WO1997012052A1

WO1997012052A1 - Transduction of hematopoietic cells by viral vector and a cationic lipid

Info

Publication number: WO1997012052A1
Application number: PCT/US1996/015580
Authority: WO
Inventors: Pervin Anklesaria; Carmel M. Lynch
Original assignee: Targeted Genetics Corporation
Priority date: 1995-09-28
Filing date: 1996-09-27
Publication date: 1997-04-03
Also published as: AU7378396A

Abstract

The present invention provides methods for the high efficiency transduction of hematopoietic cells that involve exposing the cells to combinations of viral vectors and cationic lipid adjuvants. Among hematopoietic cells, the methods are especially useful for the transduction of pluripotent hematopoietic stem cells (PHSCs) which have been difficult to transduce with high efficiency.

Description

TRANSDUCTION OF HEMATOPOIETIC CELLS BY VIRAL VECTOR AND A CATIONIC LIPID

Technical Field

The present invention is related to the genetic transduction of cells using viral vectors. More particularly, the present invention is related to methods for the genetic transduction of hematopoietic cells, especially pluripotent hematopoietic stem cells, using viral vectors.

BACKGROUND OF THR INVENTION Hematopoietic cells

The hematopoietic system is populated by cells of multiple lineages (including all cells ofthe myeloid and lymphoid systems). At the origin of hematopoiesis are the pluripotent hematopoietic stem cells (abbreviated herein as "PHSCs") which are capable of giving rise, through cell division and differentiation, to all ofthe mature cells ofthe hematopoietic system. As is known in the art, PHSCs can be obtained from a number of different sources, including bone marrow and cord blood, as well as from peripheral blood from mammals that have been mobilized with G-CSF, GM-CSF and/or other cytokines. There are two major lineages (i.e. committed progeny) derived from PHSCs: the myeloid lineage (which includes cells such as red blood cells, granulocytes, monocytes and megakaryocytes); and the lymphoid lineage (which includes cells such as B and T lymphocytes). The level of expression of various cell-surface antigens can be used as convenient indicators to distinguish between the lineages, and to distinguish between cell types and maturational stages within each lineage, as is known in the art. In murine bone marrow, for example, a population of cells that are Scal+Lin- contain PHSCs and these cells can reconstitute the hematopoietic system after introduction (i.e. "engraftment") into lethally-irradiated mice. [1] (Note that, as used herein, the suffix "-" (i.e. negative, as in "Thyl-") is used to indicate that a particular marker is either absent or is found at low levels on a particular cell (as compared to cells designated as "+" (i.e. positive) for the marker); note also that numbers in brackets refer to the numbered references which are cited below). For murine cells, such undifferentiated PHSCs represent only a small percentage (approx. 0.02 - 0.1%) ofthe total number of hematopoietic cells in the bone marrow.

In humans and non-human primates, reconstitutive hematopoietic activity is present within a population of cells that are defined as CD34+. This CD34+ population comprises approximately 1% of the total hematopoietic cell population. It has recently been demonstrated that PHSCs can be further enriched on the basis of CD38 antigen expression. In particular, the level of CD38 antigen expression increases together with the expression of early markers ofthe myeloid and lymphoid lineages. Thus, cells characterized as CD34+CD38- are believed to represent the most primitive hematopoietic sub-population which compromises about 1% ofthe CD34+ cells or about 0.01% of total hematopoietic cells. [2] Human PHSCs are also highly enriched in the CD34+Lin-Thyl+ population. [3] (The phenotypic designation "Lin-" is used to refer to an absence of cellular markers that appear on various committed cells ofthe hematopoietic system (CD38 being an example of one such "Lin" marker); as is described in further detail below.)

Functionally, PHSCs exhibit extensive proliferative potential and a capacity to differentiate into all hematopoietic lineages. Knowledge and understanding of hematopoiesis has expanded with the development of numerous in vitro and in vivo assays for hematopoietic development, with the identification and characterization of various hematopoietic growth factors and with the development of strategies for enriching for sub-populations comprising PHSCs. [4]

To further assess the functionality of human PHSC populations, several assays have been developed. These include, for example, the engraftment of human hematopoietic cells into "SCID" mice (which lack their own immune as a result of severe combined immunodeficiency disease); engraftment of human CD34+ cells and hrIL-3 transfected stromal cells into beige/nude/xid homozygous mice; and engraftment of human fetal liver and bone marrow cells into fetal sheep. [4] Studies have demonstrated that the human cells engrafted into SCID mice survived for at least 14 weeks in the bone marrow (albeit at a low percentage) and can give rise to both myeloid and lymphoid cell lineages in the SCID recipients. Human CD34+ cells, after being cultured in the presence of an allogeneic stromal layer for 3 days, survived for over 9 months following transplantation into beige/nude/xid homozygous mice. [5] Xenogeneic donor cells have also been engrafted into fetal sheep, and the resulting chimeric sheep contained human cells in both bone marrow ("BM") and peripheral blood for a period of greater than 2 years. [6] A variety of in vitro assays have also been developed for human PHSCs. These include, for example, the long-term culture initiating cell ("LTC-IC") assay [7], the high proliferative potential colony forming cell ("HPP-CFC") assay [8], and the cobblestone area forming cell ("CAFC") assay. [9-11] Human hematopoietic cells characterized by these assays are believed to possess the multipotential and proliferative properties of PHSCs.

Significance of PHSCs and the ability to transduce PHSCs

Pluripotent hematopoietic stem cells (PHSCs) are particularly attractive targets for gene transduction for several reasons. First, PHSCs can be readily obtained from a variety of sources as noted above. Second, procedures for the transplantation of PHSCs into humans and non-human animals are well established. Third, and most significant, successful transduction and subsequent clonal expansion of PHSCs would result in the presence of a transduced gene in all hematopoietic lineages, potentially lasting for the life-time ofthe PHSC transplantation recipient. This approach would thus provide a powerful therapeutic technique for treating specific inherited or acquired diseases caused by a defect in a particular gene that affects functions of cells ofthe hematopoietic system. Examples include such diseases as ADA (adenosine deaminase deficiency), Gaucher's disease, sickle cell anemia, and AIDS among others. Other potential applications would include, for example, the introduction of genes enabling hematopoietic cells to secrete a desired protein and/or altering the resistance of such cells to particular agents (including, for example, chemotherapeutic agents or infectious agents (e.g. viruses such as HIV)). Limitations of PHSC transduction methods

While PHSCs of mice have been transduced with viral vectors, transduction has generally required direct co-cultivation of target PHSCs with cell lines that produce the viruses. Such co-cultivation protocols are not approved for human gene therapy in the United States.

Limited gene transfer into PHSCs of larger animals (such as canine and primate PHSCs) has been achieved using protocols involving at least several days of co-culture with viral producer lines and with autologous stroma (or an engineered murine stromal cell line expressing human transmembrane "steel factor", abbreviated "SF"). The presence of genetically-modified cells ranged from levels of about 0.01 % to OJ % for cells of myeloid lineage. Significantly, a number of these approaches resulted in a much higher level of genetically-modified B lymphocytes (1-14%) and T lymphocytes (0.3 to 3%) compared with myeloid cells in the reconstituted recipients. Since such B and T lymphocytes are long-lived, the high percentage of transduced lymphoid cells may actually reflect transduction of committed lymphoid cells rather than the transduction of uncommitted PHSCs per se. Issues of lineage can be addressed by analysis of unique integration sites of recombinant viruses in order to trace the progeny ofa single primitive cell in re-populated recipients. The foregoing studies are reflected in a number of references. [12-21] In summary, gene transfer into PHSCs of higher-mammalian origin (such as human, non-human primate and canine PHSCs) remains very inefficient at best. Several factors may account for these low transduction frequencies. First, retroviral-based vectors are believed to require cell division in order to cross the nuclear membrane and/or integrate into the host cell genome. While PHSCs can be induced to divide in culture using various growth factors or other agents, such cycling is frequently accompanied by cellular differentiation, resulting in the reduction or loss of PHSC functionality (i.e. the ability of PHSCs to efficiently engraft and re-populate the hematopoietic system). Second, PHSCs may exhibit poor expression of viral receptors.

Studies involving gene transfer into human PHSCs (from bone marrow or mobilized peripheral blood) have utilized protocols similar to those described for large animal studies as well as other protocols which have included the use of fibronectin and/or long-term culture conditions. Such studies have assessed the clonogenic progenitor cells present at week 5 of long-term bone marrow culture. Based on data from large animal models and clinical trials, it now seems apparent that results measured using these relatively mature progenitor cells are not very predictive of success at transducing long-lived stem cells (measured after transplantation into irradiated recipients). Clinical trials using retroviral-mediated gene transfer into human hematopoietic cells have illustrated the feasibility of gene transfer. However, current protocols optimized on the basis ofthe large animal models and/or human clinical trials have failed to target PHSCs at efficiencies high enough to generate production of therapeutically significant quantities of genetically-modified PHSCs. In addition, as noted above, protocols that involve long-term culturing of PHSCs in vitro tend to effectively eliminate the desired target cells because such cycling is believed to result in cellular differentiation (and concomitant loss ofthe PHSC phenotype).

Thus, there remains a need for efficient methods useful for the genetic transduction of hematopoietic cells, particularly PHSCs.

SUMMARY OF THE INVENTION The present invention provides methods for the high efficiency transduction of hematopoietic cells that involve exposing the cells to combinations of viral vectors and cationic lipid adjuvants. Among hematopoietic cells, the methods are especially useful for the transduction of pluripotent hematopoietic stem cells (PHSCs) which have been difficult to transduce with high efficiency.

Exemplary embodiments ofthe present invention include the following:

1. A method of transducing a hematopoietic cell, comprising the steps of: (a) exposing said hematopoietic cell to a cationic lipid adjuvant; and (b) exposing said hematopoietic cell to a viral vector.

2. A method of transducing a hematopoietic cell according to embodiment 1, wherein steps (a) and (b) are carried out in a coincidental manner by exposing the hematopoietic cell to a combination ofthe cationic lipid adjuvant and the viral vector.

3. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the hematopoietic cell is a pluripotent hematopoietic stem cell (PHSC). 4. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the hematopoietic cell is a CD34+ hematopoietic cell.

5. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the hematopoietic cell is a Lin- hematopoietic cell. 6. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the hematopoietic cell is a human CD34+CD38- hematopoietic cell.

7. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the viral vector is a retroviral vector. 8. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the hematopoietic cell has been pre-stimulated with at least one cytokine.

9. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the hematopoietic cell has been pre-stimulated with a cytokine cocktail comprising at least two cytokines.

10. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, further comprising the step of subjecting the hematopoietic cell to metabolic induction of receptors for the viral vector prior to exposure ofthe hematopoietic cell to the viral vector and the cationic lipid adjuvant. 11. A method of transducing a hematopoietic cell according to embodiment 10, wherein the step of subjecting the hematopoietic cell to metabolic induction of receptors for the viral vector is conducted by depriving the hematopoietic cell of phosphate prior to exposure ofthe hematopoietic cell to the viral vector and the cationic lipid adjuvant.

12. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the cationic lipid adjuvant comprises a lipopolyamine.

13. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the cationic lipid adjuvant comprises a cationic lipid and a neutral lipid.

14. A method of transducing a hematopoietic cell according to embodiment 13, wherein the cationic lipid adjuvant comprises a lipopolyamine and a neutral lipid. 15. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the cationic lipid adjuvant comprises a lipid selected from the group consisting of DOSPA, DOTMA, DMRIE, TM-TPS and DDAB.

16. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the cationic lipid adjuvant comprises a lipid having a cholesteryl moiety.

17. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the cationic lipid adjuvant comprises cholesterol.

18. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the cationic lipid adjuvant comprises a lipid selected from the group consisting of DOSPA, TM-TPS and DMRIE.

19. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein the cationic lipid adjuvant comprises a lipid composition selected from the group consisting of DOSPA/DOPE, TM-TPS/DOPE and DMRIE/Chol.

20. A method of transducing a hematopoietic cell according to embodiment 19, wherein the cationic lipid adjuvant comprises DOSPA/DOPE.

21. A method of transducing a hematopoietic cell according to embodiment 19, wherein the cationic lipid adjuvant comprises TM-TPS DOPE. 22. A method of transducing a hematopoietic cell according to embodiment 19, wherein the cationic lipid adjuvant comprises DMRIE/Chol.

