MINI REVIEW article

Front. Pharmacol., 22 January 2019

Sec. Experimental Pharmacology and Drug Discovery

Volume 9 - 2018 | https://doi.org/10.3389/fphar.2018.01477

A Guide to PIN1 Function and Mutations Across Cancers

  • 1. Pathology Unit, IRCCS CRO Aviano-National Cancer Institute, Aviano, Italy

  • 2. Doctoral School in Molecular Biomedicine, University of Trieste, Trieste, Italy

  • 3. Doctoral School in Chemistry, University of Trieste, Trieste, Italy

  • 4. Department of Pharmacy, University of Pisa, Pisa, Italy

  • 5. Sbarro Institute for Cancer Research and Molecular Medicine, Center for Biotechnology, College of Science and Technology, Temple University, Philadelphia, PA, United States

  • 6. Department of Molecular Science and Nanosystems, Ca’ Foscari University of Venice, Venice, Italy

Abstract

PIN1 is a member of a family of peptidylprolyl isomerases that bind phosphoproteins and catalyze the rapid cis–trans isomerization of proline peptidyl bonds, resulting in an alteration of protein structure, function, and stability. PIN1 is overexpressed in human cancers, suggesting it promotes tumorigenesis, but depending on the cellular context, it also acts as a tumor suppressor. Here, we review the role of PIN1 in cancer and the regulation of PIN1 expression, and catalog the single nucleotide polymorphisms, and mutations in PIN1 gene associated with cancer. In addition, we provide a 3D model of the protein to localize the mutated residues.

Introduction

Proline is the only amino acid with the ability to adopt either a cis or trans conformation, and this isomerization is catalyzed by peptidylprolyl isomerases (PPIases). The cis-trans isomerization of proline in phosphorylated Ser/Thr-Pro motifs is catalyzed exclusively by PIN1 (peptidylprolyl cis/trans isomerase, NIMA-interacting 1) ().

PIN1-mediated isomerization is an important regulatory mechanism in human physiology and pathology: the conformational change regulates various protein functions, including the catalytic activity, the phosphorylation status, protein interaction, subcellular location, and/or protein stability ().

Structurally, PIN1 has two domains connected by a flexible linker: the N-terminal domain is called “WW” (referring to two invariant Trp residues) and targets the enzyme to pSer/Thr-Pro motifs in substrates; the C-terminal PPIase domain has the catalytic activity ().

PIN1 is involved in cellular processes such as the cell cycle, the folding of newly synthesized proteins, responses to DNA damage and stress, and immune responses (). It is overexpressed in several human cancers (), including prostate cancer (; ), breast cancer (; ; ), and oral squamous carcinomas (). However, it is still not fully understood how this enzyme participates in cancer development and progression. Several studies showed that some single nucleotide polymorphisms (SNPs) in PIN1 gene increase the risk of cancer whereas other variants operate as protective factors (; ; ; ; ). Little has been reported so far about PIN1 somatic mutations and cancer. This review summarizes the role of PIN1 in cancer and the regulation of PIN1 expression, and is an exhaustive guide to PIN1 SNPs and mutations across cancers.

Pin1 as an Oncogene or Conditional Tumor Suppressor Gene

PIN1 has been shown to be a proto-oncogene whose protein product regulates several proteins involved in cancer initiation and progression (; ). For example, PIN1 upregulates the expression of cyclin D1 at both the transcriptional and post-translational levels. At the transcriptional level, PIN1 activates transcription of the gene encoding cyclin D1 (CCND1) via two signal transduction pathways. In the Ras signaling pathway, activation of a kinase cascade leads to phosphorylation, and activation of JNK (c-Jun N-terminal kinase), which phosphorylates and activates the transcription factor c-Jun. PIN1 can bind and isomerize both phosphorylated JNK and phosphorylated c-Jun to potentiate c-Jun transcriptional activity at the CCND1 promoter ().

