RNA-binding protein HuR is an emerging target for cancer therapeutics and immune-related disorders

Abstract

Eukaryotic gene expression is regulated at multiple levels, from transcription to messenger RNA (mRNA) processing, transport, localization, turnover, and translation. RNA-binding proteins (RBPs) are critical regulators of gene expression and are involved in all aspects of mRNA processing, from splicing to translation. The RNS-binding protein Human antigen R (HuR) mainly regulates the mRNA transport from nucleus-to-cytoplasm, stability, and translation. Dysregulation of HuR is linked to diseases such as cancer, neurodegenerative disorders, and immune-related disorders. HuR targets mRNAs containing AU-rich elements at their 3’untranslated region, which encodes proteins involved in cell growth, tumorigenesis, angiogenesis, immune evasion, tumor inflammation, invasion, and metastasis. HuR overexpression has been reported in many tumor types, which led to a poor prognosis for patients. Hence, HuR is an attractive drug target for cancer therapy. Recently, there has been an intense effort to identify small molecule inhibitors for blocking HuR functions. This article reviews the current prospects of drugs that target HuR in multiple cancer types. Furthermore, we will summarize drugs that interfered with HuR-RNA interactions and established themselves as novel therapeutics. We will also highlight the significance of HuR overexpression in multiple cancers and discuss its role in immune functions. This review provides evidence of a new era of HuR-targeted small molecules that can be used for cancer therapeutics either as a monotherapy or combination with other cancer treatment modalities.

Graphical Abstract:

Targeting HuR for anticancer activity. Left panel: Overexpressed HuR is phosphorylated by kinases that bind with ARE mRNAs and promote their protein expression to facilitate tumor growth and proliferation. Right panel: Inhibition of HuR phosphorylation by small-molecule BML-277 or disruption of HuR binding with ARE mRNA using CMLD2 reduced ARE mRNA encoding protein and promoted cell death.

1. Introduction

The significance of RNA-binding protein (RBP) mediated post-transcriptional gene regulation (PTR) in eukaryotes has emerged as a necessary regulatory mechanism in gene expression. This level of PTR is mainly controlled by RBPs, which play a significant role in all aspects of RNA maturation and the regulation of messenger RNA (mRNA) stability, localization, and translation (well-reviewed in ref. [1]). Recent emerging high-throughput genomics and mass spectrometry-based methods have identified hundreds of RBPs in mammalian cells [2–5], suggesting that the human genome may contain 1,542 or more RBP-encoding genes [6]. Due to vast differences in protein motifs, RBPs recognize different RNA sequences and are functionally distinct from each other. Thus, RBPs play critical roles in controlling the fate of the RNA molecules with which they associate.

Despite ongoing efforts in identifying RBPs and their RNA interactome, many questions remain unanswered. For example, there is a lack of understanding of how RBP-RNA binding specificities are defined and how this RBP-RNA interface can be used for drug targeting by chemical and biophysical methodologies. Recently, several attempts have been made to identify small molecule chemical probes that target RBPs and interfere with RNA binding, representing valuable tools for developing novel therapeutics to block the function of RBPs [7]. Although RBPs have been considered impossible molecules for drug-targeting due to the complexity of protein structure and domains, efforts have been made to discover small molecules that target specific RBPs have prompted considerable interest. Small molecules are often designed to target RBP’s RNA-binding domains, such as the RNA-recognition motif (RRM), hnRNP K homology (KH), and the zinc-finger domain. Other examples include small molecules that target RBPs such as spliceosomal proteins for treating muscular dystrophy [8, 9], cancer activity [10], EIF4 family proteins [11], TDP-43 for the treatment of Amyotrophic Lateral Sclerosis (ALS) [12], and Human antigen R (HuR) for anti-cancer activities [13–15]. We direct the readers to the following reviews to further read RBP probes and their medicinal uses [7, 16]. Here, we review the RBP HuR’s role in cancer and immune functions, detail its RNA-binding motifs, analyze its expression across different cancers, and highlight drug-targeting opportunities that will enable us to develop HuR-dependent therapeutics for future medicine.

