The dark matter in cancer immunology: beyond the visible– unveiling multiomics pathways to breakthrough therapies

Abstract

Cancer cells orchestrate the surrounding tumor microenvironment (TME) to strike a fine balance between tissue regeneration providing them with nutrients, and tissue destruction triggered by immunogenic alarm signals. At steady state, the tenuous balance favors cancer growth. Therapies aimed at enhancing the immunogenic properties of cancer cells or the reacting immune responses can, however, revert the equilibrium to clear the host of cancer. The understanding of factors that affect this balance is progressing rapidly due to advances in high throughput technologies disclosing from previously uncharted territories new biologies referred to as “dark matter”. These advances are critical for the understanding of the true mechanisms leading to immune-mediated cancer rejection.

This review focuses on cancer genetic, epigenetic and metabolic derangements that approximate those caused by intra-cellular pathogen infection, a phenomenon referred to as “viral mimicry” (VM) and other aspects of cancer/host cells interactions unexplored in the past that enhance the VM effects. On the cancer side, VM prompts alterations of cancer cell metabolism leading to the generation of aberrant cellular products recognized as foreign by the host’s immune system. The latter are defined as “dark matter” to emphasize the powerful effects exerted by these obscure bioproducts on the TME as the mass of invisible particles can dictate the rotational period of galaxies. On the other side, a myriad of previously unappreciated factors can influence the host responses. Thus, here we propose an extended definition of dark matter beyond the limits of cancer cell-intrinsic biology, to a broader interpretation encompassing elements that influence the cellular networks within the TME.

Introduction

When Fritz Zwicky postulated and Vera Rubin demonstrated the existence of invisible particles in the universe, hence defined as “dark matter”, to explain why the rotational period of galaxies is faster than predicted by the mass of visible matter, little was known about similar mysterious forces within our cells that could help defeat a scourge closer to home.

In fact, for decades, cancer research has focused largely on genetic mutations, particularly within coding regions, signaling networks, and protein patterns involved in tumor development. However, a significant amount of tumor biology has remained unexplained by these traditional approaches. The term dark matter was originally coined to describe the discovery that the majority of non-ribosomal, non-mitochondrial-RNA arises from long, intergenic transcribed regions that could be involved in neoplastic transformation [1, 2] and it was subsequently extended to “cancer dark matter” to describe this underexplored realm comprising subtle immunological signals, obscure regulatory molecules, and complex microenvironmental interactions that conventional techniques overlook [3]. Although the term dark matter, strictly refers to cancer cell-specific alteration, in this review, however, we extend the concept of dark matter to other uncharted territories where the function of host cells within the TME may be also controlled by unconventional regulatory mechanisms. The TME is the critical battleground between cancer and its host [4], a complex ecosystem where cancer cells interact with immune and stromal cells, and the extracellular matrix [5]. It is where cancer cell survival or demise results from a multicellular crosstalk with other factors mediated by cell-to-cell interactions or secreted factors [6, 7]. This intricate interplay determines whether the immune system can effectively eliminate the tumor or whether the tumor will evade immune surveillance and progress. The TME dark matter encompasses a variety of factors including metabolic pathways, non-coding RNAs, post-translational modifications, and physical properties beyond those directly depending upon cancer cell intrinsic behavior [8] This review aims to shed light on the dark matter of cancer, exploring the multiomics features associated with it and their potential implications for cancer immunotherapy [9].

This revolution applies to immunogenic mechanisms leading to cancer eradication that have been previously obscured by the prevalent attention given to other elements such as germline or somatic alterations, enhanced mutational burden, mutated protein recognized as neo-antigens, immunogenic cell death (ICD) and other circumstantial factors, representing a phenomenology of relatively limited significance [10].

Cancer is a genetic ailment whose intricate nature stems from genetic and epigenetic aberrations, causing a dynamically heterogeneous disease that remains a leading cause of death worldwide [11]. The complexity arises from the uniqueness of each tumor influenced by interactions within the TME with patient-specific factors like germline predisposition [12, 13], lifestyle, environmental exposures and co-existing morbidities [10, 14]. Thus, cancer cell genetic instability and evolutionary heterogeneity [7, 15] together with its interactions with benign cells in the TME significantly influence cancer progression and therapeutic resistance to various modalities including those covered by the emerging realm of immune oncology (IO) [7, 16].

IO has revolutionized cancer treatment shifting the focus from targeting cancer cells directly to mobilizing the host’s immune system against them [4, 17–19]. However, these indirect approaches are limited by their erratic tumor specificity, dependency on the presence and functional status of the immune cells in the TME and the dominance of the targeted mechanism of other immune suppressive factors [7, 20]. Since only a subset of cancer patients respond to IO treatments, understanding the underlying molecular mechanisms of tumor immunogenicity and compensatory immune resistance [20] is crucial [21, 22]. Thus, the study of cancer dark matter focuses on hidden factors including subtle epigenetic modifications, non-canonical protein expression, microbial influences, and ionic imbalances play critical roles in inducing cancer cell-intrinsic immunogenicity, immune evasion, therapeutic resistance, and disease relapse. Traditional studies have characterized immune checkpoints, tumor associated antigens (TAA), and cytokine networks; yet many facets of the immune interface remain hidden. For example, subpopulations of immune cells within the TME may secrete atypical cytokines or express non-canonical markers that dampen immune activity. These subtle yet critical determinants of immune dysfunction comprise part of the cancer dark matter that current models miss.

