Preparative Enzymatic Synthesis and Characterization of the Cytoplasmic Intermediates of Murein Biosynthesis

Article abstract of DOI:10.1021/ja983850b

The first six cytoplasmic intermediates of murein biosynthesis (2-7) have been prepared in high yield (>60%) and at preparative scale (10-100 mg) using purified enzymes of the murein biosynthetic pathway (MurA-MurF). For at least three of these compounds,

Full text of DOI:10.1021/ja983850b

J. Am. Chem. Soc. 1999, 121, 1175-1178
1175
Preparative Enzymatic Synthesis and Characterization of the
Cytoplasmic Intermediates of Murein Biosynthesis
†
Sreelatha G. Reddy, Sherman T. Waddell, David W. Kuo, Kenny K. Wong,* and
David L. Pompliano*
Departments of Biochemistry and Medicinal Chemistry, Merck Research Laboratories,
Rahway, New Jersey 07065-0900
ReceiVed NoVember 6, 1998
Abstract: The first six cytoplasmic intermediates of murein biosynthesis (2-7) have been prepared in high
yield (>60%) and at preparative scale (10-100 mg) using purified enzymes of the murein biosynthetic pathway
(MurA-MurF). For at least three of these compounds, 5, 6, and 7, this is the first high-yield synthesis and
purification by any means that has been reported. Detailed spectroscopic and analytical data accompany each
of the six compounds and serve as reference standards for future work. The availability of these intermediates
will facilitate the study of the murein pathway enzymes MurA-MurF, all of which are validated antibacterial
targets.
Introduction
We also provide detailed characterization, including complete
H and C NMR assignments, elemental analysis, and mass
spectral data, for each of the intermediates, many of which
heretofore have not been chemically characterized.
1
13
Cell wall, or murein, biosynthesis is an essential and uniquely
bacterial process and thus is an attractive target for antibiotic
action. While many antibacterial compounds, such as â-lactams,
vancomycin, fosfomycin, and bacitracin, target this complex
Materials and Methods
1
2
pathway, the emergence of bacterial resistance now threatens
the effectiveness of these (and virtually all other) antibiotics.
The seriousness of the medical concern has invigorated the
search both for antibacterial compounds of novel chemical
structure as well as for new mechanisms by which to kill
bacteria. The cytoplasmic steps of murein precursor assembly,
Materials. The disodium salt of UDP-N-acetyl glucosamine, phos-
phoenol pyruvate hexaammonium salt, â-NADH, â-NADPH, DTT,
ATP, and ammonium acetate were purchased from Sigma. Pyruvate
kinase type III from rabbit muscle, L-lactate dehydrogenase type
XXIX-S from porcine muscle, L-alanine, D-glutamic acid, meso-
diaminopimelic acid, and D-alanine-D-alanine were also obtained from
Sigma. Ammonium acetate buffers were prepared by titrating with
ammonium hydroxide or acetic acid to obtain the desired pH.
3
catalyzed by the enzymes MurA through MurF (Scheme 1),
offer a wealth of unexploited targets for which permeable
inhibitors might be found.
Since temperature-sensitive mutants of each of the murA
through murF genes are lethal and each of the wild type genes
are highly conserved among a wide range of bacteria, permeable
inhibitors of any one of the mur gene products will be bac-
teriocidal and have a wide spectrum. Screening these enzymes
for such compounds, either singly or combinatorially using a
UDP-N-acetyl muramyl substrates were assayed routinely following
3
the consumption of NADH (at 340 nm wavelength, ꢀ ) 6.2 × 10
-1
-1
M
cm ) in a pyruvate kinase/lactate dehydrogenase-coupled assay
to detect the enzymatic formation of ADP (converting PEP to pyruvate
with regeneration of ATP). Reaction mixtures (400 µL) contained bis-
Tris propane (100 mM), NADH (200 µM), ATP (1 mM), PEP (10
2
mM), MgCl (5 mM), DTT (1 mM), pyruvate kinase (215 units), and
4
L-lactate dehydrogenase (995 units). The amino acids L-alanine,
D-glutamic acid, and meso-diaminopimelic acid were dissolved in water.
