CA2247131A1 - Modified rapid expansion methods ("modified-rem") for in vitro propagation of t lymphocytes - Google Patents

Abstract

The present invention provides a modified rapid expansion method (termed "lowPBMC-REM" or "modified-REM"), for quickly generating large numbers of T lymphocytes, including cytolytic and helper T lymphocytes, without using the large excesses of peripheral blood mononuclear cells (PBMC) or EBV-transformed lymphoblastoid cells (LCL) characteristic of high-PBMC-REM. Clonal expansions of greater than 500-fold can be achieved within a single stimulation cycle of about 8-14 days.

Description

W O 97/32970 PCTrUS97/03293 MOnTFIF,n RAPTn F,XPA~SION METHOnS ("MOnl~IF.n-RF,l\~") FOR rN VTTRO PROPAGATION OF T T YMPHOCYTFS

S FJFr T) OF THF ~VFNTION
This invention relates to improved methods for cllltnring T Iymphocytes, including human antigen-specific cytolytic and helper T lymphocytes. The methods of the present invention result in the very rapid and efficient expansion of T cells which are useful, for exal~nple, in cellular immlmot-h-era RACKGROUNr~
T lymphocytes are formed in the bone marrow, migrate to and mature in the thymus and then enter the peripheral blood and Iy~ halic circulation. T lymphocytes can be phenotypically subdivided into several distinct types of cells incll--ling: helper T cells, OU~ oSOI T cells, and cytotoxic T cells. T lymphocytes, unlike B lymphocytes, do not produce antibody molecules, but express a heterodimeric cell surface receptor that can recognize peptide fr~gm~onte of antigenic proteins that are ~ff~rhP~l to proteins of the ma}or histocomp~tihility complex (MHC) expressed on the surfaces of target cells; see, e.g., Abbas, A.K., Lichtm~n, A.H., and Pober, J.S., CelI~ r and Moler~ r Immlmology, 1991, esp. pages 15-16.
T lymphocytes that can be ~xp~n~e~7 according to the present invention are of particular interest in the context of cellular "immllnotherapy". As used herein, cellular immlmotherapy refers to any of a variety of techniques involving the introduction of cells of the immllne system, espe~i~lly T lymphocytes, into a patient to achieve a therapeutic benefit. Such techniques can inrlll~le by way of illllstration, '';~ O-~ uldli~/t;'' techniques (involving, e.g., the a~1minietration of T cells to a patient having a coll~.~"l,ised immlme system); ";~ n-~.~nh~nrin~" techniques (involving, e.g., the ~Amlinietr~fion of T cells to a patient in order to enhance the ability of that patient's immune system to avoid or combat a cancer or a pathogen such as a virus or bactt-rial pathogen), and "immuno-mo~ t;ng" techniques (involving, e.g., the ~1minietr~tion of T
cells to a patient in order to modulate the activity of other cells of the patient's immune system, such as in a patient affected by an autoimmlme condition).

W 097/32970 ~CTrUS97/03293 Cytotoxic T Iymphocytes (CTLs) are typically of the CD3+, CD4-, CD8+
phenotype and lyse cells that display frslgment~ of foreign antigens associated with class I
MHC molecules on their cell surfaces. CTLs that are CD3+, CD4~, CD8- have also been identified. Target cells for CTL recognition include normal cells expressing antigens after infection by viruses or other pathogens; and tumor cells that have undergone transformation and are expressing mutated proteins or are over-expressing normalproteins.
Most "helper" T cells are CD3+, C~4+, CD8-. Helper T cells recognize fragments of antigens preserited in association with class II MHC molecules, and primarily function to produce cytokines that amplify antigen-specific T and B cell responses and activate accessory immlm~ cells such as monocytes or macrophages. See, e.g., Abbas, A.K., et al., supra. Helper T cells can also participate in and/or ~ mPnt cytolytic activites.In addition to conventional helper T cells and cytolytic or "killer" T cells, it will also be useful to be able to rapidly expand other T cell populations. For example, T cells ~ es~ing the gamma/delta T cell receptor l cpl~ s~;~ll a relatively small portion of the human T cell population, but are suspected to play a role in reactivity to viral and ~ tçr pathogens as well as to tumor cells (see, e.g., W. Haas et al. 1993. Annu. Rev. Tmmlmol.
11:637). Another T cell population of potential clinical inl~o~ ce is the population of CD 1 -restricted T cells. CD 1 is an MHC-like molecule that shows limited polymorphism and, unlike classical MHC molecules which "present" antigenic peptides, CD molecules bind lipoglycans and appear to be illl~JU~ l~lll in the recognition of microbial antigens (see, e.g., P.A. Sieling et al. 1995. Science 269:~7, and E~.M. Beckman et al. 1994. Nature 372:691).
T lymphocytes are thus key cul.lpolle~ of the host immlme response to viruses, b~rt~ri~l pathogens and to tumors. The significance of 1JrUI)C-1Y functioning T cells is made quite clear by individuals with congenital, acquired or iatrogenic T cell immunodeficiency conditions (e.g., SCID, BMT, AIDS, etc.) which can result in the development of a wide variety of life~ infections or m~ n~ncies Persons with e~es that are related to a deficiency of immunologically-colll~,cLell~ T lymphocytes, or persons with conditions that can be improved by ~-imini~tering additional T Iymphocytes, can thus be bçnefitec~ by cellular immunotherapies, as referred to above. T cells for use in such therapies can be derived from the imml-nodeficient host, or from another source W O 9'7J3~970 PCTrUS97/03293 (preferably a compatible donor). The latter source is of course especially important in situations in which an immlmndeficient host has an insufficient number of T cells, or has T
cells that are insufficiently effective. In either case, it is difflcult to obtain sufficient numbers of T cells for effective a-1mini~tration; and thus target T cells must first be grown S to large numbers in vifro before ~-lmini~tration to a host.
~ After undergoing such cellular immunotherapy, hosts that previously exhibited, e.g., inadequate or absent responses to antigens ~ essed by pathogens or tumors, can express sufficient immllne responses to become resist_nt or immune to the pathogen or tumor.
Adoptive transfer of antigen-specific T cells to establish i ~ i Ly has been cl~m~nctrated to be an effective therapy for vir~ infections and tumors in animal models (reviewed in Cireenberg, P.D., Adv~nces in ~mmllnt~lo~v (1992)). For adoptive itmmllnotherapy to be eLre~;live, antigen-specific T cells usually need to be isolated and exrRn~e~l in numbers by in vitro culture, _nd following adoptive Llall.,r~ such cultured T cells must persist and function in vivo. For tr~fment of some hum_n ~ eR~es, the use in i~munotherapy of cloned antigen-specific T cells which r~ s~;lll the progeny of single cells, offers ~ignifir~nt advantages because the specificity and function of these cells can be rigorously defined and precise dose:response effects readily evRlllRt~l Riddell et al.
were the first to adoptively transfer human antigen-specific T cell clones to restore deficient ill~ iLy in hllm~n~ Riddell, S.R. et al., "Restoration of Viral I~ y in T1~mml0deficient Humans by the Adoptive Transfer of T Cell Clones", Scien~ ~ 257:238-240 (1992). In that study, Riddell et al. used adoptive immnnotherapy to restore deficient umly to cytomegalovirus in allogeneic bone marrow transplant recipients.
Cytom~g~lovirus-specific CD8+ cytotoxic T cell clones were isolated from three CMV
seropositive bone marrow donors, prop~g~te~l in vitro for 5 to 12 weeks to achieve numerical expansion of effector T cells, and then ~timini~t.ored i~ vt;ilously to the ei,~e~;live bone marrow transplant (BMT) recipients. The BMT recipients were deficient in CMV-specific; ~ l; Ly due to ablation of host T cell responses by the pre-transplant chemoradiotherapy and the delay in recovery of donor i-,l"~ ;Ly commonly observed after ~ 30 allogeneic bone n1~1OW transplant (Reusser et al. ~ , 78:1373-1380, 1991). Riddell et al. found that no toxicity was encountered and that the transferred T cell clones provided these imml~n~deficient hosts with rapid and persistent reconstitution of CD8+
cytomegalovirus-specific CTL responses.
Riddell et al. (J. Immlln(llof~y, 146:2795-2804, 1991) used the following procedure for isolating and culturing the CD8+ CMV-specific T cell clones: peripheral blood S mononuclear cells (PBMCs~ derived from the bone marrow donor were first cultured with autologous cytomegalovirus-infected fibroblasts to activate CMV-specif1c CTL
precursors. Cultured T cells were then restim~ fecl with CMV-infected fibroblasts and the cultures supplemente~l with ~-irradiated PBMCs. 2-5 U/ml of interleukin-2 (IL-2) in suitable culture media was added on days 2 and 4 after restimulation to promote expansion of CD8+ CTL (Riddell et al., J. Tmmlmol.~ 146:2795-2804, 1991). To isolate T cell clones, the polyclonal CD8+ CMV-specific T cells were plated at limiting dilution (0.3-0.6 cells/well) in 96-well round bottom wells with either CMV-infected fibroblasts as antigen-esr~ cells (Riddell, J. TmmllnoL~ 146:2795-2804, 1991), or anti-CD3 monoclonal antibody to mimic the stimllh-~ provided by antigen-prçsPntin~ cells. (Riddell, J. Imm.
Meth~-ds, 128:189-201, 1990). Then, y-irrs~ t~l peripheralbloodmononuclearcells (PBMC) and EBV-transformed lymphoblastoid cells (LCL) were added to the microwells as feeder cells. Wells positive for clonal T cell growth were evident in 10-14 days. The clonally derived cells were then propagated to large numbers initially in 48- or 24-well plates and subsequently in 12-well plates or 75-cm2 tissue culture flasks. T cell growth was promoted by restimlll~tion every 7-10 days with autologous CMV-infecte(l fibroblasts and ~-irradiated feeder cells consisting of PBMC and LCL, and the addition of 25-50 U/ml of IL-2 at 2 and 4 days after restim~ ti-~n.
A major problem that exists in the studies described above, and in general in the prior art of cl-ltnrin~ T cells, is the inability to grow large qll:~ntities of human antigen-2~ specific T cell clones in a timely f~hion It is not known if the slow growth of T cells in culture represents an inherent plol ~;l ly of the cell cycle time for human lymphocytes or the culture conditions used. For e~mple7 with the culture method used in the CMV adoptive imml-nntherapy study described above, three months were required to grow T cells to achieve the highest cell dose under study which was 1 x 109 T cells/cm2. This greatly limits the application of adoptive immllnotherapy for human viral tli~e~es and cancer since the disease process may progress during the long interval required to isolate and grow the specific T cells to be used in therapy. Based on extrapolation from animal model L~

W O9~13297~ PCTnUS97/~3293 studies (reviewed in Greenberg, P.D., Advances in Immunology, 1992), it is predicted that in llllm~n~ doses of antigen-specific T cells in the range of 109-10l~ cells may be required to ~llgment immllne responses for therapeutic benefit However, rapidly expanding antigen-specific human T cells in culture to achieve S such high cell numbers has proven to be a significant obstacle. Thus, with the exception of the study by Riddell et al., supra, (in which several months were taken to grow a sufficient number of cells) studies of adoptive immnnotherapy using antigen-specific T
cell clones have not been performed. The problem of producing large numbers of cells for adoptive immlm(>therapy was identified in U.S. Patent No. 5,057,423. In this patent, a method for isolating pure large granular lymphocytes and a method for the expansion and conversion of these large granular lymphocytes into lymphokine activated killer (LAK) cells is described. The me~ods are described as providing high levels of expansion, i.e.
up to 100-fold in 3 - 4 days of culture. Although LAK cells will lyse some types of tumor cells, they do not share with MHC-restricted T cells the ~,opellies of recognizing defined a~tigens and they do not provide immllnologic memory. Moreover, the methods used to expand LAK cells, which predomin~ntly rely on high conrentr~tions of IL-2 do notefficiently expand antigen-specific human T cells (Riddell et al., unpublished); and those methods can render T cells subject to programmed cell death (i.e. apoptosis) upon withdrawal of IL-2 or subsequent stim~ tion via the T cell lt;cc;~lol (see the discussion of the papers by Lenardo et al, and Boehrne et al., infra). Earlier methods that relied on the use of lectins, such as conc~l~v~lin A or phytohem~y~lulinill (see, e.g., Van de Griend et al., Tr~n~l~nt~tion 38: 401-406 (1984), and Van de Griend et al., J. Immlmnl. Methods 66: 285-298 (1984)), are even less s~ti~fa-~tory because the use of such non-specific stimllltc-ry lectins tends to induce a number of phenotypic changes in the stimlll~ted cells that make them quite di~,- .ll from T cells stimlll~t~fl via the CD3 lec~l ~r.
The inability to culture antigen-specific T cell clones to large numbers has in part been responsible for limiting adoptive immlln-~therapy studies for human ~liee~es such as cancer (Rosenberg, New F.n~l J. Med.. 316: 1310-1321, 1986; Rosenberg, New Fr~l J.
Med., 319:1676-1680, 1988) andHIV infection(Ho M. etal.,E~ 81:2093-2101, 1993) ~ 30 to the evaluation of activated polyclonal lymphocyte populations with poorly defined antigen specificities. In such studies, polyclonal popul~tio~ of Iymphocytes are either isolated from the blood or the tumor filtrate and cultured in high concentrations of the W O 97t32970 PCTrUS97/03293 T cell growth factor IL-2. In general, these cells have exhibited little if any MHC-restricted specificity for the pathogen or tumor and in the minority of patients that have experienced therapeutic benefit, it has been difficult to discern the effector merh~ni.~m involved. Typically, adoptive immunotherapy studies with non-specificS effector lymphocytes have a~mini~tered a~ oxilllately 2 x 101~ to 2 x 10~ cells to the patient. (See, e.g., U.S. PatentNo. 5,057,423, at column 1, lines 40-43).
The development of efflcient cell culture methods to rapidly grow T lymphocytes will be useful in both diagnostic and therapeutic applications. In diagnostic applications, the ability to rapidly expand T cells from a patient can be used, for example, to quickly generate sufficient numbers of cells for use in tests to monitor the specificity, activity, or other attributes of a patient's T lymphocytes. Moreover, the capability of rapidly achieving cell doses of lO9-10l~ cells will greatly facilitate the applicability of specific adoptive immunotherapy for the l~ of human ~ e~es There are several established methods already described for culturing cells for possible therapeutic use including methods to isolate and expand T cell clones. Typical cell culture methods for anchorage-dependent cells, (i.e., those cells that require ~tt~rhment to a substrate for cell proliferation) are limited by the amount of surface area available in culture vessels used (i.e., multi-well plates,.petri dishes, and culture flasks).
For anchorage-dependent cells, the only way to increase the number of cells grown is generally to use larger vessels with increased surface area and/or use more vessels.
However, hematopoietic cells such as T lymphocytes are anchorage-independent. They can survive and proliferate in response to the a~~ ,;ate growth factors in a suspension culture without ~ rh~"~nt to a substrate. Even with the ability to grow antigen-specific lyrnphocytes in a suspension culture, the methods reported to date have not c-)n~i~t~lltly produced rapid numerical expansion of T cell clones. For example, in a study of T cells conducted by Gillis and Watson, it was found that T cells cultured at low densities, i.e., 5 x 103 to 1 x 104 cell/ml in the presence of the T cell growth factor IL-2, proliferated rapidly over a seven day period and eventually reached a saturation density of 3-5 x 105 cells/ml.
Gillis, S. and Watson, J. "Interleukin-2 Dependent Culture of Cytolytic T Cell Lines", Tmmunolo~ical Rev.. 54:81-109 (1981). Furthermore, Gillis and Watson also found that once cells reached this saturation concentration, the cells would invariably die. Gillis et al., id.

