Recent clinical trends in Toll‐like receptor targeting therapeutics

Abstract Toll‐like receptors (TLRs) are germline‐encoded receptors that are central to innate and adaptive immune responses. Owing to their vital role in inflammation, TLRs are rational targets in clinics; thus, many ligands and biologics have been reported to overcome the progression of various inflammatory and malignant conditions and support the immune system. For each TLR, at least one, and often many, drug formulations are being evaluated. Ligands reported as stand‐alone drugs may also be reported based on their use in combinatorial therapeutics as adjuvants. Despite their profound efficacy in TLR‐modulation in preclinical studies, multiple drugs have been terminated at different stages of clinical trials. Here, TLR modulating drugs that have been evaluated in clinical trials are discussed, along with their mode of action, suggestive failure reasons, and ways to improve the clinical outcomes. This review presents recent advances in TLR‐targeting drugs and provides directions for more successful immune system manipulation.

(MAPKKK) family and forms a complex with the TAK1 binding proteins, TAB1-3. TAK1 deficiency reduces inflammatory signaling across TLRs; however, no such difference has been observed in response to a deficiency of TAB proteins. 33,34 TAK1/TABs signaling then branches into two arms: activation of nuclear factor κB (NF-κB) and MAPK. NF-κB is held inactive in the cytoplasm by inhibitor of κB (IκB), which is phosphorylated by IκB kinase α (IKKα) and IKKβ, and degraded via ubiquitin mediated-proteasomal degradation, exposing a nuclear localization signal in NF-κB, and subsequently translocating to the nucleus as reviewed by Kawai and Akira. 35 NF-κB is a hub molecule for inflammatory signals and it induces the expression of a wide array of molecules that cause inflammation, alteration in cell surface receptors, expression of pro-and anti-cancerous molecules, and perturbation in cell motility, among other responses. TAK1 also activates MAPK family members, including MKK7 and/or MKK6/3, resulting in the phosphorylation of p38 and JNK, and culminating in the activation of activated protein 1 (AP1) family transcription factors and messenger RNA (mRNA) stabilization of various genes involved in the regulation of inflammation ( Figure 1B). 4,36 TRIF-dependent signaling is a separate arm of TLR signaling perpetuated only by TLR3 and TLR4, where TRIF interacts with TRAF3 and TRAF6. TRAF6 interacts with receptor interacting protein kinase 1 (RIPK-1), which transduces the signal by activating TAK1, a crucial branch point in the TLR signaling pathway. TRAF3 activates IKK-related kinases, such as TANK-binding kinase 1 (TBK1) and IKKi, along with NEMO, and the transduced signals culminate in interferon (IFN)-regulatory factor 3 (IRF3) phosphorylation, which translocates into the nucleus after dimerization, inducing expression of type I IFN genes. 4,36 The production of IFNs is the prominent outcome of TLR3 and TLR4 pathways to counter viral infections, which in turn regulated by IRF3. Recently, it has been shown that phosphatidylinositol 5-phosphate (PtdIns5P) can regulate IRF3 activation. This inositol lipid can bind to and facilitate complex formation between IRF3 and TBK1, leading to the IRF3 phosphorylation by TBK1, situated proximally. 37 Furthermore, during viral infection, production of the inositol lipid, PtdIns5P, could be observed by evaluation of PIKfyve activity. 38

| ENDOGENOUS REGULATION OF TLR SIGNALING
Regulation of TLR signaling is achieved through various molecules that restrict it to an appropriate level to avoid any detrimental consequences in the form of autoimmune or inflammatory diseases. These regulatory molecules bind to key components of TLR signaling and quench their activities as reviewed elsewhere. 39 The MyD88-dependent pathway can be suppressed by spleen tyrosine kinase, Cbl-b, and suppressor of cellular signaling 1 (SOCS1), while the TRIF arm is negatively regulated by sterile αand armadillo-motif-containing protein (SARM) and TRAM adapter with Golgi dynamics domain (TAG). 40,41 The inhibition mechanisms of molecules can be unique or may overlap. Similarly, SOCS3 and deubiquitinating enzyme A (DUBA) negatively regulate TRAF3 42 while A20, cylindromatosis, TANK, tripartite motif 38 (TRIM38), ubiquitin-specific protease 4, and small heterodimer partner can negatively influence TRAF6 ( Figure 1B). 39,43,44 TAK1 activation is regulated by A20 and TRIM30α. 45 NF-κB is pivotal in TLR signaling; therefore, it is regulated by numerous molecules, including NF-κB inhibitor δ (NFKBID), B-cell lymphoma 3-encoded protein (BCL-3), activating transcription factor 3 (ATF3), Nurr1, and PDZ and LIM Domain 2 (PDLIM2). 46 IRF3 is an important player in TRIFdependent pathways that is suppressed by Pin1 and replication and transcription activator-associated ubiquitin ligase (RAUL). 47 Various microRNAs (miRNAs) have been implicated in mRNA level regulation of TLR signaling molecules, including miR-21, -29, -126, -146a, -155, -199a, -148/152, and -466l. 39 Moreover, cytokine mRNA stability can also be regulated by regulatory Regnase-1 and tristetraprolin. 4,48 Collectively, TLR signaling homeostasis is established and maintained by these endogenous modulators ( Figure 1B).

