Homeobox A10 in Gastrointestinal malignancies: Unraveling metastatic mechanisms and novel therapeutic opportunities

Abstract

Homeobox (HOX) genes are essential regulators of embryonic development and cellular differentiation under physiological conditions. Among the HOX gene family, HOXA10 has emerged as a critical player in gastrointestinal (GI) cancers, affecting tumor development, metastasis, disease severity, and treatment resistance. HOXA10 is a transcription factor that acts as a key regulator in embryonic development and immunomodulation. Recent research has investigated the role of HOXA10 in shaping the tumor immune microenvironment by modulating immune cell recruitment and signaling pathways that favor tumor immune evasion. In fact, our recent study identified a HOXA10-regulated 5-gene signature to classify long- and short-term survivors of pancreatic cancer, and HOXA10 expression was associated with T_reg enrichment. HOXA10 regulates genes/pathways of macrophages, regulatory T cells (T_regs), and other immune cells, thereby potentially fostering an immunosuppressive niche, increasing metastasis, and reducing the efficacy of immunotherapies. Here, we review and describe the contributions and relevance of distinct HOXA10 functions in GI cancers. We conducted a comparative analysis of HOXA10 and other HOX family proteins, highlighting their overlapping and mutually exclusive functions in GI malignancies. This review aims to provide a comprehensive overview of the role of HOXA10 in shaping the tumor and its microenvironment. We describe in detail the molecular mechanisms underlying HOXA10’s oncogenic role by integrating findings from cell lines, patient tumors, and clinical data. We identify current knowledge gaps regarding HOXA10’s role in cancer pathology and investigate approaches to target HOXA10 to improve treatment outcomes, emphasizing the potential synergistic effects of HOXA10 inhibition and immune therapies. These studies identify vulnerabilities that can be targeted by developing novel drugs and combination therapies in GI malignancies.

INTRODUCTION

Gastrointestinal (GI) cancers have some of the highest incidence rates and lowest 5-year survival rates among major cancers. These cancers encompass those of the esophagus, stomach, colorectum, liver, and pancreas (1). Due to late symptom presentation, lack of efficient diagnostics, and the highly metastatic nature of these malignancies, they are often not detected until late-stage disease. Thus, there is a dire need for improved diagnosis and treatment strategies.

The past 30 years have advanced the study of biologically relevant markers in multiple malignancies. Specifically, there has been an interest in identifying and understanding the molecular mechanisms of genes and pathways involved in disease onset and progression. Proto-oncogenes, genes that acquire oncogenic function through mutation, are among the most prominently studied of these genes. Oncogenes such as KRAS, p53, and others have attracted attention as promising drug targets or biomarkers, with impressive clinical success (2, 3). Proto-oncogenes are also recognized as the first regulatory factor preceding the development of cancer cells (4) as they encode regulatory elements for cell growth and differentiation, and genetic or epigenetic manipulation may lead to either their overexpression or underexpression. One such group of molecules, the HOX gene family, is widely recognized as a proto-oncogene (5) and has generated significant interest over the last 20 years.

In this review, we discuss the functional implications of HOXA10, a member of the HOX gene family, in solid GI malignancies. Specifically, we highlight the association between HOXA10 expression and immune modulation and explore how this information can be used to target HOXA10 within the tumor immune microenvironment. To the best of our knowledge, this is the first review to bridge the gap in our understanding of HOXA10 in the tumor-immune microenvironment of GI cancers.

HOX GENE CLUSTER ORGANIZATION

The HOX gene family consists of 39 homeobox (Hox)--containing genes that encode transcription factors that regulate embryogenesis and cellular differentiation. They are divided into four clusters in mammals, including HOXA on chromosome 7, HOXB on chromosome 17, HOXC on chromosome 12, and HOXD on chromosome 2 (6) (Figure 1A). Each cluster contains nine and eleven genes: 11 in the HOXA cluster, 10 in the HOXB cluster, and 9 in the HOXC and HOXD clusters (Figure 1B). Upon expression, these genes' spatial and temporal collinearity allows for proper body segmentation and organogenesis in animals. That is, genes closer to the 3’ end are expressed first and are responsible for cranial development, while genes closer to the 5’ end are expressed later and are responsible for caudal development (7). Interestingly, despite their different locations, the functions of these HOX families are not autonomous. It is hypothesized that this functional redundancy across clusters is due to their highly conserved DNA-binding domain and duplication events early in vertebrate evolution.

Figure 1.

