|Home | About | Journals | Submit | Contact Us | Français|
Inhibition of the PI3K/mTOR pathway is an appealing method for decreasing immunoresistance and augmenting T cell mediated immunotherapy. A major impediment to this strategy is the impact of conventional PI3K/mTOR pathway inhibitors on T cell function. In particular, Rapamycin, a well known immunosuppressant can decrease activity of the PI3K/mTOR pathway in tumor cells, but also has a profound inhibitory effect on T cells. Here we show that Honokiol, a natural dietary product isolated from an extract of seed cones from Magnolia grandiflora, can decrease PI3K/mTOR pathway mediated immunoresistance of glioma, breast and prostate cancer cell lines, without affecting critical pro-inflammatory T cell functions. Specifically, we show that at doses sufficient to downregulate levels of phospho-S6 and the negative immune regulator B7-H1 in tumor cells, Honokiol does not significantly impair T cell proliferation or pro-inflammatory cytokine production. In contrast to classic inhibitors, including LY294002, Wortmannin, AKT inhibitor III and Rapamycin, Honokiol specifically decreases PI3K/mTOR pathway activity in tumor cells, but not in freshly stimulated T cells. Collectively, our data define a unique application for Honokiol and provide the impetus to more fully elucidate the mechanism by which T cells are resistant to the effects of this particular inhibitor. Honokiol is clinically available for human testing and may serve to augment T cell mediated cancer immunotherapy.
Cancer therapy is predicated upon specificity without toxicity. The most effective therapies have been able to exploit a unique attribute of the cancer cell with minimal impact on normal tissue. In this regard immunotherapy is very appealing as an approach to target unique tumor associated antigens that are not expressed by healthy cells. Successful T cell mediated immunotherapy requires expansion of both CD4 and CD8 positive tumor specific lymphocytes, followed by target lysis, and subsequent development of a memory T cell population that can serve to target cancer recurrence. As a result, several recent clinical trials have employed anti-tumor vaccination strategies in an attempt to activate and expand tumor specific immune cells1–3. Many anti-cancer vaccination techniques have been described to evoke ex vivo T cell mediated immunity, with documented peripheral immune responses that are tumor specific4,5. These vaccines include tumor specific peptides6, dendritic cell approaches7, and tumor derived heat shock proteins8. However, in the majority of clinical trials there has been no clear correlation between systemic immune responses and clinical efficacy. T cells in the periphery can be generated in response to vaccine, but they do not consistently kill their tumor target, suggesting that an intrinsic defect in T cell responses in the tumor environment.
The discordance between peripheral immune response and clinical efficacy highlights cancer immunoresistance as an emerging problem in immunotherapy, the role of tumor cell immunoresistance. Many types of solid malignancies, including breast, prostate and brain tumors have well characterized mechanisms of immunosuppression. Local and systemic inhibition of T cell activity can be due in part to factors secreted into the local milieu9,10 or surface expression of proteins that prevent activation11 or induce apoptosis12 of effector cells. In some cases, immunoresistance is correlated with oncogenic pathway activation13. We recently uncovered a link between the PI3K/mTOR pathway and expression of the immunoresistant protein B7-H1 in glioma, breast and prostate carcinoma cells14,15. Other mechanisms of PI3K/mTOR pathway mediated cancer immunoresistance have been described16,17, suggesting that PI3K/mTOR inhibition may serve as a means to augment T cell mediated immunotherapy. However, a major impediment to this strategy is the impact of conventional PI3K/mTOR pathway inhibitors on T cell function. For example, Rapamycin, a well known immunosuppressant can decrease activity of the PI3K/mTOR pathway in tumor cells18–20, but also has a profound effect on T cells21. The ideal compound would inhibit the PI3K/mTOR in tumor targets without disrupting T cell function or proliferation.
Here we show that Honokiol, a natural product isolated from an extract of Magnolia grandiflora seed cones, can decrease PI3K/mTOR mediated immunoresistance in glioma, breast and prostate cancer cells through PI3K/mTOR pathway inhibition, without significantly affecting pro-inflammatory T cell function. First characterized for its antimicrobial activity22, Honokiol has been shown to have anti-angiogenic properties as well23. Specific anti-tumor properties include induction of caspase-dependent apoptosis in B cell chronic lymphocytic leukemia cells24, and inhibition of the bone metastatic growth of human prostate cancer cells25. Most recently, work in breast cancer cells has shown that Honokiol can attenuate the PI3K/Akt/mTOR signaling by down-regulation of Akt phosphorylation, and upregulation of PTEN expression26. In this report we show that Honokiol is effective in the prevention of PI3K/mTOR signaling pathway in glioma, breast and prostate tumor cells, including decreased expression of B7-H1. Given the role of B7-H1 in the induction of T cell apoptosis, our findings indicate that PI3K/mTOR inhibition by Honokiol can enhance T cell specific tumor responses both through increased T cell survival as well as maintenance of pro-inflammatory T cell functions. Together these data support the clinical study of Honokiol as a means to augment T cell mediated immunotherapy.
