A Potent Conformation-Constrained Synthetic Peptide

Letters Cite This: ACS Chem. Biol. 2018, 13, 2003−2009

A Potent Conformation-Constrained Synthetic Peptide Mimic of a Homeodomain Selectively Regulates Target Genes in Cells Basusree Ghosh,† Liberalis Debraj Boila,‡ Susobhan Choudhury,§ Priya Mondal,† Sayan Bhattacharjee,∥ Samir Kumar Pal,§ Amitava Sengupta,‡ and Siddhartha Roy*,† †

Department of Biophysics, Bose Institute, P-1/12 CIT Scheme VIIM, Kolkata 700054, India Stem Cell & Leukemia Lab, Cancer Biology & Inflammatory Disorder Division, Translational Research Unit of Excellence (TRUE), Indian Institute of Chemical Biology, CN-6, Sector V, Salt Lake, Kolkata 700 091, India § Department of Chemical, Biological & Macromolecular Sciences, S. N. Bose National Centre for Basic Sciences, Block JD, Sector III, Salt Lake, Kolkata 700 098, India ∥ Division of Structural Biology and Bioinformatics, CSIR-Indian Institute of Chemical Biology, 4 Raja S.C. Mullick Road, Kolkata 700 032, India

Downloaded via Basusree Ghosh on November 4, 2018 at 14:45:28 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

‡

S Supporting Information *

ABSTRACT: DNA, as a target for therapeutic intervention, remains largely unexplored. DLX-4, a homeodomain containing transcription factor, and its spliced isoforms play crucial roles in many aspects of cellular biochemistry and important roles in many diseases. A smaller peptide mimicking the homeodomain of the transcription factor DLX4 was designed and synthesized by suitable conjoining of its modified DNA-binding elements. The peptide binds to DLX-4 target sites on the regulatory region of the globin gene cluster with native-like affinity and specificity in vitro. When conjugated to cell penetrating and nuclear localization sequences, it upregulated some of the genes repressed by DLX-4 or its isoforms, such as β- and γ-globin genes in erythropoietin-induced differentiating CD34+ human hematopoietic stem/progenitor cells with high specificity by competing with the respective binding sites. Engineered peptides mimicking DNA-binding domains of transcription factors offer the potential for creating synthetic molecules for directly targeting DNA sites with high specificity.

N

globin gene repression.4 Because of the involvement of DLX-4 or its isoforms in many physiological processes and pathological conditions, we have focused our attention toward the construction of a peptide-based synthetic transcription factor that will mimic the DNA-binding homeodomain part of DLX-4 (DLX4-HD), which is common to all its spliced isoforms. Such a construct, lacking other functional modules of DLX-4 or its isoforms, may bind to targeted DNA sites and antagonize their functions, thus counteracting their gene regulation activity. In this article, we report design and synthesis of a peptide encompassing the recognition helix and the minor groove binding arm of DLX4-HD using a designed crossover loop, with the imposition of conformational constraints on the helix through the incorporation of helix promoting amino acid residues. We go on to show a high binding specificity of this synthetic molecule toward BP1 target sequences, in vitro and ex vivo.

ature has extensively used the major groove for readout of information present in the DNA sequences. Many naturally occurring transcription factors, which contain such motifs as the helix-turn-helix, the helix−loop−helix, the zinc finger and the homeodomain, use α-helices to interact with the major groove of the DNA. However, in a large number of naturally occurring transcription factors, multiple DNA binding motifs have been combined to create DNArecognition modules. Homeodomains (HD), the signature modular construct of HOX proteins, are known for their unique structure and binding pattern.1,2 DLX-4 (the DLX group of transcription factors is a class within the ANTP Homeobox gene family) or its isoforms, Beta Protein 1 (BP1) and DLX-7, play crucial roles in oncogenesis. In normal tissues, the DLX-4 isoform BP1 commonly functions as an adult βglobin gene repressor in embryonic and fetal erythroid lineages but rarely expressed in most other normal adult tissues.3−5 BP1 acts as a repressor of adult β- and δ-globin genes and, so far, this function is the best characterized.5,6 This apparent absence of expression is not due to gene deletion or translocation, but rather is mostly due to BP1-induced β© 2018 American Chemical Society

