Design and Selection of Novel C1s Inhibitors by In Silico and In Vitro Approaches

Abstract

The complement system is associated with various diseases such as inflammation or auto-immune diseases. Complement-targeted drugs could provide novel therapeutic intervention against the above diseases. C1s, a serine protease, plays an important role in the CS and could be an attractive target since it blocks the system at an early stage of the complement cascade. Designing C1 inhibitors is particularly challenging since known inhibitors are restricted to a narrow bioactive chemical space in addition selectivity over other serine proteases is an important requirement. The typical architecture of a small molecule inhibitor of C1s contains an amidine (or guanidine) residue, however, the discovery of non-amidine inhibitors might have high value, particularly if novel chemotypes and/or compounds displaying improved selectivity are identified. We applied various virtual screening approaches to identify C1s focused libraries that lack the amidine/guanidine functionalities, then the in silico generated libraries were evaluated by in vitro biological assays. While 3D structure-based methods were not suitable for virtual screening of C1s inhibitors, and a 2D similarity search did not lead to novel chemotypes, pharmacophore model generation allowed us to identify two novel chemotypes with submicromolar activities. In three screening rounds we tested altogether 89 compounds and identified 20 hit compounds (<10 μM activities; overall hit rate: 22.5%). The highest activity determined was 12 nM (1,2,4-triazole), while for the newly identified chemotypes (1,3-benzoxazin-4-one and thieno[2,3-d][1,3]oxazin-4-one) it was 241 nM and 549 nM, respectively.

Autoimmune hemolytic anemia: current knowledge and perspectives

“…In silico and in vitro research is ongoing to find new compounds that could inhibit the initial complement pathways. Studies showed that potentially the highest activity compound was 1,2,4triazole [128].…”

Application 2D Descriptors and Artificial Neural Networks for Beta-Glucosidase Inhibitors Screening

“…The compounds were classified as low-active or inactive (class 0, IC 50 > 50 μM) and significantly active (class 1, IC 50 ≤ 50 μM). This is an arbitrary criterion; however, the 50 μM threshold has been widely used for evaluating various biological activities including enzymes inhibitors [ 90 , 91 , 92 , 93 , 94 , 95 ]. All data used for model development and validation are summarized in the Supplementary Materials (Table S1) .…”

Complement C1s as a diagnostic marker and therapeutic target: Progress and propective

“…A large machine learning-based virtual screening was carried out and found a series of potential valuable inhibitors ( 66 ). Further combination of in silico and in vitro approaches identified hit compounds with new chemo-types and high potency in inhibiting C1s ( 67 ). It is interesting to determine whether these inhibitors can interfere the activity of C1s under physiological and pathological conditions.…”