Toward an understanding of the chemical etiology for DNA minor-groove recognition by polyamides

Helvetica Chimica Acta ± Vol. 85 (2002)

4485

Toward an Understanding of the Chemical Etiology for DNA

Minor-Groove Recognition by Polyamides

by Michael A.Marques , Raymond M.Doss , Adam R.Urbach , and Peter B.Dervan*

Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA, 91125

Dedicated to Professor Dieter Seebach on the occasion of his 65th birthday

Crescent-shaped polyamides composed of aromatic amino acids, i.e., 1-methyl-1H-imidazole Im, 1-methyl-

H-pyrrole Py, and 3-hydroxy-1H-pyrrole Hp, bind in the minor groove of DNA as 2 :1 and 1:1 ligand/DNA

1

complexes. DNA-Sequence specificity can be attributed to shape-selective recognition and the unique corners

or pairs of corners presented by each heterocycle(s) to the edges of the base pairs on the floor of the minor

groove. Here we examine the relationship between heterocycle structure and DNA-sequence specificity for a

family of five-membered aromatic amino acids. By means of quantitative DNase-I footprinting, the recognition

behavior of polyamides containing eight different aromatic amino acids, i.e., 1-methyl-1H-pyrazole Pz, 1H-

pyrrole Nh, 5-methylthiazole Nt, 4-methylthiazole Th, 3-methylthiophene Tn, thiophene Tp, 3-hydroxythio-

phene Ht, and furan Fr, were compared with the polyamides containing the parent-ring amino acids Py, Im, and

Hp for their ability to discriminate between the four WatsonÀCrick base pairs in the DNA minor groove.

Analysis of the data and molecular modeling showed that the geometry inherent to each heterocycle plays a

significant role in the ability of polyamides to differentiate between DNA sequences. Binding appears sensitive

to changes in curvature complementarity between the polyamide and DNA. The Tn/Py pair affords a modest 3-

fold discrimination of T¥ A vs. A ¥ T and suggests that an S-atom in the thiophene ring prefers to lie opposite T

not A.

1.Introduction. ± Many diseases are related to aberrant gene expression, and the

ability to reprogram transcription in a cell by small molecules could be important in

biology and human medicine. Minor-groove-binding polyamides, which bind prede-

termined DNA sequences offer a chemical approach to artificial gene regulation. These

molecules are based on analogues of the 1-methyl-1H-pyrrole ring of amino acid

residue Py of the natural products netropsin and distamycin A, which have been shown

to bind in the minor groove of DNA in 1:1 and 2 :1 ligand/DNA stoichiometries [1 ± 4]

(

Fig. 1). Py is specific for A ¥ Tand T¥ A base pairs due to steric exclusion of H NÀC(2)

2

of guanine (G-NH2) [1 ± 3]. Base-pair specificity can be altered by changing the

functional group(s) presented to the floor of the DNA minor groove. Stabilizing and

destabilizing interactions with the different edges of the four WatsonÀCrick base pairs

are modulated by specific H-bonds and, importantly, steric fit or shape complemen-

tarity. For example, the 1-methyl-1H-imidazole residue Im presents the DNA with the

2

N-atom and its lone pair sp orbital, which can accept a H-bond from G-NH2 [1 ± 6].

Additionally, 3-hydroxy-1H-pyrrole residue Hp presents an OH group that is sterically

accommodated opposite T not A and, in addition, can donate H-bonds to the OC(2)

of thymine [7] [8]. For discrimination of each of the WatsonÀCrick base pairs, the 2 :1

stoichiometry involving unsymmetrical antiparallel cofacial pairs appears to be the best

solution such that Im/Py is specific for G ¥ C, Py/Im for C ¥ G, Hp/Py for T¥ A, and Py/

Hp for A ¥ T.

4

486

Helvetica Chimica Acta ± Vol. 85 (2002)

Fig. 1. Structures of polyamides bound to DNA: a) 2 :1 motif determined by X-ray crystallography [8]; b) 1 :1

motif determined by NMR [19]. DNA is shown as a stick model in blue. Polyamides are shown as space-filling

models, with imidazole residues in red, hydroxypyrrole in orange, pyrrole in yellow, and aliphatic residues in white.

The pairing rules have proven useful for the recognition of hundreds of DNA

sequences by polyamides. However, sequence-dependent DNA structural variation

(such as minor-groove width) makes binding affinity and specificity at many DNA

sequences unpredictable, which leads us to continue our search for new aromatic amino

acid residues of slightly different shape (curvature and twist) for minor-groove

recognition. Importantly, we find that the Hp residue can degrade over time in the

presence of acid or free radicals, and a robust replacement for the Hp/Py pair suitable

for use in biological studies is desirable. Several five-membered heterocyclic residues

other than Py, Im, and Hp have been investigated previously, including furan Fr [9],

thiazole Nt [10][11], and 1H-pyrazole Pz [11][12], with no new specificity uncovered.

This raises the issue whether there is something −special× about the 1-methyl-1H-pyrrole

analogs Py, Im, and Hp for minor-groove recognition. We attempt here to broaden the

repertoire of aromatic five-membered heterocycles for DNA recognition by synthesiz-

ing and characterizing a family of five-membered heterocyclic carboxamides grouped

by the type of functionality directed toward the floor of the DNA minor groove

(

Fig. 2). Analogs of Py would be 1-methyl-1H-pyrazole Pz and 1H-pyrrole Nh, which

project a H-atom toward the floor of the minor groove. Analogs of Im would be 5-

methylthiazole Nt and furan Fr, which project lone pairs of electrons from the N- or O-

atom. An analog of Hp is the 3-hydroxythiophene Ht, which projects an OH group. The

4-methylthiazole Th, 3-methylthiophene Tn, and thiophene Tp project a large

Helvetica Chimica Acta ± Vol. 85 (2002)

4487

heteroatom, the S-atom to the floor of DNA and likely represent a new class for shape-

selective recognition. We chose to maintain a five-membered heterocyclic framework in

our library to retain overall the crescent shape of the polyamide ligand and to observe

the effects of small structural changes resulting from single atom substitution on DNA

base-pair specificity. One anticipates that substitution of atoms in the five-membered

ring projecting away from the DNA minor groove (i.e., the non-reading frame) will

have effects on bond lengths and bond angles of each heterocycle as well and allows us

to ask how tolerant is DNA to subtle alterations in curvature and twist of the minor-

groove-binding ligand.

Fig. 2. Family of five-membered heterocyclic amino acids studied here. Centered above is a formula showing the

five-membered heterocyclic framework with the variable positions labeled X, Y, and Z. The parent Im, Py, and

Hp residues are boxed. All residues are shown with the functionality that faces the DNA minor groove pointed

down (X).

The covalent head-to-tail linkage of polyamide subunits in a 2 :1 complex results in

a hairpin oligomer with increased DNA affinity and sequence specificity [13][14]. The

hairpin motif avoids the ambiguity of slipping between the side-to-side stacked subunits

[

15], and −locks× individual ring pairings in a predictable cofacial manner. In a formal

sense, the hairpin oligomer provides a predictable foldamer for studying the DNA

recognition characteristics of new ring pairings. We have reported previously the

sequence specificities of Py/Py, Hp/Py, Pz/Py, and Th/Py pairings at a single position

within the hairpin-polyamide-sequence context Im-Im-X-Py-g-Im-Py-Py-Py-b-Dp

(

X Py, Hp, Pz, and Th; g g-aminobutanoic acid; b b-alanine; Dp N,N-

dimethylpropane-1,3-diamine) opposite the four WatsonÀCrick base pairs within the

4

488

Helvetica Chimica Acta ± Vol. 85 (2002)

sequence context 5'-ATGGXCA-3' (X A, T, G, and C) [11][17]. We found that Pz/Py

can mimic Py/Py and, as a surprising result, Th/Py bound all four base pairs with low

affinity and no specificity. Undaunted, we broaden our data set here to include Nh, Fr,

Ht, Tp, and Tn residues each paired with Py in the same hairpin context for comparison

(

Fig. 3). For the sake of completeness and to provide a comparative analysis of

heterocycle behavior opposite the four WatsonÀCrick base pairs in both stoichiometries

Fig. 3), we analyze polyamides also in a 1:1 polyamide/DNA complex (1:1 motif) of

(

type Im-b-ImPy-b-X-b-Im-Py-b-Dp (X Py, Hp, Nh, Ht, Fr, Nt, Tn, and Th), which

Fig. 3. Schematic illustrating the examination of sequence selectivity against the four WatsonÀCrick base pairs

within a) hairpin motifs and b) 1:1 motifs. Each chemical structure has a variable residue containing X, Y, and

Z-labeled positions, which are designated in Fig. 2. The dot models shown below each chemical structure

illustrate the binding mode with the polyamide shown inside its target DNA sequence: * imidazole residue;

*

pyrrole residue; ^ b-alanine residue; semicircle . g-aminobutanoic acid turn residue connecting the

two subunits; 6novel heterocycle residue.

Helvetica Chimica Acta ± Vol. 85 (2002)

4489

bind DNA sequences 5'-AAAGAXAGAAG-3' (X A, T, G, and C) in a single

orientation [16± 19].

Quantitative DNase-I footprinting was used to determine the equilibrium

association constant for each complex. Ab initio computational modeling of the

heterocyclic amino acids was implemented to derive their inherent geometric and

electronic parameters to guide interpretation of the experimental outcome. The

combination of experiment and modeling provides insight to the origin of DNA-

sequence discrimination by polyamides.

2

.Monomer, −Dimer×, and Polyamide Synthesis. ± Polyamides were synthesized

manually on solid support by the stepwise addition of monomeric and −dimeric× Boc-

protected amino acids, 1 ± 17 (Schemes 1 ± 7). Hydroxythiophene(Ht)-containing

polyamides were prepared by deprotecting the 3-methoxy analogue Mt. Boc-protected

amino acids for Im, Hp, Py, Pz, and Th monomers were synthesized according to

previously reported procedures [11][17][20]. Syntheses of core amino-ring-carboxylic

acid alkyl ester (NH -X-OR) structures 1 ± 4 for X Fr, Nh, Tn, and Mt are shown

2

in Schemes 2 ± 4 and 6. Boc-protected monomeric amino acids 5 ± 12 suitable for

olid-phase synthesis were prepared in two steps from their NH -X-OR analogues

2

(

Schemes 1 ± 7). However, aminofuran and aminothiophene precursors were un-

reactive to coupling on solid support, and, therefore, −dimers× 13 ± 17 were pre-

formed in solution under strong acylation conditions prior to solid-phase coupling.

Synthetic schemes for all Boc-protected amino acids used in this study follows

(

Schemes 2 ± 7).

Furan Ring (Fr) Derivatives. The nitro-furan ester 18 was prepared from methyl

furan-2-carboxylate [21]. The amino ester 1 was then synthesized from 18 by treatment

with H (500 psi) and Pd/C (Scheme 2) and isolated as the free base from a AcOEt

2

solution by precipitation with hexanes. The free base is a stable crystalline solid at room

temperature. The (nitro-imidazolyl)-furan ester 19 (NO -Im-Fr-OMe) was prepared by

2

condensing the 1-methyl-4-nitro-1H-imidazol-2-yl trichloromethyl ketone (NO -Im-

2

COCCl ) with 1 in AcOEt at 358. The −dimer× product 19 is rather insoluble in AcOEt

3

and began to precipitate upon formation. Reduction of 19 with H (500 psi) and Pd/C,

2

followed by addition of 2m HCl in Et O, gave the hydrochloride salt 20 (HCl ¥ H N-Im-

2

Fr-OMe). Salt 20 was Boc-protected (Boc (tert-butoxy)carbonyl) with (Boc) O, N,N-

2

diisopropylethylamine (DIEA), and DMF at 608 for 12 ± 18 h to give 21 (Boc-Im-Fr-

OMe; see 15, with CO Me instead of CO H). Elevated temperatures and extended

2

reaction times were necessary, presumably due to the poor nucleophilicity of the NH₂

group at the imidazole ring. Saponification of 21 in 1n aqueous NaOH and MeOH at

room temperature provided the target −dimer× 15 (Boc-Im-Fr-OH).

