Kinetics of ocular and systemic antigen-specific T-cell responses elicited during murine cytomegalovirus retinitis

Cytomegalovirus (CMV) reactivation in the retina of immunocompromized patients is a cause of significant morbidity as it can lead to blindness. The adaptive immune response is critical in controlling murine CMV (MCMV) infection in MCMV-susceptible mouse strains. CD8(+) T cells limit systemic viral replication in the acute phase of infection and are essential to contain latent virus. In this study, we provide the first evaluation of the kinetics of anti-viral T-cell responses after subretinal infection with MCMV. The acute response was characterized by a rapid expansion phase, with infiltration of CD8(+) T cells into the infected retina, followed by a contraction phase. MCMV-specific T cells displayed biphasic kinetics with a first peak at day 12 and contraction by day 18 followed by sustained recruitment of these cells into the retina at later time points post-infection. MCMV-specific CD8(+) T cells were also observed in the draining cervical lymph nodes and the spleen. Presentation of viral epitopes and activation of CD8(+) T cells was widespread and could be detected in the spleen and the draining lymph nodes, but not in the retina or iris. Moreover, after intraocular infection, antigen-specific cytotoxic activity was detectable and exhibited kinetics equivalent to those observed after intraperitoneal infection with the same viral dose. These data provide novel insights of how and where immune responses are initiated when viral antigen is present in the subretinal space.

Avidity determines T-cell reactivity in abacavir hypersensitivity

The antiretroviral drug abacavir (abc) elicits severe drug hypersensitivity reactions in HLA-B*5701(+) individuals. To understand the abc-specific activation of CD8(+) T cells, we generated abc-specific T-cell clones (abc-TCCs). Abc reactivity could not be linked to the metabolism and/or processing of the drug, since abc metabolizing enzymes were not expressed in immune cells and inhibition of the proteasome in APCs did not affect TCC reactivity. Ca(2+) influx assays revealed different reactivity patterns of abc-TCCs. While all TCCs reacted to abc presented on HLA-B*5701 molecules, a minority also reacted immediately to abc in solution. Titration experiments showed that the ability to react immediately to abc correlated significantly with the TCR avidity of the T cells. Modifications of soluble abc concentrations revealed that the reactivity patterns of abc-TCCs were not fixed but dynamic. When TCCs with an intermediate TCR avidity were stimulated with increasing abc concentrations, they showed an accelerated activation kinetic. Thus, they reacted immediately to the drug, similar to the reaction of TCCs of high avidity. The observed immediate activation and the noninvolvement of the proteasome suggest that, in contrast to haptens, abc-specific T-cell stimulation does not require the formation of covalent bonds to produce a neo-antigenic determinant.

Interferon-gamma regulates idiopathic pneumonia syndrome, a Th17+CD4+ T-cell-mediated graft-versus-host disease

RATIONALE: Pulmonary complications of hematopoietic stem cell transplantation include infections and graft-versus-host diseases, such as idiopathic pneumonia syndrome (IPS). Conflicting data exist regarding the role of the interferon (IFN)-gamma-producing Th1 CD4(+) T-cell subset and IL-17A in IPS. OBJECTIVES: To determine the role of IFN-gamma and IL-17A in the establishment of pulmonary graft-versus-host disease. METHODS: A semiallogeneic murine model based on C57BL/6 x BALB/c as recipients with transplantation of BALB/c RAG2(-/-) bone marrow and transfer of different genetic knockout T cells (T-bet(-/-), IFN-gamma(-/-), IFN-gammaR(-/-)) on a BALB/c background. Lung tissue was examined for parenchymal changes and infiltrating cells by histology and fluorescence-activated cell sorter analysis. MEASUREMENTS AND MAIN RESULTS: After transfer of semiallogeneic bone marrow together with donor CD4(+) T cells lacking IFN-gamma or T-bet-a T-box transcription factor controlling Th1 commitment-we found severe inflammation in the lungs, but no enhancement in other organs. In contrast, wild-type donor CD4(+) T cells mediated minimal inflammation only, and donor CD8(+) T cells were not required for IPS development. Mechanistically, the absence of IFN-gamma or IFN-gamma signaling in pulmonary parenchymal cells promoted expansion of IL-17A-producing CD4(+) T cells and local IL-17A release. In vivo depletion of IL-17A reduced disease severity. CONCLUSIONS: One mechanism of IFN-gamma protection against IPS is negative regulation of the expansion of pathogenic IL-17A-producing CD4(+) T cells through interaction with the IFN-gamma receptor on the pulmonary parenchymal cell population.

