If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Division of Experimental Hematology and Cancer Biology, Cincinnati Children's Hospital Medical Center, Cincinnati, OHDepartment of Cancer Biology, University of Cincinnati, Cincinnati, OHDepartment of Pediatrics, Cincinnati Children's Hospital Medical Center, Cincinnati, OH
FHA domain-containing proteins are critical gatekeepers of diverse signaling pathways.
TIFA and TIFAB are critical FHA domain-containing proteins in immune-sensing pathways.
TIFA and TIFAB have diverse roles in human pathological disorders and diseases.
TIFA and TIFAB have shared and distinct characteristics.
Forkhead-associated (FHA) domain–containing proteins are widely expressed across eubacteria and in eukaryotes. FHA domains contain phosphopeptide recognition motifs, which operate in a variety of phosphorylation-dependent and -independent biological processes, including the DNA damage response, signal transduction, and regulation of the cell cycle. More recently, two FHA domain–containing proteins were discovered in mammalian cells as tumor necrosis factor receptor-associated factor (TRAF)–interacting proteins: TIFA and TIFAB. TIFA and TIFAB are important modifiers of the innate immune signaling through their regulation of TRAF proteins. Recent studies have also revealed distinct roles for TIFA and TIFAB in the context of immune cell function, chronic inflammation, hematopoiesis, and hematologic disorders. Collectively, these studies indicate the important role of TIFA- and TIFAB-dependent signaling in hematopoietic cells and their dysregulation in several human diseases. In this review, we summarize the molecular mechanisms and biological role of these FHA-domain homologues, placing them into the context of human disease.
Overview of forkhead-associated domains
Forkhead-associated (FHA) domains are versatile functional domains that mediate protein–protein interactions [
], these domains recognize primarily phosphorylated (p) threonine (Thr) motifs on a variety of target proteins. FHA domains are widely expressed in many genomes; are found in prokaryotes, eubacteria, and eukaryotes; and appear in a diverse range of proteins [
]. They are often found in tandem with other functional domains such as kinase domains or RING-fingers, aiding in substrate recognition. FHA domain–containing proteins are frequently observed in phosphorylation-dependent signaling events, including the DNA damage response, signal transduction, cell cycle progression, and cell growth [
]. These proteins serve as the interpreters responding to the activity of kinases and phosphatases, taking a decisive role in the integration of phosphorylation-dependent cellular signals. Although the molecular mechanisms of FHA domains in prokaryotes and in lower eukaryotes have been extensively studied and comprehensively reviewed [
]. Despite their diverse range of functions, FHA domains are not necessarily promiscuous in their preferences for phosphomotif recognition. On target proteins, residues proximal to the p-Thr, particularly in the +3 position, seem to dictate interactions for a given FHA domain–containing protein [
]. Thus, FHA proteins are afforded binding specificity despite the ubiquity of phosphorylation events. Moreover, some exposed FHA domain apices map to functional domains of other proteins, away from the phosphopeptide recognition site, providing evidence of phosphorylation-independent roles for FHA domain–containing proteins [
]. Many FHA domain proteins can exist as monomers but exhibit induced dimerization and subsequent oligomerization in response to phosphorylation. Because FHA domains bind and therefore shield phosphorylation sites, it has also been surmised that the presence of FHA domain–containing proteins dictate whether and how phosphorylation events are recognized or transduced. This binding function can also prevent access by phosphatases [
]. Therefore, FHA domain–containing proteins can serve as potent gatekeepers of signal transduction in response to stimuli.
