C Pharmacology & Toxicology 2001, 88, 300–303. Copyright CPrinted in Denmark . All rights reserved

ISSN 0901-9928

Comparison of the Embryotoxic Effects of Saporin, Agrostin(Type 1 Ribosome-Inactivating Proteins) and Ricin

(a Type 2 Ribosome-Inactivating Protein)Wood-Yee Chan1 and Tzi-Bun Ng2

1Department of Anatomy and 2Department of Biochemistry, Faculty of Medicine,The Chinese University of Hong Kong, Shatin, Hong Kong, China

(Received October 24, 2000; Accepted January 8, 2001)

Abstract: The effects of two type I ribosome-inactivating proteins, saporin and agrostin, and the type 2 ribosome-inactivat-ing protein ricin and its constituent A and B chains, on the development of cultured mouse embryos, were investigated.Saporin and agrostin had similar embryotoxicity which was approximately 105 times weaker than that of ricin and its Aand B chains, and the embryotoxicity of saporin and agrostin increased gradually with the dosage. In contrast, there wasan abrupt rise in embryotoxictiy of ricin-A chain and B chain as the dose was raised from 10 to 20 ng mlª1. Saporin andagrostin did not affect the heart-beat and otic placode, but at 10 ng mlª1 ricin and 40 ng mlª1 ricin-A chain and B chain,all of the treated embryos exhibited abnormalities in heart-beat, yolk sac circulation, body axis, optic placode, oticplacode, forelimb buds, branchial apparatus and cranial neural tube.

Ribosome-inactivating proteins are a family of proteinswith interesting and potentially applicable activities includ-ing antiviral, anti-human immunodeficiency virus, antipro-liferative, antitumour and immunosuppressive activities (Nget al. 1992). They also exert undesirable effects such as anti-fertility and embryotoxic activities (Ng et al. 1992). It isreported herein that the type 2 ribosome-inactivating pro-tein, ricin, which is made up of a ribosome-inactivating pro-tein chain and a lectin chain, was considerably more em-bryotoxic than the single-chained type 1 RIPs’ agrostin andsaproin. The constituent chains of ricin also had muchgreater embryotoxicity than agrostin and saporin.

Materials and Methods

Ribosome-inactivating proteins. Saporin, agrostin, ricin, ricin Achain and ricin B chain were purchased from Sigma Chemical Com-pany, St. Louis, MO, USA.

Embryo culture in the presence of ribosome-inactivating proteins.Mouse embryos were explanted and cultured using a modified ver-sion of the whole embryo culture method (Sadler 1979; Chan et al.1995). Mice on 8.5 days post coitum were sacrificed. Embryos weredissected out of the decidua in PB1 medium, Reichert’s membranewas removed, leaving intact the visceral yolk sac and the ectoplacen-tal cone. The embryos which possessed 2–4 somites were dividedinto groups of five each and cultured in 50 ml serum bottles(Wheaton). Each of the bottles contained 5 ml culture mediumwhich was composed of a 1:1 mixture of Dubecco’s modified Eagle’smedium (DMEM, Gibco) and heat-inactivated (56 æ for 30 min.) ratserum. The culture medium used had been sterilized by filtration

Author for correspondence: T. B. Ng, Department of Biochem-istry, Faculty of Medicine, The Chinese University of HongKong, Shaitn, Hong Kong, China (fax π852 2603 5123, e-mailbiochemistry/cuhk.edu.hk).

through Millipore filters (pore size 0.22 mm) and equilibrated with5% CO2 in air overnight. Various doses of the ribosome-inactivatingprotein dissolved in a small volume of phosphate-buffered salinewere added to the culture medium. The cultures were kept at 37 æ,rotated at 20–40 rpm and re-gasssed with the same gas mixtureevery 8 hr (Chan et al. 1995).

Examination of cultured embryos. Twenty-four hr after culture theembryos were transferred to warm PB1 medium for inspection ofyolk sac circulation and heart-beat. Subsequently the embryos weredissected free of embryonic membranes and washed twice withphosphate-buffered saline. The gross morphology, somite number,the presence of various organ primordia, and the existence of ab-normalities were noted. The embryos were then fixed in Bouin’sfluid for routine histological observations. For scanning electronmicroscopic studies the embryos were fixed in half-strength Karnov-sky’s fixative for 1.5 hr, post-fixed with 1% OsO4, and critical point-dried with Freon. Gold-palladium-coated specimens were examinedwith a Joel JSM-35 GF microscope which was operated at 15 KV(Chan et al. 1995).

Statistical analyses. The data were presented as means∫S.E. Thelevel of significance was set at PΩ0.05.

