Studies on the substrate specificity of Escherichia coli galactokinase

Article abstract of DOI:10.1021/ol034642d

(Martix presented) In vitro glycorandomization (IVG) technology is dependent upon the ability to rapidly synthesize sugar phosphates. Compared with chemical synthesis, enzymatic (kinase) routes to sugar phosphates would be attractive for this application. This work focuses upon the development of a high-throughput colorimetric galactokinase (GalK) assay and its application toward probing the substrate specificity and kinetic parameters of Escherichia coli GalK. The demonstrated dinitrosalicylic assay should also be generally applicable to a variety of sugar-processing enzymes.

Full text of DOI:10.1021/ol034642d

ORGANIC
LETTERS
2003
Vol. 5, No. 13
2223-2226
Studies on the Substrate Specificity of
Escherichia coli Galactokinase
Jie Yang,† Xun Fu,† Qiang Jia,‡ Jie Shen,‡ John B. Biggins,† Jiqing Jiang,†
,†
Jingjing Zhao,† Joshua J. Schmidt,† Peng G. Wang,‡ and Jon S. Thorson*
Laboratory for Biosynthetic Chemistry, Pharmaceutical Sciences DiVision,
School of Pharmacy, UniVersity of Wisconsin-Madison, 777 Highland AVenue,
Madison, Wisconsin 53705, and Department of Chemistry, Wayne State UniVersity,
Detroit, Michigan 48202
jsthorson@pharmacy.wisc.edu
Received April 14, 2003
ABSTRACT
In vitro glycorandomization (IVG) technology is dependent upon the ability to rapidly synthesize sugar phosphates. Compared with chemical
synthesis, enzymatic (kinase) routes to sugar phosphates would be attractive for this application. This work focuses upon the development
of a high-throughput colorimetric galactokinase (GalK) assay and its application toward probing the substrate specificity and kinetic parameters
of Escherichia coli GalK. The demonstrated dinitrosalicylic assay should also be generally applicable to a variety of sugar-processing enzymes.
Glycosylated bacterial metabolites represent a main class of
pharmaceutically important natural products.1 In many cases,
the sugar constituents of these molecules are critical for their
mode of action and modification of these carbohydrates can
lead to beneficial affects on drug targeting, action, and/or
pharmacology. While there are a number of routes for
altering glycosylation, one of the most promising for complex
metabolites is in vitro glycorandomization (IVG).2 This
method takes advantage of the limitless flexibility of the
chemical synthesis of unique sugar precursors with the
inherent or engineered substrate promiscuity of enzymes to
activate (nucleotidylyltransferase, “Ep”) and attach (glyco-
syltransferases, “GlyT”) these precursors to various natural
product aglycons (Scheme 1). This method has recently been
applied toward the generation of novel nonribosomal peptides
and aminocoumarin antibiotics.2e,3
Given the key role sugar phosphates play in the glyco-
randomization process, the rapid synthesis of sugar phosphate
libraries would directly contribute to the efficiency of IVG.
Many elegant synthetic routes to sugar phosphates already
exist, but a tedious process is often unavoidable.2a-c The
overall yield can be low due to multistep synthetic manipula-
tions, and the purification and/or stability of sugar phosphates
are also limiting factors. Thus, an enzymatic (kinase) route
would be an attractive alternative, as it is a one-step process
and could be coupled, in a single reaction vessel, to IVG.
† University of Wisconsin-Madison.
‡ Wayne State University.
(1) (a) Potier, P. Actual. Chim. 1999, 11, 9. (b) Weymouth-Wilson, A.
C. Nat. Prod. Rep. 1997, 14, 99. (c) Vogt, T.; Jones, P. Trends Plant Sci.
2000, 5, 380. (d) Thorson, J. S.; Hosted, T. J., Jr.; Jiang, J.; Biggins, J. B.;
Ahlert, J. Curr. Org. Chem. 2001, 5, 139. (e) Kren, V.; Martinkova, L.
Curr. Med. Chem. 2001, 8, 1303. (f) Jones, P.; Vogt, T. Planta 2001, 213,
164. (g) Keegstra, K.; Raikhel, N. Curr. Opin. Plant Biol. 2001, 4, 219. (h)
Thorson, J. S.; Vogt, T. Glycosylated Natural Products in Carbohydrate-
based Drug DiscoVery; Wong, C.-W., Ed.; Wiley-VCH: Weinheim, 2003.
(2) (a) Jiang, J.; Biggins, J. B.; Thorson, J. S. Angew. Chem., Int. Ed.
