Targeted Isolation of Antibiofilm Compounds from Halophytic Endophyte Bacillus velezensis 7NPB-3B Using LC-HR-MS-Based Metabolomics

Abstract

The discovery of new natural products has become more challenging because of the re-isolation of compounds and the lack of new sources. Microbes dwelling in extreme conditions of high salinity and temperature are huge prospects for interesting natural metabolites. In this study, the endophytic bacteria Bacillus velezensis 7NPB-3B isolated from the halophyte Salicornia brachiata was screened for its biofilm inhibition against methicillin-resistant Staphylococcus aureus (MRSA). The fractionation of the crude extract was guided by bioassay and LC-HRMS-based metabolomics using multivariate analysis. The 37 fractions obtained by high-throughput chromatography were dereplicated using an in-house MS-Excel macro coupled with the Dictionary of Natural Products database. Successive bioactivity-guided separation yielded one novel compound (1), a diketopiperazine (m/z 469.258 [M − H]⁻) with an attached saturated decanoic acid chain, and four known compounds (2–5). The compounds were identified based on 1D- and 2D-NMR and mass spectrometry. Compounds 1 and 5 exhibited excellent biofilm inhibition properties of >90% against the MRSA pathogen at minimum inhibition concentrations of 25 and 35 µg/mL, respectively. The investigation resulted in the isolation of a novel diketopiperazine from a bacterial endophyte of an untapped plant using an omics approach.

Full text

Citation: Singh, S.; Nwagwu, E.;
Young, L.; Kumar, P.; Shinde, P.B.;
Edrada-Ebel, R. Targeted Isolation of
Antibiofilm Compounds from
Halophytic Endophyte Bacillus
velezensis 7NPB-3B Using
LC-HR-MS-Based Metabolomics.
Microorganisms 2024,12, 413.
https://doi.org/10.3390/
microorganisms12020413
Academic Editor: Renato Fani
Received: 15 January 2024
Revised: 12 February 2024
Accepted: 16 February 2024
Published: 19 February 2024
Copyright: © 2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
microorganisms
Article
Targeted Isolation of Antibiofilm Compounds from
Halophytic Endophyte Bacillus velezensis 7NPB-3B Using
LC-HR-MS-Based Metabolomics
Sanju Singh 1,2,3, Elizabeth Nwagwu 3, Louise Young 3, Pankaj Kumar 1,2, Pramod B. Shinde 1, 2,*
and RuAngelie Edrada-Ebel 3, *
1Natural Products & Green Chemistry Division, CSIR-Central Salt and Marine Chemicals Research
Institute (CSIR-CSMCRI), Council of Scientific and Industrial Research (CSIR), Bhavnagar 364002, India;
sanju.singh@strath.ac.uk (S.S.); pawar.p08@gmail.com (P.K.)
2Academy of Scientific and Innovative Research (AcSIR), Ghaziabad 201002, India
3Strathclyde Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, The John Arbuthnott
Building, 161 Cathedral Street, Glasgow G4 0RE, UK; elizabeth.nwagwu@strath.ac.uk (E.N.);
louise.c.young@strath.ac.uk (L.Y.)
*Correspondence: pramodshinde@csmcri.res.in (P.B.S.); ruangelie.edrada-ebel@strath.ac.uk (R.E.-E.)
Abstract: The discovery of new natural products has become more challenging because of the re-
isolation of compounds and the lack of new sources. Microbes dwelling in extreme conditions of
high salinity and temperature are huge prospects for interesting natural metabolites. In this study,
the endophytic bacteria Bacillus velezensis 7NPB-3B isolated from the halophyte Salicornia brachiata
was screened for its biofilm inhibition against methicillin-resistant Staphylococcus aureus (MRSA). The
fractionation of the crude extract was guided by bioassay and LC-HRMS-based metabolomics using
multivariate analysis. The 37 fractions obtained by high-throughput chromatography were derepli-
cated using an in-house MS-Excel macro coupled with the Dictionary of Natural Products database.
Successive bioactivity-guided separation yielded one novel compound (1), a diketopiperazine (m/z
469.258 [M
−
H]
−
) with an attached saturated decanoic acid chain, and four known compounds (2–5).
The compounds were identified based on 1D- and 2D-NMR and mass spectrometry. Compounds
1and 5exhibited excellent biofilm inhibition properties of >90% against the MRSA pathogen at
minimum inhibition concentrations of 25 and 35
µ
g/mL, respectively. The investigation resulted in
the isolation of a novel diketopiperazine from a bacterial endophyte of an untapped plant using an
omics approach.
Keywords: antibiofilm; dereplication; endophytes; halophyte; metabolite profiling; multivariate analysis
1. Introduction
Bacterial biofilms cause serious global health concerns contributing to persistent
chronic infections. Biofilms provide a shield for pathogenic bacteria against the host
immune system by inhibiting phagocyte function and suppressing complement system ac-
tivation, thereby enhancing bacterial resistance against conventional antibiotics by around
1000-fold [
1
,
2
]. Multidrug-resistant bacteria like Staphylococcus aureus are responsible
for nosocomial and community-acquired infections globally causing numerous diseases,
such as food poisoning, skin infections, pneumonia, and septicemia. Biofilm-forming
staphylococci for instance S. epidermis and S. aureus cause biomaterial-associated infec-
tions [
3
]. Hence, it is imperative to develop new means of eradication to combat resistant
pathogens [4].
Fermentation bioprocessing technology in microbial systems has elevated their use
in the isolation of commercially important metabolites. There lies a gap in research and
investigation of beneficial compounds from extreme environments, such as marine, desert,
and marshlands. Endophytic bacteria from halophytes hold a vast antimicrobial potential
Microorganisms 2024,12, 413. https://doi.org/10.3390/microorganisms12020413 https://www.mdpi.com/journal/microorganisms

Microorganisms 2024,12, 413 2 of 20
against pathogenic fungi and bacteria [
5
]. Among various biocontrol agents, Bacillus sp.
has demonstrated strong abilities to restrict microbial pathogens of various niches by
the production of hydrolytic enzymes, peptides, and polyketides [
6
]. B. velezensis is a
recently classified species of the genus Bacillus [
7
,
8
] and is reported to exhibit antifungal
activity, plant growth promotion properties, dye detoxification, and keratinolytic, dehairing,
and proteolytic actions. The versatility extends to the production of enzymes such as
protease and cellulase, bioremediation of organophosphorus pesticides, and generation of
probiotics for animal feeds [
7
,
9
,
10
]. We hypothesized that investigation of endophytic B.
velezensis from a halophyte growing in extreme conditions can lead to the isolation of some
interesting metabolites. Compounds such as surfactin, fengycin, bacilysin, bacillibactin,
bacillaene, macrolactin, and difficidin are reported from varied B. velezensis strains [
11
–
13
].
An antimicrobial peptide YS12 (mwt. 3348 Da) was reported from B. velezensis CBSYS
12
that
exhibited antibiofilm properties against bacterial pathogens [
14
]. The presence of secondary
metabolite biosurfactants such as surfactins also contributes to the antibiofilm capacity of
B. velezensis strains [15–17].
The discovery of new potential bioactive compounds from natural niches is challeng-
ing because of the re-isolation of compounds. Metabolite profiling of crude extracts in
amalgamation with dereplication using natural product databases can facilitate rapid and
high-throughput assessment of metabolites. Dereplication benefits by screening out the
known metabolites in the crude extract to guide the isolation procedure for novel metabo-
lites [
18
]. In this study, we investigate an endophyte B. velezensis 7NPB-3B isolated from
the halophytic plant Salicornia brachiata for antibiofilm compounds using metabolomics in
concatenation with bioassay-guided separation. We performed LC-MS-based metabolomics
of the crude extract and fractions using multivariate analysis, i.e., PCA and OPLS-DA to
determine the significant metabolites responsible for the bioactivity. Putative identification
of the metabolites was carried out employing the Dictionary of Natural Products (DNP)
database. The study structurally characterized and identified the bioactive compounds
capable of effectively controlling MRSA biofilms by using NMR spectroscopy and mass
spectrometry techniques.
2. Results and Discussion
Recently, our laboratory isolated around 350 bacterial endophytes from the halophyte
Salicornia brachiata wildly distributed along the coast of Gujarat, India. The endophytic
strain of interest, 7NPB-3B was identified as Bacillus velezensis with the GenBank accession
number MT645763 based on its 16S rRNA sequence and phylogenetic analysis performed
using MEGA X [
5
]. B. velezensis 7NPB-3B displayed a broad antibacterial potential against
pathogenic bacteria, such as S. aureus MCC 2043, E. coli MCC 2412, P. aeruginosa MTCC 3541,
and X. campestris NCIM 5028 (Figure 1). B. velezensis species is a recently identified species
under the genus Bacillus [
8
]; therefore, it is important to explore its bioactive potential to
determine a distinguishing factor with respect to its contemporaries. Therefore, bioactivity-
guided isolation of antibiofilm metabolites from the strain was attempted using MRSA as
the biofilm-forming agent.
Microorganisms 2024, 12, x FOR PEER REVIEW 3 of 21
Figure 1. Preliminary antimicrobial screening of B. velezensis 7NPB-3B crude extract (CE) against
four different pathogens using disk diffusion assay at a concentration of 1 mg/mL and loading vol-
ume of 10 µL keeping DMSO as vehicle control.
For the extraction of bioactive metabolites, a large-scale culture of B. velezensis 7NPB-
3B was performed and the culture broth was extracted with ethyl acetate resulting in 13.9
g crude extract. The crude extract was partitioned between methanol and hexane to re-
move the non-polar or possible fatty acid constituents. The total ion chromatogram (TIC)
of the crude extract of B. velezensis 7NPB-3B (Figure 2) when dereplicated using the DNP
database showed the existence of known and unknown compounds present in the extract
(Table 1). Some of the putatively identified compounds have been previously isolated
from microbes including the genus Bacillus. Characteristic metabolites known from the
genus were annotated such as a series of more non-polar surfactin congeners with molec-
ular weights at ca. 1000 Da eluting after 25 min. Some ion mass peaks were found to have
“No Hit”. However, their peak areas and predicted formulas were also listed. On the other
hand, four of the dereplicated features were plant metabolites, which could indicate that
referred ion peaks may instead belong to an unreported or unknown metabolite yet to be
discovered. The dereplication study revealed that the bacterial extract possesses a wide
arena of compounds such as lipopeptides, diketopiperazines, polyketides, quinones,
amino alcohols, and macrolides isolated from microbial sources either fungi or bacteria.
The number of probable hits was filtered by biological sources that were closer to our
strain of interest, i.e., B. velezensis 7NPB-3B.
Figure 2. LC-MS total ion chromatogram (TIC) of B. velezensis 7NPB-3B crude extract in positive (—
) as well as negative mode (—) compared with media control in positive (—) and negative mode (—
). The labeled ion peaks represent some of the significant metabolites.
Figure 1. Preliminary antimicrobial screening of B. velezensis 7NPB-3B crude extract (CE) against four
different pathogens using disk diffusion assay at a concentration of 1 mg/mL and loading volume of
10 µL keeping DMSO as vehicle control.

