WO1994024277A1 - Protection of human bone marrow from high dose antifolate therapy using mutated human dihydrofolate reductase dna - Google Patents

Abstract

This invention provides a DNA vector which comprises DNA encoding a mutant, antifolate resistant, dihydrofolate reductase inserted into a site within the vector, the presence of which site is not essential for replication of the vector. This invention also provides bone marrow cells which comprise the above vector. Finally, this invention provides a method for reducing the toxic effects of antifolate therapy in a subject which comprises replacing the subject's hematopoietic cells with hematopoietic cells which comprise the above-described vector so as to reduce the toxic effects of antifolate therapy in the subject.

Description

PROTECTION OF HUMAN BONE MARROW FROM HIGH DOSE ANTIFOLATE THERAPY USING

MUTATED HUMAN DIHYDROFOLATE REDUCTASE DNA

The invention described herein was made in the course of work under grant number CA-08010 from the National Institute of Health, U.S. Department of Health and Human Services. The U.S. Government has certain rights in this invention.

BACKGROUND OF THE INVENTION

Throughout this application various publications are referenced by arabic numerals within parentheses. Full citations for these publications may be found at the end of each series of experiments. The disclosures of these publications in their entireties are hereby incorporated by reference into this application in order to more fully describe the state of the art to which this invention pertains.

Dihydrofolate Reductase and Methotrexate

Dihydrofolate reductase (DHFR, 5,6,7,8-tetrahydrofolate: NADP + oxidoreductase, EC 1.5.1.3) catalyzes the NADPH- dependent reduction of dihydrofolate to tetrahydrofolate, an essential carrier of one-carbon units in the biosynthesis of thymidylate, purine nucleotides, serine and methyl compounds (Blakly,1969, Figure 1). It is an essential enzyme in both eukaryotes and prokaryotes.

In rapidly dividing cells, the inhibition of DHFR results in the depletion of cellular tetrahydrofolates, inhibition of DNA synthesis, and cell death. Folate analogs such as methotrexate (MTX) inhibit DHFR and are thus used as antineoplastic agents. A powerful inhibitor of DHFR, methotrexate (MTX) (Figure 2A) , has become a widely used antineoplastic agent in the clinic (Bleyer et al., 1978).

The clinical value of MTX treatment is limited by two problems that represent the major obstacles to the effective treatment of neoplastic disease with MTX as well as with most other antineoplastic agents: the rapid development of resistance in tumors and toxicity in normal tissue.

Mechanisms of MTX resistance

Resistance to MTX can be either natural or can be acquired after initial response to the drug. It has long been observed that some types of tumors are responsive to MTX treatment while others are intrinsically resistant to the treatment. One possible explanation is the differences in the ability to transport MTX into the cell (Bertino et al., 1985) . An intrinsically poor ability to transport the drug into the cell may confer natural resistance to MTX. After entering the cell MTX is converted intracellularly to polyglutamylated forms by the folylpolyglutamyl synthetase. Compared with the non- polyglutamylated form of MTX, the polyglutamylated form is less ready to be effluxed from the cell, more likely to retained in the cell, and therefore may be accumulated at higher intracellular concentrations. The polyglutamylated forms also have higher affinity for other enzymes in the 1-carbon folate-dependent pathway (Chabner et al., 1985) and can be stored in the cell to exert its cytotoxicity when the cell enters the S-phase (McGuire et al., 1985). A decrease in polyglutamylation has been correlated with the natural MTX resistance of certain tumors (Curt et al 1985; Pizzorno et al., 1989; Li et al. , 1989) .

Amplification of the DHFR gene resulting in increased levels of the enzyme has been identified as one mechanism of acquired MTX resistance (Carman et al., 1984; Horns et al., 1984). Defects in MTX transport and polyglutamylation have also been found responsible for some clinical cases of acquired MTX resistance as well as in experimental systems (Sirotnak et al., 1981, Pizzorno et al., 1988) .

Mutations in the target enzyme of DHFR resulting in a decreased affinity for MTX is a mechanism for acquired resistance to MTX (see review by Simonsen, 1986) . Several mutant DHFRs have been characterized in the past few years. Some mutant DHFR enzymes have mutations in the active site of the enzyme which results in both reduced affinity for antifolate as well as a decrease in catalytic efficiency of the mutant enzyme as compared to the wild type protein (Thillet et al., 1988; Schweitzer et al., 1989; Prendergast et al., 1989).

The firεt mutant murine DHFR (mDHFR) was cloned from a MTX resistant cell line 3T6-R400 (Simonsen and Levinson,

1983) . The mutant DHFR contained a G to T point mutation at nucleotide 68 resulting in a Leu to Arg change at residue 22. Leu 22 is normally involved in hydrophobic contacts with the substrate or inhibitors and play a critical role in the function of DHFR. The presence of a charged residue at this position considerably reduced the catalytic activity of the enzyme and the binding of

MTX. Arg would protrude in the active site more than the wild-type Leu residue and could hinder the positioning of inhibitors either by its steric effect or by allowing the penetration of water molecules in the active site. The mutant enzyme exhibited a 70 fold reduction in catalytic efficiency and a 7.5 x 10⁵ fold increase in MTX Ki as compared to the wild type enzyme (Haber et al., 1981; Thillet et al., 1988). When the cDNA of the mutant enzyme was transfected into the parental cells, the mutant enzyme was able to confer MTX resistance to the transfected cells which also contain the wild type DHFR. The ability of the 3T6 enzyme to act as a dominant selective marker has been demonstrated in murine, hamster and human cells (Simonsen and Levinson, 1983; Isola et al., 1986; Banerjee et al., 1994; see review by Simonsen, 1986) . The first mutant human DHFR (hDHFR) was isolated from a MTX resistant HCT-8 human colon carcinoma cell line (Srimatkandada, et al., 1989; Schweitzer, et al., 1989). The phenylalanine at residue 31 of the hDHFR was replaced by a serine in this MTX resistant cell line. The Serine 31 mutant has a 2 fold decrease in catalytic efficiency and a 100 fold increase in MTX Kd. This mutant enzyme was able to confer MTX resistance to cells containing wild-type DHFR when the cDNA of the mutant hDHFR was transfected into these cells (Banerjee, et al., 1994) . Molecular modeling studies have shown that Phe at position 31 interacts with the p-aminobenzoyl gluta ate portion of MTX or folate (Oefner et al., 1988). Thus substituting a large hydrophobic group with a small hydrophilic group has profound effects on MTX binding. Another Phe occurs at position 34 and is also an active site residue, which makes van der Waals contact with peteridine portion and a part of the p-aminobenzoyl group of the ligands. Site directed mutagenesis at this residue with a Ser substitution generated a hDHFR mutant which has a 3 fold reduction in catalytic efficiency and a 8 x 10⁴ increase in MTX Kd (Schweitzer et al., 1989). Other mutations in the DHFR were reported in experimental systems to facilitate the study of the precise nature of the mutation (see review by Schweitzer et al., 1990).

MTX toxicity

Anorexia, progressive weight loss, bloody diarrhea, and leukopenia are the outstanding features of lethal doses of MTX. The major lesions occur in the intestinal tract and bone marrow. Swelling and cytoplasmic vacuolization of the ucosal cells of the intestinal epithelium is followed by desqua ation of epithelial cells, extrusion of plasma into the lumen of the bowel, and leukocytic infiltration of the submucosa. Terminally, the entire intestinal tract exhibits a severe hemorrhagic desquamating enteritis. Degeneration of bone marrow develops rapidly. Proliferation of erythroid precursors is inhibited, and significant proportions of primitive erythroid elements have the appearance of megaloblasts. Rapid pathological alteration in yelopoiesis also occurs, and within a few days the bone marrow becomes aplastic. The disturbance in hematopoiesis is reflected in the circulating blood by a marked granulocytopenia and reticulocytopenia and a moderate ly phopenia (Goodman and Gilman, 1980) .

βene Therapy: Current Status

Gene therapy is defined as the transfer of genetic material into the cells of an organism to treat disease. There are many potential applications of this technique to the treatment of numerous hereditary diseases caused by defects in single genes (see review by Miller, 1990) . In addition, gene therapy may be useful for acquired diseases, such as cancer or infectious disease. Achievement of efficient gene transfer and persistent gene expression is the major focus of current research.

Specific gene therapy has been accomplished in cultured cells by homologous recombination of added DNA with endogenous sequences to target genes to specific sites within the genome (Smithies et al., 1985; Thomas and Capecchi, 1987). While this technique has promise for ultimate application to gene therapy, practical considerations, such as the finite life span of normal somatic cells or the inability to isolate or grow the relevant transplantable cells, presently limit its use.

An alternative to gene replacement is the addition of genes to correct a disease, namely gene addition therapy. Gene addition therapy can engineer a cell to express a new gene which the cell does not normally express. This method is currently the most practical approach to gene therapy due to the development in methodology for highly efficient gene delivery with retroviral vectors (see below) . It has been found to be useful in application involving acquired as well as hereditary diseases (Sarver et al., 1990; Sullenger et al., 1990; Gansbacher et al., 1990, see review by Anderson, 1992) . One potential problematic aspect of gene addition therapy is the random insertion of genes into the genome which may lead to the inappropriate expression of the inserted gene or the genes near the insertion site.

Different methods of gene transfer have been employed in the past. A method of DNA transfection employs purified DNA co-precipitated with calcium phosphate or dextran sulfate and brought into direct contact with the cells. The precipitated DNA on the cell surface is then endocytosed into the cells by uncharacterized pathways (Wigler et al., 1977). The efficiency of the transfection is very low, a maximum 1% of the cells will have incorporated the transferred DNA. Usually multiple copies of the gene in tandem repeats are integrated into the host genome, which may result in uncontrollable overexpression of the transfected gene, or interruption of the normal chromosome structure. The host range of the DNA transfection is limited to a small number of cultured cell lines, while a majority of primary culture cells either can not stand the toxicity of the method or is not susceptible to it. A few other techniques have been developed using physical means to introduce genes into cells: protoplast fusion (Schaffner, 1980) , in which bacteria containing recombinant DNA are fused with eukaryotic cells, resulting in the transfer of DNA from the cytoplasm of the bacteria into the host cell; lipofection (Feigner et al., 1987), in which the positively charged lipids in liposomes complex with DNA and the lipid-DNA complex fuses with plasma membranes and transfer the DNA into the cells; and electroporation (Potter et al., 1984), in which DNA is electrophoretically transferred across the host cell membrane into the cell though pores open up by the electric field. These techniques broaden the host range of susceptible cellε, but still have a very low efficiency of gene transfer. Microinjection, in which DNA is injected directly into the nucleus of the cell, results in stable integration of DNA in a large percentage of injected cells although the method is very time consuming and the number of transformed cells is limited by the cells that can be injected.

Gene transfer procedures based on SV40, polyoma, adenovirus vectors, and vaccinia based vectors lead to efficient but transient expression of the transduced gene. Bovine papillomavirus (BPV) (Sarver et al., 1982; Dimaio et al., 1982), and Epstein-Barr virus (EBV) (Yates et al., 1984) based vectors have only limited host ranges even though they result in stable expreεεion of the tranεduced gene carried aε an epiεome in multiple copies per cell. Adeno aεsociated viruε (AAV) baεed vectorε (Hermonat and Muzyczka, 1984; Tratεchin et al., 1985) integrate into the chromosome of the host cell but the full potential of this system needs to be further explored. These viral vectors have not been shown to transduce hematopoietic stem cells effectively (see review by Karlsson, 1991) . A new system for delivering genes to cells, which relies on an antibody molecule and a chain of amino acid units to hook DNA to the outside of adenovirus, has been reported lately (Curiel et al., 1992) . But the stability of the expression of the transduced gene and the ability of the system to express the transduced gene in vivo remains to be seen.

In the past several years, retroviral-mediated gene transfer, in which the genes are delivered into the cells by retroviruses, has emerged as superior to other techniques explored in gene therapy.

Retroviral Vector Mediated Gene Transfer

Retroviral life cycle

Retroviruses are animal viruses which contain a viral RNA genome which is replicated through a DNA intermediate. Moloney murine leukemia virus (MoMLV) is an ecotropic murine leukemia retrovirus which replicates well in only mouse and rat cells. The retroviral virion contains two copies of the retroviral RNA genome (Kung et al., 1976; Bender and Davidson, 1976; Bender et al., 1978) asεociated with the gag and pol gene productε in an icosohedral viral core structure which is surrounded by a lipid bilayer (derived from the previously infected host cell) . The viral encoded env gene products are embedded in the lipid bilayer (Varmus and Swanstrom, 1984) . The interaction of env protein molecules of the virion particle with a cell εurface protein on the target cell membrane reεults in the penetration of the virus into the cell (Figure 3) . After penetration the viral RNA genome is released into the cytoplasm and is reverse transcribed into a double-stranded DNA form by the viral encoded RNA dependent DNA polymerase, the reverse transcriptase (Baltimore, 1970; Temin and Mizutani, 1970) . This viral DNA migrates to the nucleus and integrates into one of the host's chromosomes to form the provirus. This integrated provirus iε the DNA template responsible for the expression of the viral gag, pol and env genes as well as the virion RNA (Varmus and Swanstrom, 1984) . In the infected cell the viral RNA is preferentially packaged into the virion particles. This specificity is mediated by an RNA sequence on the viral RNA called the packaging signal (Mann et al., 1983). Integration of the viral genome into the cell chromosome and the formation of subsequent virus usually has no deleterious affect upon the host cell. Thus cells harboring an unrearranged MoMLV provirus are normal and healthy, and continually secrete progeny virus into the surrounding medium.

MoMLV based retroviral vectors

The MoMLV genome encodeε three genes, the gag, pol and env genes, whose protein products are needed in trans for the replication of the virus, as well as several DNA and RNA elements required in cis for the replication of the virus. These cis elements include: the viral long terminal repeats (LTRs) which are required for transcription, transcription termination and polyadenylation; the viral RNA packaging signal which is required for efficient packaging of the viral RNA into virions; and primer binding sites (PBS) required for reverse transcription of the viral RNA to DNA.

The basic principle of a MoMLV based retroviral vector is to remove the εequenceε of the genome which are required in trans and replace them with foreign εequenceε of intereεt, while retaining all cis sequences necessary for viral replication. The hybrid DNA iε then introduced into specially designed packaging cells, which harbor a retrovirus defective in cis function. Its RNA cannot be encapsulated into a virion but it can express all the viral proteins and is therefore able to complement the trims functions missing in the incoming hybrid vector DNA. The vector DNA is then reverse transcribed into a corresponding RNA which iε encapεulated into a retrovirus virion, infectiouε but replication defective (Temin, 1986; Gilboa, et al., 1986). Such packaging can generate viruε containing vector RNA with a fairly high titer of up to IO⁶ infectiouε unitε/ml (Armentano et al., 1987; Markowitz et al., 1988a, 1988b). The packaging of a retroviral vector in an amphotropic baεed packaging cell line allows for the generation of amphotropic viruε able to infect a wide range of cell types. Through the efficient viral infection process the foreign gene iε inserted into the cell chromosome as if it were a viral gene (Figure 4) .

Various retroviral vector designε have been utilized in an attempt to increaεe the titer of the vector containing viruε coming from a packaging cell line aε well aε to increaεe the fidelity of expreεεion of the tranεferred geneε after infection (see review by Gilboa, 1987) . Among them is the design of vectorε with internal promoterε (VIP) (Figure 5A) . In theεe vectorε a selectable gene is expressed from the viral LTR promoter. The gene of interest is fuεed to another DNA fragment containing a promoter which is reεponsible for its expression. These vectors posεeεε the flexibility of chooεing a promoter to express the tranεduced gene moεt appropriate for a particular target cell (Enerman and Temih, 1984; Miller, et al., 1984). However, poεitioning a tranεcription unit within another active tranεcription unit often leads to the occlusion of the internal tranεcription unit (Cullen et al., 1984). Therefore placement of promoterε and genes within the retroviral vector LTR initiated transcription unit may reduce their expression. The N2 vector is a VIP type of retroviral vector based upon the MoMLV. It contains the firεt 418 baεe pairs of the gag coding sequences, as well aε the Neo resistance gene. The neo resistance gene appears to be expressed by a cryptic splicing of the vector RNA (Armentano et al., 1987). The cryptic 3'splice site was provided by the 418 base gag sequence just upstream from the Neo gene (Figure 5B) . The double copy vector (DC) used in this study is based on the N2 vector. In DC vectors, the gene of intereεt, driven by itε own promoter, waε placed outεide the retroviral vector's LTR initiated tranεcription unit to overcome the possible negative effect of the LTR transcription on the transcription initiated by the internal promoter (see Results) .

The two major advantages of retroviral vector mediated gene transfer over other means of gene transfer are its high efficiency and the broad host range. The gene maybe introduced into cells at one copy per cell in a genetically stable manner without adverse effect on the recipient cell and may efficiently infect a large proportion of the target cells. Retroviral vectors packaged in amphotropic viral particles can potentially infect a wide variety of cell types including human cells.

Gene Transfer in Hematopoietic Tissues

Gene tranεfer has been conducted in various non- hematopoietic types of cells, such as skin fibroblasts (Palmer et al., 1987), skin keratinocytes (Morgan et al., 1987; Flowers et al., 1990), hepatocytes (Wilεon et al., 1988; Anderεon et al., 1989), endothelial cells (Zwiebel et al., 1989; Wilson et al., 1989), muscle cells (Wolff et al., 1990), lymphocytes (Rosenberg et al., 1990). These cellε, however, normally have a limited lifespan and the expression of the transduced gene is therefore εhort lived.

Bone marrow as the major hematopoietic organ in adults, iε an attractive target for gene therapy. Compared to other target cellε mentioned above, bone marrow has obvious advantages as the target of gene therapy: the well developed procedures for bone marrow transplantation, the large number and wide distribution of hematopoietic cells, the existence of many diseases that affect hematopoietic cells, and most importantly, the existence of a small number of pluripotent hematopoietic stem cells (HSC) capable of both self- renewal and differentiation following transplantation into appropriately conditioned recipients. These cells and their progeny will contribute to hematopoietic reconstitution for the lifetime of the recipient.

The frequency of εtem cells has been estimated to be approximately 0.001% of nucleated marrow cells (Harrison et al., 1988). Therefore, any method of introduction of exogenous DNA needs to be extremely efficient. To date recombinant retroviral vectorε appear to be the oεt promiεing technology to tranεfer DNA into thiε rare cell type. Retroviral mediated gene tranεfer demonεtrateε a relatively high efficiency of gene transfer, stable integration of the provirus into the host cell genome, and the capacity to carry up to 10 kb of new genetic material (εee 1.3).

Gene tranεfer into bone marrow cellε of mouεe, primateε and human haε been reported by many reεearch groupε with a variety of retroviral vectorε containing genes as diverse aε hypoxanthine phosphoribosyltransferase (HPRT) , purine nucleoside phosphorylase (PNP) , adenosine deaminase (ADA) , B-globin and hematopoietic growth factors as well as drug resistance genes such as NEO^r and MTX^r (Joyner et al., 1983; Williams et al., 1984; Dick et al., 1985; Keller et al., 1985; Gruber et al., 1985; Valerio et al., 1985; Mclvor et al., 1987; Willia ε et al., 1986; Kantoff et al., 1987; Karlεson et al., 1987; Wong et al., 1987; Bender et al., 1989; Corey et al., 1990) . The presence and/or the expression of the transferred gene for longer than 4 months posttransplantation in myeloid and lymphoid tissues is generally accepted aε evidence for εtem cell infection. Sequential transplantation has also been used to demonstrate infection of stem cells with extensive repopulating capability.

Deεpite the conεiderable progreεs in achieving long-term and stable expression of transduced genes in recipients of infected pluripotent stem cells, various difficulties remain to be overcome before gene therapy can be considered a feasible treatment.

The efficiency of infection and expression of any specific vector, for example, is unpredictable without in vitro testing in a proper system. The expression of retroviral vectors in the conventionally used murine fibroblaεt cell line NIH3T3 and other cell lineε haε been repeatedly reported as failing to correlate with the expression in primary hematopoietic cells (Williams et al, 1986; Magli et al., 1987; Belmont et al., 1988; Hock et al., 1989; Li et al., 1992). The expreεεion of certain genes from certain promoters can vary widely in different cell lines or in hematopoietic cells from different species (see review by Apperley and Williams, 1990) .

The complicated kineticε of reconstitution of the hematopoietic system also presents a major obstacle for long-term expression of the transduced gene at an adequate level in the hematopoietic cells in vivo. The clonal succession of the normal hematopoiesiε, in which the sequential activation of different stem cell clones contribute to hematopoiesiε (Lemiεchka et al., 1986), and the finding that the hematopoietic syεtem consists of stem cell clones which supply progeny for long periods of time as well as those which undergo dramatic temporal changes (Snodgrass and Keller, 1987) made gene therapy difficult to achieve in this organ. 100% of the transplanted cells may have to be infected to ensure permanent correction of the disease phenotype.

One of possible way to overcome this difficulty iε to develop a selective scheme to enrich the transduced stem cells in vivo, as well as in vitro. Vectors were developed which contained not only the gene of intereεt but also a gene conferring a selectable phenotype. A bacterial transposon Tn5 neomycin phosphotranεferase gene (NEO) , which confers resistance to the drug G 18 (Southern and Berg,1982), haε been uεed aε a dominant selectable marker in different vector designs by itself or in conjunction with other genes of interest and has led to successful expression of NEO resistance in hematopoietic cells (Dick et al., 1985; Keller et al., 1985; see review by Williams, 1990). G418 is an aminoglycoside antibiotic, with a εtructure resembling gentamicin, neomycin and kanamycin. But unlike these related compounds, G418 interfereε with the function of 80S ribosomes and blocks protein εyntheεiε in eukaryotic cells (Davies et al., 1980). This aminoglycoside antibiotic can be inactivated by the bacterial phosphotransferase coded by the NEO gene. So far, no mammalian cells have been found naturally resistant to G418 unless the cells are transduced by the NEO gene, a deεirable εituation for a selection system.

There are limitations to this selection system when applied to gene transfer in hematopoietic cells in vivo. Due to its toxicity to mammalian hosts, G 18 can not be used for in vivo selection. Although in vitro selection with G418 for 48 hr before bone marrow transplantation results in the efficient removal of non-transduced cells, the long-term expression of the transduced gene waε not improved, suggesting that pre-selection eliminated long- term reconεtituting εtem cellε, either because none of the εtem cellε were infected or because they were incapable of expressing sufficient neomycin phosphotransferase at the time of selection (Karlsson et al., 1988). Other possible explanations for the failure to sustain expresεion include mutation or deletion of the tranεferred DNA εequenceε (Hauser et al., 1987) or the development of an antibody responεe to the exogenouε protein (St. Louis and Verma, 1988) .

Another approach to εelection of tranεduced hematopoietic εtem cellε would be the uεe of a selectable gene for which in vivo treatment is posεible. Purpose of Present Invention

Efficient expresεion of MTX resistance in mammalian cells via gene transfer of altered DHFR cDNAs haε important implicationε for both basic gene transfer studies and eventual clinical applications of the gene transfer technique for gene therapy. The introduction into bone marrow stem cellε of an altered murine DHFR gene (3T6 R400) reεulting in MTX reεistance enabled the selection of transduced hematopoietic cells in vivo. The recipient mice were protected from the lethal bone marrow toxicity induced by MTX, although the enrichment of εtem cellε under the particular in vivo εelection εchedule waε not obviouε (Williamε et al., 1987; Corey et al., 1990). Generation of a drug-resistant bone marrow may facilitate the development of aggresεive chemotherapeutic regimenε that otherwiεe might lead to lethal bone marrow toxicity (Bertino, 1979).

In an effort to uεe an altered DHFR gene conferring MTX reεistance as a dominant selectable marker and to determine the effect of different promoters on the expression of the mutant DHFR, MoMLV based retroviral vectors carrying the murine mutant DHFR 3T6 were constructed. Five different promoters were used and their expresεion was compared in NIH 3T3 fibroblast cell lines, three human leukemia cell lines and mouse bone marrow CFU-GM colonies. Retroviral vectorε carrying human mutant DHFR S31 and S34 were also constructed and their expression was tested and compared with the murine mutant.

The in vivo expression of mutant murine or human DHFR constructε were tested in mice to determine if protection waε conferred to the recipient mice with different MTX selection schedules which allow the demonstration of long-term as well as short term expression of the MTX resistance phenotype. Serial bone marrow transplantations also were performed. The enrichment of the MTX reεiεtant progenitor cellε waε teεted.

We alεo teεted human mutant dhfr with εerine mutationε at poεitionε 31 or 34, as well as murine mutant dhfr with a non-active site mutation at residue 15 where the wild type glycine was changed to tryptophan (G to T change at nt 46) , for their use as a selectable marker in Chinese hamster ovary cells.

SUMMARY OF THE INVENTION

Thiε invention provideε a DNA vector which compriεeε DNA encoding a mutant, antifolate reεiεtant, dihydrofolate reductase inserted into a εite within the vector, the preεence of which εite is not esεential for replication of the vector.

This invention further provides the above-deεcribed DNA vector, wherein the mutant dihydrofolate reductaεe haε substantially the same amino acid sequence as naturally occurring human dihydrofolate reductase.

In an embodiment, the mutant dihydrofolate reductase differs from naturally occurring human dihydrofolate reductase by virtue of the preεence of a serine residue at position 31 or 34. In another embodiment, the mutant dihydrofolate reductase differs from naturally occurring human dihydrofolate reductase by virtue of the presence of a tryptophan residue at position 15.

This invention provides the above-described DNA vector, wherein the 5'end of the cDNA encoding a mutant dihydrofolate reductase is operatively linked to a promoter sequence and the 3'end of the cDNA to a polyA sequence.

This invention also provides a human cell which compriseε the above-described vector or retroviral vector. The human cell may be a hematopoietic human cell or bone marrow cell.

This invention also provides a method for reducing the toxic effects of antifolate therapy in a subject which comprises replacing the subject's hematopoietic cells with hematopoietic cells which comprise the above- described vector or retroviral vector so as to reduce the toxic effects of antifolate therapy in the εubject.

Thiε invention provideε a method for introducing a selectable marker into a mammalian cell which compriseε tranεfecting the cell with DNA encoding a mutant dihydrofolate reductase capable of increasing the antifolate resistance when introduced into a cell.

Finally, thiε invention provideε a method for εelecting mammalian cellε expressing protein of interest which comprises a. introducing into the cells a DNA molecule comprising DNA encoding the protein of interest and DNA encoding a mutant dihydrofolate reductase capable of increasing the antifolate resistance when introduced into a cell; b. culturing the resulting transfected cells; and c. selecting cells which express mutant dihydrofolate reductase, so as to obtain cells which express the protein of interest.

DETAILED DESCRIPTION OF THE FIGURES

Figure 1. The role of folate coenzymes in the syntheεiε of thymidylate, purine nucleotideε, and methionine.

Abbreviationε : FH2, 7,8-dihydrofolate; FH4, 5,6,7,8-tetrahydrofolate; CH3FH4, 5- methyltetrahydrofolate; CHOFH4, 5- f ormyltetrahydrof olate ; CHFH4 , 5 , 10- methenyltetrahydrof olate ; CH2FH , 5 , 10- methylenetetrahydrof olate; S-AM, S- adenoεylmethionine .

Figε. 2A-2B: Structure of Methotrexate (MTX) and dihydrofolate (Fig. 2A) ; Tranεport, polyglutamation and mechanism of action of MTX (Fig. 2B) . Abbreviations: same as Figure 1; FGS, folylpolyglutamyl synthetase; (G)n, polyglutamated formε with variouε number of the glutamates.

Figure 3. The retroviral life cycle. The retroviral virion contains two molecules of the single stranded (sε) RNA genome. The viruε infectε a εuεceptible cell and the RNA enters the cell and is reverεe transcribed into double stranded (ds) DNA which migrates to the cell nucleus and iε integrated into the chromoεome of a host cell to form the provirus. Viral RNA and mRNAs, encoding viral proteins, are transcribed from the integrated provirus. The m NAε are translated to viral proteins which are used to encapsulate the viral RNAs as the virus buds off from the cell. Figure 4 The use of a packaging cell line in retroviral vector mediated gene tranεfer. MoMLV vector viruε DNA contains the cis elements necesεary for retroviral replication εuch as the LTR sequences and packaging signal sequence, +. In the vector DNA the viral geneε have been replaced with sequences of interest. A packaging cell line produces the viral gag, pol, and env proteinε necessary in trans for retroviral replication. These proteins are expressed from templates that lack the MoMLV packaging signal. When a retroviral vector DNA is tran fected into and tranεcribed in the packaging cell, transcripts which contain the viral packaging signal, ψ, are encapsulated by the viral proteinε εupplied by the packaging cell and are secreted as vector containing virus. Vector containing virus can be used to infect a target cell through the pathway normally used by the retrovirus. Thiε process leads to the stable integration of a vector provirus in the target cell.

