Single-molecule detection of 5-hydroxymethylcytosine in DNA through chemical modification and nanopore analysis

Article abstract of DOI:10.1002/anie.201300413

DNA threading the needle: A new method of single-molecule detection of 5-hydroxymethylcytosine (5hmC) in DNA has been developed. Selective thiol substitution of 5hmC (giving SMC) in a single-step, bisulfite-mediated reaction (see scheme) allows the incorporation of a peptide (yellow sphere) or biotin into DNA. Modified 5hmC bases can be readily distinguished at the single-molecule level using protein nanopore analysis. Copyright

Full text of DOI:10.1002/anie.201300413

.
Angewandte
Communications
DOI: 10.1002/anie.201300413
Epigenetic Markers
Single-Molecule Detection of 5-Hydroxymethylcytosine
in DNA through Chemical Modification and Nanopore
Analysis**
Wen-Wu Li, Lingzhi Gong, and Hagan Bayley*
Angewandte
Chemie
4350
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4350 –4355

Angewandte
Chemie
DNA methylation is an important
epigenetic modification of the
genome. For example, 5-methylcy-
tosine (5mC) plays a key role in the
regulation of gene expression asso-
ciated with development and
tumorigenesis.[1,2] Recently, 5-
hydroxymethylcytosine
(5hmC),[3,4]
5-formylcytosine
(5fC),[5–8] and 5-carboxylcyto-
sine[6,7] have also been found in
mammalian DNA. Changes in the
level and distribution of 5hmC
bases may be associated with the
maintenance and differentiation of
embryonic stem (ES) cells[8] and
transcriptional regulation in those
cells.[9] 5hmC has also been shown
to play a role in the regulation of
chromatin structure and gene
expression in central nervous
system cells.[10]
It is important, therefore, to
develop methods for the detection
of epigenetic bases in genomic
DNA.[11] A method based on bisul-
fite treatment, PCR amplification,
and nucleotide sequence analysis
Figure 1. Bisulfite reactions of cytosine derivatives. a) Bisulfite (>4.0m) at high temperature
(condition 1) converts C into U;[12] bisulfite (0.05–0.6m) in the presence of excess of thiol at lower
temperature (condition 2) does not affect C. b) Under either condition 1 or 2, 5mC is unreactive.
c) Under condition 1, 5hmC is converted into CMS; under condition 2, 5hmC is converted into CMS
as the minor product and SMC as the major product.
has become routine for the detection of 5mC in DNA.[12,13]
Bisulfite treatment converts cytosine (C) to uracil (U;
Figure 1a), whereas 5mC remains intact (Figure 1b).[14–16]
However, this method is unable to differentiate 5hmC[17,18]
from 5mC, because 5hmC is converted into cytosine 5-
methylenesulfonate (CMS) during the process,[19] which
behaves as C residues during PCR amplification (Figure 1c).
To overcome this problem, an oxidative bisulfite sequencing
method has been developed recently to locate 5hmC with
single base resolution.[20] Selective chemical oxidation of
5hmC to 5fC by potassium perruthenate (KRuO4) enables
bisulfite conversion of 5fC into uracil. Sequences of oxidized/
bisulfite-treated DNA are compared with untreated and
bisulfite-only sequences to distinguish C, 5mC, and 5hmC. In
a second combined enzymatic and chemical approach, 5hmC
is converted into C, and C and 5mC into U, before
sequencing.[21] The enrichment of 5hmC-containing DNA
for targeted sequencing has also been accomplished by
enzymatic modification of 5hmC with b-glucosyltransferase
followed by chemical modification with a biotin tag.[9,22–24]
DNA methyltransferase-directed thiol and selenol substitu-
tion of the hydroxy group of 5hmC in DNA has also enabled
subsequent modification with a biotin- or fluorescein-con-
taining group.[25] However, the use of expensive enzymes with
restricted recognition motifs, the need for the synthesis of
analogues of enzyme co-factors or substrates, and the multiple
steps required for nucleobase modification are drawbacks of
these approaches.
Single-molecule sequencing of modified bases in unam-
plified DNA is an attractive alternative to this method.
