Clinical implications of tristetraprolin (TTP) modulation in the treatment of inflammatory diseases

Abstract

Abnormal regulation of pro-inflammatory cytokine and chemokine mediators can contribute to the excess inflammation characteristic of many autoimmune diseases, such as rheumatoid arthritis, psoriasis, Crohn’s disease, type 1 diabetes, and many others. The tristetraprolin (TTP) family consists of a small group of related RNA-binding proteins that bind to preferred AU-rich binding sites within the 3’-untranslated regions of specific mRNAs to promote mRNA deadenylation and decay. TTP deficient mice develop a severe systemic inflammatory syndrome consisting of arthritis, myeloid hyperplasia, dermatitis, autoimmunity and cachexia, due at least in part to the to the excess accumulation of proinflammatory chemokine and cytokine mRNAs and their encoded proteins. To investigate the possibility that increased TTP expression or activity might have a beneficial effect on inflammatory diseases, at least two mouse models have been developed that provide proof of principle that increasing TTP activity can promote the decay of pro-inflammatory and other relevant transcripts, and decrease the severity of mouse models of inflammatory disease. Animal studies of this type are summarized here, and we briefly review the prospects for harnessing these insights for the development of TTP-based anti-inflammatory treatments in humans.

1. Introduction.

Tristetraprolin (TTP) or zinc finger 36 (ZFP36) family proteins are a small group of mRNA binding and destabilizing proteins that are expressed in almost all eukaryotes. Three TTP family genes, all derived from a common ancestral gene, are expressed in both humans and mice (Zfp36, Zfp36l1, and Zfp36l2). Zfp36, encoding TTP, is probably the best characterized member of the family. Zfp36 or TTP knockout (KO) mice develop a spontaneous systemic inflammatory syndrome with erosive arthritis, myeloid hyperplasia, cachexia, and dermatitis, due at least in part to the excess accumulation of cytokine and chemokine mRNAs and their expressed proteins in the absence of TTP. Conversely, TTP overexpression studies in mice have shown that regulated TTP overexpression protects mice from several models of inflammatory diseases. These data suggest that increasing TTP expression in humans could be a novel therapeutic approach for treating human diseases involving inflammation and abnormal cytokine and chemokine expression.

This review will briefly summarize the roles and mechanisms of TTP family members in regulating cytokine expression. We will then review studies published to date that support the concept that increasing TTP mRNA stability, by genetically removing an instability element in its own mRNA, increases TTP protein expression in a regulated manner in the entire mouse, which in turn can protect mice from several different inflammation-driven disease models. We will also briefly summarize other studies that showed that increasing the activity of TTP, by inhibiting its phosphorylation at certain sites, can also protect mice from inflammatory damage in certain disease models. Finally, we will speculate about the potential relevance of these mouse studies to possible future applications of TTP modulation as a therapeutic approach in human inflammatory diseases.

2. Relevance of RNA-binding proteins (RBPs) to inflammatory gene regulation.

Post-transcriptional control of gene expression is vital for the regulation of cytokine and chemokine expression (Baou, Norton, & Murphy, 2011; Fu & Blackshear, 2017; Fukao & Fujiwara, 2017; Gerstberger, Hafner, & Tuschl, 2014; Kovarik, Ebner, & Sedlyarov, 2017; Makita, Takatori, & Nakajima, 2021). Certain classes of RNA-binding proteins (RBPs) bind to cis-elements in mRNA to regulate the stability of specific transcripts, which determines the ultimate amounts of mRNA and translated proteins in the cell. Defects in or changes in expression of such regulatory proteins can lead to abnormal accumulation of specific transcripts, such as inflammatory cytokines and chemokines. Increased levels of pro-inflammatory cytokines, such as tumor necrosis factor (TNF), are associated with the chronic inflammation characteristic of many autoimmune diseases, such as rheumatoid arthritis, multiple sclerosis, psoriasis, uveitis, Crohn’s disease, and periodontitis (Moudgil & Choubey, 2011). Inflammation is also involved in many forms of cancer (Nagarsheth, Wicha, & Zou, 2017). Novel classes of therapeutics that harness these pathways could be useful in treatment of many of these inflammatory diseases.

Tristetraprolin (TTP) or zinc finger protein 36 (ZFP36) family proteins comprise a small group of RNA-binding proteins that bind to specific transcripts, including cytokine mRNAs, and promote mRNA decay (Brooks & Blackshear, 2013; Lai, Wells, Perera, & Blackshear, 2019; Sanduja, Blanco, & Dixon, 2011; Wells, Hicks, Perera, & Blackshear, 2015; Wells, Perera, & Blackshear, 2017). TTP family proteins have been found in almost all eukaryotes studied to date, although the number of TTP proteins expressed in a given species varies considerably; for example, only one protein is found in many invertebrates and in the urochordate Ciona instestinalis, while teleost fish express approximately seven family members. Three TTP family genes appear to be conserved in all mammals (including mice and humans): Zfp36, Zfp36l1, and Zfp36l2, which code for their respective proteins: TTP or ZFP36, ZFP36L1, and ZFP36L2 (Lai, Stumpo, & Blackshear, 1990; Stumpo, et al., 2009; Stumpo, et al., 2004; Taylor, et al., 1996). Knocking out each of these family members results in unique phenotypes in mice: spontaneous chronic inflammation (Zfp36) (Taylor, et al., 1996), embryonic lethality (Zfp36l1) (Stumpo, et al., 2004), and hematopoietic failure resulting in early death (Zfp36l2) (Stumpo, et al., 2009). Mice, rats and many other rodents also express a fourth TTP family gene, Zfp36l3, which is only expressed in the placenta and yolk sac (Blackshear, et al., 2005; Frederick, Ramos, & Blackshear, 2008; Gingerich, et al., 2016; Stumpo, et al., 2016). All three of the widely distributed mammalian TTP proteins are expressed in nearly all tissues, with the expression patterns varying by cell-type; in some cases, as with TTP itself, expression can be transient and highly responsive to environmental stimuli (Cao, Tuttle, & Blackshear, 2004; Lai, Parker, Grissom, Stumpo, & Blackshear, 2006; Lai, et al., 1990). However, the focus of this review will be on TTP (ZFP36) itself.

Basal levels of TTP mRNA and protein are normally very low in cultured fibroblasts and macrophages, but upon stimulation with serum, mitogens, or other stimuli, TTP mRNA and protein levels rapidly increase (Lai, et al., 1990). In human diploid fibroblasts, TTP mRNA peaks at 30–60 minutes after stimulation, and then decreases rapidly; the protein increases with similar kinetics but then remains elevated for several hours (Qiu, Abey, et al., 2015). Similar results were also reported in mouse embryonic fibroblasts and mouse bone marrow-derived macrophages (BMDM) (Lai, Stumpo, Qiu, Faccio, & Blackshear, 2018). There are multiple adenylate-uridylate (AU)-rich binding potential TTP family member binding sites within the 3’-untranslated regions (3’-UTR) of all of the family members’ mRNAs, which probably contribute to their mRNA lability. In the case of TTP, recent work with a point mutant in mice has shown that TTP can regulate its own mRNA stability by binding to these AU-rich elements (AREs) and promoting the decay of its own mRNA (Lai, et al., 2018).

In addition, TTP expression and activity are thought to be regulated by the post-translational phosphorylation of TTP (reviewed in (Brooks & Blackshear, 2013)). Early studies demonstrated serine phosphorylation of TTP in cultured fibroblasts that could be stimulated by growth factors but not cyclic AMP-elevating agents, and implicated members of the mitogen-activated protein (MAP) kinases in this phenomenon (Taylor, Thompson, Lai, & Blackshear, 1995). These authors concluded that “…These results suggest that TTP function might be regulated by mitogens at the level of reversible phosphorylation” (Taylor, et al., 1995). Later studies suggested that TTP activity and perhaps expression levels could be regulated by multisite phosphorylation (Cao, Deterding, & Blackshear, 2007, 2014; Cao, et al., 2006; Carballo, et al., 2001; Clark & Dean, 2016; Gaestel, 2013; Rezcallah, Al-Mazi, & Ammit, 2021; Sandler & Stoecklin, 2008). TTP phosphorylation may reduce binding of TTP to AREs, resulting in the possible stabilization of TTP target mRNAs (Carballo, et al., 2001; Ross, et al., 2015). Importantly in the current context, phosphorylation at specific sites can lead to the interaction of phosphorylated TTP with cellular 14–3-3 proteins, which might have the effect of sequestering active TTP and preventing its access to its mRNA targets (Ross, et al., 2015; Tiedje, et al., 2016).

TTP family proteins generally contain two characteristic domains that are thought to be involved in their mechanism of action, as shown in Fig. 1. The defining domain is the characteristic tandem zinc finger (TZF) domain, which has two CX₈CX₅CX₃H motifs that are responsible for binding to its preferred mRNA target sequence, the nonameric motif UUAUUUAUU, which is present in the 3’-UTR of target mRNAs (Lai, et al., 2018; Lai, Wells, et al., 2019; Wells, Washington, et al., 2015). The 18 amino acid spacing between the zinc fingers is highly conserved in TTP family proteins, as are the (R/K)YKTELC lead-in sequences to each zinc finger (Hudson, Martinez-Yamout, Dyson, & Wright, 2004; Lai, Perera, Hicks, & Blackshear, 2014). TTP (ZFP36), ZFP36L1, and ZFP36L2 all bind tightly to ARE-containing mRNAs and promote mRNA deadenylation and decay in cell-based transfection experiments and cell-free deadenylation assays (Lai, Kennington, & Blackshear, 2003). Mammalian TTP family proteins also contain a C-terminal CNOT1 binding domain (NOT1 BD), which has been proposed to facilitate association with the CCR4-NOT deadenylase complex to promote mRNA deadenylation and eventual degradation of target transcripts (Fabian, et al., 2013; Lai, Stumpo, et al., 2019).