Figs. 5A-5B. Retroviral vector designε. All retroviral vectors contain the retroviral LTRs and packaging signal sequence, ψ. Fig. 5A. The vectors with internal promoters (VIP) allow for the expresεion of two genes, but from different promoters. One gene, often a marker protein encoding gene, iε expressed from the viral LTR promoter, and the εecond gene is expressed from an

SUBSTITUTE SHEET (RULE 2δj internally encoded promoter. Fig. 5B. The N2 retroviral vector contains the first 418 baεe pairε of the MoMLV gag gene. The marker gene, NEO, is expressed from a spliced RNA which uses the MoMLV splice donor site (SD) and a cryptic splice acceptor site (SA) (Armentano et al., 1987) . The unspliced RNA contains the MoMLV packaging signal and represents the vector containing virion RNA.

Figs. 6A-6D. Structure of N2A (Fig. 6A) , N2AP (Fig.

6B) , DC/AC (Fig. 6C) , DC/SV (Fig. 6C) , DC/AD/R (Fig. 6D) , DC/SV/R (Fig. 6D) , DC/TK/R (Fig. 6D) and DC/CMV/R (Fig. 6D) .

In N2A, the inserted polylinker at U3 region containε Apal(A), Bglll(B) , SnaBI(Sn), SacII(S), MluI(M) reεtriciton sites. The poly(A) signal sequence was cloned in the antiparallel orientation of the LTR transcriptional unit in N2AP. The SV 40 early promoter (SV) or the human β- actin promoter (AC) was cloned in N2A in the parallel orientation of the LTR transcriptional unit. The SV promoter and the human adenosine deaminaεe promoter (AD) or the HSV thy idine kinase promoter (TK) or the cytomegalovirus promoter (CMV) was cloned in N2AP in the antiparallel orientation.

Figs. 7A-7E. Structure of DC vectors carrying mutant mDHFR cDNA. Mutant mDHFR was obtained from pFR400 (Simonsen and Levinson, 1983) (Fig. 7A) . By inserting different restriction fragments of pFR400 into the SnaBI site of different vectors in either the parallel or antiparallel orientation, the vectors carrying mutant DHFR cDNA were generated: N2A + PvuII-

SacII(parallel)=DC/SV-mDHFR (Fig.7B) ; N2A + PvuII-Sall(antiparallel)=DC/SV/R-mDHFR (Fig. 7C) ; DC/AC + Hindlll- Ncol(parallel)=DC/AC-mDHFR (Fig. 7C) ; DC/AD/R, DC/TK/R and DC/CMV/R + Hindlll-

Ncol(antiparallel)=DC/AD/R-mDHFR, DC/TK/R- mDHFR and DC/CMV/R-mDHFR (Fig. 7D) .

Figs. 8A-8H. Structure of DC vectors carrying mutant hDHFR cDNA. The full length mutant hDHFR cDNAs (S31 or S34) with modified 5'ends were obtained from pKT7HDR (Schweitzer et al., 1989). The Ncol and Hindlll fragment (about 800bp) was cloned into the SnaBI εite of DC/SV in parallel orientation, and of DC/SV/R and DC/AD/R in antiparallel orientation, generating DC/SV-hDHFR31/34 (Fig. 8B) , DC/SV/R-hDHFR31/34 (Fig. 8C) and DC/AD/R-hDHFR31/34 (Fig. 8C) . The Ncol and Bglll fragment (about 655 bp) was cloned similarly into the SnaBI site of DC/SV in parallel orientation, and of DC/SV/R and DC/AD/R in antiparallel orientation, generating DC/SV-hDHFR31 NB (Fig. 8D) , DC/SV/R-hDHFR31 NB (Fig. 8E) and DC/AD/R-hDHFR31 NB (Fig. 8E) . The mutant hDHFR cDNA (S31) with non-modified 5'end was obtained from pSV4HDR. The Hindlll and Bglll fragment was cloned into the SnaBI site of DC/SV in parallel

C- 'x

SUBSTITUTE SHEET (RULE 26 orientation, and of DC/SV/R and DC/AD/R in antiparallel orientation, generating DC/SV-hDHFR31 HB (Fig. 8G) , DC/SV/R- hDHFR31 HB (Fig. 8H) and DC/AD/R-hDHFR31 HB (Fig. 8H) .

Figure 9. The reverse transcriptase activity in the 3T3 cell lines transduced by retroviral vectors. Parental and transduced 3T3 cells were passed 4 times in culture over 2-3 weeks before the culture media were collected and centrifuged at 3,000 RPM at 4*C for 10 min. the supernatants were mixed with RT cocktail and the RT assay were performed with the Oligo (dT) primer and α³²P-dTTP. The reaction mixture was filtered and washed, and the filter paper waε exposed to x-ray film. Parental 3T3 cells and ecotropic producer line were used as negative and positive control reεpectively. The numbers following the vectorε indicate individual viruε producer cloneε from which the infectious supernatants were collected and used in the infection.

Figure 10. Role of the U3 region of the LTR in retroviral replication. The R region of the viral LTRs is present in both ends of the viral RNA. The U5 and U3 region of the viral LTR are only present in one copy in either the 5' or 3' end of the viral RNA respectively. Therefore, the single copy of the U5 region serves as the template in reverse tranεcription for both copies of the U5 region of the linear DNA form of the viral genome. Similarly, the single copy of the U3 region present in the 3' end of the viral RNA serveε aε the template for both the 5' and 3' LTR copieε of the U3 region in the linear DNA form of the viral genome and the proviruε. The proviral DNA encodeε a tranεcription unit which iε responsible for the production of the viral RNA. Transcription initiateε in the 5' LTR R region and terminateε in the 3' LTR R region. The primer binding εite (PBS) and the poly (A) εignal sequence (An) are also illustrated.

Figure 11. Possible mesεenger RNA εpecies in target cells infected by DC retroviral vectors carrying a foreign gene. In target cells, the gene of interest inserted in the U3 region is duplicated. The gene inserted in the 5' LTR is outside the retroviral transcriptional unit. The DHFR mRNAs tranεcriptε are initiated from the minigene promoter. The viral RNA and NEO RNA tranεcriptε are initiated from LTR promoter. Higher molecular weight tranεcriptε are poεεibly initiated from the minigene promoter and read through the proviral DNA.

Figs. 12A-12B. Expression of the mDHFR in NIH 3T3 cells.

Total cellular RNA was isolated from 3T3 cells infected with different retroviral vectors carrying mutant mDHFR and fractionated on Oligo(dT) cellulose columnε. The poly (A) fraction waε εubjected to electrophoreεiε in an agaroεe/formaldehyde gel, blotted on a nylon membrane, and hybridized with a ³²P- labeled mDHFR probe (Fig. 12A) or a human glyceraldehyde-3-phosphate dehydrogenase (GAPD) probe (Fig. 12B) to control for loading. The number following the vectorε indicate individual viruε producer cloneε from which the infectiouε supernatants were collected and used in the infection. The 1.1 kb and 0.8 kb transcripts are the DHFR transcriptε from the internal promoters.

Figε. 13A-13B. Northern analysis of the expression of the mutant mDHFR in NIH 3T3 cellε infected with viral εupernatantε of five individual producer lineε of DC/SV-mDHFR and DC/SV/R- mDHFR. The RNA blot waε hybridized with a ³²P-labeled mDHFR probe (Fig. 13A) or a GAPD probe (Fig. 13B) . The number following the vectors indicate individual virus producer clones from which the infectious supernatants were collected and used in the infection. The 1.1 kb and 0.8 kb tranεcriptε are the DHFR tranεcriptε from the internal promoterε.

Figs. 14A-14B. RNA blot analysis of the expression of the mutant mDHFR in CEM, K562 and Raji cells infected with different retroviral vectors carrying mutant mDHFR. The RNA blot waε hybridized with a ³²P-labeled mDHFR probe (Fig. 14A) or a GAPD probe (Fig. 14B) . The number following the vectors indicate individual virus producer clones from which the infectious supernatantε were collected and uεed in the infection. The 1.1 kb and 0.8 kb tranεcripts are the DHFR transcriptε from the internal promoterε.

Figs. 15A-15B. Southern analysis of the integrated proviral structure. DNA iεolated from Raji, CEM and K562 cellε infected with different retroviral vectorε carrying mutant mDHFR waε digested with Dral, electrophoresed in agarose gels, blotted, and hybridized with a ³²P-labled mDHFR probe (Fig. 15A) , and a NEO probe (Fig.

15B) aε deεcribed in Material and Methods. The number following the vectors indicate individual viruε producer cloneε from which the infectiouε supernatantε were collected and used in the infection;

Raji:plasmid and K562:plasmid are the positive controls with vector DNA plasmid added to the DNA of noninfected parental cells. The distortion of the DNA bands especially in the K562 panel is due to the high salt content of the DNA prepared using CsCl gradient.

Figure 16. Dose response to MTX of CFU-GM in different selection conditions. The number of CFU-GM colonies per 4 X 10⁵ mouεe bone marrow cells was counted after culturing for 14 days in 3 different selection medium with various concentrations of MTX. The number of CFU- GM colonies in the absence of MTX was used as control.

Figure 17. Protocol for DHFR gene tranεfer to marrow progenitors of the mouse and εelection with MTX. Donor marrow (1-2 X IO⁷ cells per donor mouse) was cocultured for 48 hr with the pre-irradiated (1500 R, 2 hr earlier) parental packaging line (AM12) or viral producer lineε before being tranεplanted into the recipient mice irradiated with 900 R 24 hr earlier. MTX selection waε started either 24 hr later with low-dose schedule or 4 weeks later with delayed high-dose schedule.

Figs. 18A-18B. Survival after BMT and low-dose MTX selection. Irradiated recipients were transplanted with transduced (cocultured With DC/AD/R-mDHFR or DC/SV/R-mDHFR) or untransduced (cocultured with AMI2) bone marrow on day 0. The low-doεe MTX selection started on day 1. In experiment A, 1.5 mg/kg twice a week was administered ip for the first week and 3 mg/kg twice a week ip for the next few weeks. In experiment B, 2 mg/kg twice a week ip was given for the first week and 5 mg/kg twice a week ip for the next few weeks. Both experiments had 8 to 10 recipient mice in each of the three BMT groups. After day 30, one of the survived animals was sacrificed at intervalε to perform CFU-GM aεsay or for biochemistry test.

SUBSTITUTE SHEET (RULE 265 Figs. 19A-19B. Changes in hematocrit and white blood cell count (WBC) after BMT and Low-doεe MTX εelection. The hematocrits in survived recipients of untranεduced marrow (control) or transduced marrow (DC/SV/R- mDHFR, DC/AD/R-mDHFR) before the BMT (day 0) and day 13 or day 28 after BMT and MTX selection are shown in Fig. 19A. No animal survived in the control group at day 28. The WBC counts are εhown in Fig.

19B. Each BMT group had 8 to 10 animalε and the εtandard deviations are also εhown in the figure.

Figs. 20A-20C. The survival in a series of BMT under low- dose MTX selection. The MTX selection schedule of the primary BMT (Fig. 20A) was 1.5 mg/kg twice a week for the first week, 3 mg/kg twice a week for the next few weeks. The secondary (Fig. 20B) and tertiary (Fig. 20C) recipients were selected with MTX 2 mg/kg twice for the first week, 5 mg/kg twice per week for the next few weeks. The survival of the control for the primary BMT (irradiated mice which received untransduced marrow) was uεed aε a reference in the secondary and tertiary BMT survival curves.

Figs. 21A-21B. MTX resistant CFU-GM colonies after in vivo MTX selection following BMT. Bone marrow cells from the primary recipientε (Fig. 21A) and from εecondary recipientε (Fig. 21B) were obtained at intervals during the in vivo MTX selection (2 mg/kg

SUBS -p-^a-"! i i I L Ε SHEET (RULE 26) twice for the first week, and 5 mg/kg twice a week for the remaining weeks after BMT) . The CFU-GM colonies per 4 X IO⁵ bone marrow were counted after culturing in the absence or preεence of 100 nM MTX for 14 dayε. The percentage of reεiεtant colonieε (%) was calculated by dividing the number of MTX resiεtant colonieε with the colonieε formed in the absence of MTX. Bone marrow cellε from normal mice without

BMT and MTX in vivo treatment were uεed aε the control.

Figε. 22A-22B. The εurvival of BMT and Delayed high-doεe MTX εelection. The number of recipientε surviving the delayed high-dose MTX selection in either the control group (receiving the untransduced marrow) or the group receiving DC/SV/R-mDHFR transduced marrow are shown in Fig. 22A. The high- dose selection started 4 weeks after the BMT with MTX of 100 mg/kg twice a week, i.p. for 4 weeks and increased to 200 mg/kg twice a week, i.p. for 6 weeks. There were 3 recipients in the control group and 5 in the DC/SV/R-mDHFR group.

Figs. 23A-23C. PCR blot of mouse tissues after BMT and

MTX selection. Genomic DNA from mouεe tissues was prepared 8 weeks after BMT and low-dose MTX selection (Fig. 23A) . NEOl and NE02 were used as primers in the PCR reaction (40 cycleε of 94°C 1 min, 55°C 1 min, and 72°C 1 min) . The productε were subjected to electropheresiε in an agarose

SUBSTITUTE SHEET (RULE 26 gel, blotted to a nylon membrane, and hybridized with a ³²P-labelled NEO probe. The 415 bp hybridization is the specific NEO fragment amplified. DNA from normal mouse tissues were used aε control. Genomic DNA extracted from mouεe tissues 4 months after BMT with delayed high-dose MTX selection was alεo analyεed by a PCR blot with NEO primerε and NEO probe (Fig. 23B) . DNA extractε from the mouεe receiving untransduced marrow (AM12) were used as controls. The genomic DNA from 3T3 cells transduced by mutant mDHFR were subjected to PCR blot analysis under similar conditions (Fig. 23C) . DNA extractε from a AM12 mouεe were used aε controls.

Figure 24, PCR blot analysiε of MTX reεistant CFU-GM colonies after BMT and in vivo MTX selection. Bone marrow cells from a normal mouse or recipients of BMT with either untransduced marrow (AM12) or transduced marrow (DC/SV/R-mDHFR, primary or secondary) were used in the CFU-GM assay in the absence or the presence of 100 nM MTX. Genomic DNA was pooled from 5 to 6 CFU-GM colonies resiεtant to MTX in the recipients of transduced marrow, and from colonies of a normal mouse or the recipients of AM12 marrow grown in the absence of MTX. NEOl and NE02 were uεed aε primerε in the PCR analysis and a NEO probe was used for hybridization (see legend of Figure 17) . Figs. 25A-25B. Sequence analysis of mouse tissues after

BMT and MTX selection. Genomic DNA from peripheral blood cells 8 months after secondary BMT and Low-dose MTX selection (Fig. 25A) , and from spleen and liver 5 weeks after primary BMT and low-dose MTX selection (Fig. 25B) , was amplified by asymmetric PCR using GT-NC1 and M301 primers. The PCR product was sequenced with a M210 primer by the dideoxy chain termination method. The four lanes of the sequencing gel, read from left to right are A, C, G, T bases. The arrow points to the mutation of A to C in the non-coding strand.

Figs. 26A-26B. Southern analysiε of the mouse tissues after BMT and MTX selection. Genomic DNA extracted from tissues of mice receiving transduced marrow (DC/SV/R-mDHFR or

DC/AD/R-mDHFR) or untransduced marrow (AM12) , or of a normal mouse (NM) was digested with Dral, electrophoresed in agarose gels, blotted and hybridized with a ³P-labelled mDHFR probe (Fig. 26A) , and a NEO probe (Fig. 26B) . 3T3:plasmid is the positive control with vector DNA plasmid added to the DNA of noninfected parental 3T3 cells. 3T3 cells infected with either of the two DC vectors are used as control for 100% integration of a εingle copy of the proviral DNA. All tiεεues were obtained from primary BMT unless otherwise indicated as secondary (2) BMT. Figure 27. The detection limit of the southern analysiε. DNA extracted from 3T3 cells transduced by DC/AD/R-mDHFR was digested with Dral, diluted to the indicated percentage in DNA of the untransduced 3T3, electrophoresed in an agarose gel, blotted, and hybridized with a ³²P- labelled NEO probe. The vector plasmid DNA was used as control.

Figs. 28A-28B. RNA and DNA analysiε of the expreεεion and the proviral structure of the retroviral vectorε carrying the full length mutant hDHFR cDNA. The poly A fraction of total cellular RNA isolated from 3T3 cells infected with retroviral vectorε (G418 reεiεtant, gr, with the exception of DC/SV-hDHFR31/mr, εee legend of Table 5 for detail) carrying the full length of mutant hDHFR cDNA, was subjected to electrophoresis in agaroεe/formaldehyde gel, blotted to a nylon membrane, and hybridized with a mixture of ³²P-labelled mDHFR and hDHFR cDNA probeε (Fig. 28A) . The 3T3 cell lines infected with retroviral vectors carrying mDHFR or the vector alone were used as control for the experiment. The 1.2 and 0.9 kb markers indicate the expected length of the hDHFR tranεcriptε. DNA iεolated from 3T3 cellε tranεduced by retroviral vectorε carrying full length mutant hDHFR cDNA waε digested with Dral, waε electrophoreεed in agarose gels, blotted, and hybridized with a 32P- labelled NEO probe (Fig. 28B) . The parental 3T3 cellε and 3T3 cells infected with the vector alone were used as negative controls and the 3T3:plasmid is the positive control with hDHFR carrying vector plaεmid DNA added to the DNA of uninfected parental cells.

Figs. 29A-29B. Immunoprecipitation of the DHFR enzyme protein in the 3T3 cells transduced by retroviral vectors carrying full length mutant hDHFR cDNA.

Fig. 29A. A rabbit anti-human polyclonal antibody againεt DHFR waε titered in a CHO cell line that lackε the DHFR (DG44) . The ratios at the top indicate the dilution of the antibody against a ³⁵S-labelled cell extract. The DG44-hDHFR is the DG44 cell line that had been transduced by hDHFR. The molecular weight markers are on the right and the arrow points to the 22 kd

DHFR enzyme precipitation. Fig. 29B. Cell extracts from 3T3 cell lines parental or infected with retroviral vector alone or with the vectorε carrying the full length mutant hDHFR were incubated with the Ab in a 50 to 1 ratio. Equal amountε of radioactivity of the precipitates were electrophoresed on a 15% SDS polyacrylamide gel. The DG44-hDHFR line was used aε control for the experiment. The arrow pointε to the 22 kd DHFR enzyme precipitated. Gr εtandε for G418 resistance; mr stands for MTX resistance (see Table 5 legend and the text for detail) .

SUBSTITUTE SHEET (RULE 25^' Figs. 30A-30B. RNA analysis of the expression of the retroviral vectors carrying less than full length mutant hDHFR cDNA. The RNA blot of the poly A fraction of the cellular RNA from the 3T3 cell lines transduced by the vectors carrying lesε than full length of mutant hDHFR CDNA (hDHFR31HB or hDHFR31NB) were hybridized to a ³²P-labelled hDHFR probe (Fig. 30A) and a GAPDH probe (Fig. 3OB) . The numbers under each vector indicate individual virus producer clones from which the infectious supernatant was collected and used in the infection. The conditions in which the infected 3T3 cells were selected are designated as gr for

G 18 resiεtant or mr for MTX reεiεtant. The arrowε point to the expected 1.1 and 0.8 kb messages from the internal promoters. The two 3T3 cell lines transduced by the mDHFR were used as control though the mDHFR did not hybridized well with the hDHFR probe. The parental 3T3 cell line waε used as the negative control.

Figε. 31A-31B. DNA analysiε of theSproviral structure of the retroviral vectorε carrying leεε than the full length of the mutant °hDHFR. Genomic DNA extracted from the 3T3 cell lineε infected with the viral vectorε εelected either with G418 (gr) or MTX (mr) waε digested with Dral and separated on an agarose gel, blotted and hybridized to a 32P-labelled NEO probe. The parental 3T3 cell line was used as a negative control

SUBSTITUTE SHEE and 3T3:plasmid is the positive control in which the vector plasmid DNA was added to the parental 3T3 DNA. The molecular size markers are on the left. The numbers under different vectors indicate individual virus producer clones from which the infectious supernatant was collected and used in the infection.

Figs. 32A-32B. The εurvival of BMT with mutant hDHFR and

MTX εelection. The εurvival of mice receiving untransduced marrow (control) or transduced marrow (DC/SV-hDHFR31HB) under the low-dose schedule (2 mg/kg, twice a week for the first week, 5 mg/kg twice a week for the next 6 weeks) (Fig. 32A) or the delayed high-dose schedule (no MTX for first 4 weeks, 200 mg/kg twice a week for the next 7 weeks) (Fig. 32B) of MTX εelection are εhown. The animals surviving 7 weeks of the low-dose εelection were subjected to the high-dose εelection (200 mg/kg twice a week) for 5 weekε.

Figure 33. PCR blot of the MTX reεiεtant CFU-GM colonieε after BMT with mutant hDHFR and in vivo MTX εelection. Genomic DNA from the pooled CFU-GM colonieε (5 to 6) was amplified with H250 and GT-NC1 primers (40 cycle of 94°C 1 min, 55°C 1 min, and 72°C 1 min) , electrophoresed on an agarose gel, blotted to a membrane, and hybridized to a hDHFR cDNA probe. The arrow points to the εize of the εpecific hDHFR fragment

amplified.

Figure 34, pSV5 Expreεεion Vector Containing the dhfr Insert Cloned into the Ncol and Hindlll Sites. The dhfr insert is placed downεtream of the SV40 early promoter and iε followed by poly A+ εignal containing sequences. Some important restriction sites within the plasmid vector are shown. Vector Construction: Generation of the Ser31 and the Ser34 mutants by site directed mutageneεiε has been described (Schweitzer et al., 1989a). The expresεion vectorε containing the mutant dhfr cDNAs as well as the wild type were constructed as follows: The plasmid pHD80 containing human dhfr cDNA was obtained from G. Attardi, California Inst. of Tech. , CA. The pHD80 plasmid was used as a template for amplification of the wild type human dhfr cDNA by the polymerase chain reaction (PCR) . Amplification of the Ser31 and Ser34 cDNAs were carried out by the polymerase chain reaction (PCR) using the oligomers DHFR24 and pSV3' and the Ser31 and Ser34 cDNA insertε aε templateε. The DHFR24 PCR primer for the 5'end of the h- hfr cDNA containε a Ncol site centered at ATG start codon and anneals to the noncoding strand between nucleotides -8 and 24. The pSV3' PCR primer for the 3'end of h-dhfr cDNA contains a Hindlll εite attached to nucleotideε 638-609 of the cDNA. pSV3' 5' CGATCGA GGATCC C AAGCTT ACCTTTT 3' (Sequence ID No. 1)

Hindlll 5 DHFR24 5 '

ATCATCCCATGGTTGGTTCGCTAAACTGCATCG 3' (Sequence ID No. 2)

Ncol In vitro amplification using PCR was

10 carried out, and the product was restricted with Ncol and Hindlll. The pSV2 plasmid vector was modified to generate Ncol and Hindlll sites. The h-dhfr cDNAs were then cloned into the

15 Ncol and Hindlll sites of the modified vector. The resulting plasmid expresεion vector waε termed pSV5. For cloning the mouse Arg22 mutant dhfr cDNA a similar approach was taken. The cDNA waε

20 amplified by PCR from the 3T6R400 dhfr cDNA template (preεent in the vector pFR400 obtained from C. Simonsen and A. Levinson, Genetech, CA) uεing the primerε M5'NcoI (which annealε to the noncoding

25 strand between nucleotides -9 and +15 taking the A of the first ATG as +1) and M3' Hindlll (which anneals to the coding strand between nucleotides 610 and 643) . The mouse dhfr cDNA was of the same length

30 as the human dhfr cDNAs. The primers had the following sequenceε: M5' Ncol 5' GCTGCCATCCATGGTTCGACCATTG 3' (Sequence ID No . 3 )

Ncol

35 M 3 ' H i n d l l l 5 ' TCTAAAGCCAGCAAAAGCTTCATGGTCTTATAA 3 ' (Sequence ID No . 4 )

Hindlll After digestion with Ncol and Hindlll the Arg22 dhfr cDNA was alεo cloned into the pSV5 vector. Thiε was done so that compariεonε between the mutant dhfr cDNAε could be carried out after tranεfection uεing the same expression vector syεtem. Competent DH5 alpha cellε were transformed with the pSV5 plasmidε containing the reεpective inserts. Ampicillin resiεtant colonieε were iεolated and analyzed for the preεence of the correct insert. For transfection studies, plasmidε were iεolated from the transformed bacteria uεing a midi prep kit (Qiagen, Inc., CA) . Plasmids were purified by two rounds of phenol chloroform extraction and ethanol precipitation prior to transfection.

Figs. 35A-35B. Southern Blot Analysiε of Ncol and Hindlll

Digested Genomic DNA Isolated from Transfected Cells Hybridized with ^P Labelled Human dhfr. Panel A: Lanes 1,

2, and 3 represent DNA isolated from Ser31 transfected cells grown in 100,500 and 1000 nM MTX; lanes 4, 5, and 6 represent DNA from Arg22 transfected cells grown in 100,500 and 1000 nM MTX; lanes 7, 8, and

9 represent DNA from Ser34 transfected cells grown in 100,500 and 1000 nM MTX, respectively; while lane 10 repreεentε DNA from control CHO cells transfected with the neoR only (Fig. 35A) . Panel B shows the εame εamples hybridized with hamster dhfr probe (Fig. 25B) . Autoradiogram exposure time was 24h. Signal in lane 7, panel A was too weak to εhow any viεible 5 band at 24h; longer exposure (72h) showed a visible band. For Southern analyεis, 10 μg of genomic DNA for each cell line was incubated with restriction enzymes Hindlll and Ncol in 10 x universal buffer for 18h

10 at 37*C. The restricted DNA was electrophoreεed on a 0.8% Tris Acetate EDTA (TAE) agaroεe gel (SeaKem GTG, FMC) and transferred to Nytran (S&S, MA) overnight in 10 x SSC transfer solution by

15 capillary transfer. The DNA was crosε linked to the membrane by UV exposure (UV Stratalinker 1800, Stratagene) and then hybridized to a ³²P-labeled human dhfr cDNA. Labeling of probe was done by

20 random priming using the random priming kit from Boehringer Mannheim and alpha-³PdCTP (> 3000 Ci/m mole, NEN) . Calcium phosphate mediated gene tranεfer waε carried out uεing the mammalian

25 transfection kit (Stratagene, CA) according to the manufacturers directionε. The wild type and mutant dhfr containing plasmids (20 μg) were cotransfected with plasmids harboring the neomycin resistance

30 gene in a ratio of 20:1 (i.e., dhfrzneoR

= 20:1). MTX resistant colonies as well as G-418 resiεtant colonies were scored after 14 days. A ³P-labeled human dhfr probe was used for hybridization, and the

35 blot was washed twice at room temperature in 1 x SSC/0.1% SDS for 30 min. and once at 55°C in 0.1 x SSC/1.0% SDS for 20 min.

Figs. 36A-36B. Northern Blot Analyεiε of Total Cellular RNA from Transfected and Control CHO

Cells. Panel A shows ethidium bromide staining of the nylon filter after transfer (Fig. 36A) . Panel B shows the result of hybridization of the filter with a ³²P labelled h-dhfr probe (Fig. 36B) .

