3'-nitrophenylpropyloxycarbonyl (NPPOC) protecting groups for high-fidelity automated 5' --> 3' photochemical DNA synthesis.

Article abstract of DOI:10.1021/ol0069150

[structure: see text]. The most powerful DNA microarrays would be prepared by photolithography with free 3'-ends that could be processed enzymatically. A photoremovable group that could be removed in quantitative yield would ensure high purity of the synt

Full text of DOI:10.1021/ol0069150

ORGANIC
LETTERS
2001
Vol. 3, No. 8
1105-1108
3′-Nitrophenylpropyloxycarbonyl
(NPPOC) Protecting Groups for
High-Fidelity Automated 5′ f 3′
Photochemical DNA Synthesis
Michael C. Pirrung,* Laixin Wang, and Michael P. Montague-Smith
Department of Chemistry, LeVine Science Research Center, Duke UniVersity,
Box 90317, Durham, North Carolina 27708
pirrung@chem.duke.edu
Received November 22, 2000 (Revised Manuscript Received March 13, 2001)
ABSTRACT
The most powerful DNA microarrays would be prepared by photolithography with free 3′-ends that could be processed enzymatically. A
photoremovable group that could be removed in quantitative yield would ensure high purity of the synthesized probes. We have developed
new pyrimidine building blocks for 5′ f 3′ DNA synthesis with high cycle yields using the NPPOC (3′-nitrophenylpropyloxycarbonyl) protecting
group. These phosphoramidites were proved in automated photochemical DNA synthesis on a modified synthesizer.
DNA microarray technology has had a profound effect on
the way nucleic acid analysis is conducted1 and, conse-
quently, on the way biological research is performed in the
post-genomic era. Many different problems are being ad-
dressed through single-nucleotide polymorphism (SNP)
genotyping,2 RNA/gene expression analysis,3 mutation detec-
tion/diagnostics,4 genomic DNA sequencing and resequenc-
ing,5 and population genetics.6
DNA is formed one base at a time in situ. Many more
methods are now available to fabricate microarrays, involving
either in situ DNA synthesis by spatially directed delivery
methods such as ink-jet printing or spotting of prepared DNA
sequences (both short, synthetic oligonucleotides (<50 bases)
and longer, biologically derived DNAs, such as PCR
products).9 The size of the elements in physically directed
DNA arrays is inherently limited by the volume of reagent
that can be accurately delivered. Thus far, volumes of ∼100
pL have been dispensed, creating spot sizes of ∼100 µm
diameter after spreading on the surface. For comparison,
arrays prepared photolithographically currently have 20 µm
diameter probe sites, giving them 25-fold more probes per
unit area. Improvements in the lithographic technology may
also be forthcoming, as semiconductor manufacturing cur-
rently uses 0.3 µm photolithography for memory chip
production. Photolithographic DNA synthesis is becoming
more convenient with replacement of masks by digital
programmable spatial light director devices.10 It also has the
Initial excitement about the field was generated7 using
arrays prepared by light-directed chemical synthesis,8 where
(1) Gerhold, D.; Rushmore, T.; Caskey, C. T. Trends Biochem. Sci. 1999,
24, 168-73.
(2) Guo, Z.; Guilfoyle, R. A.; Thiel, A. J.; Wang, R.; Smith, L. M. Nucleic
Acids Res. 1994, 22, 5456-5465.
(3) Lockhart, D. J.; Winzeler, E. A. Nature 2000, 405, 827-36. Sohail,
M.; Akhtar, S.; Southern, E. M. RNA 1999, 5, 646-55.
(4) Whitcombe, D.; Newton, C. R.; Little, S. Curr. Opin. Biotechnol.
1998, 9, 602-608.
(5) Hacia J. G. Nat. Genet. 1999, 21 (Suppl), 42-7. Kurg, A.; Tonisson,
N.; Georgiou, I.; Shumaker, J.; Tollett, J.; Metspalu, A. Genet. Test. 2000,
4, 1-7.
(6) Cronin, M. T.; Fucini, R. V.; Kim, S. M.; Masino, R. S.; Wespi, R.
M.; Miyada, C. G. Hum. Mutat. 1996, 7, 244-55. Kozal, M. J.; Shah, N.;
Shen, N.; Yang, R.; Fucini, R.; Merigan, T. C.; Richman, D. D.; Morris,
D.; Hubbell, E.; Chee, M.; Gingeras, T. R. Nat. Med. 1996, 2, 753-759.
(7) Pease, A. C.; Solas, D.; Sullivan, E. J.; Cronin, M. T.; Holmes, C.
P.; Fodor, S. P. A. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 5022-5026.
(8) Fodor, S. P. A.; Read, J. L.; Pirrung, M. C.; Stryer, L.; Liu, A. T.;
Solas, D. Science 1991, 251, 767.
(9) Schena, M.; Shalon, D.; Davis, R. W.; Brown, P. O. Science 1995,
270, 467-470.
10.1021/ol0069150 CCC: $20.00 © 2001 American Chemical Society
Published on Web 03/29/2001

