HOX and MEINOX in cellular plasticity, fibrosis, and cancer

Abstract

HOX transcription factors and their cofactors, MEINOX, are critical regulators of positional identity and cellular plasticity. While their functions are essential during embryonic development, they also play key roles in maintaining adult tissue homeostasis. Dysregulation of HOX and MEINOX has been implicated in the pathogenesis of various diseases, including fibrosis and cancer. This review explores the contributions of HOX and MEINOX to dedifferentiation and cellular reprogramming, processes that drive fibrotic disease onset and cancer progression. It also addresses their role in extracellular matrix remodeling in these conditions. Particular attention is given to their involvement in epithelial-mesenchymal transition, where altered HOX and MEINOX expression promotes phenotypic plasticity, cancer invasiveness, and fibrotic tissue remodeling. By integrating these perspectives, this review underscores the significance of HOX-MEINOX dysregulation and altered positional identity in disease progression. Targeting this dysregulation may offer innovative strategies to modulate epithelial-mesenchymal transition and extracellular matrix dynamics, presenting new therapeutic opportunities for combating fibrosis and cancer.

Key Words: HOX; MEIS; MEINOX; PKNOX; Fibrosis; Cancer; Molecular signaling pathways

Core Tip: HOX transcription factors and MEINOX cofactors critically regulate positional identity and cellular plasticity, fundamental processes in development and adult tissue homeostasis. Dysregulation of these factors profoundly influences extracellular matrix remodeling and epithelial-mesenchymal transition, promoting fibrosis progression and cancer invasiveness. By highlighting specific HOX-MEINOX-mediated mechanisms underlying disease pathology, including their interactions with key signaling pathways (transforming growth factor-beta, Wnt, Notch), this review identifies novel therapeutic targets. Modulating HOX-MEINOX activity could offer innovative strategies to reverse aberrant epithelial-mesenchymal transition and extracellular matrix remodeling, bridging translational gaps toward effective treatments for fibrotic diseases and cancer.

INTRODUCTION

Understanding how cells determine their positional identity and choose between retaining or relinquishing plasticity is akin to decoding nature’s blueprint for life, with profound implications for both development and disease. Among the molecular regulators guiding these fundamental biological decisions, HOX transcription factors and their cofactors, PBX and MEINOX, have emerged as essential regulators that coordinate positional identity and cellular plasticity throughout embryonic development and persistently into adult life[1].

HOX genes and their cofactors, initially characterized by their canonical roles in segmentation during embryogenesis, also exhibit diverse non-canonical functions crucial for adult tissue homeostasis, including roles in cell-cycle regulation, apoptosis, DNA repair, and metabolism[2-7]. Throughout development, these transcription factors are essential in establishing cellular positional identity by determining specific differentiation pathways on a cell-by-cell basis along the craniocaudal axis[8]. Closely intertwined with positional identity is the concept of cellular plasticity - the capacity of cells to alter their phenotype in response to physiological or environmental stimuli - which is also modulated by HOX-MEINOX interactions[9]. Thus, these molecular interactions serve as critical architects, orchestrating precise developmental patterns during embryogenesis while ensuring the necessary cellular flexibility for effective tissue repair and regeneration in adulthood.

However, the delicate balance maintained by these transcriptional complexes can become disrupted, potentially leading to pathological conditions characterized by altered cell identity and plasticity, notably in fibrosis and cancer[10]. In these diseases, dysregulated HOX-MEINOX expression is one of the key factors promoting pathological dedifferentiation, uncontrolled proliferation, and aberrant remodeling of the extracellular matrix (ECM), processes fundamental to fibrotic disease progression and cancer metastasis[2,11-14]. This review investigates how HOX and MEINOX dysregulation can contribute to disease development by altering positional identity and cellular plasticity and discusses their value as potential therapeutic targets for managing fibrosis and cancer progression.

HOX AND MEINOX: MOLECULAR FUNCTIONS AND MECHANISMS

Homeobox sequences are defined as a highly conserved DNA motif (about 180 base pairs) encoding a 60 amino-acid DNA binding region called homeodomains. HOX genes, which is a subset of homeodomain-containing genes encode a family of transcription factors known for their roles in specifying developmental patterns along the craniocaudal axis during embryogenesis[15,16]. In humans, HOX gene family consists of 39 protein-encoding genes organized into four clusters (HOXA, HOXB, HOXC and HOXD) located on distinct chromosomes, at loci 7p15, 17q21, 12q13, and 2q31 (Figure 1)[17-19]. Each cluster contains 13 paralog genes, occurred by segmental duplications, gene duplications and translocations during evolution[20]. HOX genes are numbered according to their sequential activation from 1 to 13 and localized in the 3’ to 5’ direction (Figure 1)[21]. Alongside these protein-encoding genes, 231 non-coding RNA (ncRNA) genes have been identified within four HOX loci which contains an approximate total of 500 kilobases[22-25]. This genomic organization facilitates temporal and spatial collinearity, a defining feature of HOX gene expression, where genes positioned at the 3’ end of clusters are expressed earlier and in more anterior regions, while genes at the 5’ end are expressed later and more posteriorly (Figure 1)[17,19]. Generally, 3’ anterior HOX genes can induce the 5’ posterior HOX genes expression, and the posterior genes can inhibit the anterior genes, which is defined as posterior prevalence or dominance[26]. Majority of HOX ncRNAs also show a coordination along the body development axes and may affect HOX gene expression in cis or trans[22]. For example, long ncRNA (lncRNA) HOTAIR, which is transcribed from HOXC locus, functions as a trans repressor of HOXD locus by promoting H3K27 trimethylation through its interaction with polycomb repressive complex 2 complex[22]. On the other hand, lncRNA HOXA transcript at the distal tip (HOTTIP), transcribed from the 5’ end of HOXA locus, cis activates the 5’ HOXA genes by promoting H3K4 trimethylation through its interaction with mixed-lineage leukemia complex[27]. Despite highly preserved nature of HOX gene sequences, species-specific differences are also observed, mainly in the loss or retention of particular HOX paralogs, which notably correlates with different regenerative capacities observed across species[19].

Figure 1 Organization and collinearity of human HOX gene clusters. This figure illustrates the genomic organization and transcriptional regulation of human HOX gene clusters (HOXA, HOXB, HOXC, HOXD). Each cluster is arranged linearly, displaying protein-coding HOX genes (rectangles), long non-coding RNA genes (rounded rectangles), and microRNA genes (diamonds). The transcriptional direction of genes within each cluster is indicated, highlighting the spatiotemporal collinearity characteristic of HOX genes, whereby genes located more 3’ are expressed earlier in development and in more anterior body segments, while genes positioned towards the 5’ end are expressed later and in progressively posterior regions. This principle, known as the “collinearity” rule, is represented by arrows depicting the relationships between the direction of transcription, developmental timing, and the anteroposterior positioning along the human body axis. Created in BioRender.

Through their homeodomains, HOX transcription factors bind to 5’-TAAT-3’ DNA sequence, which is abundantly found in the genome[28]. Despite the abundance of the HOX binding sequence along the genome, the mechanisms regulated by HOX genes display precise and specific genetic regulations, thereby constituting a long-standing paradox called “HOX specificity paradox”[29]. Further research into this paradox has revealed that HOX transcription factors perform their specific and precise regulatory roles through their interactions with cofactors, particularly PBX and MEINOX [MEIS and PKNOX (formerly named as PREP)] family of proteins[30-34]. PBX and MEINOX proteins belong to the three amino acid loop extension (TALE) family of homeodomain proteins, characterized by a conserved three-amino-acid insertion in their homeodomain. The MEIS family consists of three MEIS proteins, namely MEIS1, MEIS2 and MEIS3, and the PKNOX family consists of PKNOX1 and PKNOX2 proteins. PBX proteins (PBX1-4), specifically PBX1, have also been shown to function in nuclear localization of HOX and MEINOX proteins[35-37]. PBX or PBX-MEINOX complexes contribute to DNA-binding specificity and affinity of HOX protein binding to target DNA sequences, through forming either dimeric or trimeric complexes (Figure 2). DNA binding affinities of monomeric PKNOX/PREP and MEIS are weak; however, binding as a dimer with PBX enhances their affinity. An excellent study by Penkov et al[31] focused on DNA binding sites and motifs of PREP1 and MEIS, unveiling their divergent genomic binding patterns and proposing HOX paralog groups 1-9 as TALE-interactive and HOX paralog groups 10-13 as TALE-noninteractive regions. These protein complexes not only increase transcriptional precision but also resolve the long-standing HOX specificity paradox - how highly similar HOX proteins achieve distinct functional outcomes - by modulating DNA-binding specificity and transcriptional outputs through cooperative interactions.

Figure 2 HOX-PBX-MEINOX DNA binding affinity and transcriptional effectiveness. HOX proteins form stable dimers with PBX cofactors, which exhibit limited transcriptional activity. The addition of MEINOX cofactors (MEIS and PKNOX), facilitating the formation of HOX-PBX-MEINOX trimers, significantly enhances both the DNA-binding specificity and transcriptional output[32,34,195,196]. In this review, PBX proteins are considered a constant, essential component required for complete functional activity, while HOX and MEINOX proteins are treated as dynamic variables reflecting biological complexity and disease-specific regulatory states. Created in BioRender.

During embryogenesis, positional identity established by HOX genes instructs specific differentiation pathways at a remarkably fine cellular resolution. This positional information, generated with the “HOX code” of the cells, is critical not only for embryonic segmentation but also for the maintenance and functional specialization of adult tissues[38,39]. The collective, cell-specific transcriptional state of HOX genes across distinct cell types is termed the “HOXOME”[40]. Throughout adulthood, HOX and MEINOX cofactors remain actively engaged in maintaining tissue homeostasis, where they dynamically regulate cell differentiation and plasticity in response to physiological signals or environmental stressors[1,38,41]. Thus, HOX and MEINOX proteins serve as vital molecular orchestrators, continuously balancing cell identity stability with the requisite flexibility for adaptive tissue responses. In addition to canonical transcriptional regulation, HOX and MEINOX proteins - either as part of HOX-MEINOX complexes or as standalone molecules - exert diverse non-canonical functions in cells, influencing processes such as DNA repair, cell-cycle control, apoptosis and metabolic pathways[42-47]. These multifaceted roles highlight the versatile and context-dependent nature of HOX and MEINOX proteins, allowing them to influence cellular behaviors beyond mere positional patterning.

A comprehensive understanding of the molecular intricacies governing HOX-MEINOX interactions, including the balance between canonical and non-canonical roles and their broader implications in positional identity and cellular plasticity, provides crucial insights into developmental biology and disease pathology. It is well-established that PBX cofactors are essential and constant components required to facilitate the formation of transcriptionally active HOX-PBX-MEINOX trimeric complexes[48]. While HOX proteins can form dimers with PBX alone, these complexes exhibit limited DNA-binding affinity and transcriptional activity. The inclusion of MEINOX cofactors significantly enhances both DNA-binding specificity and transcriptional output[30,31]. Although HOX-PBX dimers have been implicated in cancer biology and represent potential therapeutic targets[49,50], our review intentionally maintains PBX as a constant cofactor, thus allowing focused exploration of the dynamic and disease-specific regulatory roles of HOX and MEINOX proteins. Subsequent sections will delve deeper into these concepts, exploring how disruptions in these finely tuned molecular processes specifically involving HOX and MEINOX contribute directly to diseases such as fibrosis and cancer, providing a robust foundation for identifying targeted therapeutic opportunities.

HOX AND MEINOX: ROLE IN CELLULAR PLASTICITY

Positional identity established during embryonic development and maintained throughout adulthood significantly influences cellular plasticity[51-53]. Importantly, cellular plasticity arises as a direct consequence of these positional cues, emphasizing the fundamental link between spatial identity and cellular adaptability. Positional identity defined by HOX and their MEINOX co-factors, guides immediate developmental cell fates and has lasting implications for the plasticity of adult tissues, particularly adult stem cells and their niches[54].

HOX and MEINOX proteins help reconfigure chromatin states and activate lineage-inappropriate genes in response to spatial cues, particularly in embryonic patterning and regenerative contexts. Such a process can be observed in both limb formation and regeneration in axolotl, where positional information driven by HOXA and MEIS expression also drives plastic responses[55,56]. In a higher-order vertebrate model organism, Raines et al[57] demonstrated that simultaneous frameshift mutations in six HOX genes (Hoxa9, 10, 11 and Hoxd9, 10, 11) in mice resulted in significant limb skeletal abnormalities, including reduced ulna and radius, underscoring the critical role of these genes in limb development and patterning.