Lane 1 representε RNA from control CHO cellε; lane 2 represents RNA from Ser31 transfected CHO cells; lane 3 repreεents RNA from Ser34 transfected cells; and lane 4 representε RNA from Arg 22 transfected cells. The presence of the smaller message(ε) in lanes 2, 3, and 4 (between 0.6 and 0.7 Kb) indicate that the transfected Ser31 and Ser34 as well as the Arg22 dhfr cDNAs are expressed. Total amount of RNA loaded in each lane was 50 μg. RNA was isolated from cells derived from a single colony growing at 100 nM MTX for the Ser31 and Ser34 as well aε the Arg22 transfectants. For the control CHO cells RNA was isolated from a culture expanded from a single colony growing in 750 μg/ml of G-418. For Northern analysiε, total cellular RNA waε extracted from cellε by the guanidinium thiocyanate-phenol method (Chomczynεki and Sacchi, 1987) according to the manufacturers instructions (RNAzol, Cinna Biotecx, TX) . The RNA was electrophoresed on agarose formaldehyde gel using 1 x MOPS

SUBSTITUTE SHEET (RULE rr) buffer. The RNA was then transferred to Nytran (S&S, MA) by capillary transfer overnight and hybridized to a radiolabelled full length human dhfr cDNA. After hybridization at 42^βC overnight all blots were washed for 30 min. at room temperature in 1 x SSC/0.1% SDS and then for 20 min. at 55^βC in 0.1 x SSC/0.1% SDS.

Figure 37, Northern blot analyεis of RNA isolated from untransfected cellε and from cells transfected with the Trp15 mutant murine dhfr.

Figure 38, Southern blot analysis of DNA isolated from untransfected cells and from cellε transfected with the Trpl5 mutant murine dhfr.

i— . Λ

SUBSTITUTE SHEET (RULE ά) DETAILED DESCRIPTION OF THE INVENTION

This invention provides a DNA vector which comprises DNA encoding a mutant, antifolate resistant, dihydrofolate reductase inserted into a εite within the vector, the preεence of which εite iε not eεεential for replication of the vector.

Thiε invention further provideε the above-deεcribed DNA vector, wherein the mutant dihydrofolate reductase has substantially the same amino acid sequence as naturally occurring human dihydrofolate reductase.

In an embodiment, the mutant dihydrofolate reductase differs from naturally occurring human dihydrofolate reductase by virtue of the presence of a serine residue at position 31 or 34.

In another embodiment, the mutant dihydrofolate reductase differε from naturally occurring human dihydrofolate reductase by virtue of the presence of a tryptophan residue at position 15.

This invention also provides the above-deεcribed DNA vector, wherein the 5'end of the DNA encoding a mutant dihydrofolate reductase is operatively linked to a promoter sequence and the 3'end of the cDNA to a polyA sequence.

In an embodiment, the promoter sequence is an SV40 promoter.

This invention also provides a plasmid which comprises the above-described vector. In an embodiment, the plasmid is designated pSV5-Ser31 h-

DHFR (pSV5-Ser31) . This plasmid contains DNA encoding a mutant dihydrofolate reductase with a serine residue at position 31. This plasmid also contains SV40 promoter and poly A sequences. Plasmid, pSV5-Ser31 h-DHFR waε deposited with the American Type Culture Collection

(ATCC), 12301 Parklawn Drive, Rockville, Maryland 20852,

U.S.A. on April 9, 1993 under the provisions of the

Budapest Treaty for the International Recognition of the Deposit of Microorganism for the Purposes of Patent

Procedure. The plasmid was accorded ATCC accession number 75441.

In another embodiment, the plasmid is designated pSV5- Ser34 h-DHFR (pSV5-Ser34) . This plasmid contains DNA encoding a mutant dihydrofolate reductaεe with a serine residue at position 34. This plasmid also contains SV40 promoter and poly A sequences. Plasmid, pSV5-Ser34 h-

DHFR was deposited with the American Type Culture Collection (ATCC) , 12301 Parklawn Drive, Rockville,

Maryland 20852, U.S.A. on April 9, 1993 under the provisions of the Budapest Treaty for the International

Recognition of the Depoεit of Microorganism for the

Purposeε of Patent Procedure. The plasmid was accorded ATCC accession number 69276.

In another embodiment, the above-described vector is a retroviral DNA vector. In a further embodiment, the retroviral vector comprises DNA from a retrovirus corresponding to a 5' long terminal repeat, a 3' long terminal repeat and a packaging signal. In a still further embodiment, the site at which the DNA encoding a mutant dihydrofolate reductase inserted is in the 3' long terminal repeat of the retroviral vector. This invention further provides plasmids which comprises the above-described retroviral vectors.

In an embodiment, the plaεmid which compriεes a retroviral vector which compriseε DNA encoding a mutant dihydrofolate reductaεe capable of increaεing the antifolate reεiεtance when introduced into a cell is designated pDC SV S31 h-DHFR. This plasmid contain DNA which codeε for a mutant dihydrofolate reductaεe with serine at poεition 31. Plasmid, pDC SV S31 h-DHFR waε depoεited with the American Type Culture Collection (ATCC), 12301 Parklawn Drive, Rockville, Maryland 20852, U.S.A. on April 9, 1993 under the proviεionε of the Budapest Treaty for the International Recognition of the Deposit of Microorganism for the Purposes of Patent Procedure. The plasmid was accorded ATCC accesεion number 75440.

This invention also provideε a mammalian retroviral producer cell which compriεeε the above-deεcribed vectorε or plaεmids.

This invention further provides human cell which compriseε the above-described vectors or plasmidε. In an embodiment, the human cell is a hematopoietic cell. In another embodiment, the human cell is a bone marrow cell.

This invention also provides a method for reducing the toxic effects of antifolate therapy in a εubject which compriεeε replacing the subject's hematopoietic cells with hematopoietic cellε which comprised the above- deεcribed vectorε or plasmids so as to reduce the toxic effects of antifolate therapy in the subject. In an embodiment, the antifolate iε methotrexate. Thiε invention also provides a method for introducing a selectable marker into a mammalian cell which compriseε transfecting the cell with DNA encoding a mutant dihydrofolate reductase capable of increasing the antifolate reεiεtance when introduced into a cell.

Finally thiε invention provideε a method for εelecting mammalian cells expressing protein of interest which comprises: a.introducing into the cells a DNA molecule comprising DNA encoding the protein of interest and DNA encoding a mutant dihydrofolate reductase capable of increasing the antifolate resistance when introduced into a cell; b. culturing the resulting transfected cells; and c. selecting cells which express mutant dihydrofolate reductase, so as to obtain cellε which expreεε the protein of interest. In an embodiment, the DNA molecule of step (a) of the above method is part of a retroviral vector.

This invention will be better understood from the Experimental Details which follow. However, one skilled in the art will readily appreciate that the specific methods and results discussed are merely illustrative of the invention as described more fully in the claims which follow thereafter.

EXPERIMENTAL DETAILS

First Series of Experiments

MATERIAL AND METHODS

Materials

Chemicals Common laboratory chemicals were of the highest purity commercially available and were obtained from Mallinckrodt, Fisher, Sigma, or Bio-Rad. The following is a list of special reagents used:

From Boehringer Mannheim: DNA markers, RNAase, dNTPs, ATP, Proteinase K, Klenow fragment, T4 DNA ligase. Reverse transcriptase, PMSF, Bovine serum albumin (pentax fraction V)

From GIBCO: Trypan-blue, G418, Penicillin-streptomycin, Trypsin

From Lederle: Methotrexate

From Pharmacia: Protein A sepharose CL-4B, Poly A, Oligo dT From Sigma: Thymidine phosphorylase, Polybrene, Diethyl pyrocarbonate, Lysozyme, Ampicillin, Tetracycline, Dithiothreitol, β-Mercaptoethanol, Sarkosyl, Dimethyl εulfoxide, Hypoxanthine, Xanthine, Mycophenolic acid, Hygromycin B, MOPS, Ethidium bromide, Salmon sperm DNA, Polyvinylpyrolidone, Sodiu -deoxycholate, Leupeptin

From United States Biochemicals: T7 DNA polymeraεe (Sequenase version 2.0) Medium

From central medium laboratory of SKI unleεε otherwiεe indicated: IMDM, DME, RPMI, PBS, fetal bovine εerum (Hy Clone Laboratory)

Restriction Enzymes

From Boehringer Mannheim: Alul, Apal, BamHI, Bglll, Dral, EcoRI, Hindlll, Mlul, Ncol, PvuII, Sad, SacII, Sail, S al, SnaBI, Xbal, Xhol

Radioactive Isotopes

From Amerεham: α-^P-dCTP, -∞P-dTTP, α-³²P-dATP, ³⁵S-Met, stored at -20°C

others

Bacteria: E.coli (JM109, Stratagene*)

Cell lineε: NIH3T3, murine fibroblast cell line; CEM, human T lymphoblastoid cell line (Foley et al., 1965);

K562, human multipotential, hematopoietic malignant cell line (Lozzio and Lozzio,1975) ; Raji, Burkitt lymphoma derived lymphoblast-like cell line (Pulvertaft, 1964) ;

CHO, Chinese hamster ovary line (Puck, T.T. , 1958) ;

DG44, CHO line lacking DHFR activity (Urlaub and Chasin,

1980) ; E86, ecotropic retroviral packaging cell line (Markowitz et al., 1988a); AM12, amphotropic packaging cell line (Markowitz et al., 1988b);. WEHI-3B, murine myelomonocytic leukemia cell line (Ralph and Nakoinz,

1977) .

Antibody: Rabbit anti-human DHFR polyclonal antibody prepared by Dr. Srimatkandada

Animal: mouse (CBA/J 7-11 week old, male)

Retroviral ector construction (General Techniques) Plasmid DNA preparation

A) miniscale plaεmid DNA preparation: modified from Holmes and Quigley (1981) . Bacteria were grown in 2 ml L-Broth (1% Bacto tryptone, 0.5% Bacto Yeast, 0.5% NaCl) containing 50 μg/ml Ampicillin in a 4 ml εnapcap polypropylene tube overnight at 37°C in an incubator εhaker (220-230 RPM), tranεferred to a 1.5 ml eppendorf tube and pelleted for 2 min in a microfuge. The pellet waε resuspended in 200 μl of lysis solution (8% Sucrose, 0.5% Triton X-100, 50 mM EDTA, 10 mM Tris-Cl, pH 8.0, and with freshly added Lyεozyme 0.75 mg/ml) . The suspension was boiled for 1 min and centrifuged for 10 min at room temperature in a microfuge. The pellet was removed with a toothpick and the DNA was precipitated from the supernatant by addition of 2.5 M sodium acetate to a 0.25 M final concentration and 250 μl isopropanol. After incubation for 10 min at -20°C, the DNA was pelleted by spinning in a microfuge for 10 min. The pellet was washed with 70% ethanol and resuspended in 30 μl TE (10 mM Tris-HCl, pH 7.5, 1 mM EDTA) with 1 μg /μl RNAεe and incubated at 37°C for 30 min. The DNA solution at this stage may be used for restriction digestion and gel analysis.

B) Large scale plasmid DNA preparation: The alkaline extraction procedure (Birnboim and Dolly, 1979) with some modifications was followed to prepare plasmid DNA from bacteria. 250 ml L-Broth with 50 μg/ml Ampicillin was inoculated with 2.5 ml (1/100) of 5-6 hr bacterial culture (as described in 2.2.1 A) and incubated overnight at 37°C in a shaker (220-230 RPM) . The bacterial culture was tranεferred to a 250 ml plastic bottle and centrifuged at 6000 RPM for 10 min at 4°C. The pellet was resuspended in 5 ml ice-cold Alkaline Lysiε Solution I (ALSI: 50 mM Glucoεe, 25 mM Triε-HCl, pH 8.0, 10 mM EDTA), containing 5 mg/ml lyεozyme freshly dissolved and was transferred into a 50 ml plastic tube. 10 ml ALS II (0.2 N NaOH, 1% SDS, freεhly made) waε added and the mixture chilled on ice for 10 min before the addition of 7.5 ml of ice-cold ALS III (5 M potassium acetate, pH 4.8) . The mixture waε chilled on ice for another 30 min after thorough mixing. The mixture waε then εpun at 16,000 RPM for 40 min at 4°C. The DNA waε precipitated out from the supernatant by the addition of 12 ml isopropanol and incubation at room temperature for 15 min, and then by centrifugation at 16,000 RPM for 15 min at room temperature. The DNA pellet waε resuspended in 5 ml TE with 20 μg/ml RNAse and incubated at 37°C for 30 min.

A Quiagen column-500 (Quiagen Ine) waε used to purify the plasmid DNA according to the protocol provided by manufacturer with εome modificationε. 5.5 ml of 5 M NaCl and 2.5 ml of 1 M MOPS, pH 7.0, waε added to the DNA εolution prepared from a bacteria culture (maximum 500 ml) in 33 ml total volume. The mixture waε passed at a maximum flow rate of 3 ml/min through the Quiagen pack- 500 column pre-equilibrated with 5 ml Quiagen Buffer A (400 mM NaCl, 50 mM MOPS, 15% ethanol, pH 7.0). The column was then washed with 20 ml Quiagen Buffer C (IM NaCl, 50 mM MOPS, 15% ethanol, pH 7.0). The plaεmid DNA waε eluded from the column with 5 ml Quiagen Buffer F (1.5 M NaCl, 50 mM MOPS and 15% ethanol, pH 7.5) at a maximum flow rate of 2 ml/min. The purified plasmid DNA was precipitated from the eluate by the addition of 4 ml (4/5 volume) of isopropanol, freezing at -20°C for 30 min, and centrifugation (10,000 RPM for 15 min at 4°C) . The DNA pellet was resuspended in 1 ml TE and phenol extraction was performed twice followed once by phenol- chloroform extraction (in 1:1 ratio). Traceε of phenol and chloroform were removed by ethanol precipitation (250 mM sodium acetate and 2.5 volumes of ethanol). After incubation at -70°C for 30 min the DNA was pelleted by spinning in a microfuge for 15 min at 4°C. The pellet was washed with 70% ethanol and resuspended in TE.

Restriction enzyme digestion

A) miniscale digestion with restriction enzymes 0.2-1 μg of DNA in 17 μl water was mixed with 2 μl 10 x digestion buffer in a sterile Eppendorf tube. 0.5 to 1 μl of restriction enzyme was added to the mixture. The mixture was incubated at 37°C for 1 hr before addition of EDTA to a final concentration of 10 mM to stop the reaction. The DNA solution can be used directly to analyze the digeεtion pattern on a minigel (see 2.2.3).

When more than one restriction enzyme was used in the digestion, the enzymes were added to the digestion mixture at the same time provided the ionic strength of the digestion buffers recommended for each enzyme was the same. If not, the enzyme requiring low ionic strength digestion condition was added to the mixture first. After the first digestion was completed, the ionic strength in the mixture was increaεed according to the requirement of the εecond enzyme and εo on. In caεe of εome enzymes with altered cleavage εequence specificity under non-optimal conditions ("star" activity, Polisky, B., et al., 1975), heat-inactivation (65°C, 20 min) waε performed to inactivate the enzyme after the firεt reaction was completed before increasing the ionic εtrength for the next enzyme.

B) Large-εcale reεtriction digeεtion

For purpoεeε of preparation or in Southern analyεiε, more than 10 μg DNA waε usually uεed for digeεtion. DNA was digested at a final concentration of leεε than 0.1 μg/μl, with 1/10 volume of 10 X digestion buffer and with 4-5 units of restriction enzyme per μg of DNA. The reaction was usually performed overnight at 37°C and terminated by addition of 10 mM EDTA.

Purification of DNA by gel electrophoresis

A) Preparation of agarose gel (Maniatis et al, 1982) The agarose powder was added to electrophoresiε buffer (0.5 X TBE: 45 mM Tris-borate, 45 mM boric acid, 1 mM EDTA, pH 8.0) at a final concentration of 1%. The mixture was heated in a microwave oven until the agarose dissolved. After cooling to 50°C, ethidium bromide (from a stock solution of 10 mg/ml in water, stored at 4°C in a light-proof bottle) was added to a final concentration of 0.5 μg/ml. The solution was poured into the gel mold (50-60 ml for minigel of 4 x 6", 200-300 ml for larger gel of 6 x 10") and kept at room temperature for 30 min until the agarose was solidified.

B) Agarose gel electrophoresis

The gel was immersed in electrophoresis buffer (0.5 x TBE) in the electrophoresiε tank to which a power εupply waε connected. When the electrophoreεiε waε performed uεing the large gel, the buffer waε recirculated by a pump. For separation of DNA fragments generated by large-εcale reεtriction digeεtion, phenol-chloroform extraction and ethanol precipitation (εee 2.2.1 B) were carried out before the DNA waε loaded on the gel with 1/6 volume of 6 x loading buffer (0.25% bromophenol blue, 0.25% xylenecyanol, 30% glycerol in water).

The DNA fragments were visualized by fluorescence of the ethidium bromide (Sharp et al, 1973) which emits at 590 nm in the red- orange region of the viεible spectrum when irradiated with UV light of 360 nm wavelength. The desired DNA fragment waε cut out with a razor blade and the DNA waε electroeluted from the agaroεe gel by the following method.

C) Electro-elution of the separated DNA fragment

An electro-elutor (Model Uea, International Biotechnologies Ins.) was used to elute the DNA fragment from the agarose gel. The gel slices were placed in the wells connected to channels filled with 100 μl 10 M ammonium acetate. The power supply was connected to the electroelutor box filled with 0.5 X TBE buffer. After the elution, the DNA fragment trapped in the channel was carefully removed and was precipitated by addition of ethanol, freezing at -20°C followed by centrifugation. The DNA pellet was resuspended in TE and another round ethanol precipitation with 0.25 M sodium acetate was carried out to remove traces of the ammonium acetate salt from the DNA solution.

D) Quantitation of DNA was done either by measuring the absorption at 260 nm (The DNA concentration in the solution was calculated by the following formula: OD at 260 nm x 50 x dilution factor = [DNA] μg/ml) or by comparing the denεity of the red-orange fluoreεcence of the ethidium bromide bound to the DNA fragment on a minigel with the density of a DNA marker of known quantity.

Klenow and ligation

Blunt end ligation waε used in retroviral construction. A Klenow reaction was conducted to fill the recesεed 3' endε and to degrade the protruding 3' ends (Wartell and Reznikoff, 1980) . The Klenow fragment of DNA polymerase I (E.coli) was used in the reaction, which has the 5' to 3' DNA polymerase activity and the 3' to 5' exonuclease activity. The reaction mixture contained 0.2 mM of each deoxynucleotide, 1/10 volume of 10 x Klenow buffer (500 mM Tris-HCl, 100 mM MgCl₂, 1 mM DTT, pH 7.5), 1-5 μg of the purified DNA fragment, and 1-5 U Klenow enzyme in a total volume of 200 μl. After incubation at room temperature for 30 to 60 min, the reaction was stopped by adding EDTA to 10 mM. Then, phenol-chloroform extraction and ethanol precipitation were performed to remove the unincorporated nucleotides as well aε the protein before proceeding to the ligation reaction.

The ligation reaction waε catalyzed by T4 DNA ligase (Weiss et al., 1968). The reaction mixture contained 1 mM ATP, 1/10 volume of the 10 x ligation buffer (250 mM Tris-HCl, 100 mM MgCl₂, 50 mM DTT, pH 7.5) , the insert and the vector DNA resuspended in water in a molar ratio of 20-40 to 1 (up to 1 μg DNA per 10 μl) , and 1 U ligase in a total volume of 10 μl. The reaction was performed at room temperature overnight. The mixture was diluted 3 times to transform bacteria. In εome cases two sequential ethanol precipitations first with 3 M ammonium acetate and then with 250 mM sodium acetate were performed before the DNA ligation solution was used for bacterial tranεformation. Bacterial transformation

Electroporation was used to transform E.coli (JM109 bacteria εtrain) . Preparation of the competent cellε and the electro-tranεformation waε carried out according to the protocol provided by the manufacturer (Bio-Rad) .

A) Preparation of cellε

One liter of L-broth was inoculated by 10 ml (1/100 volume) of a fresh overnight bacterial culture. The cells were grown at 37°C with εhaking (220-230 RPM) for several hours until the OD reading at 600 nm was between 0.5 and 1 (when the cells were in the log phase of growth) . The cells were then harvested by chilling the flask on ice for 15 min and centrifuging in a prechilled rotor at 4,000 x g for 15 min at 4°c. The pellets were resuspended in an equal volume of ice-cold sterile water and were centrifuged again at the same setting. The pellets were then resuspended in half volume of ice-cold water and the above centrifugation was repeated. The pellets then were reεuspended in 20 ml 10% glycerol and recentrifuged. The cells were resuspended in a final volume of 2-3 ml in 10% glycerol at a concentration of 3 x 10¹⁰ cells/ml. The suspension was either frozen in aliquots on dry ice and stored at -70°C, or was used immediately for the electro-transformation.

B) Electro-transformation

The cells were thawed at room temperature and kept at 4°C on ice. 40 μl of the cell suspension was mixed with 1 to 2 μl of recombinant DNA from the ligation reaction in a cold 1.5 ml polypropylene tube and incubated on ice for 1 min. The DNA waε in a low ionic strength buffer (TE) . The Gene Pulser apparatus was set at 25 μF and 2.5 kV and the Pulse Controller at 200 Ω. The electroporation was performed in a pre-chilled 0.2 cm electroporation cuvette with a single pulse. The time constant was between 4.5 to 5 msec. Immediately after electroporation 1 ml of SOC (2% Bacto tryptone, 0.5% Bacto yeast extract, 10 mM NaCl, 2.5 mM KC1, 10 mM MgCl₂, 10 mM MgS0₄, 20 mM glucose) was added to the cuvette and the cellε were quickly reεuεpended and transferred to 4 ml polypropylene tube and incubated at 37°C for 1 hr with 220-230 RPM shaking. The cellε were then plated on εelective plates (1% Bacto Tryptone, 0.5% Bacto Yeast, 0.5% NaCl, 1.5% Bacto Agar, containing 50 μg/ml Ampicillin) .

Diagnosis of the clones containing the correct construct

Bacterial colonieε growing on the εelective plateε were picked with εterile toothpicks. These individual colonies were used to inoculate 2 ml L-Broth and the miniscale plasmid DNA preparation procedure was followed to isolate 30 μl DNA solution (2.2.1 A). 4 to 5 μl of this solution was used in a 20 μl restriction digestion mixture (2.2.2). At least 2 restriction digestion patternε were chosen to determine if the structure of the construct was correct. Usually it was determined first whether the construct contained the insert gene, and then the direction of the insert waε determined, which waε essential for checking constructs generated by blunt-end ligation.

Freezing and storage of the bacterial colonies

Two bacterial colonies containing the correct construct were grown up individually in 2 ml selective medium (2.2.1A) for 4 to 5 hr and were frozen in 15% glycerol at -70°C in 1 ml aliquots. The Construction of Retroviral Vectors Carrying a Mutant DHFR CDNA

Construction of retroviral vectors (Figs. 6A-6D)

The Moloney murine leukemia virus-based N2 retroviral vector(Armentano et al., 1987), which contains the bacterial transpoεon Tn5 neomycin reεistance gene (NEO) , was modified by the inεertion of a 52-bp polylinker (containing the unique reεtriction sites 5'-ApaI-BglII- SnaBI-ScaII-MluI-3') into the Nhel restriction εite present in the U3 region of the 3'LTR. The polylinker- modified N2 vector was designated as N2A (Hantzopoulos et al., 1989). The N2A vector waε further modified by inεertion of a 275 bp poly A fragment into the Apal reεtriction εite in the anti-parallel orientation of the viral transcriptional unit; the poly A fragment was obtained from plaεmid PBC12/CMV/IL2 (Ganεbacher et al., 1990) by reεtriction digeεtion with Smal and EcoRI, followed by Alul. The modified vector containing the poly A fragment in the anti-parallel orientation iε deεignated aε N2AP. The SV40 early (SV) promoter waε cloned into the Bglll εite of N2A in a parallel orientation, generating the vector DC/SV, and into the Mlul εite of N2AP in an anti-parallel orientation, generating the vector DC/SV/R (DC εtandε for double copy, see Result 3.1.1; R stands for reverse). The SV promoter fragment was from pFR400 (Simonsen and Levinεon, 1983) , digested with Kpnl and Xbal. The human β-actin promoter was cloned into the Bglll site of N2A vector, generating the vector DC/AC; the promoter fragment was from pl4T-B17 (Gunning et al., 1987), digested with BamHI and Sad. The human adenoεine deaminase (AD) promoter fragment was cloned into the Mlul site of N2AP in the anti-parallel orientation, generating the vector DC/AD/R; the AD promoter fragment was from the 2.2 ADA plasmid (Wiginton et al., 1986). It waε digested with Sspl/Ncol, followed by Mung bean nucleaεe and Klenow modification to eliminate the ATG codon in the Sεpl/Ncol fragment. The Herpes virus thymidine kinase (TK) promoter and the cytomegalovirus (CMV) promoter were cloned similarly into the Mlul εite of N2AP, generating the vector DC/TK/R and DC/CMV/R. The TK promoter was obtained from the pHSV-106 plasmid (Mcknight and Gavis, 1980) , digeεted with BamHI and Bglll. The CMV promoter waε obtained from pBC140, digeεted with Hindi and Xhol.

Construction of retroviral vectors carrying murine mutant DHFR (Figs. 7A-7E)

The murine mutant DHFR cDNA was obtained from pFR 00, containing the 3T6 DHFR with a T to G point mutation at nucleotide 68 resulting in a Leu to Arg change at reεidue 22 (Simonsen and Levinson, 1983) . In the construct deεignated aε DC/SV-mDHFR, the DHFR cDNA pluε SV40 promoter waε excised from pFR400 by restriction enzyme PvuII and SacII and waε cloned into the SnaBI site in N2A in a parallel orientation. In the construct designated as DC/SV/R-mDHFR, the minigene fragment containing the poly A εignal waε excised from pFR400 by the restriction enzymes PvuII and Sail and was cloned into the SnaBI site of N2A in an anti-parallel orientation. DHFR minigene fragments in both the DC/SV-mDHFR and DC/SV-mDHFR constructs contained the SV40 promoter.

In the construct designated aε DC/AC-mDHFR, the DHFR fragment excised from pFR400 by Hindlll and Ncol restriction enzymes was cloned into the SnaBI site of DC/AC in the parallel orientation. In the construct designated as DC/AD/R-mDHFR, DC/TK/R-mDHFR and DC/CMV/R-

SUBSTITUTE SHEET (RULE ££ } mDHFR, the DHFR fragment was cloned into the SnaBI site of DC/AD/R, DC/TK/R and DC/CMV/R in a reverse orientation.

Construction of retroviral vectors carrying mutant human DHFRS (Figs. 8A-8H)

The mutant hDHFRs uεed in theεe constructs were S31 and S34 which contain a T to G point mutation at nucleotide 95 or 104 resulting in a Phe to Ser change at residue 31 or 34 respectively (Schweitzer, B.I. et al., 1989).

The full length cDNA of mutant hDHFR including the 560bp coding region and 240 bp 3' untranslated region was obtained from pKT7HDR (Schweitzer, B.I. et al., 1989). The cDNAε were excised by restriction enzyme Ncol and Hindlll, were blunt ended by the Klenow reaction and cloned into the SnaBI site of DC/SV in a parallel orientation, and the SnaBI site of DC/SV/R and DC/AD/R in antiparallel orientation. These constructs were named DC/SV-hDHFR31/34, DC/SV/R-hDHFR31/34, and DC/AD/R- hDHFR31/34 respectively.

The mutant hDFHR cDNA (Phe to Ser at residue 31) lesε than full length, containing the coding region and 95 bp of 3' untranεlated region, were obtained from pKT7HDR by reεtriction digeεtion with Ncol and Bglll. The fragments were Klenowed and cloned into the SnaBI εite of DC/SV, DC/SV/R, and DC/AD/R, generating DC/SV-hDHFR31 NB, DC/SV/R-hDHFR31 NB, and DC/AD/R-hDHFR31 NB.

In the pKT7HDR construct, the 5' untranεlated region of the cDNA immediately before the εtarting codon ATG waε modified to generate a Ncol εite for cloning and mutageneεiε experimentε (Schweitzer, B.I. et al., 1989).

