Introduction

Glycogen synthase kinase-3β (GSK-3β) is a highly conserved multifunctional serine/threonine (Ser/Thr) protein kinase widely expressed in many tissues, which was first identified over 30 years ago [1]. GSK-3β was originally thought to regulate hepatic glycolysis by phosphorylating and inhibiting hepatic glycogen synthase, but now more studies have shown that GSK-3β was involved in multiple key cell-signaling pathways and had various phosphorylation targets [2].

Due to its versatility, GSK-3β plays important roles in the occurrence and development of human key diseases (Fig. 1), such as cancer, Alzheimer’s disease (AD), Parkinson’s disease, stroke, inflammation, type-II diabetes, and so on [3,4,5,6], therefore, GSK-3β inhibitors could be employed to target these kinds of diseases. For example, many small-molecule inhibitors of GSK-3β were explored for their effects in cancer treatment [7]. Recently, Sahin et al. outlined the prospects of GSK-3β in cancer therapy and summarized the preclinical and early clinical results of these potential anti-cancer GSK-3β inhibitors [8]. As a key regulator of cognitive function, GSK-3β has been involved in the process of neurodegenerative diseases, such as AD [9, 10]. Studies have shown that inhibiting GSK-3β can improve memory deficits and cognitive function of the aging mice model [11]. Besides, inhibition of GSK-3β is thought to reduce acute myocardial infarction [12], suppress inflammation [13], protect the lung from acute injury [14], delay pulmonary fibrosis [15], and improve severe cardiac dysfunction [16]. Moreover, some recent studies indicated that inhibition of GSK-3β could prevent liver fibrosis [17], treat vascular dementia [18], and target rheumatic disease [19].

Fig. 1
figure 1

The correlations of GSK-3β with various diseases

Full size image

The discovery of GSK-3β inhibitors in pharmaceutical companies and academic institutes has been carried out over the years, and its advances in the research of GSK-3β inhibitors were reviewed by several articles [20,21,22,23]. Despite massive efforts in the research and development of GSK-3β inhibitors in the past decades, only an irreversible and time-dependent GSK-3β inhibitor, tideglusib (a TDZD compound, IC50 = 100 nM), has entered the phase II clinical trials for the treatment of AD and progressive supranuclear palsy [24, 25]. The current dilemma pushes pharmaceutical chemists to look for new selective and reversible GSK-3β inhibitors, which might be derived from virtual screening, rational drug design, or the bioassay of natural products database. Herein, the present view aims to summarize the recent development of new GSK-3β inhibitors reported over the past five years (2015–2019) and to provide information for the future discovery and design of GSK-3β inhibitors.

The recent development of GSK-3β inhibitors

Curcumin-derived GSK-3β inhibitors

Curcumin is a β-diketone natural product isolated from the rhizomes of Curcuma longa L [26], and it has been investigated as an effective and safe versatile molecule for preventing and treating various disorders [27, 28]. Based on the analysis of structural features and biological profile of curcumin, Belluti et al. envisioned that the α,β-unsaturated carbonyl moiety of curcumin could covalently interact with the key residue (Cys199) of GSK-3β, and thus prepared a small library of curcumin-based analogs [29]. The bioassay results indicated that some of the analogs, such as 1a–1f shown in Table 1, were potent dual-target inhibitors, which not only target GSK-3β (IC50 values of 0.04 ± 0.01–2.69 ± 1.01) but also BACE-1 (IC50 values of 0.53 ± 0.27–2.78 ± 0.44) in a potent and balanced way, making curcumin a promising scaffold for developing anti-AD drug candidates by dual targeting GSK-3β and BACE-1. The structure–activity relationship (SAR) study proved that the tautomeric enol form is an important feature to achieve good chemical stability and metabolic stability. Among them, the double-sided p-methoxy-substituted compound 1a showed the best inhibitory activity with an IC50 value of 0.53 μM. However, the activity was found to decrease slightly when p-methoxy group in 1a was replaced by other substituents.