During embryogenesis, HOX and MEINOX proteins contribute profoundly to establishing adult stem cell niches and maintaining their long-term integrity. The positional identity encoded by HOX genes ensures the correct establishment of these niches, which subsequently maintain the plasticity of resident adult stem cells. For example, Abd-B, a Hox gene in drosophila, was found to be the key driver behind testicular stem cell niche development through integrin-modulated cellular positioning[58]. In addition to niche, HOX and MEINOX proteins have been shown to be directly crucial in the acquisition and maintenance of stemness in various adult stem cells. For instance, Turan et al[59] developed small molecule inhibitors targeting MEIS1, a member of the MEINOX family, and demonstrated that these inhibitors modulate hematopoietic stem cell activity, underscoring the pivotal role of MEIS1 in maintaining hematopoietic stem cell function. Similarly, Bradaschia-Correa et al[60] found that HOX gene expression in periosteal stem/progenitor cells is essential for preserving their multipotency, with suppression of HOX genes leading to a loss of stemness and altered differentiation pathways. Together, these findings underscore the critical role of HOX and MEINOX factors not only in establishing the spatial framework of adult stem cell niches but also in preserving the intrinsic plasticity and self-renewal capacity of the stem cells they support.

The concept of a “HOX code” is particularly evident in mesenchymal stem cells (MSCs) derived from different anatomical sources, where unique HOX gene expression patterns contribute to their differentiation potential and plasticity. Research published by Khajeh et al[41] have demonstrated that MSCs derived from dental pulp and bone marrow exhibit highly tissue-specific HOX expression profiles, which persist during in vitro expansion and directly impact their plasticity. Multiple studies have also found that MSCs derived from these sources also display variability in osteogenic, adipogenic and chondrogenic capabilities, hinting at a possible correlation between their HOX codes and differentiation potentials[41,61,62]. A review by Steens and Klein[63] further consolidates this perspective by highlighting that tissue-specific HOX gene signatures not only reflect the developmental origin of MSCs but also serve as predictors of their differentiation trajectories and functional responsiveness to microenvironmental cues. Together, these findings indicate the role of the HOX code as a molecular framework that integrates developmental history with functional plasticity in MSCs, offering critical insight into their behavior and therapeutic potential.

Additionally, interactions between HOX-MEINOX proteins and their crosstalk with major signaling pathways, including Wnt, transforming growth factor-beta (TGF-β), and Notch, are fundamental in mediating cellular plasticity. These delicate frameworks integrate external signals with intrinsic transcriptional programs, fine-tuning the balance between pluripotency and differentiation. For instance, Notch signaling has been shown to drive HOX gene activation in human neuro-mesodermal progenitors, where its inhibition resulted in impaired HOX expression and disrupted posterior axial identity, demonstrating a direct regulatory role in early fate decisions[64]. Similarly, Wnt signaling influences HOX gene expression during stem cell differentiation through β-catenin stabilization, reinforcing HOX-dependent regulation of stem cell identity and lineage commitment[65,66]. HOXC8 was discovered to be promoting proliferation and migration in non-small cell lung cancer (NSCLC) cells by transcriptionally upregulating TGF-β1, highlighting a direct regulatory role of HOXC8 in TGF-β signaling pathways[67]. Elkouby et al[68] have demonstrated a direct regulatory link between the Wnt signaling pathway and MEIS proteins, showing that Wnt3a signaling negatively regulates MEIS3 expression during hindbrain development. The crosstalk between HOX-MEINOX proteins and these pathways enables cells to dynamically respond to developmental and environmental cues, ensuring robust regulation of cellular plasticity across different tissues and developmental stages. Understanding these intricate signaling networks is essential, as disruption of these finely-tuned interactions between HOX-MEINOX proteins and developmental pathways can contribute significantly to pathological conditions such as fibrosis and cancer, which will be explored further in the following sections.

HOX AND MEINOX IN FIBROSIS

Fibrosis is characterized by excessive accumulation and deposition of ECM components, particularly collagen fibers, leading to progressive disruption of normal tissue architecture and functional integrity. This pathological remodeling occurs as a maladaptive response to chronic injury or inflammation, driven by key molecular mechanisms including aberrant ECM remodeling, epithelial-mesenchymal transition (EMT), and differentiation of fibroblasts into activated myofibroblasts. These processes collectively result in persistent ECM synthesis, increased tissue stiffness, and impaired organ function. Fibrosis is not restricted to a single organ but occurs broadly across diverse tissues such as the liver, lungs, kidneys, heart, and skin. In this context, elucidating the roles of developmental transcription factors, such as HOX proteins and their MEINOX cofactors, is particularly promising given their critical involvement in positional identity and cellular plasticity. To further elaborate on how positional identity and cellular plasticity mechanistically contributes to fibrosis progression, we will discuss these concepts in relation to key molecular mechanisms of fibrosis.

Disruption or loss of positional identity can profoundly impact tissue homeostasis, steering cells toward aberrant differentiation pathways and pathological states, such as fibrosis. Through transcriptomic analysis, Forte et al[69] demonstrated that adult mouse fibroblasts retain distinct positional identities from embryogenesis, suggesting disruptions to these identities could directly contribute to tissue-specific fibrotic responses. Single-cell RNA sequencing of human dermal fibroblasts has also revealed distinct fibroblast subpopulations, each with unique positional identities that strongly correlate with their different fibrotic capacity[70,71]. In humans, multiple studies performed on various cohorts have also identified differences in the prevalence of keloid formation on different body parts. Keloid or hypertrophic scar formation has been found to be more prevalent in upper body regions such as shoulders and upper torso, when compared to lower regions such as hips and legs[72-75]. Curiously, this difference in prevalence also overlaps with the HOX expression gradients along the craniocaudal axis. Such an overlap hints at a possible relationship between the concept of dermal fibroblasts “HOX code” and the prevalence of fibrotic tissue repair in skin[76]. Conversely, the difference in injury-induced tension responses in different body parts was also suggested for these regional changes in keloid formation as well[72,77-79].

Complementary to positional identity, cellular plasticity - another critical process influenced by HOX-MEINOX proteins - allows cells to dynamically transition between differentiated and more flexible states, facilitating normal tissue repair and regeneration. However, dysregulated plasticity can initiate maladaptive responses, notably fibrosis, exemplified by excessive activation of EMT and persistent differentiation of fibroblasts or MSCs into contractile myofibroblasts. These maladaptive transitions result in excessive ECM accumulation, increased tissue stiffness, and impaired organ function[80-83]. Conversely, impaired cellular plasticity in disease states such as liver fibrosis, myelofibrosis, and idiopathic pulmonary fibrosis (IPF) can hinder effective tissue repair, perpetuating chronic tissue damage and further exacerbating fibrotic progression[84-88]. Thus, maintaining positional identity and proper cellular plasticity is critical, as disruptions in these processes constitute major mechanistic drivers of fibrosis. Dysregulation of HOX and MEINOX transcription factors significantly influences critical molecular mechanisms underlying fibrosis, including ECM remodeling, EMT, fibroblast functions, and myofibroblast differentiation.

Aberrant HOX gene expression directly affects ECM remodeling, as illustrated by Hahn et al[79], who found that fibroblasts from fibrotic skin lesions exhibited altered HOX gene expression patterns, notably reduced HOXA9 levels, which modulated ECM-regulating gene expression and fibroblast behavior. Similarly, Su et al[89] demonstrated that the lncRNA HOXA11-AS promoted ECM accumulation and fibrosis progression by regulating the miR-205-5p/FOXM1 signaling axis, highlighting a regulatory mechanism through HOX-related ncRNAs in ECM synthesis.

HOX-MEINOX dysregulation also significantly modulates EMT processes. Li et al[90] demonstrated that human umbilical cord MSC-derived exosomal miR-27b attenuates subretinal fibrosis by suppressing EMT through targeting HOXC6, indicating a key role of HOX transcription factors in controlling epithelial plasticity during fibrosis progression. In parallel, Wasson et al[91] found that the lncRNA HOTAIR, encoded from the HOXC locus on 12q13, activates NOTCH by driving enhancer of zeste homolog 2 (EZH2) to specific DNA regions that enhance methylation of the miRNA-34a locus and, in this way, myofibroblasts are activated in systemic sclerosis.

Finally, several HOX and MEINOX proteins play crucial roles in fibroblast-to-myofibroblast differentiation, a hallmark event in fibrosis. Wang et al[92] reported that the lncRNA HOTAIR facilitates high glucose-induced mesangial cell proliferation, fibrosis, and oxidative stress in diabetic nephropathy via regulating the miR-147a/Wnt2b axis, contributing to myofibroblast differentiation. Recent findings also highlight the role of the MEINOX transcription factor PKNOX2 as a critical regulator in this differentiation pathway[11,93]. Chen et al[11] demonstrated through single-nucleus RNA sequencing that reduced PKNOX2 expression correlated with a shift from physiological fibroblast activation to pathological myofibroblast differentiation in cardiac fibrosis, suggesting a protective role of PKNOX2 against fibrotic remodeling. Additionally, Miyake et al[93] identified PKNOX2 as a pivotal regulator of myofibroblast differentiation during renal fibrosis, further emphasizing the broad significance of MEINOX transcription factors in fibrotic pathologies. In our previous study, PKNOX2 expression was noted to be decreased in bone marrow MSCs of Fanconi anemia patients, a genetic disease associated with myelofibrosis[43,94].