Distinct requirements for activation-induced cell surface expression of preformed Fas/CD95 ligand and cytolytic granule markers in T cells

Fas (CD95/Apo-1) ligand is a potent inducer of apoptosis and one of the major killing effector mechanisms of cytotoxic T cells. Thus, Fas ligand activity has to be tightly regulated, involving various transcriptional and post-transcriptional processes. For example, preformed Fas ligand is stored in secretory lysosomes of activated T cells, and rapidly released by degranulation upon reactivation. In this study, we analyzed the minimal requirements for activation-induced degranulation of Fas ligand. T cell receptor activation can be mimicked by calcium ionophore and phorbol ester. Unexpectedly, we found that stimulation with phorbol ester alone is sufficient to trigger Fas ligand release, whereas calcium ionophore is neither sufficient nor necessary. The relevance of this process was confirmed in primary CD4(+) and CD8(+) T cells and NK cells. Although the activation of protein kinase(s) was absolutely required for Fas ligand degranulation, protein kinase C or A were not involved. Previous reports have shown that preformed Fas ligand co-localizes with other markers of cytolytic granules. We found, however, that the activation-induced degranulation of Fas ligand has distinct requirements and involves different mechanisms than those of the granule markers CD63 and CD107a/Lamp-1. We conclude that activation-induced degranulation of Fas ligand in cytotoxic lymphocytes is differently regulated than other classical cytotoxic granule proteins.

T cell-mediated immunoregulation in the gastrointestinal tract

In the intestinal tract, only a single layer of epithelial cells separates innate and adaptive immune effector cells from a vast amount of antigens. Here, the immune system faces a considerable challenge in tolerating commensal flora and dietary antigens while preventing the dissemination of potential pathogens. Failure to tightly control immune reactions may result in detrimental inflammation. In this respect, 'conventional' regulatory CD4(+) T cells, including naturally occurring and adaptive CD4(+) CD25(+) Foxp3(+) T cells, Th3 and Tr1 cells, have recently been the focus of considerable attention. However, regulatory mechanisms in the intestinal mucosa are highly complex, including adaptations of nonhaematopoietic cells and innate immune cells as well as the presence of unconventional T cells with regulatory properties such as resident TCRgammadelta or TCRalphabeta CD8(+) intraepithelial lymphocytes. This review aims to summarize the currently available knowledge on conventional and unconventional regulatory T cell subsets (Tregs), with special emphasis on clinical data and the potential role or malfunctioning of Tregs in four major human gastrointestinal diseases, i.e. inflammatory bowel diseases, coeliac disease, food allergy and colorectal cancer. We conclude that the clinical data confirms some but not all of the findings derived from experimental animal models.

Programmed death 1 signaling on chronic myeloid leukemia-specific T cells results in T-cell exhaustion and disease progression

Chronic myeloid leukemia (CML) is a malignant myeloproliferative disease with a characteristic chronic phase (cp) of several years before progression to blast crisis (bc). The immune system may contribute to disease control in CML. We analyzed leukemia-specific immune responses in cpCML and bcCML in a retroviral-induced murine CML model. In the presence of cpCML and bcCML expressing the glycoprotein of lymphocytic choriomeningitis virus as a model leukemia antigen, leukemia-specific cytotoxic T lymphocytes (CTLs) became exhausted. They maintained only limited cytotoxic activity, and did not produce interferon-gamma or tumor necrosis factor-alpha or expand after restimulation. CML-specific CTLs were characterized by high expression of programmed death 1 (PD-1), whereas CML cells expressed PD-ligand 1 (PD-L1). Blocking the PD-1/PD-L1 interaction by generating bcCML in PD-1-deficient mice or by repetitive administration of alphaPD-L1 antibody prolonged survival. In addition, we found that PD-1 is up-regulated on CD8(+) T cells from CML patients. Taken together, our results suggest that blocking the PD-1/PD-L1 interaction may restore the function of CML-specific CTLs and may represent a novel therapeutic approach for CML.

