Synthetic progress in cMyc-Max oncoprotein miniaturization: Semi-online monitoring gives solid-phase access to hydrophobic b(-HLH-)ZIP peptidosteroid tweezers

FULL PAPER

DOI: 10.1002/ejoc.201201235

Synthetic Progress in cMyc-Max Oncoprotein Miniaturization:

Semi-Online Monitoring Gives Solid-Phase Access to Hydrophobic

b(-HLH-)ZIP Peptidosteroid Tweezers

Dieter Verzele^[a]and Annemieke Madder*^[a]

Dedicated to Professor Dr. Pierre J. De Clercq on the occasion of his 65th birthday

Keywords: Solid-phase synthesis / Peptides / Peptidosteroids / Bioorganic chemistry / Protein models / Leucine zippers

Miniature versions of basic leucine zipper (bZIP) and basic

helix–loop–helix zipper (b-HLH-ZIP) transcription factors are

promising tools for molecular dissection of the genetic infor-

mation in a post-genomic context. Despite the opportunities

of genome interfering agents based on certain oncogenic zip-

per proteins, structural mimicry of transcription factors is a

delicate undertaking, and experimental fine-tuning through

bottom-up organic chemistry could benefit from solid-phase/

library approaches. Involved in a variety of human patholo-

gies, we became interested in the miniaturization of the

cMyc-Max b-HLH-ZIP oncoprotein, and herein elaborate on

our synthetic progress in that direction. A bile acid scaffold

was successfully employed as artificial dimerization interface

in this new type of transcription factor model. Orthogonality

of the applied Alloc/Boc/Fmoc chemistries allowed the syn-

thesis of both homo- and heterodimeric peptidosteroid conju-

gates, covalently restricted with defined geometrical proper-

ties. Recognition peptides were assembled through standard

Fmoc/tBu solid-phase peptide synthesis (SPPS) chemistry,

assisted by automated procedures for consecutive chain

elongation on solid support. Invaluable to monitor present

strategy, a photocleavable linker allowed rapid, yet detailed

analysis of side chain protected peptide intermediates, liber-

ated from the sampled resin, by reverse-phase HPLC and

MALDI-TOF-MS. By decorating each scaffold position with

two basic region peptides in a 2ϫ2 design, a first generation

of unprecedented b(-HLH-)ZIP peptidosteroids was effi-

ciently obtained. As such, a versatile methodology amenable

to library generation is presented.

Although sharing a similar mode of binding DNA target

sites with the scrutinized basic leucine zipper (bZIP) equiva-

lents, artificial mimics of these basic helix–loop–helix zipper

(b-HLH-ZIP) TFs are conspicuously few. Despite the ex-

ceptional total chemical synthesis (172 residues) yielding a

covalently stitched replica of the vertebrate cMyc-Max

(proto-)oncoprotein dimer achieved by Kent et al. in 1995

(Figure 1),^[6]translation of this precedent towards miniature

peptide derivatives is long overdue. A supplementary loop

in the replaceable HLH-ZIP dimerization interface being

the only salient difference, development of synthetic models

of the cMyc and Max proteins carves out a niche for inno-

vative initiatives. The preponderance of leucine zipper de-

signs has been further supplemented by the singular MyoD-

MyoD mimic, communicated by Morii et al. before finally

resorting to the GCN4 bZIP-standard (vide infra).^[7]De-

spite the lack of a zipper region in the dimerization inter-

Introduction

Aberrant modulation of gene expression at the transcrip-

tional level is at the origin of numerous diseases. By dysreg-

ulating cell growth and triggering cell proliferation, various

oncoproteins carry out their biological functions as tran-

scription factors (TFs),^[1]critically relying on their DNA-

binding capacities. Directly involved in human tumorigene-

sis and cancer, the pivotal Myc member^[2]of the surround-

ing Myc/Max/Mad network^[3]is a prominent example of

how these proteins are promising targets towards novel che-

motherapeutics.^[4,5]The deregulated proliferation of Myc by

overexpression has been implicated in the progression of

many aggressive malignancies, including Burkitt’s lym-

phoma, neuroblastomas, and small-cell lung cancers.

[a] Laboratory for Organic and Biomimetic Chemistry,

Department of Organic Chemistry, Faculty of Sciences, Ghent face, similarities between the loop-containing b-HLH-ZIP

and this b-HLH type of TF^[8]offer excellent precedent for

University,

Krijgslaan 281 (S4), 9000 Ghent, Belgium

the current work.

Fax: +32-9-264-4998

E-mail: annemieke.madder@ugent.be

In spite of the advances in the de novo development of

such peptide miniatures, progress in this area has been slow

compared to the wealth of available high-resolution data.

Homepage: www.orgchem.ugent.be

Supporting information for this article is available on the

WWW under http://dx.doi.org/10.1002/ejoc.201201235.

Eur. J. Org. Chem. 2013, 673–687

673

D. Verzele, A. Madder

FULL PAPER

Figure 1. Similar bipartite arrangement of bZIP (left) and b-HLH-ZIP (right) transcription factors, pictorial illustration of the original

artificial counterparts by Kim et al. and Kent et al., and our heterodimeric peptidosteroid design (box, not scaled).

Only a handful of bottom-up synthetic approaches have work, there exists no doubt about the virtues of organic

succeeded in mimicking the DNA-recognizing potential of synthesis for these objectives.

naturally occurring TFs by minimized versions. Deduction

The eukaryotic bZIP and b-HLH-ZIP motifs (Figure 1)

of reliable binding principles or recognition rules has been are among the simplest protein structures able to bind the

hampered by the large variability of TF folds and discrimi- DNA major groove in a sequence-specific way.^{[12c,19–21]}Re-

nation mechanisms. A universal pairing code for the re- lying on the formation of non-covalent dimers, most of the

cognition between natural amino acids (AAs) and nucleic direct contacts between the TF residues and the DNA nu-

acids has yet to be generalized, and only tentative guidelines cleobases are made within the recognition α-helices of the

have been established.^[9,10]The fragile balance between so-called basic region, which fork across the major groove

structural minimization and biophysical outcome addition- to grip the double helix in a tweezer-like fashion. The mod-

ally impedes the rational design of downsized mimics and ularity and tractability of the zipper-type proteins has made

conspires against biomimetic efforts, which demand dedi- them attractive frameworks for the development of artificial

cated trial-and-error tuning of the empirical constraints. So- counterparts. The bipartite arrangement readily suggests

lid-phase/library initiatives are therefore desirable. Remark- design opportunities, as illustrated in Figure 1, and peptides

ably, various b-HLH and b-HLH-ZIP (i.e., loop-contain- derived from the basic region can serve as the simplest mod-

ing) proteins display only limited DNA affinity and/or spec- ules to target specific DNA sequences. In this respect, albeit

ificity outside of living cells, despite their precise physiologi- conceptual, such an approach makes the bZIP, b-HLH-ZIP,

cal regulation in vivo. In this respect, Myc (proto-)oncopro- and b-HLH proteins uniform in terms of design.

teins are often called “enigmatic”, as they hide their details

Various innovative strategies have been implemented to

during many in vitro assays.^[11]Given (a) the influence of meet thermodynamic requirements^[22]in miniaturization ef-

loop-projecting DNA contacts,^[12](b) the ambiguous role of forts.^[23]Given the vital role of dimerization to mediate tight

loop-associated flexibility,^[13](c) the constraining/stabilizing major groove recognition,^[24]the extended C-terminal di-

interactions with essential accessory mediators of the tran- merization domain has often been replaced by artificial

scriptional machinery in vivo,^[14]and (d) the particular reg- connectors, both covalent and non-covalent. Showing a rare

ulatory subtleness of these proteins, the scarcity of success- structural transparency despite the debatable practical rele-

ful literature precedents might hint towards the daunting vance, the most progress has by far been made with the

nature of the corresponding miniaturization attempts.^[15,16]

classical GCN4 leucine zipper basic region peptide, as men-

Considering the variety of challenging opportunities, the tioned earlier. In conjunction with (total synthesis) en-

present work aims to make contributions to this research deavors towards larger constructs, and complemented by

area by expanding the repertoire of miniature TF models. numerous biotechnological and semisynthetic approaches,

Besides the simple desire to mimic Nature, such peptide several so-called minimalistic strategies have been reported,

probes hold the potential to assess the implications of the next to the development of models other than the zipper-

exquisite selectivity of natural TFs on the control of gene type.^[25,26]The current contribution exploits the chemical

expression. An emerging area in the current post-genomic approach, using bottom-up organic synthesis to develop C-

era, we previously highlighted the pharmacological signifi- terminally dimerized b(-HLH-)ZIP-like models of reduced

cance of such TF mimics.^[17,18]By allowing the unlimited size, simultaneously targeting the DNA major groove by

introduction of artificial elements into the peptidic frame- two peptide strands comprising natural residues. In the

674

www.eurjoc.org

Eur. J. Org. Chem. 2013, 673–687

Synthetic Progress in cMyc-Max Oncoprotein Miniaturization

seminal 1990 publication of Kim et al., a simple disulfide

bridge sufficed for tethering the GCN4 basic region pep-

tides (Figure 1).^[27,28]Since then, a substantial number of

attempts have been directed towards the incorporation of

dedicated surrogate modules with increasing complexity.

These include an N^α-,N^ε-lysine linkage (Ebright et al.),^[29]

a

bis(terpyridyl)iron(II) coordination complex (Schepartz

et al.),^[30]a β-cyclodextrin:adamantyl host–guest inclusion

complex and bridged enantiomeric biphenyl derivatives

(Morii et al.),^[31]and a photoresponsive azobenzene moiety

(Mascareñas et al.).^[32]

For the understanding of DNA–protein recognition and

the principles of TF mimicry, macro/supramolecular as-

semblies with specific architectural features and/or physico-

chemical properties are still needed. While a decade has

passed since Mascareñas’ homodimeric GCN4-GCN4 pho-

totrigger, the contrasting void of HLH counterparts stimu-

lated initiatives from our own group. We herein report on

our methodological progress towards the first generation of

downsized b(-HLH-)ZIP models with an emphasis on the

therapeutically relevant cMyc-Max oncoprotein.

Results and Discussion

Scheme 1. Preparing for the linear SPPS strategy: Synthesis of so-

lid-supported scaffold starting material 5 through immobilization

of steroid scaffold 1 via photolabile linker resin 3. Reagents and

conditions: (i) PyBOP, DIPEA, DMF, r.t., double coupling: 3 +

5 h. (ii) Piperidine in DMF (20% v/v), r.t., triple deprotection: 2 +

5 + 15 min. (iii) (a) Scaffold 1, PyBOP, DIPEA, DMF, r.t., single

coupling overnight; (b) 1-AcIm, CHCl₃, r.t., overnight. (iv) TFA in

DCM (20% v/v), r.t., triple deprotection: 2 + 5 + 15 min (washing

included DIPEA in DMF: 10% v/v).

b(-HLH-)ZIP Peptidosteroids: Preparing for a Linear,

Monitored SPPS Approach

The present work harnessed a steroid framework as an

artificial dimerization interface displaying defined geomet-

rical features (Figure 1 and Scheme 1). Among the variety

of molecular scaffolds (templates) used in supramolecular

chemistry, steroids in general and bile acids in particular

are versatile synthons for applications based on coopera-

tivity and multivalency.^[33]The interest in these carbocycles directional orientation (so-called α-stereochemistry) and

is readily explained by their unique combination of biocom- suitable spatial arrangement (ca. 7 Å apart) of the C3–C12

patibility, chirality, high availability, and various function- steroid appendages thus offers an attractive starting point

alization patterns that can be modified in a tunable manner. to seize the DNA duplex, and thereby, their slight diver-

