Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
212~987
WO 93/17033 - PCI/US93/01375
Inhibition of Cell Adhesion by Chemically-Defined Oligosaccharides, Their Derivatives,
~Iime~ics, and Antiboclies Directed Thereto
Cross-Reference To Related APPlications
This application is a continuation-in-part of pending U.S.
Application Ser. No. 07/950720 ~iled 25 S~ptember 1992; which is
a continuation-in-part of pending U.S. ~pplication Ser. No.
07/836978 filed 19 February 1992; which is a continuation-in-part
of pending U.S. Application Ser. No. 07/78996g filed
; ~: 12:November 1991; which is a continuation-in-part of pending U.S.
,
: ~ Application Ser. No. 07/724983 filed 2 July lg91; which is a
` c~ntinuation-in-part o~ U.S. application Serial No. 07/575539
filed 30 August 1990 (abandoned).
All ~ive applications expressly are incorporated he~ein by
ref2rence.
:~ :
Technical Field
~: ~ The present invention is directed generally to the
-
`:
15~: inhibition of:tumor: cell metastases and inv~siveness and of
inflammatory processes:based on the inhibition of adhesion of
~`~ tumor cells or inflammatory leukocytes to specific types of
cells. More specifically, the in~ention is directed to such
inhibition through: the~ use of ~umor-associated carbohydrate
WO93/17033 2 1 2 9 9 8 7 PCT/US93/01~
antigens, leukocyte-associated carbohydrate antigens,
oligosaccharide derivatives thereof, mimetics of the
tumor-associated carbohydrate antigens, leukocyte-associated
carbohydrate antigens and antibodies directed to the
tumor-associated carbohydrate antigens.
Backqround of the Invention
Despite enormous in~estment of financial and human
resources, cancer remains one of the major causes of death.
Current cancer therapies cure only about half o~ all patients who
develop a malignant tumor. In most human malignancies,
metastasis is the major cause of death.
Metastasis is the formation of secondary tumor colonies at
one or more distant sites. Metastasis is a multistep process of
which tumor invasion is the f irst step. Tumor cells locally
in~ade host tissue barriers, such as the epithelial basem~nt
~: ~ membrane, to reach the interstitial stroma where th~ tumor cells
~ .
gain acces~ to blood vessels (or lymphatic chann~ls) for f~ther
~; di~semination.: After invading the endothelial layer of the
vessel wall, the circulating tumor c~lls are dislodged into the
circulation and arrest in the precapillary venules of the target
~ org~n by adhering to endothelial cell lumenal surfaces or exposed
:: basement membranes. The tumor cells again invade the vascular
::
wall to enter~the organ p~renchyma. Finally, the ~xtravasat2d
;~ . tumor cell grows in a tissue di~ferent from where the tumor
: 25 originated. ~ :
2 -
2~29987
WO 93/17033 PCr/US93tO137~
In most human malignancies, distant metastases often are too
small to be detected at the time the primary tumor is treated.
Furthermore, widespread initiation of metastatic colonies usually
occurs before clinical symptoms of metastatic disease are
evident. The size of the metastases, age of the patient,
disp~rsed anatomic location and heterogeneous composition all are
factors that hinder surgical removal of tumors and limit the
concentration of anticancer drugs that can be deli~ered to the
metastatic colonies.
Due to difficulties in the current approaches for treating
and preventing metastases, ~h~re is a need in the art for
improved methods and composi~ions capable of inhibiting the
metastasis potential of tumor cells. The present in~ention fills
those needs and further provides other related advantages.
on the other hand, there are common mechanisms between the
initiation of the inflammatory process and metastasis. For
example, both processes are triggered by adhesion of cells,
leukocytes in the former case and tumor cells in the latter, to
microvascular endothelial cells followed by transendo ~ elial
migra~ion o~ th~ leukocytes or tumor c~lls into the tissue
spaces. Both processes are enhanced by activating platelets.
Both processes~ are mediated strongly by specific types of
carbohydrata~, such as tumor associated carbohydrate antigens
~TAC~ or leukocyte associated carbohydrate antigens (LACA~.
Some TACA'~ share s~ructures with L~CA's.
The instant invention is directed to and based on the
inhibition of cell adhesion, for example, through TACA's or
LACA~s, using, for example, antibody thereto.
- 3 -
W093/17033 2 12 g 9 ~ 7 PCT/US93/01~-~
Summary of the Invention
Briefly stated, the instant invention pr~vides compositions
and methods of inhibiting metastatic potential and in~asiveness
of tumor cells based on blocking tumor cell adhesion by
carbohydrate structures or antibodies directed thereto. The
instant invention also relates to compositions and methods of
~;; inhibiting inflammation potential of leukocytes based on blocking
leukocyte adhesion by carbohydrate structures or antibodies
directed thereto. The rationale for the approach is to block
(a) carbohydrate to carbohydrate interaction; (b) carbohydrate
to selectin interaction; or (c) both. For example:
i) In model experiments with mouse melanoma Bl6
variants with high and low metastatic potential, high
metastatic variants, BL6 and FlO, express more ~M3
~- lS than low-metastatic or non-metastatic variants, Fl or
; Wa4. Adhesion of ~high metastatic ~ariants ko
endothelial cells is greater than wi~h low met~tatic
variants ~and the adhesion is inhibited by
Me-~-lactoside, GM3 or LacCer (each within liposomes)
2~0~ or oth~r~lactoside~; derivations. The sugars and
derivatives also inhibit 816 melanoma metastatic
potential. Such is an example of (a) abo~e, that is,
interfering~ with a carbohydrate to carbohydrate
interaction.~
,~
~ 4 -
: ~
.- W093/17033 2 1 2 ~ 9 ~ 7 PC~/US93/0137~
ii) As to human cancer, patients whose primary tumors
express defined tumor-associated carbohydrate
antigens, such as H/~eY/Le (d~fined ~y monoclonal
antibody MIA-15-5), sialosyl Tn (defined by monQclonal
an~ibody TKH2) or sialosyl-LeX (defined by monoclonal
antibody FH6, SN~3 or S~H4), had a much shorter
survival rate than those patients whose primary tumors
do not express or which weakly express ~hose antigens.
iii) Those tumor-associated carbohydrate antigens (GM3
lQ in the mouse melanoma model and H/~Rb/~eY, sialosyl-LeX
or sialosyl-Tn in human tumors) are essentially
adhesion molecules which are reco~nized by target
cells, particularly platelets or endothelial ~ells.
Such is an example of a combination approach, that is,
interfering with (a) and (b).
: iv) I~teraction o~ tumor cells with endothelial cells
and :plat~l~ts is mediated by LECCAM (or sel~tin),
EL~- or G~P-140, which are expressed on acti~ated
endothelial cells and activated platelets.
~: 20 Sialosyl-LeK antigen has been knvwn to be recognized by
those LECCAM's. Such is an example of (b), af~ecting
' a carbohydrate to se1ectin inter~ction.
v) ~ GMP-l40, whose expre~sio~ on platelet or
endothe1ial cells is induced by thrombin, ADP or (~MP)
~phorbol ester~ may play an important role in
;~ - 5 -
W093/17033 2 ~29 9 8~ PCTIU~93/01~^
platelet-tumor cell interaction and mediate tumor cell
metastases. While the epitope recognized by that
selectin was identified previously as sialosyl-Le
(Polley et al., Proc. Natl. Acad. Sci. 8~:6224, 1991),
it has been found that sialosyl-Le~ ~also known as
monosialosyl-Lea ~, monosialosyl-Le~ II (a positional.
isomer o~ sialosyl-LR~) and disialosyl-Le~ also are
recognized by ~MP-140. GM~-140 binds to sialosyl Le~
better than to sialosyl-LeX. Such is another example
of process (b).
vi) EL~M-l, whose expression on endothelial cells is
induced by interleukin-1, TGF~, TNF~ or
lipopolysaccharide, may play an important role in
endothelial cell-leukocyte and endothelial cell-tumor
cell interaction, mediate tumor cell metastasis,
mediate endotheli~l cell-l~ukocyte interactions and
m~diate transendothelial migration of leukocytes and
tumor cells. While the epitopes recognized by that
sel~ction previo~sly were identified as sialosyl-LeX
and ~i~losyl-Lea (Phillips et al., Science 250:1~30,
: 1990; Berg et al., JO Biol. Chem. 266:14869, 1991;
Takada et al~, Biochem. Biophys~ ResO Comm~n. 179:713,
~991), it ha~ ~een found that the selectin epitopes
also are internally sialyla~ed, penultimate
fucosylated typ~ 1 or type 2 chains, such as
monosialosyl-Le II and disialosyl-Le , particularly in
a dynamic flow system. But the binding phenomenon is
-- 6
, ^WO 93/17033 21 ~ 3 9 8 7 PCTlOS93/0137~
vi.brant. Under static: or low shear stress dynamic
conditions, EI.AM-l (also known as E-selectin)
recognizes primarily a2-3 sialylated and atl~ or a~1-4
fuc:osylated carbohydrates, such as SLeX and SLe~.
However, under middle to high shear stress dynamic
conditions, molecules having formulae ~I) or (II),
~ee, for example, Figure 20, such as LeX/SLe~, play an
important role in providing high a~f inity ~inding
sites to E-se~ectin. That role is p~rti ::ularly
evident under high shear stress conditiolls.
~tii) Human c:o~on tumor cells showing differential
expression of metast~tic potential in nude mice showed
a close cs:~rrelation with the expression of
sialosyl-LeX, i . e.,, cells with high metastatic
IS potent~al expressed high levels of sialosyl-LeX and
vice versa.
viii) ~dhe~ion of E~selectin-expressing cells to S~eX is
enhanced greatly when SLeX i~ mixed in liposomes with
various quantities O~ LQX. Hence, greatly enhanced
adhesion was observ~d not only with hybrid SLeX/Lex,
but al~o with mixed glyco-liposomes with SLeX and LRX.
ix) Human endothelial cells are char~cterized by high
expres~ion of H (Fuc~1~2Gal) and many types of human
cancers are characterized by expression of LeY, H or
Leb de~ined by monoclonal antibody MIA-15 5.
W093/17033 PCT/US93/O1?'~
Interaction of H with LeY or H with H has been
established clearly, ther~fore, those human tumors
expressing H/LeY/Le may adhere to H-expressing
endothelial cells which are mediated by LeY-H or H-H
interaction. Such is an example of process ~a), that
is affecting a carbohydra~e to carbohydrate
interaction.
x) Monoclonal anti~ody MIA-15-5 directed to H/LeY/Leb
~ inhibited lung metastasis of highly ~etastatic FlO and
BL6 ~ariant cells in the mouse. Furthermore,
monoclonal antibody FH7 directed to disialosyl-Le~ and
monosialosyl-LR~ II inhibited adhesion of human cancer
: : cells expressing those antigens in a dynamic flow
system.
Based on tho~e an~ other various observatlons an~
considerations, the instant invention provides the following:
:~ .
a~ Compositions and me*hods ~or inhibiting tumor
cell metastasis based on tumor ~ell adhesion mediated
y carbohydrate antigen by such oligosaccharides
: XO ~ comprising GM3, H, LeY, Le , monosialosyl~Le ~SLe )~
Len ~ LeX ~ hybrid sugars, such as, LeX~SLex hybrids
Structure l~ in Figure ~O3, monosialosyl-LeU I (SLe ),
~monosialosyl-Lea II, sialosyl Tn, lactosyl and other
:
structures; as depicted in structures ~-14, in
Example 3.
- 8 -
- . W~93/17033 2 1 2 ~ 9 8 7 PCT/US93/0~37~
b) Compounds, such as those set forth in Figure 20,
which can be LRX/SLex hybrids, or an appropriate
mixture of the relev.ant components, such as LRX and
SLRX, provide high affinity adhesio~ binding sites,
particularly under high shear stress conditions in a
dynamic flow system. Hence, such compounds block
E-selectin-mediated adhesion of t~mor cells or
leukocytes to endothelial cells.
c) Oligosaccharide derivatives bas~d on those
structures and linked to an appropriate carrier.
d) 01igosaccharide derivatives whose sugar
: str~ctures ~re modified appropriately showing better
blocking activity of tumor cell adhesion based on
oIigosaccharide-lectin : (selectin; LECCAM) or
QligosaGcharide~oligosaccharide interaction.
~, ~
e) UtiIization of antibodies recognizing' those
oligosacrharides ~comprising and representing
:: : : : : :
: tumor-associated carbohydrate antigells also may
inhibit tumor cell adhesion to endothelial cells,
~ 20 ~ p~atelets or target cells, and may inhibit metastasis.
:i ~
e) ~ :~ Vtilization of combination~ of antibodies
: recognizing those oligosaccharides involv2d in cell
: ~ :
~ a~lesion and representing ~umor-associated antigensO
: ~ :
: :
:
: ' - g _
::: :
WO 93/17033 r PCl/US93/01~
2~299~ 1
Thus, in one aspect of the instant invention, a method for
inhibiting tumor cell metastasis potential or inflammation within
a biologic preparation is provided. The method comprises
incubating the biologic preparation with at ~east one a~ent
selected ~rom the group consisting of (a) tumor-associated
carbohydrate antigens (or leukocyte-associated carbohydrate
antigens) that exhibit differential prognostic si~nificance,
(b) antibodies that specifically bind to those antigens,
(c) oligosaccharide components of those antigens, (d) conjugates
~ of those antigens or oligosaccharides and (e) mimetics of the
tumor-associated carbohydrate antigens (or leukocyte-associated
carbohydrate antigens), the agent inhibiting the metastasis
potential of the praparation. Suitable biologic preparations
include cell cultures and biologic fluids~
~15 : Another aspect of~the instant invention provides a method
:~ for inhibiting metastasis potential of tumor cells or
inflammation in a warm-blooded animal. The method comprises
administering to a warm-blooded animal an effective amount of at
least :one agent~elected~; from the group consisting ~ (a)
20~ ~ tumor-associated~carbohydrate:antigens (or leukocyte-associated
carbo~ydrate antigens)~ that exhibit differential prognostic
significance, ~b) antibodies ~that specifically bind to those
antigens, ~cj~ oligosaccharide components of those antigens,
(d) conjugates of those antigens or oligosaccharide components
: .
2~5~ and~ (e): mimetics~ of the tumor-associated carbohydrate antigens
or leukocyte-associated~carbohydrate antigens), the agent
inhibiting tu~or cell metastasis potential or inflammation
potential. : ~ ~
-- 1 0 --
W093/17033 212 9 9 ~ 7 PCT~US93/01375
Wîthin a related aspect, the instant invention provides a
variety of glycoconjugates useful for prolonging the in vivo
half-life of oligosaccharide components, The conjugates comprise
an oligosaccharide coupled to polyethyleneglycol.
Additional oligosaccharide ~omponents for use within the
methods and c~mpositions of the instant invention include
lactose, lacto-N-tetrose, methyl ~-D-lactoside and phenyl
~D-thiolactosideO Oligosaccharide components may be used
individually or in combination with one another.
The instant invention further provides a variety of methods
for inhibiting GMP~140-mediated or ~LAM-l-mediated cell
aggregation or adhesion causing metastasis at a tumor site and
inflammatory responses at a site.
One such method inhibits GMP-140-mediated or ELAM-1-mediated
cell aggregation or adhesion within a biologic preparation and
comprises incubating the biologio prepara~ion with at least one
: agent selected from the group consisting of: (a) a hybrid æugar
~olecule, such as one comprising LeX and SLeX (Struc~ure 1 o~
:
~:~ Figure 20, a branched~type II ahain); (b) a mixture ~ the
:~ 20 components of the hybrid ~ugar of (a), such as, ~ex and SLeX;
: ~c) monosialosyl-Lea I, ~ Le~, L2X~ mono~ialosyl-Le~ II,
: : ~ disialosyl-L~ or sialosyl ~x; (d) antibodies that specifically
: bind to a hybrid sugar,~ such as LeX/SLex, or to the componen~s
'~ therecf; (e) antibodies that specifically bind to
: 25 monosialosyl-Le^ I, ~R~, LeX, monosialosyl-Le~ II, disialosyl--Le~
or sialosyl LeX; (f) oligosaccharide components of hybrid
sugars, su h as LeX~S~ex, monosialosyl-Le~ I, Le~, ~e ,
monosialosyl-~ea II, disialosyl-Lea or sialosyl LeX; (g)
-- 11 --
W~93/t7033~ 12 9 98~ P~T/US93/01~7
conjugates of hybrid sugars, such as SLeX/Lex, monosialosyl-Le~
I, Le0, LeX, monosialosyl~LR~ II, disialosyl-Le or sialosyl LRX
or of the oligosaccharide components; and ~h) mimetics of hybrid
sugars, such as LRX/SLex, monosialosyl-Le I, Lea , LR ~
monosialosyl-Lea II, disialosyl-LR or sialosyl Le , said agent
inhibiting the cell aggregation or adhesion.
Another such method inhibits GMP-l40-mediatsd or
ELAM-l-mediated cell aggregation or adhesion at a tumor cell sr
inflammatory site in a warm-blooded animal thereby reducing
metastatic potential or in~lammation ~t the site and comprises
administering to the warm-blooded animal an ePfective amount of
at least one agent selected from the group consisting of: (a)
hybrid sugar, such as, SLeX/LeXî (b) a mixture of the components
~ of a hybrid sugar ~a~, such .~s, LeX and SLeX;
15 (C) ~0~05ialo~yl - Le I ~ Lea Lex mo~osialosyl-Le II,
disia}osyl LRa or sialosyl LeX; (d~ antibodies that specifically
bind to a hybrid æugar, such as LeX/SLex, monosialosyl-Le~ I, Le~,
eX, monosialosyl-Lea II, disialosyl-Le~ or sialosyl LeX; (e) a
mi~ure of antibodies, paxticul~rly to the components of ~ hybrid
; 20 sugar, such as to ~8X, SLeX, Le~ or SLe"; (f) oligosaccharide
co~ponents of a hybrid sugar, such as I.eX/SLe", ~onosialosyl-Le"
:~ I, Le~, LeX, ~monosialosyl-Le II, disialosyl-I,si or sialosyl Lei;
(g) c:onjugates of a hybrid sugar, such as, LeiX/SLe
~' monosialosyl~ I, Lea~ monosialosyl~ II, disialc: syl~
or sialosyl LsX or: of ~ the oligosaccharide components; and
(h) mimetics of a hybrid sugar, such as, L~X~SLe
monosialosyl-Leia I~ Lea~ LeiX~ monosialosyl-Lein II~ disialosyl-Leia
~ 12 --
- WO93/17033 ~12 9 9 8 7 PCT/US93/0137~
or sialosyl LeX, the agent reducing the metastatic potential at
~he tumor cell site or inflammation in the warm-blooded animal.
