CA2176566A1

CA2176566A1 - Method of binding recognizing substances to liposomes

Info

Publication number: CA2176566A1
Application number: CA002176566A
Authority: CA
Inventors: Rimona Margalit
Original assignee: Individual
Current assignee: Baxter International Inc
Priority date: 1994-10-03
Filing date: 1995-10-03
Publication date: 1996-04-11
Also published as: EP0731691A1; WO1996010394A1; AU3894895A; US5603872A; US5846561A; AU699887B2; JPH09511252A

Abstract

Recognizing substances, including epidermal growth factor, gelatin, collagen and hyaluronic acid, have been covalently bound to liposomal surfaces and utilized to attach liposomes onto a cellular or an extracellular matrix (ECM) target site. These "bioadhesive"
liposomes offer several advantages including the mutual protection of both the drug and biological environment; an increase in drug bioavailability and retention at the target site; and improved adherence or adhesion to the designated target site

Description

2~ 76566 PCT~S95113176 . ,....~ I .
Descri~tion Method of Binding - - toT-;r R~iC~ a to Related ~ A~
This is a C~n~;nl~Ation-In-PArt of: (a) Application 8erial No. 08/058,442 which i8 a ~'~mt;nl~n~ n of Application Serial No. 655,576 filed February 14, 1991; (b) Applic~ltion Serial No. 07/951,351 which i~ a File Wrapper C~nt;nl~tion of Application SeriAl No. 655,879 filed February 14, 1991; (c) Application Serial No. 07/960,196 which i5 a File Wrapper r~n~;nll~tion of Application Serial No. 07/655, 878 filed F~~ y 14, 1991; and, (d) Application Serial No. 07/978, 985 which i8 a C~nt;n~ntion-In-Part of Application Serial No. 655,013 filed February 14, 1991.
Terhn~ ~l Field The present invention relates to the preparation of microscopic drug delivery systems ~MDDS) u~ ;
drug-encapsulating h;~A~lh~ive l;F- -.
Bacl~rou~d Art Currently, the topical and local admini~tration of a drug can be in its free form, diaaolved or diaperaed in a suitable diluent, or in a vehicle such a~ a cream, gel or o;n~ t. By definition, "topicala administration ;n~ n~
non-invasive drug administration, while "local" includes invaaive, i.e., through a localized injection or infuaion.
r ~ of ~h-~. ~reut; ~ or deaignated targets for topical or local drug administration include burns; wounds; bone injuries; ocular, skin, intranaaal and buccal in$ections;
ocular chronic situationa auch aa gl-v~ ; intraperitoneal infectiona, tumors and metaat~aia; and topically and locally ~ a A tumora . Several ~l; ff~ l tiea exist with WO96/10394 ~ f~ F~111J.,,3~1~116 0 either the topical or local administration of a drug in its free form. For example, short retention of the drug at the designated site of administratioA reduces the efficacy of the treatment and requires freguent dosing. E--~oD~e of the free form drug to the biolo~;r~l environment in the topical or local region can result in drug degradation, tran8formatlon into inactive entitiea and n~n~; nrriminating nd uncontrollable diatribution of the drug. Such degr~dation and uncontrollable distribution of the drug can result in toxicity issues, lln~l~sirnh~e side effects and 1088 of effica¢y.
Microscopic drug delivery systems (MDDS) have been developed for i _ ~ved drug administration relative to administration of drugs in their free form. Drug-loaded MDDS can perform as sustained or controlled release drug depots. By providlng a mutual protection of the drug and the htolcr;;c~l environment, MDDS reduces drug degradation or inactivation. As a system for controlled release of a drug, MDDS i ~ ve8 drug efficacy and allows reduction in the L-~ y of dosing. Since the rh~ ok;netics of free drug release from depots of MDDS are different than from directly-administered drug, MDDS provides an additional measure to reduce toxicity and undesirahle side ef f ects .
MDDS is divided into two basic classes: particulate systems, such as cells, mi~ h_-e8, viral envelopes and liposomes; or nonpar~; rlll nt~ systems which are macromolecules such as proteins or synthetic polymers.
T~;rOI ~ have been studied as drug carriers and offer a range of advantages relative to other MDDS systems.
r _ ~21 of natl rally-occurring materials which are bic_ _ ~tible and bio~r~rnf~hle, 1 ;~ - are used to biolo~ lly active n!aterials for a variety of ~JI.L~03e3. Having a variety of layers, si2es, surface Wo96110394 ~ 21 76566 ~ ,J1I~1/6 ~ 3 charges and compositions, numerous LJLoc~l~LL~8 for 1 ;roE
prepar~tion and for drug ^nr~r8~1 AtiOn within them have been developed, some of which have been sc~led up to industrial levels.
T.;p~ can be ~ n^d to act as sustained release drug depots and, in certain applications, aid drug access across cell me~branes. Their ability to protect encapsulated drugs and various other characteristics make 1 ;, -_ - a popular choice in developing NDDS, with respect to the pr~vious practices of free drug adT~inistration.
Despite the advantages offered, u~; 1; ~t;~n of drug-~n~-nrsul s~ting l ;F-- ~ does pose some ~l; ff~ l ties.
For example, 1 ;F-- ~~ as MDDS have limited targeting ahilities, limited retention and stahility in circulation, potential toxicity upon chronic adT~inistration and inability to extr:~vasate. Binding of ~:Ly L y~sin to 1 ;~ o= - has been studied ~8 a model for binding substances to 1 ;ro~ ~ surfaces. ~ o~n; ~;n~ substances, ;n. l1-~ nt;hO~ , glycoproteins and lectins have been bound to 1 ;F~- 1 surfaces in an attempt to confer target specificity to the 1 ;ro~ a. rr~nr~ntrating on gystemic applications and in vivo studies, these previous efforts discuss methods of binding recosn; ~;n~ substances with 1;L~= ~ and the effectiveness of ~uch ';f;ed liposomes.
Although the bonding of these r~o~n;~;n~ su~L.L~ ce~ to liposomes oc~u~ ~d, the resulting '; f; ed 1;A ~~ 1r did not perform aa hoped, particularly during in vivo studies.
Other difficulties are presented when 11t;l;~;n~ these reco3n; ~;ns substances. For example, ~nt;ho~ can be p~tient spe~f;c and, therefore, add cost to the drug therapy .
In addition to the problems outl; n^d above, the prior art has failed to A;~lose an Gff;~i~nt and effective method of making bio~h^~ive 1 ;L ~~ ~ useful for scaling-WO96/10394 , r ~ ~ ~ PCrNS9~113176 up to ~n industrial level. In "Prepara.tion of EGF Labeled T.; ro~ - - and Their Uptake by Hepatocytes, " Ishii et al ., R;ocho~n;cn1 and Biophysical Research C ;r~tions, Vol.
160, pp. 732-36, 1989 (nIshii et al.n), the authors 5 describe uptake of 3GF-bearing 1 iror - by liver cells in suspension. In the preparation of their 1 ;, -- --, Ishii et ~1., disclose a ~L~c~-lULe involving at least four different steps, each individually involving at least two more sub-steps. These ~teps include further purification 10 by column chromatography, which can be ~;ff;~lt to scale-up to an industrial level. Fur~h~ _e, not only is this process _ ~ ~, but each additional step contributes to a 1088 of material or poRsihle inactivation of the EGF. It h~s been reported that the biological ~ctivity of EGF is 15 d~ t upon the conservation of the native ~nfon~~tion of EGF, to which the ~l;alllf;~lP bonds are critical. In binding EGF to 1 ;F:- ~~, Ishii et al. exploited the existence of the ~;alllf;t3P bonds. Sp~c;fir~11y, EGF was bound to the l;F-- 1 surf~ce by the ~l;Rl~1f;~lP bridge 20 linkage using a heterobifunctional CrO881 ;nk;n~
reagent,N-hydroxysuccinimidyl-3- (2-pyrid-yldithio) propionate. The complex chemistry of this process results in LY~L~dU~ L8 whose effect on drug delivery and toxicity Are unknown, possibly resulting in inactivation of the EGF.
25 Further, the comp ~ ex proce_s described by Ishii et al .
would be virtually; _ -~-;hlP to ~ 1; ah in an aseptic environment, as required in a 1 ;F- ~ process.
Prior to the development of the present invention, a need existed for a 1 ;E-- having targeting and retention 30 abilities to a target organ or tissue. Sper;f;r~lly, there remains a need for the dev~1 ~ t of a l'bioR~lh~Rive"
1;,-e comprising ~ - having an effective rq~o3n;~;n~ substance attached thereto. Prior to the present invention, a need also existed for an ef_icient ~ Wo 96/10394 ~ 2 1 7 6 5 6 6 PcTlus9S/13176 method for binding reco~n;~in~ substances to a l;ro6 thereby producing a bi~ih- lve liposome, using fewer steps than those de~cribed in the prior art.
5 r~inclo~ re of IlLver-tion According to the pre~ellt invention, nff;~ nt meth~ logiea have been developed to effectively bind various recoJn;~;n~ ~L-,L~-cel3. These include, and are not limited to, coll~^n, gelatin, hyaluronic acid and 10 ~ri~ l growth factor to l ;, - l surfaces thereby forming lxio~ih~ive 1 ;, o~ ~. Further, the method~ of the present invention employ fewer steps than known in the art, thus making auch methods more efficient and cost-effective on a commercial sclle. Further, the process described in 15 the present invention ~voids the risk of inactivating the reco~r~ ~;ng substance during creation of the bio~-lh~ive l iro6 - . The bio~rih~ive l ;, - - of the present invention have spec; f; ~ ity for and the ability to a&ere to the designated target area for sustained release of the 20 l;F-- -'8 therapeutic contents.
The; _ _ved process of the p~ssent invention ;n~31~d~
adding a rec-~Jn; ~; n~ subs tance to a l; E ~ _ -; adding a crossl; n~; n~ reagent to the mixture of the l; E i~ ~ ~ and rero~Jni~;n~ ~uL,i~.ce; and, J~13~ n~ the mixture to 25 incubate for a period of time to form the bio~ ihe~ive l;F-- . By modifying regular l;E- --~ through covalent bonding of certain reco~n; ~in~ substance to the 1 ;ros~ l ~urface, the reco~Jn;~;n~ sub~tancea can be u~ ' a3 an adhesive or glue tc attach and retain the ~;f;~d l;F-~
30 onto ~ target area despite c~ r nd fluid dynamics.These ~b;~ h~n;ven l;E-- ~ offer potential advantages as a NDDS for the administration of drugs.

