Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Int. J. lmmunopharmac., Vol. 16, No. 8, pp.

641-650, 1994
Elsevier Science Ltd
Pergamon International Society for lmmunopharmacology
Printed in Great Britain

0192 - 0561(94)E0022-F

ANTIALLERGIC ACTIVITY OF TYLOGENIN, A NOVEL STEROIDAL


COMPOUND FROM T Y L O P H O R A S YL VA TICA

JOHN N. GNABRE,* MARILYN J. HALONEN, t DAVID G. MARTIN* and JACOB L. PINNAS ~H

*Department of Biology, The Johns Hopkins University, Baltimore, MD 21218, U.S.A.;


Departments of *Pharmacology, and ~Medicine, University of Arizona Health Sciences Center,
Tucson, AZ 85724, U.S.A.; and *Division of Cancer Research, The Upjon Company,
Kalamazoo, MI 49001, U.S.A.

(Received 15 July 1993 and in final form 20 January 1994)

Abstract - - Tylogenin, a steroidal aglycone generated by acid hydrolysis from two seasonal glycosides
occurring in Tylophora sylvatica, inhibits IgE-induced basophil mediator release for allergic reactions. In the
rabbit basophil-dependent serotonin release (BDSR) assay system, the inhibitory activity of tylogenin (geom
mean tc50 = 39/aM) was significantly (/><0.05) greater than that of its parent glycosides, tylophoroside
(geom mean IC50263/~M) and acetyltylophoroside (geom mean ICs0 331 /~M), and that of dexamethasone
(geom mean ICso = 912/aM). The activity of tylogenin was found to increase with the incubation time. In the
human leukocyte-dependent histamine release (LDHR) test model, the glycosides had only a minimal
activity. In contrast, tylogenin, with a geom mean ICs0 = 49/JM, exerted a significantly (P<0.05) greater
potency than dexamethasone (ICs0 = 257 ~M). These results suggest that tylogenin could represent a new
class of antiallergic agents.

Since the identification of IgE antibody and its ground of traditional uses of natural products (Chen
association with immediate hypersensitivity reactions & Schmidt, 1923; Howell & Altounyan, 1967). This
(Ishizaka & Ishizaka, 1967), considerable effort has approach served as a basis for investigating the
been devoted to the understanding o f the interaction antiallergic activity of the African medicinal plant,
of specific allergens and IgE bound to receptors on Tylophora sylvatica. This plant is successfully used
mast cells and basophils, leading to mediator release in the coastal region of the Ivory Coast for the
in allergic and inflammatory reactions (Lichtenstein management of asthma and various allergic
et al., 1984). These mediators include histamine, disorders. The traditional formulation of the remedy
platelet-activating factor (PAF) and the eicosanoids consists of a paste made with plant leaves and
(leukotrienes and prostaglandins) (Marone, 1985). thoroughly mixed with lemon juice prior to oral
The pharmacological control or prevention of the administration. Tylophora species have been used
release of these mediators is regarded as a m a j o r for centuries in traditional medicine by many
stride in the management o f these disorders. cultures. In Indian folk medicine, T. asthmatica and
Many therapeutic agents including antiasthmatic T. indica are used for the management of asthma,
drugs have been discovered from ethnopharmacolo- and are sources o f the isoquinoline alkaloids,
gical inquiries aimed at the verification on a scientific tylophorine and tylophorinine (Ratnagiriswaran &

HAuthor to whom correspondence should be addressed, at: Allergy Center of Arizona, 1601 N. Tucson Blvd, #32, Tucson,
AZ 85716, U.S.A.
Abbreviations -- BDSR: Basophil-dependent serotonin release; LDHR: leukocyte-dependent histamine release; 3H-5HT:
tritiated 5-hydroxytryptamine (serotonin); HRP: horseradish peroxidase; BSA: bovine serum albumin; LSM: lymphocyte
separating medium; HNMT: histamine N-methyltransferase; HEPES" N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic
acid]; ACM" albumin-calcium-magnesium; LSD: least significant difference; ICs0: inhibitory concentration inducing 50°7o
response; TFA: trifluoroacetic acid; CCC: countercurrent chromatography; Tol: C H C I 3 : C C I n : M e O H : H z O :
Toluene:chloroform : carbon tetrachloride : water.

