Thiadiazole derivatives as anticancer agents

In spite of substantial progress made toward understanding cancer pathogenesis, this disease remains one of the leading causes of mortality. Thus, there is an urgent need to develop novel, more effective anticancer therapeutics. Thiadiazole ring is a versatile scaffold widely studied in medicinal chemistry. Mesoionic character of this ring allows thiadiazole-containing compounds to cross cellular membrane and interact strongly with biological targets. Consequently, these compounds exert a broad spectrum of biological activities. This review presents the current state of knowledge on thiadiazole derivatives that demonstrate in vitro and/or in vivo efficacy across the cancer models with an emphasis on targets of action. The influence of the substituent on the compounds’ activity is depicted. Furthermore, the results from clinical trials assessing thiadiazole-containing drugs in cancer patients are summarized.


Introduction
According to the most recent data provided by the International Agency for Research on Cancer (IARC), 18.1 million new cases and 9.6 million cancer deaths were registered worldwide in 2018 [1]. Due to the population aging and growth, the number of new cancer cases is expected to increase. Although a substantial progress was made in the understanding of molecular biology of particular cancer types, and numerous potential specific therapeutic targets were identified in recent years, there is an urgent necessity for the development of improved anticancer therapeutic strategies.
Thiadiazole is a five-membered heterocyclic compound containing one sulfur and two nitrogen atoms. It occurs in nature in four isoforms: 1,2,3-thiadiazole, 1,2,4-thiadizaole, 1,2,5-thiadiazole and 1,3,4-thiadiazole (Fig. 1). Taking into account that thiadiazole is the bioisostere of pyrimidine and oxadiazole, it is not surprising that compounds bearing this moiety present a broad spectrum of pharmacological properties, including antiviral, antibacterial, antifungal, antiparasitic, anti-inflammatory and anticancer activities [2]. Due to the mesoionic nature, thiadiazoles are able to cross the cellular membranes. Their relatively good liposolubility is most likely attributed to the presence of the sulphur atom [3]. The thiadiazole-containing drugs, including diuretics acetazolamide and methazolamide or antibiotics cefazedone and cefazolin sodium, are already used in clinics. Accumulating evidence has also revealed numerous thiadiazole derivatives that display anticancer activities in various in vitro and in vivo models (summarized in Table 1). Moreover, several thiadiazole-containing compounds have moved into clinical trials either as single agents or in combination with existing anticancer drugs (summarized in Table 2).

Inhibitors of tubulin polymerization
Microtubules are cytoskeleton filamentous proteins built by tubulin. They are involved in numerous cellular processes such as intracellular transport, cell signaling, mitosis, cellular integrity and gene expression, but also contribute to polarity and shape of cells. A growing body of evidence documents anticancer activity of different heterocyclic compounds inhibiting tubulin polymerization [4].

3
Wu and co-workers focused on analogs of the cis stilbene derivative combretastatin A-4 (CA-4), an anticancer agent which binds to tubulin and inhibits microtubule polymerization. Cis configuration of the double bond in olefin group and 3,4,5-trimethoxyphenyl group are crucial for the CA-4's activity. The newly designed and synthetized analogs contained 1,2,3-thiadiazole instead of the CA-4's olefin group. They exhibited a diverse cytotoxicity against human myeloid leukemia HL-60 cell line, human colon adenocarcinoma HCT-116 cell line, and immortalized human microvascular endothelial (HMEC-1) cells. In all three cell lines, several tested compounds presented cytotoxic activity similar to that of CA-4 or lower, but still considerable (IC 50 ranging from 13.4 to 86.6 nM). Of note, if the 3,4,5-trimethoxyphenyl was at 4th position in 1,2,3-thiadiazole, six out of nine tested compounds displayed significant activity in all three cell lines, while if this substituent was at 5th position, only one out of nine compounds was cytotoxic. These compounds inhibited tubulin polymerization with activities quantitatively similar to those of CA-4 and arrested the cell cycle at G2/M phase. Two of these derivatives significantly reduced tumor growth in mice S180 sarcoma model with the inhibition rate comparable or even higher to that of CA-4 [5].

Inhibitors of Hsp90
The other group of 1,2,3-thiadiazole derivatives appeared to block the activity of heat shock protein 90 (Hsp90). Hsp90 displays a chaperone activity and controls the folding of numerous proteins. Inhibition of its activity results in the degradation of several oncoproteins. A growing body of evidence shows that tumor cells are more susceptible to blocking of Hsp90 compared to normal cells, therefore, this protein seems to be a promising anticancer target [6]. Cikotine et al. synthesized a series of 5-aryl-4-(5-substituted-2-4-dihydroxyphenyl)-1,2,3-thiadiazoles bearing at the position 5 of the thiadiazole ring one of the following groups: 4-MeOC 6 H 4 , 4-EtOC 6 H 4 , 4-MeC 6 H 4 or 3,4-di-MeOC 6 H 3 and either chloro-or ethyl-substituent at the position 5 of the dihydroxyphenyl moiety. Each of these derivatives tightly bound to Hsp90 (Fig. 2), and the strongest binding (dissociation constant (Kd) of 4.8 nM) displayed compound 3b, bearing a 4-EtOC 6 H 4 substituent at position 5 of thiadiazole and a chlorine atom at position 5 of dihydroxyphenyl. Each of the derivatives significantly inhibited viability of both human cervical carcinoma HeLa and osteosarcoma U2OS cells and the most potent inhibitor appeared to be compound 3e bearing 4-EtOC 6 H 4 group at the position 5 of the thiadiazole ring and an ethyl substituent at the position 5 of dihydroxyphenyl. The GI 50 values of this compound were 0.70 μM for HeLa and 0.69 μM for U2OS cells, respectively. Of note, the other group of 5-aryl-4-(5substituted-2-4-dihydroxyphenyl)-1,2,3-thiadiazoles bearing chloro-substituent at position 3 of the dihydroxyphenyl moiety did not bind to Hsp90 and was a very weak inhibitor of cancer cell viability. Most likely, the presence of this chloro-substituent prevented the formation of the extensive H-bonding network which in turn led to a lack of activity [7].
Three of the compounds, confirmed to bind Hsp90 most effectively, were subjected to further analysis. They exhibited antiproliferative activity against human colon cancer HCT-116 cells with GI 50 values ranging from 3.2 to 4.6 μM. Treatment of HCT-116 cells with each of the compounds resulted in a depletion of Hsp90 client proteins, CRAF, ERBB2 and CDK4, confirming that antiproliferative activity was linked to the inhibition of Hsp90 activity. Furthermore, such treatment caused upregulation of Hsp27 and Hsp72 expression, suggesting an induction of the heat shock response. Moreover, an increase in PARP cleavage evoked by the tested compounds indicated the induction of apoptosis [8].

