Methods and compositions for activation or expansion of T lymphocytes

Abstract

The invention features a complex that binds to, stimulates, and expands a desired T cell population and facilitates separation of the target population from a sample. The complex can be gently dissociated from the binding unit after separating the desired T cell population, representing a safe and efficient approach for processing T cells for clinical use. Invention also provides methods of using such complexes as part of adoptive T cell therapy systems.

Claims

1. A polymeric moiety comprising: a hydrogel of an alginic acid-polyethylene glycol (PEG) copolymer, a signal 1 stimulus configured to induce T cell expansion, and a cation; wherein the alginic acid-PEG copolymer changes from a solid matrix into a solution or suspension in response to a sufficient decrease of cationic concentration in the environment of the copolymer.

2. The polymeric moiety of claim 1, wherein the decrease in the cationic concentration in the environment of the polymer is caused by the presence of EDTA, EGTA, sodium citrate, BAPTA, crown ether, cryptand, phenanthroline sulfonate, dipyridyl sulfonate, dioxane, DME, diglyme, or triglyme.

3. The polymeric moiety of claim 1, wherein the cation is Li.sup.+, Mg.sup.2+, Ca.sup.2+, Sr.sup.2+, Ba.sup.2+, Zn.sup.2+, Cu.sup.2+, or Al.sup.3+.

4. The polymeric moiety of claim 1, wherein the hydrogel has an elastic modulus of less than 100,000 pascals (Pa).

5. The polymeric moiety of claim 1, wherein the hydrogel has at least one cross-sectional dimension of between about 50 nm and about 50 m.

6. The polymeric moiety of claim 1, wherein the hydrogel is substantially spherical and has a diameter of between about 1 and 20 m.

7. The polymeric moiety of claim 1, wherein the signal 1 stimulus is covalently attached to an alginic acid domain of the alginic acid-PEG copolymer.

8. The polymeric moiety of claim 1, further comprising a signal 2 stimulus.

9. The polymeric moiety of claim 8, wherein the molar ratio of the signal 1 stimulus and the signal 2 stimulus is between about 1:100 and about 100:1.

10. The polymeric moiety of claim 9, wherein the molar ratio of the signal 1 stimulus and the signal 2 stimulus is about 1:1.

11. The polymeric moiety of claim 1, wherein the signal 1 stimulus is antigen-specific.

12. The polymeric moiety of claim 1, wherein the signal 1 stimulus is an antibody or antigen-binding fragment thereof.

13. The polymeric moiety of claim 12, wherein the antibody or antigen binding fragment thereof is a monoclonal antibody or antigen-binding fragment thereof, a Fab, a humanized antibody or antigen-binding fragment thereof, a bispecific antibody or antigen-binding fragment thereof, a monovalent antibody or antigen-binding fragment thereof, a chimeric antibody or antigen-binding fragment thereof, a single-chain Fv molecule, a bispecific single chain Fv ((scFv) 2) molecule, a domain antibody, a diabody, a triabody, an affibody, a domain antibody, a SMIP, a nanobody, a Fv fragment, a Fab fragment, a F(ab) 2 molecule, or a tandem scFv (taFv) fragment.

14. The polymeric moiety of claim 12, wherein the signal 1 stimulus is anti CD3.

15. The polymeric moiety of claim 8, wherein the signal 1 stimulus is anti-CD3 and/or the signal 2 stimulus is anti-CD28.

16. The polymeric moiety of claim 1, wherein the signal 1 stimulus is covalently attached to the polymeric moiety.

17. The polymeric moiety of claim 1, wherein the hydrogel comprises an average of at least one signal 1 stimulus per square m of surface area.

18. A polymeric moiety comprising: (i) a hydrogel of an alginic acid-PEG copolymer and Ca.sup.2+, and (ii) anti-CD3 and anti-CD28; wherein the alginic acid-PEG copolymer changes from a solid matrix into a solution or suspension in response to a sufficient decrease of Ca.sup.2+ concentration in the environment of the copolymer.

19. The polymeric moiety of claim 1, further comprising a separation unit, wherein the separation unit dissociates from the polymeric moiety upon the change of the alginic acid-PEG copolymer from the solid matrix into the solution or suspension.

20. The polymeric moiety of claim 19, wherein the separation unit is entrapped within the alginic acid-PEG copolymer.

21. The polymeric moiety of claim 19, wherein the separation unit comprises a magnetic material.

22. The polymeric moiety of claim 21, wherein the magnetic material is part of a nanoparticle or a microparticle.

23. The polymeric moiety of claim 18, further comprising a separation unit, wherein the separation unit dissociates from the polymeric moiety upon the change of the alginic acid-PEG copolymer from the solid matrix into the solution or suspension.

24. The polymeric moiety of claim 23, wherein the separation unit is entrapped within the alginic acid-PEG copolymer.

25. The polymeric moiety of claim 23, wherein the separation unit comprises a magnetic material.

26. The polymeric moiety of claim 25, wherein the magnetic material is part of a nanoparticle or a microparticle.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1: Workflow for the manufacture of CAR T cells using anti-CD3/anti-CD28 complexes for T Cell activation and expansion. Post expansion, the complex is dissolved away leaving T Cells magnetic-label free increasing viability and reducing negative host response.

