The Par-4-GRP78 TRAIL, more Twists and Turns

Abstract

GRP78, also referred to as BiP, is an essential molecular chaperone and a master regulator of the unfolded protein response. Traditionally, GRP78 is regarded as localized in the lumen of the endoplasmic reticulum (ER). However, recent findings revealed that a subfraction of GRP78 can localize to the surface of specific cell types. Furthermore, preferential expression of GRP78 on the surface of tumor cells but not in normal organs suggests that surface GRP78 can serve both as a target as well as mediator for cancer-specific therapy. Recent reports further established that GRP78 forms complexes with specific proteins on the cell surface and plays an important role in signaling, impacting cell survival and proliferation. Burikhanov et al. (Cell 2009, 138:377-88) reported that Par-4, generally regarded as a cytosolic and nuclear protein that promotes cell death via the mitochondrial cell death pathway, is spontaneously secreted by normal and cancer cells and this process is enhanced by ER stress or with addition of TRAIL. It is proposed that ER stress, induced by extracellular insults such as TRAIL, causes translocation of the Par-4-GRP78 complex from the ER to the plasma membrane, and through a positive feedback loop, extracellular Par-4 binds to cell surface GRP78 and activates the extrinsic apoptotic pathway. In this journal club, we discuss some open questions and how these new findings integrate with current understanding of GRP78 function in vivo.

DISCUSSION

A recent Cell paper by Burikhanov et al.¹ described an unanticipated observation that the tumor suppressor Par-4, previously known to act as a cytosolic and nuclear protein that promotes cell death via the intrinsic cell death pathway, is secreted outside the cell, and this is enhanced shortly upon exposure to endoplasmic reticulum (ER) stress-inducing agent or with addition of TRAIL. What started as a potentially promiscuous in vitro interaction between Par-4 and ER chaperone GRP78 leads to the unraveling of a new apoptotic pathway with the discovery that extracellular Par-4 binds to cell surface GRP78 and activates the extrinsic apoptotic pathway. Importantly, apoptosis inducible by TRAIL is dependent on extracellular Par-4 signaling via cell surface GRP78.

These novel observations also raise new questions. For example, the molecular basis responsible for Par-4 translocation into the ER lumen requires resolution. Since the Par-4 protein sequence does not contain any hydrophobic stretch for potential ER-targeting and translocation signal, is it possible that the highly charged residues of the SAC domain of Par-4 act as a cell penetrating peptide? It will be interesting to test directly whether this domain and Par-4 can functionally direct ER translocation. Is it possible that the Par-4 clone gets spliced? Another intriguing point is that surface expression of GRP78 is dependent on intracellular Par-4 and if it is pre-bound to surface GRP78, how does extracellular Par-4 bind to GRP78 to trigger apoptosis? Are different GRP78 binding sites involved?

Another scenario for the results reported is that Par-4 does not actually enter the ER but rather interacts with GRP78 through transmembrane ER protein complexes. This could account for the co-immunoprecipitation and co-staining data, as there are precedents that cytosolic proteins such as the pro-apoptotic BH3 protein BIK and caspase-7 that localize to the ER membrane form complex with GRP78 in these same assays.^2,3 Interestingly, Par-4 contains two putative N-linked glycosylation sites at amino acids 257 and 328. Thus, in principle, if Par-4 does enter the ER, these sites should be glycoslyated and upon tunicamycin treatment, the modification will be blocked. Nonetheless, in Burikhanov et al.,¹ the electrophoretic mobility of Par-4 does not appear to be affected by tunicamycin treatment. To resolve this, it will be informative to attach a genuine ER signal sequence to Par-4 and test whether these sites can be glycosylated as predicted.

The interaction between Par-4 and GRP78 highlights the emerging signaling network mediated by GRP78, which in addition to being a key regulator of the unfolded protein response as a major ER protein, may determine a wide range of anti- or pro-apoptotic activities as a cell surface bridge protein. The repertoire of partner proteins of GRP78 is likely to expand even more with the recent discovery of a cytosolic isoform of GRP78 (GRP78va), generated by ER-stress induced alternative splicing.⁴ Interestingly, GRP78va has the ability to modulate PERK signaling suggested by Burikhanov et al.¹ to play a role in caspase-8-dependent apoptosis by extracellular Par-4 and TRAIL.

