α-enolase: a promising therapeutic and diagnostic tumor target

Authors

  • Michela Capello,

    1. Department of Medicine and Experimental Oncology, Center for Experimental Research and Medical Studies (CeRMS), San Giovanni Battista Hospital, University of Turin, Italy
    Search for more papers by this author
  • Sammy Ferri-Borgogno,

    1. Department of Medicine and Experimental Oncology, Center for Experimental Research and Medical Studies (CeRMS), San Giovanni Battista Hospital, University of Turin, Italy
    Search for more papers by this author
  • Paola Cappello,

    1. Department of Medicine and Experimental Oncology, Center for Experimental Research and Medical Studies (CeRMS), San Giovanni Battista Hospital, University of Turin, Italy
    Search for more papers by this author
  • Francesco Novelli

    1. Department of Medicine and Experimental Oncology, Center for Experimental Research and Medical Studies (CeRMS), San Giovanni Battista Hospital, University of Turin, Italy
    Search for more papers by this author

F. Novelli, Center for Experimental Research and Medical Studies (CeRMS), San Giovanni Battista Hospital, Via Cherasco 15, 10126 Turin, Italy
Fax: +39 011 633 6887
Tel: +39 011 633 4463
E-mail: franco.novelli@unito.it

Abstract

α-enolase (ENOA) is a metabolic enzyme involved in the synthesis of pyruvate. It also acts as a plasminogen receptor and thus mediates activation of plasmin and extracellular matrix degradation. In tumor cells, ΕΝΟΑ is upregulated and supports anaerobic proliferation (Warburg effect), it is expressed at the cell surface, where it promotes cancer invasion, and is subjected to a specific array of post-translational modifications, namely acetylation, methylation and phosphorylation. Both ENOA overexpression and its post-translational modifications could be of diagnostic and prognostic value in cancer. This review will discuss recent information on the biochemical, proteomics and immunological characterization of ENOA, particularly its ability to trigger a specific humoral and cellular immune response. In our opinion, this information can pave the way for effective new therapeutic and diagnostic strategies to counteract the growth of the most aggressive human disease.

Abbreviations
EGFR

epidermal growth factor receptor

ENOA

α-enolase

ERK

extracellular signal-regulated kinase

MBP-1

c-myc promoter-binding protein

MHC

major histocompatibility complex

MMP

matrix metalloproteinase

PAI-1

plasminogen activator inhibitor-1

PTM

post-translational modification

TAA

tumor-associated antigen

tPA

tissue-type plasminogen activator

uPA

urokinase-type plasminogen activator

uPAR

urokinase-type plasminogen activator receptor

Introduction

Enolase is a metalloenzyme that catalyzes the dehydration of 2-phospho-d-glycerate to phosphoenolpyruvate in the second half of the glycolytic pathway. In the reverse reaction (anabolic pathway), which occurs during gluconeogenesis, the enzyme catalyzes the hydration of phosphoenolpyruvate to 2-phospho-d-glycerate [1,2]. Enolase is found from archaebacteria to mammals, and its sequence is highly conserved [3]. In mammals, three genes, ENO1, ENO2 and ENO3 encode for three isoforms of the enzyme, α-enolase (ENOA), γ-enolase and β-enolase, respectively, with high sequence identity [4–6]. The expression of these isoforms is tissue specific: ENOA is present in almost all adult tissues, β-enolase is expressed in muscle tissues and γ-enolase is found in neurons and neuroendocrine tissues [1,7–9]. The monomer of ENOA consists of a smaller N-terminal domain (residues 1–133) and a larger C-terminal domain (residues 141–431). In eukarya, enzymatically active enolase consists of a dimeric form in which two subunits face each other in an antiparallel manner [1,10]; some eubacterial enolases, by contrast, are octameric [11]. Enolase can form homo- or heterodimers, such as αα, αβ, ββ, αγ and γγ [1].

Apart from its enzymatic activity, in many prokaryotic and eukaryotic cells, ENOA is expressed on the cell surface, where it acts as a plasminogen receptor promoting cell migration and cancer metastasis [12–23]. Moreover, ENO1 can be translated into a 37 kDa protein, c-myc promoter-binding protein (MBP-1), by using an alternative start codon [24]. MBP-1 lacks the first 96 residues of ENOA and localizes in the nucleus, where it binds to the c-myc P2 promoter and acts as a transcription repressor, leading to tumor suppression [25–27]. ENOA associates with MBP-1 in the transcriptional regulation of the oncogene c-myc [28].