Single-molecule, real-time differentiation of 5hmC and 5mC
has been achieved based on their influence on the rate of
a polymerase engaged in sequencing-by-synthesis, but the
signals are context-dependent and the error rates are high.[26]
Nanopore methods are a promising alternative for single-
molecule DNA sequencing[27,28] and sequencing of the four
canonical bases (G, A, C, and T) has recently been accom-
plished.[29–31] In the experiments that demonstrated the
feasibility of base identification for strand sequencing,
biotinylated DNA strands were immobilized within an
engineered protein pore after binding to streptavidin, which
prevented translocation through the pore.[32–34] In this way,
5hmC and 5mC could be differentiated from G, A, C, or T.[35]
Herein, we report an alternative approach in which 5hmC is
chemically modified in situ, permitting the enrichment of rare
5hmC-containing sequences and their identification with
a protein nanopore at the single-molecule level.
[*] Dr. W. W. Li, L. Gong, Prof. H. Bayley
Department of Chemistry, University of Oxford
Mansfield Road, Oxford, OX1 3TA (UK)
E-mail: hagan.bayley@chem.ox.ac.uk
Dr. W. W. Li
Institute for Science and Technology in Medicine, Keele University
Thornburrow Drive, Stoke-on-Trent, ST4 7QB (UK)
[**] We thank Oxford Nanopore Technology for financial support.
Supporting information for this article (including synthesis and
characterization of sulfur-substituted 5hmC derivatives, the modi-
fication of 5hmC in ssDNA, UPLC-MS analysis, and single-channel
electrical recordings) is available on the WWW under http://dx.doi.
org/10.1002/anie.201300413.
Bisulfite converts 5hmC into CMS by transformation of
the 5-hydroxy group to a sulfonate (Figure 1c; Supporting
Angew. Chem. Int. Ed. 2013, 52, 4350 –4355
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4351

.
Angewandte
Communications
Information Figure S1a).[19] We speculated that CMS is
produced through formation of an exo-methylene intermedi-
ate (Figure S1b), similar to those implicated in the mecha-
nisms of the enzymes T4 deoxycytidylate hydroxymethy-
lase,[36] thiaminase,[37] and thymidylate synthase,[38] and sub-
sequent attack by the nucleophilic sulfite anion (Figure S1b).
If this mechanism is correct, a nucleophilic thiolate should
trap the exo-methylene intermediate to form the correspond-
ing 5-thiomethyl derivative (Figure 1c; Figure S1b). We
tested this idea by using two model compounds (5hmC and
5-hydroxymethyl-2’-deoxycytidine (5hmdC)). The reaction of
5hmC with excess of glutathione (GSH) or 1-thio-b-d-
glucopyranose in the presence of bisulfite in D2O at pD 5.0
yielded CMS (25%) and the sulfur-substituted 5hmC adducts
reaction products of C and 5hmC after bisulfite treatment
with a high concentration of bisulfite at a high temperature
(condition 1, Figure 1) in these four ssDNAs were character-
ized by ultra-performance liquid chromatography mass spec-
trometry (UPLC-MS). Measurement of the masses of the
unmodified ssDNAs, the stable intermediate products (after
bisulfite treatment), and the final modification products (after
alkali treatment; Figure S6) enabled us to determine the
number of 5hmC and C bases as CMS and U, respectively
(Figure S7, Table S3, and Supporting Information). CMS is
stable under both acidic and basic conditions, whereas the 5,6-
dihydrouracil-6-sulfonate intermediate is only stable under
acidic or neutral conditions (Figure 1c).[14,15] 5mC is unaltered
in the overall reaction (Figure 1b).
Next, the 5hmC of ssDNA-1 was modified at pH 5.0 by
bisulfite at a lower concentration and temperature in the
presence of excess GSH at 428C for 36–48 hours (condition 2)
followed by alkali treatment (Figure 1 and 2a). UPLC-MS
showed the formation of an adduct of GSH and ssDNA
(ssDNA-1-SG, yield approximately 70%) and a sulfonated
DNA adduct (ssDNA-1-CMS, yield approximately 30%;
based on the relative intensities of the MS peaks; Fig-
ure 3a,b). At 708C for four hours or 428C for 15 hours, the
reactions were incomplete. No conversion of C into U
occurred, and 5mC was also unchanged. In a control experi-
ment with ssDNA-4, in which the 5hmC and 5mC of ssDNA-
1 are replaced by 5mC and C, respectively, no modification of
1
(SMC, 75%; Figure S1a). H NMR spectroscopy (Figure S2)
showed that the reactions were complete after 18 hours at pD
5.0 and 428C. The 5hmC and 5hmdC adducts were isolated by
HPLC and characterized by UV/Vis absorption spectroscopy,
negative and positive ion mode electrospray ionization MS,
and NMR spectroscopy (Figures S3,S4; Tables S1,S2).
