Content uploaded by Ahmed Mekkawy
Author content
All content in this area was uploaded by Ahmed Mekkawy on Oct 30, 2021
Content may be subject to copyright.
Am J Transl Res 2021;13(5):4309-4321
www.ajtr.org /ISSN:1943-8141/AJTR0124020
Original Article
Comparison of proteolytic, cytotoxic and anticoagulant
properties of chromatographically fractionated
bromelain to un-fractionated bromelain
Samina Badar1, Mohamed Azarkan2, Ahmed H Mekkawy1,3, Javed Akhter1,3, Krishna Pillai1,3, Rachida El
Mahyaoui2, Kevin Ke1, Lauren Cavanaugh4, David L Morris1,3
1Department of Surgery, University of New South Wales, St. George Hospital, Kogarah, NSW, Australia; 2Service
de Chimie Générale I (CP 609), Protein Chemistry Unit, Faculty of Medicine, Université Libre de Bruxelles (ULB),
Brussels, Belgium; 3Mucpharm Pty Ltd, Australia; 4Haematology Department, St. George Hospital, Kogarah, NSW,
Australia
Received October 12, 2020; Accepted March 2, 2021; Epub May 15, 2021; Published May 30, 2021
Abstract: Bromelain consisting of a number of proteolytic enzymes possess anticancer and thrombotic properties.
Hence, four chromatically separated fractions were examined for their proteolytic, anticancer and antithrombotic ac-
tivity. Bromelain fractions were separated using ion-exchange column chromatography. Proteolytic properties were
assessed using standard azocasein assay. Anticancer properties were rst assessed using four different cell lines
PANC-1, HEP 2B, HEP 3G and OVCAR-3 on cells grown in 96 well plates. Subsequently, fraction 2 and fraction 3
combined with gemcitabine were tested in ASPC-1 cells. Then cytotoxicity of fraction 3 was compared to bromelain
in combination with doxorubicin and N-acetylcysteine on HEP G2 and HEP 3B cells. Finally, the anticoagulation ef-
fect of fraction 3 or bromelain combined with N-acetylcysteine was evaluated using human blood. Fraction 3 showed
the highest proteolytic activity (5% greater than standard bromelain) whilst others were less active. Cytotoxicity as
assessed by IC50 indicated fraction 3 to be the most potent whilst the others did not follow their proteolytic potency
order. OVCAR-3 was the most sensitive amongst the cell lines. Fraction 3 showed higher potency in combination with
gemcitabine in ASPC-1 cells compared to fraction 2. Similarly, fraction 3 in combination with doxorubicin showed
higher toxicity when compared to bromelain. Fraction 3 or bromelain only showed thrombolytic activity in combina-
tion with N-acetylcysteine. Fraction 3 may be developed for clinical use since it showed better cytotoxicity compared
to bromelain.
Keywords: Bromelain, proteolysis, cytotoxic, coagulation, cancer
Introduction
Bromelain, an enzymic extract from the fruits
and stem of pineapple plant (Ananas comosus)
containing proteases, phosphatases, hydroxy-
lases, peroxidases, glycoproteins etc, has
shown anti-tumoural, anticoagulant, anticancer
and a variety of therapeutic properties in a
number of studies [1-4]. Further, it has muco-
lytic properties and in combination with
N-acetylcysteine serves as an efcient muco-
lytic and anticancer agent [5-7]. Currently, it is
undergoing clinical evaluation for the treatment
of mucinous tumours secreted by a rare dis-
ease known as pseudomyxoma peritonei [8, 9].
More recently, we have shown that the addition
of bromelain or N-acetylcysteine or their combi-
nations with cytotoxics such as gemcitabine,
doxorubicin, oxaliplatin etc, can potentiate the
action of these chemotherapeutic agents in
pancreatic cancer cell lines and in other can-
cers [10] with indication that the effective dos-
age of these chemotherapeutic agents may be
dramatically reduced.
Since bromelain is made up of a number of
enzymes and other proteinaceous component,
researchers have separated them into fractions
[11] in the hope of conning certain enzymic
and therapeutic properties exclusively to these
fractions. Hence, we have also separated bro-
melain into four fractions with varying proteo-
Comparison between fractionated and un-fractionated bromelain
4310 Am J Transl Res 2021;13(5):4309-4321
lytic activity. Although these fractions are com-
ponents of bromelain, they do not necessarily
share similar proteolytic potency, anti-tumoural
and blood anti-coagulant properties. Since
the anticancer effect of bromelain has been
attributed to its proteolytic properties [12], we
rst examined this property using chromato-
graphically separated bromelain fractions bo-
th, in their catalytically competent and irrevers-
ibly inactivated forms, with comparison to un-
fractionated bromelain after which, we exam-
ined the cytotoxic potency in four different
cell lines such as pancreatic cancer (ASPC-1),
hepatocellular cancer (HEP 3B, HEP G2) and
ovarian cancer (OVCAR 3), in vitro. Their poten-
cy may enable us to select fractions that have
good potential as chemotherapeutic adjuvant
agents. Further, investigation of fractions with
high anti-tumoural activity in combination with
gemcitabine or doxorubicin, two common cyto-
toxics, along with the addition of N-acetylcy-
steine was carried out in the hope of develop-
ing a more effective anti-cancer agent.
Although the blood anticoagulant properties of
bromelain are quite well known [13, 14], there
are paradoxical reports suggest that bromelain
may also cause minor coagulative disorder
especially when delivered intraperitoneally in
mice [15]. The intraperitoneal delivery of a com-
bination of bromelain with N-acetylcysteine for
mucinous tumours have resulted in elevation of
inammatory cytokines indicating that brome-
lain may act by degrading the surface layers of
the peritoneum. The extrinsic pathway in blood
coagulation may be at play in this instance
since the intrinsic coagulative pathway is prob-
ably inhibited by bromelain [16]. The antithrom-
botic properties of bromelain have been quite
well investigated in several studies and it has
been shown to interfere with the coagulation
cascade at several crucial points (intrinsic
pathway and the common pathway), thus serv-
ing as an anticoagulant [17, 18]. Since the anti-
coagulative action of bromelain has been linked
to its proteolytic properties, we investigated
the antithrombotic properties of bromelain
fraction with the highest proteolytic activity to
give an indication of its safety in patients on
anticoagulation therapy.
