| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
1 Division of Hepatology and Gene Therapy, School of Medicine, Fundacion para la Investigacion Medica Aplicada (FIMA), University of Navarra, Pamplona, Spain; 2 Jilin University, Department of Pathology, Changchun, China; and 3 Institüt für Mikrobiologie, Biochemie und Genetik, Friedrich-Alexander Universität, Erlangen-Nurnberg, Germany
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
(IFN-
), activation of T-lymphocytes and natural killer cells, and inhibition of angiogenesis (4, 5, 6)
. In animal models of hepatic tumors, intratumor administration of first generation adenoviral vectors expressing IL-12 can efficiently eradicate the neoplasm (6
, 7)
. This strategy, however, has demonstrated little efficacy in pilot clinical trials in patients with advanced primary and metastatic liver cancer, seemingly as a result of low transduction efficiency and very short duration of gene expression (8)
. The transfer of vectors encoding IL-12 to liver cells may have the advantage of generating high intrahepatic levels of IL-12 with lower systemic concentrations, thus increasing the therapeutic properties of the cytokine while lessening unwanted side effects. It is necessary, therefore, to develop vectors enabling tight regulation of IL-12 for its production at adequate levels and duration. Pharmacological control of gene expression can be achieved by using the so-called "on/off" regulatory systems. Most of them are composed of two expression units: (a) one bearing the gene of interest under the control of an inducible promoter; and (b) the other carrying a constitutively expressed chimeric trans-activator protein able to bind a specific drug and mediate the activation or repression of the inducible promoter activity. Depending on the nature of the inducer drug, five systems adequate for their use in mammalians have been described to date: the Tet on/off (9 , 10) ; the Pip on/off (11) ; and the antiprogestin-dependent (12) , ecdysone-dependent (13 , 14) , and rapamycin-dependent (15 , 16) gene switch systems. Among them, the Tet-on system exhibits the best features for applications in patients because: (a) the inducer doxycycline (Dox) is well tolerated in humans and has been widely used as an antibiotic; (b) Dox is liposoluble and has considerable tissue penetration; (c) it can be given orally and permits rapid gene induction/silencing switch in vivo in a dose-dependent manner (17) ; (d) this system has been studied in the context of numerous viral and nonviral vectors to regulate expression of various genes, and results are well documented (17, 18, 19, 20, 21, 22, 23) ; and (e) the level of gene expression in individual cells correlates directly with the dose of inducer, allowing a graded transcriptional response (24) . Despite these advantages, a main drawback of this system is its relatively high basal expression. This feature would limit its use for regulation of therapeutic proteins, which are potentially toxic.
In the Tet-on system, the reverse repressor of tetracycline operon (rtetR) was fused to the herpes simplex virus VP16 transcriptional factor to generate the reverse tetracycline-controlled trans-activator (rtTA), which interacts with the inducible promoter in the presence of tetracycline (or analogues as Dox) and activates transcription (10) . The inducible promoter is composed of seven copies of the Tet operator tetO fused to a cytomegalovirus minimal promoter region (CMVm). To reduce the basal expression, two kinds of chimeric repressors were developed: (a) one is the bacterial tetR protein fused to the repressor domain (KRAB) of the human Kox-1 protein (25) ; and (b) the other is the tetR fused to the human Mad-1 domain, which is involved in the recruitment of mSin3-histone deacetylase complex (26) . Both of these repressors interact with the inducible promoter in the absence of Dox and actively silence the basal expression in culture cells and in vivo (26, 27, 28) . Additionally, modifications on the original rtTA were made. One of the mutants, named rtTA2s-M2, binds with much lower efficiency to the tetO regions than rtTA in the noninduced state, and its VP16 domain was shortened to avoid cell toxicity (23 , 29 , 30) . Moreover, rtTA2s-M2 shows higher sensitivity to Dox than rtTA, and it is able to induce the same expression levels with 10 times less inductor dose (29) . Best results for gene regulation in vivo were obtained when combined repressor and modified trans-activator proteins together (28 , 31) .
