Received: 10 July 2020 Revised: 1 December 2020 Accepted: 13 January 2021

DOI: 10.1002/pca.3031

RESEARCH ARTICLE

In vivo anti-inflammatory activity of Fabaceae species extracts

screened by a new ex vivo assay using human whole blood

Welton Rosa1 | Olívia da Silva Domingos1 | Paula Pio de Oliveira Salem1 |

Ivo Santana Caldas2 | Michael Murgu3 | Jo~
ao Henrique Ghilardi Lago 4
|
Patricia Sartorelli5 | Danielle Ferreira Dias1 |
1
Daniela Aparecida Chagas-Paula | Marisi Gomes Soares1

1
Institute of Chemistry – Federal University of
Alfenas – Unifal-MG, Alfenas, MG, Brazil Abstract
2
Department of Pathology and Parasitology, Introduction: Plants have been considered a promising source for discovering new
Federal University of Alfenas – Unifal-MG,
compounds with pharmacological activities. The Fabaceae family comprises a large
Alfenas, MG, Brazil
3
Waters Corporation, Alphaville, S~
ao Paulo, variety of species that produce substances with diverse therapeutic potential, includ-
SP, Brazil ing anti-inflammatory activity. The limitations of current anti-inflammatories generate
4
Centre for Human and Natural Sciences –
the need to research new anti-inflammatory structures with higher efficacy as well as
Federal University of ABC – UFABC, S~
ao
Paulo, SP, Brazil develop methods for screening multiple samples, reliably and ethically, to assess such
5
Institute of Environmental, Chemical and therapeutic properties.
Pharmaceutical Sciences, Federal University of
S~ao Paulo – UNIFESP, S~ao Paulo, SP, Brazil
Objective: Validate and apply a quantification method for prostaglandin E2 (PGE2)
production from an ex vivo assay in human blood in order to screen anti-inflammatory
Correspondence
Marisi Gomes Soares, Institute of Chemistry –
activity present in many Fabaceae species extracts.
Federal University of Alfenas – Unifal-MG, Methods: Human blood was incubated with extracts from 47 Fabaceae species. After
Alfenas, MG, Brazil.
Email: marisigs@gmail.com
lipopolysaccharide (LPS)-induced inflammation, PGE2 was quantified in the plasma by
liquid chromatography with tandem mass spectrometry (LC–MS/MS). The extracts
Funding information
Coordenaç~ao de Aperfeiçoamento de Pessoal
that presented PGE2 production inhibition were further assessed through in vivo
de Nível Superior – Brazil (CAPES) – Finance assay and then chemically characterised through an analysis of ultra-performance liq-
Code 001; Fundaç~ao de Amparo à Pesquisa do
Estado de Minas Gerais – Brazil (FAPEMIG)
uid chromatography electrospray ionisation quadrupole time-of-flight tandem mass
APQ-02353-17; Conselho Nacional de spectrometry (UPLC-ESI-QTOF-MS2) data.
Desenvolvimento Científico e Tecnológico –
Brazil (CNPq) 427497/2018-3; Financiadora
Results: The new ex vivo anti-inflammatory assay showed that five out of the
de Estudos e Projetos - Brazil (FINEP) 47 Fabaceae species inhibited PGE2 production. Results from an in vivo assay and the
metabolic profile of the active extracts supported the anti-inflammatory potential of
four species.
Conclusion: The quantification method for PGE2 demonstrated fast, sensitive, pre-
cise, and accurate results. The new ex vivo anti-inflammatory assay comprised a great,
reliable, and ethical approach for the screening of a large number of samples before
an in vivo bioassay. Additionally, the four active extracts in both ex vivo and in vivo
assays may be useful for the development of more efficient anti-inflammatory drugs.

KEYWORDS
dereplication, ear edema, LC–MS/MS, LPS, PGE2, plants, plasma samples, SPE, UPLC-ESI-
QTOF-MS2

Phytochemical Analysis. 2021;1–25. wileyonlinelibrary.com/journal/pca © 2021 John Wiley & Sons, Ltd. 1
2 ROSA ET AL.

1 | I N T RO DU CT I O N activities of Fabaceae species have been reported in the literature as

fungicidal, leishmanicidal, antimicrobial, and potential activity against
1
Anti-inflammatories are used as medicine around the world. Specific gastrointestinal, respiratory, and hepatitis diseases.12,13 Anti-
substances involved in inflammatory processes are naturally associ- inflammatory effects are also related to many species including
ated with important physiological functions under homeostasis. These Glycine max L., Phaseolus vulgaris L., Pisum sativum L., Codariocalyx
substances are mainly derived from arachidonic acid (AA) metabolism motorius, Tetrapleura tetraptera, and Piptadenia stipulacea.14–17
2–5
and are known as eicosanoids. Upon an inflammatory stimuli, AA is Besides the pharmacological aspects, many Fabaceae species are con-
released in high levels from the cell membrane by the action of the sumed as food and are well-known for their nutritional importance.18
phospholipase A2 (PLA2) enzyme. Then, the AA is metabolised in Thus, fast and efficient methods are necessary in order to screen a
inflammatory mediators by different enzymatic pathways such as large number of samples and discover more efficient anti-
cyclooxygenases (COX), lipoxygenases (LOX), and cytochrome P450 inflammatory compounds, especially among plant species in the
(CYP). Mediators produced by the COX-pathway such as prostaglan- Fabaceae family. Furthermore, high-performance liquid chromatogra-
din, prostacyclin, and thromboxane are responsible for the classic phy (HPLC) associated to electrospray ionisation high-resolution mass
symptoms associated with inflammations, i.e. pain, fever, redness, and spectrometry (ESI-HRMS) data may be a powerful tool to analyse bio-
swelling, whereas those produced by the LOX-pathway, such as the active samples and guide further phytochemical studies.
leukotrienes and lipoxins, lead mainly to the recruitment of inflamma-
tory cells.6 Available anti-inflammatory compounds can be classified
as a steroidal drug (SAID) or a non-steroidal drug (NSAID). SAIDs act 2 | EX PE RI MENT AL
mainly as inhibitors of the PLA2 enzyme, preventing AA liberation
and, consequently, the formation of metabolites from the COX and 2.1 | Solvents and reagents
LOX-pathways. In this sense, SAIDs are the most effective drugs in
anti-inflammatory therapy and also the most used drugs, although All of the solvents used for sample preparation and analyses by
7
they present many significant side effects. However, NSAIDs can act LC–MS/MS and ultra-performance liquid chromatography
selectively in COX or LOX as an enzymatic cascade, inhibiting the pro- electrospray ionisation quadrupole time-of-flight tandem mass spec-
duction of the AA-derived mediators that are responsible for the main trometry (UPLC-ESI-QTOF-MS2) were HPLC grade solvents, including
inflammation signs, as previously mentioned. However, NSAIDs still acetonitrile, hexane (Sigma-Aldrich, St Louis, MO, USA), methanol,
present limited efficacy in some inflammatory disorders, especially ethanol (Tedia, Fairfield, OH, USA), and glacial acetic acid (J.T. Baker,
chronic ones, and important side effects.6,8 In regards to the demand Radnor, PA, USA). Standards for PGE2, chloramphenicol (CAP), dexa-
for developing new anti-inflammatory drugs that are more efficient methasone, croton oil, LPS from Escherichia coli O26:B6, formic acid,
and have reduced side effects, it is important to search for new active and indomethacin were purchased from Sigma-Aldrich. Purification of
compounds from different sources. In this sense, many assays screen- plasma samples was performed through solid-phase extraction (SPE)
ing several types of compounds and samples are available. Some of using a Supelco Manifold System and Supelclean™ LC-18 SPE car-
them comprise enzymatic kits for in vitro assessment, which are usu- tridges (with 500 mg sorbent and 2.8 mL). Simple pure Microfilters of
ally expensive and may have a time-consuming import process. There polytetrafluoroethylene (PTFE, 45 μm size pore) were used to filter
are also in vivo assays that, although well established, require the use the plant extracts that were resuspended after the liquid–liquid parti-
of a large number of animals, which is not practical when testing many tion cleanup with hexane HPLC grade solvents (Sigma-Aldrich).
samples that do not have a preliminary assessment of pharmacological
potential.9 Based on these limitations, this study suggests the use of
human whole blood stimulated by lipopolysaccharide (LPS) for ex vivo 2.2 | Plant material
assessment of the anti-inflammatory potential of a large number of
samples, by measuring the inhibitory effect in prostaglandin E2 (PGE2) Leaves from 47 different species of Fabaceae were collected in
production. Since PGE2 is one of the main inflammatory biomarkers, Alfenas, Minas Gerais State, in the Atlantic Forest biome, Brazil. The
the proposed anti-inflammatory assay comprises a simple test that species collected for this study were: Acacia polyphylla DC. (1), Albizia
effectively mimics an in vivo environment since it occurs in whole lebbeck (L.) Benth. (2), Albizia polycephala (Benth.) Killip ex Record
blood where many cells and other mediators are present. Thus, the (3), Apuleia leiocarpa (Vogel) J.F. Macbr. (4), Arachis hypogaea L. (5),
anti-inflammatory potential of different samples of plant extracts from Bauhinia galpinii N.E.Br. (6), Bauhinia longifolia (Bong.) Steud. (7),
Brazilian species in the Fabaceae family was assessed. The PGE2 pro- Bauhinia purpurea L. (8), Bauhinia rufa (Bong.) Steud. (9), Bauhinia
duced was analysed and quantified by liquid chromatography with variegata L. (10), Caesalpinia echinata Lam. (11), Libidibia ferrea (Mart.
tandem mass spectrometry (LC–MS/MS), through a robust and sensi- ex Tul.) L.P. Queiroz (12), Poincianella pluviosa var. peltophoroides
tive detection method. (Benth.) L.P. Queiroz (13), Caesalpinia pulcherrima (L.) Sw. (14),
10
Fabaceae is the third-largest family of plants in the world, Copaifera langsdorffii Desf. (15), Dalbergia villosa (Benth.) Benth. (16),
including a wide range of secondary metabolites, such as alkaloids, fla- Delonix regia (Bojer ex Hook.) Raf. (17), Enterolobium contortisiliquum
vonoids, terpenoids, among others.11 Different pharmacological (Vell.) Morong (18), Glycine max (L.) Merr. (19), Holocalyx balansae
ROSA ET AL. 3

Micheli (20), Hymenaea courbaril L. (21), Inga cylindrica (Vell.) Mart. authorised by the Research Ethics Committee at the Federal Univer-
(22), Inga edulis Mart. (23), Inga semialata (Vell.) Mart. (24), Inga vera sity of Alfenas (Unifal-MG) (number 89325818.1.0000.5142). Blood
Willd. (25), Lablab purpureus (L.) Sweet (26), Leucaena leucocephala was collected from donors who reported to not have used anti-
(Lam.) de Wit (27), Machaerium aculeatum Raddi (28), Machaerium inflammatory drugs or CAP at least 15 days before the date of the
dimorphandrum Hoehne (29), Machaerium isadelphum (E.Mey.) Standl. blood collection. Fresh blood samples were collected in vacuum tubes
(30), Machaerium nyctitans (Vell.) Benth. (31), Machaerium stipitatum containing sodium heparin immediately before incubation in the
(DC.) Vogel (32), Machaerium villosum Vogel (33), Mimosa ex vivo assay.
caesalpiniifolia Benth. (34), Peltophorum dubium (Spreng.) Taub. (35),
Phaseolus vulgaris L. (36), Piptadenia gonoacantha (Mart.) J.F. Macbr.
(37), Platycyamus regnellii Benth. (38), Platypodium elegans Vogel (39), 2.6 | Ex vivo anti-inflammatory assay in human
Pogonophora schomburgkiana Miers ex Benth. (40), Senna macranthera whole blood
(DC. ex Collad.) H.S. Irwin & Barneby (41), Stryphnodendron
adstringens (Mart.) Coville (42), Sweetia fruticosa Spreng. (43), Tachigali The ex vivo anti-inflammatory assay in whole blood was adapted
rugosa (Mart. ex Benth.) Zarucchi & Pipoly (44), Tamarindus from previously described methodologies.20,21 Fresh and heparinised
indica L. (45), Vicia faba L. (46), and Zollernia ilicifolia (Brongn.) Vogel human whole blood (200 μL) was incubated with the tested material
(47). Voucher specimens (Supporting Information Table S1) of each samples (reference drugs or plant extracts) in 96 well plates. After
studied plant were deposited in the herbarium at the Federal Univer- this, the LPS solution (in PBS) was added at a final concentration of
sity of Alfenas (Unifal-MG). All species were registered on the 10 μg/mL for an inflammatory stimulation. Reference drugs (dexa-
National System for the Management of Genetic Heritage and Associ- methasone and indomethacin) were initially diluted in ethanol–water
ated Traditional Knowledge – SisGen (A642B9B). (1:1, v/v) as stock solutions at 1 mg/mL due to their insolubility in
pure water. They were further diluted in PBS and were used as pos-
itive controls at a final concentration of 1 μg/mL. The plant extracts
2.3 | Crude extracts from plant material were all tested with a 10 μg/mL final concentration. Negative con-
trol was taken as whole blood with LPS and without any treatment.
Leaves from each studied plant species were dried for 72 h at 50
C The final volume in each well was 250 μL and all dilutions of refer-
and then powdered. Crude extract from each plant species was pre- ence drugs, negative control, or plant extracts used PBS. The assay
19
pared based on a previous study. The extraction system composi- was performed with four replicates for each sample tested. After
tion was ethanol–water (3:1, v/v) with 0.1% glacial acetic acid (1:50, 24 h of incubation at 37
C and 5% carbon dioxide (CO2) atmo-
w/v plant material/solvent mixture in g/mL) using approximately 2 g sphere, the plate was centrifuged at 1000 rpm for 5 min and 4
C,
of dried plant material. After 24 h of dark incubation, at 25
C, and and the plasma of each replicate was collected.20,21 All instruments
125 rpm in an orbital Shaker, each extract was filtered, and the sol- and PBS solutions used were sterilised before the assay was
vents were eliminated by evaporation under reduced pressure performed.
followed by lyophilisation.

