Antibodies specific for disease-associated antigens (DAA) expressed in non-malignant diseases reveal potential new tumor-associated antigens (TAA) for immunotherapy or immunoprevention

Abstract

Immune responses to a large number of mutated and non-mutated tumor antigens have been studied in an attempt to unravel the highly complex immune response to cancer. Better understanding of both the effectors and the targets of successful immunosurveillance can inform various immunotherapeutic approaches, which can strengthen or replace natural immunosurveillance that a tumor has managed to escape. In this review we highlight targets of antibodies generated in the context of diseases other than cancer, such as asthma, allergies, autoimmune disorders, inflammation and infections, where the antibody presence correlates either with an increased or a reduced lifetime risk of cancer. We focus on their target antigens, self-molecules abnormally expressed on diseased cells or cross-reactive with exogenous antigens and found on cancer cells as tumor associated antigens (TAA). We refer to them as disease-associated antigens (DAA). We review 4 distinct categories of antibodies according to their target DAA, their origin and their reported impact on cancer risk: natural antibodies, autoantibodies, long-term memory antibodies and allergy-associated antibodies. Increased understanding and focus on their specific targets could enable a more rational choice of antigens for both therapeutic and preventative cancer vaccines and other more effective and less toxic cancer immunotherapies.

1. Introduction

Immunosurveillance and elimination of cancer is an important function of the immune system. Even when tumors manage to escape immune control, the presence of immune infiltrates in the primary tumors is associated with lower recurrence rates or longer progression-free survival, suggesting continued immunosurveillance [1]. Immunosurveillance is acting on the earliest premalignant lesions and can result in one of three potential outcomes: tumor elimination, considered to be the most frequent outcome; equilibrium, where the tumor and the immune system interact over a long period of time without evidence of clinical disease; and escape, where the tumor escapes immune control and becomes a clinical disease [2]. The ability to escape from immune recognition and control is now recognized as an important hallmark of cancer [3].

A lot that is known about anti-tumor immunity is derived from the failed phase of immunosurveillance and this knowledge has led to immunotherapies designed to help the immune system regain control of the disease, establish an equilibrium or achieve complete cancer elimination [4]. These therapies, known as checkpoint inhibitors, have shown impressive results but still in a relatively small number of patients and restricted to only a few cancer types [5–7]. Immunoprevention is another anti-cancer immune-based strategy, still very early in development, that aims to strengthen natural immunosurveillance and lower the likelihood of cancer escape [8]. In the context of infectious diseases, prevention through vaccination has had an enormous impact on improved health around the world [9]. However, with many deadly infectious diseases under control, cancer is becoming a leading cause of death with the newest and most successful therapies available only in the most affluent countries due to their high costs. A successful effort to develop vaccines for the prevention of various cancers might result in a victory similar to the one observed over infectious diseases.

The major barrier to the development of preventative cancer vaccines has been the lack of antigens that would be predictably expressed on future tumors and against which immune responses elicited through vaccination would be safe, destroying arising tumors but not normal tissues. In this review, we highlight target antigens of antibodies that are present constitutively or elicited in immunological contexts other than cancer, such as acute inflammatory states, infections, auto-immune diseases or allergies. Some of their target antigens, to which we refer to as disease-associated antigens (DAA), could be highly appropriate for the development of safe preventative cancer vaccines or therapies because they undergo transient changes in expression and post-translational modifications similar to what is observed on malignant cells where they are considered to be tumor-associated antigens (TAA).

2. Tumor antigens recognized by the immune repertoire of cancer patients

An all-out effort using state-of-the-art cellular and molecular techniques to study immune responses in cancer patients against their tumors, resulted in the discovery and characterization of hundreds of molecules of different types recognized on tumors by human T cells and antibodies [10]. The first category of molecules are the tumor-specific antigens that include mutated neoantigens unique to each tumor [11], products of oncogene mutations such as K-ras and N-ras, or gene translocations and fusions, such as BCR-ABL, shared by many tumors [12], and oncofetal antigens, shared by many tumors but not expressed in normal adult tissues [13]. Immune responses against antigens derived from shared oncogene mutations and oncofetal antigens have been found to be targets of both humoral and cellular immunity in cancer patients [14,15] and have been tested as therapeutic cancer vaccines. They are likely to also be appropriate antigens in preventative cancer vaccines as their expression in certain tumor types is highly predictable. Evidence of immunity against mutated neoantigens has been found in melanoma [16], breast [17] and lung cancers [18], however those mutations are random and unpredictable and thus not suitable for broadly applicable immunotherapies or preventative vaccines.

