Mammalian p38 mitogen-activated kinases (MAPKs) are activated by cellular stresses and inflammatory cytokines and are critical for normal immune and inflammatory response (63). Several studies showed HBV infection could activate mitogen-activated protein kinases (MAPKs), including p38 MAPK kinases, c-Jun N-terminal kinase (JNK), and extracellular signal-regulated kinases (ERKs) (64–67). p38 MAPK, but not JNK or ERK1/2, was significantly phosphorylated in Huh7 cells after HBV infection and thereby promote intracellular HBV replication (68) with the involvement of STAT3 (67). p38 MAKP also acts as a negative regulator of type I IFN pathway through unfolded protein response (UPR)-stimulated priming phosphorylation of IFNAR1, ensuring IFNAR1 down-regulation and type I IFN signaling attenuation (38). Given this knowledge as well as the important role of HBV in p38 MAPKs production, we hypothesized that negative regulation of p38 MAPKs should stabilize IFNAR1 and improve IFN therapy efficacy against chronic HBV infection, which requires further investigation.

Matrix metalloproteinase 9 (MMP-9) is expressed in normal leukocyte and transformed cells, and is associated with IFNAR1 degradation (39). IFNAR1 degradation can be promoted by either ligand binding and cellular signaling induced ubiquitination-dependent pathway, or Ser532 phosphorylation and p38 kinase activity requiring ligand-independent way (69). MMP-9 can be activated by HBV in CHB patients (70, 71) and in turn facilitates HBV replication through attenuating IFN/JAK/STAT signaling, interacting with IFNAR1 to promoting its phosphorylation, ubiquitination, subcellular distribution, and degradation of IFNAR1 through lysosome pathways in Ser532 phosphorylation and p38 independent manner, as well as blocking its binding to IFN-α (39). However, a recent study found type I IFN strongly promotes the clearance of MMPs (72), thus, further study is needed to investigate the role of MMPs in the IFN-α therapy mediated antiviral efficacy against HBV infection.

Interferon-induced transmembrane proteins (IFITMs) are a family of small proteins that localize in the plasma and endolysosomal membranes (73) and were originally described based on their expression after IFN therapy (74). The human IFITM family comprises five members, including immune-related IFITM1, IFITM2, and IFITM3, as well as IFITM5 and IFITM10 with an unknown role in immunity (75). Brass et al. first described IFITM proteins as antiviral factors that target the early life cycle steps of several viruses (76). HCV infection could be restricted by IFITM proteins by targeting the endocytosed HCV virion for lysosomal degradation (77). However, our previous study indicated IFITM2 was shuttled by exosomes to DCs and inhibit IFN-α synthesis via inhibiting phosphorylation of extracellular signal-regulated kinase (ERK), TANK-binding kinase 1 (TBK1), and interferon regulatory factor 3 (IRF3), thereby blocking anti-HBV efficacy of exogenous IFN-α in chronic HBV infection (40). Thus, down-regulating IFITM2 may be a potential strategy to improve the therapeutic response to IFN-α treatment.

Most negative regulators function in the various biological processes in IFN production and signaling pathways, including inhibit IFN expression, disturb IFNAR function, and suppress the expression of ISGs which serve as anti-HBV molecules (Table 1).

Ubiquitin-specific protease 18 (USP18, also known as UBP43) and Interferon stimulated gene 15 (ISG15) have been studied extensively for their negative regulation of type I IFN signaling. Various studies showed that ISG15/USP18 signaling activation is involved in IFN therapy non-response in both HBV and HCV patients (78, 79). Responding to viral infection or IFN stimulation, both USP18 and ISG15 are strongly upregulated (80). USP18 belongs to the ubiquitin protease family, which is responsible for removing ubiquitin or ubiquitin-like proteins from their conjugated substrates (81). ISG15 is a ubiquitin-like modifier that binds other proteins in a process called ISGylation, and its antiviral activity was first observed using a recombinant chimeric Sindbis virus and IFNAR^−/− mice model (82). ISG15 conjugated to various cellular substrates via an enzymatic cascade: E1 activating enzyme UbE1L (83), E2-conjugating enzyme UbCH8 (84, 85), and various E3 ligase (86, 87). ISG15 targets host and viral proteins and impacts diverse cellular pathways, including RNA splicing, chromatin remodeling/polymerase II transcription, cytoskeleton organization and regulation, stress responses, and translation mainly in a conjugation-dependent manner (88, 89). By deISGylating ISG15, USP18 was stabilized by unconjugated ISG15 via preventing its S-phase kinase-associated protein (SKP2)-induced ubiquitylation, therefore mediates downregulation of type I IFN signaling during HBV infection (41). Independent of its deconjugating activity, USP18 also competes with JAK1 for binding to the IFNAR2 and therefore impede IFN signaling (42). Because USP18 and ISG15 are inducible by viral factors, these inhibitory proteins may be important mediators in reducing IFN therapy efficacy by negatively regulating IFN-α signaling and host immune against chronic HBV infection.

