| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Fourth Department of Internal Medicine, Sapporo Medical University School of Medicine, Chuo-ku, Sapporo, Hokkaido, 060-8543, Japan
1Correspondence: Fourth Department of Internal Medicine, Sapporo Medical University School of Medicine, South-1, West-16, Chuo-ku, Sapporo, Hokkaido, 060-8543, Japan. E-mail: niitsu{at}sapmed.ac.jp
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: plasmid DNA • gene therapy • transferrin receptor • tumor targeting • metastasis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. By virtue of this unique intracellular behavior, molecules
conjugated to Tf may be efficiently and repeatedly delivered into
cells. Most normal cells other than immature erythroid cells are low in
Tf-R expression, but actively proliferating tumor cells exhibit high
levels (3
4
5)
. This suggests that targeting mediated by
Tf-R could provide a method for delivery of anti-cancer agents into
tumor cells in vivo as well as in vitro. In fact,
diphtheria toxin (6
, 7)
, Ricin A (8)
,
ribosome inactive protein (SO-6) (9)
, RNases
(10)
, neocarzinostatin (11
, 12)
, and
adriamycin (13)
conjugated to Tf have been successfully
delivered into tumor cells via Tf receptor. Previous attempts to
deliver genes by this method, however, have been unsuccessful. Wagner
et al. constructed Tf-gene conjugates using polylysine as a binding
moiety (14
15
16
17)
, but the conjugates tended to form
aggregates due to polycationic interaction (18)
; after
being internalized by the cell, they were routed mostly to the
lysosomal fraction, bypassing the recycling process, and underwent
degradation. Thus, the use of chloroquine or glycerol (19
, 20)
, which block activity of lysosomal enzymes, and an inactive
adenovirus (21
22
23
24)
, which facilitates the release of DNA
from the endosome to cytoplasm, were needed to attain efficient
transfection. Because of the toxicity of chloroquine, immunogenicity of
the adenovirus and nonspecific binding of polycationic molecules to the
cell surface, in vivo gene delivery using polylysine
conjugates is considered impractical. Alternatively, Cheng et al.
(25
, 26)
used a Tf tagged liposome in which the gene was
encapsulated prior to administration. However, administration was
performed intra-tumorally, not systemically. In the present
investigation, we used biotin-avidin bridging, which is
stoichiometrically controllable, to avoid aggregate formation,
and Tf-R affinity chromatography, which keeps the receptor binding
motif of Tf intact. Furthermore, a biotin reagent with disulfide bonds
was specifically chosen for this experiment to allow for intracellular
dissociation of DNA from Tf (7
8
9
10
11
12
13
, 27
28
29)
. The conjugate
constructed by this method was shown to provide high transfection
efficiency in vitro against tumor cells and specific
delivery of genes by systemic administration to disseminated tumors
in vivo, as revealed by successful systemic suicide gene
therapy. |
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The 5E7 was coupled
to Affi-Gel Hz (Bio-Rad, Hercules, Calif.) at a concentration of 2 mg
protein/ml of gel according to the manufacturer’s instructions.
Application of Tf-R-Tf complex to the 5E7 immobilized Affi-Gel Hz
column
Crude Tf-R-Tf complex extracted from human placenta (100 g)
according to the method of Turkewitz (30)
was applied to 4
ml of Sepharose CL-4B (Pharmacia, Piscataway, N.J.) column (20 cm x 1.5 cm, Bio-Rad) to remove any large aggregates, then the elutant
was applied to 2 ml of 5E7 immobilized Affi-Gel Hz column (5 cm x
1 cm, Bio-Rad). The column was washed with 100 ml of KPi/NaCl (10 mM
potassium phosphate, pH 7.5, 150 mM NaCl), 0.2% Triton X-100, 100 ml
of KPi/NaCl and then with 10 ml of KPi/NaCl, 10 mM
3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS). To
detach Tf from the complex, 2 ml of 50 mM sodium citrate, pH 4.9, 100
mM NaCl, 10 mM CHAPS, and 1 mM desferrioxamine mesylate (Ciba Geigy,
Summit, N.J.) were passed through the column. The column was then
washed with 10 ml of 50 mM hydroxyethylpiperazine-N-2-ethanesulfonic
acid (HEPES), pH 7.5, 100 mM NaCl, 10 mM CHAPS and subsequently with 2
ml of 2 M KCl, 50 mM HEPES, pH 7.5, 2 M KCl, 10 mM CHAPS.
