Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)

ARTICLE
Curcumin and folic acid abrogated methotrexate induced
vascular endothelial dysfunction
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université Laval Bibliotheque on 11/21/15

Himanshu Sankrityayan and Anuradha S. Majumdar

Abstract: Methotrexate, an antifolate drug widely used in rheumatoid arthritis, psoriasis, and cancer, is known to cause vascular
endothelial dysfunction by causing hyperhomocysteinemia, direct injury to endothelium or by increasing the oxidative stress
(raising levels of 7,8-dihydrobiopterin). Curcumin is a naturally occurring polyphenol with strong antioxidant and anti-
inflammatory action and therapeutic spectra similar to that of methotrexate. This study was performed to evaluate the effects
of curcumin on methotrexate induced vascular endothelial dysfunction and also compare its effect with that produced by
folic acid (0.072 ␮g·g−1·day−1, p.o., 2 weeks) per se and in combination. Male Wistar rats were exposed to methotrexate
(0.35 mg·kg−1·day−1, i.p.) for 2 weeks to induce endothelial dysfunction. Methotrexate exposure led to shedding of endothelium,
decreased vascular reactivity, increased oxidative stress, decreased serum nitrite levels, and increase in aortic collagen deposi-
tion. Curcumin (200 mg·kg−1·day−1 and 400 mg·kg−1·day−1, p.o.) for 4 weeks prevented the increase in oxidative stress, decrease
in serum nitrite, aortic collagen deposition, and also vascular reactivity. The effects were comparable with those produced by
folic acid therapy. The study shows that curcumin, when concomitantly administered with methotrexate, abrogated its vascular
side effects by preventing an increase in oxidative stress and abating any reduction in physiological nitric oxide levels.

Key words: methotrexate, curcumin, vascular endothelial dysfunction, folic acid, nitric oxide, oxidative stress.

Résumé : Le méthotrexate, un médicament antifolique largement utilisé contre l’arthrite rhumatoïde, le psoriasis et le cancer,
For personal use only.

est connu pour causer une dysfonction endothéliale vasculaire en provoquant une hyperhomocystéinémie, un dommage direct
à l’endothélium ou en accroissant le stress oxydant (augmentant les niveaux de 7,8-dihydrobioptérine). La curcumine est un
polyphénol naturel qui exerce de fortes actions antioxydante et anti-inflammatoire, et qui possède un spectre thérapeutique
similaire à celui du méthotrexate. Cette étude a été réalisée afin d’évaluer les effets de la curcumine sur la dysfonction
endothéliale vasculaire induite par le méthotrexate, et de comparer son effet à celui de l’acide folique (0,072 ␮g·g−1·jour−1, PO,
pendant 2 semaines), seul ou en combinaison. Des rats Wistar mâles ont été exposés au méthotrexate (0,35 mg·kg−1·jour−1, i.p.)
pendant 2 semaines afin d’induire une dysfonction endothéliale. L’exposition au méthotrexate provoquait une desquamation
endothéliale, une diminution de la réactivité vasculaire, une augmentation du stress oxydant, une diminution des niveaux de
nitrites sériques et une augmentation du dépôt de collagène aortique. La curcumine (200 mg·kg−1·jour−1 et 400 mg·kg−1·jour−1,
PO) pendant 4 semaines prévenait l’augmentation du stress oxydant, la diminution du nitrite sérique, le dépôt de collagène
aortique et la réactivité vasculaire. Les effets étaient comparables à ceux produits par une thérapie à l’acide folique. Cette étude
montre que la curcumine, lorsqu’administrée de manière concomitante au méthotrexate, atténuait les effets secondaires
vasculaires en prévenant l’augmentation du stress oxydant et en abolissant toute réduction des niveaux physiologiques d’oxyde
nitrique. [Traduit par la Rédaction]

Mots-clés : méthotrexate, curcumine, dysfonction endothéliale vasculaire, acide folique, oxyde nitrique, stress oxydant.

Introduction Methotrexate, which is one of the most commonly used disease

Vascular endothelial dysfunction (VED) is the state of imbalance
between the vasorelaxation and vasoconstriction. It particularly
involves impairment in the vasodilatory action of the endothe-
lium. Reduced bioavailability of vasodilatory factors (especially
nitric oxide (NO)) is considered to be the main characteristic of
endothelial dysfunction (Davignon and Ganz 2004; Hadi et al.
2005; Sena et al. 2013). It has been established that an interaction
between reactive oxygen species (ROS) and NO leads to the re-
duced bioavailability of NO (Hsieh et al. 2014; Kojda and Harrison
1999). VED is known to be linked with various forms of cardiovas-
cular diseases including hypertension, coronary artery disease,
chronic heart failure, peripheral vascular disease, diabetes, and
chronic kidney failure (Rajendran et al. 2013). In fact, many stud-
ies are pointing to VED as a potential surrogate marker for the
cardiovascular diseases (Anderson 2006; Cohn et al. 2004; Tardif
et al. 2006).
modifying antirheumatic and anticancer drugs, may cause endo-
thelial dysfunction either by direct injury to endothelium (Merkle
et al. 2000; Soultati et al. 2012), increase in homocysteine levels
(Hoekstra et al. 2005; Svardal et al. 1988), or by an increase in the
levels of 7,8-dihydrobiopterin (BH2) (Crabtree et al. 2011).
Methotrexate was reported to cause direct endothelial injury at
higher doses (15 mg·m−2 on day 1 and 10 mg·m−2 on days 3, 6, and 11)
in 6- to 8-week-old female mice (Zeng et al. 2008). In another study,
low dose methotrexate (0.350 mg·kg−1·day−1, i.p. for 10 days) was
reported to increase the level of homocysteine, which itself is
considered as an independent marker of cardiovascular diseases
and VED (Svardal et al. 1988). Although methotrexate may cause
VED by different mechanisms, the most important is develop-
ment of oxidative stress and reduction of NO. NO is known to play
a key role in maintaining the endothelial function. Endothelial

Received 7 April 2015. Accepted 18 June 2015.

H. Sankrityayan and A.S. Majumdar. Bombay College of Pharmacy, Department of Pharmacology, University of Mumbai, 400098 Mumbai, India.
Corresponding author: Anuradha S. Majumdar (e-mail: anuradha.majumdar@bcp.edu.in).