Miscellaneous 1,2,3-thiadiazole derivatives
Aside from the derivatives presented above, there are also some other compounds containing 1,2,3-thiadiazole moiety which display an anticancer activity, but their molecular targets remain unknown. Among a series of d-ring fused 1,2,3-thiadiazole dehydroepiandrosterone (DHEA) derivatives, the most potent compounds 22, 23 and 25 presented antitumor activity against human breast cancer T47D cells with IC 50 values ranging between 0.042 and 0.058 μM. These values were comparable to that of reference drug adriamycin (IC 50 = 0.04 μM). Of note, compound 25 possessed a considerable selectivity towards T47D cells. Its activity against the other breast cancer cell lines, MDA-MB-231 and MCF-7, as well as human

Derivatives of 1,2,4-thiadiazole
Resveratrol is a naturally occurring trans stilbene with anticancer and cancer chemopreventive potential. However, due to rapid and extensive metabolism, its bioavailability is poor. A growing body of evidence indicates that various chemical modifications may significantly improve resveratrol's bioavailability and potency against different types of cancer [12]. Mayhoub and co-workers replaced the trans stilbene ethylenic bridge of the resveratrol scaffold with a 1,2,4-thiadiazole heterocycle and next modified the substituents on the two aromatic rings and evaluated cytotoxicity of these compounds against human breast cancer MCF-7 cell line. These compounds displayed the IC 50 values ranging from 4.7 to 39.8 μM and the most potent turned out to be 3-hydroxyderivative termed 3jj. Although modifications introduced to the compounds' structure altered their potency and specificity against three enzymes inhibited by resveratrol, i.e. aromatase, NF-κB and quinone reductase 1, it did not translate to the anticancer activity towards MCF-7 cells [13].
Derivatives based on 3-substituted benzo [4,5] imidazo [1,2-d] [1,2,4]thiadiazole linked with a polymethylene spacer to α-bromoacryloyl amido benzoheterocycles are another example of compounds carrying 1,2,4-thiadiazole moiety with an anticancer activity. These compounds showed a considerable activity in human myeloid leukemia cell lines HL-60 and U937 (IC 50 values from 0.24 to 1.72 μM), as well as in melanoma SK-MEL-1 cell line (IC 50 from 2.09 to 8.95 μM). The most active were compounds 1 and 2 containing N-unsubstituted indole in the benzoheterocycle. Comparing these two derivatives, the anticancer activity decreased on increasing the length of the chain from three (compound 1) to six (compound 2) methylene units. Among the derivatives characterized by the same alkyl chain (6 methylene units), the greatest potency was exhibited by compound 2 bearing α-bromoacrylamidoindole moiety (IC 50 values from 0.40 to 3.02 μM). A replacement of indole (compound 2) with N-methyl indole (compound 3) decreased an anticancer activity. Further decrease was observed when indole was replaced with benzofuran (compound 4) and benzothiophene (compound 5). Treatment with each of compounds induced apoptosis in HL-60 and U937 cells, although most effective anticancer compound 1 appeared to be a less potent apoptosis inducer compared to the rest of derivatives, suggesting that this compound may trigger also some other mechanisms underlying cell death [14].
Further analysis revealed that NSC745885 was also effective in oral cancer cell lines, but anticancer concentrations of this compound, up to 4 μM, did not significantly affect the viability of normal lung fibroblasts. Treatment with NSC745885 reduced tumor growth iv vivo, but did not decrease the bodyweight of the animals. This compound displayed anti-tumor efficiency similar to that of doxorubicin, an anthraquinone derivative, but higher safety. Induction of apoptosis was observed upon treatment with NSC745885 both in vitro and in vivo (Fig. 3) [16].
In another study from the same group, 2.5 μM NSC745885 reduced the viability of bladder cancer T24 and MBT2 cell lines and induced G2/M cell cycle arrest, but did not harm immortalized normal urothelial cells SV-HUC-1 or normal fibroblasts 3T3. Of note, the same inhibitory effect on viability was observed when the cancer cells were treated either with 2.5 μM doxorubicin or 40 μM emodin, a natural anthraquinone, but both of these compounds presented significant toxicity towards SV-HUC-1 and 3T3 cells.
Moreover, treatment with 2.5 μM NSC745885 suppressed the viability of multi-drug resistant bladder cancer MGH-U1R cells. This compound also displayed a remarkable antitumor activity in vivo. NSC745885 treatment downregulated expression of the enhancer of zeste homolog 2 (EZH2) in cancer cells, but not SV-HUC-1 cells. EZH2 is a member of the polycomb repressive complexes 2 (PRC2) that catalyzes the methylation of histone H3 lysine 27 (H3K27), which in turn mediates chromatin compaction. Overexpression of EZH2 is observed in numerous cancer of different origin and its inactivation was therapeutically effective in several cancer models [17]. Of note, the other components of PRC2 and the global H3K27 methylation, modified by EZH2, remained unaffected by NSC745885 treatment. The up-regulation of EZH2-silenced tumor suppressor genes, CDKN1C, DAB2IP, and WNT5a, was found in NSC745885-treated cancer cells, but not SV-HUC-1 cells (Fig. 3). More detailed molecular analysis revealed that suppression of the cancer cells' viability by NSC745885 treatment was indeed causatively linked to downregulation of EZH2 [18].
The same group further investigated anticancer activity of NSC745885 and its 4-chloro derivative, NSC757963. While melanoma and ovarian cancer cell lines were particularly sensitive to NSC745885 (GI 50 values ranging from 0.55 to 7.71 μM), leukemia cell lines were notably susceptible to NSC757963 (GI 50 values from 0.33 to 3.15 μM). Colon and non-small cell lung cancer cell lines presented the lowest sensitivity to NSC745885 (GI 50 values from 1.28 to 19.00 μM) and NSC757963 (GI 50 values from 1.43 to > 100 μM). Either of compounds inhibited translocation of NF-κB to the nucleus, which in turn suppressed constitutive activation of this transcription factor. Docking studies revealed favorable binding of both compounds to the ATP site of the N-terminal kinase domain of the IKKβ subunit, a very potent activator of NF-κB [19].
In a subsequent analysis, the same group examined an anticancer activity of a series of sulfur-substituted anthra [1,2-  10 μM concentration of this compound was relatively not toxic towards human non-cancerous cell lines, as it reduced the viability of urothelial SV-HUC-1 cells to 80%, prostatic stromal myofibroblasts WMPY-1 to 60%, and prostate epithelial cells RWPE-1 to 90%. At the same time, 10 μM doxorubicin decreased the viability of SV-HUC-1, WMOY-1 and RWPE-1 cells to 50%, 20%, and 40%, respectively. In prostate cancer DU-145 cells, NSC763968 induced apoptosis and inhibited phosphorylation of ERK and p38 kinases [20]. As both ERK and p38 pathways are involved in the pathogenesis of cancers of different origin [21,22], it is tempting to speculate that NSC763968 treatment inhibits those pathways, which in turn contributes to the anticancer activity of this compound.
In very recent research from the same group, a nitrogensubstituted anthra[1,2-c] [1,2,5]thiadiazole-6,11-dione derivative, RV-59, turned out to be particularly toxic to human colon cancer HCT116 cells with high cytoplasmic level of a transcription factor Nrf2 (cNrf2). A significant contribution of a cytoplasmic, but not nuclear, localization of Nrf2 to aggressive phenotype of colorectal cancer cells has previously been documented by the same group [23]. The IC 50 value of RV-59 was 3.55 μM for HCT116 cells with high cNrf2, and 16.81 μM for Nrf2-knockdown HCT116 cells. The IC 50 value of 5-FU, a drug used to treat colon cancer [24], was 17.74 μM for the cells with high cNrf2, and 5.35 μM for the cells not expressing Nrf2. These results indicate that RV-59 predominantly kills and overcomes cNrf2mediated resistance to 5-FU. Further analysis clearly showed that RV-59 remarkably suppressed xenograft tumor growth induced by the cells with cNrf2-mediated 5-FU resistance. Of note, RV-59 treatment did not affect the body weights of the animals, suggesting that this drug may not be toxic to normal cells [25].