Table 1 Curcumin-based analogs 1a–1f
Full size table

Amaryllidaceae alkaloids

Amaryllidaceae alkaloids are a group of nitrogen-containing polycyclic compounds produced exclusively by the plants of the Amaryllidaceae family. As one of the best known amaryllidaceae alkaloids, galanthamine is a potent acetylcholinesterase (AChE) inhibitor clinically used in the treatment of AD. Recently, Cahlikova and coworker explored the potency of amaryllidaceae alkaloids on GSK-3β inhibition [30]. Consequently, twenty-eight amaryllidaceae alkaloids of seven structural types obtained from different Amaryllidaceae plants were tested for their inhibitory activity against GSK-3β, and three alkaloids including caranine (2a), 9-O-desmethylhomolycorine (2b), and masonin (2c) were found to be the most active with IC50 values around 30 μM (Table 2). Since amaryllidaceae alkaloids can be easily obtained in large quantities from natural resources, they could serve as parent compounds used in further structural modification and biological optimization.

Table 2 Amaryllidaceae alkaloids GSK-3β inhibitors 2a–2c
Full size table

6-C-glycosyl flavonoids

Isoorientin (3, Fig. 2) as known 6-C-glycosoflavone isolated Corn silks based on bioassay-guided method, was recently found to be a new non-ATP competitive GSK-3β inhibitors with IC50 values of 185 μM [31]. This compound could in vitro inhibit GSK-3β-mediated tau hyperphosphorylation and prevent β-amyloid (Aβ)-induced neuronal cytotoxicity in SH-SY5Y cells at the tested concentration of 112 μM, suggesting that compound 3 represents a promising lead candidate for novel anti-AD drug development. To increase the druggable potency of isoorientin, the bioactive improvement including structure–activity relationship (SAR), in vitro and in vivo study of isoorientin analogs was performed by Li’s group [24]. Finally, a new class of selective substrate-competitive GSK-3β inhibitors was developed. In cells, these new inhibitors exhibited membrane-passive permeability, and attenuated GSK-3β-mediated tau protein hyperphosphorylation and Aβ-induced neurotoxicity. SAR analysis and computer simulation proved that the hydrophobicity, π-cation and orthogonal multipolar interaction of compound 4 (IC50 of 0.59 μM) with substrate sites of the enzyme were the key factors for GSK-3β inhibition and selectivity.

Fig. 2
figure 2

Structures of isoorientin (3) and its derivative 4

Full size image

Isonicotinamides

A set of isonicotinamides 6 [32] shown in Fig. 3 were derived from GSK-3β inhibitor 5 by processing pyrrolopyridinone core [33]. Replacement of the lactam part with the pyridine ring was expected to keep the hydrogen-bond interaction with Lys-85 of GSK-3β and improve CNS penetration. In addition, the pyrrole moiety was further replaced with an amide group to make the rigid tricyclic scaffold more flexible. These new synthetic derivatives showed good GSK-3β and cell potency. However, most of the potent compounds had poor metabolic stability in rodent liver microsomes or had a high KBV efflux ratio, except compounds 6a–6h. These isonicotinamides had good Caco-2 permeability and excellent selectivity towards GSK-3β over almost the other 400 kinases. Although these compounds had a high in vivo clearance rate in the triple-transgenic AD mouse model, they exhibited good in vivo efficacy to lower tau protein phosphorylation levels.

Fig. 3
figure 3

Design and structures of the 3-pyridinyl isonicotinamides 6

Full size image

GSK-3β inhibitors with indazole core

To discover GSK-3β inhibitors acting as mood stabilizers, Reggiani and coworkers identified a novel 1H-indazole-3-carboxamide GSK-3β inhibitor 7 through virtual screening [34, 35]. Based on the analysis of the X-ray co-crystal structure of 7 with GSK-3β, the hit 7 was structurally optimized by changing the substituents on C-5 of indazole and piperidine ring [34], which led to the production of a series of more potent N-[(1-alkylpiperidin-4-yl)methyl]-1H-indazole-carboxamide ATP-competitive GSK-3β inhibitors 8 (Fig. 4).