PIN1 also stimulates cyclin D1 expression via the Wnt / β-catenin pathway. Briefly, in unstimulated cells, a complex composed of adenomatous polyposis coli (APC), glycogen synthase kinase 3β (GSK-3β), and other proteins keeps cytosolic levels of β-catenin low by triggering this protein’s phosphorylation, ubiquitination and degradation. When extracellular Wnt proteins activate their receptor (composed of a Frizzled receptor and other proteins), GSK-3β is displaced from the complex so β-catenin can accumulate and translocate to the nucleus. There, β-catenin binds transcription factors and other co-activators in a transcription complex that activates CCND1 and other Wnt target genes (). PIN1 and β-catenin levels are strictly correlated. PIN1 inhibits the APC-dependent exporting of β-catenin from the nucleus to the cytoplasm and cytoplasmic degradation of β-catenin, thereby β-catenin accumulates in the nucleus where it activates the transcription of genes such as CCND1 ().

At the protein level, PIN1 isomerizes cyclin D1; this protein modification has a stabilizing effect (). Cyclin D1 then accumulates in the nucleus, where in concert with other proteins it drives cell cycle progression (; ; ). The cyclin D1 activation as downstream target suggests that PIN1 coordinates different events of cell cycle, by acting as molecular timer, and that the overexpression of PIN1 in cancer leads to uncontrolled cell cycle.

Other oncogenic proteins stabilized by being isomerized by PIN1 are Akt (also called protein kinase B), retinoblastoma-associated protein (pRb), and myeloid cell leukemia 1 protein (MCL-1). PIN1 isomerization of Akt is critical for activation of the Akt signaling cascade that in turn activates the transcription of genes encoding cyclin D1, p53 and IKK-NFκB. In cancer cells, high levels of PIN1 amplify the activation of the Akt cascade and thus enhance tumor progression (). PIN1 isomerization of pRb facilitates its binding to CDK–cyclin complexes in mid- to late G1. As a result, pRb is hyperphosphorylated and orchestrates cell proliferation by allowing the expression of genes that mediate entry into the S phase via the E2F transcription factor. In cancer, PIN1 overexpression leads to pRb pathway iperactivation (, ). Finally, isomerization of MCL-1 causes a conformational change that may stabilize the protein and enhance its anti-apoptotic function. Briefly, MCL-1 is phosphorylated by GSK-3β, facilitating MCL-1 association with the E3 ligase β-TrCP. The interaction between MCL-1 and the GSK-3β–E3 ligase β-TrCP complex leads to MCL-1 ubiquitination and degradation (). PIN1-mediated isomerization may prevent MCL-1 association with the GSK-3β–E3 ligase β-TrCP complex, blocking MCL-1 degradation, but further studies are required ().

Finally, PIN1 isomerizes two transcription factors, namely NF-κB, increasing its nuclear retention (), and STAT3, promoting its transactivation (; ). These two proteins are involved in inflammation-induced carcinogenesis and are constitutively activated in several cancers (). PIN1’s action on these transcription factors enhances the transcription of genes encoding cyclin D1, c-Myc and Bcl-2 (; ).

Altogether, these results feature PIN1 as a tumor promoter, but Yeh and Means described PIN1 as a “conditional” tumor suppressor () and successive studies support this theory. Indeed, PIN1 can induce apoptosis, prevent genomic instability, and promote the ubiquitin-dependent proteolysis of many oncogenic proteins. All these processes limit tumor progression.

In stress conditions, PIN1 induces apoptosis via p53 and p73 (). Moreover, a study on murine embryonic fibroblasts showed that PIN1 prevents p53-dependent genomic instability ().

PIN1 is involved in the ubiquitin-dependent proteolysis of Myc, Bcl-6 and cyclin E (; ; ; ). For example, PIN1 binds to doubly phosphorylated Myc on Thr58 and Ser62. The conformational change facilitates Myc dephosphorylation on Ser62 by protein phosphatase 2 (PP2A), which allows Myc ubiquitination and degradation by the proteasome (). Additionally, upon DNA damage, the kinase ATM phosphorylates Bcl-6 that in turn becomes a substrate for PIN1. Bcl-6 isomerization signals its degradation by the ubiquitin-proteasome system ().

Finally, PIN1 regulates the degradation of cyclin E during the G0/G1-S phase transition of the cell cycle permitting to cell cycle to proceed correctly (). Experiments in which PIN1 expression was down-regulated showed increased steady-state levels of cyclin E, the arrest of cells in G1/S phase, genomic instability, and tumoral transformation ().

