Abstract
Demonstrations of both pro-apoptotic and pro-survival abilities of Fas (TNFRSF6/CD95/APO-1) have led to a shift from the exclusive “Fas apoptosis” to “Fas multisignals” paradigm and the acceptance that Fas-related therapies face a major challenge, as it remains unclear what determines the mode of Fas signaling. Through protein evolution analysis, which reveals unconventional substitutions of Fas tyrosine during divergent evolution, evolution-guided tyrosine-phosphorylated Fas proxy, and site-specific phosphorylation detection, we show that the Fas signaling outcome is determined by the tyrosine phosphorylation status of its death domain. The phosphorylation dominantly turns off the Fas-mediated apoptotic signal, while turning on the pro-survival signal. We show that while phosphorylations at Y232 and Y291 share some common functions, their contributions to Fas signaling differ at several levels. The findings that Fas tyrosine phosphorylation is regulated by Src family kinases (SFKs) and the phosphatase SHP-1 and that Y291 phosphorylation primes clathrin-dependent Fas endocytosis, which contributes to Fas pro-survival signaling, reveals for the first time the mechanistic link between SFK/SHP-1-dependent Fas tyrosine phosphorylation, internalization route, and signaling choice. We also demonstrate that levels of phosphorylated Y232 and Y291 differ among human cancer types and differentially respond to anticancer therapy, suggesting context-dependent involvement of Fas phosphorylation in cancer. This report provides a new insight into the control of TNF receptor multisignaling by receptor phosphorylation and its implication in cancer biology, which brings us a step closer to overcoming the challenge in handling Fas signaling in treatments of cancer as well as other pathologies such as autoimmune and degenerative diseases.
Signalling by the tumor necrosis factor receptor (TNFR) superfamily member Fas can promote either survival or death of a cell, but the mechanism underlying this choice is unclear. This study reveals that the outcome of Fas signalling (death versus survival) is determined by the tyrosine phosphorylation status of its death domain.
Author Summary
The versatility of the tumor necrosis factor receptor superfamily members in cell fate regulation is well illustrated by the dual signaling generated by one of the most extensively studied members of the family, Fas (CD95/TNFSFR6). Upon binding its ligand, Fas is able to elicit both pro-death and pro-survival signals. Until now, we have lacked mechanistic knowledge about when and how one signaling output of Fas is favored over the other. We demonstrate here that the outcome of Fas signaling is determined by the phosphorylation status of two tyrosine residues (Y232 and Y291) within the death domain. Dephosphorylation of Fas tyrosines by SHP-1 tyrosine phosphatase turns on the pro-apoptotic signal whereas the tyrosine phosphorylation by Src family kinases (SFKs) turns off the pro-apoptotic signal and turns on the pro-survival signal. Furthermore, we provide evidence that Fas tyrosine phosphorylation status may vary among different cancer types and influence the response to anti-cancer treatments. This information reveals an opportunity to use the screening of Fas tyrosine phosphorylation, a newly discovered direct molecular indicator of Fas functional output, to aid the design of Fas-related cancer therapies.
Introduction
Fas (CD95/APO-1/TNFRSF6), a tumor necrosis factor (TNF) receptor superfamily member, is a well-known apoptosis activator. The binding with Fas ligand (FasL) can lead to the recruitment of Fas-associated protein with death domain (FADD) and procaspase-8, forming the death-inducing signaling complex (DISC). This results in the activation of the caspase cascade and, ultimately, apoptosis [1]. Fas was mainly considered as a tumor suppressor thanks to its familiar ability to promote programmed cell death (apoptosis). However, accumulating evidence supports a significant role of Fas in the alternative non-death signaling leading to cell survival, proliferation, motility, epithelial-mesenchymal transition, cancer growth, and metastasis in some contexts [2]. While such conditional multisignaling of Fas has also been well demonstrated in several cancer models, including colon cancer [3–5], the mechanism controlling these multisignals is unclear.
Fas multiple signaling implies an efficient molecular switch mechanism that lends itself to a flexible formation of different signaling complexes depending on the type of signal being transmitted. One of such mechanisms to be considered is tyrosine phosphorylation. While it has been shown for almost two decades that both tyrosines in the intracellular domain, Y232 and Y291, can be phosphorylated [6], the role of their phosphorylation is not understood. Due to the lack of tools for functional analysis and site-specific phospho-tyrosine (pY) detection, there have been only few conflicting reports that infer functions of Fas phosphorylation [7–8] and, thus, limiting our understanding of the functions of each pY in Fas signaling.
Taking a unique approach based on Fas protein evolution analysis, evolution-guided Fas pY proxy, and site-specific detection of Fas pY, we show for the first time that Fas death domain tyrosine phosphorylation is a dominant anti-apoptosis and a pro-survival mechanism. We discover that while phosphorylations at Y232 and Y291 share the anti-apoptotic function, they differ in terms of structural requirement and other functions. We also reveal the regulation of Fas pY by SFK/SHP-1-based system and the functions of death domain pY in the control of DISC formation and clathrin-dependent Fas endocytosis. Furthermore, we present the implication of the pY-based regulation of Fas signaling in different human cancers along with potential means to predict Fas signaling modes, which is crucial for Fas-related therapeutic design to achieve clinical success.
Results
Evolution-Guided Analysis and Site-Directed Mutagenesis Demonstrate that Death Domain Tyrosine Phosphorylation Is Dispensable for Fas-Induced Apoptosis
To date, the active role of each pY of the Fas death domain in apoptosis induction by FasL in the human cell system is unknown. To approach this issue, we turned toward the evolution of Fas protein as a guide by considering the substitutions of amino acids at these phosphorylation sites during the course of evolution. Multiple sequence alignment of Fas proteins from vertebrates illustrates that the side chain size and aromatic ring feature are highly conserved at position 232. This, however, is not the case for position 291, where neither the size nor the aromatic side chain of tyrosine is a substitution criterion (Fig 1A; positions Y232 and Y291 of human Fas are used as references to indicate corresponding amino acid positions in other species throughout the text). Notably, substitution of Y by a small amino acid, cysteine (C), is common among primates (particularly in old world monkeys) and rodents, which are relatively close to hominoids (apes, including human). Further in evolutional distance, one can also observe the substitution of Y291 by a small amino acid, alanine (A), in some fishes, including coelacanth (the living fossil) and cod.