The well-spaced array of selectively addressable functionali- gence^[37]is a further asset. Ligated onto O-alkylated cholic

ties, distributed around the tetracyclic backbone, and the acid derivatives in solution, the increased α-helical content

curved cavity profile resulting from the cis-A/B ring junc- of the homotrimeric miniproteins (up to 3ϫ37 = 111 AAs)

tion are ideally suited for receptor design. Pioneered by Still of Wang et al. bears special mentioning in the current

et al.,^[34]peptidosteroid libraries and macrocycles are now b(-HLH-)ZIP story.^[38]Mutual induction of α-helicity by

accessible.^[35]Further interest has been cast on the im- the proximal, codirectional peptide strands might thus be

proved pharmacological characteristics of peptides, pro- anticipated.^[39]Next to enhancing stability of the recogni-

teins, and Lipinski structures as drug candidates in a so- tion helix, selective binding can be increased upon restric-

called Trojan Horse strategy upon conjugation.^{[33b–33e,36]}

tion of the conformational flexibility of the residues that

Compared to the large volume of high-resolution data, form specific contacts with the nucleobases.^[13,14,26]

solved structures of HLH-containing proteins are relatively Through a series of homo- and heterodimeric GCN4-based

sparse, and the X-ray views of cMyc-Max and Mad-Max designs, the importance and subtleness of scaffold geome-

heterodimeric complexes by Burley et al. are a particularly try, rigidity, and chirality was assessed by both the

important achievement.^[12c]From a structural point of Schepartz and Morii groups.^[30,31]Interestingly, the small

view, the extended dimerization interface of the natural cis-azobenzene unit of the molecular switch^[32]designed by

(HLH-)zipper proteins is regarded as a supramolecular Mascareñas et al. shows dimensions and stiffness compar-

scaffold to preorganize, stabilize, and present the relatively able to our steroid.

small, N-terminal recognition α-helices in an appropriate

A concise route towards dipodal building block 1 (C3α-

geometrical positioning relative to the DNA duplex. Be- NHAlloc, C12α-NHBoc-diamino-5β-cholan-24-oic acid;

cause dimerization is a prerequisite, proper docking of the Scheme 1), prepared in just six simple steps from commer-

peptides in the DNA major groove is another necessity. Co- cially available deoxycholic acid (3α,12α-dihydroxy-5β-cho-

Eur. J. Org. Chem. 2013, 673–687

www.eurjoc.org

675

D. Verzele, A. Madder

FULL PAPER

lan-24-oic acid), was developed in our group.^[40,41]Basic re- tionalization of the C12 position should precede C3

gion peptides have now been constructed at both the C3α derivatization because of steric constraints. Already suffer-

and C12α amine groups through stepwise, linear solid- ing from steric impediments of the steroid framework, ad-

phase peptide synthesis (SPPS).^[42,43]The natural spacer ditional hindrance upon prior C3 derivatization could

and carboxylic acid moiety at the C24 position conveniently render the C12 position inaccessible. Acidic NHBoc cleav-

enable anchoring of our steroid scaffold to a solid-phase age prior to peptide generation therefore accommodates the

resin. Compatible with a broad range of chemical condi- use of standard AAs, which are side chain protected by

tions, the multidimensional protecting group (PG) strategy acid-labile moieties in the Fmoc/tBu SPPS approach. Man-

thereby hinged on the introduction of Holmes’ nitrovera- ual attachment of the very first residue to solid-supported 5

tryl-based photosensitive linker 2 (4-{4-[1-(9-fluorenyl- preceded the automated generation of the first basic region

methoxycarbonylamino)ethyl]-2-methoxy-5-nitrophenoxy}- peptides at the C12 position of the resin-bound scaffold,

butanoic acid)^[44]to yield supported counterpart 3 immobi- and this ensured adequate coupling at the hindered steroid

lized on TentaGel NH₂. With light as a reagent, photolytic appendage (Scheme 2). A standard PyBOP [benzotriazol-1-

cleavage of a chemical bond by the absorption of a photon yl-N-oxy-tris(pyrrolidino)phosphonium

hexafluorophos-

can only occur through a limited number of pathways, phate] mediated procedure provided straightforward intro-

which results in superior orthogonality, and hence, minimi- duction of glutamine and asparagine, which were both side

zation of premature loss and controlled liberation of the chain protected with a bulky trityl moiety. Consecutive

attached species at any given time. Fine-tuned through dif- automated chain elongation from the C terminus to the N

ferent substituents by Holmes et al.,^[45]this o-nitrobenzyl- terminus resulted in side chain protected, monomeric inter-

derived linker shows excellent photoreactivity, with release mediates. After N^α-Fmoc deprotection of 6 and 7, synthesis

of primary carboxamide species upon irradiation with UV of the GCN4 peptide at resin 8 and cMyc peptide at 9

light.^[46]

yielded peptidosteroids 10 and 11, respectively. Although

To save on precious scaffold material, a single coupling the 23-residue GCN4 peptide by Kim et al.^[51]has become

of a slight excess amount of the substrate was performed popular, as minimal sequence is required for sequence-spe-

overnight. Quantitative resin derivatization was monitored cific binding by α-helix formation, the significantly shorter

by common color tests^[47]by visually detecting resin-bound MyoD mimic of Morii et al., truncated at both the N and

amino groups with specific reagents. Whereas a simple C termini, supports the potential of the current cMyc/Max

TNBS (2,4,6-trinitrobenzenesulfonic acid) test allowed the peptide length (only 16 AAs), containing the essential

conversion to be followed until construct 4 was obtained, DNA-contacting interface and key residues.

the in-house-developed NF31 test was required to follow

reactions at the hindered C12 amino group.^[48]Considering

the intricacies of the desired b(-HLH-)ZIP peptidosteroids,

thorough verification of intermediate compounds is, how-

ever, of paramount importance. Therefore, evaluation was

complemented by ESI-MS, which was ultimately substi-

tuted by MALDI-TOF-MS and reverse-phase (RP)-HPLC

analysis upon subsequent decoration. The unique reactivity

of the photocleavable handle furnishes a convenient system

for advanced monitoring in a semi-online fashion.^[49]Sam-

ples for detailed analysis are readily available upon simple

irradiation of resin aliquots in an appropriate solvent. Only

minute quantities are needed for proper evaluation at every

stage of the synthesis. The mild cleavage conditions thereby

allow the non-destructive release of side chain protected in-

termediates. Although Moss et al. recently disclosed an op-

timized procedure for mass spectrometric analysis of pro-

tected, synthetic peptides,^[50]these compounds remain a

Scheme 2. Synthesis of monomeric C12 peptidosteroid intermedi-

ates: manual introduction of initial residue and automated SPPS

challenge for RP-HPLC and mass spectrometry, and the

of the first strand at the C12 steroid position.^[52]Reagents and con-

analytical “dos and don’ts” regarding our hydrophobic

macromolecules form a vital part of the present contri-

bution.

ditions: (i) N^α-Fmoc-protected Gln(Trt) or Asn(Trt), PyBOP,

DIPEA, DMF, r.t., double coupling: 4 + 4 h. (ii) Piperidine in

DMF (20% v/v), r.t., triple deprotection: 2 + 5 + 15 min. (iii) Until

completion (with intermediate sampling): (a) N^α-Fmoc-protected

residue, HBTU, DIPEA, NMP (+ DMF trace), r.t., single coupling:

3 h; (b) piperidine in NMP (40% v/v), r.t., triple deprotection: 2 +

5 + 15 min.

Aiming for Diversity in a 2

؋

2 Dimeric Peptidosteroid

Library Setup

From C12 Monomeric Peptidosteroids ...

Relying on the established orthogonality of the C3 on the solid phase often depends on minutiae,^[43]different

NHAlloc and C12 NHBoc scaffold differentiation, func- parameters of the automated protocol were carefully con-

Given that the outcome of an extended linear procedure

676

www.eurjoc.org

Eur. J. Org. Chem. 2013, 673–687

Synthetic Progress in cMyc-Max Oncoprotein Miniaturization

Table 1. Semi-online monitoring: side chain protected C12 peptidosteroid intermediates cleaved from consecutive resin samples.^[a–c]

[a] During automated synthesis and (intermediate) cleavage (i.e., analysis), peptides were protected with standard side chain PGs.^[52]

Peptidosteroids were released (and analyzed) as C24 carboxamides upon UV-light irradiation of the corresponding resin,^[46]which was

sampled after Fmoc deprotection of the N terminus. [b] See Figure 2 for RP-HPLC details of the cMyc synthesis. [c] See Supporting

Information for further RP-HPLC details of these semi-online monitoring efforts: Table S4 & Figures S9–S11.

sidered, and the integrities of the peptides were monitored tion of a C18/100 Å stationary phase resulted in severe tail-

at regular sequence intervals. Given that standardized SPPS ing and ultimate failure of analysis, once reaching/exceeding

protocols are inherently restricted to the routine application ca. 16 residues,^[57]in contrast to non-scaffolded counter-

of a sequential assembly of 50 residues on average, the par- parts. This outcome reflects the augmented lypophilicity

ticular steroid moiety could further impair synthetic viabi- and/or bulkiness of the peptidosteroid conjugates, and in-

lity. Close proximity of the covalently constrained peptides

might considerably affect the synthesis, which would pre-

vent reliable derivatization as a result of significant hin-

drance and/or interstrand interactions. The hydrophobic

nature of the carbocyclic framework could result in particu-

lar physicochemical features and further complicate both

preparation and analysis. Next to the so-called pseudodi-

lution effect of low-loaded resins, solvation and accessibility

of the generated peptides might benefit from using NMP

(N-methylpyrrolidone) as a solvent.^[53]Trying to ensure

adequate derivatization while obviating double coupling

steps, an extended 3-h coupling period was allowed,^[54]

whereas the concentration of the HBTU {1-[bis(dimethyl-

amino)methylene]-1H-benzotriazolium hexafluorophosphate

3-oxide} coupling mixture (0.16 m of active species) was

considered the central parameter for reliable generation of

the extended sequences.^[55]Presented in Table 1, stop com-

mands were incorporated at particular positions of the

automated protocol. Sampling after NHFmoc deprotection

should enhance compatibility of the side chain protected

intermediates with the analytical techniques. The unprotec-

ted amino terminus facilitates both elution in reverse-phase

chromatography and ionization during mass spectrometry.

The essential crystallization during sample preparation for

MALDI-TOF-MS might additionally be affected by the

presence of the polar N-terminal moiety.

Monitoring of the current protocol by consecutive, pho-

tolytically cleaved samples^[56]proved efficient and allowed

reliable evaluation of the synthetic outcome. Initial

MALDI-TOF-MS data confirmed the identity of the de-

sired intermediate and final peptides. Furthermore, excel-

lent results were obtained with RP-HPLC (Figure 2), and

analysis of the protected peptide segments relying on

chromatography offers a valuable tool for facilitated moni-

toring and evaluation in cases where MALDI-TOF-MS

lacks reliability or fails (see the Supporting Information for

further RP-HPLC details of these semi-online monitoring

efforts: Table S4 & Figures S9–S11). Whereas side chain

PGs allow sensitive UV detection, free N termini assure

proper elution of the hydrophobic compounds. However,

Figure 2.

Semi-online

monitoring:

RP-HPLC

(254 nm,

gradient 1)^[58]of side chain protected C12 peptidosteroid interme-

diates from the synthesis of cMyc monomer 11 cleaved from con-

secutive resin samples. See Table 1: (a) Sample 1, (b) Sample 2, and

(c) Sample 3 (i.e., complete C12 monomer) analyzed with the C18/

100 Å column, and (d) Sample 3 reanalyzed with the C4/300 Å col-

demonstrating the influence of the steroid moiety, applica- umn.

