Identification of the new chymotrypsin inhibitor micropeptin 996 by metabolomics-guided analysis

Article abstract of DOI:10.1016/j.tetlet.2018.01.087

An untargeted metabolomics approach was used to investigate a cultured strain of Microcystis aeruginosa (UTEX LB2386) known to be a prolific producer of a diverse class of cyanopeptides. Identification of a putative new compound with a molecular weight of 996 led to the purification and structure elucidation of this new member of the micropeptin class of cyanopeptides. Micropeptin 996 displayed potent inhibition of the serine protease enzyme chymotrypisin relative to structurally related members of this class.

Full text of DOI:10.1016/j.tetlet.2018.01.087

Tetrahedron Letters 59 (2018) 934–937
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Identification of the new chymotrypsin inhibitor micropeptin 996 by
metabolomics-guided analysis
Wendy K. Strangman a, Allison K. Stewart a, Megan C. Herring b, Jeffrey L.C. Wright a,b,
⇑
a MARBIONC, University of North Carolina Wilmington, 5600 Marvin Moss Lane Wilmington, NC 28409, USA
b Department of Chemistry and Biochemistry, University of North Carolina Wilmington, 601 S. College Road Wilmington, NC 28403, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
An untargeted metabolomics approach was used to investigate a cultured strain of Microcystis aeruginosa
(UTEX LB2386) known to be a prolific producer of a diverse class of cyanopeptides. Identification of a
putative new compound with a molecular weight of 996 led to the purification and structure elucidation
of this new member of the micropeptin class of cyanopeptides. Micropeptin 996 displayed potent inhi-
bition of the serine protease enzyme chymotrypisin relative to structurally related members of this class.
Ó 2018 Published by Elsevier Ltd.
Received 15 December 2017
Revised 24 January 2018
Accepted 28 January 2018
Available online 31 January 2018
Keywords:
Cyanobacteria
Micropeptin
Metabolomics
Chymotrypsin inhibitor
Introduction
or fatty acids connected to the ring through the amide nitrogen
of the lactone forming threonine. Structurally similar compounds
Cyanobacteria are extremely prolific producers of biologically
active small molecules. Some freshwater genera such as Microcystis
are of enhanced interest due to their propensity for forming harm-
ful algal blooms (HABs) with high concentrations of toxins. In
addition to their acute toxins such as microcystins, cylindrosper-
mopsins, and anatoxin, many species also produce a wide variety
of other peptide natural products that act as potent and selective
inhibitors of proteases.1 In the environment, this inhibitory activity
is thought to protect the cyanobacteria by interfering with the
digestive enzymes of crustacean grazers.2 These compounds are
also of great interest in the search for new small molecule therapeu-
tics and for chemical probes of biological systems as proteases are
known to play key roles in infections as well as in cardiovascular
diseases and cancer.3–6
Micropeptins are members of the cyanopeptolin family of com-
pounds and are produced by species of Microcystis.7 This family is
characterized by a 6-amino acid-membered cyclic core cyclized
through a depsi-linkage between the side chain hydroxyl of the
N-terminus amino acid (almost always threonine) with the car-
boxy terminus of the adjacent residue. They are further defined
by the presence of the unique 3-amino-6-hydroxy piperadone
(Ahp) moiety and a side chain containing 1–4 amino, hydroxyl,
have been described from other freshwater cyanobacterial genera
including oscillapeptins (Oscillatoria), planktopeptins (Plank-
tothrix), nostopeptins (Nostoc), and aeruginopeptins (M. aerugi-
nosa).7–13Members have also been isolated from marine
cyanobacteria (eg. Lyngbyastatins14 and somamides15 from Lyng-
bya; symplocamide from Symploca.6 Partially due to their modular
non-ribosomal peptide synthase (NRPS) biosynthetic origin, there
is a high degree of variability within this structural class with over
150 congeners isolated to date.16
In our continuing search for novel chemistry and molecules with
potential therapeutic value from cultured microorganisms, we
employed the emerging technique of non-targeted metabolomic
screening to detect new compounds present in our cyanobacterial
culture collection. Metabolomics is a method for comprehensively
assessing the small-molecule content of a complex biological sys-
tem.17 We applied a UPLC-QTOF MS metabolomics method towards
laboratory cultures of the M. aeruginosa strain UTEX LB2386
(Supplemental). Originally isolated from Little Rideau Lake, Ontario,
Canada, this strain is not reported as acutely toxic and does not pro-
duce microcystins. Previous large-scale extractions of this strain
followed by traditional chromatographic separation, compound
purification, and structure confirmation by NMR indicated that this
strain is a prolific producer of several classes of bioactive cyanopep-
tides (Fig. 1). These include 3 new chlorinated microginins with the
rare 3-amino-2-hydroxy- octanoic acid (Ahoa) side chain18 as well
⇑
Corresponding author.
E-mail address: wrightj@uncw.edu (J.L.C. Wright).
https://doi.org/10.1016/j.tetlet.2018.01.087
0040-4039/Ó 2018 Published by Elsevier Ltd.