Materials and methods
Bromelain was purchased from Challenge Pty
Ltd, Taiwan, China and Sigma-Aldrich. All other
reagents used in this study are of analytical
grade and purchased from Sigma-Aldrich.
Iodoacetamide was purchased from Merck.
Fractionation and purication of Ananas como-
sus stem proteases
Stem bromelain proteases (Sigma-Aldrich, Ref.
B4882; 3 units/mg protein) were fractionated
as described previously [19]. Briey, stem bro-
melain powder was suspended in 100 mM
sodium acetate buffer, pH 5.0, in the presen-
ce of the reversible thiol-blocking reagent
S-methyl methanethiosulfate (MMTS, Sigma-
Aldrich, Ref. 64306) under constant moderate
stirring for 2 hours at 4°C. MMTS is added to
prevent autolysis and/or irreversible oxidation
of the catalytic cysteine residues of stem bro-
melain proteases. The resulting suspension
was ultracentrifuged (35000 × g, 4°C, 30 min)
and the supernatant constituting the total solu-
ble protein fraction (TE) was applied onto a
home-made SP-Sepharose Fast Flow column
(13 × 2.5 cm internal diameter; GE Healthcare)
pre-equilibrated with 100 mM sodium acetate
buffer pH 5.0 and eluted with a linear concen-
tration gradient from 100 mM to 800 mM sodi-
um acetate buffer, pH 5.0. The unbound mate-
rial was washed away with ten column volumes
of the pre-equilibrating buffer and elution of the
bound proteins was performed with a linear
concentration gradient of sodium acetate buf-
fer pH 5.0. The chromatographic fractions were
assayed for amidase activity using DL-BAPNA
(Sigma-Aldrich, Ref. B4875) as a substrate as
previously described [20]. The chromatograph-
ic fractions constituting the different proteases
were pooled according to their amidase activity
prole, concentrated by ultraltration and
exhaustively dialyzed against water at 4°C.
These fractions were lyophilized and stored at
-20°C until use, in their reversibly inhibited
forms, where the catalytic cysteine is S-thio-
methylated. In this form, the fractions become
fully active upon addition of an activator, such
us L-cysteine.
Irreversible inhibition of stem bromelain frac-
tions
Stem bromelain total extract (ET) and chro-
matographically obtained fractions (F1, F2, F3
and F4) were rst activated with dithiothreitol
(DTT, Sigma-Aldrich, Ref. 43815) at 5 mM nal
concentration for 10 minutes and subsequent-
ly irreversibly inhibited with a large excess of
Comparison between fractionated and un-fractionated bromelain
4311 Am J Transl Res 2021;13(5):4309-4321
iodoacetamide (Merck, Ref. 407710) at 10 to
20 mM nal concentration, until the inhibition
was completely achieved. After removing the
iodoacetamide excess by exhaustive dialysis,
residual proteolytic activity was checked uoro-
metrically using appropriate substrates [19].
The irreversibly inhibited samples were lyophi-
lized and stored at -20°C until use.
Proteolytic properties of fractions (F1-4) in
comparison to un-fractionated bromelain
Two hundred μg/ml of the different forms of
bromelain (fractions 1-4, un-fractionated Sig-
ma-Aldrich and Challenge bromelains) were
prepared in 1X phosphate buffer saline at pH
7.0, containing 5.0 mM L-cysteine. To 250.0 μl
of the different bromelain solutions was added
250.0 μl of azocasein (1.0% (w/v) solution in
distilled water). The mixture was agitated in a
shaker for 30 minutes at room temperature
(23°C) and then 1.5 ml of trichloroacetic acid
(5.0% (w/v) solution in distilled water) was
added to the mixture and vortexed. The precipi-
tated azocasein was centrifuged at 2500 rpm
for 6 minutes and 150.0 μl of supernatant was
pipetted out into a 400 μl microwell. To this,
150.0 μl of 1.0 N NaOH solution was added
and the absorbance at 440 nm was measured
[21]. The absorbance for the four fractions were
then compared to assess their relative proteo-
lytic activities.
Cytotoxic effect of the four bromelain fractions
(F1-4)
The cancer cells were grown in cell culture
medium (RPMI) containing 10.0% foetal bovine
serum in humidier at 37°C with 5.0% carbon
dioxide, following a standard protocol. After
three passages the cells were trypsinised and
then seeded into a 96 well plate at 4000 cells/
well and allowed to anchor overnight. The cells
were then treated with varying concentrations
of each fraction (F1-F4) and un-fractionated
bromelain (control) in RPMI and incubated at
37°C, for 72 hours, at the end of which the
media was decanted, and the plates were xed
using cold 10.0% trichloroacetic acid for 30
minutes at 4°C. The plates were then washed
with tap water and dried overnight at 23°C,
after which they were subjected to sulfhorda-
mine B (SRB) assay following a standard proto-
col [22]. From the absorbance at 510 nm, the
50% inhibitory concentrations (IC50) in μg/ml
were determined.
Cytotoxic effect of bromelain fractions 2 or 3 in
combination with cytotoxic drugs
The tumour cell line ASPC-1 was seeded as
before in a 96 well plate and treated with vary-
ing concentrations of bromelain fractions 2 or 3
and in combinations with the cytotoxic drug
gemcitabine and then incubated over 72 hours
at the end of which they were subjected to the
SRB assay [22] as before with cells viability
assessment at 510 nm. Similarly, the tumour
cells HEP-G2 and HEP 3B were treated with
varying concentrations of doxorubicin in combi-
nation with bromelain and NAC or with fraction
3 and NAC.
Interestingly, when irreversibly inactivated un-
fractionated bromelain and chromatographi-
cally separated fractions 1, 2, 3 and 4 were
assayed in the presence of L-cysteine on ASPC-
1 cells in the range of 0-100.0 μg/ml, no cyto-
toxic activity was detected. These data clearly
demonstrated that the cytotoxic effects of bro-
melain fractions are linked to their proteolytic
activity (data not shown).
Determination of antithrombotic properties of
fraction 3
Prothrombin time (PT), activated partial throm-
boplastin time (APTT), international normalis-
ed ratio: INR, F-10 (%): factor 10 as a percent-
age: The citrated blood samples prior to, and
after addition of, bromelain fraction 3 were
centrifuged rst at 150 × g for 20 minutes and
then at 1200 × g for 10 minutes. The resulting
platelet poor plasma (PPP) was then obtained
which was used for the determination of PT,
INR, APTT and F10 using STA Neoptimal 10,
STA TriniCLOT aPTT S, STA Decient X kits
(Diagnostica Stago Inc), respectively. All run on
instrument Diagnostica Stago STA-R Evolution
(Diagnostica Stago Inc).