(2) FIG. 2: Light microscope image showing complexes of the present invention, which include a magnetic bead substrate (i.e., one or more magnetic particles) encapsulated within an alginic acid hydrogel.

(3) FIG. 3: A schematic showing complexes of the present invention, which includes, in addition to a magnetic bead substrate, anti-CD3 and anti-CD28 surface moieties.

(4) FIG. 4: A timeline showing a sequence of procedures involved in an exemplary protocol for setting up an experiment using the complexes of the invention.

(5) FIG. 5: A graph showing the induction of human T cell expansion by complexes of the invention in comparison with control beads (anti-CD3/CD28-coated beads) and untreated controls.

(6) FIGS. 6A and 6B: Flow cytometry plots showing the phenotype of T cells expanded by complexes of the invention. FIG. 6A shows the expression of CD3 versus the uptake of propidium iodide (PI) as a marker of cell death after a twelve-day expansion protocol. FIG. 6B shows the relative expression of CD8 versus CD4 after a twelve-day expansion protocol.

(7) FIGS. 7A and 7B: Flow cytometry plots showing the phenotype of T cells expanded by control beads (anti-CD3/CD28-coated beads). FIG. 7A shows the expression of CD3 versus the uptake of propidium iodide (PI) as a marker of cell death after a twelve-day expansion protocol. FIG. 7B shows the relative expression of CD8 versus CD4 after a twelve-day expansion protocol.

(8) FIGS. 8A and 8B: Bar graphs showing the percent (%) change in populations of CD4 cells (FIG. 8A) and CD8 cells (FIG. 8B) over the course of expansion with control beads versus complexes of the invention. The data were derived from flow cytometry data analyzed as shown in FIGS. 6 and 7.

(9) FIGS. 9A-9E: Flow cytometry plots showing the phenotype of T cells at day 0 (a starting population).

(10) FIG. 9A shows expression of CD8 versus CD4. FIG. 9B shows expression of CD45RA versus CD3. FIG. 9C shows expression of CD45RO versus CD3. FIG. 9D shows expression of CD62L versus CD45RA. FIG. 9E shows expression of CCR7 versus CD45RA.

(11) FIGS. 10A-10E: Flow cytometry plots showing the phenotype of control bead-expanded T cells at day 12 of culture. FIG. 10A shows expression of CD8 versus CD4. FIG. 10B shows expression of CD45RA versus CD3. FIG. 10C shows expression of CD45RO versus CD3. FIG. 10D shows expression of CD62L versus CD45RA. FIG. 10E shows expression of CCR7 versus CD45RA.

(12) FIGS. 11A-11E: Flow cytometry plots showing the phenotype of T cells expanded by the complexes of the invention at day 12 of culture. FIG. 11A shows expression of CD8 versus CD4. FIG. 11B shows expression of CD45RA versus CD3. FIG. 11C shows expression of CD45RO versus CD3. FIG. 11D shows expression of CD62L versus CD45RA. FIG. 11E shows expression of CCR7 versus CD45RA.

(13) FIGS. 12A and 12B: Bar graphs showing the percent of CD3.sup.+CD45RO.sup.+ cells at day 0 versus day 6 of expansion with control beads versus complexes of the invention. The data were derived from flow cytometry data analyzed as shown in FIGS. 10 and 11.

(14) FIG. 13: Bar graph showing the percent of CD62L.sup.+CD45RA.sup.+ (nave) T cells at day 0 versus day 6 of expansion with control beads versus complexes of the invention. The data were derived from flow cytometry data analyzed as shown in FIGS. 10 and 11.

(15) FIG. 14: Bar graph showing the percent of CCR7.sup.+CD45RA.sup.+ (nave) T cells at day 0 versus day 6 of expansion with control beads versus complexes of the invention. The data were derived from flow cytometry data analyzed as shown in FIGS. 10 and 11.

(16) FIG. 15: Bar graph showing the concentration of interferon-gamma (IFN-) secreted at day 6 of expansion with control beads versus complexes of the invention, as measured by enzyme-linked immunosorbant assay (ELISA).

(17) FIG. 16: Bar graph showing the concentration of interleukin-4 (IL-4) secreted at day 6 of expansion with control beads versus complexes of the invention, as measured by enzyme-linked ELISA.

(18) FIG. 17: A series of light microscope images showing T cells expanded by complexes of the invention in comparison with control beads.

DETAILED DESCRIPTION OF THE INVENTION

(19) The invention provides a complex that binds to, stimulates, and expands a desired T cell population and facilitates separation of the target population from a sample. The complex can be gently dissociated from the binding unit after separating the desired T cell population, representing a safe and efficient approach for processing T cells for clinical use. The invention also provides methods of using such complexes as part of adoptive T cell therapy systems.

(20) Binding Unit

(21) The invention features a complex having a binding unit and a separation unit. The binding unit includes a binding moiety that binds a component on a T cell surface. The binding moiety is linked to a polymeric moiety that includes a cation-sensitive hydrogel. The binding moiety and the polymeric moiety, together, form the binding unit.