The reaction is initiated by the addition of 5 µL of Mur enzymes (typical
protein concentration of 2 mg/mL). The extents of reaction of substrates
reconstituted murein pathway of MurA through MurF (Scheme
1
), requires pure pathway intermediates. Unfortunately, many
of these intermediates are not readily available to support the
necessary mechanistic and structural studies of these enzymes.
Here, we report a unified protocol for the rapid synthesis and
purification of 10-100 mg quantities of murein substrates 2-7.
5
2
and 3 were assayed by adding MurB, MurC, and L-alanine, whereas
the extent of reactions of substrates 4, 5, 6, and 7 were assayed by
adding only reaction mixtures and assaying for ADP, since ADP is
produced in the reactions.
1H NMR spectra were recorded on either a Bruker AMX-500 (500
MHz) or a Varian Mercury-400 (400 MHz) spectrometer. Chemical
shifts are reported in ppm from trimethylsilane with the solvent
*
To whom correspondence should be addressed. E-mail: kenny wong@
merck.com or david pompliano@merck.com.
†
Department of Medicinal Chemistry.
(
1) Ghuysen, J.-M., Hackenbeck, R., Eds. Bacterial Cell Wall. New
ComprehensiVe Biochemistry; Elsevier: Amsterdam, 1994; Vol. 27.
2) (a) Goetz, A.; Yu, V. L. Curr. Opin. Infect. Dis. 1997, 10, 319-
23. (b) Chadwick, D. J., Goode, J., Eds. Antibiotic Resistance: Origins,
resonance as the internal standard (CDCl
3
: δ 7.24 ppm, CD OD: δ
3
(
3
.25 ppm, D O: δ 4.80 ppm). Data are reported as follows: chemical
2
3
shift, multiplicity (s ) singlet, d ) doublet, t ) triplet, q ) quartet, p
) pentet, br ) broad, m ) multiplet), integration and coupling constants
(Hz). 13C NMR spectra were recorded on either a Bruker AMX-500
EVolution and Spread; Ciba Foundation Symposium 207, John Wiley &
Sons: New York, 1997. (c) Davies, J. Science 1994, 264, 375. (d) Cohen,
M. L. Science 1992, 257, 1050.
(
3) The capitalized gene name in roman type is used to refer to the gene
product throughout (i.e., the gene product of murA is designated MurA).
4) Wong, K. K.; Kuo, D. W.; Chabin; R. M.; Fournier, C.; Gegnas, L.
(
125 MHz) or a Bruker DPX-300 (75 MHz) spectrometer. Chemical
shifts are reported in ppm from trimethylsilane with the solvent
resonance as the internal standard (CDCl : δ 77.0 ppm, CD OD: δ
49.8 ppm). Dioxane (δ 66.5 ppm) was added to D O as internal
standard. By employing a combination of two-dimensional NMR
(
3
3
D.; Waddell, S. T.; Marsilio, F.; Leiting, B.; Pompliano, D. L. J. Am. Chem.
Soc. 1998, 120, 13527-13528.
2
(5) Bugg, T. D.; Walsh, C. T. Nat. Prod. Rep. 1992, 9, 199-215.
1
0.1021/ja983850b CCC: $18.00 © 1999 American Chemical Society
Published on Web 01/27/1999

1176 J. Am. Chem. Soc., Vol. 121, No. 6, 1999
Reddy et al.
