Note: Descriptions are shown in the official language in which they were submitted.
b ~ 51.,~9
PATAP363
G~N~TICALLY ENGIN~ZR~D ENDOTH~LIAL
OE LLS ~D USE THæ~OF
This invention relates to genetically engineered
cells, and to the use thereof. Still more
particularly, this invention relates to genetically
engineered endothelial cells and the use thereof for
e~pressing a therapeutic agent.
There have been numero~s proposal~ with respect
to genetically engineering mam~alian cells. In
general, retroviruse~ have been employed for
introducin~ genetic material into mammalian c~
Thus, there have been proposals to genetically
engineer bone marrow and hematopoietic progenitor
cell~ by the use of retroviral vectors. In general,
there have been drawbacks to the use of -quch cell~,
such as the variable ability to express certain gene~
and~or inefficient gene tran~fer. Thus, there have
been further propo~als for genetically engineering
cells capable of both long term sur~ival and ~table
e~pression including cells such a~ fibroblasts,
lymphocytes, keratinocytes and hepatocytes for u~e in
8ene therapy.
~c~o~ 9
The present invention is directed to genetically
engineered endothelial cells, and the use thereof for
expressing a heterologou~ protein. In one embodiment
the heterologous protein i9 a therapeutic agen..
According to one aspect of the present
invention, there i9 provided endothelial cells which
are transformed with at lesst one gene which encode~
for at least one heterologous protein, which is
preferably a therapeutic agent.
In accordance with another aspect of the present
invention, there is provided a solid support which
includes endothelial cells transformed with at least
one gene which encodes for at least one heterologous
protein, preferably a therapeutic a8ent. In a
preferred embodiment, the ~olit support i~ one which
i9 compatible with blood and may, for example, be in
the form of a blood vessel graft.
In accordance with yet another aspect of the
present invention, endothelisl cells which are
transformed with at least one gene which encodes for
at least one heterologous protein, preferably a
therapeutic agent are implanted in a blood vessel.
More particularly, the endothelial cells
employed in the present invention are endothelial
cells derived from a mammal. The endothelial cells
are obtained from a blood vessel. The term "blood
ve~sel" as used herein includes veins, arteries and
capillaries. Thus, the endothelial cells which are
genetically engineered include macrovascular and/or
microvascular endothelial cells.
The mammalian cells may be derived from a human
or nonhuman mammal. The endothelial cells are
preferably deri~ed from a human.
The endothelial cells are tran9formed with st
least one gene which encodes for at least one
~6~0~
het~sologous protein which is preferably a
therapeutic agent. The cells may be transformed in a
manner in which the therapeutic agent is secreted
from the transformed cells or may be transformed in a
manner in which the therapeutic a~ent remain~ in or
on the transformed cells.
The mammalian endothelial cells are transformed
with 8 suitable vector or e~pre~sion vehicle which
includes a gene for at lesst one therapeutic agent.
The vector includes a promoter for expression in
mammalian cells; for example, SV 40, LTR,
metallothionein, PGK; CMV; ADA; TK; etc. The vector
may also include a ~uitable signal sequence or
sequences for secreting the therapeutic agent from
the cells. The selection of a suitable promoter is
deemed to be within the skill of the art from the
teachings herein.
The expression vehicle or vectos is preferably a
viral vector and in particular a retroviral vector.
As representative examples of suitable viral vectors
which can be modified to include 8 gene for a
therapeutic agent, there may b~e mentioned: Harvey
Sarcoma virus; ROUS Sarcoma virus, MPSV, Moloney
murine leukemia virus, DNA viruses (adenovirus) etc.
Alternatively, the expression vehicle may be in the
for~ of a plasmid. The expre~sion vehicle may also
be in a form other than a vector; for example,
tran~formation may be accomplished by liposome
fusion, calcium phosphate or dextran sulfate
transfection; electroporation, lipofection, tungsten
particles etc. The selaction of a ~uitable vehicle
for transformation is deemed to be within the scope
of those skilled in the art from the teachings
herein.
39
In employing a retrovlral vector a~ the
expre~ion vehicle for transforming endothelial
cells, steps should be taken to eliminate and/or
minimize the chances for replication of the virus.
Various procedures are known in the art for providing
helper cells which produce viral vector particles
which are essentially free of replicating virus.
Thus, for e~ample, ~arkowitz, et al., "A Safe
Packaging Line for Gene Transfer: Separating Viral
Genes on Two Different Plasmidq," Journal of
Viroloxy, Vol. 62, No. 4, P8q- 1120-1124 (April
1988); Watanabe, et al., "Construction of a Helper
Cell Line for Avian Reticuloendotheliosi3 Virus
Cloning Vectoss, "Molecular and Cellular Biolo~Y,
Vol. 3, No. 12, p8s- 2241-2249 (Dec. 1983); Danos, et
al., "Safe and Efficient Generation of Recombinant
Retroviruses with Amphotropic and Ecotropic Host
Ranges, "Proc. Natl. Acsd. Sci. Vol. 85, pgs.