Chemical and biophysical approaches such as fluorescence polarization (FP) and fluorescence resonance energy transfer (FRET) assays are the principal methods used to determine RBP interaction with small molecules for drug screening. FP is a well-known technique to study molecular interactions, particularly those between proteins and nucleic acids. FP utilizes the polarized light emission from a 3’ or 5’ or bodily labeled RNA and monitors its binding to RBPs. The ability of FP to screen drugs relies solely on the changes in the polarization of corresponding unbound RNA, which is indicative of the small molecule blockade of RNA-RBP interaction [7]. FRET is a photophysical event in which energy is transferred between two appropriate fluorophores called donor and acceptor, appropriately oriented and distanced to study their interactions. The donor and acceptor are individually attached to RNA and RBP, respectively, and the interference of small molecule binding will decrease the emission of the acceptor fluorophore [7]. For example, to determine the spliceosome complex, FRET was employed to measure dynamic protein-protein interactions between splicing factors [17–19]. In addition to FP and FRET, another proximity assay, Alphascreen (Amplified Luminescent Proximity Homogenous Assay), has been used to detect RBP inhibitors. Recently, Alphascreen was used to identify HuR inhibitor CMLD-2 by disturbing the interaction between HuR and AU-rich element (ARE) mRNA [20].

This review will identify FDA-approved drugs that target HuR or other pending therapies for cancer. We will also summarize examples of potential RNA-targeted therapies that inhibit HuR and RNA interactions and act as antitumor drugs. We will review the posttranslational modification of HuR carried out by signaling kinases, which are targets of small molecule inhibitors that interfere with HuR functions. Finally, we will highlight small animal models of HuR and their contribution to immune dysfunctions, HuR’s cancer traits in multiple cancers, and provide insight into the mRNA-HuR targeted therapies.

2. Biology of HuR

2.1. ELAV family of proteins:

HuR belongs to the ELAV (embryonic lethal abnormal visual system) family of RNA binding proteins, present in prokaryotes and eukaryotes [21]. In Drosophila, the ELAV protein was the first discovered member of the family, followed by its homologs [22]. The ELAV family of proteins contains three RRMs, which have characteristic ribonucleoprotein consensus sequences [23]. Among 1,542 human RBPs, approximately 692 proteins are mRNA binding proteins in which 225 proteins contain canonical RRM [6], indicating that RRM is the primary domain among all the other RBP motifs. Furthermore, a search of RRM in the Protein Data Bank (PDB) yields over 500 structures and is estimated to be present in 1% of all the human proteins [24].

HuR has canonical protein motifs to bind RNA. Protein ELAVL1, alias HuR, is composed of three highly conserved canonical RRMs. All three RRMs adopt a βαββαβ topology, where a four-stranded antiparallel β-sheet is packed against two α-helices [25]. The tandem RRM1 and 2 are separated from the C-terminal RRM (RRM3) by a ~50-residue basic hinge region containing the nucleocytoplasmic export signal [26]. Both RRM 1 and 2 are the significant domains that bind ARE-containing RNAs and stabilize them [27], but RRM3 recognizes AREs and is involved in dimerization and competes with both RRM 1 and 2 for mRNA degradation [28]. Further studies demonstrated that RRM3, with hinge region of HuR, is involved in protein-protein interactions [29], protein dimerization [30], and multimerization in cancer cells [31]. Both RRM3 and hinge regions counter miRNA-mediated repression and promote miRNA-induced silencing complex release from target mRNAs [32]. Multiple studies reported that structural details of HuR and its family members are essential for RNA recognition [33–35]. Based on these structural features of HuR, several investigators attempted to develop small-molecule ligands that modulate HuR functions and block RNA binding activity (reviewed in [36, 37]). In this review, we will update and summarize the significant role of HuR in cancer and immune regulation in conjunction with the potential impact of targeting HuR for cancer therapeutics. As our understanding of HuR protein-ligand interactions grows, it is becoming evident that structure-based computer-aided drug discovery (molecular docking) using bioinformatic approaches will further develop multiple therapeutic opportunities.

2.2. Posttranslational modifications (PTM) of HuR:

Many excellent reviews have covered the different posttranslational modifications (PTM) of HuR and their modes of RNA binding and biological functions [38–40]. Therefore, we will limit the depth of our discussion for each PTM and focus our attention on drug targeting of upstream protein kinase modifiers. Figure-1 illustrates the protein structure and posttranslational modification of HuR by signaling kinases, which can serve as small molecule drug targeting enzymes to block HuR localization and functions.

Figure 1:

Posttranslational modification of HuR by signaling kinases. Schematics illustrates the protein domains and amino acids of HuR modified at the posttranslational level by signaling kinases targeted using specified small molecules. References citing the relevant studies are listed.

2.2.1. HuR phosphorylation by kinases provides an opportunity to indirectly target HuR:

HuR is phosphorylated by several kinases, which modulate its RNA binding and turnover activities. The checkpoint kinase 2 (Chk2), activated and rapidly phosphorylated in response to DNA damage, phosphorylates HuR at S88, S100, and threonine (T)118 [41]. This phosphorylation affects HuR binding with SIRT1 mRNA and triggers its mRNA decay. In a subsequent study, the authors showed that ionizing radiation-induced CHK2 phosphorylates HuR, consequently promoting dissociation of HuR from its target mRNAs, including mRNAs that encoded proapoptotic and proliferative proteins [42]. These findings provide an opportunity to target Chk2 for the control of HuR phosphorylation at the pharmacological level. Accordingly, a small-molecule inhibitor BML-277 (Inhibitor II) of Chk2 dephosphorylate HuR showed decreased mRNA association in cancer cells [43].