HEET RU iL • t:>

In order to avoid a posεible negative effect of thiε modification on the expresεion of the mutant hDHFR, the hDHFR cDNA obtained from pSV4HDR which contains the internal mutation at residue 31 (Phe to Ser) but with an unmodified 5' end waε alεo uεed. The pSV4HDR waε digested with Hindlll and Bglll and the fragment was blunt ended by the Klenow reaction and cloned into the SnaBI site of DC/SV, and DC/SV/R, generating the DC/SV- hDHFR31 HB and DC/SV/R-hDHFR31 HB constructε.

Production of Live Virus Packaging

The vector DNA waε packaged into the retrovirus by transfecting a packaging cell line with the vector DNA by electroporation using a modified protocol from the manufacturer (Bio-Rad) . Cells in log phase (80% confluence) were trypsinized and resuspended in DME medium with 10% FBS at the density of 1.5 x IO⁶ cells/ml at room temperature. 2 μg of vector DNA (in a supercoiled circular form) was mixed with 0.5 ml of the cell suspension in a 0.4 cm sterile electroporation cuvette and incubated at room temperature for 10 min. Electroporation was performed with a gene pulser (Bio- Rad) set at 200 volts and the capacitance extender set at 960 μF. The time constant under these conditions was between 23 and 27 μs. After 10 min incubation at room temperature 2 ml DME with 10% FBS waε added to the curvet and the cell suspension plated in a 60 mm petri dish and incubated at 37 °C in a C0₂ incubator for 48 hr before selection was applied. Transfected cells were selected with G418 (0.75 mg/ml) . The selection medium was changed every 3 to 4 dayε. After 8 to 10 days of εelection, the G418 reεiεtant colonieε were large enough to be iεolated by ring-cloning and expanded into producer cell lines. Both ecotropic and amphotropic producer cell lines were used to package the vector DNA. Ecotropic producer lines were obtained by direct electro-tranεfection of the E86 ecotropic packaging cell line (Markowitz et al., 1988a). Amphotropic producer lines, however, were obtained either by ecotropic virion infection (see 2.5) of an amphotropic packaging cell line AM12 (Markowitz et al., 1988b) or by direct electro-tranεfection aε deεcribed above.

Virus-containing supernatant collection

The viral producer cellε were grown in a petri diεh till 80% confluence, and the medium waε replaced by freεh medium (4 ml for a 60 mm petri dish, 10 ml for 100 mm petri dish) . 12 hr later, the medium was collected from the petri diεh and centrifuged at 3,000 RPM at 4°C for 10 min. The supernatant was carefully removed from the cell debris pelleted at the bottom of the centrifuge tube, and stored at -70°C.

Viral titering

The viral titer is defined as the number of infectiouε particles in l ml of viruε containing supernatant. Because the viral vector carries the Neo gene and DHFR^r gene which would render the infected cell reεiεtant to G 18 and MTX, the viral titer can be determined by the number of G 18 or MTX reεiεtant colonieε reεulting from the viral infection of NIH 3T3 cellε (εee 2.5.1).

The viral εupernatant waε diluted 1 to 10 and 1 to 1000 in medium containing 8 μg/ml polybrene, and 1 ml of the diluted εupernatant waε uεed for the assay. After 2-3 hr incubation and 8 to 10 dayε of G418 or MTX εelection, the number of reεiεtant colonieε waε counted and the viral titer calculated. The viral titer for both ecotropic and amphotropic producer cell lines was approximately 4 x 10⁴ - 5 x 10⁵ NEO and MTX resiεtant colony forming unitε/ml.

Freezing and storage of the producer cell lines

Producer cell lines were frozen in DME medium containing 50% FBS and 10% DMSO, and stored in liquid nitrogen.

Target Cell Infection NIH 3T3 cells

TK^'NIH 3T3 fibroblasts were plated at a density of 10⁵ cells/ 60 mm petri dish the night before the infection.

Supernatants collected from the producer lines were used to infect these cells in the presence of 8 μg polybrene per ml for 2-3 hr at 37°C. Parallel selections in G418

(0.75 mg/ml) and MTX (1-2 x IO^*7 M) started 24 hour post infection. After 8-10 days the reεiεtant colonieε were counted and the G418 reεistant colonies were pooled and expanded in drug-free media for use in subsequent experiments. The vector-transduced cell lines were free of replication-competent viruε, even after extensive culturing in vitro, as determined by absence of reverse transcriptaεe activity in the culture medium (Figure 9, εee 2.6) .

Human leukemia cells

Human leukemia cells (CEM, K562, Raji) were incubated for 3 hours with the amphotropic producer εupernatant in the preεence of 8 μg polybrene per ml. The 3 hour infection was repeated after overnight incubation in virus-free medium. The selection with G418 (0.75 mg/ml) was started ₂₄ hr after the εecond infection. After 2-3 weeks in G418 the resistant cells were expanded in drug-free medium for subsequent experiments.

Mouse bone marrow cells

Bone marrow cells from CBA/J 7-11 week old male mice were harveεted in IMDM medium and a mononucleated cell εuspension waε prepared. The bone marrow cellε were cocultured with producer cellε irradiated with 1500 R 2 hr before the coculturing. The coculture waε carried out in IMDM medium containing 20% FCS, 10% WEHI-3B conditioned medium (See 2.13) and 8 μg/ml polybrene for 48 hr in 37°C C0₂ incubator with a 1:1 εtarting ratio of marrow cellε to producer cells. The bone marrow cells were then either uεed for colony forming aεεayε or transplanted to recipient mice.

Reverse Transcriptase Assay

The activity of the reverse transcriptase was meaεured by synthesiε of poly T in the preεence of poly A template (Goff et al., 1981) . The sample to be measured waε mixed with RT cocktail containing 50 mM Tris-HCl, pH 8.0, 20 mM DTT, 0.6 mM MnCl₂, 60 mM NaCl, 0.05% Nonidet P-40, 5 μg/ml Oligo(dT) primer, 10 μg/ml poly (A) , 10 μM dTTP, 1 μl α- ³²P-dTTP (3000 Ci/mMole) . The final volume was 50 μl and the incubation was carried out at 37°c for 1-2 hr. The reaction mixture was added to DEAE filter paper which waε washed with washing solution (2 x SSC) at room temperature twice for 15 min. The filter paper was then rinsed with ethanol and exposed to x-ray film or counted in a scintillation counter. DNA Analysis

Genomic DNA preparation from cultured cell lines

Chromosomal DNA was prepared by the guanidinium isothiocyanate extraction procedure and centrifugation through a CεCl cuεhion (Chirgwin et al., 1979). Confluent cellε from two 100 mm petri dishes were washed twice with PBS. 2.5 ml GTC solution (4 M Guanidinium Thiocyanate, 25 mM Sodium Citrate, 0.5% Sarkoεyl, 0.1 M β-Mercaptoethanol, εtored in a dark bottle at room temperature) waε added to each diεh. The cell extractε were pooled after 5-10 min of shaking at room temperature. 2 g of CsCl was added to the 5 ml extract and dissolved by gentle shaking and incubation at 37°C. The cell extract was loaded on top of the 6 ml CsCl solution cuεhion (5.7 M CsCl, 0.1 M EDTA, pH 7.5) in a 14 x 89 mm centrifuge tube (Beckman) . After 16-18 hr centrifugation at 30,000 rpm in room temperature, the DNA layer was saved (about 2 ml in volume) and diluted to 5 ml with water. 10 ml ethanol was added to the solution to precipitate the DNA, which was spooled out with a curved pipet, rinsed in 70% ethanol, and resuspended in 5 ml water overnight at 37°C. The DNA was quantitated by measuring the abεorbance at 260 nm (εee 2.2.3 D) .

Genomic DNA preparation from mouse organs

Genomic DNA from whole animal organs and tissues was prepared by the proteinase K digestion method in a SDS denaturing buffer, followed by phenol-chloroform extraction (Gross-Bellard et al., 1972 and Enrietto et al., 1983). The organs were frozen in liquid nitrogen immediately after removal from the animal and ground to a fine powder with a prechilled mortar and pestle. The powdered tiεsue was suεpended in digeεtion buffer (100 mM NaCl, 10 mM Tris-HCl, pH 8, 25 mM EDTA, pH 8, 0.5% SDS, and freshly added proteinase K at 0.1 mg/ml). The εuεpension was incubated at 50°C overnight with εhaking. Phenol-chloroform extraction waε performed, followed by ethanol precipitation with 2.5 M ammonium acetate. The DNA was pelleted by centrifugation at 1700 g for 2 min. The pellet was washed with 70% ethanol, air dried and resuspended in TE buffer.

Genomic DNA preparation from CFU-GM colonies

5-6 CFU-GM colonies (50-200 cells per colony) were aspirated and pooled and diluted 1 to 5 with PBS in an 0.5 ml eppendorf tube. The cells were centrifuged at 2,000 rpm and washed once with PBS. The cells were then lysed by boiling for 5 min in 20 μl εterile water, and centrifuged in a microfuge (15,000 rpm) for 5 min. The supernatant was used as the source of DNA for PCR amplification (See 2.9.1).

Southern analysis

A) Restriction digestion, gel electrophoresiε and membrane tranεfer DNA waε digested with restriction enzymes (see 2.2.2 B) , and fractionated by electrophoresis on a 1% agarose gel (10-15 μg/line, εee 2.2.3 A,B) overnight with circulating buffer. The gel waε then soaked in alkaline solution (0.2 N NaOH, 0.6 M NaCl) with gentle shaking for 20 min to denature the DNA, rinsed in water, and equilibrated with electro-tranεfer buffer (25 mM sodium phosphate, pH 6.5) before being transferred to a nylon membrane (Biotranε, ICN) with an electroblotter (Hoefer) for 4-5 hr at room temperature with cooling water circulation. The DNA was UV cross-linked to the membrane with a UVεtratalinker (Stratagene) .

B) Syntheεiε of radio-active probe

The ³²P-labeled εpecific probe waε generated by an Oligolabelling Kit (Pharmacia) . DNA (25-50 ng) waε first denatured by boiling for 2-3 min. It was then immediately chilled on ice and mixed with hexadeoxyribonucleotides of random sequence which anneal to random εites on the DNA and serve as primers for DNA syntheεiε by the Klenow Fragment of E.coli DNA polymerase I. ³²P-dCTP and three other nonlabelled nucleotides were preεent in the synthesiε which waε carried out at 37°C for 1 hr. The labelling reaction waε stopped by addition of EDTA and the unincorporated nucleotides were separated from the labelled probe by passing the reaction mixture through a prepacked G-50 sephadex column (Boehringer Mannheim) .

C) Hybridization and washing The nylon membrane waε prehybridized in Church and Gilbert buffer (1% crystalline BSA, 1 mM EDTA, 0.5 M NaHP0₄, pH 7.2, 7% SDS) for 5-10 min at 65°C, before the ³²P-labeled specific probe was added. The hybridization was carried on overnight at 65°C. The membrane was then washed with wash I buffer (2 x SSC, 0.1% SDS) 30 min twice at room temperature and with wash II buffer (0.2 x SSC, 0.1% SDS) 30 min twice at 65°C. The washing procedure was monitored by a hand held Geiger counter. The membrane was exposed to an X-ray-sensitive film (Kodak XAR5) in the presence of intensifying screens at - 70°C.

RNA Analysis

All procedures were performed in autoclaved and 0.1% diethylpyrocarbonate (DEPC) treated glassware, and sterile disposable plasticware. All solutions were either prepared in DEPC treated water or were treated with DEPC before use. DEPC is a strong inhibitor of RNAse waε added to εolutionε or water at a concentration of 1%, and allowed to εtand for 12 hr, and then autoclaved to inactivate the remaining DEPC (Kumar and Lindberg 1972) . Gloves were used for all of the experiments.

Total cellular RNA preparation

Total cellular RNA was prepared by the guanidinium isothiocyanate extraction procedure and centrifugation through a CsCl cushion (εee 2.7.1). The RNA pellets after the 16-18 hr centrifugation were waεhed gently with 70% ethanol, air dried and resuspended in 300 μl water. Ethanol precipitation with 0.5 M NaCl was performed and the RNA was either used immediately or stored in ethanol in -20°C.

Pol (A) selection for mRNA

Total cellular RNA from 5 x IO⁷ cells was incubated with Oligo(dT) cellulose (Collaborative Research Incorporated) for 2-3 hr with gentle shaking at room temperature, at a concentration of 25 mg cellulose/ml of the loading buffer

(500 mM NaCl, 20 mM Tris-HCl, pH 7.4, 10 mM EDTA, 0.2%

SDS) . The RNA-Oligo suspension was packed a in plastic dispoεable column (Bio-Rad) and waεhed with Oligo(dT) waεhing buffer (100 mM NaCl, 10 mM Tri-HCl, pH 7.4, 1 mM

EDTA, 0.2% SDS). The poly (A) fractions were eluted at

37°C with elution buffer (10 mM Tris-HCl, pH 7.4, 1 mM

EDTA, 0.2% SDS). Ethanol precipitation was performed to pellet the poly (A)^* RNA. The RNA was quantitated by meaεuring the abεorbance at 260 nm with the following formula: O_260rm x 30 x dilution factor = [RNA] μg/ml.

Northern analysis

The poly (A)* RNA fraction was resuεpended in εample buffer (50% formamide, 2.2 M formaldehyde, 1 X running buffer, 0.4% bromphenol blue), loaded on agarose/formaldehyde gel (1% agarose, 2.2 M formaldehyde) , and subjected to electrophoresis at 40 volts with circulating running buffer (20 mM MOPS, 5 mM NaAc, 1 mM EDTA) overnight at room temperature. The RNA was transferred to a nylon membrane and hybridized to a ∞P labelled probe (see 2.7.3).

Polymerase Chain Reaction (PCR) and sequencing PCR analysis

PCR analysiε (Mullis, K.B. and Faloona F.A. , 1987) waε carried out in a DNA thermal cycler (Perkin Elmer Cetuε) , for 40 cycles in 50 ul reaction mixtures containing genomic DNA prepared from mouse tissues or from CFU-GM colonies (see 2.7.2, 2.7.3), 1.25 mM of each dNTP, 1 X PCR buffer (50 mM KC1, 10 mM Tris-HCl, pH 8.3, 1.5 mM MgCl₂, 0.01% gelatin), 1 μl of each primer (300 ng/μl) , and 0.7 μl Taq polymerase (Perkin Elmer Cetuε) . The sequence of primers used in various experiments are deεcribed in Reεults. The reaction mixture was overlayed with 50 μl of mineral oil to prevent evaporation and the following thermocycle profiles were programmed: one cycle of initial denaturation at 94°C for 3 min, annealing at 55°C for 2 min, and extenεion at 72°C for 2 min was followed by 40 cycles of 94°C 1 min, 55°C 1 min, and 72°C 1 min. Detection of the products was carried out by agarose gel electrophoreεiε and by blot hybridization analyεiε (see 2 . 7 .4 ) .

Sequencing

Direct sequencing of the PCR products was carried out using moεtly εingle εtranded DNA productε generated from asymmetric PCR using a ratio of 1 to 50 of the primers (Dicker, A.P. et al., 1989). Sequence analysis was performed by the dideoxy chain termination method (Sanger, F.W. et al., 1977) with α- ³⁵S dATP using a modified T7 DNA polymerase (Sequenase version 2.0, United States Biochemicals) according to the manufacturer's instructions. After completion of the reaction, the products were heated to 95°C for 3 min, cooled on ice and loaded on a prewarmed 6% polyacrylamide urea gel. Electrophoresis was carried out at 60 W for 2 hr after which the gel was dried and exposed to Kodak XAR5 film for 16 hr.

DHFR Protein Immunoprecipitation

Cell labelling and cell extraction

Cells of 80-90% confluence were incubated in Met-free medium for 1 hr in a 37°c C0₂ incubator and the medium was replaced by Met-free medium containing 0.3 mCi/ml ³⁵S-Met (0.5 ml/60 mm petri dish). After incubation for 3-4 hr with occasional shaking, the cells were washed 2 times with PBS, and the cell extracts made by adding 0.6 ml ice cold lysiε buffer (50 mM Triε-HCl, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1% Triton X-100, 1% Na-Deoxycholate, 0.1% SDS) with freεhly added protease inhibitors (PMSF 0.2 mg/ml, Leupeptin 0.05 mg/ml) per 60 mm dish. The lyεed cells were sonicated in an ice bath, 30 sec x 2 with a 10 sec interval and centrifuged for 30 min in a microfuge at 4°C. The supernatantε (about 0.5 ml) were εtored at -70°C or uεed freεh.

TCA precipitation

TCA precipitation was used to quantitate the concentration of the labelled protein. 5 μl of cell extract waε mixed with 100 μl BSA (1 mg/ml) and 1 ml 10% TCA and incubated on ice for 30 min. The εample waε paεεed through Whatman filter paper (934 AH) pre- equilibrated with 10% TCA. The filter was rinsed twice with 10% TCA and 95% ethanol, air dried and 5 ml scintillation fluid (BCS, Amersham) added and radioactivity measured in a scintillation counter.

Preparation of protein A sepharose-Ab complex

Rabbit anti-human DHFR antibody was used for immunoprecipitations. Protein A sepharose (PAS) was allowed to swell in sepharose buffer (20 mM Tris-HCl, pH 7.5) at a concentration of 6 mg/ml and washed 3 times in the same buffer by centrifugation at 1,700 g for 3 min. The antibody was added to the εepharoεe suspension at various dilution and the mixture was rotated at 4°C for 2-3 hr. The protein A sepharose-Ab complex was pelleted and washed 3 times with Wash A buffer (20 mM Tris-HC, pH 7.5, 150 mM NaCl, 10% glycerol, 0.1% Triton X-100). Aliquots (0.5 ml) were pipetted into 1.5 ml eppendorf tubes and spun down.

PAS-Ab-Antigen complex formation and purification

PAS-Ab complex pellets were suspended in the equal amount of labelled cell extracts (see 2.8.2) in a volume of 0.5 ml. The mixtures were rotated at 4°C overnight to allow the Ag-Ab reaction to take place. The PAS-Ab-Ag complex was pelleted by centrifugation at 4°C for 5 min in a microfuge. The pellet was washed with 1 ml Waεh B buffer once (50 mM Tris-HCl, pH 8.0, 500 mM NaCl, 5 mM EDTA, 0.2% Triton X-100), Wash C buffer 3 timeε (50 mM Tris- HCl, pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.1% Triton X-100, 0.1% SDS), and Waεh D buffer once (10 mM Triε-HCl, pH 8.0, 0.1% Triton X-100).

SDS polyacrylamide gel electrophoresis

The PAS-Ab-Ag washed pellet was reεuεpended in εample buffer (250 mM Triε-HCl, pH 6.8, 50% glycerol, 5% SDS, 5% β-mercaptoethanol, 0.25% bromophenol blue, 20 μl/sample) , boiled for 10 min, chilled on ice, spun briefly, and the supernatant was loaded on an acrylamide minigel (7 cm x 8 cm) , consisting of the separating gel (15% acrylamide, 0.375 M Tris-HCl, pH 8.8, 0.1% SDS, 0.4% N,N'-Methylene- bis-acrylamide (Bis)) and the stacking gel(3% acrylamide, 0.125 M Tris-HCl, pH 6.8, 0.1% SDS, 0.08% Bis). The gel electrophoresiε waε run at 200 voltε, for 45-50 min in SDS-PAGE buffer (25 mM Triε-HCl, pH 8.3, 0.192 M Glycine, 0.1% SDS). The gel waε fixed in 7% Acetic acid and 25% Methanol for 20 min at room temperature with gentle shaking, enhanced with Enhancer (NEN) for 30 min, rinεed with water, dried with a gel drier (Bio-Rad, Model 583) at 60°C for 1 hr, and expoεed to X-ray film at -70°C with intenεifying screens.

MTX cytotoxicity Assays Colony formation assay for 3T3 cells

Parental and the transduced 3T3 cells pooled from the

G418 resiεtant colonieε were plated at 10³ cell/100 mm plate. Variouε concentrations of MTX were added after 16-18 hr. Drug-containing media was changed every 3-4 dayε. The reεiεtant colonies were scored 8 to 10 days after MTX treatment.

Cell growth inhibition assay for leukemia cells

Exponentially growing parental and transduced human leukemia cells (CEM, K562, Raji) were exposed to different concentrations of MTX at an initial density of 2xl0⁵ cells per ml. The cellε were counted after 3 days of treatment with a hemocytometer after trypan-blue staining to exclude dead cells. The fetal bovine serum used in the MTX selection medium was treated with thymidine phosphorylase at 6 units per ml for 5 min at 37°C to reduce the background in the presence of MTX (Li et al., 1990) .

Granulocyte-Macrophage (CFU-GM) assay for mouse bone marrow progenitor colonies

Immediately after coculture (see 2.5.3) and at regular intervals after bone marrow transplant(εee 2.12), CFU-GM assays were performed. 10^s marrow cells were plated in a grid petri dish (10 x 35 mm) in 2 ml of methylcellulose IMDM cocktail medium (1% α-methylcellulose, 20% FCS treated with thymidine phosphorylase, 10% WEHI-3B CM, 1% NaHC0₃, 1% Na-pyruvate, 1 μM β-mercaptoethanol, 100 units/ml penicillin, 100 μg/ml εtreptomycin, 1% essential amino acids, 1.5% nonessential amino acids, 0.5% vitamin C) . MTX of different concentrations was added to the cocktail medium. The CFU-GM colonies greater than 50 cells were scored after 12-14 days culture in a 37°C C0₂ incubator. Bone Marrow Transplantation (BMT)

2 x IO⁶ donor marrow cells suspended in 0.2 ml IMDM, after coculture with an amphotropic virus producing cell line (see 2.5.3) were injected through the tail vein into recipient mice irradiated with 900 rads 24 hr before the BMT. In vivo εelection with MTX waε performed with different doεe schedules (see Chapter 3 Results) . A typical MTX treatment protocol was as follows: one day after BMT, MTX was given twice weekly i.p. at 2 mg/kg body weight for the firεt week and 5 mg/kg for the next 7 weeks. Hematocrit, WBC countε, body weight and overall εurvival rates were monitored throughout the course of the experiment. The recipient mice were provided with water containing streptomycin (1 g/liter) .

Cell Line Maintenance

The murine TK^' 3T3 fibroblast lines, and the E86 and AM12 packaging lineε were maintained in DME with 10% FBS. Human leukemia lineε CEM, K562 and Raji were maintained in RPMI with 10%, 10%, and 15% FBS, reεpectively. 100 units/ml penicillin and 100 μg /ml streptomycin were also present in the media. Bone marrow culture for retroviral infection (see 2.5.3) and for the CFU-GM assay (see 2.9.3) was performed in IMDM media.

The E86 and AM12 cell lines were selected for 48 hr in selection media after long term storage to maintain the efficiency of packaging. The E86 cell lines were selected in HXM medium (DME with 10% FBS, and 15 μg/ml Hypoxanthine, 250 μg/ml Xanthine, 25 μg/ml Mycophenolic acid) and the AM12 cell lines were selected in HXMB medium (HXM medium with the addition of 200 μg/ml Hygromycin B) . WEHI-3B cells were grown in IMDM medium with 10% FCS. After 3-4 dayε growth (initial concentration 3 x 10⁵ cellε/ml) , the cellε were εpun down at 1,200 rpm and the εupernatant waε collected (WEHI-3B conditioned medium (CM)) and stored at -20°C.

Data Calculation and Analysis

The IC50s were calculated from cell survival εtudies in the absence and presence of drug using the Median Effect Equation (Chou and Talalay, 1984).

EXPERIMENTAL RESULTS

Vector Design

The construction of double copy (DC) vectors

The retroviral vectors used to introduce the altered DHFR gene into recipient cellε are baεed on the high titer retroviral vector N2, which in turn waε derived from the moloney murine leukemia virus containing the bacterial transpoεon Tn5 neomycin resistance gene (NEO) under the control of the viral LTR promoter (Armentano et al., 1987) . The N2 vector was modified by the insertion of a 52-bp polylinker (containing the unique restriction siteε 5'-ApaI-BglII-SnaBI-ScaII-MluI-3') into the U3 region of the 3'LTR. The polylinker-modified N2 vector waε designated as N2A (Hantzopoulos et al., 1989). The U3 region of the MoMLV 3'LTR serves as the template for the syntheεiε of both the 5' and 3' U3 regions of the provirus LTRs (Figure 10, see review by Varmus and Swanstrom, 1982) . The foreign sequence inserted at thiε position will be duplicated to the U3 region of the 5'LTR upstream of the LTR initiated transcription unit. Therefore such vectors were termed double copy or DC vectors because in the integrated provirus the foreign sequences are present in two copies (Hantzopoulus et al., 1989; Sullenger et al., 1990). The N2A vector waε further modified by inserting a poly A fragment in the Apal site in the polylinker to allow insertion of foreign sequence in the anti-parallel orientation of the LTR transcription.

In this εtudy, the SV40, human β-actin, human ADA, herpeε viruε TK, and the cytomegalovirus promoters were cloned into the polylinker region in either a parallel or an anti-parallel orientation, generating DC vectors designated as DC/SV, DC/AC, DC/SV/R, DC/AD/R, DC/TK/R and DC/CMV/R. The design of these vectors iε outlined in Figs. 6A-6D and described in detail in Materials and Methods (2.3).

The mesεenger RNA εpecies that would be transcribed from the integrated proviruε of the DC vector carrying foreign sequences are shown in Figure 11.

The construction of retroviral vectors carrying a murine mutant DHFR cDNA.

Figs. 7A-7E shows the cloning strategy used to generate retroviral constructs that contain a murine 3T6 DHFR with a Leu to Arg mutation at residue 22 of the DHFR enzyme. In the DC/SV/R-mDHFR construct, the DHFR cDNA should be transcribed from the SV40 promoter in the parallel orientation with a predicted mRNA length of 1.1 kb. In the DC/SV-mDHFR conεtruct the DHFR iε transcribed from the same promoter with similar length message as in DC/SV-mDHFR, but in the antiparallel orientation of the viral transcription unit. In DC/AD/R-mDHFR, DC/TK/R- mDHFR and DC/CMV/R-mDHFR, the DHFR transcriptional unit is in the antiparallel orientation and the length of the predicted transcripts is 0.8 kb.

Vector DNA waε converted into correεponding viruε by the procedure described in Material and Methods.

Note that for all the constructs described above, the promoter-DHFR fragments were inεerted into the 3'LTR of N2A or N2AP so that in the infected cell the promoter- DHFR template will be duplicated and be present in both LTRs of the proviral DNA as discussed above. The importance of this duplication is that the second copy of the promoter-DHFR template preεent in the 5'LTR is placed outεide the viral transcription unit, therefore avoiding possible adverse effects of the active LTR promoter on the promoter-mDHFR transcriptional unit (Emerman and Temin, 1984). This strategy has been previously shown to greatly facilitate the expression of recombinant genes (Hantzopoulos et al., 1989 and Sullenger et al., 1990). Furthermore, in constructε in which the promoter-DHFR was cloned in antiparallel orientation of the viral LTR transcription, the two copies of the template are transcribed against LTR transcription, further reducing the negative effects of the LTR transcriptional unit (Emerman and Temin,1984, Proudfoot, 1986 and Cullen et al., 1984). A similar design of retroviral vectors haε resulted in the enhanced expresεion of the β-globin gene (Karlεεon et al.,1987).

The viral titers determined by the number of G418 resistant colonies following the infection of virus with both parallel or antiparallel constructs were approximately the same, ranging from 4x10* to 5xl0⁵ CFU/ml.

Expression and Resistance Produced bv the Mutant DHFR under Control of Different Promoters in NIH 3T3 Cells

MTX resistance of the transduced 3T3 cells

The MTX resistance level of the G418 resistant 3T3 cells transduced by the mutant mDHFR viral constructε waε determined by meaεuring the inhibitory effect of MTX on colony formation of 3T3 cellε. MTX concentrationε that inhibit 50% of colony formation of the tranεduced cellε (IC50) were compared with the non-transduced parental cells to determine the resistance ratio. The DC/SV-mDHFR and DC/SV/R-mDHFR constructε afforded εimilar levels of reεiεtance and conferred the highest level of resistance when compared to constructε with promoterε other than SV40 (Table 1, εee below) .

Although the infection of the 3T3 cellε with the mutant DHFR conεtructε give approximately equal number of reεiεtant colonieε in either G418 or MTX εelection, only the G418 resistant cells were pooled and tested for the level of MTX resistance (IC50) to avoid the complication of increasing MTX reεiεtance due to other mechaniεmε that might arise during the process of selection.