Figure 1.

Highly Conserved Domains of Tristetraprolin (TTP) Family Proteins. Mammalian TTP family proteins often contain a nuclear export sequence (NES). In the case of TTP, as shown here, this is at the extreme amino terminus; in other family members this is at the C-terminus. Also shown are the defining CCCH tandem zinc finger (TZF) domain, and the C-terminal NOT1-binding domain (NOT1 BD). The TZF domains bind to mRNA targets at the preferred nonameric binding site (UUAUUUAUU) to promote mRNA deadenylation and decay, thought to be at least in part mediated by the interaction of the protein with the CCR4-NOT complex via the TTP NOT1 BD.

As mentioned above, the whole body deletion in mice of Zfp36, the gene coding for TTP, leads to the spontaneous development of a chronic inflammatory syndrome characterized by polyarticular arthritis, dermatitis, myeloid hyperplasia, autoimmunity, and failure to gain weight (Taylor, et al., 1996). When 10-day old Zfp36 KO mice were treated with neutralizing antibodies raised against mouse TNF, the Zfp36 KO mice maintained similar growth curves to their wild-type (WT) and Zfp36 heterozygous littermates. In addition, the Zfp36 KO mice treated with antibodies did not develop arthritis, myeloid hyperplasia, and other aspects of the inflammatory syndrome that were readily observed in Zfp36 KO mice injected with PBS (Taylor, et al., 1996). In another approach to this question, Zfp36 KO mice interbred with TNF receptor 1 KO mice also did not develop the various aspects of the TTP-deficiency inflammatory syndrome (Carballo & Blackshear, 2001). Taken together, the data suggested that excess accumulation of TNF is largely responsible for the phenotype observed in Zfp36 KO mice, and that by neutralizing TNF activity, either genetically or by neutralizing antibodies, the phenotype can be virtually prevented. Other models of TNF excess often result in similar phenotypes, e.g., (Kontoyiannis, Pasparakis, Pizarro, Cominelli, & Kollias, 1999).

However, that other pro-inflammatory cytokines might play major roles in this “TTP-deficiency syndrome” was suggested by the observation that genetic deletion of the IL17/IL23 axis also protected mice from developing the TTP-deficiency phenotype (Molle, et al., 2013). Interbreeding TTP KO mice with mice lacking IL23 receptors also protected them from the development of the TTP deficiency phenotype (DJ Stumpo and PJ Blackshear, unpublished data). These data suggest that increasing the activity of the “TTP axis” might not have exactly the same effects as neutralizing the activity of TNF, currently used widely for the treatment of human inflammatory diseases.

It was then later shown that TTP can bind to its preferred binding site, UUAUUUAUU, in 3’-UTRs of other target mRNAs (Lai, et al., 1999a; Lai, et al., 2003). For example, Carballo et al. investigated the effect of TTP deficiency on levels of the mRNA encoding granulocyte-macrophage colony- stimulating factor (GM-CSF), a potent cytokine involved in many aspects of biology. These studies were initiated because of the presence in the Csf2 mRNA, the transcript encoding GM-CSF, of multiple copies of the UUAUUUAUU instability motif mentioned above (Carballo, Lai, & Blackshear, 2000). Not only did they identify the Csf2 transcript as a direct target of TTP, but these studies established that TTP could promote the deadenylation of that transcript, then as now thought to be the rate-limiting step in mRNA decay.

Since then, many other chemokines and cytokines, as well as other transcripts, have been identified as well-documented or potential TTP targets (summarized in (Brooks & Blackshear, 2013)). Besides Tnf and Csf2 mRNAs, these include Cxcl1 (Qiu, Lai, Bradbury, Zeldin, & Blackshear, 2015), Cxcl2 (Datta, et al., 2008; Qiu, Lai, et al., 2015), Il10 (Stoecklin, et al., 2008), Il17 (Lee, et al., 2012), Ccl3 (Kang, et al., 2011), and Il23 (Molle, et al., 2013). Thus, in the absence of TTP, Tnf mRNA and other mRNA targets, particularly other pro-inflammatory cytokines and chemokines, accumulate abnormally, leading to the development of the inflammation and leukocyte infiltration readily observed in Zfp36 KO mice. Zfp36 KO mice have increased serum levels of TNF and other cytokines. In addition, cells from Zfp36 KO mice (such as macrophages and fibroblasts) have increased accumulation and secretion of the cytokine and chemokines whose transcripts are targeted by TTP (Carballo, Lai, & Blackshear, 1998; Lai, et al., 1999b; Lai, et al., 2006).

The structure of the TZF domain from the human TTP family member ZFP36L2 (TIS11D), in complex with the 9-mer UUAUUUAUU, shows that the TZF domain interacts with the RNA through stacking interactions between the RNA bases and the aromatic rings of hydrophobic amino acids, as well as electrostatic interactions (Hudson, et al., 2004). RNA-binding assays have shown that mutating any one of the eight cysteine or histidine residues within the TTP TZF domain completely eliminates high-affinity RNA-binding (Lai, et al., 1999b; Lai, et al., 2018). In addition, when these TZF point mutations were made in co-transfection assays used to measure the activity of TTP on the decay of Tnf, Csf2, or Il3 mRNAs, the effect of TTP to promote decay of these transcripts was inhibited (Lai, Kennington, & Blackshear, 2002).

In order to determine if the biological activity of TTP is dependent on the ability of the TTP TZF domain to bind to RNA, as suggested by these experiments in cell culture and cell-free systems, Lai et al. developed a knock-in mouse with a cysteine-to-arginine (C116R) point mutation within the first zinc finger of the TTP TZF domain (Lai, et al., 2018). This mutation was knocked into the endogenous gene in order to make a mouse that otherwise normally expressed TTP protein from its normal genetic locus and otherwise normal mRNA, except that one mutated zinc-coordinating residue was present in one of its TZF domains. C116R TTP mRNA and protein were expressed at roughly the same levels as WT TTP mRNA and protein in the basal state. Briefly, the important conclusions from this study were that: 1. Mutating a single zinc coordinating residue in the TZF domain, thus eliminating mRNA binding, resulted in the same “clinical” syndrome in mice as complete TTP deficiency, suggesting that most or all of TTP’s physiological activities stem from its ability to bind mRNA; and 2. TTP has a significant but modest effect on its own mRNA stability. This study also emphasized the extreme lability of the TTP transcript: For example, in mouse fibroblasts stimulated with interleukin 1β, the half-life in the WT cells of Zfp36 mRNA was 12 minutes, and it was 17 minutes in lipopolysaccharide (LPS)-stimulated WT mouse macrophages. In other words, these data supported the concept that changes in the stability of this very labile mRNA might be expected to lead to significant changes in the accumulation of mRNA and, ultimately, protein.

Although the TZF domain is critical for RNA-binding, TTP is not thought to have intrinsic enzymatic activity, and must recruit deadenylases in order to deadenylate and degrade target mRNAs. Several papers have described interactions between TTP and members of the CCR4/NOT complex, especially CNOT1 (Bulbrook, et al., 2018; Fabian, et al., 2013; Lykke-Andersen & Wagner, 2005; Sandler, Kreth, Timmers, & Stoecklin, 2011; Webster, Stowell, & Passmore, 2019). There is a highly conserved sequence motif in the C-terminus of TTP, called the NOT1 BD, which has been shown to bind to a central domain of CNOT1 (Fabian, et al., 2013; Sandler, et al., 2011). CNOT1 is a large scaffolding protein that forms the structural basis of the CCR4-NOT complex, which includes two deadenylases (Collart, 2016; Doidge, Mittal, Aslam, & Winkler, 2012; Raisch, et al., 2019). Co-immunoprecipitation assays showed that WT TTP could pull down a CNOT1 fragment; however, TTP with the NOT1-BD deleted (TTPΔNOT1-BD) was unable to pull down the CNOT1 fragment (Lai, Stumpo, et al., 2019). These data further support a direct interaction between CNOT1 and the conserved C-terminal domain of TTP. When TTP lacking this domain (TTPΔNOT1-BD) was tested for TTP-like activity in co-transfection or deadenylation assays, TTP’s ability to promote decay of TTP targets was significantly reduced, but not completely inhibited (Lai, Stumpo, et al., 2019). This suggests that the NOT1-BD is important for TTP function, but is not as vital as the TZF domain, suggesting that there could be other interactions with deadenylases.

In order to determine the importance of the NOT1-BD in mice, Lai et al. (Lai, Stumpo, et al., 2019) developed a knock-in mouse model in which the C-terminal NOT1-BD of TTP was deleted (TTPΔNOT1-BD mice). They analyzed homozygous TTPΔNOT1-BD mice on two genetic backgrounds, as well as mixed heterozygote mice that expressed one allele of Zfp36 KO and one allele of TTPΔNOT1-BD. There were no significant differences in body weight when homozygous TTPΔNOT1-BD male and female mice were compared to WT mice, whereas the mixed heterozygotes exhibited significant decreases in body weight compared to TTP heterozygotes; however, this failure of weight gain was much less severe than that seen in the original Zfp36 KO mice. Homozygous TTPΔNOT1-BD mice showed relatively mild swelling and redness in one to several digits (dactylitis) of the front paws, but no involvement in the larger joints. This is in stark contrast to the severe peripheral joint arthritis with bone erosion and neutrophilic inflammation that is observed in Zfp36 KO and TTP C116R mice. The mixed heterozygous mice had arthritis that was more frequent and more severe than the dactylitis observed in homozygous TTPΔNOT1-BD mice. However, the arthritis observed in mixed heterozygotes consisted of soft tissue swelling and infiltration by neutrophils with minimal bone or cartilage destruction, which was far less severe than the arthritis seen in Zfp36 KO and C116R mice. Homozygous TTPΔNOT1-BD mice were less severely affected than Zfp36 KO mice, but mixed heterozygotes were more severely affected than heterozygous Zfp36 KO mice. This indicated that the ΔNOT1-BD allele is a hypomorphic allele, meaning that there is increased phenotype severity when it is placed in trans from a null allele.