Received: May 27, 2018 Accepted: July 2, 2018 Published: July 2, 2018 2003

DOI: 10.1021/acschembio.8b00488 ACS Chem. Biol. 2018, 13, 2003−2009

Letters

ACS Chemical Biology

Figure 1. Schematic representation of the peptide design: (a) HD region of the DLX5 protein. Blue and green colors demarcate the binding and the nonbinding regions of the protein, respectively. W48 in the helix and Y8 in the arm (in space-filling models) interact with each other, both in the free and the DNA-bound state. (b) A stick model of the noninteracting amino acids that were substituted with AiB in synBP1B (also highlighted with magenta color in panel (c)). The green ribbonlike band is the linker that connects the helix and the arm (c) Sequence alignment of BP1-HD and DLX5-HD and design of synBP1B. (d) A model of synBP1B peptide bound to DNA. The linker (shown in green) was grafted to span the distance between the C-terminal end of the arm and the N-terminal end of the recognition helix.

■

RESULTS AND DISCUSSION Design of the Synthetic Transcription Factor. The homeodomain consists of an N-terminal arm and three alpha helices. The arm and the third helix of the HD, dubbed the recognition helix, bind to the DNA minor and the major groove, respectively, thus imparting a major part of its DNA binding affinity and specificity (see Figure 1a).7 The third helix inserts into the major groove of the DNA. The N-terminal arm has a flexible conformation and binds to the minor groove in a shape-dependent manner.8,9 Although the helix by itself was shown to be a good DNA-binder in a homeodomain protein,10 given the well-known role of the N-terminal arm in homeodomain proteins, it is very likely that the presence of both the N-terminal arm and the third helix is important for re-

engineering a proper synthetic transcription factor that would be capable of functioning properly inside the cell. Thus, we have attempted to combine the recognition helix with the Nterminal arm in an appropriate manner to create a synthetic transcription factor capable of functioning effectively in vivo. The importance of imposing conformational constraints on the helical part of this type of peptides has been previously reported.11 We impose conformational constraints on the helical part by incorporating α-aminoisobutyric acid (AiB) at suitable positions.12,13 The AiB incorporation sites were chosen using the crystal structure of DLX5, the only available structure of a DLX family member whose HD crystal structure in complex with the target DNA is known (PDB 2DJN, 4RDU). AiB substitution points were chosen on the non2004


Letters


Figure 2. (a) Far UV CD spectra showing secondary structure of synBP1 and synBP1B; (b) binding isotherm of BGS1 with BP1-HD; (c) binding isotherm of BGS1 with synBP1B; (d) binding isotherm of BGS1 with synBP1Bm; (e) binding isotherm of BGS2 with synBP1. All the isotherms were determined using fluorescence anisotropy. Each set (average of five independent experiments) was performed at 4 °C with 5 nM initial concentration of BGS1 or BGS2. Error bars indicate the standard error of the mean. Note that, at the concentrations where the synBP1B binds to BGS2, synBP1 does not show any significant binding.

points of the loop, such that the distance between those points was ∼12.6 Å (see Figure S1b in the Supporting Information). The designed peptide preserves the DLX4-HD protein’s DNA binding regions while eliminating two additional helices (Figure S1c in the Supporting Information), thus making the construct small enough to be chemically synthesizable in a straightforward manner. Secondary Structure of the Peptide and Binding Studies. All designed peptides were synthesized by solidphase peptide synthesis and purified by HPLC (see Table S1 in the Supporting Information). The AiB-substituted synthetic transcription factor will be referred to as synBP1B. Another peptide was synthesized that lacked the AiB substitutions (had wild-type amino acids at the substituted positions) but has an otherwise identical primary sequence (synBP1). (See Table S1 for the peptide sequences.) Secondary structural preferences of

