- PI3 K:
Phosphatidylinositol-3 phosphate kinase
The mechanisms that regulate basal T cell size and metabolic activity are uncertain. Since the phosphatidylinositol-3 phosphate kinase (PI3 K) and Akt (PKB) pathway has been shown in model organisms to regulate both cell size and metabolism, we generated transgenic mice expressing a constitutively active form of Akt (myristoylated Akt, mAkt) in T cells. Naive transgenic T cells were enlarged and had increased rates of glycolysis compared to control T cells. In addition, mAkt transgenic T cells resisted death-by-neglect upon in vitro culture. Upon activation, mAkt-transgenic T cells were less dependent than control cells on costimulation through CD28 and could both grow rapidly and secrete cytokines in the absence of CD28 ligation. In addition, transgenic expression of mAkt led to the accumulation of CD4 T cells and B cells with age. Many aged mAkt-transgenic mice also developed autoimmunity with immunoglobulin deposits on kidney glomeruli and displayed increased incidence of lymphoma. Together, these data show that Akt activation is sufficient to increase basal T cell size and metabolism. Enhancement of T cell metabolism by Akt and more rapid CD28-independent T cell growth may contribute to the accumulation of excess immune cells and the development of lymphoma and autoimmunity.
To allow proper development and to maintain proper tissue homeostasis, it is important that cell growth and proliferation be balanced with cell death. If lymphocyte growth and proliferation are allowed to proceed in excess, oncogenesis 1 or autoimmunity can occur 2. In contrast, if too much apoptosis occurs, degenerative or immunodeficiency disease may arise 3–5. To regulate this balance, tissue-specific growth factors are required to provide cell survival and growth signals 6. In the absence of growth factors, cells from multicellular organisms undergo death-by-neglect, while excess growth factors can promote cell growth and subsequent proliferation. For T cells, growth factors that regulate cell survival and growth vary depending on developmental and activation state. Resting T cells require signals from the T cell receptor (TCR) or the interleukin-7 receptor (IL-7R) to prevent death-by-neglect both in vivo and in vitro7–10. During T cell activation, signals from the TCR and the costimulatory molecule CD28 are requiredto stimulate maximal cell survival and proliferation 11. If costimulation through CD28 is not available, TCR signals fail to lead to significant cell proliferation and can instead lead tocell death 12–16.
It is now appreciated that growth factors can directly affect cell growth independently of cell proliferation 17. One means by which growth factors may regulate cell growth is through control of cellular metabolism 18–21. Availability of growth factors is a critical element determining T cell rates of glycolysis and size 22. If T cells are removed from TCR and IL-7R ligands, expression of the glucose transporter Glut1 is decreased, cells atrophy, and glycolytic rates decrease prior to cell death-by-neglect 22, 23. In contrast to the decline in metabolism that occurs in neglect, the metabolic rate of T cells undergoing activation is significantly increased 24. Mitogenic TCR signals with CD28-mediated costimulation cause T cells to induce expression of Glut1, increase rates of glycolysis, and grow rapidly. These events are dependent on CD28 signal transduction because in the absence of CD28-mediated costimulation, Glut1 is poorly induced and T cells fail to increase glycolysis and proliferate 25.
The serine/threonine kinase Akt (PKB) is utilized in a variety of signal transduction pathways from T cell growth factors such as IL-7R 26, and CD28 costimulation 27, 28. Akt exists as an inactive cytosolic enzyme that becomes recruited to the cell surface through its pleckstrin homology domain following the generation of phosphatidylinositol-3-phosphates by phosphatidy-linositol-3 phosphate kinase (PI3 K). After recruitment to the plasma membrane, Akt becomes phosphorylated and activated by PDK1. When active, Akt phosphorylates a variety of targets including glycogen synthase kinase 3β (GSK3β), BAD, and Forkhead family member transcription factors 29. Phosphorylation of these targets by Akt inhibits their activities. In addition to inhibiting protein function, however, Akt also promotes increased cellular glycolysis and has been shown to promote increased cell size in both Drosophila and in transgenic mice that express constitutively active Akt in pancreatic β cells or cardiac myocytes 30–33.