Scheme 1
techniques, including COSY (proton-proton correlation), HMQC (one-
bond proton-carbon correlation), and HMBC (two- and three-bond
proton-carbon correlation), it was possible to assign the proton and
carbon spectra for each of the murein enzyme substrates with few
ambiguities. In the proton spectrum, many resonances overlap, par-
ticularly in the region from δ 4.5 to δ 3.5, and peaks corresponding to
individual protons are sometimes impossible to discern in the one-
dimensional proton spectrum. In these cases, chemical shifts were
assigned via cross-peaks in the HMQC spectrum. Occasionally, precise
determination of chemical shift was impossible due to overlap of
HMQC cross-peaks, and in these cases the chemical shifts for the
overlapping protons are given as a range. Microanalytical data were
obtained from Robertson Microlit Laboratories, Inc. NJ. Mass spectra
were recorded on a Finnigan MAT TSQ700 triple quadropole mass
spectrometer for electrospray ionization (ESI). HPLC analysis was
performed using YMC AX300 weak anion-exchange column (4.6 mm
internal diameter × 250 mm length) and run using an isocratic gradient
mmol), PEP (114 mg, 427 mmol), â-NADPH (237 mg, 284 mmol),
and DTT (4 mmol) was prepared in a round-bottom flask and flushed
with nitrogen for 30 min. MurA (0.056 mg/mL) and MurB (0.056 mg/
mL) were added, which initiated the reaction, and the reaction flask
was flushed with nitrogen for 10 min. The extent of the reaction was
determined by assaying for 3 using MurC and the pyruvate kinase/
lactate dehydrogenase-coupled assay. The reaction was complete within
5 h at 37 °C or overnight at 25 °C. The workup of the reaction was
carried out as in the case of 2 except that the MonoQ anion-exchange
column (10 × 10 cm Pharmacia) was eluted with a nonlinear gradient
4
from 0.02 to 1.0 M NH OAc, pH 5.0. The active fractions were pooled,
concentrated, and lyophilized. The lyophilized flocculent powder was
dissolved in water and lyophilized three times to remove ammonium
acetate. Finally, the residue was dissolved in water and stored at -80
°C. Product 3 (130 mg, 87%) was analyzed by FAB mass spectrometry,
1
13
HPLC, H and C NMR, and elemental analysis. MS (FD) m/z 680
(100%, M + H); Anal. Calcd for formula C20 : C, 29.27;
31 3 19 2
H N O P
of 300 mM KH
2
PO4 pH 3.5 (1.5 mL/min over 25 min). The reaction
H, 6.14; N, 10.24; P, 7.55. Found: C, 29.35; H, 5.86; N, 10.56; P,
7.24.
UDP-N-Acetylmuramyl-L-alanine (4). Purified 3 was used as a
starting material to synthesize 4. The reaction mixture contained TAPS
mixtures were concentrated by overnight lyophilization with Virtis
Benchtop 3L lyophilizer. All UV/vis spectra were recorded using a
HP8425A diode-array UV/visible spectrophotometer (Hewlett-Packard)
equipped with a thermostated, multicell-transport cell holder connected
to a RM6 Lauda thermostat.
2
(50 mM, pH 8.0), ATP (475 mmol), MgCl (190 mmol), DTT (95
mmol), L-alanine (380 mmol), and 3 (123.5 mmol). The MBP fusion
MurC (0.2 mg/mL) was added to initiate the reaction. The extent of
the reaction was followed by quantitating the amount of ADP present
in the reaction mixture using the pyruvate kinase/lactate dehydrogenase-
coupled assay. The reaction was complete within 5 h at 37 °C. The
workup of the reaction was carried out as in the case of 2 except that
the Mono Q anion-exchange column (10 × 10 cm, Pharmacia) was
Synthesis of UDP-N-Acetylglucosamine Enolpyruvate (2). A
reaction mixture containing TAPS (100 mM, pH 7.5), UDP-N-acetyl
glucosamine (1) (32.5 mg, 50 mmol), and PEP (18.7 mg, 70 mmol)
was prepared in a round-bottom flask. MurA (0.056 mg/mL) was added
to the reaction mixture and incubated at 37 °C for 5 h. The extent of
the reaction was determined by assaying for 2 with MurB and NADPH,
following the decrease in absorbance at 340 nm. The reaction was
complete within 5 h at 37 °C. The reaction mixture was filtered through
an Amicon YM 10 membrane to remove the enzymes. Part of the
reaction mixture (5 mL) was applied to a MonoQ anion-exchange
column (10 × 10 cm, Pharmacia) and eluted with a nonlinear gradient
4
eluted with a nonlinear gradient from 0.02 to 1 M NH OAc, pH 9.0.
The active fractions were pooled, concentrated, and lyophilized. The
lyophilized material was dissolved in 20 mL of water, and 5 mL of it
was applied to the pre-equilibrated MonoQ column with a nonlinear
4
gradient of 0.02-1 M NH OAc, pH 5.0 buffer. Fractions containing
4
from 0.02 to 1.0 M NH OAc, pH 8.0. The active fractions were pooled,
the product were pooled and lyophilized. The lyophilized flocculent
powder was dissolved in water and lyophilized three times to remove
ammonium acetate. Finally, the residue was dissolved in water and
stored at -80 °C. Product 4 (66 mg, 79%) was analyzed by FAB mass
concentrated, and lyophilized. The lyophilized flocculent powder was
dissolved in water and lyophilized three times to remove the ammonium
acetate buffer. The residue was dissolved in water and stored at -80
1
13
°
C. Product 2 (30 mg, 92%) was analyzed by FAB mass spectrometry,
spectrometry, HPLC, H and C NMR, and elemental analysis. MS
(FD) m/z 751.1 (60%, M + H); Anal. Calcd for C23 : C,
1
13
HPLC, H and C NMR. MS (FD) m/z 679 (100%, M + H).