Alternatively, −dimer× 22 (Boc-b-Fr-OMe; see 13, with CO Me instead of CO H)

2

was synthesized from 1 by coupling to the symmetrical anhydride of Boc-b-alanine in

DMF, DIEA, and N,N-dimethylpyridin-4-amine (DMAP) (Scheme 2). The anhydride

was pre-formed in minutes with dicyclohexylcarbodiimide (DCC) in CH Cl at room

2

temperature. The amino-furan ester 1 was then added as a solution in DMF and DIEA,

followed by the transacylation catalyst DMAP. The 5-amino group at the furan ring is

significantly unreactive [22], and attempts to couple it on solid support with reagents

such as DCC and HOBt (1-hydroxy-1H-benzotriazol), HBTU (2-(1H-benzotriazol-1-

4

490

Helvetica Chimica Acta ± Vol. 85 (2002)

Scheme 1. a) Formation of Monomer and −Dimer× Boc-protected Amino Acids. b) Monomeric and −Dimeric×

Units

i) For Py (X CÀH, YNÀMe, Z CÀH, R Me), Im (X N, Y NÀMe, Z CÀH, R Et), Op (X

CÀOMe, YNÀMe, Z C ÀH, R Et), Th (X S, Y CÀMe, Z N, R Et), Pz (X CÀH, YNÀMe,

Z CÀH, R Et), Nt (X N, Y CÀMe, Z S, R H), and Nh (X NÀH, YZ CÀH, R Et),

(

Boc)

2

O, N,N-diisopropylethylamine (DIEA), DMF, 608, 12 ± 18 h; for Mt (X CÀOMe, YS, Z CÀH,

R Me), (Boc)

2

O, DIEA, Et

3

N, CH

2

Cl

2

, 6 08 , 12 ± 18 h. ii) For Py (X CÀH, YNÀMe, Z CÀH, R Me),

Im (X N, Y NÀMe, Z CÀH, R Et), Op (X CÀOMe, YNÀMe, Z CÀH, R Et), Th (X S, Y

CÀMe, Z N, R OEt), Pz (X CÀH, YNÀMe, Z CÀH, R Et), and Nh (X NÀH, YZ CÀH,

R Et), 1n NaOH, MeOH, r.t., 3 ± 4 h; for Nt (X N, Y CÀMe, Z S, R H), 1n NaOH, MeOH, r.t., 1 h;

for Mt (X CÀOMe, YS, Z CÀH, R Me), KOH, MeOH, 608, 4 ±6 h. iii) For Fr (X O, Y Z CÀH,

R Me), Tn (X S, Y CÀMe, Z CÀH, R Me), and Tp (X S, Y Z CÀH, R Me), O

2

N-Im-

COCCl

CÀH, R Me), and Tp (X S, Y Z CÀH, R Me), H

YZ CÀH, R Me), Tn (X S, Y CÀMe, Z CÀH, R Me), and Tp (X S, Y Z CÀH, R Me),

Boc)

O, DIEA, DMF, 608, 12 ± 18 h. vi) For Fr (X O, Y Z CÀH, R Me), 1n NaOH, MeOH, r.t., 3 h; for

Tn (X S, Y CÀMe, Z CÀH, R Me) and Tp (X S, Y Z CÀH, R Me), 1n NaOH, MeOH, 608, 4 ±

h. vii) For Fr (X O, Y Z CÀH, R Me) and Tn (X S, Y CÀMe, Z CÀH, R Me), (Boc-b-Ala) O,

DMF, DIEA, 408, 12 ± 18 h. viii) For Fr (X O, Y Z CÀH, R Me), 1n NaOH, MeOH, r.t., 3 h; for Tn

X S, Y CÀMe, Z CÀH, R Me), 1n NaOH, MeOH, 608, 4 ±6 h.

3

, AcOEt, DIEA, 358, 10 ± 12 h. iv) For Fr (X O, Y Z CÀH, R Me), Tn (X S, Y CÀMe, Z

2

, 10% Pd/C, AcOEt, r.t., 1.5 h. v) For Fr (X O,

(

2

6

2

(

yl)-1,1,3,3-tetramethyluronium hexafluorophosphate), HATU, PyBrOP, PyBOP, and

TFFH, were unsuccessful. Formation of the Boc-b-Fr-OMe −dimer× 22 occurred slowly

at elevated temperature. Heating above 408 did not affect the rate of reaction. Several

different solvent systems were tried, but DMF was found to be optimal. Saponification

of 22 was accomplished with a mixture of MeOH and 1n NaOH at room temperature to

provide the target 13 (Boc-b-Fr-OH), which is suitable for standard solid-phase

protocols [20].

1

-H-Pyrrole Ring (Nh) Derivatives. Pyrrolyl trichloromethyl ketone 23 was

prepared by adding 1H-pyrrole to a mixture of trichloroacetyl chloride in Et O at 08,

2

then warming to room temperature and stirring overnight, followed by precipitation

from hexanes. Nitration of 23 was accomplished with Ac O and nitric acid at À408 to

2

provide the 5-nitro regioisomer 24 as the major product (Scheme 3). Regiocontrol of

the nitration appears to depend significantly on the reaction temperature. Higher

temperatures provided unfavorable mixtures of 4-nitro and 5-nitro regioisomers.

Treatment of trichloromethyl ketone 24 with EtONa/EtOH at room temperature gave

the nitro-pyrrole ester 25 (NO -Nh-OEt) in good yield. Reduction of 25 with H₂

2

(

500 psi) and Pd/C, followed by the addition of 2m HCl in Et O, provided the

2

Helvetica Chimica Acta ± Vol. 85 (2002)

4491

Scheme 1 (cont.)

hydrochloride salt 2 (HCl ¥ H N-Nh-OEt), which was then Boc-protected ((Boc) O,

2

DIEA, and DMF) to yield 26 (Boc-Nh-OEt) which was saponified (1n NaOH, MeOH,

room temperature) to the final monomer unit 12 (Boc-Nh-OH).

3

-Methylthiophene Ring (Tn) Derivatives. The acyclic precursor to the thiophene-

ring system was prepared by a Knoevenagel reaction involving acetoacetate and

cyanoacetic acid, providing 27 as a mixture of (E)- and (Z)-regioisomers in moderate

yield after vacuum distillation. Treatment of 27 with sulfur flakes and Et NH in EtOH

2

yielded the cyclized aminothiophene, and addition of conc. HCl solution precipitated

the hydrochloride salt 3 (HCl ¥ H N-Tn-OMe) (Scheme 4). Formation of the (nitro-

2

imidazolyl)-thiophene ester 28 (NO -Im-Tn-OMe), followed by reduction and Boc-

2

protection to provide 29 (HCl ¥ H N-Im-Tn-OMe) and 30 (Boc-Im-Tn-OMe; see 16,

2

with CO Me instead of CO H), respectively, was accomplished by the procedures

2

described above for the furan compounds 19, 20, and 21. The Boc-b-Tn-OMe −dimer× 31

see 14; with CO Me instead of CO H) was prepared as described above for 22, by

(

2

coupling 3 with the symmetric anhydride of Boc-b-alanine. The NH group at the

2

thiophene ring displays low reactivity comparable to furan. Elevated temperature was

4

492

Helvetica Chimica Acta ± Vol. 85 (2002)

Scheme 2

i)

H2N

CO2Me

O2N

CO2Me

O

1

8

ii) iii)

iv)

O

N

CO2Me

O

BocHN

N

CO2H

H

O

H

N

1

9

O2N

13

v)

O

N

vi) vii)

N

H

CO2Me

O

N

CO2H

O

H

N

2

0

1

5

HCl·H2N

BocHN

i) H

2

, Pd/C, AcOEt, 500 psi, r.t. ii) (Boc-b-Ala)

2

O, DMF, DIEA, DMAP, 408. iii) 1n NaOH, r.t. iv) NO

2

-Im-

COCCl

3

, AcOEt, 358. v) H

2

, Pd/C, AcOEt, 500 psi, r.t. vi) (Boc) O, DIEA, DMF, 608. vii) 1n NaOH, MeOH,

2

r.t.

necessary to completely saponify methyl esters 30 and 31. Saponification was carried

out in 1n NaOH and MeOH at 608 for 4 ± 6h to give the −dimers× 16 (Boc-Im-Tn-OH)

and 14 (Boc-b-Tn-OH), respectively.

Thiophene Ring (Tp) Derivatives. Treatment of commercially available 5-nitro-

thiophene-2-carboxaldehyde in acetone with a mixture of NaOCl and Na HPO in

2

4

H O, gave the nitro-thiophene acid 32 (NO -Tp-OH). Esterification of 32 by refluxing

2

for 48 h in a mixture of H SO and MeOH provided the nitro ester 33 (NO -Tp-OMe)

2

4

2

(

Scheme 5). Reduction of 33 with a mixture of tin(II) chloride dihydrate and HCl in

EtOH gave the hydrochloride salt 34 (HCl ¥ H N-Tp-OMe). Formation of the −dimer×

2

3

5 (NO -Im-Tp-OMe), followed by reduction to 36 (HCl ¥ H N-Im-Tp-OMe), Boc-

2

protection to 37 (Boc-Im-Tp-OMe; see 17, with CO Me instead of CO H, and

2

saponification to the target −dimer× 17 (Boc-Im-Tp-OH) proceeded as described above

for furan compounds 19, 20, 21, and 15.

Helvetica Chimica Acta ± Vol. 85 (2002)

4493

Scheme 3

i)

COCCl3

CO2Et

CO2H

N

O2N

N

H

2

3

24

ii)

iii)

CO2Et

HCl·H2N

O2N

N

H

2

25

iv)

v)

CO2Et

BocHN

N

BocHN

N

H

2

6

12

i) HNO

3

, Ac

2

O, À408. ii) EtONa, EtOH, reflux. iii) H

2 2 2

, Pd/C, DMF, HCl, Et O, r.t. iv) (Boc) O, DIEA, DMAP,

DMF, r.t. v) 1n NaOH, MeOH, r.t.

3

-Methoxythiophene Ring (Mt) Derivatives. The hydroxythiophene methyl ester 38

(

Ht-OMe) was synthesized by a cyclization reaction between methyl thioglycolate and

methyl-2-chloroacrylate in MeONa/MeOH [23]. Ester 38 was nitrated with a mixture

of conc. H SO and HNO at À108 to give 4-nitro-3-hydroxythiophene ester 39 (NO -

2

4

3

2

Ht-OMe) as the major regioisomer after column chromatography (Scheme 6).

Treatment of 39 with CH N in Et O afforded the methyl ether 40 (NO -Mt-OMe;

2

see 39, with OMe instead of OH) in near quantitative yield (Ht 3-hydroxythiophene

residue, Mt 3-methoxythiophene residue). Reduction of 40 with a mixture of tin(II)

chloride dihydrate, HCl, and EtOH gave the hydrochloride salt 4 (HCl ¥ H N-Mt-

2

OMe). Boc-protection was accomplished by heating a mixture of 4, (Boc) O, Et N, and

2

3

CH Cl at 608 for 12 h !41 (Boc-Mt-OMe; see 11, with CO Me instead of CO H).

2

Saponification of 41 was achieved with methanolic KOH and heating at 508 for 6h to

give the final monomer 11 (Boc-Mt-OH).

5

-Methylthiazole Ring (Nt) Derivative. The Boc-protected 2-amino-5-methylthia-

zole acid 10 (Boc-Nt-OH) was synthesized on a multi-gram scale by brominating 2-

4

494

Helvetica Chimica Acta ± Vol. 85 (2002)

Scheme 4

CN

O

i)

CO2Me

HCl·H2N

MeO

S

2

7

3

ii) iii)

iv)

O

N

CO2Me

S

H

BocHN

N

CO2H

N

S

H

2

8

O2N

1

4

v)

O

vi) vii)

N

CO2Me

N

CO2H

S

H

N

2

9

16

HCl·H2N

BocHN

i) S, Et

COCCl

2

NH, EtOH, r.t. ii) (Boc-b-Ala)

, AcOEt, 358. v) H , Pd/C, AcOEt, 500 psi, r.t. vi) (Boc)

08.

2

O, DMF, DIEA, DMAP, 408, iii) 1n NaOH, MeOH, 608. iv) NO

2

-Im-

3

2

O, DIEA, DMF, 608. vii) 1n NaOH, MeOH,

6

oxobutanoic acid, followed by condensation with thiourea and Boc-protection of the

amine (Scheme 7). For best results, the Br should be added dropwise over at least 2 h,

2

as the reaction is autocatalytic and highly exothermic. Also, the thiourea should be

added in small portions with vigorous stirring. Boc-protection was accomplished by

dissolving the crude material in DMF and DIEA, followed by the addition of (Boc) O

2

and stirring at 608 for 12 h. The material was then stirred in a solution of MeOH and 1n

NaOH for ester saponification to provide the target compound 10 (Boc-Nt-OH).

Polyamides. Hairpin and 1:1 motif polyamides were synthesized manually from

Boc-b-Pam resin in a stepwise fashion with Boc-protected monomeric and −dimeric×

amino acids (Scheme 8) according to solid-phase protocols [20]. Polyamides containing

3-methoxythiophene (Mt) were deprotected by treatment with sodium benzenethio-

late in DMF (1008, 2 h) to provide the Ht analogues after HPLC purification.

Helvetica Chimica Acta ± Vol. 85 (2002)

4495

Scheme 5

i)

CO2H

CO2Me

O2N

S

3

2

33

ii)

O

N

iii)

N

H

CO2Me

S

CO Me

2

HCl·H2N

S

N

3

5

O2N

34

iv)

O

N

v) vi)

N

H

CO2Me

S

N

H

CO2H

S

N

3

6

HCl·H2N

17

BocHN

i) H

2

SO

4

, MeOH, reflux. ii) SnCl

2

¥ 2 H

2

O, 95% EtOH, 37% HCl soln., 358. iii) NO

2 3

-Im-COCCl , AcOEt, 358.

iv) H , Pd/C, AcOEt, 500 psi, r.t. v) (Boc)

2

O, DIEA, DMF, 608. vi) 1n NaOH, MeOH, 608.

3

.DNA Affinity and Sequence Specificity in the Hairpin Motif. ± Quantitative

DNase-I-footprinting titrations [24] were carried out for the following polyamides on

the 278-base-pair PCR product of plasmid pDHN1 [11]: Im-Im-Im-Py-g-Im-Py-Py-Py-

b-Dp (Im/Py pair; 42), Im-Im-Nh-Py-g-Im-Py-Py-Py-b-Dp (Nh/Py pair; 43), Im-Im-

Tn-Py-g-Im-Py-Py-Py-b-Dp (Tn/Py pair; 44), Im-Im-Tp-Py-g-Im-Py-Py-Py-b-Dp (Tp/

Py pair; 45), Im-Im-Ht-Py-g-Im-Py-Py-Py-b-Dp (Ht/Py pair; 46), and Im-Im-Fr-Py-g-

Im-Py-Py-Py-b-Dp (Fr/Py pair; 47). The DNA-sequence specificity of each polyamide

at a single ring-pairing position (bolded in the sequences listed above) was determined

by varying a single DNA base pair within the parent-sequence context, 5'-TGGXCA-3',

to all four WatsonÀCrick base pairs (X A, T, G, C) and comparing the relative

affinities of the resulting complexes (Fig. 4). The variable base-pair position was

installed opposite the novel heterocycle/pyrrole pairing in question, according to

previously reported specificity studies on eight-ring hairpin polyamides [7][11].

