SK1-I (BML258), 1 (2R, 3S, 4E)-N-methyl-5-(4-pentylphenyl)-2-aminopent-4-ene-1, 3-diol, is sphingosine analogues discovered in 2008 by Paugh et al. (Figure 3A) (Paugh et al., 2008). SK1-I is a competitive and selective SphK1 inhibitor with Ki value of 10 μM (Pitman and Pitson, 2010), and has been widely used to elucidate the role of SphK1 in cancer. In the mouse model of breast cancer, SK1-I decreased serum S1P level, stimulated cancer cell apoptosis, and reduced angiogenesis and lymphangiogenesis (Nagahashi et al., 2012). SK1-I also reduced S1P levels and increased the expression of Cer derivatives in human leukemia U937 cells by inhibiting SphK1, which is related to the decrease of ERK1/2 and Akt signals (Paugh et al., 2008). Recent studies have shown that SK1-I increased the transcriptional activity of tumor suppressor protein TP53 and the expression of pro-apoptotic members of the downstream BCL2 family, thereby enhancing autophagy and cancer cell death (including colon cancer and breast cancer) in a SphK1 dependent manner (Lima et al., 2018). In addition, SK1-I negatively regulates the expression of MMPs and cell migration and invasion, which is related to the alleviation of various pregnancy related diseases (Chahar et al., 2021). Compared with other SphK1 inhibitors, SK1-I has high solubility and allowed its delivery in vivo, which makes it an ideal SphK1 inhibitor and used in a wide range of animal models. However, some studies have found that SK1-I plays a cytotoxic role in AML models (Paugh et al., 2008), suggesting that more research is needed for the cytotoxic effects to be applied in disease treatment.

LCL351 and LCL146 are potent SphK1 inhibitors that are Sph analogues. Unlike the Sph structure, their polar head amino group is replaced by a guanidine group. The chemical names are L-erythro-2-N-(1’-carboxamidine)- sphingosine hydrochloride and D-erythro-2-N-(1’-carboxamidine)-sphingosine hydrochloride, respectively (Figure 3B). They act as two erythro diastereoisomers, showing different inhibitory potency due to different stereochemistry at C2 and C3 positions. LCL351 was found to be more potent than LCLl46, suggesting a stereochemical interaction with target cells. LCL351 was identified as a SphK1 selective inhibitor with Ki values of 5 and 50 μM against SphK1 and SphK2, respectively. In vitro studies showed that LCL351 induced SphK1 degradation and had no effect on cell death and cell cycle, thus it may not cause side effects (Sharma, 2011). Studies in a mouse model of IBD found that LCL351 reduced plasma S1P levels as well as the expression of pro-inflammatory cytokine, alleviated neutrophil infiltration and immune responses, and exerted therapeutic effects (Pulkoski-Gross et al., 2017).

This novel class of SphK1 selective inhibitors was engineered on the basis of Sph to design a series of compounds that replace the aminodiol headpiece of Sph with a serine amide (Figure 3C). Such a structure not only increases the affinity of new molecules to SphK1, the carboxylic acid of serine also provides a convenient synthetic site to bind various mimetics for the lipophilic tail of Sph. The activity was significantly increased when L-threonine was used as the polar headpiece, and the resulting compound 6ag was nearly 10 times higher. To explore the effect of the distance between the terminal alcohol and the amide group, a series of compounds were prepared and found that the greater the distance, the lower the inhibitor potency. Moreover, the potent of S-enantiomer 9ab was approximately 40 times that of R-enantiomer (50 nM vs. 2.2 μM), indicating that the stereochemistry of homoserine analogues has a significant effect on activity. Compound 12aa is a modification of the polar headpiece with 3-hydroxyproline that is much more potent. In this series of inhibitors, the amide functionality was critical for inhibitor potency and significantly more potent than the previously reported SphK1 inhibitor N, N-dimethylsphingosine. Although no more studies on this series of compounds have been reported, 3-hydroxyproline provides more possibilities for the development of selective inhibitors to improve the activity and ADME properties in vitro (Xiang et al., 2009).

The design and biological activity evaluation of a series of amide SphKs inhibitors were reported by Frank W et al. (Figure 3D) (Foss et al., 2009). It is worth noting that VPC45129 is the first alcohol compound with significant activity on SphK1. This special compound containing alcohol is attractive because few synthetic analogues are active on SphK1. Compound VPC94075 has dual inhibitory effects on SphK1 and SphK2 with Ki values of 55 and 20 μM, respectively, and has the effects of reducing S1P level and anti-proliferation in vitro. Therefore, a series of amidine-based compounds were synthesized and optimized as SphKs inhibitors with VPC94075 as the leading compound.