Inhibitors of inosine monophosphate dehydrogenase (IMPDH)
One of the first studies documenting anticancer activity of 1,3,4-thiadiazole derivatives was published by Shapiro et al. in 1957 [26]. In this paper, 2-ethylamino-1,3,4thiadiazole (EATDA), an analog of niacin, inhibited the growth of mammary adenocarcinomas induced in mice. The tumor-inhibitory effect of this compound was prevented by the prior injection of nicotinamide, supporting the evidence that it is a niacin antagonist. Moreover, the addition of EATDA to the combination of 8-azaguanine, deoxypyridoxine and testosterone improved anticancer activity of this three-drug combination [26].
EATDA and 2-amino-1,3,4-thiadiazole (ATDA) displayed anti-cancer properties in mice surviving systemic leukemia L1210. These compounds not only inhibited tumor growth but also prolonged survival time of the animals. The anti-leukemic activity and host toxicity of either compound were blocked by the administration of nicotinamide [27]. Later study revealed that treatment of L1210-bearing mice with ATDA, further referred to as NSC 4728 (Fig. 4), diminished adenine and guanine ribonucleotide pools and increased uridine triphosphate (UTP) and inosine monophosphate (IMP) pools in the tumor cells. This effect was prevented by simultaneous administration of nicotinamide. Taking into account that nicotinamide prevents anti-leukemic activity of ATDA, it might suggest that the inhibition of guanosine monophosphate synthesis was related to the anti-leukemic action of this compound [28]. In further mechanistic studies, NSC 4728 and its derivatives appeared to be potent inhibitors of IMP dehydrogenase (EC 1.2.1.14), an enzyme involved in the conversion of IMP to xanthosine monophosphate (XMP), a substrate for the production of guanosine monophosphate (GMP) (Fig. 2) [29].
In phase I clinical trials, NSC 4728 was administered daily, twice a week or weekly in a range of doses from 2 to 200 mg/m 2 to 42 patients suffering from different cancers. One patient with squamous cell carcinoma of the lung and one with squamous cell carcinoma of the cervical esophagus exhibited a partial response to a total dose of 550 or 575 mg/ m 2 , respectively. In either patient, a progression of disease was observed after 2 months despite continued treatment. In phase II, 125 mg/m 2 of NSC 4728 was given once a week to 29 patients with non-small cell lung carcinoma. A partial response was observed in one patient after receiving a total of 1200 mg/m 2 and another one after two doses of 125 mg/ m 2 of NSC 4728. In either patient, the progression was observed despite continued treatment. Stomatitis was the most common adverse effect, but dermatitis, nausea vomiting, lethargy, and hyperuricemia were observed as well [30].
Out of 46 patients with metastatic renal cell carcinoma one patient, treated with NSC 4728 in a daily dose of 125 mg/m 2 , experienced partial remission. Leukopenia, thrombocytopenia and anemia were the most common side effects in patients enrolled in this phase II clinical trials [31]. Asbury and co-workers conducted several phase II clinical trials of NSC 4728 in a dose of 125 mg/m 2 weekly in patients with different tumor types. Partial response to treatment was observed in 12% of patients with advanced colon cancer. Gastrointestinal toxicity was severe in l6% of patients [32]. A partial response was also observed in 5% of patients with squamous carcinoma of the cervix and 28% of patients had stable disease. The patients exhibited mild renal toxicity and a single life-threatening toxic episode was observed [33]. Complete response was observed in 8% of patients with non-squamous cervical carcinoma, and 35% had stable disease. Anemia and emesis were the main side effects [34]. Partial response was exhibited by 5% of patients with mixed mesodermal tumors of the uterine corpus. Severe nausea and anemia were often [35]. No response was observed in patients with either advanced or recurrent leiomyosarcoma [36] or advanced epidermoid carcinoma of the esophagus [37] treated in the same way. Clinical trials with NSC 4728 are summarized in Table 1.