6460-6464 (Sept. 1988); and Bosselman, et al.,
"Replication-Defective Chimeric Helper Proviruses and
Factors Affecting Generation of Competent Virus:
Expression of Moloney Murine LeuXemia Virus
Structural Genes via the Metallothionein Promoter,
"Molecular and Cellular Biolo~, Vol. 7, No. 5, pg~.
1797-1806 (May 1987) di~clo~e procedures for
producing a helper cell which minimize~ the chances
for producing a viral particle which includes
replicating virus. This procedure and other
procedures may be employed for genetically
engineering with endothelial cells by use of a
retrovirsl vector.
The endothelisl cells which are to be
genetically engineered in accordance with the present
invention may be derived from a mam~al, and as
hereinabove indicated, quch endothelial cells msy be
)51~39
obtainet from an appropriate blood ve~el, ~uch a~ an
artery, vein or capillary. The proc~dure for
obtaining ~uch endothelial cells from the blood
vessel of a mammsl are generally ~nown in the art and
a representative procedure i~ disclo~ed in the
E~amples.
She inventi~n will be further described with
respect to endothelial cells genetically engineered
with a gene for a therapeutic sgent; however, the
scope of the invention is not to be limited thereby.
For example, the endothelial cells may be genetically
engineered with a gene for a protein which is not a
therapeutic aBent; for example, a marker protein,
such as beta-galactosidase.
The endothelial cells are genetlcally engineered
to include a gene for a therapeutic agent by the use
of an appropriate vector, with the vector preferably
being a retroviral vector. A representative
proceture for genetically engineering endothelial
cells by the u~e of a retroviral vector is described
in the e~amples, and such general procedure and
others may be employed for int~roducing other genes
into mammalian endothelial cells. Thus, as described
in the E~amples, the procedure basically involves
introduction of an appropriate promoter and DNA for
the de8ired therapeutic agent into an appropriate
retroviral vector. In addition to the promoter and
the gene for the therapeutic agent, other material
may be included in the vector such as a selection
gene; for example a neomycin resistance gene; a
sequence for enhancing expression, etc.
The appropriate vector now containing a gene for
at least one desired therapeutic agent i9 employed
for transducing mammalian entothelial cells by
procedures generally available in the art.
X~53~9~
-B-
In accordance with an aspect of the present
invention, genetically en8ineered endothelial c~lls
and in particular those genetlcally engineered with
at least on~ gene for at least one therapeutic a8ent
may be supported on a solid support. The solid
support is preferably one which ls biocompatible with
blood whereby the solid support including the
genetically engineered endothellal cells may be
placed in communication with the blood system of a
patient. Thus, for example, the solid support msy be
employed in an e~tracorporeal device or implanted in
a blood vessel (the term implant in a blood vessel
includes a by-psss or a shunt for a blood vessel).
The implantation may take the form of a blood vecsel
graft (the term graft includes a shunt or bypass).
It is to be understood, however, that the solid
~upport may take a variety of forms, such as pads,
strips, gels, etc. and is not limited to grafts.
The genetically engineered mammalian endothelial
cells, which include a gene for a therapeutic agent,
may be implsnted in a blood vessel of a mammal. The
mammalian endothelial cells wh~ich are genetically
engineered in accordance with the present invention
are derived from a mammal, and the transformed
endothelial cell~ are implanted in a blood vessel of
a mammal of the ~ame species. In a preferred
embotiment~ the genetically engineered mammalian
endothelial cells are implanted in the blood vessel
of a host from which the cells were originally
derived; i.e., autologous cells are employed. Thus
in a preferred embodiment, endothelial cells are
derived from a blood vessel of a patient, genetically
engineered to include a gene for at lea~t one
therapeutic agent and the genetically engineered
cells are implanted in a blood ves~el of the patient
. . X~)0~199
fron which they were derived. In this manner,
autologous genetically engineered endothelial cells
are employed for in vivo production of a therapeutic
agent for treatment of a patient, i.e., gene therapy.
It is to be understood that the genetically
engineered endothelial cells may be implanted in a
blood ves~el on a solid support or implanted directly
onto a blood vessel (without the use of a ~olid
support). It is also to be understood that the
endothelial cells may be placed on a solid support in
an e~tracorporeal device in communication with the
blood sy~tem.
In accordance with a preferred embodiment, the
genetically engineered mammalian cells are implanted
in 8 blood vessel by providing a blood vessel graft
which includes the genetically engineered endothelial
cell3. The graft now including genetically
en8ineered endothelial cells may be inserted intc a
blood vessel of a ho~t. Thus, in accordance with one
aspect of the present invention, there i9 providet a
blood vessel graft which includes genetically
engineered endothelial cells which are suitable for
use in a mammalian host, which may be a human or
nonhuman mammal.
The blood vessel graft may be any one of a wide
variety of vascular grafts and such grafts may be of
various sizes. The graft may be used in a vein, an
artery, or a capillary. The selection of appropriate
grafts is deemed to be within the scope of those
skilled in the art from the teaching~ herein.
