GRPling with PTEN

Regulation of the Pten tumor suppressor is complex and mediated by varied mechanisms. In this issue of Blood, Wey and colleagues show in a biallelic conditional knockout mouse model of GRP78 and PTEN that heterozygous loss of Grp78 suppresses leukemogenesis mediated by Pten.¹  These findings suggest a novel manner of down-regulation of PI 3 kinase signaling that may have potential therapeutic benefit.

PTEN (Phosphatase and tensin homolog on chromosome 10) is a major human tumor suppressor genes that is frequently inactivated through mutation, deletion, or promoter methylation in tumors such as glioblastoma, endometrial, breast, thyroid, and prostate cancers (reviewed in Cully et al,²  and Martelli et al³ ). In addition, germ line mutations of PTEN lead to PTEN hamartoma tumor syndrome, a cancer predisposition condition. PTEN is a lipid phosphatase that has catalytic activity on the D-3 phosphate of the active lipid second messenger phosphatidylinositol 3,4,5-triphosphate (PIP₃).⁴ 

PTEN mRNA expression is regulated by promoter hypermethylation, a diverse range of transcriptional control and miRNA.³  All phosphatase enzymes possess a cysteine nucleophile at the catalytic site that is subject to oxidation. Serine and threonine phosphorylation of PTEN at the carboxy terminus locks PTEN in a stable conformation that reduces membrane localization and enzymatic activity. PTEN is also nuclear localized and the function and stability of PTEN in different, unique cellular environments has not been extensively investigated.

Gene-targeting experiments revealed that Pten^−/− mice had defective endodermal, ectodermal, and mesodermal differentiation,^5,6  suggesting that Pten is required for mouse embryogenesis. Heterozygous animals developed germ cell, gonadostromal, thyroid, and colon tumors. Murine embryo fibroblasts isolated from Pten^−/− mice showed enhanced Akt phosphorylation, suggesting that Pten is a negative regulator of PI 3 kinase signaling.⁷  Conditional inactivation of Pten in the hematopoietic lineage results in a myeloproliferative disease followed by onset of leukemia with a latency of 4 to 6 weeks.⁸ 

GRP78 (Glucose-regulated Protein of 78 kDa) is a member of the HSP70 (Heat-shock protein 70) gene family, and is thought to be an endoplasmic reticulum (ER) chaperone and a marker of ER stress (reviewed in Ni et al⁹ ). Recent studies suggest GRP78 may be found outside the ER, especially in transformed cancer cell lines. GRP78 can be found secreted, at the membrane, in the cytosol, within mitochondria or in the nucleus. α₂-macroglobulin is proposed to bind GRP78 and couple to PI 3 kinase pathway activation in tumor cells. Therefore, the possibility exists that GRP78 could regulate members of the PI 3 kinase pathway, including PTEN.

Wey et al initially tested this hypothesis by performing compound crosses using conditional alleles of Grp78 and Pten crossed with a probasin-Cre reporter.¹⁰  Loss of Grp78 did not affect the development of the prostatic epithelium. However, homozygous deletion of Grp78 in a Pten-deficient background had a profound effect on prostate adenocarcinoma development. Importantly, Pten^fl/flGrp78^fl/fl mice displayed absent phosphorylation of Akt whereas Pten^fl/flGrp78^+/+ mice had robust phospho-Akt staining in dorsolateral prostate sections.

Wey et al extended these studies. They showed that biallelic targeting of Pten^fl/fl and Grp78^fl/+ in an Mx1-Cre background resulted in an increase in disease latency from 4 weeks in Pten^fl/fl animals to 7 weeks in Pten^fl/flGrp78^fl/+ mice. Heterozygous loss of Grp78 suppressed blast cell formation in the bone marrow, whereas spleen weight and percent of Lin⁻Sca1⁺Kit⁺ cells were intermediate in Pten^fl/flGrp78^fl/+ comparing Pten^fl/fl and wild-type mice.

Phosphorylation of Akt and the downstream target of Akt signaling, the S6 kinase, were increased in Pten^fl/fl bone marrow cells. Loss of one copy of Grp78 resulted in suppression of Akt and S6 kinase phosphorylation in Pten^fl/flGrp78^fl/+ bone marrow. These data were complemented by siRNA experiments that showed knockdown of GRP78 results in decreased phosphorylation of AKT in the HL60 cell line. Wey and colleagues also performed in vitro studies demonstrating that overexpression of GRP78 in NB4 cells protects cells from AraC-induced death, whereas siRNA-targeting GRP78 results in potentiation of AraC-mediated apoptosis.

GRP78 is overexpressed in acute myeloid leukemia, chronic myeloid leukemia, acute lymphoid leukemia, and chronic lymphoid leukemia patient specimens examined in this study. Overexpression of GRP78 was confirmed in primary AML patients. However, the expression of other HSP70 family members was not examined in this paper and it is unclear whether the expression of other HSP70 family members is altered in human leukemia.

The conditional knockout data from Wey et al pose several interesting questions. Is the wide distribution of GRP78 outside of the ER in cancer cell lines a function of overexpression, increased ER stress, or an artifact of cellular transformation? Are these data specific to GRP78 or are there roles for other ER chaperones as has been suggested for HSP90-dependent regulation of BCR-ABL? Recent studies suggest a nuclear role for both PTEN and GRP78. Are these proteins colocalized in the nucleus and does this impact the observations presented in this study?

What is the role of GRP78 in oncology? GRP78 is a tumor antigen and highly expressed in metastatic pancreatic cancer.¹¹  GRP78 was successfully targeted in a ligand-directed therapeutic regimen using synthetic chimeric peptides expressing an apoptotic sequence in mouse models of prostate and breast cancer.¹²  A cyclic peptide targeting cell-surface GRP78 induced killing of melanoma cells in a GRP78-dependent manner.¹³  The successful targeting of GRP78 and the ability of GRP78 to affect PI 3 kinase pathway activation is worthy of additional studies to investigate the role of GRP78 in hematopoietic disease and beyond.