Mapping the landscape of potentially primordial informational oligomers: Oligodipeptides and oligodipeptoids tagged with triazines as recognition elements

Article abstract of DOI:10.1002/anie.200603207

(Chemical Equation Presented) Pairing up: Oligodipeptide, oligodeoxy-dipeptide, or oligodipeptoid backbones tagged with the 2,4-diaminotriazine nucleus pair strongly with complementary DNA and RNA. This is in sharp contrast with the behavior of the 2,4-dioxotriazine nucleus, which does not act as a nucleo-base in these systems.

Full text of DOI:10.1002/anie.200603207

Communications
DOI: 10.1002/anie.200603207
Base Pairing
Mapping the Landscape of Potentially Primordial Informational
Oligomers: Oligodipeptides and Oligodipeptoids Tagged with Triazines
as Recognition Elements**
Gopi Kumar Mittapalli, Kondreddi Ravinder Reddy, Hui Xiong, Omar Munoz, Bo Han,
Francesco De Riccardis, Ramanarayanan Krishnamurthy,* and Albert Eschenmoser*
One of the challenges that current thinking on the problem of
biogenesis poses to contemporary organic chemistry is to
map, by chemical synthesis, the landscape of the chemical
structures of potentially primordial informational[2] oligom-
ers. Knowledge of such a landscape in terms of chemical facts
will offer a guide on one of the directions along which the
search for the chemistry of lifeꢀs origin has to proceed. The
very existence of such a landscape is to be assumed in view of
the outcome of the extensive experimental studies toward a
chemical etiology of nucleic acid structure[3] as well as of the
comprehensive search in various laboratories for nucleic acid
substitutes that are functionally competent with respect to the
goals of the medicinal antisense project.[4]
It has long been conjectured that RNA was not the first
genetic system, a notion adumbrating a transient existence of
primordial genetic systems that have preceded RNA.[5,6] In
fact, the capability of informational Watson–Crick base
pairing is surprisingly widespread among oligomer systems
of quite different backbone structures, ranging from non-
canonical oligonucleotides and oligonucleosides[3,7] to oligo-
amides[8] and to oligopeptides.[9] Furthermore, Watson–Crick
pairing is found to mediate a variety of “base-pairing
languages” that are orthogonal to each other in the sense
that base sequences bound to a given type of oligomer
backbone may be devoid of the capability to communicate
with corresponding complementary base sequences attached
to another type of backbone and yet are capable of efficient
intrasystem Watson–Crick base pairing in their own “langua-
ge”.[3,9c–d,10] The primary chemical criterion for deciding
whether a given candidate system may belong to the land-
scape of primordial oligomers refers primarily to a systemꢀs
generational[11] properties in reference to boundary condi-
tions of the type considered in prebiotic chemistry.[12] For our
work, we are adopting a pragmatic criterion according to
which an oligomer system can qualify as potentially primor-
dial, if it is generationally (not necessarily structurally) at least
as simple as, yet preferably, simpler than, the structure types
of oligonucleotides, oligopeptides, and oligosacharides within
constraints that refer to the nature of starting materials,
reaction types, and reaction conditions deemed to be pre-
biotically realistic.[13] This is structurally far broader a
selection criterion than that used in our earlier studies,[3] in
as far as candidates may belong to any type of oligomer
structure as long they satisfy the premises and deserve to be
called informational.[2] Backbones may be achiral or chiral;
backbone chirality as such is not considered a reason for
excluding a system from being assigned to the landscape, the
underlying presumption being that replication of a chiral
informational oligomer would be intrinsically chiroselec-
tive[14] and, therefore, an oligomer system capable of replica-
tion would eventually become homochiral through the very
process of replicating.[15]
The majority of the experimental work on nucleic acid
alternatives carried out thus far has concentrated on back-
bone variation, making conservative use of the family of
canonical nucleobases as recognition elements. Studies in
various laboratories dealing with nucleobase variation[17] have
referred to alternative bases that were chosen for study
without reference to etiological constraints.[18] Yet, there are
heterocycle families that can be envisaged to have been
capable of acting as alternative nucleobases while, at the same
time, are derived from the same type of starting materials
from which the canonical nucleobases are supposed to have
been formed.[19] Of special interest is the possibility that such
heterocycles may offer ways of becoming attached to back-
bones by reactions that are different from the notorious
nucleosidation process required for nucleosides to be formed
from aldosugars and nucleobases, a reaction type that, after
half a century of prebiotic chemistry, has still not been
convincingly demonstrated to be capable of proceeding
efficiently under reaction conditions deemed to be potentially
primordial.[5,20]
[*] Dr. G. K. Mittapalli, Dr. K. R. Reddy, Dr. H. Xiong, Dr. O. Munoz,
Prof. Dr. F. De Riccardis, Prof. Dr. R. Krishnamurthy,
Prof. Dr. A. Eschenmoser
Department of Chemistry and
The Skaggs Institute for Chemical Biology
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
Fax: (+1)858-784-9573
E-mail: rkrishna@scripps.edu
Eschenmoser@org.chem.ethz.ch
Dr. B. Han, Prof. Dr. A. Eschenmoser
Laboratorium für Organische Chemie
Eidgenössische Technische Hochschule
HCI Hönggerberg, 8093 Zürich (Switzerland)
Fax: (+41)1-632-1043
[**] Chemistry of a-Aminonitriles, Part 45. For part 44, see referen-
ce [1a]; for part 43, see reference [1b]. This work was supported by
the Skaggs Research Foundation. G.K.M., K.R.R., H.X., O.M., and
F.D.R. are Skaggs Postdoctoral Fellows.
