Note: Descriptions are shown in the official language in which they were submitted.
Process Eor the preparation of_antitumoral glycoproteins
modified on their carbohydrate units.
The present invention relates to new antitumoral
glycoproteins whose carbohydrate u~its are modified
S by oxidation with the periodate ion.
~ lore particularly, but without implying a limita-
tion, the present invention r~fers to the new glyco-
proteins which inactivate ribosomes and have a pro-
longed action.
The term "glycoprotein which inactivates ribo-
somes", as used in the present description and also in
the claims, denotes any substance which carries
saccharide units belonging to the class of macromole-
cules which inactivate ribosomes and consequently in-
hibit protein synthesis in eucaryotic cells, as well as
any fragment of the said substance which possesses the
same inactivating property, it being possible for the
said glycoprotein which inactivates ribosomes to be of
natural or biosynthetic origin, being derived from a
cell whose genotype has been modified for this purpose,
and it also being possible Eor the said glycoprotein
which inactivates ribosomes to be modified on the func-
tional groups of its amino acids so that it can easily
be coupled with an antibody.
In the description, the expression "glycoprotein
which inactivates ribosomes" will be denoced by the
symbols GPIR.
In the description, the term "periodate" denotes
the I04 ion, which is also referred to in the litera-
ture as "metaperiodate".
The glycoproteins which inactivate ribosomes(GPIR) are especially useful as intermediates in the
preparation of immunotoxins by coupling with antibodies.
U.S. Patent 4 340 535 and French Patent Applica-
tions no. 81/07596 and no. 81!~1836 describe the pre-
paration of anticancer products called conjugates,
.,v ,
:,. -, ,. :
:: ':
: ~- ' ,., , :
~ -:
37~
-- 2
which are obtained by the coupling, by means of a
covalent bond, of the A chain of ricin with antibodies
or antibody fragments directed against antigens carried
by the cell to be destroyed. The products of this type
have been designated, and are designated in the present
App]lcation, under the generic name of immunotoxins.
Conjugates analogous to the previously des-
cribed immunotoxins containing the A chain oE ricin are
~nown which are also suitable as anticancer drugs and
result from the coupling, by means of a covalent bond,
of antibodies or antibody fragments with other glyco-
proteins ~hich inactivate ribosomes, such as, in par-
ticular, the gelonine extracted from Gelonium multi-
florum (Eur. J. Biochem., 1981, 116, 447-454; Cancer
Res., 1984, 44, 129-133) or the inhibitor extracted
from Momordica charantia (MOM) (U.S. Patent 4 368 149~.
These glycoproteins which inactivate ribosomes
(GPIR), and which have properties similar to those of
the A chain of ricin, are substances with a molecular
weight of the order of magnitude of 20,000 and 30,000
(Cancer Survey, 1982, 1, 489-520).
It is also known that the cytotoxic activity
of these immunotoxins can be potentiated by a variety
of adjuvant substances such as ammonium ~salts, various
amines or certain carboxylic ionophores such as
monensin and nigericin.~ - ~
HoweYer, the therapeutlc effects of immuno-
toxins, whether activated or not, can only manifest
themselves fully inasmuch as the immunotoxin~is
capable, through its antlbody part, o~ becoming
localized in vivo, in the active form, on the~target
cells to~be destroyed (si~ne qua non~c~ondit~ton for an~y
expression of immunotoxin activity). The capacity of
the immunotoxin to become localized on the targe~ de-
pends first and foremost~on the ab~l1ty of~ the immuno-
:,
~9~
-- 3 ~
toxin to remain in the bloodstream and the extracellu-
lar fluids, in the active form, for sufficient lengths
of time for it to reach its target cells and in suffi-
ciently large concentrations for the degree of occu-
pation of the corresponding antigenic sites to be high.
~ umerous studies have made it possible to es-
tablish the plasma elimination kinetics of immunotoxins
after intravenous injection into different animal
models. It is apparent that, after injection, the
plasma level of biologically active immunotoxin de-
creases very rapidly and very substantiallyO Thus, in
a typical case involving rabbits, in a model using an
immunotoxin synthesized by coupling, with the aid of
an arm containing a disulfide bridge, the A chain of
ricin with a monoclonal antibody directed against the
antigen T65 of human T lymphocytes (antibody T101), it
appears that 97% of the immunotoxin present in the
bloodstream at time O after injection disappears in 30
minutes and 99.9% disappears in 17 hours. This rapid
2~ disappearance of the immunotoxin quite obviously de-
tracts from the expression of its complete cytotoxic
capacity by preventing the immunotoxin from saturating,
for a prolonged period, a~high proportion of the target
antigens carried by the cells to be destroyed. ~ore-
~5 over, comparison of the plasma elimination kinetics ofthe immunotoxins with those of the corresponding non-
conjugated antibodies sho~s that, on the contrary, the
antibodies remain in the plasma at a high level for
relatively long periods, as is well known. Now, there
is always a certain residual level of non-conjugated
antibodies, even in the most highly purified~immuno-
toxin preparations. Through the effect of the differ-
ential elimination rates of immunotoxins and~anti-
bodies, the non-conjugated antibodies, which are
initially in a very small m1nority, gradually become a
. :
- ~ ~
7~
majority after a few hours and these antibodies there-
fore gradually become, by competition, powerf~1l an-
tagonists for the fi~ation of the immunotoxins to
their targets.
The advantage of increasing the plasma persis-
tence of immuno~oxins, in the active form, in order to
increase both the duration and the degree of occupation
of the target antigens, and consequently to improve the
therapeutic effects of the immunotoxins, is therefore
clearly apparent.
Furthermore, in vivo localization experiments
on the immunotoxin containing the A chain of ricin,
radiolabelled and then injected into animals without a
specific target, have shown that, in the first few
lS minutes after injection, the conjugate becomes local-
ized preferentially in the liver. The same applies to
the A chain of ricin, which follows the same pattern
when it is injected in the uncoupled form. This
strongly suggests that the immunotoxin fixes in the
liver via the cytotoxic sub-unit which it contains.
It is known that the A chain of ricin is a glycoprotein
whose polyosidic groups include especially mannose
residues and N-acetylglucosamine residues, some mannose
residues being in terminal positions (Agri~ Biol.
Chem., 1978, 42, 501). The existence, in the liver, of
receptors capable of recognizing glycoproteins having
these terminal mannose residues has also been est~ab~
lished. Moreover, it has been~shown that the glyco-
proteins recognized by these~re~ceptors - the latter
bein8 present es~sentially on the Kupffer cel~ls -
are rapidly elimin~at:ed from the bl~oodstream~by;~fixation
to these cells~, which~metabolize them. This is par~
ticular]y well documented in the case of beta-glucuro-
nidase and also i~n the cas~e of; r1bonuclease B; (Arch.
Biochem. Biophys.,~1978, 188,~418, Adva~nces in;~
: ~ : : : : : :
- : ~
Enzymology, published by A. Meister, New York, 1974;
Pediat. Res., 1977, ll, 8l6).
Taken as a whole, these data show that the
rapid elimination of immunotoxins containing the A
chain of ricin can be explained by the recognition
of the mannose residues of the A chain of ricin by
the liver cells and in particular the Kupffer cells.
The studies of plasma elimination kinetics
carried out on other GPIRs, for example gelonine or
~lO~I, after intravenous injection into the animal,
have shown that, as in the case of the A chain of
ricin, the plasma level of GPIR decreases very rapidly
and very substantially after injection. Thus, in a
typical case involving rabbits, after the injection of
gelonine purified by the method described (J. Biol.
Chem., 1980, 255, 6947-6953), it appears that 93% of
the gelonine present in the bloodstream at time 0
after injection disappears in 1 hour and 99.99% dis-
` appears in 24 hours.
It is known that the oxidation of osidic
structures, including those contained in glycoproteins,
with periodate ions causes the scission of the carbon
chain wherever two adjacent carbon atoms carry primary
or secondary hydroxyls. If the two adjacent hydro-
xyls are secondary, as is generally the case in the
cyclic oses present in GPIRs, oxidation produces~t~o
aldehyde groups on the carbons between which the
scission has taken place.
It has now been found that when the carbo-
hydrate units of an antitumoral glycoprotein are mod-
ified by oxidation with periodate ions, the biological
activity of the sald glyc~oproteLn~remains~substantlally -
unchanged.
~ It has also been found,~ absolutely unexpectedly,
that lf the carbohydra~te units~of a glycoproteln which
::
.
' ,
'
`, -:
-- 6 --
inactivates ribosomes are modified by oxidation with
periodate ions, a new glycoprotein which inactivates
libosomes is obtained, the said new glycoprotein having
the dual property of retaining its biological activit-
ies and of being eliminated very slowly from the blood-
stream in vivo.
~ n in-depth biochemical study of oxidized and
native GPIRs has made it possible to demonstrate that
the oxidation of GPIRs with periodate involves ex-
cluslvely the osidic part of the GPIRs and has noaction on the sequence of the amino acids
constituting their peptide part.
These new glycoproteins which inactivate
ribosomes and have a prolonged action are referred`to
below by the symbols GPIR-La.
Finally, it has been found that when these new
glycoproteins which inactivate ribosomes and have a
prolonged action are coupled with antibodies, the con-
jugates obtained retain the biological properties
known for immunotoxins and have slow plasma elimination
kinetics.
The present invention therefore relates, by way
of new products, to structurally modified, antitumoral
glycoproteins whose carbohydrate units are modified by
oxidation with the periodate ion.