In addition to USP18 and ISG15, the suppressor of cytokine signaling (SOCS) proteins and cytokine-inducible SH2 (CIS) also play important roles in downregulating IFN-α signaling, particularly the JAK-STAT signaling pathway (90), and suppress type I IFN production. To date, SOCS/CIS family includes eight proteins (SOCS 1-7 and CIS) which sharing similar structures: a central Src-homology 2 (SH2) domain, a C-terminal SOCS box, and a highly variable N-terminal region (91). The SOCS box functions to recruit an E3 ubiquitin ligase complex, so the SOCS box-containing proteins can act as substrate-recognition modules to mediate the polyubiquitination and subsequent degradation of substrate proteins by the 26S proteasome (92). Responding to IFN signaling and cytokine stimulation, SOCS 1-3 and CIS proteins are induced and negatively regulate the JAK-STAT-mediated cytokine signal responsible for its production in a classic negative feedback loop (30, 45), while SOCS4-7 primarily focus on regulating the growth factor receptor signal (93, 94). SOCS1 and SOCS3, the most widely studied proteins in this family, possess a kinase inhibitory region (KIR) at the N-terminal region that inhibits JAK tyrosine kinase activity by acting as a pseudo-substrate (95). At the same time, SOCS1 is related to IFNAR1 specific signals, thereby abrogating tyrosine phosphorylation of STAT1 and reducing the duration of antiviral genes expression (46). Apart from regulating the IFN-α signaling pathway, SOCS proteins can suppress type I IFN production by binding and degrading key molecules in the TLR signaling pathway, including myeloid differentiation factor 88 (MYD88)-adaptor-like (Mal) protein (43) and IRF7 (44). HBV increased SOCS1 and SOCS3 expression and resulting in sustained STAT3 activation (96, 97). Thus, the SOCS family, especially SOCS1 and SOCS3, inhibit both type I IFN production and IFN-α signaling pathway in HBV infection and provides targets for improving IFN-α-mediated antiviral effect.

Epidermal growth factor (EGF) receptor (EGFR) is a transmembrane protein that belongs to the ErbB family of receptors and is well-studied for its association with a number of cancers. EGFR has recently been shown to be a host entry cofactor triggering HBV internalization (98). Inhibition of EGFR activity enhances the antiviral efficacy of IFN-α by both inducing SOCS3 expression to reduce IFN-α-induced STAT3 activity and enhancing STAT1-mediated antiviral response. Moreover, EGFR inhibitors inhibited the replication, antigen syntheses, and cccDNA reservoir of HBV (47). Thus, EGFR inhibitors could be an important candidate that facilitates IFN-α therapy against HBV infection.

Protein tyrosine phosphatases, known as PTPs, is a large family that encodes enzymes that are divided into the classical, phosphotyrosine (pTyr)-specific phosphatases and the dual specificity phosphatases (DSPs) (99), regulating the phosphorylation state of many important signaling molecules. The overexpression of HBx or NF-κB led to increased SHP2 expression, which is a ubiquitously expressed PTP, via NF-κB binding to the Shp2 promoter (100). SHP2, encoded by Ptpn11, containing two SH2 domains and acts as a negative regulator of IFN-induced STAT activation in a protein kinase Cβ (PKCβ)-dependent pathway (48, 49). PTP1B is another PTP protein and is capable of binding and dephosphorylating IFNAR1 and thereby suppress type I IFN signaling in human cells. PTP1B inhibitors robustly augmented the type I IFN-mediated antiviral effects against HCV (50). However, the knowledge on PTP1B expression and its function needs further study in chronic HBV infection. Currently, it is estimated that as many as 125 PTP genes are in the human genome, and most of them have not been identified (101). Thus, additional studies are required to identify PTP proteins that suppress the IFN-α signal pathway and therefore attenuate therapy in chronic HBV infection.

