Abbreviations
AKT1:
RAC-alpha serine/threonine-protein kinase
BCL2:
B cell leukaemia/lymphoma 2
CXCL:
chemokine (C-X-C motif) ligand
ECM:
extracellular matrix
EGFR:
epidermal growth factor receptor
H&E:
haematoxylin and eosin
IHC:
immunohistochemistry
IL:
interleukin
MAP3K14:
mitogen-activated protein kinase kinase kinase 14
MAPK:
mitogen-activated protein kinase
MMP9:
matrix metallopeptidase 9
NFκB:
nuclear factor kappa-light-chain-enhancer of activated B
NOAEL:
no observed-adverse-effect level
NSCLC:
non-small cell lung cancer
PI3K:
phosphatidylinositol 3-kinase
RAF:
RAF proto-oncogene serine/threonine-protein kinase
VEGF:
vascular endothelial growth factor
VEGFA:
vascular endothelial growth factor alpha
Introduction
Cancer progression relies upon malignant cell proliferation and the generation of
new blood vessels (angiogenesis) to sustain survival and invasion (metastasis). Lung
and liver cancers are amongst the leading causes of cancer-related deaths worldwide
(about 2 million new cases annually). Erlotinib is an FDA approved EGFR tyrosine-kinase
inhibitor for treating locally advanced or metastatic NSCLC [1]. In a phase III study, Erlotinib significantly improved the overall survival, relative
to the supportive care for refractory stage IIIB/IV NSCLC [2]
[3]. However, Erlotinib often causes side effects such as weakness, diarrhoea, rash,
shortness of breath, cough, fever and loss of appetite, dry eyes, unusual eyelash
growth, swollen cornea, extreme tiredness, and nausea. Sometimes it causes more serious
side effects such as interstitial lung disease, liver and kidney damage, gastrointestinal
perforation, blistering and skin peeling, and bleeding and clotting problems, which
may lead to a heart attack, stroke, and death [4].
The leaves of Morinda citrifolia L. (Rubiaceae) or noni (America)/mengkudu (Malaysia) are often used as vegetables
or salads. The M. citrifolia leaves contain epicatechin and scopoletin, purported to have immune-modulating [5], antioxidant, liver protective and wound healing effects without any acute, subacute,
and subchronic oral toxicity [6]. The NOAEL of oral M. citrifolia leaves ethanolic extract is 1 000 mg/kg [7]. M. citrifolia is beneficial for wound infections, pain, arthritis, swellings, and homeostasis.
Morinda fruit has been reported to have immunostimulant, antioxidant, anticancer,
and anti-inflammatory properties [8]
[9]
[10].
Patients and caregivers often use diets as complementary therapy to the prescribed
anticancer drugs. This report investigated the mechanisms (antiproliferation, anti-metastasis
and possibly antiangiogenesis) by which M. citrifolia leaf extract (standardised to scopoletin and epicatechin) could prevent the spread
of cancer in lung and liver metastasised tumours in vivo.
Results and Discussion
The control cancer-induced mice ([Fig. 1a]) displayed two adenomatous growths in the pulmonary parenchyma of the lungs with
large and medium-sized pseudostratified tumour cell clusters (black arrow) together
with goblet-like granules within their cytoplasm (yellow arrow). They also showed
two irregular tumour invasive glandular structures consisting of pseudostratified
lepidic growth (cells) with a pink amorphous proteinaceous secretion within the glandular
lumen ([Fig. 1b], black arrow). The light arrow shows free red blood cells within the alveolar lumen
and pulmonary parenchyma. The cancer-induced mice treated with 300 mg extract/kg body
weight had smaller tumours, with almost similar lung morphology to the control normal
mice, and indicated a 41% better antimetastatic effect than the 50 mg Erlotinib/kg
body weight treatment.
Fig. 1 The histological images of the metastasised cancers in the lungs and livers. [H&E;
x100; x200 (Untreated cancer a and b only)]. The adenomatous growth-like patterns in the lung were seen in the untreated
control cancer group. The poorly differentiated metastasised tumour cells in the liver
tumours were seen in the untreated control cancer group; (H&E; x200).
The metastasised liver tumours of the control cancer-induced mice ([Fig. 1c]) showed 60% tumour infiltration (metastasis), which is indicated by poorly differentiated
tumour cells and some areas of necrosis. After 3 weeks, the metastasised tumours in
the control cancer-induced mice were ~290 mm3, while tumours in the mice treated with 150 and 300 mg extract/kg body weight and
50 mg Erlotinib/kg body weight were significantly smaller (50, 95, and 87% smaller,
respectively). The mice treated with 150 mg extract/kg body weight showed tumour cell
infiltration to the liver and metastatic foci formation, while the Erlotinib-treated
mice showed hyperchromatic nuclei and hepatocytes cytoplasmic vacuolisation. The mice
treated with 300 mg extract/kg body weight showed no such changes.
