Chemistry & Biology

Previews

DNA Made of Purines Only

Christoph Arenz1 and Oliver Seitz1,*
1
Institut für Chemie, Humboldt Universität zu Berlin, Brook-Taylor-Str. 2, 12489 Berlin, Germany
*Correspondence: [email protected]
DOI 10.1016/j.chembiol.2007.05.001

The DNA double helix, containing both purine and pyrimidine bases, has evolved as the universal
genetic system. In this issue of Chemistry & Biology, Battersby and colleagues describe the
formation of double helix that is comprised solely of naturally occurring purine-nucleotides [1].

The double helical structure of DNA onal base pairs that may expand the matic ring [11]. In double helical
has fascinated researchers ever since repertoire of letters in the genetic al- xDNA each of the enlarged bases is
its discovery [2]. Chemists, biologists, phabet [3–5]. Analogs of nucleobases pairing with a natural nucleobase.
and physicists have used their reper- have been introduced into DNA to Thus, xDNA has eight bases instead
toire of methods to answer key ques- probe interactions between DNA and of the four bases in natural DNA. A
tions about the chemical and structural proteins [6, 7]. However, these studies central feature of xDNA is the enlarged
nature of the genetic system. How- did not intend to change the double distance between the two phosphate
ever, it remains difficult to answer helix architecture of the DNA mole- backbones of the double helix. Inter-
some fundamental questions. Why cule. A drastic redesign of duplex estingly, the extended DNA duplex
has nature chosen DNA as the univer- architecture has been shown by Es- was found to be more stable than the
sal genetic system and why have the chenmoser and colleagues [8]. They natural Watson-Crick structure, pre-
bases adenine (A), thymine (T), cyto- discovered a new base-pairing mode sumably due to the contribution of
sine (C), and guanine (G) evolved as present in the structure termed homo- p-stacking.
the letters of the genetic alphabet? DNA, and demonstrated the occur- Kool’s version of a size-expanded
Why this particular double helix archi- rence of Watson-Crick pairs between DNA relied on the incorporation of
tecture? One way to approach these guanine and isoguanine as well as artificial nucleobases. In this issue of
questions is to use our current knowl- 2,6-diaminopurine and xanthine. The Chemistry & Biology, Battersby and
edge in the design and synthesis of backbone of homoDNA is comprised colleagues provide ample evidence
possible alternatives and to study of a 60 /40 -linked hexopyranosyl for the existence of a size-expanded
whether DNA function can be emu- sugar, which replaces the 50 /30 - duplex structure that is made of com-
lated. Such studies put our under- linked 20 -deoxyribose unit in natural ponents that can be found in the pool
standing of DNA structure and func- DNA. The distance between the two of the naturally occurring ribonucleo-
tion to the test. A particular challenge sugar C atoms that connect the nucle- sides [1]. They introduce an all-purine
is the design of a completely new obases with the sugar phosphate DNA in which the natural pyrimidine
genetic system, in which the complete backbone is larger in purine-purine bases thymidine and cytosine have
set of ‘‘natural’’ base pairs, A-T, T-A, than in purine-pyrimidine base pairs. been replaced by hypoxanthine (H)
G-C, and C-G, has been replaced by It was reasoned that the nonhelical, and isoguanine (J) (Figure 1). Thermal
the alternatives. Battersby and col- ladder-type structure of homoDNA denaturation studies, that involved
leagues have succeeded in demon- facilitated the accommodation of the the use of fluorescent probes, sug-
strating the function of a new base- size-expanded purine-purine pairs. gested a specific base-pairing be-
pairing principle. According to their Watson-Crick pairing between purines tween adenine and hypoxanthine and
results, published in this issue, purine- was expected to be hindered in DNA, between guanine and isoguanine.
purine pairing enables a new mode of due to the constraints induced by the More sophisticated hybridization ex-
DNA duplex association. Interestingly, double helical structure. Similar opin- periments that employed 7-deazaade-
the components of the new DNA du- ion was also adopted by Diederichsen nine instead of adenine and hypoxan-
plex architectures all occur in natural and coworkers, who demonstrated thine instead of guanine revealed that
nucleotides, which evokes the ques- Watson-Crick interactions between the interaction between the noncanon-
tion about a possible intermediacy of purines in linear alanyl-peptide nucleic ical base pairs was likely to proceed
purine-purine pairing in the early acids [9]. However, a seminal paper in a Watson-Crick analog manner.
stages of molecular evolution. from Kool and colleagues revealed Prerequisite for this kind of interaction
The formation of base pairs is at the that size-expanded base pairs can be is the N3-H-tautomer of isoguanine,
heart of DNA function. Thus, a consid- incorporated also into DNA double he- which is less preferred in aqueous
erable amount of research has been lical structures, termed xDNA (Figure 1) solution but obviously stabilized upon
devoted to the exploration of alterna- [10]. They synthesized four new DNA double strand formation.
tive base pairs. Significant work has nucleotides, each of them differing Although it can be expected that the
been invested in the design of orthog- from the native ones by an extra aro- all-purine DNA duplex has a widened