W 09~132970 PCTnU~97J032g3 Another study reported three di~ferent methods for establishing murine T Iymphocytes in long-terrn culture. Paul et al., reported that the method most widely used is to grow T lymphocytes from immlmi7~1 donors for several weeks or more in the presence of antigen and antigen-presenting cells (APCs) to provide the requisite T cell receptor signal and co-stimulatory signals, and with the addition of exogenous growth factors before ~LL~ ling to clone them, Paul, W.E., et al., "Long-term growth and cloning of non-transformed lymphocytes", Nature, 294:697-699, (1981). T cells specific for protein antigens are then cloned by limiting dilution with antigen and irradiated spleen cells as a source of APCs. ~ second method involved growing T cells as colonies in so~
agar as soon as possible aftPr taking the cells from an immllni7~d donor. The T cells were stim~ tecl in an initial suspension culture with antigen and a source of APCs, usually irr~ ted spleen cells. In this second approach, it was found that, after 3 days, the cells were distributed in the upper layer of a two-layer soft agar culture system. The colonies were picked from day 4 to 8 and then exp~nl1ed in long-term cultures. The third approach involved selecting cells for their functional properties rather than their ~ntigenic specificity and then growing them with a series of dirr~.~iLl~ irradiated feeder cells and growth factor Co~ u~ . Paul, W.E. et al., "Long-term growth and cloning of non-bransformed lymphocytes", Nature, 294:697-699, (1981). It is a~p~u~-l that with each of these methods, it is not possible to expand individual T cell clones from a single cell to l 09-10l~ cells in a timely manner. Thus, despite the ability to clone antigen-specific T
cells, and convincing evidence of the thc~;ulic efficacy of T cell clones in accepted animal models, the technical difficulty in clllt lring human T cells to large numbers has impeded the clinical evaluation and application of cellular immlmotherapeutic procedures.
Yet another concern with cultured T cells is that they must remain capable of functioning in vivo in order to be useful in immlm~therapeutic procedures. In particular, it has been observed that antigen-specific T cells which were grown long term in culture in higl~ concentrations of IL-2 may develop cell cycle abnorm~liti~s and lose the ability to return to a ~uiescent phase when IL-2 is withdrawn. In contrast, the normal cell cycle consists of four successive phases: rnitosis (or "M" phase) and three phases which make up ~ 30 ~e "interphase" stage. During the M phase, the cell undergoes nuclear division and cytokinesis. The interphase stage consists of the Gl phase in which the biosynthetic activities resume at a high rate after mitosis, the S phase in which DNA synthesis occurs and the G2 phase which continues until mitosis commences. While in the (:i l phase, some cells appear to cease progressing through the division cycle; and are said to be in a "resting" or quiescent state (denoted as the "G0" state). Certain environment~l factors (such as a lack of growth factors in serum or confluence of cell cultures) may cause cells to enter the quiescent state. Once the factor is restored, the cell should resume its normal progress through the cycle. However, cells grown in culture may be unable to enter the quiescent phase when the growth factor is removed, rçslllting in the death of these cells.
This growth factor depçntl~n- e is particularly relevant to cultured T cells. T lymphocytes that are exposed over a long term to high concentrations of IL-2 to promote cell growth often will die by a process called apoptosis if IL-2 is removed or if they are subsequently stim~ te~l through the T cell receptor, i.e., if they encounter specific antigens. (see, e.g., Lenardo M.J., ~a~, 353:858-861, 1991, Boehme S.A. and Lenardo M.J., Fur. J
~mml1n~ 23:1552-1560, 1992). Therefore, the culture methods used to propagate LAK
cells or TIL-cells, and prior methods to culture T cells which predo...;,.~ ly rely on high long-term concentrations of IL-2 to promote expansion in vitro, may render many of the cells susceptible to apoptosis, thus limiting or elimin~tin~ their usefulness for cellular immllnotherapy.
It may also be advantageous in cellular immllnotherapy studies to use gene methods to insert foreign DNA into the T cells to provide a genetic marker, to f~-~ilit~te evaluation of in vivo migration and survival of transferred cells, or to confer functions that may improve the safety and efficacy of transferred T cells. An established method for stable gene Lld~ into nn~mm~ n cells is the use of amphotropic retroviral vectors (see, e.g., Miller AD, Cllrrent Topics in Microbiology ~n~1 Immlml-lo~v, 158:1-24, 1992). The stable integration of genes into the target cell using retrovirus vectors requires that the cell be actively cycling, specifically that these cells transit M phase of the cell cycle. Prior studies have introduced a marker gene into a small proportion of polyclonal T cells driven to proliferate with high doses of IL-2, and these cells were reinfused into hl~m~n~ as tumor therapy and provided a means of following the in vivo survival of transferred cells.
(Rosenberg et al. New F~l J. Med.~ 323:570-578, 1990). However, for human T cells (which cycle slowly when grown with standard techniques) the efficiency of stable gene transfer is very low, in the range of 0.1-1% of T cells. (Springett CM et al. J. Virolo~v~
63:3865-69, 1989). Culture methods which more efficiently recruit the target T cells ~nto wo g ll32970 PCTIUS97/03293 the S and G2-M phases of the cell cycle may increase the efficiency of gene modification using retrovirus-mediated gene transfer (Roe T. et al., FMP~O J, 2:2099-2108, 1993), thus improving the prospects fo~r using genetically-modified T cells in cellular immunotherapy or using T cells to deliver defective genes in genetic deficiency diseases.
The rapid expansion method described by S. Riddell et al. (in PCT Publication WO96/06929, published 07 March 1996), hereinafter referred to as "high-PBMC REM" or "h!p-REM" was developed to provide functional, antigen-specific T cell clones ffir use in clinical adoptive immnn~therapy protocols. The hp-REM protocol was designed to provide m~im~l T cell expansion in a limited amount of time without loss of T cell function and specificity. Generally, the hp-REM protocol involves the steps of adding an initial T lymphocyte population to a culture medium in vitro; adding to the culture medium a disproportionately large number of non-dividing peripheral blood mononuclear cells ("PBMC") as feeder cells such that the resulting population of cells contains at least about 40 PBMC feeder cells (preferably at least about 200, more preferably at least about 400) for each T lymphocyte in the initial population to be exr~n~le~; and incubating the culture. In pL~,r~llcd embo-liment~ of the hp-REM protocol, the T cells to be expanded are also exposed to a disproportionately large number of EBV-transformed lymphoblastoid cells ("LCL"), to an anti-CD3 monoclonal antibody (e.g., OKT3) (to activate the T cells via the T cell antigen receptor), and to the T cell growth factor interleukin-2 (IL-2).
In the hp-REM protocol, T cells are generally e~p~n~l~cl using a vast excess of feeder cells conei.etin~ of periI h~r~l blood mononuclear cells (PBMC) and possibly also EBV-transformed lymphoblastoid cells (EBV-LCL). T cells to be t~xr~n~e~l typically represent less than about 0.2% of the cells in the hp-REM culture method. As described, the T cells can be activated through the T cell antigen lcc~lc,r using an anti-CD3 2~ monoclonal antibody (e.g. OKT3~ and T cell proliferation can be in~ cet1 using IL-2.
Such hp-REM culture conditions were reported to result in a level of T cell expansion 100 to 200-fold greater than that reported by others.
However, for most uses, it would be preferable to avoid the use of large excesses of feeder cells (i.e. PBMC and EBV-LCL) in the pl~dlion of T cells destined for clinical use. For example, PBMCs are derived from human blood and could represent a potential source of adventitious agents (e.g. human imunodeficiency virus, type 1 and 2;
hurnan T cell leukemia virus I, type 1 and 2; and hepatitis virus, such as hepatitis B, C and W097/32970 PCTrUS97/03293 G), and EBV-LCL could represent a potential source of Epstein-Barr virus. In addition, the large-scale application of the hp-REM protocol would require a large supply of human peripheral blood to provide adequate numbers of feeder cells.
It would therefore be particularly advantageous to reduce the numbers of such feeder cells required or to replace them entirely. With these concems in mind, the methods of the present invention (hereinafter referred to as "low-PBMC-REM" or "modified-REM") are clesi~n~l to achieve rapid in vitro expansion of T cells without using the vast excess of PBMC and/or EBV-LCL feeder cells that are the key characteristic of the hp-REM protocol.
R~TFF SuMM~ARy OF THF rNVF~TION
This invention provides a method for rapidly producing large numbers of T cells,including human antigen-specific cytolytic and helper T cells, isolated from an initial population of T cells, without using the vast excess of PBMC and/or EBV-LCL feeder cells that are the key characteristic of the hp-REM protocol. While the methods of the present invention are applicable to the rapid expansion of T lymphocytes, generally, the rapid expansion method will be especially advantageous in situations in which anindividual T cell clone must be exr~n~le~l to provide a large population of T lymphocytes.
Thus, the present invention provides an especially i~ Ol l~l~ tool in the context of human adoptive immllnt~therapy, as has ~een ex~mrlified in studies (using hp-R~M, described below) involving human bone marrow transplant recipients at the Fred ~lltf~hin~on Cancer Research Center. The present invention also provides a method to improve the efficiency of stable gene transfer into T lymphocytes, as exemplified below.
Accordingly, one object of the invention is to rapidly expand T lyrnphocytes to large numbers in vitro without using the vast excess of PBMC and/or EBV-LCL feeder cells that are the key characteristic of the hp-REM protocol. Such rapidly expanded T cell populations can be used, inter alia, for infusion into individuals for the purpose of co~ ing a specific immllne response, as exemplified herein. T cells that can be ~xp~n-le~l using the present invention include any of the various T lymphocyte populations described herein (see, e.g., the discussion above l~g~dillg CTLs, helper T cells and other T lymphocytes, and the potential uses of such cells in imn~unotherapeutic techniques).

-W 09~S3297~ PCTnUS97/03293 Another object of the invention is to use the method to grow T cells in a mannerwhich f~rilit~t~5 the stable introduction of foreign genetic material which can be used to alter the function of T cells to be used in cellular immtlnotherapies, as described above, or to otherwise overcome for a defective or inadequate gene in the host.
S A number of preferred embodiments of the present invention are described in the following enumeration:
1. A method (referred to herein as "low-PBMC-REM" or "modified-REM") for rapidly exr~ntling an initial T Iymphocyte population in culture medium in vitro, comprising the steps of: adding an initial T Iymphocyte population to a culture m~ ]m in vitro; adding to the culture medium a non-dividing m~mmz~ n cell line ~c~lessil-g at least one T-cell-stimlIl~t~ ry component, wherein said cell line is not an EBV-transformed Iymphoblastoid cell line (LCL), and incubating the culture. REM cultures will generally be incubated under conditions of temperature and the like that are suitable for the growth of T lymphocytes. For the growth of human T Iymphocytes, for example, the te~ eldlule will generally be at least about 25 degrees Celsius, preferably at least about 30 degrees, more preferably about 37 degrees. Descriptions of suitable media and other culture conditions are well-known in the art, and are also exemplified herein.
2. A rapid ~xr~n~inn method according to the preceding item, wherein said T-cell-stimIlI~tory component is selected from the group cnn~i~tin~ of an Fc-r ,ec~Lor, a cell ~-lh~ n-~c.cessory molecule and a cytokine.

3. A rapid expansion method according to any of the prece~1ing items, wherein said T-cell-stim~ tory component is selected from the group consisting of an Fc- y receptor, a cell adhesion-accessory molecule and a cytokine, and wh~ said initial T
lymphocyte population is t~xr~ntled at least 200-fold after an incubation period of less than about two weeks.

4. A rapid expansion method according to any of the prece-lin~ items, wherein said T-cell-stimulatory coll,yonelll is selected from the group consisting of an Fc-y . receptor, a cell adhesion-accessory molecule and a cytokine, and wherein said initial T
lymphocyte population is e~r~ndç~1 at least ~00-fold after an incubation period of less than ~ 30 about two weeks.

5. A rapid expansion method according to any of the preceding items, wherein said T-cell-st;mIll~t- ry component is selected from the group consisting of an Fc-~y _ PCTrUS97/03293 reeeptor, a eell adhesion-accessory molecule and a cytokine, and wherein said initial T
lymphocyte population is expanded at least 1 0~)0-fold after an incubation period of less than about two weeks.

6. A rapid expansion method according to any of the preceding items, further comprising the step of adding anti-CD3 monoclonal antibody to the culture mediumwherein the coneentration of anti-CD3 monoclonal antibody is at least about 1.0 ng/ml.
Typically, a concentration of about 10 ng/ml is employed although much lower levels can be used, as illustrated below.

7. A rapid expansion method according to any of the preceding items, further comprising the step of adding IL-2 to the culture medium, wherein the concentration of IL-2 is at least about 10 units/ml. Typically, a concentration of about 25 units/ml is used.
Preferably, the incubation is continued for at least about 9 days and wherein the step of adding IL-2 to the culture medium is repeated after each 3-5 day interval. Typieally, IL-2 is added on day 0, again on day 5 or 6, and again on day 8 or 9.

8. A rapid expansion method according to any of the preceding items, wherein said ~ n eell line eomprises at least one eell type that is present at a frequeney at least twiee that found in human peripheral blood mononuelear cells (human PBMCs), preferably at least three times, at least ten times, or at least fifty times the frequency generally found in human PBMCs.

9. A rapid expansion method according to any of the preceding items, wherein said T-cell-stim~ tory eomponent is selected from the group eonsisting of an Fe-~
reeeptor and a eell adhesion-aeeessory molecule.

10. A rapid expansion method according to any of the preceding items, wherein said T-cell-stim~ tory component is seleeted from the group con~i~ting of a cell adhesion-~cees~ory molecule and a cytokine.

11. A rapid expansion method aeeording to any of the preeeding items, wherein said T-eell-stim~ tory eomponent is seleeted from the group e-)n~i~ting of an Fe-~
reeeptor and a eytokine.

12. A rapid expansion method aceording to any of the preeeding items, wherein said m~mm~ n cell line expresses a cell adhesion-accessory molecule.

13. A rapid expansion method according to any of the preee~ing items, wherein said eell adhesion-aeeessory molecule is seleeted from the group eon~i~tin~ of Class II
1~

wo 97~32970 PCTJUS97103293 MHC, Class I MHC, ICAM 1, ICAM 2, ICAM 3, CD58, CD72, fibronectin, ligand to CD27, CD80, CD86 and hyaluronate.