TRAF-interacting proteins with FHA domains
Innate immunity is the first line of defense against microbial and viral infection. The pattern recognition receptors (PRRs) on innate immune cells recognize pathogen-associated molecular patterns (PAMPs) derived from various pathogens, thus resulting in a robust and rapid induction of an innate immune response [
]. During infection, conserved and recognizable PAMPs allow innate immune cells to mount efficient protective responses to ensure that infection does not progress. Discovery of a range of PRRs, including the family of Toll-like receptors (TLRs), was a major breakthrough in the field of immunity. Thus, identification of new PAMPs, PRRs, and their mechanisms of signal transduction are of priority for understanding the complexity of host–pathogen interactions [
]. Two FHA domain–containing proteins—tumor necrosis factor receptor-associated factor (TRAF)–interacting protein with an FHA domain (TIFA) and its closely related homologue, TRAF-interacting protein with an FHA domain containing protein B (TIFAB)—have recently been identified as signal transducers in innate immune mechanisms (Table 1) [
]. In particular, TIFA and TIFAB were discovered to regulate key members of the TRAF family of proteins, which are ubiquitin ligases that associate with surface receptors and synthesize ubiquitin chains to transduce extracellular inflammatory stimuli. Namely, TRAF2 and TRAF6 are established interactors with TIFA and are mediators of the NF-κB pathway [
]. After their initial discoveries, a number of studies have unveiled additional roles for TIFA and TIFAB in the innate immune system, but also in epithelial, endothelial, and hematopoietic cells. Of note, TIFAB has been implicated in a subset of myeloid malignancies, including myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML) [
]. Compared with TIFA, very little is known about TIFAB, yet disrupting the expression of either of these FHA proteins yields potent cellular consequences. The discovery of TIFA and TIFAB has laid the groundwork for understanding intermolecular and intramolecular fine-tuning of TRAF proteins in inflammation-related signaling. Given the increasingly appreciated correlation between chronic inflammation and infection and other morbidities such as cancer and cardiovascular diseases, TIFA and TIFAB are gaining traction as suspects underlying their pathogenesis (Table 2). In this review, we summarize current literature to delineate the shared versus distinct mechanisms of the FHA domain–containing proteins TIFA and TIFAB, while placing them in the context of immune signaling, inflammation, hematopoietic disorders, and other diseases.
Table 1Comparison of TIFA and TIFAB
Chromosome 3 in mice and 4q25 in humans
Chromosome 13 in mice and 5q31 in humans
184 amino acids
161 amino acids
Phosphorylation/FHA recognition site
Lacks a homologous threonine residue at position 9
TRAF binding site
Binding to TRAF2 requires 1–162 amino acids to activate TNFR2/NF-κB signaling Binding to TRAF6 via E178 activates TLR/NF-κB signaling
Does not bind to TRAF2 Binding to TRAF6 via C-terminus suppresses TLR/NF-kB signaling
Enriched in HSCs, myeloid progenitors, and B cells
Enriched in myeloid progenitors and monocytes
TIFA is a positive regulator of innate immune and inflammatory pathways TIFA oligomerization is required for NF-κB activation No documented evidence of binding to USP15 ALPK1 induces p-Thr9-TIFA to activate NF-κB
TIFAB is a negative regulator of innate immune and inflammatory pathways May not require oligomerization for its biological function Binds to C-terminal domain of active USP15 No documented evidence of binding to ALPK1
] identified TIFA as a TRAF2 binding protein. Located at chromosome 3 in mice and 4q25 in humans, the TIFA gene encodes a small protein containing a conserved FHA domain and a p-Thr recognition site (Figure 1). Though relatively small and lacking any intrinsic activity, TIFA prompted interest because of its ability to bind not only TRAF2, but also TRAF6, and potently activate innate immune response through NF-κB and activator protein 1 (AP-1) transcription factors [
]. TIFAB shares DNA sequence homology and harbors a conserved FHA domain but was quickly revealed to exert opposing cellular functions as compared with TIFA. Unlike what is observed in many other FHA domain–containing proteins, the FHA domains of TIFA or TIFAB do not occur in tandem with any other characterized functional protein domains. However, TIFA and TIFAB exhibit high sequence conservation in vertebrates, indicating evolutionarily important cellular functions.
TIFA and TIFAB: structure to function
Human TIFA is a protein of 184 amino acids, containing an FHA domain flanked by short N- and C-terminal sequences (Figure 1). TIFA binds to both TRAF2 and TRAF6, which play important roles in inflammatory signaling [
Identification of TIFA as an adapter protein that links tumor necrosis factor receptor-associated factor 6 (TRAF6) to interleukin-1 (IL-1) receptor-associated kinase-1 (IRAK-1) in IL-1 receptor signaling.
