RUNX1 : Runt-related transcription factor 1; AML1; Acute myeloid leukemia gene 1; PEBP2αB; Polyomavirus enhancer-binding protein 2 alpha B subunit; CBFA2; Core-binding factor subunit alpha-2
RUNX2 : Runt-related transcription factor 2; AML3; Acute myeloid leukemia gene 3; PEBP2αA; Polyomavirus enhancer-binding protein 2 alpha A subunit; CBFA1; Core-binding factor subunit alpha-3
RUNX3 : Runt-related transcription factor 3; AML2; Acute myeloid leukemia gene 2; PEBP2αC; Polyomavirus enhancer-binding protein 2 alpha C subunit; CBFA3; Core-binding factor subunit alpha-3
RUNX was discovered in the early 1990s by several independent studies. The first RUNX family member to be cloned, Drosophila runt, encodes nuclear protein with no known function and has roles in segmentation, sex determination, and neurogenesis (Kania et al. 1990). Shortly after that, the human RUNX1 was sequenced as part of the t(8;21) (q22;q22) chromosomal translocation, a common recurrent abnormality in acute myelogenous leukemia (AML) (Miyoshi et al. 1991). In 1993, isolation of Runx2 cDNA of a polyomavirus enhancer-binding protein identified RUNX proteins as sequence-specific DNA-binding transcription factors (Ogawa et al. 1993a). It is notable that RUNX expression is strongly induced in differentiating embryonal carcinoma cells and Ha-ras transformed NIH3T3 (Ogawa et al. 1993a). Sequence comparison of these three genes led to the founding of a new transcription factor family, distinguished by the evolutionarily conserved Runt domain (Kagoshima et al. 1993). Subsequently, the third member, RUNX3, was identified (Levanon et al. 1994; Bae et al. 1995). These initial findings are profound indications of RUNX’s roles in stem cells, differentiation, and oncogenesis. RUNX protein was found not to bind DNA by itself but it was heterodimerized with a partner protein, CBFβ (Ogawa et al. 1993b; Wang et al. 1993). Both DNA tumor virus (polyomavirus) and RNA tumor virus (murine sarcoma virus) use RUNX proteins for their life cycle. Identification of oncogenic fusion gene, CBFB-MYH11 (SMMHC), that induces acute myeloid leukemia (Liu et al. 1993) established the role of CBFβ as a part of obligatory partner of the RUNX family.
Determination of the three-dimensional structure of the Runt domain which is a 128 aa long evolutionarily conserved domain required for DNA binding and heterodimerization with CBFβ formally determined that RUNX is a new family of transcription factor that has an S-type immunoglobulin fold. The three-dimensional structure of CBFβ was also determined. The crystal structure of the Runt domain together with CBFβ bound to DNA was also reported.
RUNX genes were initially called by different names by each researcher but they now have common names, RUNX1, RUNX2, and RUNX3 and CBFβ (see synonyms).
Most, if not all, metazoans possess RUNX genes. The discovery of two Runx genes in the holozoan Capsaspora owczarzaki indicates that RUNX originated in a unicellular organism. Primordial function of RUNX could be to control cell growth by orchestrating transcriptional programs in response to environmental cues such as various stresses. All other attributes of RUNX (e.g., lineage specification) were acquired later in evolution.
Functional Domains of RUNX Proteins
Each of RUNX1, RUNX2, and RUX3 is regulated by two promoters: distal (P1) and proximal (P2). Alternative splicing of the RUNX transcripts generate isoforms with different structure and properties.
Each RUNX protein has a highly conserved DNA-binding domain (Runt domain) and divergent C-terminus containing activation and repression domains. Major forms of each RUNX proteins have VWRPY sequence which is an inhibitory domain and interacts with TLE, mammalian homolog of Drosophila Graucho. The Runt domain is also required for interaction with core-binding factor subunit beta (CBFβ) which displaces the inhibitory domain and stabilizes the interaction of RUNX with the DNA.