The second category are tumor-associated antigens (TAA), which includes differentiation antigens and overexpressed and post-translationally abnormally modified antigens. Differentiation antigens are encoded by genes with a tissue-specific expression, found therefore on specific tumor types and the corresponding healthy tissue. They include CEA [19], PSA [20], tyrosinase [21], mammoglobin A [22], gp100 [23] and melanA/MART-1 [24]. Overexpressed antigens include a large variety of molecules such as MUC1 [25], cyclin B1 [26], HER2/neu, hTERT [27], survivin [28] and mesothelin [29], which are overexpressed in many epithelial tumors due to tumor-specific increase in their gene transcription, gene amplification and/or protein stability [30]. Overexpression, and in many instances abnormal post-translational modifications of these molecules in cancer, generate predictable neoepitopes that are known to elicit immunity in cancer patients and could be candidate antigens for preventative cancer vaccines and other immunotherapeutic strategies.

3. Anti-tumor immune repertoire of healthy individuals with no history of cancer

T cells recognizing well-known TAAs have been be found in healthy individuals who never experienced cancer. For instance, a study observed that healthy HLA- A*0201- positive individuals showed a similar mean frequency of CD8+ cells recognizing a tyrosinase peptide YMDGTMSQV with melanoma patients [31]. Similarly, T cells against melanA/MART-1 were found in 8 % of healthy donors, albeit 95% were naïve compared to melanoma patients where one-third of these T cells were of the effector memory type [32,33]. Another study found that both young and old healthy individuals with no history of cancer have cyclin B1-specific memory CD4 and CD8 T cells as well as antibodies [34]. T cell responses were also found against gp100 [35] and MAGE-10 [36] but in fewer healthy individuals. The same is true for humoral immunity. Antibodies against a variety of TAAs such as MUC1 [37], HER2-neu [38] CEA [39] and Cyclin B1[34], have been reported in healthy individuals with no history of cancer. It is hypothesized that these immune responses could have been acquired in the context of other immunological conditions in which abnormal self-antigens, constitutively expressed on malignant cells as TAA, are transiently abnormally expressed on tissue affected by acute infectious diseases. Therefore, we have proposed that these antigens currently referred to as TAA should be renamed disease-associated antigens (DAA).

Antibodies binding to a variety of exogenous antigens, such as those on bacteria, viruses, and fungi, as well as self-antigens such as nucleic acids, phospholipids, erythrocytes, serum proteins, cellular components, insulin or thyroglobulin, account for a significant proportion of immunoglobulins in healthy individuals [40]. In addition, conditions such as autoimmune diseases and allergic diseases are characterized by specific sets of antibodies. Their role in anti-tumor immunity had not been postulated until recently when epidemiological evidence began to emerge of possible associations between the histories of allergies or exposure to inflammatory infectious diseases, antibody responses generated in those contexts and modulation of cancer risk [41]. Figure 1 shows different categories of antibodies circulating in cancer-free individuals that recognize DAA that have also been identified as TAA, which are aberrantly expressed in various immunological contexts.

Figure 1: Antibodies made in non-cancer settings and their association with cancer risk.

The figure specifies antibody isotypes, the known target antigens, known functions and correlations of their presence with cancer risk.

3.1. TAA recognized by natural antibodies

Traditionally natural antibodies are described as “pre-immune” antibodies that arise independently of known exposures to pathogens or deliberate immunization. The majority of natural antibodies are of IgM isotype produced by spontaneously arising CD5+ B-1 cell clones and are encoded by V(D)J genes that lack somatic mutations [42,43]. They are broadly cross-reactive and despite the low-to-moderate intrinsic affinity of their antigen-binding sites, natural IgM antibodies possess a high overall binding avidity, a feature that is responsible for a particularly effective binding of antigens with a repetitive structure [40]. While their role in health or disease remains unclear, they are thought to play a role in the early recognition of bacterial and viral infections [44]. They could therefore provide a first line defense that protects the organism during the extended time needed for specific antibody responses to be mounted [45]. Natural antibodies and other IgM antibodies have also been shown to bind to post-transcriptionally modified cell surface antigens, recognizing the conserved structures of carbohydrate epitopes [46–48]. Finally, natural antibodies that bind to self-antigens, such as Thy-1 glycoprotein [49] and phosphorylcholine [50], are thought to be initiators of auto-immune responses by providing templates for the development of pathogenic autoantibodies.

