Clinical Research Progress of Small Molecule Compounds Targeting Nrf2 for Treating Inflammation-Related Diseases

Studies have found that inflammation is a symptom of various diseases, such as coronavirus disease 2019 (COVID-19) and rheumatoid arthritis (RA); it is also the source of other diseases, such as Alzheimer’s disease (AD), Parkinson’s disease (PD), lupus erythematosus (LE), and liver damage. Nrf2 (nuclear factor erythroid 2-related factor 2) is an important multifunctional transcription factor in cells and plays a central regulatory role in cellular defense mechanisms. In recent years, several studies have found a strong association between the activation of Nrf2 and the fight against inflammation-related diseases. A number of small molecule compounds targeting Nrf2 have entered clinical research. This article reviews the research status of small molecule compounds that are in clinical trials for the treatment of COVID-19, rheumatoid arthritis, Alzheimer’s disease, Parkinson’s disease, lupus erythematosus, and liver injury.


Introduction to Nrf2 Function
Oxidative stress refers to the imbalance between oxidation and antioxidants and is caused by the production of reactive oxygen species (ROS) in the body, resulting in oxidative damage to tissue and cells. The Nrf2/Keap1 pathway is the principal protective response to oxidative and electrophilic stresses. Kelch-like ECH-associated protein 1 (Keap1) is a component of the Cullin 3 (CUL3)-based E3 ubiquitin ligase complex and controls the stability and accumulation of Nrf2 [1][2][3][4][5][6][7]. Normally, Nrf2 exists in the cytoplasm under the regulation of Keap1 and maintains low activity in a normal physiological state. When cells are stimulated by oxidative stress, Nrf2 detaches from Keap1 and translocates into the nucleus to form heterodimers with musculoaponeurotic fibrosarcoma (MAF), bind antioxidant response element (ARE), and activate the expression of Nrf2 target genes (Phase II detoxification enzymes and antioxidant enzyme genes), such as heme-oxigenase-1(HMOX-1), NAD (P) H-quinone oxidoreductase 1 (NQO1) and glutamate cysteine ligase (GCL), glutathione S-transferase(GST), superoxide dismutase (SOD), γ-glutamyl cysteine synthetase (γ-GCS), glutathione peroxidase (GSH-Px), γ-glutamyl cysteine synthetase catalytic subunit(GCLC), γ-glutamyl cysteine synthetase modifier subunit(GCLM), etc. [8,9]. The functions of the proteins they encode are as follows: HO-1 is encoded by the HMOX-1 gene, which catalyzes the decomposition of heme with cytochrome P450 to produce biliverdin, etc., and then biliverdin is converted into bilirubin. Both biliverdin and bilirubin have antioxidant and immunomodulatory properties [10]. NQO1 protects cells from the harmful effects of quinone redox cycling [11]. GCL consists of GCLC and GCLM and is the rate-limiting enzyme in the glutathione biosynthetic pathway. GST mainly catalyzes the covalent combination of various chemicals and their metabolites with the sulfhydryl group of glutathione (GSH), making electrophilic compounds into hydrophilic substances, which are easy to excrete [12,13]. SOD catalytically converts the superoxide radical to hydrogen peroxide (H 2 O 2 ), constituting the first line of defense against oxidative stress [14]. γ-GCS catalyzes the rate-limiting biosynthesis of GSH, an abundant physiological antioxidant that plays important roles in regulating oxidative stress. GSH-Px specifically catalyzes the reaction of GSH with ROS, thereby protecting cells from ROS damage [15,16]. Nuclear factor kappa B (NF-kB) is closely related to the regulation of inflammation by participating in the activation of genes encoding proinflammatory cytokines, growth factors, and inducible enzymes, such as interleukin 1 beta (IL-1β), interleukin 6 (IL-6), tumor necrosis factor alpha (TNF-α), and inducible nitric oxide synthase (iNOS). Nrf2 reduces inflammatory response by inhibiting the activity of NF-kB through the Nrf2-ARE pathway and by directly inhibiting the activity of NF-kB and the expression of proinflammatory cytokine genes ( Figure 1) [17]. Numerous studies have shown that Nrf2 and NF-kB play important roles in regulating cancer responses to chemotherapy [18,19] and the immune/inflammatory cancer microenvironment in almost all types of cancer [20].