Small amino acid substitutions for Y291 in closely related species do not appear to impact the apoptotic functions of Fas. Previous work has shown that, like in human and mouse, Fas in cynomolgus monkey and rat that carries a C at position 291 could signal apoptosis upon ligation with an agonistic antibody [10] and FasL [8,11], respectively. The observation that Y at the position 291 is interchangeable with C among closely related species whose Fas can function as an apoptosis inducer suggests that the presence of Y at this position and, thus, its phosphorylation is not essential for Fas apoptotic signal. Our observation that in several human cell types unphosphorylated mutants (Y232F and Y291F) could transmit apoptotic signals supports this conclusion (Figs 1B, S1 and S2).
Evolution-Guided, Site-Directed Mutagenesis Reveals the Anti-Apoptotic Role of Death Domain Tyrosine Phosphorylation
The above-mentioned results led us to hypothesize that the advantage of Fas death domain pY, if it occurred through evolution, was to provide a reversible switch from the apoptotic signal to other signals, e.g., the survival signal. However, to clarify whether Fas pY plays active roles in cellular processes, an ability to induce and maintain, or mimic the properties of, the phosphorylated state of amino acid residues of interest in cells is required.
A comparative genomic study of Raf kinases shows that pY could have evolved from smaller acidic amino acids such as aspartic acid (D) or glutamic acid (E) [12]. Thus, substituting D or E for pY may mimic the phosphorylated state of some proteins [13–16]. However, such substitutions require careful consideration to ensure that the observed results are not due to the change in the amino acid size.
The common substitution between small amino acids and Y at position 291, but not at position 232, of Fas in vertebrates suggests that, depending on the sites, net charge can be more important than the details of the side chain structure. To investigate this issue, we performed evolution-guided, site-directed mutagenesis to examine the functional effects of the following amino acid substitutions on Fas: 1. Size reduction by (a) substituting Y232 with C, as observed in reptile and in cases of human autoimmune lymphoproliferative syndrome (ALPS) [17], (b) substituting Y291 with A and C, as observed in fish and mammals respectively; and 2. negative charge addition by substituting Y232 and Y291 with the acidic D, which has a size comparable to C, the most common small amino acid substituting for Y in evolution of Fas. The features of amino acids used in the site-directed mutagenesis are summarized in S3 Fig.
We observed that introducing a negative charge by Y232D mutation rescued cells from FasL-induced cell death (Fig 2A and 2B). However, side chain size reduction by Y232C mutation also, to a lesser extent, rescued the cells, suggesting the importance of side chain size and aromatic ring of Y232 in Fas signaling. This is in accord with the high conservation of aromatic amino acids at this position in vertebrates (Fig 1A). Since the complete rescue observed in Y232D-carrying cells could be a combined effect of the added negative charge and reduced side chain size, substituting a small acidic amino acid for pY (as a single measure) may not provide an adequate proxy for functional studies of phospho-Y232 (pY232).
However, the situation differed for the 291 position where, similar to Y291F, size-reduction substitutions (Y291A and Y291C) had no impact on the apoptotic function (Fig 2C and 2D) and the formation of the DISC of Fas (Fig 2E). In contrast, the negative-charge substitution, Y291D, completely abolished FasL-induced apoptosis and DISC formation (Fig 2C and 2E). This indicates that the abolition of Fas apoptotic signaling was due to the negative charge of aspartic acid but not to its small size. Of note is that while Y291A mutation did not impact the apoptotic function of Fas, we observed an increase in cleaved fragments of caspase 8 in the DISC from stimulated Y291A cell lysate. This phenomenon did not cause any spontaneous cell death (Fig 2C) and could be related to non-apoptotic function of caspase 8, as it is now well-known that the death-effector domains (DEDs) containing proteins, including caspases, not only regulate apoptosis but also other forms of cell death, including necroptosis, as well as other important cellular processes such as autophagy and inflammation (see review [18]). Further studies are required to explore the roles and effects of basal activation of caspase 8 in Y291A-containing cells. Overall, these results demonstrate that pY291 and the aromatic side chain at this position are dispensable for cell death signaling and that the net charge at this site is more important for the protein's function than the detail of the side chain structure.
Death Domain Phosphorylation Confers Inter- and Intramolecular-Dominant Anti-Apoptotic Activity to Fas
Having established that Y291D mutation emulated the negative charge of pY291 independently of the side chain size reduction and loss of aromatic ring, we used the mutant as a proxy to examine how adding negative charge to this site by phosphorylation could modulate Fas signaling.
To further demonstrate the anti-apoptotic effect exerted by the negatively charged 291 residue in the Fas signaling, we introduced Y291D mutant Fas (death-off) into cells that stably expressed wild-type, Y232F, or Y291F mutant Fas (death-on) whose apoptotic capacity had been established (Fig 2). We found that the expression of Y291D Fas exhibited a clear dominant-negative effect on all “death-on” Fas species investigated, reducing the level of dead cells by approximately 50% (Fig 3A). In line with these data, when we introduced the “death-on” Fas species to cells stably expressing Y291D, a clear reduction in apoptosis-inducing capacity of the “death-on” Fas was observed when compared to cells expressing the control vector (Fig 3B). The data suggest that a subset of Fas that is phosphorylated can extend its inhibitory effect to an unphosphorylated Fas population, rendering it inefficient in inducing apoptosis.
Since both Y232 and Y291 can be phosphorylated, we examined whether phosphorylation of both tyrosines is required to turn off the death signal. Double mutations of Y232 and Y291 in SW480 and SW620 cells (Fig 3C and 3D) showed that negative-charge mutation at the 291 position alone was sufficient to completely block Fas-mediated cell death. The cell death inhibition in Y232F/Y291D mutant cells shows that maintaining unphosphorylated Y232 could not override the cell death blockage by Y291D mutation. This implies that FasL-induced apoptosis is rendered possible only when Y291 is dephosphorylated and that pY291 exerts dominant-negative effect on this apoptotic process.
