A point mutation converts dihydroneopterin aldolase to a cofactor-independent oxygenase

Article abstract of DOI:10.1021/ja063455i

Dihydroneopterin aldolase (DHNA) catalyzes the conversion of 7,8-dihydroneopterin (1) to 6-hydroxymethyl-7,8-dihydropterin (4) in the folate biosynthetic pathway. Substitution of a conserved tyrosine residue at the active site of DHNA by phenylalanine converts the enzyme to a cofactor-independent oxygenase, which generates mainly 7,8-dihydroxanthopterin (6) rather than 4. 6 is generated via the same enol intermediate as in the wild-type enzyme-catalyzed reaction, but this species undergoes an oxygenation reaction to form 6. The conserved tyrosine residue plays only a minor role in the formation of the enol reaction intermediate but a critical role in the protonation of the enol intermediate to form 4.

Full text of DOI:10.1021/ja063455i

Published on Web 09/19/2006
A Point Mutation Converts Dihydroneopterin Aldolase to a
Cofactor-Independent Oxygenase
Yi Wang,‡ Gwynyth Scherperel,§ Kade D. Roberts,§ A. Daniel Jones,‡,§
Gavin E. Reid,*,§,‡ and Honggao Yan*,‡
Contribution from the Department of Biochemistry and Molecular Biology and Department of
Chemistry, Michigan State UniVersity, East Lansing, Michigan 48824
Received May 17, 2006; E-mail: yanh@msu.edu; reid@chemistry.msu.edu
Abstract: Dihydroneopterin aldolase (DHNA) catalyzes the conversion of 7,8-dihydroneopterin (1) to
6-hydroxymethyl-7,8-dihydropterin (4) in the folate biosynthetic pathway. Substitution of a conserved tyrosine
residue at the active site of DHNA by phenylalanine converts the enzyme to a cofactor-independent
oxygenase, which generates mainly 7,8-dihydroxanthopterin (6) rather than 4. 6 is generated via the same
enol intermediate as in the wild-type enzyme-catalyzed reaction, but this species undergoes an oxygenation
reaction to form 6. The conserved tyrosine residue plays only a minor role in the formation of the enol
reaction intermediate but a critical role in the protonation of the enol intermediate to form 4.
Introduction
Dihydroneopterin aldolase (DHNA) catalyzes the conversion
of 7,8-dihydroneopterin (1, Figure 1) to 6-hydroxymethyl-7,8-
dihydropterin (4) in the folate biosynthetic pathway, one of
principal targets for developing antimicrobial agents.1 Like other
enzymes, such as dihydropteroate synthase and dihydrofolate
reductase, in the folate pathway,2,3 DHNA is an attractive target
for developing antimicrobial agents.4 DHNA is also a unique
aldolase in that it requires neither the formation of a Schiff base
between the substrate and the enzyme nor metal ions for
catalysis,5 and in addition to the aldolase reaction, DHNA also
catalyzes the epimerization at the 2′-carbon of 1 to generate
7,8-dihydromonapterin6 (5). Interestingly, DHNAs from Sta-
phylococcus aureus (SaDHNA) and Escherichia coli (EcDH-
NA), representatives of DHNAs from Gram-positive and Gram-
negative bacteria, respectively, have significant differences in
binding and catalytic properties.
One of the conserved residues at the active site of DHNA is
a tyrosine residue, Y54 in SaDHNA and Y53 in EcDHNA.
According to the crystal structure of SaDHNA in complex with
4,7 the pterin ring is stacked with the phenol ring of Y54 (Figure
2). The hydroxyl group of the phenol ring of Y54 is hydrogen-
bonded to the amino group of the putative general base of K100
and the hydroxyl of the 6-hydroxymethyl group of the bound
4. To investigate the role of the conserved tyrosine residue in
catalysis, we replaced Y54 of SaDHNA and Y53 of EcDHNA
‡ Department of Biochemistry and Molecular Biology.
§ Department of Chemistry.
(1) Walsh, C. Nat. ReV. Microbiol. 2003, 1, 65-70.
(2) Bermingham, A.; Derrick, J. P. Bioessays 2002, 24, 637-648.
(3) Kompis, I. M.; Islam, K.; Then, R. L. Chem. ReV. 2005, 105, 593-620.
(4) Sanders, W. J.; et al. J. Med. Chem. 2004, 47, 1709-1718.
(5) Mathis, J. B.; Brown, G. M. J. Biol. Chem. 1970, 245, 3015-3025.
(6) Haussmann, C.; Rohdich, F.; Schmidt, E.; Bacher, A.; Richter, F. J. Biol.
Chem. 1998, 273, 17418-17424.
Figure 1. Reactions catalyzed by DHNA and related pterin compounds.
with a phenylalanine residue by site-directed mutagenesis. The
single-point mutations convert the aldolases to oxygenases,
which generate mainly 7,8-dihydroxanthopterin (6) rather than
(7) Hennig, M.; D’Arcy, A.; Hampele, I. C.; Page, M. G. P.; Oefner, C.; Dale,
G. E. Nat. Struct. Biol. 1998, 5, 357-362.
9
13216
J. AM. CHEM. SOC. 2006, 128, 13216-13223
10.1021/ja063455i CCC: $33.50 © 2006 American Chemical Society

Structure−Function Relationships of DHNA
A R T I C L E S
µM SaY54F was titrated with a stock solution of 8 or 9. To determine
the Kd values of the binding of 10 to SaY54F and the binding of 8, 9,
and 10 to EcY53F, a 2 mL solution containing 1 µM of one of the
ligands (8, 9, or 10) was titrated with a stock solution of SaY54F or
EcY53F. The Kd values were obtained by nonlinear least-squares fitting
of the titration data as previously described.9
Stopped-Flow Analysis. Stopped-flow experiments were performed
on an Applied Photophysics SX.18MV-R stopped-flow spectrofluo-
rometer at 25 °C. One syringe contained the protein (SaY54F or
EcY53F), and the other contained 8, 9, or 10. The protein concentration
was 1 or 2 µM, and the ligand concentrations ranged from 5 to 60 µM.