2.7 | Obtaining plasma samples based on the

2.4 | Sample preparation of plant extracts for an method validation tests
ex vivo anti-inflammatory assay and analysis by UPLC-
ESI-QTOF-MS2 Plasma samples obtained from heparinised fresh whole blood of dif-
ferent volunteers were mixed in order to obtain the homogenisation
Dried crude extracts (1 mg) were resuspended in pure water (1 mg/mL) of plasma matrix to be used in method validation tests.
and submitted to an ultrasound bath and vortex. The cleanup was per-
formed twice for each sample using 300 μL of hexane to eliminate
non-polar compounds. Then, cleaned-up extracts were filtered in a 2.8 | Plasma sample preparation for analyses with
PTFE microfilter and stored at −20
C. They were then injected in a LC–MS/MS
UPLC-ESI-QTOF-MS2 system or diluted with phosphate-buffered
saline (PBS) to the whole blood anti-inflammatory assay. Plasma sample preparation was performed according to a previously
described methodology.22 Protein precipitation was performed adding
500 μL of the mixture methanol–acetonitrile (1:1, v/v) to 100 μL of
2.5 | Observations regarding human ethics plasma. After homogenisation in a vortex, precipitated proteins were
eliminated through centrifugation for 10 min at 6000 rpm and 4
C.
Peripheral venous blood was obtained from both male and female Supernatants obtained were diluted in water to decrease the
healthy volunteers between 20 and 30 years old. The volunteers methanol–acetonitrile amount to 10%. Sample purification was
signed their informed consent for participation in the research conducted in SPE C18 cartridges. The first step using SPE consisted in
4 ROSA ET AL.

conditioning cartridges with 2 mL of methanol and 2 mL of water con- 2.11 | In vivo anti-inflammatory assay for the ear
taining 0.1% of acetic acid. After that, the sample was loaded in the edema inhibition assessment
cartridge and was filled with 2 mL of 0.1% of an acetic acid solution.
The analyte (PGE2) was eluted from the cartridge with 2 mL of metha- This assay was carried out with male Swiss mice, chosen according to
nol containing 0.1% of acetic acid. Purified PGE2 samples obtained a previously described methodology.23,24 The Ethics Committee for
were completely dried for 6 h, without heating and protected the Use of Animals at the Federal University of Alfenas (Unifal-MG)
from the light. Then, each sample was resuspended in 100 μL of approved the protocol (number 16/2016). Initially, cutaneous inflam-
a CAP solution containing 25 ng/mL in acetonitrile and analysed by mation was induced in the left ear of the mice (n = 8) with the topical
LC–MS/MS. Resuspended samples were kept on ice or in a freezer application of croton oil (5% v/v) dissolved in acetone (vehicle),
until the exact moment of the analysis. according to the established protocols by Tubaro et al. and Santos
et al.23,24 Only acetone was applied in the right ear. The topical
application of croton oil or its predominant phorbol ester
2.9 | Validation of the PGE2 quantification method [12-O-tetradecanoylphorbol-13-acetate] (TPA) leads to an acute
in plasma samples inflammatory reaction characterised by vasodilatation, swelling, and
polymorphonuclear leukocyte infiltration to the tissue.25 Although
Before validation, stock solutions of PGE2 and CAP (1 ppm) were pre- croton oil and TPA are effective as phlogistic agents for anti-
pared in acetonitrile. CAP was diluted to a 25 ng/mL solution in aceto- inflammatory assay, croton oil is cheaper than TPA. Our study intends
nitrile to be used as the internal standard (IS). A mix of plasma to develop an approach that is the cheapest and simplest for the
samples from fresh whole blood was pooled with PGE2 standards assessment of the anti-inflammatory activity.
solutions with different concentrations, ranging from 400 to Therefore, 30 min after the application of 20 μL of the irritant
0.195 ng/mL, for constructing calibration curves in triplicates. Quality solution (croton oil), topical treatment took place with the plant
control (QC) samples were prepared at 30, 100, and 200 ng/mL in the extracts or reference drugs (positive controls). Reference drugs (dexa-
plasma and solution. All figures of merit in the quantification method methasone and indomethacin), as well as tested extracts, were all
(such as linearity, accuracy, precision, matrix effects, recovery, and diluted in acetone (vehicle) at 0.5 mg/ear.23,24 Animals from the nega-
stability) were determined. tive control group received only the vehicle acetone as treatment.
Animals were euthanised by inhalation of isoflurane 6 h after the
inflammation was induced. A 6 mm-diameter ear fragment excision on
2.10 | LC–MS/MS instrument parameters for PGE2 both ears was made on each animal. The edema quantification was
quantification in plasma samples determined by the difference in weight between the fragments of the
left ears and their respective right ears.
LC–MS/MS analyses were carried out in a liquid chromatograph
Shimadzu Prominence with two LC-20 AD pumps, a DGU-20A3
degasser, SIL-20A HT autosampler, CTO-20A oven, and CBM-20A 2.12 | UPLC-ESI-QTOF-MS2 analyses of the active
controller coupled to a mass spectrometer Shimadzu LC-8030 (triple extracts on ex vivo assay
quadrupole analyser) performing ESI.
Chromatographic separation was conducted at 30
C on a The five extracts of active species on ex vivo assay were analysed in a
Poroshell 120 EC-C18 (Agilent, Santa Clara, CA, USA) reversed-phase UPLC-ESI-QTOF-MS2 system for their metabolic profile determina-
column (150 mm and 4.6 mm i.d.) with 2.7 μm particles coupled to a tion. A volume of 10 μL of these samples was injected in a Waters
reversed-phase pre-column. Two mobile phases were used: 0.1% of Acquity UPLC using a column ACQUITY UPLC HSS T3 (2.1 mm ×
formic acid (phase A) and acetonitrile (phase B). A gradient 100 mm, 1.8 μm size particle) in an oven at 45
C. A binary mobile
chromatographic method was used with a total flow of 0.3 mL/min: phase system was used: 0.1% formic acid aqueous solution (phase A)
mobile phase B varied from 40% to 100% in 3 min; held for 1 min at and acetonitrile (phase B). The total flow was 0.5 mL/min in a gradient
100%. Then, the method returned to phase B with 40% in 0.50 min method: phase B varied from 2 to 95% for 8 min; phase B was kept at
and was kept for 4 min for column reconditioning, resulting in 8.5 min 95% for 0.6 min; after that, phase B varied from 95 to 2% in 0.1 min
for the analysis of each sample. All injection volumes were 20 μL. and then the 2% isocratic in phase B was kept for 1.3 min, totaling
Processed samples were kept on ice or in a freezer until the analysis 10 min for each sample.
was performed by LC–MS/MS. PGE2 and IS detection in the mass The ULPC system was coupled with a mass spectrometer XEVO-
spectrometer was performed using selected reaction monitoring G2XSQTOF (Waters, Milford, MA, USA) equipped with an ESI source.
(SRM) in the negative mode [M − H]− with nitrogen (N2) as the drying The spectrometer used leucine enkephaline as a calibrator for the
(15 L/min at 450
C) and nebuliser gas (2 L/min). Transitions of TOF analyser. The lockspray infusion flow rate was 50 μL/min, and
mass-to-charge (m/z) ratios monitored by the SRM method were lockspray capillary was 1 kV. Detection time was 10 min in full scan
optimised to improve sensitivity during the detection of these com- mode with a mass range from 100 to 900 Da, for the positive mode,
pounds (Table S2). and from 100 to 1000 Da, for the negative mode. In positive mode,
ROSA ET AL. 5

the capillary voltage was set as 3 kV, source temperature was 130
C, clearly differentiated by the RT and SRM methods. The analysis pro-
drying gas was 450
C with a flow of 900 L/h, and the cone gas flow vided for the sensitive and specific detection of these compounds.
was set to 50 L/h. In negative mode, the capillary voltage was 2 kV SRM methods are well known for enabling higher sensitivity with
and the other parameters were the same as those used in the positive lower detection and quantification limits, which becomes a prominent
mode. Automatised MS2 analyses were performed with argon as a technique in quantification methods.28 SRM chromatograms of the
collision gas with 6 and 15 kV in negative and positive modes, spiked plasma with PGE2 and IS, and blank plasma sample, are shown
respectively. The data obtained were processed with the MassLynx in Figure 1. There were no peaks identified during the RTs for PGE2
4.1 (Waters) software. or IS in blank plasma samples after the injection of the upper limit of
quantification (ULOQ) (Figure 1). Therefore, no carry-over effect was
considered significant in the method proposed. Besides this, the com-
2.13 | Determining the metabolic profile of active parison of the SRM chromatograms showed that PGE2 and IS were
extracts in an ex vivo assay absent or undetectable in samples that did not have an inflammation
induced (blank samples).
Chromatographic and HRMS data of the five active extracts were The matrix effect was measured by comparing the PGE2 and IS
processed in the Masslynx 4.1 software (Waters) for peak detection, response between QC samples prepared with plasma (biological
deconvolution, m/z and retention time (RT) correction, deisotope, and matrix) and solutions with the same concentration levels. Matrix
noise elimination. Peak area related to each compound characterised effect variation was expressed as a coefficient of variation (CV %) of
by RT and m/z was used to compare the metabolic profile of the the normalised matrix factor (NMF) values (Table S3). For each con-
extract samples. centration tested, the CV % value was low and acceptable according
The main peaks detected were identified through dereplication to the required validation limits guidelines.29,30 The plasma matrix did
comparing the m/z with comprehensive databases according to the not interfere in the detection of both substances (PGE2 and IS) since
Dictionary of Natural Products© (DNP), METLIN,26 and SciFinder. A there was no sign in the chromatogram regarding the correspondent
further comparison was made with in silico fragmentation patterns RTs in the blank plasma sample (Figure 1). Thereby, these results
provided by MetFrag Web tool,27 which was combined with other corroborate that the plasma matrix did not significantly affect the
databases such as KEGG, LipidMaps and PubChem. The search for detection of PGE2 in plasma samples from the ex vivo anti-
candidates on databases and the fragment comparisons provided for a inflammatory assay.
maximum error of 10 ppm and the compounds were identified consid- An analytical curve for PGE2 was obtained by the mean value of
ering the m/z of the adducts [M + H]+ and [M + Na]+ in positive and three individual curves constructed based on 11 calibration standards
[M − H]− in the negative modes. from spiked plasma samples ranging from 400 to 0.195 ng/mL. The
calibration curve was plotted using peak-area ratios between PGE2 to
IS versus the correspondent concentration of PGE2. The linear regres-
2.14 | Statistical analyses sion equation was y = 0.04341x + 0.07256 and the obtained correla-
tion coefficient (r2 = 0.9996) showed the linear fit of the curve.
Calculations of validation and quantification parameters for PGE2 Precision and accuracy were determined with four different concen-
were performed using Microsoft Excel (Microsoft® Office® 2010, trations for intraday and interday experiments. Precision was calcu-
Microsoft Corporation, Redmond, WA, USA), and the experimental lated as a percentage of the relative standard deviation (% RSD) and
results of the ex vivo anti-inflammatory assay in human whole blood accuracy as a relative error percentage (% RE). All results presented
and in vivo anti-inflammatory assay in mice were statistically analysed values that are according to limits from the validation guidelines29,30
via one-way analysis of variance (ANOVA) followed by Dunnett's mul- and are shown in Table S4. The concentration of 6.25 ng/mL of PGE2,
tiple comparison test on GraphPad Prism® 6 (GraphPad Software©, assessed in interday and intraday experiments to determine precision
La Jolla, CA, USA). and accuracy, was taken as the lower limit of quantification (LLOQ)
since this was the lowest concentration tested that showed reliable
precision and accuracy values. Recovery rates were calculated com-
3 | RESULTS AND DISCUSSION paring values of PGE2 in QC samples in the three concentrations in
the plasma and solution. The method presented recoveries that
3.1 | Validation of the PGE2 quantification method ranged from 40% to 45%, approximately, as shown in Table S5.
in plasma samples Preparation of plasma samples using SPE presented recovery
mean rates of around 42%, good precision, and accuracy (Tables S4
Chromatographic and SRM methods were determined and optimised and S5), according to results obtained by Galv~
ao et al.22 by using a
to improve the resolution between PGE2 and IS peaks with the lowest small number of steps for PGE2 purification and smaller amounts of
total running time (8.5 min) as well as a greater detection sensitivity. solvents. Thus, the sample preparation method reported here consists
The parameters in the analysis method for PGE2 quantification are in a cheaper and faster procedure compared to other procedures
shown in Table S2. Results showed that PGE2 and IS peaks could be described in the literature,22 providing suitable values for figures of
6 ROSA ET AL.

