Mass spectrometry provides accurate characterization of two

Document Sample
Mass spectrometry provides accurate characterization of two Powered By Docstoc

              Mass spectrometry provides accurate
             characterization of two genetic marker
                   types in Bacillus anthracis
             Matthew N. Van Ert1, Steven A. Hofstadler2, Yun Jiang2, Joseph D. Busch1,
       David M. Wagner1, Jared J. Drader2, David J. Ecker2, James C. Hannis2, Lynn Y. Huynh1,
                    James M. Schupp1, Tatum S. Simonson1, and Paul Keim1,3

                                                   BioTechniques 37:642-651 (October 2004)

    Epidemiological and forensic analyses of bioterrorism events involving Bacillus anthracis could be improved if both variable num-
    ber tandem repeats (VNTRs) and single nucleotide polymorphisms (SNPs) could be combined on a single analysis platform. Here
    we present the use of electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (ESI-FTICR-MS) to
    characterize 24 alleles from 6 VNTR loci and 11 alleles from 7 SNP loci in B. anthracis. The results obtained with ESI-FTICR-
    MS were consistent with independent results obtained from traditional approaches using electrophoretic detection of fluorescent
    products. However, ESI-FTICR-MS improves on the traditional approaches because it does not require fluorescent labeling of PCR
    products, minimizes post-PCR processing, obviates electrophoresis, and provides unambiguous base composition of both SNP and
    VNTR PCR products. In addition, ESI-FTICR-MS allows both marker types to be examined simultaneously and at a rate of approxi-
    mately 1 sample per min. This technology represents a significant advance in our ability to rapidly characterize B. anthracis isolates
    using VNTR and SNP loci.

INTRODUCTION                                    diversity (8,9). SNPs, discovered by se-          a methodology that minimizes post-
                                                quencing the entire protective antigen            PCR processing and evaluates SNP and
   Current interest in Bacillus anthra-         gene (pagA) from diverse B. anthra-               VNTR loci on a single platform.
cis, the etiological agent of the disease       cis isolates (3), have also been used to             Mass spectrometry is rapidly emerg-
anthrax, stems from its potential use as        determine phylogenetic relationships              ing as a sensitive and accurate method
a bioterrorism or biowarfare agent. The         within this pathogen. Appropriately,              for characterizing PCR products. While
use of B. anthracis spores in the mail-         multiple locus VNTR analysis (MLVA)               PCR products have been analyzed on
related bioterrorism events of 2001 em-         and pagA SNP typing played an impor-              several different mass spectrometry
phasized the need for rapid, high-reso-         tant role in differentiating and identify-        platforms including quadrupole (11–
lution molecular typing methods for             ing B. anthracis strains following the            14), time-of-flight (TOF; Reference
this pathogen. Until recently, however,         2001 bioterrorism events (10). Unfor-             15), and quadrupole ion traps (16,17),
development of these methods has been           tunately, current methods for generat-            no other platform can simultaneously
hampered by the highly monomorphic              ing data for SNP and VNTR markers                 provide the resolution and mass accu-
nature of the B. anthracis genome (1).          require two different reagent systems,            racy obtainable on the Fourier trans-
   Variable number tandem repeats               electrophoresis-based detection, and, in          form ion cyclotron resonance (FTICR)
(VNTRs) and single nucleotide poly-             the case of SNPs, extensive post-PCR              platform. Electrospray ionization Fou-
morphisms (SNPs) represent sources of           processing that is not only cumbersome            rier transform ion cyclotron resonance
variation in the genome and are recog-          but also introduces a potential source of         mass spectrometry (ESI-FTICR-MS)
nized as powerful tools for examining           laboratory contamination. Furthermore,            has been used to detect double-strand-
genetic relationships within B. anthra-         performing separate assays to collect             ed PCR products at the attomolar level
cis (2–7). VNTRs have been used suc-            data on SNP and VNTR loci can con-                (18), genotype simple and compound
cessfully to discriminate among closely         tribute to difficulties in sample and data        short tandem repeat sequences (18–20),
related strains of B. anthracis and have        tracking that can ultimately impact data          characterize PCR products generated
facilitated detailed epidemiological            quality. Rapid genotyping of B. anthra-           from amplification of 16S to 23S rDNA
analyses of local patterns of anthrax           cis would, therefore, be facilitated by           intergenic spacer regions (ISRs) from

       1Northern Arizona    University, Flagstaff, AZ, 2Isis Pharmaceuticals, Inc., Carlsbad, CA, and 3TGen, Phoenix, AZ, USA