To enhance our understanding of the mechanistic role of HOX-MEINOX molecules in fibrosis, we will briefly mention significant fibrosis-associated signaling pathways and their mechanistic contributions to fibrosis. TGF-β signaling is known as the central pathway driving fibrosis through fibroblast activation, myofibroblast differentiation, and ECM production[95-97]. Wnt/β-catenin signaling pathway is aberrantly activated in fibrosis and sustains fibroblast proliferation and myofibroblast activation, often in synergy with TGF-β[98,99]. Platelet derived growth factor (PDGF) signaling potently stimulates fibroblast migration, proliferation, and ECM production, thereby expanding the population of matrix-producing myofibroblasts in fibrotic lesions[100,101]. Notch signaling contributes to fibrosis by promoting fibroblast-to-myofibroblast transition and enhancing the release of collagen in a variety of organs[102-104]. Several recent findings provide mechanistic insight into how HOX-MEINOX molecules influence these pathways during fibrotic processes, outlined in the subsequent examples. Figure 3 provides a visual guide for how these HOX-MEINOX molecules partake in these pathways and the development of resulting fibrosis-related mechanisms. In a study by Wang et al[105], HOXA11 overexpression in gastric cancer cell lines were found to drive PDGF-BB and TGF-β1 paracrine secretion, which in turn induced fibrotic activation of peritoneal mesothelial cells to adopt a myofibroblast phenotype and produce excessive ECM. Interestingly, HOXA11-AS overexpression, which is a lncRNA regulating expression of HOXA11, has also been observed to increase the TGF-β1 gene expression in mouse cardiac fibrosis model[106]. HOXA11-AS have also been implicated in acceleration of keloid formation via acting as a sponge of miR-124-3p, which is a microRNA repressing TGFβ receptor 1 gene (TGFBR1) expression, thus upregulating TGFβ/phosphatidylinositol 3-kinase-protein kinase B (PI3K-Akt) signaling, inhibiting apoptosis of keloid fibroblasts, thereby contributing to keloid formation[107]. A study by Ganesan et al[108] have also identified that in myelofibrosis, MSCs have aberrantly high expression of HOXB7 as a result of TGF-β1 stabilizing β-catenin, which then binds to the HOXB7 promoter region, increasing HOXB7 expression. HOXB7 elevation has also been observed to induce downstream fibrogenic genes [e.g., alpha-smooth muscle actin (α-SMA)] expression[108]. HOTAIR, which is another lncRNA transcribed from the HOXC locus, was found to promote cardiac fibroblast activation via Wnt signaling by upregulating unconventional prefoldin RPB5 interactor 1 gene (URI1) expression[12]. HOTAIR has also been identified as a key factor in the unilateral ureteral obstruction (UUO) model of renal interstitial fibrosis, by acting as a competing endogenous RNA for miR-124, which is a post-transcriptional repressor of Jagged1 (JAG1) ligand of Notch signaling[109]. Thus, by reducing the miR-124 availability in an in vitro kidney fibrosis model, HOTAIR overexpression results in an increase in α-SMA and fibronectin expression, through the Notch signaling pathway, but a decrease in E-cadherin[109]. HOXA9 has also been observed to decrease in keloid lesions, and its overexpression has been shown to reduce cell migration, increase matrix metalloproteinase 3 (MMP3) expression as well as regulate Wnt pathway inhibitors, reducing planar cell polarity protein 1 gene (PRICKLE1) and increasing dickkopf-1 gene (DKK1) expression. HOTTIP, a lncRNA transcribed from HOXA locus, have been implicated in lung fibrosis via downregulating miR-744-5p expression, thereby increasing the polypyrimidine tract binding protein 1 gene (PTBP1) expression[110]. PTBP1 is a downstream effector of TGF-β1 and its overproduction has been shown to increase transcription of fibrosis-related ECM proteins, such as collagen type 1, collagen type 3 and fibronectin[111]. Additionally, HOTTIP has been shown to downregulate miR-148a, which is an inhibitor of TGFBR1 and TGFBR2, in a mouse hepatic stellate cell activation model, thereby contributing to liver fibrosis[112]. miR-10a and miR-10b, microRNAs transcribed from HOXB and HOXD loci respectively, play a role in TGF-β signaling through directly targeting TGFBR1 mRNA, thereby reducing its expression[113]. In diabetic mice and diabetic kidney disease models, renal miR-10a/b were downregulated while TGFBR1 was upregulated and restoring miR-10a/b via lentiviral delivery in streptozocin-diabetic mice reduced TGFBR1 levels, Smad3 activation, and fibrosis markers (fibronectin, α-SMA), attenuating collagen deposition[114]. Conversely, miR-10 family has been found to contribute to chronic kidney disease fibrosis by targeting vasohibin 1 mRNA, which is an inhibitor of TGFB1[115,116]. Another HOX locus, which transcribes miR-196a and 196b, has been identified as targeting TGFBR2 mRNA[117]. In a UUO-induced mouse renal fibrosis model, miR-196a/b levels have rapidly decreased, coinciding with TGFBR2 expression increase, and enforcing miR-196a/b expression in UUO mice suppressed Smad2/3 phosphorylation and lowered α-SMA and collagen-I deposition, mitigating fibrosis. A MEINOX protein, MEIS1, is upregulated in kidney myofibroblasts of human chronic kidney disease patients and mouse models, with a clear negative correlation with fibrosis severity[6]. Mechanistically, MEIS1 increases transcription of protein tyrosine phosphatase receptor J gene (PTPRJ), which is a negative regulator of TGFBR1[6,118]. Curiously, fibroblast-specific forced overexpression of MEIS1 have been found to ameliorate renal fibrosis in multiple kidney fibrosis models (UUO, ischemia-reperfusion, folic acid injury), hinting to its fibroprotective capabilities[6]. These described molecular relationships between HOX-MEINOX molecules, fibrosis-associated pathways, and resultant fibrogenic mechanisms are concisely summarized in Table 1. A deeper reflection of literature with these molecules proposed mechanism of action in fibrosis can be found in Table 2 for protein coding HOX genes, Table 3 for non-coding HOX genes and Table 4 for MEINOX genes.

Figure 3 HOX-MEINOX molecules’ involvement in fibrosis-related pathways. HOX (yellow boxes) and MEINOX (gray boxes), associated long non-coding RNAs (orange boxes) and HOX-associated microRNAs (blue boxes) are involved in different steps in fibrosis-related signaling pathways, resulting in various fibrosis-related responses. Created in BioRender. TGF: Transforming growth factor; PDGF: Platelet derived growth factor; VASH: Vasohibin; PTBP1: Polypyrimidine tract binding protein 1; JAG1: Jagged1; EZH2: Enhancer of zeste homolog 2; ECM: Extracellular matrix; DKK1: Dickkopf-1.

Table 1 HOX and MEINOX molecules in fibrosis related mechanisms and associated pathways.

Mechanism	HOX-MEINOX molecule	Associated pathway(s)	Effect of molecule on pathway	Ref.
Myofibroblast differentiation	HOXA9	Wnt/β-catenin	Inhibition	[79]
	miR-196a/b	TGF-β	Inhibition	[117]
	miR-10a/b	TGF-β	Inhibition	[197]
	HOTTIP	TGF-β	Stimulation	[112]
ECM production	HOTTIP	TGF-β	Inhibition	[110]
	HOXA11	PDGF	Stimulation	[105]
	HOXA11	TGF-β	Stimulation	[106]
	HOXB7	Wnt/β-catenin	Stimulation	[108]
	HOXA9	Wnt/β-catenin	Stimulation	[79]
	HOXA11-AS	TGF-β	Stimulation	[107]
Fibroblast migration	HOXA11	PDGF	Stimulation	[106]
Fibroblast survival	MEIS1	TGF-β	Inhibition	[6]
	HOTTIP	TGF-β	Stimulation	[110,112]
Fibroblast activation	miR-10a/b	TGF-β	Stimulation	[114,197]
	HOXA11	TGF-β	Stimulation	[105,107]
	HOTAIR	Notch	Stimulation	[91]

TGF: Transforming growth factor; PDGF: Platelet derived growth factor; ECM: Extracellular matrix.

Table 2 The effects of protein-coding HOX genes in various fibrotic diseases.

HOX genes	Disease	Mode of function in fibrosis	Proof of concept or mechanism of action	Ref.
HOXA2	Liver fibrosis	Anti-fibrotic	Gene silencing of HOXA2 via DNA hypermethylation correlates with advanced fibrosis in chronic hepatitis B	[198]
	Lung fibrosis	Pro-fibrotic	Expression is significantly upregulated in mesenchymal stromal cells from patients with progressive idiopathic pulmonary fibrosis	[199]
HOXA5	Lung fibrosis	Pro-fibrotic	HOXA5 drives DNM3OS transcription, which recruits EZH2 to suppress TSC2, promoting fibroblast proliferation, migration, and ECM gene expression	[169]
HOXA9	Skin fibrosis	Anti-fibrotic	In keloids, overexpression of HOXA9 reduces cell migration, increases MMP3 expression, regulates Wnt pathway inhibitors (reducing PRICKLE1, increasing DKK1 expression)	[79]
HOXB7	Lung fibrosis	Pro-fibrotic	Found to be upregulated in IPF patients’ lung tissues	[175]
HOXB13	Liver fibrosis	Pro-fibrotic	The number of HOXB13+ cells in fibrotic liver increases. HOXB13 expression is correlated with increased hepatic inflammatory activity, but not with fibrosis stages	[165]
HOXC8	Liver fibrosis	Pro-fibrotic	Inhibition of HOXC8 suppressed the hepatic stellate cell activation and the expression of fibrosis-associated genes (α-SMA and COL1A1)	[187]
HOXD10	Kidney fibrosis	Anti-fibrotic	In a mouse kidney fibrosis model, HOXD10 overexpression significantly reduced collagen deposition and renal dysfunction	[200]

DNM3OS: Dynamin 3 opposite strand; EZH2: Enhancer of zeste homolog 2; TSC2: Tuberous sclerosis complex subunit 2; ECM: Extracellular matrix; MMP: Matrix metalloproteinase; DKK1: Dickkopf-1; IPF: Idiopathic pulmonary fibrosis; α-SMA: Alpha-smooth muscle actin; COL1A1: Collagen type 1.

Table 3 The effects of non-coding HOX genes in various fibrotic diseases.

HOX genes	Disease	Mode of function in fibrosis	Proof of concept or mechanism of action	Ref.
HOTAIR	Liver fibrosis	Pro-fibrotic	HOTAIR upregulation was found to be promoting liver fibrosis in mouse and cell line models. Arsenic exposure was found to induce hepatic fibrosis via T-cell expression of HOTAIR	[177,201,202]
	Lung fibrosis	Pro-fibrotic	Acts as a competing ceRNA to regulate MMP2 expression during paraquat-induced lung epithelial-mesenchymal transition, promoting fibrosis	[203]
	Kidney fibrosis	Pro-fibrotic	In a rat model with TGFβ-treated kidney cells, HOTAIR was elevated in renal fibrosis, driving epithelial-mesenchymal transition via the Notch pathway	[109]
	Cardiac fibrosis	Pro-fibrotic	Upregulated in atrial fibrillation and promotes pathological fibrosis in the atrium. In Ang II-treated atrial fibroblasts, HOTAIR and Wnt5a levels increased, and HOTAIR knockdown inhibited fibroblast proliferation, migration, and expression of collagen I/III and α-SMA	[12]
	Skin fibrosis	Pro-fibrotic	In systemic sclerosis dermal fibroblasts, HOTAIR is aberrantly upregulated and induces a pro-fibrotic gene program. Overexpression of HOTAIR in healthy skin fibroblasts caused epigenetic silencing of miR-34a via EZH2 (H3K27me3), which activated Notch signaling and secondarily upregulated the Hedgehog effector GLI2	[91,204]
HOTAIRM1	Lung fibrosis	Pro-fibrotic	Hypoxia-exposed alveolar epithelial cells secrete pro-fibrotic exosomes enriched in lncRNA HOTAIRM1	[205]
HOTTIP	Liver fibrosis	Pro-fibrotic	Promotes hepatic stellate cell activation through increased SRF expression by sponging miR-150	[206]
	Liver fibrosis	Pro-fibrotic	Highly upregulated in fibrotic livers and activated hepatic stellate cells. Increased HOTTIP acts as a sponge for miR-148a, relieving repression of TGFβ receptor genes. Consequently, TGFBR1/2 levels rise, driving HSC activation and collagen production	[112]
	Lung fibrosis	Pro-fibrotic	Enhances lung fibrosis by regulating the miR-744-5p/PTBP1 signaling axis	[110]
HOXA11-AS	Liver fibrosis	Pro-fibrotic	Upregulated in a mouse model of ischemia/reperfusion-induced liver fibrosis and exacerbates fibrosis via a PTBP1/HDAC4 mechanism	[167]
	Cardiac fibrosis	Pro-fibrotic	Drives TGFβ1-mediated cardiac fibroblast activation. Overexpression of HOXA11-AS in mouse cardiac fibroblasts significantly increased TGF-β1 expression and downstream Smad signaling, promoting fibroblast proliferation, colony formation, and invasion, whereas HOXA11-AS knockdown had opposite effects	[106]
	Skin fibrosis	Pro-fibrotic	Significantly overexpressed in keloid tissue and fibroblasts, promotes abnormal scar formation by acting as a ceRNA	[207]
miR-10a	Liver fibrosis	Pro-fibrotic	In a CCl4-induced mouse liver fibrosis model, miR-10a was found to exacerbate fibrosis. Overexpression of miR-10a increased TGF-β1/Smad3 signaling and collagen-I expression, whereas miR-10a inhibition ameliorated liver fibrosis	[197]
miR-10a/b	Kidney fibrosis	Pro-fibrotic	miR-10 family (miR-10a/10b, located in HOX clusters) is aberrantly expressed in chronic kidney disease. In fibrotic mouse kidneys, miR-10a/b levels are elevated and contribute to fibrosis by targeting VASH1, an anti-angiogenic factor that also inhibits TGF-β/Smad signaling	[114]
miR-196a	Lung fibrosis	Anti-fibrotic	In a bleomycin-induced fibrosis model, lncRNA H19 was upregulated and sponged miR-196a, thereby relieving suppression of COL1A1. Silencing H19 reduced fibroblast activation and collagen deposition, an effect reversed by miR-196a inhibition	[208]
miR-615-5p	Liver fibrosis	Pro-fibrotic	miR-615-5p expression is significantly upregulated in cirrhotic livers and in plasma of patients with advanced fibrosis compared to healthy controls	[209]

ceRNA: Competitive endogenous RNA; MMP: Matrix metalloproteinase; TGF: Transforming growth factor; α-SMA: Alpha-smooth muscle actin; EZH2: Enhancer of zeste homolog 2; GLI2: Glioma-associated oncogene homolog 2; SRF: Serum response factor; TGFBR1/2: Transforming growth factor-beta receptor 1/2; PTBP1: Polypyrimidine tract-binding protein 1; HDAC4: Histone deacetylation 4; VASH1: Vasohibin 1; lncRNA: Long non-coding RNA; COL1A1: Collagen type 1.

