Posttransplant lymphoproliferative disorder (PTLD)-associated Epstein–Barr virus (EBV)+ B cell lymphomas are serious complications of solid organ and bone marrow transplantation. The EBV protein LMP2a, a B cell receptor (BCR) mimic, provides survival signals to virally infected cells through Syk tyrosine kinase. Therefore, we explored whether Syk inhibition is a viable therapeutic strategy for EBV-associated PTLD. We have shown that R406, the active metabolite of the Syk inhibitor fostamatinib, induces apoptosis and cell cycle arrest while decreasing downstream phosphatidylinositol-3′-kinase (PI3K)/Akt signaling in EBV+ B cell lymphoma PTLD lines in vitro. However, Syk inhibition did not inhibit or delay the in vivo growth of solid tumors established from EBV-infected B cell lines. Instead, we observed tumor growth in adjacent inguinal lymph nodes exclusively in fostamatinib-treated animals. In contrast, direct inhibition of PI3K/Akt significantly reduced tumor burden in a xenogeneic mouse model of PTLD without evidence of tumor growth in adjacent inguinal lymph nodes. Taken together, our data indicate that Syk activates PI3K/Akt signaling which is required for survival of EBV+ B cell lymphomas. PI3K/Akt signaling may be a promising therapeutic target for PTLD, and other EBV-associated malignancies.
EBV is a ubiquitous, B cell lymphotrophic human γ-herpesvirus . In immunocompetent individuals infection is controlled by a vigorous cytotoxic T lymphocyte (CTL) response. However, EBV can transform normal B cells and promote lymphomagenesis when the CTL response is debilitated as in PTLD. PTLD-associated B cell lymphomas can express a series of EBV latent genes, including six EBV nuclear antigens (EBNAs), both small RNAs (EBERs) and three latent membrane proteins, LMP1, LMP2a, LMP2b. In addition to participating in malignant transformation, these viral genes promote evasion of immune recognition, maintain viral latency and hijack production of the autocrine growth factor, cellular IL-10 [2, 3]. Since over 90% of PTLD-associated lymphomas are EBV-positive , the virus itself and the mechanisms by which EBV drives lymphomagenesis are attractive targets for development of new therapeutics for PTLD.
LMP2a is a 54 kDa protein expressed in the membrane of EBV-infected B cells and a functional, constitutively active mimic of the BCR [1, 5]. LMP2a provides constitutive, BCR-like signals primarily through recruitment and activation of Syk tyrosine kinase. Syk subsequently activates downstream signaling pathways including the PI3K/Akt and Erk mitogen-activated protein kinase (MAPK) pathways. Whether Syk or PI3K/Akt activation is required for EBV+ B cell lymphomagenesis, tumor survival and proliferation is unknown. Here, we used small molecule inhibitors of Syk and PI3K/Akt to probe whether Syk or PI3K/Akt activation is required for the survival and proliferation of EBV+ B cell lymphomas. We also assessed the efficacy of Syk and PI3K/Akt inhibition on tumor burden in a xenotransplantation model of EBV-associated PTLD.
Materials and Methods
Fostamatinib is a soluble, orally dosed pro-drug that is rapidly and extensively converted to the active metabolite R406. Both fostamatinib and R406 are Syk inhibitors. Fostamatinib-containing chow and the active component R406 were kindly provided by Rigel Pharmaceuticals and AstraZeneca. For cell culture studies, R406 (10 mM in dimethylsulfoxide [DMSO]) was aliquotted and stored at −80°C. The small molecule inhibitors of Erk (PD98059, 20 mM) and PI3K/Akt (LY294002, 20 mM) (Calbiochem; EMD Millipore, Billerica, MA, USA) were diluted in DMSO (Sigma-Aldrich, St. Louis, MO, USA) to the indicated concentrations. The following antibodies were used: anti-Erk and anti-phospho-Erk (Tyr204) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-Akt, phospho-Akt (Ser473), B cell linker protein (BLNK) and phospho-BLNK (Tyr96) (Cell Signaling Technology, Danvers, MA, USA), anti-β-actin (Sigma-Aldrich), horseradish peroxidase (HRP)-conjugated goat anti-rabbit and HRP-conjugated donkey antimouse antibodies (Jackson Immunoresearch Labs, Reston, VA, USA).