Results

All control embryos developed normally in vitro. No differ-ences in the extent of morphogenesis were detected betweenthe in vivo and in vitro controls. The control embryos ex-hibited a powerful heart-beat, a prominent yolk sac circula-tion, optic placodes, otic placodes, a closed cranial tube andan increase in somite number (fig. 1A). Embryos exposedto 50 mg mlª1 saporin, 50 mg mlª1 agrostin, 1 ng mlª1 ricin,10 ng mlª1 ricin-A chain and 10 ng mlª1 ricin-B chain weresimilar to control embryos. At 100 mg mlª1 saporin andagrostin (fig. 1B), however, abnormalities started to appear

301EMBRYOTOXIC EFFECTS OF SAPORIN, AGROSTIN AND RICIN

Fig. 1. Scanning electron micrographs of embryos 24 hr after treat-ment with ribosome-inactivating protein in vitro. Bar, 100 mm. (A)Lateral view of an in vitro control mouse embryo, which had beentreated with phosphate-buffered saline only. Note the normal bodyaxis, the presence of the expanded forebrain (F) and midbrain (M)and the normal branchial apparatus (BA), otic placode (Ot), devel-oping heart (H), forelimb bud (L) and somites (S). (B) Lateral viewof a mouse embryo treated with 100 mg/ml of agrostin. The embryohas abnormal body axis, small branchial apparatus (BA) and heart(H), open cranial neural tube (arrowhead) and absence of forelimblimb. (C) Lateral view of a mouse embryo treated with 200 mg/mlof saporin. The body axis of the embryo is twisted; the cranial neu-ral tube is open (arrowhead); and the branchial apparatus (BA),otic placode (Ot) and heart (H) are abnormal. (D) Dorsal view ofa mouse embryo treated with 5 ng/ml of ricin. The body axis istwisted; the otic placode is enlarged (Ot); the cranial neural tube isopen (arrowheads); and the somites are morphologically indis-tinguishable.

in the embryos. The body axis became distorted, forelimbbuds did not appear and the cranial neural tube remainedopen. At 200 and 300 mg mlª1 of either saporin or agrostin,there was an increase in the incidence of abnormal embryos,and these embryos demonstrated abnormalities in the yolksac circulation, body axis, forelimb buds, branchial appar-atus and cranial neural tube (fig. 1C). Agrostin was slightlymore embryotoxic than saporin as revealed by a compari-son of the percentages of abnormalities at 100 and 200 mgmlª1 of the ribosome-inactivating protein (table 1). Agros-tin but not saporin induced abnormalities in the optic plac-odes, but neither of them affected the heart-beat and oticplacodes. More than two-thirds of the embryos treated with

5 ng mlª1 ricin, 20 ng mlª1 ricin-A chain and 20 ng mlª1

ricin-B chain were abnormal and abnormalities were de-tected in all of the features examined including heart-beat,yolk sac circulation, body axis, optic placode, otic placode,forelimb buds, branchial apparatus and cranial neural tube(fig. 1D). All of the embryos treated with 10, 40 and 40 ngmlª1 of ricin, ricin-A chain and ricin-B chain respectivelywere abnormal; exhibiting all of the aforementioned abnor-malities. Ricin-A chain at 40 ng mlª1 appeared to induce aslightly higher incidence of abnormalities in heartbeat, yolksac circulation, body axis, optic placode, forelimb buds andbranchial apparatus than ricin-B chain (table 1). The toxiceffects of the ribosome inactivating proteins on mouse em-bryos and HeLa cells are compared in table 2. HeLa cellswere more sensitive than mouse embryos to ricin. However,HeLa cells and mouse embryos had similar sensitivities toagrostin, saporin and ricin A.

Discussion

The conditions used for embryo culture in the present studywere optimal as evidenced by the minimal number of abnor-mal embryos, the normal embryonic morphology and theusual increase in somite number in the in vitro controlgroups.

The type I ribosome-inactivating proteins saporin andagrostin exerted a much milder, approximately 105 timeslower, embryotoxicity than the type II ribosome-inactivat-ing protein ricin and its constituent A and B chains.Saporin and agrostin did not affect the heart-beat and theotic placode even when tested at 300 mg mlª1. In contrast,abnormalities were seen in all of the tissues examined whenthe mouse embryos were exposed to 5, 20 and 20 ng mlª1

of ricin, ricin-A chain and ricin-B chain respectively. Themuch greater embryotoxicity of ricin, a type II ribosome-inactivating protein, was due to binding of its lectin chainto cells facilitating entry of its ribosome-inactivating pro-tein chain into the cell. The constituent lectin chain andribosome-inactivating protein chain of ricin were also muchmore toxic than the type I ribosome-inactivating proteins.