VPC96091, an effective selective SphK1 inhibitor, was found to be further developed through the above inhibitors. It chemical name is (S)-1-(4-dodecylbenzoyl)pyrrolidine-2-carboximidamide hydrochloride, which is characterized by a terminal α-substituted amino group linked with 4-alkyl phenyl via amide bond. The Ki values for SphK1 and SphK2 were 0.10 and 1.50 μM, respectively. Selective inhibition of SphK1 by VPC96091 reduced epidermal growth factor (EGF) driven S1P levels and increased Akt/ERK phosphorylation in human leukemia U937 cells and mice model (Kharel et al., 2011). In the patent reported by the University of Virginia, a series of SphKs inhibitors with amide bonds are described, including aminooxime analogues 5. It is different from VPC96091 because the terminal α-substituted amino group is linked by a linear alkyl group (University of Virginia Patent Foundation, 2014). The conversion of amidines into amidoximes can improve the oral bioavailability of drugs according to the prodrug principle (Kotthaus et al., 2011).

SLP7111228, an effective selective inhibitor of SphK1 with Ki value of 48 nM, which was obtained after modification of SphK2 inhibitor SLP120701. Its chemical name is (S)-2-((3-(4-octylphenyl)-1,2,4-oxadiazol-5-yl)methyl)pyrrolidine-1-carboximidamide hydrochloride and its structure is based on guanidine, which was published by Patwardhan et al. (Figure 3E) (Patwardhan et al., 2015). Studies have confirmed that SLP7111228 can reduce the level of S1P in U937 cells, and the same effect has been proved in vivo studies of mice and rats models. Another compound in the same series, SLP080801, is defined as a selective SphK2 inhibitor that improves S1P levels in the circulation of mice (Kharel et al., 2012). The inhibition or gene deletion of SphK2 is the key to the increased accumulation of S1P in the blood, but its mechanism is still unclear. This suggests that this increase may be regulated by the dynamic balance between SphK1 and SphK2, which are responsible for the synthesis and elimination of S1P in the blood in some way.

Compounds 23 and 28 are the most interesting of the amido derivatives, which are derived from VPC45129 step by step (Figure 3F). Mathews et al. found moderate inhibition of SphKs after catalytic ring-opening and re-arrangement. Then, the methyl group was replaced by cyclopropyl ring to increase the steric bulk of the junction region, and some cyclopropyl derivatives with different tail lengths were evaluated in order to determine the optimal tail length. It is worth noting that these compounds retain the unique chemical properties of amides, which is very important for inhibitory potencies. Finally, an effective dual SphK1/2 inhibitor (Compound 23) and an effective selective SphK1 inhibitor (Compound 28) were identified after evaluating the inhibitor potency, selectivity and in vitro activity of the inhibitors. The Ki values of compound 23 for SphK1 and SphK2 were 0.2 and 0.5 μM, respectively. Compound 28 showed good selectivity for SphK1, with Ki values of 0.3 μM for SphK1 and 6 μM for SphK2. Specificity evaluation did not find the inhibitory effect of the two compounds on diacylglycerol kinases (γ, δ1, ζ) and protein kinase C family, PKC α. The availability in smooth muscle cell-based disease models demonstrates the pharmacodynamic potential of these novel SphKs inhibitors. In addition, more studies have begun to identify pharmacophores related to aminosphingosine kinase inhibitors (Mathews et al., 2010).

Compound 82 is a selective SphK1 inhibitor obtained by optimizing the structure of dual SphK1/2 inhibitor SKI-Ⅱ. It competitively inhibits SphK1 with IC50 of 0.02 and 0.10 μM for SphK1 and SphK2, respectively (Figure 3G) (Gustin et al., 2013). Hydrogen bonds are formed by the aminoalcohol portion of compound 82 and two key residues of Asp in SphK1, which play a key role in the interaction between the compound and the target molecule, just like PF-543 (Wang et al., 2013). Asp178 forms hydrogen bonds with nitrogen on the piperidine ring and hydroxyl outside the piperidine ring, respectively. Asp81 also forms hydrogen bonds with hydroxyl on the piperidine ring. In addition, in vivo and in vitro pharmacokinetic parameters showed good mean residence time (MRT) and the bioavailability of 32% (Gustin et al., 2013). Pharmacological studies have shown that compound 82 can reduce the level of S1P in human breast and melanoma cell lines, but has no effect on the growth of tumor cells (Rex et al., 2013). Unfortunately, these results have not been confirmed in vivo studies.