642 BioTechniques                                                                                                          Vol. 37, No. 4 (2004)
Table 1. PCR Product Size, Position, and States for Seven pagA SNPs
                                                                                                                          Base Compositions as
 SNP                                        pagA Nucleotide                             Frequency in 23                Determined by ESI-FTICR-MS
 Name                   Product Size           Positiona              SNP State            Isolatesb                      [SNP State (A:G:C:T)]c

 PAGA01                      77                     1998                 T↔C                  16↔7                              T (17:13:14:33)
                                                                                                                                C (17:13:15:32)

 PAGA02                      84                     2883                 A↔G                  0↔23f                             G (25:21:14:24)

 PAGA03d                     65                     3481                 C↔T                  0↔23f                            TT (29:12:9:15)d
 PAGA04d                     65                     3496                                                                       TC (29:12:10:14)d
                                                                         T↔C                  2↔21

 PAGA05e                     69                     3602                 T↔C                  14↔9                            TT (33:10:8:18)e
                                                                                                                              CT (33:10:9:17)e
 PAGA06e                     69                     3606                 C↔T                  1↔22                            CC (33:10:10:16)e

 PAGA07                      63                     3672                 G↔A                  0↔23f                              A (29:9:4:21)

 SNP, single nucleotide polymorphism; ESI-FTICR-MS, electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry.
 aAs reported previously by Price et al. (3), nucleotide positions are based on the 4235-bp pXO1 sequence containing pag in its entirety (GenBank® accession no.

 bDNA from one isolate (pXO1 negative) in this study did not support amplification of pagA loci.
 cThe reported base compositions of the products are the only ones consistent with the measured molecular masses and the mass measurement uncertainties.
 dPAGA03 and PAGA04 are separated by 15 nucleotide positions in a single PCR product. Base changes and product base compositions are reported sequen-

  tially for each SNP.
 ePAGA05 and PAGA06 are separated by 4 nucleotide positions in a single PCR product. Base changes and product base compositions are reported sequentially

  for each SNP.
 fPAGA02 mutation has only been observed in one South African Bacillus anthracis isolate. PAGA03 and PAGA07 mutations are observed only in Sverdlovsk

  clinical samples (3).

Bacillus cereus strains (21), and de-                  described heat lysis method (6). Brief-                 than 100 bp (Table 1). Similarly, only
tect heterogeneity in 16S to 23S rDNA                  ly, a 1.0-μL inoculating loop was used                  VNTR products between 100–200 bp
ISRs within the genome of a single B.                  to transfer a portion of a B. anthracis                 were examined in this study (Table 2).
cereus strain (21,22).                                 colony into 200 μL of TE [Tris-HCl,
    As first demonstrated by the groups                pH 8.0, 1.0 mM ethylenediamine tetra-                   pagA SNP PCR
of Aaserud et al. (23) and Muddiman et                 acetate (EDTA)]. The colony material
al. (24), accurate mass measurements                   was dispersed by vortex mixing and                         The PCR conditions for the pagA
obtained by high-performance mass                      was incubated at 95°C for 20 min. Fol-                  SNP PCR products (Table 1) were as
spectrometry can be used to unambigu-                  lowing heat lysis, cellular debris was                  follows: initial denaturation at 95°C for
ously derive base compositions from                    removed by centrifugal filtration using                 5 min, followed by 34 cycles of 95°C
double-stranded DNA constructs using                   an Ultra-Free®-MC GV .22 μm cen-                        for 20 s, 60°C for 30 s, and 72°C for
the mathematical constraints imposed                   trifugal filter unit (Millipore, Billerica,             30 s. Each 20-μL reaction contained 1
by the complementary nature of the                     MA, USA) at 5000× g for 5 min. The                      μL of template DNA, 1× PCR buffer,
two strands. More recently, Jiang and                  filtrate was diluted 1:10 and was used                  3.0 mM MgCl2, 0.1 mM dNTPs, 0.05
Hofstadler (25) demonstrated that high-                as a template to support subsequent                     U Platinum® Taq DNA polymerase
resolution ESI-FTICR-MS measure-                       PCR for VNTR and SNP analyses.                          (Invitrogen, Carlsbad, CA, USA), and
ments could be used to unambiguously                                                                           0.2 μM of both the forward and reverse
calculate the base composition of 120                  ESI-FTICR-MS PCR Product                                primers. Following a 2-fold dilution of
bp PCR products in a high-throughput                   Design                                                  products in molecular-grade water, 20
modality. Here we demonstrate the po-                                                                          μL were removed for ESI-FTICR-MS
tential of ESI-FTICR-MS for rapid ge-                     PCR product characterization via                     analysis, and the remaining 20 μL were
notyping of B. anthracis strains using                 ESI-FTICR-MS is most effective for                      used for downstream single-base exten-
VNTR and SNP loci and compare this                     products with fragment sizes less than                  sion analysis.
technology to current methodologies.                   200 bp because of the increased diffi-
                                                       culty in generating single-stranded spe-                Single-Base Extension and Genotype
                                                       cies in the gas phase for larger products               Analyses
MATERIALS AND METHODS                                  and the inherent difficulty in obtaining
                                                       isotopic resolution for higher molecu-                     Single-base extension analysis was
DNA Isolation                                          lar weight species. Thus, to simplify                   performed according to the ABI Prism®
                                                       analysis of VNTR and SNP products in                    SNaPshot™ protocol (Applied Biosys-
   DNA was obtained from 24 diverse                    the same mass spectrometry run, SNP                     tems, Foster City, CA, USA). Briefly,
B. anthracis isolates using a previously               PCR products were designed at less                      the single nucleotide primer extension
Vol. 37, No. 4 (2004)                                                                                                                         BioTechniques 643