Nrf2 and Coronavirus Disease 2019 (COVID-19)
COVID-19 is a complex infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Clinical evidence has shown that the main symptoms of COVID-19 can include acute infection of the respiratory tract, as well as inflammatory reactions of multiple organs [21,22]. SARS-CoV-2 enters cells by first binding to angiotensin-converting enzyme 2 (ACE2), followed by cleavage of the virus spike protein by transmembrane protease serine 2 (TMPRSS2) [23]. Nrf2 located in the nucleus can directly inhibit the expression of ACE2 and TMPRSS2 on the cell surface to reduce SARS-CoV-2 entry into cells; it can also block the replication of the viral genome by mediating the production of type I interferons (IFN-I) by HO-1 [24][25][26]. Through these two mechanisms, the Nrf2 signaling pathway can effectively reduce SARS-CoV-2 infection (Figure 2). At present, seven Nrf2 agonists have entered clinical trials as COVID-19 treatments (Scheme 1 and Table 1). These compounds are discussed below.
Epigallocatechin gallate (EGCG) (1, 50% inhibitory concentration (IC 50 ) is 7.51 µM [27]) (3CL Protease (M pro ) (SARS-CoV-2) Assay Kit [BPS bioscience, https://bpsbioscience.com/ (accessed on 3 August 2022)] using the fluorescence method) is the main component of green tea polyphenols; it is a catechin monomer isolated from tea and is a flavanol compound. EGCG has entered phase 2/3 clinical trials. According to the strength of oxidation, the order of oxidation of EGCG and its three derivatives ((Epigalloeatechin)EGC, (Epieatechin gallate) ECG, (Epieatechin) EC) is: EGCG > EGC > ECG > EC [27][28][29][30][31][32][33][34][35]. Therefore, it can be speculated that the 3 ,4 ,5 -trihydroxyl group of the B ring in the EGCG structure is important for its antioxidant capacity. The gallic acid ester of the D ring also contributes to its antioxidant capacity, and studies have shown that the two may be involved in metal chelation. Multiple phenolic hydroxyl groups endow EGCG with robust antioxidant activity, high hydrophilicity, and active properties, so its stability should be fully considered when designing a drug.
Sulforaphane (2, IC 50 is 2.4 µM ) (quantification of viral RNA from SARS-CoV-2infected human intestinal Caco-2 cells treated with SFN using qPCR) is produced through the hydrolysis of glucosinolates, which are found in cruciferous vegetables, such as cabbage and radishes [36][37][38][39][40][41][42][43]. Sulforaphane has entered phase 2 clinical trials. The bioavailability of sulforaphane is approximately 80%. However, its disadvantages are its instigation of strong irritation, volatility, and sensitivity to temperature and pH. The isothiocyanate group is the pharmacophore of sulforaphane. Converting methylsulfinyl to an acetyl group or N-methylformamide reduces its antioxidant capacity; however, when converted to squaramide, its antioxidant capacity is 25 times greater than that of sulforaphane [44]. Therefore, it is possible to modify the structure of the butyl carbon chain of sulforaphane to make it less irritating and more stable.