23. A method of transducing a hematopoietic cell according to embodiment 8, wherein said cytokine is selected from the group consisting of interleukin-l, interleukin-3, interleukin-6, FLT3L and Steel factor. 24. A method of transducing a hematopoietic cell according to embodiment 9, wherein the hematopoietic cell has been pre-stimulated with at least two cytokines selected from the group consisting of interleukin-l, interleukin-3, interleukin-6, FLT3L and Steel factor.

25. A method of transducing a hematopoietic cell according to embodiment 24, wherein the hematopoietic cell has been pre-stimulated with a cytokine cocktail comprising interleukin-3, interleukin-6 and Steel factor. 26. A method of transducing a hematopoietic cell according to embodiment 24, wherein the hematopoietic cell has been pre-stimulated with a cytokine cocktail comprising interleukin-l, interleukin-6, Steel factor and FLT3L.

27. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein said viral vector comprises a therapeutic gene.

28. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, wherein said viral vector comprises a marker gene selected from the group consisting ofa detectable marker gene and a selectable marker gene.

29. A method of transducing a hematopoietic cell according to any ofthe preceding embodiments, further comprising after steps (a) and (b), the following step:

(c) selecting a hematopoietic cell that has been transduced by the viral vector.

30. A method of transducing a hematopoietic cell according to embodiment 28, wherein the viral vector comprises a selectable marker gene and said step of selecting a hematopoietic cell that has been transduced by the viral vector is conducted by exposing the hematopoietic cell to a selective agent.

31. A method of transducing a hematopoietic cell according to embodiment 28, wherein the viral vector comprises a gene encoding a detectable cell surface marker and said step of selecting a hematopoietic cell that has been transduced by the viral vector is conducted by detecting expression ofthe cell surface marker on the hematopoietic cell. 32. A hematopoietic cell transduced by the method of one of embodiments 1 to

31 , and progeny thereof.

33. A method of treating a patient for a disease condition, comprising the steps of: (1) transducing a hematopoietic cell according to the method of one of embodiments 1 to 31 , wherein the viral vector comprises a therapeutic gene; and (2) administering a transduced hematopoietic cell of step (1) to the patient.

It should be noted that while the present invention is especially useful for the transduction of hematopoietic cells, particularly PHSCs (which have been difficult to transduce with high efficiency), the present invention is also useful for the transduction of other (non-hematopoietic) target cells (i.e. by using cationic lipid adjuvants in conjunction with viral vectors, as described and illustrated herein). DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS Definitions

"Pluripotent hematopoietic stem cells" (abbreviated "PHSCs") are cells which are capable of giving rise, through cell division and differentiation, to all ofthe mature cells of the hematopoietic system. PHSCs include progenitor cells with significant though limited capacity for self-renewal, and still more primitive cells possessing long-term and/or multilineage re-populating ability in a transplanted mammalian host. A variety of methods have been described for assaying progenitor cells and their functionality, as described herein and in the art. "Hematopoietic cells" include the various mature cells ofthe myeloid and lymphoid systems (including lymphocytes and other blood cells), as well as pluripotent hematopoietic stem cells.

"Host cell", "recipient cell", "target cell", and other such terms denote higher mammalian cells, most preferably human cells, which can be or have been used as recipients for transduction by viral vectors, and include the progeny ofthe original cell which has been transduced. It is understood that the progeny ofa single cell may not necessarily be completely identical (in morphology or in genomic or total DNA complement) to the original parent cell, due to natural, accidental or deliberate mutation. "Transduction," as used herein, refers to the introduction of a polynucleotide of interest into a host cell by viral-mediated delivery, such as by infecting the host cell with a viral vector carrying the polynucleotide.

A "viral vector," as used herein, refers to a virus that is capable of mediating transfer of a polynucleotide from the virus to a host cell, in a process referred to as transduction. The transferred polynucleotide may be stably or transiently maintained in the host cell. In preferred examples of such vectors, the transferred polynucleotide contains terminal repeats allowing it to be stably integrated into a replicon ofthe host cell (such as nuclear or mitochondrial DNA). Both DNA and RNA viruses are known which can be used to mediate transduction using the methods ofthe present invention. Indeed, a large variety of such viral vectors are well known in the art and are widely available. Where long-term maintenance ofthe transduced gene is desirable, RNA "retroviruses" are presently the most preferred class of such viral vectors. The gene or genes to be transferred can comprise any nucleotide sequence(s) that it is desirable to transfer to the host cells. Such genes might include, for example, therapeutic genes as well as detectable and/or selectable marker genes.

"Retroviruses" are a class of viruses which use RNA-directed DNA polymerase, or "reverse transcriptase," to copy a viral RNA genome into a double-stranded DNA intermediate which can be incorporated into chromosomal DNA of an avian or mammalian host cell. Many such retroviruses are known to those skilled in the art and are described, for example, in Weiss et al., eds, RNA Tumor Viruses. 2d ed., Cold Spring Harbor, New York (1984 and 1985). Plasmids containing retroviral genomes are also widely available, from the American Type Culture Collection (ATCC) and other sources. The nucleic acid sequences of a large number of these viruses are known and are generally available from databases such as GENBANK, for example.

A "polynucleotide" refers to a polymeric form of nucleotides of any length, eidier ribonucleotides or deoxyribonucleotides, or analogs thereof. This term refers only to the primary structure ofthe molecule. Thus, double- and single-stranded DNA, as well as double- and single- stranded RNA are included. It also includes various modified polynucleotides.

A "gene" refers to a polynucleotide or portion of a polynucleotide comprising a sequence that encodes a protein. For most situations, it is desirable for the gene to also comprise a promoter operably linked to the coding sequence in order to effectively promote transcription. Enhancers, repressors and odier regulatory sequences may also be included in order to modulate activity ofthe gene, as is well known in the art (see, e.g., the references cited below).

A "detectable marker gene" is a gene that allows cells carrying the gene to be specifically detected (i.e. to be distinguished from cells which do not carry the marker gene). A large variety of such marker genes are known in the art. Preferred examples thereof are detectable marker genes which encode proteins appearing on cellular surfaces, thereby facilitating simplified and rapid detection and/or cellular sorting. By way of illustration, we utilized an alkaline phosphatase ("AP") gene as a detectable marker; which allowed cells transduced with a vector carrying the AP gene to be detected and/or sorted based on expression of AP on the surface of transduced cells.

A "selectable marker gene" is a gene that allows cells carrying the gene to be specifically selected for or against, in the presence ofa corresponding selective agent. By way of illustration, an antibiotic resistance gene can be used as a positive selectable marker gene that allows a host cell to be positively selected for in the presence ofthe corresponding antibiotic. A variety of positive and negative selectable markers are known in the art, some of which are described below. A "cationic lipid" comprises a hydrocarbon chain (or "tail") attached to a positively-charged head group. The hydrocarbon tail may be linear or branched, may be aliphatic and/or aromatic in structure, and may be saturated or unsaturated. The head group may possess a single net positive charge (monovalently cationic) or multiple net positive charges (polyvalently cationic). A large variety of such cationic lipids have been described in the art and are readily synthesized and/or are widely available from commercial sources.

A "lipopolyamine" is a type of cationic lipid that has, as part of its head group, a polyamine or analog thereof. Polyamines are compounds containing at least two amino groups; including, for example, compounds such as spermine, spermidine and putrescine. In the case of spermine, for example (which is the polyamine incorporated into the

"lipofectamine" reagent described below), the srjermine moiety consists of an aliphatic hydrocarbon having charged amino groups incorporated in its backbone (see, e.g., GIBCO/BRL publication FOCUS, 15:73, 1993, and Behr, Bioconj. Chem., 5:382, 1994). A "cationic lipid adjuvant" is a molecule or combination of molecules comprising a cationic lipid. A cationic lipid adjuvant may comprise, for example, a mixture of a different cationic lipids. A cationic lipid adjuvant may also comprise one or more neutral lipids. A variety of such neutral lipids have been described and are widely available. A variety of lipid combinations are also widely available. As an illustrative example referred to below, "Lipofectamine" comprises a lipopolyamine (i.e. "DOSPA") as the cationic lipid, and also comprises a neutral lipid (i.e. "DOPE") ["DOSPA" is 2,3- dioleoyloxy-N-[2(spermine carboxamido)ethyl]-N,N-dimethyl-l -propanaminium trifluoroacetate; and "DOPE" is dioleoyl-phosphatidylethanolamine]. Cationic lipid adjuvants may also comprise other, non-lipid molecules.

A "cytokine" refers to a polypeptide that is a soluble intercellular signalling molecule, including, for example, interleukins, interferons and colony stimulating factors

(CSFs). Preferred classes of cytokines for use with the present invention include interleukins, colony stimulating factors and other cytokines that stimulate cell division in pluripotent hematopoietic stem cells, as described below.

A "therapeutic gene" refers to a nucleotide sequence that is capable, when transferred to a patient, of eliciting a prophylactic, curative or other beneficial effect in the patient.

"Treatment" or "therapy" as used herein refers to administering, to a patient, cells or other agents (or combinations thereof) that are capable of eliciting a prophylactic, curative or other beneficial effect in the patient.

A "patient" as used herein refers to a higher mammal, preferably a human.

References

The practice ofthe present invention will employ, unless otherwise indicated, conventional techniques of molecular biology, microbiology, recombinant DNA, immunology and stem cell biology, which are within the skill ofthe art. Such techniques are explained fully in the literature. Sss. e.g., MOLECULAR CLONING: A LABORATORY MANUAL, (J. Sambrook et al., Cold Spring Harbor Laboratory, Cold

Spring Harbor, N.Y., 1989); CURRENT PROTOCOLS IN MOLECULAR BIOLOGY (F.M. Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Siedman, J.A. Smith, and K. Struhl, eds., 1987 and updated); ESSENTIAL MOLECULAR BIOLOGY (T.A. Brown, ed., IRL Press, Oxford, 1991); GENE EXPRESSION TECHNOLOGY (Goeddel, ed., Academic Press, San Diego, 1991 ); METHODS FOR CLONING AND ANALYSIS OF

EUKARYOTIC GENES (A. Bothwell et al. (eds), Bartlett Publ., Boston, 1990); GENE TRANSFER AND EXPRESSION (M. Kriegler, Stockton Press, New York, 1990); RECOMBINANT DNA METHODOLOGY (R. Wu et al. (eds.), Academic Press, San Diego, 1989); PCR: A PRACTICAL APPROACH (M.J. McPherson et al., IRL Press at Oxford University Press, 1991 ); CELL CULTURE FOR BIOCHEMISTS (R.L.P. Adams ed., Elsevier Science Publishers, Amsterdam, 1990); GENE TRANSFER VECTORS FOR MAMMALIAN CELLS (J.M. Miller and M.P. Calos eds. 1987); MAMMALIAN CELL BIOTECHNOLOGY (M. Butler (ed.), 1991); ANIMAL CELL CULTURE (J.W. Pollard et al. (eds.) Humana Press, Clifton, N.J., 1990); CULTURE OF ANIMAL CELLS (RJ. Freshney et al. (ed.), Alan R. Liss, New York, 1987); FLOW

CYTOMETRY AND SORTING (M.R. Mela ed et al. (eds.), Wiley-Liss, New York, 1990); the series METHODS IN ENZYMOLOGY (Academic Press, Inc.); HANDBOOK OF EXPERIMENTAL IMMUNOLOGY, (D.M. Weir and CC. Blackwell, Eds.); CELLULAR AND MOLECULAR IMMUNOLOGY (A.K. Abbas, A.H. Lichtman and J.S. Prober, Saunders, 1991, 1994); CURRENT PROTOCOLS IN IMMUNOLOGY (J.E. Coligan, A.M. Kruisbeek, D.H. Margulies, E.M. Shevach and W. Strober, eds.,

1991); the series ANNUAL REVIEW OF IMMUNOLOGY; and the series ADVANCES IN IMMUNOLOGY;

Additional references describing cationic lipids, viral vectors, cytokines, hematopoietic cells and other components that may be used in the methods ofthe present invention include the following: TECHNIQUES OF LIPIDOLOGY (M. Kates et al.,

Elsevier Science Publishers, Amsterdam, 1986); ADVANCES IN LIPID METHODOLOGY (W.W. Christie, The Oily Press, Ayr, Scotland, 1992); A LIPID GLOSSARY (F.D. Gunstone et al., The Oily Press, Ayr, Scotland, 1992); Fasman, G.D. (ed.) CRC PRACTICAL HANDBOOK OF BIOCHEMISTRY AND MOLECULAR BIOLOGY (CRC Press, Boca Raton, FL, 1989); GENE TRANSFER VECTORS FOR

MAMMALIAN CELLS (J.M. Miller and M.P. Calos eds. 1987); PRACTICAL MOLECULAR BIOLOGY, VIRAL VECTORS FOR GENE EXPRESSION (M.K.L. Collins (ed.), Humana Press, Clifton, N.J., 1991); GENE TRANSFER AND EXPRESSION PROTOCOLS (E.J. Murray (ed.), Humana Press, Clifton, N.J., 1991); R. Callard and A. Gearing, THE CYTOKINE FACTS BOOK (Academic Press, 1994);

GUIDEBOOK TO CYTOKINES AND THEIR RECEPTORS, (N.A. Nicola, ed., Oxford University Press, 1994); Gordon, A.S. (ed.), REGULATION OF HEMATOPOIESIS, (Appleton, New York, 1970); Baum, S.J., Ledney, G.D., and Kahn, A. (eds.), EXPERIMENTAL HEMATOLOGY TODAY (Springer-Verlag, 1981); McCulloch, E.D., CELL CULTURE TECHNIQUES - CLINICS IN HAEMATOLOGY, Vol. 13, No.