W0 96/10394 ` ~ 7 6 S ~ ~ PCr/US9S/13176 Brief DegcriDtion of r~rA~rin~o FIG. l shows the binding of bioA~h~o;ve 1 ;F-- ~
(EGF- '; fie~; open double triangle) and regular 1 ;F ~ ~
(asterisk) of the LUVET type to A431 cells in culture (in 5 monolayers), as d~ t upon 1;~ -_ ~ cnnt~nl ration.
Bound 1 ;F-- ~~, denoted as B, are in units of ng EGF per 106 cells. Free ligand con~-~n~ration, denoted as L, are in units of ng EGF per lO' cells for blo~h~oive liposome (first row of L values) and in unita of umoles lipid per lO' 10 cell8 for the regular ~ ;po_ ~ (second row of L values) .
FIG. 2 shows a time course of the binding of bio~h~o;Ve ~;F - (collagen- '~ f; ~C~.) of the MLV type to A431 cells in culture (in monolayers) . Coll ~n ia tritium-labeled. The fraction of liposomes relative to the 15 amount present in the initial reaction mixture at zero-time which is cell-~asociated is det~rm;n~d over time.
FIG. 3 shows the binding of bio~h~o,ive 1 ;F- ~~
(coll~ n- -';fio~7) and regular 1;,-_ ~ of the MLV type to A431 cells in culture (in monolayers). Collagen is 20 tritium-labeled ('-H) and 1;, _ ~~ are 1~-C labeled. Bound liposomes, denoted as B, are in units of 3-H DPM per lOscella (left scale) and in units of 1~-C DPM per lOs cells (right scale) . Free ligand c~n~ ~ntration, denoted as L, ~re in unit~ of 3-H or ~-C DPM per lOs cells. Bi~ hPoive 25 liposome with ~ g~n labeled ia depicted with open double triangles; bio~3h~o-ive 1 ;F- with the l;, _ - labeled is depicted with crosses; and, regular 1 ;ro~ - is depicted with asterisks.
FIG. 4 shows a achematic drawing of the experimental 30 aetup for atudying the effects of fluid dynamics on cultures of adherent cells having bio~lh~o;ve 1 ;roF --attached thereto.

wo 96110394 ~ $ 2 1 7 6 5 6 6 1 l~u~ é~
Best llode for CarrYen~ Out the Invention While the invention is susceptible Of A ~ -~' t in many different forms, there i8 shown in the drawings and will herein be described in detail pre$erred ' -'; ~ of 5 the invention with the u d~. D L~ding that the present disclosure is to be c~n~i~^red as an e _l;f;r~tion of the pr;nr~;rl~R of the invention and is not ~nt~n~d to limit the broad aspect of the invention to the : ' - '; ~ R
illustrated .
According to the present invention, various r~CO~n; 7!;n5 aubstances have been covalently bound to 1;, o_ 1 surfaces through the cro~6l ;nlr;n~ of ~emine residues. T-;ror -F, in particular, mult;l~ r vesicles (MI.V) or lln;l~ r vegicles such as micro lR;f~ed 15 l; F - 7 (~EL) and large lln; l ~ r 1; ~ LWET) ~
each c~nt~;n;ng pho8phatidyleth~n~ m;n-- (PE), have been p c~Lcd by es~hl;~hsd plocclu aR. The ;n~ n of PE in the 1 ;F- provides an active functional residue, a primary ~emine, on the 1 ;FC- 1 surface for crogr~l ;nk;n~
2 0 ~u ~o~
R~co3n;7;n~ DubDLcu~ccs have been successfully linked with PE-l;r -. R~c-o~n;~;ng substances useful in the present invention include collagen, gelatin, hyaluronic acid (HA) and ~ri~l~r---l growth factor (EGF). Using 25 commercially availahle gelatin and collagen, these protein-re~o~n; ~;n~ substances were linked to the 1;,3= --through amine residues. Hyaluronic acid is a natural polymer with alternating units of N-acetyl glucoseamine and glucoronic acid. Using a crosRl ;nk;n~T reagent, hyaluronic 30 acid offers carboxylic acid residues as functional groups for covalent hinding. The N-acetyl~ osAA-n;n-~ cont~;nR
hydroYyl units of the type -CH,-OH which can be ~Y;~;~ ' to aldehydes, thereby offering an additional method of crosRl ;nk;n~ hyaluronic acid to the 1 ;~ 0= 1 surface in WO 96/10394 ' ~ f 7 6 ~ 6 6 PCr/U595/13176 --the absence of a crosRl ;nl~;n~ reagent. EGF i8 a polypeptide. Although urogastrone and EGF are r~o~n; ~ed as biolo~ l esiuivalent~, both purified uroga3trone or EGF
mouse were used as r~co~n~ ~nl ~ L~.ces . When used in 5 the specification and claims, the term "EGF~ means either urogastrone or epid~~l growth factor regardless of the source .
EGF stimulates cell growth and proliferation through inter~ction with an EGF roc~toL. EGF receptora are 10 distributed on the c~ll surface of various organs and are present in burns, wounds, and other designated targets of MDDS such as ocu~ Ar, dermal and tumors . Accordingly, EGF- ~;f;~rl l;F-- ~ pot~n~lly offer ~ff;~i~n~y as drug carriers to target sites, i.e., organs or tissuea, 15 expressing the EGF receptors.
~ Co~n;~;ng ~ Lc.~ccs are bound covalently to discrete sites on the 1 ;E-- ~ surf~ces. The number and surface density of these sites will be dictated by the 1;, 2= formulation a~d the 1;}-_ ~ type- The 1;F--20 surfaces may also have sites for noncovalent association.Covalent binding is ~ont~l as noncovalent binding might result in ~; R80~ tion of re~o~n; ~;n~ substances from the 1 ;F-= - at the site of administration since the 1 ;ro= R
and the bio~lh~ive counterparts of the target site (the a5 bioP~h~sive matter) compete for the re~o~n;~;ng substances.
Such ~; Rso~i~tion would reverse the administered modified l;, - = - - into re~ular, non- '; f; ~C~ l; L - - - ~ thereby defeating the purpose of administration of the modified ~;ro-_ --.
To form covalent conjugates of reco~n;~;n~ substances and 1 ;~ -, croR~l ;nk;n~ reagents have been studied for ef f ectiveness and b; r _ ~ tibility . CroRsl; nk; n~ reagents include glutAraldehyde (GAD), bifunctional oxir~ne (OXR), ethylene glycol diglycidyl ether (EGDE), and a water -Wo 96110394 ~' ! 7~` ~ ~ 2 1 7 6 5 6 6 Pcr/usg5/l3176 , ~