Equilibrium association constants (K ) for eight-ring polyamides containing Hp/Py,

a

Py/Py, Pz/Py, and Th/Py pairings against the four DNA sites used in this study have

been reported [11][17] and are included in Table 1 for comparison with values

4

496

Helvetica Chimica Acta ± Vol. 85 (2002)

Scheme 6

S

i)

O2N

CO2Me

OH

3

9

3

8

ii) iii)

S

iv) v)

BocHN

CO2H

HCl·H2N

CO2Me

OMe

1

4

i) HNO

3

, H

2

SO

4

, 08. ii) CH

2

N

2

, r.t. iii) SnCl

2

¥ 2 H

2

O, HCl, EtOH, 408. iv) (Boc)

2

O, CF

3

2 2

COOH, CH Cl , 6 08 .

v) 1n NaOH, MeOH, 508.

Scheme 7

O

Br

i) ii) iii)

S

HO2C

BocHN

CO2H

N

O

1

0

2 4 2

i) CSN H , neat, r.t. ii) (Boc) O, DIEA, DMF, DMAP, 608. iii) 1n NaOH, MeOH, 358.

presented here. As expected, polyamide 42 (Im/Py pair) exhibited single-site specificity

8

À1

(

5'-TGGXCA-3', X G) at modest affinity (K 4.5 ¥ 10 m ) with fourteen-fold

a

preference over X C and at least 400-fold preference over X A, T. Polyamide 43

10

À1

(

Nh/Py pair) bound with high affinity to the X A, T sites (K ꢀ10 m ) in preference

a

to X G, C by about ten-fold. Hairpin 44 (Tn/Py pair) bound with high affinity to X

9

À1

T, A (K ꢀ10 m ), with a 3-fold preference for T¥ A > A ¥ T, and 800-fold preference

a

over the X G, C sites. The thiophene analogue, Tp, lacking the methyl group was

examined to probe possible effects on DNA binding caused by the 3-methyl group. The

Tp/Py pair (45) was found to display similar recognition properties as the Tn/Py pair

(44). Remarkably, hairpins containing Ht/Py and Fr/Py pairs (46 and 47, resp.)

demonstrated no binding to the designed sites at concentrations up to 1 mm.

4

.DNA Affinity and Sequence Specificity in the 1:1 Motif. ± Quantitative DNase-

I-footprinting titrations were carried out for the following polyamides on the 298-base-

Helvetica Chimica Acta ± Vol. 85 (2002)

4497

Fig. 4. Quantitative DNase-I-footprinting experiments in the hairpin motif for a) polyamides 43, b) polyamide

4, c) polyamide 46, and d) polyamide 47 on the 278-base-pair, 5'-end-labelled PCR product of plasmid DHN1.

Lane 1, intact DNA; Lane 2, G reaction; Lane 3, A reaction; Lane 4, DNase-I standard; Lanes 5 ± 15, 1 pm,

pm, 10 pm, 30 pm, 100pm, 300 pm, 1 nm, 3 nm, 10 nm, 30 nm, and 100 nm polyamide, resp. Each footprinting gel

4

3

is accompanied by the following: chemical structure of the residue of interest (left, top) and, for a) and b),

binding isotherms for the four designed sites (left, bottom). qnorm values were obtained according to published

methods [24][26]. A binding model for the haipin motif is shown centered at the top as a dot model with the

polyamide bound to its target DNA sequence: * imidazole residue; * pyrrole residue; ^ b-alanine

residue; . semicircle g-aminobutanoic acid turn residue connecting the two subunits; 6novel heterocycle

residue.

4

498

Helvetica Chimica Acta ± Vol. 85 (2002)

Scheme 8. Solid-Phase Synthesis of Im-Im-X-Py-g-Py-Py-Py-b-Dp (arrows up from center) and Im-b-ImPy-b-

X-b-ImPy-b-Dp (arrows down from center) starting from commercially available Boc-b-Pam resin

Helvetica Chimica Acta ± Vol. 85 (2002)

4499

À1

a b

Table 1. Hairpin Motif: K

a

[M ] ) )

Pair^c)

A ¥ T

T¥ A

G ¥ C

C ¥ G

6

8

7

Im/Py (42)

Hp/Py

Py/Py

Nh/Py (43)

Pz/Py

Th/Py

Tn/Py (44)

Tp/Py (45)

Ht/Py (46)

Fr/Py (47)

ꢁ10

4.5 (Æ0.7) ¥ 10

3.2 (Æ0.5) ¥ 10

7

9

7

8

7

8.1 (Æ1.9) ¥ 10

1.6( Æ0.3) ¥ 10

5.5 (Æ1.5) ¥ 10

7.9 (Æ2.1) ¥ 10

2.5 (Æ0.9) ¥ 10

9

8

3.1 (Æ0.7) ¥ 10

8.5 (Æ0.3) ¥ 10

4.7 (Æ0.4) ¥ 10

2.2 (Æ0.6) ¥ 10

9

10

1.1 (Æ0.1) ¥ 10

9.2 (Æ0.1) ¥ 10

8.2 (Æ0.4) ¥ 10

9

7

1.0 (Æ0.5) ¥ 10

2.0 (Æ0.3) ¥ 10

ꢁ2 ¥ 10

7

ꢁ2 ¥ 10

8

9

6

8.0 (Æ0.4) ¥ 10

2.7 (Æ0.2) ¥ 10

ꢁ10

8

9

6

3.8 (Æ0.5) ¥ 10

1.0 (Æ0.3) ¥ 10

ꢁ10

6

ꢁ10

6

ꢁ10

^a) Values reported are the mean values from at least three DNase-I-footprint titration experiments, with the

b

standard deviation given in parentheses. ) Assays were performed at 228 in a buffer of 10 mm Tris ¥ HCl, 10 mm

c

KCl, 10 mm MgCl

2

, and 5 mm CaCl

2

at pH 7.0. ) The number in parentheses indicates the compound containing

the unique pairing.

pair PCR product of pAU8 [30]: Im-b-Im-Py-b-Py-b-Im-Py-b-Dp (48), Im-b-Im-Py-b-

Hp-b-Im-Py-b-Dp (49), Im-b-Im-Py-b-Nh-b-Im-Py-b-Dp (50), Im-b-Im-Py-b-Ht-b-

Im-Py-b-Dp (51), Im-b-Im-Py-b-Fr-b-Im-Py-b-Dp (52), Im-b-Im-Py-b-Nt-b-Im-Py-b-

Dp (53), Im-b-Im-Py-b-Tn-b-Im-Py-b-Dp (54), and Im-b-Im-Py-b-Th-b-Im-Py-b-Dp

(

55). The sequence specificity of each polyamide at a single carboxamide position

bolded in the sequences listed above) was determined by varying a single base pair

within the parent DNA sequence context, 5'-AAAGAXAAGAG-3', to all four

WatsonÀCrick base pairs (X A, T, G, C) and comparing the relative affinities of the

resulting complexes (Figs. 5 and 6). The variable base-pair position was installed

opposite the novel heterocycle in question, according to previously described

specificity studies on 1:1 polyamide/DNA complexes [17].

Equilibrium association constants (K ) for 1:1 polyamides containing Im, Py, and

a

Hp residues tested against the four WatsonÀCrick base pairs have been reported.

However, in that study, only the Im specificity experiment was performed at the central

residue, as with the new polyamides reported here. Therefore, new polyamides

containing Py and Hp residues at the central position have been included in the study

for a more controlled comparison. Polyamide 48 (Py) demonstrated very high affinity

10

À1

(

K ꢀ6 ¥ 10 m ) at the X A, T sites (5'-AAAGAXAAGAG-3') with a 5- to 10-fold

a

i) 80% CF

CF COOH/CH

vii) Boc-g-OH, HBTU, DIEA, DMF. viii) 80% CF

x) 80% CF COOH/CH Cl , 0.4m PhSH. xi) Boc-X-OH, HBTU, DIEA, DMF. xii) 80% CF

.4m PhSH. xiii) Boc-Im-OH, HBTU, DIEA, DMF. xiv) Im-COCCl , DIEA, DMF. xv) 80% CF

CH Cl , 0.4m PhSH. xvi) Boc-X-OH, HBTU, DIEA, DMF. xvii) 80% CF COOH/CH Cl , 0.4m PhSH. xviii)

Im-COCCl , DIEA, DMF. xix) Im-COCCl , DIEA, DMF. xx) N,N-Dimethylpropane-1,3-diamine, 858. xxi)

0% CF COOH/CH Cl , 0.4m PhSH. xxii) Boc-Py-OBt, DIEA, DMF. xxiii) 80% CF COOH/CH Cl , 0.4m

PhSH. xxiv) Boc-b-Im-OH [19], HBTU, DIEA, DMF. xxv) 80% CF COOH/CH Cl , 0.4m PhSH. xxvi) Boc-X-

OH, HBTU, DIEA, DMF. xxvii) 80% CF COOH/CH Cl , 0.4m PhSH. xxviii) Boc-X-OH, HBTU, DIEA,

DMF. xxix) Elongation and termination according to standard procedures [19].

3

COOH/CH

Cl , 0.4m PhSH. v) Boc-Im-OH, HBTU, DIEA, DMF. vi) 80% CF

COOH/CH Cl , 0.4m PhSH. ix) Boc-Py-OBt, DIEA, DMF.

COOH/CH Cl

COOH/

2

Cl

2

, 0.4m PhSH. ii) Boc-Py-OBt, DIEA, DMF. iii) Repeat steps i) and ii) twice. iv) 80%

3

2

3

COOH/CH Cl , 0.4m PhSH.

2

3

2

3

2

3

2

,

0

3

2

3

2

3

8

3

2

3

2

3

2

3

2

4

500

Helvetica Chimica Acta ± Vol. 85 (2002)

Fig. 5. Quantitative DNase-I-footprinting experiments for a) polyamide 48, b) polyamide 49, c) polyamide 50,

and d) polyamide 51 on the 298-base-pair, 5'-end-labelled PCR product of plasmid pAU8. a) and b): Lane 1,

intact DNA; Lane 2, G reaction; Lane 3, A reaction; Lane 4, DNase-I standard; Lanes 5 ± 15, 100 fm, 300 fm,

1

pm, 3 pm, 10 pm, 30 pm, 100 pm, 300 pm, 1 nm, 3 nm, and 10 nm polyamide, resp. c): Lane 1, intact DNA; Lane 2,

G reaction; Lane 3, A reaction; Lane 4, DNase I standard; Lanes 5 ± 15, 1 pm, 3 pm, 10 pm, 30 pm, 100 pm,

00 pm, 1 nm, 3 nm, 10 nm, 30 nm, and 100 nm polyamide, resp. d) Lane 1, intact DNA; Lane 2, G reaction;

Lane 3, A reaction; Lane 4, DNase-I standard; Lanes 5 ± 15, 3 pm, 10 pm, 30 pm, 100 pm, 300 pm, 1 nm, 3 nm,

0 nm, 30 nm, 100 nm, and 300 nm polyamide, resp. Each footprinting gel is accompanied by the following:

3

1

chemical structure of the residue of interest (left, top) and binding isotherm for the four designed sites (left,

bottom). qnorm values were obtained according to published methods [24]. A binding model for the 1 :1 motif is

shown centered at the top as a dot model with the polyamide bound to its target DNA sequence: * imidazole

residue; * pyrrole residue; ^ b-alanine residue; 6novel heterocycle residue.

Helvetica Chimica Acta ± Vol. 85 (2002)

4501

Fig. 6. Quantitative DNase-I-footprinting experiments for a) polyamide 52, b) polyamide 53, c) polyamide 54,

and d) polyamide 55, on the 298-base-pair, 5'-end-labelled PCR product of plasmid pAU8. Lane 1, intact DNA;

Lane 2, G reaction; Lane 3, A reaction; Lane 4, DNase-I standard; Lanes 5 ± 15, 100 fm, 300 fm, 1 pm, 3 pm,

1

0 pm, 30 pm, 100 pm, 300 pm, 1 nm, 3 nm, and 10 nm polyamide resp. Each footprinting gel is accompanied by

the following: chemical structure of the residue of interest (left, top) and binding isotherms for the four

designed sites (left, bottom). Isotherms for c) and d) were generated from gels run out to a final concentration

of 1 mm (not shown). qnorm values were obtained according to published methods [24]. A binding model for the

1

:1 motif is shown centered at the top as a dot model with the polyamide bound to its target DNA sequence:

imidazole residue; * pyrrole residue; ^ b-alanine residue; 6novel heterocycle residue.

*

4

502

Helvetica Chimica Acta ± Vol. 85 (2002)

preference over X G, C (Table 2). Polyamide 49 (Hp) bound with lower affinity

9

À1

(

1

K ꢀ3 ¥ 10 m ) but with similar specificity to 48, preferring X A, T> G, C by 5- to

a

0-fold. The Nh-containing polyamide 50 bound with very high affinity to the X A, T

10

À1

sites (K 7.5 ¥ 10 m ) but with a mere 3- to 5-fold selectivity over the high-affinity

a

X G, C sites. Polyamide 51 (Ht) bound with subnanomolar affinities to the X A, T

sites, similar to 49 but with ꢂ40-fold specificity for X A, T> G, C. Polyamide 52 (Fr)

10

À1

showed high affinity for the X A, T sites (K ꢀ10 m ) with a small 2- to 4-fold

a

preference over X G, C. The 5-methylthiazole-containing polyamide 53 (Nt), which

places the thiazole-ring N-atom into the floor of the minor groove, bound all four sites

9

À1

with similar high affinities (K ꢀ5 ¥ 10 m ). Thiophene-containing polyamide 54 (Tn)

a

showed specificity for A, T vs. G, C and a preference for a single A ¥ T site. The 4-

methylthiazole-containing polyamide 55 (Th), which places the thiazole-ring S-atom

into the floor of the minor groove, bound with similar X A, T affinity as 54 (Tn) but

with >400-fold preference over X G, C. In all cases, binding isotherms fit well to an

n 1 Hill equation, supporting a 1:1 polyamide/DNA stoichiometry (Figs. 4 ± 6).