ENOA is a surface plasminogen-binding receptor in tumors

In breast, lung and pancreatic neoplasia, ENOA is localized on the surface of cancer cells [29–31], whereas in melanoma and nonsmall cell lung carcinoma cells it can also be secreted by exosomes [32,33]. How ENOA is displayed on the cell surface remains unknown. Many glycolytic enzymes and cytosolic proteins that lack N-terminal signal peptide reach the surface of eukaryotic cells [34]. In mammal cells, some export routes of unconventional protein secretion have been postulated: membrane blebbing, membrane flip-flop, endosomal recycling or a plasma membrane transporter [35]. One possibility is that phosphoinositides recruit ENOA and translocate it to the cell surface [36]. It is not known if surface ENOA is also present as a monomer. As the monomeric form is catalytically inefficient it could be available to interact with other proteins that mediate its transport to the cell surface [37]. However, in breast cancer cells, surface ENOA maintains its catalytic activity, suggesting that cell surface localization does not affect this function [31].

Cell surface ENOA is one of the many plasminogen-binding molecules that include actin [38], gp330 [39], cytokeratin 8 [40], histidine-proline rich glycoprotein [41], glyceraldehyde-3-phosphate dehydrogenase [42], annexin II [43], histone H2B [44] and gangliosides [14]. ENOA and most of these proteins have C-terminal lysines predominantly responsible for plasminogen activation [45]. Interaction of the plasminogen lysine-binding sites with ENOA is dependent upon recognition of ENOA C-terminal lysines K420, K422 and K434 [14]. In view of the surface potential of the human ENOA crystal structure, an additional plasminogen binding site that includes K256 has been proposed [10].

Binding with ENOA lysyl residues leads to activation of plasminogen to plasmin by the proteolytic action of either tissue-type (tPA) or urokinase-type (uPA) plasminogen activators [19,46]. Plasmin is a serine protease with a broad spectrum substrate, including fibrin, extracellular matrix components (laminin, fibronectin) and proteins involved in extracellular matrix degradation (matrix metalloproteinases, such as MMP3) [47–50]. Binding of plasminogen to the cell surface has profibrinolytic consequences: enhancement of plasminogen activation, protection of plasmin from its inhibitor α2-antiplasmin and enhancement of the proteolytic activity of cell-bound plasmin [13,51]. Proteolysis mediated by cell-associated plasmin contributes to both physiological processes, such as tissue remodeling and embryogenesis, and to pathophysiological processes, such as cell invasion, metastasis and inflammatory response [19,45]. A noteworthy positive correlation exists between elevated levels of plasminogen activation and malignancy [46,52]. Higher expression levels of uPA and/or plasminogen activator inhibitor-1 (PAI-1) in tumor tissues correlate with aggressiveness and poor prognosis. ENOA takes part, together with urokinase plasminogen activator receptor (uPAR), integrins and some cytoskeletal proteins, in a multiprotein complex, called metastasome, responsible for adhesion, migration and proliferation in ovarian cancer cells [53]. In human follicular thyroid carcinoma cells, retinoic acid causes a decrease in ENOA levels that coincides with their reduced motility [54], and cell surface ENOA is enhanced in breast cancer cells rendered superinvasive following paclitaxel treatment [55].

In pancreatic cancer patients, deregulated expression of many proteins involved in the plasminogen pro-fibrinolytic cascade (annexin A2, PAI-2, uPA, uPAR, MMP-1 and MMP-10) correlates with survival [56–59]. In the same tumor, tPA activates a mitogenic signal mediated by extracellular signal-regulated kinase (ERK)-1/2 through epidermal growth factor receptor (EGFR) and annexin A2 [60,61]. These proteins probably form a complex that also includes ENOA, as it has been pulled down with annexin A2, cytokeratin 8 and tPA in raft membrane fractions of pancreatic cancer cells [62].

