DNA binding cooperativity and are functionally relevant in a PC model.

a–d Kinetics of PEG10 DNA binding by OC2, OC2SQ, OC2N and OC2RR. All proteins are titrated at concentrations 75, 125, 175, 225 and 275 nM, and representative curves are shown (n = 3; technical replicates). The association and dissociation phases are separated by a dotted line. e Plot showing the proliferation of LNCaP cells upon stable expression of ectopic wild-type and mutant OC2 compared to cells with endogenous OC2 (vector control) at 48 and 72 h. Data are presented as mean values ± SD. Two-sample t-test was used for statistical analysis (n = 4; biological replicates). At 48 h, ****P(OC2) < 0.0001, P(OC2SQ) = 0.64, P(OC2N) = 0.67, P(OC2RR) = 0.26, and at 72 h, ***P(OC2) = 0.00011, P(OC2SQ) = 0.09, P(OC2N) = 0.31, P(OC2RR) = 0.33. f-g Relative mRNA levels of AR target genes KLK3, NKX3-1 and TMPRSS2 and NEPC marker genes NSE, PEG10 and SYP, upon stable overexpression of ectopic wild-type and mutant OC2 compared to cells with endogenous OC2 (vector control). Data are presented as mean values ± SD. Two-sample t-test was used for statistical analysis (n = 3; biological replicates). In case of AR targets, for KLK3, **P(OC2) = 0.008, P(OC2SQ) = 0.53, **P(OC2N) = 0.009, **P(OC2RR) = 0.007; for NKX3-1, ***P(OC2) = 0.0007, **P(OC2SQ) = 0.008, P(OC2N) = 0.19, P(OC2RR) = 0.62; and for TMPRSS2, ****P(OC2) < 0.0001, ****P(OC2SQ) < 0.0001, P(OC2N) = 0.93, P(OC2RR) = 0.16. In case of NE markers, for NSE, *P(OC2) = 0.02, *P(OC2SQ) = 0.02, P(OC2N) = 0.45, *P(OC2RR) = 0.04; for PEG10, **P(OC2) = 0.003, *P(OC2SQ) = 0.01, P(OC2N) = 0.14, ***P(OC2RR) = 0.0008; and for SYP, *P(OC2) = 0.01, P(OC2SQ) = 0.09, P(OC2N) = 0.2, P(OC2RR) = 0.09. Source data are provided as a Source Data file.