Results
Fractionation and purication of Ananas como-
sus stem proteases
This rst step of ion-exchange chromatography
leads to four major protein populations accord-
ing to amidase activity measurements (Figure
Comparison between fractionated and un-fractionated bromelain
4312 Am J Transl Res 2021;13(5):4309-4321
1). According to SDS-PAGE analysis (Inset
Figure 1), fractions 1 and 4 are nearly homoge-
neous and pure. For F2 and F3 fractions, we
can see that in addition to the major protein
band at around 24 kDa which corresponds to
basic bromelain isoforms, another protein with
F1 (<21%), F4 (<37%) and F2 (<40%) had activi-
ties that were less compared to the bromelain
(C) control. Fractions F2 and F4 had almost
similar proteolytic activities. Hence, the order
of proteolytic activities with highest are F3, F1,
F4 and F2. (Figure 3 and Table 1).
Figure 1. Fractionation on SP-Sepharose Fast Flow of the whole soluble pro-
tein fraction from Ananas comosus stem. Fractions of 14.0 mL were col-
lected at ow rate of 60.0 mL/h and analyzed by absorbance measurements
at 280 nm (●), Na+ concentration (dotted line), and amidase activity (nkat/
chromatographic fraction) against DL-BAPNA (▲). Inset, SDS-PAGE: lane 1
and 7: molecular weight standard, lane 2: whole soluble protein fraction,
lane 3-6: fractions F1-F4, respectively.
Figure 2. Afnity chromatography on D-mannose-Agarose of F3 fraction ob-
tained after ion exchange chromatography on SP-Sepharose FF. The arrow
indicates the starting elution with D-mannose. Inset: SDS-PAGE. Lane 1: mo-
lecular weight standard, lane 2: F3 fraction obtained after SP-Sepharose FF,
lane 3: afnity chromatography ow through fraction (basic bromelain), lane
4: afnity chromatography retained fraction (lectin) and lane 5: afnity chro-
matography F2 ow through fraction.
an apparent molecular we-
ight of 14 kDa is also visible.
This later corresponds to the
Ananas comosus lectin [23].
F2 and F3 fractions were thus
further submitted to afnity
chromatography on a man-
nose-agarose support [23] to
obtain basic bromelain-enri-
ched fractions. SDS-PAGE ex-
periments (F3 taken as exam-
ple in Figure 2) clearly indicat-
ed that the ow through frac-
tions contained only the basic
bromelain isoforms (lane 3 in
Inset of Figure 2), lectin (lane
4 in Inset of Figure 2) being
specically retained on the
afnity support.
For the anti-antitumor stud-
ies, fractions F1 and F4 were
used without further puri-
cation, while for fractions F2
and F3, the mannose-agarose
ow through fractions were
assayed. Mass spectrometry
and N-terminal sequencing
analyses showed that F1 con-
tained two acidic bromelain
isoforms, F2 contained poorly
active basic bromelain iso-
forms, F3 contained highly
active basic bromelain iso-
forms and F4 contained anan-
ain isoforms [19].
Comparison of proteolytic
properties of fractions (F1-4)
in comparison to un-fraction-
ated bromelain
Proteolytic activities assess-
ed with equivalent quantities
of the various fractions indi-
cates that F3 has the high-
est activity, about 5.0% high-
er than the control bromela-
in (un-fractionated). Fractions
Comparison between fractionated and un-fractionated bromelain
4313 Am J Transl Res 2021;13(5):4309-4321
Cytotoxic effect of the four bromelain fractions
(F1-4)
In ASPC-1 cells, F3 is comparably more potent
than any of the other fractions and it is more
than twice as potent when compared to brome-
lain (C) (4.22 vs 10.00 μg/ml). Noticeably, F2 is
slightly more potent (about 20%) compared to
bromelain (C) (8.10 vs 10.00 μg/ml). When
compared to bromelain (C), F1 and F4 have
reduced potency (Figure 4A).
In HEP 3B cells, a similar trend was seen with
the different fractions, F3 being more potent,
about twice as potent as control bromelain (C)
(5.11 vs 11.98 μg/ml), whilst F2 is similar to
control bromelain (C). For the remaining frac-
tions, F1 is about 5 times weaker than control
bromelain (C) whilst F4 is about four times as
weak as control bromelain (C) (Figure 4B).
Fraction 3 shows superior cytotoxicity in HEP
G2 cells and is twice as potent as control bro-
melain (C) (5.77 vs 11.67 μg/ml), whilst F2
shows similar potency to control bromelain (C).
F1 shows almost about six-fold reduction in
cytotoxicity when compared to control brome-
lain (C) (72.00 vs 11.67 μg/ml) (Figure 4C).
A similar trend is shown in OVCAR 3, F3 is again
2.5 times as potent as control bromelain (C)
whilst F2 is only slightly more potent compared
to control bromelain (C) (3.71 vs 4.73 μg/ml).
The weakest fraction is again F1 followed by F4
(Figure 4D).
Hence, there seems to be trend in cytotoxicity
in the different fractions, indicating that F3 is
considerably more potent compared to the rest
whilst F2 seems to be much more potent com-
pared to F4 and F1 (Table 2).
Comparing the potency of F3 with that of F2 or
un-fractionated bromelain in the different cell
lines seems to indicate that F3 is twice as
potent as F2. Further there is indication (based
on IC50 values) that OVCAR 3 (ovarian tumour
cells) is more sensitive to these two fractions,
indicating that certain oncogenic cellular fea-
tures are targeted by these fractions (Figure
4E).
Cytotoxic effect of bromelain fractions 2 or 3 in
combination with gemcitabine on ASPC-1 cells
and doxorubicin in combination with NAC and
bromelain or fraction 3 on ASPC-1, HEP-G2
and HEP 3B cells
In ASPC-1 cells (pancreatic cancer cells), frac-
tion 3 has slightly higher cytotoxicity in combi-
nation with gemcitabine (5.0 μM) at the con-
centrations investigated (Table 3A). At 5.00
and 10.0 μg/ml, fraction 3 shows a slightly
greater difference in cytotoxicity. Table 3B indi-
cates that at 2.5 and 5.0 μg/ml concentration,
in the presence of 2.5 μM doxorubicin, un-frac-
tionated bromelain showed slightly higher cyto-
toxicity compared to fraction 3. However, at
10.0 μg/ml, fraction 3 exhibited a much higher
cytotoxicity (18% more) on hepatic cancer cells
(HEP G2) when combined with doxorubicin (2.5
μM). Table 3C indicates that in the presence of
Figure 3. The picture indicates the relative proteolytic
activities as compared to standard bromelain. BR (L)
= bromelain from Sigma-Aldrich; BR (C) = bromelain
from Challenge.