Natural antibodies have been recently investigated for their role in the recognition of oligosaccharides expressed on tumor cells. For example, IgM and IgA antibodies directed against several TAA, including MUC1 and CA125, have been found in healthy colostrum and maternal and newborn cord sera [51]. Another recent study also reported a protective role for B-1 cells through production of natural antibodies reactive to the TAA NeuGcGM3 ganglioside, in a clinical study of stomach cancer [52]. In addition, the fully germline-encoded IgM antibody PAM-1 has been found to react with a post-transcriptionally modified form of the cysteine-rich fibroblast growth factor receptor 1 (CFR-1) that is over-expressed on epithelial cancers [53].

These data suggest that natural antibodies that recognize specific TAA could be protective against cancers expressing these TAA. In support of that hypothesis, it has been found that the natural antibody SC-1 that binds to carbohydrate residues on the CD55 receptor can induce apoptosis of stomach cancer cells [54]. Another study used two-dimensional immunoblotting to detect antigens on the 4T1 mouse breast cancer cell line that were recognized by natural IgM from three different strains of healthy mice. The authors demonstrated that natural antibodies had different patterns of reactivity to tumor antigens depending on the mouse strain and that natural antibody profiles correlated with differential susceptibility to breast cancer. This last result also suggested that the genetic background of individuals can influence the reactivity of natural antibodies to tumor antigens [55]. Nevertheless, this work implies that immune responses against these antigens are safe and protective and therefore could be boosted or elicited de novo for breast cancer therapy or prevention.

3.2. TAA recognized by pathogenic autoantibodies

When the threshold for self/non-self-discrimination is altered, it can result in a spectrum of pathologies where homeostatic pathways related to cell clearance, antigen-receptor signaling or cell effector functions are disturbed. Autoimmune diseases are considered to be a result of the breakdown of central tolerance to self-antigens established in the thymus during T cell development through clonal deletion that eliminates T cells that recognize self-peptides. For B cells, this process involves receptor editing that focuses further the specificity of B cells through a secondary recombination of antibody genes [56]. When this process is impaired, self-reactive and potentially pathogenic responses can be mounted.

Pathogenic autoantibodies are continuously produced in the setting of symptomatic autoimmune diseases and are usually of high-affinity and somatically mutated IgGs or IgAs. The presence of autoantibodies in the serum is a well-accepted diagnostic marker of many auto-immune diseases including type I diabetes, thyroiditis, celiac disease and rheumatic diseases [57]. They have a wide spectrum of action, from mimicking stimulation of hormone-dependent receptors [58], blocking neural transmission, inducing altered signaling and inflammation, to triggering cell lysis and neutrophil activation [59]. Proteomic microarrays have served as a high-throughput screening method for the identification of antigens recognized by autoantibodies. Dozens of autoantigens have been characterized in autoimmune diseases such as systemic lupus erythematosus (SLE) [60], scleroderma (SSc) [61] and rheumatoid arthritis (RA) [62].

Interestingly, some of the autoantigens shared by multiple autoimmune diseases, are also considered to be TAA. For example, a protein fragment from topoisomerase I (TOP1) targeted by autoantibodies in SLE and SSc, was recently identified as a novel TAA associated with early stage esophageal squamous cell carcinoma, gastric cancer, colorectal cancer and non-small-cell lung cancer [63]. An increase in expression of the essential centromere protein-A (CENP-A), has been associated with SSc, higher grade cancers and relapse in estrogen receptor-positive breast cancer [64]. Finally, increased production of TAA CA15–3 (breast carcinoma), CA19–9 (pancreatic carcinoma) and CA125 (ovarian carcinoma), was observed in RA, SLE and SSc [65].