This and the following Communication[45] report results of
exploratory work on the pairing properties of heterocycles
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
2470
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2470 –2477

Angewandte
Chemie
that are derivatives of 2,4-diamino- and 2,4-dioxotriazines in
one of the studies, and derivatives of the family of 2,4-
disubstituted 5-aminopyrimidines in the other. Herein we
describe oligomer systems that contain backbones of the
structure types of oligodipeptides, oligodeoxodipeptides, and
oligodipeptoids tagged with triazines as recognition elements
that are derived from carboxyl functions of a-amino acids, for
example, aspartic, glutamic, a,b-diamino propionic, and
imino diacetic acids.[21,22] Scheme 1 depicts the chemical
structure of the four types of oligomer systems A–D
(exemplified as 2,4-diaminotriazine derivatives A(TNN)–D-
(TNN)) that this report deals with.
1,3,5-Triazines are a family of heterocycles that have been
widely investigated in chemical molecular-recognition stud-
ies.[23] With a few exceptions,[23e–g] these studies refer to the
behavior of triazines as recognition elements in an organic
medium and a few investigations dealt with triazines as
alternative nucleobases in nucleic acid chemistry.[24] 1,3,5-
Triazines are generationally simple heterocycles in the sense
that they can be derived from carboxylic acid functions with
activating agents and reaction partners under conditions that
are deemed to be compatible with the constraints of prebiotic
chemistry.[21] We became involved in the chemistry of this type
of recognition element in the context of our recent work on
allopurines.[25] The choice of oligoamide backbones for testing
triazines as recognition elements in oligomers was prompted
by the vast amount of information that has become available
in the wake of Nielsenꢀs discovery of the pairing properties of
peptide nucleic acids (PNAs)[4b,26] involving a wide variety of
variants[4c,f,8] as well as genuine peptide analogues of nucleic
acids.[9] We were further attracted to this project by
Scheme 1. Structures of the four triazine-tagged oligomeric systems
dealt with in this report. End groups are either free NH2 and CO2H or
CH3CONH and CONH2. Asp=aspartyl; Glu=glutamyl; IDA=iminodi-
acetic acid; EDA=ethylenediamine; (TN,N)=(2,4-diamino)triazin-6-yl.
an enticingly simple and rough[3d] way to extrap-
olate the (idealized) pairing conformations of A-
and B-type DNA to potential pairing conforma-
tions of oligomer systems that contain noncyclic
backbones tagged with noncanonical recognition
elements, such as triazines or any other potential
nucleobase alternative. Scheme 2 and its caption
explain how this was done.
The types of triazine-tagged oligomers A–C
depicted in Scheme 1 are built of monomer units
that are dipeptides, a requirement that corresponds
to a basic constitutional complementarity between
the structure types of oligonucleotides and oligo-
peptides[27] and conspicuously follows from the
conformational reasoning referred to above
Scheme 2. An approximate derivation of an (idealized) pairing conformation of
triazine-tagged aliphatic oligomer backbones from the (idealized) (ꢀg/ꢀg)-pairing
conformation of A-type DNA (and RNA). The two bonds of the latter conformation
with torsion angles of 1808 (b and e) correspond to bond positions of an aliphatic
backbone thread in which double bonds (or equivalents thereof, for example,
amide bonds) can be accommodated. Also compatible with this conformation are
electronegative centers in the solid-sphere positions of the (saturated) aliphatic
backbone where they satisfy the gauche effect. Alternation of such electronegative
centers with, for example, amide bonds, represents a third possibility. The figure
also depicts positioning and lengths of the side chain that bears the recognition
element such that the pairing axis of the triazine ring approximates that of a
pyrimidine ring in the pairing conformation of oligonucleotides. An alternative
pairing conformation of an aliphatic backbone can be derived analogously from the
(t/ꢀg)-pairing conformation of A-type DNA or the (g/t)-conformation of B-type
DNA (for details, see reference [3d]).
(Scheme 2). In all three systems, the repeating
unit contains (in addition to the recognition
element) a free carboxyl group, the function of
which is to ensure solubility of oligomers in
aqueous solution, whereas in oligomers of type D,
solubility is promoted by the presence of an
ammonium group in the backboneꢀs monomer
unit. The peptoid oligomers derived from imino-
diacetic acid (type C) are achiral;[28] the other
oligomers are derived from the corresponding
(enantiopure) l-a-amino acids.
Given that the goal of our studies was primarily
the screening of pairing capabilities, the synthesis
Angew. Chem. Int. Ed. 2007, 46, 2470 –2477
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2471

Communications
Scheme 3. Synthesis of triazine-tagged amino acids: a) 0.2m DCC
(1.1 equiv), 0.2m HOBt (1.1 equiv), dry DMF, 12 h, RT; b) 0.2m
biguanide (2.0 equiv),[29] 2 h, RT; c) 0.4m diphenyl imidodicarbonate
(2.0 equiv),[30] 0.4m DBU (2.0 equiv), dry CH2Cl2, 2 h, RT. Bn=benzyl,
Boc=tert-butoxycarbonyl, DBU=1,8-diazabicyclo[5.4.0]undec-7-ene,
DCC=N,N’-dicyclohexylcarbodiimide, DMF=N,N-dimethylforma-
mide, HOBt=N-hydroxybenzotriazole.
Figure 1. UV spectroscopic Tm curves of selected duplexes formed by a) 2,4-diamino-
triazine-tagged dodecamer sequences from oligomeric systems A–D with RNA
(r(T12)) and b) 2,4-diketotriazine-tagged hexadecamer sequences, AspGlu(TOO)16,
with complementary DNA (d), RNA (r), and TNA (t) sequences. For conditions of
measurement refer to the caption for Table 1. No self-pairing was observed for
individual partner stands, except in the case of AspGlu(TNN)16 (see Table 1,
entry 11). The asterisk ( ) indicates measurement in 0.15m NaCl.
