Specificity of lysine: N6-hydroxylase: A hypothesis for a reactive substrate intermediate in the catalytic mechanism

Article abstract of DOI:10.1006/bioo.1996.0034

The recombinant cytoplasmic preparation of lysine: N⁶-hydroxylase catalyzes the conversion of L-lysine to its N⁶-hydroxy derivative when supplemented with the cofactors NADPH and FAD. A number of lysine analogs reflecting minor alterations in the inherent structural features of the amino acid as well compounds with relatively high affinity for lysine binding domains in other proteins were examined for their ability to serve as substrates of lysine: N⁶-hydroxylase. These studies have revealed that the enzyme does not tolerate any change in the structural features of L-lysine, its preferred substrate, with the exception of the replacement of the C(γ)H₂-methylene group by sulfur, as in (S)-2-aminoethyl-L-cysteine. L-Norleucine is a potent inhibitor of the enzyme while L-norvaline and L-α-aminobutyric acid do not exhibit such effect, indicating the importance of a C₄ hydrophobic side chain for effective interaction with the enzyme. Among the N-alkyl amides of hydrophobic amino acids, only L-norleucine methylamide and L-α-aminobutyric acid ethylamide serve as moderate inhibitors of lysine: N⁶-hydroxylase. Based on the enzyme's stringent substrate specificity, a mechanism involving the conversion of L-lysine to 2-aminocaprolactam prior to its oxygenation by the 4α-peroxyflavin intermediate in the catalytic cycle is proposed.

Full text of DOI:10.1006/bioo.1996.0034

BIOORGANIC CHEMISTRY 24, 401–416 (1996)
ARTICLE NO. 0034
Specificity of Lysine : N6-Hydroxylase: A Hypothesis for a
Reactive Substrate Intermediate in the Catalytic Mechanism
2
L. MARRONE,1 S. SIEMANN,1 M. BEECROFT
, AND T. VISWANATHA
Department of Chemistry, University of Waterloo, Waterloo, Ontario N2L 3G1 Canada
Received July 22, 1996
The recombinant cytoplasmic preparation of lysine : N6-hydroxlase catalyzes the conversion
of L
-lysine to its N6-hydroxy derivative when supplemented with the cofactors NADPH and
FAD. A number of lysine analogs reflecting minor alterations in the inherent structural
features of the amino acid as well compounds with relatively high affinity for lysine binding
domains in other proteins were examined for their ability to serve as substrates of lysine : N6-
hydroxylase. These studies have revealed that the enzyme does not tolerate any change in
the structural features of
of the CͲH2-methylene group by sulfur, as in (S)-2-aminoethyl-
potent inhibitor of the enzyme while -norvaline and -Ͱ-aminobutyric acid do not exhibit
such effect, indicating the importance of a C4 hydrophobic side chain for effective interaction
with the enzyme. Among the N-alkyl amides of hydrophobic amino acids, only -norleucine
methylamide and
-Ͱ-aminobutyric acid ethylamide serve as moderate inhibitors of lysine : N6-
hydroxylase. Based on the enzyme’s stringent substrate specificity, a mechanism involving the
conversion of -lysine to 2-aminocaprolactam prior to its oxygenation by the 4a-peroxyflavin
intermediate in the catalytic cycle is proposed.
L-lysine, its preferred substrate, with the exception of the replacement
L
-cysteine. -Norleucine is a
L
L
L
L
L
L
1996 Academic Press, Inc.
INTRODUCTION
Lysine : N6-hydroxylase catalyzes the conversion of
L
-lysine into its N6-hydroxy
derivative, the initial step in the biosynthesis of the siderophore aerobactin (1–4).
The wild-type enzyme is membrane associated, a feature that has precluded the
elucidation of the molecular basis for the catalytic mechanism of the protein (4, 5).
However, recombinant cytoplasmic preparations of lysine : N6-hydroxylase, rIucD,3
have been developed, and these forms require NADPH as well as FAD for their
1 L. Marrone and S. Siemann contributed equally to this work.
2 To whom correspondence should be addressed. Fax: (519) 746-0435. E-mail: tviswan@chemistry.
watstar.uwaterloo.ca.
3 Abbreviations used: DCC, N,NЈ-dicyclohexyl carbodiimide; DCU, N,NЈ-dicyclohexyl urea; dec.,
decomposition; DTT, dithiothreitol; ESMS, electrospray mass spectrometry; FAD, flavin adenine dinu-
cleotide (oxidized form); G-6-P, glucose-6-phosphate; G-6-P deH2 , glucose-6-phosphate dehydrogenase;
HAbuNHEt, Ͱ-aminobutyric acid ethylamide; HNleNH2 , norleucineamide; HNleNHMe, norleucine
methylamide; HNvaNHMe, norvaline methylamide; NADPH, nicotinamide adenine dinucleotide phos-
phate (reduced from); NADPϩ, nicotinamide adenine dinucleotide phosphate (oxidized form); nd, not
determined; rIucD, recombinant cytoplasmic lysine : N6-hydroxylase; t-Boc, N-tert-butoxycarbonyl; t-
BocAbuNHEt, t-Boc-protected Ͱ-aminobutyric acid ethylamide; t-BocNleNHMe, t-Boc-protected nor-
leucine methylamide; t-BocNvaNHMe, t-Boc-protected norvaline methylamide.
401
0045-2068/96 $18.00
Copyright 1996 by Academic Press, Inc.
All rights of reproduction in any form reserved.