Apart from the above-mentioned molecules, other molecules involved in the down-regulation of IFN therapy efficacy include NEIL3, TDG, and NT5C3, but the exact biological mechanisms underlying the influence of these molecules in IFN-α treatment response remain to be determined. It is reported that positive response to IFN-α treatment in CHB patients is associated with a lower level of NEIL3 and thymine DNA glycosylase (TDG), two intra-hepatic base excision repair (BER) genes (102), but the underlying mechanism is unknown. MiR-122, a miRNA that has been demonstrated to inhibit HBV replication by directly targeting the HBV pre-genomic RNA sequence or by indirectly modulating HO-1 and CCNG1/p53 pathways (103), could be sequestered by an ISG, NT5C3 (51). Therefore, NT5C3 could efficiently inhibit miR-122 and suppress IFN therapeutic efficacy. However, the function of NT5C3 and miR-122 need further confirmation in clinical research. In addition, numerous molecules have been shown to negatively regulate type I IFN signaling, including DCST1, PIAS1, and so on (32, 104), although further studies are needed to clarify their function in HBV infection.

In general, the anti-HBV activity of type I IFN is mediated through binding to their unique receptors, following by JAK-STAT pathway activation and ultimately ISGs production. Cell-surface IFNAR consists of 2 major subunits: IFNAR1 and IFNAR2. However, soluble IFNAR was reported as an inhibitory factor of IFN therapy in both HBV and HCV infection (52, 105). The expression of IFNAR1 and IFNAR2 in lymphocytes and monocytes was increased in CHB patients which was negatively correlated to plasma glutathione S-transferase (GST), a class of cytosolic enzyme participated in the protection of cell from reactive oxygen species (ROS), suggesting that oxidative stress play an important role in the upregulation of IFNAR in CHB patients (37, 52). Oxidative stress could modulate protein misfolding and then disrupt the biological protein conformation. Therefore, oxidative stress may influence IFNAR function, further studies are required to find out the mechanism.

TNF is a pleiotropic cytokine that exerts homeostatic and pathogenic bioactivities (106). Elevated levels of TNF-α are observed in serum and hepatocytes of patients with acute or chronic hepatitis B (107), possibly through ERKs and NF-κB pathway (108). TNF-α may play a role in downregulating HBV replication in hepatocytes by non-cytopathic mechanisms and synergizing with IFN in suppressing viral replication (109, 110). On the other hand, TNF-α was reported as a negative regulator during IFN-α production determined by HBsAg (55). It is also reported that the injections of synthetic TNF-α could inhibit IFN-α-induced signals in the liver and both SOCS3 and SHP2 contributes to the inhibitory effect (53). However, anti-TNF-α-agents may lead to enhanced HBV replication and cause reactivation of HBV infection in HBsAg carrier and Occult carrier (anti-HBc+) patients (111). Consequently, the ultimate impact of TNFα-mediated effects on IFN therapy against HBV is a question that remains to be resolved.

Interleukin 10 (IL-10), also known as cytokine synthesis inhibitory factor (CSIF), is a pleiotropic cytokine with anti-inflammatory and immunosuppressive activities. HBcAg stimulated IL-10 production by CD4+ and CD8+ cells, as well as by monocytes from CHB patients, which serves as a viral strategy to downregulate the host immune response and allow viral persistence in the host (112). HBsAg inhibited the production of IFN-α by pDCs through the induction of monocytes that secreted IL-10 (55). IL-10 also suppresses NK cell IFN-γ production without altering cytotoxicity or death-ligand expression (56). In addition, SOCS1 and SOCS3 are both produced in response to IL-10, and both function as negative regulators during JAK-STAT signaling (54). On the other hand, evidence suggests that heterogeneity in the promoter region of the IL-10 gene has a role in determining the initial response of HBV infection to IFN-α therapy. IL-10 variants were more frequent among virologically sustained response (SR) compared with non-responders (NR) to IFN-α-2b. Carriage of the−592A allele,−592A/A genotype, and−1082/1819/-592 ATA haplotype was associated with SR (113).