In this study, we aimed at: (a) reducing the basal activity of the Tet-on system by making additional modifications without using repressor proteins; (b) constructing a liver-specific single vector carrying the necessary elements of the regulatory system in a short DNA sequence with minimal leakage and high level of inducible gene expression; and (c) using this vector to control expression of IL-12 in the liver and analyzing its antitumoral activity in animal models of metastatic liver cancer.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Plasmid Constructions.
pTRE-luc was generated by inserting the luciferase gene from pGL3 basic (Promega) into HindIII-XbaI site of pTRE2 (Clontech). CMVm promoter (positions 319438 in pTRE2) was substituted by Palb and Pcore by SmaI-HindIII fragment replacement. Palbm was generated by PCR amplification of a region lacking the 112 5'-end of Palb using forward primer: GAGGTAACCCGGGTTAATGATCTACA that has a SmaI site and following the same conditions as described (33)
. rtTA2s-M2 sequence was isolated from pUHrT621 (29)
with BamHI-EcoRI and placed under the control of EIIPa1AT (L1), EalbPa1AT (L2), and Phpx (L3) promoters. To construct single vectors carrying two genes oriented in opposite direction, the XhoI-XbaI 2,203-bp fragment containing the tetO7 (position 7318 in pTRE2)-Palb-luciferase cassette was excised from pO7Palb-luc and ligated into the unique SalI site of plasmids L1-rtM2, L2-rtM2, and L3-rtM2. To construct the tandem-oriented single plasmids, the XhoI-AseI 3,379-bp fragment was excised from pO7Palb-luc and ligated in L2-rtM2 linearized with MluI to generate pTonL2(T)-luc. The other two plasmids were generated from the previous one by excision of the NotI-MluI 3,568-bp Tet-inducible cassette and subcloning into L1-rtM2 or L3-rtM2 plasmids, digested with same enzymes. The hIL-12- and mIL-12-containing vectors were obtained by exchange of luciferase gene in pTonL2(T)-luc using HindIII and partial digestion with XbaI. The mouse IL-12-coding sequence was obtained from pBS/IL-12 (34)
by XhoI-SpeI digestion. The p35 and p40 subunits of hIL-12 were obtained by PCR of cDNA from lipopolysaccharide (10 µg/ml)-activated leukocytes. Primers used were: CTGCAGACCATGGGTCCAGCGCGCAGCCTCCT and CTGCAGTTAGGAAGCATTCAGATAGCTCGTCA for p35; and CCATGGG-TCACCAGCAGTTGGTCAT and GATATCTAACTGCAGGGCACAGAT for p40. The 675- (p35) and 994-bp (p40) genes were linked into pBSKII(+) (Stratagene) with internal ribosomal entry site (IRES) from pCITE-1 (Novagen). The SpeI-XhoI p35-IRES-p40 fragment was subcloned in pTonL2(T)-luc as indicated for mIL-12. Sequencing of DNA was routinely used for cloning verification.
Cell Transfections and Luciferase Detection.
Cells growing in 12-well dishes to 5060% confluence were transiently transfected by the calcium phosphate precipitation method. Plasmids were purified with Concert Nucleid Acid Purification System (Life Technologies, Inc.), and an equimolar amount of DNA corresponding to 1 µg of pTRE-luc was used. When single and separate plasmids were used in the same experiment, pBSKII(+) was included to adjust the amount of DNA. Always, 10 ng of pRL-SV40 (Promega) were added to each reaction to monitor transfection efficiency. Post-transfection culture medium (16 h) was replaced with DMEM 10% fetal bovine serum with or without Dox (Clontech). Cells were recovered at 48 h, washed twice with PBS, and lysed in 250 µl of Passive Lysis Buffer (Promega). Firefly and renilla luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega). Protein concentration was calculated using the Bio-Rad protein assay. Experiments were repeated at least three times. For in vitro mIL-12 and hIL-12 measurement in supernatants, cell transfection was done with 2 µg of pTonL2(T)-hIL12 and pTonL2(T)-mIL12 plasmids in six-wells plates. Western blot analysis to detect luciferase protein was performed exactly as described previously (33)
using goat antifirefly luciferase polyclonal antibody (Promega).
Animal Manipulation and Model of Liver Metastasis.
Each plasmid DNA (20 µg) resuspended in 1.6 ml of saline was injected into mice using the hydrodynamics-based procedure (35)
. Dox was administrated i.p dissolved in 200 µl of saline. Blood samples were obtained by retro-orbital bleeding, and serum was recovered by centrifugation at 10,000 rpm for 10 min and stored at 20°C until protein measurement. For the liver cancer model, 150,000 CT26 cells in 50 µl of saline were inoculated in two different areas of the main liver lobe. Mice were laparotomyzed under anesthesia, and tumor growth was monitored by measurement of two perpendicular diameters using a precision caliper. Survival was checked daily, and animals were euthanazed only if moribund.