Table 4 The effects of MEINOX genes in various fibrotic diseases.

MEINOX genes	Disease	Mode of function in fibrosis	Proof of concept or mechanism of action	Ref.
MEIS1	Kidney fibrosis	Pro-fibrotic	MEIS1 is strongly induced in PDGFRβ+ pericytes as they transition into myofibroblasts during acute and chronic kidney injury	[210]
	Kidney fibrosis	Anti-fibrotic	In mice, fibroblast-specific MEIS1 overexpression inhibited myofibroblast activation and attenuated renal fibrosis, whereas MEIS1 knockout in fibroblasts worsened fibrosis	[6]
PKNOX2	Kidney fibrosis	Pro-fibrotic	TGF-β1 treatment induces PKNOX2 in fibroblasts, and PKNOX2 appears to support fibrogenesis by enhancing fibroblast survival	[93]
	Cardiac fibrosis	Anti-fibrotic	In healthy human heart, PKNOX2 is associated with normal fibroblast activation, but in failing heart, its expression drops during fibroblast-to-myofibroblast transition	[11]
	Myelofibrosis	Anti-fibrotic	PKNOX2 expression is significantly downregulated in Fanconi anemia patients’ bone marrow MSCs compared to healthy controls	[43]

PDGFRβ: Platelet-derived growth factor receptor beta; TGF: Transforming growth factor; MSC: Mesenchymal stem cell.

Collectively, these findings emphasize the pivotal roles of HOX-MEINOX transcriptional complexes as central integrators of multiple fibrosis-driving signaling pathways. Through dynamic transcriptional regulation, these complexes mediate crosstalk among critical fibrotic signaling cascades, including TGF-β, Wnt, PDGF, and Notch. For example, HOXB7 exemplifies the direct convergence of TGF-β signaling with Wnt/β-catenin activity, where TGF-β1-induced stabilization of β-catenin results in enhanced HOXB7 expression, subsequently driving profibrotic outcomes[108]. Similarly, lncRNA HOTAIR connects TGF-β and Notch pathways, further sustaining myofibroblast differentiation and ECM deposition[109]. This intricate signaling network not only perpetuates fibrosis progression through synergistic effects on EMT and ECM remodeling but also underscores the therapeutic promise of targeting HOX-MEINOX complexes as a strategy capable of concurrently modulating several key pathogenic pathways.

HOX AND MEINOX IN CANCER

As previously discussed, HOX and MEINOX molecules play critical roles in the regulation of positional identity and cellular plasticity - fundamental biological concepts necessary for proper development, differentiation, and adult tissue maintenance. When dysregulated, these molecules are capable of profound disruption of tissue homeostasis, thus creating a fertile ground for pathological transformations. In the context of cancer, disruption of cellular plasticity and loss or aberration of positional identity can drive oncogenesis as well as malignant progression, metastasis, and therapeutic resistance. Cancer cells often exploit cellular plasticity to dynamically transition between differentiated states and highly plastic, stem-like phenotypes, facilitating adaptation to changing microenvironments and therapeutic challenges. Similarly, the disruption or misinterpretation of positional cues by dysregulated HOX and MEINOX expression not only alters cell fate decisions but also promotes invasive and metastatic properties, characteristic of aggressive malignancies. These positional disruptions may also facilitate abnormal signaling interactions between cancer cells and their surrounding stroma, reshaping the tumor microenvironment (TME) to favor malignancy[119]. This section will explore how HOX and MEINOX dysregulation mechanistically contributes to cancer by influencing positional identity and cellular plasticity, then will focus on current insights into the research involving various cancers and HOX and MEINOX molecules.

HOX and MEINOX proteins can profoundly influence cancer progression through their disruption of positional identity. In normal tissues, positional identity ensures cells differentiate and function correctly within their anatomical context, guided by tightly controlled HOX-MEINOX gene expression patterns established during embryogenesis[38,63,120]. However, the aberrant expression of HOX and MEINOX proteins in cancer disrupts this positional code, leading cells to adopt inappropriate differentiation states and acquire malignant phenotypes[121-123]. For instance, HOXD11 dysregulation has been shown to drive penile squamous cell carcinoma metastasis and cell invasion capabilities through ECM remodeling molecules fibronectin 1/MMP2/MMP9, towards more invasive phenotype by altering positional information critical for maintaining epithelial integrity[124]. Furthermore, dysregulated positional identity can influence the TME, promoting abnormal interactions with stromal and immune cells, thereby enhancing tumor growth and metastatic potential. In a meta-analysis study focused on HOX-related lncRNAs, Shao et al[125] have found that dysregulation of HOTAIRM1 and HOXB-AS1 has shown strong correlation with oncogenesis via involvement in immune regulation across a variety of cancer types[125]. Specifically, altered positional identity mediated by HOX-MEINOX may contribute to an immunosuppressive environment by modulating immune cell recruitment and polarization, thus facilitating immune evasion. For example, HOXC6 was found to be significantly upregulated in colorectal cancer (CRC) primary tumors, with elevated HOXC6 expression correlating strongly with the cytokine pathway and expression of chemokines, the infiltration ratio of immune cells, expression of immune checkpoint markers, and poorer survival outcomes[126]. Coculture experiments revealed that overexpression of HOXC6 attracted more CD8+ T cells by upregulating specific chemokines, but also downregulated interferon-γ, indicating the cytotoxic function of CD8+ T cells might be exhausted by the downregulation of interferon-γ[126]. Additionally, tumor cells overexpressing HOXC6 exhibited increased immunogenicity and lower expression of mismatch repair gene MLH1, which indicated its significant roles in reorganizing the TME, further highlighting its potential impact on tumor biology and prognosis[126]. A recent pilot study by Pfeiferová et al[76] revealed that cancer-associated fibroblasts (CAFs) originated from the ectomesenchyme-dependent areas, MSCs isolated from glioblastomas and secondary brain tumours exhibited negligible HOX gene expression. Despite the growing knowledge of CAFs on cancer, the impact of their HOX code is an ongoing area of research. An elegantly designed study by Yang et al[127] have used publicly available single-cell RNA sequencing data of human endometrial cancer samples to establish the concept of “HOX score”, by utilizing machine learning algorithms. Within the same study, Yang et al[127] have also identified that high HOX score is correlated with the immune-excluded phenotype whereas the immune-inflamed phenotype characterized by high anti-tumor immune cell infiltration, has a lower HOX score. Collectively, these examples illustrate how positional identity dysregulation via altered HOX-MEINOX expression fundamentally contributes to cancer progression, underscoring the importance of further exploring their roles not only in modulating cellular plasticity and tumorigenesis, but also tumor immunity.

Complementary to positional identity, cellular plasticity is also substantially influenced by HOX and MEINOX dysregulation in cancer. Under normal physiological conditions, cellular plasticity enables dynamic transitions between differentiated and stem-like states, facilitating tissue repair and homeostasis[128]. However, in cancer, aberrantly elevated or impaired plasticity driven by altered HOX-MEINOX expression can promote malignancy by enhancing cancer stem cell (CSC) populations, metastatic dissemination, and resistance to therapy[129]. Dysregulated HOX and MEINOX expressions can activate key plasticity-regulating signaling pathways such as Notch, Wnt, and Hedgehog, driving cancer cells toward more aggressive, therapy-resistant phenotypes[130-132]. For example, HOTAIR was found to be one of the critical regulators of cancer stemness and malignant progression of cutaneous squamous cell carcinoma[133]. In a study, MEIS1 was found to be highly upregulated in glioma stem cells compared to normal and differentiated glioma cells, and inhibition of MEIS1 through short hairpin RNA significantly reduced glioma stem cell growth both in vitro and in vivo, thus showing that its upregulation is related to cancer stemness[134]. In another study, overexpression of HOXA4 and HOXA9 genes were found to be promoting self-renewal, thus contributing to colon CSC overpopulation[135]. Furthermore, increased cellular plasticity mediated by HOX and MEINOX dysregulation can facilitate EMT, promoting invasive and metastatic behavior of cancer cells. Case in point, HOXA10 was found to be the mediator of EMT to promote gastric cancer metastasis through TGFβ2/Smad/methyltransferase-like 3 signaling pathways[136]. Moreover, EMT driven by HOX-MEINOX dysregulation can enhance cancer cell interaction with stromal cells, further reshaping the TME to promote metastatic niche formation. For example, CAFs-derived exosomes with HOXD11 overexpression were found to promote ovarian cancer cell angiogenesis[137]. Similarly, cancer-associated EMT can increase cellular heterogeneity within tumors, augmenting therapeutic resistance by providing multiple cellular states resistant to targeted therapies. In a study, HOXC13 was found to be the main driver behind tumor heterogeneity in castration-resistant prostate cancer[138]. Thus, elucidating the molecular mechanisms by which HOX and MEINOX proteins disrupt positional identity and cellular plasticity in cancer can reveal critical insights into tumor progression and uncover novel therapeutic targets.

To further our understanding of how HOX-MEINOX proteins contribute to cancer, we will briefly highlight major cancer-associated signaling pathways and their mechanistic roles in tumor progression. The Wnt/β-catenin pathway promotes tumor growth, metastasis, and stemness through aberrant activation in numerous cancers. Similarly, TGF-β signaling has a dual role, suppressing early-stage tumor development but promoting EMT, invasion, and metastasis in advanced cancers. Notch signaling enhances tumor cell survival, stemness, and metastatic potential, while PDGF signaling strongly drives proliferation, angiogenesis, and metastasis. Several recent studies provide critical insights into how HOX-MEINOX proteins regulate these signaling cascades in various cancers, outlined in the following examples. Figure 4 also provides a visual map for the effects of following HOX-MEINOX molecules on these pathways and the resultant cancer hallmark mechanisms. In breast cancer, the HOXD-derived lncRNA HAGLR sponges miR-335-3p, thus upregulating Wnt2 and activating Wnt signaling to drive triple-negative breast cancer proliferation and invasion[139]. In glioblastoma, HOXC-derived lncRNA HOTAIR activates Wnt/β-catenin signaling via a miR-214-dependent network, conferring resistance to temozolomide chemotherapy[140]. Similarly, MEIS2 isoforms (C/D) enhance hepatocellular carcinoma (HCC) progression by activating Wnt/β-catenin and Yes-associated protein pathways, respectively, promoting aggressive tumor phenotypes[141]. Multiple HOX proteins modulate TGF-β signaling during cancer progression. HOXA10 induces EMT and metastasis in gastric cancer via direct activation of the TGF-β signaling axis involving Smad and methyltransferase-like 3 (METTL3) pathways[136]. HOXA13 similarly promotes HCC invasion and metastasis, partly through regulation of EMT-related transcription factors and signaling cascades such as TGF-β/bone morphogenetic protein 7 (BMP-7)[142]. Conversely, miR-10a-5p (from the HOXB locus) directly targets and suppresses TGFβ receptor signaling in HCC, functioning as a critical tumor-suppressor microRNA that inhibits migration and metastasis[143,144]. Wnt signaling pathway is modulated by HOX-associated lncRNAs, such as HOTAIR, whose elevated expression in glioblastoma leads to an increase in β-catenin protein level via sponging miR-214-3p[140]. HOXA11-AS, a lncRNA transcribed from the HOXA locus, promotes glioma progression by activating pathways β-catenin/c-Myc cascade through increasing collagen triple helix repeat containing-1 gene ( CTHRC1) expression by sponging let-7b-5p[145]. PDGF signaling, critical for tumor growth and metastatic dissemination, is influenced by HOX dysregulation. HOXA11 overexpression in NSCLC strongly correlates with advanced disease and poor prognosis, likely through mechanisms involving PDGF and related proliferative pathways[146]. Similarly, HOXA1 drives cervical carcinoma progression by directly transactivating glycolytic enzymes ENO1 and PGK1, indirectly affecting PDGF and other growth-factor signaling cascades via metabolic rewiring[147]. Complex regulatory scenarios emerge in other cancers as well. In CRC, MEIS3 overexpression correlates with invasion and metastasis, especially at tumor invasive fronts, suggesting activation of pro-invasive signaling pathways potentially linked to Wnt or TGF-β[148]. HOXB13 plays a context-dependent role in prostate cancer, acting as an androgen receptor cofactor in early stages, but its loss later promotes metastasis through altered lipid metabolism and metastatic pathway activation[149]. Additionally, miR-196a/b (from HOX clusters) is upregulated in HCC and oral squamous cell carcinoma, driving aggressive tumor behavior by suppressing tumor suppressor targets, including suppressors of cytokine signaling 2 (SOCS2), thereby amplifying Janus kinase/signal transducer and activator of the transcription (JAK/STAT) signaling[150,151]. Tumor suppressive functions are also clearly documented. HOXA5 downregulation in NSCLC correlates with poorer patient survival, while restoration of HOXA5 represses Wnt/β-catenin signaling and inhibits cancer progression[146,152]. PKNOX2 acts as a tumor suppressor in lung and gastric cancers, inhibiting proliferation via suppression of phosphatidylinositol 3-kinase/protein kinase B/mechanistic target of rapamycin (PI3K/AKT/mTOR) and stabilization of p53 pathways, respectively[42,153]. Similarly, PKNOX1 downregulation strongly correlates with melanoma progression, indicating its broad tumor-suppressive potential[154,155]. These molecular connections between HOX-MEINOX transcription factors and major cancer-related signaling pathways, along with their final effects on cancer hallmark processes are concisely summarized in Table 5. A more-detailed snapshot of the current literature findings with HOX-MEINOX molecules and their specific mechanism of actions in various cancers can be found in Table 6 for protein coding HOX genes, Table 7 for non-coding HOX genes and Table 8 for MEINOX genes.