RB-005, chemical name 1-(4-octylphenethyl)piperidin-4-amine, is a derivative obtained from the route of synthesizing FTY-720 from 4-octylphenylethanol (Figure 3H) (Baek et al., 2013a). RB-005 was identified as a selective SphK1 inhibitor based on Sph with IC50 of 3.6 µM. The compound has an n-octylphenyl group linked in a 2-carbon tether to the nitrogen of 4-hydroxypiperidine. This small change in tertiary amine structure is responsible for supporting that RB-005 maintains the selectivity for SphK1. The hydroxyl groups and size of heterocycles are also crucial for selectivity, such as RB-042, which is converted into a dual inhibitor of SphK1/2 with the IC50 of 2 µM for both isoforms. The specificity of RB-005 for SphK1 was established by a study that found that RB-005 induced SphK1 proteasome degradation in human pulmonary artery smooth muscle cells, which was reversed by proteasome inhibitor MG132 (Baek et al., 2013b). Other studies found that RB-005 inhibition of Cer synthase has an effect on lung and heart remodeling in the hypoxic model of pulmonary hypertension mice (MacRitchie et al., 2016). However, more evidence is needed to prove that the affinity for SphK2 and other bioactive enzymes.

These are a series of selective inhibitors of SphK1 based on the framework of 2-piperidine thiazole, which were published by Merck Serono (Figure 3I) (Frank, 2013). These compounds are usually occupied by a 5,5,8,8-tetramethyl tetralin (Compound 1) at the 4-position of thiazole ring. On this basis, the structure is modified to achieve the goal of additional diversity by linking piperidine at 1-position, replacing piperidine with piperazine and various alkyl groups (Compound 2), or replacing pentaaryl group in previous patents with 2,6-disubstituted pyridine (Compound 3). The publisher has declared that the IC50 value of the above series of compounds is in the range of 1–1,000 nM, but the Ki value has not been described. Therefore, it is difficult to compare the potency of the target compound without specific IC50 and Ki values. Unfortunately, it is unclear whether the action of such compounds is competitive with ATP or Sph, or non-competitive, nor is the use of these SphK1 inhibitors mentioned. But it is exciting that the patent mainly states the treatment of these SphK1 inhibitors for cancer and RA, which makes people look forward to the research on mechanism and pharmacological activity in the near future (Lynch et al., 2016).

PF-543, developed by high throughput screening and medicinal chemistry optimization, is the most effective selective SphK1 inhibitor described so far (Figure 3J). Its selectivity is 100 times higher than that of SphK2 (Ki = 4.3 nM) discovered and reported by Pfizer (Schnute et al., 2012). Its chemical name is (2R)-1-[[4-[[3-Methyl-5-[(phenylsulfonyl)methyl]phenoxy]methyl]phenyl]methyl]-2-pyrrolidinemethanol hydrochloride, belonging to pyrrolidine analogues. Structurally, the basic skeleton of the compound is composed of a tertiary amine which form part of pyrrolidine ring, and the rest maintains the same 1,2-amino alcohol motif as Sph. A recent study shows that pyrrolidine inhibitors, especially with 2-hydroxymethylpyrrolidine structure, may be the key to the inhibitory potency (Li et al., 2021). Subsequently, the crystal structure of a hSphK1 binding to PF-543 was reported by Wang et al. (2014b). It was clarified that the inhibitor was bound in the SphK1 substrate pocket with a J-shaped structure. Specifically, the terminal phenylsulphonyl ring occupied the hydrophobic pocket formed by residues, including Phe374 and Leu347, 354 and 405. The (R)-2- (hydroxymethyl) - pyrrolidine head group is rotated to match the lipid primary hydroxyl group for phosphorylation, in which the hydroxyl and pyrrolidine nitrogen forms hydrogen bonds with the side-chain of Asp264. Finally, the central aromatic ring and its substituted methyl groups interact with Phe and Leu respectively. However, toluene group did not seem to be an essential group, which has no significant effect on SphK1 inhibition and anticancer activity (Kim et al., 2020). Compared with SphK2, SphK1 has three different residues in lipid binding sites, of which Phe374 and terminal phenyl ring may be the most closely bound part. The substitution of Cys in SphK2 by Phe374 may be an important reason for the high selectivity to SphK1 rather than SphK2.