Determination of IL-12 and Alanine Aminotransferase (ALT) Levels.
OptE1A human IL-12 (p70) and OptE1A mouse IL-12 (p70) ELISA kits (both from BD Bioscience PharMingen, San Diego, CA) were used for hIL-12 and mIL-12 determination. The limit of detection was 5 and 50 pg/ml, respectively. Values of mIL-12 obtained with PharMingen kit are usually higher than those obtained with another commercial kit from Endogen. ALT was measured using a Hitachi 911 Automatic Analyzer.
Statistical Analyses.
All analyses were done using SPSS version 9.0 software (Chicago, IL) with Ps of <0.05 considered to be statistically significant. The test used in each case is specified in the figure legends.
| RESULTS |
|---|
|
|
|---|
|
70 times when cotransfected with CMV-rtM2 in Hep3B (Fig. 1D)
Construction of Single Vectors for Liver-Specific Regulable Gene Expression.
To restrict expression to hepatocytes, we used a liver-specific promoter to direct rtM2. Because the amount of trans-activator produced can affect the leakiness of the system (39)
, we tested different promoters with graded activity to select the sequence, allowing more inducibility while keeping minimal basal expression. Taking advantage of our chimeric promoters characterized previously (33)
, we have chosen three of them, named EIIPa1AT, EalbPa1AT (human
-1 antitrypsin promoter fused to the hepatitis B virus EII enhancer or mouse albumin distal sequence, respectively), and Phpx (promoter of human hemopexin gene). EIIPa1AT (here, L1) is two times stronger than CMV in HepG2 cells, followed by EalbPa1AT (here, L2), with 70% of CMV activity and Phpx (here, L3) with <1% of CMV function (33)
.
Because the relative orientation of the expression cassettes can influence gene regulation and leakage of the Tet-on system (23
, 28)
, we constructed single vectors carrying both transcription units (L1, L2, or L3-rtM2 and pO7Palb-luc) in two different orientations: tandem or opposite (Fig. 2A)
. Fold induction of luciferase activity in the tandem/opposite constructs was
20/30, 35/40, and 1000/500 for the L1-, L2-, and L3-containing plasmids, respectively (Fig. 2B)
. Similar results were obtained in Hep3B cells (data not shown). These data showed that: (a) the basal, as well as induced expression, are directly proportional to the strength of the promoter used to direct rtM2 expression; (b) when both units are organized in tandem, gene expression was higher than when opposite oriented; and (c) fold induction was inversely proportional to the basal expression (maximal fold induction with L3-containing plasmids).
|
10 times lower compared with the induced protein in hepatic cells (Fig. 2, B and C)
10 times more active in 293 than HepG2 cells (33)
, we conclude that these plasmids are strongly attenuated in nonhepatic cells. To compare activity of the original and modified Tet-responsive promoters in these cells, we performed equimolar cotransfections with L1-rtM2 and pTRE-luc or pO7Palb-luc. About 40 times lower expression levels (with Dox) were obtained with pO7Palb-luc compared with the original pTRE-luc plasmid (Fig. 2C)
Regulation of IL-12 and Dox Dose Dependence Study in Vivo.
Among all vectors (Fig. 2A)
, we have selected pTonL2(T)-luc to control IL-12 expression, because: (a) the induced expression levels were high, whereas basal activity was acceptable; and (b) L2 promoter allows a more stable gene expression in the liver compared with L1 and L3 (33)
. Thus, we have substituted the luciferase gene in pTonL2(T)-luc by the coding sequences of the human and murine IL-12, resulting in plasmids pTonL2(T)-hIL12 and pTonL2(T)-mIL12, respectively. The response of both vectors to increasing doses of Dox was studied in HepG2 cells. Our results demonstrate that the system responds to very low doses of Dox (0.1 ng/ml) and that saturation takes place between 0.1 and 1 µg/ml (data not shown). Human IL-12 is devoid of most of its known biological functions in rodents (40)
, and in accordance to that, we were unable to detect IFN-
, a mayor mediator of IL-12 activity (41)
in mice treated with hIL-12 (data not shown). Therefore, we used this protein to study the regulation and kinetics of the inducible system. To this end, we transferred the pTonL2(T)-hIL12 vector to the liver of BALB/c mice using the hydrodynamics-based procedure (35
, 42)
. Fifteen days after plasmid injection, different doses of Dox were given by i.p. route. No cytokine was detected in serum in the absence of the drug. With the lowest Dox dose tested (0.5 mg/kg), hIL-12 levels were
100 pg/ml, and the maximum levels (28003500 pg/ml) were achieved with 5 and 50 mg/kg, respectively (Fig. 3)
.
|
|
Efficacy of mIL-12 Expression in the Liver for the Treatment of Liver Cancer.