The kinase Cyclin-dependent kinase-1 (CDK1) dependent phosphorylation of HuR at S202 facilitates its interaction with protein 14–3-3 for nuclear retention [44]. However, when CDK1 is inactive and amino acid S202 is dephosphorylated, HuR translocates from the nucleus to the cytoplasm. Interestingly, inhibition of CDK1 by pharmacologic agent (Cdk1-specific inhibitor, CGP74514A) promotes HuR translocation to the cytoplasm and stabilizes CDKN1A mRNA, a tumor suppressor gene [45]. Thus, pharmacological targeting of kinases that phosphorylate HuR is an alternative strategy to target HuR in cancer cells at the therapeutic level.

The protein kinase C (PKC) family members phosphorylate HuR and promote its cytoplasmic export and control of mRNA stability [46–48] of encoding proteins involved in various cancer pathways. Among many PKC isoforms, PKCα phosphorylates S158 and S221 residues of HuR to promote its cytoplasmic localization to enhance its binding to COX2 mRNA leading to increased mRNA stability, COX2 protein, and prostaglandin E2 biosynthesis [48]. However, the authors demonstrated that inhibition of PKCα by small molecules such as gö6976, CGP 41251, or rottlerin prevents cytoplasmic HuR in an ATP-dependent manner, suggesting that targeting PKCα may indirectly control HuR-mediated gene expression [48]. Consequently, the same group demonstrated that HuR phosphorylation at serines 221 and 318 confers binding to COX2 mRNA, which PKCδ differentially regulates in human mesangial cells [47]. Since both PKC isoforms phosphorylate HuR at different sites, it is most likely that those serine residues induce oligomerization of HuR and regulate its RNA-binding capacity [49]. However, pharmacological inhibition of PKCδ by rottlerin abolished the S318 phosphorylation of HuR, and its binding to target mRNAs in colon cancer cells demonstrated that targeting PKC isoforms may alter HuR functions in multiple cancer cells [50, 51]. The p38MAPK also participates in HuR phosphorylation at Thr-118 and promotes its translocation to the cytoplasm to stabilize p21 mRNA [52]. Further studies also supported that p38 MAPK-mediated Thr-118 phosphorylation of HuR plays a crucial role in regulating the COX-2 expression during inflammation [53]. Concurrently, blocking p38MAPK by small-molecule with SB203580 abolished the effect of HuR phosphorylation and target COX-2 expression [53]. HuR phosphorylation by other kinases and their coordinated regulation of HuR target mRNAs are well-reviewed (please see ref. [38]).

2.2.2. Proteolytic cleavage of HuR:

We [54–56] and others [57–60] have shown that apoptotic stress stimuli promote HuR translocation to the cytoplasm and proteolytic cleavage of HuR at the aspartate D226 by caspases 3 and 7, subsequently generating two active HuR cleavage products, HuR-CP1 and HuR-CP2. Furthermore, our group showed that HuR-CP1 alone facilitates cell death, whereas the non-cleavable isoform of HuR enhances cancer cell proliferation [56]. We further showed that specific inhibition of caspase-3 by a small molecule compound, NSC321205, blocks proteolytic degradation of HuR and promotes the oral epithelial membrane integrity after ionizing radiation treatment [55]. These findings demonstrated that the use of caspase inhibitor NSC321205 indirectly supports HuR functions by establishing HuR’s role in epithelial homeostasis. However, more specific epithelial-specific HuR gain- or loss-of-function studies are required to fully understand the biology of HuR and its cleavage products in epithelial homeostasis.

2.2.3. Ubiquitinylation (protein degradation) of HuR:

HuR undergoes Lys-182 residue-dependent proteolytic degradation by proteosomes during heat shock [61]. However, this proteolytic degradation is opposed by phosphorylation of HuR by chk2 [61]. Further studies demonstrated that the tumor suppressor esophageal cancer-related gene 2 (ECRG2) was shown to increase in response to DNA damage, in turn favoring ubiquitination of HuR at K182 and HuR degradation [62]. HuR is also noted for the enzyme ubiquitin E3 ligase βTrCP1 mediated degradation by glycolytic stress in prostate cancer cells [51]. Hence, an agonist approach to activating ECRG2 or βTrCP1 may promote HuR ubiquitinylation, which may provide a therapeutic window of opportunity to target HuR in cancer cells. Interestingly, studies demonstrated that ubiquitination of HuR with a short K29 chain attached to Lys-313 and Lys-326 did not affect protein stability but promoted the disassociation of HuR from its target mRNAs [63]. Thus, altering ubiquitinylation and de-ubiquitinylation may regulate HuR’s expression and RNA-binding activity in cancer cells.