Although in most caQes a synthetic vascular graft is
preferred, it is possible within the spirit and scope
of the present invention to provide a blood ves~el
derived from a host with genetically engineered
endothelial cells and then graft such blood ves~el
3~ 9
bac~ into the host. Thus the term ~raft i~clude~
nstural and synthetic graft 9 .
The graft may be provided with genetically
engineered endothelial cells in accordance with the
present invention by ~eeding the genetically
engineered endothelial cells onto a ~uitable blood
vessel graft. Represèntstive graft and procedure iq
disclosed in the Examples. The present invention is
not limited to such grafts and procedures. Other
grafts and procedures for qeeding endothelial cells
onto the graft are known in the art ant may be used
in the present invention. For e~ample, Herring et al
Eds. Endothelial Seeding in Vascular Surgery (Grune &
Stratton, Inc. Orlando 1987); Ziller et al Eds.
Endothelialization of Vasculsr Grafts, 1st European
Workshop~ on Advanced Technologies in Yascular
Surgery, Vienna, Nov. 5-6 ~Karger, Basel, 1986). As
representative graft materials, there may be
mentioned polye~ters (for example DACRON); expanded
polytetrafluroethylene (Gore-Tex); polyurethane~;
coated polyurethanes; such as a silicone coated
polyurethane msnufactured by C~orvita corporation in
Miami, Florida; tubular slotted stainle~s ~teel
stents (Johnson and Johnson) which are coated with a
substrate to permit adhesion of the endothelial cells
to the stents; natural blood vessels, etc.
The graft, now including geneticslly engineered
endothelial cells, may then be inserted into a blood
vessel of 8 host. The procedures for placing 8 graft
in an appropriate blood vessel are generally known in
the art, and such procedures are spplicable to the
present invention.
Alternatively, endothelial cells may be removed
from a blood vessel of 8 patient, genetically
19C~
eng~n~er~d and returned to a blood vessel of the
patient, without u~e of an implantable ~olid support.
The endothelial cells, which sre genetically
engineered with an appropriate therapeutic agent, may
be genetically engineered in a manner such that the
therapeutic agent i9 secreted into the blood, whereby
~uch therapeutic agent may e~ert its effect upon
cells and tissues either in the immediate vicinity or
in more di~tal locations. Alternatively, the
therapeutic agent may not be secreted from the cell~,
and exert its effect within or on the genetically
engineered endothelial cells upon substances that
diffuse into the cell. Thus, for example, adenosine
deaminase (ADA) may function within the cell to
inactivate adenosine, a toxic metabolite that
accumulates in severe combined immunodeficiency
~yndrome; phenylalanine hytroxylase may function
within a cell to inactivate phenylalanine, a to~ic
metabolite in phenylketonuria, etc.
As hereinabove indicated, the endothelial cells
are transformed with a gene for at least one
heterolo~ou~ protein, prefersbly a therapeutic agent.
The term therapeutic agent is u~ed in its broadest
sen~e and means any agent or material which has a
beneficial effect on the ho~t. The therapeutic agent
may be in the form of one or more protein~. As
repre~entative examples, there may be mentioned:
CD-4; Factor VIII, Factor IX, von Willebrand Factor,
TPA; urokins~e; hirudin; the interferon~; tumor
necrosis factor, the interleukins, hematopoietic
growth factors (G-CSF, GM-CSF, IL3 erythropoietin),
antibodies, glucocerebrosidsse; ADA; phenylalanlne
hydroxyla~e, human growth hormone, in~ulin, etc. The
selection of a suitable gene is deemed to be within
~t~05~9~
-10-
the 5cope ~f those skilled in t:he art from the
teachings herein.
In u~ing the genetlcally engineered endothellal
cell~, it iq possible to employ a mi~ture of
endothelial cells which includes endothelial cells
genetically enBineered with a gene for a first
therapeutic agent and endothelisl cells genetically
engineered with a gene for a second therapeutic
agent. It is al~o poqsible to transform individual
endothelial cells with more than one gene.
The genetically engineered endothelial cells may
be implsnted in a blood vessel alone or in
combination with other genetlcally engineered
endothelial cell or with other genetically
engineered cells, such as smooth muscle cells,
fibroblasts, glial cells, keratinocytes, etc.
The use of genetically engineered endothelial
cellq permits a therapeutic a8ent to be introduced
directly into the blood. As a result of the location
of the endothelial cell~ in immediate contact with
the circulating blood, the survival and delivery of a
therapeutic agent is facilitat~ed.
The genetically engineered endothelial cells (by
selection of high producing clonal populations and/or
the use of vectors with enhanced expression) may be
employed to produce, in vivo, therapeutically
effective amounts of a desired therapeutic aBent for
treatlng a patient. In determining the number of
cells to be implanted, factors such as the half life
of the therapeutic agent; volume of the va~cular
system; production rate of the therapeutic agent by
the cells; and the desired dosage level are
con~idered. The selection of such vectors and cells
is dependent on the therapeutic agent and i9 deemed
2~05~99
to be w$thin the scope of those skilled in the art
from the teachings herein.
The drawing i9 a schematic illustration of
vectors used in the present invention.