*
Scheme 4. The monomeric (5), dipeptidic (6), dipeptoidic (7), and
monopeptidic (8) building blocks used in the solid-support synthesis
of oligomers of type A–D (see Scheme 1). CBz=benzyloxycarbonyl,
Fmoc=9-fluorenylmethoxycarbonyl.
of specific members of the oligomer systems followed
conventional methodologies of chemical peptide and
peptoid synthesis with no consideration for etiological
constraints. Scheme 3 exemplifies, in the case of gluta-
mic acid, the tagging procedures, and Scheme 4 depicts
the four different types of monomer units 5, 6, 7, and 8
that served as starting materials for the synthesis of the
oligomers of type A–D, which were prepared by
standard automated solid-phase peptide synthesis.
Notably, all these manipulations could be conducted
without the need for protecting the amino or oxo groups
of the triazine moiety.[31]
Figure 2. Temperature-dependent CD spectra of duplexes formed by the cross-
pairing of 2,4-diaminotriazine-tagged A–D dodecamer sequences (see
Scheme 1) with the complementary DNA and RNA sequences: a) AspGlu-
(TNN)12 + d(T12), Tm =508C; b) AspGlu(TNN)12 + r(T12), Tm =558C;
c) IDAbackbone(TNN)12 + d(T12), Tm =208C; d) EDAbackbone(TNN)12 + r(T12),
Tm =258C. Measurements were made with the respective A, B, C, or D (5 mm)
and 5 mm complementary DNA/RNA sequence in 1m NaCl , 10 mm NaH2PO4,
0.1 mm Na2EDTA, pH 7.0; d) in 0.15m NaCl. Temperature increments in 58C
steps; Tm(CD) at approximately 270 nm.
Most of our exploratory screening of the pairing proper-
ties was done with homobase sequences containing the 2,4-
2472
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2470 –2477

Angewandte
Chemie
Table 1: Tm data of triazine-tagged oligomeric sequences derived from backbones depicted in Scheme 1.
Entry
Pairing system[a]
Tm(UV)
Tm(CD)
Comments[c]
(l [nm])[b]
(l [nm])[b]
2
1
2
3
4
H NOCAspGlu(TNN)6 + poly-d(T)
+ poly-r(U)
41.7 (260)
28.6 (260)
50.0 (260)
35.8 (260)
32.2 (260)
28.2 (260)
42.8 (260)
31.1 (260)
0.15m NaCl
0.15m NaCl
0.15m NaCl
0.15m NaCl
+ d(T)12
+ r(T)12
H NOCAspGlu(TNN)12
<0 (250)
59.4 (260)
52.2 (260)
65.0 (260)
51.4 (260)
53.8 (260)
49.2 (260)
57.2 (260)
51.2 (260)
35.8 (260)
2
5
6
+ poly-d(T)
+ poly-r(U)
0.15m NaCl
0.15m NaCl
0.15m NaCl
0.15m NaCl
7
8
+ d(T)12
50 (270)
50 (270)
9
+ r(T)12
10
11
+ t(T)12
AcNHAspGlu(TNN)16
ꢁ21 (250)
34.3 (250)
50.8 (260)
54.0 (260)
50.0 (260)
22 (222)
self-pairing of (TNN)16-mer
100 mm
12
13
14
+ poly-d(T)
+ poly-r(U)
+ d(T)16
49 (260)
51 (260)
Ref. [40]
Ref. [40]
1:1 duplex by Job plot (UV); Ref. [40]
15
16
17
18
AcNHAspGlu(TOO)12 + d(D)12
+ poly-r(A)
19.4 (240)
<0 (240)
ꢁ14 (240)
ꢁ13 (240)
<10 (240)
<0 (240)
Ref. [44]
Ref. [44]
Ref. [44]
+ t(D)12
+
+
NH2 AspGlu(TNN)12
1:1 duplex by Job plot (UV); Ref. [44]
pH 5.0
Ref. [44]
H NOCAspAsp(TNN)12
2
19
20
21
AcNHAspGlu(TOO)16 + d(A)16
+ d(D)16
<0 (240)
<0 (240)
25.5 (240)
25.3 (240)
25.7 (240)
24.3 (240)
18.4 (240)
<10 (240)
<0 (240)
32.5 (240)
38.6 (240)
ꢁ11 (240)
Ref. [44]
pH 5.0, acetate buffer solution; Ref. [44]
2:1 triplex by Job plot (UV); Ref. [44]
pH 8.0; Ref. [44]
25 (250)
pH 6.0; Ref. [44]
pH 5.5, acetate buffer solution; Ref. [44]
pH 5, acetate buffer solution; Ref. [44]
Ref. [44]
Ref. [44]
Ref. [44]
22
23
24
+ poly-r(A)
+ r(A)16
+ t(D)16
50+50 mm
NH2
25
26
+
AspGlu(TNN)16
13 (252)
10+10 mm, 1:1 duplex by Job plot (UV); Refs. [40,44]
AcNHAspGlu[(TNN)(TOO)]6
<0 (240)
<0 (240)
Ref. [44]
100 mm
27
28
HOOCAspAsp(TNN)6 + poly-d(T)
+ poly-r(U)
<10 (260)
ꢁ10 (260)
29
30
31
32
HOOCAspAsp(TNN)12 + poly-d(T)
+ poly-r(U)
26.6 (260)
33.1 (260)
11.6 (260)
28.6 (260)
24.4 (260)
25 (270)
30 (270)
+ d(T)12
+ r(T)12
30 (270)
0.15m NaCl
NH2
33
34
IDAbackbone(TNN)8 + poly-d(T)