When the pD of the reaction was increased from 5.0 to 7.0,
the reaction rate decreased (Figure S2). This suggests that
protonation of the N-3 atom of 5hmC at pD 5.0, to initiate
formation of the exo-methylene intermediate, is more impor-
tant than deprotonation of the thiol (pKa = 9.42 for GSH[39]),
which would provide a higher concentration of the nucleo-
philic thiolate. In another possible pathway, bisulfite reacts
with the 5-hydroxy group of 5hmC to form a sulfite ester that
subsequently undergoes substitution by either a nucleophilic
thiolate or a second sulfite anion (Figure S1c). Under the
same conditions, G, A, T, and 5mC (as nucleosides) were
unaltered (data not shown). Remarkably, in the presence of
GSH and bisulfite, the nucleobase C in
any nucleobase (including
C into U conversion) was
observed, based on UPLC-MS analysis (data not shown).
The chemistry of ssDNA-1 and ssDNA-4 is therefore
consistent with that of the model mononucleotides (Figur-
es S1,S5).
cytidine underwent the initial attack by
bisulfite to form 5,6-dihydrocytidine-6-sul-
fonate,[14,15] but did not undergo hydrolytic
deamination to generate U at the rela-
tively low concentration of bisulfite and
low temperature that was used.[15,40] Thiol
derivatives (e.g. glutathione) do not react
with 2’-deoxycytidine and do not affect the
formation of 5,6-dihydrocytidine-6-sulfo-
nate (NMR data not shown). Bisulfite is
known to participate in the deamination
step.[40] During subsequent alkali treat-
ment (adjusted to pH 13 with aqueous
10n NaOH), the 5,6-dihydrocytosine-6-
sulfonate nucleobase was transformed
back to C rather than to U (Figure 1a;
Figure S5).
Next, the modification of 5hmC in
single-stranded DNA (ssDNA) by bisulfite
under conditions 1 and 2 (Figure 1) was
Figure 2. Bisulfite-mediated biotinylation of 5hmC in ssDNA. a) Modification of 5hmC with
glutathione (condition 2, Figure 1) and subsequent modification of the primary amino group
of the adduct. G, A, T, C and 5mC remain intact, while the 5-hydroxy group of 5hmC is
substituted by sulfite to form DNA-CMS (ꢀ30%) and DNA-SG (ꢀ70%). In a second step,
the primary amino group of glutathione is modified by biotin- or fluorescein-N-hydroxysulfo-
succinimidyl ester (shown as spheres). b) Direct bisulfite-mediated biotinylation with N-
biotinyl-l-cysteine (condition 2, Figure 1).
carried out. Four 28-mer ssDNAs contain-
ing 5mC and 5hmC were used: 5’-ACTG-
TATCAXCTGGTCCTGTATYTAATA-
3’; 1, X = 5hmC and Y= 5mC; 2, X =
5hmC and Y= C; 3, X = 5hmC and Y=
5hmC; 4, X = 5mC and Y= C. First, the
4352
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4350 –4355

Angewandte
Chemie
Figure 3. UPLC-MS characterization of ssDNA-1, and its modified products. a) Mass spectrum of unmodified ssDNA-1. b) Mass spectrum of
a single UPLC peak containing a mixture of ssDNA-1-CMS and ssDNA-1-SG, which were not separated. c) Mass spectrum of a UPLC peak of
a biotinamidyl hexanoate glutathione DNA adduct (ssDNA-1-SG-Biotin) from the two-step modification approach (Figure 2a). d) Mass spectrum
of a biotinamidyl cysteinyl DNA (ssDNA-1-Cys-Biotin) adduct from the single-step modification (Figure 2b) of ssDNA-1.
