| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles |
Division of Endocrinology, Metabolism, and Molecular Medicine (E.J.L., F.M., T.K., J.L.J.) and Department of Microbiology/Immunobiology (B.T.), Northwestern University Medical School, Chicago, Illinois 60611
Address correspondence and requests for reprints to: J. Larry Jameson, M.D., Ph.D., Division of Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Medical School, 303 East Chicago Avenue, Chicago, Illinois 60611. E-mail: ljameson{at}nwu.edu
Abstract
Management of Cushing’s disease remains challenging, despite advances in its diagnosis and treatment. Here, we describe a strategy for targeting the expression of toxic genes to ACTH-producing tumor cells using adenoviral vectors. The POMC promoter was used to express either a marker gene (ß-galactosidase) or a toxic gene [herpes simplex virus thymidine kinase (TK)]. In ACTH-producing AtT20 cells, infection with recombinant adenoviruses containing the POMC promoter (AdPOMCGal; AdPOMCTK) led to high-level gene expression. Stereotactic injection of AdPOMCGal into the rat pituitary resulted in localized expression of the ß-galactosidase transgene in corticotrope cells. Cytotoxicity studies were performed using the TK-containing vectors and treatment with ganciclovir. AdPOMCTK caused greater than 95% cytotoxicity of AtT20 cells at a viral dose (multiplicity of infection, 5 plaque-forming units/cell) that induced minimal toxicity using control viruses. No cellular toxicity was seen using a nonpituitary cell line (T47D breast tumor cells). AtT20 cells transplanted into nude mice induced features of Cushing’s syndrome and were used as an in vivo model of ACTH-producing tumors. Injection of the AdPOMCTK virus caused significant regression of the transplanted AtT20 tumors. These studies suggest that the POMC promoter may provide a useful gene therapy strategy for the adjunctive treatment of pituitary tumors causing ACTH-dependent Cushing’s syndrome.
CUSHING’S DISEASE IS caused by ACTH-producing pituitary tumors. It can be cured by surgical resection in 75–80% of cases when the tumors are small (<1 cm) (1). Some surgical failures are the consequence of difficult localization of the adenomas, either preoperatively or at the time of surgery (2, 3). The cure rate for ACTH-producing macroadenomas is less than 50–60% (4). Medical treatments and radiation have been used as adjunctive therapies, but with incomplete efficacy (5, 6).
Recently, potential gene therapy strategies have been described for
pituitary adenomas using adenoviral vectors that use pituitary
hormone-specific promoters to selectively express marker genes or toxic
genes (7, 8, 9, 10, 11, 12). For example, targeted expression of the
herpes simplex virus thymidine kinase (HSV-TK) gene induced
cytotoxicity in GH- or glycoprotein hormone 
-subunit-producing tumor
cells following the administration of ganciclovir (GCV)
(9).
Based on these initial studies, we reasoned that a similar strategy could be applied to the corticotrope cell. However, there is less information about the sequence requirements and expression characteristics of the POMC gene, which encodes ACTH (13). Important features of a promoter for targeted expression using adenovirus vectors include: 1) a high level expression; 2) selective expression in the desired cell type; and 3) the promoter sequences must be small enough to fit into the adenovirus genome. Here, we describe experiments using adenoviruses containing the POMC promoter to target toxic gene expression to ACTH-secreting pituitary tumor cell lines and transplanted tumors.
Materials and Methods
Cell culture
AtT-20/D16v-F2 (hereafter referred to AtT20) cells, GH3 cells, and HEK293 embryonic kidney cell lines were obtained from the American Type Culture Collection (Manassas, VA). T47D breast cancer cells (provided by V. Craig Jordan, Northwestern University Medical School, Chicago, IL) were used as a negative control. AtT20 and GH3 cells were grown in DMEM/F12 containing 10% FBS. HEK293 cells were maintained in DMEM supplemented with 10% FBS. T47D cells were cultured in RPMI with 10% FBS. All media contained penicillin-streptomycin (100 U/mL to 100 µg/mL).
