Introduction
Liver disorders are one of the major health care concerns worldwide1 mostly because of chronic liver diseases such as nonalcoholic/metabolically associated fatty liver disease (NAFLD/MAFLD), alcoholic liver disease (ALD) and viral hepatitis.2,3 In addition, acute liver disease can be associated with high mortality most frequently caused by drug associated liver injury, especially in Western countries.4 Autoimmune hepatitis (AIH) is also a risk factor for the development of liver cirrhosis and hepatocellular carcinoma.5 Therefore, exploring new therapeutic options for treatment of liver disease has become increasingly important in the past couple of decades. Given that inflammation, whether acute or chronic, and the production of proinflammatory cytokines play a key role in the progression of liver disease, it should not come as a surprise that the spotlight of recent pharmacotherapeutic research has been directed to immune processes and the development of molecules with immunomodulatory properties.6–8 In this review, we have focused on the interleukin (IL) 1 cytokine superfamily as an important player in the development of liver damage regardless of etiology.9
Pathobiological effects of the IL1 family
The IL1 superfamily consists of 11 members of IL1 superfamily cytokines and 10 members of IL1 superfamily receptors and is divided into three subfamilies, the IL1 subfamily (IL1α, IL1β, and IL33, and IL1Ra), the IL18 subfamily (IL18 and IL37), and the IL36 subfamily (IL36α, β, γ, and IL38). They are primarily associated with inflammation injury; yet some of the members also improve defensive mechanisms and build immune response to infection. However, most of the IL1 family have nonspecific features. These cytokines may function as pro-inflammatory (IL1α, IL1β, and IL33) or anti-inflammatory (IL1Ra, IL36Ra, IL37, or IL38) cytokines; IL18 can act as either a pro- or anti-inflammatory cytokine10–12 depending on the microenvironment. IL1 receptors (ILRs) consist of ligand binding subunits IL1R1, ST2, IL18Rα and IL36R, signaling subunits IL1RAcP, IL18Rβ, and a single immunoglobulin IL1-related receptor (SIGIRRs), alternatively named TIR8. SIGIRR/TIR8 has a regulatory function and is considered as an orphan receptor.10 ILRs are comprised of two subunits, an extracellular immunoglobulin-like domains and an intracellular Toll/Interleukin1R (TIR) domain responsible for oligomerization of IL1R subunits after cell stimulation. Subsequently, MyD88 activates nuclear factor-kappa B (NF-κB) and mitogen-activated protein kinases) such as p38 and JNK pathways eliciting inflammation (Fig. 1).10
IL1 superfamily members lack a signaling peptide for excretion. For instance, activation of IL1β, IL18, and IL37 depends on caspase-1, which is triggered by the NOD-like receptor family, pyrin domain containing 3 (NLRP3)-inflammasome, converting procaspase-1 into the active caspase.12 In contrast, IL1α is a biologically active precursor and is activated in liver necrosis.11 IL1α and IL33 have dual functions. They not only prevent inflammation induced by proapoptotic signals, but also act as proinflammatory factors following tissue necrosis, as part of a damage-associated molecular pattern or DAMP.11
NAFLD
NAFLD is a serious public health issue because of its high incidence and increased risk of its progression to liver cirrhosis and hepatocellular carcinoma.13 NAFLD consists of nonalcoholic fatty liver (NAFL), characterized by accumulation of triglycerides in the absence of inflammation. Nonalcoholic steatohepatitis (NASH), a more severe form of NAFLD characterized by cell damage and infiltration by inflammatory cells.14 In NAFL, liver damage is usually absent or insignificant because inflammation and pyroptosis are absent or mild. In NASH, stage inflammation and pyroptosis are more serious, and the damage is significant.15 In recent years, researchers have been increasingly interested in the association of NAFLD with inflammasomes, mostly NLRP3, and to some extent NLRP1, which is less understood, and pyroptosis.