W.K. Strangman et al. / Tetrahedron Letters 59 (2018) 934–937
935
as the previously described aeruginosamides B and C,19 ferintoic
acid A,20 and several large molecular weight compounds assumed
to be members of the microviridin class of cyanopeptides.21 Given
our knowledge of the metabolites produced by this culture using
a traditional extraction and purification procedure, we employed
a non-targeted metabolomics approach to identify any additional
metabolites that remained undetected.
molecular formula of C52H68N8O12 for 1. Characteristic signals in
the 1D NMR spectrum including a singlet hydroxyl resonance at
dH 6.06, a quartet ester methine shifted downfield at dH 5.43, a sin-
glet N-methyl resonance (dH 2.78) and a methyl doublet at dH 1.21
suggested that this compound was a member of the cyanopeptolin
family of cyanopeptides. As it was isolated from a strain of Micro-
cystis aeruginosa, 1 is more precisely called a micropeptin. Further
2D NMR analysis including COSY, TOCSY, HSQC and HMBC allowed
us to assign the component amino acids of this compound in
sequence as valine, N-methyl phenyl alanine, phenylalanine, amino
hydroxyl piperidone (Ahp), homotyrosine, threonine, glutamine,
and butyric acid (Fig. 2 a and Table 1, please see Supplemental
for full correlation data table and NMR spectra). Further 2D NMR
analysis of the HMBC and ROESY spectra allowed the full planar
assignment of the cyclic depsipeptide including the side chain
extending from the threonine amine.
These assignments were further confirmed through analysis of
the high-resolution MS-MS fragmentation data. Fragment ions cor-
responding to reported values for Ahp-Phe-MePhe- H2O (m/z
404.1978), Ahp-Phe-H2O (m/z 243.1127), and Phe-MePhe-H2O
(m/z 292.1136) confirm the connectivity of residues 2–413, (Sup-
plemental). Additional ions at m/z 282.1446 and 199.1099 can be
ascribed to Htyr-Thr-H2O and BTA-Gln-H2O respectively.
Results and discussion
To perform the metabolomics analysis, replicate cultures were
grown, chemically extracted, and processed through SPE cartridges
prior to UPLC-HRQTOF spectral acquisition (see Supplemental for
experimental details). The chromatographic and mass spectral data
were then imported into Progenesis QI software for spectral pro-
cessing which resulted in a data set of 8079 ions. This set was then
filtered to remove compounds with anova p-values > 0.05 and
those with a highest mean value in the solvent blank, bringing
the total down to 5528. These compound ions were then exported
to the program EZinfo and subjected to orthogonal partial least
square discriminant statistical analysis (OPLS-DA) to construct an
S-plot comparing the cellular extracts to the media components.
A loadings value of >0.07 was chosen and 20 compound ions with
values greater than this were tagged and imported back into Proge-
nesis QI. Manual annotation of these ions allowed us to identify
ions falsely identified as separate compounds by the software but
were typically high intensity fragment ions from the most abun-
dant compounds produced by the organism. As others have noted,
there is a quite large dynamic range in the production of natural
products16 and samples loaded on to the QToF at concentrations
high enough to see the more minor compounds often result in false
positives around the major metabolites. Detailed analysis of the
untargeted metabolomics screen of cells vs the media blank rapidly
highlighted the compounds previously described from this strain
(Fig. 1). Additionally, a compound with an [M + H]+ m/z of 997.5
was identified and yielded no matches in our search of published
(Progenesis, Scifinder, MarinLit and Antibase) and internal data-
bases of compounds and was targeted for purification and struc-
ture determination.
Structure elucidation
Micropeptin 996 was isolated as a clear glassy material. High
resolution ToF mass spectrometry of the purified compound pro-
vided a molecular mass of [M + H]+ m/z 997.5032 indicating a
Fig. 2. Micropeptin 996 (1) structure with (a) key ROESY and HMBC correlations (b)
ROESY correlations for Ahp stereochemistry. BTA refers to butyric acid.
Fig. 1. S-Plot analysis expansion and representative classes of top 20 cyanopeptide mass features identified (red-box, color coded - micropeptin 996 is denoted by a pink star).
Each square corresponds to a high-resolution mass feature enabling accurate database searching.