While it was evident that dephosphorylation at both Y232 and Y291 (Y232F/Y291F) allowed efficient FasL-induced apoptosis (Fig 3C and 3D), it was unclear whether double tyrosine dephosphorylation was essential for this signal or if single dephosphorylation at Y291 sufficed for the apoptotic signal to proceed. While Y232D mutation partially presented the anti-apoptotic effect of side chain size reduction, the additional anti-apoptotic effect of negative charge at 232 site could still be observed (Fig 2A and 2B). Thus, the observation that FasL-induced cell death in cells carrying Y232D/Y291F mutants remained completely blocked (Fig 3C and 3D) points toward the likelihood that pY232 also exerts dominant-negative effect on FasL-induced apoptosis. That the single dephosphorylation at 291 residue could not override the anti-apoptotic effect of negative-charge addition at 232 residue suggests that double dephosphorylation at both Y232 and Y291 is required for FasL-induced apoptosis. Fig 3E depicts some dominant-negative scenarios in which: 1. an intramolecular dominant-negative effect on apoptosis occurs in a Fas molecule when at least one of the death domain tyrosines is phosphorylated; and 2. an intermolecular dominant-negative effect occurs when Fas molecules carrying at least one death domain pY dominant-negatively suppress the apoptotic function of Fas molecules in the pro-apoptotic state (i.e., having both death domain tyrosines dephosphorylated).
Y291 Phosphorylation Primes Clathrin-Mediated Endocytosis of Fas
The 291YDTL motif of Fas has been suggested as a putative tyrosine (Y)-based sorting motif (Yxxϕ; ϕ, a bulky amino acid; x, any amino acid) for clathrin-dependent endocytosis (CDE) [19]. The Y in the motif is essential for binding to μ2 subunit of AP-2 and, in most cases, cannot be substituted by other aromatic acid residues or pY (review [20]). Our data demonstrated that neither Y291D nor Y291F mutation impaired FasL internalization (Fig 4A). Moreover, Y291D Fas expression resulted in a more efficient FasL uptake, indicating that the added negative charge favored the process and, thus, raising doubt regarding the function of Y291 as a critical component of a Y-based sorting motif for CDE. To address these issues, we investigated the involvement of CDE in FasL/Fas uptake. Our synchronized internalization study by immunofluorescence confirmed that neither Y291D nor Y291F inhibited FasL uptake upon its engagement and that the uptake was more efficient with Y291D mutant (Fig 4B). The rapid FasL uptake (within 10 minutes of activation) was accompanied by the transport of a population of Fas to the perinuclear region (Fig 4B). In cells carrying wild type and Y291F Fas, the inhibition of CDE by overexpression of a truncated form of AP180 protein (AP180-C), which blocks the recruitment of clathrin to the plasma membrane [21], caused only a small delay in FasL uptake, which was completed by 30 min of activation. This indicated that the inhibition of Fas/FasL uptake by CDE could be compensated by an alternative pathway, as we previously reported [22]. In contrast, AP180-C expression in cells expressing Y291D Fas strongly inhibited the FasL/Fas complex uptake along with the transport of Y291D Fas to the perinuclear region. Similarly, disrupting dynamin-dependent endocytosis by dynasore, a potent inhibitor to dynamin GTPase activity, led to a strong reduction of FasL/Fas uptake in cells carrying Y291D Fas but not in those carrying wild-type or Y291F Fas (Fig 4C). These data demonstrate that FasL uptake and perinuclear transport of Y291D Fas relied on dynamin-dependent CDE. Unlike in the case of wild-type and Y291F Fas, in which dynamin-independent, clathrin-independent endocytosis (CIE) could compensate for the CDE blockage, the negatively charged Y291D committed the internalization of Fas to CDE. This implies that the constitutive pY291 engages Fas trafficking to CDE exclusively, hence the loss of the flexibility to carryout FasL-activated trafficking via compensatory CIE processes.
Our observation that Y291D even promoted CDE in these cells suggests that Y291 may participate in CDE as a part of another sorting motif. We analyzed amino acids flanking Y291 and found the similarity between the acidic dileucine (LL) sorting motif (D/E)xxxL(L/I) and Fas sequence 289EAYDTLI295. It is possible that the negative charge of Y291D mutation promoted the interaction between Fas and CDE adaptor proteins that binds the LL motif, such as AP2. Coimmunoprecipitation showed that overexpression of Y291D Fas, but not wild-type or Y291F Fas, increased the association of Fas with AP2 (Fig 4D), suggesting that the negative charge of pY291 may promote the sorting function of the LL motif. This is in line with the importance of phosphorylation in the LL motif in the internalization process, which has been previously reported [23–27]. That inhibiting dynamin-dependent CDE with dynasore sensitized cells to FasL-induced cell death (Fig 4E) also suggests that the function of pY291 in dynamin-dependent CDE contributed to the pro-survival signal of Fas.
Death Domain Tyrosine Phosphorylation Is Vital to FasL-Induced Non-Death Signaling of Fas
The contribution of Fas signaling to colorectal cell proliferation was demonstrated as Fas knockdown by siRNA led to a decrease in BrdU incorporation (Fig 5A) and sublethal doses of FasL increased viability of the cells (S6A Fig). To determine the role of Y232 and Y291 phosphorylation in FasL-induced proliferation by site-directed mutagenesis while minimizing the interference from endogenous Fas in SW480 cells, we used stable SW480 cell lines overexpressing Fas proteins that carried silent mutations in the region targeted by an siRNA against Fas. This was to allow the reduction of background signals from endogenous Fas while maintaining that of overexpressed Fas. In cells treated by control siRNA, the FasL-induced proliferation depended on Fas pY, since abolition of tyrosine phosphorylation by the expression of Y232F and Y291F Fas reduced BrdU incorporation, while mimicking the negative charge of pY by Y291D Fas expression did not (Fig 5B). The specific effects of Fas pY mutations in proliferation were confirmed in cells treated with Fas siRNA to reduce background proliferative signals from endogenous Fas. Following Fas siRNA treatment, the FasL-induced proliferation was reduced in control cells, while it increased significantly in cells carrying Y291D mutation, suggesting an active role of pY291 in a proliferative signal of Fas. As found for cells not treated with Fas siRNA, FasL-induced proliferation decreased in cell carrying Y232F and Y291F mutation that were subjected to Fas siRNA treatment, demonstrating a strong inhibitory role of Fas Y232 and Y291 dephosphorylation in FasL-induced proliferation. The importance of pY291 in FasL-induced proliferation was also supported by our observation that the expression of Y291D Fas led to an increase in viability when cells were treated with sublethal doses of FasL, while the expression of Y291F produced the opposite effect (S6B and S6C Fig).