The EBV–Burkitt's lymphoma line Ramos was obtained from ATCC. Cell lines AB5, JB7 and ZD3 are spontaneously derived EBV+ B lymphoblastoid cell lines (SLCL) from peripheral blood (JB7, ZD3) or lymph nodes (AB5) of patients diagnosed with EBV+ PTLD [2, 3, 6, 7]. The EBV–Burkitt's lymphoma line BL41 and the EBV-infected line BL41.B95 were provided by Dr. Elliot Kieff (Harvard Medical School). Cell lines were maintained in a 5% CO2 humidified 37°C incubator in RPMI-1640 supplemented with 10% heat-inactivated fetal calf serum (Serum Source International, Charlotte, NC, USA), and 50 units/mL penicillin-streptomycin (Invitrogen, Carlsbad, CA, USA).
Apoptosis and cell cycle analysis
Cells (0.5–1×106 cells/mL) were plated in serial dilutions of small molecule inhibitors R406, LY294002, or PD98059 (0–10 μM), or equivalent amounts of vehicle (DMSO, 0–1:1000). Drug and media were replenished after 48 h of a total of 96 h in culture at 37°C. Cell cycle analysis using propidium iodide was as previously described [4, 7]. The percentage of apoptotic cells was determined using Annexin V-enhanced GFP (EGFP) apoptosis detection kits (BioVision) per the manufacturer's instructions. Data were analyzed on a FACScan flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA).
Cells (2×106 cells/mL) were cultured with small molecule inhibitors or DMSO vehicle. Tumor tissue was recovered during necropsy. After manual dissociation of tumor, single cell suspensions were collected after passing through a 70 μM cell strainer (BD Biosciences). Cells were recovered and washed with PBS containing 1 mM sodium orthovanadate. Lysate preparation and western blots were performed as previously described .
Phosflow for phospho-BLNK
BCR crosslinking was initiated by addition of F(ab)2 fragments of anti-IgM (10 μg/mL), in the presence of R406 or vehicle. Fixation and permeabilization of cells was performed as described . Cells were stained with antiphospho-BLNK-phycoerythrin (PE) (BD Pharmingen, San Jose, CA, USA) before collection on a FACScan (BD Biosciences).
In vivo xenogeneic model of PTLD
Six-week-old, male nonobese diabetic (NOD) severe combined immunodeficiency disease (SCID), NOD.CB17-Prkdcscid/J mice were injected subcutaneously (s.c.) over the right flank with 5–10×106 EBV+ SLCL on day 0. For Syk inhibition studies, mice were placed on control chow or fostamatinib-containing chow the day prior to tumor cell injection and maintained on chow for the duration of the experiment. For PI3K/Akt inhibition studies, mice were injected intraperitoneally (i.p.) with 50 mg/kg LY294002 or DMSO vehicle three times weekly. Tumors were measured in two perpendicular dimensions along the plane of the body every 2–3 days using Vernier calipers. Tumor volume was calculated as described previously . The slope of the growth curves yielded the tumor growth rates (s) and was calculated by s = ln(v(t1)/v(t2))/(t2-t1), where v(t1) and v(t2) are the tumor volumes at timepoint 1 and 2, respectively (t1 and t2). Tumor doubling times were calculated as ln(2)/s. At the time of killing, gross examination of the animal was performed to assess for tumor. All experimental procedures were performed in accordance with a Stanford Institutional Animal Care and Use Committee approved protocol.
Histology of tumor tissue
Tumor tissue was recovered at the time of killing, fixed in 10% neutral buffered formalin and embedded in paraffin. Serial sections of 4 μm were cut from individual paraffin blocks, deparaffinized in xylene and hydrated in a series of graded alcohol solutions. Sections were stained with hematoxylin and eosin (H&E) and subjugated to immunohistochemistry or labeled by in situ hybridization for EBV RNA using automated immunohistologic instruments (Ventana XT, Ventana Medical Systems, Tucson, AZ, USA). Primary antibodies to human CD20 (L26, 1:1000; Dako, Glostrup, Denmark) and Ki-67 (rabbit polyclonal, 1:1000; Dako) were used. In situ hybridization for EBV was carried out using the EBER inform small RNA probe cocktail and developed using ISH iVIEW nitroblue tetrazolium (Ventana Medical Systems).
Syk inhibition induces apoptosis of EBV+ SLCL in vitro
BCR crosslinking activates Syk, which can be measured by phosphorylation of the Syk substrate BLNK [10, 11]. In Ramos B lymphoma cells, BCR crosslinking induces robust phosphorylation of BLNK, which was ablated by addition of the Syk inhibitor R406 (Figure 1A). Additionally, R406 significantly reduces constitutive Syk signaling in EBV+ cell lines derived from patients with PTLD, termed SLCL . Therefore, R406 inhibits Syk activation.