F I G U R E 1 Chromatograms obtained by SRM detection for PGE2 and IS from plasma sample in the upper limit of quantification (ULOQ)
(A) and from blank plasma sample after injection of the sample in the upper limit of quantification (B)

merit according to validation guidelines.29 Processed plasma samples

were kept on ice until the exact moment of the injection and analysis
with LC–MS/MS. Thus, PGE2 stability in these samples was deter-
mined in two conditions: short-term, comprising the time of prepara-
tion and analysis of a batch with 10 samples in a chromatographic run,
and long-term, including the time in which all processed samples were
kept on ice from the moment when they were processed until their
injection. Results (Table S6) showed that PGE2 had no substantial deg-
radation during these periods in the plasma samples analysed.

3.2 | Optimisation of dexamethasone

concentration as a positive control

An initial dexamethasone test (positive control) was performed and

compared with the negative control (LPS-induced blood without
treatment) and non-induced controls (without LPS), containing only
blood and PBS (non-induced 1) or with ethanol at 0.05% (v/v) (non- F I G U R E 2 Production of PGE2 (in ng/mL of plasma) analyzed by
induced 2). This test was performed in four replicates and all samples LC-MS/MS, prior to human whole blood antiinflammatory assay, to
determine the reference drug concentration to be used as the positive
tested were diluted with a PBS solution to a final volume of 250 μL in
control. The test was performed in four replicates (n = 4) and the
each well, according to procedures described earlier. Dexamethasone groups were compared using one-way ANOVA followed by Dunnett’s
was tested with two concentrations (1 and 5 μg/mL) with multiple comparison test. Bars represent mean ± standard deviation
LPS-induced blood. Figure 2 shows the results of PGE2 levels (SD) and * statistical difference with the negative control (p ≤ 0.05)
(ng/mL of plasma). Dexamethasone at 1 μg/mL demonstrated PGE2 and #(p ≤ 0.05) compared with dexamethasone at 1 μg/mL
production inhibition, as efficiently as dexamethasone at 5 μg/mL
when statistically compared with the negative control group. These screening assay of plant extracts. Also, it was possible to verify that
results led to the use of a lower concentration of reference drugs there was no remarkable PGE2 production in non-induced blood sam-
(dexamethasone and indomethacin) as positive controls in the ples (without LPS inflammatory stimuli) compared with the negative
ROSA ET AL. 7

control (Figure 2). This fact showed that LPS is really effective and the results of the statistical comparison of PGE2 concentration (ng/mL
necessary to induce PGE2 production for the whole blood ex vivo of plasma) in those samples.
assay. Moreover, the final concentration of ethanol (0.05%, v/v) in Among the 47 tested extracts, only 11 of them had a PGE2 con-
positive controls did not influence PGE2 production since a consider- centration that was lower than 12 ng/mL (Table 1). A comparative sta-
able reduction in the production of this analyte when compared with tistical analysis of these 11 extracts with negative control and a
negative control was not observed. Lastly, ethanol at 0.05% (v/v) did reference drug (indomethacin) (Figure 3) pointed out lower values of
not affect the reference drug's efficacy. PGE2 production for five samples when compared to the negative
control indicating their anti-inflammatory potential: Acacia polyphylla
(1), Poincianella pluviosa (13), Enterolobium contortisiliquum (18), and
3.3 | Assessment of the anti-inflammatory Holocalyx balansae (20), Peltophorum dubium (35). Tested extracts
potential of plant extracts using ex vivo assay in were at 10 μg/mL and reference drugs (dexamethasone and indo-
human whole blood methacin) at 1 μg/mL. Butterweck and Nahrstedt9 have suggested
that for in vitro assays, the acceptable concentration for biological
An anti-inflammatory assay in human whole blood was performed activities of extracts should be up to 100 μg/mL. Thus, the relatively
with 47 extracts from leaves of different species from the Fabaceae lower concentration of plant-derived samples that were active in this
family. Fresh blood was incubated in groups of four replicates for each test reflects their promising anti-inflammatory potential since the
plant extract or the reference drugs – dexamethasone and indometha- ex vivo assay is closer to the in vivo approach than the in vitro.31
cin at 1 μg/mL (positive controls) – together with LPS (using PBS for
dilution in all of them), according to the methodology described ear-
lier. Results of this anti-inflammatory activity screening were obtained 3.4 | Assessment of ear edema inhibition through
by comparing the PGE2 production between the extract samples and an in vivo anti-inflammatory assay
the controls (negative and positive). Table 1 shows the mean ± stan-
dard deviation (SD) of PGE2 production for each sample. From the An assessment of in vivo anti-inflammatory activity in mice was fur-
values of PGE2 produced in the blood induced by LPS and treated ther performed with those active samples in the ex vivo blood assay.
with different plant extracts (Table 1), it is possible to observe that The test assessed the inhibition potential for croton oil-induced ear
some plant species induced PGE2 production at higher quantities than edema. Table 2 shows the percentage of inhibition of an ear edema
the negative control. Thus, only the samples that presented a PGE2 for the tested samples (considering the mean edema) and Figure 4,
mean concentration that was lower than 12 ng/mL were statistically provides the comparative statistical analysis with a negative and posi-
compared with negative control and indomethacin by one-way tive control (dexamethasone) using one-way ANOVA followed by
ANOVA followed by Dunnett's multiple comparisons. Figure 3 shows Dunnett's multiple comparison. Results showed (Table 2 and Figure 4)

T A B L E 1 Prostaglandin E2 (PGE2)
Sample [PGE2] (ng/mL) Sample [PGE2] (ng/mL) Sample [PGE2] (ng/mL)
levels (ng/mL of plasma) in whole blood
lipopolysaccharide (LPS)-stimulated and Negative control 8.95 ± 1.27 15 33.98 ± 18.70 32 44.71 ± 31.41
treated with different plant extracts Dexamethasone 5.37 ± 0.55a 16 32.60 ± 13.62 33 41.24 ± 20.16
samples or reference drugs (mean ± Indomethacin 3.86 ± 0.29 a
17 13.58 ± 3.90 34 7.41 ± 1.38a
standard deviation, n = 4)
1 5.57 ± 0.69a 18 5.58 ± 1.17a 35 5.18 ± 0.66a
a
2 11.42 ± 3.85 19 29.41 ± 14.84 36 19.98 ± 11.55
3 18.66 ± 8.14 20 8.37 ± 5.06a 37 35.91 ± 18.81
4 17.27 ± 7.66 21 27.35 ± 11.13 38 48.10 ± 20.51
5 21.35 ± 7.26 22 12.72 ± 4.57 39 51.88 ± 23.73
6 46.54 ± 22.88 23 19.31 ± 8.82 40 12.50 ± 5.16
7 46.27 ± 23.43 24 37.54 ± 17.50 41 17.00 ± 6.10
8 13.92 ± 4.09 25 41.48 ± 19.95 42 8.01 ± 0.71a
9 13.30 ± 4.63 26 9.26 ± 2.21a 43 20.46 ± 24.62
10 23.27 ± 11.27 27 13.73 ± 5.23 44 12.91 ± 5.99
11 20.63 ± 12.64 28 13.38 ± 4.22 45 9.23 ± 1.75a
12 16.36 ± 6.23 29 41.77 ± 24.38 46 13.36 ± 3.84
13 6.62 ± 0.42a 30 28.95 ± 3.05 47 28.53 ± 14.12
14 8.57 ± 1.21a 31 39.51 ± 20.50 — —
a
Values of mean of PGE2 concentration lower than 12 ng/mL.
8 ROSA ET AL.

F I G U R E 3 Comparison of PGE2 levels (ng/mL) produced in

human whole blood induced by LPS and treated with different plant F I G U R E 4 In vivo anti-inflammatory activity in mice by ear edema
extract samples or reference drugs (dexamethasone and inhibition of the active samples screened by the human whole blood
indomethacin). The screening for determination of samples’ anti- anti-inflammatory assay. The test was performed in eight replicates
inflammatory activity was performed in four replicates (n = 4) and the (n = 8) and the groups were compared using one-way ANOVA
groups were compared using one-way ANOVA followed by Dunnett’s followed by Dunnett’s multiple comparison test. Bars represent mean
multiple comparison test. Bars represent mean ± SD. Statistical ± SD. Statistical difference is presented by *(p ≤ 0.05) compared with
difference is presented by *(p ≤ 0.05) compared with the negative the negative control and #(p ≤ 0.05) compared with dexamethasone
control and #(p ≤ 0.05) compared with indomethacin

T A B L E 2 Ear edema inhibition of Fabaceae samples tested in the

in vivo anti-inflammatory assay in mice (n = 8) Anti-inflammatory activity in both ex vivo and in vivo assays with
these four Fabaceae species in our study emphasise their great and
Percentage
promising pharmacological potential. Poincianella pluviosa has been
Sample inhibition (%)
reported for ear edema inhibition as well as cell recruitment.32 A pro-
Negative control —
teinase inhibitor from seeds of Enterolobium contortisiliquum pres-
Dexamethasone 64.10a
ented a significant reduction of the number of molecules, enzymes,
Indomethacin 61.89a and cells involved in the inflammation such as nitric oxide synthase
Acacia polyphylla (1) 50.12a (NOS) enzymes (iNOS and eNOS-positive).33 Studies of D-pinitol and
Poincianella pluviosa var. peltophoroides (13) 70.90a,b kaempferitrin isolated from Holocalyx balansae also have anti-
Enterolobium contortisiliquum (18) 41.52a inflammatory activity.34,35 Even though some Acacia species have
Holocalyx balansae (20) 41.26a already been studied in regards to their anti-inflammatory activity
Peltophorum dubium (35) 34.62 with extracts and isolated substances,36,37 this is the first time that
a
the extract of the Acacia polyphylla species has an anti-inflammatory
Samples statistically different from the negative control compared by
one-way analysis of variance (ANOVA) followed by Dunnetts's multiple activity reported.
comparison test (P ≤ 0.05).
b
Data from Domingos et al.32
3.5 | Metabolic profile of active extracts in ex vivo
human blood assay
that four out of five active species in the ex vivo human blood anti-
inflammatory assay presented remarkable inhibition for ear edema in To investigate and compare the chemical composition of the five
mice: Acacia polyphylla (1), Poincianella pluviosa (13), Enterolobium active plant extracts on the validated ex vivo human whole blood
contortisiliquum (18), and Holocalyx balansae (20). Croton-oil induced assay, the UPLC-ESI-QTOF-MS2 data were registered for both
ear edema inhibition for an ethanolic extract of Poincianella pluviosa ionisation modes. Most of the peaks were detected mainly in the neg-
has already been reported in the literature32 with 70.9%. Therefore, ative mode (Figure 5).
the in vivo anti-inflammatory activity was not assessed in our work or Many peaks could be identified through dereplication using com-
included in a comparative statistical analyses with one-way ANOVA prehensive databases and fragmentation pattern comparisons.27,38
(Figure 4). Identified compounds from the five active extracts in the ex vivo assay
ROSA ET AL. 9