2.2.4. Methylation:

The protein coactivator-associated arginine methyltransferase 1 (CARM1/PRMT4) methylates HuR at the arginine residue R217 in macrophages [64], human cervical carcinoma cells [65], and human embryonic stem cells (hESCs) [66] and led HuR to promote mRNA stabilization of its targets. As CARM1 and other protein methyltransferase inhibitors are used to treat a variety of cancers [67, 68], inhibition of CARM1 with small molecules may have a significant impact on HuR functions.

3. HuR in cancers:

While increased expression of HuR is reported in many cancers and provides a poor prognosis, the role and function of HuR in cancers have been well-reviewed by many publications [69–72]. Hence, we limit the detailed review of the HuR’s role in cancers. For the past 20 years, more than a hundred papers have detailed the role of HuR in cancers using in vitro cell culture methods with fewer in vivo studies. We have summarized the essential cancer types in which HuR and its target mRNAs encoding proteins play a critical role in carcinogenesis (Table 1).

Table 1:

A representative list of HuR associated cancer types and their role on prognosis.

Cancer Type	Association between High HuR Expression and Clinicopathological Features	Pathology and Prognosis	References
Breast (DCIS)	↑ tumor grade; ↑ aggressiveness; More frequently observed in PR(-ve) cases	Elevated HuR expression observed in atypical ductal hyperplasia (ADH) and ductal in situ carcinomas (DCIS) compared to normal controls, associated with high grade, progesterone receptor negativity, and microinvasion and/or tumor-positive sentinel nodes of the DCIS.	[137]
Breast (DIC)	↑ histological grade; ↑ ductal tumor grade; More frequently observed in PR(-ve), ER(-ve) cases; ↑ p53; ↓ OS	In non-BRCA1/2 breast cancer patients, cytoplasmic HuR protein expression is present in 39.4% of the BRCA1/2 negative patients and associated with high tumor grade and ductal type of the tumor. However, in patients with BRCA1 or BRCA2 mutations, cytoplasmic HuR expression is more frequent (62.7% for BRCA1 and 61.7% for BRCA2) and associated with poor survival.	[138]
Esophageal squamous	46.6% of specimens had positive cytoplasmic staining for HuR.	Cytoplasmic HuR expression was positively associated with lymph node metastasis, depth of tumor invasion, advanced tumor stage, and exhibits a low 5-year survival rate. Patients positive for cytoplasmic HuR expression had a cumulative 5-year survival rate of 25.3 %, whereas it was 43.8 % for patients negative for cytoplasmic HuR expression.	[139]
Gallbladder carcinoma	↑ advanced tumor stage, histological grade, vascular and perineurial invasion, tumor necrosis, Ki-67 labeling index; ↑ cyclin A expression; ↓ DSS & DFS	Nuclear HuR expression correlated with poor disease-free survival, but cytoplasmic HuR expression was strongly associated with inferior disease-specific survival and disease-free survival.	[140]
Gastointestinal stromal tumors (GISTs)	Epithelioid histology, larger tumor size, NIH risk category and nuclear expression of Ki67 and cyclin A. Cytoplasmic HuR and cyclin A overexpression were strongly associated with worse DFS.	Both HuR cytoplasmic expression (P < 0.001) and cyclin A overexpression (P < 0.001) are strongly associated with worse disease-free survival in GISTs patients.	[141]
Lung carcinoma	↑ lymph-node involvement at presentation; ↑ risk of death and metastasis	High cytoplasmic HuR expression is directly associated with the risk of death and metastasis. The levels of nuclear and cytoplasmic HuR were allowed the authors to discriminate between patients with the highest risk of metastasis and death with lung adenocarcinomas.	[142]
Meningioma	↑ tumor grade; ↓ PFS & OS; HuR knockdown decreases meningioma cell growth and hypoxia-mediated resistance	High HuR cytoplasmic expression correlated with tumor grade and negatively with progression-free and overall survival of meningioma patients.	[143]
Oral squamous cell carcinoma	↑ IAP2; ↓ OS	Positive cytoplasmic HuR expression was significantly associated with positive cellular inhibitors of apoptosis protein-2 (cIAP2) in high-grade tumors, which highly correlated with poor patient survival.	[144]
Oral squamous cell carcinoma	↑ Lymph node metastasis	The expression of cytoplasmic HuR is correlated positively with IMP3 and negatively with p53, significantly associated with the risk of death and poor patient survival.	[145]
Ovarian cancer (metastatic highgrade serous carcinoma)	↑ HuR mRNA and protein level	Higher HuR protein expression was associated with higher serum Cancer Antigen 125 levels at diagnosis. HuR protein as detected in the nucleus and cytoplasm of tumor cells in 98% and 58% respectively, of the effusion samples. Higher HuR mRNA levels were significantly associated with poor OS.	[146]
Pancreatic cancer	In pre-treatment tumor samples, Deoxycytidine kinase (dCK) and HuR cytoplasmic expression were strongly correlated.	The dCK levels were prognostic and had predictive value for sensitivity to 5-FU. However, cytoplasmic HuR expression was not associated with either prognostic or predictive to 5-FU treatment.	[147]
Urinary tract urothelial carcinoma	Cytoplasmic HuR and nuclear cyclin A expressions were correlated with disease-specific survival (DSS), metastasis-free survival (MeFS), urinary bladder recurrence-free survival (UBRFS), and treatment response.	Increased Tumor grade, lymph node metastasis, histological grade, vascular and perineurial invasion and cyclin A expression; decreased DSS, MeFS, UBRFS; High-risk patients (pT3 or pT4 with/without nodal metastasis) with high cytoplasmic HuR had better DSS if given adjuvant chemotherapy.	[148]