14. A rapid expansion method according to any of the prece~lin~ items, wherein said m~mm~ n cell line expresses a cytokine. Preferably the cytokine is an interleukin.
S 15. A rapid expansion method according to any of the prece~1inE~ items, wherein - said T-cell-s~im~ tory component is a molecule that binds to CD21.
16. A rapid expansion me~od according to any of the prece~lin~ items, wherein said cytokine is selected'from the group consisting of IL-l, IL-2, IL-4, IL-6, IL-7, IL-12 and IL-15.
17. A rapid exp~n~i on method according to any of the precedin~ items, further compri.~in~ the step of adding a soluble T-cell-stimnl~tory factor to the culture medium.
18. A rapid expansion method according to any of the prece~1in~ items, wherein said soluble T-cell-stim~ tory factor is selected from the group consisting of a cytokine, an antibody specific for a T cell surface component, and an antibody specific for a co,~on~-,l capable of binding to a T cell surface component.
19. A rapid expansion method according to any of the prece~lin~ items, wherein said soluble T-cell-stimnl~fory factor is a cytokine selected from the group con~i~tin~ of IL-l, IL-2, IL-4, IL-6, IL-7, IL-12 and IL-15.
20. A rapid expansion method according to any of the prece~1in~ items, wherein said soluble T-cell-stimlll~tory factor is an antibody specific for a T cell surface component, and wherein said T cell surface colllpollent is selected from the group con~i~ting of CD4, CD8, CDl la, CD2, CD5, CD49d, CD27, CD28 and CD44.
21. A rapid expansion method according to any of the precer1ing items, wherein said soluble T-cell-stim~ tory factor is an antibody specific for a component capable of binding to a T cell surface component, and wherein said T cell surface component is selected fromthe group consisting o~CD4, CD8, CDlla, CD2, CD5, CD49d, CD27, CD28 and CD44.
22. A rapid expansion method according to any of the preceding items, wht;
said soluble T-cell-stimul~tory factor is a molecule that binds to CD21.
~ 3{) 23. A rapid expansion method according to any of the prece~lin items, wherein said molecule that binds to CD21 is an anti-CD21 antibody.
l3 _ 24. A rapid expansion method according to any of the preceding items, further comprising the step of adding to the culture a multiplicity of peripheral blood mononuclear cells (PBMCs). Preferably, PBMC are irradiated with gamma rays in the range of about 3000 to 3600 rads, more preferably at about 3300 rads.
25. A rapid expansion method according to any of the preceding items, wherein the ratio of PBMCs to initial T cells to be exp~n-1e~ is less than about 40~
26. A rapid expansion method according to any of the preceding items, wherein the ratio of PBMCs to initial T cells to be exr~n-lf cl is less than about 10:1.27. A rapid expansion method according to any of the preceding items, wherein the ratio of PBMCs to initial T cells to be exp~n~l~d is less than about 3 :1.
28. A rapid expansion method according to any of the preceding items, further compti~in~ the step of adding to the culture a multiplicity of EBV-transformed lymphoblastoid cells (LCLs). Preferably, PBMC are irradiated with gamma rays in the range of about 6000 to 10,000 rads, more preferably at about 8000 rads.
29. A rapid expansion method according to any of the preceding items, wherein the ratio of LCI,s to initial T cells to be exp~n-le~l is less than about 10:1.
30. A rapid expansion method according to any of the preceding items, wherein the initial T lymphocyte population comprises at least one human CD8+ antigen-specific ~;yLoloxic T lymphocyte (CTL). In ~ d embo-lim~nt~ of the present invention, theCTL is specific for an antigen present on a human tumor or encoded by a pathogen such as a virus or bacterium.
31. A rapid ~xp~n~ n method according to any of the preceding items, wherein the initial T lymphocyte population compri~es at least one human CD4+ antigen-specific helper T lymphocyte.
32. A method of genetically transducing a human T cell, compri.~ing the steps of: adding an initial T lymphocyte population to a culture llle-liu.ll in vitro; adding to the cul~re medium a non-EBV-transformed m~mm~ n cell line expressing a T-cell-stim~ t-~ry component; and incubating the culture; and adding a vector to the culture mediurn. A vector refers to a unit of DNA or RNA in a form which is capable of being introduced into a target cell. Transduction is used generally to refer to the introduction of such exogenous DNA or RNA into a target cell and includes the introduction of heterologous DNA or RNA sequences into tar~get cells by, e.g., viral infection and W 097/32970 PCTnUSg7/032g3 ele~ opoldlion. A currently pl~r~lled method of transducing T Iymphocytes is to use rekoviral vectors, as exemplified herein.
33. A genetic tr~n~ ction method according to item 32, wherein the vector is a retroviral vector cont~ining a selectable marker providing resistance to an inhibitory compound that inhibits T Iymphocytes, and wherein the method further comprises the steps of: co..l; ~ g incubation of the culture for at least one day after addition of the retroviral vector, and adding said inhibitory compound to the culture medium after said continued incubation step. Preferably, the retroviral vector contains both a positive and a negative selectable marker. Preferred positive selectable markers are derived from genes selected from the group c--n~i~ting of hph, neo, and ~pt, and ple~cll.d negative selectable m~rkers are derived from genes selected from the group con~i~inE~ of cytosine ~1ç~min~e, HSV-I TK, VZV TK, HPRT, APRT and ~pt. Especially ~ ell~,d markers are b~ifunctional selectable fusion genes wherein the positive selectable marker is derived from ~h or neo, and the negative selectable marker is derived from cytosine ~le~ e or a TK
gene.
34. A genetic tr~n~ ction method according to any of items 32-33, further compri~in~ adding a multiplicity of human PBMCs.
35. A rapid expansion method according to any of items 32-34, wh~ileill the ratio of PBMCs to initial T cells is less than about 40: 1.
36. A genetic tr~n~dllction method according to any of items 32-35, further comrri~ing adding non-dividing EBV-transformed lymphoblastoid cells ~LCL).
37. A rapid expansion method according to any of items 32-36, wherein the r atio of LCL to initial T cells is less than about 10: 1.
38. A method of generating a R~M cell line capable of promoting rapid expansion of an initial T lymphocyte population in vi~ro, compri.~ing the steps of:
depleting one or more cell types from a human PBMC population to produce a cell-type-depleted PBMC population, using said cell-type-depleted PBMC population in place of non-depleted PBMCs in an hp-REM protocol to determine the contribution of the depleted cell type to the activity provided by the non-depleted PBMCs, identifying a T cell stim~ tory activity provided by said depleted cell type, and L~ nillg a m~mm~ n cell line with a gene allowing expression of said T cell stim~ tory activity.
lS

39. A method of generating a REM cell line according to item 38, wherein said T-cell-stim~ t--ry component is selected from the group consisting of an Fc- y receptor, a cell ~(1hesion-~ccessory molecule and a cytokine.
40. A REM cell line capable of stimulating rapid expansion of an initial T
lymphocyte population in vitro, coml~ri~ing a m~mm~ n cell line generated according to a method according to the preceding item 38 or item 39.
41. A REM cell line according to item 40, wherein said cell line expresses a cell adhesion-accessory molecule.
42. A E~EM cell line according to any of items 40-41, wherein said cell adhesion-accessory molecule is selected from the group consisting of Class II MHC, Class I MHC, ICAM 1, ICAM 2, ICAM 3, CD58, CD72, fibronectin, ligand to CD27, CD80, CD86 and hyall~lon~le.
43. A REM cell line according to any of items 40-42, wherein said cell line expresses an Fc-~ receptor.
44. A REM cell line according to any of items 40-43, wherein said cell line expresses at least one T cell stim~ tory cytokine.
45. A REM cell line according to any of items 40-44, wherein said T cell stim~ tory cytokine is selected from the group consisting of IL-l, lL-2, IL-6, I~-7, IL-12 and IL-15.
46. A REM cell line according to any of items 40-44, wherein said cell line c;~ ;S~es a molecule that binds CD21. As used herein, a molecule that binds CD21 can be a natural or synthetic molecule known or ~l~tPrmin(~d to bind to the CD21 cell surface detr.. " ,i.~ Molecules known to bind to CD21 include anti-CD21 antibodies, as well as molecules such as C3d, C3dg, iC3b and EBV gp350/220, and derivatives thereof.
47. A culture medium capable of rapidly e~r~n~ling an initial T lymphocyte population in vitro compri~ing a REM cell line according to any of items 40-46.
48. A culture medium according to item 47, further comrri~ing an exogenous cytokine.
49. A culture medium according to any of items 47-48, further comprising a mnltirlicity of exogenous cytokines, wherein said multiplicity compri~çs at least one interleukin. 16 WO 9~1132970 PCT~US97/03293 50. A culture medium according to ally of items 47-49, wherein said interleukin is selected from the group consisting of IL-l, IL-2, IL-6, IL-7, IL-12 alld IL-15.
51. A culture medium according to any of items 47-50, further comprising a molecule that binds to CD21. As used herein, a molecule that binds CD21 can be a natural or synthetic molecule known or determined to bind to the CD21 cell surface determin~nt - MLolecules known to bind to CD21 include anti-CD21 antibodies, as well as molecules such as C3d, C3dg, iC3b and EBV gp350/220, and derivatives thereof that bind to CD21.
52. A culture medium according to item 51 "wherein said molecule that binds to CD21 is an anti-CD~l antibody.
53. A culture medium according to any of items 49-52, further comprising an anti-CD3 monoclonal antibody.

p~TATT Fl) DF~CRTPTION OF PRFFFRRF~ EMROT Tl~FNTS Al~D
,~PpT TCATIONS OF T~F ~VFNTION
The invention described herein provides methods for rapidly .o~p~nr1ing populations of T lymphocytes, including human cytotoxic T lymphocytes and helperT lymphocytes, which can be particularly useful in cellular immunotherapy of human ~li5e~es, without using the vast excess of PBMC and/or EBV-LCL feeder cells that are the key characteristic of the hp-REM protocol.
The T cells will be referred to as "target T cells". In general, target T cells are added in small numbers to a culture vessel co~ standard growth mediurn that has been supplçment~d with components that stim~ t~ rapid exr~n~i- n in vitro (REM) as dP,scribe-l herein. Preferably, human recombinant IL-2 or another suitable IL-2 p~ al~lion is added in low concentrations at 3-5 day intervals (typically on "day 0" (i.e. at culture initiation) or "day 1" (the day following initi~i~tion), again on day 5 or 6, and again on day 8 or 9). REM protocols result in a rapid expansion of T cells, typically in the range of a ~00- to 3000-fold expansion v~rithin 8 to 14 days. Such methods can thus achieve -xp~n~i-n rates that are a~roxi",ately 100- to 1000-fold more efficient for each~iml~ ion cycle than previously-described methods of culturing human T cells.
~ 30 Furthermore, REM protocols are applicable to the rapid exp~n~i~ n of any T cell sub-population including helper T cells and cytolytic T cells; and to T cell clones of many different antigenic specificities (e.g., to cytolytic or helper T cells specific for CMV, HIV, or other viral, b~cteri~l, or tumor-derived antigens). In addition, REM protocols can be used for both small scale growth (e.g. to rapidly expand T cells from 104 to 107 cells); or for large-scale expansions (e.g. to rapidly expand T cells from 106 to greater than 101~
cells), depending on the size of culture vessel chosen.
S REM protocols thus make it possible to efficiently expand T cell clones for use in adoptive immunotherapies by dr~m~tic~lly shortening the time required to grow the nurnbers of cells required to restore, enhance, or modulate human i.",.,.",ily. In the study by Riddell et al. (Science, 257:238-240, 1992), once T cell clones were isolated it was nece~ry to culture ~e clones for twelve weeks and to pool multiple clones to achieve the highest ~lmini~tered cell dose of 1 x 109 CD8+ CMV-specific T cells/m2 body surface area. Using REM protocols, the expansion of individual T cell clones to greater than 109 cells can be accomplished in less than three weeks.
With respect to the rapid expansion methods (i.e. "REM" technology), the following abbreviations are used to distinguish the various REM protocols referred to herein. The basic Riddell protocol (as described above and in the cited Riddell patent application), which uses a disproportionately large number of PBMC feeder cells (and preferably also EBV-LCL feeder cells) is referred to as "high-PBMC REM" or simply "hp-REM". Conversely, the methods of the present invention, which do not employ such large excçsses of PBMC feeder cells (and preferably no EBV-LCL feeder cells) arereferred to as "low-PBMC REM" or "modified-REM". Such methods are described in detail below.
The practice of the present invention will employ, unless otherwise indicated, collvt;llLional techniques of molecular biology, microbiology, cell biology, recombinant DNA, and immlmcllogy, which are within the skill of the art. Such techniques areexplained fillly in the liL~d~ulG. See e.g., Sambrook, Fritsch, and M~ni~t;~, Molecllklr Cloni~: A Labol~lo-y Manual. Second Edition (1989); Anim~l Cell Cul~lre (R.I.
Freshney, Ed., 1987); GPne Tr~n~fer Vectors for Mamm~ n Cells (J.M. Miller and M.P.
Calos eds. 1987); ~n~lhook of Fxper;ment~l Tmmlm~-lo~y, (D.M. Weir and C.C.
Blackwell, Eds.); Current Protocols in Molecular Biology (F.M. Ausubel, R. Brent, R.E.
Kin~t- n, D.D. Moore, J.G. Siedman, J.A. Smith, and K. Struhl, eds., 1987); Current Protocols in Lm--munology (J.E. Coligan, A.M. Kruisbeek, D.H. Margulies, E.M. Shevach f~

PCTn~S971~3293 and W. Strober, eds., 1991); Oli~onucleotide Synthesis (M.J. Gait Ed., 1984), and the series Methods ;n F.n7ymolo~y (Academic Press, Inc.).
All patents, patent applications, and publications mentioned herein, both supra and infra, are hereby incorporated herein by reference.
S As an aid in unders~ntlin~ this invention, the following is a list of some abbreviations commonly used herein:
CTL cytotoxic T lymphocyte(s) APC antigen-prest~ntin~ cell(s) CMV cytomegalovirus HIV lluman immun~deficiency virus EBV Epstein Barr virus hIL-2 human interleukin-2 MHC major histocompatibility complex PBMC peripheral blood mononuclear cell(s) EBV-LCL EBV-transformed lymphoblastoid cell line (sometimes abbreviated as simply "LCL") PBS phosphate buffered solution REM rapid expansion method hp-REM high-PBMC REM
lp-REM low-PBMC or "modified" REM