À1

a b

Table 2. 1 : 1 Motif: K

a

[m ] ) )

G ¥ C

Ring^c)

A ¥ T

T¥ A

C ¥ G

1

0

10

9

10

9

Im

2.5 (Æ0.2) ¥ 10

7.2 (Æ0.3) ¥ 10

1.1 (Æ0.1) ¥ 10

2.6( Æ0.4) ¥ 10

3.2 (Æ0.4) ¥ 10

5.3 (Æ0.5) ¥ 10

1.3 (Æ0.3) ¥ 10

10

Py (48)

Hp (49)

Nh (50)

Ht (51)

Fr (52)

Nt (53)

Tn (54)

Th (55)

5.3 (Æ0.1) ¥ 10

9.4 (Æ0.2) ¥ 10

9

1

9

8

3.9 (Æ0.1) ¥ 10

2.5 (Æ0.3) ¥ 10

1.9 (Æ0.5) ¥ 10

2.3 (Æ0.1) ¥ 10

0

10

7

10

7

7.5 (Æ0.2) ¥ 10

2.8 (Æ0.5) ¥ 10

2.2 (Æ0.5) ¥ 10

5.4 (Æ0.9) ¥ 10

3.0 (Æ0.2) ¥ 10

7.4 (Æ0.1) ¥ 10

1.6( Æ0.2) ¥ 10

3.8 (Æ1.3) ¥ 10

9

1

9

1

9

1.6( Æ0.6) ¥ 10

3.7 (Æ0.7) ¥ 10

5.0 (Æ0.5) ¥ 10

4.2 (Æ0.6) ¥ 10

8.3 (Æ0.2) ¥ 10

7.4 (Æ0.5) ¥ 10

10

9

7

6

1.0 (Æ1.3) ¥ 10

4.4 (Æ0.5) ¥ 10

9

2.9 (Æ0.6) ¥ 10

8.0 (Æ1.3) ¥ 10

8.1 (Æ0.4) ¥ 10

6.5 (Æ0.5) ¥ 10

9

7

6

5.7 (Æ0.4) ¥ 10

1

0

9

1.5 (Æ0.2) ¥ 10

3.0 (Æ0.7) ¥ 10

^a) Values reported are the mean values from at least three DNase-I-footprint titration experiments, with the

b

standard deviation given in parentheses. ) Assays were performed at 228 in a buffer of 10 mm Tris ¥ HCl, 10 mm

c

KCl, 10 mm MgCl

2

, and 5 mm CaCl

2

at pH 7.0. ) The number in parentheses indicates the compound containing

the unique pairing.

5

.Molecular-Modeling Calculations. ± Modeling calculations were preformed with

the −Spartan Essential× software package [25]. Each ring was first minimized by means

of an AM1 model, followed by ab initio calculations by means of the HartreeÀFock

model and a 6-31G* polarization basis set. Each heterocycle exhibited a unique

geometric and electronic profile (Fig. 7). Bonding geometries for the imidazole,

pyrrole, and 3-hydroxypyrrole residues were in excellent agreement with coordinates

derived from X-ray structures of polyamides containing these heterocycles [7][9]. The

overall curvature of each monomer was calculated to be the sweep angle (q) created by

the theoretical intersection of the two ring-to-amide bonds in each ring. The structures

were ranked by increasing q as follows: Fr > Nt > Ht > Nh > Im > Py > Hp > Tn >

Pz > Tp > Th. The ring atom in closest proximity to the floor of the DNA minor

groove was examined for partial charge. The structures were ranked by decreasing

partial charge on this atom as follows: Hp > Ht > Nh > Pz > Py > Tn Tp > Th > Fr >

Helvetica Chimica Acta ± Vol. 85 (2002)

4503

Fig. 7. Geoetric and electrostatic profiles for eleven heterocyclic amino acids, derived from ab initio molecular-

modeling calculations with −Spartan Essential× software [25]. Left: Schematic illustrating the ami-

deÀringÀamide angle of curvature, q. X, Y, and Z denote variable functionality at the different ring positions

for each heterocycle (see Fig. 2). Right: functional groups at X, Y, and Z, along with the angle q and the

electrostatic partial charge at X, for Ht, Nh, Py, Hp, and Pz, the positive charge at X is listed for the H-atom.

Im. Four-ring subunits containing the sequence Im-Im-X-Py (X Py, Pz, Nh, Im, Fr,

Hp, Ht, Th, Tn, and Tp) were constructed and subjected to AM1 and ab initio

calculations as described above to examine overall subunit curvature and planarity

(

Figs. 8 and 9).

6

.Discussion. ± Here we explore the effects of varying single atom positions in five-

membered aromatic heterocycles on the ability of polyamides to discriminate the four

WatsonÀCrick base pairs in the minor groove of DNA. In this experimental design, the

incremental scheme of DNA and polyamide sequence allows for the comparison of

binding affinities for a {4} Â{10} array of complexes containing unique combinations of

{

(

WatsonÀCrick base pair} Â{five-membered aromatic heterocycle} at a single position

Tables 1 and 2). This quantitative analysis combined with computational modeling of

the different heterocycles has led to insight into the etiology of DNA-sequence

discrimination by polyamides.

The success of Im/Py and Hp/Py pairs at discriminating between the four

WatsonÀCrick base pairs has been attributed to the shape of the functional groups

directed toward the floor of the minor groove. Therefore, the heterocycles discussed

here will be divided into groups based on the types of −bumps and holes× presented to

the complementary bumps and holes created by each WatsonÀCrick base pair on the

floor of the DNA minor groove. For the sake of clarity and brevity, hairpin polymaides

will be referred to in bold as their unique amino acid pair, e.g., Tn/Py for polyamide Im-

Im-Tn-Py-g-Im-Py-Py-Py-b-Dp (44). The 1:1 polyamides will be referred to in bold as

their unique heterocycle, e.g., Tn for polyamide Im-b-Im-Py-b-Tn-b-Im-Py-b-Dp (54).

DNA sequences will be identified as the variable base position within each motif, e.g.,

X G in the hairpin motif for 5'-TGGXCA-3'.

The Pairing Rules. Discrimination of the four WatsonÀCrick base pairs by means of

unsymmetrical cofacial pairs of aromatic amino acids has proven to be a key insight for

4

504

Helvetica Chimica Acta ± Vol. 85 (2002)

Fig. 8. a) Schematic illustrating the curvatures of four-ring polyamide subunits containing Tn, Py, and Fr

heterocycles with respect to one another and the DNAhelix ; b) ab initio models of polyamide subunits (Im-Im-X-

Py, X Tn, Py, and Fr) superimposed to demonstrate the significant difference in curvature resulting from atomic

substitution (H-atoms are not shown)

minor-groove DNA recognition. The four-ring contiguous subunits described here (e.g.

Im-Py-Py-Py) are p-conjugated, which limits their conformational flexibility. Con-

sequently, small changes in individual ring curvature can cause greater effects on

overall oligomer curvature.

Py/Py, Pz, Py, and Nh/Py present an H-atom with positive potential to the minor-

groove floor. As reported previously, Py/Py exhibits preference for the X A, T

binding sites with >10-fold selectivity for X A, T> G, C. Pz/Py behaves with similar

affinity as Py/Py, but greater discrimination for A, T vs. G, C. Nh/Py is similar to Py/Py

and binds all sites with higher affinity than Py/Py and Pz/Py. The A ¥ T/T¥ A preference

for the pairs Py/Py, Pz/Py, and Nh/Py originate from a steric interaction with G-NH2,

yet there are subtle differences among these compounds, which undoubtedly derive

from their slightly different shape. Py is known to be over-curved with respect to the

DNA helix [6][8][27 ± 29]. The calculations described here provide an amide-ring-

amide intersection angle (q in Fig. 7) of 1468 and a partial positive charge at the

HÀC(3) of 0.21. By contrast, Pz is somewhat less curved than Py (q 1518) with

similar charge, and Nh is considerably more curved (q 1368) with a greater charge of

0.34. The reduced curvature of Pz/Py should make it more complementary to the

DNA curvature, and therefore, the steric HÀC(4) to G-NH2 clash would be

exacerbated, resulting in lower affinity for the X G, C sites. Nh/Py is more curved

Helvetica Chimica Acta ± Vol. 85 (2002)

4505

Fig. 9. Ab initio models of four-ring polyamide subunits (Im-Im-X-Py, X Ht, Hp, Th, and Tn): a) Im-Im-Ht-

Py subunit demonstrating a dihedral (shown as a curved arrow) created at the ringÀcarboxamide juncture due to

destabilizing eclipsing interaction (shown as red arcs) between HOÀC(3) and the proximal carboxamide proton;

b) Im-Im-Hp-Py subunit demonstrating coplanarity of the contiguous ring system; c) Im-Im-Th-Py subunit

showing the negative isopotential surface (lone-pair density) in red and the dihedral (shown as a curved arrow)

resulting from lone-pair repulsion interaction between the thiazole N-atom and the proximal carbonyl group;

d) Im-Im-Tn-Py subunit showing the negative isopotential surface (lone-pair density) in red and the small

dihedral (shown as a curved arrow) resulting from a destabilizing eclipsing interaction (shown as red arcs)

between the thiophene Me group and the proximal carboxamide. Atomic substitution of N to CÀH for Th to Tn

removes lone-pair repulsion interaction between the ring N-atom and the proximal carbonyl group.

and more charged, thereby reducing negative steric effects and increasing binding

affinity.

2

Im/Py and Fr/Py present an N- or O-atom with sp lone-pair electrons directed

toward the minor-groove floor. Their DNA-recognition behavior is strikingly different.

Im/Py preferentially targets X G. On the other hand, Fr/Py shows a complete loss of

DNA-binding affinity. Based on the established principles for G ¥ C recognition by an

Im/Py pair, it was not unreasonable to expect that the Fr/Py pair could be a positive

recognition element for G ¥ C as well, with the Fr O-atom acting as a H-bond acceptor

to G-NH2. However, calculations for Fr reveal tight over-curvature, with the amide-

ring-amide angle decreased by more than ten degrees with respect to Im and more than

twenty degrees with respect to Py. This property causes a pronounced effect on the

4

506

Helvetica Chimica Acta ± Vol. 85 (2002)

entire Im-Im-Fr-Py subunit (Fig. 8), such that complex formation is no longer

energetically favorable. This view is further reinforced by results in the more flexible

1

:1 motif, wherein the Fr-containing polyamide binds all designed sites with high affinity.

Hp/Py and Ht/Py present an OH group (HOÀC(3)) to the minor-groove floor. Hp/

Py displays a preference for X T that breaks the A ¥ T/T¥ A degeneracy of Py/Py pairs.

The Hp/Py pair presents OH opposite T not A. Structural studies reveal the origin of

this specificity to reside in the H-bond formed between HOÀC(3) of Hp and OC(2)

of thymine (T-O2), and to shape recognition of the asymmetric cleft in the T¥ A base

pair [8][29]. However, a loss in affinity is typically observed for polyamides containing

Hp/Py pairs compared to the Py/Py pair revealing that there is an energetic penalty for

the gain in selectivity [30]. This loss in affinity −trade-off× may be attributed to

unfavorable steric interactions between the HOÀC(3) and the minor-groove floor or

perhaps differential solvation between Hp/Py and Py/Py in H O. The Ht/Py pair was

2

designed with the hope to improve upon Hp/Py by increasing the ring curvature, thus

reducing the steric interaction while maintaining the HOÀC(3) to T-O2 H-bond.

Remarkably, DNase-I footprinting for Ht/Py reveals a complete loss in binding affinity.

As seen with Fr/Py, increased ligand curvature may be responsible for disrupting

hairpin binding. Ab initio calculations on the Im-Im-Ht-Py subunit reveal an

unfavorable eclipsing interaction between HOÀC(3) and the proximal carboxamide

proton. As shown in Fig. 9, this steric clash may force rotation about the ringÀamide

bond, which would twist the subunit out of plane. Given the snug fit of stacked hairpin

subunits within the DNA minor groove, a large distortion in ligand planarity may not be

tolerated. Ht in the flexible 1:1 motif binds its target sites with high affinity, which

further underscores the pronounced effects of ring geometry within the conformation-

ally constrained hairpin motif.

2

Th/Py, Tn/Py, and Tp/Py pairs present a large S-atom with an sp lone pair to the

minor-groove floor. The thiazole analog Th/Py has been shown previously to afford

poor affinity overall and no discrimination of the WatsonÀCrick base pairs [11]. It was

thought that the thiazole S-atom was too large to be easily accommodated within the

closely packed cofacial rings in the hairpin ¥ DNA complex. However, ab initio

calculations on thiazolecarboxamide reveal an unfavorable interaction between the

lone pair of the thiazole N(3) and the proximal carbonyl O-atom (Fig. 9). Although

this interaction exists for the pyrazole N(2), the effect on thiazole is much greater due

to the large S-atom forcing the thiazole N-atom into closer proximity with the

carboxamide. Consequently, the polyamide subunit may twist out of plane about the

amideÀring bond to alleviate electronic strain, as shown in Fig. 9. As with Ht/Py, the

diminished DNA binding affinity of Th/Py may be due to its non-planar conformation.

These effects are not observed for Th in the 1:1 motif.

The negative electronic interaction and hence non-planar conformation should be

alleviated if the N-atom on the back corner of thiazole were replaced with CÀH, as in

the case of thiophene (Tn and Tp). Remarkably, Tn/Py binds to the X A and T sites

with high affinity, but with no observable binding to X G, C. There is a modest 3-fold

preference for T> A. Therefore, the S-atom is not too large to be accommodated within

the tightly packed hairpin ¥ DNA complex. It appears that the large S-atom of

thiophene prefers to sit opposite T not A for the Tn/Py pair in this sequence context.