Fig. 4
figure 4

1H-indazole-3-carboxamide GSK-3β inhibitors 7 and 8

Full size image

The analysis of the available SAR information suggested that the benzyl- or pyridine-substituted piperidine were favorable for the kinase inhibitory activity (IC50 values ranging from 0.64 to 0.25 μM). Besides, 5-methoxy, 5-aryl, or 5-heteroaryl in indole ring could significantly increase the inhibitory effect, while 5-aryl or 5-heteroaryl substituents are more preferable compared with 5-methoxy group. In contrast, the compounds having methyl or and 4-methoxyphenyl in piperidine ring showed a significant loss of efficacy. A total of nine most potent compounds 8a–8f and 8h–8j were later selected to be evaluated for their inhibitory effect against Tau protein phosphorylation in cells. The results showed that all of them, except for 8i and 8j, showed potent inhibitory effect with IC50 in the low micromolar range.

Benzimidazole core- and oxazolo[4,5-b]pyridine core-based GSK-3β inhibitors

The combination of benzimidazole core with different functional groups has led to various potent GSK-3β inhibitors. Recently, Hamid’ group made a fruitful contribution to the development of benzimidazole-derived GSK-3β inhibitors. In 2016, they reported two types of benzimidazole GSK-3β inhibitors based thiadiazole (9) and carbohydrazide (10) conjugates (Fig. 5), and evaluated their anti-depressant activity. Benzimidazoles 9a, 9b, and 10a–10c were found to be the most potent inhibitors with IC50 values ranging from 72 to 107 nM, of which 10c was shown to have significant in vivo antidepressant activity in tail suspension and forced swim tests. In the same year, this group prepared a series of benzimidazole-based 1,3,4-oxadiazole-1,2,3-triazole conjugates as GSK-3β inhibitors (11) with in vivo anti-depressant activity [36]. The comparison of bioassay results of compounds 11a–11f indicated that most of the substituents (R) on triazole were beneficial for inhibitory activity, but when R was 2-bromophenyl or 2-methylphenyl, the activity decreased significantly. In 2018, this research group synthesized another set of benzimidazole-based inhibitors (12) with 1,3,4-oxadiazole carboxamides moiety as a side chain [37]. Among 19 synthesized analogs, compounds 12a–12d showed significant inhibitory effects with IC50 values range of 0.13, 0.22, 0.20, and 0.14 μM, respectively, while compound 12g has the worst inhibitory activity (IC50 value of 2.04 μM).

Fig. 5
figure 5

GSK-3β inhibitors 9–14 reported by Hamid’ and workers

Full size image

In addition, the same research group also reported a novel type of GSK-3β inhibitors with oxazolo[4,5-b]pyridine-2-one core [38], which were designed based on their previous studies on anti-inflammatory benzoxazolinone and 2-mercaptobenzoxazole-based 1,2,3-triazole conjugates [39,40,41]. Among these conjugates, compounds 13a–13d were found to be the strongest inhibitors with IC50 values of 0.19, 0.23, 0.31, and 0.37 μM, respectively. SAR analysis shows that pyridyl or 2-halophenyl has positive effect on the activity than phenyl or 2-halophenyl group. Further bioactivity study showed that compounds 13a–13d had significant in vivo anti-inflammatory activity by inhibiting the pro-inflammatory mediators, such as NO, TNF-α, IL-1β, and IL-6.

In 2017, the same group further optimized oxazolo[4,5-b]pyridine-2-one scaffold, and synthesized seventeen novel oxazolo[4,5-b]pyridine-based piperazinamides as GSK-3β inhibitors [42]. Of these synthesized analogs, compounds 14a–14d had the best inhibitory activity against GSK-3β with IC50 values ranging from 0.53, 0.34, 0.39, and 0.47 μM, respectively, while compounds 14e–14h showed poor inhibitory activity, indicating that the type and position of substituents had a greater impact on their activities. Amongst them, 14b was found to inhibit the pro-inflammatory mediators, such as TNF-α, IL-1β, and IL-6, indicating the great values of these oxazolopyridine-based GSK-3β inhibitors in the development of anti-inflammatory agents.