Two approaches were explored for the incorporation of
biotin into ssDNA at the 5hmC site. First, the N-terminal
primary amine of the GSH in ssDNA-1-SG was selectively
modified with biotin N-hydroxysulfosuccinimidyl esters[25,41]
with yields ranging from 50% to 95%, depending on the
length of the linker in the reagent (Figure 2a and 3c). The
overall yield of biotinylation of 5hmC-containing DNA over
two steps was 30–65%. Similarly, fluorescein can be incorpo-
rated using N-hydroxysuccinimidyl esters with yields of
around 60% (data not shown). Second, a synthetic biotinyl
cysteine derivative (Supporting Information) was used for
a single-step incorporation of biotin at 5hmC sites in ssDNA-
1 and a 100-mer ssDNA with yields ranging from 35–55%, as
demonstrated by UPLC-MS (Figure 3d; Figure S8). Impor-
tantly, the biotinylated ssDNAs can be enriched by using
immobilized streptavidin (Figure S9).[25]
The wild-type staphylococcal a-hemolysin (aHL)[42] pro-
tein nanopore (which is not optimized for nucleobase
detection)[33] was then used to detect modified 5hmC in
a
ssDNA, using the streptavidin immobilization
approach.[33,34] The 40-mer ssDNA (DNA40-hmC, Figure 4a;
Figure S10) contained a fragment of the POU5F gene (5’-
TATACACAGGCXGATGTGGG-3’) from human embry-
onic stem cells,[13] in which a 5mC at position 9 with respect to
the 3’-end (X) was replaced by 5hmC. Biotinylation at the 3’-
end placed the 5hmC in the immobilized DNA40-hmC at the
recognition site near the constriction in the lumen of the aHL
pore.[33] In the presence of DNA40-hmC and streptavidin, the
open pore current (IO) was decreased to a new level (IB;
Figure 4b). The histogram of IRES% values (IRES% = (IB/
IO)100) shows a main peak with IRES% = 11.8 Æ 0.1 (Fig-
ure 4b,c), which is set at zero for further comparison with
Angew. Chem. Int. Ed. 2013, 52, 4350 –4355
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4353

.
Angewandte
Communications
as 5fC, where the aldehyde group
could be specifically modified by
hydrazine and aminooxy deriva-
tives.[5,8]
Received: January 17, 2013
Published online: April 4, 2013
Keywords: 5-hydroxymethylcytosine ·
.
DNA methylation · epigenetics ·
gene sequencing · nanopores
[1] A. Bird, Nature 2007, 447, 396.
[2] E. N. Gal-Yam, Y. Saito, G. Egger,
P. A. Jones, Annu. Rev. Med. 2008,
59, 267.
[3] S. Kriaucionis, N. Heintz, Science
2009, 324, 929.
Figure 4. Detection of modified 5hmC in an immobilized ssDNA with the wild-type aHL pore.
a) Scheme of the immobilized 40-mer ssDNA (5’-TTTTTTTTTTTTTTTTTTTTTATACACAGGCX-
GATGTGGG-TEG-biotin-3’). X=5hmC (DNA40-hmC, top) or X=glutathione adduct of 5hmC
(DNA40-SG, bottom; Figure 2a and 3b). TEG-biotin is a biotin moiety attached to a 15-atom mixed
polarity triethylene glycol spacer.[33] b) Current levels for an aHL pore before and after blockade with
DNA40-hmC only (top) or DNA40-hmC, DNA40-CMS and DNA40-SG (bottom), after addition of the
modified products to the recording chamber (Supporting Information). The recordings were made at
+160 mV in 1m KCl, 20 mm Tris·HCl, containing 0.1 mm EDTA, pH 8.0. c) Histograms of residual
current levels for an aHL pore interrogated with DNA40-hmC (top), and an aHL pore interrogated
with DNA40-hmC, DNA40-SG and DNA40-CMS (bottom). The value of IRES% for DNA40-hmC was set
equal to 0.
[4] M. Tahiliani, K. P. Koh, Y. Shen,
W. A. Pastor, H. Bandukwala, Y.
Brudno, S. Agarwal, L. M. Iyer,
D. R. Liu, L. Aravind, A. Rao,
Science 2009, 324, 930.
[5] T. Pfaffeneder, B. Hackner, M.
Truss, M. Munzel, M. Muller,
C. A. Deiml, C. Hagemeier, T.
Carell, Angew. Chem. 2011, 123,
7146; Angew. Chem. Int. Ed. 2011,
50, 7008.
[6] Y. F. He, B. Z. Li, Z. Li, P. Liu, Y.
Wang, Q. Tang, J. Ding, Y. Jia, Z.
Chen, L. Li, Y. Sun, X. Li, Q. Dai,
C. X. Song, K. Zhang, C. He, G. L. Xu, Science 2011, 333, 1303.
[7] S. Ito, L. Shen, Q. Dai, S. C. Wu, L. B. Collins, J. A. Swenberg, C.
He, Y. Zhang, Science 2011, 333, 1300.