To analyze the antitumor effect of IL-12 expressed in the liver parenchyma, we first developed a mouse model of multifocal liver metastasis. We established that 150,000 colon cancer CT26 cells injected in syngenic BALB/c mouse liver were able to develop a tumor of
200 mm3 in 7 days (Fig. 5A)
. Thus, we used these conditions in our subsequent studies. Plasmids pTonL2(T)-mIL12, pTonL2(T)-luc, or vehicle (saline) were administrated 1 week before injection of the tumor cells (to escape the initial period postvector injection that causes noninduced expression of mIL-12 attributable to hydrodynamic liver stress.4
One week after tumor implantation, engraftment of cancer cells was checked by laparotomy, and six groups were established: (a) saline (no plasmid); (b) Luc-D50:pTonL2(T)-luc + 50 mg/kg Dox; (c) IL12Dox:pTonL2(T)-mIL12 without Dox; (d) IL12-D2.5:pTonL2(T)-mIL12 + 2.5 mg/kg Dox; and (e) IL12-D50:pTonL2(T)-mIL12 + 50 mg/kg Dox. Administration of Dox was performed every 24 h for 10 days. Sixteen days after initiation of Dox treatment, mice were laparotomized to evaluate the progression of tumor growth. All animals that had received pTonL2(T)-mIL12 plus Dox showed no evidence of tumor (Fig. 5, A and B)
and long-term survival (Fig. 5C)
. On the contrary, animals given saline or Luc-D50 developed big tumors, and all animals died (or were euthanized) between days 37 and 52 after tumor implantation. Interestingly, four of nine animals from the IL12-Dox group rejected the tumor and survived the entire duration of the study. Serum levels of mIL-12 in this group were undetectable, as well as groups Saline and Luc-D50. Only groups IL12-D2.5 and IL12-D50 expressed the cytokine, which reached
350 and 1400 ng/ml at day 5, respectively (Fig. 6A)
. To evaluate the liver toxicity of the therapy, ALT levels were measured in the same serum samples. As shown in Fig. 6
B, mice treated with pTonL2(T)-mIL12 exhibited a slight increase of ALT only when administrated 2.5 or 50 mg/kg Dox. Control animals, as well as IL12-Dox, showed no significant differences in ALT levels.
|
|
| DISCUSSION |
|---|
|
|
|---|
The Tet-on system is composed of a transcriptional activator that binds to a minimal promoter in the presence of Dox and induces gene expression. Although this system has great potential for gene therapy purposes, its widespread use is limited by its relatively high basal activity. Two improved trans-activator proteins, rtTA2s-S2 and rtTA2s-M2, were recently developed to generate a system displaying lower background activity and higher window of Dox-dependent gene induction as compared with the original rtTA activator (23 , 29 , 31) . Among them, rtTA2s-M2 (rtM2) exerted better characteristics regarding Dox sensitivity and inducibility, although this protein still needs to be combined with a repressor protein (tTS) to minimize gene expression in the noninduced state (28 , 31) . When including repressor-coding sequences, the resulting increase of the DNA size of all of the system would limit its use in vectors with restricted cloning capacity. Moreover, the presence of a constitutive repressor protein could affect cellular functions by interacting with endogenous factors (25) . For these reasons, we attempted to modify the system to reduce basal expression in the liver avoiding repressor molecules.
To this aim, we have changed elements in the Tet-responsive promoter because we found that its intrinsic activity greatly contributed to leakiness of the system. In fact, the substantial activity in 293 cells of its CMVm region, which contains the TATA box and also several binding sites for cellular transcriptional factors (9
, 44)
, seemed to be responsible for the difficulties in generating adenoviral vectors carrying a proapoptotic factor regulated with the Tet-off system (45)
. Thus, by simply substituting CMVm by a liver-specific, but very weak, mouse albumin promoter (Palb), we found a 3550 times reduction in basal expression in human cell lines (Figs. 1
and 2
). A shorter version of the albumin promoter of 84 bp lacking the 5'-region showed higher activity in the same cell lines, indicating a possible role in promoter suppression of the deleted region.
