The tumor microenvironment (TME) is a complex structure comprising immune cells, stromal cells, blood vessels, and extracellular matrix [1]. Immune cells play crucial roles in the TME and are typically categorized as adaptive and innate immune cells. Dendritic cells (DCs) are key intermediates proficient in antigen presentation, bridging the gap between innate and adaptive immunity [2]. DCs express receptors capable of recognizing diverse danger signals, including pathogens and altered cells, such as tumor cells [3]. Following antigen capture, activated DCs serve as specialized antigen-presenting cells [4]. They process both self and non-self antigens, subsequently presenting them to naïve T lymphocytes. These T lymphocytes, in turn, initiate antigen-specific immune responses while concurrently regulating tolerance and immunity [4].

The efficacy of the anti-tumor immune response relies on the cross-presentation of tumor-derived antigens by DCs to T cells, resulting in a predominant T cell-mediated cellular immune response [3, 5]. However, DCs infiltrating the TME display a heterogeneous nature characterized by variations in surface markers, migration patterns, localization, and cytokine production [6, 7]. Furthermore, distinct TME conditions can exert an influence on the effector functions of DCs, alter their phenotypic characteristics, and induce dysfunction and tolerance [8]. For example, studies have demonstrated that tumor-infiltrating DCs exhibit decreased expression of co-stimulatory molecules such as CD86 and CD80 [9], while concurrently displaying heightened expression of immune inhibitory molecules such as programmed cell death 1 ligand 1 (PD-L1) [10]. Therefore, comprehending the diversity of tumor-infiltrating DCs is crucial for the development of improved strategies for cancer immunotherapy.

Traditional bulk genomic and transcriptome analyses average signals across diverse cell groups, thereby hindering the identification of specific cell types and states [11]. Single-cell sequencing, however, offers the capacity to reveal transcriptomic cellular heterogeneity at a single-cell resolution, thereby exposing subpopulation structures that may remain indistinct in bulk RNA sequencing [11, 12]. The heterogeneity of DCs at the single-cell scale has been extensively explored in recent reviews [13]. In this context, we provide a succinct summary of DC subpopulations within the TME, elucidating their functions in various TME contexts, their roles in tumorigenesis and development, and their significance in ongoing anti-tumor therapies. This review centers on their tumor-related immune responses or pathways and their potential utility as predictive markers for therapeutic targeting.

Cancer is characterized by its inherent heterogeneity and the complex composition of the TME [1]. Both tumor heterogeneity and the TME play crucial roles in tumor initiation, progression, invasion, metastasis, and drug resistance [14]. Bulk RNA sequencing technology primarily reveals an average gene expression profile within a sample, which poses challenges in comprehending tumor heterogeneity and the TME [11, 12]. The emergence and advent of single-cell sequencing technology have provided an opportunity to deconstruct the TME by discerning discrete cellular subpopulations, thus facilitating a more profound understanding of the intricate TME [11, 12]. In contrast to bulk sequencing, single-cell sequencing offers several distinct advantages, including its capacity to characterize cell subtypes and their relative frequencies within a sample, identify actively expressed genes within individual cells or cell types, and investigate communication between cells or cell types [15]. Recent advancements in single-cell sequencing techniques have undergone rapid development, with various applications and a primary focus on single-cell RNA sequencing (scRNA-seq). The critical determinant of success in single-cell studies lies in the preparation of high-quality single-cell suspensions. The process of single-cell suspension preparation encompasses density centrifugation for blood samples and mechanical enzymatic dissociation for solid tissues. Specific enzymes or mixtures are employed to facilitate effective cell separation, followed by DNase I treatment to minimize clumping. The choice of enzymes employed in various tumor models may exhibit slight variations depending on the tissue type. While Type IV collagenase is the standard choice in most scenarios, specific tissues such as the pancreas and intestine necessitate the utilization of alternative enzymes such as collagenase P and collagenase I [16]. In recent times, mixed enzyme products, such as Miltenyi Biotec's gentleMACS™ Dissociator, have become the preferred choice in the field and are frequently utilized in cancer studies for the preparation of single-cell suspensions [17]. This product exhibits effectiveness in dissociating tissues from various human and mouse tumor models, following meticulously designed procedures tailored to each tumor type. In summary, the acquisition of single-cell suspensions from diverse tumor models has evolved into a straightforward process. Subsequently, these suspensions are filtered through a mesh or strainer prior to single-cell capture. Short processing times are imperative to prevent gene expression variation and protect sensitive cells from damage. Alternatively, nuclear RNA sequencing is employed to alleviate biases stemming from cell type composition, particularly advantageous for intricate tissues such as interconnected adult neuronal tissues. This approach proves optimal for delicate cell types, such as differentiated neurons, providing valuable insights into their gene expression profiles, all without necessitating the isolation of intact cells [16]. In accordance with their experimental designs, researchers may find it necessary to augment or deplete specific cell types to increase the overall count of cells of interest in the final sequencing library. For instance, the analysis of specific immune responses may mandate the enrichment of immune cells, whereas cancer-related investigations may entail the exclusion of immune cells to boost the overall count of tumor cells.