Table 1. The relative increase or decrease
in proteolytic activity of different bromelain
fractions (F1-F4) as compared to control
bromelain (L)
OD (440 nm) Z
Bromelain (L) 0.705
F1 0.555 <21
F2 0.417 <40
F3 0.736 >5
F4 0.441 <37
Fraction 3 shows a 5.0% increase whilst the others all
show lesser proteolytic activity (<37-<21). Comparative
proteolytic activity of fraction to control (Z) is calculated
as follows: X = [OD (fraction)/OD (control) × 100]; 100-X
= Y.
Comparison between fractionated and un-fractionated bromelain
4314 Am J Transl Res 2021;13(5):4309-4321
7.0 mM NAC in combination of 2.5 μM doxorubi-
cin, there is no difference in cytotoxicity be-
tween un-fractionated bromelain and fraction
3 at concentrations of 2.50 and 5.00 μg /ml
on HEP G2 cells (hepatic cancer cells). However,
at 10.0 μg/ml bromelain fraction 3 showed
greater activity (10% higher). The addition of
Determination of antithrombotic properties of
fraction 3
The prothrombin time (PT) indicates that the
addition of bromelain (C) in the range of 5.0-
10.0 μg/ml to blood had no effect; however the
addition of 10.0 mM NAC to bromelain (5.0-
Figure 4. The picture show the IC50
(μg/ml) of unfractionated bromelain
(C), bromelain fractions 1, 2, 3 and
4 in different cell lines (A) ASPC-1,
(B) HEP 3B, (C) HEP G2 and (D) OV-
CAR 3 cell lines. (E) shows IC50 (μg/
ml) of bromelain fraction 3 in ASPC-
1, HEP 3B, HEP G2 and OVCAR 3 cell
lines.
Table 2. Inhibitory concentration (IC50) μg/ml
FRACTION ASPC-1 HEP 3B HEP G2 OVCAR 3
F1 36.34 57. 3 1 72.00 16.95
F2 8.10 12.24 11.15 3.71
F3 4.22 5.11 5.77 1.80
F4 16.53 37.14 22.52 11.38
Bromelain (C) 10.00 11.98 11.67 4.73
The IC50 values for the different fractions indicate that their cytotoxicity var-
ies both, with the fractions and the cell lines.
Table 3A. ASPC-1cells after 72 hours treatment with gem-
citabine (5.0 µM) in combination with bromelain (C) or brome-
lain fractions 3
Concentration of BR (C)
or BR fraction 3 (μg/ml)
% alive
BR (C)
% alive
(F3)
% reduction
BR (C)
% reduction
(F3)
2.50 31.0 28.8 69.0 71. 2
5.00 31.1 24.1 68.9 75.9
10.0 26.1 18.9 73.9 81.1
2.50-10.0 μg/ml bromelain (C) or
fraction 3 with doxorubicin (2.5
μM) showed superior cytotoxic
performance by fraction 3 on
HEP 3B cells (hepatic cancer ce-
lls). A 20.0% increase in cytotoxic
efciency is obtained with fraction
3 compared to the un-fractionated
bromelain (C) (Table 3D). In the
presence of 7.0 mM NAC and 2.5
μM doxorubicin, at concentrations
ranging from 2.5-10.0 μg/ml,
there is only a marginal difference
in cytotoxicity between the un-
fractionated bromelain (C) and
fraction 3 (Table 3E). Figure 5
shows the cytotoxic effects of frac-
tionated and un-fractionated bro-
melain combination therapies on
pancreatic and hepatic cells in
vitro.
Comparison between fractionated and un-fractionated bromelain
4315 Am J Transl Res 2021;13(5):4309-4321
10.0 μg/ml) showed a signicant increase in
PT, indicating that the clotting mechanism will
be delayed. The addition of fraction 3 also did
not alter the PT whilst with the addition of NAC
10.0 mM to 10.0 μg/ml of fraction 3 increased
the PT substantially (21.6 sec). The addition of
10.0 mM of NAC had a substantial increase in
PT (25.4 sec). This indicates that both, brome-
lain (C) and fraction 3 should be combined to
NAC to obtain an anticoagulant effect.
Compared to control APTT, the clotting time
with the addition of bromelain (C) (5-10 μg/ml
seems to have a very minor effect on the APTT
(30.6-29.9 sec vs 31.5 sec for control). The
addition of 10.0 mM NAC to 5.0 or 10.0 μg/ml
(%), whilst the addition of NAC (10.0 mM) to bro-
melain (C) (5.0-10.0 μg/ml) showed a substan-
tial drop in F-10 (%) (82-79.5), indicating an
anticoagulant effect. Similarly, F3 on its own
had no effect whereas in combination with
10.0 mM NAC it showed an anticoagulant effect
(F-10% = 72). The addition of 10.0 mM NAC
showed a high drop in F-10 (%), indicating that
it has a considerable anticoagulant effect
(Table 4).