Resveratrol (3,IC 50 is 0.98 µM [45]) (detection of the inhibitory activity of resveratrol on COX-2 enzymes using a COX-2 inhibitor screening kit using the fluorescence method) is found in plants such as blackberries, peanuts, and grapes and has good anti-inflammatory and anti-SARS-CoV-2 activities as an Nrf2 agonist [45][46][47][48][49]; this compound has entered phase 3 clinical trials. The resveratrol-ibuprofen combination, in which the hydroxyl group on the B ring is monosubstituted, has a more significant anti-inflammatory effect than either compound given alone [50]. Pei Ling et al. once discussed the relationship between antioxidation and chemical structure of resveratrol and its analogues (Scheme 1) and put forward the concept of hydrogen-donating ability. Using quantum chemical calculations based on the density functional theory (DFT), they calculated the hydrogen-donating ability of these compounds. The order of these compounds is C > D > B > G > E > A > F, and the antioxidation of these compounds is positively correlated with their hydrogen-donating ability [51]. Dimethyl fumarate (DMF) (4, IC 50 is 9.30 µM [52]) (detection of the inhibitory activity of DMF on IL-6 produced by lipopolysaccharide-induced dTHP-1 cell line using IL-6 commercial kits (Perkin Elmer) and the fluorescence method) is a US Food and Drug Administration (FDA)-approved synthetic drug used as an anti-inflammatory therapeutic for multiple sclerosis (MS) via Nrf2 inhibition of pathogenic inflammation [53,54], and it is currently in phase 2/3 clinical trials. Isosorbide di-(methyl fumarate) (IDMF), with a central isosorbide moiety and two methyl fumarate groups, can partially replicate DMF activity and is nonirritating and nonsensitizing when applied to the skin [55]. When the carboxyl groups at both ends of the DMF structure were changed to 4-chlorophenyl ester, Ar-NH-, and Ar-CH2-NH-, anti-inflammatory efficacy was greatly improved [56]. Based on the above, we preliminarily hypothesize that the intermediate chain ketone-ene-ketone structure of DMF may be an essential group for its activity and that the carboxyl groups at both ends can be transformed to synergistically improve the detoxification of the compound.  Children below 5 years: 2 pills of medicine Nanocurcumin 6C once a week for first 2 months, followed by once in 2 weeks. Individuals 5 years or above: 4 pills of medicine Nanocurcumin 6C once a week for first 2 months, followed by once in 2 weeks. ArtemiC is an oromucosal medical spray composed ofartemisinin (6 mg/mL), curcumin (20 mg/mL), frankincense (15 mg/mL) and vitamin C (60 mg/mL); spray administration two times a day on days 1 and 2. Each dose contains 1 mL (10 puffs/pushes on the spray bottle), total daily dose 2 mL (20 puffs/pushes on the spray bottle). The total treatment is 40 puffs over two days.   Two curcumin-piperine capsules (500 mg curcumin + 5 mg piperine) will be given daily for 2 weeks after lunch and dinner. ) (Researchers pulsed bone-marrow-derived dendritic cells (BMDCs) for 1 h with curcumin before stimulation with the TLR7 ligand R837 followed by ATP to investigate IL-1β production.) is a diketone compound extracted from the rhizomes of plants in the Zingiberaceae and Araceae families [54,58,59]. It is in phase 4 clinical trials. The unsaturated carbon chain and hydroxyl group on the benzene ring of curcumin are extremely important for its anti-inflammatory activity. The alkoxy group next to the phenol group and the benzene ring substituted by the strong electron withdrawing group of the ortho-diphenol hydroxyl group can increase its anti-inflammatory ability [60,61]. The hydrophobicity and rapid metabolism of curcumin lead to poor bioavailability. Some studies have structurally modified the phenolic hydroxyl groups at both ends to transform them into ether, which effectively slowed the metabolism of the compound [62].
Fluoxetine (6, IC 50 is 10 µM [63]) (Different concentrations of fluoxetine were added to Vero E6 cell cultures along with SARS-CoV-2, and the levels of infectious particles in culture supernatants were detected by incubation.) is a synthetic drug that was first approved in Belgium in 1986 for the treatment of depression. It is one of the few classic clinical drugs with Nrf2 agonistic effects [63][64][65][66]. It has entered phase 3 clinical trials. There is almost no anti-inflammatory activity when the methylamino group in fluoxetine is replaced by pyrrolidine, imidazole or piperidine, but there is equivalent activity when replaced by morpholine, piperazine, or N-methylpiperazine [67]. In the case of removing the trifluoromethyl benzene ring and replacing the methylamino group with morpholine, there is no anti-inflammatory activity when the ether bond is replaced by a hydroxyl group and an oxime group; when it is replaced by a ketone group, the activity is comparable to that of fluoxetine. There is increased anti-inflammatory activity after the introduction of trifluoromethyl to the carbon [68]. Therefore, it can be speculated that fluoxetine, methylamino groups, trifluoro-methylbenzene rings, and ether linkage are the key groups that affect the activity of fluoxetine analogues.