Interleukin 8 (IL-8), or chemokine (C-X-C motif) ligand 8 (CXCL8) is a pro-inflammatory chemokine produced by various cell types to recruit leukocytes to the sites of infection or tissue injury (114). It has been demonstrated that serum IL-8 amounts is elevated during HBV reactivation, and HBeAg-negative patients had significantly higher levels of IL-8 transcripts in the liver than HBeAg-positive patients (115). Accumulating evidence indicates that IL-8 may contribute to counteracting IFN-α antiviral action (116, 117). During the IFN challenge, IL-8 expression induced by HBV can impair the ability of endogenous IFN-α to inhibit early stages of viral replication, thus facilitate viral persistence, and can also contribute to the poor response to IFN-α treatment (57). IL-8 is induced by HBV through various mechanisms, but the mechanism underlying the IL-8 affecting the anti-HBV function of IFN-α therapy still needs further discussion.

Interleukin-28B (IL-28B), also known as IFN-λ3, has been reported by numerous studies that its genetic polymorphisms are related to the therapeutic efficacy of PEG-IFN-α. Sonneveld et al. reported the proportions of IL-28B genotypes were 77, 19, and 5% for AA/AG/GG at rs12980275 and also for CC/CT/TT at rs12979860, respectively (118). rs12980275 genotype AA and rs12979860 genotype CC were associated with a higher probability of HBeAg seroconversion (118). A meta-analysis summarized that IL-28B rs12979860 CC genotype and rs8099917 TT genotype indicated a better treatment response than non-CC and non-TT genotypes for PEG-IFN-α in patients with CHB (58).

MicroRNAs (miRNAs) are a class of non-coding small RNAs that play important roles in modulating gene expression mainly by base-pairing to the 3' translational region (3' UTR) of mRNAs. miRNAs have been reported to be widely involved in antiviral immunity and some studies show miRNA plays a role in HBV-host interaction. HBV could change the expression profiles of cellular miRNA and escape host immune clearance (119). It was shown that nearly 90% of the total miRNAs in the liver are comprised of ~10 miRNAs and miR-122 is the major contribution with 50–70% of the miRNA profile (120). Zhang et al. identified 11 response-related miRNAs, including let-7a, miR-30a, miR-1290, miR-106b, miR-1224-5p, miR-939, miR-1281, miR-198, let-7f, miR-22, and miR-638, by performing miRNA microarray analysis of plasma samples and liver biopsy samples from CHB patients who received IFN therapy and using the OneR feature ranking and incremental feather selection method to ultimately establish a prediction model for predicting the initial response of HBV patients receiving IFN injections (121). Not all of the 11 miRNAs in the predicted model is negatively associated with the IFN-α therapy efficacy. Results obtained by Tan et al. suggest that the aberrant expression of miRNAs was associated with different therapy outcomes, and monitoring the fluctuation patterns of miRNAs was important for predicting the IFN-α therapeutic effects. They also identified 11 miRNAs to be associated with the effect of HBV therapy, namely, let-7d-5p, let-7f-5p, let-7i-5p, miR-122-5p, miR-126-3p, miR-1307-3p, miR-181a-5p, miR-21-5p, miR-425-5p, miR-652-3p, and miR-320a-3p, with 9 miRNAs significantly declined after therapy in the responder group: let-7d-5p, let-7f-5p, let-7i-5p, miR-126-3p, miR-1307-3p, miR-181a-5p, miR-21-5p, miR-425-5p, and miR-652-3p (122).

miRNAs could modulate the IFN-α therapy effect in multiple processes, including type I IFN production, IFN signaling pathway, as well as ISGs transcription and translation. The effect of miRNA on IFN and its signaling pathway has been well-illustrated in many reviews (123, 124), indicating that miR-130a, let-7b, miR-26a, miR-24, miR-146a, and many other miRNAs could down-regulate the effect of IFN therapy. However, only a few miRNAs are studied extensively in CHB patients. HBV infection promotes miR-146a transcription, which suppressed STAT1 and ultimately attenuated the production of type I IFN-induced antiviral factors, as well as impairing T cell function, resulting in IFN resistance and immune defects during chronic HBV infection (59, 60). miR-3613-3p was upregulated in the CHB patients suppressed IFN-induced anti-HBV immune response by targeting cytidine monophosphate kinase 1 (CMPK1) (61), a kinase responsible for the metabolism of CMP, UMP, and deoxycytidine analogs (125), and ultimately decrease the relative expressions of IFN-α and IFN-β. miR-21 has been reported to be associated with the effect of HBV therapy (122) and promote an anti-inflammatory response by increasing IL-10 production through down-regulating programmed cell death 4 (PDCD4) (126). But the detailed mechanism of miR-21 affecting the anti-HBV effect of IFN-α therapy still needs further investigation.