4. HuR as an immune regulator:

Given the critical roles played by HuR in tumor progression in a variety of cancers, understanding its role in the tumor microenvironment (TME) and the immune response is essential to investigate the potential for HuR as a target for cancer therapeutics. HuR is the most often studied RBP in the immune system. Conditional deletion of HuR impairs the B-cell population in the bone marrow and the periphery leading to reduced serum immunoglobulins (IgG) [73, 74]. HuR is also critical for T-cell development, and mice lacking HuR develop egress from the thymus, which results in lymphopenia in the periphery [75]. HuR expression and functions are cell-specific [76], as it controls both pro-inflammatory and anti-inflammatory cytokines. For example, myeloid cell-specific deletion of HuR resulted in an inflammatory phenotype with an enhanced inflammatory response in chemically induced colitis [77]. Similarly, heterozygous HuR mice showed steadily decreased mRNA levels of Gata3, Il4, and Il13 without affecting protein expression; however, HuR KO homozygous mice showed increased expression of Il2, Il4, and Il13 mRNA and protein [78]. The distinct differences observed between these studies indicate that intrinsic differences in immune cell types determine HuR immune cell functions.

Another example of HuR expression is correlated with increased Tnf mRNA stability and TNF synthesis in macrophages [79, 80]. However, in B-cells, HuR controls the oxidative response induced upon mitogen activation, which allows the cells for the metabolic switch and cell growth [74]. When HuR is absent in B-lymphocyte development and differentiation (HuR^fl/fl Mb1Cre mice), the germinal center (GC) reaction is not formed, and affinity maturation is severely compromised [73, 74]. HuR, in association with other RBPs like PCBP1, promotes the production of pro-inflammatory cytokines in T cells and contributes to the development of T-cell mediated inflammation [81, 82]. Thus, these collective findings indicate that HuR is critical for immunosurveillance and promotes a hijack of an immune response in cancer cells.

5. Targeting HuR through RNA interference:

Although HuR is overexpressed in various cancers, miRNAs can control its expression at post-transcriptional gene regulation. The miRNAs miR-519, miR-125a, miR-16, miR-34a suppress HuR levels either by RNA stability or translational repression [83–86]. Hence, a miRNA-mediated manipulation of HuR is feasible using an antisense oligonucleotide targeted approach. We have observed that smoking-induced downregulation of miR-133a directly controls the expression of its targets HuR and EGFR in human papillomavirus-infected patient samples and oral pharyngeal carcinoma cells [87]. In addition to miRNAs, HuR silencing is carried out by delivering siRNAs targeting HuR using a nanoparticle formulation. For example, a folate or transferrin receptor-targeted nanoparticle formulated with siRNAs against HuR reduced lung cancer cell growth, migration, and invasion [88, 89]. Another study demonstrated that siRNA conjugated with cholesterol was loaded onto exosomal vesicles for the delivery and silencing of HuR [90]. HuR is silenced in the retina of diabetic retinopathy rats, in which cationic liposomes loaded with siRNAs target HuR and result in downregulation of VEGF and retinal thickness [91]. In another study, siRNAs targeting HuR were conjugated with Cy3-labeled DNA dendrimer nanocarrier delivered into ovarian cancer xenografts and showed antitumor activity [92]. Additional findings have demonstrated that siRNAs against HuR inhibit the expression of HuR and its targets, including Bcl-2, to repress the survival of colorectal cancer cells [93]. In other settings, an antisense oligonucleotide (ASO) knockdown of HuR reduces neuropathic pain through changes in target gene expression to inhibit microglia-mediated spinal neuroinflammation and promote an anti-inflammatory and neuroprotectant response [94]. Further, multiple sclerosis (MS) model ASO silencing of HuR exhibited improved experimental autoimmune encephalomyelitis in mice-related motor dysfunction, pain hypersensitivity, and body weight loss. Hence, targeting HuR using ASO is a promising approach to improve neurological disturbances in MS patients [95]. Together, these findings demonstrated that the possibility of siRNA or ASO therapeutics has the favorable effect of targeting HuR in multiple cell systems.