DNA-interacting face of the recognition helix to constrain the peptide sequence to a helical conformation (see Figures 1b and 1c).12 To suitably connect the DNA major groove binding helix, the third helix, and the minor groove binding N-terminal arm, a peptide sequence that can crossover properly from the major groove to the minor groove was needed (see Figure 1d). Upon searching through the Protein Database, it was observed that Serum Response Factor (SRF) (PDB 1SRS) has a natural loop sequence (Figure S1a in the Supporting Information, Residues 146−152 (IKMEFID)) that connects the DNA binding arm and a major groove binding helix14 and has a distance of ∼12.6 Å between the N- and C-termini. To bring this loop into the context of this homeodomain mimicking synthetic transcription factor, we chose suitable amino acids in the DLX5 structure (one on the third helix and one on the N-terminal arm), which will be the connecting 2005


Letters


to the target sites.16,18 The conserved tryptophan and the tyrosine/phenylalanine in the arm serve as important components of the hydrophobic core of the homeodomains. The interaction between the tryptophan indole ring NH and π electron density of tyrosine through electron transfer may be important for imparting stability to this hydrophobic core and also the reason behind the signature fluorescence quenching of the tryptophan in free homeodomains.18,19 Therefore, the maintenance of a proper spatial relationship of the arm and the recognition helix in the free and DNA-bound state of DLX4HD may be attributed to the direct interaction between Y8 and W48. Thus, a quenching of the tryptophan may be taken as a surrogate signature for the proper recognition mode of the specific DNA sequence. We have thus explored the effect of binding of synBP1B to BGS2 on the tryptophan fluorescence of BP1-HD. The inset of Figure 3a shows the decay of fluorescence of W48 in free BP1-

synBP1B and synBP1 were explored using circular dichroism spectroscopy. synBP1 shows a broad minimum at ∼216 nm, characteristic of β-strands. In contrast, an α-helical conformational preference was observed for synBP1B, as indicated by the presence of clear negative peaks at ∼204 and 225 nm in the far-UV CD spectra, a pattern similar to that of the α-helical spectra. To buttress our arguments, we also present the NH− NH region NOE spectra of synBP1B, which shows clear N, N +1 connectivity in the region of the recognition helix, characteristics of α-helices (Figure S2 in the Supporting Information).15 Although the far-UV CD spectra of synBP1B does not exactly correspond to that of a classical α-helix and the MRE values are lower than a fully formed α-helix of all Lamino acids, it may not be fully unexpected, because of the presence of four achiral AiB residues. Taken together with the NMR data, it suggests a shift in conformational preference from the β-strand-like conformation to that of a helical conformation (see Figure 2a). BP1 binds to two sites upstream of the β-globin promoter region at −550/−527 (silencer 1; henceforth called BGS1) and −302/−294 (silencer 2; henceforth called BGS2)6 in K562 cell (human erythroleukemia cell line). We have explored binding of synBP1B in vitro to both of these regions by fluorescence anisotropy. For BP1-HD, a somewhat higher binding affinity for BGS2 (see Figure S3a in the Supporting Information; Kd = 11.2 ± 2.49 nM) was observed than for BGS1 (see Figure 2b; Kd = 33.8 ± 4.63 nM). The values are similar to those determined for the entire protein.5 The corresponding values of dissociation constants for synBP1B were 31.7 ± 5.75 nM for BGS1 and 23.6 ± 6.81 nM for BGS2 (see Figure 2c, as well as Figure S3b and Table S2 in the Supporting Information). These values are comparable to that of the BP1-HD under identical conditions. The in vitro specificity of the peptide for the target sequence was judged through two different sets of binding experiments, using mutant oligonucleotides and mutant peptides. Generally, HDs bind to a core AT-rich tetramer. Mutant oligonucleotides were designed by substituting two base pairs of the DNA core sequence, TATT to TGCT (BGS1 to BGS1M) and CAAT to CGCT (BGS2 to BGS2M)16,17 (see Table S3 in the Supporting Information). synBP1B binds to both BGS1M and BGS2M with significantly weaker affinity than the wildtype sequence (see Figures S2c and S2d and Table S2). A mutant peptide (synBP1Bm) was also synthesized by substituting a few critical amino acid residues at the putative protein−DNA interface with alanine (Table S1). synBP1Bm binds to both BGS1 and BGS2 (Figure 2d and Table S2) oligonucleotides with almost 7-fold lower affinity than that of the synBP1B (Table S2). Apart from the sequence mutant synBP1Bm, we have used synBP1 (lacking AiB residues) as a structural mutant (lacking a preformed α-helix). synBP1 showed negligible binding affinity toward BGS2, suggesting that the α-helical structure of the recognition helix is crucial for higher affinity binding (Figure 2e and Table S2). DNA Binding Mode of BP1-HD and synBP1B. Whether this stripped-down version of the homeodomain interacts with its target DNA site in a manner similar to the parent homeodomain was addressed by time-resolved fluorescence and NMR spectroscopy. The binding mode of homeodomains of different DNA-binding proteins to its target DNA appears to be structurally conserved. The conserved tryptophan in the third helix of the homeodomain (the recognition helix) is crucial for the structural stability of the domain and its binding