26.7 (270)
19.7 (270)
+ poly-r(U)
NH2
35
36
37
IDAbackbone(TNN)12 + poly-d(T)
35.2 (270)
29.1 (270)
22.7 (270)
35 (277)
25 (265)
20 (265)
1:1 duplex by Job plot (CD)
+ poly-r(U)
+ d(T)12
Angew. Chem. Int. Ed. 2007, 46, 2470 –2477
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2473

Communications
Table 1: (Continued)
Entry
Pairing system[a]
Tm(UV)
Tm(CD)
Comments[c]
1:2 ratio
(l [nm])[b]
(l [nm])[b]
38
39
+ r(T)12
27.7 (270)
<0 (240)
+
NH2IDAbackbone(TOO)6
NH2
NH2
40
41
IDAbackbone(TOO)6 + poly-d(A)
<0 (240)
<0 (240)
Ref. [44]
Ref. [44]
+ poly-r(A)
42
IDAbackbone(TOO)7 + t(D)12
<0 (240)
Ref. [44]
NH2
43
44
45
46
EDAbackbone(TNN)12 + poly-d(T)
+ poly-r(U)
47.4 (270)
31.7 (270)
33.1 (270)
30.9 (270)
45 (275)
30 (275)
30 (275)
30 (275)
0.15m NaCl
0.15m NaCl
0.15m NaCl
0.15m NaCl
+ d(T)12
+ r(T)12
[a] Oligodipeptide sequences are written from the COOH terminus (except when stated otherwise) and oligodipeptoid and oligoamide sequences are
written from the NH2 terminus, with every second amino acid residue tagged with the heterocycle; Asp=aspartyl; Glu=glutamyl; IDA=iminodiacetic
acid; EDA=ethylenediamine; (TNN)=(2,4-diamino)triazin-6-yl; (TOO)=(2,4-dioxo)triazin-6-yl; (TNO)=(2-amino-4-oxo)triazin-6-yl; A=adenine;
D=2,6-diaminopurine; T=thymine; U=uracil; d=DNA; r=RNA, t=TNA. [b] Measurements were made at the indicated wavelength (nm), the
pairing systems (each ꢁ5 mm) in phosphate buffer solution (1m NaCl, 10 mm NaH2PO4, 0.1 mm Na2EDTA; pH 7.0) except when stated otherwise. Tm
values [8C] are derived from maxima of the first derivative of the heating curve (Kaleidagraph software). [c] In acetate buffer solution (1m NaCl, 10 mm
NaOAc/HOAc, 0.1 mm EDTA). The end NH2 group of oligopeptide sequences 11, 15, 20, and 26 are acetylated.
diaminotriazine nucleus. Not unexpected (see Scheme 2), yet
nevertheless remarkable was the finding that such homobase
sequences in all four oligomer systems show efficient cross-
pairing with complementary sequences of RNA and DNA in
aqueous solution. This is shown by the Tm curves reproduced
in Figure 1a, the CD curves of Figure 2, and the Tm data
summarized in Table 1. Furthermore, the Tm data illustrate
that, in apparent agreement with the conformational reason-
ing about the directionality of base-pairing axes, the glutamic
acid tagged homo-oligomers AspGlu(TNN)n=6,12 cross-pair
with the complementary counterparts of RNA and DNA
more strongly than the corresponding aspartic acid tagged
homo-oligomers AspAsp(TNN)n=6,12. Additionally, a corre-
sponding strand with no methylene group in the backbone-
triazine linker, EDAbackbone(TNN)12, again shows more effi-
cient cross-pairing.[32] Job plots[33] (see the Supporting Infor-
mation) demonstrate the absence of triplex formation for the
cross-pairing of AspGlu(TNN)16 with d(T16) and of
IDAbackbone(TNN)12 with r(T12). The affinities of AspGlu(TNN)
sequences for RNA and DNA are comparable, with RNA
prevailing slightly. The cross-pairing capability extends to
(3’!2’)-a-l-threofuranosyl oligonucleotides (TNA)[3b] as a
partner, yet as the Tm value of the interaction between
AspGlu(TNN)12 and t(T12) (Table 1, entry 10) indicates, is
somewhat weaker than that found in natural systems.
than a steric incompatibility of triazine–purine base pairs
within hetero-backbone duplexes (as compared with triazine–
pyrimidine base pairs). We came to this conclusion as we
observed unambiguous (yet still comparably weak) base
pairing of both AspGlu(TOO)–dodeca- and hexadecamers
with the corresponding homobase sequences of DNA and
TNA containing a 2,6-diaminopurine instead of adenine as
the complementary base (Figure 1b).