641
642 J. N. GNABREet al.
Venkatachalam, 1935). More recently, these generated by trifluoroacetic acid (TFA) hydrolysis of
alkaloids have been shown to possess antiasthmatic each parent compound for 2 rain at room
and antiinflammatory activity (Gopalakrishnan et temperature. Purification was achieved by (CCC)
al., 1980). An unnamed alkaloid has also been with a mixture of the solvent system, tolu-
extracted from T. sylvatica (Burkill, 1985). ene : chloroform : carbon tetrachloride : metha-
These observations fostered further investigation nol : water (tol : CHC13 : CC14 : MeOH : H20) in
of this African plant species to seek additional the ratio 8 : 2 : 2 : 8.5 : 1.5 (v/v), as indicated on
compounds with potential antiasthmatic activity. Fig. 1. Tylogenin occurs as 44.7% of the starting
Based on phytopharmacological criteria of simp- weight of its parent compound acetyltylophoroside.
licity, reproducibility, sensibility (Farnsworth, 1966) Tylophora compounds were characterized as
and the need for using a minimal amount of plant previously described (Gnabre et al., 1991).
extract, the basophil dependent serotonin release
(BDSR) technique was utilized herein for assay-
guided isolation of the active constituents of T. Preparation o f antigen and test compounds
sylvatica. These studies led to the isolation and The antigen for inducing IgE synthesis, horse-
characterization of tylophoroside and acetyltylopho- radish peroxidase (HRP) (Type II; Sigma Chemical
roside, two chemically related seasonal glycosides Co., St Louis, MO, U.S.A.), was prepared as a
which appear to be novel natural steroidal glycosides 10 mg/ml stock solution in 0.85% sodium chloride,
(Gnabre et al., 1991). Based on the traditional sterilized by Millipore filtration, and stored at
formulation of the antiallergic remedy, a mild acidic - 2 0 ° C . Ketotifen, cromolyn sodium, dexametha-
hydrolysis of these glycosides was undertaken to sone and prednisolone were purchased from Sigma.
generate the more active genin denoted tylogenin. Tylocrebrine was a generous gift from Dr Jack Cole,
Tylogenin is a three-ring steroid-like molecule former Dean of the College of Pharmacy, University
thought to represent a partially biodegraded steroid of Arizona. Tylogenin and dexamethasone were
(Devon & Scott, 1972; Gnabre et al., 1991). solubilized in 0.5% and 1% ethanol, respectively,
We report the activity of tylogenin and compare it prior to dilution in the buffer systems for the
with antiallergic and antiinflammatory drugs appropriate concentrations.
currently in clinical use. We postulate that tylogenin
may represent a novel antiallergic chemotype for
drug development. Its structural and functional Animal immunization f o r source o f IgE-bound
differences with glucocorticosteroids may have basophils
important clinical relevance.
Randomly bred Californian rabbits of both sexes
were immunized to produce specific anti-HRP
antibodies of the IgE class using an immunization
schedule beginning within 24 h after birth and
EXPERIMENTAL PROCEDURES periodically thereafter until 3 months of age
(Halonen et al., 1976).
Preparation o f Tylophora glycosides and tylogenin
The Tylophora glycosides, tylophoroside and its
monoacetate acetyltylophoroside, are seasonal com- Inhibition o f basophil-dependent serotonin release
ponents of a polar fraction obtained by methanol (BDSR ) in rabbit leukocytes
extraction of hexane-treated plant materials. The The BDSR technique was performed as described
two glycosides, respectively, with Rf 0.25 and 0.30 on previously (Pinckard et al., 1977). Briefly, blood
C18 TLC plates developed in 70% methanol, were from a sensitized rabbit was collected in 3.8%
purified using a combination of C18 column citrate, labeled with 3H-serotonin and washed free of
chromatography and countercurrent chromato- plasma. Replicate 250/al samples of blood cells in
graphy (CCC) with the Ito coil (Ito, 1986). The Tyrode's solution containing 0.25% gelatin and
studies were monitored stepwise by TLC densito- 0.0013 M Ca (Tyr-gel-Ca) were incubated for 2 h in
metry with UV detection at 220 nm, and the BDSR a water bath at 37°C with 200 ¢1 of the test
assay. The estimated proportion (% yield) based on compound appropriately diluted in Tyr-gel-Ca, or
the weight of the starting dry plant materials was with 200/al buffer alone (control samples). Fifty
0.01% and 0.011%, respectively. Tylogenin, the microliters of previously determined optimal
common steroidal genin to both glycosides, was concentration of antigen (antigen-treated samples) or
Tylogenin Inhibition of Mediator Release 643

Dry season glycoside 1: Tylophoroside (R = H) OMe


C48H72022 mol wt 1000 21~.~ O
18 /-20
Rainy season glycoside 2: Acetyltylophoroside (R = Ac) ~ || M,,,('
11"~
17 15
15~"-"
/ O
C50H74023 mol wt 1042 19

--.o:c
o
L-diginose | A Ii
6, ~ . ~ O ~ 1' OAc OAc

I OMe
O
t-acovenose /

D-glucose

HOJ~N,~jO\
6
~
[ OMe [
1"
TFA
2 min
room temp

.o" \ - OR
H O ~ ~ r" t Hydrolysis products
OH

I CCC
: ~ CC14:MeOH:8.5
2 H20)1.5

Sugars Compound 3: Tylogenin OMe


C2sHagOHj tool wt 535

O \ ~ , O~~Ae ~ O

O l i ',
OAc OAc
Fig. 1. Stereostructures of Tylophora compounds and generation of tylogenin. The parent glycoside, acetyltylophoroside,
was treated with trifluoroacetic acid (TFA) for 2 rain at room temperature. Tylogenin was purified from the hydrolysis
products by countercurrent chromatography (CCC) using the solvent system toluene: chloroform:carbon
tetrachloride : methanol : water (tol : CHCI3 : CC14 : MeOH : HzO) in the ratio 8 : 2 : 2 : 8.5 : 1.5 (v/v).