| TLRS AND DISEASES
TLRs are involved in a wide spectrum of diseases that either directly or indirectly exacerbate the conditions. In recent years, many endeavors have been dedicated to delineate this relationship and compile data regarding the ANWAR ET AL.

| 1057
TLR involvement in various diseases. 4,20,[49][50][51] Here, we would like to present a brief overview of how TLRs influence the pathobiology of inflammatory, autoimmune, and cancerous diseases. [22][23][24]52 Sepsis is the worst outcome of host-pathogen interaction and is the leading cause of death in United States. 53,54 The infection by Gram-positive and Gram-negative bacteria equally contribute to the development of sepsis where exaggerated immune response lead to multiorgan failure and septic shock. 55 These bacteria harbor ligands that trigger TLR2 and TLR4; particularly, the presence of LPS significantly contributes into sepsis development. The septic shock is due to the body immune response rather than infection itself. 56 For sepsis management, various TLR inhibitors are evaluated clinically, and new modalities are being devised recently. 57 Chronic pulmonary obstructive disease (COPD) is characterized by the poor reversible air flow and bronchial inflammation. 52,58 This condition can also be worsened when TLRs react in response to viral infections. It has been observed that the COPD patients exhibit higher inflammatory cytokines, TNF-α and CCL5 in infections. 59 Among various treatments, the inhibition of TLRs can also be an approach to curb the COPD. 49 The involvement of TLRs in RA, an inflammatory disease, is well known. The exact mechanism of RA initiation is yet debatable; however, it is believed that the PAMPs from commensal flora is crucial for RA initiation. 60 After initial insult, an autocrine loop perpetuates that increases matrix metalloproteinases (MMP) and worsen the damage. Moreover, the DNA and peptidoglycan from intestinal bacteria have also been observed in RA joints. 61 This result in damaged cells that will release DAMPs such as RNAs, HMGB1, S100-A8; the presence of such molecules activate TLRs that over-inflame the situation.
SLE is an autoimmune disease that featured autoantibodies against double-stranded DNA and nucleic acidbound proteins that served both as diagnostic and prognostic markers; however, the initial events are still a mystery. 62 SLE patients manifest deficiency in clearing apoptotic cells that promote the formation of the immune complex (IC), and these ICs can trigger the endosomal TLRs. The role of TLR7 (inflammatory) and TLR9 (protective) in SLE can be different due to variation in study samples among different studies; nevertheless, TLR9 −/− murine models displayed higher TLR7-mediated inflammation concluding a regulatory role of TLR9. [63][64][65] An autoimmune disease where the immune system destroys the fluid secreting glands, for instance, the salivary gland, has a potential TLR involvement and is known as Sjogren's syndrome (SS). The patients with SS exhibit higher TLR expression, with increased expression of inflammatory cytokines in response to TLR7 and TLR9 activation. 66,67 TLRs participation in cancers act as double-edge swords; their activation can regress the tumor growth or conversely promote the tumor cells. 20,67 The accumulating data strongly advocate both aspects. Furthermore, it is now well-acknowledged that the inflammation and cancer are strongly correlated in various diseases. 68 Similarly, organs with higher PAMPs density such as gastrointestinal tract and skin are prone to TLR-mediated oncogenesis along with the organs that expose to indirect TLR agonist such as the liver. The dual role of TLRs in cancers has a significant correlation with the length and amplitude of receptor activation. TLR4 has been reported to promote colon cancer, and its deficiency can alleviate the inflammation as well as tumor burden. 69,70 The liver cancer has also been related to TLR4 activation 71 ; however, its role may be context dependent in skin cancer. 72,73 Similarly, TLRs are also critical for the cellular transformation in breast cancers, as reviewed before, 14 can critically modulate the metabolism in the tumor microenvironment, 74 and can regulate other signaling networks to favor pro-or anti-tumor outcomes. [75][76][77] 5 | TLR LIGANDS: ADJUVANTS VS DRUGS TLR signaling activates innate immunity and assists in shaping adaptive immunity. Hence, TLR ligands are attractive for use in immunotherapy and are primarily exploited as adjuvants to specifically trigger humoral and/or cellmediated responses as reviewed elsewhere. 78,79 They can also magnify the immune response toward certain poorly antigenic targets. Therefore, in the majority of clinical trials, TLR ligands are evaluated as adjuvants.
The number of trials that involve TLR ligands as adjuvants (64%) are double than those considering TLR ligands as drugs (35%). This highlights the immune-therapy role of TLRs in various diseases and their potential utilization for further exploration for immunomodulation therapy. Additionally, TLR activation can also alter other signaling pathways and it is desirable to cotarget multiple pathways with the aim of achieving improved treatment efficacy.