]. Transient-transfection experiments performed with deletion and substitution mutants of TIFA indicate that TIFA harbors a TRAF6-binding motif in its C-terminus and, surprisingly, that the FHA domain is entirely dispensable for the TIFA–TRAF6 interaction [
]. Yet, the FHA domain in TIFA is still required for subsequent NF-κB activation. Binding of TIFA to TRAF6 alone is insufficient to activate this pathway. Combined with observations that TIFA is detected at multiple molecular weight fractions in size-exclusion liquid chromatography, this evidence points to higher-order TIFA structures that are critical to activation of the NF-κB pathway [
]. Biochemical studies revealed primarily that TIFA exists as a homodimer and, secondarily, that phosphorylation of Thr-9 stabilizes the FHA domain interactions of different TIFA homodimers and not within the FHA domains of monomers [
]. Thus, pThr-9 on TIFA initiates a rapid cascade of phosphorylation, oligomerization, and subsequent auto-ubiquitination of TRAF6 in TIFAsomes (Figure 2B). These findings support a model in which “head-to-tail binding” between different TIFA dimers promotes oligomerization while leaving the C-terminal TRAF6 binding site exposed, thus enabling the TRAF6 trimers to bind to phosphorylated TIFA hexamers, creating a large signaling scaffold that potently drives downstream signaling (Figure 2B). Detailed structural studies support these findings, indicating that TRAF6 trimers are bound to the axis of a phosphorylated TIFA hexamer. This finding postulates that TIFA oligomerization would be important for the alignment of multiple full-length TRAF6 molecules to form a large signalosome (Figure 2B) [
] provided evidence for direct binding between TIFA and the TRAF domain of TRAF6. By mass spectrometry, a total of 77 TIFAsome-associated proteins were identified. These include a number of core NF-κB pathway components, including TRAF2, TAB2, and TRIM21 [
]. Thus, it is possible that the oligomerization events originating with the FHA domain of TIFA create a graded threshold for robust innate immune activation, through either TRAF6 or TRAF2, thus permitting rapid signal amplification only under sustained pathway stimulation.
TIFAB is 161 amino acids in length and also contains the FHA domain flanked by short N- and C-terminal sequences. TIFAB garnered interest because of its location on the commonly deleted region of chromosome 5q (del(5q)) in AML and MDS [
] found that, like TIFA, the key amino acid residues mediating effects on TRAF6 lie in the C-terminus of TIFAB. However, in stark contrast to TIFA, disrupting this C-terminal domain of TIFAB de-repressed TRAF6 and instead prevented TIFAB-mediated inhibition of NF-κB signaling (Figure 2A). Further mechanistic studies solidified these observations, revealing that TIFAB can bind TRAF6, inhibit TRAF6 protein levels, and ultimately impinge on downstream signaling events, including the activation of NF-κB signaling. Binding of TIFAB to TRAF6 occurs in the cytoplasm and across multiple model systems, including normal and leukemic hematopoietic cells. However, unlike TIFA, TIFAB does not appear to bind or regulate TRAF2, as it affects neither TRAF2 protein levels nor TRAF2-induced NF-κB activation [
]. Despite their sequence similarities, TIFA mediates TRAF6-induced NF-κB signaling, while in contrast, TIFAB suppresses TRAF6-induced NF-κB signaling.
To gain additional insight into the functional role of TIFAB in del(5q) MDS/AML, a proteomics screen was performed to identify TIFAB-binding proteins. Mass spectrometry analysis confirmed that TIFAB binds endogenous TRAF6 in a human del(5q) AML cell line, but these studies also revealed that TIFAB readily binds the ubiquitin-specific peptidase USP15 [
]. A series of biochemical assays and co-immunoprecipitation experiments confirmed that TIFAB binds USP15, and this appears to be partly dependent on the deubiquitinating (DUB) activity of USP15. USP15 harbors histidine and cysteine residues in its C-terminal hydrolase domain [
]. TIFAB exhibited a preference for binding active USP15 enzyme rather than its inactive cysteine mutant. Moreover, TIFAB likely binds the C-terminal hydrolase domain of USP15, thus placing the binding of TIFAB near the DUB active site. Cell-free DUB assays indicated that TIFAB further promotes the DUB activity of USP15. Collectively, these studies revealed that FHA domain–containing proteins, which typically mediate protein–protein interactions, are capable of fine-tuning the activity of a deubiquitinating enzyme. Although several reports have identified phosphorylated threonine residues on USP15 [
], future studies will be required to dissect the degree of involvement of the TIFAB FHA domain in mediating these interactions.