F I G U R E 5 Chromatograms obtained by analyses in UPLC-ESI-QTOF-MS2 in the active species’ negative mode in the ex vivo anti-
inflammatory assay

are shown in Table 3. A heatmap in Supporting Information Figure S1 (18), and Holocalyx balansae (20). The concentration of those glycosides
presents a qualitative comparison of the identified compounds among was lower in Peltophorum dubium (35) extract, and this extract demon-
the extracts based on the peak area. strated ex vivo activity but was not active in the in vivo test.
A diversity of phenolic compounds was identified in the five Other polyphenols such as gallic acid, ellagic acid, and their deriv-
extracts, such as different subclasses of flavonoids, glycosides or agly- atives were identified mainly in Poincianella pluviosa (13) extract.
cones, and hydroxybenzoic acid derivatives (Table 3 and Figure S1). These compounds have been described with well-documented anti-
Many flavonoids have an antioxidant activity due to their hydroxyl inflammatory activity with different mechanisms.41–45 An overview of
groups in different positions in the structures and have great all phenolic compounds identified in our work shows that there were
39
anti-inflammatory potential. Quercetin, myricetin, luteolin, taxifolin, fewer and lower concentrations in Peltophorum dubium (35). These
glycosides or aglycones derivatives, presented an inhibitory results suggested that this class of substances may be important to
activity during the release of some pro-inflammatory mediators in in vivo anti-inflammatory activity for the four Fabaceae species, mainly
LPS-induced tests using isolated cells. Kaempferol glycosides presented quercetin and kaempferol glycosides.
weak inhibition on nitric oxide (NO) and PGE2 production in Saponins also represent a significant number of compounds
LPS-activated isolated cells. However, these kaempferol glycosides characterised in the active samples in both ex vivo and in vivo anti-
demonstrated good anti-inflammatory activity during the edema test – inflammatory assays in this study. In general, they were almost all
higher than quercetin glycosides.40 In our study, most of the identified detected in medium to very high intensities at least in one of the spe-
quercetin and kaempferol glycosides derivatives were common in active cies analysed in each of the genera, with most of them present in the
extracts in both ex vivo blood assay and in vivo edema assay − Acacia genera Acacia and Holocalyx. Saponins also have anti-inflammatory
polyphylla (1), Poincianella pluviosa (13), Enterolobium contortisiliquum activity, mainly those derived from oleanolic acid.46,47
10 ROSA ET AL.

TABLE 3 Compounds identified from active extracts dereplication in the ex vivo human whole blood assay

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
C1 [M + H]+ 0.383 182.9618 170.9256; 154.9657; 5.35 C4H6O2S3 Sulphur compound 1 1, 13,
138.9715; 18,
124.9386; 108.9611 20,
35
C2 [M – H]− 0.476 165.0389 152.9924; 147.0292; −6.15 C5H10O6 L-Xylonic acid 1, 13,
134.8664; 131.8282 18,
20,
35
C3 [M – H]− 0.476 195.0485 177.0377; 164.9839; 2.68 C10H12O2S Sulphur compound 2 1, 13,
159.0290; 152.9841 18,
20,
35
C4 [M – H]− 0.519 191.0546 179.0540; 177.0024; −5.05 C7H12O6 Quinic acid 1, 13,
173.0444; 171.0292; 18,
165.0390 20,
35
C5 [M – H]− 0.783 191.0199 179.0547; 175.0237; 3.77 C6H8O7 Succinate derivative 1, 13,
173.0440; 163.0592; 18,
161.0421; 157.0123 20,
35
C6 [M – H]− 0.891 204.0860 186.0737; 175.0273; −5.87 C8H15NO5 Amine dideoxyglycoside 13
169.0125; 165.0164;
159.0291; 158.0799
C7 [M + H]+ 0.933 332.1332 152.0689; 142.0853; −4.05 C14H21NO8 5’-O-beta-D– 13
136.0728; 124.0739; Glucosylpyridoxine
108.0792
C8 [M – H]− 1.374 737.2349 705.2072; 661.2184; 7.60 C34H42O18 Dihydroxymethylflavone 1, 13,
517.0815; 449.1273; triglycoside 18,
337.0184; 305.0855; 20,
249.0382; 205.0494; 35
193.0116
C9 [M – H]− 1.429 331.0651 300.9974; 275.0175; −4.30 C13H16O10 Galloylglucoside 13
271.0441; 241.0362;
211.0227; 169.0123
C10 [M – H]− 1.480 169.0130 157.0131; 153.0177; −4.14 C7H6O5 Gallic acid 13
151.0022
C11 [M – H]− 1.499 685.1825 667.1721; 635.0555; 8.22 C33H34O16 Flavonoid diglycoside 1, 13,
581.1365; 539.1256; derivative 18,
521.1138; 490.1547; 20,
432.1481; 363.0922; 35
345.0803; 303.0695;
217.0333; 175.0220
C12 [M – H]− 1.612 369.1388 325.1058; 261.0862; −2.41 C14H26O11 Dyssacaride 1, 18
237.0958; 204.0860;
160.0384
C13 [M – H]− 1.774 331.0654 313.0548; 300.9972; −3.40 C13H16O10 β-Glucogallin 13, 18,
271.0442; 211.0231; 20,
169.0119; 168.0046; 35
149.9942
C14 [M – H]− 1.787 343.0655 331.0656; 313.0545; −2.99 C14H16O10 Theogallin 13
300.9973; 275.0179;
247.0230; 211.0232;
191.0537; 173.0439;
169.0124
C15 [M + Na]+ 1.810 443.1659 281.1121; 263. 1,002; −5.17 C22H28O8 Lactone 20
245.0898; 217. 0957;

(Continues)
ROSA ET AL. 11

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
199.0848; 179.0724;
171.0901; 156.0708;
133.0742; 120.0792
C16 [M + Na]+ 1.823 166.0847 147.0428; 138.0538; 1.81 C7H13NO2 Stachydrine 13
124.0383; 120.0793;
153.0165
C17 [M – H]− 1.909 865.1989 577.1348; 543.0937; 1.05 C45H38O18 Proanthocyanidin 1 18
451.1024; 425.0863;
407.0762; 371.0611;
289.0697; 287.0540;
245.0435; 243.0277;
209.0282
C18 [M – H]− 1.946 315.0704 269.0429; 257.0436; −3.84 C13H16O9 Galloyl glycoside 13, 18,
243.0273; 229.0482; derivative 20,
209.0471; 194.0802; 35
189.0022; 175.0371;
161.0219; 152.0085
C19 [M – H]− 2.004 255.0497 245.0435; 209.0307; −3.06 C11H12O7 Piscidic acid 18, 20
193.0485; 179.0331;
165.0535; 161.0224
C20 [M – H]− 2.011 477.1238 449.0875; 407.0763; −1.33 C19H26O14 Dihidroxybenzoic acid 13, 18,
289.0700; 255.0489; diglycoside 35
168.0042; 153.0169
C21 [M – H]− 2.053 380.1540 320.1963; 304.1646; −7.92 C16H23N5O6 trans-Zeatin-7-beta-D- 20, 35
289.0689; 275.1389; glucoside
261.1222; 247.1063;
204.0640; 191.0170;
167.0327
C22 [M – H]− 2.067 433.0980 415.0860; 371.0604; −2.51 C18H26O8S2 Sulphur compound 3 1, 20
301.0555; 279.1068;
209.0287; 191.0184;
168.0044; 149.9943
C23 [M – H]− 2.123 447.1136 401.1072; 373.1116; −0.60 C18H24O13 Dihydroxybenzoate 13, 18,
341.0854; 315.0703; diglycoside 20
271.0808; 255.0492;
223.0573; 207.0282;
195.0629; 163.0370;
152.0090
C24 [M – H]− 2.132 461.1287 409.0769; 381.0413; −1.78 C19H26O13 Hidroxybenzoic acid 13, 18,
357.0457; 337.0191; diglycoside derivative 35
313.0550; 300.9979;
275.0176; 247.0231;
219.0284; 197.0441;
169.0124; 152.0097
C25 [M – H]− 2.159 495.0771 463.0509; 449.1273; −0.78 C21H20O14 Digalloylquinic acid 13
425.0142; 365.0174;
313.0546; 300.9975;
275.0182; 247.0231;
169.0126
C26 [M – H]− 2.189 355.0655 337.0554; 289.0695; −2.89 C15H16O10 Caffeic acid glucuronide 1, 18,
245.0793; 209.0282; 20
191.0177; 177.0177;
161.0226
C27 [M – H]− 2.194 219.0484 191.0535; 183.0274; −9.50 C8H12O7 Citric acid 35
179.0323; 173.0063
C28 [M + H]+ 2.203 146.0586 2.86 C9H7NO Indole-3-carbaldehyde 13, 18

(Continues)
12 ROSA ET AL.

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
130.0637; 127.0378;
123.0426; 118.0635;
117.0567; 115.0530

C29 [M – H]− 2.281 285.0603 249.0388; 247.0233; −2.61 C12H14O8 Uralenneoside 13

229.0121; 219.0281;
211.0233; 203.0331;
191.0315; 179.0328;
169.0123; 161.0224;
152.0097
C30 [M – H]− 2.282 313.0907 269.1006; 255.0480; −5.25 C14H18O8 Mandelic acid O-beta-D- 18, 20
248.9836; 221.0435; glucopyranoside
203.0317; 195.0546;
161.0430; 159.0284
C31 [M + H]+ 2.331 277.1656 247.0590; 163.0378; −3.07 C14H20N4O2 p-Coumaroylagmatine 18
147.0431; 139.0380;
119.0482
C32 [M – H]− 2.360 356.0974 300.9973; 289.0701; −2.13 C15H19NO9 Leucodopachrome 13
275.0179; 247.0232; glucoside
229.0122; 217.0125;
203.0327; 191.0325;
169.0126
C33 [M – H]− 2.3858 635.0876 613.0460; 605.0786; −1.33 C27H24O18 Trigalloyl glucoside 13
483.0769; 463.0509;
425.0139; 313.0547;
300.9973; 275.0181;
273.0027; 247.0231;
245.0075; 169.0126
C34 [M + H]+ 2.388 177.0540 163.0380; 149.0586; −6.61 C10H8O3 Herniarin 18
145.0275; 139.0383;
134.0349; 127.0382;
123.0431; 117.0326
C35 [M – H]− 2.427 417.1032 385.0804; 355.0673; −0.25 C17H22O12 Hydroxybenzoic acid 1, 18,
300.0539; 289.0697; diglycoside 20
285.0602; 255.0504;
241.0701; 209.0296;
191.0185; 179.0329;
152.0098
C36 [M – H]− 2.431 801.0792 757.0895; 745.0314; 0.66 C34H26O23 Mallotinic acid 13
633.0720; 613.0468;
603.0616; 463.0505;
425.0138; 316.9926;
300.9974; 273.0029;
247.0233; 169.0128
C37 [M + H]+ 2.438 867.2165 725.2104; 697.1581; 3.29 C45H38O18 Arecatannin A1 or B1 13, 18
611.1627; 579.1506;
437.0871; 425.0869;
409.0912; 360.1439;
301.0703; 289.0701;
247.0595; 208.0960;
191.0696; 163.0381;
139.0382
C38 [M – H]− 2.442 577.1350 451.1031; 425.0872; 0.68 C30H26O12 Procyanidin B5 1, 13,
407.0765; 355.0655; 18,
339.0858; 299.0555; 35
289.0703; 255.0282;
245.0792; 245.0426;
203.0693; 161.0224
(Continues)
ROSA ET AL. 13