2001, 40, 1502. (b) Jiang, J.; Biggins, J. B.; Thorson, J. S. J. Am. Chem.
Soc. 2000, 122, 6803. (c) Barton, W. A.; Biggins, J. B.; Jiang, J.; Thorson,
J. S.; Nikolov, D. B. Proc. Natl. Acad. Sci. 2002, 99, 13397. (d) Barton,
W. A.; Biggins, J. B.; Lesniak, J.; Jeffrey, P. D.; Jiang, J.; Rajashankar, K.
R.; Thorson, J. S.; Nikolov, D. B. Nature Struct. Biol. 2001, 8, 545. (e)
Thorson, J. S.; Barton, W. A.; Hoffmeister, D.; Albermann, C.; Nikolov,
D. B. ChemBioChem 2003, manuscript in press. (f) Jiang, J.; Albermann,
C.; Thorson, J. S. ChemBioChem. 2003, manuscript in press.
(3) (a) Solenberg, P. J.; Matsushima, P.; Stack, D. R.; Wilkie, S. C.;
Thompson, R. C.; Baltz, R. H. Chem. Biol. 1997, 4, 195. (b) Losey, H.;
Jiang, J.; Biggins, J. B.; Oberthur, M.; Ye, X.-Y.; Dong, S. D.; Kahne, D.;
Thorson, J. S.; Walsh, C. T. Chem. Biol. 2002, 9, 1305. (c) Albermann, C.;
Fu, X.; Jiang, J.; Thorson, J. S. Unpublished. (d) Albermann, C.; Soriano,
A.; Jiang, J.; Vollmer, H.; Biggins, J. B.; Barton, W. A.; Lesniak, J.; Nikolov,
D. B.; Thorson, J. S. Org. Lett. 2003, 5, 933.
10.1021/ol034642d CCC: $25.00 © 2003 American Chemical Society
Published on Web 06/06/2003

Scheme 1. Schematic Pathway for in Vitro
Scheme 2. Schematic Reaction of Dinitrosalicylic (DNS) Acid
Glycorandomization
Reagent and a Reducing Sugar
N-His6-GalK fusion was used for directly for all studies
presented.7 Our colorimetric assay for GalK is based upon
an oxidation-reduction reaction between DNS acid (oxidant)
and a reducing sugar (Scheme 2). The reduced product
3-amino-5-nitrosalicylate has a red-brown color, which can
be quantified by the absorbance at 575 nm (ꢀ575 ) 758 M-1
cm-1).
One such kinase, galactokinse (GalK), catalyzes the forma-
tion of R-D-galactose-1-phosphate (Gal-1-P) from galactose
and ATP. Previous studies have focused upon purification
and characterization of GalK from various sources.4 Limited
substrate specificity studies reveal that galactose, 2-deoxy-
galactose, galactosamine, ATP, 2′-dATP, and 3′-dATP are
possible substrates for certain members of this class.
In an effort to expand upon these studies, herein we
describe a novel high-throughput assay based on dinitrosali-
cylic (DNS) acid,5 and the application of this assay toward
the determination of the previously undetermined substrate
specificity and kinetic parameters of GalK from Escherichia
coli. This work reveals that native E. coli GalK has a limited
substrate scope but clearly provides a foundation from which
to launch the pursuit of directed evolution experiments
designed to enhance the promiscuity of GalK.
To test the utility of this assay for GalK, galactose (8 mM)
and ATP (10 mM) were incubated with GalK at 30 °C.8 At
different time points, the free sugar/sugar-phosphate ratio
was assessed by both TLC9 and the DNS assay.10 As a
control, assays without GalK gave no change in absorbance
at 575 nm, while assays without galactose gave no absor-
bance at 575 nm. Moreover, DNS assays containing solely
the GalK product galactose-1-phosphate lacked absorbance
at 575 nm, suggesting that product sugar phosphates should
not affect the DNS assay. As confirmation, the product of
The galK gene from E. coli K-12 was subcloned into pET-
(7) GalK-His6 was purified via a cobalt affinity column at 4 °C. The
crude cell extract, in sodium phosphate buffer, was sonicated 5 × 45 s on
ice. The solution was passed through a cobalt resin column (50 mM CoCl2).
The bound recombinant proteins were eluted by elution buffer (200 mM
imidazole in sodium phosphate buffer). The active fractions were concen-
trated and further resolved via gel filtration chromatography. The fractions
containing homogeneous GalK were collected and concentrated, and the
final concentration of protein was determined by Bio-Rad protein assay.