All concentrations reported are those after the mixing of the two syringe
solutions. Fluorescence traces were obtained with an excitation
wavelength of 360 nm and a filter with a cutoff of 395 nm for emission.
Apparent rate constants were obtained by nonlinear least-squares fitting
of the data to a single-exponential equation and were replotted against
the ligand concentrations. The association and dissociation constants
were obtained by linear regression of the apparent rate constants vs
ligand concentration data.
Kinetic Assay. The kinetic experiments were performed manually.
Both 1 and DHNA were dissolved in a buffer containing 100 mM Tris-
HCl, 1 mM ethylenediaminetetraacetic acid, and 5 mM dithiothreitol,
pH 8.3. The reactions were initiated by the addition of DHNA and
stopped with 1 N HCl. The stopped reaction mixtures were processed
and separated by HPLC as previously described.6
NMR Spectroscopy. NMR measurements were made at 25 °C with
a Varian Inova 600 spectrometer. The initial NMR sample contained
1 mM 1 and 1 mM tris(2-carboxyethyl) phosphine (TCEP) in 100 mM
sodium phosphate buffer, pH 8.3 (pH meter reading without correction
for deuterium isotope effects), made with 90% H2O and 10% D2O.
The reaction was initiated with 1 µM SaDHNA or 3 µM SaY54F. NMR
spectra were recorded before and after the addition of the enzyme. The
spectral width for the NMR data was 12000 Hz, with the carrier
frequency at the HDO resonance. The solvent resonance was suppressed
by presaturation. Each FID was composed of 16K data points with 64
transients. The delay between successive transients was 1.7 s. The time
domain data were processed by zero-filling to 32K points, exponential
multiplication (1 Hz), and Fourier transformation. Chemical shifts were
referenced to the internal standard sodium 2-dimethyl-2-silapentane-
5-sulfonate sodium salt.
Mass Spectrometry (MS). All MS experiments were performed
using a Thermo model LTQ linear ion trap mass spectrometer. The
initial 300 µL samples contained 200 µM 1 and 4 µM SaDHNA or 10
µM SaY54F in 5 mM ammonium carbonate, pH 8. Twenty-microliter
aliquots of the reaction mixtures were taken out at 1-min intervals and
mixed with 80 µL of a solution containing 1% acetic acid and 50%
methanol. The samples were further diluted with the same solution and
then introduced into the mass spectrometer at a flow rate of 0.5 µL/
min by nanoelectrospray ionization (nESI). The following nESI
conditions were used: spray voltage, 1.8 kV; heated capillary temper-
ature, 200 °C; capillary voltage, -10 V; and tube lens voltage, -50
V. Collision-induced dissociation (CID) tandem mass spectrometry
(MS/MS and MS3) spectra were acquired at an activation q value of
0.25 using an isolation width of 1 or 2 Da (to monoisotopically isolate
the precursor ion), a normalized collision energy of 30-40%, and an
activation time of 30 or 300 ms. The values were chosen such that the
gentlest conditions were used in order to completely dissociate the
selected precursor ion (i.e., some precursor ions required a larger
normalized collision energy and/or longer activation time). The MS,
MS/MS, and MS3 product ion spectra shown are the averages of 60
individual mass analysis scans.
Figure 2. Polar interactions of 4 with SaDHNA.
4. 6 is generated via the same enol intermediate (2) as in the
wild-type enzyme-catalyzed reaction, but this species undergoes
an oxygenation reaction to form 6. The conserved tyrosine
residue plays only a minor role in the formation of the enol
reaction intermediate but a critical role in the protonation of
the enol intermediate to form 4.
Experimental Procedures
Materials. Compounds 1, 4, 5, 6, 7 (6-formyl-7,8-dihydropterin), 8
(D-neopterin), 9 (L-monapterin), and 10 (6-hydroxymethylpterin) were
purchased from Schircks Laboratories. Other chemicals were from
Sigma or Aldrich. Pfu DNA polymerase was purchased from Strategene.
Other molecular biology reagents were from New England Biolab.
Site-Directed Mutagenesis and Protein Purification. The mutants
SaY54F and EcY53F, where Y54 of SaDHNA and Y53 of EcDHNA
were replaced by a phenylalanine residue, respectively, were made by
a polymerase chain reaction-based method using high-fidelity pfu DNA
polymerase according to a protocol developed by Stratagene. The
forward and reverse primers for making the SaY54F mutant were 5′-
GTT ATTGATACAGTTCATTTTGGTGAAGTGTTCGAAGAG G-3′
and 5′-CCTCTTCGAACA CTTCACCAAAATGAACTGTATCAA-
TAAC-3, respectively. The forward and reverse primers for making
the EcY53F mutant were 5′-CGGATTGCCTCAGTTTCGCTGACAT-
TGCAGAAAC-3′ and 5′-GTTTCTGCAATGTCAGCGAAACTGAG-
GCAATCCG-3′, respectively. The mutants were selected by DNA
sequencing. To ensure that there were no unintended mutations in the
mutants, the entire coding sequences of the mutated genes were
determined.