3, 5-Diamino-N-substituted benzamide-based GSK-3β inhibitors

In 2019, Zhong and coworkers [43] reported the virtual screening, design and synthesis of seventy 3,5-diamino-N-substituted benzamides as GSK-3β inhibitors 15 (Fig. 6). All these compounds reserve the 3,5-diaminobenzamide fragment but having different substituents to the amide group. One of them, compound 15a containing N-3-(trifluoromethyl)phenyl substituent has good selectivity and inhibitory effect towards GSK-3β (IC50 = 5.2 μM). Also this compound has the highest cytotoxic effect on human colon cancer cells (HCT-116). Further in vivo experiments demonstrated that 15a has an acceptable selectivity and ADME characters, suggesting its significant guiding value for novel GSK-3β inhibitors used in anticancer candidates discovery. SAR study on this type of inhibitors can be summarized as follows: (1) When the substituent R in the amide group of 15 is a benzene ring, meta-substituent on benzene is the most optimal, followed by para-substituent, and then ortho-substituent; (2) electron-withdrawing group on benzene instead of electron-donating group lead to the increase of activity; (3) the type and number of substituents on the benzene ring have a more significant effect on the compound’s cytotoxicity, for example, the bulky substituent could increase the cytotoxicity towards tumor cells.

Fig. 6
figure 6

Structures of 3,5-diamino-N-substituted benzamides 15

Full size image

Maleimide-derived GSK-3β inhibitors

Among various reported GSK-3β inhibitors, maleimide-based analogs, such as SB415286 (anilinomaleimide) and SB216763 (arylindolemaleimide), are well-known representative potent, ATP-competitive GSK-3β inhibitors. These types of inhibitors have been shown to have potential applications in the treatment of many diseases, such as cancer, diabetes, and AD [44]. Up to date, lots of maleimide-derived GSK-3β inhibitors have been revealed and structurally modified to develop chemotherapeutical drugs.

Kalam and coworkers [45] synthesized a series of anilinomaleimide-based GSK-3β inhibitors (Fig. 7). Most of these synthetic compounds showed moderate to strong inhibitory activity against GSK-3β, with 1/3 of them having IC50 values less than 1 µM. The preliminary SAR study indicated that replacement of H (R2) on the maleimide core by methyl led to the obvious loss of activity. In addition, the substituents (R1 and R3) on the benzene ring also have influence on their activity, for example, the compounds with an electron-donating group (MeO–) showing much better GSK-3β inhibitory activity than those with electron-withdrawing groups (–NO2, –F, –Cl). Among these derivatives, compound 16b exhibited the highest activity with an IC50 value of 0.09 µM. The further functional study proved that compounds 16b and 16c exhibited significant antidepressant activity by 1.4-fold than the positive control fluoxetine.

Fig. 7
figure 7

Structures of anilinomaleimide-based GSK-3β inhibitors 16

Full size image

Pan et al. discovered a series of (aza)indolyl maleimide-based GSK-3β covalent inhibitors, which were achieved by optimizing non-covalent interactions and reactive groups [46]. Among these inhibitors, compounds 17a–17d in Table 3 showed good selectivity towards GSK-3β, of which compound 17a (IC50 = 6 nM) containing a mild α-fluoromethyl amide reactive group could effectively inhibit the phosphorylation of tau protein and update the expression of β-catenin in living cells.

Table 3 The structures of (aza)indolyl maleimide-based GSK-3β inhibitors
Full size table

A series of new 3-(furo[2,3-b]pyridin-3-yl)-4-(1H-indol-3-yl)maleimide GSK-3β inhibitors (Table 4) was designed and synthesized by the same group [47]. Most compounds showed favorable inhibitory activities against GSK-3β. Based on bioassay results of these compounds, introduction of an alkyl at N1-position of the indole ring made the activity of 18b–18j increase by several folds compared with 18a. With respect to the effects of halogen substitution on activity, it was clearly revealed that the compounds with halogen on 5-position of indole ring obviously showed better activity than those with halogen on the 6-position. In addition, compounds 18g–18i with 3-(1H-imidazol-1-yl)propyl group at N1-position also exhibited potent inhibition on GSK-3β and could inhibit the cellular GSK-3β activity in primary neurons to reduce GSK-3β substrate tau phosphorylation at Ser396. Among them, inhibitor 18i could reduce infarct size and improve the neurological deficit in the cerebral ischemia animal model, suggesting its potential neuroprotective activity against brain ischemic stroke.

