Note: Descriptions are shown in the official language in which they were submitted.
~ 3
1 BACKGROUND OF THE INVENTION
2 The present ;nvention relates to conjugates of therapeutic
3 agents for the treatment of disease ;n mammals.
The term therapeutic agent includes not only the large
number of drugs used in the treatment of disease, but also a large
6 number of enzymes which are used ;n the treatment of genetic, metabolic
and malignant diseases. These diseases ;nclude a wide range of Inborn
8 Errors of Metabol~sm in which specific enzymes in the body are either
g deficient or deFective. Suitable enzyme replacement might provide an
appropriate therapy. In other diseases, cancer for example, certain cells
1l have been shown to be sensitive to a spec;f;c enzyme. For example,
12 in Acute Lymphocytic Leukemia the tumor cells are sensitive to the
enzyme L-asparaginase. The tumor cells;have an absolute requirement for
exogenous L-asparagine and cannot survive in the presence of
L-asparaginase, the enzyme responsible for the removal of the sub-
16 strate. Enzyme treatment of this type ;s usually termed enzyme therapy.
There are problems associated with the injection of many
18 of the therapeutic agents, especially enzymes, into the body. Firstly,
19 the therapeutic agent may ~e foreign to the body and can therefore
cause the body's immune system-to mount a rejection reaction. The
21 therapeutic agen~ is then rapidly cleared from the body following each
22 subsequent administration. This rejection reaction, in its severest
~3 form called an anaphylactic reaction, may threaten the life of the
24 recipient
Secondly, especially in the case of enzymes, the therapeut;c
26 agent is frequently heat labile and sensitive to proteolysis by other
27 circulating enzymes. Rapid bioinactivation of the therapeutic agent
28 can occur. This may necessitate repeated administration of large
29 dsse oF the therapeutic agent. In addition to the expense and bother
which this causes, there is al~o the increased risk of complicating the
31 above described immunological reaction, thereby increasing the possibility
32 of anaphylaxis.
- 2 -
5 0
l Thirdly, it ;s often necessary or desirable to deliver or
2 target the therapeutic agent to specific body cells or organs requiring3 action. For instance, in many enzyme defiiciency diseases the defect
4 is intracellular. Due to the lack of a functioning enzyme, the
substrate which accumulates in the body cell is compartementalized in
6 intracellular organelles termed lysosomes. Frequently, the substrate
7 is stored only in specific tissues or organs. Thus to effect enzyme
8 replacement it may be necessary to target the enzymes to these specific9 sites. This has proven to be a major limitation to the treatment of
such fatal childhood énzyme deficiency diseases as lipid storage
11 diseases and glycogen storage diseases.
12 Sorne, but not all, of the above problems have been solved13 in the prior art by conjugating the therapeutic agent with a carrier.
14 Carriers are divided into two groups, those that cause highly specific
binding to cell-surface receptors and those that do not. The former
16 type of carrier will hereinafter be referred to as a targeting agent.
17 Targeting of drugs is well docurnented, see for example
18 the revlew article by M. J. Poznansky and L. G. Cleland in Drwg
19 Delivery Systems, ed. Julianog Oxford University Press, New York,
1980, pg. 253 - 315. Targeting agents are molecules, frequently bio-
21 logical macromolecules, which bind to specific receptor sites on the sur-
22 faces of body cells. Known targeting agents include serum hormones,
23 antibodies against cell surface antigens, and lectins.
24 It is also known in the prior art to protect therapeutic
agents by conjugation with an appropriate carrier. Conjugates of
26 albumin and therapeutic agents are documented in the above-referenced
27 review article and in an article by M. J. Poznansky and D. Bhardwaj in
28 Canadian Journal oF Physiology and Pharmacology, 58 , 1980, pg. 322-325.
.
s .;
- ~ 3 6~15(~
1 These albumin conjugates have been shown to be both non-immunogenic and
2 non-antigenic, see for instance M. H. Remy and M. J. Poznansky, The Lancet,
3 July 8, 1978, pg. 68 - 70. The albumin ;s believed to mask the antigenic
sites on the therapeutic agent such that the recip;ent recogn;zes the
conjugate as self and therefore does not mount an ;mmune response. Further,
6 these albumin conjugates have been shown to be more resistant to bioin-
activation than was the free therapeutic agent.
8 To my knowledge, no attempt has been made to target these
g albumin-protected therapeutic agents. Prior to the present invention
it was not known whether a targeting agent would be effective in targeting
11 such a large and complex molecule. Further, it was not known whether12 attachment of a targeting agent would interfereiw;th the biological
13 act;vity of the albumin-therapeutic agent conjugate.
:
14 SUMMARY OF THE INVENTION
I have discovered that conjugates of albumin and therapeutic
16 agent can be made targetable by chem;cally linking the conjugate to a
17 targeting agent having binding specificity for receptor sites on body
18 cells against which it is desirable to direct the therapeutic agent.
19 Known target;ng agents, including serum hormones, cell-surface directed
antibodies, and lectins, are suitable for this purpose.