HABA assay
Human Tf (Miles, Naperville, Ill.) saturated with iron and
dissolved at 5 mg/ml in phosphate-buffered saline (PBS) was reacted
with various molar ratios of NHS-SS-biotin (Pierce, Rockford, Ill.).
After being allowed to stand for 3 h at room temperature, the
reaction mixtures were dialyzed against PBS to remove any unreacted
biotin. Stoichiometry of biotin to Tf was determined using
4'-hydroxyazobenzene-2-carboxylic acid (HABA) dye (Sigma, St. Louis,
Mo.) (31)
. A preparation of biotinylated Tf with a ratio
of 1.2:1 was chosen for conjugate formation.
4. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
(SDS-PAGE) of biotinylated Tf bound to Tf-R
The samples (beads) obtained from each step described above were
boiled for 3 min in a buffer containing 1% SDS and 5%
2-mercaptoethanol. Two sets of electrophoreses were carried out on a
discontinuous mini gel (Daiichi Pure Chemicals, Tokyo, Japan). One set
was stained with the Silver Stain Plus Kit (Bio-Rad) and the other set
was electroblotted onto nitrocellulose membrane (BA85, Schleicher &
Schuell, Dassel, Germany) to visualize biotinylated bands with the
enzyme-catalyzed light reaction method using an Imaging
High-chemiluminescence Detection Kit (Toyobo, Japan) according to the
manufacturer’s instructions.
Expression plasmid DNA
ß-Galactosidase with nuclear localization signal (nlacZ)
expressing plasmids containing the CAG promoter (pCAGnlacZ) was
constructed by inserting the nlacZ gene into the EcoRI site
of pCAGGS (32)
. A plasmid (pCAHSVTK) of the HSV-TK gene
driven by the CAG promoter was kindly provided by H. Hamada (Cancer
institute Japanese Foundation for Cancer Research). Green fluorescence
protein (GFP) expression plasmid containing the CMV promoter (pEGFP-C1)
was purchased from Clontech (Palo Alto, Calif.).
Biotinylation of plasmids
Biotinylation of the plasmids (100 µg) was carried out by
using various concentrations of PHOTOPROBE (S-S) Biotin (Vector
Laboratories, Burlingame, Calif.; 0.1–1 mg/ml) according to the
manufacturer’s instructions.
Conjugate formation on the column
Five hundred micrograms of biotinylated Tf was applied to 200
µl of Tf-R immobilized column, which was washed with PBS after 1 h of incubation at 4°C. To this column, 20 µl of streptavidin
(Pierce) dissolved in PBS at a concentration of 5 mg/ml was applied.
After incubation for 1 h at 4°C, the column was washed with 10
ml of PBS, and 200 µl of PBS containing 200 µg of biotinylated
plasmid was added. To block the unreacted streptavidin, 10 µl of PBS
containing 50 µg of biotin (Pierce) was added to the column and
incubation continued for another 30 min. The conjugate was eluted with
2 ml of 50 mM sodium citrate, pH 4.9, 100 mM NaCl, 10 mM CHAPS, 1 mM
deferoxamine mesylate. The eluted apo Tf-DNA conjugate was saturated
with iron using Fe-NTA (1)
and dialyzed against PBS.