The instant invention alsQ provides a method o~ inhibiting
GMP~l40-m~diated or ELAM-~-mediated cell aggrega~ion or adhesion
5 at a site of inf lammation in a warm-blood~d animal thereby
reducing inflammatory potential at the site and comprises
administering to warm-blooded animal an ef~ective amount of at
least one agent selected from the group consisting of (a) a
h~brid sugar, such as, LeX/SLRx; (b) an appropriate mixture of
sugars which a~e the components of a hybrid sugar (a), such as,
LeX and sLex; (c) monosialosyl-Le~ I, LR, Lex
monosialosyl-Le~ II, disialosyl-Lea or sialosyl LeX;
~d) antibodies that specifically bin~ to a hybrid sugar, such
as, LeX~SLex~, monosialosyl-Le~ I, Len, LeX~ monosialosyl-LR~ II,
disialosyl-Le~ or sialosyl LeX; (e) a mixture of antibodies,
partiaularly to the components of a hybrid sugar, such as to LRX,
S~eX, Le~ and SLea; (f) oligosaccharide components of a hybrid
: ~ : sugar, such: as, SLeX/Lex, monosialosyl-L2~ I, Le , Le ~
monosi~IosylLe3 II, disialosyl-Le~ or ~ialosyl~ Le ;
20 (9) con3ugates of ~ a ~hybrid sugar, such as SLs'~/LeX,
~; monosialosyl-Le~ I, Le~ LeX, monosialosyl-Lea II, disialosyl-Lsh
: or sialo8yl LeX or: of the oligosaccharide co~ponents; and
: (h) mi~etic6 of a hybrid sugar, such as LRX/SLex
mono~ialosyl-Lea I, Le~, ~eX~ monosia}osyl-~Ra II, disialosyl-Le~
or sialosyl LeX~ the agent reducing the inflammakory potential at
:
the inflammatory site in the warm blooded animal.
~: In another aspect, the instant invention provides a method
for identifying a tumor associated carbohydrate antigen (TACA)
.
: 13 -
wo 93J17033 21 PCT/US93/01~-~
epitope to which lectin activity of GMP-140 is directed,
comprising: ~A) cons~ructing a fluorescent probe comprising
fluorescent plastic ~eads coated with the TACA epitope suspected
of being targeted by GMP-140; (B) incubating the fluorescent
probe with a suspension of platelets; and (C) determining the
degree of binding of the ~luorescent probe to the platelets.
Those and other aspects of the instant invention will become
evident on reference to the following detailed description and
al~acnea urawlngs.
:
Brief Description of the Drawinqs
Figure 1 graphically illustrates the effects of methyl
D-lact~side or methyl~-D-thiolactoside on the number and size
of lung colony deposits of BL6 cells. BL6 cells were
preincubated with control medium, 0.1 M methyl ~-D lactoside
; 15 (nMe~ lactoside") or O.I M phenyl ~-D-thiolactoside
phe-~-S-lactoside)~. Twenty thousand cells were injected
ntravenously into C57Bl mic~. Lung colony numbers were c~nted
at 21 days and colonies were classified on the basis of diameter
l m~ vs. < 1 mm), as~indicated~for each bar. Colony numbers
2~0~ are expressed per single lung. Number of experiments ("n") is
indicated in parentheses.;
::
'~ Figure 2 graphically illustrates the e~ect of prior
~: :
~ administration of methyl ~-D-lactoside on the number and size of
:: :: .~ : ;
lung colony;deposits~of BL6 cell~. Methyl ~-D-lactoside (1 ml
25~ dose) was injected intraperitoneally into C57~1 mice. After
10 minutes, BL6 melanoma cells were injected intravenously. Lung
14 -
:; ~:: :
.
^. WO 93/17033 2 1 2 9 9 8 7 PCr/US93/01375
colonies were counted and sized at 19 days. Group A represents
control animals (not administered with methyl ~-D-lactoside) and
groups B and C represent animals injected with 0.25 M and 0.5 M
methyl ~-D-lactoside, respectively. For each group, column l
S represents the total number of colonies, column 2 the number of
colonies with diameter > 1 mm and column 3 the number of colonies
with diameter ~ 1 mm. Number of experiments is expressed as "n".
Figure 3 graphically illustrates survival of cancer patients
with or without expression of a defined tumor-associated
carbohydrate antigen (TACA) in the tumors. Panel 3A represents
the expression of H/LeY/Leb antigen in lung squamous cell
carcinoma as determined by monoclonal antibody MIA-15-5.
Panel 3B represe~ts sialosyl-LeX expr~ssion in colonic cancer
using antibody FH6. Panel 3C represents sialosyl-Tn ex~ression
in colonic cancer using antibody TKH2. Panel 3D represents
sialosyl-Tn level in sera of ovarian cancer patients.
Figure 4 qraphically illustrates that melanoma cell adhesion
on LacCer is based on ~M3-LacCer interaction. The order of
metastati potential is BL6~FlO>Fl>>Wa4. Panel 4A sho~s the
: 20 order of ~elanoma cell adhesion on a LacCer-coated solid phase.
Panel 4B sh~ws the order of melanoma cell adhesion on
~acCer/Fibronectin (FN) co-coat~d solid phase. Panel 4C shows
integrin-dependent adhesio~.
Figure 5 graphîcally illustrates the melanoma cell (BL6)
2S adhes~ion on LacCer (Panel SA) an~ on endothelial cells (HuVEC)
(Panel 5B) is inhibited by Lac~er and GM3.
Figure 6 graphically illustrates the metastasis-inhibiting
effect of methyl(Me)-~-lactoside. Tumor cells were injected
W093~17033 2 1 ~99 8 ~ PCT/US93/01~
intravenously, followed by intraperitoneal injection of:
PBS ~A); 0.25 M Me-~-lactoside (B); 0.5 M Me-~-lactoside (C);
0.5 M lactose (D); 0.25 M N-acetyllactosamine (E); or 0.5 M
Me-~-galactoside (F).
S Figure 7 graphically illustrates ~-LeY and H-H interaction.
Panel 7A shows Hl-liposome binding to various glycolipids.
Panel 7B shows LeY-liposome binding to various glycolipids.
Figures 8A-8D are flow cytometric profiles o~ non-activated
~Panels 8A and 8C) and activated (Panels 8B and 8D) platelets
with anti-GMP-140 monoclonal antibody.
Figure 9 graphically illustrates the binding indices of
platelets with fluorescent beads coated with various GSL's. The
hatched bars represent non-activated platelets and the open bars
repr~sent activated platelets.
Figure lO graphically illustrates the e~fects of various
monoclonal antibodies: on binding of activated platelets to
: sialosyl-Le~-coated beads~ The abscis~a represents the percent
~: : inhibition. Column 1 r~present~ anti-GMP-140-mAb, IOP62;
: :
colu~n 2 represents anti-sialosyl-Led monoclonal an~body,
CAl9-9; column 3 represents anti-sialosyl-LeX monoclonal
; antîbody~ SNH4; and column :4 represents normal mouse IgG.
Figures llA-llD illustrate e~perimental systems
~ demonstrating dynamic adhesion of cells in a flow system.
'~ Panel llA shows ~he structure of the laminar flow chamber~
~: 25 Panel llB depicts a cro~s section of a laminar chamber in which
: the flow chamber body tl63 i~ affixed tightly with the cover slip
(3~ on which cells or adhesion molecules (9) are fixed.
Panel llC shows the entire as~embly of the recording system.
:: ~
- 16 -
: :
.
-~ WO93/17033 2 ~ 2 9 9 8 7 PCT/US93/0137~
Panel llD is a schematic presentation of the flow of tumor cells
in su~pension passing over the cell layer or adhesion molecules.
Figure 12 is a graph showing the effect of various
monoclonal antibodies on adhesion of human colon carcinoma
S Colo205 cells to interleukin-l-activated human umbilical vein
endothelial cells in a dynamic ~low system. Open circles
represent a mixture o~ irrelevant mouse IgG plus IgM (control),
the solid triangles represent monoclonal antibody CAl9-9 directed
to monosialosyl-Lea I, the open triangles represent monoclonal
antibody SNH4 directed to sialosyl-LeX, the solid circles
represent monoclonal antibody FH7 directed to monosialosyl-Le~ II
and disialosyl-LeU and the solid squares represent a mixture of
irrelevant mouse IgG plus IgM and non-activated endothelial
cellsO
Figure 13 depicts binding o~ mAb's to HL60 cells and the
effect of sialidase thereon. Binding activity was determined by
fl~w cytometry. Abscissa: log fluore~cence intensi~y.
Ordinate: relative cell numb~r. Panel A: Solid line, cells
stained with mAb SNH4 as primary antibodyO ~ot~ed line, ~trol
~0 cells stained with ~ou~e IgG plus IgM ~l0 ~g/ml~ as primaxy
antibody. Pan~l B: mAb SNH3 as primary antibody; control as in
Pane A. Panel C: Solid ~ine, cells treated with Newcastle
Disease ~irus (NDV) sialidase and then stained with mAb SNH4.
Dotted line, control cells (as in Panel A, after sialidase
kreatment). Panel D: N~ ~ialidase followed by mAb S~H3;
~control as in Panel C. Panel E: Vibrio cholerae (VC~ sialidase
followed by m~b SNH4. Panel F: VC sialidase followed by mAb
- 17 -
WO93/17033 2 1 299 ~ PcT/us93/o1?~;
SNH3. Note that expression of SLRX ~defined ~y both SNH3 and
SNH4) was abolished completely by both NDV and VC sialidases.
Figure lq depicts . adhesion of HL60 cells to
E-selectin-coated plates in a static system. Abscissa, type of
treatment. Ordinate, percent cell adhesion relative to untreated
control cell~. Panel A: effects of various sialidases.
Panel B: ef~ects of anti-LeX and anti-SLeX mAbls alone and in
combination (incubated 90 min at 37-C). Panel ~: ~ffects of ND~
sialida~e plus mA~. Panel A: NDV sialid~se (which cleaves a2~3
sialosyl at a terminal Gal, eliminates the SLeX ~tructure and
abolishes reactivity with mAb's SNH3 and SNH4, see Figure 13, but
did not abolish adhesion. VC and Arthrobacter ureafaciens (AV)
:sîalîdases did abolish adhesion. Panel B: anti-SLeX mAb's were
less effective than anti-LeX mAb's. Combinations of both types
of mAb's were most effec~ive. Panel C: Adhesion was inhibited
; ~ ; most effectively by NDV sialidase plus anti-LeX mAb.
: ~Figure l5 depicts adhesion of HL60 cells to
E-selectin-coated plates in a dynamic flow system. Truncated
E-selectin was coated onto marked areas (diameter o~ bout
0.5~c~) on plastio plates and adhesion under defined wall shear
~ str~sses was assayed as described herein. Abscissa, shear stress
dynes/cmZ). Ordina~e, n ~ er of cell~ adhered within 3 min~
Panel A: hollow circle, contr~l (untreated) cells; solid
~ triangle, cells treated with NDV sialidase; solid circle, VC
:~:25 sialidas~; and hollow triangle, ~U sialidase. Panel B: hollow
:circle, control; $ol~d triangle, cell~ cultured in medium
containing anti-SL~ IgG3 mAb SNH4; solid circle, anti-Le IgM mAb
FH2; and: hollow: triangle, anti-Le IgG3 ~Ab 5Hl.
: - 18 -
: :
-. WO93/17033 212 9 9 8 7 PCT/US93/0137~
Panel C: hollow circle, control; solid triangle, NDV sialidase;
solid circle, mAb SH1; and hollow triangle, NDV sialidase plus
mAb SHl. Panel D: hollow circlç, control; solid circle, mixture
(1~1) of mAb's SNH4 and FH2; and hollow triangle, mixture (1:1)
of mAb's SMH4 and SH1. Cleavage of ~2-~3 sialosylation at a
terminal ~al by NDV sialidase reduced adhesion somewhat, however
adhesion remained at low shear stress. In contrast, VC and AU
sialidases strongly inhibited adhesion (Pan21 15A) indicating the
importance of internal sialosylation (which is unaffected by NDV
sialida~e). That observation is substantiated by observations
that (i) NDV sialidase plus mAb SHl strongly inhibited adhesion
and (ii3 combination of anti SLe mAb SNH4 plus anti-LRX mAb's
FH2 or SHl inhibited adhesion more s~rongly than SNH4 a~one
(Panels l5B and l5D).
Figure 16 depicts reactivity of Colo201 cells with ~ar.ious
mAb's, with or without sialidase treatment. Colo~01 cells were
~ reac~ive strongly with anti-SLeA I mAb~s CA19-9 and NKH1
: : ~Panel A~, anti ~2~ ~Ab CA3F4 (Panel B) and anti-SLe~ II mAb FH7
(Panel C).~ Reactivity with CA19-9 was decreased by NDV si~idase
: 20 ~Panel b) and abolish~d by VC sialidase ~Panel G). Reacti~i~y
with CA3F4 Was increased slightly ~y NDV and VC sialidases
~Pan~ls E and H)~ ~eactiv~ty with FH7 was unchanged ~y NDV
~ialidase (Panel Fj :and decreased slightly by VC sialidase
(Panel I).
Fiyure 17 depicts adhesion of Colo20~ cells to
~-selectin-coated plates in a sta~ic system. Abscissa and
ordinate as in FigurQ 14. Panel A: effects of various
sialidases (90 min. incubation, 37 C). Panel B: effects of
- 19 -
.
WO93/17032 ~299 ~ PCT/US93/O~
sialidases (18 hr. incubation, 37C), cells were first fixed with
O.5% paraformaldehyde for lO minutes at room temperature.
Pa~el C: effects of si~lidases followed by mAb's. NDV
sialidase, which cleaves a2~3 sialosyl at terminal Gal, did not
affect adhesion, whereas VC ~nd AU sialidases, which cleave
: sialic acid residues regardless of location, abolished adhesion
(Panel B). In PaneI C, most effective inhibition was ob~erved
with VC or AU sialidase plus mAb CA3F4.
; ~Figure 18 depicts adh~sion of Colo201 cells to
E-selectin-coated plates in a dynamic flow system. The adhesion
assay is as described herein. ~bscissa and ordinate as in
Figure 15. Panel A: hollow circle, control; solid circle, NDV
sialidase; hollow triangle, AU sialidase; and solid triangle, VC
sialidase. Panel B: hollow circle, control; solid circle,
15~ anti-SLe~ I mAb CAl9-9: hollow triangle, anti-SLea II mAb FH7;
: and solid triangle, anti-Le mAb CA3F4. Panel C: hollow circle,
::: control; solid circle, CA3F4; solid triangle, VC sialidase;
hollow inverted triangle, VC sialidase plus CAl9-9; and hollow
triangle, VC sialidase plus~ CA3F4 (note ~hat adhesion waS most
2Q ~ strongly inhibited~by th~t combination). Panel D: hollow circle,
control solid triangle, ND~ sialidase; solid inverted ~riangle,
CA3F4; hollow inverted triangle, NDV sialidase plus CAl9-9; solid
~ circle, NDV sialidase ~plus FH7; and hollow triangle, NDV
: ~ialidase plus CA3F4 (note ~that adhesion was inhibîted most
;25~ ~trongly:by that com~ination). :
Figure l9 depicts~ the effect of Newcastle Disease Virus
NDV) sialidase, Vibrio cholerae (VC) sialidase or mAb's SNH4 or
SHl on HL60 binding to ELAM-coated plates in a dynamia flow
~ - 20 -
::
- W~93/17033 2 1 2 $ 9 8 7 P~T/US~3/0137~
sys~em under various shear strength conditions. The ordinate
represents per cent cell binding relative to untreated control
cells. The antibodies were used at 15 ~g/ml, NDV sialidase at
002 U/ml and VC sialidase at 0.1 U/ml. Each point represents the
S mean of three experiments. Number of untreated cells bound at
Shear StreSSeS Of 15.5, 7.75, 3.13, 1.56 and 0.78 dynes/cm2 were
4.5, 27/ 109.6 206.2 and 283.8 cells/mm2, respectively.
Figure 20 depicts VariOUS branched sugars. The hYbrid
SUgar, LeX/SLeX, iS dePiCted aS StrUCtUre 1. The glycolip.ids
COntaining SUCh a StrUCtUre Were iSO1ated frOm CO1On CarCinOma
OX Were PrePared frOm G8 gang1iO5ide presented in Structure 5
Origina11Y fOUnd in hUman erYthrOCYteS (Watanabe et a1., J. BiO1.
Chem., 254:8223, 1979) bY enZYme Cata1YZed a1~3 fUCOSY1atiOn.
StrUCtUre 2 WaS Obtained bY ~1~3 fUCOSY1atiOn Of C~mPOUnd 6
Origina11Y Obtained frOm hUman P1aCenta. StrUCtUre 2 hOWeVer did
nOt eXhibit hi~h affini~Y binding tO E-Se1eGtin. StrUCtUreS 3
and 4 dePiCt ana1OgS With high affinitY binding SiteS haVing LeX
Sia1Y1-Ga1~1~3Ga1NaC Within th~ Same mO1eCU1e (StrUCtUre 3),
Or the hYbrid m~IeCU1e Le~/SLe~, the POSitiOna1 i50m~r Of
StrUCtUre 1~
~igUre 21 dePiOtS the re1atiVe adheSiOn Of NS-1 Ce11S
eXPre8Sing E-Se1eCtin On YariOUS "glyco-liposomes" COated On a
P1aStiC SUrfaCe. Pane1 21A 5hOWS the reSU1t Of SUCh re1atiVe
adheSiOn in ~ dYnamiC f1QW Setting Under midd1e ~hear Stre~S
COnditiOnS (7.75 dYneS/~m2~. The firSt S~V~n barS indiCate
relative adhe~;ion of NS-1 cells to SLeX on each glycoliposome as
indicated . Cpd I is structure 1 of ~igure 2 0 and Cpd II is
structure 2 of Figure 20. Bars 8-10 show a mixture o~ LeX Wîth
-- 21 --
WO93/17033 ~ g9~r~ PCl/U593/01?-~
different types of compounds as indicated. The value of relative
adhesion is expressed in comparison with the adhesion of
SLe -liposome as 100~. Values represent the mean of five
determinations. Panel 21B i~dicates the same relative adhesi.on
of NS-l cells at high shear stress conditions (11.8 dynes/cm2).
The value is expressed in texms of the adhesion on SLeX-coated
plates. Values represent the mean of five determinations.