Next-generation tissue microarray (ngTMA) increases the quality of biomarker studies: an example using CD3, CD8, and CD45RO in the tumor microenvironment of six different solid tumor types

Background Tissue microarray (TMA) technology revolutionized the investigation of potential biomarkers from paraffin-embedded tissues. However, conventional TMA construction is laborious, time-consuming and imprecise. Next-generation tissue microarrays (ngTMA) combine histological expertise with digital pathology and automated tissue microarraying. The aim of this study was to test the feasibility of ngTMA for the investigation of biomarkers within the tumor microenvironment (tumor center and invasion front) of six tumor types, using CD3, CD8 and CD45RO as an example. Methods Ten cases each of malignant melanoma, lung, breast, gastric, prostate and colorectal cancers were reviewed. The most representative H&E slide was scanned and uploaded onto a digital slide management platform. Slides were viewed and seven TMA annotations of 1 mm in diameter were placed directly onto the digital slide. Different colors were used to identify the exact regions in normal tissue (n = 1), tumor center (n = 2), tumor front (n = 2), and tumor microenvironment at invasion front (n = 2) for subsequent punching. Donor blocks were loaded into an automated tissue microarrayer. Images of the donor block were superimposed with annotated digital slides. Exact annotated regions were punched out of each donor block and transferred into a TMA block. 420 tissue cores created two ngTMA blocks. H&E staining and immunohistochemistry for CD3, CD8 and CD45RO were performed. Results All 60 slides were scanned automatically (total time < 10 hours), uploaded and viewed. Annotation time was 1 hour. The 60 donor blocks were loaded into the tissue microarrayer, simultaneously. Alignment of donor block images and digital slides was possible in less than 2 minutes/case. Automated punching of tissue cores and transfer took 12 seconds/core. Total ngTMA construction time was 1.4 hours. Stains for H&E and CD3, CD8 and CD45RO highlighted the precision with which ngTMA could capture regions of tumor-stroma interaction of each cancer and the T-lymphocytic immune reaction within the tumor microenvironment. Conclusion Based on a manual selection criteria, ngTMA is able to precisely capture histological zones or cell types of interest in a precise and accurate way, aiding the pathological study of the tumor microenvironment. This approach would be advantageous for visualizing proteins, DNA, mRNA and microRNAs in specific cell types using in situ hybridization techniques.

Clinical impact of programmed cell death ligand 1 expression in colorectal cancer

BACKGROUND Programmed cell death 1 (PD-1) receptor triggering by PD ligand 1 (PD-L1) inhibits T cell activation. PD-L1 expression was detected in different malignancies and associated with poor prognosis. Therapeutic antibodies inhibiting PD-1/PD-L1 interaction have been developed. MATERIALS AND METHODS A tissue microarray (n=1491) including healthy colon mucosa and clinically annotated colorectal cancer (CRC) specimens was stained with two PD-L1 specific antibody preparations. Surgically excised CRC specimens were enzymatically digested and analysed for cluster of differentiation 8 (CD8) and PD-1 expression. RESULTS Strong PD-L1 expression was observed in 37% of mismatch repair (MMR)-proficient and in 29% of MMR-deficient CRC. In MMR-proficient CRC strong PD-L1 expression correlated with infiltration by CD8(+) lymphocytes (P=0.0001) which did not express PD-1. In univariate analysis, strong PD-L1 expression in MMR-proficient CRC was significantly associated with early T stage, absence of lymph node metastases, lower tumour grade, absence of vascular invasion and significantly improved survival in training (P=0.0001) and validation (P=0.03) sets. A similar trend (P=0.052) was also detectable in multivariate analysis including age, sex, T stage, N stage, tumour grade, vascular invasion, invasive margin and MMR status. Interestingly, programmed death receptor ligand 1 (PDL-1) and interferon (IFN)-�� gene expression, as detected by quantitative reverse transcriptase polymerase chain reaction (RT-PCR) in fresh frozen CRC specimens (n=42) were found to be significantly associated (r=0.33, P=0.03). CONCLUSION PD-L1 expression is paradoxically associated with improved survival in MMR-proficient CRC.

