β-Lactam compounds act on penicillin-recognizing enzymes via acylation of the hydroxyl group of an active site serine. When the resulting acyl enzyme is kinetically stable, as in the case of a penicillin-binding protein (PBP), the biosynthesis of a bacterial cell wall is inhibited, and death of the organism results. The de novo design of an antibacterial agent targeted to a PBP might be possible if the three-dimensional structural requirements of the equilibrium (i.e, fit) and catalytic (i.e. reactivity) steps of the aforementioned enzymatic process could be determined. For a model of the active site of a PBP from Streptomyces R61, the use of molecular mechanics calculations to treat "fit," and ab initio molecular orbital calculations to treat "reactivity," leads to the idea that the carboxyl group (G1) and the amide N-H (G2) of the antibiotic are hydrogen bonded to a lysine amino group and a valine carbonyl group in the enzyme–substrate complex. These two hydrogen bonds place the serine hydroxyl group on the convex face of the antibiotic, in position for attack on the β-lactam ring by a neutral reaction, catalyzed by water, that involves a direct proton transfer to the β-lactam nitrogen. Molecular orbital calculations of structure–reactivity relations associated with this mechanism suggest that C=N is bioisosteric to the β-lactam N-C(=O), comparable to a β-lactam in its reactivity with an alcohol, and that the product RO(C-N)H is formed essentially irreversibly (−ΔE > 10 kcal/mol). Accordingly, structures containing a G1 and a G2 separated by a C=N, and positioned in different ways with respect to this functional group, have been synthesized computationally and examined for their ability to fit to the PBP model. This strategy identified a 2H-5,6-dihydro-1,4-thiazine substituted by hydroxyl and carboxyl groups as a target for chemical synthesis. However, exploratory experiments suggested that the C=N of this compound equilibrates with endocyclic and exocyclic enamine tautomers. This required that the C2 position be substituted, and that the hydroxyl group not be attached to the carbon atom adjacent to the C=N. These conditions are met in a 2,2-dimethyl-3-(2-hydroxypropyl)-1,4-thiazine, which also exhibits the necessary fit to the PBP model. Two epimers of this compound have been synthesized, from D- and L-serine. The compound derived from L-serine is not active. The compound derived from D-serine exhibits antibacterial activity, but is unstable, and binding studies with PBP's have not been performed. It is hoped that these studies can be carried out if modification of the lead structure leads to compounds with improved chemical stability.
β-内酰胺化合物通过酰化活性位点丝氨酸的羟基作用于青霉素识别酶。当产生的酰酶在动力学上是稳定的,如青霉素结合蛋白(PBP)的情况下,细菌细胞壁的生物合成被抑制,导致生物体的死亡。如果能确定青霉素结合蛋白的三维结构要求(即平衡的适合性)和催化(即反应性)步骤的需求,可能会实现针对PBP的抗菌剂的全新设计。对来自链霉菌R61的PBP的活性位点模型,使用分子力学计算处理“适合性”,并使用从头算分子轨道计算处理“反应性”,得出一个想法,即抗生素的羧基(G1)和酰胺N-H(G2)与酶-底物复合物中的赖氨酸氨基团和缬氨酸羰基团形成氢键。这两个氢键将丝氨酸羟基置于抗生素的凸面上,处于通过中性反应由水催化的直接质子转移至β-内酰胺氮原子的位置。与这种机制相关的结构-反应关系的分子轨道计算表明,C=N与β-内酰胺N-C(=O)是生物等同物,与醇类反应时与β-内酰胺相当,产物RO(C-N)H基本上是不可逆的(−ΔE > 10 kcal/mol)。因此,含有由C=N分隔的G1和G2的结构,并以不同方式相对于这个功能团定位的结构已经通过计算合成并检验其适合于PBP模型的能力。这种策略确定了一个被羟基和羧基取代的2H-5,6-二氢-1,4-噻嗪作为化学合成的目标。然而,探索性实验表明,该化合物的C=N与内环和外环烯胺互变异构体平衡。这要求对C2位置进行取代,并且羟基不能连接到邻近C=N的碳原子。这些条件在2,2-二甲基-3-(2-羟基丙基)-1,4-噻嗪中得到满足,该化合物还展现出与PBP模型的必要适合性。这种化合物的两个对映体已经从D-丝氨酸和L-丝氨酸合成。从L-丝氨酸衍生的化合物不活跃。从D-丝氨酸衍生的化合物表现出抗菌活性,但不稳定,并且尚未进行与PBP的结合研究。希望如果引物结构的修改导致具有改善化学稳定性的化合物,这些研究可以进行。