ENOA is a tumor-associated antigen (TAA)

TAAs are self-proteins that can trigger multiple specific immune responses in the autologous host [63]. Activation of the immune system against TAAs occurs at an early stage of tumorigenesis, as illustrated by the detection of high titers of autoantibodies in patients with early-stage cancer [64], and correlates with the progression of malignant transformation [65]. It is not entirely clear how TAAs are able to trigger humoral responses, especially as many of those discovered so far are intracellular proteins, but are thought to be altered in a way that renders the proteins immunogenic [66,67]. Several hypotheses have been proposed: these self-proteins could be overexpressed, mutated, misfolded, aberrantly degraded or localized so that autoreactive immune responses in cancer patients are induced [65,68,69]. Moreover TAAs that have undergone post-translational modifications (PTMs) (e.g. glycosylation, phosphorylation, acetylation, oxidation and proteolytic cleavage) may be perceived as foreign by the immune system [66–68]. The immune response against such immunogenic epitopes of TAAs induces the production of autoantibodies as serological biomarkers for cancers [70]. Both its overexpression in tumors and its ability to induce a humoral and/or cellular immune response in cancer patients classify ENOA as a true TAA.

ENOA expression is increased in tumors

The overexpression of ENOA is associated with tumor development through a process known as aerobic glycolysis or the Warburg effect [71]. Warburg observed that cancer cells consume more glucose than normal cells and generate ATP by converting pyruvate to lactic acid, even in the presence of a normal oxygen supply [72]. The mechanism of the Warburg effect was uncertain until the recent identification of upregulation of glycolytic enzymes by hypoxia-inducible factor. When a solid tumor exceeds 1 mm3, its cells face hypoxic stress due to slow angiogenesis [73,74]. Because the ENO1 promoter contains a hypoxia responsive element [75,76], ENOA is upregulated at the mRNA and/or protein level in several tumors, including brain [77], breast [78–83], cervix [77,84,85], colon [77,86,87], eye [77], gastric [77,88,89], head and neck [90,91], kidney [77], leukemia [92], liver [77,93,94], lung [77,95–99], muscle [77], ovary [77,100], pancreas [29,77,101,102], prostate [77,103], skin [104] and testis [77] (Table 1). Results from a bioinformatic study support a correlation between ENOA expression and tumorigenicity [52,77]. Moreover, ENOA’s enzymatic activity may also be increased in breast tumor tissue, especially in metastatic sites [82,83]. Increased ENOA expression can influence chemotherapy treatments, as shown in estrogen receptor-positive breast tumors, where it induces tamoxifen resistance [78], and in colorectal carcinoma cells, where it is overexpressed after 5-fluorouracil administration [87].

Table 1.   Expression of ENOA, the immune response to it and clinical correlations in cancer.
CancerENOA enhanced expressionImmune response to ENOAClinical correlations
  1. Percentages indicate the reported frequencies of enhanced ENOA mRNA, protein and enzymatic activity or the frequencies of anti-ENOA Ig.

  2. m, mRNA; p, protein; e, enzymatic activity; Ab, antibody production; T, T cell response; DP, disease progression; DFI, disease-free interval; M, malignancy; OS, overall survival; PFS, progression-free survival.

Brainm [77]  
Breastm (68%), p, e (100%) [78–83]Ab [69,125]DP, DFI, M [69,78]
Cervixm, p [77,84,85]  
Colonm, p [77,86,87]  
Eyem [77]  
Gastricm (73%), p [77,88,89]  
Head and neckm (68%), p [90,91]Ab (79%) [91,123,124], T [131,132]OS, PFS [91]
Kidneym [77]  
Leukemiap (> 50%) [92]Ab (33–86%) [120,121] 
Liverm, p (17–80%) [77,93,94] M [93,94]
Lungm, p (79–100%) [77,95–99]Ab (7–80%) [30,69,96,99,126–129]DP, OS, PFS [69,99]
Musclem [77]  
Ovarym, p [77,100]  
Pancreasm (100%), p (82–90%) [29,77,101,102]Ab (62%) [119], T [29]OS, PFS [119]
Prostatem, p (100%) [77,103]  
Skinm [104]Ab (38–100%) [104,122] 
Testism [77]  

ENOA PTMs in tumors

PTMs are common mechanisms that control signal transduction, protein-protein interaction and translocation [105,106]. Reversed-phase liquid chromatography, nanospray tandem mass spectrometry has been used to characterize ENOA PTMs in several cancer and normal cell lines (Table 2) (http://www.uniprot.org/uniprot/P06733) [107–115].