The apparent multistep signal amplification of TIFA presents several stages at which the innate immune signaling may be fine-tuned by negative feedback, which is a process critical for preventing sustained activation and hyper-inflammation under stress. As we now understand, TIFAB is a negative regulator of innate immune and inflammatory pathways. Despite the fact that the N-terminal, FHA and C-terminal domains of TIFA and TIFAB share DNA sequence similarity, there is notable variability at the amino acid level, and TIFAB exerts opposite effects on NF-κB signaling as compared with TIFA. Adding to the ambiguity of TIFA and TIFAB functions is the fact that they lack intrinsic catalytic functions; they merely dictate protein–protein interactions and, in the case of TIFA, the assembly of signalosomes. TIFA and TIFAB are also quite small, limiting the likelihood that they would act monomerically as molecular scaffolds, which seems to explain the requirement of TIFA oligomers for NF-kB activation. TIFAB lacks a homologous p-Thr residue at position 9 (Figure 1), indicating that it may not require oligomerization for its biological function, further separating it from the function of TIFA. Future biochemical and structural studies will help clarify this distinction. Although TIFA and TIFAB had both been studied in vitro and in a subset of immune and hematopoietic cells, it is unknown whether these FHA domain–containing proteins played a functional role in hematopoiesis or other biological systems.
Biology of TIFA
Although TIFA is a ubiquitously expressed cytoplasmic protein within the hematopoietic system, TIFA expression patterns indicate a lineage bias toward lymphoid cell fates. TIFA expression is slightly enriched in hematopoietic stem cells (HSCs) and myeloid progenitors, with robust expression in B cells and reduced expression in monocytes [
] (Figure 3). Yet, thus far, the functional effects of TIFA expression on hematopoiesis have not been reported. However, several studies on the cellular effects mediated by TIFA reveal a diverse range of biological functions, even in cell types that are not hematopoietic in origin. On the basis of several reports, it is known that TIFA is functionally important in cell types that exhibit epithelial barrier function, including gastric and intestinal epithelial cells [
]. Most recently, TIFA has been reported to sense bacterial intermediate by-products of lipopolysaccharide (LPS) via a cytosolic surveillance pathway triggering the NF-κB response, thus playing an important role in innate immune cells [
]. TIFA-mediated innate immune signaling is critical for transducing intracellular signals to mount an inflammatory immune response that recruits immune cells to sites of bacterial infection or tissue damage.
Three independent yet complementary studies reported that the human pathogen Helicobacter pylori promotes NF-κB and interleukin (IL)-8 responses in gastric epithelial cells via the activation of a novel innate immune signaling pathway, dependent on TIFA, TRAF6, and α-kinase 1 (ALPK1) [
]. In response to the bacteria H. pylori, TIFA mediates NF-κB responses to the monosaccharide heptose 1,7-bisphosphate, a metabolic intermediate in the LPS biosynthesis pathway of gram-negative bacteria. Subsequently, it was reported that the induction of TIFA oligomerization and IL-8 responses by LPS-containing gram-negative bacteria is dependent on the function of ALPK1, a neighboring member of TIFA on human chromosome 4 [
]. By using combinatorial genome-wide screens within the bacteria and their host cells, Zhou et al. identified ALPK1 as a cytosolic innate immune receptor for the LPS metabolite, ADP-β-D-manno-heptose (ADP-Hep) [
]. ADP-Hep triggered association of ALPK1, TIFA, and TRAF6, indicating that complex formation is required for sensing ADP-Hep and activating NF-κB. However, whether ALPK1 directly phosphorylates TIFA at Thr-9 remains to be resolved.