(22) T Cell Binding Moieties

(23) In one aspect, the invention features a binding moiety that can bind to a surface component of a T cell. Preferably, this binding event can lead to signal transduction within the T cell, resulting in activation and/or proliferation of the target cell. Several T cell surface molecules are known to have these downstream effects. In general, any ligand that induces clustering of the T cell receptor (signal 1) is able to stimulate that T cell to proliferate and, depending on a secondary signal (signal 2), differentiate towards a particular functional phenotype. Preferably, the signal 1 agent of the present invention is anti-CD3, and the signal 2 agent of the present invention is anti-CD28. Other examples of signal 1 stimuli include MHC-I or MHC-II, and agonists of various other T cell receptor components known in the art. Other examples of signal 2 stimuli include antigen-presenting cell surface molecules that are agonists of co-stimulatory molecules (e.g., CD80, CD86, CD40, ICOSL, CD70, OX40L, 4-1BBL, GITRL, LIGHT, TIM3, TIM4, ICAM1, or LFA3), antibodies toward T cell costimulatory surface molecules other than CD28 (e.g., CD40L, ICOS, CD27, OX40, 4-1 BB, GITR, HVEM, Galectin 9, TIM-1, LFA-1, and CD2), antigen-presenting cell surface molecules that are agonists of co-inhibitory molecules (e.g., CD80, CD86, PD-L1, PD-L2, B7-H3, B7-H4, HVEM, ILT3, or ILT4), and antibodies against co-inhibitory molecules (e.g., CDTL-4, PD-1, BTLA, or CD160). Signal 2 stimulation has been shown to affect multiple aspects of T cell activation. It lowers the concentration of anti-CD3 required to induce a proliferative response in cultures and enhances cytokine production to direct T cell differentiation pathways. Importantly, costimulation helps activate cytolytic potential of CD8 T cells. Other molecules, including but not limited to CD2 and CD137, can be targeted in activation protocols used to activate and expand various T cell populations, such nave and memory T cells, T helper cells, regulatory T cells, Natural Killer T cells and Cytotoxic T Lymphocytes from mouse and human samples. Additional examples of antibodies, ligands, and other agents useful as signal 1 and signal 2 stimuli for use in the present invention are described in WO2003/024989.

(24) Antigen-Specific T Cell Engagement

(25) In one aspect of the invention, a binding unit may bind a T cell receptor in an antigen-specific manner analogous to the binding that occurs between the T cell receptor and a peptide-MHC on an antigen presenting cell. Synthetic methods to engage a T cell in an antigen specific manner (i.e., in the absence of a natural antigen presenting cell) include MHC class I and MHC class II multimers (e.g., dimers, tetramers, and dextramers). Illustrative examples are described in U.S. Pat. No. 7,202,349 and U.S. 2009/0061478. Multimerization of MHC-peptide complexes functions to enhance avidity of interaction between peptide-MHC and T cells, which increases the potency of signal 1 transduction. Another means to achieve multivalent presentation of antigen-specific T cell receptor ligands is by tethering the ligands to a surface of a complex. Affinity, in contrast to avidity, describes the strength of binding of each individual molecule. The affinity of peptide-MHC for a T cell receptor can vary dramatically and dictates the downstream effect of signal 1 stimuli, as discussed below. Antigens, and peptides thereof, for use in the present invention include, but are not limited to melanoma antigen recognized by T cells (MART-1), melanoma GP100, the breast cancer antigen, Her-2/Neu, and mucin antigens. Other relevant antigens and sources thereof are described, for example, in U.S. Pat. No. 8,637,307.

(26) Effects of the Degree of T Cell Binding

(27) The relative degree of signal 1 and signal 2 binding can influence TCR signal transduction and lead to varying downstream phenotypic effects. For example, high affinity with low avidity signal 1, in the absence of signal 2, primes a nave CD4 T cell to turn on Foxp3 expression and differentiate toward a regulatory phenotype (Gottschalk et al., Journal of Experimental Medicine 207 (2010): 1701).

(28) Alternatively, in response to a high avidity signal 1 stimulus without sufficient signal 2 stimulus, a nave T cell may be more likely to undergo exhaustion, leading to functional anergy (see Ferris et al., J Immunol August 15; 193 (2014): 1525-1530). Either of these effects would be undesireable in the context of cancer adoptive immunotherapy but may be helpful in when priming regulatory T cells for autoimmune treatment. Thus, the relative contribution of signal 1 and signal 2 can be rationally modulated, depending on the application, by exposing the functionalized complex to a desired ratio of binding moieties.

(29) The binding moieties may be coupled to the same surface or to separate surfaces. In a preferred embodiment, a signal 1 stimulus and a signal 2 stimulus are immobilized on the surface at a 1:1 ratio. In certain aspects of the present invention, a signal 1 stimulus and a signal 2 stimulus are immobilized on the surface at a ratio of other than 1:1 (e.g., between about 1:100 and about 100:1, between about 1:10 and 10:1, between about 1:2 and 2:1). In other aspects, the ratio of signal 1 stimulus to signal 2 stimulus immobilized on the surface is greater than 1:1, or less than 1:1.

(30) The effect of relative strength of signaling between signal 1 and signal 2 also depends on the activation state of the target T cell. For instance, nave T cells react differently to T cell receptor stimulation and costimulation than antigen-experienced T cells. A skilled artisan would appreciate this effect when selecting a configuration of binding units of the present invention, especially in cases of chronic infection or aberrant immune tolerance, and configure the binding unit accordingly.