Characterization and optimization of antigen-specific T cell responses during ex vivo expansion of melanoma tumor-infiltrating lymphocytes

Treatment of metastatic melanoma with tumor reactive T cells (adoptive T cell therapy, ACT) is a promising approach associated with a high clinical response rate. However, further optimization of this treatment modality is required to increase the clinical response after this therapy. ACT in melanoma involves an initial phase (pre-REP) of tumor-infiltrating lymphocyte (TIL) expansion ex vivo from tumor isolates followed by a second phase, ���rapid expansion protocol��� (REP) generating the billions of cells used as the TIL infusion product. The main question addressed in this thesis was how the currently used REP affected the responsiveness of the CD8+ T cells to defined melanoma antigens. We hypothesized that the REP drives the TIL to further differentiate and become hyporesponsive to antigen restimulation, therefore, proper cytokine treatment or other ways to expand TIL is required to improve upon this outcome. We evaluated the response of CD8+ TIL to melanoma antigen restimulation using MART-1 peptide-pulsed mature DC in vitro. Post-REP TILs were mostly hypo-responsive with poor proliferation and higher apoptosis. Phenotypic analysis revealed that the expression of CD28 was significantly reduced in post-REP TILs. By sorting experiment and microarray analysis, we confirmed that the few CD28+ post-REP TILs had superior survival capacity and proliferated after restimulation. We then went on to investigate methods to maintain CD28 expression during the REP and improve TIL responsiveness. Firstly, IL-15 and IL-21 were found to synergize in maintaining TIL CD28 expression and antigenic responsiveness during REP. Secondly, we found IL-15 was superior as compared to IL-2 in supporting the long-term expansion of antigen-specific CD8+ TIL after restimulation. These results suggest that current expansion protocols used for adoptive T-cell therapy in melanoma yield largely hyporesponsive products containing CD8+ T cells unable to respond in vivo to re-stimulation with antigen. A modification of our current approaches by using IL-15+IL-21 as supporting cytokines in the REP, or/and administration of IL-15 instead of IL-2 after TIL infusion, may enhance the anti-tumor efficacy and long-term persistence of infused T cells in vivo.

CD8+ T cells and NK cells from blister fluid of an EEM/SJS overlap highly express Fas ligand and Granulysin

INTRODUCTION Erythema exsudativum multiforme majus (EEMM) and Stevens-Johnson Syndrome (SJS) are severe cutaneous reaction patterns caused by infections or drug hypersensitivity. The mechanism by which widespread keratinocyte death is mediated by the immune system in EEMM/SJS are still to be elucidated. Here, we characterized the blister cells isolated from a patient with EEMM/SJS overlap and investigated its cause. METHODS Clinical classification of the cutaneous eruption was done according to the consensus definition of severe blistering skin reactions and histological analysis. Common infectious causes of EEMM were investigated using standard clinical techniques. T cell reactivity for potentially causative drugs was assessed by lymphocyte transformation tests (LTT). Lymphocytes isolated from blister fluid were analyzed for their expression of activation markers and cytotoxic molecules using flow cytometry. RESULTS The healthy 58 year-old woman suffered from mild respiratory tract infection and therefore started treatment with the secretolytic drug Ambroxol. One week later, she presented with large palmar and plantar blisters, painful mucosal erosions, and flat atypical target lesions and maculae on the trunc, thus showing the clinical picture of an EEMM/SJS overlap (Fig. 1). This diagnosis was supported by histology, where also eosinophils were found to infiltrate the upper dermis, thus pointing towards a cutaneous adverse drug reaction (cADR). Analysis of blister cells showed that they mainly consisted of CD8+ and CD4+ T cells and a smaller population of NK cells. Both the CD8+ T cells and the NK cells were highly activated and expressed Fas ligand and the cytotoxic molecule granulysin (Fig. 2). In addition, in comparison to NK cells from PBMC, NK cells in blister fluids strongly upregulated the expression of the skin-homing chemokine receptor CCR4 (Fig 4). Surprisingly, the LTT performed on PBMCs in the acute phase was positive for Ambroxol (SI=2.9) whereas a LTT from a healthy but exposed individual did not show unspecific proliferation. Laboratory tests for common infectious causes of EEMM were negative (HSV-1/-2, M. pneumoniae, Parvovirus B19). However, 6 weeks later, specific proliferation to Ambroxol could no longer be observed in the LTT (Fig 4.).