Table 2. VNTR Loci Information and Size Data for Observed Alleles as Detected by Gel Electrophoresis and ESI-FTICR-MS
                                                                     Allele Size Ranges                Observed Allele Sizes and Base
                                                                      as Determined by                 Compositions as Determined by
 VNTR Locus             Repeat Size             Motif                Gel Electrophoresis                      ESI-FTICR-MS
                           (bp)                                              (bp)                             [bp (A:G:C:T)]a

 BaVNTR12                    2                   AT                     112.44–112.55                           113 (39:20:19:35)
                                                                        114.45–114.57                           115 (40:20:19:36)

 pXO2                        2                   AT                         134.96                              135 (38:22:26:49)
                                                                        136.92–137.03                           137 (39:22:26:50)
                                                                        138.98–139.04                           139 (40:22:26:51)
                                                                        140.98–141.15                           141 (41:22:26:52)
                                                                            143.06                              143 (42:22:26:53)
                                                                            155.14                              155 (48:22:26:59)

 pXO1                        3                   AAT                        119.33                              120 (51:14:15:40)
                                                                            122.43                              123 (53:14:15:41)
                                                                            125.42                              126 (55:14:15:42)
                                                                        128.48–128.57                           129 (57:14:15:43)
                                                                        131.52–131.54                           132 (59:14:15:44)
                                                                        134.53–134.64                           135 (61:14:15:45)
                                                                            143.61                              144 (67:14:15:48)

 CG3                         5                 TAATA                    152.92–152.93                           153 (61:14:23:55)
                                                                        157.87–158.05                           158 (64:14:23:57)

 BaVNTR35                    6                TGATTG                        103.03                              103 (28:13:19:43)
                                                                        109.13–109.26                           109 (29:15:19:46)
                                                                         115.43–115.5                           115 (30:17:19:49)
                                                                            121.78                              121 (31:19:19:52)

 vrrB2                       9              CAACAATAT                    151.4–151.62                           153 (54:21:51:27)
                                                                        159.50–159.67                           162 (59:21:53:29)
                                                                        168.49–168.63                           171 (64:21:55:31)

 VNTR, variable number tandem repeats; ESI-FTICR-MS, electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry.
 aThe reported base compositions of the VNTR products are the only ones consistent with the measured molecular masses and the mass measurement


method involves the use of an unlabeled          DNA polymerase, and 0.2 μM of both               Prism 377 Automated Fluorescent DNA
oligonucleotide primer that is designed          the forward and reverse primers. The             Sequencer. GeneScan® software (Ap-
to anneal directly adjacent to the nu-           PCR thermal cycling parameters for               plied Biosystems) was used to analyze
cleotide base of interest. An extension          VNTR loci pXO1, pXO2, vrrB2, and                 the gel images and assign sizes to the
reaction is then performed in the pres-          CG3 were identical. Reactions were               alleles. Custom macros (available upon
ence of four possible dye-labeled ter-           incubated for 5 min at 94°C, followed            request from the corresponding author,
minators, resulting in the incorporation         by 34 cycles of 94°C for 20 s, 60°C for          Paul Keim) created in Genotyper® soft-
of a single fluorophore-labeled ddNTP.           20 s, and 65°C for 20 s. The final step          ware (Applied Biosystems), which use
We detected the labeled primer exten-            was 65°C for 5 min. PCR thermal cy-              fixed bin size assignments, allowed for
sion products by electrophoresis on an           cling parameters for BaVNTR12 and                the automated scoring of alleles.
ABI Prism 3100 Genetic Analyzer (Ap-             BaVNTR35 involved an incubation for
plied Biosystems), and genotypes were            5 min at 94°C; 35 cycles of 94°C for 30          Electrospray Ionization Fourier
manually assigned via differential base          s, 53.9°C for 30 s, and 72°C for 45 s;           Transform Ion Cyclotron Resonance
incorporation at the mutation site.              with a final step of 72°C for 5 min.             Mass Spectrometry