(31) Size and Shape of the Complex

(32) The complex may be formed into any shape compatible with contacting the surface of a T cell. For this reason, however, it is often preferable for a complex to have a high surface area to volume ratio to maximize the available binding surface. Artificial antigen presenting cell platforms having different sizes and shapes have been evaluated for various functional benefits (see Fadel et al., Nano Letters 8 (2008): 2070-2076; Sunshine et al., Biomaterials 35 (2014): 269-277). The shape of the structure may be any suitable shape, such as elongated like a wire, tubular, i.e., having a lumen, planar, or spherical. An immune synapse can range in size from less than 50 nm to about 20 m. Preferably, the complex of the present invention is spherical, and the size is between 50 nm and 20 m. For example, the complex can be the size of antigen presenting cells, such as dendritic cells or macrophage, which range from about 10-20 m and 10-20 m, respectively. Alternatively, the complex can include a larger matrix, such as a porous scaffold, which can be mechanically exposed, e.g. dipped, into a suspension of cells. Methods for synthesizing such scaffolds, including by using alginate, are known in the art.

(33) Polymeric Moiety

(34) The binding unit of the present invention can be linked to the separation unit by conjugation to a polymeric moiety. Preferably, this conjugation occurs through a covalent reaction.

(35) Polymeric moieties useful in the present invention include hydrogels that can be dissolved by changing the ionic composition of their environment. Hydrogels of the invention are formed from alginic acid conjugated to a polyalkylene oxide, e.g., PEG, as generally described in WO 2012/106658. Conjugation of alginic acid to multifunctional PEG confers greater mechanical strength compared with that achieved solely by ionic crosslinking of alginate. Thus, the mechanical properties (e.g. stiffness) can be modulated by increasing the number of functional groups on each PEG molecule. PEG is useful as part of the present invention because of its superior hydrophilic properties, which prevent protein adsorption to the complex of the present invention. Adsorption of serum proteins onto the complex can result in aberrant signaling pathways in adjacent cells, such as those caused by Fc receptor engagement. The hydrophilic properties of PEG are also functional to maintain a high diffusivity within the complex interior, such that ionic chelators can rapidly access the complex interior to quickly sequester rigidity-maintaining cations. Thus, incorporation of branched PEG molecules within the hydrogel ensures a rapid dissolution of the hydrogel structure upon exposure to appropriate stimuli. Alternatively, any other biocompatible hydrophilic polymer (e.g. polyvinyl pyrrolidone, polyvinyl alcohol, and copolymers thereof) can be substituted (see, e.g., U.S. Pat. No. 7,214,245).

(36) A polymeric moiety of the invention can have one or more mechanical properties (e.g., elastic modulus, Young's modulus, compression modulus, or stiffness) suitable for T cell expansion. For example, the mechanical properties of the polymeric moiety can be tailored to be suitable to allow an population of T cells (e.g., a population containing expanded T cells) to retain one or more characteristics of a nave phenotype (e.g., as described in the Methods section) after contacting a complex of the invention. In some cases, the elastic modulus of the polymeric moiety is between 100 pascals (Pa) and 100,000,000 Pa (e.g., from 100 Pa to 1,000 Pa, from 1,000 Pa to 10,000 Pa, between 10,000 Pa to 100,000 Pa, between 100,000 Pa and 1,000,000 Pa, 1,000,000 Pa and 10,000,000 Pa, or 10,000,000 Pa and 100,000,000 Pa, e.g., less than 1,000,000 Pa, less than 900,000 Pa, less than 800,000 Pa, less than 700,000 Pa, less than 600,000 Pa, less than 500,000 Pa, less than 400,000 Pa, less than 300,000 Pa, less than 200,000 Pa, less than 100,000 Pa, less than 50,000 Pa, or less than 10,000 Pa).

(37) A reference to alginic acid is also a reference to a salt form, e.g., sodium alginate, unless otherwise noted. Suitable alginic acid is 20 kDa medium viscosity alginic acid. Alginic acid can be present in the polymer moiety in a percentage (e.g., a weight-by-volume percentage) between 0.01 and 10% (e.g., from 0.1% to 0.15%, from 0.15% to 0.2%, from 0.2% to 0.3%, from 0.3% to 0.4%, from 0.4% to 0.5%, from 0.5% to 0.6%, from 0.6% to 0.7%, from 0.7% to 0.8%, from 0.8% to 0.9%, from 0.9% to 1.0%, from 1.0% to 2%, from 2% to 5%, from 5% to 7.5%, or from 7.5% to 10%, e.g., about 0.2%, 0.25%, 0.3%, 0.35%, 0.4%, 0.45%, 0.5%, 0.6%, 0.7%, 0.75%, 0.8%, 0.9%, 1.0%, 1.5%, 2.0%, 2.5%, 3%, 3.5%, 4%, 4.5%, 5%, or 10%). The alginic acid content of the polymeric moiety will influence the stiffness of the polymeric moiety according to known principles (e.g., cation content of the polymeric moiety). The stiffness of alginate polymeric moieties can be varied while maintaining a constant or near-constant density according to known methods.

(38) Polyalkylene oxides, e.g., PEG and polypropylene oxide, are known in the art. Linear or branched, e.g., 4-arm or 8-arm, polyalkylene oxides, e.g., PEG, may be employed. The polyalkylene oxide, e.g., PEG, preferably has a molecular weight between 10 kDa and 20 kDa. An exemplary ratio of polyalkylene oxide, e.g., PEG, to alginic acid is 1:2 by weight.

