Derivatization chemistry of the double-decker dicobalt sandwich ion targeted to design biologically active substances

Abstract

A synthesis of the first double-decker sandwich ion [(1',2'-C2B9H11)-3,3'-Co-(1,2-C2B8H10)-6,3"-Co-(1",2"-C2B9H11)]2– (DD2–) derivatives is described, having been developed in connection with our search for biologically active substances. A series of B-substituted hydroxyl derivatives was prepared by direct hydroxylation of the ion using aqueous sulfuric acid. Two isomers of monohydroxy derivatives were isolated. The main product was substituted at the central “canastide” ion fragment, whereas the substitution site for the minor isomer corresponded to a B(8) atom of one of the terminal 11-vertex dicarbollide parts. Similarly, the disubstitution occurred slightly more preferentially on the “canastide” fragment providing the main isomeric derivative with a symmetric structure. The cesium salt of this ion was characterized by X-ray diffraction. Two other isomeric species have one substituent sitting on the “canastide” ion and the second present on the dicarbollide ligand in apart or syn-geometric arrangement. A new zwitterion anion [(1',2'-C2B9H11)-3',3-Co-(8-(CH2)4O-1,2-C2B8H9)-6,3"-Co-(1",2"-C2B9H11)-]1– was prepared by the reaction of the parent ion with tetrahydrofuran (THF), activated by BF3·OEt2. This new compound serves as a versatile building block for constructing organic derivatives, as exemplified by the ring cleavage by various amines or phenolate ions and the synthesis of a basic series of compounds of general formulation [(1',2'-C2B9H11)-3',3-Co-(8-X-(CH2)4O-1,2-C2B8H9)-6,3"-Co-(1",2"-C2B9H11)]n– where the organic end-groups X adjacent to the “canastide” moiety via a B-oxatetramethylene spacer corresponds to C4H9NH2, NC5H5, N(C2H5)3, (C6H5)3P (n = 1), or (4-t-Bu-C6H4-1-O)– and (2-CH3O-C6H4O)– (n = 2). We show that dicluster compounds with two identical DD2– anion units or asymmetric molecules containing two different clusters, the cobalt bis(dicarbollide) and the DD2– anion, are accessible using this building block. All compounds were characterized by high-resolution NMR (1H, 13C, and 11B) and mass spectrometry. Some of the compounds were tested by in vitro assay for their ability to inhibit the HIV-protease (HIV-PR) enzyme. The majority of the tested species proved substantially high activity toward the HIV-PR, exhibiting on the other hand a noncompetitive mechanism of the inhibition.