Protein concentration of cellular extracts was measured using Protein Assay reagent (Bio-Rad) relative to bovine serum albumin standards. Protein (30 µg) was subjected to SDS-PAGE and electroblotted onto Immobilon-P membrane (Millipore). The membrane was blocked at 4 °C overnight and incubated with mouse monoclonal antibody to α-tubulin, p70 S6K1 or phospho-S6K1 (Thr-389); rabbit monoclonal antibody to phospho-AKT (S473) (Abcam, Cambridge, MA); or goat polyclonal antibody to B7-H1 (Santa Cruz Biotechnology) followed by incubation with isotype specific secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA). Bound protein was visualized using chemiluminescence western blotting detection reagents (Amersham, Piscataway, NJ). Densitometric measurements were acquired and α-tubulin was used to verify equal loading.
1×106 tumor or T cells were incubated for 48 hours with the following inhibitors: LY294002 (100µM) (Ly294), Wortmannin (100µM) (WORT), AKTIII inhibitor (50µM) (AKTIII), Honokiol (20µM) (HNK) or Rapamycin (100µM) (Rapa). Honokiol was a gift from Dr. Jack Arbiser at Emory University, all other inhibitors were obtained from Calbiochem, San Diego, CA). All inhibitors were reconstituted in DMSO (vehicle). Where indicated, vehicle treatment refers to cells treated only with DMSO for the same duration.
Primary blood lymphocytes (PBL) were isolated from whole blood by Ficoll centrifugation in accordance with the UCSF Committee for Human Research approved protocol # H47634-01655-34. Informed consent was obtained for collection of PBLs from patients with non-malignant tumors. PBLs were enriched for CD3+ T cells using negative magnetic cell selection according to manufacturer’s instructions (StemCell Technologies, British Columbia). CD3+ cells were stimulated for 24 hours with 5µg/ml plate bound anti-human CD3 (OKT3) and 2µg/ml soluble anti-human CD28 (CD28.2) prior to coculture. 5×104 tumor cells were plated in a 96 well round bottom plate and allowed to adhere overnight, 5×104 enriched and stimulated CD3+ cells were then added in a final volume of 200µl.
Cells were cocultured as described above for 4 hours and subsequently surface stained on ice with CD45 APC (BD Pharmingen, San Diego, CA) in PBS/ 2% BSA for 20 minutes and washed three times. Cells were then stained with Annexin V FITC and 7-AAD (BD Pharmingen) for 10 minutes and fixed in 1% formaldehyde containing 5µg/ml 7-AAD and analyzed by flow cytometry. Cells treated with 1 µM staurospirine (Sigma Aldrich, St. Louis, MO) served as a positive control for apoptosis.
CD3+ cells were enriched as described above and stained with10µM carboxyfluorescein succinamidyl ester (CFSE) for 10 minutes at 37°C. Cells were quenched with 100% FBS and washed three times in serum containing growth media (RPMI 1640). Cells were then CD3/CD28 stimulated for 72 hours, stained with CD45 APC, fixed in 2% paraformaldehyde and analyzed by flow cytometry.
Following CD3/CD28 stimulation in the presence of Brefeldin A (Sigma-Aldrich), CD3+ cells were surface stained with CD8 FITC, CD4 PerCP (BD Pharmingen) and/or CD25 APC (eBioscience, San Diego, CA) on ice for 20 minutes and washed three times. Cells were then fixed in 2% paraformaldehyde for 1 hour. CD3+ cells were then stained using IFNγ APC or IL-17 PE for the detection of intracellular cytokines (eBioscience) in saponin containing buffer (PermWash, BD Pharmingen). Cells were subsequently analyzed by flow cytometry.
Data for all figures were collected from multiple independent experiments performed in triplicate. Representative experimental data is shown. Statistical significance was determined using a two-tailed student’s t test with p < 0.01. Error bars represent +/− standard deviation.
