In summary, we have used three-dimensional molecular models of the E. coli DHFR-MTX complex to design analogues of TMP that not only had significantly higher affinity for DHFR than that of TMP but also furnished useful information on the binding mode of this class of inhibitor in solution. The postulated binding mode was then verified by X-ray crystallographic studies of TMP and two of these analogues in complex with E. coli DHFR. Although these analogues were not as effective as TMP as broad-spectrum antibacterials, we feel that this study amply demonstrates the considerable potential of this approach to inhibitor design. In summary, members of 2,3,6-trisubstituted clonidine-like 2-(phenylimino)imidazolidines show potential for pronounced hypotensive activity following systemic administration. This new class of derivatives needs further exploration with respect to their substituent allowance on the phenyl ring. Recently, we suggested1 that the biological activity of chloramphenicol (la)-an antibiotic inhibiting procaryotic ribosomal protein synthesis-can be explained in terms of retro-inverso relationship2 to the amino acid moiety of another strong inhibitor puromycin3. It should be possible to extend this hypothesis to other antibiotics, for example, sparsomycin4, that carry an acylamido function attached to the asymmetric (D) carbon of the substituted propanol moiety and interfere with ribosomal protein synthesis. This approach would lead to novel synthetic antibiotics by a simple interchange of the relevant N-acyl residues. We now report on the first case5 of such a hybrid antibiotic (lb) derived from a combination of chloramphenicol (la) and sparsomycin (1c)-hence, the suggested name sparsophenicol-which is indeed a strong inhibitor of ribosomal peptide synthesis.