21 While it might have been eY~pected that the albumin carrier22 molecule would mask or interfere with the targetability.~ the-targeting
23 agent once it was l;nked to the therapeutic agent - albumin conjugate,
24 th;s was Found not to be the case. The targeting agent, linked to the
conjugate, was found to reta;n its ab;lity to specifically deliver the
26 conjugate to speciF;c cellular surface receptors. Further, the targeting27 agent was Found not to interfere with the ab;lity of the body cell to
28 utilize the therapeut;c agent once the conjugate was delivered to the cell.
29 This was not a predictable property of the conjugates of the present
invention.
-- 4 --
~ ~ 6 ~
1 The albumin carrier ;s included ~n the conjugate in an
2 amount sufficient to mask the antigenici~ty of the therapeutic agent.
3 The term 'to mask the antigenic;ty' ;s meant to infer that the conjugate
4 does not illicit an immune response. The al~umin used is most preferably
homologous to the mammal intended as the recipient. Since in most
6 cases the intended recipient is human, human serum albumin is preferred.
7 Other sources of albumin, for example bov;ne serum albumin and horse
8 serum albumin are useful for therapy in cattle and horses respectively.
9 As explained, the targeting agent allows for delivery of
the conjugates to specific cellular surface receptors. The cho;ce of
11 targeting a~ent therefore depends on the desired site of delivery. The
12 present invention exemplifies the use of insulin, immunoglobulin G,
ant~ibody against human pancreatic tumor cells, and antibody against
14 hepatocytes as targeting agents. Insulin ;s an example of a serum
hormone while the rest are examples of cell-surface directed antibodies.
16 The therapeutic agent in the conjugates is chosen from the
17 known drugs and enzymes used in the treatment of disease in mammals.
18 The conjugates of the present invention are illustrated with three
19 examples of therapeutic agents, ~-1,4-glucosidase9 superoxide dismutase,
and L-asparaginase. Each of these enzymes is chem;cally linked to
21 carrier albumin and one of the above-listed target;ng agents to
22 demonstrate the targetability and effectiveness of the conjugates.
23 A deficiency of the a~l,4-glucosidase enzyme in humans is
24 known as Pompe's disease or Type II glycogen-storage disease. The
deficiency causes death, usually by cardio-respiratory failure, before
26 the age of 2 years. The absence of this enzyme results in the intra-
27 cellular accumulation of glycogen in what are believed to be lysosomes in
28 the liver and in both respiratory and cardiac muscle. The conjugates
29 of the present invention with ~-1,4-glucosidase are shown to retain
the enzyme activity after cross-linking to the albumin carr;er and the
31 targeting agent. The conjugates are also shown to be non-immunogenic.
.:
1 Further, the conjugates are shown to greatly enhance the amount of the
2 ~-1,4-glucosidase enzyme delivered to muscle cells.
,
3 As mentioned previously, L-asparaginase has the potential
, .
4 of being an important therapeutic agent against Acute Lymphocytic
Leukemia. Conjugates of L-asparaginase, albumin and targeting agent
6 are shown to be non-immunogenic, biologically active and targetable to
7 specific cell types~ for example human pancreatic tumor cells.
8 Superoxide dismutase (SOD) may be used as a therapeutic
g agent in the treatment of rheumatoid arthritis where it may act to
lo reduce superoxide free radicals. The conjugate of SOD, albumin and
1l antibodies against hyaluronic acid may be targeted to joint tissue known
12 to be susceptible to this Form of arthritis.
The process used to chemically link the therapeutic agent
to the carrier albumin and in turn to the targeting agent tlepends
on the functional groups of the particular therapeutic agent and
16 targeting agent. The linking process should be chosen to preserve
water solubility of the final conjugate product, to preserve the
18 activity of the therapeutic agent, and also to preserye the site
19 specific binding capabilîty of the targeting agent. In most cases
the l;nking process utilizes a cross-linking agent between the therapeutic
21 agent and the carrier albumin and between the carrier albumin and the
22 targeting agent. Glutaraldehyde, sodium periodate and water soluble
23 carbodiimide are exemplified as suitable cross-linking agents in the
24 specific conjugates illustrated herein.
.~
- 6 -
1 Broadly stated the invention provides a novel composition
2 oF matter, comprising~ in water soluble, sterile and non-immunogen;c
3 form, a therapeutic agent chemically linked to an albumin carrier,
4 and a targeting agent chemically linked to the albumin, the amount of
albumin beîng sufficient to mask the antigenicity of the therapeutic
6 agent, and the targeting agent having binding specificity for receptor
7 sites on cells against which it is desirable to direct the therapeutic
8 agent.
g The present invention also provides a process for
producing a water soluble, sterile and non-immunogenic conjugate of
11 a therapeutic agent with an albumin carrier and a targeting agent,
12 comprising the steps of: (a) chemically linking the th0rapeutic
13 agent with an amount of the albumin carrier sufficient to mask the
antigenicity of the therapeutic agent; (b) chemically linking the
resulting complex of the therapeutic agent and the albumin with a
16 targeting agent having binding speci-Ficity for recptor sites on cells
17 against which it is desirable to direct the therapeutic agent~ said
18 binding specificity being retained after the linking reaction.
1 :i6~5Q
:.
1 DESCRIPTION OF THE PREFERRED EMBODIMENT
2 The three component conjugates ~f the present in~ention
3 are useful in delivering therapeu~ic agents to specific body
4 cells, t.issues or organs in mammals for the treatment of disease.