NLRP3 inflammasomes are associated with various pathological events in different organs including fibrosis in the liver, heart, kidneys, lungs, and others.16 In the liver, activation of NLRP3 inflammasomes stimulates activation of caspase-1, leading to pyroptosis.15 NLRP3 recognizes microbial and non-microbial signals of cell damage, and in NAFLD it is activated by lipotoxic ceramides17 and triggers aseptic inflammation18 by transferring the signal to apoptotic-related spot protein to activate caspase-1, a key processing mediator of interleukin 1 family of cytokines and gasdermin D (GSDMD) cleavage.12,19 GSDMD-N (cleaved GSDMD) then regulates adipogenesis by activating the NF-κB signaling pathway and increases secretion of inflammatory cytokines.20
Pyroptosis is a form of programmed cell death, different form apoptosis and autophagy, triggered by proinflammatory signals, and dependent on inflammatory caspase-1 and caspases4, 5, and 11, with a series of inflammatory responses.15 It is characterized by the creation of membrane pores that dissipate ion gradients of the cells allowing influx of water, cell swelling, osmotic dissolution, and release of proinflammatory substances inside of the cell, including IL1β, IL18, IL33), IL37, high mobility group protein box-1, and heat shock protein .7,21–23
The involvement of NLRP3 inflammasome activation in the severity of NAFLD has been elucidated by numerous animal studies. Inflammation and fibrogenesis in liver damage, was reduced in NLRP3 knockout mice fed a choline-deficient amino acid diet, moreover arsenic trioxide induced pyroptosis by NLRP3 activation through cytoplasmic cathepsin that led to NAFLD development.24 MCC950, a selective inhibitor of NLRP3, significantly suppressed inflammation and fibrosis in NAFLD by reducing expression of caspase-1 and monocyte chemoattractant protein-1, IL1β and IL6 levels, and hindered migration of neutrophiles and macrophages in obese diabetic mice.25 Levels of IL33, also processed by NLRP3 inflammasomes, were increased in serum of mice fed a high-fat diet, and administration of IL33 to the mice attenuated hepatic steatosis but increased fibrosis.26,27 Anakinra, an IL1 receptor antagonist, as a treatment in type 2 diabetes patients, resulted in a significant decrease of inflammation and insulin resistance. In the treatment of ethanol-induced liver injury, it resulted in a significant reduction of hepatic inflammation, steatosis, and neutrophil infiltration. This raises the possibility of its potential use for treatment of NAFLD.28–30 The evidence is consistent with other reports that inhibition of NLRP3 inflammasomes and GSDMD significantly reduced inflammation and fibrosis by regulating pyroptosis pathways.31–34
Another inflammasome important for the development of NAFLD is NLRP1. It is activated in nonhematopoietic cells and interacts with caspase-1, caspase-5, and most likely with apoptosis-associated speck-like protein containing a C-terminal caspase recruitment domain to form an inflammasome that activates both ILβ and IL18,7,35,36 but with a preference for IL18, at least in insulin-responsive tissue like adipocytes, muscle, and liver.35 Although the effects of activation of this inflammasome are not completely clear in the development of NAFLD, it has been shown that IL18 has protective effects in animal models of NAFLD.37 However, that was not confirmed in type 2 diabetes patients or in obese children, in whom IL18 had the opposite correlation.17,38 Henao-Mejia et al.39 reported that inflammasomes and their effector protein IL18 negatively regulated NAFLD/NASH progression by modulation of the gut microbiota and gut leakage. In mouse models associated with inflammasome-deficiency, IL18 changed the configuration of gut microbiota in a way that exacerbated hepatic steatosis and inflammation through influx of TLR4 and TLR9 agonists into the portal circulation, activating TNF-α expression that driving NASH progression.39
Furthermore, the anti-inflammatory cytokine IL37, was found to cause increases of circulating adiponectin and insulin sensitivity in mice transgenic for human IL37 fed a high-fat diet and in mRNA expression in human adipose tissue was correlated with insulin sensitivity.40 In the same transgenic mice fed ethanol, IL37 expression was lower than in pair-fed transgenic mice with the same extent of liver damage. In patients with alcoholic steatohepatitis, IL37 levels were lower than they were in patients with NAFLD.41 It is important to note that no mouse homolog of IL37 has been described, and for that reason, only transgenic expression of human IL37 allows study its effects in a mouse model.