Homeobox (HOX) gene family. (A) Each of the HOX gene clusters, including HOXA, HOXB, HOXC, and HOXD, is located on chromosomes 7, 17, 12, and 2, respectively, in humans, and on chromosomes 6, 11, 15, and 2, in mice. (B) HOXA, HOXB, HOXC, and HOXD clusters contain 11, 10, 9, and 9 genes, respectively, likely due to duplication and loss during evolution. Each gene contains two exons and a homeodomain, which is highly conserved and acts as the DNA-binding domain of HOX proteins. Paralogs between clusters, such as the HOX10 paralogs, are highly synonymous in sequence. Isoforms are also common, and those with longer sequences are often the functional forms of each gene. Figure created using BioRender.

In humans, HOX genes are comprised of two exons and one intron, which may produce multiple isoforms. An integral part of this protein, called the homeodomain, is a highly conserved helix-turn-helix motif of 60 amino acids responsible for DNA binding (8). Several subgroups of three amino acid loop extension (TALE) homeodomain proteins, including MEIS and PBX, bind to the hexapeptide motif located slightly upstream of the HOX homeodomain and act as cofactors to aid in the stability and specificity of HOX DNA binding (9-13). Interestingly, binding sites for MEIS and PBX class members have been identified adjacent to HOX binding sites at gene promoters and enhancers, and studies have demonstrated that these DNA-binding proteins form complexes and bind to target DNA sequences in tandem (14).

Despite efforts to identify downstream transcriptional targets of HOX genes, they remain largely unknown today. Many that have been predicted in the literature have unfortunately not been confirmed using a rigorous experimental approach (15, 16). Mainly, there is a lack of evidence supporting HOX's direct regulation of these genes, suggesting that they may be regulated more directly by HOX's downstream targets (17). However, there is evidence supporting the direct regulation of key genes by HOX, particularly in relation to disease development. In fact, among the known downstream targets of HOX gene clusters, which include TP53, ITGB3, VEGF, IGFBP-1, PTN, and NCAM (17, 18), many are aberrantly expressed in disease and are involved in cancer-associated signaling pathways (19).

Despite the vital functions of HOX genes during embryonic development, their expression in adult tissues is commonly associated with various disease states. Specifically, these genes have demonstrated a strong propensity for overexpression in tumor tissues and an association with the development and progression of cancer (Table 1). This is supported by our knowledge of HOX gene targets, which are concomitantly dysregulated in these cases. Namely, HOX genes and their transcriptional targets have been largely associated with oncogenic functions, including cancer cell proliferation, invasion, migration, metastasis, and resistance to apoptosis (5). However, some HOX genes, including HOXB13 and HOXA5, have also demonstrated tumor-suppressive behavior in colon and esophageal cancer (20, 21), indicating that the role of HOX genes is both cancer and HOX-specific. To gain clearer insight into the overlapping and distinct roles of HOX family genes, we conducted an extensive literature review and summarized the main functions of each gene in Table 1.

Table 1:

Differential expression and function analysis of different HOX family members in GI cancers

Malignancies	Protein Expression	Gene	Tumor Promoter/Tumor Suppressor
Esophageal	Upregulated	HOXA1	Tumor promoter
		HOXA5	Therapy (Cisplatin) resistance
		HOXA7	Independent Prognostic Biomarker
		HOXA9	Tumor promoter
		HOXA10	Apoptosis suppressor
		HOXA13	Tumorigenesis and metastasis
		HOXB7	Enhanced radio resistance
		HOXC6	Tumor promoter
	Downregulated	HOXD10	Tumor suppressor
Colorectal cancer	Upregulated	HOXA6	Tumor promoter
		HOXA9	Lymph node metastasis
		HOXA10	Poor prognosis
		HOXA13	Tumor promoter
		HOXB4	Increased infiltration of TAM’s
		HOXB5	Metastasis
		HOXB7	Prognostic factor
		HOXB8	Proliferation and metastasis
		HOXB9	Poor survival
		HOXC8	EMT
		HOXC9	Poor prognosis
		HOXC10	Recruitment of MDSCs
	Downregulated	HOXA5	Counteracts stem cell traits
		HOXB13	Downregulates TCF4 and c-myc
		HOXD8	Apoptotic inducer
		HOXD10	Tumor suppressor
Gastric cancer	Upregulated	HOXA1	Tumor Promoter
		HOXA6	Metastasis
		HOXA11	Lymphatic metastasis
		HOXA13	Chemoresistance
		HOXB5	Tumor promoter
		HOXB7	Poor prognosis
		HOXB8	EMT
		HOXC6	Its mRNA utility as biomarker
		HOXC9	Prognostic marker for poor survival
		HOXC10	Tumor promoter
		HOXD4	Poor prognosis
		HOXD9	Tumor promoter
	Downregulated	HOXA5	Tumor suppressor
		HOXB9	Tumor suppressor
Hepatocellular carcinoma	Upregulated	HOXA9	Tumor promoter
		HOXA10	Tumor promoter
		HOXA13	Poorly differentiated tumors with resistance to sorafenib
		HOXB5	Tumor promoter
		HOXB7	Poor prognosis
		HOXB9	EMT
		HOXC6	EMT, poor survival
		HOXC10	Tumor promoter
		HOXD3	Angiogenesis
	Downregulated	HOXB4	Its overexpression increased T cell infiltration
		HOXD10	Tumor suppressor
Pancreatic cancer	Upregulated	HOXA3	Tumor promoter
		HOXA9	Stemness contributor
		HOXA10	Immune regulation
		HOXB5	Tumor promoter
		HOXB7	Survival predictor
		HOXB8	Tumor promoter
		HOXC4	Tumor promoter by LDHA-mediated glycolysis
	Downregulated	HOXB9	Cell cycle arrest at the G0/G1 phase
		HOXD10	Tumor suppressor