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
C39 [M – H]− 2.480 865.1987 755.1810; 739.1681; 0.81 C45H38O18 Proanthocyanidin 2 13, 18
723.1934; 695.1418;
587.1198; 577.1351;
543.0944; 525.0820;
451.1022; 425.0867;
407.0764; 339.0848;
299.0549; 289.0702;
287.0544; 243.0282;
161.0220
C40 [M – H]− 2.486 353.0860 325.0909; 305.0645; −3.57 C16H18O9 Chlorogenic acid 35
289.0693; 267.0686;
249.0723; 219.0631;
195.0639; 191.0538;
179.0323; 173.0426;
163.0369
C41 [M – H]− 2.488 291.0131 235.9430; 207.9446; −3.42 C13H8O8 Brevifolincarboxylic acid 13
190.9418; 174.9682
C42 [M – H]− 2.533 595.1665 505.1344; 487.1233; 0.34 C27H32O15 Isobutrin 1, 18,
475.1235; 457.1130; 20
415.1022; 385.0916;
355.0810; 343.0806;
325.0703; 313.0703;
265.0338; 239.0547
C43 [M – H]− 2.610 755.2061 609.1467; 610.1500; 3.48 C33H40O20 Apigenin triglycoside 1, 20,
593.1531; 431.0952; derivative 35
430.0891; 369.0803;
299.0750; 285.0374;
284.0306; 239.0539
C44 [M – H]− 2.638 439.0864 417.0786; 395.0963; −2.86 C19H20O12 Malonyl chlorogenic acid 35
378.0040; 353.0864;
335.0748; 289.0695;
233.0641; 191.0535;
173.0432
C45 [M – H]− 2.656 593.1506 575.1267; 503.1190; −0.08 C27H30O15 Apigenin diglycoside 1, 13,
473.1083; 413.0868; derivative 18,
395.0762; 383.0761; 20
353.0654; 325.0699;
297.0749; 289.0700
C46 [M + H]+ 2.690 783.0703 483.0766; 471.0197; 2.80 C34H22O22 Punicalin 13
453.0094; 427.0296;
321.0239; 291.0861;
249.1113; 139.0383;
123.0432
C47 [M + H]+ 2.695 471.0204 453.0096; 427.0300; 0.91 C21H10O13 Valoneic acid dilactone 13
321.0243; 301.0707;
261.0391; 249.0391;
177.0538; 153.0175;
123.0430
C48 [M – H]− 2.716 289.0704 275.0179; 249.0387; −2.82 C15H14O6 Catechin 1, 13,
247.0230; 229.0118; 18,
205.0485; 203.0330; 35
187.0381; 175.0382;
169.0123; 159.0426
C49 [M – H]− 2.720 951.0746 933.0648; 915.0547; 0.65 C41H28O27 Geraniin 13
765.0580; 631.0578;
613.0474; 463.0511;
445.0404; 316.9929;

(Continues)
14 ROSA ET AL.

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
300.9978; 273.0028;
247.0234
C50 [M – H]− 2.723 239.0561 219.0281; 217.0481; 2.24 C11H12O6 2-Succinyl-6-hydroxy- 18
205.0480; 203.0697; 2,4-cyclohexadiene-
189.0050; 177.0172; 1-carboxylate
175.0370; 165.0168;
161.0221
C51 [M – H]− 2.730 863.1839 711.1355; 693.1248; 1.80 C45H36O18 Cinnamtannin D1 1, 18,
595.1656; 573.1031; 20
559.0876; 451.1017;
411.0709; 313.0810;
299.0545; 289.0699;
285.0389; 193.0488
C52 [M + H]+ 2.745 495.4271 421.3412; 350.2680; −0.64 C28H54N4O3 Sphingolipid derivative 1 18, 35
282.2416; 264.2307;
266.2458; 214.1900;
171.1477; 169.1321;
129.1374
C53 [M – H]− 2.752 465.1029 450.9932; 425.0138; −0.87 C21H22O12 Taxifolin glucoside 13
413.0869; 345.1172;
316.0210; 300.9975;
285.0390; 275.0181;
247.0234; 191.0538;
151.0018
C54 [M – H]− 2.759 337.0910 327.0701; 307.1001; −3.99 C16H18O8 p-Coumaroylquinic acid 35
289.0695; 233.0646;
191.0537; 173.0431
C55 [M – H]− 2.809 577.1344 559.1246; 473.1087; −0.36 C30H26O12 Proanthocyanidin 3 18
453.0650; 451.1024;
425.0866; 407.0765;
389.1809; 383.0764;
325.0699; 297.0748;
289.0703; 287.0547;
245.0800; 161.0228
C56 [M – H]− 2.839 953.0896 935.0787; 909.0678; −0.03 C41H30O27 Chebulagic acid 13
829.1099; 801.0793;
783.0688; 633.0727;
481.0617; 463.0513;
337.0189; 300.9976;
275.0182; 205.0490
C57 [M – H]− 2.840 563.1410 545.1300; 503.1190; 1.62 C26H28O14 Isovitexin 200 -O- 1, 18,
473.1082; 443.0974; arabinoside 20
425.0870; 413.0869;
383.0762; 353.0654;
325.0702; 297.0751
C58 [M + Na]+ 2.844 391.0810 379.0816; 361.0716; 4.16 C20H16O7 12-Deoxynogalonic acid 18
363.0861; 373.0715;
351.0855; 349.0710;
337.0711; 325.0707;
321.0754; 307.0604;
295.0599; 283.0598
C59 [M – H]− 2.859 447.0930 429.0816; 383.0761; 0.58 C21H20O11 Apigenin glycoside 1, 13,
357.0605; 327.0499; derivative 1 20
311.0547; 297.0396;
285.0389; 269.0442;
250.0704

(Continues)
ROSA ET AL. 15

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
C60 [M – H]− 2.874 299.0764 263.0550; 250.0705; −0.99 C13H16O8 Hydroxybenzoic acid 20
233.0448; 221.0433; glucoside
205.0357; 191.0326;
177.0385
C61 [M + Na]+ 2.879 329.0657 311.0538; 300.0594; −2.14 C18H12NO4 Quinoline derivative 1, 13
299.0543; 283.0590;
165.0169
C62 [M – H]− 2.881 479.0821 450.9931; 316.0213; −0.98 C21H20O13 Myricetin glucoside 13
300.9977; 271.0228;
247.0231; 214.0258
C63 [M – H]− 2.903 593.1512 563.1423; 473.1094; 0.93 C27H30O15 Quercetin 3-methyl ether 1, 18,
447.0945; 430.0905; 7-rhamnoside-30 - 20
383.0768; 353.0657; xyloside
325.0703; 299.0764;
285.0386; 255.0281
C64 [M – H]− 2.916 461.1086 443.1001; 383.0762; 0.46 C22H22O11 6-Methoxyluteolin 7- 1, 20
341.0656; 327.0494; rhamnoside
313.0355; 298.0470;
285.0389; 193.0128;
175.0020
C65 [M – H]− 2.925 739.2094 679.1515; 635.1616; 1.14 C33H40O19 Kaempferol 1, 20,
577.1585; 563.1409; 3-isorhamninoside 35
473.1082; 433.1135;
383.0763; 353.0656;
325.0866; 313.0769;
293.0443; 283.0596;
271.0602; 193.0127
C66 [M + H]+ 2.930 383.3743 309.2885; 287.0539; −1.79 C22H46N4O Pithecolobine 35
238.2151; 196.2043;
171.1476; 169.1321;
153.0167; 147.0425;
129.1371; 112.1108
C67 [M – H]− 2.960 379.1017 359.1474; 329.1365; −3.19 C18H20O9 4-Hydroxy-trans- 35
300.0254; 289.0700; cinnamoyl 2-O,3-O-
271.0223; 255.0277; diacetyl-beta-D-
233.0644; 191.0542; xylopyranoside
173.0433
C68 [M – H]− 3.009 615.0990 497.0350; 463.0864; 0.63 C28H24O16 Quercetin galloylglucoside 13
445.0397; 343.0438;
313.0549; 301.0247;
300.0260; 273.0030;
255.0284; 169.0121
C69 [M + H]+ 3.031 303.0498 295.0600; 283.0597; −2.24 C15H10O7 Quercetin 13, 18,
267.0645; 165.0171; 20,
163.0380; 153.0171 35
C70 [M – H]− 3.031 609.1460 577.1521; 503.1192; 0.71 C27H30O16 Quercetin 3-O- 1, 13,
461.1088; 445.1138; rhamnoside 7-O- 18,
417.1027; 371.0772; glucoside 20,
341.0661; 325.0706; 35
298.0465; 297.0397;
282.0521; 269.0446;
193.0494
C71 [M – H]− 3.044 577.1572 457.1176; 413.0873; 2.54 C27H30O14 Apigenin 7-rutinoside 1, 13,
387.0924; 353.0657; 18,
341.0659; 323.0553; 20
311.0552; 293.0447;
269.0442; 249.0601

(Continues)
16 ROSA ET AL.

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
C72 [M – H]− 3.059 607.1674 581.2246; 575.1361; 1.81 C28H32O15 Flavonoid glycoside 1, 18,
487.1245; 445.1133; 20
427.1030; 419.1703;
324.0627; 307.0600;
292.0363
C73 [M – H]− 3.085 431.0975 357.0963; 341.0659; −0.75 C21H20O10 Apigenin 7-O-beta-D- 1, 13,
323.0552; 311.0554; glucoside 18,
295.0601; 283.0601; 20
269.0443; 239.0698;
224.0466; 197.0584;
183.0433
C74 [M – H]− 3.094 300.9976 283.0596; 273.0023; −2.81 C14H6O8 Ellagic acid 13
257.0070; 245.0071;
229.0120; 217.0119;
201.0170; 185.0222;
173.0226
C75 [M – H]− 3.107 463.0877 301.0274; 300.0261; 0.10 C21H20O12 Quercetin glucoside 13, 18,
271.0234; 255.0283; 20,
243.0282; 227.0330 35
C76 [M – H]− 3.198 593.1511 545.1307; 503.1196; 0.76 C27H30O15 Luteolin diglycoside 1, 18,
473.1098; 459.1290; derivative 20
425.0873; 413.0877;
399.0924; 353.0661;
343.0816; 313.0726;
293.0446; 282.0523;
267.0710; 245.0922;
203.0809; 193.0491
C77 [M – H]− 3.233 951.0741 907.0858; 829.1099; 0.13 C41H28O27 Pentagalloyl glycoside 13
765.0575; 693.2404;
633.0737; 593.1509;
539.2130; 491.1917;
477.1033; 463.0537;
359.1488; 315.0709;
300.9985; 273.0027;
169.0127
C78 [M – H]− 3.266 599.1045 537.1961; 517.1920; 1.34 C28H24O15 Quercetin 13
463.0521; 431.0971; galloylrhamnoside
399.1650; 345.0603;
313.0536; 300.9978;
285.0384; 273.0025;
247.0226; 211.0238;
169.0125
C79 [M + H]+ 3.272 585.1249 453.0089; 337.0699; 0.79 C28H24O14 2”-O-Galloylisovitexin 13
313.0703; 295.0596;
283.0598; 267.0645;
163.0377; 153.0173
C80 [M – H]− 3.291 433.0764 301.0327; 300.0256; −1.59 C20H18O11 Quercetin 3-glycoside 35
271.0228; 255.0277;
243.0275; 227.0325;
178.9960
C81 [M – H]− 3.296 707.1836 645.1839; 605.1520; 1.78 C32H36O18 Kaempferide 1
563.1415; 545.1308; 3-rhamnoside-7-(600 -
473.1092; 455.0984; succinylglucoside)
443.0982; 413.0877;
395.0770; 353.0662;
341.0662; 323.0556;
293.0449; 281.0450
(Continues)
ROSA ET AL. 17