Table 2.   ENOA PTMs in normal and cancer tissues. Asp, aspartate; Glu, glutamate; Lys, lysine; Ser, serine; Thr, threonine; Tyr, tyrosine; numbers refer to the position of each residue in the ENOA amino acid sequence.
Cell typeAcetylationMethylationPhosphorylationReference
ResiduePositionResiduePositionResiduePosition
Embryonic kidney    Tyr57 111
    Ser63
Normal pancreasLys64, 71, 80, 81, 89, 92, 126, 193, 202, 228, 233, 281, 335, 343, 358, 406, 420Asp23, 91, 203, 209, 274, 299, 300, 378Ser419 115
Glu21, 45, 48, 86, 88, 96, 101, 187, 210, 219, 250, 293, 375, 377, 415, 416  
Cervix carcinomaLys71  Thr72 112–114
  Ser254, 263
Colon cancerSer2     http://www.uniprot.org/uniprot/P06733#ref14
LeukemiaLys5, 60, 64, 71, 80, 81, 89, 126, 193, 199, 221, 228, 233, 256, 281, 285, 343, 406, 420  Ser37, 40, 281 107–109
  Thr41, 390
  Tyr44, 287
Lung cancer    Tyr44, 287 110
Pancreatic cancerLys28, 64, 71, 80, 81, 89, 92, 103, 105, 126, 193, 202, 221, 228, 233, 239, 256, 262, 281, 285, 330, 335, 343, 358, 406, 420Asp23, 91, 203, 209, 266, 274, 286, 294, 297, 299, 300, 378, 383Ser419 115
Glu21, 45, 48,86, 88, 96, 101, 167, 187, 210, 219, 222, 225, 250, 293, 352, 375, 377, 414, 415, 416

Acetylation, methylation and phosphorylation are the main PTMs (Table 2). Acetylation was found in cervix and colon cancer, leukemia, normal pancreatic ducts and tumoral pancreatic cells. Fourteen acetylated lysine residues are common to leukemia, pancreatic cancer and normal pancreas, and one of them is the only acetylated residue in cervix tumor. Three acetylations are common to both leukemia and pancreatic cancer, whereas three are specific for normal and tumoral pancreatic cells. However, six specific acetylated lysines were found in pancreatic cancer cells, and three in leukemia. The only acetylated serine identified is specific for colon cancer (Table 2).

Methylation has been assessed in normal and tumoral pancreas only. Twenty-four aspartate and glutamate residues were found in both cell types. However, five aspartates and five glutamates are specifically methylated only in pancreatic cancer (Table 2).

Phosphorylation is the PTM that displays the most specific pattern in each cell line. Two serine and one threonine residues were specifically found in cervix cancer, one threonine and one serine in embryonic kidney, three serines and two threonines in leukemia; whereas two tyrosine residues were found in both leukemia and lung cancer and one serine in both tumoral and normal pancreas.

ENOA in tumor cells is subjected to more acetylation, methylation and phoshorylation than in normal tissues, indicating that many PTMs are associated with cancer development and some are specific for each kind of tissue or cancer. This can reflect the specific activation of pro-mitogenic signaling pathways in tumor cells. In many cases, PTMs regulate the stability and functions of proteins; for example, in metabolic enzymes, acetylation acts as an on/off switch mechanism [116], whereas methylation on carboxylate side-chains enhances hydrophobicity by increasing the affinity of proteins for phospholipids [115]. We speculate that PTMs are important mechanisms in the regulation of ENOA functions, localization and immunogenicity.

ENOA induces a specific immune response in tumors

Several TAAs induce the production of IgG autoantibody in cancer patients via an integrated immune response triggered by CD4+ T cells, CD8+ T cells and B cells. TAAs released by secretion, shedding or tumor cell lysis are captured by antigen presenting cells, processed and presented by either major histocompatibility complex (MHC) class I or MHC class II molecules for priming and activation of CD8+ and CD4+ T cells, respectively. Uptake of antigen by B cells also occurs and is driven by membrane Ig, leading to MHC class II antigen presentation to CD4+ T cells. Activated CD4 + T cells, through the secretion of appropriate cytokines, trigger B cells to produce IgG against the same TAA [117], and CD8+ T cells to differentiate into TAA-specific cytotoxic T lymphocytes. In vivo maintenance and survival of TAA-specific cytotoxic T lymphocytes is also dependent on cytokines released by CD4 +  T cells [118]. This coordinated immune response suggests that IgGs against TAA are not only a diagnostic tool, but also allow the selection of TAAs for cancer immunotherapy.

In many cancer patients, including pancreatic [119], leukemia [120,121], melanoma [104,122], head and neck [91,123,124], breast [69,125] and lung [30,69,96,99,126–129], ENOA has been shown to induce autoantibody production (Table 1). In pancreatic cancer patients, autoantibodies to ENOA are directed against two upregulated isoforms phosphorylated in Ser 419 [115,119] (Table 2). Protein phosphorylation increases the affinity of peptides for MHC molecules that can be recognized by T cells [130].