402
MARRONE ET AL.
catalytic function (6, 7). Preliminary studies have also indicated that the recombinant
forms of the enzyme are specific for -lysine and its analog (S)-2-aminoethyl-
L
L-
cysteine serving as the preferred substrates (6, 7). Current investigations concern
a systematic study of various substrate analogs to serve as either a substrate or an
inhibitor with the aim of gaining an insight into the details of the mechanism
operative in the catalytic process mediated by rIucD.
MATERIALS AND METHODS
General
NMR (1H) spectra were recorded on a Bruker AM-250 spectrometer in either
CDCl3 or D2O with tetramethylsilane as an internal standard (ͳ ϭ 0 ppm).
Mass spectra were recorded with a VG/Quattro II triple quadrupole mass
spectrometer using electrospray ionization and THF or H2O as the solvent.
Ultraviolet–visible spectra were recorded on a Beckman DU 640 spectrophotome-
ter. Melting points were determined in open capillaries on a Mel-Temp apparatus
(Laboratory Devices, U.S.A.) and are not corrected. Thin-layer chromatography
was performed on precoated silica gel sheets (TLC aluminum sheets, silica gel
60F254 , precoated sheets, 20 ϫ 20 cm, layer thickness 0.2 mm by Merck, Darmstadt,
Germany) using an appropriate solvent system. The location of the components
was established by exposing the plates to ninhydrin as follows: the sheets were
sprayed with ninhydrin reagent (0.1% w/v in ethanol) and dried at 80ЊC.
Compounds containing free amino groups appeared as colored spots with charac-
teristic Rf values. In the case of t-Boc-protected amino acids the sheets were
exposed to trifluoroacetic acid vapor for 1 min, dried if necessary, and finally
developed with ninhydrin, as described above.
Materials
p-Aminobenzamidine dihydrochloride, benzamidine hydrochloride, DTT, ethyl-
amine (70% aqueous solution), FAD, G-6-P, G-6-P deH2 , N-hydroxysuccinimide,
L
-lysine, methylamine hydrochloride, ͱ-NADPϩ sodium salt, ͱ-NADPH tetraso-
dium salt, ninhydrin, DL-norleucine, and
L(ϩ)-norleucine were purchased from
Sigma Chemical Company (St. Louis, MO). Trans-4,5-dehydro-
L
-lysine, -norleu-
L
cine methylester, and all t-Boc-protected amino acids were obtained from Bachem
Bioscience Inc. (Philadelphia, PA). Trifluoroacetic acid, iodine, and Ͱ-naphthyla-
mine were purchased from BDH Chemical Company (Toronto, ON, Canada). 1-
Cyclohexyl-3-(2-morpholinoethyl) carbodiimide method-p-toluenesulfonate, hy-
drochloric acid (4 ) in 1,4-dioxane, and sulfanilic acid were obtained from Aldrich
M
(Milwaukee, WI). Dowex 50W-X8 (200–400 mesh, Hϩ form) was purchased from
Bio-Rad Laboratories (Richmond, CA). DCC was purchased from Fluka Chemie
AG (Buchs, Switzerland). All inorganic salts and organic solvents were purchased
from J. T. Baker Chemical Company (Phillipsburg, NJ). Ͱ-N-Methyl- -lysine was
L
kindly provided by Prof. N. L. Benoiton (Department of Biochemistry, University
of Ottawa, Canada).

LYSINE : N6-HYDROXYLASE
403
The isolation and purification of rIucD was achieved by employing procedures
documented in the literature (6, 7). The concentration of the enzyme was determined
1
1
spectrophotometrically using a molar extinction coefficient of 6.75 ϫ 104 MϪ cmϪ
at 280 nm (8).
Determination of Lysine : N6-Hydroxlyase Activity (Hydroxylamine Assay)
The protocol for the determination of lysine : N6-hydroxylase activity was similar
to that described previously (9). The assay mixture, in a final volume of 5.0 ml,
consisted of 100 m
(1 m ), FAD (40 Ȑ
and enzyme (100 n
M
potassium phosphate buffer (pH 7.0), DTT (400 Ȑ
M
),
L
-lysine
M
M
), NADPϩ (160 ȐM), G-6-P (800 Ȑ
M
), G-6-P deH2 (1 unit),
M
). Following incubation at 37ЊC for 15 min, the reaction was
stopped by adding 2.5 ml of a slurry of Dowex 50W-X8 resin (200–400 mesh, Hϩ-
form). After the termination of the reaction, the entire mixture was transferred
into a glass column (1.2 ϫ 25 cm), the resin washed with 40 ml HCl (0.2
M), and
the product eluted with 20 ml HCl (6 ). The eluate was then taken to dryness
M
under reduced pressure. The resulting residue was dissolved in 5.0 ml of distilled
water, and an aliquot (4.5 ml) was finally used for N6-hydroxylysine determination
(10) by the sequential addition of the following reagents: sodium acetate (0.5 ml;
48%, w/v), sulfanilic acid (0.5 ml; 1% w/w in acetic acid (25% v/v)), and iodine (0.2
ml; 1.3% w/v in glacial acetic acid). After standing for 5 min at room temperature,
the excess iodine was removed by treatment with sodium thiosulfate (0.2 ml; 0.1
M), followed by the addition of Ͱ-naphthylamine (0.1 ml; 0.6% w/v in acetic acid
(30%, v/v)). The absorbance at 520 nm was recorded after standing at room tempera-
ture for 20 min.
The effect of desired compounds on the catalytic function of rIucD was deter-
mined by their inclusion in the assay at a final concentration of 1 mM.
NADPH Oxidation
The assay mixture, in a final volume of 3.0 ml, consisted of NADPH (200 Ȑ
FAD (40 Ȑ ), potassium phosphate buffer (100 m , pH 7.0), DTT (400 Ȑ ), and
rIucD (370 n ). The decrease in NADPH concentration was monitored spectropho-
tometrically at 340 nm at 23ЊC.
M),
M
M
M
M
Synthesis of t-Boc
L-Amino Acid Alkylamides
The procedure concerning the first step in the synthesis of t-Boc
L-amino acid
alkylamides (t-BocAbuNHEt, t-BocNvaNHMe, and t-BocNleNHMe), the conver-
sion of a t-Boc amino acid into its N-hydroxysuccinimide ester, was adapted from
the literature (11, 12). The t-Boc-protected -amino acid (1 mmol) and N-hydroxy-
L
succinimide (2 mmol) were dissolved in 5 ml THF, and solid DCC (1.1 mmol) was
added slowly to the solution at Ϫ5ЊC. The mixture was stirred at Ϫ5ЊC for 1 h.
After 2 h of stirring at room temperature, insoluble DCU was removed by filtration,
and the filtrate was stored overnight at Ϫ20ЊC and filtered again. The purity of
the activated esters was established by thin-layer chromatography using CH2Cl2/
methanol (9/1), and ESMS. The filtrate containing the N-hydroxysuccinimide ester
was cooled to 0ЊC. In the case of methyl amides, methylamine hydrochloride (1.4