Table 4 GSK-3β inhibitory activity of 3-(furo[2,3-b]pyridin-3-yl)-4-(1H-indol-3-yl)maleimides 18
Full size table

To find more potent and selective GSK-3β inhibitors, Pang et al. designed a series of novel 4,5-bisindolyl-1,2,4-triazolones based on the analysis of the interaction mode of bishetero-maleimide with the hinge residues (Asp133 and Val135) of GSK-3β (Fig. 8) [48]. This series of inhibitors could significantly reduce GSK-3β substrate Tau phosphorylation at Ser396 in primary neurons by inhibiting cellular GSK-3β. In this series of thirteen synthetic compounds in total, compounds 19c, 19e, and 19g showed the most potent GSK-3β inhibitory activity. By the comparison of the activity of 19a–19h, it was concluded that the halogen substitution on 5-position is the most optimal for increasing inhibitory activity towards GSK-3β. In addition, compounds 19c, 19e, and 19g showed promising neuroprotective activity by preventing neuronal cells from glutamate-induced death.

Fig. 8
figure 8

Design strategy and structures of 4,5-bisindolyl-1,2,4-triazol-3-one GSK-3β inhibitors 19

Full size image

GSK-3β inhibitors with pyrimidin-4-one core

A set of pyrimidin-4-one-1,2,3-triazole conjugates as GSK-3β inhibitors have been synthesized by Hamid et al. using the click-chemistry approach [49]. The target compounds 20 (Table 5) were constructed from the linkage of pyrimidine-4-one core with triazole ring through a sulfur bond. Among these novel conjugates, four of them, 20a–20d, showed improved inhibitory activity towards GSK-3β. The result of SAR analysis clearly indicates that when the substituent attached on triazole ring is methoxyphenyl, ethoxyphenyl, or pyridine, the compound’s inhibitory activity will significantly decrease.

Table 5 Structures of pyrimidin-4-one-1,2,3-triazole GSK-3β inhibitors
Full size table

Later, the same research group reported another study on the GSK-3β inhibitors derived from the connection of substituted piperazine ring to pyrimidin-4-one backbone (Fig. 9) [50]. From the result of GSK-3β inhibition bioassay, the activity was increased in accordance with the following changes made in the SAR analysis shown in Fig. 9. Among these pyrimidin-4-one analogs, compounds 21a–21d showed the most potent inhibition on GSK-3β with IC50 values ranging from 83 to 127 nM.

Fig. 9
figure 9

Structures of GSK-3β inhibitors 21 and SAR analysis

Full size image

Watanabe et al. prepared and evaluated a similar series of promising pyrimidin-4-one derivatives containing piperazine moiety as GSK-3β inhibitors (Fig. 10) [51]. The SAR analysis shown in Fig. 10 indicated that (1) the analogs with (R)-methyl group had more potent activity and (2) the analogs with pyrimidinyl ring (22a) seemed to have more preferable GSK-3β inhibitory activity than those with pyridyl ring (22b). The further replacement of methyl by isopropyl, phenyl, benzyl, acyl, alkoxycarbonyl, and sulfonyl on the nitrogen atom of the piperazine resulted into a more potent compound 22c (IC50 = 1.4 nM) with moderate metabolic activity and CYP2D6 inhibition. Then, compound 22c was chosen as a hit compound for further optimization, which led to the generation of more potent GSK-3β inhibitors. From the bioassay results, it was shown that the substituents (such as –F, –Cl, –MeO) on the phenyl group generally increased the activity, while the nitrile group made the activity decrease, however, the position effect of substituents on the activity was unclear. Among them, analog 22j showed the most potent activity towards GSK-3β (IC50 = 0.1 nM) with good in vitro and in vivo pharmacokinetic profiles, and could significantly decrease Tau phosphorylation in mice model. Later, the same Watanabe’s group reported another set of GSK-3β inhibitors possessing phenylpiperazinyl-pyrimidin-4-one framework [52], of which 23a and 23b with 4-fluoro-2-methoxy group displayed the best inhibitory activity towards GSK-3β.

Fig. 10
figure 10

Structures of GSK-3β inhibitors 22 and 23

Full size image

Benzoxazinone- and indole-based AK/GSK-3β dual inhibitors

Since adenosine kinase (AK) and GSK-3β are both involved in neurodegenerative disorders, in 2017 Brindisi et al. developed the first class of AK/GSK-3β dual inhibitors with benzoxazinone (24) or indole (25) scaffolds. Their preliminary SAR analysis is displayed in Fig. 11 [53].