Discussion
The anticancer properties of bromelain have
been mainly attributed to its proteolytic com-
ponent [12] although it contains a number of
Table 3B. HEPG2 cells after 72 hours treatment with brome-
lain fraction 3 and doxorubicin (2.5 µM)
Concentration of BR (C)
or BR fraction 3 (μg/ml)
% alive
BR (C)
% alive
(F3)
% reduction
BR (C)
% reduction
(F3)
2.50 25.4 28.6 74.6 71. 4
5.00 24.4 29.5 75.6 70.5
10.0 23.9 7.60 76.1 92.4
Table 3C. HEPG2 cells after 72 hours treatment with BR Frac-
tion 3, NAC (7.0 mM) and doxorubicin (2.5 µM)
Concentration of BR(C)
or BR fraction 3 (μg/ml)
% alive
BR (C)
% alive
(F3)
% reduction
BR (C)
% reduction
(F3)
2.50 20.5 20.8 79.5 79.2
5.00 19.0 18.7 81.0 81.3
10.0 15.5 4.20 84.5 95.8
Table 3D. HEP3B cells after 72 hours treatment with BR (C)
and fraction 3 in combination with doxorubicin (2.5 µM)
Concentration of BR (C)
or BR fraction 3 (μg/mL)
% alive
BR (C)
% alive
(F3)
% reduction
BR (C)
% reduction
(F3)
2.50 27. 1 20.1 72.9 79.9
5.00 26.1 20.9 73.9 79.1
10.0 21.9 2.10 78.1 97.9
Table 3E. HEP3B cells after 72 hours treatment with BR (C) or
bromelain fraction 3 in combination with NAC (7.0 mM) and
doxorubicin (2.5 µM)
Concentration of BR
(C) or F 3 (μg/mL)
% alive
BR (C)
% alive
(F3)
% reduction
BR(C) % reduction (F3)
2.50 5.20 3.50 94.8 96.5
5.00 4.90 3.40 95.1 96.6
10.0 5.10 2.20 94.9 9 7. 8
bromelain (C) showed an increase
in APTT (32.9-32.7 sec), indicat-
ing that there is an anticoagulant
effect.
The addition of F3 showed a reduc-
tion of APTT value, 29.6 sec as
compared to control value of 31.5
sec, similar to that obtained for
bromelain (C). The addition of 10.0
mM NAC to 10.0 μg/ml of fraction
3 showed a small increase in APTT
(33.4 sec vs 31.5 sec (control)),
indicating a minor anticoagulant
effect. However, the addition of
10.0 mM NAC by itself gave a
value of APTT of 37.3 sec that
shows a modest anticoagulant
effect.
The INR values indicated that bro-
melain (C) (5.0-10.0 μg/ml) had
no effect on coagulation time,
whilst the addition of NAC (10.0
mM) to bromelain (C) (5.0-10.0
μg/ml) showed a substantial in-
crease in INR that is indicative of
delay in clotting (anticoagulant
effect). Likewise, F3 on its own
had no effect whilst in combina-
tion with 10.0 mM NAC it showed
an anticoagulant effect (INR =
1.59). Finally, the addition of 10.0
mM NAC showed a high INR, indi-
cating that it has a substantial
anticoagulant effect (Table 4).
The addition of bromelain (C) (5.0-
10.0 μg/ml) had no effect on F-10
Comparison between fractionated and un-fractionated bromelain
4316 Am J Transl Res 2021;13(5):4309-4321
Figure 5. (A and B) show the cytotoxic effect of fractions 2 or 3 in combination with gemcitabine on ASPC-1 cell line.
(C-F) show the cytotoxic effect of doxorubicin with NAC in combination with either fraction 3 or standard bromelain
(C) on hepatic cancer cell lines HEP G2 and HEP 3B.
Table 4. It shows blood parameters that were determined in the presence of different concentrations
of control bromelain (C) (BR) in μg/ml alone, control bromelain (C) (BR) + N-acetylcysteine mM (NAC),
fraction 3 (F3) alone, fraction 3 + NAC, and NAC alone
Parameter Control BR (5) BR (10)
BR (5)
+
NAC (10)
BR (10)
+
NAC (10)
F3 (10)
F3 (10)
+
NAC (10)
NAC (10)
PT (sec) 14.4 14.5 14.4 19.5 19.8 14.7 21.6 25.4
APTT (sec) 31.5 30.6 29.9 32.9 32.7 29.6 33.4 37. 3
INR 1.04 1.05 1.04 1.43 1.45 1.06 1.59 1.88
F-10 (%) 115 115 115 82 79.5 108.5 72 55
PT: partial prothrombin time; APTT: activated partial prothromboplastin time. INR: International normalised ratio (patient PT/PT
(control)), F-10: factor 10. Data highlighted in red indicate anticoagulant effect.
Comparison between fractionated and un-fractionated bromelain
4317 Am J Transl Res 2021;13(5):4309-4321
other enzymes such as carbohydrases, lectins,
phosphatases, hydroxylases, peroxidases and
glycosylases, etc [1, 4]. This is mainly due to the
digestive nature of proteases that can disinte-
grate protein components of numerous recep-
tors, signalling molecules and cancer related
molecules that provide resistance, metastasis
and accelerated replication in cancer cells [24].
Further, various mucins in cancer cells, both
secretory and transmembrane, are glycopro-
teins providing a barrier to drug penetration in
chemotherapy besides inducing accelerated
replication and resistance [25]. These muci-
nous glycoproteins are disintegrated by a com-
bination of bromelain and N-acetylcysteine
with disruption of their cellular biochemical and
physiological role [6, 7] and this mixture is cur-
rently used for treating mucinous tumours
secreted by a rare cancer known as pseudo-
myxoma peritonei [9]. Besides, bromelain has
also antithrombotic properties that affects vari-
ous components in the coagulation cascade
[17, 18]. Since bromelain is composed of sev-
eral enzymes, fractions with proteolytic activity
were isolated and their anticancer and anti-
thrombotic properties were evaluated in the
present study.
Proteolytic activity as assessed using the azo-
casein assay indicated that fraction 3 had the
best activity compared to control (5% higher).
However, the second potent fraction was F1 fol-
lowed by F4 and then F2. F1 was 21% less
active compared to control whilst F2 and F4
were almost of similar potency (40 and 37%
less). The differential proteolytic activities may
be related to the different enzyme composition
such as highly active basic bromelain isoforms
(F3), poorly active basic bromelain isoforms
(F2), ananain (F4) and acidic bromelain iso-
forms (F1) [19]. This indicates that if cytotoxici-
ty was mainly related to proteolysis, then the
following investigation that relates to inhibitory
concentration (IC50) of the four fractions
should follow a similar order of activity, although
other cellular features found within the cell
types may inuence cytotoxicity.
Cytotoxicity as assessed in pancreatic cancer
cells (ASPC-1) indicated that fraction 3 was the
most potent with the lowest IC50 value (twice
as potent as control bromelain (C)) indicating
that basic bromelain isoforms performs well in
a cellular environment with pH 7.0, followed by
poorly active basic bromelain isoforms (F2) that
may be again indicative of pH effect on their
proteolytic activity and hence their cytotoxicity.