Consensus DNA-binding motif is 5′-PuACCPuCa-3′ (Pu indicates purine). All three RUNX proteins bind to the same sequence. The region C-terminal to the Runt domain is a scaffold for multiple protein interactions and the basis of functional diversity. RUNX proteins do not have strong transactivation function by themselves. They cooperate with other transcription factors for strong transactivation function. A typical example is cooperation between RUNX1 and ETS1. These two proteins do not bind to DNA well by themselves due to their own inhibitory domains. When DNA-binding sites of these two proteins are closely located, RUNX1 and ETS1 cooperatively bind to DNA and transactivate target genes strongly. Following cytokine stimulation, ERK1 and ERK2 phosphorylate RUNX1, resulting in enhanced interaction of RUNX1 with p300 and increased RUNX1 transactivation activity. The region containing ERK phosphorylation sites is conserved in RUNX2 but not in RUNX3. This suggests differential regulation of RUNX family members by ERK (Ito et al. 2015).
The Interplay of RUNX Proteins
The presence of several RUNX sites in the RUNX promoters suggests cross regulation or autoregulation of RUNX proteins; RUNX family members are expressed independently. Levels of RUNX1 and RUNX3 mRNAs are inversely correlated in B cells. Levels of RUNX1 and RUNX2 mRNAs are inversely correlated in skeletal development; RUNX2 interacts with its own promoter and RUNX1 promoter. RUNX1 is inversely correlated with RUNX2 and RUNX3 expression in breast cancer.
Interaction with Signaling Pathways
All three RUNX proteins are regulated by most of the developmental regulator pathways. Direct interaction with the effectors of TGF-β pathway, the Smad proteins, is a common feature of all mammalian RUNX proteins. TGF-β receptor phosphorylates receptor-regulated Smad2 and Smad3 which interact with RUNX proteins. Then they form complex with co-Smad, Smad4, and bind to DNA to transcriptionally regulate downstream target genes (Hanai et al. 1999).
In response to TGF-β, RUNX3 directly activates the transcription of growth inhibitor CDKN1A (also known as p21 CIP ) (Chi et al. 2005) and proapoptotic BIM in the stomach. While all RUNX proteins have the ability to regulate p21 CIP1 transcription through the multiple RUNX consensus-binding sequence in the p21 CIP1 promoter, the downstream effects are different. RUNX1 regulates the p21 CIP1 promoter in a cell type-dependent manner. It transactivates the p21 CIP1 promoter in myeloid leukemia cells and represses the p21 CIP1 promoter in NIH3T3 fibroblast. RUNX2 repressed the CDKN1A promoter and attenuated TGF-β1-mediated growth inhibition and apoptosis in vascular endothelial cell. TGF-β also upregulates Indian Hedgehog (IHH) mediated by Runx2 and Smad.
RUNX proteins also interact with BMP pathway specific Smad1 and Smad5 (Hanai et al. 1999).
RUNX proteins act both upstream and downstream of the Wnt pathways. RUNX1 modulates bidirectional LEF1-Wnt signaling between epithelial and mesenchymal cells and is implicated in the regulation of regional microenvironments in hair follicle development. Wnt signaling directly induces Runx2 expression to promote osteogenesis. Binding of β-catenin-TCF7 complex to the Runx2 promoter induces Runx2 transcription in mouse multipotent mesenchymal and osteoprogenitor cells.
Runx3 interacts with the nuclear effector of Wnt signaling, TCF4 and β-catenin complex through DNA-binding domains of both Runx3 and TCF4. By this mechanism, Runx3 inhibits the DNA-binding ability of TCF4 and, hence, inhibits Wnt signaling activity. Since DNA-binding domains of both Runx and TCF families are well conserved, each member of both transcription factor families is likely to interact with each other. Consistent with this mechanism, a Wnt activity is markedly elevated in the Runx3 −/− epithelial cells, such as Cdx2, Myc, and CD44 (Ito et al. 2008).
RUNX1 and estrogen receptor-a (ERα) cooperatively activate estrogen-dependent oncogenic signaling. On the other hand, RUNX2 interacts with the DNA-binding domain of ERα to reduce the activity of ERα on the promoter of ESR1 (which encodes ERα).
RUNX3 and ERα expression levels are inversely correlated in breast cancer cells.
RUX3 mediates the ubiquitylation and degradation of ERα, leading to reduction of ERα-dependent proliferation and tumorigenic potential. Conversely, estrogen induces hypermethylation of the Runx3 promoter in mammosphere-derived cells, indicating that the oncogenicity of estrogen might partly stem from inactivation of Runx3.