Generation of recombinant adenovirus vectors
The recombinant adenoviral vector has a backbone derived from
adenovirus type 5 (Ad5 309/356) in which the E3 regions have been
deleted. A cassette containing the POMC promoter and the
ß-galactosidase, or the HSV-TK, genes were inserted in place of the
E1 deletion. The structures of the recombinant adenoviral vectors are
shown in Fig. 1
. The rat POMC promoter
(provided by Dr. Malcom J. Low, Vollum Institute, Oregon Health Science
University, Portland, OR) consists of nucleotides -706 to + 64 from
the rat POMC 5'-flanking region. The recombinant adenoviruses
containing the HSV-TK or Escherichia coli ß-galactosidase
genes fused to the POMC promoters (AdPOMCTK and AdPOMCGal) were
generated as described previously (9). Recombinant
adenovirus containing the cytomegalovirus (CMV) promoter (654 bp of the
immediate-early enhancer region of the CMV gene) and the E.
coli ß-galactosidase gene (AdCMVGal) was used as a positive
control for expression and detection of ß-galactosidase activity.
Individual clones were purified and titrated using plaque assays as
described previously (14). The expression cassette was
confirmed in the adenoviral vectors by DNA sequencing.
|
Transduction efficiency of adenoviral vectors in pituitary tumor cell lines was tested using AdCMVGal and AdPOMCGal. Cells were plated in 12-well culture plates at a density of 2 x 105 cells/well. The next day, cell lines were infected by the addition of viral solutions to cell monolayers and incubation at 37 C for 1 h with brief agitation every 15 min. After the addition of new culture medium, infected cells were returned to the 37 C incubator and the media was changed 24 h later. For studies using a range of multiplicity of infection (MOI), triplicate wells were infected with each virus at MOI of 1, 5, 10, or 20 plaque-forming units (PFU) per cell for 48 h. Cells were fixed with 1% glutaraldehyde for 10 min, washed with phosphate-buffered saline (PBS; pH 7.4), and then incubated with X-gal substrate solution (10 mM potassium ferrocyanide, 10 mM potassium ferricyanide, 1 mM MgCl2, 0.2% NP40, and 0.1% X-Gal in PBS) at 37 C for 2 h to evaluate ß-galactosidase expression.
Triplicate wells of infected cells were also used to assay ß-galactosidase activity using O-nitrophenyl ß-D-galactopyranoside as a substrate (Sigma, St. Louis, MO). Culture media was aspirated, cell lysis solution was added, and lysates were mixed with the O-nitrophenyl ß-D-galactopyranoside substrate solution and incubated in 37 C for 2 h. The reaction was stopped with 100 µL of 1 M Na2CO3. Absorption was measured at 405 nm, and ß-galactosidase activity was calculated using a standard curve and expressed as mU/mg cellular protein.
Stereotactic injection of adenoviral vectors into the rat pituitary
Sprague Dawley rats (200–250 g) were anesthetized with ketamine (75 mg/kg), xylazine (4 mg/kg), and acepromazine (0.75 mg/kg) and placed in a stereotactic frame. After midline incision of the scalp, the vertex area was exposed, and two small openings were made using a dental drill according to the following coordinates: 2.5 mm rostral to ear–bar line of the frame and 0.4 mm lateral to sagittal suture. A stainless steel cannula (30 gauge) was inserted stereotactically through the opening 10 mm ventral to the dura to reach the anterior pituitary. The solution containing the adenovirus was infused into both lobes of the pituitary at a rate of 2 µL per minute via a Harvard apparatus compact infusion pump (model 975; Harvard, Millis, MS). The total infusion volume was 20 µL (5 x 108 PFU). The cannula was left in for 10 min after infusion to permeate the adenoviral solution into pituitary tissue. Rats were monitored for 4–6 h after recovery from anesthesia. Animals that exhibited signs of pain or distress were killed immediately. Morbidity or mortality related to stereotactic injection was uncommon (<5%). Animals were killed 5 days after injection, and the pituitaries were removed and embedded in paraffin blocks for immunohistochemistry. All studies involving the use of rats and nude mice were approved by the Northwestern University Medical School Animal Care and Use Committee.