UDP-N-Acetylmuramic Acid (3). A reaction mixture containing
TAPS (50 mM, pH 8.0), UDP-N-acetyl glucosamine (150 mg, 230
36 4 20 2
H N O P
29.23; H, 6.29; N, 11.86; P, 6.55. Found: C, 29.55; H, 6.22; N, 12.18;
P, 6.18.

Cytoplasmic Intermediates of Murein Biosynthesis
J. Am. Chem. Soc., Vol. 121, No. 6, 1999 1177
Results and Discussion
UDP-N-acetylmuramyl-L-alanyl-γ-D-glutamic Acid (5). Purified
3
was used as a starting material to synthesize 5. The reaction mixture
contained TAPS (50 mM, pH 8.0), ATP (600 mmol), MgCl (240
mmol), DTT (120 mmol), L-alanine (480 mmol), D-glutamic acid (600
mmol), and 3 (120 mmol). MBP fusion MurC (0.2 mg/ml) and MurD
2
Although cell wall precursors 3-7 will accumulate in bacteria
treated with specific inhibitors of the murein pathway, purifica-
tion of these compounds is inefficient and labor intensive.6
Chemical synthesis is not an easy route either, given the stereo-
chemical complexity of these intermediates, and only 1 and (the
(0.25 mg/ml) were added to initiate the reaction. The progress of the
reaction at 37 °C was followed by assaying for ADP activity as in the
case of 4. The reaction was complete within 5 h. The workup of the
reaction was carried out as in the case of 2 except that the MonoQ
anion-exchange column (10 × 10 cm, Pharmacia) was eluted with a
7
8
lysyl-containing version of) 7 have been prepared synthetically.
Although enzymatic syntheses and purifications of 2, 3, and 4
have been reported, the HPLC purification protocol limited the
4
nonlinear gradient from 0.02 to 1.0 M NH OAc, pH 9.0. The active
9
amount of material that could be processed. A larger scale,
fractions were pooled concentrated and lyophilized. The lyophilized
material was dissolved in 20 mL of water and 5 mL of it was applied
to the preequilibrated MonoQ column with a nonlinear gradient of
10
open-column chromatography method to purify 3 using trieth-
ylamine ammonium bicarbonate (TEAB) as solvent produced,
in our hands, a persistent contaminant of 317 amu (as revealed
0
4
.02-1M NH OAc pH 9.0 buffers. Fractions containing the product
1
1
were pooled and lyophilized. The lyophilized flocculent powder was
dissolved in water and lyophilized three times to remove ammonium
acetate. Finally, the residue was dissolved in water and stored at -80
by FAB mass spectral analysis). No high yield synthesis and
purification of 5, 6, or 7 by any method has been reported.
The drawbacks of literature protocols compelled us to devise
a generalized procedure for producing large quantities of each
cell wall precursor. Since recombinant Escherichia coli mur
genes can be overexpressed in soluble form in E. coli and
°
C. Product 5 (60 mg, 74%) was analyzed by FAB mass spectrom-
1
13
etry, HPLC, H and C NMR, and elemental analysis. MS (FD)
m/z 880.2 (100%, M + H); Anal. Calcd for C28 : C, 29.38;
43 5 23 2
H N O P
H, 6.43; N, 13.46; P, 5.41. Found: C, 29.38; H, 6.36; N, 13.05;
P, 5.13.
1
2
purified in gram quantities, enzymatic synthesis appeared to
be the most efficient way to proceed. In short, the target
compound was prepared by incubating 1 with the Mur enzyme-
(s) and their cosubstrates that preceded the target compound in
the metabolic sequence. Each intermediate could be produced
in turn, capitalizing on the in situ synthesis of the prior
UDP-N-Acetylmuramyl-L-alanyl-γ-D-glutamyl-m-diaminopimel-
ic Acid (6). To synthesize 6, purified 3 was used as a starting material.