VNTR PCR                                         VNTR Electrophoresis and                            Prior to mass spectrometric analysis,
                                                 Genotype Analysis                                PCR products were purified and desalt-
   VNTR loci yielding products short-                                                             ed using a 96-well automated purifica-
er than 200 bp in size (Table 2) were               The VNTR PCR products obtained                tion protocol as previously described
amplified using PCR. Each 20-μL re-              from each B. anthracis isolate were              (25). Briefly, products from single 20-
action contained 1.0 μL of template              pooled equally and diluted 1:6 in mo-            μL aliquots of crude PCRs were bound
DNA, 1× PCR buffer, 2.0 mM MgCl2,                lecular-grade water. Products were elec-         to weak anion exchange ZipTips® (Mil-
0.2 mM dNTPs, 0.04 U Platinum Taq                trophoretically analyzed using an ABI            lipore, Bedford, MA, USA) and rinsed

644 BioTechniques                                                                                                           Vol. 37, No. 4 (2004)

repeatedly with 40 mM NH4HCO3,                          a CTC HTS PAL autosampler (LEAP                         mode whereby ions were accumulated
followed by multiple rinses with a so-                  Technologies, Carrboro, NC, USA),                       in the hexapole ion reservoir simultane-
lution containing 20% MeOH. Elution                     which was triggered by the FTICR data                   ously with ion detection in the trapped
of the final purified/desalted PCR prod-                station. Samples were injected with an                  ion cell. Following a 1.2-ms transfer
ucts was accomplished with a 10-μL                      integrated fluidics handling system,                    event, during which ions were trans-
aliquot of 1 M NH4OH, with each elu-                    which allowed for differential control                  ferred to the trapped ion cell, ions were
ent directly dispensed into a well of a                 of flow rate for fast rinsing between                   subjected to a 1.6-ms chirp excitation
96-well plate. Prior to analysis by ESI-                sample injections (26). During analy-                   corresponding to 500–8000 m/z (mass/
MS, the eluent was diluted 1:1 with a                   sis, PCR products were introduced                       charge ratio). Data were acquired over
solution containing 50% MeOH and 50                     into the mass spectrometer at a flow                    an m/z range of 500–5000 (1 M data
mM piperidine/imidizole. A small oli-                   rate of 1.25 μL/min. A source poten-                    points over a 225-kHz bandwidth).
gonucleotide (SH2; 5′-CGTGCATG-                         tial of -6 kV was applied to produce a
GCGG-3′) was added into the solution                    stable ion current. Countercurrent dry-
as an internal mass standard at a final                 ing gas was heated at 130°C to assist                   RESULTS
concentration of 50 nM.                                 desolvation. Ions were accumulated in
    A modified Apex II 70e (Bruker                      an external ion reservoir comprising an                    ESI-FTICR-MS analysis correctly
Daltonics, Billerica, MA, USA) ac-                      rf-only hexapole, a skimmer cone, and                   identified all 24 VNTR alleles (Table 2)
tively shielded FTICR-MS was utilized                   an auxiliary electrode for 2.3 s prior                  and all 11 pagA SNP alleles (Table 1),
in negative ionization mode. Desalted                   to transfer into the trapped ion cell for               which were also independently deter-
sample aliquots were injected directly                  mass analysis. Spectral acquisition was                 mined by electrophoretic-based analyses.
from sealed 96-well microplates using                   performed in the continuous duty cycle                  Figure 1A includes an electrophoretic gel

Figure 1. Electrophoretic and ESI-FTICR-MS analyses of three alleles in the pXO2 VNTR locus. (A) Electrophoretic gel image illustrating size data for
6 VNTR loci across 24 diverse isolates of Bacillus anthracis. The gel lanes and isolate identifications are sequential from left to right, and isolates/lanes 1, 5,
10, 15, and 20 are labeled for reference purposes. Each gel lane contains all VNTR products for a given isolate. The box and zoom highlight 3 different alleles
for the pXO2 locus, which differ by 2 bp. (B) Deconvoluted monoisotopic molecular weights and corresponding base compositions for the same 3 alleles
obtained using ESI-FTICR-MS analysis. Note that alleles differ only in the AT repeat. The products are monoadenylated (+A) due to the terminal transferase
activity of the Taq DNA polymerase. Additional peaks are present in the spectra, representing the nonadenylated and mono-adenylated complementary strand
of the product. The base compositions of the VNTR products given in the figure are the only ones consistent with the measured molecular masses and the mass
measurement uncertainties. VNTR, variable number tandem repeats; ESI-FTICR-MS, electrospray ionization Fourier transform ion cyclotron resonance mass
spectrometry; MW, molecular weight.