Can. J. Physiol. Pharmacol. 94: 1–8 (2016) dx.doi.org/10.1139/cjpp-2015-0156 Published at www.nrcresearchpress.com/cjpp on xx xxx xxxx.
 Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
2 Can. J. Physiol. Pharmacol. Vol. 94, 2016
nitric oxide synthase (eNOS) is responsible for the regulation of
NO levels in the endothelium. Methotrexate treatment hinders
with the bioavailability of BH4 (tetrahydrobiopterin) by oxidizing
it into BH2 and also inhibiting the enzyme DHFR (dihydrofolate-
reductase), which is responsible for the regeneration of BH4 from
BH2 as shown in a study by Crabtree et al. 2011. They used wild type,
BH4 deficient (hph1) mice with 3 low doses of methotrexate
(2 mg·kg−1, i.p.) each at an interval of 48 h. It caused an attenuated
eNOS activity. In another study, a single dose of methotrexate
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université Laval Bibliotheque on 11/21/15
(5 mg·kg−1, i.p.) in male Wistar rats reduced the BH4 to BH2
(Noguchi et al. 2011). Thus, methotrexate ultimately elevates the level
of BH2 and decreases the level of BH4 in the aorta. Endothelial BH4
bioavailability is vital for maintaining the balance between NO and
superoxide production. BH4 is responsible for eNOS coupling and
production of NO whereas BH2 causes eNOS uncoupling (Crabtree
et al. 2011; Noguchi et al. 2011). Earlier studies had shown that exter-
nal supplementation of BH4 improves endothelial dysfunction
(Fukuda et al. 2002; Mittermayer et al. 2005; Stroes et al. 1997).
Curcumin is an active polyphenolic constituent from Curcuma
longa having a wide spectra of therapeutic effects in diseases in-
cluding cardiovascular disorders, neurodegenerative disorders,
diabetes, arthritis, psoriasis, cancer, etc. (Aggarwal and Harikumar
2009). Methotrexate is supposed to produce most of its side effects
due to development of oxidative stress. So, curcumin being a
strong antioxidant (Boonla et al. 2014; Fleenor et al. 2013) may be
used to curb its side effects. According to a recent study, curcumin
not only reduced the side effects of methotrexate but also accen-
tuated its therapeutic effect (Banji et al. 2011). Curcumin is also
known to prevent endothelial dysfunction due to hypercholes-
For personal use only.
terolemia (Sudjarwo et al. 2011), hyperhomocysteinemia, and hy-
perlipidaemia (Kapoor et al. 2008). The aim of this study was to
explore the potential of curcumin in abrogating the detrimental
effect of methotrexate on vasculature. Further, we aimed to ex-
plore whether it acts in concert with folic acid in reducing the
endothelial dysfunction induced by methotrexate.
Materials and methods
Chemicals
All the chemicals used in the study were purchased from Sigma–
Aldrich (St. Louis, Missouri, USA) except methotrexate, curcumin,
and folic acid. Methotrexate was a gift sample (Mac Chem Pvt. Ltd.,
Mumbai, India). It was dissolved in freshly prepared phosphate buff-
ered saline (pH 7.4). Curcumin was obtained as a gift sample (K. Patel
Phytoextractives Pvt. Ltd., Mumbai, India) and administered as a sus-
pension in 0.5% sodium carboxymethylcellulose. Folic acid was also a
gift sample (Martin and Brown Biosciences, Baddi, India) and was
dissolved in water.
Animals
42 male Wistar rats were purchased from an animal house
(Bombay Veterinary College, Parel, Mumbai, India). The experi-
mental room was maintained under standard conditions of tem-
perature (25 ± 2 °C) and relative humidity (55% ± 10%). The
animals were subjected to a 12 h light – 12 h dark cycle. They had
access to rat chow and fresh drinking water ad libitum through-
out the study. The housing conditions and experimental protocols
(CPCSEA- BCP/2011-02/08 dated 20 August 2011) were in accordance
with the guidelines of Institutional Animal Ethics Committee
(IAEC) of Bombay College of Pharmacy, Mumbai, India.
Study design
42 male Wistar rats, 10–12 weeks old weighing around 200–
220 g were used. They were divided into 7 groups (Groups I–VII),
each having 6 animals (Table 1).
Group I rats (positive control group) were administered metho-
trexate (0.35 mg·kg−1·day−1, i.p.) for 2 weeks and Group II rats
(vehicle control group) were given phosphate buffered saline (pH
7.4) for 2 weeks from the 3rd week onwards. The study period
was 4 weeks and had 5 treatment groups (III, IV, V, VI, and VII).
Rats from Groups III, IV, and V were administered curcumin
(100 mg·kg−1·day−1, 200 mg·kg−1·day−1, and 400 mg·kg−1·day−1, re-
spectively; p.o.) along with methotrexate (0.35 mg·kg−1·day−1, i.p.).
The dose of curcumin was based on a similar study of endothelial
cell protection (Sudjarwo et al. 2011). Methotrexate exposure was
stopped after completion of 2 weeks of the 4-week-long study
period while curcumin was continued till the end of the study
period. Group VI rats were administered with methotrexate
(0.35 mg·kg−1·day−1, i.p.) for 2 weeks and then with folic acid
(0.072 ␮g·g−1·day−1, p.o.) for the next 2 weeks. The dose of folic acid
was based on previous study and is equivalent to a human dose of
5 mg·70 kg−1·day−1 (Clarke et al. 2006). Group VII rats were admin-
istered with methotrexate (0.35 mg·kg−1·day−1, i.p.) for 2 weeks
and then with folic acid (0.072 ␮g·g−1·day−1, p.o.) for the next
2 weeks while receiving curcumin (200 mg·kg−1·day−1) for the en-
tire 4 weeks of the study period (Fig. 1).
Curcumin was administered half an hour prior to methotrexate
administration. Rats from all the groups were sacrificed at the end of
the 4th week. Serum was used for the estimation of oxidative stress
parameters including SOD (superoxide dismutase) activity, GSH (re-
duced glutathione), and MDA (malondialdehyde) levels. Serum was
also used for the estimation of NO levels. Thoracic aorta was used to
assess the vascular reactivity and collagen deposition apart from
histopathological examinations for morphological changes.
Sample collection
The rats were anaesthetized using ketamine:xylazine (80:20 mg·
(kg body mass)−1 i.p.) after treatment and the blood was withdrawn
through the abdominal aorta. It was allowed to clot and the serum
was separated by centrifugation (3340g for 15 min). The serum was
preserved at –20 °C until analysis. The aorta was immediately
removed and placed in modified Krebs–Henseleit buffer gassed
with 95% O2 and 5% CO2. A segment of aorta was fixed in 10%
formalin for histopathological examination. The aorta was also
separated to measure the collagen content.
Vascular reactivity
The thoracic aorta was cleaned of the adhering perivascular fats
and remaining connective tissues. The aorta was then cut into small
rings of 3–5 mm. The rings were mounted in between 2 stirrups and
placed in a 20 mL organ bath filled with modified Krebs–Henseleit
buffer. The composition of the buffer was as follows (in mmol·L−1):
NaCl, 118; KCl, 4.7; MgSO4, 1.2; KH2PO4, 1.2; NaHCO3, 25; CaCl2, 2;
glucose, 10. The aortic rings were kept at a tension of 1 g. The rings
were allowed to equilibrate for a period of 60 min. The buffer was
replaced every 15–20 min. The rings were stretched to its maximum
by exposing it to phenylephrine thrice (0.1 ␮mol). After, they were
washed thoroughly and allowed to reach the initial tension and
again left for equilibration (Lan et al. 2011).
Vascular reactivity was assessed by recording the relaxation pro-
ducedbyacetylcholineaddedcumulatively(10−8–10−4 mol·L−1)inphen-
ylephrine (0.3 ␮mol·L−1) pre-contracted rings. A submaximal dose of
phenylephrine (0.3 ␮mol·L−1) was added to the organ bath and the
rings were allowed to contract until it reached the plateau. It was
followed by the cumulative addition of acetylcholine. The change in
the tension was recorded by Powerlab data acquisition system using
the software Labchart 7 Pro (AD Instruments, Australia).
Aortic collagen
Aortic collagen is based on the determination of hydroxyproline
levels in the sample (Woessner 1983). The aorta was hydrolysed by
hydrochloric acid, followed by addition of methyl indicator (0.002%),
and finally titrated with NaOH (2.5 mol·L−1). This solution was used
for the estimation of hydroxyproline. The sample solution was
mixed with 0.01 mol·L−1 CuSO4, 2.5 mol·L−1 NaOH, and 6% H2O2. The
above reagents were mixed and shaken occasionally for about 5 min
and was placed in a water bath at 80 °C for 5 min. Then, the mixture
Published by NRC Research Press
 Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Sankrityayan and Majumdar 3