soluble c~rho~l;;m;de, preferably 1-ethyl-3-(3-dimethyl~m;n~r opyl) ~-9rho~1;;m;~ (EDC) . Through the complex chemistry of crss~l ~nk;n~, link~ge of the amine residues of the re~o~n; ~;n~ ~ I-~ce and 1 ;F-- ~ is 5 e8t ~hl; nh~d .
An; _ L~-t feature o~ the present invention is the binding between the newly-created bi~~'h~nive 1;L~_ ~ and potential biolog; C91 target sites . Biological target sites ~re divided into two classes. The first class ~ - _-n8--n 0 , _ --Itn of the extrs~ r matrix (ECM). The ~CM can be v;n~sl;-ed as a network made of a variety o~ _ ~n~
which is not cast loose in a living system, but is connected at some of its points to cells. ECM is found -n~rn-nth cells, above cells, ;n between layers of cells, 15 and in between cells in a layer. The second class of potential targets are ~c: ' -''-d receptors.
A complete Arco~nt~n~ of bir~ding entities has been dete~rm;n~l by the previou~ly known multi-term T.r ~r Isotherm equation, as applied for the quantitative 20 description of the relat~r~nnh;r between the free and t vari~bles:
n Bmax~ [L]
B = ~ __________ (1) 2S i~1 Kdi + [L]
where n is the nuT~ber of dif f erent types of binding entities that a cel l~ r or ~n ECM system target site has for a spe~;f~ reCo~n;~n~ substance; [L] is the 30 ~nr~-ntration of free ligand, which can be re~o~n;~;
substance, free 1 ;F-- ~~ or h,iO?~3h~n;ve 1 ;L ~~ ~~; B is the total quantity of bound re~o~n;~;n r DuLDI_ Ce per given number of cells or guantity of ECM, at a given [L]; and, Bm xi and Kdi are the total number of siteD of a given 35 entity ~nd the c~ ;n~ equilibrium ~3; n130~ iAtion W096li0394 ;~\ ~ 76~ PCIIUS95113176 ~

con~tant. B and B,,." are n~l~-7; 7ed for the same nu~ber of cells or guantity of ECN.
For caE~es in which r6ce~tor~ and non-r6c~toL cell membrane r _ ~ participate in the re~ogn; 7;n~
5 aubatance binding and in which the ~3;nao~ntion conatant of the non-specific binding ia a~ffi~iPntly large with respect to the free ligand ~n~Pn~ration, e9uation 1 can take the _orm:
n-l Bmaxl [L]
B= ~ _____~ + R", tL] (2) di + [L]
where the laat term, R", ~L], i8 the contribution of the non-aper;f;~ binding to B and R~, i9 the ratio of Bmax to Rd c.,LL.=~ l;n~ to the non-specific binding.
"Beat-fit" valuea for parametera n, Bmaxl and ICdl are obtained by computer-aided data analyaia, according to equ~tion~ (1) and/or (2) above, applying PL~C-1UL~S of n~lnl ~nP~r regreasion analyaia.
The interaction of the h;n~lh~Aive 1~L- r with potential biolo~ l targeta haa been eR~hl; ~h~d through the uae of cultures of A431 cella, in monolayers, as a h~ olo~ l model . This well-est~hl; ~hPd cell line, originating from human Pri-l id carcinoma, ia enriched with EGF receptora, and ~8 a nolayer, alao providea ECM.
A431 cells have been repeatedly used for atudy of the interaction of free EGF and ita receptor. A431 cella have been ahown to have three claaaes of EGF receptors, differing in their a__initiea ~nd popul~tiona. The firRt of these claaaea ia the ultra-high affinity sites with an equilibrium ~ o~ i ~ tion conatant of 0 . 07 nM and a population of 150-4000 aites per cell. The next clasa i8 the high af_inity aitea with an equilibrium ~; ~ao~; ~tion constant of 0 . 7 nM and a popul ~ n of 1. 5 x 105 sites per cell . The _inal class ia the low af f inity aitea with an Wo 96110394 `~ ~ ~ 11 2 1 7 6 5 6 ti /vv~J113116 eguilibrium ~3; n~o~ ;~tion constant of 5 .9 nM and a population of 2 x 106 sites per cell. Because o_ their affinity for EGF, A431 is particularly useful for demon~3trating the targeting ability of EGF ';f;ed 1;L-- __ However, targeting of b;o~7h~1ve 1;~- -F
having other types of re~o~n;~;ns substances has also been demonstrated with this cell line, as shown in the following r The " level of covalent binding" as reported in the r len below is defined ~ the quantity of bios~3h~ive ligand, such as collagen, gelatin, hyaluronic acid or EGF
bound to a given quantity of lipid in the f inal product since the most accurate quantitative measure of 1;~ a_ --is in terms of lipid quantities. For a given lipid quantity, different 1 ;~ -~ types will yield different quantities of 1; L :- . Therefore, similar initial ratioa of EGF to lipid for different 11,-_ - types should not be expected to yield the same level of binding. Another factor which would yield different results for different 1 ;ror -- even under the same initial EGF to lipid ratios, is the differences in particle size, therefore in curvature, nu~ber and A"C~ 3;h; 1 ~ ty of PE sites on the surface of the 1;}- . Therefore, comparisons among 1 Iroc ~ types should be avoided.
The effects of. the incresse in the EGF/lipid ratios in the presence of a cros~l Ink;n~ reagent are shown below in Tables 1 and 2. Generally, an increase in the level of binding occurs with the increase in initial EGF/lipid ratios regardless of which cro~l ;nk;ng reagent is used.
At the lower end of the EGF/lipid ratios, the level of covalent binding increases ~;~rn;f;~ sn~ly. Beyond initial c~n~nt~ation ratios of 25 ng EGF/uMoles lipid, the increase of binding is less sign~f;~snt. Noncovalently WO 96/10394 - - ; " 5 ~ `- 2 ~ 7 6 ~ 3 bound product i8 removed as excess unreacted material and does not appear in the reported results.
r 1~
All of the f-~lll 'n~ Examples using ~ollr__~, gelatin md 3GF J~D the re~o~n; ~;n~r substances, were prepnred according to the method described in Example One. Slight A;f;~at;~nn, aR described in the r ~ , were required for the r l~R using hyaluronic acid as the r~co~n; ~;n~
s~ La~ce. ~rhe re~o~n; ~;n~ DULDL~UC~8 are assayed by traces of radioactive or flu~,ldaa_ lt labels.
Alternatively, the lipids are ~ssayed by colorimetric methods . DeterminAtion of the protein reeo~n; ~;n~
su~DLculc~s can be done by the Lowry pLOC~ lu~e, while free 15 BA and 1 iF - ~ bound ~A can be det~m;n~ by the Alcian Blue method .
Examr,le One EGP is added to a PE-l ;L __ sample and the mixture 20 is buffered by a phosphate buffer saline solut;~n ~PBS~ to plI of 7.2. For drug-c~n~";n~n~ liposomes, drug encapsulaticn was pe r ' in a ll;n~ so3~ n also of PBS. C~n~nt ration ratios of EGF to lipid are shown in Table 1. Aliquots from a 25% solu~ n of the cro~ ;nk;n~
25 reagent glutaraldehyde (GAD) are added at a ratio of 10ul per 1 ml EGF/PE-l ;ror mixture. Tn~ hA~ n for a desired period (24-72 hours) is let~d at either room temperature without stirring or at 370C with stirring.
Dep~n~;n~ upon the l;Fo- ~ used, excess unreacted material 30 waR removed, preferably through high speed centrifugation for one hour at 4C and 27000xg or ultrahigh centrifugation for one to two hours, at 4C and 250000xg foll~ ~ by several repeat~d w-Rh;n~R with EGF-free PBS. Column WO96/10394 t, ~ 2 ~ 7 6 5 6 6 P~l/u- ~l3l~6 chromatography or dialysis against PBS may also be used in pl~ce of the centrifugation.