404
MARRONE ET AL.
mmol) was dissolved in 2 ml H2O, the pH adjusted to 11.0 with KOH, and slowly
added to the filtrate. In the synthesis of t-BocAbuNHEt, 100 Ȑl (approx. 1.2 mmol)
ethylamine (70% in H2O) was diluted to 2 ml with H2O and slowly added to the
filtrate. The reaction mixture was stirred for 2 min at 0ЊC, 15 min at room tempera-
ture, and finally taken to dryness under reduced pressure. The residue was dissolved
in a minimal amount of ethanol and treated with water to facilitate the precipitation
of DCU that could not be removed from the preparation in the previous step. After
the removal of the insoluble material, the filtrate was taken to dryness, and the
residue was suspended in H2O. The t-Boc amino acid amides were extracted with
ethylacetate, and the organic phase was subsequently washed with H2O and satu-
rated NaCl solution. After drying over anhydrous Na2SO4 , the organic solvent was
filtered and taken to dryness under reduced pressure. The purity of the amides
was established by thin-layer chromatography employing CH2Cl2/methanol (9/1),
1
ESMS, and H NMR (250 MHz, CDCl3). t-BocAbuNHEt: yield: 70%; Rf ϭ 0.58;
1H NMR (CDCl3): ͳ 0.93 (t, 3 H, CͲH3), 1.13 (t, 3 H, CH3CH2N), 1.44 (s, 9 H,
(CH3)3C), 1.75 (m, 2 H, CͱH2), 3.30 (m, 2 H, CH2N), 3.93 (m, 1 H, CͰH), 5.0 (band,
1 H, NH), 6.0 (band, 1 H, NH); ESMS: 231.1 (Calcd MHϩ 231.3). t-BocNvaNHMe:
1
yield: 70%; Rf ϭ 0.52; H NMR (CDCl3): ͳ 0.93 (t, 3 H, CͳH3), 1.3–1.5 (m, 2 H,
CͲH2), 1.44 (s, 9 H, (CH3)3C), 1.8 (m, 2 H, CͱH2), 2.81 (d, 3 H, CH3N), 4.04 (m, 1
H, CͰH), 5.0 (band, 1 H, NH), 6.1 (band, 1 H, NH); ESMS: 231.1 (Calcd MHϩ
1
231.3). t-BocNleNHMe: yield: 72%; Rf ϭ 0.60; H NMR (CDCl3): ͳ 0.9 (t, 3 H,
C H3), 1.2–1.5 (m, 4 H, CͲH2CͳH2), 1.5 (s, 9 H, (CH3)3C), 1.9 (m, 2 H, CͱH2), 2.85
␧
(d, 3 H, CH3N), 4.1 (m, 1 H, CͰH), 5.0 (band, 1 H, NH), 6.1 (band, 1 H, NH);
ESMS: 245.1 (Calcd MHϩ 245.3).
Deprotection of t-Boc
The residue containing the t-Boc-protected amino acid amides was dissolved in
4.0 HCl in 1,4-dioxane under nitrogen atmosphere at Ϫ5ЊC. After stirring for 30
L-Amino Acid Amides
M
min at room temperature, the reaction mixture was taken to dryness under reduced
pressure. Recrystallization was achieved from ethanol/THF. In the case of
HAbuNHEt, however, attempts to crystallize the compound have been unsuccess-
ful. The purity of the deprotected amides was established by thin-layer chromatogra-
phy using n-butanol/acetic acid/H2O (4/1/2), ESMS, and 1H NMR (250 MHz, D2O).
1
HAbuNHEt: yield: 90%; Rf ϭ 0.48; H NMR (D2O): ͳ 0.81 (t, 3 H, CͲH3), 0.97 (t,
3 H, CH3CH2N), 1.72 (m, 2 H, CͱH2), 3.1 (m, 2 H, CH2N), 3.7 (t, 1 H, CͰH); ESMS:
1
131.1 (Calcd MHϩ 131.2). HNvaNHMe: yield: 92%; mp 190–191ЊC; Rf ϭ 0.48; H
NMR (D2O): ͳ 0.77 (t, 3 H, CͳH3), 1.2 (m, 2 H, CͲH2), 1.63 (m, 2 H, CͱH2), 2.62
(s, 3 H, CH3N), 3.76 (t, 1 H, CͰH); ESMS: 131.1 (Calcd MHϩ 131.2). HNleNHMe:
yield: 93%; mp 210–211ЊC; Rf ϭ 0.54; 1H NMR (D2O): ͳ 0.72 (t, 3 H, C H3), 1.1–1.3
␧
(m, 4 H, CͲH2CͳH2), 1.68 (m, 2 H, CͱH2), 2.62 (s, 3 H, CH3N), 3.76 (t, 1 H, CͰH);
ESMS: 145.1 (Calcd MHϩ 145.2).
Synthesis of
-Norleucine amide was prepared according to the procedure employed for the
synthesis of -leucine amide (13). -Norleucine methylester (764 mg, 4.2 mmol)
L
-Norleucine Amide
L
L
L

LYSINE : N6-HYDROXYLASE
405
was added to methanol, saturated with NH3 at Ϫ78ЊC. The methanolic ammonia
solution was prepared by adding approximately 75 ml cold anhydrous methanol
(Ϫ78ЊC) to liquid NH3, which had been condensed previously into a 250-ml round-
bottom flask on a dry ice/ethanol bath. After the addition of the -norleucine
L
methylester, the dry ice/ethanol bath was removed, and the mixture was allowed
to reach room temperature. After 24 h, the reaction was terminated by evaporating
the solvent under reduced pressure. The amide was finally crystallized from
methanol/THF. The purity of the compound was established by thin-layer chroma-
tography using n-butanol/acetic acid/H2O (4/1/2), ESMS, and 1H NMR (250 MHz,
1
D2O). Yield: 85%; mp 200–205ЊC dec.; Rf ϭ 0.58; H NMR (D2O): ͳ 0.72 (t, 3 H,
C H3), 1.1–1.3 (m, 4 H, CͲH2CͳH2), 1.7 (m, 2 H, CͱH2), 3.83 (t, 1 H, CͰH); ESMS:
␧
131.0 (Calcd MHϩ 131.2).
Synthesis of S-Propargyl-
The protocol for the preparation of S-propargyl-
documented for the synthesis of S-alkyl- -cysteine (14).
-Cysteine hydrochloride (1 g, 5.72 mmol) was dissolved in an aqueous solution
L
-Cysteine
L
-cysteine was similar to that
L
L
of Ba(OH)2 (33.3 ml, 0.2
M), and approximately 20 ml of ethanol (absolute) was
added. After the addition of propargyl bromide (1.7 ml, 19.1 mmol), the mixture
was stirred overnight under nitrogen atmosphere. The barium ions were removed
quantitatively with the aid of sulfuric acid (15 ml, 10%), followed by the addition
of absolute ethanol (3–4 vol). After standing overnight at 4ЊC, S-propargyl- -
L
cysteine was recovered as a precipitate. The purity of the compound was established
by thin-layer chromatography using n-butanol/acetic acid/H2O (4/1/5) as well as
1
1H NMR (250 MHz, D2O). Yield: 85%; mp 250ЊC dec.; Rf ϭ 0.44; H NMR
(D2O): ͳ 2.97 (t, 1 H, J ϭ 2.4, HCIC), 3.43 (dd, 1 H, J1 ϭ 7.8, J2 ϭ 14.8, CͱH2),
3.59 (dd, 1 H, J1 ϭ 4.2, J1 ϭ 14.8, CͱH2), 3.66 (d, 2 H, J ϭ 2.4, CICUCH2),
4.27 (dd, 1 H, J1 ϭ 4.2, J2 ϭ 7.8, CͰH).
Synthesis of (S)-2-Aminoethyl-DL-Homocysteine
(S)-2-Aminoethyl-DL-homocysteine was prepared from DL-homocysteine ac-
cording to the procedure employed for the synthesis of (S)-2-aminoethylcysteine
(15). DL-Homocysteine was generated by treatment of DL-homocysteine thiolactone
1
hydrochloride (Sigma) with KOH (1
M
). Yield: 64%; H NMR (250 MHz, D2O): ͳ
2.00 (m, 2 H, CͱH2), 2.55 (t, 2 H, CͲH2), 2.74 (t, 2 H, CH2CH2NHϩ3 ), 3.08 (t, 2 H,
CH2NHϩ3 ), 3.71 (t, 1 H, CͰH).
RESULTS
Effect of Substrate Analogs on the Catalytic Function of rIucD
The influence of various compounds serving as analogs of -lysine was investigated
L
in order to gain an insight into the structure–function relationship of lysine : N6-
hydroxylase. Most of the compounds chosen for these studies share certain structural