(8) Sugar solution (0.5 mL, 8 mM), ATP (10 mM), and MgCl2 (5 mM)
in 50 mM sodium phosphate buffer (pH ) 7.5) was incubated at 30 °C for
5 min, and then GalK (final concentration is 5.6 µM) was added to the
reaction system. At various time points, aliquots were analyzed via
chromatography (ref 9) and the DNS assay (ref 10).
(9) Reaction was monitored by TLC using 10 mM tetrabutylammonium
hydroxide in 80% aqueous acetonitrile (anisaldehyde stain). Rf values of
galactose, ATP, and galactose-1-phosphate were 0.51, 0.12, and 0.28,
respectively. LC-MS analysis was also consistent with the formation of
Gal-1-P.
(10) During the reaction process, 50 µL of reaction solution was taken
out periodically and transferred to 100 µL of DNS reagent solution (45mM)
to assess the reaction progress. Each DNS reaction solution was placed in
a bath of boiling water for 15 min and subsequently cooled to room
temperature followed by the addition of 15 µL of a 40% potassium sodium
tartrate solution to stabilize the color. The absorbance at 575 nm was
recorded, and a plot of change in absorbance at 575 nm as a function of
time was obtained.
15b vector and overexpressed in BL21 (DE3).6 The purified
(4) (a) Lavine, J. E.; Cantlay, E.; Roberts, C. T., Jr.; Morse, D. E.
Biochim. Biophys. Acta 1982, 717 (1), 76. (b) Dey, P. M. Eur. J. Biochem.
1983, 136 (1), 155. (c) Foglietti, M. J.; Percheron, F. Biochimie 1976, 58,
8 (5), 499. (d) Blume, K.; Beutler, E. J. Biol. Chem. 1971, 246, 6507. (e)
Schell, M. A.; Wilson, D. B. J. Biol. Chem. 1977, 252, 1162. (f) Srivastava,
S. K.; Blume, K.; Loon C. V.; Beutler, E. Arch. Biochem. Biophys. 1972,
150, 191. (g) Vorgias, C. E.; Lemaire, H. G.; Wilson, K. S. Protein Expr.
Purif. 1991, 2 (5-6), 330. (h) Ballard, F. J. Methods Enzymol. 1975, 42,
43. (i) Timson, D. J.; Reece, R. J. Biochimie 2002, 84, 265. (j) Howard, S.
M.; Heinrich, M. R. Arch. Biochem. Biophys. 1965, 110, 395. (k) Thomas,
P.; Bessell, E. M.; Westwood, J. H. Biochem. J. 1974, 139, 661.
(5) (a) Miller, G. L. Anal. Chem. 1959, 31, 426. (b) Sumner, J. B. J.
Biol. Chem. 1921, 47, 5. (c) Sumner, J. B.; Sisler, E. B. Arch. Biochem.
1944, 4, 333.
(6) GalK gene was cloned from E. coli K-12 and inserted into pET-15b
downstream of His-6 coding sequence after digestion with NdeI and BamHI.
pGalK/pET15b-BL21 (DE3) was grown on LBAmp100 plate overnight at 37
°C. A single colony was transferred into 100 mL of LBAmp100 medium and
grown at 37 °C overnight, which was used to inoculate 3 L of superbroth
and incubated at 37 °C with shaking. At OD600 ) 0.6, 1 M IPTG was added
(1 mM final concentration) and the culture was incubated overnight. The
cells were harvested by centrifugation and washed twice with sodium
phosphate buffer.
2224
Org. Lett., Vol. 5, No. 13, 2003

Figure 2. Progress curve of the decrease in free sugar substrate
concentration as a function of reaction time.
4-deoxy-D-galactose, 9; 6-azido-6-deoxy-D-galactose, 12; and
6-amino-6-deoxy-D-galactose, 13) were synthesized, while
the other six (2-deoxy-D-galactose, 2; 2-amino-deoxy-D-
galactose, 3; 2-N-acetamido-2-deoxy-D-galactose, 4; D-gu-
lose, 7; D-glucose, 10; 6-deoxy-D-galactose, 11) were com-
mercially available. Of the synthetically derived analogues,
5 and 8 were prepared as previously described,2b,13j-l while
the routes to 6, 9, 12, and 13 are illustrated in Scheme 3.13
The DNS assay was then used to monitor the reaction
progress.8,10 As a control, assays without GalK and assays
without sugars were also analyzed for each substrate. We
found the DNS assay to be very sensitive to the change in
sugar concentration, amenable to a multiwell plate format,
and, as a first approximation, easily discernible by the naked
eye (Figure 3). This assay is also advantageous as it directly
Figure 1. Potential GalK substrates tested in this study.