These new findings also raise the important issue of whether surface GRP78-mediated cell apoptosis in vitro mimics therapeutic in vivo targeting. Studies on surface GRP78 expression performed in tissue culture cells show a great amount of variability, ranging from no surface expression⁵ to robust expression of GRP78 in prostate cancer PC-3 cells,¹ differential surface expression in a variety of cell lines⁶ to positive reactivity with GRP78 binding peptide motifs whether or not they are tumor cells.⁷ Nonetheless, normal organs appear to express few, if any, cell surface GRP78. Recently, it is reported that a human monoclonal IgM antibody (SAM-6) isolated from a gastric cancer patient binds to O-glycosylated GRP78 specifically expressed on the surface of malignant tissues.⁸ Considering that the levels of specific O-glycosylation enzymes strongly correlate with sensitivity to TRAIL, and O-glycosylation promotes TRAIL-stimulated clustering of its receptors and apoptosis,⁹ it will be interesting to examine whether O-glycosylation or other post-translational modifications of GRP78 influences its response to Par-4 and other binding partners.

A key question regarding the functional interaction between Par-4 and GRP78 is how it impacts cancer progression and acquired resistance. As pointed out by Hart and El-Deiry¹⁰ in the Cell Preview, given the requirement of GRP78 in embryonic survival, conditional deletion of GRP78 in somatic cells and tumors in vivo is needed to further establish its potential pro-apoptotic function. Indeed, conditional knockout of GRP78 has recently been achieved in two somatic cell types with distinct consequences. Whereas GRP78 is absolutely required for the survival of Purkinje cells which are specialized, terminally differentiated neuronal cells,¹¹ targeted knockout of GRP78 in mouse prostate epithelial cells does not affect prostate gland development.¹² In prostate cancer, GRP78 upregulation associates with development of castration-resistance, recurrence and worse overall survival. Using a novel mouse model in which both Grp78 and the tumor suppressor gene Pten undergo biallelic inactivation specifically in the postnatal mouse prostate epithelium, we discovered that Grp78 knockout potently suppresses prostate tumorigenesis.¹² In contrast to invasive carcinoma observed in the Pten null prostate, mouse prostates with simultaneous homozygous inactivation of Grp78 exhibit normal histology and cytology, and only partial intra-epithelial neoplasia for heterozygous Grp78 inactivation.¹²

Superficially, these results appear to contradict the pro-apoptotic function of cell surface GRP78 in the PC-3 human prostate epithelial cancer cells reported by Burikhanov et al.,¹ as loss of the extracellular Par-4 GRP78 apoptotic pathway is expected to lead to more aggressive tumor. Recent studies suggest that GRP78 could exert opposing regulatory function on intracellular versus extracellular Par-4 activity. It is well established that loss of PTEN function results in a constitutive activation of the PI3K/AKT pro-survival pathway. Corresponding with various reports using cultured cells that AKT activation is suppressed by downregulation of GRP78, AKT activation in Pten null prostate epithelium is inhibited by Grp78 homozygous deletion.¹² Since AKT activation leads to binding, phosphorylation and inactivation of the pro-apoptotic function of intracellular Par-4,¹³ inhibition of AKT activation by GRP78 knockdown could thus trigger intracellular Par-4-dependent apoptosis, leading to tumor suppression.

Another consideration is that the majority of cellular GRP78 resides in the ER lumen, with multiple protective functions including prevention of protein misfolding, maintenance of ER Ca²⁺ homeostasis, suppression of BIK and caspase-7 activation and mediation of stress-induced autophagy.^14,15 Interestingly, a recent report shows that GRP78 can be secreted by specific tumor cells, leading to induction of pro-survival and suppression of pro-apoptotic signaling in endothelial cells.¹⁶ Furthermore, cytosolic GRP78va and surface GRP78 have been shown to mediate pro-survival and pro-oncogenic functions. For example, the extracellular signaling protein Cripto forms complex with cell surface GRP78 and enhances tumor growth via inhibition of TGF-β signaling.¹⁷ GRP78 also associates with GPI-anchored T cadherin on the surface of human vascular endothelial cells and promotes cell survival.¹⁸ Interestingly, Cripto, T-cadherin and Par-4 all interact with the same N-terminal region of surface GRP78. Collectively, these exciting new developments point to the pivotal role of cell surface GRP78 in determining cell viability, with the outcome likely to be dependent on the type and concentration of partner proteins, cell or tumor types and genetic lesions. The story of extracellular Par-4 and surface GRP78 also supports the emerging notion of functional importance of minor subpopulations of various proteins generated by different means to influence fundamental biological processes.