In Vitro Reconstitution of Bacterial DMSP Biosynthesis

Article abstract of DOI:10.1002/anie.201814662

Dimethylsulfoniopropionate (DMSP) is one of the most abundant sulfur metabolites in marine environments. The biosynthesis of DMSP and its degradation to dimethylsulfide are important links in the planetary sulfur cycle. Herein, the first complete description of a DMSP biosynthetic pathway is provided by the in vitro reconstitution of four enzymes from Streptomyces mobaraensis. The isolation of DMSP from S. mobaraensis cells grown at high salinity confirmed that this actinobacterium is indeed is a DMSP-producing organism. The described DMSP biosynthesis follows the same route as that previously described for angiosperm plants. Despite this chemical congruence, limited sequence similarity between plant and bacterial enzymes suggests that the two biosynthetic activities emerged by convergent evolution.

Full text of DOI:10.1002/anie.201814662

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: In vitro reconstitution of bacterial DMSP biosynthesis
Authors: Florian P. Seebeck and Cangsong Liao
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201814662
Angew. Chem. 10.1002/ange.201814662
Link to VoR: http://dx.doi.org/10.1002/anie.201814662
http://dx.doi.org/10.1002/ange.201814662

10.1002/anie.201814662
Angewandte Chemie International Edition
COMMUNICATION
In vitro reconstitution of bacterial DMSP biosynthesis
Cangsong Liao [a] and Florian P. Seebeck[a]*
Abstract: Dimethylsulfoniopropionate is one of the most abundant
sulfur metabolites in marine environments. Biosynthesis of DMSP
and degradation to dimethylsulfide are important links in the
planetary sulfur cycle. In this report we provide the first complete
description of a DMSP biosynthetic pathway by in vitro reconstitution
of four enzymes from Streptomyces mobaraensis. Isolation of DMSP
from S. mobaraensis cells grown at high salinity confirmed that this
actinobacterium is indeed is a DMSP producing organism. The
described DMSP biosynthesis follows same route as previously
described in angiosperm plants. Despite this chemical congruence,
limited sequence similarity between plant and bacterial enzymes
suggests that that the two biosynthetic activities emerged by
convergent evolution.
and many a-proteobacteria,
Met
S
NH2
O
OH
SAM
SAH
MetDC
MSMT
CO2
1a
2a
4a
O
NH2
NH2
OH
OH
S
S
S
O
O
SMMDC
CO2
Dimethylsulfoniopropionate (DMSP) is one of the most abundant
organosulfur compounds in marine environments with an
estimated biogenic production of 109 tons annually.[1] DMSP is
an important precursor for the production of dimethyl sulfide
(DMS) which in turn is oxidized to dimethylsulfoxide (DMSO) by
biotic or abiotic processes.[2] The portion of DMS that escapes to
the atmosphere represents the largest natural source of airborne
sulfur, second only to anthropogenic contributions.[3] DMS
oxidation products aggregate to aerosols that can effect cloud
formation with significant impact on the local and global
climate.[4] The recent discovery that the oxidation product of
DMSP – dimethylsulfoxonium propionate (DMSOP) – is also an
abundant marine product added another potentially important
link in the marine organosulfur cycle.[5]
2b
3
O
1b
OH
NH2
OH
OH
S
S
S
O
O
SAM
pyruvate
alanine
DMSPAAT
DsyB
SAH
1c
4b
OH
O
O
OH
O
2c
S
S
OH
S
+
NAD
As a cellular metabolite, DMSP has been implicated as a
protectant against osmotic, oxidative, or thermal stress,[6] as
signaling molecule,[7] or simply as a sulfur source for marine
microorganisms.[8] One difficulty in studying the physiological
DMSPADH
NADH
DMSP
O
function of DMSP is that the
genetic basis for DMSP
Until recently, DMSP
S
OH
biosynthesis is poorly understood.
