WO2000050571A1 - Improved method of culturing t cells - Google Patents

Abstract

The invention provides a method of enhancing the stimulation and proliferation of T cells. The method comprises culturing a culture containing T cells at a pH of about 6.8 to about 7.2 during the entire length of the culture and/or culturing the culture at an oxygen level substantially less than about 20 %, preferably about 5 %, for at least the time required for stimulation of the culture. The invention also provides a method of cellular immunotherapy and a method of gene therapy. These methods comprise administering an effective number of the autologous or allogenic T cells cultured according to the method of the invention to a patient in need thereof.

Description

IMPROVED METHOD OF CULTURING T CELLS

FIELD OF THE INVENTION This invention relates to methods of culturing T cells. In particular, the invention relates to an improved method of culturing T cells so as to enhance the proliferation (expansion) of the T cells. The invention also relates to methods of cellular immunotherapy and gene therapy which utilize T cells.

BACKGROUND OF THE INVENTION

Cellular immunotherapy, whereby a patient is treated with large doses of autologous ex v/vo-expanded immune cells in order to eradicate aberrant cells, such as malignant or virally-infected cells, offers an alternative treatment for such diseases (see Chang etal, J. Clin. O«co/., 15(2):796-807 (1997);Plautzetal.,_ Newro_«/r ., 89(l):42- 51 (1998); Rooney et al., Blood, 92(5): 1549-55 (1998)). The effectiveness of these therapies is considered to be dose dependent, as patient survival increases proportionately with larger doses of effector cells (Cheever et al., Immunol. Rev., 157:177-94 (1997)). As a result, treatment regimens require large numbers of cells, typically 10⁹- 10¹¹ cells per patient, making expansion of the cells by ex vivo cell culture necessary. In expanding T cells for immunotherapy it is important not only to generate large numbers of cells, but also to ensure that the cell product is of high quality such that it will be biologically active upon transfusion back to the patient. Thus, optimization of the cell culture parameters used for expansion of these cells is a crucial issue for the success of cellular immunotherapy. Currently, the majority of cellular immunotherapy protocols make use of static flasks or gas-permeable tissue culture bags housed in standard incubators with a gas atmosphere of 5% CO₂ and air (about 20% O₂ or a tension of 140 mmHg). Such static culture conditions result in spatial and temporal variations in many culture parameters which may be suboptimal, and may lead to decreased cell-expansion potential, thus requiring a longer time for culture completion.

Hematopoietic and immune cells experience a large spectrum of oxygen tension environments due to the physiology and vasculature of the bone marrow and the lympoid organs (Pennathur-Das et al., Blood, 69:899-907 (1987); Guyton et al, Textbook of Medical Physiology, W.B. Saunders Co., Philadelphia, PA, 1996). The mean oxygen tension in the hematopoietic and lymphoid organ tissues is closer to 40 mm Hg (or 5% of the oxygen in the gas atmosphere), while the anatomical architecture of these organs suggests substantial intraorgan gradients that might expose cells to even lower oxygen tensions. In recognition of this physiological oxygen tension milieu, several studies have established the profound effect of lower oxygen tensions on hematopoietic cells.

Oxygen effects on immune cells have been examined, but with variable results. Studies of myeloid hematopoietic cells have shown that greater cell proliferation is observed in a 5% O₂ environment (Roller et al, Exp. Hematol. , 20:264-70 (1992)). More recent work (Laluppa et al, Exp. Hematol., 26:835-43 (1998)) suggests that expansion and differentiation of different myeloid cell lineages are optimized under different oxygen conditions. Megakaryocytopoiesis and erythropoiesis are enhanced under high oxygen tensions (20%), while granulopoiesis is enhanced by low oxygen tensions (5%). In contrast, it has been reported that lower oxygen tensions may not have beneficial effects on lymphocyte cultures. Loeffler et al. (Loeffler et al, Br. J. Cancer, 66:619-22 (1992)) reported a significant reduction of lymphocyte proliferation under 1% O₂ when compared to 20% O₂. Naldini et al. (Naldini et al, J. CellPhysiol, 173:335-42 (1997)) reported that peripheral blood mononuclear cells (PBMCs) are less susceptible to phytohemagglutinin (PHA) activation under 2% O₂. Anderson et al. (Anderson et al, J. Cell. Physiol, 72:149-52 (1968)) reported maximal thymidine incorporation by PHA-stimulated lymphocytes in a 20% O₂ environment, over a range of 0 to 70% O₂ concentrations. Krieger et al. (Krieger et al, Int. J. Immunopharmac, 18:545-52 (1996)) reported greater proliferation of peripheral blood mononuclear cells (PBMCs) under 5% O₂ conditions than 20% O₂ when stimulated with concanavalin A (Con A) or pokeweed mitogen (PWM), but no differential oxygen effects when stimulated with PHA or Staphylococcal enterotoxin

B (SEB). It is worth noting that in all of these experiments no exogenous interleukin 2 (IL-2) was added to the cultures. It is also important to note that the optimal oxygen concentration for proliferation may not necessarily be the same as for other cell functions. For the case of PBMCs stimulated only with high levels of IL-2 in order to generate lymphokine activated killer (LAK) cells, Ishizaka et al. (Ishizaka et al. , Immunol. Letters,

32:209-14 (1992)) reported no difference in proliferation of the cells, but a reduction in the killing activity of the cells under 2% O₂ when compared to 20% O₂. Clearly, optimization of the culture oxygen tension for expansion of T cells is crucial for generation of healthy functional T cells for use in immunotherapy protocols.

It is well known that extracellular pH is an important parameter in the culture of mammalian cells. Culture pH is known to have many diverse effects on cells including effects on proliferation rate (Akatov et al, Exp. Cell. Res., 160(2):412-8 (1985); Loeffler et al, Br. J. Cancer, 66:619-622 (1992)), differentiation (McAdams et al, Br. J. Haematol, 103(2):317-25 (1998); Endo et l,LeukRes., 18(l):49-54 (1994); Fischkoff et al, J. Exp. Med, 160(l):179-96 (1984)), metabolism (McDowell et al., Biotechnol. Prog., 14(4):567-72 (1998); Miller et al, Biotechnol. Bioeng, 32:965-977 (1988);

McQueen et al, Biotechnol. Bioeng., 35:1067-1077 (1990)), and protein synthesis and glycosylation (Borys et al, Biotechnology (NY), ll(6):720-4 (1993); England et al, Am. J. Physiol, 260(2 Pt l):C277-82 (1991); Gaitanaki et al., FEBS Lett, 260(l):42-4 (1990)). Given that extracellular pH can affect cultured cells in so many ways, it is especially crucial to examine its effects when the cultured cells themselves are the therapeutic product.

Obviously, once the optimal pH has been determined, the ability to establish and maintain that condition within the culture system is also crucial. Well-developed pH control systems exist for homogeneous stirred culture bioreactors, but such stirred systems have not yet been widely employed in the field of cellular therapies. Static culture devices, such as T flasks and tissue culture bags, which are commonly used, allow for the formation of nutrient, metabolite, oxygen, and pH gradients, especially as the cell density increases. Akatov et αl, (Akatov et al, Exp. Cell. Res., 160(2):412-8 (1985) found that in a static culture of Chinese hamster fibroblasts at a density of 10⁶ cells/cm², the pH within the pericellular microenvironment dropped 1.2 units in 6 hours, as compared to the bulk medium pH. Piret et al. (Piret et al, Can. J. Chem. Eng., 69:421-428 (1991)) have shown that large gradients, both axial and radial in orientation, exist within hollow fiber reactor culture devices, which have also been used in T cell expansion protocols (Freedman et al., J. Immunol. Methods, 167:145-160 (1994); Knazek et al, J. Immunol. Methods, 127:29-37 (1990)). Because of the difficulties in adequately controlling pH within the culture devices most commonly employed today, it is important to understand the impact of pH on the cultured cells.

One might assume that the optimal culture pH for the ex vivo expansion of T cells should closely match that experienced by the T cells in vivo. However, it is likely that the T cells experience a range of pH values in vivo as they migrate from the blood stream, into the lymph nodes, and out to the sites of infection or malignant disease. Arterial blood has a normal pH of 7.4, other tissues have pH values in the range of 7.0 - 7.4, and solid tumor masses (which may contain tumor infiltrating lymphocytes (TIL)) may have pH values as low as 6.2 - 6.5 due to the accumulation of metabolic waste products as a result of poor vascular perfusion. Loeffler etal. (Loeffler _X .,Br. J. Cancer, 66:619-622 (1992)) have shown that the proliferation of murine lymphocytes is pH-dependent. They studied a range of pH values from 6.4 to 1.4 and found optimal proliferation (based on ³H- thymidine incorporation) at pH 7.0 and greatly reduced proliferation at pH values of 6.7 and 6.4. No data were reported between values of 7.0 and 7.4 in their study. The majority of T cell expansion protocols in the literature do not report the pH of the culture medium. The pHs of two commercially-available serum-free media commonly used for T cell culture (AIM V, GTBCO BRL, Grand Island, NY, and X-VIVO 20, BioWhittaker, Walkersville, MD) were measured after equilibration in a 5% CO₂ incubator to be 7.3 (AIM V) and 7.4 (X-V-TVO 20) (see Example 2 below). Expression of certain surface receptors has been shown to be affected by pH.

McDowell and Papoutsakis showed that decreasing culture pH increased the CD 13 receptor surface content on HL60 cells over a range of 7.0 to 7.4 and that this increase in expression was not associated with changes in the messenger RNA levels for the receptor (McDowell et al, Biotechnol. Prog, 14(4):567-72 (1998)). Katafuchi et al. (Katafuchi et al, Am. J. Physiol, 264:C1345-C1349 (1993)) demonstrated that the surface concentration of atrial natriuretic peptide (ANP) receptor, present on endothelial cells, varied more than three orders of magnitude between cells cultured at 7.0 and 7.7. Proper expression of cell surface receptors is crucial for T cells, as they are required for many key functions in vivo including target cell recognition, activation of the T cell, and trafficking within the body. Gene therapy can be defined as a therapeutic modality whereby the transfer of genetic material into some cells of a patient can bring about a therapeutic outcome of temporary or lasting duration (Anderson, Science, 256:808-813 (1992); Mulligan, Science, 260:926-932 (1993)). The genetic material can be delivered to patient cells either in vivo (i.e., the cells are, and remain, part of the patient's body when the genetic material is delivered to them) or ex vivo (i.e., specific cells are removed from a patient, the genetic material is delivered to these cells, and the genetically-modified cells are then returned to the patient). In ex vivo gene therapy, before the genetically-modified cells are returned to the patient, the cells may or may not be expanded prior to, or after, the introduction of the genetic material. However, in most cases, ex vivo cell expansion is used as part of the gene therapy protocol (Anderson, Science, 256:808-813 (1992); Mulligan, Science, 260:926-932 (1993); Ferrari et al, Science, 251:1363-1366 (1991); Jolly, Cancer Gene Therapy, 1:51-64 (1994); Crystal, Science, 270:404-410 (1995)). A gene therapy vector is the means by which one or multiple genes are delivered to cells in vivo or ex vivo. A vector is made up of genetic material suitably constructed for the introduction into the cells, and the expression of, the necessary genetic material. A large variety of vectors have been proposed, designed, constructed, and tested, including vectors of viral and non-viral origin (Anderson, Science, 256:808-813 (1992); Mulligan, Science, 260:926-932 (1993); Ferrari et al, Science, 251:1363-1366 (1991); Jolly, Cancer Gene Therapy, 1:51-64 (1994); Crystal, Science, 270:404-410 (1995)) . Various cell types have been used as part of ex vivo gene therapy protocols, including hematopoietic stem and progenitor cells, T cells, tumor cells, fibroblasts, and hepatocytes, to treat a wide spectrum of diseases including various malignancies, AIDS, AD A-deficient SOD, liver failure, familial hyperchlolesterolemia, and cystic fibrosis (Anderson, Science, 256:808-813 (1992); Mulligan, Science, 260:926-932 (1993); Ferrari et al, Science,