[Fig. 2] shows the control cancer-induced mice overexpressed the EGFR in the metastasised
tumours (for both lung and liver). The mice treated with 300 mg extract/kg body weight
suppressed the EGFR expression more effectively than those treated with 50 mg Erlotinib/kg
body weight and 150 mg extract/kg body weight. The lung tumours expressed MMP9, while
the liver tumours expressed integrin-β1 ([Fig. 3]). These expressions were less severe in the extract- and Erlotinib-treated groups.
The mice treated with 300 mg extract/kg body weight had downregulated expression of
MMP9 and integrin-β1 in the tumours to near normal healthy levels. Integrin-β1 is the most copiously expressed integrin in NSCLC [11]. High MMP9 expression is an indicator for aggressive tumour growth in NSCLC [12].
Fig. 2 Immunohistochemical staining for EGFR in the metastasised tumours in the lungs and
liver (lung IHC x200; liver IHC x100). Means with different superscript letters within
the graph are significantly different (p<0.05).
Fig. 3 Immunohistochemical staining for MMP9 in the metastasised tumours in the lungs and
integrin-β1 staining in the livers (x100). Means with different superscript letters within the
graph are significantly different (p<0.05).
The 300 mg extract/kg body weight treatment effectively downregulated the BCL2, AKT1,
VEGFA, and MAP3K14 expressions in the lung tumours by more than threefold ([Fig. 4a]). The 50 mg/kg Erlotinib treatment only significantly downregulated the BCL2, AKT1,
and MAPK1 expressions (but not VEGFA or MAP3K14), while the 150 mg/kg extract treatment
only downregulated BCL2 and MAP3K14 expressions (relative to the untreated control
cancer group). In the liver, the 300 mg/kg extract treatment also significantly (p<0.05)
downregulated BCL2, AKT1, VEGFA, and MAPK1 ([Fig. 4b]) more effectively than the 50 mg/kg Erlotinib treatment.
Fig. 4 Mouse mRNA expressions in the lung a, liver b, and tumours c, and schematic representation of the signalling pathways involved in the inhibition
of angiogenesis and metastasis of lung adenocarcinoma by the Morinda leaf extract.
*Significant difference (p<0.05) between control and treatment groups. Adapted from
https://www.qiagen.com/my/shop/genes-and-pathways/pathway-details/.
This vegetable extract showed no adverse changes in the mice behaviour, body, or food
and water intake at the given dose, with a reported oral NOAEL of 1 000 mg/kg [7], unlike most cancer chemotherapy. The 300 mg/kg extract dose is equivalent to consuming
about 100–125 g of fresh leaves daily for a 50-kg adult (FDA animal dose to human
conversion guidelines).
Epicatechin and scopoletin are common compounds in edible leaves and may be a potential
dietary therapy against carcinogenesis. Scopoletin or epicatechin when used alone
produces weak cytotoxic activities towards various cancer cell lines. The IC50 of scopoletin for A549 lung cancer was above 100 µM [13]. Scopoletin was reported to boost the apoptosis of human prostate tumour PC3 cells
[14] and the human leukaemia cell line HL-60 [15], while (−)-epicatechin could inhibit growth and induced apoptosis in SW480 human
colon cancer cells [16]. When combined, epicatechin and scopoletin may synergistically suppress cell proliferation.
Scopoletin possesses antiangiogenic properties by inhibiting (a) endothelial cells
growth and migration, (b) extracellular signal-regulated kinase (ERK) 1/2 activation,
(c) tube formation, and (d) VEGF expression through NFκB [13]. Scopoletin, which is a coumarin, also helped activate protein kinase C (PKC) to
induce normal T lymphocytes cell proliferation without mitogen stimulus [17].
(−)-Epicatechin enhanced curcumin apoptotic effects towards human lung cancer cells
[18]. It also impaired angiogenesis, arrested metastasis through metalloproteinases inhibition,
and helped reverse multidrug resistance [19]. Erlotinib’s antiangiogenic properties were demonstrated by its ability to inhibit
human umbilical vein endothelial cells (HUVECs) growth and xenograft vessel density
[20]. Kaempferol in the leaf extract can also inhibit cancer [5].
The VEGF/VEGF-receptor pathway involving the PI3K/AKT and RAS/RAF/MAPK pathways is
important for cancer cell proliferation, angiogenesis, migration, and invasion [21]. The VEGF and PI3K/AKT/MTOR pathway is critical for the fibronectin-integrin effects
on proliferation. The control cancer-induced mice overexpressed EGFR in both lung
and liver tumours (brown IHC staining and strong positive signal by DAB visualisation).
The leaf extract dose-dependently restored the mice tumours towards normal healthy
conditions by suppressing cancer cell proliferation and metastasis through the suppression
of EGFR, MMP9, and integrin-β1 activities, and by inhibiting the VEGF/EGFR/NFκB signalling pathway.