production was believed to be exclusive to phytoplankton,
macroalgae, coral and coastal angiosperms.[1a] The recent
demonstration that numerous marine a-proteobacteria also
make DMSP completely changed this perspective.[9] Four
different pathways for DMSP biosynthesis have been proposed
based on isotope labeling and feeding experiments.[1a, 10] All four
pathways start with methionine, but differ in the sequence by
which S-methylation, decarboxylation, transamination and
oxidation occur (Figure 1). In green algae, coral, phytoplankton
Figure 1. DMSP biosynthesis in coral, green algae, phytoplankton and many
a-proteobacteria follows pathway
1
(1a
–
1c).[9-10] Spartina alterniflora
11]
(saltmarsh cordgrass),[10c,
and the actinobacterium Streptomyces
mobaraensis produce DMSP via pathway 2 (2a – 2c, red, this work). A third
and a fourth pathway has been postulated in Wollastonia biflora (beach daisy,
via 3)[10d] and in Dinoflagellata (4a – 4b) respectively.[10b]
the pathway starts with transamination followed by reduction,
methylation, oxidation and then decarboxylation (Figure 1, route
1). Pathways described in angiosperms start with S-methylation
and end with oxidation with transamination and decarboxylation
occurring either as separate (route 2) or combined (route 3)
steps. Detection of 3-(methylthio)propylamine (4a) in the
dinoflagellate Crypthecodinium cohnii raised the possibility of a
forth biosynthetic pathway (route 4).
[a]
C. Liao and F.P. Seebeck*
Department for Chemistry
University of Basel
Mattenstrasse 24a, 4002, Basel, Switzerland
E-mail: florian.seebeck@unibas.ch
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/anie.201814662
Angewandte Chemie International Edition
COMMUNICATION
1
3
4
2
MSMT
NH2
NH2
3
4
OH
OH
S
S
1
+
2
O
O
2a
Met
SMMDC
3
1
4
2
NH2
O
3
DMSPAAT
4
1
S
+
S
+
2
2c
2b
DMSPADH
spontaneous
3
4
2
O
1
4
+
3
O
S
S
OH
2
5
6
DMSP
3.8
3.6
3.4
3.2
3.0
[ppm]
2.8
2.6
2.4
2.2
1
Figure 2. Product analysis of in vitro reconstituted DMSP biosynthesis by H NMR spectroscopy. DMSP (blue) and the biosynthetic
intermediates SMM (2a, red) and DMSPA (2b, green) were produced by the concerted activities of MSMT, SMMDC, DMSPAAT and
DMSPADH. In the absence of DMSPDH the aldehyde 2c rapidly decays to DMS (5) and acrolein (6) (Figure S8).
Despite the apparent diversity of DMSP biosynthetic
pathways, information on the involved enzymes is limited.[1a]
Only the methionine S-methyltransferase MSMT (route 2 and
Table 1[a]
kcat,Met
kcat/KM,Met
kcat,SAM
kcat/KM,SAM
3),[12a]
and
the
methylthiohydroxybutryate
S-
enzyme
MSMTStr
MSMTRho
[s-1]
[M-1s-1]
[s-1]
[M-1s-1]
methyltransferase DsyB (route 1),[9, 14] have been identified
and directly linked to DMSP production. Complete sets of
DMSP biosynthetic enzymes are unknown.[15] In this report
we describe the full in vitro reconstitution of DMSP
biosynthesis from Streptomyces mobaraensis. This four-
enzyme cascade reaction follows the same route as
described for plants (route 2). However, limited sequence
similarities to plant enzymes suggest that the bacterial DMSP
biosynthesis emerged by convergent evolution.
0.09
0.23
0.27
50
23
0.1
230
180
0.28
0.27
[b]
MSMTWoll
1900
7900
[a]
The given values of the parameters represent averages from multiple
independent measurements with a standard deviation less than 20% of the
average value. kcat,Met and kcat,SAM were determined in the presence of 2
mM of SAM or a 20 mM of MET respectively. Parameters for MSMTWoll
[b]
were adapted from reference [12a]. The published value of Vmax = 2.7
nkat/mg was converted to kcat with a calculated mass for MSMTWoll of 120
kDa.