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
C82 [M – H]− 3.328 447.0923 426.1396; 337.0934; −0.98 C21H20O11 Apigenin glycoside 13,
284.0314; 255.0284; derivative 2 20,35
227.0333; 193.0487
C83 [M – H]− 3.3332 319.0445 300.9976; 284.0309; −2.81 C15H12O8 Ethyl 13
273.0024; 255.0285; brevifolincarboxylate
245.0077; 227.0335;
217.0127; 189.0174;
169.0126
C84 [M – H]− 3.417 575.1396 473.1082; 431.0978; −0.84 C27H28O14 Hydroxymethylglutaroyl 1, 18
413.0870; 341.0659; vitexin
311.0553; 283.0599
C85 [M – H]− 3.438 589.1564 527.1558; 487.1246; 1.13 C28H30O14 Flavonoid glucoside 1
445.1139; 427.1031; derivative
337.0710; 325.0710;
293.0446; 282.0523;
231.0289; 217.0127;
193.0492
C86 [M – H]− 3.494 625.1195 583.1089; 539.2129; 0.24 C30H26O15 Quercetin 13
553.1927; 487.1241; caffeylgalactoside
463.0872; 401.1801;
316.0213; 300.0260;
271.0250; 255.0287;
161.0227
C87 [M – H]− 3.501 593.1529 551.1415; 523.2186; 3.79 C27H30O15 Kaempferol 3-glucosyl-(1 1, 18,
489.1051; 429.1031; ! 2)-rhamnoside 20
361.1648; 285.0391;
257.0434; 229.0488
C88 [M + H]+ 3.522 271.0596 229.0483; 177.0543; −3.87 C15H10O5 Genistein 13
163.0378; 153.0170;
147.0427; 145.0273;
135.0424; 121.0270
C89 [M – H]− 3.630 315.0134 299.9898; 270.9868; −2.21 C15H8O8 3-O-methylellagic acid 13
242.9920; 227.1273;
216.0047; 200.0097;
183.1371; 160.0146
C90 [M – H]− 3.631 521.1304 461.1098; 431.0984; 1.69 C24H26O13 Dihydroxyflavone 1, 13,
399.0926; 367.1015; glycoside derivative 20
337.0933; 314.0421;
285.0402; 218.0441;
193.0481; 152.0091
C91 [M – H]− 3.651 431.0978 337.0913; 301.0703; −0.06 C21H20O10 Kaempferol rhamnoside 1, 20
285.0386; 284.0313;
255.0283; 227.0332;
193.0485
C92 [M + H]+ 3.694 218.2105 207.0641; 177.0533; −6.89 C12H27NO2 Sphingolipid derivative 2 1, 13,
167.0689; 152.0597; 18,
147.0425; 137.0584; 20,
127.0384 35
C93 [M – H]− 3.744 609.1254 567.1152; 535.3100; 1.58 C30H26O14 Quercetin p- 13
463.0874; 301.0318; coumarylglucoside
300.0260; 271.0236;
255.0284; 191.0542;
169.0124
C94 [M + H]+ 3.793 262.2369 207.0640; 200.1997; −5.03 C14H31NO3 Sphingolipid derivative 3 1, 13,
177.0536; 167.0690; 18,
147.0426; 137.0585 20,
35

(Continues)
18 ROSA ET AL.

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
C95 [M + H]+ 3.886 306.2637 287.0542; 265.1534; −2.40 C16H35NO4 Sphingolipid derivative 4 13, 18,
249.0777; 244.2261; 20,
207.0639; 191.0708; 35
177.0534; 167.0690;
163.0738; 147.0427;
137.0584; 127.0383
C96 [M + Na]+ 3.972 439.2303 287.0546; 207.0640; −1.12 C21H36O8 Atractyloside 20
177.0535; 167.0693;
147.0430; 137.0580
C97 [M – H]− 4.057 285.0390 265.0699; 241.0486; −3.21 C15H10O6 Kaempferol 1, 13,
217.0488; 205.0488; 18,
199.0383; 175.0378; 35
163.0382; 151.0017;
145.0277
C98 [M + H]+ 4.071 439.4375 365.3516; 294.2782; −0.20 C26H54N4O Budmunchiamine L4 18, 35
252.2674; 226.2514;
171.1479; 169.1323;
129.1374; 112.1109
C99 [M – H]− 4.110 359.0753 337.0898; 285.0383; −3.88 C18H16O8 6-C-Methylmyricetin 3,400 - 20
237.0382; 223.0587; dimethyl ether
193.0480; 165.0531
C100 [M – H]− 4.250 329.0291 314.0052; 300.0251; −1.96 C16H10O8 Di-O-methylellagic acid 13
298.9821; 270.9888;
255.0280; 242.9926;
214.9972; 187.0018
C101 [M + H]+ 4.300 465.4530 391.3674; 320.2937; −0.51 C28H56N4O Budmunchiamine L6 35
278.2824; 252.2671;
171.1479; 169.1322;
129.1374; 112.1108
C102 [M – H]− 4.322 971.4875 957.5084; 799.2125; 2.39 C48H76O20 Oleanolic acid derivative 1, 20,
776.2411; 645.3642; saponin 1 35
632.1992; 417.1174;
403.1022; 223.0590;
193.0483
C103 [M – H]− 4.336 271.0600 247.0228; 235.0606; −2.40 C15H12O5 Trihydroxyflavanone 13
227.1278; 223.0597;
205.0487; 195.0650;
191.0556; 183.1370;
179.0328; 175.0376;
169.0126; 163.0379;
161.0225
C104 [M – H]− 4.371 811.4488 667.2599; 455.0981; 0.99 C42H68O15 Saponin 1 1
427.1031; 399.0929;
353.1022; 341.1014;
293.0442; 223.0597;
193.0496
C105 [M – H]− 4.372 591.1717 535.1230; 489.1386; 0.53 C28H32O14 Biochanin A 7-O- 20
447.1279; 403.1023; rutinoside
327.0851; 285.0742;
270.0509; 193.0479
C106 [M – H]− 4.428 781.4379 763.4270; 657.3633; 0.60 C41H66O14 Saponin 2 1
631.3848; 607.3325;
587.2687; 549.3648;
503.2501; 495.3313;
473.3619; 427.1063;
413.0878; 401.1416;
353.1010; 341.1020;

(Continues)
ROSA ET AL. 19

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
282.0515; 249.0593;
223.0599; 193.0491;
183.1374
C107 [M + H]+ 4.471 271.0593 243.0639; 221.0703; −4.98 C15H10O5 Apigenin 13
207.0654; 177.0550;
163.0382; 153.0170;
147.0428; 137.0584;
119.0476
C108 [M – H]− 4.531 285.0404 271.0235; 265.0679; 1.70 C15H10O6 Tetrahydroxyflavanone 13, 20
255.0269; 227.0332;
211.1312; 193.0118;
192.0049; 183.1376;
177.0172; 171.1007;
169.0118; 163.0383
C109 [M + H]+ 4.557 467.4683 393.3825; 322.3094; −1.25 C28H58N4O Budmunchiamine G 35
299.1475; 281.1371;
254.2827; 169.1321
C110 [M – H]− 4.564 955.4934 937.4824; 893.4912; 3.29 C48H76O19 Oleanolic acid derivative 1, 20
697.1772; 681.1830; saponin 2
629.3695; 539.3724;
475.1228; 447.0967;
327.2152; 285.0384
C111 [M – H]− 4.566 327.2159 301.0694; 285.0381; −3.82 C18H32O5 Trihydroxyoctadienoic 1, 13,
255.0285; 234.1112; acid 18,
223.0585; 211.1313; 20,
201.1105; 193.0480 35
C112 [M + H]+ 4.620 290.2681 271.0218; 228.2306; −4.89 C16H35NO3 Sphingolipid derivative 5 1, 13,
207.0635; 184.2040; 18,
175.0378; 167.0696 20,
35
C113 [M – H]− 4.679 537.0820 443.0394; 417.0601; −0.33 C30H18O10 Amentoflavone 13
399.0494; 375.0494;
331.0591; 309.0386
C114 [M – H]− 4.757 403.1017 390.0015; 331.2462; −3.00 C20H20O9 Chalconaringenin 20 - 20
313.2362; 301.0690; xyloside
281.0640; 263.0540;
243.9798; 223.0585;
193.0477; 189.0527
C115 [M – H]− 4.828 329.2319 301.0691; 283.0591; −2.73 C18H34O5 Trihydroxyoctadecenoic 1, 13,
268.0355; 249.1082; acid 18,
243.0633; 229.1429; 20,
223.0584; 211.1317; 35
193.0482; 183.1362;
171.1000; 152.0087
C116 [M – H]− 4.988 287.2207 249.1077; 241.0083; −5.34 C16H32O4 Dihydroxy palmitic acid 1, 13,
223.0580; 211.1319; 20
197.1154; 193.0844;
183.1362; 178.0594;
163.0385
C117 [M – H]− 5.021 343.0444 328.0208; 312.9970; −2.90 C17H12O8 Tri-O-methylellagic acid 13
297.9737; 285.0022;
269.9785; 257.0076;
241.9832; 213.9888;
197.9935; 185.9944
C118 [M + H]+ 5.113 495.4994 473.4194; 421.4138; −1.59 C30H62N4O 14-Normethyl 35
350.3405; 308.3294; budmunchiamine K
282.3138; 171.1477;

(Continues)
20 ROSA ET AL.

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
169.1321; 129.1371;
112.1105
C119 [M – H]− 5.297 987.5172 923.4851; 879.5095; 0.73 C49H80O20 Oleanolic acid derivative 1, 20
795.4528; 733.4539; saponin 3
615.3891; 597.3791;
492.2708; 339.1999;
247.0808; 205.0713
C120 [M + H]+ 5.307 274.2736 256.2624; 242.2454; −3.66 C16H35NO2 Hexadecasphinganine 1, 13,
230.2466; 212.2358; 18,
201.0447; 106.0850; 20,
102.0899 35
C121 [M + H]+ 5.363 318.2996 300.2884; 256.2622; −3.83 C18H39NO3 Phytosphingosine 1, 13,
212.2353; 146.1161 18,
20,
35
C122 [M + H]+ 5.420 362.3260 300.2886; 282.3142; −2.85 C20H43NO4 Aminoicosane derivative 1, 13,
256.2624; 171.1476; 18,
169.1322; 146.1162; 35
132.1004
C123 [M – H]− 5.445 795.4536 765.4419; 733.4523; 0.65 C42H68O14 Saponin 3 1
687.3235; 615.3902;
543.2814; 457.3669;
339.1931; 311.1676;
205.0703
C124 [M – H]− 5.560 765.4430 719.3996; 655.3333; 0.63 C41H66O13 Saponin 4 1
633.4008; 615.3903;
595.2790; 543.2807;
457.3678; 339.1977;
325.1829; 269.1305
C125 [M + H]+ 5.563 789.4419 767.4611; 635.4167; −0.79 C43H64O13 Saponin 5 1
617.4054; 599.3955;
441.3727; 423.3624;
331.0635; 203.1782
C126 [M – H]− 5.655 985.5009 921.4840; 849.4618; 0.08 C49H78O20 Oleanolic acid derivative 1, 20
791.4202; 765.4402; saponin 4
719.3997; 675.4089;
633.3984; 593.2633;
327.2162; 183.0108
C127 [M – H]− 5.656 283.0588 268.0350; 267.0275; −6.54 C16H12O5 Dihydroxy 20
239.0325; 223.0372; methoxyflavone
211.0374; 195.0423;
183.0087; 167.0478
C128 [M – H]− 5.726 779.4581 717.4579; 631.3855; −0.09 C42H68O13 Saponin 6 1, 20
599.3946; 581.3843;
509.3994; 441.3714;
339.1971; 327.2161;
309.2050; 297.1318;
255.0290; 221.1521;
183.0108
C129 [M – H]− 5.779 895.5058 877.4943; 861.4457; 0.32 C47H76O16 Oleanolic acid derivative 1, 20
779.4579; 721.2403; saponin 5
687.4446; 633.3999;
599.3944; 577.1953;
559.1861; 541.1774;
509.3873; 415.1447;
397.1340; 383.1180;
339.1983; 327.2153;