YAP (Hippo Pathway)
RUNX proteins have highly conserved PPPYP sequence (PY Motif) at C-terminal region. This motif is known to interact with WW domain-containing proteins. One of such proteins, yes-associated protein (YAP), is a transcription coactivator and interacts with RUNX proteins through these protein-protein interaction motifs (Yagi et al. 1999). YAP is a nuclear effector of the Hippo pathway and, when phosphorylated, it translocates into nucleus and interacts with TEAD (TEA domain transcription factor). YAP-TEAD4 functions as an oncogene in several cancer types, such as breast and gastric cancer. The Runt domain of RUNX3 interacts with the DNA-binding domain of TEAD4, resulting in attenuation of TEAD4 DNA-binding activity and downregulation of TEAD–YAP-mediated transcription. RUNX3 negatively regulates the oncogenic TEAD–YAP complex in gastric cancer. Various YAP-TEAD target genes (e.g., collagen type XII and calpain 6) that were involved in metastasis and apoptosis were suppressed by RUNX3. Smurf is an ubiquitin ligase and another WW domain-containing protein that interacts with RUNX. RUNX proteins are degraded by interacting with Smurf. When RUNX is acetylated by p300, proteolytic degradation is blocked.
Epithelial-Mesenchymal Transition (EMT)
Epithelial cells are in close contact with each other and organized in sheets with apico-basal polarity. They are linked tightly through cell-cell junctional complexes: adherent junctions, desmosomes, and tight junctions. Mesenchymal cells are loosely organized. Their main purpose is to act as connective tissues that provide structural support to the epithelia. Mesenchymal cells can migrate especially during development. During normal development, epithelial cells in some tissue lose junctional complex to acquire mesenchymal cell property. This process is called epithelial mesenchymal transition (EMT). Following EMT, cells often undergo reverse process, mesenchymal epithelial transition (MET). RUNX2 is involved in the formation of cardiac valves in the embryonic heart. RUNX3 is transcriptionally regulated by the Notch pathway and its nuclear effectors, CBF-1/Suppressor of Hairless/Lag1 (CSL) and Mastermind-like (MAML-1). Runx3 sustains the long-term expression of Snail2 to maintain EMT-transformed endothelial cells in mesenchymal state. EMT is one of the important factors for solid tumor progression and invasion. The loss of Runx3 in gastric epithelial cells led to the induction of EMT, resulting in a subpopulation of cells which acquired tumorigenic, stem cell-like properties. RUNX3 therefore protects gastric epithelial cells from aberrant TGFβ signaling and subsequent reprogramming by EMT.
RUNX1 in Normal Hematopoiesis
Runx1-positive cells are present in hemogenic endothelial cells and hematopoietic cell clusters in the dorsal aorta of the Aorta-Gonado-Mesonephros (AGM) region at E9.5–E11.5 in mouse embryos. Hematopoietic stem cells are developed by budding from the vessel wall by the process called endothelial-to-hematopoietic cell transition (EHT). Runx1 and Cbfb are indispensable for this process (Okuda et al. 1996).
RUNX1 and Leukemia
RUNX1 is frequently involved in human leukemias. Prevalent type of RUNX1 genetic alterations is chromosomal translocations. RUNX1 fusion proteins identified are: RUNX1-ETO produced by t(8;21), associated with AML (Miyoshi et al. 1991), and TEL-RUNX1 produced by t(12;21) observed in childhood acute lymphoblastic leukemia (ALL). Most RUNX1 fusion proteins retain their N-terminal and Runt domain but lack the C-terminal regulatory region. Therefore, the fusion proteins can bind to DNA but are defective in transcriptional regulation. They compete with wild-type RUNX1 for DNA binding. RUNX1 fusion proteins are widely believed to be dominant negative mutants and this is the common underlying mechanism for leukemogenesis. The germline deletion of RUNX1 is associated with familial platelet disorder with predisposition to AML (FPD/AML) (Song et al. 1999).