The therapeutic agents which can be delivered in this form
are usually those drugs and enzymes for which the characteristics of
7 avoiding immunological reactivity by antigenicity masking and site
~ specific targeting are desired. Exemplary drugs include known
9 alkylating agents, antibiotics, antiparasitic agents and chelating
agents. The enzymes useful in the conjugates include th~se enzymes
11 causitive of enzyme deficiency diseases and those enzymes ;ntended
12 for enzyme therapy. A partial list of enzyme deficiency diseases
13 together with the enzyme responsible for the deficiency is included in
14 Table I. Exemplary enzymes for enzyme therapy purposes include
L-asparaginase, uricase and superoxide dismutase. Persons skilled
16 in the art will know of other therapeutic agents which would be
17 desirably conjugated in accordance with the present invention.
1 5 0
1 TABLE I
2 DOCUMENTED ENZYME DEFICIENCY DISEASES
3 Disease Enzyme
.
4 Acatalasia Catalase
Albinism Tyrosinase
6 Alcaptonuria Homogentisic Acid Oxidase
7 Cholesteryl Ester Deficiency Lecithin Cholesterol Acyltransferase
8 Cystathioninuria Cystathionase
9 Disaccharide Intolerance III Lactase
Fructose Intolerance Fructose-l-Phosphate Aldolase
11 Fructosuria Hepatic Fructokinase
12 Galactosemia Galactose-l-Phosphate Uridyl Transferase
13 Gangliosidosis (GMl) ~-Galactosidase A, B, C,
14 Gaucher's Disease Glucocerebrosidase
~6PD Deficiency Glucose-6-Phosphate Dehydrogenase
16 Glycogen Storage Disease I Glucose-6-Phosphatase
17 Glycogen Storage Disease II a-l ,4-Glucosidase
18 Glycogen Storage Disease III Amylo-1,6-Glucosidase
19 Glycogen Storage Disease V Muscle Phosphorylase
Glycogen Storage Disease VI Liver Phosphory1ase
21 Glycogen Storage Disease VII Muscle Phosphofructokinase
22 Glycogen Storage Disease VIII Liver Phosphorylase Kinase
23 Hemolytic Anemia Glucose-6-Phosphate Dehydrogenase
24 Hemolytic Anemia Phosphoglycerate Kinase
Hemolytic Anemia Pyruvate Kinase
26 Histidinemia Histidase
27 Homocytinuria I Cystathionine Synthetase
28 Hydroxyprolinemia Hydroxyproline Oxidase
29 Hyperlipoproteinemia II Lipoprotein Lipase
Hyperlysinemia Lysine-Ketoglutarate Reductase
I ~ 6 ~ ~ 5 ~
1 TABLE I (Continued)
2 _isease_ Enzyme
3 Hypoglycemia (Acidosis) Fructose-1,6-Diphosphatase
4 Immunodeficiency Disease Adenosine Deaminase
Intestinal Lactase Deficiency Lactase
6 Krabbe's Disease A ~-Galactosidase
7 Lesch~Nyhan Syndrome Hypoxanthine-Guanine Phosphoribosyl
8 Transferase
9 Mannosidosis ~-Mannosidase
Maple Sugar Urine Disease Keto Acid Decarboxylase
11 Metachromatic Leu.kodystrophy ArylsulPatase A
12 Muoopolysaccharidosis I a-L Iduronidase
13 Mucopolysaccharidosis III Heparin Sulphate Sulphatase
14 Mucopolysaccharidosis VI Arylsulfatase B
Mucopolysaccharidosis VII ~-Glucoronidase
16 Niemann Pick Disease Sphingomyelinase
17 Orotic Aciduria II Orotidylic Decarboxylase
18 Pentosuria L-Xylulose Reductase
19 Phenylketonuria Phenylalanine Decarboxylase
Pyruvate Carboxylase Def. Pyruvate Carboxylase
21 Richner-Hanhart Syndrome Tyrosine Aminotransferase
22 Sandhoff's Disease Hexosaminidase A, B
23 Tay-Sachs Disease Hexosaminidase A
24 Tyrosinem~a Tyrosine Transaminase
Xanth;nuria Xanthine Oxidase
-- 10 --
l l S8 ~
1 The albumin carrier is included in the conjugate in a molar
2 excess to the therapeutic agent in order to mask the ant;genicity of
3 the therapeutic agent. Homologous albumin is most preferably used in
4 the conjugate so as not to cause the conjugate to triyger an immune
response ;n the recipient mammal. Heterologous albumin may be used only
6 if it does not illicit a pronounced immunological response in the
7 recipient mammal.
8 The particular targeting agent included in the conjugate
9 is one which has binding specificity for specific receptor sites on
cells against which it is desirable to direct the therapeutic
11 agent. The targeting agent is selected from serum hormones, cell-
12 surface directed antibodies and lectins which are known to have receptor
13 sites on specific body cells. Exemplary of suitable targeting agents
14 are insulin, glucagon, epidermal growth factor, low-density lipoprotein,
human chorionic gonadrotropin, thyroid stimulating hormone,
16 asialoglycoproteins, mannosyl-term;nal glycoproteins, endorphins,
17 enkephalins, transferrin, melanotropin, cell-surface directed anti-
18 bodies (e.g. antibodies against tumor specific antigens , cell surface
19 antigens~, cell surface receptors), human growth hormone, a 2-
macroglobulin, melanotropin, plant and human lectins (e.g. peanut lectin,
21 wheat germ lectin, concanavilin A, protein A), and galactose term;nal
22 glycoproteins. The amount of targeting agent included in the conjugate
23 has not been found to be critical; however, a molar excess of the
24 targeting agent to the therapeutic agent has been found to improve the
delivery of the conjugate to the desired cells.