PF-543 induced protein degradation of SphK1, which reduced the level of S1P and increased Sph. However, PF-543 had no effect on Cer level, suggesting that the lack of the ability to induce apoptosis (Byun et al., 2013; Wang et al., 2013). A recent study has proved that PF-543 reduces apoptosis in the lungs of mice after acute ethanol intoxication, and inhibits neutrophil infiltration and the release of inflammatory cytokines to reduce lung injury (Chen et al., 2021b). As the most effective SphK1 inhibitor so far, PF-543 can inhibit inflammation in RA model (Deng et al., 2021), ulcerative colitis model (Liu and Jiang, 2020) and mouse pulmonary hypertension hypoxia model (Ha et al., 2020) in vivo and in vitro, mainly by inhibiting the release of inflammatory cytokines and the change of cell biological function. In addition to inflammatory response, PF-543 plays a therapeutic role in angiogenesis in the pathogenesis of RA and microvascular leakage induced by sepsis (Zhong et al., 2020). Although these evidences show the beneficial effects of PF-543 on the pharmacological inhibition of SphK1, it is worth noting that the administration concentration of PF-543 needs to be further confirmed to ensure the specificity of SphK1.

A recent study found that a SphK1 selective inhibitor CHJ01 showed a unique therapeutic effect on RA. CHJ01 is an analogue of jaspine B, which is a naturally occurring anhydrophytosphingosine derivative isolated from the Okinawan marine sponge Pachastrissa sp. and Jaspis sp (Figure 3K). It can reduce intracellular S1P level and increase Cer level by inhibiting SphK1 (Kuroda et al., 2002; Salma et al., 2009). The hydrochloride of CHJ01 obtained by structural optimization has a good inhibitory effect on SphK1, but almost no inhibitory effect on SphK2 with the IC50 of 8.89 µM. Structurally, the replacement of the furan ring by a pyrrole ring in CHJ01 may account for the stronger activity. Pharmacological experiments showed that CHJ01 exhibited anti-inflammatory effects similar to those of methotrexate in vitro and in vivo by reducing the swelling volume, arthritis score, spleen index and IL-1β, TNF-α, IL-6 levels in arthritis model rats, which contributed to the significant improvement of RA symptoms (Chen et al., 2021a).

A series of potent SphKs inhibitors based on an N-(5-alkyloxadiazol- 3-yl)benzyl)-3-hydroxypyrrolidine-2-carboxamide scaffold proposed by Genzyme, structurally similar to PF-543 but with better metabolic stability (Figure 3L) (Xiang et al., 2010). For example, the pyrrolidine group`of compound 17 was replaced by a hydroxyl group instead of methanol, in contrast to PF-543. Further optimization of compound 17 by Xiang et al. found that shortening of the straight alkyl chain leads to loss of inhibitor viability. In order to maintain activity, the attachment of a two-carbon alkyl linker between the oxadiazole ring and the carbocyclic ring increased activity as well as solubility. The resulting compound 51 is characterized by a cyclopentethyl group attached on the oxadiazole ring, chemical name (2S,3S)-N-((S)-1-(4-(5-(2-cyclopentylethyl)-1,2,4-oxadiazol-3-yl)phenyl)ethyl)-3-hydroxypyrroli-dine-2-carboxamide, showed excellent inhibitory effect on SphK1 with the IC50 of 0.058 μM. The pharmacokinetics study of compound 51 found the moderate oral bioavailability, qualified half-life in blood circulation, and good internal clearance, particularly for human liver microsomes (Table 1).

SKI-I, (N’-[(2-hydroxy-1-naphtyl)methylene]-3-(2-naphthyl)-1H-pyrazole-5-carbohydrazide), is also a competitive Sph inhibitor of SphK1, which was found by French et al. (Figure 4A) (French et al., 2003). SKI-I inhibit SphK1 competitively with IC50 of 1.2 μM, but it also inhibit SphK2 with similar affinity, and cross react with ERK2, PKC and PI3K (Hengst et al., 2010). Many in vitro and in vivo studies showed that SKI-I not only down-regulates S1P level and up-regulates Cer level, but also induces apoptosis and autophagy in T24 bladder cancer cells and mouse embryonic fibroblast (Young et al., 2012). In addition, the compound showed anti-tumor activity in mouse melanoma and breast cancer xenograft models (French et al., 2006).