The RUNX1 heterodimerization partner, CBF-β, is also mutated in human leukemias by chromosome 16 inversion, inv(16)(p13;q22), observed in AML-M4Eo subtype. This inversion generates a fusion gene, CBFB-MYH11, encoding CBF-β-smooth muscle myosin heavy chain (SMMHC) fusion protein, which acts as a dominant repressor for RUNX1 function. RUNX1-ETO knock-in mouse does not induce leukemia. Additional genetic changes appear to be required. RUNX1-ETO is generally considered to be a dominant-negative form of all RUNX proteins by interfering with the DNA binding (Speck and Gilliland 2002). However, it is also reported that normal allele of RUNX1 is required for the survival of leukemic cells caused by RUNX1-ETO fusion gene. CBFB-MYH11 knock-in mouse together with oncogenic N-ras spontaneously induces leukemia.
Roles of RUNX1 in Tissue Stem Cells
Runx1 has an essential role in hematopoietic stem cells. A 270 bp Runx1 enhancer element located between two promoters of Runx1 gene was identified that drives Runx1 expression in hematopoietic stem cells. This element, termed eR1, was subsequently found to drive Runx1 expression in the rapidly proliferating cells of stomach located in the isthmus (narrowing) of both corpus (main body) and antrum (close to duodenum) which have been predicted to be stem/progenitor cells of stomach.
RUNX2 in Osteogenesis
Runx2 is required for osteoblast differentiation (Komori et al. 1997). Runx2 transcriptionally activates Sp7 which is a crucial transcription factor for osteoblast differentiation and bone matrix genes including Spp1, Ibsp, and Bglap2. Runx2 is also required for chondrocyte maturation, and Runx3 has a redundant function with Runx2 in chondrocyte maturation. Runx2 regulates the expression of Col10a1, Spp1, Ibsp, and Mmp13 in chondrocytes. It regulates chondrocyte proliferation through the regulation of Ihh expression. Runx2 enhances osteoclastogenesis by regulating Rankl. The expression of Runx2 in osteoblasts is regulated by a 343-bp enhancer located upstream of the P1 promoter. Thus, Runx2 is a multifunctional transcription factor that is essential for skeletal development, and Cbfβ regulates skeletal development by modulating the stability and transcriptional activity of Runx family proteins.
Germline mutation of RUNX2 causes Cleido-Cranial Dysplasia (Lee et al. 1997).
RUNX in Lymphocytes
The Runx family is one of the ancestral regulators of hematopoiesis and involved in shaping the immune system. They are involved in innate and acquired lymphoid cells in vertebrates. There are two types of T cell antigen receptor, αβTCR and γδTCR. RUNX1 is involved in the enhancer activity of all TCR. Skin-specific γδT cell, called dendritic epidermal T cells (DETC), is regulated by RUNX3.
In the development of αβT cells, CD4+ and CD8+ single positive cells are derived from CD4+CD8+ double positive thymocytes. When CD8+ single positive cytotoxic lineage T cells are developed, a transcriptional silencer is responsible to repress CD4. RUNX3 is primarily responsible for this silencer activity.
For the development of helper lineage specific CD4+ cells, major transcription factor is ThPOK whose expression is regulated by RUNX1. ThPOK attenuates RUNX3 activity in CD4+ cells. In CD8+ cells, ThPOK is also silenced by ThPOK silencer by RUNX3. Cross antagonism between ThPOK and RUNX3 serves as a central mechanism in CD4 helper and CD8 cytotoxic T cell development.
Another αβT cell subset generated from CD4+/CD8+ double positive cells are regulatory T cells (Treg). Treg plays an essential role in immune tolerance. FOXP3, a member of forkhead-box transcription factor, is essential in generating Treg. RUNX1 is required for the maintenance of FOXP3 level.
Another αβT cells’ subset is CD8αα IEL (intraepithelial lymphocyte). RUNX3 is required for their generation. These T cells are unique in that they are not CD8αβ heterodimer and localized in the space between gut epithelial cells.
For the development and function of Innate Lymphoid Cells (ILC), RUNX proteins have important roles. All ILC are developed from common lymphoid precursor cells (CLP). Runx1 and Runx3 are expressed at extremely high levels in PLZF+ innate lymphoid cell precursor (ILCP). Of three ILC subtypes, Runx3 is indispensable to all type I ILC population (Collins et al. 2009).