Bardoxolone methyl (7, IC 50 is 5.81 µM [69]) (detection of the inhibitory activity of Bardoxolone methyl on SARS-CoV-2 3CLpro with Thr-Ser-Ala-Val Leu-Gln-pNA-substrate by using absorbance at 390 nm) is a semisynthetic pentacyclic triterpenoid derived from oleanolic acid [69][70][71]. It has entered phase 2 clinical trials. Suqing Zheng et al. synthesized a series of monocyclic cyanoketene compounds and tested their anti-inflammatory ability. The study showed that the pharmacophore in semisynthetic pentacyclic triterpenoids is not pentacyclic triterpenoid and has nonenolized cyanoketenes rather than a tricyclic skeleton [72]. It is speculated that the A-rings are necessary for the anti-inflammatory activity of bardoxolone methyl. They function as Michael receptors, and the single-ring structure is more potent than the penta-ring structure. One study showed that removal of C-24 at the C-4 position of the A ring led to higher biological activity and that transforming methyl 28-carboxylate into ethylamide or trifluoroethylamide improved drug delivery to the brain [73,74]. Thus, the structural modification of C-28 is expected to alter its pharmacokinetic properties.

Nrf2 and Rheumatoid Arthritis
Rheumatoid arthritis (RA) is a chronic autoimmune disease of unknown etiology and affects approximately 0.5-1.0% of the world's population. It often presents with joint involvement, synovitis, and intra-articular cartilage damage [75,76]. It is thought that the etiology of RA is closely related to one's living environment, genetics, immunity, and other factors. Individuals with genetic factors are affected by their living environment, stress, and other factors, which induce abnormal responses in the innate and adaptive immune systems, leading to the destruction of immune tolerance and thus stimulating an inflammatory response [77,78]. The main pathological feature of RA is inflammation leading to articular cartilage damage caused by cartilage degradation. Many studies have shown that Nrf2 activation is a promising method for the treatment of RA [79]. The Kelch-Nrf2/ARE signal transduction pathway can have beneficial anti-inflammatory and antioxidant effects and can regulate oxidative stress in RA. At its core, increased Nrf2 activity can regulate mitochondrial function and limit the production of mitochondrial ROS after activation of this pathway [80] (Figure 3). At present, two Nrf2 agonists have entered clinical research for rheumatoid arthritis (Scheme 1 and Table 2).

Nrf2 and Alzheimer's Disease
Senile plaques formed through the accumulation of β-amyloid(Aβ) and neurofibrillary tangles caused by hyperphosphorylation of tau protein are important pathological features of AD [93]. AD affects more than 50 million people. There are various pathogenic hypotheses of AD, such as the cholinergic hypothesis, the Aβ toxicity hypothesis, the tau protein hypothesis, and the inflammation hypothesis, but the pathogenesis of AD still must be elucidated [94]. A recent experiment showed that Chlamydia pneumoniae infection is closely related to AD pathogenesis. Chlamydia pneumoniae was shown to enter the nasal cavity of mice and rapidly infect the olfactory and trigeminal nerves, which connect to the brain through the olfactory bulb and brain stem, respectively. Microglia and astrocytes (macrophages of the central nervous system (CNS)) can respond to and engulf bacteria. However, Chlamydia pneumoniae can evade destruction by phagocytes and infect glial cells by forming inclusion bodies in these cells. Following infection, activated microglia and astrocytes secrete proinflammatory cytokines, including IL-1β, TNFα, and IL-6, which are neurotoxic and directly increase Aβ production by activating β-site amyloid-precursorcleaving enzyme (BACE). On one hand, activated microglia reduce the accumulation of Aβ in the brain by increasing their phagocytosis, clearance, and degradation. On the other hand, the continuous activation of microglia caused by their binding to Aβ can increase the production of inflammatory mediators, which further amplifies the neuroinflammatory response, leading to chronic inflammation and AD [95][96][97][98] (Figure 4).  In animal models of AD, Nrf2 inhibits its expression by binding to AREs in the BACE promoter and inhibits Aβ production. It can also induce nuclear dot protein 52 (NDP52) by binding to AREs in the NDP52 promoter, thereby reducing p-tau levels in AD [99][100][101]. Therefore, the activation of Nrf2 by drug intervention may play a positive role in treating AD patients.