Figure 22 depicts the relative adhesion of NS-l cells
expressing E-selectin on various glycoliposomes coated on plastic
plates at different shear stress conditions. CPD I and CPD II
are structures 1 and 2 o~ Figure 20. Enhancement of adhesion on
CPD I-coated plates was noted only at middle to high shear stress
conditions. The ordinate indicates ~he relatiYe adhesion as
compared with that of the SLeX liposome. The abscissa indicates
the wall shear stress in dynamic flow in dynes/cm2. DSI
represents disialosyl-I antigen.
Figure 23 depicts cell numbers bound per ~uare millimeter
: on various glycoliposomes coated on a plastic surface with
different glycolipid concentrations. Note that structuré 1 of
Figure 20 adheres E-selectin-expressi~g cells much more avidly
than on SLeX-coated plates at high shear stress. The difference
iæ not as stsrk at low shear stress. The ordinate indicates the
number of cells ~ound per millimeter and the abscissa indicates
glycolipid concentration in ~m. Each point is the mean of five
determination~
Figure 24~ depicts adhesion of NS-l cells expressing
E-selectin on glycoliposomes having a mixture o~ SLeX and various
other glycolipids. The ordinate shows the number of cells
- 22 -
-- WO g3~17033 2 1 2 9 9 ~ 7 PCT/US93/01375
adhered per field. The solid circle is SLeX + SPG. The hollow
circle is SLeX ~ H2. The solid triangle is S~e ~ LeX. The
hollow triangle is SLeX + LeY. Each point is the mean of five
de~erminations.
Detailed DescriPtion of the Invention
As noted above, the instant invention in one aspect is
directed ko m~tho~s and compositions for the inhibition of tumor
c~ll metastasis potential and invasivene~s. Numerous tumor cells
posses~ the ability to metastasize, i.e., to ~orm a secondary
tumor colony at a distant site. Sourceæ of malignant tumor cells
include melanoma, lung, breast, colorectal and urogenital
: cancers, such as bladder and prostate ca~cers. Within the
instant invention, ~he metastasis potential of tumor cells,
(i.e., the ability of tumor cells to metastasize) may be
inhibited through the use of (a) tumor-associated carbohydrate
antigens (TACA's, a used hersin TACA is meant to include LACA);
; (b) antibodie~ dirPcted to those TACA's; (c~ oligosaccharide
: components of those TACA's; (d) conjugates of s~ch TACA's or of
oligosaccharide components of 9uch TACA's, æuch as multivalent
~ conjugates of lysyllysine or TACA-bearing glycosphingolipid (GSL)
:~ ~ liposomes; or te) mimetic of the TACA's. Generally, unless
indicated to the contrary, tumor cells and leukocytes are
substantial equivalents inasmuch as both bind to endothelial
cells by carbohydrate structures.
; TAC~ epitopes play essential roles in tllmor cell adhesion
:~ through interactiun with endothelial cells, platelets and
basement membranes, whereby tumor metastasis and invasion may
- 23 -
WO 93/1~033 ~29 ~ PCI ~US~3/01 ~
occur. The mechanism of adhesion may be based on carbohydrate
(CHO) CHO-CH0 interaction, CHO~lectin interaction or CHO-selectin
~amily interaction.
Adhesion of various tumor cells to non-activated endothelial
cells is mediated initially by carbohydrate to carbohydrate
interactions, which in turn, trigger activation of endothelial
cells to expre~s selectins, such as ELAM-1 and GMP-140, Kojima
& Hakomori, J. Biol. Chem., 266:17552, 1991; Kojima et al., J.
Biol. Chem., 267:17264, 1992; Hakomori, Histochem. J., 24:771,
1992. Subsequently, adhesion of various tumor cells to activated
~ndoth~lial cells and platelets is mediated primarily by the
LECCAM or selectin superfamily (e.g., ELAM-1 and GMP-140).
Tumor cell adhesion mediated by sial~syl-LeX is inhibited by
anti-sialosyl-LeX monoclonal antibodies (FH6, CSLEX, SNH3 and
SNH4) and tumor cell adhesion mediated by monosialosyl-Le I is
inhibited by monoclonal antibodies (CAl9-9, CSLEA, NKHl and NKH2)
directed to that epitope. Handa et al., Biochem. BiophysO Pces.
Commun. 181:1223, 1991; Kojima et al,,, Biochem. Biophys. Res.
Commun. 182 0 ~88 , 1992 ; Halcomori , Histochem~ J ., 24 : 771 ,,;~992 .
The adhesion of Colo2û5 tumor cells, which express
predominantly type 1 chain sialos~ and to a lesser extent
sialosyl-LeX, to endothelial ce~ l~ is inhibited by
anti~sialc)syl~ monoclonal antibody and to a }esser extent by
anti-sialosyl-le't monoclonal antibody . Tho~3e f indings suggest
that no~: vnly sialosyl-LeX, but also ~;ialosyl-Lee, are the
mport:ant ligands recognized by EI~l-l and GMP-140 (previously
termed CD62 or PADGEM and also known as E-selectin and
P-selectin)
-- 24 --
:
.~ WO93/1~033 2 1 2 ~ 9 8 7 PCT/US93/0137~
It is known now that adhesion of tumor cells to activated
endothelial cells is based al50 on recognition of
monosialosyl-Le~ II and disialosyl-Le . Both monosialosyl-Le II
and disialosyl-Le~ are defined by monoclonal antibody FH7, which
is known to inhibit strongly adhesion of various types of
epithelial cancer cells (particularly colorectal,
gastrointestinal and pancreatic) to activated endothelial cells
or platelets ~ia selectins.
In particular, GMP-140 is the major selectin (LECCAM)
located on a-granules of platelets or Weibel-Pallade bodies of
endothelial cells (EC's). on activation of those cells, GMP-140
is redistributed rapidly to the cell surface, where it plays an
important role in adhesion of platelets or EC's to certain
carbohydrate epitopes~expressed on blood cells or tumor cells,
resulting in aggregation of platelets or tumor cells, or adhesion
~thereof to capillary endothelia. GMP-140-mediated cell adhesion
is believed by the instant inventors to be involved in initiation
; of~ ~etastatic deposition of tumor cells and i.nitiation of
~ inflammptory processes. -;'
20~ ~ Also, ~ELA~-l is expressed on endothelial cells after
activation with ~interleukin~l, TGF-~, TNF- or
li opolysaccharide.~ ~ELAM-l-mediated cell adhesion also is
believed to be involved in initiation of metastatic deposition
:::
of tumor cells.
: ~
~ 25 ~ Thus, the instant invention in another aspect is directed
`
to inhibiting GMP-140-mediated ox ELAM-l-mediated cell
aggregation~or adhesion, especially at tumor cell sites. Within
the instant invention, GMP-140-mediated or EL~M-l-mediated cell
:
- 25 -
:
.
WO93/17033 ~299 PCT/US93/01~
aggregation or adhesion can be inhibited through the use of
(a) a hybrid sugar, such as LeX/SLex; (b) a mixture of sugars
whi~h are the components of a hybrid sugar (a), such as LeX and
S~e ; (c) monosialosyl-Le I~ Le , Le , monosialosyl-Le II,
S disialosyl-Lea or sialosyl LeX; (d) antibodies that specifically
bind to a hybrid sugar (a), sUch as LeX/SLex, monosialosyl-LR4 I,
Lea ~ LeX ~ monosialosyl-~ea II, disialosyl-Le~ or sialosyl LeX;
(e) a mixture of antibodies, particularly to the components of
a hybrid sugar, such as to SLeX and LeX; (~) oligosaccharide
components of a hybrid sugar, such as LeX/SLex, monosialo~yl-Le~
I, Le~, LeX, monosialosyl-Le~ II, disialosyl-Le~ or sialosyl LeX;
(g) conjugates o~ a hybrid sugar, such as LeX/SLRx,
~onosialosyl-Le~ I, Len~ LeX~ monosialosyl-Le II, disialosyl~Le~
or sialosyl LeX or of the oligosaccharide components; and
(h) mimetics of a hybrid su~ar, such as LeX/SLex
monosialosyl-Lea I, Leh, LeX, monosialosyl-Lea II, disialosyl-LR
: ~: x
or sialosyl Le.
Within the instant invention, tumor metastasis and invasion ~:
is inhibited by blocking tumor c:ell adhesion ~4ereby
~ 20 a:ignificantly reducing or eliminating the spread of metastatic
:~ ~ cells.
P,lso within the instant invention, tumor metastasis and
~ invasion is minimized by inhibiting~ GMP-140-mediated tumor
: : : cell aggregation or adhesion at a tumor site due to: (a)
25~ adhesion of tumor cells to platelets, (b) adhesion of tumor cells
to tumor cells via platelets,- (c) adhesion of tumor cells to EC's
~: via platelets and (d) adhesion of tumor cells to EC's directly
: via G~P-140; and (2) ELAM-l-mediated tumor .cell aggregation or
: - ~6 -
: :
. W093/t7033 2 1 2 ~ 9 8 7 PCT/US93/0137~
adhesion at a tumor site due to adhesion of cells to EC e s
directly via EL~M-l.
Further within the instant invention, inflammation is
minimized by inhibiting GMP~140-mediated leukocyte aggr~gation,
adhesion or migration at a potential site of inflammation due to:
(a) adhesion of ~eukocytes to platelets, (b) adhesion of
leukocytes to endothelial cells (EC) via platelets, (c) adhesion
of leukocyt~s to EC's directly via selectin and (d)
transendothelial migration of leukocytes.
TACA's suitable for use within the instant invention are
those showing differenti~l prognostic significance (i.e., TACA's
that may be correlated clearly with invasive or metastatic
potential). Within the context o~ the instant invention, such
TA~A's may be distinguished through a comparison of invasiveness,
metas~asis and clinic2l prognosis of similar tumors showing
expression vs. non-expression of such TACA's. Pre~erred TACA'æ
for use within the present invention include H/LeY/Leb,
sialosyl-LeX (SA-LeX or S~eX), L~, LeX, monosialosyl-Le~ I (5Le~
or SA-Le) and sialosyl-Tn (SA-Tn or STn). Derivatives ~ such
ao ~ACA'~ include hybrid sugar, such as LeX/S~ex, dimeri LeX,
sialosyl dimeric~ rifuscosyl LeY, disialo yl~ and
monosialosyl Le' II.
~ s noted above, ~ACA's for use within the instarlt invention
; exhibit a differential prognos~ic significance. By way of
:2:5 example, such a differential prognostic significance may be
illustrated by the fact:that tumors expres~ing H/LRY~Leb antigens
(as defined by monoclonal antibody M~A-15-5) showed much worse
patient prognosis than tumors not expressing.those antigens. For
- 27 -
W093/17033 2~ 2 g9~ I PCl/US93/01~-~
instance; as shown in Figure 3~, patients with squamous cell lung
carcinoma expressing H/LeY/Le had only an 11% survival over a
5-year period (i.e., 8~% died) whereas comparable patients not
expressing H/LeY/Leb had an approximat~ly 62% survival over the
same period.
Similar results were obtained for tumors showing expression
vs. non-expression of sialosyl-LeX and sialosyl Tn antigens.
More specifically, as shown in Figure 3B, patients with colonic
cancer e~pressing sialosyl-LRX had only a 15% survival over a
5~year period, whereas comparable patients not expressing that
antigen had an approximately 50~ survival over that period. In
a separate study, ~he 5-year survival o~ patients with
early-ætage colonic cancer not expres5ing sialo~yl-Tn was 100%,
as compared to 75% for patien~s ~ho expressed sialosyl-Tn ~see
Figure 3~). A~ shown in Figure 3D, similar but more obvious
differences were observed in patients with ovarian cancer showing
expre~sion vs. non-expression of sialosyl-Tn antigen.
Also as noted above, antibodies or a mixture of antibodies
to suitable TACA ' may be employed within the context,of the
instant inVention. As usea herein, such antibodies include both
monoclonal and polyclonal antibodies and may be intact molecules,
: a fragment of such a molecule or a functional equivalent thereof.
Th~ antibody may be engineered genetically~ Examples o~ antibody
fra~ments include F(ab'~2, Fab', Fab and Fv.
: 25 Briefly, polyclonal antibodies may be produced by
immunization of an animal and subseguent collection of sera
therefrom~ Immunization is accomplished~ for example, by a
systemic administration, such as by subcutaneous, intrasplenic
- 28 -
-- W093/17033 2 1 2 9 9 ~ 7 PCT/US93/Ot37~
or intramuscular injection, into a rabbit, rat or mouse. It is
pre~erred generally to follow the initial immunization with one
or more booster immunizations prior to sera collection. Such
methodology is well known and described in a number of
references.
While polyclonal antibodies may be employed in the present
invention, monoclonal antibodies are preferred. Monoclonal
antibodies suitable for use within the instant invention include
`~ those of murine or human origin, or chimeric antibodies such as
~10 those which combine portions of both human and murine antibodies
(i.e~, antigen binding region of murine antibody plus constant
regions of human antibody). Human and chimeric antibodies may
be produced using methods known by those skilled in the art.
; Human antibodies and ohimeric human-mouse ant~bodies are
15~ ~ advantageous because suoh antibodies are less likely than murine
antibodies to cause the~production of anti-antibodies when
~ : : :
administered clinically.~
Monoclonal antibodies may be produced generally by the
~ method of Rohler and~ M~ilstein (Nature 256:495/ 1975; 9~r. J.
;~ 20 ~ I~munol. 6:511, 1976), as~well as by various techniques which
modify the initial method :of~Kohler and Milstein (see Harlow and
Lane~(eds.)~,~"Antibodies~ A Laboratory Manual", Cold Spring
` Harbor~ Laboratory, ~1988, ~which is herein incorporated by
` ; referenoe in its entirety).
~25 ~ Briefly,~ the ;;lymph" nodes and/or -~pleen of an animal
immunized~ with one of the~ TACA's or the oligosaacharide
co~ponents thereof are~ fused with myelo~a cells to form hybrid
cell lines ("hybridomas" or "clones"). Each hybridoma secretes
29 -
::: ~ ` : `
:~
~9 9 PCr/US93/01~ ~
a sing1e type of i~nunoglobulin and, like the myeloma cells, has
the potential for indefinite cell division. It may be desirable
to coup1e su~h mo1ecu1es to a carrier to increase immunogenicity.
Suitable carriers include keyhole 1impet hemocyanin,
thyrog10bu1in, bovine serum albumin and derivatives thereof.
An alternative to the production of monoclonal antibodies
via hybridomas is the creation of monoclonal antibodies
expression libraries using bacteriophage and bacteria (e.g.,
Sastry et al., Proc. Matl. Acad. Sci. USA 8~:5728, 1989; Huse
et al., Science 246:1275, 1289). Se1ection of antibodies
exhibiting appropriate specificity may be performed in a variety
of ways which will be evident to those skilled in the art.
Representative examples of monoc10na1 antibodies suitable
~or u e within the present invention include MIA-15 5 (Miyake ~
15~ Hakvmori, Biochem. 30:3328, 19g1~, as well as the monoc10na1
~ntibodies cited in Hakomori, Advances In C~ncer Research
52:257-331, 1989.
As discussed above, oligosaccharide compon~nts of suitable
: TACA's also may be used ~in the instant invention~ ~ used
.:
: 20 herein, the term "oligosa charide" includes natura11y deri~ed
oligosaccharides, synthetically prepared ~nd mimetic derivati~es
o~ either, including port1ons of a TACA oligosaccharide
component. - ~
~ Additional oligosaccharide components useful in the instant
:~ 25 invention include lactose and 1actose derivatives, such as methyl
D-lactoside, 1act-N-tetrose (Ga1~3G1cNAc~1~3Ga1~1~4G1c) and
ph~nyl ~-D-thiolactoside. For example, b~th compounds were found
to inhibit melanoma cell metastasis in the mous~ lun~. Other
- 30 -
.
2129987
-- WO 93/1703~ PCI/US93tO137~
lactose derivatives also may be used, including ethyl or phenyl
lactoside and methyl or ethyl thiolactoside. It is believed that
such lactose derivatives block binding of melan~ma cells to EC's
by inhibiting melanoma cell GM3 ganglioside interaction with
S lactosyl ceramide of the EC's.
Other oligosaccharide components suitable for inhibiting
metastasis potential of cells of a particular ~umor may be
identified based on determining the structure of specific
carbohydrate chain(s) which are involved in the ability of the
tumor to metastasize. The identification of
carbohydrate-containing molecules involved in the ability of a
tumor to metastasize may be accomplished in a variety of ways,
including through the use of glycosidases and inhibitors of
glycosyltransferases.
~15 The structure of carbohydrates bound to either lipids or
proteins may be determined based on degradation, mass
spectrometry, including electron-impact direct~probe (EI) and
fast ato~ bombardment (FAB), and methylation analysis (techniques
~:: described, for example,: in Nudelman et al., J. Biol_~Chem.
261OS487, 1986). Degradation analysis may be accomplished
ahemiaally and/or enzymatically, e.g., by glycosida~es. The
: car~ohydrate sequence suggested~by degradation analysis may be
determined by methylation analysis (Hakomori, J. Bio~hem. 55:205,
~ 196~) followed b~ chemical ionization mass spectrDmetry of
25~ per~ethylated sugars (Stellner et al.l ~rch. Biochem. ~iophys.
: ~55:464, 1974; Levery et al., Meth. Enzymol. 138:13, 1987).
: ~ Alternatively,:or in conjunction with those techniques, EI
mass spectrometry may be performed on permethylated glycans or
: '
w093/1703~ ~ 2 9 9 ~ ~ PCT/ US93/01-~7~
after the appropriate degradation of intact glycans
(Kannagi et al., J. Biol. Chem. 259:8444, 1984; Nudelman et al.,
J. Biol. Chem. 263:13942, l~88~. Homogeneity of the carbohydrate
sequence may be demonstrated based on various chemical and
S physical criteria, including proton NMR spectroscopy of intact
or methylated glycans and FAB mass spectrometry~ Once the
carbohydrate sequence has been determined, it will be `evident to
those of ordinary skill in the art to select an appropriate
~ oligosaccharide for inhibiting the metastasis potential of a
: lO tumor cell.
As briefly discussed above, conjugates of suitable TACA's
or oligosaccharide components thereof, such as multi~alent
conjugates with lysyllysine or TACA-bearing glycosphingolipid
. .
~ tGSL) liposomes tor glyco-liposomes), also may be used in the
::
~: ~l5 instant invention.
The components of the conjugate may be coupled covalently
to one another either directly or via a linker group. A direct
re w tion between components~ is possible when each possesses a
;substituent capable of reacting~with the other~ For ex~mple, a
nucleophilic group, such as::an amino or sulfhydryl group, on one
:component may~be:capable:of reacting with a carbonyl-containing
group,:such as an anhydride or an acyl halide, or with an alkyl
group containing a:1eaving group, e.g., a halide, on the other.