Previously, we showed that inhibition of Syk diminishes the proliferation of EBV+ SLCL . While proliferation of five of six EBV+ SLCL was sensitive to R406 at submicromolar concentrations, proliferation of the JB7 EBV+ SLCL was more resistant to Syk inhibition . To determine whether the decrease in proliferation was due to cell cycle arrest or apoptosis, SLCL were cultured in the presence of R406 and cell cycle analyses was performed by propidium iodide staining. R406 induced apoptosis (Figures 1B and S1) and cell cycle arrest (Figure 1C) in five of six of the EBV+ SLCL, as shown for the representative line AB5. Notably, R406 was efficacious at inducing both apoptosis and cell cycle arrest at submicromolar concentrations, with exception of the JB7 cell line. These data indicate that R406 can inhibit cellular growth of most EBV+ SLCL through induction of both cell cycle arrest and apoptosis.
Syk inhibition does not affect in vivo tumor burden or tumor formation
We next asked if Syk inhibition affects tumor formation or tumor burden of EBV+ SLCL in vivo. We previously reported that injection of EBV+ PTLD lines into the flank of immunodeficient SCID mice yields tumors that are phenotypically and histologically indistinguishable from primary PTLD tumors . We used this xenogeneic model of PTLD in NOD.SCID mice to determine whether the Syk inhibitor fostamatinib, the prodrug of R406, can inhibit tumor growth of B cell lymphoma lines in vivo.
NOD.SCID mice were given control chow or fostamatinib-containing chow beginning the day prior to tumor inoculation. Control mice (n = 5) injected with the EBV+ SLCL AB5 (in vitro R406 IC50 = 0.625 μM)  developed palpable solid tumors in the third week after inoculation (Figure 2A and S2). On day 33 postinjection, the average tumor volume in control animals was 1344 ± 550 mm3 (mean ± SD). Fostamatinib-treated mice injected with AB5 (n = 6) also developed solid tumors by the third week postinjection (Figure 2A and S2). On day 33 postinjection, the average tumor volume in the fostamatinib-treated mice was 1649 ± 590 mm3, and was not significantly different (p = 0.4) than control. Indeed, tumors in the control group grew at a rate of 0.18 mm3/day with a doubling time of 3.77 days, while tumors in the treated group grew at a rate of 0.2 mm3/day with a doubling time of 3.5 days, for days 28–37. Two additional cohorts of NOD.SCID mice injected with the AB5 cells (5 × 106) were monitored and no differences in time tumor formation or in tumor burden were observed between control (n = 7) and fostamatinib-treated (n = 9) groups (data not shown). Furthermore, no change in time to tumor formation or tumor burden was observed in fostamatinib-treated mice that were injected with another EBV+ SLCL, ZD3 (in vitro R406 IC50 = 0.156 μM)  (data not shown). These data suggest that the formation or growth of solid tumors established from EBV+ SLCL was not affected by fostamatinib.
Serum and tissue samples from mice were analyzed to confirm the presence of active metabolite R406 in the blood and tumor, respectively. R406 was detected in the serum (Figure 2B) and tumor site (data not shown) of animals receiving fostamatinib chow at levels above the reported IC50 values . To confirm R406 activity at the tumor site of treated animals, tumors were recovered at the time of killing from control and fostamatinib-treated animals and tumor lysates were analyzed for the presence of phosphorylated BLNK (pBLNK) by western blot. Indeed, we observed a reduction in pBLNK in fostamatinib-treated tumors, compared to control tumors (Figure 2C). These data suggest that Syk signaling was indeed impaired in fostamatinib-treated animals.
Histological profiles of control and treated tumors
We compared the histologic features of tumors from control and fostamatinib-treated mice. The EBV+ SLCL AB5 grown in NOD.SCID mice show a diffuse proliferation of large pleomorphic cells morphologically comparable to human diffuse B cell lymphoma (Figure 3A). Like the tumors from control mice, the tumors from fostamatinib-treated mice comprised diffuse sheets of atypical, large lymphoid cells that infiltrate normal structures, such as skeletal muscle. There was minimal cellular infiltrate into tumor sites of both control and treated tumors. Nearly 100% of the tumor tissue stained positive for human CD20 and the EBV small RNA, EBER (Figure 3A), indicating that the tumors are of human B cell origin and EBV+. Both control and treated tumors displayed abundant mitotic activity and karyorrhexis, and displayed foci of necrosis. We detected no significant difference in the amount of necrosis assessed in H&E stained sections of tumors from control and fostamatinib-treated mice (Figure 3B) (p = 0.19), nor was there a difference in the number of Ki-67+ proliferating cells (Figure 3A, 3C) (p = 0.13). These findings indicate that Syk inhibition does not affect the morphology or proliferative characteristics of EBV+ B cell lymphomas in vivo.