A "cytokine," as used herein, refers to any of a variety of intercellular ~i~n~lin~
molecules (the best known of which are involved in the regulation of 1~5 "1ll~ n somatic cells). A number of families of cytokines, both growth promoting and growth inhibitory in their effects, have been charact~ri7~d including, for eY~mrle: int~rlellkin~ (such as IL-la, IL-l ~, IL-2, IL-3, IL-4, IL-S, IL-6, IL-7, IL-8, IL-9 (P40), IL-l 0, IL-l 1, IL-12, IL-13, IL-14iand IL-15), CSF-type cytokines such as GM-CSF, G-CSF, M-CSF, LIF, EPO, TPO
("thrombopoietin"), TNF-oc, and TNF-O; i~ "r~lons (such as IFN-a, IFN-,B, IFN-~;cytokines of the TGF-~ family (such as TGF-,B 1, TGF-,B2, TGF-,B3, inhibin A, inhibin B, activin A, activin B~; growth factors (such as EGF, VEGF, SCF ("stem cell factor" or "steel factor"), TGF-a, aFGF, bFGF, KGF, PDGF-A, PDGF-B, PD-ECGF, INS, IGF-I, IGF-II, NGF-,~); a-type intercrine cytokines (such as IL-8, GRO/MGSA, PF-4, Iq PBP/CTAP/,~TG, IP-10, MIP-2, KC, 9E3); and ,~-type intercrine cytokines (such asMCAF, ACT-2/PAT 744/G26, LD-78/PAT 464, RANTES, G26, I309, JE, TCA3, MIP-la,B, CRG-2); and chemotactic factors (such as NAP-l, MCP-l, MIP-loc, MIP-l,B, Ml[P-2, SIS,~, SIS~, SIS~, PF-4, PBP, yIP-10, MGSA). A number of other cytokines are also known to those of skill in the art. The sources, characteristics, targets and effector activities of these cytokines have been described and, for many of the cytokines, the DNA
sequences encoding the molecules are also known; see, e.g., R. Callard & A. Gearing, The Cytokine Facts Rook (Academic Press, 1994~, and the particular publications reviewed and/or cited therein, which are hereby incorporated by reference in their entirety. As referenced in catalogs such as The Cytokine Facts Book, many of the DNA and/or protein sequences encoding such cytokines are also generally available from sequence databases such as GENBANK (DNA); and/or SWISSPROT (protein). Typically, cloned DNA
encoding such cytokines will already be available as plasmids, although it is also possible to synth~si7~ polynucleotides encoding the cytokines based upon the published sequence information. Polynucleotides encoding the cytokines can also be obtained using polymerase chain reaction (PCR) methodology, as (l~sçribed in the art. ~, e.g., Mullis &
Faloona, Met. Fn7~molo~y, 155: 355 (1987). The detection, pllrific~tion, and chara~ ion of cytokine~, including assays for identifying new cytokines effective upon a given cell type, have also been described in a number of publications as well as the references referredto herein. See, e.g.,Lyn~h~kin.qc ~n~ , 1987; and DeMaeyer, E., et al., "Illle~ ,lls and Other Regulatory Cytokines," (John Wiley & Sons 1988).
A m~mm~ n ~'cell line", as used herein, refers to a population of m~mm~ n cells (preferably human cells) that have undergone repeated propagation in vitro; as distinguished from "primary cells" taken from an individual such as a human. Generally, a m~mm~ n cell line will have been prop~g~te-1 in vilro for at least about 10 gelleldlions, more typically at least about 40 generations, most typically at least about 100 generations.
Most preferably, the m~mm~ n cell line can be propagated and m~int~inPrl long terrn t (i.e., at least several months in vi~ro, preferably at least a year). Such cell lines would include, but are not limited to, "clonal" lines (in which all of cells of the population are derived from a single ancestral cell). Conversely, a mixed peripheral blood population such as PBMCs would not constitute a m~mm~ n cell line. A m~mm~ n cell line for ~0 W 097132970 PCTrUS97JO3293 use in the present invention may, however, contain a cell type found in peripheral blood but in that case the cell type will generally be present at a frequency much higher than is normally found in human peripheral blood mononuclear cells (at least twice the frequency generally found in human peripheral blood mononuclear cells, preferably at least five ti~nes, at least ten times, at least twenty times or at least fifty times the frequency generalTy ~ found in human peripheral blood mononuclear cells). A particular "cell type" might be, for example, one of the cell types typically found in peripheral blood (such as B
lymphocytes, monocytes, cytotoxic T lymphocytes, helper T lymphocytes, granulocytes, eosinophils or NK cells); or of a cell type not normally found in peripheral blood (such as fibroblasts, endothelial cells, etc.); or a more specific subpopulation of such a cell type (e.g. a subpopulation tl1at is relatively homogeneous with respect to antigen-specificity or t~res~ion of a particular receptor). Thus, a cell line might be relatively homogeneous with respect to attributes such as antigen-specificity or cell surface lec~l~ /lig~n~ls, as discussed in more detail below. By way of illustration, a receptor-specific monocyte line refers to a population of cells in vitro in which the majority of cells are monocytes poc~e~ing a particular cell surface receptor (which cell line might have been obtained for example by transforming a population of monocytes with genes t;~lC;S~ g the particular receptor). Again, by way of illustration, an antigen-specific CTL cell line refers to a population of cells in vitro in which the majority of cells are cytotoxic T Iymphocytes specific for a particular antigen such as a viral, b~ct~ l or tumor antigen (which cell line might have been obtained for c~mple by exposing a population of T cells to repeated stim~ tion with a particular antigen and subsequently enriching for antigen-specific CTLs).
Preferably, such a cell line for use with the present invention will be rendered non-dividing prior to use in the modified-REM culture (e.g., by irradiation). However, one can ~ltern~tively (or in addition) employ a cell line that is dividing (preferably at a rate similar to or slower than the expanding T cells) but which can be subsequently elin in~tçd by ~irtue of its having a negative selectable marker (e.g., a suicide gene that can be used to inhibit or kill cells carrying the gene, or a cell surface marker that can be used to isolate and/or elimin~te cells carrying the marker). In the latter case, the cell line can be allowed to expand to some degree in the REM culture before being negatively selected.
~ l -Preferably, m~mm~ n cell lines to be used with the present invention are relatively homogeneous lines (i.e. at least 50% of the cells are of a particular cell type, more preferably at least 70%, at least 90%, at least 95% or at least 99% of the cells are of a particular cell type). It should be noted, however, that T cells to be expanded by exposure to such a cell line might also be exposed to additional cell lines ~at the same time or in sequence). Thus, by way of illustration, a modified-REM culture (cont~ining a T
lymphocyte population to be expanded) might be exposed to one mslmm~ n cell line or to several such lines. For modified-REM, T cells to be expanded will be exposed to at least one such m~mm~ n cell line and/or to a non-cellular nli~Lul~; of factors (including, e.g., cytokines, antibodies, soluble lig~n(1s, etc.), as discussed herein.
The T cells to be propagated in culture (i.e., the "target" T-cells) can be obtained from the subject to be treated. ~ltP.rn~tively, T cells can be obtained from a source other than the subject to be treated, in which case the recipient and transferred cells are preferably immunologically compatible (or the receipient is otherwise made immuno-tolerant of the transferred cells). Typically, the target T cells are derived from tissue, bone marrow, fetal tissue, or peripheral blood. Preferably, the cells are derived from peripheral blood. If the T cells are derived from tissues, single cell suspensions can be ~ ~ed using a suitable medium or diluent.
Mononuclear cells co~ g the T lymphocytes can be isolated from the heterogenous population according to any of the methods well known in the art. As illustrative examples, Ficoll-Hypaque gradient centrifil~ti~-n, fluorescence-activated cell sorting (FACs), p~nnin~ on monoclonal antibody coated plates, and/or magnetic separation techniques can be used (separately or in combination) to obtain purified populations of cells for expansion according to the present invention. Antigen-specific T cells can be i~ol~tetl by standard culture techniques known in the art involving initial activation of antigen-specific T cell precursors by stim~ tion with antigen-~~s~l.L;~-g cells and, for a clonal population, by limiting dilution cultures using techniques known in the art, such as those described in Riddell and Greenberg (J. Tmmlmol. Me~.~ 128:189-201, 1990); and Riddell et al. (J. Tmmunol., 146:2795-2804, 1991). See also, the Examples below. T cell clones isolated in microwells in limiting dilution cultures typically have e~r~n-le~l from a single cell to 2 x 104 to 5 x 105 cells after 14 days.

WQ 91132s7û PCTIUS97l03293 For expansion, T cells can be placed in a~-vyliate culture media in plastic culture vessels with T cell stimulatory components as described herein. The initial phase of rapid expansion is generally calTied out in a culture vessel, the size of which depends upon the number of target cells, and which may typically be a 25 cm2 flask. The size of the culture S vessel used for subsequent cycles of T cell expansion depends on the starting number of ~ T cells and the number of cells needed. Typical starting cell nurnbers for dirre~c~ sized culture vessels are as follows: 5x104 to 2xlOs ~lo~illlately 25cm2 flask; 2xlOs to 5xlOs approximately 752cm flask; 5xlOs to lx106 - approximately 225-cm2 flask; and lxl o6 to 2xl o6 roller bottle. The approximate initial volurne of media used with each flask is: 25 cm2 - 20-30 ml; 75 cm2 - 60-90 ml; 225 cm2 - 100-200 ml; roller bottle - 500 ml.
For even larger-scale exr~n~inns, a variety of culture means can be used, including for example, spinner flasks, cell culture bags, and bioreactors (such as hollow-fiber bioreactors).
As used herein, "feeder cells" are accessory cells that provide co-stim~ ting fi~nctions in conjunction with T cell receptor activation (which can be achieved by ligation of the T cell receptor complex with anti-CD3 monoclonal antibody). PBMC feeder cells for use in REM can be obtained by techniques known in the art, for exarnple by k~phoresis, which is a standard medical procedure with minim~l risks (see, e.g.,Weaver et al.,3~ 82:1981-1984, 1993), and these feeder cells can be stored by clyc~yl~s~ lion in liquid nitrogen until use. LCL can be generated from perirhpral blood B cells by tr~n~form~tiQn with ~BV, for example the B95-8 strain of EBV, using standard methods (see, e.g., Crossland et al., J. Tmmunol. 146:4414-20, 1991), or by spontaneous ouL~owlh in the presence of cyclosporin A. Such LCL cells will grow rapidly and i~ efinitely in culture.
Prior to adding any feeder cells to the culture vessel (whether PBMCs or cells derived from a cell line as described herein), such feeder cells are preferably prevented from undergoing mitosis. Techniques for preventing mitosis are well known in the art and include, for example irradiation. For example, any PBMCs can be irradiated with gamma rays in the range of about 3000 to 4000 rads (preferably PBMCs are irradiated at about 3600 rads); any LCL can be irradiated with gamma rays in the range of about 6000-12,000 rads (preferably LCL are irradiated at about 10,000 rads); and any cells derived from other PCT~US97/03293 cell lines can also be irradiated with gamma rays in the range of about 6000-12,000 rads As discussed above, negatively selectable feeder cells can also be used.
Since the antigen specificity of the T cell clone is generally defined prior to expanding the cell in the culture system, either autologous or allogeneic feeder cells can be used to support T cell growth. The ability to use allogeneic feeder cells is important in situations in which the patient is infected with a virus that is present in PBMC, e.g., HIV, that could therefore co"~l";~te the T cell cultures. In such circllmct~n-çs, the use of allogeneic feeder cells derived from an individual that is screened and deemed to be a suitable blood donor by American Red Cross criteria can be used in the culture method.
The T cell receptor activation signal (normally provided by antigen and antigen-,el~ g cells) can be achieved by the addition anti-CD3 monoclonal antibodies to the culture system. The anti-CD3 monoclonal antibody most commonly used is "OKT3", which is commercially available from Ortho Ph~rm~ellticals in a f~ tion suitable for clinical use. The use of anti-CD3 ("aCD3") mAb rather than antigen as a means oflig~tin~ the T cell receptor bypasses the need to have a source of antigen-pres~ nting cells, which for virus-specific T cells would require ~ large numbers of suitable autologous cells and infecting these cells in vitro with high titer virus. A concentration of anti-CD3 monoclonal antibody of at least about 0.5 ng/ml, preferably at least about 1 ng/ml, more preferably at least about 2 ng/ml, promoted the rapid exr~n~ n of the T cells such that a 500- to 3000-fold expansion can be achieved within about 10 to 13 days of growth. Typically, a concentration of about 10 ng/ml anti-CD3 monoclonal antibody was used.
Of course, as an ~l L~ ; ve to anti-CD3 monoclonal antibody, the T cell lece~
can be a~;liv~ted and the cells stim~ tç-l by the addition of antigen-~ s~ ;..g cells, as described in Riddell et al., J. Tmml1no1. 146:2795-2904, 1991. Suitable antigen-~ ,s~ ;... g cells include, for example, viral infected cells, tumor cells, and cells pulsed with the relevant peptide antigen.
The culture media for use in the methods of the invention can be any of the commercially available media, preferably one col~t~ : RPMI, 25 mM HEPES, 25 ~M
2-mercaptoethanol, 4 mM L-g1l11~mine, and 11% human AB serum. Fetal calf serum can be sub~Li~u~ed for human AB serum. Preferably, after addition of any feeder cells, anti-CD3 monoclonal antibody, and culture media are added to the target T cells, and the WO 97l32970 PCT/US97103293 mixture is allowed to incubate at 37~C in a 5% CO2 humidified atmosphere under standard cell culture conditions which are well known in the art. Typically, such conditions may include venting, and addition of CO2 if necessary (e.g., 5% CO2, in a humidifiedincubator).
Preferably, the medium is also supplemen1ed with interleukin-2 (IL-2). Typically- recombinant human IL-2 is used, although a functional equivalent thereof may also be used. Preferably, IL-2 is added on day 1, and is re-added at 3-5 day intervals. Thus, IL-2 was generally added on day 1, on day 5 or 6, and again on day 8 or 9. Expansion can be improved by using an IL-2 concentration of at least about 5 U/ml, more preferably at least about 10 U/ml. Generally, a concentration of about 25 U/ml can be used.
As described in Riddell et al., supra, antigen-specific T cells P,~Cp~n~lerl using REM
r.etained their antigen-specific functionality. For example, four dirrt;lGil~ HIV-specific CD8+ cytotoxic T cell clones retained the ability to kill virus-infected cells ~le~hlg the relevant antigen (i.e. HIV), and did not acquire non-specific cytolytic activities against irrelevant virus-infected or transformed target cells. Similarly, four diLrt;~ CMV-specific CD8+ cytotoxic T cell clones retained the ability to kill CMV-infected cells, and did not acquire non-specific cytolytic activities against irrelevant virus-infected or transformed target cells. These char~cteri~tics were also applicable to CD4+ helper T
cells. Thus, antigen-specific CD4+ T cells propagated using REM retained the ability to proliferate in response to the ~I,plo~,liate viral antigens and a~l.~pliale antigen-presenting cells (APC). Furthermore, antigen-specific T cells cultured under REM were also capable of ent~ring a qni~scent non-dividing phase of the cell cycle; and were capable of . . ,~i . ,; . .~ viable for at least 4 weeks in vitro. Thus, aliquots of T cells can be removed from the cultures at the end of a stim~ tiQn cycle (generally day 12-14), and placed in a 2~ eulture vessel with a roughly equal number of irradiated PBMC (without anti-CD3 mAb, antigen or IL-2).
The ~ 1iti~)n of irradiated PBMC as feeder cells during storage of expanded populations improved the ability of the T cells to enter a resting phase and to remain viable. Preferably, the ratio of PBMC feeder cells to resting T cells during storage is at least about 2:1. Without the addition of PBMC feeder cells, viability of the T cells generally drops ~ignifi(~.~ntly (typically to levels of about 10% or less).