Eur. J. Org. Chem. 2013, 673–687

www.eurjoc.org

677

D. Verzele, A. Madder

FULL PAPER

creased stationary phase polarity and/or pore-size seemed equivalent of the palladium catalyst was used to improve

necessary. Application of a C4/300 Å stationary phase in workability at the μmol scale. In the complex chemical con-

that case provided adequate, sharp elution of both 10 and text of the present peptidosteroid intermediates, it was

11.

opted to rely on the firmly established procedure of Guibé

et al.,^[62]which involves palladium-catalyzed hydrostanno-

lysis with tributyltin hydride as the nucleophilic scavenger

... through Key C3 Manipulations ...

bZIP and b-HLH-ZIP proteins have many members that in the presence of a proton source. To further suppress un-

form homo- and heterodimers, a feature that significantly desired back-allylation and ensure rapid conversion, a sup-

expands the repertoire of DNA-binding specificities from a plementary morpholine scavenger was added in excess

limited number of protein partners (i.e., combinatorial gene amount. Whereas a variety of proton donors has been re-

regulation).^[24]Although transcription regulation by zipper- ported, simple application of moist DCM (dichlorometh-

type proteins is often mediated by heterodimerization, the ane) provided rapid degradation of the deprotection inter-

majority of synthetic miniature designs has focused on the mediates. Next to satisfying MALDI-TOF-MS analysis, ex-

homodimeric bZIP-GCN4 protein, as covered earlier. As a cellent RP-HPLC results with the C18/100 Å column were

result of orthogonal N protection, a unique feature of the obtained (Figure 3). Removal of the lipophilic NHAlloc

applied scaffold is the straightforward solid-phase genera- moiety and/or liberation of the hydrophilic C3 amino group

tion of homo- and heterodimeric peptidosteroid tweezers. appreciably influenced chromatographic behavior, com-

Steroid conjugation further confers enhanced biostability^[59]pared to the precluded C18/100 Å elution of both 10 and

(next to bioavailability) compared to non-scaffolded peptide 11 (and capped counterparts 12 and 13).

counterparts, as mentioned above,^[60]and the associated

partners are covalently secured. Whereas natural dimeriza-

tion networks are governed by non-covalent interactions,

the present strategy allows a stable connection between

both partners, which facilitates further studies. The auto-

mated solid-phase approach should be amenable to the par-

allel synthesis of a b(-HLH-)ZIP library, and this offers the

significant advantage to accelerate the meticulous adjust-

ment of recognition capabilities. As the sole literature pre-

cedent of such initiatives, the value of a library strategy in

TF mimicry was noted in 1996 by Ebright et al., who re-

ported comparable but less-sophisticated bZIP heterodi-

mers with an N^α-,N^ε-lysine linkage on solid phase.^[29]To

explore the potential of solid-phase methodology for gener-

ating molecular diversity, a 2ϫ2 combination of the GCN4

bZIP and cMyc/Max b-HLH-ZIP monomeric basic region

peptides to afford four different, dimeric peptidosteroids

(vide infra: Scheme 4) was applied. After parallel C12 deco-

ration of construct 5 with the GCN4 or cMyc sequence,

both resins were portioned into two new batches prior to

C3 assembly. At the C3 position, once again either the

cMyc strand or the related Max strand was synthesized.

Structurally, the obtained peptidosteroids consist of either

16 + 16 or 16 + 23 residues, and this represents either

homo- or heterodimeric constructs in a quasi- or non-sym-

metrical arrangement. Whereas the former represent minia-

ture models of natural b-HLH-ZIP combinations, the latter

are non-natural conjugates of a yeast bZIP-peptide and a

biologically unrelated mammalian b-HLH-ZIP-strand.^[61]

Both the bZIP + b-HLH-ZIP combinations and the syn-

thetic miniature models of the naturally occurring b-HLH-

ZIP Myc-Max network are unprecedented.

Strategically, prior capping of the C12 N terminus is

mandatory to proceed towards synthesis of the second

Figure 3. From C12 monomers to C3–C12 dimers through N-ter-

minal acetylation and key C3 NHAlloc removal: RP-HPLC (C18/

100 Å, 214 nm, gradient 1)^[58]of (a) GCN4-appended 14 and

(b) cMyc-appended 15. Reagents and conditions: (i) 1-AcIm,

CHCl₃, r.t., overnight. (ii) Pd(PPh₃)₄, Bu₃SnH, morpholine, DCM,

strand at the C3 position. Mild overnight treatment of res-

ins 10 and 11 with an excess amount of 1-AcIm (1-acetyl-

imidazole) in CHCl₃provided satisfying acetylation, fur-

nishing 12 and 13, respectively (Figure 3). Subsequent re-

moval of the C3 NHAlloc PG proceeded smoothly, yet one r.t., single deprotection: 2 h.

678

www.eurjoc.org

Eur. J. Org. Chem. 2013, 673–687

Synthetic Progress in cMyc-Max Oncoprotein Miniaturization

Scheme 3. From C12 monomers to C3–C12 dimers through key attachment of the first C3 residue. Reagents and conditions: (i) Piperidine

in DMF (20% v/v), r.t., triple deprotection: 2 + 5 + 15 min. (ii) N^α-Fmoc-protected Arg(Pbf), PyBOP, DIPEA, DMF, r.t., double coupling:

2.5 + 5 h.

Different batches of Alloc-deprotected resins 14 and 15 shown that manual PyBOP-mediated attachment of a sub-

were C3 derivatized with either protected aspartate or pro- sequent arginine residue at compound 20 provided straight-

tected asparagine towards compounds 16–19 (Scheme 3). forward formation of 21. Although confirming particular

Manual attachment of the first residues again preceded demands of the constrained C3 position, a single residue

automated peptide assembly, and couplings were carefully seemed to sufficiently protrude for reliable further derivati-

evaluated. RP-HPLC equipped with the C18/100 Å station- zation by the automated procedure. Remaining resins 16–

ary phase allowed the consumption of starting materials 14 18 were precautionary N capped by 1-AcIm prior to N^α-

and 15 to be assessed. Surprisingly, a double PyBOP-medi- Fmoc deprotection to yield 22–24, respectively. At this piv-

ated procedure failed to quantitatively derivatize the C3 otal C3 position, once again the C4/300 Å stationary phase

amino group, in contrast to the previous outcome at the demonstrated general compatibility with all intermediates,

hindered C12 appendage. Although the C3 position of the which vouches for its superior verification and monitoring

current scaffold has generally been considered the most re- (see Supporting Information for RP-HPLC details and tab-

active site, accessibility might be significantly impaired by ular summary: Table S5).

the proximal, side chain protected first strand present in

Relying on the satisfying synthesis of the monomeric in-

this particular constrained design. Further PyBOP-based termediates, the current procedure was further exploited for

attempts were unsuccessful. Gratifyingly, application of the automated generation of the second peptide strand at

the renowned HATU {1-[bis(dimethylamino)methylene]- the C3 position (Scheme 4). Synthetic feasibility could how-

1H-1,2,3-triazolo[4,5-b]pyridinium hexafluorophosphate 3- ever be abrogated by the particular polarity and complexity

oxide} reagent provided increased coupling. However, of the macromolecular targets. Bearing structural similari-

whereas quantitative conversion of 15 into 19 was suggested ties with so-called helical bundles, interchain interactions

by RP-HPLC and MALDI-TOF-MS (see Figure S20, Sup- between both peptides could render aggregated strands in-

porting Information), subsequent N^α-Fmoc deprotection accessible for adequate chain elongation.^[42,64]Attempting

still indicated the presence of contaminating starting mate- to continue the monitoring strategy, samples were again

rial 15 (see Figure S22, Supporting Information). Double, subjected to RP-HPLC and MALDI-TOF-MS. Not sur-

perhaps triple, HATU coupling might therefore be consid- prisingly, incorporation of five additional residues at 25 def-

ered for future research. Liberation of the N^α-amino group initely prevented proper elution on both the C18/100 Å and

restored the proper eluting behavior and confirmed deriva- C4/300 Å stationary phase and detection by MALDI-TOF-

tized 20 as the major compound. Additionally, prior purifi- MS.

cation of samples by analytical RP-HPLC and subsequent

... to C3–C12 Dimeric Peptidosteroids

lyophilization enhanced the crystallinity of the solid spots,

improving MALDI-TOF-MS analysis.^[63]Whereas data

Further attention was therefore immediately focused on

indicate preferential application of the α-cyanohydroxy- subsequent acidic deprotection towards the desired, dimeric

cinnamic acid matrix for analyzing previous, monomeric peptidosteroid compounds (Scheme 4). RP-HPLC and

compounds, monitoring of the current C3 manipulation is MALDI-TOF-MS analysis benefits from liberation of the

better accommodated by 2,5-dihydroxybenzoic acid.

numerous hydrophilic side chain functionalities. Although

Considering these challenges, automated generation of aromatic substitution by the liberated electrophiles^[65]could

the second strand might compromise significant steric con- significantly alter cleavage efficiency, resemblance of the

straints. Satisfyingly, in a preliminary test reaction it was employed photocleavable linker to the Rink amide {4-[(2,4-

Eur. J. Org. Chem. 2013, 673–687

www.eurjoc.org

679

D. Verzele, A. Madder

FULL PAPER

dimethoxyphenyl)(Fmoc-amino)methyl]phenoxyacetic acid}

Considering the relevance of the cMyc-Max heterodimer,

and HMPB [4-(4-hydroxymethyl-3-methoxyphenoxy)but- further attention was focused on peptidosteroid 31. In con-

yric acid] linkers supported its acid stability.^[66]Both com- trast to straightforward single (3 h) irradiation of analytical

pound recovery and integrity could be compromised by (the resin samples in small test tubes, scaled photolysis requires

combination of) increased photostability, altered reactivity a careful experimental setup. In view of the generally ob-

profile, and/or premature loss. Thus, to avoid the incontrol- served opacity of samples containing larger compounds

lable occurrence of side reactions at the photolabile linker upon UV cleavage in ACN, solubility issues might hamper

and the loss of precious compounds at this advanced stage, the efficient release of our hydrophobic, side chain pro-

application of an inverse photolysis–acidolysis protocol was tected peptidosteroids. Therefore, the influence of different

elaborated without further assessment of acidolytic linker solvents on the photolytic outcome was shortly assessed.

cleavage. Upon the usual UV cleavage of side chain pro- ACN, EtOH, and acetone provided satisfying results,

tected samples 26–29 in ACN (acetonitrile), the resulting whereas samples in THF, dioxane (+2% DMSO) and

opaque solutions were separated from the resin, and the iPrOH showed lower purity. Direct exposure of the floating

evaporated material was subjected to the TIS (triisoprop- resin to UV irradiation and/or generation of radical species

ylsilane) containing acidolytic cocktail at room temperature might account for complete degradation in CHCl₃.^[69]Con-

for ca. 2 h. Upon concentration^[67]and Et₂O trituration, sidering the particular demands of our fully protected mac-

flocculation seemed to benefit from warming of the samples romolecular conjugates, EtOH was selected over usually

to room temperature and a short (Ͻ1 min) period of sonic- employed ACN for further experiments. While matching the

ation and intense vortexing. Consistent with the multitude desirable intermediate boiling point, decreased sample

of side chain protected residues in general and arginine- opacity indicated an enhanced solubility. The sufficiently

protecting Pbf (2,2,4,6,7-pentamethyldihydrobenzofuran-5- low volatility of this (regularly replenished) solvent must

sulfonyl) groups in particular, incomplete deprotection was prevent complete evaporation and minimize potential risks

observed in the RP-HPLC and MALDI-TOF-MS data. To of directly irradiating dry resin/compound with intense UV

ensure complete deprotection, yet prevent decomposition, light.^[56]Peptide integrity further benefitted from several

either 50 °C for 4 h or 60 °C for 2.5 h was applied to these successive, rather short (max. 4 h) periods of photolysis

partially deprotected compounds. Figure 4 (a) and (b) with intermediate isolation, while providing sufficient cleav-

shows the excellent MALDI-TOF-MS and RP-HPLC data age recovery. Although extensive rounds of UV cleavage

of crude precipitate 31. Desired peptidosteroid dimer 31 might be tempting yield-wise, RP-HPLC evaluation indi-

was obtained as the major component, mainly contami- cated an appreciable decrease in compound liberation in

nated by an acceptable amount of the C3-capped deletion later samples, with a relative increase in contaminating spe-

sequence, which was anticipated from the above-discussed cies.

experiments.^[68]Comparable results were obtained for the

larger, GCN4-containing conjugates 32 and 33.

After single reagent B deprotection of cleaved 27 at an

elevated temperature overnight, crude 31 was obtained in

Scheme 4. Towards the final C3–C12 dimers in a 2ϫ2 b(-HLH-)ZIP peptidosteroid library format. Reagents and conditions: (i) Until

completion: (a) N^α-Fmoc-protected residue, HBTU, DIPEA, NMP (+ DMF trace), r.t., single coupling: 3 h; (b) piperidine in NMP (40%

v/v), r.t., triple deprotection: 2 + 5 + 15 min. (ii) UV-light irradiation (365 nm) either on small scale (ACN, single cleavage: 3 h) or

preparatory scale (EtOH, triple cleavage: 3.5 + 2.5 + 3.5 h). (iii) Reagent B (88% v/v TFA + 5% H₂O + 5% phenol + 2% TIS) under

various conditions (small scale: r.t. for ca. 2 h + 50–60 °C for 2.5–4 h; or preparatory scale: 60 °C overnight) (see Experimental Section

for details).