The present invention re:Lates more particularly
to glycoproteins which inactivate ribosomes, whose
carbohydrate units are modified by oxidat1on~with the
periodate ion and which have the same activity as a~nd a
longer half-life than the unmodified glycoprotein.~
The invention preferential~ly relates~to glyco-
proteins which inactivate~ribosomes~and have~a pro-
longed action and which are obtained by~treatment of
a glycoprot~e1n which~inac~tivates ribosomes, t~he~thiol
groups of which are optionally~protected,~with an
:`' ` ` `~ ,' ` .
': ~ ` ~ ' ;. ' ~
:~. . ..
: ';,, ' :
~, ` - : :
aqueous solution of an alkali metal periodate, for a
period of 0.2 to 24 hours, at a temperature of O to
15C and in the absence of light, unblocking of the
thiol groups, if appropriate, and isolation oE the
final product by known methods.
Any antitumoral olycoprotein can be modified
on its carbohydrate units by reaction with the
periodate ion in accordance with the known methods.
The glycoproteins which inactivate ribosomes
and which are used as preferred starting materials
for oxidation with periodate ions, according to the
invention, are all GPIRs, such as the A chain of ricin,
which are in themselves only very slightly cytotoxic
because they cannot fix to cells, but which, on the
lS other hand, after coupling with an antihody recog-
nizing particular cells, become highly cytotoxic to-
wards these cells once the antibody has recognized its
target.
Representative starting compounds are the A
chain of ricin, gelonine and the substance extracted
~rom Momordica charantia (MOM), as obtained by ex-
traction.
Other GPIRs which are usefu] as starting
matcrials Eor oxidation with periodate ions are as
25 follo~ls:
Dianthin 30 from Dianthus
caryophyllus
Dianthin 32 from " "
~grostin A Erom Agrostemma githago
30 Agrostin B from " "
Agrostin C from
HCI from Hura crepitans
Asparagus officinalis from Asparagus
inhibitor officinalis
The same substances produced b;osynthetically
~2~7~
by cells whose genotype has been modified for this pur-
pose are also suitable compounds.
Fragments of the above GPIRs, provided they
retain all or part of the property of inactivating
ribosomes which characterizes the GPIR from which they
are derived, can also be used as starting materials.
The A chain of native ricin in which at least
one of the thiol groups is protec`ted is a preferred
starting compound.
Recent studies have demonstrated that -the A
chain of ricin comprises 2 constituents denoted by Al
and A2, ~hich differ especially in their polysaccha-
ride units. The experiments which have been carried
out on the 2 constituents of the A chain have made
it possible to demonstrate that periodate oxidation
takes place in a similar way on the Al and A2 chains
and gives these 2 constituents identical properties
of improving the pharmacokinetics.
The preparation oE:~ the pure A chaln of rlcin
is described in U.S. Pa~tent 4 340 535. Gelonine and
~IOM are also described in p~rior art.
Protection of the~thiol~ groups of the starting~
GPIRs is desirable when the~sald;~thiol groups are those
which are to be used for coupling with the antibody.
~ ~ If other functional~groups are;used for the
coupling, for example the~;phenolic hyd~roxyl of the
tyrosines or amino groups~-~or;~carboxy~l groups of the ;
GPIRj protection can'be;carr~ied~out or not~
Blocklng~is car~ried~out~by reaction~wlth an
agent capab~le of substituting the~SH~gr~oups with a
~rad~ical~-~which ca~n;~subs~equ~ently~be~remo~ved~by,~re;duction~
or~thiol/disu`lfide~exchang~e;~ for~exa~mple~2,2~ dini~tro~
5~,5'~-dit~hiod~b~enzolc:aci~d~(DTNB)~or~3-~(p;yrid~i~n:-2-~yl~
disulfanyl)pro~pl~onlc~aci~d o~r alter~atlvely~dl~pyridyl~
2,2'-disulfide~o~r~d;ipyridyl-~4~,~4~-d~lsulflde.~In the~
- ~. : . : : ~
absence of such a treatment, the free thiols may dis-
appear during the oxidation reaction, in which case they
cannot be totally regenerated. The excess bloc}cing
agent is removed by dialysls or any other appropriate
treatment.
The periodate oxidation reaction is carried out
at an acid pH of between 3 and 7, preferably of between
5 and 6.5.
The periodate is used in excess; more particu-
larly, the concentration of alkali metal periodate isgreater than the concentration of the vicinal diols
capable of being oxidized; concentrations of 10 to 50
mM in respect of sodium periodate for concentrations of
1 to 10 mg/ml of cytotoxic sub-unit are suitable. Thè
i5 treatment, carried out at a temperature of between 0 and
15C, preferably of between 1 and 5C, and in the dark,
takes between 0.2 and 24 hours.
The reaction is stopped by the addition of a
reagent which consumes the remaining periodate t for
; 20 example an excess of ethylene glycol, and the by-pro-
ducts are removed by dialysis or by any other appro-
priate treatment. The product obtained at the end of
the reaction is isolated by conventional techniques.
If the thiol groups of the starting material
have been blocked, unblocking is effected by the known
methods, for example by reaction with a reducing agent
capable of freeing the previously blocked thiol group,
such as 2-mercaptoethanol, giving the new~glycoprotein
which inactivates ribosomes and has a prolonged action,
ready to be used for example for coupling with an anti-
body to give an immunotoxin.
In the case of the A chain of ricin, the~result-
ing new molecule (reEerred to by the symbols A~-La)
possesses the following main properties: ~
- a mole~colar weight~which 15 not signlf1cant1y~d1ffer-
,
-:
:
-
~;~4~
-- 10 --
ent from that of the native A chain. As far as it ispossible to see by polyacrylamide gradient electro-
phoresis, this modification process only produces
polymers of the protein in a very small quantity
and does not produce any degradation products.
- a proportion of free thiol groups greater than 0.7
per mol.
- an immunoreactivity to~ards rabbit antibodies in-
hibiting the A chain of ricin ~hich is indistinguish-
able from the immunoreactivity of the native A chain.- an inhibitory activity on the protein synthesis in
an acellular model which is grea~er than 50% of that
caused by an equal quantity of native A chain.
- finally, after a single intravenous administration
to ra-bbits at a dose of about 0.4 mg/kg of body
weight, the plasma level of the prolonged-action A chain
(A-la) 23 hours after injection is greater than 10% of
the level present at time zero (as against 0.015% for the
native A chain at this time) i.e. an increase in the
plasma level by a factor very much greater than 500 .
Likewise in the case of gelonine, the molecule
obtained by periodate oxidation possesses the following
main properties:
- a molecular weight which is not significantly differ-
ent~from that of the native gelonine.- an immunoreactivity towards anti-gelonine rabbit
antibodies which is indistinguishable from that of~the
native gelonine.
- finaIly, after a single intravenous administration to
rabbits at a dose~of about 0.3 mg/kg of body weight, the
plasma Ievel o~ the modified gelonine 24~hollrs after
injection is greater than 3% of the level~present at
time zero (as against 0.01%~for~ the nat:ive~gelon~in~e at
this time) i.e. an increase in~ th~e~plasma leve~l by a
factor greater ~than 200 .
:: ~
:: ::~ ::
"'
The preparation of the conjugates or immuno-
toxins from the glycoproteins which inactivate ribo-
somes and have a prolonged action is carried out by any
process suitably chosen from the range of processes
descril)e~ in IJ.S. Patent 4 340 535. If the chosen
cytotoxic sub-unit naturally contains at least one
thiol making it suitable for coupling, this group
will preferably be used by reaction with the antibody
or antibody fragment carrying an activated disulfide
group. If the chosen cy,totoxic sub-unit does not
naturally possess a thiol group making it suitable
for coupling, at least one functional group carrying
a free thiol can preferably be introduced artificially
into the said sub-unit by any known process and the
coupling can be continued as indicated above. The
introduction of the said functional group can take
place either before the oxidation step with periodate
ions, in which case it will be necessary for the thiol
radical to be blocked during the oxidation step and
then unblocked after this step, or after the oxidation
step.
This gives modified immunotoxins which have
acquired a new character as regards their pharmaco-
kinetic properties. ~lore particularly, by appropriate
modification of the cytotoxic sub-unit, it has been
possible to add to the specific cytotoxicity properties
of immunotoxins, without interfering with them, a new
property which is just as intrinsic, namely the capa-
city to show slow plasma elimination kinetics.
The examples which follow;provid,e a c~learer
understanding of the~invention without l~imiting its
scoye.
Exam~
Oxidation of the methylat~ed A chain in whïch the SH
groups are blocked w~ith N-ethy~lma~leimide.
:
~'~-: -
,
. .' ~
~ - .:, ` ' ~ .:. ' ' :
- 12 -
I - Preparation of the correctly functionalized A chain
of ricin
1) Hexamethylation of the A chain
The methylation reaction is carried out at 0C,
with stirring, :in 0.2 M borate buffer of pH 10, by the
method of Means and Feeney (Bi.ochemistry 7, 2192
(1968)). 20 mg of tritiated borohydride (containing
9.5 mCi/mmol) are add.ed to 35 ml of A chain (3 mg/ml),
followed by 350 microliters of 6% formaldehyde (added
in five 70 microliter portions spread over a period of
30 minutes).