Nrf2 and Parkinson's Disease
Parkinson's disease (PD) is a chronic progressive nervous system disease. In late-stage PD, extreme tremors, motor retardation, muscle stiffness, and loss of balance occur [117]. In sporadic and familial PD, α-synuclein(α-syn) aggregates into Lewy bodies and Lewy neurites, which are cytotoxic to dopaminergic neurons and can lead to mitosis and enhance mitochondrial autophagy [118]. The increase in dopamine may affect mitochondrial function, increase ROS levels, affect Nrf2 activity, alter the response to antioxidant damage [119][120][121], and promote the progressive production and accumulation of Aβ [122]. These effects lead to dysregulated cellular function. However, Nrf2 activation can neutralize ROS, inhibit inflammatory processes, and restore cellular redox balance [123][124][125][126][127]. In PD, there are decreased protein expression levels of phosphatase and tensin homolog (PTEN)-induced kinase (PINK) and Parkin protein; the decreases in these proteins affect mitochondrial function, induce depolarization and fragmentation and reduce adenosine triphosphate (ATP) concentrations ( Figure 5) [128]. These changes will affect synaptic function, leading to neurodegeneration and cognitive impairment [124][125][126][127].The Nrf2 upregulation induced by antioxidant therapy was shown to enhance thioredoxin-1(TrX-1), inhibit the formation of nucleotide-binding domain leucine-rich repeat-related (NLR) family pyrin domain-containing 3 (NLRP3) inflammatory bodies and improve neuronal apoptosis in amyloid precursor protein plus presenilin-1 (APP/PS1) mice [129]. Although some mechanisms are not fully understood, Nrf2 can be considered a useful therapeutic target for PD [130].     Table 4).

Scheme 2. Structures of Vitamin D3.
Vitamin D3 (8, IC 50 is 2.1 µM [131]) (the ability of VD3 in C3H10T1/2 fibroblasts to down-regulate gli1mrna expression in a dose-dependent manner) is an important regulator of bone metabolism and calcium and phosphorus balance. It is converted from 1α-hydroxylase to its active metabolite 1,25(OH)2D and is currently in phase 4 clinical trials [132,133]. The 1-hydroxy group and the 10-position exomethylene group play important roles in maintaining the activity of the compound. Most research has focused on the modification of side chains and the A ring [134]. Compound (2) can cross the blood-brain barrier. The treatment with 0.1% glucoraphanin pellets preserved dopaminergic neurons from neurodegeneration [135,136]. Currently, in phase 2 clinical trials, Compound (1) (20 µmol/L) acts by upregulating antioxidase activity [137], and it can effectively scavenge H2O2 [138]. EGCG is currently in phase 2 clinical trials for the treatment of PD. Monoamine oxidase (MAO) regulates the local levels of neurotransmitters such as dopamine, norepinephrine and serotonin, and (3) has a selective inhibitory effect on MAO-A [139]. Compound (3) is currently in phase 1 clinical trials.

Nrf2 and Lupus Erythematosus
Systemic lupus erythematosus (SLE) is a chronic disease characterized by the loss of immune tolerance. SLE has a variety of clinical manifestations, the main sign of which is the production of autoantibodies that cause tissue damage [140]. Toll-like receptor 9 (TLR9) is an important bridge linking innate and adaptive immunity. For example, when the body is subjected to specific external stimuli, TLR9 activates the NF-κB pathway, leading to inflammation. T helper type 17 (Th17) cells are major proinflammatory T cells involved in the regulation of lupus nephritis (LN) through multiple mechanisms.