Immunohistochemistry and terminal deoxynucleotidyltransferase-mediated UTP end labeling (TUNEL) assay
The cell specificity of ß-galactosidase expression was assessed by performing an X-gal enzymatic assay, followed by immunohistochemistry for individual pituitary hormones. Pituitaries were fixed for 1 h in 4% paraformaldehyde in pH 7.2 sodium phosphate buffer and incubated at room temperature overnight in X-Gal solution (0.01% X-Gal). After a 3-h fixation in buffered formaldehyde, the samples were embedded in paraffin and 5-µm sections were prepared. Immunohistochemical analyses were performed using the ABC kit (Vector Laboratories, Inc., Burlingame, CA). After deparaffinization, the specimens were treated with 3% hydrogen peroxide in absolute methanol and preincubated with horse serum (Vector Laboratories). Sections were incubated with primary antibodies specific for pituitary hormones: rabbit antihuman GH (prediluted; Zymed Laboratories, Inc., San Francisco, CA), rabbit antihuman ACTH (prediluted; Zymed Laboratories, Inc.), rabbit antihuman PRL (prediluted; Zymed Laboratories, Inc.), rabbit antirat LHß (1:250; National Hormone and Pituitary Program, NIDDK, NIH, Rockville, MD), and rabbit antirat TSHß (1:500; NIDDK). Incubation with primary antibodies was performed for 2 h at room temperature. After washing with Tris-buffered saline/0.025% Tween, staining was performed using biotinylated secondary antibodies and streptavidin-peroxidase (Vector Laboratories, Inc.) according to the manufacturer’s protocol. Nova-Red (Vector Laboratories, Inc.) or 3-amino-9-ethylcarbazole substrate (DAKO Corp., Carpinteria, CA) was used as a chromogen.
For the detection of ß-galactosidase expression in AtT20 tumors injected with AdPOMCGal, deparaffinized slides (5-µm section) were incubated with rabbit polyclonal anti-ß-galactosidase (5 µg/mL; CLONTECH Laboratories, Inc., Palo Alto, CA) for 1 h at room temperature. Staining was completed as described above.
To detect apoptosis in AtT20 tumors injected with adenoviral vectors, 5-µm sections of paraffin-embedded tissue were used. After deparaffinization, slides were placed in a jar containing 10 mM sodium citrate buffer (pH 6.0), irradiated in a microwave oven (1000 W) for 5 min, and cooled to room temperature. The TUNEL assay was performed using an In Situ Cell Death Detection Kit (Roche Molecular Biochemicals, Indianapolis, IN). DAPI was used for counterstaining. The images were obtained using a Nikon Eclipse E400 microscope (Nikon, Tokyo, Japan).
GCV sensitivity of AdPOMCTK-infected cells
The sensitivity of adenovirus-infected cells to GCVwas measured using a nonradioactive cell proliferation assay according to the manufacturer’s protocol (CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay; Promega Corp., Madison, WI). One day after plating 5 x 103 cells in triplicate wells of 96-well plates, adenoviral vectors were infected at different MOI (0, 0.1, 0.5, 1, 2.5, 5, 10, 100, and 500 PFU/cell). One hour after infection, increasing concentrations (0, 0.5, 2.5, and 5 µg/mL) of GCV were added and fresh medium containing GCV was added every 2 days; cell viability was assayed at day 4. A solution (20 µL) containing 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium/phenazine methosulfate was added to each well, and incubation continued at 37 C for 2 h. The absorbance at 490 nm was recorded using a Microplate plate reader (Emax; Molecular Devices, Sunnyvale, CA). The percentage of cell survival (mean ± SD) was calculated as the absorbance at 490 nm in the GCV-treated cells divided by that in the cells without GCV treatment. For long-term studies of growth in infected tumor cells, a MOI of 5 was used and 2.5 µg GCV was added to each well. Cell viability was assayed at days 0, 2, 3, 4, 6, 8 after adding GCV.
Treatment of tumor-bearing mice with adenoviral vectors in vivo
AtT20 cells (1 x 108) were injected into the flank area of adult (12 weeks of age) athymic female nude mice (Harlan-Sprague Dawley, Indianapolis, IN). Ten days after injection of AtT20 cells, tumors of about 0.3–0.5 cm size in diameter developed. To investigate the efficiency of gene delivery, AdPOMCGal (1.0 x 109 PFU) in a total volume of 50 µL dialysis buffer was injected into three sites of the tumor on two successive days. Tumors were removed and embedded in tissue-freezing medium (Tissue-TeK OCT compound; Miles Inc., Kankakee, IL) for the preparation of frozen sections 4 days after injection of AdPOMCGal. Sections (20 µm) of frozen tumor tissue were stained with X-gal solution as described previously (10).