Bacterial DMSP biosynthetic genes. In a BLAST search
using plant MSMTs as a query we identified fourteen bacterial
homologs that are encoded in conjunction with two PLP-
dependent enzymes (fold-types I and III) (Table S1).[16] The
relative position and direction of the three open reading
frames is fully conserved (Figure S2). Most of these loci also
encode a putative NAD-dependent dehydrogenase attached
either before or after the conserved three-gene cluster
(Figure S2). Exceptions are Rhodovulum sp. P5 which
encodes all four genes clustered on a plasmid, and S.
mobaraensis which encodes the dehydrogenase in a different
locus. Nevertheless, conserved colocalization of these genes
across species from different bacterial phyla provided a first
indication that the corresponding enzymes form a functional
unit. To test the hypothesis that these genes enable DMSP
biosynthesis, we produced the corresponding enzymes from
S. mobaraensis as fusions with N-terminal His-tags in
Escherichia coli and examined their catalytic activity.
sp. JA431 (MSMTRho) are indeed bona fide Met-methylating
enzymes. To this end, we measured the rates of S-
methylmethionine (SMM, 2a, Figure 1) production in reactions
containing 1 mM methionine, 1 mM SAM and 10 uM of either
enzyme. The products S-adenosylhomocysteine (SAH) and
SMM were identified by chromatographic comparison with
authentic samples (Figure S3). SMM was also identified by
electrospray ionization high-resolution mass spectrometry
(ESI HRMS, m/z calc: 164.0740; obs: 164.0740) and 1H NMR
(Figure 2, Table S2). To compare the bacterial enzymes with
the plant MSMT from Wollastonia biflora (beach daisy,
MSMTWoll),[12a] we also determined the Michaelis-Menten
parameters and the substrate-binding mechanism of MSMTStr
and MSMTRho (Table 1, Figure S4 and S5). This analysis
revealed that the bacterial enzymes are 40-fold less efficient
that the plant enzyme, mostly due to higher KM values for
both Met and SAM. However, we note that KM values in the
millimolar range are not unusual for Met-utilizing enzymes.[12b,
Kinetic characterization of bacterial MSMTs. Because of
the low sequence similarity to plant MSMTs (24 % sequence
identity, seqID), we first needed to test whether the putative
MSMT from S. mobaraensis (MSMTStr) and from Rhodobacter
17]
Plant and bacterial MSMTs also differ in their substrate-
binding mechanism. Binding of Met and SAM to MSMTRho
follows a random order, whereas binding to MSMTWoll was
This article is protected by copyright. All rights reserved.

10.1002/anie.201814662
Angewandte Chemie International Edition
COMMUNICATION
reported to follow an obligatory sequence with SAM as the
leading substrate (Figure S6). These different kinetic
behaviors may be an indication, that bacterial and plant
enzymes were adapted to function under different conditions.
oxidase instead of a PLP-dependent transaminase as a
catalyst for this step.[19]
A further indication that plant and bacterial DMSP production
emerged from different origins comes from the observation
that bacterial DMSP biosynthetic enzymes are more related
to bacterial enzymes with other functions, rather than to plant
enzymes. SMMDC from S. mobaraensis shares up to 50 %
seqID with putative bacterial diaminopimelic acid
decarboxylases. DMSPAAT shares up to 60 % seqID with
putative bacterial transaminases, and DMSPADH shares up
In vitro reconstitution of DMSP biosynthesis. In a next
step, we examined the contribution of the three remaining
enzymes to DMSP biosynthesis. 1 mM SMM was incubated
with 10 uM of the putative PLP-dependent decarboxylase
(SMMCD, Figure 1). This reaction converted all SMM to
dimethylsulfoniopropylamine (DMSPamine, 2b) as inferred by
ESI HRMS (m/z calc: 120.0841; obs: 120.0841) and by 1H
NMR (Figure 2, Table S2). The same reaction supplemented
with 10 uM of the putative aminotransferase (DMSPAAT) and
pyruvate, consumed all 2b and produced one equivalent of
alanine (Table S2), suggesting that DMSPAAT transferred
the amino group from 2b to pyruvate. The second expected
product dimethylsulfoniopropanal (DMSPaldehyde, 2c) could
not be detected because this compound spontaneously
decayed to acrolein by elimination of DMS (Figure S7).[18] In
contrast, a reaction containing SMMCD, DMSPAAT together
to 60
% seqID with putative glycine betaine aldehyde
dehydrogenases. By contrast, plant proteins are significantly
less related to SMMDC (< 30 % seqID), DMSPAAT (< 35 %
seqID) or DMSPADH (< 40 % seqID). Although MSMTStr and
MSMTWoll catalyze the same reaction they are only distantly
related (24% seqID). MSMTStr is much more related to MSMT
homologs (> 50 % seqID) from bacteria that contain no
discernable homologs of SMMDC and DMSPAAT (< 30%
seqID). These sequence relationships suggest that DMSP
production in bacteria via route 2 emerged by adaptation of
bacterial enzymes from other pathways rather than from gene
exchanges with plants.