RUNX in Dorsal Root Ganglion
Dorsal root ganglion (DRG) is one of the tissues in which Runx1 and Runx3 are expressed at the highest levels among the entire body. Runx1 is expressed in both the central and peripheral nervous system of mouse embryo. In central nervous system, Runx1 is expressed in selective populations of somatic motor neurons in the spinal code and in cholinergic branchial and visceral motor neurons in the hindbrain such as dorsal vegal nucleus and nucleus ambiguous. In the peripheral nervous system, Runx1 is localized to DRG and selective cranial ganglia, including trigeminal (V) and vestibulocochlear (VIII) ganglia and the glossopharyngeal-vegal (IX–X) ganglia complex. On the other hand, Runx3 is confined to the peripheral nervous system, specifically to DRG and cranial ganglia.
There are three major subpopulations of DRG neurons: nociceptive, mechanoreceptive, and proprioceptive. Runx1 and Tunx3 are synthesized initially in TrkA+ nociceptice and TrkC+ proprioceptice neurons, respectively. Later, TrkB/TrkC hybrid neurons are formed. When TrkB and TrkC expressing cells are segregated, Runx3 represses TrkB and DRG neurons acquire TrkC+ identity.
It has been suggested that Runx1 controls nearly all known marker genes critical for nociceptive functions. This is similar to the global control function of Runx1 in hematopoietic stem cell formation and Runx2’s role of osteoblast maturation.
Dysregulation of RUNX in Solid Tumors
RUNX1 is one of the significantly mutated genes in luminal breast cancer. Missense mutations at the Runt domain of RUNX1 and its binding partner, CBFB, are clear indications that the DNA-binding ability/transcriptional activity of RUNX1 influence breast cancer growth. RUNX1 is highly expressed in a broad range of epithelial tumors, such as those of the skin, esophagus, lung, colon, and the breast. RUNX1 is critical regulator of stem cell homeostasis at the skin follicles. RUNX1-expressing hair follicle stem cells are detected at the origin of skin tumors. This suggests that RUNX1 is essential for tumor initiation and maintenance in the skin.
RUNX2 is not significantly mutated in cancer. Overexpression of RUNX2 is frequently observed in bone, breast, and prostate cancers, suggesting that enhanced RUNX2 activity contributes to oncogenic growth in such tissues. For example, human tissue microarray revealed that RUNX2 expression is elevated in triple negative (i.e., oestrogen receptor (ER)/progesterone receptor (PR)/HER2 negative) breast cancers and associated with a poor survival rate.
A comprehensive molecular analysis of gastric adenocarcinoma revealed key dysregulated pathways and putative drivers of various subtypes of gastric cancer. The significantly mutated genes included those in the KRAS, β-catenin, TGF-β signaling, p53, Fanconi anemia, and mitotic pathways. RUNX3 is significantly involved in most of these pathways. The mutation of RUNX3, R122C, was identified in gastric cancer.
The increased proliferation and suppressed apoptosis of stomach epithelial cells of Runx3 knockout mice are attributed to defective TGF-β-mediated apoptosis.
RUNX and p53
Induction of cellular senescence by the ARF-p53 pathway constitutes one of the most robust defense mechanisms against oncogenic Ras. All three RUNX genes depend on the ARF-p53 pathway to induce senescence in mouse embryo fibroblast (MEF). Moreover, oncogenic H-RASG12V failed to induce senescence in Runx2−/− MEF. These results indicate that the ability to counter oncogenic signal through oncogene induced senescence is conserved in all three RUNX members and constitute an important antitumor property of the RUNX family. The discovery of RUNX-binding sites in the ARF promoter and the ability of RUNX1 to induce ARF transcription strongly indicate that RUNX-function upstream of the ARF-p53 pathway as a tumor suppressor. The fact that oncogenic RUNX1-ETO is associated with decreased levels of ARF mRNA in AML cells further reinforces RUNX as an upstream regulator of p53 (Linggi et al. 2002). It has become increasingly apparent that p53 is not activated by low levels of oncogenic stimuli. In K-rasG12D-driven mouse models of lung cancer, restoration of p53 expression in established tumor did not result in complete tumor regression. Malignant adenocarcinoma, but not adenoma, were specifically regressed by p53 restoration, despite the fact that adenoma will progress to adenocarcinoma (Feldser et al. 2010; Junttila et al. 2010). Then Runx3 was found to prevent adenoma formation in mouse model of lung cancer. Inactivation of Runx3 alone induces adenoma and combination of activation of oncogenic K-ras and inactivation of Runx3 rapidly induces adenocarcinoma. The results suggest that Runx3 is a gatekeeper of lung adenoma formation (Lee et al. 2013).