EGF-LIPOSOME rR~ T TllRTNG BY GAD
ngEGF/uMOLE LIPID (a) LIPOSONE TYPE INITIAI FINAL
Ml~V O . 080 0 . 009 MI.V 0.309 0.006 MI,V O . 347 0 . 016 MEI O . 071 0 . 004 MEL 0 .106 0 . 009 ~EL 0 .141 0 . 025 LUVET 0 . 016 0 . 003 (a) EGF assayed by a radioactive tracer.
r l e Two EGF is cros~l ink--~i with PE-l ;E-- samples following the same ~L-~c~ LLe as in Example 1. rl~nront~ation ratios of labeled EGF to lipid are shown in Table 2.

EGF-LIPOSoME rR~Ssr, BY GAD
ng EGF/uMOLES LIPID (a) MI,V 0.26 0.07 Mr-V 0.78 0.16 Ml~V 1. 60 0 .21 MI,V 6.00 0.31 MI,V 24.70 0.35 (a) EGF ass~yed by a fl.loresc.:~lt tracer .
r 1~ Three Reaction mixtures of EGF and PE-l ;L "- - were ed as above and buffered by PBS to pH 7 or by 0.5N
carbonate buffer to p~ 9. rr~nt~n~ations ratios of EGF to 45 lipid are shown in Table 3 . The crossl ;n~;n~ reagent EGDE

4 " ~ I 6 5 ~i ~ PCr/US95/13176 was ~dded in 0 . 2 - 1. 0 ml volumes to buf f ered reaction mixture~ of 2 . 5 - 3 . 0 ml volumes . Incubation periods were completed for 10-24 hours at 370C with stirring. ~r~n~l;n~
upon the l ;E ~- ~ u~ed, excess unreacted material was 5 removed through high speed centrifugationa and w-Rh~n~ or dialysis against PBS .

EGF-LIPOSONE t~T~r~9.~T.TNRTNG BY EGDE
/EGF/uMOLE LIPID (a) IIIPOS07.~E TYPE INITT~T FINAL EJII m~ EGDE
D~V (b) 0.45 0.0078 9 500 7.~V 3.72 0.90 9 500 MEL 0.10 0.012 9 500 MEL 0.10 0.0098 9 1000 7.~EL (a) 0.12 0.0022 7 200 7.~EL 1.78 0.47 9 500 (a) EGF a~sayed by a radioactive tracer.
(b) Initial ratios were increased by decreasing lipid n-~Pn1~natiOn.

From these results, the preferred pH of 9 and ~lauantity of cros~l ;nk;n~ reagent of 500 mg has been det~;nP~.
l e Four - -Gelatin wa~ cro~l ;nkPd to PE-l~E- - following the ~ame ~ ,c~ as in Example One.

~ WO96110394 ~ 2~ 76566 r-~/u~ ~lJl/6 TAB~E 4 GEI.ATIN-LIPOSONE rRnSST~TNIrT-, BY GAD
T.; F - _ - ugGelatin/uMole Lipid Incubation TlrPe Initial Final Period (a) MEL 21 0 . 02 Short MEL 63 0 . 24 Short MEL 127 0.26 Short MEL 21 15 Long MEL 23 14 Long MEL 2 5 18 Long MEL 6 3 4 3 Long MEL 187 208 Long MI.V 18 0 . 2 4 Long MLY 66 0 . 67 Long MI.V 281 2.6 Long MBV 556 6.4 Long MIV 1140 13 Long MIV 2350 13 Long MI.V 3440 24 Long MIV 5830 26 Long (a) Incubation Periods: "Short" i8 5 minutes; "Long" is 24 - 4 8 hours .
r 1 e Five Collagen i8 cro~38l ;n~ cl to PE-NI-V samples with GAD
~ollowing the same ~oce.l-l e as in Exallple 1 except incubation was at 4C, at "Long" incubation periods .

cnT~T~ T~N-LIposoME ~'Rn9~T-TNRTNg BY GAD
T.~ F ~ - ugCollagen/uMole Lipid ~ Initial Final MI.V 1 . 64 0 . 90 40 MI-V 2.06 1.18 MI.V 5.01 2.20 MIV 8 . 96 5 . 07 MIV 9.83 6.78 MIV 9.86 6.02 45 MI-V 10.68 8.20 MIV 18.79 11.55 MI.V 2 0 . 0 0 14 . 14 W096/10394 "~ . 2~17b~ 1i6 ~

r le Six Aqueous 8Olut; ~n~ of HA was pre-activated by incubation with water-aoluble cs~rho~;;m;r7~ EDC. The 5 - _ ts were mixed to yield a preparation system of HA
and EDC each ~t final c~n~ntrations of 1.7 mg/ml. The plI
of the preparation system was adjusted to 3 by titration with lN ~ICl. The preparation system was incubated for a variety of time periods at 370C with stirring. Table 6 lO shows an example of variation in the pre-incubation time period for reacting }IA with EDC. A pre-incubation period of 3 hours is preferred to activate the carboxylic reaidues of ~A.