Chemistry & Biology 14, May 2007 ª2007 Elsevier Ltd All rights reserved 467
 Chemistry & Biology

Previews

DNA. Thus, the all-purine DNA may

be of utility as an orthogonal base-
pairing system in the design of probes
for applications in biotechnology and
materials science. A hallmark of Bat-
tersby’s all-purine DNA is the fact
that all of its building blocks occur in
nature, which inevitably raises the
question whether the Watson-Crick
interaction between A and H and G
and J has played a role in an early pro-
genitor self-replicating system. Since
RNA is not only able to act as a genetic
storage molecule but can also act as
a biocatalyst, the hypothesis of the
‘‘RNA world’’ as the progenitor of life
as we know it has found wide accep-
tance [12]. However, since experimen-
tal attempts to demonstrate the ability
of pyrimidine nucleotides to oligomer-
ize onto a complementary template
have run into difficulties, a simpler
ancestral nucleic acid preceding
RNA has been suggested [13]. Nearly
twenty years ago, Wächtershäuser
speculated that such an ancestral
nucleic acid might be an all-purine
nucleic acid [14]. Battersby and col-
leagues now show that chimeric struc-
tures composed of purine-pyrimidine
and purine-purine base pairs are pos-
sible, which could reflect a hypothetic
evolutionary transition between an an-
cestral and the modern genetic sys-
tem. At least this article in Chemistry
& Biology provides some experimental
data that might lead us to re-engage in
this old debate. Why is our modern ge-
netic system still seemingly superior to
all other possible systems? This ques-
tion remains wide open, regardless of
the putative role proposed for an all-
purine ancestral genetic system.

Figure 1. Enlarged Watson-Crick Base Pairs

(A) In naturally occurring DNA, purines form Watson-Crick base pairs with pyrimidines. REFERENCES
(B) In Eschenmoser’s linear homoDNA [8] the purines 2,6-diaminopurine (D) and guanine (G) can
pair with the purines xanthine (X) and isoguanine (I) in a Watson-Crick fashion. 1. Battersby, T.R., Albalos, M., and Friesen-
(C) Size-expanded Watson-Crick base pairs in helical DNA duplexes have been shown by Kool [10]. hahn, M.J. (2007). Chem. Biol. 14, this
(D) In this issue, Battersby and colleagues introduce purine-purine pairing as an alternative mode of issue, 525–531.
DNA duplex association between components that all occur in nature.
2. Watson, J.D., and Crick, F.H. (1953).
Nature 171, 737–738.
diameter resembling Kool’s xDNA [10] between A and H when compared to
3. Piccirilli, J.A., Krauch, T., Moroney, S.E.,
and should therefore be similarly stabi- the canonical A–T interaction. and Benner, S.A. (1990). Nature 343, 33–37.
lized by an additional contribution of In contrast to the canonical du-
base stacking, all-purine DNA duplex plexes, the all-purine DNA duplexes, 4. Abou-Zied, O.K., Jimenez, R., and Romes-
berg, F.E. (2001). J. Am. Chem. Soc. 123,
was found to be substantially less sta- although they may carry the same con- 4613–4614.
ble than canonical DNA duplexes. This tent of information, are probably not
5. Hirao, I., Harada, Y., Kimoto, M., Mitsui, T.,
may be due to a reduced hydration recognized by DNA-directed enzymes Fujiwara, T., and Yokoyama, S. (2004).
and a missing electrostatic interaction or small molecules targeting natural J. Am. Chem. Soc. 126, 13298–13305.

468 Chemistry & Biology 14, May 2007 ª2007 Elsevier Ltd All rights reserved
Chemistry & Biology

Previews
6. Jiang, Y.L., Stivers, J.T., and Song, F. moser, A. (1998). Helv. Chim. Acta 81, 11. Gao, J., Liu, H., and Kool, E.T. (2005). An-
(2002). Biochemistry 41, 11248–11254. 375–474. gew. Chem. Int. Ed. Engl. 44, 3118–3122.