In pancreatic cancer, ENOA elicits a CD4+ and CD8+ T cell response both in vitro and in vivo [29]. Anti-MHC class I Ig inhibited the cytotoxic activity of ENOA-stimulated CD8+ T cell against pancreatic tumor cells, but no MHC class I restricted peptide of ENOA has been identified so far. Moreover, in pancreatic ductal adenocarcinoma patients, production of anti-ENOA IgG is correlated with the ability of T cells to be activated in response to the protein [29], thus confirming the induction of a T and B cell integrated antitumor activation against ENOA. In oral squamous cell carcinoma, an HLA-DR8-restricted peptide (amino acid residues 321–336) of human ENOA recognized by CD4+ T cell and able to confer cytotoxic susceptibility has been identified [131,132].

Clinical correlations

The diagnostic and prognostic value of ENOA expression and production of autoantibodies to it has been illustrated in several tumors (Table 1). In breast cancer, enhanced ENOA expression is correlated with greater tumor size, poor nodal status and a shorter disease-free interval [78]. In head and neck and nonsmall cell lung cancer, patients with high ENOA expression had significantly poorer clinical outcomes than low expressers, including shorter overall- and progression-free survival [91,99]. In hepatocellular cancer, expression of ENOA increased with tumor de-differentiation and correlated positively with venous invasion [93,94].

In breast and lung cancer patients, anti-ENOA autoantibodies are decreased in the advanced stages of the disease [69]. In pancreatic cancer, detection of autoantibodies against Ser 419 phosphorylated ENOA usefully complemented the diagnostic performance of serum CA19.9 levels up to 95%. The presence of this humoral response was also correlated with a longer progression-free survival upon gemcitabine treatment and overall survival, supporting the clinical significance of phosphorylated ENOA autoantibodies [119]. The concept that autoantibody levels can also function as markers for the diagnosis and prognosis of cancers has been extensively pursued [69,133].

Conclusions

Taken as a whole, these findings illustrate the multifunctional properties of ENOA in tumorigenesis, and its key implications in cancer proliferation, invasion and immune response. In cancer cells, ENOA is overexpressed and localizes on their surface, where it acts as a key protein in tumor metastasis, promoting cellular metabolism in anaerobic conditions and driving tumor invasion through plasminogen activation and extracellular matrix degradation. It also displays a characteristic pattern of PTMs, namely acetylation, methylation and phosphorylation, that regulate protein functions and immunogenicity. In several kinds of tumor, patients develop an integrated response of CD4+, CD8+ T cells and B cells against ENOA, together with anti-ENOA autoantibodies in their sera. Clinical correlations propose ENOA as a novel target for cancer immunotherapy. In pancreatic cancer, for example, the pancreas-specific Ser 419 phosphorylated ENOA is upregulated and induces the production of autoantibodies with diagnostic and prognostic value (Fig. 1).

Figure 1.

 Production of autoantibodies to phosphorylated ENOA in pancreatic cancer. ENOA is overexpressed in tumor cells compared with normal tissues and it is present on the surface of different cell types where it acts as a plasminogen receptor. ENOA is phosphorylated on Ser 419 only in pancreatic tissues, the overexpression of this post-translationally modified ENOA in tumor condition induces the production of autoantibodies with clinical relevance in pancreatic cancer patients.

Acknowledgements

The authors thank Dr W. Zhou for discussion on the role of post-translational modifications in the regulation of protein functions and Dr J. Iliffe who critically reviewed the manuscript. This work was supported in part by grants from the Associazione Italiana Ricerca sul Cancro (AIRC); Fondazione San Paolo (Special Project Oncology); Ministero della Salute: Progetto strategico, ISS-ACC, Progetto integrato Oncologia; Regione Piemonte: Ricerca Industriale e Sviluppo Precompetitivo (BIOPRO and ONCOPROT), Ricerca Industriale ‘Converging Technologies’ (BIOTHER), Progetti strategici su tematiche di interesse regionale o sovra regionale (IMMONC), Ricerca Sanitaria Finalizzata, Ricerca Sanitaria Applicata; Ribovax Biotechnologies (Geneva, Switzerland) and Fondazione Italiana Ricerca sul Cancro (FIRC).

Ancillary