Taken together, these studies reveal that the differential functions of HOX family genes stem from their interactions with distinct cofactors, epigenetic modulation, and diverse downstream targets. Their involvement in tumor progression, particularly concerning therapy resistance and immune regulation, is further emphasized by the identification of specific HOX gene signatures as prognostic markers. Conversely, downregulated HOX genes with tumor-suppressive functions contribute to tumor regression by inducing cell cycle arrest and enhancing immune regulation upon overexpression.

Interestingly, the HOX10 paralogous group (HOXA10, HOXC10, HOXD10) is among the few HOX cluster paralogs that share common downstream targets, specifically cytokine and NF-κB signaling and apoptosis (5). One possible reason for the target alignment among these three genes is that they also share two conserved motifs at either end of the homeodomain, which are proposed to alter their DNA-binding affinity (22). Further, HOXC10 and HOXA10 have also been shown to regulate key immunosuppressive genes (23, 24). The expression of the HOXA family genes has also been commonly associated with tumor progression and differential survival, and importantly, HOXA genes are also important for GI development in the early embryo (25). The impact of HOXA10, one of the 5’ HOX proteins coded on chromosome 7, has been of particular interest in GI cancers.

HOXA10 IN GASTROINTESTINAL CANCERS

Esophageal

Esophageal cancer has one of the lowest 5-year survival rates among major cancers, which is primarily due to its asymptomatic nature and resulting late-stage diagnosis (1). Several studies have demonstrated HOXA10 expression in esophageal cancer. In fact, there is a positive correlation between HOXA10 expression and esophageal tumor stage and grade, with the highest expression observed in stage IV tumors (26) (Figure 2A & 2B). In line with this observation, HOXA10 has also been shown to activate the p38 and ERK MAPK signaling pathways, which are associated with enhanced cell proliferation, decreased apoptosis, and metastasis in esophageal cancer (27).

Figure 2.

HOXA10 expression in gastrointestinal (GI) cancers. (A) Box plots of HOXA10 expression in GI cancer tissues versus the normal. Expression of HOXA10 is significantly higher in colorectal, esophageal, liver, pancreatic, and stomach cancers than in normal tissue. (B) Stage-wise expression of HOXA10 in GI cancers. HOXA10 expression is significantly heightened throughout the progression of esophageal, liver, pancreatic, and stomach cancers, and maintains stable expression throughout colorectal cancer progression. Data downloaded from MiPanda, a platform that includes GTEX and TCGA datasets, were used to generate these plots in R. The p-values were calculated using the Wilcoxon signed-rank test.

Interestingly, high expression of HOXA10 is also correlated with several factors that impact patient outcomes in esophageal cancers, including decreased sensitivity to chemotherapy, including the alkylating agent cisplatin, increased tumor mutational burden, and microsatellite instability (26). As chemotherapy resistance is both an intrinsic and acquired characteristic of many esophageal tumors, and HOXA10 has been correlated with both treatment resistance and immune infiltration (26), it is posited that HOXA10 may contribute to drug resistance and thereby affect patient outcomes. Specifically, HOXA10 may be involved in the progression of this disease and may also drive the inefficiency of immunotherapy regimens, as it correlates with increased levels of immunosuppressive regulatory T cells (T_regs) and decreased levels of proinflammatory M1 macrophages (26).

HOXA10 long non-coding RNA (HOXA10-AS) has also been shown to promote the malignant phenotype of esophageal cancer. In fact, HOXA10-AS stabilizes the ribosomal binding protein FMR1 that is shared between HOXA10 and HOXA10-AS, which preserves the direct upregulation of HOXA10 by HOXA10-AS. This interaction supports the migration, invasion, and overall tumor growth of esophageal cancer (28). Together, these studies suggest that HOXA10 plays a tumor-promoting role in esophageal cancer and contributes to disease progression through several different mechanisms (Figure 3).

Figure 3.