Microorganisms 2024,12, 413 3 of 20
For the extraction of bioactive metabolites, a large-scale culture of B. velezensis 7NPB-
3B was performed and the culture broth was extracted with ethyl acetate resulting in 13.9 g
crude extract. The crude extract was partitioned between methanol and hexane to remove
the non-polar or possible fatty acid constituents. The total ion chromatogram (TIC) of
the crude extract of B. velezensis 7NPB-3B (Figure 2) when dereplicated using the DNP
database showed the existence of known and unknown compounds present in the extract
(Table 1). Some of the putatively identified compounds have been previously isolated from
microbes including the genus Bacillus. Characteristic metabolites known from the genus
were annotated such as a series of more non-polar surfactin congeners with molecular
weights at ca. 1000 Da eluting after 25 min. Some ion mass peaks were found to have
“No Hit”. However, their peak areas and predicted formulas were also listed. On the
other hand, four of the dereplicated features were plant metabolites, which could indicate
that referred ion peaks may instead belong to an unreported or unknown metabolite yet
to be discovered. The dereplication study revealed that the bacterial extract possesses a
wide arena of compounds such as lipopeptides, diketopiperazines, polyketides, quinones,
amino alcohols, and macrolides isolated from microbial sources either fungi or bacteria.
The number of probable hits was filtered by biological sources that were closer to our strain
of interest, i.e., B. velezensis 7NPB-3B.
Microorganisms 2024, 12, x FOR PEER REVIEW 3 of 21
Figure 1. Preliminary antimicrobial screening of B. velezensis 7NPB-3B crude extract (CE) against
four different pathogens using disk diffusion assay at a concentration of 1 mg/mL and loading vol-
ume of 10 µL keeping DMSO as vehicle control.
For the extraction of bioactive metabolites, a large-scale culture of B. velezensis 7NPB-
3B was performed and the culture broth was extracted with ethyl acetate resulting in 13.9
g crude extract. The crude extract was partitioned between methanol and hexane to re-
move the non-polar or possible fatty acid constituents. The total ion chromatogram (TIC)
of the crude extract of B. velezensis 7NPB-3B (Figure 2) when dereplicated using the DNP
database showed the existence of known and unknown compounds present in the extract
(Table 1). Some of the putatively identified compounds have been previously isolated
from microbes including the genus Bacillus. Characteristic metabolites known from the
genus were annotated such as a series of more non-polar surfactin congeners with molec-
ular weights at ca. 1000 Da eluting after 25 min. Some ion mass peaks were found to have
“No Hit”. However, their peak areas and predicted formulas were also listed. On the other
hand, four of the dereplicated features were plant metabolites, which could indicate that
referred ion peaks may instead belong to an unreported or unknown metabolite yet to be
discovered. The dereplication study revealed that the bacterial extract possesses a wide
arena of compounds such as lipopeptides, diketopiperazines, polyketides, quinones,
amino alcohols, and macrolides isolated from microbial sources either fungi or bacteria.
The number of probable hits was filtered by biological sources that were closer to our
strain of interest, i.e., B. velezensis 7NPB-3B.
Figure 2. LC-MS total ion chromatogram (TIC) of B. velezensis 7NPB-3B crude extract in positive (—
) as well as negative mode (—) compared with media control in positive (—) and negative mode (—
). The labeled ion peaks represent some of the significant metabolites.
Figure 2. LC-MS total ion chromatogram (TIC) of B. velezensis 7NPB-3B crude extract in positive (—)
as well as negative mode (—) compared with media control in positive (—) and negative mode (—).
The labeled ion peaks represent some of the significant metabolites.
The methanolic extract (13.4 g) was fractionated using normal phase flash chromatogra-
phy resulting in 16 different fractions with the last fraction being the most polar (Scheme 1).
To proceed with efficient isolation of bioactive metabolites, these fractions were tested for
their inhibitory activity against planktonic cells of MRSA ATCC 43300 and for their capacity
to destabilize the formed MRSA biofilms. Only two fractions F9 and F15 inhibited about
50% planktonic MRSA at a concentration of 100
µ
g/mL, whereas excellent dispersion of
formed biofilms of up to 90% was observed with fractions F2, F3, F4, F10, F11, F13, and
F15 (Scheme 1and Figure 3A). Based on the results of the post-biofilm inhibition assay,
the bioactive fractions were further purified by either preparatory TLC or Biotage flash
chromatography. Moreover, the metabolite profile of the fractions was dereplicated using
LC-HRMS before multivariate analysis. After the successive high-throughput chromato-

Microorganisms 2024,12, 413 4 of 20
graphic separation, a total of 37 subfractions were obtained. All the subfractions were
tested for their anti-MRSA activity and post-biofilm inhibition of MRSA biofilms. Only
two subfractions F10.6.1 and F17 exhibited activity >50% against planktonic MRSA cells
at a concentration of 100
µ
g/mL. In the case of post-biofilm inhibition, nine subfractions,
i.e., F2.5, F3.3, F10.5.2, F11.2, F11.3, F13.2, F15.4.1, F16.3, and F16.4, displayed >90% inhi-
bition at the significant concentration of 100
µ
g/mL (Figure 3B). The pure fractions were
then subjected to NMR analysis for identification and structure elucidation considering
their bioactivity.
Table 1. The predicted compounds from the TIC of the crude extract of B. velezensis 7NPB-3B after
dereplication using the DNP database. The listed features are arranged according to their retention
time (RT) in min, while the selected masses are the peaks visible in the TIC. Highlighted features
in grey were hits only found in plants, which may instead indicate an unknown compound. The
highlighted row in green indicates the isolated metabolite in this study.
Mzmine
ID *
RT
(min) MS m/zExact
Mass
Chemical
Formula Name
Tolerance
Source Peak Area
P48233 1.19 170.084 169.076 C7H12N3P No hits
P21878 1.50 212.091 211.084 C10H13 NO4N-salicyloylserinol 1.12 Streptomyces
hygroscopicus 2,076,169.45
P20118 4.65 211.148 210.140 C16H18 1,7-hexadecadiene-
10,12,14-triyne 0.006
Flower heads of
Chrysanthemum
leucanthemum
2,859,745.14
P25741 6.66 213.148 212.141 C12 H20O3
10-hydroxy-10-
methyl-2-undecen-4-
olide
0.55
Marine bacteria
Streptomyces strain
MO2750,
Streptoverticillium
luteoverticillatum
strain 11014
738,381.18
N12863 6.93 284.130 285.138 No hits No hits 376,501.26
P48148 7.65 247.148 246.140 C19H18 1,7-diphenyl-1,3,5-
heptatriene −0.10
Rhizomes of Curcuma
comosa 120,097.45
P46440 8.24 227.179 226.172 C17H22
1,3,7,9,13-
heptadecapentaen-
11-yne
−0.52 Various Asteraceae
species 75,738.67
N22740 8.76 261.149 262.157 C16 H22O3
Paecilocin A or D
Spartinoxide
Solanapyrone H
Versiol
1.02
Marine-derived
Paecilomyces varioti
Marine-derived
Phaeosphaeria spartina
Marine-derived
Microsphaeropsis sp.
Stamm 6288
Marine-derived
Aspergillus versicolor,
Penicillium
decumbens, P.
striatisporum
537,461.67
P1561 10.79 250.112 249.104
C
10
H
21
NO
2
P
2
C3H11N11 O3No hits 2.16 ×108
P259 14.24 367.226 366.219 C24 H30O3Diorcinol K −0.79
Marine-derived
Aspergillus sp.
CUGB-F046
2.25 ×109
N23146 15.07 501.251 502.258 C30H30 N8No hits 1.61 ×109
N2762 16.88 559.283 560.290 C31H45O7POxydifficidin
Proticin 0.03 Bacillus subtilis 4.99 ×108

Microorganisms 2024,12, 413 5 of 20
Table 1. Cont.
Mzmine
ID *
RT
(min) MS m/zExact
Mass
Chemical
Formula Name
Tolerance
Source Peak Area
N53 18.84 469.258
470.265
3.91 ppm
1.80 ppm
2.18 ppm
C28H38 O6
DBE = 10
C24H34 N6O4
DBE = 11
C23H38 N2O8
DBE = 6
7-O-(2E-butenoyl)
macrolactin A
No hits
No hits
−0.19 Marine-derived
Bacillus subtilis B5 5.39 ×109
P10773 20.10 154.990 153.982 C6H3O3P No hits 274,191.89
P34624 20.77 358.331 357.324 C21H43 NO3
2-amino-
heneicosene-1,3,4-
triol
−0.58
Helichrysum
cameroonense
Parinari hypochrysea
Acnistus arborescens
14,120.53
P14139 23.81 447.325 446.318 C31 H42 O2
8′-apo-β-caroten-8′-
al; 8′-carboxylic acid,
Me ester
−0.27 Staphylococcus aureus 8.37 ×108
P41851 25.36 447.325 446.318 C31 H42 O2Same as P14139 −0.10 Staphylococcus aureus 5.59 ×108
P38446 26.33 994.643 993.636
C
50
H
87
N
7
O
13 Surfactin B1,
4-L-alanine analog 0.50 Bacillus subtilis S499 3.31 ×108
P1275 27.32 1008.659 1007.652
C
51
H
89
N
7
O
13
Surfactin C1,
4-L-alanine analog;
Surfactin A1,A2,A3
0.70 Bacillus subtilis S499 1.67 ×109
P10881 27.7 1008.659 1007.652
C
51
H
89
N
7
O
13
Surfactin C1,
4-L-alanine analog;
surfactin A1,A2,A4
0.59 Bacillus subtilis S500 4.70 ×108
P1149 28.6 256.263 255.256 C16H33NO
11E,12-dihydro-2S-
amino-11,15-
hexadecadien-3R-ol
−0.70 Pseudodistoma
obscurum 3.61 ×107
P211 28.89 1022.675 1021.667
C
52
H
91
N
7
O
13
Surfactin C1,
7-L-valine analog,
surfactin B1, B2
0.49
Bacillus subtilis and
Bacillus pumilus
KMM 456
2.16 ×109
P310 29.27 1022.675 1021.667
C
54
H
98
N
5
O
7
P
3
C
44
H
97
N
9
O
13
P
2
C
49
H
100
NO
18
P
C
47
H
88
N
15
O
8
P
C
49
H
95
N
13
O
2
P
4
C
64
H
99
NOP
4
C
56
H
105
N
3
OP
6
C
57
H
89
N
11
O
2
P
2
C
52
H
91
N
7
O
13
C
46
H
104
N
7
O
7
P
5
No hits 3.51 ×109
P6562 30.13 1034.675 1033.668
C
57
H
105
N
3
OP
6
C
50
H
95
N
13
O
2
P
4
C
32
H
13
N
2
O
9
P
15
C
47
H
104
N
7
O
7
P
5
C
45
H
97
N
9
O
13
P
2
C
28
H
21
N
4
O
2
P
19
C
55
H
98
N
5
O
7
P
3
C
54
H
87
N
11
O
9
C
42
H
101
N
15
O
2
P
6
C
50
H
100
NO
18
P
No hits 2,879,678.11
P6206 31.14 1036.691 1035.684
C
53
H
93
N
7
O
13
Surfactin B2, 1-Me
ester, surfactin C1
7-L-isoleucine analog,
C2 7-L-valine analog
0.88 Bacillus sp. 1.12 ×108
P15049 32.6 415.357 414.349 C28 H46O27-methoxycholesta-
5,22-dien-3-ol 0.41 Bryozoa Cryptosula
pallasiana 318,610.49
P24946 37.61 217.158 216.151 C15H20 O Anaephene C −1.40 Marine-derived
Hormoscilla sp. 1,503,069.04
* The letters P and N represent the ionization mode.

