Suppression of linear side products by macromolecular crowding in nonribosomal enterobactin biosynthesis

Article abstract of DOI:10.1021/ol7030153

Nonribosomal enterobactin synthetase of Escherichia coli was found to prematurely release a large amount of linear precursors in an in vitro reconstitution. However, these side products are suppressed to negligible levels by polymeric cosolvents that create macromolecular crowding, a prominent feature of the intracellular environment. These findings show that macromolecular crowding is essential to normal functioning of the nonribosomal peptide synthetase and suggest that it may be crucial to btotechnological utilization of similar enzyme systems.

Full text of DOI:10.1021/ol7030153

ORGANIC
LETTERS
2008
Vol. 10, No. 4
649-652
Suppression of Linear Side Products by
Macromolecular Crowding in
Nonribosomal Enterobactin Biosynthesis
Zu-Feng Guo, Ming Jiang, Suilan Zheng, and Zhihong Guo*
Department of Chemistry, Center for Cancer Research, The Hong Kong UniVersity of
Science and Technology, Clear Water Bay, Kowloon, Hong Kong SAR, China
chguo@ust.hk
Received December 14, 2007
ABSTRACT
Nonribosomal enterobactin synthetase of Escherichia coli was found to prematurely release a large amount of linear precursors in an in vitro
reconstitution. However, these side products are suppressed to negligible levels by polymeric cosolvents that create macromolecular crowding,
a prominent feature of the intracellular environment. These findings show that macromolecular crowding is essential to normal functioning of
the nonribosomal peptide synthetase and suggest that it may be crucial to biotechnological utilization of similar enzyme systems.
Nonribosomal peptide synthetases (NRPSs) are large, modu-
lar proteins responsible for biosynthesis of many medicinally
important peptide natural products.1 There is intense interest
in using these synthetases in biomedical research because
of their ability to assemble complex natural peptide products
and analogues from simple amino acids. For example, the
cyclosporine synthetase has been used in isotope-labeling
of cyclosporine A and synthesis of various analogues of the
natural product.2 Enniatin synthetase has also been used for
analogue synthesis.3 However, the cyclosporine synthetase
has been found to prematurely release precursors in the cell-
free reconstitution.4 In addition, the stability and synthetic
efficiency of the synthetase are less than desirable.2 Similar
problems were found in cell-free reconstitution of dozens
of other NRPSs.5
As a step to overcome these problems, we chose entero-
bactin synthetase as a model to investigate the factors
influencing the activities of these multifunctional enzymes.
We noticed that the previous cell-free reconstitution of
NRPSs was exclusively carried out in dilute solution with a
(2) (a) Billich, A.; Zocher, R. J. Biol. Chem. 1987, 262, 17258. (b) Lawen,
A.; Traber, R.; Geyl, D.; Zocher, R.; Kleinkauf, H. J. Antibiot. 1989, 42,
1283. (c) Traber, R.; Hofmann, H.; Kobel, H. J. Antibiot. (Tokyo) 1989,
42, 591. (d) Kleinkauf, H.; Dittmann, J.; Lawen, A.; Biomed. Biochim. Acta
1991, 50, S219. (e) Lawen, A.; Traber, R.; Geyl, D. J. Biol. Chem. 1991,
266, 15567. (f) Lawen, A.; Dittmann, J.; Schmidt, B.; Riesner, D.; Kleinkauf,
H. Biochim. 1992, 74, 511. (g) Hensens, O. D.; White, R. F.; Goegelman,
R. T.; Inamine, E. S.; Patchett, A. A. J. Antibiot. (Tokyo) 1992, 45, 133.
(h) Lawen, A.; Traber, R.; Reuille, R.; Ponelle, M. Biochem. J. 1994, 300,
395.
(3) (a) Krause, M.; Lindemann, A.; Glinski, M.; Hornbogen, T.; Bonse,
G.; Jeschke, P.; Thielking, G.; Gau, W.; Kleinkauf, H.; Zocher, R. J.
Antibiot. (Tokyo) 2001, 54, 797. (b) Feifel, S. C.; Schmiederer, T.;
Hornbogen, T.; Berg, H.; Su¨ssmuth, R. D.; Zocher, R. ChemBioChem 2007,
8, 1767.
(1) (a) Finking, R.; Marahiel, M. A. Annu. ReV. Microbiol. 2004, 58,
453. (b) Sieber, S. A.; Marahiel, M. A. Chem. ReV. 2005, 105. (c) Fischbach,
M. A.; Walsh, C. T. Chem. ReV. 2006, 106, 3468.
(4) Zocher, R.; Nihira, T.; Paul, E.; Madry, N.; Peeters, H.; Kleinkauf,
H.; Keller, U. Biochemistry 1986, 25, 550.
10.1021/ol7030153 CCC: $40.75
© 2008 American Chemical Society
Published on Web 01/23/2008