2 (W.B. Saunders, Eastbourne, England, 1984); J.H. Jandl (ed.), BLOOD: TEXTBOOK OF HEMATOLOGY (Little, Brown, Boston, 1987); and LYMPHOCYTES A PRACTICAL APPROACH (G.G.B. Klaus, (ed.), IRL Press, Oxford, England, 1987).

Bibliography of Articles Cited bv Reference Number

1. Spangrude, G.J., et al. (1988) Science, 241:58-62.

2. Terstappen, L.W.M.M., et al. (1991) Blficd, 77:1218-1227. 3. Murray, L., et al. (1995) BJaoj, 85:368-378.

4. Orlic, D., & Bodine, D.M. (1994) BJββd, 84:3991-3994.

5. Nolta, J. A., et al. ( 1995) BJafid, 86: 101 - 110.

6. Zanjani E.D. et al. (1992) J. Clin. Invest. 89: 1178 et seq. 7. Sutherland. H.J.. et al. ( 1990Ϊ Proc. Natl. Acad. Sci. USA. 87r9,:3584-3588

8. McNiece, I.K., et al. (1990) Int. J. Cell Cloning. 8:146 et seq.

9. Ploemacher, R.E., et al. (1989) Blood. 74:2755 et seq.

10. Reading C. et al. ( 1994) Experimental Hematol, 22:406 et seq.

11. Neben S., et al. (1993) Experimental Hematol. 21 :438 et seq. 12. Camel. A., et al. f 1993^{^}1 Experimental Hematology 21 :585-591.

13. Xu, L.C, et al. (1995) Proc, Natl. Acad, Sci. USA, 92:4372-4376.

14. Barquinero, J., et al. (1995) Biojid, 85:1195-1201.

15. Kiem, H.-P., et al. (1994) Blfloi, 83:1467-1473.

16. Schuening, F.G., et al. (1991) BJc_cd, 78:2568-2576. 17. Stead, R.B., et al. (1988) Blood, 71:742-747.

18. Bodine, D.M., et al. (1993) BJcod, 82:1975-1980.

19. Flasshove, M., et al. (1995) BJflβd., 85:566-574.

20. Dunbar, C.E., et al. (1995) BJβfid, 85:3048-3057.

21. Brenner M. et al (1993): The Lancet 342:1134 et seq. 22. Boyum A. (1964) Halme 204:793 et seq.

23. Curtis, B., et al. (1991) Proc. Natl. Acad. Sci. USA. 88:5809 et seq.

24. Kavanaugh M.P. et al. ( 1994) Proc. Natl. Acad. Sci. USA 91 : 7071 et seq.

25. Clowes, M.M., et al. (1994) J. Clin. Invest.. 93:644-651.

26. Miller, A.D. et al. (1991) Proc. Natl. Acad. Sci. USA 65:2220-2224. 27. Miller A.D. and Buttimore, C. (1986) Mol. Cel Biol. 6:2895 et seq.

28. U.S. Patent Application Serial No. 08/471 ,083, filed 6/6/95; continuation of USSN 08/105,471, filed 8/11/93; continuation of USSN 07/660,616, filed 2/22/91.

29. U.S. Patent Serial No. 4,861,719, issued 8/29/89.

30. Crooks, G.M. and Kohn, D.B. (1993). Blood 82:3290 et seq.

All patents, patent applications, and publications mentioned herein, both supra and infra, are hereby incoφorated herein by reference. Detailed Description ofthe Preferred Embodiments

The present invention provides high efficiency methods for the viral-mediated transduction of hematopoietic cells, particularly pluripotent hematopoietic stem cells ("PHSCs"). As described above, the ability to efficiently transduce PHSCs is particularly significant for a number of therapeutic applications, especially applications of gene therapy in which the transduced PHSCs can be re-introduced to patients and thereby used to re-populate portions ofthe hematopoietic system with genetically-modified cells. During successive stages of embryonic development, the major site of hematopoiesis shifts from the yolk sac to the fetal liver and then to the marrow of developing bones. Hematopoiesis is normally limited to bone marrow during adult life and involves patterns of proliferation and differentiation that ultimately yield all ofthe various cells ofthe hematopoietic system.

As is known in the art, PHSCs can be obtained from a number of different sources, including bone marrow and cord blood, as well as from peripheral blood from mammals that have been mobilized with G-CSF, GM-CSF and/or other cytokines. [4] If desired, PHSCs can be isolated by various enrichment protocols which eliminate various populations of other cells. A convenient initial enrichment method involves the isolation of mononuclear cells (which contain PHSCs and other hematopoietic cells) using a Ficoll- Hypaque gradient [22] (see, also, the Examples below).

Enrichment experiments are also useful for elucidating the functional characteristics of PHSC populations. In the mouse, for example, the most primitive sub¬ populations of PHSCs are generally characterized as Scal+Lin-Thyl-. As few as one hundred such isolated cells can reconstitute both the myeloid and the lymphoid compartments oflethally-irradiated transplantation recipients. Further enrichment ofthe

Scal+Lin-Thyl- population on the basis of dyes such as rhodamine 123 can be used to further fractionate PHSCs into resting and activated subsets that differ in their ability to proliferate and to reconstitute lethally-irradiated primary and secondary recipients. Rhodamine 123, for example, is a mitochondrial vital dye which therefore preferentially stains active versus resting cells. Highly-enriched murine PHSCs can thus be obtained by isolating cells that are Scal+Lin-Thyl- and are poorly-stained with rhodamine 123). In humans and non-human primates, reconstitutive hematopoietic activity is present within a population of cells mat are defined as CD34+ (which population comprises approximately 1% ofthe total hematopoietic cell population).

Several general protocols are available for further purifying PHSCs to the extent desired. These include but are not limited to "counter centrifugal elutriation" [4], immuno-based selection procedures (using, e.g., flasks coated with antibodies to particular markers such as CD34 or CD38), affinity columns to enrich for cells stained with labeled antibodies to cellular markers, and multiparametric flow cytometry sorting, as described herein. A presently preferred method for enriching human PHSCs involves immuno-based affinity columns as illustrated below.

If desired, the most primitive sub-populations of human PHSCs can be further isolated by enriching for CD34+ cells that are also "Lin-". The Lin- phenotype refers to cells that lack the various lineage-specific markers ("Lin markers") that appear on committed cells ofthe hematopoietic system (including the various committed progenitor cells as well as the mature cells ofthe myeloid and lymphoid lineages).

Thus, multiparametric sorting (e.g. by flow cytometry) can be used to selectively eliminate cells expressing various markers, and can therefore be used to enrich for progressively more primitive sub-populations of PHSCs to the extent that is desirable for particular applications. A number of markers that can be used are found among the various "CD" (or "cluster of differentiation") markers that become expressed on various committed hematopoietic cells. [See, e.g., Knapp, W. et al. (eds.), "Leukocyte Typing IV. White Cell Differentiation Antigens", (Oxford, 1989); and Abbas, A.K., et al. "Cellular and Molecular Immunology (W.B. Saunders, 1991, 1994)].

By way of illustration, there are a variety of markers that first appear on less- primitive cells ofthe hematopoietic system, and which can therefore be used to eliminate particular later-arising sub-populations to the extent that is desirable. Such later markers include, e.g., CD4, CD8, CD11, CD 18, CD 19, CD21, CD45RA and numerous others.

Markers that arise at relatively earlier points during the hematopoietic differentiation lineages can be used for eliminating relatively larger numbers of less- primitive cells. For example, the use of Lin markers such as CD3, CD10, CD33, CD38,

CD71 and/or HLA-DR could be used to eliminate all but the very primitive PHSCs (i.e. Lin- PHSCs) which do not bear such markers. The use of such markers in multiparametric sorting is known in the art; and is illustrated below (using CD38 by way of example). As noted above, it is also possible to use various dyes such as rhodamine 123 to distinguish between active and quiescent cells, thereby further enriching for particular subsets of PHSCs. By way of example, if it is desired to limit transduction to relatively primitive

PHSCs, then a population of cells such as CD34+ cells can be further enriched on the basis ofthe expression of a Lin marker such as CD38. The level of CD38 antigen expression increases together with the expression of very early markers ofthe myeloid and lymphoid lineages. Thus, cells characterized as CD34+CD38- represent a very primitive hematopoietic sub-population (which comprises about 1% of the CD34+ cells or about 0.01% of total hematopoietic cells). The isolation of such primitive PHSCs and the subsequent transduction of those PHSCs using the methods ofthe present invention are illustrated in the Examples below.

PHSCs are generally quiescent cells. However, for efficient transduction using retroviruses, it is generally necessary for the target cells to be in an actively dividing phase ofthe cell cycle. Numerous investigators have demonstrated that mammalian cells, including CD34+ cells, can be expanded in vitro using cultures supplemented with various cytokines. A large variety of such cytokines are widely available and new cytokines are regularly being characterized. Receptors for a number of cytokines are known or believed to be expressed on various hematopoietic cells including PHSCs. In addition, in vitro assays can be readily employed to assess the effects of various cytokines on cellular proliferation, differentiation and/or susceptibility to viral transduction. For the transduction of PHSCs, we presently prefer cytokines that stimulate PHSC proliferation without substantially effecting differentiation. A variety of such stimulatory cytokines are known and other cytokines can readily be tested for their ability to stimulate PHSC division. [See, e.g., N.A. Nicola (ed.), "Guidebook to Cytokines and Their Receptors" (Oxford University Press, 1994)]. Presently preferred examples of such cytokines include interleukins (such as IL-l, IL-3, IL-6 and IL-l 1); colony stimulating factors (or "CSFs") (such as G-CSF (granulocyte CSF), GM-CSF (granulocyte-macrophage CSF) and M-CSF (monocyte-macrophage CSF); and other cytokines such as "FLT3L" (also known as

"FLT3-ligand"), "SF" (also known as "Steel Factor" or "stem cell growth factor") and "LIF" (leukemia inhibitory factor). Various experimental data indicate that combinations of cytokines (referred to as cytokine "cocktails") can be particularly effective for inducing proliferation of hematopoietic cells, including PHSCs. As illustrative examples, described in detail below, we have employed a number of different cytokines for pre-stimulation of PHSCs prior to exposure ofthe PHSCs to viral vectors and cationic lipid adjuvants. For example, we have used a first cocktail of cytokines (referred to as "16SF") that includes IL-l, IL-6, SF and FLT3L. We have also used another cocktail of cytokines (referred to as "36S") that includes IL-3, IL-6 and SF. Combinations of cytokines can also be provided as fusion proteins, as is known in the art. For example, "PIXY" is a recombinant fusion of IL-3 and GM-CSF available from Immunex Coφoration, Seattle, Washington. [23] For treating human cells, we typically use human recombinant versions of cytokines (abbreviated "hr" herein, as in "hrIL-1").