There was a significant decrease in the mRNA decay of TTP target transcripts, such as Tnf, Cxcl1, and Cxcl2 mRNAs, in BMDM from heterozygous TTPΔNOT1-BD/Zfp36 KO mice after LPS and actinomycin D (ActD) treatment, when compared to BMDM from heterozygous Zfp36 KO/Zfp36 WT mice. However, this decrease in mRNA decay was less dramatic than that seen in BMDM from Zfp36 KO or TTP C116R mice. These data suggest that TTP with a C-terminal deletion of the CNOT1 binding domain is less active in vivo than WT TTP, but still has more mRNA destabilizing activity than null alleles of TTP. However, it appears that both the TZF domain and the NOT1 BD are necessary for normal TTP function in the mouse.

Many studies have also focused on the functions of TTP in specific cell types. Given the predominant role of TNF in the TTP deficiency syndrome, as described above, it has been hypothesized that knocking out TTP specifically in myeloid cells could be responsible for the entire pathogenesis of the TTP-deficiency phenotype. However, although myeloid-specific deletion of TTP in mice resulted elevations of serum TNF as expected, the mice exhibited no overt phenotype under normal vivarium conditions, suggesting that other cell types may also contribute to the TTP-deficiency syndrome, or that other TTP family members might compensate in myeloid cells (Qiu, Stumpo, & Blackshear, 2012). Despite this apparently normal phenotype in a normal laboratory environment, these mice were nonetheless hypersensitive to a low dose LPS challenge, and developed severe and lethal endotoxemia after an LPS dose that had minimal effects on the WT controls (Qiu, et al., 2012). Similar results with a similar model were reported by another group (Kratochvill, et al., 2011).

Work by the Goriely group has also investigated the role of TTP in myeloid, dendritic, and epidermal cells. Interestingly, mice with a complete deficiency of TTP in epidermal cells, such as keratinocytes, developed a spontaneous phenotype with arthritis, dermatitis, and systemic inflammation, similar to some aspects of the whole body TTP-deficiency phenotype (Andrianne, et al., 2017). In other cells, intestinal epithelial cell-specific TTP expression plays an important role in acute colitis (Eshelman, et al., 2019). In addition, TTP in lung alveolar epithelial cells may play an important role in regulating acute lung injury (Thorley, et al., 2007).

Overall, the data from the various loss of function mouse models demonstrate that TTP is vital for the normal regulation of biosynthesis and secretion of many cytokines and chemokines in many cell types in mice, and that disrupting TTP function leads to a severe form of inflammation that is similar to that observed in many human inflammatory diseases. The Zfp36 gene and TTP protein are highly conserved between mice and humans, and human mutations in TTP have been associated with rheumatoid arthritis and other autoimmune diseases (Carrick, et al., 2006; Suzuki, et al., 2008). Taken together, these data suggest that treatments that increase TTP levels or activity might be beneficial in the treatment of inflammatory diseases in humans, but these treatments might be required to modulate the activity of the TTP axis, not only in myeloid cells, but other cell types as well.

3. Effects of regulated overexpression of TTP on mouse models of inflammatory disease.

To address the question of whether increased endogenous levels of TTP in the intact mouse could be beneficial in the prevention and/or treatment of inflammatory diseases, Patial et al. generated a mouse model in which endogenous TTP expression was overexpressed in a regulated manner, referred to as TTPΔARE mice (Patial, et al., 2016). The 3’-UTR of TTP mRNA is highly conserved between mice and humans, and contains many AU-rich regions that could potentially regulate the stability of TTP mRNA and thus its expression. Fig. 2 shows the entire 3’-UTR of both human and mouse TTP mRNA, including the 136-base conserved AU-region in the 3’-UTR of mouse Zfp36 mRNA that was deleted using a targeted knock-in approach in the germline. With this approach, Zfp36 gene expression would still be regulated under its endogenous promoter and other controlling elements, but the removal of this potential instability element in TTP mRNA might be expected to stabilize the mRNA and increase protein expression.

Figure 2.

Schematic representation of the mouse Zfp36 gene and mRNA, and sequence alignment of mouse and human Zfp36 3’-UTRs. (A) The Zfp36 gene contains a promoter (not shown) and single intron shown in yellow, and mRNA coding exons shown in blue. (B) The mouse Zfp36 mRNA consists of a very short 5’UTR (blue), the protein coding region shown in green, and the 3’-UTR (blue). The segment of the 3’UTR that was deleted in the TTPΔARE mice is indicated in orange (ARE). (C) Shown is an alignment of the 3’UTRs from the human and mouse 3’UTRs. There is a an AU-rich region in both species that has particularly high sequence conservation, and this is indicted in the mouse transcript in red type. This was the 136 base AU-rich region that was deleted in the TTPΔARE mice.

Patial and colleagues found that the TTPΔARE mice had normal body weights, blood chemistries, complete blood counts, and bone marrow cell counts, and exhibited no obvious anatomical or cellular abnormalities by histopathologic and immunohistochemical analysis (Patial, et al., 2016). TTP mRNA expression in BMDM from TTPΔARE mice was significantly increased, both under unstimulated conditions and after LPS treatment compared to BMDM from WT mice (Patial, et al., 2016). TTP mRNA decay rates were significantly decreased in TTPΔARE BMDM after LPS and ActD treatment compared to WT BMDM; similar findings were observed in embryonic fibroblasts from the same animals. TTP protein expression was also markedly increased in TTPΔARE BMDM before and after LPS treatment, as well as in tissues from TTPΔARE mice, such as liver, spleen, and thymus (Patial, et al., 2016). These data suggest that deleting an AU-rich region from the Zfp36 mRNA 3’-UTR did indeed increase its stability, and led to the increased expression of TTP protein in cells and tissues from the TTPΔARE mice. In addition, the expression levels of some TTP targets, such as Tnf, Il1β, and Cxcl2 mRNAs, were lower in LPS-stimulated BMDM from TTPΔARE mice compared to BMDM from WT mice (Patial, et al., 2016). However, other probable direct targets were not changed in LPS-treated BMDM compared to WT BMDM, including Il10, Il6, and Il23a mRNAs, suggesting that the increased TTP expression may not result in measurable increased decay and decreased accumulation of all TTP targets under these conditions.

These foundational studies established several things about this approach. First, it seemed apparent that this rather large deletion in the TTP mRNA 3’UTR was compatible with a more or less normal life in the laboratory mouse, even in homozygotes. Second, this genetic deletion clearly increased the stability of the TTP mRNA in cells derived from the homozygous mice, resulting in increases in steady state and inducible levels of both mRNA and protein. Third, these increases in protein could be seen in many mouse tissues using anti-TTP antibodies. Taken together, these findings suggested that these mice might represent a useful animal model in which to test the hypothesis that the increases in TTP protein expression seen in the TTPΔARE mice, under otherwise normal physiological control, could prevent or decrease the severity of several models of inflammatory diseases. The following pages contain summaries of recent attempts to test this hypothesis in a number of models of inflammatory disease in mice, summarized in Table 1.

Table 1.

Summary of disease models tested against TTP ΔARE mice.

Animal Model Tested	Human Disease Equivalent	Result	Reference
LPS-induced endotoxemia	Septicemia and shock	TTPΔARE mice still have elevated TNF levels, but some cytokine levels were decreased in TTPΔARE mice compared to WT mice.	(Patial, et al., 2016)
Collagen antibody-induced arthritis (CAIA)	Rheumatoid arthritis	TTPΔARE mice were almost completely protected.	(Patial, et al., 2016)
Imiquimod (IMQ)-induced dermatitis	Psoriasis	TTPΔARE mice were largely protected from hyperkeratosis and neutrophil infiltration.	(Patial, et al., 2016)
Experimental autoimmune encephalitis (EAE)	Multiple sclerosis	All TTPΔARE mice survived the EAE protocol and were almost completely protected from clinical EAE.	(Patial, et al., 2016)
Alveolar Bone Loss	Periodontitis	TTPΔARE mice were protected against alveolar bone loss and inflammation.	(Steinkamp, et al., 2018)
Experimental autoimmune uveitis (EAU)	Autoimmune uveitis	Homozygous TTPΔARE mice were completely protected, and heterozygous mice were partly protected, from EAU.	(Xu, et al., 2020)
LPS-induced acute lung injury	Acute Respiratory Distress Syndrome (ARDS)	TTPΔARE mice were protected against LPS-induced acute lung injury, whereas overexpression of TTP in hematopoietic progenitor cells partially protected against LPS-induced acute lung injury.	(Choudhary, et al., 2020)
DMBA/TPA-induced papilloma formation	Skin carcinogenesis	TTPΔARE mice were protected from chemical-induced papilloma development.	(Assabban, et al., 2021)
Adrenalectomy-induced gastric metaplasia	Gastric metaplasia	Heterozygous and homozygous TTPΔARE mice were protected from adrenalectomy-induced gastric metaplasia.	(Busada, et al., 2021)

3.1. LPS-induced endotoxemia.

The first attempt to determine a possible anti-inflammatory effect of the TTPΔARE mutation used the LPS-induced model of endotoxemia. This is a well-established model for septic shock after a bacterial infection, and many aspects of the clinical syndrome are thought to be due to the excess accumulation of TNF (Plociennikowska, Hromada-Judycka, Borzecka, & Kwiatkowska, 2015). It had been shown previously that TTP deficiency specifically in myeloid cells resulted in mice that had elevated serum levels of TNF, both under normal conditions and after a low dose LPS challenge. These mice developed a lethal endotoxemia syndrome, whereas WT mice studied in parallel readily recovered after low dose LPS treatment (Qiu, et al., 2012).