Microorganisms 2024,12, 413 6 of 20
Microorganisms 2024, 12, x FOR PEER REVIEW 6 of 21
P15049 32.6 415.357 414.349 C28H46O2 7-methoxycho-
lesta-5,22-dien-3-ol 0.41 Bryozoa Cryptosula
pallasiana 318,610.49
P24946 37.61 217.158 216.151 C15H20O Anaephene C −1.40 Marine-derived Hor-
moscilla sp. 1,503,069.04
* The letters P and N represent the ionization mode.
The methanolic extract (13.4 g) was fractionated using normal phase flash chroma-
tography resulting in 16 different fractions with the last fraction being the most polar
(Scheme 1). To proceed with efficient isolation of bioactive metabolites, these fractions
were tested for their inhibitory activity against planktonic cells of MRSA ATCC 43300 and
for their capacity to destabilize the formed MRSA biofilms. Only two fractions F9 and F15
inhibited about 50% planktonic MRSA at a concentration of 100 µg/mL, whereas excellent
dispersion of formed biofilms of up to 90% was observed with fractions F2, F3, F4, F10,
F11, F13, and F15 (Scheme 1 and Figure 3A). Based on the results of the post-biofilm inhi-
bition assay, the bioactive fractions were further purified by either preparatory TLC or
Biotage flash chromatography. Moreover, the metabolite profile of the fractions was
dereplicated using LC-HRMS before multivariate analysis. After the successive high-
throughput chromatographic separation, a total of 37 subfractions were obtained. All the
subfractions were tested for their anti-MRSA activity and post-biofilm inhibition of MRSA
biofilms. Only two subfractions F10.6.1 and F17 exhibited activity >50% against planktonic
MRSA cells at a concentration of 100 µg/mL. In the case of post-biofilm inhibition, nine
subfractions, i.e., F2.5, F3.3, F10.5.2, F11.2, F11.3, F13.2, F15.4.1, F16.3, and F16.4, displayed
>90% inhibition at the significant concentration of 100 µg/mL (Figure 3B). The pure frac-
tions were then subjected to NMR analysis for identification and structure elucidation
considering their bioactivity.
Scheme 1. Isolation procedure of bioactive compounds from the endophytic bacteria B. velezensis
7NPB-3B.
Scheme 1. Isolation procedure of bioactive compounds from the endophytic bacteria B. velezensis
7NPB-3B.
Microorganisms 2024, 12, x FOR PEER REVIEW 7 of 21
Figure 3. Bioassay against methicillin-resistant S. aureus. (A) Inhibition of planktonic MRSA (top)
and biofilms (bottom) on treatment by fractions of B. velezensis 7NPB-3B. (B) Inhibition of planktonic
MRSA (top) and biofilms (bottom) on treatment by the respective subfractions. The results shown
were an average from three replicates. The red frames highlights the fractions with bioactivity >90%.
Secondary metabolites present in the B. velezensis 7NPB-3B were tentatively derepli-
cated using high-resolution LC-MS against the Dictionary of Natural Products (DNP ver-
sion 2023) database. The relationship between the bioactive fractions and their metabolites
was statistically evaluated using SIMCA. The unsupervised PCA scores plot showed the
uniqueness and similarity of metabolites present in the fractions. Variable samples that
clustered together indicated similarity in their chemical profile, while the dispersal of the
fractions implied the differences. The farther the samples from the center axis of the scores
plot, the higher the metabolite diversity for the respective fractions (Figure 4A). F9 was
observed to be the outlier in the PCA analysis suggesting a unique chemical profile of the
sample. The supervised multivariate analysis demonstrated by the OPLS-DA scores plots
shown in Figure 4C depicted the differences between the two groups of fractions, F2, F3,
F4, F5, F10, F11, F13, and F15 as the active fractions and F1, F6, F7, F8, F9, F12, F14, F16,
and F17 for the inactive fractions. The active fractions were discriminated by ions peaks
at m/z [M + H]+ 245.127 at 6.40 min, 296.148 at 6.77 min, 349.215 at 15.83 min, 367.225 at
17.10 min, 385.236 at 15.95 min, and 420.272 at 16.26 min, while for [M−H]−, discriminating
ions peaks were observed at m/z 174.055 at 6.59 min, 447.239 at 16.23 min, and 501.250 at
14.78 min. Dereplication data of these ion peaks are listed in Table 2. It was noticeable that
none of the surfactin congeners earlier detected from the TIC of the fractions were per-
ceived as target metabolites for antibiofilm activity. Bioactive metabolites were detected
between 200 and 500 Da with the occurrence of macrolactin congeners as the typical Ba-
cillus secondary metabolites for ions peaks at m/z 385.236 [M + H]+ and 501.250 [M−H]−
[19].
Figure 3. Bioassay against methicillin-resistant S. aureus. (A) Inhibition of planktonic MRSA (top)
and biofilms (bottom) on treatment by fractions of B. velezensis 7NPB-3B. (B) Inhibition of planktonic
MRSA (top) and biofilms (bottom) on treatment by the respective subfractions. The results shown
were an average from three replicates. The red frames highlights the fractions with bioactivity > 90%.

Microorganisms 2024,12, 413 7 of 20
Secondary metabolites present in the B. velezensis 7NPB-3B were tentatively derepli-
cated using high-resolution LC-MS against the Dictionary of Natural Products (DNP
version 2023) database. The relationship between the bioactive fractions and their metabo-
lites was statistically evaluated using SIMCA. The unsupervised PCA scores plot showed
the uniqueness and similarity of metabolites present in the fractions. Variable samples
that clustered together indicated similarity in their chemical profile, while the dispersal of
the fractions implied the differences. The farther the samples from the center axis of the
scores plot, the higher the metabolite diversity for the respective fractions (Figure 4A). F9
was observed to be the outlier in the PCA analysis suggesting a unique chemical profile of
the sample. The supervised multivariate analysis demonstrated by the OPLS-DA scores
plots shown in Figure 4C depicted the differences between the two groups of fractions,
F2, F3, F4, F5, F10, F11, F13, and F15 as the active fractions and F1, F6, F7, F8, F9, F12,
F14, F16, and F17 for the inactive fractions. The active fractions were discriminated by
ions peaks at m/z[M + H]
+
245.127 at 6.40 min, 296.148 at 6.77 min, 349.215 at 15.83 min,
367.225 at 17.10 min, 385.236 at 15.95 min, and 420.272 at 16.26 min, while for [M
−
H]
−
,
discriminating ions peaks were observed at m/z174.055 at 6.59 min, 447.239 at 16.23 min,
and 501.250 at 14.78 min. Dereplication data of these ion peaks are listed in Table 2. It
was noticeable that none of the surfactin congeners earlier detected from the TIC of the
fractions were perceived as target metabolites for antibiofilm activity. Bioactive metabolites
were detected between 200 and 500 Da with the occurrence of macrolactin congeners as the
typical Bacillus secondary metabolites for ions peaks at m/z385.236 [M + H]
+
and 501.250
[M −H]−[19].
Microorganisms 2024, 12, x FOR PEER REVIEW 8 of 22
MRSA (top) and biofilms (bottom) on treatment by the respective subfractions. The results shown
were an average from three replicates. The red frames highlights the fractions with bioactivity >90%.
Secondary metabolites present in the B. velezensis 7NPB-3B were tentatively
dereplicated using high-resolution LC-MS against the Dictionary of Natural Products
(DNP version 2023) database. The relationship between the bioactive fractions and their
metabolites was statistically evaluated using SIMCA. The unsupervised PCA scores plot
showed the uniqueness and similarity of metabolites present in the fractions. Variable
samples that clustered together indicated similarity in their chemical profile, while the
dispersal of the fractions implied the differences. The farther the samples from the center
axis of the scores plot, the higher the metabolite diversity for the respective fractions
(Figure 4A). F9 was observed to be the outlier in the PCA analysis suggesting a unique
chemical profile of the sample. The supervised multivariate analysis demonstrated by the
OPLS-DA scores plots shown in Figure 4C depicted the differences between the two
groups of fractions, F2, F3, F4, F5, F10, F11, F13, and F15 as the active fractions and F1, F6,
F7, F8, F9, F12, F14, F16, and F17 for the inactive fractions. The active fractions were
discriminated by ions peaks at m/z [M + H]+ 245.127 at 6.40 min, 296.148 at 6.77 min, 349.215
at 15.83 min, 367.225 at 17.10 min, 385.236 at 15.95 min, and 420.272 at 16.26 min, while for
[M−H]−, discriminating ions peaks were observed at m/z 174.055 at 6.59 min, 447.239 at
16.23 min, and 501.250 at 14.78 min. Dereplication data of these ion peaks are listed in
Table 2. It was noticeable that none of the surfactin congeners earlier detected from the
TIC of the fractions were perceived as target metabolites for antibiofilm activity. Bioactive
metabolites were detected between 200 and 500 Da with the occurrence of macrolactin
congeners as the typical Bacillus secondary metabolites for ions peaks at m/z 385.236 [M +
H]+ and 501.250 [M−H]− [19].
Figure 4. Multivariate analysis of the mass spectral data of fractions obtained from the organic
extract of Bacillus velezensis 7NPB-3B. (A) Unsupervised principal component analysis (PCA) scores
and (B) loadings plots show moderate separation between the datasets. (C) Supervised orthogonal
projections to latent structures discriminant analysis (OPLS-DA) scores plot and (D) loadings S-plot
differentiating the metabolites between active and inactive fractions listed in Table 2.
Commented [M2]: There are parts of Figures 5A
and 5c in the right is missing (x-axis), please
provide a complete version of this figure. Also,
please considering using scientific notation
instead of exponential notation for number that
reaches five figures. Thank you!
Commented [RE3R2]: No parts are missing in
Figures 4A and 4C to affect the interpretation of
the data. The plots we generated the software
SIMCA and the numbers cannot be altered as
generated the software. The missing end numbers
on the x-axis were then deleted.
Figure 4. Multivariate analysis of the mass spectral data of fractions obtained from the organic
extract of Bacillus velezensis 7NPB-3B. (A) Unsupervised principal component analysis (PCA) scores
and (B) loadings plots show moderate separation between the datasets. (C) Supervised orthogonal
projections to latent structures discriminant analysis (OPLS-DA) scores plot and (D) loadings S-plot
differentiating the metabolites between active and inactive fractions listed in Table 2.

Microorganisms 2024,12, 413 8 of 20
The antibiofilm compounds discriminated by multivariate analysis were dereplicated
with DNP and listed in Table 2. The OPLS-DA scores plot of the subfractions indicated
a unique chemistry for F13.2 (Figure 5A), which separated it from the rest of the active
subfractions with a mass ion peak at m/z863.06 [M
−
H]
−
; however, its molecular formula
cannot be predicted following the seven heuristic rules [
20
]. In parallel to the separation
of F13.2 from most of the active fractions, there was also an overlapping of weaker an-
tibiofilm active fractions to the inactive right quadrant yielding low fitness R2 (0.65) and
predictability Q2 (0.08) scores. The percentage variability scores were also low at only
0.04% between groups and 24% within groups, where the variability score must be higher
between groups to indicate good separation. As presented in Table 3, the discriminating
features from the active quadrant were not found in any of the fractions overlapping with
those of the inactive quadrant; these subfractions (F2-5, F10-2-5, F10-4-4, F10-5-1, F10-5-2,
F10-5-4, F11-2, F11-3, F13-3, F16-1, F16-4, and F17) were then excluded. This improved the
R2 and Q2 scores to 0.99 and 0.46. Due to the low predictability score, this was validated
through an observed vs. predicted plot (Figure 5B) to reveal the observed versus predicted
values of the active Y-variable. Except for F11-4, F13-4, F13-4C, and F16-2, all the variables
must fall close to the 45-degree line to attain a good model. But the R2 of the regression
line afforded a goodness of fit of 0.99 cross-validating the predicted Y-variables for the
active components.
Table 2. Dereplicated antibiofilm target metabolites from the OPLS-DA loadings plot of bioactive
fractions using DNP. The highlighted row indicates the isolated and elucidated compound in this
study. * The letters P and N represent the ionization mode.
Primary ID * m/zRT (min)
Fraction
Found
Exact
Mass
Molecular
Formula Dereplicated Compound Biological Source
N2994
174.055
6.59 FR4 175.0622 No
prediction
P6029
245.127
6.40 FR10
FR11 244.12051 C14H16 N2O2Cyclo(phenylalanylprolyl)
(5)
Produced by the
marine bacterial
strains CF-20 and
C-148 obtained from
molluscs
P29373
296.148
6.77 FR4 295.1411 C15H21 NO5Coronafacic acid; L-serine
amide
Produced by
Pseudomonas syringae
P55
349.215
15.83
FR3
FR4
FR5
348.2075 C18H33 ClO410R-chloro-plakortether B Plakortis simplex
P16
367.225
15.93
FR3
FR4
FR5
366.2181 C24H30 O3Diorcinol K
Marine-derived
Aspergillus sp.
CUGB-F046
P239
385.236
15.37 FR3 384.2286 C24H32 O47,13-epoxymacrolactin A Marine-derived
Bacillus subtilis B5
P6477
420.272
16.26 FR3 419.2652 C24H37 NO5Aspochalasin E
Marine-derived
Aspergillus sp.
XS-2009-0B15
N3227
447.239
16.23
FR3
FR4
FR5
448.2467 C25H36 O7Antibiotic L 731128 Sporormiella
intermedia MG 5447
N26607 501.25 14.78 FR5 502.2576 C28 H38O87-succinoyl-macrolactin A
Bacillus
amyloliquefaciens
NJN-6,B. subtilis
DSM 16696,marine
Bacillus sp. Sc026