Whether the presence of autoantibodies directed against TAA is associated with increased or reduced risk of cancer is unclear. Recent data showed an increased risk of non-Hodgkin’s lymphoma, lung, vaginal and thyroid malignancies in patients with SLE. In the same patient population, however, decrease in breast and prostate cancer risk was linked to the presence of autoantibodies [66]. Other studies showed that patients with scleroderma and autoantibodies directed against RPC1 are at increased risk for cancer [67], but in this case a reverse scenario was suggested where prior cancer, and presumably the immune responses against it, triggered scleroderma [68].

3.3. TAA recognized by long-term memory antibodies

Antibodies produced spontaneously by cancer patients against TAA are often referred to in the literature as natural or autoantibodies. However, those are generally adaptive IgM produced by B2 cells or IgG and other isotypes produced by plasmacytes and memory B cells and therefore differ from natural antibodies described earlier. In addition, by comparing sera from patients with breast cancer versus patients with a symptomatic autoimmune disease, it was demonstrated that the antibody repertoire of breast cancer patients was different from the autoantibody repertoire found in rheumatic autoimmune diseases [69]. Together, this suggests that autoantibodies observed in cancer patients should be considered as memory antibodies produced spontaneously during an acquired response to self-antigens differentially presented by the tumor and maintained as a memory response.

The presence of memory antibodies recognizing various tumor antigens in a large number of healthy individuals with no history of cancer have been a puzzle for cancer immunologists. We describe below different immunological contexts that could lead to their production and we discuss the potential relevance of these antibodies and of their target antigens to health, diseases and cancer.

3.3.1. TAA on premalignant lesions

It is assumed that many tissues harbor at some point beginnings of cancer, but that most premalignant cells are readily cleared by the immune system preventing clinical diagnosis of cancer. Even if not completely eliminated, some premalignant lesions can also persist for decades without symptoms and under efficacious immune control [2]. Therefore, antibodies to TAA in healthy individuals could represent active immunosurveillance against dormant precancerous lesions that are difficult to detect by standard means, or a successful memory response to previously eliminated lesions. For example, p53 mutations or overexpression are early events in the progression of tobacco-related oral malignancies and antibodies directed against p53 were observed in 30 % of patients with premalignant lesions [70]. Similarly, antibodies against cyclin B1 were found in the sera of 4.8% of patients with benign prostatic hyperplasia, a premalignant precursor of prostate cancer [71], and also in patients with benign lung diseases, especially heavy smokers [72]. Circulating MUC1-specific IgGs have been found in some patients with intraductal papillary mucinous neoplasms [73]. Finally, in a mouse model, authors showed that low levels of IgG antibodies against the SV40 T-antigen (derived from an oncogenic virus) were detectable in the serum of LoxP-Tag mice without macroscopically visible tumors [74].

The presence of these memory antibodies could confer protective immunity against future lesions and also decrease the risk of progression toward a more advanced stage of malignancy. Supporting this idea, it was shown that patients with asymptomatic monoclonal gammopathy, a precursor state of multiple myeloma (MM), who had antibodies specific for the MM tumor antigen SOX2, had a reduced risk for progression to multiple myeloma [75]. These antibodies are of interest because they could serve as biomarkers for early detection of cancer that cannot be seen otherwise [76]. More importantly, however, target antigens of these antibodies could be especially appropriate to include in vaccines to further boost and maintain this apparently safe and protective immunity.

3.3.2. TAA induced by infectious inflammatory events

There is a very large body of literature in the field of cancer epidemiology that shows that strong febrile infections, especially those that characterize common childhood diseases, can profoundly reduce life-time risk of various cancers. For example, case–control studies in lymphoma, stomach, colorectal, breast, and ovarian cancer found that childhood diseases such as chicken pox and pertussis, as well as repeated cold and influenza infections throughout life, significantly decreased life-time risk for these cancers [77,78]. In addition, a large case–control study in patients with melanoma from six European countries showed reduction of melanoma risk with increasing numbers of febrile viral infections experienced early in life [79,80].

This very strong protection has not yet been fully supported by any specific mechanism with the exception of studies showing the association between a history of febrile infections, the presence of antibodies against known TAA in healthy patients and their risk of developing cancer. Cramer et al. showed that the likelihood of having anti-MUC1 antibodies increased from 24% for women with 0 or 1 inflammatory conditions to 51% for women with five or more conditions, such as mastitis, pelvic surgery and intrauterine device use [81]. The same study showed that a decrease in ovarian cancer risk correlated with the number of inflammatory conditions. These results were then confirmed by a prospective nested case-control study [82]. Finally, it was found that women experiencing mumps early in life had higher serum levels of anti-MUC1 which was associated with a drastically reduced risk of ovarian cancer later in life [37,41].