936
W.K. Strangman et al. / Tetrahedron Letters 59 (2018) 934–937
Table 1
tion for structure-activity relationship analysis for this group of
compounds.6,12,23 In general, the nature of the amino acid residue
directly following Ahp in the peptide ring strongly influences the
resulting bioactivity specificity for trypsin vs. chymotrypsin. Basic
residues such as l-Arg at this position indicate a likelihood for tryp-
sin inhibition while bulky hydrophobic residues such as l-Phe sug-
gest a specificity for chymotrypsin (although this correlation is less
rigid for chymotrypsin).6 Compound 1 is unusual in that it contains
the rare l-homotyrosine amino acid in this position which has only
been observed in four other chymotrypsin inhibiting cyanopep-
tolins (oscillapeptins A, B (7-Mhtyr), and E and micropeptin
NMR Spectroscopic data (500 MHz for 1H, 125 MHz for 13C, DMSO-d6) for Cyanopep-
tolin 996.
Position
dH mult, J (Hz)
dC mult
Val 1
2
3
4
5
NH
NMePhe 1
2
171.8, C
55.7, CH
30.6, CH
19.3, CH3
17.2, CH3
4.74, dd (9.7, 4.5)
2.07, m
0.89, d (3.0)
0.73, d (6.9)
7.42, d (9.7)
168.9, C
5.01, m
3.23, m
2.85, m
60.6 (CH)
33.7 (CH2)
3a
KT946; (IC50’s = 2.2 lM, 2.1 lM, 3 lM, and 21.5 lM respec-
3b
tively).10,24 To examine its inhibitory activity, compound 1 was
screened in triplicate against both trypsin and chymotrypsin
enzymes. As expected, it showed no inhibition against trypsin.
However, it proved to be a potent inhibitor of chymotrypsin with
10
137.8, C
20,60
30,50
40
7.24, d (7.2)
7.39, t (7.2)
7.30, t (7.2)
2.78, s
129.6, CH
128.6, CH
126.6, CH
30.3, CH3
170.3, C
N-Me
Phe 1
2
an IC50 value of 0.64
tolin 963 standard25 that was used as a control in this assay and
has an IC50 of 1.01 M (see Supplemental for bioassay details and
data). structure/activity comparison analysis of reported
lM which is more potent than the cyanopep-
4.71, dd (12.0, 4.2)
2.82, m
49.9 (CH)
35.2 (CH2)
l
3a
3b
1.64, m
A
10
136.5, C
cyanopeptolins with chymotrypsin inhibition revealed that com-
pound 1 is structurally most similar to micropeptin KR998 with
differences lying at positions 2 and 5 (NMe-l-Phe and l-Htyr in 1
vs. NME-l-Tyr and l-Tyr in micropeptin KR998).26 Micropeptin
KR988 reportedly displayed no inhibition of the trypsin enzyme
while it was somewhat less active than compound 1 against chy-
20,60
30,50
40
6.76, d (7.1)
7.15, d (7.1)
7.12, d (7.1)
129.3 (CH)
127.6 (CH)
126.2 (CH)
169.0, C
Ahp 1
2
3a
3.58, m
48.4, CH
2.40, m
21.5, CH2
3b
1.55, m
motrypsin with an IC50 of 5.9 lM. This report underscores the
4a
1.66, m
1.49, m
5.02, m
29.2, CH2
73.5, CH
advantage of using a non-targeted metabolomics approach to
assist in the discovery of minor metabolites present in a complex
extract of an organism.
4b
5
NH
OH
HTyr 1
2
7.09, d (8.8)
6.04, d (3.3)
169.9, C
51.0, CH
32.1, CH2
Acknowledgements
4.03, dd (9.3, 2.6)
2.12, m
3a
3b
1.60, m
2.45, m
2.27, m
The authors would like to thank Mrs. Eve Wright for culturing of
cyanobacterial strain UTEX LB2386 and Mr. Michael Recchia for
performing the structure-activity literature search. They would
also like to acknowledge funding from the UNCW MARBIONC pro-
gram and the UNCW Carl B. Brown Trust to J.L.C.W.
4a
30.1, CH2
4b
10
130.9, C
20,60
30,50
40
6.86, d (8.5)
6.63, d (8.5)
129.3, CH
115.0, CH
155.4, C
NH