W 097/32970 PCTrUS97/03293 As described in Riddell et al., supra, T cells e~p~n~lçcl by REM assumed a smalIround morphology and 60-95% r~m~ine~l viable by trypan blue dye exclusion even after 28 days in culture. T cells propagated by hp-REM also entered a resting phase upon IL-2 withdrawal; and they did not undergo prograrnmed cell death (i.e. apoptosis) upon restimulation via the antigen-specific T cell receptor. Upon restim~ tion (e.g. with anti-CD3 mAb or antigen), the T cells reacquired responsiveness to IL-2, and can enter the S
and G2 phases of the cell cycle and increased in cell number. Such characteristics are believed to be important for in vivo survival of the cells and for the efficacy of cellular immunotherapy. In contrast, certain previously-described methods for the propagation of T cells have been reported to cause apoptotic cell death in a proportion of cells after cytokine withdrawal or T cell receptor r~ctim~ tion (see, e.g., Boehrne SA and Lenardo MJ, F.llr. J. Imm-mol., 23:1552-1560, 1992).
There are a number of different ch.;. ,f . ,~ . ,ces in which the introduction of functional genes into T cells to be used in immunotherapy may be desirable. For ç~mple, the introduced gene or genes may improve the efficacy of therapy by promoting the viability and/or function of transferred T cells, or they may provide a genetic marker to permit selection and/or evaluation of in vivo survival or migration; or they may incorporate fimrtilm~ that improve the safety of immllnl~therapy, for example, by making the cell ~.lscc;~Lible to negative selection in vivo as described by Lupton S.D. et al., Mol. ~ncl Cell ~iQL, 11:6 (1991); and Riddell et al., Hurnan Gene Therapy 3:319-338 (1992); see also the publications of WO/92 08796 and WO/94 28143 by Lupton et al., describing the use o~
bifunctional selectable fusion genes derived from fusing a dominant positive selectable marker with a negative selectable marker.
Various infection techniques have been developed which utilize recombinant infectious virus particles for gene delivery. This represe~ a ~ lly prefci.led approach to the transduction of T lymphocytes of the present invention. The viral vectors which hav~ been used in this way include virus vectors derived from simian virus 40 (SV40) (see, e.g., Karlsson et al., Proc. Natl. Acad. Sci. USA 84 82:158, 1985); adenoviruses (see, e.g., Karlsson et al., EMBO J. 5:2377, 1986); adeno-associated virus (AAV) (see, e.g., B.J.
Carter, Current Opinion in Biotechnology 1992, 3:533-539), and retroviruses (see, e.g., Coffin, 1985, pp. 17-71 in Weiss et al. (eds.), RNA Tumor Viruses, 2nd ed., Vol. 2, Cold Spring Harbor Laboratory, New York). Thus, gene transfer and e~l~ssion methods are WO 9~1132970 PCT/US97/03293 n~merous but essentially function to introduce and express genetic material in m:~mm~ n cells. A number of the above techniques have been used to transduce hematopoietic or lyrnphoid cells, including calcium phosphate transfection (see, e.g., Berman et al., supra, 1984); protoplast fusion (see, e.g., Deans et al., supra, 1984); electroporation (see, e.g., S Cann et al., Oncogene 3:123, 1988); and infection with recombinant adenovirus (see, e.g., Karlsson et al., supra, Reuther et al., Mol. Cell. Biol. 6:123, 1986); adeno-associated virus (see, e.g., LaFace et al., supra); and retrovirus vectors (see e.g., Overell et al., Oncogene 4: 1425, 1989). Primary T lymphocytes have been s~cce~.cfully trSm~ re~l by electroporation (see, e.g., Cann et al., supra, 1988) and by retroviral infection (see e.g., Ni~hih~r?. et al., Cancer Res. 48:4730, 1988, Kasid et al., supra, 1990; and Riddell, S. et al., Human Gene Therapy 3:319-338, 1992).
Retroviral vectors provide a highly efficient method for gene transfer into eukaryotic cells. Moreover, L~l~ovil~l integration takes place in a controlled fashion and results in the stable integration of one or a few copies of the new genetic inforrnation per cell.
Retroviruses are a class of viruses which replicate using a virus-encoded, RNA-d~ ;d DNA polymerase, or reverse ~ s~ se, to replicate a viral RNA genome to provide a double-stranded DNA intermediate which is incorporated into chromosomal DNA of an avian or m~mm~ n host cell. Most lc;LLovil~l vectors are derived from murine retroviruses. Retroviruses adaptable for use in accordance with the present invention can, however, be derived fi~om any avian or m~nnm~ n cell source. These retroviruses are p~referably amphotropic, me~nin~ that they are capable of infecting host cells of several species, including hllm~n~ A t~h~r~rt~ristic feature of ~ ovhal genomes (and retroviral vectors used as described herein) is the retroviral long t~rmin~l repeat, or LTR, which is an untr~ncl~t~l region of about 600 base pairs found in slightly variant forms at the 5' and 3' ends of the retroviral genome. When incorporated into DNA as a provirus, the retroviral I,TR includes a short direct repeat sequence at each end and signals for initiation of transcription by RNA polymerase II and 3' cleavage and polyadenylation of RNA
transcripts. The LTR contains all other cis-acting sequences n~ce~ry for viral replication.
A "provirus" refers to the DNA reverse transcript of a retrovirus which is stably integrated into chromosomal DNA in a suitable host cell, or a cloned copy thereof, or a PCTrUS97/03293 cloned copy of unintegrated intermediate forrns of retroviral DNA. Forward transcription of the provirus and assembly into infectious virus occurs in the presence of an al)prupl;ate helper virus or in a cell line cont~ining d~ )pliate sequences enabling encapsidation without coincident production of a cont~min~ting helper virus. Mann et al. (Cell 33:153, 1983) describe the development of cell lines (e.g., ~2) which can be used to produce helper-free stocks of recombinant retrovirus. These cells lines contain integrated ~ ovildl genomes which lack sequences reguired in cis for encapsidation, but which provide all necess~u~y gene product in trans to produce intact virions. The RNA transcribed from the integrated mutant provirus cannot itself be packaged, but these cells can enczipsi~l~te RNA
transcribed from a recombinant retrovirus introduced into the same cell. The resulting virus particles are infectious, but replication-defective, rendering them useful vectors which are unable to produce infectious virus following intro~ ctil n into a cell lacking the complement~ry genetic information enabling encapsidation. Encapsidation in a cell line harboring Irans-acting elements encoding an ecokopic viral envelope (e.g., 'Y2) provides ecotropic (limited host range) progeny virus. Alternatively, assembly in a cell line c~ nt~ining amphotropic ps~rk~ging genes (e.g., PA3 17, ATCC CRL 9078; Miller and Buttimore, Mol. Cell. Biol. 6:2895, 1986) provides a~ Lll~ic (broad host range) progeny virus. Such packing cell lines provide the n~cç~ry retroviral gag, pol and env proteins in ~rans. This ~LLdle~y results in the production of r~ vildl particles which are highly infectious for m~mm~ n cells, while being incapable of further replication after they have integrated into the genome of the target cell. The product of the env gene is responsible for the binding of the retrovirus to viral receptors on the surface of the target cell and therefore flet~rrnin~s the host range ofthe ~ vil.ls. The PA 317 cells produce vvhdl particles with an amphotropic envelope protein, which can tr~n~ çe cells of human and other species origin. Other pz~ ging cell lines produce particles withecotropic envelope proteins, which are able to transduce only mouse and rat cells.
Numerous rc;llovildl vector constructs have been used successfully to express many foreign genes (see, e.g., Coffin, in Weiss et al. (eds. ), RNA Tumor Viruses, 2nd ed., vol. 2 (Cold Spring Harbor Laboratory, New York, 1985, pp. 17-71). Retroviral vectors with inserted sequences are generally functional, and few sequences that are con~i~t~n~ly inhibitory for lelrovildl infection have been identified. Functional polyadenylation motifs inhibit le1lovildl replication by blocking retroviral RNA synthesis, and there is an upper WO 97/32970 P~T/IJS97l03293 size limit of approximately 11 kb of sequence which can be packaged into retroviral particles (Coffin, supra, 1985); however, the presence of multiple internal promoters, initially thought to be problematic (Coffin, supra, 1985), was found to be well tolerated in several letlovildl constructs (Overell et al., Mol. Cell. Biol. 8:1803, 1983).
S Retroviral vectors have been used as genetic tags by several groups to follow the - development of murine hematopoietic stem cells which have been kansduced in vi~ro with retrovirus vectors and transplanted into recipient mice (Williams et al., Nature 310:476, 1984; Dick et al., Cell 42:71, 1985, Keller et al., Nature 318:149, 1985). These studies have demonstrated tha~ the infected hematopoietic cells reconstitute ~e hematopoietic and lymphoid tissue of the recipient animals and that the cells display a normal developmental potential in vivo. The marked cells can be vi~ i7P~ using any of a number of molecular biological techniques which can demonstrate the presence of the retroviral vector sequences, most notably Southern analysis and PCR (polymerase chain reaction). The ability to mark cells genetically using leLlovildl vectors is also useful in clinical settings in which the technique can be used to track grafts of autologous cells. This approach has already been used to track TILs (tumLor-infiltrating lymphocytes) in patients given TIL
therapy for tPrrnin~l cancer trç~mPnt by Rosenberg et al. (N. Engl. J. Med. 323:570, 1990). The transduction of these cells with the marker gene was not associated with in vitro cellular dysfunction (Kasid et al., Proc. Natl. Acad. Sci. USA 87:473, 1990).
Many gene products have been ~ essed in retroviral vectors. This can either be achieved by placing the sequences to be ~ ssed under the tran~crir~tional control of the p~romoter incorporated in the retroviral LTR, or by placing them under the control of a heterologous promoter inserted between the LTRs. The latter strategy provides a way of co~ es~ g a dominant selectable marker gene in the vector, thus allowing selection of cells which are expressing specific vector sequences.
It is conte-llplated that ov~ ession of a stim~ tory factor (for example, a lymphokine or a cytokine) may be toxic to the treated individual. Therefore, it is within the scope of the invention to include gene segment~ that cause the T cells of the invention to be susceptible to negative selection in vivo. By "negative selection" is meant that the - 30 infused cell can be elimin~te~l as a result of a change in the in vivo condition of the individual. The negative selectable phenotype may result from the insertion of a gene that confers st;l~ilivily to an ~-lmini~tered agent, for example, a compound. Negative W O 97/32970 PCT~US97/03293 selectable genes are known in the art, and include, inter alia the following: the Herpes simplex virus type I thymidine kinase (HSV-I TK) gene (Wigler et al., Cell 1 1:223, 1977) which confers ganciclovir sensitivity; the cellular hypo~nthine phosphribosyltransferase (HPRT) gene, the cellular adenine phosphoribosyltransferase (APRT) gene, bacterial S cytosine cle~min~ç, (Mullenetal., Proc. Natl. Acad. Sci. USA. 89:33 (1992)).
~n addition, it is u~seful to include in the T cells a positive marker that enables the selection of cells of the negative selectable phenotype in vitro. The positive selectable marker may be a gene which, upon being introduced into the host cell expresses adominant phenotype p~ lilLing positive selection of cells carrying the gene. Genes of this type are known in the art, and include, inter alia, hy~ ycin-B phosphotransferase gene (hph) which confers resistance to hy~,lullly~;hl B, the aminoglycoside phosphotransferase gene (neo or aph) from Tn5 which codes for resistance to the antibiotic G418, the dihydlofolate retlllr,t~e (DHFR) gene, the adenosine cle~ e gene (ADA), and the multi-drug rçsi~t~nce (MDR) gene.
Preferably, the positive selectable marker and the negative selectable element are linked such that loss of the negative selectable çlem~-nt necçss~rily also is accompanied by loss of the positive selectable marker. Even more preferably, the positive and negative selectable markers are fused so that loss of one obligatorily leads to loss of the other. An example of a fused polynucleotide thal yields as an t;~les~ion product a polypeptide that confers both the desired positive and negative selection ÇeaLules described above is a hy~rolllychl phosphotransferase thymidine kinase fusion gene (HyTK). Expression of this gene yields a polypeptide that confers hy~rolllycin B resistance for positive selection in vitro, and ganciclovir sen~ilivily for negative selection in vivo. See Lupton S.D., et ~L, Mol. andCell. Biolo,~ 3374-3378, 1991. Inaddition,inpreferredembo~liment~,the polynucleotides of the invention encoding the chimeric receptors are in retroviral vectors con~ining the fused gene, particularly those that confer hyglolllycill B reci~t~nce for positive selection in vitro, and ganciclovir sellsi~ivily for negative selection in vivo, for example the HyTK retroviral vector described in Lupton, S.D. et al. (1991), supra. See also the publications of PCT/US91/08442 and PCT/US94/05601, by S.D. Lupton, describing the use of bifunctional selectable fusion genes derived from fusing a ~lomin~nt positive selectable markers with negative selectable m~rker.~.
3~

W ~9'7132970 P~T~U~97J03293 Preferred positive selectable markers are derived from genes selected from the group consisting of hph, neo, and ~pt, and plGf~,led negative selectable markers are derived from genes selected from the group c~n~icting of cytosine cle~min~e, HSV-I TK, ~IZV TK, HPRT, ~PRT and ~pt. Especially plGf~ d markers are bifunctional selectable S fusion genes wherein the positive selectable marker is derived from hph or neo, and the ~ negative selectable marker is derived from cytosine de~m;n~e or a TK gene.
A variety of methods can be employed for transducing T Iymphocytes, as is well known in the art. Typically, one can carry out retroviral transductions as follows: on day 1 after stim~ tion using REM as described herein, one can provide the cells with 20-30 units/ml IL-2; on day 1, 2, or 3, one half of the m~ m can be replaced with retroviral supern~t~nt prepared according to standard methods and the cultures supplemented with 5 g/ml polybrene and 20-30 unitslml IL-2; on day 4, the cells are washed and placed in fresh culture medium supplement~c~ with 20-30 units/ml IL-2, on day 5, the exposure to retrovirus can be repeated; on day 6, the cells can be placed in selective medium (c~ ing, e.g., an antibiotic corresponding to an antibiotic resistance gene provided in the lG~Iuvildl vector) supplemente~l with 30 unitslml IL-2; on day 13, viable cells can be sep~ Gd from dead cells using Ficoll EIypaque density gradient separation and then the viable cells can be subcloned using REM.
Using an antigen-specific CTLs (5~1, an EBV-specific CD8+ clonal line) and a le~ovildl vector (LAPSN, Clowes et al., 1994, J. Clin. Invest. 93:644 (which allowed for monitoring of ~lk~line phosphatase ~ es~ion by flow cytometry)), high transduction frequencies can be achieved when the cells are exposed to vector on day 1, 2 or 3 after initiation of REM.
As described above, T cells l,l~al~ed according to the invention can be used to restore,enhance,and/ormodulatei.. ~ yinrecipientindividuals. By";~n~ y"is meant a lee~nin~ of one or more physical symptoms associated with a response to infection by a pathogen, or to a tumor, to which the lymphocyte response is directed. The amount of cells ~1mini~tered is usually in the range present in normal individuals with ily to the pathogen. Thus, CD8+ CD4- cells are usually a~lmini~t~red by infusion, v/ith each infusion in a range of at least 1 o6 to 1 ol~ cells/m2, preferably in the range of at least 107 to 109 cells/m2. The clones may be ~lmini~tf~red by a single infusion, or by multiple infusions over a range of time. However, since dirrelG~ll individuals are expected ~1 W O 97/32970 PCTrUS97/03293 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 ~tten(1inF physician, and can be determined by routine e~min~tion. The generation of sufficient levels of T lymphocytes (including cytotoxic T lymphocytes and/or helper T
lymphocytes) is readily achievable using the rapid expansion method of the present invention, as exemplified herein.
It has also been observed that T cells (~xp~n~le~l using REM exhibited very highlevels of tr~n~dl~ction using vectors such as retroviral vectors which will be of great use in the contexts of cellular immlm~therapy and gene therapy using lymphocytes.
The examples in Riddell et al., supra, exemplify the basic REM protocol (i.e. hp-REM), and also help to illustrate the general applications of REM technology to the preparation and use of e~p~n~(l T cell populations and, in that regard, exemplify techniques and principles that can also be applied in the context of modified-REM.
The exarnples below illustrate e~emrl~ry mo-lific~ti--n~ of the REM technology according to the present invention (i.e. modified-REM), to enable a reduction orelimin~tion of the PBMC and/or EBV-LCL feeder cells that are ch~r~ct.ori~tic of the hp-REM protocol.
All of the examples presented below are provided as a further guide to the practitioner of oldi~l~ y skill in the art, and are not to be construed as limiting the invention in any way.