These observations in humans have been modeled in mice. In C57BL6 mice, a repeated infection with the influenza virus yielded numerous post-infection IgG antibodies and T cells directed against antigens on a mouse lung tumor cell line 3LL and protection from 3LL challenge 3 months post last influenza infection. These antigens were found to be non-mutated self-molecules constitutively overexpressed on 3LL cells and transiently overexpressed on infected mouse lungs and therefore can be considered as DAA [83]. Inflammation that accompanies infections and early carcinogenesis is likely to be the main inducer of changes of expression in many self-antigens on infected cells. It was recently confirmed that mimicking acute inflammation in vitro was sufficient to induce the transient expression of TAAs EEF2, SerpinB3 and NAMPT in normal mammary epithelial cells (Jacqueline et al, in review).

Vaccination with peptides derived from these antigens transiently abnormally expressed on infected tissues and then later re-expressed on malignant tissues as TAA could be the best and safest promising strategy to reduce the risk of developing cancer [41]. Indeed, in the study with influenza mentioned above, it was shown that vaccination based on a cocktail of DAA was able to protect mice from tumor challenge [83].

MUC1, as a prototype DAA that is also a TAA, has already been tested for immunogenicity and safety in preclinical mouse models as well as in therapeutic and a preventative vaccine trials. In human MUC1 transgenic mice, MUC1 vaccine was able to stimulate a cell-mediated immune response, to decrease inflammation and abnoral MUC1 expression in tumors, which together demonstrated its efficiency as a prevention strategy for breast cancer [84] and colon cancer [85,86]. A preventive vaccine based on the MUC1 antigen has also been tested in a premalignant setting in individuals at high risk for colon cancer. This vaccine elicited strong immunity in 43% of the participants without any side-effects and a long-term memory measured by high antibody responses to a booster injection at one year [87]. Later, it was shown that these antibodies were highly tumor specific [88].

The success of the immunoprevention based on MUC1 supports the idea that DAA shared by infected cells and cancer cells should be investigated for their potential in successful immunosurvellance and immunoprevention. A global preventive strategy could be envisioned where the lack of DAA-specific immune memory due to limited early exposure to infectious inflammatory events could be compensated by a vaccine composed of a cocktail of DAA [89]. As these DAA are expressed in other conditions such as viral infections or inflammatory diseases, it could also open a new opportunity for the development of vaccines that confer “universal” protection.

3.3.3. TAA as molecular mimicry: the microbiota

There is some evidence that antibodies generated against antigens presented by virus and bacteria can cross-react with some self-antigens, a phenomenon known as “molecular mimicry.” [90]. Such cross-reactivities have been documented in autoimmune diseases and associated with an increased risk of developing the disease [91]. For example, anti-dsDNA antibodies were shown to bind to the sequence ARVLWRATH from human cytochrome B561 and to the sequence RAGTDEGFG from one of the transcriptional regulators from Burkholderia sp. suggesting the involvement of molecular mimicry in SLE-related diseases [92]. Another example is glomerulonephritis where autoantibodies generated against a sequence of amino acids of the bacterial adhesin FimH (present in Gram-negative enteropathogens) also recognize the human lysosomal-associated membrane protein (hLAMP-2)[93].

Molecular mimicry is increasingly investigated in the context of cancer and whether pre-existing immunity to microbial peptides could eliminate tumor cells that express a cross-reactive TAA is still an open question. Significant homologies have been found between amino acid sequences found in some TAA and in the microbiome [94]. For instance, the peptide GLEREGFTF that arises from a mutation in CSMD1, highly expressed in melanoma, has 80% similarity to a known Burkholderia pseudomallei antigen [95]. Another tumor antigen MAGE-A6 was found to contain an epitope which was highly homologous with a peptide within Mycoplasma penetrans HF-2 permease (MPHF2) protein [96]. Finally, a number of viral and bacterial peptides have been shown to have a potential cross-reactivity with the tumor antigen Melan-A [97].