It may be desirable to couple covalently components via a
25~: 1inker~9roup.~ A lin~-r~group can serve to increase the chemical
:: ~ reac~ivity of a substituent and: thus increase the coupling
~ efficiency. ~n increase :in chemical reactivity also may
; :~ : facilitate the use of~functional groups on components which would
~ ~:
~ ~ - 32 -
212~9~7
~-, ~093/17033 PCT/US93/0137~
not otherwise be possible. For example, a carboxyl group may be
activated. Activation of a carboxyl group includes formation of
an "active ester", such as a succinimidyl ester. The term
"active ester" is known to refer to esters Which are highly
reactive in nucleophilic su~stitution reactions.
Alternatively, it may be desirable to produce conjugates in
which the components are linked non-covalently. For example, one
or more TACA's may be incorporated into the outer surface of
glycosphingolipid (GSL) liposomes.
It may be desirable to increase the in vivo half life of an
oligosaccharide. ~s disclosed in the instant invention,
oligosaccharides may be coupled to (i.e. t coval~ntly bonded to)
a straight-chain amphophilic polymer, such as polyethyleneglycol.
A represe~tative example of a method for producing an
oligosaccharide-poly~thyleneglycol conjugate is the reaction of
: an oligosaccharide, which has been derivatized to contain a
uccinimidyl group, with a polyethyleneglycol having a terminal
amino group. The latter compound has a general formula of
NH -~CH2CH2-O)n-CH3, where n typically averages ~ 44.7
(iaR~ mslecular weight of ~bout 2,000) to 112.9 ~i.e., molecular
weight of about 5,000~)~
Addition~lly, because the cell adhesion mediated by
selectins, EL~M-1 Qr GMP-140, is based on recognition of
sialylated and fucosylated lactos~ries type 1 and type 2 chains
by a lectin se~uence domain present at the N~t~rminal region of
the selectin molecules,: any structure which may show more
effective blocking activity of the lectin domain than naturally
occurring epitopes are useful in the present invention. Such
- 33 -
WO93/17033 ~ PCT~US93/01~
9~l
unnatura~ synthetic compound~, termed "mimetic~", of, for
example, sialosyl-~eX or sialosyl-Lea I or II, which mimic the
surface structure of naturally occurring epitopes but show better
blocking activity o~ carbohydrate-dependent adhesion, can be
considered.
Examples of useful mimetics include, but ar~ not limited to,
sialosyl-LeX or monosialosyl-~e0 I or II having
trifluoro-L-fuco~e, N-trifluoro-acetyl-glucosamine or a
heterocyclic or aromatic ring structure having a sialic acid
analog and fucose analog at the same distance and spacial
configuration as those found in naturally occurring sialosyl-LeX,
monosialosyl-Lea I and II, or the H/LeY/Leb structure having
trifluoro-L-fucose, N-trifluoro-acetyl-glucosamine or sialosyl-Tn
analogs containing N-tri~luoro-acetyl-neuraminic acid.
Thus, a modified carbohydr~te epitope, or any other
~'~imetic" mimicking the sur~ace structure of a carbohydrate
epitope, which blocks cell adhesion through tum~r-associated
carbohydrates more e~ficiently than a naturally occurring epitope
is within the scope of th~ instant invention. -~
The inhibition of metastasis potential of tumor cells and
GMP-140-mediated or ELaN-l-~ediated cell aggragation or adhesion
:
` have a variety of in vitro and in vivo uses, e.g., treatment of
: isolated tumor cells or tumor-bearing hosts and treatment of
di~ease processes involving GMP-140 or ELAM 1.
Regarding in vitro aspects, as noted above, the instant
invent~ion pro~ides a method for inhibiting tumor cell metastasis
potential within a biologic preparation. The ~ethod comprises
incubating a biologic preparation with at least one agent
- 34 - .
-~ WO93/17033 2 1 2 9 9 8 7 PCT/US93/01375
selected from the group consisting of (a) tumor-associated
carbohydrate antigens that exhibit differential prognostic
significance, (b) antibodies that specifically bind to those
antigens, (c~ oligosaccharide components of those antigens,
5 (d) conjugates of those antigens or oligosaccharide components
and (e) mimetics of the tumor-associated carbohydrate antigens,
the agent inhibiting the metastasis potential o~ the preparation.
Regarding further in vitro aspects, the instant invention
also provides a method for inhibiting GMP-l40-mediated or
EL~M-l-~ediated cell aggregation or adhesion in a biologic
preparation. The method comprises incubating ths biologic
preparation with at least one agent selected from the group
consisting o~ (a) a hybrid sugar, such as Le~/SLeX; ~b) sugar
~ component~ of a hybrid sugar ta), such as ~ex and SLeX;
:~ 15 (c) monosialosyl-Le- I, Le~ LRX monosialosyl-~e~
di~ialosyl-~R~ or sialosyl LeX; ~d) antibodies that specifically
bind to a hybrid sugar, ~uch as LeX/SLex or to the component
sugars thereof, monosialosyl-Le~I, Le~, LeX, monosialosyl ~ea II~
di~ialosyl-L~ or sialosyl LeX; (e) oligosaccharide co~bnents
vf a hybrid sugar,:suc~ a~ /SLeXO monosialosyl-Le~ I, ~e~, LeX,
; monos~alosyl-Le' II, diæialo~yl-Le3 or ~ialosyl LeX; (f)
aonjugates of a ~hybrid Rugar ~ such as LeX/SL~x,
monosialosyl-Le~ I,:;Le', LeX, monosialosyl-Le~ II, disialosyl-LeD
or sialosyl LeX or o f the oligosaccharide components; and
(g) mimetics of ~a :hybrid sugar, such as SLeX/Lex,
nosialosyl-Le~ Le-, LeX, monosialosyl-Le II, disialosyl-Le
or sialosyl LeX, the agent inhibiting the cell aggregation or
adhesion.
~ - 35 -
W093/17033 ~ ~L~99a~ PCT/US93/0 ~
Suitable biologic preparations include cell cultures and
cell suspensions in biologic fluids, such as blood, urine, lymph,
synovial and cerebrospinal fluid. TACA's, oligosaccharides or
conjugates thereof generally will be incubated at a final
concentration of about O.l to l M, and typically at about 0.2 to
0.5 M. Incubation is performed typically for 5 to 15 minutes at
37-C. After treatment of a biologic preparation, the preparation
may be injected or implanted in an animal, e.g./ to confirm
effectiveness of the inhibition of metastasis potential.
10The instant in~ention also provides a method for inhibiting
tumor cell metastasis potential in a wa~m-blooded animal, such
as a human. The method comprises administering to a warm-blooded
animal an effective amount of at least one agent selected from
the group consisting of (a) tumor-associated carbohydrate
antigens that exhibit differential prognostic significance,
tb) antibodi~s that specifically bind to those antigens,
c) oligosaccharide components of those antigens, (d) conjugates
of those antigens: or the oligosaccharide components and
(e) m~imetics of monosialosyl-Le I, Le , Le , monosialosyl~ea II,
.
~ 2~ disialosyl-Ie~ or ~sialosyl LeX, the agent inhibiting the
: ~
metastasis potential of the preparation.
Similarly, the instant invention also provides a met~od for
inhibiting GMP-140-mediated or ELAM-l-mediated cell aggregation
or adhesion at a tumor cell site in a warm-blooded animal. The
method~ comprises~administering to a warm-blosde~ animal an
- èffective amount of at~least one agent selected from the group
consisting ~of (a)~ a hybrid sugar, such as Le~/S~eX;
; (b) component sugars of a hybrid sugar ~a), such as LeX and
~ - 36 -
_~ W093/17033 2 1 2 9 9 ~ 7 P~T/US93/0137~
SLe ; (c) mo~osialosyl-Le I, Le , Le , monoslalosyl-Le II,
disialosyl-Led or sialosyl LRX; (d) antibodies that specifically
bind to a hybrid sugar, such as LeX/SLeX, monosialosyl-Le~ I, Le~,
LeX, monosialosyl-Le~ II, disialosyl-Le~ or sialosyl LeX;
(e) oligosaccharide components of a hybrid sugar, such a~
LeX/S~eX, monosialosyl-Le~ I, monosialosyl-LR~ II, Le~, LeX,
disialoæyl-Lea or sialosyl LeX; (f) conjugates of a hybrid
sugar, such as LeX/SLe , monosialosyl-Le I, Le , Le ,
monosialosyl Le~ disialosyl-Le or sialosyl Le or of the
oligosaccharide components; and (g) mimetics of a hybrid sugar,
: such as LeX/SLe~, monosialosyl-Le~ I, LeA, Lex
monosialosyl-Lea II, disialosyl-Lea or sialosyl LeX, said agent
reducing the metastatic potential at the tumor site in the
warm-blooded animal.
,
The instant invention also provides a method for inhibiting
: GMP 140-mediated cell aggregation or adhesion at an inf~ammation
site in a warm-blooded animal.
; The method~comprises: administering to warm-blooded animal
an effective~amount of at least one agent selected from the~group
20~ consisting ~o~f: ~ (a) :~a: hybrid sugar, such as LRX/SLex;
(b)~ component sugars~of a hyb~id :sugar (a), such as LeX and
S~eX~ (c)~ ~onosialosyl-Lea I, Le~, LeX, monosialosyl-Le~ II,
disialosyl-IA~ or sia10syl~LeX; (d) antibodies that specifically
bind to 8 hybrid'sugar,;~such as LeXJSLex, monosialosyl-Le~ I, ~e~,
25~ LRX,; ;~onosialosyl-LeD~ II:, disialosyl-Le or sialosyl LeX;
:(e~ o1igosaccharide :components of a hybrid sugar, such as
SLeX,~ monosialosyl-~e' I, Le~, LeX, monosialosyl-Le II,
::
disialosyl-Le~ or si:alosyl LeX; (f) conjugates of a hybrid
37 -
:~: : : :
W~93/17033 ~ 99~ PC~/US93/0~-7~
sugar, such as SLe /LR , monosialosyl-Le I, Le , Le ,
monosialosyl-Le~ II, disialosyl-LR~ or sialosyl LsX or of the
oligosaccharide components; and (g) mimetics of a hybrid sugar,
such as LeX/SLex, monosialosyl-LR~ I, Le , LR ,
monosialosyl-Le~ II, disialosyl-Le~ or sialosyl LRX, the agent
reducing the inflammatory potential at the inflammatory site in
the warm-blooded animal.
For ~oth method~, TACA's, oligosaccharides or conjugates
thereof generally will be administered at ~ concentration of
about O.l to l M and typically at about 0.2 to Q.5 M. It will be
evident to those skilled in the art how to determine the optimal
effective dose for a particular substance, e.g., based on in
vitro and in vivo studies in non-human ani~als. A variety of
: routes of administration may be used. Typically, admini~tration
will be intravenous or intraca~i~ary, e.g., în the pleural or
peritoneal cavities, in the bed of a resected tumor or at a site
of inflammation.
A TACA, antibody, oligosaccharide or derivative as discussed
aboYe may be administered in combination with a pharmaceu~cally
acceptable carrier or diluent, such as physiologic salins.
Moreover, the agents that inhibit or reduce metastatic potential
may be administered in combination with an immunotherapeutic or
chemotherapeutic substance, and the agents that reduce
inflammato~y po~ential may be administered in combination with
an anti-inflammatory substance.
:~ When ~ combination of such an agent and a substance is
d~sirad, each compound may bs administered sequentially,
simultaneously or co~bined and administered as a single
- 38 -
,~-....................... W093~17033 2 1 2 9 9 8 7 PCT/US9310137~
composition. Diagnostic techniques, such as CAT scans, may be
performed prior to and subsequent to administration to confirm
the effectiveness of the inhibition of metastatic potential or
inflammatory potential.
One in vitro system for measuring adhesion or aggregation
of tumor cells to other cells ~e~g. EC's), or for determining
successful inhibition of adhesion or aggregation is a dynamic
flow system similar to that described by M. B. ~awrence et al.
: (Blood 70:1284, 1987) and which is shown in Figures llA, llB, llC
;: 10 and llD.
A parallel-plate laminar flow chamber ~ hown upside down
for convenience) connected to a pressure pump ~2) via tubing ~18)
iæ used to simulate the flow shear stresses present in
physiological microvascular environments. The flow chamber
:
~:: 15 ~consists of a plastic or :glass cover slip (3) resting on a
chamber body ~16~ on which a parallel, transparent plastia
,
surface t~) is attached with a~rubber or silicone gasket 15);
there is a 114 ~m~gap;between the two surfaces, and this gap is
connected to an inlet 810t ~C~ connected to an inlet manif~ld 18)
;~: 20~ ~and~outlet slot: ~7) connected to an outlet manifold tl9)
(Figure~ A). ~A laminar flow:with defined rate and wall shear
stress is achieved: by~manipulation of the pressure pump 12),
~:~ which is connected to the inlet manifold (8) of the flow chamber
via :tubing ~18~.~ Figure llB depicts the configuration of an
~25:~ asse~bled:flow chamber ~
: To~fix together the:;:cover slip ~3) and the chamber body ~6S
:very tightly, there~is:a:continuous circular grooved space ~20)
: ~ on:the:periphery~of the chamber body 116). The circular, grooved
: : :- 39 -
'~
W093/17033 PCT/US93/Ol'~
,09sa~
spa nects to a vacuum pump by placement of the rubber or
silicone rubber gasket (5) with cover slip (3) on top. Thus, by
applying va~uum (21) in (20) the c~ver slip (3) and chamber body
~16) are affixed strongly and immovable. The thin inlet and
outlet slots in the chamber body (16) open to the inlet and
outlet manifold~, respectively. The outlet manifold is connected
to a pressure pump (~) which can be operated in either a negative
or positive mode.
: Cells (e.g., endothelial cells) are grown o~ either a glass
:: 10 or plastic cover slip ~3), or various adhesion molecules are
affixed on ~3), and a tumor cell suspension in medium flows from
inlet manifold (16) to outlet manifold ~19). The structure of
the flow chamber (1) in Figure llB is shown upside down for
convenience. The chamber is placed under an inverted microscop~
:~15 : stage, right side up (Figure llC), and the flow of tumor cells
over tho cell layer (e.g.~, endothelial cell layer) is observed
~ under the microsGopo.~ The observed pattern of rolling and
:~ : stopping (i.e., pattern :of adhosion) of tumor cells can b~
recorded on videotape.
20~ Turning ~to Figure llC, khe cells ~9) are grown as a
monolayer, or adhosion moloaules are affix~d, on the cover slip
(3) and a laminar flow of~tumor cell suspension ~ aintained
in a vessel in a water~bath (17), is passed through the chamber
via tubing ~18). Coll:movements are observed under an inverted
:~ 25 phaso-contrast micro8cope ~ ~10) and recorded by time-lapse
videocassette recorder ~ll) using a video camera ~12) and a
:digital image procossor ~l3). Adhesion is observed as rolling
followed by stopping of cells. Number of cells bound during a
- 40 -
~- W093/17033 2 1 2 ~ 9 8 7 PCT/US~3/01375
set time, e.g. 3 minutes, at different shear stresses, e.g., from
0~4 to 4.8 dynes/cm2, are counted from several fields recorded on
videotape (Figure llB). Wall shear stress (T) is calculated as
3~Q/2ba2, where ~ = coefficient of viscosity, e.g. 1.0 cP,
Q = volumetric flow rate (cm3/sec), a = half channel height, e.g.
5.7 x 103 cm, and b = channel width, e.g. 1.3 cm.
Figure llD schematically shows laminar flow of tumor cell
suspe~sion ~1~) through a chamber in which one surface is coated
with endothelial cells ~9). Rolling or stopped cells ~15) are
observed under an inverted microscope and recorded on videotape,
as described above. The arrows indicate the direction of flow
of the tumor cell suspension ~14).
As mentioned above,~the in~tant invention also pro~ides a
method for identifying a TACA epitope to which lectin activity
~15 of a selectin, such as GMP-140, is directed.
Previously, the TACA epitopes we~e studied based on the
~ i~nhibitory effect of ~arious glycosphingolipids (GSL's), GSL
; ~ ~ oligosaccharides or GS~-containing liposomes on adhesion of blood
cells or tumor cells to a solid phase (e~g., a plastic ~urface)
~;~20 ~oated wlth activated platelets. In practice, that meant coating
a solid phase with~gelatin,~which was in turn was c~ated with
aativated~ platelets;~ platelets bind readily to a gelatin-coated
solid phase via GpIIb/IIIa, the major platelet integrin receptor.
In studies using that method, binding of promyelocytic
25 ~ ieukemia HL60 cells;;to platelet-coated solid phase was inhibited
~by liposomes~containing a sialosyl-LeX determinant, but not by
,
liposomes containing sialosylparagloboside (SPG), and only weakly
by liposomes containing al~3 fucosylated type 2 chain (Le
~ - 41 -
: : :
W093/17~ ,99a~ ' PCr/US93/01-'S
~see Table 1 in Example 3 below). ~hose results suggested that
sialosyl-LeX is the carbohydrate epitope defined by GMP-140
(Polley et al., Proc. Natl. Acad. Sci. USA 88: 6224, 1991).
However, the me~ od described above had an important
limitation: no cell line which expresses exclusively type 1
chain GSL is available. Myelogenous cell line ~L60 and monocytic
cell line U937 express exclusively type 2 chain and little, if
any, type 1 chain. On the other hand, all known human tumor cell
lines derived from colonic, gastric or lung carcinoma express
both ~ype 1 and type 2 chains. For those reasons, it is
di~ficult to determin~ the real epitope to which GMP-140 binds.
TQ address that problem, a new methodology was developed, as
described below.
Fluorescent plasti~ (e.g. polystyrene) beads
(diameter ~ O.S ~m~ are coated with GS~. GSL's are known to be
adsorbed strongly on such beads, which allows construction o~
fiuorescent probes containing sp~cific GSL's. Platelets
(activated or non-activated) are incubated with such GSL-coat~d
beadæ, follQwed by determi~ation of platelet fluor~cenae
intensi~y by ~low ~yto~etryO
Using hat method, acti~ated platelets were ~ound to show
muah ~ronger binding to fluorescent beads coated with
monosialosyl ~LRa I (see Table 3) than to bead~ coated with any
related GS~. The binding o~ pl~telets to sialosyl-~2~-coated
b~ads was inhibited~by anti-GMP-140 monoclonal antibody or
anti-sialosyl-Le monoclonal antibody, but not by
anti-sialosyl LRX monoclonal antibody. Although binding of
activated platelets to sialosyl-LeX-coated beads was observable,
42 -
~-.. , WO~3ll7033 2 1 2 9 9 ~ 7 PCT/US93~01375
the level of binding was much lower than bi~ding to
sialosyl-LR~-coated beads~ Those results indicate that the
primary epitope structure d~fined by GMP-140 is ~ialosyl-LR ,
rather than sialosyl-~eXO
Of course, other epitope structures defined by a selectin,
such as ~MP-140, can be identified using the instant inventi~e
method.
