Flexibility of flanking DNA is a key determinant of transcription factor affinity for the core motif

Abstract

Selective gene regulation is mediated by recognition of specific DNA sequences by transcription factors (TFs). The extremely challenging task of searching out specific cognate DNA binding sites among several million putative sites within the eukaryotic genome is achieved by complex molecular recognition mechanisms. Elements of this recognition code include the core binding sequence, the flanking sequence context, and the shape and conformational flexibility of the composite binding site. To unravel the extent to which DNA flexibility modulates TF binding, in this study, we employed experimentally guided molecular dynamics simulations of ternary complex of closely related Hox heterodimers Exd-Ubx and Exd-Scr with DNA. Results demonstrate that flexibility signatures embedded in the flanking sequences impact TF binding at the cognate binding site. A DNA sequence has intrinsic shape and flexibility features. While shape features are localized, our analyses reveal that flexibility features of the flanking sequences percolate several basepairs and allosterically modulate TF binding at the core. We also show that lack of flexibility in the motif context can render the cognate site resistant to protein-induced shape changes and subsequently lower TF binding affinity. Overall, this study suggests that flexibility-guided DNA shape, and not merely the static shape, is a key unexplored component of the complex DNA-TF recognition code.

Figure 1

The conserved Exd-Ubx binding interface. (a) Recognition and binding of DNA consensus motif 5′-TGATTTAT-3′ by Exd (green)-Ubx (ice blue) heterodimer. The recognition helices of Exd and Ubx (indicated by ^∗) bind to opposite major groove faces of the DNA consensus site. The Ubx NTA binds to the minor groove of the consensus site. The crystal structure of Exd-DNA-Ubx was obtained from the PDB (1b8i). Missing regions critical for DNA-protein and protein-protein interactions were modeled (highlighted in opaque shade). DNA core motif (TGATTTAT) is in off-white and the 4-mer flanks in gray. (b) Interfacing residues between Exd and Ubx are depicted in surface mode and their (c) percentage occurrence during the simulation period is depicted using the Exd-HH-Ubx complex as a representative example. All interactions were calculated for the last 400 ns of the simulation time. CTE, C-terminal extension; H1/2/3, homeodomain 1/2/3; TALE, three-amino acid loop extension; HX, hexapeptide; NTA, N-terminal arm. To see this figure in color, go online.

Figure 2

The flexible Ubx-DNA binding interface. Time evolution of binding of the Ubx N-terminal arm (NTA) with the cognate DNA motif (TGATTTAT) in all six simulated systems. The NTA is color coded based on the simulation time. To illustrate NTA:DNA binding dynamics, snapshots of MD runs were saved every 10 ns between 100 and 500 ns. The 20-mer DNA sequences used for simulation in complex with Exd-Ubx homeodomain are listed (see Table 1 for details). To see this figure in color, go online.

Figure 3

Persistence of H-bond between Arg5 (R5) of Ubx and the DNA cognate sequence (T₅G₆A₇T₈T₉T₁₀A₁₁T₁₂) during the entire simulation period in all simulated systems. The guanidino group of Arg5 can H-bond with T-O2 (black) and with A-N3 (green). Simultaneous occurrence of black and green dots indicates formation of R5-mediated H-bond bridge between T-O2 and A-N3. H-bonds between the cognate site and other arginines (R2 and R10) of the N-terminal arm are indicated in brown, as described in the schematic. An N-H-O/N angle cutoff of 120° and an N-H … O/N distance cutoff of 3.5Å was used to capture the dynamic R5:DNA interaction. Bases in the antisense strand are indicated by (‘). To see this figure in color, go online.

Figure 4

2D shape space explored by DNA in apo and complexed forms. (a) Structural sampling by the cognate site of apo-HH DNA (CCGAT₅G₆A₇T₈T₉T₁₀A₁₁T₁₂GGCC) shown for selected basepairs steps. Fraction of occurrence of a particular structural variant is determined from the entire simulation period of 200 ns for apo DNA. The dotted line is used to guide the eye to perceive the minor groove width profile of the consensus site. Time evolution of shape transition of the (b) HH and (c) HL DNA in complex with Exd-Ubx heterodimer for basepair steps that are involved in binding of Ubx R5 residue and their immediate neighbors (T₅G₆A₇T₈T₉T₁₀A₁₁T₁₂). The basepair step in question is highlighted in boldface/underline. To see this figure in color, go online.