Fig. 11
figure 11

SAR analysis of benzoxazinone- and indole-based AK/GSK-3β dual inhibitors 24 and 25

Full size image

2,4-Dihydropyrano[2,3-c]pyrazole-based GSK-3β inhibitor/Nrf2 inducer

The dual-target treatment strategy for AD has attracted more and more attention due to the complex pathogenesis of AD. In 2017, León et al. [54] reported the design of the first dual GSK-3β inhibitor/Nrf2 inducer, 2,4-dihydropyrano[2,3-c]pyrazoles 28 (Fig. 12), through analyzing the chemical features of the known GSK-3β inhibitors 26 and 27. Among the twenty-one synthetic analogs, compounds 28a–28l showed an inhibitory effect on GSK-3β. The SAR study shows that methoxyphenyl- and nitrophenyl-derivatives have better inhibitory activity (IC50 < 10 μM) than those having phenyl or methylphenyl substituents. In addition, the substituting position of halogen groups on phenyl also has an effect on the activity of compounds, which can be clearly demonstrated by the comparison of the activity of compounds 28c–28k. Furthermore, some of these compounds were found to induce the Nrf2 phase II antioxidant and anti-inflammatory pathway at micromolar concentrations. Thus, this series of 2,4-dihydropyrano[2,3-c]pyrazole compounds were considered as multi-targeted leads for the development of anti-AD drugs.

Fig. 12
figure 12

Structures of 2,4-dihydropyrano[2,3-c]pyrazoles 28 as the first dual GSK-3β inhibitors/Nrf2 inducers

Full size image

2,4-Dihydropyrano[2,3-c]pyrazole-based GSK-3β inhibitor/Nrf2 inducer

Due to that GSK-3β and casein kinase 2 (CK2) are both responsible for the phosphorylation of the tumor suppressor protein (PTEN), Vasu et al. designed and prepared a set of novel dual kinase CK2/GSK-3β inhibitors with 4,5,6,7-tetrahydrobenzo[d]thiazole core (Fig. 13) to maximize the effect the kinase inhibitors [55]. The bioassay results showed that compound 29a had the highest dual kinase inhibitory activity, with IC50 values of 1.9 μM against CK2 and 0.67 μM against GSK-3β. Other analogs 29b–29f also showed good dual CK2/GSK-3β inhibition. The SAR analysis indicated that the presence of a carboxyl group at the meta-position of the phenyl ring played a vital role in dual kinase inhibition.

Fig. 13
figure 13

Dual kinase GSK-3β/CK2 inhibitors 29

Full size image

Benzo[e]isoindole-1,3-dione-based selective GSK-3β inhibitors

Quan et al. reported a series of benzo[e]isoindole-1,3-dione derivatives (Fig. 14) as selective GSK-3β inhibitors through rational drug design [56]. By analyzing the detailed structure inhibitors of GSK-3β and cyclin-dependent kinase 2 (CDK2), they used compound 30 as a starting point to derive a series of inhibitors 31 by introducing a sulfonyl group, which was considered to interact with the positively charged residues Arg141 and Arg144 at the helix D of GSK-3β. Thereinto, five of them (31a–31e) showed high selectivity against GSK-3β over CDK2, and could significantly activate the cellular Wnt/β-catenin pathway.

Fig. 14
figure 14

Benzo[e]isoindole-1,3-dione selective GSK-3β inhibitors 30 and 31

Full size image

Benzothiazinone-based GSK-3β inhibitors

Most of the known GSK-3β inhibitors are reported to be bound to the ATP-binding pocket of GSK-3β. To avoid the adverse effects induced by ATP-competitive GSK-3β inhibitors, Ye and coworkers synthesized a bunch of novel non-ATP competitive GSK-3β inhibitors with benzothiazinones core (Table 6) based on their previous research [57]. These compounds displayed potent inhibition on GSK-3β with low micromolar activities, of which analogs 32a–32d and 32h–32k showed good potency toward GSK-3β. The poor performance of the remaining compounds may be due to the extension of the alkyl carbon chain. The antitumor bioassay indicated that most of these compounds had moderate in vitro anti-proliferative against A2780 and OVCA433 cancer cell lines.