The following Examples further illustrate the
present invention; however, the scope ~f the
invention i9 not to be limited thereby. In the
Examples, unless otherwise specified, restriction
enzyme digests, ligations, transformations, etc. may
be performed as described in Molecular Clonin~, A
Laboratory Manual by Maniatis et 81.
E~ample 1
A. To construct the pG2N retroviral vector of the
drawing u~ed to genetically engineer endothelial
cells to produce rat growth hormone, an SV40 promoted
neomycin resistance gene and a rst growth hormone
cDNA were placed into the pB2 retroviral vector
(Laboratory of Molecular Hematology, NI~). A growth
hormone cDNA was obtained by digesting the plasmid
RGH-l (Nature 270, 494 (1977)) with Xho I and Mae III
restriction endonucleases (~oehringer Mannheim
Biochemicals). This rat growth.hormone cDNA was
eletrophoretically i~olated out of an agaorse gel and
purified via binding/elution to glass beads,
Geneclean (BI0 101, LaJolla, California). This
growth hormone cDNA was then blunted using the large
fra~ment of DNA polymerase (Klenow) (New England
Biolabs) and nucleotide triphosphates 8S recommended
by the manufacturer. This fragment was then purified
with Geneclean.
The B2 vector was constructed in order to
replace the NeoR gene in N2, ~M.A. Eglitis, P.
Kantoff, E. Gilboa, W.F. Anderson, Science 230, 1395
(1985); D. Armentano et al., J. Virol, 61, 1647
(1987) and shown in the draw$ng] with a multiple
~)0~)19~
-12-
cloning ~lte. N2 was first digested with Eco RI,
thereby releasing both the 5' and 3' LTRs with the
ad~oining MoMLV flanXing sequences. The 3' LTR
fra8ment was ligated into th^ EcoRI ~ite of the
plasmid GEM4 (Promega Biotech). The 5' LTR fra8ment
with its flanking gag sequence was then digested with
Cls I, Hind III linkers were sdded, and the fragment
wa~ inserted into the-Hind III ~ite of pGEM4.
The pB2 ~ector wa3 digested with the HincII
restriction endonuclease (New England ~iolab~), and
pho~phstased using calf alkaline phosphatase.
(Boehringer Mannheim Biochemicals). The pB2 plaYmid
W85 then purified with Geneclean. The pB2 vector and
the rat growth hormone cDNA were then ligated using
T4 ligase (New England Biolabs). The ligation was
then transformed into competent DH5 bacteris (Bethe~
da Research Labs). Colonie~ were then screened for a
growth hormone cDNA containing vector. The new
vector was called pG2. pG2 was then digested with
BamHI (New England Biolabs), purified with Geneclean
(Bio 101), and blunt ended with the Klenow fragment
(New England Biolabs). A 340 ~sse pair SV40 promoted
neomycin resistance gene fragment wa~ isolated from
the pSV2CAT plasmid (ATCC accession number 37155) by
di8e~ting with PvuII and HindIlI (New England
Biolsb~). This fragment wa~ i~olated by agarose gel
electrophoresis ant purified with Geneclean. The
SV40-neomycin resistance fragment W8~ then ligated
using T4 ligase (New England Biolsbs) with pG2 and
tran~formed into DH5 competent bacteria per the
manufacturers in~truction (B~L). Colonies were
screened and the resulting pla9mid con~truct was
called pG2N.
5~9
-13-
The SA~ vector shown ln th~ drawing was obtained
as described in Proc. Na~l. Acad. Sci. USA 83:6563
(1986).
The recombinant vectors (N2,SAX, G2N) used in
the study were each separately transfected into the
currently available retroviral vector packaging cell
lines, including the amphotropic packaging lines,
PA12 (Science 225:630 (19~4) and PA317 (Mol. Cell.
Biol 6:2895 (1986), and the ecotropic line, P~i2
(Cell 33:153 (1983). These lines were developed in
order to allow the production of helper virus-free
retroviral vector particle~.
Aortic endothelial cells were obtained from New
Zealand White rabbits (2-5 kilograms) by methods
described previously for obtaining endothelial cell~
from bovine pulmonary artery (U.S. Ryan, M. Mortara,
C. Whitaker, Tissue ~ Cell 12, 619 (1980)). The
rabbit was anesthetized (1 ml sodium pentobarbital)
and the aorta was removed and placed in Hanks
buffered saline containing 3X antibiotics. The
vessel wa~ slit longitudinally and the luminal
surface was ~craped with a #}1 scalpel blade taking
care to scrape each area only once. The initial
isolates were grown in Ryan Red medium [Ryan et al J.
Tissue Cult. Method~, 10:3 (1986)], purified by
selection of endothelial "islands" and pas~aged with
a rubber policeman. Passsged cells were grown in
Rrimaria 25 cm2 flasks.
A confluent 100 mm ti9sue culture dish (Costar)
was harvested with a cell scraper. Following the
dispersal of the cells by titurating 10-20 times with
a 5 ml pipet, the cells were plated into 2-100 mm
tissue culture dishes with 8 ml Ryan Red medium.