Table 1. Grouping of animals for the study with treatment description.

Group No. of
No. Group Treatment description rats (n)
I Positive control Rats were administered MTX in PBS (0.35 mg·kg−1·day−1 6
for 2 weeks), i.p.
II Vehicle control Rats were administered PBS, i.p. for the last 2 weeks 6
III Curcumin (100 mg·kg−1, p.o.) Rats were administered curcumin suspension (100 mg·kg−1·day−1) 6
for 4 weeks, p.o. and MTX for 2 weeks
IV Curcumin (200 mg·kg−1, p.o.) Rats were administered curcumin suspension (200 mg·kg−1·day−1) 6
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université Laval Bibliotheque on 11/21/15

for 4 weeks p.o. and MTX for 2 weeks

V Curcumin (400 mg·kg−1, p.o.) Rats were administered curcumin suspension (400 mg·kg−1·day−1) 6
for 4 weeks, p.o. and MTX for 2 weeks
VI Folic acid Rats were administered MTX for 2 weeks followed by folic acid 6
(0.072 ␮g·g−1·day−1, p.o.) administration for another 2 weeks
VII Folic acid + curcumin Rats were administered curcumin suspension (200 mg·kg−1·day−1 p.o.) 6
for 4 weeks and MTX for 2 weeks followed by folic acid, p.o. for
another 2 weeks
Note: MTX, methotrexate; PBS, phosphate-buffered saline.

Fig. 1. Pictorial representation of the study design followed during the experimental procedure. (A color version of this figure is available
through the journal Web site at http://www.nrcresearchpress.com/doi/10.1139/cjpp-2015-0156.)

Week 0 Week 2
N=6 GROUPS

II
For personal use only.

III, IV, V

VII

VII

Week 4

Key:

: Methotrexate (0.35 mg.kg-1, i. p)

: Phosphate buffered saline (0.35 mg.kg-1, i. p)

: Curcumin (100 mg.kg-1, 200 mg.kg-1 and 400 mg.kg-1, p. o.)

: Folic acid (0.072 µg.g-1, p. o.)

was cooled followed by addition of 1.5 mol·L−1 H2SO4, 5% DMBA (p-
dimethylaminobenzaldehyde). The tubes containing the mixture
were again heated in a water bath at 70 °C for a period of about
15 min followed by cooling under tap water. Absorbance of the solu-
tion was taken at 540 nm. The above method gave us the concentra-
tion of hydroxyproline in the sample. Collagen concentration was
obtained by using the following equation (Neuman and Logan 1950):
Collagen content (%)
Micrograms hydroxyproline in 1 mL of hydrolysate
⫽
Micrograms tissue represented in 1 mL of hydrolysate
× 7.46 × 100
Serum biochemical assays
Serum SOD activity
For this assay, carbonate buffer (pH 10.7), EDTA solution
(0.1 mmol·L−1), and serum sample were mixed and the reaction
was initiated by the addition of 0.5 mL of epinephrine solution. In
this assay, epinephrine is oxidized to form adrenochrome and the
absorbance is taken at 480 nm. SOD itself gets oxidized by react-
ing with O2− formed during epinephrine oxidation and converts it
into hydrogen peroxide and thus inhibits the formation of adre-
nochrome. The ability of SOD to inhibit the autoxidation of epi-
nephrine at pH 10.2 has been used as the basis of a convenient and
sensitive assay for this enzyme (Misra and Fridovich 1972). The