(39) Alginic acid naturally possesses multiple carboxyl groups that provide convenient groups for conjugation to polyalkylene oxide, e.g., PEG, and/or binding moieties. The polyalkylene oxide, e.g., PEG, and binding moiety will naturally possess or be modified to possess an appropriate group to conjugate to a carboxyl group. Suitable groups include amine groups, which are often found in binding moieties that include amino acids or can be introduced into binding moieties and polyalkylene oxides, e.g., PEG. For example, amine-terminated polyalkylene oxide, e.g., PEG, can be employed. In other embodiments, a linker may be use to conjugate appropriate groups on the polyalkylene oxide, e.g., PEG, or binding moiety to carboxyl groups on the alginic acid. In the hydrogel, a single polyalkylene oxide, e.g., PEG, may be conjugated to one or more alginic acid molecules. When a polyalkylene oxide binds to more than one alginic acid, the number of such crosslinks in the composition may or may not be sufficient to form a gel. The binding moiety can bind to either the alginic acid directly or to a polyalkylene oxide, e.g., PEG, bound to alginic acid.

(40) The hydrogel forms by noncovalent crosslinking of the alginic acid with a cation, e.g., Li.sup.+, Mg.sup.2+, Ca.sup.2+, Sr.sup.2+, Ba.sup.2+, Zn.sup.2+, Cu.sup.2+, or Al.sup.3+. A preferred cation is Ca.sup.2+. Gelation of hydrogels of the invention may be reserved by contact with a chelator for the cation, e.g., EDTA, EGTA, sodium citrate, BAPTA, crown ether, cryptand, phenanthroline sulfonate, dipyridyl sulfonate, dioxane, DME, diglyme, or triglyme. Preferably, the chelator is a bioinert molecule such as EDTA, which is well-known not to interfere with cell growth and proliferation pathways at concentrations relevant for complex dissolution.

(41) Separation Unit

(42) The complex of the present invention can be formulated by gelation of the hydrogel around a separation unit. For example, the hydrogel may be formulated as a coating on the outer surface of the separation unit. Preferably, the separation unit can include multiple smaller structures that are encapsulated within the hydrogel matrix, as shown in FIGS. 2 and 3.

(43) The separation unit can be any structure that enables the user to physically manipulate the position of the target cells, once bound with the complex of the invention. Preferably, the separation unit can include one or more magnetic elements to enable physical manipulation, separation, or purification of target cells. Preferably, the magnetic element is encapsulated within a polymeric or hydrogel structure, such that its movement along a magnetic field functions to pull the complex, along with any attached cell, through the field. The magnetic element can be one or more microparticles or nanoparticles having magnetic properties. Methods for making magnetic microparticles and nanoparticles and integrating them within polymeric complexes are described in the art (WO 2003/071561, WO 2003/086660, WO 2004/060580, U.S. Pat. No. 7,691,285, U.S. P.G. Pub. 2004/0099954, WO 2001/017662, WO 2002/043708). Magnetic particles can be ferrite, oxide, or metallic, and they can be superparamagnetic. Magnetic nanoparticles are often iron oxides, such as magnetite (Fe.sub.3O.sub.4) or maghemite (Fe.sub.3O.sub.3). Magnetic particles can have a shell, such as a polymer shell. The precipitation of iron oxides can occur under non-aqueous conditions (U.S. Pat. No. 4,677,027) and subsequently be converted to aqueous conditions (U.S. Pat. No. 5,160,725) or can occur exclusively in aqueous solutions (U.S. Pat. No. 4,329,241). Aqueous formulations may preferable in biological applications because of toxicological considerations (U.S. Pat. No. 4,101,435). Wet chemical synthesis of the iron oxide crystallites can precede coating with polymer components (core-shell method) or can occur in the presence of the polymer (one-pot method).

(44) In one aspect of the invention, the separation unit can be removed from the target T cell population. This separation can occur in response to exposure of the complex to a cation chelator as discussed above, which results in release of the separation unit from the binding unit.

(45) Synthesis

(46) The compositions of the present invention can be synthesized by methods currently known in the art. Preferably, complexes can be formulated as microparticles by one of several known methods (e.g. microemulsion). Techniques useful in encapsulating nanoparticles (i.e. magnetic nanoparticles) within polymeric microparticles are disclosed, for example, in U.S. 2011/0229580. Other methods that can be used to entrap a separation unit (i.e. magnetic particles) within a hydrogel include coating by, e.g., dipping, painting, or spraying, with a liquid alginic acid composition and then contacting with cations for crosslinking, e.g., by dipping painting or spraying. Methods for synthesizing alginate beads for various applications are known in the art (see, for example, Hatch, et al., Langmuir 27 (2011): 4257-4264 and Sosnik, ISRN Pharmaceutics (2014):1-17). Alginate PEG co polymer is synthesized using aminated PEG combined in a batch reaction of Alginic acid, EDC, and sulfo NHS. Antibody conjugation is performed using standard bio conjugation techniques including malidimide/thiol and EDC/NHS linking. Other useful conjugation methods are described in Hermanson et al., (2013) Bioconjugate Techniques: Academic Press.