In addition to promoting FasL-induced proliferation, phosphorylation of death domain tyrosine also promoted FasL-induced cell migration. Using Boyden chamber migration assay, we found that cells that overexpressed Y291D Fas mutant exhibited increased migration ability induced by FasL when compared to control cells and cells that expressed unphosphorylable mutants (Fig 5C).
Phosphorylation of Y232 and Y291 Is Regulated by Src Family Kinases (SFKs) and Protein Tyrosine Phosphatase SHP-1
Based on previously suggested involvement of SFKs in Fas signaling [11,28–29], we examined their influence on FasL-induced cell death. Inhibiting SFKs by a Src family kinase inhibitor, PP2, sensitized cells to FasL-induced cell death (Fig 6A). On the other hand, PP2 also significantly reduces FasL-induced proliferation (Fig 6B). The potentiation of FasL-induced cell death and the inhibition of FasL-induced proliferation by PP2 implied the role of SFKs in the phosphorylation of Fas death domain tyrosines. To further identify the SFKs involved, we subjected the cells to siRNA against Src and Yes-1 and found that suppressing either Src or Yes-1 could somewhat reduce the levels of pY232 and pY291 Fas (Fig 6C). However, the effect was more pronounced when both Src and Yes-1 were simultaneously suppressed. This reflects the redundancy of SFK activities in the Fas tyrosine phosphorylation process, since suppression of Src or Yes-1 alone was not as efficient as suppressing both kinases in reducing pY232 and pY291 Fas. The functional redundancy among SFKs is well recognized [30], and this notion is also supported by our observation that overexpression of either Src or Yes-1 could increase the level of pY232 and pY291 Fas (S12 Fig).
The protein tyrosine phosphatase, SHP-1, has been implicated in Fas signaling [31]. Therefore, we investigated its involvement in Fas tyrosine phosphorylation process. We found that inhibiting SHP-1 activity by protein tyrosine phosphatase inhibitor I (PTPiI) protected the cells from FasL-induced cell death (Fig 6D) while promoting FasL-induced proliferation (Fig 6E). This implied that SHP-1 might negatively regulate the phosphorylation of Fas death domain tyrosines. We further confirmed the role of SHP-1 in Fas dephosphorylation by demonstrating that overexpressing SHP-1 protein decreased pY232 and pY291 levels (Fig 6F) while suppressing SHP-1 expression by siRNA resulted in the opposite effect (Fig 6G).
Fas Tyrosine Phosphorylation Profiles Can Correlate to Different Contexts of Human Cancers
By comparing several colon cell lines, we found that the relative Fas pY levels tended to increase with the cancer progression (Fig 7A and 7B), implying that Fas pY might correlate to some contexts of human cancers. We therefore examined the relative levels of pY232 and pY291 in malignant tissues when compared to corresponding normal tissues of patients diagnosed with different types of cancers. We found that most patients having cancer of the colon, breast, or ovary also had an increased level of pY232 and/or pY291. However, this was not the case in patients having cancer of the cervix or lungs (Fig 7C). These diverse Fas pY profiles in different cancer types suggests that Fas signaling modes may be cancer type-dependent. Additional evidence supporting the involvement of Fas pY in human cancer comes from our observation that pY291 Fas levels decreased while pY232 Fas levels increased in the majority of rectal tumors after radiotherapy (± concurrent chemotherapy, Fig 7D), suggesting distinct regulation and functions of pY232 and pY291 in Fas signaling in rectal cancer in response to cancer therapy.
Discussion
Fas Death Domain Tyrosine Phosphorylation and Outcome of Fas Signaling
The functions of different phospho-tyrosines of Fas have not been distinguished to date. Using comparative genomics to guide functional analysis of Fas pY, in conjunction with site-specific detection of the phosphorylated death domain tyrosines, Y232 and Y291, we show that the phosphorylation of both death domain tyrosines in human Fas is dispensable for FasL-induced apoptosis. Our findings in human colorectal cells and B-cells that demonstrate this claim are well supported by evolution data and functional data from other animal cell models, including macaques [10] and rats [8,11]. Guided by comparative genomics, which reveals an unconventional cysteine substitution for Y291 in primates and rodents, we show that the net charge at this site is more important for the protein's function than the size or details of the amino acid side chain. This has created the possibility of using acidic amino acid substitution as a proxy for pY291 and of demonstrating that pY291 is, rather, a pro-survival mechanism that confers apoptosis resistance and proliferative advantage while its dephosphorylation permits apoptotic process.
The substitution of C for Y at 291 residue in old world monkeys is unique among the three Fas tyrosines found in primates. It suggests a low structural requirement from this tyrosine and that its role is, rather, in functional specificity. This amino acid exchange in Fas orthologs in closely related species may appear surprising and drastic considering amino acid sizes and properties. However, it serves as an example that, for certain protein functions, Y can be exchanged for a small, non-aromatic amino acid and that such Y (or pY) may have evolved from a smaller amino acid, as previously demonstrated [12]. Our finding that small amino acid substitutions occur at pY sites of Caspase 8 and are common in FAP-1 (S7 and S8 Figs) supports this point of view. The shift from small amino acids to (p)Y of Fas and other proteins in the Fas signaling pathway in primates implies a preferential shift to pY switch systems that can confer their functional plasticity and specificity in these species.