ALD
ALD includes acute and chronic forms that can progress to liver fibrosis or cirrhosis. Alcohol causes increased production of the proinflammatory cytokine ILβ through activation of the inflammasome NLRP3-caspase 1.42,43 In addition, microRNA-148a, which is responsible for the inhibition of NLRP3 inflammasomes is decreased by alcohol consumption through the transcriptional regulator forkhead box protein O1. A recently identified target molecule of microRNA -148a, thioredoxin-interacting protein, was found to be overexpressed during ALD-induced inflammation in the liver through NLRP3 inflammasome activation and pyroptosis.44 ILβ also triggers invariant natural killer T lymphocyte activation leading to polymorphonucleocyte invasion and further liver damage.45,46 At the same time, several DAMPs such as ATP and uric acid are produced by hepatocytes,47 further promoting liver damage. Development of new therapeutic options to block the IL1/IL1R signaling pathways seems reasonable. For now, anakinra and canakinumab, drugs used to treat other inflammatory diseases, but not liver disease, have that mechanism of action. Anakinra is an ILR antagonist with an excellent safety profile, and is used to treat adult rheumatoid arthritis by blocking the biologic activity of IL1.48 The results of a study that found blockage of IL1 signaling caused reduced liver inflammation and increased in liver regeneration in a mouse model of acute-on-chronic liver injury induced by ethanol also support the hypothesis.49 In another animal study, administration of IL1Ra led to inhibition of IL1β signaling by down-regulation of Caspase-1 activity and inflammasome activation, thus reducing liver steatosis, inflammation, and damage. Administration of anakinra, an antagonist of IL1α and β receptors was more effective than inhibition of IL1β alone.28
Considering that human studies are lacking, data from the Defeat Alcoholic Steatohepatitis (DASH) study, a multicenter, randomized, double-blind controlled trial are eagerly awaited. The primary objective is assessment of the safety and efficacy of a combination of an ILR1 antagonist, anakinra to suppress acute inflammation, pentoxifylline for hepatorenal syndrome prevention, and zinc sulfate compared with methylprednisolone, a standard of care in alcoholic steatohepatitis (ASH).29 The results of phase 2 trials demonstrating the superiority of combination therapy regarding the survival rate after 3 and 6 months compared with glucocorticoid therapy are encouraging.50 Other treatment options such as canakinumab, which targets IL1β and not IL1α seem to be less favorable compared with anakinra for treatment of liver disease.48
On the other hand, IL18 has shown a proinflammatory role in ALD by promoting inflammation and intestinal cell permeability in animal models.51 However, a study by Khanova et al.52 using RNA sequencing and proteomic analyses in a mouse binge-drinking model, showed that the CASP11/4- GSDMD pathway was associated with pyroptosis in ASH that was promoted by IL18 deficiency, indicating dual properties of IL18.52 Thus, depending on the microenvironment, IL18 has the potential to either promote or inhibit inflammation and liver damage, but studies in humans are lacking.
The IL1RL1 chain (also called ST2 or suppression of tumorigenicity 2, T1/ST2, or IL1-R4) is also a potential therapeutic target. IL33 is a soluble form of a decoy receptor shown to correlate with ALD severity in human patients.48 In the early stages of the disease, ST2 has a protective role mediated by NF-κB inhibition in liver macrophages. It is independent of IL33, as was shown in an animal model comparing alcohol-induced liver injury, inflammation, and hepatic macrophage activation in wild-type, IL33−/− and ST2−/− mice. However, in the same study, which included ALD patients, only individuals with severely decompensated ALD had increases in serum IL33 and ST2.53,54 Hence, ST2/IL33 potentially has a dual mode of action that is protective in the early stages of disease and damaging as liver injury and inflammation progress.
In a study investigating IL37 in humans and an animal model, IL37 transgenic mice had decreased expression of IL37 compared with pair-fed transgenic mice. Moreover, infusion of human recombinant IL37 improved liver inflammation in a mouse binge-drinking model of ALD. In addition, IL37 expression was compared in liver samples of NAFLD and ASH patients confirming, the anti-inflammatory activity of IL37 in ASH patients, as its expression was decreased when compared to NAFLD patients.41 Enhancing IL-37 action could present a possible therapeutic option in treating ALD.