Apart from many ongoing trials, Food and Drug Administration approved TLR ligands, MPLA 80 (TLR4 agonist), and imiquimod 81 (TLR7 agonist) could be highlighted to address adjuvant or drug roles. MPLA has been used in various vaccine formulations, for instance, Fendrix (Hepatitis B vaccine, GSK), as an adjuvant and imiquimod is famously used to cure viral diseases as a drug. 82,83 The majority of TLRs produce redundant responses (inflammatory vs antiviral); however, there are slight, but distinct, differences in outcomes. 84 These differences are largely attributable to the relative roles of ligands and tissue-dependent TLRs expression. 85,86 F I G U R E 2 TLRs targeting ligands with respect to their relative clinical trials and disease conditions. A, The total number of clinical trials, activators (including agonists) and inhibitors (including antagonists), and the diversity of ligands are presented. The majority of ligands have been extensively pursued in different diseases, making it difficult to determine their exact numbers. The data indicate that total number of clinical trials exceeds the number of active drugs, suggesting the use of single drugs in multiple clinical trials. B, Clinical trial data showing the current status of drugs targeting TLRs from the disease perspectives. TLR ligands have been evaluated in multiple diseases including cancers, immune disorders, and viral and bacterial diseases. The largest proportion of clinical trials focuses on cancers, followed by immune disorders. "Mixed" indicates those cases where cancer and immune disease have been targeted simultaneously. The category "general" covers vaccination, clinical trials involving healthy volunteers, and those that are not covered by prior instances. This data was gathered from the clinical trials website (clinicaltrials.gov) using various keywords (cancers, immune disorders, TLR, TLRs, TLR1, TLR2, TLR3,  TLR4, TLR5, TLR6, TLR7, TLR8, TLR9, and adjuvant) from June 2017 to Jan 2018. TLR, Toll-like receptor Given their vital roles in pathogen clearance, inflammation induction, and cancer pathogenesis, TLRs are attractive targets for manipulation of the immune system in favor of the patients. Therefore, many research centers and pharmaceutical companies are attempting to develop TLRs modulators (Figure 2A). Scaffolds of naturally occurring modulators are ideal candidates for targeting these receptors; thus, these have been heavily investigated in clinical studies and are emerging as a fruitful approach in clinical trials. An exhaustive search of the literature also supports this notion.
6.1 | TLR2 (TLR1/TLR6) TLR2, in combination with TLR1 or TLR6, detects the lipoproteins, diacyl lipid, or triacyl lipid, respectively, which makes it unique in forming functional heterodimers with other TLRs. Further, TLR2 interacts with modified proteins such as glyco-and lipoproteins, peptidoglycan, and zymosan, allowing it to detect a variety of PAMPs. 87 This heterogeneity in TLR2 PAMP detection ranges across all types of pathogens, including viruses, bacteria, fungi, and parasites. The TLR2 expression has been detected in immune, endothelial, and epithelial cells, 88 indicating that it is a functionally ubiquitous molecule. The homodimerization of TLR2 has been reported; however, further studies are required to confirm these findings. 8,12,89 The ubiquitous nature and pivotal role of TLR2 make it an attractive drugtarget for various diseases; consequently, many clinical trials have been initiated to evaluate the efficacy of various lipopeptide derivates. Compounds being evaluated in clinical trials include lipopeptides, lipoproteins, oxidized lowdensity lipoproteins, and TLR2 specific humanized IgG4 antibody, either alone or in various combinations ( Table 2).
The most recent ligands, such as CBLB612 (synthetic lipopeptide TLR2 agonist), ISA-201 (peptide agonist for TLR2), OPN-305 (TLR2 antagonizing IgG4 monoclonal antibody), are in phase 2 in clinical trials primarily for oncogenic therapy, and act both as drugs and as adjuvants ( Figure 2B). 90,91 The chemical constituents of these molecules are either lipoprotein or protein derivates, indicating that TLR2 can be targeted using mimetics of its natural ligands. This is not an absolute requirement; however, it is useful to note the existing therapeutic trend while targeting TLR2. Moreover, other than OPN-305, the majority of molecules in phase 2 trials are agonists of TLR2, highlighting the importance of TLR2 activation in the context of malignancies. Small molecule-based therapeutics have potential side effects that can be overcome by the application of biologics, including monoclonal antibodies (OPN-305). 90 The inhibition of TLR2 overactivation using OPN-305 has potential applications in the treatment of inflammatory diseases. TLR2 in TLR2/1 or TLR2/6 complexes exhibits a cavity on the binding junction of its convex side that allows the docking of Pam3CSK4 and other TLR2-modulating ligands. 12,92 Pam3CSK4 has two esters and one amide bound lipid chains. The ester chains interact with TLR2, while the amide bound lipid chain can be accommodated into the hydrophobic cavity provided by TLR1 ( Figure 3). 12,93 The TLR2/1 complex can further be stabilized by interprotein hydrogen bonding and hydrophobic interactions. 12 The hydrophobic cavity in TLR1 has been mutated with bulky amino acids (Met338 and Leu360 to Phe338/360) in TLR6 to make binding of any lipid chain unfavorable, explaining the diacyl requirement for TLR2/6 complex formation.