In a careful spectroscopic study of the 5-aza analogue of
uracil, Jonas et al.[35] could not find any evidence for this
compound to exist in aqueous solution in a tautomeric form
other than the dilactam form, and for the solid state of
another 2,4-dioxo triazine derivative, the lactam form is
confirmed by an X-ray study.[36] The comparison of the pKa
value of 2,4-dioxotriazines with that of 2,4-dioxopyrimidines
is of significance: the first dissociation constant of the 6-(Boc-
aminomethyl)-2,4-dioxo-1,3,5-triazine[37] in aqueous solution,
as determined by the pH-dependent UV spectrum,[37] was
found to be 6.0, that is, approximately 4 pK units lower than
the pK value of uracil.[38] Remarkably, the observed Tm value
of the duplex AspGlu(TOO)16–d(D16) appears essentially
unchanged at pH values in the range of pH 5.5 to 8
(Table 1, entry 21), whereas its UV spectrum indicates that
the dioxotriazine rings in the duplex are deprotonated, at
least in part, at pH 8 and protonated at pH 5.[37,39]
Homobase sequences containing the 2,4-dioxotriazine
base complementary to the 2,4-diaminotriazine base were
synthesized with the backbones of the oligomer systems A, B,
and C. To our surprise and (transient) disappointment,[34] the
2,4-dioxotriazine nucleus in these oligomers is found to act as
a drastically weaker pairing partner to adenine in DNA and
RNA than the 2,4-diaminotriazine base acts toward thymine
in the natural backbones (Figure 1b). No cross-pairing was
discernable from Tm or CD curves of a (1:1) mixtures of the
homobase sequence AspGlu(TOO)16 and either r(A16) or
d(A16) (Table 1). This behavior is assumed to be caused by the
electronic properties of the 2,4-dioxotriazine nucleus rather
The pronounced imbalance in the pairing behavior of the
2,4-diamino and the 2,4-dioxo member of the 1,3,5-triazine
family (see Figure 1 and Table 1) augurs badly for the
relevance of triazine heterocycles as a family of recognition
elements of potentially primordial informational oligomer
systems. Indeed, intrasystem base pairing between the
homobase dodecamer and hexadecamer sequences AspGlu-
(TNN)n and AspGlu(TOO)n (n = 12 and 16) under standard
conditions were found to be very weak (UV, CD) and
comparable in strength with self-pairing of the hexadecamer
AspGlu(TNN)16 (Table 1 and Figure 3).[40] The (formally self-
complementary) dodecamer containing the two bases in
2474 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2470 –2477

Angewandte
Chemie
dioxotriazine rings.[37] In comparison, the 2,4-diaminotriazine
ring is very stable.
The systematic screening of the pairing properties of
oligomer systems, which according to the criterion of genera-
tional simplicity could be members of a potentially primordial
landscape, will perhaps more often than not lead to a negation
of such a membership on the basis of functional criteria. Our
observations show this to be the case, presumably irrespective
of backbone constitution, for oligomers containing 2,4-dioxo
derivatives of 1,3,5-triazines as recognition elements. On the
other hand, they also show that 2,4-diaminotriazine in
dipeptidic oligomers undergoes strong base pairing in aque-
ous solution with 2,4-dioxopyrimidines bound to the DNA or
RNA backbones. We consider this remarkable difference in
base-pairing fitness of the 2,4-dioxo derivatives of 1,3,5-
triazine compared with 2,4-dioxo derivatives of 1,3-pyrimi-
dine to be relevant to etiological reasoning on nucleic acid
structure. The observations described in the following com-
munication[45] corroborate this notion by providing further
insight into the constitutional and physicochemical prereq-
uisites that heterocycles must fulfill to be capable of efficient
base-pairing in aqueous solution.
Received: August 7, 2006
Published online: November 17, 2006
Keywords: base pairing · DNA · oligonucleotides · RNA ·
.
triazines
Figure 3. Intrasystem pairing experiments. a) Tm(UV) curves of the
duplex formed from cross-pairing of AspGlu(TNN)16 with AspGlu-
(TOO)16; also shown is the behavior of individual partner strands;
b) CD spectrum of the corresponding duplex; c) Job plot[33] showing
the 1:1 ratio of the pairing partners in the duplex formed (at 08C). CD
spectrum of AspGlu(TNN)16 alone shows self-pairing, Tm =228C at
222 nm (see the Supporting Information), whereas AspGlu(TOO)16
alone shows no temperature dependence in the CD spectrum.
Measurements were made with AspGlu(TNN)16 and AspGlu(TOO)16
(cꢁ10 mm, 1:1) in 1m NaCl, 10 mm NaH2PO4, 0.1 mm Na2EDTA,
pH 7.0. CD temperature increments in 58C steps.
[1] a) T. Wagner, B. Han, G. Koch, R. Krishnamurthy, A. Eschen-
moser, Helv. Chim. Acta 2005, 88, 1960; b) B. Han, V. Rajwanshi,
J. Nandy, R. Krishnamurthy, A. Eschenmoser, Synlett 2005, 744.
[2] In the present context, an oligomer system is considered to be
“informational” if it can accommodate constitutional diversity in
the positioning of recognition elements such that any specific
constitutional pattern of them can be “read” intermolecularly
and (in principle) reproduced by self-replication.
[3] a) M. Beier, F. Reck, T. Wagner, R. Krishnamurthy, A.
Eschenmoser, Science 1999, 283, 699; b) K.-U. Schöning, P.
Scholz, S. Guntha, X. Wu, R. Krishnamurthy, A. Eschenmoser,
Science 2000, 290, 1347; c) A. Eschenmoser, Science 1999, 284,
2118; d) A. Eschenmoser, Chimia 2005, 59, 836.
[4] a) P. Herdewijn, Liebigs Ann. 1996, 1337; b) P. E. Nielsen, Acc.
Chem. Res. 1999, 32, 624; c) K. N. Ganesh, P. Nielsen, Curr. Org.
Chem. 2000, 4, 931; d) M. Meldgaard, J. Wengel, J. Chem. Soc.
Perkin Trans. 1 2000, 3539; e) C. Leumann, Bioorg. Med. Chem.
2002, 10, 841; f) V. A. Kumar, K. N. Ganesh, Acc. Chem. Res.
2005, 38, 404.
[5] a) A. G. Cairns-Smith in Genetic Takeover and the Mineral
Origin of Life, Cambridge University Press, Cambridge, 1982;
b) G. F. Joyce, A. W. Schwartz, S. L. Miller, L. E. Orgel, Proc.