Fibrosis
Hepatic fibrosis is a major characteristic of chronic inflammatory liver disease progression and a risk factor for development of hepatocellular carcinoma (HCC). Immunoregulatory mediators such as cytokines, including the IL-1 family, play an important role in fibrinogenesis. Activation of NOD-like receptor NLRP3 inflammasomes has been identified as important factor in hepatocyte pyroptosis, liver inflammation and fibrosis, which can initiate and facilitate progression of fibrosis.55 These findings propose blockade of NLRP3 pathway as a therapeutic target to reduce liver inflammation and fibrosis. IL-1 and its role in hepatic fibrosis has been extensively investigated. Gieling et al. conducted an in vivo study which found that IL-1 receptor-deficient mice exhibited decreased hepatic tissue damage and reduced fibrogenesis, indicating that IL-1 participates in the progression from liver injury to fibrosis.56 In a similar study, mice with steatosis induced by a high-fat diet, and deficient in either IL1α or IL1β had a significantly reduced transformation of steatosis to fibrosis. The result supports neutralizing IL1α and IL1β as a potential therapeutic option in the progression of liver fibrosis.57 Anakinra, an IL1R antagonist, has shown significant, beneficial modulation of liver inflammation and fibrosis in several in vivo studies.7,28,58 On the other hand, in animal and human studies, IL33 activated hepatic stellate cells and worsened fibrosis.59,60
HCC
The IL1 family participates in signaling pathways in tumorigenesis. The most extensively studied family members are IL1 and IL18. An epidemiological study in South Korea showed IL1β polymorphisms were associated with either increased or decreased HCC risk.61 IL1α is produced in hepatocytes damaged by reactive oxygen species, and promoted carcinogenesis in a mouse model of carcinogen-induced liver cancer. Targeting IL1R signaling may thus be a preventive or therapeutic option in HCC.62 Bermekimab, an IL1α-specific monoclonal antibody, was recently used in a phase III trial in treatment of metastatic colorectal cancer. The study showed no survival benefit of bermekimab, but cancer-associated cachexia was improved.63 There is strong evidence that the IL18/IL18R axis is a checkpoint in immunological processes regulating carcinogenesis.9 Absence of IL18 production leads to loss of antitumor activity, partially because of the absence of the FasL-dependent cytotoxicity of hepatic natural killer (NK) cells. Reduced production of IL18 is also associated with increased liver metastasis of colorectal cancer.64 Currently, there are several clinical trials targeting the IL18 signaling pathway, including recombinant IL18 and a monoclonal anti-IL18 neutralizing antibody.65 Elevated IL33 has been detected in HCC patients, and several animal model studies demonstrated antitumoral and antimetastatic activity of IL33. Some studies found decreased IL33 levels and its diminished effects as a protective factor in HCC, highlighting the need for further research of the mechanisms.66 Regarding IL37, current data suggests that it has antitumor activity in HCC, with strong evidence associating elevated hepatic levels with improved survival.67
Drug-induced liver injury
Acute liver injury is also mediated by IL1 superfamily members. In murine-model studies of IL1α and IL1β knockout mice, acute liver injury was diminished compared with wild-type mice.68 In addition, IL1869 and IL3370,71 promoted acetaminophen-induced liver injury, and IL36 exhibited a protective role by induction of CCL20, a protective chemokine.72 Furthermore, in animal studies, IL37 was shown to have a dual function, with protection through TNF-α inhibition, and destruction by increasing liver injury.73,74 In acetaminophen-induced liver injury, increased production of IL1β and IL18 by Kupffer cells have shown to induce IFN-γ and TNF-α secretion by Th1 and NK cells, resulting in acute drug-induced liver injury.69
Viral hepatitis (A, B, C)
