Symmetric molecules with 1,4-triazole moieties as potent inhibitors of tumour-associated lactate dehydrogenase-A.

A series of symmetric molecules incorporating aryl or pyridyl moieties as central core and 1,4-substituted triazoles as a side bridge was synthesised. The new compounds were investigated as lactate dehydro-genase (LDH, EC 1.1.1.27) inhibitors. The cancer associated LDHA isoform was inhibited with IC50 = 117-174 µM. Seven compounds exhibited better LDHA inhibition (IC50 117-136 µM) compared to known LDH inhibitor - galloflavin (IC50 157 µM).

Introduction

The lactate dehydrogenase (LDH, EC 1.1.1.27) is one of the most abundant proteins and it is expressed in all tissues. The main function of LDH is interconversion of lactate and pyruvate with accompanying interconversion of NAD+ and NADH.

Three LDH isoforms are present in humans – LDHA, LDHB and LDHC. LDHA and LDHB are expressed in all cells, whereas LDHC is produced only in testis. In the active form, LDH is a tetramer formed of LDHA and LDHB in various ratios making five tetramers: LDH1 (4 × LDHB), LDH2 (1 × LDHA/3 × LDHB), LDH3 (2 × LDHA/2 × LDHB), LDH4 (1 × LDHA/3 × LDHB) and LDH5 (4 × LDHA).

In normal cells, predominant is LDHB where it converts lactate to pyruvate with interconversion of NAD+ into NADH, which allows cells to use lactate as a nutrient source for oxidative metabolism, and/or for gluconeogenesis. LDHA is the predominant isoform found in skeletal muscle and other highly glycolytic tissues. In contrast to LDHB, LDHA has a higher affinity for pyruvate, that is, LDHA and LDH5 tetramer in particular predominantly converts pyruvate to lactate with consumption of one NADH molecule to produce NAD+ which in turn is essential in glycolysis. Cancer cells mainly generate energy through glycolysis even in the presence of normal oxygen pressure. Since the LDHA is the final enzyme in glycolysis pathway where generated NAD+ is necessary for continued high glycolysis rate in cancer cells (Warburg effect), LDHA is an important supporter of glucose metabolism in cancer cells and can affect tumourigenesis and metastasis. Additionally, elevated levels of LDHA are markers of many tumours, the majority of them are highly glycolytic, and high LDHA levels are related with poor prognosis, for instance in several human malignancies. Therefore, LDHA is defined as anticancer drug target,. Notably, a limited number of LDHA inhibitors is reported in the literature so far. In several papers, very promising LDHA inhibition results have been reported. As a one of the first inhibitors with low-micromolar inhibition activity (Ki = 13.5 μM) compound FX11 (Figure 1) along with two similar compounds was reported in 2010. Utilising NMR and SPR (surface plasmon resonance) fragment based hit identification technique and further optimisation of structure a series with low- and submicromolar LDHA inhibition was obtained with best lead A (IC50 = 0.27 μM). In the other study, also using fragment-based hit identification, series of new LDHA inhibitors was obtained. In this series, IC50 values ranged from 59 to 0.12 μM, where the best inhibition was observed for compound B. It was noted that in this series carbohydrate moiety in the middle of the molecule and its stereochemistry had significant influence on the inhibition activity. In very recent work through docking-based virtual screening, several potential LDHA inhibitors were identified. One of the compounds identified (C) showed very good LDHA inhibition potency in vitro with an IC50 value of 0.33 μM.

Figure 1.

Examples of chemical structures of known LDHA inhibitors.

Results and discussions

Here, we report the synthesis of symmetric molecules incorporating aryl or pyridyl moieties as central core and 1,4-substituted triazoles as a side bridge and they evaluation as LDHA inhibitors.

At the beginning of our study based on literature data, we assumed that V-shape structures are beneficiary for good LDHA inhibition. Our assumption was based on published X-ray structures, for instance for B–LDHA complex (PDB code 4I9H). It was also obvious that besides V-shape of the molecule and appropriate length of V-“arms” the terminal groups have to be able to make hydrogen bonding, therefore we chose carboxylic, sulphonamide and nitro groups as terminal ones for this study.

Chemistry

The synthesis of desired inhibitors 7a-c was started from commercially available 1,3-dibromobenzene (1) which was reacted with trimethylsilylacetylene in Sonogashira reaction to provide bis-TMS protected derivative 2 (Scheme 1). Following deprotection with KF afforded building block 3 in good yield over two steps. Azides 6a-c necessary for Cu-mediated click reaction were prepared in two steps from commercially available anilines 4a-c. Acylation of anilines 4a-c with chloroacetyl chloride afforded chlorides 5a-c in good yields and following treatment with NaN3 provided azide building blocks 6a-c also in good yields. Reaction of building block 3 with 6a-c under acidic click reaction condition provided inhibitors 7a-c.

Scheme 1.

Reagents and conditions: (i) trimethylsilylacetylene, Pd2(PPh3)2, CuI, i-Pr2NH, THF, 70 °C; (ii) KF, MeOH/THF, rt, 69% for two steps; (iii) chloroacetyl chloride, K2CO3, THF, 0 °C, 5a (98%), 5b (84%), 5c (89%); (iv) NaN3, DMF, rt, 6a (97%), 6b (68%), 6c (95%); (v) CuSO4.5H2O, sodium ascorbate, AcOH, DMF/H2O, rt, 7a (32%), 7b (71%), 7c (41%). All detailed experimental procedures are provided in the Supplemental data.

The same strategy as described earlier was utilised for the synthesis of inhibitors 11a and 11b. First, aminomethylphenyl derivatives 8a,b were acylated with chloroacelyl chloride to obtain intermediates 9a and 9b, which were converted into corresponding azides 10a,b by treatment with NaN3 (Scheme 2). Following reaction of azides 10a,b with building block 3 under acidic click reaction conditions provided desired inhibitors 11a and 11b.