advantage of being a parallel, combinatorial process, where
the number of sequences prepared far exceeds the number
of steps in the synthesis. Spatial delivery, on the other hand,
is an inherently serial process. The most complex DNA
microarrays known (>105 probe sites) are commercially
prepared using photolithography.
operate on such mismatches with greatly reduced efficiency
and rate. When there is a perfect match between probe and
target site (X & Y ≡ complements), reaction with a
triphosphate occurs (J & K ≡ complements, in ligation R )
DNA chain (deoxyoligonucleotide 5′-triphosphate), in primer
extension R ) H or OH (nucleotide 5′-triphosphate)). The
covalent attachment of J to the probe permits stringent
washing steps that minimize background in imaging of the
probe site. While equilibrium binding may permit the
formation of end-mismatched duplexes, they are not attached
and do not contribute to background.
DNA microarrays are most commonly used for hybridiza-
tion to analyte nucleic acid (DNA or RNA). With a plethora
of oligonucleotide probes on an array and a high-complexity
target significantly longer than the probes, a number of
different hybrids (perfect match and single-base mismatch)
might be formed at each probe site. The difference in their
stability may be quite small, depending on the location of
the mismatch, with mismatches at the end of the probe being
hardest to discriminate against. Analysis of the hybridization
of target to the chip is primarily based on perfect match
hybrids, so signal from mismatches could confound the
analysis. Interestingly, chips prepared in situ by photo-
lithography with a high density of probe sequences within
each probe site have an unusual, nonclassical binding
behavior (as compared to hybridization events in solution)
that makes the discrimination against mismatches greater than
might be expected.11 This is indeed fortunate, as low cycle
yields (∼90%) in commercial DNA chip production lead to
impure probe sequences that could further confound the
analysis.12 Thus, better methods for fabrication and use of
DNA arrays are needed to give high discrimination against
false signals.
Primer extension methods for DNA arrays have been
reported by other groups13 as well as our own APEX effort.14
Ligation methods are also known.15 Both require a free 3′-
hydroxyl. Conventional DNA synthesis, as used in current
photolithographic array preparation, attaches probes to the
array at their 3′-end, preventing ligation or primer extension
from being performed. Ligation/primer extension methods
have therefore been limited to arrays made by spotting.
Methods to synthesize DNA starting from the 5′-end are
known.16 We earlier reported efforts to develop new methods
for reverse photochemical DNA synthesis.17
Approaches to increase the fidelity of photolithographic
DNA microarray preparation have focused on improved
photochemically removable protecting groups. Pfleiderer
developed the NPPOC (nitrophenylpropyloxycarbonyl) group
that is deprotected by a â-elimination reaction18 (Scheme
Approaches to increase the fidelity of microarray analysis
are based on enzymatic processing (Scheme 1). They exploit
(10) Singh-Gasson, S.; Green, R. D.; Yue, Y.; Nelson, C.; Blattner, F.;
Sussman, M. R.; Cerrina, F. Nat. Biotechnol. 1999, 17, 974-978. LeProust,
E.; Pellois, J. P.; Yu, P.; Zhang, H.; Gao, X.; Srivannavit, O.; Gulari, E.;
Zhou, X. J. Comb. Chem. 2000, 2, 349.
(11) Forman, J. E.; Walton, I. D.; Stern, D.; Rava, R. P.; Trulson, M. O.
ACS Symp. Ser. 1998, 682, 206-228.
Scheme 1. Enzymatic Processing on DNA Microarrays
(12) Pirrung, M. C.; Bradley, J.-C. J. Org. Chem. 1995, 60, 6270. Pirrung,
M. C.; Fallon, L.; McGall, G. J. Org. Chem. 1998, 63, 241.
(13) Broude, N. E., Sano, T.; Smith, C. L.; Cantor, C. R. Proc. Natl.
Acad. Sci. U.S.A. 1994, 91, 3072-3076. Livshits, M. A.; Forentive, V. L.;
Mirzabekov, A. D. J. Biomol. Struct. Dynam. 1994, 11, 783-795. Dubiley,
S.; Kirillov, E.; Mirzabekov, A. Nucleic Acids Res. 1999, 27, e19. Tonisson,
N.; Kurg, A.; Kaasik, K.; Lohmussaar, E.; Metspalu, A. Clin. Chem. Lab.
Med. 2000, 38, 165-70.
(14) Shumaker, J. M.; Metspalu, A.; Caskey, C. T. Laboratory Protocol
for Mutation Detection Landegren, U., Ed.; Oxford University Press:
Oxford, U.K., 1996; pp 93-95. Tollett, J. J.; Kurg, A.; Shah, A.; Roa, B.
B.; Richards, C. S.; Nye, S. H.; Pirrung, M.; Metspalu, A.; Shumaker, J.
M. Am. J. Hum. Gen. 1997, 61 (Suppl. S), 1322. Kurg, A.; Tollett, J. J.;
Shah, A.; Roa, B. B.; Richards, C. S.; Nye, S. H.; Pirrung, M.; Metspalu,
A.; Shumaker, J. M. Am. J. Hum. Gen. 1997, 61 (Suppl. S), 68. Pirrung,
M. C.; Connors, R. V.; Montague-Smith, M. P.; Odenbaugh, A. L.; Walcott,
N. G.; Tollett, J. J. J. Am. Chem. Soc. 2000, 122, 1873.
(15) Drmanac, R.; Drmanac, S.; Strezoska, Z.; Pauneska, T.; Labat, I.;
Zeremski, M.; Snoddy, J.; Funkhouser, W. K.; Koop, B.; Hood, L.;
Crkvenjakov, R. Science 1993, 260, 1649-1652. Gunderson, K. L.; Huang,
X. C.; Morris, M. S.; Lipshutz, R. J.; Lockhart, D. J.; Chee, M. S. Genome
Res. 1998, 8, 1142-1153. Drmanac, S.; Kita, D.; Labat, I.; Hauser, B.;
Schmidt, C.; Burczak, J. D.; Drmanac, R. Nat. Biotechnol. 1998, 16, 54-
8.
(16) Horne, D. A.; Dervan, P. B. J. Am. Chem. Soc. 1990, 112, 2435-
2437. Roble, J.; Pedroso, E.; Grandas, A. Nucleic Acids Res. 1995, 23,
4151. Shchepinov, M.; Case-Green, S.; Southern, E. Nucleic Acids Res.
1997, 25, 1155. Hudson, R.; Robidoux, S.; Damha, M. Tetrahedron Lett.
1998, 39, 1299.
(17) Pirrung, M. C.; Fallon, L.; Lever, D. C.; Shuey, S. W. J. Org. Chem.
1996, 61, 2129.
(18) Giegrich, H.; Eisele-Buhler, S.; Hermann, C.; Kvasyuk, E.; Charuba-
la, R.; Pfleiderer, W. Nucleosides Nucleotides 1998, 17, 1987. Hasan, A.;
Stengele, K.-P.; Gegrich, H.; Cornwell, P.; Isham, K.; Sachleben, R.;
Pfleiderer, W. Foote, R. Tetrahedron 1997, 53, 4247.
the fact that enzymes that replicate and join DNA have
evolved to maintain a very low error rate. While hybridiza-
tion can occur to give an imperfect match at the reaction
site (X & Y * complements), polymerase and ligase enzymes
1106
Org. Lett., Vol. 3, No. 8, 2001