Microorganisms 2024,12, 413 9 of 20
Microorganisms 2024, 12, x FOR PEER REVIEW 10 of 22
and F4.2. Amongst the detected antibiofilm features included diketopiperazines
derivatives cyclo(tryptophanyltyrosyl), cyclo(isoleucylisoleucyl), and
cyclo(isoleucylprolyl) or their corresponding leucyl isomers found at m/z 348.136
[M−H]−/350.149 [M + H]+, 227.174 [M + H]+, and 211.144 [M + H]+, respectively, compatible
with the exact masses 349.143, 226.167, and 349.143 Da. Bacillus species have been reported
to produce compounds of the diketopiperazines class [21–23]. The features present in the
active quadrant also included the ion peak at m/z 447.239, putatively identified as
Antibiotic L731128, earlier described from Sporormiella intermedia MG 5447 [24]. A related
ion peak was preliminarily screened from the antibiofilm fractions FR3, FR4, and FR5
eluting at 16.23 min (Table 2). Another ion peak at m/z 557.331 [M + H]+ provided a
compound hit for turnagainolide A, a known antibacterial earlier isolated from the
marine-derived Bacillus sp. RJA2194 [25].
Figure 5. OPLS-DA of the mass spectral data of semi-purified subfractions of the active fractions
shown in Figure 4. (A) Scores between active and inactive fractions. (B) The observed vs. predicted
plot displays the observed versus predicted values of the active Y-variable. With a good model, all
the points must fall close to this 45-degree line. The R2 of the regression line afforded a goodness of
fit close to 1.0. (C) Loadings plot differentiating between the metabolites of active and inactive
fractions. Encircled features represent the discriminating antibiofilm-active metabolites listed in
Tab le 2. ( D) Extracted loadings plot for metabolites between 200 and 300 Da to show the positions
of the isolated compounds indicated by a star on the left quadrant for the active samples.
Meanwhile, the ion peaks at m/z 270.205, 469.258, and 471.273 remained unidentified
after the dereplication of active fractions (Table 3 and Figure S1). The ion peak at m/z
270.205 [M + H]+ yielded a molecular formula prediction for C11H23N7O with a DBE of 4
for the exact mass of 269.197 Da. The ion peaks at m/z 469.258 [M−H]− and 471.270 [M +
H]+ eluting at 18.91 and 18.81 min, respectively, are compatible with the exact mass 470.265
with predicted molecular formulas of C23H38N2O8 (DBE = 6), C24H34N6O4 (DBE = 11), and
C28H38O6 (DBE = 10) with an accuracy of 2.18, 1.80, and 3.91 ppm, respectively. A similar
differentiating ion peak at m/z 469.258 [M−H]− with a retention time of 18.84 min was
detected from the TIC of the crude extracts (Table 1). The latter prediction of C28H38O6
Commented [M4]: There is a part of Figure 5A in
the right is missing (x-axis 300….), please provide
a complete version of this figure. Also, please
considering using scientific notation instead of
exponential notation for number that reaches five
figures. Thank you!
Commented [RE5R4]: No parts are missing in
Figures 5A to affect the interpretation of the data.
The plots we generated the software SIMCA and
the numbers cannot be altered as generated the
software. The missing end numbers on the x-axis
were then deleted.
Figure 5. OPLS-DA of the mass spectral data of semi-purified subfractions of the active fractions
shown in Figure 4. (A) Scores between active and inactive fractions. (B) The observed vs. predicted
plot displays the observed versus predicted values of the active Y-variable. With a good model, all
the points must fall close to this 45-degree line. The R2 of the regression line afforded a goodness of
fit close to 1.0. (C) Loadings plot differentiating between the metabolites of active and inactive frac-
tions. Encircled features represent the discriminating antibiofilm-active metabolites listed in Table 2.
(D) Extracted loadings plot for metabolites between 200 and 300 Da to show the positions of the
isolated compounds indicated by a star on the left quadrant for the active samples.
Table 3. Dereplicated antibiofilm target metabolites from the OPLS-DA loadings plot of subfractions
using DNP. Features are arranged and listed according to their ion peaks at m/zin Da. Highlighted
rows indicate the isolated and elucidated compounds in this study. Structures of dereplicated
compounds are shown in Figure S1.
Primary ID * m/zRT
(min)
Subfraction
Exact
Mass
Molecular
Formula
Dereplicated
Compound Biological Source
P5 211.144 5.14 F6-2 210.136 C11H18N2O2Cyclo(isoleucylprolyl)
Marine-derived
Pseudomonas
aeruginosa and Vibrio
parahaemolyticus
Marine bacterial
strains CF-20 and
C-148 isolated from
larvae of the mollusc
Pecten maximus
P22 211.144 5.19 F6s 210.136 C11 H18 N2O2Same as P5
P151 227.174 7.41 F4-2 226.167 C12H22N2O2
Cyclo(isoleucylisoleucyl)
Marine-derived
Paecilomyces
marquandii
P185 227.175 8.12 F10-2-4 226.167 C12 H22N2O2Same as P151
P318 270.205 10.76 F5-5 269.197 C11H23 N7O
DBE = 4 No hits
N32 348.136 6.75 F15-3-5 349.143 C20 H19N3O3Cyclo(tryptophanyl-
tyrosyl)
Marine-derived
Aspergillus niger
EN-13 and from B.
subtilis B38
P108 350.149 6.75 F15-3-5 349.142 [M + H]+of
N32 Same as N32

Microorganisms 2024,12, 413 10 of 20
Table 3. Cont.
Primary ID * m/zRT
(min)
Subfraction
Exact
Mass
Molecular
Formula
Dereplicated
Compound Biological Source
N157 447.239 14.21 F5-5 448.246 C25H36O7Antibiotic L 731128 Sporormiella
intermedia MG 5447
N289 469.258 18.54 F15-4-1 470.265 Same as
N299
N299 469.258 18.91 F15-4-1 470.265 [M −H]−of
P802 No hits
P802 471.273 18.81 F15-4-1
470.265
(∆ppm)
(2.18)
(1.80)
(3.91)
C23H38 N2O8
DBE = 6
C24H34 N6O4
DBE = 11
C28H38 O6
DBE = 10
No hits
No hits
7-O-(2E-butenoyl)
macrolactin A
Marine-derived
Bacillus subtilis B5
P611 557.331 15.98 F3-3 556.324 C30H44 N4O6
DBE = 11 Turnagainolide A
Marine-derived
Bacillus sp. RJA2194
and Microascus sp.
098059A.
Arthrobacter sp.
PGVB1 and
Streptomyces
rutgersensis T00
N40 863.067 7.12 F13-2 864.074 No
prediction
N59 863.067 8.26 F13-2 864.074 No
prediction
* The letters P and N represent the ionization mode.
The OPLS-DA loadings plot showed the distribution of discriminating metabolites at
the ends of both active and inactive quadrants (Figure 5C), further confirming the bioactive
masses to be present in the subfractions F15.4.1, F13.2, F10.2.4, F15.3.5, F3.3, F5.5, and
F4.2. Amongst the detected antibiofilm features included diketopiperazines derivatives
cyclo(tryptophanyltyrosyl), cyclo(isoleucylisoleucyl), and cyclo(isoleucylprolyl) or their
corresponding leucyl isomers found at m/z348.136 [M
−
H]
−
/350.149 [M + H]
+
, 227.174
[M + H]
+
, and 211.144 [M + H]
+
, respectively, compatible with the exact masses 349.143,
226.167, and 349.143 Da. Bacillus species have been reported to produce compounds of the
diketopiperazines class [
21
–
23
]. The features present in the active quadrant also included
the ion peak at m/z447.239, putatively identified as Antibiotic L731128, earlier described
from Sporormiella intermedia MG 5447 [
24
]. A related ion peak was preliminarily screened
from the antibiofilm fractions FR3, FR4, and FR5 eluting at 16.23 min (Table 2). Another ion
peak at m/z557.331 [M + H]
+
provided a compound hit for turnagainolide A, a known
antibacterial earlier isolated from the marine-derived Bacillus sp. RJA2194 [25].
Meanwhile, the ion peaks at m/z270.205, 469.258, and 471.273 remained unidentified
after the dereplication of active fractions (Table 3and Figure S1). The ion peak at m/z
270.205 [M + H]
+
yielded a molecular formula prediction for C
11
H
23
N
7
O with a DBE of 4 for
the exact mass of 269.197 Da. The ion peaks at m/z469.258 [M
−
H]
−
and 471.270
[M + H]+
eluting at 18.91 and 18.81 min, respectively, are compatible with the exact mass 470.265
with predicted molecular formulas of C
23
H
38
N
2
O
8
(DBE = 6), C
24
H
34
N
6
O
4
(
DBE = 11
),
and C
28
H
38
O
6
(DBE = 10) with an accuracy of 2.18, 1.80, and 3.91 ppm, respectively.
A similar differentiating ion peak at m/z469.258 [M
−
H]
−
with a retention time of
18.84 min was detected from the TIC of the crude extracts (Table 1). The latter prediction of
C
28
H
38
O
6
(
DBE = 10
) was dereplicated as 7-O-(2E-butenoyl) macrolactin A, which was also
previously reported from the marine-derived Bacillus subtilis B5 [
19
]. However, the proton

Microorganisms 2024,12, 413 11 of 20
NMR of the semi-purified fraction F15-4-1 did not match the structure of 7-O-(2E-butenoyl)
macrolactin A.
The pre-defined bioactive metabolites obtained by multivariate analysis were able
to guide the further purification and isolation of the antibiofilm compounds. During the
purification step, the isolation of the compound with an ion peak at m/z469.258
[M −H]−
and retention time at 18.8 min was targeted. In parallel, a series of diketopiperazine
compounds and mixtures of other low-molecular-weight compounds ranging between
200 and 300 Da are being detected on the active fractions. These were again subjected
to OPLS-DA, and these compounds were indeed positioning themselves on the active
quadrants of the loadings plots as demonstrated in Figure 5D.
The study aimed to isolate antibiofilm compounds using metabolomics and bioassay-
guided separation from the halophytic endophyte B. velezensis 7NPB-3B. Repeated chro-
matographic separation of the organic crude extract afforded a new metabolite (1) and four
other previously isolated cyclic dipeptides (2to 5). The structure elucidation and identifica-
tion using 1D- and 2D-NMR with mass spectrometry led to identification of compounds
1(m/z469.358 [M
−
H]
−
),cyclo-Phe-Leu (m/z259.145 [M
−
H]
−
)2, cyclo-Phe-Val (m/z
245.128 [M
−
H]
−
)3, cyclo-Leu-Val (m/z213.159 [M + H]
+
)4, and cyclo-Phe-Pro (m/z
243.112 [M
−
H]
−
)5(Figure 5). The presence of some cyclic dipeptides also known as
diketopiperazines 2to 5was initially predicted through LC-MS dereplication. Thus, such
results validated the accuracy of the dereplication process. The structures of the cyclic
dipeptides were elucidated by 1D and 2D NMR as well as corroborated with the literature
(Table S1 and Figure S2A) [
26
–
30
]. Comparison of the CD spectra of compounds 2to 5
(Figure S2B) to those found in the literature followed the 2S,2
′
S(3S,6Sfor 2to 4and 3S,8aS
for 5in their IUPAC nomenclature) [31–34].
Compound 1: Compound 1(50 mg) was isolated as a dark brown oil, and its molecular
formula was determined by high-resolution mass spectrometry as C
23
H
38
N
2
O
8
based on
the ion peak at m/zof 469.258 [M
−
H]
−
with 6 degrees of unsaturation and an optical
rotation of [
α
]
20−
23.60
◦
(c 1.00, MeOH). The compound showed UV (MeOH) absorption
maxima at 207 nm (log ε4.24) and 260 nm (log ε3.43).
The
1
H NMR (DMSO-d
6
) analysis of compound 1(Table 4) exhibited the presence
of three spin systems. The
1
H NMR spectrum (Figure S3) indicated the presence of a
dipeptide group with characteristic alpha proton signal at
δH
4.20 (m, H-2), two methylene
group protons at
δH
1.83 (m, H-4a), 2.04 (m, H-4b) and 2.89 (dtd, H-5a), 2.76 (m, H-5b),
respectively, corresponding to a glutamic acid moiety terminating with a carboxylic acid
group. Similarly, the other amino acid of the dipeptide corresponds to an unusual lysine
unit detected with characteristic amine protons signals at
δH
8.23 (d, J = 8.3 Hz), alpha
proton at
δH
4.18 (m, H-8), four methylene groups at
δH
1.84 (m, H-10), 2.15 (ddt, H-11,
H-12), 1.86 (m, H-11, H-12), and 3.38 (m, H-13). However, the amine group of lysine was
instead substituted with an acid moiety. The downfield shift of the H-13 methylene group
is due to the presence of the electron-withdrawing acid group at C-14. Both amino acids
were connected to form a diketopiperazine ring (DKP). Furthermore, the broad singlet
at
δH
1.23 (H-17 to H-22) was observed indicating the presence of an alkyl fatty acid
chain. Detailed analysis of DEPT NMR spectrum (Figure S4) along with HMBC spectrum
(Figure S5) showed the presence of 5 carbonyl carbons at
δC
165.2 ppm and 170.1 ppm
corresponding to the amide bonds (C-1, C-7), in DKP ring as well for acid groups (C-6,
C-14, C-24) along the side chains. Two alpha carbon signals of the two amino acids showed
their characteristic carbon chemical shifts at 53.5 ppm (C-2) and 58.5 ppm (C-8). The
methylene carbons of glutamic acid had
δC
resonances at 48.8 (C-4), 22.4 (C-5), and the
same for a characteristic lysine unit with
δC
at 22.3 (C-10), 27.7 (C-11), 22.3 (C-12), and 44.9
(C-13). However, the downfield shifts of C-5, C-13, and C-23 pertained to the existence of a
carboxylic acid in their periphery. The methylene groups for the fatty acid chain appeared
at
δC
28.9.
1
H-
1
H COSY couplings (Figure S6) from H-2 to H-4 and H-5 along with HMBC
correlations of H-4 to C-2, C-5, C-6 and H-2 to C-5 and C-1 confirmed the presence of
glutamic acid. Similarly, COSY couplings from H-8 to H-9, H-10 to H-11, and H-12 to H-13