Vascular cell adhesion molecule 1 expression by biliary epithelium promotes persistence of inflammation by inhibiting effector T-cell apoptosis

Chronic hepatitis occurs when effector lymphocytes are recruited to the liver from blood and retained in tissue to interact with target cells, such as hepatocytes or bile ducts (BDs). Vascular cell adhesion molecule 1 (VCAM-1; CD106), a member of the immunoglobulin superfamily, supports leukocyte adhesion by binding a4b1 integrins and is critical for the recruitment of monocytes and lymphocytes during inflammation. We detected VCAM-1 on cholangiocytes in chronic liver disease (CLD) and hypothesized that biliary expression of VCAM-1 contributes to the persistence of liver inflammation. Hence, in this study, we examined whether cholangiocyte expression of VCAM-1 promotes the survival of intrahepatic a4b1 expressing effector T cells. We examined interactions between primary human cholangiocytes and isolated intrahepatic T cells ex vivo and in vivo using the Ova-bil antigen-driven murine model of biliary inflammation. VCAM-1 was detected on BDs in CLDs (primary biliary cirrhosis, primary sclerosing cholangitis, alcoholic liver disease, and chronic hepatitis C), and human cholangiocytes expressed VCAM-1 in response to tumor necrosis factor alpha alone or in combination with CD40L or interleukin-17. Liver-derived T cells adhered to cholangiocytes in vitro by a4b1, which resulted in signaling through nuclear factor kappa B p65, protein kinase B1, and p38 mitogen-activated protein kinase phosphorylation. This led to increased mitochondrial B-cell lymphoma 2 accumulation and decreased activation of caspase 3, causing increased cell survival. We confirmed our findings in a murine model of hepatobiliary inflammation where inhibition of VCAM-1 decreased liver inflammation by reducing lymphocyte recruitment and increasing CD8 and T helper 17 CD4 Tcell survival. Conclusions: VCAM-1 expression by cholangiocytes contributes to persistent inflammation by conferring a survival signal to a4b1 expressing proinflammatory T lymphocytes in CLD.

Suppression of peripheral blood mononuclear cell proliferation by a periodontopathic bacteria {\it Actinobacillus actinomycetemcomitans\/}

Actinobacillus actinomycetemcomitans (Aa) is a gram-negative coccobacillus implicated as a major pathogen in juvenile periodontitis. The immunosuppressive activity of a sonic extract (designated 100SN) derived from Aa was investigated. 100SN suppressed spontaneous proliferation as well as proliferative response to the mitogens, PHA and PWM, of human peripheral blood mononuclear cells (PBMC). 100SN-induced suppression of PHA-stimulated proliferation was heat-sensitive, inactivated by pronase and trypsin, dose-dependent and non-cytotoxic. There were no significant changes in the CD4$\sp+$ or CD8$\sp+$ subsets of PBMC after 7-day incubation with 100SN. There was a trend toward increased levels of the CD4$\sp+$CD45R$\sp{\rm hi}$CDw29$\sp{\rm lo}$ (naive cells, associated with suppressor-inducer activity) and CD4$\sp+$CDw29$\sp{\rm hi}$CD45R$\sp{\rm lo}$ (memory cells, associated with helper-inducer activity) subsets. The target of 100SN appeared to be the non-adherent cells and suppression by 100SN could not be reversed by indomethacin (IDM), the cyclo-oxygenase inhibitor of prostaglandin (PG) synthesis. The mechanism of 100SN-induced suppression was studied in terms of inhibition involving IL-2-regulated T cell proliferation and the results point to the possibility that suppression occurred subsequent to IL-2 receptor binding.^ The suppressive activity observed could occur through multiple mechanisms including cell-cell; contact or release of soluble factors. Supernatants derived from 7-day cultures of PBMC and 100SN (designated CSN-A) were able to suppress proliferative response of PBMC to PHA without affecting cell viability. Analysis of CSN-A showed that it contained PGE2 and soluble IL-2 receptors. Suppression by CSN-A could be partially overcome by either IDM or exogenous IL-2. Significant suppression was also maintained when both IDM and exogenous IL-2 were added at the same time. These findings suggest that PGE2 and soluble IL-2 receptors contribute to the suppression observed but other suppressive cytokine(s) may be involved. Collectively, the data indicate that a factor derived from oral bacteria associated with juvenile periodontitis have profound effects on cellular immune responses, and that these effects may be partially mediated by secondary factors produced by the host in response to the bacteria. ^

Cytotoxic CD8+ T cells stimulate hematopoietic progenitors by promoting cytokine release from bone marrow mesenchymal stromal cells.