EFFECTS OF PRE - TNC TT~TTON
E~a-LIPOSOME BINDING (a) PRE-INCIJBATION PERIOD mg ~IABound/mmole (~ours) Li~id O O
3 22.8 $ 0.9 24 20.9 $ 2.8 (a) T~F~- are LUVET or MBV.
Incubation of the complete reaction mixture was for 24 hours, at 370C, p~
3 with the addition of borate buffer.
After the pre-in~h~ n period, PE-l;F~_ - samples were added and fol 1 ` by the addition of a O . lM borate buffer ~t p~ 8.5. The ~IA/PE-l~ mixture was incubated at 370C in a shaker bath for 24 hour~. Removal of excess unbound R~ and reagents was by ul~r~r~ntrifugation and washing~. Initial and final concentrations of HA/lipid are reported in Table 3.

w0 96ll0394 ~ ' 2 1 7 6 5 6 6 ~ b r 1 e 6even - - -Various parametera affect the s~c~sfll1 binding of HA
to PE-1;F-- -- when using EDC as the cr~asl~nk;n~ reagent.
These parameters include a pre-incubation ~L~C6~ L~, pH of - 5 the reaction mixture, use o~ buffer solut;~n in the ;nl llha~;rn gystem and the contact ~rea between 1 ;F-- ~
~nd HA. Tables 7 and 8 provide data on v7.riations of these p~ rameters .
TAB~E 7 EFFECTg OF pH, BUFFER
PRE- INC~BATION AND CONTACT
AREA ON COVALENT BONDING
OF HA AND l .L~C ~ ~ (' Borate HA-T ;F-- mg HA Bound/
H Buf f er Contact Area mmole LiT~id 4 . 5 (b) __ _ Narrow 3 .1 + 0 . 6 2 0 4 . 5 - - - Narrow 5 . 2 + O . 5 4 . 5 - - - Wide 7 . 6 + 3 . 9 4 . 5 Added Wide 19 . O + O . 9 3 . O Added Wide 2 6 . 5 + O . 9 (a) Using MLV and EDC, three hours of pre-incubation (see exception below), 24 hours incubation of complete reaction mixture, both at 370C.
(b) No pre-incubation, pH listed is for the incubation of 3 0 the complete reaction mixture .
Ex~le EiGht A reaction mixture of HA, dimethyl 8111fr~Y;~3~ (DMgo) and acetic anhydride were stirred at room temperature for 24 hours. At the end of this period, the mixture was transferred to a dialysis sac and dialyzed ag~inst water over 48 hours. Activated HA was completely rc~ eL~_d from the sac as ~ts~m;n~ by the Alcian Blue method. Activated HA was incubated with PE-l;F _ -- in 0.5M 1~9 h~nste buffer at a pH of 9 for 24 hours in a shaker bath at 370C. Adding sodium borohydride as a reducing agent, portions of the WO96/10394 ,~ ; 21 ~6~ r~ 6 o activated 3A/PE-l;F-_ ~ mixture were incubated for an additional two hours. Removal of exces~ unbound HA and reagent~ was by centrifug~tion and ~~~h;n~. Cr~nr~nt --~tion rAtioa o~ ~ctivated-HA to lipid ~re shown in Table 8.

COVALENT 3INDING OF ~a TO Ll~C_ rRng~T T~T~TI-HA ~ ACTIVATED-~A (a) mg HA/mmoles Lipid MethQ~oloq~r Initial E~;L
Crossl; nk~ ng With EDC 1000 27 3A Activation with DSMO/acetic anhydride, with reduction 974 86 3A Activation with DSMO/acetic anhydride, without reduction 974 113 (a) T.;~ ~ ~ were MLV
The covalent bonding of the re~o~n; ~;ng substance~i, 30 gelatin, collagen, hyaluronic acid and EGF to l ;ror surfaces can be achieved. Noncovalently bound product i~
removed as xces~ unre~cted material and does not appear in the reported results. Preferably, protein-r~o~n; 7;
sub~tancea such as gel~tin, col 1 g~n, and EGF are 35 cov~lently bonded to PE-l ;F- - through amine residues with the crosEl Tnk;n~ reagent GAD.
The bonding of hyaluronic acid to PE-l ~ can be completed either in the presence or absence of a croa~l ;nk;n~ reagent. In the ~,- cee..ce of a reagent, 40 preferably EDC, a pH of 3 in the pre-incubation system is pref erred . A 3 -hour approximate tiTne period i~ pref erred for pre-incubation of the hyaluronic acid and cro8131 ;nk;n~

~ W096/10394 `~ 2 PCI/US95/13176 reagent. The addition of a O.lM borate buffer at pH of 8.5 to the incubation system offers a positive contribution to the binding step. rh~ n~ the reaction mixture vessel in the binding step fron test tubes to flasks, thereby - 5 incre~sing the area of contact between 1 ;F-- ~~ and hyaluronic acid did not adversely ef ~ect the binding results. Bonding of hyaluronic acid to PE-l ~F-- F
without a crossl ink;n~ reagent is preferably conpleted by pre-activation of hyaluronic acid and an ;n~llhat;~n period of 24 houra at a reaction mixture pH of 9.
Exam~le Nine To compare the binding ability of regular 1;, _ -and bio~7h~sive 1 ~F-= -, A431 cell cultures were grown in monolayers, in flasks, applying usual ~,ce-l-- ea for this cell line. Two to three days prior to an experiment, the cells were seeded into multiwell culture plates and the experiments were done when the systems were confluent.
For F,.. ~oses of assaying the modlfied l ;F - - -, the 20 EGF-reco~; 7~n~ substance was labeled with a generally known radioactive marker. Preparation of EGF----';f;ed LWET was completed as previously ~ cu~ d. Prior to the addition o~ a reaction mixture of EGF- ';f;-~ l;p~08 --, free 1 ;ros~ ~ or free EGF, media was removed fron the A431 25 cells and the cells were washed with ~ binding buffer. The reaction mixture and cells were incubated for 1-2 hours, at room temperature. Upon dilution and withdrawal of the reaction mixture at the end of incubation, 2-3 s~lrce~sive .h In~FI with a binding buffer, of the wells were 30 completed. Lysis of cells or det~' t of cells from the wells was then foll~ ~ ~' by withdrawal and collection of the well content, denoted as the cell fraction. Assays o~ the cell ~raction were completed by label ~o~nt;n~ o~ the WO96/10394 ' ' ` ' ~ 21 76~ PCrNS95113176 fraction as _ ~ with the collnt;n~ of the immediate products created through the preparation process.
A comparison between the binding o~ $ree 1 ;roc - and EGF- ';f;P~l l;E-- ~ to the A431 cells is illustrated in 5 Figure 1. The EGr ~f~ed l;F-- - adhere to the A431 c:ells c~n~;~3Plably better th2~n free l;Fs_ ~ as no free 1 ;ros~ -- were found at cell fraction. It is 3peculated that if ~ree 1;,-_ ~ do associate with the cell3, the ~;lut~n brought by the ~ lh;n~Jn is sllff~;Pnt to cau3e 10 quantitative ~ 30~ tion.
r 1P Ten Binding studies of EGF ';f;ed l;F- ~~ to A431 cells were carried out A8 described in Example 9 and the 15 data were ,~Loc~33a~l according to e~uation (1) above. The experimental conditions were such that the contribution of non-3pe~jf;~ binding was nP~ ;hle. Indeed, the data were found to fit unambiguously with a single type of binding site for each l;F- system studied. Results for several 20 systems are listed in Table 9.

BINDING p~ 7q OF RT~I~nT~RqIVE
Ll~OS TO A431 CELLS IN CIJITURE
25 RIo~nR~IVE ~ SITES PER CELL
LIPOSOME SYSTEM (nM) (x10-5) EGF-MLV 0.60 + 0.017 0.17 + 0.03 30 EGF-MLV 5.03 + 1.9 1.07 + 0.03 EGF-LWET 2.91 + 0.003 0.18 + 0.001 EGF-MEL 0.04 + 0.007 0.042+0.0042 35 EGF-MEL 0.40 + 0.13 3.7 + o.go EGF-MEL 0.48 + 0.05 0.28 + 0.01 Each bio~rlhp~ve 1 ~po~ - system is a di~erent preparation; r~o~n;~;n~ substance in each system is EGF.