the GalK reaction coelutes on TLC or HPLC with Gal-1-P
and was also confirmed via LC-MS. Use of the DNS assay
allowed a progress curve of the change in sugar concentration
as a function of time to be obtained by converting the
absorbance to the corresponding sugar concentration by using
standard curves (Figure 2).11 The initial velocity was
determined by the slope value of the linear phase in the
progress curve, and the kinetic parameters for galactose and
ATP were determined (Vmax ) 1.65 µmol min-1 mg-1, and
Km for galactose and ATP were calculated to be 2.1 and 2.5
mM, respectively).12 The results are comparable with previ-
ously published kinetic parameters for GalK from different
species (0.5-1.5 mM for Km of galactose and 0.15-5 mM
for Km of ATP).4
(13) (a) 3-Methyl-R-D-galactose (6). 1H NMR (CD3OD): δ 4.56-4.48
and 4.14-4.08 (m, 1H), 3.84-3.34 (m, 5H), 1.40-1.14 (m, 3H). 13C NMR
(CD3OD): δ 96.2, 74.9, 74.4, 74.0, 73.7, 73.7, 17.8. ESI-MS: calcd for
C7H14O6Na, 217.1; found, m/z 217.1 [M + Na]+. (b) 4-Azido-4-deoxy-R-
1
D-galactose (9). H NMR (D2O): δ 4.57 (d, J ) 7.2 Hz, 1H), 3.49 (dd, J
) 7.2, 9.8 Hz, 1H), 3.93 (dd, J ) 9.8, 3.9 Hz, 1H), 4.00 (d, J ) 3.9 Hz,
1H), 3.74 (m, 1H), 3.76 (m, 2H). 13C NMR (D2O): δ 96.738, 72.092,
73.459, 62.935, 73.765, 61.239. MS: calcd for C6H11N3O5, 205.1; found,
To test the utility of this assay toward other putative
substrates, a sugar library, including galactose (1) and other
monosaccharide analogues, was tested (Figure 1). Six of the
alternative free sugars examined (3-deoxy-D-galactose, 5;
3-C-methyl-D-galactose, 6; 4-deoxy-D-galactose, 8; 4-azido-
1
m/z 203.9 [M - H]-. (c) 6-Azido-6-deoxy-R-D-galactose (12). H NMR
(D2O): δ 5.28 (d, J ) 3.7 Hz, 1H), 3.81 (dd, J ) 3.7, 10.3 Hz, 1H), 3.87
(dd, J ) 10.3, 3.2 Hz, 1H), 3.97 (dd, J ) 3.2, 0.8 Hz, 1H), 4.21 (m, 1H),
3.52 (m, 2H). 13C NMR (D2O): δ 92.609, 68.421, 69.266, 69.879, 69.166,
51.149. MS: calcd for C6H11N3O5, 205.1; found, m/z 206 [M + H]+. (d)
6-Amino-6-deoxy-R-D-galactose (13). 1H NMR (D2O): δ 5.30 (d, J )
3.5 Hz, 1H), 3.81 (dd, J ) 3.5, 10.3 Hz, 1H), 3.87 (dd, J ) 10.3, 3.2 Hz,
1H), 3.99 (d, J ) 3.2 Hz, 1H), 4.27 (m, 1H), 3.35 (m, 1H), 3.26 (m, 1H).
13C NMR (D2O): δ 92.580, 68.274, 69.131, 70.064, 66.445, 40.579. MS:
calcd for C6H13NO5, 179.1; found, m/z 180.0 [M + H]+. (e) May, J. A.,
Jr.; Sartorelli, A. C. J. Med. Chem. 1979, 22, 2(8), 971. (f) Cappi, M. W.;
Moree, W. J.; Qiao, L.; Marron, T. G.; Weitz-Schmidt, G.; Wong, C.-H.
Bioorg. Med. Chem. 1997, 5 (2), 283. (g) Ziegler, T.; Eckhardt, E.; Strayle,
J.; Herzog, H. Carbohydr. Res. 1994, 253, 167. (h) Zhang, Z.; Magnusson,
G. J. Org. Chem. 1996, 61, 2383. (i) Lichtenthaler, F. W.; Oberthur, M.;
Peters, S. Eur. J. Org. Chem. 2001, 20, 3849. (j) Andreana, P. R.; Sanders,
T.; Janczuk, A.; Warrick, J.; Wang, P. G. Tetrahedron Lett. 2002, 43, 6525.