Therefore, homozygous TTPΔARE mice and their controls were injected with three different doses of LPS, to see if they were protected from LPS-induced endotoxemia. The TTPΔARE mice did not have decreased peak levels of TNF in the serum after the LPS injections compared to WT, nor was subsequent rate of decrease of serum TNF altered in the mutant animals. However, it appeared that serum levels of IL-10, the product of another TTP target transcript, were lower and decreased faster in the TTPΔARE mice than in the WT controls after the highest dose of LPS challenge. These data suggested that the acute TNF response to injected LPS was not affected by the increased levels of TTP, but the apparent modulation of other cytokines suggested that the increased TTP might be playing a role in regulating at least some cytokine responses to LPS.

3.2. Collagen antibody-induced arthritis (CAIA).

In addition to the acute LPS injection model, more chronic models of inflammatory disease in the mouse were tested in the same study. Rheumatoid arthritis is a chronic, debilitating disease that causes synovitis leading to cartilage and bone erosion, and is estimated to affect over 1 million adults each year in the United States (Hunter, et al., 2017). Collagen antibody-induced arthritis (CAIA) is a widely used mouse model of inflammatory arthritis that resembles some aspects of human rheumatoid arthritis (Caplazi, et al., 2015; Holmdahl, et al., 1989; Khachigian, 2006; Staines & Wooley, 1994). In this model, arthritis is stimulated by treating mice with a cocktail of monoclonal antibodies raised against type II collagen, followed by treatment with an endotoxin, such as Escherichia coli-derived LPS. After treatment, arthritis is rapidly induced, and mice are monitored daily with a clinical disease severity scale for 1–2 weeks, after which paws and joints are collected for RNA and histology (Khachigian, 2006).

When TTPΔARE and WT mice were subjected to the standard CAIA protocol, both WT and TTPΔARE mice initially lost weight after the LPS injection; however, the TTPΔARE mice recovered to their pre-injection weight by day 9, whereas the WT mice did not recover to their original body weight. All WT mice had high clinical arthritis scores, with marked redness, swelling, and ankylosis of the tarsal joints and other joints from day 6 on, while the TTPΔARE mice had no or few clinical signs (Patial, et al., 2016). Histopathological analysis of paws and joints of the WT mice after CAIA induction showed very evident disease, with multiple joints affected per mouse, whereas only two of the seven TTPΔARE mice exhibited any signs of disease, with minimal signs of disease affecting only one or two joints per mouse. Histopathology scores for the four joints combined were consistently and markedly higher in WT mice compared with TTPΔARE mice. These data suggested that this model of regulated overexpression of TTP almost completely protected mice from developing CAIA.

It was presumed that the mechanism of protection would be through down-regulation of TTP target transcripts encoding pro-inflammatory cytokines and other mediators, and this was supported by decreased serum levels of IL-6 and G-CSF in the TTPΔARE mice on day 7. G-CSF in particular is interesting in this context, since G-CSF levels are elevated in Zfp36 KO mouse plasma (Kaplan, et al., 2011) and G-CSF KO mice are protected from CAIA (Lawlor, et al., 2004), implicating G-CSF in the pathogenesis of arthritis in this model. Csf1 mRNA, encoding G-CSF, contains two UAUUUAU sequences and other potential TTP binding sites, suggesting the possibility of direct TTP binding to Csf1. Similarly, IL-6 has also been implicated in the pathogenesis of rheumatoid arthritis, and IL-6 inhibitors (sarilumab and tocilizumab) have been used for the treatment of rheumatoid arthritis (Narazaki, Tanaka, & Kishimoto, 2017; Ogata, Kato, Higa, & Yoshizaki, 2019). Il6 mRNA is also thought to be a direct TTP target, and IL-6 levels are increased in both Zfp36 KO and TTP C116R mice compared to WT mice (Lai, et al., 2018; Zhao, Liu, D’Silva, & Kirkwood, 2011).

In the same model, gene expression profiling was performed on mRNA from whole joints of the WT and TTPΔARE mice after CAIA induction (Patial, et al., 2016). At least five of the downregulated transcripts (Ccl2, Cd20, Cxcl1, Il1β, and Il6) are known or suspected TTP targets, with typical TTP binding sites within their 3’UTRs. These data support the concept that there was down-regulation of inflammatory gene expression in the TTPΔARE mice, which was in turn responsible for the prevention of CAIA.

3.3. Imiquimod-induced dermatitis.

Imiquimod (IMQ)-induced dermatitis is a well-established mouse model of human psoriasis (Flutter & Nestle, 2013; Moos, Mohebiany, Waisman, & Kurschus, 2019; Singh, Zhang, Hwang, & Farber, 2019; van der Fits, et al., 2009). Skin inflammation with epidermal thickening, similar to that observed in psoriasis, is induced in mice by applying imiquimod (IMQ) cream onto shaved skin for five consecutive days (Flutter & Nestle, 2013). Skinfold thickness is then analyzed. In order to determine if regulated overexpression of TTP could prevent or ameliorate this form of dermatitis, WT and TTPΔARE mice were treated with IMQ, and skin thickness, skin histopathology, and gene expression analysis were evaluated at the endpoint of the study (Patial, et al., 2016). A modest decrease in skinfold thickness was observed in the TTPΔARE mice compared to WT mice, while the decrease in epidermal thickness observed on histology was more marked. Subepidermal infiltration of inflammatory cells, shown by hematoxylin and eosin (H and E) staining and immunostaining for neutrophils, was markedly decreased in the TTPΔARE mice treated with IMQ compared to the WT mice treated in parallel (Patial, et al., 2016). Overall, the WT mice exhibited more severe epidermal hyperplasia, parakeratosis, and dermal inflammation, whereas the TTPΔARE mice showed more pronounced hyperkeratosis.

Gene expression analysis of skin from the TTPΔARE and WT mice after treatment with IMQ showed that eleven inflammatory transcripts were down-regulated in skin from the TTPΔARE mice compared to that of WT mice. Among these were Cxcl1, Il12β, Il1α, and Myc mRNAs, which are all known or suspected TTP targets. These data showed that although the TTPΔARE mice were still susceptible to IMQ-induced dermatitis, the response was much less severe than that of the WT mice, possibly due in part to decreased inflammatory gene expression, seen in skin of the TTPΔARE animals.

3.4. Experimental autoimmune encephalitis (EAE).

Experimental autoimmune encephalitis (EAE) is a widely used mouse model for multiple sclerosis (MS) in which 17–18-week-old mice are immunized with myelin oligodendrocyte glycoprotein (MOG_35–55), a minor component of myelin, followed by treatment with pertussis toxin (Miller & Karpus, 2007). Animals are then monitored for body weight loss and clinical signs of neurological impairment for 30 days. To determine whether the TTPΔARE mice would be less susceptible to this version of EAE than WT mice, mice of both genotypes were induced with EAE and monitored by observers blinded to the genotypes of the mice (Patial, et al., 2016). When the genotype code was broken at the end of this experiment, it was apparent that the WT mice began losing body weight by about day 8 after immunization, whereas the TTPΔARE mice did not lose body weight. The WT mice started exhibiting clinical signs of EAE by about day 10 that got progressively more severe over time, whereas TTPΔARE mice did not show any clinical signs until about 5 days later. By day 20, three of the WT mice required euthanasia, and the remaining nine mice showed severe clinical impairment. In contrast, by day 20, only four of the ten TTPΔARE mice had mild clinical signs, and none required euthanasia. By day 30, eight of the original eleven WT mice remained and all exhibited severe EAE, whereas all ten of the original ten TTPΔARE mice remained and only half of them had clinical signs of mild EAE. The mean cumulative disease scores and the mean maximal disease scores for the TTPΔARE mice were much lower than those of the WT mice. These data demonstrated that the TTPΔARE mice were markedly resistant to the induction of EAE in this specific model compared to WT mice, raising the possibility that increasing TTP expression in humans could be a potential therapeutic approach in the treatment of multiple sclerosis.

3.5. Alveolar bone loss and periodontitis.

Periodontitis is the sixth most prevalent disease in the world and affects nearly half of the American adult population (Eke, et al., 2015). Periodontitis is a chronic inflammatory disease that leads to epithelial and connective tissue degradation and alveolar bone loss when the balance between pro- and anti-inflammatory factors is compromised. Since TTP tightly regulates pro-inflammatory mRNA stability, it was hypothesized that TTP could play a role in alveolar bone homeostasis. Zfp36 KO mice exhibited severe alveolar bone loss as determined by micro-CT scans of maxillae (Steinkamp, et al., 2018). This was associated with increased inflammation, upregulated osteoclast cellular endpoints, shown by tartrate-resistant acid phosphate (TRAP) staining, increased cervical lymph node size, and adaptive cellular immune profiles (Steinkamp, et al., 2018).

Carboxy-terminal collagen crosslinks 1, a bone turnover marker, was up-regulated in Zfp36 KO versus WT mice over time. However, there was no difference in osteocalcin expression, a non-collagenous protein produced by osteoblasts, suggesting that TTP plays a role in bone turnover, but does not significantly regulate osteoblastic bone formation. TNF enhances osteoclastogenesis, inhibits osteoblastogenesis, and is a well- known TTP target. In this study, serum TNF levels were significantly increased in Zfp36 KO mice compared to WT mice at all time points.