Microorganisms 2024,12, 413 12 of 20
supported by HMBC correlations H-8 to C-7, C-10; H-10 to C-11; H-11 to C-7, C-8, C-12;
H-12 to C-13, C-14; and H-13 to C-11, C-10 confirmed prevalence of the atypical lysine
unit. The diketopiperazine ring formation is confirmed by the HMBC correlation of the H-9
amine proton of the unusual amino acid to C-1 and C-2, i.e., carbonyl and alpha carbon of
glutamic acid, respectively. The four-bond HMBC correlation of H-16 and C-14 reveals the
connection of the fatty acid alkyl chain with diketopiperazine moiety and cross-peaks from
H-22 to C-24 further verify the carboxylic acid attached at the end of a saturated alkyl chain
that is decanoic also known as capric acid. Capric acid (C10:0) has been reported to exhibit
potent antibacterial activity against the Gram-negative bacteria Chlamydia trachomatis upon
treatment with 10 mM of the compound [35].
Table 4. 1H NMR (400 MHz) and 13 C NMR (150 MHz) data of compound 1 (DMSO-d6).
SN Type δH(ppm, Mult. J HZ) δC(ppm) HMBC
1 CO 165.2
2 CH 4.20 (m) 53.5 C1, C4, C5
3 N
4 CH22.89 (dtd, J= 12.9, 10.6, 10.7, 5.6 Hz)
2.76 (m)48.8 C-2, C-5, C-25
5 CH22.04 (m)
1.83 (m)22.4 C-2, C-4, C-6
6 CO 165.2
7 CO 170.1
8 CH 4.18 (m) 58.5 C-9, C-11
9 NH8.23 (d,J= 8.3 Hz) C-1, C-2, C-8,
C-10
10 CH21.84 (m) 22.3 C-8, C-9, C-11
11 CH22.15 (ddt, J= 11.5, 9.8, 4.0 Hz)
1.86 (m)27.7 C-7, C-8, C-10,
C-12
12 CH22.15 (ddt, J= 11.5, 9.8, 4.0 Hz)
1.84 (m)22.3 C-13, C-14
13 CH23.38 (m) 44.9 C10, C-11, C-12
14 CO 170.1
15 CH23.31 (m) 62.3
16 CH21.83 (m) 22.4 C-14, C-17
17 CH21.23 (s) 28.9
18 CH21.23 (s) 28.9
19 CH21.23 (s) 28.9
20 CH21.23 (s) 28.9
21 CH21.23 (s) 28.9
22 CH21.23 (s) 28.9 C-24
23 CH23.38 (m) 44.9
24 CO 170.1
25 CH32.54 (s) 37.9 C-4
The location of the methyl group was elucidated to be at N-3 based on the
4
J-HMBC
correlation from H-25 to C-4 (Figures 6, S5 and S6). The chemical shifts of CH
2
-4 were
typical to those of a
γ
-position to a carbonyl unit at
δH
2.89 and 2.76,
δC
48.8, which were
downfield to that of the
β
-position that was shifted upfield at
δH
2.04 and 1.83,
δC
22.4.
This was further corroborated by the presence of an amine proton on the atypical amino
acid unit and its absence on the glutamic acid moiety [
36
,
37
]. The analysis of 1D- and
2D-NMR spectra confirmed the planar structure of 1(Figures 6and S4–S8). The structure
of 1was further validated by the MS/MS fragmentation pattern that resulted in fragments
m/z453.26 and 299.25 (Figures 7B and S8). Compound 1was elucidated as 10-((5-(5-(2-
carboxyethyl)-4-methyl-3,6-dioxopiperazin-2-yl)pentanoyl)oxy)decanoic acid. The ECD
spectrum for compound 1is shown in Figure S9.

Microorganisms 2024,12, 413 13 of 20
Microorganisms 2024, 12, x FOR PEER REVIEW 13 of 21
NMR spectra confirmed the planar structure of 1 (Figures 6 and S4–S8). The structure of 1
was further validated by the MS/MS fragmentation pattern that resulted in fragments m/z
453.26 and 299.25 (Figures 7B and S8). Compound 1 was elucidated as 10-((5-(5-(2-carbox-
yethyl)-4-methyl-3,6-dioxopiperazin-2-yl)pentanoyl)oxy)decanoic acid. The ECD spec-
trum for compound 1 is shown in Figure S9.
HN
NH
O
O
HN
NH
O
O
(2)
(3)
HN
N
O
O
(4)
(5)
NH
HN
O
O
N
N
H
O
HO
O
O
O
O
OH
O
(1)
1
3
4
67
914 24
25
Figure 6. Structures of isolated compounds from B. velezensis 7NPB-3B.
N
N
H
O
HO
O
O
O
O
OH
N
N
H
O
HO
O
O
O
O
Chemical Formula: C23H37N2O7+
Exact Mass: 453.2595
N
N
H
O
HO
O
O
OH
O
OH
Chemical Formula: C13H20N2O6•+
Exact Mass: 300.1316
O
O
Chemical Formula: C23H38N2O8
Exact Mass: 470.26
O
N
N
H
O
HO
O
O
O
O
OH
O
(B)
(A)
COSY
Key HMBC
Figure 6. Structures of isolated compounds from B. velezensis 7NPB-3B.
Microorganisms 2024, 12, x FOR PEER REVIEW 14 of 22
Figure 7. (A) COSY and key HMBC correlations of compound 1. (B) The fragmentation pattern of
compound 1 as obtained by MS/MS fragmentation confirming the structure of 1.
Diketopiperazines are well-reported for their antibiofilm abilities. These molecules
are reported to influence LuxR-mediated quorum sensing systems and offer an alternate
mechanism for disrupting bacterial communication systems [38]. Compound 1 and cyclo-
(Phe-Pro) 5 were found to exhibit biofilm dispersal ability at 100 µg/mL against MRSA
biofilms via alamar blue assay. Upon further testing, the minimal biofilm eradication
concentration (MBEC) for 1 and 5 was determined to be 25 µg/mL and 35 µg/mL,
respectively, keeping the positive control ciprofloxacin with an MBEC of 3.125 mg/mL
against MRSA ATCC 43300 (Figures 8 and S10). The MBEC calculation validated that the
1 and 5 have a potent biofilm eradication capacity in comparison to the positive control
and hence shall be tested further for the development as a lead molecule. Compounds 2
to 4 also exhibited post-biofilm inhibition potential, but the activities were not so
noteworthy. Compound 5 contains proline amino acid in contrast to 2 to 4; therefore, it
can be noted that proline may be contributing towards the management of biofilms by
diketopiperazines. There have been multiple reports where the diketopiperazine cyclo-(L-
Phe-L-Pro) was found to modulate/interfere with the cell signaling of quorum sensing
systems [39–41]. The ECD of compound 5 is comparable to that of a L, L configuration as
Figure 7. (A) COSY and key HMBC correlations of compound 1. (B) The fragmentation pattern of
compound 1as obtained by MS/MS fragmentation confirming the structure of 1.