It is unclear what role Akt may play in hematopoietic cells to promote cellular metabolism, growth, and activation. To address this issue, transgenic mice were generated that express a constitutively active form of Akt specifically in T cells. Naive T cells expressing active Akt were found to be significantly enlarged compared to non-transgenic control T cells. Resting mAkt-transgenic T cells had an increased glycolytic rate and resisted death-by-neglect. Akt-transgenic T cells were also found to be less dependent on CD28-mediated costimulation for cell growth, proliferation, and cytokine production. This increased potential for cellular activation may have contributed to the accumulation of memory phenotype CD44high T cells and significant increases of both CD4+ T and B cell numbers over time. In addition, Akt-transgenic mice were found to develop both autoimmunity and lymphoid tumors as they aged. Together, these data demonstrate that Akt can affect primary T cell size, survival, and activation requirements. These changes correlate with an increased propensity to develop autoimmunity and malignancy.
2.1 Generation and expression of mAkt transgene
To analyze the role of Akt in T cell metabolism and size, transgenic mice were generated to express a hemaglutinin (HA)-tagged and myristoylated Akt (mAkt) specifically in T cells. Myristoylation of Akt causes its association with the plasma membrane, resulting in its constitutive activation. In vitro, mAkt has been shown to prevent cell death-by-neglect and to increase glycolytic metabolism 25, 34, 35, possibly through its action on Glut1 expression and cell surface localization 22. HA-tagged mAkt was cloned into an expression vector driven by a 3.1-kb segment that includes both proximal and distal elements of the of lck promoter 36 and the CD2 enhancer. Five founder lines were generated. As previously reported, most founder lines led to animals with rapid progressive thymic lymphoma 37. However, in one line (line 42) animals consistently reached adulthood prior to the development of lymphoma. Surviving pups were backcrossed to C57BL/6 and their progeny used for the study of mAkt on T cell size, metabolism, and function. To analyze transgene expression, thymocytes and purified splenic T cells were lysed and probed by Western blot for the presence and activation of the transgene (Fig. 1). Blotting for the transgene-specific HA-tag showed that the mAkt transgene was expressed in both thymocytes and peripheral T cells. In addition, transgene-expressed mAkt was active, as evidenced by its phosphorylation on T308 and S473 (Fig. 1 and data not shown) and the increased phosphorylation of the Akt substrates GSK3α and β (Fig. 1).
2.2 mAkt T cells are enlarged and have increased glycolytic rates
Because Akt expression has been shown to increase cell size and metabolism in Drosophila and in murine pancreatic β cells and cardiac myocytes 30–33, T cell size and glycolysis were analyzed in mAkt-transgenic mice. To measure the size of CD4 and CD8 T cells, splenocytes from control and mAkt transgenic mice were isolated and analyzed flow cytometrically to determine the forward light scatter, an indicator of cell size (Fig. 2A). Because activated cells are larger than resting cells and mAkt transgene expression may affect cellular activation, CD44+ cells were excluded from analysis to allow comparison of only resting cells. For both CD4 and CD8 T cell lineages, mAkt transgene expression was found to result in increased forward light scatter of resting T cells, suggesting that activated Akt led to an increase in basal T cell size. To ensure that flow cytometric measurement of forward scatter reflected physical cell size differences, purified T cells were analyzed with a particle size analyzer (Coulter Z2). As described 23, non-transgenic T cells had a mean volume of 130±5 fL (Fig. 2B). Akt-transgenic T cells, however, were significantly larger, with a mean volume of 143±6 fL (p<0.0005). In addition to affecting cell size, Akt activation has been shown to affect glucose uptake and metabolism 34, 35, 38. Therefore, the glycolytic rate of purified control and mAkt-transgenic T cells was also determined (Fig. 3). Freshly isolated resting T cells expressing constitutively active Akt had a significantly higher rate of glycolysis than non-transgenic T cells, with glycolytic rates of 4.2±0.4 and 3.2±0.6 nmoles glucose consumed/106 cells/hour, respectively (p<0.05).