SKI-Ⅱ, 2-(p-hydroxyanilino)-4-(p-chlorophenyl)thiazole, was identified from a high throughput screening by French et al. (Figure 4B) (French et al., 2003). As non-lipid small molecule compound with dual SphK1/2 inhibitors, SKI-Ⅱ showed stronger inhibitory effect of SphK1 than SphK2 (Ki (SphK1) = 16 μM, Ki(SphK2) = 8 μM). Similar to SKI-I, SKI-Ⅱ attenuates SphK1 signaling by triggering lysosomal degradation of SphK1 in different cell types, mainly by binding to an allosteric site, which is different from the enzyme sites promoting polyubiquitination and proteasome degradation of SphK1 (Loveridge et al., 2010; Ren et al., 2010). SKI-Ⅱ has no inhibitory effect on ERK2, PKC and PI3K compared with SKI-I. SKI-Ⅱ induces apoptosis and inhibits proliferation and migration by increasing Cer/Sph and reducing S1P levels (Bien-Möller et al., 2016; Sun and Wang, 2021). The effect of reducing inflammation and acute myeloid leukemia (AML) in vivo has been reported due to oral bioavailability of SKI-Ⅱ (Yang et al., 2015). Recent studies have shown that SKI-Ⅱ inhibit the activity of dihydroceramide desaturase-1 (Des1) with Ki of 0.3 μM, which is related to the prevention of oxidative stress (Cingolani et al., 2014; Noack et al., 2014). SKI-Ⅱ stabilize the redox sensitive transcription factor nuclear factor-erythroid-2-related factor 2 (Nrf2) and play a protective role in diseases caused by oxidative stress by destroying the negative regulator Keap1 (McNaughton et al., 2016). Aurelio et al. further proved that the anti-proliferative effect of SKI-Ⅱ and its analogues is mainly due to Des1 inhibition rather than the binding of allosteric sites (Aurelio et al., 2016).

MP-A08, chemical name 4-methyl-N-[2-[[2-[(4-methylphenyl)sulfonylamino]phenyl]iminomethyl]phenyl]benzenesulfona-mide, found by homology modeling of the ATP binding site of SphK1 (Figure 4C) (Pitman et al., 2015). MP-A08 is an ATP competitive dual SphK1/2 inhibitor with corresponding Ki values of 27 and 7 μM. This indicates that MP-A08 targets both SphK1 and SphK2, but has higher affinity for SphK2. Some SphKs inhibitors, such as SKI-Ⅱ and PF-543, have been found to inhibit SphK1 in cells by targeting proteasome degradation, which seems to be a feature of many Sph competitive SphKs inhibitors. However, MP-A08 had no effect on SphK1 degradation, which further proved that the inhibitor had higher affinity with SphK2. On the other hand, early SphK1 inhibitors mostly acted against the Sph binding pocket, which is highly likely to have off-target effects due to the molecular structure retaining Sph characteristics (Cingolani et al., 2014). Whereas the selectivity for SphK1 and SphK2 is higher than other kinases, because MP-A08 targeting the ATP-binding pocket is structurally different from protein kinases and almost all other lipid kinases. In brief, MP-A08, an inhibitor developed to target the ATP-binding pocket of SphK1, which not only takes advantage of the known divergence of the SphK1 ATP-binding site from other protein kinases to improves selectivity, but also overcomes off-target effects common to molecules like Sph. MP-A08 increased Sph and Cer levels and decreased S1P levels, shifting tumor cells from an anti-apoptotic, pro-proliferative to a pro-apoptotic and anti-proliferative phenotype and blocking the survival of multiple cancer cell lines (Pitman et al., 2015).

The D, L-threo-dihydrosphingosine (DHS), a SphK1 competitive inhibitor with Ki of 3–6 μM, is the first SphK1 inhibitor reported in the literature (Figure 4D) (Coward et al., 2009). Later studies have shown that this compound can also participate in sphingolipid metabolism pathway as SphK2 substrate, and inhibit other kinases (such as PKC-α), to produce other non-target effects. DHS, also known as Safingol, has a large number of experimental researches before clinical trials. It enhances the anti-tumor effect of various chemotherapeutic drugs (such as doxorubicin) in vitro by inducing apoptosis (Schwartz et al., 1997). As the first SphKs inhibitor to enter clinical trials as an anticancer agent, it has obvious anticancer activity in vitro and can be safely administered in combination with cisplatin (Dickson et al., 2011).

DMS, N,N-dimethyl-D-erythro-sphingosine, is the first direct SphKs inhibitor with Ki of 16 μM at SphK1, which is similar to DHS (Figure 4E) (Yatomi et al., 1996). It blocks the activity of the two isozymes by competing with the natural substrate Sph. DMS inhibit the growth of tumor, induce apoptosis of cancer cells, and block PKC signaling (Edsall et al., 1998). In addition, DMS has been shown to inhibit SphK2 and ceramide kinase (CerK), which makes it unable to act on SphK1 specifically. Although the inhibitors have been found to have therapeutic effects on many types of cancer cells and tumors in vitro and in vivo, but severe hemolysis in mice is triggered (Sweeney et al., 1996; Sah et al., 2021). Obviously, the non-targeting effect of sphingolipid analogue inhibitors makes it unable to be an ideal SphKs inhibitor (Table 1).