The mutant proteins were purified as previously described (Wang
et al., unpublished). Briefly, SaY5F was purified to homogeneity by a
Ni-NTA column followed by a Bio-Gel A-0.5m gel column (Bio-Rad).
EcY53F was purified by a (diethylamino)ethyl-cellulose column
(Whatman) followed by a Bio-Gel A-0.5m gel column. The purities of
the protein preparations were checked by sodium dodecyl sulfate-
polyacrylamide gel electrophoresis. The amino acid sequences of the
purified proteins were confirmed by “top-down” tandem mass spec-
trometry.8 The purified proteins were concentrated, dialyzed, lyophi-
lized, and stored at -80 °C.
Equilibrium Binding Studies. The procedures for the equilibrium
binding studies of DHNAs were essentially the same as previously
described for the similar studies of 6-hydroxymethyl-7,8-dihydropterin
pyrophosphokinase using a Spex FluoroMax-2 fluorometer.9,10 Briefly,
proteins and ligands were all dissolved in 100 mM Tris-HCl, pH 8.3,
and the titration experiments were performed in a single cuvette with
a Spex FluoroMax-2 fluorometer at 24 °C. To determine the Kd values
of the binding of 8 and 9 to SaY54F, a 2 mL solution containing 10
Gas Chromatography/Mass Spectrometry (GC/MS) of Glycola-
ldehyde (3). The derivatization protocol is a modified version of a
procedure developed for analysis of plant metabolites,11 with N-(tert-
butyldimethylsilyl)-N-methyltrifluoroacetamide (MTBSTFA) substituted
for N-methyl-N-(trimethylsilyl)trifluoroacetamide as needed to form tert-
(8) Scherperel, G.; Yan, H. G.; Wang, Y.; Reid, G. E. Analyst 2006, 131, 291-
302.
(9) Li, Y.; Gong, Y.; Shi, G.; Blaszczyk, J.; Ji, X.; Yan, H. Biochemistry 2002,
41, 8777-8783.
(10) Shi, G.; Gong, Y.; Savchenko, A.; Zeikus, J. G.; Xiao, B.; Ji, X.; Yan, H.
Biochim. Biophys. Acta 2000, 1478, 289-299.
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13217

A R T I C L E S
Wang et al.
Table 1. Binding Constants of S. aureus and E. coli DHNAs and
YfF Mutantsa
butyldimethylsilyl (TBDMS) ether derivatives instead of the trimeth-
ylsilyl derivatives that were reported previously. The advantage is that
the TBDMS derivatives give more conclusive evidence of the product’s
molecular mass in the form of a strong [M - 57]+ ion in the mass
spectrum. The reaction contained 200 µM 1 and 4 µM SaDHNA or
6.3 µM SaY54F in 5 mM ammonium carbonate, pH 8.3. The reaction
was stopped after 20 min by adding 1 N HCl and centrifuged for 5
min with a microcentrifuge. The supernatant collected was then
lyophilized overnight to dryness. The dried residue was incubated for
1 h with 10 µL of methoxyamine hydrochloride in pyridine (10 mg/
mL) and MTBSTFA. GC/MS analyses were performed on an Agilent
5973 mass spectrometer coupled to a model 6890 gas chromatograph.
All analyses were performed using 70 eV electron ionization. Separa-
tions were performed using a 30 m HP-5MS column and helium (36
cm/s) as carrier gas. The column temperature was programmed from
50 (4 min hold) to 285 °C at 10 deg/min.
HPLC Identification of Formic Acid. Formic acid was identified
as previously described.12 One-milliliter reaction mixtures contained
200 µM 1 and 10 µM SaDHNA or SaY54F in 100 mM HEPES-KOH,
pH 8.3, and were mixed with 100 µL of 3.2 M H2SO4 after 20 min of
incubation at 24 °C. The solutions were centrifuged, filtered through a
0.45 µm syringe filter, and injected into a 7.8 mm × 300 mm Aminex
HPX-87H column. Standard compounds, including formic acid, acetic
acid, glycolic acid, glyceric acid, succinic acid, formaldehyde, glyco-
laldehyde (3), and glyceraldehyde, were treated the same way as the
enzymatic reaction mixture. The column was eluted with 4 mM
H2SO4 at a flow rate of 0.6 mL/min, and the elution was monitored by
a Waters 2487 UV detector at 210 nm and a Waters 410 differential
refractometer.
SaDHNA
SaY54F
EcDHNA
EcY53F
8
9
Kd 18 ( 2
k1
k-1 4.5 ( 0.1
4.5 ( 0.9
0.77 ( 0.06
0.20 ( 0.01
0.26 ( 0.01
0.060 ( 0.006
0.24 ( 0.01 0.099 ( 0.005 0.32 ( 0.02
0.44 ( 0.05
0.29 ( 0.03
Kd 13 ( 1
3.7 ( 0.5
2.6 ( 0.06
0.26 ( 0.01
0.58 ( 0.03
0.40 ( 0.01
0.28 ( 0.01
0.11 ( 0.01
k1
0.29 ( 0.02 0.13 ( 0.01
k-1 4.2 ( 0.2
0.80 ( 0.09
10 Kd 24 ( 0.2
k1
k-1 10 ( 0.5
3.8 ( 0.9
0.10 ( 0.01
0.55 ( 0.04
0.40 ( 0.02
0.79 ( 0.01
0.45 ( 0.02 0.68 ( 0.01
2.9 ( 0.06
0.062 ( 0.001 0.26 ( 0.01
a Kd in µM, k1 in µM-1 s-1, and k-1 in s-1
.
of 4. The changes were mainly caused by the changes in the
dissociation rate constants. The results suggested that neither
the hydroxyl of Y54 of SaDHNA nor that of Y53 of EcDHNA
is critically important for the binding of substrate or products.