Since osteoclastogenic activity and inflammation activity were altered, these investigators further studied the mechanisms involved in TTP’s role in alveolar bone maintenance. When periodontal homeostasis is dysregulated, innate immune cell types become activated within the oral cavity and drain into the cervical lymph nodes to elicit an immune response from the adaptive and humoral immune systems. Zfp36 KO mice had significantly increased cervical lymph node size and increased numbers of plasma cells in the medulla and distended sinusoids compared to WT mice. The memory B-cell populations, activated and inactivated dendritic cell populations, and monocyte populations were elevated in Zfp36 KO mice compared with WT mice, and differences were seen in the microbiomes of WT and Zfp36 KO mice. These data suggest that TTP plays an important role in regulating osteoclastogenic activity and inflammation in normal alveolar bone maintenance.

In order to determine whether increased TTP expression could prevent alveolar bone loss during normal aging, WT and TTPΔARE mice were aged (3 months, 6 months, and 9 months) and the maxillae were analyzed by micro-CT, histomorphometric analysis, and TRAP staining. TTPΔARE mice exhibited 13.4% less bone turnover than WT mice. Histomorphometric analysis of H and E-stained maxillae showed normal to minimal inflammation in both groups, suggesting that overexpression of TTP may have been protective against alveolar inflammation. TRAP-stained maxillae sections showed similar numbers and sizes of osteoclasts between WT and TTPΔARE mice, suggesting that TTP overexpression does not alter cellular commitment to an osteoclast population. These data suggested that TTP overexpression may protect mice against inflammation-driven bone-resorption. In addition, TTP appears to be a key molecule in the processes of alveolar bone maintenance and health, and could be a potential therapeutic target for modulation of host inflammatory responses in the prevention of alveolar bone loss.

3.6. Autoimmune uveitis.

Non-infectious uveitis is a common cause of blindness that is thought to be caused by cellular autoimmune processes and mediated by cytokines. Non-infectious uveitis conditions are estimated to cause 10–15% of cases of blindness in western countries (Durrani, et al., 2004; Nussenblatt, 1990). Uveitis presents as localized eye inflammation, but the term ‘uveitis’ refers to a family of eye conditions characterized by intraocular inflammation, that includes: sympathetic ophthalmia, “birdshot” chorioretinopathy, Behcet’s disease, sarcoidosis, and Cogt-Koyanagi-Haradi diease (Caspi, 2010; Nussenblatt & Gery, 1996). Typical current treatments for uveitis involve glucocorticoids or cyclosporin. However, “biologic” therapies targeting TNF have been used more recently, since TNF levels were found to be increased in aqueous humor of patients with uveitis (Becker & Davis, 2005; Murphy, et al., 2004).

The pathogenesis and ocular changes observed in these diseases are very similar to the pathological and ocular changes seen in rodents with experimental autoimmune uveitis (EAU). EAU is an autoimmune disease mediated by immunopathogenic T cells, and can be induced in mice by either immunizing mice with the retinal antigen, interphotoreceptor retinoid-binding protein (IRBP), or by adoptive transfer of T-cells sensitized against IRBP (Caspi, et al., 1988; Luger, et al., 2008; Rizzo, et al., 1996). In a recent study, both WT and TTPΔARE mice were immunized by subcutaneous injection of emulsion containing bovine IRBP and human IRBP uveitogenic peptide and pertussin toxin, and the development of EAU was monitored by fundoscopy on days 9, 11, and 13 post-immunization to determine disease onset and progression using an established eye inflammation scale (Xu, et al., 2020).

Representative fundus images from WT and TTPΔARE mice showed moderate to severe disease changes in all WT mice, whereas very little or no disease was found in the homozygous TTPΔARE mice. Histopathological analyses of tissues (22 mice per group) showed highly significant differences in inflammation and tissue damage between WT and TTPΔARE mice. Heterozygous mice developed EAU and associated immunological responses at levels intermediate between homozygous TTPΔARE mice and WT controls, suggesting that the induction of EAU is dependent on the gene dose of Zfp36. Draining lymph node cells (lymphocytes) cultured with IRBP from WT mice produced significantly higher levels of the pro-inflammatory cytokines (IFN-γ, IL-17, IL-6, and TNF) than homozygous or heterozygous TTPΔARE mice. Il6 and Tnf mRNAs are well-established TTP targets, suggesting that the regulated TTP overexpression in the TTPΔARE mice prevents EAU by decreasing the expression of cytokines encoded by TTP target mRNAs. In addition, draining lymph node cells cultured with IRBP from WT mice produced significantly lower levels of the anti-inflammatory cytokine IL-10 compared to homozygous or heterozygous TTPΔARE mice.

In order to determine the involvement of T regulatory (Treg) cells in resistance of TTPΔARE mice to EAU, the authors tested the frequencies and suppressive function of Treg cells from spleen and draining lymph node cells from immunized TTPΔARE and WT mice. Flow cytometry analysis showed higher percentages of FoxP3+ cells in spleen cells, lymph node cells, and peripheral blood from TTPΔARE mice than in WT mice. In addition, a well-established Treg suppression assay showed that Treg cells from TTPΔARE mice were more suppressive than the WT controls on a per-cell basis.

As another measure of T cell activity, they measured IRBP-specific antibody levels in sera from each group of mice by ELISA. The levels of IgG and IgG1 antibodies against uveitogenic protein in sera from WT mice were significantly higher than the levels in sera from homozygous TTPΔARE mice, whereas antibody levels from heterozygous mice were intermediate. IgG2a is a marker for Th1 responder cells, and essentially no IgG2a antibody to IRBP was detected in sera from homozygous TTPΔARE mice, and intermediate levels of IgG2a were detected in sera from heterozygous TTPΔARE mice, compared to the high levels of IgG2a antibody to IRBP seen in the WT mice.

Next, they used an adoptive transfer system in which EAU can be transferred by injecting activated lymphocytes specific to IRBP or its pathogenic peptide into the WT or TTPΔARE mice (Mattapallil, et al., 2015). EAU from adoptive transfer generally occurs much faster (5 days) than the conventional disease caused by immunization (9 days). The levels of disease activity measured by fundoscopy and histological analysis after adoptive transfer were similar in the recipient TTPΔARE and WT mice, meaning that effector cytokines produced by donor cells were necessary and sufficient to induce pathology. These data demonstrate that TTPΔARE mice can develop EAU similarly to their WT counterparts when the disease is mediated by activated lymphocytes, and that the protection from disease is primarily through the decreased reactivity of TTPΔARE lymphocytes.

To determine whether the resistance to the development of EAU was due to a defect in the ability of T cells to respond to specific antigenic stimuli presented on antigen-presenting cells, the authors mimicked the process of antigenic stimulation by incubating naïve T-cells with antibodies against CD3 and CD28 to determine whether T cells from homozygous TTPΔARE mice were capable of being stimulated. There were no changes between the responses of the cells from WT or TTPΔARE mice, as determined by proliferation or cytokine release. This indicated that the inability of TTPΔARE mice to develop EAU is not due to the inability of their T cells to respond via T cell receptor activation.

Since antigen presentation by antigen-presenting cells to T-lymphocytes can determine the magnitude of the immune response, they examined the ability of antigen-presenting cells from TTPΔARE mice to prime naïve T cells from OVA-TCR transgenic OTII mice. TTPΔARE antigen presenting cells were significantly less efficient compared to WT antigen-presenting cells in priming naïve T cells, suggesting that deficiency in antigen presentation plays a role in the dampened immune responses of the TTPΔARE mice.

Overall, these studies suggest that elevated, regulated systemic levels of TTP could inhibit the pathogenic processes involved in EAU, a model of inflammatory eye diseases. It was proposed that EAU was prevented through the ability of increased levels of TTP in many cell types to restrict the pathogenic cytokine response after immunization, through destabilization of their mRNAs.

3.7. Acute lung injury in mice overexpressing TTP in non-hematopoietic cells.

Acute lung injury is characterized by elevated levels of pro-inflammatory mediators, exaggerated neutrophil recruitment, a compromised pulmonary epithelial-endothelial barrier, and increased vascular permeability. Non-cardiogenic pulmonary edema causes excess accumulation of edematous fluid and inflammatory cells in the alveolar spaces, resulting in hypoxemia in the acute respiratory distress syndrome, often requiring mechanical ventilation. Both acute lung injury and the acute respiratory distress syndrome have high rates of mortality, but many aspects of their pathogenesis remain unclear (Butt, Kurdowska, & Allen, 2016; Matthay & Zemans, 2011).

Lungs from Zfp36 KO mice have rare foci of leukocyte infiltration, but overall have low spontaneous inflammation (Qiu, Lai, et al., 2015). However, lungs from mice with the combined deletion of TTP and TNFR1 have no inflammation, suggesting that TNF is implicated in the lung leukocyte infiltration that is rarely seen in Zfp36 KO mice (Qiu, Lai, et al., 2015). To investigate the role of TTP in regulating a model of acute lung inflammation, WT, Zfp36 KO, and TTPΔARE mice were subjected to LPS-induced acute lung injury, using an oropharyngeal aspiration approach, and the mice were monitored for signs of respiratory distress after the LPS challenge (Choudhary, et al., 2020). Both Zfp36 KO and WT mice exhibited increased infiltration of immune cells after LPS administration, and the total number of recovered immune cells (neutrophils, lymphocytes, and macrophages) in LPS-challenged Zfp36 KO mice was fourfold higher compared to LPS-challenged WT mice. There were also red blood cells in the bronchoalveolar lavage fluid (BALF) of Zfp36 KO mice, indicating apparent injury to the pulmonary vascular barrier. LPS-challenged WT mice had mild to moderate lung consolidation (~26%), two- to four-fold increases in alveolar septal thickening, moderate perivascular and airspace edema, and perivascular inflammation. In contrast, LPS-challenged Zfp36 KO mice had severe consolidation (~90%), with infiltration of neutrophils, edema, fibrin, and airspace hemorrhage within the airway and alveolar lumen, multifocal loss of bronchiolar epithelium with infiltration of neutrophils and red blood cells within the bronchiolar lumen, and moderate to severe perivascular edema and inflammation. Furthermore, approximately half of the LPS-challenged Zfp36 KO mice died from the LPS challenge before 72 hours and were excluded from the analysis. These data suggest that systemic loss of TTP leads to extreme susceptibility of mice to LPS-induced acute lung injury (Choudhary, et al., 2020).