The excess reagent is removed by continuous
dialysis against 125 m~l phosphate buffer of pH 7 (40 l
at 300 ml/h). After dialysis, the protein solution is
centrifuged. 36.5 ml of hexamethylated A chain con-
taining 2.6 mg/ml are collected.
2) Blocking with N-ethylmaleimide
The natural Sl-l of the hexamethylated A chain
is blocked by the method described in Methods in
En~ymology 11, S~l (1967). To do this, the A chain of
ricin obtained i.n the previous step is incubate(l for
2 hours at 30C ln the presence of 20 equivalents of
N-ethylmal.eimide per mo:L of A chain. The excess re-
agent is removed by continuous d:ialysis against 125
m~ phosphate buffer of pH 7, which is renewed for 20
hours at a rate of 500 ml/hour.After concentration, there is obtained
l3 ml of a solution of ~ chain of ricin containin~ 7 m /ml ~n~
no longer possessing thiol groups which can be deter-
- mined by ELLMAN's reagent. The product thus obtained
is subsequently called hexamethylated li chain (NE~I).
3) Periodate oxidation
6 ml of the solution of hexamethylated A chain
(NEM) obtained above are treated with NaI04 (12.8 mg)
for 40 minutes in the dark, at pH 4.5 and at 0C. The reaction is
stopped by the addition of 600 microliters of ethylene
- 13 -
glycol and the reaction medlurn is dialyzed contlnuously
against 0.1 M carbonate buffer of pH 10 (20 h at 500
ml/h)-
II - Enzymatic activity of the prolon~ed-action A chain.
measured on an acellular model
The fundamental biological property of the A
chain of ricin is to inhibit protein synthesis in cells
by degradation of the ribosomal sub-unit 60S.
The in vitro protocol involves the use of appro-
priately complemented, subce]lular fractions of ratliver capable of incorporating 14C-phenylalanine in
the presence of an artificial messenger RNA: poly-
uridylic acid.
The procedure employed for preparing the sub-
cellular fractions and measuring the incorporation of
C-phenylalanine is an adaptation of the method des-
cribed in Biochemica Biophysica Acta 1973, 312, 608-
615, using both a microsomal fraction and a cytosol
fraction of the rat hepatocytes. The sample contain-
ing the A chain is introduced in the form of a solutionappropriately diluted in a 50 mM Tris HCl buffer of
pH 7.6 containing 0.2% of 2-mercaptoethanol and 15
micrograms/ml of bovine serum albumin.
The count data are used to calculate, relative
to a control medium without inhibitor, the percentage
inhibition of the incorporation o 1 C-phenylala~nine
into the proteins for each reaction medium containing
A chain of ricin.
The inhibitory activlty was determined. An
IC50 of 2.7 ~10 l~mol/l i9 observed~for the~oxldized
A chain~ The IC50 of thç control A chain in~the~ex-
periment is 1.03-10 l mol/l; therefore,~the modifica-
tion does not cause a los~s of a~ct1v~ity of the A~ch~in.
E~ample 2:
This example demonstrates the slow elimination
, ~
:
: - , .
- 14 -
of the A chain of ricin modified with sodium periodate,
after intravenous injection into the animal.
I - Modification of the A chain of rlcin ~lith sodium
periodate
l) Blocking of the natural SH with DTNB
The A chain of ricin was prepared and purified
in the manner indicated in U.S. Patent 4 340 535. 20
equivalents oE a solution of 2,2'-dinitro-5,5'-dithio-
dibenzoic acid (~TNB), i.e. 385 microliters of a 0.1
~I solution of DT~B in a 125 mM phosphate buffer of pH
7 (this solution is brought to pH 7 with sodium hydro-
xide), are added to 10 ml of a solution of A chain of
ricin containing 5.6 mg/ml (with 0.84 thiol group per
A chain) in PBS buffer (a buffer 20 mM in respect of
phosphate and 150 mM in respect of NaCl, of pH 7).
Incubation is left to proceed for 20 minutes at 20C.
The solution is then dialyzed against PBS buffer at
4C to give 53 mg of A chain blocked on the thiol group,
as a solution containing 5 mg/ml.
2) Periodate oxidation of the blocked A chain
120 microliters of a 0.5 M solution of sodium
periodate in ~ater are added to 6 ml of a solution con-
taining 5 mg/ml of blocked A chain, brought to pH 6
~ith l M acetic acid. Incubation is left to proceed
for 16 hours at 4C in the dark. The oxidation reaction
is stopped by the addition of 620 microliters of a 1 M
aqueous solution of ethylene glycol. After incubation
for 15 minutes at 20C, the reaction medium is dialyzed
at 4C against PBS buffer. The periodate oxidation
produces a slight precipitate of protein, which is re-
moved by centrifugation at lO,000 x g for 30 minutes.
This gives 24 mg of oxidized blocked A chain at a con~
centration of 3.4 mg/ml.
3~ Unblocking of the thiol groups
2-Mercaptoethanol is added as a reducing agent,
' : :
. ` ` ' , ~ ~ ` ~ :
:
~ - , .
, ~ ~
,
7~
- 15 -
at a final concentration of 1%, to 6 ml of oxidized
blocked A chain containing 3.4 mg/ml in PBS buffer.
Inc~lbation is left to proceed for 1 hour at 20C. The
solution is then dialyzed against PBS buffer at 4C.
This gives 19 mg of oxidized A chain at a concentration
of 2.8 mg/ml.
Using the DTNB technique (Methods in Enzymology,
1972, 25, 457 (Academic Press)), it is determined that
- the modified A chain obtained has 0.70 free thiol group
per mol. The molecular weight of the modified A chain
is 30,000+3,000, determined by polyacrylamide gradient
electrophoresis in the presence of sodium dodecyl-
sulfate.
The previously obtained preparation of A chain
in which the polysaccharide units have been oxidized
~as studied for its enzymatic activities in the in-
hibition of protein synthesis and for its pharmaco-
kinetic properties.
II - Enzymatic activit~ of the prolonoed-action A chain,
measured on an acellular modeI
The inhibitory activity ~as determined by the
technique described in example 1. An IC50 of 3 10 ]
mol/l is observed for the oxidized A chain. The IC50
of the control A chain in the experiment isl~2 10 l
moi/l; therefore, the modification-does not cause a
loss of acti~ity of the A chain.
III - Pharmacokinetic~properties of the prolonged-action
A chain (A-La)
The A chain is administered to rabbits by means
of a single injection into a vein in the ear. The
quantity of A chain~injected corresponds to~0.415 mg/kg.
Blood samples are taken at ~intervals on heparin. ~The
plasmas are~analyz;ed w1th the ald of a radioimmunometric
test designa;ted bel~ow~b~y~ the~abbreviation RIM-1. ~ ~
This technique~has the advantage of d~etermining
; :, ~:
. :
, :
..
~: '
7~
- 16 -
the A chain wi-thout modifying it. This deter~ination
is carried out in microtitration plates (for example:
"NUNC-TSP screening s'ystem" from Poly Labo Block
France), the lid of ~hich carries hyperabsorbent spikes
which dip into the cavities in the base. These spikes
constitute the solid phases. Ewe antibodies inhibiting
A chain of ricin (designated below by the abbreviation
Acl), purified by affinity chromatography, are absorbed
on the solid phases. For this purpose, 200 microliters
of a solution of Acl containing 10 micrograms/ml in PBS
buffer are divided up into the cavities. The spikes are
brought into contact firstly with the solution of AcI
for 24 h at 4C and'then with fetal calf serum for 3 h
at 20C in order to saturate all-the fixation sites.
The saturated immunoabsorbent is then brought into con-
tact for 3 h at 20C with the plasma samples to be de-
termined at different dilutions, or with solutions of A
chain o'f known concentrations in order to establish the
calibration curve. After washing with a PBS buffer, the
immunoabsorbent is brought into contact for 2 h at 20C
with the ewe antibodies inhibiting A chain of ricin,
which have been purified by affinity chromatography and'
radiolabeled (designated below by the abbreviation Ac2).
The radiolabeling of the Ac2 is effected with iodine-
125 in the presence of chloramine T by the method ofGreenwood and Hunter (Biochem J., 1963, 89, 114); the
specific activity of the radiolabeled A'c2 antibodies is
5 to 10 microcuries/microgram. 10 cpm of radiolabeled
Ac2 antibodies are introduced as 200 microliters into a
PBS buffer containing 0.1% of bovine serum albumin.
After washing in PBS~buffer, the spikes are detached and
the quantity of bound Ac2 is~measured by counting the
radioactivity. The concentration of A chain in the
samples to be determinéd is measured by reference to the
calibration curve established by introducing the A chai'n
.
.
~ ' ~
. '. : ' . ' '
7~
- 17 -
at different known concentrations. When prolonged-
action A chain is injected into the animal, this same
prolonged-action A chain is used to establish the
corresponding calibration curve.
The values of the concentration of A chain in
the blood plasma measured by this technique are repro-
ducible and reliable. The detection threshold is 1
nanogram/ml. A study of the reproducibility within
and bet~een experiments gives coefficients of variation
of less than 10% for concentration values within the
range from 1 to 200 nanograms/ml.
The results of these experiments are represented
in the form of curves in which the time, expressed in
hours, is plotted on the abscissa and the plasma con-
centration of the product measured, recorded in pe~ centof the theoretical plasma concentration at time zero, is
plotted on a logarithmic scale on the ordinate. This
value, called the "relative plasma concentration" (RPC),
is calculated using the following expression:
concentration measured at time t
x 100
RPC =
quantity injectedlplasma volume
The plasma volume is considered to be equal to
36 ml/kg of the animal's body weight.