Agarose gel electrophoresis and immunoblotting
Aliquots containing 100 ng of conjugates were electrophoresed
together with control samples (GFP plasmid, biotinylated Tf,
streptavidin) on 0.8% agarose gel, stained with ethidium bromide, and
transferred onto the nitrocellulose membranes by capillary action using
10x solution of sodium citrate (SSC). Blots were incubated with rabbit
anti-human Tf antibody (1:500, DAKO, Glostrup, Denmark) or rabbit
anti-streptavidin antibody (1:1000, Chemicon International Inc., El
Segundo, Calif.), followed by incubation with horseradish
peroxidase-linked donkey anti-rabbit antibody, and developed in ECL
system (Amersham, Arlington Heights, Ill.).
Gene transfection with Tf conjugate
K562 cells (human erythroleukemia cell line, ATCC, Rockville,
Md.) that highly expressed Tf-R [95% on fluorescein-activated cell
sorting (FACS); data not shown], M7609 cells (human colonic cancer
cell line; 79% on FACS; ref 33
), TMK-1 cells (human
gastric cancer cell line; 67% on FACS, generous gift of E. Tahara,
Hiroshima University School of Medicine, Japan); and HEL (human
embryonic lung cell lines 3% on FACS, ATCC) were plated in a 24-well
plates (Costar, Cambridge, Mass.) at 5 x
105 cells per well with 1 ml of RPMI 1640
(Gibco-BRL, Grand Island, N.Y.) containing 10% fetal calf serum. One
hundred microliters of the conjugates containing 5 µg of DNA (GFP
gene, nlacZ gene, HSV-TK gene) was added to the cells in the absence or
presence of excess transferrin (10 µg/ml) with or without 100 µM of
chloroquine (Sigma) (Fig. 2A
). Cells were cultured until
gene expression was assayed.
|
ß-Galactosidase gene transfection with adenovirus vector and
retrovirus vector
A recombinant adenovirus expressing nlacZ (Ad-LacZ) was
generated as described before (34)
. K562 cells were
incubated with Ad-LacZ at a multiplicity of infection (MOI) of 100 for
1 h at 37°C, then the transfected cells were cultured at 37°C
for another 2 days. A retrovirus plasmid encoding nlacZ gene (pRxZpN)
was provided by H. Hamada. A retrovirus producer cell 
CRIP was
cloned according to our previous report (35)
. K562 cells
were infected with this vector (8x104 CFU/ml)
for 4 h at 37°C as described previously (35)
.
Lipofection
BOSC 23 cells (5x105), which are known to
be highly transfectable (36)
, were transfected with 2 µg
of biotinylated GFP or nlacZ plasmids using Lipofectin (Life
Technologies, Gaithersburg, Md.) as described in the manufacturer’s
instructions. Lipofection of K562 cell with nlacZ plasmid was carried
out in the same manner.
Assays for expression of ß-galactosidase and HSV-TK
Transfection efficiency of ß-galactosidase was determined by
visually counting X-gal (37)
-positive cells under a
microscope. The activity of ß-galactosidase was measured by using
ß-galactosidase enzyme assay system (Promega, Madison, Wis.). HSV-TK
activity was estimated by viable transfectants with MTT tetrazolium
(Sigma) conversion assay (38)
after incubation for 5 days
in the presence of ganciclovir (GCV) (2–200 µM, Syntex, Palo Alto,
Calif.).
Pulse-chase study
Radioiodination of the Tf was carried out by the lactoperoxidase
method (11)
. The specific activity of Tf was 1.5
µCi/µg protein. The 125I-Tf-nlacZ DNA
conjugate was constructed as described above. Pulse-chase study of
125I-transferrin-DNA conjugate in K562 was
performed following the methods described earlier for
125I-Tf.