2.3 mAkt-transgenic T cells resist death-by-neglect
The ability of Akt to promote glucose metabolism has been associated with its ability to protect cells from death upon removal from growth factor 34, 35. Because mAkt-transgenic T cells had increased overall rates of glycolysis, T cell death-by-neglect was assayed by culturing purified spleen and lymph node T cells in the absence of stimulation and cell viability was determined over time. While both control and mAkt-transgenic T cells underwent cell death, both CD4 and CD8 mAkt-transgenic T cells were found to survive better when placed in suspension culture (Fig. 4 and data not shown). Unlike surviving control cells, which atrophied and became smaller, the mean size of surviving mAkt transgenic T cells was maintained in culture over time.
2.4 Akt promotes CD28-independent T cell growth, proliferation, and cytokine production
Optimal mitogenic stimulation of T cells requires signals through the antigen receptor as well as costimulation through CD28. One role CD28 costimulation plays in T cell activation is to promote rapid up-regulation of glycolysis 25. Because CD28 activates Akt, the dependence of mAkt-transgenic T cells on CD28 for growth and proliferation was examined (Fig. 5). Purified T cells from control and mAkt-transgenic mice were labeled with CFSE and cultured in the absence of stimulation, on plates coated with anti-CD3, or on plates coated with anti-CD3 and anti-CD28. After 3 days in vitro, T cells were analyzed flow cytometrically to determine the sizes of T cells that had not divided and T cells that had undergone a single division (Fig. 5A). While mAkt T cells were larger than control T cells, both populations of cells failed to proliferate in the absence of TCR engagement. Stimulation of control and mAkt-transgenic T cells with anti-CD3 was sufficient to promote blastogenesis (Fig. 5A). When T cells that had completed a cell division following stimulation with anti-CD3 alone were analyzed, mAkt-transgenic T cells had grown to a larger size than control cells. Stimulation of control T cells required CD3 and CD28 to achieve maximal cell growth within one cell division. In contrast, mAkt-transgenic T cells grew to maximal size with their first cell division when stimulated with CD3 only and the increase in size was not further enhanced by CD28 costimulation.
The ability of mAkt-transgenic T cells to achieve maximal cell growth in the absence of CD28 costimulation suggested that mAkt-transgenic T cells may be CD28-independent for cell proliferation and cytokine production. Purified CFSE-labeled T cells were cultured in the absence of stimulation, with anti-CD3 only, or with anti-CD3 and anti-CD28 together. T cells expressing the mAkt-transgene were found to be independent of CD28 costimulation for cell division (Fig. 5B), with 65.3% and 76.3% cells having undergone two or greater cell divisions in anti-CD3 only and anti-CD3 plus anti-CD28 stimulated populations, respectively. Non-transgenic T cells, in contrast, underwent maximal cell divisions only when stimulated with both CD3 and CD28, with 22.4% and 56.0% cells having undergone two or greater cell divisions in anti-CD3 only and anti-CD3 plus anti-CD28 stimulated populations, respectively. Production of the cytokines IL-2 (Fig. 5C) and IFN-γ (Fig. 5D) by mAkt T cells was partially independent of CD28, with tenfold higher levels of both IL-2 and IFN-γ produced by mAkt transgenic T cells compared to control cells when stimulated with only CD3. However, expression of the Akt transgene did not induce the production of the Th2 cytokine IL-4 in these cultures (data not shown).
2.5 mAkt-transgenic mice accumulate memory phenotype CD4 T cells and B cells
The ability of mAkt-transgenic T cells to be stimulated in vitro to grow and divide independent of CD28 suggested that mAkt T cells may become more easily activated invivo. The in vivo accumulation of T cells and expression of the T cell activation marker CD44 was examined in control and mAkt-transgenic mice. Young mAkt-transgenic mice were found to have normal numbers and frequency of T cells when CD4 and CD8 splenic T cells were quantitated flow cytometrically (Fig. 6A and data not shown). T cell phenotype in mAkt-transgenic animals became altered over time and CD44high CD4 and CD8 T cells accumulated with age (Fig. 6B and data not shown). While control 1-year-old mice were found to maintain numbers of B cells, CD4, and CD8 T cells identical to those of 8-week-old control mice, 1-year-old mAkt-transgenic mice were found to have a large accumulation of CD4 T cells as well as B cells (Fig. 6C). CD8 T cells in aged mAkt-transgenic mice were only modestly increased in number.