680

www.eurjoc.org

Eur. J. Org. Chem. 2013, 673–687

Synthetic Progress in cMyc-Max Oncoprotein Miniaturization

62% isolated yield (as the TFA salt) over the photolysis–

acidolysis steps. As the main contaminant, the defective C3-

capped monomer showed a pronounced difference in reten-

tion time. Obtained in satisfying integrity according to Fig-

ure 4 (c), a finishing RP-HPLC purification of the crude

material yielded our first miniature cMyc-Max tweezer

model.

Conclusions

The potential of cMyc-Max downsizing was noticed

shortly after its discovery in the 1990s.^[26]Peptide minia-

tures as TF mimics might rival engineered protein or oligo-

nucleotide approaches in terms of biostability/availability,

yet outscore small-molecule, intercalator, or minor groove

approaches in terms of target selectivity/generality.^[18]

Prior establishment of a preparatory methodology is

however a prerequisite for experimental evaluation of at-

tractive designs, preferably approached by a library format.

A simple SPPS strategy proved viable for the sequential

generation of basic region peptides at the C12 and C3 posi-

tions of the scaffold module, thereby covalently restraining

the appendages. Accommodated by a key NHAlloc protec-

tant at the pivotal C3 position, both homo- and heterodi-

meric peptidosteroids were assembled by standard Fmoc/

tBu chemistry through automated procedures. Enabled by

the facilitated photolytic release of minute samples at regu-

lar stages, an elaborate monitoring strategy focused on ded-

icated RP-HPLC and MALDI-TOF-MS analysis of side

chain protected intermediates. Comprising recognition se-

quences of the natural GCN4, cMyc, and Max proteins,

batch-wise incorporation of two basic region peptides at

each scaffold position in a 2ϫ2 combination design ef-

ficiently furnished a unique collection of first generation

b(-HLH-)ZIP peptidosteroid models. Non-interfering with

the binding process, the C24 handle might further be used

to introduce measurable or switchable modules, which is de-

sirable for advanced applications in the postgenomic era

with a chemical biology mindset.^[17]

Implementation of the here-reported solid-phase pro-

cedures into a parallel library setup should maximize the

odds to identify recognition tendencies, facilitate the ratio-

nalization of design clues, and accelerate the miniaturiza-

tion of the enigmatic loop-type TF class. Now equipped

with a synthetic platform, current prototype contribution

supplements the paucity of literature precedents, gives a

taste of the many opportunities ahead, and provides a

methodological framework from which systematic refine-

ments can depart.

Experimental Section

Figure 4. The first miniature cMyc-Max peptidosteroid tweezer

model: (a) MALDI-TOF-MS (DHB) and (b) RP-HPLC (C18/

100 Å, 214 nm, gradient 1)^[58]of crude 31, isolated after treatment

with reagent B for ca. 2 h at r.t. + an additional 4 h at 50 °C.

(c) RP-HPLC (C18/100 Å, 214 nm, gradient 1)^[58]and complete

structure of purified 31 (calcd. EM 4443.6, MW 4446.1; C3-NHAc

capped deletion monomer: calcd. EM 2530.5, MW 2532.0).

General Information: All organic solvents and chemical reagents

were acquired from commercial sources and used without further

purification or drying. Extra-dry DMF (N,N-dimethylformamide;

with molecular sieves, H₂O Ͻ50 ppm) was used during manual

couplings and manual Fmoc deprotections. When utilizing this sol-

Eur. J. Org. Chem. 2013, 673–687

www.eurjoc.org

681

D. Verzele, A. Madder

FULL PAPER

vent for resin washing and during robot-assisted automated SPPS,

peptide synthesis grade was used. The same applies for NMP.

HPLC grade quality was employed for all other organic solvents

(involved in, e.g., washing, color testing, Alloc/Boc deprotection,

UV Fmoc determination, NAc capping, and photolysis). H₂O met

the Milli-Q grade standard. DIPEA (N,N-diisopropylethylamine)

was supplied as redistilled (i.e., dry), whereas tetrakis(triphenyl-

phosphane)palladium(0) was 99% pure (Aldrich). TentaGel NH₂

resin (90 μm, manufacturer’s loading: 0.28 mmolg^–1) was obtained

from Merck Novabiochem. All chiral α-amino acids used in this

paper possessed the l-configuration. Throughout this work, N^α-

Fmoc-protected residues with standard acid-sensitive side chain

PGs were used: Asp(OtBu) [D], Glu(OtBu) [E], His(Trt) [H],

Lys(Boc) [K], Asn(Trt) [N], Gln(Trt) [Q], Arg(Pbf) [R], Ser(tBu) [S],

Thr(tBu) [T]. Adapted from Holmes et al., photocleavable linker 2

was synthesized prior to the present work and employed as

such.^[44,45]The same applies to our steroid scaffold building block

1.^[40]

3%). LC-(TIC)-MS analysis (reverse-phase) was performed with

diode array detection by using a Phenomenex Luna C18 (2) 100 Å

column (250ϫ4.6 mm, 5 μ, at 35 °C), with hyphenation to an ESI

single quadrupole MS detector type VL.

A flow rate of

1.0 mLmin^–1was applied. UV detection was done at 214, 254, 280,

310, and 360 nm simultaneously, and mass detection operated in

the positive mode. Through a binary solvent system composed of

(A =) H₂O + HCOOH (0.1%) or H₂O + NH₄OAc (5 mm) and

(B =) ACN as the mobile phase, linear gradient elution was per-

formed by applying gradients similar to the above RP-HPLC pro-

files (yet with shorter 2 min preflushing phase).

It is noted that compound numbers were used interchangeably be-

tween the solid-supported molecules and the liberated counter-

parts. Compounds were released (and analyzed) as C24 carbox-

amides upon UV irradiation of the corresponding resin. Unless

stated otherwise, all analytical data refer to crude (i.e., non-RP-

HPLC-purified) samples, which were, at most, subjected to tritura-

tion with Et₂O. Empirical formulae refer to neutral molecules.

Automated peptide synthesis was performed on a fully-automated

SYRO Multiple Peptide Synthesizer robot, equipped with a vor-

texing unit for the 24-reactor block (MultiSynTech GmbH). Reac-

tions were open to the atmosphere, executed at ambient tempera-

ture and shielded from light. Photolysis was carried out on a small

scale with a 4 W Bioblock Scientific compact UV lamp, set at

365 nm. For large scale cleavage, a 451 W UV ACE glass incorpo-

rated 7225–34 immersion lamp equipped with a Schott WG320 UV

cut-off filter was used. The bead suspension was thereby continu-

ously agitated by N₂bubbling, and the solvent was regularly replen-

ished to prevent direct resin irradiation by complete solvent evapo-

ration. All samples were cleaved at a distance of ca. 1 cm from the

lamp.

Standard Procedure for Small-Scale Photolytic Cleavage: Either

aiming for analytical verification/monitoring or proceeding

towards compound isolation, a resin sample (1–5 mg) was transfer-

red to a miniature glass test tube (600 μL, 35ϫ6 mm Ø) and sus-

pended in an appropriate solvent (4–8 drops), usually ACN or

EtOH. The tube was flushed with argon and sealed with a septum.

Placed near-horizontally at a distance of ca. 1 cm from the small-

scale UV lamp (365 nm), the beads were irradiated for 3 h, with

occasional manual homogenization of the resin suspension. Using

a glass syringe with a narrow bore cemented needle, the resulting

solution was carefully separated from the resin, ready for further

manipulation or analysis.

Synthesis of Resin-Bound Steroid Scaffold Starting Constructs

RP-HPLC analysis and purification was performed with diode ar-

ray detection by using a Phenomenex Luna C18 (2) 100 Å column

(250ϫ4.6 mm, 5 μ, at 35 °C) or a Phenomenex Jupiter C4/300 Å

column (250ϫ4.6 mm, 5 μ, at 35 °C) by applying a flow rate of

1.0 mLmin^–1. Signals at 214, 254, 280, 310, and 360 nm were simul-

taneously detected. Through a binary solvent system composed of

(A =) H₂O + TFA (0.1%) and (B =) ACN as the mobile phase,

linear gradient elution was performed: after injection (and shown

in the chromatographic output) the column was flushed with x%

B for 3 min, followed by a linear increase in B (vs. A) to 100% in

y min, finishing by flushing with 100% B for 5 min, after which

the gradient was returned to x% B in 0.5 min, concluding the cycle

by flushing with x% B for 3 min. Gradient 1 refers therein to (x,y)

= (0, 15), or a 0 to 100% linear increase in B (vs. A) over 15 min.

Gradient 2 refers therein to (x,y) = (75, 15), or a 75 to 100% linear

increase in B (vs. A) over 15 min. In addition to these standard

gradients, all others follow the same profile. For brevity, only the

linear increase in B vs. A is quoted in the remainder of this manu-

script. ESI-MS spectra were recorded with a quadrupole ion trap

LC mass spectrometer, equipped with electrospray ionization.

MeOH/H₂O (4/1Ϯ0.1% formic acid) was used as carrier solution.

All data were collected in the positive mode, at a capillary tempera-

ture of 250 °C. MALDI-TOF-MS spectra were acquired with a

high-performance nitrogen laser (337 nm), using the positive and

reflectron mode with delayed extraction. All measurements were

calibrated against MePEOH (M_n≈ 2000, PD = 1.06), spotted from

a MeOH (2 mgmL^–1) solution. The following matrix solutions were

utilized (made in microtubes, stored in freezer, carefully defrosted

and homogenized upon use): DHB: 2,5-dihydroxybenzoic acid

(98.0% pure, 10 mg) + ACN (500 μL) + H₂O (470 μL) + TFA_aq.

(30 μL, 3%); α-CHCA: α-cyano-4-hydroxycinnamic acid (99%

Coupling of Photocleavable Linker 2 to TentaGel NH₂Resin Ǟ Con-

struct 3: Weighed in a fritted glass reactor, TentaGel NH₂resin

(1.0 g, manufacturer’s loading: 0.28 mmolg^–1, 0.28 mmol) was pre-

swollen in dry DMF (10 mL) for 1 h. After filtration under reduced

pressure, the beads were resuspended in dry DMF (4 mL) and pho-

tocleavable linker 2 (437.3 mg, 3 equiv.) was added, followed by dry

DIPEA (280 μL, 6 equiv.) and dry DMF (3 mL). Upon addition of

PyBOP (437.3 mg, 3 equiv.) and dry DMF (3 mL), the reaction ves-

sel was flushed with argon, wrapped in foil, and agitated for 3 h at

room temperature. During agitation, complete dissolution of the

photocleavable linker was observed, which generated an intense

orange mixture. Excess amount of the reagents and the solvent were

removed by filtration under reduced pressure, and the resin was

extensively washed with DMF, ACN, and DCM. After an ad-

ditional DMF washing step, the reaction was repeated, agitating

for 5 h. TNBS test:^[47]colorless (pale orange after ca. 30–60 min as

a result of test-induced NHFmoc deprotection); NF31 test:^[48]pale

pinkish-colorless (TentaGel NH₂starting resin: either intense

orange or intense red).

NHFmoc Deprotection of Construct 3: After an initial DMF wash-

ing step, resin 3 (0.28 mmol) was successively treated for 2, 5, and

15 min with a piperidine solution in DMF (20% v/v, 6 mL) at am-

bient temperature by applying intermediate filtration under re-

duced pressure and washing with DMF. The final resin was ad-

ditionally washed with ACN and DCM. TNBS test:^[47]intense

orange (coloration delayed as a result of hindrance by the α-methyl

group); NF31 test:^[48]intense dark red.