RNA analysis of the transduced 3T3 cells

The poly A fraction of the total cellular RNA isolated from NIH3T3 cell lines infected with mutant mDHFR vectors was subjected to electrophoresis on 1% formaldehyde- agarose gels, transferred to a nylon membrane, and hybridized with a 3T6 DHFR probe (Figure 12A) . The two low-molecular weight RNA species of 1.1 kb and 0.8 kb are the DHFR transcripts from the internal promoters. The high molecular weight species are transcripts from the viral LTR promoter in the spliced or nonspliced formε or the read-through from the internal promoters (Figure 11) . The same RNA blots were rehybridized to a human glyceraldehyde-3-phosphate dehydrogenaεe (GAPDH) probe as a quantitative control of the amount of RNA loaded in each lane (Figure 12B) .

The high molecular weight RNA species in the DC/SV/R- mDHFR construct waε several fold less than the DHFR transcriptε from the internal promoter, when compared with the DC/SV-mDHFR construct which had a similar level of transcription for both species. To further investigate the phenomenon, RNA from 3T3 cellε infected with virus-containing supernatant collected from four additional individual producer lines of both DC/SV-mDHFR and DC/SV/R-mDHFR waε analyzed (Figs. 13A-13B) . A similar pattern of inhibition of the high molecular weight εpecieε tranεcripts in the DC/SV/R-mDHFR constructs was observed, indicating that antiparallel tranεcription from the internal promoter may have inhibitory effects on the LTR transcription.

Expression and Resiεtance Produced bv the Mutant DHFR under Control of Different Promoters in Human Leukemia Cell Lines

MTX resistance of transduced leukemia cells

The MTX resistance level of the transduced leukemia cells was determined by a cell growth inhibition assay described in Material and Methods. The MTX IC50s were calculated and compared to that of the parental cells (Table 2) . Different levels of reεiεtance were observed in different cell lines with different constructs. The DC/TK/R-mDHFR construct produced the highest level of resistance and transcriptε in K562 and Raji cellε, but not in the CEM cell line. In the CEM cell line the DC/AC-mDHFR produced the higheεt level (2.1 fold) of MTX resistance.

RNA and DNA analysis of the transduced leukemia cells.

RNA analysiε by Northern blotting of the tranεduced leukemia cell lineε by variouε conεtructε containing mutant DHFR iε εhown in Figs. 14A-14B. The mDHFR transcriptε from the internal promoter have the expected length of 1.1 and 0.8 kb. The εtructure of the proviral DNA in the infected leukemia cells was analyzed by DNA blotting. The Dral restriction enzyme has a unique digeεtion εite in the mDHFR cDNA. Thuε only faithful duplication of the mDHFR cDNA sequence in the vector can generate a DNA fragment approximately 4 kb upon digestion of the genomic DNA with Dral, as was observed for all the conεtructε in the three leukemia cell lineε (Figε. 15A-15B) . The exact size of the 4 kb fragment varied since it consiεted of the vector sequence preεent between the two dral sites in the U3 regionε. Thiε includeε one copy of the mDHFR cDNA, one copy of the NEO gene, and one copy of the promoter in the construct (0.4 kb for SV, 0.8 kb for AD and TK, 1.3 kb for AC) and one copy of the additional poly (A) εignal εequence (0.6 kb for DC/SV/R, 0.3 kb for DC/AD/R and DC/TK/R) for the antiparallel conεtructε (εee 2.3.1 and Figε. 6A-6D) . In Raji cellε, additional minor lower molecular weight bandε were preεent. Thee significance of these bands iε not known.

Expression and Resistance Produced bv the Mutant DHFR under Control of Different Promoters in Murine Bone Marrow CFU-GM Colonies

Conditions for MTX selection in the CFU-GM assay.

The dose-response of CFU-GM to MTX in different selection media was carried out to establish conditions for MTX selection. Five different concentrations of MTX (0, 10, 100, 1,000, and 10,000 nM) were tested in medium containing regular FBS or dialysed FBS or FBS treated with thymidine phoεphorylaεe (TP) (εee 2.11). The medium containing 100 nM MTX and TP treated FBS in which untransduced bone marrow cellε did not produce background colonieε waε chosen aε the εelection condition for the MTX reεiεtance teεt (Figure 16) .

Comparison of the expression of the mDHFR under control of different promoters in the CFU-GM colony assay.

The MTX resiεtance level of the bone marrow cellε tranεduced through coculture with the viral producer cells was determined by the CFU-GM colony aεεay (Table 3) . Conεtructε with the SV promoter and the AD promoter produced a higher percentage of MTX reεiεtant colonieε than constructs with the TK or AC promoter. Based on these resultε the DC/SV/R-mDHFR and DC/AD/R-mDHFR conεtructε were used in the in vivo bone marrow transplantation studies.

Table 1. MTX resiεtance level in 3T3 cellε infected with mutant mDHFR in retroviral vectorε.

NIH 3T3 cellε were infected with the 3T6 DHFR in retroviral vectorε aε indicated. The numbers following the vector indicate different producer lines from which the infectiouε supernatants were collected. Infection and subsequent G418 selection were performed as deεcribed in Material and Methods. The G418 resistant colonies were pooled and expanded in drug-free medium. The G418 resiεtant cellε were incubated with different concentrationε of MTX and MTX resistant colonies were counted as described in Material and Methods. The IC₅₀ value waε calculated (Chou and Talalay, 1984), uεing 5 to 7 MTX concentrations. The IC₅₀ values shown in the table are the average of 3-5 independent experiments. vector MTX IC₅₀ ^"(nM) resistance ratio none 7.2 1

DC/SV-mDHFR 1.1 79.3 11.1

DC/SV-mDHFR 1.5 73.7 10.3

DC/SV-mDHFR 2 . 1 147.4 20.6

DC/SV/R-mDHFR 1.1 118.5 16.6 DC/SV/R-mDHFR 1.6 56.3 7.9

DC/SV/R-mDHFR 2.2 135.0 18.9

DC/AD/R-mDHFR 2 17.5 2.5

DC/AD/R-mDHFR 7 11.7 1.6

DC/TK/R-mDHFR 5 13.5 1.9 DC/TK/R-mDHFR 9 20.5 2.9

DC/AC-mDHFR 10 13.5 1.9

DC/CMV/R-mDHFR 1 9.2 1.3

DC/CMV/R-mDHFR 2 12.4 1.7

* Average of 3-5 experiments.

Table 2. MTX reεistance level in leukemia cells infected with mutant mDHFR in retroviral vectors.

vector CEM K562 Rai

IC50+SD ratio IC50+SD ratio IC50+SD ratio (nM) (nM) (nM) none 49.7+ 1 50.8+ 1 34.1+ 1 10.5 3.2 2.8

DC/SV- mDHFRl.1 61.7+ 1.2 104.9+ 2.1 61.0+ 1.8 7.6 56.2 1.1

DC/SV/R- mDHFRl.1 67.8+ 1.4 123+ 2.4 97.1+ 2.8 1.7 14.6 7.6

DC/AD/R- mDHFR2.6 84.5+ 1.7 94.6+ 1.9 54.0+ 1.6 3.1 1.9 7.2

DC/TK/R- mDHFR5.2 49.2+ 1 172.1+ 3.4 190+ 5.6 4.7 70.9 132

DC/AC- mDHFR 10/2.3 106.3+ 2.1 91.7+ 1.8 50.3+ 1.5

3.3 30.3 4.9

CD50s are the average of 3 experiments.

Three human leukemia cell lines (CEM, K562, Raji) were infected with 3T6 DHFR in retroviral vectorε. The numberε following the vector indicate different producer lineε from which the infectious εupernatant waε collected. Infection and subsequent G418 selection were performed as deεcribed in Material and Methodε. The G418 resistant cells were pooled and expanded in drug-free medium before being seeded at a density of 2x10^s cells per ml and expoεed to different concentrations of MTX. Each concentration was tested in triplicate. Cellε were counted after 3 dayε of exposure. The IC₅₀ values shown are the average of 3 independent experiments.

Table 3. MTX resiεtant CFU-GM of bone marrow cellε infected with mutant mDHFR in retroviral vectors. Mouse bone marrow cells were infected with 3T6 mutant DHFR in retroviral vectors. The numberε following the vector indicate different producer lines cocultured with the marrow cells. Immediately after coculture the CFU-GM assay was performed in the presence or absence of MTX as described in Material and Methods. The percentage of resiεtant colonieε waε calculated by dividing the MTX reεiεtant colony number with the colony number in the absence of MTX. Untransduced bone marrow waε uεed aε control.

Table 3

vector No. of colonieε percentage of -MTX +MTX resistant colonies (100 nM) control 394 0 0

DC/SV-mDHFRl.l 393 46 11.7

DC/SV/R-mDHFRl.1 386 48 12.4

DC/AD/R-mDHFR2.6 379 51 13.5

DC/TK/R-mDHFR5.2 336 9 2.7

DC/AC-mDHFR10/2.3 300 7 2.3

MTX Resistance Developed in Mice After the Transplantation of the Bone Marrow Cells Transduced by DC/SV/R-mDHFR and DC/AD/R-mDHFR

Protocol for in vivo bone marrow transplantation (BMT) studies (Figure 17).

The donor bone marrow cells were cocultured 48 h with the pre-irradiated producer cell lines: DC/SV/R-mDHFR and DC/AD/R-mDHFR, and AM12, the parental packaging cell line used as a control. After coculture, 2 X IO⁶ bone marrow cells were transplanted into each recipient mouse irradiated with 900 R 24 hr before transplantation. MTX treatment was started after the BMT, with the following dose schedules:

A) Low-dose: 48 hr after BMT, MTX of 2 mg per kg body weight, was adminiεtered i.p. twice for the firεt week, and 5 mg per kg body weight twice a week for the reεt of the experiment.

B) Delayed high-doεe: MTX treatment εtarted 4 week after the BMT at a doεe of 200 mg per kg body weight twice a week, i.p.

MTX resistance in mice after gene transfer.

Recipient mice were transplanted with tranεduced or untransduced marrow cells and treated with MTX under low- doεe εelection schedule. The survival rates of the recipient mice from two experiments were shown in Figs. 18A-18B. Mice receiving marrow cells cocultured with the AM12 control line did not survive the low-dose selection and died in the first 30 days of the selection, while more than 80% of mice receiving marrow cells transduced by DC/SV/R-mDHFR or DC/AD/R-mDHFR survived. Mice without BMT after 900R irradiation died within 2 weeks with or without MTX selection, while all irradiated mice with BMT but without MTX selection were alive (data not shown) . The changes of the hematocrit and the white blood cell count (WBC) following the tranεplantation and treatment with MTX are εeen in Figε. 19A-19B. At day 13, the hematocrit and the WBC were decreaεed markedly in the control animals. There was also a decrease in the transduced animals. At day 28 post transplant, complete recovery of the hematocrit and the WBC were noted in the εurviving animals with transduced marrow.

Deaths that occurred in the control group (AM12, non- virus producing) or the group receiving tranεduced marrow were associated with severe anemia, Gl bleeding and marked weight loss.

The surviving mice with the transduced marrow were either treated with a lower dose of MTX (5 mg/kg, twice a week) for over 6 months without evidence of toxicity or used as the donor for the εecond generation tranεplantation.

The second transplant was carried out 5 weeks after the primary BMT, uεing marrow from DC/SV/R-mDHFR mice. The secondary recipients were treated as before with low dose MTX, and all recipients survived the selection. A third tranεplantation waε performed 5 monthε after the εecond tranεplant. Marrow from the εecondary recipientε waε uεed. The tertiary recipientε were treated as before with low dose MTX and animalε εurvived for longer than 30 dayε (Figε. 20A-20C) .

Sequential CFU-GM assays were carried out in mice receiving tranεduced marrow at 20, 34, 45 days after

SUBSTITUTE SHEET (RULE 26} primary BMT and 30 and 138 days after the secondary BMT. The in vivo MTX selection resulted in an enrichment for MTX reεiεtant colonies, shown by the increased percentage of reεistant CFU-GM colonies with the increase of MTX selection time (Figs. 21A-21B) .

Recipient mice transplanted with DC/SV/R-mDHFR infected marrow were able to survive the high-dose MTX selection that was started 4 weeks after BMT (Figure 22A) . The control group died within 2 weeks of MTX treatment at 100 mg/kg, twice a week, while 3/5 mice in the DC/SV/R-mDHFR group survived 4 weeks of 100 mg/kg twice a week and 6 weeks of 200 mg/kg twice a week MTX. The MTX toxicity on normal mouse without irradiation or BMT was εhown in Figure 22B. 4/4 normal mouεe died within 4 weekε of MTX treatment at 200 mg/kg twice a week.

Demonstration of the integration of retroviral vector carrying the mutant mDHFR in the recipient mice

Genomic DNA from the εpleen, liver, peripheral blood cellε (PBC) and bone marrow were extracted aε deεcribed in the Material and Methods (2.7) . Primerε NEOl and NE02 (Table 4) were uεed to amplify the NEO gene by PCR in the proviral DNA, generating a 415 bp fragment. The productε were analyεed by agaroεe gel electrophoreεiε, then tranεferred to a nylon membrane and hybridized to a NEO probe. The preεence of NEO sequence in mouse tissues at different time points after BMT with different MTX selection schedules are shown in Figs. 23A-23C. Table 4. The sequences of primers uεed for PCR amplification and sequencing. primer εequence annealε to Neol 5'-GGAAGCCGGTCTTGTCGATC-3' NEO(coding) (Sequence ID No. 5)

Neo2 5'-CGAAATCTCGTGATGGCAGG-3' NEO(non-coding) (Sequence ID No. 6)

M301 5'-TGCCAATTCCGGTTGTTCAAT-3' mDHFR (coding) (Sequence ID No. 7)

M210 5'-TCTGTCCTTTAAAGGTCG-3' mDHFR (coding) (Sequence ID No. 8)

H250 5'-GAGGTTCCTTGAGTTCTCTGC-3' hDHFR (coding) (Sequence ID No. 9)

GT-NC1 5'-CCTCGGCCTCTGAGCTAT-3' SV40 promoter (NC, nt -50 to -68)

(Sequence ID No. 10)

The integration of the proviral DNA in MTX reεiεtant CFU- GM colonies of recipient mice after the primary BMT and the secondary BMT were also analysed by PCR. DNA from 5- 6 colonies were extracted and amplified with NEOl and NE02 and hybridized to NEO probe (Figure 24) .

DNA from PBC of a secondary recipient 8 months after BMT, and from εpleen and liver 5 weekε after primary BMT was amplified by asymmetric PCR (Dicker et al., 1989, 2.9.2) with primer GT-NC1, which covered a region spanning part of the SV40 promoter in the vector, and primer M301 in the mDHFR (Table 4) . The PCR product was sequenced with primer M210 and revealed the presence of the T to G point mutation in the mDHFR of the PBC in the recipient mouεe 8 months after the initial BMT (Figs. 25A-25B) . The direct εouthern analyεiε of the genomic DNA extracted from tiεεueε of the recipientε, however, did not show the integration of the proviral DNA after repeated trials (Figs. 26A-26B) . One of the posεible explanations for this reεult iε that only a small percentage of cells in recipient spleen and bone marrow tissues contained the proviral DNA. The detection limit of the southern analysiε conducted was between 10% and 3.3% (Figure 27), which would indicate that less than 3.3% of cellε in the recipient tiεεueε contained the proviral DNA.

The doεe reεponse of normal mice (without irradiation or BMT) to MTX toxicity was shown in B. The animals were injected (ip) with MTX at the indicated weekly doεeε, which were divided into 2 and adminiεtrated twice a week, for 4 weekε. There were 4 mice in each group.

construction of Retroviral Vectors Carrying Human Mutant DHFR (Figures 8A-8H)

Double copy vectors carrying full length cDNAs of human mutant DHFR (hDHFR) with mutations at residues 31 or 34 (Phe to Ser) under control of the SV40 promoter and the human ADA promoter were constructed and are deεignated aε DC/SV-hDHFR, DC/SV/R-hDHFR, and DC/AD/R-hDHFR (see 2.3.3). The full length cDNA contains 560 bp coding region and 240 bp 3' non-coding region. The SV40 and ADA promoter were chosen because of the studies described above using the 3T6 mDHFR. The expected length of hDHFR transcriptε is 1.2 kb from the parallel promoter and 1.0 kb from the antiparallel promoters.

The εame two promoterε were uεed in conεtructing DC vectorε carrying leεε than full length cDNA of the mutant hDHFR containing the coding region and 95 bp 3'non-coding region. Constructs carrying the cDNA fragment between Ncol and Bglll site from the pKT7HDR, which was modified immediately before the ATG εtarting codon (see 2.3.3), were prepared and deεignated aε DC/SV-hDHFR(NB) , DC/SV/R- hDHFR(NB) and DC/AD/R-hDHFR(NB) . Constructs carrying the cDNA fragment between Hindlll and Bglll site from pSV4HDR without modification before ATG were alεo prepared and named aε DC/SV-hDHFR(HB) , DC/SV/R-hDHFR(HB) and DC/AD/R- hDHFR(HB) (Figε. 8A-8H) . The expected length of the hDHFR tranεcript iε 1.0 kb from the parallel promoter and 0.8 kb from the antiparallel promoters.

These constructs were packaged to corresponding virus by the procedure described in Material and Methods (2.4) .

Expreεεion _ ύ WLX Resistance Produced bv Viral

Constructε Containing Mutant hDHFR cDNAs in 3T3 Cells

The expresεion of full length cDNA of hDHFR in tranεduced 3T3 cells

3T3 cells infected with DC/SV-hDHFR, DC/SV/R-hDHFR or DC/AD/R-hDHFR were incubated with either G418 (0.75 mg/ml) or MTX (1.5 X 10'⁷M) for 8-10 days. Cells infected with DC/SV-hDHFR were able to survive the selection by either G418 or MTX, while cells infected with DC/SV/R- hDHFR or DC/AD/R-hDHFR were able to survive only under the G418 selection but not the MTX selection. The MTX resiεtance level of theεe transduced 3T3 cells was determined by measuring the inhibitory effect of MTX on the colony formation of 3T3 cells. The IC50 of the cells transduced by mutant hDHFR were compared to that of the parental 3T3 cellε and the cellε transduced by wild type hDHFR in DC/SV vector constructs. Only the cells transduced by DC/SV-hDHFR31 had survived MTX selection (DC/SV-hDHFR31/mr, mr indicates MTX resiεtant) and were found to be reεiεtant to MTX (Table 5) . Table 5. IC₅₀ for MTX in 3T3 cells infected with full length mutant hDHFR cDNA in retroviral vectors. 3T3 cells were infected with the retroviral vectors carrying full length hDHFR (wild type, S31 or S34 mutatation) . The infected cells were εelected in either G 18 (0.75 mg/ml) or MTX (1.5 X 107 M) . While cellε tranεduced by all constructs survived the G418 selection, only the DC/SV- hDHFR31 tranεduced 3T3 cellε εurvived the MTX selection. The resiεtant cellε from each infection and εelection (mr εtandε for MTX resistance) were pooled, expanded in drug- free medium, and plated out in different concentrations of MTX. The resiεtant colonieε were counted and the IC_S0 values calculated, using 5 to 7 MTX concentrations.

Table 5

vector MTX IC₅₀(nM)

DC/SV-hDHFR 19.0

DC/SV-hDHFR31 24.9 DC/SV-hDHFR31/mr 74.3

DC/SV/R-hDHFR31 10.8

DC/AD/R-hDHFR31 15.2

DC/SV-hDHFR34 13.1

DC/SV/R-hDHFR34 15.5 DC/AD/R-hDHFR34 20.8 Northern analyεis of the poly (A) fraction of the total cellular RNA from these transduced 3T3 cell lines is shown in Figure 28A. The cell lines transduced by DC/SV- mDHFR and DC/AD/R-mDHFR were used as controls in the experiment. The length of the mDHFR transcriptε in these two constructε are 1.1 and 0.8 kb. The length of hDHFR tranεcriptε from DC/SV-hDHFR, DC/SV/R-hDHFR and DC/AD/R-hDHFR were εhorter than the expected length of 1.2 and 1.0 kb with the exception of DC/SV-hDHFR/mr. A full length tranεcript of 1.2 kb iε preεent in DC/SV- hDHFR/mr aε well aε the truncated message.

The DNA analysis of these transduced cell lines did not reveal any gross recombination in the proviral DNA (Figure 28B) .

Immunoprecipitation of the DHFR enzyme protein with a polyclonal antibody in the transduced cell lineε was carried out to determine whether the transcripts were translated to a full length or shorter protein. The antibody was titered on DG44 cells transduced by hDHFR. The DG44 cell line is a CHO cell line lacking endogenous DHFR (Urlaub and Chasin, 1980) . The ³⁵S-labelled cell extractε were incubated with the antibody bound to protein A sepharose and the precipitated proteins were εeparated on 15% SDS polyacrylamide gel. The 22 kd DHFR protein waε detected at dilution of 1:50 (antibody:cell extract) . The 40 kd band iε a non-εpecific protein precipitation aε it iε alεo preεent in the non- transformed DG44 cells (Figure 29A) . The immunoprecipitation of the transduced 3T3 cell lines is shown in Figure 29B. Unlike the DG44 line, 3T3 cellε contain endogenouε mDHFR to which the polyclonal antibody to hDHFR alεo cross-reacts. By using the 40 kd non- εpecific precipitation (band 2) aε a control for loading, the relative ratio of the denεity of the 22 kd εpecific precipitation (band 1) to that of the band 2 waε calculated, giving a semiquantitative measurement for the DHFR in the transduced cells. The DC/SV-hDHFR/mr which express the full length message at a detectable level produced a higher level of DHFR protein than untransduced 3T3 or 3T3 cell transduced by DC/SV/R vector without hDHFR gene. No smaller proteins were detected, indicating only the full length hDHFR mesεage waε able to be tranεlated into the protein which could be precipitated by the antibody.

Expression and resistance produced by the less than full length cDNA of mutant hDHFR

The MTX resistance level of the 3T3 cells transduced by less than full length cDNA of mutant hDHFR was determined by measuring the IC50 of MTX on colony formation of 3T3 cells. The MTX IC50S on DC/SV-hDHFR31(HB) , DC/SV- hDHFR31(NB) and DC/AD/R-hDHFR31(NB) were significantly higher than the untransduced parental 3T3 cell line or the 3T3 cell lines tranεduced by the wild type hDHFR (Table 6) .

Table 6. MTX reεiεtance level in 3T3 cellε infected with less than full length mutant hDHFR cDNA in retroviral vectors. The 3T3 cells infected with retroviral vectors carrying less than full length mutant hDHFR cDNA (S31) were selected in G418 selection medium, and the resiεtant colonieε were pooled. The inhibitory effect of MTX on the colony formation of theεe G418 resistant cells was measured aε deεcribed in Table 5. The IC₅₀ valueε were the average of 3 experimentε and the εtandard deviations are shown. The parental 3T3 cell line and the cell line transduced by wild type hDHFR in retroviral vector (DC/SV-hDHFR) were used as controls.

Table 6

vector MTX IC₅₀+SD resiεtance ratio

(nM)

3T3 16.0+3.0 1

DC/SV-hDHFR 19.0 1.2

DC/SV-hDHFR31(HB) 77.0^*±17.7 4.8 DC/SV/R-hDHFR31(HB) 45.5+16.6 2.8

DC/SV-hDHFR31(NB) 67.7^*+14.1 4.2

DC/SV/R-hDHFR31(NB) 20.4+4.0 1.3

DC/AD/R-hDHFR31(NB) 54.7^*+15.8 3.4

'significantly different from 3T3 parental line

Semiquantitation of the DHFR protein in these cell lines were carried out by immnunoprecipitation (see Material and Methods, and 3.7.1). The DC/SV-hDHFR31(HB) , and DC/SV-hDHFR31(NB) and DC/AD/R-hDHFR31(NB) have a higher DHFR content than parental 3T3 cellε (Table 7) . The DHFR protein content εeems to correlate well with the MTX resiεtance level of theεe cell lineε.

Table 7. Quantitation of DHFR by immunoprecipitation from 3T3 cellε infected with mutant hDHFR in retroviral vectorε. The immunoprecipitation of the DHFR enzyme protein from the transduced 3T3 cells is described in Material and Methods (2.10) and in the legend of Figs. 23A-23C. The intensities of the two bands (bandl of 22 kd and band2 of 40 kd) on the X-ray film were measured by Gelscan XL (2.0) (LKB). The ratio of bandl over band2 was calculated to εemiquantitate the DHFR enzyme in the cell.

Table 7

vector bandl/band2'

3T3 1.2

DC/SV-hDHFR31(HB) 2.4 DC/SV/R-hDHFR31(HB) 1.0 DC/SV-hDHFR31(NB) 2.4 DC/SV/R-hDHFR31(NB) 1.3 DC/AD/R-hDHFR31(NB) 1.7

*bandl: specific DHFR precipitation band2: non-εpecific precipitation

The RNA and DNA analysis of the following transduced cell lineε were performed: εix DC/SV-hDHFR31(HB) lines (infected by viral supernatant collected from three producer lines, and selected by either G418 or MTX) , three DC/SV/R-hDHFR31(HB) lines (infected by viral εupernatant collected from three producer lineε, and εurviving the εelection of G418) , four DC/SV-hDHFR31(NB) lines (infected by viral supernatant of two producer lines, and εurviving the selection by either G418 or MTX), one DC/SV/R-hDHFR31(NB) line (infected by viral supernatant of one producer lineε, and εurviving the G418 εelection), and two DC/AD/R-hDHFR31(NB) lineε (infected by viral supernatant of two producer lineε, and εurviving the G418 selection) .

Figs. 30A-30B shows the RNA analysis of these cell lines. The DC/SV-mDHFR and DC/AD/R-mDHFR were uεed aε positive controls for the DHFR transcript from the internal promoters and were 1.1 and 0.8 kb respectively. The 3T3 cell line was uεed aε the negative control. The poly A fraction of total cellular RNA was separated on formaldehyde-agarose gel, transferred to a nylon membrane and hybridized with hDHFR31(HB) cDNA probe (A) and rehybridized with GAPDH (B) . The 1.1 kb DHFR transcript from the parallel promoter was detected in all six cell lines of DC/SV-hDHFR31(HB) and in the four cell lines of DC/SV-hDHFR31(NB) in lesε abundance. The 0.8 kb DHFR tranεcript from the reverεe promoter, waε detected in DC/AD/R-hDHFR31(NB) , but not in DC/SV/R-hDHFR31(HB) nor in DC/SV/R-hDHFR31(NB) . Instead a truncated meεεage of 0.4 kb waε observed from the reverse promoters, but in less abundance than the truncated message observed with the full length cDNA of hDHFR (Figs. 28A-28B) .

The DNA analyεiε of theεe cell lineε iε εhown in Figureε

SUBSTITUTESHEET IRULE 26) 31A-31B. The genomic DNA was digeεted with Dral which cuts once within the hDHFR cDNA, generating a single band of about 4 kb. No gross recombination was observed in any cell line, indicating that the observed truncation of the message occurred at the RNA level rather than at the DNA level.

Expression of Mutant hDHFR in Murine Bone Marrow Cells in CFU-GM Colonies

The MTX resiεtance level of the bone marrow cells transduced through coculture with DC/SV-hDHFR31(HB) and DC/SV-hDHFR31(NB) constructs was determined by CFU-GM assay and compared with the DC/SV/R-mDHFR. The two hDHFR constructs produced similar level of resiεtance to MTX aε the mDHFR (Table 8). DC/SV-hDHFR31(HB) waε chosen for the in vivo bone marrow transplantation εtudieε.

Table 8. MTX reεistant CFU-GM of bone marrow cells infected with mutant hDHFR in retroviral vectorε.

The CFU-GM asεay of the mouεe bone marrow cellε infected with the mutant hDHFR in retroviral vectors was performed in the absence or the presence of 100 nM MTX. The percentage of resistant colonies was calculated by dividing the number of MTX resiεtant colonieε by the colonieε formed in the absence of MTX.