251:1363-1366 (1991); Crystal, Science, 270:404-410 (1995)). The use of T cells of various origins has been quite widespread. T cells are, thus, likely to be one of the most frequently used cells for ex vivo gene therapy, and T cell expansion will be necessary for most such gene therapy protocols. SUMMARY OF THE INVENTION

The invention provides an improved method of culturing T cells. The method comprises providing a sample of cells, at least some of which are T cells, placing the cells in a culture medium, and culturing the cells in the culture medium for a selected period of time under conditions effective to cause proliferation of the T cells. The conditions effective to cause proliferation of the T cells comprise including a T cell stimulant in the culture for at least a period of time sufficient to stimulate the T cells. The conditions further comprise: (i) maintaining the pH of the culture at from about 6.8 to about 7.2 during the entire length of the culture; (ii) maintaining the oxygen level of the culture at substantially less than about 20% for at least the period of time sufficient to stimulate the

T cells with the T cell stimulant, preferably for the entire length of the culture; or (Hi) maintaining the pH of the culture at from about 6.8 to about 7.2 during the entire length of the culture and maintaining the oxygen level of the culture at substantially less than about 20% for at least the period of time sufficient to stimulate the T cells with the T cell stimulant Use of this pH range, oxygen level, or both, substantially increases the stimulation (activation) and proliferation (expansion) of the T cells as compared to the pH (about 7.3-7.4) and oxygen level (about 20%) used in standard T cell cultures.

The invention also provides a method of cellular immunotherapy. This method comprises administering an effective number of autologous or allogeneic T cells to a patient in need thereof, the T cells having been cultured as described in the previous paragraph.

Finally, the invention provides a method of gene therapy. This method comprises administering an effective number of genetically-modified autologous or allogeneic T cells to a patient in need thereof, the T cells having been cultured as described above.

BRIEF DESCRIPTION OF THE DRAWINGS

Figures 1A-B. Growth curves of T clles from an "O₂ crossover" experiment

(sample D as identified in Tables I-III below). Cells were stimulated six days prior to time

0 on the plots with either PHA (Figure 1 A) or anti-CD3 monoclonal antibody (mAb) (Figure IB), and then cultured at either 5% or 20% O₂. At time 0, the activated cells were split in half and either maintained in their original O₂ condition or switched to the alternate condition. In Figures 1A-B, • = stimulation in 5% O₂ and culture in 5% O₂, Δ = stimulation in 5% O₂ and culture in 20% O₂, □ = stimulation in 20% O₂ and culture in 5% O₂, and ▼ = stimulation in 20% O₂ and culture in 20% O₂.

Figures 2A-B. Figure 2A is a Graph showing percent of apoptotic cells (as determined by the TUNEL assay) as a function of time in cultures under 5% and 20% O₂ environments. Figure 2B shows the growth curves from this culture, where the boxed time points correspond to the samples in Figure 2A. In Figures 2A-B, π = 5% O₂ and

• = 20% O₂

Figures 3A-B. Graphs showing expression kinetics of CD25 (IL-2 receptor, IL- 2R) as percent of cells brighter than an isotype control (Figure 3 A) and amount of receptor per cell (Figure 3B) (quantified as PE molecules bound by antibody staining using flow cytometry and the Quantibrite calibration beads) on those cells designated as CD25⁺ in Figure 3 A. Note the log scale on the vertical axis. In Figures 3 A-B, □ = 5% O₂ and

• = 20% O₂. Figure 4. Graph showing the effect of O₂ on specific metabolic rates, q_gh,^ and

^_lactose- (mean ± SEM for n = 5 blood samples; includes both PHA and anti-CD3 mAb activated cells). Solid bars — T cells stimulated and cultured at 5% O₂. Crosshatched bars — T cells stimulated and cultured at 20% O₂. Means are averaged over the entire length of each culture and then averaged over all five samples. * means p < 0.001 for 5% versus 20% O₂. Figure 5. Graph showing the effect of O₂ and autologous patient plasma on T cell expansion using a commercially available serum-free medium (AIM V). Peripheral blood mononuclear cells from a hemochromatosis patient were stimulated simultaneously with both anti-CD3 and anti-CD28 antibodies (100 ng/ml each) and expanded under either 5% or 20% O₂ levels in the presence of IL-2, with or without 2% autologous patient plasma. In Figure 5, 0 = 5% O₂, no plasma, □ = 5% O₂, with plasma,

• = 20% O₂, no plasma, ■ = 20% O₂, with plasma.

Figure 6. Graph showing the effect of O₂ and autologous patient plasma on the expansion of CD4+ and CD8+ T cells using a commercially available serum-free medium

(AIM V). Peripheral blood mononuclear cells from a hemochromatosis patient (same as in the culture performed in Figure 5) were stimulated simultaneously with both anti-CD3 and anti-CD28 antibodies (100 ng/ml each) and expanded under either 5% or 20% O₂ levels in the presence of IL-2, with or without 2% autologous patient plasma. In Figure 6, O = 5% O₂, no plasma, □ = 5% O₂, with plasma, • = 20% O₂, no plasma, ■ = 20% O₂, with plasma.

Figure 7. Graph showing % apoptotic cells (as determined by the TUNEL assay) in the cultures of Figures 5 and 6. In Figure 7, O = 5% O₂, no plasma, □ = 5% O₂, with plasma, • = 20% O₂, no plasma, ■ = 20% O₂, with plasma.

Figure 8. Graph of average ratios of fold expansions for T cells stimulated with PHA and cultured at different pH values. Error bars = standard error of the mean (S.E.M). Low = pH 7.0, medium = pH 7.2, and high = pH 1.4. Figures 9A-B. Representative growth curves showing decreased total proliferation due to PHA stimulation and culture at pH 1.4 as compared to pH 7.2 or pH 7.0 for two samples with varying extents of expansion. In Figures 9A-B, O = pH 7.0, ■ = pH 7.2, and X = pH 7.4.

Figures lOA-B. Graphs showing slower downregulation kinetics of IL-2R expression at pH 7.0 for both percentage of IL-2R positive cells (Figure 10A) and mean fluorescence intensity of the IL-2R positive fraction (Figure 10B) (note the log scale of the vertical axis). In Figures 10A-B, O = pH 7.0, ■ = pH 7.2, and X = pH 7.4.

Figure 11. Graph of percent apoptotic cells (determined by A nexin V method) in a representative culture experiment showing increased levels of apoptosis at higher pH values. In Figure 11, O = pH 7.0, ■ = pH 7.2, and X = pH 7.4.

DETAILED DESCRIPTION OF THE PRESENTLY-PREFERRED EMBODIMENTS

As is well known, T cells are lymphocytes that mature in the thymus. They are responsible for cell-mediated immunity and act as regulators of the immune response. Mature T cells can be divided into two subsets that differ in function on the basis of surface antigenic determinants - CD4 and CD8. CD4+ T cells recognize antigen associated with Class II histocompatibilty (MHC) molecules, and CD8+ T cells recognize antigen in associated with Class I MHC molecules. Most CD4+ cells are helper T cells or are responsible for delayed-type hypersensitivity reactions. Most CD8+ T cells are cytotoxic and/or suppressor cells. To culture T cells according to the invention, a sample of cells containing at least some T cells is obtained. T cells are found in the lymph nodes, spleen, thymus, peripheral blood, peritoneal fluid, and other tissues and fluids of the body (e.g., infiltrating T cells from tumors), and T cells from any source may be used. The sample of cells is removed from the desired source, and a single-cell suspension is prepared, all by methods well known in the art. See, e.g., Freedman et al, J. Immunol. Methods, 167: 145-160 (1994); Riddell and Greenberg, J. Immunol. Methods, 128:189-201 (1990); Robinet et al, J. Hematother, 7:205-215 (1998): Chang et al, J. Clin. Oncol., 15:796-807 (1997). The cell suspension, containing T cells and other cells, can be used as such in the culture, or the T cells can be purified, partially or completely, as is also well known in the art. For instance, the source of the cells will be peripheral blood in many cases, and the most common, and the currently preferred, method of T cell purification for peripheral blood is apheresis (a density gradient-based separation of peripheral blood yielding T cells in the mononuclear cell fraction which may be achieved using a continuous process). See, e.g., Peshwa et al, Biotechnol. Bioeng., 50:529-540 (1996); Robinet et al., J. Hematother.,

7:205-215 (1998). T cell purification can be performed using antibody-based techniques to selectively capture the desired T cell fraction or to remove undesired populations, such as B cells. See, e.g., Levine et al., J. Hematother., 7:437-448 (1998). Subpopulations of T cells can also be used in the cultures of the invention. Methods of preparing subpopulations of T cells are well known. For instance, antibody-based methods employing antibodies to a selected T cell surface antigen (e.g., CD4 or CD8) can be used to purify a subpopulation of T cells (e.g., anti-CD4 antibodies can be used to purify either the CD4+ or CD8+ subpopulation).

The present invention provides a method of culturing the sample of cells under conditions effective to enhance the proliferation of the T cells in it as compared to standard T cell cultures. The conditions that have been found to enhance the proliferation of T cells are: (i) culturing the T-cell-containing sample at an optimum pH; (ii) culturing the T-cell-containing sample at an optimum oxygen level; or, preferably, (iii) culturing the T-cell-containing sample at both an optimum pH and an optimum oxygen level. The optimum pH for cultures containing T cells has been found to be from about

6.8 to about 7.2, preferably about 7.0. By "about 7.0," etc. is meant 7.0 ± the measurement error of the instrument being used. The pH of the culture should be measured and maintained on a regular basis. Generally, measurement of the pH once per day will be sufficient. Methods for continuous pH measurement and control using probes inserted into the culture vessel are well known. See, e.g., McDowell and Papoutsakis, Biotechnol. Prog, 14: 567-572 (1998). The pH can also be measured by taking a sample of the culture medium and measuring its pH with a pH meter. The pH of the culture should be adjusted as needed to keep the pH in the optimum pH range during the entire culture period. The pH can be adjusted by adding base or acid, such as 1 MNaOH or 1M HCl The pH can also be adjusted by the addition or removal of CO₂ in media containing a bicarbonate buffer. Also, prior to feeding the cultures with fresh culture medium and/or subdividing the cultures, the pH of the fresh culture medium should be measured and adjusted to be within the optimum pH range, if necessary.