Conversely, p53 could act upstream of RUNX. During the differentiation of keratinocyte in human interfollicular epidermis, cooperation between p53 and p63 is required for the precise regulation of RUNX1 expression.
RUNX Proteins as Oncogene
Retrovirus-mediated insertional mutagenesis in mouse revealed that overexpression of RUNX genes during the development of Myc-driven T cell lymphoma. It was suggested that RUNX genes cooperate with Myc during oncogenesis, overriding p53-Arf-mediated tumor suppression in thymocytes (Blyth et al. 2005). Moreover, RUNX1 – through its stimulation of STAT3 signaling – was implicated in the development of skin squamous cell carcinoma (SCC), oral SCC, and ovarian cancer. In pancreatic cancer, Runx3 functions as a driver of metastasis depending on the dose of Smad4. RUNX protein can be a tumor suppressor or an oncogene.
EBV and Immortalization of B Cell
In B lymphocytes, RUNX1 and RUNX3 are expressed throughout B-cell development. Epstein-Barr virus (EBV) is capable of immortalizing B cells and RUNX1 and RUNX3 are involved in this process. In EBV-infected cells RUNX1 levels are low and RUNX3 levels are high. RUNX1 in these cells is repressed by RUNX3 to relieve RUNX1-mediated growth repression. RUNX3 is upregulated by the EBV-encoded EBNA2 protein and represses RUNX1 transcription through RUNX sites in the RUNX1 P1 promoter. EBNA2 activates RUNX3 transcription through an 18 kb upstream super-enhancer in collaboration of EBNA2 and Notch DNA-binding partner RBP-J. This super-enhancer also directs RUNX3 activation by two further RBP-J-associated EBV TFs, EBNA3B and 3C.
HIV and CBFβ
In order to escape from host immune system, viruses have a cunning strategy. The HIV-1 protein Vif downregulates the human RNA editing factor, APOBEC3 family, by targeting them for proteolysis by the ubiquitin-proteasome pathway. A partner of RUNX transcription factor, CBF-β, acts as a template to assist in Vif folding and allow for assembly of an APOBEC3-targeting E3 ligase complex. In uninfected cells, CBF-β is an essential binding partner of RUNX transcription factors. By hijacking CBF-β, Vif reduces the function of RUNX1 which is essential in T cell function. Expression of APOBEC3 family members are also perturbed by reduction of RUNX1 function.
Induction of RUNX by Oncogene
Runx3 expression is highly dynamic and changes according to environmental cues. In response to oxidative and osmotic stress, the Caenorhabditis elegans RUNX homolog RNT-1 protein is rapidly stabilized in the intestine. In humans, chemotherapeutic and DNA-damaging agent doxorubicin induced RUNX3 expression in cultured cell lines. Oncogenic stress, such as expression of mutant K-Ras, induced Runx3 expression in human embryonic kidney HEK293 cells reinforced the notion that stress response is a fundamental, as well as evolutionarily conserved, function of RUNX3. RUNX3 is transcriptionally activated by oncogenic K-Ras to mediate the expression of p53. RUNX3 may serve as a monitor of K-Ras activity, physiological vs. oncogenic levels. It might be that all RUNX genes are activated by stress or other oncogenic signals to protect normal cells from becoming cancerous.
The RUNX family consists of evolutionarily old genes that function as transcription factors in developmental regulation and are frequently involved in cancer. There are three mammalian genes, RUNX1, RUNX2, and RUNX3. RUNX1 is essential for hematopoiesis and frequently involved in chromosome translocation in leukemia. RUNX2 is essential for skeletal development. RUNX3’s major roles are T cell development as well as neuronal development. RUNX3 is associated with various solid tumors as a tumor suppressor and in some cases as an oncogene. CBFβ is an obligatory subunit for all three RUNX proteins for strong DNA binding. CBFβ is also involved in leukemia as a fusion gene. They are regulated by most of the developmental signals.