Published by NRC Research Press

 Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
4 Can. J. Physiol. Pharmacol. Vol. 94, 2016
results were expressed in the form of SOD unit activity per milli-
gram of protein. One unit of SOD activity is equal to 50% inhibi-
tion of adrenochrome formation.
Serum MDA
MDA forms a condensation product with 2-thiobarbituric acid
in the ratio of 1:2. The absorbance of this complex is measured
usually by spectrophotometry (Lovri et al. 2008). Two millilitres of
the clear supernatant of sample was added to 2 mL of freshly
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université Laval Bibliotheque on 11/21/15
prepared thiobarbituric acid (TBA). The resulting mixture was in-
cubated at 95 °C for 60 min. The mixture was then cooled under
tap water and centrifuged at 1520g for about 10 min. Absorbance
of the supernatant was taken at 530 nm. The concentration was
expressed as nanomoles of MDA per gram of protein.
Serum GSH
GSH was measured by reaction with 5, 5=-dithiobis (2-nitrobenzoic
acid) (DTNB) to form a compound that absorbs at 412 nm (Ellman
1959). The DTNB reacts with the thiol compounds leading to the
formation of p-nitrophenol anion, which is highly colored and can
be useful in the measurement of the thiol concentration in the
given sample. The concentration of the GSH was obtained by in-
terpolation of the standard plot of GSH concentration with absor-
bance. The concentration was expressed as microgram of GSH per
gram of protein.
Serum nitrite
It was performed by reduction of the nitrate to nitrite by a metal
with subsequent determination of nitrite by diazotization of
sulfanilamide and its coupling with n-(1-naphthyl) ethylenediamine
For personal use only.
(Cortas and Wakid 1990). The serum sample was deproteinized
with the addition of trichloroacetic acid. Serum supernatant
(100 ␮L) was taken and suitably diluted to 1 mL with distilled
water. This was followed by the addition of copper coated gran-
ules to the solution which was kept for incubation at 37 °C for a
period of 90 min. The solution was filtered. Griess reagent (100 ␮L)
was mixed with 300 ␮L of the sample and 2.6 mL of water. The
mixture was incubated for 30 min in the dark. Absorbance was
taken at 540 nm.
Histopathological examination of thoracic aorta
Tissues specimens were collected from rats of each treatment
group. After collection, the tissues were immediately preserved in
the 10% neutral buffered formalin, processed by routine method
for histological observation. Processed tissue were sectioned (at
5 ␮m) and taken on the clean glass slides and stained by hematoxylin
and eosin and observed under microscopes at different magnifi-
cations. Transverse sections of aorta were examined to check any
histological microscopic alterations of pathological significance.
Statistical analysis
All values are expressed as mean ± SEM (n = 6). Statistics was
applied using Graph pad PRISM 5 software (Graph pad Software
Inc., San Diego, California, USA). Statistical analysis was per-
formed by using one-way analysis of variance (ANOVA), followed
by Bonferroni’s Multiple Comparison Test (***, p < 0.05 when com-
pared with the vehicle control group (Group II); #, p < 0.05 when
compared with the positive control group (Group I); @, p < 0.05 when
compared with curcumin (400 mg·kg−1) treated group (Group V)).
Results
Effect of different treatments on percentage relaxation of
aorta
Methotrexate exposure (Group I) caused a significant decrease
in the vascular reactivity of the aorta which was shown by reduced
relaxation (36.69% ± 4.87, p < 0.05) by acetylcholine on phenyl-
ephrine pre-contracted rings as compared with the relaxation
showed by the control group (87.14% ± 10.38). When the groups
Fig. 2. Effect of different treatments on ex vivo vascular reactivity.
The change in the percentage relaxation of thoracic aorta to the
increasing dose of acetylcholine in phenylephrine precontracted
aortic rings were recorded. Data are shown as mean ± SEM (n = 6).
***, p < 0.05 when compared with the vehicle control group; #, p < 0.05
when compared with the positive control group; @, p < 0.05 when
compared with curcumin (400 mg·kg−1) group.
were treated with curcumin in doses 200 mg and 400 mg (Group
IV and Group V) the impairment in relaxation was reduced signif-
icantly (59.05% ± 6.13% and 62.38% ± 8.85%, p < 0.05) as compared
with Group I. Folic acid per se (Group VI) and in combination with
curcumin, 200 mg·kg−1 (Group VII) significantly improved the ace-
tylcholine induced aortic relaxation (77.1% ± 7.53% and 89.18% ±
5.53%, p < 0.05) when compared with methotrexate exposed group
(Group I) (Fig. 2).
Effect of different treatments on aortic collagen
Changes in the group exposed to methotrexate (Group I) are
shown in Fig. 3, which clearly indicates an increase in aortic col-
lagen levels (20.88% ± 1.22%, p < 0.05). Curcumin prevented the
aortic collagen deposition. The levels of aortic collagen in Group
IV (17.39% ± 1.58%, p < 0.05) and V (17.04% ± 1.45%, p < 0.05) were
statistically different from Group I. Folic acid administration per
se (Group VI) also prevented increase in the levels of aortic colla-
gen (16.62% ± 0.51%, p < 0.05). A similar reduction in aortic colla-
gen levels (15.98% ± 0.87%, p < 0.05) was observed when folic acid
was administered in combination with 200 mg·kg−1 of curcumin
(Group VII) (Fig. 3).
Effect of different treatments on serum SOD activity
Methotrexate exposure (Group I) led to reduced serum SOD
activity (0.72 UA·mg−1 ± 0.03, p < 0.05). In Group IV and Group V,
it was observed that curcumin improved the serum SOD activ-
ity significantly, the values being (0.57 UA·mg−1 ± 0.03 and
0.59 UA·mg−1 ± 0.01, p < 0.05) respectively for the two groups. Folic
acid (0.072 ␮g·g−1·day−1) per se (Group VI) and in combination with
curcumin, 200 mg·kg−1 (Group VII) showed significant improve-
ment in SOD activity (0.66 UA·mg−1 ± 0.02 and 0.71 UA·mg−1 ± 0.01,
p < 0.05) when compared with the methotrexate exposed and
curcumin, 100 mg·kg−1 (Group III) treated group (Fig. 4A).
Effect of different treatments on serum MDA
Methotrexate exposure (Group I) caused a rise in the levels of
serum MDA concentration (67.70 nmol·g−1 ± 4.66, p < 0.05) as
compared with vehicle control group (19.05 nmol·g−1 ± 3.49) indi-
Published by NRC Research Press
 Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Sankrityayan and Majumdar 5