Cancer in itself is not a threat to our species since it prevails past the reproductive cycle. Therefore, immune reactions against cancer are not an evolutionary prerequisite representing rather an epiphenomenon occasionally triggered when cancer cell biology mimics infection against which the immune system is directed to protect the survival of the specie [23]. Thus, cancer cell-specific derangements that approximate those caused by intra-cellular pathogen infection referred to as “viral mimicry” (VM) predominantly dependent of the expression of aberrant forms and amounts of nucleic acids and production of aberrant peptides offer a better mechanistic explanation for cancer-intrinsic immunogenicity. VM should not be confused with the concept of immunogenic cell death (ICD) that, as later discussed) implies the implies cancer cell death followed by the release of additional immunogenic factors specific related to the process of cellular decay that vary according to the various mechanisms leading to ICD [23–25]. However, since VM can cause demise of stressed cancer cells, the two phenomena can overlap, particularly when therapeutics that can enhance VM lead to cancer cell death.

Current Understanding of cancer immunogenicity and the host’s reaction

Before entering the unexplored world of the dark matter, it is important, however, to clarify the current understanding of the determinism of cancer immunogenicity and the epiphenomena associated with it [10].

Cancer cell intrinsic immunogenicity drives the infiltration of immune effector cells in the TME, in particular CD8-expressing T cells [19]. Immune-infiltrated cancers are characterized by the expression of a transcriptional signature termed the immunologic constant of rejection (ICR) that indicates the presence of cytotoxic T cells polarized toward an effector TH1 phenotype [4, 26–29]. These cancers are prone to respond to anti-cancer immunotherapy resulting in their complete eradication [4, 7, 19, 20, 30–46].

The presence of T cells and the ICR is driven primarily by VM, which is a constitutive, non-physiological, cancer cell-privileged status triggered by endogenous stimuli [47]. The ICR represents the host’s secondary reaction to cancer cell-intrinsic VM that leads to the activation of interferon-stimulated genes (ISGs) and generates chemo-attractive and proinflammatory signals responsible for the recruitment and activation of T cells, just as accidents attract the emergency response system and not vice versa. Several mechanisms can induce or amplify VM including (1) the epigenetic dysregulation of the transcription of endogenous retroviruses (ERVs) that results in the accumulation of double-stranded RNA (dsRNA) [47] and (2) DNA damage repair defects and/or mitochondrial stress that respectively lead to the release of nuclear double-stranded DNA (dsDNA) or mitochondrial double-stranded DNA (mt-dsDNA) [48].

The constitutive expression of ISGs by cancer cell lines expanded in vitro without exposure to extrinsic interferon-stimulating signals demonstrates that VM is a phenomenon intrinsic to cancer cell biology (unpublished observation). Therefore, pharmacological enhancement of VM is cancer-specific because VM does not occur normally in benign non-pathogen infected tissues [47]. Similarly, cancer dark matter also referred to as “epitranscriptome” is a cancer-specific phenomenon triggering the aberrant production of short protein products absent in benign tissues that are highly immunogenic because that are not expressed by benign cells and, therefore, escape central tolerance during thymic selection [3, 49].

VM and dark matter do not automatically imply cellular death and, therefore, should be distinguished from ICD [23, 24]. However, pharmacological enhancement of VM can induce cancer cell death [50–52]. Thus, VM-inducing or enhancing agents can amplify the anti-tumor immune response by adding mechanisms related to ICD such as the release of TAAs. At the same time, ICD can further enhance the effect of VM and of the dark matter by the release by dying cancer cells of damage associated molecular patterns (DAMPs) [23, 24].

The mechanisms inducing VM and ICD are conserved across different cancer types independent of their ontogeny and affect about 2/3 of neoplasms across indications [46]. Thus, targeting ISGs-expressing tumors can effectively treat a large proportion of cancers.

Dark matter in cancer immunology

Strictly speaking, cancer dark matter refers to a newly discovered level of control over cancer gene activity that is not related to DNA mutations, but rather to changes in gene expression and function dependent upon epigenetic alterations. These changes can contribute to cancer’s evolution without affecting the DNA sequence itself and include aberrations in regulatory elements, untranslated regions, splice sites, non-coding RNA and synonymous mutations [3, 53]. Thus, current interest focuses not only on the coding regions of the genome, but also genomic and proteomic profiles related to the dark matter [3], including noncoding sequences and noncanonical proteins [54]. These elements named non-canonical open reading frames (ncORFs) or transcribed ultra-conserved regions (T-UCRs) play a significant role in cell proliferation, resistance to therapy and other oncogenic processes [55]. T-UCRs are long non-coding RNAs (lncRNAs) transcribed from DNA and are highly conserved across mammalian genomes. ncORFs are translated genomic regions beside the traditionally defined, canonical protein-coding sequences, that encode functional peptides sequences. ncORFs can be in various positions, including 1) upstream of canonical coding sequences (uORFs), 2) within the 5’ or 3’ untranslated regions (UTRs) of mRNAs, 3) in lncRNAs and 4) in circular RNAs (circRNAs).

Advanced technologies such as whole genome next-generation sequencing (NGS) enable the exploration of non-coding elements of the genome like ncORFs and other regulatory variants that influence tumor progression. The ncORFs are crucial elements in cellular regulation due to their ability to regulate signaling pathways [56]. For instance, ncORFs modulate cell growth and suppress the capability of cytotoxic T lymphocytes to recognize tumor cells [57]. CRISPR/Cas9 analysis of ncORFs provided more in-depth information regarding their role in cancer progression [58]. Moreover, proteins derived from ncORFs are being investigated in IO as tumor-specific neoantigens and represent promising therapeutic targets for their high degree of cancer specificity.

Multiomics technologies in Cancer research

The acquisition of malignant traits requires cancer cells to undergo extensive metabolic reprogramming, enabling them to proliferate uncontrollably evading normal cellular regulation mechanisms [59]. Cancer cells adapt their metabolic programs to support increased proliferation, survival, and metastasis, often creating a microenvironment characterized by nutrient depletion, hypoxia, and acidosis [60]. Genomics provides insights into the genetic mutations and structural variations that drive cancer development and influence immune responses [61]. Transcriptomics elucidates the gene expression patterns that determine cellular phenotypes and responses to immunotherapy. Proteomics identifies the proteins expressed and modified in cancer cells and immune cells. Metabolomics characterizes the metabolic landscape of the TME, also revealing metabolic vulnerabilities that can be exploited for therapy [62]. By integrating these data, researchers can gain a holistic understanding of the mechanisms that govern cancer-immune interactions enabling the identification of novel biomarkers, therapeutic targets, and strategies to overcome resistance to immunotherapy [63, 64]. Such approaches have been extensively discussed for several cancers, and we refer the readers to several recent cancer-specific reviews including breast [65], esophageal [66] colon [67], gastric [68], and lung [69] as a few examples.