Coupling of Steroid Scaffold 1 to Solid-Supported Photocleavable

Linker Ǟ Compound 4: The resulting construct (0.28 mmol), pre-

pure, 10 mg) + ACN (500 μL) + H₂O (400 μL) + TFA_aq.(100 μL, swollen with an additional washing step with dry DMF, was sus-

682 Eur. J. Org. Chem. 2013, 673–687

www.eurjoc.org

Synthetic Progress in cMyc-Max Oncoprotein Miniaturization

pended in dry DMF (6 mL). Steroid scaffold (193.3 mg,

1

NHFmoc Deprotection of 6–7 Ǟ Compounds 8–9: Before automated

peptide synthesis, resins 6 and 7 were manually NHFmoc depro-

tected by using the procedure described above. After deprotection,

the resins were dried under high vacuum. TNBS test:^[47]rapidly

intense orange/reddish; NF31 test:^[48]rapidly intense dark red. A

small sample was photolytically cleaved (365 nm) for analysis. For

8: ESI-MS [EM calcd. for C₅₂H₆₉N₅O₅843.5 (MW 844.1)]: m/z (%)

= 844.4 (100) [M + H]⁺, 867.6 (21) [M + Na]⁺, 243.2 (60) [Trt]⁺,

1688.0 (9) [2M + H]⁺. RP-HPLC (Luna C18/100 Å, gradient 1): t_R

= 18.4 min. For 9: ESI-MS [EM calcd. for C₅₁H₆₇N₅O₅829.5 (MW

830.1)]: m/z (%) = 830.4 (100) [M + H]⁺, 852.5 (52) [M + Na]⁺,

243.2 (79) [Trt]⁺, 1659.3 (6) [2M + H]⁺, 1672.7 (4) [2M + Na]⁺.

RP-HPLC (Luna C18/100 Å, gradient 1): t_R= 18.4 min.

1.2 equiv.) and dry DIPEA (115 μL, 2.4 equiv.) were added, fol-

lowed by dry DMF (3 mL). Addition of PyBOP (175.0 mg,

1.2 equiv.) and dry DMF (3 mL) completed the reaction mixture.

Rapid dissolution of the solid compounds was observed, and the

argon-flushed vessel was gently agitated and shielded from light.

Upon overnight reaction at room temperature, the excess amounts

of the reagents and the solvent were removed under reduced pres-

sure, and the solid-phase beads were extensively washed with DMF,

ACN, and DCM. Single overnight coupling was followed by a pre-

cautionary N-capping step: the resulting resin was washed with

CHCl₃as a preswelling step, followed the by addition of CHCl₃

(8 mL) and 1-AcIm (308.5 mg, 10 equiv.). The argon-flushed reac-

tion mixture was gently agitated overnight at room temperature

and shielded from light. The solution was discarded by filtration

under reduced pressure, and the resin was extensively washed with

CHCl₃. TNBS^[47]and NF31 test:^[48]colorless (executed before cap-

ping). A small sample was photolytically cleaved (365 nm) for

analysis. ESI-MS [E(xact) M(ass) calcd. for C₃₃H₅₅N₃O₅573.4

(M(ol.)W(t.) 573.8)]: m/z (%) = 474.2 (100) [M – Boc]⁺, 596.2 (28)

[M + Na]⁺, 573.7 (10) [M + H]⁺, 1146.6 (2) [2M + H]⁺.

Robot-Assisted SPPS of Side Chain Protected b(-HLH-)ZIP Pep-

tidosteroids

Automated Synthesis of (Side Chain Protected) C12 Peptidosteroid

Intermediates: Compounds 8–9 Ǟ Compounds 10–11: Resins 8

(30.4 mg, 0.07 mmolg^–1, 2.1 μmol scale) and

9

(30.1 mg,

0.1 mmolg^–1, 3 μmol scale) were transferred to smaller (2 mL) sy-

ringe reactors. Experimental loadings of these resins were calcu-

lated from the actual values of UV-Fmoc quantified precursors 6

and 7 (see data above), translating to 0.07 and 0.1 mmolg^–1for 8

and 9, respectively. A detailed overview of the automated cycle pro-

gram is included in the Supporting Information (Table S3). In sum-

mary, after 15 min of preswelling in NMP (312 μL), the resin beads

were filtered, and the desired peptide sequences were generated

from the C to N terminus through a repetitive coupling–deprotec-

tion cycle by using the Fmoc/tBu SPPS methodology: NMP solu-

tions of N^α-Fmoc-protected residues (0.5 m), DIPEA (0.9 m), and

HBTU (0.5 m, + 5% DMF) were sequentially added in equal vol-

umes (104 μL each) to resins 8 and 9. The resulting coupling mix-

tures (312 μL, 0.16 m of active species, 25 and 17.5 equiv. respec-

tively, + 2% DMF) were allowed to react for a 3-h single coupling

period, with gentle vortexing at regular intervals, after which the

excess amount of the reagents and the solvent were removed under

reduced pressure, and the resins were extensively washed with NMP

(9ϫ312 μL). Translated from the manual procedure described

above, subsequent automated NHFmoc deprotection was effected

by successive treatment (2, 5, and 15 min) of the resin beads with

a piperidine solution in NMP (40% v/v, 312 μL), with filtration

under reduced pressure and extensive washing with NMP

(6ϫ312 μL) after every deprotection cycle. These coupling–depro-

tection steps were repeated until completion. In addition to the

intermediate samples taken for the semi-online monitoring pur-

poses as described, to evaluate the complete sequence a small sam-

ple was photolytically cleaved (365 nm) for analysis. For 10:

MALDI-TOF-MS [(α-CHCA): EM calcd. for C₂₇₈H₄₁₂N₄₈O₅₆S₆

5510.9 (MW 5514.9)]: m/z = 5537.6 [M + Na]⁺, 5553.5 [M + K]⁺.

RP-HPLC (Jupiter C4/300 Å, gradient 1): t_R= 23.2 min. For 11:

MALDI-TOF-MS [(α-CHCA): EM calcd. for C₂₈₅H₃₆₉N₄₁O₄₃S₅

5213.7 (MW 5217.6)]: m/z = 5238.7 [M + Na]⁺, 5255.8 [M + K]⁺.

RP-HPLC (Jupiter C4/300 Å, gradient 1): t_R= 23.1 min.

C12 NHBoc Deprotection of 4 Ǟ Compound 5: Following an initial

washing/preswelling step with DCM, resin 4 (0.28 mmol) was suc-

cessively treated with a TFA solution in DCM (20% v/v, 6 mL) for

2, 5, and 15 min at room temperature. DCM was used for interme-

diate washing, and the final resin was further purified by consecu-

tive rinsing with DCM, DMF, DIPEA/DMF (10% v/v), DMF,

ACN, DCM, and Et₂O, removed by filtration under reduced pres-

sure. The resin beads were thoroughly dried under high vacuum.

TNBS test:^[47]light orange after prolonged reaction (significant

hindrance); NF31 test:^[48]intense red. A small sample was photolyt-

ically cleaved (365 nm) for analysis. ESI-MS [EM calcd. for

C₂₈H₄₇N₃O₃473.4 (MW 473.7)]: m/z (%) = 474.3 (100) [M + H]⁺,

931.5 (6) [2M – NH₃]⁺, 457.2 (4) [M – NH₃]⁺. LC-TIC-MS [Luna

C18/100 Å, gradient 1 (A = HCOOH_aq.)]: t_R= 11.8 min.

Coupling of the First Residue to the C12 Position of 5 Ǟ Compounds

6–7: Two portions of scaffold-bearing resin 5 (50 mg each, theoreti-

cal loading calcd. from TentaGel: 0.23 mmolg^–1, 0.0115 mmol)

were weighed in fritted syringe reactors (5 mL), washed with dry

DMF, filtered under reduced pressure, and resuspended in dry

DMF (1 mL). Dry DIPEA (8.0 μL, 4 equiv.), either NHFmoc-pro-

tected Gln(Trt) or Asn(Trt) (28.1 or 27.4 mg, respectively, 4 equiv.)

and PyBOP (23.9 mg, 4 equiv.) were consecutively added, observing

rapid dissolution of the latter. The reactors were flushed with argon

and wrapped in foil. After vortexing the reaction mixtures at room

temperature for 4 h, the excess amounts of the reagents and the

solvent were removed under reduced pressure, and the resins were

washed with DMF, ACN, and DCM. After an additional washing

step with dry DMF, the couplings were repeated. TNBS test:^[47]

colorless (pale orange after ca. 30–60 min, but biased by the hin-

drance in the starting material on the one hand, yet TNBS test

induced premature NHFmoc deprotection on the other hand);

NF31 test:^[48]pale pinkish-colorless. UV Fmoc^[70]estimated load-

ing: For 6: 0.072 mmolg^–1; for 7: 0.096 mmolg^–1. A small sample

was photolytically cleaved (365 nm) for analysis. For 6: ESI-MS

[EM calcd. for C₆₇H₇₉N₅O₇1065.6 (MW 1066.4)]: m/z (%) = 1066.2

Manual Incorporation of the N-Terminal Acetyl Cap at the C12

Strand of (Side Chain Protected) Peptidosteroid Intermediates 10–

11 Ǟ Compounds 12–13: In the same tubes (2 mL) employed during

the automated peptide assembly, resin 10 (2.1 μmol) and 11

(23) [M + H]⁺, 1088.5 (77) [M + Na]⁺, 1105.6 (15) [M + K]⁺, 243.2 (3.0 μmol) were suspended in CHCl₃(200 μL each) and 1-AcIm

(100) [Trt]⁺. LC-MS [Luna C18/100 Å, gradient 2 (A

=

(2.4 and 3.3 mg, respectively, 10 equiv.) was added. Flushed with

argon and wrapped in foil, the reactors were agitated overnight at

room temperature. The excess amounts of the reagent and the sol-

vent were removed under reduced pressure, and the resins were

washed with CHCl₃, DMF, MeOH, and DCM. A small sample

NH₄OAc_aq.)]: t_R= 16.6 min. RP-HPLC (Luna C18/100 Å, gradi-

ent 2): t_R= 19.1 min. For 7: ESI-MS [EM calcd. for C₆₆H₇₇N₅O₇

1051.6 (MW 1052.4)]: m/z (%) = 1052.0 (34) [M + H]⁺, 1074.6 (50)

[M + Na]⁺, 1090.6 (13) [M + K]⁺, 243.2 (100) [Trt]⁺.

Eur. J. Org. Chem. 2013, 673–687

www.eurjoc.org

683

D. Verzele, A. Madder

FULL PAPER

was photolytically cleaved (365 nm) for analysis. For 12: MALDI-

groups at the C3 scaffold position prior to Fmoc removal of the

TOF-MS [(α-CHCA): EM calcd. for C₂₈₀H₄₁₄N₄₈O₅₇S₆5552.9 appended residue and automated elongation of the second strand,

(MW 5557.0)]: m/z = 5304.8 [M – Pbf + H]⁺. For 13: MALDI- thereby securing the original lack of decisive monitoring.

TOF-MS [(α-CHCA): EM calcd. for C₂₈₇H₃₇₁N₄₁O₄₄S₅5255.7

Manual Test Introduction of the Second Residue of the Second

Strand at (Side Chain Protected) Peptidosteroid Intermediate 19 (via

20) Ǟ Compound 21: To assess the influence of steric impediment,

resin-bound compound 19 was manually derivatized with an ad-

(MW 5259.6)]: m/z = 4755.1 [M – 2Pbf + H]⁺, 5007.3 [M – Pbf +

H]⁺, 5029.9 [M – Pbf + Na]⁺, 5283.2 [M + Na]⁺, 5297.4 [M +

K]⁺. RP-HPLC (Jupiter C4/300 Å, gradient 1): t_R= 23.4 min.

ditional residue. First, the NHFmoc protectant was removed as

described to yield N^α-deprotected compound 20. A small sample

was photolytically cleaved (365 nm) for analysis. For 20

(C₃₀₆H₃₈₇N₄₃O₄₄S₅, calcd. EM 5527.8, MW 5531.9): RP-HPLC

(Luna C18/100 Å, gradient 1): t_R= 22.7 min. After an initial wash-

ing step with dry DMF, resulting resin 20 was resuspended in dry

DMF (100 μL) and NHFmoc-protected Arg(Pbf) (2.7 mg,

4 equiv.), dry DIPEA (1.4 μL, 8 equiv.), and PyBOP (2.0 mg,

4 equiv.) were added. The reaction was further homogenized by

additional DMF (20 μL), flushed with argon, shielded from light,

and vortexed for 2.5 h at ambient temperature. The excess amount

of the reagents and the solvent were removed under reduced pres-

sure, and the reaction was repeated (5 h). A small sample was pho-

tolytically cleaved (365 nm) for analysis. For 21 (C₃₄₀H₄₂₅N₄₇O₅₀S₆,

calcd. EM 6158.0, MW 6162.7): RP-HPLC (Jupiter C4/300 Å, gra-

dient 1): t_R= 23.6 min.