To our surprise, there was a dramatic drop in the total
fluorescence intensity in the HPLC analysis of the reaction
mixtures generated by either SaY54F or EcY53F as the reaction
progressed, suggesting that the substrate was converted to a
nonfluorescent compound at the chosen excitation and emission
wavelengths and the mutant enzyme-catalyzed reactions gener-
ated different products.
NMR Analysis. To identify the products of the mutant-
catalyzed reactions, we analyzed the reaction mixtures by NMR.
Representative spectra are shown in Figure 3, and the chemical
shifts of selected protons of related compounds are summarized
in Table 2. As shown in Figure 3, 3 was generated by both the
wild-type enzyme- and the mutant-catalyzed reactions. While
the wild-type enzyme-catalyzed reaction generated 4 along with
3 (Figure 3A), the mutant-catalyzed reaction generated little 4
(Figure 3B). The major product of the mutant-catalyzed reaction
was a compound with a proton peak at 4.12 ppm, which was
identified as 6, as detailed in the following section; 6 was a
minor product in the wild-type enzyme-catalyzed reaction. The
NMR spectra of the reaction mixtures generated by the E. coli
enzymes (data not shown) were very similar to those of the
reaction mixtures generated by the S. aureus enzymes. Because
both SaDHNA and SaY54F were saturated with the substrate 1
under the experimental conditions, their turnover numbers (kcat)
could be calculated on the basis of the enzyme concentration,
the initial substrate concentration, and the NMR peak intensities
of the substrate and products. The calculated kcat values for the
formation of 4 and 6 were 0.20 and 5.7 × 10-3 s-1, respectively,
for SaDHNA, and 3.4 × 10-4 and 0.018 s-1, respectively, for
SaY54F. The rate for the formation of 6 was ∼3% of that for
the formation of 4 in the SaDHNA-catalyzed reaction. The
relative rates were reversed in the SaY54F-catalyzed reaction,
and the rate for the formation of 4 was ∼2% of that for the
formation of 6.
Oxygen Consumption Assay. Oxygen consumption was measured
using a Gilson 5/6H oxygraph with a Clark oxygen electrode at 25 °C
as previously described.13 The initial 1.8 mL samples contained 200
µM 1 in 100 mM HEPES-KOH, pH 8.2. The reactions were initiated
by 8 µM SaDHNA or SaY54F at 25 °C. The turnover number for
SaY54F was calculated using the slope of the initial linear part of the
data obtained by linear regression.
Results
Biochemical Analysis. Previously we established the ther-
modynamic and kinetic framework for the structure-function
studies of SaDHNA and EcDHNA (Wang et al., unpublished).
We performed the biochemical characterizations of the mutants
SaY54F and EcY53F using the same strategy. The binding steps
were mimicked with the substrate and product analogues 8, 9,
and 10. The only difference between the substrate and product
analogues and the corresponding substrate and products is that
between C7 and N8 is a double bond in the analogues and a
single bond in the substrate and products. 8, 9, and 10 are
excellent substrate and product analogues (Wang et al., unpub-
lished). The Kd values were measured by equilibrium binding
experiments using fluorometry. The association and dissociation
rate constants were determined by kinetic binding experiments
using stopped-flow fluorometry. Representative results are
shown in Figures S1 and S2 for the equilibrium and kinetic
binding experiments, respectively. The complete results are
summarized in Table 1. For SaY54F, the mutation increased
the affinities of the ligands for the enzyme by a factor of ∼4-
6. For EcY53F, the mutation increased the affinities of 8 and 9
by a factor of 4-6, but decreased the affinity of 10 by a factor
MS Analysis. The identity of the major product of the
SaY54F-catalyzed reaction that gave rise to the 4.12 ppm peak
in the NMR spectrum could not be determined by 1D proton
NMR spectroscopy. To identify the reaction product, we
analyzed the reaction mixtures generated by both the wild-type
and the mutant enzymes by MS. As expected, the primary
product generated by SaDHNA was different from that gener-
ated by SaY54F (Figure 4). The mass spectrum of the reaction
involving SaDHNA showed the formation of a product ion at
(11) Roessner, U.; Wagner, C.; Kopka, J.; Trethewey, R. N.; Willmitzer, L.
Plant J. 2000, 23, 131-142.
(12) McKinlay, J. B.; Zeikus, J. G.; Vieille, C. Appl. EnViron. Microbiol. 2005,
71, 6651-6656.
(13) Bratton, M. R.; Pressler, M. A.; Hosler, J. P. Biochemistry 1999, 38, 16236-
16245.
9
13218 J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006

Structure−Function Relationships of DHNA
A R T I C L E S
Figure 4. ESI-MS analysis of the mixtures generated by a 10 min reaction
of SaDHNA (A) and SaY54F (B). The substrate 1 is observed at m/z 256
in both spectra. The region from m/z 180-300 has been magnified (×5)
for clarity.
spectrum obtained from the standard (compare parts A and B
of Figure S3). The most abundant product ion in both parts A
and B of Figure S3 at m/z 178 is most likely due to the loss of
water from the precursor ion. To obtain further confirmation of
the expected product ion structure, MS3 was then performed
on the m/z 178 product ion (Figure S3C,D). As expected, the
experimental and standard MS3 spectra were essentially identi-
cal. The most abundant product ions at m/z 161 and 150 are
most likely due to the loss of NH3 and CO, respectively. Based
on an MS4 experiment (data not shown), the MS3 product ion
seen at m/z 179 was formed via an ion-molecule reaction
between the ion at m/z 161 and a water molecule. Taken
together, the MS/MS and MS3 spectra provided strong evidence
that the product seen at m/z 196 from the reaction of 1 with
SaDHNA is 4.