Some cytokines are also capable of up-regulating the expression of viral receptors. Such cytokines can be useful for further enhancing the transduction efficiencies by making the target cells more susceptible to infection. For example, in the case of hrIL-3, it has been demonstrated that treatment of target cells results in an up-regulation of receptors for amphotropic retroviruses [30].

Normal cellular proliferation and development appears to involve shifting balances of positive and negative regulatory factors such as cytokines. As a corollary to the addition of positive stimulatory factors, it is also possible to treat cells with agents that tend to inactivate or eliminate negatively acting factors including down-regulatory cytokines. As an illustrative example, one can treat cells in the presence of antibodies directed to TGF-beta which is found in serum and is known to down-regulate the proliferation of PHSCs. For use in the present invention, target cells are generally pre-stimulated widi cytokines for a relatively short period (typically about one day) prior to exposure ofthe cells to the cationic lipid adjuvant and the viral vector. Generally, the cytokines are also included in the subsequent transduction stage.

As will be appreciated by those of skill in the art, the use of any agents such as cytokines would typically be preceded by an optimization assay in which the effects of varying levels of cytokines or other agents can be assessed and thus optimized. Preliminarily, such an assay might test varying levels of cytokines in the range of, e.g., 1- 300 ng/ml.

The methods ofthe present invention utilize combinations of viral vectors and cationic lipid adjuvants. Viral vectors can be of any ofa variety of types, including for example, RNA viral vectors such as retroviral vectors, and DNA viral vectors, such as vectors based on Epstein-Barr virus (EBV), adenovirus, adeno-associated virus (AAV) and heφes simplex virus (HSV). Retroviral vectors have been particularly preferred mediators of transduction because such vectors can generally become stably integrated into the genome ofthe recipient cell. A large variety of viral vectors, including retroviral vectors, are available and well known in the art. Suitable vectors will be those for which receptors exist on the desired target cell. The host range of viral vectors is principally determined by the particular molecules found on me outer surface ofthe viral particle. A number of viruses (often referred to generically as "amphotropic" viruses) are known to have a very wide host range of infectivity. However, it should be noted that while such viruses may infect distantly-related species, the efficiency of transducing particular target cells (such as human hematopoietic cells) may be greater when using viruses that are more specific (i.e. specific for the species being targeted or for a group of more closely related species such as the various primate species). The reasons for the greater transduction efficiencies are often not clear, but in some cases the increased efficiency may be due to receptor affinity and/or concentration on the cellular surface.

As is well known in the art, and illustrated herein, viral vectors such as retroviruses can be readily "pseudotyped" using various cell lines to produce viruses with different surfaces and thereby with varying host specificity. By way of illustration, we have infected human PHSCs with viral vectors that had been pseudotyped with either amphotropic or primate-specific viral coats (e.g. GaLV (Gibbon ape leukemia virus). As described in detail below, it is believed that human PHSCs express a greater concentration of receptors for GaLV and, consistent with those observations, we observed that the use of GaLV-pseudotyped vectors resulted in comparatively higher transduction efficiencies. The viral vectors used in the present invention comprise a "proviral" nucleic acid

(i.e. proviral RNA or DNA) which is associated with, and therefore deliverable by, the viral particle. In addition to genes of interest (which are to be transduced to target cells) 97/12052 PCI7US96/15580

the proviral nucleic acid will generally include the "packaging signal" that allows the proviral nucleic acid to be packaged in the virus particle and the long terminal repeats (LTRs) that allow the proviral nucleic acid to become effectively integrated into the target cell genome. The LTRs are positioned at eidier end ofthe proviral nucleic acid and also generally contain regulatory sequences such as promoters and/or enhancers that affect expression of genes within the proviral nucleic acid. The gene or genes of interest will also typically be operably linked to a suitable promoter which can be constitutive, cell- type specific, stage-specific and/or modulatable. Enhancers, such as those from other viruses (e.g. Friend virus and GaLV), can also be included. The specific regulatory sequences employed will depend on the particular needs ofthe user (depending, for example, on the level of expression desired, and whether cell-specific, stage-specific or modulatable expression is desired).

The viral vector can comprise any gene of interest that is to be transduced to the target cell, including, for example, marker genes and or therapeutic genes. Thus, the vector can contain a detectable marker gene that allows the transduced cells to be distinguished from other cells. Such "marked" target cells can be detected after introduction into a patient (thereby allowing monitoring ofthe spread and/or persistence of PHSCs in vivo). Selectable marker genes can also be included, as discussed below.

The viral vectors used in the present invention can also comprise (in place of or in addition to a marker gene) a dierapeutic gene that is used to alter the activity ofthe transduced target cell so that the target cell and/or its progeny have a beneficial effect on a patient receiving such cells. By way of illustration, a typical example would be a PHSC that has been transduced with a therapeutic gene that enhances the level of a beneficial protein or other agent in the PHSC and/or its progeny, or that reduces the level ofa deleterious protein or other agent in the PHSC and/or its progeny, or that provides resistance to a cytotoxic or other harmful agent.

As another basic illustrative example, it will be possible using the present invention to transduce a target cell with a gene or genes that encode secreted proteins or that encode proteins involved in the secretion of other agents from the target cell or its progeny, which secreted proteins or other agents have a beneficial effect on the recipient patient. As yet another illustrative example, it will be possible using the present invention to transduce a target cell with a gene or genes that affect the interaction between a target cell and/or its progeny and other cells in the recipient patient. By way of illustration, the therapeutic gene might render the transduced cells and/or their progeny more or less susceptible to activation by otlier cells, more of less dependent on "helper" cell interactions, more or less dependent on exogenous cytokines, more or less resistant to a chemomerapeutic agent, or more or less resistant to an infectious agent (such as a virus), or a toxic agent such as a chemotherapeutic drug), to name just a few examples.

In preferred embodiments, the viral vectors will comprise one or more selectable genes that can be used to select cells that have been transduced with the vector. By way of illustration, an antibiotic resistance gene can be used as a positive selectable marker gene that allows a host cell to be positively selected for in die presence of me corresponding antibiotic.

The retroviral vectors may also contain one or more detectable markers. A variety of such markers are known, including, by way of illustration, the bacterial beta- galactosidase (lacZ) gene; the human alkaline phosphatase ("AP") gene and genes encoding various cellular surface markers; which have been used as reporter molecules both in vitro and in vivo. Since alkaline phosphatase is expressed on the cell surface (and antibodies directed against it are available commercially), the AP gene offers several advantages including rapid and quantifiable detection of expression in the target population (such as in transduced CD34+CD38- cells) by flow cytometry; efficient and non-toxic selection of transduced target cells by fluorescence-activated cell sorting (FACS) or various immuno-based selection procedures; and convenient tracking of transduced cells and their progeny both in vitro and in vivo. For applications involving gene erapy, it may also be advantageous for the vector to comprise a "suicide" gene diat allows recipient cells to be selectively eliminated at will. A suicide gene is a type of negative selectable marker gene that causes host cells to be inhibited or eliminated in the presence of me corresponding selective agent. Such suicide genes can thereby be used to selectively eliminate me host cells should mat become necessary or desirable. One particularly preferred type of marker gene is a

"bifunctional selectable fusion gene" comprising both a positive selectable marker and a negative selectable marker fused together as a single in-frame fusion product (tiiereby ensuring that positively-selectable cells remain subject to negative selection as well); see, e.g., the publications of S.D. Lupton et al. including Mol. and Cell. Biol. 11:3374-3378 (1991); PCT Publication WO92/08796 (29 May 1992); and PCT Publication WO94/28143 (8 Dec. 1994). A large variety of such positive and negative selectable markers are known in the art and are widely available.

As described above, pluripotent hematopoietic stem cells have been significantly more difficult to target than many other cell types. Using the present invention, however, it is possible to target PHSCs with high efficiency, even with short-term culturing periods (which are believed to maximize preservation of PHSC functionality). Indeed, as illustrated in the Examples below, it is possible to achieve transduction rates of greater than 30% by employing the methods ofthe present invention.

Enhancing the efficiency of viral transduction using die present invention can also be extremely beneficial for the transduction of omer cells (including the more mature cells ofthe hematopoietic system); especially in situations in which a large number of transduced cells may be required (for transplantation to a patient for example) or in which the cells are negatively affected by long-term culturing (and, therefore, an efficient short- term transduction method is highly desirable). The memods ofthe present invention can also be applied to the viral transduction of non-hematopoietic cells.

The memods ofthe present invention involve exposing target cells to a viral vector and to a cationic lipid adjuvant. As described above, cationic lipid adjuvants contain at least one cationic lipid molecule, which essentially comprises a hydrocarbon chain (or "tail") attached to a positively-charged head group. The hydrocarbon tail may be linear or branched, may be aliphatic and/or aromatic in structure, and may be saturated or unsaturated. The head group may possess a single net positive charge (monovalently cationic) or multiple net positive charges (polyvalently cationic). A lipopolyamine is a type of cationic lipid tiiat has, as part of its head group, a polyamine or analog thereof. Polyamines are compounds containing at least two amino groups; including, for example, compounds such as spermine, spermidine and putrescine. In the case of spermine, for example (which is the polyamine incoφorated into the "lipofectamine" reagent described below), the spermine moiety consists of an aliphatic hydrocarbon having charged amino groups incoφorated in its backbone (see, e.g., GIBCO/BRL publication FOCUS, 15:73, 1993, and Behr, Bioconj. Chem., 5:382, 1994). A large variety of such cationic lipids have been described in the art and are readily synthesized and/or are widely available from commercial sources. (See, e.g.., Rose et al., Biotechniques 10:520, 1991; Feigner et al., J. Biol Chem 269:2550, 1994; Remy et al., Bioconj. Chem., 5:647, 1994; reviewed in Behr, Bioconj. Chem., 5:382, 1994; and the various commercial sources of cationic lipids and lipid combinations).

A cationic lipid adjuvant may comprise, for example, a mixture of different cationic lipids. A cationic lipid adjuvant may also comprise one or more neutral lipids. A variety of such neutral lipids have been described and are widely available. A variety of prepared combinations of lipids are also widely available (see, e.g., the DOTMA/DOPE formulations described by Feigner et al. (PNAS 84:7413, 1987), sold as "lipofectin" by

GIBCO/BRL; various modified lipopolyamines, sold as "lipofectamine" by GIBCO/BRL; and the "Transfectam" reagents sold by Promega. Cationic lipid adjuvants may also comprise other, non-lipid molecules.

As described and illustrated herein, high efficiency transduction using a viral vector can be accomplished by exposing a hematopoietic cell (preferably a PHSC pre- stimulated with one or more cytokines) to both the viral vector and at least one cationic lipid adjuvant. Generally, the cells are exposed to both the lipid adjuvant and the viral vector at approximately the same time (and they may of course be applied in combination). It is also possible to pre-incubate the virus with the lipid adjuvant (followed by introduction to the target cells), and/or to pre-incubate the cells with the lipid adjuvant (followed by exposure to the viral vector).

As an example ofthe use ofthe present invention, we have illustrated the high efficiency genetic transduction of PHSCs using a commercially-available cationic lipid mixture known as "lipofectamine." Lipofectamine, available from GIBCO/BRL for example, comprises a lipopolyamine (DOSPA) as the cationic lipid, and also comprises a neutral lipid (DOPE), as described above.

Rapid testing of various cationic lipid combinations and/or any ofthe other components ofthe methods described herein can be easily accomplished using procedures such as those illustrated in the Examples below. For example, the effect of a particular agent on the efficiency of transduction can be readily assessed by using transducing cells

(e.g. PHSCs or "PHSC-surrogates" as described below) with a viral vector carrying a selectable marker (such as an antibiotic resistance gene) or an easily detectable marker (such as a protein that is readily detectable on cell surfaces). Examples of both sorts of assays are provided below; using the neo gene and the AP gene, respectively. A variety of such techniques and associated markers are well known in the art. The use of readily screenable selectable/detectable markers to quantify transduction efficiencies can be employed to rapidly test and optimize any ofthe variations described herein or obvious to those of skill in the art. Such variations and routine optimization might include, for example, the use of various pre-stimulatory cytokine cocktails, the use of various cationic lipid combinations, the use of different viral vectors, the optimization of culture conditions, and the timing of pre-stimulation and exposure of cells to vector and lipids. As will be clear to those of skill in the art, other factors besides transduction frequency may influence the choice of particular conditions, depending on the particular use to which the transduced cells will be put. For PHSCs, it will generally be preferable for the cells to retain as much normal functionality as possible (subject of course to any desired modifications associated with the transduction). In such cases, it is generally preferable to minimize the necessary time in culture since it is believed that long passages in vitro can lead to loss or alteration of PHSC functionalities. There are a variety of tests that can be conducted to ensure that the transduced cells retain such functionalities, as described herein.