Since myeloid-specific TTP deficiency leads to severe endotoxemia after a low dose LPS-challenge (Qiu, et al., 2012), Choudhary et al. investigated a possible role for myeloid cell TTP in the pathogenesis of acute lung injury. LPS administration to myeloid-specific Zfp36 KO mice and control mice resulted in increased numbers of immune cells in both groups, but the increase was significantly greater in the myeloid-specific Zfp36 KO mice than in the controls. The increase in cellular infiltration was approximately threefold less in LPS-challenged myeloid-specific Zfp36 KO mice when compared to the LPS-challenged whole body Zfp36 KO mice. The neutrophil counts in BALF were similar between LPS-challenged myeloid-specific Zfp36 KO mice and control groups; however, macrophage counts were significantly increased in the LPS-challenged, myeloid-specific Zfp36 KO mice compared to the control groups. Histopathological analysis showed comparable levels of lung consolidation, with widespread inflammatory cellular infiltrates within the airspaces, of both myeloid-specific Zfp36 KO and control mice. Unlike the LPS-challenged whole body Zfp36 KO mice, there was no airspace hemorrhage in the myeloid-specific Zfp36 KO mice, and all myeloid-specific Zfp36 KO mice survived the LPS challenge. Overall, the LPS challenge of myeloid-specific Zfp36 KO mice resulted in less severe acute lung injury than the LPS challenge of whole body Zfp36 KO mice.

To determine whether regulated systemic over-expression of TTP could protect against LPS-induced acute lung injury, TTPΔARE mice were challenged with LPS and evaluated for the development of acute lung injury. BALF from LPS-challenged TTPΔARE mice contained significantly fewer immune cells compared to that from LPS-challenged WT mice at all time points. Histologically, LPS-challenged WT mice exhibited ~33% lung consolidation, with increased infiltration of immune cells within the airspaces, bronchiolitis, perivascular edema, and inflammation. In comparison, LPS-challenged TTPΔARE mice exhibited less bronchiolitis and perivascular edema, but there were no significant differences in lung consolidation, perivascular inflammation, or airspace edema compared to the LPS-challenged WT mice.

In order to determine whether hematopoietic cell TTP over-expression is essential for the protection of TTPΔARE mice from this model of acute lung injury, these authors tested the effects of transplantation of hematopoietic progenitor cells (HPCs) from the various mouse genotypes. In one set of experiments, irradiated WT mice received HPCs from either WT, TTPΔARE, or Zfp36 KO mice. Overall, they found that transplantation of Zfp36 KO HPCs still resulted in increases of severity of acute lung injury, but that overexpression of TTP in the TTPΔARE HPCs did not offer much protection against LPS-induced acute lung injury.

To determine whether non-HPCs from TTPΔARE mice could confer protective effects against LPS-induced acute lung injury, they generated chimeras in which irradiated TTPΔARE mice received HPCs from WT, TTPΔARE, or Zfp36 KO mice. Compared to the WT recipient chimeras, LPS-challenged TTPΔARE recipient chimeras exhibited significantly less cellular recruitment (including total cells, neutrophils, and macrophages). Of the various combinations examined, the LPS-challenged Zfp36 KO→TTPΔARE mice (i.e., Zfp36 KO HPCs were transplanted into irradiated TTPΔARE recipient mice), showed significantly greater cellular recruitment, compared to the WT→TTPΔARE and LPS-challenged TTPΔARE→TTPΔARE chimeras. However, the cellular recruitment in the Zfp36 KO→TTPΔARE lungs was still significantly lower than that seen in the Zfp36 KO→WT chimeras. Histologically, lung consolidation was reduced by half in the WT→TTPΔARE and TTPΔARE→TTPΔARE mice, compared to the TTPΔARE→WT and WT→WT mice. Overall, the histological findings (septal thickening, cellular infiltration, mild bronchiolitis, perivascular edema, inflammation, and airspace edema) in the TTPΔARE recipient mice were mild. These data suggest that the increased expression of TTP present in the lung non-hematopoietic cell populations from the TTPΔARE mice could protect them against acute lung injury, but that this protection was somewhat compromised in the absence of TTP in the transplanted HPCs.

They also tested what would happen if TTPΔARE HPCs were transplanted into Zfp36 KO recipients. The BALF cell counts observed were significantly higher than in any of the other chimeras, but were still lower than in the LPS-challenged Zfp36 KO mice. In addition, none of the LPS-challenged TTPΔARE→Zfp36 KO chimeras died, compared to the 50% mortality of LPS-challenged Zfp36 KO mice. These data suggested that loss of TTP in lung non-HPC populations severely worsens LPS-induced acute lung injury, but that overexpression of TTP in HPCs offers partial protection from severe acute lung injury.

Summarizing these experiments, it appeared that TTP overexpression in HPCs is partially protective against LPS-induced acute lung injury when non-HPCs express at least basal levels of TTP. However, when TTP levels are depleted in non-HPCs, overexpression of TTP in HPCs is not sufficient to protect against LPS-induced acute lung injury. Over-expression of TTP in non-HPCs was sufficient to protect against LPS-induced acute lung injury, even when the HPCs were deficient in TTP. However, overexpression of TTP in non-HPCs provided the greatest protection when at least WT levels of TTP were expressed in the HPCs. The non-HPC population in the lungs consists of a several cell types, including epithelial cells, endothelial cells, and fibroblasts, but it is still unknown which of these is involved in the apparently protective effect of increased TTP in acute lung injury.

The authors (Choudhary, et al., 2020) proposed a model in which an endotoxin such as LPS damages the epithelial cells within the lungs, resulting in exaggerated releases of pro-inflammatory mediators and neutrophilic chemo-attractants, causing further lung damage and neutrophilic infiltration. TTP is a vital intracellular regulator of the expression of many of these pro-inflammatory mediators and neutrophilic chemo-attractants, and TTP overexpression in non-HPCs may confer protection against lung injury by suppressing mRNAs encoding pro-inflammatory mediators and chemo-attractants. They proposed that increasing TTP expression could be a beneficial therapeutic approach for patients with acute lung injury, and TTP levels within the lung might possibly represent a prognostic indicator for the severity of acute lung injury and respiratory distress syndromes.

3.8. Skin papilloma development.

Altered post-transcriptional mechanisms, such as modulation of mRNA stability, can affect cancer-related gene expression pathways. TTP has been found to be dysregulated in several human cancers. For example, in breast tumor cell lines, miR-29a can regulate TTP expression, affecting epithelial polarity and metastasis (Gebeshuber, Zatloukal, & Martinez, 2009). Another study showed that certain ZFP36 polymorphisms were associated with a poor prognosis in breast cancer patients (Upadhyay, Dixit, Manvar, Chaturvedi, & Pandey, 2013). In hepatocellular carcinoma cell lines, hypermethylation of the Zfp36 promoter favors tumor growth (Sohn, et al., 2010). In glioma cell lines, hyperphosphorylated forms of TTP predominate, potentially resulting in increased VEGF and IL8 mRNA stability (Suswam, et al., 2008). Myc oncoprotein directly suppresses TTP expression, leading to abnormal overexpression of ARE-containing mRNAs (Rounbehler, et al., 2012). Oncogenic Ras signaling can also modulate TTP activity, leading to increased PD-L1 expression, indicating that this pathway may negatively affect anti-tumoral immune responses (Torricelli, Jardim, Guglielmetti, Patel, & Coelho, 2017).

As discussed above, TTP also plays a major role in the control of inflammation, often a feature of tumor development. In macrophages and dendritic cells, TTP controls production of key inflammatory cytokines such as TNF, IL-1β, IL-6, and IL-23 (Carballo, et al., 1998; Kratochvill, et al., 2011; Taylor, et al., 1996). TTP can also control tumor-associated inflammation and oncogenic pathways in epidermal cells. For example, in keratinocytes, TTP can regulate skin immune homeostasis by regulating TNF production (Andrianne, et al., 2017), and in human malignant keratinocytes, ZFP36 is down-regulated (Assabban, et al., 2021). In addition, keratinocyte-specific deletion of TTP (Zfp36^ΔEP), leads to the spontaneous development of a phenotype including dermatitis, arthritis, and failure to gain weight. This phenotype was ameliorated by simultaneous deletion of Tnf in keratinocytes as well (Zfp36^ΔEPTnf^ΔEP). Mice with keratinocyte-specific deletion of TTP (but not myeloid cell-specific (Zfp36^ΔM) or dendritic cell-specific (Zfp36^ΔDC) deletion of TTP) exhibited more severe skin inflammation after imiquimod treatment than WT mice treated with the same dose of imiquimod (Andrianne, et al., 2017). In addition, Zfp36^ΔEP KO mice, but not Zfp36^ΔM or Zfp36^ΔDC mice, exhibited extreme sensitivity to DMBA/TPA-induced tumor formation in mice (Assabban, et al., 2021).