Figure 4 HOX-MEINOX molecules’ involvement in cancer-related pathways. HOX (yellow boxes) and MEINOX (grey boxes), HOX associated long non-coding RNAs (orange boxes) and HOX-associated microRNAs (blue boxes) are involved in different steps in cancer-related signaling pathways, partaking in suppression or maintenance of different cancer hallmark processes. Created in BioRender. TGF: Transforming growth factor; PI3K: Phosphatidylinositol 3-kinase; LATS1: Large tumor suppressor 1; YAP: Yes-associated protein; TAZ: Transcriptional co-activator with PDZ-binding motif; IGFBP5: Insulin-like growth factor-binding protein 5; IGF: Insulin-like growth factor; JAK2: Janus kinase 2; SOCS2: Suppressors of cytokine signaling 2; JNK: C-Jun N-terminal kinase; NME4: Nucleoside diphosphate kinase 4; STAT: Signal transducer and activator of the transcription; MGMT: O6-methylguanine-methyltransferase; CTHRC1: Collagen triple helix repeat containing-1; MAPK: Mitogen-activated protein kinases.

Table 5 HOX and MEINOX molecules in cancer hallmark mechanisms and associated pathways.

Mechanism	HOX-MEINOX molecule	Associated pathway(s)	Effect of molecule on pathway	Ref.
Invasion	miR-10a/b	TGF-β	Inhibition	[113,143]
	HOXA10	TGF-β	Stimulation	[136]
	HOXA13	Wnt/β-catenin, TGF-β	Stimulation	[211]
	MEIS2D	Hippo	Inhibition	[141]
	MEIS2C, HAGLR, HOTAIR	Wnt/β-catenin	Stimulation	[139-141]
	HOXA5	Wnt/β-catenin	Inhibition	[212]
	miR-196a/b	MAPK	Stimulation	[151]
Chemoresistance	PKNOX2	PI3K	Inhibition	[42]
	MEIS2C, HAGLR, HOTAIR, HOXA13	Wnt/β-catenin	Stimulation	[139-141,211]
	HOXA5	Wnt/β-catenin	Inhibition	[212]
Proliferation	PKNOX2, HOXA5	p53	Stimulation	[152,153,174,212]
	MEIS2C, HAGLR, HOTAIR	Wnt/β-catenin	Stimulation	[139-141]
	HOXA13	Wnt/β-catenin, TGF-β	Stimulation	[211]
	miR-196a/b	JAK/STAT	Stimulation	[150]
	HOXA11-AS	β-catenin/c-Myc	Stimulation	[145]
	MEIS2D	Hippo	Inhibition	[141]
Transformation	HOXA10	TGF-β	Stimulation	[136]
Evading apoptosis	miR-196a/b	JAK/STAT, MAPK	Stimulation	[150,151]
	HOXA11-AS	β-catenin/c-Myc	Stimulation	[145]
	PKNOX2	PI3K	Inhibition	[42]
	PKNOX2	p53	Stimulation	[153]
	HOXA5	p53	Stimulation	[152,212]
Immune evasion	HOXA11-AS	β-catenin/c-Myc	Stimulation	[145]

TGF: Transforming growth factor; MAPK: Mitogen-activated protein kinases; PI3K: Phosphatidylinositol 3-kinase; JAK: Janus kinase; STAT: Signal transducer and activator of the transcription.

Table 6 The effects of protein-coding HOX genes in various cancers.

HOX genes	Disease	Mode of expression	Mode of function in cancer	Proof of concept or mechanism of action	Ref.
HOXA1	Breast cancer	Upregulated	Oncogenic	HOXA1 is overexpressed in breast tumors and correlates with advanced disease and poor patient survival. HOXA1 knockdown in breast cancer cells induces cell cycle arrest and apoptosis, suggesting it promotes tumor growth	[213]
	Cervical cancer	Upregulated	Oncogenic	HOXA1 is highly expressed in cervical carcinoma. HOXA1 directly transactivates glycolytic enzymes ENO1 and PGK1, enhancing aerobic glycolysis and promoting cervical cancer cell growth and metastasis. HOXA1 knockdown impairs tumor growth and increases chemosensitivity	[147]
HOXA2	Breast cancer	Downregulated	Tumor suppressor	HOXA2 is frequently silenced by promoter hypermethylation in breast tumors. Restoring HOXA2 inhibits breast cancer cell proliferation and motility	[198]
HOXA3	OSCC	Dysregulated	Context-dependent	HOXA3 shows stage-specific expression changes in oral tumorigenesis; upregulated in dysplastic lesions but then downregulated in advanced OSCC. Hypermethylation of the HOXA3 3’UTR in OSCC was linked to worse overall survival	[214]
HOXA5	Colorectal cancer	Downregulated	Tumor suppressor	HOXA5 is frequently hypermethylated and silenced in colorectal cancers. Loss of HOXA5 correlates with poor differentiation; demethylation can restore HOXA5 expression, supporting a tumor-suppressor role	[193]
	NSCLC	Downregulated	Tumor suppressor	HOXA5 exhibits reduced expression in NSCLC. A meta-analysis found that high HOXA5 is associated with increased overall survival in NSCLC, consistent with a tumor-suppressor function	[146,152]
	HCC	Downregulated	Tumor suppressor	HOXA5 is significantly downregulated in HCC tissues. Low HOXA5 levels associate with larger tumor size, high AFP, and predict worse overall and recurrence-free survival. HOXA5 acts as a tumor suppressor: Restoring HOXA5 (or inhibiting its upstream repressor miR-130b-3p) restrains HCC angiogenesis and growth. In HCC cells, loss of HOXA5 Leads to increased VEGF and microvessel formation, whereas HOXA5 overexpression inhibits these pro-tumorigenic processes	[215]
HOXA7	ESCC	Upregulated	Oncogenic	HOXA7 is one of several HOX genes significantly overexpressed in ESCC tumor tissue (vs normal esophagus). High HOXA7 Levels are associated with worse overall survival in ESCC patients	[216]
HOXA9	AML	Upregulated	Oncogenic	Overexpressed in > 50% of AML cases; drives leukemogenesis and correlates with poor prognosis. HOXA9 forms a complex with SAFB to repress differentiation genes, and its disruption induces differentiation and apoptosis	[217]
	NSCLC	Downregulated	Oncogenic	Transient transfection of HOXA9 into H23 Lung cancer cells resulted in the inhibition of cell migration but not proliferation	[218]
	HCC	Upregulated	Oncogenic	HOXA9 is dramatically upregulated in HCC and its high expression predicts poor patient survival. HOXA9 may be controlled by RPL38 and is involved in epigenetic and immune-regulatory networks in HCC. Targeting HOXA9 (e.g., via siRNA) led to suppressed tumor growth and induced apoptosis in vitro	[188]
HOXA10	HCC	Upregulated	Oncogenic	HOXA10 is one of the most overexpressed HOX genes in HCC and liver tumor-initiating cells. A long noncoding RNA “lncHOXA10” drives HOXA10 transcription, which in turn promotes self-renewal of liver cancer stem cells and tumorigenesis. Knocking out HOXA10 impairs sphere formation and tumor propagation, confirming its tumor-promoting role	[219]
	PDAC	Upregulated	Oncogenic	HOXA10 is significantly overexpressed in PDAC and is associated with higher tumor stage and shorter survival. HOXA10 overactivity drives pancreatic cancer progression by directly activating the NF-κB signaling pathway, thereby promoting tumor cell proliferation and invasion. HOXA10 silencing can reduce PDAC cell aggressiveness	[220]
HOXA11	NSCLC	Upregulated	Oncogenic	HOXA11 is overexpressed in some NSCLC cohorts and has been linked to worse outcomes. HOXA11 was identified as an independent predictor of poor overall survival in NSCLC	[146]
HOXA13	HCC	Upregulated	Oncogenic	HOXA13 is significantly overexpressed in HCC. High HOXA13 correlates with advanced disease and poor outcome - patients with elevated HOXA13 had more metastases and shorter survival	[142]
HOXB5	HCC	Upregulated	Oncogenic	HOXB5 is aberrantly elevated in HCC. Its high expression correlates with poor differentiation, higher stage, and worse prognosis. HOXB5 acts as a metastasis promoter: It transactivates FGFR4 and CXCL1 to drive HCC cell invasion and myeloid suppressor cell recruitment. Knockdown of HOXB5 or its targets suppresses lung and liver metastases in mice, confirming HOXB5 as a pro-metastatic oncogene in HCC	[189]
HOXB7	HCC	Upregulated	Oncogenic	HOXB7 is highly overexpressed in HCC tumors compared to normal liver. HOXB7 overexpression correlates with poor patient survival and aggressive disease. HOXB7 enhances proliferation, sphere formation (stemness), migration and invasion of HCC cells, while HOXB7 knockdown has the opposite effect. Mechanistically, HOXB7 activates the AKT pathway; it upregulates c-Myc and Slug to promote EMT and cancer stem cell traits	[190]
HOXB13	HCC	Upregulated	Oncogenic	HOXB13 is overexpressed in a subset of HCC cases. High HOXB13 has been shown to enhance HCC cell proliferation, metastasis, and chemoresistance	[166]
	PCa	Downregulated	Tumor suppressor	HOXB13 plays a complex role in prostate cancer; In early, androgen-dependent disease it is an important AR cofactor, but in castration-resistant PCa, loss of HOXB13 drives metastasis. HOXB13 recruits HDAC3 to suppress lipid biosynthesis; loss of HOXB13 (or the germline G84E mutant) causes abnormal lipid accumulation, which increases cell motility and metastasis	[149,191]
HOXC4/HOXC6	PCa	Upregulated	Oncogenic	HOXC4 is overexpressed in prostate tumors, especially in high-grade disease. Genomic analyses identified HOXC4 (and HOXC6) as part of a gene signature associated with aggressive prostate cancer. HOXC4 overexpression has been linked to increased proliferation of prostate cancer cells, and HOXC4/HOXC6 mRNA in urine has been tested as a biomarker to predict high-risk prostate cancer	[191]
HOXC13	ESCC	Upregulated	Oncogenic	HOXC13 is significantly overexpressed in ESCC tumors relative to normal tissue	[221]
HOXD10 (through miR-10b)	Gastric cancer	Downregulated	Tumor suppressor	HOXD10 expression is reduced through miR-10b in gastric cancer, and its low expression correlates with better outcomes. Restoration of HOXD10 can inhibit tumor cell migration	[222,223]
HOXD10/HOXC9	PTC	Downregulated	Tumor suppressor	HOXD10 and HOXC9 are significantly downregulated in PTC compared to normal thyroids. Lower HOXD10 and HOXC9 expression in PTC is associated with greater invasiveness - including higher incidence of lymph node metastasis and extrathyroid extension	[224]
Multiple HOX genes (HOXA6, HOXC6, HOXD9/HOXD10/HOXD13)	HCC	Upregulated	Oncogenic	A systematic analysis found widespread upregulation of HOX family genes in HCC. Notably, HOXA6, HOXC6, HOXD9, HOXD10, and HOXD13 were among the most overexpressed and each was an independent risk factor for poor overall survival. HCC tissues show higher total HOX mRNA levels than normal liver, reflecting a global reactivation of HOX clusters in liver carcinogenesis	[225]

OSCC: Oral squamous cell carcinoma; UTR: Untranslated region; NSCLC: Non small cell lung cancer; HCC: Hepatocellular carcinoma; AFP: Alpha-fetoprotein; VEGF: Vascular endothelial growth factor; ESCC: Esophageal squamous cell carcinoma; AML: Acute myeloid leukemia; SAFB: Scaffold attachment factor B; siRNA: Small interfering RNA; PDAC: Pancreatic ductal adenocarcinoma; NF-κB: Nuclear factor-kappa B; FGFR4: Fibroblast growth factor-4 receptor; CXCL1: C-X-C motif ligand 1; AKT: Protein kinase B; EMT: Epithelial-mesenchymal transition; PCa: Prostate cancer; PTC: Papillary thyroid cancer.