| TLR3
TLR3 forms homodimer and signals in an exclusively TRIF-dependent manner in response to viral infections (double-stranded RNA [dsRNA]) and stimulates the production of IFNs. The only known agonist for this TLR is poly-ICLC (and its derivatives), which is being investigated in various clinical trials. 94,95 The success and ubiquitous nature of poly-ICLC led to the belief that this was the only realistic possibility for targeting this TLR; however, recent studies have identified other small molecules that can either inhibit or activate TLR3. 96,97 These alternatives will not be available for clinical trials for a considerable period of time. There are a few clinical trials that involve anti-TLR3 antibody to evaluate its efficacy in healthy individuals and asthmatic patients. 98 The success of these proof-of-concept studies will lay the foundation of antibody-based endosomal TLRs targeting in various diseases.
However, in rhinoviral infection, the antibody could not demonstrate any improvement in asthmatic condition. 99 Targeting of TLR3 is currently used as adjuvant therapy, along with other drugs or vaccines, against a variety of cancers; nonetheless, the sole activation of TLR3 to curb any disease has yet to be successfully explored (Table 3).
F I G U R E 3 TLRs with bound ligands. The ligand binding mechanism of the extracellular TLRs (left, TLR1, 2, 4, 5, 6) and endosomal TLRs (right, TLR3, 7, 8) has been presented. Each monomer has been labeled; however, for the homodimers, the other monomer has been labeled with asterisk (*). In the case of TLR5, flagellin-bound single monomer has been given. The respective protein databank (PDB) ID has also been given at the bottom of each structure. TLRs recognize several molecules, including protein, lipopeptide, small molecules and nucleic acids, and the bound ligand with individual TLR has been shown in 2D interaction diagram. The color code for the 2D interaction is given at the bottom of the figure Poly-ICLC is a synthetic complex of polyinosinic-polycytidylic acid (nucleic acid mimetics and pathway intermediates), carboxymethylcellulose, and poly-L-lysine (stabilizers). As dsRNA is a natural ligand with relatively low stability, its mimetics could be an affordable means of activating this TLR. Activation of TLR3 depends on dsRNA binding at two opposite sides of its ectodomain, which favorably relocates the C-terminus of the ECD to facilitate further interactions and increased stabilization. 100,101 TLR3 interacts with the nucleotide backbone, rather than nucleotide bases, which explains its activation via multiple nucleotide combinations ( Figure 3). 100