Fraction 4 that is composed of ananain iso-
forms is less active whilst fraction 1 (acidic bro-
melain isoforms) is least active. A similar trend
existed in the other cell lines investigated such
as hepatocellular cancer cells and the ovarian
cell line. However, the trend in IC50 value as
indicative of proteolytic potency only showed
that F3 was the most active whilst fraction 4
also conformed to its order of proteolytic activ-
ity. Fractions 1 and 2 did not follow the order of
their proteolytic activity in cellular cytotoxicity.
Further, the proteolytic potency of un-fraction-
ated bromelain and F3 were almost equal (F3
exceeding by 5%) but their IC50 values were
very dissimilar and not proportional to their pro-
teolytic activity. Hence, these ndings suggest
that other anti-tumour intrinsic factors other
than their proteolytic properties contribute to
their cytotoxic potency. Amongst the four cell
lines investigated, OVCAR 3 was the most sen-
sitive with the lowest IC50 values for all the
fractions investigated. This suggests the pres-
ence of certain cellular features that are prone
to the action of bromelain which needs further
investigation. However, it is known that OVCAR
3 has high expression of mucins that are prone
to the hydrolytic action of bromelain [26]. The
second most sensitive cell line is the pancreat-
ic tumour cells, ASPC-1. These cells are also
highly mucinous expressing mainly of trans-
membrane types (MUC1, MUC16, etc.) [26, 27].
HEP 3B and HEP G2 that are of hepatic origin
appear to be almost equally affected by the
fractions but less sensitive (higher IC50 val-
ues). Further hepatic cell lines may have meta-
bolic enzymes that are capable of deactivating
bromelain and its chromatographically sepa-
rated fractions since the liver is primarily
involved in detoxifying xenobiotics [28].
Subsequent investigation of fractions 2 and 3
in combination with gemcitabine (5.0 μM) in
pancreatic cancer cells indicated that at low
concentration of fractions (2.5 μg /ml), the dif-
ference in % cellular inhibition was almost simi-
lar whilst at higher concentrations (5.0 and
10.0 μg/ml), fraction 3 was more potent. This
may suggest the presence of certain other
molecules that at higher concentrations may
enable the increase in cytotoxic potency. In
hepatocellular cell lines (HEP G2), the combina-
Comparison between fractionated and un-fractionated bromelain
4318 Am J Transl Res 2021;13(5):4309-4321
tion of fraction 3 or un-fractionated bromelain
(C) with doxorubicin (2.5 mM) indicated that un-
fractionated bromelain (C) was slightly more
cytotoxic at 2.0-5.0 μg/ml, however at 10.0 μg/
ml, fraction 3 was signicantly more potent,
again indicating the presence of other mole-
cules at trace amount that at certain concen-
trations provide this enhancement of cytotoxic-
ity. A similar scenario existed when NAC (7.0
mM) was included in the treatment. NAC is
known to enhance the cytotoxicity of bromelain
in our previous studies in both hepatic and pan-
creatic cancer cell lines [10].
In the case of HEP 3B, the addition of doxorubi-
cin (2.5 μM) indicated that fraction 3 was more
potent at the three concentrations investigated
when compared to un-fractionated bromelain
(C). However, the addition of NAC (7.0 mM) to
the mixture indicated that fraction 3 was only
marginally more potent. This indicates that
fraction 3 is relatively less enhanced by NAC as
compared to un-fractionated bromelain that is
made up of composite enzymes.
The variation of cytotoxic response of these
hepatocellular cell lines to un-fractionated bro-
melain (C) and fraction 3 may be partially attrib-
uted to their individual cellular features [29].
On the other hand, the ratio of the agents that
have been used either as dual or triple combi-
nations may play a role in their efciency. The
combination of these agents is only efcient at
certain molar ratio as it has been shown previ-
ou sly [10].
A comparison of proteolytic activity to cytotoxic-
ity indicates that highly active basic isoforms of
bromelain present in F3, with high proteolytic
activity, shows the highest cytotoxicity as indi-
cated by the IC50 values in all the cell lines
tive order of proteolytic activity of the assayed
fractions. Finally, although F1 that is composed
of acidic bromelains has relatively very high
proteolytic activity (second most activity
amongst the fractions), it seems to show low
cytotoxicity. This again indicates that there is
not usually a relationship between proteolytic
activity and cytotoxicity (Table 5).
The basic bromelain isoforms seem to perform
very well as cytotoxic agents perhaps owing to
their basic optimum proteolytic activity at pH
7.0, whilst ananain and acidic isoforms may
also be dependent on acidic environment for
maximal proteolytic activity and hence their
weaker cytotoxicity.
There are various reports indicating the anti-
thrombotic properties of bromelain, as a total
extract, and its potential as an anticoagulant
for treatment of blood coagulative disorders
[17, 18]. However, earlier evaluation on the pro-
teolytic activity of fraction 3 showed that its
activity is only slightly higher (by 5%) than un-
fractionated bromelain, hence we did not
expect to see a large variation in any of the
coagulative parameters that we measured. The
PT (prothrombin time) indicates that F3 was
similar to controls and to un-fractionated bro-
melain (5.0-10.0 μg/ml). The APTT showed no
change with addition of 10.0 μg/ml of brome-
lain. With the addition of 10.0 μg/ml of F3, the
difference was being very small. Comparing the
INR values indicated that the addition of bro-
melain (5.0-10.0 μg/ml) did not alter the values
compared to a control which is similar to F3.
Finally, the F-10 (%) also showed no difference
as compared to un-fractionated bromelain with
the addition of 5.0-10.0 μg/ml bromelain, how-
ever, fraction 3 showed a very small drop (108),
Table 5. It shows the order of proteolytic activities arranged from
highest to lowest in relation to their cytotoxicity, along with their
main components
Order of proteolytic
activity
Order of
Cytotoxicity Main components
Fraction 3 Fraction 3 Highly active basic bromelain isoforms
Fraction 1 Fraction 2 Poorly active basic bromelain isoforms
Fraction 4 Fraction 4 Ananain isoforms
Fraction 2 Fraction 1 Acidic bromelain isoforms
The basic bromelain isoforms seem to have a higher cytotoxicity in all the cell lines
inestigated.
investigated. The second most
cytotoxic component is F2
composed of poorly active
basic isoforms of bromelain
but with low proteolytic activi-
ty (<40% compared to con-
trol), hence indicating that
proteolytic activity does not
really corresponds with degree
of cytotoxicity. The third most
cytotoxic component is F4
containing ananain and it cor-
responds with the compara-
Comparison between fractionated and un-fractionated bromelain
4319 Am J Transl Res 2021;13(5):4309-4321
again indicating that there is no signicant anti-
coagulative action.
