Fragmentation pathway studies of oligonucleotides in matrix-assisted laser desorption/ionization mass spectrometry by charge tagging and H/D exchange
Chau-Wen Chou
0
1
3
4
Patrick A. Limbach
0
1
2
3
Richard B. Cole
0
1
3
0
Published online October 24, 2002 Address reprint requests to Dr. C.-W. Chou,
Department of Chemistry, University of New Orleans
, 2000 Lakeshore Drive, New Orleans,
LA 70148, USA
1
Department of Chemistry, University of New Orleans
, New Orleans,
Louisiana, USA
2
Present address: Department of Chemistry, University of Cincinnati
, Cincinnati,
OH 45221
. This article is dedicated to Professor Peter Williams of Arizona State University in celebration of his 60th birthday
3
Department of Chemistry, Louisiana State University
, Baton Rouge,
Louisiana, USA
4
Also at the Department of Chemistry, University of New Orleans
, New Orleans,
LA 70148
The desorption and decompositions of synthesized oligonucleotides bearing fixed charge sites have been investigated by linear, delayed-extraction, reflecting and post-source decay mode matrix-assisted laser desorption/ionization (MALDI) mass spectrometry. In contrast to the conventional [M H] forms of unmodified molecules where a proton is likely attached to a nucleobase, here the charge is fixed at one of the termini. In this case the observed fragment ions always incorporate the charge-tag. H/D exchange experiments provide no evidence for intramolecular migration of protons on the phosphate backbone to initiate the fragmentation event. New unique pathways of proton migration from the ribose have been elucidated and are rationalized by a charge-remote fragmentation pathway. (J Am Soc Mass Spectrom 2002, 13, 1407-1417) 2002 American Society for Mass Spectrometry
-
Ttrometry has been an active area for many years
he analysis of oligonucleotides using mass
specdue to the demands for a rapid and accurate
DNA sequencing technique as well as for the
identification and characterization of modified
oligonucleotides [1, 2]. Since Matrix-Assisted Laser Desorption/
Ionization (MALDI) has been introduced, [3, 4] much
effort has been devoted to the development of
oligonucleotide analysis by MALDI mass spectrometry. To
date, Tang et al. [5] and Liu et al. [6] have reported the
analysis of PCR products 500 to 600 bases using
UV-MALDI (at 337/355 nm), and Hillenkamp and
co-workers extended the upper limit of DNA and RNA
analyses up to 2180 bases using IR-MALDI [7].
Applications such as analyzing short tandem repeats [8],
gene-defect diseases [9 11], and genotyping of single
nucleotide polymorphisms [12] have been attempted.
Despite these successes, oligonucleotide analysis by
MALDI-MS has been limited primarily by the low ion
abundances of larger oligonucleotides. Routine analysis
by UV-MALDI is still limited to approximately 50 bases
in cases where good mass spectrometry performance
(e.g., resolution and reproducibility) is of high
importance [13]. The vast majority of the analyses of
oligonucleotides by MALDI is limited to the identification and
characterization of short chain oligonucleotides, such as
antisense drugs [14, 15].
The most probable cause for the limited success at
very high masses is that larger oligonucleotide ions are
unstable and dissociate rapidly after the desorption/
ionization step. A large number of studies have been
undertaken to better understand the fragmentation
process [16 34]. Generally, the severity of fragmentation is
matrix-dependent, and a matrix of choice is
3-hydroxypicolinic acid (3-HPA) [35, 36] (or the two
component matrix-3HPA/picolinic acid [5]) because it
results in a lesser amount of fragmentation and enables
the analysis of larger oligonucleotides. The popular
matrices for peptides and proteins, such as sinapinic
acid (SA) instrument facility tend to produce significant
fragmentation when used for oligonucleotide analysis.
Despite the matrix dependency on the degree of
decomposition, the types of observed fragment ions remain
similar for almost all matrices that have been studied.
The stability of molecular ions of oligodeoxynucleotides
is highly dependent on their base composition. Studies
of homopolymer and mixed bases of oligonucleotides
reveal that fragmentation often occurs at the location of
a guanosine, adenosine, or cytidine base (in decreasing
order of fragmentation severity) [22, 37, 38]. The laser
wavelength does not seem to be a determining factor.
3-HPA is an effective matrix for oligonucleotides at 266,
307, 337, and 355 nm. Even though IR photons have less
energy, IR absorbing matrices such as succinic acid do
not offer as much success as 3-HPA in the UV range [21,
39].
Several hypotheses have been proposed to
rationalize observed fragmentation pathways, and all of the
proposed mechanisms involve protonation of a
nucleobase as the initiating step [22, 24, 39]. It has been
proposed that proton transfer occurs from either the
neighboring acidic phosphodiester groups to the
nucleobase, thus forming short-lived zwitterionic
intermediates, or from matrix ions to form stable zwitterions
upon protonation of the nucleobase. The N-glycosidic
bond is then polarized and weakened, which initiates
base elimination leading to strand scission along the
phosphodiester backbone. These hypotheses nicely
explain the correlation of proton affinities (PA) of the four
nucleobases to the observed fragmentation patterns:
Gua (229.3 kcal/mol) Ade (225.3 kcal/mol) Cyt
(227.0 kcal/mol) Thy (210.5 kcal/mol) [40]. Evidence
supporting this ranking comes from the absence of
prompt fragment ions from poly-(T)n desorbed from a
variety of matrices in linear time-of-flight mass
spectrometry. The 7-deaza purine analogs of guanosine and
adenosine that are less acid labile in aqueous solutions
compared to the native forms exhibit higher stability (as
[M H] ) in MALDI [24, 41]. Thermodynamic
arguments have been used to support the influence of the
matrix on the fragmentation process. 3-HPA has the
highest proton affinity (214.1214.5 kcal/mol) and
therefore, when protonated, has the least capability to
transfer that proton to an oligonucleotide compared to
other common oligonucleotide matrices such as THAP
(210.8 kcal/mol) and 2,5-DHBA (201208 kcal/mol) [25,
42 44].
To probe the fragmentation mechanisms in more
detail, Hettich et al. used MALDI Fourier Transform Ion
Cyclotron Resonance Mass Spectrometer (FTICR-MS) to
characterize the structures of native and modified
oligonucleotides through prompt and collision-induced
dissociation fragmentation [45]. They elucidated the
structures of fragment ions (metastable and stable
prompt ions) and proposed fragmentation pathways.
Gross et al. have investigated fragmentation pathways
of metastable ions of dideoxynucleotide tetramers using
post-source decay (PSD) and H/D exchange in a
delayed extraction reflecting MALDI-TOF [28, 33]. They
reported many fragment ions that are not observed in
regular TOF-MS, such as y- and z-series ions. The H/D
results revealed convincing details of the bond
cleavages by clarifying hydrogen transfers. Wang et al.
compared similarities and differences of T-rich
oligonucleoti (...truncated)