Scheme 1. Non-ribosomal Enterobactin Biosynthesis and the Premature Release of the Precursorsa
a A, adenylation domain; C, condensation domain; ArCP, aryl carrier protein; PCP, peptidyl carrier protein; TE, thioesterase domain;
vertical squiggle line, phosphopantetheinyl.
low macromolecular content. This in vitro condition differs
significantly from the highly crowded intracellular environ-
ment,6 where biomacromolecules occupy a large proportion
of volume and cause the excluded volume effect or macro-
molecular crowding.7 We created in vitro crowding condi-
tions to mimic the intracellular environment and studied the
effects of macromolecular crowding on the activities of the
enterobactin NRPS. It was found that macromolecular
crowding suppresses premature release of precursors from
the synthetase. This finding provides the first experimental
evidence for the importance of the crowded intracellular
environment to catalytic properties of a nonribosomal peptide
synthetase.8
antibodies to be 1.4, 0.35, and 0.65 µM, respectively, in cells
induced for enterobactin synthesis by iron starvation (Figure
S2, Supporting Information). These proteins were prepared
through recombinant expression according to the reported
methods.10 The half-saturation concentrations of the recom-
binant EntB and EntE and the maximum reaction rate of the
synthetase are consistent with the previously reported values11
(Figures S6 and S7, Supporting Information). However, the
synthetase containing EntB, EntE, and EntF at their physi-
ological concentrations in a dilute buffer produced a large
amount of linear side products, which include the aminolytic
release product of DHB-Ser trimer by Tris base (TT) and
DHB-serine monomer (M), dimer (D), and trimer (T) (Figure
1A).12 The correct enterobactin product was only 44% of
the total turnover products. This result demonstrates that the
enterobactin NRPS is not different from other NRPSs in
being prone to release the precursors prematurely.4,5 This
conclusion does not contradict a previous investigation in
which no linear side products were found in the initial 10
min at a low EntF concentration and high EntB/EntF ratio.11
Actually, the linear products were formed with a mass
The enterobactin synthetase is a two-module NRPS
responsible for siderophore biosynthesis in E. coli under iron-
deficient conditions (Scheme 1).9 Its protein components
EntB, EntE, and EntF were determined by rabbit polyclonal
(5) (a) von Do¨hren, H.; Keller, U.; Vater, J.; Zocher, R. Chem. ReV.
1997, 97, 2675. (b) Kleinkauf, H.; von Do¨hren, H. Prog. Drug. Res. 1997,
48, 27.
(6) (a) Fulton, A. B. Cell 1982, 30, 345. (b) Zimmerman, S. B.; Trach,
S. O. J. Mol. Biol. 1991, 222, 599.
(7) (a) Zimmerman, S. B.; Minton, A. P. Annu. ReV. Biophys. Biomol.
Struct. 1993, 22, 27. (b) Minton, A. P. J. Biol. Chem. 2001, 276, 10577.
(c) Ellis, R. J. TIBS 2001, 26, 597. (d) Minton, A. P. J. Pharm. Sci. 2005,
94, 1668. (d) Minton, A. P. J. Cell Sci. 2006, 119, 2863.
(8) Nonionic small molecule detergents have been reported to enhance
the macrocyclization activity of the putative terminal thioesterase domain
of tyrocidine synthetase (see: Yeh, E.; Lin, H.; Clugston, S. L.; Kohli, R.
M.; Walsh, C. T. Chem. Biol. 2004, 11, 1573). This detergent effect is
different from the crowding effects discussed here.
(10) (a) Rusnak, F.; Faraci, W. S.; Walsh, C. T. Biochemistry 1989, 28,
6827. (b) Reichert, J.; Sakaitani, M.; Walsh, C. T. Protein Sci. 1992, 1,
549. (c) Gehring, A. M.; Bradley, K. A.; Walsh, C. T. Biochemistry 1997,
36, 8495.
(11) Gehring, A. M.; Mori, I.; Walsh, C. T. Biochemistry 1998, 37, 2648.
(12) The synthetase components were purified to homogeneity as shown
in Figure S1 (Supporting Information). EntB and EntF were correctly
pantotetheinylated as shown by MALDI-ToF spectra in Figures S3 and S4
(Supporting Information). The linear side products and enterobactin were
identified by ESI-ToF mass spectrometry as shown in Figure S5 (Supporting
Information).
(9) (a) Walsh, C. T.; Liu, J.; Rusnak, F.; Sakaitani, M. Chem. ReV. 1990,
90, 1105. (b) Crosa, J. H.; Walsh, C. T. Microbiol. Mol. Biol. ReV. 2002,
66, 223.
650
Org. Lett., Vol. 10, No. 4, 2008