Functions of Fas Death Domain Tyrosines
We show that pY232 and pY291 of Fas have common features. They (1) are dispensable for Fas-induced apoptosis and (2) dominant-negatively inhibit apoptosis. Using the proxy Y291D, we also show that pY291 can promote FasL-induced cell proliferation and, thus, present pY291 as a reversible anti-apoptotic/pro-survival switch of Fas. This switch mechanism involves the function of pY291 in preventing FasL-induced DISC formation (Fig 2E) and promoting CDE of Fas (Fig 4).
Previous work showed that the Y291F mutation of human Fas (hFas) in murine cells inhibited the downregulation of the hFas-antibody complex [19]. Since the 291YDTL motif is consistent with the Y-based sorting motif for CDE, one may infer that this decrease in the antibody-induced Fas downregulation was caused by Y → F substitution in the motif. However, we found that, in human cells, Y291F mutation did not inhibit the FasL uptake (Fig 4). Likewise, Y283F mutation of murine Fas (corresponding to Y291F in hFas) in murine T cells did not inhibit the downregulation of the surface Fas-FasL complex (S9 Fig). Additionally, using the Y291D mutation, we further provide evidence suggesting that pY291 could enhance the uptake of the receptor via AP2-mediated CDE, which was important to its anti-apoptotic role. The fact that substitution of Y by other amino acids allowed Fas/FasL uptake by CDE suggests that the FasL-induced Fas uptake did not depend on Y291 as a part of the Y-based sorting motif but possibly of other motifs such as the acidic LL motif. Our finding is in line with previous reports for Vpu protein from HIV-1 subtype C [32], in which the tyrosine was dispensable for the protein's cell surface transport but important for viral replication, while the LL motif was crucial for cell surface transport.
Additionally, we offer an insight into distinct functions of Y232 and Y291 and their respective phosphorylation, which has not been addressed thus far: (1) the aromatic side chain of Y232 may contribute more to the structural integrity of the protein than that of Y291 (Fig 1A); (2) the size and details of the hydroxyl aromatic side chain of Y232 are essential for the functions of Fas, while the charge of Y291 is more important than the size and details of the aromatic side chain (Fig 2); and (3) pY232 and pY291 are distinctly regulated in different types of cancer (see below).
Regulation of Death Domain Tyrosine Phosphorylation of Fas by SFKs and SHP-1
SFKs are important mediators of tumor cell proliferation and survival and are involved in Fas signaling (Fig 6A, [33–34]). Yet, how activities of SFKs exert an effect on Fas has been unclear thus far. We reveal an intricate regulation of Fas death domain phosphorylation by SFKs, leading to the inhibition of the apoptotic signal of Fas, and, thus, provide the first mechanistic link between SFKs, major drivers of tumor development and progression, and the control of Fas multisignaling. This is of clinical significance because, in tumors from glioblastoma multiforme patients, the expression of Yes and phosphorylation of SFKs, as well as an enhanced FasL expression, were observed in the zone of tumor–host interaction, suggesting their roles in glioma invasion [34]. This is in concert with the concept presented herein that, in cancer, Fas-mediated survival signaling is promoted by SFK-dependent tyrosine phosphorylation.
Concerning the Fas pY dephosphorylation, it has been proposed that SHP-1 might be involved in this process in neutrophil, since it is associated with wild-type human Fas but not Y291A mutant in mouse lymphoblastic cells, and human SHP-1 from several cell lines could associate with phosphorylated peptides corresponding to the YxxL motif of death receptors [7]. We show that, in colorectal cells, pY232 and pY291 dephosphorylation is mediated by SHP-1, which has been shown to effectively dephosphorylate Src substrates [35] and to negatively regulate colonic cells proliferation [36]. Our data support the notion that the pY-based proliferative/apoptotic switch system of Fas is regulated via phosphorylation by SFKs and dephosphorylation by SHP-1, similar to that of caspase-8 [37–38]. This emphasizes the importance of the SFKs/SHP-1-based phosphorylation/dephosphorylation mechanism in the Fas multisignaling pathway.
Regulation and Functions of Death Domain Tyrosines and Their Implications in Cancer
Multimodal signaling of Fas has been demonstrated in many cancer cell types, including colon [39], breast [40], and glioblastoma [34]. Currently, both pro-apoptotic and pro-survival roles of Fas are bases of therapeutics that aim either to activate Fas signaling (APO010 agent targeting extracellular domain of Fas) [41] or to inhibit Fas signaling triggered by FasL (APG101 targeting FasL) [42]. These approaches face a major challenge, since it has been unclear what determines the outcome of Fas signaling and when one role of Fas will dominate the other. Our observation that colon, breast, and ovarian malignant tissues from most patients we tested had higher levels of pY232 and/or pY291 than their corresponding normal tissues suggests the probability that the pro-survival signal of Fas may dominantly operate in these cancers. On the other hand, the opposite observation for tumors from lung cancer patients suggests the probability that Fas may not contribute to the pro-survival signal in lung cancer (Fig 7C). Furthermore, we provide evidence that pY232 and pY291 can be distinctly regulated in cancer by showing opposing changes in their levels in rectal tumors in response to radiotherapy (RT ± chemotherapy). The reduction of pY291 following the treatment may suggest a decrease in the FasL-induced pro-proliferative/anti-apoptotic signal of Fas conferred by pY291, whereas the increase in pY232 may suggest an increase in other signaling events that involve the function of pY232. This may include the involvement of pY232 in the cell cycle phase, since we also observed a Fas-dependent G2/M accumulation that depended on the phenolic hydroxyl group of Y232 (Figs 2A and S10). This G2/M accumulation was associated neither with the resistance to FasL-induced apoptosis (Fig 2) nor an increase in FasL-induced cell proliferation (Fig 5B), suggesting an additional role of pY232 that is independent of FasL-induced apoptotic and proliferative functions of the protein.
Data presented here were obtained from a small number of patients. Thus, generalizations about the various extents of Fas phosphorylation in different types of cancer should be made with caution. However, our data revealing that the outcome of Fas signaling is determined by its pY status of the death domain and that the Fas pY status may differ among different cancer types and may respond to anticancer treatment provide a basis for further studies in larger sets of human cancer samples, as well as an opportunity to develop a practical means to predict the outcome of Fas signaling in different pathologies that can lead to the use of Fas pY screening to aid Fas-related therapeutic design and maximize the chance of therapeutic success.