Table 8

vector NO. Of colonies percentage of -MTX +MTX resiεtant colonieε (100 nM)

control 657 47 7.2

DC/SV/R-mDHFR 691 222 32.1

DC/SV-hDHFR31(HB) 763 234 31.0

DC/SV-hDHFR31(NB) 665 180 27.0

Z2 Resistance Developed in Mice After the

Transplantation of the Bone Marrow cells Transduced bv Mutant hDHFR

Mice transplanted with the marrow tranεduced by DC/SV- hDHFR31(HB) survived the MTX selection of both low-dose (A) or delayed high-dose scheduleε (B) , while control mice transplanted with untransduced marrow died (Figs. 32A-32B) .

The integration of the proviral DNA in MTX reεiεtant CFU- GM colonieε of recipient mice 5 weeks after the BMT and low-dose MTX selection was analysed by PCR blotting. The DNA pooled from 5 to 6 colonies was amplified uεing H250 (annealling to hDHFR) and GT-NC1 (annealling to SV40 promoter,) as primers and then was hybridized to a labelled hDHFR probe (Table 4) . A 300bp fragment waε detected aε expected from the PCR reaction (Figure 33) .

Experimentε are in progreεs to follow the surviving recipientε, to conduct εecondary BMT, and to demonstrate the integration of the mutant hDHFR in the tissues of the recipient mice.

EXPERIMENTAL DISCUSSION

Vector Design and the Expression of the Altered Murine DHFR in Vitro from the DC Vectors

Efficient expression of a variant DHFR conferring MTX resiεtance iε a neceεεary requirement to protect tranεduced cellε from MTX toxicity. In previouε εtudies involving the expression of a mutant mDHFR in various retroviral vectors, the mutant gene was expressed from the viral LTR promoter or from a SV40 promoter situated within the LTR transcriptional unit (Williams et al., 1987, Corey et al., 1990, Kwok et al., 1986, Schuening et al., 1989, Hock and Miller, 1986). The activity of promoters iε often reduced however, when placed downεtream from an active promoter (Emerman and Temin, 1984) . The εelection of a gene driven by one promoter may reεult in the suppression of the expression of another gene driven by the other promoter. Bowtell et al (1988) have also observed that retroviruseε carrying two genes, one transcribed from the LTR and the other from the SV40 promoter are poorly transcribed in vivo, even in the absence of selection and despite the presence of the provirus in the host hematopoietic cellε. It is possible that the suppression and the low expresεion of the εecond gene waε due to the fact that in both cases the gene driven by the internal promoter was εituated within the same transcriptional unit of the LTR promoter, rather than the mere sequential presence of the two transcriptional unitε. To overcome this suppression, we used a double copy vector which has a 5' duplication of the recombinant gene inserted at the 3' LTR (Figs. 6A-6D, 11) . The 5' duplication is outside the LTR transcriptional unit. Thiε vector deεign haε been εhown to express the inserted gene at a 10-20 fold higher level

it →*

# than the vector with the promoter cloned within the LTR transcriptional unit (Hantzopoulos et al., 1989).

To determine the effect of different promoters on the expression of the altered mDHFR gene in different cell types, five different promoters were cloned in the double copy vector and their expresεion waε compared in 3T3 fibroblaεt cells, three human leukemia cell lines and mouεe bone marrow.

In 3T3 cellε, the mutant mDHFR iε tranεcribed from the minigene promoters in DC/SV-mDHFR, DC/SV/R-mDHFR, DC/AD/R-mDHFR, DC/TK/R-mDHFR and DC/AC-mDHFR constructε, with the exception of the CMV promoter (Figs. 12A-12B) . The MTX reεiεtance level aε teεted by the colony formation assay correlates well with the mRNA level of tranεcription from the internal promoter, even though a low level of translation of the mutant enzyme from the polycistronic message cannot be excluded (Kaufman et al., 1987) .

The DC/SV/R-DHFR construct which contains the recombinant transcriptional unit in the opposite orientation to viral tranεcription, haε a εimilar or εlightly higher expreεεion of the mutant DHFR from the internal promoter and a 5-10 fold decreaεe of tranεcription from the LTR promoter, when compared to DC/SV-DHFR which containε the tranεcriptional unit in the εame orientation as virus transcription (Figure 11, Figs. 12A-12B) . The mechanism for the decrease in the viral LTR mesεage level iε not known. In a εtudy with a replication-competent MoMLV vector which has a double copy recombinant mutant DHFR under the control of the SV40 promoter, the transcription unit cloned in the opposite orientation of the LTR gave no virus production (Stuhlmann et al., 1989), indicating _* tranεcription in the opposite orientation of the LTR transcription might have an inhibitory effect on viral transcription, meεεage εtability or tranεlation efficiency, reεulting in no or low viral production. In our εtudy, however, DC/SV-mDHFR and DC/SV/R-mDHFR conεtructε have εimilar viral titerε, and gave essentially the same number of MTX reεiεtant colonies in infected 3T3 cells. Conεtructε εuch aε DC/AD/R-mDHFR and DC/TK/R-mDHFR which contain the recombinant tranεcriptional unit in the opposite orientation of the LTR also do not have reduced viral production. These results indicate the orientation of the transcription of the recombinant gene has very little effect on viral production.

The results from the transduced human leukemia cell lines show that these constructε can give lower but reaεonable levels of MTX resistance in human cell lines of hematopoeitic lineage, but there was no correlation with the results obtained in the 3T3 cell line, as regards effectiveness of the various promoter constructε.

The results from the transduced bone marrow cells further indicate that the mouεe fibroblast cell lines such as NIH 3T3 cannot be used to assess the effectiveness of retroviral mediated gene transfer, and that the activity of retroviral vector encoded promoters vary in an unpredictable manner and is probably modulated by the transduced cell type.

The lack of correlation of the expresεion of the mutant DHFR between different cell types is not a total εurprise. The SV40, TK, metallothionein (MT) , c-fos, CMV and adenoviruε EIA promoterε were reported to lack or show very low level of expresεion of the human ADA cDNA in primary murine hematopoietic cellε deεpite excellent expreεεion in fibroblaεt and hematopoietic cell lineε (Lim et al., 1987; Mclvor et al., 1987; Belmont et al., 1988) . The preεent εtudy iε the firεt direct compariεon of the expreεεion of a particular gene (altered DHFR) driven by different promoterε in the context of the same vector design in four cell lineε and in primary hematopoietic cellε.

The Expression of the MTX Resistance Phenotype Induced by Altered mDHFR in Vivo,

The animal εurvival data after BMT with low-doεe MTX selection clearly demonstrated that the MTX reεiεtant phenotype waε preεent in the animals receiving the bone marrow tranεduced by altered mDHFR. The hematocrit and white blood cell count of theεe animalε returned to normal 4 weekε after BMT and MTX treatment, while all control animals died of anemia, Gl bleeding and marked weight loss within the same time.

It was somewhat surprising that the transduced bone marrow not only protected the animal from the marrow toxicity of MTX but Gl toxicity as well. Similar observations were reported from other research groups (Williams et al., 1987; Corey et al., 1990). The prevention of leukopenia by the transduced gene probably contributes to the protection of the Gl tract.

The continued expresεion of the retroviruε-mediated tranεfer of the altered DHFR and the infection of stem cells with extensive repopulating capability was demonstrated by the serial transplantations of transduced bone marrow cellε. The design of the delayed high-dose MTX treatment schedule takes into consideration the increased MTX tolerance in animals 4 weeks after BMT with a relatively normal cell count compared with the animals juεt after BMT, εo that the long term expression of the MTX resiεtant phenotype can been measured. The design alεo allows testing whether the MTX resiεtance conferred to bone marrow by gene tranεfer can protect the animal from the toxicity of a dose higher than the regular therapeutic dose of MTX. In normal animals (animalε without irradiation or BMT) , both the Gl toxicity and marrow toxicity contribute to the lethal toxicity of MTX (1.1.2) . Our preliminary data showed that the protection was maintained even after challenge with a high dose of MTX.

The integration of the retroviral vector in the hematopoietic cells of the recipient mice iε the necessary requirement for the stable and long-term expression of the mutant mDHFR. PCR analysis showed the presence of the NEO gene carried by the viral vector in both the tissues of the recipient mice and in the MTX resiεtant CFU-GM colonies of the recipient mice marrow. The sequence analysis of the PCR amplification product of the vector DHFR gene product confirmed the existence of the mutant DHFR in the recipient mice.

The negative southern analysis in the primary and secondary BMT recipient is difficult to explain. The obvious conclusion is that less than detectable level of cells (3.3%) in the bone marrow and. spleen of the recipient examined were transduced by the mutant mDHFR but εomehow conferred the reεistance observed in vivo. Possible explanations include the following: the products of the enzyme reaction conducted by the mutant mDHFR in the transduced cells in the presence of MTX might be transported out the transduced cells to rescue other sensitive cells; among the small number of the transduced cells there were progenitor stem cells that could repopulate the sequential transplantation recipients but gave rise to untranεduced progenieε through DNA sequence deletion. The posεibility of a small number of transduced non-multipotential progenitor cells or even mature cells with long life span being carried over in the serial BMT was not excluded. The in vivo experiments and the Southern analysis were repeated several timeε using various controls to exclude the presence of artifacts. The use of a chemotherapeutic agent, 5-fluorouracil (5-FU) to destroy later stage dividing progenitor cells in the donor animal of the sequential BMT might help to exclude the poεsibility that only the later stage progenitor cells were transduced.

5-FU has also been used in several laboratories in primary gene transfer to enrich the number of primitive cells in cycle since the finding that 5-FU can increase the proportion of stem cells by deεtroying more mature dividing cellε (Hodgson and Bradley, 1979) . 5-FU was given to the donor mice εeveral dayε (1-5 days) before harvesting the bone marrow. Because retroviral integration in the target cell genome requires the division of the host cell, the 5-FU treatment, which results in the depletion of the mature cells thuε more primitive cellε were forced into cycle at the time of infection, were believed to increaεe the efficiency of gene transfer into the stem cells (Lerner and Harrison 1990) . But 5-FU pretreatment is not an absolute requirement for gene transfer into stem cellε (Belmont, 1990) . The use of hematopoietic growth factors such as IL-1, IL- 3 and IL-6, to treat the target cells before or in the process of the infection waε reported to increaεe the efficiency of the infection into the stem cells (Lim et al., 1989; Dick et al., 1985). In our experiment, the WEHI conditioned medium, which contains IL-3 and small amountε of other growth factorε, was used in the coculture infection to improve the efficiency of infection.

Vector Construction and the Expression of the Human Altered DHFR in Vitro

The DC vectors carrying the full length hDHFR cDNA did not express the MTX resistance phenotype effectively except for DC/SV-hDHFR31/mr. Northern analysis demonstrated that the hDHFR messages from the internal promoters were truncated with the exception of DC/SV- hDHFR31/mr. The full length hDHFR cDNA in the DC vectors contained a 250 bp 3' untranslated region, in which there were a few potential poly (A) signal sequences. The shorter length of the message might be due to the early termination of the transcription by the poly (A) signal sequence within the cDNA rather through the poly (A) signal sequence provided by the DC vector. The shorter message, however, was not translated into the hDHFR protein in the host cell (3T3) as demonεtrated by immunoprecipitation with a polyclonal antibody against hDHFR.

There waε a modification immediately before the ATG codon in the 5'of the full length hDHFR cDNA. The modification was generated to create a Ncol restriction εite for cloning and did not εhow any adverεe effect on the expreεεion of the hDHFR in a bacterial expreεεion εyεtem (Schweitzer et al., 1989).

To improve the expreεεion of the hDHFR, we constructed the DC vectors carrying the shorter length of the hDHFR cDNA to exclude the potential poly (A) signal sequences. These constructs indeed expressed the mutant hDHFR more effectively with the exception of the DC/SV/R-hDHFR. The mesεage of expected length waε tranεcribed preεumably using the vector poly (A) signal, and the hDHFR protein waε tranεlated. The 5' modification before the ATG starting codon seemed to be a less significant factor than the 3' poly (A) signals in regulation of the expresεion of hDHFR.

The expression of the altered hDHFR (S31) was compared with the altered mDHFR (Arg 22) in the CFU-GM asεay after the coculture infection of the murine bone marrow cells with the amphotropic producer lines. The percentage of the resistant CFU-GM colonies of two hDHFR constructε teεted waε εimilar to the conεtruct DC/SV/R-mDHFR which had been shown to confer MTX resistance in vitro and in vivo.

The Expression of the MTX Resistance Phenotype induced by Altered hDHFR in Vivo

Comparable with the in vivo result of the mDHFR gene transfer, 6/9 mice transplanted with marrow transduced by mutant hDHFR (S31) survived the low-doεe MTX εelection while all mice in the control group died within 3 weekε of the MTX treatment. The surviving . mice were then subjected to high dose MTX treatment (200 mg/kg, twice a week) for 5 additional weekε after BMT with the MTX reεiεtant phenotype persisting during the treatment. Another group of mice treated with delayed high-dose MTX εhowed a εimilar result in that the mice with transduced marrow survived, while the control mice died. Theεe reεults indicate that MTX resiεtance waε conferred to the recipient mice through transplantation of hDHFR transduced marrow.

The preεence of the altered hDHFR in the recipient mice bone marrow was demonstrated by PCR blot analyεiε of the genomic DNA pooled from the MTX resistant CFU-GM colonies. The Southern analyεis of the tissues from the recipient mice (data not shown) did not show the integration of the proviral DNA as in the case with the mutant mDHFR gene transfer. Common factors might be responsible for the negative Southern analysiε in both cases.

Significance of the Study and Future Work Vector construct

One of major efforts in the study was an attempt to optimize the design of the retroviral vectors to improve the expression of the transduced gene. DC vector design used in this study seems to have the advantage of achieving equal in vitro expresεion of the two recombinant geneε in the vector, i.e. the NEO and the DHFR. The in vivo expreεεion of the NEO gene waε not teεted due to the known cytotoxicity of G418. The expreεεion of DHFR driven by five different promoterε was compared in the DC vector construct in different cell lines and murine bone marrow. The reεult of this study was consiεtent with the idea that the expreεεion of a gene under control of a certain promoter cannot be predicted in bone marrow cellε baεed on data from a different cell type or in a different context. In cloning the hDHFR cDNA into the retroviral vector, the 3' untranεlated region resulted in a truncated message that was neither translated into hDHFR protein, nor confered MTX reεiεtance. The impact of thiε finding on optimizing retroviral vector constructs needs to be borne in mind in the future when attempting to improve expresεion.

The infection efficiency of the DC vector conεtructε is between 10-30% measured by the MTX resistant CFU-GM colonies after coculture infection. No significant improvement waε observed when compared to other MoMLV based vectorε carrying mutant DHFR reported before (Hock and Miller, 1986; Kwok et al., 1986; Schuening et al., 1989) . This may partly due to the intrinsic character of the MoMLV vector, rather than particular vector conεtruct deεign. MoMLV induces T-cell lymphomas in new born-NFS mice, while Friend murine leukemia virus (FrLV) induces erythroleukemia. U3 region of the viral LTR were shown to be the primary determinant of the distinct disease specificities of the two virus. A 200 base direct repeat and a short 3' adjacent GC-rich segment within the U3 region encode the enhancer function for both viruε and the exchange of the region resulted in almoεt complete exchange of the disease specificities of the virus (Li et al., 1987; Golemis et al., 1989). A study done by Holland et al. (1987) showed that replacement of the enhancer segment of the MoMLV with the corresponding fragment of the FrLV improved the expression of the NEO gene carried in the vector in hematopoietic progenitor colonies (GM, BFU-E and GEMM) . Theεe results suggest that MoMLV based vectorε may not be the ideal vector of choice for expreεεion in hematopoietic cellε. The study on Moloney sarcoma virus (MoSV) which transforms fibroblastε in vitro and myeloproliferative εarcoma viruε (MPSV) which transform HSC and other hematopoietic progenitors in vivo as well aε fibroblaεtε in vitro • - Ill - showed similar results. The enhancer region within U3 of MPSV confered the hematopoietic tissue εpecificity (Stocking et al., 1985). The in vivo study by Corey et al. (1990) demonstrated that there is a small advantage of using Fr/Moloney hybrid LTR for expression in myeloid hematopoietic cells. The enhancer of the FrLV and MPSV can be further εtudied by deletion and replacement analyεiε to define the εpecific function of different εegmentε of the enhancer. This knowledge can be used to modify the DC vector with the enhancer fragment that is shown to improve the hematopoietic tissue specificity.

MTX as a selecting agent, in vitro and in vivo

MTX resistance as a dominant selectable marker has several advantages over NEO resistance, namely, MTX selection can be conducted in vivo; the in vivo expression of the MTX resistance is very stable (Williams et al., 1987, Corey et al.,1990 and thiε study); and the MTX reεiεtance phenotype iε readily selectable. In spite of these advantages, the in vitro and in vivo MTX selection can still be problematic. Unlike G418 which selectε for a tranεduced bacterial gene without background reεiεtance, the thymidine kinaεe preεent in mammalian cellε can salvage thymidine to syntheεize dTMP de novo and thuε by-pass MTX inhibition, and give rise to background resistance. The use of thymidine phosphorylase to treat the fetal bovine serum in the culture medium in this εtudy successfully reduced the background in the presence of MTX and made the εelection system more sensitive.

In vivo MTX selection, on the other hand, was limited by the change in the MTX εenεitivity of the hematopoietic cells in the recipient mice after BMT. The MTX in vivo selection uεed in previous studies (low-dose εelection) waε only effective before the hematopoietic system recovered from the irradiation and BMT. The delayed high-dose selection developed in this study allowed for the selection of prolonged expresεion of a MTX reεiεtance phenotype, and indicated that early εelection waε not required to εelect for MTX resistance. The hypothesis that MTX resiεtant bone marrow will enable patients to tolerate higher doses of MTX to treat neoplaεms, may be tested with the delayed high-doεe MTX εelection εyεtem.

The study on altered hDHFR (S31) confirmed the report that thiε variant DHFR can be use as a dominant selectable marker (Schweitzer et al., 1989; Banerjee et al., 1994).

This study shows that MTX resistance offers an attractive choice of a selectable marker in gene transfer studies, as well as an opportunity to protect bone marrow from MTX toxicity in patients treated with this drug for neoplasms or immunosuppression purposes.

The complexity of the hematopoietic system

The discrepancy of the in vivo study between the survival data at the phenotipic level and the result of Southern analysiε at the molecular level shows the complexity of the hematopoietic system. A better understanding of the syεtem iε essential in order to achieve successful gene therapy. Recent advances in purification of hematopoietic εtem cellε (HSC) and the interaction of HSC with the microenvironment may lead to improved efficiency of gene tranεfer to HSC.

Besides the known techniqueε to enrich the HSC in cycle (the 5-FU pretreatment of the donor and the uεe of growth factorε) , techniqueε to purify hematopoietic εtem cells using monoclonal antibodies to εtem cell εpecific antigens (CD34+ for human and other primates, Thy-1 and H-2K for mice etc.) have been developed (Spangrude et al., 1988; Szilvassy et al., 1989; Berenεon et al., 1988). Small numbers of the purified cells were reported to be able to reconstitute irradiated animals. Recently, εucceεεful engraftment of bone marrow after infuεion of purified CD34+ marrow cellε into patients has been reported (Berenson et al., 1991). It is alεo poεsible to use long term cultures of the hematopoietic cells to improve the efficiency of retroviral infection (Schuening et al., 1989). In addition, immortalized bone marrow stromal lineε with selectable phenotypes, which are capable of supporting hematopoiesis in long-term culture, have been developed and may facilitate both the retroviral infection and selection of hematopoietic cells in vitro (Williams et al., 1988; Paul, et al., 1991). For example, pre-established irradiated autologous marrow stromal layer was recently shown to enhance cell-free retroviral vector transduction of human bone marrow long- term culture initiating cells (Moore et al., 1992).

Some new growth factorε that may play a role in the regulation of hematopoiesis are also under intensive study. C-kit ligand (mast cell growth factor, MGF) was reported to εynergistically interact with a number of cytokines, such as IL-3, granulocyte-monocyte colony- stimulating factor (GM-CSF) , and directly augment the proliferative capacity of primitive human hematopoietic progenitor cells (Brandt et al., 1992). The leukemia inhibitory factor was shown to be able to improve the survival of hematopoietic stem cells during culture of bone marrow with vector-producing fibroblastε reεulting in a 10 fold increase of infection efficiency, allowing long-term expresεion of the vector encoded gene (Fletcher et al., 1991). The results generated from these studies indicate that combination of various growth factors can improve the efficiency of the retroviral-mediated gene tranεfer.

The optimization of the deεign of retroviral constructs, the perfection of the selection regimen and the improvement in the efficiency of the HSC infection will be the major areas of future efforts to develop MTX resistance in mice by retroviral gene transfer of altered DHFR genes, and form the basis for studies in patients.

CONCLUSIONS

Double copy retroviral vectors were constructed that contained altered cDNA for mDHFR (3T6) and hDHFR (S31) under the control of different promoters.

These constructε were able to expreεε the mutant DHFR and confer different levelε of MTX resistance to the infected cells (3T3 fibroblast cell line; CEM, K562, Raji leukemia cell lineε; and murine bone marrow cellε) in vitro.

Studieε in irradiated mice transplanted with bone marrow cells infected with retroviral constructs indicated that protection was af orded to MTX cytotoxicity for prolonged periods of time.

REFERENCES OF THE PRECEDING SECTIONS

Anderson, K.D., Thompεon, J.A., DiPietro, J.M. , Montgomery, K.T. , Reid, L.M. and Anderεon, W.F. (1989). Gene expreεεion in implanted rat hepatocyteε following retroviral-mediated gene tranεfer. Somatic Cell Mol Genet 15, 215-227

Anderεon, W.F. (1992) . Human Gene Therapy. Science, 256,808-813

Apperley,J.F. , Luskey,B.D. and Williams,D.A. (1991) . Retroviral gene transfer of human adenosine deaminaεe in murine hematopoietic cells: Effect of selectable marker sequence on long-term expression. Blood,78:310-317

Apperley, J.F. and Williams, D.A. (1990) . Annotation- Gene therapy: current status and future directions. British J Haematology 75, 148-155

Armentano,D. , Yu,S.-F., Kantoff,P.W. , Von Ruden,T • •. I Anderson,W.F. and Gilboa,E. (1987) . Effect of internal viral sequences on the utility of retroviral vectors. J.Viorl. 61,1647-1650

Banerjee,D., Schweitzer,B. , Volkenandt,M. , Li,M., Waltham,M., Mineishi,S. Zhao,S.C. and Bertino,J.R. Transfection with a cDNA encoding a Ser31 or Ser34 mutant human dihydrofolate reductase into Chineεe Hamεter Ovary and mouεe marrow progenitor cellε conferε methotrexate reεistance. Gene 139, 1994 (submitted, 1992), 269-274.

Belmont, J.W., Macgregor, G.R. , Wager-Smith, K. ,

Fletcher, F.A. Moore, K.A. , Hawking, D., Villalon, D. , Chang, S.M.W. and Caskey, CT. (1988). Expreεεion of human adenoεine deaminase in murine hematopoietic cells. Mol Cel Biol. 8,5116-5125

Bender, M.A. , Gelinaε, R.E. and Miller, A.D. (1989). A majority of mice εhow long-term expression of a huamn β- globin gene after retrovirus transfer into hematopoietic stem cells. Mol Cell Biol. 9,1426-1434

Berenεon, R.J. , Bensinger, W.I., Hill, R.S., Andrews, R.G., Garcia-Lopez, J., Kalamasz, D.F., Still, B.J., Spitzer, G., Buchner, CD., Bernstein, I.D. and Thomaε, E.D. (1991). Engrafment after infusion of CD34+ marrow cellε in patients with breast cancer or neuroblastoma. Blood, 77,1717-1722

Berenson, R.J., Andrews, R.G., Bensinger, W.I., Kalamasz, D. , Buckner, CC, and Bernstein, I.D. (1988). Antigen CD34+ marrow cells engraft lethally irradiated baboons. J Clin Invest 81, 951-955

Bertino, J.R. , Srimatkandada,S. , Carman, M.D., Mini, E., Jastreboff, M. , Moroson, B.A. and Dube, S.K. (1985). Mechanisms of drug resiεtance in human leukemia, in Modern Trends in Human Leukemia VI. Neth, R. , Gallo, R.C., Greaveε, M.F. Janka, G., edε. Berlin.

Bertino,J.R. (1979) . Toward improved selectivity in cancer chemotherapy. Cancer Res.39,293-304

Birnboim, H.C and Doly, J. (1979). A rapid alkaline extraction procedure for screening recombinant plaεmid DNA. Nucleic Acid Reε. 7, 1513-1523

Blakly,R.L. (1969) . The biochemistry of folic acid and related peteridines (North-Holland, Amsterdam) _*

Bleyer,W.A. (1978) . The clinical pharmacology of methotrexate. Cancer 41,36-51

Bowtell, D.D.L. , Cory, S., Johnson, G.R. and Gonda, T.J. (1988) . Comparison of expreεεion in hematopoietic cells by retroviral vectors carrying two genes. J Viol. 62, 2454-2473

Brandt, J., Briddell, R.A. , Srour, E.F., Leemhuis, T.B. and Hoffman, R. (1992) . Role of c-kit ligand in the expansion of human hematopoietic progenitor cells. Blood, 79,634-641

Carman, M.D., Schornagel, J.H., Rivest, R.S., Srimatkandada , S., Portlock, CS. and Bertino, J.R. (1984) . Clinical reεistance to methotrexate due to gene amplification. J. Clin. Oncol. 2:7-11

Chabner, B.A. , Allegra, C.J., Curt, G.A., Clendeninh, N.J. , Baram,J. , Koizumi, S., Drake, J.C, and Jolivet, J. (1985) . Polyglutamylation of methotrexate. Iε methotrexate a prodrug? J. Clin. Inveεt. 76:907-912

Chirgwin,J.M. , Przybyla,A.E. , MacDonald,R.J. and Rutter,W.J. (1979) . Isolation of biologically active ribonucleic acid from source enriched in ribonuclease. Biochemiεtry 18,5294-5299

Corey, CA. , Deεilva, A., Holland, C. and Williams, D.A. (1990). Serial transplantation of methotrexate-reεiεtant bone marrow: Protection of murine recipients from drug toxicity by progeny of transduced stem cells. Blood 75,337-343

Chou, T-C. and Talalay, P. (1984). Quantitative analysis _* of doεe-effect relationεhipε: The combined effeσtε of multiple drugs or enzyme inhibitors, in Advances in Enzyme Regulation Ed. Weber,G. Pergamon Preεs, vol 22,27- 55

Cullen,B.R., Lomedico,P.T. and Ju,G. (1984). Transcriptional interference in avivan retroviruses— implications for the promoter inεertion model of leukaemogeneεis. Nature 307,241-245

Curiel, D.T., Wagner, E., Cotten, M. , Birnstiel, M.L. , Agarwal, S., Li, C.-M., Loechel, S. and Hu, P-C (1992). High efficiency gene transfer mediated by adenovirus coupled to DNa-polylysine. Human Gene Therapy. 3,147-154

Curt, G.A. , Jolivet, J._» Carney,D.N., Bailey,B.D., Drake, J.C, Clendeninn, N.J. , and Chabner,B.A. (1985). Determinants of the sensitivity of human small-cell lung cancer cell lines to methotrexate. J. Clin. Invest. 76:1323-1329

Davies, J. and Jiminez, A. (1980) . A new selective agent for eukaryotic cloning vectors. Am J Trop Med Hyg 29:suppl 1089-1092

Dicker, A.P., Volkenandt, M. , Adamo, A., Barreda, C. and Bertino, J.R. (1989) . Sequence analysiε of a human gene responsible for drug resiεtance: A rapid method for manual and automated direct sequencing of products generated by the polymerase chain reaction. BioTechniqueε 7, 830-838

Dick, J.E., Magli, M.C, Huεzar, D. , Phillips, R.A. and

Bernstein, A. (1985) . Introduction of a selectable gene in primitive stem cells capable of long-term _* reconstitution of the hematopoietic system of W/W^v mice. Cell, 42, 71-79

Dimaio, D. , Treisman, R. and Maniatis, T. (1982). Bovine papillomaviruε vector that propagateε as a plasmid in both mouse and bacterial cells. Proc Natl Acad Sci USA. 79,4030-4034 ^•

Emerman,M. and Temin,H.M. (1984) . Genes with promoters in retroviruε vectors can be independently suppressed by an epigenetic mechanism. Cell 39,459-467

Feigner, P.L. , Gadek, T.R., Holom, M. , Roman, R. , Chan, H.W. , Wenz, M. , Northrop, J.P., Ringold, G.M. , and Danielsen, M. , (1987). Lipofection: a highly efficient lipid-mediated DNA transfection procedure. Proc Natl Acad Sci USA. 84, 7413-7417

Fletcher, F.A. , Moore, K.A., Ashkenazi, M. Vrieε, P.D., Overbeek, P.A. , Williams, D.E. and Belmont, J.W. (1991). Leukemia inhibitory factor improveε εurvival of retroviral vector-infected hematopoietic εtem cellε in vitro, allowing efficient long-term expression of vector- encoded human adenosine deaminaεe in vivo. J. Exp. Med. 174,837-845

Flowers, M.E.D., Stockεchlaeder, M.A.R. , Schuening, F.G., Niederwieser, D. , Hackman, R. , Miller, A.D., Storb, R. (1990) . Long-term transplantation of canine keratinocytes made resistant to G418 through retrovirus-mediated gene transfer. Proc Natl Acad Sci USA 87,2349-2353

Foley,G.E., Lazarus,H. , Farber,S., Uzman,B.G., Boone,B.A. and McCarthy, R.E. (1965) . Continuous culture of human lymphoblasts from peripheral blood of a child with acute leukemia. Cancer 18,522-529

Gansbacher, B. , Zier,K., Danielε,B., Cronin,K., Bannerji,R. and Gilboa,E. (1990) . Interleukin 2 gene tranεfer into tumor cellε abrogateε tumorigenicity and induceε protective immunity. J.Exp.Med. 172,1217-1224

Gilboa, E. , Eglitiε, M.A. , Kantoff,ube, P.W. and

Anderεon, W.F. (1986). Transfer and expresεion of cloned geneε uεing retroviral vectorε. BioTechniqueε 4:504-512.