In order to determine whether the increased TTP expression seen in the TTPΔARE mice could prevent early phases of tumor development, (Assabban, et al., 2021) used a classic two-step skin chemical carcinogenesis model (Abel, Angel, Kiguchi, & DiGiovanni, 2009; Cataisson, et al., 2012) in WT and TTPΔARE mice, and monitored the mice for papilloma development. Upon treatment with 7,12-dimethylbenz[a]anthracene (DMBA) followed by bi-weekly 12–0 tetradecanoylphorbol-13-acetate (TPA) application over the course of 10–18 weeks, all control mice developed multiple papillomas. In contrast, tumor burden was delayed and decreased in the TTPΔARE mice, with half of the animals remaining tumor-free after 20 weeks of treatment.

RNA-Seq analysis of treated adjacent non-tumoral whole skin samples from both TTPΔARE and WT mice showed that 1144 genes were statistically significantly differentially expressed more than two-fold (232 up-regulated and 912 down-regulated genes) in the TTPΔARE mice. Zfp36 mRNA expression was increased by about 3-fold in the skin of the TTPΔARE mice, whereas Zfp36l1 and Zfp36l2 mRNA were largely equivalent in both groups. There were strong enrichments for genes involved in innate immunity, myeloid cells, and inflammation-related pathways in the transcripts that were down-regulated in the TTPΔARE samples, suggesting that TTP overexpression restricts inflammation in these chronically stimulated skin samples. Some known TTP target transcripts, such as Cxcl2 and Il23a, and many myeloid cell transcripts (Cd163, Cd14, Fcer1g, Csfr1, Tlrs) were found to be decreased in the TTPΔARE samples, suggesting the possibility of decreased recruitment and possible activation of innate immune cells.

One principal conclusion of this study, in relation to this review, is that keratinocyte TTP deficiency in mice leads to greatly increased susceptibility to this model of chemical-induced skin carcinogenesis. Conversely, whole body regulated overexpression of TTP in the TTPΔARE mice leads to marked protection against papilloma development in this experimental model. Although the detailed mechanisms of these effects are likely to be complex, the data suggest that increasing TTP in one or more cell types might be a worthwhile therapeutic goal in inflammatory diseases of the skin, and possibly skin cancer.

3.9. Adrenalectomy-induced gastric metaplasia.

Chronic inflammation is strongly correlated with gastric cancer development, and gastric adenocarcinoma is the third leading cause of cancer deaths worldwide (Siegel, Naishadham, & Jemal, 2012). Although histopathological gut changes in TTP KO mice under normal conditions appear to be minimal, it has been suggested that TTP KO mice have increased intestinal inflammatory markers and changes in the microbiota (La, et al., 2021). Excessive expression of pro-inflammatory cytokines, such as IFNλ, TNF, and IL-1β, can induce spasmolytic polypeptide-expressing metaplasia (SPEM) and dysplasia development in experimental models (Osaki, et al., 2019; Oshima, Oshima, Matsunaga, & Taketo, 2005; Petersen, Mills, & Goldenring, 2017; Syu, et al., 2012; Tu, et al., 2008). A relatively new model of gastric inflammation and SPEM development occurs in response to removal of endogenous glucocorticoids by adrenalectomy (ADX) (Busada, Peterson, et al., 2021; Busada, et al., 2019). To determine whether TTP overexpression could protect the stomach from gastric inflammation and SPEM development, WT and TTPΔARE mice were subjected to the standard adrenalectomy model protocol, and their stomachs were analyzed for inflammation and SPEM development (Busada, Khadka, et al., 2021b).

In this study, both heterozygous and homozygous TTPΔARE mice were completely protected from ADX-induced gastric inflammation and SPEM; in contrast, the WT mice developed prominent inflammation in the gastric corpus 2 months after ADX. The overall gross morphology of heterozygous and homozygous TTPΔARE after ADX was indistinguishable from that of sham-operated mice. Characteristics of SPEM development in this model include loss of the mature chief cell marker MIST1 and expansion of GSII+ cells. TTPΔARE mice exhibited no changes in parietal cells or chief cells after ADX, whereas WT mice lost 82% of their parietal cells and 99% of mature chief cells, and also showed mucous cell hyperplasia within the gastric corpus, as demonstrated by increased GSII+ lectin staining. CD44v9 is a SPEM marker, and there was widespread staining of CD44v9 in ADX WT mice, that was not observed in the gastric glands of ADX TTPΔARE mice. WT mice 2 months after ADX showed significant proliferation of KI67+ cells in the neck and base; however, cell proliferation was unchanged in the TTPΔARE heterozygous and homozygous mice. Several SPEM markers (Cftr, Wfdc2, and Olfm4) were increased in ADX WT mice, but were not increased in the TTPΔARE mice. Previous studies have indicated that macrophages may be required to induce SPEM development (Busada, et al., 2019; Petersen, et al., 2014). WT mice 2 months after ADX exhibited 4.7-fold increases in gastric macrophages, and 28-fold increases in gastric eosinophils, whereas the TTPΔARE mice exhibited no such increases in inflammatory cells.

RNA-Seq of stomach tissues showed that there were significant increases in inflammatory gene expression 5 days after ADX in both WT and TTPΔARE mice, with slightly less enrichment in the inflammatory response pathway transcripts in the TTPΔARE mice, suggesting that there might be greater activation of inflammatory response pathways in WT mice after ADX. There were 760 differentially expressed genes between the sham WT and ADX WT mice, whereas there were 490 differentially expressed genes between the sham TTPΔARE and the ADX TTPΔARE mice, of which 189 were differentially regulated in both the WT and TTPΔARE. The transcripts up-regulated after ADX in WT mice included well-established TTP targets such as Tnf mRNA, as well as other inflammatory genes associated with SPEM such as Il13. Il13 mRNA up-regulation after ADX in the WT mice was decreased by the over-expression of TTP in the TTPΔARE mice, suggesting that TTP might suppress Il13, thus potentially disrupting macrophage activation.

Tnf has long been associated with inflammatory disease within the gastrointestinal tract and may increase the risk of gastric cancer (Delgado & Brunner, 2019). To determine whether Tnf overexpression after ADX was responsible for the gastric inflammation observed in the WT mice, Busada et. al. (Busada, Khadka, et al., 2021a) adrenalectomized Tnf KO mice. In this experiment, the Tnf KO mice were only partially protected from SPEM, with some regions of the gastric corpus appearing normal, and some regions appeared identical to the WT mice after ADX. These findings suggest that Tnf does contribute to SPEM development, but is not the only inflammatory factor involved. Therefore, it appears that overexpression of TTP can prevent SPEM development through a broader anti-inflammatory mechanism, rather than solely by promoting the decay of Tnf mRNA.

To determine whether TTP could prevent SPEM in another model that is not thought to involve inflammation, the authors (Busada, Khadka, et al., 2021a) treated the WT and TTPΔARE mice with high-dose tamoxifen, which induces SPEM by killing parietal cells through a largely non-inflammatory mechanism (Huh, et al., 2012; Saenz, Burclaff, & Mills, 2016). There were no differences in stomach pathology between the tamoxifen-treated WT and TTPΔARE mice, and loss of MIST1 staining was observed in both genotypes (Busada, Khadka, et al., 2021a). These data suggest that the TTP overexpression found in the TTPΔARE mice is likely to prevent SPEM development through a largely anti-inflammatory mechanism. Overall, this study suggests that treatments that increase TTP protein expression in one or more relevant cell types could be beneficial for the treatment of gastric inflammation and the prevention of gastric cancer.

4. TTP phosphorylation modifications in models of inflammatory diseases.

In addition to the transcriptional and post-transcriptional regulation of TTP expression discussed above, TTP expression and activity may also be controlled by the extensive post-translational phosphorylation of TTP on numerous serine residues. In 2015, Clark and colleagues (Ross, et al., 2015) described a new mouse model in which two specific phosphorylation sites (serines 52 and 178) of mouse TTP were replaced by non-phosphorylatable alanine residues, creating a mutant mouse allele that they called Zfp36^aa. The development of these mice followed previous studies, summarized in (Ross, et al., 2015), in which they demonstrated that these two residues were normally phosphorylated in response to p38 MAPK activation of the downstream kinase MAPK-activated protein kinase 2 (MK2). According to this model, these phosphorylations should prevent the recruitment of CCR4/CAF1, resulting in a decrease of TTP activity and hence the stabilization of TTP target mRNAs (Chrestensen, et al., 2004; Mahtani, et al., 2001; Marchese, et al., 2010; Ronkina, et al., 2019). As noted above, TTP phosphorylation was also thought to reduce binding of TTP to ARE-containing mRNAs (Carballo, et al., 2001; Hitti, et al., 2006).

In these mice, TTPaa appeared to be constitutively degraded by the proteasome, resulting in relatively low level expression of the modified TTP protein (Ross, et al., 2015). Nonetheless, the expression of several inflammatory mediators was decreased in LPS-treated macrophages from these Zfp36^aa/aa mice (Ross, et al., 2015), and the mice appeared to be protected from LPS-induced experimental endotoxemia. Specifically, the Zfp36^aa/aa mice were protected from LPS-induced organ damage, shown by histopathology, and serum markers of kidney damage (creatinine and blood urea nitrogen) and liver damage (transaminase). In addition, the Zfp36^aa/aa mice exhibited decreased levels of serum cytokines after LPS treatment, including TNF, CXCL1, CXCL2, and IL-10, in comparison to WT mice (Ross, et al., 2015).