:
~ ~ ~ 8~
1 To prepare the thearpeutic agent - albumin-targeting agent
2 conjugates, the therapeutic agent is first chemically linked to the
3 molar excess of albumin. The result;ng theYapeut;c agent-album;n
4 conjugate ;s thereafter chemically linked to the preferred molar excess
of targeting agent. This order oF linking is preFerably used so that
6 the binding sites of the targeting agent rema~n substantially clear
7 for binding to receptor sites.
8 The functionality of the therapeutic agent and targeting
g agent usually require that a cross-linking agent be used for each of
the chemical linking steps. The particular cross-linking agent chosen
11 will of course depend on the functional;ty of the specific components
12 being linked. The cross-linking agents will usually utilize carboxyl
13 groups, amino groups, sulfhydral groups or sugar residues on one or
14 both of the components to be linked.
A large number of cross-linking agents are known, see
16 for ex~mple the previously referenced review article by Poznansky and
17 Cleland (1980). A partial list of suitable cross-linking agents includes
18 glutaraldehyde, water-soluble carbodiimides, sodium periodate
19 (periodate oxidation), dithiothreotol (disulfide reduction), diisocyanate,
cyanuric chloride, mixed anhydrides, imidoesters, bisdiazobenzidine,
21 cyanogen bromlde, p, p'-difluoro-m, m'-dinitrophenyl sulphone, N-
22 succininidyl-4~iodoacetylaminobenzoate, and diazonium salts.
23 The conditions for the cross-link;ng reaction, for
24 example pH, temperature and degree of cross-linking are chosen such
that the biological activity of the therapeutic agent, the binding
26 specificity of the targeting agent and the water solubility of the final
27 conjugate are maintained. Each of these characteristics of the final
28 conjugate can be tested for and the cross-linking conditions adjusted
29 accordingly by persons skilled in the art.
I 3 &~3 1 5 ~3
1 The conjugates of the present invention are utilized in a
2 sterile form to treat diseases in warm blooded mammals. To that end
3 the conjugate is injected intramuscularly, subcutaneously, intravenously,4 intraperitoneally, intracranially or intradermally, depending on the
desired site of action, into the recipient patient. Alternatively
6 there may be topical applications oF~:the con~ugates. The dosage used
7 will be dependent on such factors as the type and severity of the
8 disease, the size and species of the recipient patient, the toxicity
9 of the therapeutic agent and the degree of targeting attained by the
conjugate. The dosage can therefore be worked out by routine experiments
1l with each of the conjugates.
12 The present invention is exemplified by the following
13 specific embodiments which are meant to be merely illustrative and not
14 limitative of the invention.
Example 1
16 L-Aparaginase-Albumin-Insulin Conjugate Cross-Linked with Glutaraldehyde
17 and Carbodiimide
18 L-Asparaginase (5 mg) obtained from E. co~ was chemically
19 linked to homologous albumin (25 mg3 obtained from mouse or human
by react~on with 50 ~ 1 of 25% glutaraldehyde in 4 ml of phosphate
21 buffered saline (0.1 M potassium phosphate pH 6.8). The reaction
22 was performed at 4C for 4 hours in the presence of asparagine (5 mg)
23 in order to protect the active site of the enzyme. The reaction was
24 halted with the addition of glycine (50 mg). The product was then
separated from the unreacted monomeric components by dialysis, pressure
26 ultrafiltration or molecular sieve chromatography. The isolated product
27 was then cross-linked to bovine or porcine (15 mg) insulin by reacting
28 the same with water soluble 1-ethyl-3-(3-dimethylaminopropyl)
I 1 6S ~ 5 ~
1 carbodiimide HCl ~ECDI - 10 mg~ at 4C for 2 hours. The end product
2 was separated and purified by molecular sieve ch w matography ~hich
3 ind~cated that the final product had a molecular ~eight ranging from
4 9 x 105 to 1.4 x 106 The calculated mole ratio-was 1.5:15:90,
L-asparaginase:albumin:insulin.
6 The resulting L-asparaginase-album;n-insulin conjugate was
7 assayed for enzyme activity in accordance with the technique of Mashburn
8 and Wriston, Arch. Biochem. Biophys., 1964~ 105 , pg. 450 452. The
g product (enzyme:albumin:insulin = l:lO:Ç0, based on starting quantities)
was found to retain about 70% of the starting enzyme activity, as
11 reported in Table II. This represen~s a significant amount of enzyme
12 activity since, as will be shown below, the enzyme is now in a
13 protected form.
14 To demonstrate the resistance of the product conjugate to
proteolytic inactivation, equal amounts of the enzyme in free and
16 conjugated form were incubated with 5 units of trypsin (Sigma Chemical,
17 St. Louis, Mic.). The enzyme activity monitored as a function of time
18 is reported in Table III. The results show that the enzyme conjugated
19 in accordance with this invention was much more resistant to b;oinactivation
than was the free enzyme.