with
the
aldehyde
dehydrogenase
(DMSPaldehyde
dehydrogenase, DMSPADH) and NAD+, converted 80 % of
SMM to DMSP (m/z calc: 135.0472; obs: 135.0474, 1H NMR:
Figure 2, Table S2). Finally, a reaction combining these three
enzymes with MSMT and SAM was able to catalyze the
entire four-step transformation of Met to DMSP (Figure S8).
This experiment represents the first in vitro reconstitution of
DMSP biosynthesis.
This finding conflicts with the expectation that secondary
metabolite production lines are usually disseminated by
horizontal gene transfer.[13] Exchange of ready-to-use genetic
instructions between different species generally occurs far
more frequently than de novo emergence of new enzyme
activity, let alone entire enzyme cascades. However,
emergence of DMSP biosynthetic activity may be an
unusually simple evolutionary process because of the
adaptability of the involved enzyme types. PLP-dependent
enzymes are often promiscuous and may adapt substrate
and reaction specificities requiring only minimal sequence
changes.[20] Aldehyde dehydrogenases too are often
characterized by limited substrate specificity not least
because bacteria require unspecific aldehyde dehydrogenase
activity as a protection against electrophilic stress.[21] Hence,
onset of DMSP production and accumulation in an adapting
cell may take as little as acquiring an MSMT coding gene by
horizontal gene-transfer and downregulation of activities that
consume SMM through competing pathways.[22] This
simplicity provides a plausible answer for why bacterial and
plant DMSP production emerged independently. By extension,
we would not be surprised to find additional unrelated DMSP
biosynthetic loci as more genome sequences are becoming
available.
Production of DMSP by S. mobaraensis. The in vitro
activity of these four enzymes suggest that S. mobaraensis
can produce DMSP. To test this expectation, we cultivated S.
mobaraensis (DMSZ DSM 40847) in liquid culture in a
medium containing glucose, yeast-, and malt-extract (GYM-
Streptomyces Medium) supplemented with 0, 0.4 or 0.8 M
NaCl. Ethanolic extracts from cells grown in 0.8 M indeed
contained DMSP as inferred by 1H NMR spectroscopy and by
comparison with authentic DMSP (Figure S9). Quantification
of the dimethylsulfonium proton signal indicated a cellular
DMSP concentration of 0.50 ± 0.06 mM (Figure S10). In
contrast, cells grown at low or medium salinity contained less
than 50 uM DMSP (Figure S11). Induction of DMSP
biosynthesis by osmotic stress is consistent with the idea that
this metabolite serves as an osmolyte.[6b, 9] Indeed, most of
the fourteen bacterial species that contain close homologs of
the DMSP biosynthetic enzymes from S. mobaraensis were
isolated from saline or hypersaline habitats (Table S1).
Conclusions. In this report we described the first in vitro
reconstitution of DMSP biosynthesis using four enzymes from
S. mobaraensis. Identification of these enzymes lead to the
Independent origin of bacterial DMSP production. DMSP
biosynthesis via SMM (route 2) was also observed in
angiosperms such as W. biflora[10d] and Spartina alterniflora
(saltmarsh cordgrass).[10c] However, the following evidence
suggests that plant and bacterial DMSP biosynthetic activities
may have emerged independently. Since the plant enzymes
transforming SMM to DMSP are not known, the descriptions
of these pathways are solely based on the structures of
detected metabolites. In W. biflora 2b could not be detected
as an intermediate, which lead to the proposition that
deamination and decarboxylation may be coupled in a single
enzymatic step (route 3).[10d] In S. alterniflora the conversion
of 2b to 2c was found to be O2-dependent,[10c] implicating an
correct prediction that this actinobacterium is
a DMSP
producing organism. Although the bacterial pathway follows
the same chemical steps as described in plants,[11] limited
sequence similarity to plant enzymes suggest that the two
pathways emerged through convergent evolution. The broad
distribution of MSMT-like enzymes, the adaptability of PLP-
dependent enzymes and the promiscuity of aldehyde
dehydrogenases may have been key factors making de novo
assembly of a DMSP production line in bacteria more efficient
than horizontal gene transfer from plants. The discovery of
angiosperm-like DMSP biosynthesis in bacteria also lends
This article is protected by copyright. All rights reserved.