| TLR4
TLR4 is the only TLR that can function both at the cell membrane and in the endosome, and that can signal through MyD88-and TRIF-dependent pathways. 102,103 This has led to the evolution of additional precautions, such as an extensive ligand detection mechanism (cluster of differentiation 14, lipid binding protein, and coreceptor myeloid differentiation factor 2 [MD2]) and signal propagation mechanism (requirement of MAL for MyD88 and TRIFrelated adapter molecule [TRAM] for TRIF signaling pathways).
Among TLRs, only TLR4 has a suitable ligand binding pocket provided by MD2, rather than by the ectodomain of TLR4. This also provides an additional means of TLR4 (in)activation, 104 since MD2 has a large hydrophobic cavity and lipid A derivatives are suitable binding molecules; however, other methods, such as disruption of MD2 binding with TLR4 or inhibition of interaction with the activating ligand, have also been explored. 105 In case of MD2 and lipid A interaction, the lipid with six acyl chains (lipid VI-A) can fully occupy the pocket and reorient the side chain of F126 amino acid into the binding cavity. This creates a favorable environment for the other TLR4/MD2 to dock properly ( Figure 3). 106 However, in case of a lower number of acyl chains, their binding is unable to reorient the side chain that, in turns, creates a steric hindrance for other heterodimer, leading to TLR4 inhibition. 11 (Table 4). In addition to inhibition of MD2mediated TLR4 signaling, the interaction of HMGB1 with TLR4 has also been considered in recent trials to improve the efficacy of anticancer drugs (https://clinicaltrials.gov/ct2/show/NCT02995655). Besides direct modulation, the addition of a constitutively active form of TLR4 as a vaccine substitute (https://clinicaltrials.gov/ct2/show/ NCT02888756) and inhibition of dipeptidyl peptidase-4 (DPP4) that induces IL-6 expression through TLR4 are also the subjects of therapeutic evaluations. 111

| TLR5
TLR5 detects the bacterial monomeric flagella and mounts an immune response. 112 120 In the majority of therapeutic settings, ligands for TLR5 act as adjuvants rather than as stand-alone drugs, enhancing the efficacy and potency of vaccine candidates (  118,121 Recently, the crystal structure of zebrafish TLR5 with flagellin was determined, providing insights into its mode of activation ( Figure 3). 122 The leucine-rich repeat 9 (LRR9) region in TLR5 has a critical role, and Arg89, Glu114, and Leu93 from flagellin form a hotspot with chemical ANWAR ET AL. Proinflammatory responses, such as expression of IL-12, TNF-α, and macrophage inflammatory proteins-1α (MIP-1α) were enhanced by TLR8 agonism when compared with TLR7, leading to characteristic differential cell induction profiles. TLR8 can also be activated by ssRNA as natural ligand and by VTX-2337 (motolimod), a synthetic small molecule selective for TLR8 and is being evaluated in clinical trials. 124,130 TLR8 is a less studied receptor, as its roles overlap with those of TLR7, with which it shares multiple features. When treated with VTX-2337, TLR8 stimulates TNF-α and IL-12 production at lower concentrations in human PBMCs. It also induces TNF-α and IL-12 secretion from monocytes and myeloid DCs through the NF-κB pathway. IFNγ secretion was observed when NK cells were treated with VTX-2337, which can enhance the lytic capability and antibody-dependent cell-mediated cytotoxicity of NK cells. 130 VTX-2337 also improves the efficacy of pegylated liposomal doxorubicin in treatment of ovarian cancer in a mouse model with humanized immune system that has been reconstituted with human CD34 + cells. 131 This is the only ligand molecule that has been actively evaluated for treatment of a variety of cancers, including head and neck cancer, colorectal, pancreatic, melanoma, breast, renal cell carcinoma, nonsmall-cell lung carcinoma, and other solid neoplasms. In the majority of cases, VTX-2337 was used in combination with other drugs; however, it is also being evaluated as a stand-alone drug for treatment of lymphoma. 130 TLR7 and TLR8 share similar activation patterns, both have z-loops involved in ssRNA recognition, and both possess two binding sites; the first binding site binds guanosine and uridine in TLR7 and TLR8, respectively, while the second binds ssRNA in both cases ( Figure 3). 132 In TLR7 ssRNA binding primes the receptor for guanosine binding and subsequent dimerization, while synthetic molecules, such as R848, can activate TLR7 without the need for ssRNA. 133 | 1077 target TLR9 are either nucleotides or nucleotide derivatives. There are various types of CpG DNAs that are being evaluated in different trials for treatment of diverse conditions. AZD1419 is a C-type CpG-based inhaled TLR9 agonist for treatment of asthma and to stimulate IFNs production in lungs. This treatment was classified as welltolerated and safe in phase 1 human trials with potential disease-modifying characteristics and is a promising new therapeutic for use in various immune diseases. 138 CYT003 was initially found to be effective; however, its effects were not confirmed in phase 2 clinical trials where 35 patients were treated with varying doses of CYT003. 139 Another TLR9 agonist, EMD 1201081, was evaluated in phase 2, open-label, randomized trial in patients with head and neck cancer, and was found to be ineffective in the tested dose regimen. 140 GNKG168 is another CpG-based molecule that can induce CD8 + T cell antitumor cytotoxic responses; however, it was withdrawn in clinical phase 1 because of sponsor reluctance to further support the study 141 (NCT01743807) (Tables 7 and 9).
Similar to other TLRs, TLR9 forms a symmetrical complex with CpG-DNA; nonetheless, during inhibitory DNA interactions, it remains in a monomeric form. CpG-DNA binding with TLR9 is symmetric and they form a stoichiometric complex of 2:2, as DNA is recognized by both TLR9 monomers, particularly via the amino-terminal fragment (LRRNT-LRR10) from one protomer and the carboxy-terminal fragment (LRR20-LRR22) from the other. 142 CpG-DNA-based TLR9 inhibition is mediated by binding to the concave surface formed by LRR2-LRR10, thereby inhibiting its signaling.