Figure 3. (a) Picosecond-resolved fluorescence transients of Trp residue in free synBP1B peptide and BGS2-synBP1B complex. Inset shows time-resolved transients of free BP1-HD protein. (b) Chemical shift difference of αH protons between BGS2 bound synBP1B (1:10 mol/mol) and free synBP1B. (c) synBP1B sequence with numbering used in this study.

HD. The observed fluorescence lifetime is short compared to tryptophan lifetimes seen in most proteins (see Figure 3a, inset). It is likely that this short lifetime is due to the decay of the excited state through nonradiative processes, presumably due to the rapid electron transfer from W to Y, as mentioned previously. synBP1B, on the other hand, shows a much longer fluorescence lifetime (see Figure 3a and Table S4 in the Supporting Information). This is likely due to synBP1B not having a BP1-HD-ike tertiary structure and Y8 not being in 2006


Letters


Figure 4. (a) Schematic representation of chromatin affinity precipitation assay. Stepwise isolation of peptide bound cross-linked DNA with Streptavidine Agarose (SA) beads. (b) Primer selection covering two known promoter binding sites of BP1 protein in the globin LCR. Red letters show the sequence of two primers and gray highlight indicates β-globin silencer1 and silencer2 regions upstream of the β-globin promoter region. (c) Extent of BIO-synBP1B-dR6 binding upon pull-down and amplification of the pulled-down DNA using primers described above. Lane A is untreated K562 cell chromatin fragment, Lane B is BIO-synBP1B-dR6 treated fragment after 12 h, Lane C is the same peptide as that decribed for Lane B after 24 h, Lane D is a synBP1B-dR6 (without Biotin)-treated fragment after 12 h, Lane E is a BIO-synBP1-dR6-treated fragment after 12 h. (d−f) RTPCR analysis of mRNA expression levels of β-globin (panel (d)), γ-globin (panel (e)), and δ-globin genes (panel (f)) in CD34+ HSPCs isolated from human cord blood. “C” in the graphs represent CD34+ cells cultured with erythropoietin (Epo) after overnight prestimulation. “P(48)” represents CD34+ cells cultured with Epo for 48 h and then treated with 10 μM synBP1B-dR6 for another 48 h. “P(96)” represents CD34+ cells cultured with Epo for 96 h and treated with 10 μM synBP1B-dR6 for another 48 h.