Microorganisms 2024,12, 413 14 of 20
Diketopiperazines are well-reported for their antibiofilm abilities. These molecules
are reported to influence LuxR-mediated quorum sensing systems and offer an alternate
mechanism for disrupting bacterial communication systems [38]. Compound 1and cyclo-
(Phe-Pro) 5were found to exhibit biofilm dispersal ability at 100
µ
g/mL against MRSA
biofilms via alamar blue assay. Upon further testing, the minimal biofilm eradication con-
centration (MBEC) for 1and 5was determined to be 25
µ
g/mL and 35
µ
g/mL, respectively,
keeping the positive control ciprofloxacin with an MBEC of 3.125 mg/mL against MRSA
ATCC 43300 (Figure 8and Figure S10). The MBEC calculation validated that the 1and 5
have a potent biofilm eradication capacity in comparison to the positive control and hence
shall be tested further for the development as a lead molecule. Compounds 2to 4also exhib-
ited post-biofilm inhibition potential, but the activities were not so noteworthy. Compound
5contains proline amino acid in contrast to 2to 4; therefore, it can be noted that proline
may be contributing towards the management of biofilms by diketopiperazines. There
have been multiple reports where the diketopiperazine cyclo-(L-Phe-L-Pro) was found to
modulate/interfere with the cell signaling of quorum sensing systems [
39
–
41
]. The ECD
of compound 5is comparable to that of a L,Lconfiguration as reported by Domzalski
et al. [
31
]. In another report, cyclo-(L-Phe-L-Pro) inhibited the Ptst promoter of agr quorum
sensing system of S. aureus biofilms [
42
]. Furthermore, for the bioactivity against planktonic
MRSA ATCC 43300 cells, subfraction F10.6.2 and fraction F17 demonstrated positive results.
LC-MS of F10.6.2 indicated the presence of lipopeptide surfactins with masses (m/z1006
[M
−
H]
−
, 1020 [M
−
H]
−
, 1034 [M
−
H]
−
, 1048 [M
−
H]
−
, 1074 [M + K]
+
, and 1088
[M + K]+),
but they were not purified in this work. Surfactin analogs from Bacillus sp. are
well known for their antibacterial potential [
5
]. Compounds like lipopeptide surfactins
and an antimicrobial peptide YS12 from B. velezensis exhibited antibiofilm activity against
multiple clinical pathogens including MRSA [
15
] but were found inactive in our test or-
ganisms in this study. However, diketopiperazines with antibiofilm activities have not yet
been reported from the endophytic B. velezensis strains. Furthermore, cyclic dipeptides
reported from other strains of B. velezensis mostly exhibit bioactivity against fungal plant
pathogens [
43
–
45
]. We report 1(m/z469 [M
−
H]
−
) and 2–4for the first time from an
endophytic B. velezensis strain. Therefore, the strain isolated from an untapped niche led
to the identification of new compounds, confirming the effects of the environment on the
chemical ecology of the bacteria.
Microorganisms 2024, 12, x FOR PEER REVIEW 14 of 21
Figure 7. (A) COSY and key HMBC correlations of compound 1. (B) The fragmentation pattern of
compound 1 as obtained by MS/MS fragmentation confirming the structure of 1.
Diketopiperazines are well-reported for their antibiofilm abilities. These molecules
are reported to influence LuxR-mediated quorum sensing systems and offer an alternate
mechanism for disrupting bacterial communication systems [38]. Compound 1 and cyclo-
(Phe-Pro) 5 were found to exhibit biofilm dispersal ability at 100 µg/mL against MRSA
biofilms via alamar blue assay. Upon further testing, the minimal biofilm eradication con-
centration (MBEC) for 1 and 5 was determined to be 25 µg/mL and 35 µg/mL, respectively,
keeping the positive control ciprofloxacin with an MBEC of 3.125 mg/mL against MRSA
ATCC 43300 (Figures 8 and S10). The MBEC calculation validated that the 1 and 5 have a
potent biofilm eradication capacity in comparison to the positive control and hence shall
be tested further for the development as a lead molecule. Compounds 2 to 4 also exhibited
post-biofilm inhibition potential, but the activities were not so noteworthy. Compound 5
contains proline amino acid in contrast to 2 to 4; therefore, it can be noted that proline may
be contributing towards the management of biofilms by diketopiperazines. There have
been multiple reports where the diketopiperazine cyclo-(L-Phe-L-Pro) was found to mod-
ulate/interfere with the cell signaling of quorum sensing systems [39–41]. The ECD of
compound 5 is comparable to that of a L, L configuration as reported by Domzalski et al.
[31]. In another report, cyclo-(L-Phe-L-Pro) inhibited the Ptst promoter of agr quorum
sensing system of S. aureus biofilms [42]. Furthermore, for the bioactivity against plank-
tonic MRSA ATCC 43300 cells, subfraction F10.6.2 and fraction F17 demonstrated positive
results. LC-MS of F10.6.2 indicated the presence of lipopeptide surfactins with masses (m/z
1006 [M−H]−, 1020 [M−H]−, 1034 [M−H]−, 1048 [M−H]−, 1074 [M + K]+, and 1088 [M + K]+),
but they were not purified in this work. Surfactin analogs from Bacillus sp. are well known
for their antibacterial potential [5]. Compounds like lipopeptide surfactins and an antimi-
crobial peptide YS12 from B. velezensis exhibited antibiofilm activity against multiple clin-
ical pathogens including MRSA [15] but were found inactive in our test organisms in this
study. However, diketopiperazines with antibiofilm activities have not yet been reported
from the endophytic B. velezensis strains. Furthermore, cyclic dipeptides reported from
other strains of B. velezensis mostly exhibit bioactivity against fungal plant pathogens [43–
45]. We report 1 (m/z 469 [M−H]−) and 2–4 for the first time from an endophytic B. velezensis
strain. Therefore, the strain isolated from an untapped niche led to the identification of
new compounds, confirming the effects of the environment on the chemical ecology of the
bacteria.
Figure 8. Post-biofilm MBEC determination of (A) 1 and (B) 5 against MRSA ATCC 43300.
Figure 8. Post-biofilm MBEC determination of (A)1and (B)5against MRSA ATCC 43300.
3. Materials and Methods
3.1. Isolation and Identification of the Strain
The plant Salicornia brachiata Roxb. was collected from the salt marshland of New
Port (21
◦
45
′
15.7
′′
N, 072
◦
14
′
01.4
′′
E) from Bhavnagar district, Gujarat, India. The isolation
and identification of endophytes are described in the work by Singh et al. (2021). Briefly,
the plant material was surface-sterilized, cut aseptically, and then inoculated in 8 different
media for the growth of diverse bacteria. The isolates were screened based on their
bioactive potential against a panel of pathogens. Selected bioactive endophytes were
further identified using 16s rRNA sequencing [5].

Microorganisms 2024,12, 413 15 of 20
3.2. Fermentation and Extraction
The large-scale liquid-state fermentation of B. velezensis 7NPB-3B was carried out to
obtain the crude extract for the isolation of compounds. The seed broth of the strain was
prepared using Luria Bertani (LB) broth media (1.0% Tryptone, 0.5% yeast extract, 1.0%
sodium chloride), pH (7.0–7.5), in a 500 mL conical flask and incubated at 30
◦
C, 150 rpm
for 24 h. Thereafter, a 5% (v/v) concentration of seed culture was used to inoculate 60
×
1 L
conical flasks containing 400 mL of LB production medium with final pH 7 (autoclaved
at 121
◦
C, 15 min). The culture was incubated at 30
◦
C, 150 rpm for 5 days wherein the
incubation period was decided based on time-dependent antimicrobial activity against S.
aureus MCC 2043. Thereafter each flask was extracted with an equal amount of ethyl acetate
to partition the organic metabolites of culture into the solvent. The organic layer obtained
was pooled and concentrated using a rotary evaporator to obtain the crude extract.
3.3. Isolation of the Bioactive Metabolites
The crude extract (13.9 g) was partitioned with a solution of 10% n-hexane and 90%
methanol to remove fatty acids. The extract containing methanol-soluble compounds
(13.40 g) was collected for further isolation work. The fractionation of the methanol extract
was accomplished using MPLC (Buchi Flash, Flawil, Switzerland). Linear gradient elution
was employed with ethyl acetate (A), hexane (B), and methanol (C) as the mobile phase
at a flow rate of 100 mL/min. The sample was mixed with celite and loaded on top of
a pre-packed silica column (20–45
µ
m, 23
×
110 mm, Silica VersaPak cartridge, Sigma
Aldrich, St. Louis, MO, United States [
2
]). It was connected to a Buchi Pump Manager
C-615 coupled to binary pumps (Buchi Modules C-601, Flawil, Switzerland); 100% B was
run for 5 min, followed by 100% B to 100% A for 60 min, and finished with 100% B for
the last 15 min. Still, the extract color was visible in celite; therefore, an extra wash of
15 min was given with a mobile phase of 90% A and 10% C. The total run time was 95 min.
Fractions were collected manually for every 30 s, i.e., 50 mL in collecting flasks. Fractions
were run on silica TLC plates (TLC Silica gel 60 RP-18 F254s, Merck, Darmstadt, Germany),
visualized in UV light (
λ
= 254 nm), and the ones with similar TLC profiles were pooled
together, yielding a total of 16 fractions. The fractions were analyzed using LC-MS for
dereplication study and tested for biofilm inhibition. Depending on the weights obtained
and the bioactivity of the fractions, further purification was performed. Fractions 2 to 5 were
further separated using preparative TLC, and other bioactive fractions were purified with
Biotage high-throughput flash chromatography (Biotage Isolera One, Uppsala, Sweden)
using normal phase pre-packed column (SNAP, Biotage Sweden Ab, Biotage, LLC, MORE,
Uppsala, Sweden). Finally, a total of 36 subfractions were obtained after consecutive
fractionation, which were analyzed for their biofilm inhibition properties. The structure
elucidation was performed for the purified compounds using 1D NMR and 2D NMR
(Bruker
1
H 400 MHz,
13
C 150 MHz) spectroscopy. Samples were prepared by dissolving
5 mg of bacterial extract or fractions in 600
µ
L DMSO-d6 (Sigma-Aldrich, Dorset, UK).
Spectra were processed using MestReNova (Mnova 14.2.0) software (Mestrelab Research,
Santiago de Compostela, Spain).
Data-dependent MS
2
and MS
3
experiments were carried out using a Finnigan LTQ Or-
bitrap coupled to a Surveyor Plus HPLC pump (Thermo Scientific, Bremen, Germany) and
autosampler (Thermo Fisher, Bremen, Germany) in positive and negative ionization modes
using a mass range of m/z100–2000 and 30,000 resolutions. The capillary temperature
was 270
◦
C, the ion spray voltage was 4.5 kV, the capillary voltage was 35 V, the tube lens
voltage was 110 V, and the sheath and auxiliary gas flow rates were 50 and 15, respectively
(units not specified by the manufacturer). Multi-fragmentation (MS
n
) experiments were
accomplished on an Orbitrap analyzer. CID (collision-induced dissociation) was utilized
with a normalized collision energy of 35%, activation Q of 0.250 ms, and activation time
of 30,000 ms applied on ions of most intense, second most intense, and third most intense
peaks for MS
2
and MS
3
, respectively, at an isolation width of 3 microns with 5 microscans.

Microorganisms 2024,12, 413 16 of 20
Resolution was at 15,000 m/
∆
m50%, while the minimum ion signal threshold was set to
500. Fragment mass tolerance for molecular formula detection was set at ±5 ppm.
3.4. In Vitro Biofilm Inhibition Assay
AlamarBlue
®
biofilm viability assay was adopted to check the antibiofilm potential
of fractions of B. velezensis 7NPB-3B obtained after MPLC separation. Stock solutions of
each of the extract/fraction were prepared with 1 mg of the samples dissolved in 100
µ
L of
biological-grade DMSO to get a concentration of 10
µ
g/
µ
L. Assay plates were prepared
by adding 200
µ
L of LB broth pre-inoculated with log phase culture (1
×
10
6
cfu/mL) of
biofilm-forming Staphylococcus aureus MRSA ATCC 43300. A control containing 200
µ
L
of LB without inoculation by the pathogen was maintained. The plates were incubated
at 35
◦
C in a shake incubator at a speed of 140 rpm for 24 h. After 24 h, the planktonic
cells were removed carefully by pipetting and washed with sterile 0.1 M PBS such that the
formed biofilms were not distorted. Thereafter, the plate was seeded with subfractions
such that the final concentration of the sample in each well was 100
µ
g/mL and the volume
was made up to 200
µ
L by adding LB media. DMSO was used as a negative control, while
ciprofloxacin served as a positive control. Finally, the newly seeded plate was placed into a
shaking incubator and agitated for 23 h, at 35
◦
C, 200 rpm. After incubation, the wells were
rinsed with 0.1 M PBS such that the formed biofilms were not distorted. After washing,
90
µ
L of DMEM and 10
µ
L of AlamarBlue
®
were added to the wells. The plates were
further incubated for 1.5 h in a shaking incubator at 35
◦
C covered with foil. Absorbance
reading was taken at 560 nm excitation wavelength and 590 nm emission wavelength
after incubation. The readings were processed in an EXCEL sheet, and final graphs were
produced after processing the results with GraphPad Prism 10.1.0.
Similarly, compounds 1and 5were further assayed to determine their MBEC val-
ues against biofilm-forming MRSA (ATCC 43300). A dilution plate was prepared for
each compound with a concentration range of 200 to 1.56
µ
g/mL by double dilution
method. Ciprofloxacin was used as a positive control, with a concentration range of 100 to
0.78 µg/mL. The MBEC values were determined using graph pad prism 10.1.0.
3.5. LC-MS-Dependent Metabolomics and Dereplication
3.5.1. LC-HRMS
For LC-MS analysis, a 1 mg/mL sample was prepared in a 1:1 MeOH–ACN solvent
system. Experiments were carried out using an Exactive mass spectrometer with an
electrospray ionization source attached to an Accela 600 HPLC pump with an Accela
autosampler and UV/Vis detector (Thermo Scientific, Bremen, Germany). The mass
accuracy was set to less than 3.0 ppm. The Orbitrap mass analyzer can limit the mass
error to
±
3.0 ppm. Mass spectrometry was carried out over a mass range of 100–2000 m/z
in positive and negative ionization modes with a spray voltage of 4.5 kV and capillary
temperature of 270
◦
C. About 10
µ
L was injected from each vial, at a flow rate of 300
µ
L/min.
The column used was an ACE5 C18 column (5
µ
m
×
75 mm
×
3 mm) (Hichrom Limited,
Reading, UK). A binary gradient method was utilized using the two solvents, A (water and
0.1% formic acid) and B (MeCN and 0.1% formic acid). The gradient was carried out for
45 min and the program followed; at zero minutes A = 90% and B = 10%, at 30 min
A = 0%
and B = 100%, at 36 min A = 90% and B = 10% until end at 45 min. The UV absorption
wavelength was set at 254 nm, the sample tray temperature was maintained at 4
◦
C, and
the column at 20
◦
C. The samples were run sequentially, with solvent and media blanks
analyzed first. LC-MS data were acquired using Xcalibur version 2.2.
3.5.2. Metabolic Profiling Studies
LC–MS data were used for metabolomic profiling studies following the protocol of
Macintyre et al. (2014) [
46
]. Raw data were first converted to mzML file format using the
software ProteoWizard 3 [
47
]. Data processing was carried out using the software MZMine
3 [
48
,
49
]. The data were processed while maintaining the specific parameters as mentioned