Scheme 2.

Reagents and conditions: (i) chloroacetyl chloride, K2CO3, THF, 0 °C to rt, 9a (33%), 9b (69%); (ii) NaN3, DMF, rt, 10a (78%), 10b (77); (iii) 3, CuSO4.5H2O, sodium ascorbate, AcOH, DMF/H2O, rt, 11a (17%), 11b (59%).

Bis-acetylenylpyridine building block 14 was prepared in similar way to compound 3, where 3,5-dibromopyridine 12 was first reacted with TMS-acetylene under Sonogashira reaction conditions and then TMS were eliminated in obtained intermediate 13 by treatment with K2CO3 proving building block 14 in 67% yield over two steps (Scheme 3).

Scheme 3.

Reagents and conditions: (i) trimethylsilylacetylene, Pd2(PPh3)2, CuI, Et3N, 75 °C; (ii) K2CO3, MeOH/THF, rt, 67% for two steps; (iii) 6a-c, CuSO4.5H2O, sodium ascorbate, AcOH, DMF/H2O, rt, 15a (85%), 15b (39%), 15c (73%); (iv) 10a,b, CuSO4.5H2O, sodium ascorbate, AcOH, DMF/H2O, rt, 16a (63%), 16b (46%).

The reaction between building blocks 13 and 6a-c under acidic click reaction conditions provided inhibitors 15a-c (Scheme 3). In turn, treatment of 13 by azides 10a,b under the same condition provided inhibitors 16a,b with extended bridge.

And finally, for the synthesis of inhibitors 20a-c and 21a,b, anisole building block 19 was synthesised in analogy to 3 and 14 as above, starting synthesis from dibromoanisole 17. Bis-TMS protected intermediate 18 was obtained under Sonogashira reaction conditions and following deprotection by potassium hydroxide provided building block 19 in 61% yield over two steps (Scheme 4).

Scheme 4.

Reagents and conditions: (i) trimethylsilylacetylene, Pd2(PPh3)2, CuI, Et3N, THF 65 °C; (ii) KOH, MeOH/THF, rt, 61% for two steps; (iii) 6a-c, CuSO4.5H2O, sodium ascorbate, AcOH, DMF/H2O, rt, 20a (64%), 20b (56%), 20c (47%); (iv) 10a,b, CuSO4.5H2O, sodium ascorbate, AcOH, DMF/H2O, rt, 21a (31%), 21b (85%).

Treatment of 19 with azides 6a-c under the same condition as described above provided inhibitors 20a-c (Scheme 4). In turn, treatment of 19 with azides 10a,b afforded desired inhibitors 21a,b.

The structures of all new compounds synthesised were fully approved by 1H and 13C NMR, IR and HRMS data (see Supplemental data).

Inhibition studies

All 15 symmetrical compounds were evaluated for their ability to inhibit LDHA, for the comparison galloflavin as a known isoform nonselective LDHA inhibitor was used. Our choice of galloflavin was based on the fact that it is extensively studied as potent anticancer agents in recent years; additionally, galloflavin is already commercially available.

Even though all compounds exhibited similar LDHA inhibition and it is difficult to perform structure-activity-relationship (SAR) analysis, we can divide this compound into two groups. First one, the most active compounds (7a, 7b, 15a, 15b, 16b, 20b and 21b) showed better LDHA inhibition compared to galloflalvin, ranging IC50 values from 117 to 136 µM for compounds in first group, whereas galloflavin has IC50 = 157 µM (Table 1). The most active compound in this group was 16b (IC50 = 117 µM), which incorporated pyridyl moiety as a central core and carboxylic groups as a terminal ones. Second group of compounds (7c, 11a, 11b, 15c, 16a, 20a, 20c and 21a) exhibited equal or weaker inhibition compared to galloflavin, ranging IC50 values from 156 to 174 µM. The most active compounds in this group were 11a, 11b and 15c (IC50 158, 156 and 156 µM, respectively). Similar activity of this three compounds is hard to explain, where compounds 11a and 11b have phenyl ring as a central core and sulphonamide and carboxyl groups as terminal ones on aminomethylphenyl moieties, but compounds 15c has pyridyl central moiety and nitro groups as terminal one on shorter, aniline containing, bridge.

Table 1.

Inhibition data of human LDHA.

Compound	Inhibition, IC50 (μM)*
7a	128
7b	120
7c	172
11a	158
11b	156
15a	136
15b	125
15c	156
16a	174
16b	117
20a	165
20b	127
20c	163
21a	173
21b	128
Galloflavin	157
Blank	0

Mean from three different assays, by colorimetric assay measuring the absorbance at 450 nm (errors were in the range of ±5–10% of the reported values).

We hypothesise that the putative interaction of the inhibitors described in this work with LDHA is similar to the published one, that is, inhibitors bound in the active centre close to conserved residues involved in the catalytic processing of LDHA substrates. We also assume that at the same time the NADH cofactor is bound in the active centre near the inhibitor molecule. This binding region of the protein is known to undergo considerable conformational changes during the catalytic cycle, that is, it has rather high flexibility. Therefore, inhibitor binding might be unstable and this reflects in IC50 values obtained.

Conclusions

In conclusion, a series of new symmetric molecules have been designed and synthesises as potent LDHA inhibitors. The compounds synthesised exhibited promising in vitro LDHA inhibition activity, where seven compounds were better inhibitors (IC50 117–136 µM) as known LDH inhibitor – galloflavin (IC50 157 µM), and other eight showed equal or slightly lower inhibitory activity (IC50 156–174 µM) as galloflavin. The results obtained are promising base for further development of novel LDH inhibitors.