Genomic landscape of the TME

Genomic analyses have revealed a wide range of genetic mutations, copy number variations, and structural rearrangements in both cancer cells and surrounding stromal and immune cells [70]. These alterations including also traits related to genomic instability, DNA repair pathways, including nucleotide excision repair, mismatch repair, and homologous recombination can influence various aspects of cancer biology, including cell growth, survival, metastasis, and response to therapy [71–73]. Several studies have shown the germline and somatic alteration of cancer determine at least in part the immune landscape of the TME and, most specifically, the presence of absence of immune effector mechanisms [12, 19, 27–29, 74]. However, most studies have demonstrated that factors beyond the somatic patterns of cancer cells contribute both to cancer immune landscapes and its predisposition to response to immunotherapy treatment. These include germline predisposition, and influence of external factors such as the microbiome, and individual patients behavioral traits such as nutritional habits and anamnestic conditions while other factors such as mutational burden and the expression of mutated protein products play a relatively minor role [10, 13, 29, 75].

Epigenetics and immune modulation

Epigenetic modifications, encompassing aberrant DNA methylation, histone acetylation, and microRNA expression, orchestrate a complex symphony of gene expression regulation, profoundly influencing immune cell function within the TME [76]. Tumor cells exploit epigenetic mechanisms to silence tumor suppressor genes, activate oncogenes, and evade immune recognition [77]. This epigenetic reprogramming extends to immune cells infiltrating the tumor, where alterations in DNA methylation patterns, histone modifications, and microRNA expression can impair their effector functions and promote immune tolerance. For instance, epigenetic silencing of genes involved in antigen presentation such as those encoding major histocompatibility complex (MHC) proteins can enable cancer cells to escape T-cell–mediated destruction [78]. Similarly, the expression of immune checkpoints such as programmed death-ligand 1 (PD-L1), expressed on the surface of tumor cells and its ligand programmed cell death-1 (PD-1) expressed by T cells can be regulated by several mechanisms including histone modification and remodeling, microRNAs, long noncoding RNAs, and post-translational modifications (PTM) [79]. Epigenetic silencing of genes encoding for chemokines and cytokines can disrupt immune cell trafficking and recruitment to the tumor site, while aberrant expression of microRNAs can suppress the expression of key immune signaling molecules. DNA methylation patterns can influence the expression of genes involved in immune cell differentiation, activation, and function. For instance, methylation of the promoter region of interferon (IFN)-γ can suppress its expression, impairing T cell-mediated anti-tumor immunity. Histone modifications, such as acetylation and methylation, can also regulate gene expression and impact immune cell function. This has been particularly well described with adoptive T cell products, whose function and level of differentiation is strongly influenced by epigenetic alterations [80, 81], Moreover, epigenetic modification have been shown to affect resident T cell metabolism and function [82, 83]. Histone modifications regulate a variety of processes, including DNA replication, DNA damage repair, and transcription. Histone acetylation generally promotes gene transcription, while histone methylation can have either activating or repressive effects depending on the specific histone residue modified. These modifications can affect the TME and the response of cancer to immunotherapy [84, 85]. MicroRNAs, small non-coding RNA molecules, play also a crucial role in regulating gene expression by binding to mRNAs and inhibiting their translation. Aberrant microRNA expression has been implicated in various cancers and can affect immune cell function by targeting genes involved in immune signaling, activation, and differentiation [86, 87].

Alterations in DNA methylation are broadly observed in cancer resulting in disturbed gene expression and they can be modified by external factors such as diet [88]. Ancient endogenous retroviruses normally repressed in the mammalian genome are reactivated by VM mostly in response to global DNA hypomethylation [89]. In particular, intergenic regions are primarily affected by genomic instability., Hypermethylation of the promoters of tumour suppressor genes often accompanies the global reduction in DNA methylation. This paradoxical methylation of CpG islands in regulatory regions of canonical genes as well as TE leads to transcriptional-silencing [89–91] and suppression of gene expression [92] related to the oncogenic process. The use of inhibitors of DNA methylation, such as azacitidine and decitabine, can reverse epigenetic silencing and restore the expression of tumor suppressor genes [90]. Histone deacetylase inhibitors can also modulate gene expression and enhance anti-tumor immunity.

The integration of RNA sequencing (RNA-seq) and epigenetic chromatin immune-precipitation sequencing (ChIP-seq) revealed how non-coding transcripts contribute to neoplastic transformation and development. Jen et al. [93] demonstrated that the transcription factor Oct4 regulates lncRNAs like NEAT1 and MALAT1 by directly binding to the regulatory region of these lncRNAs. Luciferase reporter assays, quantitative polymerase chain reaction (q-PCR) and ChIP-PCR experiments validated Oct4’s role in enhancing NEAT1 and MALAT1 expression. The functional consequences of this interactions revealed that NEAT1 or MALAT1 overexpression promotes lung cancer cell proliferation, migration, and invasion, whereas their knockdown leads to opposite effects.