646 BioTechniques                                                                                                                            Vol. 37, No. 4 (2004)

image with size data for the 6 VNTR loci         DNA from one isolate did not support              nonadenylated forward-strand prod-
examined across 24 B. anthracis isolates.        amplification of any of the pagA SNP              uct in Figure 2B (molecular weight =
On the gel image, adjacent VNTR loci             loci or the VNTR pXO1 locus (Figure               23617.875 Da) within a 1 ppm mass
are labeled with different fluorophores,         1A, isolate #3), indicating that the iso-         measurement uncertainty (±0.002 Da)
which allowed for separation among sim-          late was pXO1 negative. In total, the             and 20 base compositions consistent
ilar-sized products from different loci. In      pagA SNPs resolved the 23 pXO1 posi-              with the nonadenylated reverse-strand
total, the 24 unique VNTR alleles sizes,         tive B. anthracis isolates into 5 unique          product (23802.066 ± 0.002 Da). Taking
as measured by GeneScan, ranged from             genotypes (data not shown).                       into account that the base compositions
103–171 bp in length (Table 2). The red              The mass measurement accuracy of              of the two strands are complementary,
box and zoom in Figure 1A indicate size          the 7 Tesla FTICR instrument (Apex                the list of putative base compositions
data for 3 alleles from the pXO2 VNTR            70e ESI-FTICR mass spectrometer;                  can be culled, leaving only possibili-
locus. Figure 1B gives an example of the         Bruker Daltonics) is generally better             ties in which the base composition of
corresponding data output from ESI-FTI-          than 1.5 ppm when an internal mass                the forward strand is complementary
CR-MS analysis of these same 3 alleles,          standard is used to post-calibrate each           to that of the reverse strand. The use
including size data and base pair com-           spectrum. The monoisotopic molecular              of Taq DNA polymerase somewhat
position. It is important to note that al-       weights for each strand are derived by            complicates this situation because the
though the products illustrated in Figure        an “averagine-like” fitting routine (27)          individual strands are adenylated and
1B are labeled with a fluorescent moiety         that fits the observed isotope envelope           are therefore, strictly speaking, not
to accommodate fluorescent detection on          to the distribution expected for a DNA            complementary. This situation is rem-
an electrophoretic platform, ESI-FTICR-          molecule of the approximate measured              edied by driving the mono-adenylation
MS does not require labeled products.            mass. For PCR products in this size               to completion, adjusting the measured
The deconvoluted monoisotopic mo-                range, an accurate molecular weight               molecular weights by the mass of an
lecular weights and corresponding base           determination of an individual strand             adenosine (313.0576 Da), and using
pair compositions (Figure 2B) accurately         is insufficient for base composition de-          the adjusted masses to calculate the
reflect the two base pair size differences       termination. For example, there are 41            base composition.
among these three alleles and the known          base compositions consistent with the                 While Muddiman and coworkers
repeat structure of this locus
(Table 2, ±AT). For all VNTR
alleles, the sequence composi-
tions determined by ESI-FTI-
CR-MS were consistent with
the size differences measured
by GeneScan and predicted
by the base composition of the
repeat structures of the loci
(Table 2). Collectively, the 6
VNTR loci resolved the 24
B. anthracis isolates into 23
unique genotypes (data not
    Figure 2 compares data
output from the ABI Prism
SNaPshot assay (Figure 2A)
and ESI-FTICR-MS analy-
sis (Figure 2B) of 2 alleles in
the pagA01 SNP locus. Both
techniques provided unam-
biguous scoring of the two-
allele states, however, unlike
the SNaPshot assay, ESI-
FTICR-MS also provides
very precise measurements
of the molecular mass of the
different products. For all 11 Figure 2. Electrophoretic and ESI-FTICR-MS analyses of two alleles in the PAGA01 SNP locus. (A) Electro-
SNP alleles, the base compo- pherograms indicating results of single-base extension analysis of both alleles (C, T) of the Bacillus anthracis pagA01
                                                                     mass spectra and                                                    derived base
sitions determined by ESI- locus. (B) DeconvolutedTESI-FTICRthe pagA01 locus. corresponding monoisotopic molecular weights and in the figure
                                   compositions of C and alleles of                   The base compositions of the SNP products given
FTICR-MS analysis were are the only ones consistent with the measured molecular masses and the mass measurement uncertainties. ESI-FTICR-
consistent with SNaPshot MS, electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry; SNP, single nucleotide poly-
single-base extension results. morphism.
648 BioTechniques                                                                                                                Vol. 37, No. 4 (2004)
                    RESEARCH REPORT