Overall, we provide the delineation of the pY-based control of Fas signaling, revealing differential evolutional criteria of the two death domain tyrosines, their regulatory elements, mechanistic links between this molecular switch system and the cellular outcome, and the implications in cancer. This information has far-reaching consequences, not only in cancer contexts but also in other pathologies in which Fas signaling is involved.
Materials and Methods
Ethic Statement
The rectal tumors were obtained from patients providing informed consent under protocols approved by the Clinic Institutional Review Board of the Département Recherche Clinique Innovation et Statistiques (DRIS)–Centre Antoine LACASSAGNE, Nice, France.
Materials and additional detailed methods are listed in S1 Text.
Flow Cytometry Analysis of Unsynchronized Fas/FasL Complex Internalization
SW480 cells (2.5 x 105 cells/well) were seeded in a 24-well plate for 24 h. The medium was then replaced with fresh RPMI+0.1% BSA containing 1 μg/ml mouse anti-Flag (M2)+ 1 μg/ml donkey anti-mouse Alexa Fluor 647 with or without 100 ng/ml FasL. The cells were then incubated at 37°C for a specified time to allow FasL-triggered stimulation. Thereafter, the plate was transferred to an iced water basin and the activation was stopped by adding ice-cold PBS. FasL that remained on the cell surface was removed using an ice-cold acid-stripping buffer (50 mM glycine, 100 mM NaCl, pH3). After washing with ice-cold PBS, cells were analyzed for the internalized FasL (crosslinked with mouse anti-Flag antibody and Alexa Fluor 647 anti-mouse antibody) by flow cytometer (LSRFortessa, Becton Dickinson). To assess the degree of FasL internalization, the median fluorescence intensity of the background control cells (treated with anti-Flag and Alexa Fluor 647 without FasL) was subtracted from the median fluorescence intensity of the FasL-treated cells to obtain the absolute fluorescence intensity of the detected internalized FasL in the cells.
Synchronized Fas/FasL Complex Internalization Analysis by Immunofluorescence Microscopy
SW480 Cells (2.5 x 105 cells/well) were seeded in RPMI+10% FBS on coverslip in 24-well plate for 24 h. Cells were then cooled down to 0°C in a refrigerating chamber for 45 min in RPMI+10% FBS+10 mM Hepes. The medium was then replaced with ice-cold RPMI+0.1% BSA+10 mM Hepes (internalization medium) containing 200 ng/ml FasL+1μg/ml mouse anti-FLAG (M2)+1 μg/ml Alexa Fluor-647 anti-mouse antibody, and cells were incubated at 0°C in a refrigerating chamber for 1 h to allow FasL binding. The medium containing excess FasL was then removed, and internalization medium at 0°C was added to the well. The coverslips containing cells for control condition were kept at 0°C in the refrigerating chamber throughout the experiment, while coverslips containing cells destined for activation were rapidly transferred to another 24-well plate containing 0.25 ml of internalization medium per well (maintained at 37°C in an incubator) to trigger the internalization of Fas/FasL complexes. After incubation at 37°C for indicated times, the plate was rapidly transferred from the incubator to the refrigerating chamber, and 1 ml of PBS at 0°C was added to each well to stop the activation. Cells were then rapidly fixed with ice-cold 4% paraformaldehyde for 20 min on ice and counter-stained with DAPI for nuclear detection before mounting on a slide in the presence of mounting medium (Fluoromount, Sigma-Aldrich). Fluorescence images were taken using a spinning disk confocal microscope (Olympus/Andor/Yokogawa system) with a 100×oil/1.4 numerical aperture objective lens. Images were deconvolved with Huygens software (Scientific Volume Imaging).
Mobility Shift Detection of Phospho-Proteins
Phosphate affinity SDS-PAGE was performed using 7.5% polyacrylamide gels containing 10 μM acrylamide-pendant Phos-Tag (Wako), according to the manufacturer's instructions. Highly phosphorylated proteins migrate more slowly through the gel than less phosphorylated proteins, allowing protein separation based on phosphorylation states.
Supporting Information
Acknowledgments
We thank the technical platforms for imaging (PRISM), cytometry, and biochemistry (IBV- CNRS UMR 7277- INSERM U1091-UNS) for their technical assistance and services.
Abbreviations
- ALPS
autoimmune lymphoproliferative syndrome
- CDE
clathrin-dependent endocytosis
- CIE
clathrin-independent endocytosis
- CIP
calf intestinal phosphatase
- DED
death-effector domain
- DISC
death-inducing signalling complex
- FasL
Fas ligand
- hFas
human Fas
- FADD
Fas-associated protein with death domain
- LL
dileucine
- PI
propidium iodine
- PTPiI
protein tyrosine phosphatase inhibitor I
- pY
phospho-tyrosine
- pY232
phospho-Y232
- RT
radiotherapy
- sFasL
soluble FasL
- SFK
Src family kinase
- TNF
tumor necrosis factor
Data Availability
All relevant data are within the paper and its Supporting Information files.