21 To test the binding specificity of the enzyme conjugate,
22 the-end product was labelled covalently with 125Iodine and then incubated
23 with mouse spleen cells. The % b;nd;ng of the enzyme was determined
24 after 20 minutes at 24C in accordance with the procedure of J. R. Gavin,
et al., Proc. Natl. Acad. Sci., U.S.A., 71, (1974), 84 - 90. The
26 results, as reported in Table IV, show that the enzyme conjugate binds
27 to the spleen cells, which are known to possess insulin receptors, to
28 a much greater extent than does the free L-asparaginase or the L-
29 asparaginase-albumin conjugate. The insulin is therefore shown to be
an effective targeting agent. Conjugates having insulin to albumin
- 14 -
~ 5 ~
1 molar ratios ranging from 1:1 to 1:10 (data is given for 1:6) were
2 found to retain the binding qualities of the insulin.
3 The L-asparaginase-albumin conjugates (absent the -targetlng
4 agent~ were shown to be non-immunogPnic in both tissue culture and whole
animal experiments in accordance w;th techniques reported by Remy and
6 Poznansky, The Lancet ii, Jul~ 8, 1978, pg. 68 - 70. The addition of
7 insulin to the conjugates does not affect the immunogenicity of the
8 of the resulting conjugate. The immune response to the free L-
9 asparaginase enzyme is compared to the immune response from the L-
asparaginase-albumin conjugates in Table V~ The L-asparaginase-albumin
11 conjugates tested had ratios of enzyme to albumin ranging from 1:5
12 to 1:20. As indicated in the table a molar excess of 10:1 albumin to
13 enzyme is sufficient to mask any antigenicity of the enzyme.
,
"
- 15 -
~ ~ ~ 8 1 5
1 TABLE II
2ACTIVITY OF ENZYME AND ENZYME CONJUGATES
3 % Activity
4 Enzyme Preparation Enzyme Activ;ty Retained
L-asparag;nase 400 units~mg**
6 L-asparag;nase-albumin 310 units/mg** 77.5
7 L-asparaginase-albumin-IgG 290 units/mg** 72.5
8 L-asparaginase-albumin-(Fab')2 . 301 units/mg** 69.8
- 9 L-asparaginase-albumin-insulin 280 units/mg** 70.0
~-1,4-glucos;dase 8.5 units/mg***
11 ~-1,4-glucosldase-albumin 6.6 units/mg*** 77.6
12 ~-1,4-glucosidase-albumin-IgG 6.0 units/mg*** 70.5
13 ~-1,4-glucos;dase-albumin-insulin 5.8 un;ts/mg*** 68.2
:
.
14 * % activity retained calculated on a per mg enzyme basis
** E.CoR~ L-asparaginase assayed as per (Mashburn & Wriston, 1964,
16 Arch. Biochem. Biophys. 105, 450 - 452.
17 *** 1, 4-glucosidase assayed as per (de Barsy et al., 1972, Eur. J.
18 Biochem. 31, 156-165) Data is for ~ 4-glucosidase from human
19 placenta.
',
.~
- 16 -
~ 5
1 TABLE III.
2 TRYPSIN SENSITIVITY OF L-ASPARAGINASE-ALBUMIN-INSULIN CONJUGATES
3 .. .~ - T 1/2 at 37C ~ *
4 Enzyme Preparat;on 5 Un;ts Trypsin
L-asparaginase 10 min.
6 L-asparaginase-albumin-insulin 120 m;n.
7 * Equal amounts of enzyme in free and polymeric form were incubated
8 wi~h 5 units of Trypsin (Sigma Chemical, St. Louis MI~ and the enzyme
9 act;vity was monitored as a functiorl of time as per Tabl~e II.
.,
.~
,.,.~
- 17 -
B8~50
1 Example 2
2 L-Asparaginase-Albumin-Insulin Conjugate~ Cros~-Linked with Glutaraldehyde
3 Following the procedure oF Example 1, L-asparaginase waslinked to homologous albumin with the glutaraldehyde cross-linking
agPnt. The resulting enzyme-albumin complex was then cross-linked to
6 insulin using the same glutaraldehyde cross-linking conditions. A
7 molar ra~io of 1:14:60 of L-asparaginase:albumin:insulin was used. The
8 separated conjugate had a molecular weight of 1.2 x 106 Daltons.
g The physiological properties of this conjugate were very1~ similar to those of the conjugates produced in accordance with Example 1.
: .
~ 11 Example 3
;
12 ~-1,4-Glucosidase-Albumin-Insulin Conjugate Cross-Linked with Glutaraldehyde
13 or Carbodiimide
Human placental ~-1,4-glucosidase (2 mg, 150 units) was
cross-linked with human albumin (20 mg) using 5 ~g of glutaraldehyde.
. ~
16 The reaction conditions of Example 1 were maintained except that the
enzyme substrate included was p-nitrophenyl glucoside. The resulting
18 enzyme-album;n complex was then cross-linked`to insulin (2 mg) us;ng
g either glutaraldehyde or ECDI, again in accordance with the conditions
of Example 1. The separated end product had a molecular weight of
21 1 x 106 Daltons when a molar ratio of 1:12:12 of en y me:albumin:insulin
22 was used. Some properties~of these conjugates are shown in Tables II,
28 IV and V.
24 The end produc~ was tested for enzyme activity in accordance
with the procedure of de Barsy et al., Eur. J. Biochem.,31, 1972,
26 pg. 156 - 165. As indicated in Table II, the conjugate retained
27 enzyme act~vity against an artificial substrate p-nitrophenyl glucose,
28 against maltose and against its natural substrate human glycogen
29 (from liver).