W O 97~3297~ PCT~US97/03293 ~X ~MP~ , ple 1 T~e Contribution of Monocyte Fc-~ Receptors in 12~id F.~ n~ion The PBMC feeder cells, which are used in large excess to drive hp-REM, are a ~ heterogeneous population of cells including B lymphocytes, T lymphocytes, monocytes, macrophages, and granulocytes, and Natural Killer ("NK") cells.
One of the activities believed to be supplied by PBMCs in the hp-REM protocol isthe provision of Fc-~ receptors which can bind to the Fc portion of IgG antibodymolecules. In particular, it is believed that T cell activation in the hp-REM protocol can be mediated by binding of Fc- y receptors ("Fc~R") on monocytes within the PBMC
population to the Fc portion of anti-CD3 antibody (e.g. OKT3), which can thereby be "presented" by the monocytes to T cells within the population to be e~p~nclerl Following such activation, T cells are believed to be capable of initi~tin~ an "autocrine" growth st;m~ tory cycle in which activated T cells both secrete growth-stimlll~tory cytokines and also increase the ~ ,ei,~ion of cell surface receptors for such cytokines. Supplying Fc-~
receptors, or otherwise ~rr~ v~ly pl'eS~~ g anti-CD3 antibody, is thus believed to function in the initi~tion and promotion of T cell expansion.
Col, 1~, ...il~ and 4u~lLiryillg the contribution of monocyte Fc-~ receptors in the hp-REM protocol can be accomplished by depleting monocytes from the PBMC feeder cells.
Peripheral blood mononuclear cells can be obtained from any of a variety of sources, as described above. For the following examples, buffy coat layers (derived from healthy human donors) were obtained from a Red Cross blood bank. PBMCs were isolated using a Ficoll gradient, washed and stored in cell culture media (at 4 degrees Celsius) using standard techniques as referred to above.
A variety of techniques can be used for se~aldlillg out various cell types from a mixed population such as PBMCs. By way of illustration, depletion of the monocyte/macrophage population was performed using sephadex C3-10 chromatography(see, e.g., Section 3.6 in "Current Protocols in Tmmllnology" (Wiley Interscience, 1992)).
~Ionocyte depletion was monitored by flow cytometry following staining with FITC-conjugated anti-CD14 monoclonal antibody (available from, e.g., PharmaGen). CD14 33 W 097/32970 PCTrUS97/03293 ~;;x~ e~.sion before depletion was about 7.4% of total cells. Following depletion, it was about 1.5%.
In order to assess the impact of depleting monocytes on the ability of PBMCs to promote rapid expansion, a standard hp-REM protocol was used and monocyte-depleted S PBMCs were compared with non-depleted PBMCs.
For purposes of illustration, a CTL line ((1e~i~n~te-l "27EB") was ~ paLed by procedures analogous to those described above. Briefly, PBMCs were obtained from an individual blood sample and were cultured with EBV-LCL derived from the same individual. After two weeks, CD8+ CTLs were isolated by "panning" with a flaslc coated with anti-CD8 antibodies (e.g. the AIS-CD8+ "CELLector flask" from Applied Immune Sciences).
For all of these illustrative examples, PBMCs were gamma-irradiated at 3600 rads(using a Cs-137 source) and EBV-LCL were irradiated at 10,000 rads.
Cultures were generally m~int~inPd as described above for hp-REM except that 10% fetal calf serum was used in place of human serum; and IL-2 was used at 25 units/ml and was generally first added on "day 0" (as opposed to 1 day after culture initi~ti~n), and then at 3-5 day intervals (generally, on day 5, and then again on day 8), as otherwise described above for the hp-REM protocol. OKT3 was generally used at about 10 ng/ml.
Cells were typically harvested and quantified after 14 days of culture.
For this example, cultures were established with 5x104 CTL ("27EB", as describedabove), SX1O6 irradiated EBV-LCL (i.e. a 100:1 excess over CTLs) (~.~ared and irradiated as described above), 10 ng/ml OKT3, 25 U/ml IL-2 and 2.5x107 irradiated PBMC (i.e. a 500:1 excess over CTLs) (either monocyte-depleted or nondepleted, pl~ cd and irradiated as described above). Typical control cultures would include, for example, cultures without any added CTL. After 14 days of culture, cells were h~ ~ lt;d and quantified.
In the case of the standard hp-REM protocol using nondepleted PBMCs, a~ ,xi.l.ately 4.86x107 T cells were recovered, repr~s~nting an expansion of approximately 882-fold.
As shown in TABLE 1, when monocyte-depleted PBMCs were used instead, only about 2.7x107 T cells were recovered, indicating that the Pxp~n~jon rate had dropped to about 55% of ~e control rate. ~3 y W 09~132970 PCTnUS97/03293 TABL~ 1 PBMC Cells Method of depletion T Cell % Control T cell Recovery FxpPn~ion Nondepleted PBMC none 4.86xlQ7 100%
Monocyte-depleted PBMC Seph~ column 2.70x~07 55%

The above results provided fur~er indication that monocytes within the PBMC
population ~ uelltly conkibute significantly to the ability of PBMCs to bring about the rapid expansion of T cells. In the following example, the ability to provide FcrR activity or its equivalent from a source other than PBMCs as a means for reducing the dependence of REM on large excesses of PBMC iEeeder cells.
~le 2 Replacçm~nt of MQ ~ le Fc-ry Receptor ~ctivity in Mo(lified-Rh~
As discussed above, Fc- y receptors found on monocytes are believed to be responsible for a ~i~nific~nt portion of the stim~ tory activity supplied by PBMCs in the presence of antibodies such as anti-CD3 antibody (e.g. OKT3).
Having identified a stim~ tory component supplied by the heterogeneous PBMC
population, it is possible to reduce the depçn-lence on PBMCs themselves by providing that activity (or its equivalent) from another source. Preferably, for use in modified-REM
as described herein, the i~ntifiecl stimlll~tory activity will be provided by a m~mm~ n cell line or as a non-cellular additive to the REM culture. Tll~ 1; ve examples of both are provided below.

a. Use of In~mm~ n cell lines jn rnodified-REM
Cell lines t;x~res~ing one of more identified T cell stimlll~tc)ry activities provided by PBMCs (or LCL) can be effectively used to reduce the dependence of REM on such PBMC (or LCL) feeder cells. Where such a cell line is to be incorporated into the protocol, it is ~lerclable~ as described above, that the cell line not be a potential source of 3~-W 097/32970 PCT~US97/03293 adventitious agents such as viruses. Accordingly, the supplemental cell line used is preferably not an EBV-kansformed cell line (such as EBV-LCL). Also, for the rapid expansion of human T cells, it is generally preferable to use a cell line derived from a higher m~mm~l, especially a primate, most preferably a human.
S ~mm ~ n cell lines expressing Fc-~ receptors have been described in the lil~,aLule and can be obtained from a variety of sources. For example, a number of human tumor lines have been demonstrated to express Fc-~ receptors (see, e.g., R.J. Looney et al., J.Immunol.136:1641 (1986) (describing K562, an erythrolellkemi~ cell line); S.J.
Collins at el. 1977. Nature 270:347 (1977) (describing HL60, a promyelocytic cell line);
C. Lozzio and B. Lozzio, Blood 45:321 (1975) (describing U937, a histiocytic lymphoma cell line); and G.R. Crabtree et al., Cancer Res. 38:4268 (1979).
Cell lines expressing Fc- y receptors can also be readily prepared using standard molecular biological techniques. By way of illustration, Fcl~R-positive cell lines can be obtained by immortalizing cells that already express Fc- y receptors using any of a variety of well-known techniques for transforming m~mm~ n cells. Altelll~iv-ely, an existing cell line such as a human cell line can be genetically modified to express Fc-~ receptors by introducing genes encoding Fc-~ lc;ctplol~ into the cells. Thus, a cell line of choice, such as a hurnan cell line already ~lw~ g a stimulatory colllpollent such as a cytokine or a cell adhesion-accessory molecule (both of which are discussed below), can be further modified by introduction of genes encoding an Fc~yR.
By way of illustration, human monocytes ~ ly express two distinct Fc-~
ec~lor types ("Fc~RI" and "Fc~RII") which differ in their affinity for IgG antibody bin~ling, (see, e.g., Ravetch and Kinet Annu. Review of Tmmunol. 9:457-492, 1991). In particular, Fc yRI is generally a high affinity ~ec~ol (Ka=108 - 109 ) while Fc~RII is generally a lower affinity receptor (Ka=107). The genes for both Fc~RI and FcrRII have been identified and cloned. Previous studies have demonsl~lt;d that fibroblasts expressing the Fc~RlI receptor following gene transfer could ~rre~;lively restore anti-CD3-dependent proliferation of monocyte-depleted T lymphocyte cultures (see, e.g., Peltz et al., J.
~mmlmol. 141:1891 (1988)). Thus, a cell line genetically modified to express FcyR should be capable of supplying a significant portion of the stimulatory activity supplied by PBMCs in the hp-REM protocol. Use of such a cell line, potentially in conjunction with WO 9~J32970 PCT/US97/03293 other components such as cytokines or adhesion-accessory molecules as described below, would thereby enable a decrease in the nurnber of PBMCs required for rapid expansion.

b. Use of non-cellular additives for modified-RFM
In addition to providing T cell stimulatory components by way of a cell line, such as described above, it will be possible to provide a number of components (or their functional equivalents) as non-cellular additives to the modified-REM culture mediurn.
l hus, as a different alternative to the Fc yR activity a~a~ lly contributed by PBMC
rnonocytes, it will be possible to provide a substitute or structural equivalent for Fc~R
activity. For example, an ~lt~rn~tive means for achieving "presentation" of antibodies such as anti-CD3 antibodies to T cells is to conjugate such antibodies to beads (such as seph~e~ beads or m~gn~tic beads).
In order to assess the ability of such bead-conjugated antibodies to substitute for soluble antibodies (which are presumably presented via Fc~R), we con~l~cte~ ~;;x~ iments to ~lettormine whether anti-CD3-conjugated m~gnPti(~ beads can effectively replace soluble anti-CD3 monoclonal antibodies in the Rl~M protocol.
Anti-CD3-conjugated m~netic beads ("BioMag anti-CD3") were obtained from Pelc~L;ve Diagnostics. The particles used were a~ xh~lately 1 jlm in size and had covalently ~tt~h~rl anti-CD3 monoclonal antibodies loaded at a~ illlately 20 ,ugantibody/lxlO7 beads. A range of beads was chosen to a~p.vxilllate the number ofantigen ~les~ g cells ("APCs") çstim~tP~1 to be present within the PBMC population used in hp-REM.
For 4u~liryh~g the actual impact on hp-REM, T cell expansion cultures c~ ;..i.-g5x104 CTL, 5X1O6 irr~ t~d EBV-LCL, 2.5x107 irr~ ted allogeneic PBMC and 25 U/ml IL-2 were established, ec~t-.nti~lly as described above in Example 1. Either 10 ng/ml of soluble anti-CD3 antibody (OKT3) or various quantities of anti-CD3-conjugated beads w~ e added as the T cell activation reagent. Cell cultures were e~r~ntle(l and T cells counted, ess~nti~lly as described for Example 1.
The results are shown in TABLE 2. While T cell expansion using anti-CD3-con~ugated beads was somewhat less than with soluble OKT3 (in the range of about 80%), the results suggest that antibody-coated beads would be capable of in(~ inp substantial levels of T cell activatiorl/proliferation within a rnodified REM protocol. More detailed W O 97/32970 PCT~US97/03293 quantification of the relative role of anti-CD3 presentation as compared to other activities potentially provided by APCs within the PBMC population can be readily obtained by assessing the expansion rates obtainable with anti-CD3 beads using APC-depleted cultures, either in the presence or absence of various cytokines or other soluble stimulatory factors (which are described in more detail below).

Anti-CD3 source T Cell Recovely T Cell F.~p~n~ion Soluble anti-CD3 Ab (OKT3) 6.52x107 1304-fold lx107 BioMag anti-CD3 5.36x107 1071-fold 5X106 BioMag anti-CD3 5.0x107 1000-fold 2.5x106 BioMag anti-CD3 5.25x107 1051-fold Comp~ri~ons of the relative eff1ciency of providing various PBMC repl~emcllt co~ onents, as described herein, can be readily achieved by standard titration analyses in which the various components are added back at varying concentrations to a PBMC-limited REM culture (i.e. a culture in which PBMCs are included at a sub-optimal level).
By way of illustration, experim~nt~ described below sl~esse~l the impact of adding various combinations of exogenous cytokines to sub-optimi7.ocl hp-REM cultures in which the PBMC population had been reduced to one-half or one-quarter of an optimal starting level.
Analogous assays can be readily performed for other components such as cells ex~l~ ssillg Fc~R or a&esion-accessory components, anti-CD3-conjugated beads, and/or other soluble stim~ tQry factors (such as monoclonal antibodies directed to T cell surface co~ )onents), as described in more detail below.

n~le 3 The Contribu~ion of lB Ce~l~ jn Rapid ~ n~ior In order to 4u~lliry the contribution of B lymphocytes to the stim~ supplied by PBMCs, we ~ mined the relative ability of B-cell-depleted PBMC populations to support REM. 3 8 W~ 9 7l32970 PCTIIJS97/03293 Isolation of cells such as B cells (or other cells referred to herein) can be conveniently achieved using antibodies directed to a cell surface marker known to be present on the cells to be depleted. A variety of such markers are well known, including fhe various "CD'7 or "cluster of di~ferentiation" markers; and antibodies to many such markers are readily obtainable. Also, for many such markers, beads conjugated with the ~ antibodies are readily available and can greatly facilitate cell separation.
By way of illustration, CD19 is a well-known cell surface marker for B
lymphocytes. Magnetic'beads that had been conjugated with anti-CD19 antibodies were obtained from Dynal, and were used to deplete a PBMC population of B cells, following standard procedures as described by the mi mlf~cturer.
Depletion was evaluated by fluorescence activated cell sorting ("FACS") after CDl9 st~-ining The PBMC population was estimated to contain approximately 12% B
cells prior to depletion and less than 1% B cells after depletion.
For testing the impact of B cell depletion on hp-REM, T cell expansion cultures CO~ it~ 5x104 CTL, 5x106 irradiated EBV-LCL, 2.5x107 irradiated allogeneic PBMC
(B-cell-depleted or nondepleted), 10 ng/ml OKT3 and 25 U/ml IL-2 were established; and, after 14 days, T cells were harvested and quantified as described above. The results, shown in TABLE 3, suggested that B cells also contribute to the ~tim~ ting activity supplied by PBMCs.