Signal transducer and activator of transcription 3 (STAT3) directly regulates interleukin-17 (IL-17) expression and suppresses cytokine signaling 3 (Socs3), which negatively regulates Th17 differentiation by downregulating STAT3 phosphorylation. Nrf2 inhibits Th17 differentiation and reduces STAT3 phosphorylation by upregulating Socs3 expression ( Figure 6) [141]. SLE can affect bone metabolism and serum electrolysis through renal impairment and by disturbing endocrine homeostasis [142]. In general, dysimmunity, oxidative stress, and inflammation are the key pathogenic features of SLE and LN [143,144]. Preventing SLE development in humans might be facilitated by activating the Nrf2 pathway and applying other antioxidant therapies. Three Nrf2 agonists have entered clinical research trials as a treatment for SLE (Schemes 1-3 and Table 5).  Compound (5) can inhibit inflammatory pathways, neutralize free radicals, and inhibit ROS production [145,146]. It is currently in phase 2 clinical trials as an immunomodulator for the treatment of SLE. Clinical trials of (8) for the treatment of SLE are in phase 2.
β-Aminoarteether maleate (SM934) (9, IC 50 is 1.24 µM [147]) (immunosuppression method of spleen cell proliferation induced by Con and LPS) is a water-soluble derivative of artemisinin. SM934 can inhibit TLR7/9 expression [148,149], renal antibody production, and the accumulation of inflammatory cytokines [150]. This compound has entered phase 2 clinical trials. The peroxyl bridge of SM934 is the key group enabling its functionality. The aminoethyl group in the structure increases its water solubility, reduces toxicity and side effects, and enhances efficacy [151].

Nrf2 and Liver Injury
The liver is the largest digestive gland in the human body. It has basic functions such as secreting bile, breaking down sugars and storing glycogen, detoxification, phagocytosis, and defense. Oxidative stress caused by drugs, viruses, alcohol, and other factors is the main cause of liver damage, which can further aggravate drug-induced liver damage, fatty liver, viral hepatitis, autoimmune liver disease, liver fibrosis, and primary liver cancer. The Nrf2 pathway is widely involved in many aspects of the body's defense against oxidative stress, such as detoxification, anti-inflammatory processes, and the regulation of cellular metabolism [152][153][154][155][156].

Role of Nrf2 in Nonalcoholic Fatty Liver Disease (NAFLD)
NAFLD is the most common chronic liver disease worldwide and is mainly characterized by a clinicopathological syndrome of excessive deposition of fat in liver cells. It can be caused by excessive alcohol intake and other liver-damaging factors. There are many structural and functional abnormalities in the mitochondria of NAFLD patients [157] that lead to the overproduction of ROS and cytokines. This triggers lipid peroxidation, and the generated ROS and lipid peroxidation products further damage mitochondrial function [158] in a vicious cycle (Figure 7). There is currently no definitive drug treatment for NAFLD.  Table 6).  Liraglutide increases the concentrations of Sestrin2 and Nrf2 and improves obesityrelated NAFLD [159]. It is currently in phase 4 clinical trials. Resveratrol (3) was shown to attenuate methylation of the Nrf2 promoter in the liver of mice fed a high-fat diet (HFD) and attenuated NAFLD through epigenetic modification of Nrf2 signaling [160]. This compound is now in phase 2/3 clinical trials. Curcumin (5) treatment significantly alleviated liver steatosis in mice fed an HFD, reversed abnormal serum biochemical parameters, and increased the metabolic capacity to effectively restore the Nrf2-FXR-LXR pathway [161]. Curcumin is currently in phase 2/3 clinical trials.
In addition, a variety of natural Nrf2 activators such as aucubin [162], ginkgolide B [163], and limonin [164] can also alleviate NAFLD by regulating lipid metabolism and oxidative stress in hepatocytes. However, these compounds require further clinical investigation.