(47) Methods of Use

(48) The invention features a method of activating a population of T cells using the complex described above. In one aspect of the invention, T cell activation or expansion can be performed by isolation of T cells and subsequent stimulation followed and/or expansion. Prior to expansion, a source of T cells can obtained from a subject. T cells can be obtained from a number of sources, including peripheral blood mononuclear cells, bone marrow, lymph node tissue, spleen tissue, and tumors. In certain embodiments of the present invention, a T cell line available in the art may be used. T cells can also be obtained from a unit of blood collected from a subject using any number of techniques known to the skilled artisan, such as Ficoll separation or through a PERCOLL gradient. Cells from the circulating blood of an individual can be obtained by apheresis or leukapheresis. The apheresis product typically contains lymphocytes including T cells, monocytes, granulocytes, B cells, other nucleated white blood cells, red blood cells, and platelets. The cells collected by apheresis may be washed to remove the plasma fraction and to place the cells in an appropriate buffer or media for subsequent processing steps. T cells can be enriched prior to treatment with the complex of the present invention through conventional techniques such as magnetic bead negative selection or fluorescence activated cell sorting (FACS). T cells can be antigen-specific, antigen-nonspecific, or tumor specific. They can include populations of CD4 cells, CD8 cells, NK T cells, or, for autoimmune or transplant rejection therapies, regulatory T cells.

(49) In addition to purification steps, T cells that have been obtained from a subject may be further processed prior to or after incubation with the complex of the present invention. For example, cells can undergo genetic engineering to equip them with certain functional characteristics, such as in the process of chimeric antigen receptor (CAR) engineering, which equips a patient's cells to recognize cancer antigens. In this case, the CAR engineering procedure may be performed prior to treatment with the present complexes in order to expand the initially small number of CAR T cells into a substantial activated population. Other genetic procedures used for adoptive therapy are discussed in Rosenburg et al. (Nat Rev Cancer 8 (2008): 299-308). In other cases, such as those not involving genetic modification, such as bispecific T cell engager (BITE) technology (see WO2011/057124), such procedures may be performed after the cells have been expanded using the present invention. Other non-genetic processing procedures include but are not limited to treatment with IL-2, IL-4, IL-7, IL-10, IL-12, IL-15, IL-21, or TGF-.

(50) Sample Treatment

(51) In one aspect of the invention, the starting T cell population is incubated with the complexes. The ratio of complexes to starting cell number will depend on the size and shape of the complex. For this purpose, it will be understood by a person of skill in the art that the ratio of surface area between target T cells and complexes is a significant factor governing the degree of resulting T cell activation. Preferably, the surface area ratio between target cells and complexes is between about 1:100 and about 100:1 (e.g., about 1:100, about 1:80, about 1:50, about 1:25, about 1:10, about 1:5, about 1:4, about 1:3, about 1:2, about 1:1, about 2:1, about 3:1, about 4:1, about 5:1, about 10:1, about 25:1, about 50:1, about 80:1, or about 100:1). Other factors which should be considered when incubating starting T cells with complexes are the density of binding moieties on the complex surfaces, the specific binding moieties chosen (i.e., their binding affinity, EC50, receptor concentration on T cells, etc.). Additionally, it will be appreciated that as T cells proliferate over the course of multiple days, their volumetric requirements will increase, which may limit the total volume that can be allocated to complexes.

(52) Incubation Procedures

(53) Methods of the present invention include incubating T cells with complexes in a suitable vessel. Such cell-culture vessels are known in the art and can include cell culture plates, flasks, or bioreactors of any suitable size. Cell culture vessels are preferably sterile, and may be configured for optimal gas exchange or media exchange, such as perfusion capable systems, which are known in the art. T cells can be seeded at a concentration of between about 0.210.sup.6 and 1010.sup.6 cells/ml. Complexes may require special ionic conditions, e.g., to maintain a solid structure in solution. For example, cell culture media can be supplemented with ions, such as Ca.sup.2+, through addition of salts, e.g., CaCl.sub.2. Ions can be present at any physiologically suitable concentration (e.g., 1.0 nM-100 mM, e.g., 1.0 M to 10 mM, e.g., 0.1 mM, 0.2 mM, 0.5 mM, 1.0 mM, 2 mM, 3 mM, 4 mM, 5 mM, 6 mM, 7 mM, 8 mM, 9 mM, or 10 mM).

(54) Kinetics

(55) Ex vivo T cell expansion protocols are well-developed, especially for human samples, and can often yield cell number increases in the hundred-fold range, over the course of multiple weeks of culture. Thus, multiple rounds of expansion may be required to overcome constraints on physical space and media nutrient depletion. To accommodate these constraints, the complexes of the present invention can be dissolved (i.e., the separation unit removed) multiple times (e.g., once, twice, or 3-12 times) over the course of a T cell expansion protocol. Such dissolution/re-administration cycles can be achieved by washing out cation chelators from the media by centrifugation or other known methods after each dissolution. Alternatively, additional complexes can be introduced into the cultures without removing separation units of the existing complexes, thus bypassing the need for media changes. If necessary, beads encapsulating magnetic particles, or excess magnetic particles alone, can be removed from cultures using well-established magnetic separation procedures.