W096/10394 ` ~~ 76566 PCr/US95113176 An EGF- ';f;ed 1;,- is ~n~ P~ably larger than ~nd different from free EGF, which is expected to a~fect the binding parameters. For a given class o~ receptors, the magnitudes of the ~; ~so~i~tion constants for 5 EGF- ';fied l;p-- systems are expected to be similar to or higher than those of free EGF. For a given cla~s of receptors, the number of receptors per cell that are available for the EGF- ';f;Pd 1;~ is expected to be egual to or lower than the number of available for free 10 EGF. Based on these cr~nR; ~qP ations, the binding data of the present example f it with the r~c_,l,tol classes of ultra-high and high affinities.
Regardless of the specific cell-A~o~-; AtP~l binding entity involved, the binding data listed in Table 9 show 15 that EGF- ';f;ed l;E-- - bind to this cP~ system wlth high affinity and with a sufficient number of sites for these ';f;ed l;ro- -- to perform as the desired bio~lh~ive 1 ;F-= ~ -20 ExamT~le Eleven Binding collagen- '; fied l;r~ to A431 cell~ was c~rried out ~8Pnt~ ,~1 ly according to the procedures detailed above. The A431 cell line is not known to contain receptors for collagen. The interaction of either free 25 collagen or l;ror lly bound collagen with the A431 cell line is expected to result from association of collagen with - __ Pnts within the extra~ Pllul~ matrix. Referring to Figure 2, incubation periods up to 4 hours were completed with 3 hours being the optimal period f or 3 0 binding .
Quantitative evaluation~ of binding of collagen- ~;f;ed liposomes to A431 cells in culture are compared to regular 1 ;F ~= ~ and _ l; f; P.l in Figure 3 .
The data were processed according to eguation (1) above.

W096/10394 ~ , 2f 76~6~ F~~ tl~l/6 ~

Through double 1 ~hel ~ n~J~ 3 -H-collagen and 14 -C -cholesterol, it wAs possible to monitor the collagen and 1 ;E - ~
simultsne~ ly. The binding of the collagen-modified l; } - - - - to the cells is greater than the binding of the cc, ~ regul~r 1; ,, ~
For free and collagen-modified 1 ;rr,~ -, the binding entities are of the extrarelll~ls m~trix type of cell-associated entity. Aa in the case of EGF- '~f;~,l l;roC -~ r~89~ in Example 10, the ~ sor;~tion 10 constant for collagen- '~f;.o~l ~;ror ~ is expected to be similar to or higher than those of f ree collagen .
Likewise, the nu~ber of available sites in the extra~Pl 1--1 9r matrix available for col l 95C~ f; ed 1;, _ - is expected to be similar to or lower than free 15 collagen. The example given in Table 10 fits with these r~n~i~lerations. The data for free collagen demonstrate that binding of this bion~lh~R;ve re~r, r;7;~g ~st.z~ce to thia cellular system does occur and is a measurable F?~ 9~ which can be ~L~ctaa~l to yield quantitative and 20 meilningful parameters. ~ , the data in Table 10 show quite clearly that the binding of collagen-modif ied liposomes to this aoll~ls system is of ~-lff;r;~l~tly high affinity and with a large enough nu~ber of sites, for the collagen-~-';f;e~ l;E- - to perform as the desired 25 h;~c-~h~;ve 1 iror BINDING P~T AMRTRT~q OF FREE RECOGNIZING
SUBSTAN''ES AND RTr~nTTRCIVE LI~OSOME
TO A431 CELLS IN CI~TURE
8Ttl~nT~RqIvE l~d N~MBER OF SITES
~IPOSOME SYSTEM (uN) rxlO-s) FREE corT~RN 8.5 + 2.3 179 + 11 COT~T.~RN-MI,V 67.6 + 31.35 548 + 160 w0 96ll0394 , ~ ~ ~ s F~ ~ .31 ,6 The b;o~3ho~3;ve 1;L- ~ of this invention are n~d to perform as site-adherent, site-retained, sustained drug release depots. In vivo, the dynamics of the hi olo~; C91 system can prematurely det~ch a drug 5 delivery system from its target site. The dynamics in ~Lue8tion are cellular dynamic8 due to proliferation, migration and demise of cells in the area to which the delivery system adheres. This type of dynamics may be seen in wounds, burns and tumors . The second dynamic is f luid 10 dynamics due to the flow of body fluids over an area where the h;on~lh~3ive 1 ;r~F- - may adhere. Effects of fluid dynamics are expected to vary tloron~l;n~ on the bodily location and disease being treated. For example, the flow is relatively slow in ther~peutic targets, such as wounds, 15 mild in a target such as the peritoneal cavity and ~ast in a target such as the ocular area. In addition, in producing the hi on~lhosive l ;F-- ~ of the present invention, it wa~ ;, L-l~t to determine whether the addition of the reco~n;~lng substance to the ~urface of the 20 l;E-- ~ would interfere with release of the on~-rs~lnted drug. Examples Twelve and Thirteen demonstrate the usefulness of bion~ho~ive liposomes as sustained release depots, in that att~ ' t of the desired re~o~n; ~;n~
substance does not interfere with release of the 25 encapsulated drug, in this case, the antibiotic Cefazolin.
Examples Fourteen and Fifteen focus on the site-re~;nnhil;ty of h;n~ho~ive 1;}-- -~ in view of colllllnr ~nd f luid dynamics, respectively .
3 0 Example Twelve MUlt; 1 -11 n-- 1 ;F - ~ (MLV), at a c~n~n~ration of 30mg/ml were prepared as described in previous Example~
with one ';f;rntion: in order to preserve drug stability, all incubations and procedures of the Cef~l;n-c~ntn;n;n~