                    (28) have shown several methods by           product, ESI-FTICR-MS could accu-
                    which the nontemplated adenylation           rately and reliably detect these sequence
                    can be eliminated using alternative en-      variants. Thus, ESI-FTICR-MS has the
                    zymes such as Pfu and kodakaraensis          potential to characterize both VNTR and
                    (KOD), we chose to use products gen-         SNP loci more accurately than tradition-
                    erated under identical conditions (in-       al electrophoretic-based methods.
                    cluding fluorescently labeled primers)           In addition to potential increased
                    to allow a direct comparison of the two      accuracy, ESI-FTICR-MS provides
                    detection approaches. Consequently,          several other advantages over tradi-
                    the base compositions of the deconvo-        tional VNTR and SNP analysis plat-
                    luted spectra shown in Figure 2B are         forms. (i) ESI-FTICR-MS does not
                    annotated with “+A” to be consistent         require product labeling, so there is no
                    with the measured molecular weights.         need for expensive fluorescently la-
                    This precision allowed the unambigu-         beled amplification primers for VNTR
                    ous determination of the single change       analysis. (ii) Other than a simple de-
                    in base pair composition. ESI-FTICR-         salting step, there is no post-PCR pro-
                    MS also characterized the products on        cessing of the reaction, which is re-
                    both the forward and reverse strands,        quired by single-base extension assays
                    which provided two measurements for          for SNP analysis. (iii) Analysis is rap-
                    each SNP allele (Figure 2B).                 id, and complete analysis of 96 sam-
                                                                 ples requires approximately 96 min.
                                                                 (iv) Although ESI-FTICR-MS is cur-
                    DISCUSSION                                   rently only applicable to products less
                                                                 than 200 bp in size, multiple SNP and
                        Our results indicate that ESI-FTICR-     VNTR loci can be analyzed simultane-
                    MS is just as accurate as traditional,       ously, which promises to significantly
                    electrophoretic-based platforms in iden-     increase the throughput of these analy-
                    tifying alleles at VNTR and SNP loci in      ses. Preliminary studies in our labora-
                    B. anthracis. Furthermore, ESI-FTICR-        tory suggest that similar analyses can
                    MS provides complete base composi-           be performed on a low-cost bench-top
                    tion of the VNTR and SNP products.           ESI-TOF mass spectrometer; while
                    This additional information will allow       not offering the same mass resolu-
                    researchers to overcome several prob-        tion as the FTICR platform, the mass
                    lems with the current methodologies.         accuracy of the ESI-TOF instrument
                    Although not illustrated in this study, we   (approximately 10 ppm) is sufficient
                    have observed that, in rare cases, some      to provide unambiguous base compo-
                    VNTR loci may have multiple repeat           sitions for products less than 200 bp.
                    motifs. If one of the repeat motifs is a     Broad deployment of this genotyping
                    multiple of the other motif (e.g., 2 and 4   protocol on an ESI-TOF-based sys-
                    bp), they can produce alleles with iden-     tem is an attractive alternative to ESI-
                    tical sizes but different sequence com-      FTICR because of the significantly
                    positions. These differences are indistin-   smaller footprint and cost of the TOF
                    guishable using electrophoretic analyses     instruments. Further validation studies
                    and would be scored as the same allele.      are presently under way, which will be
                    Through accurate base composition            reported in detail elsewhere.
                    analysis, ESI-FTICR-MS could render              In conclusion, we have demonstrat-
                    these products distinguishable from one      ed that ESI-FTICR-MS is a potentially
                    another, thereby increasing genetic reso-    powerful method for high-throughput
                    lution. While analyzing large, diverse       genotyping of B. anthracis. The use
                    strain collections of B. anthracis and ge-   of ESI-FTICR-MS to simultaneously
                    netic near-neighbors, we have observed       examine multiple genetic marker types
                    that products may contain additional         represents an attractive alternative to