| TLR10-13
Other than TLR1-9, humans also have TLR10 and TLR11, whereas they lack TLR12 and TLR13. 143 The expression of TLR10 has been confirmed in humans (spleen, lymph node, B cells, monocytes, and neutrophils) 144 ; nonetheless, its function and specific ligand are yet to be determined. Recently, it was suggested that TLR10 may act as an antiinflammatory TLR, rather than a conventional inflammatory receptor and that it modulates TLR2-mediated responses through the formation of heterodimers with TLR1 or TLR6. 145 Humans have a pseudogene homologous to TLR11 that includes a premature stop codon, resulting in lack of protein expression. 146 TLR11 and TLR12 have been studied in mouse and they have shown to detect profilin from Toxoplasma gondii and be capable of forming heterodimers. 143

| INTERDEPENDENT AND CROSS-TALK AMONG TLR PATHWAYS
Since TLRs overlap in their structures and signaling pathways, it is rational to assume that one single ligand can activate multiple TLRs; however, this is less common among plasma membrane expressed TLRs, TLR2/1, TLR2/6, TLR4, and TLR5, and there are a few ligands that can share targets, particularly for TLR2 and TLR4. This situation is very common among endosomal TLRs, partly because they are all involved in sensing nucleic acids, and endosomes have a specific pH range that is also thought to contribute to their activation. Various ligands exert their actions on multiple endosomal TLRs (eg, TLR7/8 or TLR7/8/9), which may imply a combination of multiple pathways in their activity, a common mode of activation, and, to some extent, H + interference of these TLRs being a common factor 147,148 (Tables 7 and 10). TLR7 and TLR8 detect ssRNA, which may explain why one ligand is equally effective against both TLRs. Some studies have also explored the independent targeting for either TLR7 149 or TLR8. 150  as reduced survival, exaggerated cytokine responses, and salmonella hepatitis, while TLR2 deficiency produces the opposite effects. Deficiency of either TLR may disrupt NK cell cytotoxicity, and IFN-γ and ROS production. 151 Synergism is very common in TLRs. When monocyte-derived DCs have been triggered with a TLR8 ligand, TLR3 or TLR4 are also activated, resulting in expression of IL-6, IL-10, IL-12, and TNF-α elevation. These results were also confirmed by increased binding of IRF and signal transducers and activators of transcription (STAT) transcription factors to their respective DNA binding sites, which was abolished when NF-κB, p38, and phosphoinositide 3-kinase (PI3K) inhibitors were used. 152 These data suggest that co-operation among TLRs is perpetuated, not only at the top level but also among different signaling pathways to ensure proper and balanced expressions of target genes.
Synergy and tolerance of TLRs are long-established and are critical to the innate immune response. The coadministration of LPS (TLR4 agonist) and MALP-2 (TLR2 agonist) to mouse macrophages resulted in increased TNF-α production. 153 Repeated treatment with LPS or MALP-2 resulted in a hyporesponse, also termed tolerance.
Intriguingly, pretreatment with any ligand results in lower responses on exposure to the second ligand. 153  for the induction of this differential response, and its expression is modulated by IL-7. 154 The mycobacterium extract, 155,156 and autophagosome-enriched cancer vaccine (DRibbles), 157 which likely contain multiple biological molecules and can trigger numerous TLRs, is being evaluated in clinical trials; however, caution is required when considering the use of such substances in the clinic due to synergy and differential responsiveness of TLRs to various ligands. Moreover, DC vaccines that have been matured using TLR ligands are also therapeutically relevant, owing to the use of TLR ligands in their production. 158,159