WO96/10394 ,~;~";r~ 2t7b;~ p~ s9s/l3l76 ~

systems were performed at 5C. ~l~fnY~l;n, at a cnn-~ntrntion of 15 mg/ml was introduced into the system through the PBS swelling 8OlUt;~n. The 1 ;ro~ were incubated with EGF and with GAD, i~ the PBS buffer, as 5 previously described. The reaction mixture was constantly stirred for 48 hours at 5C. At the end of the incubation, the 1 ;F-- - were separAted from excess unbound EGF and byproducts by high speed centrifugation for one hour at 4C
~Ind 27000xg. The 1;,-- - pellet was subjected to thrQe 10 consecutive washes which consisted of suspending the pellet in PBS ~nt~;n;n~ 15mg/ml Co~n~ol ;n and centrifuging as above. The drug was ;n~ ed in the wash buffer in order to prevent 1088 of the on~ ~r~ n ted drug in the process .
Control (i.e., regular, non-b;~ h~ive) ~lAfA~ ;n_ 15 ~n~-~rn~l a ted 1; L ~ ~ ~ F were prepared according to the same ,LOC8~ without the addition of the r~o~n; ~;
l~iLD L~lce .
The kinQtics of Cefn7~1 ;n release from these EGF-- '; f; ed bio~lh~ive liposomes as well ~8 from the control 20 1 ;E-~ -~ were studied, and the data ~Locesscd, according to the following equation:
_=fl* (l-exp ( -k,t) + f2* (1-exp ( -kzt) ) wherein "f" denotes the fraction of the total drug in the system present in the dialysate at time=t, kl and k~ are the 25 r-~te constants for diffusion of the l~n~n~-nr8l~l nted and the encapsulated drug, respectively, and f1 and f, are the initial (i.e., at zero time) distr;hut;t~n~q (in fractions) of the total drug in the system, between the l~n~n~rs~ ted and the ~n~-nr~lnted pools, respectively, f, and fz sum to 3 0 unity . The data f it the case of two drug pools, one of ~n~rgl~lated drug and a second of l-n~nrS~rs~lated drug. The results showing drug encapsulation eff;c;F~n~y and rate constant for release of the ~n~rs~ ted drug were as follows: for the control 1;L-- ~~ encapsulation F-ff;~ nr-y W096ll0394 ~ 2 1 76566 r~l/u~Jl/b was 54(+11)%, and the rate constant for release of the ~nt~rs~lated drug was 0.19(+0.004)hours~l. For the bioS~4h~qive l ;F- the ~n~ rs~ tion efficiency was 71(+7)96, and the rate constant for release of the ~n~-~rs~lAted drug was 0.21(+0.0005)hours~l.
3xa~1e Thirteen C~fA7ol;n-~n~-~rs~ ted mult;l -11~ 1;,-- -- were prepared as described above. For surface ~ ~;f;~ation into 10 bio~lh~nive 1 ;F-- -, the 1 ;E ~= - were incubated with collagen and GAD, in the PBS buffer which also ~nt~;n~d 15 mg/ml C~f:-7O1 ;n. The reaction mixture was constantly stirred for 24 hours at 5C. Two sets of 1 ;ros~ -- were prepared. In Set I, the final c~n~ ~ntrations were 25 mg/ml 15 lipid, 2 mg/ml coll ~n and 10 ul/ml glutaraldehyde . In 8et II, the l;FC= c~n~nt~ation was 150 mg/ml. At the end of the incubation, the 1 ;~ _ - were separated from the excess uDbound collagen and l.~y~c-h.-lLs by high speed centrifugation for one hour at 4C, 27000xg. For each Set, 20 the 1 ;, -~ 1 pellet was subjected to three c~nne~u~;ve ~-nh;nsn which consisted of s~np~n~;n~ the pellet in PBS
c~ntA;n;n~ 15mg/ml C~f~7Ol ;n and centrifuging as described before. For Set I, the final pellet was s~np~n~ d in PBS
buffer plus C~f~7Ol-ln~ while for Set II, the final pellet 25 was s~np~n~4 in drug-free PBS.
The kinetics of C~fS~7ol ~n release was studied as described previously. The data fit the case of two drug pools: one of encapsulated drug, and a second of ~n~nt~r~ul~ted drug. The parameters det~;n~cl for Set I
30 were: (a) the fraction of 1 ;r -- encapsulated C-~fs~70l ;n at time zero was 52 (+3)96, which in this case, was also the encapsulation efficiency: and, (b) the rate constant for release of the ~nr~rsul~ted drug was 0.033(+0.013)hours~l.
The parameters for Set II were: (a) the fraction of W096110394 ~ 2~ i76~ii6~5 PCTN~9~/13176 l;F- ~n~ s~rEI~lA'ed C~f~Ql;n at time zero was 81(+1)%;
and, (b) the rate constant for release of the ~n~ ar~ ted drug was 0 . 0093 (+0 . 0009) hours~~ .
The results of thege two r ~ indicate: (l) the 5 method of creating bi~P~h~ive l;F- ~ can be carried out at low temperatures, thereby ~Yp~n~l;n~ the range of drugs which can be encapsulated and delivered in bioA~he~ive l;rnr systems: and, (2) the surface '~f;cation of the - through the addition of a re~osni 7;n~ substance, lO does not impair the ability of the liposome to release its encapsulated drug, thereby creating a ~ustained release depot in a bio~lh~;ve 1 ;, a_ . In addition, Example Thirteen shows that the 1 ;, ~ n~-~n~ation can be used as a tool to ~--n;r~ te the rate of relea~e of the 15 ~n~S~r~~ tecl drug, going from a half-life of 0.88 days to a half-life of 3.1 days upon an increa~e in the liposome concentration. Based on these f;n~l;n~q, such antibiotic-~n~ rEIlllAt;n~ bio~h~3ive l;roL 3 are suitable for use in the topical treatment of infected wounds and burns, as well 20 al3, in other topi~lly and regionally ~c~ ;hle injury-nnd non-injury-related infections, wlth the ability to match the rate of drug relea~e to the re~uirements of the therapy, through ~election of the 1 ;E-= ~ dose range.
25 Exam~le Fourteen Bio~lh~ive 1;, :_ - having hyaluronic acid as the bi on~lh~ive ligand, were prepared as previously described.
Monolayers o_ A431 cell li~e, seeded in 24-well culture plates using known ~Loce-lu~es, serve as models for the in 30 vivo areas to which the 1 ;L __ - are d^~i~n~d to adhere and retain. The b ~ h~ive 1 ;ro~ - were incubated with the cell monolayers in complete cell growth media, at 37C, for a period of 28 hours, which is well beyond the incubation period of 2-3 hours needed to attain equilibrium .

Wo 96/10394 ~ 2 ~ 7 ~ 5 6 6 PCT/U595/13176 binding. At d~ n~ted periods, the media in several wells was aspirated and the wells were washed three times with PBS at pH 7.2, c~nl ~;n;n~ 0.296 BSA. The cells were detached ~rom the wells through trypsinization and 5 ~ubjected to the following assays: (1) determination of the nuTnber of viable cells/well, using Trypan Blue Method; and, (2) determination of the guantity of cells (or monolayer)-associated 1 ;F= ~ using rAdioactive ~o~nt;n~. The results appear in Table 11 below.