The optimum oxygen level for T cell cultures has been found to be substantially less than about 20% O₂, the level of oxygen employed in standard T cell cultures. Preferably the oxygen level employed for the T cell cultures of the invention is from about

2% O₂ to about 10%) O₂, more preferably from about 3% O₂ to about 7% O₂, most preferably about 5% O₂. The oxygen level of the culture should be kept at the optimum level for at least the period of T cell stimulation (see below). After this time, the oxygen level can be increased (up to about 20%), but it is preferably kept at the optimum level for the entire period of the culture. By "substantially less than about 20% O₂" is meant that level of oxygen below 20% O₂ giving statistically significant better T cell proliferation than obtained using 20% O₂. By "about 5% O₂," etc. is meant 5% + the measurement error of the machine being used to measure oxygen level. Percent (%) oxygen refers to the oxygen concentration in the gas phase with which the culture medium is equilibrated (i.e., the culture headspace or incubator environment).

Typically, T cell cultures are cultured in an incubator having an atmosphere consisting of 5% CO₂ and 95% air, which has an oxygen concentration of about 20%. Thus, an inert gas, such as N₂, must be added to the incubator to lower the O₂ content so that the final atmosphere in the incubator will contain a level of oxygen in the optimum range described above, most preferably about 5% O₂. Incubators containing oxygen sensors and mechanisms for controlling oxygen concentration are commercially available from, e.g., Forma Scientific. The oxygen level of the T cell culture should be in equilibrium with that of the atmosphere in the incubator. In addition, the oxygen level of the culture may be measured on regular basis during the culture period. Methods for continuous oxygen measurement and control using probes inserted into the culture vessel are well known (available from Ingold Electrodes, Wilmington, MA), and in a bioreactor- type culture vessel, gases can be fed directly into the headspace for control. Alternatively, samples can be removed from the culture, and the oxygen level can be measured using a blood gas analyzer (available from Instrumentation Laboratories). During sampling procedures, care should be taken to minimize exposure of the cultures to air. Oxygen levels can be controlled in cultures maintained in an incubator by adjusting the gas environment inside the incubator. The oxygen level should be adjusted as needed based on the regular measurements. Also, prior to feeding the cultures with fresh culture medium and/or subdividing the cultures, the oxygen level of the fresh culture medium should be measured and adjusted if necessary. Finally, it is crucial that the oxygen level in the culture be reasonably uniform, as would be afforded by suitable mixing of the culture.

As noted above, culturing T-cell-containing cultures using an optimum pH and an optimum oxygen level enhances the proliferation of the T cells in the cultures as compared to standard T cell cultures. For instance, culturing of peripheral blood mononuclear cells (PBMC) at a pH of about 7.0 or about 7.2 for the entire length of the culture resulted in over a three-fold greater expansion of the T cells as compared to culturing PBMC at a pH of about 1.4. Standard T cell cultures are performed at a pH of about 7.3-7.4. Similarly, culturing PBMC at an oxygen level of about 5% over the entire length of the culture resulted in a 5.8-fold greater expansion of the cultures as compared to culturing PBMC at an oxygen level of about 20%. Standard T cell cultures are performed at an oxygen level of about 20%. It is expected that the use of both an optimum pH and an optimum oxygen level will give multiplicative results (e.g., about 17-18 fold greater expansion for PBMC as compared to standard culture conditions).

Other than the use of the optimum pH and/or optimum oxygen level as described above, the conditions, equipment, media, and other reagents used in the T cell cultures of the invention will be those well known in the art. See Freedman et al., J. Immunol. Methods, 167:145-160 (1994) for a review of these conditions, equipment, media, and reagents. Some of these conditions and reagents will be discussed briefly.

The cultures are initiated using a particular inoculum density. The inoculum density refers to the quantity of mononuclear cells (MNCs) per unit volume in the cell sample used to initiate the cultures. Methods of determining the number of MNCs present in a sample are well known in the art. Preferably, the MNCs are counted using a hemocytometer, Coulter Counter or similar apparatus. The inoculum density will vary depending on the source of the cells, the culture device, and the other culture conditions. Acceptable inoculum densities are known or can be determined empirically. Culture media suitable for T cell cultures are well known. Either serum-containing or serum-free medium can be used for the T cell cultures. Serum-containing medium generally gives higher total cell expansion. However, if a more defined medium is desired, as is the case for most clinical applications, acceptable expansion can be obtained using serum-free medium. Preferably, however, serum or plasma obtained from the patient that is to receive the T cells (autologous serum or plasma) will be used for clinical applications.

Alternatively, allogeneic serum or plasma can be used in cultures for clinical applications. As described above, the culture medium should be adjusted to the optimum pH and/or optimum oxygen level prior to adding cells.

One or more T cell stimulants must be used in the T cell cultures. T cell stimulants are materials which stimulate the proliferation of T cells (T cell activation) . Suitable T cell stimulants, effective amounts of them, and methods of using them are well known. For instance, polyclonal mitogens, such as phytohemagglutinin (PHA), concanavalin A (Con A), pokeweed mitogen (PWM), and Staphylococcal enterotoxin B (SEB), can be used. Also, antibodies specific for antigens on the surfaces of T cells whose binding stimulates T cells can be used. Suitable antibodies include anti-CD3 antibodies and anti-CD28 antibodies. Soluble antibodies may be used or the antibodies may be immobilized on various solid supports, such as beads of various sizes and made of various materials, as is well known in the art. T cell stimulants further include antigen-presenting cells. Suitable antigen-presenting cells include dendritic cells, macrophages, B cells, and tumor cells, which may or may not be genetically modified (Peshwa et al. , Biotechnol. Bioeng. ,

50:529-540 (1996); Greenberg et al, Annals N.Y. Acad. Set, 636:184-195 (1991)). Further, high doses of interleukin 2 can be used to stimulate T cells. Finally, combinations of one or more T cell stimulants can be used (e.g., a combination of PHA and an anti-CD3 antibody or a combination of an anti-CD3 antibody and an anti-CD28 antibody). The T cell stimulant must be kept in the culture for at least a time sufficient to stimulate the T cells. Such times are known or can be determined empirically.

A cytokine is also preferably used in the T cell cultures. Cytokines activate T cells and/or promote the growth and/or proliferation of T cells. Many suitable cytokines are known in the art. Preferred and crucial is interleukin 2 (IL-2) which promotes the growth and proliferation of all T cells. Recombinant IL-2 is commercially available from Chiron Corp. IL-2 is also a component of T cell conditioned media, also known as autologous lymphokine mixtures, which can be used in the culture of T cells (Gold et al, Eur. J. Cancer, 31 A:698-708 (1995)). Other cytokines that have been shown to have a positive effect on T cell proliferation and can be used in T cell cultures, include IL-lβ (Baxevanis et al, Br. J. Cancer, 70:625-630 (1994), IL-4 (Kawakami et al, J. Immunother. , 14:336- 347 (1993), and IL-12 (Kobayasbi et al, J. Exp. Med, 127:827-845 (1989)). Effective amounts of cytokines are known or can be determined empirically.

Suitable protocols for feeding and/or subdividing T cell cultures are known in the art or can be determined empirically. In general, the higher the inoculum density and the higher the cell density present in a culture, the more often feeding and/or subdivision of the culture (whether static or stirred) will be required.

The T-cell-containing cultures of the invention are cultured for a selected period of time. The selected period of time may be a set period of time (e.g. , 14-21 days) or may be the period of time necessary to obtain a desired number of T cells. In the latter case, the number of T cells in the culture will be monitored periodically during the culture (e.g., daily) by methods well known in the art (see, e.g., the Examples below). Stimulation of the T cells with one or more T cell stimulants as described above induces proliferation which will typically last for 2-3 weeks, and can be repeated if desired.

T cells, from both autologous and allogeneic sources, can be used in cellular immunotherapy and gene therapy protocols. For a recent review describing several potential clinical applications for T cell therapies, see Greenberg and Riddell, Science,

285: 546-551 (1999), the complete disclosure of which is incorporated herein by reference. Also see those references discussed in the Background section, above. Atypical protocol would include harvesting the starting cell material (usually peripheral blood) from a patient, purifying the mononuclear cells (this fraction contains the T cells) by apheresis and then culturing the cells as described above. Serum-free medium can be used to culture the T cells, but, preferably, the patient's plasma or serum is added to the culture medium at an effective concentration (2% - 5% has been found to give good results). The cells would be cultured until the desired number of T cells is obtained. Then, the cells would be washed and resuspended in saline, or another similar solution, for administration to the patient. For gene therapy protocols, the cells could be genetically modified prior to commencing the culture, after stimulation with the T cell stimulant(s), or after the culture is completed.

EXAMPLES Example 1 : Effect of Oxygen Level on T Cell Cultures In this study, PBMCs were stimulated with either PHA, an anti-CD3 monoclonal antibody (mAb), or a combination of an anti-CD3 mAb and an anti-CD28 mAb 5% (low) or 20% (high) oxygen atmospheres. After stimulation, cells were cultured in the presence of IL-2 under either low or high oxygen conditions. Levels of proliferation, apoptosis and expression of certain cell-surface receptors and metabolic rates were determined. It was found that T cells stimulated and grown under 5% O₂ exhibited higher proliferation rates and a mean of 5.8 fold greater total expansion over T cells grown under 20% O₂. Stimulation under 5% O₂ produced a lasting proliferative effect even after a switch to 20% O₂. Examination of apoptosis by the flow cytometry-based TUNEL assay showed a mean of 2.9 fold greater percentage of apoptotic cells under 20% O₂. Flow- cytometric analysis of the IL-2 receptor (IL-2R, CD25) showed that the normal downregulation kinetics following stimulation-induced CD25 upregulation were slowed under 5% O₂, such that the 5% O₂ cultures had a greater number of CD25+ cells, and those CD25+ cells expressed an average of 41% higher levels of CD25 receptor per cell. The beneficial effects of low (5%) versus high (20%) O₂ levels were found for three different activation methods, in both serum-free (with or without autologous patient plasma) and serum-containing media. No significant effects of the O₂ level were observed on other surface antigens (CD3, CD28, and CD62L) examined. The key metabolic parameters, specific glucose uptake rate and specific lactate production rate, were both increased by a mean of 47% under 5% O₂. Beyond the physiological significance, improved T cell proliferation under 5% O₂ allows for decreased culture times in expanding T cells for cellular immunotherapies. Evidence of the increased IL-2R expression and reduced apoptosis levels under 5% O₂ may help explain this phenomenon.