We first aimed to demonstrate that Honokiol could inhibit activation of the PI(3)K/mTOR pathway in PTEN deficient glioma cell lines. We found that in U87 and U251 glioma cell lines, treatment with Honokiol inhibited PI(3) kinase activation as demonstrated by a decrease in phospho AKT, phospho S6 kinase and phospho S6 proteins (Figure 1A). As expected, inhibition of this pathway was associated with a decrease in B7-H1 protein expression. Inhibition of this pathway following Honokiol treatment was similar to that observed following treatment with other well characterized inhibitors of the PI3K/mTOR pathway, LY294002, Wortmannin, AKT inhibitor III and Rapamycin. We have previously demonstrated that a decrease in B7-H1 protein expression on tumor cells promotes the survival of activated T cells through inhibition of B7-H1 mediated apoptosis14. To determine if Honokiol treatment mediates a similar effect, we compared T cell survival following coculture with treated U87 and U251 cell lines before and after exposure to Honokiol (Figure 1B). We found a significant decrease (p<0.01) in the percentage of T cells undergoing apoptosis when compared to cocultured cells treated with vehicle alone. The percentage of T cells undergoing apoptosis following coculture with U87 or U251 was reduced to levels comparable to those observed following treatment with any other PI3K/mTOR pathway inhibitor tested (Average= 6.59% and 10.7%, respectively), suggesting that the decrease in PI3K/mTOR pathway activation observed by western blot in Figure 1A has functional consequences on tumor cell mediated induction of T cell apoptosis.
To determine if this observation was a phenomenon restricted to PTEN deficient glioma cells, we next examined the effect of Honokiol treatment on PTEN deficient breast (Figure 2A) and prostate cell lines (Figure 2B). As observed in glioma cell lines, Honokiol treatment of BT549 and PC-3 cell lines significantly reduced downstream phosphorylated components of the PI3K/mTOR pathway to levels similar to those observed with treatment of other well-characterized PI3K inhibitors. Following coculture with stimulated T cells, Honokiol treated tumor cells also significantly reduce the induction of T cell apoptosis as measured by Annexin V staining to a similar degree as the other inhibitors tested (Figure 2B). Interestingly, the decrease in T cell apoptosis following coculture with Honokiol treated BT549 cells was significantly less (p<0.05) than that observed following treatment with Ly294002, Wortmannin, AKT inhibitor III or Rapamycin, suggesting that Honokiol either promotes T cell survival, or better inhibits tumor cell mediated T cell apoptosis.
One possible explanation for the reduction in T cell apoptosis following coculture with Honokiol treated tumor cells is that the treatment is toxic to tumor cells, reducing the number of tumor cells in coculture. To address this possibility, we treated tumor cells with doses of Honokiol, Ly294002, Wortmannin, AKT inhibitor III or Rapamycin at doses sufficient to inhibit PI3K/mTOR signaling and assessed the amount of induced tumor cell death prior to coculture. We found similar levels of Honokiol was no more toxic to tumor cells than the other inhibitors (Figure 3A), suggesting that preferential killing of tumor cells by Honokiol does not account for the reduction in T cell apoptosis observed in Figure 1. When inhibitor concentrations were maintained throughout T cell-tumor cell coculture (Figure 3B), we found that T cell apoptosis was decreased in Honokiol treated tumor cells, while in the presence of Rapamycin, Wortmannin and AKT inhibitor III, T cell death was significantly increased. This may be due to induction of T cell apoptosis as a result of inhibitor toxicity. Ly294002 treatment also increases T cell apoptosis to a lesser degree, despite all of the inhibitors reducing tumor cell survival and B7-H1 expression on tumor cells (Figure 1).
Having demonstrated that Honokiol mediated similar effects on tumor cell signaling and a reduction in tumor cell mediated T cell apoptosis, we sought to determine the effects of Honokiol on activated T cell responses. We found that in contrast to other inhibitors, Honokiol supported proliferation of a high percentage of T cells following stimulation (Figure 4). The slight reduction of proliferating T cells from 25.9% to 22.9 % in response to vehicle treatment (Figure 4B) was not statistically significant (p=0.58), suggesting a minimal effect of Honokiol on T cell expansion. All other PI3K/mTOR inhibitors tested significantly reduced levels of proliferation to below that observed in unstimulated cells (8.83%), suggesting that baseline proliferation in patient lymphocytes is also affected by these inhibitors. Inclusion of inhibitors during the coculture with tumor cells did impair T cell proliferation (data not shown), suggesting that the observed effects are due to anti-inflammatory properties of tumor cells and their effects on T cell proliferation, rather than the presence of Honokiol in the coculture. It has been reported, however, that Honokiol has anti-inflammatory properties in T cells, inhibiting the proliferation of the lipopolysaccharide responsive CTLL-2 cell line26. To determine if we could detect a similar inhibition of proliferation in primary T cells, we measured T cell proliferation following increasing doses of Honokiol treatment. At doses sufficient to inhibit PI3K/mTOR signaling in tumor cells (20µM), T cell proliferation is relatively unaffected. However, at doses approaching 80µM, T cell proliferation can be inhibited, suggesting that Honokiol does have the potential to act as an anti-inflammatory agent at relatively high doses.