PBMC Cells Method of Depletion T Cell % Control T Cell Recu~ r--Nondepleted PBMC none 4.28x107 100%
B-cell-depleted PBMC Anti-CD19 Magnetic Beads 9.8x106 23%

The decreased levels of T cell expansion in this and fhe preceding ~ r,l ,,..~nf-~
suggests a role for monocytes and B cells as antigen pl~S~ ; l Ig cells ("APCs") in the hp-REM protocol. (The inability to inhibit T cell expansion to the levels observed in mple 1 to the levels observed in this experiment may be a result of differences in cell depletion by the various methods used and/or the presence of small numbers of APC~

PCTrUS97103293 undetected by the assays used. It should also be noted that monocyte depletion as measured in Example 1 only reduced the monocyte population from about 7.5% to about 1.5%.) There are a number of other well-known techniques that can be used to deplete various cell types from the PBMC population and that can therefore be used to provide additional confirm~tion and quantification of the results described herein. Thus, for example, nylon wool can be used to remove both monocytes and B cells (as well as any fibroblasts) from the PBMC population. The replacernent of various APC activities in modified-REM is further described in the following example.
~mple 4 Replacement of Various APC Activities in Modified-l~F.l~[
The results obtained in the B-cell-depletion and monocyte-depletion experiments,described above, indicated that putative APCs in the PBMC population appear to contribute to the stimulus supplied by PBMCs. As described in Examples 1-2, the role of Fc~R activity in presentation of anti-CD3 antibody is expected to account for some portion of the activity provided by l~ulalive APCs. Such Fc~R activity can be supplied by another (non-PBMC) source, e.g. a cell line ~ S~illg Fc~R or anti-CD3-conjugated beads, as also described above.
It is believed that APCs within the PBMC population also contribute adhesion-~cç~-)ry molecules and stim~ tQry cytokines that would be ç~recte~l to further enhzlnc.e the activation/proliferation process. (Furthermore, as described below, T lymphocytes within the PBMC population are also expected to produce st;mlll~t~-ry cytokines as a re:sult of activation via the anti-CD3 antibody.) The roles of such adhesion-~cceesory molecules and cytokines are described in more detail in the examples below.

E~La~
The Contrihuffon of Cytokines in REM
Although anti-CD3 antibody (e.g. OKT3) is used to activate and induce the proliferation of T cell clones for their in vitro expansion, the ~-irradiated feeder PBMC
population also contains a substantial population of T lymphocytes that are believed to be activatable by the anti-CD3 antibody. While Lsuch irradiated feeder cells are inc~rable of W 097132970 PCTnUS97103293 dividing, their activation via anti-CD3 antibody is believed to result in the secretion of multiple cytokines which can provide additional Iympho-proliferative signals. For example, in addition to IL-2, anti-CD3 activation of T cells is believed to result in the secretion of other stimul~tory cytokines including IL-l a and ,B, IL-6, IL-8, GM-CSF, S IFN-a and TNFa and ,B (du Moulin et. al. 1994. Cytotechn- logy 15:365). It is believed ~ that the secretion of one or more of those cytokines can contribute substantially to the prolif~ldli~e stimulus provided by PBMCs within the hp-REM protocol. Of the nurnerous other cytokines that have been characterized, a number of these are known to stim~ t~- the growth of T cells, including, for example, IL-7 and IL-l 5. Others can be readily screened for their ability to enhance T cell proliferation and for their relative ability to reduce the dependence of REM on large numbers of PBMCs, as described herein.
By way of illustration, we analyzed the ability of a number of exogenously-supplied cytokines to reconstitute T cell expansion in REM cultures in which the numbers of PBMC feeder cells had been reduced to sub-optimal levels, in order to 4u~lliry the potential role of such cytokines in promoting REM.
Following procedures e~sl~nti~lly analogous to those described above, cultures cont~inin~ sx104 CTL, 5X106 EBV-LCL, 10 ng/ml OKT3 and 25 U/ml IL-2 were established with either 2.5x107 irr~ ted PBMC (100% control), 1.25x107 irr~ t~
PBMC (50%), or 6.12xlo6 irradiated PBMC ~25%).
It is believed that a number of cytokines can act synergistically with IL-2 to promote T cell proliferation. In this illustrative experiment, the following exogenous cytokines were added to the cultures either alone or in various combinations as described:
IL-l (40 U/ml), IL-4 (200 U/rnl), IL-6 (500 U/ml) and IL-12 (20 U/ml).
The results, shown in TABLE 4, confirm~d that such cytokines can subst~nti~lly enhance T cell expansion when PBMC populations are reduced to sub-optimal levels. It is not unexpected that expansion levels were not returned to that observed with the optimal nurnber of PBMC feeders, because the PBMC population is believed to supply additional sfim~ tory activities as described herein. The data suggest that replacement of IL-4 with IL-12 in a cytokine cocktail may further enh~nee proliferation. The properties, sources, and DNA and protein sequences of many such cytokines are described in cytokine reference books such as "The Cytokine Facts Book" by R. Callard et al., supra. To take a single example for purposes of illustration, IL4-12 is known to be a heterodimeric cytokine W 097/32970 PCTrUS97/03293 comprising two peptide chains (p35 and p40) that induces IFNr production by T
lymphocytes and co-stimulates the proliferation of peripheral blood lymphocytes. IL-12 also stim~ tes proliferation and differentiation of TH 1 T lymphocytes, and is known to be produced by B cells, monocytes/macrophages, and B lymphoblastoid cells. The complete amino acid sequences for both the p35 and p40 chains are known and available on Genbank (Accession numbers provided in Callard). The IL-12 l~c~lor has also beencharacterized (id.).
Additional cytokines and cocktails thereof can readily be tested in an analogous marmer, and a comr~ri~on of stim~ tory cocktails can then be made using even lower levels of PBMCs.

Added Cytokine(s) PBMC Cell Recovery Fxpsn~ion Percent Control IL-2 2.5x107 4.4x107 882-fold 100%
IL-2 1.25x107 2.8x107 564-fold 64%
IL-2 6.12x106 1.8x107 360-fold 41%
IL-2 + IL-1 1.25x107 3.2x107 648-fold 73%
IL-2 + IL-l 6.12x106 2.4x107 486-fold 55%
IL-2 + IL-4 1.25x107 2.0x107 402-fold 46%
IL-2 + IL-4 6.12x106 1.5x107 295-fold 33%

IL-2 + IL-6 1.25x107 3.2x107 636-fold 72%
IL-2 + IL-6 6.12x106 3.0x107 606-fold 69%
IL-2 + IL-12 1.25x107 4.65x107 930-fold 105%
IL-2 + IL-12 6.12x106 2.88x107 558-fold 63%
IL-2 +IL-l + 1.25x107 3.8x107 768-fold 87%
IL-4 + IL-6 IL-2+IL-1 + 6.12x106 3.1x107 618-fold 70%
IL-4 + IL-6 W 097132970 PCTnUS97JO3293 Further evidence that soluble components of the feeder cell sup~rn~t~n~ can provide an effective stimulus for low-PBMC REM was obtained by reducing the PBMCpopulation to sub-optimal levels and using a REM supern~t~nt to provide soluble ~tim~ tQry signals.
Briefly, a standard hp-REM protocol was perforrned as described above, using an anligen-specific CTL clone and p~lro~ g a 48-hour REM expansion with PBMC
(500:1), EBV-LCB (100:1), anti-CD3 antibody (10 ng/ml), and recombinant human IL-2 (25 units/ml). After 48 hours, the cells were harvested and the supernatant ("REM
su]?ern~1~nt") was e~minecl as a source of soluble stim~ tory factors in a REM expansion in which PBMC were reduced to sub-optimal levels (i.e. 1/2, 1/4 or 1/8 of optimal or "SOP").
The results, as shown in TABLE 5, confirm that such soluble factors can provide an effective stim~ tory signal in the context of low-PBMC REM. In particular, a large proportion of the reduction in fold proliferation levels observed when PBMC are reduced can be overcome by using the REM sl ll)t~ k~'ll in place of the standard medium. In addition, the more the PBMC were reduced (i.e. to 1/8 of 0~11,illlUIll), the greater was the observed effect from using the R~M sup~ (1022-fold average expansion using the RlEM ~u~ , .I versus 359-fold expansion without). Such ~u~e. . .~ and/or their components such as individual cytokines or "cocktails" thereof can thus be used to reduce the need for conrllloting REM with large excesses of ~eeder cells such as PBMCs).

Medium P3BMC Avg. Fold Std.
Proliferation Dev.
SOP MEDIUM SOP 1255 ~160 1/2 SOP 1178 ~64 1/4 SOP 996 ~23 1/8 SOP 359 ~29 48 HR. REM SUP. SOP 1253 ~144 1/2 ~OP 1218 ~73 1/4 SOP 1 178 ~89 1/8 SOP 1022 ~77 lt3 W 0~7/32970 PCT~US97/03293 Example 6 Replacement of Cytokine Activity in Mo-lified-R~
As described above, a large number of cytokines have been described and are widely available, including a number of cytokines that are known to stimulate T
lymphocytes. As will be a~ale,lL to those of skill in the art, such cytokines (whether or not they were previously known to stimul~te T cells) can be readily tested for their ability to ~llgm~Mt rapid expansion using methods such as those above. In addition, for any of the rapid expansion techniques described herein, the resulting exp~n~lçcl T cells can be monitored for the mz~infen~nee of various desired characteristics, using methods such as those illustrated above for hp-REM.
Cytokines to be used in modified-REM can be introduced to the target T cells in any of several ways as illustrated herein. Thus, for example, one or more cytokines can be added to the medium, as exemplified above. ~lt~rn~tively, or in addition, cytokines can also be supplied by cells secreting the cytokines into the REM m~ m Thus, by way of illustration, a m~mm~ n cell line known to secrete a particular cytokine or combination of cytokines can be used. ~lt~rn~tively, a m~mm~ n cell line that does not already secrete a particular cytokine (or that secretes it at suboptimal levels) can be readily modified by introducing a gene encoding the desired cytokine. As is well known, the gene can be placed under the control of any of a variety of promoters (as ~It~ tives to its nri~in~l promoter) so that ~ res~ion of the cytokine can be controlled to maximize its effe~;livt;lless. The entire sequences for a large number of cytokines are known and encoding DNA is often available. Many such sequences are published in nucleic acid and/or protein databases (such as GenEMBL, GENBANK or Swissprot); see, e.g., theCytokine Facts Book, R.E. Callard et al., Academic Press, 1994). Also, as described 2~ above, such additional m~mm~ n cell lines can be modified to provide several T cell stimulatory activities at once.

ple 7 The Rnle of Accessory~ hesion Mol- r~les in Rapid F xp~n~ion As discussed above, APCs such as monocytes and B cells also provide other T cellco-stimulatory signals which serve to enh~nre T cell activation/proliferation. Thus, while T cell activation involves the specific recognition of MHC-bound antigenic peptides on the W O 97132970 PCT~US97/03293 surface of APCs (which interact with the T cell receptor/CD3 complex), a number of antigen-non-specific receptor:ligand interactions between APCs and T cells can further enh7~nc e T cell activatio n/proliferation. In particular, APCs express ligands for a variety of receptors on T cells, and it appears that T cell activation/proliferation is the result of a combination of signals delivered through the T cell receptor and other ~i~n~ling molecules.
A nurnber of such receptor:ligand interactions have already been identified and, for a number of those, inhibition of the receptor:ligar d interactions have been reported to inhibit T cell proliferation and cytokine secretion. By way of illustration, a nurnber of rec~eptor:ligand pairs that are considered likely to play a role in T cell activation/proliferation are listed in TABLE 6 below.

Receptor ~T cell) Ligand (APC) CD4 Class II MHC
CD8 Class I M~C
CDl la (LFA-l) CD54 (ICAM-l) and ICAM 2 ~ 3 CD2 CD58 (LFA-3) CD49d (VLA-4) fibronectin (FN) CD27 ligand to CD27 CD28 CD80 (B7.1) and CD86 (B7.2) CD44 hyaluronate While many of these molecules have been reported to function in adhesion (enhancing cell:cell and/or cell:substrate interactions), many have also been shown to deliver T cell co-stim~ tory signals such as enh~ncing intracellular calcium and the activation of PI and PKC (see, e.g., Geppert et al. 1990. Tmmlmol. Reviews 117:5-66).
The inter~ctions of such adhesion-accessory molecules as described above have been shown to positively enh~nee activation of resting T lymphocytes. Antibodies which bind these accessory molecules have been shown, under specific conditions, to provide T
cell activation signals (see, e.g., the references cited below). Also, the addition of purified ~eces~ory molecule ligands ICAM-l and LFA~ ~(ligands for CD 11 a and CD2 W 097/32970 PCTrUS97/03293 respectively) to purified T cells being stim~ te~l with anti-CD3 monoclonal antibody has been shown to provide co-stimulatory signals for T cell activation and proliferation (see, e.g., Semnani et al. 1994. J. Exp. Med. 180:2125).
Thus, various antibodies directed against CD4 and CD8 are capable of either inhibiting T cell activation (see, e.g., I. Bank and L. Chess. 1985. J. Exp. Med. 162:1194;
G.A. van Seventer. 1986. Eur. J. Tmmllnl~l. 16:1363) or synergizing with anti-CD3 mAb to induce T cell proliferation (see, e.g., F. Emmrich et al. 1986. PNAS 83:8298; T. Owens et al. 1987. PNAS 84:9209; K. Saizawa et al. 1987. Nature 328:260). As is well known by those of skill in the art, a collection of antibodies raised against a particular antigen would be expected to contain antibodies binding to a variety of diLr~rell~ sites on the antigen.
A number of studies have shown that antibodies to other adhesion-accessory molecules are capable of ~ mentin~ T cell stimlll~t;on/proliferation. By way of illustration, see, e.g., J.A. Ledbetter et al. 1985. J. Immunol 135:2331 (antibodies directed to CD5 and CD28 augment anti-CD3-in~ ced T cell proliferation); P.J. Martin et al. 1986.
J. Tmmunol. 136:3282 (antibodies to CD28 ~n~m~nt anti-cD3-inAllred T cell proliferation); R. Galandrini et al . 1993. J. Tmmlml l. 150:4225, and Y. Shimizu. 1989. J.
Immunol 143:2457 (antibodies directed against CD44 ~llgment anti-CD3 in~ re~l T cell proliferation); S.C. Meur et al. 1984. Cell 36:897 ~antibodies directed against the T11.2 and T11.3 epitopes of CD2 stim~ t~ T cell proliferation); R. van Lier. 1987. J. Imrnunol.
139:158g (antibodies directed against CD27 ~ m~nt anti-CD3-in~ ce~ T cell proliferation); Bossy et al. 1995. Eur. J. Irnmunol. 25:459 (antibodies to CD50 (ICAM-3) ment anti-CD3-indueed T cell proliferation); M.C. Wacholtz et al. 1989. J. Exp. Med.
170:431 (antibodies directed to LFA-l augment anti-CD3-in(1nrecl proliferation when the two antibodies are crosslinked on the T cell surface); G.A. van Seventer et. al. 1990. J.
Tmmlm~l. 144:4579 (purified ICAM-l immobilized on plastic with anti-CD3 mAb cc-stimnl~tes T cell proliferation via the LFA-l molecule); Y. Shimizu et al. 1990. J.
Tmmlmol 145:59 (purified fibronectin on plastic with anti-CD3 mAb co-stimulates T cell proliferation, and antibodies to VLA4 and VLA5 inhibited this activity indicating the role of VLA4 and VLA5 as co-stim~ tory T cell receptors); N.K. Damle et al. 1992. J.
Tmmlm~l. 148:1985 (soluble ICAM-l, B7-1, LFA-3 and VCAM ~llgm~nt anti-CD3-infll1ee~1 T cell proliferation). ~ 6 W~ 9 Jl3~970 PCT/US97J03293 Quantification of the relative contribution of such adhesion-accessory factors within the REM protocol can be readily accomplished using deletion techniques and titration experiments in PBMC-limited hp-REM assays analogous to those illustrated above for the combinations of various cytokines.

l~ mple 8 Replnc~ t of Adhesion-Accessory Molecllle Activity in Modified-RF,l~l In an analogous manner to the modifications described above, and perhaps in combination with such modifications, the REM protocol can thus be modified to include a ~ h~racterized cell line expressing high levels of these receptor ligands (obtained by, e.g., gene modification of a cell line of choice or by the identification of established cell lines already e~cpressing such molecules). It is also possible to utilize antibodies directed against accessory molecules known to induce signal tr~n~ ( tion and/or to use purified accessory ligand molecules as means of sub~ g for the corresponding activity provided by the PBMC feeder cells, thereby enabling a reduction in the number of PBMCs required to drive REM.

