Resistance to β-lactam antibiotics in Gram-negatives producing metallo-β-lactamases (MBLs) represents a major medical threat and there is an extremely urgent need to develop clinically useful inhibitors. We previously reported the original binding mode of 5-substituted-4-amino/H-1,2,4-triazole-3-thione compounds in the catalytic site of an MBL. Moreover, we showed that, although moderately potent, they represented a promising basis for the development of broad-spectrum MBL inhibitors. Here, we synthesized and characterized a large number of 4-amino-1,2,4-triazole-3-thione-derived Schiff bases. Compared to the previous series, the presence of an aryl moiety at position 4 afforded an average 10-fold increase in potency. Among 90 synthetic compounds, more than half inhibited at least one of the six tested MBLs (L1, VIM-4, VIM-2, NDM-1, IMP-1, CphA) with K_i values in the μM to sub-μM range. Several were broad-spectrum inhibitors, also inhibiting the most clinically relevant VIM-2 and NDM-1. Active compounds generally contained halogenated, bicyclic aryl or phenolic moieties at position 5, and one substituent among o-benzoic, 2,4-dihydroxyphenyl, p-benzyloxyphenyl or 3-(m-benzoyl)-phenyl at position 4. The crystallographic structure of VIM-2 in complex with an inhibitor showed the expected binding between the triazole-thione moiety and the dinuclear centre and also revealed a network of interactions involving Phe61, Tyr67, Trp87 and the conserved Asn233. Microbiological analysis suggested that the potentiation activity of the compounds was limited by poor outer membrane penetration or efflux. This was supported by the ability of one compound to restore the susceptibility of an NDM-1-producing E. coli clinical strain toward several β-lactams in the presence only of a sub-inhibitory concentration of colistin, a permeabilizing agent. Finally, some compounds were tested against the structurally similar di-zinc human glyoxalase II and found weaker inhibitors of the latter enzyme, thus showing a promising selectivity towards MBLs.

产生金属β-内酰胺酶（MBL）的革兰氏阴性菌对β-内酰胺抗生素的耐药性代表着重大的医学威胁，并且迫切需要开发临床上有用的抑制剂。我们先前报道了MBL催化位点中的5个取代的4-氨基/ H-1,2,4-三唑-3-硫酮化合物的原始结合模式。此外，我们表明，尽管具有中等效力，但它们代表了广谱MBL抑制剂开发的有希望的基础。在这里，我们合成和表征了大量的4-氨基-1,2,4-三唑-3-硫酮衍生的席夫碱。与先前的系列相比，在4位上存在芳基部分可使效力平均提高10倍。在90种合成化合物中，超过一半的化合物抑制了至少六个被测MBL（L1，VIM-4，VIM-2，NDM-1，IMP-1，K _i的值在μM至亚μM范围内。几种是广谱抑制剂，也可以抑制临床上最相关的VIM-2和NDM-1。活性化合物通常在位置5处包含卤代，双环芳基或酚基部分，以及在邻苯甲酸，2,4-二羟基苯基，对苄氧基苯基或3-（m-苯甲酰基）-苯基在第4位。与抑制剂形成复合物的VIM-2晶体结构显示，三唑-硫酮部分与双核中心之间具有预期的结合，并且还揭示了涉及Phe61，Tyr67，Trp87和Phe61的相互作用网络。保守的Asn233。微生物学分析表明，化合物的增强活性受到不良的外膜渗透或流出的限制。一种化合物恢复生产NDM-1的大肠杆菌的敏感性的能力证明了这一点。仅在亚抑制浓度的粘菌素（一种渗透剂）存在下，针对几种β-内酰胺的临床菌株。最后，对某些化合物针对结构相似的二锌人乙二醛酶II进行了测试，发现了后者酶的抑制剂较弱，因此显示出对MBL的有希望的选择性。