METHODS AND MATERIALS Cells and culture conditions. Cells for this study were obtained from two different populations. Samples of leukapheresis (continuous-flow process for blood cell separation) products from non-hematological cancer patients who had undergone stem cell mobilization procedures consisting of treatment with G-CSF with or without chemotherapy (Response Oncology, Memphis, TN) were used, as they represent a potentially clinically relevant population. Samples were also obtained from otherwise healthy donors undergoing therapeutic phlebotomies for treatment of hemochromatosis

(Evanston Hospital, Evanston, IL). Most hemochromatosis donors were in the "maintenance" phase of treatment, and therefore did not have highly elevated serum iron levels. All cells were obtained after informed consent under protocols approved by the respective Institutional Review Boards. The mobilized apheresis product samples were from patients with non-hematological malignant disease, and although their potential exposure to chemotherapeutic agents could alter lymphocyte proliferation and susceptibility to oxidative damage, this nonetheless represents a population of significant interest as potential recipients of adoptive immunotherapy. The PBMCs from the hemochromatosis patient samples were collected using a Histopaque density gradient, and were washed in culture medium prior to stimulation, which removes most of the iron- enriched serum and erythrocytes. Cells were seeded at lxl 0⁶ cells/ml and cultured for 5- 10 days in T-flasks using RPMI with 100 U/ml IL-2 (Chiron: Emeryville, CA), 10% FBS (Hyclone: Logan, UT), 2 mM glutamine, 1 mM sodium pyruvate, 0J mM non-essential amino acids, 25 mMHEPES, 100 U/ml penicillin, 100 μg/ml streptomycin and either 5 μg/ml PHA or 20 ng/ml of soluble anti-human CD3 monoclonal antibody (Pharmingin:

San Diego, CA) for activation. Following this cell activation period, the cells were observed to be >90% CD3⁺ by flow cytometry. Unless otherwise noted, all reagents were from Sigma Chemical Co. (St. Louis, MO).

Preparation of autologous patient plasma. Whole patient blood was centrifuged for 10 minutes at 500 x g. The supernatant was extracted, placed into a new tube, and centrifuged for 20 minutes at 1800 x g to remove any platelets. The plasma was extracted without disturbing the platelet pellet and centrifuged for an additional 20 minutes at 1800 x g. Plasma was stored at -20°C.

Culture protocols and O_; evaluation. Following activation, cells were expanded in 100 ml spinner flask (Bellco Glass, Vineland: NJ - model #1965) cultures in order to ensure a homogeneous culture environment. The spinner flasks were agitated at 60 rpm on stir plates in incubators that allowed for control of gas phase O₂ as well as CO₂ concentrations (Forma Scientific: Marietta, OH). Low O₂ concentrations (5% in the gas phase) were obtained using a nitrogen purge and CO₂ was maintained at 5%. Using a dissolved O₂ probe (Ingold, Columbus, OH) inserted into a spinner flask, it was determined that following sampling, where a culture is briefly exposed to 20% O₂, it takes about 4-5 hours for the culture to completely re-equilibrate to the low O₂ level. Therefore, sampling of the cultures was performed only once a day to minimize exposure of the low O₂ cultures to elevated O₂ concentrations. Cell concentrations were maintained between approximately 2xl0⁵ and 2xl0⁶ cells/ml by diluting the cultures using media which had been pre-equilibrated to the desired O₂ concentration. Counting of nucleated cells was carried out using a Coulter Multisizer (Coulter Electronics, Hialeah, FL) after treatment with Cetrimide (Sigma, St. Louis, MO) to lyse the cells and release the nuclei. The impact of oxygen level on T cell proliferation was assessed based on differences in proliferation rate and total fold expansion resulting from the initial stimulation with PHA or anti-CD3 mAb. These differences can be readily observed on plots of T cell expansion versus time. Cells were stimulated only once, and experiments concluded when cell proliferation stopped (usually 2 to 3 weeks). Total fold expansion was calculated based on percentage of CD3⁺ cells in the initial blood sample.

Glucose and lactate metabolic analysis. Glucose and lactate concentrations were determined using a YSI model 2700 Biochemistry Analyzer (Yellow Springs Instrument

Co., Yellow Springs, OH). Specific metabolic uptake and production rates (q^ and q^ in units of mmol (glucose or lactate)/cell hr) were determined by plotting the glucose or lactate concentration versus the integral under the cell growth curve (cumulative cell- hours) (Rennard et al, Biotechnol. Lett., 10:91-96 (1988)). This method was used over relatively small 2-4 day segments of the experiments where periods of constant specific metabolic rates, yield a straight line with a slope equal to that rate. For more detailed kinetic analysis of q^ and c^ changes over the course of a culture the second-order central slope method was used as previously described (Collins et al., Biotechnol. Bioeng. , 55:693-700 (1997)). The yield coefficient describing the ratio of lactate production to glucose consumption, Y,^, was calculated as q_{la6A I}. The yield coefficients ¥*_____& and Y_mmoi _kc/ceu were calculated as the respective ratios of changes in cell, glucose, or lactate concentrations over specific time intervals.

TUNEL assay. Flow cytometric determination of the apoptotic fraction of cells was performed using the terminal deoxy transferase mediated dUTP nick end labeling (TUNEL) assay (Li et al, Cytometry, 20:172-180 (1995)) (In Situ Cell Death Detection Kit, Fluorescein, Boehringer Mannheim, Indianapolis, IN). The basis of this assay is enzymatic incorporation of fluorescein-conjugated nucleotides into the DNA strand breaks characteristic of apoptotic cells. Briefly, cells were washed in PBS containing 1% BSA, fixed in 1% methanol-free formaldehyde, washed in PBS, and then resuspended in 70% ethanol and stored at -20 °C until further processing. After all of the samples had been collected, cells were rewashed in PBS with 1% BSA, then resuspended in the label and enzyme solutions from the kit, and incubated at 37 °C for 1 hr. Cells were then washed and resuspended in PBS with 1% BSA for analysis by flow cytometry. As a positive control, cells were incubated in DNAse (lmg/ml) following the ethanol permeabilization to induce strand breaks. Acridine Orange Ethidium Bromide (AO EtBr) assay. Some samples were also analyzed for apoptosis using the method described in Mercille et al, Biotechnol. Bioeng. , 44: 1140-54 (1994)). Cells were labeled with a dye solution containing acridine orange (Molecular Probes, Eugene, OR) (100 μg/ml) and ethidium bromide (100 μg/ml), which stain the DNA and allow for discrimination between apoptotic and non-apoptotic cells, based on breakdown of the nuclear morphology, as well as the viability of those populations. A volume of 8 μl of the dye solution was mixed with 60 μl of cell suspension, and viewed using a hemacytometer and fluorescence microscope (Diastar, Reichert-Jung). At least 200 cells were counted per sample at each timepoint.

Surface Antigen Staining and Flow Cvtometric Analyses. Triplicate samples of approximately 5x10⁵ cells were washed twice with cold PBA (phosphate buffered saline containing 1% BSA and 0.1% sodium azide). The pellets were resuspended in 100 μl of

PBA containing an amount of monoclonal antibody (Becton Dickinson: San Jose, CA) that had been previously titrated to be saturating (CD3 - 25 μl, CD25 - 15 μl, CD28 - 20 μl, CD62L - 20 μl). Samples were incubated at 4 °C for 30 minutes, washed twice more with cold PBA, and then resuspended in 500 μl PBA for immediate analysis. Flow cytometric measurements were performed using a Becton Dickinson FACScan cytometer equipped with a 15 mW, 488 nm air cooled argon-ion laser. Approximately 7500 cells were analyzed per sample. Data acquisition was performed using the FACScan Research software and then analyzed using CellQuest version 1.2 (Becton Dickinson). The QuantiBRITE (Becton Dickinson) fluorescence quantitation beads were used at each acquisition session in order to provide a consistent calibration measure to relate fluorescence intensity to the number of phycoerythrin (PE) molecules conjugated to the cell by the antibody staining process. While this does not necessarily give the exact number of receptors per cell (as the ratio of antibody binding to receptor is not known), the number of PE molecules per cell is proportional to the number of receptors per cell. This allowed for direct quantitative comparison of values of fluorescence intensity from samples stained and analyzed on different days in order to obtain kinetic information about surface receptor expression levels. Propidium iodide (PI) (2 μg/ml) was added to one unstained sample on each day in order to exclude dead cells from the analyses. A gate on a plot of side scatter vs. forward scatter (SSC vs. FSC), which corresponded to the PI negative population, was established from this sample and applied to all of the other samples on that day. This means of excluding dead cells (as opposed to adding PI to every sample) allowed for better comparison of fluorescence intensity with the QuantiBRITE calibration beads, as no PI was added to the bead samples. RESULTS Reduced O-, levels enhance T cell activation and expansion. One of the culture protocols used to investigate the effects of oxygen concentration during stimulation and culture of T cells was the oxygen "crossover" culture experiments. PBMCs were activated with either PHA or anti-CD3 mAb and then cultured under 5% or 20% O₂ gas- phase environments. After approximately a week, half of the cells were maintained in the original culture environment and the other half was switched to the alternate O₂ condition. Figures 1 A-B are representative examples of these "crossover" cultures which show the increased rate and greater extent of proliferation achieved by activating and culturing T cells in the 5% O₂ environment for both PHA and anti-CD3 mAb activated cells.

Additionally, it appears that activation in the 5% O₂ environment provides a lasting protective effect, such that cells activated under the 5% O₂ and switched to 20% O₂ proliferate better than the cells that were both stimulated and maintained under 20% O₂. Similarly, cells stimulated under 20% O₂ and then switched to 5% O₂ failed to proliferate as well as cells that were both activated and cultured under 5% O₂. Thus, the low O₂ environment improves the activation of the T cells. The greater proliferation seen in the anti-CD3 mAb-stimulated culture versus the PHA-stimulated culture is not a consistent effect and seems to vary between patient samples. Table I summarizes the differences in total fold expansion achieved for cultures of eleven experiments (9 blood samples) where cells were activated and cultured under 5% O₂ versus those activated and cultured under

20% O₂. Activation and culture under the low O₂ environment allows for a mean±S.E.M. of 5.8±0.95 fold greater total expansion over that achieved by activation and culture under the high O₂ condition. The extent of variability in fold expansion between different donor samples, while large, is not unlike the variability shown in a number of other works (Levine et al, J. Hematother., 7:437-448 (1998); Robinet et al, J. Hematother., 7:205-

215 (1998); Knazek et al, J. Immuol. Methods, 127:29-37 (1990)). There are no statistically significant differences in total fold expansion (based on Student's t-tests) between the two sources ofblood samples (non-hematological cancer or hemochromatosis patients) under 5% (p=0.44) or 20% (p=0.70) O₂ or in the ratio of proliferation between the two O₂ conditions (p=0.64). Experiments were also conducted to determine if a further decrease in O₂ concentration would provide an additional benefit in cell proliferation. Cells were activated with PHA and cultured for a week in T flasks in the 20% O₂ environment. The cells were then transferred into spinner flasks in incubators with 2.5%, 5%, and 20% O₂. Similarly increased proliferation was observed in the 2.5% and 5% O₂ cultures, with both being superior to the 20% O₂ culture (n=2; data not shown). Therefore no additional benefit was seen from culturing the cells under a 2.5% O₂ atmosphere.