~,y~n~plc 9 l~eplace~nt of Additional Stimulatory Activ;t;~ Provided by Fl~V-T,C~, While EBV-LCL do not appear to be sufficient for achieving m~rim~l T cell exr~n~ n, they are capable of augmenting expansion in the hp-REM protocol. Analysis of EBV-LCL has indicated that they express adhesion molecules such as LFA-l, ICAM-l, and LFA-3, as well as Fc~R. In addition, EBV-LCL secrete IL-l (Liu et al. Cell.
Tmmunol. 108:64-75, 1987) and IL-12 ~Kobayashi et al., 1989. J. Exp. Med. 170:827), both of which are also secreted by APCs.
As described above, it is believed that such components can be readily supplied by other sources - thereby reducing the need for the large numbers of PBMC and/or EBV-LCL feeder cells characteristic of hp-REM.

W 097132970 PCTrUS97/03293 .Y~ple 10 The Use of anti-CD21 ~ntibody ill Modified REM
CD21 is an accessory molecule expressed on mature B lymphocytes and, at low levels, on T Iymphocytes. We examined the ability of a molecule that binds to CD21 to provide a stimulatory signal in the context of modified REM.
In a first set of ~ hllents, we used plate-bound anti-CD21 antibody to examine the ability to provide a stim~ tory signal in modified REM in which the EBV-LCL feeder population was completely elimin~te-l Two different antigen-specific CTL clones ("R7"
which is alloantigen-specif1c, and "1 lE2" which is EBV-specific) were tested in a modified REM procedure in which EBV-LCL were ~.lillli,~i.lrrl but other components were m~int~ined as described above (PBMC at 500: 1, IL-2 at 25 units/ml). Anti-CD21 antibodies are available from comrnercial sources. We used the anti-CD21 antibody available from Ph~rmingen. Anti-CD3 antibody was also used, and was bound to plates, as with anti-CD21. Cultures were exp~ncle~l over a two week standard REM cycle, es~nti~lly as described above.
The data, as shown in TABLE 7, revealed that the inclusion of anti-CD21 antibodyresulted in a large increase in the fold proliferation obtainable without the use of EBV'-LCL (to 650% of control and 408% of control for R7 and l lE2, respectively).
A second set of experiments, performed using soluble anti-CD21 antibody, provided additional confirm~tQry data. In particular, a range of anti-CD21 concentrations was used in REM as above, except that both anti-CD21 and anti-CD3 were supplied as soluble antibodies (anti-CD21 at concc;lllla~ions ranging from 0 ng/ml to 1.75 ng/ml, anti-CD3 at 10 ng/ml).
As shown in TABLE 8, the removal of all EBV-LCL feeder cells from the cultures resulted in a substantial reduction in the average fold proliferation (to 10% of control and 14% of control for R7 and 1 lE2, respectively). The addition of even small amounts of anti-CD21 antibody to the culture media resulted in a large increase in fold proliferation (to 72% of conkol and 57% of control for R7 and 1 lE2, respectively).
While anti-CD21 antibody provides a conveninet method for enhancing the stim~ t- ry signal, it is also possible to stimlll~te CD21 in other ways. For example, in addition to anti-CD21 antibody, other molecules that can be used to bind to CD21 include C3d, C3dg, iC3b and gp350/220 of EBV (see e.g., W. Timens et al., pages 516-518 in WC) 97~32970 PCT/US97/03293 '~Leucocyte Typing V, White Cell Differentiation Antigens," Schlossman, S.F., et al.
(eds.), Oxford University Press, Ox~ord, 1995). Also, as described above, while such T
cell s~im~ fQry components can be provided as soluble factors in the modif~ed REM
medium, they can also be provided by a cell line included in the medium (e.g., a cell line lhat secretes or presents a molecule that binds to CD21).

Clolle Specificity Stimulation Fold %
Proliferation Control R7 Alloantigen anti-CD3 96 100%
anti-CD3 ~ a~ti-CD21 624 650%
llE2 ¦ EBV ¦ anti-CD3 1 48 100%
¦ anti-CD3 + anti-CD21 ¦ 196 408%

TAI~LE 8 Clone Condition Fold % Control Proliferation R7 SOP l~M 900 100%
1.75 ng/ml anti-CD21 228 25%
1.25 ng/ml anti-CD21 156 17%
0.625 ng/ml anti-CD21516 57%
0.325 ng/ml anti-CD21372 41%
O ng/ml anti-CD21 90 10%
11 E2 SOP REM 420 100%
1.75 ng/ml anti-CD21 96 24%
1.25 ng/ml anti-CD21 192 48%
0.625 ng/ml anti-CD21288 72%
0.325 ng/ml anti-CD21132 33%
" O ng/ml anti-CD21 60 14%

Claims (53)

1. A method for rapidly expanding an initial T lymphocyte population in culture medium in vitro, comprising the steps of:
adding an initial T lymphocyte population to a culture medium in vitro, adding to the culture medium a non-dividing mammalian cell line expressing at least one T-cell-stimulatory component, wherein said cell line is not an EBV-transformed lymphoblastoid cell line (LCL), and incubating the culture.

2. A rapid expansion method of claim 1, wherein said T-cell-stimulatory component is selected from the group consisting of an Fc-.gamma. receptor, a cell adhesion-accessory molecule and a cytokine.

3. A rapid expansion method of claim 1, wherein said T-cell-stimulatory component is selected from the group consisting of an Fc-.gamma. receptor, a cell adhesion-accessory molecule and a cytokine, and wherein said initial T lymphocyte population is expanded at least 200-fold after an incubation period of less than about two weeks.

4. A rapid expansion method of claim 1, wherein said T-cell-stimulatory component is selected from the group consisting of an Fc-y receptor, a cell adhesion-accessory molecule and a cytokine, and wherein said initial T lymphocyte population is expanded at least 500-fold after an incubation period of less than about two weeks.

5. A rapid expansion method of claim 1, wherein said T-cell-stimulatory component is selected from the group consisting of an Fc-.gamma. receptor, a cell adhesion-accessory molecule and a cytokine, and wherein said initial T lymphocyte population is expanded at least 1000-fold after an incubation period of less than about two weeks.

6. A rapid expansion method of claim 1, further comprising the step of adding anti-CD3 monoclonal antibody to the culture medium wherein the concentration of anti-CD3 monoclonal antibody is at least about 1.0 ng/ml.

7. A rapid expansion method of claim 1, further comprising the step of adding IL-2 to the culture medium, wherein the concentration of IL-2 is at least about 10 units/ml.

8. A rapid expansion method of claim 1, wherein said mammalian cell line comprises at least one cell type that is present at a frequency at least three times that found in human peripheral blood mononuclear cells (human PBMCs).

9. A rapid expansion method of claim 1, wherein said T-cell-stimulatory component is selected from the group consisting of an Fc-.gamma. receptor and a cell adhesion-accessory molecule.

10. A rapid expansion method of claim 1, wherein said T-cell-stimulatory component is selected from the group consisting of a cell adhesion-accessory molecule and a cytokine.

11. A rapid expansion method of claim 1, wherein said T-cell-stimulatory component is selected from the group consisting of an Fc-.gamma. receptor and a cytokine.

12. A rapid expansion method of claim 1, wherein said mammalian cell line expresses a cell adhesion-accessory molecule.

13. A rapid expansion method of claim 12, wherein said cell adhesion-accessory molecule is selected from the group consisting of Class II MHC, Class I MHC, ICAM 1, ICAM 2, ICAM 3, CD58, CD72, fibronectin, ligand to CD27, CD80, CD86 and hyaluronate.

14. A rapid expansion method of claim 1, wherein said mammalian cell line expresses a cytokine.

15. A rapid expansion method of claim 1, wherein said T-cell-stimulatory component is a molecule that binds to CD21.

16. A rapid expansion method of claim 14, wherein said cytokine is selected from the group consisting of IL-1, IL-2, IL-4, IL-6, IL-7, IL-12 and IL-15.

17. A rapid expansion method of claim 1, further comprising the step of adding a soluble T-cell-stimulatory factor to the culture medium.

18. A rapid expansion method of claim 17, wherein said soluble T-cell-stimulatory factor is selected from the group consisting of a cytokine, an antibody specificfor a T cell surface component, and an antibody specific for a component capable of binding to a T cell surface component.

19. A rapid expansion method of claim 17, wherein said soluble T-cell-stimulatory factor is a cytokine selected from the group consisting of IL-1, IL-2, IL-4, IL-6, IL-7, IL-12 and IL-15.

20. A rapid expansion method of claim 17, wherein said soluble T-cell-stimulatory factor is an antibody specific for a T cell surface component, and wherein said T cell surface component is selected from the group consisting of CD4, CD8, CD11a, CD2, CD5, CD49d, CD27, CD28 and CD44.

21. A rapid expansion method of claim 17, wherein said soluble T-cell-stimulatory factor is an antibody specific for a component capable of binding to a T cell surface component, and wherein said T cell surface component is selected from the group consisting of CD4, CD8, CD11a, CD2, CD5, CD49d, CD27, CD28 and CD44.

22. A rapid expansion method of claim 17, wherein said soluble T-cell-stimulatory factor is a molecule that binds to CD21.

23. A rapid expansion method of claim 22, wherein said molecule that binds to CD21 is an anti-CD21 antibody.

24. A rapid expansion method of claim 1, further comprising the step of adding to the culture a multiplicity of peripheral blood mononuclear cells (PBMCs).

25. A rapid expansion method of claim 24, wherein the ratio of PBMCs to initial T cells to be expanded is less than about 40:1.

26. A rapid expansion method of claim 24, wherein the ratio of PBMCs to initial T cells to be expanded is less than about 10:1.

27. A rapid expansion method of claim 24, wherein the ratio of PBMCs to initial T cells to be expanded is less than about 3:1.

28. A rapid expansion method of claim 1, further comprising the step of adding to the culture a multiplicity of EBV-transformed lymphoblastoid cells (LCLs).

29. A rapid expansion method of claim 28, wherein the ratio of LCLs to initial T cells to be expanded is less than about 10:1.

30. A rapid expansion method of claim 1, wherein the initial T lymphocyte population comprises at least one human CD8+ antigen-specific cytotoxic T lymphocyte (CTL).

31. A rapid expansion method of claim 1, wherein the initial T lymphocyte population comprises at least one human CD4+ antigen-specific helper T lymphocyte.

32. A method of genetically transducing a human T cell, comprising the steps of: adding an initial T lymphocyte population to a culture medium in vitro; adding to the culture medium a non-EBV-transformed mammalian cell line expressing a T-cell-stimulatory component; and incubating the culture; and adding a vector to the culture medium.

33. A genetic transduction method of claim 32, wherein the vector is a retroviral vector containing a selectable marker providing resistance to an inhibitory compound that inhibits T lymphocytes, and wherein the method further comprises the steps of: continuing incubation of the culture for at least one day after addition of the retroviral vector; and adding said inhibitory compound to the culture medium after said continued incubation step.

34. A genetic transduction method of claim 32, further comprising adding a multiplicity of human PBMCs.

35. A genetic transduction method of claim 34, wherein the ratio of PBMCs to initial T cells is less than about 40:1.

36. A genetic transduction method of claim 32, further comprising adding non-dividing EBV-transformed lymphoblastoid cells (LCL).

37. A genetic transduction method of claim 36, wherein the ratio of LCL to initial T cells is less than about 10:1.

38. A method of generating a REM cell line capable of promoting rapid expansion of an initial T lymphocyte population in vitro, comprising the steps of:
depleting one or more cell types from a human PBMC population to produce a cell-type-depleted PBMC population, using said cell-type-depleted PBMC population in place of non-depleted PBMCs in an hp-REM protocol to determine the contribution of the depleted cell type to the activity provided by the non-depleted PBMCs, identifying a T cell stimulatory activity provided by said depleted cell type, and transforming a mammalian cell line with a gene allowing expression of said T cell stimulatory activity.

39. A method of generating a REM cell line according to claim 38, wherein said T-cell-stimulatory component is selected from the group consisting of an Fc-.gamma.
receptor, a cell adhesion-accessory molecule and a cytokine.

40. A REM cell line capable of stimulating rapid expansion of an initial T
lymphocyte population in vitro, comprising a mammalian cell line generated according to the method of claim 38.

41. A REM cell line according to claim 40, wherein said cell line expresses a cell adhesion-accessory molecule.

42. A REM cell line according to claim 41, wherein said cell adhesion-accessory molecule is selected from the group consisting of Class II MHC, Class I MHC, ICAM 1, ICAM 2, ICAM 3, CD58, CD72, fibronectin, ligand to CD27, CD80, CD86 and hyaluronate.

43. A REM cell line according to claim 40, wherein said cell line expresses an Fc-.gamma. receptor.

44. A REM cell line according to claim 40, wherein said cell line expresses at least one T cell stimulatory cytokine.

45. A REM cell line according to claim 44, wherein said T cell stimulatory cytokine is selected from the group consisting of IL-1, IL-2, IL-6, IL-7, IL-12 and IL-15.

46. A REM cell line according to claim 40, wherein said cell line expresses a molecule that binds to CD21.

47. A culture medium capable of rapidly expanding an initial T lymphocyte population in vitro comprising a REM cell line according to claim 40.

48. A culture medium according to claim 47, further comprising an exogenous cytokine.

49. A culture medium according to claim 47, further comprising a multiplicity of exogenous cytokines, wherein said multiplicity comprises at least one interleukin.

50. A culture medium according to claim 49, wherein said interleukin is selected from the group consisting of IL-1, IL-2, IL-6, IL-7, IL-12 and IL-15.

51. A culture medium according to claim 47, further comprising a molecule that binds to CD21.

52. A culture medium according to claim 51, wherein said molecule that binds to CD21 is an anti-CD21 antibody.

53. A culture medium according to claim 49, further comprising an anti-CD3 monoclonal antibody.