This same type of analysis was also performed in order to
determine the identity of the m/z 182 ion in Figure 4B. MS/MS
and MS3 spectra of the ion were obtained and compared to those
of the proposed reaction product, 6. An examination of Figure
S4 indicated that the major ion observed by CID MS/MS of
the ion at m/z 182 obtained from the SaY54F-catalyzed reaction
is identical to the major ion seen in the MS/MS product ion
spectrum obtained from the standard (compare parts A and B
of Figure S4). The m/z 154 product ion in both parts A and B
of Figure S4 is most likely due to the loss of CO from the
precursor ion. MS3 was then performed on the m/z 154 ion
(Figure S4C,D). Again, the experimental and standard spectra
were essentially the same. The major MS3 product ion, seen at
m/z 126, corresponds to the loss of CO, while the second most
intense ion, seen at m/z 137, corresponds to the loss of NH3.
As was the case for the previous analysis of the SaDHNA-
catalyzed reaction, these spectra together provided strong
Figure 3. NMR analysis of the reactions catalyzed by SaDHNA (A) and
SaY54F (B). The initial NMR sample contained 1 mM 1 and 1 mM TCEP
in 100 mM sodium phosphate buffer, pH 8.3, made with 90% H2O and
10% D2O. The reaction was initiated with 1 µM SaDHNA (A) or
3 µM SaY54F (B). Peak identities are as follows: a, 7H and 1′H of 1; b,
7H and the 6-hydroxymethyl group of 4; c, 7H of 6; d, from 3 and TCEP;
e, 2′H of 5; f, 2′H of 1; g, 3′Ha of 1; h, 3′Ha of 5; i, 3′Hb of 1 and 5; and
j, from 3.
Table 2. Chemical Shifts of Selected Protons of Compounds
Related to the Reactions Catalyzed by the Wild-Type and Mutant
DHNAsa
compds
7H
1
′
H
2
′
H
3
′
Ha
3′
Hb
other
1
5
4
6
3
4.22 (d) 4.22 (d) 3.90 (m) 3.80 (m) 3.66 (m)
4.22 (d) 4.27 (d) 3.94 (m) 3.72 (m) 3.66 (m)
4.18 (s)
4.12 (s)
4.18 (s)
3.50 (d)
4.02 (m)
3 + TCEP
a The chemical shifts were from those of the commercial compounds in
50 mM sodium phosphate buffer, pH 8.3 (pH meter reading without
correction for deuterium isotope effects), made with 90% H2O and 10%
D2O. The peak at 4.02 ppm was present only in the presence of both 3 and
TCEP. The 4.18 ppm peak of 4 also belongs to the 6-hydroxymethyl group.
m/z 196, while the spectrum of the reaction involving SaY54F
showed the formation of a product ion at m/z 182.
To determine the identity of the m/z 196 ion in Figure 4A,
MS/MS and MS3 spectra were obtained and compared to those
of the expected reaction product 4. As shown in Figure S3, the
major ions observed by CID MS/MS of the ion at m/z 196
obtained from the SaDHNA-catalyzed reaction were identical
to the major product ions seen in the MS/MS product ion
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13219

A R T I C L E S
Wang et al.
evidence that the product seen at m/z 182 from the reaction of
1 with SaY54F is 6.
The m/z 256 ion (Figure 4) was found to correspond to 1
(Figure S5), as well as probably partly due to 5, which was
generated by the enzymatic reactions. 1 and 5 differ only in
the stereochemistry of their 2′-carbons and have the same
molecular weight. The m/z 194 ion (Figure 4) was found to
correspond to 10 and 7, because the MS/MS and MS3 spectra
of the m/z 194 species were the sums of those of 10 and 7
standards, with 10 contributing more than 7 (Figure S6). The
existence of 7 was also confirmed by spectrophotometry, in
which 7 has a characteristic absorption peak at 420 nm (data
not shown).
GC/MS Confirmation of 3. Since the major product of the
SaY54F-catalyzed reaction was 6, not 4, it was necessary to
confirm the identity of 3 as indicated by the NMR analysis
(Figure 3). To this end, we derivatized the reaction products of
SaDHNA and SaY54F with methoxyamine and MTBSTFA to
form the tert-butyldimethylsilyl ether of the methoxime. 3 was
then detected by GC/MS as shown in Figure S7. The enzymatic
reaction product (Figure S7B) was identified by comparison with
the commercially available standard compound (Figure S7A).
The retention time of the standard is 12.3 min (data not shown).
The molecular weight of its derivative is 203 Da, but the
molecular ion was too weak to be observed, as is typical for
these derivatives. The dominant fragment ions in the mass
spectrum appeared at m/z 146 ([M - C4H9]+) and 188 ([M -
CH3]+). The two reaction mixtures generated by SaDHNA and
SaY54F both contained a peak with the same retention time as
that of the standard 3. The mass spectrum of the derivative (the
peak at 12.3 min) from the SaY54F-catalyzed reaction (Figure
S7B), which was identical to that of the SaDHNA-catalyzed
reaction (data not shown), also showed two dominant ions at
m/z 146 and 188. This result and that of the NMR analysis
clearly indicated that the SaY54F-catalyzed reaction generates
3, as does the SaDHNA-catalyzed reaction.