This experimental finding is opposite the result predicted by Lown and Dickerson from

Helvetica Chimica Acta ± Vol. 85 (2002)

4507

model building [31]. The reduced curvature of thiophene should exacerbate steric

effects between the S-atom and the minor-groove floor, sterically permissive for A, T

recognition but resulting in a >800-fold loss in binding affinity at the X G, C sites. An

ab inito calculation of the Im-Im-Tn-Py subunit reveals a likely steric interaction

between the thiophene Me group and the proximal carboxamide O-atom, which may

force the subunit to twist slightly out of plane. Although structural data on polyamide ¥

DNA complexes reveals a tolerance to small amideÀring dihedrals, a thiophene

without the Me substituent (Tp) and the Tp/Py pair were tested as a control. Tp/Py

exhibited virtually identical DNA recognition behavior as Tn/Py.

In retrospect, in our search for a Hp/Py replacement, our group had great hopes for

the Ht/Py to distinguish T¥ A from A ¥ T. Yet this new Ht ring system was a complete

disappointment. On the other hand, the earlier negative result with the thiazole Th/Py

pair suggested that the S-atom on the corner of the ring was not a viable recognition

element and that the thiophene Tn/Py or Tp/Py pairs were not likely to lead to T¥ A vs.

A ¥ T discrimination. The fact that Tn or Tp prefers to bind opposite T not A is the most

significant unanticipated lead to emerge from the study.

The 1:1 Motif. The 1:1 motif has emerged recently as a way to target certain purine-

rich DNA sequences (e.g., GAGAA) with high affinity [16][17]. Although the cofacial

pairing of rings in the hairpin motif offers a greater chance to differentiate between

WatsonÀCrick base pairs, the single-subunit ¥ DNA complexes of the 1:1 motif provide

a relatively flexible system for the exploration of novel shape-selective recognition

elements. Due to the conformational freedom imparted by the b-alanine residues,

changes in heterocycle geometry do not have such a pronounced impact on the rest of

the molecule. Therefore, specificity may be more difficult to achieve in this motif. In

fact, all 1:1 polyamides described here bind with high affinity to the X A, T sites but

with varying degrees of X A, T> G, C specificity. Structural studies reveal an

important register of amide NH groups with the purine N(3) and pyrimidine OC(2)

groups on the floor of the DNA minor groove [19]. Given this alignment as a driving

force for DNA recognition in the 1:1 motif, one may view the subtle differences in

heterocycle curvature as merely placing the central ring atom (X in Fig. 7) closer to or

farther from the DNA. In this view, increasing the ring curvature decreases the

polyamide ¥ DNA intimacy, thereby diminishing DNA specificity. The results presented

here fit well within this model.

Py and Nh present a H-atom with a positive potential to the minor-groove floor.

Both compounds exhibit a modest 3- to 5-fold selectivity for X A, T> G, C, but Nh

binds with higher affinity to all sites. The selectivity is probably due to the unfavorable

steric X to G-NH2 interaction (X HÀC(3) for Py and HÀN(1) for Nh) postulated

for netropsin and supported by recent NMR studies [1][19]. The higher affinity for Nh

may be attributed to a combination of greater positive charge at HÀN(1) and higher

ring curvature, both of which should reduce specificity.

2

Im, Fr, and Nt present a small atom with an sp lone pair directed toward the minor-

groove floor. Im has been reported previously [16][17], binding all sites with high

affinity and displaying virtually no discrimination between sites. Fr and Nt behave quite

similarly. It is likely that the small atom (N for Im and Nt or O for Fr) presented to the

DNA provides no steric clash with G-NH2, and therefore, all sites are bound with

similarly high affinity.

4

508

Helvetica Chimica Acta ± Vol. 85 (2002)

Hp and Ht present an OH group to the DNA minor groove. In a different

polyamide context, Hp discriminated between A ¥ Tand T¥ A base pairs in the 1:1 motif

30]. In this case, Hp is flanked on both sides by b-alanine residues, and specificity is

[

lost. This loss may be attributed to a larger degree of conformational freedom afforded

to the Hp ring by the aliphatic linkers [17]. Nonetheless, both Hp and Ht exhibit

significant X A, T> G, C specificity, as expected from a negative HOÀC(3) to G-

NH2 steric clash. Ht is slightly more specific, which may result from the non-planarity

of this ligand as discussed above for Ht/Py in the hairpin motif.

2

Tn and Th present an S-atom with an sp lone pair to the DNA minor groove. These

compounds exhibit substantial X A, T> G, C specificity ranging from ꢂ70 to ꢂ

2300-fold. This remarkable selectivity may be attributed to the decreased curvature of

thiazole and thiophene rings, which forces a more intimate interaction of the large S-

atom and the minor-groove floor. In the case of X G, C, this interaction is very

negative, resulting in a dramatic loss in binding affinity. Th is more specific than Tn,

which is probably due to the curvature-induced non-planarity of Im-Im-Tn-Py, as

discussed above for the hairpin motif (Fig. 7).

7.Conclusions. ± Our understanding of the origin of DNA-sequence discrimination

by polyamides has been improved by combining the tools of quantitative DNase-I

footprinting and computational molecular modeling to establish a correlation between

polyamide structure and DNA-sequence specificity (Tables 3 and 4). We believe that

the footprinting results are best explained by differences in the overall heterocycle

structure. Each heterocyclic amino acid has an inherently unique shape, which results in

varying degrees of curvature complementarity between the polyamide and the DNA

minor groove. Given that pyrrole is over-curved with respect to the DNA helix,

reducing heterocycle curvature should increase the polyamide-DNA fit. Consequently,

the polyamide would have greater sensitivity to changes in DNA structure and,

therefore, greater DNA-sequence selectivity. On the other hand, increasing heterocycle

curvature should decrease sensitivity to changes in DNA sequence. In addition, over-

curvature can induce ligand non-planarity deriving from destabilizing eclipsing

interactions. These results suggest that merely considering the functional group facing

Table 3. Pairing Specificity of the Hairpin Motive

Specificity for

Pair

A ¥ T

T¥ A

G ¥ C

C ¥ G

Lone pair to floor

Im/Py

Fr/Py

À

CÀH or NÀH to floor

Py/Py

Nh/Py

Pz/Py

À

CÀOH to floor

Hp/Py

Ht/Py

À

S to floor

Th/Py

Tn/Py

Tp/Py

À

Helvetica Chimica Acta ± Vol. 85 (2002)

4509

C ¥ G

Table 4. Specificity in the 1 : 1 Motif

Pair

A ¥ T

T¥ A

G ¥ C

Im

Py

Hp

Nh

Ht

Fr

Nt

Tn

Th

À

the minor-groove floor is insufficient for an accurate prediction of DNA-recognition

behavior.

Curvature effects are amplified in contiguous-ring polyamides, where continuous p-

conjugation limits conformational flexibility. Furthermore, the packing of cofacial

polyamide subunits in the minor groove, as with the hairpin motif, provides additional

preorganization. The dense functional array offered to the DNA by ring pairs affords

greater promise for sequence discrimination. By contrast, the 1:1 motif offers a high

degree of flexibility but a less dense functional array and, therefore, a lower capacity for

DNA-sequence selectivity.

In retrospect, it is remarkable that a library of eight five-membered heterocycles

(and hence, new heterocycle pairs) reveals very little new leads for sequence

discrimination. This implies that the solution to DNA recognition by Im/Py, Py/Im,

Hp/Py, and Py/Hp could be a narrow structural window. Perhaps the most novel and

useful lead is the possibility that the Tn/Py pair will distinguish T¥ A from A ¥ T. The

next step is to test multiple Tn/Py pairs in different A,T-rich contexts to validate

whether this is a potential breakthrough or not. This study is underway and results will

be reported in due course.

Experimental Part

1. General. N,N-Dimethylformamide (DMF), N,N-diisopropylethylamine (DIEA), thiophenol, Et

2

NH,

N,N-dimethylpropane-1,3-diamine (Dp), Et N, methyl furane-2-carboxylate, oxobutanoic acid, methyl aceto-

3

acetate, cyanoacetic acid, trichloroacetyl chloride, 1H-pyrrole, Na metal, methyl thioglycolate, methyl 2-

chloroacrylate, tin(II) chloride dihydrate, and thiourea were purchased from Aldrich. [(tert-Butoxy)carbonyl]-

b-alanine-(4-carbonylaminomethyl)-benzyl-ester-copoly(styrene-divinylbenzene)resin (Boc-b-Pam resin), di-

cyclohexylcarbodiimide (DCC), 1-hydroxy-1H-benzotriazole (HOBt), 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetra-

methyluronium hexafluorophosphate (HBTU), N,N-dimethylpyridin-4-amine (DMAP), and Boc-b-alanine

3

were purchased from NOVABiochem . CF COOH was purchased from Halocarbon. All other solvents were

reagent grade from EM. Oligonucleotide inserts were synthesized by the Biopolymer Synthesis Center at the

California Institute of Technology. Glycogen (20 mg/ml), dNTPs (PCR nucleotide mix), and all enzymes, unless

otherwise stated, were purchased from Boehringer-Mannheim. pUC19 was purchased from New England

3

2

Biolabs, and deoxyadenosine [g- P]triphosphate was provided by ICN. Calf thymus DNA (sonicated,

deproteinized) and DNase I (7500 units/mL, FPLC pure) were from Amersham Pharmacia. AmpliTaq DNA

polymerase was from Perkin-Elmer and used with the provided buffers. Tris ¥ HCl, DTT, RNase-free H

2

O, and

0

.5m EDTA were from United States Biochemical. Calcium chloride, potassium chloride, and magnesium

chloride were purchased from Fluka. Tris borate EDTA buffer was from GIBCO, and bromophenol blue was

from Acros. All reagents were used without further purification. Column chromatography (CC): silica gel 60,

4

510

Helvetica Chimica Acta ± Vol. 85 (2002)

EM. TLC: J. T. Baker. UV Spectra: Hewlett-Packard 8452A diode array spectrophotometer. IR Spectra: in

À1

cm . NMR Spectra: Varian spectrometer at 300 MHz, (D

6

)DMSO or CDCl

3

solns.; d in ppm rel. to residual

solvent, J in Hz. High-resolution FAB- and EI-MS: recorded at the Mass Spectroscopy Laboratory at the

University of California, Los Angeles; in m/z. Matrix-assisted, laser desorption/ionization time of flight

(

MALDI-TOF)-MS: recorded at the Protein and Peptide Microanalytical Facility at the California Institute of

Technology.

2

. Monomer and −Dimer× Synthesis. 5-Aminofuran-2-carboxylic Acid Methyl Ester (NH -Fr-OMe; 1).

2

Methyl 5-nitrofuran-2-carboxylate (18) was prepared by published methods [21] in 84% yield. TLC (hexanes/

1

13

AcOEt 5 :2). R

(D

f

0.7. H-NMR ((D

6

)DMSO): 7.78 (d, J 3.9, 1 H); 7.59 (d, J 3.9, 1 H); 3.38 (s, 3 H). C-NMR

(

6

)DMSO): 157.7; 152.8; 144.5; 120.4; 113.6; 53.5. EI-MS: 171.117 ( M , C

6

H

5

NO

5

; calc. 171.117).

A mixture of 18 (3 g, 17.5 mmol) and 10% Pd/C (0.3 g) in AcOEt (25 ml) was hydrogenated in a Parr

apparatus at 500 psi and r.t. for 1.5 h (TLC monitoring). The mixture was filtered over a 2.5-cm pad of Celite to

remove Pd/C. The filtrate was cooled to À208, and hexanes were added until a white precipitate was formed.

The precipitate (2.1 g) was collected by vacuum filtration and washed with Et

2

O: 1 (85%). TLC (hexanes/

1

AcOEt 5 :2): R

(D

f

0.25. IR (film): 3395, 3322, 1684, 1627, 1528, 1441, 1338, 1297, 1199, 1153. H-NMR

1

3

(

1

6

)DMSO): 7.13 (d, J 3.6, 1 H); 6.60 (s, 2 H); 5.09 (d, J 3.6, 1 H); 3.66 (s, 3 H). C-NMR (CDCl

3

): 123.8;

23.0; 121.9; 88.3; 86.3. EI-MS: 141.042 (M , C

6

H

7

NO

3

; calc. 141.042).

Methyl 5-{[(1-Methyl-4-nitro-1H-imidazol-2-yl)carbonyl]amino}furan-2-carboxylate (NO

2

-Im-Fr-OMe;

1

9). A mixture of 1 (1 g, 7.08 mmol), NO -Im-COCCl (2.31 g, 8.5 mmol), DIEA (1.1 g, 1.48 ml, 8.5 mmol),

2

3

and AcOEt (14 ml) was stirred at 358 for 12 h. The mixture was cooled to r.t., and sufficient hexanes were added

to completely precipitate a pale yellow solid. The precipitate was collected by vacuum filtration and washed with

cold MeOH and Et

2

O: 19 (1.18 g, in 57%). TLC (hexanes/AcOEt 5 :2): R

f

0.20. IR (film): 3195, 3133, 1731,

1

3

1

676, 1553, 1541, 1525, 1442, 1387, 1313, 1146. H-NMR ((D

6

13

)DMSO): 12.17 (br., 1 H); 8.65 (s, 1 H); 7.34 (d, J

.6, 1 H); 6.52 (d, J 3.6, 1 H); 4.01 (s, 3 H); 3.77 (s, 3 H). C-NMR ((D

6

)DMSO): 158.6; 155.8; 150.0; 145.0;

37.3; 137.0; 127.8; 121.4; 99.0; 52.3; 37.3. EI-MS: 294.060 (M , C11

H

10

N

4

O

6

; calc. 294.060).