Fig. 3. Effect of different treatments on aortic collagen. Data are
shown as mean ± SEM (n = 6). ***, p < 0.05 when compared with the
vehicle control group; #, p < 0.05 when compared with the positive
control group; @, p < 0.05 when compared with curcumin (400 mg·kg−1)
group.
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université Laval Bibliotheque on 11/21/15
Fig. 4. Effect of different treatments on oxidative stress. (A) Change
in the activity of serum superoxide dismutase. (B) Change in the
serum malondialdehyde (MDA) concentration (C) Change in the
serum reduced glutathione (GSH) concentration. Data are shown as
mean ± SEM (n = 6). ***, p < 0.05 when compared with the vehicle
control group; #, p < 0.05 when compared with the positive control
group; @, p < 0.05 when compared with curcumin (400 mg·kg−1)
group. gm, gram.

cating increased lipid peroxidation. In Group IV, the levels of

serum MDA (30.82 nmol·g−1 ± 4.19, p < 0.05) were statistically
For personal use only.

different from Group I. Group V also showed a similar reduction

(30.26 nmol·g−1 ± 4.19, p < 0.05) in the levels of serum MDA. Groups
treated with folic acid (0.072 ␮g·g−1·day−1) per se (Group VI) and in
combination with curcumin, 200 mg·kg−1 (Group VII) also showed
significant reduction in the levels of serum MDA (28.20 nmol·g−1 ±
0.94 and 15.17 nmol·g−1 ± 2.08, p < 0.05 respectively) (Fig. 4B).

Effect of different treatments on serum GSH

Methotrexate exposure (Group I) resulted in decreased levels of
serum reduced GSH (129.81 ␮g·g−1 ± 8.16, p < 0.05) as compared
with vehicle control group (253.61 ␮g·g−1 ± 8.16). The groups ad-
ministered with curcumin in doses 200 mg·kg−1 and 400 mg·kg−1
(Group IV and Group V) showed a significant rise in the level of
reduced GSH (188.28 ␮g·g−1 ± 3.98 and 209.67 ␮g·g−1 ± 6.60,
p < 0.05). The change in the group administered with curcumin
100 mg·kg−1 (Group III) was not significant (135.43 ␮g·g−1 ± 4.68) as
compared with Group I. The groups treated with folic acid per se
(Group VI) and along with curcumin, 200 mg·kg−1 (Group VII)
revealed significant rise in the levels of reduced GSH (232.64 ␮g·g−1 ±
7.42 ␮g·g−1 and 247.44 ␮g·g−1 ± 4.92, p < 0.05) (Fig. 4C).

Effect of different treatments on serum nitrite

It was found that the levels of serum nitrite were significantly
lower (108.5 ␮mol·L−1·g−1 ± 6.79, p < 0.05) in the methotrexate
exposed group as compared with the vehicle control group
(227.40 ␮mol·L−1·g−1 ± 6.62). Curcumin prevented the decrease in
the levels of serum nitrite both in Group IV (192.80 ␮mol·L−1·g−1 ±
6.28) and Group V (202.38 ␮mol<·L−1·g−1 ± 6.49, p < 0.05). Folic acid
per se (Group VI) treatment increased the serum concentrations of
nitrite (223.16 ␮mol·L−1·g−1 ± 9.44, p < 0.05). The combination
treated Group VII also showed a significant rise in the levels of
serum nitrite (237.6 ␮mol·l−1·g−1 ± 7.65, p < 0.05) (Fig. 5).

Effect of different treatments on morphology of the

endothelium
Methotrexate exposure (Group I) injured the endothelial layer
as shown in Fig. 6a. It was found that endothelium was shed on
few places. Curcumin administration in doses 200 mg·kg−1 and
400 mg·kg−1 (Group IV and Group V) prevented methotrexate in-