Table 7 The effects of non-coding HOX genes in various cancers.

HOX genes	Disease	Mode of expression	Mode of function in cancer	Proof of concept or mechanism of action	Ref.
HAGLR	Breast cancer	Upregulated	Oncogenic	HAGLR is highly expressed in triple-negative breast cancer. HAGLR sponges miR-335-3p, leading to upregulation of Wnt2 and activation of Wnt signaling to promote TNBC cell proliferation, invasion, and tumor growth	[139]
HOTAIR	Glioblastoma (brain cancer)	Upregulated	Oncogenic	The lncRNA HOTAIR is highly upregulated in temozolomide-resistant glioblastoma. HOTAIR overexpression in GBM cells activates Wnt/β-catenin signaling and increases MGMT levels (via a HOTAIR/miR-214/β-catenin network), conferring chemoresistance. Silencing HOTAIR restores TMZ sensitivity, indicating HOTAIR drives drug resistance and tumor progression	[140]
	HCC	Upregulated	Oncogenic	The HOXC-derived lncRNA HOTAIR is overexpressed in HCC tissues and is strongly linked to cancer progression. High HOTAIR levels associate with advanced TNM stage, vascular invasion, poor differentiation, and shorter overall and relapse-free survival. HOTAIR is higher in tumors than normal liver and correlates with aggressive phenotypes	[226]
HOTAIRM1	AML	Upregulated	Oncogenic	HOTAIRM1 is significantly overexpressed in AML with NPM1 mutation and promotes leukemic cell proliferation. High HOTAIRM1 is associated with poorer outcomes in intermediate-risk AML	[227]
HOTTIP	HCC	Upregulated	Oncogenic	HOTTIP is highly expressed in HCC tumors. HOTTIP expression strongly correlates with HOXA13 Levels, forming a positive feedback loop. High HOTTIP is associated with increased metastasis. Silencing HOTTIP in HCC cells reduces HOXA13 and inhibits cell proliferation	[142]
	SCLC	Upregulated	Oncogenic	HOTTIP is focally amplified and overexpressed in SCLC, correlating with advanced stage and poor prognosis. HOTTIP acts as an oncogene by sponging miR-574-5p and upregulating EZH1, thereby promoting SCLC cell proliferation and cell-cycle progression. Knocking down HOTTIP in SCLC models impairs tumor growth	[228]
HOXA11-AS	Glioma (high-grade)	Upregulated	Oncogenic	HOXA11-AS is significantly overexpressed in high-grade gliomas and correlates with poor prognosis. It acts as a ceRNA, sponging tumor-suppressive let-7b-5p to upregulate CTHRC1/c-Myc, also scaffolds with c-Jun to activate the TPL2-MEK1/2-ERK1/2 pathway	[145]
	HCC	Upregulated	Oncogenic	HOXA11-AS is significantly upregulated in HCC tumors and cell lines. It acts as a ceRNA, sponging miR-506-3p, thereby de-repressing the EMT transcription factor Slug. Through this miR-506/Slug axis, HOXA11-AS promotes HCC cell proliferation, invasion and epithelial-mesenchymal transition. Knockdown of HOXA11-AS inhibits these malignant behaviors	[168]
HOXB-AS1	HCC	Upregulated	Oncogenic	HOXB-AS1 is highly upregulated in HCC tissues and patient serum. Silencing HOXB-AS1 in HCC cell lines (Hep3B, Huh7) markedly reduces proliferation, migration, and invasion. Clinically, high HOXB-AS1 is associated with lower survival	[229]
miR-10a	PDAC	Upregulated	Oncogenic	miR-10a is overexpressed in a subset of pancreatic cancers and promotes an invasive, metastatic phenotype. In PDAC cells, miR-10a-5p enhances migration and invasion and its inhibition reduces metastasis. Notably, retinoic acid receptor antagonists can repress miR-10a, leading to reduced invasion. miR-10a likely exerts its pro-metastatic effect by suppressing HOX genes (e.g., HOXB1/B3) that restrain motility	[230]
	HCC	Downregulated	Tumor suppressor	miR-10a plays a context-dependent role in HCC. It is significantly downregulated in HCC tissues and cell lines, especially in metastatic tumors. miR-10a-5p acts as a tumor-suppressor miR: Restoring miR-10a-5p inhibits HCC cell migration, invasion, and EMT, both in vitro and in vivo. It directly targets spindle and kinetochore-associated protein 1 (SKA1), leading to its mRNA degradation and suppression of pro-metastatic signaling. Thus, loss of miR-10a in HCC unleashes metastasis, whereas its presence restrains tumor spread	[143]
miR-10b	Breast cancer	Upregulated	Oncogenic	miR-10b is significantly overexpressed in metastatic breast cancer cells and was shown to initiate tumor invasion and metastasis. Overexpression of miR-10b drives cell migration/invasion, whereas silencing miR-10b in mouse models inhibits metastasis	[113,182,184,231]
	HCC	Upregulated	Oncogenic	miR-10b is markedly overexpressed in HCC tumor samples and cell lines. Higher miR-10b levels associate with metastatic potential. Overexpression of miR-10b enhances HCC cell proliferation, migration and invasion, whereas inhibition of miR-10b reduces invasiveness. Mechanistically, miR-10b targets and downregulates HOXD10, which in turn upregulates pro-migratory genes RhoC, uPAR, MMP2/9	[232]
miR-196a/b	OSCC	Upregulated	Oncogenic	miR-196a and miR-196b are significantly overexpressed in OSCCs. High miR-196 levels promote cancer cell migration and invasion, contributing to an invasive tumor phenotype. Clinically, elevated miR-196a/b in oral tumors correlates with advanced tumor stage and nodal metastasis. These “metastamiRs” target multiple genes (including HOX genes) to drive tumor progression	[151]
	HCC	Upregulated	Oncogenic	The miR-196a and miR-196b are upregulated in HCC tissues and cell lines. High miR-196a/b levels are linked to aggressive tumor features. miR-196a/b act as oncogenic miRNAs by targeting negative regulators of growth. miR-196a/b directly suppresses SOCS2, a tumor suppressor that normally inhibits the JAK/STAT pathway. Consequently, miR-196 overactivity leads to unchecked JAK/STAT signaling and tumor progression. Experimental downregulation of miR-196a or miR-196b was shown to inhibit HCC proliferation and metastasis by de-repressing SOCS2 and dampening JAK/STAT signaling. Clinically, elevated miR-196a is associated with HCC metastasis and poor prognosis	[150]

TNBC: Triple negative breast cancer; lncRNA: Long non-coding RNA; GBM: Glioblastoma; MGMT: O6-methylguanine-methyltransferase; TMZ: Temozolomide; HCC: Hepatocellular carcinoma; TNM: Tumor-node-metastasis; AML: Acute myeloid leukemia; NPM1: Nucleophosmin 1; SCLC: Small cell lung cancer; EZH1: Enhancer of zeste homolog 1; ceRNA: Competitive endogenous RNA; CTHRC1: Collagen triple helix repeat containing-1; TPL2: Tumor progression locus 2; MEK1/2: Mitogen-activated protein kinase kinase1/2; ERK1/2: Extracellular signal-regulated kinase1/2; EMT: Epithelial-mesenchymal transition; PDAC: Pancreatic ductal adenocarcinoma; uPAR: Urokinase plasminogen activator receptor; MMP: Matrix metalloproteinase; OSCC: Oral squamous cell carcinoma; miRNA: MicroRNA; SOCS2: Suppressors of cytokine signaling 2; JAK: Janus kinase; STAT: Signal transducer and activator of the transcription.

Table 8 The effects of MEINOX genes in various cancers.

HOX genes	Disease	Mode of expression	Mode of function in cancer	Proof of concept or mechanism of action	Ref.
MEIS1	AML	Upregulated	Oncogenic	MEIS1 is highly expressed in AML and promotes a stem cell-like program. High MEIS1 Levels predict shorter survival and chemo-resistance. Targeting MEIS1/PBX interaction is explored as therapy	[233]
	HCC	Downregulated	Tumor suppressor	MEIS1 is often decreased in HCC. Patients with higher MEIS1 experienced significantly longer time-to-progression after ablation. In a rodent HCC model, adding MEIS1 enhanced the tumor-killing effect of radioablation. MEIS1 acts as a negative regulator of HCC, and low MEIS1 permits aggressive tumor behavior, while high MEIS1 is favorable for prognosis	[234]
MEIS2	Breast cancer	Downregulated	Tumor suppressor	MEIS2 acts as a tumor suppressor in breast cancer. Its expression is reduced in breast tumors, and MEIS2 loss correlates with tumor progression. In cell-line and xenograft models, restoring MEIS2 suppresses proliferation and invasion through downregulation of IL10	[235]
MEIS2C/D	HCC	Upregulated	Oncogenic	The MEIS2 isoforms C and D are overexpressed in HCC tumors vs adjacent liver. Elevated MEIS2C/D correlates with worse prognosis in HCC patients. MEIS2C/D knockdown markedly inhibits HCC cell proliferation, migration, and invasion (in vitro and in mice), whereas MEIS2 overexpression accelerates tumor growth. MEIS2C activates Wnt/β-catenin signaling (with CDC73), and MEIS2D activates YAP by suppressing Hippo signaling - together promoting HCC progression	[141]
MEIS3	CRC	Upregulated	Oncogenic	MEIS3 is overexpressed at the invasive front of CRC tumors and in tumor buds. Higher MEIS3 correlates with advanced stage and worse 5-year disease-free survival. Functional assays showed MEIS3 promotes CRC cell migration and invasion	[148]
	HCC	Upregulated	Oncogenic	MEIS3 is aberrantly expressed in HCC and has been implicated as a pro-metastatic factor. High MEIS3 expression promotes HCC cell migration and invasion and is associated with higher recurrence rates in postoperative patients	[236]
PKNOX1	Melanoma (cutaneous)	Downregulated	Tumor suppressor	PKNOX1 is absent or strongly downregulated in about 70% of diverse human cancers. In mouse models, PKNOX1 deficiency leads to spontaneous development of lymphomas and carcinomas, confirming a tumor-suppressor role. In melanoma, a specific lncRNA (lnc-PKNOX1-1), encoded from PKNOX1 gene locus shown to inhibit melanoma progression	[155]
PKNOX2	AML	Downregulated	Tumor suppressor	In a mouse AML model, PKNOX2 was downregulated in KRAS-mutant model	[171]
	NSCLC	Downregulated	Tumor suppressor	PKNOX2 is frequently downregulated in lung cancer via promoter methylation. Restoring PKNOX2 suppresses NSCLC cell proliferation by inhibiting the PI3K/AKT/mTOR pathway. Low PKNOX2 is associated with poorer prognosis in lung cancer, and PKNOX2 is proposed as a tumor suppressor in both lung and gastric cancers	[42]
	GC	Downregulated	Tumor suppressor	PKNOX2 expression is often silenced via hypermethylation in GC. In vivo, PKNOX2 activates the transcription of IGFBP5 and stabilizes p53, thereby inhibiting gastric tumor growth. Low PKNOX2 in GC is linked to increased proliferation and poor prognosis	[153]

AML: Acute myeloid leukemia; HCC: Hepatocellular carcinoma; IL: Interleukin; CDC37: Cell division cycle 37; YAP: Yes-associated protein; CRC: Colorectal cancer; lncRNA: Long non-coding RNA; NSCLC: Non small cell lung cancer; PI3K: Phosphoinositide 3-kinase; AKT: Protein kinase B; mTOR: Mammalian target of rapamycin; GC: Gastric cancer; IGFBP5: Insulin-like growth factor-binding protein 5.