Cells transduced according to the methods ofthe present invention can be used in vitro for any of a variety of situations in which it is desirable to have genetically- modified cells. Cells transduced according to the methods ofthe present invention can also be used for administration to patients. The amount of cells administered will generally be in the range present in normal individuals. Typically, administrations would be between about lxl 0⁴ cells/kg and lxl 0⁸ cells/kg. However, since different individuals are expected to vary in responsiveness, the type and amount of cells infused, as well as the number of infusions and the time range over which multiple infusions are given are determined by the attending physician on the basis of routine examination criteria. Generally, it would be expected that smaller doses may be administered initially, with potentially increasing doses over time. The examples presented below are provided as a further guide to the practitioner of ordinary skill in this art, and are not to be construed as limiting the invention in any way. EXAMPLES Example 1 Obtaining Human Pluripotent Hematopoietic Stem Cells

Human bone marrow (HBM) was obtained from normal volunteers. Mononuclear cells (MNCs) were obtained by Ficoll-Hypaque gradient separation, following standard protocols. [22]

CD34+ cells were obtained from mononuclear cells using an avidin column to separate cells stained with biotin-labeled anti-CD34 antibodies (we used the "Ceprate- LC" kit available from CellPro, Seattle, Washington, according to the manufacturer's directions). In brief, MNCs were stained with biotin-labeled anti-CD34 antibodies for 20 min., then washed and run over an avidin column. The column flow-through contained CD34- cells. The CD34+ cells were then harvested by gentle squeezing ofthe column, (note: in experimental protocols, minute is typically abbreviated "min.", hour is typically abbreviated "h" and day is typically abbreviated "d"). The percentages of total CD34+, CD34+CD38+ and CD34+CD38- cells were quantified using flow cytometry. Briefly, aliquots ofthe CD34+ (column-enriched) cells were stained by standard methods using mouse anti-(human CD34) antibodies labeled with either FITC (fluoro-isothiocyanate) or PE (phycoerythrin), or mouse anti-(human CD38) antibodies labeled with PE. All antibodies were obtained from Becton-Dickinson; and stained cells were analyzed on a Becton-Dickinson FACSCAN. In some experiments, cells were stained with anti-(CD34FITC), anti-(CD38PE) and/or anti- (CD90Cychrome) (to monitor Thyl expression).

To quantify multipotential and committed progenitors in the CD34+ population, cells were plated (at about 600-1200 cells/ml) in a standard colony assay using semi-solid methylcellulose supplemented with cytokines. We generally used "PIXY" (a combination of IL-3 and GM-CSF provided as a recombinant fusion protein), Steel Factor ("SF") and erythropoietin ("EPO") at standard concentrations of lOng/ml, lOOng/ml and 2.5U/ml, respectively. All cytokines for human CD34+ cells were obtained from Immunex Coφoration, Seattle, Washington. (As discussed above, other cytokines can also be used, at predetermined optimal concentrations.)

Six separate experiments were performed and in each assay at least 3 wells were plated. Colonies were counted 14 days after plating. The range of CFU-GM (granulocyte-macrophage) progenitors observed was 193-1437, for BFU-E (erythrocyte) it was 60-830 and for CFU-GEMM (granulocyte-erythrocyte-macrophage- megakaryocyte) it was 19-257 (all per 5000 cells plated in PIXY+SF+EPO).

A comparison of progenitors formed in PIXY+FLT3L+EPO was also performed. FLT3L (abbreviated "F") is considered to be a cytokine that preferentially stimulates very primitive cells among PHSCs. As expected, a lower number of progenitors was observed when SF was replaced by FLT3L.

Example 2 Obtaining Canine Pluripotent Hematopoietic Stem Cells

Bone marrow aspirates were obtained from normal dogs maintained at the Fred Hutchinson Cancer Research Center in Seattle, Washington. Mononuclear cells (MNCs) were obtained from canine bone marrow (CBM) by Ficoll-Hypaque gradient separation, as for human bone marrow. No CD34-based enrichment of canine MNCs was performed as antibodies to canine CD34 protein are not presently available. However, other enrichment procedures can be performed, as described below. The MNCs were quantified for the number of progenitors by plating in a standard colony assay, essentially as described above. The number of CFU-GM progenitors obtained was about 195 per 5000 cells plated. Typically 10,000 cells/ml were plated and the cytokines used were hrPIXY, canine rG-CSF, canine rSF and hEPO (at 25ng/ml for PIXY, G-CSF and SF, and 2.5U/ml for EPO). (As with human CD34+ cells, those of skill in the art will recognize that other cytokines can be used, at predetermined optimal concentrations.) For stimulating PHSCs from a particular mammal, it is generally preferable to use cytokines derived from the same species since the activity of cytokines derived from other species may be sub- optimal.

For further enrichment for PHSCs, normal canines can be treated with the drug "Cytoxan" at about 30mg/Kg by intravenous infusion approximately seven days prior to bone marrow harvest. Cytoxan is a cell-cycle-specific drug that selectively eliminates actively dividing cells. Since PHSCs are usually quiescent, they are relatively resistant to cytoxan which can thus be used to enrich for PHSCs (even in the absence of specific antibodies to cell surface markers such as CD34 and CD38). Other cell-cycle-specific drugs (including many chemotherapeutic agents and or anti-viral agents) can be similarly utilized, in vivo or in vitro, including for example, 5-fluorouridine (5-FU) which has been frequently used to inhibit actively-dividing mammalian cells (including murine, canine and human cells).

Example 3

Preparing a Viral Vector

As an initial illustration ofthe present invention, we used the "LAPSN" retroviral vector that has been described in detail. [25] Essentially, the LAPSN vector contains a human placental alkaline phosphatase ("hPLAP") reporter gene (for use as a detectable marker) under the transcriptional control of the MoMLV-LTR. The vector also contains a neo dominant selectable marker gene which is transcribed from an internal SV40 promoter.

The nature ofthe viral envelope is determined by the type of cell line that is used to produce the virus. We have produced both amphotropic and GaLV-pseudotyped LAPSN viral vectors using PA317/LAPSNC1 and PG13/LAPSNC9 producer lines, respectively, as described. [26-29] As is well known in the art, any of a variety of selectable genes, detectable genes and/or therapeutic genes can be cloned into such viral vectors using standard molecular biological techniques.

Viral supernatant was prepared in "IMDM" medium (from GIBCO) supplemented with 10% FCS (fetal calf serum) and 2 mM giutamine. The medium used to collect viral supernatant did not contain any antibiotics such as penicillin or streptomycin. To make retroviral supernatant the following method was used. The producer cell line (such as PA317/LAPSN) was seeded in flasks at a density in the range about 1 to 8xl0⁴ cells/cm² (generally about 4xl0⁴ cells/cm²) in "DMEM" medium (GIBCO) supplemented with 10% FCS. After incubation at 37 degrees Celsius for about 24-28 hours, medium was removed from the flasks and fresh IMDM supplemented with 10% FCS^'and 2 mM giutamine was added at about 0J3 mls/cm . After about 16 hours of incubation, viral supernatants were harvested and filtered through a 0.45 micron filter. Viral preparations were titered on HeLa cells according to standard methods. Aliquots of viral supernatants were routinely stored at minus 70°C

It is generally preferable to minimize the amounts of fetal calf serum used in the preparation of viral supernatants (and during PHSC pre-stimulation and infection according to the present invention) because FCS is believed to contain components that tend to sequester the cationic lipids used herein, thereby reducing their effective concentration. For these experiments, we reduced the level of FCS from about 25% to about 10% (although it should be possible to reduce the serum level even further or to eliminate it entirely).

Example 4

Optimizing Components for Transduction

Stable gene transfer into PHSCs using retroviral vectors was optimized by ensuring that the target cell population was induced to proliferate and that the concentration of retroviral vectors was optimal. Target cell proliferation was preferably induced using cytokine cocktails, as described below. Differing amounts of viral supernatant were applied to cells in order to optimize that component ofthe transduction. To determine the optimal concentration of retroviral vector, viral supernatant at a titer of about 1x10 CFU/ml (as tested on HeLa cells) was used at increasing volumes starting with 50 microliters up to 1 ml. The final volume was in all cases no more than 1 ml and for lower volumes of viral supernatant the volume was made up with regular medium. The viral supernatant was made in IMDM supplemented with 10% fetal calf serum and giutamine (2 mM), without antibiotics. A linear relationship was observed between the percentage of transduced cells obtained (as measured by AP expression and

PCR analysis) and the amount of retroviral supernatant used. Under our transduction conditions and the current retroviral titer, it was determined that optimal transduction was obtained when about 1.0ml of viral supernatant was used for every l-5xl0⁴ human CD34+ cells. (Higher volumes of viral supernatant were not preferred given the relatively large volume being applied to the cells). Optimization of cytokine cocktails and other potential components of transduction are described below. Example 5

Pre-stimulation of PHSCs with Cvtokine Cocktails

Human CD34+ cells, prepared as described above, were plated at l-2xl0⁴ cells/ml in IMDM supplemented with 25% FCS, 2 mM giutamine and various cytokines, and then incubated for a period of about 18-24 hours. For puφoses of illustration, two different "cocktails" of cytokines were used for pre-stimulation of PHSCs. A first cytokine cocktail (called "16SF") comprised hrlL- lbeta, hrIL-6, hrSF and hrFLT3L. A second cytokine cocktail (called "36S") comprised hrIL-3, hrIL-6 and hrSF. (Other combinations of cytokines may also be used, as discussed above).

Useful cytokine concentrations typically range from about 1 ng/ml up to several hundred ng/ml. Typically, we used hrIL-3 in the range of about 5-25ng/ml, hrlL-lbeta and hrIL-6 in the range of about 25-50ng/ml, and hrSF and hrFLT3L in the range of about 50-1 OOng/ml. Our preferred concentrations were based on the number of cells that traversed through S phase during the pre-stimulation and infection periods. The percentage of cells traversing through S phase was determined by measuring BrDU (bromo-deoxyuridine) incoφoration. Briefly l-2xl0⁴ CD34+ cells/ml were incubated for 18-24 h with BrDU (10-50 micromolar, preferably about 30 micromolar) in the presence of various cytokines. At the end ofthe 18-24h incubation period, cells were removed and cytospun on glass coverslips. Cells that had incoφorated BrDU were detected using a mouse anti-BrDU antibody labeled with FITC. These antibodies were purchased from Becton Dickinson and the staining procedure used was as provided by the manufacturer. Cells were visualized under a fluorescence microscope and cells were considered to have traversed through S phase if they were labeled with anti-BrdU. The percentage of cells traversing S phase were as follows: in the 16SF cocktail, the average number was about 34% (with the range being about 17-76%); and in the 36S cocktail the average number was about 36% (with the range being about 30-50%).

The data indicate that different cytokine combinations can be used to stimulate proliferation of human CD34+ cells. The preference for particular cytokine cocktails would generally be influenced by the relative ability ofthe cocktail to induce proliferation without inducing substantial differentiation (thereby maximizing PHSC functionality), and possibly also to up-regulate receptors for the viral vector used (since a number of cytokines can apparently up-regulate such cell surface receptors).

To monitor cell expansion after treatment with various cytokines, cells were plated at varying cell densities and the increase in the total cell number and the CD34+ cell population was measured at day 4 after plating. Under the conditions used, cells maintained at a density of about 1 -5x10 cells underwent a total cell expansion of 6-fold with a 2-fold increase in CD34+ cell population. Increasing or decreasing the concentrations of cells per ml did not result in any greater expansion of CD34+ cells. Since proliferation of cells is a prerequisite for successful vector integration, cells were generally plated at densities of about l-5xl0⁴. For expansion studies, we used IMDM supplemented with 25% FCS, 2 mM giutamine, and various cytokines as described.