Role of Nrf2 in Viral Hepatitis
The core protein and nonstructural protein 5A (NS5A) of hepatitis C virus (HCV) cause mitochondrial dysfunction in hepatocytes, and the resulting expression of cytochrome P450 2E1 (CYP2E1) and NADPH-oxidase (NOX) produces a large amount of ROS [165,166]. HCV core protein and NS5A can also activate Nrf2 to alleviate HCV [167], while HCV can cause MAF to translocate and bind to extranuclear nonstructural protein 3 (NS3), which then binds to Nrf2 in the cytoplasm, preventing Nrf2 from entering the nucleus [168][169][170]. The hepatitis B x protein (HBx) of hepatitis B virus (HBV) can alter a variety of mitochondriarelated functions and is an important cause of mitochondrial dysfunction [171]. HBV can enhance the interaction between p62 and Keap1 to form the HBx-p62-Keap1 complex in the cytoplasm, thereby promoting Nrf2 expression [172] (Figure 8). Currently, silymarin is the only Nrf2-related compound that has entered clinical trials for the treatment of HCV (Scheme 4 and Table 7). Silymarin (10, IC 50 is 1.70 µM [173]) (measure metabolite concentrations with a fluorescence spectrometer at an excitation wavelength of 409 nm and an emission wavelength of 530 nm; the positive control runs on the same plate), refers to a class of flavonoid lignans extracted from the fruit and seeds of the Compositae herb, Silybum marianum; these lignans contain dihydroflavonols and phenylpropanoid derivatives [174]. Silymarin has entered phase 3 clinical trials [175]. Multiple phenolic hydroxyl and methoxy groups endow silymarin with good antioxidant activity. The introduction of methoxy groups on the B ring and the E ring improves the ability of silymarin to scavenge superoxide free radicals and 2,2-diphenyl-1-picrylhydrazyl (DPPH) free radicals [176]. Esterification of the 3-or 23-hydroxyl of silymarin significantly improves its solubility, but its biological activity is reduced [177].

Role of Nrf2 in Primary Biliary Cholangitis
Primary biliary cholangitis (PBC) is an organ-specific chronic and cholestatic autoimmune liver disease. Nrf2 protein concentrations are elevated in PBC patients, but Nrf2 gene expression is significantly decreased, and Keap1 and p62 protein concentrations are significantly increased [178,179]. Aberrant Nrf2/Keap1 system integrity may affect the self-defense mechanism against oxidative stress in PBC.
Currently, only one Nrf2-related compound has entered clinical trials for PBC (Scheme 5 and Table 8). Ursodeoxycholic acid (11, IC 50 is 30.82 µM [180]) (structure of primary and secondary bile acids as well as corresponding potency in differential scanning fluorimetry binding and cell rounding assays) is a bile acid compound [180]. It is the only drug approved by the US FDA for the treatment of PBC, and it is still the first-line drug for the treatment of PBC. It is currently in phase 4 clinical trials. The 3 and 7 phenolic hydroxyls endow ursodeoxycholic acid with antioxidant activity; this compound also enhances Nrf2 activation in hepatocytes of PBC patients and increases thioredoxin (TRX) and thioredoxin reductase 1 (TrxR1) proteins, thereby relieving PBC [181]. Ursodeoxycholic acid derivatives modified by glycine at position 24 have strong antioxidant effects and fewer toxic side effects than the parent compound [182,183]. The 24-position carboxylic acid is substituted by a heterocycle to obtain a ursodeoxycholic acid derivative that can selectively deliver NO to the liver, significantly increase the concentration of cyclic guanosine 3 , 5 -monophosphate (cGMP) in the liver, and effectively inhibit various inflammatory factors, such as interleukin and tumor necrosis factor [184]. This derivative has a good therapeutic effect for the treatment of liver damage and the associated inflammation.

Role of Nrf2 in Liver Fibrosis
Globally, the number of people with liver fibrosis is expected to increase from 740 million in 2017 to 821 million in 2022. An important cause of liver fibrosis is the activation of hepatic stellate cells (HSC). Ruart et al. found that damage to sinusoidal endothelial cells during acute liver injury aggravates oxidative stress and activates stellate cells to promote liver fibrosis; furthermore, autophagy-impaired liver sinusoidal endothelial cells (LSEC) can cause ROS accumulation and elevated p62 levels, which activates the upregulation of Nrf2 and its target genes [185].