In this context, the antibody profile of healthy individuals may also be profoundly shaped by their microbiome. Especially, the reactivity of IgAs that coat the intestinal barrier and are principally involved in the maintenance of microbiota homeostasis [98]. A recent study showed that intestinal IgAs are natural polyreactive antibodies with innate-like recognition properties that can help the host adapt to the dynamic arrays of antigens encountered at mucosal surfaces [99]. They also found that antibodies with microbiota-specificity can recognize antigens such as cardiolipin that has been associated with mitochondrial dysfunction following KRAS mutation [100]. Although the cross-reactivity of these polyreactive IgA with other well-known TAA has not been investigated, such mechanism could be particularly relevant in the context of colorectal and other gastrointestinal cancers.

Even though most of the studies showed tumor-promoting effects of the microbiota, antitumor effects have also been observed and gut microbiota is now recognized to play an important role in shaping systemic immune response and in immunotherapeutic interventions (see [101]). For instance, some of the bacteria, such as Akkermansia muciniphila, Bacteroides fragilis, Bifidobacterium spp. and Faecalibacterium spp. have been associated with anticancer immune responses in both preclinical tumor models and in patients with cancer [102]. Finally, the adoptive transfer of B. fragilis-specific CD4+ T cells was found to reduce the growth of MCA205 fibrosarcomas [103]. It is tempting to postulate that a cross-reactive, (tumor) antigen-specific immune response is one of the mechanisms of this protection. If confirmed, these antigens would make an exceptionally important candidates for therapeutic or preventative vaccines as well as targets for adoptive immunotherapy.

3.4. TAA recognized by allergy-associated IgE

Allergic diseases, such as allergic rhinitis, eczema, hives, asthma, and food allergies, are the most common chronic health conditions in the world. Allergy symptoms can range from mild to serious, life-threatening allergic reactions (anaphylaxis). Allergy and autoimmunity result from dysregulation of the immune system. Whereas autoimmune diseases are related to the Th1 pathway, allergies are classically Th2-driven. Exposure of allergic individuals to exogenous allergens leads to immediate inflammation caused by degranulation of mast cells via IgE-allergen immune complexes and the release of inflammatory mediators, proteases and pro-inflammatory cytokines [104]. In developed countries, ~35% of adult population has increased levels of IgE specific for common environmental allergens [105]. To our knowledge, only one study has investigated the antibody profile of cancer-free patients for the presence of IgE recognizing tumor antigens. Zennaro et al. recently described a microarray approach to examine immune responses to tumor antigens in populations with cancer or allergy and found low levels of IgE recognizing the TAA EGFR and Her2/neu in the sera of allergic patients [106].

The paradoxical relationship between cancer and allergies has interested epidemiology, oncology and immunology researchers for several decades. Even though detrimental effects of asthma are now well supported in lung and bladder cancer [107], IgEs have not been implicated. The main hypothesis is that the increased risk of cancer is caused by tissue remodeling and inflammation-induced DNA damage on cells [108]. In contrast, several pieces of evidence support a strong inverse associations between IgE and pancreatic, lung and colorectal cancer, glioma and hematological malignancies [109–111]. For instance, serum-levels of IgE have been inversely associated with melanoma risk in men and women combined, and with the risk of breast and gynecological cancers combined [112]. Furthermore, in a clinical study higher levels of polyclonal IgE in non-allergic individuals were directly correlated with lower disease incidence and better survival from multiple myeloma [113]. Similarly, the survival of mice with engrafted mammary tumors was positively associated with the pre-tumor challenge levels of IgE [114].

Numerous studies have been conducted to understand the mechanisms behind the protective effect of IgE [115]. The main hypothesis to explain the impact of IgE-mediated allergies on the occurrence of cancer proposes that individuals with allergy have a general hyper-responsiveness of the immune system, which could lead to decreased cancer risk due to the production of tumor-specific IgE and the mobilization of antitumor effectors such as eosinophils, basophils, and mast cells [108,116]. Accordingly, it has been shown that IgE directed against the folate receptor, a TAA over-expressed in ovarian and stomach cancer, is able to increase tumor infiltration of activated macrophages and to reduce lung metastases [117]. In an hFcεRI transgenic mouse model of breast carcinoma, administration of anti-hMUC1 IgE activated allergic effector cells that led to 25–30% reduction in tumor growth [118]. Lastly, an engineered Her-2/neu specific IgE was able to trigger in vitro degranulation of basophils expressing human FcεRI in the presence of murine mammary carcinoma cells that express human Her-2/neu [119]. In the context of prevention, known TAA-specific IgEs have been elicited by vaccination and shown to be extremely efficient as anti-tumor responses [120–122]. Therefore, it may be particularly relevant to further characterize IgE repertoires found in individuals with allergies and search for IgEs directed against TAA to better understand the association between allergies and cancer risk and develop new therapeutic and preventative strategies.