The reaction mixture contained TAPS (50 mM, pH 8.0), ATP (225
2
mmol), MgCl (90 mmol), DTT (45 mmol), L-alanine (135 mmol),
D-glutamic acid (135 mmol), meso-diaminopimelic acid (135 mmol),
and 3 (22.5 mmol). MBP fusion MurC (0.2 mg/mL), MurD (0.25
mg /mL), and MurE (0.25 mg/mL) were added to initiate the reac-
tion. The progress of the reaction was followed by assaying for ADP
activity as in the case of 4 and 5. The reaction was complete within
13
intermediate. Thus, synthesis of 2 from 1 was straightforward
using MurA (Table 1). Developing a one-pot, two-enzyme
(MurA and MurB) synthesis of 3 from 1 to bypass the
purification of 2 was complicated by the sensitivity of MurB
5
h at 37 °C. The work up of the reaction was carried out as in case
of 2 except MonoQ anion-exchange column (10 × 10 cm Pharm-
acia) was eluted with a nonlinear gradient from 0.02 to 1 M NH OAc,
9
d
to substrate and product inhibition, which precluded the use
of high substrate concentration in the reaction. Using high
concentrations of MurB mitigated the substrate inhibition
problem but also revealed an intrinsic NADPH oxidase activity
4
pH 5.0. The active fractions were pooled, concentrated and lyophilized.
The lyophilized material was dissolved in 20 mL of water, and 5 mL
of it was applied to the preequilibrated MonoQ column with a nonlinear
gradient of 0.02-1.0 M NH
4
OAc, pH 9.0 buffers. Fractions containing
(6) (a) Flouret, B.; Mengin-Lecreulx, D.; van Heijenoort. J. Anal.
Biochem. 1981, 114, 59-63. (b) Kohlrausch, U.; Holtje, J. V. Fems.
Microbiol. Lett. 1991, 78, 253-258. (c) Allen, N. E.; Hobbs, J. N., Jr.;
Richardson, J. M.; Riggin, R. M. Fems. Microbiol. Lett. 1992, 98, 109-
product were pooled and lyophilized. The lyophilized flocculent powder
was dissolved in water and lyophilized three times to remove am-
monium acetate. Finally, the residue was dissolved in water and stored
at -80 °C. Product 6 (10.4 mg, 68%) was analyzed by FAB mass
1
16. (c) Michaud, C.; Blanot, D.; Flouret, B.; van Heijenoort, J. Eur. J.
Biochem. 1987, 166, 631-637. (d) Mizuno, Y.; Ito, E. J. Biol. Chem. 1968,
1
13
spectrometry, HPLC, and H and C NMR. MS (FD) m/z 1052.9
: C, 34.29; H, 6.49;
243, 2665-2672. (e) Anwar, R. Vlaovic, M. Biochem. Cell Biol. 1986, 64,
2
97-303.
54 7 26 2
(100%, M + H). Anal. Calcd for C35H N O P
(
7) Within eubacteria, the third position of the pentapeptide side chain
N, 13.71; P, 5.05. Found: C, 34.70; H, 6.29; N, 13.55; P, 4.65.
of 7 is the most variable. There is no easy formula to predict which amino
acid residue occurs in this position for a given bacterial species. Of the
more than nine different amino acid residues known to occupy this position,
the most common is meso-diaminopimelate (as in 7), which is probably
present in all Gram-negative bacteria. L-Lysine occurs frequently in the
MurNAc-pentapeptide of Gram-positive cocci. For a review, see: Schleifer,
K. H.; Kandler, O. Bacteriol. ReV. 1972, 36, 407-477.
UDP-N-Acetylmuramyl-L-alanyl-γ-D-glutamyl-m-diaminopime-
lyl-D-alanyl-D-alanine (7). To synthesize 7, purified 3 was used as a
starting material. The reaction mixture contained TAPS (50 mM, pH
8
2
.0), ATP (175 mmol), MgCl (70 mmol), DTT (35 mmol), L-alanine
(
(
105 mmol), D-glutamic acid (105 mmol), meso-diaminopimelic acid
105 mmol), D-alanine-D-alanine (105 mmol), and 3 (17.5 mmol). MBP
(8) (a) Heidlas, J. E.; Lees, W. L.; Pale, P. Whitesides, G. M. J. Org.