ELAM-1 (E-selectin3 is expressed on the sur~ace of act ~ated
endothelial cells. E~A~ 1 has a carbohydrate-binding domain at
~0 the am.ino terminal region and indeed ELAN-1 is known to bind SLeX
and SLeG
Those conclusions were based on the observations that
adheæion betwe~n tumor cells or leukocytes to activated human
E~'s was inhibited by SLeX glycolipid or oligosaccharide, but not
by other tested glycolipids or oligosaccharides available at that
ti~e (Phillips e~ al., Science ~50: 1130, 1990) 4 Later, the
above-noted adhe~ion was found to be inhibited by SLe~ I, the
: :positional isomer of SLeX gly~olipid, as well (Berg et al., J.
;~
Biol. Chem., 266:14869~, lg91; Takada et al., Biochem. Blophys.
~Re~. Commun.~ 189:713,: 1991~
That adhesion re~ction was claimed to be inhibited ~y IgG3
~: anti-SLeX ~Ab (Ph~llips et al., supra). However, there are other
st ~ atural variant~ related to SLeX and SLe~ together with m~b's
directed to those variants and the reported studies are based on
inhibition of selec~in-dependen~ adhesion by assumed epitope
structure(s).
~ The instant invention is a result of systematic studies on
selectin dependent adh sion under static and dynamic
'
: ~ 43 - :
~.
WO93/l70~3~99~ PCT/USs3/o~~7~
circumstances. For example, the methods employed include,
(i) adhesion of tumor cells to IL-l-activated human umbilical
cord endothelial cells (HUVEC); (ii) adhesion of tumor cells to
E~selectin-coated solid supports, for example, by using
recombinant ELAM-l; (iii) adhesion of fluorescent particulate
solid supports coated with glycoliposomes with a~tivated
platelets or NUVEC' 5 expressing P-selectin or E-selectin; and
(i~) adhesion of NS-1 myeloma cells, transfected with E-selectin
coding sequences and permanently expressing E-selectin onto
plates coated with glycoliposomes.
Hence, the systems (i), (ii) and (iii) were employed to
assess the effect on adhesion of various mAb's directed to SLeX,
SLe- I, SLe~ II, LeX, Le~ and related structures; comblnations of
such mAb's; sialidases with various substrate specificities; or
; 15~ combinations of varicus sialidases and mAb'5. The method of (iv)
was~ used to compare the intensity of adhesion under dynamic
conditions.
`Specifically, the instan~ invention r~lates to carbohydrates
defined ~by formulae (I~, (II) and (III) below whi'~ are
~20: ~ characterized by internal sialosyl residues or a branched
` structure. ~: ~
Formula (I) relates to a type 1 or extended type 1 chain
with internal ~2~6 sialosyl:substitutions and an ~1~4 fucosyl
substitution.
~ ~ NeuAc~2
:: Gal~1~3GlcNAc0l~3Gal0l~3GlcNAc01~]n3Gal~l~ (I)
30~ ~1 t ~ : R3
F UC~ 1
: - 44 -
; ~ ~
,~ WO93/17033 PCT/US93/0137~
2I 2~98 7
In formula (I), R1 is H or a sialic acid residue in ~2~3
linkage; R2 is H or a sialic acid residue in ~2~6 linkage; n is
eq~l to or greater than 0: and R3 is H or a fuco~yl residue in
al~4 linkage.
S Formula (II) relates to a type 2 chain structure with
internal sialosyl and fucosyl substitutions.
NeuAc~2
6 12 l2 R~
10Gal~1~4GlcN~t~3Gal~1~4GlcNAc~1~]n3~al~1~ (~I)
R~ t R4
Fuco~l
In formula (II), R2 is as defined for Pormula I, R4 is H or
a fucosyl residue in al~3 linkage and R~ i~ H, a sialic acid
residue in ~2-3 linkage, NeuAc~2~8NeuAc in ~2~3 linkage or
R4~NeuAc in ~2~3 link~gQ, wherein R6 is one ~r more sugars other
than a sialic acid residue and n is equal to or greater than 0.
Formula (III) relates to a type 2 chain s~ructure which is
a hybrid molecule comprising a branch wherein each branch
co~prises an epi~ope of :a single carbohydrate antigen as
disclosed;herein. Hence, as used herein, a hybrid molecul~ does
not~necessarily comprise the entirety of the two component sugars
that co~prise the hybrid. Inst~ad, the hybrid comprises the
:: :
~25 ~ epi~opes of the component sugars. Hence, xeferring to Figure ~Q,
s ~ucture 1 compri~es the~epitopes of LeX and SLeX, however it
will be noted that with reference to the diagrammatic ~tructure~
of the various su~ars: se~ forth herei~below, not all of the LeX
: or SL~X molecules are found in the hybrid. As to the SLeX
. 30 portion of the hybrid, only the terminal galactose and
: glucosamine, together~with the attached fucosyl and sialic acid
- 45 -
W093/17033~ ~99~ I PCT/US93/01~
residues, of the intact SLeX molecule comprise the hybrid.
Similarly, for LR , only the epitope generating terminal three
~ugar residues comprise the hybrid. As used herein, epitope is
that portion of the sugar which interacts in the adhesion
phenomenon.
R101 ~
R8--R~ ( III )
R11l ~3
In formula III, each of ~10 and R11 comprises galactose,
lO Gal~ 4GlcNAc: or Gal,Bl ~3GlcNAc; R8 comprises Gal or GalNAc; and
R~ comprises lactosyl ceramide or an 0-linked sugar.
Additionally, R10 and R11 may comprise fucosyl and sialic acid
residues. The hybrid structures are ident~fi~d by the respective
epitopes contained therein. Hence, structure 1 of Figure 20 i~
denoted SLeX/Lex or LRX/SLex
Formula I is based on inhibition by various mAb's and
`~ sialidas~s and combinations thereo~ of E-selectin-dependent
adhesion of tumor cells (e.g. t Colo201 cells) which ~press
exclu~ively type 1 chsin, i.e~, Gal~ 3GlcNAc~1~3Gal, repeats
thereof and ~ubsti~utions thereof. E-~electin-dependent adhesion
: of Colo20I cells was inhibited only minimally by m~b C~l9~9
(directed to 5 ~) and moderately inhi~ited by mAb FH7 (direck~d
to disialosyl Le8 and monosialosyl ~e~ . Colo201 adhesion was
i~hibited most strongly by mAb ~A3F4 (directed to monosialosy~
Le~ II and L2a or by a combination of CAl9-9 plus CA3F4.
Specific reactivities ~f FH7 with disialosyl Le~ and monosialosyl
- 46 -
S~-! WO 93/17033 2 1 2 ~ 9 ~ 7 PCT/US93/0137S
Le II, and of CA3F4 with monosialosyl Lea II, were described
previously (Nudelman et al., J. Biol. Chem., 261: 5487, 1986).
Further evidence for the epitope structures was based on the
following observations. Treatment of Colo201 cells with
5: Newcastle Disease Virus (NDV) sialidase, which cleaves
NeuAc~2~3Gal (R1 in Formula I) only slighted inhibited
E-selectin-dependent adhesion, but treatment with Arthrobacter
ureafaciens (AU or often denoted as AV in the Figures) or Vibrio
cholerae (VC) sialidases, both which cleave NeuAc~2~6 linkage to
lO~ GlcNAc or Gal (i.e., R2 in Formula I), com~letely inhibited such
adhesion. Thus, involvement of internally 2~6 sialosylated
structures in the adhesion is clear. NDV sialidase in
combination with mAb's C~l9-9 or CA3F4 strongly inhibited the
~ adhesion. The results described above were obtained in both
;~ 15 static and dynamic adhesion systems, described herein.
How~ver, the binding dynamics of selectins is vibrant, as
revealed ~in dynamic flow systems which simulate more closely
physiologic conditions, that is, for example, leukocytes or tumor
c d ls can be ~ovinq at cons~iderable speed in large and unoccluded
~2~0~ small vessels and at a slower speed in occluded ves~els and in
ti88ue spaces:. Under~static conditions cell interactions may be
~ ;mediated~by interaction with a first set of molecules that share
:~:: a common characteriætic, whereas under non-static conditions,
cell interactions may be mediated by interaction with a second
~: set 5f molecules ~hat share a common characteristic, different
from that shared by the~first set of molecules. Furthermore,
;under non-static conditions,~ the binding requirements may vary
depending on the speed at~which the cells are moving.
47 -
: ~
W093/17033 ? ~9~ PCT/US93/O~'S
~ or example, E-selectin (ELAM)-mediated adhesion of HL60
cells is d~pendent on different carbohydrate structures when the
cells are reacted in a stationary or slow moving setting or are
reacted while the cells are in rapidly moving setting. Under
static or low shear condit.ions ELAM binds preferentially to ~2~3
sialylated and ~ 3 fucosylated structures, such as SL2X, while
under high shear conditions, ELAM preferentially binds to other
structures, such as Le~, LeY, H and to various hybrid structures,
such as LeX/SLex.
Formula II is based on inhibition by various mAB's and
sialidases and combinations thereo~ of E-selectin-dependént
adhesion of HL60 tumor cells, which express only type 2 chain,
i.e. Gal~1~4GlcNAc~1~3Gal and repeats thereof, and substitutions
thereof. Treatment of HL60 cells with N~V sialida~e, which
cleaves NeuAc~2~3~al (Rl in Formula ~ completely abolished
reactivity of the cells with anti-SLeX mAb'c, although the cells
xemained strongly adherent ~o E-selectin-coated plat~s and to
~ activa~ed EC's. Complete inhibition of adhesion to E-seleatin
: or EC's re~uired treatment with AU or VC sialidase, which~cleaves
Neu~o~6 linked to GlcNAc or Gal (i.e., R2 in Formula II, in
addition to NeuAc~2~6 as shown in Formula II). The R~ group is
susceptible to cleavage by ~U and VC sialidase but not by N~V
sialidase.
Further evidence for Formula II was provided by observed
ef~ects of various mAb'~ on E-selectin-dependent H~60 cell
: adhesion. The adhesion was inhibited strongly by NDV sialidase
in ~mbinatlon with anti-LRX mAb S~1, or by anti-SLeX mAb SNH4 in
combination with SHl.
- 48 -
- - WO 93tl 7033 2 1 ~ 9 9 8 7 PCl /US93/0137-
The relevance of compounds of structure III was deduced
through the use of various mAb's and sialidases on
E-selectin-dependent adhesion of tumor cells which express type 2
chain sugars but al50 with E-selectin-trans~ected NS-l cell
S adhesion to glycoliposomes. For example, NDV sialidase treatment
o~ HL-60 cells, which removes NeuAc~2~6Gal, completely abolished
reactivity of cells with anti-S~RX mAb although the cells
remained adherent to E-selectin plates and activated end~thelial
cells.
10:Adhesion was inhibited effectively with a co~bination of
:! ~mAb~s direct~d to LeX and SLeX-
:In contrast to type 1 chain structures whose internally
~ sialosylated structure is known (Nudelman et al., supra) type 2
:~ chain structures with internally sialic acid residues were
15~ :hitherto unknown. Data presented in the instant application
indicate the natural occurrence of such epitopes.
: The~structures bindable to EL~M-1 can be synthesized using
known techniques. Thus, for example, the carbohydrates can be
~synthesizèd~ chemically using known and commercially available
ZO~ reagents or can be synthesized using known and available enzymes
to effect~the~appropriate linkage.~ For example, known sialosyl
transferases and fucosyl transferases can be used to derivatiæe
the b~sic carbohydrat~ backbone.:
Alternatively, ;the:carbohydrates bindable to ELAM-l can be
~ isolated using EL~-1 as an absorbent. For example, purified
: ELAM~ cells expressing ELAM-l:or membrane preparations o~ cells
~; :expressing EL~M~l can~be used.:~The ELAM-l can be immobilized to
:~ a solid phase~ such~as~ an~inert bead matrix or the inside wall
49 _
:: :
~29981
WO 93/17033 PCr/l)S93/01 ~ ~
of a vessel, to enhance separation. Then suitable carbohydrates
bindable to ELAM-1, such as extracts of HL60 or Colo201 cells
obtained by known techniques, are exposed to the ELAM-1 affinity
matrix. Following a washing procedure to remove unwanted and
non-specifically bound components, the ELAM-1 together with
carbohydrates bindable thereto are collected. The carbohydrates
bound to the EL~M-1 are separated ~rom the EL~M-l, fsr example,
by altering the salt concentration of ~he holding buffer, and
collected. The various carbohydrate species can be discriminated
using known procedures, such as chromatography.
Also, cells known to express predomi~antly type 1 chain
structures or type 2 chain structures are grown and membrane
preparations are obtained therefrom using known technigues. The
glycolipid -and glycoprotei2~ fraction o~ the me~brane prep is
obtained using ~nown techniquec and exposed to an affi~ity column
wherein antibodies directed to carbohydrate epitopes, such as
those describ~d herein, are affixed to a matrix, such as agarose
beads, to form an affinity matrix. In an affinity chromatography
procedure, th~ bound ~aterials are eluted and separated~~hrther
~ by known tec~ni~ues, such as HLPC and TLC.
When using TLC, ~he separated molecules in the separation
medium can be exposed to ELAM-l expressing cells that are
.
labeIled to serve as a tag, for example, the cells can be
labelled metabolically with a radioisotope. The
ELAM-l-expressing cells will bind to the respective sites of the
separation medium wher~ separated ELAM-l epitopes are found. The
TLC ~atrix can be autoradiographed to locate such sites of cell
binding ~o identify ELAM-l epitope-bearing molecules. The
- 5 t:) -
-~ WO93/17033 2 1 ~ 9 9 ~ 7 PCT/US93/0137~
respective sites of the TLC matrix can be excised and the
molecules extracted.
As noted hereinabove, the carbohydrates of formulae I and
II can be derivatized to provide oligosaacharides with more
desirable therapeutic properties. Thus, portions of the
structures comprising formula I or II can be substituted, for
exa~ple, with sulfur-containing sugars or fluorine containin~
sugars. The oligosaccharide derivatives can be prepared using
~ the methods disclosed hereinabove but substitutîng for the
naturally occurring components the appropriate reagent comprising
an altered substituent~ such as 6-trifluoro-fucosyl which is
incorporated into either of formula I or II as the fucosyl
residues.
The carbohydrates bindable to ELAM-l can be used as
immunogens to obtain ~ntibodies bindable to the carbohydrates
bi:ndable to ELAM~ ither polyclonal or mono~lonal antibodies
can be generated, using methods such as those described
hereinabo~e, and in the references cited herein, which are
incorporated by ~eference. ~onoclonal antibodies are pre~erred.
~: :20 : Because ~ ~ -1 may ~erve to mediate inter~ellular
interactions, interruption of binding between ELAM-l and
carbohydrates bindable th~reto will be beneficial. Thus,
ca~bohydrates bindable to ELAM-l, ELAM-l, antibody to EL~M~l or
antibody to carbohydrates bindable to EL~M-l, for example, can
b~:us~d to interrupt binding between ELA~ol and carbohydrates
bindable~ther~to. The carbohydrates bindable to EL~M~l~ ELA~
antibody to EL~M-l or antibody to carbohydrates bindable to
EL~M-l are administered in therapeutically e~fective amounts and
- 51 -
W0~3/17~33 ~99 PCT/US93/01--~
via routes that are determinable readily and routinely practicing
settled methods of the pharmaceutic arts.
As noted in formulae (I).and (II), the terminal sialic acid
i~ not essential in a carbohydrate bindable to ELAM-l. Key
element~ held in common are the terminal galactose, glucosamine,
~2-6sialic acid and fucose residues. Thus, antibodies capable
of binding to such a structure are effective in inhibiting
ELAM-l-mediated interactions. Suitable antibodies are CA3FA and
FH7.
Compounds of formula (III), for example, LeX/SLex, wherein
relevant epitopes comprised the branched chain structure were
identified clearly as comprising high affinity binding sites for
ELAM-l under high hear stress conditions. However, ~uch
structures can show~less binding ability than simple SLeX to
EL~M-1 at low shear stress conditions or under static conditionsO
Using that hybrid it is noted that the terminal galactose ~1-3
lin~ed fucose to Glc~Ac at one branch and an ~2~3 linked ~ialic
acid and ~1~3 linked fucose at the other branch are critical
: sites on that hybrid structure. Hence, antibo~ies bin~able to
~0 ~aX~ such as, SH-1 and FH-Z, and to SLeX, such as F~-6, SN~-4 and
SNH-3, are e~fective~aooperatively in inhibiting EL~M-1-mediated
adh~ion at high sh2ar stress conditions.
: ~a~y epitopes recognized by ELAM and GMP-140 are carried by
O-linked sugar chains and selectin-depe~dent cell adhesion ¢an
be blocked by inhibitors of 0-glycosylation (Kojima et al.,
Biochem. Biophys. ResO Commun. 182:1288, 1992-~. Hence, it often
i5 pr~ferable:to have compounds of formulae (I~, (II) and (~II)
carried on 0-linked carbohydrate chains. -
- 5~ -
WO 93/17033 212 9 9 8 ~ Pcr/us93/ol37~
A further means of interrupting ELAM-l mediated interactions
is using a combination of carbohydrates or antibodies ~o
interfere with El~-l binding to relevant carbohydrates. The
carbohydrates or antibodies are related to EL~M-l or
5 carbohydrates bindable thereto or in certain circumstances may
be aarbohydrates or antibodies that are not specifically those
carbohydrates believed to bind E~M-l. Far example, a
combination of antibodies directed to SLeX and LeX is effective
in inhibiting ELAM-l interaction. Suitable SLeX antibodies are
10~ SNH3 and SNH4; and suitable LeX antibodies are SHl and FH2. The
skilled artisan can determine other suitable combinations
practicing the methods taught herein using reagents disclosed
herein, with particular attention drawn to the working examples
set forth hereinbelow.
: ~ 15 ~ The foIlowing examples are offered by way of illustration
and not by way of limitation.