N~ION OF CEL~-A~SOrT~TRn T TPOSOMES
Time Levels of Viable Bound Periods Cells/Well (x105) T~;L-- 7 (hours) (nmoles of lipid/105 viable cells) 0 3.77(+0.52) N/D
3.4(+0.73) 12(+2) 4 2.7(+0.59) 34(+71 11 2 .2 (+0.27) 37 (+8) 23 2 . 9 (+0 . 86) N/D
28 4 .4 (+0 . 93) 34 (+9) The data above in~ te that once the l;p-- binding reaches eoluilibrium (within 2-3 hours), the level of bound 35 l ;, o_ - remains constant despite the c~ events taking place, i.e., demise, proliferation, migration of the cells. The dat~ provide direct experimental support of the ability of the bio~7h~;ve 1 ;ro~ - to be retained at a site to which they will be administered, despite c~
40 dynamics, a significant attribute o~ the bio~tlh~;ve l ;E- ~-WO 96/10394 = ~ ' 2 ~ ~ 6 ~ ~ PCT/US95113176--ExamT2le Fif teen R; o~Ahc~give 1 ;p~ having hyaluronic acid as thereco~n;~;n~ substance and the A431 cell culture they bound to, were y~ ar~d accordlng to thQ pLvc~duL~ described in 5 Example Fourteen.
To test the effects of fluid dynamics on site retention re~uired ~m experimental setup which would make it possible to flow fluid at a controlled r~te and for designated perioda over a monolayer of cells to which the 10 l~F:- ~ are bound. The ~"~ign~l setup, as shown in Figure 4, ;n~ d a peristaltic pump connected to a culture flask ~nt~;n;n~ a nolayer of cells, using aseptic conditions. Selection of the rate of fluid to which the ~rea of interest is expo5ed, will depend upon the 15 spe~ ;f;c l;roc 1 system tested and on the future th_ ~_ I ic objective. The object of this Example was to determine whether fluid dynamics put site-ret~nt;on in a given cell-l ~ ~ ~ system at risk at all, by deliberately testing at a high fluid flow rate. The flow rate was set 20 at 0.64 ml/minute, using phosphate-buffered saline c~ntn;n;n~ 0.296 bovine serum albumin, at a p~ of 7.2. A
flow rate of l minute sufficed for removal of more than 90%
of the l;F- -- that r~ ;n~d unbound at eSr~il;hrium.
This was completed by additional in-; ~ t~l removals of 4%
25 and 2% brought about by ~lows of 5 minutes and lO minutes, respectively, leaving the 1 ;F- 7 that were bound at ~;l;hrium attached to the cell monolayer. Increaslng the flow tlme beyond lO minutes (for example, to 15 minutes) did not generate ~ny additional removal of 1 ;ros -- from 30 the cell monolayer. A second experiment ~ , ed the retention of regular (non-hio~qh~ive) l; ro_ - to bir~h~Aive l;ror ~-, after a lO minute ~low rate of 0.64 ml/minute. The retention of bio~lh~ive liposomes was Wog6/lo3s4 `? ~ 2176566 r~l,U~ 3l,6 found to be at least 2 fold higher th~n that o regular ; ro~
The above data clearly indicate that the bio:~h~ive 1 ;L O= -- of the present invention possess the ability to 5 be retained at the site to which they will be administered, for prolonged periods of time despite fluid dynamics.
Furl h~ e, the data indicate that once the fraction of that remain unbound at eq~; 1 ;hrium is removed, the bio~h~ive 1 ;F-= - rem in bound to the cell 10 monolayer ~t a level close to the eSr~;l;hrium binding, ~urviving expo~ure to ~r~nt;n~ u~ fluid ~low at a rate well ~bove the dynamics of body fluids expected in the biological targets. This feature ~nh~nl~e~ the novel delivery and retention ability of the bio~lhsa;ve liposomes 15 of the present invention.
While the preferred: - '; ta have been described, various ~; f; cAtions and substitute~ may be made without departing from the scope of the invention. For example, the pre-activation of the carboxylic residues of hyaluronic 20 acid could be ~ . letec7 with dicyclohexyl~rho~;;m;d~ or with N,N'-~ ;n;m;dyl ~-~rh-~n~te. Additionally, the mouse EGF and human urogastrone used in the disclosed could be substituted with growth factors from other natural or synthetic ~ources. Accordingly, it is to 25 be under~tood that the invention has been described by way of 111u~tr tio= =d not l~ t~t10~.

Claims

CLAIM:

1. A process for producing bioadhesive liposomes comprising the steps of:
(a) providing a reaction vessel containing a liposome having phosphatidylethanolamine;
(b) admixing a recongnizing substance component to the reaction vessel to form a reaction mixture with the liposome;
(c) buffering the reaction mixture;
(d) admixing a crosslinking reagent to the buffered reaction mixture; and, (e) incubating the buffered reaction mixture and reagent for a period of time sufficient to form the bioadhesive liposomes.

2. The process of Claim 1 wherein the liposome is selected from the group consisting of multilamellar vesicles and unilamellar vesicles.

3. The process of Claim 2 wherein the unilamellar vesicles include microemulsified liposomes and large unilamellar liposomes.

4. The process of Claim 1 wherein the recognizing substance component is selected from the group consisting of collagen, gelatin and a growth factor.

5. The process of Claim 4 wherein the growth factor is epidermal growth factor.

6. The process of Claim 4 wherein the growth factor is urogastrone.

7. The process of Claim 1 wherein the crosslinking reagent is glutaraldehyde .

8. The process of Claim 1 wherein the crosslinking reagent is a bifunctional oxirane.

9. The process of Claim 1 where in the crosslinking reagent is ethylene glycol diglycidyl ether.

10. A bioadhesive liposome produced according to the method of Claim 1.

11. A process for creating a bioadhesive liposome formed by covalently bonding a recognizing substance to a liposome creating a bioadhesive liposome comprising the steps of:
(a) providing a reaction vessel containing a liposome having phosphatidylethanolamine;
(b) admixing a recognizing substance selected from the group consisting of collagen, gelatin and growth factor to the reaction vessel to form a reaction mixture with the liposome;
(c) buffering the reaction mixture;
(d) admixing a crosslinking reagent to the buffered reaction mixture;
(e) incubating the buffered reaction mixture and reagent for a period of time sufficient to form the covalent bonding of amine residues of the recognizing substance and liposome to create a bioadhesive liposome; and, (f) separating and removing of excess unreacted and noncovalently bound materials from the bioadhesive liposome.

12. The process of Claim 11 wherein the liposome is selected from the group consisting of multilamellar vesicles and unilamellar vesicles.

13. The process of Claim 12 wherein the unilamellar vesicles include microemulsified liposomes and large unilamellar vesicles.

14. The process of Claim 11 wherein the growth factor is epidermal growth factor.

15. The process of Claim 11 wherein the growth factor is urogastrone.

16. The process of Claim 11 wherein the crosslinking reagent is glutaraldehyde.

17. The process of Claim 11 wherein the crosslinking reagent is a bifunctional oxirane.

18. The process of Claim 11 where in the crosslinking reagent is ethylene glycol diglycidyl ether.

19. A bioadhesive liposome produced according to the method of Claim 11.

20. A process for covalently bonding a recognizing substance which is hyaluronic acid to a liposome creating a bioadhesive liposome comprising the steps of:
(a) providing a reaction vessel containing a liposome having phosphatidylethanolamine;
(b) activating the hyaluronic acid by pre-incubation in acidic pH with a crosslinker;

(C) admixing the activated hyaluronic acid to the reaction vessel;
(d) buffering the reaction mixture of the liposome and the activated hyaluronic acid to a basic pH;
(e) incubating the buffered reaction mixture for a period of time sufficient for the bioadhesive liposome to form.

21. The process of Claim 20 wherein the hyaluronic acid may be activated with a mixture consisting of DMSO and acetic anyhdride at an acidic pH.

22. A bioadhesive liposome produced from the process of Claim 20.

23 . A bioadhesive liposome comprising a liposome covalently linked to a recognizing substance by a crosslinking reagent.

24. The bioadhesive liposome of Claim 23 wherein the liposome includes phosphatidylethanolamine.

25 . The bioadhesive liposome of Claim 23 wherein the liposome is selected from the group consisting of multulamellar vesicles and unilamellar vesicles including microemulsified liposomes and large unilamellar liposomes.

26. The bioadhesive liposome of Claim 23 wherein the recognizing substance is selected from the group consisting of collagen, gelatin, hyaluronic acid and epidermal growth factor.

27 . The bioadhesive liposome of Claim 26 wherein the recognizing substance component is urogastrone.

28 . The bioadhesive liposome of Claim 23 wherein the crosslinking reagent is glutaraldehyde.

29. The bioadhesive liposome of Claim 23 wherein the crosslinking reagent is a bifunctional oxirane .

30. The bioadhesive liposome of Claim 22 wherein the crosslinking reagent is ethylene glycol diglycidyl ether.

31. A bioadhesive liposome designed for the sustained release of a therapeutic substance comprising a liposome component, the theraptic substance encapsulated by the liposome component, and a recognizing substance component bonded to the liposomal surface.

32. The bioadhesive liposome of Claim 31 wherein the recognizing substance component confers to the liposome component target specificity for and retention at a designated biological target site for release of the therapeutic substance.

33. The bioadhesive liposome of Claim 32 wherein the recognizing substance component confers to the liposome component target specificity for and retention at the designated biological target site for release of the therapeutic substance from the liposome component without interference from the attachement of the recognizing substance.

34. The bioadhesive liposome of Claim 31 wherein the recognizing substance component is selected from the group consisting of gelatin, collagen, hyaluronic acid and epidermal growth factor.