A Divergent Substrate-Binding Loop within the Pro-oncogenic Protein Anterior Gradient-2 Forms a Docking Site for Reptin

Abstract

Anterior gradient-2 (AGR2) functions in a range of biological systems, including goblet cell formation, limb regeneration, inhibition of p53, and metastasis. There are no well-validated binding proteins for AGR2 protein despite the wealth of data implicating an important cellular function in vertebrates. The yeast two-hybrid system was used to isolate the ATP binding protein Reptin as an AGR2-interacting protein. AGR2 formed a stable complex in human cell lysates with Reptin, thus validating Reptin as an AGR2 binding protein in cells. Reptin was also shown to be overproduced in a panel of primary breast cancer biopsy specimens, relative to normal adjacent tissue from the same patient, suggesting a role in cancer growth in vivo. Mutations were made at the two ATP binding motifs in Reptin to evaluate the effects of ATP on Reptin–AGR2 complex stability. Loss-of-ATP binding mutations at the Walker A motif (K83A) or gain-of-ATP binding mutations at the Walker B motif (D299N) resulted in Reptin mutants with altered oligomerization, thermostability, and AGR2 binding properties. These data indicate that the two ATP binding motifs of Reptin play a role in regulating the stability of the AGR2–Reptin complex. The minimal region of AGR2 interacting with Reptin was localized using overlapping peptide libraries derived from the AGR2 protein sequence. The Reptin docking site was mapped to a divergent octapeptide loop in the AGR2 superfamily between amino acids 104 and 111. Mutations at codon Y104 or F111 in full-length AGR2 destabilized the binding of Reptin. These data highlight the existence of a protein docking motif on AGR2 and an ATP-regulated peptide-binding activity for Reptin. This knowledge has implications for isolating other AGR2-interacting proteins, for developing assays to isolate small molecules that target the Reptin ATP binding site, and for measuring the effects of the Reptin–AGR2 complex in cancer cell growth.

Introduction

Anterior gradient-2 (AGR2) is a protein whose function is proving to play an increasingly critical role in a diverse range of biological systems, including vertebrate tissue development, inflammatory tissue injury responses, and cancer progression. AGR2 was identified initially as a secretory factor expressed in the anterior region of the dorsal ectoderm in Xenopus laevis embryos, where it was postulated to mediate the specification of dorsoanterior ectodermal fate, particularly in the formation of the cement gland.1, 2 AGR2 was subsequently cloned as a gene whose expression is induced by the estrogen receptor α,³ and subsequent studies in primary breast carcinomas have also shown significant associations between AGR2 expression and estrogen receptor-α positivity or tamoxifen resistance.4, 5 Clinical studies have shown that the AGR2 protein is overexpressed in a wide range of human cancers, including carcinomas of the esophagus, pancreas, breast, prostate, and lung.4, 6, 7, 8, 9 More biological studies in cell lines have shown a significant role for AGR2 in tumor-associated pathways, including tumor growth, cellular transformation, cell migration, limb regeneration, and metastasis.8, 10, 11, 12

AGR2 protein was also identified as part of a clinical proteomics screen aimed at discovering novel inhibitors of the tumor suppressor p53, and it was subsequently validated as a potent inhibitor of p53 activity and of the p53-dependent response to DNA damage.¹¹ The latter data provide a specific oncogenic pathway into which AGR2 integrates; however, the signaling mechanisms that drive AGR2 to inhibit p53 are not defined. Although there are no well-validated binding proteins in human cells that can explain how AGR2 can act as a pro-oncogenic protein, peptide aptamer screens have identified a specific peptide-binding activity for the AGR2 protein for peptides containing an (S/T)xIΦΦ consensus motif, suggesting that the AGR2 protein might prove to have a peptide groove able to interact with cellular proteins containing such a consensus motif.¹³ Furthermore, penetratin peptides linked to this AGR2-binding (S/T)xIΦΦ motif or EGFP fusions to the (S/T)xIΦΦ motif can stabilize AGR2 in cells and stimulate p53 activity, indicating that the AGR2 protein can interact with this peptide consensus motif in vivo.¹⁴ A yeast two-hybrid screen has also been used to previously identify the prometastatic proteins C4.4 and DYS1 as interactors of AGR2¹⁵; however, there was no biological validation of C4.4 and DYS1 as a bona fide protein–protein interaction in human cells. However, potential extracellular receptor functions for AGR2 in human cells remain possible, because an interaction between the newt extracellular receptor PROD1 and newt AGR2 was identified using a yeast two-hybrid screen and validated to demonstrate a direct signaling role for AGR2 in amphibian limb regeneration.¹⁰

Although the general biochemical functions of AGR2 in human cells remain undefined, AGR2 is part of the protein disulfide isomerase (PDI) superfamily that contains core thioredoxin folds (CxxC or CxxS motif), which have the potential to act as molecular chaperones that regulate protein folding via regulation of disulfide-bond formation.¹⁶ There are five protein members of this family: TRX1 (thought to be predominantly the nuclear thioredoxin), TRX2 (thought to be predominantly the mitochondrial thioredoxin), endoplasmic reticulum (ER) protein 18 (ERP18; the ancestral protein in the AGR2/AGR3 group that has potent reducing potential),¹⁷ AGR2, and the AGR2 ortholog AGR3. AGR2 and AGR3 are confined to vertebrates, and both have the CxxS core motif instead of the CxxC motif of TRX1, TRX2, and ERP18.¹⁸ The majority of PDIs/ERPs harbor a typical H/KDEL ER retrieval signal. A putative ER retention sequence that has been shown to regulate the intracellular localization of AGR2 in human cells has been identified at the C-terminus of AGR2.¹⁴ It is therefore possible that at least one function of the AGR2 is to act as a PDI and hence as a protein molecular chaperone. A recent study has confirmed that AGR2 is essential for the production of the intestinal mucin MUC2, a cysteine-rich glycoprotein that forms the protective mucus gel lining the intestine. The cysteine residue within the AGR2 thioredoxin-like domain was shown to form a mixed disulfide bond with a cysteine in the N-terminus or C-terminus of MUC2 as it is being processed.¹⁹ However, there are currently no established biochemical mechanisms to explain the function or the regulation of AGR2 protein.

In this report, a yeast two-hybrid screen was used to identify a potentially novel interacting protein for the AGR2 protein from a human breast cancer library. A protein named Reptin was identified by the yeast two-hybrid screen and validated as an interacting protein of AGR2 in human cells. Reptin is a highly conserved member of the AAA+ family that can be found in numerous multiprotein complexes linked to transcription, DNA damage response, and nonsense-mediated RNA decay.20, 21, 22, 23, 24, 25 This protein is a member of the highly conserved RuvBl1/2 superfamily containing ATP binding motifs and DNA binding and helicase functions and an ability to form biologically relevant protein–protein interactions with proteins implicated in cancer, including Myc, Tip60, APPL1, Pontin, and telomerase holoenzyme complexes.20, 23, 26, 27, 28, 29 The validation of Reptin as an AGR2 binding protein gives rise to a potentially novel signaling complex involved in prometastatic cancer development. Our mapping of the determinants that mediate a specific Reptin–AGR2 protein–protein complex in vitro provides biochemical insights for understanding how the Reptin–AGR2 complex can be regulated in cells. These data also provide ideas for development of in vitro enzyme assays for the screening of small molecules that might be used to disrupt the AGR2–Reptin complex in cancer cells as potential therapeutic leads.