Natl. Acad. Sci. USA 1987, 84, 4398; c) A. Eschenmoser, Origins
Life Evol. Biosphere 1997, 27, 535.
[6] See also, however, the opposite viewpoint: a) J. P. Ferris, Origins
Life Evol. Biosphere 1993, 23, 307; b) W. Huang, J. P. Ferris,
Chem. Commun. 2003, 1458.
alternating positions AspGlu[(TNN)(TOO)]6 fails to show a
Tm value through either UV or CD measurements (Table 1,
entry 26). Exploratory tests with homo-2,4-dioxotriazine
sequences in the oligomer systems AspAsp and IDAbackbone
had been equally discouraging.[41]
The unfitness of the pair of 2,4-diamino- and 2,4-
dioxotriazine nuclei for the efficient base-pairing in aqueous
solution is exacerbated by the observation that oligomer
strands containing 2,4-dioxotriazines are found to be appre-
ciably more sensitive toward hydrolysis than corresponding
strands that contain 2,4-dioxopyrimidines.[42] Although UV
spectroscopic Tm determinations involving AspGlu(TOO)16
strands within the temperature range of 0–608C (18minꢀ1)
did not reveal any instability of the substrate, subjecting the
dioxo 16mer alone to two heating cycles reaching the
temperature of 908C under otherwise identical conditions
led to a partial constitutional change involving the UV-active
recognition elements which, on the basis of UV , HPLC, and
MS analysis, is to be interpreted as hydrolytic opening of
[7] a) R. O. Dempcy, O. Almarsson, T. C. Bruice, Proc. Natl. Acad.
Sci. USA 1994, 91, 7864; b) L. Kerremans, G. Schepers, J.
Rozenski, R. Busson, A. Van Aerschot, P. Herdewijn, Org. Lett.
2001, 3, 4129; c) M. Tarköy, M. Bolli, C. Leumann, Helv. Chim.
Acta 1994, 77, 716; d) S. K. Singh, P. Nielsen, A. A. Koshkin, J.
Angew. Chem. Int. Ed. 2007, 46, 2470 –2477
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2475

Communications
Wengel, Chem. Commun. 1998, 455; e) L. Zhang, A. Peritz, E.
Meggers, J. Am. Chem. Soc. 2005, 127, 4174.
[20] a) R. A. Sanchez, L. E. Orgel, J. Mol. Biol. 1970, 47, 531;
b) W. D. Fuller, R. A. Sanchez, L. E. Orgel, J. Mol. Evol. 1972, 1,
249; c) G. Zubay, T. Mui, Origins Life Evol. Biosphere 2001, 31,
87.
[8] a) N. K. Sharma, K. N. Ganesh, Chem. Commun. 2003, 2484;
b) T. H. S. Tan, D. T. Hickman, J. Morral, I. G. Beadham, J.
Mickelfield, Chem. Commun. 2004, 516; c) P. von Matt, A.
De Messmaker, U. Pieles, W. Zürcher, K.-H. Altmann, Tetrahe-
dron Lett. 1999, 40, 2899; d) T. Vilaivan, G. Lowe, J. Am. Chem.
Soc. 2002, 124, 9326; e) R. A. Goodnow, S. Tam, D. L. Pruess, W.
McComas, Tetrahedron Lett. 1997, 38, 3199.
[9] a) M. Kuwahara, M. Arimitsu, M. Shigeuasu, N. Saeki, M. Sisido,
J. Am. Chem. Soc. 2001, 123, 4653; b) A. Lenzi, G. Reginato, M.
Taddei, E. Trifilieff, Tetrahedron Lett. 1995, 36, 1717; c) U.
Diederichsen, Angew. Chem. Int. Ed. Engl. 1996, 35, 445;
d) A. M. Brückner, H. W. Schmitt, U. Diederichsen, Helv. Chim.
Acta 2002, 85, 3855; e) M. Fuji, K. Yoshida, J. Hidaka, T. Ohtsu,
Chem. Commun. 1998, 717; f) Y. Huang, S. Dey, X. Zhang, F.
Sönnichsen, P. Garner, J. Am. Chem. Soc. 2004, 126, 4626; g) H.
Umemiya, K. Komatsu, T. Yamzaki, H. Kagechika, K. Shudo, Y.
Hashimoto, Nucleic Acids Symp. Ser. 1995, 34, 37.
[21] a) C. G. Overberger, F. W. Michelotti, P. M. Carabateas, J. Am.
Chem. Soc. 1957, 79, 941; b) H. Bredereck, F. Effenberger, A.
Hoffmann, M. Hajek, Angew. Chem. 1963, 75, 825; Angew.
Chem. Int. Ed. Engl. 1963, 2, 655; c) E. Kuwano, E. Taniguchi, K.
Maekawa, Agric. Biol. Chem. 1973, 37,423; d) J. K. Simons,
M. R. Saxton, Org. Synth. Collect. Vol. IV, Wiley, New York,
1963, pp. 78; e) E. M. Smolin, L. Rapoport, The Chemistry of
Heterocyclic Compounds, Vol. 13 (Ed.: A. Weissberg), Wiley,
New York, 1959; f) J. T. Thurston, US 2,460,397, Jan. 18, 1949.
[22] The reaction of biguanide with a-amino acid esters to afford 6-
alkylamino-substituted 2,4-diamino-1,3,5-triazines has been
mentioned in a patent by Thurston.[21f] Kuwano et al.[21c] have
described the preparation
a variety of 2,4-diaminotriazine
derivatives by the reaction of morpholinobiguanide and dime-
thylbiguanide with a-amino acid esters.