Intravenous (i.v.) administration of the Tf-GFP plasmid conjugate
to the mice bearing subcutaneously (s.c.) tumors
K562 cells (1x107) were s.c. injected
into the flank of BALB/c athymic nude mice (6-wk-old female, Charles
River Japan) after pretreating with a combination of
anti-ganglio-N-tetraosylceramide (anti-ASGM1) and sublethal radiation
to suppress their residual immune system, which may present a barrier
for functional engraftment of human cells (39)
The
polysera (250 µg; Wako Chemicals, Kyoto, Japan) were administered
intraperitoneally on day -1 and once every 7 days thereafter.
Sublethal (3 Gy) gamma radiation was administered from
137Cs source (Gammacell) on day 0.
Three weeks after inoculation, the K562 tumor grew to 
100
mm3. A total of 20 µg of Tf-GFP-plasmid
conjugate in 100 µl volume was injected through the tail vein. At
1 h and on days 3, 5, and 7 after injection, mice were killed, and
each organ and the tumor were dissected for further analyses.
Polymerase chain reaction (PCR) for GFP transgene and its mRNA
Each organ and the tumor was homogenized to extract plasmid DNA
by an alkaline lysis procedure. GFP gene was amplified using primers
5'-CTTCAAGGACGACGGCAACTAC-3' and 5'-ACTGGGTGCTCAGGTAGTGGTT-3', yielding
a 317 bp fragment. PCR was performed by 1 µg of extracted DNA in a
total volume of 50 µl at 95°C/30s, 62°C/30s, and 72°C/60s (40
cycles). The RNA was extracted from frozen tissue using RNeasy Mini Kit
(Qiagen, Chatsworth, Calif.) and treated with RNase free DNase (Life
Technologies) according to the manufacture’s instructions. cDNA was
synthesized using Super Script II RNaseH-reverse transcriptase (RT)
(Life Technologies) at 200 U/1 µg of RNA. PCR was performed as
described above. RNA of glyceraldehyde-3-phosphate dehydrogenase
(GAPDH) was PCR-amplified from each sample to demonstrate the integrity
of the cDNA synthesized using primers 5'-ATGGTGAAGGTCGGTGTGAACGGA-3'
and 5'-GCCTACATGGCCTCCAAGGAGTAA-3', yielding a 1002 bp fragment. PCR
products were electrophoresed on 2% agarose gel and visualized by
ethidium bromide staining.
TaqMan real time quantitative RT-PCR assay for mRNAs of GFP and
GAPDH
To quantitate the GFP and GAPDH mRNA in each organ and the
tumor, we performed TaqMan real time quantitative RT-PCR assay using
the ABI PRISM 7700 sequence detector system (Applied Biosystems, Foster
City, Calif.). The primer set of GFP has been described above. TaqMan
probe for GFP was 5'-CAAGCAGAAGAACGGCATCAAGGTGA-3'. The
GAPDH target primers were 5'-CTTCACCACCATGGAGAAGGC-3' and
5'-GGCATGGACTGTGGGTCATGAG-3'. TaqMan probe for GAPDH was
3'-CCTGGCCAAGGTCATCCATGACAACTTT-5'. Otherwise the procedure of TaqMan
RT-PCR assay followed our previous report (40
, 41)
.
Examination of GFP expression in the tissues
Resected tumor and muscle were embedded in optimal cutting
temperature compound (OTC compound, Miles, Elkhard, Ind.) and snap
frozen in isopentane chilled in liquid nitrogen. Cryostat sections 5
µ thick were prepared, and immediately examined and photographed
using a Bio-Rad MRC-1024 confocal imaging system equipped with a
krypton/argon laser.
Protocol for suicide gene therapy
For this study, three sets of experiments with severe combined
immune-deficient (SCID) mice (6-wk-old female, Charles River Japan)
were conducted. In all, there were nine groups with five mice per
group, each receiving pretreatment with anti-ASGM1 (day -1) and
radiation (day 0) following the methods described earlier. Experiment A
was conducted on one group to confirm the presence of metastasis. On
day 0, K562 cells (1x107) were injected into the
tail veins immediately after radiation pretreatment, and on day 22 the
mice were killed and examined.