- 18 -
1 1 ~ 8 1 5 ~
1 The binding specificity of the enzyme conjugate was tested
2 by covalently labell;ng the conjugate ~ith ~25Iod;ne and then incubating
the conjugate with mouse spleen cells cr chick em~ryonic muscle cells
4 ;n accordance with the procedure ind;cated in Example 1. The data ;n
Table IV ;llustrates the bindin~ of the enzyme-albumin-insulin conjugate
~ to cells known to be high in insulin receptor activity. By subcelluler
-~ 7 fractionation of the tissue, ;t was found that the conjuga-te had been
8 internalized by the cell and was associated with a lysosomal fraction.
g The enzyme can be located within the cell in a fraction rich in acid
phosphatase act;vity known to be contained within lysosomes. Thus
1l the insulin targeting agent was shown not to interfere with the abllity
.
`;~ 12 of the body cell to utilize the ~-1,4-glucosidase therapeutic agent.
'`
:..
,,~ ,
.
'~
. .
19
i ~ 6 ~ ~ 5
1 TABLE IV
2 P~EFERENTIAL 3INDING OF INSULIN-CONJUGATED
3 ENZYME-AL3UMrN CONJUGATES TO SPLEEN CELLS
4 AND TO MUSCLE GELLS
.
Spleen Cells Muscle Cells*
6 Enzyme Preparation % Uptake
7 L-~sparaginase 0.81
8 L-Qsparaginase-Albumin Z.78
`~ g L-Asparaginase-Albumin-Insulin 22.70
" .,
~-1,4-Glucosidase 6.5 8.7
11 ~-1,4-Glucosidase-Albumin 6.9 9.0
12 ~ 4-Glucosidase-Albumin-Insulin 28.1 3U.l
13 * Enzyme preparations were labelled covalently with 125Iodine and
14 then co-incubated with either mouse spleen cells or chick embryonic
muscle cells an~ the % binding of the enzyme determined after 20
16 min. at 24C.
- 20 -
~ 5
1 TABLE V
2 IMMUNOGENICITY OF ENZYME AND ENZrME-ALBUMIN* CONJUGhTES
3 Enzyme Preparation _mmunogenicity**
4 L-Asparaginase*** ~+~
L-Asparaginase-Albumin (1:5)
6 L-Asparaginase-Albumin (1:10)
7 L-Asparaginase-Albumin (1:20)
8 ~-1,4-Glucosidase**** ~+
g ~-1,4-Glucosidase-Albumin (1:10)
~-1,4-Glucosidase-Albumin-Insulin (1:10:60)
11 * Homologous albumin is used in all cases: rabbit albumin if
12 immuno~enicity testing is to be performed in rabbits and mouse
13 albumin if testing is performed in mice. Testing is as described
14 (Remy & Poznansky, The Lancet, ii, (1978), 68 - 70).
** Immunogenicity scored on radioimmunoassay as - no reaction,
16 ~ slight reaction to **~ strong reaction.
17 *** L-asparaginase is from E. co~.
18 **** ~-1,4-Glucosidase is from human placenta
- 21 -
1 ~ 6 ~ ~ 5
1 Example 4
2 L-Asparaginase-Albumin-IgG Conjugate Cross-Linked with Glutaraldehyde
3 and Sodium Periodate
4 L-asparag~nase-albumin conjugates were prepared using
glutaraldehyde as a cross-linking agent and the conditions set forth
6 in Example 1 ~2.5 mg L-asparaginase~ 20 mg ~lbumin, 5 ~g glutaraldehyde
7 and 5 mg L-asparagine). The resulting conjugate was then cross-linked
8 to a monoclonal antibody (5.0 mg of anti-H-2k, an antibody against the
g mouse histocompatibility antigen H-2k, prepared according to Kennett,
R. H. et al. (1980) "Monoclonal Antibody Hybridomas: A New Dimension
11 in Biological Analyses. Plenum Press, New York) using 5 mg of sodium
12 periodate for 2 hours at 4C. The cross-linking procedure utilizes a
sugar residue on the Fc fragment oF the antibody and the amino groups
on the enzyme-albumin conjugate. The resultant conjugate was separated
as described in Example 1 and was found to have a^molecular weight of
16 1.1 x 106 Daltons and a calculated mole ratio of 1:10:2, L-
17 asparaginase:albumin:anti-H-2k antibody.