406
MARRONE ET AL.
features inherent in
L
-lysine, which consists of a positively charged amino group
(A) at the Ͱ-carbon as well as a negatively charged carboxyl group (B) at the
same position as indicated in Fig. 1. Furthermore, -lysine is characterized by a
L
hydrophobic carbon side chain composed of four methylene groups (Cn), and a
positively charged amino group at the end of this chain (D). Other compounds
were chosen in view of their documented ability to interact with lysine binding
domains in other proteins. This latter class includes: (i) tranexamic acid (trans-4-
(aminomethyl)-cyclohexanecarboxylic acid) and ␧-aminocaproic acid, components
known to bind lysine specific kringles in plasminogen (16); and (ii) benzamidine
as well as p-aminobenzamidine, which serve as inhibitors of trypsin (17), a protease
with remarkable specificity for lysine (and arginine) substrates (18).
As indicated in Table 1, lysine : N6-hydroxylase appears to be stringently specific
with regard to its substrate. However, (S)-2-aminoethyl-
taining the same structural features (A, B, Cn , and D) as the preferred substrate
-lysine, was found to be hydroxylated by the enzyme in significant yields. The
L-cysteine, an analog con-
L
substitution of the methylene group at the Ͳ-position by a sulfur atom still provides
the necessary flexibility and hydrophobicity of the side chain. Neither (S)-2-amino-
ethyl-DL-homocysteine (one additional methylene group), nor
L
-ethionine and S-
propargyl- -cysteine, which are devoid of the ␧-amino function present in lysine,
L
was found to serve as a potent substrate/inhibitor of rIucD. The importance of the
presence of an unmodified ␧-amino group gains further support from the observation
that neither ␧-N-methyl-L-lysine nor ␧-N-acetyl-
L-lysine is capable of being hydrox-
ylated by the enzyme. However, methylation of
L
-lysine at the Ͱ-amino function
(A) allows the analog to promote NADPH-oxidation as well as to serve as a
substrate of lysine : N6-hydroxylase, although only 20% as effective as the preferred
substrate. These observations show that a free, unmodified Ͱ-amino group is not
absolutely necessary for substrate recognition by rIucD, whereas the modification
of the ␧-amino function leads to a total loss of the ability to serve as a substrate,
F
IG. 1. Structural features inherent in L-lysine. A and D represent the positively charged amino group
at the Ͱ and ␧ carbon, respectively. B symbolizes the negatively charged Ͱ-carboxyl function. Cn represents
the hydrophobic side chain composed of four methylene groups (n ϭ 4).

LYSINE : N6-HYDROXYLASE
407
TABLE 1
Influence of Various Effectors on the Enzymatic Activity of rIucD
Lysine : N6-
hydroxylase
activity (%)c
Structural featuresb
Effectora
A
B
Cn
D
ϩL
-Lysine
ϪL
-Lysine
None
(S)-2-Aminoethyl-
Ͱ-N-Methyl- -lysine
-Lysine
␧-N-Methyl-
␧-N-Acetyl-
DL-Ͱ-Aminocaprylic acid
-Ethionine
trans-4,5-Dehydro-
6-Aminocaproic acid
100
105
103
100
93
0
100
21
0
0
0
0
nd
0
0
0
0
0
0
0
0
0
nd
0
L
-cysteine
NHϩ3
NHϩ2 CH3
NHϩ3
NHϩ3
NHϩ3
NHϩ3
NHϩ3
NHϩ3
H
COϪ2
COϪ2
COϪ2
COϪ2
COϪ2
COϪ2
COϪ2
COϪ2
COϪ2
CONH2
CO2CH3
H
4d
4
4
4
4
NHϩ3
NHϩ3
L
D
NHϩ3
L
-lysine
-lysine
NHϩ2 CH3
NHCOCH3
CH3
L
100
4
100 (2 m
M
M
)
)
L
4e
4f
4
CH3
85
100
100
L
-lysine
NHϩ3
NHϩ3
L
-Lysineamide
NHϩ3
NHϩ3
NHϩ3
NHϩ3
NHϩ3
NHϩ3
NHϩ3
NHϩ3
NHϩ3
4
4
NHϩ3
100 (2 m
100
L
-Lysinemethylester
NHϩ3
1,6-Diaminohexane
(S)-Aminoethyl-DL-homocysteine
Ͱ,␧-Diaminopimelic acid
4
NHϩ3
100
80
95
100
100
100
COϪ2
COϪ2
COϪ2
COϪ2
COϪ2
COϪ2
5b
3
NHϩ3
CH(NHϩ3 )(COϪ2 )
NHϩ3
L
-Ornithine
-Arginine
3
L
3
NHC(NH2)ϩ2
CICH
NHϩ3
S-Propargyl-
L-cysteine
3a
2
2,4-Diaminobutyric acid
100 (2 m
M
)
a All compounds listed below were purchased from Sigma Chem. Co., except Ͱ-N-methyl-
L-lysine, trans-4,5-dehydro-
L-lysine, (S)-aminoethyl-DL-homocysteine, and S-propargyl-L-cysteine (see Materials and Methods).
b The structural features of the effectors (A, B, Cn , D) are illustrated in Fig. 1. The subscript n indicates the number
of methylene groups present in the hydrophobic side chain of these compounds.
c The enzymatic activity of rIucD was determined as described under Materials and Methods. The concentration of
the effectors was 1 m
M unless further specified.
d CͲH2 replaced by sulfur.
e CͳH2 replaced by sulfur.
f CͲH2UCͳH2 replaced by CͲHuCͳH (trans-conformation).
suggesting the detrimental nature of any type of modification at this location. Even
the presence of a carboxyl group in the vicinity of the ␧-amino group, as in the
case of Ͱ,␧-diaminopimelic acid, renders the compound incapable of interacting
with the enzyme.
Alteration in the site B of the substrate, either by deletion of the carboxy group
(as in 1,6-diaminohexane) or by its modification (as in lysine amide or lysine meth-
ylester), results in a total loss of the ability to serve as a substrate.
The importance of the flexible hydrophobic backbone consisting of four methyl-
ene groups is documented by the observation that analogs with identical structural
features (A, B, and D) but a change in the number of methylene groups in the
side chain (as in
L
-2,4-diaminobutyric acid, -ornithine, and (S)-2-aminoethyl-DL-
L
homocysteine) do not serve as substrates/inhibitors of rIucD to any significant
extent. Hence, a C4 chain length appears to be essential for a compound to function
as a substrate, with a replacement of the methylene group at the Ͳ-position by
sulfur being tolerated ((S)-2-aminoethyl-
L-cysteine). Furthermore, a decrease in
the flexibility of the hydrophobic side chain, as in the case of trans-4,5-dehydro-
L-