Figure 5. ESI-MS time course analysis of the DHNP reactions catalyzed
by SaDHNA (A) and SaY54F (B). (b, m/z 256; 9, m/z 194; 2, m/z 196;
[, m/z 182).
catalyzed reaction generates formic acid but the wild-type
enzyme-catalyzed reaction does not.
Is 6 Derived from 4? Since the C1′-C2′ bond is cleaved
first in the generation of DXHP, which is the same for the
normal aldolase reaction for the formation of 4, it is logical to
ask whether 6 is derived from 4. To address this question, we
analyzed the time course of the DHNP reaction by MS. The
results are shown in Figure 5. The ions at m/z 256, 196, and
182 were due to 1, 4, and 6, respectively, as described earlier.
The m/z 194 ions were due to a large part from 10 and to a
small part from 7. The results showed again that the major
product of the wild-type enzyme-catalyzed reaction was 4 as
expected, but the major product of the mutant-catalyzed reaction
was 6. Furthermore, there was no accumulation of 4 in the
mutant-catalyzed reaction, suggesting that 6 was not derived
from 4, unless the conversion of 4 to 6 was much faster than
the formation of 4. To confirm that 6 was not formed from the
rapid conversion of 4, 4 was incubated with SaY54F and
analyzed by MS. If 6 was derived from the rapid conversion of
4, then over time these spectra should show a decrease in the
intensity of m/z 196 and an increase in the intensity of m/z 182.
There was not, however, any production of the ion at m/z 182
seen in these spectra (not shown), indicating that DXHP was
not derived from 4.
The Source of Oxygen. Finally, we considered the source
of the new oxygen at the C6 position of 6. Was it from water
or from molecular oxygen dissolved in the buffer? To address
this issue, we ran the DHNP reaction in buffer prepared with
18O-water and analyzed the reaction mixture with MS. With
18O-water, if the oxygen was from the solvent, the mass
spectrum should show a shift in the m/z of the protonated
product ion from 182 to 184. The MS spectra of the reaction
mixtures obtained with 18O-water as the solvent, however, were
the same as the spectra obtained when the reaction was run in
unlabeled solvent (data not shown), indicating that there was
HPLC Identification of Formic Acid. Because 6 contains
one carbon less than 4, the SaY54F-catalyzed reaction must
generate a one-carbon species, most likely formic acid. To
identify the one-carbon species, we analyzed the 7,8-dihydro-
D-neopterin (DHNP) reaction mixtures generated by SaDHNA
and SaY54F by HPLC using an Aminex HPX-87H column. The
unknown compounds in the reaction mixtures were identified
by comparing their retention times with those of the standard
compounds, as shown in Figure S8. This HPLC run of standard
compounds (blue line) included succinic acid at 13.5 min, formic
acid at 14.9 min, and acetic acid at 16 min. It also showed a
peak at 17.5 min, probably due to the HEPES buffer, which
also appeared in all chromatograms, including those of the
SaY54F solution (green line), the reaction mixture generated
by SaY54F (red line), the reaction mixture generated by
SaDHNA (cyan line), and the buffer alone (data not shown).
The chromatogram of the reaction mixture generated by SaY54F
showed an intense peak at 14.9 min, which was the same as
the retention time for standard formic acid. The small shoulder
at the left side was probably due to the protein. The intense
peak in the chromatogram of the reaction mixture generated by
SaY54F was absent in the chromatogram of the reaction mixture
generated by SaDHNA. The result indicated that the mutant-
9
13220 J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006

Structure−Function Relationships of DHNA
A R T I C L E S
in the aldolase reaction? The fact that either mutant-catalyzed
reaction generates GA, and furthermore that about equal moles
of GA and 6 are generated (Figure 5), strongly indicates that 6
is generated via the cleavage of the C1′-C2′ bond. The
formation of GA was confirmed by the derivatization of the
reaction mixtures followed by GC/MS analysis (Figure S7).
Since the first step in the generation of 6 is the cleavage of the
C1′-C2′ bond, as in the wild-type enzyme-catalyzed reaction,
the next question is whether 6 is derived from 4. The MS
analysis of the time course of the mutant-catalyzed reaction
shows that there is no accumulation of 4 in the reaction (Figure
5). This indicates that 6 is not derived from 4, unless the
conversion of 4 to 6 is very rapid relative to the generation of
4. This was addressed by mixing standard 4 with the mutant
and following the possible reaction by MS. The result indicates
that 4 cannot be converted to 6 under the experimental
conditions. Since the cleavage of the C1′-C2′ bond generates
an intermediate with one carbon more than 6, the conversion
of the intermediate to 6 must generate a one-carbon species,
which was identified as formic acid by comparing the HPLC
chromatograms of the reaction mixtures generated by the mutant
and wild-type enzymes with those of standard organic acids
and aldehydes (Figure S8). The remaining question is where
the oxygen atom at the C6 position of 6 comes from. Is it from
water or from oxygen molecules dissolved in the buffer? This
issue was addressed by the 18O-water experiment and the oxygen
consumption assay. The 18O-water experiment eliminates water
as the source for the oxygen newly attached to the pterin ring.
The oxygen consumption assay indicates that molecular oxygen
is the source for the oxygenation reaction, because significant
oxygen consumption occurs only in the mutant-catalyzed
reaction (Figure 6) and the turnover number calculated from
the oxygen consumption data is comparable to that for the
formation of 6 calculated from the NMR data.
The above analysis leads us to propose a chemical mechanism
for the mutant-catalyzed reaction, as depicted in Figure 7. The
initial step in the formation of 6 is the generation of the same
enol intermediate as in the wild-type enzyme-catalyzed reaction,
which involves the cleavage of the C1′-C2′ bond and yields
approximately equal moles of GA and 6, as described earlier.