                    unexpected SNPs in close proximity           traditional methods in terms of cost-ef-
                    to the target SNP. In these situations,      fectiveness, efficacy, throughput, and
                    single-base extension assays will not        information content. This technology
                    detect these additional sequence vari-       will significantly advance the ability
                    ants, which may interfere with the assay     to accurately and rapidly identify and
                    and lead to errors. However, because it      characterize isolates of B. anthracis
                    interrogates the entire sequence of the      and other pathogens.
650 BioTechniques                                                                       Vol. 37, No. 4 (2004)
ACKNOWLEDGEMENTS                                          L.W. Mayer, and T. Popovic. 2002. Molecular            ESI-FTICR mass spectrometry. Anal. Chem.
                                                          subtyping of Bacillus anthracis and the 2001           70:1203-1207.
                                                          bioterrorism-associated anthrax outbreak.           22.Muddiman, D.C., D.S. Wunschel, C. Liu, L.
   We would like to thank Jeff Henrik-                    Emerg. Infect. Dis. 8:1111-1116.                       Pasa-Tolic, K.F. Fox, A. Fox, G.A. Anderson,
son, Chris Allender, and Ryan Easter-                  11.Krahmer, M.T., Y.A. Johnson, J.J. Walters,             and R.D. Smith. 1996. Characterization of
day for their technical assistance. This                  K.F. Fox, A. Fox, and M. Nagpal. 1999. Elec-           PCR products from Bacilli using electrospray
work was supported by funding from                        trospray quadrupole mass spectrometry analy-           ionization FTICR mass spectrometry. Anal.
the U.S. Department of Energy, Chemi-                     sis of model oligonucleotides and polymerase           Chem. 68:3705-3712.
                                                          chain reaction products: determination of base      23.Aaserud, D.J., N.L. Kelleher, D.P. Little, and
cal, and Biological Nonproliferation                      substitutions, nucleotide additions/deletions,         F.W. McLafferty. 1996. Accurate base com-
Program and by the Federal Bureau of                      and chemical modifications. Anal. Chem.                position of double-strand DNA by mass spec-
Investigation.                                            71:2893-2900.                                          trometry. J. Am. Soc. Mass Spectrom. 7:1266-
                                                       12.Johnson, Y.A., M. Nagpal, M.T. Krahmer,                1269.
                                                          K.F. Fox, and A. Fox. 2000. Precise molecular       24.Muddiman, D.C., G.A. Anderson, S.A. Hof-
                                                          weight determination of PCR products of the            stadler, and R.D. Smith. 1997. Length and
                                                          rRNA intergenic spacer region using electro-           base composition of PCR-amplified nucleic
STATEMENT                                                 spray quadrupole mass spectrometry for dif-            acids using mass measurements from elec-
                                                          ferentiation of B. subtilis and B. atrophaeus,         trospray ionization mass spectrometry. Anal.
   The authors declare no competing                       closely related species of bacilli. J. Microbiol.      Chem. 69:1543-1549.
interests.                                                Methods 40:241-254.                                 25.Jiang, Y. and S.A. Hofstadler. 2003. A highly
                                                       13.Walters, J.J., W. Muhammad, K.F. Fox, A.               efficient and automated method of purifying
                                                          Fox, D. Xie, K.E. Creek, and L. Pirisi. 2001.          and desalting PCR products for analysis by
                                                          Genotyping single nucleotide polymorphisms             electrospray ionization mass spectrometry.
REFERENCES                                                using intact polymerase chain reaction prod-           Anal. Biochem. 316:50-57.
                                                          ucts by electrospray quadrupole mass spec-          26.Sannes-Lowery, K.A., J.J. Drader, R.H.
 1.Keim, P., A. Kalif, J. Schupp, K. Hill, S.E.           trometry. Rapid Commun. Mass Spectrom.                 Griffey, and S.A. Hofstadler. 2000. Fourier
   Travis, K. Richmond, D.M. Adair, and M.                15:1752-1759.                                          transform ion cyclotron resonance mass spec-
   Hugh-Jones. 1997. Molecular evolution and           14.Naito, Y., K. Ishikawa, Y. Koga, T. Tsuneyo-           trometry as a high throughput affinity screen
   diversity in Bacillus anthracis as detected            shi, and H. Terunuma. 1995. Molecular mass             to identify RNA-binding ligands. Trends Anal.
   by amplified fragment length polymorphism              measurement of polymerase chain reaction               Chem. 19:481-491.