Manual Removal of the C3 NHAlloc Protecting Group in (Side

Chain Protected) Peptidosteroid Intermediates 12–13 Ǟ Compounds

14–15: Prior to the reaction, resins 12 and 13 were additionally

washed with DCM and Et₂O, followed by careful drying under

high vacuum. Experimental loadings of these resins were calculated

from the actual values of UV-Fmoc quantified precursors 6 and 7

(see data above), translating to 0.053 and 0.07 mmolg^–1for 12 an

13, respectively. Progressing towards our 2ϫ2 solid-phase library

setup, these resins were both divided into two fritted syringe reac-

tors to obtain two batches of each resin (12: 20.6 mg, 1.1 μmol and

11.9 mg, 0.63 μmol; 13: 14.7 mg, 1.0 μmol and 25.6 mg, 1.8 μmol).

As illustrated for the third batch, the following chemicals were se-

quentially added: DCM (100 μL), morpholine (8.1 μL, 90 equiv.),

Bu₃SnH (2.7 μL, 10 equiv.), Pd(PPh₃)₄(1.3 mg, 1 equiv.), and

DCM (150 μL). The resulting yellow-orange reaction mixtures were

flushed with argon, shielded from light and vortexed at room tem-

perature for 2 h. The excess amounts of the reagents and the sol-

vent were removed by filtration under reduced pressure, and the

resins were thoroughly washed with DCM, DMF, MeOH, and

DCM. TNBS test:^[47]intense orange/red after ca. 30 min. A small

sample was photolytically cleaved (365 nm) for analysis. For 14:

Manual Deprotection of the NHFmoc Protecting Group of (Side

Chain Protected) Peptidosteroid Intermediates 16–18 and 21 Ǟ

Compounds 22–24 and 25: Before automated assembly of the sec-

ond peptide strand at the C3 position of resins 22–24 and 25,

NHFmoc deprotection of 16–18 and 21, respectively, was manually

performed, by adopting the above-described procedure. A small

sample was photolytically cleaved (365 nm) for analysis. For 24

(C₂₉₁H₃₈₀N₄₂O₄₅S₅, calcd. EM 5342.7, MW 5346.7): RP-HPLC

(Luna C18/100 Å, gradient 2): t_R= 19.3 min. For 25: An analytical

sample was purified by RP-HPLC for MALDI-TOF-MS [(DHB):

EM calcd. for C₃₂₅H₄₁₅N₄₇O₄₈S₆5936.0 (MW 5940.4)]: m/z =

5734.9 [M – Trt + K]⁺, 5719.2 [M – Trt + Na]⁺, 5697.3 [M – Trt

+ H]⁺, 5444.0 [M – Trt – Pbf + H]⁺. RP-HPLC (Luna C18/100 Å,

gradient 2): t_R= 20.6 min.

MALDI-TOF-MS [(DHB): EM calcd. for

C₂₇₆H₄₁₀N₄₈O₅₅S₆

5468.9 (MW 5472.9)]: m/z = 5494.1 [M + Na]⁺, 5510.1 [M + K]⁺,

5218.0 [M – Pbf + H]⁺. RP-HPLC (Luna C18/100 Å, gradient 1):

t_R= 23.1 min. For 15: MALDI-TOF-MS [(DHB): EM calcd. for

C

₂₈₃H₃₆₇N₄₁O₄₂S₅5171.6 (MW 5175.5)]: m/z = 5196.7 [M + Na]⁺,

5214.5 [M + K]⁺, 4944.3 [M – Pbf + Na]⁺, 4920.5 [M – Pbf +

H]⁺. RP-HPLC (Luna C18/100 Å, gradient 1): t_R= 22.2 min.

Manual Introduction of the First Residue of the Second Strand at the

Available C3 NH₂Group in (Side Chain Protected) Peptidosteroid

Intermediates 14 and 15 Ǟ Compounds 16–17 and 18–19: The ob-

tained NHAlloc-deprotected resins were washed with dry DMF

prior to reaction. Again illustrated for the above batch (towards

19), the resin beads were resuspended in dry DMF (60 μL), fol-

lowed by consecutive addition of dry DIPEA (0.7 μL, 4 equiv.),

NHFmoc-protected Asn(Trt) (2.4 mg, 4 equiv.), PyBOP (2.0 mg,

4 equiv.), and dry DMF (60 μL). The solid reagents readily dis-

solved, and the reaction mixture was flushed with argon, shielded

from light, and vortexed at room temperature for 2 h. After the

reaction, the excess amounts of the reagents and the solvent were

removed by filtration under reduced pressure, and the resin was

rinsed with DMF, MeOH, and DCM. The procedure was repeated

(reaction for 3 h), yet incomplete reaction was observed upon

analysis, even after further repetition. Although the inferior analyt-

ical outcome of the other batches prevented conclusive interpre-

tation, diagnostic results were obtained for 19, as discussed above

and evidenced in the Supporting Information. Fortunately,

satisfying conversion was obtained by a similar 2-h (single) treat-

ment with HATU as the coupling reagent (1.6 mg, 4 equiv.). A

small sample was photolytically cleaved (365 nm) for analysis. For

19: MALDI-TOF-MS [(DHB): EM calcd. for C₃₂₁H₃₉₇N₄₃O₄₆S₅

5749.9 (MW 5754.2)]: m/z = 5776.4 [M + Na]⁺, 5792.3 [M + K]⁺.

RP-HPLC (Jupiter C4/300 Å, gradient 1): t_R= 23.5 min. In con-

trast to resin 19, which was further derivatized as presented below,

resins 16–18 were subjected to precautionary N-capping (adopting

the above-described procedure) of potentially underivatized amino

Automated Synthesis of (Side Chain Protected) C12–C3 Peptido-

steroid Dimers: Compounds 25 and 24–22 Ǟ Compounds 26 and 27–

29: Prior to reaction, resins 25 and 24–22 were additionally washed

with Et₂O, followed by careful drying under high vacuum and

transferred to small (2 mL) syringe reactors. Experimental loadings

of these resins were calculated from the actual values of UV-Fmoc

quantified precursors 7 and 6 (see data above), translating to 0.067

and 0.051 mmolg^–1for 24–25 and 22–23, respectively. These resins

were subjected to the same automated procedure as described

above (0.16 m of active species, 3-h single couplings), with scaled

conditions (0.2–1.4 μmol scale). A detailed overview of the auto-

mated cycle program is included in the Supporting Information

(Table S6).

Photolysis–Acidolysis towards Final Deprotected Peptidosteroids

Preparation of Final Analytical Samples of Side Chain Deprotected

Peptidosteroids 30–33: Prior to UV cleavage, resins 26–29 were

washed with Et₂O and dried under high vacuum. As described

above, the resins (3–4 mg) were photolytically cleaved (3 h) in ACN,

with occasional swirling of the resin suspensions. The resulting

opaque whitish solutions were transferred to microtubes (1.5 mL),

aided by MeOH and sonication. The solutions were evaporated

under reduced pressure. The yellowish/white, oil-like residues were

further dried under high vacuum. Reagent B (88% v/v TFA + 5%

H₂O + 5% phenol + 2% TIS, 100 μL) was added to each residue,

and the microtubes were sonicated. Appearance of an intense yel-

684

www.eurjoc.org

Eur. J. Org. Chem. 2013, 673–687

Synthetic Progress in cMyc-Max Oncoprotein Miniaturization

low color was rapidly observed. The samples were vortexed at room

temperature for ca. 2 h. The mixtures were evaporated under re-

duced pressure and Et₂O (20 drops) was added (at room tempera-

ture) to the resulting orange residues, readily observing floccu-

lation. After sonicating, intensely vortexing, and reflocculation, the

microtubes were centrifuged (20 °C, 10 min, 4500 rpm). The clear

supernatants were carefully removed with a narrow bore cemented

glass syringe and the Et₂O washing was repeated. Residual Et₂O

was removed by gentle argon flushing. The white-greyish solids

were dissolved in MeOH (100 μL, sonication) and analysis demon-

strated incomplete removal of side chain PGs as discussed. There-

fore, the solutions were evaporated under reduced pressure and the

acidolysis procedure was repeated, yet partially deprotected C12-

cMyc–C3-Max 31 was heated at 50 °C for 4 h, whereas partially

deprotected cMyc-cMyc 30, GCN4-cMyc 32, and GCN4-Max 33

were heated at 60 °C for 2.5 h. Satisfying results were obtained for

compound 31, 32, and 33, whereas compound 30 was sacrificed

along the experimental optimizations. For 31: Vide infra. For 32:

MALDI-TOF-MS [(DHB): EM calcd. for C₂₁₈H₃₈₆N₈₆O₅₆5105.0

(MW 5107.9)]: m/z = 5107.1 [M + H]⁺, 5090.7 [M – H₂O + H]⁺.

RP-HPLC (Luna C18/100 Å, gradient 1): t_R= 12.5 min. For 33:

MALDI-TOF-MS [(DHB): EM calcd. for C₂₁₄H₃₇₄N₈₃O₅₆5002.9

(MW 5005.8)]: m/z = 5003.7 [M + H]⁺, 4986.4 [M – H₂O + H]⁺.

RP-HPLC (Luna C18/100 Å, gradient 1): t_R= 12.5 min.

ear increase of B vs. A from 0 to 100% over 30 min, 214 nm: t_R=

15.7 min). For 31: MALDI-TOF-MS [(DHB): EM calcd. for

C₁₉₀H₃₂₉N₇₆O₄₈4443.6 (MW 4446.1)]: m/z = 4445.1 [M + H]⁺,

4428.2 [M – H₂O + H]⁺. RP-HPLC (Luna C18/100 Å, gradient 1):

t_R= 12.3 min.

Supporting Information (see footnote on the first page of this arti-

cle): Details of the robot-assisted SPPS protocols, additional infor-

mation of the monitoring/analytical efforts, and supporting data

for compound evaluation (including overview tables, RP-HPLC

chromatograms, and ESI/MALDI-TOF mass spectra).

Acknowledgments

This research was financially supported by Ghent University (BOF

011D16403)

and

the

Research

Foundation

Flanders

(FWO KAN 1.5.186.03; G.0347.04N). We gratefully thank

Prof. Dr. José L. Mascareñas, Dr. M. Eugenio Vázquez, and Dr.

Olalla Vázquez (University of Santiago de Compostela, Spain) for

useful discussions during a short research stay in their laboratory.

Suggestions from Prof. Takashi Morii (Kyoto University, Japan)

are further acknowledged. Technical assistance by Jan L. Goeman

is appreciated. Noël Hosten and Jos Van den Begin are thanked for

the synthesis of the photocleavable linker.

Scaled Preparation of Final Side Chain Deprotected cMyc-Max

Peptidosteroid 31: As discussed above, further emphasis was put on

cMyc-Max model 31, for which a scaled photolysis–acidolysis was

performed. Prior to reaction, resin 27 was additionally washed with

DCM and Et₂O, followed by careful drying under high vacuum.

The resin was transferred to a small fritted glass reactor. The exper-

imental loading of this resin was calculated from the actual value

of UV-Fmoc quantified precursor 7 (see data above), translating to

0.054 mmolg^–1, and cMyc-Max bearing resin 27 (18.2 mg,

0.983 μmol) was suspended in EtOH (2 mL). Left for 15 min to

allow preswelling, this resin was irradiated by the large-scale UV

lamp for 3.5 h, with N₂bubbling for agitation. At regular intervals

(ca. 1 h), EtOH (1 mL) was added to prevent complete evaporation.