EXAMPLES
: :
: : :
Example l
SYNTHESIS OF IACTOSE DERIVATIVES
:
~ ~:20 A. Methyl ,6-D-lactoside
:: ; j
:~ : XeptaacetyllactosylilQidate ~Zimmermann et al., J. Carbohydr.
Chem. 7:435, ~ 1988):~ was ~ reacted with methanol in dry ~:
dichloromethane ~containing trimethylsilyl
; trifluoromethanesulfonate according to astandard procedure
53
:: : :
.
WO~3/17033 PCT/US93/0'~
( ~ ndler & Schmit, Liebigs. Ann. Chem. 1984:1826, 1984).
Purification by silica gel column chromatography (toluene/EtOAc,
1:1 by vol.), followed by de-O-acetylation with 0.01 M sodium
methoxide, gaYe methyl ~-D-lactoside in 68~ yield from the
imidate: m.p. 211-212C (lit. ~05-C, Smith ~ van Cleve, J. Am.
Chem. Soc. 77:3159, ~955); ~D + 1.3 (C 6.9, H20) (lit. + 1,c
5.0, H20), ibid.
B. Phenyl ~-D-thiolactoside
Lactose octaacetate (Hudson & Kunz, J. Am. Chem Soc.
47:2052, 1926~ was treated with thiophenol and SnCl4
(Nicolaou et al., J~ Am. Chem Soc~ 110:7910, 1988) in
dichlorometha~e at 0-C to gi~e phenyl heptaacetyl
~-D-thiolactoside in 80~ yield. The prsduct was deacetylated
with Na~Me in MeOH a~d neutralized with Amberlys~ 15.
Puri~ication of the product on a BioGel P-2 column using water
as an eluent, followed by lyophilization of the sugar containing
fraction, left phenyl ~-D-thiolactoside as a white a~rphou~
: powder.
~. ~acto-N-t~trose
.~
The oligosaccharide (Gal~1~3GlcN~c~1~3Gal~1~4Glc) was
prepared from human milk by pretreatment with ethanol and
recycling BioGel P-2 column chromatography wi~h water as eluent
followed by reversed-phase (C18) high pressure liquid
chromatography with water (Dua & Bush~ Anal. Biochem. 133:1,
- 54 -
. W093/17033 212 9 9 ~ 7 PCT/US93tO137~
1983)0 The 1~_NMR spectrum superimposed that of the authentic
sample (BioCarb Chemicals, Lund, Sweden).
D~ The polyethyleneglycol derivati~e of ~-D-lactoside
The reaction scheme is as set forth below:
~lc o~ A~.~o~(C'' )3 N~
~cO
(CY.~ C ~ C!, C !, 3
n- Y~i,
2 ~ :;
:
C)Ac ~oAc
\~G~o--(C~ U2~.P70,~,~C
_
`~
C112)3_CO~ --(C~-~r~2
-- }10
q
J_
: - 55 -
W093/li033 99~ CT/US93/01~
The polyethyleneglycol derivative of ~-D-lactoside was
prepared from readily available 3-succinimidooxycarbonylpropyl
0 (2, 3, 4, 6-tetra-O-acetyl-O-~-D-galactopyranoSyl)~ 4)-
2,3,6-tri-O-acetyl-~-D-g~copyranoside 1 and polyethyleneglycol
5methyl ether ~average ~.W. 2000; Aldrich Chemical, Milwaukee, WI)
h~ving a terminal amino group 2 (Zalipsky et al., Eur~ Polym. J.
19:1177, 1983~. Trea~ment of 1 (100 mg, 0.12 mmol) and
2 (163 mg, 0.082 mmol) in dry N,N~dimethylformamide (2 ml) at
room temperature for 2 hours gave, after chromatography on LR-20
10with acetone as an eluent, the ~-D-lactoside heptaacetate 3 in
91% yield: ~]D-5.3 (C 0.5, chloroform). A subsequent
saponification of 3 with 0.05 M sodium hydroxide at room
temperature for one hour, followed by lyophilization, afforded
the desired lactoside 4 quantitatîvely: [~D-2.4-(c 1.0,
15chlorofor~).
Example 2
EFFECT OF LACTOSE AND L~CTOSE DERIVATIYES ON
MET~ST~TIC POTENTIAL OF B16 MEL~NOMA CELLS
The highly meta~tatic BL6 clone of the B16 melanoma cell
20line was obtained originally from Dr. Jean Starkey (Montana State
Univ., Bozeman, MT) and clones~were reselected in syngeneic CS7Bl
:~ : mice ~according to metastatic potential. C57Bl mice were
~; maintained in plastic; csges under filtered air atmosphere and
provided with water and food pellets ad lib. Cells were cultured
2Sin RP~I 1640 supplemented with 2 mM glutamine and 10~ fetal calf
- 56 -
'
,-.. WO93/17033 ~ 212 9 9 8 7 PCT/US93/0137~
serum (FCS), and detached with phosphate buffered saline (PBS)
containing 2 mM EDTA. Viability was inferred by a trypan blue
exclusion test.
A suspension of B~6 cells (1-3 x 10 cells/ml RPMI 1640
medium) was prepared and aliquots were incubated in the presence
or absence of various oligosaccharides a~ various concentrations,
at 37C for 5-10 minutes. Following incubation, typically,
3 x 104 or 2 x 104 cells (with or without oligosaccharide
pretreatment) per 200 ~1 were injected via a tail vein into
8-week-old female mice~ After 18-21 days, the mice were killed,
the lungs were fixed in 10~ formaldehyde in PBS (pH 7.4) and
tumor cell colonies were counted under a di~secting microscope,
thereby providing background values of metastatic melanoma colony
number in lung under those conditions. Data on the number and
the size of colonies were treated statistically by an analysis
of variance (ANOVA) proaedure. CO1QnieS with a diameter of 1 mm
or greater were considered larga-size and those with a dia~eter
~: less than 1 mm were considered small-sîze.
For one experiment, BL6 cells were incubated withr ~arious
concentrations o~ ~actose, l~cto-N-t~tro~e
(Gal~1~3GlcNAc~1~3Gal~1~4GL~), methyl ~ actoside or phenyl
: ~-D-thiolactoside for various durations~ In the majority of
: experiments, ~ concentration of 0.1 M was used and cells were
incubated at 37C for 10 minutes, separated from the
~ sugar-containing medium:by mild centrifugation at 400 x g for
~10 minutes, resuspended in~RP~I 1640 and injected (3 x 104 cells
in 0.2 ml suspension) via a tail vein. For some experiments,
: 2 x 104 cells were injected and coloniec were counted at 2~ days.
- 57 -
WO93/17~ PCT/US93/01~'S
Viability and cell growth ability of BL6 cells after incubation
in various sugar solutions were tested by trypan blue exclusion
test, by plating in RPMI 164Q culture u~der norma~ conditions
in vitro as well as by subcutaneous inoculation in age-matched
C57Bl mice to test tumor growth.
Lactose and lacto-N-tetrose showed 26~ and 36% reductions,
respectively, of meta~tatic colonies in lung when BL6 cells were
preincubated with those sugars followed by intra~enous injection
of cells under identical conditions. Treatment vf BL6 cells with
O.l M, O.Ol M or O.OOS M methyl ~-D-lactoside. under the same
conditions as above resulted in (respectively) a 43~, 16% ~nd 8%
rsduction of metastatic lung colony number compared to control.
The significant reduction caused by 0.~ M methyl ~-D-lactosid~
was reprod~ced in three separate experiments and the reductîon
was found to be consistently between 35% and 45%.
, .
In a second, independent series of experiments, treatment
with methyl ~-e-lactoside or phenyl ~-D-thiolactoside under
dif~erent conditions also produced a signi~icant reduction of
metastatic colonization~ i.e., tot~l colony number was-~educed
to 35:% or 50% o~ control values following preincubation with
methyl ~ lactoside or:phenyl ~-D-thiolactoside, respectively.
Reduction of larger-sizc colonies was more apparent than that of
smaller colonies in all experi~ents, particularly those with
phenyl ~-D-thiolactoside (Figure l). Methyl ~ lactoside and
phenyl ~-D-thiolactoside both showed a slight i~ vitro
stimulatory effec~ on cell number increase and on thymidine
i~corporation. Thus, the inhibito~y effect on tumor deposition
is not related to the efPect on cell growth in vitro or in vi~o.
- 58 -
WO93t17033 212 9 9 8 7 PCT/US93/0137~
In a separate experiment, the effect of methyl ~-D-lactoside
on melanoma cell me~astasis was determined after administration
of the oligosaccharide, followed by inoculation with tumor cells.
Specifically, a one ml dose of methyl ~ D-lactoside (at a
concentration of 0.25 M or 0.5 M~ was injected intraperitone~lly
in mice. After 10 minutes, B16 melanoma cells were injected
intravenously. Lung colonies were counted 19 days latPr.
Injection of methyl ~-D-lactoside in advance o~ inoculation with
tumor cells resulted in a significant reduction of lung
metastatic colony formation (Figure 2).
In a separate experiment, mouse melanoma B16 ~ariants
showing different degrees of metastatic potential
(BL6/F10/Fl/Wa4) showed the same order of expre~sion o~ GM3
ganglioside, which was previously identified as a
~elanoma-associated antigen (Hirabayashi et al., J. Biol. Chem.
: 260:13328, 1985; Nores et al., ~. Immunol. 139:3171, 1987)~ GM3
interacts with LacCer, which is highly expressed on endothelial
~ cells. The order of adhesion of the B16 variants onto
; LacC~r-coated solid phase or onto endothelial cells wa~`~lso in
th~ ame order as me~astatic potential ~MP~. In contrast,
integrin-dependent adhesion of the ~16 variants was approximately
~qual for BL6, F10 and Fl (see Figure 4~. Those observations
suggest that B16 adhesion o~ LzcCer is based on molecular
: G~3-LzcCer interaction. It also has been demonstrated that B16
~ 25 melanoma adhesion on endothelial cells is inhibited not only by
: methyl-~-lacto~ide but also by LacCer liposome~ Gg3Cer liposome,
and G~3 liposome (see Figure 53.
- 59 -
J
WO93/17033 PCT/US93/01~
sa~
In addition, the observations on the metastasis-inhibitory
effect of methyl-~-lactoside noted above have been extended to
separate methyl-~-lactoside injection, i.e., tumor cells were
injected intravenously, ollowed by ~ntraperitoneal injection of
S methyl-~-lactoside. In those experiments, injection of
0.25-0.5 M methyl-~-lactoside reduced lung metastatic colony
number by 40%-70% ~see Figure 6; A = PBS control, B = 0.25 M
Me-~-lactoside; C = 0~5 M Me-~-lactoside; D - 0.5 M lactose;
E = 0.25 M N-acetyllactoæamine; F = 0.5 M Me-~-galactoside;
intraperitoneal injection).
Capillary endothelial cells are strongly reactive with
antibodies directed to H/LeY/Leb, such as antibody MIA-15-5.
;~ That observation comports with the earlier observa~ions that Ulex
~Europ. I stains endothelial cells, Holthoer et al., Lab. Invest.
45:391, 1~81; 47:60, 1982.
Liposomes comprising H-l or LeY were made and exposed to
plates to which ~arious glycolipids had been affixed at a range
: ~ ,
of concentrations.
As noted in Figure 7, N-bearing liposomes bound to-~ or LeY
coated onto plates. On the other hand LeY-bearing liposomes were
found to bind only~to~H-aoated plates. H and paragloboside are
` related,~ the only dif~ference being the presence of a texminal
fucos~e residue in H. ~
Hence, cells expressing H, LeY or Leb can adhere to
25 ~ endothel`ial cells expressing H and possibly to LeY as well.
Those types of interactions~may be the first step in tumor cell
~ ,
to endothelial cell~àdhesion.
60 -
~, W~93/17033 21239~ 7 PCT/US93/0137~
Example 3
EXPRESSION OF S~A~OSYL-DIMERIC LEX
ON HUMAN L~NG ADENOC~RCINOMA CELL LINES AND
METASTATIC POTENTIAL
S KUM-LK-2 is a human non-adenocarcinoma cell line
characterized by producing spontaneous lung metastasis in nude
mice. After screening 35 human carcinoma cell lines grown in
nude mice, only that cell line produced metastatic deposits in
nude mouse lung~ KUM-LK-2 was u~ed as the parent cell line to
obtain, by lîmi~ing dilution technique, sub-cell lines producing
lung metastasis on IV injection.
The procedure for the limitin~ dilution technique was as
followsc KU~ 2 was cultured in RPMI 1640 medium (GIBC0, Grand
Island ~, NY) supplemented with 1~% FCS ~Hyclone, Logan, UT~ at
37 C in a 5% C02/95% a7 r atmosphere. Cells were trea~ed briefly
with 2 3nM EDT~ solution and washed twice with RPMI 1640 to make
a single cell suspensivn in RP~I with 10% FCS. Cell vi~bility
~as ~ 9~% as d~termined by trypan blue exclu~ion staining. A
c@ll SU pension ~ontaining 1 cell per 100 ~l wa~ transferred to
20 each well of a 96-well mic:rotiter pla~e ~Corning Glass WorksO
Corning, NY) and cultured continuously ~or 24 hours. Each well
then was examined by phase contrast microscopy.
Th~ee c:ell lines (HAL-8, HALo24 and ~IAL-333 with different
metastatic poten~ial ("MP") were selected out of 2S clones
25 obtained by limiting dilution technigue on the basis of stable
cell morphology. The 25 clones were select d originally from 63
-- 61 --
W093/17033 PCT/US93/01~'~
~2g9~ .
clon showing stable morphology as well as consistent in vitro
cell growth.
All of the clones produced spon~aneous lung metastasis.
However, on I.V. injection, clear differences were observed among
the clones in terms of lung metastatic deposit formation. Two
clon~s with high MP, five with low MP and 18 with no MP were
distinguished.
Through repeated selection by I.V. inJection of the clones,
the most stable sub-cell lines showing consistent MP were
established. Those were HAL-8, H~L-33 and HAI,-24, showi~g high,
: low and no MP, resp~ctively, to nu/nu mouse lung (see Table 1
below). Jud~ing by macroscopic and microscopic examination, none
of the three sub-cell lines showed me~astasis in other organs or
lymph nodes. The sub-cell lines represent stable variants
original}y present in K~M-LK-2. Based on chromosome analysis,
the sub~lones are independent.
~: :
- 62 -
- :
,~., W093/17033 2 1 2 9 9 8 7 PCT/US93/~137
Table 1
Metastatic potential of HAL-8, HAL-24 and HAL-33 in nude mice. A
Clone # generations #lung nodu~es on
S day 56
_
HAL-8 15 15~8 (8-23)
22 15.0 (10-22)
46 1~.3 (11-25)
;10 HA~-24 15 0
22
46 0
HAL-33 15 4.3 ~3-7)
22 5.1 (2-8)
15: 46 5~8 ~3-8)
llude rnice ~ere inj~cted S2 x 105 cells) vi~ th~ t~il vein ~t v~rious generation ti0es as indicated.
~ifty-six ~ f~er injection, mic~ r ere killed ~nd rr2tl~st~tk nodules on l~g ~urf~ce l~ere counted ulder
dissecting microscope.
bbe~n o~ 6 ~ ls ~rsr~e in p~renthe~es)
.
: ~:
:: :
The celI ~urface~expression o~ various carbohydrate ~itopes
was ~nalyzed ~by cyto~luorometry using various monoclonal
antibQdies: (mAb'~) direct~d to: Le~ (mAb SHl), sialosyl-LeiX
~:: 25~ mAb~SNH4), sialosyl-dimeric LeX (mAb FH6), T (mAb HH8),
Tn (mAb lE3) and sialosyl-Tn ~mAb TKH2). Al~ a~tibodies used
were culture supernatants ~rom the respectiYe hybridomas,
,:
~adjusted as 10 ~g/ml of immunoglobulin. The structures of
sialosyl LeX (structure~l~, sialosyl-dimerio-~eX (struictursi 2),
~; 30 dimeric~LRX (structure 3), trifucosyl-LeY ~strUcture 4),
: Le ~structure 5:), H (:structure 6), SA-Le I (structure 7~,
.
~: :
63 - . .
':
21~99~3~
WO93/17033 PCr/US93/01 ~Ç.
SA-Tn (structure 8), disialosyl-Le (structure 9),
monosialosyl-Lea II (structure 10), GM3 (structure 11),
S-PG (structure 12), ~ex (structure 13) and Le~ (structure 14)
are shown below. R represents a carrier molecule.
Structure 1:
NeuAc~2~3Gal~1~4GlcN~c~1~3~Gal~1~4GlcNAc~1~3]nGal~1~4GLc~l~R
~ (n>0)
Fuc~1
Structure 2:
NeuAc~2~3Gal~1-4GlcNAc~1-3Gal~1~4GlcNAc~1~3Gal~ GLc~l~R
t t
Fuc~1 Fuc~1
St~ucture 3:
Gal~1~4GlcNAc~1-3Gal~1~4GlcNAc~1~3Gal~1~4Glc~l~R
3 3
t t
;Fuc~l Fuca1
~0 Structure 4: .
:: Gal~1~4GlcNAc~3Gal~1~4GlcNAc~1~3Gal~1~4Glc~ltR
2 ~ 3
t t ~ (The al~2 fucose can be
~uc~l Fuc~1 Fuc~l replaced by ~)
Structure 5:
Fuc~1~2Gal~1~3GlcNAc~1-3Gal~l~R
t
Fuc~l
- 64 -
.~, WO93/17033 212 9 9 8 7 PCT/US93/0137~
Structure 6:
Fucal~2Gal~1~3GlcNAc~1~3Gal~l-R
Structure 7:
NeuAca2~3Gal~1~3GlcNAc~1~3Gal~l~R
Fuc~
:
Structure 8:
N~uAc~2~6GalNAcal~0-Ser/Thr
Stxucture 9:
NeuAc~2
: ,~
NeuAc~2 3Gal~1~3GlcNAc~1~3Gal~l~R
t
Fuc~l
Struct~ure lO.
20 ~NeuAc~:2 :
:
Gal,Bl ~3GlcNAc,~1~3Gal~ R~
:
t
Fuo~1 :
Structure 11 : ~
euAca2~al~l~4Glc~l~Cer
Structure 12: ~
NeuAc~2~3~al~l~4Glc~Ac~1~3Gal~l-R :
: .