10.1002/anie.201814662
Angewandte Chemie International Edition
COMMUNICATION
further support to the notion that this biosynthetic activity is
far more common than previously thought.[9]
Hirose, T. Okajima, T. Tamura, K. Soda, K. Inagaki, J Biochem.
2017, 161, 389 - 398.
[13]
[14]
aM. Syvanen, Annu Rev Genet. 2012, 46, 341 - 358; bR. G.
Beiko, T. J. Harlow, M. A. Ragan, Proc Natl Acad Sci U S A.
2005, 102, 14332 - 14337; cB. Lacroix, V. Citovsky, MBio 2016,
7, e00863 - 00816.
Acknowledgements
L.C. was a recipient of a SSSTC & CSC postdoctoral
fellowship; F.P.S. is supported by the “Professur für
Molekulare Bionik”. This project was supported by European
Research Council (ERC-2013-StG 336559) and the
“Innovationsraum Biokatalyse”
A. R. J. Curson, B. T. Williams, B. J. Pinchbeck, L. P. Sims, A.
B. Martínez, P. P. L. Rivera, D. Kumaresan, E. Mercadé, L. G.
Spurgin, O. Carrión, S. Moxon, R. A. Cattolico, U.
Kuzhiumparambil, P. Guagliardo, P. L. Clode, J. B. Raina, J. D.
Todd, Nat Microbiol. 2018, 3, 430 - 439.
[15]
H. Rousseau, M. Rousseau-Gueutin, X. Dauvergne, J. Boutte,
G. Simon, N. Marnet, A. Bouchereau, S. Guiheneuf, J. P.
Bazureau, J. Morice, S. Ravanel, F. Cabello-Hurtado, A.
Ainouche, A. Salmon, J. F. Wendel, ., M. L. Ainouche, Mol
Phylogenet Evol. 2017, 114, 401 - 414.
Keywords: DMSP • biosynthesis • pyridoxal phosphate • S-
Methylmethionine • methyltransferase • carbon-sulfur bond
[1]
aJ. S. Dickschat, P. Rabe, C. A. Citron, Org Biomol Chem.
2015, 13, 1954 - 1968; bH. A. Bullock, H. Luo, W. B. Whitman,
Front Microbiol. 2017, 8, 637; cM. A. Moran, C. R. Reisch, R. P.
Kiene, W. B. Whitman, Ann Rev Mar Sci. 2012, 4, 523 - 542;
dJ. D. Todd, A. R. Curson, M. J. Sullivan, A. W. Johnston, Nat
Rev Microbiol. 2012, 9, 849 - 859; eS. M. Sievert, R. P. LKiene,
H. N. H. Schulz-Vogt, F., Oceanography 2007, 20, 117 - 123;
fK. B. Ksionzek, O. J. Lechtenfeld, S. L. McCallister, P.
Schmitt-Kopplin, J. K. Geuer, W. Geibert, B. P. Koch, Science
2016, 354, 456 - 459.
[16]
[17]
J. N. Jansonius, Curr Opin Struct Biol. 1998, 8, 759 - 769.
aF. Rébeillé, S. Jabrin, R. Bligny, K. Loizeau, B. Gambonnet, V.
Van Wilder, R. Douce, S. Ravanel, Proc Natl Acad Sci U S A.
2006, 103, 15687 - 15692; bJ. Heilbronn, J. Wilson, B. J.
Berger, J. Bacteriol. 1999, 181, 1739 - 1747.
[18]
C. Trossat, K. D. Nolte, A. D. Hanson, Plant Physiol. 1996, 111,
965 - 973.