Methyl 5-{[(4-Amino-1-methyl-1H-imidazol-2-yl)carbonyl]amino}furan-2-carboxylate Hydrochloride

HCl ¥ H N-Im-Fr-OMe; 20). A mixture of 19 (1 g, 3.4 mmol) and 10% Pd/C (0.2 g) in AcOEt (7 ml) was

hydrogenated in a Parr apparatus at 500 psi and r.t. for 1.5 h. The mixture was filtered over a 2.5-cm pad of Celite

to remove the Pd/C. The AcOEt was evaporated and 2m HCl in Et O was added to give 20 (0.62 g, 61%). TLC

0.15 (amine), 0.0 (salt). IR (film): 2985, 3008, 1708, 1689, 1537, 1318, 1193, 1142, 1119,

(

2

(

hexanes/AcOEt 5 :2): R

018, 756, 668. H-NMR ((D

.96( s, 3 H); 3.76( s, 3 H). C-NMR ((D

1.6; 35.8. EI-MS: 264.086 ( M , C11

f

1

3

5

6

13

)DMSO): 11.73 (br., 1 H); 7.54 (s, 1 H); 7.33 (d, J 3.6, 1 H); 6.51 (s, J 3.6, 1 H);

6

)DMSO): 157.9; 155.0; 149.5; 136.3; 134.8; 129.5; 120.9; 118.7; 97.9;

H

12

N

4

O

4

; calc. 264.086).

Methyl 5-{{{4-{[(tert-Butoxy)carbonyl]amino}-1-methyl-1H-imidazol-2-yl}carbonyl}amino}furan-2-car-

boxylate (Boc-Im-Fr-OMe; 21). A mixture of 20 (0.5 g, 1.6mmol), (Boc) O (545 mg, 2.4 mmol), and DIEA

258 mg, 348 ml, 2 mmol) in DMF (5 ml) was stirred at 608 for 18 h. The mixture was added to ice-water (0.5 l)

and the precipitate extracted with AcOEt (50 ml). The org. layer was dried (Na SO ) and evaporated, and the

2

(

2

4

resulting residue was subjected to CC (hexanes/AcOEt 1:1): 21 (376mg, 62 %). White solid. TLC (hexanes/

1

AcOEt 1:1): R

f

0.65. IR (film): 3243, 2978, 1722, 1577, 1533, 1436, 1368, 1311, 1163, 1138, 755. H-NMR

(

7

(D

s, 3 H); 1.44 (s, 9 H). C-NMR ((D

9.8; 52.3; 35.9; 28.9. EI-MS: 364.138 (M , C16

6

)DMSO): 11.11 (s, 1 H); 9.52 (s, 1 H); 7.33 (d, J 3.3, 1 H); 6.47 (d, J 3.3, 1 H); 3.90 (s, 3 H); 3.76

13

6

)DMSO): 158.6; 156.1; 153.6; 150.4; 137.7; 136.8; 133.1; 121.7; 115.3; 97.8;

H

20

N

4

O

6

; calc. for 364.138).

5

-{{{4-{[(tert-Butoxy)carbonyl]amino}-1-methyl-1H-imidazol-2-yl}carbonyl}amino}furan-2-carboxylic

Acid (Boc-Im-Fr-OH; 15). A mixture of 21 (0.3 g, 0.82 mmol), 1n NaOH (5 ml), and MeOH (1 ml) was stirred

at r.t. for 3 h (TLC monitoring). The MeOH was evaporated and the aq. layer carefully adjusted to pH 2 with 1n

HCl. The milky white precipitate was extracted with AcOEt, and the extract was dried (Na

2

4

SO ) and

evaporated: 15 (270 mg, 94%). Fine white powder. TLC (hexanes/AcOEt 1:1, 10% AcOH): R

f

0.5. IR (film):

1

3

231, 2917, 2856, 1688, 1542, 1311, 1259, 1163, 1119. H-NMR ((D

6

)DMSO): 10.94 (s, 1 H); 9.52 (s, 1 H); 7.30

1

3

(

1

s, 1 H); 7.22 (d, J 3.6, 1 H); 6.44 (d, J 3.6, 1 H); 3.91 (s, 3 H); 1.44 (s, 9 H). C-NMR ((D

6

)DMSO): 160.1;

59.3; 149.9; 136.5; 135.5; 133.2; 123.6; 115.4; 101.4; 97.0; 82.3; 35.9; 28.9. EI-MS: 350.123 (M , C15

calc. 350.123).

Methyl 5-{{3-{[(tert-Butoxy)carbonyl]amino}propanoyl}amino}furan-2-carboxylate (Boc-b-Fr-OMe; 22).

A mixture of Boc-b-alanine (3.22 g, 17 mmol) and DCC (1.75 g, 8.5 mmol) in CH Cl (25 ml) was stirred at r.t.

for 30 min. To the above mixture was added 1 (0.6g, 4.25 mmol) as a soln. in DMF (5 ml) and DIEA (0.741 ml,

.55 g, 4.25 mmol), followed by the addition of DMAP (0.155 g, 1.27 mmol). The mixture was heated to 408,

stirred overnight, and then filtered to remove the dicyclohexylurea. The filtrate was poured into ice-water (0.5 l)

H

18

N

4

O

6

;

2

0

Helvetica Chimica Acta ± Vol. 85 (2002)

4511

upon which time a crude white precipitate formed. The crude precipitate was extracted with AcOEt and

subjected to CC (hexanes/AcOEt 5 :2): 22 (1.1 g, 85%). Flaky white solid. TLC (hexanes/AcOEt 5 :2) R

f

0.2.

1

IR (film): 3372, 3234, 3036, 1957, 1728. H-NMR ((D

6

)DMSO): 11.58 (s, 2 H); 7.29 (d, J 3.8, 1 H); 6.86

13

(

s, 2 H); 6.36 (d, J 3.8, 1 H); 3.76( s, 3 H); 3.32 (q, J 6, 2 H); 2.69 (t, J 6, 3 H); 1.35 (s, 9 H). C-NMR

(D

6

)DMSO): 167.9; 159.4; 155.8; 150.1; 137.0; 120.8; 95.2; 78.3; 51.6; 36.1; 35.8; 29.0. EI-MS: 312.132 (M ,

C

14

H

20

N

2

O

6

; calc. 312.132).

5-{{3-{[(tert-Butoxy)carbonyl]amino}propanoyl}amino}furan-2-carboxylic Acid (Boc-b-Fr-OH; 13). As

described for 15, with 22 (1.1 g, 3.52 mmol), 1n NaOH (15 ml), and MeOH (5 ml) for 4 h: 13 (0.97 g, 92%). Off-

1

white solid. TLC (hexanes/AcOEt 5 :2, 10% AcOH): R

f

0.6. IR (film): 3321, 3270, 3979, 1684, 1522. H-NMR

(

3

(D

6

)DMSO): 11.28 (s, 2 H); 6.98 (d, J 3.6, 1 H); 6.82 (s, 2 H); 6.23 (d, J 3.6, 1 H); 3.19 (q, J 6, 2 H); 2.45

13

t, J 6, 2 H); 1.35 (s, 9 H). C-NMR ((D

6

)DMSO): 168.7; 159.7; 156.0; 150.6; 137.1; 120.8; 95.8; 78.3; 36.9;

6.6; 29.0. EI-MS: 298.117 (M , C13

H

18

N

2

O

6

; calc. 298.116).

2,2,2-Trichloro-1-(1H-pyrrol-2-yl)ethan-1-one (23). A soln. of 1H-pyrrole (20.6ml, 20 g, 298 mmol) and

Et

2

O (86ml) was added dropwise to trichloroacetyl chloride (71.9 ml, 117 g, 64 4 mmol) with stirring at 0 8. The

mixture was allowed to warm to r.t. and stirred overnight. The solvent was evaporated, and the residue was

reprecipitated from hexanes: 23 (24 g, 38%). White solid. TLC (hexanes/AcOEt 5 :2): R

f

0.75. IR (film): 3322,

1

656, 1388, 1136, 1035, 953, 842, 808, 754, 733, 688. H-NMR ((D

6

)DMSO): 12.4 (s, 1 H); 7.32 (m, J 2.1, 1 H);

1

3

7

.29 (m, J 2.1, 1 H); 6.34 (m, J 2.1, 1 H). C-NMR ((D

6

)DMSO): 172.5; 130.0; 122.3; 121.9; 112.0; 95.9. EI-

MS: 210.936( M , C

,2,2-Trichloro-1-(5-nitro-1H-pyrrol-2-yl)ethan-1-one (24). A soln. of 23 (20 g, 95 mmol) and Ac

111 ml) was cooled to À408 and treated dropwise with 70% nitric acid (8.24 ml) over 2 h. After completion of

6 4 3

H Cl NO ; calc. for 210.936).

2

O

(

addition, the mixture was warmed to r.t. over 2 h and then cooled back down to À408. Sufficient ice-water was

added to pecipitate 24 (16.5 g, 68%). White solid. TLC (hexanes/AcOEt 5 :2): R

f

0.6. IR (film): 3316, 1676,

1

551, 1518, 1405, 1379, 1317. H-NMR ((D

6

)DMSO): 13.62 (s, 1 H); 8.33 (d, J 3, 1 H); 7.66 (d, J 3, 1 H).

1

3

C-NMR ((D

55.921).

6

)DMSO): 173.4; 137.6; 128.4; 122.0; 115.0; 94.4. EI-MS: 255.921 ( M , C

6

H

3

Cl

3

N

2

O

3

; calc.

2

Ethyl 5-Nitro-1H-pyrrole-2-carboxylate (NO

35 ml) at r.t. was added EtONa (4 g, 59 mmol). The mixture was stirred for 2 h, then quenched with H

2

-Nh-OEt; 25). To a mixture of 24 (10 g, 39 mmol) in EtOH

SO , and

(

2

4

cooled to 08. Ice-water was added (0.5 l) to precipitate 25 (7 g, 97%). Tan solid. TLC (hexanes/AcOEt 5 :2):

1

R

(

f

0.5. IR (film): 3263, 3152, 2979, 1687, 1565, 1508, 1365, 1323, 1207, 1017, 752. H-NMR ((D

6

)DMSO): 8.04

1

3

d, J 1.5, 1 H); 7.23 (d, J 1.5, 1 H); 4.29 (q, J 7.2, 2 H); 1.29 (t, J 7.2, 3 H). C-NMR ((D

6

)DMSO): 160.0;

1

37.2; 124.9; 123.6; 110.1; 61.4; 14.9. EI-MS: 184.048 ( M , C

7

H

8

N

2

O

4

; calc. 184.048).

N-Nh-OEt; 2). A mixture of 25 (3 g,

6mmol) and 10% Pd/C (0.3 g) in AcOEt (25 ml) was hydrogenated in a Parr apparatus at 500 psi and r.t. for

Ethyl 5-Amino-1H-pyrrole-2-carboxylate Hydrochloride (HCl ¥ H

2

1

.5 h. The mixture was filtered over a 2.5-cm pad of Celite to remove Pd/C. The filtrate was cooled to 208, and

HCl in Et O was added. Upon addition, the hydrochloride salt precipitated and was collected by vacuum

filtration: 2 (2.4 g, 78%). White solid. TLC (hexanes/AcOEt 1:1): R 0.15 (amine), 0.0 (hydrochloride). IR

2

f

1

(

1

(

6

film): 2914, 1694, 1495, 1429, 1376, 1345, 1284, 1224, 1106, 1020, 965. H-NMR ((D )DMSO): 12.21 (s, 1 H);

1

3

0.13 (s, 1 H); 7.10 (d, J 1.8, 1 H); 6.74 (d, J 1.8, 1 H); 4.24 (q, J 7.2, 2 H); 1.26( t, J 7.2, 3 H). C-NMR

(D

6

)DMSO): 159.6; 121.3; 117.7; 115.6; 109.2; 60.0; 14.3. EI-MS: 154.074 ( M , C

7

H

10

N

2

O

2

; calc. 154.074).

Ethyl 5-{[(tert-Butoxy)carbonyl]amino}-1H-pyrrole-2-carboxylate (Boc-Nh-OEt; 26). As described for 21,

with 2 (2 g, 11 mmol), (Boc) O (3.6g, 1 6. 5 mmol), DIEA (2.1 ml, 1.56g, 12.1 mmol), and DMF (15 ml) for

2 h. Workup with ice-water (1 l) and AcOEt (150 ml). CC (hexanes/AcOEt 5 :2) afforded 26 (2 g, 72%). Flaky

2

1

white solid. TLC (hexanes/AcOEt 5 :2): R

f

0.65. IR (film): 3296, 1683, 1570, 1384, 1315, 1264, 1249. H-NMR

(

(D

6

)DMSO): 11.48 (s, 1 H); 9.06( s, 1 H); 6.93 (d, J 1.8, 1 H); 6.58 (d, J 1.8, 1 H); 4.21 (q, J 7.2, 2 H); 1.41

1

3

s, 9 H); 1.24 (t, J 7.2, 3 H). C-NMR ((D

6

)DMSO): 160.1; 152.5; 119.1; 112.4; 105.2; 78.3; 59.4; 28.2; 14.4. EI-

MS: 254.127 (M , C12

H

18

N

2

O

4

; calc. 254.127).

5-{[(tert-Butoxy)carbonyl]amino}-1H-pyrrole-2-carboxylic Acid (Boc-Nh-OH; 12). As described for 15

with 26 (2 g, 7.9 mmol), 1n NaOH (15 ml), and MeOH (5 ml) for 3 h: 12 (1.6g, 92%). Off-white solid. TLC

(

1

(

C

hexanes/AcOEt 5 :2, 10% AcOH): R

f

0.5. IR (film): 3329, 3153, 2969, 1691, 1586, 1549, 1434, 1374, 1250, 1167,

1

117, 1057, 961, 762. H-NMR ((D

6

)DMSO): 11.32 (s, 1 H); 9.01 (s, 1 H); 6.88 (s, 1 H); 6.52 (s, 1 H); 1.41

1

3

s, 9 H). C-NMR ((D

6

)DMSO): 162.3; 153.3; 125.4; 120.7; 112.6; 106.0; 79.0; 28.9. EI-MS: 226.095 (M ,

10

H

14

N

2

O

4

; calc. 226.095).