Aberrant transcription may affect various components of the dark matter such as T-UCRs [55, 94]. T-UCRs are conserved regions of the genome that are transcribed but do not encode proteins. These regions play key regulatory roles, especially in processes related to genomic stability and gene expression regulation. In tumors, many T-UCRs have been found to be deregulated, causing disruption of cellular homeostasis [95]. In some adenocarcinomas, deregulation of specific T-UCRs contributes to the survival and migration of tumor cells [96] and they have been also involved in modulation of responsiveness to various therapeutic modalities [97]. T-UCRs regulate gene expression by inhibiting or activating signaling pathways that influence the cell cycle and apoptosis. ncORFs erratically produce non-canonical proteins or peptides specifically in tumor tissues [98]. ncORFs present in brain tumors such as glioblastoma can modulate the cellular response to stress and induce resistance to chemotherapeutic agents [99]. Recent studies have revealed that many ncORFs act as regulators of processes involved in angiogenesis and cell cycle [56]. Although ncORFs result from epigenetic dysregulation rather than somatic mutations, their expression may generate “neoantigens” because they product are not expressed by normal benign tissues throughout the host’s life cycle. These neoantigens offer new opportunities for personalized immunotherapies, as they are recognized as foreign by the patient’s immune system [98]. Thus, neoantigens derived from ncORFs and T-UCRs is a significant area of development [100]. These neoantigens are specific to tumor cells providing optimal targets for immunotherapy (Fig. 1) [101].

Fig. 1.

Non-canonical peptide recognition by anti-tumor immunity and its application in cancer vaccines and CAR T cell therapy. Tumor cells generate non-canonical peptides that together with MHC class I are recognized by dendritic cells (DCs) and subsequently presented to T cells, therefore initiating an adaptive immune response. In the lymph nodes, DCs activate naïve T cells leading to their differentiation into effector T cells capable of targeting and eliminating tumor cells. Simultaneously, T helper cells activate B cells, then promoting antibody production against tumor antigens. Identified non-canonical peptides can be exploited for therapeutic applications, including cancer vaccine development (top right), toward the stimulation of a targeted immune response. Additionally, non-canonical peptides can be used in CAR T cell engineering (bottom right) to generate modified T cells with enhanced tumor recognition and cytotoxicity. Overall, the integration of non-canonical peptides into cancer vaccines and CAR T cell therapies may represent a promising approach for both improving anti-tumor immunity and overcoming immune evasion mechanisms

Transcriptional profiling

Transcriptional profiling allows for comprehensive analysis of gene expression, alternative splicing, and RNA editing, providing a holistic view of the tumor’s molecular landscape revolutionizing for instance the understanding of immune checkpoint expression patterns in both cancer and immune cells [102] and their biological relevance [20, 46]. By analyzing the transcriptomic profiles of tumor samples, researchers can identify potential targets for immunotherapies such as checkpoint blockade and predict patient responses [30, 103]. Transcriptional profiling of immune cells infiltrating the TME can reveal their activation status, functional orientation, and potential for exhaustion [4, 26, 104–107]. This approach allows for the identification of therapeutic targets and the development of personalized immunotherapies [108]. Single-cell RNA sequencing provides a high-resolution view of gene expression in individual cells, allowing researchers to identify rare cell populations and understand cellular heterogeneity within the TME [109]. By analyzing the transcriptomic profiles of individual cells, researchers can gain insights into the mechanisms of immune evasion [110]. Furthermore, the identification of aberrantly expressed genes can substantially increase the proportion of patients whose care can be individualized based on molecular information [111].

Proteomics and the hidden cancer immunome

Proteomics examines simultaneously multiple protein expression patterns defining how biological systems respond to distinct stimuli [112, 113] providing insights into signaling pathways concomitantly activated or suppressed in cancer and immune cells that are not detected by molecular biology methods [114, 115] providing unprecedented views of protein interactions [116]. However, the integration of proteomics with genomics and transcriptomics, can provide a more complete understanding of the signaling process enacting cancer biology and immune responses [117] generating entire patterns of information [118]. Quantitative proteomics refines the picture by identifying differences in individual samples through relative and absolute quantification of proteins and peptides [115, 119].

The intricate interplay between the epigenome and proteome constitutes a major leap in our understanding of cancer’s root causes, extending beyond the conventional mutation-centric view. The functional consequences of epigenetic alterations are most pronounced at the proteome level. Modern proteomics, powered by mass spectrometry and data-independent acquisition technologies, has begun to catalog how epigenetic modifications correlate with altered protein expression or PTMs. Table 1 summarizes some key epigenetic markers with their corresponding proteomic signatures and roles in tumorigenesis.

Table 1.

Key epigenetic markers with their corresponding proteomic signatures and roles in tumorigenesis

Epigenetic Marker	Associated Proteomic Signature	Role in Tumorigenesis
H3K27ac	Upregulation of growth-promoting transcription factors	Enhances oncogene expression and tumor aggressiveness
H3K4me3	Activation of proteins involved in cell cycle progression	Correlates with active transcription and proliferation
DNA Methylation	Suppression of tumor suppressor proteins	Leads to gene silencing and malignant transformation

Mass spectrometry enables the fragmentation of peptides, revealing amino acid sequences and allowing detailed analysis of PTMs unique in different tumors [120]. Proteomic analysis, including mass spectrometry, has expanded the potential to detect unconventional proteins [121] and its the integration of proteomics and genomics, referred to as proteo-genomics, allows the correlation of genomic mutations with protein expression [122] and the identification of peptide sequences that may escape conventional methods, such as those produced by ncORFs [123]. High-resolution mass spectrometry is an essential tool for identifying neoantigens derived from ncORFs and non-coding regions, thus providing a comprehensive mapping of peptides presented by MHC-class I molecules on the surface of tumor cells [124, 125]. These peptides, which are not expressed by normal cells, can activate T lymphocytes, making them ideal targets for personalized vaccines and CAR-T therapies [126].

High-resolution imaging mass spectrometry, combined with transcriptomics, can identify specific lipid signatures associated with breast cancer invasiveness [127]. Lipid metabolism can regulate ncORF expression and influence the function of ncORF-derived microproteins.