However, the addition of NAC showed signi-
cant difference, an increase in PT, APTT and
INR values and a substantial drop in F-10 (%)
values that is indicative of positive anticoagula-
tive action. Likewise, the addition of NAC to un-
fractionated bromelain in blood caused a sig-
nicant increase in PT, APTT and INR values
and a drop in F-10 (%) that is again indicative of
anticoagulative action. Hence, NAC has sub-
stantial anticoagulative properties. This antico-
agulative action is mainly due to its antioxidant
properties that prevents the aggregation of
platelets [30]. Further, it is able to disrupt the
disulde linkage found in brin brils in blood
clot owing to its antioxidant properties [31].
Similarly, many other antioxidants increase the
bleeding time [32]. Hence, from the present
antithrombotic parameters measured, F3 does
not have any signicant anti-coagulant proper-
ties that may affect patients who are on warfa-
rin or other anticoagulant therapy. However,
one has to be careful since these measure-
ments represent the intrinsic coagulation path-
way. Studies with intraperitoneal delivery of
bromelain in mice have indicated bleeding in
the peritoneal cavity, owing to the abrasive
action of bromelain on the peritoneal surface
cells where the extrinsic pathway comes into
action that unies with the common pathway of
coagulation (Figure 6). The common pathway
where factor X, prothrombin II and plasminogen
come into action to generate brin bres may
be affected by un-fractionated bromelain and
fraction 3, and this needs further investigation.
Although bromelain affects the kinin system
(bradykinin, prekallikrin, etc) of the intrinsic
coagulation parameters will be affected and
hence increase bleeding time (Figure 6).
Therefore, the current study indicates that bro-
melain fractions (F1-F4) have different proteo-
lytic activities that may affect their perfor-
mance as an anti-cancer agent in the various
cell lines investigated. However, their proteo-
lytic potency does not appear to correlate well
with their degree of anticancer action suggest-
ing that these chromatically separated frac-
tions do not only carry proteolytic enzymes but
may have other components that affect the
viability of tumour cells. Fraction 3 performs
slightly better in combinations with gemcitabine
when compared to fraction 2 in ASPC-1 cells
(pancreatic cancer). In combination with doxo-
rubicin, fraction 3 performs slightly better than
standard (un-fractionated) bromelain. On the
whole, as a single agent, bromelain (5.0-10.0
μg/ml) or fraction 3 (10.0 μg/ml), does not
show any antithrombotic properties, however,
in combination with NAC, they exhibit anti-
thrombotic properties. Hence, for therapeutic
purposes, a combination regime of NAC with
either fraction 3 or bromelain may affect indi-
viduals who are on antithrombotic medication.
The reason why only fraction 3, composed of
basic bromelain isoforms, seems to mimic the
action of un-fractionated bromelain may be
attributed to the fact that basic bromelain iso-
forms represent the major components of the
bromelain total extract. The assessment of the
dose of the bromelain complex mixture to be
used for therapeutic purposes thus remains an
important factor to be elucidated. Using a well
characterized fraction, e.g. F3, will thus be a
nice alternative and perspective. Actually, it
Figure 6. Coagulation cascade in the common pathway. Bromelain and frac-
tion 3 may degrade factor X, prothrombin II and plasminogen and hence
delay coagulation.
coagulation pathway that con-
nects with the common path-
way, our investigation does
not seem to suggest that
the system is affected at the
concentrations of bromelain
and fraction 3 used in this
study. Further it also does
not affect the common path-
way. However, the presence
of NAC is crucial in affecting
the coagulation parameters
that we have monitored cur-
rently and hence when it is
combined with bromelain, the
Comparison between fractionated and un-fractionated bromelain
4320 Am J Transl Res 2021;13(5):4309-4321
remains a challenge to understand at the
molecular level the numerous effects of stem
bromelain proteases, highlighting the need for
more detailed studies combining both, cytotox-
icity and proteomics investigations to identify
potential specic targets affected by these
promising proteases.
Acknowledgements
We would like to thank Mrs Laetitia Bolle for
here technical assistance during the prepara-
tion of bromelain fractions. This research is
partly funded by Mucpharm Pty Ltd, Australia.
Professor David L Morris is the co-inventor and
assignee of the combination of Bromelain and
Acetylcysteine in cancer patent licence and
director of the spin-off company, Mucpharm Pty
Ltd. Dr Javed Akhter, Dr Krishna Pillai and Dr
Ahmed H Mekkawy are employees of Mucpharm
Pty Ltd.
Disclosure of conict of interest
None.
Address correspondence to: David L Morris, Depart-
ment of Surgery, St George Hospital and University
of New Sout h Wale s, Sy dn ey, Aust ralia. E- ma il : david.
morris@unsw.edu.au
References
[1] Maurer HR. Bromelain: biochemistry, pharma-
cology and medical use. Cell Mol Life Sci 2001;
58: 1234-1245.
[2] Pillai K, Akhter J, Chua TC and Morris DL. Anti-
cancer property of bromelain with therapeutic
potential in malignant peritoneal mesothelio-
ma. Cancer Invest 2013; 31: 241-250.
[3] Singer AJ, Taira BR, Anderson R, McClain SA
and Rosenberg L. The effects of rapid enzy-
matic debridement of deep partial-thickness
burns with Debrase® on wound reepithelial-
ization in swine. J Burn Care Res 2010; 31:
795-802.
[4] Chobotova K, Vernallis AB and Majid FA. Bro-
melain’s activity and potential as an anti-can-
cer agent: current evidence and perspectives.
Cancer Lett 2010; 290: 148-156.
[5] Amini A, Ehteda A, Masoumi Moghaddam S,
Akhter J, Pillai K and Morris DL. Cytotoxic ef-
fects of bromelain in human gastrointestinal
carcinoma cell lines (MKN45, KATO-III, HT29-
5F12, and HT29-5M21). Onco Targets Ther
2013; 6: 403-409.
[6] Pillai K, Ehteda A, Akhter J, Chua TC and Morris
DL. Anticancer effect of bromelain with and
without cisplatin or 5-FU on malignant perito-
neal mesothelioma cells. Anticancer Drugs
2014; 25: 150-160.