fusion MurC (0.2 mg/mL), MurD (0.25 mg/mL), MurE (0.25 mg/mL)
and MurF (0.25 mg/mL), were added to initiate the reaction. The
progress of the reaction was followed by assaying for ADP activity
as in the case of 4, 5, and 6. The reaction was complete within 5 h
at 37 °C. The workup of the reaction was carried out as in the case of
Chem. 1992, 57, 146-151. (b) Hitchcock, S. A.; Eid, C. N.; Aikins, J. A.;
Zia-Ebrahimi, M.; Blaszczak, L. C. J. Am. Chem. Soc. 1998, 120, 1916-
1917.
(
9) (a) Falk, P. J.; Ervin, K. M.; Volk, K. S.; Ho, H. Biochemistry 1996,
3
5, 1417-1422. (b) Benson, T. E.; Marquardt, J. L.; Marquardt, A. C.;
Etzkorn, F. A.; Walsh, C. T. Biochemistry 1993, 32, 2024-2030. (c) Ho,
H. T.; Falk, P. J.; Ervin, K. M.; Krishnan, B. S.; Discotto, L. F.; Dougherty,
T. J.; Pucci, M. J. Biochemistry 1995, 34, 2464-2470. (d) Dhalla, A. M.;
Yanchunas, J., Jr.; Ho, H.-T.; Falk, P. J.; Villafranca, J. J.; Robertson, J.
G. Biochemistry 1995, 34, 5390-5402.
2
except that the MonoQ anion-exchange column (10 × 10 cm,
Pharmacia) was eluted with a nonlinear gradient from 0.02 to 1 M
NH OAc, pH 9.0. The active fractions were pooled, concentrated, and
4
lyophilized. The lyophilized material was dissolved in 20 mL of water,
and 5 mL of it was appilied to the pre-equilibrated MonoQ column
with a nonlinear gradient of 0.02-1.0 M NH OAc, pH 5.0 buffers.
4
(
10) Jin, H.; Emanuele, J. J., Jr.; Fairman, R.; Robertson, J. G.; Hail, M.
E.; Ho, H.-T.; Falk, P. J.; Villafranca, J. Biochemistry 1996, 35, 1423-1431.
11) Commercially available TEAB buffer and triethylamine base both
(
Fractions containing the product were pooled and lyophilized. The
lyophilized flocculent powder was dissolved in water and lyophilized
three times to remove ammonium acetate. Finally, the residue was
dissolved in water and stored at -80 °C. Product 7 (9.3 mg, 61%) was
contained a contaminant of the same mass. We suspect that this contaminant
is a product of light-activated polymerization of triethylamine that yields a
nonvolatile compound. Because of this problem, the TEAB solvent system
was not used to purify the murein substrates.
1
13
(12) Pryor, K. D.; Leiting, B. Protein Expression Purif. 1997, 10, 309-
19.
analyzed by FAB mass spectrometry, HPLC, and H and C NMR.
MS (FD) m/z 1194.4 (100%, M + H). Anal. Calcd for C41
C, 34.65; H, 6.81; N, 14.78; P, 4.36. Found: C, 34.81; H, 6.46; N,
4.52; P, 3.90.
3
64 9 28 2
H N O P :
(13) 1 is commercially available or it can be prepared: ref 8a or Leiting,
B.; Pryor, K. D.; Eveland, S. S.; Anderson, M. S. Anal. Biochem. 1998,
256, 185-191.
1

1
178 J. Am. Chem. Soc., Vol. 121, No. 6, 1999
Table 1
substrate
Reddy et al.
scale (mg)
yield (%)
2
3
4
5
6
7
30
130
66
60
10.4
9.3
92
87
79
74
68
61
of MurB (constituting 10% of the normal reaction rate). Because
of this side reaction, high levels of NADPH would be required,
adding difficulty to the purification problem. However, by
performing the reaction under argon, conversion of 1 to 3 was
accomplished in >95% yield on 100 mg scale (Table 1).