the BGS2 oligonucleotide duplex was added, the fluorescence lifetime of the tryptophan in synBP1B decreased quite dramatically (Figure 3a and Table S4), indicating the existence of efficient nonradiative processes. The fluorescence lifetime of the tryptophan in the BGS2-synBP1B complex approaches that of the free BP1-HD. We suggest that this high degree of

close proximity of W48 for excited-state electron transfer to occur. It is quite likely that the first and second helices of BP1HD help to stabilize the tertiary structure by interactions with other residues and promoting the W−Y interaction; whereas, synBP1B, lacking some of these residues, is devoid of such stabilizing interactions, causing W and Y to move apart. When 2007


Letters


h. Specificity of the pull-down was established since, without Biotin, synBP1B-dR6 and BIO-synBP1-dR6 were unable to pull-down any significant amount of the fragments containing the amplified region (see Figure 4c). In order to observe the effect of this synthetic transcription factor on HSPCs and its successor cellsand ultimately study their future applicability in proper disease modelshumancord-blood-derived CD34+ hematopoietic stem cells and its ex vivo-differentiated progenitors were studied. In vitro differentiation of HSPCs using erythropoietin is a well-established procedure, and this process has been studied in detail.20 CD34+ cells do not express either the β- or γ-globin genes, perhaps because of repressive epigenetic modifications. After ∼4 days of erythropoietin treatment, both the β- and γ-globin genes start expressing.20 We chose to measure gene expression at two time points around this window, to measure the effects of synBP1B-dR6. We observed ∼1.5-fold β-globin upregulation after 48 h Epo treatment, followed by 48 h of further incubation with 10 μM synBP1B-dR6 (Figure 4d). Prolonging of Epo treatment prior to peptide application from 48 h to 96 h showed further augmentation of β-globin gene activation, up to ∼2-fold (Figure 4d). No change was observed for the δ-globin gene regulation under the same conditions (Figure 4f). However, the γ-globin gene expression underwent rapid increase after 96 h of Epo treatment, followed by 48 h of incubation with synBP1B-dR6 (Figure 4e). As fetal γ-globin expression is believed to be a promising weapon against some types of hemoglobinopathies,21 the observations reported here may indicate a future course of action to ameliorate some of the problems associated with such diseases. Transcription factors play a vital role in many pathogenic processes and specifically bind to sites on the DNA. They are the critical elements that regulate genes and, thus, underpin much of cellular processes. However, it is difficult to target them directly. The effectiveness of the construct reported here in counteracting the repressor activity of a homeodomain transcription factor, DLX-4/BP1, in human-cord-bloodderived hematopoietic stem and progenitor cells suggest that such types of constructs may offer a way to target aberrant transcription factors belonging to the homeodomain class.