Table 6 In vitro inhibitory activities of 32 against GSK-3β kinase, and A2780, and OVCA433 cancer cell lines
Full size table

7-Chloro-9H-pyrimido[4,5-b]indole-based GSK-3β inhibitors

Recently, Koch et al. reported the optimization of tofacitinib-derived 3-[(3R,4R)-3-[(7-chloro-9H-pyrimido[4,5-b]indol-4-yl)(methyl)amino]-4-methylpiperidin-1-yl]-3-oxopropanenitrile (33) resulting into a novel class of GSK-3β inhibitors (34) by the rigidization of cyanoacetyl piperidine moiety (Fig. 15) [58]. The SAR study showed that when these compounds had a methyl group as a substituent (R2), their inhibitory activities were decreased, while the formation of the pyrrolidine ring will lead the activity significantly increase. For example, compound 34a showed the strongest inhibitory effect on GSK-3β with an IC50 value of 0.130 ± 0.008 μM, about 19-fold over the hit 33. Molecular docking analysis indicated the importance of nitrile side chain for the activity of this series of inhibitors.

Fig. 15
figure 15

Development of new type of GSK-3β inhibitors 34 derived from hit 33

Full size image

3-Substituted 2-oxoindole GSK-3β inhibitors

A series of 3-substituted 2-oxoindole derivatives 35a–35× and 36a–36g (Fig. 16) as novel GSK-3β were disclosed by Lozinskaya and coworkers [59]. Abundant substituent groups intuitively show the increase or decrease effect between substituents and inhibitory activity, of them, compound 32a (R = H; aryl = pyridin-2-yl) was the most potent GSK-3β inhibitor with an IC50 value of 4.19 nM. This compound exhibited no significant cytotoxicity towards leucocyte at 10 μM, and it further showed a high in vivo anti-diabetic efficacy in streptozotocin-treated obese rat model.

Fig. 16
figure 16

Series of 3-substituted 2-oxoindole-derived GSK-3β inhibitors 35 and 36

Full size image

2,3-Diaminopyridine moiety-containing GSK-3β inhibitors

To find multifunctional anti-AD molecules, a series of 2,3-diaminopyridine moiety-containing novel GSK-3β inhibitors (Fig. 17) with metal chelation activity were designed and synthesized by Liu et al. [60]. The bioassay results revealed that the amide derivatives 37a–37f showed moderate potency against GSK-3β and weak chelating ability towards Cu2+, Zn2+, and Al3+, while the imine derivatives 38a–38c were potent dual GSK-3β inhibitors (IC50 = 38–72 nM)/selective Cu2+ and Al3+ chelates. In addition, imines 38a–38c could inhibit the phosphorylation of tau protein and have neuroprotective activity by preventing nerve SH-SY5Y and PC12 cells from Cu2+-Aβ1-42- and H2O2-induced cell injury.

Fig. 17
figure 17

2,3-Diaminopyridine moiety-containing GSK-3β inhibitors/metal chelates 37 and 38

Full size image

Conclusion

GSK-3β was involved in the occurrence and development of various human diseases, such as cancer, AD, inflammation, type-II diabetes, mood disorders, etc. Thus, the development of GSK-3β inhibitors attracted more and more interest from the academic and industrial areas. The present review provided the recent development of GSK-3β inhibitors, including their design strategy, SAR, and in vitro or in vivo biological functional study targeting different diseases, especially cancer and AD. The information summarized herein is helpful for the design of new GSK-3β inhibitors.

As yet, many GSK-3β inhibitors have been discovered, however, only one GSK-3β inhibitor, tideglusib, has entered the phase II clinical trials, meaning that these available promising GSK-3β inhibitors still have great room for structural/biological improvement and more new types of inhibitors are waiting for exploration. Currently, the reversible non-ATP competitive GSK-3β inhibitors, such as allosteric modulators and substrate competitive inhibitors, were more attractive to medicinal chemists due to the advantages, such as higher selectivity and better in vivo efficacy, less adverse effects, and potential therapeutical effects. Besides, the dual/or multi-target GSK-3β inhibitors were also pursued by medicinal chemists with the understanding of both GSK-3β and other targets in multifactorial diseases, especially for AD and cancer associated with complicated pathogenesis. With the ongoing effect and interests towards GSK-3β inhibitors development, more GSK-3β inhibitors with great value in clinical application, either derived from natural product optimization, rational drug design, or drug virtual screening, will be disclosed.