Inhibitors of carbonic anhydrase (CA)
Some of the 1,3,4-thiadiazole derivatives turned out to inhibit the activity of human carbonic anhydrase (CA; EC 4. 2. 1. 1). This zinc metalloprotein catalyzes CO 2 / HCO 3 − interconversion and is thereby involved in several physiological and pathological processes. So far, 14 humans CA isoforms have been identified and each of them serves as biomarker for various diseases, including cancers of several origin [38].
A CA inhibitor acetazolamide (5-acetamido-1,3,4-thiadiazole-2-sulfonamide, AZA) (Fig. 2, Fig. 5) displaying diuretic effects is already used in clinics. There are two currently ongoing phase I clinical trials with AZA (Diamox) in cancer patients. In one of them AZA in combination with platinum and etoposide-based radiochemotherapy is evaluated in patients with small-cell lung cancer (NCT03467360). In the other one, AZA in combination with temozolomide is assessed in patients with malignant glioma (NCT03011671) ( Table 1).
Supuran and Scozzafava documented inhibitory effect of the other 1,3,4-thiadiazole-2-sulfonamide derivatives on CA isoforms [39]. Nanomolar concentrations of compounds 10-13 remarkably inhibited the activity of CA II and CA IV but were less potent towards CA I. Nanomolar concentrations of the urea/thiobiguanide derivatives 14-16 inhibited CA I and CA II and to the lesser extend CA IV. Most of these compounds were more potent CA inhibitors than AZA. Compounds 14-16 were much more cytotoxic (GI 50 values from 12 to 70 μM) than compounds 10-13 which in most cancer cell lines displayed GI 50 values > 100 μM. The exception was the OVCAR-4 ovarian cancer cell line, susceptible to compounds 10 and 11 (GI 50 values of 0.5 and 0.1 μM, respectively). The exact mechanism underlying anticancer activity of these derivatives was not elucidated, but the authors postulated that the acidification of the intracellular environment resulted from CA inhibition might be of crucial importance [39].
In the later study from the same group, compound 14, a biphenyl-disulfonamide derivative bearing 5-amino-1,3,4thiadiazole-2-sulfonamide, inhibited CA II and CA IX, and to the lesser extend also CA XII and CA I. This compound displayed cytotoxicity towards human colon cancer HCT116 cell line with GI 50 value of 3.789 μg/mL but was less potent against non-small cell lung cancer H460 and breast cancer MCF7 cells [40].

Inhibitors of glutaminase (GA)
Glutamine (Gln) plays the versatile and crucial role in tumors regardless of the driving oncogene or tissue of origin. Increased metabolism of Gln, of which the first step is catabolized by glutaminase (GA; EC 3.5.1.2), is a hallmark of cancer. Therefore, GA is a potential therapeutic target for different cancer types [41].
Treatment with 2 μM 10 μM of bis-2-(5-phenylacetamido-1,3,4-thiadiazol-2-yl)ethyl sulfide (BPTES), a specific inhibitor of kidney type GA (GLS), diminished the proliferation of human Burkitt lymphoma P493 cells [42]. Higher concentrations (10-40 μM) of this compound decreased the viability of glioblastoma [43], non-small cell lung cancer [44] and acute myeloid leukemia [45] cells. Moreover, BPTES injected daily in a dose of 12.5 mg/kg body weight inhibited lymphoma and hepatocellular carcinoma xenograft growth [42,46]. Poor drug-like molecular properties of BPTES, mainly extremely poor aqueous solubility (< 1 μg/ mL) ruled out the feasibility of further development of this compound as a therapeutic agent. Of a number of BPTESderived GLS inhibitors displaying much better drug-like properties compared to BPTES recently synthesized [47], only those already tested in cancer models will be presented below.
A  Schematic chemical structure of acetazolamide (5-acetamido-1,3,4-thiadiazole-2-sulfonamide, AZA). Sulphur, nitrogen, hydrogen, and oxygen atoms are marked as yellow, blue, green or red circles, respectively compound 6, exhibited potency similar to that of BPTES, but much better aqueous solubility of 13 μg/mL. Treatment with 20 μM concentration of this compound significantly attenuated the growth of lymphoma B P493 cells in vitro as well as in a mouse xenograft model while injected daily in doses of 12.5 mg/kg [48].
In EGFR mutant non-small cell lung cancer HCC827 cells, treatment with 300 nM CB-839 resulted in approximately 50% reduction of the viability. Administration of this compound twice daily in a dose of 200 mg/kg inhibited xenograft growth. Furthermore, the combination of an EGFR inhibitor, erlotinib, and CB-839 cooperated to inhibit the growth of HCC827 cells in vitro and in vivo [50]. For the other lung cancer cell lines, A427, A549 and H460, the CB-839 ED 50 values for inhibition of colony formation were 9.1, 27.0 and 217 nM, respectively. Treatment with 1 μM CB-839 increased response of H460 cells to radiation. The short-term CB-839 dosing at 200 mg/kg did not affect xenograft tumor growth, but the combination of CB-839 administration with the radiation dose of 12 Gy reduced tumor growth by 15-30% [51].
Treatment with 500 nM CB-839 inhibited the viability of NF1 mutant/null malignant peripheral nerve sheath tumor cell lines (MPNST), ST8814 and S462. Administration of 200 mg/kg of CB-839 significantly suppressed the volume of MPNST xenografts [52]. In chondrosarcoma cell lines, CB-839 activity correlated with the status of isocitrate dehydrogenase 1/2 (IDH1/2), as the cells carrying IDH1/2 mutation were more susceptible to this compound than the wild type cells [53]. In a very recent study, CB-839 administered twice daily at dose 200 mg/kg significantly reduced undifferentiated pleomorphic sarcoma (UPS) tumor growth and weight but did not alter animal weights [54].
Taken together, data presented above clearly indicate that CB-839 exhibits high therapeutic potential. Indeed, this drug was evaluated either as a single agent in patients with leukemia (NCT02071927), hematologic malignancies (NCT02071888), and solid tumors (NCT02071862), as well as in combination with paclitaxel in patients with triplenegative breast cancer (NCT03057600) or in combination with panitumumab and irinotecan in patients with colorectal cancer (NCT03263429). However, detailed results of any of those trials are not yet available. Numerous clinical trials evaluating anticancer activity of CB-839 are still ongoing (Table 1).
Recently, 2-phenyl-N-(5-(4-((5-(2-phenylacetamido)-1,3,4-thiadiazol-2-yl)amino)piperidin-1-yl)-1,3,4-thiadiazol-2-yl)acetamide (UPGL00004), another analog of BPTES, turned out to be a promising anti-cancer agent. In this compound, previously termed 7c, the flexible region of BPTES or CB-839 has been replaced by relatively rigid heterocyclic core [55]. UPGL00004 inhibited the enzymatic activity of GLS more potently than BPTES and displayed binding affinity for this protein similar to that of CB-839 (Fig. 2). Crystallographic studies revealed that UPGL00004 occupied the same binding site as CB-839 or BPTES and that all three compounds inhibited the enzymatic activity of GLS via a similar allosteric mechanism. The anticancer activity of UPGL00004 was examined in a triple-negative breast cancer patient-derived tumor graft model. Neither UPGL00004 in a dose 1 mg/kg body weight nor approved for the treatment of metastatic breast cancer bevacizumab in a dose 2.5 mg/ kg body weight reduced tumor growth. However, a combination of these compounds completely prevented an increase in tumor size during the course of treatment [56].