To study the cytotoxic effect of AdPOMCTK on AtT20 tumors, nude mice were divided into 4 groups: 1) AdPOMCTK injection and GCV treatment (n = 6); 2) AdPOMCTK injection without GCV treatment (n = 6); 3) AdPOMCGal injection and GCV treatment (n = 6); and 4) injection of dialysis buffer used for adenovirus preparation with PBS treatment for GCV control (n = 6). Adenoviral vectors (1.0 x 109 PFU) were injected into a growing tumor from three directions on 2 successive days. The following day, 100 mg/kg GCV was administered ip once daily for 10 days. The size of the tumor was measured with calipers in three dimensions every 2 days. Tumor size is presented as mm3 (product of 3.14/6 x length x width x width). Tumors were removed and embedded in paraffin.
Results
Gene transfer efficiency of AdPOMCGal and AdCMVGal
The efficiency of adenoviral vector-mediated gene transfer to
corticotrope-derived AtT20 cells was assessed after infection with
AdPOMCGal or AdCMVGal adenoviruses. ß-galactosidase expression
was detected by X-gal staining of infected cells. Consistent with the
strong activity of the CMV promoter, 95–100% of AtT20 cells were
stained blue at 48 h after infection with AdCMVGal (MOI of 5
PFU/cell) (Fig. 2A). Using the adenovirus
containing the POMC promoter-driven ß-galactosidase (AdPOMCGal),
25–35% of AtT20 cells were stained at 48 h after infection using
10 PFU/cell (Fig. 2B). The percentage of these cells expressing
ß-galactosidase increased with time and reached 75–85% by 3–4
days after infection (Fig. 2C). A greater fraction of cells (85–95%)
were positive for ß-galactosidase expression using 20 PFU/cell (data
not shown).
|
).
|
To assess whether expression of POMC-driven ß-galactosidase was
specific for different pituitary hormone-producing cell types,
adenoviral vectors (5 x 108 PFUs) were
injected directly into pituitary glands using a stereotactic device.
After 5 days, the injected pituitaries were stained for
ß-galactosidase activity and immunohistochemical staining was
performed in the paraffin-embedded pituitary. High levels of
ß-galactosidase expression were seen in 5-µm sections of
pituitaries injected with AdCMVGal (Fig. 4A-a), whereas injection of AdPOMCGal
resulted in less intense and more scattered staining of
ß-galactosidase around the injection site (Fig. 4A-b). The number of
cells expressing ß-galactosidase in the AdPOMCGal-injected
pituitaries was about 10–15% of that in the AdCMVGal-injected
pituitaries (Fig. 4A). Immunohistochemical analysis for pituitary
hormones revealed that AdPOMCGal transgene expression was consistently
restricted to corticotropes. AdPOMCGal transgene expression was not
seen in somatotropes, lactotropes, gonadotropes, or thyrotropes (Fig. 4B).
|
Adenoviral vectors expressing the POMC-driven TK gene were used to
investigate the ability to confer cytotoxicity to different tumor cell
lines. Adenovirus infected cells were treated for 4 days with a range
of GCV doses to activate TK-mediated toxicity. Cell viability was
determined using a cell proliferation assay. Without viral infection,
doses of GCV higher than 10 µg/mL induced a partial cytopathic effect
(10–15% cell death) in AtT20 cells (data not shown). Infection with
the ß-galactosidase-expressing vectors served as negative controls.
In the absence of GCV, AtT20 cells infected with AdPOMCTK retained
viability, except at very high MOI (100 PFU/cell or greater). GCV
treatment of the AdPOMCTK-infected AtT20 cells induced striking
toxicity. For example, at an MOI of 1 PFU/cell, 85% of
AdPOMCTK-infected AtT20 cells were killed by treatment with 5 µg/mL
GCV (Fig. 5A). The cytotoxicity in AtT20
cells increased in proportion to the amount of virus and in response to
increasing amounts of GCV. Complete cell death occurred at doses of 5
PFU/cell and 2.5 µg/mL GCV (Fig. 5A). Little, or no, cytopathic
effect was seen with the control AdPOMCGal viruses, except at very high
doses of the vectors (>100 PFU/cell) (Fig. 5B).
|
, GCV-induced cytotoxicity was lower than that seen in
AtT20 cells, which actively express the POMC promoter. For example, at
a MOI of 1 PFU/cell, 85% cell death occurred in AtT20 cells infected
with AdPOMCTK and treated with 5 µg/mL GCV, whereas only 20% of cell
death occurred in GH3 cells at the same dose of AdPOMCTK (Fig. 5C).