Some miRNAs exhibit an immune-modulating or IFN therapy enhancing effect in HBV infection. HBeAg could elevate miR-212-3p expression in macrophages, and thereby decrease the production of cytokines through targeting MAPK1 (127). Higher pretreatment plasma miR-301a-3p and miR-145-5p levels were observed in CHB patients who achieved combined response or HBsAg loss following peg-IFN/NA therapy compared to non-responders (120). miR-155 enhances innate antiviral immunity against HBV through promoting JAK/STAT signaling pathway by targeting SOCS1 and mildly inhibits HBV infection in human hepatoma cells (128). In conclusion, miRNAs play an important role in regulating the efficiency of IFN therapy in CHB patients, and could act as predictive factors of IFN therapy outcome, or be used for discovery for better treatment strategies. But the mechanism underlying these miRNAs calls for more attention.

Long non-coding RNAs (lncRNAs) are a subset of non-coding RNAs >200 nucleotides in length and have recently been suggested to act as both positive and negative regulators in viral infections utilizing various mechanisms, mediated by their specific sequences or structural motifs that bind DNA, RNA, or protein (129). Recent study suggests that HBx-LINE1, a hybrid RNA transcript of the human LINE1 and the HBV-encoded X gene can serve as a sponge for sequestering miR-122 and may therefore promote HBV expression and replication, but whether HBx-LINE1 function during IFN-α treatment is still unknown (130). On the other hand, lncRNA#32 has been reported to be positively associated with type I IFN signaling in HBV infection by regulating APOBEC3A and APOBEC3G expression (131). Given the important and complicated roles of lncRNAs in IFN anti-HBV treatment, more extensive studies are required to develop a more comprehensive profile of lncRNAs in HBV infection and IFN therapy.

Circular RNAs (circRNAs) are closed long non-coding RNAs and can bind miRNA and act as a miRNA sponge to regulate gene expression, they may also play a role in negatively regulating IFN therapy outcomes. Among top ten circRNAs significantly up-regulated in CHB compared to normal controls, circRNA has-circ-0004812 impairs IFN-induced immune response by sponging miR-1287-5p to regulate Follistatin-related protein (FSTL) 1 in chronic hepatitis B (62, 132). Several circRNAs have been demonstrated to function in HBV infection, including hsa_circ_0005389, hsa_circ_0000038, hsa_circ_0000650, hsa_circ_4099, hsa_circ_0000976, hsa_circ_0007750, hsa_circ_0027089, and hsa_circ_100338, but it is still unknown whether they function in IFN-α therapy of HBV infection (133, 134). Therefore, due to presently poor understanding of circRNAs expression, function and regulation, further studies are desired to identify the mechanisms behind different circRNA regulation patterns associated with IFN-α treatment in HBV infection.

Patient age is a host factor that is related to IFN-α therapy responsiveness in Chronic HBV infection. Generally, it is believed that younger individuals are more likely to develop a sustained virological response to IFN-α treatment than older individuals. Bonino et al. reported that patients under 40 years of age had higher combined response rates than patients over 40 years of age, regardless of which treatment they received (with or without lamivudine) (p < 0.024; OR: 1.3 per 10-year decrease; 95% CI: 1.0–1.7), indicating that aging imparts a negative influence in the IFN-α treatment efficacy (135). Similar results were showed in other clinical researches (136). Aging enhances the vulnerability to liver injury mainly by increasing oxidative stress and inflammatory response, accelerating cellular senescence, as well as suppressing mitochondrial function (137). A retrospective study showed that age parameters were also associated with significant histological abnormalities in patients with persistently normal ALT levels (PNALT) (155). Moreover, impairments of cellular, humoral, and innate immunity in the elderly may be another explanation responsible for decreased virological response to IFN-α therapy in older patients (138).