7. Beuck, C., Singh, I., Bhattacharya, A., 12. Orgel, L.E. (1968). J. Mol. Biol. 38, 381–
9. Hoffmann, M.F.H., Bruckner, A.M., Hupp,
Hecker, W., Parmar, V.S., Seitz, O., and 393.
T., Engels, B., and Diederichsen, U.
Weinhold, E. (2003). Angew. Chem. Int.
(2000). Helv. Chim. Acta 83, 2580–2593.
Ed. Engl. 42, 3958–3960. 13. Orgel, L.E. (1986). Orig. Life Evol. Biosph.
17, 27–34.
8. Groebke, K., Hunziker, J., Fraser, W., 10. Liu, H., Gao, J., Lynch, S.R., Saito, Y.D.,
Peng, L., Diederichsen, U., Zimmermann, Maynard, L., and Kool, E.T. (2003). Sci- 14. Wächtershäuser, G. (1988). Proc. Natl.
K., Holzner, A., Leumann, C., and Eschen- ence 302, 868–871. Acad. Sci. USA 85, 1134–1135.

Exploiting Green Treasures

Stephanie Grond1,* and Guido Meurer2,*
1
Georg-August Universität Göttingen, Institute of Organic and Biomolecular Chemistry, Tammannstr. 2, D-37077, Göttingen
2
B.R.A.I.N AG, Unit Head Microbial Production Technologies, Darmstädter Str. 34, D-64673 Zwingenberg, Germany
*Correspondence: [email protected] (S.G.), [email protected] (G.M.)
DOI 10.1016/j.chembiol.2007.05.002

In this issue of Chemistry & Biology, Ishida and colleagues [1] report on the characterization of new
aeruginoside metabolites and the respective NRPS gene cluster in the cyanobacterium Planktothrix
agardhii.

Cyanobacteria are among the most In this issue of Chemistry & Biology its ecological and physiological rele-
talented, culturable micro-organisms [1], two analogs of cyanobacterial vance; although, Fujii and coworkers
from which novel, structurally diverse, aeruginosides, produced by the genus [5] gave some initial insights.
biochemically active natural products Oscillatoria (syn. Planktothrix), are re- The approach taken by Ishida and
have been isolated. Many of the com- ported. More than 200 compounds colleagues [1] is a sophisticated
pounds are small nontoxic or toxic originate from different species of the application of the general knowledge
peptides with unusual amino acids, order Oscillatoriales [2]. In the last about the organization of nonriboso-
polyketides, or hybrids of different decade, the toxic Planktothrix agardhii mal peptide synthetase (NRPS) bio-
biosynthetic pathways [2, 3]. received considerable attention, and synthetic gene clusters in cyanobacte-
Cyanobacteria are structurally di- was recognized as a prolific source of ria to probe a yet unexplored part of
verse, geographically widespread in novel metabolites [5]. Several com- the metabolite pattern of P. agardhii
freshwater, marine, and terrestrial hab- pounds from different chemical clas- strain CYA 126/8. The method used el-
itats, and some genera are nitrogen ses are currently known. These include egantly turns around a classic concept
fixing and are therefore of great impor- the first aeruginosins (205A and 205B) in natural product research. Typically,
tance for the natural balance of the reported from P. agardhii that were a biosynthetic gene cluster is identified
ecology. Due to the production of highly highlighted as glycopeptides inhibiting long after the respective metabolite is
active hepatotoxins and neurotoxins, serine proteases [6], several variants of identified and fully characterized in
increasingly common cyanobacterial the hepatotoxin complex microcystin a chemical-pharmaceutical screening
water blooms are of serious concern isolated from a single strain [7], and program [8]. In the new report, a
in freshwater reservoirs or lakes [2, 3, the multicyclic microviridins that are NRPS gene sequence was identified
4]. Investigating these toxins and other the largest known cyanobacterial oli- by a degenerate PCR approach in
bioproducts with a focus on cancer gopeptides with characteristic ester Planktothrix and used for insertional
drug discovery and chemical ecology and secondary amino bonds [3]. How- mutagenesis with a chloramphenicol
is a promising research field where ever, the full spectrum of chemical resistence cassette to give evidence
culturable cyanobacteria play an out- compounds produced by different iso- for the involvement of the genes within
standing role. Diverse metabolites lates of the P. agardhii species has not the cluster in putative peptide biosyn-
with relevant pharmaceutical activities been cataloged. Perhaps more inter- thesis. Importantly, parallel cultivation
have been isolated from cyanobacteria. estingly, the whole metabolite pattern of wild-type cells and mutant cells
However, the true relevance of these of a single strain as a distinct cocktail called attention to previously unknown
highly active agents for the producer of bioactive agents has not been metabolites. Subsequent careful
strains remains elusive at present. investigated in detail with respect to chemical analysis identified two new

Chemistry & Biology 14, May 2007 ª2007 Elsevier Ltd All rights reserved 469