18 Using methods similar to those described in Example 19 the
19 resulting conjuyate was found to retain enzyme activity (see Table II)
toward the substrate L-asparagine. Two types of experiments indicate
21 that the conjugate binds preferentially to cells which possess the
22 corresponding H-2k antigen but not to cells which possess a different
23 histocompatibility antigen, the H-2d antigen. When 1 ~g of conjugate
24 (labelled wlth 125I) is incubated w;th 5 x 106 Balb/Ccr spleen cells
which contain the H-2d antigen less than 2% of the conjugate binds to
26 the cells. When the same amount of conjugate is incubated with the same
27 number of spleen cells from C3H mice which contain the H-2k antigen
28 45% of the conjugate was found bound to the cells after a 2 hour incubation
29 at 18C. This exper;ment is an in-vitro experiment where the cells
were grown in a tissue culture flask. Table VI demonstrates that the
~ ~8~5~
1 preferential binding of the targeted conjugates persists in whole
2 animal experiments. A significant increase in the retent;on of 125I~
3 labelled conjugate is observed ~hen the conjugate (containing the anti-H-
4 2k antibody) is injected ;nto Balb/Ccr m;ce (wh;ch possess the H-2d
antigen) which have been innoculated with tumor cells (RC3HED cells
6 which possess the H-2k antigen). Th;s suggests strongly that the
7 conjugate is binding preferentially to the tumor cells in the whole
8 animal experiment. The use of the histocompatibility ant;gen (H-2d
9 and H-2k in mice) presents a convenient cell surface antigen commonly
found in varying mouse strains. The analagous use of human histo-
11 compatibility antigens as cell surface targets (i.e. using the cor-
12 responding antibody) might be expected to be useful if this provides
13 a proper target for a given enzyme or drug therapy.
- 23 -
~ 3 ~
1 TABLE VI.
2IN-VIVO TARGETING OF L-ASPARAGINASE CO JUGATES TO TUMOR CELLS
3% 125I-Labelled Enzyme Remaining
4 15 h 24 h 48 h
Free L~Asparaginase . 9% 4% 3%
6 L-Asparaginase-Albumin 41% 17% 11%
7 L-Asparaginase-Albumin-Anti-
8 H-2k Antibody 74% 51% 37%
9Balb/Ccr mice possessing the H-2d antigen were injected with C63HED
tumor cells which possess the H-2k antigen. 125I-labelled enzyme
11 preparations were injected intravenously ;nto the mice and the %
12 lahel remaining determined after varying time periods.
~'
.
':
.
- 2~ -
---
~ 1 6 ~ ~ ~ 0
l Example S
2 L-Asparag;nase-Albumin-(Fab'~2 Fragment.of IgG Conjugates
. . .
3 L-Asparaginase-al~umin c~njugates~ were prepared using the
4 glutaraldehyde cross-linking agent and the conditions set forth in
Example 4. The resulting conjugate was then cross-linked to an equal
6 molar quantity of the (Fab')2 fragment of the monoclonal anti-H-2k
7 antibody using either sodium periodate or ECDI as the cross-linking
8 agent. Tbe resulting end conjugate had a molecular weight of about l x
9 106 Daltons. The procedure was exactly~as in Examplé 4 except~that
only the (Fab' )2 fragment of the ant,i,body molecule was use~d.
ll In testing procedures similar to those of Examples 1 and 4,
12 the enzyme-albumin-(Fab')2 conjugate was found to retain both L-
13 asparaginase activity (see Table II) and binding affinity to cells
14 possessing the H-2k antigen. The bind;ng specificity was shown to
exist in both t;ssue culture and whole animal (mouse) exper1ments as
16 dçscribed in Example 4.
17 The IgG targeting agent used in Example 4 included both
18 the Fc fragments and the (Fab')2 fragments. The use of the ~Fab')2
l9 fragments only in the present example is preferred since conjugates of
the (Fab')2 fragments retain the ligand properties of the antibody
21 for binding specifically to antigen receptors~ while the possibility
22 of the conjugate binding to the less specific Fc receptors is removed.
23 Fc receptors are found on a wider range of cell ~ypes. Further, use of
24 the (Fab')2 fragment in place of the entire IgG molecule has the added
advantage o~ rendering the entire polymeric complex less immunogenic
26 because of the absence of the Fc fragment.
t. ~ V
l Example 6
2 ~-1,4-Glucosidase-Albumin-Immunoglobulin Conjuyates
3 Conjugates of ~-1,4-glucosidase and albumin were produced
4 in accordance with the conditions of Example 3. The conjugates were
then chemically linked to a heterologous rabbit immunoglobulin
6 preparation that had been prepared against isolated rat hepatocytes.
7 The mole ratio was 1:10:2~ ~-1,4-glucosidase:albumin:immunoglobulin.
8 Both the ECDI or glutaraldehyde cross-link;ng condi~ions were found to
9 be successful in this linking step. The separated enzyme-albumin-immuno-
gl~bul;in conjugate had a ~ ecular weight of 1.2 x 106 Daltons and was
1l composed of an average~of one molec~n e enzyme, twelve molecules albumin
12 and one and a half molecules~of antibody.
13 Testing procedures similar to those used in the previous
14 examples showed that the final conjugate retained enzymatic activity
(see Table II). The complex also was shown to be taken up preferentially
16 by rat hepatocytes over other cell types in whole animal experiments
17 (see Table VII). Comparison of the data in Table VII for the ~-1,4-
18 glucosidase-albumin-immunoglobulin (control) conjugate and the ~-1,4-
19 glucosidase-albumin ilmmunoglobulin (anti-hepatocyte) conjugate shows that
preferential uptake by hepatocytes occurs only when the immunoglobulin
21 molecule is directed against hepatocytes.