In experiment B, three groups of mice were used to measure the efficacy of treatment with HSV-TK DNA Tf conjugate and GCV. On day 0, two groups received K562 cells (1x107) through injection into the tail veins immediately after radiation pretreatment. On day 22, a total of 20 µg of HSV-TK DNA complexed with Tf was injected into the tail veins of one of these two groups and once a day for the next 3 days, followed by 7 consecutive days of GCV treatment. The second group received no further treatment after receiving the K562 cells. The third group received neither K562 cells nor any other subsequent treatment. On day 35, the mice in all three groups were killed, and each organ and tumor was dissected for analysis.
The remaining five groups of mice were used for experiment C to assess the survival rates under various treatments. All five groups received K562 cells (1x107) through injection into the tail veins immediately after radiation pretreatment on day 0. Beginning on day 22, all five groups began a 4 day treatment, followed by a 7 day treatment under the distribution: group 1, Tf-HSV-TK, GCV; group 2, Tf-HSV-TK, PBS; group 3, HSV-TK plasmid, PBS; group 4, PBS, GCV; group 5, PBS, PBS.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 43)
. We therefore applied Tf-Tf-R complex extracted from human
placenta to the column with immobilized 5E7 and then liberated Tf from
Tf-R using an iron chelator (deferoxamine mesylate). Gel
electrophoresis of the eluate and Affi-Gel Hz beads dissolved in sample
solution clearly demonstrated bands corresponding to Tf, Tf-R and 5E7
antibodies (Fig. 1A
).
|
Stoichiometry of biotinylated Tf
Because unregulated biotinylation of Tf provides multiple
interacting sites for streptavidin, causing aggregation of the
conjugates (44)
, stoichiometry of biotinylation was
determined. The optimal stoichiometric ratio of biotin per Tf is 1:1;
by reacting 3 molar biotin with 1 molar Tf, we produced a 1.2:1 ratio
as revealed by HABA assay (31)
. To facilitate transgene
dissociation from the conjugate after endocytosis, for this
biotinylation we used a biotin reagent with s-s bonds linking to Tf so
that once in the endocytic pathway, the s-s bonds could be readily
cleaved by intracellular reductase (e.g., protein disulfide isomerase;
refs 27
28
29
), allowing the DNA to proceed to the nucleus
and express its gene. After application of thus biotinylated Tf, the
column (Affi-Gel Hz beads) was examined by SDS-PAGE, transblotting, and
chemiluminescent reaction using streptavidin-biotinylated alkaline
phosphatase (SA-AP). A band of biotinylated Tf, which reacted with the
SA-AP probe, was clearly observed with bands of Tf-R and the 5E7
antibody (Fig. 1A
).
Expression of biotinylated plasmids in BOSC23 cells
Photolabeling of DNA with biotin, which can occur at a ratio of up
to 1 biotin molecule per 200 bp nucleotides (45)
, may
bring about a loss of transgene function. BOSC 23 cells were
transfected with nlacZ plasmids biotinylated, using a disulfide biotin
reagent in various molar ratios and then examined for their expression
of ß-galactosidase by X-gal staining. The expression was drastically
impaired when the molar ratio of biotin/1 kbp of DNA exceeded 1.0.
(Fig. 1B
). Therefore, plasmids with a molar ratio of 0.5
were used for further experiments. Genes of HSV-TK and GFP were also
biotinylated in the same way.