There was a gradual increase in embryotoxicity ofsaporin and agrostin as the dosage of the ribosome-inactiv-ating protein increased. However, the initial parts of thedose-response curves for the embryotoxicity of ricin and itsconstituent A and B chains were much steeper in slope. Ric-in-B chain appeared to have a slightly weaker embryotoxic-ity than ricin-A chain.

The present study demonstrated a difference in em-bryotoxic potency between type I and type II ribosome-in-activating proteins. A previous study on the type I ribo-some-inactivating proteins, luffaculin, luffin-a, luffin-b andmomorcochin has demonstrated difference in their abilityto induce toxic effects on the mouse embryo, the rankingbeing luffin-b.luffaculin.luffin-a and momorcochin(Chan et al. 1994).

The cell-free translation-inhibitory potencies of the ribo-some-inactivating proteins have been reported as follows:

WOOD-YEE CHAN AND TZI-BUN NG302

Tab

le1.

Dev

elop

men

tof

8.5-

days

mou

seem

bryo

sun

der

the

influ

ence

ofri

boso

me-

inac

tiva

ting

prot

eins

.

Abn

orm

alit

ies

(%)b

Num

ber

Num

ber

ofIn

itia

lF

inal

ofab

norm

alA

bnro

mal

Twis

ted

Abn

orm

alA

bnor

mal

Abs

ence

Abn

orm

alO

pen

Con

cent

rati

onem

bryo

sso

mit

eso

mit

eem

bryo

sA

bnor

mal

yolk

sac

body

opti

cot

icof

fore

limb

bran

chia

lcr

ania

lTr

eatm

ent

(mg/

ml)

exam

ined

num

ber

num

bera

(%)b

heat

beat

circ

ulat

ion

axis

plac

ode

plac

ode

buds

appa

ratu

sne

ural

tube

Con

trol

:In

vivo

–34

–23

.4∫

1.1

3(9

)0

(0)

0(0

)0

(0)

0(0

)0

(0)

3(9

)3

(9)

3(9

)In

vitr

o5

mlP

BS

306–

822

.8∫

1.6

3(1

0)0

(0)

0(0

)0

(0)

1(3

)0

(0)

1(3

)1

(3)

3(1

0)

Exp

erim

enta

l:

Sapo

rin

5012

6–8

22.2

∫1.

91

(8)

0(0

)0

(0)

0(0

)0

(0)

0(0

)1

(8)

0(0

)1

(8)

100

106–

821

.5∫

2.1

2(2

0)0

(0)

1(1

0)2

(20)

*1

(10)

0(0

)2

(20)

*2

(20)

*2

(20)

*20

015

6–8

19.1

∫2.

77

(47)

*1

(7)

3(2

0)*

3(2

0)*

1(7

)1

(7)

7(4

7)*

7(4

7)*

7(4

7)*

300

106–

820

.8∫

3.7

6(6

0)*

1(1

0)6

(60)

*6

(60)

*1

(10)

1(1

0)6

(60)

*6

(60)

*6

(60)

*

Agr

osti

n50

126–

823

.0∫

1.1

2(1

7)0

(0)

0(0

)1

(8)

1(0

)0

(0)

2(1

7)1

(8)

2(1

7)10

010

6–8

19.0

∫2.

23

(30)

*1

(10)

1(1

0)2

(20)

*2

(20)

*0

(0)

3(3

0)*

1(1

0)3

(30)

*20

014

6–8

19.1

∫2.

97

(50)

*1

(7)

4(2

9)*

4(2

9)*

2(1

4)*

0(0

)5

(36)

*4

(29)

*7

(50)

*30

012

6–8

18.8

∫1.

7*7

(58)

*1

(8)

7(5

8)*

5(4

2)*

2(1

7)*

0(0

)7

(58)

*3

(25)

*7

(58)

*

Ric

in1¿

10ª

310

6–8

21.4

∫1.

21

(10)

0(0

)0

(0)

0(0

)0

(0)

0(0

)1

(10)

0(0

)1

(10)

5¿10

ª3

126–

816

.6∫

2.1*

8(6

7)*

8(6

7)*

8(6

7)*

8(6

7)*

8(6

7)*

2(1

7)*

8(6

7)*

8(6

7)*

8(6

7)*

10¿

10ª

312

6–8

–d12

(100

)*12

(100

)*12

(100

)*12

(100

)*12

(100

)*12

(100

)*12

(100

)*12

(100

)*12

(100

)*

Ric

in10

¿10

ª3

106–

819

.8∫

1.2

1(1

0)0

(0)

0(0

)0

(0)

0(0

)0

(0)

0(0

)1

(10)

1(1

0)A

-cha

in20

¿10

ª3

106–

816

.0∫

1.3*

7(7

0)*

7(7

0)*

7(7

0)*

7(7

0)*

6(6

0)*

2(2

0)*

7(7

0)*

7(7

0)*

5(5

0)*

40¿

10ª

310

6–8

16.8

∫1.