deprotection during automated synthesis. Commercial con-
trolled-pore glass-5′-T-3′-DMTr support (Glen Research) was
loaded into a synthesis cartridge fabricated from polymer
filters and a clear fluoropolymer PFA tube and then depro-
tected by manual delivery of 3% TCA/acetonitrile. An ABI
392 automated DNA synthesizer was modified such that an
external events control relay usually used to trigger a fraction
collector (to assess dimethoxytrityl cation release) instead
opens a shutter in an optical train (Figure 1). This optical
train delivers ∼4 mW/cm2 of 365 nm light from an arc lamp
through a liquid light guide to the cartridge, of which ∼25%
is transmitted to the support. The synthesis program is
modified so the usual acid deprotection step is replaced with
the photochemical removal of an NPPOC group in the
presence of base. The base solution (50 mM piperidine in
CH3CN) replaces the acid solution on the synthesizer. The
synthesizer triggers the external event relay during the
deprotection step instead of afterward.
Earlier work on 5′ f 3′ synthesis with DMTr protection
showed that increased coupling time is required compared
to 5′-DMTr-amidites, presumably as a result of the lower
nucleophilicity of the 3′-hydroxyl group. We verified this
observation with commercial 3′-DMTr-5′-amidites, showed
that 120 s coupling provides >99% cycle yield, and used
this coupling time in our photochemical reverse synthesis.
In initial experiments to optimize photochemical 5′ f 3′
synthesis, T4 was prepared on a 0.2 µmol scale (enabling
product analysis by HPLC and MS) starting with cpg-5′-T-
3′-OH. Coupling with 1 and photochemical deprotection were
performed twice, and the terminal T was added by coupling
with 1 (i.e., 2.5 cycles). The T4 was cleaved and fully
deprotected with ammonia and directly analyzed by reverse-
phase HPLC (Supelcosil LC-18S (5 µm, 1.6 mm × 250 mm),
20% f 40% linear gradient of 25 mM Et3NHOAc, pH 6.5
in 2:3 CH3CN/H2O), which gives single-nucleotide resolution
up to decanucleotides. The deprotection step was examined
at increasing irradiation times, with >98% cycle yields
observed in 20 min at ∼1 mW/cm2, 365 nm. The piperidine
solution was refreshed 10× to remove the styrene byproduct
and agitate the support, a measure that would not be
necessary during DNA array fabrication. Substitution of DBU
for piperidine, as reported by Beier and Hoheisel, led to
reduced product yield and purity. This problem was traced
to dark deprotection. A T3 synthesis omitting irradiation in
the deprotection step using DBU as base still gives T3 (5%
yield), while with piperidine as base only T2 is observed.
Several oligodeoxynucleotides were prepared by 5′ f 3′
synthesis using phosphoramidites 1 and 2 and analyzed by
HPLC, with both the detection of truncation sequences and
absolute peak integration against authentic calibration stan-
dards providing cycle and overall yields (Table 1). MALDI-
TOF MS confirmed the fidelity of these products.
Scheme 2. Photochemical Removal of NPPOC Groups As
Reported by Pfleiderer18 Occurs by â-Elimination
2). Beier and Hoheisel reported nucleoside phosphoramidites
with NPPOC groups at the 5′ position that can be deprotected
photochemically in the presence of limited amounts of amine
base (piperidine or DBU).19 Arrays with free 5′-ends result
from DNA synthesis with these NPPOC amidites. They
undergo coupling and photochemical deprotection in es-
sentially quantitative cycle yields, far superior to amidites
with other photoremovable groups such as MeNPOC. Given
this high-fidelity method of DNA array synthesis and the
interest in high-fidelity array analysis methods requiring free
3′-ends, we have adapted the NPPOC method to reverse
DNA synthesis.
Two pyrimidine 5′-phosphoramidites 1 and 2 (Chart 1)
were prepared from the 5′-tert-butyldiphenylsilylated nucleo-
sides by treatment with NPPOC-Cl (generated from phosgene
and the alcohol, 0 °C f rt, 12 h, pyridine, 88-92%),
deprotection with triethylamine trihydrofluoride20 (10 equiv,
THF, rt, 12 h, 91%), and treatment with â-cyanoethyl N,N-
bis(diisopropylamino)chlorophosphine (i-Pr2EtN, 3 h, CH2-
Cl2, 0 °C f RT, 70-77%). The resulting amidites were
purified by precipitation from methylene chloride/hexane.
Chart 1. Pyrimidine Nucleoside NPPOC Phosphoramidites
In conclusion, these novel phosphoramidite building blocks
provide oligonucleotides in high cycle yields via reverse (5′
f 3′) light-based solid-phase synthesis. Under the photo-
chemical deprotection conditions, isobutyroyl deoxycytidine
Testing of these phosphoramidites as a prelude to array
preparation requires a method to perform photochemical
(20) Pirrung, M. C.; Shuey, S. W.; Lever, D. C.; Fallon, L. Bioorg. Med.
Chem. Lett. 1994, 4, 1345.
(19) Beier, M.; Hoheisel, J. D. Nucleic Acids Res. 2000, 28, e11.
Org. Lett., Vol. 3, No. 8, 2001
1107