Inhibitors of histone deacetylase (HDAC)
Histone deacetylase (HDAC, EC 3.5.1.98) removes acetyl groups from DNA-binding histone proteins, which in turn . Sulphur, nitrogen, hydrogen, oxygen or fluorine atoms are marked as yellow, blue, green, red or purple circles, respectively decreases chromatin accessibility for transcription factors and blocks the transcription. The human HDAC family consists of 18 proteins divided into 4 classes. These proteins modulate, among others, the transcription of genes encoding proteins involved in carcinogenesis [57].
Guan and co-workers designed and synthesized a series of amino-1,3,4-thiadiazole-based hydroxamic acid derivatives with different linkers and substitution in thiadiazole ring. Both the length of the linker chain and the substitution in 1,3,4-thiadiazole turned out to be important for the HDAC inhibitory activity. Compounds with the linker comprised of five or six methylene units inhibited HDAC in the nanomolar range, while the rest derivatives showed only micromolar activity. Compounds with the phenyl or benzyl substitution in 1,3,4-thiadiazole were more potent than the those substituted with phenethyl or (E)-styryl. The linker between zinc-binding group and 1,3,4-thiadiazole ring was more important for the anti-HDAC activity than the substitutions in 1,3,4-thiadiazole moiety. Compound 6i presented an increased HDAC inhibitory activity (IC 50 0.089 μM) compared to that of suberoylanilide hydroxamic acid (SAHA), an HDAC inhibitor (IC 50 0.15 μM). Docking studies revealed that this compound had a similar binding mode to SAHA in the active site of HDAC1. Treatment with each of four most potent HDAC inhibitors (IC 50 0.089-0.26 μM) decreased the viability of the human breast cancer MDA-MB-231 cells (IC 50 2.98-6.14 μM) as well as chronic myelogenous leukemia K562 cells (IC 50 6.75-12.9 μM). The IC 50 values displayed in MDA-MB-231 and K562 cells by SAHA were 1.32 and 1.69 μM, respectively [59].
In the next study, the same group attempted to increase the anticancer activity of 1,3,4-thiadiazole bearing hydroxamates. Substitution of the phenyl ring did not increase the HDAC inhibitory activity. Three important findings appeared from this part of the experiments: i. phenyl substitution with a bulky group resulted in the loss of the HDAC inhibitory activity; ii. compounds with an electron-withdrawing group exhibited poorer HDAC inhibition than those with an electron-donating group; iii. following tendency in enzymatic potency was found: para-< meta-< ortho-substitution. Replacement of the phenyl ring with a naphthalenyl group led to a loss in HDAC inhibition, while analogues containing pyridine or thiophene displayed higher or similar activity (IC 50 286-411 nM) compared to SAHA (IC 50 416 nM). Compound 35 bearing thiophen-2-yl and 6 carbon units in the linker appeared to be the more active against three cancer cell lines, MDA-MB-231, K562 and human prostate cancer PC3 (IC 50 1.21, 1.56 and 3.6 μM, respectively) compared to SAHA (IC 50 2.29, 1.61 and 5.79 μM, respectively) [60].
A series of 5-substitutedphenyl-1,3,4-thiadiazole-based hydroxamic acids was also evaluated by Nam and co-workers [61]. In this study, compound 5a, of which the HDAC inhibitory activity (Fig. 2), but not cytotoxicity has been previously examined [59], showed strong cytotoxicity against human colon cancer SW620, breast cancer MCF7, prostate cancer PC3, pancreas cancer AsPC1 and lung cancer NCI-H460 cells (IC 50 0.70, 1.80, 0.88, 2.71, 1.07 μM, respectively). Analogs bearing one halogen atom on the phenyl ring, either at position 2, 3 or 4, presented the cytotoxicity comparable or slightly higher to that of compound 5a. Substitution of a chlorine atom at position 2 was the most favorable for cytotoxicity (IC 50 between 0.11 and 1.23 μM). In the same assay, SAHA displayed IC 50 values between 2.77 and 6.42 μM. Of note, the introduction of an additional chlorine at position 6 or nitro substituent at position 2 or 4 remarkably decreased cytotoxicity. Western blot analysis revealed that treatment with compounds showing cytotoxicity comparable or higher to that of SAHA increased level of histone acetylation, suggesting that the anticancer activity of those compounds might be linked to their HDAC inhibitory activity. Results of docking studies indicated that two compounds 5b (bearing 2-chlorophenyl) and 5c (bearing 3-chlorophenyl) had higher binding affinities to HDAC8 compared to SAHA [61].