'25 Figure 1 shows the plasma elimination curve, as
a function of time, for the A chain of native ricin in-
jected intravenously. This curve (curve l) has two
phases: in the first phase~, the product d1sappears very~
rapidly from the blood'stream since only 0.1% of~the dose
administered remains in the plasma 3 hours after injec~
tion. In the second phase, the decrease is slower.
When~the A ;chain has been;oxidized on i~ts poly-
saccharide units,~the eliminati~on profile is profo'undly
modlfled: the first-eliminaéion phase -;which~is res~
ponsible for the disappear~ance;~of~the majority of the;
: : : ::
., . :
: ~ , :
7~
product - is practically suppressed, which leads to a
considerable increase in the plasma leve]s of A chain.
Twenty hours after injection, the concentration of the
oxidized A chain is 600 times greater than in the case
of the unmodified A chain (curve 2).
E~ample 3:
This example demonstrates the effect of
periodate o~idation on the pharmacokinetic properties
of the A chain blocked with NEM.
1) - Modification of the A chain of ricin
a) Blocking of the natural SH with N-ethylmaleimide
40 ml of an aqueous solution of A chain of ricin
containing 8 mg/ml (i.e. 4.1 micromol of A chain) are
treated with an aqueous solution of 2-mercaptoethanol
so that the final concentration is 1 per cent.
The solution is left to stand for one hour and
then dialyzed continuously against 125 mM phosphate
buffer of pH 7, which is renewed for 40 hours at a rate
of 300 ml/hour. Using Ellman's method, 0.9 equivalent
oE SH was determined per mol of A chain of ricin.
This SH group is blocked with N-ethylmaleimide
by the method described in Methods in Enzymology, ll,
541 (1967). To do this, the A chain of ricin obtained
in the previous step is incubated for 2 hours at 30C
in the presence of 20 equivalents of N-ethylmaleimide
per mol of A chain.~ The excess reagent is removed by
continuous dialysis against 125 mM phosphate buffer of
pH 7, which is renewed for 20 hours at a rate of 500
ml/hour. This gives 35 ml of a solution of A chain of
ricin containing 7 mg/ml and no longer possessing thiol
groups which can be determined by Ellman's reagent. The
product thus obtained is subsequently called A chain
(NE~I).
b) Periodate oxidation of~the A chain (NEM)
Periodate oxidation of the A chaln (NE~I) is~
, :
~:
' ~' ~ ' ' :
:
7~
-- 19 --
carried out using the procedure indicated in example 2.
2) - Properties of the oxidized A chain (NEM)
a) Enzymatic activity of the oxidized A chain (NEM)
The inhibitory activity on the protein syn-
thesis was determined using the procedure described inexample l. The enzymatic properties are found to be
maintained ~ith an IC50 of 4.3-10 10 mol/l for the
oxidized A chain (NEM).
b) Pharmacokinetic properties of the oxidized A chain
(NEM)
The oxidized or unoxidized A chain (NEM) is
administered to rabbits by a single injection into a
vein in the ear. The quantity of A chain injected
corresponds to 0.100 mg/kg. The plasma samples collect-
ed at time 23 h are analyzed using the immunometric
test RI~-1 as described in example 2. The results are
shown in the table below:
I
Plasma concentration 23 h after
injection:
_
A chain (NEM) 0.01%
20 Oxidized A chain ~ 8%
(NEM)
Twenty~three hours after injection, the con-
centration of the oxidized A chain (NEM) is 800 times
greater than in the case of the unmodified A chain (NEM)~
Example 4:
This example demonstrates the importance of
the duration of the oxidative treatment on the pharmaco-
kinetic properties of the oxidized A chain.
Six preparations of oxidized A chain are pre-
pared using the procedure lndicated in example~2, exceptfor the duration of the sodium periodate treatment. The
treatment times are as Eollows: zero (reaction stopped
.
. ~ :
.
- 20 -
immediately with ethylene glycol), 20 minutes, 40 min-
utes, 2.5 hours, 4 hours and 18 hours.
These various preparations are injected into
rabbits and the relative plasma concentration of the A
chain is measured after 23 hours by the same procedure
as in example 1.
The results are shown in figure 2. These results
indicate that 1) the increase in the plasma level of the
A chain is indeed due to periodate oxidation because,
when the reaction is stopped immediately, the plasma
` concentration of A chain is-identical to that obtained
for the native A chain, and 2) it is necessary for the
duration of this reaction to be relatively long in order
to obtain optimum effects.
Example 5:
This example demonstrates the importance of the
duration of the oxidative treatment on the pharmaco-
kinetic properties-of the methylated A chain blocked
with NE~I.
1) - Preparation of the functionalized A chain of ricin
a) Blocking oE the natural SH of the A chain with N-
ethylmaleimide
The natural SH of the A chain is blocked with
N-ethylmaleimide by the same procedure as t`hat described
~5 in example l.
b) Methylation of the A chain
The methylation reaction is carried out at 0C,
with stirring, in 0.2 M borate buffer of pH 10, by the
method of Means and Feeney (Biochemistry~7~j 2192, 1968).
38 mg of tritiated borohydride (containing 47 mCi/mmol)
are added to 65.5 ml of A chain (N~EM) (3 mg/ml),;~follo~-
ed by l ml of 6% formaldehyde~added in~five~200 mlcro-
liter portions spread over a period~of 30 minutes~
The excess reagent is r~emoved by discontinuous
35 dialysis agains;t 125 m~ phosphate buffer of~pil 7~(40 ml).
~: :
. . ~ :
- ' ' '
.. ~ .
7~
- 21 -
After dialysis, the protein solution is centrifuged.
63 ml of methylated A chain containing 3 mg/ml are
collected.
- c) Periodate oxidation
Six preparations of methylated A chain (NEM)
are oxidized using the procedure indicated in example
l,except for the duration of the sodium periodate
treatment. The treatment times are as follows: ~ero
(reaction stopped immediately with ethylene glycol),
lO minutes, 40 minutes, 2.5 hours, 4 hours and 18 hours.
These various preparations are injected into
rabbits and the relative plasma concentration of the A
chain is measured after 23 hours by the same procedure
as in example 2.
The results are shown in figure 3, curve 2.
These results indicate that, as for the A chain (curve
1 ) :
l. the increase in the plasma level of the methylated
A chain (NEM) is indeed due to periodate oxidation
because, when the reaction is stopped immediately, the
plasma concentration of methylated A chain (NEM) is
identical to that obtained for the A chain ;
and
2. it is necessary for the duration of this reaction to
be relatively long in order to obtain optimum effects.
Example 6:
This example demonstrates that, when carried
out separately on the two constituent molecular variants of the A~
chain (Al chain and A2 chain), the oxidation reaction
produces effects on each oE the two isomers which~are
analogous to those described in example 2 for the
chain of ricin.
l) - Separation of the Al and A2 chains
28 ml of; A cha~in~containing lO.~mg/ml~(309 mgl
in 125 mM phosphate~buf~fer of pU~7.0~are deposited o~n~a
c ~ : :
. .
,
- 22 -
column of 112 ml of concanavalin A/sepharose, equili-
brated in the same bufrer. The A1 chain is obtained in
the first peak by washing with the same buffer; the A2
chain is eluted with 0.1 M borate buffer of pH 6.0,
which is 0.5 M in respect of NaC] and 0.1 ~ in res?ect
of alpha-methylmannoside.
This gives 184 mg of Al chain and 103 mg of A2
chain.
The Al and A2 chains are concentrated by ultra-
filtration under nitrogen pressure; the A2 chain isdialyzed against 125 m~l phosphate buffer of p~l 7Ø
Analysis of the A chain by acrylamide gel
gradient electrophoresis with SDS shows the presence of
2 bands of different intensity, corresponding to molec-
ular weights of 30,000 and 33,000. The Al chain corres-
ponds to the band of stronger intensity and of MW
30,000 and the A2 chain corresponds to the band of weak
intensity and of ~IW 33,000.
2) - ~lodification o the Al and A2 chains of ricin with
sodium periodate
This modification is effected as described in
example 2. The preparations of A chain in which the
polysaccharide units have been oxidized were studied
or their enzymatic activities in the inhibition of
protein synthesis and for their pharmacokinetic prop-
erties.
3) -Enzymatic activities of the prolonged-action Al
and A2 chains, measured on an acellular model
The inhibitory activity was determined as des-
cribed in`example 1. The IC 0 observed is equal to2.1-10 10 mol/l and 2.1.10 1~ mol/l for the~oxidized
Al and A2 chains respectively. The IC50 values of the
native Al and A2 chains, which a~e the controls in the
experiment, are 1.9-10 10 mol/l and 1 10-~1 mol/l
respectively. Thereeore, the modlf1cation of the
:
:
;,, '~
- 23 -
separate ~ariants of the A chain does not cause a loss of
their enzymatic activity.
4) - Pharmacokinetic properties of the prolonged-action
Al and A2 chains (Al-La, A2-La)
The Al or Al-La chain or the A2 or A2-La chain
is administered to rabbits by a single injection into a
vein in the ear (415 micrograms of A chain/kg). The
plasma samples collected after 20 hours are analyzed
with the aid of the immunometric test RIM-l (see
e~ample 2). The results are shown in the table below.
The values for the A and A-La chains are indicated by
way of comparison.
Relative plasma concentration
20 hours after injection
.