Gilboa, E. (1987) . retroviruε vectors and their uses in molecular biology. BioEssays. 5,252-258

Goff,S., Traktman,P. and Baltimore,D. (1981). Isolation and properties of moloney murine leukemia virus mutantε: uεe of a rapid assay for release of virion reverse transcriptase. J.Virol. 38,239-248

Golemis, E. , Li, Y., Fredrickεon, T.N., Hartley, J.W. and Hopkinε, N. (1989) . Diεtinct εegmentε within the enhancer region collaborate to εpecify the type of leukemia induced by nondefective Friend and Moloney viruεeε. J Virol. 63,328-337

Goodman and Gilman (1980). Folic Acid Analogε. In The Pharmacological Basis of Therapeutics . Macmillan

Publiεhing Co., INC. New York, p:1272-1276

Gruber, H.E., Finley, K.D., Herεhberg, R.M. , Katzman, S.S., Laikind, P.K. and Seegmiller, J.E. (1985). Retroviral vector-mediated gene transfer into human hematopoietic progenitor cells. Science, 230,1057-1061

Gunning,P., Leavitt,J., Muscat,G., Ng,S-Y. and Kedes,L. (1987) . A human β-actin expression vector syεtem directs high-level accumulation of antisenεe transcripts. Proc.Natl.Acad.Sci.USA 84,4831-4835

Haber,D.A., Beverly,S.M. , Kiely,M.L. and Schimke,R.T. (1981) . Properties of an altered dihydrofolate reductase encoded by amplified genes in cultured mouse fibroblasts. J. Biol. Chem. 256,9501-9510

Hantzopoulos,P.A. , Sullenger,B.A. , Ungers,G. and Gilboa,E. (1989) . Improved gene expreεεion upon tranεfer of the adenoεine deaminaεe minigene outεide the tranεcriptional unit of a retroviral vector. Proc.Natl.Acad.Sci.U.S.A. 86,3519-3523

Harrison, D.E., Astle, CM. and Lerner, C (1988). Number and continuous proliferative patterns of transplanted primitive immunohematopoietic stem cells. Proc Natl Acad Sci USA. 85,822-826

Hauser, J., Levin, A.S. and Dickson, K. (1987) . Unique pattern of point mutations arising after gene transfer into mammalian cells. EMBO J. 6, 63-67

Hermonat, P.L. and Muzyczka, N. (1984). Use of adeno- associated virus as a mammalian DNA cloning vector: Transduction of neomycin reεiεtance into mammalian tiεsue culture cells. Proc Natl Acad Sci USA. 81; 6466-6470

Hock, R.A. , Miller, A.D. and Osborne,W.R.A. (1989) . Expreεεion of human adenoεine deaminaεe from various strong promoters after gene transfer into human hematopoietic cell lines. Blood, 74,876-881

Hock, R.A. and Miller,A.D. (1986) . Retrovirus-mediated _* tranεfer and expreεsion of drug resiεtance geneε in human hematopoeitic progenitor cellε. Nature 320,275-277

Holland, CA. , Ankleεaria, P., Sakakeeny, M.A. and Greenberger, J.S. (1987). Proc Natl Acad Sci USA. 84, 8662-8666

Holmeε, D.S. and Quigley, M. (1981). A rapid boiling method for the preparation of bacterial plasmids. Anal. Biochem. 114, 193-197

Horns, R.C, Dower, W.J., and Schimke, R.T. (1984). Gene amplification in a leukemia patient treated with methotrexate. J. Clin. Oncol. 2:2-7

Isola, L.M. and Gordon, J.W. (1986) . Syεtemic reεistance to methotrexate in transgenic mice carrying a mutant dihydrofolate reductase gene. Proc.Natl.Acad.Sci. U.S.A. 83,9621-9625

Joyner, A., Keller, G. , Phillips, R.A. and Bernstein, A. (1983) . Retroviruε tranεfer of a bacterial gene into mouεe hematopoietic progenitor cellε. Nature, 305,556-558

Kantoff, P.W., Gillio, A.P., McLachlin, J.R., Bordigonon, C, Eglitiε, M.a., Kernan, N.A., Moen, R.C, Kohn, D.B., Yu, S.F., Karson, E. , Karlsson, S., Zwiebel, J.A. , Gilboa, E., Blaese, R.M. , Neinhuis, A., O'Reilly, R.J. and Anderεon, W.F. (1987). Expreεεion of human adenoεine deaminaεe in non-human primateε after retroviral mediated gene tranεfer. J Exp Med. 166, 219-234

Karlsson,S. , Papyannopoulou,T. , Schweigerr,S.G. ,

Stamatoyannopoulos, S. and Nienhuiε,A.W. (1987). Retroviral-mediated tranεfer of genomic globin genes leads to regulated production of RNA and protein. Proc.Natl.Acad.Sci.USA 84,2411-2415

Karlsson, S., Bodine, D.M. , Perry, L. , Papayannopoulou, T. and Nienhuis, A.W. (1988) . Expression of the human beta globin gene following retroviral mediated transfer into multipotential hematopoietic progenitors of mice. Proc Natl Acad Sci USA. 85, 6062-6066

Kaufman,R.J. , Murtha,P. and Davieε,M.V. (1987). Tranεlational efficiency of polyciεtronic mRNAε and their utilization to express heterologous geneε in mammalian cellε. EMBO J.6:187-193

Keller, G. , Paige, C, Gilboa, E. , and Wagner, E.F. (1985) . Expression of a foreign gene in myeloid and lymphoid cellε derived from multipotent haematopoietic procursors. Nature, 318,149-154

Kwok,W.W. , Schuening,F. , Stead,R.B. and Miller,A.D. (1986). Retroviral transfer of genes into canine hemopoietic progenitor cells in culture: A model for human gene therapy. Proc.Natl.Acad.Sci.USA 83,4552-4555

Lemiεchka, I.R., Raulet, D.H. and Mulligan, R.C (1986). Developmental potential and dynamic behaviour of hematopoietic stem cells. Cell, 45, 917-927

Lerner, C. and Harrison, D.E. (1990). 5-fluorouracil spareε hemopoietic εtem cellε responsible for long-term repopulation. Exp Hematol 18,114-118

Li,M. , Zhao,S. , Banerjee,D. , Hantzopoulos,P. , Gilboa,E. and Bertino,J.R. (1990) Enhanced selectivity of methotrexate resistant CFU-GM colonies from transfected murine bone marrow in thymidine phoεphorylase treated serum. Blood 76 (10), εuppl 1,938

Li, M. , Hantzopouloε, P.A. , Banerjee, D. , Zhao, S.C, Schweitzer, B.I., Gilboa, E. and Bertino, J.R. (1992). Compariεon of the expression of a mutant dihydrofolate reductase under control of different internal promoters in retroviral vectors. Human Gene Therapy (in preεε) .

Li, W.W. , Lin, J.T., and Bertino, J.R. (1989). Antifolate reεistance: in vitro vs. in vivo. Tapiero, H. , Rovert, J., and Lampidis, T.J., eds. Colloque INSERM, vol.191, Anticancer Drugs. p329-339, John Libbey Eurotext Lid. , Paris.

Li, Y., Golemis, E. , Hartley, J.W. and Hopkins, N. (1987) . Disease specificity of nondefeective Friend and Moloney murine leukemia viruses is controlled by a smal number of nucleotides. J. Virol. 61, 693-700

Lim, B. , Apperley, J.F., Orkin, S.H. and Williams, D.A. (1989) . Long-term expreεεion of human adenoεine deaminaεe in mice transplanted with retrovirus-infected hematopoietic εtem cells. Proc Natl Acad Sci USA. 86, 8892-8896

Lozzio,C.B.and Lozzio,B.B. (1975). Human chronic myelogenous leukemia cell line with poεitive Philadelphia chromosome. Blood 45,321-334

Magli, M.C., Dick, J.E., Huszar, D. , Bernstein, A. and Phillipε, R.A. (1987) . Modulation of gene expreεεion in multiple hematopoietic cell lineageε following retroviral vector gene tranεfer. Proc Natl Acad Sci USA. 84, 789-793 _*

Maniatiε, T. , Fritsch, E.F. and Sambrook, J. (1982) . Molecular Cloning, a laboratory manual. Cold Spring Harbor Laboratory

Markowitz,D. , Goff,S. and Bank,A. (1988). A safe packaging line for gene transfer: separating viral genes on two different plasmids. J.Virol. 62,1120-1124

Markowitz,D. , Goff,S. and Bank,A. (1988) . Construction and use of a safe and efficient amphotropic packaging cell line. Virology 167,400-406

McGuire, J.J., Mini, E. , Hsieh,P., and Bertino, J.R, (1985) . The role of methotrexate polyglutamates in methotrexate and sequential methotrexate-5-fluorouracil mediated cell kill. Cancer Res. 45:6395-6400

Mclvor, R.S., Johnson, M.J., Miller, A.D., Pitts, S., Williams, S.R., Valerio, D. , Martin, D.W. and Verma, I.M. (1987) . Human purine nucleoside phosphorylase and adenosine deaminaεe gene tranεfer in cultured cellε and murine hematopoietic εtem cellε by uεing recombinant amphotropic retroviruses. Mol Cel Biol. 7, 833-846

Mcknight,S.L. and Gavis,E.R. (1980) . Expresεion of the herpes thymidine kinase gene in Xenopus laevis oocytes: an assay for the study of deletion mutants constructed in vitro. Nucleic Acid Res. 8,5931-5948

Miller, A.D., Ong, E.S., Rsosenfeld, M.G.; Verma, I.M. and Evans, R.M. (1984). Infectious and selectable retrovirus containing an inducible rat growth hormone minigene. Science 225:993-997

Miller, A.D. (1990). Progresε toward human gene therapy. Blood, 76, 271-278

Moore, K.A. , Deisεenroth, A.B., Reading, CL. , Williamε, D.E. and Belmont, J.W. (1992). Stromal εupport enhanceε cell-free retroviral vector tranεduction of human bone marrow long-term culture-initiating cellε. Blood, 79,1993-1399

Morgan, J.R.,, Barrandon, Y., Green, H. , Mulligan, R.D.(1987). Expression of an exogenous growth hormone gene by transplantable human epidermal cellε. Science, 237, 1476-1479

Mullis, K.B. and Faloona F.A. (1987). Specific synthesis of DNA in vitro via a polymerase catalyzed chain reaction. Meth. Enzym. 155, 335-350

Narayanan, R. , Jastreboff, M.M. and Bertino, J.R. (1986). Development of an amphotropic, high-titer retrovirus vector expressing the dihydrofolate reductase gene and conferring methotrexate resistance. Gene, 48,71-80

Pizzorno, G. , Mini, E., Coronnello, M. , McGuire, J.J., Moroson, B.A. , Cashmore, A.R. , Dreyer, R.N. , Lin, J.T. , Mazzei, T., Periti, P., and Bertino, J.R. (1988). Impaired polyglutamylation of methotrexate aε a cause of reεistance in CCRF-CEM cells after short-term, high-dose treatment with this drug. Cancer Res. 48:2149-2155

Oefner, C, D'Arcy, A., and Winkler, F.K. (1988). Crystal structure of h-DHFR complexed with folate. Eur. J. Biochem. 174, 377-385

Palmer, T.D., Hock, R.A. , Osborne, W.R. , Miller, A.D. (1987) . Efficient retroviruε-mediated tranεfer and expreεεeion of a human adenoεine deaminaεe gene in diploid skin fibroblasts from an adenosine deaminase- deficient human. Proc Natl Acad Sci USA 84,1055-1059

Paul, S.R., Yang, Y-C , Donahue, R.E., Goldring, S. and Williams, D.A. (1991) . Stromal cell-associated hematopoiesiε: immortalization and characterization of a primate bone marrow-derived εtromal cell line. Blood, 77,1723-1733

Pizzorno, G. , Chang, Y-M. , McGuire, J.J., and Bertino, J.R. (1989). Inherent reεistance of human squamous carcinoma cell lines to methotrexate as a result of decreased polyglutamylation of this drug. Cancer Res. 49:5275-5280

Polisky, B. , Greene, P., Garfin, D.E., McCarthy, B.J.,

Goodman, H.M. and Boyer, H.W. (1975) . Specificity of substrate recognition by the EcoRI restriction endonuclease. Proc. Natl. Acad. Sci. U.S.A. 72, 3310-3314

Prendergast, N.J., Appleman, J.R., Delcamp, T.J., Blakley, R.L. and Freiεheim, J.H. Effects of conversion of Phenylalanine -31 to Leucine on the function of human dihydrofolate reductase. Biochemistry. 28:4645-4650 (1989) .

Proudfoot,N.J. (1986) . Transcriptional interference and termination between duplicated α-globin gene constructs suggests a novel mechanism for gene regulation. Nature 322,562-565

Puck, T.T. , Cieciura, S.J. and Robinson, A. (1958).

Genetics of somatic mammalian cells III: Long-term cultivation of euploid cells from human and animal subjectε. J. Exp. Med. 108,945-955

Pulvertaft,R.J.V. (1964) . Cytology of Burkitt'ε tumour (African Lymphoma) . Lancet 1,238-240

Ralph, P. and Nakoinz (1977) . Direct toxic effects of immunopotentiators on monocytic, myelomonocytic, and histiocytic or macrophage tumor cells in culture. Cancer Res. 37, 546-550

Rosenbergerg, S.A., Aebersold, P., Cornetta, K. , Kaεid, A., Morgan, R.A. , Moen, R. , Karson, E.M. , Lotze, M.T., Yang, J.C, Topalian, S.L. , Merino, M.J., Culver, K. , Miller, A.D., Blaese, R.M. and Anderson, W.F. (1990). Gene transfer into humans—Immunotherapy of patients with advanced melanoma, using tumor-infiltrating lymphocytes modified by retroviral gene transduction. N Engl J Med. 323,570-578

Sanger, F.W. , Nicklen,S. and Couleon, A.R. (1977). DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. U.S.A. 74, 5463-5467

Sarver, N. , Cantin, E.M. , Chang, P.S., Zaia, J.A., Ladne, P.A. , Stephens, D.A. and Rossi, J.J. (1990). Ribozymes as potential anti-HIV-1 therapeutic agents. Science 247:1222-1225

Sarver, N., Byrne, J.C, and Howley, P.M. (1982) . Transformation and replication in mouse cells of a bovine papillomavirus-pML2 plaεmid vector that can be reεcued in bacteria. Proc Natl Acad Sci USA. 79,7147-7151

Schaffner, W. (1980) . Direct tranεfer of cloned genes from bacteria to mammalian cells. Proc. Natl., Acad. Sci . USA. 77 , 2163-2167

Schuening,F.G. , Storb,R., Stead,R.B., Goehle,S., Nash,R. and Miller,A.D. (1989) . Improved retroviral transfer of genes into canine hemotopoietic progenitor cells kept in long-term marrow culture. Blood 74,152-155

Schweitzer,B.I. , Dicker,A.P. and Bertino, J.R. (1990). Dihydrofolate reductase as a therapeutic target. FASEB 4:2441-2452

Schweitzer,B.I. , Srimatkandada,S. , Gritsman,H., Sheridan,R. , Vendataraghavan,R. and Bertino, J.R. (1989) . Probing the role of two hydrophobic active εite residues in the human dihydrofolate reductase by εite-directed mutagenesis. J Bio.Chem. 264:20786-20795

Sharp, P. A., Sugden, B. and Sambrook, J. (1973). Detection of two reεtriction endonuclease activities in Haemophilus parainfluenzae using analytical agarose. Biochemistry 12,3055-3063

Simonsen, CC (1986). Drug-resiεtant dihydrofolate reductases. in The Molecular Genetics of Mammalian Cellε . Malacinski, G. , Simonsen, CC and Shepard, M. edε. Machillan Publishing Co., N.Y. p:99-128

Simonsen,CC. and Levinson,A.D. (1983) . Isolation and expresεion of an altered mouεe dihydrofolate reductaεe CDNA. Proc.Natl.Acad.Sci. U.S.A.80,2495-2499

Sirotnak, F.M. , Moccio, D.M. , Kelleher, L.E. , and Goutas,

L.J. (1981) . Relative frequency and kinetic properties of transport defective phenotypes among methotrexate- resiεtant L1210 clonal cellε derived in vivo. Cancer Res. 41,4447-4452

Smithies, O. , Gregg, R.G., Boggs, S.S., Koralewεki, M.A. and Kucherlapati, R.S. (1985). Inεertion of DNA εequenceε into the human chromoεomal β-globin locuε by homologous recombination. Nature 317:230-234

Snodgrass, R. and Keller, G. (1987) . Clonal fluctuation within the hematopoietic syεtem of mice reconstituted with retrovirus-infected cellε. EMBO J. 6,3955-3960

Spangrude, G.J., Heimfeld, S. and Weissman, I.L. (1988). Purification and characterization of mouse hematopoietic stem cells. Science, 241, 58-62

Southern,P.J. and Berg,P. (1982) . Transformation of mammalian cells to antibiotic resiεtance with a bacterial gene under control of the SV40 early promoter. J.Mol.Appl.Genet. 1:327-341

Srimatkandada, S., Schweitzer, B.I., Moroson, B.A. , Dube, S., and Bertino, J.R. (1989). Amplification of a polymorphic dihydrofolate reductaεe gene expressing an enzyme with decreased binding to methotrexate in a human colon carcinoma cell line, HCT 8R4, resiεtant to thiε drug. J. Biol. Chem. 264, 3524-3528

St. Louis, D. and Verma, I.M. (1988) . An alternative approach to somatic cell gene therapy. Proc Natl Acad Sci USA. 85, 3150-3154

Stocking, C , Kollek, R. , Bergholz, U. and Ostertag, W.

(1985) . Long terminal repeat εequenceε impart hematopoietic transformation properties to the myeloproliferative sarcoma virus. Proc Natl Acad Sci USA. 82, 5746-5750

Stuhlmann,H. , Jaaenisch,R. and Mulligan,R.C. (1989). Conεtruction and properties of replication-competent murine retroviral vectors encoding methotrexate resistance. Mol.Cel.Bio. 9:100-108

Sullenger,B.A. , Lee,T.C, Smith,CA. , Ungers,G.E.and Gilboa,E. (1990) . Expression of chimeric tRNA-antisenεe tranεcriptε renders NIH3T3 cells highly resistant to Moloney murine leukemia virus replication. Mol.Cell.Biol. 10,6512-6523

Sullenger, B.A. , Gallardo, H.F. Ungers, G.E. and Gilboa, E. (1990) . Overexpreεεion of TAR εequenceε renderε cellε resistant to human immunodeficiency virus replication. Cell, 63,601-608

Temin, H. M. (1986) . Retrovirus vectors for gene transfer; efficient integration into and expression of exognous DNA in vertebrate cell genomeε. In Gene Trims fer (ed R. Kucherlapati) . Plenum Preεε, New York.

Thillet, J., Absil, J., Stone, S.R. and Pictet, R. (1988). Site directed mutagenesis of mouse dihydrofolate reductase. J. Biol. Chem. 263, 12500-12508

Thomas, K.R. and Capecchi, M.R. (1987) . Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell, 51, 503-512

Tratschin, J-D. , Miller I.L., Smith, M.G., Carter, B.J. (1985) . Adeno-aεεociated virus vector for high-frequency integration, expresεion, and rescue of geneε in mammalian cells. Mol Cell Biol 5, 3251-3260 Urlaub, G. and Chaεin, L.A. (1980) . Isolation of Chinese hamεter ovary cell mutantε deficient in dihydrofolate reductaεe activity. Proc. Natl., Acad. Sci. USA. 77, 4216-4220

Valerio, D. , Duybeεteyn, M.G.C and van der Eb, A.J. (1985) . Introduction of εequenceε encoding functional human deaminaεe into mouεe cellε uεing a retroviral εhuttle εyεtem. Gene, 34, 163-168

Varmus,H. and Swanstrom,L. (1982) . In RNA Tumor Virus , eds. Weiεs,R., Teich,N., Varmus,H. and Coffin, . (Cold Spring Harbor Lab., Cold Spring Harbor, N.Y.), pp 369-512

Wartell, R.M. and Reznikoff, W.S. (1980). Cloning DNA restriction endonuclease fragments with protruding, single-stranded ends. Gene 9, 307-319

Weiss, B. , Jacquemin-Sablon, A., Live, T.R., Fareed, G.C and Richardson, CC (1968). Enzymatic breakage and joining of deoxyribonucleic acid. VI. Further purification and properties of polynucleotide ligase from E.coli infected with bacteriophage T4. J. Biol. Chem. 243,4543-4555

Wiginton,D.A. , Kaplan,D.J., States,J.C, Akeson,A.L. , Perme,CM., Bilyk,I.J., Vaughn,A.J., Lattier,D.L. and Hutton,J.J. (1986) . Complete sequence and structure of the gene for human adenosine deaminase. Biochemiεtry 25,8234-8244

Wigler, M. , Silverεtein, S., Lee, L. , Pellicer, A., Chen, V. and Axel, R. (1977). Transfer of purified Herpes virus thymidine kinase gene to cultured mouse cell. Cell, 11, 223-232 Williams,D.A. (1990) . Expresεion of introduced genetic εequences in hematopoietic cells following retroviral- mediated gene transfer. Hum.Gene Ther.1,229-239

Williams,D.A. , Hεieh,K., Deεilva,A. and Mulligan,R.C. (1987) . Protection of bone marrow tranεplant recipients from lethal doses of methotrexate by the generation of methotrexate-resiεtant bone marrow. J.Exp.Med. 166,210- 218

Williamε, D.A. , Orkin, S.H. and Mulligan, R.C. (1986). Retrovirus-mediated transfer of human adenoεine deaminaεe gene εequences into cellε in culture and into murine hematopoietic cellε in vivo. Proc Natl Acad Sci USA. 83, 2566-2570

Williams, D.A. , Lemischka, I.R. , Nathan, D.G. and Mulligan, R.C. (1984). Introduction of new genetic material into pluripotent stem cells of the mouse. Nature, 310, 476-480

Williamε, D.A. , Rosenblatt, M.F., Beier, D.R. and Cone, R.D. (1988). Generation of murine stromal cell lineε supporting hematopoietic stem celll proliferation by uεe of recombinant retroviruε vectors encoding simian virus 40 large T antigen. Mol Cel Biol. 8, 3864-3871

Wilson, J.M., Jefferson, D.M. , Chowdhury, J.R., Novikoff, P.M., Johnston, D.E. and Mulligan, R.C. (1988) Retroviruε-mediated tranεduction of adult hepatocyteε. Proc Natl Acad Sci USA, 85,3014-3018

Wilεon, J.M. , Birinyi, L.K., Salomon, R.N. , Libby, P.,

Callow, A.D. and Mulligan, R.C. (1989). Implantation of vascular grafts lined with genetically modified endothelial cells. Science 244, 1344-1346

Wong, P.M.C, Chung, S.-W. and Nienhuis, A.W. (1987). Retroviral transfer and expression of the interleukin-3 gene in hematopoietic cells. Gene and Development, 1, 358-365

Yates, J. , Warren, N. , Reiεman, D. and Sugden, B. (1984). A ciε-acting element from the Epεtein-Barr viral genome that permits stable replication of recombinant plasmidε in latently infected cellε. Proc Natl Acad Sci USA. 81, 3806-3810

Zwiebel, J.A., Freeman, S.M., Kantoff, P.W. , Cornetta, K. , Ryan, U.S. and Anderson, W.F. (1989). High-level recombinant gene expresεion in rabbit endothelial cellε transduced by retroviral vectorε. Science, 243, 220-222

second Series of Experiments

Transfection of a novel 8er31 and Ser3 Mutant Human Dihydrofolate Reductase cDNA

Two novel mutant human dihydrofolate reductase (DHFR) cDNAε encoding either a Ser31 or a Ser34 mutation in a mammalian expreεsion vector were transfected into Chinese hamεter ovary (CHO) cellε. Transfection of the Ser31 and Ser34 mutant dhfrs into DHFR^" cells converted them into the DHFR* phenotype. Furthermore, transfection of theεe mutantε into wild type CHO cellε made them reεistant to high levels of methotrexate (MTX) thus indicating that these variants can act aε dominant selectable markers. Southern blot analysis and polymerase chain reaction amplifications confirmed that the transfected plasmids were integrated into the CHO DNA. Gene copy number analysiε revealed that both the Ser31 and the Ser34 mutantε are amplifiable in increasing concentrations of MTX. Theεe mutated human dhfr cDNAε may be worthy of further study as dominant selectable genes for gene therapy.

Dihydrofolate reductase (DHFR E.C.I.5.1.3) catalyzes the reduction of folate and dihydrofolate to tetrahydrofolate, an esεential cofactor in the synthesis of purines and thymidylate (Blakley, 1984) . Methotrexate

(MTX) is a powerful inhibitor of DHFR and is used as an antineoplastic agent in the clinic, although the use of MTX is limited due to emergence of drug reεiεtant tumor cells (Sobrero and Bertino, 1986) . Simonsen and Levinson

(1983) demonstrated the usefulness of an altered mouse dhfr (subεtitution of Arg for Leu at residue 22) from a

MTX reεiεtant 3T6 cell line employing a plaεmid expreεεion vector (pFR400) aε a dominant selectable _* marker in gene tranεfer εtudies. Thiε altered enzyme has been used in both in vitro and in vivo gene transfer studieε (Cline et al., 1980; Williamε et al., 1987; Carr et al., 1983; Iεola and Gordon, 1986; Corey et al., 1990) . While the altered 3T6 DHFR iε markedly impaired in binding to MTX, it has limitations as a selectable marker in that it has poor catalytic activity (Haber et al., 1981; Thillet et al., 1988).

An earlier report (Srimatkandada et al., 1989) from this laboratory characterized a variant human DHFR from a MTX resiεtant human colon carcinoma line, HCT-8R. The alteration waε shown to be a single amino acid change at position 31 where the wild type Phe was changed to a Ser residue. Molecular modeling studies have shown that Phe at position 31 makes van der Waals interactions with both the pteridine and the p-aminobenzoyl moieties of MTX (Oefner et al., 1987). Thus, substituting a large hydrophobic group for a small hydrophilic group has profound effectε on MTX binding. Another Phe occurε at position 34 and is also an active site residue. In a previous report from thiε laboratory, εite directed mutagenesis was used to generate both of these mutants with either the Phe at 31 or 34 changed to Ser (Schweitzer et al., 1989a). Cloning and expression in bacteria of these altered Ser31 and Ser34 cDNAs revealed that theεe DHFRs exhibited decreased MTX binding properties when compared to the wild type human DHFR. However, unlike the 3T6 altered DHFR, both these enzymes had good catalytic activity (Schweitzer et al., 1989b). In the preεent communication we report the uεe of theεe two altered human dhfr cDNAs, the Ser31 and the Ser34 mutants, as dominant selectable markerε in CHO cellε, and compare the results obtained with the Arg22 3T6 mutant CDNA. EXPERIMENTAL RESULTS AND DISCUSSION Vectorε

The mammalian expreεεion vectors were constructed as deεcribed in the legendε of Figure 34. Direct εequence analyεiε of the PCR amplified SV40 promoter and dhfr cDNA verified the εequence of the conεtruct. Sequence analysis revealed that all the cDNAs cloned in the expression vectors contained the desired sequence and did not contain any other mutations (data not shown) .