Regulation Of PIN1 Expression

Transcriptional Regulation

PIN1 transcription can be activated by the E2F transcription factor or by Notch1 binding to the PIN1 promoter (; ). In this way PIN1 sustains the transformed phenotype induced by E2F or Notch1 activation.

PIN1 transcription can also be suppressed by the tumor suppressor gene BRCA1 (). BRCA1 associates with several proteins to regulate DNA repair response. In cancer, BRCA1 is often mutated and lost such function, thereby cells accumulates DNA damage ().

Recently, micro-RNAs (miRNAs) have been identified as regulators of PIN1 expression. For instance, miR200c binds to a conserved region in the 3′-untranslated region (UTR) of PIN1 mRNA and prevents its translation (). Mutations in this region of PIN1 can prevent the repressive effects of miR200c (). miRNA-200b, and miR-296-5p also bind the 3′ UTR of PIN1 mRNA and down-regulate its expression. In cancer cells, both these miRNAs were found to be underexpressed, allowing PIN1 to sustain tumor progression (; ).

Post-translational Regulation

Depending on the physiological or pathological conditions, the activity of proteins is regulated by post-translational modifications. For PIN1, there is evidence of post-translational modification by phosphorylation, ubiquitination, SUMOylation and oxidization at specific sites.

Phosphorylation of PIN1 on Ser16 in the WW domain suppresses its ability to interact with its substrates (). At least three kinases can phosphorylate this residue: protein kinase A (), ribosomal S6 kinase 2 (), and aurora kinase A (). Phosphorylation on Ser71 in the PPIase domain inhibits the protein’s enzymatic activity (). Phosphorylation on Ser65 by polo-like kinase (Plk1) () and on Ser138 by mixed-lineage kinase 3 increases PIN1’s catalytic activity and nuclear translocation (). Plk1-mediated Ser65 phosphorylation is suggested to regulate PIN1 turnover. It induced PIN1 deubiquitination and stabilization, while the absence of Plk1 enhanced the ubiquitination and degradation of PIN1 ().

SUMOylation of Lys6 in the WW domain and Lys63 in the PPIase domain abolishes PIN1’s enzymatic activity and oncogenic functions (). However, deSUMOylation of these two domains by SUMO1/sentrin specific peptidase 1 (SENP1) restores PIN1’s activity. SENP1 overexpression increases the levels of deSUMOylated PIN1 and in turn the ability of PIN1 to induce centrosome amplification and cell transformation ().

Finally, under conditions of oxidative stress, PIN1 is oxidized on Cys113 in the catalytic site, inhibiting its enzymatic activity ().

PIN1 Single Nucleotide Polymorphisms and Cancer Risk

Several SNPs, located in the promoter or coding region of PIN1, are associated with cancer risk. The PIN1 variants rs2233678 (c.–842G>C) and rs2233679 (c.–667T>C), both located in the promoter, and the synonymous change rs2233682 (G>A; p.Gln33Gln) in exon 2 of the coding region, have been widely investigated.

The −842C allele of rs2233678 was found to confer a significantly lower risk of cancer (odds ratio, 0.75) in a meta-analysis of 11 studies (9280 participants) of patients with esophageal carcinoma, nasopharyngeal carcinoma, laryngeal squamous cell carcinoma, lung cancer, breast cancer, squamous cell carcinoma of the head, and neck or hepatocellular carcinoma (HCC), and matched healthy controls (). Seven of the included studies had found that the allele reduced risk, while four found no association with cancer risk.

A study of 209 patients with oral squamous cell carcinoma and 444 controls did not find an association between the −842G>C polymorphism and cancer risk ().

The −667T>C polymorphism has been found to not associate with esophageal carcinoma (), breast cancer (), or squamous cell carcinoma of the head and neck (), whereas it did associate with a lower risk of nasopharyngeal carcinoma (), and a higher risk of oral squamous cell carcinoma ().

Finally, the synonymous change Gln33Gln was not found to associate with the risk of breast cancer (), or squamous cell carcinoma of the head and neck (). However, a higher risk of HCC was found among carriers of the Gln33Gln variant in a Chinese population ().