Following stimulation of T cells, several activation markers are upregulated on the cell surface, including the IL-2 receptor, CD25. CD25 is an activation marker, but also functionally important in immune responses as activated T cell proliferation is dependent on IL-2 signaling through CD25. To determine if the difference in T cell proliferation observed following inhibitor treatment could be attributed to differential expression of CD25, we examined the surface expression on CD4 negative, CD8 positive, T cells (Figure 5A). While Honokiol does significantly decrease the percentage of CD25 expressing cells (p<0.01) relative to cells treated with vehicle only, we find that 55.5% of stimulated CD8 T cells express CD25, compared to only 9.9% following Wortmannin treatment, and 27.0% following AKT inhibitor III treatment, which could explain the differences in T cell proliferation. Interestingly, Ly294002 treatment and Honokiol treatment had similar effects on CD25 expression (51.5%), suggesting a CD25 independent mechanism for abrogation of T cell proliferation observed in Ly294002 treated T cells.
Successful immunotherapy is dependent upon pro-inflammatory CD8 T cell responses characterized by either IFNγ or IL-17 production27,28. To determine if Honokiol treatment prevented pro-inflammatory cytokine production in response to stimulation, we next measured the effect on cytokine production by CD8 T cells. We found that Honokiol treatment did not significantly impair the ability of CD8 T cells to produce IFNγ (Figure 5B) and IL-17 (Figure 5C) when compared to T cells treated with vehicle alone. In contrast, all of the other inhibitors tested significantly reduced the percentage of pro-inflammatory cytokine producing cells. Importantly, the amount of cytokine produced on a per cell basis was similar to that observed in untreated cells following stimulation (Figure 5B and 5C). T cell cytokine responses during tumor cell coculture were further decreased in the presence of inhibitor (data not shown), indicating that tumor cells with reduced PI3K/mTOR have an intrinsic ability to impair T cell pro-inflammatory responses. This suggests that Honokiol treatment would best function in conjunction with other treatments aimed to offset the anti-inflammatory tumor microenvironment. Together, these data indicate that Honokiol treatment at concentrations sufficient to downregulate PI3K/mTOR pathway signaling, and B7-H1 expression, has a minimal effect on the pro-inflammatory response of adaptive immune effector cells.
We assessed the levels of phosphorylated S6, an endpoint of PI3K/mTOR signaling in T-cells following treatment with Honokiol to determine if the retention of proliferation and pro-inflammatory cytokine production occurs in spite of impaired PI3K/mTOR /S6 signaling in T cells. Whereas AKT inhibitor III treatment of T cells decreased levels of phospho-S6 following stimulation, Honokiol treatment did not (Figure 6A). However, Honokiol treatment of the PTEN deficient glioma cell line U87, did result in a 60% reduction of phospho-S6 levels (Figure 6B). Because activation of S6 is important for protein translation and activated T cell function, these data suggest that T cells may have a redundant mechanism to activate S6 despite the potent PI3K/mTOR pathway inhibition by Honokiol.
T cell mediated cancer immunotherapy requires that a tumor specific CD8 T cell can identify and effectively target the cancer cell. Local and systemic immunoresistance is therefore a significant impediment to successful immunotherapy. Tumor cells have several mechanisms that facilitate immunoresistance and immunosuppression, including TGFβ secretion29 or CTLA-4 expression30. We have previously linked the expression of B7-H1 protein to activation of the PI3K/mTOR pathway14,15, identifying this oncogenic pathway as a potential target for overcoming one known mechanism of immunoresistance in glioma, breast and prostate cancer cells, resulting in T cell apoptosis. Here we have shown that conventional inhibitors (LY294002, Wortmannin, Rapamycin, AKTIII) can reduce T cell apoptosis in coculture with PTEN deficient tumor cell lines, but that these inhibitors all have a significant impact on T cell pro-inflammatory function. Specifically, these inhibitors reduced T cell proliferation, decreased the percentage of CD8 positive T cells expressing interferon gamma and IL-17, as well as decreased the percentage of T cells expressing the activation marker CD25. Accordingly, the strategy of augmenting T cell mediated immunotherapy by inhibiting the PI3K/mTOR pathway using these conventional agents is likely to be minimally effective due to side effects on T cell function.