Transfection and Selection

Two types of CHO cells were used in the gene tranεfer experiments. CHO DG44 cells lacking endogenous DHFR activity cannot grow in F12 without HGT, and colony formation in this medium by cells transfected with the wild type or mutant h-dhfr cDNA indicates successful gene transfer. In this DHFR^* cell line the wild type h-dhfr (HDR) was able to induce the highest number of colonies as compared to the Ser31, the Ser34, and the Arg22 mutantε (Table 9) .

Table 9

COLONY FORMATION IN MTX BY TRANSFECTED CHO DG44 CELLS^* nM MTX HDR Ser31 Ser34 Arg22 Mock

1072 ± 955 ± 902 ± 820 ±

223 147 98.1 93.8

100 0 280 ± 153 ± 122 ±

35.5* 14.4 15.9

500 0 115 ± 70 ± 52 ±

28.8$ 9.5 11.1

1000 0 23 ± 32 ± 43 +

4.7 8.1 2.5

^*Colony formation in F12 without HGT waε measured after 14 days in the presence and absence of MTX. An average of four independent determinations. Expresεed aε colonieε per μg DNA per 10⁶ cellε.

*p < 0.05 as compared to Arg22.

Chinese hamster ovary (CHO) cells lacking DHFR (CHO DG44, obtained from Dr. L. Chasin) were cultivated in complete F12 medium supplemented with 10% fetal bovine serum (FBS) . The MTX sensitive parental CHO cells were grown in RPMI-1640 medium supplemented 10% FBS. All MTX selections were carried out in F12 medium without hypoxanthine, glycine, and thymidine (F12 without HGT) . All cell culture media were supplemented with L-glutamine and penicillin/streptomycin. For MTX selection, dialyzed FBS was uεed in place of FBS. Succesεful gene tranεfer waε alεo demonεtrated by the emergence of G418 resistant colonies when the neo^r gene was cotranεfected. For MTX resistance, 5 x 10⁴ cells from each group were plated in 100 mm petri plates and exposed to various concentrations of MTX. Colonies were scored after 14 days. It was observed that the Ser31 mutant gave rise to the highest number of colonies at lower MTX concentrationε. The Ser34 and the Arg22 tranεfections gave higher number of colonies at MTX concentrations above 500 nM.

To study the mutant h-dhfr cDNAs aε dominant εelectable markerε the level of MTX reεiεtance in wild type CHO cellε (CHO) was determined by selection of transfected cells in various concentrationε of MTX. Colonieε were scored after 14 days (Table 10) .

Table 10

COLONY FORMATION IN MTX BY TRANSFECTED CHO CELLS^*

nM MTX HDR Ser31 Ser34 Arg22

100 0 216 ± 30.5 144 ± 12.8 183 ± 15.2

500 0 66 ± 5.3 73 ± 5.5 120 + 17.3

1000 0 46 ± 4.7 57 ± 11.1 92 + 8.1

^*Colony formation was measured after 14 days of exposure to MTX as described in the experimental εection. An average of four independent determinations. Expressed as colonies per μg DNA per 10⁶ cellε. The Ser31 and the Ser34 mutants were capable of acting as dominant selectable markers, as was the murine dhfr Arg22, in CHO cells. The Ser31 mutant dhfr gave rise to the higheεt number of colonies at 100 nM MTX indicating that at lower MTX doseε it waε capable of acting aε a better εelectable mutant. In order to eliminate the poεεibility that reεiεtance to MTX in theεe tranεfected cellε waε due to other resistance mechanismε, e.g., poor uptake (Sirotnak et al., 1989) and/or defective polyglutamylation of MTX (Pizzorno et al., 1989) , another antifolate, trimetrexate (TMTX) which uses an alternate transport mechanism and cannot be polyglutamylated (Kamen et al., 1984), was used to determine resistance. It was observed that cells originally resistant to 1000 nM MTX (cloned from a single colony growing in 1000 nM MTX) were also resiεtant to TMTX, thuε strengthening the conclusion that resistance to MTX was probably due to presence of an altered DHFR in these transfected cells and not due to poor uptake or defective polyglutamylation of MTX.

Theεe results provide evidence that the Ser31 and Ser34 h-dhfr mutants can act as dominant selectable markers in CHO cells in vitro. Both these mutants convert the DHFR^' to the DHFR^* phenotype when transfected into the DHFR^" DG44 cells. Southern blot analysiε, PCR amplification, and direct sequence analysis of the amplified product provide evidence for chromosomal integration of the tranεfected plasmid expresεion vectorε. Cross resistance to TMTX further strengthens the argument that drug resiεtance in these cells iε due to the expreεεion of the tranεfected altered DHFRs. The Ser31 and Ser34 mutants are comparable to the Arg22 mutant aε dominant εelectable markerε in CHO cells. #

Amplification of Transfected Plasmids

Genomic DNA isolated from cellε cloned from a εingle colony growing at 100 nM MTX (for the Ser31 and Ser34 cDNA tranεfected cellε) waε digeεted with Ncol and Hindlll enzymeε and analyzed by Southern blotting. The Southern blot demonεtrated the presence of unique fragments of approximate size 600 bp generated by restriction with Hindlll and Ncol (Figs. 35A-35B) . When cells resiεtant to 100 nM MTX were εelected in 500 nM MTX and then in 1000 nM MTX, amplification of the integrated plasmid occurred. The Ser31 and the Ser34 containing plaεmidε were found to amplify in increaεing concentrationε of MTX. Gene copy number analysis revealed approximately 5, 35, and 40 copies of the plasmid DNA for the Ser31 transfectantε grown in 100 nM, 500 nM, and 1000 nM MTX, respectively, and 2 , 5 , and 8 copies for the Ser34 tranεfectantε grown in εimilar MTX concentrationε. The Arg22 tranεfectantε alεo showed amplification of the transfected plasmid in increasing MTX concentrations (15, 20, and 25 copieε for 100 nM, 500 nM, and 1000 nM MTX, reεpectively) .

Expreεεion of the Tranεfected cDNAε

Northern analyεiε of total cellular RNA iεolated from CHO, CHO Ser31, CHO Ser34, and CHO Arg22 cellε (the Ser31, Ser34, and the Arg22 tranεfected cellε derived from εingle colonies growing in 100 nM MTX) showed that the transfected cells contained shorter approximately 600 bp and 700 bp DHFR messages which are not normally present in CHO cells (which have 2.35, 1.8, and 1.3 kb sized messages) . The sizes probably result from usage of different poly A+ signals present within the vector plasmid. The two major bandε preεent in all laneε indicated by arrowε repreεent the 28S and the 18S ribosomal RNA species which crosε reactε with the probe. Of all the dhfr mutations reported εo far, only three have been uεed in gene tranεfer studies as candidate dominant selectable markers. Mclvor and Simonsen (1990) reported that the altered murine DHFR characterized from the mouse L5178 Y cell line (subεtitution of Trp for Phe at reεidue 31) may be of uεe in gene transfer studieε. Huεεein et al. (1992) have εhown that the Leu to Phe change at codon 22 of the mouεe dhfr cDNA reεults in an enzyme also capable of acting in a dominant selectable fashion and may also be of use in gene transfer studieε. However, thiε altered enzyme impartε a low level of MTX resistance (100 nM) . The Arg22 and the Trp31 mutants, although they impart a higher level of drug resistance, are catalytically poor enzymes. According to Thillet et al. (1988), mouse DHFR with the Leu to Arg mutation at residue 22 has a 700-fold reduction in catalytic efficiency although it haε a 7.5 x 10⁵-fold increaεe in MTX K, aε compared to the wild type mouse DHFR. In contrast the Ser31 mutant human DHFR has only a 2-fold decrease in catalytic efficiency (Schweitzer et al., 1989 a and b) and a 100-fold increaεe in MTX K, , while the Ser34 mutant haε a 70-fold reduction in catalytic efficiency and a 2 x 10* increase in MTX K, (Schweitzer et al., 1989b). Thuε, despite having higher affinities for MTX than the Arg22 enzyme, the Ser31 and the Ser34 DHFRε may function more efficiently in the transfected cells.

The report that the murine Ser31 mutant dhfr waε not able to behave aε a dominant εelectable marker in CHO cellε (Thillet and Pictet, 1990) is difficult to explain. However, in the same report, the authors were unable to observe MTX resiεtance in Arg22 dhfr tranεfected CHO cellε. Thiε suggests that the experimental conditions were not optimal for the selection of MTX resiεtant colonieε aε the Arg22 dhfr variant haε been reproducibly εhown to be a dominant εelectable marker both in vitro and in vivo (Simonεen and Levinεon, 1983; Carr et al., 1983; Williamε et al., 1987; Iεola and Gordon, 1986; Corey et al., 1990; and the preεent report). Altered dhfr cDNAε may be uεeful in gene tranεfer work in two ways, (1) they can act as dominant εelectable markerε and allow for selection of otherwise nonselectable genes, and (2) they can impart resiεtance to tranεfected cellε such as bone marrow progenitor cells which can permit high dose antifolate chemotherapy of tumors of non-hematologic origin εenεitive to high dose antifolates.

For the first use the ideal mutant dhfr should encode an enzyme with high catalytic efficiency and should amplify in increasing levels of MTX. Toward this goal the Ser31 appearε to be a superior selectable marker than all the other mutantε reported so far (Arg22, Phe22, and Trp31) because it has high catalytic activity, iε relatively resistant to MTX which allows for εtarting selection at fairly high doses of MTX (not possible for the Phe22 mutant) , and is readily a plifiable. The Arg22 and the Trp31 mutants on the other hand have a very poor catalytic activity and would be of limited uεe aε a selectable marker, as the production of a relatively little amount of theεe enzymes would be sufficient to impart resiεtance and the plasmid would not need to be amplified in order to produce more of the DHFR enzyme. Toward the second goal the Arg22 and the Ser3 mutant dhfrs appear equally able to confer MTX resiεtance.

Concluεions

1) Two mutant human dhfrs, the Ser31 and the Ser34, have been shown to act in a dominant selectable manner in tranεfected CHO cells. Selection of transfected cellε in increaεing concentrationε of #

MTX is posεible uεing these mutants. The Ser31 mutant at lower MTX concentrations generates a higher number of colonies than either the human Ser34 or the murine Arg22 mutant. 2) The Ser31 and the Ser34 mutant human dhfr cDNAs are amplifiable in increasing concentrations of MTX. This should allow amplification of cotransfected nonselectable geneε.

Transfection of a tryptophan-15 mutant murine dhfr cDNA into CHO and mouse narrow progenitor cells

Materials and Methods CHO DG44 and wild type CHO cells were obtained from Dr. L. Chasin, Columbia University, New York as deεcribed above and in Urlaub et al., 1983. The murine dhfr cDNA containing the point mutation at nucleotide 46 (amino acid residue 15) waε exciεed from the bacterial expreεεion vector pKT7 (Schweitzer et al., 1989) with the enzymes Ncol and Hindlll and cloned into the εame reεtriction siteε in the mammalian expreεsion vector pSV5, as described above for the Ser31 and Ser34 mutants. Calcium-phosphate mediated transfectionε of pSV5Trpl5 and pWLNeo into both CHO DG44 and CHO S cellε were performed using the mammalian transfection kit obtained from Stratagene (LaJolla, California) according to the manufacturer's instructions. Selection of successful transfectantε for the DG44 cells were carried out in F12 media lacking hypoxanthine, glycine and thymidine (HGT) as deεcribed above. For εelection of tranεfected CHO S cellε in various MTX concentrationε, 10% dialyzed fetal calf εerum waε uεed. G-418 reεiεtant cell which were εelected from each batch of tranεfection (in 750μg/ml of drug) served as controls. Comparisonε of colony formation in selection media were made between the murine TrplS and the Arg 22 mutants and the human Ser31 mutant DHFR cDNA. For Northern blot analysiε lOμg of total RNA iεolated from tranεfected clones were electrophoreεed on a 1.0% agarose/2M formaldehyde gel. For Southern blot analysis lOμg of genomic DNA was digested with Ncol and Hindlll and electrophoreεed on a 0.8% agarose /TBE gel. All hybridizations were carried out at 42*C with formamide in the hybridization solution. Murine dhfr cDNA containing the point mutation at nt 46 (G to T) was labeled with alpha ³²p-dCTP by the random primer method using a random priming kit (Boehringer Mannheim, IN) to a specific activity of 10^β cpm/μg DNA. All blots were washed for 30 min at 37*C in 1XSSC/1.0%SDS and for another 60 min. at 55*C in 0.1XSSC/1.0%SDS. The blots were then exposed to x-ray film for autoradiography. DOTAP (Boehringer Mannheim, IN) mediated transfection of pSV5Trpl5, pSV5Arg22 and PSV5Ser31 plasmidε into murine bone marrow cellε waε performed according to the manufacturer's inεtructionε. Bone marrow cellε from CBAJ (7-11 weekε) mice were harvested in Iscove's modified Dulbecco'ε medium (IMDM) and a mononuclear cell suspension was prepared by Ficoll hypaque separation. lOμg plasmid DNA mixed with the DOTAP tranεfection reagent waε then added to 2xl0⁶ mononuclear cellε in IMDM and 20% fetal bovine serum and incubated for 38 hours at 37*C After incubation, the bone marrow cells were harvested and uεed for the CFU-GM aεεay in the preεence and absence of MTX as described above.

Reεults and Discuεεion

Transfection of Trpl5 mutant murine DHFR cDNA into DHFR- DG44 cells converted them to the DHFR^* phenotype which suggested that the altered enzyme was catalytically functional (Table 11) . Furthermore, the Trpl5 dhfr cDNA was capable of acting as a dominant selectable marker in CHO S cells, as MTX resiεtant colonies were obtained after transfection and selection in increasing concentrationε of MTX (Table 12) . Northern blot analyεeε of RNA isolated from cells obtained from individual MTX reεiεtant colonieε revealed that the transfected Trpl5 cDNA was expreεεed at high levelε (Fig. 37) . Southern blot analyεis indicated that the plasmid was incorporated into genomic DNA of the tranεfected cellε. The transfected mutant dhfr cDNA was amplified readily by increaεing the εelection preεεure i.e. the MTX concentration within a period of eight weekε as shown by increase in intensity of the bands on the Southern blot (Fig. 38). Quantitation of the gene copy number showed that it increased by approximately 5 and 23 fold over a period of four weeks after selection in 5μM and 15μM MTX respectively.

Other studieε from our laboratory have εhown that the murine mutant dhfr haε a markedly reduced affinity for MTX but is catalytically as efficient as the wild type enzyme (Dicker et al., 1993). It haε been demonstrated that the Trp15 variant dhfr is less stable than the wild type protein, suggesting that there may be a need to compensate for this instability by increasing copy number so that overall enzyme function and hence the resiεtance phenotype remains the same. This may explain the amplifiability of the Trpl5 mutant cDNA in response to increasing selection pressure. In the in vitro CFU-GM asεay in preεence of 10^"7M MTX, the Trpl5 cDNA tranεfected bone marrow cellε gave riεe to reεistance colonies and were comparable to the resiεtance levelε of thoεe imparted by the Arg 22 and the Ser31 cDNAε (Table 13) . Like the Ser31 and the Ser34 mutantε, the Trp15 mutant cDNA alεo appears to be a uεeful dominant εelectable marker in gene transfer studies.

Our studieε herein uεed a murine mutant dhfr. In the murine εystem the triplet code for the amino acid glycine is GGG. To encode tryptophan (TGG) only a εingle baεe change of G to T iε required. In the human dhfr however the triplet code for glycine at reεidue 15 is GGC while that for tryptophan is TGG. Thus, a two base change is required to encode tryptophan instead of glycine. Aε a two baεe change iε leεε likely than a single base change, iεolation of a naturally occurring Trpl5 mutant verεion in antifolate reεiεtant human dhfr iε leεε likely. cDNA encoding the human Trp 15 dhfr mutation may be constructed by Polymerase Chain Reaction (PCR) amplified mutagenesiε using a mutation εpecific oligonucleotide primer.

Table 11. COLONY FORMATION IM HGT BY TRANSFECTED CHO DG CELLS ¹

¹colonies expressed as per μg DNA/10⁶ cells

^*these were nonviable colonieε aε they did not propagate upon subcloning in selection media lacking HGT.

Table 12. COLONY FORMATION BY TRANSFECTED CHO 8 CELLS IN MTX¹

'colonies expressed as per μg DNA/10⁶ cells

Table 13. CFU-GM A88AY IN THE PRESENCE AND ABSENCE OF MTX

REFERENCES OF THE SECOND SERIES OF EXPERIMENTS

1. Blakley, R.L. (1984) Dihydrofolate reductaεe in R.L. Blakley & S.J. Benkovic, eds. Folates and Pterins. John Wiley & Sons. New York. Vol 1 pp. 191-253.

2. Carr, F. , Medina, W.D., Dube, S.K., and Bertino, J.R. (1983) Genetic transformation of murine bone marrow cells to methotrexate resiεtance. Blood 62:180-185.

3. Chomczynεki, P., and Sacchi, N. (1987) Single-εtep method of RNA isolation by acid guanidinium thiocyanate phenol chloroform extraction. Anal. Biochem. 162:156-159.

4. Cline, M.J., Stang, H. , Mercola, K. , Morse, L. , Ruprecht, R. , Browne, J., and Salse, W. (1980) Gene transfer in intact animalε. Nature 284:422-425.

5. Corey, CA. , DeSilva, A.D., Holland, CA. , and Williamε, D.A. (1990) . Serial tranεplantation of methotrexate resistant bone marrow: protection of murine recipients from drug toxicity by progeny of transduced stem cells. Blood 76:337-343.

6. Dicker, A.P., Waltham, M. , Volkenandt, M. , Schweitzer, B.I., Otter, G.M., Schmid, F. , Sirotnak, F.M. and Bertino, J.R. Methotrexate resistance in an in vivo mouse tumor due to a non active site dihydrofolate reductase mutation. Proc. Natl. Acad. Sci. U.S.A. 90:11797-11801 (1993).

7. Haber, D.A. , Beverly, S.M., Kiely, M.L. , and Schimke, R.T. (1981) Propertieε of an altered dihydrofolate reductaεe encoded by amplified geneε in cultured mouεe fibroblasts. J. Biol. Chem. 256:9501-9510.

8. Hussein, A., Lewis, D. , Yu, M. , and Melera, P.W. (1992) Construction of a dominant εelectable marker uεing a novel dihydrofolate reductaεe. Gene 112:179-188.

9. Iεola, L.M. , and Gordon, J.W. (1986) Systemic reεiεtance to methotrexate in tranεgenic mice carrying a mutant dihydrofolate reductase gene. Proc. Natl. Acad. Sci. USA 83:9621-9625.

10. Kamen, B.A. , Eibl, B. , Cashmore, A., and Bertino, J.R. (1984) Uptake arid efficacy of trimetrexate {TMQ, 2,4 diamino-5 methy1-6-[ (3,4,5 trimethoxyl) ] quinazoline) Biochem. Pharmacol. 33:1697-1699.

11. Mclvor, R.S., and Simonεen, CC. (1990) Isolation and characterization of a variant dihydrofolate reductase cDNA from the methotrexate resistant murine L5178Y cells. Nucleic Acids Reε. 18:7025-7032.

12. Oefner, C, D'Arcy, A., and Winkler, F.K. (1988) Cryεtal εtructure of h-DHFR complexed with folate. Eur. J. Biochem. 74:377-385.

13. Pizzorno, G. , Chang, Y.-M., McGuire, J.J., and Bertino, J.R. (1989) Inherent reεiεtance of human εquamouε carcinoma cell lines to MTX aε a reεult of decreased polyglutamylation of this drug. Cancer. Res. 49:5275-5280. 14. Schweitzer, B.I., Srimatkandada, S., Gritsman, H. , Sheridan, R. , Venkataraghavan, R. , and Bertino, J.R. (1989) Probing the role of two hydrophobic active site residues in dihydrofolate reductase by site directed mutageneεiε. J. Biol. Chem. 264:20786-20795.

15. Schweitzer, B.I., Gritεman, H. , Dicker, A.P., Volkenandt, M. , and Bertino, J.R. (1989b) Mutations at hydrophobic residues in dihydrofolate reductase in Curtius, H.C, Ghisla, S., and Blau, N. (eds). Chemistry and Biology of Pteridines. deGruyter, New York, pp. 760-764.

16. Simonsen, CC (1986) Drug resistant dihydrofolate reductases. la Malacinski, G., Simonsen, CC, and Shepard, M. (eds) . The Molecular Genetics of Mammalian Cells. Macmillan Pub. Co., New York, pp. 99-128.

17. Simonsen, CC, and Levinson, A.D. (1983) Isolation and expresεion of an altered dihydrofolate reductase cDNA. Proc. Natl. Acad. Sci. USA 80:2495-2499.

18. sirotnak, F.M. , Moccio, D.M. , Kelleher, L.E., and Goutas, L.J. (1989) Relative frequency and kinetic properties of transport defective phenotypes among L1210 clonal cell lines derived in vivo. Cancer Res. 41:4447-4452.

19. Sobrero, A., and Bertino, J.R. (1986) Clinical aspects of drug resistance. Cancer Surveys 5:93-107.

20. Srimatkandada, S., Schweitzer, B.I., Moroson, B.A. , Dube, S., and Bertino, J.R. (1989) Amplification of a polymorphic dihydrofolate reductase gene expreεsing an enzyme with a decreased binding to MTX in a human colon carcinoma cell line HCT-8R4, reεiεtant to thiε drug. J. Biol. Chem. 264:3524-3528.

21. Thillet, J. , Abεil, J., Stone, S.R, and Pictet, R. (1988) Site directed mutagenesis of mouse dihydrofolate reductase. J. Biol. Chem. 263:12500-12508.

22. Thillet, J., and Pictet, R. (1990) Tranεfection of DHFR- and DHFR+ mammalian cells using methotrexate resiεtant mutantε of mouεe dihydrofolate reductaεe. FEBS. Lettε. 269:450-453.

23. Urlaub, G. , Kaε, E. , Carothers, A.M. and Chasin, L.A. Deletion of the diploid dihydrofolate reductase locus from cultured mammalian cells. Cell. 33:405- 412 (1983).

24. Williamε, D.A. , Hεieh, K. , DeSilva, A., and Mulligan, R.C. (1987) Protection of bone marrow tranεplant recipientε from lethal doses of methotrexate by the generation of methotrexate resiεtant bone marrow. J. Exp. Med. 166:210-218.

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT: Bertino, Joseph R. Gilboa, Eli Li, Ming-Xia Schweitzer, Barry I Banerjee, Debabrata Zhao, Shi-Cheng.

(ii) TITLE OF INVENTION: PROTECTION OF HUMAN BONE MARROW FROM HIGH DOSE ANTIFOLATE THERAPY USING MUTATED HUMAN DIHYDROFOLATE REDUCTASE DNA

(iii) NUMBER OF SEQUENCES: 10

(iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: Cooper & Dunham

(B) STREET: 30 Rockefeller Plaza

(C) CITY: New York

(D) STATE: New York

(E) COUNTRY: U.S.A.

(F) ZIP: 10112

(v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFTWARE: Patentln Release #1.0, Version #1.25

(vi) CURRENT APPLICATION DATA:

(A) APPLICATION NUMBER: Not Yet Known

(B) FILING DATE: 13-APR-1994

(C) CLASSIFICATION:

(vii) PRIOR APPLICATION DATA:

(A) APPLICATION NUMBER: 08/049,284

(B) FILING DATE: 13-APR-1993

(viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: White, John P.

(B) REGISTRATION NUMBER: 28,678

(C) REFERENCE/DOCKET NUMBER: 43366A/JPW/KLK

(ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: (212) 977-9550

(B) TELEFAX: (212) 664-0525

(C) TELEX: 422523 COOP UI

(2) INFORMATION FOR SEQ ID NO:l:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 27 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO #

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:l: CGATCGAGGA TCCCAAGCTT ACCTTTT 2

(2) INFORMATION FOR SEQ ID NO:2:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 33 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:2: ATCATCCCAT GGTTGGTTCG CTAAACTGCA TCG 3

(2) INFORMATION FOR SEQ ID NO:3:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 25 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:3: GCTGCCATCC ATGGTTCGAC CATTG 2

(2) INFORMATION FOR SEQ ID NO: :

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 33 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: : TCTAAAGCCA GCAAAAGCTT CATGGTCTTA TAA 3

(2) INFORMATION FOR SEQ ID NO:5:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 20 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:5: GGAAGCCGGT CTTGTCGATC 2

(2) INFORMATION FOR SEQ ID NO:6:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 20 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:6: CGAAATCTCG TGATGGCAGG 2

(2) INFORMATION FOR SEQ ID NO:7:

(1) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 21 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:7: TGCCAATTCC GGTTGTTCAA T 2 (2) INFORMATION FOR SEQ ID NO:8:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 18 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:8: TCTGTCCTTT AAAGGTCG 1

(2) INFORMATION FOR SEQ ID NO: :

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 21 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:9: GAGGTTCCTT GAGTTCTCTG C 2

(2) INFORMATION FOR SEQ ID NO:10:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 18 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic) (iii) HYPOTHETICAL: NO (iv) ANTI-SENSE: NO

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10: CCTCGGCCTC TGAGCTAT 1

Claims

* What is claimed is:

1. A DNA vector which comprises DNA encoding a mutant, antifolate resistant, dihydrofolate reductase inserted into a site within the vector, the presence of which site is not essential for replication of the vector.

2. A DNA vector of claim 1, wherein the mutant dihydrofolate reductase has substantially the same amino acid sequence as naturally occurring human dihydrofolate reductase.

3. A DNA vector of claim 2, wherein the mutant dihydrofolate reductase differs from naturally occurring human dihydrofolate reductase by virtue of the presence of a serine residue at position 31 or 34.

4. A DNA vector of claim 2, wherein the mutant dihydrofolate reductase differs from naturally occurring human dihydrofolate reductase by virtue of the presence of a tryptophan residue at position 15.

5. A DNA vector of claim 1, wherein the DNA encoding the mutant dihydrofolate reductase is operatively linked at its 5' end to a promoter sequence and at its 3'end to a polyA sequence.

6. A DNA vector of claim 5, wherein the promoter sequence is an SV40 promoter.

7. A plasmid which comprises the vector of claim 1. _*

8. The plasmid of claim 7 designated pSV5-Ser31 h-DHFR (ATCC Accession No. 75441) .

9. The plasmid of claim 8 designated pSV5-Ser34 h-DHFR (ATCC Accession No. 69276) .

10. A retroviral DNA vector of claim 1.

11. A retroviral vector of claim 10, wherein the vector comprises DNA from a retrovirus corresponding to a

5' long terminal repeat, a 3' long terminal repeat, and a packaging signal.

12. A retroviral vector of claim 11, wherein the site at which the DNA encoding the mutant dihydrofolate reductase inserted is in the 3 ' long terminal repeat.

13. A plasmid which comprises the retroviral vector of claim 12.

14. The plasmid of claim 13 designated pDC SV S31 h-DHFR (ATCC Accession No. 75440) .

15. A mammalian retroviral producer cell which comprises the vector claim 1 or the plasmid of claim 13 or 14.

16. A human cell which comprises the vector of claim 1 or the plasmid of claim 13 or 14.

17. A hematopoietic human cell of claim 16.

18. A bone marrow cell of claim 17.

19. A method for reducing the toxic effects of antifolate therapy in a subject which comprises replacing the subject's hematopoietic cells with hematopoietic cells of claim 17 so as to reduce the toxic effects of antifolate therapy in the subject.

20. A method of claim 19, wherein the antifolate is methotrexate.