Inhibitors of kinesin spindle protein
The mitotic kinesins are the proteins responsible for force generators in the process of cell division. The most extensively studied of these proteins is Eg5 (also known as KIF11, kinesin-5 or KSP). Due to its role in cell division, Eg5 is a potential cancer-selective therapeutic target. Indeed, overexpression of this protein is observed in tumors of different origin [62].
Based on the results of a phenotype-based forward chemical genetics screen, Nakai and co-workers selected K858 (N-(4-Acetyl-4,5-dihydro-5-methyl-5-phenyl-1,3,4-thiadiazol-2-yl)acetamide) as an antimitotic agent. This compound induced mitotic arrest, caspase-dependent apoptosis, and cell growth inhibition in human colon cancer HCT-116 cells, but had no effect on microtubule polymerization. K858 blocked centrosome separation and induced the formation of a monopolar spindle during mitosis as well as inhibited the ATPase activity of Eg5 with an IC 50 of 1.3 μM. Of note, this compound appeared to be 150-fold more selective for Eg5 than other members of the kinesin superfamily. It should be also emphasized that K858 induced mitotic cell death in HCT-116 cells, but not in non-cancerous retinal pigment epithelial ARPE-19 cells. Moreover, treatment of mice with 150 mg/kg K858 suppressed tumor growth in a human A2780 ovarian cancer model but no overt evidence of toxicity was found [63].
K858 and its derivative bearing an ethyl moiety at C5 position of the thiadiazole ring, compound 33, displayed significant antitumor activity against human prostate cancer PC3 and melanoma SK-MEL-5 and SK-MEL-28 cells. Both K858 and compound 33 presented the ability to inhibit Eg5 enzymatic activity (Fig. 2) [64]. K858 diminished the viability and induced apoptosis in human breast cancer MCF7, MDA-MB231, BT474 and SKBR3 cells [65] as well as in human glioblastoma U87 and U251 cells [66].
Filanesib has already been evaluated in clinical trials in patients with different cancers (summarized in Table 1). Filanesib demonstrated an acceptable safety profile at dose levels up to 4.5 mg/m 2 in patients with advanced myeloid leukemia (AML). Partial response was observed in 3% and stable disease in 28% of patients. Drug-related serious adverse events, mainly mucositis and neutropenic fever, were observed in 28% of patients and led to study discontinuation in 8% [70]. No partial or complete response was noted in patients with advanced solid tumors treated with filanesib at dose levels up to 2.50 mg/m 2 [71].
In a cohort of 55 patients with multiple myeloma (MM), a combination of filanesib plus bortezomib and dexamethasone demonstrated a favorable safety profile. The overall response rate (ORR) was 20%, the clinical benefit rate (CBR) was 33%, and the disease control rate was 65% [72]. Later, filanesib 1.50 mg/m 2 /day alone or in combination with dexamethasone was evaluated in phase 2 trials in 25 patients with MM. Filanesib has single-agent an ORR of 16% and a clinically meaningful CBR of 23%. The response rates in filanesib/dexamethasone population were also clinically relevant (ORR 15%; CBR 20%) [73]. The results of the most recent clinical trials with filanesib in patients with MM remain unknown.
In phase I of trials in patients with solid tumors (NCT01358019), the recommended dose of litronesib was  ). Sulphur, nitrogen, hydrogen and oxygen atoms are marked as yellow, blue, green or red circles, respectively determined to be 5 mg/m 2 /day. No tumor responses were observed in this study [75]. In more recent trials, partial response to litronesib plus pegfilgrastim was observed in 2% of patients with advanced solid tumors and 20% of patients maintained stable disease (NCT01214629; NCT01214642) [76]. Phase II trials evaluating litronesib in patients with different cancer types (NCT01416389; NCT01025284; NCT01059643) have recently been completed and are summarized in Table 1.

Inhibitors of tubulin polymerization
Kamal and co-workers synthesized a series of compounds with imidazothiadiazole linked with a 3,4,5-trimethoxyphenyl ring, an indolinone ring, and a phenyl group. These imidazo[2,1-b] [1,3,4]thiadiazolindolin-2-ones showed considerable cytotoxicity, with IC 50 values ranging from 1. indolin-2-one) appeared to be most potent with IC 50 values ranging from 1.1 to 1.6 μM and from 2.5 to 2.9 μM, respectively. Compounds 7 and 11 presented anti-tubulin polymerization activity with IC 50 of 0.15 and 1.23 μM, respectively (Fig. 2). Treatment of A549 cells with either of these compounds decreased the level of a polymerized fraction of tubulin and increased the level of its soluble fraction. Docking studies showed that compounds 7 and 11 bound in the colchicine binding site of polymerized tubulin [77].
The same group synthesized a series of conjugates with a core unit of imidazothiadiazole linked with a cyclopropyl ring, an oxindole moiety and an aryl ring. Among these conjugates, compounds 7 ((E)-3-((2-cyclopropyl-6-(4methoxyphenyl)imidazo [2,1-b] [1,3,4] [1,3,4] thiadiazol-5-yl)methylene)indolin-2-one) exhibited cytotoxicity towards the cell lines used in the previous study with GI 50 values from 0.13 to 3.8 μΜ. They also displayed the anti-tubulin activity (IC 50 between 2.8 and 5.6 μM). Treatment with these compounds decreased the level of polymerized fraction of tubulin, induced the cell cycle arrest in the G2/M phase and apoptosis. Similarly to the compounds described in the earlier report, also conjugates 7, 14 and 15 of this study bound in the colchicine-binding site of the tubulin [78].

Abl kinase inhibitors
The Abelson tyrosine kinase (Abl) regulates cytoskeletal dynamics, organelle trafficking, cell proliferation and survival. Its contribution to the initiation and progression of leukemia is relatively well understood, but recent studies indicate the involvement of this protein in pathogenesis of solid tumors as well [79]. Some of the Abl inhibitors turned out to inhibit also Src kinase, the other protein involved in tumor pathogenesis [80].
Radi and co-workers synthesized a series of substituted benzoylamino-2-[(4-benzyl)thio]-1,3,4-thiadiazole which turned out to inhibit either both Abl and Src kinases, or only one of them. The Abl inhibitory activity of those compounds ranged between 0.044 and 1.26 μM, and their Src inhibitory activity was between 0.064 and 1.137 μM (Fig. 2). In the same assay, imatinib, used as a reference drug, displayed inhibitory activity with IC 50