Methyl 4-Cyano-3-methylbut-3-enoate (27). A mixture of acetoacetate (30 g, 258 mmol), cyanoacetic acid

24 g, 284 mmol), NH OAc (3.98 g, 51.6 mmol), AcOH (6.65 ml, 6.98 g, 116 mmol), and benzene (75 ml) was

stirred for 12 h at 1458 (round-bottom flask, DeanÀStark apparatus, condenser). The mixture was allowed to

(

4

512

Helvetica Chimica Acta ± Vol. 85 (2002)

4

soln. (0.3 l), dried (MgSO ), and evaporated. The crude

cool to r.t., washed with brine (0.3 l), sat. NaHCO

3

product was distilled at 608/0.1 Torr: 27 (23 g, 65%), (E)/(Z) mixture. Clear liquid. IR (film): 2957, 2221, 1741,

1

(

1

437. H-NMR ((D

6

)DMSO): 5.69 (q, J 0.6, 1 H); 5.62 (q, J 0.6, 1 H); 3.61 (s, 3 H); 3.60 (s, 3 H); 3.42

13

s, 2 H); 3.35 (d, J 1.2, 2 H); 2.01 (d, J 1.2, 3 H); 1.93 (d, J 1.2, 3 H). C-NMR ((D

6

)DMSO): 170.1; 169.5;

58.4; 158.1; 117.4; 117.3; 99.6; 99.4; 52.8; 52.7; 42.8; 41.3; 23.6; 21.7. EI-MS: 139.063 ( M , C

39.063).

Methyl 5-Amino-3-methylthiophene-2-carboxylate Hydrochloride (HCl ¥ H

18.7 ml, 13.2 g, 181 mmol) was added dropwise to a mixture of 27 (23 g, 165 mmol) and S-flakes (5.28 g,

65 mmol) in EtOH (130 ml) and stirred at r.t. for 3 h. The mixture was concentrated to a minimal volume and

placed in an ice bath. Conc. HCl was slowly added to the mixture to give a light orange solid. The precipitate was

collected by vacuum filtration and washed repeatedly with Et O: 3 (19 g, 68%). TLC (hexanes/AcOEt 5 :2):

0.55 (amine), 0.0 (hydrochloride). IR (film): 3422, 3339, 3204, 2849, 1713, 1677, 1546, 1462, 1269, 1187, 1092.

7

H

9

NO

2

; calc.

2

N-Tn-OMe; 3). Et

2

NH

(

1

2

R

1

f

13

H-NMR ((D

46.7; 145.5; 114.9; 114.7; 52.0; 16.6. EI-MS: 171.035 (M , C

Methyl 3-Methyl-5-{[(1-methyl-4-nitro-1H-imidazol-2-yl)carbonyl]amino}thiophene-2-carboxylate (NO

6 6

)DMSO): 6.91 (s, 2 H); 5.76( s, 1 H); 3.61 (s, 3 H); 2.62 (s, 3 H). C-NMR ((D )DMSO): 163.5;

7 9 2

H NO S ; calc. 171.035).

1

2

-

Im-Tn-OMe; 28). As described for 19, from 3 (1 g, 4.8 mmol): 28 (0.87 g, 56%). Yellow solid. TLC (hexanes/

1

AcOEt 5 :2): R

8

1

f

0.60. IR (film): 3125, 1649, 1543, 1506, 1382, 1312, 1267. H-NMR ((D

6

)DMSO): 12.43 (s, 1 H);

.66 (s, 1 H); 6.95 (s, 1 H); 4.03 (s, 3 H); 3.79 (s, 3 H); 2.41 (s, 3 H). ¹³C-NMR ((D

44.7; 143.6; 136.9; 128.0; 118.8; 117.2; 52.2; 37.5; 16.6. EI-MS: 324.053 (M , C12

)DMSO): 163.5; 155.7; 145.1;

6

H

12

N

4

O

5

S ; calc. 324.053).

Methyl 5-{[(4-Amino-1-methyl-1H-imidazol-2-yl)carbonyl]amino}-3-methylthiophene-2-carboxylate hy-

drochloride (HCl ¥ H N-Im-Tn-OMe; 29). As described for 20, from 28 (0.5 g, 1.5 mmol): 29 (321 mg, 63%).

2

Pale yellow solid. TLC (hexanes/AcOEt 5 :2): R

f

0.25 (amine), 0.0 (hydrochloride). IR (film): 3294, 1735, 1674,

)DMSO): 12.14 (s, 1 H); 7.47 (s, 1 H); 6.95 (s, 1 H); 3.99

1

562, 1520, 1440, 1407, 1267, 1185, 1091. H-NMR ((D

6

1

3

(

s, 1 H); 3.73 (s, 1 H); 2.39 (s, 1 H). C-NMR ((D

6

)DMSO): 177.9; 63.4; 157.6; 155.1; 144.1; 128.0; 118.6; 116.1;

; calc. for 294.079).

5

2.2; 35.4; 16.4. EI-MS: 294.079 (M , C12

H

14

N

4

O

3

Methyl 5-{{{4-{[(tert-Butoxy)carbonyl]amino}-1-methyl-1H-imidazol-2-yl}carbonyl}amino}-3-methylthio-

phene-2-carboxylate (Boc-Im-Tn-OMe; 30). As described for 21, from 29 (300 mg, 0.91 mmol): 30 (239 mg,

6

1

7%). Pale yellow solid. TLC (hexanes/AcOEt 1:1): R

f

0.7. IR (film): 3424, 3219, 2995, 1750, 1704, 1677, 1571,

1

251, 1141. H-NMR ((D

6

)DMSO): 11.78 (s, 1 H); 9.33 (s, 1 H); 7.32 (s, 1 H); 6.90 (s, 1 H); 3.91 (s, 1 H); 3.73

1

3

(

1

s, 1 H); 2.39 (s, 1 H); 1.43 (s, 9 H). C-NMR ((D

15.6; 115.3; 79.4; 51.3; 35.2; 26.9; 15.9. EI-MS: 394.131 ( M , C17

-{{{4-{[(tert-Butoxy)carbonyl]amino}-1-methyl-1H-imidazol-2-yl}carbonyl}amino}-3-methylthiophene-2-

carboxylic Acid (Boc-Im-Tn-OH; 16). As described for 15, with 30 (200 mg, 0.507 mmol), MeOH (1 ml), and 1n

6

)DMSO): 162.8; 155.7; 152.8; 144.0; 143.4; 136.7; 132.4; 117.0;

22 4 5

H N O S ; calc. 394.131).

5

NaOH (5 ml) at 608 for 6h: 16 (158 mg, 82%). Pale tan solid. TLC (hexanes/AcOEt 1:1, 10% AcOH): R

f

0.8.

1

IR (film): 3400, 2976, 3231, 2961, 1722, 1678, 1589, 1253, 1179, 1091. H-NMR ((D

6

)DMSO): 11.68 (s, 1 H); 9.35

1

3

(

6

s, 1 H); 7.32 (s, 1 H); 6.87 (s, 1 H); 3.93 (s, 3 H); 2.38 (s, 3 H); 1.44 (s, 9 H). C-NMR ((D )DMSO): 164.0;

20 4 5

H N O S ;

1

58.3; 155.6; 142.8; 136.7; 117.3; 117.1; 115.2; 99.4; 81.4; 35.2; 28.1; 15.8. EI-MS: 380.115 ( M , C16

calc. 380.115).

Methyl 5-{{3-{[(tert-Butoxy)carbonyl]amino}propanoyl}amino}-3-methylthiophene-2-carboxylate (Boc-b-

Tn-OMe; 31). A mixture of Boc-b-alanine (1 g, 5.28 mmol) and DCC (545 mg, 2.64 mmol) in CH Cl (10 ml)

was stirred at r.t. for 30 min. The mixture was then filtered into a round-bottom flask containing 3 (382 mg,

.8 mmol), DIEA (322 ml, 239 mg, 1.8 mmol), DMAP (100 mg, 0.8 mmol), and DMF (8 ml). The mixture was

2

1

heated at 458 (TLC monitoring). Additional symmetrical anhydride (1.4 equiv.) was added every 8 h, as needed,

until completion of the reaction. The mixture was then added to brine (0.2 l) and extracted twice with AcOEt

(

3 2 4

50 ml). The org. layer was washed with sat. NaHCO soln. (0.1 l) and 10 mm HCl (0.1 l), dried (Na SO ), and

evaporated, and the crude residue subjected to CC (hexanes/AcOEt 5 :2): 31 (392 mg, 62%). Flaky white

1

powder. TLC (hexanes/AcOEt 5 :2): R

f

0.32. IR (film): 3348, 3450, 2981, 1684, 1568, 1522, 1445, 1252. H-NMR

(

CDCl

C-NMR (CDCl

42.124 (M , C15

3

): 10.11 (s, 1 H); 6.49 (s, 1 H); 3.79 (s, 3 H); 3.49 (q, J 6, 2 H); 2.64 (t, J 6, 2 H); 1.40 (s, 9 H).

1

3

): 168.7; 164.1; 157.0; 144.9; 143.4; 128.4; 117.3; 116.1; 80.4; 51.7; 37.0; 34.8; 16.4. EI-MS:

3

H

22

N

2

O

5

S ; calc. 342.124).

5

-{{3-{[(tert-Butoxy)carbonyl]amino}propanoyl}amino}-3-methylthiophene-2-carboxylic Acid (Boc-b-Tn-

OH; 14). As described for 15, with 31 (200 mg, 0.58 mmol), MeOH (4 ml), and 1n NaOH (15 ml) at 608 for 6h:

1

2

4 (180 mg, 94%). Light yellow solid. TLC (hexanes/AcOEt 5 :2, 10% AcOH): R

f

0.5. IR (film): 3255, 2976,

1

954, 1674, 1569, 1522, 1445, 1253. H-NMR ((D

6

)DMSO): 12.35 (s, 1 H); 11.42 (s, 1 H); 6.89 (s, 1 H); 6.48

1

3

(

s, 1 H); 3.21 (q, J 6, 2 H); 2.50 (t, J 6, 2 H); 2.37 (s, 3 H); 1.35 (s, 9 H). C-NMR ((D

6

)DMSO): 168.2;

Helvetica Chimica Acta ± Vol. 85 (2002)

4513

1

3

64.1; 155.4; 143.3; 143.1; 116.8; 115.0; 77.7; 36.3; 35.6; 28.2; 15.7. EI-MS: 328.109 (M , C14H N O S ; calc.

28.109).

20 2 5

5-Nitrothiophene-2-carboxylic Acid (NO

2

-Tp-OH; 32). A mixture of NaOCl (26.35 g, 291 mmol) and

Na

2

HPO ¥ H O (30.3 g, 219 mmol) in H O (250 ml) was added dropwise to a soln. of commercially available 5-

4

2

nitrothiophene-2-carboxaldehyde (5 g, 31.8 mmol) in acetone (0.6l) at r.t. Upon completion of addition, TLC

showed total consumption of the starting aldehyde. The mixture was washed with hexanes (0.1 l) and acidified to

pH 2 with 1n HCl. The mixture was extracted with Et

evaporated: 32 (3.74 g, 68%). White solid. TLC (hexanes/AcOEt 5 :2, 10% AcOH): R

2

O (3 Â0.1 l) and the extract dried (Na

2

SO

4

) and

f

0.55. IR (film): 3118,

1

3

1

109, 2876, 1688, 1680, 1512, 1350, 1336, 1274. H-NMR ((D

H). C-NMR ((D

72.978).

6

)DMSO): 8.12 (d, J 4.2, 1 H); 7.73 (d, J 4.2,

1

3

6 5 3 4

)DMSO): 162.2; 154.5; 141.0; 132.6; 130.5. EI-MS: 172.978 (M , C H NO S ; calc.

Methyl 5-Nitrothiophene-2-carboxylate (NO

2

-Tp-OMe; 33). A mixture of 32 (3.5 g, 20.2 mmol), conc.

H

2

SO (0.2 g, 110 ml, 2.0 mmol), and MeOH (50 ml) was refluxed for 48 h. The MeOH was evaporated and the

4

residue neutralized with 1n NaOH. The mixture was extracted with AcOEt (2 Â0.1 l) and the extract dried

(

Na

2 4 f

SO ) and evaporated: 33 (3.4 g, 91%). Crystalline white solid. TLC (hexanes/AcOEt 1:1): R 0.8. IR

1

film): 3476, 3115, 1730, 1705, 1535, 1508, 1423, 1360, 1282, 1250, 1191, 997, 856, 748, 732. H-NMR

1

3

(D

38.4; 133.3; 130.4; 54.0. EI-MS: 186.994 (M , C

Methyl 5-Aminothiophene-2-carboxylate Hydrochloride (HCl ¥ H

5.8 ml) was added dropwise to a mixture of 33 (0.3 g, 1.6mmol) and SnCl

5% EtOH (5.8 ml) at r.t. Sufficient cooling was necessary to keep the reaction temp. under 358. The mixture

6

)DMSO): 8.21 (d, J 3.9, 1 H); 7.80 (d, J 3.9, 1 H); 3.87 (s, 3 H). C-NMR ((D

6

)DMSO): 161.2; 155.0;

1

6

H

5

NO

4

S ; calc. 186.994).

N-Tp-OMe; 34). Conc. HCl soln.

¥ 2 H O (2.43 g, 12.8 mmol) in

2

(

9

2

was stirred at 358 for 2 h. The EtOH was evaporated and the aq. layer washed with hexanes (2 Â50 ml). The aq.

layer was neutralized with 1n NaOH to pH 9 ( !milky white emulsion), the mixture was extracted several times

2 4

with AcOEt (50 ml), and the org. phase was dried (Na SO ) and evaporated to give a thin yellow film. Addition

of 2m HCl in Et

2

O gave 34 (220 mg, 71%). White solid. TLC (hexanes/AcOEt 1 :1): R

f

0.55 (amine), 0.0

1

(

hydrochloride). IR (film): 3219, 1731, 1706, 1471, 1272, 1088, 739. H-NMR ((D

6

)DMSO): 7.33 (d, J 4.5,

1

3

1

H); 6.70 (s, 2 H); 5.87 (d, J 4.5, 1 H); 3.65 (s, 3 H). C-NMR ((D

6

)DMSO): 163.5; 162.8; 136.2; 105.1; 51.8.