Reduced O_? levels reduce T cell apoptosis. In an effort to understand the mechanism by which the cultures maintained under 5% O₂ were outperforming the cultures maintained under 20% O₂, the kinetics of apoptosis over the course of cultures under the two conditionswere examined. The fractions of apoptotic cells were identified by flow cytometry using the TUNEL assay and also by fluorescence microscopy, using a dye solution of acridine orange and ethidium bromide (AO/EtBr) to stain the DNA. Figures 2A-B show a representative example of the kinetics of the apoptotic populations, as detected by the TUNEL assay, for T cells which were stimulated and cultured under

5% or 20% O₂. While the AO/EtBr assay almost always identifies a greater fraction of cells as being apoptotic, both assays showed that T cells cultured under 20% O₂ have a greater fraction of apoptotic cells than those cultured under 5% O₂. Table II summarizes the apoptosis data from nine culture experiments (8 blood samples). The values for each condition represent the averages of multiple timepoints measured using the TUNEL and

AO/EtBr assays during the cultures. Apoptosis timepoints measured prior to 4 days following stimulation were excluded due to the large amount of cell death from non-T cell populations. Average results show an increase in apoptosis levels in the 20% O₂ cultures of 2.9 fold based on the TUNEL assay, and 1.4 fold based on the AO EtBr assay. The extent of the increase in apoptosis levels in the 20% O₂ cultures (as measured by the

TUNEL assay) does differ significantly (p=0.012, based on Student's t-test) between the sources ofblood samples (non-hematological cancer or hemochromatosis patients). The mean for the non-hematological cancer samples is a 4.4 fold increase, while the mean for the hemochromatosis patient samples is 1.8. It is not clear at this point why this difference is observed. O₂ level alters the levels of expression of the IL-2 receptor (CD25Y Whether the expression levels of several surface receptors known to be important for in vivo T cell function were affected by O₂ concentration was also examined. Kinetic analysis of CD25,

CD3, CD28, and CD62L expression was carried out using flow cytometry over the course of the cell expansion.

CD25 is the interleukin-2 receptor alpha chain (IL-2R-α). Stimulation of T cells upregulates the expression of CD25, with peak levels occurring between days 2 and 8 depending upon the stimulation method (Hviid et al, J. Clin. Lab. Immunol., 40: 163-71 (1993); Biselli et al, Scand. J. Immunol., 35:439-447 (1992); Caruso et al, Cytometry, 27:71-76 (1997)). CD25 expression levels then decrease back to amounts equivalent to background (isotype control) staining over the next several weeks of culture. For a representative experiment, Figure 3A shows that as CD25 is downregulated over the course of the culture, T cells growing under 5% O₂ had a greater fraction of cells which were positive for CD25 (defined as staining brighter than an isotype control) than the 20% O₂ cultures. Figure 3B, from the same experiment, shows that the CD25 receptor content

(per cell designated as CD25⁺) was higher on the 5% O₂ cultured cells than the 20% O₂ cultured cells during the course of the culture. Table DI summarizes the increased expression levels of CD25 on T cells in the 5% O₂ cultures for six culture experiments (4 blood samples), two of which were stimulated separately with both PHA and anti-CD3 mAb. Over the time period during culture in which CD25 was being downregulated, the samples cultured under low O₂ had an average of 56% greater fraction of cells which were positive for CD25. This difference was statistically significant, with p<l 0^"6 as determined by the Student's paired t-test. Additionally, the average level of CD25 receptor content per cell during this time was also found to be 41% greater in the samples cultured under low O₂. This difference was also statistically significant (Student's paired t-test) with p=0.004.

Effect of O-, on CD3. CD28 and CD62L expression. The effect of O₂ level on CD3 expression was examined because of its key role as a signal transduction molecule leading to T cell activation. Analysis of CD3 expression levels on T cells cultured under 5% and 20% O₂ showed no consistent variation with O₂. However, while the expression levels of CD3 declined slightly over the course of the culture period, there were consistent increases in CD3 expression levels after each feeding with fresh medium (data not shown). This suggests that accumulation or depletion of some factor in the culture medium is causing downregulation of CD3 expression.

The effect of O₂ level on the expression of CD28 and CD62L was also examined. CD28 is an important costimulatory molecule expressed on the majority of T cells and its expression is enhanced by activation (June et al., Immunol. Today, 11:211-216 (1990); Turka et al, J. Immunol, 144:1646-1653 (1990)). CD62L, also known as L-selectin, is a key protein involved in regulation of trafficking of T cells to the lymph nodes. Comparison of expression levels between T cells grown under 5% and 20% O₂ revealed no consistent trends for either CD28 or CD62L (data not shown).

O-, level affects the metabolism of T cells. Basic knowledge regarding the metabolism of T cells in culture is important for further optimization of culture media and feeding strategies for T cell ex vivo expansion, therefore examined the effect of O₂ level on on the metabolic characteristics of T cells in culture was examined. Important metabolic parameters including the specific rates of glucose uptake (q^ and lactate production (q^), as well as several key yield coefficients (Y_{hc/ u}, Y_ceiw_mmoi_gi_u. Y_mmoii_{ac ∞}ii) are vital for understanding the basic metabolism and the impact of O₂ level (Miller and Reddy, in Doyle and Griffiths, eds., Cell and Tissue Culture: Laboratory Procedures in Biotechnology, pages 133-159 (John Wiley & Sons, New York, 1998)). These parameters were examined from cultures of T cells stimulated and grown under either 5% or 20%> O₂. Over the duration of a culture, as the proliferation of the cells slowed, q^ and q^ decreased by as much as 10 fold. Both parameters were consistently higher in the culture stimulated and cultured under 5% O₂. Both the q^ and the q__c were significantly higher under 5% O₂ culture conditions (p<0.001, as determined by the Student's paired t-test) for five patient samples which included both PHA and anti-CD3 mAb stimulated cultures (see Figure 4). The mean±S.E.M. q^ values were 0.38± 0.07xl0^"10 mmol/cell/hr for the 20%) O₂ cultures and 0.56±0.65xl0^'10 mmol/cell/hr for the 5% O₂ cultures, which is a 47%) increase. The mean±S.E.M. q^ values were 0.73±0. lOxlO^"10 mmol/cell/hr for the 20% O₂ cultures and 1.07±0.13xlO-¹⁰ mmol/cell/hr for the 5% O₂ cultures, which is also a 47% increase. The mean values depicted in Figure 4 represent averages over the entire duration of the cultures and over five blood samples. The apparent yield of lactate from glucose, Y^,,, is indicative of the extent to which glucose is being metabolized anaerobically. The theoretical maximum yield is 2.0, as only two molecules of lactate can be obtained from a single molecule of glucose; however, production of lactate from other substrates such as glutamine can result in values larger than 2.0 (Miller and Reddy, in Doyle and Griffiths, eds., Cell and Tissue

Culture: Laboratory Procedures in Biotechnology, pages 133-159 (John Wiley & Sons, New York, 1998)). No statistically significant difference (Student's paired t-test) was found in Y^,, under the two O₂ concentrations (mean±S.E.M. of 1.9±0.05 for 5% O₂ cultures and mean±S.E.M. of 2.07±0.11 for 20% O₂ cultures). The

_lateen were also calculated over the six experiments. No significant differences (Student's paired t-test) in these parameters were observed between cultures under 5% versus 20% O₂. For Y_{ceϋsΛnmol g!U} the mean±S.E.M. was 3.0±0.15xl0⁸ for 5% O₂ cultures, and 3.2±0.27xl0⁸ for 20% O₂ cultures. For Y^n^n, the mean±S.E.M. was 6.7±0.34xl0^"9 for 5% O₂ cultures, and 1.9±0.76xl0^-8 for 20% O₂ cultures. One of the important technical issues in performing these experiments was to limit sampling of the cultures (and hence exposure to elevated O₂ levels) to once per day. In the initial efforts, more frequent sampling was used in order to better capture the growth kinetics; however after measuring the time needed to re-equilibrate to 5% O₂ following sampling (approx. 4-5 hrs), it became clear that this did not allow sufficient time for exposure to the low O₂ environment.

The beneficial effects of reduced O levels are observed when also using serum- free media, with or without autologous patient plasma, and when using other T-cell activation methods. All of the experiments reported above were conducted using a medium (RPMI) containing serum (fetal bovine serum, FBS). The effects of low oxygen levels on the expansion of T cells and the two T cell subpopulations (CD4+ and CD8+ subtypes) were also examined when using a serum-free medium (AIM V, GH3CO BRL, Gaithersburg, MD) supplemented with or without 2% (v/v) autologous patient plasma. Expansion of each subtype, or both subtypes, of T cells is likely to be useful in various cellular immunotherapies and gene therapies (Greenberg and Riddell, Science, 285:546- 551 (1999)). In addition, most, if not all, cellular immunotherapies and gene therapies will require the use of clinically approved serum-free media, such as AIM V. In the past, autologous patient plasma or serum has been used to enhance the expansion and improve the cellular properties of T cells (Trimble et al, Biotechnol. Bioeng, 50:521-528 (1996)).

It was found that 2% and 5% autologous plasma each had the same beneficial effect (data not shown). Therefore, 2% plasma was used in all subsequent experiments, including the experiments shown in Figures 5, 6 and 7.

As shown in Figure 5, a low (5%) O₂ level had a major beneficial effect on T cell expansion as compared to the currently used standard 20% O₂ level in serum-free medium, with or without autologous patient plasma. Specifically, the low oxygen level (5%) resulted in an improved expansion of 353% with plasma, and 6120% without plasma, over that obtained under 20% O₂. Total expansions obtained were:

5% O₂, with plasma 40,053-fold expansion

20%) O₂, with plasma 11,371 -fold expansion

5% O₂, no plasma 4,345-fold expansion

20%) O₂, no plasma 71 -fold expansion. Similar results were obtained for the two T cell subtypes, CD4+ and CD8+ (see

Figure 6). Although the ratio of CD4+ to CD8+ T cells varied with the culture time, because of the large overall T cell expansion under the low O₂ level, the beneficial effects of low versus high O₂ levels on the expansion of both CD4+ and CD8+ T cells were observed. Both CD4+ and CD8+ subpopulations experienced large cellular expansions, and the low oxygen level (5%) resulted in an improved expansion of both CD4+ and

CD 8+ T cells, with or without plasma, as compared to the high, standard oxygen level (20%).

As in the other experiments described above, a low oxygen level (5%) resulted in substantially reduced T cell apoptosis compared to 20% O₂, especially in the absence of autologous plasma (see Figure 7). Also as in the other experiments, the IL-2R was retained at higher levels during the course of expansion under a low (5%) O₂ level, with or without autologous plasma (data not shown).

The effects of O₂ levels on T cell stimulation and expansion when T cells were stimulated using both anti-CD3 and anti-CD28 antibodies (at 100 ng/ml each) simultaneously were also examined. Using both anti-CD3 and anti-CD28 antibodies simultaneously is one of the most potent means of T-cell stimulation (Greenberg and Riddell, Science, 285:546-551 (1999)), and when combined with reduced O₂ levels in the present experiments, the beneficial effects on T-cell stimulation and expansion were very large (see Figure 5). The data of Figures 5, 6 and 7 show that the beneficial effects of a low (5%) O₂ level are independent of the type of stimulant (mitogen or antigen-presenting cells) used to stimulate (activate) T cells.

DISCUSSION The data demonstrate the profound positive effect of low O₂ level (which is more representative of the in vivo tissue environment) on the ex vivo expansion of T cells. In addition to the physiological significance of this finding, the ability to significantly increase the rate and extent of proliferation by culturing under 5% O₂ conditions is an important improvement over the current methods of culturing T cells in a 20% O₂ atmosphere. The beneficial effects of low O₂ culture have been shown here to exist for T cells stimulated either with PHA, with an anti-CD3 monoclonal antibody, or with a combination of anti- CD3 and anti-CD28 monoclonal antibodies.