2.6 Mice expressing the mAkt transgene develop both tumors and autoimmunity
The enhanced T cell growth and accumulation of memory phenotype T cells in mAkt-transgenic mice could contribute to an enhanced incidence of lymphoma and the development of autoimmune disease. To determine incidence of lymphoma caused by the mAkt transgene, cohorts of transgenic and control mice were followed for a 1-year period. The majority of mAkt-transgenic mice were found to die as a result of thymic lymphoblastic lymphoma between 100 and 200 days of age (Fig. 7A). No control mice died over this period of time. Surviving mAkt-transgenic mice after this time period also displayed evidence of autoimmunity. Kidneys from 8-week- and 1-year-old control mice and surviving 1-year-old mAkt-transgenic mice were analyzed for immunoglobulin deposition by immunofluorescence after staining with anti-mouse immunoglobulin-FITC (Fig. 7B). Young control mice had no detectable immunoglobulin deposition. While control mice aged 1 year had small amounts of detectable immunoglobulin deposits, aged mAkt mice had substantial quantities of immunoglobulin in kidney sections. Glomeruli of mAkt-transgenic mice, in particular, were found to stain brightly with anti-mouse immunoglobulin. This deposition of immunoglobulin was not due to excessive hypergammaglobulinemia in mAkt-transgenic animals because serum immunoglobulin levels of all isotypes in mAkt-transgenic mice were not significantly different than aged control mice.
The serine/threonine kinase Akt has been shown to promote increased cell size and metabolism in Drosophila imaginal disc cells 30, pancreatic β cells 31, and cardiomyocytes 32, 33. Here, we show that constitutively active Akt can also directly affect T cell size, growth, and function. T cells expressing a form of Akt made constitutively active by myristoylation were enlarged and maintained an increased basal glycolytic rate yet remained in a non-proliferating state. Transgenic T cells were also capable of rapid growth, proliferation, and cytokine production in the absence of CD28 co-stimulation. Ultimately, transgenic mice were found to accumulate memory-phenotype CD4 T cells, B cells, and develop both tumors and autoimmunity. These data show that Akt can directly regulate T cell size, metabolism, and function.
The findings reported here corroborate and expand findings from previous studies analyzing the role of Akt in T cell activation and function. Similar to Malstrom et al. 37, the mAkt transgenic lines described here developed lymphoblastic lymphomas. Retroviral transduction of Akt in vitro has also been shown to increase T cell activation and production of IL-2 and IFN-γ but not IL-4 39. In addition, T cell-specific Akt-transgenic mice have been previously described to accumulate CD4 T cells and B cells with age and to develop autoimmunity 40. mAkt-transgenic T cells described here were less resistant to death than those described by Jones et al. 41 to a variety of stimulations, including dexamethasone (data not shown) and death-by-neglect. A possible explanation for this difference is the lack of detectable Bcl-xL induction in transgenic thymocytes described here (data not shown) compared those of Jones et al. 41. Overall, the findings described here share the late developmental effects of constitutive Akt activation in T cells previously described.
A novel feature of Akt-transgenic mice reported here is the increase in resting peripheral T cell size, metabolism, and enhanced growth of transgenic T cells upon antigenic stimulation which has not been examined in previous studies. Transgenic T cells described here were also found to have increased rates of glycolysis, consistent with in vitro observations that Akt can promote increased nutrient uptake and glycolysis 35, 42. The mechanism of this increased resting T cell size is uncertain, but the PI3 K/Akt pathway has been shown to play an important role in cell size regulation in other cell types 43–45. In addition, the Akt-induced increase in glucose metabolism can contribute to increased cell size directly by promoting the generation of metabolic substrates for biosynthesis, cell growth, and maintenance of increased biomass.