The VEGFA activates endothelial cells to produce MMPs that break down the stroma and
ECM proteins [22] for angiogenesis and metastasis. The MMP9 are activated by fibronectin via the PI3K/AKT
or RAS/RHO/MAP pathway [23]. The vegetable extract dose-dependently suppressed angiogenesis-related mRNA expressions
(MMP9, VEGFA, and MAPK1), indicating reduced ECM degradation and remodelling. MMP9
is linked to the vascular remodelling and aggressive invasion of lung cancers [21]. The 300 mg extract /kg body weight treatment downregulated MAP3K14 (or NFκB-inducing kinase; NIK) that consequently helped inhibit downstream metastasis genes
including MMP9, VEGF, and the urokinase-type plasminogen activator receptor (uPAR)
[24]. The 50 mg Erlotinib/kg body weight treatment did not show this effect.
The leaf extract also dose-dependently suppressed BCL2 and AKT1 expressions. The VEGF
induces BCL2 expression. The upregulated BCL2 from the endothelial cells subsequently
initiated the nuclear factor of kappa light polypeptide gene enhancer in the B cells
inhibitor (IκB)/NFκB-dependent pathway, which elevated proangiogenic IL8 and CXCL1 expressions [25]. High AKT1 activation in all NSCLC subtypes induced endothelial cell migration via
nitric oxide signalling [26]. AKT regulated tumour angiogenesis through downstream targets such as the mammalian
target of rapamycin (mTOR)/p70S6K1 signalling axis, Forkhead box O (FOXO) inhibition,
VEGF mRNA upregulation, nitric oxide synthase induction, and/or glycogen synthase
kinase 3 beta (GSK3β) inhibition [27].
The extract dose-dependently downregulated integrin-β1 expression, which is necessary for angiogenesis and metastasis. Integrins are transmembrane
signal transduction receptors for attachment from the ECM to the cells. The lung cancer
cells are protected against apoptosis by the activation of integrin-β1 via ECM proteins (including fibronectin) [20]. In NSCLC, overexpression of integrin α5β1 was negatively associated with patient survival [28]. Activation of integrin-β1 on endothelial cells would trigger the transcription of a gene repertoire related
to angiogenesis [heparin-binding epidermal growth factor-like growth factor (HB-EGF),
IL8, CXCL1], adhesion [vascular cell adhesion molecule (VCAM), E-selectin], signal
transduction (NFκB), and coagulation (tissue factor) [29]. Integrins modulate the angiogenesis-related cell signalling pathways of transmembrane
protein kinases, such as receptor tyrosine kinases (RTK) [30].
The M. citrifolia (Noni) leaf extract was shown to inhibit proliferation and induced apoptosis in A549
cells (IC50=23.47 μg/mL) and mouse Lewis (LL2) lung carcinoma cells (IC50=5.50 μg/mL) in vitro by arresting the cancer cell cycle at G0/G1 phases and significantly increasing caspase-3/-8
without changing caspase-9 levels [31]. The A549 is an NSCLC cell line, which is the most common lung adenocarcinoma subtype
(with high mortality rate).
The vegetable previously demonstrated anticancer properties by increasing the proapoptotic
(TRP53) genes, downregulating the pro-tumourigenesis genes (BIRC5, JAK2/STAT3/STAT5A),
increasing anti-inflammatory biomarkers [IL4, IL10 and glucocorticoid receptor (NR3C1)],
enhancing the antioxidant NFE2L2 [nuclear factor (erythroid-derived 2)-like 2]-dependent
responses against oxidative injuries, modulating the immune responses (increasing
blood lymphocytes, spleen tissues B cells, T cells, and natural killer cells), enhancing
the tumour suppressor gene PTEN (phosphatase and tensin homolog), inhibiting cellular
tumour growth genes (MDM2, RAF1, MTOR) [5], inducing G0/G1 cell cycle arrest and the extrinsic apoptosis pathway, reducing
inflammatory markers cyclooxygenase 2 (COX2), increasing inflammatory cells clearance,
enhancing the efflux of inflamed tissues, suppressing oedema accumulation, and inhibiting
oxidative stress [31]. The vegetable extract was not cytotoxic on MRC5 normal lung cells (IC50>100.00 μg/mL).
These results demonstrated the Morinda leaf effects on metastasised cancer tissues
microstructure, indicating the suppression of cancer cell proliferation and vascular
and tissue remodelling, which were confirmed by mRNA expressions. The 300 mg extract
/kg body weight treatment appeared more effective than the 50 mg Erlotinib /kg body
weight treatment for most of the parameters measured.