These findings are consistent with the levels of ß-galactosidase
expression by the POMC promoter in GH3 cells (Fig. 3). In contrast,
T47D breast cancer cells infected with AdPOMCTK did not show GCV
sensitivity even at doses as high as 500 PFU/cell (Fig. 5C). These
results indicate that the AdPOMCTK virus confers dose-dependent GCV
sensitivity to pituitary cells, but not to breast cancer cell
lines.
Long-term effects of the adenoviral vectors on cell growth were also
assessed. Cells were infected with adenoviral vector (0 or 5 PFU/cell)
and cell viability was determined during an 8-day period with or
without treatment of 2.5 µg/mL GCV. In AtT20 cells infected with
AdPOMCTK (MOI of 5 PFU/cell), there was complete growth inhibition
after GCV treatment, whereas growth inhibition was not seen without GCV
treatment (Fig. 6). Growth inhibition was
not a nonspecific effect of viral infection or GCV as the growth of
AtT20 cells was not affected by treatment with AdPOMCGal (5 PFU/cell)
and/or GCV (2.5 µg/mL).
|
AtT20 cells were injected sc into nude mice to develop an in
vivo model for assessing the effects of the recombinant
adenoviruses. Tumors (71.9 ± 37.4 mm; Ref. 3)
developed 10 days after injection. Nude mice with At20 cell tumors
showed several Cushingnoid-like features, including thin skin, abnormal
fat deposition, and redistribution to the posterior neck, axilla, and
groin (Fig. 7A).
|
-a and b). Based on these results, the
same dose of adenoviral vectors was used to investigate the cytotoxic
effect of AdPOMCTK in vivo. Injection of the AtT20 cell
tumors with AdPOMCTK virus alone had no effect on tumor size (Fig. 7C).
However, treatment with GCV caused dramatic regression of the
AdPOMCTK-infected tumors (Fig. 7B). Apoptotic features including
disruption of cellular architecture and condensation of nuclei were
evident in tumors treated with AdPOMCTK and GCV (Fig. 8B-a). The TUNEL
assay was also positive in these tumors. (Fig. 8B-c). These features
were not seen in control tumors (Fig. 8B-b, d, and f).
|
). In contrast, all six tumors injected
with AdPOMCTK, but not treated with GCV, increased in size (152.0
± 40.1 mm3) (Fig. 9B). Tumors injected with the
control virus (AdPOMCGal) in the presence of GCV (153.1 ± 53.4
mm3) (Fig. 9C), or with vehicle instead of virus
(221.8 ± 95.8 mm3) (Fig. 9D), also
increased in size.
|
Here, we demonstrate the feasibility of targeted expression of a toxic gene (TK) to mouse pituitary tumor cells producing ACTH using a recombinant adenovirus containing the corticotrope-specific POMC promoter.
The expression of toxic genes has been widely considered as a gene therapy strategy for the treatment of tumors (15, 16, 17). In most cases, these strategies have used strong, ubiquitously expressed viral promoters to drive expression of the toxic genes. Although this approach has been encouraging, there are two important limitations: 1) most types of malignancies are highly invasive and/or metastatic, making it difficult to achieve clinically relevant tumoricidal effects; and 2) the viral promoters are active and, therefore, toxic in normal cells. Pituitary tumors represent a good model for the development of gene therapy approaches because they are usually benign neoplasms that are restricted anatomically to the region of the sella turcica. In addition, the tumors are characterized by the expression of a number of genes encoding receptors, transcription factors, and hormones that are highly restricted to the various pituitary cell types (18, 19, 20). Finally, the efficacy of adjunctive gene therapy might be monitored using radiologic (magnetic resonance imaging) as well as hormonal measurements.