6. Small molecules targeting HuR:

The first report to develop inhibitors for HuR used a confocal fluctuation spectroscopy assay with recombinant HuR (both RRM1 and RRM2), and a tetramethylrhodamine (TMR) labeled ARE-RNA using a 50,000-compound library from Novartis [15]. In a competitive binding assay between HuR and ARE-RNA complex, the investigators identified MS-444, okicenone, and dehydromutactin as small molecules that block HuR binding with ARE-RNA. Out of the three compounds, MS-444 has been widely used as an inhibitor of HuR by many laboratories for anti-cancer activity [96–98], including melanoma [99], colorectal [100], pancreatic carcinoma [101], and glioma [102] by in vitro and other xenograft studies. However, the failure of in vivo tumor models to establish changes in HuR-mediated TME studies minimizes the promise of MS-444 as a potent HuR inhibitor. It is uncertain whether Novartis will continue the studies related to MS-444. Using a confocal nanoscanning screening method, a selective inhibitor compound, H1N, was identified against HuR RRM3. The compound H1N belongs to the dicarboxylic acid scaffold sub-library, which provides an opportunity to identify an ATP-binding pocket associated with 3′-terminal adenosyltransferase activity in HuR RRM3 [103].

Two independent studies revealed that the bioactive flavonoid quercetin disrupts the HuR binding with TNFα and IL-6 mRNA using electrophoretic mobility shift assays (EMSA), suggesting that both mRNAs are destabilized in the presence of quercetin [104, 105]. Further screening by PerkinElmer demonstrated that cetylpyridinium chloride and mitoxantrone prevented the formation of HuR-TNFα ARE complex [106]. In addition, mitoxantrone disrupts the interaction between HuR and SOX2 mRNA, which results in translational activation of SOX2 in mesenchymal stem cells [107]. Further studies from the same screening approach established the inhibitor DHTS (15, 16-dihydrotanshinone-I), which disrupts the binding between HuR and TNFα mRNA in the nanomolar range. Furthermore, treatment with DHTS exhibited anti-cancer activity in breast cancer cells [108], reduced colon cancer cell growth and proliferation [109], and cytotoxicity towards glioma cells [31]. Based on the structure of DHTS, treatment of its derivatives tanshinone mimic 6a and 6n reduced the cell growth of breast and pancreatic cancer cells. However, they are less potent when compared to DHTS [110]. In another study, using an actin-depolymerizing macrolide latrunculin A, or blebbistatin, an inhibitor of myosin II ATPase activity, the authors showed reduction in the cytoplasmic HuR content of HepG2 and Huh7 hepatocellular carcinoma (HCC) cells, which concomitantly changes the intracellular HuR localization, and markedly decreased the half-lives of COX-2, cyclin A and cyclin D1 resulting in a significant reduction in their expression levels in HepG2 cells [111].

Wu et al. used an FP assay optimized for high throughput screening (HTS) to screen 6000 compounds for HuR/ARE-mRNA disruptors [112]. The HTS analysis yielded six compounds of coumarin derivatives, CMLD1–6, in which CMLD-2 was the most potent HuR/ARE disruptor identified. CMLD-2 showed promising antitumor activity in lung, breast, colon, and thyroid cancers [20, 113, 114], in addition to triggering apoptosis and autophagy through deregulation of HuR targets Bcl-2 and XIAP [112]. Further efforts using FP-based assays identified small molecule azaphilones, derivatives of the fungal natural product asperbenzaldehyde, which disrupt the binding between HuR and c-Fos ARE mRNA [115]. An FDA-approved drug, suramin, an anti-trypanosomal drug, competitively binds with HuR and exerts antitumor activity in oral cancer cells [116]. Trichostatin A (TSA) and 5-Aza 2’deoxycytidine (AZA), two well-characterized pharmacologic inhibitors of histone deacetylation and DNA methylation, affect HuR translocation in ER-negative breast cancer cell lines, modulate HuR dependent ER mRNA [117] to reduce tumor burden [117].