Materials and Methods
Human lung adenocarcinoma (A549) cell lines were cultured in Kaighn's Modified Ham's
F-12 (F-12K) medium (ATCC) [5] containing 10% fetal bovine serum (PAA) and 1% of 100 µg/mL penicillin and streptomycin
(Biowest) in a humidified incubator at 37°C with 5% carbon dioxide.
Extraction and chemical analysis
The M. citrifolia leaves were identified and authenticated by the Biodiversity Unit, Institute of Bioscience,
UPM (Voucher No. SK2322/14). The leaves were dried and mixed in a ratio (w/v) of 1:5
with 50% ethanol in water. The total yield after extraction three times was 13.61%.
The extracts were analysed using HPLC (Waters 2996) with an Atlantis C18 column (4.6 mm×250 mm;
5 µm, Waters Corp.) maintained at 25°C. HPLC grade MeOH, acetonitrile (MeCN), and
analytical grade trifluoroacetic acid (TFA) were obtained from Merck. The mobile phase
consisted of three solvents: A: MeCN, B: MeOH, and C: 0.1% TFA in H2O (v/v), programmed consecutively in linear gradients as follows: 0 min, 10% A, 10%
B, and 80% C; 15 min, 20% A, 20% B, and 60% C; 26 min, 40% A, 40% B, and 20% C; 28-39 min,
50% A, 50% B, and 0% C; and 40-45 min, 10% A, 10% B, and 80% C. The elution was run
at a flow rate of 1.0 mL min−1 with a 50-µL sample injection volume and a UV spectra detector set at 210 and 450 nm.
Pure standards [scopoletin, and (−)-epicatechin] were purchased from Sigma-Aldrich.
The extract was standardised to scopoletin (2.2%, retention time, Rt=12.02 min) and
epicatechin (3.4%, Rt=9.17 min) as the main compounds, which were qualitatively and
quantitatively identified via the retention times and calibrated standard plots. Spiking
with scopoletin and epicatechin produced sharp extended peaks at the specified retention
times and their presence was further confirmed with LC-MS (Fig. 1S, Supporting Information). The epicatechin isomer in the extract was not determined because the HPLC retention
times between isomers are usually either very close to each other or they may overlap,
and LC-MS only confirmed the molecular weight of the parent and daughter molecules.
There have not been many reports on the difference in biological activities between
the epicatechin isomers. The isomer of the epicatechin may be determined in the future,
since the scope of this work was on the animal studies and biological activities.
Known chemical compounds in the M. citrifolia leaf have been reported and are shown in Table 1S, Supporting Information.
Animal studies
Male Balb/c mice (6 weeks old, weighing 19–20 g) from Faculty of Veterinary Medicine,
University Putra Malaysia, were given standard chow and water, and kept in a 12-h
light/12-h dark cycle [5]. The study was approved by the Institutional Animal Care and Use Committee (UPM/IACUC/AUP-R016/2013).
The A549 cells (2×107 in 100 µL PBS) were injected subcutaneously into the mice backs [32]. When the metastasised lung tumour size reached 100 mm3, 14 days after implantation, the mice were grouped (n=10) accordingly: (1) Control
healthy, (2) Cancer-induced untreated control (saline vehicle only), (3) Cancer-induced
and treated with 50 mg Erlotinib/kg body weight (orally gavaged daily), (4 and 5)
Cancer-induced and treated with 150 or 300 mg extract/kg body weight. After 21 days,
they were sacrificed via intraperitoneal injection of ketamine HCl (100 mg/kg) and
xylazine (10 mg/kg) and the tumour volume was measured [5]. The lung and liver tissues were snap frozen in liquid nitrogen for gene expression
analysis, while some were fixed in 10% formalin and embedded in paraffin for H&E and
IHC examination [5]. The IHC primary antibody kits (ChemMate DAKO EnVision Detection Kit, Peroxidase/DAB,
Rabbit/Mouse) were for (a) anti-EGFR (ab15669 from Abcam), (b) anti-MMP9 (Dako Corporation),
and (c) anti-integrin (Novus) [31].
The tissue mRNA was isolated using Trizol (Invitrogen) and analysed quantitatively
for VEGFA, AKT1, BCL2, MAP3K14, and MAPK1. The Custom RT2 Profiler PCR Array (CAPM11988), RT2 SYBR Green qPCR Mastermix, RT2 First Strand Kit, RNase-Free DNase Set (SuperArray Bioscience Corporation), and Data
Analysis version 3.5 (SABiosciences) were used, with heat shock protein 90 alpha (cytosolic),
class B member 1 (HSP90AB1; NM_008302), and glyceraldehyde-3-phosphate dehydrogenase
(GAPDH; NM_008084) as housekeeping genes for mRNA analysis [5].
All data are the mean±standard deviation and were analysed by one-way analysis of
variance (ANOVA) and Duncan’s test for significant differences (p<0.05) using IBM
SPSS Statistics 21 software.