The O₂ crossover experiments, shown in Figures 1 A-B, were initially designed to investigate potential adaptation to damaging reactive oxygen species (ROS) by the T cells in the 20% O₂ environment during the activation period. The initial expectation was that one might observe an increased difference in proliferation between the 5% and 20% O₂ cultures if the cells were not exposed to the 20% O₂ level during activation. Thus, cells stimulated in the 5% O₂ environment would have been more susceptible to the damaging effects of the 20% O₂ when they were switched into that condition. What was observed however was the opposite. Cells stimulated in the 5% O₂ environment grew better after being switched into the 20% O₂ condition than the cells that were stimulated in the 20% O₂ environment and maintained there. This lasting protective effect seems to imply that the reason for the observed differences in proliferation rates is not simply the increased oxidative damage that the cells may suffer in the 20% O₂ environment. This is consistent with observations made by Zuckerberg et al. (Zuckerberg et al, Crit. CarMed., 22: 197- 203 (1994)) who found that even brief hypoxic exposure had lasting effects on T cell function and IL-2 mRNA levels. Several studies have linked apoptosis to increased levels of ROS. These studies include a proposed mechanism relating expression of p53 to increases in ROS which lead to cell death (Polyak et al, Nature, 389:300-305 (1997)), and a proposed mechanism of action for Bcl-2 BY antioxidant pathway (Hockenbery et al, Cell, 75:241-51 (1993); (Kane et al, Science, 262: 1274-1277 (1993)). The results show consistently higher levels of apoptosis in the 20% O₂ cultures. While the increase in percentage of apoptotic cells is consistent, it is not clear if this small difference in apoptotic death is enough to account for the large differences observed in proliferation rates between the 5% and 20% O₂ cultures. It is known that dissolved O₂ concentration can have dramatic regulatory effects on the expression of a number of different proteins expressed by lymphocytes and other cells including IL-2 (Zuckerberg et al, Crit. Car Med, 22:197-203 (1994)), erythropoietin (Jelkman, Physiol. Rev., 72:449-489 (1992)), EL-1 (Ghezzi et al., Cytokine, 3: 189-94 (1991)), tumor necrosis factor (Id), CD1 lb (Scannell et al, J. Surg. Res., 59:141-145 (1995)), CD18 (Id.), CD44 (Hasan et al, Br. J. Cancer, 77:1799-1805

(1998)) and neural cell adhesion molecule (Id.). Because the interactions of T cells by their surface receptors with target cells, cytokines, and localization target molecules are so crucial for their in vivo function, it was examined whether culture under different O₂ levels would affect the expression levels of key surface receptors. CD3, CD28, and CD62L exhibited no differences in expression due to O₂ concentration. This can be viewed as a positive result, as the increased proliferation under 5% O₂ is not offset by reduced expression of these key receptors. For CD25, the slowed downregulation kinetics observed during culture under 5% O₂ may help explain the increased growth rates and extent of proliferation. A simple set of mathematical analyses was carried out in order to examine the possible impact on proliferation of the increased levels of apoptosis and faster downregulation kinetics of CD25 under the 20% O₂ atmosphere. The impact of the greater apoptotic fraction observed under 20% O₂ was assessed by making projections of the expected cell numbers at successive generations using the assumption that the apoptotic fraction would not divide. For these calculations the averages of apoptotic cell fractions from both the TUNEL and AO/EtBr assays over the whole culture duration were used: 9% apoptotic cells for the 5% O₂ cultures and 14% apoptotic cells for the 20% O₂ cultures. After eight generations the cells cultured under 5% O₂ would be predicted to have a 1.6 fold greater expansion over those cultures under 20% O₂. A similar analysis was performed based on the assumption that cells identified as CD25 negative (less bright than the isotype control) would not divide. Because CD25 expression was measured once a day, it was assumed that one doubling occurred per day such that the CD25 positive fraction measured on a particular day was the fraction that was assumed to divide. Using this method and averaging the results from five sets of experimental data, after eight generations the cells under 5% O₂ would be predicted to have a 1.9 ± 0J4 (mean ± SD) fold greater expansion over those cultures under 20% O₂. Coupled together, these two effects predict a 3.0 ± .2 fold greater number of cells in the 5% O₂ cultures after 8 doublings. Obviously these effects may not be truly independent, as removal of growth factors can induce apoptosis. However, the predicted effects are in the same general range as the observed effects. Additionally, it has been indicated that the rate of T cell proliferation may also be dependent on the expressed surface concentration of IL-2 receptor, with those expressing higher surface concentrations entering the cell cycle more rapidly (Cantrell et al, Science, 224:1312-1316 (1984); Burke et al., Cell Immunol, 178:42-52 (1997)). Therefore, T cells cultured under 5% O₂, which possess a higher CD25 surface receptor content per cell, may be proliferating faster and also adding to this effect.

The reported effects of O₂ concentration on the glucose metabolism of mammalian cells are somewhat variable. Miller et al, J. Cell PhysioL, 132:524-30 (1987) reported decreasing specific glucose uptake rates with decreasing pO₂ over a range of 0.1% to 20% O₂ for a murine hybridoma cell line grown in continuous culture. Jan et al, Biotechnol. Bioeng, 54: 153-64 (1997) also reported decreasing specific glucose uptake and lactate production rates with decreasing pO₂ over a range of 2% to 25% O₂ with a murine hybridoma cell line grown in continuous culture. Lin et al., Biotechnol. Bioeng., 42:339- 50 (1993) reported increasing specific glucose uptake and lactate production rates with decreasing pO₂ for a recombinant CHO cell line grown in continuous culture. The data also show an increased specific glucose uptake and lactate production rates in cultures grown at low (5%) O₂ level as compared to the cultures grown at high (20%) O₂ tension. The variation in the rates of metabolism of T lymphocytes over the course of expansion is quite dramatic and it is important to understand the kinetics of this process as one attempts to develop optimal feeding protocols. The observation that proliferating lymphocytes have increased metabolic activity is well known (Brand et al, Immunobiol, 173:23-34 (1986); O'Rourke and Rider, Biochem. Biophys. Ada, 1010:342-345 (1989); Roos and Loos, Biochem. Biophys. Ada, 222:565-582 (1970)), and it is believed that these increases are necessary to meet the energy demands of rapid cell division such as formation of new biomass. No differences in the examined metabolic yields were observed, although it might be expected that changes in the availability of O₂ to the cells would affect Y^^. However, activated lymphocytes are known to use glycolysis extensively as their source of energy (Brand et al, Immunobiol, 173:23-34 (1986).

Ex vivo stimulation and expansion of T cells in a 5% O₂ level environment offers clear advantages over culturing under the standard 20% O₂ environment. Cells cultured under the 5% O₂ level have reduced levels of apoptosis and have increased expression of IL-2R. The increased proliferative capacity under the lower O₂ level is an important process improvement toward producing the large numbers of cells needed for cellular immunotherapy applications. TABLE I: Summary of Effect of 0₂ Level on T Cell Expansion

Total fold expansion of T cells from 9 blood samples stimulated with either PHA or anti-CD3 mAb, and maintained under either 5% or 20% oxygen. Expansion was based on the fraction of CD3⁺ cells in the initital samples.

* = sample from non-hematolgical cancer patient mobilized peripheral blood apheresis product

§ = sample from hemochromatosis patient therapeutic phlebotomy product TABLE II: Summary of the Effect of 0₂ Level on T Cell Apoptosis in Culture

For each culture, the reported value is the percentage of apoptotic cells as detected by either the flow cytometry-based TUNEL assay orthe fluoresence microscopy-based (AO/EtBr) assay. The values are the means of all measurements during that culture. No measurements were made during the first four days of culture due to the extensive apoptotic cell death of non-T cell populations.

* = sample from non-hematolgical cancer patient mobilized peripheral blood apheresis product

§ = sample from hemochromatosis patient therapeutic phlebotomy product

TABLE III: Summary of Effect of 0₂ Level on CD25 (IL-2R) Expression Levels

For each culture the given ratio is the mean ratio of CD25 expression for 5% 0₂ : 20% O_z averaged over all of the timepoints taken during a culture ± standard deviation. The overall means are the averages of the means from each culture ± SEM.

¹ Percent positive values indicate the fraction of cells staining brighter than the isotype control.

² Values are quantified as PE molecules bound via antibody staining using flow cytometry and the Quantibrite calibration beads.

* = sample from non-hematolgical cancer patient mobilized peripheral blood apheresis product

§ = sample from hemochromatosis patient therapeutic phlebotomy product Example 2: Effect of pH on T Cell Proliferation

In this study, the effects of pH on T cell proliferation and expression of the IL-2 receptor (IL-2R or CD25) were examined. PBMCs were stimulated with PHA and cultured at pH values of 7.0, 7.2, or 7.4. The effects of pH on the cells were studied over the two to three weeks of proliferation resulting from the PHA stimulation, rather than using single timepoint assays, in order to examine the culture kinetics over realistic time scales for ex vivo expansion.

A greater than three fold increase in the proliferation capacity of the T cells was observed for the pH 7.0 and 7.2 cultures as compared to the pH 7.4 cultures. The culture pH also affected the kinetics of the Interleukin-2 receptor (IL-2R)(CD25) downregulation process. The faster receptor downregulation in both the pH 7.2 and 1.4 cultures resulted in a greater than two fold higher fraction of IL-2R⁺ cells in the pH 7.0 cultures. Culture pH also significantly increased the fraction of apoptotic cells in the higher pH cultures, with 27% more apoptosis in the pH 7.4 cultures than the 7.2 cultures, and 49% more apoptosis in the pH 7.4 cultures than the 7.0 cultures. These effects on IL-2R expression and cellular apoptosis may help explain the observed effects of pH on T cell proliferation.

METHODS AND MATERIALS Cells and Culture Medium. Primary blood cells were obtained from patient samples from otherwise healthy donors undergoing therapeutic phlebotomies for treatment of hemochromatosis (Evanston Hospital, Evanston, IL). Most donors were in the "maintenance" phase of treatment, and therefore did not have highly elevated serum iron levels. All samples were obtained after informed consent under protocols approved by an Institutional Review Board. Peripheral blood mononuclear cells (PBMCs) from the hemochromatosis patient samples were collected using a Histopaque density gradient.

These cells were seeded at 3x10 cells/ml and cultured for 3-4 days in T-flasks using the following medium: RPMI with 100 U/ml IL-2 (Chiron: Emeryville, CA), 10% FBS (Hyclone: Logan, UT), 2 mM glutamine, 1 mM sodium pyruvate, 0J mM non-essential amino acids, 25 mMHepes, 100 U/ml penicillin, 100 μg/ml streptomycin and containing 5 μg/ml phytohemagglutinin (PHA). The pH of the media was adjusted by addition of 1

N NaOH or 1 N HCl in predetermined amounts based on titrations performed in the same 5% CO₂ environments used in the culture experiments. The media were allowed to equilibrate in the 5% CO₂ incubators overnight before use. Once the cells had begun proliferating they were subcultured to a density of 5x10⁴ cells/ml using the above pH- adjusted, pre-equilibrated media without PHA. The cultures were maintained between 5xl0⁴ and 2.5xl0⁵ cells/ml throughout the experiment to avoid accumulation of lactic acid, which would alter the pH. The cultures were maintained in a 37 °C incubator with an atmosphere of 5% CO₂ in air and 95% relative humidity. Unless otherwise noted, all reagents were from Sigma Chemical Co. (St. Louis, MO).