(k) Thomas, S. S.; Plenkiewicz, J.; Ison, E. R.; Bols, M.; Zou, W.; Szarek,
W. A.; Kisilevsky, R. Biochim. Biophys. Acta 1995, 1272 (1), 37. (l) Kucar,
S.; Zamocky, J.; Bauer, S. Collect. Czech. Chem. Commun. 1975, 40 (2),
457.
(11) Standard curve was prepared by making a series of standards
(sugar: 0.5, 1, 2, 2.5, 3, 3.5, 4, 5, 6, 7, and 8 mM respectively) and ATP
(10 mM) in sodium phosphate buffer (final volume 50 µL and pH ) 7.5)
and submitting them to the DNS assay (ref 10).
(12) To determine the Km for galactose, the sugar concentration was
varied over a range of 1-6 mM, and ATP was kept at a constant saturating
concentration of 10 mM. A series of reactions were set up, and the DNS
assay was monitored reaction as mentioned above. The initial velocity was
determined by the slope of the linear phase in the progress curve. Km of
ATP was determined where ATP was a variable substrate (0.5-10 mM)
and galactose concentration was fixed at a saturating concentration of 10
mM. Kinetics data were analyzed by Enzyme Kinetics Module, Version
1.1. The saturation plots and Lineweaver-Burke plots are shown in
Supporting Information.
Org. Lett., Vol. 5, No. 13, 2003
2225

Scheme 3. Preparation of 3-C-Methyl-D-galactose,
4-Azido-4-deoxy-D-galactose, and 6-Azido- and
6-Amino-D-galactosea
Figure 3. Reaction progress monitored by DNS assay. Lane 1: 1.
Lane 2: 2. Lane 3: 3. Lane 4: 5. Lane 5: 11. Lane 6: 10.
consistent with galactokinases from other species,4 moder-
ately active (a <10-fold decrease in kcat) with substrates
substituted at C-2 (2, kcat ) 30 min-1, Km ) 3.6 mM; 3, kcat
) 11.7 min-1, Km ) 2.9 mM). However, the extent of
tolerance of C-2 alteration (e.g., increasing C-2 steric bulk,
4) is limited. GalK is also weakly active (a <20-fold decrease
in kcat) with substrates lacking hydrogen-bonding potential
at either C-3 or C-6 (5, kcat ) 5.1 min-1, Km ) 6.4 mM; 11,
kcat ) 2.9 min-1, Km ) 4.9 mM), extending the known
substrates for this class of enzyme.
It is interesting to note that the lack of tolerance for C-4
changes in the substrate, yet given the typically high
concentrations of cellular glucose, this stringent C-4 control
clearly supports the need for the alternative (and typically
highly regulated) metabolic entry of glucose via the dual
action of hexokinase and phosphoglucomutase. Moreover,
the demonstrated DNS assay can be used for the rapid
screening of various sugar processing enzymes and presents
a convenient screen for future directed evolution work of
GalK as a new component of IVG.
a Reaction conditions: (a) BzCl, pyridine; (b) Tf2O, then NaN3,
DMF; (c) MeONa, then 2 N H2SO4; (d) anhydrous CuSO4, acetone,
concentrated H2SO4; (e) anhydrous pyridine, TsCl; (f) NaN3, DMF;
(g) 80% CF3COOH; (h) 10% Pd/C, H2; (i) SnO(Bu)2, PMBCl; (j)
BnBr, NaH, then CAN; (k) Dess-Martin oxidant; (l) CH3MgBr;
(m) 10% Pd/C, H2.
follows substrate disappearance and thereby eliminates the
chance of false positive results, which have been observed
in other sugar kinase-coupled assays.14
The substrate specificity studies revealed that GalK is
highly active with D-galactose (Figure 2), as expected, and,
Supporting Information Available: Saturation plots,
Lineweaver-Burk plots, and synthetic details for 5, 6, 9,
12, 13, and 16-24. This material is available free of charge
via the Internet at http://pubs.acs.org.
(14) Chenault, H. K.; Mandes, R. F.; Hornberger, K. R. J. Org. Chem.
1997, 62, 331.
OL034642D
2226
Org. Lett., Vol. 5, No. 13, 2003