There have also been reports of TIFA-dependent NF-κB activation in endothelial cells under oxidative stress. This oxidative stress-induced activation, in turn, upregulates components of an NLRP3-containing inflammasome, leading to a feed-forward mechanism whereby TIFA interacts with caspase-1 and amplifies the release of pro-inflammatory cytokines like IL-1β [
]. In endothelial cells, TIFA is phosphorylated by AKT, which also raises the possibility that TIFA may have additional roles in cellular metabolism and protein synthesis in response to AKT signaling. Expression of TIFA in endothelial cells was further validated by data from single-cell RNA sequencing, where TIFA expression was strictly limited to LepR+ stroma and VE-Cadherin+ endothelial cells in the bone marrow (BM) [
]. In liver cells, TIFA was reported to be one of the most significantly upregulated genes in a model of hemorrhagic shock. After hypoxia–reoxygenation, TLR4/MyD88 signaling induces TIFA interaction with TRAF6 and IRAK1, leading to rapid release of extracellular factors, such as tumor necrosis factor (TNF)-α. TIFA-mediated release of TNF-α is a predictor of survival during hemorrhage [
], who found that the matrix protein of a paramyxovirus, Newcastle disease virus (NDV), promotes NDV replication by reducing TIFA expression, and thus, downregulating TIFA/TRAF6/NF-κB–mediated production of cytokines [
]. Mechanistically, Aurora A was found to be essential for Thr-9 phosphorylation of TIFA and NF-κB activation. Conversely, TIFA inhibition led to reduced leukemic growth and chemoresistance via anti-apoptotic signaling through BCL2. During genotoxic stress leading to DNA damage, TIFA exhibited nuclear translocation and accumulated on damaged chromatin [
]. The authors found that DNA damage induces TIFA phosphorylation at Thr-9 and that this phosphorylation event is crucial for its enrichment on damaged chromatin and subsequent NF-κB activation. Moreover, TIFA, along with TRAF2, relayed the DNA damage signals by stimulating ubiquitination of the NF-κB essential modulator (NEMO), whose sumoylation, phosphorylation, and ubiquitination were critical for NF-κB's response to DNA damage.
In multiple myeloma cells, TIFA-mediated NF-κB activation during DNA damage inhibited cell proliferation, suggesting a novel tumor suppressive function of TIFA. Similar tumor suppressive functions of TIFA were also noted in human hepatocellular carcinoma wherein TIFA expression was found to be significantly lower in primary liver biopsies [
]. The mechanistic basis of TIFA-mediated suppression of hepatocellular carcinoma was supported by the observation that ectopic TIFA overexpression leads to caspase-dependent apoptosis via the TRAF6-binding C-terminal residues (178–184 amino acids). Although the molecular basis of NF-κB activation by TIFA and TRAF6 has been determined, the question of whether TIFA-dependent NF-κB signaling is activated mainly by IL-1β, TNF-α, and/or other unidentified stimulation still remains to be elucidated.
On the contrary, in lung adenocarcinoma, TIFA expression was found to be significantly higher and positively correlated with reduced overall survival of lung cancer patients [
]. Pro-inflammatory cytokines such as IL-1β and TNF-α were also increased, and inhibition of TIFA/NF-κB led to reduced infiltration of inflammatory cells and amelioration of cardiac toxicity. These studies suggest a critical role for TIFA in cardiac remodeling during myocardial infarction.