At present, only one Nrf2-related compound has entered clinical research for the treatment of hepatic fibrosis (Scheme 6 and Table 9). Candesartan (12, IC 50 is 3.59 µM [186]) (Immunofluorescence was conducted with mouse anti-OC43 N protein antibody and followed by Alexa Flour 488 and DAPI. The IC 50 was calculated using automated image analysis software), is an angiotensin II (Ang II) receptor antagonist. Recent studies have found that candesartan's antihepatic fibrosis effect occurs partly through the activation of Nrf2 and its downstream target genes [187]. Candesartan is in phase 3 clinical trials. One study found that the substitution of 2-ethoxy increases its antioxidant activity, and the substitution of the carboxyl group at the 4-position of the benzene ring increases its water solubility and improves its pharmacokinetic properties.  In addition, sitagliptin [188,189], liraglutide [190], and mulberrin [191] can reduce aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels in mouse serum and alleviate stellate cell activation and liver fibrosis. These compounds are currently being investigated.

Conclusions
In conclusion, 12 of small molecule compounds targeting Nrf2 have entered clinical research for the treatment of inflammation-related diseases. According to different sources, these compounds can be divided into natural products and repurposed drugs.
Compounds 1, 2, 3, 5, 7, 9 and 10 are the chemical constituents of natural plants or their structural modifications. Among them, EGCG (1) and sulforaphane (2) are in clinical trials for the treatment of COVID-19, Alzheimer's disease, and Parkinson's disease; resveratrol (3) is in clinical trials for the treatment of COVID-19, Alzheimer's disease, Parkinson's disease, and liver injury; curcumin (5) is in clinical trials for the treatment of COVID-19 and rheumatoid arthritis; oleanolic acid derivatives (7) are in clinical trials for the treatment of COVID-19; artemisinin derivative SM934 (9) is in clinical trials for the treatment of lupus erythematosus. Silymarin (10) is in clinical trials for the treatment of viral hepatitis. Compounds 4, 6, 8 and 12 are the repurposed drugs. Dimethyl fumarate (4) is used in the treatment of multiple sclerosis (MS) in the United States, Europe, and other countries. Now, it is in clinical trials for the treatment of COVID-19 and rheumatoid arthritis; the antidepressant drug fluoxetine (6) is in clinical trials for the treatment of COVID-19. Vitamin D3 (8) is in clinical trials for the treatment of Parkinson's disease and lupus erythematosus; the bile acid ursodeoxycholic acid (11), a drug used for the treatment of gallstone diseases, is now used in clinical trials for the treatment of autoimmune liver disease. Candesartan (12), a lipid-lowering drug, is in the clinical research stage for liver fibrosis.
It should be noted that there is now some genuine structural information about the NRF-2/KEAP system from crystallographic studies carried out in China which identifies a nucleophilic addition of a thiol on the protein target to a Michael acceptor in the drug molecule [192]. Several of the compounds described in this review contain either Michael acceptors or other electrophilic groups to which a thiol would add. Furthermore, the possibility of blocking the NRF-2/KEAP interaction with small molecule drugs has been discussed in some detail by the Strathclyde group led by Harnett [193]. These research results of the specific interaction between the small molecule drugs and target proteins provide a valuable basis for the further design of new drugs targeting Nrf2.
Drugs with various structural types that target Nrf2 have achieved promising clinical experimental results, which confirms the good drug ability of these compounds that target Nrf2. The therapeutic areas involved are diverse, and clinical drugs are scarce, so the development of related new drugs is of great value and significance. However, on the whole, the total number of compounds entering clinical research in this field is small, the structural types are not sufficiently rich, and the IC 50 values of these compounds that have entered the clinical stage are all several to tens of µM, and further improvement of the activity is needed. Therefore, with the help of computer-supported drug design methods that optimize the structural characteristics of target proteins and by focusing on natural product components and their structural modifications, the design and development of highly active and selective Nrf2 agonists will provide the possibility for the discovery of novel drug molecules in the future.