Molecular mechanisms of HOXA10-mediated signaling pathways and downstream targets in different GI cancers. HOXA10 plays variable roles in the progression of esophageal, gastric, liver, pancreatic, and colorectal cancers. A largely pro-tumoral role is observed in esophageal, gastric, liver, and pancreatic cancers, while anti-tumoral function is seen in colorectal cancer. Figure created using BioRender.

Gastric

Overexpression of HOXA10 has been reported in gastric cancer patients with poor prognosis, and several studies have indicated its potential as a prognostic marker when combined with CD44 or miR-196b (29, 30). Its expression is often observed in the early stages of gastric cancer and persists throughout disease progression (Figure 2A & 2B). This pattern of expression suggests that HOXA10 may play a role in the initiation and development of gastric cancer (29). However, there are conflicting reports about the correlation between HOXA10 expression and disease severity. One study reported a positive relationship between HOXA10 expression and the prognosis of patients with the intestinal mucin phenotype and found that HOXA10 conferred antitumor characteristics in gastric cancer cells (31). This discrepancy could be attributed to subtype-specific effects of HOXA10 expression or the variability of test sets, sample sizes, and patient data. Therefore, further research is needed to determine whether subtypes differ in their roles in disease progression.

More recently, the importance of HOXA10 in gastric cancer metastasis was studied by Song et al. (32). Although HOXA10 is increased in gastric cancer, this group found even higher expression in patients with lymph node metastases. Further research showed that HOXA10 enhanced gastric cancer cell migration and invasion in vitro. Similar results were seen in live models, with more metastases in mice with HOXA10-positive tumors. Notably, when examining cells with HOXA10 overexpression and knockdown, there was a loss of the epithelial marker E-cadherin, increased levels of mesenchymal markers N-cadherin and vimentin, and higher levels of EMT transcription factors Snail and Slug in cells with high HOXA10. Further analysis revealed that HOXA10 directly regulates TGFβ2, a key player in EMT, affecting downstream TGFβ/SMAD signaling, METTL3 expression, and N6-methyladenosine (m6A) modifications. Additional studies suggest that HOXA10 may also promote anti-apoptotic features in gastric cancer cells by increasing BCL2 expression through transcriptional regulation. Notably, gastric cancer cells with HOXA10 overexpression show higher BCL2 levels and lower levels of pro-apoptotic proteins BAX, cleaved PARP, Caspase-9, and Caspase-3 compared to cells with reduced HOXA10, indicating that HOXA10 could regulate the intrinsic apoptosis pathway by binding the BCL2 promoter (33). More research on the downstream targets of HOXA10 is needed to understand better the HOXA10-BCL2 axis and the pathways involved in HOXA10-driven gastric cancer behavior. Overall, these findings suggest that HOXA10 may enhance gastric cancer progression, metastasis, and EMT by regulating key signaling pathways at the transcriptional level. Supporting this, evidence indicates that HOXA10 may promote gastric cancer by activating the JAK1/STAT3 signaling pathway (34), as STAT3 mediates TGFβ-induced EMT in various cancers (35-37) (Figure 3).

Previous studies have established a significant correlation between CD44 expression and poor survival in gastric cancer patients (38). In light of this, one study examined the relationship between CD44 and HOXA10 in gastric cancer. Their initial findings determined that CD44 and HOXA10 are co-expressed in gastric cancer and that CD44 may be a downstream target of HOXA10, with their signaling pathways interacting through moesin (30), a member of the ERM protein family that plays a key role in cell-cell signaling (39). Further analysis also found a positive association between HOXA10 and CD44 co-expression and tumor invasion, metastasis, and poor prognosis (30), suggesting that HOXA10 may promote tumor progression by regulating CD44.

Colorectal

Given that HOX genes are involved in tumorigenesis of multiple cancer types, the significant and consistent upregulation of HOXA10 throughout colorectal cancer development has been a point of interest in colorectal cancer research (Figure 2A & 2B). Interestingly, HOXA10 expression in colorectal cancer has previously been shown to be inversely correlated with PTEN (40), a tumor suppressor often inactivated in colorectal cancer (41). It is also important to note that PTEN has more recently been linked to immune responses in the tumor microenvironment. Specifically, PTEN deficiency has been linked to increased tumor infiltration of immunosuppressive T_regs, M2 macrophages, and MDSCs (42). Given the association between HOXA10 and PTEN expression, the predicted downstream regulation of PTEN by HOXA10 (43, 44), and its high expression in colorectal cancer, HOXA10 may impact disease development and patient survival through downstream immune regulation (Figure 3).