PIN1 Somatic Mutations in Cancer Tissues

Because of the lack of published papers reporting on PIN1 somatic mutations, we obtained deposited genetic data on these mutations in different tumor types from the cBioPortal for Cancer Genomics and COSMIC (Catalog of Somatic Mutations in Cancer). cBioPortal is the main resource for the analysis of large-scale cancer genomics datasets (; ). COSMIC, a database of mutations reported in the scientific literature or from the Cancer Genome Project, permits researchers to explore the effects of somatic mutations in cancer ().

The data obtained from cBioPortal regarded 11,000 cancer cases collected for genomic characterization in December 2013 and 32,555 cases in 61 primary sites of cancer (retrieved on June 13, 2018) (; ). The data from COSMIC refer to 41,924 unique samples from patients with different types of cancer including skin, breast, intestinal, lung, liver, prostate, and stomach cancer (retrieved on June 14, 2018) (). Altogether, the data revealed the existence of 32 somatic mutations affecting 29 unique residues in the coding region of PIN1 gene (Table 1). Five mutations affect the WW domain (residues 1–39), two are in the flexible linker (residues 35–53), and 25 affect the PPIase domain (residues 50–163). Twenty-three are missense mutations, four are synonymous mutations (they have no effect on PIN1 function), and four are nonsense mutations (R21∗, S71∗, S108∗, and E163∗). Overall, 17 mutations were predicted to be pathogenic by the Functional Analysis through Hidden Markov Models (FATHMM) filter in COSMIC and three to be deleterious by the Sorting Intolerant from Tolerant (SIFT) algorithm in cBioPortal. The others are predicted to be tolerated.

Table 1

AA changeTypePredicted functional consequencePosition (GRCh37)Nucleotide changeCancerPatients n§Frequency %Reference or study identifier
G20GSplicingNonea9949113C>TSkin cutaneous melanoma11210.83
R21∗NonsensePathogenicb9949114C>TSkin cutaneous melanoma23660.27
Q33KMissensePathogenicb9949150C>ASkin212150.08
R36PMissensePathogenicb9949160G>CLarge intestine214820.07
G39CMissenseDeleteriousa9949168G>THCC33730.54;
G39CMissensePathogenicb9949168G>TSCLC2422.38
S42IMissenseNoneb9949178G>TLarge intestine214820.07
Q49QSynonymousNoneb9949200G>ASkin212150.08COSU540c
V55IMissensePathogenicb9949216G>AER+ breast cancer221030.05
S71*NonsenseNonea9949265C>ASarcoma32470.40;
S71LMissensePathogenicb9949265C>TSkin212150.08
S71SSynonymousNeutralb9949266G>AStomach adenocarcinoma27900.13COSU541c
E100DMissenseTolerateda9958734G>TCRC42240.45
E104KMissenseTolerateda9958744G>ANSCLC511440.09
S105FMissenseNoneb9958748C>TLarge intestine214820.07
S108∗NonsensePathogenicb9958757C>ASkin212150.08
D112NMissensePathogenicb9958768G>ALarge intestine214820.07
A124VMissensePathogenicb9958805C>TStomach adenocarcinoma22890.35COSU541c
P133LMissensePathogenicb9959781C>TDesmoplastic melanoma2205.00
F134SMissenseDeleteriousa9959784T>CNeuroendocrine prostate cancer6811.23
S138SSynonymousNeutralb9959797G>ALarge intestine214820.07
F139SMissensePathogenicb9959799T>CCervical squamous cell carcinoma21940.52COSU415c
T143MMissenseDeleteriousa9959811C>TAdeno-cortical carcinoma3901.11;
G144EMissensePathogenicb9959814G>AHCC218160.06COSU381c
E145KMissenseTolerateda9959816G>AHead & neck squamous cell carcinoma35100.20;
G148RMissensePathogenicb9959825G>CEsophagus-stomach cancers25180.19
G148GSynonymousNoneb9959827G>ABiliary tract cancer23660.27COSU658c
P149SMissensePathogenicb9959828C>TSkin212150.08
T152MMissensePathogenicb9959838C>TCRC26190.16
S154FMissenseDeleteriousa9959844C>TNSCLC511440.09
H157YMissensePathogenicb9959852C>TSkin212150.08
T162IMissensePathogenicb9959868C>TLarge intestine214820.07
E163*NonsenseNoneb9959870G>TSquamous cell carcinoma218350.05COSU583c

PIN1 mutations in cancer.