quenching upon DNA binding is due to the close proximity of W and Y in the peptide−DNA complex and is indicative of the formation of a BP1-HD-like configuration around the tryptophan. Thus, we conclude that the binding mode of synBP1B in the BGS2-synBP1B complex is similar to that of the BP1-HD. However, we cannot fully rule out the possibility that the proximity of the tryptophan to the DNA is partly responsible for rapid depolarization. To explore the roles of other amino acid residues in DNA recognition, most of the peptide amino acids (Figure S4 and Table S5 in the Supporting Information) in synBP1B was assigned using a combination of TOCSY and NOESY. After the addition of BGS2 at a substoichiometric concentration (at a molar ratio of 1:0.1), we observed chemical shift changes of some residues, the magnitude of which was dependent on the ratio (Figure S4f). At this ratio, only the synBP1B peaks are visible at the recorded signal-to-noise ratio and their changes reflect binding to the DNA, as a result of averaging of chemical shifts of the free and the bound forms under prevailing fast exchange conditions on NMR timescale. Some of the residues showing chemical shift perturbation belong to the N-terminal arm, as well as to the recognition helix (Figure 3b and Figure S4). R5, K6, and Y11 of the arm underwent substantial chemical shift change on BGS2 addition, of which the Y11 change was by far the largest. Although we could not identify the Y11 αH peak (possibly coincides with the water peak), we found a prominent shift of the βH of Y11 (Figure S4e). This large change of the chemical shift was presumably due to dual effects of diminishing the arm dynamicity and Trp−Tyr electronic interaction. Our previous observation of proteinlike tryptophan to tyrosine electron transfer in the BGS2 bound synBP1B in time-resolved fluorescence spectra supports this NMR observation. R5 and K6 of the arm and T21, Q28, N29, S32, and K33 of the recognition helix also experienced significant chemical shift change upon BGS2 addition (Figure 3b). These experiments strongly underlined that synBP1B retains the most important facets of homeodomain−DNA interaction, including interactions of the two important components of DNA recognition: the arm and the recognition helix with the DNA. Gene Regulation by synBP1B in the Cell. To test the functionality of synBP1B ex vivo, we have attached a nuclear localization sequence (PKKKRKV) and a cell penetrating sequence, hexa-D-Arginine, for localization to the cell nucleus. Henceforth, this peptide will be called synBP1B-dR6. Confocal microscopic images indicated the significant nuclear localization of carboxyfluorescein tagged synBP1B-dR6 (see Figure S5 in the Supporting Information). We explored whether synBP1B can occupy the BP1 binding site on the globin gene promoter region. To check that, we have used a chromatin affinity precipitation assay (CHAP) using a biotin labeled synBP1B-dR6 (BIO-synBP1B-dR6) (Figure 4a) and applied it to K562 cells. After peptide cross-linking, the fragmented DNA was adsorbed onto streptavidin-agarose and de-cross-linked. The purified DNA then was amplified with the primers (BGP) covering two known protein binding sites upstream of the βglobin promoter region in K562 cells, in semiquantitative-PCR (see Figure 4b, as well as Table S6 in the Supporting Information). The PCR products indicate increased pull-down of DNA fragments containing the globin gene regulatory region and may conclude that there is accumulation of BIOsynBP1B-dR6 in the same chromatin fragment. Peptide incubation for 24 h shows greater accumulation than for 12

■ ■

METHODS

Please see the Supporting Information for the full methods section.

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acschembio.8b00488. Experimental procedures, and supporting figures and tables (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Siddhartha Roy: 0000-0003-2915-513X Funding

S.R. acknowledges funding from JC Bose Fellowship and DST, Govt. of India. A.S. acknowledges funding from CSIR, DBT (BT/PR13023/MED/31/311/2015) and Ramalingaswami Fellowship. B.G. acknowledges CSIR for support. L.D.B. is 2008


Letters


revealed by high-resolution analysis of sequence preferences. Cell 133, 1266−1276. (18) Subramaniam, V., Jovin, T. M., and Rivera-Pomar, R. V. (2001) Aromatic amino acids are critical for stability of the bicoid homeodomain. J. Biol. Chem. 276, 21506−21511. (19) Shang, Z., Isaac, V. E., Li, H., Patel, L., Catron, K. M., Curran, T., Montelione, G. T., and Abate, C. (1994) Design of a″ minimAl″ homeodomain: the N-terminal arm modulates DNA binding affinity and stabilizes homeodomain structure. Proc. Natl. Acad. Sci. U. S. A. 91, 8373−8377. (20) Wojda, U., Noel, P., and Miller, J. L. (2002) Fetal and adult hemoglobin production during adult erythropoiesis: coordinate expression correlates with cell proliferation. Blood 99, 3005−3013. (21) Wilber, A., Nienhuis, A. W., and Persons, D. A. (2011) Transcriptional regulation of fetal to adult hemoglobin switching: new therapeutic opportunities. Blood 117, 3945−3953.