Tf plasmid conjugate
After sequential application of streptavidin and biotinylated
plasmids (nlacZ, HSV-tK, GFP) to the column immobilized with
biotinylated Tf, unreacted streptavidin was blocked with excess biotin
and the conjugate was eluted with deferoxamine mesylate. In Fig. 1C
, agarose gel electrophoresis of Tf-nlacZ gene conjugate
is shown. No aggregates were detected by immunoblotting of the
conjugate, and ethidium bromide staining disclosed two discrete bands
corresponding to supercoiled and linear DNA. Both bands transblotted on
nitrocellulose membranes reacted with anti-Tf antibody and
anti-streptavidin antibody. Nearly identical patterns were observed
with genes of HSV-TK and GFP (data not shown). The overall yield was
40–50% of the applied biotinylated DNA. The final product was stable
for at least 14 days at 4°C. When the conjugate was constructed
without Tf-R affinity chromatography (i.e., constructed in an aqueous
solution), electrophoresis revealed only an aggregated band (Fig. 1C
).
Expression of ß-galactosidase in the Tf-nlacZ conjugate-treated
human tumor cell
K562 cells, which have a high expression of Tf-R (95% on FACS:
data not shown), were treated with Tf-nlacZ conjugate for various
incubation periods and examined for their expression levels of
ß-galactosidase by measuring enzyme activity (Fig. 2A
). At day 1 of incubation, some appreciable enzyme activity
(3.8±0.6 milli unit/µg protein) was already evident, which gradually
increased thereafter until reaching a peak (5.7±0.5 milliunit/µg
protein) at day 4. It should be noted that this transfection was
accomplished with almost equal efficiency as that with the aid of the
lysosomotropic agent chloroquine, which has been reported by others to
prevent lysosomal degradation (14
15
16
17
18)
of ligand-plasmid
conjugates, but because of its toxicity is only applicable for in
vitro studies. Upon addition of excess Tf (100-fold molar excess),
expression of the conjugate was significantly suppressed at days 1 and
2, indicating specific uptake of conjugate via Tf-R. However,
suppression of enzyme activity by excess Tf was not observed after day
3. This rather peculiar phenomenon may be explained by assuming that
conjugated Tf is a much poorer iron donor than native Tf; therefore,
the latter had become depleted of its iron by day 3 and no longer
effectively competed with the former. An alternative explanation is
that the conjugated DNA could be taken up by an another low affinity
receptor system, such as a scavenger receptor, that is not competed by
Tf. The specific uptake of the conjugate via Tf-R was further confirmed
by examining the relationship between levels of Tf-R expression and
enzyme activities in three different cell types transfected with this
conjugate (Fig. 2B
). M7609, a human colonic cancer cell
line; TMK-1, a human gastric cancer cell line; and HEL, a human
embryonic lung cell line (normal diploid cell), having Tf-R expressions
of 79%, 67% and 3%, exhibited ßgal activity of 4.9 ± 0.8
milliunits/µg protein, 4.0 ± 0.3 milliunits/µg protein, and
0.3 ± 0.2 milliunits/µ protein, respectively, at day 4. Thus
Tf-R mediated specific transfection by this conjugate was verified.
Comparison of transfection efficiency of this method with other
currently prevailing gene transduction means was then carried out at
their respective optimal conditions as indicated in Materials and
Methods. As shown in Fig. 2C
, our method yielded much higher
levels of ß gal staining and enzyme activity than those of
lipofection and retroviral transduction, but a slightly lower
expression of the enzyme than with the adenoviral method, which is
known to be highly cytotoxic and immunogenetic, precluding it from
systemic in vivo administration.
Intracellular kinetics of 125I-Tf-nlacZ conjugate in
K562 cells
To examine the intracellular behavior of
125I-Tf-nlacZ conjugate, a kinetics study
(12)
of 125I-labeled conjugate was
conducted in K562 cells (Fig. 3
). Radiolabeled conjugates on the cell surface disappeared rapidly
(within 1 min) and reciprocally appeared intracellularly, forming a
small peak at 510 min in a similar fashion as native Tf. Thereafter,
extracellular radioactivity in the medium increased, reaching a plateau
at 30 min. Approximately 63% of this extracellular radioactivity was
trichloroacetic acid (TCA) precipitable (nondegraded form of
conjugate), suggesting that some appreciable proportion of the
conjugates is repeatedly recycled, delivering genes to the cell.