This model was also used to study the role of TTP’s phosphorylation on its ability to regulate cytokines and chemokines in the lung, and protect against inflammatory lung damage. In previous studies, regulation of TTP activity by p38 MAPK resulted in a biphasic response of TNF-induced IL-6 expression in human bronchial smooth muscle cells (Prabhala, et al., 2015), whereas dephosphorylation of TTP resulted in reduced expression of IL-6 and IL-8 in A549 lung epithelial cells (Rahman, et al., 2016). Furthermore, the Zfp36^aa/aa mice were protected from cigarette smoke-induced chronic obstructive pulmonary disease (COPD) (Nair, et al., 2019). Specifically, after four days of cigarette smoke exposure, Zfp36^aa/aa mice exhibited decreased numbers of airway neutrophils and lymphocytes, and cytokine mRNA levels were decreased in comparison to WT controls (Nair, et al., 2019). After longer term cigarette smoke exposure, the Zfp36^aa/aa mice had improved lung function and reduced pulmonary inflammation, airway remodeling, and emphysema-like alveolar enlargement compared to WT controls (Nair, et al., 2019).

As an extension of this model, Nair et al. treated normal mice with the protein phosphatase 2A activator AAL_(S) and the proteasome inhibitor bortezomib in order to activate TTP before cigarette smoke exposure. Bortezomib and bortezomib plus AAL_(S) treatment significantly reduced the inflammation and numbers of neutrophils and lymphocytes in the lungs of cigarette smoke-exposed mice compared to vehicle-treated mice (Nair, et al., 2019). In addition, the expression of TTP target mRNAs, such as Il6, Tnf, Cxcl1, and Cxcl2 mRNAs, was significantly increased in the vehicle-treated mice compared to the bortezomib plus AAL_(S) treated, cigarette smoke-exposed mice (Nair, et al., 2019). Taken together, these data suggest that prevention of TTP phosphorylation by these methods can decrease the severity of LPS-induced endotoxemia and can suppress cigarette smoke-induced experimental COPD. In other words, increasing the activity of TTP by inhibiting its phosphorylation could be another valuable approach for the treatment of inflammatory diseases in general, and specifically COPD.

5. Conclusions: Future prospects for therapeutic interventions involving TTP.

Autoimmune diseases include a large range of debilitating conditions including rheumatoid arthritis, psoriasis and psoriatic arthritis, multiple sclerosis, Crohn’s disease, ankylosing spondylitis, dermatomyositis, type 1 diabetes, autoimmune uveitis, and many others. The pathogenesis of many of these diseases is due, at least in part, to abnormal secretion of pro-inflammatory mediators (Cho & Feldman, 2015; Wang, Wang, & Gershwin, 2015). These conditions often respond to nonsteroidal anti-inflammatory drugs, glucocorticoids, or disease modifying anti-rheumatic drugs, such as methotrexate, but, in severe cases, these drugs are not sufficient to eliminate disease activity and can still have significant side effects (Abbasi, et al., 2019; Lin, Anzaghe, & Schulke, 2020; Steinman, Merrill, McInnes, & Peakman, 2012).

Recently, biologic therapies, especially TNF inhibitors, have become mainstream treatments (Danese, Vuitton, & Peyrin-Biroulet, 2015; van Schouwenburg, Rispens, & Wolbink, 2013; Yao, et al., 2014). However, compared to many conventional types of treatment, biologic therapies are expensive, can be difficult to produce, often have reduced shelf-life, require injection or infusions, and are accompanied by increased risk of infection (Atzeni, et al., 2018; Patel & Khan, 2017; Rubbert-Roth, et al., 2019). Since TTP is a normal, physiological regulator of cytokine and chemokine biosynthesis and secretion, increasing TTP expression might be a way to address multiple inflammatory pathways simultaneously in order to reduce inflammation and disease pathogenesis. As illustrated in Fig. 3, TTP has anti-inflammatory functions in many cell types, which form a complex and interacting network in the intact organism. As described above, proof of principle experiments involving the TTPΔARE mice demonstrated that regulated, overexpression of TTP from its endogenous genetic locus could protect mice from several models of inflammatory disease of different types, including skin pappillomatosis, with no apparent adverse effects on the mice. The data accumulated so far suggest that increasing TTP concentrations in essentially all the cells of a human patient could have beneficial effects in the treatment of human inflammatory diseases. However, the practical question is how to accomplish this in humans without undue toxicity.

Figure 3.

Schematic representation of the complex network of intercellular interactions mediated by TTP. TTP is expressed in many cell types and controls the stability of specific mRNAs in many cells, many of which encode cytokines and chemokines that regulate interactions with other cells through a network of interactions. Increasing levels of TTP coordinately throughout the body appears to have an anti-inflammatory effect in many cell types, mainly by decreasing the biosynthesis and secretiton of these chemokines and cytokines, and prevents or inhibits the pathogenic processes in many different inflammatory disease models in mice. Therapeutic approaches that increase TTP expression throughout the body, or in specific cell types, could be useful in the treatment of human inflammatory diseases.

One approach would be to use the conventional strategy of identifying small molecules using cell-based screening programs to find compounds that affect TTP levels in cells and tissues by any of the known modes of regulation, including increasing gene transcription, increasing mRNA stability, increasing protein stability, or a combination of these mechanisms. Cell-based screening programs of this type are currently underway. It might also be possible, through the use of siRNA or CRISPR-based screens, to identify genetic modifiers of TTP expression, some of which might themselves represent drug targets. These types of screens are also underway, and may yield interesting information about the pathways that regulate TTP expression, in addition to the practical possibilities of identifying novel drug targets.

A different approach to increasing TTP levels by increasing its mRNA stability would be to use an antisense RNA approach to block ARE instability motifs from access by decay-promoting proteins. TTP or ZFP36 mRNA might be particularly amenable to such an approach, given its very rapid turnover in cells. With recent advances in RNA delivery and RNA drug stabilization, RNA-based drugs have been successfully used as vaccines, and antisense oligonucleotides have been used as therapeutics for several diseases, e.g., muscular dystrophy (Bogdanik, et al., 2015; Kaczmarek, Kowalski, & Anderson, 2017; Liang, et al., 2016). Steric blocking antisense oligonucleotides could potentially be designed with sequence specificity to target the ARE regions of TTP mRNA to block and stabilize the transcript. Such studies are also ongoing.

Other possibilities for altering TTP activity include identifying small molecules that could target co-activator proteins that have been identified that increase the affinity of TTP for its mRNA targets, such as AUF1 (Kedar, Zucconi, Wilson, & Blackshear, 2012). In addition, as described above, altering the phosphorylation state of key serine residues within TTP, with or without proteosome inhibitors, might be an effective approach to taking advantage of its anti-inflammatory properties in the treatment of human diseases (Nair, et al., 2019; Ross, et al., 2015).

It is also possible that cell- or lineage-specific overexpression of TTP could have beneficial therapeutic effects. In early studies, bone marrow transplantation from Zfp36 KO mice into immune-deficient mice resulted in recapitulation of the original Zfp36 KO phenotype after a lag of several months (Carballo, Gilkeson, & Blackshear, 1997). Therefore, it might be possible to alter TTP mRNA stability in hematopoietic cells alone, possibly through CRISPR-CAS technology or other gene editing approaches, which are already being used to modify human CD34+ hematopoietic stem cells (Hendel, et al., 2015; Mandal, et al., 2014). As described above in the context of an acute lung injury model (Choudhary, et al., 2020), transplantation of hematopoietic progenitors from TTPΔARE mice into irradiated recipients was not sufficient to completely prevent the development of disease in this particular model, but it remains possible that TTP overexpression of this type in autologous hematopoietic stem cell transplantation might be a worthwhile approach in some conditions.

In conclusion, TTP loss of function studies in mice showed that TTP deficiency results in a serious, systemic inflammatory syndrome that appears to be largely due to hyper-secretion of TNF and other cytokines and chemokines. In contrast, TTP gain of function studies demonstrated that increased expression or activity of TTP, either by increasing TTP mRNA stability in the TTPΔARE mice, or by increasing TTP activation by preventing TTP phosphorylation in the Zfp36^aa/aa mice, can result in decreased susceptibility to inflammatory diseases in several mouse models. These studies demonstrate that TTP plays a vital role in regulating cytokine and chemokine homeostasis, and support the concept that modulating TTP expression and/or activity could serve as a therapeutic intervention in many types of human disease that are accompanied by inflammation.

Abbreviations:

TTP

Tristetraprolin

ZFP36

zinc finger 36

RBP

RNA-binding protein

TNF

tumor necrosis factor

BMDM

bone marrow-derived macrophages

AU

adenylate-uridylate

3’-UTR

3’-untranslated regions

ARE

AU-rich element

MAP

mitogen activated protein

TZF

tandem zinc finger

NOT1 BD

NOT1 binding domain

WT

wild-type

GM-CSF

granulocyte macrophage colony-stimulating factor

LPS

lipopolysaccharide

ActD

actinomycin D

CAIA

collagen antibody-induced arthritis

IMQ

imiquimod

H and E

hematoxylin and eosin

EAE

experimental autoimmune encephalitis

MOG

myelin oligodendrocyte glycoprotein

TRAP

tartrate-resistant acid phosphate

EAU

experimental autoimmune uveitis

IRBP

interphotoreceptor retinoid-binding protein

Treg

T regulatory

BALF

bronchoalveolar lavage fluid

HPCs

hematopoietic progenitor cells

DMBA

7,12-dimethylbenz[a]anthracene

TPA

12–0 tetradecanoylphorbol-13-acetate

SPEM

spasmolytic polypeptide-expressing metaplasia

ADX

adrenalectomy

MAPKAPK2

MAPK-activated protein kinase 2

COPD

chronic obstructive pulmonary disease