(56) In one aspect of the present invention, the T cell expansion protocol proceeds for a predetermined length of time that is suitable to generate a desired number of T cells, a representative phenotype, or both. Alternatively, the phenotypic properties of T-cell populations of the present invention can be monitored by a variety of methods, and the isolation can be performed after the desired phenotype is acquire. Relevant phenotypes are metabolic changes such as biochemical or morphological changes (e.g., change in frequency of cell division, change in cytokine expression profile, change in median cell diameter, change in surface molecule expression, or change in cellular motility). Assays for monitoring such changes include standard flow cytometry methods, ELISA, microscopy, migration assays, metabolic assays, and other techniques known by those skilled in the art.

(57) In some cases, the methods of the present invention allow for expansion of a population of cells containing nave T cells, central memory T cells, and/or effector memory cells. As discussed above, the composition of the polymeric moiety can influence the percentage and/or total number of nave T cells, central memory T cells, and/or effector memory cells present in a population (e.g., an expanded population). For example, by culturing a starting population of T cells with a complex having a low-elastic modulus polymeric moiety (e.g., an alginic acid-based polymeric moiety having an elastic modulus less than 100,000 Pa, less than 50,000 Pa, less than 10,000 Pa, less than 5,000 Pa, or less than 1,000 Pa, at one or more points in time over the course of incubation with the cells) a resulting population of T cells may include a high percentage of nave T cells (e.g., nave CD4.sup.+ T cells or nave CD8.sup.+ T cells). In some instances, nave T cells are characterized as having one or more properties (surface markers or secreted cytokines) of Table I (see Examples).

(58) Purification

(59) Alginic acid provides a physical anchor to entrap the separation unit. In the presence of a cation, e.g., calcium, alginic acid is crosslinked and solid. Upon completion of T cell processing, the hydrogel can be dissolved using a chelator to release the separation unit from the T cell. In some aspects, the separation unit includes magnetic particles that are not permissible as part of clinical samples (e.g. the resulting activated T cell suspension). By releasing the separation unit (e.g. magnetic particles) by cation chelation, the resulting T cells are efficiently cleaned of potentially harmful impurities and ready for infusion into a patient in need thereof.

(60) EDTA is a well-characterized calcium chelator of use in the present invention. EDTA can be used at 2 mM in the form of a physiologically insert buffer that preferably also contains 137 mM NaCl, 2.7 mM KCl, 25 mM Hepes and 0.75 mM Na.sub.2H.sub.2PO.sub.4.2H.sub.2O.

(61) Cell culture media can be exchanged for the ion chelator solution, e.g., EDTA buffer, potentially by centrifugation and subsequent resuspension of cell pellet, e.g., in EDTA buffer. Other methods may also be used. The cell/ion chelator suspension may also be agitated, e.g., by pipetting or vortexing for about 5 seconds. Cell suspension can be placed on a magnetic stand for long enough to allow separation of all magnetic particles from cells (5 minutes is typical). During this step, the hydrogel dissolves, cells are released, and the magnetic particles are attracted to the side of the vessel being used (e.g. tube). With the cell suspension still on the magnetic stand, the supernatant can be changed to a clean tube, while magnetic particles can be left behind. This step can be repeated again in the initial tube in order to recover any cells that remain attached to magnetic particles. The isolated cells can then be returned to cell culture media or isotonic solution (e.g., for administration into a patient).

EXAMPLES

(62) FIG. 1 shows a workflow for the manufacture of chimeric antigen receptor (CAR) T cells using CD3/CD28 complexes for T cell activation and expansion. First, T cells can be isolated from the patient and engineered to express CARs that recognize cancer cells. At this point, anti-CD3/anti-CD28 complexes can be added to CAR T cells, which will facilitate activation and expansion. Post expansion, the complex can be dissolved away leaving T cells magnetic-label free, increasing viability and reducing the potential for a negative host response.

(63) FIG. 2 is a light microscope image showing complexes of the present invention, which include a magnetic bead substrate coated with an alginic acid hydrogel. The hydrogel is a novel dissolvable polymer, which has the ability to be conjugated with ligands to facilitate the attachment and subsequent release of targeted cell types.

(64) FIG. 3 is a schematic of a complex of the present invention, which includes a magnetic bead substrate coated with an alginic acid hydrogel, which has been conjugated with anti-CD3 and anti-CD28 ligands.

(65) FIG. 4 shows an exemplary timeline of procedures carried out to prepare a T cell population with hydrogel complexes of the present invention. At day 0, cells are seeded in a 24 well plate with hydrogel complexes. At day 2, additional media is added to the cells and the culture is mixed by gently pipetting up and down five times. At day 5, cells are counted and transferred to a six-well plate, where they are cultured in experiment-specific conditions or allowed to continue growth.

(66) FIGS. 5-17 show characteristics of T cell populations treated using methods of the invention, in comparison to conventional methods involving commercially available anti-CD3/CD28 magnetic beads (control). Primary human T lymphocytes were seeded at a density of 110.sup.6 cells/mL and cultured in advanced RPMI medium supplemented with fetal bovine serum, glutamate, HEPES, and recombinant human IL-2. For the hydrogel complex-treated cells, the media was further supplemented with 2 mM CaCl.sub.2. On day one of culture, cells were stimulated with either control beads at a concentration of one bead per cell or with two different formulations of hydrogel complexes coated with anti-CD3 and anti-CD28 antibodies (Affymetrix/ebioscience). Subsequently, media was replenished every 2-3 days without further addition of control beads or hydrogel complexes. In response to stimulation, T cells expanded potently and in a sustained manner (FIG. 5). Cells proliferated to a similar degree whether they were treated with control beads or hydrogel complexes. In unstimulated cells, no cell expansion occurred.