Table 2 illustrates the list of HuR inhibitors demonstrating the antitumor activity, and promising, potent drugs are listed in the order of their discovery. Figure 2 shows small molecules that are either interrupting HuR binding with RNA or inhibiting HuR activity.

Table 2:

Small molecules targeting HuR show anticancer activity.

Small-molecules	Drug action on HuR	Method of identification	References
Mitoxantrone and 15,16- dihydritanshinone-1	Blocks HuR association with target mRNAs	Amplified luminescent proximity homogeneous assay (AL-PHA Screen)	[123]
Tanshinone and derivatives	Blocks HuR association with target mRNAs	ALPHA Screen	[127]
MS-444	HuR translocation	In vitro/ in vivo	[13, 91]
5-aza 2” deoxycytidine/ trichostatin A	HuR translocation	In vitro/ in vivo	[134]
Blebbistatin	HuR trafficking	In vitro/ in vivo	[128]
CMLD2	HuR target mRNA binding inhibitor	In vitro only	[105]
Tanshinone II	HuR target mRNA binding inhibitor	In vitro only	[125]
Suramin	HuR target mRNA binding inhibitor	In vitro/ in vivo	[133]
NSC# 651084	HuR target mRNA binding inhibitor	In vitro	[12]

Figure 2:

Small molecule inhibitors of HuR. Small molecules interrupting RNA binding or protein localization of HuR are displayed.

7. Small animal models of HuR provide a therapeutic window of opportunity to treat human diseases:

Recent efforts from multiple groups showing a conditional knockout of HuR have a significant impact on cell survival. Timothy Hla’s group established the first reported HuR flox mice [118] and showed deletion of HuR induced atrophy of hematopoietic organs, extensive loss of intestinal villi, obstructive enterocolitis, and lethality within ten days. The same HuR flox mice (Jax mice stock number # 021431) model was used to generate multiple tissue-specific HuR silencing, such as adipose and vascular smooth muscle [119, 120], intestinal epithelium [121], skeletal muscle [122], hepatic tissue [123], cardiac-specific [124], and mouse neurons [125]. For example, fat-specific knockout of HuR greatly enhances adipogenic gene program in fatty tissues, accompanied by a systemic glucose intolerance and insulin resistance [126]. In another study, neuron-specific HuR-deficient mice developed a phenotype consisting of poor balance, decreased movement, and strength [127]. These findings demonstrated that deletion of HuR will have a complex differential post-transcriptional regulation and biological functions in a tissue-specific manner. A complete gene knockout differs primarily from the inhibition of protein activity because the phenotypes obtained from gene knockouts are likely to be different from those obtained after inhibitor treatment. It is evident that genetic silencing of HuR in murine is embryonically lethal and disturbs cellular development and normal homeostasis [118]. However, these effects were not observed when mice were treated with a HuR inhibitor MS-444 [100]. Therefore, drug targeting of HuR is more efficient than genetic silencing in cancer cells.

As HuR belongs to the ELAV family of RBP, its complete knockout is embryonically lethal. Hence, tissue-specific HuR deletion is critical to determine its functions. For example, heterozygous HuR conditional (OX40-Cre HuR^fl/+) KO in T cells had decreased steady-state levels of Gata3, Il4, and Il13 mRNAs with minor changes at the protein level. However, homozygous HuR conditional (OX40-Cre HuR^fl/fl) KO mice showed increased Il2, Il4, and Il13 mRNA and protein via different mechanisms [128]. These findings demonstrate that the dosage of HuR is critical for the management of cytokines in the immune system. In another study, vascular HuR^SMKO (smooth muscle-specific HuR knockout) mice showed hypertension and cardiac hypertrophy [129]. The functional abnormalities in HuR^SMKO mice were attributed to decreased mRNA and protein levels of RGS (regulator of G-protein signaling) protein(s) RGS2, RGS4, and RGS5, which increased intracellular calcium levels and elevated blood pressure. Consistently, the degree of intracellular calcium ion increase and blood pressure in HuR-deficient smooth muscle cells was reduced by overexpression of RGS2, RGS4, or RGS5 [129]. Since HuR has an alleged role in the post-transcriptional activation of inflammatory mRNAs, an inducible increase of HuR in murine innate compartments suppresses inflammatory responses in vivo [130]. In macrophages, HuR overexpression induced the translational silencing of specific cytokine mRNAs despite positive or nominal effects on their corresponding turnover. The study demonstrated that HuR acts in a pleiotropic fashion in inflammation through its functional interactions with specific mRNA subsets and negative post-transcriptional modules [130]. Taken together, knowledge gained from HuR KO mice models further supports the use of HuR inhibitors in clinical settings. However, the lack of studies related to HuR’s role in the TME, and immune crosstalk provides a positive framework for us to block HuR activity and study its role in antitumor functions in vivo using mouse models.