~ 65 -
~ ,
:
WO93/17033 P~T/US93/017
~ ~t~
Str ure 13:
Gal~1~4Glc~Ac~1~3Gal~l~R
Fuc~1
Structure 14:
Gal~1~3GlcNAc~1~3Gal~l~R
t
FUc~l
: Cells were detached from culture flasks with 0.25% trypsin,
2mM EDTA solu~ion and 1 x 10 cells were prepared ~or each mAb
treatment. Cells were incubated with a mAb for 1 hour at 4-C and
washed 2 times with RPMI 1640. Goat anti-mouse IgG or IgM-FITC
(Boehringer-Mannh~im, Indianapolis, }N), diluted 50 times with
: PBS, then was added and incubated 30 minutes at 4-C. Finally,
cells were washed 3 times, resuspended with PBS and analyzed in
: an EPICS PR~FILE flow cytometer (Epics, Hialeah, FL~. The
experi~ents were repeated with three different cell gene~tions~
: 20 Patterns of expression of six carbohydrate epitopes (defined
by the respective mAb's) on sub-cell lines HAL-8, H~L-24 and
: HAL-33 showed nearly identical pro~iles (as did the protein
profiles for the thrce sub-cell lines) except in the case of
~ialosyl-dimeric-LeX. In particular, H~L-8, HAL-24 and HAL 33
were found to e~press~ highly and equally sialosyl-LRX and
sialosyl-Tn structures. Each of ~he three l~nes exp~essed low ~
~uantities of LeX and Tn, and did not express T. In contrast,
:
.
~ - ~6 ~
.
~ WO g3/17033 2 1 2 9 3 8 7 PCT/US93/~137~
expression of sialosyl-dimeric LeXwas high on HAL 8, moderate on
HAL-33 and low on HAL-24.
The release o~ sialosy~ residues was assessed in the
following manner. Cells were detached using 2 mM EDTA in PBS,
washed and resuspended in 9 volumes of PBS. One ml of cell
suspension was incubated 5 minutes at 37'C with 0.2 U/ml of
Clostridium perfringens sialidase (type X, Sigma Chemical Co.,
St. Louis, MO). After incubation, cells were washed three times,
resuspended with RPMI l640 and investigated for MP and expression
of sialosyl-dimeric-L2X. MP o~ ~AL 8 and HAL-33 was inhibited
completely by sialidase treatment of cells ~see Table 2 ~elow).
Expression of sialosyl-dimeric-LeX appears to play an important
role in blood-borne metastasis.
Table 2
: 15 : Effect of sialidase treatment on me~astatic potential
~ of clones HAL-8 and -33 . a
: Trea~ent Clone # lung nodules on day 56
Control (PBS) ~AL-8 15.3 (9-24)
HaL-33 4.6 (3-7)
Sialidase : NAL-8 0
HAL-33 0
:
_
llude ~llice ~ere injected 52 x 105 celts) vi~ the tail vein. Fifty-six d~ys ~Ifter injection, ~ice
~ere killed ~ t~static nodules ~ l~g surface ~ere coulted u~er dissectin~ ~icroscope.
b~ean of 6 ~niR~ls tr~n~e in par~nthescs~.
: .
:; : :
~: :
W~ ~3/17033PCr/US93/Oi~-5
~99~'~
EXAMPLE 4
IDENTIFICATION OF CARBOHYDRATE EPITOPES CAPABLE
OF BINDING TO THE LECTIN DOMAIN OF GPIP-14 0
Platelets were isolated from "platelet-rich plasma" obtained
from the Oregon Red Cross (Portland, OR). Contaminating red
blood cells were remo~ed by centrifugation at 80 x g for 10 min.
Platelets were centrifuged at 300 x g for 10 min and suspended
in Tyrode1s buffer (pH 6.5) containing 22 mM citrate buffer with
0~35% bovine serum albumin (BSA). The platelet suspension
(1 x 108~ml) was incubated (p~I 7.21 37~C, 5 min) after addition
of thrombin ~f~nal concentration 1 U/ml). The mixture then was
incubat~d at 37C for 10 min without stirring. The
thrombin-artiYated plate~ets were fixed with an equal ~olume of
2% formaldehyde in phosphate-buffered saline (PBS) r pH 702, and
washed 2 x with PBS containing 1% BSA. Activated platelets (but
not non-activated platelets) showed strong reactivity with
2.5 ~g/ml anti GMP-140 mAb AC1.2 (iso~yp~ IgGt;
Beckton-Dickinson, San Jose, CA) when incubat~d at 37 C for
30 ~in., followed by reaction with 50 ~1 of fluorescenc~-labeled
goat anti-mou~e Ig (Tago, Burlingame, C~3. Flow cytometric
profiles of activated vs. non-activated platelets with mab ACl.
are shown in ~igures 8A-8D.
Activat2d and non-activated platelets were flxed with
parafor~aldehyde in Ca2-free PBS, pH 7.2, washed 2 x with
Ca2 -c~ntaining PBS with 1~ BSA, resuspended in CA2-PBS with
1% BSA and 0.1% azide and the number of platelets adjusted to
- 68 ~
~WO93/17033 2 1 2 9 9 8 7 PCT/VS93/~137-
~lxl0 /ml. The cell suspension was stored at 4C and the binding
assay performed within 24 hrO
Fluorescent polystyrene latex beads were obtained from
Molecular Probe, Inc., Eugene, OR~ The beads were yellow-green
5fluorescent beads with a sulfate group at the surface, diameter
0.5 ~m (actually 0.4B6 ~m). Beads (l x l0 3 in 30 ~l ETOH were
added to lO ~g of GSL solution in 200 ~l C=M, mixèd well and
dried under an N~ ~tream. The residue was resuspended in 200 ~l
ethanol, sonicated briefly and dried under a N2 stream. The
10dried residue was suspended in 2 ml CaZ-PBS with 3% BSA and 0.1%
: azide, sonicated for l0 min and allowed to stand at 37C for
60 mi~ to block the bead surface with BSA. The suspension was
centrifuged at 3000 x g for l0 min, the bead pellet was washed
: 2 x with Ca2 PBS containing 1% BSA and azide and finally
l5suspendsd in 500 ~l of the same medium and stored at 4-C.
Twenty ~l of platelet (non-activated or activa~ed)
uspension~ paraformaldehyde-fixed and containing ~2 x 107
platelets, was mixed with l0 ~l of fluorescent GSL-coated beads,
cont~ining ~2 x 107 beads, mixed well and allowed to s~and at
2037~C for 30 min. The platelet suspension was mixed wi$h 200 ~l
Ca2 -PBS and~ aDaly ed~by; flow cytomet~y (EPICS Profile~ Coulter
ytometry, Hialeah, FL).
.
Flow cytometric analyses of platelets alone and beads alone
were performed for setting a gating to include most of the
25signals produced by platelets and excluding signals produced by
free beads. The binding index (BI) was calculated as mean
fluorescence intensity (MFI) of platelets incubated with
fluorescent GSL-coated beads divided by ~FI of platelets
- 69 -
'
W0~3/~ Pcl`/us93/o~
~incubat~d with fluorescent non-GSL-coated (control) beads. BI
values for various GS~'s are shown in Table 3 and in Fiyure 9.
In Figure 9, the hatched bars represent non-activated platelets
and the open bars represent activated platelets. The ratio of
the binding index (BI) ~f activated/non-activated platelets for
SA-LR , SA-Le , SPG, GM3 and LeX also is shown in the "Ratio A/NA"
column.
Table 3
Bindi~g index of thrombin-activated platelets to
GSL-coated, sulfate-containing polystyrene beads
__
:GSL Activated platelets Ratio (Activated/
Non-activated)
,
GM3 1.0 ~ 0.1 0.7 + 0.5
SA-LeX 4.2 + 1.0 4.2 ~ 0.5
SA-LR 8.1 + 1.0 6.1 + 0.2
SPG 0:.8 + 0.2 0.7 ~ 0.2
LeX 1.1 ~ 0.3 ~.0 + 0.3
:~0
D
V~lue8 represent a~ar~ of four ~eparate exp~riments. ,;~
. :
:
':
mAb's affected platelet binding to fluorescen~ GSL-coated
~: ~ beads. Platelets were incubated with anti-GMP-140 mAb IOP62
(Immunotech, Marseille, France~ at 37-C for 30 min and a binding
as ay was performed using GSL-coated beads, as dsscribed
hereinabove. Non pecific mouse IgG (10 ~g/ml3 was used in a
control binding assay.
: - 70
~! WO 93/17033 2 1 2 9 9 ~ 7 PCT/US93/01375
Also, lo ~1 of SA-Le~-coated beads (2 x 107) were incubated
with 20 ~1 of anti-SA-Le~ mAb CA19-9 (20 ~g/ml) (mouse IgG1;
Signet Laboratories, Dedham, MA) at room temperature for 60 min
and used Por the plat~let binding assay. Anti-SA-LeX mAb SNH4
S and non-specific mouse IgG were used as controls.
The results are shown in Figure 10, where the abscissa
represents percent inhibition and column 1 represents
anti-GMP-14 0 mAb IOP62, column 2 represents anti-SA-Le mAb
CA19-9 (alternative mAb's are NKHl and NKH2), column 3 represents
~nti-5A-LeX mAb SNH4 and colu~n 4 represents normal mouse IgG.
Activated platelets showed high expression of GMP-140 as
evidenced by high reactivity with anti-CD62 mAb (Figures 8A-8D).
Activated platelets expressing ~MP-140 showed strong binding with
fluorescent beads coated with SA-LeX (Figure 9).
Binding of platelets to beads coated with SA-LeX was
observed but to a mu~h lower degree than with SA-Le~ (Figure 9).
No binding was observed to bead~ coated with other GSL's.
Furth r, the binding of platelets to SA-Leæ coated beads was
inhibited by anti-~P 140 mA~ and anti-SA-Le~ mAb, ~u~ ~not by
20 anti-S~ eXnL~b (FiguFe 10)-
'
- 71 -
W~93tl7033 ~ P~r/usg3/
EXAMPLE S
EFFECT OF YARIOUS MONOCLONAL ANTIBODIES ON ADHESION
OF HUM~ CO~ON CARCINOMA COL0205 CELI.S TO
INTERLEUKIN-l-ACTIVATED HUM~ UMBILICAL
S VEI~ ENDOT~E:LIAL CELI,S IN A DYNAMIC Fl.OW SYSTEM
Adhesion w~s measured usin~ the dynamic flow experimental
system shown in Figures l~A to ~lD. The number o~ cells bound
during 3 minutes at different shear stresses, for example, from
O . 4 to 4 .. 8 dynes/cm3, were counted f rom several f ields recorded
on videotape. The coefficient of viscosity was 1. 0 P, the hal:E
channel height was 5.7 X 10 ~ cm and th~ channel width was
1.3 cm.
Using that system, various human tumors and monoclonal
antibodies directed to various tumor-associated carbohydrate
antigens were studied. The results of one ~tudy, a~hesion of
human colon carcinoma Colo20S cells to ~ctivated human
endoth~lial cells, is ~hown in Figure lZ, where the ab~cissa
rep.resents wall shear stress (dynes/cm~) and the ordinate
represents c~ll adhesion (x 102/field).
In:Figure 12, th~ symbols are as follows: open circles,
mixture o* ir.relevant mouse IgG plus XgM (control); solid
triangles, monoclonal antibody CAl9 9 directed to
mono~ialosyl-LQa I; open triangles, monoclonal antibody SNH4
direct~d tv ~ialosyl-LeX; solid circles, monoclonal antibody F~7
directed to mono~ialosyl-Le~ II and disialosyl-Le~; and ~olid
- 72 -
,r ~ WO 93J17033 212 9 9 8 7 PC~/lJS93101375
sc~uares, mixture of irrelevant mouse IgG plus IgM and
non-activated endothelial cells.
The results show that adhesion of Colo205 cells to activated
endothelial cells was inhibited most strongly by antibody FH7,
particularly at high wall shear stress ~5-10 dyne~/cm2). In
contrast, antibody CA19-9 had no inhibitory e~ect. The findings
suggest that tumor cell adhesion to endothelial cells may proceed
via înteraction between monosialosyl-Le~ II or disialosyl-Le and
interleukin-l-activated selectin.
EXAMPLE 6
SEI.ECTIN-DEPENDENT ADH13SION OF HL60 CELLS
C's (Cell System~, Rirkland, WA) were ultured to
con~lu~ncy in 48-weil plates (Costar, Cambridge, MA) and
stim~lated with 1 U/ml IL-l for 4 hr. Non-si~ulated HUVEC's wer~
15~ used as a control. Expression of E-selectin (ELAM-l) on
IL-l stimulated HUVECi~ was confirmed by reac~iv~y with
anti-E-6electin m~b 387 ~IgG28) ~Graber et al. J. Imm. 145:819,
1990). HL60 and Colo201 cells were labeled metabolically by
culture in the pre~ence of ~3H3;thymidine after pretreatment with
glycosylation modifier and added to HUVEC-coated plates. After
15 min incubation~ plates were washed with PBS and adherent cell
number estimated by~conversion from radioacti~ity count. In
another set of experiments, 96-well plates (Falcon, Lincoln, ~J)
: were coated wi~h 0.1-1 ~g/ml o~ ~ truncated, recom~inant
E-selectin lacking transme~brane and cytoplasmic dom~ins (Shimizu
- 73 -
W093/17033 PCT/~S93/0- ~
2 ~2 9 9 8 ~t al. Nature 349:799, 19~1) for 18 hr. Plates then were coated
with 1% BSA, washed with PBS and coated with
metabolically-labeled,glycosy~ation-modifiedcells,asdescribed
above. After 60 min incubation, plates were washed with PBS and
adherent cell number estimated by conversion from radioactivity
count~
Assays of cell a~hesion to activated or native platelets
coated and fixed on 48-well plates were performed as previously
d~scribed (Handa et al., Biochemistry 30:11682, l991). HL60
cells were pretreated with 2mM benzyl-a-GaeNAc for 72 hr and
labeled with ~3H]thymidine. After washing with PBS 9 lxlO~ cells
were added to each well and plates were incubated for 30 min at
room temp. ~fter wa hing to remove unbound cells, bound cell~
were detached with trypsin and count~d by liquid scintillation
co~nter. Platelet~ bound on plates were incubated with
anti-P-sele~tin m~b IOP-62 (1~ 6 dilution) (Immunotech,
Marseille, France) at room temp for 30 min, followed by addition
of HL60 cells, to evaluate dependence o~ adhesion on P selectin
exprassion. Non specigic mouse IgG was used as control~
dhesi~n assay in a dynamic flow system
A parallel-plate laminar flow chamber connected to an
infusion pump (~odel 935, Harvard Apparatus, Cambridge, MA) was
used ~o si~ulate the flow shear stresses present in physiological
microvascular e~vironments. ~he flow chamber consists of a glass
plate on which a parallel, transparent plastic surface is
attached with a Silastic rubber gasket; there is a 114 ~m gap
- 74 -
~- Wo93/17033 2 1 2 ~ 9 8 7 PCT/~S93/0137~
between the two surfaces and the gap is connected to an inlet and
outlet.
A laminar flow with defined r~te and wall shear stress is
achieved by manipulation of the infusion pump, which is connected
to the inlet of the flow chamber. EC's are grown as a monolayer,
or adhesion molecules are coated, on the glass plate, and a
laminar flow o~ a cell suspension is pas~ed through the chamber.
Cell movements are observed under inverted phase-contrast
microscope (Diaphot-TMD Nikon) and recorded by time-lapse
videocassette recorder. ~dhesion is observed as rolling followed
by stopping of cells. Number of cells bound during 3 min at
different shear stres~es ~rom, for example, 0.4 to 4.8 dynes/cm2
or 0.76 to 15.5 dynes/cm2 were counted from several fields
~ recorded on videotape~ Wall shear stress (T) was calculated by
: 15 the equation of Lawrence et al. (Blood 75:227, l990):
T = 3~Q/2ba2
where ~= coefficient of viscosity ~l.0 cP), Q= volume~ic flow
rate (cm3/sec), a= half channel height (for the experiments
: reported herein, 5.7~x 103 cm) and b= channel width (1.3 cm)~
:
,
HL60 adhesion to E-selectin-coated ~lates under static conditions
Promyelocyti~ leukemia cell line HL60 has been shown to
express only type ~ chain: and sialosylated/fucosylated
deriva~ives as probed by speci~ic mAb's (Symington et al.,
J. Immunol. 13~:2498, 19~5) and has been extensively used as a
~ -- 75 -
WO93/17033 PCT/US93/0~
9~ model of leukocyte adhesion mediated by E-selectin and P selectin
(Phillips et al. Science 250:1130, 1990; Polley et al. Proc.
Natl. Acad. Sci. USA 88:622.4, 1991; Handa et al~ Biochem.
Biophys~ Res. Commun. 181:1223, 1991; Kojima et al. Biochem.
Biophys~ Res. Commun. 182:1288, 1992).
When HL60 cells were treated with Newcastle Disease Virus
(NDV) or Vibrio cholerae (VC~ sialidase, reacti~ity of cells with
mAb's SNH3 and SNH4 was abolished (Figure 13).
:: E-æelectindependent HL60 adhesion was reduced by only about
; 10 20-50% after treatment with NDV sialidase, w~ereas adhesion of
the same cells treated with Vibrio or Arthrobacter ureafaciens
~AU~ sialidase was reduced to about 5-10% o~ control values or
essentially abolished. (NDV, VC and AU sialidase were effective
equally in eliminating SLeX expr0ssio~ on HL60 cells.~
NDV sialidase eliminates only the ~2~3 sialosyl residue
linked to the terminal Ga} whereas both Vibrio and Art~roacter
sialidasa completely eliminate terminal and internal si~lic acid
residues, notably, ~2~6 linked sialic acid residues. The
~: ~ findlngs indicate that SLeX ~nd SLe~ are nok the sole epi~pes of
E-selectin and P-selectin.
Effects of various mAb~s on ~-selectin-dependent HL60
adhesion were tested. Anti-S~X mAb's SNH3 and SNH4 produced
strong HL60 cell aggregation, even ~nder carefully-controlled
conditions. Therefore, the degree of inhibition of HL60 adhesion
:~ ~ 25 by SNH3 or SNH4 varied considerably since aggregated cells tend
: to detach from~E-selectin-coated plates.
~; In general, the degree of inhibition by those mAb's was
minimal compared to the degree of inhibition previously described
76
~:
~ W~3/17033 2 1 2 ~ 9 8 7 PCTtUS93/0137~
by Phillips et al. (supra). Anti-LRX mAb's SH1 (IgG3) (Singhal
et al. Cancer Res. 50:1375, lg~o) and FH2 (IgM) (Fukushi et Al.
J4 Biol. Chem. 25g:4681, 1984) produced a consistently higher
degree of inhibition than mAb's SNH3 and SNH4.
A mixture of SNH3 or SNH4 with SH1 or with FH2 produced
stronger inhibition than any of the mAb's alone. Strongest
inhibition was produced with a mixture of SNH4 and SH1
(Figure 14).