(67) After a twelve-day T cell expansion, cells were stained for viability using propidium iodide. Cell viability was high (greater than 99%) in both hydrogel complex-expanded T cells (FIG. 6A) and control bead-expanded T cells (FIG. 7A). Live samples of each expanded population were then analyzed for surface expression of T cell markers CD3, CD4, and CD8 (FIGS. 6-8). Under each expansion condition, greater than 97% of cells expressed CD3 (FIGS. 6A and 7A). Cells expanded using hydrogel complex formulation #1 were 33.24% CD4-positive and 62.84% CD8-positive (FIG. 6B). T cells expanded by control beads preferentially expressed CD4, relative to T cells expanded by hydrogel beads (FIG. 7B). Control bead-expanded T cells were 88.67% CD4-positive and 7.18% CD8-positive (FIG. 7B). The percentage of cells expressing neither CD4 nor CD8 was less than 5% in all conditions tested. The effect of treatment on CD8 and CD4 expression by T cells is further exemplified by FIGS. 8A and 8B, which show the percent change in CD4 and CD8 expression, respectively, at day 6 of expansion relative to the initial expression levels (e.g., of the starting population at day 0). In hydrogel complex-expanded groups, the percentage of CD4.sup.+ T cells decreased by more than 10% from day 0 to day 6, whereas the percentage of CD4.sup.+ T cells in the control bead-treated groups increased by about 4% (FIG. 8A). Furthermore, the percentage of CD8.sup.+ T cells in the hydrogel complex-expanded groups increased by more than 30%, whereas the CD8.sup.+ population of control-treated cells diminished by more than 10% (FIG. 8B).

(68) As summarized in Table 1, below, one current model of T cell development suggest that CD45RA is expressed on nave T cells and that after antigen experience, central memory T cells and effector memory T cells gain expression of CD45RO and lose expression of CD45RA. According to this model, CD45RA.sup.+ cells that simultaneously express CD62L or CCR7 are characterized as nave T cells.

(69) TABLE-US-00001 TABLE 1 Current model of memory T cell subtype marker expression Secretion T cell subtype Surface marker expression profile profile Nave T cells CD45RA.sup.+ CD45RO.sup. CCR7.sup.+ CD62L.sup.+ IL-4.sup. IFN-.sup. Central memory CD45RA.sup. CD45RO.sup.+ CCR7.sup.+ CD62L.sup.+ IL-4.sup. IFN-.sup. T cells Effector memory CD45RA.sup. CD45RO.sup.+ CCR7.sup. CD62L.sup. IL-4.sup.+ IFN-.sup.+ T cells

(70) Six days after induction of T cell expansion with control beads (FIGS. 10A-10E) or hydrogel complexes (FIGS. 11A-11E), each expanded population was analyzed via flow cytometry for surface expression of CD3, CD4, CD8, CD45RO, CD45RA, CD62L, and CCR7 and compared with the expression profiles of cells at day 0 (FIGS. 9A-9E). In the CD3.sup.+ expanded population, a greater percentage of cells treated with hydrogel complexes were CD45RA.sup.+, relative to cells treated with control beads (FIG. 12B), suggesting that treatment with hydrogel complexes preferentially yields nave T cells. To confirm this observation, dual expression of CD62L and CD45RA was assessed. In the CD3.sup.+ expanded population, a greater percentage of cells treated with hydrogel complexes were CD62L.sup.+CD45RA.sup.+, relative to cells treated with control beads (FIG. 13). Dual expression of CCR7 and CD45RA was also measured in the CD3.sup.+ expanded population, and a greater percentage of cells treated with hydrogel complexes were CCR7.sup.+CD45RA.sup.+, relative to cells treated with control beads (FIG. 14). Thus, hydrogel complexes induced a higher percentage of T cells having a nave phenotype than control beads, as determined by surface marker expression profile.

(71) As noted in Table 1, the secretion of certain effector cytokines, such as IL-4 and IFN-, can be indicative of a T cell's effector memory phenotype. Six days after induction of T cell expansion with control beads or hydrogel complexes, samples of cell culture medium were harvested and analyzed for the presence of IL-4 and IFN- by ELISA. Medium from hydrogel complex-treated cultures contained lower quantities of IFN- (FIG. 15) and IL-4 (FIG. 16), relative to control bead-treated cultures, reflecting a smaller population of effector memory T cells within the hydrogel complex expanded population.

(72) Lastly, it was demonstrated that separation of expanded T cells from hydrogel complexes can be performed simply and efficiently. The top panels of FIG. 17 show hydrogel complex-expanded T cell cultures following hydrogel dissolution and removal. After 7 days of culture in the presence of hydrogel complexes, 5 mM EDTA was added to the culture medium, and cells were briefly agitated. Following agitation, cultures were placed on a cell separation magnetic stand (Quad Technologies) for 5 minutes. The supernatant was then removed and replated. As shown, hydrogel complex-expanded cells were free of magnetic particles. The bottom panels of FIG. 17 show control bead-expanded T cells following an analogous separation process.

(73) Other embodiments are in the claims.