Contradictorily, HOXA10 has also been identified as a functional repressor of integrin β1, a key mediator of extracellular matrix degradation, and its expression in colorectal cancer is suppressed by miR-27b-3p, thereby promoting integrin β1-mediated cell migration and invasion (45). A similar study determined that miR-544a also functionally represses HOXA10 in colorectal cancer, which directly heightens the invasive and metastatic capabilities of colorectal cancer (46). Further, HOXA10 is a recognized negative transcriptional regulator of 3-phosphoglycerate dehydrogenase (PHGDH) in murine and human endometrium (47). A recent study demonstrated the significant role of monoubiquitinated PHGDH in colorectal tumor growth and metastasis by enhancing S-adenosylmethionine levels (48, 49). Serine palmitoyltransferase (SPT) mediates the synthesis of sphingolipids through serine utilization and promotes the generation of toxic 1-deoxysphingolipid using alanine as a substrate under serine restriction conditions. In colorectal cancer, inhibition of PHGDH decreased circulating serine levels and enhanced deoxysphingolipid accumulation, thereby reducing tumor growth (49). While limited exploration has examined the direct consequences of HOXA10 transcriptional regulation on PHGDH-mediated metastasis in colorectal cancer, this evidence aligns with studies suggesting an antitumor function of HOXA10 in colorectal cancer. Altogether, these investigations indicate that HOXA10 may play complex and diverse roles in colorectal cancer progression, particularly in the primary tumor versus metastatic sites.

Hepatic

Overexpression of HOXA10 is also observed in liver cancers (Figures 2A & 2B), particularly in liver tumor-initiating cells (TICs), also known as cancer stem cells or tumor-initiating stem cells (50). These initiating cells are crucial to tumor development and progression, particularly because of their self-renewal and differentiation capabilities (51). Interestingly, they are also a key component of therapy resistance and immune escape (52). Specifically, crosstalk between TICs and the tumor microenvironment leads to the generation of immunosuppressive signals on antigen-presenting cells, and promotes elevated NF-κB and Wnt signaling activation, both pathways that are associated with therapy resistance (53) The discovery that HOXA10 plays a major role in the promotion of TIC self-renewal (50) further supports indications that HOXA10 may be associated with the immunosuppressive environment in cancer, as TICs have a well-documented capacity for immune evasion (54-56).

In liver cancer, unchecked cell division is driven by several factors, including dysregulated cell cycle proteins, mutated oncogenes, and mutated tumor suppressor genes. Among these, p53 regulation is of particular interest (57). p53, a tumor suppressor, performs important functions in cell cycle progression, apoptosis, and DNA repair. The dysregulation of p53 by HOXA10 via HDAC1 suppresses p53 activity, thereby enhancing cell proliferation in liver cancer cells (57). Besides its function in these processes, p53 is also implicated in multiple important physiological functions, including immune function (58, 59).

HOXA10 is further implicated in the progression of liver cancer through its mechanism of action in the suppression of hepatitis B virus replication in hepatocytes (60). It is reported that HOXA10 has two modes of action to reduce HBV replication, including competing with STAT3 for binding to the promoter region and binding to p38 MAPK, recruiting SHP-1 to catalyze the dephosphorylation of p38 MAPK/STAT3 (60). This is supported by a study that found SHP-1 was a key element in the suppression of JAK1/STAT3 and p38 MAPK phosphorylation signals and downstream immunosuppression in mesenchymal stem cells (61). However, in cancer, these pathways have very different implications. STAT3 signaling is of particular interest in liver cancer due to its diverse functions in this disease state, particularly those associated with autophagy, epithelial-to-mesenchymal transition, radio resistance, and chemoresistance, all of which promote cancer progression and development (62). Similarly, one isoform of p38 MAPK, p38γ, has been shown to contribute significantly to cell cycle progression and liver tumorigenesis. p38γ may also allow cancer cells to exit quiescence independent of cyclin-dependent kinases (63, 64). Besides these, SHP-1, which HOXA10 recruits in alternative disease states, also plays a key role as a tumor suppressor in liver cancer development and progression. Specifically, SHP-1 has demonstrated its capacity to inhibit several signaling pathways, including STAT3, NF-kB, and AKT, all of which are important in cancer development (65). Taken together, these findings indicate that HOXA10 may play a dynamic role in immune regulation of the tumor microenvironment in liver cancer through multiple mechanisms and may act as an upstream regulator of tumor suppressors and key cancer-promoting signaling pathways.

Pancreatic

Unlike other GI cancers, our understanding of HOXA10 and its role in pancreatic cancer is largely unknown. Its expression has been reported to be statistically higher in pancreatic tumors than in the normal pancreas and to increase in a stage-wise manner throughout disease development (66) (Figure 2A & 2B). Recently, we identified an HOXA10-associated prognostic subtype in pancreatic cancer patients and demonstrated the ability to differentiate between short- and long-term survivors (67). We, the authors, further investigated the role of this HOXA10 signature in pancreatic disease biology, with an emphasis on its impact on the immune microenvironment. However, attempts by a few groups have been made to elucidate the role of HOXA10 in pancreatic cancer development and progression (15, 66, 68).