Figure 5

Time evolution of structural transition of the (a) LH1 and (b) LL1 DNA in complex with Exd-Ubx heterodimer for basepair steps that are involved in binding of Ubx R5 residue and their immediate neighbors (T₅G₆A₇T₈T₉T₁₀A₁₁T₁₂). The basepair step in question is highlighted in boldface/underline. To see this figure in color, go online.

Figure 6

Sequence of events leading up to the formation of the canonical Arg5:DNA H-bond in the HL system (CCGAT₅G₆A₇T₈T₉T₁₀A₁₁T₁₂AAAA). The TA step in HL is held in an unfavorable conformation until (a) the disruption of a highly conserved H-bond between A11 of the consensus motif (TATAAAAA) and Asn51 of the Ubx α3 recognition helix (inset) at ∼370 ns (b) permits the TA step to freely assume its preferred large positive roll conformation, (c) leading to an increase in interphosphate distance on the major groove edge and finally enabling the neighboring TT steps to rapidly transform to a narrow minor groove width conformation to (d) form the canonical H-bond with Arg5. (a) and (b) are truncated at 400 ns since no changes were observed in these parameters beyond this time point. The interphosphate distance was calculated using the NUPARM suite (64). All other calculations were carried out using cpptraj v18.01 (66). To see this figure in color, go online.

Figure 7

Water-bridge dynamics as a measure of DNA flexibility. Persistence of water-mediated cross-strand H-bonds bridging (a–c) T8:T7ʹ and (d–f) T9:A8ʹ bases in the consensus site (5′-T₅G₆A₇T₈T₉T₁₀A₁₁T₁₂-3′) of HH, LH1, and LL1 sequences. “+” indicates the presence of a water-mediated H-bond bridge, i.e., a single water molecule simultaneously H-bonded to T8-O2 and T7′-O2 atoms (T8:T7ʹ bridge) or T9-O2 and A8′-N3 atoms (T9:A8ʹ). The presence of contiguous patches of + indicate highly persistent H-bond bridges (indicated by ^∗). Discrete + symbols indicate H-bond bridges that are highly transient. Persistence of H-bond bridges increased with increasing number of A tracts (HH < LH1 < LL1) with lifetimes extending up to several tens of nanoseconds in the A tract containing sequences (LH1 and LL1). A schematic of the DNA duplex with cross-strand H-bond bridges (red lines) is given for reference. A stringent angle cut-off of 135° and distance cut-off of 3.0 Å between heavy atoms was used to define H-bonds. The H-bond bridges zip up the minor groove of LL1 DNA in a cooperative fashion. (g) Cross-correlation matrix of the step-wise minor groove width of the LL1-DNA shows positive correlation between the minor groove width of the central A tract (A9ʹ) and the 5′ A tract flank (A3). To see this figure in color, go online.

Figure 8

Flanking DNA sequence diminishes subtle shape distinctions between similar cognate sites. (a) Structural sampling by the cognate site of Scr LL1 DNA in the apo form shows preference for characteristic minor groove width profile with groove constriction at two bp steps (AAAAA₅G₆A₇T₈T₉A₁₀A₁₁T₁₂AAAA). (b) The same shape profile is lost in the Exd-Scr-DNA complex, due to which the conserved His-12 residue of Scr is ejected from the groove within 25 ns of the simulation (inset). Conserved R3 and R5 residues continue to persist in H-bonding with respective bp steps (insets to bp step panels). In the complexed form with Exd-Scr, the cognate site instead exhibits stark resemblance with the groove width profile exhibited by (c) the DNA binding motif in the Exd-Ubx-DNA complex. An N-H-O/N angle cutoff of 120° and an N-H … O/N distance cutoff of 3.5 Å was used to capture the dynamic Arg (R3/R5):Thy-O₂ interactions. Water-mediated interactions between His-12 and DNA were calculated using an angle cutoff of 135° and a distance cutoff of 3.5 Å (see Fig. 4 for all other details). To see this figure in color, go online.