(Continues)
ROSA ET AL. 21

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
309.2054; 293.2108;
220.1451; 205.0697;
183.0108
C130 [M – H]− 5.955 633.4003 615.3890; 577.1977; 0.06 C36H58O9 Saponin 7 1
559.1862; 457.3665;
437.3408; 409.3448;
397.1328; 339.1988;
325.1838; 311.1667;
293.2108; 182.0108
C131 [M + H]+ 6.034 302.3045 284.2932; 258.2769; −4.64 C18H39NO2 Sphinganine 1, 13,
171.1471; 169.1324; 18,
129.1370; 112.1102; 20,
106.0849 35
C132 [M + H]+ 6.076 346.3311 328.3199; 284.2935; −2.94 C20H43NO3 Sphingolipid derivative 6 1, 13,
240.2668; 171.1471; 18,
102.0896 35
C133 [M + Na]+ 6.169 699.3572 571.2428; 537.3038; 0.60 C33H56O14 Fatty acid glycoside 1, 13,
497.3121; 437.1928; 35
406.1804; 365.1064;
353.2686; 333.1508;
307.2621; 261.2208;
243.2097; 201.1649;
173.1307; 145.1001;
135.1157; 119.0841
C134 [M + Na]+ 6.212 437.1931 405.1660; 349.1807; −2.08 C24H30O6 Magnoshinin 1, 13,
314.7950; 305.1582; 18,
303.1157; 201.0464; 20,
135.0793; 119.0843; 35
117.0689; 115.0519
C135 [M – H]− 6.691 293.2101 279.2310; 277.2161; −5.35 C18H30O3 Octadecenoic acid 1, 20
265.1459; 255.2309; derivative
197.0255; 191.0060;
183.0103
C136 [M + H]+ 6.704 330.3360 312.3263; 307.2627; −3.64 C20H43NO2 Sphingolipid derivative 7 1, 13,
177.1105; 133.0849 20,
35
C137 [M + H]+ 6.906 356.3518 338.3413; 307.2624; −2.96 C22H45NO2 Sphingolipid derivative 8 1, 13
177.1118; 133.0850
C138 [M + H]+ 7.036 496.3395 459.2491; 313.2729; −1.64 C24H50NO7P Fatty acid phosphate 1
307.2629; 184.0723; derivative
177.1102; 133.0848;
124.9980
C139 [M – H]− 7.196 571.2879 397.1334; 391.2238; −1.08 C32H44O9 Triterpenoid derivative 1 1, 13
339.1985; 325.1836;
315.0474; 311.1666;
299.0416; 279.2309;
255.2310; 241.0097;
225.0058; 183.0102
C140 [M – H]− 7.349 311.1665 287.2209; 283.2605; 5.72 C20H24O3 Naphtalenedione 1, 13,
279.2308; 255.2305; derivative 18,
239.0683; 225.0564; 20,
207.0387; 197.0247; 35
191.0037; 183.0093;
170.0012
C141 [M – H]− 7.670 555.2836 483.2720; 397.1336; 5.51 C28H44O11 Saponin 8 1, 13,
339.1979; 325.1833; 20
311.1663; 299.0429;

(Continues)
22 ROSA ET AL.

TABLE 3 (Continued)

Retention Parent Error Molecular Identity or class of the

Identifier Ionisation time (min) ion (m/z) Fragment ionsa (m/z) (ppm) formula compound Speciesb
279.2304; 255.2317;
225.0059; 206.9955;
183.0106; 164.9848
C142 [M – H]− 7.881 483.2719 397.1340; 339.1979; −1.19 C22H45O9P Fatty acid phosphate 1
325.1832; 311.1667;
255.2314; 183.0108
C143 [M + H]+ 8.109 425.2146 365.1938; 341.3041; −6.93 C22H32O8 Ester terpenoid 1, 13,
335.2576; 307.2625; 18,
261.2202; 283.2614; 20,
184.0726; 177.1111; 35
165.0898;
133.0855
C144 [M + H]+ 8.302 607.2549 481.2605; 439.3420; −1.25 C35H34N4O6 Pheophorbide B 1, 13
393.2079; 351.2891;
333.2784; 307.2630;
263.2358; 245.2255;
221.1376; 195.1215;
177.1112; 133.0853
C145 [M + H]+ 8.505 797.5203 621.2721; 547.2347; −0.09 C47H72O10 Fatty acid glycoside 1, 18,
519.2395; 307.2631; derivative 20
263.2361; 245.2257;
177.1114; 133.0853
C146a [M + H]+ 9.001 593.2780 533.2551; 505.2263; 2.70 C35H36N4O5 Pheophorbide A 1, 13,
461.2326; 447.2177; 20,
433.2369; 307.2628; 35
177.1113; 133.0853

Dereplication of main detected peaks in chromatograms was made by comparison of m/z of parent ions with comprehensive databases as Dictionary of
Natural Products© (DNP), METLIN, and SciFinder. Further comparison was made with in silico fragmentation patterns provided by MetFrag Web tool
combined to other databases such as KEGG, LipidMaps and PubChem.
a
Fragment ions with m/z highlighted are those matched with databases (KEGG, LipidMaps or PubChem) with deviation up to 10 ppm.
b
The presence of putative identified compounds compared with mass databases in the extracts was based on peak area values above 2000.

Some sphingolipid derivatives could also be identified as sphingo- transcription of different genes for the production of several proteins
sine, hexadecasphinganine, phytosphingosine. In the literature, involved in inflammatory processes such as cytokines and chemokines
phytosphingosine and some derivatives are reported to have anti- [tumour necrosis factor-alpha (TNFα), interleukin-1 (IL-1), IL-6, IL-8],
inflammatory activity.48,49 iNOS, and COX.54 LPS causes inflammatory stimulation in a systemic
Other compounds with anti-inflammatory activity were also iden- way, involving the production of different entities like molecules and
tified in some Fabaceae species analysed as caffeic and quinic acids enzymes. These can lead to the production of other inflammatory
50 51 52
and their derivatives, fatty acids, and pheophorbides A and B. mediators as well as the recruitment of other types of cells as
Dereplication of the five active species of Fabaceae suggested neutrophils.53–55 Thus, the methodology used in the ex vivo human
that the identified compounds may act against inflammation alto- blood assay may be used for the assessment of different constituents
gether. Therefore, phytochemical studies with these five species may from LPS-induced inflammatory process, like COX and iNOS enzymes,
be carried out to guide the isolation of substances with unknown cytokines, and others, by enzyme immunoassays.20,21,56 Using fresh
structures and/or anti-inflammatory activity. human whole blood to estimate anti-inflammatory activity also pro-
vides advantages since it mimics the in vivo micro-environment with a
proper number of cells and nutrients.57 In this sense, the ex vivo whole
3.6 | Advantages and disadvantages of the human blood assay of our work presented an easy, robust, representative,
whole blood ex vivo assay and versatile methodology for the assessment of the anti-
inflammatory potential of a large number of samples.
Inflammation-induced by LPS is caused through the stimulation of dif- The reduction of PGE2 production observed in the ex vivo assay
ferent cell types from the autoimmune response, such as monocytes performed herein identifies the species from the Fabaceae family that
and macrophages.53 In these cells, LPS mainly activates the transcrip- can inhibit COX-pathway and/or AA liberation in inflammatory pro-
tion factor NF-κB, which moves to the nucleus and promotes the cesses. Inactive species in the ex vivo bioassay do not need to
ROSA ET AL. 23

necessarily be inactive in the in vivo anti-inflammatory test9 since we AC KNOW LEDG EME NT S
evaluated only one of the main inflammation mediators in the ex vivo The authors would like to thank Maria Ângela Rodrigues and all of the
assay and only one dose. Thus, inactive species in this ex vivo assay researchers and professors from the Department of Pathology and
may be active in higher doses or through other mechanisms of action. Parasitology at the Federal University of Alfenas (Unifal-MG) for LPS
Another limitation of this new whole blood ex vivo assay is related to donations and the use of their equipment. The authors are grateful to
the prior extraction of extract samples with hexane. This step can researcher MSc Jo~ao Pedro Elias Costa, and Professors Dr Marcelo
eliminate most of the non-polar compounds, even those with possible Polo and Dr Flavio Nunes Ramos, for all the support they provided
anti-inflammatory activity, such as fatty acids, steroids, etc. Thus, the during the collection and identification of plants during this research.
applicability domain of this method is to assess samples with medium The authors would also like to thank the Central Laboratory of Clinical
to high polarity compounds mostly. Analysis – LACEN (Unifal-MG) for providing the structure and techni-
Many reports describe the quantification of PGE2 or other cal support during the blood collection process and the financial sup-
inflammation mediators in human whole blood. Brideau et al.20 for port granted by the Coordenaç~
ao de Aperfeiçoamento de Pessoal de
example, used LPS-induced human blood to access the efficacy of Nível Superior – Brazil (CAPES) – Finance Code 001, Fundaç~ao de
selectivity COX-2 inhibitors. They used 500 μL of blood for each Amparo à Pesquisa do Estado de Minas Gerais – Brazil (FAPEMIG)
sample replicate, LPS at 100 μg/mL for the final concentration for APQ-02353-17, Conselho Nacional de Desenvolvimento Científico e
blood stimulation, and the detection of PGE2 used a radioimmuno- Tecnológico – Brazil (CNPq) 427497/2018-3 and FINEP, enabling this
assay kit. Liu et al.21 evaluated cytokines secretion in whole blood research. The authors also thank CNPq for the scientific research
phytohemagglutinin (PHA)-induced in response to treatments with awards of PS and JHGL.
immunosuppressants. They used 200 μL (for each sample replicate)
of human blood PHA-stimulated to detect cytokines with multiplex OR CID
immunoassay kits. Galv~ao et al.22 assessed the levels of eicosanoids Welton Rosa https://orcid.org/0000-0002-4825-1568
produced also in human whole blood upon stimulus with healthy Olívia da Silva Domingos https://orcid.org/0000-0003-3371-5777
and sick individuals. They assessed the difference in levels of Paula Pio de Oliveira Salem https://orcid.org/0000-0002-5842-
22 eicosanoids (including PGE2) detected by LC–MS/MS, by using 0150
1 mL of whole blood of sick and healthy volunteers for inflamma- Ivo Santana Caldas https://orcid.org/0000-0002-4937-2425
tory stimulation (for each sample replicate) and without any treat- Michael Murgu https://orcid.org/0000-0002-7489-7637
ment. In our work, we optimised and validated a novel methodology ~ Henrique Ghilardi Lago
Joao https://orcid.org/0000-0002-1193-
using less human whole blood (200 μL for each sample replicate) 8374
than the median used by the previous works. Additionally, the Patricia Sartorelli https://orcid.org/0000-0002-1624-8145
human blood was successfully stimulated with lower LPS concentra- Danielle Ferreira Dias https://orcid.org/0000-0001-9129-4734
tion (10 μg/mL) and treated with different samples (reference drugs Daniela Aparecida Chagas-Paula https://orcid.org/0000-0003-
or plant extracts samples). Comparisons of PGE2 levels using 2274-4919
LC–MS/MS detection made it possible to assess the anti- Marisi Gomes Soares https://orcid.org/0000-0001-9221-9867
inflammatory activity among tested samples as a screening method
before in vivo anti-inflammatory activity assessment. The use of less RE FE RE NCE S
blood and lower LPS concentration for blood stimulation, in our 1. Varga Z, Kriška M, Kristová V, Petrová M. Analysis of non-steroidal
method, allows us to test a large number of samples, with a cheaper anti-inflammatory drug use in hospitalized patients and perception of
their risk. Interdiscip Toxicol. 2013;6(3):141-144.
method and using less volumes of blood. Furthermore, considering
2. Hangai S, Ao T, Kimura Y, et al. PGE2 induced in and released by dying
the availability of LC–MS/MS equipment, our study suggests a cells functions as an inhibitory DAMP. PNAS. 2016;113(14):3844-3849.
cheaper method for anti-inflammatory activity screening of samples 3. Antonelli M, Kushner I. It's time to redefine inflammation. FASEB J.
compared to those which depend on immunoassay kits. 2017;31(5):1787-1791.
4. Montrose DC, Nakanishi M, Murphy RC, et al. The role of PGE2 in
The PGE2 quantification method from LPS-induced human whole
intestinal inflammation and tumorigenesis. Prostaglandins Other Lipid
blood using LC–MS/MS analyses proved to be fast, sensitive, precise, Mediat. 2015;117:26-36.
and accurate. Furthermore, the use of a large number of animals was 5. Saeki K, Yokomizo T. Identification, signaling, and functions of LTB4
avoided since few species were guided for in vivo testing from the receptors. Semin Immunol. 2017;33:30-36.
6. Meirer K, Steinhilber D, Proschak E. Inhibitors of the arachidonic acid
ex vivo experiment.
cascade: interfering with multiple pathways. Basic Clin Pharmacol
Therefore, the use of both ex vivo and in vivo anti-inflammatory Toxicol. 2014;114(1):83-91.
assays in this work comprised a great, reliable, and ethically accept- 7. McLain RF, Kapural L, Mekhail NA. Epidural steroid therapy for back
able approach for screening a large number of samples. From these and leg pain: mechanisms of action and efficacy. Spine J. 2005;5(2):
bioassays, the phytochemical study from those active plant extracts 191-201.
8. Machado GC, Maher CG, Ferreira PH, Day RO, Pinheiro MB,
may be guided by UPLC-ESI-QTOF-MS2 analysis to isolate, character-
Ferreira ML. Non-steroidal anti-inflammatory drugs for spinal pain: a
ise and identify more effective compounds in the anti-inflammatory systematic review and meta-analysis. Ann Rheum Dis. 2017;76(7):
therapy. 1269-1278.
24 ROSA ET AL.