[8] E. A. Raiber, D. Beraldi, G. Ficz, H. E. Burgess, M. R. Branco, P.
Murat, D. Oxley, M. J. Booth, W. Reik, S. Balasubramanian,
Genome Biol. 2012, 13, R69.
[9] W. A. Pastor, U. J. Pape, Y. Huang, H. R. Henderson, R. Lister,
M. Ko, E. M. McLoughlin, Y. Brudno, S. Mahapatra, P.
Kapranov, M. Tahiliani, G. Q. Daley, X. S. Liu, J. R. Ecker,
P. M. Milos, S. Agarwal, A. Rao, Nature 2011, 473, 394.
[10] M. Mellꢀn, P. Ayata, S. Dewell, S. Kriaucionis, N. Heintz, Cell
2012, 151, 1417.
[11] M. Mꢁnzel, D. Globisch, T. Carell, Angew. Chem. 2011, 123,
6588; Angew. Chem. Int. Ed. 2011, 50, 6460.
[12] S. J. Clark, A. Statham, C. Stirzaker, P. L. Molloy, M. Frommer,
Nat. Protoc. 2006, 1, 2353.
[13] R. Lister, M. Pelizzola, R. H. Dowen, R. D. Hawkins, G. Hon, J.
Tonti-Filippini, J. R. Nery, L. Lee, Z. Ye, Q. M. Ngo, L. Edsall, J.
Antosiewicz-Bourget, R. Stewart, V. Ruotti, A. H. Millar, J. A.
Thomson, B. Ren, J. R. Ecker, Nature 2009, 462, 315.
[14] H. Hayatsu, Y. Wataya, K. Kazushige, J. Am. Chem. Soc. 1970,
92, 724.
[15] R. Shapiro, R. E. Servis, M. Welcher, J. Am. Chem. Soc. 1970, 92,
422.
[16] M. Frommer, L. E. McDonald, D. S. Millar, C. M. Collis, F. Watt,
G. W. Grigg, P. L. Molloy, C. L. Paul, Proc. Natl. Acad. Sci. USA
1992, 89, 1827.
[17] Y. Huang, W. A. Pastor, Y. Shen, M. Tahiliani, D. R. Liu, A. Rao,
PLoS One 2010, 5, e8888.
[18] S. G. Jin, S. Kadam, G. P. Pfeifer, Nucleic Acids Res. 2010, 38,
e125.
modified ssDNAs. When DNA40-hmC was modified by
reaction with glutathione and bisulfite (Figure S10) and
added to a recording chamber already containing DNA40-
hmC, the histogram showed two additional peaks correspond-
ing to a glutathione-ssDNA conjugate (DNA40-SG) and
a sulfonated ssDNA conjugate (DNA40-CMS; Figure 4c).
The peak with the larger shift (DIRES% = À2.7 Æ 0.1) relative
to DNA40-hmC was tentatively assigned to DNA40-SG,
because of the larger size of GSH compared to the sulfonate
group. In control experiments with the 40-mer ssDNAs where
position 9 was 5mC, G, A, C, or T, histograms of the electrical
recordings showed no change (data not shown) after treat-
ment with glutathione and bisulfite and subsequent addition
of alkali. Therefore, the combination of chemical modifica-
tion and nanopore analysis enables us to detect 5hmC within
a ssDNA using the wild-type aHL pore.
In summary, we have described a simple non-enzymatic
modification of 5hmC, bisulfite-mediated thiol substitution,
which allows coupling to peptides, fluorescein, and biotin.
This single-step biotinylation procedure should be a useful
way to enrich 5hmC-containing DNA fragments, and com-
pares favorably with alternative approaches that use
enzymes.[9,24,25] By nanopore analysis, the modified 5hmC
residues can easily be distinguished from other nucleobases,
which provides a basis for single-molecule sequencing to
localize epigenetic markers in unamplified DNA. The
approach might be extended to other epigenetic bases such
[19] H. Hayatsu, M. Shiragami, Biochemistry 1979, 18, 632.
4354
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4350 –4355

Angewandte
Chemie
[20] M. J. Booth, M. R. Branco, G. Ficz, D. Oxley, F. Krueger, W.
Reik, S. Balasubramanian, Science 2012, 336, 934.
[21] M. Yu, G. C. Hon, K. E. Szulwach, C. X. Song, L. Zhang, A.
Kim, X. Li, Q. Dai, Y. Shen, B. Park, J. H. Min, P. Jin, B. Ren, C.