[23] a) T. K. Park, J. Schroeder, J. Rebek Jr., J. Am. Chem. Soc. 1991,
113, 5125; b) T. K. Park, J. Rebek Jr., J. Am. Chem. Soc. 1992,
114, 4529; c) G. M. Whitesides, E. R. Simanek, J. P. Mathias,
C. T. Seto, D. N. Chin, M. M. Mammen, D. M. Gordon, Acc.
Chem. Res. 1995, 28, 37; d) K. Kurihara, K. Ohto, Y. Honda, T.
Kunitake, J. Am. Chem. Soc. 1991, 113, 681; e) K. Ariga, T.
Kunitake, Acc. Chem. Res. 1998, 31, 371; f) H. Asanuma, T. Ban,
S. Gotoh, T. Hishiya, M. Komiyama, Macromolecules 1998, 31,
371; g) T. Kawasaki, M. Tokuhiro, N, Kimizuka, T. Kunitake, J.
Am. Chem. Soc. 2001, 123, 6792; h) O. Slinchenko, A. Rachkov,
H. Miyachi, M. Ogiso, N. Minoura, Biosens. Bioelectron. 2004,
20, 1091.
[10] H. Wippo, F. Reck, R. Kudick, M. Ramaseshan, M. Bolli, G.
Ceulemans, R. Krishnamurthy, A. Eschenmoser, Bioorg. Med.
Chem. 2001, 9, 2411.
[11] Generational refers to generating a given molecule.
[12] a) S. L. Miller, L. E. Orgel, The Origins of Life on the Earth,
Prentice-Hall, Englewood Cliffs, NJ, 1974; b) L. E. Orgel, Crit.
Rev. Biochem. Mol. Biol. 2004, 39, 99.
[13] What is to be meant as prebiotically realistic is and necessarily
remains a matter of continuing debate, one that, however, can
progressively be subjected to experimental scrutiny, for example,
see: K. Nelson, M. Levy, S. L. Miller, Proc. Natl. Acad. Sci. 2000,
97, 3868.
[24] a) P. Pithova, A. Piskala, J. Pitha, F. Sorm, Collect. Czech. Chem.
Commun. 1965, 30, 90; b) M. Winkley, R. K. Robins, J. Org.
Chem. 1970, 35, 491; c) H. Dugas, B. J. Blackburn, R. K. Robins,
R. Deslauriers, I. C. P. Smith, J. Am. Chem. Soc. 1971, 93, 3468;
d) T. A. Riley, W. J. Hennen, N. K. Dalley, B. E. Wilson, R. K.
Robins, J. Heterocycl. Chem. 1986, 23, 1709; e) D. Gasparutto, S.
Da Cruz, A.-G. Boudat, M. Jaquinod, J. Cadet, Chem. Res.
Toxicol. 1999, 12, 630; f) R. G. Garcia, A. S. Brank, J. K.
Christman, V. E. Marquez, R. Erija, Antisense Nucleic Acid
Drug Dev. 2001, 11, 369; g) J. A. Beisler, J. Med. Chem. 1978, 21,
204; h) M. Hysell, J. S. Siegel, Y. Tor, Org. Biomol. Chem. 2005,
3, 2946.
[25] B. Han, B. Jaun, R. Krishnamurthy, A. Eschenmoser, Org. Lett.
2004, 6, 3691.
[26] P. E. Nielsen, M. Egholm, R. H. Berg, O. Buchardt, Science 1991,
254, 1497.
[27] H. Moser, A. Fliri, A. Steiger, G. Costello, J. Schreiber, A.
Eschenmoser, Helv. Chim. Acta 1986, 69, 1224.
[28] To include in our study peptoid oligomers as achiral variants of
the di-aspartate derived oligomers of type B has been proposed
by F.D.R. (A.E.).
[14] a) M. Bolli, R. Micura, A. Eschenmoser, Chem. Biol. 1997, 4,
309; b) A. Saghatelian, Y. Yokobayashi, K. Soltani, M. R.
Ghadiri, Nature 2001, 409, 797; c) I. A. Kozlov, L. E. Orgel,
P. E. Nielsen, Angew. Chem. Int. Ed. 2000, 112, 4462; Angew.
Chem. Int. Ed. 2000, 39, 4292.
[15] See also, however, the opposite viewpoint in reference [16].
[16] a) W. A Bonner, Origins Life Evol. Biosphere 1992, 21, 407; b) S.
Pizarello, A. Weber, Science 2004, 303, 1151.
[17] a) N. J. Leonard, R. A. Laursen, J. Am. Chem. Soc. 1963, 85,
2026; b) N. J. Leonard, A. G. Morrice, M. A. Sprecker, J. Org.
Chem. 1975, 40, 356; c) A. G. Morrice, M. A. Sprecker, N. J.
Leonard, J. Org. Chem. 1975, 40, 363; d) J. A. Piccirilli, T.
Krauch, S. E. Moroney, S. E. Benner, Nature 1990, 343, 33;
e) S. A. Benner, Acc. Chem. Res. 2004, 37, 784; f) F. Seela, H.
Debelak, J. Org. Chem. 2001, 66, 3303; g) J. He, F. Seela, Helv.
Chim. Acta 2002, 85, 1340; h) F. Seela, R. Kröschel, Org. Biomol.
Chem. 2003, 1, 3900; i) B. A. Schweitzer, E. T. Kool, J. Am.
Chem. Soc. 1995, 117, 1863; j) E. T. Kool, Acc. Chem. Res. 2002,
35, 936; k) D. L. McMinn, A. K. Ogawa, Y. Wu, J. Liu, P. G.
Schultz, F. E. Romesberg, J. Am. Chem. Soc. 1999, 121, 11585;
l) A. H. Henry, F. E. Romesberg, Curr. Opin. Chem. Biol. 2003,
7, 727; m) I. Luyten, P. Herdewijn, Eur. J. Med. Chem. 1998, 33,
515.