(Those of skill will recognize that other media can be used so long as they support growth of hematopoietic cells.) Both the 16SF and 36S cytokine cocktails were used in various experiments described below.

For transduction, the human CD34+ cells were incubated continuously with the various cytokine cocktails during both the pre-stimulation and infection periods. Each of these periods was approximately 18-24h in duration, with the CD34+ cells being in culture for a maximum of about two days. At the end of these culture periods, aliquots of the infected cells were tested for transduction by monitoring the acquisition of selectable and/or detectable markers, as described below. Additional aliqouts ofthe transduced cells can be frozen in liquid nitrogen and then transplanted into patients after fulfilling release criteria.

Example 6

Optimizing Cationic Lipid Adjuvant and Transduction of Human PHSCs Using an Amphotropic Viral Vector

Human CD34+ cells derived from bone marrow were prepared as described above. Cells were pre-stimulated for 18-24h in cytokine cocktail (16SF) and then transduced with 1.0ml of ampho-pseudotyped LAPSN vector for every 1x10 cells. The range of cell concentration used was 1 - 10x10⁴ cells and thus the range of viral supernatant used was 1 to lOmls. The titer ofthe viral supernatant was about 1x10^s

CFU/ml as measured by G 148 resistant colonies on HeLa cells.

For transduction of human PHSCs according to the present invention, viral supernatants were combined with a cationic lipid adjuvant. For puφoses of illustration, we selected a cationic lipid combination "lipofectamine" (abbreviated "LPF") which is available commercially from GIBCO/BRL.

To preliminarily optimize the concentration ofthe cationic lipid adjuvant (in this case lipofectamine), it is convenient to use a "PHSC-surrogate" cell line which can be readily obtained by isolating spontaneously immortalized variants of PHSCs. In this case, we used a surrogate cell line (KMT2) that was derived by spontaneous immortalization of CD34+CD38- cells derived from cord blood (the KMT2 line had been identified by Dr. T. Suda in Japan, and is available from the Japanese cell line depository). The KMT2 cell line is phenotypically CD34+CD38- and was maintained in medium supplemented with

PIXY at a preferred concentration of about 5ng/ml. Based on preliminary transduction results with KMT2 (using the Ampho/LAPSN vector and lipofectamine (LPF) at concentrations ranging from 0 to 80 micrograms/ml), we determined that the optimal transduction (as measured by expression of AP) was achieved with about 10-40 micrograms/ml LPF, more preferably with about 20-30 micrograms/ml LPF.

Lipofectamine at higher concentrations was found to be somewhat toxic to cells. We typically add LPF to the viral supernatant before introducing the combination ofthe lipid and virus to the cells. However, the cells might also (or alternatively) be pre-treated with LPF. For puφoses of comparison with our cationic lipid adjuvant, we conducted side- by-side experiments using polybrene or protamine sulfate at concentrations of about 4 to 8 micrograms/ml. Higher concentrations of polybrene were found to be toxic to the cells.

Our preferred transduction protocol was as follows. First, the viral supernatant was thawed. Then a cationic lipid adjuvant ("CL") (such as lipofectamine at about 30 micrograms/ml) was added to the viral supernatant. We next added a cytokine cocktail, as described above for pre-stimulation. Generally we used the same cytokine cocktail that was used during the pre-stimulation protocol. Next, we placed 1-2 x10 cells (in 50 microliters) in a 9 to 25cm² container (either flasks or plates). We then added one ml of the viral supernatant that had been supplemented with cationic lipid adjuvant and cytokines. The cells were usually at a final density of about 1 -2x10 cells per ml.

Typically the total number of cells transduced was about 5-10x10 cells in a final volume of 5-10mls of viral supernatant supplemented with adjuvants and cytokines. Preferably, the cell concentration should be less than about 2xl0⁴ cells/ml and they should be at a density of about 1 cell/cm² (range of OJ - 1.5 cells/cm²). The cells were then transferred to a 37°C incubator at 5% CO₂. (The temperature may range from 32-37°C, and CO₂ from 4 to 10%, depending on the medium of choice.) The time of infection generally ranges from about 16-26h; typically we use infection periods of about 18-24h. If the cell numbers that need to be transduced are greater than about 1-1 OxlO⁶, then cells can be transduced twice during the 24h infection period. By way of example, one can incubate 2xl0⁶ cells in 50mls of viral supernatant supplemented with adjuvant and cytokines at the appropriate concentrations; then, at the end ofa 4-8h incubation (preferably about 6h), another 50mls of viral supernatant supplemented with cationic lipid adjuvant and cytokines can be added. Thus, at this point the cell concentration would be about 2x10 cells/ml of viral supernatant.

At the end ofthe infection period, cells were washed three times with IMDM supplemented with 10-25% FCS. If at this point the experiment involves transplantation into an animal or human patient, the cells can be resuspended in endotoxin-free PBS (total of about 20-100mls, preferably about 60mls). The percentage of CD34+ cells still in culture at this point should be within about 80-100% ofthe original percentage.

If samples are being analyzed for frequency of transduction, cells are replated in IMDM supplemented with 25% FCS and cytokine cocktails as used earlier. Cells are then incubated for an additional period of about 1 -2 days (preferably about two days) and then analyzed for expression of markers (e.g. AP or neomycin resistance by standard histochemical assays and or G418 resistant colonies in a CFU-GM assay). For G418- resistant CFU-GM assays, our preferred cytokine cocktail was PIXY+SF+EPO (as described above) and the G418 concentration was in the range of about 800-2000 micrograms/ml of active G418. We typically use 1200 micrograms/ml of active G418.

For assay puφoses, we also assessed the relationship between the percentage of transduced cells as determined by AP expression as compared with G418-resistant CFU- GM in a side-by-side experiment. There was a very close correlation between the two and we therefore preferred to use AP expression since it was more convenient. AP data were also corifirmed using standard PCR analysis using primers for the neomycin gene.

The results from several experiments are summarized in Table 1. For those experiments, enriched CD34+ cells from human bone marrow were pre-stimulated in cytokine cocktail 16SF for 18-24h and then infected with PA317/LAPSN viral supernatant in the presence of 16SF and the indicated adjuvant. Lipofectamine was used as the cationic lipid adjuvant at 30 micrograms/ml; and, for comparison, polybrene ("PB") was used at a concentration of 4 micrograms/ml. After incubation, CFU-GM assays were performed according to standard protocols in PIXY+SF+EPO. G418 was used at 1200 micrograms/ml (at this concentration there were no detectable G418-resistant CFU-GM obtained from cells exposed to supernatant from control PG13 cells not making retroviral vector). The percentages of transduced cells (by AP expression or G418-resistance in a CFU-GM assay) were determined at approximately three days (72h) after infection. Three experiments were run using polybrene and five experiments were run using the cationic lipid adjuvant LPF. Ofthe three polybrene experiments, only one yielded any G418-resistant colonies in CFU-GM assays. In contrast, when transduction was carried using a combination of a viral vector and a cationic lipid adjuvant (LPF), G418-resistant colonies were obtained from 100% of experimental samples. The overall results and transduction frequencies are illustrated in Table 1.

These results indicate that genetic transduction of PHSCs according to the present invention (using a combination of a viral vector and a cationic lipid adjuvant) results in greatly enhanced transduction efficiencies, with very short incubation periods, and with transduced cells being obtained from all samples treated. Cells infected without an adjuvant exhibited no transduction. While one set of polybrene-treated samples exhibited some transduction, the results indicate that the use ofa cationic lipid adjuvant was not only much more effective at transduction but also yielded a far greater number of transduced cells (even as measured at three days). The preferential proliferation of cells treated with a cationic lipid adjuvant as compared to polybrene was even more striking when examined over a longer term, as described in detail below.

Example 7

Transduction of Human PHSCs ffrom Bone Marrow and from Mobilized Peripheral Bloods Using a Second Viral Vector fGaLV-pseudotvped LAPSN in Combination with Various Cationic Lipid Adjuyants

(a) Transduction of Human PHSCs from Bone Marrow Using GaLV-pseudotvped LAPSN in Combination with LPF as a Cationic Lipid Adjuvant

Cell surface receptors for the GaLV retrovirus are believed to be present on bone maπow cells at a higher level than receptors for amphotropic viruses. [24] We therefore assessed the frequency of infection of human CD34+ HBM cells using a GaLV- pseudotyped LAPSN viral vector produced in the PG13 cell line. All experimental protocols were essentially similar to those described above using the amphotropic LAPSN viral supernatant produced in the PA317 cell line. In these experiments, transduction of human CD34+ cells was compared using two different cytokine cocktails, 16SF and 36S as described above. Again, transduction according to the present invention (using LPF as a cationic lipid adjuvant) was compared to transduction using polybrene. Experimental procedures were essentially as described above for transduction with the amphotropic viral vector. The results of transductions using GaLV-pseudotyped vectors are shown in T_ablg_2. As with the amphotropic vector, there was a significant enhancement in the efficiency of transduction of PHSCs when the viral vector was combined with a cationic lipid adjuvant. And, again, the difference was even more striking when the total number of transduced cells was compared with the total number obtained after polybrene treatment.

fb) Transduction of Human PHSCs from Bone Marrow Using GaLV-pseudotvped

LAPSN in Combination with Various Cationic Lipid Adjuvants

The ability to enhance the efficiency of transducing PHSCs using viral vectors in conjunction with cationic lipid adjuvants was further confirmed using other cationic lipids. Thus, in follow-up examples, conducted essentially as described above, we performed transductions in which "DOSPA/DOPE" (LPF as described above) was compared with various other cationic lipid compositions, including the following: "TM- TPS/DOPE", "DMRIE/Chol" (also known as "DMRIE/C"), "DOTMA/DOPE", "DMRIE/DOPE" and "DDAB/DOPE", as well as spermine, spermidine and "polyethylemine 800K". ["DOSPA/DOPE" (also refeπed to herein as "LPF" and described above) is a 3:1 molar ratio (w/w) of DOSPA (2,3-dioleoyloxy-N-[2(spermine carboxamido)ethyl]-N,N-dimethyl-l-propanaminium trifluoroacetate) and DOPE (dioleoyl phosphatidylethanolamine), see, e.g., Hawley-Nelson, P. et al. (1993) Focus

15:7). "TM-TPS/DOPE" (also known as "CellFECTTN") is a 1J.5 molar ratio (w/w) of TM-TPS (N.N'y.N¹"- tet^amethyl-N,N^,,N^,I, N^ιπ-tetrapalmitylspermine) and DOPE. TM-TPS has been described by, e.g., Luckow, V.A. et al. (1993) J Virol. 67:4566. "DMRIE/Chol" (also known as "DMRIE/C") is a 1 : 1 molar ratio (w/w) of DMRIE (dimyristyloxypropyl-3-dimethy-hydroxyethyl ammonium, see, e.g., Ciccarone, V. et al.

(1995) Focus 17:84) and Cholesterol. "DOTMA DOPE" (also known as "Lipofectin") is a 1:1 molar ratio (w/w) of DOTMA (N-[l-(2,3-dioleoyloxy)propyl]-N,N,N- trimethylammonium chloride) and DOPE; see, e.g., Feigner, P.L. et al. (1987) Proc. Natl. Acad. Sci, 84:7413. "DDAB/DOPE" (also known as "LipofectACE") is a 1:2.5 molar ratio (w/w) of DDAB (dimethyl dioctadecylammonium bromide) and DOPE; see, e.g.,

Whitt, M.A. et al. (1991) Focus 13:8. "DMRIE/DOPE" is a 1:1 molar ratio of DMRIE and DOPE; see, e.g., Feigner, J. et al. (1994) J. Biol. Chem 269:2550-2561, and Harrison, G. S. et al. (1995) BioTechniques 19: 816-823.]

As described above, DOSPA/DOPE (LPF) was generally used at a concentration of about 20-30 micrograms/ml (for the following experiments we used a concentration of

30 micrograms/ml). Other cationic lipids tested were used at the following concentrations: TM-TPS/DOPE (10 micrograms/ml), DMRIE/Chol (30 micrograms/ml), DOTMA/DOPE (10 micrograms/ml), DDAB/DOPE (30 micrograms/ml), DMRIE/DOPE (10 micrograms/ml), spermine (0.03-3 micrograms/ml), spermidine (0.3 -3 micrograms/ml), Polyethylemine 800K (0.1-30 pM, from Fluka, see Boussif, O. et al.