   markers. J. Bacteriol. 179:818-824.                    products amplified from human blood DNA             27.Senko, M.W., S.C. Beu, and F.W. McLaf-
 2.Andersen, G.L., J.M. Simchock, and K.H.                by electrospray ionization mass spectrom-              ferty. 1995. Determination of monoisotopic
   Wilson. 1996. Identification of a region of            etry. Rapid Commun. Mass Spectrom. 9:1484-             masses and ion populations for large biomol-
   genetic variability among Bacillus anthra-             1486.                                                  ecules from resolved isotopic distributions. J.
   cis strains and related species. J. Bacteriol.      15.Deforce, D.L., J. Raymackers, L. Meheus,               Am. Soc. Mass Spectrom. 6:229-233.
   178:377-384.                                           F. Van Wijnendaele, A. De Leenheer, and G.          28.Benson, L.M., A.P. Null, and D.C. Muddi-
 3.Price, L.B., M. Hugh-Jones, P.J. Jackson,              Van den Eckhout. 1998. Characterization of             man. 2003. Advantages of thermococcus ko-
   and P. Keim. 1999. Genetic diversity in the            DNA oligonucleotides by coupling of capillary          dakaraensis (KOD) DNA polymerase for PCR-
   protective antigen gene of Bacillus anthracis.         zone electrophoresis to electrospray ioniza-           mass spectrometry based analyses. J. Am. Soc.
   J. Bacteriol. 181:2358-2362.                           tion Q-TOF mass spectrometry. Anal. Chem.              Mass Spectrom. 14:601-604.
 4.Le Fèche, P., Y. Hauck, L. Onteniente, A.              70:3060-3068.
   Prieur, F. Denoeud, V. Ramisse, P. Sylves-          16.M.T. Krahmer, J.J. Walters, K.F. Fox, A.
   tre, G. Benson, et al. 2001. A tandem repeats          Fox, K.E. Creek, L. Pirisi, D.S. Wunschel,          Received 9 February 2004; accepted
   database for bacterial genomes: application to         R.D. Smith, et al. 2000. MS for identification      11 May 2004.
   the genotyping of Yersinia pestis and Bacillus         of single nucleotide polymorphisms and MS/
   anthracis. BMC Microbiol. 1:2.                         MS for discrimination of isomeric PCR prod-
                                                          ucts. Anal. Chem. 72:4033-4040.                     Address correspondence to:
 5.Sylvestre, P., E. Couture-Tosi, and M. Mock.
   2003. Polymorphism in the collagen-like re-         17.Hahner, S., A. Schneider, A. Ingendoh, and J.       Paul Keim
   gion of the Bacillus anthracis BclA protein            Mosner. 2000. Analysis of short tandem repeat       Department of Biological Sciences
   leads to variation in exosprium filament length.       polymorphisms by electrospray ion trap mass         Northern Arizona University
   J. Bacteriol. 185:1555-1563.                           spectrometry. Nucleic Acids Res. 28:E82             Flagstaff, AZ 86011-5640, USA
 6.Keim, P., L.B. Price, A.M. Klevytska, K.L.          18.Hannis, J.C. and D.C. Muddiman. 2001.               e-mail:
   Smith, J. M. Schupp, R. Okinaka, P.J. Jack-            Detection of double-stranded PCR amplicons          or
   son, and M. Hugh-Jones. 2000. Multiple-lo-             at the attomole level electrosprayed from low       Steven A. Hofstadler
   cus variable-number tandem repeat analysis             nanomolar solutions using FT-ICR mass spec-         Ibis Therapeutics, A Division of Isis
   reveals genetic relationships within Bacillus          trometry. Fresenius J. Anal. Chem. 369:246-           Pharmaceuticals
   anthracis. J. Bacteriol. 182:2928-2936.                251.                                                2292 Faraday Avenue
 7.Keim, P., M.N. Van Ert, T. Pearson, A.J. Vo-        19.Hannis, J.C. and D.C. Muddiman. 1999. Ac-           Carlsbad, CA 92008, USA
   gler, L.Y. Huynh, and D.M. Wagner. 2004.               curate characterization of the tyrosine hydrox-     e-mail:
   Anthrax molecular epidemiology and forensics:          ylase forensic allele 9.3 through development
   using the appropriate marker for different evolu-      of electrospray ionization fourier transform ion
   tionary scales. Infect. Genet. Evol. 4:205-213.        cyclotron resonance mass spectrometry. Rapid
 8.Patra, G., J. Vaissaire, M. Weber-Levy, C.             Commun. Mass Spectrom. 13:954-962.
   Le Doujet, and M. Mock. 1998. Molecular             20.Null, A.P., J.C. Hannis, and D.C. Muddiman.
   characterization of Bacillus strains involved in       2001. Genotyping of simple and compound
   outbreaks of anthrax in France in 1997. J. Clin.       short tandem repeat loci using electrospray
   Microbiol. 36:3412-1563.                               ionization fourier transform ion cyclotron
 9.Smith, K.L., V. DeVos, H. Bryden, L.B. Price,          resonance mass spectrometry. Anal. Chem.
   M. Hugh-Jones, and P. Keim. 2000. Bacillus             73:4514-4521.
   anthracis diversity in Kruger National Park. J.     21.Wunschel, D.S., D.C. Muddiman, K.F. Fox,
   Clin. Microbiol. 38:3780-3784.                         A. Fox, and R.D. Smith.1998. Heterogeneity
10.Hoffmaster, A.R., C.C. Fitzgerald, E. Ribot,           in Bacillus cereus PCR products detected by

Vol. 37, No. 4 (2004)                                                                                                                        BioTechniques 651