[18] Systematic contributions to nucleobase variation have been
made by Leonard,[17a–c] Benner,[17d,e] Seela,[17f–h] Kool[17i,j] and
Romesberg[17k,l]. The criteria for choosing nucleobase variants in
these studies were different from those used in the present
project.
[19] a) J. Oro, A. Kimball, Biochem. Biophys. Res. Commun. 1960, 2,
407; b) J. P. Ferris, L. E. Orgel, J. Am. Chem. Soc. 1966, 88, 1074;
c) J. P. Ferris, R. A. Sanchez, L. E. Orgel, J. Mol. Biol. 1968, 33,
693; d) A. B. Voet, A. W. Schwartz, Bioorg. Chem. 1983, 12, 8;
e) M. P. Robertson, S. L. Miller, Nature 1995, 375, 772; f) S.
Miyakawa, H. J. Cleaves, S. L. Miller, Origins Life Evol.
Biosphere 2002, 32, 209.
[29] S. N. Holter, W. C. Fernelius, Inorg. Synth. 1963, 7, 58.
[30] a) Y. Watanabe, H. Usui, S. Kobayashi, H. Yoshiwara, T.
Shibano, T. Tanaka, Y. Morishima, M. Yasuoka, M. Kanao, J.
Med. Chem. 1992, 35, 189; b) H. Usui, Y. Watanabe, M. Kanao, J.
Heterocycl. Chem. 1993, 30, 551.
[31] All compounds were characterized by 1H and 13C NMR spec-
troscopy and mass spectrometry after purification by column
chromatography on silica gel. Oligodipeptides and oligodipep-
toids were synthesized on an Expedite 8909 Nucleic Acid
Synthesizer (Perseptive Biosystems) using a modified PNA-
protocol, purified by HPLC (ion exchange) to a minimal purity
of 95%% and checked by MALDI-TOF-MS (see Supporting
Information).
[32] It should be noted, however, that oligomer D(TN,N)12 differs
from the others carrying a positive charge in its backbone. We
2476 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2470 –2477

Angewandte
Chemie
have not pursued further studies with this system because
oligomers from this series were found to be constitutionally
relatively unstable. (lability of the backbone toward strand-
scission by b-elimination of the ammonium residue toward the
a-position of the triazine ring).[37]
[41] A homododecamer sequence in the oligomer system A tagged
with 2-amino-4-oxotriazine, AspGlu(TNO)12 showed weak self-
pairing with a Tm = 16.18C (c ꢁ 100 mm, 1m NaCl) and weak
cross-pairing with poly-r(C) (Tm = 14.48C, c ꢁ 5 + 5 mm, 1m
NaCl).
[33] C. R. Cantor, P. R. Schimmel in Biophysical Chemistry, Free-
man, San Francisco, CA, 1980, pp. 1135 – 1139; Part III (The
Behavior of Biological Macromolecules).
[34] The reason why the disappointment was “transient” will become
clear in the following paper. That a single replacement of
cytosine by 5-aza-cytosine in a DNA duplex lowers the Tm value
has been reported in reference [24f].
[35] J. Jonas, M. Horak, A. Piskala, J. Gut, Collect. Czech. Chem.
Commun. 1962, 27, 2754.
[36] P. R. Lowe, C. H. Schwalbe, Acta Crystallogr. Sect. C 1998, 54,
1936.
[42] The lower stability of the 5-aza-cytosine heterocycle in a
nucleoside (unsubstituted at position C6) against hydrolysis as
compared with cytosine has been reported in reference [24g]. 5-
Aza-uracil (unsubstituted at carbon C6) is known to be unstable
against hydrolysis.[24a]
[43] R. M. C. Dawson, D. C. Elliot, W. H. Elliot, K. M. Jones, Data
for Biochemical Research, Oxford, Clarendon Press, 1959.
[44] The temperature-dependent UV spectra of dodecamer AspGlu-
(TOO)12 displays an isosbestic point at 240 nm (both at pH 5.0
and 7.0).[37] This wavelength was chosen for all Tm(UV) melting
point measurements involving oligomers containing AspGlu-
(TOO) to avoid interference from variations in absorbance. The
variations are owing to a change in the equilibrium between the
protonated and deprotonated forms of AspGlu(TOO) because of
temperature increments. Measuring at 240 nm ensures observa-
tion of the change in absorption owing to the temperature
dependence of a process (taken to be base pairing) different
from nucleobase deprotonation.
[37] See the Supporting Information.
[38] K. Nakanishi, N. Suzuki, F. Yamazaki, Bull. Chem. Soc. Jpn.
1961, 34, 53.
[39] The apparent independence of the Tm value from the pH value
between pH 8–6 (Table 1, entry 21) may result from an overlap
of at least two factors: deprotonation of the dioxotriazine and
the protonation of the 2,6-diaminopurine (pKa DH+ ꢁ 4.4[37]),
the latter taking over below pH 6. The Tm(UV) measurement
with d(A16) (pKa of AH+ ꢁ 3.8[43]) at pH 5.0 does not show a Tm
value above 08C.
[45] G. K. Mittapalli, Y. M. Osornio, M. A. Guerrero, K. R. Reddy,
R. Krishnamurthy, A. Eschenmoser, Angew. Chem. 2007, 119,
2530; Angew. Chem. Int. Ed. 2007, 46, 2478.
[40] Self-pairing of AspGlu(TNN)16 (see the Supporting Information)
may lower the Tm(UV) value.
Angew. Chem. Int. Ed. 2007, 46, 2470 –2477
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2477