408
MARRONE ET AL.
lysine, leads to the total loss of the ability both to promote NADPH-oxidation and
to undergo hydroxylation. Moreover, none of the other compounds (benzamidine,
p-aminobenzamidine, and tranexamic acid) was utilized as a substrate by lysine : N6-
hydroxylase, providing further support for the essential role of a flexible hydropho-
bic backbone. It is pertinent to note that the small extent of inhibition observed in
a few instances would appear to fall in the range of the experimental error of the
assays, based on either lysine : N6-hydroxlation or NADPH-oxidation.
L
-Norleucine, a compound characterized by its C4 hydrophobic side chain, was
found to be the best effector of rIucD, inhibiting both NADPH-oxidation and
lysine : N6-hydroxylation.
-Norleucine analogs, such as -Ͱ-aminobutyric acid, -
L
L
L
norvaline, and DL-Ͱ-aminocaprylic acid, which differ in the number of methylene
groups present in the hydrophobic side chain, failed to exert any adverse effect on
the reaction mediated by rIucD (Tables 1 and 2), providing further evidence for
the stringent specificity of the enzyme for a C4 backbone.
N-Alkylamides of -Norleucine and Analogs
L
Since a C4 hydrophobic side chain (Cn) plays a vital role in the binding of
substrate(s) or effector(s) to lysine : N6-hydroxylase, experiments were undertaken
to determine if a provision of alkyl side chains on the carboxyl side of norleucine
and its analogs would result in a promotion of their ability to serve as substrates/
inhibitors of the enzyme. Hence, alkyl amides of norleucine, norvaline, and Ͱ-
aminobutyric acid were synthesized, and their influence on rIucD-mediated ly-
sine : N6-hydroxylation was investigated. Since little can be found in the literature
on the synthesis of methyl and ethylamides of hydrophobic amino acids, a few
general comments are appropriate concerning the limitations and problems in
achieving the synthesis of the N-alkyl amides prior to the discussion of their effect
TABLE 2
Influence of Norleucine and Its Analogs on the Catalytic Function of rIucDa
Relative activity of
Relative activity of
NADPH oxidation (%)
Effector
lysine : N6-hydroxylation (%)
None
100
50
100
100
75
100
87
100
78
100
18
100
100
nd
100
82
100
35
L
L
L
(ϩ)-Norleucine
-Norvaline
-Ͱ-Aminobutyric acid
DL-Norleucine
-Norleucine methylester
L
HAbuNHEt
HNvaNHMe
HNleNHMe
HNleNH2
100
62
a The relative activity of lysine : N6-hydroxylation and NADPH-oxidation was determined
as described under Materials and Methods. The concentration of the desired compound was
1 m
M M in NADPH-oxidation experiments.
in the case of lysine : N6-hydroxylation, and 5 m

LYSINE : N6-HYDROXYLASE
409
on the enzyme. These are: (i) the standard procedure(s) for peptide or amide
synthesis employing t-Boc amino acid, DCC, Et3N, and the desired alkyl amine
hydrochloride in CH2Cl2 was ineffective, primarily due to the insolubility of the
amine hydrochloride. Repetition of the reaction in a heterogeneous solvent phase
with the t-Boc amino acid and DCC in CH2Cl2 as well as the free base of the amine
(generated by treatment with KOH) in the aqueous phase failed to produce the
desired product in reasonable yields. (ii) An alternative procedure involving the use
of a water-soluble carbodiimide, 1-cyclohexyl-3-(2-morpholinoethyl) carbodiimide
(19), for achieving the synthesis of alkyl amide derivatives of amino acids was
employed, and the reaction was performed in a medium of acetonitrile (20). How-
ever, the yield of the product was found to be low, primarily due to the limited
solubility of the reactants in acetonitrile. The procedure which resulted in good
yields of the desired amides comprises three steps (Scheme 1): (i) the conversion
of the t-Boc amino acid to its N-hydroxy succinimide ester, a highly valuable interme-
diate in peptide synthesis (11, 12, 21, 22); (ii) the conversion of the t-Boc amino
acid N-hydroxysuccinimide ester to the amide; and (iii) the removal of the t-Boc
protecting group using HCl in dioxane rather than TFA, which because of water
contamination, was found to be less effective as a deprotectant.
The influence of
rIucD-mediated NADPH-oxidation is shown in Table 2 and Fig. 2.
represents the most potent inhibitor among the compounds examined in this study.
Both -norleucine methylamide (HNleNHMe) and -Ͱ-aminobutyric acid ethylam-
ide (HAbuNHEt) served as moderate inhibitors of the enzyme, whereas -norvaline
methylamide (HNvaNHMe) did not exhibit any influence on rIucD-mediated
L
-norleucine, its analogs, and their alkylamide derivatives on
L-Norleucine
L
L
L
NADPH-oxidation. With the exception of -norleucineamide (NHleNH2), a similar
L
pattern of results was obtained when the enzyme-mediated process of lysine : N6-
hydroxylation was monitored in the presence of these compounds (Table 2).
DISCUSSION
A comprehensive study of the ability of a number of compounds to serve as
substrates for lysine : N6-hydroxylase has revealed that the enzyme is stringently
specific in its function.
L
-Lysine and (S)-2-aminoethyl- -cysteine are the only known
L
compounds that are hydroxylated by the enzyme. Besides its unique substrate
specificity, lysine : N6-hydroxylase also exhibits exclusive preference for its cofactors,
NADPH and FAD. The enzyme-mediated processes of NADPH-oxidation and
lysine : N6-hydroxylation appear to be coupled (7, 23). However, the coupling of
the two above-mentioned processes catalyzed by rIucD does not appear to be as
rigid as that noted in the case of other flavin-dependent monooxygenases. For
example, in the case of p-hydroxybenzoate hydroxylase, the rate of NADPH oxida-
tion is stimulated by a factor of 104–105 in the presence of substrate, p-hydroxyben-
zoate (24). In contrast, the substrate, -lysine, elicits only a moderate, approximately
L
10- to 15-fold increase in the rate of NADPH-oxidation mediated by lysine : N6-
1
hydroxylase (8). Furthermore, the low kcat value (0.1–0.2 sϪ ) of lysine : N6-hydroxy-
1
lase relative to that of p-hydroxybenzoate hydroxylase, approximately 45 sϪ (25),

