α/β-Hydrolase Domain (ABHD) Inhibitors as New Potential Therapeutic Options against Lipid-Related Diseases

Much of the experimental evidence in the literature has linked altered lipid metabolism to severe diseases such as cancer, obesity, cardiovascular pathologies, diabetes, and neurodegenerative diseases. Therefore, targeting key effectors of the dysregulated lipid metabolism may represent an effective strategy to counteract these pathological conditions. In this context, α/β-hydrolase domain (ABHD) enzymes represent an important and diversified family of proteins, which are involved in the complex environment of lipid signaling, metabolism, and regulation. Moreover, some members of the ABHD family play an important role in the endocannabinoid system, being designated to terminate the signaling of the key endocannabinoid regulator 2-arachidonoylglycerol. This Perspective summarizes the research progress in the development of ABHD inhibitors and modulators: design strategies, structure–activity relationships, action mechanisms, and biological studies of the main ABHD ligands will be highlighted.


INTRODUCTION
Endocannabinoids 2-arachidonoylglycerol (2-AG) and Narachidonoylethanolamine (anandamide, AEA) are endogenous lipid molecules activating the two G protein-coupled cannabinoid receptors 1 and 2 (CB1R and CB2R). 2-AG and AEA are produced following stimulation from phospholipid precursors present in the cell membranes and immediately metabolized after their activation of specific signaling pathways by specific lipases. 1 Therefore, the endocannaboinoid system (ECS) includes also enzymes controlling endocannabinoid levels and the most important is fatty acid amide hydrolase (FAAH), mainly responsible for the hydrolysis of AEA and monoacylglycerol lipase (MAGL), which is designated for 2-AG inactivation. 2 In this context, it is noteworthy to introduce a family of endocannabinoid-degrading enzymes which is progressively attracting more interest by the scientific community: the α/β-hydrolase domain (ABHD) enzymes. Muccioli et al. provided the first evidence that not only does MAGL hydrolyze 2-AG, since they found that MAGL was not expressed in the mouse microglial cell line, BV-2, but also a 2-AG hydrolyzing activity was present. 3 In the same year, ABHD6 and ABHD12 were identified by activity-based protein profiling (ABPP). 2 In particular, 85% of brain 2-AG hydrolase activity can be ascribed to MAGL, and the remaining 15% is mostly performed by ABHD6 and ABHD12 (4% and 9%, respectively).
Besides ABHD6 and ABHD12 which are related to ECS, many other ABHD enzymes have been identified and they showed specific physiological functions as regulators of lipid metabolism and signal transduction. Their association to human diseases of altered lipid metabolism will be explained in detail in the following specific sections.
Importantly, all ABHD enzymes belong to the α/βhydrolase fold superfamily, 4 which includes many different hydrolytic enzymes and shares a common three-dimensional feature since members of this family contain eight β-strands with the second antiparallel strand. The β sheets are surrounded on both sides by α helices and loops connecting the eight sheets. Each member of this family derives its hydrolytic activity from a highly conserved catalytic triad, characterized by the sequence: (a) nucleophile residue (serine, cysteine, or aspartate) located in the nucleophilic elbow in the loop following strand β5; (b) acid residue (glutamate or aspartate) after strand β7; (c) histidine residue located after the last β strand. The active site can be covered by a dynamic lid. Most of the ABHD enzymes are also endowed with acyltransferase activity due to the conserved His-XXXX-Asp region (X is any amino acid). 5,6 It is noteworthy to underline that the α/β-hydrolase fold superfamily is a very large multifaceted protein family which includes more than 50 enzymes possessing different names. Nevertheless, the present Perspective is focused on those members of this superfamily which are usually named ABHD enzymes, with the aim of highlighting the therapeutic potential of this group of proteins.
Despite the fact that ABHD enzymes are attractive targets for novel therapies targeting cancer and metabolic diseases, 7,8 the research field concerning the development of inhibitors/ modulators of these ABHDs is still quite unexplored. A greater interest has been devoted to ABHD6 and ABHD12 inhibitors, due to their involvement in the ECS. In fact, CB1R and CB2R are involved in many physiological and pathological processes; therefore, beneficial effects derive from their modulation. Nevertheless, it is well-known that their direct activation is associated with many drawbacks such as receptor desensitization and abuse potential. For this reason, more recent therapeutic approaches are directed toward their indirect stimulation by the inhibition of endocannabinoid degradation. 9 While a growing number of selective and potent inhibitors of FAAH and MAGL have been published or patented in the last decades, the discovered ABHD6 and ABHD12 inhibitors are still in their beginning, since the amount of inhibitors is limited and few of them have been the object of extensive studies.
Many developed ABHD inhibitors reported in the literature and reviewed here were characterized by activity-based protein profiling (ABPP), because ABPP is a proteomic technology used to determine not only the activity in cells and tissues but also the selectivity of ABHD inhibitors in an unbiased proteome-wide fashion. A variety of applications of ABPP have been developed in the last decades, since ABPP combines different scientific disciplines. In order to speed up the drug discovery process, ABPP is able to test inhibitors against many enzymes in parallel, and thus, potency and selectivity can be determined in a saving-time approach. 10 ABPP relies on the design of small-molecule probes that covalently label the active site of families of enzymes in complex proteomes. In particular, these probes possess (a) a "warhead" that is a chemical portion targeting conserved structural features present in active sites of an enzyme family, such as electrophilic groups binding conserved active-site nucleophile serine of serine hydrolase enzymes, and (b) a reporter tag, to facilitate target characterization, i.e., fluorophores, biotin, and alkynes or azides (which can be modified by Huisgen 1,3-dipolar cycloaddition). 11 Experimental read-out techniques such as gel-based methods or LC-MS approaches are usually adopted for analyzing probetreated proteomes.
All ABHD proteins are reviewed in Table 1 and Figure 1, in which their main features are summarized. In this Perspective, inhibitors and modulators of ABHDs will be reviewed, classifying the compounds on the basis of the specific inhibited ABHD enzyme and on the different chemical families, with a special focus on the specific roles of each ABHD enzyme. Additionally, specific attention will be dedicated to the patented ligands of ABHDs, in particular to those that have not been reviewed elsewhere as dual MAGL/ABHDs inhibitors. 12 bone marrow, fat, skin arachidonoyl-CoA, oleoyl-CoA, 1-oleoyl-lysophosphatidic, triacylglycerols ABHD6 small intestine, duodenum, spleen, brain, brown adipose tissue, kidney, liver, skin, ovary diacylglycerols, 1(3)-monoacylglycerols with saturated medium or long acyl chains, 2arachidonoylglycerol, lysophosphatidylinositols, bis(monoacylglycero)phosphate ABHD7 brain − a ABHD8 brain, testis − a ABHD9 skin, esophagus epoxyeicosatrienoic acids, 9,10-epoxyoctadecamonoenoic acids, leukotoxin, linoleate-derived epoxy-alcohols ABHD10 kidney, thyroid S-palmitoyl substrates ABHD11 skeletal muscle, colon, prostate, small intestine, thyroid triacylglycerols, 2-oxoglutarate ABHD12 ubiquitous expression, brain 2-arachidonoylglycerol, 1(3)-isomer of arachidonoylglycerol, unsaturated C20:4 monoacylglycerols, lysophosphatidylserine lipids ABHD12B skin − a ABHD13 ubiquitous expression − a ABHD14A adrenal glands, brain, kidney, thyroid − a ABHD14B ubiquitous expression p-nitrophenyl butyrate ABHD15 fat, liver − a ABHD16A ubiquitous expression, skeletal muscle, brain and platelets medium-chain saturated monoacylglycerols, 1-linoleylglycerol, 15-deoxy-Δ 12,14 -prostaglandin J2-2-glycerol ester ABHD16B testis − a it is ubiquitously expressed, mainly in liver and stomach. ABHD2 is considered a triacylglycerol lipase and an ester hydrolase. 14 It is overexpressed in human androgen-sensitive prostate cancer tissues since lipid metabolism plays a key role in the development and progression of this type of tumor. Moreover, high ABHD2 expression is correlated with resistance to docetaxel-based chemotherapy. 15 Deletion of the ABHD2 gene was correlated to anoikis resistance in highgrade serous ovarian cancer (HGSOC), thus promoting a malignant phenotype and poor prognosis. 16 Furthermore, ABHD2 was shown to be involved in many diseases such as Hepatitis B virus propagation, 17 since its downregulation using antisense oligonucleotides blocked Hepatitis B virus replication and expression without affecting host cell physiology. ABHD2 plays a key role in monocyte/macrophage recruitment, therefore influencing the development of chronic diseases such as atherosclerosis and emphysema. In particular, ABHD2 deficiency induced emphysema, due to increased macrophage infiltration, increased inflammatory cytokines and enhanced apoptosis because ABHD2 is important to maintain lung structural integrity. 18 With regard to its involvement in the pathogenesis of atherosclerosis, ABHD2 genetic deficiency enhances the migration of vascular smooth muscle cells, which is one of the causes of this vascular disease. 19 In addition, ABHD2 expression was significantly increased in parallel with the differentiation from monocyte into macrophage, and macrophages of atherosclerotic lesions abundantly expressed ABHD2. 20 High expression of ABHD2 in spermatozoa revealed the ability of this protein to bind progesterone, triggering 2-AG degradation, thus revealing that progesteronemediated activation of ABHD2 finally stimulates sperm activation. 21 The same mechanism was induced by pregnenolone sulfate: similarly to progesterone, it activated calcium channel of sperm by ABHD2 binding. 22 Finally, ABHD2 proved to be involved in the regulation of calcium release from the endoplasmic reticulum (ER). 23 2.2. Inhibitors. Very recently, Baggelaar and his research group conducted an ABPP screening based on ABHD proteins and a library composed of 207 lipase inhibitors to identify selective ABHD2 inhibitors. 24 Urea derivative 1 ( Figure 2) exerted a notable activity on ABHD2 (pIC 50 = 5.50) with no other off-targets in mouse testis proteome. This selectivity assay was performed in this specific proteome, since ABHD2 has an important role in sperm fertility. In order to analyze this aspect, inhibitor 1 was evaluated for its capacity to reduce progesterone-induced acrosome reaction (AR) in vitro, which is an important calcium-dependent process for the fertilization of mammalian eggs by spermatozoa, and it is stimulated by many molecules including progesterone. Compound 1 reduced progesterone-induced AR in vitro in a concentration-dependent manner, by blocking calcium increase induced by progesterone, thus confirming that ABHD2 finely tunes intracellular calcium levels in mouse sperm. These results suggest that urea derivative 1 could represent an interesting starting compound to develop new ABHD2 inhibitors as perspective novel contraceptives.

ABHD3
3.1. Biochemical Features and Biological Roles. ABHD3, previously known as lung α/β-hydrolase 3 (LABH3), 25 is a poorly characterized 409-residue (46 kDa) serine hydrolase highly expressed in appendix, colon, gall bladder, lymph nodes, stomach, thyroid, small intestine, duodenum, whose biochemical or physiological functions are still scarcely known. ABHD3 showed a multifaceted role in the catabolism of medium-chain phospholipids, that is distinct from those of other known phospholipases, as demonstrated in metabolomic studies. 26 In fact, ABHD3 showed a good specificity toward phosphatidylcholines (PCs) containing C14 acyl chain and oxidatively truncated phospholipids over other phospholipids. ABHD3 was shown to be upregulated in a series of pathological conditions: in human ovarian cancer cell lines exposed to standard chemotherapeutic drugs (cisplatin, paclitaxel or topotecan), 27 in breast cancer tumors, as a proapoptotic gene, 28 in a human osteosarcoma cell line overexpressing the tumor suppressor gene HIC1 (Hypermethylated in Cancer 1) 29 and in mice hippocampus after low-intensity exercise alone and/or in combination with the natural antioxidant carotenoid astaxanthin, revealing an antioxidant function of ABHD3. 30 Conversely, ABHD3 is downregulated in peripheral blood mononuclear cells of patients affected by Crohn's disease 31 and in a rat model of glaucoma characterized by early optic nerve head, which is the principal site of initial axonal injury. 32

Inhibitors.
Tan and collaborators performed a competitive ABPP screening on a library of synthesized αand β-aminocyano N-methyliminodiacetic acid-containing (MIDA) boronates in mouse brain proteome. 33 Several compounds belonging to this class exhibited ABHD3 inhibition, but further studies on HEK293T (human embryonic kidney cells) lysates overexpressing ABHD3 showed that the most active and selective ABHD3 inhibitor was β-aminocyano(MIDA)boronate 2 ( Figure 3), with an IC 50 value of 0.14 μM in vitro. With regard to 2 selectivity, SDS-PAGE analysis of tissue proteomes was able to identify only a limited number of serine hydrolases. Consequently, the authors further investigated the selectivity of 2 by MS-based ABPP using stable isotope labeling with amino acids in cell culture (SILAC). This technique allowed to confirm the selectivity of boronate 2 on ABHD3 (>95% of blockade at 0.5 μM) without detecting any activity over 60 additional serine hydrolases in human colon cancer cell line SW620. A structure−activity relationship analysis revealed the importance of the phenylamide portion, the cyano group, and the fluorine atom of 2 for inhibition potency. Importantly, the boron atom is fundamental for ABHD3 covalent inhibition, and the MIDA boronate portion seemed to increase cell permeability or stability in cells when compared to the free boronic acid analogue, proving to be resistant to hydrolytic cleavage under neutral conditions during the ABPP experiments. Metabolomic studies of 2 confirmed the previous findings that ABHD3 inhibition leads to an increase of medium-chain PCs in human cells.
It is also noteworthy to cite 1,3-dicarbonyl derivatives 3 and 4 ( Figure 3) identified in the previously mentioned screening aimed at finding new ABHD2 inhibitors: 24 these two compounds proved to selectively inhibit ABHD3 over the other tested ABHD enzymes.
The research group of Cravatt synthesized a small library of N-hydroxyhydantoin carbamates and screened them through competitive ABPP on serine hydrolases. 34 Compounds 5 (ABC47, Figure 3) and 6 (ABC34, Figure 3) demonstrated a good activity on ABHD3 (IC 50 = 0.1 and 7.6 μM, respectively), but a more potent inhibition was observed on ABHD4 (IC 50 = 0.03 and 0.1 μM, respectively) in mouse brain membrane proteome. However, ABPP-SILAC experiments in human PC3 cells highlighted that 5 and 6 inhibited not only ABHD3 and ABHD4, but they also had four additional offtargets: ABHD6, hormone-sensitive lipase (HSL), phospholipase A2 Group VII (PLA2G7), and carboxylesterase 2 (CES2). This study suggests that the N-hydroxyhydantoin carbamate scaffold could be finely optimized to achieve the inhibition activity toward the desired serine hydrolase.

ABHD4
4.1. Biochemical Features and Biological Roles. Human ABHD4 is composed of 342 residues (39 kDa) and is prevalently expressed in testis and gall bladder. ABHD4 is a lysophospholipase/phospholipase B first identified in 2006 as the enzyme responsible for the deacylation of N-acyl phosphatidylethanolamines and lyso-N-acyl phosphatidylethanolamines to generate glycerophospho-N-acyl ethanolamines, which are intermediates for the biosynthesis of N-acyl ethanolamines, an important group of signaling lipids including anandamide. 35 Later, biochemical and in vivo studies revealed that brain N-acyl lysophosphatidylserines are also substrates of ABHD4. 36 ABHD4 has a beneficial role in a fibrosarcoma model, limiting cell proliferation. 37 ABHD4 is a regulator of anoikis, which is a programmed cell death of anchoragedependent cells when they detach from the extracellular matrix, and resistance to anoikis usually leads to cancer metastases. Genetic deletion of ABHD4 induced anoikis resistance in prostate cells as well as nasopharyngeal and ovarian cancer cells; however, the exact mechanism was not yet elucidated. 38 Very recently, Laśzlóet al. found that ABHD4 is a necessary mediator for the elimination of pathologically detached cells in embryonic brain, confirming that downregulation of ABHD4 may induce resistance to anoikis. 39 4.2. Inhibitors. Very few ABHD4 inhibitors are reported in literature: the most potent are the previously mentioned compounds 5 and 6 ( Figure 4) identified by Cognetta et al.
These N-hydroxyhydantoin carbamates displayed IC 50 values in the submicromolar range (IC 50 = 0.03 and 0.1 μM for 5 and 6, respectively) in mouse brain membrane proteome analyzed by gel-based ABPP, although they are not highly selective for ABHD4, because of their additional inhibition activity on ABHD3 (subsection 3.2). Interestingly, some analogues of compound 6 were further developed as probes for gel-based detection of ABHD4 in ABPP experiments; however, their discussion is out the scope of this perspective. Cognetta et al. identify other ABHD4 ligands, unfortunately none of them were selective for ABHD4 nor reached a greater inhibition potency than compounds 5 and 6. 34   a well characterized member of this class of ABHDs. It is a 349-residue protein (39 kDa) mainly expressed in bone marrow, fat, and skin. The mutation of ABHD5 gene causes the human Chanarin-Dorfman Syndrome or Neutral Lipid Storage Disease with Ichthyosis (NLSDI), which is a rare autosomal recessive disorder characterized by the presence of intracellular accumulation of triacylglycerol (TG) droplets in many tissues. Multiple organs and tissues are affected by this syndrome, since patients suffering of NLSDI manifest ichthyosis and sometimes liver steatosis with hepatomegaly, muscle weakness (or myopathy), ataxia, neurosensory hearing loss, subcapsular cataracts, nystagmus, strabismus, and mental retardation. 40,41 ABHD5 mutation is also related to a rare heritable form of nonalcoholic fatty liver disease (NAFLD), a severe health disease associated with significant morbidity and mortality. 42,43 In ABHD5, the nucleophilic serine is substituted by asparagine; therefore, ABHD5 itself is not able to hydrolyze triacylglycerols, but it coactivates adipose triglyceride lipase (ATGL), an important TG hydrolase which catalyzes the formation of glycerol and free fatty acids. 44 Mutations in both ATGL and ABHD5 cause the "neutral lipid storage disease" characterized by massive accumulation of TG in various tissues. Knockout of ABHD5 in mice resulted in an excessive lipid storage due to defective activation of ATGL-mediated TG hydrolysis. In fact, newborn mice showed a condition similar to human NLSDI, with severe hepatic steatosis and a defective skin permeability barrier. These studies have highlighted that ABHD5 exhibits a crucial role in cellular TG catabolism by its regulation on ATGL activity. 45 Differently, the use of antisense oligonucleotides to inhibit ABHD5 expression in adult mice induced severe hepatic steatosis, but at the same time prevented high-fat diet-induced obesity and insulin resistance. 46 Conversely, when mice were genetically deprived of ATGL, they showed a massive accumulation of lipids in several tissues and the inability to mobilize these fat stores, along with an increase in insulin sensitivity, glucose use, and tolerance. 47 A further study confirmed that ABHD5 knockdown by antisense oligonucleotides paradoxically improved hepatic insulin signaling, reducing diet-induced stress kinase activation, thus highlighting an important role of ABHD5 in mediating inflammatory responses. 48 ABHD5 overexpression in mice did not prevent the development of diet-induced obesity; therefore, the ATGL activation induced by ABHD5 is not a determining factor for lipolysis. 49 Therefore, despite the involvement of both ABHD5 and ATGL in TG hydrolysis, experimental evidence suggests distinct roles of these two proteins. 50 ABHD5 displayed acyl-CoA-dependent acyltransferase activity to lysophosphatidic acid, showing a preference for unsaturated species of acyl-CoA, such as arachidonoyl-CoA, oleoyl-CoA, and 1-oleoyl-lysophosphatidic acid. 6 ABHD5 was found to be located in the lipid droplets in adipocytes, thanks to the interaction with perilipin-1 (PLIN1 or perilipin-A), which is expressed almost exclusively in adipocytes, 51 and it is designated to the breakdown of TG in lipid droplets via its phosphorylation. A mutation of ABHD5, as in Chanarin-Dorfman syndrome, determines a weakening of the ABHD5 binding to PLIN1, suggesting that the loss of this interaction could induce this syndrome. 52 Lipolytic stimulation by catecholamines triggers the phosphorylation of PLIN1, 53 disrupts the complex ABHD5/PLIN1, thus inducing release and translocation of ABHD5 from the lipid droplets surface into the cytosol, enabling it to activate ATGL-mediated lipolysis. 54 The structure of C-terminal moiety of PLIN1 is of crucial importance, because mutations affecting this region proved to make PLIN1 unable to sequester ABHD5, thus triggering ATGL activation and resulting in increased basal lipolysis. 55 Another isoform of this protein, perilipin-5 (PLIN5 or Mldp), is highly expressed in tissues characterized by high rates of fatty acid oxidation, such as heart, skeletal muscle, and liver, and PLIN5 was able to bind both ABHD5 and ATGL, but not both the protein at the same time. 56 Both PLIN5 and ABHD5 were observed on the surface of cardiomyocyte lipid droplets, and their interaction was promoted by lipid loading. 57 Cardiac PLIN5 overexpression regulated ATGL-mediated TG catabolism under regulation of protein kinase A, but PLIN5 does not constantly impair cardiac lipolysis. 58 Patatin Like Phospholipase Domain Containing 3 (PNPLA3, also known as adiponutrin) interacts with ABHD5 competing with ATGL, so preventing its activation and their binding was much stronger than the interaction of ABHD5 with ATGL. Importantly, PNPLA3 suppressed ABHD5-dependent lipolysis in brown adipocytes. 59,60 ABHD5 is involved in cancer development: its reduced expression was detected in metastatic castrationresistant prostate cancer and colorectal tumors, in which ABHD5 deficiency induced epithelial to mesenchymal transition and promoted Warburg effect; thus ABHD5 acts as a tumor suppressor. 61,62 Differently, ABHD5 expression was increased in tumor-associated macrophages in colorectal cancer, and ABHD5 facilitated cancer growth by suppression of spermidine synthase-dependent spermidine production, since spermidine exerts an inhibitory effect on the growth of colorectal cells. 63 Later, the same authors proved that ABHD5 expressed in macrophages displayed an antimetastatic effect mediated by matrix metalloproteinases, and this opposite finding was justified by the observation that tumor-associated macrophages exhibited heterogeneous expression of ABHD5 and that subgroup of macrophages with low ABHD5 expression was found to be correlated with the invasive behavior of the tumor. 64 However, the role of ABHD5 in tumors is quite controversial: other studies reported that ABHD5 was overexpressed in prostate cancer cells and ABHD5 genetic deletion decreased growth of prostate cancer cells by inducing apoptosis. 65 Recent studies demonstrated that overexpression of ABHD5 induces cell cycle arrest at the G1 phase and blocks cell proliferation in prostate cancer cells by inhibition of protein synthesis mediated by mTOR complex 1 (mTORC1); therefore, activation of ABHD5 by ligands may represent a promising therapeutic option against cancer. 66 ABHD5 was found to be overexpressed and exerted a protumorigenic role in endometrial cancer by involving the AKT signaling pathway. 67 Travers et al. provided the first evidence of serine protease activity of ABHD5. Histone deacetylases (HDACs) act as repressors of cardiomyocyte hypertrophy through association with the pro-hypertrophic transcription factor myocyte enhancer factor-2 (MEF2). Catecholamine-induced stimulation of β-adrenergic receptors leads to activation of protein kinase A, which triggers the cleavage of HDAC4, with the subsequent production of an amino-terminal polypeptide of HDAC4, and ABHD5 was identified as the one responsible for HDAC4 proteolysis. This series of events ultimately ends with the inhibition of MEF2 transcriptional activity, with resulting protective effects in cultured cardiomyocytes and diabetic hearts, in turn identifying a cardioprotective role for ABHD5. In vivo studies confirmed that ABHD5 lacking mice displayed cardiomyopathy typically associated with neutral lipid storage disease. 68 70 The authors considered previous studies assessing that PLIN1 suppresses lipolysis by binding ABHD5, thus preventing ABHD5-mediated activation of ATGL. On the other side, phosphorylation of PLIN1 by protein kinase A led to ABHD5 release, which activates ATGL, thus promoting lipolysis in adipocytes. 53 High-throughput screening identified two compounds able to disrupt the interaction between ABHD5 and PLIN1 or PLIN5 in the absence of protein kinase A activation: the thiaza-tricyclic urea 7 (SR-4995, Figure 5) and the sulfonyl piperazine 8 (SR-4559, Figure 5). These two derivatives prevented the binding of ABHD5 to PLIN1, with IC 50 values of 200 and 510 nM, respectively. The newly developed ligands 7 and 8 directly bound to ABHD5 and were shown to be potent and specific allosteric modulators of this enzyme. In brown adipocytes, 7 quickly disrupted the complex between ABHD5 and PLIN5. Inhibitors 7 and 8 were also tested in adipocytes and muscles to evaluate their effects on lipolysis, and they rapidly stimulated lipolysis, displaying EC 50 values of 4−7 μM. ABHD5 knockdown experiments highlighted that ABHD5 deletion abolished the efficacy of synthetic ligands 7 and 8 of stimulating lipolysis. Moreover, these two compounds promoted dissociation of ABHD5 from PLIN1 or PLIN5, without affecting the ABHD5 capacity to activate ATGL. These two compounds, together with 9 (SR-3420, Figure 5), another thiaza-tricyclic urea derivative differing from 7 only in the presence of the 1,3-(trifluoromethyl)benzene substituent at the end of the alkyl-urea chain, were subjected to further biological experiments. 71 Compound 9 was more effective in inducing lipolysis than 7 or 8 in white and brown adipocytes. Activation of ABHD5 by 9 potently inhibited mTORC1, thus blocking mTORC1 signaling and inhibiting the anabolism of cancer cells. 66 Inhibitor 9 regulated the interaction between ABHD5 and PNPLA3 by increasing this interaction. 59 Compounds 7 and 8 were also the object of a patent dating from 2016, claiming small molecules as modulators of cellular lipolysis. 72 The authors declared that these modulators, by facilitating fat catabolism, may be used as therapeutic agents to cure diabetes, obesity, cardiovascular diseases but also cancer. Moreover, these derivatives could increase the content of skin barrier lipids upon topical application. Structure−activity relationship (SAR) studies focusing on the thiaza-tricyclic urea scaffold highlighted that the urea at C4 of the tricyclic ring was fundamental for the activity, while the shift at C3 caused a loss of activity; moreover, the replacement of the urea group with esters, amides and N-dialkyl ureas at C4 was detrimental for ABHD5 activity. The activity against this enzyme increased by introducing long alkyl chains on the urea group (i.e., n-butyl chain in 7), in particular those ending with aryl groups, while the presence of an oxygen atom in this side chain decreased activity. In addition, small substituents were preferred on the nitrogen atom of the amidic group (i.e., methyl group of 7). SAR studies on the sulfonyl-piperazine scaffold demonstrated that the length of the linker between the sulfonyl-piperazine moiety in 8 and the benzofuran ring could be slightly increased but this modification decreased the inhibition potency. For what concerns the benzofuran ring, it tolerated alkyl substituents such as methyl group as in 8 and a benzothiazole ring without electron-donating substituents was also allowed. Moreover, if the benzofuran was connected at C2 to the rest of the molecule, the potency was reduced.
6. ABHD6 6.1. Biochemical Features and Biological Roles. Serine hydrolase ABHD6 is a 337-amino acid protein (38 kDa) with its catalytic triad composed of Ser148-Asp278-His306. It is an integral membrane protein possessing a N-terminal transmembrane helix 73 and is ubiquitously expressed, in particular in brain (cerebellum, frontal cortex, hippocampus, and striatum), 74 small intestine (duodenum), brown adipose tissue, 75 spleen, skin liver, kidney, and ovary. 76 Moreover, female hormones such as estradiol and progesterone can promote the overexpression of ABHD6 in immune cells. 77 ABHD6 is an important enzyme not only in the central nervous system but also in peripheral tissues, and it is involved in many physiological and pathological states. 78−80 ABHD6 is significantly expressed in several cancer cell lines, such as bone, prostate, and leukocyte tumor cell lines. 76 ABHD6 expression is increased in Ewing family tumors (EFT), thus representing a possible diagnostic and/or therapeutic target for this disease, although ABHD6 knockdown in EFT cell lines did not result in a decreased proliferative activity or increased apoptosis rate. 81 Human pancreatic ductal adenocarcinoma (PDAC) cell lines displayed an high expression of ABHD6, and this enzyme was considered the driving force for the metastatic potential of PDAC cells. 82 ABHD6 is an important oncogene in non-smallcell lung carcinoma (NSCLC) cells, 83 since ABHD6 silencing reduced migration and invasion in vitro as well as metastatic potential and tumor growth in vivo. Differently, ABHD6 was identified as an antioncogene in hepatocellular carcinoma (HCC). 84 A recent study revealed a diacylglycerol lipase (DAGL) activity for ABHD6 in Neuro-2a cells. 85 A study identified ABHD6 as the main monoacylglycerol lipase present in pancreatic islet β-cells, in which glucose-stimulated insulin secretion is amplified by ABHD6 inhibition. This effect was ascribed to reduced hydrolysis of 1-monoacylglycerols, which activated the protein Munc13-1 (a key exocytotic effector), thus triggering insulin secretion. 86,87 Deprivation of ABHD6 in mice fed with a high-fat diet induced a reduction of weight gain and liver steatosis, an improved glucose tolerance and insulin sensitivity, an enhanced locomotor activity, and browning of white adipose tissues. In particular, the mechanism of adipose browning behind ABHD6 suppression seems to involve an increase in 1-monoacylglycerols (MAGs), which causes peroxisome proliferator-activated receptors α and γ (PPARα and PPARγ) activation. 88 A study was focused on the role of ABHD6 in the central control of energy homeostasis. ABHD6 knockdown in neurons of the ventromedial hypothalamus in mice led to impaired adaptive responses to high-fat feeding, dieting, and cold exposure, thus underlining the importance of ABHD6 in maintaining a good flexibility in energetic metabolism. 89 Some studies highlighted the correlation between ABHD6 expression and the pathogenesis of Epstein−Barr virus (EBV)-related diseases 90 and the autoimmune disease systemic lupus erythematosus. 91 As anticipated, the main substrate of ABHD6 is 2-AG: 2,92,93 ABHD6 controls 2-AG at the site of 2-AG production (postsynaptic), differently MAGL exerts the control at the site of CB1R (presynaptic). The intracellular orientation of ABHD6 is strategic to regulate 2-AG production at the site of its formation. ABHD6 preferentially cleaves MAGs possessing saturated acyl chains, with medium or long chains, with a preference for 1(3)-isomers compared to 2-isomers. 94 Considering that ABHD6 increases the formation of arachidonic acid by hydrolyzing 2-AG, it is easy to explain its involvement in inflammatory processes. ABHD6 inhibition reduces lipopolysaccharide (LPS)-induced macrophage activation by increasing 2-AG levels in vitro, since 2-AG oxygenation by cyclooxygenase-2 (COX-2) led to the formation of antiinflammatory prostaglandin D 2 -glycerol ester (PGD 2 -G). ABHD6 was also able to reduce LPS-induced inflammation in mice without provoking the typical central effects of MAGL inhibition 95,96 (cannabinoid behavioral and functional antagonism of the endocannabinoid system due to chronic MAGL inhibition). 97 The role of ABHD6 in peripheral tissues was established by Thomas et al. using antisense oligonucleotides to knock down the enzyme in vivo. ABHD6 proved to be implicated in lipid metabolism, since ABHD6 inhibition resulted in the accumulation of lysophosphatidylglycerol (LPG) and phosphatidylglycerol (PG). It exerted a protecting activity from high-fat-diet-induced obesity, hepatic steatosis, hyperglycemia, hyperinsulinemia and it improved both glucose and insulin tolerance in mice. Therefore, ABHD6 contributes to the development of the metabolic syndrome. 75 ABHD6 is implicated in lysophosphatidylinositols (LPI) metabolism in J774 macrophages as ABHD6 inhibition led to an increase of the levels of all LPI. The effect of ABHD6 inhibition was investigated in LPS-activated J774 cells to study the role of this enzyme in the response of an inflammatory setting. The authors of this study observed an increase in 20:4 LPI levels, therefore ABHD6 could be involved in the hydrolysis of 20:4 LPI; however, extensive studies are still needed to clarify the complex metabolic pathways of LPI. 98 Bis(monoacylglycero)phosphate (BMP), a phospholipid present in the intraluminal vesicles of late endosomes and lysosomes exerting a fundamental role in degradation and sorting of lipids, was identified as a substrate of ABHD6, thus revealing a role for ABHD6 in the late endosomal/lysosomal lipid sorting. 99 A more recent study pointed out that ABHD6 affected circulating BMP levels both in mice and humans; consequently deletion of ABHD6 led to increased BMP concentrations without provoking lysosomal storage disorders (LSDs). 100 These studies suggest that ABHD6 is a key regulator of different classes of lipids. High expression of ABHD6 was detected in an animal model of multiple sclerosis (cuprizone model of nonimmune dependent demyelination) and pharmacological blockade of this hydrolase partially attenuated demyelination and astrogliosis. 101 The role of ABHD6 was investigated in another animal model of multiple sclerosis, the experimental autoimmune encephalomyelitis (EAE): the use of an ABHD6 inhibitor remarkably ameliorated the clinical signs of EAE, exerting an anti-inflammatory and neuroprotective action. 102 However, more recently, the therapeutic efficacy of the pharmacological blockade of ABHD6 in improving the clinical signs of EAE was discredited, considering that the ABHD6 inhibition resulted only in a modest slowdown of EAE progression. 103 Wei and co-workers performed studies regarding the involvement of ABHD6 and α-amino-3hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-type glutamate receptors (AMPARs), which are tetrameric receptors formed by GluA1−4 subunits. ABHD6 inhibited the glutamate-induced currents of GluA1-, GluA2-, and GluA3containing AMPARs, by binding to GluA1−3 C-terminal regions. 104,105 Pharmacological ABHD6 inhibition in a mouse model of traumatic brain injury had multiple positive effects: it improved motor coordination and working memory performance due to a reduction of brain lesions, neuroinflammation, neurodegeneration and blood-brain dysfunctions. 106 An antiepileptic role was reported for ABHD6: ABHD6 pharmacological inhibition reduced pentylenetetrazole-induced seizures and also blocked spontaneous seizures in R6/2 mice, a genetic model of Huntington's disease characterized by dysregulated endocannabinoid signaling. This study suggests that the observed anticonvulsive effect was independent of cannabinoid receptors, but it involved GABA A receptors; however, further experiments are needed to confirm the abovementioned mechanism of action. 107 6.2. Inhibitors. 6.2.1. Carbamate Derivatives. Cravatt and collaborators performed a competitive ABPP in COS-7 cells transfected with the human ABHD6 and a library of known carbamate serine hydrolase inhibitors, with the aim of demonstrating that ABPP can be applied for the identification of potent and selective inhibitors for serine hydrolases. 108 On the basis of this strategy, carbamate 10 ( Figure 6) was the most potent and selective inhibitor of this library on ABHD6 (IC 50 = Figure 6. Carbamate-based ABHD6 inhibitors. Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective 350 nM). Compound 10 was further optimized to improve ABHD6 inhibition potency and among the 20 newly synthesized derivatives compound 11 (WWL70, Figure 6), which differs from compound 10 only for the presence of a pcarboxamide group in the para position on the terminal phenyl ring, showed the highest ABHD6 inhibitory activity, with an IC 50 value of 70 nM, still maintaining an excellent selectivity. Compound 11 was widely investigated in further pharmacological studies. It inhibited of about 50% the [ 3 H]-2-AG hydrolysis in homogenates prepared from neurons in primary culture, whereas the inhibition of [ 3 H]-2-AG hydrolysis was reduced to about 20% in homogenates prepared from adult mouse brain, without exerting significant effects in homogenates prepared from microglia in primary culture. These findings are consistent with the fact that ABHD6 activity is greater in neurons in primary culture than in adult mouse brain and ABHD6 expression is very low in microglia in primary culture. 93 Further pharmacological evaluations highlighted the potential therapeutic role of 11 in animal models of traumatic brain injury and experimental autoimmune encephalomyelitis, as previously described. 102, 106 Kiritoshi et al. tested 11 in an arthritis pain model: 11, by increasing 2-AG levels and hence activating CB1R, rescued the metabotropic glutamate receptor 5 (mGluR5) activity with a consequent restore of the medial prefrontal cortex output and cognitive function, in addition it reduced pain in the animal model. 109 Tanaka et al. reported that the anti-inflammatory and neuroprotective properties of 11 were not attributable to ABHD6 inhibition but to its interference with the metabolic pathway from arachidonic acid to prostaglandin E2 (PGE 2 ). 110,111 In particular, derivative 11 blocked PGE 2 production and the expression of COX-2 and microsomial prostaglandin E synthase-1/2 (mPGES-1/2), the metabolic enzymes necessary for PGE 2 production from arachidonic acid, in LPS-activated microglia cells and in an animal model of neuropathic pain (chronic constriction injury of the mouse sciatic nerve), thus proving its possible use for the treatment of inflammatory diseases and neuropathic pain.
Madiraju et al. deposited three patents showing that ABHD6 activity is tightly correlated to insulin secretion and to conversion of white into brown adipose tissue. 112−114 In particular, the three patents described ABHD6 inhibitors which promoted insulin secretion by increasing the accumulation of MAGs and that may be useful for the treatment of type-2 diabetes, insulin resistance and metabolic syndrome. In a cell-based model for insulin secretion, regulation, and pancreatic islet β-cell function studies, the carbamate derivative 11 and the related analogues 12 and 13 ( Figure 6) showed 95%, 98%, and 95% of ABHD6 inhibition, respectively, when tested at 10 μM. Moreover, they displayed an increased percentage of insulin secretion compared to control. Finally, compound 11 exerted a benefic effect on mice blood glucose level by increasing plasma insulin concentrations, thus confirming its potential application for treating type-2 diabetes and any other conditions associated with a low level of insulin secretion/production. 113 A novel carbamate-based compound, 14 (WWL123, Figure  6), was discovered by Cravatt's research group in 2010 by an ABPP screening. 115 Compound 14 is a selective ABHD6 inhibitor (IC 50 = 0.43 μM), which also maintained its selective inhibitory activity on ABHD6 in vivo (mice treated with 5−20 mg/kg, i.p., 4 h). Carbamate 14, thanks to its high permeability to the blood-brain barrier, exerted an antiepileptic activity in vivo, as previously described. 107 The 1,2,5-thiadiazole carbamate scaffold, present in potent inhibitors of lysosomal acid lipase, was properly optimized by Patel et al. to develop selective ABHD6 inhibitors, considering that many carbamate-based compounds were found to efficiently inhibit enzymes of the ECS. 116 The most potent ABHD6 inhibitor of this class was 15 (JZP-430, Figure 6), possessing a carbamate moiety linked to a saturated eightmembered ring in position 3 of the thiadiazole ring and a morpholine ring in position 4 of the central heterocycle, in order to balance the increased lipophilicity determined by the big ring size on the other position. Compound 15 showed an IC 50 value of 44 nM in lysates of HEK293 cells transiently expressing human ABHD6 and it was also able to inhibit ABHD6 in competitive ABPP of the mouse brain membrane proteome. Derivative 15 was endowed with a good selectivity for ABHD6 over FAAH (only 18% inhibition when tested at 10 μM concentration), and it maintained only a negligible residual activity on lysosomal acid lipase (<20% when tested at 10 μM concentration), without exerting any appreciable activity on cannabinoid receptors, ABHD12 and MAGL. As expected, 15 inhibits ABHD6 by an irreversible mechanism of action. The class of 1,2,5-thiadiazole carbamates was subjected to comparative molecular field analysis (CoMFA) and molecular dynamic (MD) studies on a homology model of ABHD6. 117 This study highlighted that the most important bond was the hydrogen bond established between the carbonyl group of 15 and the Phe80 backbone, one of the two residues forming the oxyanion hole, thus demonstrating the proper fitting of the compound in this region of the enzyme. During MD simulations, the ABHD6-15 complex was quite stable; however, the distance between the carbonyl group of the ligand and the Phe80 backbone increased during the simulation, thus weakening the hydrogen bond. On the other hand, the formation of an additional hydrogen bond between Ser148 and the carbonyl group of the inhibitor promoted the covalent bond necessary for the irreversible inhibition of the enzyme.
In the previously mentioned screening of Cognetta et al., Nhydroxyhydantoin carbamate 16 (MJN193, Figure 6), characterized by an isopropyl group and a N-substituted piperazine on the hydantoin moiety, showed a considerable activity and selectivity for ABHD6. 34 In 2017, Abide Therapeutics, Inc. patented a series of dual lipoprotein-associated phospholipase A2 (Lp-PLA2) and ABHD6 inhibitors for the treatment of several pathological conditions such as multiple sclerosis, ischemia, traumatic brain injury, Alzheimer's disease, Parkinson's disease, amyotrophic lateral sclerosis, cancer, and diabetes. 118 These newly developed Lp-PLA2/ABHD6 inhibitors show the common chemical structure of 2,5-dioxoimidazolidin-1-yl phenylpiperazine-1-carboxylates, resembling N-hydroxyhydantoin 16. They were tested in vitro (ABPP assays) to evaluate their inhibition potency on both enzymes, and they did not show selectivity for ABHD6. Representative compound 17 ( Figure  6) showed an IC 50 value lower than 100 nM for ABHD6 and between 100 nM and 1 μM for Lp-PLA2.
In the same year, a series of spirocyclic-fused carbamates as modulators of MAGL/ABHD6 was reported by the same company for the treatment of pain. 119 These dual MAGL/ ABHD6 inhibitors were characterized by a hexafluoropropan-2-yl piperidine-1-carboxylate moiety. The most promising derivatives for what concerns ABHD6 inhibition potency were compounds 18, 19, and 20 ( Figure 6). All the three spirocyclic-fused carbamates proved to be slightly selective for ABHD6 versus MAGL and FAAH. Indeed, they showed IC 50 values lower than 100 nM on ABHD6, between 100 and 1000 nM on MAGL and greater than 1000 nM on FAAH. Compounds 18,19, and 20 displayed a MAGL and ABHD6 inhibition activity greater than or equal to 75% at 1 μM, on the contrary FAAH inhibition activity was lower than 25% when tested in competitive ABPP assays in mouse brain membrane fraction. No in vivo data were available for these three inhibitors.
6.2.2. Triazole Urea Derivatives. The 1,2,3-triazole urea scaffold is a typical feature of serine hydrolase inhibitors, 120 in particular in 2012, the research group of Prof. Cravatt focused on this scaffold to develop new DAGL inhibitors. In this screening campaign, the piperidyl-1,2,3-triaziole urea 21 (KT195, Figure 7) was identified as a selective ABHD6 inhibitor (IC 50 = 10 nM) in competitive ABPP with a marginal cross-reactivity against DAGLβ, 121,122 and it was predicted to irreversibly bind to the enzyme, by carbamoylating the enzyme's serine nucleophile. In Neuro-2a cells, 21 confirmed its inhibition activity by fully blocking ABHD6 with an IC 50 value of 1 nM and a negligible inhibition of DAGLβ. Similarly, in peritoneal macrophages from inhibitor-treated mice, 21 inhibited ABHD6 and lowered interleukin-1β secretion from LPS-treated macrophages; however, two carboxylesterases (CES3 and CES2G) and lysosomal phospholipase A2 group XV (PLA2G15) were identified as off targets of this compound. Compound 21 was further studied to evaluate its potential role to block necrotic cell death. 123 This ABHD6 inhibitor was able to attenuate necrotic cell death of cultured fibroblasts by preventing mitochondrial calcium uptake and permeability transition pore formation. In addition to the above-mentioned off-targets, 21 also blocked ER calcium release and cell death by targeting the nucleophilic serine in ABHD2. 23 On the basis of ABHD6 inhibitor 21 and with the aim of obtaining new selective and central nervous system (CNS)active inhibitors of DAGLα and β, Ogasawara and collaborators synthesized a new triazole urea 22 (DO53, Figure 7), characterized by a 2-phenyl-piperazine moiety instead of the 2phenyl-piperidine group of 21. 124 After intraperitoneal administration to C57BL/6 mice, it showed a good selectivity on ABHD6 together with inhibition of PLA2G7, and a low potency on the original DAGL targets. The selectivity profile of 22 was more extensively elucidated by ABPP experiments coupled to quantitative high-resolution mass spectrometry: it confirmed a negligible activity against DAGLs, but it showed notable cross-reactivity with many other targets, such as ABHD2, ABHD3, carboxylesterase CES1C, and the platelet activating factor acetylhydrolase 2 (PAFAH2). In the same research program, the potent DAGLα and β inhibitor 23 (DH376, Figure 7) showed an undesired ABHD6 inhibition activity both in vitro and in vivo and also a cross-reactivity with carboxylesterase CES1C and HSL. 124,125 Later, 23 was used to identify the enzymes responsible for 2-AG production during retinoic acid (RA)-induced neurite outgrowth of murine neuroblastoma Neuro-2a cells. 85 The terminal alkyne group present in the chemical structure of 23 was used in a "click chemistry" approach to introduce reporter tags, which allowed one to visualize by a chemical proteomic strategy the targets of 23 in Neuro-2a cells. ABHD6 and DAGLβ were identified as the only targets and ABHD6 was found to hydrolyze diacylglycerols, thus contributing to the production of 2-AG during RA-induced differentiation of Neuro-2a cells, since 23 blocked 2-AG production and reduced neuronal differentiation.
Compound 21 was structurally optimized by Hsu et al. to improve its potency, selectivity and in vivo activity toward ABHD6. In this new series of irreversible piperidyl-1,2,3triazole urea inhibitors, compounds 24 (KT182, Figure 7), 25 (KT185, Figure 7), and 26 (KT203, Figure 7) showed a remarkable inhibitory activity against ABHD6, with IC 50 values of 1.7, 1.3, and 0.82 nM, respectively, corresponding to 0.24, 0.21, and 0.31 nM, when their potencies were measured in situ in Neuro-2a cells. None of them exerted any significant offtarget activity. 126 In these compounds, polar substituents were added in meta position of the biphenyl moiety (R 2 group, Figure 5), such as hydroxymethyl (24), piperidine-amide (25), or carboxylic acid (26). The quantitative mass-spectrometrybased proteomic method ABBP-SILAC was applied to verify their activities: both 24 and 26 inhibited >90% of ABHD6 activity, while 24 blocked >80% of ABHD6 activity in Neuro-2a cells. In addition, the three developed inhibitors did not show any considerable cross-reactivity toward a panel of serine hydrolases present in Neuro-2a cell line, confirming their selectivity for ABHD6 in living cells. Compounds 24 and 26 were also tested in vivo when intraperitoneally administered in mice: both compounds were effective in blocking ABHD6 in the liver at the higher tested dose (1 mg/kg), and only 24 reached the same effect in the brain, probably due to the carboxylic acid of 26 which hinders its brain penetration. A mild systemic inhibitory effect on a plasma esterase carboxylesterase-1 (CES1) was detected only for 24. Encouraging results were observed with compound 25, that proved to be an orally bioavailable and selective ABHD6 inhibitor in vivo, even if complete ABHD6 inhibition was only observed at higher dose (40 mg/kg). As anticipated in subsection 6.1, Manterola and co-workers used 24 in the Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective cuprizone model of nonimmune dependent demyelination, 101 because of its ability to cross the blood-brain barrier and its selectivity in vivo after intraperitoneal administration. After the promising results of this first evaluation, the use of ABHD6 inhibitors was reassessed in multiple sclerosis by testing compounds 24 and 26, showing different CNS permeability. 103 The administration of systemically active inhibitor 24 modestly attenuated the neurological disability of the EAE; on the contrary, the peripherally active inhibitor 26 was not effective in ameliorating the clinical signs of EAE. Both compounds 24 and 26 did not attenuate inflammatory responses associated with tissue damage in the chronic phase of EAE, and the chronic treatment with 24 caused the desensitization of brain CB1R. All together, these results suggest that ABHD6 blockade has only a moderate therapeutic effect in this model of demyelination. A series of dual ABHD6 and DAGLα inhibitors were recently published by Deng et al. 127 Their strategy aimed at finding dual inhibitors as potential therapeutic agents to treat metabolic and neurodegenerative diseases. This series of dual inhibitors bear the chiral hydroxylated 2-benzylpiperdine scaffold with a triazole urea moiety. Surprisingly, some of them including compound 27 (Figure 7), showed a good combination of inhibition activity of ABHD6 (pIC 50 = 6.6 in membranes from HEK293T cells expressing recombinant human ABHD6; 83% inhibition in ABPP experiments) and selectivity for ABHD6 versus DAGLα (4-fold) and other serine hydrolases such as FAAH and MAGL.
In conclusion, 1,2,3-triazole urea represents a suitable scaffold to design irreversible inhibitors of serine hydrolases, thanks to the electrophilic carbonyl group which promotes the binding to the nucleophile active site serine. To date, all DAGL inhibitors reported in literature also inhibit ABHD6, 128 and this aspect may be exploited as the starting point to develop new selective ABHD6 inhibitors, by conveniently modifying this chemotype. Moreover, some activity-based probes (i.e., compounds binding to the enzyme covalently, useful to detect the amount of the enzyme present in a biological system) based on the triazole-urea scaffold were developed to target ABHD6, thus highlighting the high versatility of this chemical core. 129 6.2.3. Other ABHD6 Inhibitors. In 2011, Marrs and coworkers designed a series of esters by replacing the glycerol polar head of 2-AG with various oxygenated heterocycles. 130 The ester derivative 28 (UCM710, Figure 8), characterized by an oxirane moiety, proved to be a potent dual inhibitor of ABHD6/FAAH (IC 50 values of 2.4 and 4.0 μM, respectively) when tested in neuron homogenates, without inhibiting MAGL nor binding to cannabinoid receptors. Additionally, it was able to efficiently inhibit 2-AG and EAE hydrolysis also in intact neurons, although without reaching the maximum activity (60% and 30% inhibition of AEA and 2-AG hydrolysis, respectively). The unique pharmacological profile of 28 may be determined by its chemical structure, which mimics the natural substrates of the target enzymes, thus likely the oxirane group cannot fit into the active site of MAGL which is covered by the cap domain, differently from ABHD6 and FAAH that lack the cap domain necessary for the substrate recognition and interaction.
In 2014, Janssen and collaborators developed a series of glycine sulfonamides as novel DAGLα inhibitors. A member of this chemical class, compound 29 (LEI106, Figure 8), acts as a submicromolar dual ABHD6/DAGLα inhibitor. 131 In the colorimetric biochemical assay performed in HEK293 membranes overexpressing human DAGLα, the sulfonamide 29 showed an IC 50 value of 18 nM and inhibited the hydrolysis of DAGLα natural substrate, [ 14 C]-sn-1-oleoyl-2-arachidonoylglycerol, with a K i value of 0.7 μM. After observing an off target in brain membrane homogenate-based assays, a specific biochemical human ABHD6 activity assay revealed that 29 inhibited ABHD6 with a K i value of 0.8 μM.
The 1,3,4-oxadiazol-2-one scaffold is widely adopted for the discovery of serine hydrolase inhibitors. Patel et al. optimized the 1,3,4-oxadiazol-2-one 30 (Figure 8), 132 in order to develop new potent and selective ABHD6 inhibitors. Compound 30 was previously synthesized by their research group in a discovery campaign of FAAH inhibitors, but 30 selectively inhibited human ABHD6 (about 40% inhibition at 1 μM),  Figure 8), which exerted a notable AHBD6 inhibition with an IC 50 value of 216 nM. 133 The free amino group in the meta position seemed to be essential for the activity on ABHD6, since its protection or shift led to detrimental decreases of inhibition activity. Compound 31 was selective for ABHD6 when tested at 10 μM concentration, with any notable activity on other members of the ECS (FAAH, MAGL, ABHD12, and cannabinoid receptors). This novel and selective ABHD6 inhibitor interacts with the enzyme through an irreversible mechanism, as suggested by dilution assays and further confirmed by molecular docking studies. Docking of 31 underlined that the compound was located in the oxyanion hole, thus the carbonyl group of the inhibitor was suggested in proximity of the nucleophilic Ser148. Additionally, the importance of the free amino group on the benzyl moiety was explained by considering its involvement in hydrogen bonds with the side chains of Glu190 and Glu253. 1,2-Dihydro-2-oxo-pyridine-3-carboxamides were developed as potential CB2R ligands; however, this scaffold furnished a very promising ABHD6 inhibitor. 134 4-Methyl-5-bromo-2substituted pyridine 32 ( Figure 8) bound not only to both cannabinoid receptors as expected (K i values of 113 and 606 nM for CB1R and CB2R, respectively) but exhibited a remarkable inhibition activity of ABHD6 enzyme with an IC 50 value of 530 nM, exerting also inhibitory activity against anandamide cell uptake (IC 50 = 620 nM), without affecting FAAH.
In 2021, a study about the role of 2-AG protection of the retina against the excitatory amino acid AMPA involved two ABHD6 inhibitors: AM12100 which was selective for ABHD6 (IC 50 = 8 nM) and AM11920 which was a dual MAGL and ABHD6 inhibitor (IC 50 values of 12.1 and 6.0 nM, respectively). 135 The structures of both inhibitors are not disclosed yet. Interestingly, both compounds exerted a neuroprotective effect in the animal retinal model of AMPA excitotoxicity, but the selective ABHD6 inhibitor was less effective, thus leading to the conclusion that the dual inhibition exerted by AM11920 induced a more evident 2-AG increase and therefore it showed a better pharmacological profile.
It is noteworthy to add in this section a series of dual inhibitors of human ABHD6 and ABHD12 (ABHD12 will be analyzed in detail in section 10) discovered in 2014 by Kaczor et al. The authors screened an in-house library of heterocyclic compounds, 136 leading to six weak inhibitors, pyrazole-based derivatives 33−37 ( Figure 8) and thiosemicarbazide compound 38 (Figure 8). The remaining enzymatic activity on each enzyme was measured as a percentage compared to control and ranged from 65.3 to 84.2% for ABHD6 and 78.4 to 85.4% for ABHD12. Despite their low inhibition activity on both ABHDs, these heterocycles could represent a starting point for further structural modifications to tune their activity selectively on ABHD6 or ABHD12.

ABHD9
7.1. Biochemical Features and Biological Roles. ABHD9, also named epoxide hydrolase 3 (EPHX3), is a 360-amino acid protein (41 kDa) characterized by the presence of a nucleophilic aspartate in place of a serine. ABHD9 is prevalently expressed in skin and esophagus. ABHD9 was renamed EPHX3 after studies in which it displayed epoxide hydrolase activity against epoxyeicosatrienoic acids and 9,10-epoxyoctadecamonoenoic acids in vitro. 137 Nevertheless, in a more recent in vivo study, genetic silencing of ABHD9 had no significant effects on the metabolism of fatty acid epoxides and did not alter LPS-induced lung inflammation or functional recovery after ischemia/reperfusion injury, that are two models regulated by epoxyeicosatrienoic acids. 138 ABHD9-mediated hydrolysis of leukotoxin led to the production of a metabolite which was identified as a strong mediator of acute respiratory distress syndrome (ARDS). 137 ABHD9 seems to be involved in cancer, since ABHD9 expression has been reported to be downregulated in tumors, such as prostate cancer, 139,140 melanoma, 141 B cell tumor, 142 gastric cancer, 143 salivary gland adenoid cystic carcinoma, 144 oral squamous cell carcinoma, 145 head and neck squamous cell carcinoma, 146 and colorectal carcinoma. 147 ABHD9 was considered a potential ichthyosis-related gene. 148 The role of ABHD9 in the regulation of skin barrier function was confirmed by other studies; in fact, ABHD9 was found to be involved in the production of epidermis-related linoleate triols, considering that it is highly expressed in the external cells of human epidermis. 149 Moreover, ABHD9 hydrolyzes linoleatederived epoxy-alcohols esterified in skin ceramides in vivo. 150 7.2. Inhibitors. In 2012 Decker and colleagues tested a class of N,N′-disubstituted urea derivatives, which previously were considered inhibitors of mammalian soluble epoxide hydrolase, on ABHD9. 137 Among these N,N'-disubstituted urea derivatives, 1-(1-acetylpiperidin-4-yl)-3-(4-(trifluoromethoxy)phenyl)urea 39 (TPAU, Figure 9), 1cyclohexyl-3-dodecylurea 40 (CDU, Figure 9), and 12-(3adamantan-1-yl-ureido)-dodecanoic acid 41 (AUDA, Figure 9) were the most active inhibitors on ABHD9, with IC 50 values of 75, 80, and 100 nM, respectively. These findings could be a starting point for the development of new ABHD9 inhibitors able to better elucidate their possible use as new therapeutic agents.

ABHD10
8.1. Biochemical Features and Biological Roles. ABHD10 is a 306-residue protein (34 kDa), ubiquitously expressed yet prevalent in kidney and thyroid. Proteomic studies located ABHD10 in the mitochondria. 151 Some studies in the literature described the involvement of ABHD10 in drug metabolism. 152 ABHD10 plays a key role in the metabolism of the immunosuppressant mycophenolate mofetil (MMF), because it led to the deglucuronidation in human liver of acyl glucuronide metabolite (AcMPAG), potentially responsible for some MMF-induced adverse effects such as leucopenia or gastrointestinal toxicity; therefore, ABHD10 exerted a detoxifying effect. 153 A similar detoxifying activity Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective was observed in the case of probenecid acyl glucuronide (PRAG), which is the main metabolite of the uricosuric agent probenecid, that can provoke severe allergic or anaphylactic reactions, as ABHD10 catalyzed PRAG deglucuronidation in human liver. 154 S-Depalmitoylase activity was observed for ABHD10; in particular ABHD10 acts on peroxiredoxin-5 (PRDX5), a key antioxidant protein and therefore ABHD10 can be included in the acyl protein thioesterases (APT) family of regulatory proteins. 155 8.2. Inhibitors. Cravatt and his research group, with the aim to identify new serine hydrolase inhibitors, discovered a series of aza-β-lactams (ABLs), which efficiently inhibited the mammalian serine hydrolase protein-phosphatase methylesterase-1 (PME-1). 156 Further structural optimization led to the identification of 42 (R enantiomer, ABL117, Figure 10), which inhibited both PME-1 and ABHD10 with IC 50 values of 250 and 210 nM, respectively. Thereafter, in order to improve ABHD10 inhibition, the authors performed a SAR evaluation of the ABL scaffold. Using bulky substituents as O-alkyl groups on the carbamates or shifting the methyl group to the para position of the benzene ring increased potency for ABHD10, as demonstrated by compound 43 (R enantiomer, ABL303, also named ML257, 157,158 Figure 10), which showed an augmented inhibition potency on ABHD10 (IC 50 = 30 nM) and a marked selectivity over ABHD6, prolyl endopeptidase (PREP), PME-1, and other serine hydrolases. Aza-β-lactam 43 maintained a notable activity in living Neuro-2a cells with an IC 50 value of 21 nM, without exhibiting off-targets when tested at 1 μM concentration and no appreciable inhibition of PME-1 at 10 μM. The quantitative mass spectrometry-based proteomic method ABPP-SILAC was employed to further test 43 in Neuro-2a cells: it selectively and near-completely inhibited ABHD10 (>95%), and thus, 43 was the first discovered potent ABHD10 inhibitor. Compound 43 is an irreversible inhibitor, acting via aza-β-lactam ring opening and subsequent serine acylation. 157 The same research group combined the flavagline rocaglate core, typical of natural compounds isolated from the genus Aglaia, characterized by a cyclopenta[b]-benzofuran structure, with a β-lactone scaffold, to give a class of rocaglate-derived βlactones as potential serine hydrolases inhibitors. 159 The most interesting derivative of this series is compound 44 (both enantiomers (+)-44 and (−)-44 are reported in Figure 10). Unfortunately, ABPP in proteomes deriving from human cancer cell lines (PC3 and LNCaP) and mouse tissues (brain, liver, and testes) and ABPP-SILAC analysis on PC3 cells pointed out that β-lactone 44 inhibited different serine hydrolases including not only ABHD10 (IC 50 value of about 100 nM) but also cathepsin A (CTSA), retinoid-inducible serine carboxypeptidase 1 (SCPEP1), and acyl-CoA thioesterase 1/2 (ACOT1/2). In particular, the pure (−)-44 enantiomer was shown to be responsible for most of the ABHD10 and ACOT1/2 inhibition activity in competitive APBB assay on PC3 cells. The authors hypothesized that compound (−)-44 irreversibly inhibits the target hydrolase, by acylation of the active site nucleophilic serine, as has been reported for other β-lactones.
In 2012, Adachi et al. employed MIDA-boronates to identify new inhibitors of ABHD10 and serine carboxypeptidase (CPVL). 160 Alkyl(MIDA)boronate 45 ( Figure 10) was tested in ABPP experiments in PC3 cell proteome: its conversion into the corresponding boronic acid was evident in buffer after 2 h of incubation. Therefore, the inhibition required the decomposition of the (MIDA)boronate portion to the free boronic acid. Compound 45 induced a complete inhibition of ABHD10 with few off-targets and showed a near-complete inactivation of ABHD10 at 10 μM and of ACOT1/2 at 100 μM. CPVL inhibition was confirmed in ABPP-SILAC assays, in which 45 inhibited by more than 95% both ABHD10 and CPVL, at 25 μM. 9. ABHD11 9.1. Biochemical Features and Biological Roles. ABHD11 or Williams-Beuren syndrome chromosomal region 21 protein (WBSCR21) or PP1226 is a 315-amino acid protein (35 kDa). ABHD11 is a mitochondrial protein, mainly found in skeletal muscle, 161 but it is an ubiquitous protein with higher expression in colon, prostate, small intestine, and thyroid. Its alternative name WBSCR21 originates from the fact that ABHD10 is among the deleted genes in Williams-Beuren syndrome, a severe neurodevelopmental disorder characterized by several diseases and abnormalities, concerning both physical and cognitive aspects. 162 ABHD11 expression was reduced in white adipose tissue in mice fed with a high-fat diet as well as in HSL knockout mice. On the contrary, treatment with the antidiabetic drug rosiglitazone increased its expression; however, other analyzed lipases and esterases were unaffected. Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective Therefore, the importance of these changes needs further elucidation. 163 ABHD11 is involved in cancer aggressiveness, since increased ABHD11 is a predictive biomarker of metastases in lung adenocarcinoma. 164 In breast cancer, ABHD11 was downregulated in paclitaxel-resistant MCF7/ PacR cells (68% compared to MCF7 cells), 165 but it was also related to breast cancer malignancy. 166 169 Recently, a role for ABHD11 in embryonic stem cell (ESC) maintenance was highlighted, determining that ABHD11 is important for self-renewal and metabolic homeostasis of ESC. 170 The ABHD11 locus also encodes for long noncoding RNA, named ABHD11-antisense (ABHD11-AS1), whose increased expression was observed in gastric, 171 colorectal, 172 pancreatic, 173 endometrial, 174 nonsmall-cell lung, 175 papillary thyroid, 176 and ovarian cancer. 177 9.2. Inhibitors. 9.2.1. Carbamate Derivatives. As mentioned before, the carbamate scaffold is very common among serine hydrolase inhibitors. During the screening study performed by Cravatt's group to identify new inhibitors of serine hydrolases, during the discovery of compound 14, compound 46 (WWL151, Figure 11) was identified as a mild inhibitor of ABHD11 (IC 50 = 5.3 μM), however highly selective, likely due to the unicity of its seven-membered azepane ring, compared to other carbamate derivatives with broad spectrum activity on the panel of serine hydrolases. The substitution with a 2-ethylpiperidine ring proved to be successful, giving rise to a more potent inhibitor 47 (WWL222, Figure 11) which selectively blocked ABHD11 (IC 50 = 170 nM) without any activity against other serine hydrolases. 115 Carbamate 47 was also very efficacious and selective in vivo when administered intraperitoneally in mice at 10 mg/kg. 9.2.2. Urea Derivatives. In 2010, Cravatt and colleagues carried out a fluorescence polarization-based competitive ABPP high throughput screening study to discover new inhibitors of lysophospholipase 1 and 2 (LYPLA1 and LYPLA2). 178 During this study, performed on a library of triazole urea-based compounds, the authors serendipitously identified the racemic compound 48 (ML226, Figure 12) as a remarkably potent (IC 50 = 15 nM) and selective (≥100-fold over more than 20 serine hydrolases) ABHD11 inhibitor, with a residual activity on N-acylaminoacyl-peptide hydrolase (APEH, 50% inhibition at 1.5 μM). The mode of action of 48 was assessed by LC-MS/MS studies, which revealed a covalent modification of the catalytic Ser141 of ABHD11, in which the triazole ring acts as the leaving group. A close analogue of 48, derivative 49 178 (AA44-2, Figure 12), bearing a bulkier methoxymethyl group instead of the ethyl group in 2position of the piperidine ring, showed an improved ABHD11 inhibition with an IC 50 value of 1 nM, still maintaining a high selectivity versus other serine hydrolases and no activity on APEH. 120 These properties were confirmed by ABPP-SILAC analysis in living mouse T-cells: treatment with 49 resulted in a blockade greater than 95% of ABHD11 activity at the concentration of 3 nM with no cross-reactivity over other 40 serine hydrolases observed in T-cells.
In 2016, Navia-Paldanius and collaborators screened more than 200 in-house synthesized compounds designed to target serine hydrolases by using competitive ABPP tests. 179 This screening led to the identification of three isoxazol-5(2H)-onecontaining urea derivatives exerting a nanomolar potency against human ABHD11: 50 (JZP-228), 51 (JZP-245), and 52 (JZP-249) reported in Figure 12 showed IC 50 values of 2.4, 3.4, and 2.3 nM, respectively. The three compounds were assessed in a competitive ABPP assay among the serine hydrolases of mouse whole brain membrane in order to evaluate their selectivity. All of them completely blocked ABHD11 activity when tested at 100 nM concentration; however, 50 inhibited an additional protein band, migrating at ∼60 kDa, attributable to FAAH. Moreover, 50 was previously found to be a HSL inhibitor with a reported IC 50 value of 14 nM. 180 The three inhibitors were tested in competitive ABPP with lysates of prostate cancer LNCaP and VCaP cells (both expressing FAAH) and PC3 cells (not expressing FAAH). Predictably, compounds 51 and 52 selectively blocked ABHD11, differently from 50 which confirmed its activity on FAAH in LNCaP and VCaP cells. Inhibitor 51 was further investigated in competitive ABPP: at 0.1 μM, 51 inhibited ABHD11 in all tested proteomes (mouse whole brain membranes, prostate cancer cell lysates, and mitochondrial fraction of brown fat and testicle), but at higher concentrations  (1−10 μM) it also inhibited FAAH. Additionally, at 10 μM, 51 showed as off-targets ABHD6 and the serine hydrolase KIAA1363 in mouse whole brain membrane proteome. The cytotoxic effect of urea 51 was evaluated in prostate cancer cells: it reduced proliferation of the nonaggressive cell line LNCaP, but it was poorly effective on the aggressive cell line PC3. Nevertheless, in LNCaP and VCaP cells, 51 acted as a dual inhibitor targeting both ABHD11 and FAAH with a similar potency. Navia-Paldanius et al. built an ABHD11 homology model in order to better understand the interactions between these urea-based compounds in the enzyme active site. The docking studies suggested that the inhibitors properly fitted the active site of ABHD11, where they established π−π interactions. It was postulated a possible irreversible inhibition mechanism, through active site serine acylation, in which the isoxazol-5(2H)-one ring behaves as the leaving group.
10. ABHD12 10.1. Biochemical Features and Biological Roles. ABHD12 is also known as ABHD12A, c20orf22, or 2arachidonoylglerol hydrolase, and it is a 398-residue protein (45 kDa). From a structural point of view, ABHD12 is a singlepass integral membrane protein, possessing a N-terminal transmembrane helix, which points its active site toward the extracellular space, and its catalytic triad is Ser246-Asp333-His372, as discovered by site-directed mutagenesis studies. 73, 92 The ubiquitously expressed ABHD12 has the highest expression in the brain (especially in microglia), and it is localized to the ER membrane in the mammalian brain, 181 where it is responsible for about 9% of 2-AG hydrolysis, together with MAGL and ABHD6. 2 ABHD12 is also present in macrophages and osteoclasts. Studies of substrate specificity revealed that ABHD12 prefers the 1(3)-isomer of arachidonoylglycerol over 2-AG and unsaturated C20:4 MAGs over C18:2 MAGs. 73 It was found that ABHD12 required glycosylation for optimal activity and it showed a strong preference for very-long-chain lipid substrates, such as lysophosphatidylserine (lysoPS) lipids. 181 Furthermore, in the brain, ABHD12 hydrolyzes oxidizedphosphatidylserine, which is considered an apoptotic signal, under severe inflammatory stress. 182 Mutations of ABHD12 were found to be related to the etiology of some pathologies, such as the neurodegenerative disorder called polyneuropathy, hearing loss, ataxia, retinitis pigmentosa, and cataract "PHARC", likely due to impaired 2-AG metabolism. 183,184 Other studies suggest that PHARC may be induced by a dysregulated lysoPS lipase activity which is typical of ABHD12, since ABHD12 deficient mice displayed increased proinflammatory lysoPS lipid levels and neurobehavioral abnormalities similar to those of the PHARC phenotype. 185 Together with ABHD16A, ABHD12 dynamically regulates lysoPS metabolism: ABHD16A contributes to the production of both cellular and secreted lysoPS starting from phosphatidylserine (PS), and ABHD12 preferentially controls degradation of secreted lysoPS to glycerophosphoserine, thus exerting complementary roles. 186,187 A recent study in ABHD12 knockout mice reveals an upregulation of lipids deriving from arachidonic acid in the brain, thus suggesting that neuroinflammation may contribute to the development of PHARC-like symptoms. 188 Dysfunctional ABHD12 has been linked to a variant of PHARC named Usher syndrome 3 (USH3), an autosomal recessive genetically heterogeneous disorder, characterized by congenital sensorineural hearing impairment and retinitis pigmentosa. 189,190 Some tumor types showed an increased ABHD12 expression, such as in colorectal cancer 191 and in breast cancer MCF7 and MDA-MB-231 cell lines ABHD12 knockdown reduced cell growth, proliferation, and invasiveness. 192 10.2. Inhibitors. 10.2.1. Natural Compounds. Encouraged by the fact that some natural triterpenes exerted a certain inhibition activity on hydrolases (i.e., MAGL and ABHD6), such as pristimerin 69 193 (Figure 17), Parkkari et al. performed a screening of triterpene and triterpenoid derivatives by purchasing 15 commercially available compounds. The inhibition data were determined in lysates of HEK293 cells transiently overexpressing human ABHD12 and revealed that the oleanane derivative maslinic acid 53 (Figure 13) was the most potent ABHD12 inhibitor of this series, showing an IC 50 value of 1.3 μM. 194 A preliminary SAR study revealed that the presence of a carboxylate in position 17 in combination with small hydrophobic groups such as the methyl groups at position 4 determined a good inhibition activity. The screening of triterpene derivatives continued with a series of synthetic betulinic acid derivatives: among them, triterpene 54 ( Figure  13), bearing an indole heterocycle fused with the central core in the place of the two hydroxyl groups of maslinic acid 53, showed good inhibition of ABHD12 (IC 50 = 0.9 μM). The authors enriched the SAR relative to this class of derivatives, since it was evident that the presence of hydrogen bond donors or acceptors at position 3 was required for an optimal inhibition activity. Later, the inhibition mechanism for the best two compounds was investigated: they proved to inhibit ABHD12 in a reversible manner, as tested by a dilution assay of the enzyme−inhibitor complex. Moreover, compounds 53 and 54 were tested in ABPP of HEK293 cell lysates and mouse brain membrane preparations and proved to be selective for ABHD12 over ABHD6, MAGL, FAAH, CB1R, and CB2R.
A cycloartane-type triterpene derivative 55 ( Figure 13) isolated from Euphorbia pterococca proved to be a moderate ABHD12 inhibitor (IC 50 = 11.6 μM); however, it was surprisingly selective, since it did not affect ABHD6, MAGL, and FAAH enzymes. 195 10.2.2. Synthetic Compounds. In 2019, the research group of Cravatt developed a thiourea derivative, 56 (DO264, Figure  14), which proved to efficiently and selectively inhibit ABHD12 in vitro and in vivo. 196,197 An initial HTS procedure based on an innovative fluorescence assay, which measures the ABHD12-mediated hydrolysis of lysophosphatidic acid, was inhibition, since the block of the enzyme occurred at a lower concentration of the inhibitor, and this fact was confirmed testing 56 on different cell lines, when exposed to 1 μM 56. An excellent ABHD12 inhibition was confirmed in C57BL/6 mice treated with 56 by intraperitoneal or oral administration, observing only a low inhibition of ABHD2 and phospholipase A2 group VI (PLA2G6). 56-treated mice exhibited increased levels of brain lysoPS and 20:4 PS lipids, similarly to the changes observed in ABHD12 knockout mice, although they did not show any auditory defects, which are typical symptoms of PHARC disease. Moreover, 56-treated and ABHD12 knockout mice exhibited a heightened immunological responses in a lymphocytic choriomeningitis virus (LCMV) clone 13 infection animal model, thus highlighting that ABHD12 may have an immunosuppressive function. A recently discovered effect of compound 56 is its ability to enhance ferroptotic death, a particular form of cell death defined by peroxidation of membrane phospholipids, in a similar fashion to what observed in ABHD12 knockout mice. 198 Inhibitor 56 and structurally similar thiourea derivatives are also reported in a patent of the Scripps Research Institute, in which these ABHD12 inhibitors were claimed as useful for the treatment of cancer, neuropsychiatric disorders, and neurodegenerative, autoimmune, neuroinflammatory, and infectious diseases. 199 Besides 56, the most promising compounds are compounds 57, 58, and 59 ( Figure 14), which displayed IC 50 values lower than or equal to 100 nM for ABHD12 inhibition in competitive ABPP assays and in a substrate-based assay by using HEK293T cells overexpressing ABHD12.
In 2020, Lundbeck La Jolla Research Center Inc. published a patent including pyridinyl urea derivatives as ABHD12 inhibitors, which may be useful for the treatment of cancer and infectious, neurodegenerative, autoimmune, and neuroinflammatory diseases. 200 The most potent inhibitor of this series was compound 60 (Figure 14), possessing the same structure of compound 57 (Figure 14) with the exception of the urea instead of the thiourea moiety. This pyridinyl urea inhibited mouse brain ABHD12 with an IC 50 value lower than or equal to 100 nM, and it showed an ABHD12 inhibition activity greater than or equal to 75% when tested at 1 μM in mouse brain membrane proteomes.

ABHD16A
11.1. Biochemical Features and Biological Roles. ABHD16A is also named Human Lymphocyte Antigen Bassociated transcript 5 (BAT5), and it is composed of 558 residues (63 kDa). It is a poorly known serine hydrolase, whose highest expression was observed in skeletal muscle and brain. 201 ABHD16A is localized in the plasma membranes in human platelets and mouse megakaryocytes. 202 It is palmitoylated; however, further investigation about this modification was not performed. 203 The substrate preference for ABHD16A was investigated by Savinainen et al., and it was found that ABHD16A acts as a hydrolase for medium-chain saturated MAGs, long-chain unsaturated MAGs (in particular 1linoleylglycerol, 1-LG) as well as the 15-deoxy-Δ 12,14prostaglandin J2-2-glycerol ester (15d-PGJ2-G). 204 Polymorphisms of ABHD16A are correlated to Kawasaki syndrome, a disease characterized by vascular inflammation, which may cause coronary artery aneurysm formations and cardiac complications. 205 In pigs, the polymorphism of ABHD16A was related with back fat thickness, thus suggesting its potential role as a marker associated with obesity. 206 As already anticipated in subsection 9.1, ABHD16A is implicated in immunoregulation together with ABHD12, since both regulate lysoPS metabolism in vivo. In particular, ABHD16A regulates the lysoPS-induced release of proinflammatory cytokines from macrophages. 186 The involvement of ABHD16 in immunoregulation originates from studies regarding its gene location, considering that ABHD16A belongs to a cluster of genes within the human major histocompatibility complex class III. 207,208 Moreover, the expression of ABHD16A could influence the immunogenicity of bone marrow cells in mice. 209 11.2. Inhibitors. In 2014, the first ABHD16A inhibitor was reported, the β-lactone palmostatin B 61 (Figure 15), which inhibited the hydrolysis of the MAG substrate 1-LG inhibitor in HEK293 lysates transfected with human ABHD16A, with an IC 50 value of 100 nM. 204 Considering thatpalmostatin B 61 was first developed as a LYPLA1 inhibitor (IC 50 = 670 nM), 210 further assays to determine its selectivity disclosed that compound 61 dose-dependently inactivated not only LYPLA1/2 and ABHD16A, but also ABHD12 (IC 50   Competitive ABPP assays on mouse brain membrane proteome suggested that, at 1 μM concentration, 1,3,4oxadiazol-2(3H)-ones 63 selectively inhibited ABHD16A with no cross-reactivity over other serine hydrolases such as FAAH, KIAA1363, MAGL, and LYPLA1/2 as they were offtargets of 62, yet analogue 64 showed only activity against KIAA1363.
Considering the electrophilic nature of the β-lactone group, which is suitable to bind to the active site nucleophilic serine, a screening of α-alkylidene-β-lactone-based library of compounds 211 led to the identification of derivative 65 (KC01, Figure 15) as an inhibitor of human and mouse ABHD16A, with IC 50 values in the range 0.2−0.5 μM in competitive ABPP assays. 186 A comparable inhibition activity was detected when a PS substrate was used in the membrane proteome of ABHD16A-transfected HEK293T cells, obtaining IC 50 values of 90 and 520 nM for human and mouse ABHD16A, respectively. Further assays were performed in the membrane fraction of COLO205 (colon cancer), K562 (leukemia), and MCF7 (breast cancer) cancer cells, confirming the in situ inhibition of ABHD16A by 65. Quantitative ABPP-MS experiments revealed that compound 65 also inhibited ABHD2 (94% at a concentration of 1 μM) as well as many other targets such as ABHD3, Patatin Like Phospholipase Domain Containing 4 (PNPLA4), PAFAH2, ABHD6, ABHD13, ABHD11, and LYPLA1 although to a minor extent (14−80%), thus revealing a low selectivity for ABHD16A. Furthermore, treatment of COLO205, K562, and MCF7 cells with 65 showed significant reductions in the levels of cellular lysoPSs compared to untreated cells, whereas different lipids were unaffected. Importantly, pretreatment of macrophages with 65 reduced the lysoPS-induced cytokine release, thus affecting immune response. Additionally, inhibitor 65 lowered the elevated lysoPS secretion observed in ABHD12-null cells derived from a PHARC subject. Compound 65 was also patented in 2016 by the Scripps Research Institute and the University of Connecticut. 212 Pharmacological investigation of 65 suggested that ABHD16A plays a key role in the production of lysoPSs in both mammalian cells and in vivo and it should be a suitable target for the development of therapeutic agents to treat PHARC and other neuroinflammatory disorders.
The most recently discovered ABHD16A inhibitor is a diterpenoid of the abietane family, compound 66 (Figure 15), which was selected as the most promising inhibitor from a screening of an in-house library of 50 similar derivatives. 213 Compound 66 led to a 23% remaining activity of ABHD16A when tested in lysates of HEK293 cells transfected with ABHD16A (IC 50 value of 3.4 μM). It was selective over ABHD12 and demonstrated a reversible inhibition, as proved by dilution assays. Interestingly, the authors hypothesized allosteric binding of the compound because 66 reached an incomplete ABHD16A inhibition in all used assay conditions. 12. ABHD17A, ABHD17B, AND ABHD17C 12.1. Biochemical Features and Biological Roles. Very little information is known about ABHD17 enzymes, which are broadly expressed in several cell types. They are localized in the plasma membrane and are identified as proteins able to depalmitoylate N-Ras. 214 Palmitoylation is a biological process in which proteins are modified by the addition of palmitate, thus directing them to the cellular compartments where they play their specific functions. N-Ras is a protein that can promote the development of cancer, so the protein depalmitoylases ABHDs may influence tumor growth.
12.2. Inhibitors. Few inhibitors are reported in the literature for ABHD17 proteins, and among them the first discovered compounds are the unselective palmostatin B 61 (Figure 15), 210 the analogue palmostatin M 67 (Figure 16), 215 and hexadecylfluorophosphonate HDFP 68 ( Figure 16). 216 Palmostatin M 67 has the same β-lactone-based scaffold of palmostatin B, and it is metabolically unstable; moreover, it is unselective, since it also inhibits APT. 215 Hexadecylfluorophosphonate HDFP 68 lacks selectivity, since it inhibits many different targets, above all fatty acid synthase (FASN), FAAH, neutral cholesterol ester hydrolase 1 (AADACL1), MAGL, and LYPLA1/2. 216 In 2021, Remsberg et al. performed a gel-based ABPP screening in a native mouse brain proteome of a serine hydrolase-directed compound library developed at Lundbeck La Jolla Research Center, Inc., followed by a chemical  Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective optimization process aimed at improving selectivity and potency. This strategy led to the identification of a more selective pan-ABHD17 inhibitor ABD957 69 ( Figure 16). 217 Compound 69 blocked covalently (likely due to its ureacontaining moiety) more than 90% all the ABHD17 enzymes in MS-based ABPP experiments in OCI-AML3 (leukemia) cells, with a reported IC 50 value against ABHD17B of 0.21 μM in lysates of HEK293T cells. Nevertheless, compound 69 cannot be defined as highly selective, due to a residual inhibition activity on CES1/2, ABHD6, and ABHD13. The importance of 69 derives from its contribution to the palmitoylation/depalmitoylation cycle, because it partially impaired N-Ras depalmitoylation in human acute myeloid leukemia (AML) cells, mainly affecting plasma membraneassociated dynamically palmitolylated proteins. Its therapeutic significance is linked to the impairment of NRAS mutant cancer cell growth, since the antiproliferative effect was abrogated in cells lacking ABHD17A and ABHD17B, thus confirming the ABHD17 inhibition by 69.

UNSELECTIVE ABHD INHIBITORS
Some well-known nonselective ABHD inhibitors that deserve to be mentioned are included in this section. The most important are methylarachidonoyl fluorophosphonate (MAFP 70, Figure 17), the approved antiobesity drug tetrahydrolipstatin (THL, Orlistat, 71 in Figure 17), the natural compound triterpene pristimerin 72 (Figure 17), and compound RHC-80267 73 (Figure 17), and, with the exception of 71, these compounds are generally used as pharmacological probes for biological studies rather than hit compounds to be developed as potential new drugs, because of their broad-spectrum activity. MAFP 70, a FAAH inhibitor (IC 50 = 2.5 nM), 218 inhibited not only ABHD6 (IC 50 = 0.017 μM) but also ABHD12 (IC 50 = 0.087 μM) 73 as well as other serine hydrolases (i.e., KIAA1363 and MAGL), and it was a cannabinoid receptor modulator. This compound was also reported in a patent dated 2017 concerning the progesterone activation of ABHD2. 219

CONCLUSIONS AND FUTURE PERSPECTIVES
The family of serine hydrolases is one of the most diversified and numerous existing classes of enzymes. More than 200 enzymes belong to this class and all share a typical feature that is the presence of an active site serine, which is fundamental to catalyze the hydrolysis of substrates. Among serine hydrolases, ABHD proteins play many important roles in a wide range of pathophysiological processes and, thanks to their multifaceted roles, they have progressively acquired more importance in the scientific community. In this Perspective, the main members of the ABHD family are described for what concerns their biochemical role as well as their involvement in human diseases. Many ABHDs are involved in lipid metabolism, thus affecting obesity and fat-related diseases; moreover, some ABHD mutations are often correlated to rare genetic diseases. ABHD6 and ABHD12 are strictly connected to ECS since they are implicated in 2-AG metabolism. In addition, immunoregulation, cancer, and neurodegeneration are affected by downregulation or overexpression of ABHD enzymes. Since 2007, many compounds targeting ABHD enzymes have been developed with the aim of blocking ABHD enzymes. Considering all the enzymes of the ABHD family, the research efforts led to a quite limited number of inhibitors, restricted to ABHD2, ABHD3, ABHD4, ABHD5, ABHD6, ABHD9, ABHD10, ABHD11, ABHD12, ABHD16A, and ABHD17A-C likely due to a scarce characterization of the remaining ABHD enzymes. An exception is represented by modulators of ABHD5 7−9 which are efficient modulators of cellular lipolysis; however, they do not block ABHD catalytic activity but rather they are allosteric ABHD5 ligands. Future research lines should be aimed at characterizing the other less known members of the ABHD family, that are ABHD1, ABHD7, ABHD8, ABHD12B, ABHD13, ABHD14A, ABHD14B, Journal of Medicinal Chemistry pubs.acs.org/jmc Perspective ABHD15, ABHD16B, and ABHD18, to completely understand the complex biological roles of ABHD proteins. The identification of selective ligands for these proteins may help to fill a knowledge gap on endogenous lipid biosynthesis and metabolism as well as physiological and pathophysiological functions of these proteins in humans.
Other ABHD enzymes play key biological roles, although no selective ligands have yet been developed, and therefore, they are not extensively reviewed in this Perspective. For instance, ABHD1, whose biochemical function is not fully discovered, is ubiquitously expressed in murine and human tissues, reaching the highest expression in testis. 221 ABHD1 overexpression was observed in a renal cell line: it contributed to the reduction of reactive oxygen species formation by NADPH oxidase, thus contributing to the protection against oxidative stress. For this reason, upregulation of ABHD1 in D5 dopamine receptor deficient mice, which develop hypertension and increased systemic oxidative stress, may contribute to the protective mechanism against oxidative stress. 222 The herein reported ABHD-targeting compounds belong to several chemical classes ( Figure 18). It is evident that most of them are urea-or carbamate-based molecules, since they exploit the intrinsic reactivity of active-site nucleophilic serine by promoting its acylation. Nevertheless, also lactones and lactams were expected to irreversibly bind to ABHD enzymes. Differently, compounds derived from natural sources, MIDAboronates, and heterocycle-containing derivatives in general act as reversible inhibitors.
The structural similarity among ABHD proteins increases the risk of developing compounds inhibiting more ABHD enzymes, thus losing selectivity, and this may represent an obstacle to the development of an ideal clinical candidate. Nevertheless, considering the overlapping roles of some ABHD enzymes, the low selectivity of most ABHD inhibitors could be considered an opportunity to simultaneously interfere with a pathophysiological process by blocking different strategic targets.