OH
Thr 1
2
3
4
NH
Gln 1
2
8.45, d (9.3)
9.11, s
A. Supplementary data
169.3, C
54.8, CH
71.7, CH
17.6, CH3
4.64, m
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.tetlet.2018.01.087.
These data include MOL files and InChiKeys of the most important
compounds described in this article.
5.43, dd (8.0, 6.6)
1.21, d (6.6)
8.01, d (9.0)
172.2, C
51.9, CH
27.8, CH2
4.42, m
1.93, m
1.74, m
2.17, m
3a
References
3b
4
5
31.6, CH2
173.9, C
1. Chlipala EG, Mo S, Orjala J. Curr Drug Targets. 2011;12:1654–1673.
2. Gademann K, Portmann C, Blom JF, Zeder M, Jüttner F.
2010;73:980–984.
J Nat Prod.
NH
NH2
BTA 1
2
3
4
8.05, d (7.9)
7.31, s
3. Southan C. Drug Discovery Today. 2001;6:681–688.
172.0, C
4. Prassas I, Eissa A, Poda G, Diamandis EP. Nat Rev Drug Discovery. 2015;14:183.
5. Vandooren J, Opdenakker G, Loadman PM, Edwards DR. Adv Drug Delivery Rev.
2016;97:144–155.
6. Linington RG, Edwards DJ, Shuman CF, McPhail KL, Matainaho T, Gerwick WH. J
Nat Prod. 2007;71:22–27.
2.12, m
1.52, m
0.87, d (3.5)
37.1, CH2
18.5. CH2
13.4, CH3
7. Okino T, Murakami M, Haraguchi R, Munekata H, Matsuda H, Yamaguchi K.
Tetrahedron Lett. 1993;34:8131–8134.
Acid hydrolysis of 1 and subsequent derivatization with Mar-
fey’s reagent22 followed by UPLC-ESIMS analysis9 and comparison
to standards established all of the amino acids possessed the L con-
figuration. Oxidation of 1 with Jones reagent followed by acid
hydrolysis, Marfey’s derivatization and UPLC-ESIMS analysis liber-
ated an l-glutamic acid identifying a 2S-configuration for the Ahp
moiety. The configuration of the Ahp C-5 was assigned as R based
on ROESY correlations (Fig. 2b).
8. Martin C, Oberer L, Ino T, Konig WA, Busch M, Weckesser J.
1993;46:1550–1556.
J Antibiot.
9. Harada K-I, Fujii K, Mayumi T, et al. Tetrahedron Lett. 1995;36:1515–1518.
10. Itou Y, Ishida K, Shin HJ, Murakami M. Tetrahedron. 1999;55:6871–6882.
11. Okino T, Qi S, Matsuda H, Murakami M, Yamaguchi K.
1997;60:158–161.
J Nat Prod.
12. Grach-Pogrebinsky O, Sedmak B, Carmeli S. Tetrahedron. 2003;59:8329–8336.
13. Welker M, Maršalek B, Šejnohová L, Von Doehren H. Peptides.
2006;27:2090–2103.
14. Harrigan GG, Yoshida WY, Moore RE, et al. J Nat Prod. 1998;61:1221–1225.
15. Nogle LM, Williamson RT, Gerwick WH. J Nat Prod. 2001;64:716–719.
16. Elkobi-Peer S, Carmeli S. Mar Drugs. 2015;13:2347–2375.
As mentioned above, a large number of cyanopeptolin-type
compounds have been isolated. This gives an exceptional founda-

W.K. Strangman et al. / Tetrahedron Letters 59 (2018) 934–937
937
17. Dettmer K, Aronov PA, Hammock BD. Mass Spec Rev. 2007;26:51–78.
22. Marfey P. Carlsberg Res Commun. 1984;49:591–596.
18. Strangman WK, Wright JLC. Tetrahedron Lett. 2016;57:1801–1803.
19. Leikoski N, Liu L, Jokela J, et al. Chem Biol. 2013;20:1033–1043.
20. Williams DE, Craig M, Holmes CF, Andersen RJ. J Nat Prod. 1996;59:570–575.
21. Ishitsuka MO, Kusumi T, Kakisawa H, Kaya K, Watanabe MM. J Am Chem Soc.
1990;112:8180–8182.
23. Salvador LA, Taori K, Biggs JS, et al. J Med Chem. 2013;56:1276–1290.
24. Beresovsky D, Hadas O, Livne A, Sukenik A, Kaplan A, Carmeli S. Israel J Chem.
2006;46:79–87.
25. Bister B, Keller S, Baumann HI, et al. J Nat Prod. 2004;67:1755–1757.
26. Bladt TT, Kalifa-Aviv S, Larsen TO, Carmeli S. Tetrahedron. 2014;70:936–943.