§Number of patients in cancer study; ∗The mutation inserts a stop codon; aSIFT algorithm in cBioPortal; bFATHMM filter in COSMIC; cCOSMIC study identifier; 1Skin cutaneous melanoma-cBioPortal; 2COSMIC database; 3TCGA, PanCancer Atlas-cBioPortal; 4TCGA, Colorectal Adenocarcinoma-cBioPortal; 5Pan-lung cancer-cBioPortal; 6Neuroendocrine Prostate cancer-cBioPortal (Trento/Cornell/Broad). CRC, colorectal cancer; HCC, hepatocellular carcinoma; NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer.

Figure 1 illustrates the positions of PIN1 somatic mutations that alter the protein’s primary sequence and are predicted to be pathogenic or deleterious. This model is based on the X-ray crystal structure of PIN1 bound to a non-natural peptide inhibitor (Protein Data Bank accession code, 2ITK) (). The Q33K and R36P mutations (orange) are in the WW domain, as is G39C that is not shown because it belongs to a peptide loop missing from the X-ray structure. All the other mutations are found in the PPIase domain: F134S, S154F, and H157Y (magenta) interact with the enzyme’s substrates (; ) while S71L, D112N, P133L, and T152M (yellow) are indirectly involved in the interaction with substrates (; ; ). The other mutations (green) do not interact with substrates and thus could have a role in interactions with other proteins or in PIN1 protein folding. Among them, F139S is within the PPIase domain interface (S138 to R142) that is involved in interdomain communication and regulates the function of PIN1 upon substrate binding (; ).

FIGURE 1

).

PIN1 Inhibitors

In the last decade, some PIN1 inhibitors were discovered by industries and academic research groups. Compounds 1 () and 2 (; Figure 2), developed by Pfizer Worldwide Research & Development, are nanomolar PIN1 inhibitors (Ki = 6 and 890 nM for compounds 1 and 2, respectively), as confirmed by X-ray structures that showed their binding to the enzyme. Unfortunately, both of them failed to be active in cell-based assays, likely due to the phosphate or carboxylate groups which were necessary to properly interact in a charged pocket of the enzyme, but strongly limited cell membrane permeability.

FIGURE 2

The α-Amino acid-derived compounds 3 and 4 (Figure 2) were discovered by researchers at Vernalis (R&D) Ltd. (). These compounds maintained the carboxylate group, which is the moiety necessary for an optimal interaction in the enzyme binding site, and the aromatic portions were varied, inserting benzimidazole (3), or naphthalene (4), since they lied in a hydrophobic region of the protein. Compound 3 showed an IC50 value in the submicromolar range (0.13 μM), but it was unable to be active in cellular assays. Nevertheless, with the aim of reducing the polar surface area, its analog 4 was about twenty-fold less potent on the isolated enzyme (IC50 = 2.6 μM), but it gained activity in PC3 prostate cancer cells in which it reduced the proliferation.

Due to the problems encountered with the benzimidazole/naphthyl-based compounds, a further development in the search for PIN1 inhibitors at Vernalis (R&D) Ltd, consisted in the phenyl-imidazole derivatives. As a result, compound 5 (Figure 2) reached an optimal balance between inhibition activity on PIN1 (IC50 = 0.83 μM) and antiproliferative activity in PC3 cells (GI50 = 13 μM) ().

The dihydrothiazole series, exemplified by compound 6 (Figure 2), was discovered in 2014 at Pfizer Worldwide Research & Development: in this chemical class, the amide group present in all the PIN1 inhibitors previously reported in literature was substituted by a dihydrothiazole ring, bearing a carboxylic acid moiety, with aim of decreasing polar surface area, which is a factor strongly influencing cellular permeability (; ). They showed micromolar inhibition potencies (IC50 values ranging from 1.9 to 27 μM) on PIN1 and they were able to reduce the proliferation of colon cancer cells (HT29) ().