Several factors contribute to the difficulty of annotating noncanonical peptides. First, their typically short length makes detection challenging. Second, ncORF translation often follows unconventional mechanisms, sometimes initiating from alternative start codons. Third, many noncanonical peptides lack strict conservation across species, which is crucial for determining biological relevance. Identifying peptide-coding ncORFs requires a combination of computational predictions, ribosome profiling, and mass spectrometry. Furthermore, validating and characterizing the resulting micro-peptides primarily relies on proteomic approaches to elucidate their functional roles [128]. For instance, this proteogenomic approach successfully characterized the TME of lung cancer assessing of the different compositions of immune cells and neoantigens with prognostic implications [129].

Proteomics offers a snapshot of protein expression and modifications that shape cellular behavior far beyond what transcriptomics can reveal. In cancer, PTMs such as phosphorylation, ubiquitination, and glycosylation often play a critical role in modulating immune responses. High-resolution proteomic profiling uncovers heterogeneity in protein networks governing immune checkpoints and the extracellular matrix. These unexpected variations can lead to altered interactions between tumor cells and immune effector cells potentially creating niches where cancer cells prosper. Such findings imply that a subset of the proteome, heretofore largely unstudied, constitutes part of the immunologically relevant cancer dark matter [5, 130–132].

While MS is a powerful tool for identifying and quantifying compounds based on their mass-to-charge ratio, it can face challenges when analyzing complex samples due to interference and limited sensitivity. Tandem mass spectrometry (MS/MS) overcomes these limitations by introducing an additional stage of fragmentation, where selected precursor ions are broken down and the resulting products are analyzed [133]. This process significantly improves the ability to differentiate structurally similar compounds, enhances sensitivity in detecting trace amounts, and provides valuable structural information for more accurate identification. These features make MS/MS indispensable when single-stage MS is insufficient for detecting low-abundance analytes in complex matrices [134].

Electrospray Ionization mass spectrometry (ESI-MS) offers even greater capability to explore cancer dark matter, including elusive metabolites [135]. While MS/MS is indispensable for obtaining high-resolution structural information, ESI-MS stands out for its exceptional sensitivity in detecting polar and fragile biomolecules [136]. A major advantage of ESI-MS lies in its ability to ionize biomolecules in a soft manner, preserving their native structure and enabling the identification of extremely low-abundance species. Moreover, when coupled with MS/MS and sophisticated computational approaches such as competitive fragmentation modeling, ESI-MS significantly improves the ability to predict and characterize previously unannotated metabolites and peptides, expanding the current landscape of molecular oncology [137]. By incorporating ESI-MS into multi-omics strategies, scientists can shed light on previously hidden aspects of tumor biology, ultimately transforming the dark matter of cancer into actionable insights for diagnostics and therapeutic development [138].

Sequential windowed acquisition of all theoretical mass spectra (SWATH-MS) combines a highly specific data-independent acquisition method with a novel targeted data extraction strategy to mine the resulting fragment ion datasets. SWATH-MS has been widely used to compare protein expression and PTMs [139] enabling, for instance, the identification of key proteomic signatures induced by anti-androgen therapy in prostate cancer [140].

Sequence-based proteomics allows for quantitative analysis of protein expression and PTMs, providing a more in-depth understanding of the role of proteins in various biological processes and diseases. Aptamer based proteomics is also used to explore many diseases states [141]. Aptamers are chemically synthesized small, single-stranded nucleic acid molecules that can bind to specific target molecules with high affinity and specificity, similar to antibodies. They are known for their unique three-dimensional structures that enable them to interact with their targets. Aptamer-based proteomics can accurately capture and quantify proteins. Analyzing the proteomic profiles of saliva samples can identify potential protein biomarkers of disease [142]. Antibody based proteomics are popular due to high throughput and ability to multiplex. Using ESI-MS, 1,381 proteins were identified [143]. Free flow electrophoresis with linear ion trap MS/MS identified 437 proteins in whole human saliva [142]. Data-independent acquisition and proteogenomics are also emerging technologies that are used to characterize the proteome [141]. All these emerging technology will play a revolutionary role in providing mechanistic insights about cell-to-cell interactions in the TME.

Ionomics and metabolic imprints on immunity

Ionomics is the comprehensive analysis of elemental and trace mineral concentrations in biological tissues. Emerging evidence suggests that trace elements play critical roles in modulating immune cell activity and influencing the TME and affecting the natural or therapeutic history of cancer [144–146]. Similarly, metabolic derangement affecting amino acid metabolism can affect the TME through metabolic reprogramming [147]. Disruption of ion homeostasis within immune cells can impair their ability to effectively target and eliminate cancer cells. Elements such as zinc, copper, and iron are critically involved in immune cell function and the activity of enzymes that regulate both epigenetic modifications and protein stability. Perturbations in ion homeostasis have been linked to changes in the TME [148]. Recent investigations suggest that ionomic imbalances may serve as both markers and mediators of the immunologically active but yet uncharacterized components of cancer dark matter. Integrating ionomic data with other omics could, therefore, provide valuable insights into novel mechanisms of tumor resistance and relapse.

Metabolomics and lipidomics

Metabolic reprogramming is a hallmark of cancer, allowing tumor cells to meet their bioenergetic and biosynthetic demands for rapid proliferation and survival [149]. Cancer cells often exhibit increased glycolysis, even in the presence of oxygen, and altered glutamine metabolism, fatty acid metabolism, and mitochondrial function [9, 150, 151] (Bi et al., 2018). These metabolic alterations not only fuel cancer cell growth but also influence immune cell function within the TME [152–154]. The Warburg effect, where cancer cells preferentially use glycolysis over oxidative phosphorylation, is a well-known example of metabolic reprogramming in cancer [155]. Comprehensive analysis of the metabolomics and transcriptomics uncovered dysregulated networks in cancer [156]. Additional layers of biological complexity reside in the realms of lipidomics that may affect various aspect of cancer cell behavior [157]. Metabolomic profiling can unveil the signatures of immune cell activation and metabolic stress within tumors, while lipidomic analyses help decipher the composition of cellular membranes that govern immune receptor function and signal transduction [158, 159]. Together, these emerging fields enrich our understanding of cancer biology by capturing biochemical details that conventional genomic analyses miss. The integration of these diverse omics into a comprehensive systems biology approach is essential for mapping the full spectrum of cancer dark matter, which in turn may reveal the elusive factors driving tumor relapse and treatment failure.