Published by NRC Research Press

 Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
6 Can. J. Physiol. Pharmacol. Vol. 94, 2016
Fig. 5. Effect of different treatments on serum nitrite concentration.
Data are shown as mean ± SEM (n = 6). ***, p < 0.05 when compared
with the vehicle control group; #, p < 0.05 when compared with the
positive control group; @, p < 0.05 when compared with curcumin
(400 mg·kg−1) group. M, mol·L−1.
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université Laval Bibliotheque on 11/21/15
duced damage to the endothelium. Folic acid per se (Group VI) and
in combination with curcumin, 200 mg·kg−1 (Group VII) also pre-
For personal use only.
vented the damage to endothelial layer (Fig. 6).
Discussion
Vascular endothelium is a lining once considered to be just like
a layer of cellophane with no other function than exchange of
water and electrolytes. However, it is now established as one of
the most complex and physiologically important systems (Rajendran
et al. 2013). VED and cardiovascular pathologies are known to be
interconnected. Methotrexate is an antimetabolite widely used in
diseases such as rheumatoid arthritis, psoriasis, and chemother-
apy, but is frequently discontinued because of its several side
effects and induction of VED is one such limiting effect. The role
of curcumin in ameliorating the hepatotoxic (Banji et al. 2011) and
nephrotoxic effects (Morsy et al. 2013) of methotrexate has already
been established in certain studies. It has been reported that cur-
cumin not only alleviated the toxicity of methotrexate but also
showed synergy in its anti-arthritic effect (Banji et al. 2011). How-
ever, curcumin has still not received its status as a therapeutic
agent because of its low relative bioavailability. According to pre-
vious pharmacokinetic studies of curcumin in rodents and hu-
mans, the factors responsible for its low bioavailability include
poor absorption, high rate of metabolism, and rapid systemic
elimination. The serum concentration of curcumin peaks at about
1 h after oral administration and declines within 12 h. The serum
and tissue levels of curcumin, as well as its absorption and elimi-
nation half-lives in rats and humans, can not be compared directly
(Anand et al. 2007).
The objective of our study was to reveal the role of curcumin in
protecting the endothelium from the methotrexate induced VED.
Along with it, we also aimed to investigate the role of combina-
tion of folic acid and curcumin because folic acid is a drug that is,
most of the time, prescribed to be taken with methotrexate to
curb its side effects.
The VED caused by methotrexate can be evaluated by measur-
ing the vascular reactivity of the blood vessels. In the methotrex-
ate exposed group (Group I), the vasorelaxation was impaired to a
great extent when compared with the vehicle controlled group
(Group II). This can be attributed to direct injury imposed by meth-
otrexate on the endothelial layer as shown in the histopathologi-
cal analysis. Due to methotrexate treatment, the endothelial layer
was shed on few places, which is why acetylcholine could not
produce desired relaxation. The mechanism of injury is not clear
but it is supposed to be because of increased generation of ROS
which damages the endothelial layer. Although in the groups
treated with curcumin in doses 200 mg·kg−1 and 400 mg·kg−1
(Group IV and Group V), there was an improvement in the injury
status of endothelium as depicted in the histopathological exam-
ination. The group administered with folic acid (Group VI) also
inhibited the action of methotrexate by replenishing the depleted
folate levels. The group administered with the combination of
folic acid and curcumin in dose 200 mg·kg−1 (Group VII) was also
found equally better in abating the adverse impact of methotrex-
ate on vasculature. NO bioavailability is established as the most
vital factor in the development of endothelial dysfunction. It is
because of its major role in the determination of basal vascular
tone. Methotrexate is supposed to lower the NO bioavailability by
increasing the oxidative stress and uncoupling of eNOS, which
also leads to the generation of ROS. This was pretty evident from
the results of methotrexate group, which showed reduced serum
concentration of nitrite. Here, too, groups administered with cur-
cumin in doses 200 mg·kg−1 and 400 mg·kg−1 (Group IV and Group V)
showed an increase in the concentration of nitrite. Folic acid per
se (Group VI) and in combination with curcumin, 200 mg·kg−1
(Group VII) also showed a significant rise in the levels of nitrite
when compared with the methotrexate exposed group. Metho-
trexate exposure also caused an increase in oxidative stress,
which plays an important role in the development of VED. SOD is
the metallo-enzyme that is responsible for the dismutation of
superoxides into oxygen and hydrogen peroxide, thus protecting
the organs from its ill effects. The reduction in SOD activity also
has an impact on nitrite levels because the superoxide generated
reacts with the nitrite, thus reducing its bioavailability. The group
exposed to methotrexate (Group I) showed a decrease in the activ-
ity of serum SOD resulting in increased oxidative stress when
compared with the vehicle control group (Group II). The groups
treated with curcumin in doses 200 mg·kg−1 and 400 mg·kg−1
(Group IV and Group V) showed a significant rise in the activity of
SOD. The groups administered with folic acid (Group VI) and the
combination of curcumin with folic acid (Group VII) also showed
similar effects.
During oxidative stress, lipid peroxidation is also reported to in-
crease. The oxidative modification of lipids is considered to be a
fundamental process in the development of atherosclerosis (Esterbauer
et al. 1993). The low density lipoproteins after being oxidized become
a factor initiating atherogenesis and other adverse biological pro-
cesses, which includes induction of endothelial dysfunction. Our
study corresponds to the previous studies and we also found that, in
the groups treated with methotrexate, serum levels of MDA had
increased considerably. Curcumin in the doses 200 mg·kg−1 and
400 mg·kg−1 (Group IV and Group V) significantly reduced the levels
of serum MDA. Folic acid (Group VI) also reduced the concentration
of MDA. The combination group of curcumin and folic acid (Group
VII) also showed significantly better results.
The levels of reduced GSH were lowered when the groups were
administered with methotrexate. GSH is a thiol that is involved in
the protection of cells from ROS. Curcumin in the doses 200 mg·kg−1
and 400 mg·kg−1 (Group IV and Group V) was effective in significantly
increasing the levels of GSH as compared with the methotrexate
exposed (Group I) and curcumin, 100 mg·kg−1 (Group III) treated
groups. Folic acid (Group VI) also led to a significant rise in the con-
centration of GSH. Moreover, combination of folic acid and cur-
cumin (Group VII) also raised the levels of GSH significantly.
So, the above results are in complete agreement with the fact
that endothelial dysfunction is often associated with pronounced
oxidative stress that is due to, at least in part, increased superox-
ide (O2−), which increases degradation of NO by reacting with it
and hence reducing its bioavailability. This causes uncoupling of
Published by NRC Research Press
 Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
Sankrityayan and Majumdar 7

Fig. 6. Representative photomicrographs of hematoxylin and eosin (400×) stained sections of aorta of (a) Positive control group (Group I) rat
showing broken and disorganized endothelium (DED); (b) Vehicle control group (Group II) rat showing normal endothelium (NE) without any
disruptions; (c) Curcumin 100 mg·kg−1 group (Group III) rat showing disorganized endothelium (DED) at a place and disorganized elastic fibres
(DEF); (d) Curcumin 200 mg·kg−1 group (Group IV) rat showing NE and tunica media (TM); (e) Curcumin 400 mg·kg−1 group (Group V) rat
showing NE and TM; (f) Folic acid per se 0.072 ␮g·g−1·day−1 (Group VI) rat showing NE without any morphological change and (g) Folic acid
(0.072 ␮g·g−1·day−1) + curcumin 200 mg·kg−1 (Group VII) rat showing NE. (A color version of this figure is available through the journal Web
site at http://www.nrcresearchpress.com/doi/10.1139/cjpp-2015-0156.)
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université Laval Bibliotheque on 11/21/15
For personal use only.