| FAILED CLINICAL TRIALS
The proportion of failures of clinical trials depends on the clinical stage, as well as the type of disease; particularly, failure at phase 3 is an impediment to the development of successful therapy for various diseases and TLRs are no exception. For example, eritoran, a TLR4 antagonist, that was being evaluated for treatment of sepsis could not meet its target end-point in phase 3 when data from~2000 patients were analyzed. 160 Among the reasons of failure of eritoran, there were oversights in study design, patient population differences, improved patient care methods, and mixed bacterial infections. 160 Similarly, imiquimod, a TLR7 agonist, produced a divergent result in phase 3 when evaluated for treatment of the skin disorder, molluscum contagiosum (MC) lesions, in children. 161 Imiquimod was first approved by the Food and Drug Administration in 1997 for treatment of genital warts. This approval has prejudiced its subsequent off-label use as the treatment of MC in children, since it was already shown to be effective against viral-based diseases and its use is supported by several research and clinical investigations. 82,83 This off-labeled use of imiquimod is still debatable. 83

| 1081
Lack of recruitment (23%) and unstated reasons (such as unknown reasons for termination, unable to begin the study, unavailability of a drug, or protocol modification resulting in cessation of a trial; 26%) comprise the majority of reasons for failure of clinical trials targeting TLRs, followed by safety issues (18%) and financial concerns (where a sponsor withdraws the drug; 15%). Moreover, 15% of trials do not show any efficacy in subsequent clinical trials.
The inadequate understanding of the biology of TLRs may also contribute to drug failure, 160 underlining the need for further studies, increased understanding of the theoretical background of disease etiology and progression, modification of protocols to address problems, and trial redesign.
While many factors contribute to a failed clinical trial, a common reason underlying failure is a lack of serious focus on biomarker discovery and implementation. There is a clear trend of success among those trials including biomarker selection (25.9%) compared with those lacking selection biomarkers (8.4%). 164 There are several methods that can be used to reduce failure rates, such as early identification of false drug candidates, stratification of patients, development of diagnostics, proper use of pharmacogenomics through machine learning, and other analysis tools that can provide improvements and efficiencies in patient categorization. Focusing on neglected disease areas can also help to reduce the failure burden. In recent years, the Drugs for Neglected Diseases Initiative has approved six treatments within a decade, with many more in the pipeline. This is not only dramatically reducing the cost of drug development, but also providing hope for individuals affected by neglected diseases and incentivizing the pharmaceutical industry to continue their search for new drugs.
Financial and commercial reasons are also major contributors to trial failures because sponsors are "unwilling," or there are "failure to pursue" investigational drugs for commercially important diseases. This can be reduced if pharmaceutical companies focus on diseases that lack adequate therapeutic intervention, as drugs that show positive effects will soon be marketable. Additionally, if such a trial does fail, it has a lower cost impact on the company.
Drug development is a lengthy process that starts with lead molecule identification and progresses through optimization, animal modeling studies, pharmacokinetic and pharmacodynamic studies, and preclinical and clinical stage trials. Therefore, if a drug fails to show any effect or shows toxicity in clinical studies, there must have been a series of oversights during earlier experimental stages. It is hard to give a single reason for any failure and failures may encompass complex issues, such as the use of subjective, composite, or surrogate endpoints. 165 Moreover, biases in outcome reporting and publications; underreporting of adverse events; failure to select an appropriate patient group; preference for relative outcomes, rather than absolute values; no defined core outcome sets; lack of transparency and basic science; inappropriate study population size; and lack of data integrity are among the reasons for trial failures.
Finally, during clinical trials involving humans, factors that influence the drug metabolism, distribution, and secretion are diverse that predispose the pharmacokinetically and pharmacodynamically optimized drug molecules to failure.