Because Cushing’s disease remains difficult to cure in many patients, despite transsphenoidal surgery, we used the POMC promoter in an effort to target gene expression to ACTH-secreting cells. Initial efforts using the human POMC promoter were unsuccessful because of frequent DNA recombination during construction of the recombinant adenovirus and low level expression, at least using promoter sequences between -706 and +64 (data not shown). Although it may be possible to target expression using the human POMC promoter, additional studies will be required to identify the necessary promoter regulatory elements within the limitations of the carrying capacity of the adenovirus or other viral vectors. Studies using 0.8 kb of the rat POMC promoter were successful, however. This gene is expressed primarily in anterior pituitary corticotropes and intermediate lobe melanotropes. However, POMC is also expressed in a number of nonpituitary tissues including brain, testis, ovary, adrenal medulla, placenta, lung, circulating monocytes, and macrophages (21). This broad pattern of tissue expression is a theoretical disadvantage to the use of this promoter, although nonpituitary expression seems to be controlled by distinct gene regulatory regions (22).
The choice of promoter sequence is, therefore, important to maintain restricted expression. In transgenic mice, pituitary-specific expression of the POMC gene is controlled by cis-acting DNA sequences localized within 300 bp of the transcriptional start site (23). POMC promoter fragments extending from either -706 or -480 to + 64 bp resulted in higher levels of transgene penetrance in both corticotropes and melanotropes (24, 25, 26, 27). However, the -706 to + 64 bp POMC promoter was also active in a transiently transfected, enriched population of primary granulosa cells and in the testis of transgenic animals (25). These results suggest that promoter sequences active in the pituitary may also drive expression in some nonpituitary tissues. Studies using 4000 bp of POMC promoter sequence suggested a greater degree of pituitary specificity (28). The choice of promoter fragments for the current study was subject to practical limitations involving the amount of DNA that can be inserted into the modified adenoviral genome. With other viral vectors, it may be possible to examine the specificity of additional POMC promoter sequences. In addition to efforts to achieve pituitary selectivity, it is also interesting to consider whether vectors might be developed to mimic the "ectopic" expression of POMC as occurs in certain lung cancers, for example.
Using tumor cell lines, the rat POMC promoter seems to be relatively specific for expression in AtT20 cells. For example, AdPOMCGal expression was much lower in the nonpituitary T47D breast cancer cells. Consistent with the ß-galactosidase results, there was no GCV-induced cytotoxicity in AdPOMCTK-infected T47D cells. In GH and PRL-secreting GH3 cells, AdPOMCGal expression was about 10% of that seen in AtT20 cells. In transient expression studies using -480 to + 64 of the POMC promoter, others have also observed low POMC promoter activity in GH3 cells (29).
Using stereotactic injection of adenoviral vectors, we found that ß-galactosidase expression of the AdPOMCGal-injected pituitary was 10–15% that of the AdCMVGal-injected pituitary. This level of expression corresponds to the proportion of corticotrope cells in the pituitary gland. These results suggest that pituitary expression of the POMC promoter may be more specific in vivo than in pituitary tumor cell lines. This finding is not entirely unexpected because tumor cell lines are not fully differentiated and sometimes permit leaky gene expression. When AdPOMCGal (1 x 1010 PFU) was injected into the tail vein, no expression was detected in the livers of rats (data not shown). By comparison, a similar amount of AdCMVGal was strongly expressed in liver, lung, and other organs. Given the high affinity of adenovirus to the liver, these data confirm that the POMC promoter strongly restricts gene expression and should minimize collateral cytotoxicity. Moreover, local administration of the virus is also likely to limit its dissemination to other tissues. The rodent has a prominent intermediate lobe that contains abundant melanotropes. Although the POMC promoter is also active in melanotropes, we did not observe ß-galactosidase expression in the melanotropes, most likely because the stereotactic injection was localized primarily to the anterior lobe.
The cytotoxic effects of the HSV-TK gene can be activated by treatment with synthetic nucleosides such as acyclovir or GCV (17). In the setting of cancer gene therapy, GCV has usually been used because it is more effective and less toxic than acyclovir (15, 30). GCV is monophosphorylated by HSV-TK. An endogenous cellular kinase rapidly catalyzes subsequent phosphorylations steps (31, 32), leading to production of the highly toxic GCV triphosphate. The incorporation of the false nucleotide results in base pair mismatches, DNA fragmentation, sister chromatid exchange, and lethal genome instability (33, 34). In addition, GCV inhibits cellular DNA polymerases, further blocking DNA synthesis (35).