Biology of TIFAB
TIFAB is implicated in myeloid malignancies as it resides within the commonly deleted region in del(5q) MDS/AML. Interestingly, genetic alterations involving TIFA have not been associated with any hematologic malignancies. We previously reported that TIFAB is highly expressed in the more immature fraction of normal human BM CD34+ stem/progenitor cells as compared with the mature cell fraction of CD34– BM cells, while TIFAB expression is significantly decreased in BM CD34+ cells isolated from del(5q) MDS patients [
]. The reported function of TIFAB in innate immune signaling and its genetic deletion in del(5q) MDS warranted further studies. Guided by previous reports implicating dysregulation of innate immune signaling via TRAF6 in MDS [
] conducted functional studies to evaluate the consequences of TIFAB deletion on TLR/TRAF6/NF-κB signaling in hematopoietic cells and on its role in the pathogenesis of MDS. These studies ultimately revealed divergent functional roles of TIFAB from TIFA on TRAF6-dependent innate immune activation, with detrimental consequences for normal hematopoiesis. Consistent with TIFAB being a negative regulator of innate immune signaling, deletion of TIFAB in murine BM hematopoietic stem and progenitor cells (HSPCs) resulted in enrichment of genes implicated in interferon response, LPS-induced inflammation, and NF-κB activation. Moreover, BM transplantation models revealed that manipulating TIFAB and, therefore, the innate immune response resulted in clear HSPC defects in vivo. TIFAB-deficient HSPCs are outcompeted by wild-type HSPCs in competitive BM transplantations, indicating that these effects are largely cell intrinsic. Moreover, lethally irradiated recipient mice engrafted with TIFAB-deficient BM cells developed progressive peripheral blood cytopenias, myeloid dysplasia, and altered myeloid differentiation. A subset of these mice also succumbed to BM failure. These findings are consistent with depression of TRAF6 signaling in hematopoietic cells following deletion of miR-146a, a microRNA within the commonly deleted region on chromosome 5q in MDS. miR-146a normally represses TRAF6 and IRAK1 expression as a means of dampening TLR-mediated NF-κB activation. However, reduced expression of miR-146a and the corresponding increase in TRAF6 expression in mouse hematopoietic cells resulted in thrombocytosis, mild neutropenia, megakaryocytic dysplasia, and development of BM failure or progression to AML, which are hematopoietic defects reminiscent of human del(5q) MDS. TIFAB and miR-146a are neighboring genes on chromosome 5q and are co-deleted in nearly all cases of del(5q) MDS and AML [
]. When TIFAB and miR-146a were co-deleted in mouse hematopoietic cells, de-repression of TRAF6 mRNA and protein expression was exacerbated as compared with what occurred after deletion of TIFAB or miR-146a alone. Combined deletion of TIFAB and miR-146a in BM cells resulted in a more rapid and penetrant BM failure-like disease as compared with singular deletion of TIFAB or miR-146a, thus more accurately recapitulating the pathogenesis of del(5q) MDS. These seminal studies indicated that deletion of TIFAB, as observed in human del(5q) MDS, substantially impairs normal hematopoiesis.
Given that TIFAB played an important role in hematopoiesis and is implicated in hematologic malignancies associated with del(5q), Niederkorn et al. [
] used tandem-affinity purification and proteomic analysis to identify TIFAB-interacting proteins in del(5q) AML cells. They reported a TIFAB interactome in the del(5q) AML cell line (HL-60), which has helped to clarify key signaling networks affected by TIFAB. Putative TIFAB-interacting proteins appeared in a number of biological pathways. These pathways, predictably, represented innate immunity and inflammatory signaling, but surprisingly, there was also over-representation of proteins involved in metabolism, the unfolded protein response, iron homeostasis and reactive oxygen species regulation, and apoptosis. A number of these TIFAB-interacting proteins appeared to function in ubiquitin-dependent regulatory pathways, partly explained by the interaction of TIFAB with USP15. In parallel, gene expression analyses of TIFAB-deficient HSPCs indicated that many of these same pathways were affected at the RNA level when TIFAB was deleted. Collectively, this group of biological functions unveiled by the TIFAB interactome and by RNA expression analyses constitute what is broadly considered the cellular stress response. The aforementioned studies raised outstanding questions about how TIFAB could affect HSPC function in normal versus stressed hematopoiesis. In the prior study, deletion of TIFAB resulted in blood and BM defects including expanded HSPC compartments, aberrant myeloid differentiation, progressive cytopenias, and BM failure [
]. These hematopoietic defects were observed only after transplantation of TIFAB-deficient BM cells into lethally irradiated mice, suggesting that the stress of the transplantation model is required to initiate the phenotype of TIFAB-deficient HSPC in vivo. On the basis of these observations, it was unclear how TIFAB-deficient HSPCs would respond to an environmental challenge. To answer this question, Niederkorn et al. [
] subjected TIFAB-deficient HSPCs to a variety of stressors, both in vitro and in vivo, which mimicked conditions of DNA damage, viral infection, and chemotherapy. The deletion of TIFAB sensitized murine HSPCs to all of these cellular stressors, which significantly impaired hematopoietic progenitor function of TIFAB-deficient cells. The impaired function of TIFAB-deficient HSPCs following exposure to these cellular stressors could be rescued by suppressing p53 expression. An analysis of p53 protein in TIFAB-deficient HSPCs revealed that p53 is markedly elevated despite the fact that NF-κB signaling is also increased, suggesting that TIFAB may regulate p53 signaling. These findings uncovered the existence of TIFAB-related functions in hematopoietic cells that diverge from—or occur in parallel to—the regulation of TRAF6-dependent innate immune signaling.