Funding Statement
This work was supported by institutional funds from the Centre National de la Recherche Scientifique (CNRS) and the Institut National de la Santé et de la Recherche Médicale (INSERM), and by grants from the Institut National du Cancer (INCa; PLBIO09-317); the University of Nice, the Agence Nationale de la Recherche (ANR-10-BLAN-1226; ANR-11-LABX-0028-01). KL was supported by funding from the Canceropole PACA, and LTN was supported by a scholarship from the ministry and education of training of the socialist republic of Vietnam. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1. Kischkel FC, Hellbardt S, Behrmann I, Germer M, Pawlita M, Krammer PH, et al. Cytotoxicity-dependent APO-1 (Fas/CD95)-associated proteins form a death-inducing signaling complex (DISC) with the receptor. Embo J. 1995;14(22):5579–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Peter ME, Budd RC, Desbarats J, Hedrick SM, Hueber AO, Newell MK, et al. The CD95 receptor: apoptosis revisited. Cell. 2007;129(3):447–50. 10.1016/j.cell.2007.04.031 . [DOI] [PubMed] [Google Scholar]
- 3. Hoogwater FJ, Steller EJ, Westendorp BF, Borel Rinkes IH, Kranenburg O. CD95 signaling in colorectal cancer. Biochim Biophys Acta. 2012;1826(1):189–98. 10.1016/j.bbcan.2012.03.007 . [DOI] [PubMed] [Google Scholar]
- 4. Zheng HX, Cai YD, Wang YD, Cui XB, Xie TT, Li WJ, et al. Fas signaling promotes motility and metastasis through epithelial-mesenchymal transition in gastrointestinal cancer. Oncogene. 2013;32(9):1183–92. 10.1038/onc.2012.126 . [DOI] [PubMed] [Google Scholar]
- 5. Chen L, Park SM, Tumanov AV, Hau A, Sawada K, Feig C, et al. CD95 promotes tumour growth. Nature. 2010;465(7297):492–6. 10.1038/nature09075 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Gradl G, Grandison P, Lindridge E, Wang Y, Watson J, Rudert F. The CD95 (Fas/APO-1) receptor is phosphorylated in vitro and in vivo and constitutively associates with several cellular proteins. Apoptosis. 1996;1:131–40. [Google Scholar]
- 7. Daigle I, Yousefi S, Colonna M, Green DR, Simon HU. Death receptors bind SHP-1 and block cytokine-induced anti-apoptotic signaling in neutrophils. Nat Med. 2002;8(1):61–7. 10.1038/nm0102-61 . [DOI] [PubMed] [Google Scholar]
- 8. Reinehr R, Schliess F, Häussinger D. Hyperosmolarity and CD95L trigger CD95/EGF receptor association and tyrosine phosphorylation of CD95 as prerequisites for CD95 membrane trafficking and DISC formation. FASEB J. 2003;17(6):731–3. 10.1096/fj.02-0915fje . [DOI] [PubMed] [Google Scholar]
- 9. Wang L, Yang JK, Kabaleeswaran V, Rice AJ, Cruz AC, Park AY, et al. The Fas-FADD death domain complex structure reveals the basis of DISC assembly and disease mutations. Nat Struct Mol Biol. 2010;17(11):1324–9. 10.1038/nsmb.1920 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Murayama Y, Terao K, Inoue-Murayama M. Molecular cloning and characterization of cynomolgus monkey Fas. Hum Immunol. 2000;61(5):474–85. . [DOI] [PubMed] [Google Scholar]
- 11. Reinehr R, Becker S, Höngen A, Haüssinger D. The Src family kinase Yes triggers hyperosmotic activation of the epidermal growth factor receptor and CD95. J Biol Chem. 2004;279(23):23977–87. 10.1074/jbc.M401519200 . [DOI] [PubMed] [Google Scholar]
- 12. Pearlman SM, Serber Z, Ferrell JE. A mechanism for the evolution of phosphorylation sites. Cell. 2011;147(4):934–46. 10.1016/j.cell.2011.08.052 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Cuevas BD, Lu Y, Mao M, Zhang J, LaPushin R, Siminovitch K, et al. Tyrosine phosphorylation of p85 relieves its inhibitory activity on phosphatidylinositol 3-kinase. J Biol Chem. 2001;276(29):27455–61. 10.1074/jbc.M100556200 . [DOI] [PubMed] [Google Scholar]
- 14. Trakul N, Menard RE, Schade GR, Qian Z, Rosner MR. Raf kinase inhibitory protein regulates Raf-1 but not B-Raf kinase activation. J Biol Chem. 2005;280(26):24931–40. 10.1074/jbc.M413929200 . [DOI] [PubMed] [Google Scholar]
- 15. Wu Y, Spencer SD, Lasky LA. Tyrosine phosphorylation regulates the SH3-mediated binding of the Wiskott-Aldrich syndrome protein to PSTPIP, a cytoskeletal-associated protein. J Biol Chem. 1998;273(10):5765–70. . [DOI] [PubMed] [Google Scholar]
- 16. Zhang B, Tavaré JM, Ellis L, Roth RA. The regulatory role of known tyrosine autophosphorylation sites of the insulin receptor kinase domain. An assessment by replacement with neutral and negatively charged amino acids. J Biol Chem. 1991;266(2):990–6. . [PubMed] [Google Scholar]
- 17. Bettinardi A, Brugnoni D, Quiròs-Roldan E, Malagoli A, La Grutta S, Correra A, et al. Missense mutations in the Fas gene resulting in autoimmune lymphoproliferative syndrome: a molecular and immunological analysis. Blood. 1997;89(3):902–9. . [PubMed] [Google Scholar]
- 18. Riley JS, Malik A, Holohan C, Longley DB. DED or alive: assembly and regulation of the death effector domain complexes. Cell Death Dis. 2015;6:e1866 10.1038/cddis.2015.213 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Lee KH, Feig C, Tchikov V, Schickel R, Hallas C, Schütze S, et al. The role of receptor internalization in CD95 signaling. EMBO J. 2006;25(5):1009–23. 10.1038/sj.emboj.7601016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Bonifacino JS, Dell'Angelica EC. Molecular bases for the recognition of tyrosine-based sorting signals. J Cell Biol. 1999;145(5):923–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Ford MG, Pearse BM, Higgins MK, Vallis Y, Owen DJ, Gibson A, et al. Simultaneous binding of PtdIns(4,5)P2 and clathrin by AP180 in the nucleation of clathrin lattices on membranes. Science. 2001;291(5506):1051–5. . [DOI] [PubMed] [Google Scholar]