Microbiomics

The gut is colonized by a vast number of microorganisms, including bacteria and fungi [160]. These microorganisms perform a range of essential functions, including nutrient metabolism, pathogen defense, and the modulation of immune responses [161]. This ecosystem is generally referred to as microbiome and the discipline studying it referred to as metagenomics [162, 163].

The human microbiome, encompassing both gut and tumor-resident microbial communities, has emerged as a critical regulator of immune responses and a key determinant of cancer development and treatment outcomes [164–167] including the efficacy of checkpoint inhibitor therapy (CPI) by modulating systemic inflammation and immune cell activation [10, 165, 168]. It has also been demonstrated that the microbiota can regulate the metabolism of certain cancer treatments, thereby influencing the therapeutic response and the onset of related side effects [169].

Certain gut bacterial species can enhance anti-tumor immunity by modulating the composition and function of immune cells in the TME, while others can promote tumor growth and metastasis [29, 170]. Moreover, experimental alterations of the microbiome can modify responsiveness to immunotherapy [164] This crosstalk between microbiomics and the host immune system adds another layer to cancer’s dark matter, revealing that bacteria and their by-products may quietly orchestrate aspects of immune evasion and therapeutic resistance. We recently reported that Intratumoral presence of specific bacteria affect the immune landscape of the TME and has profound prognostic and therapeutic implications [29].

The complex interplay between the gut microbiome and the host immune system involves a multifaceted dialogue mediated by microbial metabolites, immune cell signaling, and epigenetic modifications [171]. The gut microbiome can also affect the metabolism of chemotherapeutic drugs, influencing their efficacy and toxicity [172]. Dysbiosis can also cause the translocation of bacteria [173]). Specific bacterial species, such as Bifidobacterium and Akkermansia muciniphila, have been associated with improved responses to anti-PD-1 therapy, while others, such as Fusobacterium nucleatum, have been linked to immune evasion and resistance [174]. Intratumoral presence of Ruminococcus bromii has been strongly associated with poor prognosis and altered TME in several cancers [29] while its fecal presence with reduced response to CPI [10, 175].

The homeostasis of gut microbiota also affects the efficacy and side effects of common cancer treatments, such as chemotherapy, radiotherapy, and immunotherapy [176]. This dysbiosis affects oncogenesis, tumor progression, and response to cancer therapy [177]. By producing a variety of metabolites, including short-chain fatty acids, the gut microbiome can modulate immune cell function and influence the TME [178]. These metabolites can have both pro-inflammatory and anti-inflammatory effects, depending on the specific metabolite and the context of the tumor [179].

Modulating the gut microbiome through dietary interventions, fecal microbiota transplantation, or the administration of probiotics represents a promising approach to enhance anti-tumor immunity and improve cancer treatment outcomes [180, 181]. The colonic microbiome is responsible for co-metabolism of dietary and environmental compounds that enter the colon, and the resultant metabolites produced can have either carcinogenic or tumor suppressive properties as well as regulating microbial growth and diversity [182]. Gut metabolites can exert direct effects not only in the intestine but also to modulate the function of cells in remote organs [178].

The study of microbial communities has undergone significant advancements, starting from the initial use of 16 S rRNA sequencing to the adoption of shotgun metagenomics. However, a new era has emerged with the advent of long-read sequencing, which offers substantial improvements over its predecessor, short-read sequencing because allow the accrual of continuous genomic information allowing more accurate identification of quantification of bacterial species [183]. Recently new bioinformatics techniques coupled with de novo assembly-based metatranscriptomic analysis on fecal samples from 29 patients with non-small cell lung cancer undergoing CPI therapy could predict therapy outcomes and progression free survival with higher accuracy than standard methods [75]. These reports suggests that the field of metagenomics is still rapidly evolving to provide a better assessment of the role of the microbiota as a contributor of cancer dark matter.

Methodological integration: from genomics to proteomics

The integration of multiomics data provides a holistic perspective on the complex interplay between cancer cells and the immune system [156, 158, 184] requiring robust bioinformatics tools and statistical methods to handle the complexity of the data [185–188]. By combining genomics, transcriptomics, proteomics, and metabolomics data, researchers can gain deeper insights into the mechanisms driving cancer development and immune evasion [189]. Analyzing multiple omics data from the same sample helps researchers understand how changes in the genome can affect the transcriptome, proteome, and metabolome [190] highlighting the interaction between biological layers that could be responsible for causing complex phenotypes or diseases [191]. This approach allows for the identification of novel biomarkers and therapeutic targets that may not be apparent when analyzing individual omics datasets [192].

The integration of multi-omics data requires sophisticated computational approaches to handle the high dimensionality and complexity of the data [193]. Graph convolutional networks have emerged as a powerful tool for integrating multi-omics data, allowing researchers to identify complex relationships between different omics layers and improve patient classification [194]. Deep learning algorithms can identify complex patterns in multi-omics data and predict clinical outcomes. Integrating single-cell RNA sequencing and spatial data provides insights into exploring diverse tissue architectures and interactions in a complicated network [102] improving survival prediction by applying complementary and consensus principles [195].

