news and views

removing the abundant (micromolar) Skeptics could argue that the systemic to fully appreciate the physiopathological
cysteine seems an impervious task for the control mechanisms would rapidly coun- implications of these modifications.
relatively few Tregs. The following mecha- teract these intercellular redox-based dia- 1. Shevach, E.M. Immunity 30, 636–645 (2009).
nisms are more likely: preventing cystine logs. However, in the constrained synapse, 2. Yan, Z., Garg, S.K., Kipnis, J. & Banerjee, R.
Nat. Chem. Biol. 5, 721–723 (2009).
uptake or GSH release and hydrolysis by cysteine, other nonprotein thiols and redox
3. Angelini, G. et al. Proc. Natl. Acad. Sci. USA 99,
DCs, or causing cysteine oxidation (Fig. 1). enzymes could reach high concentrations. 1491–1496 (2002).
A mechanistic comprehension of the phe- Also, in other tissues formed by cells tightly 4. Castellani, P., Angelini, G., Delfino, L., Matucci, A.
& Rubartelli, A. Eur. J. Immunol. 38, 2419–2425
nomena described remains a crucial task, as bound to each other, redox waves could be (2008).
it could identify new targets to modulate DC transmitted and could serve regulatory roles. 5. Kazama, H. et al. Immunity 29, 21–32 (2008).
and/or Treg activity. Nonetheless, the skeptic point of view is rea- 6. Lim, S.Y., Raftery, M.J., Goyette, J., Hsu, K. & Geczy,
C.L. J. Leukoc. Biol. 86, 577–587 (2009).
Another open question is whether and sonable: perhaps it is not surprising that such 7. Garin, M.I. et al. Blood 109, 2058–2065 (2007).
how the changes in the extracellular redox circuits have been described in inflammation 8. Blakytny, R. & Erkell, L.J. Int. J. Biochem. Cell Biol.
38, 1363–1373 (2006).
state are restricted in the areas of DC–T cell and immunity, where signals ought to be clear
9. Matsue, H. et al. J. Immunol. 171, 3010–3018
interaction. If this were not the case, the and perceptible, but brief and short-range. (2003).
nutrient-rich and reducing halo surrounding The identification of the redox-sensitive 10. Banjac, A. et al. Oncogene 27, 1618–1628
(2008).
DCs could activate bystander cells, promot- targets in Teffs and DCs, and in other systems 11. Rubartelli, A. & Sitia, R. Antioxid. Redox Signal.
ing unwanted responses. An exciting possi- where the release of cysteine/GSH serves a published online, doi:10.1089/ars.2009.2377 (26
bility is the polarized secretion of thiol-active regulatory role, is clearly a priority. Likewise, March 2009).
12. Tassi, S. et al. J. Immunol. 183, 1456–1462
enzymes, restricted to the immune synapsis, the development of dyes reporting on inter- (2009).
as described for IL-1β (ref. 13). cellular redox changes is needed in order 13. Gardella, S. et al. Blood 98, 2152–2159 (2001).

DNA-catalyzed hydrolysis of DNA phosphodiesters

Mostafa I Fekry & Kent S Gates
Hydrolysis of the phosphodiester linkages in DNA is a notoriously difficult reaction. Deoxyribozymes that use Zn2+ and
Mn2+ ions to accelerate this reaction by a factor of one hundred million (or more) have been identified and characterized.

Hydrolysis of the DNA backbone is an more hydrolytically labile. Indeed, an early cleaves an amide bond in a peptide. This is a
important reaction in biology and in the report that ascribed thermal cleavage of DNA substantial challenge due to the fact that the
laboratory manipulation of genetic mate- to phosphodiester hydrolysis may have inad- inherent hydrolytic lability of amides is about
rial. While many enzymes catalyze the vertently measured the rate of strand cleavage 20-fold lower than that of RNA phosphodi-
sequence-selective hydrolysis of DNA, it has stemming from hydrolysis of the glycosidic esters8,9. In the experiment, Watson-Crick
been difficult to design nonprotein catalysts bonds that hold the nucleobases to the sugar- base pairing was used to “clamp” a peptide
for this purpose. In this issue1, Chandra et phosphate backbone4. More recently, scien- substrate segment across from a 40-nucleo-
al. describe in vitro selection experiments tists have used cleverly designed non-nucleic tide stretch of randomized DNA sequence
that identified DNA catalysts that effectively acid model compounds to estimate the (Fig. 1b). In this manner, a vast number of
carry out sequence-specific hydrolysis of hydrolytic lability of DNA phosphodiester potential DNA catalysts were rapidly screened
DNA phosphodiesters. linkages5. The best current estimate for the for amidase activity. Polyacrylamide gel elec-
The uncatalyzed hydrolysis of DNA phos- half-life of hydrolytic cleavage of the DNA trophoresis was used to isolate the ‘winning’
phodiesters is thermodynamically favorable backbone rests at about 30,000,000 years5. sequences that caused strand cleavage in the
(∆G°′ = –5.3 kcal mol–1)(ref. 2) but kineti- At this rate, only about two phosphodiesters region opposing the 40-nucleotide catalytic
cally very slow (Fig. 1a). The negative charge within the 3 billion base pairs of DNA in a domain. These sequences were amplified and
on the phosphodiester group protects it from human cell are expected to undergo spon- re-ligated to a substrate strand, and the pro-
attack by hydroxide and other nucleophiles3. taneous hydrolysis per week. By way of cess was repeated. After 10 rounds of this selec-
In fact, the DNA hydrolysis reaction is so slow comparison, it is estimated that 10,000 nucle- tion process designed to enrich for sequences
that it has been difficult to measure the rate otide glycosidic bonds undergo spontaneous capable of cleaving the substrate strand, the
constant accurately. Attempts to measure hydrolysis per cell per day6. pool of winning sequences was able to cause
uncatalyzed phosphodiester hydrolysis in Deoxyribozymes—short segments of DNA 16% cleavage of the substrate strand over
native DNA have been doomed by the fact capable of catalyzing chemical reactions— the course of a 2-h incubation (37 °C, pH
that other bonds in the biopolymer are much have been developed for a number of reac- 7.4) in a buffer containing MgCl2 (40 mM),
tions, including the hydrolytic cleavage of RNA MnCl2 (20 mM), ZnCl2 (1 mM) and NaCl
Mostafa I. Fekry and Kent S. Gates are in the phosphodiesters, hydrolysis of the glycosidic (150 mM). From this collection, nine unique
Departments of Chemistry and Biochemistry, bonds holding guanine residues to the DNA deoxyribozymes were identified and four were
University of Missouri, Columbia, Missouri, backbone7 and oxidative cleavage of DNA1. In characterized in detail.
USA. their current work1, Chandra et al. initially set Surprisingly, all of the deoxyribozymes iden-
e-mail: [email protected] out to develop a deoxyribozyme catalyst that tified in this study cleave DNA phosphodiester