Syk inhibition results in lymphoma growth in inguinal lymph nodes adjacent to the primary tumor site
We observed that fostamatinib-treated animals displayed tumor growth in inguinal lymph nodes adjacent to the site of tumor inoculation, whereas the majority of control animals did not (Figure 3D and Table 1). Histological analysis of these adjacent inguinal lymph nodes revealed that they were composed of human CD20+, EBER+ tumor cells. There was no difference in necrosis or proliferation in these structures compared to the primary tumor site (data not shown). Interestingly, the EBV+ SLCL JB7 that was resistant to R406 in vitro, was highly metastatic in vivo in both control and fostamatinib-treated mice. In control mice, JB7 tumors metastasized primarily to the liver and abdomen. Fostamatinib-treated, JB7-injected animals displayed fewer abdominal metastases but qualitatively larger adjacent inguinal lymph nodes than control animals (Table 1). Tumor growth at distant sites was minimal in animals injected with the other cell lines. Specifically, four enlarged axillary lymph nodes and one liver metastasis were observed in the fostamatinib-treated mice injected with the SLCL AB5, and one enlarged axillary lymph node was observed in both the control and fostamatinib-treated mice injected with the SLCL ZD3. Tumor cells within the adjacent inguinal lymph nodes were detected more frequently in fostamatinib-treated animals than control animals inoculated with either R406-sensitive (AB5, p<0.0001 and ZD3, p = 0.05) or R406-resistant (BL41, p = 0.06 and BL41.B95, p = 0.06) cells (Table 1). In prior studies we showed that BL41 and BL41.B95 are resistant to effects of Syk inhibition by R406 in vitro . Enlarged inguinal lymph nodes with CD20+ EBER+ tumor cells were found in both the EBV-line BL41, and its EBV-infected counterpart BL41.B95 (Table 1). These data suggest that Syk inhibition is associated with lymphoma localization in the lymph node adjacent to the primary tumor, regardless of in vitro sensitivity to R406 or the status of EBV infection.
Table 1. Sites of tumor growth away from primary injection
Adjacent lymph node
Other lymph node
R406 IC50 (μM)
*p < 0.0001;
**p = 0.05;
***p = 0.06 from Fisher's exact test; + lymph nodes qualitatively larger than control.
Treatment of the EBV+ SLCL AB5 with R406 reduced the phosphorylation of both Akt and Erk in vitro (Figure 4A, right panel). We next asked whether either of these pathways participate in survival of EBV+ PTLD-derived B cell lymphomas. EBV+ SLCL were cultured in the presence of the PI3K/Akt inhibitor LY294002 or the Erk MAPK inhibitor PD98059 and apoptosis was analyzed by Annexin V-EGFP staining. PI3K/Akt inhibition induced a time-dependent increase in apoptosis, while Erk inhibition had no effect on apoptosis (Figure 4B). These data suggest that while both PI3K/Akt and Erk MAPK are activated downstream of Syk, PI3K/Akt, but not Erk MAPK, contributes to survival of EBV+ SLCL.
PI3K/Akt inhibition reduces tumor burden in vivo
We next asked if PI3K/Akt inhibition affects tumor burden in mice inoculated with EBV+ SLCL. In the first experiment, NOD.SCID mice were injected with 5×106 cells of the EBV+ SLCL AB5 on day 0. Treatment with either 50 mg/kg LY294002 (n = 6) or an equivalent amount of vehicle (n = 6) began before any evidence of palpable tumor and continued three times weekly throughout the study. We observed tumor formation in four of six control animals (Figure S3). In contrast, tumor formation was detected in only one of six animals treated with the PI3K/Akt inhibitor LY294002 (Figure S3), and that tumor was the smallest of all tumors identified in vehicle or treated animals at the time of necropsy.