Tyr-gel-Ca alone (untreated-antigen controls) were to 250 ~1 of similarly treated cells with or without the
added to duplicate tubes containing compound- test compound. All tubes were incubated for 15 min
treated cells. Total releasable serotonin was at 37°C and centrifuged at 1500 × g for 20 rain at
determined by adding 50/A of thrombin (200 U/ml) 15°C. One-hundred microliter aliquots of super-
644 J. N. GNABREet al.
natant were assayed for released 3H-serotonin determined optimal concentration of Bermuda grass
by liquid scintillation counting. Antigen-induced pollen extract (Hollister-Stier, # 1142, Spokane,
serotonin release was expressed as the percentage of WA, U.S.A.), (antigen-treated samples) or 50 tal
total releasable serotonin. Inhibition of antigen- HEPES-ACM buffer alone (untreated-antigen
induced release was calculated as follows: controls). All samples were then incubated for
45 min at 37°C. Total releasable histamine was
°70 Inhibition = 100-(100) [Compound-treated Ag% determined by 10 min boiling of a similarly treated
release-Compound-treated Control °70 release)/(Un- sample without a test compound. Cells were
treated Ag % release-Untreated control % release)] immediately centrifuged at 200 × g for 5 rain at
10°C. A 200/al total of supernatants was transferred
into new polypropylene tubes and stored at - 7 0 ° C
until analyzed for histamine content.
Inhibition o f histamine release in human leukoeytes Histamine determination by the N-methyltrans-
(LDHR ) using N-methyl transferase assay ferase assay. Samples were thawed and appro-
Preparation o f peripheral-blood leukocytes priately diluted with (bis)distilled, deionized and
f r o m allergic donors. Atopic patients were autoclaved water, to provide samples in the
evaluated by skin prick test with grass pollen appropriate substrate concentration. Histamine
allergen and selected for the histamine release content was then determined using a sensitive
assay after informed consent was obtained. Fifty radioenzymatic assay previously described
milliliters of venous blood were collected in Vacu- (Verhurg et al., 1982). Spontaneous histamine
tainer sterile glass tubes containing 1.50 m g / m l release (control) and antigen-induced histamine
EDTA and separated over F i c o l l - H y p a q u e release were expressed as the percentage of total
solution with a modification of the method releasable histamine obtained by 10 min boiling of
described elsewhere (Sampson & Buckley, 1981). a control sample. Percentage inhibition of antigen-
Briefly, 25 ml of the anticoagulated blood were induced histamine release was calculated as in the
mixed with equal volume of HEPES-buffered BDSR assay.
saline (0.01 M, pH 7.4) containing 0.3 mg of
human serum albumin per milliliter (HEPES-A).
Seven milliliters of the blood mixture were Curve fitting and statistical analysis
carefully layered above 3 ml of Lymphocyte
The nonlinear mathematical curve fitting model
Separation Medium (LSM) (Organon Teknika
utilized for calculating IC~0in the BDSR and LDHR
Company) in 17 x 120 mm conical sterile
assays was the log-logistic dose - response model (De
polypropylene tubes. Blood samples were then
Lean et al., 1978):
centrifuged at 400 x g for 20 min at 20°C. The
leukocyte band formed at the plasma/LSM
interface was carefully aspirated and transferred
into 12 m m x 75 mm borosilicate culture tubes, Y = 100 / [1 + (lCso/dosey'],
resuspended in cold HEPES-A buffer, then
centrifuged at 200 x g for 10 min at 10°C. This
wash step was repeated three times after which the where Y = % inhibition, ICs0 = inhibitory concen-
final cell pellet was resuspended and adjusted to tration inducing 50°7o response, dose = concentra-
1.5 - 2.5 x 106 cells/ml in HEPES-buffered saline tion of test compound, and b = factor related to
containing 2 mM Ca 2- and 1 mM Mg ~ (HEPES- slope of the dose-response curve.
ACM). This cell suspension was saved for Curve fitting was performed with MINSQ, a
inhibition of histamine release. Nonlinear Parameter Estimation and Model
Inhibition o f histamine release. Duplicate 300/al Development (MicroMath Scientific Software, Salt
aliquots of the leukocyte suspension (containing 1.5 Lake City, UT 84121-3144).
to 3 x 103 basophils) were incubated for 4 h at For the statistical analysis of the data, activities of
37°C, in polypropylene tubes containing either the compounds were compared using the Least
100 ul of various concentrations of test compounds Significant Difference (LSD) test of the log Its0
appropriately diluted in HEPES-ACM buffer values. The derived Tcs0 are expressed as geometric
(compound-treated samples), or HEPES-ACM mean (Hancock et al., 1988) with the geometric
buffer alone (control samples). Cell samples were standard deviation following and given between
then challenged with either 50 tal of a previously parentheses.
Tylogenin Inhibition of Mediator Release 645