Akt was also shown here to promote CD28-independent cell growth upon antigenic stimulation. CD28 signals in part through activation of PI3 K and Akt to mediate T cell costimulation and thus increase cytokine production by stabilizing RNA 46, promote survival by increasing expression of the anti-apoptotic protein Bcl-xL 12–16, and to increase glycolytic metabolism 25, 34, 35. The increase in glycolytic metabolism caused by CD28 costimulation appears to be in excess of that required by the T cell for growth, because a significant quantity of glucose is disposed of as lactate rather than used for oxidative phosphorylation or biosynthesis. This seeming waste of energetic and biosynthetic resources occurs despite the high level of energetic demand placed upon T cells after stimulation 24. CD28 has been suggested to promote glycolysis directly to produce an anticipatory metabolic reserve in excess of the high demands of growing and proliferating T cells 25. The observation that mAkt-transgenic T cells had an increased rate of glycolysis and increased cell growth independent of CD28 supports the idea that Akt activation is sufficient to promote increased glycolytic metabolism and is a component of CD28-mediated cell growth regulation.
An outstanding question is what role increased cellular metabolism may play in T cell activation and development of tumors or autoimmunity. The accumulation of memory phenotype T cells in mAkt-transgenic mice described here could, in principle, be due to non-metabolic outcomes of Akt activation, such as NF-κB activation 47–50. Alternatively, thisaccumulation could be due to a decreased threshold of cells for inappropriate cell growth due to increased basal cellular metabolism. Akt-mediated cell survival has been shown to depend on its ability to promote glucose metabolism 34, 35. How much glucose metabolism can directly promote Akt-like phenotypes in T cells, however, is unknown. In future work, therefore, it will be important to analyze the role of directly increased glucose metabolism in T cells to determine if this is sufficient to promote T cell activation and the development of tumors and autoimmunity.
4 Materials and methods
4.1 Generation of transgenic mice
To generate mAkt-transgenic mice, murine Akt1 with a N-terminal fusion to the Src myristoylation sequence and a C-terminal hemaglutinin (HA)-tag 35, 51, wascloned into pLck.E2. This transgenic expression vector contains the Lck promoter 36, a human growth hormone mini-gene containing splice donor and acceptor sites and a poly-A sequence, and the CD2 enhancer (gift of J. Leiden, Abbot Laboratories, Abbot Park, IL). The expression construct was linearized and was microinjected by the University of Pennsylvania School of MedicineTransgenic and Chimeric Mouse Facility. Five founders were identified by Southern hybridization with a mAkt probe. Four founders died from thymic lymphoma prior to breeding. The remaining founder (line 42) was crossed to C57BL/6J (Jackson Laboratory; Bar Harbor, ME). With the exception of the experiments on aged mice, which were backcrossed to C57BL/6J three generations, mice used in these experiments were backcrossed at least five generations to C57BL/6J.
4.2 Western blots
Cell suspensions were prepared from non-transgenic and mAkt-transgenic thymic lobes and purified splenic T cells. Peripheral T cells were purified from non-transgenic and mAkt-transgenic splenocyte suspensions by negative selection as described 22 (StemCell Technologies; Vancouver, British Columbia). Cells were lysed in NET-N (20 mM Tris (pH 8.0), 100 mM NaCl, 1 mM EDTA, 0.5% Nonidet P-40) supplemented with protease (Roche; Basel, Switzerland) and phosphatase inhibitor cocktails (Sigma; St. Louis, MO). Two million cell equivalents were loaded per lane for SDS-PAGE. Blots were probed with anti-HA (Roche), anti-phospho-Akt T308 (ABR-Affinity Bioreagents; Golden, CO), anti-Akt (Cell Signaling Technology; Beverly, MA), anti-phospho-GSK3α/β T21/9 (Cell Signaling Technology), or anti-GSKβ (Cell Signaling Technology) followed by anti-rat-HRP (Roche) or anti-rabbit-HRP (Santa Cruz Biotechnology; Santa Cruz, CA).