The role of IL1 superfamily members in viral hepatitis-induced liver inflammation has been widely studied and documented. Chronic hepatitis B virus (HBV) and hepatitis C virus (HCV) infection can lead to liver fibrosis, cirrhosis, and HCC. In vitro studies in cell cultures have shown that monocyte-derived human macrophages, peripheral blood mononuclear cell – derived primary human macrophages, and Kupffer cells incubated with HCV demonstrated enhanced IL18 and/or IL1β production through mechanisms involving NF-κB signaling, caspase-1 activation, and NLRP3 inflammasomes. Strategies targeting those interleukins may offer new therapeutic options to reduce hepatic inflammation induced by HCV infection.75,76 Increased secretion of IL18 has also been observed in patients with hepatitis A virus (HAV) infection. A rare fulminant form of viral hepatitis has been reported in patients infection with HAV has been reported in patients with HAV infection and IL18 binding protein (IL18BP) deficiency. IL18BP acts as inhibitory ligand, and its absence has been associated with uncontrolled NK cell activation by IL18 resulting in hepatotoxicity, thus highlighting its potential in treatment and prevention of HAV induced acute liver.77 However, IL18-mediated stimulation of T cells, NK cells, and NKT cells leading to IFN-γ production has shown to significantly inhibit HBV replication, suggesting that IL18 has potential therapeutic value in HBV infected patients.78 Elevation of another member of the IL1 superfamily, IL33 has been observed in patients with HBV and HCV, especially in those with the most severe forms of hepatitis. It has been suggested that it increases liver inflammation through activation of monocytes, TNFα, IL6, and IL1β.79
AIH
AIH is a chronic inflammatory liver disease with poorly understood pathophysiological mechanisms. Studies in animal models with concanavalin A – induced hepatitis show rapid neosynthesis of IL33, which demonstrates protective activity , possibly due to induction of anti-apoptotic factors and recruitment of Treg which might be an important mechanism of liver repair.80 Another study suggests that NLRP3 inflammasome-induced IL1β production has an important role in the pathogenesis of concanavalin A – induced hepatitis, providing valuable findings regarding new therapeutic strategies for AIH by blocking NLRP3 inflammasome and IL1β.81
Therapeutic perspectives in liver disease
Current therapies for liver diseases are still quite modest, especially with regard to ALD or NAFLD. Therapeutic options targeting specific members of the IL1 superfamily seem to be very promising in the development of new drugs, and are summarized in Table 1.25,27–30,39–41,49,51,53,54,58–60,65,66,69,82–84 In addition, some cytokines/receptors have anti-inflammatory action (IL37), and others like IL1α, IL1β, and IL18 have pro-inflammatory activity, which could be very useful by enabling us to act in two opposite ways on inflammatory liver disease, depending on whether agonist or antagonist properties are activated. Also, several cytokines share the same receptor. Hence, by its stimulation or inhibition, it is possible to influence several inflammatory processes mediated by those molecules. For example, IL1RAcP is shared by IL1α, IL1β, IL33, and IL36. Interestingly, some family members, ST2 / IL33. Depending on the stage of damage or inflammation, it may have a protective effect in the early phase by ST2 activation. It may worsen inflammation and accelerate progression of fibrosis in the late phase by increased IL33 secretion. However, the story behind the IL1 superfamily is not simple, as the activation of IL1β and IL18 involves not only the classical NLRP3/inflammatory caspase-1 cytokine activation pathway, but also neutrophil serine proteases (NSPs), as shown in Figure 2, which explains why inhibition of NLRP3 and NLRP1 inflammasomes had low potency.85,86 Thus, in the future, development of therapeutic options focus on targeting all of the mediators involved in the activation signaling pathway of all or several pro-inflammatory cytokines, like alpha-1 antitrypsin an inhibitor of NSPs that protects against NAFLD development in animal models.6,87,88 However, a recombinant human IL1Ra, anakinra, has shown promising results in treating ALD, NAFLD in diabetic patients, and fibrosis. In addition, cankinumab, a human monoclonal anti-IL1β antibody demonstrated groundbreaking outcomes in the CANTOS trial, preventing atherosclerosis progression and reducing cardiovascular events.82 Given that administration of canakinumab in diabetic patients led to an improvement in hyperglycemia over a period of 1 year, it may have potential as an NAFLD treatment.83
Table 1Potential therapeutic options of IL1 family members in liver disease
| Target pathway | Drug | Disease | Study model | Main findings | Author, year |
|---|
| Inhibition of ILα and ILβ receptors | Anakinra | ALD | Mouse | Reduction of liver steatosis, inflammation, and damage | Iracheta-Vellve et al., 2017;49 Petrasek et al., 201228 |
