DESIGN AND CHARACTERIZATION OF A FLUOROGENIC SUBSTRATE SELECTIVELY HYDROLYZED BY STROMELYSIN-1 (MATRIX METALLOPROTEINASE-3)

Members of the matrix metalloproteinase (MMP) family have been implica ted in disease states such as arthritis, periodontal disease, and tumo r cell invasion and metastasis. Stromelysin 1 (MMP-3) has a broad subs trate specificity and participates in the activation of several MMP zy mogens. We examined known sequences of MMP-3 cleavage sites in natural peptides and proteins and compared sequence specificities of MMP-3 an d interstitial collagenase (MMP-1) in order to design fluorogenic subs trates that (i) would be hydrolyzed rapidly by MMP-3, (ii) would discr iminate between MMP-3 and MMP-1, and (iii) could be monitored continuo usly without interference from MMP amino acid residues. Designed subst rates were then screened for activity toward MMP-1, gelatinase A (MMP- 2), MMP-3, and gelatinase B (MMP-9). The first of these substrates, NF F-1 Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Lys-(Dnp)-Gly, where Mca is (7 -methoxycoumarin-4-yl)acetyl and Dnp is 2,4-dinitrophenyl), was hydrol yzed equally well by MMP-3 and MMP-2 (K-cat/K-m similar to 11,000 S-1 M(-1)) MMP-1 had 25% of the activity of MMP-3 toward NFF-1. The second substrate, NEFF-2 (Mca-Arg-Pro-Lys-Pro-Tyr-Ala-Nva-Trp- Met-Lys(Dnp)- NH2, where Nva is norvaline), was hydrolyzed 60 times more rapidly by MMP-3 (k(cat)/K-m = 59,400 s(-1) M(-1)) than MMP-1. Unfortunately, NFF -2 showed little discrimination between MMP-3, MMP-2 (k(cat)/K-m = 54, 000 s(-1) M(-1)), and MMP-9 (k(cat)/K-m = 55,300 s(-1) M(-1)). The thi rd substrate, NFF-3 Arg-Pro-Lys-Pro-Val-Glu-Nva-Trp-Arg-Lys(Dnp)-NH2), was hydrolyzed rapidly by MMP-3 (K-cat/K-m = 218,000 s(-1) M(-1)) and very slowly by MMP-9 (k(cat)/K-m = 10,100 s(-1) M(-1)), but there was no significant hydrolysis by MMP-1 and MMP-2. NFF-3 is the first docu mented synthetic substrate hydrolyzed by only certain members of the M MP family and thus has important application for the discrimination of MMP-3 activity from that of other MMPs. Although NFF-3 was designed b y assuming that substrate subsites were independent and hence free ene rgy changes derived from single mutation experiments were additive, we found discrepancies between predicted and experimental k(cat)/K-m val ues, one on the order of 2000-5000. Thus, the design of additional dis criminatory MMP substrates may require approaches other than assuming additive free energy changes, such as screening synthetic libraries an d consideration of secondary and tertiary structures of substrates and the enzyme.