Deprotonation of N5 of the enol intermediate generates a
carbanion species, which donates a single electron to molecular
oxygen to form a caged radical pair. The caged radicals react
to generate the peroxide ion, which leads to the formation of 6
and formic acid. The participation of molecular oxygen in the
mutant-catalyzed reaction has been confirmed by the oxygen
consumption assay. The reaction path for the formation of 6
from the enol intermediate is speculative, but the formation of
formic acid has been confirmed by HPLC. The proposed
mechanism also accounts for the formation of the minor product
7. Alternative mechanisms involve protein radicals (in contrast
to just substrate radicals). Both mechanisms have been proposed
for cofactor-independent oxygenases,14 but the proposed mech-
anism or variations thereof that involve mainly general acid and
base catalysis are more likely, particularly considering that
DHNA is evolved for general acid and base catalysis.
Figure 6. Oxygen consumption by the reactions catalyzed by SaDHNA
and SaY54F. The initial samples contained 200 µM 1 in 100 mM HEPES-
KOH, pH 8.2, and the reactions were initiated by 8 µM SaDHNA or SaY54F
as marked by the vertical lines.
no 18O incorporation in any of the products and the oxygen
incorporated in the product 6 does not come from water. We
then measured the oxygen consumption of the reactions
catalyzed by SaY54F and SaDHNA. The results are shown in
Figure 6. The amounts of wild-type and mutant enzymes were
both 8 µM. Under the conditions, nearly all of substrate 1 was
converted to products. The results showed that only the mutant-
catalyzed reaction consumes a significant amount of oxygen,
indicating that SaY54F is an oxygenase and the source of
oxygen for the oxygenation reaction is molecular oxygen
dissolved in the buffer. We calculated the turnover number for
the oxygenation reaction from the linear portion of the oxygen
consumption curve, which was 0.027 s-1, in close agreement
with the turnover number for the formation of 6 calculated from
the NMR time course of the SaY54F-catalyzed reaction (0.018
s-1), indicating that one oxygen molecule is used in the
formation of one molecule of 6. The extra oxygen consumption
was probably due to the formation of 7 and 10.
Discussion
SaY54F and EcY53F Are Oxygenases. By a variety of
means, we have shown that the major product of the mutant
(either SaY54F or EcY53F)-catalyzed reaction is 6, rather than
4 as in the wild-type enzyme-catalyzed reaction. The turnover
number for the formation of 6 is >50-fold that for the formation
of the normal product 4 for SaY54F. The initial clue to the
surprising properties of the mutant enzymes was a dramatic drop
of the fluorescence intensities of the reaction mixtures. The
NMR analysis of the reaction mixtures revealed that the mutant
1
enzymes generate a new compound with a H NMR signal at
4.12 ppm and little of the normal product 4 (Figure 3). The
new compound was also a major product in the MS analysis of
the reaction mixtures generated by the mutant SaY54F (Figure
4) and was identified to be 6 by comparing its MS/MS and
MS3 spectra with those of its standard (Figure S4). This raises
the immediate question of how 6 is generated. Is 6 generated
via the cleavage of the C6-C1′ bond or the C1′-C2′ bond as
The formation of 6 is enzyme-catalyzed and is not due to
the spontaneous oxidation of the dissociated reaction intermedi-
ate 2. The dissociation rate constants of substrate and product
(14) Fetzner, S. Appl. Microbiol. Biotechnol. 2002, 60, 243-257.
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13221

A R T I C L E S
Wang et al.
enzymes seem to evolve to utilize carbanion reactivity toward
dioxygen for the oxygenation reactions,14 which might be a
general catalytic strategy for the family of cofactor-independent
oxygenases. SaY54F and EcY54F are excellent examples for
the evolution of such a cofactor-independent oxygenase. Inter-
estingly, both DHNA4,7,20,21 and urate oxidase22,23 have a tunnel-
like structure called a “T-fold” with active sites located between
subunits.24 The bound ligands are stacked with a conserved
phenylalanine residue in urate oxidase and the conserved
tyrosine in DHNA that was substituted with a phenylalanine
residue in this study. Thus, individual folds are used by nature
to catalyze many different kinds of reactions, as there might be
no more than 1000 unique protein folds. Hence, it is possible
to convert one enzyme to another by protein engineering. A
recent example is the protein engineering of cycloartenol
synthase, in which two site-directed mutations (H477N and
I481V) convert the enzyme to lanosterol synthase.25 In the
present work, a single tyrosine-to-phenylalanine mutation
converts DHNA to an oxygenase. In both cases, the protein folds
are most likely preserved.
The Role of the Conserved Tyrosine Residue. The bio-
chemical analysis of the two mutants SaY54F and EcY53F
indicates that the hydroxyl group of the phenol ring of the
conserved tyrosine residue plays only a minor role in the
physical steps of the enzymatic reaction. In contrast, the
hydroxyl group plays a critical role in catalysis. The hallmark
of the DHNA-catalyzed reaction is general acid and base
catalysis. The general acid and base involved in the protonation
and deprotonation of N5 are most likely a water molecule and
its conjugated base, based on the published crystal structures7
and our own unpublished crystal structures. The deprotonation
of the 2′-hydroxyl group is most likely the function of the
conserved lysine residue, K100 in SaDHNA and K98 in
EcDHNA. The conserved tyrosine residue may play a minor
role if any in this step, because the formation of the enol
intermediate is not impaired by the mutations to any great extent,
as evidenced by the high kcat for the formation of 6 by the mutant
SaY54F, which is 1/10 of the kcat value for the formation of 4
by the wild-type enzyme. On the other hand, the major product
of the mutant-catalyzed reaction is 6, not 4, suggesting that the
protonation of the enol intermediate to generate 4 is greatly
impaired and the conserved tyrosine residue may play a critical
role in this step. The high oxygenase activities of the mutants
are likely due to the tendency of the enol intermediate to react
with molecular oxygen and the available general acid and base
for the generation of the enol intermediate and its subsequent
oxygenation, as illustrated in Figure 7. The conserved tyrosine
residue is located at the bottom of the active site. The active
site is likely to be accessible to molecular oxygen, even in the
wild-type enzyme. The wild-type enzyme is an efficient aldolase
Figure 7. Proposed chemical mechanism for the generation of 6 and 7 by
SaY54F and EcY53F. For simplicity, many steps are indicated by single
arrows, irrespective of their reversibility.