- 22. Chakrabandhu K, Huault S, Garmy N, Fantini J, Stebe E, Mailfert S, et al. The extracellular glycosphingolipid-binding motif of Fas defines its internalization route, mode and outcome of signals upon activation by ligand. Cell Death Differ. 2008;15(12):1824–37. 10.1038/cdd.2008.115 . [DOI] [PubMed] [Google Scholar]
- 23. Hinners I, Wendler F, Fei H, Thomas L, Thomas G, Tooze SA. AP-1 recruitment to VAMP4 is modulated by phosphorylation-dependent binding of PACS-1. EMBO Rep. 2003;4(12):1182–9. 10.1038/sj.embor.7400018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kato Y, Misra S, Puertollano R, Hurley JH, Bonifacino JS. Phosphoregulation of sorting signal-VHS domain interactions by a direct electrostatic mechanism. Nat Struct Biol. 2002;9(7):532–6. 10.1038/nsb807 . [DOI] [PubMed] [Google Scholar]
- 25. Krantz DE, Waites C, Oorschot V, Liu Y, Wilson RI, Tan PK, et al. A phosphorylation site regulates sorting of the vesicular acetylcholine transporter to dense core vesicles. J Cell Biol. 2000;149(2):379–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Lukashova V, Szabó EZ, Jinadasa T, Mokhov A, Litchfield DW, Orlowski J. CK2 phosphorylation of an acidic Ser/Thr di-isoleucine motif in the Na+/H+ exchanger NHE5 isoform promotes association with beta-arrestin2 and endocytosis. J Biol Chem. 2011;286(13):11456–68. 10.1074/jbc.M110.182881 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Pitcher C, Höning S, Fingerhut A, Bowers K, Marsh M. Cluster of differentiation antigen 4 (CD4) endocytosis and adaptor complex binding require activation of the CD4 endocytosis signal by serine phosphorylation. Mol Biol Cell. 1999;10(3):677–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Reinehr R, Becker S, Wettstein M, Häussinger D. Involvement of the Src family kinase yes in bile salt-induced apoptosis. Gastroenterology. 2004;127(5):1540–57. . [DOI] [PubMed] [Google Scholar]
- 29. Reinehr R, Häussinger D. CD95 death receptor and epidermal growth factor receptor (EGFR) in liver cell apoptosis and regeneration. Arch Biochem Biophys. 2012;518(1):2–7. 10.1016/j.abb.2011.12.004 . [DOI] [PubMed] [Google Scholar]
- 30. Thomas SM, Brugge JS. Cellular functions regulated by Src family kinases. Annu Rev Cell Dev Biol. 1997;13:513–609. . [DOI] [PubMed] [Google Scholar]
- 31. Koncz G, Kerekes K, Chakrabandhu K, Hueber AO. Regulating Vav1 phosphorylation by the SHP-1 tyrosine phosphatase is a fine-tuning mechanism for the negative regulation of DISC formation and Fas-mediated cell death signaling. Cell Death Differ. 2008;15(3):494–503. 10.1038/sj.cdd.4402282 . [DOI] [PubMed] [Google Scholar]
- 32. Ruiz A, Hill MS, Schmitt K, Guatelli J, Stephens EB. Requirements of the membrane proximal tyrosine and dileucine-based sorting signals for efficient transport of the subtype C Vpu protein to the plasma membrane and in virus release. Virology. 2008;378(1):58–68. 10.1016/j.virol.2008.05.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Atkinson EA, Ostergaard H, Kane K, Pinkoski MJ, Caputo A, Olszowy MW, et al. A physical interaction between the cell death protein Fas and the tyrosine kinase p59fynT. J Biol Chem. 1996;271(11):5968–71. . [DOI] [PubMed] [Google Scholar]
- 34. Kleber S, Sancho-Martinez I, Wiestler B, Beisel A, Gieffers C, Hill O, et al. Yes and PI3K bind CD95 to signal invasion of glioblastoma. Cancer Cell. 2008;13(3):235–48. 10.1016/j.ccr.2008.02.003 . [DOI] [PubMed] [Google Scholar]
- 35. Frank C, Burkhardt C, Imhof D, Ringel J, Zschornig O, Wieligmann K, et al. Effective dephosphorylation of Src substrates by SHP-1. J Biol Chem. 2004;279(12):11375–83. . [DOI] [PubMed] [Google Scholar]
- 36. Duchesne C, Charland S, Asselin C, Nahmias C, Rivard N. Negative regulation of beta-catenin signaling by tyrosine phosphatase SHP-1 in intestinal epithelial cells. J Biol Chem. 2003;278(16):14274–83. . [DOI] [PubMed] [Google Scholar]
- 37. Cursi S, Rufini A, Stagni V, Condo I, Matafora V, Bachi A, et al. Src kinase phosphorylates Caspase-8 on Tyr380: a novel mechanism of apoptosis suppression. EMBO J. 2006;25(9):1895–905. . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Jia SH, Parodo J, Kapus A, Rotstein OD, Marshall JC. Dynamic regulation of neutrophil survival through tyrosine phosphorylation or dephosphorylation of caspase-8. J Biol Chem. 2008;283(9):5402–13. 10.1074/jbc.M706462200 . [DOI] [PubMed] [Google Scholar]
- 39. Li H, Fan X, Stoicov C, Liu JH, Zubair S, Tsai E, et al. Human and mouse colon cancer utilizes CD95 signaling for local growth and metastatic spread to liver. Gastroenterology. 2009;137(3):934–44, 44.e1–4. 10.1053/j.gastro.2009.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Malleter M, Tauzin S, Bessede A, Castellano R, Goubard A, Godey F, et al. CD95L cell surface cleavage triggers a prometastatic signaling pathway in triple-negative breast cancer. Cancer Res. 2013;73(22):6711–21. 10.1158/0008-5472.CAN-13-1794 [DOI] [PubMed] [Google Scholar]
- 41.A Phase I Dose Finding Study of APO010 in Patients With Solid Tumors (AP1001) [Internet]. National Institutes of Health. NCT00437736. http://clinicaltrials.gov/ct2/show/NCT00437736?term=apo010&rank=1.
- 42. Bendszus M, Debus J, Wick W ea. APG101_CD_002: A phase II, randomized, open-label, multicenter study of weekly APG101 plus reirradiation versus reirradiation in the treatment of patients with recurrent glioblastoma. J Clin Oncol 2012. p. 2012;30(Suppl. 15). Abstract 34. [Google Scholar]
- 43. Chakrabandhu K, Hérincs Z, Huault S, Dost B, Peng L, Conchonaud F, et al. Palmitoylation is required for efficient Fas cell death signaling. EMBO J. 2007;26(1):209–20. 10.1038/sj.emboj.7601456 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All relevant data are within the paper and its Supporting Information files.