The findings discussed here highlight HOX-MEINOX transcriptional networks as essential nodes integrating various oncogenic signaling pathways. Dysregulated HOX-MEINOX complexes modulate the crosstalk between critical cancer-driving signals such as Wnt, TGF-β, Notch, and PDGF, collectively influencing tumor progression, metastasis, and therapeutic resistance. For instance, HOXA10 demonstrates direct crosstalk by activating TGF-β signaling pathways, promoting EMT, while MEIS isoforms facilitate HCC progression through concurrent activation of Wnt/β-catenin and Hippo/Yes-associated protein pathways[136,141]. Furthermore, ncRNAs such as HOTAIR underscore additional complexity, simultaneously influencing Notch and Wnt signaling pathways to confer drug resistance in glioblastoma[140]. Thus, therapeutic interventions targeting HOX-MEINOX complexes represent promising approaches for simultaneously disrupting multiple oncogenic pathways, potentially overcoming limitations of therapies focused on single signaling axes.

CROSSROADS OF FIBROSIS AND CANCER: A SHARED MECHANISTIC PERSPECTIVE THROUGH FIBROSIS-DRIVEN CANCERS

Fibrosis and cancer are increasingly recognized as interconnected pathological processes, sharing common molecular mechanisms and frequently existing as sequential or overlapping disease states. Chronic fibrotic conditions such as liver fibrosis, IPF, and primary myelofibrosis significantly heighten the risk for developing HCC, lung cancer, and acute myeloid leukemia (AML), respectively. Central to the transition from chronic fibrosis to malignant transformation across these diverse tissues stand three fundamental mechanistic drivers: Loss of positional identity, altered cellular plasticity, and extensive ECM remodeling. Critically, these processes are tightly regulated by developmental transcription factors, particularly HOX proteins and their MEINOX cofactors, whose aberrant expression profoundly disrupts tissue homeostasis. This section will discuss how HOX-MEINOX dysregulation may partake in cancer initiation or suppression at fibrotic sites, by separately examining the roles of positional identity disruption and cellular plasticity alterations, exemplifying each from liver fibrosis, IPF, chronic pancreatitis, and primary myelofibrosis contexts.

To briefly illustrate the primary mechanisms linking fibrosis and cancer, we summarize the pathological transitions observed in several fibrosis-associated malignancies. Persistent hepatic inflammation and chronic activation of hepatic stellate cells during liver fibrosis disrupt liver architecture and positional identity, facilitating hepatocyte transformation and development of HCC[156-159]. Similarly, chronic epithelial injury and sustained fibroblast activation in IPF result in persistent ECM remodeling and altered epithelial cell plasticity, significantly increasing susceptibility to lung cancer[160-162]. Furthermore, sustained bone marrow fibrosis, chronic inflammation, and disrupted positional cues within the hematopoietic niche in primary myelofibrosis drive aberrant stem cell plasticity and genetic instability, ultimately fostering progression to AML[94,163,164]. Together, these examples highlight the common mechanistic framework, centered around disrupted positional identity and altered cellular plasticity, that underpins the fibrosis-to-cancer continuum across diverse tissues.

To further elucidate the specific roles of HOX and MEINOX proteins in establishing or disrupting positional identity during the progression from fibrosis to cancer, we will explore relevant common molecular examples across these distinct disease contexts. HOXB13 overexpression has been observed in liver fibrosis with correlation to increased hepatic inflammatory activity, and HOXB13 was overexpressed in a subset of HCC cases, where its overexpression was associated with cell proliferation, metastasis and chemoresistance, thereby suggesting the possibility of HOXB13 playing a role in obtaining a more plastic cellular state during liver fibrosis to HCC transformation[165,166]. Another example that has a role in fibrosis-to-cancer transformation is HOXA11-AS, which is upregulated in liver fibrosis and partakes in profibrotic ECM remodeling, has been implicated in the acceleration of cell proliferation and EMT in HCC[167,168]. However, miR-10a, which was found to exacerbate liver fibrosis through TGF-β1/Smad3 signaling pathway, can also act as a tumor suppressor miR in HCC through inhibition of cell migration, invasion and EMT, thereby pointing out a molecular signal can support fibrosis but suppress the plasticity-related cancer hallmarks at the same organ[143,144]. Curiously, in the context of lung fibrosis, HOXA5 was identified as a pro-fibrotic gene, promoting fibroblast proliferation, migration and ECM gene expression but HOXA5 has reduced expression in NSCLC with high expression being associated with increased overall survival, thereby underlining the fact that a molecular signal can increase cellular plasticity in fibrosis but suppress the cellular plasticity in cancer of the same tissue[152,169]. In the context of bone marrow, PKNOX2 was found to be downregulated in Fanconi anemia bone marrow-MSCs, a genetic disease originating from mutations in DNA repair genes, with an increased predisposition to bone marrow fibrosis, solid tumours and hematological cancers[43,170]. In another study, PKNOX2 was found to be significantly downregulated in the KRAS mutant mouse leukemia model, suggesting that PKNOX2 downregulation in the bone marrow niche may be one of the factors in the transformation of bone marrow fibrosis to AML[171]. Thus, identifying these nuanced, context-dependent roles of HOX and MEINOX factors in positional identity and cellular plasticity provides essential mechanistic insights that can guide future therapeutic strategies aimed at interrupting the fibrosis-to-cancer continuum.

THERAPEUTIC IMPLICATIONS AND FUTURE DIRECTIONS

As our understanding of HOX and MEINOX biology deepens, the prospect of targeting these pathways for therapeutic benefit becomes increasingly tangible. Therapeutic strategies targeting HOX and MEINOX factors in fibrosis emerge from recent insights, although much work remains to translate them from bench to bedside. Several studies suggest that restoring the expression or function of certain HOX genes may counteract fibrogenesis[172-174]. For instance, re-expression of HOXA5 in fibrotic kidneys restrains the JAG1-Notch pathway and mitigates collagen deposition, whereas loss of HOXA5 (via promoter hypermethylation) unleashes JAG1 and Notch signaling to drive kidney fibrosis[172]. Similarly, HOXA13 appears protective in renal fibrosis by enhancing BMP-7 (an anti-fibrotic signal) and suppressing TGF-β1-driven transcriptional programs[173]. These findings raise the possibility that epigenetic drugs (e.g., DNA demethylation agents or retinoids) could be used to reinstate anti-fibrotic HOX activity. Supporting this, 9-cis retinoic acid was shown to upregulate HOXA5 in dermal scar fibroblasts and thereby activate p53, leading to reduced proliferation and collagen production[174]. On the other hand, inhibiting HOX genes that are pro-fibrotic might be beneficial. HOXB7, for example, is markedly elevated in IPF and correlates with more severe lung scarring[175]. TGF-β can induce HOXB7 in mesenchymal cells, linking this HOX factor to the core TGF-β fibrotic pathway[108]. Targeting HOXB7 or disrupting its interaction with PBX cofactors with molecules such as HXR9 could potentially limit fibroblast activation, although such approaches have so far been tested mainly in cancer models[176]. Overall, modulating HOX/MEINOX activity in fibrosis - either by reactivating HOX genes that restrain pathological matrix remodeling or by blocking those that exacerbate it - is a tantalizing concept that warrants further validation in animal models. Basic research is still needed to confirm that these interventions can be delivered safely and effectively, given the developmental roles of HOX genes and the risk of off-target effects on normal tissue repair.

ncRNAs associated with HOX and MEINOX pathways also represent promising therapeutic targets in fibrosis. The HOX-regulated lncRNAs are one example: HOXA5 was recently found to induce the lncRNA dynamin 3 opposite strand (DNM3OS) in lung fibroblasts, which in turn recruits the epigenetic repressor EZH2 to silence tuberous sclerosis complex subunit 2 (TSC2) and drive profibrotic proliferation and myofibroblast differentiation[169]. This suggests that inhibiting DNM3OS (i.e., with antisense oligonucleotides) or interrupting the HOXA5-DNM3OS-EZH2 axis could reduce fibroblast activation in lung fibrosis. Likewise, HOTAIR, a lncRNA well-known for reprogramming chromatin in cancer, is upregulated by TGF-β in hepatic stellate cells and promotes their fibrogenic activation, and HOTAIR silencing reduces collagen production in liver fibrosis[177]. Targeting such lncRNAs that tie developmental gene networks to fibrotic gene expression offers a novel angle for antifibrotic therapy. MicroRNAs related to HOX pathways are another avenue. For example, the miR-10 family (embedded in HOX clusters) is elevated in renal and cardiac fibrosis and promotes TGF-β/Smad3 signaling; inhibition of miR-10a/b in fibrotic mouse models blunted Smad3 activation and organ fibrosis[115]. Therapeutic microRNA inhibitors (antagomiRs) against miR-10a have shown potential to attenuate fibrosis in preclinical studies, although their cell-specific effects need careful evaluation. Conversely, in contexts where a microRNA suppresses an anti-fibrotic HOX, one might envision delivering microRNA mimics to restore that brake on fibrosis. Canonical pathways like TGF-β, Wnt, and Notch that intersect with HOX and MEINOX signaling remain prime targets as well. Indeed, many HOX-driven fibrogenic effects (HOXB7 induction, HOXA5 loss, etc.) converge on TGF-β or Notch, so combining a HOX-centered intervention with a TGF-β or Notch inhibitor could yield synergistic suppression of fibrosis. In all cases, however, these strategies are still at an early stage. Further mechanistic dissection and in vivo proof-of-concept studies are required to ensure that manipulating a developmental regulator or its associated ncRNA will reverse fibrosis without impairing normal tissue homeostasis - a balance that will be critical before any clinical translation.

Aberrant expression of HOX and TALE transcription factors in cancer provides opportunities for targeted intervention, albeit ones that demand deeper mechanistic understanding. Many cancers display HOX/MEIS dysregulation, with certain HOX genes acting as oncogenes that sustain proliferation, survival, and plasticity of tumor cells[176]. This has spurred efforts to directly target HOX protein function. One promising approach is to disrupt HOX interaction with PBX cofactors, thereby crippling the transcriptional complexes that drive malignant programs. The peptide inhibitor HXR9, which competitively blocks HOX/PBX dimerization, induces apoptosis in a broad range of cancer cell types in vitro[178]. Remarkably, HXR9 and its improved derivatives (such as HTL-001) have shown selective toxicity to cancer cells over normal cells and even demonstrated anti-tumor efficacy in vivo (including improved survival in mouse models)[176,178]. These findings underscore the potential of HOX-PBX antagonism as a therapeutic strategy. In principle, such agents could be applied to HOX-addicted solid tumors (e.g., HOXA9/MEIS-high acute leukemias or HOXB7-driven melanomas) to trigger cancer cell death. Beyond peptides, future development of small molecules or stapled peptides that disrupt HOX-MEIS/PBX interfaces could broaden the practicality of this strategy. Another avenue is to modulate HOX gene expression in tumors. For example, some tumor-suppressive HOX genes are epigenetically silenced in cancer; reactivating HOXA5 (often methylation-silenced in breast and lung cancers) might restore its normal constraints on cell growth and survival, such as p53 pathway activation and induction of epithelial differentiation[174,179]. Conversely, knocking down oncogenic HOX (via small interfering RNA, short hairpin RNA or CRISPR approaches) in tumors where they are critical could impair tumor maintenance - although delivering gene therapy to all tumor cells remains challenging. Importantly, because HOX proteins are master regulators of cell identity, a precision targeting approach is needed: Complete systemic inhibition of HOX activity might be toxic, but tumor-specific delivery of HOX modulators (through nanoparticle carriers or tumor-homing peptides) could maximize the therapeutic index. This concept is exemplified by newer strategies in glioblastoma, where blood-brain barrier-penetrant peptide vectors have been used to deliver HOX/PBX inhibitors specifically to brain tumors[178]. Moving forward, more preclinical studies should investigate the consequences of HOX and/or MEINOX targeting in various cancers and identify any resistance mechanisms (e.g., redundant pathways or feedback upregulation of other developmental genes) that could arise. Such studies will inform combination therapies - for instance, pairing a HOX inhibitor with conventional chemotherapy or immune checkpoint blockade to eliminate both the bulk tumor cells and the HOX-driven CSC subpopulations.