Lipoxygenase inhibitors
Lipoxygenases (LOX, EC 1.13.11) catalyze the oxygenation of the polyunsaturated fatty acids (PUFAs) to form the hydroxyeicosatetraenoic acids (HETEs). Among these enzymes, 15-lipoxygense-1 (15-LOX-1) was found to be involved in the pathogenesis of tumors of different origin, therefore, its targeting may contribute to cancer treatment [83].
Aliabadi and co-workers synthesized a series of compounds bearing 1,3,4-thiadiazole and phthalimide residues and examined the inhibitory effect of these compounds towards 15-LOX-1 (Fig. 2). Compound 4d with meta positioning of the methoxy group afforded the highest inhibitory effect (38%) but did not display a remarkable activity against human prostate cancer PC3, colon adenocarcinoma HT29 or neuroblastoma SKNMC cells. The most toxic towards HT29 cells (IC 50 10.91 μM) was compound 4f with ortho positioning of the fluorine atom. This compound inhibited the activity of 15-LOX-1 in 31%. Compound bearing the fluorine substituent at meta-position displayed some cytotoxicity in SKNMC cells (IC 50 of 50.2 μM) and inhibited the activity of 15-LOX-1 in 26%. Nitro containing derivatives (4a, 4b, 4c) and compound 4 k with para positioning of the chlorine substituent did not show any inhibitory activity against 15-LOX-1, although the latter was cytotoxic towards HT29 and SKNMC cells (IC 50 of 24.06 and 69.7 μM, respectively) [84].
The hybrids of 1,3,4-thiadiazole and chalcone containing phenolic moiety are the other example of molecules binding to DNA. These derivatives exerted the strong cytotoxic activity against leukemia HL-60 cells with IC 50 values in a range from 6.92 to 16.35 μM. Some of the compounds, termed 5a, 5f, 5 h, 5 l, and 5 m, presented cytotoxicity towards cervical cancer HeLa cells, with IC 50 values from 9.12 to 12.72 μM. Lung carcinoma A549 cells were much less susceptible to those compounds (IC 50 values between 21.80 and 92.14 μM). Unfortunately, more sensitive turned out be normal lung cells MRC-5 (IC 50 values from 18.56 to 81.33 μM). The electron-donating and electron-withdrawing groups of the acetophenone moiety seemed to have no influence on the cytotoxic activity against cancer cells, suggesting that the thiadiazole-chalcone pharmacophore played the crucial role in the cytotoxicity of these compounds. The derivatives most potent in HeLa cells, 5a ((E)-N- prop-1-en-1-yl)benzamide), were subjected to further analyses. Treatment with these compounds caused G2/M cell cycle arrest, triggered caspasedependent apoptosis and induced DNA damage [88].
Conformational changes in DNA topology, necessary for transcription, replication and recombination of genetic material, are catalyzed by topoisomerases (Topo). Inhibition of the activity of these enzymes reduces DNA synthesis and cell division. Aside from the molecules that inhibit the catalytic activity of Topo, there are also Topo poisons converting this enzyme into a cell poison, leading to irreversible damage of genetic material [89]. Plech and co-workers synthesized 1,3,4-thiadizole derivatives of which two (compound 3 and 4) stabilized the DNA-TopoII cleavable complex, thus acting as topoII poisons. Among them, compound 3, 2-(4-bromophenylamino)-5-(2,4-dichlorophenyl)-1,3,4-thiadiazole, displayed toxicity towards breast cancer MCF-7 and MDA-MB-231 cells (IC 50 values of 120 and 70 μM, respectively), but did not harm fibroblasts. It diminished DNA synthesis in cancer cells, but not in fibroblasts, and inhibited the activity of topoII (Fig. 2) [90].