Cytotoxic CD8(+) T cells (CTLs) play a major role in host defense against intracellular pathogens, but a complete clearance of pathogens and return to homeostasis requires the regulated interplay of the innate and acquired immune systems. Here, we show that interferon �� (IFN��) secreted by effector CTLs stimulates hematopoiesis at the level of early multipotent hematopoietic progenitor cells and induces myeloid differentiation. IFN�� did not primarily affect hematopoietic stem or progenitor cells directly. Instead, it promoted the release of hematopoietic cytokines, including interleukin 6 from bone marrow mesenchymal stromal cells (MSCs) in the hematopoietic stem cell niche, which in turn reduced the expression of the transcription factors Runx-1 and Cebp�� in early hematopoietic progenitor cells and increased myeloid differentiation. Therefore, our study indicates that, during an acute viral infection, CTLs indirectly modulate early multipotent hematopoietic progenitors via MSCs in order to trigger the temporary activation of emergency myelopoiesis and promote clearance of the infection.

Abacavir induced T cell reactivity from drug na��ve individuals shares features of allo-immune responses.

Abacavir hypersensitivity is a severe hypersensitivity reaction which occurs exclusively in carriers of the HLA-B*57���01 allele. In vitro culture of PBMC with abacavir results in the outgrowth of abacavir-reacting CD8+ T cells, which release IFN�� and are cytotoxic. How this immune response is induced and what is recognized by these T cells is still a matter of debate. We analyzed the conditions required to develop an abacavir-dependent T cell response in vitro. The abacavir reactivity was independent of co-stimulatory signals, as neither DC maturation nor release of inflammatory cytokines were observed upon abacavir exposure. Abacavir induced T cells arose in the absence of professional APC and stemmed from na��ve and memory compartments. These features are reminiscent of allo-reactivity. Screening for allo-reactivity revealed that about 5% of generated T cell clones (n���=���136) from three donors were allo-reactive exclusively to the related HLA-B*58���01. The addition of peptides which can bind to the HLA-B*57���01-abacavir complex and to HLA-B*58���01 during the induction phase increased the proportion of HLA-B*58���01 allo-reactive T cell clones from 5% to 42%. In conclusion, abacavir can alter the HLA-B*57���01-peptide complex in a way that mimics an allo-allele ('altered self-allele') and create the potential for robust T cell responses.

VEGFR2 Signaling Prevents Colorectal Cancer Cell Senescence to Promote Tumorigenesis in Mice With Colitis.

BACKGROUND & AIMS Senescence prevents cellular transformation. We investigated whether vascular endothelial growth factor (VEGF) signaling via its receptor, VEGFR2, regulates senescence and proliferation of tumor cells in mice with colitis-associated cancer (CAC). METHODS CAC was induced in VEGFR2(��IEC) mice, which do not express VEGFR2 in the intestinal epithelium, and VEGFR2(fl/fl) mice (controls) by administration of azoxymethane followed by dextran sodium sulfate. Tumor development and inflammation were determined by endoscopy. Colorectal tissues were collected for immunoblot, immunohistochemical, and quantitative polymerase chain reaction analyses. Findings from mouse tissues were confirmed in human HCT116 colorectal cancer cells. We analyzed colorectal tumor samples from patients before and after treatment with bevacizumab. RESULTS After colitis induction, VEGFR2(��IEC) mice developed significantly fewer tumors than control mice. A greater number of intestinal tumor cells from VEGFR2(��IEC) mice were in senescence than tumor cells from control mice.��We found VEGFR2 to activate phosphatidylinositol-4,5-bisphosphate-3-kinase and AKT, resulting in inactivation of p21 in HCT116 cells. Inhibitors of VEGFR2 and AKT induced senescence in HCT116 cells. Tumor cell senescence promoted an anti-tumor immune response by CD8(+) T cells in mice. Patients whose tumor samples showed an increase in the proportion of senescent cells after treatment with bevacizumab had longer progression-free survival than patients in which the proportion of senescent tumor cells did not change before and after treatment. CONCLUSIONS Inhibition of VEGFR2 signaling leads to senescence of human and mouse colorectal cancer cells. VEGFR2 interacts with phosphatidylinositol-4,5-bisphosphate-3-kinase and AKT to inactivate p21. Colorectal tumor senescence and p21 level correlate with patient survival during treatment with bevacizumab.