[19]
[20]
J. P. Klinman, Biochim Biophys Acta. 2003, 164, 131 - 137.
aM. D. Toscano, K. J. Woycechowsky, D. Hilvert, Angew.
Chem. Int. Ed. 2007, 46, 3212-3236; bV. W. Soo, Y.
Yosaatmadja, C. J. Squire, W. M. Patrick, J. Biol. Chem. 2016,
291, 19873 - 19887.
[2]
aP. A. Lee, S. J. Mora, ATMOSPHERE-OCEAN 1999, 37, 439
- 456; bU. Alcolombri, S. Ben-Dor, E. Feldmesser, Y. Levin, D.
S. Tawfik, A. Vardi, Science 2015, 348, 1466 - 1469.
M. Chin, D. J. Jacob, J. Geophys. Res. 1996, 101, 18691 -
18699.
[3]
[4]
[5]
[6]
[21]
S. Singh, C. Brocker, V. Koppaka, Y. Chen, B. C. Jackson, A.
Matsumoto, D. C. Thompson, V. Vasiliou, Free Radic Biol Med.
2013, 56, 89 - 101.
R. J. Charlson, J. E. Lovelock, M. O. Andreae, S. G. Warren,
Nature 1987, 326, 655 - 661.
[22]
.
M. Thanbichler, B. Neuhierl, A. Böck, J. Bacteriol. 1999, 181,
662 - 665.
K. Thume, B. Gebser, L. Chen, N. Meyer, D. J. Kieber, G.
Pohnert, Nature 2018, 563, 412 - 415.
aW. Sunda, D. J. Kieber, R. P. Kiene, S. Huntsman, Nature
2002, 418, 317 - 320; bD. T. Welsh, FEMS Microbiol Rev. 2000,
24; cJ. B. Raina, D. M. Tapiolas, S. Forêt, A. Lutz, D. Abrego, J.
Ceh, F. O. Seneca, P. L. Clode, D. G. Bourne, B. L. Willis, C. A.
Motti, Nature 2013, 502, 677 - 680.
[7]
[8]
[9]
J. R. Seymour, R. Simó, T. Ahmed, R. Stocker, Science 2010,
329, 342 - 345.
H. J. Tripp, J. B. Kitner, M. S. Schwalbach, J. W. Dacey, L. J.
Wilhelm, S. J. Giovannoni, Nature 2008, 452, 741 - 744.
A. R. Curson, J. Liu, A. Bermejo Martínez, R. T. Green, Y.
Chan, O. Carrión, B. T. Williams, S. H. Zhang, G. P. Yang , P.
C. Bulman Page, X. H. Zhang, J. D. Todd, Nat Microbiol. 2017,
2, 17009.
[10]
aD. A. Gage, D. Rhodes, K. D. Nolte, W. A. Hicks, T. Leustek,
A. J. Cooper, A. D. Hanson, Nature 1997, 387, 891 - 894; bH.
Kitaguchi, A. Uchida, Y. Ishida, Fish. Sci. 1999, 65, 613 - 617;
cM. G. Kocsis, A. D. Hanson, Plant Physiol. 2000, 123, 1153 -
1161; dD. Rhodes, D. A. Gage, A. Cooper, A. D. Hanson, Plant
Physiol. 1997, 115.
[11]
[12]
M. G. Kocsis, K. D. Nolte, D. Rhodes, T. L. Shen, D. A. Gage,
H. A.D., Plant. Physiol. 1998, 117, 273 - 281.
aF. James, K. D. Nolte, A. D. Hanson, J. Biol. Chem. 1995,
270, 22344 - 22350; bM. Hayashi, A. Okada, K. Yamamoto, T.
Okugochi, C. Kusaka, D. Kudou, M. Nemoto, J. Inagaki, Y.
This article is protected by copyright. All rights reserved.

10.1002/anie.201814662
Angewandte Chemie International Edition
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 1:
COMMUNICATION
Text for Table of Contents
Cangsong Liao and Florian P. Seebeck*
Page No. – Page No.
In vitro reconstitution of bacterial
DMSP biosynthesis
NH2
O
NH2
O
NH2
OH
OH
S
S
S
O
O
S
OH
S
This article is protected by copyright. All rights reserved.