4*10

(100

)*10

(100

)*10

(100

)*10

(100

)*10

(100

)*2

(20)

*10

(100

)*9

(90)

*6

(60)

*

Ric

in10

¿10

ª3

106–

820

.9∫

1.8

1(1

0)0

(0)

0(0

)0

(0)

0(0

)0

(0)

0(0

)1

(10)

B-c

hain

20¿

10ª

310

6–8

16.4

∫1.

4*7

(70)

*7

(70)

*7

(70)

*7

(70)

*6

(60)

*0

(0)

7(7

0)*

4(4

0)*

4(4

0)*

40¿

10ª

310

6–8

17.2

∫1.

7*10

(100

)*6

(60)

*9

(90)

*9

(90)

*6

(60)

*0

(0)

9(9

0)*

6(6

0)*

6(6

0)*

aV

alue

sar

em

ean∫

S.E

.b

Num

ber

inpa

rent

hese

sre

pres

ents

perc

enta

ge.

cP

BS:

Pho

spha

te-b

uffe

red

salin

e(p

HΩ

7.4)

.d

Som

ites

wer

eno

tdi

scer

nibl

ein

all

the

embr

yos

exam

ined

.*

P,

0.05

,si

gnifi

cant

lydi

ffer

ent

from

the

invi

voco

ntro

lva

lue

bySt

uden

ts’

t-te

stor

Chi

-squ

are

test

.

303EMBRYOTOXIC EFFECTS OF SAPORIN, AGROSTIN AND RICIN

Table 2.

Comparison of the concentrations of ribosome-inactivating pro-teins required to inhibit protein synthesis in HeLa cells and producetoxic effects on cultured mouse embryos.

Protein synthesis inhibition Embryotoxicityb

in whole HeLa cellsa (EC50 in mg/ml)(IC50 in mg/ml)

Agrostin 234 200Saporin 98 200Ricin 0.033¿10ª3 5¿10ª3

Ricin A-chain 12¿10ª3 20¿10ª3

a: data from Barbieri et al. (1993), b: data from table 1.

momorcochin and luffaculin (IC50Ω0.12 nM).agrostin andsaporin (0.5 nM).luffin a (IC50Ω1 nM).luffin b (IC50Ω4nM) (Barbieri et. al. 1993). However, the ranking is differ-ent from that for their embryotoxicity. When taken to-gether, the data suggest a lack of correlation between therelative potencies of ribosome inactivating proteins in dif-ferent assay systems, indicating different requirements forthe various biological activities of ribosome inactivatingproteins.

The two type 1 ribosome-inactivating proteins saporinand agrostin, and the A-chain of the type 2 ribosome-inac-tivating protein ricin, produced inhibitory effects on HeLacells and cultured mouse embryos with similar IC50/EC50

values. In contrast, HeLa cells were more sensitive than cul-tured mouse embryos to ricin.

In view of the toxicity of type 1 and type 2 ribosome-inactivating proteins, ribosome inactivating protein-basedimmunotoxins have been developed (Ghetie & Vitetta1994).

AcknowledgementsWe thank the Research Grants Council for award of an

earmarked grant and Ms. Fion Yung for excellent sec-retarial assistance.

References

Barbieri, L., M. G. Battelli & F. Stirpe: Ribosome-inactivating pro-teins from plants. Biochim. Biophys. Acta. 1993, 1154, 237–282.

Chan, W. Y., T. B. Ng & H. W. Yeung: Differential abilities of theribosome inactivating proteins luffaculin, luffins and momorco-chin to induce abnormalities in developing mouse embryos invitro. Gen. Pharmac 1994, 25, 363–367.

Chan, W. Y., T. B. Ng & P. C. Shaw: Mouse embryonic developmentand tumor cell growth under the influence of trichosanthin (aribosome inactivating protein) and its muteins. Teratogen. Car-cinogen. Mutagen. 1995, 15, 259–268.

Ghetie, M. & E. S. Vitetta: Recent developments in immunotoxintherapy. Curr. Opinion Immunol. 1994, 6, 707–714.

Ng, T. B., W. Y. Chan & H. W. Yeung: Proteins with abortifacient,immunomodulatory, antitumor and anti-AIDS activities fromCucurbitaceae plants. Gen. Pharm. 1992, 23, 575–590.

Sadler, T. W.: Culture of early somite mouse embryos during or-ganogenesis. J. Embryol. Exp. Morphol. 1979, 49, 17–25.