4.3 Cell culture and flow cytometry and glycolysis
Purified T cells from non-transgenic or mAkt-transgenic mice were cultured in RPMI 1640 (Life Technologies; Grand Island, NY) supplemented with 10% fetal calf serum (Gemini Bio-Products; Woodland, CA). T cell sizes were determined by electrical displacement (Coulter Z2; Beckman-Coulter, Miami, FL) 15–30 min after purification to allow cells to achieve osmotic equilibrium. To test death-by-neglect, purified T cells were cultured in the absence of stimulation and cell viabilities were determined at various time points flow cytometrically by propidium iodide exclusion (Molecular Probes). In some experiments, T cells were labeled with CFSE (Molecular Probes; Eugene, OR) as described 52. CFSE-labeled T cells were then cultured in uncoated wells or wellsthat had been coated with 5 μg/ml anti-CD3 (clone 145–2C11; BD-PharMingen; San Diego, CA) alone or coated with 5 μg/ml anti-CD3 and 5 μg/ml anti-CD28 (clone 37.51; BD-PharMingen). After3 days of stimulation, cells were harvested and analyzed flow cytometrically for CFSE levels and forward light scatter. Similar results were obtained by analysis of cells at days 1 and 2 of stimulation (datanot shown). Cells were analyzed flow cytometrically with a FACSCalibur and CellQuest Software (BD-Biosciences; San Jose, CA). Antibodies used include: anti-CD3 (clone 145–2C11; BD-PharMingen), anti-CD4 (clone H129–19; BD-PharMingen); anti-CD8 (clone 53–6.7; BD-PharMingen), anti-CD44 (clone IM7; BD-PharMingen); anti-B220 (clone RA3–6B2; BD-PharMingen). Cell numbers were determined by multiplying the percentage that each cell type represented by the total cell number as determined by hemacytometer. Glycolytic rates of purified T cells were measured as described 23 1 h after initial purification and culture.
To determine T cell production of IL-2, IFN-γ, and IL-4, supernatants of T cell cultures were harvested at day 2 and recombinant mouse IL-2, IFN-γ, and IL-4 standards (all BD-PharMingen) were incubated on plates that had been coated with anti-IL-2 (1 μg/ml), anti-IFN-γ (2 μg/ml), or anti-IL-4 (4 μg/ml) (all anti-sera were purified rat anti-mouse; BD-PharMingen). After thorough washing, plates were incubated with biotinylated rat anti-mouse IL-2 (1 μg/ml), IFN-γ (1 μg/ml) or IL-4 (2 μg/ml) anti-sera (all BD-PharMingen), followed by avidin-alkaline phosphatase (Jackson Immunoresearch Laboratories; West Grove, PA), and phosphatase substrate (Sigma). Determination of immunoglobulin isotype expression was performed using a mouse immunoglobulin isotyping kit according to manufacturer's instructions (BD-PharMingen).
Kidneys were flash-frozen with liquid nitrogen in OCT compound (Tissue-Tek; Sakura Finetek, Torrance, CA) and 10-μm sections were cut. Sections were blocked with PBS containing 2% FCS and stained with FITC conjugated goat anti-mouse immunoglobulin (BD-PharMingen) in PBS 2% FCS. Slides were washed, mounted, and analyzed microscopically at 200× total magnification using a Nikon E800 microscope (Optical Apparatus; Ardmore, PA), Micromax digital camera (Princeton Instruments; Trenton, NJ), and Metamorph 4.5 imaging software (Universal Imaging; Downington, PA).
We would like to acknowledge the Thompson laboratory with particular thanks to Drs. Kenneth Frauwirth and David Plas. We also thank the University of Pennsylvania School of Medicine Transgenic and Chimeric Mouse Facility for generation of the mAkt-transgenic mice utilized in these studies. This work was supported by grants from the National Cancer Institute. J. C. R. was also supported through a Howard Temin K01 Career Development Award from the National Cancer Institute.