410
MARRONE ET AL.
S
CHEME 1. Steps in the synthesis of alkylamides of nonaromatic hydrophobic amino acids. (1) Conver-
sion of the t-Boc protected amino acids ( -norleucine, -norvaline, and -Ͱ-aminobutyric acid) into their
L
L
L
N-hydroxysuccinimide esters. (2) Conversion of the N-hydroxysuccinimide esters into alkylamides. (3)
Removal of the protecting group.
would appear to correlate with the differences in the substrate-induced rate enhance-
ment of NADPH oxidation in the two enzyme systems. Such vast differences in
the rates of NADPH oxidation and the kcat values in these two enzyme systems do
not appear to be due to a change in the nature of the hydroxylating species, since
4a-peroxyflavin has been shown to be the oxygenating species (26) in a number of
flavin-dependent monooxygenases. These include such diverse systems as p-

LYSINE : N6-HYDROXYLASE
411
F
IG. 2. Influence of synthetic analogs of
lyzed by rIucD. The typical assay mixture consisted of FAD (40 Ȑ
phosphate (100 m , pH 7.0), as well as rIucD (370 n ), and the reaction was monitored at 340 nm,
23ЊC. (1) No Lys, (2) Lys, (3) Lys ϩ -norleucine, (4) Lys ϩ HNleNHMe, (5) Lys ϩ HNleNH2 , (6)
Lys ϩ HAbuNHEt. Norleucine or its analog (5 m ) was added at t ϭ 0. Lysine was added after t ϭ
50 s, as indicated by the arrow. HNvaNHMe, -norleucine methylester, -norvaline, and -Ͱ-aminobutyric
acid were found to have no inhibitory effect.
L
-norleucine and precursors on the NADPH-oxidation cata-
M
), NADPH (200 Ȑ ), and potassium
M
M
M
L
M
L
L
L
hydroxybenzoate hydroxylase (27, 28), phenol hydroxylase (29–32), melilotate hy-
droxylase (33), bacterial luciferase (34), 2-methyl-3-hydroxypyridine-5-carboxylate
oxygenase (35), and cyclohexanone monooxygenase (36). Depending on the nature
of the substrate, the 4a-peroxyflavin intermediate can function either as an elec-
trophile or a nucleophile in facilitating the transfer of its distal oxygen to the
substrate. The former situation prevails in the case of monooxygenases like p-
hydroxybenzoate hydroxylase (37), while the latter occurs in the oxygenation of
cyclohexanone (36). These mechanisms are illustrated in Scheme 2. In light of these
observations, and those obtained with model systems (38, 39), it is not unreasonable
to expect the 4a-peroxyflavin intermediate to function as an electrophile in the
lysine : N6-hydroxylation process. This view derives additional support from a previ-
ous report which provides evidence for the exclusion of a cytochrome P450 and/
or a metalloenzyme system in the lysine : N6-hydroxylation mediated by the wild-
type IucD (5).
The above-mentioned considerations raise a question concerning the basis for
the low rate of enhancement of NADPH-oxidation and the low kcat value for the
reaction catalyzed by lysine : N6-hydroxylase. The rate enhancement of enzyme-

412
MARRONE ET AL.
S
CHEME 2. Mechanism of hydroxylation mediated by FAD containing monooxygenases. (A) FAD-
4a-OOH as an electrophile in the hydroxylation of p-hydroxybenzoate mediated by p-hydroxybenzoate
hydroxylase. (B) FAD-4a-OOH as a nucleophile in the oxygenation of ketones by ketone monooxy-
genases.
catalyzed reaction relative to that in solution can be achieved by lowering the
activation barrier of the process either through stabilization of transition state and/
or by increasing the energy content or the reactivity of the substrate(s). Serine
proteases serve as typical examples of the former route (40, 41) while the 4a-
peroxyflavin-dependent hydroxylation of aromatic substrates exemplifies the latter
option (42, 43). Furthermore, similar catalytic efficiency in the hydroxylation of
substrates with differing reactivities can be achieved by appropriate adjustment of
the dielectric constant of the active site and of the proximity of the reactants (42,
43). These considerations raise the possibility that the low rate of NADPH-oxidation
by the substrate and the consequent low kcat value of the rIucD-mediated N-
hydroxylation process may be a reflection of the low reactivity of lysine in its
extended conformation. This view derives support from the following observations:
(i) trans-4,5-dehydro- -lysine, which is characterized by a nonflexible, extended
L
conformation, is not a substrate of the enzyme. This situation is in distinct contrast
to that noted in the case of trypsin, which utilizes the ester and amides of the trans

LYSINE : N6-HYDROXYLASE
413
S
CHEME 3. Proposed mechanism for rIucD-mediated lysine : N6-hydroxylation. The interaction of 2-
aminocaprolactam with 4a-peroxyflavin species is shown. The general mechanism indicating the sequence
of events in rIucD-mediated lysine : N6-hydroxylation has been presented in a previous report (8).
4,5-dehydro- -lysine as substrates, but not those of the cis-isomer (44). (ii) Studies
L
with model 4a-peroxyflavins have shown that the reaction with primary amines is
slow relative to that of secondary and tertiary amines, and may result in the destruc-
tion of the flavin cofactor (39). In light of these considerations we propose that
lysine : N6-hydroxylase participates in the enhancement of the reactivity of
-lysine
L
by its cyclization to 2-aminocaprolactam, a transient intermediate which reacts
rapidly with the 4a-peroxyflavin species, to yield N6-hydroxylysine via the hydroxa-
mate intermediate (Scheme 3). Unorthodox as it may appear at first glance, it is
pertinent to note that although the N-hydroxylation process is an obligatory event
in the biosynthesis of hydroxamate siderophores, there appears to be no precise
order with regard to the sequence of events leading to the production of hydroxa-
mate constituents of the siderophores. In the case of ferrichrome (45), rhodotorulic
acid (46, 47), and hadacidin (48), available evidence indicates that N-hydroxylation
precedes the acylation reaction. In contrast, the sequence of events appears to be
reversed, i.e., acylation reaction occurs prior to N-hydroxylation, in the biosynthesis