One such study has shown that HOXA10 promotes tumorigenesis in pancreatic cancer cells by activating canonical NF-kB signaling (66). The authors specifically showed that HOXA10 mediates upregulation of the IKKβ gene, thereby sustaining NF-kB signaling in a TNFα-independent manner. Additionally, HOXA10 has been shown to enhance pancreatic cancer cell invasion by inducing TGFβ2-p38 MAPK-dependent MMP-3 expression (15). NF-kB signaling is well characterized as a multifaceted transcriptional regulator in oncogenesis, including its role in cancer immunity, and p38 MAPK signaling has been associated with increased levels of intratumoral immunosuppressive myeloid cells and chemoresistance in pancreatic cancer (69) (Figure 3). These findings indicate a tumor-promoting role of HOXA10 in pancreatic cancer cells and suggest the potential to understand the role of this axis in immune regulation within the tumor microenvironment.

These studies indicate that HOXA10 plays a predominantly tumor-promoting role in GI cancers. Interestingly, this function of HOXA10, particularly as it relates to signaling pathways, is also observed in prostate (70), lung (71), brain (72), and ovarian (73) cancers. Specifically, similar to the regulatory role of HOXA10 on TGFβ2 observed in gastric cancer, a RFX6/HOXA10/TGFβ2 regulatory axis that promotes cancer cell proliferation, EMT, metastasis, and therapy resistance has been identified in prostate cancer (70). Similarly, the downstream effects of HOXA10 on m6A modifications and JAK/STAT signaling in gastric cancer (32, 34) was also observed in ovarian cancer, ultimately culminating in therapy resistance due to activation of JAK2/STAT3 signaling via HOXA10/ALKBH5-mediated JAK2 m6A demethylation. Lastly, the regulatory role of HOXA10 on HDAC1 has been observed in liver cancer (57) and in lung cancer (71), showing that HOXA10-induced upregulation of HDAC1 promotes lung tumor progression through the DNMT1/KLF4 axis. It is worth mentioning that a tumor suppressor function similar to what is observed in colorectal cancer is also seen in breast cancer (74). However, the mechanism of HOXA10 does not appear to overlap with that of these cancers. Taken together, these studies demonstrate the cancer-dependent function of HOXA10 and suggest that further exploration of its mechanism is needed.

HOXA10-MEDIATED CHANGES IN IMMUNE LANDSCAPE

The influence of HOXA10 on immunoregulation was first observed by Yao and colleagues during a study aimed at understanding the underlying mechanism behind human embryo implantation failure. This group found that HOXA10 regulates local immunosuppression, specifically a downregulation of T lymphocytes in the uterine stroma, an important phenomenon required for successful implantation (75). Importantly, immunosuppression is also a well-defined characteristic of GI cancers, including those of the esophagus, stomach, colorectum, liver, and pancreas (76-80). In fact, our recent study found that a HOXA10-associated prognostic gene signature is associated with increased levels of immunosuppressive T_regs and a decrease in the levels of CD8⁺ T cells in pancreatic tumor data analyzed using CIBERSORT (67). This observation is supported by our preliminary data showing HOXA10-dependent modulation of T lymphocyte populations in murine orthotopic pancreatic tumors (unpublished). The aforementioned GI studies, particularly those in pancreatic cancer, suggest a potential role for HOXA10 in immune modulation in GI cancers. Therefore, we discuss here studies that supplement our understanding of the regulation of immune pathways through HOXA10. As previously mentioned, HOXA10 activates the p38 MAPK signaling pathway in esophageal cancer (27) and pancreatic cancer (15), promoting metastasis and chemoresistance. Interestingly, p38 MAPKs also play a key role in T cell development. Persistent activation of the p38 MAPK signaling pathway in immature thymocytes blocks their differentiation into CD4⁺CD8⁺ double-positive progenitor cells (81, 82), and the activation of ERK signaling is critical to support T cell differentiation and CD8⁺ T cell effector function (83). Interestingly, HOXA10 is associated with increased levels of T_regs, and its associated prognostic gene signature is associated with decreased CD8⁺ T cells in pancreatic tumors (67). Taken together, these studies suggest that HOXA10 expression supports progression and the immunosuppressive tumor microenvironment in esophageal and pancreatic cancers by activating p38 MAPK signaling. Intratumoral enrichment of T_regs is a commonly observed phenomenon in esophageal cancer (26), gastric cancer (84), colon cancer (85), liver cancer (86), and pancreatic cancer (87). While the percentage of infiltrating T_regs varies across these cancers, intratumoral Tregs are associated with poorer prognosis, an immunosuppressive phenotype, and decreased immunotherapy efficacy. In esophageal cancer, HOXA10 has been associated with increased infiltration of Tregs, implicating it as a protumorigenic immunosuppressive modulator (26). Interestingly, moesin, a key molecule in the HOXA10 interactome in gastric cancer, is also an important regulator of T_reg differentiation and TGFβ/SMAD signaling (88), implicating this group of molecules in reduced antitumor immunity and enhanced tumor development (30).