Murein ligases MurC through MurF each catalyze an ATP-
dependent addition of an amino acid to the free carboxylate of
its substrate (Scheme 1). To synthesize 4-7, we used 3 instead
of 1 as the starting material to reduce the number and quantity
of the reaction byproducts that need to be eliminated during
the purification. Typical reactions included 3 (0.5-1.5 mM),
the relevant murein ligases, and a high concentration of ATP
(5 mM) to provide enough energy equivalence for the ligases
as well as to drive the reaction to completion (Table 1). A
potential worry is the susceptibility of MurD and MurF to
substrate inhibition by high concentrations of UDP-N-acetyl-
muramyl-peptide (MurNAc-peptide) intermediates.14 This con-
cern can be overcome by running the reaction under conditions
where the flux through each biosynthetic enzyme is roughly
equal, allowing no buildup of any intermediate.4
Although running reactions is easy using enzymes, purifying
the water-soluble target product from a reaction mixture con-
taining closely related, highly charged reactants and byproducts
is a real challenge. In general, after removing the enzymes by
ultrafiltration and lyophilizing the reaction mixture, the desired
murein intermediate was purified by FPLC anion-exchange
chromatography (MonoQ) using an ammonium acetate solvent
buffered at various hydrogen ion concentrations (pH 5.0, 8.0,
or 9.0). All of the murein intermediates were purified varying
only the pH in this chromatography system.
Figure 1. Representative spectroscopic data for murein pathway
1
13
intermediates. (A) H NMR spectrum of 5, (B) C NMR spectrum of
(C) FAB mass spectrum of 5.
5
carbon correlation (HMQC), and two- and three-bond proton-
carbon correlation (HMBC) allowed for the complete proton
and carbon peak assignments for 2-7 (see Supporting Informa-
tion).
Compounds 2 and 3 were purified using ammonium acetate
solvent at pH 8.0 and 5.0, respectively. For the purification of
Other than a recent description of the Park nucleotide,8b this
paper is the first to describe the detailed characterization of six
murein pathway intermediates (2-7). The rapid synthesis (from
commercially available 1) and purification of these cell wall
precursors will facilitate the study of the early steps of murein
biosynthesis, a known target for antibiotic action. In addition,
radio-labeled cell wall intermediates are easily prepared by
including the relevant radio-labeled amino acid in the reaction
mixture. More importantly, the availability of all murein
substrates should allow for rapid analysis of inhibitors from
murein pathway screens, to identify their enzyme target and
elucidate their mechanisms of action.
4
and 5, and especially of 6 and 7, good separation from ADP
and ATP is crucial because, as the number of ligase steps
performed in one pot increases, so does the amount of ADP
byproduct in the final mixture. Resolving MurNAc-peptide
products 4, 5, 6, or 7 from ATP and ADP required two passages
through the anion-exchange column, the first time using
ammonium acetate at pH 9.0 and then at pH 5.0, or vice versa.
After solvent was removed from fractions containing target
compound, the solid residue was dissolved in water and stored
in aliquots at -80 °C. The murein intermediates are stable for
at least one year.
Since few murein substrates have been spectroscopically
characterized, each of the purified precursors was subjected to
a battery of analytical inspections. Using H and C NMR,
HPLC, FAB mass spectrometry, and elemental analysis, the
murein substrates were all in the range of 96-99% pure (Figure
Acknowledgment. The authors thank Barbara Leiting
(MRL) for the recombinant stage I enzymes used for the
synthesis of all the murein substrates, Tracey D. Klatt (MRL)
for assistance in the mass spectral analysis, Renee M. Chabin
(MRL) for HPLC analysis, and John W. Kozarich (MRL) for
his unbridled encouragement and resolute support.
1
13
1
). Comparison of the concentration derived from UV spectro-
-
1
-1
scopic measurement (using ꢀ262 ) 10 100 M cm ) with that
derived from an enzymatic end-point assay gave excellent agree-
ment. A combination of two-dimensional NMR techniques in-
cluding proton-proton correlation (COSY), one-bond proton-
Supporting Information Available: Analytical and spectral
data for all murein substrates, including complete proton and
carbon NMR assignments, are provided (PDF). This material
is available free of charge via the Internet at http://pubs.acs.org.
(14) Anderson, M. S.; Eveland, S. S.; Onishi, H. R.; Pompliano, D. L.
Biochemistry 1996, 35, 16264-16269.
JA983850B