414
MARRONE ET AL.
of mycobactins (49, 50) and pulcherriminic acid (51, 52). Indeed, the production
of an acyl intermediate has been noted to occur during the production of mycobac-
tin (49).
The next logical question concerns the mechanism involved in the conversion of
L-lysine to its lactam derivative, a reaction that involves condensation of the Ͱ-
carboxylic function with the ␧-amino group of the amino acid. Such amide or peptide
bond formation, a nonribosomal event in this case, could be expected to involve
ATP-dependent activation of the carboxyl function (53, 54). However, since ly-
sine : N6-hydroxylase does not require ATP for its function, the bond formation
reaction would have to be a consequence of the lysine binding in such a way as to
sterically compress the Ͱ-carboxylic and ␧-amino functions in close proximity of
each other. The importance of binding energy in the compensation for the loss in
entropy has been elegantly reviewed (55). Furthermore, the vast enhancement in
the rates of bond formation that ensues due to the restricted rotation and stereo
population control has been documented (56, 57).
It should be emphasized that these proposals, attractive as they may seem, fall
in the realm of conjecture at the present time in view of the lack of concrete
evidence in their support. Experiments to generate 2-amino caprolactam from -
L
lysine methyl ester for an assessment of its ability to function as a substrate have
not been successful, primarily in view of its facile conversion to lysine under the
conditions of the assay. The reactivity of the free hydroxylamine function has been
well documented (58–63), and it is toxic by virtue of its ability to interact with a
variety of active acyl intermediates essential for cellular function. The participation
of a lactam in the N-hydroxylation reaction would result in the formation of a
hydroxamate derivative, which would be less toxic than its parent hydroxylamine
function. A mechanism for the N-hydroxylation process, relying on the initial con-
version of the substrate to a lactam, may serve as an effective means for minimizing
the accumulation of the undesirable hydroxylamine function in addition to that
available by the posthydroxylation acylation step in the biosynthetic pathway of
aerobactin (2–4). An added benefit of such a lactam intermediate would be not
only that the reactivity of the compound is enhanced, but also that it minimizes
the possibility of the destruction of the cofactor that serves as the source of the
oxygenating species (39).
Both wild-type and recombinant forms of lysine : N6-hydroxylase are inhibited
by
tyric acid, do not serve as inhibitors, indicating the importance of the C4 hydrophobic
side chain for an effective interaction with the enzyme. N-Alkyl amides of
L
-norleucine. Lower homologs of the amino acid,
L
-norvaline and
L
-Ͱ-aminobu-
L
-
norleucine and analogs were synthesized and their ability to serve as either substrates
or inhibitors of the enzyme was investigated. This approach was based on the
premise that the alkyl side chains on both sides of the amide bond could provide
the structural motif needed for eliciting the conformation essential for the catalytic
function of the enzyme. However, none of these compounds was able to serve as
a substrate for rIucD. Both norleucine methylamide and the Ͱ-aminobutyric acid
ethylamide appeared to function as moderate inhibitors of the enzyme, whereas
norvaline methylamide was devoid of this property.
In conclusion, we reiterate that the proposals regarding the involvement of a

LYSINE : N6-HYDROXYLASE
415
lactam intermediate in the catalytic mechanism of rIucD are tentative and experi-
ments are being pursued to their test validity.
ACKNOWLEDGMENTS
The authors are indebted to Mr. J. Gaspar for the synthesis of (S)-2-aminoethyl-DL-homocysteine,
and Dr. G. I. Dmitrienko for his advice in the synthesis of this compound and S-propargyl-L-cysteine.
The technical assistance of L. Taylor in ESMS analyses is greatly appreciated. The research was supported
by the National Sciences and Engineering Research Council of Canada.
REFERENCES
1. GIBSON, F., AND
2. GROSS, R., ENGELBRECHT, F., AND
ORENZO, V., BINDEREIF, A., PAW, B. H., AND
M
AGRATH, D. J. (1969) Biochim. Biophys. Acta 192, 175–184.
B
RAUN, V. (1985) Mol. Gen. Genet. 201, 204–212.
3. DE
L
N
EILANDS, J. B. (1986) J. Bacteriol. 165, 570–578.
4. VISWANATHA, T., SZCZEPAN, E. W., AND
M
URRAY, G. J. (1987) in Iron Transport in Microbes,
Plants and Animals (Neilands, J. B., Van der Helm, D., and Winkelmann, G., Eds.), pp. 117–132,
Springer-Verlag, New York.
5. GOH, C. J., SZCZEPAN, E. W., WRIGHT, G., MENHART, N., HONEK, J. F., AND
VISWANATHAN, T.
(1989) Bioorg. Chem. 17, 13–27.
6. THARIATH, A. M., SOCHA, D., VALVANO, M. A., AND
589–596.
V
ISWANATHA, T. (1993) J. Bacteriol. 175,
7. THARIATH, A. M., FATUM, K. L., VALVANO, M. A., AND
Acta 1203, 27–35.
VISWANATHA, T. (1993) Biochem. Biophys.
8. MARRONE, L., BEECROFT, M., AND
VISWANATHA, T. (1996) Bioorg. Chem. 24, 304–317.
9. PARNIAK, M. A., JACKSON, G. E. D., MURRAY, G. J., AND
VISWANATHA, T. (1979) Biochim. Biophys.
Acta 569, 99–108.
10. TOMLINSON, G., CRUICKSHANK, W. H., AND
11. KONIG, W., AND EIGER, R. (1969) Liebigs Ann. Chem. 727, 125–129.
12. WEYGAND, F., HOFFMANN, D., ND WUNSCH, E. (1966) Z. Naturforsch. 21b, 426–428.
VISWANATHA, T. (1971) Anal. Biochem. 44, 670–679.
G
¨
A
¨
13. GREENSTEIN, J. P., AND
New York.
WINITZ, M. (1961) in Chemistry of the Amino Acids, Vol. 2, p. 1188, Wiley,
14. DU
V
IGNEAUD, V., LORING, H. S., AND
C
RAFT, H. A. (1934) J. Biol. Chem. 105, 481.
ARCO, C., MUNDOVI, B., AND ZZONE, G. F. (1953) Experientia XI, 61–62.
ASTELLINO, F. J. (1990) J. Biol. Chem. 265, 5482–5486.
HAW, E. (1965) J. Biol. Chem. 240, 1579–1585.
CHWERT, G. W. (1950) Chem. Rev. 46, 69–153.
19. SHEEHAN, J. C., CRUICKSHANK, P. A., AND OSHART, G. L. (1961) J. Org. Chem. 26, 2525–2528.
20. SHEEHAN, J. C., AND LAVKA, J. J. (1956) J. Org. Chem. 21, 439–441.
21. ANDERSON, G. W., ZIMMERMAN, J. E., AND ALLAHAN, F. M. (1963) J. Am. Chem. Soc. 85, 3039.
22. ANDERSON, G. W., ZIMMERMAN, J. E., AND ALLAHAN, F. M. (1964) J. Am. Chem. Soc. 86, 1839–1842.
23. THARIATH, A. M., VALVANO, M. A., AND ISWANATHA, T. (1994) in The Development of Iron
15. CAVALLINI, D., D
16. SEHL, L. C., AND
E
M
A
C
17. MARES-GUIA, M., AND
18. NEURATH, H., AND
S
S
B
H
C
C
V
Chelators for Clinical Use (Bergeron, R. J., and Brittenham, G. M., Eds.), pp. 169–186, CRC Press,
Boca Raton.
24. HUSAIN, M., AND
25. ENTSCH, B., PALFEY, B. A., BALLOU, D. P., AND
26. MASSEY, V. (1994) J. Biol. Chem. 269, 22459–22462.
27. ENTSCH, B., AND ALLOU, D. P. (1989) Biochim. Biophys. Acta 999, 313–322.
28. SCHOPFER, L. M., WESSIAK, A., AND ASSEY, V. (1991) J. Biol. Chem. 266, 13080–13085.
29. DETMER, K., AND ASSEY, V. (1985) J. Biol. Chem. 260, 5998–6005.
30. TAYLOR, M. G., AND ASSEY, V. (1990) J. Biol. Chem. 265, 13687–13694.
M
ASSEY, V. (1979) J. Biol. Chem. 254, 6657–6666.
M
ASSEY, V. (1991) J. Biol. Chem. 266, 17341–17349.
B
M
M
M