100 A Tylocrebrine /
/
90 - - • T. sylvatica (total extract) / /
80 - * Cromolyn sodium ~ ~ /

¢. 70 -
"~ 60
so
4o
30 -

" 10
r.
= 0 III ~l J l lllJ
-- 5 l0 30 100 300 1000 6000

Concentration, ~xg/ml
Fig. 2. Comparative activities between plant extract, tylocrebrine and cromolyn in BDSR assay. Various concentrations of
dried total plant extract were appropriately prepared in assay buffer and filtered with a 22/a Mill•pore filter. Replicate
250 ~1 samples of washed blood from a sensitized IgE anti-HRP producing rabbit were incubated with 200/al of
various concentrations of test products for 15 min, in a waterbath at 37°C. Cells were then challenged with 50/A of
3 × 10-4 mg/ml antigen (HRP) or buffer (control samples) or 200 U/ml thrombin (100% release), followed by another
15 rain incubation period. Samples were immediately centrifuged and 100/al supernatant were transferred into vials for
scintillation count. When indicated, spontaneous release was <6% and control antigen-induced serotonin release was
63 _+4.5% (mean_+ S.E.M.). Percent inhibition was calculated as previously described in Experimental Procedures.
• Hemolysis was observed with cromolyn at a concentration of 6 mg/ml, invalidating this result.

RESULTS slight reduction in antigen-induced IPB contraction


(results not shown). However, the amount of plant
Choice o f a suitable screening-method for the extract needed in the baths precluded more studies
evaluation o f the ant•allergic activity of plant total but the results led to the selection of a screening
extract method measuring directly antigen-induced mediator
A series of prospective studies was initially under- release (i.e. the BDSR and the H L D H R techniques).
taken on dichloromethane : methanol (1 : 1 v/v) dry
plant extract referred to as plant total extract in this
paper. These preliminary studies aimed at the Effects of plant extract on rabbit basophil mediator
selection of an adequate screening technique which release (BDRS)
would be utilized as a reference method for a full- Along with the plant extract, two agents were
scale isolation of plant active constituents. The concomitantly assayed with the BDSR method: the
choice of such a method was based on the standard ant•allergic agent, cromolyn sodium and
phytopharmacological criteria such as simplicity, tylocrebrine, an isoquinoline alkaloid isomeric with
sensitivity, reproducibility, selectivity (Farnsworth, tylophorine isolated from Indian Tylophora species
1966) and the minimal amount of test materials and known for its ant•asthmatic activity (Gopala-
required by the assay. These studies involved the krishnan et al., 1980). The results of these studies,
serum homologous passive cutaneous anaphylaxis illustrated in Fig. 2, indicate a d o s e - r e s p o n s e
(PCA) for potential inhibitory activity of skin mast inhibition of mediator release by the plant extract
cells, the isolated rabbit intrapulmonary bronchi and tylocrebrine. The threshold response (7.7°7o)
(IPB) for mediator antagonism and inhibition of with plant extract was obtained with 30/~g/ml and
mediator release from airways mast cells, and the maximum activity (98.5%) at 6000 ~g/ml. The plant
basophil-dependent serotonin release (BDSR) for extract had an estimated tcs0 of 450/~g/ml, whereas
inhibition of peripheral basophil mediator release. tylocrebrine had an ICs0of 250 ~g/ml. Cromolyn was
There was no evidence of P C A inhibition by plant only minimally active in this assay. The comparative
total extract in three rabbits. In the isolated IPB analysis of the results of these prospective studies
from actively sensitized IgE-producing rabbits, no indicated that the BDSR assay was the most suitable
antagonism of the bronchoconstrictive effects of phytochemical screening method for the isolation of
histamine was apparent. In contrast, there was a the active constituents of T. sylvatica.
646 J. N. GNABREet al.

100
90
~t

/
80

70
¢-. 60
50
40
30

20

10
I I L I L L I I
1 2 3 4 5 6 7 8

Time, hours
Fig. 3. Time course studies of tylogenin in BDSR assay. Tylogenin was solubilized in 0.5070 ethanol and adjusted to
10 ~ M. Replicate 250/al samples of washed blood from a sensitized rabbit were preincubated with 200 pl of compounds at
various timepoints after which cells were challenged with 50 tal of 3 z 10 4 mg/ml antigen (HRP) or buffer control
(spontaneous release) or 200 U/ml thrombin (maximum release), followed by a 15 min incubation period. Cells were
immediately centrifuged and 100/al supernatant were transferred into vials for scintillation count. Spontaneous release was
<12% after 8 h incubation. Control antigen-induced serotonin release was 63 + 4.5% (mean _+ S.E.M.). Percent inhibition
was calculated as previously described in Experimental Procedures. Each point represents the mean of eight determinations
from two experiments.