CD44, a downstream target of HOXA10 (30), is associated with immune escape through transcriptional upregulation of PD-L1 (89), a checkpoint inhibitor expressed on tumor cells. CD44 is well established for its expression in GI cancers of the esophagus (90), colon (91), stomach (92), liver (93), and pancreas (94), and has been specifically implicated in tumor cell invasion (95-97), metastasis (98-100), EMT (101, 102), and the poor prognoses (93, 100, 102-104) of these cancers. However, limited efforts have been made to elucidate the role of HOXA10-mediated CD44 expression in the immune microenvironment and antitumor immunity in GI cancers, despite corroborating evidence that CD44 indirectly suppresses immune infiltration by upregulating PD-L1 in other cancers (89, 105, 106). Therefore, studies are warranted to understand the role of HOXA10-mediated immune exhaustion via CD44.

The association between PTEN deficiency and immune evasion (107-109) supplies further evidence that HOXA10 may impact immune infiltrates. In particular, upregulation of PI3K signaling and reduced CD8⁺ T cell infiltration have been observed when PTEN expression is ablated (110). PTEN loss also induces downstream PD-L1 expression in tumor-associated macrophages (TAMs), TAM-mediated inhibition of CD8+ T cells, and increased metastatic potential, collectively leading to immune evasion and poor prognoses in colorectal cancer (107, 111). These effects are likely attributable to HOXA10-mediated STAT3 signaling (34), since PTEN is one of its downstream targets. Remarkably, STAT3 has also been reported to bind to the IL-6 promoter and create a positive-feedback loop, leading to increased levels of IL-6 (112, 113) and increased induction of CD44 expression (114), which may lend the immunosuppressive characteristics observed in colorectal cancer cells (115) and promote the epithelial-mesenchymal phenotype (116). This is supported by research demonstrating the crucial role of PTEN suppression in promoting EMT in colorectal cancer (117) and by the observed increase in HOXA10 expression in lymph metastases (40). Together, these suggest that HOXA10 may play a key role in EMT and immunosuppression in colorectal cancer.

CIBERSORT data in our recent study also revealed an association between an HOXA10-associated prognostic gene signature and increased levels of immunosuppressive M2 macrophages in pancreatic tumors (67). Preliminary data indicate that HOXA10-dependent modulation of myeloid cell populations occurs in murine orthotopic pancreatic tumors (unpublished), and this may contribute to the classical immunosuppressive microenvironment of pancreatic tumors. Importantly, subpopulations of myeloid cells, such as macrophages, are often observed in the tumor microenvironment of other GI cancers.

Macrophages, including M1 and M2 polarized macrophages, have been shown to exert antitumor and pro-tumor effects, respectively. Interestingly, HOXA10 has been associated with promoting immunosuppressive antigen-presenting cells and suppressing proinflammatory cells. In esophageal cancer tissues, HOXA10 is correlated with decreased infiltration of proinflammatory M1 macrophages (26), a distinguishing feature associated with poorer survival in these cancers (118). In liver cancer, it is proposed that HOXA10 may influence M2 polarization of tumor-associated macrophages via the KAT2B/HOXA10 axis (119), in which KAT2B, an immune-infiltration-associated biomarker (120), is induced by p53 (121). In breast cancer, studies show that p53 activation induces DD1α expression, a receptor-like transmembrane protein of the immunoglobulin superfamily. This promotes macrophage-mediated engulfment of dead apoptotic cells and proper immune surveillance. Interestingly, the expression of immune checkpoint coinhibitory molecules PD-1 and PD-L1 also increased in response to p53 activation. DD1α has also been shown to inhibit T cell activation, suggesting that inhibition of p53 by HOXA10 in liver cancer may lead to differential modulation of the two branches of the immune system by DD1α and PD-1 and PD-L1 (122, 123). Namely, decreased expression of these proteins may inhibit the innate immune system while promoting the adaptive immune system (58). Together, these studies have demonstrated that HOXA10 may be associated with the tumor immune microenvironment, specifically by inhibiting proinflammatory immune cells and promoting immunosuppressive phenotypes.