analogues for SaY54F are lower than those for SaDHNA.
Therefore, the dissociation rate constant for the reaction
intermediate is most likely higher for the wild-type enzyme than
for the mutant enzyme, although its value is not known for either
the wild-type or the mutant enzyme. On the other hand, the
rate for the formation of 6 is ∼3% of that for the formation of
4 in the wild-type enzyme-catalyzed reaction. The relative rates
are reversed in the mutant enzyme-catalyzed reaction, where
the rate for the formation of 4 is ∼2% of that for the formation
of 7,8-dihydroxanthopterin (DHXP). The dissociated intermedi-
ate 2 most likely tautomerizes to form 4 rather than reacting
with oxygen to form 6.
The key step in the formation of 6 from the enol intermediate
2 is the generation of the carbanion species, which involves
the deprotonation of N5 as in the wild-type enzyme-catalyzed
reaction and the block of the protonation of the hydroxymeth-
ylene group by the mutation of the active-site tyrosine residue.
Carbanions are prone to react with oxygen, and several
enzymatic reactions involving carbanion intermediates indeed
have oxygen-consuming side reactions.15,16 The formation of
the peroxide ion is reminiscent of the initial steps of the reactions
catalyzed by 1H-3-hydroxy-4-oxoquinaldine 2,4-dioxygenase17
and urate oxidase,18,19 neither of which has any cofactor. These
(19) Imhoff, R. D.; Power, N. P.; Borrok, M. J.; Tipton, P. A. Biochemistry
2003, 42, 4094-4100.
(20) Bauer, S.; Schott, A. K.; Illarionova, V.; Bacher, A.; Huber, R.; Fischer,
M. J. Mol. Biol. 2004, 339, 967-979.
(21) Goulding, C. W.; Apostol, M. I.; Sawaya, M. R.; Phillips, M.; Parseghian,
A.; Eisenberg, D. J. Mol. Biol. 2005, 349, 61-72.
(22) Colloc’h, N.; ElHajji, M.; Bachet, B.; Lhermite, G.; Schiltz, M.; Prange,
T.; Castro, B.; Mornon, J. P. Nat. Struct. Biol. 1997, 4, 947-952.
(23) Retailleau, P.; Colloc’h, N.; Vivares, D.; Bonnete, F.; Castro, B.; El Hajji,
M.; Mornon, J. P.; Monard, G.; Prange, T. Acta Crystallogr. Sect. D: Biol.
Crystallogr. 2004, 60, 453-462.
(15) Abell, L. M.; Schloss, J. V. Biochemistry 1991, 30, 7883-7887.
(16) Hixon, M.; Sinerius, G.; Schneider, A.; Walter, C.; Fessner, W. D.; Schloss,
J. V. FEBS Lett. 1996, 392, 281-284.
(17) Frerichs-Deeken, U.; Ranguelova, K.; Kappl, R.; Huttermann, J.; Fetzner,
S. Biochemistry 2004, 43, 14485-14499.
(24) Colloc’h, N.; Poupon, A.; Mornon, J. P. Proteins 2000, 39, 142-154.
(25) Lodeiro, S.; Schulz-Gasch, T.; Matsuda, S. P. T. J. Am. Chem. Soc. 2005,
127, 14132-14133.
(18) Sarma, A. D.; Tipton, P. A. J. Am. Chem. Soc. 2000, 122, 11252-11253.
9
13222 J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006

Structure−Function Relationships of DHNA
A R T I C L E S
rather than an oxygenase, because it has a general acid that can
efficiently protonate the enol intermediate to form 4.
Robert P. Hausinger, J. Gregory Zeikus, and Kathleen A. Gallo
for discussions and suggestions.
Acknowledgment. This work was supported in part by NIH
grant GM51901 (H.Y.) and Michigan State University Intra-
mural Research Grant 04-IRGP-117 (G.E.R.). This study made
use of a Varian INOVA-600 NMR spectrometer at Michigan
State University, funded in part by NSF Grant BIR9512253.
We are grateful to Dr. Yue Li for assistance in the site-directed
mutagenesis experiments, Dr. Denise Mills for assistance in the
oxygen consumption assay, Mr. James B. McKinlay for as-
sistance in the HPLC identification of formic acid, the MSU
Mass Spectrometry Facility for GC/MS analysis, and Drs.
Supporting Information Available: Figures S1-S8, showing
the binding of binding of 10 to EcY53F at equilibrium, the
stopped-flow analysis of the binding of 9 to EcY53F, the ESI-
MS/MS identification of 1, 4, 6, 7, and 10, the GC/MS
identification of 3, and the HPLC identification of formic acid;
complete ref 4. This material is available free of charge via the
Internet at http://pubs.acs.org.
JA063455I
9
J. AM. CHEM. SOC. VOL. 128, NO. 40, 2006 13223