in the turnover products. In comparison, the enterobactin
percentage increase was only 6% in 30% sucrose solution
with a comparable polarity to 30% Ficoll 70 (Figure 1A,B),
indicating that the side-product suppressing effect of Ficoll
was not due to polarity change caused by the crowding agent.
Besides Ficoll 70, three other commonly used crowding
agents, namely dextran 75, bovine serum albumin (BSA),
and poly(ethylene glycol) (PEG6000), were also found to
suppress the linear side products and increase the enterobactin
percentage in the turnover products to a similar extent (Figure
1A,B). Noticeably, like Ficoll 70, all of these crowding
agents seems to only reduce the production of the linear side
products; production of the normal enterobactin product of
the synthetase is minimally affected (Figure 1A). This side-
product suppressing effect from all of the structurally distinct
crowders can only be caused by the common interaction of
all of the crowding agents with the synthetase, which is the
nonspecific repulsion interactionsthe only influence exerted
by macromolecular crowding.7 This effect, therefore, is a
true macromolecular crowding effect. As shown in Figure
1B, PEG6000 is able to achieve a high enterobactin-
enhancing effect at a much lower concentration in compari-
son to other crowding agents, probably due to its additional
interaction with the proteins as demonstrated in a previous
investigation.14
All the crowding agents were also found to decrease the
reaction rate of the in vitro enterobactin synthesis (Figure
2), while the reaction rate was less affected in 30% sucrose
Figure 1. Effects of macromolecular crowding on the nonribosomal
enterobactin synthesis. (A) HPLC analysis of the products from
the reactions in the absence and presence of a crowding agent; (B)
Variation of enterobactin fraction in the total turnover product with
the concentration of a crowder or sucrose. See text for peak labels.
IS: internal standard ) ethyl 3,4-dihydroxybenzoate.
percentile yield of at least 25% when the reaction was
extended beyond 10 min under those conditions.13
An inert polymer of sucrose recommended to mimic the
intracellular crowding,7c Ficoll 70, was included in the
reaction medium to examine its effects on the in vitro
enterobactin synthesis. Interestingly, it was found that the
linear side products were suppressed to negligible levels in
30% (w/v) Ficoll 70 solution (Figure 1A). In the meantime,
the synthesis of enterobactin was little affected (see Table
1S (Supporting Information) for details), resulting in a
percentage increase from 44% in noncrowded solution to
>95% in 30% Ficoll 70 solution (Figure 1B) for enterobactin
Figure 2. Crowding agents decrease the rate of AMP generation
in the nonribosomal enterobactin synthesis.
solution. This rate reduction is consistent with the HPLC
observation (Figure 1A) that normal production of entero-
bactin is little affected while production of the linear side
products is suppressed by the crowding agents. In contrast,
this rate reduction is unlikely due to the diffusion-impeding
effect of crowding7 on components of the synthetase because
the reaction rate is far from the diffusion limit.
(13) The EntB saturation reported in ref 11 is able to reduce the side
products from >80% to 25% at a low EntF concentration, but unable to
completely suppress their formation. It actually increases the side product
formation at the physiological EntF concentration. See Figure S8 in the
Supporting Information for more details. The premature release of linear
side products is universally suppressed by 30% Ficoll 70 to < 5%
irrespective of the composition of the synthetase.
(14) Minton, A. P. Mol. Cell. Biochem. 1983, 55, 119.
Org. Lett., Vol. 10, No. 4, 2008
651