22 The prepared enzyme-albumin-immunoglobulin conjugates were
23 found to give similar results whether the intact immunoglobulin
24 molecules or only the (Fab')2 fragments were used as the targeting agent.
- 26-
~ 16~.5~
,~
1 TABLE VII
2 TARGETING OF ENZYME-ALgUMrN-IMMUNOGLOBULIN
3 CONJUGATES TO RAT HEPATOCYTES
. ,
Enzyme Preparation Hepatocyte!Kupffer Cells
~-1,4-Glucosidase 0.10
6 ~-1,4~Glucosidase-Albumin 0.17
7 ~-1,4-Glucosidase-Albumin-Immunoglobulin (Control)* 0.16
8 Immunoglobulin (Anti-Hepatocyte) 0.91
9 ~-1,4-Glucosidase-Albumin-Immunoglobulin
(Anti-Hepatocyte) 1.23
.,
11 125I-labelled enzyme and enzyme conjugate preparations were injected
12 into rats at time zero . After 90% of the label had cleared from the
circulation, the liver was excised. The Kup~fer cells and hepatocytes
were then separated and the percent label in each fraction was determined.
,:
* Control immunoglobulin was one which was not directed against rat
16 hepatocytes.
,
- 27 -
I ~ 3 L 5 (i
1 Example 7
2 L-Aspara~inase-Albumin-Human Pancreatic Tumor Cell Antibody Conjugates
3 L-Asparaginase was cros~s-linked with human serum albumin
4 and antibody in accordance with the pracedure set forth in Example 4
with the exception that the antibody was a monoclonal antibody directed
6 against human pancreatic tumor cells. Human pancreatic tumor cells
7 were grown in suspension culture in accordance with the techniques
described by Yunis, A. A. et al., Int. J. Cancer, 19, (1977), pg. 128.
9 Monoclonal antibodies against the cells were produced as described
in the Kennett reference cited in Example 4.
11 In accordance with the previous test procedures, the enzyme
12 conjugate was shown to retain enzyme activity and to be non-i~munogenic
13 and resistant to bioinactivation. These test results were similar to
14 those obtained for the L-asparaginase conjugates of Example 4. The
enzyme-albumin-antibody conjugates were found to be significantly
16 more cytotoxic to human pancreatic tumor cells grown in tissue culture
17 than L-asparaginase alone, L-asparaginase linked to albumin,
18 L-asparaginase linked to albumin and a non-specific antibody or non-
19 specific monoclonal antibody, or the monoclonal antibody against
pancreatic tumor cell itself (see Table VIII).
21 This is an important finding s;nce there is no known
22 effect;ve treatment for cancer of the pancreas and yet Yunis and co-workers
23 in Int. J. Can.g 19, ~1977)j pg. 128 - 135, have demonstrated that human
24 pancreatic tumor cells in tissue culture are asparaginase sensitive.
- 136~0
l TABLE ~III
-
2 CYTOTOXICITY OF L-ASPARA~INASE-CONJUGATES TO
3 HUMAN PANCREATIC TUMO~ CELLS G~OWN IN
4 T~SSUE CULTURE
~ ' Dose Required to Inhibit
6 Enzyme PreParation Growth for Three Days
7 L-asparaginase 0.08 Units*
8 L-asparaginase-albumin 0.02 Units*
9 L-asparaginase-albumin~antibody (contrcl) 0.03 Units*
L-asparaginase-albumin-antibody (expt.) 0.005 Units*
11 (expt. = monoclonal ant;body against
12 pancrea~ic (human) tumor cells)
`'
13 A number (5 x 105) of human pancreatic tumor cells were seeded in a
14 t1ssue culture flask at time zero and the dose required to completely
inhibit tumor cell growth over a period of 3 days for the different
^ 16 enzyme prepara~ions was determined. The monoclonal antihody alone was
`~ 17 ineffective at inhibiting tumor cell growth at the concentration used
18 in the conjugate.
'''';
19 * Units are defined in the Mashburn paper cited in Example 1.
- 29 -
1 3 ~ ~ ~ S V
1 Example 8
2 Superox~de Dismutase-Album;n-Hyaluronic Acid Ant;body Conjugates
3 In a manner analogous to Example 13 glutaraldehyde was
4 used to link the enzyme superoxide dismutase (from hog liver) to
albumin to antibcdies against hyaluronic acid (rabbit antisera). Using
6 a molar ratio of 1:10~1 of enzyme to albumin to antibody a conjugate
7 having a molecular weight of 1.1 x 106 was formed.
8 In test procedures analogous to those of Example 1, the
9 conjugate was found to resist bioinact;vation and to be non-immunogenic.
In addition3 the conjugate showed a high affinity for the substrate
11 hyaluronic acid.
12 The benefit of cross-linking superoxide dismutase to albumin13 has been shown previously, see Wong, Cleland and Poznansky, Agents and
14 Actions, (1980), 10, pg. 231 - 244. The present conjugates with the
antibody against hyaluronic acid can be targeted against sites con-
16 taining hyaluronic acid to reduce inflammation associated with rheumatoid17 arthrit;s.
18 While the present invention has been disclosed in connection19 with the preferred embodiment thereof, it should be understood that
Z0 there may be other embodiments which fall within the spirit and scope
21 of the invention as defined by the following claims.
- 30 -