The sensitivity of TIFAB-deficient HSPCs to cellular stressors implicated USP15 as a relevant binding partner of TIFAB. USP15, which was identified as the top binding partner of TIFAB in AML cells, is a deubiquitinating enzyme that removes ubiquitin moieties from its substrates. By removing ubiquitin from its targets, USP15 either protects them from proteasomal degradation or disrupts their ubiquitin-dependent cellular signaling pathways [
]. Importantly, two substrates of USP15 were previously implicated in the cellular stress response, MDM2 and KEAP1. MDM2 is an E3 ligase that normally sequesters p53 and tags it for proteasomal degradation, preventing apoptosis or cell cycle arrest. KEAP1 is the E3 ligase in a cullin–RING complex that sequesters NRF2 and prevents a cellular response to oxidative stress. USP15 removes ubiquitin chains from both MDM2 and KEAP1, thus preventing their proteasome-mediated degradation and, consequently, promoting repression of apoptosis and the antioxidant response, respectively [
]. Although the deletion of TIFAB in murine BM cells had little effect on levels of USP15 itself, a substantial reduction in MDM2 and KEAP1 could be observed. Ultimately the impairment of MDM2 and KEAP1 unleashed the cellular stress response through p53 in TIFAB-deficient HSPCs, thus sensitizing them to multiple cytotoxic insults [
]. Given that TIFAB and USP15 are both highly expressed in hematopoietic cells, it is possible that their interaction represents a tissue-specific, fine-tuned regulatory mechanism during hematopoiesis. This underlying biology indicates that the deletion of TIFAB uncouples pro-survival cues from the innate immune pathway and pro-death cues from the p53 pathway in hematopoiesis. Without cellular stress, these conflicting intrinsic mechanisms are balanced and the TIFAB-deficient HSPCs persist. However, under stress, the absence of TIFAB tips the balance of hematopoietic cells in favor of cell death.