Gender is another important host factor associated with IFN-mediated antiviral efficacy against HBV infection. It is summarized that females are more likely to achieve HBeAg seroconversion after IFN therapy comparing to males (10, 139), which may be an effect of sex hormones. Androgen-androgen receptor (AR) complex activates HBV RNA transcription through binding to the 2 response elements inside Enh I, while estrogen and/or its receptor suppressed HBV replication (156). The expression level of estrogen receptors (ESRs) in the cytosol of peripheral blood mononuclear cells (PBMCs) was significantly lower in patients with CHB than in healthy controls, and ESR1 mRNA expression at week 4 of PEG IFN-α treatment was a better indicator for treatment response although the underlying mechanism is still unknown (140). Further studies are required to clarify the effect of sex hormones in IFN-α therapy.

Alanine transaminase (ALT) is a transaminase enzyme and is clinically measured as a biomarker of liver injury. ALT is associated with significant histological characteristics to some extent in HBeAg-positive and -negative patients and its normalization are related to favorable long-term CHB infection treatment outcomes. Several studies indicated that a high baseline ALT level can be a predictor of sustained treatment response (10, 139, 141).

Bile acids are steroid acids and were found to be related to the IFN-α treatment. Elevated hydrophobic bile acid concentration impairs IFN-α treatment in patients with CHB by inhibiting Jak1- and Tyk2-phosphorylation, resulting in a decreased mRNA and protein expression of ISGs, including myxovirus resistance protein A (MxA) and dsRNA-activated protein kinase (PKR) (142). The uptake of bile acid into hepatocytes is induced by Na⁺-taurocholate cotransporting polypeptide (NTCP), which also plays an essential role in HBV entry into the target cells. HBV infection may promote the expression of genes involved in lipid and bile acid metabolism by binding to NTCP and limits its function, thus promoting compensatory BA synthesis (157).

High body mass index (BMI) or obese status is related to the presence of steatosis in CHB patients. Clinical researches showed that the occurrence of hepatic steatosis is significantly high in CHB patients, decreasing the sustained viral response (SVR) rates to PEG-IFN (143–145). Recent studies indicated that obesity attenuates and prolongs the type I IFN response that shows antiviral inefficacy. Obese individuals have a high level of blood leptin, resulting in the upregulation of the SOCS3. The elevated leptin and SOCS3 levels could reduce type I IFN response and cause other immune dysfunction associated with T cells and B cells in people with obesity (146).

The relationship between HBV infection and insulin resistance (IR) has not been completely elucidated, but it is reported that HBV infection was associated with increased IR (158). Researches illustrated that IR was associated with virological response to HBeAg-positive immune-reactive CHB patients' therapy with IFN-α mainly through downregulation of IFN-γ, TNF-α and multiple cytokines (147, 148). On the other hand, IFN-α was reported to be a catastrophic feature of the diabetic islet microenvironment, although the conclusion is controversial (159, 160). Therefore, the relationship of diabetes and IFN-α treatment is complicated and need further study.

Alcohol abuse causes rapid progression of liver disease in HBV infected patients and allows HBV to persist chronically mainly by increasing HBV replication, inducing oxidative stress, and suppressing the immune response. As reported, alcohol reduces the number of T-cells and changes the ratio of T-cell types, resulting in decreased T-cell activation and function, thereby impair IFN production (149). On the other hand, the combination of HBV and ethanol metabolites impairs IFN-γ-signaling through the JAK-STAT1 pathway (149). Moreover, alcohol and HBV synergistically promote hepatic steatosis (161), which has been reported as another negative host factor of IFN therapy against HBV infection.

Endogenous anti-IFN antibodies have been reported to be stimulated after the treatment with IFN. Multiple studies indicated that anti-IFN antibodies may negate the antiviral effects of IFN-α by neutralizing the effect (150–152). Anti-IFN was significantly more likely to develop in patients who received lower doses (2.5 or 5 MU/m²) of IFN than in those who received a higher dose (10 MU/m²) (151). Another study showed that the anti-IFN activity may negatively influence the effect of the IFN therapy on CHB patients at the early stages of the therapy, but the appearance of the anti-IFN activity at the end of a long-term IFN therapy has no significant influence on the treatment outcome (152).