Extensive transcriptomic information can be acquired through high-throughput scRNA-seq technology. Various downstream analysis tools facilitate the examination of both intra- and inter-tumor heterogeneity, mechanisms underlying tumor invasion and metastasis, TME characteristics, and the design of future treatment strategies. Corresponding bioinformatics methods have advanced to accommodate the complexities of scRNA-seq data, which are characterized by high dimensionality and the expression of numerous genes in each cell. Dimensionality reduction and clustering techniques empower researchers to categorize DCs into subpopulations with enhanced precision, thereby providing insights into the heterogeneity of traditional subtypes [18]. DCs exhibit intricate and diverse origins and developmental trajectories. Pseudotime trajectory analysis offers a means to elucidate the evolutionary progression of cells through gradual changes in gene expression. It can be employed to track cell lineage as well as to investigate the origins and differentiation of DCs [15]. Cellular communication through ligand–receptor interactions is linked to tumor progression in the TME [19]. Multiple analytical tools based on scRNA-seq data have the potential to reveal previously unexplored cellular receptor–ligand interactions critical for identifying prospective therapeutic targets [19]. The correlation among immune scores, prognosis, and responses to diverse treatments has been established [20, 21]. scRNA-seq provides an unprecedented level of resolution in characterizing infiltrating immune cells compared to conventional immune scoring methodologies, thereby enhancing the precision of prognosis and predictions for immunotherapy responses [22]. In addition, single-cell technology provides intricate details pertaining to individual cells across various dimensions. For example, the Cellular Indexing of Transcriptomes and Epitopes by Sequencing method enables simultaneous unbiased transcriptional profiling and antibody-based detection of protein markers in thousands of cells [23]. Single-cell analysis encompasses methylation patterns, histone modifications, chromatin accessibility, and T cell receptor repertoires, contributing valuable insights to cancer research from diverse perspectives [24,25,26,27,28]. The emergence of spatial transcriptomics allows for the simultaneous acquisition of cellular transcriptome data and information regarding cell locations, furnishing spatially informative datasets for TME investigations and addressing previous limitations in single-cell sequencing [11, 29, 30]. The TME consists of various cell types that frequently participate in well-organized spatial interactions [29, 30]. Deciphering this intricate spatial architecture enables us to grasp the mechanisms through which tumor cells communicate with each other, evade immune surveillance, and contribute to cancer progression. Therefore, investigating gene expression in a spatial framework offers a holistic comprehension of tumor initiation and facilitates the development of efficacious therapeutic strategies. These robust methodologies can assist in elucidating the heterogeneity of tumor-infiltrating DCs, thus offering comprehensive insights into cancer immunology research.

DCs represent a diverse group of immune cells, categorized into distinct subsets based on various criteria, including ontogeny, phenotypic characteristics, tissue distribution, and transcriptional profiles [6, 7]. DCs can be categorized into conventional or classical DCs (cDCs), which encompass type I cDCs (cDC1s) and type II cDCs (cDC2s), plasmacytoid DCs (pDCs), monocyte-derived DCs (moDCs), and LAMP3+ DCs. cDC1s excel in intracellular antigen processing and presentation, playing a crucial role in orchestrating anti-tumor immune responses. Their mechanism involves the cross-presentation of tumor-associated antigens to CD8+ T lymphocytes, which recognize these antigens through major histocompatibility complex (MHC) class I signaling [31]. Conversely, cDC2s efficiently present antigens associated with MHC II to CD4+ T cells, thereby promoting various T-helper (Th) cell responses, such as Th1, Th2, and Th17 cell polarization [31]. pDCs are major producers of type I interferons (IFNs) and are primarily involved in antiviral and antitumor immune responses [32]. moDCs represent a distinct subset that undergoes differentiation in response to inflammatory signals and is recruited to sites of inflammation, including the TME [33]. LAMP3+ DCs have been identified at the single-cell level and are distinguished by their immunoregulatory properties and migratory characteristics [10, 34].

Tumor-infiltrating DC states have been delineated through scRNA-seq across various human malignancies, encompassing breast cancer [35,36,37], hepatocellular carcinoma (HCC) [24, 34, 38, 39], colorectal cancer (CRC) [40, 41], non-small cell lung cancer (NSCLC) [9, 42,43,44,45,46], nasopharyngeal carcinoma (NPC) [47,48,49], esophageal squamous cell carcinoma [50, 51], glioma [52], cervical cancer [53, 54], gallbladder carcinoma [55], ovarian cancer [56,57,58], oral cancer [59], and gastric cancer (GC) [20, 60,61,62,63]. Pan-cancer analysis has indicated an enrichment of LAMP3+ DCs and pDCs in tumors, with both normal tissues and tumors demonstrating a comparable abundance of cDC2s and cDC1s. Among tumor tissues, cDC2s predominate [10, 64, 65]. The abundance of LAMP3+ DCs exhibits significant variability across different cancer types [10, 64, 65]. In various human malignancies, the transcriptional profiles and frequency of cDC1s are associated with improved survival rates and enhanced responsiveness to treatment [66, 67]. However, cDC2 exhibits heterogeneity, playing roles in both anti-tumor responses and tolerance processes within various TMEs [40, 68]. LAMP3+ DCs exhibit characteristics associated with both anti-tumor immunity and tolerance [10, 34]. In addition, the abundance and function of DCs display pronounced heterogeneity in the TME at different stages, underscoring their pivotal role in tumor immunity or tolerance [63,