(1995) Proc. Natl Acad Sci 92: 7297-7301). Comparisons were made to protamine sulfate (8 micrograms/ml) and polybrene (4 micrograms/ml).

Generally, viral supernatants (from GaLV-pseudotyped LAPSN) were incubated for 5 to 10 minutes at room temperature with either cationic lipid adjuvants or polycations and were then added to CD34+ cells, using IMDM medium supplemented with 10% FCS, essentially as described above. Following infection, cells were washed three times with IMDM medium supplemented with 25% FCS and resuspended at a density of about 5 to 10 x IO⁴ cells/mL in IMDM medium supplemented with 25% FCS and cytokines (using the "36S" cocktail as described above).

The results, quantified as described above, indicated that transduction occuπed in all cases. Ofthe various cationic lipid adjuvants tested, however, DOSPA/DOPE, TM- TPS/DOPE and DMRIE/Chol were especially preferred by virtue of their very high transduction efficiencies: DOSPA/DOPE (24.5±6.2%, n=7); TM-TPS/DOPE (19.5±5.0%, n=5); and DMRIE/Chol (19.2±4.7%, n=3).

Since cholesterol and derivatives thereof can function effectively as co-lipids (see, e.g., Gao, X. and L. Huang (1995) Gene Therapy 2:710-722, and Epand, R.M. et al. WO93/05162), and cholesterol in combination with DMRIE was shown to be highly effective, it is believed that cholesterol and/or cholesterol derivatives can be usefully employed in combination with other lipids (such as DOSPA, TM-TPS, DOTMA and DDAB) in the cationic lipid adjuvants ofthe present invention. See also the recent reviews of various cationic lipids by Ledley, F. D. (1995) Human Gene Therapy 6: 1129- 1144, and Balasubramanian, R.P. et al. (1996) Gene Therapy 3: 163-172.

As noted above, some authors have previously reported transductions of CD34+ cells in cultures employing maπow stromal cells and/or fibronectin during transduction (see, e.g., Moore, et al. (1992) Blood 79:1393-1399; Chertkov, et al. (1993) Stem Cells 11 :218-227; Moritz, et al. (1994) J. Clin. Invest. 93: 1451-1457; and Xu, et al. (1995) Blood 86: 141-146).

In contrast, we have observed that the enhanced transduction efficiencies obtainable using the methods ofthe present invention are not dependent on the presence of maπow stromal cells; and, indeed, the enhanced transduction frequencies obtainable using the methods described herein were not significantly affected by the inclusion of such maπow stromal cells in the transduction cultures.

fc) High efficiency transductions of human PHSCs from mobilized peripheral blood The ability to enhance the efficiency of transducing human PHSCs using viral vectors in conjunction with cationic lipid adjuvants was further confirmed by applying the methods described above to the transduction of CD34+ cells that had been mobilized in human peripheral blood. As described above in the Detailed Description of Prefeπed Embodiments and in the cited art, PHSCs can be obtained from a number of different sources, including bone maπow and cord blood, as well as from peripheral blood from mammals that have been mobilized with, e.g., G-CSF, GM-CSF and/or other cytokines (see Ref. 4, by Orlic et al., as cited above). Peripheral blood mobilized CD34+ cells are an attractive alternative target population to bone-maπow-derived CD34+ cells for retroviral-mediated gene transduction due to ease of procurement and possibly also to their relative activation state (see, e.g., Bodine, D.M. et al. (1994) Blood 84: 1482-1491; Kiem, H.-P. et al. (1994) Blood 83: 1467-1473; Bodine, D.M. (1995) Experimental Hematology 23: 293-295; Donahue, R. et al. ( 1996) Blood 87: 1644- 1653)). Procedures for transduction were essentially performed as described above, except that mobilized peripheral blood PHSCs were used as target cells. Following standard procedures, mobilized peripheral blood cells were obtained after informed consent from cancer patients undergoing autologous cell transplant. Patients had been previously treated with standard chemotherapy drugs and lOμg/Kg granulocyte colony-stimulating factor for 7 days. Leukapheresis was performed on days 5, 6 and 7 ofthe mobilization regime. For all experiments, leukapheresis product from mobilized peripheral blood was further enriched for CD34+ cells using the Ceprate-LC Kit (CellPro, Seattle, WA), as described above. The avidin column enriched the CD34+ cell content to 69.3±10.5%, as determined by staining with mouse anti-(human CD34PE) antibody and isotype-matched controls (both from Becton

Dickinson). The human peripheral blood CD34+ cells were transduced during a 24-hour period with viral supernatants and either DOSPA/DOPE ("LPF" at 30 micrograms/ml) or protamine sulfate ("PS" at 8 micrograms/ml), as described above.

The results obtained provided further confirmation that the efficiency of transducing human PHSCs can be substantially enhanced using viral vectors in conjunction with cationic lipid adjuvants as described herein. In particular, as with the human bone-maπow-derived CD34+ cells described above, much higher transduction frequencies were obtained when peripheral blood CD34+ cells were transduced with GaLV/LAPSN and DOSPA/DOPE ("LPF") (22.9±2.8%), as compared to ampho/LAPSN and protamine sulfate ( 1.4± 1.3%). Table 2

Transduction frequency of Human CD34+ cells using GaLV/LAPSN vector

Adjuvant number of Cytokine cocktail % AP expressing cells (range) total AP expressing samples cells xlO at 72h post-infection

(AP)

Polybrene 2 16SF 11.5 (10-12) 9.15 (3-15) (PB)

«o Cationic 6 16SF 22.7 ±7.8 42.2 ±25.1 Lipid (LPF)

Polybrene 2 36S 12.4 (9-15) 8.8 (7-10) (PB)

Cationic 6 36S 34.6 ±11.7 51.5 ±26.4 Lipid (LPF)

Example 8

Evaluation of Long-term Proliferation Efficiencies Following Transduction According to the Present Invention: and Summary of Transduction Experiments

Human CD34+ cells were transduced with two types of viral vectors (Ampho/LAPSN and GaLV/LAPSN) in combination with a cationic lipid adjuvant (i.e.

LPF) or polybrene, using either of two cytokine cocktails (16SF or 36S), essentially as described above.

After infection, transduced cells were further analyzed for their ability to proliferate over a two week period. Briefly, HBM CD34+ cells were pre-stimulated and then transduced, as described above, with either GaLV/LAPSN or Ampho/LAPSN in combination with the indicated adjuvants and cytokines. After transduction, cells were plated in IMDM supplemented with 25% FCS, 2 mM giutamine and the cytokine cocktail used during the pre-stimulation and infection periods. The starting number of cells plated was similar in all groups, and the titer ofthe two viruses on HeLa cells was also similar. Cells were fed weekly and, at day 15 post-infection, cells were harvested and assayed for expression ofthe transgene (AP) as described above.

For the experiment shown in Table 3. the total numbers of cells recovered in cultures supplemented with cocktail 36S and transduced with GaLV/LAPSN in the presence of either a cationic lipid (i.e. LPF) or PB was 6.75x10 and 3.0x10 , respectively. (Approximately 7.25x10⁴ total cells were recovered from cultures that had not been treated with either the viral supernatant or an adjuvant.) The lower number of recovered cells after the addition of polybrene suggests that the use of agents such as polybrene may have deleterious effects on cells that do not appear to be associated with the use ofa cationic lipid adjuvant such as lipofectamine. The results, as shown in Table 3. confirmed that high efficiency transduction can be achieved using a cationic lipid, as described herein, and the resulting transduced cells retained their capacity for long-term proliferation. The results also confirmed that higher levels of PHSC transduction could be achieved using the GaLV-pseudotyped vector as compared to the amphotropic vector. In contrast, transductions performed without the use ofa cationic lipid not only resulted in significantly lower levels of transduction but appeared to be unaffected by the type of vector used. Table 3

Enhanced long-term proliferation of cells transduced in the presence of cationic lipids

virus pseudotype Total transduced cells (AP) day 16 post-infection in the presence of various cytokines (16SF or 36S) and adjuvant (CL or PB)

16SF (PB) 16SF (CL) 36S (PB) 36S (CL)

Ampho/LAPSN 2,480 9,983 2,496 7,100

GaLV/LAPSN 2,952 17,940 2,220 17,685

Summary of Transduction Experiments

Thus from data presented in tables 1-4 the following conclusions may be reached:

1. Highly efficient transduction of PHSCs was obtained when cells were pre- stimulated with a cytokine cocktail and then infected with a viral vector in combination with a cationic lipid adjuvant.

2. There was a significant increase (2-3 fold) in the percentage of transduced cells three days post-infection when a cationic lipid adjuvant was used as compared to polybrene.

3. There was a 3-7 fold increase in total cell number 3 days post-infection when a cationic lipid adjuvant was used as compared to polybrene.

4. G418-resistant CFU-GM colonies were obtained from only 1 out 3 sample experiments when PB was used as an adjuvant. In contrast, G418-resistant CFU-GM colonies were obtained from 100% ofthe sample experiments infected in the presence of a cationic lipid adjuvant. 5. PHSCs transduced in the presence ofa cationic lipid adjuvant had a significant proliferative advantage relative to cells transduced in the presence of polybrene. Cells from only 3 out of 5 samples expanded beyond day 3 post-infection when PB was used as an adjuvant. In contrast, 100% ofthe samples infected in the presence ofa cationic lipid adjuvant expanded beyond day 3 post-infection. In a long- term proliferation assay (measured at day 16 post-transduction), there was a 4-8 fold increase in the total number of transduced cells when a cationic lipid was used as an adjuvant compared to PB. Similar data were obtained from 5 experiments using 16SF as a cytokine cocktail and Ampho/LAPSN as a viral vector.

6. There was a distinct relationship between viral pseudotype and the enhancement observed with the use ofa cationic lipid adjuvant. For cells stimulated with

36S and transduced in the presence ofa cationic lipid adjuvant, the transduction efficiency was enhanced by an additional 3.5 fold when a GaLV-pseudotyped viral vector was utilized (as compared to an amphotropic vector). Transductions performed with other agents such as polybrene resulted in significantly lower levels of transduction and appeared to be associated with deleterious effects on target cell proliferation.

7. PCR analysis was done for all samples and results confirmed the data presented in Tables 1-3. 8. In contrast with previously-described protocols for transducing PHSCs that require at least 3-5 day infection periods and also require the addition of stromal or other accessory cells or supplemental proteins such as fibronectin, we have demonstrated that neither stromal or accessory cells nor such supplemental proteins are required for the efficient transduction of PHSCs. Indeed, using the methods ofthe present invention, we were not only able to avoid the use of such additional cells or agents, but we were also able to achieve high transduction efficiencies even with very short infection periods (generally less than about 48 hours). As noted above, minimizing the amount of time that the PHSCs are present in culture is believed to be very beneficial for preserving the phenotypic functionality of PHSCs.

Example 9

Genetic transduction of human CD34+ D 8- PHSCs using the methods ofthe present invention. As described above, it is believed that PHSCs can be further enriched on the basis of CD38 antigen expression. In particular, the level of CD38 antigen increases together with the expression of early markers ofthe myeloid, lymphoid and erythroid markers. Thus, cells in the CD34+CD38- population (about 1% ofthe CD34+ cells or 0.01% ofthe total hematopoietic cells) are believed to be the most primitive. To measure gene transduction into CD34+CD38- cells, immuno-column enriched

CD34+ cells were stained with anti-(CD34FITC) and anti-(CD38PE) and sorted on a Becton-Dickinson FACS Star machine. Sorted CD34+CD38- cells were pre-stimulated with a cytokine cocktail (we tested both 16SF and 36S) for 18-24h, as described above. Approximately 2000 pre-stimulated CD34+CD38- cells were then infected with about 1 ml of viral supernatant (PG 13/L APSN or PG 13 without LAPSN) in the presence of cationic lipids (or polybrene or protamine sulfate) for an additional 16-24h as described above. Following the infection period, cells were washed and resuspended in fresh medium supplemented with cytokines and fed weekly. At day 21 post-infection, histochemical analysis for AP expression and PCR analysis for neo sequences was done on cells from each group. With the 16SF cocktail and CL, the total number of transduced cells was 18,600 and the total cell number at d21 was 60x10⁴. With the 36S cocktail and CL, the total number of transduced cells was 2,936 and the total cell number at d21 was 6.7x10⁴. Similar PCR results were obtained from a separate experiment in which CD34+CD38- cells were stimulated with the 36S cytokine cocktail and then infected with PG13/LAPSN viral supernatant in the presence ofa cationic lipid adjuvant or PB.