Isolation o f the active plant constituents generate tylogenin (Fig. 1) which was found to be the
To ascertain whether or not the antiallergic most active chemical group of the molecule, and
activity of plant total extract was due to the presence probably the major active constituent of the plant.
of isoquinoline alkaloids as is the case in the Indian
Tylophora species, a systematic alkaloid extraction Animal BDSR assay: time-course inhibition studies
was undertaken. These studies, effectively, showed a Tylogenin is primarily considered to be a steroid-
substantial yield (0.069%) of alkaloids a n d / o r nitro- like compound. Since basophil inhibition by
gen containing compounds (results not shown). corticosteroids varies with time (Schleimer et al.,
However, this alkaloidic fraction had only minimal 1981) a study was carried out to determine the
activity in the BDSR assay (45.8% inhibition) at optimal incubation time of this c o m p o u n d with
m a x i m u m concentration (1 m g / m l ) . These results basophils. The results illustrated in Fig. 3 demon-
were further substantiated by a comparative silica gel strate a time-dependent inhibition of basophil
TLC analysis which indicated significantly different mediator release by tylogenin at a fixed concentra-
T L C patterns between the presumed alkaloids from tion (10 .-4 M). The onset of this inhibition was
Tylophora sylvatica (R~ 0.2) and tylocrebrine (Rf detectable at less than 10 rain and reached a
0.8), developed in methanol : chloroform : water m a x i m u m around 2 h. Additional studies at fixed
(65 : 25 : 10, v/v), (lower phase), and visualized by incubation timepoints (15 min and 2 h) showed a
cerium sulfate charring [2% CeSO4 (w/v) in 5.6% similar shift throughout the inhibitory d o s e -
H2SO4 (v/v)]. Therefore, tylocrebine-type alkaloids response curve obtained with a 2 h incubation
readily occurring in the Indian Tylophora species period. Tylogenin exhibited a marked increase in
appeared to be absent in the African species, potency after 2 h incubation [ICso = 38.9 ~M (1.3)],
Tylophora sylvatica. The moderately active which was significantly (P<O.O01) greater than the
glycosides, which were the major constituents of a activity exhibited after 15 rain [lcso= 347 ~M (1.6),
polar fraction, were then set on isolation target. In n = 31.
the traditional sphere, the antiasthmatic/antiallergic
remedy of T. sylvatica is a paste made from plant
leaves. The formulation is orally administered only Comparative studies between tylogenin and various
after this paste is thoroughly mixed with lemon juice. agents
Based on the assumption that acid hydrolysis from At a 2 h optimal incubation time in the BDSR
either citric acid or the stomach might be required, assay, tylogenin was compared with various
the glycosides were cleaved by mild acid hydrolysis to compounds including the standard antiallergic drug
Tylogenin Inhibition of Mediator Release 647

100 -
• Tylogenin (n = 5) X-
90 -- 13 Acetyltylphorosidc (n = 5) _ /
80 - - O Dexamethasoiae (n = 5) ~/ /~
' 7 0 - - • Prednisolone(n=4)]~/ / ~
60 --
t~ 50

-.~.~
4(I

2030 i ~ ~ - / ~ /

10

-- 0.5 1 10 30 100 300 1000


Concentration (~tm)
Fig. 4. Comparative activities between Tylophora compounds and standard corticosteroids in the BDSR assay.
Experimental conditions were as in Fig. 2. Tylogenin and dexamethasone were initially solubilized in 0.5°7o and 1%
ethanol, respectively, prior to dilution in assay buffer. All test compounds were pre-incubated 2 h prior to antigen
challenge. Percent inhibition was calculated as previously described in Experimental Procedures. There was no significant
difference between the control samples with or without ethanol. A significant difference (P<0.05) was apparent between
tylogenin and the standard corticosteroids in terms of the lCso of these compounds.

~. 100 -- • Tylogenin
90 - zx Acetyltylophorosidc /
80 • Dexamethasone

07° ./ T
E
50
•7. 40
3o
20
:'2"-
•~ 10 -- & &
= 0 I I I llll I I I I l llJI I t I I l llll I I I
-- 0.3 1 10 30 100 300
Concentration (~tm)
Fig. 5. Comparative activities between Tylophora compounds and dexamethasone in the human LDHR assay. Tylogenin
and dexamethasone were initially solubilized in 0.5% and 1% ethanol. Replicate 300 MI samples of washed mixed
leukocytes from atopic donors were preincubated with 100 ~1 of various concentrations of test compounds for 4 h at
37°C. Cells with HEPES-ACM buffer alone (spontaneous release) and maximum releasable histamine samples
(obtained by 10 min boiling) were run concomitantly along with compound-treated samples. Cells were challenged with
3 x 10 -7 mg/ml optimal allergen (Bermuda grass extracts) concentration, followed by a 45 min incubation period. All
samples were immediately centrifuged and histamine was determined on 25 IA supernatant with the N-methyltransferase
assay. Spontaneous release was <7O7o, Control antigen-induced histamine release was 47.4 + 3.2O7o (mean _+ S.E.M.).
Percent inhibition was calculated as described in Experimental Procedures.

ketotifen. This b e n z o c y c l o h e p t a t h i o p h e n e is k n o w n a measure o f the efficacy o f a drug.)