Artificial intelligence (AI) and machine learning (ML) are transforming the ability to analyze multiomics data, allowing efficient and precise interpretation of large and complex datasets [196]. Machine learning algorithms and deep neural networks, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), enhance the discovery of patterns and correlations between genomic, transcriptomic, proteomic, and immunopeptidome data [197]. These models are particularly useful for identifying ncORFs and non-canonical neoantigens in tumor tissues, as well as for forecasting immunogenicity [198] by predicting which neoantigen(s) are most likely to stimulate an effective immune response [199, 200]. The integration of artificial intelligence algorithms into genomic analysis is crucial for recognizing rare sequences that often are undetected by traditional methodologies. Artificial intelligence based proteomic analysis has also allowed the identification of unconventional proteins in pancreatic cancers, paving the way for new therapeutic options [201].

AI and ML algorithms can delve into the intricate details of tumor genomics, proteomics, and other molecular profiles, providing a deeper understanding of cancer biology [202] driving advancements in precision oncology [196, 203] AI and ML are revolutionizing cancer research by enabling the integration and analysis of vast amounts of multi-omics data [204–206] to predict how patients might respond to different treatment options, paving the way for personalized treatments that maximize efficacy while minimizing adverse effects [207]. AI-driven models have improved the accuracy of disease prediction and facilitated the identification of genetic loci associated with diseases [208] and have also been used to identify potential drug targets [204]. AI algorithms can analyze the complex interactions between the tumor and the immune system, identifying potential therapeutic targets and predicting patient response to immunotherapy [209–211]. By identifying and targeting key regulators within the dark matter of cancer immunology, researchers can develop novel immunotherapies that overcome immune resistance and improve patient outcomes [212]. The integration of AI into screening methods, such as biopsy slide examination, can also enhance treatment success rates [213]. Continuous monitoring using AI-powered systems guarantees continued surveillance for at-risk patients, making it easier to quickly detect changes in health markers [214]. However, the effectiveness of AI tools in cancer treatment relies on their accessibility and comprehensibility to biologists, oncologists, and researchers [215, 216].

Targeting dark matter: therapeutic strategies

While deeper and holistic understanding of our extended concept of cancer dark matter will provide a higher likelihood to achieve breakthrough therapeutic success, we want to focus now on current approaches related to cancer cell specific dark matter that are already investigated in the clinics.

Both, ncORFs and T-UCRs offer new diagnostic and prognostic opportunities [217] (Table 2). Altered expression of ncORFs is associated with an aggressive tumor phenotype in various cancers including colorectal and pancreatic carcinoma, suggesting that these elements can be used to assess the likelihood of recurrence and disease progression [56]. In particular, monitoring ncORFs and T-UCRs through liquid biopsies from blood or urine [218] is suitable for tumor progression monitoring [219] providing real-time assessment of the response to treatments and enabling clinicians to promptly adjust therapies [220], as they are non-invasive and easier to perform than tissue biopsies [221].

Table 2.

Advanced techniques for the identification and characterization of non-canonical peptides in cancer research

Mass Spectrometry (MS)	Identifies and quantifies peptides based on their mass-to-charge ratio (m/z).
Tandem Mass Spectrometry (MS/MS)	Provides structural information about peptides by fragmenting precursor ions and analyzing product ions.
High-Resolution Mass Spectrometry (HRMS)	Enhances sensitivity and accuracy in detecting low-abundance peptides, including tumor-specific neoantigens.
Electrospray Ionization Mass Spectrometry (ESI-MS)	Soft ionization technique that preserves the native structure of biomolecules, improving detection of fragile peptides.
SWATH-MS (Sequential Window Acquisition of All Theoretical Mass Spectra)	Data-independent acquisition method that enables comprehensive proteomic analysis, useful for detecting cancer-associated peptides.
Immunopeptidomics	Analyzes peptides presented by HLA molecules to identify tumor-specific neoantigens for immunotherapy.
Ribosome Profiling (Ribo-seq)	Identifies actively translated peptides, including those derived from non-canonical open reading frames (ncORFs).
Competitive Fragmentation Modeling (CFM)	Uses computational modeling to predict and characterize unannotated peptides and metabolites.
Proteogenomics	Integrates proteomic and genomic data to correlate peptide expression with genetic alterations in cancer.

Immunotherapy leverages the immune system to identify and eliminate malignant cells. Among different approaches vaccines against TAA offer, at least in theory, a better therapeutic index as they can be recognized by T lymphocytes as cancer-specific targets, minimizing the risk of autoimmunity [222]. Moreover, CRISPR/Cas9 technology has been used to directly modulate the expression of oncogenic ncORFs or activate tumor-suppressive T-UCRs [56] which in turn may enhance the immunogenic potential of cancer cells [47].

Beyond the ability to modulate tumor immunogenicity by identifying and manipulating ncORFs and T-UCRs for diagnostic, prognostic and monitoring purposes [223, 224], and their modulation to enhance the efficacy of immunotherapy improving antigen specificity and minimizing off-target effects [23, 24], pharmacological enhancement of VM and expression of dark matter may sensitize tumors to CPI [3]. In addition, enhancement of VM can increase the immunogenic properties of cancer cells through production of cytokines and chemokines to exert and innate pro-inflammatory milieu that fosters the recruitment and activation of effector immune cells [7].

Future perspectives

Relapse remains one of the most vexing challenges in oncology. Despite initial responses, many cancers recur due to the survival of subpopulations that harbor hidden vulnerabilities elements of cancer dark matter. Multi-omics research suggests that these resistant cells may evade detection not only through classic genetic mutations but also via epigenetic reprogramming, proteomic shifts, metabolic alterations, and changes in trace element levels. By uncovering these hidden dimensions, researchers can develop more nuanced therapeutic strategies. Combination therapies that simultaneously target genetic drivers, epigenetic regulators, proteomic networks, and metabolic circuits promise to shut down tumor escape routes and prevent relapse.