|
GFP-gene expression in the tissues of nude mice i.v. injected with
the conjugate
To readily identify the gene product in the tissues in
vivo (46)
, we chose a GFP gene to be conjugated with
Tf. Since it is well accepted that human Tf binds to murine Tf-R with a
similar affinity as to human Tf-R (47
, 48)
and it is
difficult to obtain pure murine Tf in large quantities, we used a
xenobiotic model where a human tumor (K562) was s.c. inoculated;
subsequently, the human Tf-gene conjugate (20 µg DNA) was injected
through the tail vein (Fig. 4
). At 1 h after conjugate i.v. administration, GFP-DNA signals were
identified by PCR in virtually all tissues, most likely due to
nonspecific contamination of blood that carries the gene and the
potential alternative uptake pathway as suggested before. At day 3,
tumor showed a marked positivity. Also showing positivity, though
perhaps not to the same extent, were blood and bone marrow cells and
tissues such as liver, heart, brain, spleen, skin, muscle, and colon.
This finding is compatible with the previous notion that both normal
tissues consisting of rapidly dividing cells (5
, 49)
and
some nondividing cells, such as endothelial cells of brain vessels
(50)
, express Tf-R, though their magnitude of Tf-R
expression is generally much lower than in tumor tissues. By day 7, the
PCR signal in the tumor remained as a discrete band, but those in
normal tissues (except for bone marrow and muscle, which showed only a
very faint signal) became negligible (Fig. 4A
). The
erythroid precursors in bone marrow express appreciably high levels of
Tf-R. This expression, however, decreases within a few days as they
differentiate into more mature cells (51
, 52)
. The amount
of conjugate (PCR signal) internalized into the erythroid precursor
decreased accordingly (Fig. 4A
).
|
When RT-PCR products from these tissues were examined, the only
tissue (aside from tumor) where mRNA of GFP was evident was muscle, and
even then only to a minimal extent (Fig. 4B
). Muscle is a
unique tissue whereby transfected cDNA can persist as an
extrachromosomal plasmid by virtue of its prolonged post-mitotic state
as well as its multinucleated nature (53)
. Despite the
presence of PCR signal (DNA) in erythroid cells, its RT-PCR product
(mRNA) was almost undetectable as early as day 3 (Fig. 4B
),
possibly due to predominance of hemoglobin mRNA over other mRNAs in
mature erythroid cells.
These observations were further confirmed by TaqMan RT-PCR, which is a
highly regarded quantitative method of measuring mRNA (Fig. 4C
). GFP expression revealed by fluorescence microscopy,
however, was observed only in tumor tissue and not in muscle, since the
amount of internalized cDNA in muscle was assumed to be far less than
that of tumor and was not sufficient to give rise to detectable
proteins (Fig. 4D
, E
, F
, G
).
Anti-metastatic effect of systemically administered Tf-HSV-TK gene
conjugate
To verify the practical applicability of our method, we first
examined the tumoricidal activity of Tf-HSV-TK gene conjugate on K562
cells in vitro and then tested its anti-metastatic effect
in vivo. The in vitro experiment clearly
demonstrated that cells transfected with this conjugate were
significantly more susceptible to GCV treatment than nontransfected
cells (Fig. 5A
). In a first set of in vivo experiments, we
confirmed that K562 tumors xenographed through the tail vein of SCID
mice pretreated with anti-ASGM1 and radiation formed massive metastatic
lesions in the lung, liver, ovaries etc., by day 22 (data not shown).
In a second set of in vivo experiments, Tf-HSV-TK gene
conjugates were administered at days 22–25 through the tail vein of
the mice systemically inoculated with K562 cells by the same method as
mentioned above, followed by GCV injection on days 26–32. As shown in
Fig. 5B
, 5C
, in HSV-TK-treated mice, tumor
formations had clearly diminished as compared to those in control mice.