for its multiple a n t i a n a p h y l a c t i c properties including We next c o m p a r e d tylogenin to its p a r e n t
i n h i b i t i o n of I g E - d e p e n d e n t m e d i a t o r release in glycosides a n d c o m m o n l y used corticosteroid drugs.
allergic patients ( M a r t i n & R o m e r , 1978). T h e re- T h e results o f these studies illustrated in Fig. 4 show
suits (not shown) revealed a greater efficacy for a significant difference (P<0.05) between the activity
tylogenin with m a x i m u m activity (92°70) at 300 taM, of tylogenin a n d t h a t o f acetyltylophoroside
whereas ketotifen reached a plateau (52°7o) a r o u n d [Ic50 = 331 /aM (1.8)]. There was n o significant
30 taM. (The height o f the plateau o f the difference between the ICs0 for acetyltylophoroside
dose - response curves equals m a x i m u m effect a n d is a n d tylophoroside [Ic50 = 263/aM (2.1), n = 5],
648 J. N. GNABREet al.
and the two glycosides had nearly superimposable (Hanks & Wallace, 1958). No statistically significant
concentration-response curves. The activity of difference in the percent cell viability was found
tylogenin was significantly (P<0.001) greater than between control samples and samples treated with
that of dexamethasone [tcs0 = 912/~M (2.8), n = 5] Tylophora compounds utilized at concentrations
and prednisolone []c50 = 1072 I~M (3.2), n = 4]. three times greater than the maximum concentration
Control release in the absence of the antigen and used in the LDHR assay.
the test compounds was less than 6% in the BDSR
assay. No significant difference was observed
between the control samples and 1% ethanol-treated
samples in the absence of antigen and test DISCUSSION
compounds after a 2 h incubation time period.
Mediator release for the control samples challenged The present results indicate that tylogenin
with antigen (HRP) optimal concentration possesses significant antiallergic activity. Because of
(3 × 10 4 mg/ml) in the absence of test compounds the heterogenous nature of allergic phenomena, a
was 63 + 4.5% (mean + S.E.M.). At the concentra- suitable antiaUergic screening method initially had to
tions used, Tylophora compounds had no significant be determined in pilot studies. These studies showed
effect on platelet activating factor (PAF) involved in that the active constituent from Tylophora sylvatica
the principle of this assay (see Experimental appears to act by inhibiting antigen-induced
Procedures). mediator release, but does not antagonize airway
contractility effects of these mediators once released.
Test models in the present studies included IgE-
Human LDHR assay: comparative activity between induced basophil mediator release from IgE-
tylogenin and dexamethasone producing rabbits and atopic human patients. These
Previous studies had shown that the standard methods are reliable in vitro models for the study of
corticosteroid drug dexamethasone exerts nearly hypersensitivity reactions and for evaluating
20% inhibition of histamine release from human potentially active antiallergic agents currently in
basophils after a 4 h incubation period (Schleimer et clinical use (Haydick & Ma, 1988; Schleimer et al.,
al., 1981). Tylophora compounds were tested along 1981). Human basophil histamine release has been
with dexamethasone in the human LDHR assay at considered the in vitro assay which best represents in
this particular timepoint, to compare the capacity of vivo allergic reactions (Haydik & Ma, 1988).
these agents to inhibit allergy mediator release from Tylogenin exhibited greater potency in the rabbit
Bermuda grass extract-challenged, IgE-bearing and human assay systems than did the parent
human basophils. The concentration-response glycosides. This suggests that the steroidal nucleus of
curves of the compounds illustrated in Fig. 5 show the molecule, obtained by acid hydrolysis, is the
that the glycoside had a relatively weak activity in structural determinant for specificity of the
this assay, whereas tylogenin and dexamethasone antiallergic action. The sugar side-chain (glycone)
exhibited a consistent dose-dependent inhibition of may confer decreased lipophilicity which may in part
the LDHR. The two compounds had similar activity account for the weak activity of the glycosides,
at concentrations <40 ~M. However, at concentra- provided that the mode of action of the compound
tions >40 I~M, tylogenin substantially gained in involves transport across the cell membrane. The
efficacy, reaching 92.8% inhibition at 300 I~M, traditional formulation of the remedy requires a
whereas dexamethasone displayed only 52% of treatment with lemon juice of the paste made from
inhibitory activity at this concentration. plant leaves. Whether this constitutes an intuitive
In the LDHR assay, control release in the absence attempt for cleaving the sugar side-chain remains a
of antigen and test compounds was less than 7%. No subject of speculation.
significant difference was apparent between control When assayed against ketotifen and cromolyn
mixed-leukocyte samples after 4 h incubation with or sodium, tylogenin consistently displayed greater
without 1% ethanol. Histamine release for the activity than these antiallergic drugs. The data for
control samples challenged with the optimal antigen cromolyn support previous studies indicating that
(Bermuda grass extract) concentration cromolyn does not inhibit mediator release from
(3 × 10 7 mg/ml) in the absence of test compounds human basophils (Lichtenstein et al., 1984). Gluco-
was 47 _+ 3.2% (mean + S.E.M.). corticosteroids have been termed the most effective
Cell viability studies were carried out with the known natural inhibitor of basophil mediator release
human leukocytes using the Eosin Y exclusion test (Nguyen et al., 1990). This inhibition is a time-
Tylogenin Inhibition of Mediator Release 649