A chain 0.012%
15 Oxidized A chain (A-La) 10%
Al chain 0.02%
Oxidized Al chain (Al-La) 10%
A2 chain 0.04%
Oxidiæed A2 chain (A2-La) 14%
Twenty hours after injection, the concentrations
of Al-La and A2-La are respectively 500 and 350 times
greater than in the case of Al and A2.
Example 7:
This example describes the biochemical character-
istics of the A chain and its variants,the Al chain and
A2 chain, in the native form and in the oxidized form.
The A chains used in these studies are prepared
as described in examples 2~and 6.
I - Carbohvdrate compo~iti;ons
The carbohydrate~compositions of these~proteins
are determined by~gas chromatographic analyses us1ng
:
'' ~' ~ '
:, , '
~ .
- 24 -
Clamp's method (in Glycoproteins: their composition,
structure and function (edited by A. Gottschalk), volume
5 A, p 300-321, Elsevier Publishing Co., Amsterdam,
London, New Yorlc).
The results obtained are collated in the two
tables below.
Percenta~e composition
Chains Total carbohydrates, %
Native A 5.58 + or - 0.5
10Al 4.54 + or - O.S
A2 6.24 + or - 0.5
Oxidiæed A 2.27 + or - 0.5
Al 2.07 + or - 0.5
A2 3.33 + or - 0.5
Molar composition
(On the basis of a molecular weight of 30,625, the
results are given with an average precision of + or -
0.5 residue per molecule)
_
Chains
. . I
20MonosaccharidesNative Oxidized
_
A Al A2 A Al A2
N-Acetylglucosamine 1.89 1.48 2.15 1.74 1.50 2.37
~lannose 4.6 3.40 5.2 I.43 1.29 2.26
Fucose 1.37 1.41~ 1.52 O O
Xylose 1 6 1.48 1.67 0.~6 0.48 0.6
: ~ ~
These results prove that~periodate~oxidation
has destroyed part of the~sugars o-E the~A chai~n. Per
molecule of A chaln9~there is~an~a~vqrage decrease~of
::
~ ~ : . ' :
:::: : : :
. ~
,''"~
:
- 25 -
3.17, 2.11 and 2.9~ mannose residues, the fucose residues
have completely disappeared and there is an average
decrease of 1.24, 1 and 1.05 xylose residues for the A,
Al and A2 chains respectively. The N-acetylglucosamine
residues are only slightly degraded.
II - N-Terminal sequence
The sequence of the N-terminal amino acids oE
the A chain and its Al and A2variants, in the native form
and in the o~idized form, was established with a protein
sequencer by the procedures known to those skilled in the
art. The results obtained are collated in the table
below.
Chains Sequenc of the 9 N-terminal amino acids
Native A Ile-Phe-Pro-Lys-Gln-Tyr-Pro-Ile-Ile
A1 Ile-Phe-Pro-Lys-Gln-Tyr-Pro-Ile-Ile
A2 Ile-Phe-Pro-Lys-Gln-Tyr-Pro-Ile-Ile
. .
O~idized A Ile-Phe-Pro-Lys-Gln-Tyr-Pro-Ile-Ile
Al Ile-Phe-Pro-Lys-Gln-Tyr-Pro-Ile-Ile
A2 Ile-Phe-Pro-L~Js-Cln-Tyr-Pro-Ile-Ile
. . .,
It is found that the sequences of the 9 N-terminal
amino acids of the native and oxidized A, Al and A2
chains are strictly identical to one another, which
demonstrates that the oxidative-treatment leaves the
protein chain intact. It is also found that the sequence
of the 9 N-terminal amino~ acids~ of the A chains~is
strictly identical to that previously desc~ribed by
Funatsu for the A chain of ricin (Agr~ic. Biol. Chem.,
(1979), 43, 2221).
III - Affinity on concanavalin A/sepharose ~
- The A chain, the A chain blocked with DTNB
[A(DTNB)] and the A(DTNB)-La, Al(DTNB), Al(DTNB)-La,
A2(DTNB) and A2(DTNB)-La chains are tested~by their
capacity to flx to concanav~a11n~ A~/sepharose. 1~ml~oE
:
. . . :
: ~: .
. ~ ~
: . : ~ . :
,,
-
: , ~ " ~ :
- 26 -
a solution of A chain containing about 1 mg/ml is
deposited on a column of 1 ml of concanavalin A.
Chromatography is followed by measurement of the optical
density at 280 nm. After washing with 125 mM phosphate
buffer of pH 7.0 until the first peak which is not
retained by the concanavalin A has returned to the
base line, the column is washed ith 0.1 ~I borate
buffer of pH 6.0, which is 0.5 ~l in respect of NaCl
and 0.1 ~l in respect of àlpha-methylmannoside. The
results, expressed as a percentage of the optical den-
sity at 280 nm are summari~ed in
the table below.
.
Chains % not retained % eluted by alpha- Total
by the con A methylmannoside (%)
A 53 27 80
_
A(DTNB) 66 11 77
A(DTNB)-La78 5 83
_
Al(DTNB) 80 6 86
Al(DTNB~La 73 0.4 73.4'
A2(DTNB) 25 70 95
A2(DTNB~La 64 9 73
_ _
It is known that concanavalin A has an affinity
for glycoproteins with terminal mannoses. It is foun(l
that the A chain which contains such residues can bind
to con A. This is particularly clear in the case of the
A2 chain, which is the isomer more richly substituted
with mannose. It is also found that the oxidative
treatment destroys this affinity, which is coherent with
the destruction of the sugar residues by a treatment of
this type.
I~- Determination of _he E l~o
The absorption coefficient ~t 280 Dm (E 1~is
:
- : .,. '
,
~, , : ~ .
' ' :
g~
- 27 -
the optical density at 280 nm of a solution containing
1 mg/ml, in which the protein concentration is deter-
mined by the FOLIN test with a standard range of bovine
serum albumin.
The results are summarized in the table which
follows:
.
A chain 0.6~
A(DTNB) chain 1.12
A(DTNB)-La chain 1.04
.
Al(DTNB) chain 1.07
Al(DTNB)-La chain 1.02
.
A2(DTNB) chain 0.95
A2(DTNB)-La chain 0.95
Blocking of the thiol group of the chain with
DTNB produces a substantial increase in the absorption
at 280 nm, due to the introductlon of the nitrobenzoyl
group.
After oxidation, no significant variation in
t`he absorption at 280 nm is observed, demonstrating
that oxidation has not affected amlno acids responsible
for the absorption at 280 nm.
Y - Isoelectric focusing
Analysis of the A chain by isoel~ectric~focusing
produces a set of bands with isoelectric points (pI)
which are between 7.5 and 8.0 and are identical for the
A, Al and A2 chains.
Blocking of the cysteine of the A chain with
- DTNB causes a widening of~the-bands towards the~acid
region; freeing of thé~ cysteine~-with mercaptoethanol
brings these bands ba~ck~to the location~o~f thé natlve
A chain.
Thus, a compari~son o;f~the isoelec~tric points of
: :
: .
~L~J7~
- 28 -
the native and oxidized A, A1 and A2 chains shows that,
in the ahsence oE a blocking agent, all the bands charac-
teristic oE the A chain are transferred by 0.5 pH unit
towards acid pH values. This transfer takes place with-
out overlapping of the pH regions of the native andoxidized A chains, which seems to indicate that all the
A chain molecules are affected by oxidation.
Example 8:
This example demonstrates 1) the rapid elimin-
1`0 ation of native gelonine, and 2) the slow elimination ofgelonine modified with sodium periodate, after intra-
~enous injection into the animal.
I - Modification of gelonine with sodium periodate
The gelonine was prepared and purified from
15` Gelonium multiflorum by the method which has been des-
cribed (J. Biol. Chem. (1980) 255, 6947-6953). The
oxidation reaction is carried out under the same con-
ditions as those described for the A chain of ricin in
example 2, except that the step in which the thiols are
blocked with DTNB is omitted.
In fact, as the coupling of gelonine with the
antibody is not generally performed using natural thiol
groups of the gelonine, the thiol groups will be intro-
duced artificially, after the oxidation step, by the
technique described in Cancer Res., 1984, 44, 129-133.
21 microliters of a 0.5 M solution of sodium periodate
in water are added to 1 ml of a solution containing 3
mg/ml of gelonine in PBS buffer,~brought to pll 6 with ;
1 M acetic acid. Incubation is~left ~:o proceed for 16
hours at 4C in the dark.~ The reaction is~stopped by
the addition of 105 micraliters of a 1 M a~queous~sol-
ution of ethylene glycol. ~After incubation for 15~min-
utes at 20C, the reaction medium is dialyzed at 4C
against PBS bufer. A~fter centr~1fu~gation at IO,OOQ x g
for 30 minutes, this givcs 2.9~mg oE~o~idi~zed gelonine
: ~ I
- : ~
- 29 -
at a concentration of 2.5 mg/ml.
Like the i~ chain of ricin, the fundamental prop-
erty of gelonine is to inhibit protein synthesis in
eucaryotic cells by degradation of the ribosomal sub-
,unit ~0 S (Biochem. J. (1982) 207, 505-509). In th,e
case of gelonine too, the modification due to periodate'
oxidation does not cause a loss o:E activity.
II - Pharmacokinetic properties of prolon~ed-action
oenonine
~ative gelonine or gelonine modified by the
procedures explained above is administered to rabbits by
a single injection into a vein in the ear. The quantity
of gelonine injected is between 0.3 and 0.4 mg/kg.