9. Butterweck V, Nahrstedt A. What is the best strategy for preclinical RDC no 27/2012. http://bvsms.saude.gov.br/bvs/saudelegis/anvisa/
testing of botanicals? A critical perspective. Planta Med. 2012;78(8): 2012/rdc0027_17_05_2012.html
747-754. 30. Castro AP, Kawano T, Spelta LEW, et al. In vivo schistosomicidal activ-
10. Andrade ALP, Miotto STS, Santos ÉP. A subfamília Faboideae ity of 7-epiclusianone and its quantification in the plasma of healthy
(Fabaceae Lindl.) no Parque Estadual do Guartelá, Paraná, Brasil. and Schistosoma mansoni infected mice using UPLC-MS/MS.
Hoehnea. 2009;36(4):737-768. Phytomedicine. 2018;38(8):66-73.
11. Wink M. Evolution of secondary metabolites in legumes (Fabaceae). 31. Moleiro AF, Conceiç~ao G, Leite-Moreira AF, Rocha-Sousa A. A critical
South African J Bot. 2013;89:164-175. analysis of the available in vitro and ex vivo methods to study retinal
12. Santana DB, Costa RC, Araújo RM, et al. Activity of Fabaceae species angiogenesis. J Ophthalmol. 2017;2017:1-19.
extracts against fungi and Leishmania: vatacarpan as a novel potent 32. Domingos OS, Alcântara BGV, Santos MFC, et al. Anti-inflammatory
anti-Candida agent. Brazilian J Pharmacogn. 2015;25(4):401-406. derivatives with dual mechanism of action from the metabolomic
13. Nghonjuyi NW, Tiambo CK, Taïwe GS, et al. Acute and sub-chronic screening of Poincianella pluviosa. Molecules. 2019;24(4375):1-16.
toxicity studies of three plants used in Cameroonian ethnoveterinary 33. Theodoro-Júnior OA, Righetti RF, Almeida-Reis R, et al. A plant pro-
medicine: Aloe vera (L.) Burm. f. (Xanthorrhoeaceae) leaves, Carica teinase inhibitor from Enterolobium contortisiliquum attenuates pulmo-
papaya L. (Caricaceae) seeds or leaves, and Mimosa pudica nary mechanics, inflammation and remodeling induced by elastase in
L. (Fabaceae) leaves in Kabir chicks. J Ethnopharmacol. 2016;178: mice. Int J Mol Sci. 2017;18(2):1-18.
40-49. 34. Ló SMS, Duarte MR. Morpho-anatomical study of the leaf and stem
14. Šibul F, Orčic D, Vasic M, et al. Phenolic profile, antioxidant and anti- of Pau-Alecrim: Holocalyx balansae. Brazilian J Pharmac. 2011;21(1):
inflammatory potential of herb and root extracts of seven selected 4-10.
legumes. Ind Crop Prod. 2016;83:641-653. 35. Poongothai G, Sripathi SK. A review on insulinomimetic pinitol from
15. Kim E, Yoon KD, Lee W-S, et al. Syk/Src-targeted anti-inflammatory plants. Int J Pharma Bio Sci. 2013;4(2):992-1009.
activity of Codariocalyx motorius ethanolic extract. J Ethnopharmacol. 36. Bukhari IA, Khan RA, Gilani AH, Ahmed S, Saeed SA. Analgesic,
2014;155(1):185-193. anti-inflammatory and anti-platelet activities of the methanolic
16. Ojewole JAO, Adewunmi CO. Anti-inflammatory and hypoglycaemic extract of Acacia modesta leaves. Inflammopharmacology. 2010;18
effects of Tetrapleura tetraptera (Taub) [Fabaceae] fruit aqueous (4):187-196.
extract in rats. J Ethnopharmacol. 2004;95(2–3):177-182. 37. Leal LSS, Silva RO, Araujo TSL, et al. The anti-inflammatory and anti-
17. Queiroz AC, Lira DP, Dias TDLMF, et al. The antinociceptive and anti- nociceptive effects of proteins extracted from Acacia farnesiana
inflammatory activities of Piptadenia stipulacea Benth. (Fabaceae). seeds. Rev Bras Plantas Med. 2016;18(1):38-47.
J Ethnopharmacol. 2010;128(2):377-383. 38. Menger F, Ahrens L, Wiberg K, Gago-Ferrero P. Suspect screening
18. Rahman AHMM, Parvin MIA. Study of medicinal uses on Fabaceae based on market data of polar halogenated micropollutants in river
family at Rajshahi, Bangladesh. Res Plant Sci. 2014;2(1):6-8. water affected by wastewater. J Hazard Mater. 2020;401:123377.
19. Chagas-Paula DA, Zhang T, Costa FB, Edrada-Ebel RA. A metabolomic 39. Baldim JL, Alcântara BGV, Domingos OS, et al. The correlation
approach to target compounds from the Asteraceae family for dual between chemical structures and antioxidant, prooxidant, and
COX and LOX inhibition. Metabolites. 2015;5(3):404-430. antitrypanosomatid properties of flavonoids. Oxid Med Cell Longev.
20. Brideau C, Kargman S, Liu S, et al. A human whole blood assay for 2017;2017:1-12.
clinical evaluation of biochemical efficacy of cyclooxygenase inhibi- 40. Xiao J. Dietary flavonoid aglycones and their glycosides: which show
tors. Inflamm Res. 1996;45(2):68-74. better biological significance? Crit Rev Food Sci Nutr. 2017;57(9):
21. Liu Z, Yuan X, Luo Y, et al. Evaluating the effects of immunosuppres- 1874-1905.
sants on human immunity using cytokine profiles of whole blood. 41. Choubey S, Varughese LR, Kumar V, Beniwal V. Medicinal importance
Cytokine. 2009;45(2):141-147. of gallic acid and its esters derivatives: a patent review. Pharm Pat
22. Galv~ao AF, Petta T, Flamand N, et al. Plasma eicosanoid profiles Anal. 2015;4(4):305-315.
determined by high-performance liquid chromatography coupled with 42. Derosa G, Maffioli P, Sahebkar A. Ellagic acid and its role in chronic
tandem mass spectrometry in stimulated peripheral blood from diseases. In: Gupta S, Prasad S, Aggarwal B. Anti-inflammatory
healthy individuals and sickle cell anemia patients in treatment. Anal nutraceuticals and chronic diseases. Adv Exp Med Biol. 2016;928:
Bioanal Chem. 2016;408(13):3613-3623. 473-479.
23. Tubaro A, Dri P, Delbello G, Zilli C, Loggia RD. The croton oil ear test 43. Xu GL, Liu ZZ, Xie M, et al. Salicylic acid derivatives and other compo-
revisited. Agents Actions. 1985;17(3–4):347-349. nents from Gaultheria trichoclada. Chem Nat Comp. 2016;52(2):
24. Santos MFC, Alcântara BGV, Feliciano CR, et al. New bicyclic [3.2.1] 301-303.
octane neolignans derivatives from Aniba firmula with potent in vivo 44. BenSaad LA, Kim KH, Quah CC, Kim WR, Shahimi M. Anti-
anti-inflammatory activity on account of dual inhibition of PGE2 pro- inflammatory potential of ellagic acid, gallic acid and punicalagin A
duction and cell recruitment. Phytochem Lett. 2019;30(4):31-37. and B isolated from Punica granatum. BMC Complement Altern Med.
25. Cabrini DA, Moresco HH, Imazu P, et al. Analysis of the potential top- 2017;17(47):1-10.
ical anti-inflammatory activity of Averrhoa carambola L. in mice. Evid 45. Lorenzo JM, Munekata PE, Putnik P, Kovačevic DB, Muchenje V,
Based Complement Alternat Med. 2011ID908059;1-7. Barba FJ. Sources, chemistry and biological potential of ellagitannins
26. Guijas C, Montenegro-Burke JR, Domingo-Almenara X, et al. METLIN: and ellagic acid derivatives. Stud Nat Prod Chem. 2019;60:189-221.
a technology platform for identifying knowns and unknowns. Anal 46. Pollier J, Goossens A. Oleanolic acid. Phytochemistry. 2012;77:10-15.
Chem. 2018;90(5):3156-3164. 47. Wu P, Gao H, Liu J-X, Liu L, Zhou H, Liu Z-Q. Triterpenoids saponins
27. Ruttkies C, Schymanski EL, Wolf S, Hollender J, Neumann S. MetFrag with anti-inflammatory activities from Ilex pubescens roots. Phyto-
relaunched: incorporating strategies beyond in silico fragmentation. chemistry. 2016;134:122-132.
J Chem. 2016;8(3):1-16. 48. Pavicic T, Wollenweber U, Farwick M, Korting HC. Anti-microbial and
28. Song X, Amirkhani A, Wu JX, et al. Analytical performance of nano- anti-inflammatory activity and efficacy of phytosphingosine: an
LC-SRM using nondepleted human plasma over an 18-month period. in vitro and in vivo study addressing acne vulgaris. Int J Cosmet Sci.
Proteomics. 2016;16(16):2118-2127. 2007;29(3):181-190.
29. ANVISA. Bioanalytical method validation guidelines. Brasilia: Brazilian 49. Kim B-H, Lee JM, Jung Y-G, Kim S, Kim T-Y. Phytosphingosine deriva-
National Health Surveillance Agency. ANVISA; 2012 Resolution – tives ameliorate skin inflammation by inhibiting NF-κB and JAK/STAT
ROSA ET AL. 25

signaling in keratincoytes and mice. J Invest Dermatol. 2014;134(4): 56. Bodet C, La VD, Epifano F, Grenier D. Naringenin has anti-
1023-1032. inflammatory properties in macrophage and ex vivo human whole-
50. Fraige K, Dametto AC, Zeraik ML, et al. Dereplication by HPLC- blood models. J Periodontal Res. 2008;43(4):400-407.
DAD-ESI-MS/MS and screening for biological activities of 57. Silva D, Ponte CGG, Hacker MA, Antas PRZ. A whole blood assay as
Byrsonima species (Malpighiaceae). Phytochem Anal. 2018;29(2): a simple, broad assessment of cytokines and chemokines to evaluate
196-204. human immune responses to Mycobacterium tuberculosis antigens.
51. Morais SM, Nascimento JET, Silva AAS, Honorio Junior JER, Acta Trop. 2013;127(2):75-81.
Pinheiro DCSN, Oliveira RV. Fatty acid profile and anti-
inflammatory activity of fixed plant oils. Acta Sci Veter. 2017;45
(1437):1-8. SUPPORTING INF ORMATION
52. Islam MN, Ishita IJ, Jin SE, et al. Anti-inflammatory activity of edible Additional supporting information may be found online in the
brown alga Saccharina japonica and its constituents pheophorbide A
Supporting Information section at the end of this article.
and pheophytin A in LPS-stimulated RAW 264.7 macrophage cells.
Food Chem Toxicol. 2013;55:541-548.
53. Guha M, Mackman N. LPS induction of gene expression in human
monocytes. Cell Signal. 2001;13(2):85-94. How to cite this article: Rosa W, da Silva Domingos O, de
54. Ghosh S, Hayden MS. New regulators of NF-κB in inflammation. Nat Oliveira Salem PP, et al. In vivo anti-inflammatory activity of
Rev Immunol. 2008;8(11):837-848. Fabaceae species extracts screened by a new ex vivo assay
55. Liu Z, Wang Y, Wang Y, et al. Dexmedetomidine attenuates inflamma-
using human whole blood. Phytochemical Analysis. 2021;1–25.
tory reaction in the lung tissues of septic mice by activating choliner-
gic anti-inflammatory pathway. Int Immunopharmacol. 2016;35: https://doi.org/10.1002/pca.3031
210-216.