With the aim of generating a single vector, six different constructs were made using three liver-specific promoters with graded activity (33) . The functional analysis of these vectors showed a direct correlation between the strength of the promoter that directs rtM2 and the level of noninduced and induced gene expression. In addition, we observed that basal expression correlated directly with the maximal induced protein level until saturation of the system. When the transcriptional units were oriented in the opposite orientation, lower reporter protein was obtained (as compared with tandem-oriented cassettes), probably by reduction of gene expression through transcriptional interference.
On the basis of the above data, we constructed a single vector with two expression units placed in tandem: (a) one for liver-specific expression of rtM2 using the L2 promoter; and (b) the other with the O7Palb promoter to control IL-12 synthesis. We saw no basal expression and a clear dose-dependent induction with Dox, which could be maintained for >2 months without any important reduction in the level of hIL-12. Moreover, continuous administration of the inducer for several days resulted in sustained serum levels of the protein during all this period. Administration of Dox after 3 months of plasmid injection resulted in a substantial reduction of inducible gene expression. This effect may be related to promoter silencing, loss of DNA by degradation, or simply turnover of transfected cells. The long-term persistence of gene products regulated by Tet systems in immunocompetent mice indicates that the immune response against the trans-activator is not the cause for declining gene expression with time (17 , 18 , 20 , 23 , 31 , 39) .
Using a murine model of metastatic colon cancer to the liver (a condition which in humans lacks curative therapy), we found a complete eradication of tumors in animals treated with TonL2(T)-mIL12 plus Dox (either at low or high doses) for 10 days, whereas all animals that received saline or similar vector-encoding luciferase showed progressive tumor growth and death before 2 months. The antitumor treatment occurred with acceptable toxicity because the rise of transaminases was mild, and all those mice showed long-term survival. Interestingly, four of nine animals which received TonL2(T)-mIL12 but no Dox also eliminated the tumor, suggesting that noninduced expression of mIL-12 within the liver at very low levels (not detectable in serum) does not prevent tumor engraftment but may be sufficient to activate a delayed antitumor immune response in some animals.
In conclusion, we have modified the Tet-on system by reducing the basal expression without adding additional coding elements and maintaining high inducibility in liver cells. On the basis of these modifications, we have constructed a single vector to express IL-12 within the liver. This vector enabled tight regulation of the cytokine and proved to be very efficient as tool for therapy of liver cancer. Although the plasmid here described could be used in the context of long-term expression viral vectors, progress in techniques for delivery of naked DNA to the human liver may facilitate the clinical use of these constructs, considering their easier high-scale production and better acceptance by regulatory agencies than the currently used viral vectors.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: M. Gabriela Kramer, University of Navarra-CIMA. Department of Internal Medicine, Pamplona, Spain. Fax: (34) 948-425700; E-mail: mgkramer{at}unav.es
Received 9/29/03. Revised 1/25/04. Accepted 2/ 5/04.
| REFERENCES |
|---|
|
|
|---|
production. Blood, 90: 2541-8, 1997.
) in immunocompetent mice mediated by a helper-dependent adenovirus vector. Gene Ther, 8: 1817-25, 2001.[CrossRef][Medline]
This article has been cited by other articles:
![]() |
S. Huard, R. T. Elder, D. Liang, G. Li, and R. Y. Zhao Human Immunodeficiency Virus Type 1 Vpr Induces Cell Cycle G2 Arrest through Srk1/MK2-Mediated Phosphorylation of Cdc25 J. Virol., March 15, 2008; 82(6): 2904 - 2917. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. W. Nebert Comparison of gene expression in cell culture to that in the intact animal: relevance to drugs and environmental toxicants. Focus on "Development of a transactivator in hepatoma cells that allows expression of phase I, phase II, and chemical defense genes" Am J Physiol Cell Physiol, January 1, 2006; 290(1): C37 - C41. [Full Text] [PDF] |
||||
![]() |
C. E. P. Goldring, N. R. Kitteringham, R. Jenkins, C. A. Lovatt, L. E. Randle, A. Abdullah, A. Owen, X. Liu, P. J. Butler, D. P. Williams, et al. Development of a transactivator in hepatoma cells that allows expression of phase I, phase II, and chemical defense genes Am J Physiol Cell Physiol, January 1, 2006; 290(1): C104 - C115. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |