Fibrosis in Mesothelioma: Potential Role of Lysyl Oxidases

Simple Summary Why do most drugs have limited effects in difficult-to-treat cancers such as mesothelioma? One explanation commonly given in response is that not enough drug can get into the tumor to generate an anti-tumor effect. Fibrosis is a common element of mesothelioma that causes the area around the cancer to stiffen. By itself, fibrosis makes breathing difficult and causes a poor quality of life, but it also acts as a barrier stopping drugs from getting into the tumor, and limits the immune system’s ability to detect and access the tumor to kill it. In this review, we discuss how a family of proteins called lysyl oxidases is associated with fibrosis in many disease settings, including solid tumors and mesothelioma, the current status of efforts to therapeutically target these proteins and how targeting this family may have therapeutic applicability in the treatment and management of mesothelioma. Abstract Immunotherapies (such as checkpoint inhibitors) and standard chemotherapies (such as cisplatin) have limitations in the successful treatment of malignant pleural mesothelioma (MPM). Fibrosis is the accumulation of collagen in the extracellular matrix (ECM) of tissues, making them denser than that of healthy tissues and thereby affecting drug delivery and immune cell infiltration. Moreover, fibrosis severely affects the patient’s breathing and quality of life. The production of collagen and its assembly is highly regulated by various enzymes such as lysyl oxidases. Many solid tumors aberrantly express the family of lysyl oxidases (LOX/LOXL). This review examines how LOX/LOXLs were found to be dysregulated in noncancerous and cancerous settings, discusses their roles in solid tumor fibrosis and pathogenesis and explores the role of fibrosis in the development and poor clinical outcomes of patients with MPM. We examine the current preclinical status of drugs targeting LOX/LOXLs and how the incorporation of such drugs may have therapeutic benefits in the treatment and management of patients with MPM.


Introduction
Malignant pleural mesothelioma (MPM) is an aggressive inflammatory cancer associated with exposure to asbestos [1]. Industrialized nations' use of asbestos has declined, but asbestos continues to be exported and used in developing nations. Despite its ban, environmental exposure is still widespread [1][2][3][4]. Early-stage MPM patients typically present with shortness of breath, coughing and chest pains, which is attributed to pleural thickening and pleural effusion. Pleural effusion is an accumulation of fluid in the space between the lungs and the chest wall that can be eased via thoracentesis (aspiration of fluid) or chemical pleurodesis (intrapleural administration of talc obliterating the pleural space). Computed tomography (CT) imaging is routinely performed to detect pleural fluid, present in 90% of MPM patients, although biopsies are required for a definitive diagnosis [5]. Pleural effusion is caused by an imbalance between pleural fluid production and an inadequacy of the lymphatic drainage system, which again limits respiration. As Table 1. Selected examples of known LOX/LOXL family links with various non-cancerous settings.

Idiopathic Pulmonary Fibrosis (IPF)
β-aminopropionitrile (BAPN) found to inhibit pulmonary fibrosis in a lung model of silicosis [34] LOX activity induced in bleomycin-induced lung fibrosis and alleviated by treatment with LOX/LOXL inhibitor [35][36][37] In IPF patients, elevated serum levels of LOXL2 are associated with an increased risk for disease progression. [38] LOX activity promotes the progress of EMT in a paraquat model of IPF. [39,40] A comparative analysis shows that LOX/LOXL2 are elevated in IPF fibroblasts, while LOXL2/3 activity is crucial for fibroblast-to-myofibroblast transition (FMT). [41] Elevated serum levels of LOXL2 are associated with rheumatoid arthritis -associated interstitial lung disease (RA-ILD). [42] Increased collagen fibril thickness in IPF versus non-IPF lung tissues is correlated with increased levels of LOXL1/LOXL2 protein, and a decrease in LOX protein expression. [43] Loss of LOXL1 activity prevents the development of fibrosis in a transforming growth factor-β1-induced model of pulmonary fibrosis. [44] In bleomycin-induced lung fibrosis, nuclear expression of LOXL2 appears to be a major element in the progression of lung fibrosis. [45] LOXL2 inhibitor induces collagen turnover in ex vivo lung explants from patients with IPF. [46] In a study comparing extracellular vesicles (EVs) from IPF-derived pulmonary fibroblast cell lines versus normal pulmonary fibroblast cell lines for differentially expressed proteins, LOXL1 was found in IPF EVs. [47] Kidney Fibrosis LOXL2 is expressed in compartments of renal tissue, where it appears to contribute to the progression of tubulointerstitial fibrosis. [48] LOXL2 inhibition significantly reduced interstitial fibrosis in a mouse model of renal fibrosis. [49] Elevated serum LOX and LOXL2 levels may act as a potential biomarker for kidney fibrosis.
[50] In a murine model of cyclosporine induced nephropathy, pan-LOX and LOXL2 specific inhibitors attenuated kidney damage. [51] Liver Fibrosis Increased LOX levels both in tissue and serum is associated with collagen in the extracellular space in animal models of hepatic fibrosis. [52][53][54][55] Hepatic stellate cell activation results in elevated LOX mRNA and protein in liver fibrosis. [56][57][58][59] Lysyl oxidase activity levels increase in patient serum from chronic persistent hepatitis to chronic active hepatitis to cirrhosis. [60] In a mouse model of liver fibrosis, increased steady state levels of LOXL mRNA occur early in fibrosis development. [61] Expression of LOX and LOXL2 in hepatocytes is linked to liver fibrosis. [62]  Reduced levels of miR-29b are associated with elevated levels of LOX in models of liver fibrosis. [63] In patients with severe obesity or obstructive sleep apnoea, serum levels of lysyl oxidase can act as a potential biomarker of liver fibrosis. [64] Targeting of LOXL2 is associated with anti-fibrotic effects in a mouse model of hepatic fibrosis.
[65] Links between LOXL2, insulin resistance and fibrosis accumulation in non-alcoholic fatty liver disease (NAFLD) are identified. [66] LOXL1 identified as a candidate therapeutic target for ameliorating liver fibrosis progression in cirrhosis, and inhibition of human hepatic stellate cell mediated fibrogenesis. [67,68] First demonstration that a small molecule dual inhibitor of LOXL2/3 (PXS-5153A) can ameliorate fibrosis in models of liver fibrosis and myocardial infarct [69] Patients with HCV who demonstrated sustained responses to antiviral therapy were shown to have regression in liver fibrosis associated with decreased LOXL2 expression. [70] anti-LOXL2 based therapy in a mouse model of liver fibrosis results in reduced fibrosis via accelerated collagenolytic activity by macrophages. [71] miR-15b/16 are downregulated in activated hepatic stellate cells (HSCs), and overexpression of these miRs is found to suppress LOXL1 expression in HSCs and induce a fibrogenic response. [72,73] Selective deletion of LOXL1 in HSCs in a murine NAFLD model ameliorates fibrosis, and serum levels of LOXL1 are positively correlated with histological fibrosis progression in NAFLD patients. [74] Amyotrophic IPF is a clinically challenging condition of the lung with significant interstitial fibrosis, is progressive, irreversible and has a poor prognosis with a median survival of 2-4 years [82]. Recently, a significant amount of evidence has linked the immune microenvironment to the development of fibrosis in IPF [83][84][85], but what evidence links lysyl oxidase activity to the fibrotic process? Numerous studies on bleomycin-induced lung fibrosis have confirmed the induction of LOX expression and the amelioration of fibrosis by the pan-lysyl oxidase inhibitor β-aminopropionitrile (BAPN) ( Table 1) [35][36][37]86]. Elevated activity of LOX has been confirmed for other models of IPF, for example in paraquat-induced fibrosis [39,40] (Table 1). Elevated levels of LOXL2 also play important roles in IPF with links to disease progression risk [38], increased fibroblast-to-myofibroblast transition (FMT) [41] and increased collagen fibril thickness [43]. Moreover, the nuclear localization of LOXL2 protein has been associated with the progression of lung fibrosis [45] and suggests a dynamic regulatory role (Table 1). In this regard, inhibition of LOXL2 in ex vivo lung explants from patients with IPF was shown to induce collagen turnover [46] (Table 1). LOXL1 and LOXL3 have also been identified in the pathogenesis of IPF [41,43,47] (Table 1).

Other Non-Cancerous Disease States
A significant body of evidence now also links lysyl oxidases to other non-cancerous fibrotic conditions including renal fibrosis, liver fibrosis and sclerosis (Table 1). In kidney fibrosis, LOXL2 appears to be a major factor with potential utility as a candidate biomarker for monitoring kidney fibrosis [50], and is a critical element in tubulointerstitial fibrosis [48] Cancers 2022, 14, 981 5 of 25 (Table 1). In models of renal fibrosis, therapeutically targeting LOXL2 activity is able to attenuate the main symptoms [49,51].
Within the setting of the liver, fibrosis is correlated with increased LOX levels in both the tissue and serum of patients, and there is a further association with altered collagen in the extracellular space in animal models of hepatic fibrosis [52][53][54][55]64] (Table 1). Elevated LOX levels in patient serum have since been mapped along the liver fibrosis continuum from chronic persistent hepatitis to chronic active hepatitis to cirrhosis [60]. A critical element in the increased LOX activity and fibrosis in liver tissue involves the activation of hepatic stellate cells (HSCs) [56][57][58][59].
In addition to LOX, links between liver fibrosis and LOXLs have also been identified, spanning non-alcoholic fatty liver disease (NAFLD) to cirrhosis [61,62,67,68,74]. Intriguingly, microRNAs (miRNAs) that have been found to target LOX/LOXLs have been identified in liver fibrosis [63,72,73] (Table 1). It is interesting to note that some of the miRNAs identified have both altered expression in and potential therapeutic impact on MPM. The accumulated evidence that LOX/LOXL activity plays important roles in liver fibrosis suggests that targeting these may have potential therapeutic benefit by reducing fibrosis [65,[67][68][69]71,74] (Table 1).
Sclerosis encompasses a diverse number of pathological conditions and fibrosis is associated with both amyotrophic lateral sclerosis (ALS) and systemic sclerosis. LOX expression and activity are increased in both animal models and patient samples of ALS [75,76], and LOX, LOXL2 and LOXL4 are elevated in the fibroblasts and serums of patients with systemic sclerosis [37,[77][78][79][80][81] (Table 1).

Altered Expression of Lysyl Oxidases in Cancerous Settings
Fibrosis is also commonly associated with many solid tumors [97,98]. In the following sections, we summarize the pertinent information linking LOX/LOXL activity to some of the main solid cancers in Table 2. Table 2. Selected examples of known LOX/LOXL family links with various cancers.

Breast Cancer
Lysyl oxidases associated with chemotherapy resistance in triple negative breast cancer [99] Stromal expression of LOXL2 is associated with tumor aggression and disease-specific mortality.
[104] LOX expression is significantly higher in triple negative breast cancers versus other breast cancer subtypes.
[105] LOXL2 involved with breast cancer metastasis to the lung [106] LOXL2 expression may serve as a biomarker for breast cancer and is detectable in serum and urine.
[107] LOX and LOXL proteins are located in the stromal reaction of ductal carcinoma in situ (DCIS) breast cancer. [108,109] Renal Cell Cancer (RCC) LOX and LOXL2 significantly elevated in RCC and associated with poorer overall survival (OS) [110] LOXL2 associated with migration, invasion and EMT transition of RCC [111] Primary RCC cultuure endogenously express LOX, and plays major roles in progression via activities on cellular adhesion, migration, and collagen stiffness. [112] Early demonstration of overexpression of LOX mRNA in RCC [113] Non-Small Cell Lung Cancer (NSCLC) A novel LOX polymorphism G473A is associated with increased risk for lung cancer. [114][115][116] LOXL1 promotes lung cancer tumourigenicity via collagen matrix remodelling and collagen fibre alignment in vitro and in vivo. [33]  High expression of LOX and LOXL2 mRNA and protein is associated with poor prognosis in NSCLC patients. [117][118][119][120] Low expression of the LOXL2 protein in adenocarcinomas is associated with a poorer N-stage, a higher pathological TNM stage and poorer differentiation. [121] The miR-200/ZEB1 axis drives lung cancer metastasis through LOXL2.
[122] LOX activity is associated with increased invasion and migration of hypoxic NSCLC cells.
[126] High expression of LOXL2 protein is associated with worse DFS and OS in patients with PDAC.
[127] Increased levels of LOX, LOXL1, and LOXL2 expression in PDAC are associated with poor responses to chemotherapy by limiting drug distribution, and inhibition of LOX enhances drug efficacy. [128,129] In an orthotopic PDX model of PDAC, targeting LOXL2 led to accelerated tumour growth, and poorer overall survival.
[130] Levels of LOX are a potential prognostic markers for the prognosis of pancreatic cancer patients. [131] Liver Cancer Reduced levels of miR-29a in hepatocellular carcinoma (HCC) lead to elevated expression of its known targets LOX and LOXL2. [132] A circRNA network has been identified that activates LOX transcription in HCC.
[133] High LOXL3 expression predicts poor outcomes for patients with HCC, and is correlated with immune infiltrates and T-cell activation. [134] High LOX expression is associated with an higher recurrence rate and poorer OS in patients with HCC. [135,136] LOXL2 is overexpressed in HCC and positively correlated with tumour grade, metastasis, and poor OS.
[137-139] LOXL2 is significantly overexpressed in human HCC sera and may act as a good biomarker for HCC.
[140] LOXL4 is upregulated in HCC tissues and associated with poor prognosis. Exosomal-mediated transfer of LOXL4 between HCC cells and human umbilical vein endothelial cells (HUVECs) promotes cell migration and angiogenesis, respectively

Breast Cancer
The role of collagen and fibrosis in breast cancer is well-established, particularly with respect to invasive breast cancer [142]. The initial observations were that lysyl oxidases are located in the stromal region of ductal carcinoma in situ (DCIS) breast cancer [108,109]. The highly aggressive triple negative breast cancer subtype is linked to higher LOX expression [105] (Table 2). Moreover, strong links have been found between LOXL2, metastasis [101,104,106], and higher mortality [100], and LOXL2 is a potential biomarker for cancer detection [107]. Finally in pre-clinical studies, targeting LOXL4 has been found to reduce breast cancer cell proliferation, migration, and metastasis in vitro and in vivo [102,103] ( Table 2).

Renal Cell Cancer
Intratumoral fibrosis is a common element in clear cell renal cell carcinoma (RCC) [143], and has been suggested as a potential marker of prognosis [144]. Elevated LOX gene expression occurs in RCC [113] (Table 2) and is linked to poorer overall survival (OS) [110]. Pre-clinical studies of RCC progression showed increased LOX activity is associated with cellular adhesion, migration, and collagen stiffness [112]. LOXL2 has also been found to be significantly elevated in RCC and with poorer OS [110]. Similar to LOX, LOXL2 activity affects migration, invasion and progression to EMT in RCC [111].

Pancreatic Cancer
Pancreatic cancer (PDAC) is another tumor with well-established links to fibrosis [145] ( Table 2). High expression of LOX is associated with a significantly shorter progression-free survival (PFS) [131] and poor response to chemotherapy [128,129]. The inhibition of LOX activity with a LOX-specific antibody enhanced efficacy of gemcitabine treatment in animal models [129].

Lung Cancer
High expression of LOX and LOXL2 mRNA and protein is for the most part associated with poor prognosis in non-small cell lung cancer (NSCLC) patients [117][118][119][120] (Table 2). However, one histologically-based study suggests low expression of LOXL2 is associated with a poorer N-stage, higher pathological TNM stage and poorer differentiation in NSCLC patients [121]. High expression of LOXL1 in cancer-associated fibroblasts (CAFs) has been shown to enhance NSCLC tumorigenicity [33] (Table 2), while increased LOX activity also helps promote increased invasion and migration of NSCLC [123] (Table 2).

Altered Expression of Lysyl Oxidases in Malignant Pleural Mesothelioma
Targeting fibrosis has therapeutic benefits in mesothelioma. For instance, inhibition of collagen formation by the proline analogue thiaproline delays tumor growth in a mouse model of mesothelioma [146]. More recently, the endocytic collagen receptor, uPARAP, has been identified as being a potential diagnostic marker amenable to therapeutic intervention [147]. In MPM cells derived from epithelioid and biphasic histological subtypes, invasive capacity is enhanced in fibronetin-collagen matrices [148]. Moreover, high expression of the collagen alpha type 1 (COL1A1) gene is associated with a poorer prognosis [149], while type V collagen has been linked to poorer prognosis and an altered tissue microenvironment aiding invasion [150,151].
There is a strong rationale for lysyl oxidase inhibitors as a therapeutic approach in mesothelioma due to their ant-fibrotic mechanisms [152,153], synergism with chemotherapies [1], and modulation of EMT in mesothelioma [154,155].
Using Oncomine to conduct in silico analysis of existing mesothelioma microarray datasets [156,157], we confirmed that significant mRNA overexpression of lysyl oxidases occurs in both primary MPM tissues and cell lines (Figure 1), a finding since confirmed in a study of a Korean cohort of MPM patients [158].
MPM datasets accessed via The Cancer Genome Atlas (TCGA) were analyzed using Gepia2 [159] to review the relationship between survivability and lysyl oxidase expression. When stratified by median expression, the results revealed that higher lysyl oxidase expression correlated with a decreased overall survival (OS) when compared to low lysyl oxidase expression (p = 3.6 × 10 −6 ) as shown in Figure 2. MPM datasets accessed via The Cancer Genome Atlas (TCGA) were analyzed using Gepia2 [159] to review the relationship between survivability and lysyl oxidase expression. When stratified by median expression, the results revealed that higher lysyl oxidase expression correlated with a decreased overall survival (OS) when compared to low lysyl oxidase expression (p = 3.6 × 10 −6 ) as shown in Figure 2.  [159], and stratified at the median between high expression and lox expression. When stratified at the median, the high 5 signatures group had a significantly worse OS (p = 3.6 × 10 −6 ).
Subsequently, we examined LOX activity in a cohort of n = 120 patient plasmas (n = 40 benign vs. n = 80 malignant) and demonstrated that LOX activity is significantly elevated (p = 0.0002) in the plasmas of patients with MPM ( Figure 3).  [159], and stratified at the median between high expression and lox expression. When stratified at the median, the high 5 signatures group had a significantly worse OS (p = 3.6 × 10 −6 ). Subsequently, we examined LOX activity in a cohort of n = 120 patient plasmas (n = 40 benign vs. n = 80 malignant) and demonstrated that LOX activity is significantly elevated (p = 0.0002) in the plasmas of patients with MPM ( Figure 3).
Overall, the evidence suggests that LOX/LOXLs are upregulated within mesothelioma patients, play a role in the pathogenesis of MPM and are potential therapeutic targets.

Figure 2.
Poorer OS in mesothelioma is correlated with elevated LOX/LOXL mRNA levels. A combined gene expression of LOX, LOXL1, LOXL2, LOX3 & LOXL4 (5 signatures group) was analyzed for OS using the TCGA MESO dataset in Gepia2 [159], and stratified at the median between high expression and lox expression. When stratified at the median, the high 5 signatures group had a significantly worse OS (p = 3.6 × 10 −6 ).
Subsequently, we examined LOX activity in a cohort of n = 120 patient plasmas (n = 40 benign vs. n = 80 malignant) and demonstrated that LOX activity is significantly elevated (p = 0.0002) in the plasmas of patients with MPM ( Figure 3). Overall, the evidence suggests that LOX/LOXLs are upregulated within mesothelioma patients, play a role in the pathogenesis of MPM and are potential therapeutic targets.

Inhibitors of Lysyl Oxidases
Many inhibitors that target the lysyl oxidase family have now been described. In the following sections, we briefly review these and their current clinical development.

First Generation of Lysyl Oxidase Inhibitors
The first non-specific, competitive, irreversible pan-lysyl oxidase inhibitor was originally extracted from sweet peas (Lathyrus odoratus) in 1954 [161], with the active component being called beta-aminopropionitrile (BAPN). BAPN has a simple molecular structure containing a primary amine that mimics lysyl oxidases substrates (lysine residues), which interact with cofactor lysyl tyrosylquinone (LTQ) forming a covalent bond. In the 1960s, high doses of BAPN were trialed in Scleroderma patients ( Table 3, with dosing of 1-3 g/day for 22-67 days), causing alteration in bone formations along with efficacious increases of hydroxyproline in the urine and of the dermal collagen alpha to beta ratio. A decade later, BAPN was further tested in the fibrous scar that forms in urethral stricture patients (n = 5), where a shorter time frame was utilized (1 g/day for 3 weeks) and no adverse effects were observed. The physical properties of the scar were affected, but insufficient data was collected to conclude a therapeutic effect [162]. BAPN is not considered to be a clinical candidate due to the potential of bone changes and off-target effects [160,163]. BAPN is also a substrate for the other amine oxidases (semi carbazide-sensitive amine oxidase (SSAO) and diamine oxidase (DAO), which may contribute to the side effects observed in clinical trials [160].

Polyphenols Epigallocatechin Gallate (EGCG) and Ellagic Acid (EA)
Both ellagic acid (EA) and epigallocatechin gallate (EGCG) ( Table 3) are dietary catechin polyphenols. Polyphenols are large molecules characterized by multiple phenyl groups that are highly oxidative and indiscriminate towards their lysine substrates. A variety of biological functions (superoxide radical scavenging activity, H 2 O 2 scavenging, chelating capacity and lipid peroxidation) [164] have been attributed to polyphenols, potentially due to their promiscuous nature. Numerous clinical trials (Phase 1 and 2, EGCG extensively reviewed in [165] and Table 3) have been completed with EGCG and EA for both cancer and fibrosis. High doses of polyphenols are required due to poor bioavailability (EA maximal human plasma concentration 100 nM [166], and EGCG peak concentration in human plasma 341 pmol/mL [167]). In preclinical models, high doses of EGCG (2 daily doses 750 mg/kg, i.g.) are hepatotoxic to mice. Of the multiple clinical trials run on polyphenols only, one demonstrated a reduction in biomarkers (collagen oligomeric matrix protein and periostin) and a reduction in protein expression in treated tissues (col1, snail and pSMAD, NCT03928847) [46]. The EGCG-dependent reduction in fibrotic markers is a promising result, but to date no therapeutic effects or clinical benefits have been demonstrated. Polyphenols' influence on multiple biological processes, poor bioavailability and potential off-target effects will need to be addressed before they can become a successful clinical candidate.

Copper Chelators (Tetrathimolybdate and D-Penicillamine)
Copper is an essential microelement involved in a plethora of biological processes. Unbound copper ions are a potent oxidant and serve as a cofactor in redox reactions for enzymes, including lysyl oxidases. Copper absorption and its homeostasis in the body is highly regulated and complex, particularly as an excess is cytotoxic. The concentration of free copper in human plasma ranges from 66-143 µg/mL and in a variety of cancers has been shown to be elevated to between 129-328 µg/mL [168]. Copper staining of cancer tissue indicate that fibrosis is increased [169]. Copper concentration is particularly relevant to lysyl oxidases, as sequestering of copper is crucial for catalytic activity and the protein folding of the LTQ complex. A full review of copper chelators is not within the scope of this article; please see Baldari et al. [170].
The most commonly used copper chelators in clinical trials are tetra-thimolybdate (TM) and D-penicillamine, (Table 3), which are known to be well-tolerated at low doses. The only clinical trial so far that demonstrated a reduction in fibrosis had 80% of patients withdraw from the trial due to adverse effects. Despite exceptionally high doses (D-penicillamin: 750-1000 mg/mL daily) utilized, most fibrotic tissues were unchanged, with the exception of a decrease in cardiomegaly. To advance copper chelators in clinical trials, the dosedependent toxicity needs to be addressed.

2nd Generation Inhibitors Lysyl Oxidase like 2 Antibodies (Simtuzumab)
Simtuzumab was the first humanized monoclonal antibody to target the enzymatic activity of LOXL2 to enter the clinic. Expectations were high for Simtuzumab, as it was developed as other anti-fibrotic treatments (nintedanib and pirfenidone) were being approved by the FDA for IPF. In addition, there was a prevalence of promising preclinical data (utilizing the mouse LOXL2 antibody AB0023) in both cancers and organ fibrosis [189]. However, in the clinic, despite Simtuzumab being well-tolerated, it did not show efficacy in both cancers and fibrotic diseases ( Table 3). The lack of efficacy of Simtuzumab is thought to be dependent on a number of factors-which are irrespective of the potential of targeting the lysyl oxidase family in cancer and fibrotic diseases [160].
The inability to measure lysyl oxidase activity in vivo during Simtuzumab clinical trials limits our understanding of target engagement, pharmacokinetics (including determining optimal dosing, tissue distribution and process of elimination) and pharmacodynamics, which limits the ability to optimize the clinical trials. This lack of understanding is potentially the reason why the early trials (in primary sclerosing cholangitis, IPF and NASH) were dosed with only 125 mg, which, based on preclinical studies (Table 3), was insufficient for significant target inhibition. Later cancer trials dosed patients with 700 mg for 3-6 month periods; however, these trials recruited patients irrespective of LOXL2 expression, activity or presence of fibrosis, which may have diluted efficacious responses to Simtuzumab.
Simtuzumab targets a single member of the lysyl oxidase family (LOXL2). To date, it is unclear if there is redundancy of the lysyl oxidases expression when a single member is targeted in humans. In a cancer setting, redundancy alone could explain the absence of efficacy within the clinical trials. Furthermore, it is questionable if large molecules like Simtuzumab are able to penetrate the complex impervious environment of fibrosis, which is pivotal for the success of human clinical trials in either cancer or fibrosis.
The failure of Simtuzumab was disappointing, but may be explained by the unavailability of target engagement assays leading to insufficient dosing, lack of understanding of protein turnover, lysyl oxidase family redundancy and the potential for poor tissue penetration. Despite this, Simtuzumab blazes the path for the future development of small-molecule approaches and target engagement-driven clinical trials for the lysyl oxidase inhibitors.

Lysyl Oxidase Small Molecule Inhibitors: The Way of the Future
The first generations of lysyl oxidase inhibitors have shaped the numerous clinical trials that are currently underway in cancer and fibrosis. For example, all current trials are wielding oral small-molecule enzymatic inhibitors, which should be optimal for tissue penetration (Table 3). Furthermore, target engagement assays have been developed utilizing Quanterix technology to measure low concentrations and activity of LOX and LOXL2 specifically, in both tissue and plasma [160]. The plasma LOX and LOXL2 activity has been shown to mirror recovery of activity in human tissue, providing a less invasive monitoring of target engagement. The establishment of target engagement assays has illustrated the dynamic nature of protein resynthesis, particularly for LOXL2 [160]. Despite reaching 90% target engagement, daily treatments are required to ensure complete inhibition due to the high resynthesis rates of lysyl oxidases [160]. Presently, in terms of treatment, it is still unclear the role redundancy will play within the lysyl oxidase family, but this will only be a consideration when not utilizing pan-lysyl oxidase inhibitors (e.g., PAT-1251 a LOXL2 inhibitor and PXS-5382 a dual LOXL2/LOXL3 inhibitor). Redundancy is more likely to occur in a cancer setting due to the plasticity and adaptive nature of tumors; therefore, a pan-lysyl oxidase inhibitor is generally a better approach. The new generation lysyl oxidase inhibitors have been shown to be non-toxic in human subjects in Phase 1 dose escalation trials (PXS-5505, PXS-6302, PXS-5382, PXS-5338 and PAT-1251, Table 3). Phase 1c/2 recruitment for clinical trials for Myelofibrosis has opened for both the panlysyl oxidase inhibitor (PXS-5505) and a LOXL2 specific inhibitor (PAT-1251) ( Table 3). A combination clinical trial (PXS-5505 and Atezolizumab (anti-PD-L1) and Bevacizumabin (anti-VEGF) in hepatocarcinoma will begin patient recruitment imminently. For skin fibrosis, a topical pan-lysyl oxidase inhibitor PXS-6302, will soon open to recruit patients with acute and established scars.

Discussion
Given the significant roles of lysyl oxidases in both cancerous and fibrotic disease settings, coupled with the current development of small molecule inhibitors, it is our opinion that a strong rationale has emerged for the progression to clinical trials in mesothelioma. In the previous sections, we have documented the significant roles played by lysyl oxidases in fibrosis not only in the non-cancerous, but also in the cancerous setting. We have presented evidence that lysyl oxidase expression and activity are significantly dysregulated in mesothelioma. Lysyl oxidase inhibitors have been predominantly tested as a single agent, where it is well-established that their primary mechanism of action is to prevent enzymatic cross-linking of collagen and elastins to counter fibrosis. The anti-fibrotic action of lysyl oxidase inhibitors has been demonstrated in a variety of preclinical cancer [125,190] and fibrosis models [189]. In cancer, lysyl oxidase inhibitors alone are not cytotoxic, but are thought to be ideal for combining with chemotherapies and/or immune therapies.
Chemotherapies have been shown to increase the tissue stiffness and fibrosis in tumors [125,190]. Desmoplastic pancreatic cancers are well-circumscribed by fibrotic tissue, increasing interstitial pressure and limiting immune infiltration and the penetration of chemotherapies. Lysyl inhibitors have been shown to soften tumors and have a synergistic effect with chemotherapy (Gemcitabine) to increase survival and decrease metastasis [125]. Chemotherapies are beneficial in mesothelioma patients, although they may also exacerbate lung fibrosis, further limiting their respiration and lymphatic drainage. It is anticipated that lysyl oxidase inhibitors could alleviate mesothelioma patients' fibrotic response to chemotherapies, increase immune infiltration, extend survival and increase the quality of life.
The synergistic effects of chemotherapies and lysyl oxidase inhibitors have previously been thought to be solely due to the anti-fibrotic effects. However, more recently, lysyl oxidase inhibition has impacted immune cells. Chen et al. 2019 first suggested that lysyl oxidase inhibitors prevent the chemoattraction of tumor-associated macrophages (TAMs, pro-tumorigenic) in preclinical models of glioblastoma [191]. Furthermore, in pancreatic tumors treated with both a lysyl oxidase inhibitor and immune therapy, a decrease in pro-tumorigenic TAMs and an increase in CD8+ Granzyme B+ T-cells (anti-tumorigenic) was associated with increased survival [125].
Within the setting of metastasis, it may be possible to use lysyl oxidase inhibitors to prevent metastasis. Responses to paclitaxel-based therapy in breast cancer models induced LOX levels and activity in the lungs, resulting in pulmonary ECM remodeling and a prometastatic niche [192]. Mechanistically, CD8+ T cells secreted LOX, which could be targeted with lysyl oxidase inhibitors to suppress metastasis [192].
In preclinical models of human hepatocellular carcinoma, high LOXL4 expressing macrophages have an immunosuppressive function via an interferon (IFN)-mediated signal transducer and activator of the transcription (STATs)-dependent PD-L1 activation on CD8+ T cells [193]. Disappointingly, they fail to demonstrate that lysyl oxidase inhibitors affect survival and the immunosuppressive environment. Despite this, a phase 1c/2a trial combining a lysyl oxidase pan inhibitor with standard of care (Atezolizumab plus Bevacizumab) in unresectable hepatocellular carcinoma will be underway in 2022.
Checkpoint inhibitors have been approved by the FDA for the treatment of mesothelioma [11]. Coupled with lysyl oxidase inhibitors' anti-fibrotic activity and possible regulation of immune cell infiltration [125,194], their combination has potential as a treatment option for mesothelioma. Moreover, anti-fibrotic strategies could be combined with standard first-line chemotherapies (cisplatin/pemetrexed) and conceivably be added to second-line salvage settings. However, significant preclinical studies will be required to test these promising combination therapies.
Additional remaining unanswered questions that may affect the inclusion of antifibrotic treatment strategies in mesothelioma include:

1.
Is fibrosis limiting lymphatic drainage to cause pleural fluid buildup? Can lysyl oxidases inhibitors alleviate pleural fluid buildup? 2.
Will lysyl oxidase inhibitors be synergistic with chemotherapies and immune therapies in patients? 3.
How do lysyl oxidase inhibitors affect macrophages and their interaction with T-cells? 4.
Will tumors adapt to lysyl oxidase inhibition? 5.
Is there redundancy in the lysyl oxidase family members, particularly when treated with lysyl oxidase inhibitors? 6.
Are LOX/LOXL inhibitors synergistic with immune therapies and/or chemotherapies in preclinical mesothelioma and patients?
In this review, we elaborated on the known functions of lysyl oxidases in the fibrotic setting. Overall, we believe that lysyl oxidases play important roles in the pathogenesis of mesothelioma. The current developments regarding lysyl oxidase inhibitors in the clinical setting are very encouraging and it will be exciting to see how these clinical trials unfold. Targeting lysyl oxidases has potential therapeutic implications in the management of mesothelioma.

Conclusions
One of the most significant issues facing patients with mesothelioma is a drastically reduced quality of life. A debilitating symptom of mesothelioma is dyspnea, and it is our belief that fibrosis plays a major role. Targeting lysyl oxidases to disrupt the fibrotic process may improve the mesothelioma patients' quality of life not only whilst undergoing therapy, but also in a palliative setting.
A second critical aspect of targeting lysyl oxidases is the potential for the anti-fibrotic effects to either potentiate or improve the response of patients to immune checkpoint inhibition or in combination with standard chemotherapy. Drugs that target lysyl oxidases could be combined into all existing therapeutic regimens for the treatment of mesothelioma to the potential benefit of patients.
Author Contributions: Conceptualization, L.P. and S.G.G.; data curation, S.G.G.; writing-original draft preparation, L.P. and S.G.G.; writing-review and editing, L.P. and S.G.G. All authors have read and agreed to the published version of the manuscript.  (11)).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study, and a patient consent form is available upon request from MesobanK-https://www.mesobank. com/governance/privacy-notice-for-research-participants// (accessed on 17 January 2022).

Acknowledgments:
The authors would like to thank Wolfgang Jarolimek, Ross Hamilton and Jess Stolp for their support of manuscript review and comments.

Conflicts of Interest:
Lara Perryman is an employee and shareholder of Pharmaxis Ltd. Compounds PXS-5505, PXS-6302, PXS5382 and PXS-5338 described within the manuscript are owned by Pharmaxis Ltd.