THERAPEUTIC OPPORTUNITIES UTILIZING HOXA10

Given HOXA10's prominent role in tumor development and metastasis, particularly by influencing immune response, it is relevant to ask how this phenomenon may be leveraged to enhance immunotherapies across different GI cancers. Current immunotherapies for GI cancers, such as checkpoint inhibitors, have demonstrated varying degrees of efficacy and are often combined with standard-of-care chemotherapy to improve outcomes (127-129). In the same way, it may be possible to target immunosuppressive pathways downstream of HOXA10 while simultaneously targeting key signaling axes that activate antitumor immune cell types. Using this strategy, it may be possible to focus on a single key immune cell type and target it with two distinct mechanisms. For example, the PD-1/PD-L1 and CD47/SIRPα axes may be targeted in combination, leading to concomitant improvements in macrophage activation and phagocytic ability (130). Similarly, p38 and ERK MAPK signaling pathways may be targeted whilst also using an immune checkpoint blockade to simulate T lymphocyte differentiation and activation. In this way, by focusing on a single immune cell type rather than a broader biological phenomenon, the overall efficacy of a treatment combination may exceed that of current dual treatments.

Disappointingly, only two studies to date attempt to target HOX genes, and both have yet to recruit patients. They are also both concerned with improving treatment outcomes for patients with high-HOX-expressing leukemias, which notoriously express high levels of other HOXA family members. The therapies in question, including revumenib and a combination therapy of nintedanib and azacitidine, are proposed to indirectly target HOX genes by inhibiting HOX cofactors, demethylating hypermethylated HOX genes, or inhibiting upstream or downstream cancer-promoting signaling molecules, respectively. The efficacy of these therapies is also being explored in therapy-refractory solid malignancies, including gastrointestinal, gynecological, endocrine, respiratory, skin, brain, breast, and prostate cancers. Importantly, these drugs have received limited FDA approval for use in solid malignancies.

Multiple strategies could be used to target HOXA10 directly. RNA-based therapeutics, including HOXA10-targeting antisense oligonucleotides, have shown utility in early studies for classifying HOXA10 function in endometrial tissues (131, 132). Antisense oligonucleotides have also demonstrated efficacy in targeting HOXB1 in embryonal carcinoma cells (133). This targeted molecular approach and gene therapies have demonstrated some efficacy in various disease states over the last couple of decades (134). Nevertheless, HOX-targeting strategies necessitate additional evaluation in preclinical models prior to their suitability for clinical testing.

CONCLUSIONS AND PERSPECTIVES

HOXA10 plays a multifaceted role in GI cancers, influencing tumor progression, metastasis, and the immune microenvironment. Specifically, iHOXA10 promotes proliferation, invasion, and EMT in tumor cells by modulating pathways like JAK/STAT and MAPK signaling (Figure 3). In the immune microenvironment, HOXA10 can affect the recruitment and polarization of immune cells. Thus, HOXA10 may serve as both a driver of tumor progression and a modulator of immune interactions, underscoring the importance of understanding its mechanism. HOXA10 may enhance macrophage recruitment, skew them toward an immunosuppressive M2 phenotype, and suppress CD8+ T cell infiltration, thereby dampening antitumor immunity. One proposed mechanism of HOXA10 is that it acts through epigenetic changes, a key component of gene regulation. Abnormal epigenetic changes in tumor suppressors and oncogenes can lead to unchecked cell growth and division, contributing to cancer development. In liver cancer, HOXA10 has been shown to modulate HDAC1 transcriptionally (57), promoting the deacetylation of p53 and its tumor-suppressive functions. Studies have also shown that the interaction between HOXA10 and HDAC2 represses transcription in undifferentiated myeloid cells (124), and that the interaction between HOXA10 and SIRT2 regulates gene expression patterns during development (125). Further, epigenetic modifications of the HOXA10 promoter, including hypomethylation in gastric cancer (126), enhance the proliferative and invasive capabilities of tumor cells. The influence of HOXA10 on epigenetic modulators, including HDAC1 and DNMT1 in lung cancer (71), further supports this mechanistic supposition.

Overall, this review highlights gaps in the field for further research on the mechanism of HOXA10, particularly regarding how it interacts with and regulates immune signaling pathways to shape the immune microenvironment, using genetically engineered mouse models, comprehensive single-cell RNA sequencing, and spatial transcriptomic technologies. This is especially important when considering therapeutic development, where combining an HOXA10-targeted therapy with an immune checkpoint inhibitor or other immunotherapy may improve overall efficacy. As we have discussed here, an accumulating body of literature suggests that HOX genes could be a novel target for anticancer therapy in several different cancer types. This is particularly true for HOX genes, such as HOXA10, which promote the tumorigenic and metastatic potential of cancer cells. In cancers resistant to immunotherapies due to variable infiltration of pro- and anti-tumoral immune cells, the dual use of an HOX-targeting drug may be the difference between a failed and a successful clinical evaluation. Focusing on these areas of research will provide a foundation for improving outcomes for patients with GI cancers.