The resulting solution and combined washings (EtOH) were col-

lected in a flask (100 mL), obtaining a yellowish oil-like residue

upon evaporation. The procedure was repeated twice (2.5 and

3.5 h), and the batches were isolated in distinct flasks. Redissolved

in MeOH, the isolates were transferred to smaller flasks (10 mL)

and the solutions were evaporated. The resulting residues were

dried under high vacuum. Flushed with argon and magnetically

stirred, these batches were overnight treated with reagent B (1 mL

each) at 60 °C, showing rapid dissolution of the residues. The mix-

tures were cooled to room temperature, toluene was added, and

the solutions were evaporated under reduced pressure. The orange-

brownish residues were further dried under high vacuum, and Et₂O

was added at room temperature. The mixtures were sonicated, vor-

texed, and transferred to centrifuge tubes with additional Et₂O (to-

tal of 9 mL each), resulting in flocculation of white-greyish precipi-

tates. The tubes were centrifuged (23 °C, 10 min, 5000 rpm), the

supernatant was discarded, and the Et₂O trituration was repeated

(2 mL). Redissolved in MeOH (aided by sonication), the obtained

material was transferred in the original flasks (10 mL), evaporated

under reduced pressure, and the solid residues were further dried

under high vacuum. Crude, deprotected peptidosteroid 31 was ob-

tained in 62% isolated yield (respectively 1.1 mg + 1.3 mg + 1.4 mg

= 3.8 mg, 0.606 μmol, as TFA salt: 4446.1 + 16 TFA = MW 6270.4)

over the photolysis–acidolysis steps. Upon analysis, the batches

were dissolved in MeOH and pooled, proceeding towards RP-

HPLC purification of an analytical sample (Luna C18/100 Å, lin-

[1] Selected reviews: a) P. P. Pandolfi, Oncogene 2001, 20, 3116; b)

L. H. Hurley, Nat. Rev. Cancer 2002, 2, 188; c) J. E. Darnell Jr.,

Nat. Rev. Cancer 2002, 2, 740; d) M. S. Redell, D. J. Tweardy,

Drug Discovery Today Technol. 2006, 3, 261; e) P. Brennan, R.

Donev, S. Hewamana, Mol. Biosyst. 2008, 4, 909.

[2] Which consists of related isoforms such as c-Myc (cellular), v-

Myc (viral), L-Myc (involved in lung carcinoma), and N-Myc

(involved in neuroblastoma).

[3] C. Grandori, S. M. Cowley, L. P. James, R. N. Eisenman, Annu.

Rev. Cell Dev. Biol. 2000, 16, 653.

[4] Selected reading: a) S. Pelengaris, M. Khan, G. Evan, Nat. Rev.

Cancer 2002, 2, 764; b) E. V. Prochownik, Expert Rev. Antican-

cer Ther. 2004, 4, 289; c) M. Vita, M. Henriksson, Semin. Can-

cer Biol. 2006, 16, 318.

[5] While targeting the DNA–protein interface in the present

work, several contributions interfere with the corresponding

protein–protein contacts (particularly) through small-molecule

(i.e., non-peptidic) disruptors, thereby corroborating the rele-

vance of studying TFs while acknowledging the associated dif-

ficulties. Selected reviews: a) D. L. Boger, J. Desharnais, K.

Capps, Angew. Chem. 2003, 115, 4270; Angew. Chem. Int. Ed.

2003, 42, 4138; b) W. C. Tse, D. L. Boger, Chem. Biol. 2004, 11,

1607; c) H. Arndt, Angew. Chem. 2006, 118, 4664; Angew.

Chem. Int. Ed. 2006, 45, 4552; d) A. N. Koehler, Curr. Opin.

Chem. Biol. 2010, 14, 331; e) S. J. Metallo, Curr. Opin. Chem.

Biol. 2010, 14, 481.

[6] L. E. Canne, A. R. Ferré-D’Amaré, S. K. Burley, S. B. H. Kent,

J. Am. Chem. Soc. 1995, 117, 2998.

[7] T. Morii, M. Simomura, S. Morimoto, I. Saito, J. Am. Chem.

Soc. 1993, 115, 1150.

[8] The basic region, HLH region, and leucine zipper domain co-

exist in different combinations in a total of four folds. b-HLH

and HLH-ZIP proteins complement the bZIP and b-HLH-ZIP

types, with striking parallels between all of them. The b-HLH-

ZIP proteins are closely related to the network of b-HLH coun-

terparts, either being classified as class C vs. class A/B members

in an HLH-focusing taxonomy (and HLH-ZIP = class D) or

being allocated to the alternative zipper-type group: a) N. M.

Luscombe, S. E. Austin, H. M. Berman, J. M. Thornton, Ge-

nome Biol. 2000, 1, 1; b) S. Jones, Genome Biol. 2004, 5, 1.

[9] a) C. O. Pabo, L. Nekludova, J. Mol. Biol. 2000, 301, 597; b)

N. M. Luscombe, R. A. Laskowski, J. M. Thornton, Nucleic

Acids Res. 2001, 29, 2860.

Eur. J. Org. Chem. 2013, 673–687

www.eurjoc.org

685

D. Verzele, A. Madder

FULL PAPER

[10]

[23]

[24]

The overall dimensions of an α-helix allow it to fit into the

major groove in a number of ways. This variability explains the

preference of natural proteins for targeting this wider groove

over the narrower minor one. It accordingly accounts for the

contrasting prevalence of minor groove targeting mimics, as

their less ambiguous binding mode facilitates the rationaliza-

tion of recognition/design principles. The synthetic polyamides

of Dervan et al. are prominent examples thereof. A notable ex-

ception are the (mostly, if not all, bioengineered) zinc finger

proteins binding in the major groove (with Pabo et al. as pro-

tagonist), for which an advanced level of insight and pairing

control has been achieved. Selected reviews: a) P. B. Dervan,

B. S. Edelson, Curr. Opin. Struct. Biol. 2003, 13, 284; b) S.

Negi, M. Imanishi, M. Matsumoto, Y. Sugiura, Chem. Eur. J.

2008, 14, 3236; c) A. Klug, Annu. Rev. Biochem. 2010, 79, 1.

a) K. M. Ryan, G. D. Birnier, Biochem. J. 1996, 314, 713; b)

V. H. Cowling, M. D. Cole, Semin. Cancer Biol. 2006, 16, 242.

a) R. Gurlie, T. H. Duong, K. Zakrzewska, Biopolymers 1999,

49, 313; b) R. L. Winston, J. M. Gottesfeld, Chem. Biol. 2000,

7, 245; c) S. K. Nair, S. K. Burley, Cell 2003, 112, 193; d) L.

Ma, Y. Y. Sham, K. J. Walters, H. C. Towle, Nucleic Acids Res.

2007, 35, 35.

Reviewed in: a) S. Sato, M. Hagihara, K. Sugimoto, T. Morii,

Chem. Eur. J. 2002, 8, 5066; b) M. E. Vázquez, A. M.

Caamaño, J. L. Mascareñas, Chem. Soc. Rev. 2003, 32, 338;

c) E. Pazos, J. Mosquera, M. E. Vázquez, J. L. Mascareñas,

ChemBioChem 2011, 12, 1958.

Selected reading: a) N. Jones, Cell 1990, 61, 9; b) R. Fairman,

R. K. Beran-Steed, S. J. Anthony-Cahill, J. D. Lear, W. F. Staf-

ford III, W. F. DeGrado, P. A. Benfield, S. L. Brenner, Proc.

Natl. Acad. Sci. USA 1993, 90, 10429; c) C. Park, J. L.

Campbell, W. A. Goddard III, J. Am. Chem. Soc. 1996, 118,

4235; d) S. J. Metallo, A. Schepartz, Nat. Struct. Mol. Biol.

1997, 4, 115; e) S. Sauvé, J.-F. Naud, P. Lavigne, J. Mol. Biol.

2007, 365, 1163; f) O. Ecevit, M. A. Khan, D. J. Goss, Biochem-

istry 2010, 49, 2627.

Recent examples: a) E. K. Liebler, U. Diederichsen, Org. Lett.

2004, 6, 2893; b) M. Zhou, I. Ghosh, Org. Lett. 2004, 6, 3561;

c) A. C. Beltran, P. E. Dawson, J. M. Gottesfeld, Chem-

BioChem 2005, 6, 104; d) J. Svobodová, C. Cabrele, Chem-

BioChem 2006, 7, 1164; e) O. Vázquez, M. E. Vázquez, J. B.

Blanco, L. Castedo, J. L. Mascareñas, Angew. Chem. 2007, 119,

7010; Angew. Chem. Int. Ed. 2007, 46, 6886; f) O. Vázquez,

J. B. Blanco-Canosa, E. M. Vázquez, J. Martínez-Costas, L.

Castedo, J. L. Mascareñas, ChemBioChem 2008, 9, 2822; g) J.

Xu, G. Chen, A. T. De Jong, S. H. Shahravan, J. A. Shin, J.

Am. Chem. Soc. 2009, 131, 7839; h) Y. Azuma, M. Imanishi,

T. Yoshimura, T. Kawabata, S. Futaki, Angew. Chem. 2009,

121, 6985; Angew. Chem. Int. Ed. 2009, 48, 6853; i) S. Pensato,

M. Renda, F. Leccia, M. Saviano, L. D. D’Andrea, C. Pedone,

P. V. Pedone, A. Romanelli, Biopolymers 2010, 93, 434; j) R. S.

Tu, R. Marullo, R. Pynn, R. Bitton, H. Bianco-Peled, M. V.

Tirrell, Soft Matter 2010, 6, 1035.

Minimalist strategy: as an alternative to dimerization ap-

proaches, the net unfavorable binding free energy of the down-

sizing process is compensated by preorganization (Ala-substi-

tution, “stapling”, avian pancreatic polypeptide grafting) or

conjugation to an accessory module (Dervan peptide, inter-

calator) assisting in DNA binding. Further reading in addition

to ref.^{[13,14,23b,23c,25e–25g]}, focusing on Myc: a) D. E. Fisher,

L. A. Parent, P. A. Sharp, Cell 1993, 72, 467; b) C. Takemoto,

D. E. Fisher, Gene Expression 1995, 4, 311.

[25]

[11]

[12]

[13]

[14]

E. C. Turner, C. H. Cureton, C. J. Weston, O. S. Smart, R. K.

Alleman, Chem. Biol. 2004, 11, 69.

L. Guerrero, O. S. Smart, C. J. Weston, D. C. Burns, G. A.

Woolley, R. K. Allemann, Angew. Chem. 2005, 117, 7956; An-

gew. Chem. Int. Ed. 2005, 44, 7778.

Interestingly, studies focusing on the leucine zipper region of

b-HLH-ZIP proteins further indicate the inferior stability of

their dimerizing interfaces, in contrast to the bZIP counter-

parts Jun, Fos, and GCN4: C. Muhle-Goss, T. Gibson, P.

Schuck, D. Schubert, D. Nalis, M. Nilges, A. Pastore, Biochem-

istry 1994, 33, 11296.

In this respect, the heterodimeric Jun-Fos bZIP oncoprotein

bears special mentioning, combining cMyc-like cancer rele-

vance with GCN4-like structural tractability.

D. Verzele, L. L. G. Carrette, A. Madder, Drug Discovery To-

day Technol. 2010, 7, e115.

Employing peptide/protein assemblies as promising alternatives

for small-molecule approaches acting on the DNA–protein

interface (complementing the aforementioned protein–protein

disruptors), the superior control of recognition specificity

(cited above) has already led to success stories with minor

groove binding Dervan polyamides and major groove targeting

zinc finger proteins (further stimulated by their design modu-

larity, synthetic accessibility, and uptake potential). Further

reading: a) T. C. Uil, H. J. Haisma, M. G. Rots, Nucleic Acids

Res. 2003, 31, 6064; b) A. C. Jamieson, J. C. Miller, C. O. Pabo,

Nat. Rev. Drug Discovery 2003, 2, 361; c) P. D. Blancafort, J.

Segal, C. F. Barbas III, Mol. Pharmacol. 2004, 66, 1361; d) K.

Schmitz, U. Schepers, Angew. Chem. 2004, 116, 2526; Angew.

Chem. Int. Ed. 2004, 43, 2472; e) J. T. Koh, J. Zheng, ACS

Chem. Biol. 2007, 2, 599; f) X. Cai, P. J. Gray Jr., D. D.

Von Hoff, Cancer Treat. Rev. 2009, 35, 437; g) C. Li, C. Du,

H. Tian, C. Jiang, M. Du, Y. Liu, R.-Z. Qiao, Y.-X. Jia, Y.-F.

Zhao, Chem. Eur. J. 2010, 16, 12935.

a) C. O. Pabo, R. T. Sauer, Annu. Rev. Biochem. 1992, 61, 1053;

b) I. Ghosh, S. Yao, J. Chmielewski, Comprehensive Natural

Products Chemistry (Eds.: D. Barton, K. Nakanishi), Perga-

mon, Oxford, 1999, chap. 7.13, p. 477; c) C. W. Garvie, C. Wol-

berger, Mol. Cell 2001, 8, 937.

a) H. C. Hurst, Protein Profile 1994, 1, 123; b) T. Ellenberger,

Curr. Opin. Struct. Biol. 1994, 4, 12; c) S. K. Nair, S. K. Burley,

Curr. Top. Microbiol. Immunol. 2006, 302, 123.

[15]

[26]

[16]

[17]

[18]

[27]

[28]

R. V. Talanian, C. J. McKnight, P. S. Kim, Science 1990, 249,

769.

Further used by Goddard III et al. with the homodimeric Jun

bZIP: C. Park, J. L. Campbell, W. A. Goddard III, Proc. Natl.

Acad. Sci. USA 1993, 90, 4892.

[29]

[30]

M. Pellegrini, R. H. Ebright, J. Am. Chem. Soc. 1996, 118,

5831.

Selected examples: a) B. Cuenoud, A. Schepartz, Science 1993,

259, 510; b) C. R. Palmer, L. S. Sloan, J. C. Adrian Jr., B.

Cuenoud, D. N. Paolella, A. Schepartz, J. Am. Chem. Soc.

1995, 117, 8899.

Further examples (in addition to ref.^[7,23a]): a) Y. Aizawa, Y.

Sugiura, T. Morii, Biochemistry 1999, 38, 1626; b) A. Hirata,

M. Ueno, Y. Aizawa, K. Ohkubo, T. Morii, S. Yoshikawa,

Bioorg. Med. Chem. 2005, 13, 3107.

[31]

[32]

[33]

A. M. Caamaño, M. E. Vázquez, J. Martínez-Costas, L.

Castedo, J. L. Mascareñas, Angew. Chem. 2000, 112, 3234;

Angew. Chem. Int. Ed. 2000, 39, 3104.

[19]

[20]

Selected reviews: a) P. Wallimann, T. Marti, A. Furer, F. Die-

derich, Chem. Rev. 1997, 97, 1567; b) A. Enhsen, W. Kramer,

G. Wess, Drug Discovery Today 1998, 3, 409; c) R. Maltais,

M. R. Tremblay, L. C. Ciobanu, D. Poirier, J. Comb. Chem.

2004, 6, 443; d) E. Virtanen, E. Kolehmainen, Eur. J. Org.

Chem. 2004, 3385; e) D. B. Salunke, B. G. Hazra, V. S. Pore,

Curr. Med. Chem. 2006, 13, 813; f) A. P. Davis, Molecules 2007,

12, 2106; g) Nonappa, U. Maitra, Org. Biomol. Chem. 2008, 6,

657.

[21]

[22]

T. E. Ellenberger, C. J. Brandl, K. Struhl, S. C. Harrison, Cell

1992, 71, 1223.

a) L. Jen-Jacobsen, L. E. Engler, L. A. Jacobsen, Structure

2000, 8, 1015; b) P. L. Privalov, A. I. Dragan, C. Crane-Robin-

son, K. J. Breslauer, D. P. Remeta, C. A. S. A. Minetti, J. Mol.

Biol. 2007, 365, 1.

[34]

a) R. Boyce, G. Li, H. P. Nestler, T. Suenaga, W. C. Still, J. Am.

Chem. Soc. 1994, 116, 7955; b) Y. Chen, T. Suenaga, W. C. Still,

J. Am. Chem. Soc. 1996, 118, 1813.

686

www.eurjoc.org

Eur. J. Org. Chem. 2013, 673–687

Synthetic Progress in cMyc-Max Oncoprotein Miniaturization

[35]

Guide, W. H. Freeman and Company, New York, 1992, p. 124.

[56] The detrimental effect of erroneous resin irradiation (pro-

longed and/or insufficiently immersed) is worth mentioning, re-

sulting in decomposition of compounds.

[57] Opaque UV-cleaved solutions have additionally been noticed.

[58] Further details can be found in the Experimental Section. In

brief, gradient 1 refers to a 0 to 100% linear increase of B (vs.

A) over 15 min [A = H₂O + TFA (0.1%), B = ACN; flow =

1.0 mLmin^–1]. Pre- and postflushing is included in the chroma-

tographic output.

[59] Desirable given the metabolic lability of cMyc (t_1/2= 20–

30 min).

[60] Recent example thereof (in addition to the aforementioned

ones on bile acids specifically): P. Ingallinella, E. Bianchi, N. A.

Ladwa, Y. Wang, R. Hrin, M. Veneziano, F. Bonelli, T. J.

Ketas, J. P. Moore, M. D. Miller, A. Pessi, Proc. Natl. Acad.

Sci. USA 2009, 106, 5801.

[61] These combinations bear similarities to precedents by Morii

et al. (cited earlier), in which the cMyc-cMyc association re-

sembles the inherently weak MyoD-MyoD model (ref.^[7]),

whereas the unnatural union between the bZIP GCN4 and b-

HLH-ZIP cMyc/Max strands parallels the GCN4-EMBP1

bZIP-bZIP heterodimer (ref.^[31b]).

[62] a) O. Dangles, F. Guibé, G. Balavoine, S. Lavielle, A. Marquet,

J. Org. Chem. 1987, 52, 4984; b) F. Guibé, Tetrahedron 1998,

54, 2967.

[63] a) S. L. Cohen, B. T. Chait, Anal. Chem. 1996, 68, 31; b) K. B.

Green-Church, P. A. Limbach, Anal. Chem. 1998, 70, 5322; c)

G. H. Bird, A. R. Lajmi, J. A. Shin, Anal. Chem. 2002, 74, 219;

d) A. J. Hoteling, W. J. Erb, R. J. Tyson, K. G. Owens, Anal.

Chem. 2004, 76, 5157.

[64] a) As mentioned in: S. Peluso, P. Dumy, C. Nkubana, Y. Yoko-

kawa, M. Mutter J. Org. Chem. 1999, 46, 7114; b) B. D. Larsen,

A. Holm, Int. J. Pept. Protein Res. 1994, 43, 1; c) J. P. Tam, Y.

Lu, J. Am. Chem. Soc. 1995, 117, 12058.

Recent examples from our group, in addition to further pre-

cedents in above reviews (ref.^[33]): a) C. A. Bodé, T. Bechet, E.

Prodhomme, K. Gheysen, P. Gregoir, J. C. Martins, C. P.

Muller, A. Madder, Org. Biomol. Chem. 2009, 7, 3391; b) Figa-

roli, S. A. Madder, Tetrahedron 2010, 66, 6912.

W. Kramer, Biol. Chem. 2011, 392, 77.

L. J. D’Souza, U. Maitra, J. Org. Chem. 1996, 61, 9494.

H. Li, L. Wang, Org. Biomol. Chem. 2003, 1, 3507.

a) S. Yang, S. S. Cho, Y. Levy, M. S. Cheung, H. Levine, P. G.

Wolynes, J. N. Onuchic, Proc. Natl. Acad. Sci. USA 2004, 101,

13786; b) W. K. Meisner, T. R. Sosnick, Proc. Natl. Acad. Sci.

USA 2004, 101, 13478; c) T. Kameda, S. Takada, Proc. Natl.

Acad. Sci. USA 2006, 103, 17765; d) J. Garner, M. M.

Hardling, Org. Biomol. Chem. 2007, 5, 3577.

D. Verzele, A. Madder, Eur. J. Org. Chem. 2007, 11, 1793.

D. Verzele, J. L. Goeman, A. Madder, ARKIVOC 2007, x, 325

(commemorative issue in honor of Prof. Dr. A. Krief).

I. Coin, M. Beyermann, M. Bienert, Nat. Protoc. 2007, 2, 3247.

Remarkably, although (obviously) recommending their novel

native chemical ligation alternative, Kent et al. demonstrate

identical peptide integrity and constitution with a simple, linear

SPPS approach in following example: W. Lu, M. A. Quasim,

S. B. H. Kent, J. Am. Chem. Soc. 1996, 118, 8518.

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

C. P. Holmes, D. G. Jones, J. Org. Chem. 1995, 60, 2318.

[45] C. P. Holmes, J. Org. Chem. 1997, 62, 2370.

[46] It is noted that compound numbers are used interchangeably

between solid-supported molecules and liberated counterparts.

[47] F. Gaggini, A. Porcheddu, G. Reginato, M. Rodriquez, M.

Taddei, J. Comb. Chem. 2004, 6, 805.

[48] a) A. Madder, N. Farcy, N. G. C. Hosten, H. De Muynck, P. J.

De Clercq, J. Barry, A. P. Davis, Eur. J. Org. Chem. 1999, 2787;

b) L. L. G. Carrette, D. Verzele, A. Madder, Tetrahedron Lett.

2010, 51, 2106.

[49] G. Talbo, J. D. Wade, N. Dawson, M. Manoussios, G. W. Tre-

gear, Lett. Pept. Sci. 1997, 4, 121.

[50] A. M. Schaiberger, J. A. Moss, J. Am. Soc. Mass Spectrom.

2008, 19, 614.

[65] M. Rinnová, M. Nováková, V. Kasˇicˇka, J. Jirácˇek, J. Pept. Sci.

2000, 6, 355.

[66] S. M. Sternson, S. L. Schreiber, Tetrahedron Lett. 1998, 39,

7451.

[51] R. V. Talanian, J. McKnight, R. Rutkowski, P. S. Kim, Bio-

chemistry 1992, 31, 6871.

[67] Which proved to be an important step to obtain satisfying floc-

culation upon Et₂O trituration, as further mentioned in: Merck

Novabiochem Catalog 2008/2009. Synthesis Notes, p. 3.30.

[68] Intense signals for PG species could significantly be reduced

by thorough sonication of the precipitated material in Et₂O.

Because these peaks exhibit UV activity, they are most proba-

bly explained by debris of the Pbf and Trt groups. The latter

one (the cation) might be irreversibly converted into tri-

phenylmethane by the potent TIS scavenger. Merck Nova-

biochem Catalog 2008/2009. Synthesis Notes, p. 3.32.

[69] Corroborated by Merck Novabiochem Catalog 2006/2007.

Synthesis Notes, p. 2.54.

[52] All chiral α-amino acids used in this paper possessed the l-

configuration. Throughout this work, N^α-Fmoc-protected resi-

dues with standard acid-sensitive side chain PGs were used:

Asp(OtBu) [D], Glu(OtBu) [E], His(Trt) [H], Lys(Boc) [K],

Asn(Trt) [N], Gln(Trt) [Q], Arg(Pbf) [R], Ser(tBu) [S], Thr(tBu)

[T].

[53] C. K. Taylor, P. W. Abel, M. Hulce, D. D. Smith, J. Pept. Res.

2005, 65, 84.

[54] Hindrance of the DIPEA base provides prolonged (Ͼ3 h) sta-

bility of the N^α-Fmoc protectant: C. D. Chang, M. Waki, M.

Ahmed, J. Meienhofer, E. O. Lundell, J. D. Haug, Int. J. Pept.

Protein Res. 1980, 15, 59.

[70] Protocol adapted from: B. A. Bunin, The Combinatorial Index,

Academic Press, London, UK, 1998.

[55] Efficient couplings indeed critically depend on concentration

of activated species, and a 0.15 m concentration is recom-

mended in: G. A. Grant (Ed.), Synthetic Peptides, A User’s

Received: September 17, 2012

Published Online: December 6, 2012

Eur. J. Org. Chem. 2013, 673–687

www.eurjoc.org

687

Article Doi

Synthetic progress in cMyc-Max oncoprotein miniaturization: Semi-online monitoring gives solid-phase access to hydrophobic b(-HLH-)ZIP peptidosteroid tweezers

DOI: 10.1002/ejoc.201201235

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejoc.201201235

Full text of DOI:10.1002/ejoc.201201235

Products guided by the article

R&D Labs maybe for 1425510-91-3

Relevant to this article

Hot Product