Dysregulated ncRNAs linked to HOX and MEINOX networks in cancer are also attractive targets for therapy. LncRNA HOTAIR, which is transcribed from the HOXC locus, exemplifies this: It is dramatically upregulated in diverse carcinomas and drives EMT and metastasis by reprogramming chromatin (recruiting polycomb repressive complex 2 and silencing differentiation genes)[180,181]. Several other HOX-associated lncRNAs (e.g., HOTTIP, regulating HOXA cluster genes, or HOXA11-AS) likely have similar context-dependent oncogenic roles and could be explored as therapeutic targets if they were highlighted in earlier sections. MicroRNAs that influence HOX expression or function are another layer to consider. In aggressive breast cancers, for instance, miR-10b is induced by Twist-related protein 1 and directly targets HOXD10, a HOX gene that normally inhibits cell motility; the result is increased invasion and metastasis[182,183]. This mechanism suggests that delivering an anti-miR-10b oligonucleotide could relieve the repression of HOXD10 and suppress metastasis. Indeed, nanoparticle-mediated miR-10b silencing has shown success in reducing metastatic burden in preclinical breast cancer models[184]. Likewise, other microRNAs such as miR-196 or miR-181 families, which have been noted to modulate HOX gene clusters in leukemias and solid tumors, present possible targets - either to block (if they promote an oncogenic HOX effect) or to mimic (if they downregulate a HOX oncogene). It’s worth noting that HOX and MEIS proteins also feed into major oncogenic pathways: For example, HOX-PBX complexes can transcriptionally activate components of the Wnt and TGF-β pathways in cancer. Thus, indirect strategies could involve using existing pathway inhibitors in HOX-dysregulated cancers while monitoring HOX/MEIS signatures as biomarkers of response[108]. The key for future directions in cancer therapy is specificity: Since HOX genes are minimally expressed in most adult tissues, therapies targeting them or their RNA regulators might achieve a cancer cell-specific lethal effect. However, this must be balanced against the need for more detailed mechanistic studies, as not all tumors rely on HOX genes to the same extent, and some HOX factors have context-dependent tumor-suppressive roles. Prior to clinical translation, therefore, it will be critical to stratify patients based on HOX/MEINOX expression profiles and to ensure that any HOX-targeted therapy does not inadvertently promote tumor aggressiveness by disrupting a beneficial function of these proteins in certain contexts.

The intersection of fibrosis and cancer, where chronic fibrotic injury can lead to or support the development of malignancies, presents a particularly complex scenario for HOX/MEINOX-targeted interventions. As discussed in earlier sections, fibrogenesis and tumorigenesis share common signaling axes (TGF-β, inflammatory cytokines, ECM remodeling) that induce cellular plasticity changes, and HOX transcription factors are deeply involved in orchestrating these changes. In fibrosis-driven cancers, HOX and MEINOX proteins may act as molecular links between the scar microenvironment and nascent tumor cells. For example, in the chronically injured lung (such as in IPF), loss of HOXA5 in alveolar regions leads to excessive Notch signaling and fibrosis[169,172]; interestingly, HOXA5 is also a potent suppressor of lung cancer cell invasion and a promoter of cellular differentiation via p53 and other pathways[152,174]. This suggests that maintaining HOXA5 activity might both dampen fibrosis and reduce the risk of malignant transformation in the lung. On the other hand, we must consider molecules that have dual roles depending on context. HOXA5 itself, while generally anti-tumor, can drive pro-fibrotic gene expression in a different setting (embryonic lung fibroblasts) through induction of DNM3OS lncRNA[169]. Another example is miR-10a. In renal fibrosis, miR-10a/b are upregulated and facilitate myofibroblast activation (via Smad3 phosphorylation)[115]. In parallel, the miR-10 family is implicated in cancer progression: miR-10b is a known metastamiR in breast cancer that promotes invasion by silencing HOXD10 and miR-10a is associated with high-grade gliomas under TGF-β influence[182,185]. These context-specific roles illustrate a central challenge: A molecule like miR-10a might be an appealing target to prevent both fibrosis and metastasis, but its functions in normal physiology and any tumor-suppressive effects that it might have in certain cell types need careful delineation. Similarly, targeting HOX genes that appear pathogenic in fibrosis could have unintended consequences if those same genes play protective roles in incipient cancer (or vice versa). Designing therapies for fibrosis-driven cancers will therefore require a nuanced, context-aware approach. One strategy is precision targeting of the cell type of interest. For instance, an antifibrotic therapy could be formulated to specifically hit activated fibroblasts or stellate cells (using cell-surface markers or local delivery), aiming to reverse the fibrotic program (perhaps by reactivating a protective HOX gene or inhibiting a pro-fibrotic one) without significantly entering epithelial or parenchymal cells. This could halt the progression of fibrosis and remove the proliferative, mutagenic microenvironment that fosters cancer development. If a cancer has already emerged on a fibrotic background (e.g., HCC arising in cirrhotic liver), combined or sequential therapy might be warranted, such that with an agent to reprogram the tumor stroma by targeting HOX/MEIS pathways in carcinoma-associated fibroblasts alongside direct tumor-targeting therapy. There is evidence that the tumor stroma can be “reeducated” - excessive fibrosis in tumors can sometimes be toned down to improve immune and drug access - and HOX/MEINOX factors in stromal fibroblasts could be keys to such reeducation[186]. Indeed, lncRNA HOTAIR provides a compelling bridge between fibrosis and cancer in the context of the liver: Its inhibition not only diminishes fibrosis but also might reduce the pro-tumorigenic epigenetic alterations in neighboring hepatocytes[177]. Another important consideration is timing: Intervening during the fibrotic stage (before malignancy occurrence) vs after a tumor formation may require different tactics with HOX/MEINOX modulation. Preclinical models that mimic fibrosis-to-cancer progression (such as cirrhosis models that develop liver cancer, or lung fibrosis models that lead to cancer over time) will be invaluable to test how HOX-targeted therapies can be deployed for maximum benefit. These models can also illuminate potential risks, for example, whether inhibiting a certain HOX in fibrotic tissue might inadvertently accelerate dysplasia in epithelial cells. Multidisciplinary research, integrating cancer biology, fibrosis research, and genomics, is needed to map out the context-specific roles of each HOX/MEINOX player. This will enable us to identify “druggable” nodes that are safe to hit in one context and do not cause collateral damage in another. In summary, while the overlapping pathways in fibrosis-driven cancers hint that a unified therapeutic targeting is feasible, the contextual duality of many HOX axis molecules (e.g., HOXA5, miR-10a) means we must proceed with caution and deep understanding before these therapies reach the clinic.

Disruption of the posterior prevalence, the developmental hierarchy whereby posterior HOX genes dominantly suppress the activity of more anterior and central genes, may also be one of the unifying pathogenic mechanisms in both fibrosis and cancer. In anterior or mid-body organs such as the lung, liver, breast, and colorectum, aberrant activation of posterior HOX genes can override region-appropriate anterior or central HOX programs, disturbing positional identity and cellular homeostasis. For instance, in the liver, the upregulation of HOXC8 and HOXB13 has been linked to hepatic stellate cell activation and inflammation during fibrosis[165,187]. In HCC, widespread overexpression of posterior HOX genes, including HOXA10, HOXA13, HOXB13, and HOXD13, as well as central HOX genes such as HOXB5, HOXB7, and HOXC6, has been consistently associated with advanced disease, metastasis, and poor survival[142,166,188-190]. These findings suggest an inappropriate reactivation of posterior and central HOX programs in a mid-body organ where such expression should be tightly suppressed. Conversely, in posterior organs like the prostate, central HOX genes such as HOXC4 and HOXC6 are upregulated and promote cancer progression, while HOXB13, the native posterior gene in this region, is downregulated in castration-resistant states, leading to lipid dysregulation and metastasis[138,191]. A similar pattern emerges in breast and CRCs, where anterior HOX genes like HOXA2 and HOXA5 are frequently silenced through hypermethylation, removing their tumor-suppressive effects[192,193]. In lung fibrosis, central HOX genes such as HOXA5 and HOXB7 are upregulated and promote fibroblast activation[169,175]. Curiously, HOXA5 has been observed to be downregulated in lung cancer, where its expression correlates with improved survival[146,152]. These spatially discordant expression patterns, posterior HOX overexpression in anterior organs and anterior HOX overexpression in posterior tissues, indicate that erosion of the posterior prevalence hierarchy contributes to both fibrotic remodeling and malignant transformation by destabilizing regional transcriptional identity. Therefore, strategies ameliorating the dysregulation of the posterior prevalence might be promising in combating cancer.

CONCLUSION

Across fibrosis, cancer, and fibrosis-associated cancers, a unifying theme is the reactivation of developmental gene programs - prominently involving HOX transcription factors, their TALE partners (PBX/MEINOX), and downstream epigenetic regulators - that endow cells with aberrant plasticity. In all these conditions, cells hijack embryonic patterning pathways to attain new identities: Fibroblasts become persistently myofibroblastic, and epithelial cells assume mesenchymal, stem-like traits. HOX and MEINOX proteins emerge as master regulators of these processes, integrating signals from TGF-β, Wnt, Notch, and inflammatory pathways into specific gene expression outputs that drive either fibrosis or malignancy (or both)[108,194]. This convergence suggests that precisely tuning the HOX/MEINOX circuitry could have broad therapeutic value. The goal of such precision modulation would be to recalibrate the cell-fate decisions in diseased tissues, pushing cells back toward a healthy differentiated state and away from a pro-fibrotic or pro-neoplastic state. Unlike conventional therapies that might non-selectively inhibit TGF-β or immune responses (risking significant side effects), targeting the HOX axis offers a chance to intervene at a more specific nodal point - essentially “resetting” the pathogenic transcriptome while ideally sparing normal adult functions of these pathways. The challenge lies in the specificity of delivery and effect. Going forward, innovations like cell-type-specific delivery vectors, ligand-directed therapeutics, or inducible gene regulators could allow us to target HOX/MEINOX activity with high precision in the relevant cells. For example, a drug conjugate that accumulates in fibrotic tissue and releases a HOX-modulating payload only in those cells could treat organ fibrosis without systemic toxicity. In cancers, identifying patients with HOX-driven tumors (via molecular profiling) will enable personalized deployment of HOX/PBX inhibitors or HOX-regulating RNA therapies for maximum efficacy. As research continues to unravel the HOX and MEINOX roles across diseases, this strategy of precision HOX pathway modulation stands out as a promising frontier, potentially transforming how we treat not only fibrosis and cancer as separate entities, but also the pernicious interface between them. Ultimately, leveraging the common mechanistic threads controlled by HOX/MEINOX factors could yield therapies that flexibly dial cellular identity back toward normalcy, offering hope for conditions long deemed intractable. Such an approach exemplifies the broader concept of treating complex diseases by master regulator targeting, which will require careful refinement but holds considerable promise for future translational breakthroughs.

Although this review focuses primarily on the dynamic and disease-specific roles of HOX and MEINOX proteins in fibrosis and cancer, it is crucial to acknowledge that PBX cofactors are indispensable for forming transcriptionally active HOX-PBX-MEINOX complexes. Existing literature highlights the therapeutic potential of disrupting HOX-PBX interactions in cancer biology, yet the comprehensive therapeutic modulation of these complexes will likely require a deeper understanding of PBX’s structural and functional roles across pathological contexts[49,50,176]. Therefore, future research efforts should explicitly investigate PBX deficiency/inhibition’s results, not only in transcriptional outcomes across various disease conditions, but also the effects on the disrupted subcellular localization of MEINOX proteins. By clarifying PBX’s precise regulatory roles in conjunction with HOX and MEINOX expression profiles in disease states researchers can better identify novel therapeutic targets and develop combinational strategies that simultaneously address all essential components of the HOX-PBX-MEINOX axis, potentially enhancing therapeutic efficacy and specificity.