8. Conclusions and future perspectives:

HuR functions in a wide range of biological activity in a tissue-specific manner, both in mRNA stability and translation. HuR overexpression or dysregulation, including altered ARE mRNA interactions, has been implicated in many diseases, including neurodegenerative disorders, immune disorders, and cancers [76]. Hence, HuR is attractive to chemically target and manipulate its binding to ARE mRNAs for the control of gene expression. As highlighted here and recently reviewed comprehensively [36], several FDA-approved drugs have been identified that target HuR. Despite these advances in HuR targeting, it is still unanswered how HuR targeting drugs can control tumors in an in vivo TME and immune response. The major obstacle for this nonresponsive action is the absence of suitable high throughput assays for HuR small-molecule probe discovery and the heterogeneous population of cancer cells. As we described here, the recent advancement of single-cell RNA sequencing combined with HuR manipulation may reveal which cancer cell types respond to HuR inhibition.

Targeting RBPs with specific inhibitors is still an emerging field compared to targeting other proteins like kinases and transcription factors [131, 132]. In addition to targeting HuR and altering its biological activity, small molecule inhibitors that indirectly modulate its interaction with ARE mRNAs can serve as indirect modulators of HuR (see graphical illustration). RNA is negatively charged, and RBPs are positively charged, making their interacting surface have many RNA-proteins in the interface. The disturbance of the interface makes it harder for small molecules that effectively inhibit RBPs through competitive binding studies with RNAs. Although the RNA-competitive RBP inhibitors have weaknesses, the allosteric RBP inhibitors might have low off-target effects against other RBPs and will likely attract more attention in the coming years [16]. Both FRET and FP are conventional screening methods used to detect HuR inhibitors. However, in looking beyond these assays, established methods like proximity-based labeling strategies and photo-crosslinking capture methods will support the identification of additional novel inhibitors for HuR.

The recent emergence of reports linking RBPs to the formation of phase-separated posttranscriptional condensates [133, 134] logically raises questions about how the physicochemical properties of condensates could influence the action of RBP-targeting small-molecules. Studies have indicated that drugs that target proteins have significant preferences for entering posttranscriptional condensates over other types of condensates [135]. For example, HuR, along with other RBPs, can phase separate in microtubules, which can be used as a platform to understand the mechanisms underlying liquid-liquid phase separation and their deregulation in human diseases [136]. It has also been demonstrated that certain small molecules preferentially interact with specific condensates based on their physicochemical properties [135]. Therefore, it is appropriate to evaluate the physicochemical properties associated with partitioning into condensates, which could significantly affect a small molecule’s ability to reach its target. Moreover, it will be imperative to understand the circumstances where entering a condensate is required for drug action. At present, there are many more questions than answers regarding the relationship between condensates and drug action. Still, future insights may have a significant impact on the ways we approach RBP-targeting and small-molecule design.

In conclusion, research into HuR inhibition using small molecules or siRNAs has evolved rapidly in the past decade, providing an opportunity for us to develop a monotherapy or combination therapy with anti-cancer drugs. However, many HuR inhibitors were identified and tested in vitro cell culture approaches. HuR’s critical role in the TME remains an understudied area as well as how the inhibition of HuR changes the TME using in vivo mouse models.

Abbreviations:

HuR

Human antigen R

ELAVL1

Embryonic Lethal, Abnormal Vision, Drosophila-Like 1

PTR

Posttranscriptional gene regulation

RBP

RNA-binding proteins

mRNP

messenger RNA nucleoproteins

ARE

AU-rich elements

UTR

Untranslated region

ALS

Amyotrophic Lateral Sclerosis

ARE

AU-rich elements

Chk2

Checkpoint Kinase 2

DHTS

15,16-dihydrotanshinone-I

EMSA

electrophoretic mobility shift assays

ELAVL1

Embryonic Lethal, Abnormal Vision, Drosophila-Like 1

FP

Fluorescence Polarization

FRET

fluorescence resonance energy transfer

GC

germinal center

HTS

High throughput screening

HuR

Human antigen R

mRNP

messenger RNA nucleoproteins

PDB

Protein Data Bank

PTR

Posttranscriptional gene regulation

RBP

RNA-binding proteins

RRM

RNA-recognition motif

TME

tumor microenvironment