710 volume 5 number 10 october 2009 nature chemical biology

 news and views

linkages in the substrate strand, rather than at

the targeted peptide linkages. This is especially a B B
unexpected given that the inherent hydrolytic O O
lability of the phosphodiester bonds is decid- OH O
O–
edly lower than that of amide bonds in pep-
tides5,8. The hydrolysis reactions carried out O –
O P O
by these deoxyribozymes depend on the pres-
O
ence of both Zn2+ and Mn2+ ions. The team
was able to remove the peptide segment from H2O
Zn2+
the substrate strand to create an all-DNA Mn2+
B
substrate that undergoes phosphodiester
hydrolysis with an observed rate constant of B O
2.7 h–1. This rate constant is somewhat less O– O
than that observed for strand cleavage by a O
restriction enzyme such as EcoRI (k2 ~ 2.5 O P O
s–1)10, yet it represents an impressive rate O
O
acceleration of at least 108 over the uncata-
lyzed reaction. By shortening the Watson- Peptide or DNA
Crick binding arms in the all-DNA system, b linker region
the researchers were able to facilitate product
5′ 3′
release, thus enabling true catalytic behavior Substrate strand
including substrate turnover. Presumably, Catalytic strand
thermal cycling might also conveniently be
used to facilitate substrate turnover in a sys-
tem such as this. Importantly, careful experi-

Katie Vicari
ments were done to confirm that the strand
cleavage carried out by these deoxyribozymes 40 nt
proceeds via hydrolytic rather than oxidative
mechanisms. Figure 1 Deoxyribozyme-catalyzed cleavage of DNA. (a) The chemical reaction carried out by the
Metal ions are common cofactors in the deoxyribozymes. (b) Schematic design of the deoxyribozyme-substrate complexes.
enzymatic hydrolysis of phosphates and
phosphodiesters11, with the potential to (i) Agents capable of sequence-selective DNA 4. Eigner, J., Boedtker, H. & Michaels, G. Biochim.
deliver metal-bound hydroxides and (ii) hydrolysis are important tools in molecular Biophys. Acta 51, 165–168 (1961).
5. Schroeder, G.K., Lad, C., Wyman, P., Williams, N.H.
serve as Lewis acids. It may be of interest to biology and biotechnology12. In particular, & Wolfenden, R. Proc. Natl. Acad. Sci. USA 103,
determine the molecular details regarding the single-stranded nuclease deoxyribozymes 4052–4055 (2006).
how these DNA catalysts put Zn2+ and Mn2+ described here may provide an excellent start- 6. Lindahl, T. & Nyberg, B. Biochemistry 11, 3610–
3618 (1972).
ions to work in DNA hydrolysis. It will also be ing point for the development of “restriction 7. Sheppard, T.L., Ordoukhanian, P. & Joyce, G.F. Proc.
valuable to learn whether various DNA func- deoxyribozymes” that sequence-selectively Natl. Acad. Sci. USA 97, 7802–7807 (2000).
tional groups directly play roles in catalysis. cleave duplex DNA. 8. Kahne, D. & Still, W.C. J. Am. Chem. Soc. 110,
7529–7534 (1988).
Regardless of the exact mechanism, this work 1. Chandra, M., Sachdeva, A. & Silverman, S.K. Nat. 9. Williams, N.H., Takasaki, B., Wall, M. & Chin, J. Acc.
expands the repertoire of deoxyribozymes to Chem. Biol. 5, 718–720 (2009). Chem. Res. 32, 485–493 (1999).
one of the most challenging chemical reac- 2. Dickson, K.S., Burns, C.M. & Richardson, J.P. J. Biol. 10. Wright, D.J., Jack, W.E. & Modrich, P. J. Biol. Chem.
Chem. 275, 15828–15831 (2000). 274, 31896–31902 (1999).
tions in biology. This is fundamentally inter- 3. Westheimer, F.H. Science 235, 1173–1178 11. Wilcox, D.E. Chem. Rev. 96, 2435–2458 (1996).
esting and also of potential practical value. (1987). 12. Geurts, A.M. et al. Science 325, 433 (2009).

nature chemical biology volume 5 number 10 october 2009 711