GAO Jun², QIU Xiao-Fang², WU Zong-Di² and XIE Su-Qing¹

Abstract

AIM In order to express suicide gene specifically incolorectal carcinoma cells, we constructed the recombinant retroviral vector promoted bytranscriptional regulatory sequences of carcinoembryonic antigen (CEA) gene.
METHODS A 1.3 kb EcoRⅠ/ BamHⅠ fragmentof the CD gene was inserted into pUC118, and then the HindⅢ/ EcoRⅠ CD fragmentwas isolated. A 420 bp HindⅢ/ BamHⅠ fragment from pCEA 424/2CAT was released and subcloned into pUC118to harvest HindⅢ/ EcoRⅠ CEA promoter. These two fragments were ligated with thelinearized retroviral vector GINa using T4 DNA ligase.
RESULTS Restriction enzymeSalⅠ was used to digest the recombinants toidentify a positive clone A7 and a 1.7 kb band was observed on agarose electrophoresisgel. The recombinant retroviral vector, GICEACDNa was constructed.
CONCLUSION Success inconstruction of GICEACDNa provides a basis for colorectal carcinoma-specific gene therapy.

INTRODUCTION

Cytosine deaminase (CD), a keyenzyme of the DNA pyrimidine remedial pathway in fungi and some Escherichia coli, cantransform 5-fluorocytosine (5-FC) into 5-fluorouracil (5-FU) through deaminase, which ispoisonous to tumor cells. CD gene can not be expressed in eukaryotic cells, but it can beconverted into H-CD gene, an eukaryotic expression gene, through oligonucleotide-directedmutagenesis^[1] . Using carcinoembryonic antigen (CEA) gene 5＇-transcriptional regulatory sequences (5＇TRS) to promote H-CD gene specifically, which was thencloned into retroviral vector, GINa, we constructed the colorectal carcinoma specificretroviral vector GICEACDNa. The success of the construction lays a foundation for furtherresearch of the targeting gene therapy for colorectal carcinoma.

MATERIALS AND METHODS
Target genes, vectors and TRS

Plasmid LCDSN was cloned at ourlaboratory. Plasmid pCEA424/ 2CAT containing 5＇TRSof CEA gene was provided by Professor Zimmermann (Freiburg Albert-Ludwigs University,Germany). The retroviral vector LXSN was kindly provided by Professor Dusty Miller (FredHuchison Cancer Center, USA). The vector pUC118, host strain HB101, JM109 were kept at ourlaboratory.

Reagents

All restriction endonucleases wereproducts of Shanghai Sino-American Biological Product Co.RNAase, T4 DNA ligase wereproducts of Shanghai Shisheng Co.PGE 8000 was purchased from Sigma Co.

Construction of colorectalcarcinoma tissue-specific retroviral vector

The amplified pCEA424/ 2CAT was digestedwith HindⅢand BamHⅠand purified in low melting point agarose to isolate a 420 bp 5＇TRS fragment of CEA gene which was then subcloned into thepolyclone sites of pUC118. The recombinant was digested with HindⅢand EcoRⅠto harvestHindⅢ-EcoRⅠCEA TRS. The amplified LCDSN was digested with EcoRⅠ and BamHⅠ andpurified in low melting point agarose to isolate a 1.3 kb EcoRⅠ-BamHⅠCD genefragment which was then subcloned into the polyclone sites of pUC118. The recombinant wasdigested with HindⅢ and EcoRⅠto harvest EcoRⅠ-HindⅢCD genefragment. GINa was digested with HindⅢ, atwhich the dephosphorylation was performed using calf intestinal alkaline phosphatase withT4 DNA ligase. The HindⅢ-EcoRⅠCEA promoter and EcoRⅠ-HindⅢCD genefragment were ligated with the linearized and dephosphorylated GINa vector. to transfectthe plasmids. A few plasmids were prepared by using 20 clones selected. SalⅠwas employed to define the orientation of the insertion.The positive clones which had been defined by the appearance of 1.7 kb band on agaroseelectrophoresis gel were amplified. Finally, GICEACDNa was constructed.

RESULTS

CEA TRS fragment and CD genefragment were inserted into the polyclone sites of pVC118 separately (Figure 1). Therecombinant pUC118 was digested with the corresponding endonucleases to harvest HindⅢ-EcoRⅠ CEATRS and EcoRⅠ-HindⅢ CD. The recombinant retroviral vector, GICEACDNa, was constructedby inserting CEA TRS fragment and CD fragment into the polyclone sites of GINasimultaneously (Figures 2,3).