After an overni~ht incubation, the medium was removed
snd 5 ml retroviral Yector supernatant was added with
)5~3'9
Polybrene at a final concentrstion of 8 ug/ml. After
a 2 hour incubation an additional 5 ml of Ryan Red
was added to the di~h. The cells were incubated
overni~ht and the medium was replaced with 8 ml of
Ryan Red. After another overnight incubation G418
was added to a final concentrstion of 200 ug/ml. The
cells were the fed every 3-4 d~ys with Ryan Red
containing 200 u6/ml G418. The cell were
sub3equently maintained in Ryan Red without G418.
G2N-infected RAEC that had been selected in
G418-containing growth medium were harvested with a
rubber policçmsn from 2 confluent T75 flssks and
su~pended in 5 ml of Ryan Red. The cell su~pension
was titurated 6-7X with a 6 cc syringe and a 23 gauge
needle. The cells were pelleted and resuspended in
1.25ml of Ryan Red. A vascular clamp was attached to
one of a lO cm 2 4 mm (inner diameter) Corvita grsft
(Cor~ita Corp., Miami, FL) which is a silicone coated
polyurethane graft. The cell suspenslon was vortexed
and introduced into the open end of the graft with a
3 cc syrin~e and 20 gsuge needle. A second vascular
clamp was attached to the ope~ end and the graft was
placed into a 50 ml conical tube filled with Ryan Red
medium. The conical tube was capped, wrapped in
parafilm, and placed into a roller bottle. The
roller bottle was rotated overnight at 37 degrees
Centrigrade. The ne~t day the clamps were removed
and the 8raft was placed into a 500 ml bottle
containing 150 ml of Ryan Red. The bottle was placed
in an incubator equilibrated with 5% C02, at 37
degrees C. It remained in the incubator and was
periodically rotated for the next nine days. At that
time the graft was transferred 1nto a T75 flask, fed
with fresh medium and periodically sampled for rst
growth hormone produc~ion over the ne~t four weeks.
-15-
rGH continued to be ~ecreted into the tissue
culture medium at a rate of appro~imately 1000 ng/106
cell~/day for at least 4 weeks after seeding the
graft a~ follows.
Cell o~ ~raft Rat growth hormone6
6 ~ 10 production (ng/lOE
(cells/cm2) cells/24 hours)
Dsy 13 930
Day 32 1060
B. Rabbit endothelial cells were sl~o tran~fected
with the vector SAX by the hereinabove described
procedure and quch tranfected cells were found to
expres~ human ADA.
C. The CD4 containing plasmld (pT4B, a gift of
Richard Axel of College of Physicians and Surgeonq
Columbia University, New YorX, New York) was digeqted
with the rs~triction endonucleasas Eco RI and Bam HI
New England Biolabs, Beverly MA) to release the CD4
gene which was isolated by agaro~e gel
electrophoresis ~ollowed by purification via
binding/elution to glass beads (using the geneclean
product, BIO 101, La Jolls CA in the manner
recommended by the msnufacturer). The CD4 fragment
was li~ated (usin~ T4 DNA ligase as recommended by
the suppller, New England Biolabs) into Eco RI plus
Bam HI cut Bluescript cloning vector (~tratagene Co.
La Jolla CA). The ligation was then transformed into
competent DH5 alpha bacteris (Bethesda Research Labs,
Gaithersburg MD) and white colonies were isolated and
screened for proper insert size to yield the plasmid
pCDW. To produce a suitable plasmid ba~ed e~pres~ion
vector for the CD4 gene; the plasmid SV2neo (obtained
form American Type Culture Collection, Rockville MD~
was digested with Hind 3 plus ~pa I, and a ~ynthetic
5199
-lB-
polylin~er sequence from the pUC-13 vector
(Pharamicia, Piscataway NJ) was inserted (via T4 DNA
ligase) in place of the neo 8ene of pSV2neo. This
ligation wss transformed into DH5 bscteria (Bethesda
Research Labs) and colonie3 screened for the presence
of restriction enzyme sites unique to the polylinker
to yield the vector pSVPL. The pSVPL e~pression
vector wss further modified by the insertion of an
Xho I linker (conditions and reagents supplied by,
New England Biolabs) into the Pvu II site on the 5'
side of the SV40 early region promoter to produce
pSVPLX.
The pCDW and pSVPLX pla~mids were digested with
enzymes Hind 3 plus ~ba I (New England Biolabs) and
their DNAs isolated (using the Gene Clesn product)
following agarose gel electrophoresis. Ligation of
the CD4 fra8ment into the pSVPL~ vector wa~ performed
and colonies were screenet to yield pSVCDW in which
the SV40 viru~ early region promoter is used to drive
the expression of the complete CD4 8ene product. The
next step wss to produce a form of the CD4 gene such
that it would be exported from the cell as an
extracellular product.
The production of a soluble form of CD4 was
accomplished by the use of a specially designed
oligonucleotide adaptor to produce 8 mutant form of
the CD4 gene. This adaptor has the unique property
that when inserted into the Nhe I site of the CD4
gene it produces the preci~e premature termination of
the CD4 protein amino acid sequence while
re8enerating the Nhe I site and creating a new Hpa I
site. This oligonucleotide adaptor (~ynthesized by
Midland certified resgent Co.) was produced by
annealing two phosphorylatet oligonucleotides; 1) 5'
CTAGCITGAGTGAGIT 3', 2) AACTCACTCAAG and then this
~0051~9
product wa~ llgated into the ~lte of pSVCDW. The
ligation reaction was then cleaved with Hpa I and
then ~ho I linkers were added (New England Biolabs).
The linker resction wss terminated by heating at 65C
for 15 min. and then sub~ected to digestion with Xho
I restriction endonuclease (New England Biolab~).
This reaction was then subJected to agarose gel
electrophoresi~ and the fragment containing the
SV40-CD4 adsptor i~olated (Geneclean). The
retroviral vector N2 was prepared to accept the
SV40-CD4-adaptor fragment by digestion with Xho I and
treatment wlth Calf intestinal phosphstase
(Boehringer Mannheim, Indianpoli~ IN). The ligation
of CD4 expres~ion cas~ette wa9 performed with an
insert to vector ratio of 5:1 and then transformed
into DH5 competent bacteria (Bethesda Research Labs).
Constructs were analyzed by restriction endonuclease
digestion to screen for orientation and then grown up
in large scale. The construct where the SV40 virus
promoter i~ in the same orientation as the viral LTR
promoters is known as SSC while the construction in
the reverse or reverse orientation is called SCSC.
The SSC vector is packaged into PA317 cell line
as described by Miller et al supra. to provide PA 317
cells capable of producing soluble CD4 protein.
The SSC vector packaged PA 317 cells were used
to transduce rabbit endothelial cells as hereinabove
described.
The transduced endothelial cells were found to
express soluble CD-4.
D. Collsgen sponges containing sdsorbed HBGF-I
were 9urgically implantet in the abdominal cavity of
a rat near the li~er (Science 241, 1349 (1988).
Seven (7) to ten (10) day8 post implantation, sponges
were surgically removet and dige~tet 30 to 60 min. at
i~C~05~'~9
-18-
37C with a solution of colla6enase in phosphate
buffered ~aline (1 mg/ml) using a tissue culture
incubator (5% C02). Released cells were collected by
centrifugation (10 min., 1000 RPM, 20C) and washed
once with phosphste buffered saline (PBS) and
pelleted by centrifu6stion. Cell~ were resuspended
with 30 ml of media contsining:
Ml99 media (Gibco)
ECGF (crude brain e~tract ) 7.2mg
Hepsrin (Up~ohn) 750 units
20% conditioned cellular media collected a~
supernatant from confluent dishes (48 hr.) of either
bovine sortic or human umbllicsl vein endothelial
cells
10% Fetal Calf serum (Hyclone)
3000 units Peniclllan G (Biofluids)
3000 units streptomycin sulfate (Biofluids)
and plated for 16 hour~ on 100 mm tissue culture disk
coated with fibronectin (human) using lug/cm2.
Plated cells were washed with P~S three times and fed
15ml of previously mentioned media. Media was
changed every 2 days for the duration of the
procedures .
Selected rat endothelial cells were tran~duced
with the N-7, SA~, G2N and SSC vectors by the
following procedure:
1. 2 X 106 microendothelial cells
(monolayer 80% confluent)
2. 2 X 106 viral supernatant
3. Polybrene (8ug/ml)
-Combine 1, 2, 3 in 5ml total volume for 2-3 hours at
37C (5~ C02).
-Add 20ml of ti~sue culture media for 16 hrs. at 37C
(5% C02).
~05~'3~
-19-
-Aspirnte off media (virus contsining), add fresh
culture media.
-After 48-96 hours add G418 (800ug/ml~ and culture
media.
-Select for one to two week~ changing media every two
dsys.
Cells tran~duced with N7 e~pressed neomycin,
those tran~duced with SAX expressed ADA; tho~e
transduced with G2N e~pressed rat growth hormone; and
those transduced with SSC e~pressed sCD4.
The rat endothelial cells transduced with G2N
expressed rat growth hormone in vitro as follows:
106 cells produce 3.0 Y8 after 24 hrs.; 6
,ug after 48 hr~.; and 9.0 ,ug after 72 hours.
E~ample 2
A. Endothelial cells were harvested from segments
of adult sheep jugular vein, carotid artery, and
femoral vein using the method of Jaffee et al.
J.Clin.Invest., 52:2745-2756 (1973). A total of four
vessels from three sheep were used. Identification
of har~ested endothelial cells was confirmed by their
cobblestone structure and by confirmed by their
cobbleston~ structure and by binding of the
fluorescent ligand DiI-Acetyl-LDL J.Cell.Biol,
99:2034-2040 (1984)). (Biomedicsl Technologies,
Stoughton, Massschusetts). Cells were cultured on
fibronectin (Collaborative Research, Bedford,
Ma~sachusetts) coated plastic cul~ure dishe~ (1.0
,ug/cm2 in M-199 Biofluids, Rock~ille, Maryland) with
20% fetal calf serum (Hyclone Laboratories, Logan,
Utah), 100 U/ml penicillin, 100 ~g/ml streptomycin,
and 0.25 ~ug/ml amphotericin B (Biofluids). Cells
were pas3aged using trypsin-EDTA (Biofluid~
digestion. Remo~al of ~heep vessels was done
according to protocols appro~ed by the snimsl use
~O~S~3!3
-20-
committee of the Nationsl Heart, Lung, and ~lood
Institute .