by increasing the EntB concentration under non-crowding
conditions. Surprisingly, the premature release of the precur-
sors in the enterobactin synthesis was actually increased
(Figure 3). This result clearly showed that the strengthened
EntB-EntF interaction is not a cause of the observed side-
product suppressing effect of macromolecular crowding.
Future investigation of the cause of this crowding effect
should focus on potential crowding-induced structural changes
in protein components of the synthetase, which have been
suggested to alter the functional activities of several mono-
functional enzymes.15
In summary, we have shown that macromolecular crowd-
ing suppresses premature precursor release from the entero-
bactin synthetase. Since the synthetase shares the same
modular domain structure and thiotemplate catalytic mech-
anism with other NRPSs, this crowding effect is unlikely
limited to the enzymatic synthesis of enterobactin but is more
likely general to nonribosomal synthesis of other peptide
natural products. Results presented here suggest that mac-
romolecular crowding is not only critical to the correct under-
standing of the functions of NRPSs, but also an indispensable
factor in their biotechnological utilizations.
Figure 3. Decrease of enterobactin fraction in the total turnover
product with increasingly saturating EntB at the physiological
concentration of EntF and a saturating concentration of EntE. The
reaction was carried out in 75 mM Tris‚HCl buffer (pH 7.5)
containing 0.65 µM EntF, 1.5 mM DHB, 1.5 mM serine, 10 mM
ATP, 10 mM MgCl2, 5 mM DTT, varied concentrations of EntB,
and 0.35 µM EntE (() or 2 µM EntE (0) at 37 °C for 2 h.
Acknowledgment. This work was supported by HKUST
and RGC of the government of HKSAR.
Macromolecular crowding might affect the enterobactin
synthesis through influencing the specific enzyme-substrate
interaction between EntB and EntF, based on its well-known
effect of promoting protein-protein interaction.7 This EntB-
EntF interaction was indeed slightly strengthened as indicated
by the decrease of EntB half-saturation concentration from
0.50 µM in noncrowded buffer to 0.36 µM in 30% Ficoll
70 (Figure S7, Supporting Information). To examine how
the product specificity was affected by this affinity increase,
the EntB-EntF interaction was alternatively strengthened
Supporting Information Available: Experimental pro-
cedures, Table S1, and Figures S1-S8. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL7030153
(15) (a) Casati, D. F. G.; Aon, M. A.; Iglesias, A. A. Biochem. J. 2000,
350, 139. (b) Minh, D. D. L.; Chang, C.-e.; Trylska, J.; Tozzini, V.;
McCammon, J. A. J. Am. Chem. Soc. 2006, 128, 6006. (c) Jiang, M.; Guo,
Z. J. Am. Chem. Soc. 2007, 129, 730. (d) Mora´n-Zorzano, M. T.; Viale, A.
M.; Mun˜oz, F. J.; Alonso-Casaju´s, N.; Eydall´ın, G. G.; Zugasti, B.; Baroja-
Ferna´ndez, E.; Pozueta-Romero, J. FEBS Lett. 2007, 581, 1035.
652
Org. Lett., Vol. 10, No. 4, 2008