A significant impediment is the scarcity of comprehensive and well-annotated multi-omics datasets, which are essential for the construction and validation of robust predictive models. The limited availability of high-quality data hampers the identification of reliable biomarkers and therapeutic targets, hindering the development of effective personalized cancer treatments. The lack of standardized protocols for data collection, processing, and analysis further exacerbates the challenges in integrating and comparing data across different studies. Addressing the challenges of data sharing and standardization is crucial for advancing the field of AI in cancer immunology and accelerating the development of new therapies. Moreover, the inherent interpretability of AI models presents a substantial hurdle, as understanding the biological rationale behind model predictions is crucial for clinical translation [225]. Future of cancer dark matter immunology will likely depend on the integration of big data sets and AI [226].

The integration of AI into healthcare requires careful consideration of ethical and regulatory issues to ensure patient safety and data privacy [227–229]. The development of explainable AI methods is essential to unravel the decision-making processes of AI models and gain insights into the underlying biology of cancer immunology. Overcoming these technical and ethical challenges is crucial for realizing the full potential of AI in cancer immunology and translating research findings into clinical practice [230, 231]. The evidence presented herein reinforces the notion that cancer’s “dark matter” extends far beyond the repertoire of somatic mutations.

Conclusions

The integration of multiomics has revolutionized the study of dark matter leading to the discovery of unconventional elements such as ncORFs and T-UCRs. These components play essential roles in tumor regulation, influencing processes such as cell proliferation, drug resistance, and cancer invasiveness [223]. The use of advanced technologies like mass spectrometry and machine learning has allowed the identification of complex biomarkers and TAs, significantly improving the ability to predict therapeutic responses. In particular, liquid biopsy enables real-time monitoring of molecular variations, hence offering less invasive and highly sensitive diagnostic tools, thus promoting therapy personalization [232]. Immunotherapeutic approaches, including therapies based on neoantigens derived from ncORFs, may represent a promising frontier for the treatment of refractory tumors [3]. These therapies show significant potential to improve efficacy along with reduced side effects respect to conventional therapies. The future evolution of multiomics technologies and interdisciplinary data analysis, integrated with active collaborations among oncologists, molecular biologists, and bioinformatics, will further boost the precision and effectiveness of cancer treatments.

The immunological component of cancer dark matter represents an emerging frontier with profound implications for understanding the root causes of malignancy and relapse. The integration of multiomics data—ranging from epigenetics and proteomics to microbiomics, ionomics, and beyond—offers a holistic view of tumor biology far more complex than unveiled by traditional methods. These underexplored domains hold the key to deciphering how immune escape develops and persists, ultimately driving therapeutic resistance and cancer recurrence. As research continues to peel back these hidden layers, the hope is that tailored interventions targeting multiple omic dimensions will lead to more durable treatments and better patient outcomes.

The table summarizes key methodologies used to detect and analyze non-canonical peptides, including mass spectrometry-based approaches such as high-resolution mass spectrometry (HRMS) and tandem mass spectrometry (MS/MS), which enable precise peptide identification and quantification. Immunopeptidomics plays a crucial role in identifying tumor-specific peptides presented by HLA molecules, facilitating the discovery of novel targets for immunotherapy. Ribosome profiling (Ribo-seq) allows for the detection of peptides translated from non-canonical open reading frames (ncORFs), expanding the repertoire of potential tumor-associated antigens. Computational techniques, including competitive fragmentation modeling (CFM) and proteogenomics, integrate multi-omic data to predict and validate unannotated peptides, contributing to the development of innovative cancer immunotherapies such as personalized cancer vaccines and engineered CAR T cells.

Abbreviations

AI

Artificial Intelligence

circRNA

Circular RNA

ChIP-seq

Chromatin Immunoprecipitation sequencing

CPI

Checkpoint inhibitor Therapy

DAMP

Damage-Associated Molecular Pattern

dsDNA

Double-Stranded DNA

dsRNA

Double-Stranded RNA

ERV

Endogenous retrovirus

ESI-MS

Electrospray Ionization Mass Spectrometry

ICD

Immunogenic Cell Death

ICR

Immunologic Constant of Rejection

IFN

Interferon

ISG

Interferon-Stimulated Genes

IO

Immune Oncology

lncRNA

Long non-coding RNA

MHC

Major Histocompatibility Complex

ML

Machine Learning

MS/MS

Tandem Mass Spectrometry

ncORF

Non-canonical open reading frame

NGS

Next-generation sequencing

PD-1

Programmed cell Death-1

PD-L1

Programmed Death-Ligand 1

PTM

Post-Translational Modification

qPCR

Quantitative Polymerase Chain Reaction

RNAseq

RNA Sequencing

SWATH-MS

Sequential Windowed Acquisition of all Theoretical Mass Spectra

TAA

Tumor-associated antigen

TME

Tumor Microenvironment

T-UCR

Transcribed ultra-conserved region

uORF

Upstream of canonical coding sequences ORF

UTRs

Untranslated regions

VM

Viral mimicry

Author contributions

Writing, review and editing: All authors contributed equally. All authors have read and agreed to the published version of the manuscript.

Funding

Fondazione AIRC (IG n. 27386) and Ministero dell’Università e Ricerca (Prin 2022 202282CMEA; Prin 2022 PNRR P2022MALRP) supported RL. RL and MM were supported by: (1) The Next Generation EU - project Tech4You - n. ECS0000009; (2) The National Plan for NRRP Complementary Investments - project n. PNC0000003 ‐ AdvaNced Technologies for Human‐centrEd Medicine (ANTHEM); (3) The Next Generation EU - Project Age-It: “Ageing Well in an Ageing Society” [DM 1557 11.10.2022]; (4) POS RADIOAMICA project funded by the Italian Minister of Health (CUP: H53C22000650006); (5) POS CAL.HUB.RIA project funded by the Italian Minister of Health (CUP H53C22000800006); (6) BAC PNRR UniMI - Advances in Extracellular Vesicles membrane fusion processes for enhanced loading of genetic materials ADEVEGLIO (DR 3926/06.06.2024).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

Rosamaria Lappano is Associate Editor. Joyce Hu is Managing Editor. Franco Marincola is Chief Editor. Marcello Maggiolini is Section Editor.