Cell Counting and Viability. Counting of nucleated primary cells was performed using a Coulter Multisizer (Coulter Electronics, Hialeah, FL) after treatment with Cetrimide (Sigma) to lyse the cells and release the nuclei. Viabilities were measured using the trypan blue dye exclusion method. pH Measurements. Samples of the cultures were taken once a day and measured immediately using a pH meter (Corning 340, Corning, NY). The measured values were within 0.05 pH units of the targeted values. Prior to each subculturing, the pH of the feed media was also verified.

Surface Antigen Staining and Analyses. Samples of approximately 3x10⁵ cells were washed twice with cold PBA (phosphate buffered saline containing 0.5% BSA and 0.1% sodium azide). The pellets were resuspended in 100 μl of PBA containing 15 μl of anti-CD25 monoclonal antibody (Becton Dickinson: San Jose, CA) which had been titrated to ensure saturation. Samples were incubated at 4 °C for 30 minutes, washed twice more with cold PBA, and then resuspended in 500 μl PBA for immediate analysis. Flow cytometric measurements were performed using a Becton Dickinson FACScan cytometer equipped with a 15 mW, 488 nm air cooled argon-ion laser. Approximately 10,000 cells were analyzed per sample. Data acquisition was performed using the FACScan Research software and then analyzed using CellQuest version 1.2 (Becton

Dickinson). The QuantiBRITE (Becton Dickinson) fluorescence quantitation beads were used at each acquisition session in order to provide a consistent calibration measure to relate fluorescence intensity to the number of PE molecules conjugated to the cell by the antibody staining process. While this does not necessarily give the exact number of receptors per cell (as the ratio of antibody binding to receptor is not known), the number of PE molecules per cell is proportional to the number of receptors per cell. This allowed for direct quantitative comparison of values of fluorescence intensity from samples stained and analyzed on different days in order to obtain kinetic information about surface receptor expression levels. Propidium iodide (PI) (2 μg/ml) was added to one unstained sample on each day in order to exclude dead cells from the analyses. A gate on a plot of side scatter vs. forward scatter (SSC vs. FSC), which corresponded to the PI negative population, was established from this sample and applied to all of the other samples on that day. This means of excluding dead cells (as opposed to adding PI to every sample) allowed for better comparison of fluorescence intensity with the QuantiBRITE calibration beads, as no PI was added to the bead samples.

Apoptosis Analysis. Annexin V-FITC (Pharmingen, San Diego, CA) was used in a flow cytometric-based assay to determine the fraction of apoptotic cells in the T cell cultures. Annexin V is a phospholipid binding protein that has high affinity for phosphatidylserine (PS). Early in the apoptotic process, PS is translated to the outer leaflet of the apoptotic cells, however PS is also externalized during necrosis, therefore Annexin V-FITC was used together with PI to distinguish between nonviable cells and viable apoptotic cells. Briefly samples of 200, 000 cells were washed twice with cold PB S and then suspended at 2 x 10⁶ cells/ml in a binding buffer (10 mM Hepes/NaOH, pH 7.4, 140 mM NaCl, 2.5 mM CaCl₂). Then 5 μl of the Annexin V-FITC reagent and 10 μl of PI (50 μg/ml) were added to the samples and they were incubated for 15 minutes in the dark at room temperature. Following incubation, 400 μl of binding buffer was added to the samples and they were analyzed immediately by flow cytometry.

Statistical Analyses. Analyses of the pH effects on total cell expansion, IL-2R expression, and apoptotic fractions were performed based on the relative data, using ratios of the corresponding data points from each pH condition. The ratios were calculated between each condition for all of the samples and then were compared using a two-tailed Student's t-test to determine if the mean of the ratios was significantly different from 1.

For the cell proliferation analysis, the ratios were calculated based on the total cell expansion values for each culture. For the IL-2R expression statistical analyses, the ratios of both percentage of positive cells and PE molecules per cell were calculated between the pH conditions for each of the multiple timepoints at which they were measured during the cultures. The apoptosis statistical analysis was also performed on ratios calculated from the multiple measurements made during the experimental cultures. RESULTS Proliferation of Primary Human T Cells is Affected by Culture pH. T cells from eight different donor samples were stimulated with PHA and expanded in media at pH values of either 7.0, 7.2, 1.4. Frequent feeding of the samples (every 1 to 3 days depending upon the rate of growth) using media which had been preset and pre- equilibrated to the desired pH was used to maintain the cultures within 0.05 pH units of the targeted values. The total expansion capacity of the cells from a single PHA stimulation was found to be affected by the pH of the culture environment. As is to be expected with primary T cells, a large degree of variability in total fold expansion was observed between different patient blood samples. Therefore the statistical analyses were performed on the ratios of expansion between cultures, as the relative expansion data often provide a more sensitive measure of differences in culture conditions. For example, in order to compare the expansion of the pH 7.0 cultures to the pH 7.4 cultures, the ratios of total fold expansion between the 7.0 and 7.4 cultures were calculated for each sample and were compared using a two-tailed Student's t-test to see if they varied significantly from 1. Cultures stimulated and maintained at 1.4 had significantly reduced total expansion compared to both the 7.0 (p=0.003) and 7.2 (p=0.0001) pH cultures, while cultures stimulated and maintained at pH values of either 7.0 or 7.2 did not differ significantly from each other (p=0.28). Figure 8 shows the average fold expansion ratios ± S .E.M. for the different pH culture conditions. The fold differences were calculated for each experimental culture and then averaged. Stimulation and culture at a pH of either 7.0 or 7.2 gave, on average, a greater than three fold increase in expansion as compared to stimulation and culture at pH 1.4. Figures 9A-B show the proliferation profiles from two representative samples. Figure 9A shows the results from a culture with relatively low expansion levels, and Figure 9B shows the results from a sample with relatively high expansion levels. The trend that the PBMC samples cultured at the high pH, 1.4, show lower total expansion is consistent for samples with widely varying degrees of expansion. Both cultures (Figures 9A-B) also show that during the initial days of culture, the rates of proliferation are similar between the different pH cultures. However, after day 8 in both samples (Figures 9A-B), the high pH (7.4) cultures slow down and stop growing sooner than the low and medium pH cultures. Table IV shows the expansion data for the eight samples in this study. IL-2 Receptor Expression is Altered by Variations in Culture pH. Theinterleukin- 2 receptor (IL-2R, CD25) is a surface protein expressed on T cells which is critical for proliferation of the cells. Interleukin-2 is the principle growth factor for primary T cells. Upon activation by either an antigen presenting cell or polyclonal mitogens such as PHA or an anti-CD3 monoclonal antibody, T cells experience a dramatic upregulation of IL-2R with peak levels occurring between days 2 and 8 depending upon the stimulation method (Biselli et al, Scand. J. Immunol, 35(4):439-47 (1992); Hviid et al, Clin. Lab. Immunol, 40(4): 163-71 (1993); Caruso et al, Cytometry, 27(l):71-6 (1997)). IL-2R expression levels then decrease back to amounts equivalent to background (isotype control) staining over the next several weeks of culture. The present studies show that the kinetics of the

IL-2R downregulation process are affected by the extracellular pH of the culture.

The expression of IL-2R was examined by flow cytometry and was quantified in two ways. First the percentage of the cells which were positive (defined as staining brighter than the isotype control) in their expression of IL-2R was determined. Then for that positive fraction, the average fluorescence intensity was measured and used along with the Quantibrite PE calibration beads to determine a mean number of PE molecules per cell (bound by the IL-2R antibody staining), which is proportional to the number of IL-2 receptors per cell.

Figures 10A-B show the IL-2R expression kinetics data from a representative pH experiment. The changes in both the percentage of positive cells (Figure 10A) and fluorescence intensity quantified as PE molecules per cell (Figure 10B) are depicted. The IL-2R expression data were also analyzed based on the relative differences between the 7.0, 7.2, and 7.4 pH cultures using the following ratios: 7.0:7.2, 7.0:7.4, and 7.2:7.4. For each of seven experimental cultures, multiple timepoints were analyzed for IL-2R expression. The ratios of expression levels were calculated for both the percentage of positive cells and the fluorescence intensity of the IL-2R positive cells. Statistical analyses were performed using the two-tailed Student's t-test to determine if the ratios varied significantly from 1. Analyses based on the percentage of IL-2R positive cells, show statistically significantly increased expression in the pH 7.0 cultures as compared to both the pH 7.2 and 7.4 cultures (p < 10^"6 for both cases). The difference between IL-2R expression in the pH 7.2 and 1.4 cultures was also found to be statistically significant (p=0.01). Analysis of the fluorescence intensity (translated to PE molecules per cell using the Quantibrite calibration beads) of the IL-2R positive fraction of the samples also showed statistically significantly increased expression in the pH 7.0 cultures as compared to both the pH 7.2 and 7.4 cultures (p = 1.2xl0^"8 and p = 0.001, respectively). No statistical difference was observed between the CD25 fluorescence intensities in the pH 7.2 and 7.4 cultures. Table V summarizes this analysis for all of the samples in this study.

CD3 Expression is not Altered by Variations in Culture pH. The CD3 receptor complex is the primary signal transduction molecule which is triggered upon binding of the T cell receptor with the MHC/peptide complex on a target cell, initiating activation and proliferation of the T cell. The effect of pH on the expression of CD3 was examined over the course of six experimental cultures. The expression of CD3 was evaluated in terms of fluorescence intensity, which was translated to PE molecules per cell using the Quantibrite calibration beads. No significant effect of pH on the expression of CD3 was found over the range of pH values in this study (7.0 - 1.4) (data not shown).

T Cell Apoptosis Levels are Affected by pH. The levels of apoptosis over the course of the cultures were also examined. The flow cytometry-based Annexin V assay was used to quantify the fraction of cells which had undergone extemalization of phosphatidylserine to the outer leaflet of the plasma membrane, an early marker for apoptosis. Apoptosis levels were measured daily from days 5-12, where day 0 is the day of initial stimulation. The data from one of these cultures is shown in Figure 11. Earlier days were not included because of the large extent of cell death from the non-T cell fraction of the PBMC starting samples. The apoptosis data were collected for two sample cultures and analyzed based on the relative differences between percentages of apoptotic cells at the three different pH culture conditions. The cells in the pH 1.4 cultures had 27% more apoptotic cells than the pH 7.2 cultures and 49% more apoptotic cells than the pH 7.0 cultures. These differences were statistically significant (p= 0.002 and p= 0.00001, respectively). The pH 7.2 cultures also had a statistically significant 18% increase in the fraction of apoptotic cells over the pH 7.0 cultures (p=0.0004).