Published by NRC Research Press

 Pagination not final (cite DOI) / Pagination provisoire (citer le DOI)
8 Can. J. Physiol. Pharmacol. Vol. 94, 2016
eNOS. It has been reported that polyphenols present in fruits and
vegetables are able to modulate the production of NO in vascu-
lar endothelium, contributing to the prevention of endothelial dys-
function. We found that curcumin, a strongly phenolic compound,
enhanced acetylcholine induced vasodilation, decreased lipid
peroxidation, and increased SOD activity and GSH concentra-
tion, thus maintaining the antioxidant defense system of the
body. Although we did not quantify the serum homocysteine, it
has been reported in number of previous studies that curcumin
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Université Laval Bibliotheque on 11/21/15
reduces the levels of serum homocysteine (Kapoor et al. 2008;
Ramaswami et al. 2004).
In arthritic patients on methotrexate therapy, mortality is as-
cribed to cardiovascular complications (Landewé et al. 2000). In-
crease in the levels of homocysteine, direct injury to endothelium
is considered to be the factors behind the cardiovascular harmful
effects of methotrexate. The current investigation has unearthed
the beneficial effects of curcumin in maintaining vascular integ-
rity and alleviation of oxidative stress in methotrexate induced
endothelial dysfunction in rats. It can be concluded that curcumin
and folic acid could be administered in concomitance with meth-
otrexate to abrogate its side effects and augment its therapeutic
effects. Prospective clinical studies should be designed to corrob-
orate this claim in humans.
References
Aggarwal, B.B., and Harikumar, K.B. 2009. Potential therapeutic effects of cur-
cumin, the anti-inflammatory agent, against neurodegenerative, cardiovas-
cular, pulmonary, metabolic, autoimmune and neoplastic diseases. Int. J.
Biochem. Cell Biol. 41(1): 40–59. doi:10.1016/j.biocel.2008.06.010. PMID:18662800.
Anand, P., Kunnumakkara, A.B., Newman, R.A., and Aggarwal, B.B. 2007. Bio-
For personal use only.
availability of curcumin: problems and promises. Mol. Pharmaceut. 4(6):
807–818. doi:10.1021/mp700113r. PMID:17999464.
Anderson, T.J. 2006. Arterial stiffness or endothelial dysfunction as a surrogate
marker of vascular risk. Can. J. Cardiol. 22(Suppl. B): 72B–80B. doi:10.1016/
S0828-282X(06)70990-4. PMID:16498516.
Banji, D., Pinnapureddy, J., Banji, O.J., Saidulu, A., and Hayath, M.S. 2011. Syner-
gistic activity of curcumin with methotrexate in ameliorating Freund’s Com-
plete Adjuvant induced arthritis with reduced hepatotoxicity in experimental
animals. Eur. J. Pharmacol. 668(1–2): 293–298. doi:10.1016/j.ejphar.2011.06.006.
PMID:21693118.
Boonla, O., Kukongviriyapan, U., Pakdeechote, P., Kukongviriyapan, V.,
Pannangpetch, P., Prachaney, P., and Greenwald, S.E. 2014. Curcumin im-
proves endothelial dysfunction and vascular remodeling in 2K-1C hyperten-
sive rats by raising nitric oxide availability and reducing oxidative stress.
Nitric Oxide, 42: 44–53. doi:10.1016/j.niox.2014.09.001. PMID:25194767.
Clarke, Z.L., Moat, S.J., Miller, A.L., Randall, M.D., Lewis, M.J., and Lang, D. 2006.
Differential effects of low and high dose folic acid on endothelial dysfunction
in a murine model of mild hyperhomocysteinaemia. Eur. J. Pharmacol. 551:
92–97. doi:10.1016/j.ejphar.2006.08.085. PMID:17045583.
Cohn, J.N., Quyyumi, A.A., Hollenberg, N.K., and Jamerson, K.A. 2004. Surrogate
markers for cardiovascular disease functional markers. Circulation,
109(Suppl. IV): IV-31–IV-46. doi:10.1161/01.CIR.0000133442.99186.39.
Cortas, N.K., and Wakid, N.W. 1990. Determination of inorganic nitrate in serum
and urine by a kinetic cadmium reduction method. Clin. Chem. 36(8):
1440–3. PMID:2387039.
Crabtree, M.J., Hale, A.B., and Channon, K.M. 2011. Dihydrofolate reductase pro-
tects endothelial nitric oxide synthase from uncoupling in tetrahydrobiop-
terin deficiency. Free Radic. Biol. Med. 50(11): 1639–1646. doi:10.1016/j.
freeradbiomed.2011.03.010. PMID:21402147.
Davignon, J., and Ganz, P. 2004. Role of endothelial dysfunction in atheroscle-
rosis. Circulation, 109: III-27–III-32. doi:10.1161/01.CIR.0000131515.03336.f8.
PMID:15198963.
Ellman, G.L. 1959. Tissue sulfhydryl groups. Arch. Biochem. Biophys. 82: 70–77.
doi:10.1016/0003-9861(59)90090-6. PMID:13650640.
Esterbauer, H., Wäg, G., and Puhl, H. 1993. Lipid peroxidation and its role in
atherosclerosis. Br. Med. Bull. 49(3): 566–576. PMID:8221023.
Fleenor, B.S., Sindler, A.L., Marvi, N.K., Howell, K.L., Zigler, M.L., Yoshizawa, M.,
and Seals, D.R. 2013. Curcumin ameliorates arterial dysfunction and oxida-
tive stress with aging. Exp. Gerontol. 48(2): 269–276. doi:10.1016/j.exger.2012.
10.008. PMID:23142245.
Fukuda, Y., Teragawa, H., Matsuda, K., Yamagata, T., Matsuura, H., and
Chayama, K. 2002. Tetrahydrobiopterin restores endothelial function of cor-
onary arteries in patients with hypercholesterolaemia. Heart, 87(3): 264–269.
doi:10.1136/heart.87.3.264. PMID:11847169.
Hadi, A.R.H., Cornelia, S.C., and Jassim, A.S. 2005. Endothelial dysfunction: car-
diovascular risk factors, therapy, and outcome. Vasc. Health Risk Manage.
1(3): 183–198. PMID:17319104.
Hoekstra, M., Haagsma, C.J., Doelman, C.J., and van de Laar, M.A. 2005. Intermit-
tent rises in plasma homocysteine in patients with rheumatoid arthritis
treated with higher dose methotrexate. Ann. Rheum. Dis. 64(1): 141–143.
doi:10.1136/ard.2003.019828. PMID:15608313.
Hsieh, H.J., Liu, C.A., Huang, B., Tseng, A.H.H., and Wang, D.L. 2014. Shear-
induced endothelial mechanotransduction: the interplay between reactive
oxygen species (ROS) and nitric oxide (NO) and the pathophysiological impli-
cations. J. Biomed. Sci. 21: 3. doi:10.1186/1423-0127-21-3. PMID:24410814.
Kapoor, P., Ansari, M.N., and Bhandari, U. 2008. Modulatory effect of curcumin
on methionine- induced hyperlipidemia and hyperhomocysteinemia in al-
bino rats. Ind. J. Exp. Biol. 46: 534–540. PMID:18807758.
Kojda, G., and Harrison, D. 1999. Interactions between NO and reactive oxygen
species: pathophysiological importance in atherosclerosis, hypertension, di-
abetes and heart failure. Cardiovasc. Res. 43: 562–571. doi:10.1016/S0008-
6363(99)00169-8. PMID:10690328.
Lan, T.H., Xu, Z.W., Wang, Z., Wu, Y.L., Wu, W.K., and Tan, H.M. 2011. Ginsen-
oside Rb1 prevents homocysteine-induced endothelial dysfunction via PI3K/
Akt activation and PKC inhibition. Biochem. Pharmacol. 82: 148–155. doi:10.
1016/j.bcp.2011.04.001. PMID:21515242.
Landewe, R.B.M., van den Borne, B.E., Breedveld, F.C., and Dijkmans, B.A.C.
2000. Methotrexate effects in patients with rheumatoid arthritis with cardio-
vascular comorbidity. Lancet, 355(9215): 1616–1617. doi:10.1016/S0140-6736(00)
02222-4. PMID:10821370.
Lovri, J., Mesi, M., Macan, M., Koprivanac, M., Kelava, M., and Bradamante, V.
2008. Measurement of malondialdehyde (MDA) level in rat plasma after sim-
vastatin treatment using two different analytical methods. Period. Biol.
110(1): 63–67.
Merkle, C.J., Moore, I.M., Penton, B.S., Torres, B.J., Cueny, R.K., Schaeffer, R.C., Jr., and
Montgomery, D.W. 2000. Methotrexate causes apoptosis in postmitotic endo-
thelial cells. Biol. Res. Nurs. 2(1): 5–14. doi:10.1177/109980040000200102.
PMID:11232512.
Misra, H.P., and Fridovich, I. 1972. The role of superoxide anion in the autooxi-
dation of epinephrine and a simple assay for superoxide dismutase. J. Biol.
Chem. 247(10): 3170–3175. PMID:4623845.
Mittermayer, F., Pleiner, J., Schaller, G., Zorn, S., Namiranian, K., Kapiotis, S.,
et al. 2005. Tetrahydrobiopterin corrects Escherichia coli endotoxin-induced
endothelial dysfunction. Am. J. Physiol. Heart Circ. Physiol. 289(4): H1752–
H1757. doi:10.1152/ajpheart.00057.2005. PMID:15964928.
Morsy, M.A., Ibrahim, S.A., Amin, E.F., Kamel, M.Y., Rifaai, R.A., and Hassan, M.K.
2013. Curcumin ameliorates methotrexate-induced nephrotoxicity in rats.
Adv. Pharmacol. Sci. 2013: 387071. doi:10.1155/2013/387071. PMID:24381587.
Neuman, R.E., and Logan, M.A. 1950. The determination of collagen and elastin
in tissues. J. Biol. Chem. 186: 549–556. PMID:14794650.
Noguchi, K., Hamadate, N., Matsuzaki, T., Sakanashi, M., Nakasone, J.,
Uchida, T., et al. 2011. Increasing dihydrobiopterin causes dysfunction of
endothelial nitric oxide synthase in rats in vivo. Am. J. Physiol. Heart Circ.
Physiol. 301: H721–H729. doi:10.1152/ajpheart.01089.2010. PMID:21622822.
Rajendran, P., Rengarajan, T., Thangavel, J., Nishigaki, J., Sakthisekaran, D.,
Sethi, G., and Nishigaki, I. 2013. The vascular endothelium and human dis-
eases. Int. J. Biol. Sci. 9(10): 1057–1069. doi:10.7150/ijbs.7502. PMID:24250251.
Ramaswami, G., Chai, H., Yao, Q., Lin, P.H., Lumsden, A.B., and Chen, C. 2004.
Curcumin blocks homocysteine-induced endothelial dysfunction in porcine
coronary arteries. J. Vasc. Surg. 40(6): 1216–1222. doi:10.1016/j.jvs.2004.09.021.
PMID:15622377.
Sena, C.M., Pereira, A.M., and Seiça, R. 2013. Endothelial dysfunction — a major
mediator of diabetic vascular disease. Biochim. Biophys. Acta, 1832: 2216–
2231. doi:10.1016/j.bbadis.2013.08.006. PMID:23994612.
Soultati, A., Mountzios, G., Avgerinou, C., Papaxoinis, G., Pectasides, D.,
Dimopoulos, M.A., and Papadimitriou, C. 2012. Endothelial vascular toxicity
from chemotherapeutic agents: preclinical evidence and clinical implica-
tions. Cancer Treat. Rev. 38: 473–483. doi:10.1016/j.ctrv.2011.09.002. PMID:21982720.
Stroes, E., Kastelein, J., Cosentino, F., Erkelens, W., Wever, R., Koomans, H., et al.
1997. Tetrahydrobiopterin restores endothelial function in hypercholesterol-
emia. J. Clin. Invest. 99(1): 41–46. doi:10.1172/JCI119131. PMID:9011574.
Sudjarwo, S.A., Sudjarwo, K.E., Sudjarwo, G.W., and Koerniasari. 2011. Mecha-
nisms of endothelial cell protection by curcumin in hypercholesterolemia. J.
Appl. Pharm. Sci. 1(10): 32–35.
Svardal, A.M., Ueland, P.M., Berge, R.K., Aarsland, A., Aarsaether, N.,
Lønning, P.E., and Refsum, H. 1988. Effect of methotrexate on homocysteine
and other sulfur componds in tissues of rats fed a normal or defined, choline-
deficient diet. Cancer Chemoth. Pharm. 21(4): 313–318. doi:10.1007/BF00264197.
PMID:3370739.
Tardif, J.C., Heinonen, T., Orloff, D., and Libby, P. 2006. Vascular biomarkers and
surrogates in cardiovascular disease. Circulation, 113(25): 2936–2942. doi:10.
1161/CIRCULATIONAHA.105.598987. PMID:16801474.
Woessner, J.F., Jr. 1961. The determination of hydroxyproline in tissue and protein
samples containing small proportions of this amino acid. Arch. Biochem. Bio-
phys. 93: 440–447. doi:10.1016/0003-9861(61)90291-0. PMID:13786180.
Zeng, L., Yan, Z., Ding, S., Xu, K., and Wang, L. 2008. Endothelial injury, an
intriguing effect of methotrexate and cyclophosphamide during hematopoi-
etic stem cell transplantation in mice. Transplant Proc. 40: 2670–2673. doi:
10.1016/j.transproceed.2008.06.038. PMID:18929832.
Published by NRC Research Press