Microorganisms 2024,12, 413 17 of 20
(mass detector: centroid; noise level: 10,000; minimum time span: 0.2 min; minimum
height: 1
×
106; m/ztolerance: 0.001 m/zor 5 ppm), deconvolution (chromatographic
threshold: 5%), deisotoping (m/ztolerance: 0.001 m/zor 5 ppm, retention time (RT)
tolerance: 0.1 absolute (min), maximum charge: 2, representative isotope: most intense),
filtering (m/zrange: 100–1999, RT: 0–45 min), alignment (m/ztolerance: 0.001 m/zor 5 ppm,
m/zweight: 20, RT tolerance: 5 relative %), and gap filling (m/ztolerance: 0.001 m/z
or 5 ppm, intensity tolerance: 1%). Complete software settings were as described by
Macintyre et al. [46].
The positive- and negative-ionization data were processed together
in an Excel macro that minimized the risk of missing poorly ionized compounds. The
Excel macro was used to dereplicate the samples, matching each m/zfound in the samples
with those in the Dictionary of Natural Products (DNP, version 2017) database to provide
complete information of all the putatively identified metabolites, as well as those that were
unidentified. For the dereplication, a feature ID number, ionization mode, m/z, retention
time, possible molecular formula, peak intensity, related compounds, and source of the
specific metabolites (if available) were generated.
3.5.3. Multivariate Analysis
Multivariate analyses (MVAs) that included both principal component analysis (PCA)
and orthogonal projection to latent structures discriminant analysis (OPLS-DA) for the
prediction of antibacterial metabolites were carried out using the software SIMCA ver.
17 (Umetrics, Umeå, Sweden). PCA was used to observe an overview of the variance
of secondary metabolites between different fractions generated from LC–MS data and
to detect outliers that could influence the model. OPLS-DA was also used to identify
chemically distinct samples that may yield novel and bioactive secondary metabolites.
A pareto scaling was applied to all MVAs to reduce the influence of intense peaks while
emphasizing weaker peaks that may have more biological relevance [
50
]. The Q2 and
R2 values were reported as a qualitative measure of consistency between the predicted
and original data. These values explained the goodness of the prediction of the statistical
models, representing the total explained variance and the predictive power of the models.
4. Conclusions
The study presents the isolation of antibiofilm compounds from endophytic B. velezen-
sis 7NPB-3B wherein the targeted isolation of compounds was aided by LC-MS-based
metabolomics, dereplication using DNP database, and bioactivity-guided fractionation.
To the best of our knowledge, this is the first study on the application of metabolomics
and dereplication for the isolation of bioactive compounds from endophytes of S. brachiata.
The targeted bioactive fractions were also backed by multivariate analysis, such as PLS-
DA and OPLS-DA. Compounds of class diketopiperazine were isolated from the source
that exhibited prominent inhibition of biofilms formed by an MRSA pathogen. A novel
diketopiperazine derivative (1) along with four known compounds was isolated from the
endophyte. These are the first reports of these compounds from the endophytic B. velezensis.
The work justifies the investigation of new drug lead molecules from known sources as
well as untapped niches with the help of the modern omics approach.
Supplementary Materials: The following supporting information can be downloaded at https://www.
mdpi.com/article/10.3390/microorganisms12020413/s1, Figure S1: Structures of the discriminating
dereplicated metabolites detected from the OPLS-DA plot of subfractions listed in the Table 3;
Figure S2A
: 2D correlations (COSY and HMBC) of compounds 2–5; Figure S2B: ECD spectra of
compounds 2to 5in methanol; Figure S3:
1
H NMR spectrum of 1; Figure S4: J Mod
13
C NMR
spectrum of 1; Figure S5: HMBC spectrum of 1; Figure S6: COSY spectrum of compound 1in
methanol; Figure S7: HSQC spectrum of 1; Figure S8: LC-HRMS and MS/MS fragmentation of
1; Figure S9: ECD spectrum of compound 1; Figure S10: Post-biofilm MBEC determination of
ciprofloxacin against MRSA ATCC 43300; Table S1:
1
H- and
13
C- NMR data of isolated compounds
2–5(DMSO-d6, 400 Hz).

Microorganisms 2024,12, 413 18 of 20
Author Contributions: Conceptualization, S.S., P.B.S. and R.E.-E.; methodology, S.S., E.N., L.Y.
(antibiofilm) and R.E.-E. (MVA and dereplication); software and validation, S.S., R.E.-E. and P.K.;
investigation, S.S., R.E.-E. and P.K.; data curation, S.S. and R.E.-E.; writing—original draft prepara-
tion, S.S.; writing—review and editing, R.E.-E. and P.B.S.; supervision, R.E.-E. and P.B.S.; funding
acquisition, S.S., R.E.-E. and P.B.S. All authors have read and agreed to the published version of
the manuscript.
Funding: This research was funded by the Newton-Bhabha PhD placement program 2020–2021
(Grant ID: 648104788) funded by the British Council UK and Department of Biotechnology, India.
This work was supported by the Scientific and Engineering Research Board (SERB), Department of
Science and Technology (Grant ID: ECRA/2016/000788 and EEQ/2016/000268).
Institutional Review Board Statement: Not applicable.
Data Availability Statement: No new data were created or analyzed in this study. Data sharing is
not applicable to this article.
Acknowledgments: Financial support was provided by the Newton-Bhabha Placement Program
jointly supported by the Department of Biotechnology and the British Council under the Newton-
Bhabha fund. S.S. acknowledges the CSIR-JRF fellowship Council of Industrial and Scientific Research
(CSIR). P.K. acknowledges the Department of Biotechnology, India for the DBT-JRF fellowship.
Department of Pure and Applied Chemistry, University of Strathclyde, and the centralized instrument
facility of CSIR-CSMCRI are acknowledged for providing instrumentation facilities.
Conflicts of Interest: The authors declare no conflicts of interest.
References
1. Davies, D. Understanding Biofilm Resistance to Antibacterial Agents. Nat. Rev. Drug Discov. 2003,2, 114–122. [CrossRef]
2.
Roy, R.; Tiwari, M.; Donelli, G.; Tiwari, V. Strategies for Combating Bacterial Biofilms: A Focus on Anti-Biofilm Agents and Their
Mechanisms of Action. Virulence 2018,9, 522–554. [CrossRef]
3. Gotz, F. Staphylococcus and Biofilms. Mol. Microbiol. 2002,43, 1367–1378. [CrossRef]
4.
Zhang, H.; Chen, J.; Liu, Y.; Xu, Q.; Inam, M.; He, C.; Jiang, X.; Jia, Y.; Ma, H.; Kong, L. Discovery of a Novel Antibacterial Protein
CB6-C to Target Methicillin-Resistant Staphylococcus aureus.Microb. Cell Factories 2022,21, 4. [CrossRef] [PubMed]
5.
Singh, S.; Ghadge, V.A.; Kumar, P.; Mathew, D.E.; Dhimmar, A.; Sahastrabudhe, H.; Nalli, Y.; Rathod, M.R.; Shinde, P.B.
Biodiversity and Antimicrobial Potential of Bacterial Endophytes from Halophyte Salicornia brachiata.Antonie van Leeuwenhoek
2021,114, 591–608. [CrossRef] [PubMed]
6.
Iqbal, S.; Begum, F.; Rabaan, A.A.; Aljeldah, M.; Al Shammari, B.R.; Alawfi, A.; Alshengeti, A.; Sulaiman, T.; Khan, A. Classification
and Multifaceted Potential of Secondary Metabolites Produced by Bacillus subtilis Group: A Comprehensive Review. Molecules
2023,28, 927. [CrossRef] [PubMed]
7.
Ye, M.; Tang, X.; Yang, R.; Zhang, H.; Li, F.; Tao, F.; Li, F.; Wang, Z. Characteristics and Application of a Novel Species of Bacillus:
Bacillus Velezensis.ACS Chem. Biol. 2018,13, 500–505. [CrossRef] [PubMed]
8.
Dunlap, C.A.; Kim, S.J.; Kwon, S.W.; Rooney, A.P. Bacillus velezensis is not a Later Heterotypic Synonym of Bacillus amyloliquefaciens;
Bacillus methylotrophicus,Bacillus amyloliquefaciens subsp. plantarum and ‘Bacillus oryzicola’ Are Later Heterotypic Synonyms of
Bacillus velezensis Based on Phylogenomics. Int. J. Syst. Evol. Microbiol. 2016,66, 1212–1217. [CrossRef] [PubMed]
9.
Wang, S.-Y.; Herrera-Balandrano, D.D.; Wang, Y.-X.; Shi, X.-C.; Chen, X.; Jin, Y.; Liu, F.-Q.; Laborda, P. Biocontrol Ability of the
Bacillus amyloliquefaciens Group, B. amyloliquefaciens,B. velezensis,B. nakamurai and B. siamensis, for the Management of Fungal
Postharvest Diseases: A Review. J. Agric. Food Chem. 2022,70, 6591–6616. [CrossRef] [PubMed]
10.
Ngalimat, M.S.; Yahaya, R.S.R.; Baharudin, M.M.A.; Yaminudin, S.M.; Karim, M.; Ahmad, S.A.; Sabri, S. A Review on the
Biotechnological Applications of the Operational Group Bacillus amyloliquefaciens.Microorganisms 2021,9, 614. [CrossRef]
11.
Xu, B.-H.; Lu, Y.-Q.; Ye, Z.-W.; Zheng, Q.-W.; Wei, T.; Lin, J.-F.; Guo, L.-Q. Genomics-Guided Discovery and Structure Identification
of Cyclic Lipopeptides from the Bacillus siamensis JFL15. PLoS ONE 2018,13, e0202893. [CrossRef]
12.
Abdelhamid, A.G.; Hussein, W.E.; Gerst, M.M.; Yousef, A.E. Draft Genome Sequence of Bacillus velezensis OSY-GA1, Which
Encodes Multiple Antimicrobial Metabolites and Expresses Antimicrobial Activity against Foodborne Pathogens. Microbiol.
Resour. Announc. 2019,8, e01725-18. [CrossRef]
13.
Grady, E.N.; MacDonald, J.; Ho, M.T.; Weselowski, B.; McDowell, T.; Solomon, O.; Renaud, J.; Yuan, Z.-C. Characterization and
Complete Genome Analysis of the Surfactin-Producing, Plant-Protecting Bacterium Bacillus velezensis 9D-6. BMC Microbiol. 2019,
19, 5. [CrossRef]
14.
Suchi, S.A.; Bin Nam, K.; Kim, Y.K.; Tarek, H.; Yoo, J.C. A Novel Antimicrobial Peptide YS12 Isolated from Bacillus velezensis
CBSYS12 Exerts Anti-Biofilm Properties against Drug-Resistant Bacteria. Bioprocess Biosyst. Eng. 2023,46, 813–828. [CrossRef]
[PubMed]