He, Cell 2012, 149, 1368.
[30] E. A. Manrao, I. M. Derrington, A. H. Laszlo, K. W. Langford,
M. K. Hopper, N. Gillgren, M. Pavlenok, M. Niederweis, J. H.
Gundlach, Nat. Biotechnol. 2012, 30, 349.
[31] E. C. Hayden, 2013, DOI: 10.1038/nature.2012.10051.
[32] N. Ashkenasy, J. Sanchez-Quesada, H. Bayley, M. R. Ghadiri,
Angew. Chem. 2005, 117, 1425; Angew. Chem. Int. Ed. 2005, 44,
1401.
[22] A. Szwagierczak, S. Bultmann, C. S. Schmidt, F. Spada, H.
Leonhardt, Nucleic Acids Res. 2010, 38, e181.
[23] C. X. Song, T. A. Clark, X. Y. Lu, A. Kislyuk, Q. Dai, S. W.
Turner, C. He, J. Korlach, Nat. Methods 2011, 9, 75.
[24] C. X. Song, K. E. Szulwach, Y. Fu, Q. Dai, C. Yi, X. Li, Y. Li,
C. H. Chen, W. Zhang, X. Jian, J. Wang, L. Zhang, T. J. Looney,
B. Zhang, L. A. Godley, L. M. Hicks, B. T. Lahn, P. Jin, C. He,
Nat. Biotechnol. 2011, 29, 68.
[33] D. Stoddart, A. J. Heron, E. Mikhailova, G. Maglia, H. Bayley,
Proc. Natl. Acad. Sci. USA 2009, 106, 7702.
[34] E. A. Manrao, I. M. Derrington, M. Pavlenok, M. Niederweis,
J. H. Gundlach, PLoS One 2011, 6, e25723.
[35] E. V. Wallace, D. Stoddart, A. J. Heron, E. Mikhailova, G.
Maglia, T. J. Donohoe, H. Bayley, Chem. Commun. 2010, 46,
8195.
[36] M. M. Butler, K. L. Graves, L. W. Hardy, Biochemistry 1994, 33,
10521.
ˇ
˙
[25] Z. Liutkeviciu¯te, E. Kriukiene, I. Grigaityte, V. Masevicius, S.
Klimasauskas, Angew. Chem. 2011, 123, 2138; Angew. Chem. Int.
Ed. 2011, 50, 2090.
[26] B. A. Flusberg, D. R. Webster, J. H. Lee, K. J. Travers, E. C.
Olivares, T. A. Clark, J. Korlach, S. W. Turner, Nat. Methods
2010, 7, 461.
[27] H. Bayley, Curr. Opin. Chem. Biol. 2006, 10, 628.
[28] D. Branton, D. W. Deamer, A. Marziali, H. Bayley, S. A. Benner,
T. Butler, M. Di Ventra, S. Garaj, A. Hibbs, X. Huang, S. B.
Jovanovich, P. S. Krstic, S. Lindsay, X. S. Ling, C. H. Mastran-
gelo, A. Meller, J. S. Oliver, Y. V. Pershin, J. M. Ramsey, R.
Riehn, G. V. Soni, V. Tabard-Cossa, M. Wanunu, M. Wiggin,
J. A. Schloss, Nat. Biotechnol. 2008, 26, 1146.
[37] R. Nicewonger, C. A. Costello, T. P. Begley, J. Org. Chem. 1996,
61, 4172.
[38] C. W. Carreras, D. V. Santi, Annu. Rev. Biochem. 1995, 64, 721.
[39] S. G. Tajc, B. S. Tolbert, R. Basavappa, B. L. Miller, J. Am. Chem.
Soc. 2004, 126, 10508.
[40] M. Sono, Y. Wataya, H. Hayatsu, J. Am. Chem. Soc. 1973, 95,
4745.
[41] R. P. Viscidi, C. J. Connelly, R. H. Yolken, J. Clin. Microbiol.
1986, 23, 311.
[42] W. W. Li, T. D. Claridge, Q. Li, M. R. Wormald, B. G. Davis, H.
Bayley, J. Am. Chem. Soc. 2011, 133, 1987.
[29] G. M. Cherf, K. R. Lieberman, H. Rashid, C. E. Lam, K.
Karplus, M. Akeson, Nat. Biotechnol. 2012, 30, 344.
Angew. Chem. Int. Ed. 2013, 52, 4350 –4355
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4355