| | | Human | Superiority of combination therapy (anakinra, pentoxyfilline, zinc) regarding survival rate after 3 and 6 months compared to glucocorticoid therapy | Dasarathy et al., 202029 |
| Anakinra | DMT2 | Human | Decrease in insulin resistance and inflammation | Larsen et al., 200730 |
| IL1Ra | Fibrosis | | IL1Ra worsened fibrosis in a CCl4 model, but it had protective effect in the BDL model | Meier et al., 201958 |
| Inhibition of ILβ receptor | Canakimumab | DMT2 | Human | Decrease in glucose levels during a period of 1 year | Everett et al., 201883 |
| | CVD | Human | Decrease in CVD events | Ridker et al., 201782 |
| Selective NLRP3 inflammasome inhibitor | MCC950 | NAFLD | Diabetic mouse | Suppression of inflammation and fibrosis in NAFLD | Qu et al., 201925 |
| NLRP3 inflammasome inhibitor | Sulforaphane | NAFLD | Mouse | Decrease in liver steatosis and inflammation | Yang et al., 201666 |
| IL18 | / | ALD, NAFLD | Mouse | Increasing gut permeability and gut leakage fueling inflammation | Gyongyosi et al., 2019;51 Henao-Mejia et al., 201239 |
| / | ALI | Mouse | Promoting APAP induced liver injury | Bachmann et al., 201869 |
| Monoclonal recombinant IL18/neutralizing anti-IL18 antibody | HCC | Mouse, Human | Inhibition of IL18 production increases anti-tumoral activity | Birbrair et al., 202065 |
| IL37 | Recombinant human anti-IL37 antibody | ALD | Mouse | Infusion of human recombinant IL37 improved liver inflammation | Grabherr et al., 201841 |
| / | ALD | Human | Decreased expression of IL37 in ALD patients compared to NAFLD | Grabherr et al., 201841 |
| / | NAFLD | IL37-transgenic mouse | Increased insulin sensitivity and adiponectin levels | Ballak et al., 201440 |
| / | | Humans | Insulin sensitivity correlated with expression of IL37 mRNA in fat tissue | |
| ST2/IL33 | / | ALD | Mouse, human | At early stages, ST2 has a protective role independent of IL33. In severe liver damage, IL33 increases liver injury | Wang et al., 2017;54 Sun et al., 201953 |
| | NAFLD | Mouse | IL33 diminished liver steatosis but worsened fibrosis | Gao et al., 201627 |
| / | Fibrosis | Mouse, human | Activation of HSC and induction of fibrosis | Kotsiou et al.,2018;59 Tan et al., 201860 |
| Recombinant IL233–hybrid cytokine with IL2 and IL33 properties | Possibly ALI, no studies available | Mouse | Prevents acute renal injury enhancing Treg activity | Stremska et al., 201784 |
In conclusion, most of the evidence presented in this mini review originates from preclinical studies, but evidence of the efficacy of these therapeutic options in humans is very scarce. Furthermore, the functions of all IL1 family members, including IL36 and IL38, are not fully understood. We are still a long way from using the potential therapeutic advantages of IL1 family members in routine clinical practice because of lack of clinical data, high cost, and limited availability. Thus, more studies of the function of these cytokines and whether they truly represent a valid therapeutic target are needed.
Abbreviations
- AIH:
autoimmune hepatitis
- ALD:
alcoholic liver disease
- ASH:
alcoholic steatohepatitis
- GSDMD:
gasdermin D
- GSDMD-N:
cleaved GSDMD
- HBV:
hepatitis B virus
- HCV:
hepatitis C virus
- IL:
interleukin
- ILR:
IL1 receptor
- NAFL:
nonalcoholic fatty liver
- NAFLD/MAFLD:
nonalcoholic/metabolically associated fatty liver disease
- NASH:
nonalcoholic steatohepatitis
- NF-κB:
nuclear factor-kappa B
- NK cells:
natural killer cells
- NLRP3:
NOD-like receptor family, pyrin domain containing 3
- NSP:
neutrophil serine protease
- SIGIRR:
single immunoglobulin IL1-related receptor
- TIR:
Toll/Interleukin1R
Declarations
Acknowledgement
This support of the Herman Lopata Chair in Hepatitis Research is gratefully acknowledged (GYW).
Funding
This research was funded by grant from Croatian Ministry of Science and Education dedicated to multi-institutional funding of scientific activity at the J.J. Strossmayer University of Osijek, Osijek, Croatia, grant number ZUP2018-90 (to Ines Bilić Ćurčić).
Conflict of interest
GYW has been an editor-in-chief of Journal of Clinical and Translational Hepatology since 2013. MS has been an editorial board member of Journal of Clinical and Translational Hepatology since 2013. The other authors have no conflict of interests related to this publication.
Authors’ contributions
Conceptualization (MS, RS), original draft preparation (IBC), writing, review and editing (IBC, TK, AP), figure generation (TK), supervision (MS, RS), funding acquisition (IBC), and editing for important intellectual content (GYW, AT).