TIFAB in hematopoietic and nonhematopoietic disease
In other studies, TIFAB was gaining traction as a suspect in other myeloid malignancies. Several studies focused on the biology of MDS, myeloid neoplasms, and mixed lineage leukemias (MLLs) correlated the expression of TIFAB with variable functional outcomes, hinting at potential pleiotropic functions of TIFAB in hematopoietic disease. In a retroviral insertional mutagenesis screen for mutations that would cooperate with EGR1 haploinsufficiency to produce myeloid neoplasms, a common integration site was identified proximal to the TIFAB locus [
]. In these particular EGR1± myeloid neoplasm samples, TIFAB expression tended to increase. This occurs in stark contrast to the reduced expression of TIFAB in del(5q) MDS. This dichotomy, wherein increased TIFAB expression is observed in some neoplasms while decreased expression is observed in others, underscores just how versatile a role TIFAB might be playing in normal and malignant hematopoiesis. In an MLL-AF9-induced AML model, Xiu et al. [
] found that noncanonical NF-κB signaling via NIK ultimately represses canonical NF-κB in this system, resulting in decreased leukemic stem cell (LSC) function. In NIK-suppressed LSCs, the expression of TIFAB was markedly decreased [
] evaluated the kinetics of gene expression, including TIFAB, in response to the bromodomain inhibitor, JQ1, in AML cells expressing MLL-AF9 and NRas(G12D). In this system, c-Myc transcripts are acutely repressed within 2 hours of JQ1 treatment. An RNA-interference screen identified cooperative mechanisms that could induce JQ1 resistance and rapidly restore Myc transcripts. This study reports that TIFAB transcripts in several of these samples drop acutely in response to JQ1 exposure and indeed rebound over time, paralleling the relative expression of Myc in emerging resistant populations [
The paradigm observed in TIFAB-deficient HSPCs is reminiscent of human del(5q) MDS wherein defective HSPCs have reduced progenitor output, yet they persist and expand in the BM of patients. A massive effort in the field of MDS biology has been extended to understand the molecular events that dictate the pathogenesis of these disorders, with the hope that we can reinvigorate normal hematopoiesis to prevent both BM failure and transformation to AML in patients with del(5q) MDS. Several studies indicate that this primed p53 state in del(5q) MDS cells lends a therapeutic window for the reactivation of p53 to selectively target diseased HSPCs [
] or clones harboring cooperative mutations. Studies indicate that a substantial portion of low-risk del(5q) MDS patients can acquire TP53 mutations, which are associated with transformation to AML and poor response to therapy [
] may create an environment wherein HSPC clones capable of inactivating p53 gain a competitive advantage. Independent of del(5q) aberrations, TIFAB expression is quite variable among AML subtypes. TIFAB seems to be important for the leukemic stem cell function in MLL-AF9 models, but the requirement for TIFAB in many other myeloid malignancies remains unknown.
Despite recent advances in understanding TIFAB biology and molecular mechanisms in hematopoietic cells, the function of TIFAB in other cell types is entirely unknown. Sparse mutations in the TIFAB locus have been reported in the COSMIC database for cancer mutations [
], but the functional relevance of these alterations is unclear. Likewise, the amplification, deletion, or missense mutation of the TIFAB locus is reported in only 1% of the samples characterized in the TCGA PanCancer Atlas. Curiously, genomewide association studies identified SNPs in the TIFAB locus as a susceptibility for coronary artery aneurysm in patients with Kawasaki disease, an inflammatory disorder [
]. Although it is comparatively understudied, because of its ability to fine-tune immune signaling and the cellular stress response, TIFAB remains a relevant candidate in human disease that warrants further investigation.
Studies discussed in this review indicate that the FHA domain proteins TIFA and TIFAB vary in several characteristics (Table 1) and are likely to be differentially active in hematopoietic and various disorders (Table 2). Amidst emerging fields uncovering the fascinating connections between immune cell function, chronic inflammation, and malignant conditions, TIFA and TIFAB demand further understanding. Given the fact that TIFA and TIFAB are potential therapeutic targets in AML [
], studies elucidating structural information on TIFA and TIFAB should be useful for the development of novel therapeutics against infectious diseases and possibly cancers. These small FHA domain-containing proteins are potent signal transducers with pleiotropic effects, many of which are yet to be discovered.
Conflict of interest disclosure
DTS serves on the scientific advisory board at Kurome Therapeutics. All other authors have no competing financial interests to declare.
This work was supported in parts by the National Institutes of Health ( R35HL135787 , R01DK102759 , R01DK113639 to DTS; F99CA234924 , 4T32CA117846-10 to MN), Cancer Free Kids (DTS) , EvansMDS Foundation (DTS) , Pelotonia Fellowship (MN and PA) , and Cincinnati Children's Hospital Research Foundation (DTS) . DTS is a Leukemia and Lymphoma Society Scholar.
We thank the Starczynowski laboratory for insightful suggestions and feedback.
The FHA domain: a putative nuclear signalling domain found in protein kinases and transcription factors.
Identification of TIFA as an adapter protein that links tumor necrosis factor receptor-associated factor 6 (TRAF6) to interleukin-1 (IL-1) receptor-associated kinase-1 (IRAK-1) in IL-1 receptor signaling.