| PERSPECTIVES IN TLR TARGETING
Researchers are expending extensive efforts to generate appropriate solutions for various inflammatory, autoimmune, and malignant conditions; however, the process is not straightforward, rather it is littered with unexpected events and outcomes, along with unknown obstructions that severely undermine the efforts of the research community.
In the majority of studies targeting TLRs, the investigated compounds are related to or derived from natural ligands; particularly those targeting TLR3, TLR4, TLR5, and TLR9, and somewhat those for TLR2. TLR7 and TLR8 have the benefit of being targeted by small molecules rather than ssRNA. The instability of ssRNA molecules can hinder their use for TLRs activation. However, since RNAi technology is being evaluated in more than 100 clinical trials, 166 stability should not be an issue, rather, tuning of small molecules is far easier than tuning biologics for therapeutic purposes.
Other than molecules derived from natural ligands, it is necessary to focus on the chemical space that can be used to target TLRs. 106,118 This broadening of the chemical space will provide more potent, specific, and less toxic molecules, resulting in fewer trial failures. Biologics are gaining popularity, as they have a higher ratio of success, and are comparatively safe and specific. [167][168][169][170][171] It is estimated that the biologics will soon become the norm in therapeutics, in addition to being responsible for the majority of revenue. 172 For different TLRs, the therapeutic trend can vary; however, a rise in antibody-mediated TLR inhibition (TLR2, TLR3, and TLR4) and novel molecular backbone (independent of PAMPs) have been seen in recent therapeutics.
The evaluation of various drugs for similar or different conditions is also an optimal approach, which can facilitate the development of single drugs for multiple diseases. In this context, research laboratories can screen the outcomes of phase 2 failures that have been abandoned by their sponsors to evaluate them for other symptoms. 173 Such an approach can dramatically reduce the cost, speed up the process, and will encourage pharmaceutical companies to share their data with research laboratories for application to other disease targets.
Rather than directly inhibiting TLRs, it may be more appropriate to target the transcription regulation of TLRs to suppress their expression, 174 as described in a study where the authors used GST-21 for cytokine inhibition, which could be reversed by the janus kinase 2 inhibitor, AG490. Since the majority of TLRs regulate the similar pathway, targeting of their downstream inhibitory signaling mechanisms should also be explored to further intensify the benefits of their inhibition.
Lack of clinical data is an impediment to the development of clinical research. It is estimated that approximately half of all clinical trials are not reported in either peer-reviewed journals or clinical trial websites (clinicaltrials.gov; http://apps.who.int/trialsearch/). 175,176 It is now necessary to develop additional TLR ligands that should not mimic PAMPs, explore new biomarkers for disease progression, revise protocols, and clinical trials, target small subsets of patients, improve the understanding of the basic biology of diseases, and improve final outcomes, which must legitimately refer to the progress of the disease and the effect of the compound being applied.

| CONCLUSIONS
TLRs are among the ideal targets for exploitation in immunotherapy; however, their biology still needs to be better understood in the context of target diseases. These receptors are capable of inhibiting disease pathophysiology, as well as exacerbating inflammatory diseases. Given this dual role, it is imperative to fine tune their activation using a multidrug approach. Cumulative evidence suggests the participation of TLRs in almost all diseases is unique and can be exploited by including their ligands as adjuvant treatment during regular immunotherapy or as part of other therapeutic regimens.
It is vital to create superior disease models that assist in early phase evaluation of drugs, improve diagnostics and evaluation of disease progression, and facilitate identification of novel biomarkers that reliably indicate disease progression and real-time disease monitoring. Finally, the availability of clinical trial data should be ensured to guide the scientific community in their endeavors. This would also assist in the refinement of targets and lead molecules and improve the pathophysiological manifestations of diseases. Using a combination of computational power, next-generation sequencing and proteomic data, machine learning approaches, and improved availability of results, we are hopeful that a dramatic increase in new therapeutic options for various inflammatory diseases and cancers involving TLRs and a decline in clinical trial failures will be achieved.

CONFLICTS OF INTEREST
The authors declare that there are no conflicts of interest.