Figure 1. Automated photochemical DNA synthesizer. (A) Oriel Hg/Xe 1000 W Arc lamp (source). (B) Condenser lens (2 in.). (C)
Surface-coated dichroic mirror (280-350 nm). (D) Water filter. (E) Electronically controlled shutter. (F) Narrow bandwidth (310 nm)
interference filter. (G) Fiber optic focusing lens. (H) UV/vis liquid light guide (1 m × 5 mm). (I) Fiber optic output collimating lens. (J)
DNA synthesis cartridge (target).
is unreactive. The purine building blocks remain to be added,
but this methodology promises to enable the most powerful
DNA microarray fabrication technology, photolithography,
to be combined with enzymatic processing of probes to afford
the highest fidelity in nucleic acid analysis.
Table 1. Oligodeoxynucleotides Synthesized Using 1 and 2
Acknowledgment. This work was supported by NIH
av cycle yielda
total yield
GM-46720 and an NIH postdoctoral fellowship (M.M.-S.).
sequence
5′ TTT
5′ TTTT
5′ TTTTTTTTTT
5′ TCC
98.8
98.7
98.2
98.7
97.6
96.2
83.0
97.5
Supporting Information Available: Experimental de-
scriptions. This material is available free of charge via the
Internet at http://pubs.acs.org.
a Photodeprotection was not performed after coupling of the 3′-base. The
T3 synthesis requires two couplings but only one deprotection, etc.
OL0069150
1108
Org. Lett., Vol. 3, No. 8, 2001