If SLeX is the sole epitope of HL~0 cells for E-selectin and
~0 P-selectin, anti-SLeX m~b's (e.g., SNH3 ~nd SNH4) should inhibit
completely selectin-dependent adhesion. Treatment with NDV
sialidase, which abolished reactiviky of HL60 cells with SNH3 and
SNH4, also should i~hibit E-selectin-dependent cell adhesion.
However, treatment of HL60 cell.; with NDV sialidase followed by
SNH3 or SNH4 did not further reduce adhesi~n. ~reatment with NDV
sialidase followed by anti-LeX mAb's SH1 or FH2 strongly
i~hibited E-selectin-dependent HL60 adhesion.
6~ adhesion 5O activated HU~EC's ln a static s~stem -~
The ~ame trends observed ~or HL60 adhesion to
E selectin-coated plates were observed for HL60 adhesion to
activated HUVEC~ grown~ in plates. Adhesion to HUVEC's was
' affected minim~lly by anti-SLeX ~ 's SNH2 or SNH4, in contrast
:: to the pre~ious report by Phillip~ et al. (supra). Different
preparations of the mAb's varied widely in the effect on
~L60-HWVEC adhe~ion, and:some mAb's caused strong aggregation of
HL60 cells. Anti-LeX mAbJ S SHl and FH2 show2d consistently
- 77 -
WO93t17033 PCT/US93/01~'5
~99 stronger (compared to SNH3 or 5NH4) inhibition of HL60-HUVEC
adhesion~ as did a combination of SNH4 plus SHl or FH2. NDV
sialidase did not re~uce significantly HL60-~UVEC adhesion, but
Vibrio sialidase almost abolished reactivity ~ompletely.
Coatinq of adhesion molecules or EC's on alass plates in the
dynamic flow system
For le~tins, fibronectin (FN), laminin (LN), truncated
:E-seleatin and GSL's used, 10-50 ~l of a solution having a
; concentration of 20-200~g/ml was placed on a marked area (0.5 cm
diameter) on a glass plate (38 x 75 mm; Corning Glassworks,
Corning, NY) and dried in a refrigerator ~t 4-C. ~ried plates
were immer~ed in PBS at 37-C for 1 hr and washed extensively with
several changes of PBS. For GS~ coating, GSL-liposomes were
prepared from 200 ~g GS~, 200 ~g cholesterol and 400 ~g
15~ phosphatidyl~holine in l ml;PBS. Ten ~l of GSL-liposome solution
,
was~ placèd on~a glass plate, dried at 4-C and the plates were
: washed with PBS, as descri~ed above. ,;~
The~quantity;of adsorbed molecules was detex~ined using 125I
labeling for lectins,~FN or;LN, or [3H]cholesterol labeling for
~ GSL-liposomes.~ ~Under those ~conditions, almo~t ~he entire
uantity of~protein, rega~rdless of whether FN, LN or lectin, was
: adsorbed~on the glass plate. For example, when lO0 ~g/ml FN was
::
applied~, 12.5~ f ~1.8 ng/mm2~was~adsorbed. Likewise, almost all
6~S-liposome dried on~ the glass plate was ad~orbed; e.g., when
5 ~ 20:0 ~g/ml GLS-liposome was applied, 31.3 ~ 5.2 ng GSL/mm~ was
adsorbed. : ~
~ ,
~ 78 -
WO93/17033 212 9 9 8 7 PCT/US93/0137~
EC's were coated by placing 100 ~1 of a suspension
containing 2 x 105 mous~ or human EC's on glass plates and
culturing in a C02 inc~bator at 37C until conflllency was
achie~ed.
Plates coated with adhesion molecules or EC's were affixed
in a flow chamber, and a suspen~ion of Bl6 melanoma cells was
passed through the chamber as described hereinabove B16 cells
were harvested fxom culture using 0.02% EDTA in PBS, and
suspended in PBS at a concentration of 1 x 105/ml.
HL60 adhesion to ac~iva~ed H WEC's~in a dynamic flow system
The effects of various mAb's and sialidases on HL69-HUVEC
adhesion were tested also in a dynamic flow system. Generally,
the effects of the mAb's were similar to those obs~rved with a
:: stati~ system. ~ SNH4 had no inhibitory e~fect at all undervarious shear s~resses. mA~'s SHl and FH2 showed moderate
inhibition. Again, stronges~ inhibition was obtaine~with a
~ combina~ion o~ SNH4 plus SHl or FH2. Adhesion was reduced
; moderately by NDV sialidase and almost comple~ely ~y Vi~riv or
~rthrobac~er siali~ase. See Figur~ 15.
The results set forth h~reinabove using ~L60 cells suggest
: ~ that the ~resence of sialic acid in the carbohydrat~ epitope is
;
important in providing;~binding specificity to E-selectin.
~wever, th~ sialic residue is not r~quired to be ~2~3 linked at
:
:~ ~ the terminal Gal; the~sialic acid residue alternatively could be
: 25 present at an internal location, e.g., linked to internal Gal or
- 79 -
W093/17033 PCT~VS93/0~
~9~ G1cNAc. Clearly, though, ~1-3 fucosylation at G1cNAc is
essential.
Under dynamic conditions, NDV sialidase had an inhibitory
effect only at low shear stress whereas VC or AU sialidase
significantly reduced adhesion even at high shear stress.
Anti-LeX IgG mAb SH1 strongly inhibited adhesion even at high-
: shear stress, whereas the effect of anti-SLeX IgG3 mAb SNH4 was
minimal. Strongest inhibition was produced by a combination of
ND~ sialidase plus anti-LeX mAb SH1. A mixture of anti-LeX plus
anti-SLeX mAb~s produced stronger inhibitory effect than either
mAb alone.
~- As depicted in Figure 19, at low shear stress
(<4 dynes/cm2), adhesion was inhibited significantly by NDV
. : ::
sialidase or by mAb SNH4, whereas those reaqents had no effect
at high shear stress (8-16 dynes/cm2). In contrast, VC sialidase
completely abolished adhesion at hlgh shear stress. mAb SH1
inhiblted adhesion more strongly at high than at 1OW shear
stress, but the dlfference was relatively small.
~ ~ : x
~ e ~alone~clearly is not sufficient as the E-ce1ectin
: 20: epitope. Rather, a2~3 plus a2~6 sialylated structures are
necess~ry.~ ~LeX-liposomes and LR'-llposomes blnd to ELAM-coated
plates,~ and binding by SLeX is stronger than by LeX or other
glycolipids.:How~ver,~those epit~pes must be present at the cell
~ -
;:;surface in the form df multlply 0-glycosylated mucin-type
25: ~ glycoproteins.~ LeX as;well::as~SLeX may be ~2~6 sialylated at the
internal~ Ga~ or GlcNAc;~withln the same CH0 chain, or ~2-~6
sialylation~ may ~be present at an adjacent branched structure.
"6-C ganglioside,"~which~ls an ~2-6 sialylated type 2 chain
80 -
:: :
~ W093/17033 2 1 ~ ~ 9 8 7 PCT/US93~0137~
structure with internal ~1~3 fucosylation (Hakomori ~t al.,
Biochem. Biophys. Res. Commun. 113:791, 1983), failed to bind to
E selectin. Thus, such a struçture can be excluded as a possible
ELAM epitope.
EXAMPLE 7
C0L0201 CELL ADHESION: EFFECTS OF VARIOUS
SIALIDASES AND mAb's
Colo201_ adhesion to. ~-selectinocoated _plates under static
conditions
In contrast to HL60 cells ~which express predominantly
type 2 chain structure), Colo2Q1 cells express mainly type 1
: chain, and E-selectin-dependent Colo~Ol adhesion is through
type 1 chain epitopes. Colo201 cells were treated with various
; mAb's following exposure to various sialidases and w~re assessed
~15 for residual binding. ~ Colo201 reactivit~ with mAb~CA1~-~
~ ~direc~ed to SLeaI) was inhibited almost completely by Vibrio
: ~ æialidase, and to a lesser extent by Arthrobacter and NDV
sialidases. ~ ;
: ` :
In contrast, Colo201 reactivity with mAb FH7 (directed to
di-S~ea and SLe~ ;was reduced by Arthrobacter ~ia}idase but
minimal~ly affected by Vibrio or ND~ sialidasesO Colo201
reactivity with ~mAb CA3F4 (Nudelman et al. J. Biol. Chem.
261:54E7, 1986) was enhanced by siali~ase treatment. mAb CAl9-9
81 -
W~3/17~33 PCT/US93/0~
9gQo (
~ inhibited Colo201 adhesion slightly and was influenced only
,. .
minimally by Vibrio sialidase (Figure 16).
It is possible that the 5LE~ epitope present at the surface
of Colo201 cells is organized in such as way that it i5 (i) not
S susceptible to CA19-9 for E-selectin-dependent adhesion and
(ii) not ~ensitive to sialidase treatment. The inhibitory
effects of m~b FH7 (Nudelman et al. supra) and, more strikingly,
mAb CA3F4 on Colo201 adhesion to E-selectin~coated plates were
enhanced by pretreatment of cells with Vibrio sial.idase.
Arthrobacter sialidase reduced but did not abolish Colo201
adhesion. See Figure 17.
olo201 adhesion~to E-selectin-coated Pla~es in a dvnamlc flow
svstem
In the dynamic flow system, NDV sialidase had no effect on
Colo201 adhesion, particularly at high shear stressesu
Anti-SLea I mAb rAl9-9 had no effect, where~s anti-SLea II mAb
F~7 had a moderate inhibitory effect, i~ agreement with~esults
~rom the s~atic system.
At both low and high shear stre ~es, the strongest
inhibition of :adhesion was observed ~or mAb CA3F4, which is
~: directed to LR~ with an ~2~6 sialosyl substitution at the
penultimate GlcNAc. Vibrio sialidaser which effici~ntly cleaves
~: terminal ~2-3 sialosyl linkages but is less effective at removing
internal sialic acid residues, reduced adhesion to ome exte~t
at high shear stress t but less so at low shear stress.
.
- 82 -
Z129987
f~ WO 93/17033 PCT/US93/0137~
Similarly, mAb CA3F4 inhibited adhesion strongly at high
shear stress but much less at low shear stress. A combination
of Vibrio sialidace plus mAb CA3F4 produced strong inhibition at
both high and low shear stress ~Figure l8).
A similar trend was observed for N~V sialidase, which
specifically cleaves terminal a2-3 sialosyl linkages. Colo201
adhesion, at either high or low shear stress, was not affected
by NDV sialidase alone, nor by NDV sialidase ~ollowed by
mAb CAl9-9. In contrast, adhesion was inhibited strongly, at
~10 both high and low shear stress, by N~V sialidase followed by mAB
FH7 or C~3F4.
Rolling ~elocity (~m/sec) of Colo201 cells along
E-selectin-coated plates was evaluated after treatment with
various sialidases and mAb's. At three different shear stresses,
velocity was unaffected ~y NDV sialidases and mAb CAl9-9; i.e.,
cells once stopped;(0 ~m/sec) did~ not start rolling again.
However,~ Vibrio or Arthrobaater sialidase, or mAb CA3F4, caused
.
once-stopped cells to start rolling again; the resulting velocity
depending on~shear stress. The greatest velocity resul~d from
io: treatment with~a~combination of CA3F4 plus ~ibrio sialidase.
Colo201 adheslon to activated MUYEC~s in a static sy~stem
As noted for Colo201 adhesion to E-selectin-coated plates,
mAb'~s~ CAlg-9 ;and FH7~ had~ negligible effect, but mAb CA3F4
strongly~ inhibited adhesion~. Treatment of cells with NDV
25~ sialidase,~which~cleaves terminal a2~3 sialosyl linkage, either
briefly or for 24 hr,~ did not reduce adhesion significantly.
83 -
`:: : : : ::
~99~ g3/l7o33 PCT/US93/o--'5
Treatment for 24 hr with Vi~rio sialidase completely ~bolished
adhesion. Brief treatme~lt with Vibrio and Arthrobacter sialidase
still reduced adhesion significantly.
Colo201 adhesion to activated HUVEC's i~ a dynamic_flow sYstem
Inhibition of adhesion was strongest for m~b CA3F4 and
non-existent for mAb CA19-9. mAb INH1 or ST421, directed to
unsubstituted type 1 chain (Stroud et al. J. Biol. Chem~
266:8439, 1991), ~lso caused significant inhibition. Adhesion
was unaffected by NDV sialidase but strongly inhibited by Vi~rio
sialidase.
The observed eff~cts of sialidases and mAb 9 S on Colo201
adhesion to ~-selectin-coated pla~es and to HUVEC's suggest that
~ the type 1 chain epitope recognized by E-selectin is internally
: sîalo~ylated and fucosylated.
EXAMPLE 8
, ~
Truncated, recombinant E~AM-1 lacking th~ transmembrane and
cytoplasmic domains is u~ed to coat beads, for example, capable
o~ packing into a standard chromatography columns. The EL~M--1
at a c9ncentration of 0.1-1 ~g/ml is mixed with the beads and the
' 20 mi~ture is incubated t~ allow binding o~ ELAM-l to the bead
;:~ matrix. A suit~ble incubation period is 12-24 hours at
: ~ 4^C - room temperature, The beads are washed ~o remove unbound
EI~M-l, optionally ca~ be blocked wi~h an inert carrier, such as
BSA, and washed again.
- 84 -
WO93/17033 2 1 2 9 ~ 8 7 PCT/US93/0137;
f~
The ELAM-1 coated beads can be used in a batch process or
packed into a suitably-sized column.
Cells known to carry carbohydrates blndable to ELAM-1, such
as HL60 and Colo201, are obtained. The cells are lysed to vbtain
a membrane fraction using known methods, such as repeated
freeze-thaw cycles. ~he membrane fraction is obtained, or
example, by centrifugation.
I~ the ~ell source is known to express only or predominantly
carbohydrates bindable to ELAM-1, the membrane prep may be a
suitable source without further puri~ication.
The membrane prep is treated using known methods to obtain
a membrane component preparation, and in particular, a fraction
that contains cell surface carbohydrate. The carbshydrate-rich
; fraction is mix~d with or paæsed over the ELAM-1 affinity matrix,
}5 depending on the format, the exposed matrix is washed and the
carbohydrates bound to the matrix are eluted, for example, by
: exposing the matrix to a high ~alt buffer.
The resultant preparation comprises carbohydrate bindable
o ELA~-l and the various species are separable usi~g known
: 20 techni~ues~ 6uch as TLC or HP ~
~: : EXAMPLE_s
:, ;
arbohydrates bindable to ELAM-l, either prepared chemically
using known reagents and:method~, see, for example, Example 1
hereinabove, prepared enzymatically or obtained from suitable
cells, see, ~or example, Example 8 hereinabove, or whole cells
: - 85 -
W~93/17033 PCT/US93tO~
~99~ known to express carbohydrate bindable to ~LAM-l, serve as
immunogen in suitable hosts to qenerate antibody thereto. Either
polyclonal or monoclonal antibody can be obtained and the
salection of a suitable host is premised on known methods and
preferences. The ~arbohydrates, cells, cell lysates or membrane
preps are administered to the ho5t, ~ither with or without
adjuvant, in a schedule that will generate an immun~ response.
In th~ case of polyclonal antisera, the blood is collected,
serum separated and tested.
In the case of monoclonal antibodies, the spleens of the
host animals are ~emoved and cells there~rom are fused with a
suitable myeloma cell using known techniques.
Specificity o~ the antibodies can be tracked using an ELISA
comprising, for example, purified recombinant E~AM-l and mAb 3B7
with the appropriate labeled reagents and repoxter molecules.
Antibody directed to carbohydrates o~ formulae (I), (II) and
(III) can be obtained by usi~g speci$ic carbohydrate species as
antigen and in the screening ELISA.
Alternatively, the antisera can be made "monospeci~c" by
20 absorption with cells carrying only S~eX and/or SLe or with a
~olid mat~ix to which SLeX and/or SLen is bound. The resultant
residual activity dir~cted to carbohydrates bindable to EL~M-l
can be attributed in part to antibodies directed to carbohydrates
of for~ula (I), ~II) or (III).
,
:
'
- 86 -
f^~ WO93/17033 2 12 9 9 8 7 PCT/VS93/0137
EXAMPLE 1o
NS-l cells were obtained.from the ATCC (Rockville, MD) and
maintained in RP~I 1640:~ulbecco's MEM (l:l) supplemented with
10% HI FCS. Fifty ~g of a plasmid comprising cDNA of E-selectin
in vector pCDM8 (R & D Systems, Minneapolis, MN) ~nd 5 ~g of
pSV2-neo (ATCC) were co-trans~ected into NS-l cells (l x 107) by
electroporatîon. After 48 hours iIl culture, the cells were
tran~ferred to medium containing 650 ~g/ml G418 (Gibco, Grand
Island, NY).
l0~ter 15-20 days, resulting colonie were screened for
E-selectin e ~ ression by s~aining with mAb (obtained from W.
Newman, Maryland Research Laboratories, Otsuka Pharmaceu~ical
Co~, Rockville, MD). The variant expressing the highest level
of E-selectin was isolated by panning with m~b followed by
15 li~niting di~ution to achieve clonality.
E-selectin-dependent adhesion using transfected NS-l cells
onto SLe~,~ SLeX, LeY, LeX, H-2, sialylparagloboside (SPG),
disialosyl I (structure 6 of Figure 2~ and dilaer~ c ~L~x were
com~?ared at various shear~ stress c:onditions. The number of cells
20 adher~d~ per ~ is expressed relative tc~ adhesion onto
SLeX-coated ~pIates ~whic:h is regarded as 100%.
:~Whereas a~esion~ to LeY and LeX slightly increased at high
shear stress, the absolute numbers of cells which adhered was
;~ ~ much lower~at both~low and high~shear stress conditions relative
2~ to that observed with SLRY and S~e. Adhesion with s~ructure 1
o~ Fi~ure 20 was e:nhanced at middle and high shear stress
conditionsO
-- 8 7
W093/t7033 PCT/US93/0~ ~
The high binding capacity of structure 1 of Figure 20 was
revealed further using low concentrations of glycolipids in
liposomes (glycoliposomes). Transfected NS-l adhesion on SLeX
liposomes was enhanced when LeX or ~ey was added and presented as
a mixed glycoliposome. The results o~ the experiments are
presented in Figures 21-24.
From the foregoing, it will be evident that; although
specific embodiments of the invention have been described herein
for purposes of illustration, various modifications may be made
without deviating from the spirit and scope of the invention.
All references cited herein are incorporated by reference.
.
~ - 8~ -