416
MARRONE ET AL.
31. TAYLOR, M. G., AND
MASSEY, V. (1991) J. Biol. Chem. 266, 8291–8301.
32. MAEDA-YORITA, K., AND
MASSEY, V. (1993) J. Biol. Chem. 268, 4134–4144.
33. SCHOPFER, L. M., AND ASSEY, V. (1980) J. Biol. Chem. 255, 5355–5363.
M
34. HASTINGS, J. W., BALNY, C., LEPEUCH, C., AND DOUZOU, P. (1973) Proc. Natl. Acad. Sci. USA
70, 3468–3472.
35. BRISSETTE, P., BALLOU, D. P., AND
D. E., and McCormick, D. B., Eds.), pp. 573–576, de Gruyter, Berlin.
36. RYERSON, C. C., BALLOU, D. P., AND ALSH, C. (1982) Biochemistry 21, 2644–2655.
37. PALFEY, B. A., ENTSCH, B., BALLOU, D. P., AND ASSEY, V. (1991) Biochemistry 33, 1545–1554.
MASSEY, V. (1987) in Flavins and Flavoproteins (Edmondson,
W
M
38. BALL, S., AND
39. BALL, S., AND
B
B
RUICE, T. C. (1979) J. Am. Chem. Soc. 101, 4017–4019.
RUICE, T. C. (1980) J. Am. Chem. Soc. 102, 6498–6503.
40. KRAUT, J. (1988) Science 242, 533–540.
41. WARSHALL, A., PAPAZYAN, A., AND OLLMAN, P. A. (1995) Science 269, 102–106.
42. VERVOORT, J., AND IETJENS, I. M. C. M. (1996) Biochem. Soc. Trans. 24, 127–130.
K
R
43. VERVOORT, J., RIETJENS, I. M. C. M., VAN
B
ERKEL, W. J. H., AND
VEEGER, C. (1992) Eur. J.
Biochem. 206, 479–484.
44. MIZUSAKI, K., SUGAHARA, Y., TSUNEMATSU, H., AND
MAKISUMI, S. (1986) J. Biochem. 100, 21–25.
45. EMERY, T. F. (1966) Biochemistry 5, 3694–3701.
46. ATKIN, C. L., AND
47. AKERS, H. A., LLINAS, M., AND
48. STEVENS, R. L., AND MERY, T. F. (1966) Biochemistry 5, 74–81.
NEILANDS, J. B. (1968) Biochemistry 7, 3734–3739.
N
EILANDS, J. B. (1972) Biochemistry 11, 2283–2291.
E
49. TATESON, J. E. (1970) Biochem. J. 118, 747–753.
50. SNOW, G. A. (1970) Bacteriol. Rev. 34, 99–125.
51. MACDONALD, J. C. (1965) Biochem. J. 96, 533–538.
52. UFFEN, R. L., AND
C
ANALE-PAROLA, E. (1972) J. Bacteriol. 111, 86–93.
53. LIPMANN, F. (1973) Acc. Chem. Res. 6, 361–367.
54. LIPMANN, F. (1982) in Peptide Antibiotics: Biosynthesis and Function (Kleinkauf, H., and von
Dohren, H., Eds.), pp. 23–45, de Gruyter, Berlin.
¨
55. JENCKS, W. P. (1987) in Cold Spring Harbor Symposia on Quantitative Biology, Vol. LII, pp. 65–73,
Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
56. KARLE, J. M., AND
57. MILSTEIN, S., AND
K
ARLE, I. L. (1972) J. Am. Chem. Soc. 94, 9182–9189.
C
OHEN, L. A. (1972) J. Am. Chem. Soc. 94, 9158–9165.
58. JENCKS, W. P. (1958) J. Am. Chem. Soc. 80, 4581–4584.
59. JENCKS, W. P. (1958) J. Am. Chem. Soc. 80, 4585–4588.
60. JENCKS, W. P., AND
61. EDWARDS, J. O., AND
62. AUBORT, J. D., AND
63. TOMLINSON, G., GAUDIN, J. E., AND
C
ARRIUOLO, J. (1960) J. Am. Chem. Soc. 82, 1778–1786.
EARSON, R. G. (1962) J. Am. Chem. Soc. 84, 16–24.
UDSON, R. F. (1970) Chem. Commun. 937–938.
ISWANATHA, T. (1973) Can. J. Biochem. 51, 764–771.
P
H
V