B. A murine ecotropic psckaglng line capable of
tr~nsmitting the ~-galactosida~e-containing "BAG"
vector (Proc. Nat~.Acad Sci, 84:156-160 (1987) wa~
provided by Constance Cepko (Harvard University,
CambridKe, Massachusetts). Supernstant from this
packaging line was used to generate an amphotropic
psckaging line from PA-317 cells. A human t-PA cDNA
(in plasmid pPA34'f) (J.Biol. Chem, 260
t:ll223-11230(1985)) was provided by Sandra Degan
(University of Cincinnati, Cincinnati, Ohio). This
t-PA cDNA was u~ed, through several subcloning ateps,
to construct a t-PA contalning retroviral vector,
~2NSt analogous in construction to the SA~ vector.
The corre~ponding pla~mid, based on the B2 plasmid,
(Science), 343:220-222 (1989)) W8S transfected into
GPE-86 cells, (J.Virol., 62:1120-1124 (1988)) and
supernatant from these cells, thereby, generating
amphotropic packaging clones capable of transmitting
the t-PA gene. Endothelial cell3 were transduced by
incubation for 2 hours with supernatant-containing
virions with the retrovirsl vector, along with 8
~g/ml G-418 for at least 16 days. Duplicate cultures
of cells from each vessel harvest were transduced
simultaneously with either the t-PA- or
B-galactosidase-containing retroviral vector and,
then, cultured, passaged, and selected using
identical procedures. In thls manner, the t-PA- and
B-galactosidase-transduced cells served as controls
for one another in experiments involving either
B-galactosidase activity or t-PA secretion.
C. Tubular slotted stainlesa steel 1.6-mm diameter
stents (Circulation, 76:IV-27 (1987)) (Johnson and
John~on Interventional Systems, Warren, New Jer~ey)
3~ 3
-21-
were cut at the articulstion, and each half was
seeded with endothelial cells, u~ing a modification
of the method of Van der Gei~en et al. (J.Intervent.
Cardiol, 1:109-120 (1988)). A totsl of 10 stent
segments were qeeded. Endothelial cells will not
grow on bsre metal, and therefore the application of
a substrate is necessary before cell ~eeding. A
fibronectin costing is used in vitro to allow
endothelial cell adhesion to the 3tent~. Stents were
submerged in 100 ~g/ml human fibronectin for 15
minutes at 37 and, then, transferred to
polypropylene tubes containing a suspension of
6-lOX104 endothelial cells in 0.8 ml culture medium.
The tubes were plsced in a 37 incubator containing
5% C2 and rotated 180 every 10 minute~ for 2 hours,
after which the 3tents and cell suspension were
placed in well~ of plastic tissue-culture dishes and
additional culture medium added. Covera~e of the
stent surfaces was monitored both by phase-contrast
microscopy and by incubatlon of the stents for 4
hours in medium containing DiI-Acetyl-LDL followed by
fluorescence microscopy.
D. The presence of the B-galactosidase gene product
was determined by staining with 5-Bromo-4-
chloro-3-indolyl-B-D-galactopyranoside (~-Gal) EMB0
J., 5:3133-3142 (1986)) of cells either on
tissue-culture dishes or in situ on the stents.
Levels of human t-PA were determined by enzyme-linked
immunosorbent assay (ELISA) on tissue culture
supernatants using a commercially available kit
(Thromb. Res., 41:527-535 (1986)). (American
Diagnostica, New York, New York). Supernatant to be
assayed was collected above confluent monolayers in
35-mm dishes, 48 hours after addition of 2 ml fresh
medium. For measurement of t-PA secret~on from a
1'3~ `
-22-
seeded stent, the stent wa~ transEerred to a new well
containing fresh medium ant, then, began a timed
collection of culture medium. Harvested supernatant
was centrifuged at 15,000g for 15 mlnutes to remove
cellular debris, made 0.01% with Tween-80, and frozen
at -70C until sssayed. The rate of t-PA secretion
in nanogrsms per 106 cells per 24 hours was
calculsted using a confluent cell density of 3 ~ 104
cells per cm2 of tissue culture plastic (data not
shown) .
E. Seeded stents were incubated in medium
containing DiI-Acetyl-LDL for 4 hours before
e~pansion. The stents were v~sualized by
fluorescence microscopy to confirm endothelial
coverage, snd, then, manually placed over 8 deflated
3.0-mm dismeter coronsry angioplssty balloon catheter
(Scimed Life Systems, Maple Grove, Minnesota). After
balloon inflation to 4-6 atmosphere~, resulting in
complete ~tent e~pansion, the bslloon wss deflated
and the ~tents were removed fro~ the catheter~ and,
agsin, viewed by fluorescence micro~copy.