dependent function (Schleimer et al., 1981). "biodegraded steroid" is not absolute and could be a
Tylogenin, which is considered to be a biodegraded subject of further discussion. Furthermore, the
steroid (Devon & Scott, 1972; Gnabre et al., 1991), stability of the sensitive butenolide side-chain and
was similarly affected by the time of incubation the 4fl,6o~-acetyl groups has yet to be established.
although a maximum effect was observed by 2 h, at These points would also need to be addressed to
which the compound exerted nearly a nine-fold undertake a definition before comparison with
increase in activity. corticosteroids. Nevertheless, the present study may
Previous studies had revealed that the cortico- lend credence to claims that the African medicinal
steroid, dexamethasone, inhibits human mixed- plant, T. sylvatica, possesses strong antiallergic
leukocytes with nearly 20% of activity at a micro- activity. We have demonstrated that this activity may
molar concentration after a 4 h incubation period be due in part to tylogenin, a steroidal agent with a
(Schleimer et al., 1981). This contrasts with the novel chemical structure, which inhibits in vitro
present studies in which a 10-fold increased concen- correlates of allergic disorders in actively sensitized
tration was required for dexamethasone to achieve IgE-producing rabbits and atopic human indi-
similar levels of activity at a fixed 4 h incubation viduals.
timepoint. Molecular manipulations of the pregnane nucleus
The standard corticosteroid, dexamethasone, has of corticosteroids have been attempted to develop
been shown to yield a greater activity after 12 to 24 h analogs with lesser side-effects (Scudeletti et al.,
incubation under special cell culture conditions and 1989). Notably, the 1-2 and 4-5 double bonds in the
media (Schleimer et al., 1981). Therefore, the overall A ring have been linked, respectively, to potency of
greater efficacy exhibited by tylogenin over carbohydrate regulation and antiinflammatory
dexamethasone after 2 to 4 h incubation endpoints is activity. The 9a-fluorination in dexamethasone
of unknown clinical significance. To avoid potential enhances the antiinflammatory activity but is
problems of stability of tylogenin or PAF at longer associated with marked sodium retention. Similarly,
incubation times, we chose simpler conditions with the C11 oxygen or hydroxylation in the C ring has
shorter incubation periods for comparing the two been linked to a significant antiinflammatory activity
agents within therapeutic concentration ranges for but also to an enhanced carbohydrate regulation,
dexamethasone. Furthermore, the activity of dexa- whereas the 17a-hydroxy substituent in the D ring is
methasone has also been shown to be concentration believed to be a prerequisite for the ultimate
dependent (Schleimer et al., 1981). The present study expression of antiinflammatory activity (Scudeletti et
demonstrates consistent dose-response curves for al., 1989). The lack of these pharmacophoric groups
dexamethasone and tylogenin at 2 and 4 h in tylogenin, coupled with a rapid onset of activity of
incubation periods. The remarkable linear d o s e - this novel compound, suggest the possibility of a
response curve observed with dexamethasone mode of action different from that of existing
contrasts with the biphasic nature of tylogenin corticosteroids. Tylogenin may therefore represent a
inhibition curve which shows an inflection point at new class of steroidal antiallergic and antiinflam-
30 gM. The resulting two apparent slopes of this matory agents with important clinical relevance and
curve strongly suggest the possibility of two may provide a novel antiallergic chemotype for drug
mechanisms of action which may be related to both development.
tylogenin concentration and incubation time, since
such a phenomenon was not observed after 2 h
Acknowledgements -- The authors wish to thank Dr John
incubation in the animal studies. However, the
Gaines, Director of Computer and Biostatistical Services,
mechanism underlying the action of tylogenin
for his valuable assistance with data analysis. They also
remains unknown and may differ from existing thank Professors Jean Guede Lorougnon and Laurent Ake
compounds. Assi, respectively, Chairman of the Department of Botany
Tylogenin is a novel compound and therefore its and Head of the Botanical Garden (University of Abidjan,
chemical classification as "steroid-like" or a Ivory Coast), for their expertise in botanical assistance.

REFERENCES

BURKILL,H. M. (1985). Asclepiadaceae. In The Useful Plants of West TropicalAfrica (eds Keay, R. W. J. and Hepper,
F. N.), Vol. 1, pp. 240-241. Royal Botanical Garden, Kew, London.
CHEN, K. K. & SCHM1DT,C. F. (1923). The action of ephedrine, an alkaloid from Ma Huang. Proe. Soc. exp. Biol. Med.,
21, 351-354.
650 J . N . GNABREet al.