Figure 1 Identification of the positive clones resulting from separate insertion of CD gene fragment and CEA gene fragment into pUC118.
1:1.3 kb CD gene fragment resulting from digestion of recombinant pUC118 with EcoRⅠand BamHⅠ.
2:λDNA HindⅢ marker

Figure 2 Identification of the positive clones resulting from simultaneous insertion of CD gene fragment and CEA gene fragment into GINa with SalⅠ.
1:1.7 kb fragment resulting from digestion of GICEACDNa with SalⅠ. 2:λDNA Hind Ⅲ marker.

Figure 3 Scheme of construction of G1CEACDNa.

DISCUSSION

The experience in chemotherapy fortumors indicates that 5-FU is the best choice for the treatment of colorectal carcinoma.Its killing effect in colorectal carcinoma cells is affirmative. However, the dosage of5-FU is greatly limited because of its side effects on normal cells, and it can only beused as supplementary therapy. Although mammalian cells do not contain CD gene, it can beexpressed in eukaryotic cells through oligonucleotide-directed mutagenesis^[1] . Theanti-tumor effects could be expected when the tumor cells were transferred by CD gene andexposed to 5-FC^[2,3] .
Through mediation of retroviral vector, CD gene can be expressed in both tumor cells andnormal cells, so it is far from enough to only express CD gene in eukaryotic cells. Thekey of gene therapy for tumor is targeted expression. Targeting suicide gene into tumorcells can not only increase the efficiency of gene transfer, but also protect the normalcells against the suicide gene. In this way, much more prodrugs can be used to kill tumorcells to a maximum. CEA is a tumor-associated marker especially for advanced carcinoma, sotranscriptional regulatory sequences of CEA have been used to promote tissue-specificexpression of suicide genes for tumor-specific gene therapy^[4] . Schrewe etal^[5] cloned CEA 5＇TRSfrom human genome, and considered that the transcriptional regulatory activity of pCEA424/ 2CAT was higher thanthat of other sequences. Taking pCEA424/ 2CAT as colorectal carcinoma-specific promoter, weconstructed GICEACDNa by cloning H-CD gene into polyclone sites of GINa. Up to date, theincasement and transfection have been completed. The preliminary results indicate thatLoVo cells transfected by colorectal carcinoma-specific H-CD gene are more sensitive to5-FC than cells transfected by ordinary vector free of CEA transcriptional regulatorysequences. To improve colorectal carcinoma tissue-specific gene therapy, we will carry outpreclinical research on nude mice bearing tumor.

ACKNOWLEDGMENTS We would like to thank Professor Zimmermann (Institute ofImmunology, Albert-Ludwigs University, USA) for providing us the pCEA424/ 2CAT.

¹ Department of General Surgery,Changhai Hospital, Second Military Medical University, Shanghai 200433, China
² Department of Microbiology, Institute of Basic Medical Sciences, SecondMilitary Medical University, Shanghai 200433, China
³ Department of General Surgery, Changzheng Hospital, Second Military MedicalUniversity, Shanghai 200433, China
CUI Long, MD, male, Chief of Department of General Surgery, Changhai Hospital, SecondMilitary Medical University, having 6 papers published.
Supported by National Natural Science Foundation of China, No.39600142
Correspondence to: CUI Long, Department of General Surgery, ChanghaiHospital, Second Military Medical University, Shanghai 200433, China
Received 1997-01-30 Revised 1997-11-02

REFERENCES

1 Austin EA, Huber BE. A first step in the development of genetherapy for colorectal carcinoma: cloning and expression of Escherichia coli cytosine deaminase. Mol Pharmacol, 1992;43(11):380-391
2 Mullen CA, Coale MM, Lowe R, Blease RM. Tumors expressing the cytosine deaminase suicidegene can be eliminated in vivo with 5-fluorocytosine and induce protective immunity towild type tumor. Cancer Res, 1994;54(3):1503-1506
3 Huber BE, Austin EA, Good SS, Knick VC, Tibbels S, Richards CA. In vitro antitumoractivity of 5-fluorocytosine on human colorectal carcinoma cells genetically modified toexpress cytosine deaminase. Cancer Res, 1993;53(19):4619-4626
4 Richards CA, Austin EA, Huber BE. Transcriptional regulatory sequences ofcarcinoembryonic antigen: identification and use with cytosine deaminase fortumor-specific gene therapy. Hum Gene Ther, 1995;6(7):881-893
5 Schrewe H, Thompson J, Bona M, Hefta LJF, Maruya A, Hassauer M et al. Cloning of thecomplete gene for carcinoembryonic antigen: analysis of its promoter indicates a regionconveying cell type-specific expression. Mol Cell Biol, 1990;10(7):2738-2748 (CUI Long¹, CAO Guang-Wen², WANG Yuan-He², TU Yue¹, MENG )