EI-MS: 157.020 (M , C

6

H

7

NO

2

S; calc. 157.020).

Methyl 5-{[(1-Methyl-4-nitro-1H-imidazol-2-yl)carbonyl]amino}thiophene-2-carboxylate (NO

2

-Im-Tp-

OMe; 35). As described for 19, from 34 (200 mg, 1.0 mmol): 35 (164 mg, 51%). Yellow solid. TLC (hexanes/

1

AcOEt 5 :2): R

f

0.50. IR (film): 3133, 1698, 1672, 1560, 1543, 1521, 1455, 1379, 1313, 1265, 1098, 746.9. H-NMR

(

(D

6

)DMSO): 12.55 (s, 1 H); 8.68 (s, 1 H); 7.6 4 (d , J 3.9, 1 H); 7.16( d, J 3.9, 1 H); 4.04 (s, 3 H); 3.77 (s, 3 H).

1

3

C-NMR ((D

6

)DMSO): 163.1; 155.6; 145.8; 136.8; 132.6; 128.0; 123.5; 115.3; 52.6; 37.5. EI-MS: 310.037 (M ,

C

11

H

10

N

4

O

5

S ; calc. 310.037).

Methyl 5-{[(4-Amino-1-methyl-1H-imidazol-2-yl)carbonyl]amino}thiophene-2-carboxylate Hydrochloride

HCl ¥ H N-Im-Tp-OMe; 36). As described for 20, from 35 (150 mg, 0.48 mmol): 36 (107 mg, 70%). Off-white

solid. TLC (AcOEt): R

(

2

f

0.45 (amine), 0.0 (hydrochloride). IR (film): 3344, 3204, 2954, 1691, 1673, 1561, 1511,

1

(

1

458, 1438, 1343, 1275, 1100. H-NMR ((D

d, J 4.2, 1 H); 3.99 (s, 3 H); 3.76( s, 3 H). C-NMR ((D

6

)DMSO): 12.20 (s, 1 H); 7.6 3 (d , J 4.2, 1 H); 7.40 (s, 1 H); 7.11

13

6

)DMSO): 177.4; 162.3; 159.4; 155.0; 145.4; 131.8;

12 4 3

H N O S ; calc. 280.063).

28.5; 113.8; 51.7; 35.6. EI-MS: 280.063 (M , C11

Methyl 5-{{{4-{[(tert-Butoxy)carbonyl]amino}-1-methyl-1H-imidazol-2-yl}carbonyl}amino}thiophene-2-

carboxylate (Boc-Im-Tp-OMe; 37). As described for 21, from 36 (100 mg, 0.35 mmol): 37 (90 mg, 66%).

White solid. TLC (hexanes/AcOEt 1 :1): R

f

0.75. IR (film): 3282, 2964, 1707, 1692, 1673, 1573, 1550, 1368, 1341,

1

273, 1159, 1096. H-NMR ((D

6

)DMSO): 11.91 (s, 1 H); 9.32 (s, 3 H); 7.6 2 (d , J 4.2, 1 H); 7.33 (s, 1 H); 7.07

13

(

d, J 4.2, 1 H); 3.94 (s, 3 H); 3.76( s, 3 H); 1.44 (s, 9 H). C-NMR ((D

36.7; 132.4; 131.8; 122.0; 115.4; 113.5; 51.7; 35.2; 28.1. EI-MS: 380.115 (M , C16

-{{{4-{[(tert-Butoxy)carbonyl]amino}-1-methyl-1H-imidazol-2-yl}carbonyl}amino}thiophene-2-carboxyl-

ic acid (Boc-Im-Tp-OH; 17). As described for 15, from 37 (90 mg, 0.23 mmol): 17 (80 mg, 91%). White solid.

6

)DMSO): 162.3; 155.5; 152.8; 145.7;

20 4 5

H N O S ; calc. 380.115).

1

5

TLC (hexanes/AcOEt 5 :2, 10% AcOH): R

f

0.6. IR (film): 3246, 2965, 1674, 1556, 1509, 1456, 1314, 1271, 1237,

1

162, 1102, 1023, 750. H-NMR ((D

6

)DMSO): 11.79 (s, 1 H); 9.31 (s, 1 H); 7.52 (d, J 3.9, 1 H); 7.32 (s, 1 H);

13

7

1

.04 (d, J 3.9, 1 H); 3.93 (s, 3 H); 1.43 (s, 9 H). C-NMR ((D

6

)DMSO): 163.4; 155.4; 152.8; 145.2; 136.7; 132.5;

18 4 5

H N O S ; calc. 366.100).

31.3; 123.8; 115.2; 113.4; 79.1; 35.2; 28.1. EI-MS: 366.100 (M , C15

Methyl 3-Hydroxythiophene-2-carboxylate (Ht-OMe; 38). To dry MeOH (81 ml) under N

metal (3.68 g, 304 mmol). After H evolution had stopped, the soln. was cooled to 08, and methyl thioglycolate

10 g, 179 mmol) was added dropwise. Methyl 2-chloroacrylate (10.88 g, 179 mmol) in MeOH (21 ml) was then

added dropwise ( !cloudy yellow precipitate). The soln. was allowed to warm to r.t. and stirred for 2 h,

2

was added Na

2

(

4

514

Helvetica Chimica Acta ± Vol. 85 (2002)

whereupon the precipitate turned dark brown. Evaporation gave a dark yellow solid which was treated with 4n

Cl

), and evaporated, and the residual dark oil was subjected to CC (hexanes/AcOEt 20 :1):

HCl ( !pH 2). The aq. layer was extracted with CH

2

(3 Â150 ml), the org. layer was washed with H O (3 Â

2

1

3

1

50 ml), dried (MgSO

4

8 (18.4 g, 64.7%). Clear crystalline solid. TLC (hexanes/AcOEt 20 :1): R

f

0.47. IR (film): 3334, 3112, 2955,

): 9.58 (s, 1 H); 7.59 (d, J 5.7,

1

716, 1664, 1552, 1444, 1415, 1350, 1296, 1208, 1104, 1032, 781. H-NMR (CDCl

3

1

3

H); 6.75 (d, J 4.8, 1 H); 3.90 (s, 3 H). C-NMR (CDCl

3 6 6

): 131.7; 119.4; 52.2. EI-MS: 158.004 (M , C H O S ;

calc. 158.004).

Methyl 3-Hydroxy-4-nitrothiophene-2-carboxylate (NO

to conc. H SO

(48 ml) and stirred until homogeneous. The soln. was then cooled to À10 to 08, and HNO

4.3 ml) in conc. H SO (24 ml) was added dropwise with sufficient cooling to keep the temp. below 08. After the

addition, the soln. was stirred at 08 for 3 h. The resulting black soln. was added to ice and extracted with CH Cl

) and evaporated, and the resulting residue was chromatographed

2

-Ht-OMe; 39). Ester 38 (15 g, 98 mmol) was added

2

4

3

(

2

4

2

(

3 Â150 ml), the extract was dried (MgSO

4

silica gel; hexanes/AcOEt 5 :1): 39 (8.2 g, 37.6%). Yellow solid. TLC (hexanes/AcOEt 5 :1): R

f

0.12. IR (film):

1

3

107, 1674, 1561, 1520, 1446, 1368, 1267, 1212, 1127, 974, 900, 842, 773. H-NMR (CDCl

3

): 10.15 (s, 1 H); 8.43

1

3

(

s, 1 H); 3.97 (s, 3 H). C-NMR (CDCl

02.989).

Methyl 3-Methoxy-4-nitrothiophene-2-carboxylate (NO

and THF (29.5 ml) was cooled to 08. CH (341 mg, 27 ml, 8.12 mmol) in Et

plastic funnel. After several seconds, N evolution ceased, and the soln. was allowed to warm to r.t. A few drops

of glacial AcOH were added to ensure the complete consumption of CH . The solvent was evaporated: 40

0.72. IR (film): 3115, 2960, 1725, 1555, 1506, 1453,

): 164.6; 155.6; 132.6; 53.0. EI-MS: 202.989 ( M , C H NO S ; calc.

3 5 5 5

2

-Mt-OMe; 40). A mixture of 39 (1.5 g, 7.4 mmol)

O was slowly added by means of a

2

N

2

N

2

(1.49 g, 93%). Yellow solid. TLC (hexanes/AcOEt 1:1): R

f

1

435, 1387, 1353, 1281, 1199, 1110, 1059, 954, 772. H-NMR (CDCl

3

): 8.38 (s, 1 H); 4.10 (s, 1 H); 3.94 (s, 3 H).

1

3

C-NMR (CDCl

3

): 156.5; 131.3; 63.9; 53.0. EI-MS: 217.004 (M , C

7

H

7

NO

5

S ; calc. 217.004).

2

Methyl 4-Amino-3-methoxythiophene-2-carboxylate Hydrochloride (HCl ¥ H N-Mt-OMe; 4). A mixture of

4

0 (800 mg, 3.69 mmol) and SnCl ¥ 2 H O (6.66 g, 29.5 mmol) in 95% EtOH (29.5 ml) was stirred vigorously at

2

r.t. Conc. HCl soln. (29.5 ml) was added dropwise, and the soln. was heated at 358 for 6h. The mixture was

removed from heat and adjusted to pH 9 with 4n NaOH (15 ml). The resulting white emulsion was extracted

with AcOEt (3 Â100 ml), dried (MgSO

m HCl in Et O was added to precipitate crude 4. The salt was filtered and taken directly on to the next step.

Methyl 4-{[(tert-Butoxy)carbonyl]amino}-3-methoxythiophene-2-carboxylate (Boc-Mt-OMe; 41). A mix-

ture of 4 (1.0 g, 4.47 mmol), Et N (498 mg, 0.68 ml, 4.92 mmol), and (Boc) O (1.0 g, 4.92 mmol) in CH Cl

9 ml) was stirred at 608 for 12 h. The soln. was washed with sat. NH ), and

Cl soln. (3Â) dried (MgSO

evaporated, and the resulting red solid was subjected to CC (hexanes/AcOEt 5 :1): 41 (528 mg, 41.3%). White

4

), and evaporated. After addition of a small amount of fresh AcOEt,

2

3

2

(

4

solid. TLC (hexanes/AcOEt 10 :1): R

f

0.24. IR (film): 3437, 3329, 2979, 1772, 1716, 1530, 1440, 1376, 1230, 1165,

1

081, 1055, 993, 861, 778. H-NMR (CDCl

3

): 7.58 (s, 1 H); 6.82 (s, 1 H); 4.06( s, 3 H); 3.85 (s, 3 H); 1.52 (s, 9 H).

1

3

C-NMR (CDCl

3

): 161.5; 152.6; 151.5; 130.2; 112.0; 111.2; 85.5; 81.3; 62.9; 52.3; 28.6; 28.0. EI-MS: 287.083 (M ,

C

12

H

17NO

5

S ; calc. 287.083).

4

-{[(tert-Butoxy)carbonyl]amino}-3-methoxythiophene-2-carboxylic Acid (Boc-Mt-OH; 11). A mixture of

1 (250 mg, 0.87 mmol) and KOH (48.8 mg, 0.87 mmol) in dry MeOH (1 ml), was heated to 508 for 6h. The

soln. was added to CH Cl (5 ml) and H O (5 ml). The aq. layer was washed with CH Cl

(3 Â10 ml) and

acidified to pH 3 with 1n HCl (3.5 ml). The aq. soln. was then extracted with CH Cl

(3 Â10 ml), and the extract

was dried (MgSO ) and evaporated; 11 (193 mg, 81%). Off-white solid. TLC (hexanes/AcOEt 10 :1): R 0.24.

IR (film): 3400, 1699, 1526, 1438, 1369, 1230, 1154, 1053. H-NMR (CDCl ): 7.68 (s, 1 H); 6.84 (s, 1 H); 4.08

): 165.8; 152.7; 130.4; 112.0; 111.2; 81.4; 63.3; 62.3; 28.7. EI-MS: 273.067

4

2

4

f

1

3

1

3

(

s, 3 H); 1.53 (s, 9 H). C-NMR (CDCl

3

M , C11

H15NO S ; calc. 273.067).

5

2

-{[(tert-Butoxy)carbonyl]amino}-5-methylthiazole-4-carboxylic Acid (Boc-Nt-OH; 10). The 2-oxabuta-

noic acid (10 g, 98 mmol) was treated dropwise with Br (8 ml, 25 g, 157 mmol) while stirring. Upon completion

of addition, the mixture was stirred until the red color of Br had dissipated. Thiourea (14.8 g, 196mmol) was

2

then added in portions, and stirring was continued overnight. The mixture was acidified with conc. HCl soln.,

and the precipitated hydrochloride was filtered and washed with cold EtOH. The crude solid was taken into

DMF (50 ml), then DIEA (10 ml) and (Boc)

for 12 h, then diluted with AcOEt, and washed with brine (3Â). The combined org. phase was dried (Na

and evaporated. The resulting crude oil was dissolved in MeOH (0.1 l) and 1n NaOH (0.1 l) and stirred at r.t. for

h. The MeOH was then evaporated, and the aq. layer was washed with Et

O (2 Â0.1 l). The aq. phase was

acidified to pH 2 with 1n HCl and extracted with AcOEt (3 Â0.1 l). The combined org. phase was dried

Na SO ) and evaporated: 10 (11 g, 44%). White flaky solid. TLC (hexanes/AcOEt 5 :2, 10% AcOH): R 0.4.

2

O (21.4 g, 98 mmol) were added. The mixture was stirred at 608

2

SO )

4

1

2

(

2

4

f

Article Doi

Toward an understanding of the chemical etiology for DNA minor-groove recognition by polyamides

DOI: 10.1002/hlca.200290024

Source and publish data:

Authors:

Article abstract of DOI:10.1002/hlca.200290024

Full text of DOI:10.1002/hlca.200290024

Products guided by the article

R&D Labs maybe for 22600-30-2

Relevant to this article

Hot Product