CHILTON, F. H. & LICHTENSTEIN, L. M. (1990). Lipid mediators of allergic reaction. In 1939- 1989: Fifty Years Progress
in Allergy (ed. Waskman, B. H.). Chem. Immunol., 49, 173-205.
DE LEAN, A., MUNSON, P. J. & RODBARD,D. (1978). Simultaneous analysis of families of sigmoidal curves: application to
bioassay, radioligand assay, and physiological d o s e - response curves. Am. J. Physiol., 235 (2), E 9 7 - El02.
DEVON, T. K. & SCOTT, A. I. (1972). Terpenes. In Handbook o f Naturally Occurring Compounds (eds Devon, T. K. and
Scott, A. I.), Vol. II, pp. 385-478. Academic Press, London.
FARNSWORTH, N. R. (1966). Biological and phytochemical screening of plants. J. Pharm. Sci., 55 (3), 225- 269.
GNABRE, J. N., P1NNAS, J. L., MARTIN, D. G., MIZSACK,S. A., KLOOSTERMAN,D. A., BACZYNSKYJ,L., NIELSEN, J. W.,
BATES, R. B., HOFFMANN, J. J. & KANE, V. V. (1991). Characterization of steroidal glycosides from Tylophora
sylvatica. Tetrahedron, 47, 3545- 3554.
GOPALAKRISHNAN, C., SHANKARANARAYANAN, D., NAZIMUDEEN, S.K. & KAMESWARAN, R . L . (1980). Effect of
tylophorine, a major alkaloid of Tylophora indica, on immunopathological and inflammatory reactions. Ind. Z
Med. Res., 71,940-948.
HALONEN, M., FISHER, H. K., BLAIR, C., BUTLER, C. & PINCKARD, R. N. (1976). IgE-induced respiratory and circulatory
changes during systemic anaphylaxis in the rabbit. Am. Rev. Respir. Dis., 114, 961-970.
HANCOCK, A. A., BUSH, E. N., DUSANKA, S., KYN('L, J. J. & LIN, C. T. (1988). Data normalization before statistical
analysis: keeping the horse before the cart. Trends Pharmac. Sci., 9, 2 9 - 32.
HANKS, J. H. & WALLACE, J. H. (1958). Determination of cell viability. Proc. Soc. exp. Biol., 98, 188 - 192.
HAYD1K, I. B. & MA, W. S. (1988). Basophil histamine release: assay and interpretation. Clin. Rev. Allergy, 6, 141 - 162.
HOWELL, J. B. L. & ALTOUNVAN, R. E. C. (1967). A double-blind trial of disodium cromoglycate in the treatment of
allergic bronchial asthma. Lancet, ii, 539-542.
ISHIZAKA, K. & ISHIZAKA, T. (1967). Identification of lgE antibodies as a carrier of reaginic activity. J. hnmun., 99,
1187- 1198.
ISHlZAKA, T. (1981). Analysis of triggering events in mast cells for immunoglobulin IgE-mediated histamine release. J.
Allergy Clin. Immun., 67, 90-96.
1TO, Y. (1986). High speed countercurrent chromatography. CRC Crit. Rev. Analyt. Chem., 17, 65 - 143.
LIOHTENSTEIN, L. M., SCHLEIMER, R. P., MACGLASHAN, D. W. JR, PETERS, S. P., SCHULMAN, E. S., PROUD, D.,
CRETICOS, P. S., NACLERIO, R. M. & KAGEY-SOBOTKA,A. (1984). In vitro and in vivo studies of mediator release
from human mast cells. In Asthma: Physiology, Immunology, and Treatment (eds Kay, A. B., Austen, K. F. and
Lichenstein, L. M.), pp. 1 - 8. Academic Press, London.
MARONE, G. (1985). The role of basophils and mast cells in the pathogenesis of pulmonary diseases. Int. Arch. Allergy
appl. Immun., 76 (Suppl. 1), 70-82.
MARTIN, U. & ROMER, D. (1978). The pharmacological properties of a new orally active, antiallergic compound: ketotifen,
a benzocyclopheptathiophene. Arzneim: Forsch, 5, 770-782.
NGUYEN, K. L., GILLIS~ S. & MACGLASHAN, D. W. (1990). A comparative study of releasing and nonreleasing human
basophils: nonreleasing basophils lack an early component of the signal transducing pathway that follows IgE cross-
linking. J. Allergy Clin. Immun., 85 (6), 1020- 1029.
PINCKARD, R. N., HALONEN, M., PALMER, J. D., BUTLER, C., SHAW, J. O. & HENSON, P. M. (1977). lntravascular
aggregation and pulmonary sequestration of platelets during IgE-induced systemic anaphylaxis in the rabbit.
Abrogation of lethal anaphylactic shock by platelet depletion. J. Immun., 119, 2185- 2193.
RATNAGIR1SWARAN,A. N. & VENKATACHALAM,K. (1935). The chemical examination of Tylophora asthmatica and
isolation the alkaloids tylophorine and tylophorinine. Ind. J. med. Res., 22, 433- 441.
SAMPSON, H. A. & BUC~(LEY, R. H. (1981). Human IgE synthesis in vitro: a reassessment. J. lmmun., 127, 829-834.
SCHI.EIMER, R . P . , LlCUTENSTE1N, L . M . & GILLESI'IE, E. (1981). Inhibition of basophil histamine release by
antiinflammatory steroids. Nature, 292, 454 455.
SCUDELETTI, M., CIPRANDI, G., PRONZATO, C., PASSALACQUA, G., 1MBIMBO, B., BACQUAS('O,M. & CANONI('A., G. W.
(1989). Comparison of the immunomodulation of prednisolone, dexamethasone and deflazacort on T-cell activation
pathways and gamma-interferon production. Int. J. lmmunother., 5, 85-90.
VERBURC;, K. M., BOWSHER, R. R. & HENRY, D. P. (1982). A new radioenzymatic assay for histamine using purified
histamine N-methyltransferase. Life Sei., 32, 2855- 2867.

You might also like