To test whether PI3K/Akt inhibition could affect already established tumors, animals were inoculated with 5×106 cells of the EBV+ SLCL AB5 on day 0. The day after the formation of palpable tumors, treatment with either 50 mg/kg LY294002 (n = 9) or an equivalent amount of vehicle (n = 6) was initiated and continued three times weekly throughout the experiment. By 21 days postinjection, the average tumor volume was significantly different between LY294002-treated and vehicle-treated mice (77 ± 101 mm3 versus 322 ± 105 mm3, p = 0.0006) (Figure 5A). At the time of killing, the average tumor volume was significantly smaller in the LY294002-treated group (171 ± 219 mm3 vs. 998 ± 373 mm3, p = 0.0001). Tumors in the control group grew at a rate of 0.14 mm3/day with a doubling time of 4.9 days, while tumors in the treated group grew at a rate of 0.1 mm3/day with a doubling time of 6.9 days, for days 21–29. The majority of the LY294002-treated animals displayed tumors that grew to a small size, and then either plateaued in growth or disappeared completely (Figure 5B). The two LY294002-treated tumors not fitting this pattern were of comparable size to the smallest control tumors (Figure 5B and Figure S4). These data indicate that PI3K/Akt inhibition attenuates the growth of solid tumors established from EBV+ B cell lines from patients with PTLD and is not associated with lymphoma growth in adjacent inguinal lymph nodes.
Syk inhibitors targeting tonic BCR signals have been pursued in the treatment of B cell lymphomas. Unlike other B cell lymphomas, a tonic, BCR-like signal is delivered to EBV+ B cell lymphomas through Syk activation by self-oligomerization of the viral protein LMP2a. Treatment options for EBV-associated PTLD B cell lymphomas are limited and typically involve reduction of immunosuppression, which is often ineffective and puts the transplanted organ in jeopardy of rejection. Consequently, there is a need for new, more effective therapeutics that target the lymphoma without endangering graft survival.
Using R406, the active metabolite of the Syk inhibitor fostamatinib, we provide evidence for the requirement of Syk activation in PI3K/Akt and Erk MAPK signaling, and in survival of EBV+ B cell lymphomas in vitro. We also establish that PI3K/Akt signaling, but not Erk MAPK signaling, is a critical survival pathway for PTLD-derived EBV+ B cell lymphomas. Given that Syk activation is required for survival in vitro, we expected to observe a similar benefit in vivo. However, Syk inhibition had no effect on primary tumor burden in a xenogeneic model of PTLD. Instead, Syk inhibition resulted in tumor growth in inguinal lymph nodes adjacent to the site of tumor inoculation. Importantly, PI3K/Akt signaling is a key outcome of Syk activation in PTLD, because PI3K/Akt inhibition is efficacious in reducing tumor burden without tumor growth in adjacent lymph nodes.
Deregulated PI3K/Akt signaling is characteristic of a myriad of human cancers as a result of mutations in members of the PI3K/Akt pathway, including the PI3K catalytic subunit p110α or Akt. PI3K/Akt deregulation also results from mutation, amplification, or overexpression of the PI3K-activating receptor tyrosine kinases (RTKs), such as epithelial growth factor receptor (EGFR) or human epidermal growth factor receptor 2 (HER2) . Intriguingly, preliminary data from phase I clinical trials of PI3K or Akt inhibitors in tumors displaying a mutation in the gene encoding the PI3K p110α subunit (PIK3CA), including endometrial and colorectal cancers, show few responses but rather the establishment of tumor stasis, or stable disease . In our model of EBV+ B cell lymphomas, we observed both tumor stasis and tumor regression in the majority of animals treated with the PI3K inhibitor LY294002. To our knowledge, EBV+ PTLD lymphomas do not contain mutations in either PI3K or Akt. Instead, PI3K/Akt activation is driven by the EBV latent membrane proteins, LMP2a and LMP1, and more closely resembles PI3K-activation by receptor tyrosine kinases (RTKs). Therapies targeting PI3K-activating RTKs, however, such as herceptin for HER2-positive breast cancers and gefitinib or erlotinib for EGFR signaling in non-small cell lung carcinoma, pancreatic cancer and glioblastoma have been very efficacious in the clinic. These therapies provide some of the strongest evidence for the concept of oncogene addiction, the observation that despite the complexity of cancerous cells, their growth is often impaired by inactivation of a single oncogene . Given the apoptosis we observe in vitro and the tumor regression we identified in our xenogenic model of PTLD with PI3K/Akt inhibition, our data suggest that EBV hijacks and re-wires cellular signaling pathways through LMP2a and LMP1 to create a state of oncogenic addiction to PI3K/Akt in PTLD-derived EBV+ B cell lymphomas. It will be of interest in the future to test whether oncogenic addiction to PI3K/Akt signaling is operating in EBV+ B cell lymphomas.