4. Targeting DAA/TAA for cancer therapy

Targeted therapies are the focus of much research in oncology. The development of biological therapies, such as monoclonal antibodies (mAbs) and T cells that target specific tumor antigens, has given hope for improvement of survival in many cancers [123]. Anti-cancer antibodies function through various mechanisms including antibody-dependent cellular cytotoxicity (ADCC), phagocytosis (ADCP) or complement-independent cytotoxicity (CIC), which can lead to promotion of new immune responses and long-term memory that will prevent cancer recurrence. In Figure 2, we show some of the well-known TAA that should be considered as DAA. Many of these TAA have been used as therapeutic targets on cancer cells. Trastuzumab was the first FDA-approved mAb to target Her-2/neu, a TAA over-expressed in 30% of breast cancers and in other cancers including lung, ovarian and prostate cancer. Four large randomized trials clearly showed that trastuzumab had a major effect in reducing recurrence and death in patients with early breast cancer [124–126]. A phase I clinical trial in patients with advanced cancer showed efficacy and safety of PankoMab-GEX, a glyco-optimized humanized IgG1, with high affinity for a tumor-specific glycopeptide epitope of MUC1 (TA-MUC1) [127]. We propose that the repertoire of antibodies generated in other diseases but targeting the already known DAA described above or yet to be identified, should be investigated for their safety and anti-tumor activity and developed as therapeutic mAbs for administration to cancer patients.

Figure 2:

Tumor associated antigens (TAA) that are also disease associated antigens (DAA) on non-malignant tissues.

The feasibility of adoptive T cell therapy (ACT) based on T cells targeting TAA is already being tested. A study showed that bone-marrow derived T cells stimulated with dendritic cells pulsed with peptides from Her-2/neu and MUC1 can be safely used in the treatment of breast cancer [128]. In recent years, ACT of T cells expressing a chimeric antigenic receptor (CAR) have received much attention. The extracellular antigen recognition region of CARs is composed of a single‐ chain antibody fragment derived from a monoclonal antibody and can therefore be tuned to recognize specific TAA. For example, the monoclonal antibody TAB004 recognizing the aberrantly glycosylated tumor form of MUC1 was used to generated MUC1-CAR T cells that demonstrated target-specific cytotoxicity in an in vitro model of triple-negative breast cancer [129]. A strong target-specific cytotoxic function of MUC1-CAR T cells was also observed in a xenograft model of head and neck carcinoma [130], non-small-cell lung cancer [131] and leukemia and pancreatic cancer [132]. Similarly, the therapeutic efficacy of anti-HER2 CAR T cells was found to have potent cytotoxicity in an in vitro model of glioblastoma [133]. The efficacy and the specificity of such CAR T cells was also demonstrated in vivo in melanoma xenografts grown in NOG mice [134]. Finally, it was recently reported that using HER2-CAR T cells in combination with chemotherapy was safe and associated with clinical benefit in a patient with HER2+ sarcoma [135]. Another study showed a delay in tumor progression after delivery of T cells genetically engineered to target the TAA NY-ESO-1 in a xenograft model of neuroblastoma [136]. Results obtained in CEA-transgenic mice suggested that CEA-specific CAR T cells may be effective in patients with CEA+ tumors [137]. Jiang et al. also showed that anti-mesothelin CAR T cells were able to suppress tumor growth in pancreatic carcinoma patient-derived xenograft models [138]. On the other hand, a high-affinity T cell receptor targeting MART-1 induced high anti-tumor efficacy but was associated with severe toxicity [139].

Thus, the identification of new DAA and antibodies that target them could add to the TAA repertoire on cancer cells and foster the development of more effective and highly safe immunotherapies directed against them.