CSIR Shyama Prasad Mukherjee Fellow. P.M. acknowledges UGC for support. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Mann, R. S., Lelli, K. M., and Joshi, R. (2009) Hox specificity: Unique roles for cofactors and collaborators. Curr. Top. Dev. Biol. 88, 63−101. (2) Kissinger, C. R., Liu, B., Martin-Blanco, E., Kornberg, T. B., and Pabo, C. O. (1990) Crystal structure of an engrailed homeodomainDNA complex at 2.8 Å resolution: A framework for understanding homeodomain-DNA interactions. Cell 63, 579−590. (3) Mpollo, M.-S. E. M., Beaudoin, M., Berg, P. E., Beauchemin, H., D’Agati, V., and Trudel, M. (2006) BP1 is a negative modulator of definitive erythropoiesis. Nucleic Acids Res. 34, 5232−5237. (4) Fu, S., Stevenson, H., Strovel, J. W., Haga, S. B., Stamberg, J., Do, K., and Berg, P. E. (2001) Distinct functions of two isoforms of a homeobox gene, BP1 and DLX7, in the regulation of the β-globin gene. Gene 278, 131−139. (5) Berg, P. E., Williams, D. M., Qian, R.-L., Cohen, R. B., Cao, S.-X., Mittelman, M., and Schechter, A. N. (1989) A common protein binds to two silencers 5′ to the human β-globin gene,. Nucleic Acids Res. 17, 8833−8852. (6) Ebb, D., Tang, D. C., Drew, L., Chin, K., Berg, P. E., and Rodgers, G. P. (1998) Identification of upstream regulatory elements that repress expression of adult beta-like globin genes in a primitive erythroid environment. Blood Cells, Mol., Dis. 24, 356−369. (7) Otting, G., Qian, Y., Billeter, M., Müller, M., Affolter, M., Gehring, W. J., and Wüthrich, K. (1990) Protein−DNA contacts in the structure of a homeodomain−DNA complex determined by nuclear magnetic resonance spectroscopy in solution. EMBO J. 9, 3085. (8) Joshi, R., Passner, J. M., Rohs, R., Jain, R., Sosinsky, A., Crickmore, M. A., Jacob, V., Aggarwal, A. K., Honig, B., and Mann, R. S. (2007) Functional specificity of a Hox protein mediated by the recognition of minor groove structure. Cell 131, 530−543. (9) Abe, N., Dror, I., Yang, L., Slattery, M., Zhou, T., Bussemaker, H. J., Rohs, R., and Mann, R. S. (2015) Deconvolving the recognition of DNA shape from sequence. Cell 161, 307−318. (10) Kajino, M., Fujimoto, K., and Inouye, M. (2011) Side-chain cross-linked short α-helices that behave like original proteins in biomacromolecular interactions. J. Am. Chem. Soc. 133, 656−659. (11) Chakraborty, M., and Roy, S. (2017) A Peptide-based Synthetic Transcription Factor Selectively Down-regulates the Proto-oncogene CFOS in Tumour Cells and Inhibits Proliferation. Chem. Commun. 53, 376. (12) Aravinda, S., Shamala, N., and Balaram, P. (2008) Aib residues in peptaibiotics and synthetic sequences: analysis of nonhelical conformations. Chem. Biodiversity 5, 1238−1262. (13) Banerjee, R., Basu, G., Chene, P., and Roy, S. (2002) Aib-based peptide backbone as scaffolds for helical peptide mimics. J. Pept. Res. 60, 88−94. (14) Mo, Y., Ho, W., Johnston, K., and Marmorstein, R. (2001) Crystal structure of a ternary SAP-1/SRF/c-fos SRE DNA complex. J. Mol. Biol. 314, 495−506. (15) Wuthrich, K., Billeter, M., and Braun, W. (1984) Polypeptide secondary structure determination by nuclear magnetic resonance observation of short proton-proton distances. J. Mol. Biol. 180, 715− 740. (16) Wilson, D. S., Sheng, G., Jun, S., and Desplan, C. (1996) Conservation and diversification in homeodomain-DNA interactions: A comparative genetic analysis. Proc. Natl. Acad. Sci. U. S. A. 93, 6886−6891. (17) Berger, M. F., Badis, G., Gehrke, A. R., Talukder, S., Philippakis, A. A., Pena-Castillo, L., Alleyne, T. M., Mnaimneh, S., Botvinnik, O. B., Chan, E. T., et al. (2008) Variation in homeodomain DNA binding 2009