The Syk inhibitor R406 induced both apoptosis and cell cycle arrest in a panel of PTLD-derived EBV+ B cell lines in vitro. However, we observed no significant effect of Syk inhibition on time to tumor formation or tumor burden in a xenogeneic model of PTLD, despite the presence of active drug in the serum and at the tumor site of treated animals. Syk inhibition by fostamatinib or the Syk inhibitor curcumin was efficacious in murine models of other non-PTLD B cell lymphomas including diffuse large B cell lymphoma (DLBCL) , B cell non-Hodgkin's lymphoma (NHL)  and chronic lymphocytic leukemia (CLL) . In these studies, murine tumor cell lines were administered to immunocompetent, syngeneic animals in a disseminated fashion—by i.p. injection for CLL, and by i.v. injection for DLBCL and B cell NHL. Thus, several key differences between those models and our model could account for the disparate findings. These include the varied types of lymphomas studied, the use of syngeneic murine cell lines in immune competent mice versus our use of human cell lines in immunodeficient mice and the use of disseminated tumor models versus our use of a solid tumor model.
These studies are the first to report B cell lymphoma growth in lymph nodes adjacent to the primary tumor site after Syk inhibition. This occurred regardless of in vitro sensitivity to R406 or the status of EBV infection of the cell lines tested. Plausible explanations for this observation include a role for Syk in regulating the peripheralization of lymphoma cells or homing to local lymph nodes, an off-target effect of the inhibitor fostamatinib, or a feature of the xenogeneic model of PTLD we utilized. Interestingly, in phase I/II trials with fostamatinib, a transient increase in circulating lymphocytes was observed in all patients with small lymphocytic leukemia (SLL) or chronic lymphocytic leukemia (CLL); this increase exceeded 50% of baseline circulating lymphocytes in 9 out of 13 of patients . While we did not test for a similar peripheralization of lymphoma cells, the paucity of metastases at distant sites suggests that Syk inhibition predominantly affects lymph node homing. We did not observe the accumulation of tumor cells in adjacent lymph nodes in animals treated with the PI3K/Akt inhibitor LY294002. Given that Syk also activates a variety of other signaling pathways, including Erk MAPK, Jun N-terminal kinase (JNK), protein kinase C (PKC) and nuclear factor of activated T cells (NFAT) [20-22], it is possible that another downstream pathway is involved in the accumulation of tumor cells in adjacent lymph nodes. With regard to off-target effects of fostamatinib and its active component R406, the kinases FLT3, c-Kit, Lck, JAK1, JAK3 and VEGFR2 have been reported to be inhibited at doses relevant to our study . For example, given that VEGFR2 signaling can regulate angiogenesis, tumor growth in the adjacent lymph node in fostamatinib-treated animals may reflect tumor migration from a hypoxic environment. Whether these molecules play a role in dissemination of EBV+ B cell lymphomas is unknown.
Activation of PI3K/Akt signaling is not entirely Syk-dependent; RTKs, G-protein coupled receptors, integrins and cytokine receptors can all activate the PI3K/Akt pathway independent of Syk. Therefore, it is possible we did not observe a reduction in tumor burden in fostamatinib-treated animals because PI3K/Akt signaling was concurrently activated by a Syk-independent mechanism. For example, cell-to-cell contact, cytokines, or other growth factors may provide cues that override or augment the Syk survival signal in EBV+ SLCL in vivo that are not observed in vitro. Indeed, the efficacy of Syk inhibition in CLL is attributed to attenuation of both tonic B cell signals as well as Syk-driven signals stimulated in CLL by stromal cell contact . In a broad sense, this may imply that efficacy of Syk inhibition depends on its ability to dampen signaling by a variety of oncogenic pathways, including PI3K/Akt.
Taken together, our data indicate that PI3K/AKT signaling is a critical outcome of Syk activation that drives EBV+ B cell lymphoma survival. These findings also indicate the possibility that PI3K/AKT oncogenic addiction may be exploited as a therapeutic target for PTLD, and potentially other EBV-associated malignancies.
We would like to thank Rigel, particularly Drs. Polly Pine, Ann Lowe and Yasumichi Hitoshi, and AstraZeneca for kindly providing fostamatinib and R406, the Stanford Graduate Fellowship (OLH) and NIH RO1 AI41769 grants (OMM), the Arnold and Barbara Silverman Fund for funding and Dr. Karine Piard-Ruster for technical assistance.
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.