F. Transduced sheep endothe~ial cells retained
their cobblestone structure and their ability to bind
the fluorescent ligand DiI-Acetyl-LDL. No difference
in structure was detectable between those cells that
had been transduced with the B-gslactosidase vector
snd those thst were transduced with t-PA vector.
G. Only cells in cultures trsnsduced with the
B-galsctosidase gene e~hibited deep blue cytoplasm on
staining with X-Gal. After G-418 selection, most of
the B-galctosidase-transduced cells stained deep blue
with X-Gal.
H. Endothelial cells from all four vessel~, when
transduced with the t-PA vector, secreted
immunoreactive t-PA. Rates of t-PA secretion (mesn
X~05~9
-a3-
SD of duplicste tissue culture wells, e~pres~ed as
ng/lO4 cells/24 hours) were femoral vein, 370 ~ 8;
carotid artery, 660 + 240; ~ugular vein l, 230 ~ 6;
~ugular vein 2, 200 ~ 18. t-~'A production by the
~-galacto~idase- transduced cell~ was below the lower
limit of ~ensitivity of the as3ay (i.e., less than 5
ng/lO4 cells/24 hours) in all of the supernatant~
testet.
I. Fluorescence microscopy of si~ of the seeded
stents confirmed complete co~erage of the visible
stent surfaces. When eight stents seeded with either
B-galactosidase- or t-PA-transduced endothelial cells
were stained with X-Gal, the stents covered with
B-galactosidase- carrying celLi turned blue, whereas
the stents covered with t-PA-secreting cells did not.
Measurement of human t-PA levels from the cell
culture medicum surrounding the stents confirmed that
t-PA was being ~ecreted only by the t-PA-transduced
endothelial cells. Three stents seeded with
t-PA-transduced endothelial cells, secreted 6.3, 4.8,
and 2.6 n8 t-PA/24 hours. t-PA secretion by the
B-galactosidase-transduced cel~l~ on each of three
stents, if present, was below the limit of detection
of the assay. To check the internal consistency of
our results, the measured t-PA secretion from each of
three lines of transduced cells was used both before
and after they were seeded onto stents to calculate
the surface area of the stent~. This calculation i~
based on the assumption that the density of the cells
and the rate of t-PA secretion do not change when the
cells are on the stents. A stent surface area (mean
i SD) of 48 i 19 mm was calculated, not significantly
different fsom the manufacturer's value of 42 mm2
(personal communication, John90n and Johnson
Int2rventionsl Systems, Warren, New Jersey).
X0051'39
-24-
J. ~our stent~ covered with DlI-Acetyl-LDL- stained
endothelial cells were expanded using balloon
catheters and immediately viewed with a fluorescene
micso~cope. Near-complete retention of the cell~ on
the e~terior surfaces of all four stents was
confirmed. X-Gal stainin8 of stents was confirmed.
X-Gal staining of stents carrylng B-galactosidase-
transduced cells permitted evaluatlon of cellular
retention on all ~urfaces after balloon inflation.
The stents were viewed with a dlssecting microscope,
and cellulsr retention on all surfaces was estimated.
A total of ~ight expanded stent~ were observed after
X-Gal staining, four covered with B-galacto~idase-
transduced endothelial cells. Much of the interior
lumen surface of the stents was free of cells after
balloon in1ation but that the cellular layer on the
exterior and lateral ~tent-strut surfaces was largely
intact.
Although the ~cope of the present invention is
not intended to be limited to any theoretical
reasoning, it is believed that an intravascular stent
seeded with endothelial cells ~as hereinabove
described may produce a local thrombolytic
environment in vivo. High level secretion of t-PA
ad~acent to a forming clot may permit t-PA to be
concentrated through the high affinity binding of
t-P~ to fibrin. (Thorsen, et al., Thromb D.
Haemorrh, 28:65-74 (1972)). In this manner,
fibrinolytic acti~ity would be directed to
microthrombi beginning to form on the ~tent surface
or do~nstream, thus prevent~ng the formation of
occlusive thrombi. It has been demonstrated
(Hergreuter, et al., Plast. Reconstr. Sur~.,
81:418-424 (1988~) in a rabbit model that locally
administered t-PA could abort thrombus formation on a
~O~a~99
-as-
hi6hly thrombogenic inverted artery. Intravascular
stents are far lass thrombogenic than is an inverted
vessel, and it is possible that localized delivery of
nanogram quantitie9 of t-PA will re~ult in qufficient
thrombolytic activity to prevent stent-related
thrombotic events.
It has also been t'neorized that the implantation
of genetically en8ineered endothelial cells on stent
surfaces offers a potential means of preventing
intimal hyperplasia because implanted endothelial
cells would be in direct contact with the intima and
could be engineered to secrete proteins capable of
inhibiting intimsl growth.
Numerous modifications snd variations of the
present invention are possible in light of the above
teachings; therefore, within the scope of the
appended claims, the invention may be practiced
otherwise than as particularly described.