One of the benefits of using TK as a toxic gene is its ability to generate a bystander effect (BSE). For example, it has been shown that complete regression of tumors can be observed when only 10% of the tumor cells express TK (16). This effect reflects the fact that GCV triphosphate can be passed from HSV-TK-positive cells to the untransduced cells through gap junctions. Cell-to-cell contact is, therefore, required for the BSE-mediated cell killing. In AtT20 cells infected with 5 PFU/cell of AdPOMCGal, only 10–20% of cells express ß-galactosidase. However, infection with the same amount of AdPOMCTK caused greater than 95% cell death following GCV treatment, consistent with a BSE in these corticotrope cells.
When AtT20 cells were injected sc into nude mice, visible tumors developed within a week. Although typical Cushingnoid features seen in humans were not observed in mice, there was redistribution of fat, thinning of skin, and general emaciation in mice bearing AtT20 cell tumors. These findings are somewhat reminiscent of the clinical manifestations seen in patients with ectopic production of ACTH. A similar constellation of features has been described by Leung et al. (36) in AtT20 cell-bearing mice. In early experiments, these tumors caused high mortality because of rapid tumor growth and the metabolic consequences of Cushing’s syndrome. We ultimately used fully mature adult (12 weeks of age) mice to increase survival. In addition, adenoviral vectors were introduced when tumors were small (<0.4 cm in diameter) to allow sufficient time to complete the study. The dose of AdPOMCTK virus was based on the in vitro cytotoxicity studies. ß-galactosidase expression was seen in 15–40% of cells in tumors injected with AdPOMCGal. The administration of AdPOMCTK caused significant regression of AtT20 cell tumors after 10 days of GCV treatment. All other groups, including the AdPOMCTK injection in the absence of GCV, demonstrated continued growth of the tumors. Histologic analysis revealed widespread areas of cell death in the tumor mass treated with AdPOMCTK and GCV. The activated form of GCV is highly toxic for dividing cancer cells. Thus, the rapid growth of AtT20 tumors in nude mice, in addition to the BSE of HSV-TK, might contribute to the accelerated regression of tumors treated with GCV. A recent study has shown that the HSV-TK/GCV system under the control of the human PRL promoter was not effective in suppressing PRL levels in a rat model of lactotrope-hyperplasia (11). This model may more closely resemble human pituitary tumors, which normally grow very slowly. The application of the AdPOMCTK and GCV to transgenic animal models of melanotrope (37) or corticotrope tumors (38) may, therefore, provide an alternative approach for testing this adenoviral vector.
In conclusion, we demonstrate that adenoviral vectors containing the POMC promoter can efficiently express a toxic gene in pituitary ACTH-producing tumor cells in vitro and in vivo. These, or similar adenoviral vectors, have potential for gene therapy of ACTH-producing adenomas. Effective clinical application of this strategy must await further analyses of the efficacy and safety of the recombinant adenoviruses. However, this approach seems promising, particularly in the context of a treatment that might be used in conjunction with transsphenoidal surgery, which would allow the local delivery of virus to residual tumor when complete tumor excision is not possible.
Acknowledgments
We thank Dr. Malcom J. Low for providing rat POMC promoter and A. F. Parlow (NIDDK National Hormone and Pituitary Program) for providing antibodies for pituitary hormones. We also thank Jeffrey Weiss for critical reading and helpful discussions.
Footnotes
1 Supported in part by an institutional grant from the Northwestern
Memorial Foundation and by a Center of Excellence grant from Knoll
Pharmaceuticals. 
Received January 24, 2000.
Revised September 1, 2000.
Revised November 13, 2000.
Accepted November 16, 2000.
References
, ß and 
with
very different potencies and have unique mechanisms of action. Biochemistry. 34:2504–2510.[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  M Candolfi, G Jaita, D Pisera, L Ferrari, C Barcia, C Liu, J Yu, G Liu, M G Castro, and A Seilicovich Adenoviral vectors encoding tumor necrosis factor-{alpha} and FasL induce apoptosis of normal and tumoral anterior pituitary cells. J. Endocrinol., June 1, 2006; 189(3): 681 - 690. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. J. Lee and J L. Jameson Gene therapy of pituitary diseases J. Endocrinol., June 1, 2005; 185(3): 353 - 362. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Barzon, M. Boscaro, and G. Palu Endocrine Aspects of Cancer Gene Therapy Endocr. Rev., February 1, 2004; 25(1): 1 - 44. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