Miscellaneous compounds
Anticancer activity of a series of 5-substituted 2(2,4-dihydroxyphenyl)-1,3,4-thiadiazoles was examined by Matysiak and co-workers [91]. Those compounds displayed cytotoxicity towards human cell lines derived from bladder cancer (HCV29T), non-small lung carcinoma (A549), rectal adenocarcinoma (SW707), and breast cancer (T47D). The most active against HCV29T cells were compound 7, bearing 4-(CH 3 ) 3 C-C 6 H 4 group, and compound 26, bearing 4-CH 3 O-C 6 H 4 -CH 2 O group (IC 50 values of 3.7 and 1.1 μg/ mL, respectively). Cisplatin, used as a reference drug, was much more toxic (IC 50 of 0.7 μg/mL). Compounds 7 and 26 displayed significant cytotoxicity towards SW707 (IC 50 values of 4.5 and 5.0 μg/ml, respectively), A549 (IC 50 values of 12.8 and 7.9 μg/mL, respectively) and T47D (IC 50 values of 4.0 and 3.0 μg/mL, respectively) cell lines. The activity of cisplatin was comparable against SW707 (IC 50 value of 4.9 μg/mL), higher against A549 (IC 50 value of 3.3 μg/mL), but lower against T47D (IC 50 value of 6.2 μg/mL) cells. The structure-activity relationship (SAR) analysis indicated that aryl derivatives were more active compared to alkyl derivatives, but the influence of the type of aryl ring substituent on the activity was not significant. Joining the aryl ring by means of either -CH 2 -or -OCH 2 -link also did not improve the anticancer activity of compounds [91]. A further detailed quantitative SAR analysis revealed that electron properties of 1,3,4-thiadiazole ring was the most important factor for the activity of these derivatives. The type of substitution at the fifth carbon atom changed charge distribution of this moiety. Molar refractivity (CMR) turned out to be another parameter influencing the activity of 1,3,4-thiadiazole derivatives [92].
In the next study, cytotoxicity of a series of differently substituted in N-aryl ring 2-phenyloamino-5-(2,4dihydroxyphenyl)-1,3,4-thiadiazoles against four aforementioned cell lines was examined. Compound I, with an unsubstituted amine group, showed a weak activity against HCV29T cells with IC 50 value of 205.7 μM. All N-substituted derivatives, except for compounds bearing either 4-CH 3 -3-Cl-C 6 H 3 -or 4-CH 3 O-C 6 H 4 group, were significantly more potent in this cell line and displayed IC 50 ranging from 20.7 to 115 μM. Compound VIII with iodine atom in the para-position of the N-phenyl ring was the most active of the tested derivatives, yet less potent than cisplatin which IC 50 value was 2.3 μM. This reference drug presented also considerable cytotoxicity in A549, SW707, and T47D cells (IC 50 values of 11, 16.3 and 20.7 μM, respectively). None of the tested 2-phenyloamino-5-(2,4-dihydroxyphenyl)-1,3,4thiadiazole derivatives appeared to be more active than cisplatin in either A549 or SW707 cells (IC 50 values ranging from 17.0 to 98.5 μM and from 19.5 to 96.7 μM, respectively). However, in T47D cell line, most of the tested compounds were more active (IC 50 values from 9.7 to 19.9 μM) compared to cisplatin (IC 50 value 20.7 μM). The highest activity displayed derivatives bearing either bromine or iodine atom in the para-position of aryl ring as well as 2-methyl-5-chlorophenyl derivative, suggesting an advantageous influence of electron-withdrawing (σ > 0) and hydrophobic (π > 0) substituents and polarizability of bromine and iodine atoms [93].
Among another series of N-substituted 2-amino-5-(2,4dihydroxyphenyl)-1,3,4-thiadiazoles, phenyl derivatives turned out to be relatively active, although their potency was dependent on N-aromatic ring-substitution degree. In SW707 cells, the parent compound 4 displayed cytotoxicity comparable to that of cisplatin (IC 50 values of 4.3 and 4.9 μg/mL, respectively). In three other cell lines, this compound was slightly less active compared to the reference drug. Substitution of phenyl ring with either the lipophilic electron-donating or morpholinoalkyl groups decreased the activity of the parent compound. On the contrary, the compound with the hydrophobic substituents (π > 0) of electronwithdrawing character (α > 0) appeared to be much more promising anticancer agents. Thus, the substitution of the ring with a fluorine atom in the para-position resulted in the activity towards HCV29T, SW707, T47D cells (IC 50 values 6.2, 3.6 and 4.2 μg/mL, respectively) comparable or slightly higher than parent compound 4 and cisplatin. Similarly, the compound with a chlorine atom in the meta-position displayed cytotoxicity in HCV29T, SW707 and T47D cells (IC 50 values 5.4, 3.7 and 3.9 μg/mL, respectively) comparable to that of the parent compound and the reference drug. Either modification did not increase the activity towards A549 cells. However, the substitution of the ring with two chlorine atoms in positions 2 and 4 resulted in the cytotoxicity comparable (IC 50 value 5.3 μg/mL) in A549 cells or remarkably higher (IC 50 values of 2.8 and 1.5 μg/mL, respectively) in SW707 and T47D cells to that of compound 4 and cisplatin. This compound turned out to be much less active in HCV29T cells (IC 50 value 22.8 μg/mL). These results supported the previous notion that 2-amino-1,3,4thiadiazole acts as pharmacophore of anticancer activity and 2,4-dihydroxyphenyl moiety in position 5 significantly contributes to cytotoxicity. Most likely, such a substituent not only contributes to the favorable hydrophobic-hydrophilic character but also affects electronic properties crucial in compound-target(s) interactions responsible for cytotoxic properties [94].
The promising results obtained for N-halogenphenyl derivatives resulted in a further in-depth analysis of those compounds. Among them, 2-(4-chlorophenylamino)-5-(2,4dihydroxyphenyl)-1,3,4-thiadiazole (herein referred to as CPDT) displayed the ability to diminish the viability of cancer cells. This compound turned out to be most toxic towards human thyroid carcinoma FTC238, colon carcinoma HT-29 and leukemia Jurkat cells (IC 50 values from 6.4 to 6.7 μM) and mouse teratoma P19 cells (IC 50 value 8.5 μM). Less susceptible to CPDT treatment were human breast carcinoma T47D, medulloblastoma TE671, astrocytoma MOGGCCM and rat glioma C6 cells (IC 50 values of 10.7, 15.3, 19.4 and 12.7 μM, respectively). CPDT treatment diminished cancer cell proliferation and migration but did not affect the viability of non-cancerous rat astrocytes, neurons, hepatocytes as well as human fibroblasts [95].
The other derivative, 2-(4-fluorophenyloamino)-5-(2,4dihydroxyphenyl)-1,3,4-thiadiazole, herein termed FPDT, also diminished the viability of several cancer cell lines. Threshold concentrations of this compound required to elicit cytotoxic effect were as low as 5 μM (human colon cancer HT-29 cells), 10 μM (human lung carcinoma A549 and medulloblastoma TE671 cells), and 25 μM (human neuroblastoma SK-N-AS and rat glioma C6 cells). Moreover, FPDT treatment diminished proliferation and migration of C6 and A549 cells. It should be emphasized that FPDT did not affect non-cancerous cells, as rat astrocytes, neurons and hepatocytes were resistant to this compound up to 100 μM. Furthermore, neurotoxicity caused either by serum-deprivation or glutamate treatment was ameliorated by co-exposure to FPDT, indicating a neuroprotective activity of this compound. Quantum-chemical calculations confirmed that aminothiadiazole moiety acted as pharmacophore of cytotoxic activity of FPDT, and both the 2,4-dihydroxyphenyl and para-fluorophenyl substituents intensified its properties [96].
The more detailed molecular analysis revealed that in A549 cells treatment with FPDT decreased phosphorylation level of kinases involved in tumorigenesis, MEK1/2 and ERK1/2, as well as its downstream target, a transcription factor CREB. Furthermore, FPDT treatment enhanced the expression of p27/Kip1, members of the Cip/Kip family of cyclin-dependent kinase inhibitors which block cell cycle progression through the G1/S phase [97]. Indeed, treatment of A549 cells with FPDT increased the number of cells in the G0/G1 phase and decreased the number of cells in the S and G2/M phases [98]. Clearly further studies are required to elucidate whether and to what extend downregulation of the ERK pathway contributes to the anti-cancer properties of FPDT. Of note, our unpublished data strongly indicate that anticancer activity of this compound observed in human glioblastoma cells is causatively linked to downregulation of the AKT pathway, a cascade crucial for the pathogenesis of these tumors. It is, therefore, tempting to speculate that the mechanism underlying FPDT's activity is related to some upstream signal molecules of the AKT and ERK pathways. For instance, several receptor tyrosine kinases (RTK) have been shown to activate RAS proteins, which in turn modulate both AKT and ERK pathways [99].

Conclusions
The anticancer potential of numerous thiadiazole derivatives is undeniable. However, the studies reviewed above have several limitations. First, caution needs to be taken when comparing compounds' efficacy determined by different assay methods. Second, the activity of the majority of thiadiazole derivatives was evaluated in cancer cells, while their activity in non-cancerous cells remains unknown. Third, in most of the cited studies, the compounds' efficacy was evaluated in models in vitro. Finally, the molecular targets of the majority of thiadiazole derivatives have not been identified so far. Further in-depth analyses are needed to determine compounds' selectivity, bioavailability and mode of action. Nevertheless, data depicted above clearly indicate that thiadiazole moiety may be valuable lead structure in the development of new compounds with improved anticancer activity.