[7] Pillai K, Akhter J, Chua TC and Morris DL. A for-
mulation for in situ lysis of mucin secreted in
pseudomyxoma peritonei. Int J Cancer 2014;
134: 478-486.
[8] Pillai K, Akhter J and Morris DL. Assessment of
a novel mucolytic solution for dissolving mucus
in pseudomyxoma peritonei: an ex vivo and in
vitro study. Pleura Peritoneum 2017; 2: 111-
117.
[9] Valle SJ, Akhter J, Mekkawy AH, Lodh S, Pillai
K, Badar S, Glenn D, Power M, Liauw W and
Morris DL. A novel treatment of bromelain and
acetylcysteine (BromAc) in patients with perito-
neal mucinous tumours: a phase I rst in man
study. Eur J Surg Oncol 2021; 47: 115-122.
[10] Pillai K, Mekkawy AH, Akhter J, Badar S, Dong
L, Liu AI and Morris DL. Enhancing the potency
of chemotherapeutic agents by combination
with bromelain and N-acetylcysteine - an in vi-
tro study with pancreatic and hepatic cancer
cells. Am J Transl Res 2020; 12: 7404-7419.
[11] Murachi T and Neurath H. Fractionation and
specicity studies on stem bromelain. J Biol
Chem 1960; 235: 99-107.
[12] Pavan R, Jain S, Shraddha and Kumar A. Prop-
erties and therapeutic application of brome-
lain: a review. Biotechnol Res Int 2012; 2012:
976203.
[13] Taussig SJ and Batkin S. Bromelain, the en-
zyme complex of pineapple (Ananas comosus)
and its clinical application. An update. J Ethno-
pharmacol 1988; 22: 191-203.
[14] Felton GE. Fibrinolytic and antithrombotic ac-
tion of bromelain may eliminate thrombosis in
heart patients. Med Hypotheses 1980; 6:
1123-1133.
[15] Kaur H, Corscadden K, Lott C, Elbatarny HS
and Othman M. Bromelain has paradoxical ef-
fects on blood coagulability: a study using
thromboelastography. Blood Coagul Fibrinoly-
sis 2016; 27: 745-752.
[16] Kenawy HI, Boral I and Bevington A. Comple-
ment-coagulation cross-talk: a potential medi-
ator of the physiological activation of comple-
ment by low pH. Front Immunol 2015; 6: 215.
[17] Livio M, Bertoni M and DeGaetano G. Effect of
bromelain on brinogen level, prothrombin
complex factors and platelet aggregation in
the rat: a preliminary report. Drugs Exp Clin
Res 1978; 4: 49.
[18] Lotz-Winter H. On the pharmacology of brome-
lain: an update with special regard to animal
studies on dose-dependent effects. Planta
Med 1990; 56: 249-253.
[19] Matagne A, Bolle L, El Mahyaoui R, Baeyens-
Volant D and Azarkan M. The proteolytic sys-
tem of pineapple stems revisited: purication
Comparison between fractionated and un-fractionated bromelain
4321 Am J Transl Res 2021;13(5):4309-4321
and characterization of multiple catalytically
active forms. Phytochemistry 2017; 138: 29-
51.
[20] Azarkan M, Wintjens R, Smolders N, Nijs M
and Looze Y. S-pegylthiopapain, a versatile in-
termediate for the preparation of the fully ac-
tive form of the cysteine proteinase archetype.
J Chromatogr A 1996; 724: 185-192.
[21] Coelho DF, Saturnino TP, Fernandes FF, Maz-
zola PG, Silveira E and Tambourgi EB. Azoca-
sein substrate for determination of proteolytic
activity: reexamining a traditional method us-
ing bromelain samples. Biomed Res Int 2016;
2016: 8409183.
[22] Vichai V and Kirtikara K. Sulforhodamine B
colorimetric assay for cytotoxicity screening.
Nat Protoc 2006; 1: 1112-1116.
[23] Azarkan M, Feller G, Vandenameele J, Herman
R, El Mahyaoui R, Sauvage E, Vanden Broeck
A, Matagne A, Charlier P and Kerff F. Biochem-
ical and structural characterization of a man-
nose binding jacalin-related lectin with two-
sugar binding sites from pineapple (Ananas
comosus) stem. Sci Rep 2018; 8: 11508.
[24] Wang X and Li S. Protein mislocalization:
mechanisms, functions and clinical applica-
tions in cancer. Biochim Biophys Acta 2014;
1846: 13-25.
[25] Kufe DW. Mucins in cancer: function, progno-
sis and therapy. Nat Rev Cancer 2009; 9: 874-
885.
[26] Oosterkamp HM, Scheiner L, Stefanova MC,
Lloyd KO and Finstad CL. Comparison of MUC-
1 mucin expression in epithelial and non-epi-
thelial cancer cell lines and demonstration of a
new short variant form (MUC-1/Z). Int J Cancer
1997; 72: 87-94.
[27] Matte I, Lane D, Boivin M, Rancourt C and Pi-
che A. MUC16 mucin (CA125) attenuates
TRAIL-induced apoptosis by decreasing TRAIL
receptor R2 expression and increasing c-FLIP
expression. BMC Cancer 2014; 14: 234.
[28] Gu X and Manautou JE. Molecular mecha-
nisms underlying chemical liver injury. Expert
Rev Mol Med 2012; 14: e4.
[29] Qiu GH, Xie X, Xu F, Shi X, Wang Y and Deng L.
Distinctive pharmacological differences be-
tween liver cancer cell lines HEP G2 and Hep
3B. Cytotechnology 2015; 67: 1-12.
[30] Gibson KR, Winterburn TJ, Barrett F, Sharma S,
MacRury SM and Megson IL. Therapeutic po-
tential of N-acetylcysteine as an antiplatelet
agent in patients with type-2 diabetes. Cardio-
vasc Diabetol 2011; 10: 43.
[31] Martinez de Lizarrondo S, Gakuba C, Herbig
BA, Repesse Y, Ali C, Denis CV, Lenting PJ,
Touze E, Diamond SL, Vivien D and Gauberti
M. Potent thrombolytic effect of N-acetylcyste-
ine on arterial thrombi. Circulation 2017; 136:
646-660.
[32] Violi F, Pignatelli P and Basili S. Nutrition, sup-
plements, and vitamins in platelet function
and bleeding. Circulation 2010; 121: 1033-
1044.