Blood samples are taken at intervals on heparin. The
plasmas are analyzed with the aid of a radio;mmunometric test
designated below by the abbreviation RIM-2.
This test is performed by the same technique as
used for the test RI~-l, except that the solution ~cl
here is a solution of anti-gelonine rabbit antibodies
purified by aEfinity c~hromatography, the ~c2 antibodies
beillg the same antibodies radiolabeled. The radio-
labeling procedure is ide~ntical to that described forthe technique RIM-l. The concentration of native
gelonine or modified gelonine in the samples~to be
determined is measured by reference to a calibration~
curve established by introducing native or modified
gelonine at different known concentrations. The.test
RIM-2 has the same reliabi~lit~y and reproducibility~ :
- characteristics as described fo:r the technique~RIM
The results of these'experlments arc represente~d~in the
same way~as for~the A chain, of rici.n in example 2.:
Figure 4 shows~:the~plasma elimination curves, .
. as a :function of time,~for native~gelonine and~oxidized
gelonine, injected intra:venously.: The native~ge~lonine,
like the A chain of~native~ricin, d~sappears very:ràpldly
-
:
'
7~
- 30 -
from the bloods-tream since 99.99% of the gelonine present
in the bloodstream disappears in 24 hours (curve 1).
When the gelonine has been oxidized on its polysaccharide
units, the elimination profile is profoundly modified-
?4 hours after injection, the concentration of theoxidized gelonine is 300 times greater than that of the
native gelonine (curve 2).
Thus, as for the A chain of rlcin, these results
prove that periodate oxidation has modified the sugars
involved in the recognition process responsible for the
elimination of the gelonine, to the point of preventing
this recognition.
Example 9:
This e~ample demonstrates:
1. the rapid elimination of GPIR MOM extracted
from Momordica charantia,and
2. the slow elimination of GPIR MOM modified
with sodium periodate,
after intravenous injection into the animal.
1) - Modification of GPIR MOM with sodium periodate
The GPIR MOM was prepared and purified from the
endosperm of Momordica charantia seeds by the method
which has been described (Biochem1cal Journal (1980),
186, 443-452). The pharmacokinetic properties of native
or modified MOM were established using radioactive GPIR
MOM~ The MOM is radiolabel~ed on the tyrosines~with
radioactive iodine-125 in the-presence of chloramine T.
10 microliters of a solution of radioactive iodine-125
containing lOO ~Ci/ml, and 30 microliters o~ a solution~
of chloramine T containing 2.5 mgiml in water, are added
to 100 microl1ters of~a~;solut1on~cont~ain~ing 1~m~g~/ml~of
MOM in PBS buff~er. The~reaction~is left to~proceed for
one minute at~ambient;te;mperature. The reaction is
st~opped by the addltion;of~40~0 micro~liters of a solution
o~ sodium metabisulf1te~¢ontainlng~0~.5 mg/ml.~The re-
-.
, "
... , ": . ; , -
~: .. :., .,, ,. :
7~
- 31 -
action medium is chromatographed by gel Eiltration on a
column of Sephadex*G25 with PBS buffer containing 0.1%
of gelatine in order to separate the radiolabeled protein
from the unreacted iodine. After centrifugation at
10,000 x g for 30 minutes, this gives 80 micrograms
of radiolabeled MOM at a concentration of 0.04 mg/ml,
The oxidation reaction is carried out under ~he
same conditions as described for the A chain of ricin in
example 2, except that the step in which the thiols are
~0 blocked with DT~B is omitted and the concentration of
protein is 125 times smaller (40 ~g/ml). 20 micro-
liters of a 0.5 M solution of sodium periodate in water
are added to 1 ml of a solution containing 0.04 mg/ml
of radiolabeled MO~I, brought to pH ~ with 1 M acetic
acid. Incubation is left to proceed for 16 h at 4C in
the dark. The reaction is stopped by the addition of
100 microliters of a 1 M aqueous solution of ethylene
glycol. After incubation for 15 minutes at 20C, the
reaction medium is dialyzed at 4C against PBS buffer.
After centrifugation at 10,000 x g for 30 minutes, this
gives 32 micrograms of oxidized MOM at a concentration
of 0.021 mg/m].
The new molecule thus obtained has a molecular
weight which is not significantly different from that
~5 of the native MOM. As far as it is possible to see by
polyacrylamide gradient electrophoresis after development with
coomassie blue or radioautography, the modification
process only produces protein polymersin a very small
quantity and does not produce any degradation product.
2) - Pharmacokinetic properties of prolon~ed-action MOM
The radiolabeled MOM, whether or not oxidized
by the procedures explained above, is administered to
rabbits by a single injection in a vein in the ear.
The quantity of MOM injected is between 3.50 and 3.5S
micrograms/kg, Blood samples aFe taken at intervals on
* - Trademark
.~` ..
.~ - .
.,:
- : ,
~, "
,, .. ., ~ -
- 32 -
heparin. The plasmas (200 microliters) are incubated
with trichloroacetic acid (TCA) (200 microliters at a
concentration of 25%) for 30 minutes at 4C. After
centrifugation, the radioactivity contained in the sedi-
ment which can be precipitated by the acid is deter-
mined. This method of analysis makes it possible to
measure the plasma level of the intact ~IOM molecules,
and any low-molecular degradation products whlch cannot
be precipitated by TCA are not taken into account.
The results of these experiments are represented
as the percentage of the initial radioactivity remain-
in8 in the bloodstream as a function of time. This
value, which is called the "percentage of the initial
plasma radioactivity" (% IPR), is calculated using the
following expression:
r x PV x 100
% I.P.R. =
0.2 x R
where:
r = radioactivity measured at time t in 0.2 ml
of plasma,
R = total radioactivity injected,
PV = plasma volume (considered to be equal to
36 ml/kg of the animal's body weight).
The plasma elimination curves, as a function of
time, for the oxidized or unoxidized MOM after intra-
venous injection are shown in figure 5. The MOM, like
the A chain of native ricin, disappears very rapidly
from the bloodstream since 99.9% of the MOM present~in
the bloodstream disappears in 8 hours (curve 1). When
the~IOM has been oxidized on its sugar residues, the elimin-
ation rate is reduced (c~rve 2): 8 hours after injection,level of oxidized MOM is 60 times grea~ter than that of
the unoxidized MOM. These results prove that~periodate
oxidation has modified the sugars~involved in~the
recognition process responsible for the rapid e~limin-
:
:
,
- 33 -
ation of the MOM.
Example 10:
This example demonstrates:
1. the rapid elimination of GPIR Dianthin ex-
tracted from Dianthus caryophyllus,and
2. the slow elimination of GPIR Dianthin mod-
ified with sodium periodate,
after intravenous injection into the animal.
1) - Modification of Dianthin 30 with sodium periodate
The Dianthin 30 was prepared and purified from
the leaves of Dianthus caryophyllus by the method which
has been described (Biochemical Journal (1981), 195,
339-405). The pharmacokinetic properties of oxidized
or unoxidized Dianthin 30 were established using radio-
active Dianthin. The io~ination and oxidation reactions
are carried out under the same conditions as described
for MOM in example 9.
The new oxidized Dianthin molecule thus obtained
has a molecular weight which is not significantly differ-
ent from that of the native Dianthin 30.2) - Pharmacokinetic properties of prolon~ed-action
Dianthin 30
The oxidized or unoxidized radiolabeled Dianthin
is administered to rabbits by a single injection into a
vein in the ear. The plasma level of Dianthin is meas-
ured by the same procedure as described for MOM in
example 9. Figure 6 shows the plasma elimination curves,
as a function of time, for the oxidized Dianthin (curve
2) or unoxidized Dianthin (curve 1). Dianthin 30, like
MOM, disappears very rapidly from the bloodstream since
99.9% of the quantity i;nitially present disappears in
2 hours. On the other hand, when the Dianthin 30 has
been oxidized on the carbohydrate residues, the~elimin-
ation kinetics are slowed down considerably:~ 2 hours
after injection, the Dlanthl~n level is 80 times greater
,
,
.. .
- 34 -
than that of the unoxidized Dianthin. The level of
o.Yidized Dianthin remains high beyond 24 hours (3 % of
the initial value at 24 hours).
Here again, these results prove that periodate
oxidation has modified the sugars involved in the recog-
nition process responsible for the rapid elimination of
the Dianthin.
Example ll:
Conjugate obtained by the reaction of an antibody inhib
iting human T cells (an antibody directed against the
anti~en T65), substituted by activated disulfide groups,
~ith the oxidized A chain of ricin.
a) Antibody inhibiting human T cells (or antibody T101)
This antibody was obtained by the method des-
cribed in Journal of Immunology, 1980, 125(2), 725-737.
b) Oxidized A chain of ricin: The A chain of ricin was
prepared in the manner indicated in example 2.
II) Rctivated antibodies inhibiting human T cells
20 microliters of a solution containing 60.3
mg/ml of 1-ethyl-3-dimethylaminopropyl-3-carbodiimide
are added to 100 microliters of a solution containing
~0 mg/ml of 3-(pyridin-2-yldisulfanyl)propionic acid
in tert.-butanol, and the mixture is left for 3 minutes
at ambient ~emperature. 68 microliters of the solution
thus obtained are added to 2 ml of an antibody solution
containing 8.9 mg/ml in PBS buffer. The mixture~is
stirred ~or 15 minutes at 30C and~then diaIyzed against
PBS buffer at 4C. ~fter dialysis, the protein solution
is centrifuged to give 15 mg of~activated antibody at~
a concèntration of 7.9 mg/ml. By spectrophotometric -
analysis at 343 nm of the pyridine-~2-thione released by
exchange with 2-mercaptoethanol, it is found that the
antibody obtained~carries~3.8 act:ivated mixed~disulfide
groups~per mol of antibody. ~
III) Pre~aration of the lmmuno~toxin~containln~_~roloneed-
. .