DISCUSSION The results shown here demonstrate the significant effects that small variations in extracellular culture pH can have on the ex vivo expansion of T cells. The pH values used in this study, 7.0 - 1.4, are within what would be considered a normal range for T cell culture. The fact that one can obtain, on average, a greater than three-fold increase in T cell expansion by culturing the cells between pH 7.0 and 7.2, as opposed to 7.4 shows that proper attention to this fundamental culture parameter can offer significant benefits toward the goal of producing large numbers of T cells for cellular immunotherapy protocols. In examining the literature, culture pH is rarely mentioned in the T cell expansion protocols. The pH was measured, after equilibration in a 5% CO₂ incubator, of two serum free media that are commonly used for T cell culture, AIM V (Gibco BRL) (pH = 7.3) and X-VIVO 20 (BioWhittaker) (pH = 7.4). Based on the results, these media may be at a higher than optimal pH. The effect of extracellular culture pH on the downregulation kinetics of the IL-2R may offer some insight as to the mechanism behind the observed pH effects on T cell proliferation. For the pH 7.0 cultures, the IL-2R is maintained in greater numbers on a greater fraction of the cells for a longer period during culture. This correlates with the greater expansion observed in the pH 7.0 cultures. The pH 7.4 cultures show faster IL-2R downregulation kinetics, such that more of the cells become negative for IL-2R faster during culture, which also seems to be reflected in the slower growth seen at the end of the high pH cultures. The behavior of the pH 7.2 cultures however is more difficult to explain, as they have the greater expansion profiles similar to the low pH cultures but the kinetic behavior of IL-2R expression is similar to that of the more rapidly downregulating pH 7.4 cultures. The fact that similar effects of pH were not observed for the expression of CD3 shows that the alterations in IL-2R expression are not the result of global changes in T cell protein expression due to pH.

Analysis of the effect of extracellular pH on the levels of apoptosis in the T cell cultures shows statistically significant differences between all three of the pH conditions. While the relative differences in the fractions of apoptotic cells are 20-50% in size, the actual differences in the percentages of apoptotic cells are quite small and it is difficult to assess how much they contribute to the observed differences in T cell proliferation. Nevertheless, the greater fractions of apoptotic cells with increasing extracellular pH do correspond to the reduced T cell proliferation that is observed in higher pH cultures. Finally, because typical cellular immunotherapy expansion protocols employ several weeks of culture to obtain the desired number of cells, it is important to appreciate the benefits of using cell culture to examine the kinetic behavior of the pH effects on T cell expansion, as opposed to commonly employed "snapshot" methods such as use of the ³H- thymidine incorporation assay on only a single day during culture. If the effect of pH on T cell proliferation for the cultures of Figures 9A-B had been measured using the ³H- thymidine incorporation assay during the first week of culture, it is likely that no differences would have been detected. Similarly, examination of the kinetics of the IL-2R downregulation process over the duration of the culture also provides insight into the effects of pH on T cell expansion. While the longer term cell culture-based kinetic analyses are time consuming, they offer many real benefits toward understanding the actual impact of culture parameters on the process of interest, which is expanding large numbers of functional T cells.

TABLE IV. Fold expansion values from the samples stimulated and cultured at varying pH values.

TABLE V. Summary of average fold differences in IL-2R expression at varying pH values.*

Percentage of IL-2R ⁺ Cells Fluorescence Intensity of lL-2R⁺ Cells

Sample 7.0:7.2 7.0:7.4 7.2:7.4 7.0:7.2 7.0:7.4 7.2:7.4

A 1 J3±J2 1.51 ±.28 1.321.10 0.97±.01 0.95±J 0.981.10

C 2.22±.32 1.98±.33 0.88±.05 1.17±.03 0.95±.06 0.81±.03

D 2.0±.21 2.20±.26 1 J±.06 1.33±.05 1.20±.09 0.89±.04

E 2.88±.42 2.49±.44 0.86±.05 1.40±.04 1.08±.03 0.77±.02

F 1.56±.32 2.75±.74 1.65±.13 1.03±.10 1.22±.05 1.221.08

G 1.43±.18 1.58±.28 1.08±.06 0.99±.06 0.91 ±08 0.921.07

H 1.38±.1^β 1.89±.38 1.31±J 1 1.31 ±.03 1.44±.06 1.101.02

Avg.± S.E.M. 2.03±.35 2.34±.51 1.16±.07 1.20±.05 1.111.07 0.941.05

*Each reported value is the average ± S.E.M. of IL-2R expression ratios calculated over all of the measured time points during each experiment.

Claims

WE CLAIM:

1. A method of culturing T cells comprising: a) providing a sample of cells, at least some of which are T cells; b) placing the cells in a culture medium; and c) culturing the cells in the culture medium for a selected period of time under conditions effective to cause proliferation of the T cells, said conditions comprising: i) including a T cell stimulant in the culture for at least a period of time sufficient to stimulate the T cells; and ii) maintaining the pH of the culture at from about 6.8 to about 7.2 during the entire length of the culture.

2. The method of Claim 1 wherein the pH is maintained at about 7.0.

3. The method of Claim 1 wherein the T cell stimulant is a polyclonal mitogen, an antigen-presenting cell, an antibody specific for a stimulatory antigen on the surface of the T cells, or a combination thereof.

4. The method of Claim 3 wherein the T cell stimulant is phytohemagglutinin, pokeweed mitogen, concanavalin A, Staphylococcal enterotoxinB, dendritic cells, tumor cells, macrophages, an anti-CD3 monoclonal antibody, an anti-CD28 monoclonal antibody, or a combination thereof.

5. The method of Claim 1 wherein the conditions further comprise including a cytokine in the culture.

6. The method of Claim 5 wherein the cytokine is interleukin 2.

7. The method of Claim 1 wherein the sample of cells contains CD4+ T cells, CD8+ T cells, or both.

8. The method of Claim 1 wherein the T cells are genetically modified prior to beginning the culture or after stimulation with the T cell stimulant.

9. The method of Claim 1 wherein the culture medium contains serum.

10. The method of Claim 1 wherein the culture medium is a serum-free medium.

11. The method of Claim 1 wherein the T cells are to be given to a patient and the culture medium contains autologous or allogeneic plasma or serum.

12. A method of culturing T cells comprising: a) providing a sample of cells, at least some of which are T cells; b) placing the cells in a culture medium; and c) culturing the cells in the culture medium for a selected period of time under conditions effective to cause proliferation of the T cells, said conditions comprising: i) including a T cell stimulant in the culture for at least a period of time sufficient to stimulate the T cells; and ii) maintaining the oxygen level of the culture at substantially less than about 20% for at least the period of time sufficient to stimulate the T cells with the T cell stimulant.

13. The method of Claim 12 wherein the oxygen level of the culture is maintained at substantially less than about 20% for the entire length of the culture.

14. The method of Claim 12 wherein the oxygen level of the culture is maintained at from about 2% to about 10%.

15. The method of Claim 14 wherein the oxygen level of the culture is maintained at from about 3% to about 7%.

16. The method of Claim 15 wherein the oxygen level of the culture is maintained at about 5%.

17. The method of Claim 12 wherein the T cell stimulant is a polyclonal mitogen, an antigen-presenting cell, an antibody specific for a stimulatory antigen on the surface of the T cells, or a combination thereof.

18. The method of Claim 17 wherein the T cell stimulant is phytohemagglutmin, pokeweed mitogen, concanavalin A, Staphylococcal enterotoxinB, dendritic cells, tumor cells, macrophages, an anti-CD3 monoclonal antibody, an anti-CD28 monoclonal antibody, or a combination thereof.

19. The method of Claim 12 wherein the conditions further comprise including a cytokine in the culture.

20. The method of Claim 19 wherein the cytokine is interleukin 2.

21. The method of Claim 12 wherein the sample of cells contains CD4+ T cells, CD8+ T cells, or both.

22. The method of Claim 12 wherein the T cells are genetically modified prior to beginning the culture or after stimulation with the T cell stimulant.

23. The method of Claim 12 wherein the culture medium contains serum.

24. The method of Claim 12 wherein the culture medium is a serum-free medium.

25. The method of Claim 12 wherein the T cells are to be given to a patient and the culture medium contains autologous or allogeneic plasma or serum.

26. A method of culturing T cells comprising: a) providing a sample of cells, at least some of which are T cells; b) placing the cells in a culture medium; and c) culturing the cells in the culture medium for a selected period of time under conditions effective to cause proliferation of the T cells, said conditions comprising: i) including a T cell stimulant in the culture for at least a period of time sufficient to stimulate the T cells; ii) maintaining the pH of the culture at from about 6.8 to about 7.2 during the entire length of the culture; and iii) maintaining the oxygen level of the culture at substantially less than about 20%) for at least the period of time sufficient to stimulate the T cells with the T cell stimulant.

27. The method of Claim 26 wherein the pH is maintained at about 7.0.

28. The method of Claim 27 wherein the oxygen level of the culture is maintained at substantially less than about 20% for the entire length of the culture.

29. The method of Claim 26 wherein the oxygen level of the culture is maintained at from about 2% to about 10%.

30. The method of Claim 29 wherein the oxygen level of the culture is maintained at from about 3% to about 7%.

31. The method of Claim 30 wherein the oxygen level of the culture is maintained at about 5%.

32. The method of Claim 27 wherein the oxygen level of the culture is maintained at about 5%.

33. The method of Claim 26 wherein the T cell stimulant is a polyclonal mitogen, an antigen-presenting cell, an antibody specific for a stimulatory antigen on the surface of the T cells, or a combination thereof.

34. The method of Claim 33 wherein the T cell stimulant is phytohemagglutinin, pokeweed mitogen, concanavalin A, Staphylococcal enterotoxinB, dendritic cells, tumor cells, macrophages, an anti-CD3 monoclonal antibody, an anti-CD28 monoclonal antibody, or a combination thereof.

35. The method of Claim 26 wherein the conditions further comprise including a cytokine in the culture.

36. The method of Claim 35 wherein the cytokine is interleukin 2.

37. The method of Claim 26 wherein the sample of cells contains CD4+ T cells, CD8+ T cells, or both.

38. The method of Claim 26 wherein the T cells are genetically modified prior to beginning the culture or after stimulation with the T cell stimulant.

39. The method of Claim 26 wherein the culture medium contains serum.

40. The method of Claim 26 wherein the culture medium is a serum-free medium.

41. The method of Claim 26 wherein the T cells are to be given to a patient and the culture medium contains autologous or allogeneic plasma or serum.

42. A method of cellular immunotherapy comprising: administering an effective number of autologous or allogeneic T cells to a patient in need thereof, the T cells having been cultured according to any one of Claims

1-41, wherein the selected time is a time sufficient to obtain at least the effective number of T cells to be administered to the patient.

43. The method of Claim 42 wherein the sample of cells used for the culture is obtained from the patient to whom the T cells are to be administered.

44. A method of gene therapy comprising: administering an effective number of genetically-modified autologous or allogeneic T cells to a patient in need thereof, the T cells having been cultured according to any one of Claims 1-41, wherein the selected time is a time sufficient to obtain at least the effective number of T cells to be administered to the patient.

45. The method of Claim 44 wherein the sample of cells used for the culture is obtained from the patient to whom the T cells are to be administered.