The results, summarized in Table 4. indicated that CD34+CD38- PHSCs can be effectively transduced using the methods of the present invention, and the transduced cells retain their proliferative capability. In contrast, CD34+CD38- cells could not be effectively transduced using compounds such as polybrene and protamine sulfate.

The LAPSN vector packaged in the LGPS cell line (which is packaged as a virus without any envelope protein) did not transduce CD34+CD38- cells even in the presence of cationic lipid adjuvants suggesting that the cationic lipids are enhancing transduction as mediated by viral receptors.

To our knowledge, the experiments described herein are the first demonstration of effective transduction of CD34+CD38- cells within short-term culture periods (in the range of 48 hours). Limiting the amount of time that the cells are in culture (using the present invention) is believed to be very beneficial for preserving the pluripotency of

PHSCs and their ability to engage in extensive self-renewal and proliferation.

Example 10

Further enhancement of transduction efficiency bv using CL treatment in combination with stimulation of viral receptors

The cell surface receptors for GaLV and the amphotropic retroviruses appear to be sodium-dependent phosphate symporters. Based on experiments in a rat cell line, depletion of extracellular phosphate is believed to result in an increase in effective receptor levels. [24]

We therefore conducted experiments in a PHSC-surrogate cell line (KMT2, as described above) which combined the use of our viral vector plus cationic lipid adjuvant along with metabolic induction of cellular receptors via phosphate deprivation.

For the experiments summarized in Table 5. we used GaLV-pseudotyped LAPSN and measured transduction in the presence of lipofectamine as a cationic lipid adjuvant (CL), as compared with transduction in the presence of either polybrene (PB, 4 micrograms/ml) or protamine sulfate (PS, 8 micrograms/ml). The results indicated that enhancement of cell surface receptors (via phosphate deprivation) resulted in an additional three-fold increase in transduction efficiency; which reached levels of greater than 30% when a cationic lipid adjuvant was used according to the present invention.

In a separate set of experiments, CD34+ cells prepared as described above were also subjected to phosphate deprivation prior to transduction according to the present invention. Consistent with the results observed with the PHSC-surrogate cell line, phosphate deprivation of CD34+ PHSCs resulted in an additional two-fold increase in the frequency of genetic transduction according to the present invention (as measured by PCR analysis for the neo marker, and by AP expression at 72-96 hours post-infection).

We have also conducted similar experiments using even more primitive PHSCs (i.e. CD34+CD38- PHSCs, as described above). As with the PHSC-surrogate line and the

CD34+ PHSCs, phosphate deprivation of these primitive CD38- PHSCs resulted in a similar enhancement ofthe level of genetic transduction using the methods ofthe present invention (as determined by PCR analysis). Table 5

Further enhancement of transduction efficiencies by subjecting cells to phosphate deprivation

Condition % transduced (AP Positive)

CL PB PS

+PO₄ 10.4 3.2 2.8 n=2

-PO₄ 30.8 6.5 3.7 n=3

Example 11(a)

Transduction of Canine PHSCs Using a Viral Vector in Combination with a Cationic

Lipid Adjuvant

Canine hematopoietic cells were obtained as described earlier. Transduction parameters were essentially similar to those described for the human CD34+ cells.

Transduction efficiencies were monitored by assaying AP activity.

For the short-term transduction assay shown in Table 6_T cells were pre-stimulated with a cytokine cocktail comprising hrIL-3, hrIL-6 and crSF (as abbreviated throughout, "hr" = human recombinant, and "cr" = canine recombinant). Cells were then transduced in the presence ofthe same cytokines using GaLV-pseudotyped LAPSN supplemented with a cationic lipid adjuvant (LPF at 30 micrograms/ml) or, for comparison, with protamine sulfate (PS, 8 micrograms/ml) or polybrene (PB, 4 micrograms/ml). Transduced cells were analyzed by histochemical detection of AP expression after both short-term (3 day) and long-term (21 day) cultures, the results of which are summarized in Tables 6 and 7. respectively.

The AP data presented in Table 6 confirmed that, as with human PHSCs, the use of cationic lipid adjuvants for viral transduction in accordance with the present invention resulted in significantly greater transduction efficiencies when measured after several days in medium. As shown in Table 7. the improvement obtained with the use ofa cationic lipid adjuvant according to the present invention was even more striking when stable, long- term gene transfer was measured. Only infection in the presence ofa cationic lipid adjuvant resulted in a measurable population of transduced cells as determined at 21 days post-infection (in contrast to infection in the presence of polybrene or protamine sulfate). These data confirmed the results obtained with human PHSCs summarized above, which indicated that high efficiency transduction can be achieved using a cationic lipid, as described herein, and that the resulting transduced cells retain their capacity for long-term proliferation. Table 6

Transduction of canine bone marrow mononuclear cells (short-term assay)

Adjuvant % transduced cells (at day 3 total transduced cells (xlO³) post-transduction) (at day 3 post-transduction)

PB 3.65 10.8

PS 3.4 8.4

CL 10.2 38.4

Table 7

Transduction of canine bone marrow mononuclear cells (long-term assay)

Adjuvant % transduced cells (at day total transduced cells (xlO³) 21 post-transduction) (at day 21 post-transduction)

PB 0 0

PS 0 0

CL 7J 95.8 Example 11 (b)

Transduction of Baboon PHSCs Using a Viral Vector in Combination with a Cationic

Lipid Adjuvant

Further confirmation ofthe applicability ofthe techniques described herein was obtained by transducing baboon hematopoietic cells in essentially the same manner as described above.

Baboon hematopoietic cells were obtained essentially as described above for human hematopoietic cells (see, Example 1). Transduction parameters were essentially similar to those described for the human CD34+ cells. Transduction efficiencies were monitored by assaying AP activity.

A short-term transduction assay was performed essentially as described above for canine hematopoietic cells (see Example 11(a)). Briefly, cells were pre-stimulated with a cytokine cocktail comprising hrIL-3, hrIL-6 and hrSF (as abbreviated throughout, "hr" = human recombinant). Cells were then transduced in the presence ofthe same cytokines using GaLV-pseudotyped LAPSN supplemented with a cationic lipid adjuvant (LPF at

30μg/ml) or, for comparison, with protamine sulfate (PS, 8 μg/ml). Transduced cells were analyzed by histochemical detection of AP expression after short-term (3 day) cultures.

The results obtained with baboon cells provided further confirmation ofthe applicability ofthe present invention. In particular, the percent transduction and total transduction observed using standard PS treatment was 1.5% and 3.4 xlO³, respectively.

In contrast, the percent transduction and total transduction observed using a cationic lipid adjuvant as described herein was 16.6% and 41.5 xlO , respectively (representing enhancements of approximately 11 -fold and 12-fold, respectively). These data provided further confirmation that, as with human PHSCs, the use of cationic lipid adjuvants for viral transduction in accordance with the present invention resulted in significantly enhanced transduction efficiencies.

Claims

1. A method of transducing a hematopoietic cell, comprising the steps of:

(a) exposing said hematopoietic cell to a cationic lipid adjuvant; and

(b) exposing said hematopoietic cell to a viral vector.

2. A method of transducing a hematopoietic cell according to claim 1, wherein steps (a) and (b) are carried out in a coincidental manner by exposing the hematopoietic cell to a combination ofthe cationic lipid adjuvant and the viral vector.

3. A method of transducing a hematopoietic cell according to claim 1, wherein the hematopoietic cell is a pluripotent hematopoietic stem cell (PHSC).

4. A method of transducing a hematopoietic cell according to claim 3, wherein the hematopoietic cell is a CD34+ hematopoietic cell.

5. A method of transducing a hematopoietic cell according to claim 3, wherein the hematopoietic cell is a Lin- hematopoietic cell.

6. A method of transducing a hematopoietic cell according to claim 3, wherein the hematopoietic cell is a human CD34+CD38- hematopoietic cell.

7. A method of transducing a hematopoietic cell according to claim 1, wherein the viral vector is a retroviral vector.

8. A method of transducing a hematopoietic cell according to claim 1, wherein the hematopoietic cell has been pre-stimulated with at least one cytokine.

9. A method of transducing a hematopoietic cell according to claim 1, wherein the hematopoietic cell has been pre-stimulated with a cytokine cocktail comprising at least two cytokines.

10. A method of transducing a hematopoietic cell according to claim 1, further comprising the step of subjecting the hematopoietic cell to metabolic induction of receptors for the viral vector prior to exposure ofthe hematopoietic cell to the viral vector and the cationic lipid adjuvant.

11. A method of transducing a hematopoietic cell according to claim 10, wherein the step of subjecting the hematopoietic cell to metabolic induction of receptors for the viral vector is conducted by depriving the hematopoietic cell of phosphate prior to exposure ofthe hematopoietic cell to the viral vector and the cationic lipid adjuvant.

12. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises a lipopolyamine.

13. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises a cationic lipid and a neutral lipid.

14. A method of transducing a hematopoietic cell according to claim 13, wherein the cationic lipid adjuvant comprises a lipopolyamine and a neutral lipid.

15. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises a lipid selected from the group consisting of DOSPA, DOTMA, DMRIE, TM-TPS and DDAB.

16. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises a lipid having a cholesteryl moiety.

17. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises cholesterol.

18. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises a lipid selected from the group consisting of DOSPA, TM-TPS and DMRIE.

19. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises a lipid composition selected from the group consisting of DOSPA/DOPE, TM-TPS/DOPE and DMRIE/Chol.

20. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises DOSPA DOPE.

21. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises TM-TPS/DOPE.

22. A method of transducing a hematopoietic cell according to claim 1, wherein the cationic lipid adjuvant comprises DMRIE/Chol.

23. A method of transducing a hematopoietic cell according to claim 8, wherein said cytokine is selected from the group consisting of interleukin-l, interleukin-3, interleukin-6, FLT3L and Steel factor.

24. A method of transducing a hematopoietic cell according to claim 9, wherein the hematopoietic cell has been pre-stimulated with at least two cytokines selected from the group consisting of interleukin-l, interleukin-3, interleukin-6, FLT3L and Steel factor.

25. A method of transducing a hematopoietic cell according to claim 24, wherein the hematopoietic cell has been pre-stimulated with a cytokine cocktail comprising interleukin-3, interleukin-6 and Steel factor.

26. A method of transducing a hematopoietic cell according to claim 24, wherein the hematopoietic cell has been pre-stimulated with a cytokine cocktail comprising interleukin-l, interleukin-6, Steel factor and FLT3L.

27. A method of transducing a hematopoietic cell according to claim 1, wherein said viral vector comprises a therapeutic gene.

28. A method of transducing a hematopoietic cell according to claim 1 , wherein said viral vector comprises a marker gene selected from the group consisting ofa detectable marker gene and a selectable marker gene.

29. A method of transducing a hematopoietic cell according to claim 1, further comprising after steps (a) and (b), the following step:

30. A method of transducing a hematopoietic cell according to claim 28, wherein the viral vector comprises a selectable marker gene and said step of selecting a hematopoietic cell that has been transduced by the viral vector is conducted by exposing the hematopoietic cell to a selective agent.

31. A method of transducing a hematopoietic cell according to claim 28, wherein the viral vector comprises a gene encoding a detectable cell surface marker and said step of selecting a hematopoietic cell that has been transduced by the viral vector is conducted by detecting expression ofthe cell surface marker on the hematopoietic cell.

32. A hematopoietic cell transduced by the method of claim 1 , and progeny thereof.

33. A method of treating a patient for a disease condition, comprising the steps of:

(1) transducing a hematopoietic cell according to the method of claim 1, wherein the viral vector comprises a therapeutic gene; and

(2) administering a transduced hematopoietic cell of step (1) to the patient.