A new class of PIN1 inhibitors developed by Zhao et al. consisted in thiazole derivatives bearing oxalic or acetic acid group at 4-position and, according to modeling studies, this last portion was found to be located in the charged pocket of the enzyme. Compound 7 (Figure 2) showed an IC50 value of 2.93 μM, but unfortunately no data about its activity in cancer cells was reported (). Among the recently developed PIN1 inhibitors, a selenium containing compound (compound 8, Figure 2) was identified by a novel high-throughput screening study. Compound 8 efficiently inhibited PIN1 (IC50 = 0.43 μM) and it was able to affect the proliferation of breast MDA-MB-231 cancer cells in which PIN1 is overexpressed, also reducing the viability of induced cancer stem cell-like cells ().

identified PIN1 as a target of all-trans retinoic acid (ATRA), compound 9 (Figure 2). ATRA inhibited and degraded PIN1 (Ki value of 0.82 μM), as confirmed by the co-crystal structure of ATRA with PIN1. Furthermore, ATRA was able to suppress the growth of triple-negative breast cancer and acute promyelocytic leukemia cells, both in humans and in animal models.

It is also noteworthy to mention two covalent PIN1 inhibitors. The α,β-unsaturated isothiazolone derivative 10 (Figure 3) showed a micromolar inhibition activity on the enzyme (IC50 = 6.1 μM); however, it was not selective for PIN1 also exerting a similar activity on cyclophilin (IC50 = 13.7 μM) (). The quinone-sulfonamide derivative 11 (Figure 3) covalently bound Cys 113 (IC50 value of 0.64 μM), ultimately leading to PIN1 degradation. Moreover, 11 impaired PIN1-dependent invasive behavior of breast (MDA-MB-231) and prostate (PC3) cancer cells ().

FIGURE 3

Conclusion

PIN1 is overexpressed in various cancer types and is associated with a malignant phenotype and tumor progression (; ; ). PIN1 may also have an anti-cancer role depending on the cellular context; therefore PIN1 has been called a conditional tumor suppressor gene (; ). Some SNPs in PIN1 gene were found to associate with cancer risk. Carriers of the −842C variant in the PIN1 promoter have low PIN1 protein levels and low risk for developing cancer (). Contrasting evidence has been reported for the −667C variant in the PIN1 promoter, which was found to associate with a low risk of developing nasopharyngeal carcinoma (), but a high risk for oral squamous cell carcinoma and HCC (; ).

So far, 32 somatic mutations in PIN1 gene have been found in different types of cancer. Of these, 20 are predicted to be pathogenic or deleterious. Although further studies are required, we believe that investigating the complex pattern of PIN1 gene alterations and their effects on PIN1 protein structure and function is a valid strategy for identifying new biomarkers for susceptibility to cancer and response to anti-PIN1 inhibitors.

Statements

Author contributions

All the authors wrote and approved the manuscript.

Acknowledgments

The authors are grateful to and would like to recognize the Associazione Italiana per la Ricerca sul Cancro – My First AIRC (n. 15639). Valerie Matarese provided editorial advice and scientific editing.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References

Summary

Keywords

PIN1, cancer, mutations, SNP, 3D modeling

Citation

El Boustani M, De Stefano L, Caligiuri I, Mouawad N, Granchi C, Canzonieri V, Tuccinardi T, Giordano A and Rizzolio F (2019) A Guide to PIN1 Function and Mutations Across Cancers. Front. Pharmacol. 9:1477. doi: 10.3389/fphar.2018.01477

Received

30 October 2018

Accepted

03 December 2018

Published

22 January 2019

Volume

9 - 2018

Edited by

Filippo Caraci, Università degli Studi di Catania, Italy

Reviewed by

Fabrizio Carta, Università degli Studi di Firenze, Italy; Simone Carradori, Università “G. d’Annunzio” di Chiet – Pescara, Italy

Updates

Copyright

*Correspondence: Flavio Rizzolio,

†Co-first authors

This article was submitted to Experimental Pharmacology and Drug Discovery, a section of the journal Frontiers in Pharmacology

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Article metrics