- 35 -
action A chain of ricin
2.46 Ml of modified A chain conta:ining 2.87 mg/
ml are added to 1.5 ml of the solution of activated
antibody obtained above (concentration: 7.9 mg/ml, i.e.
11.8 mg of activated antibodies) and the mixture is
incubated for 20 hours at 20C. The solution is cen-
trifuged and then purified by filtration on a Sephadex
G100 column, the optical density of the efflue~t being
measured at 280 nm. Combination of the fractions con-
taining both the antibody and the A chain gives 15 mlof immunotoxin solution containing 0.7 mg/ml, i.e. 10.5
mg. This solution contains 0.14 mg of oxidized A chain
coupled with the antibody per ml.
` The average degree of coupling in this prepara-
tion is therefore 1.2 mol of oxidized A chain per mol
of antibody.
The immunotoxin containing oxidized A chain of
ricin, IT (A-la) TlOl, obtained as indicated above, was
studied for its pharmacokinetic properties and its
specific cytotoxicity properties towards the target
cells.
Example 12: ` ~
This example demonstrates the acquisition of the
property of slow plasma elimination of the immunotoxins
containing prolonged-action A chain of ricin,~ wh`ich are
abbreviated to IT (A-La)T101.
I - Procedure ~ ~
The conjugate prepared b~y the procedure explained
in example-llis administered~to~ rabbits by a singlc in-
jection into a vein in the ear.~ The quantity injected
corresponds to 0`.415~mgtkg, expresse~d as~A~chain.
Blood samples~are take~n a~t intervals;on heparin.~ ~The~
plasmas are analyzed with~the~ aid~of a~r~adioimmunometric test
with two sitesJ ~wh1ch lS ab~brevi~ated below to RIM_3.
This t~st is ~c~r-i.e~ out bV Ihe~ ~ame~ te~hn~que
:: : : ` , : ., ~ :
,
- 36 -
as that used for the test RIM-l, except that the
solution Ac2 here is a so]ution of goat antibodies in-
hibiting mouse IgG, purified by affinity chromatography
and radiolabeled as described for the technique RIM-l.
The concentration of modified immunoto~in in the samples
to be determined is measured by reference to a calibra-
tion curve esta~lished by introducing the modified
immunotoxin at different known concentrations. The
test RI~l-3 has the same reliability and reproducibility
characteristics as described for the technique R~ -1.
By ~ay of comparison, a control study is carried
out under the same conditions with the conjugate called
IT TlOl, which is obtained by the reaction of the same
antibody T101, substituted by activated disulfide groups,
with the native A chain of ri~in. The preparation and
the cytotoxic properties of this conjugate have been
described in French Patent Application no.2 5l6 794
The results of these experiments are represented in the
same way as for the uncoupled A chain of ricin in ex-
ample 2II - Results
Figure 7 shows the plasma elimination curves, as
a function of time, for IT T101 and IT (A-La) T101, in-
jected intravenously. Twenty-four hours after injection,
the concentr&tion of active immunotoxin is 140 times
greater Eor IT (A-La) TlOl than for IT T101. This fact
demonstrates that the new pharmacokinetic properties
of the oxidized A chain are retained after coupling with
an antibody.
Example 13: ~
This example demonstrates the retention of the
specific cytotoxicity properties of IT (A-La) T101 towards
the target cells.
The fundamental bioIogical property of the A
chain of ricin is to inhibit protein synthesls in cells
: ~ :
'h
'
,
- 37 -
by degradation of the ribosomal sub-unit 60S. The
technique uses a cell model in ~hich the efEect of the
substances studied on the incorporation o~ 14C_
leucine into cancerous cells in culture is measured.
The cells used belong to the CEM cell line
derived from a human T leukemia which carries the anti-
gen T65. The cells are incubated in the presence of
the substance to be studied, and then, when incubation
has ended, the degree of incorporation of 14C-leucine
by the cells treated in this way is measured.
This measurement is made by a technique adapted
from the one described in Journal of Biological Chemistry
1974, 249(11), 3557-3562, using the tracer C-leucine
to determine the degree of protein synthesis. The radio-
activity incorporated is determined here on the wholecells isolated by filtration.
On the basis of these determinations, it is
possiblè to draw the dose/effect curves, plotting, on
the abscissa, the molar concentration of A chain in the
substances studied, and, on the ordinate, the incorpora-
tion of 14C-leucine expressed as a percentage of the
incorporation by control cells in the absence of any
substance affecting protein synthesis.
It is thus possible to determine, for each sub-
stance studied, the concentration which causes a 50% in-
hibition of the incorporation of C-leucine, or "50%
inhibitory concentration" (IC50).
Figure 8 shows the curves obtained in the same
experiment ~ith IT (A-La) T101 and the uncoupled ox-
idized A~chain in the presence of 10 mM ammoniumchloride in the incubation medium. It can be seen on
this figure that the IT (A-La) T101 has a very strong
cytotoxic activity (IC50~= 5.5-10 1 M) which is aboilt
80,000 times greater than that of the uncoupled ox-
idi~ed A chain, measur:d under the same conditions.
,
- 38 -
Example 14:
This example demonstrates the comparative cyto-
toxic efficacy oE IT (A-La) T101 and IT T101 towards
CEM target cells, measured in a clonogenic test.
Immunotoxins are dedic~ted to the eradication of
every single one of the target cells. This performance
can only be evaluated with a highly sensitive technique;
tests for the inhibition of colony formation offer this
possibility because a single surviving cell can be shown
up among several million dead cells. This is made poss-
ible by optimum culture conditions in a gelled medium,
applied to the CEM human lymphoid line.
I - Technique for measurin~ cytotoxicity_by the inh _ition
of colony formation
The medium used for cloning is the medium RPMI
1640 to which 1 mmol/l of sodium alpha-ketoglutarate,
1 mmol/l of sodium oxaloacetate, 5% of inactivated fetal
calf serum and 10% of inactivated horse serum are added. A first,
0.3% agar solution (Agarose-type VII, SIGMA laboratories)
is prepared in this medium, placed as a thin layer in
small Petri dishes and solidified at +4C. The cells
are mixed with a second, 0.275% agar solution kept at
37C, which is then deposited on the-first layer and
solidified. These concentrations of agar were chosen
after a preliminary study aimed at simultaneously opti-
- mizing the cloning efficiency,~ the size of the colonies
and the consistency oE the medium. After 15 days in
the incubator, the colonies are counted using an auto-
matic colony counter ("ARTEK", DYNATECH, U.S.A.). To
determine the cloning efficiency and thus the exactnumber of cel~ls surviving the immunotoxin trèatment, it
is essential to establish a calibration line~showing the
number of cells i-noculated as a E~anc~tion~of the number
of colonies formed. ~We~hav~e proved that~the cloning ~
efficiency indicated by this calibrat~ion-line is~prac- -
: ~ :
`:
.
~L2
- 39 -
tically unaffected by the presence of a high proportion
of dead cells, which is the situation naturally en-
countered when the cells are treated with the immuno-
toxin.
The immunotoxin treatrnent is carried out by in-
cubating the cells in the exponential growth phase and
at a concentration of 106/ml with the immunotoxin IT
(.~-La) TlOl or IT T101 at different concentrations, in
a total volume of 1 ml of the medium RP~II 1640 contain-
ing 10% of inactivated fetal calf serum and 10 mmol/l
of ammonium chloride. The incubation takes place at
37C under an atmosphere containing 570 of carbon dioxide
and ~ith horizontal shaking of the test-tubes (2500 rpm
~ith a "GIRATORY*G-2" shaker, NEW-BRUNSWICK). The cells
are then washed and different dilutions are prepared,
before mixing ~ith the agar solution, so that the number
of ceIls surviving can be measured in the zone of maxi-
mum sensitivity given by the calibration line. The
results are expressed as the absolute number of cells
~0 surviving, extrapolated from the cloning efficiency,
using the following relationship:
absolute number of cells surviving: C x_d
E
where C is the number of clones per Petri dish, d is
the dilution factor of the cell preparation examined
and E is the cloning efficiency established from the
slope of the calibration line. Each point corresponds
to the average of three tests.
II - Results
Figure 9 shows the curves of the cytotoxic
activity of the immunotoxins IT (A La) T~101 and IT
T101 on the CEM cells~, in the presence of lO m~l
~ :
ammonium chIoridet as~a function of the immun~otoxin
concentration (expressed as the~molarity of A~chain).
It is apparent that the efficacies oL these two
products are of the same order of magnitude. The
* - Trademark :: ; ::
:: ~
, ~ : ~ . :
,
., ,
- ., ' .';
.
::
- 40 -
resulting reduction in the number of cells is extremely
large in both cases since, for concentrations as loi~ as
10 11 ~l, the proportion of residual cells surviving is
of the order of 0.001% of the initial value. This effect
is highly specific since, at these concentrations, it
was proved that the uncoupled A chain or a non-specific
immunoto~in has no effect on these cells.