In a third set of in vivo experiments, a survival course of
similarly treated mice was elucidated. Mice treated with the conjugate
survived significantly longer than control mice (Fig. 6
).
|
|
Prolonged survival may occur partially because sublethal radiation blocks proliferation of normal tissues. Thus, after Tf-based suicide gene transfection, these tissues would be protected from the effects of the suicide gene product. Toxicity and possibly survival of the hosts could be affected.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 55)
have disclosed that a high level of foreign gene expression
can be achieved by simple dilution of naked DNA in a large volume of
physiological solution, which possibly allows DNA to bypass degradation
pathways involving serum nuclease and endocytosis. In our method,
dilution of the conjugate in such a large volume of solution is not
necessary (1/10 dilution used in the mentioned studies) for gene
expression, possibly because the conjugates that are detected by Tf
receptor are rapidly taken up by cells before they undergo digestion by
DNase in circulation and bypass the intracellular endocytosis pathway
via the Tf recycling process, escaping intracellular degradation. Since
there is no size limitation of the gene to be conjugated and no
apparent side effects in vivo (data not shown), this method
should be applicable to a wide range of anti-cancer genes including
suicide genes, anti-oncogenes, anti-angiogenesis genes etc. In
particular, suicide gene therapy may be a most suitable modality for
this method because prodrug, which actually triggers the death of gene
transduced cells, may be administered at certain intervals after
delivery of conjugates when the nonspecifically delivered gene in
normal tissue has been destroyed. Taken collectively, this method has
proven to be useful for efficient gene transduction in vivo
as well as in vitro gene delivery to tumor tissue.
|
ACKNOWLEDGMENTS |
|---|
Received for publication December 16, 1999.
Revision received April 10, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) antisense complementary DNA increase the sensitivity of a colon cancer cell line to adriamysine, cisplatin, melphalan, and etoposide. Cancer Res 56,3577-3582This article has been cited by other articles:
|
|
 |
|
  Z. M. Qian, H. Li, H. Sun, and K. Ho Targeted Drug Delivery via the Transferrin Receptor-Mediated Endocytosis Pathway Pharmacol. Rev., December 1, 2002; 54(4): 561 - 587. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Toyobuku, Y. Sai, I. Tamai, and A. Tsuji Enhanced Delivery of Drugs to the Liver by Adenovirus-Mediated Heterologous Expression of the Human Oligopeptide Transporter PEPT1 J. Pharmacol. Exp. Ther., June 1, 2002; 301(3): 812 - 819. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Lehtolainen, A. Taskinen, J. Laukkanen, K. J. Airenne, S. Heino, M. Lappalainen, K. Ojala, V. Marjomaki, J. F. Martin, M. S. Kulomaa, et al. Cloning and Characterization of Scavidin, a Fusion Protein for the Targeted Delivery of Biotinylated Molecules J. Biol. Chem., March 1, 2002; 277(10): 8545 - 8550. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Roulon, D. Coulaud, E. Delain, E. L. Cam, C. Helene, and C. Escude Padlock oligonucleotides as a tool for labeling superhelical DNA Nucleic Acids Res., February 1, 2002; 30(3): e12 - e12. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
-HindIII), lane 2:
nlacZ plasmid, lane 3: Tf-streptavidin-nlacZ gene conjugate, lane 4:
the conjugate constructed without Tf-R affinity chromatography, lane 5:
biotinylated Tf, lane 6 and 8: Tf-streptavidin-nlacZ gene conjugate,
lane 7: streptavidin. D) Schema of Tf-streptavidin-DNA
conjugate.
) and that of recycled conjugate precipitated
by adding 10% TCA (c; •) were determined. Cells
were then treated with 0.25 M acetic acid/0.5 M NaCl; radioactivity
released from the cell surface (a; •) and that
remaining in the cells (b; •) were also
determined.