Microorganisms 2024,12, 413 19 of 20
15.
Ali, S.A.M.; Sayyed, R.Z.; Mir, M.I.; Khan, M.Y.; Hameeda, B.; Alkhanani, M.F.; Haque, S.; Mohammad Al Tawaha, A.R.; Poczai, P.
Induction of Systemic Resistance in Maize and Antibiofilm Activity of Surfactin from Bacillus velezensis MS20. Front. Microbiol.
2022,13, 879739. [CrossRef] [PubMed]
16.
Tkachuk, N.; Zelena, L. Inhibition of Heterotrophic Bacterial Biofilm in the Soil Ferrosphere by Streptomyces spp. and Bacillus
velezensis.Biofouling 2022,38, 916–925. [CrossRef] [PubMed]
17.
Yoo, Y.; Seo, D.-H.; Lee, H.; Cho, E.-S.; Song, N.-E.; Nam, T.G.; Nam, Y.-D.; Seo, M.-J. Inhibitory Effect of Bacillus velezensis on
Biofilm Formation by Streptococcus mutans.J. Biotechnol. 2019,298, 57–63. [CrossRef]
18.
Mazlan, N.W.; Clements, C.; Edrada-Ebel, R. Targeted Isolation of Anti-Trypanosomal Naphthofuran-Quinone Compounds from
the Mangrove Plant Avicennia lanata.Mar. Drugs 2020,18, 661. [CrossRef]
19.
Yan, X.; Zhou, Y.-X.; Tang, X.-X.; Liu, X.-X.; Yi, Z.-W.; Fang, M.-J.; Wu, Z.; Jiang, F.-Q.; Qiu, Y.-K. Macrolactins from Marine-Derived
Bacillus subtilis B5 Bacteria as Inhibitors of Inducible Nitric Oxide and Cytokines Expression. Mar. Drugs 2016,14, 195. [CrossRef]
20. Kind, T.; Fiehn, O. Seven Golden Rules for Heuristic Filtering of Molecular Formulas Obtained by Accurate Mass Spectrometry.
BMC Bioinform. 2007,8, 105. [CrossRef]
21.
Kumar, N.; Mohandas, C.; Nambisan, B.; Kumar, D.R.S.; Lankalapalli, R.S. Isolation of Proline-Based Cyclic Dipeptides from
Bacillus sp. N Strain Associated with Rhabitid Entomopathogenic Nematode and Its Antimicrobial Properties. World J. Microbiol.
Biotechnol. 2013,29, 355–364. [CrossRef]
22.
Park, A.R.; Jeong, S.-I.; Jeon, H.W.; Kim, J.; Kim, N.; Ha, M.T.; Mannaa, M.; Kim, J.; Lee, C.W.; Min, B.S.; et al. A Diketopiper-
azine, Cyclo-(L-Pro-L-Ile), Derived from Bacillus thuringiensis JCK-1233 Controls Pine Wilt Disease by Elicitation of Moderate
Hypersensitive Reaction. Front. Plant Sci. 2020,11, 1023. [CrossRef]
23.
Bhattacharya, D.; Lai, T.K.; Saha, A.; Selvin, J.; Mukherjee, J. Structural Elucidation and Antimicrobial Activity of a Diketopiper-
azine Isolated from a Bacillus sp. Associated with the Marine Sponge Spongia Officinalis.Nat. Prod. Res. 2021,35, 2315–2323.
[CrossRef]
24.
Harris, G.H.; Dufresne, C.; Joshua, H.; Koch, L.A.; Zink, D.L.; Salmon, P.M.; Göklen, K.E.; Kurtz, M.M.; Rew, D.J.; Bergstrom, J.D.;
et al. Isolation, Structure Determination and Squalene Synthase Activity of L-731,120 and L-731,128, Alkyl Citrate Analogs of
Zaragozic Acids A and B. Bioorg. Med. Chem. Lett. 1995,5, 2403–2408. [CrossRef]
25.
Li, D.; Carr, G.; Zhang, Y.; Williams, D.E.; Amlani, A.; Bottriell, H.; Mui, A.L.-F.; Andersen, R.J. Turnagainolides A and B, Cyclic
Depsipeptides Produced in Culture by a Bacillus sp.: Isolation, Structure Elucidation, and Synthesis. J. Nat. Prod. 2011,74,
1093–1099. [CrossRef] [PubMed]
26.
Fdhila, F.; Vázquez, V.; Sánchez, J.L.; Riguera, R. Dd-Diketopiperazines: Antibiotics Active against Vibrio anguillarum Isolated
from Marine Bacteria Associated with Cultures of Pecten maximus.J. Nat. Prod. 2003,66, 1299–1301. [CrossRef]
27.
Harizani, M.; Katsini, E.; Georgantea, P.; Roussis, V.; Ioannou, E. New Chlorinated 2,5-Diketopiperazines from Marine-Derived
Bacteria Isolated from Sediments of the Eastern Mediterranean Sea. Molecules 2020,25, 1509. [CrossRef] [PubMed]
28. Shim, S.H. Diketopiperazines from Cultures of Rhodococcus rhodochrous.Chem. Nat. Compd. 2016,52, 1157–1159. [CrossRef]
29.
Chen, X.; Zou, J.-M.; Li, L.; Bao, K.; Zhao, Y.; Miao, C.-P.; Zhao, L.-X.; Chen, Y.-W.; Li, Y.-Q. Cyclic Peptide Secondary Metabolites
with Antifungal Activity against Root-Rot Pathogens of Panax notoginseng Produced by Streptomyces yatensis.Chem. Nat. Compd.
2021,57, 1181–1183. [CrossRef]
30.
Xiang, W.-X.; Liu, Q.; Li, X.-M.; Lu, C.-H.; Shen, Y.-M. Four Pairs of Proline-Containing Cyclic Dipeptides from Nocardiopsis sp.
HT88, an Endophytic Bacterium of Mallotus nudiflorus L. Nat. Prod. Res. 2020,34, 2219–2224. [CrossRef]
31.
Domzalski, A.; Margent, L.; Vigo, V.; Dewan, F.; Pilarsetty, N.V.K.; Xu, Y.; Kawamura, A. Unambiguous Stereochemical Assignment
of Cyclo(Phe-Pro), Cyclo(Leu-Pro), and Cyclo(Val-Pro) by Electronic Circular Dichroic Spectroscopy. Molecules 2021,26, 5981.
[CrossRef]
32.
Ding, R.; Ying, J.; Zhao, Y. An Electronic Circular Dichroism Spectroscopy Method for the Quantification of L- and D-Amino
Acids in Enantiomeric Mixtures. R. Soc. Open Sci. 2021,8, 201963. [CrossRef] [PubMed]
33.
Amdursky, N.; Stevens, M.M. Circular Dichroism of Amino Acids: Following the Structural Formation of Phenylalanine.
ChemPhysChem 2015,16, 2768–2774. [CrossRef] [PubMed]
34.
Shaala, L.A.; Youssef, D.T.A.; Badr, J.M.; Harakeh, S.M.; Genta-Jouve, G. Bioactive Diketopiperazines and Nucleoside Derivatives
from a Sponge-Derived Streptomyces Species. Mar. Drugs 2019,17, 584. [CrossRef] [PubMed]
35.
Nair, M.K.M.; Joy, J.; Vasudevan, P.; Hinckley, L.; Hoagland, T.A.; Venkitanarayanan, K.S. Antibacterial Effect of Caprylic Acid
and Monocaprylin on Major Bacterial Mastitis Pathogens. J. Dairy Sci. 2005,88, 3488–3495. [CrossRef]
36.
Impellizzeri, G.; Pappalardo, G.; Rizzarelli, E.; Tringali, C. Synthesis and High Field NMR Study of a New Cyclodipeptide–
β
-
Cyclodextrin Derivative. J. Chem. Soc. Perkin Trans. 1996,2, 1435–1440. [CrossRef]
37.
Harnden, M.R. Synthesis and Stereochemistry of 1,4-Diazabicyclo[4,3,0]Nonane-2,5,9-Triones. J. Chem. Soc. C Org. 1967, 2341–2351.
[CrossRef]
38.
de Carvalho, M.P.; Abraham, W.-R. Antimicrobial and Biofilm Inhibiting Diketopiperazines. Curr. Med. Chem. 2012,19, 3564–3577.
[CrossRef]
39.
Park, D.-K.; Lee, K.-E.; Baek, C.-H.; Kim, I.H.; Kwon, J.-H.; Lee, W.K.; Lee, K.-H.; Kim, B.-S.; Choi, S.-H.; Kim, K.-S. Cyclo(Phe-Pro)
Modulates the Expression of ompU in Vibrio Spp. J. Bacteriol. 2006,188, 2214–2221. [CrossRef]
40.
Ortiz-Castro, R.; Díaz-Pérez, C.; Martínez-Trujillo, M.; del Río, R.E.; Campos-García, J.; López-Bucio, J. Transkingdom Signaling
Based on Bacterial Cyclodipeptides with Auxin Activity in Plants. Proc. Natl. Acad. Sci. USA 2011,108, 7253–7258. [CrossRef]

Microorganisms 2024,12, 413 20 of 20
41.
Holden, M.T.G.; Ram Chhabra, S.; De Nys, R.; Stead, P.; Bainton, N.J.; Hill, P.J.; Manefield, M.; Kumar, N.; Labatte, M.; England,
D.; et al. Quorum-Sensing Cross Talk: Isolation and Chemical Characterization of Cyclic Dipeptides from Pseudomonas aeruginosa
and Other Gram-Negative Bacteria. Mol. Microbiol. 2002,33, 1254–1266. [CrossRef]
42.
Li, J.; Wang, W.; Xu, S.X.; Magarvey, N.A.; McCormick, J.K. Lactobacillus Reuteri -Produced Cyclic Dipeptides Quench Agr
-Mediated Expression of Toxic Shock Syndrome Toxin-1 in Staphylococci. Proc. Natl. Acad. Sci. USA 2011,108, 3360–3365.
[CrossRef]
43.
He, W.; Li, Y.; Luo, W.; Zhou, J.; Zhao, S.; Xu, J. Herbicidal Secondary Metabolites from Bacillus velezensis JTB8-2 against Orobanche
aegyptiaca.AMB Express 2022,12, 52. [CrossRef]
44.
Khan, M.S.; Gao, J.; Chen, X.; Zhang, M.; Yang, F.; Du, Y.; Moe, T.S.; Munir, I.; Xue, J.; Zhang, X. The Endophytic Bacteria Bacillus
velezensis Lle-9, Isolated from Lilium leucanthum, Harbors Antifungal Activity and Plant Growth-Promoting Effects. J. Microbiol.
Biotechnol. 2020,30, 668–680. [CrossRef]
45.
Syed-Ab-Rahman, S.F.; Chua, E.T.; Schenk, P.M. Characterisation and Isolation of Bioactive Compounds of Anti-Oomycete
Bacterial Isolates Inhibiting the Growth of Phytophthora capsici.Australas. Plant Pathol. 2021,50, 651–659. [CrossRef]
46.
Macintyre, L.; Zhang, T.; Viegelmann, C.; Martinez, I.; Cheng, C.; Dowdells, C.; Abdelmohsen, U.; Gernert, C.; Hentschel, U.;
Edrada-Ebel, R. Metabolomic Tools for Secondary Metabolite Discovery from Marine Microbial Symbionts. Mar. Drugs 2014,12,
3416–3448. [CrossRef]
47.
Kessner, D.; Chambers, M.; Burke, R.; Agus, D.; Mallick, P. ProteoWizard: Open Source Software for Rapid Proteomics Tools
Development. Bioinformatics 2008,24, 2534–2536. [CrossRef] [PubMed]
48.
Pluskal, T.; Castillo, S.; Villar-Briones, A.; Orešiˇc, M. MZmine 2: Modular Framework for Processing, Visualizing, and Analyzing
Mass Spectrometry-Based Molecular Profile Data. BMC Bioinform. 2010,11, 395. [CrossRef] [PubMed]
49.
Schmid, R.; Heuckeroth, S.; Korf, A.; Smirnov, A.; Myers, O.; Dyrlund, T.S.; Bushuiev, R.; Murray, K.J.; Hoffmann, N.; Lu, M.;
et al. Integrative Analysis of Multimodal Mass Spectrometry Data in MZmine 3. Nat. Biotechnol. 2023,41, 447–449. [CrossRef]
[PubMed]
50. Worley, B.; Powers, R. Multivariate Analysis in Metabolomics. Curr. Metabolomics 2012,1, 92–107. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.