This example demonstrates that IT (A-La) T101
has specific cytotoxicity properties which are virtually
identical to those of conventional IT T101.
Example 15:
Conjugate obtained by the reaction of an antibody in-
hibiting human T eells (an antibody directed against the
antigen T65), substituted by activated disulfide groups,
with the oxidized and functionalized A chain (NE~I) of
ricin, the coupling taking place between the activated
disulfide groups and the functionalized sugar residues
of the-A chain.
1) - Preparation of the immunotoxin
a) Preparation of the functionalized A chain
The A chain is blocked with N-ethylmaleimide on
its SH group and then oxidized for 18 hours by the pro-
cedure described in example 3.
CouPlino with cystamine
After dialysis against carbonate buffer of pH
9.5, 5.2 ml of a protein solution containing 4.65 mg/ml
are incubated with 18 mg of cystamine hydrochloride for
2 hours at 25C. This incubation is followed by reduc-
tion with sodium borohydride (200 equivalents per~ mol
of A chain, i.e. 156 microliters of a solutlon contain-
ing 17.6 mg in 1 ml of 0.1 N NaOH) for 2 hours at 25C.
The excess reagent is removed by continuous
dialysis a~ainst 125 m~l phosphate bufEer of pll 7. The
disulfide bridge of the fixed cystamine is then reduced
with 2-mercaptoethanol at a final con~centration of 5
: ' ' !
'' ~- ' `
'
.
- 41 -
per cent for 1 hour at 30C, this being followed by
further continuous dialysis against 125 mM phosphate
buffer of pH 7 (20 l at 300 ml/hour). After dialysis
and centrifugation, 0.25 SH per mol of A chain was
determined by Ellman's method.
b) Preparation of the antibody (see example 11)
c) Coupling reactibn -
1.5 ml of the solution of modified A'chain ofricin (i.e. 0.058 micromol) are added to 211 micro-
liters of the solution of activated antibody obtainedabove (i.e. 0.006 micromol~. The mixture is left to react for
18 hours at 30C. The reaction medium is the-n dialyzed
against P'BS buffer (10 mM in respect of phosphate and
140 m~l in respect of sodium chloride, pH 7.4). After
centrifugation and examination by polyacrylamide grad-
ient electrophoresis, it is found that the immunotoxin
obtained has an average degree of coupling for this
preparation of 0.8 A chain (NEM) per mol of antibody.
2) - Propertie$ of the immunotoxin IT (A(NEM)-La-
cysteamine) T101 ~ -
Specific cytotoxicity activity
It is found that this immunotoxin, prepared by
the procedure explained above, has a very strong cyto-
toxic activity on the CEM target cells (IC50 = 1.2~10
M, established by the method described in example 13).
b) Plasma elimination ` ~ -
The immunotoxin lS administered to rab~bits by a
single in~jection into a,veln in~the ear (50 m1~crograms
- of A chain/kg). The plasma samples~collected after 22
hours are analyzed with the aid of the immunoassay RIA-
3 (example 12). The results are sho~n in the table~
below. The values~for IT T101 are 1ndicated~by way of
comparison.
~ 42 -
_ ._
Relative plasma
concentration 22 hours
after injection
IT (A(NEM)-La-cysteamine)T101 2.4%
IT Tl01 0.08%
Twenty-two hours after injection, the concentra-
tion of IT containing modified A chain is 30 times greater
than in the case oE IT T101.
Example 16:
Conjugate obtained by the reaction of an antibody in-
hibiting human T cells (an antibody directed against the
anti8en T65), substituted by activated disulfide groups,
with the methylated, oxidized and functionalized A chain
(NEM) of ricin, the coupling taking place between the
activated disulfide groups and the modified sugar resi-
dues of the A chain.
l) - Preparation of the immunotoxin
- a) Preparation of the functlonalizèd A chain
The A chain is blocked with N-ethylmaleimide on
its SH group and then methylated and oxidized for 18
hours by the method described in example 5.
Coupling with c~stamine
After dialysis against 0.1 ~I carbonate buffer
o pH 9.5, 18.5 ml of a protein solution~containing 2.5
mg/ml are incubated with;35.6 mg of cystamine hydro-
chloride for 2 hours at 25C. This incubation is
followed by reduction with sodium borohy~dride (200 ~
equivalents per mol of A chain, i.e. 39S microIiters of
a solution containing 17.6 mg in 1 ml of 0.1 N~NaOH) for
2 hours at 25C.
The excess rèagen~t is removed by continuous
dialysis against 125~mM phosph-ate buffer of pH 7. The
disulEide bridge oL the fix~e~d cystamine is then~reduced
with 2-mercaptoethanol at a~final concentration of 5~
,. : ~ : :
,
- 43 -
per cent for 1 hour at 30~C, this being followed by
further continuous dialysis against 125 m~ phosphate
buffer of pH 7 (201 at 300 ml/hour). After dialysis
and centrifugation, 0.32 SH per mol of A chain was
determined by Ellman's method.
b) Preparation of the modified antibody
A solution containing 2.12 mg of N-succinimidyl
3-pyridin-2-yldithiopropionate in ethyl alcohol is
added to ~3.5 ml of a solution of antibody TlOlcontaining
4.4 m~/ml (i.e. 0.68 micromol). The mixture is stirred
for 30 minutes at 25C and then dialyzed against 125 mM
phosphate buffer of pH 7. After dialysis, the protein
solution is centrifuged to give 23.5 ml of a solution
containing 4.2 mg of modified antibody per ml.
By spectrophotometric analysis at 343 nm of the
pyridine-2-thione released by exchange ~ith 2-mercapto-
ethanol, it is found that the antibody obtained carries
3.2 activated mixed disulfide groups per mol of anti-
body.
c) Coupling reaction
7.3 ml of the solution of modified A chain of
ricin (i.e. 0.275 micromol) are added to 781 microliters
oE the solution of activated antibody obtained above
(i.e. 0.022 micromol). The mixture is left to react for
18 hours at 30C. The reaction medium is then dialyzed
against PBS buffer (10 mM in respect of phosphate, 120
mM in respect of sodium chloride, pH 7.4).
After centrifugation and examination by poly-
acrylamide gradient electrophoresis, it is~found that
the immunotoxin obtained has an average degree of coup-
ling for this preparation of 0.8 oxldized methylated A
chain (NEM) per mol of antibody. ~
The immunotoxin containin~ methylated oxi~d~ized
A chain (NE~) of ricin, obtained as indicated above,
~as studied for its;pharmacokinet~lc~properties and its5 specific cytotoxicity properties~toi~ards the target cells.
:
:: ~
. '' ` ~
~:, , ' :
:~ ~
- 44 -
2) - Properties of the immunotoxin IT (methylated A(NEM)-
la-cystcamine) T10l
a) Specific cytotoxicity activity
It is found that this immunotoxin, prepared by
the procedure explained above, has a very strong cyto-
toxic activity on the CEM target cells (IC50 = 7-10 12 M ,
established by the method described in example 13).
b) Plasma elimination
The immunotoxin is administered to rabbits by a
single injection into a vein in the ear (81 micrograms
of A chain/kg). The plasma samples collected after 24
hours are analyzed with the aid of the immunoassay RIA-3
(example 12). The results are shown in the table below.
The values for IT T101 are indicated by way of comparison.
Relative plasma
concentration after
22 hours
IT (methylated A(NEM)-La-cysteamine) 1.4%
Tl01
IT TlOl 0.08%
. . . _ _
Twenty-two~hours after injection, the concentra-
tion of IT containing modified A chain is 17.5 times
20` greater than in the case of IT T101.
Examplè 17:
Toxicity of the prolon~ed-action A chain injected into
mice
It was important to check the overall toxico-
logical impact of the oxidized A chain on the wholeanimal (the toxicity of the immunotoxins being of the
same order of magnitude as that of the A chain at equal
molar doses). This was done~by determining the 50%
lethal dose of the oxidized A chain, administered intra-
venously to Charles River Fran~ce CDl mice, by comparison
:~,; ; :
- ,
' '
. -
- 45 -
with that of the native A chain.
The values found are indicated in the table
which follows.
~ ~ (micrograms/mouse)
Native A chain - 550
Oxidized A chain 800
These results show that the toxicity of the
oxidized A chain is lower than that of the native A
chain. This means that, despite a considerable in-
crease in the plasma level of the A chain when thelatter has been modified by oxidation, the toxicity of
- the product is not only not increased but, on the con-
trary, substantially reduced.
The immunotoxins containing modified cytotoxic
sub-units can therefore be used as drugs in human
therapy. These modified immunotoxins can be used for
the treatment of cancerous or non-cancerous diseases
where the target cells would be recognized by the anti-
body used to prepare the immunotoxin. The optimum ad-
ministration conditions and the treatment~time willhave to be determined in each case according to the
subject and the nature of the disease to be treated.
In more general terms, antitumoral glyco-
proteins whose carbohydrate units are modified by
oxidation with the periodate ion, and which have a
longer half-life than the corresponding unmodified
antitumoral glycoproteins, are useful as drugs.
Therefore, according to a~ further feature, the
present invention relates~to antitumoral drugs~in which
- 30 an antitumoral glycoprotèin whose carbohydrate units
are modified by oxidation with the~periodate ion 1 S
brought in-to a form suitable f~or~administration by
injection and preferably intravenous administration.
`` ` ' ' , ~ ~ : ` :
'
::
.