We first considered Honokiol as a possible candidate based on its proven ability to inhibit the PI3K/mTOR pathway in other cancers26, as well as its wide use as a herbal medicine, suggesting that it would be well tolerated as a clinical agent and unlikely to negatively impact the immune system. Here we show that Honokiol can inhibit B7-H1 expression in three types of PTEN deficient tumor cells as effectively as other conventional inhibitors. In doing so, Honokiol may also reduce the expression of other immunoresistant proteins that are the product of PTEN loss in these cells, by reducing cellular S6 mediated translation. In contrast to other conventional inhibitors, Honokiol has a minimal effect on the proliferation of T cells, expression of activation markers or pro-inflammatory cytokine production. The results of these experiments provide the impetus to more fully explore the clinical utility of Honokiol as an adjunct to active immunotherapy. Together with previously published work showing in vitro and in vivo inhibition of both tumor cell proliferation25 and angiogenesis23, our data demonstrate a promising application for this natural PI3K/mTOR inhibitor that has previously not been described. The data presented here are also consistent with recently published studies using xenograft tumor models23,31, suggesting a potential broad based application in patients with cancer.
Other published reports have indicated that Honokiol has anti-inflammatory effects, inhibiting the production of inflammatory cytokines32, scavenging free radicals produced during an ongoing immune response33 in apparent contrast to our findings. However, the cytokines and responses evaluated in these studies are those produced during an innate immune response such as nitric oxide, IL-6 and TNFα, rather than those made during adaptive immune responses. One recent study suggests that Honokiol behaves instead as an anti-inflammatory agent in T cells26. Possible explanations for this discrepancy relate to the type of cells and stimulations used in the experiments. Here we describe the effects of low Honokiol doses on primary T cells stimulated with anti-CD3/CD28, while the previous work used the CTLL-2 cell line following lipopolysaccharide or Conconavalin A stimulation. Following treatment of T cells with increasing doses of Honokiol, we also find that this inhibitor can block T cell proliferation (Figure 4), suggesting that in some conditions, Honokiol may be useful as an anti-inflammatory agent. Taken together with the data presented here, Honokiol may have the potential to inhibit the nonspecific tissue damage that innate cytokines and free radicals can cause and maintain T cell function and the specificity of adaptive immune therapy at doses sufficient to prevent PI3K/mTOR signaling.
The mechanism by which T cells are apparently resistant to the effects of Honokiol remains to be elucidated. Our data suggest that PI3K/mTOR pathway is unaffected in stimulated T cells exposed to Honokiol, in contrast to tumor cells exposed to the same dose over the same time course. The resistance may relate to redundancy in signaling pathways in T cells, oncogenic addiction in transformed cells, biochemical properties that distinguish permeability of T cells from tumor cells, or a combination of multiple factors. Regardless of the specific mechanism responsible for the discordance between glioma and lymphocytes, the biochemical and functional data in this report provide new insights into the potential clinical utility of Honokiol, within the context of immunotherapy for patients harboring PTEN deficient tumors.
Further studies are needed to extend these results into the clinical realm, including the use of pre-clinical animal models. Current studies are ongoing in our laboratory to evaluate the efficacy of Honokiol treatment on the growth of glioma cells in vivo. Longitudinal studies of these animals will reveal the ability of T cells to eliminate tumor cells treated with Honokiol, prolong animal survival and reduce tumor burden. In these studies, ex vivo analysis of T cells will also determine optimal doses and treatment regimens that promote T cell pro-inflammatory function while preventing PI3K/mTOR activation in tumor cells. We anticipate that higher doses will be required to achieve the results obtained in vitro due to the complexity of the tumor microenvironment. However, the widespread dietary use of Honokiol suggests that this agent will be well tolerated at high doses. It has also been shown that a dosage of between 35 and 65 mg/kg three times a week is well tolerated in animal models and can induce tumor cell apoptosis in vivo, providing a median testable range for T cell function and toxicity in tumor bearing animals31.
In conclusion, we have identified a previously unknown feature of Honokiol that may offer a therapeutic advantage to patients being treated with T cell based immunotherapy protocols. In vivo, Honokiol treatment may decrease tumor cell survival and PI3K/mTOR mediated immunoresistance to adaptive immune responses. Taken with the current interest in T cell mediated immune therapies for patients with many types of cancer, our data indicate that Honokiol may be best considered in combination with other therapies in order to maintain immune cell function during treatment.
Financial Disclosure: JA has filed for international patent rights for Honokiol. Emory University has filed for US rights. All other authors have declared there are no conflicts of interest in regards to this work.