120 INSTRUMENTAL METHODS
upon enzyme activation; (2) a catalytic domain responsible for the enzyme ’ s
hydrolase activity — the enzyme domain studied here; and (3) a hemopexin -
like domain that may necessary for substrate recognition. Much pharmaceuti-
cal research has taken place to fi nd inhibitors of MMP because misregulation
or overexpression of the enzyme is known to be a factor in several disease
states. For instance, MMP expression and activity increases in almost every
type of human cancer. Its expression correlates with increased invasiveness,
metatheses (spreading of the tumor to other parts of the body), and decreased
survival rates in cancer patients. No MMP inhibitor has passed the clinical trial
stage of drug development because of side effects caused by their low selectiv-
ity. Researchers believe that one reason for the lack of selectivity may be due
to the inherent fl exibility of the MMP protein ’ s backbone. The Bertini group ’ s
research looked at X - ray structures of several different MMPs complexed with
different inhibitors fi nding that certain loop regions in the protein show
enhanced mobility and/or conformational heterogeneity. The researchers then
looked at the structures of various MMPs studied by NMR, fi nding similar
results — increased mobility and conformational heterogeneity in certain loop
regions.
The matrix metalloproteinase discussed here — MMP12 — consists of 159 aa
residues (numbers 105 – 263) and has a molecular weight of 16.7 kDa. The
inhibitor N - isobutyl - N - [4 - methoxyphenylsulfonyl]glycyl hydroxamic acid,
NNGH, is complexed to the protein. The X - ray crystallographic structure, at
1.34 - Å resolution, is deposited as PDB: 1RMZ, and the NMR structures are
deposited as PDB: 1YCM (20 structures of lowest energy) and 1Z3J (mini-
mized average structure). The root mean square deviation (RMSD) between
the backbone (BB) atoms of the X - ray and NMR structures differ with total
BB RMSD = 1.37 Å. The RMSD is much less in α - helix or β - sheet domains
with BB RMSD = 0.65 Å. Dissimilarity increases substantially in the loop
domains with BB RMSD = 1.64 Å. Dissimilarity in the loop regions is indicated
in Figure 3.18 using PyMOL ’ s preset “ B factor putty ” setting that shows how
much structural uncertainty is occurring in two ways: (1) by thickness of
ribbon — that is, the thicker the ribbon, the more the uncertainty or mobility
of position; and (2) by color — that is, blue - green = least uncertainty or fl exibil-
ity; yellow - orange = more fl exibility; red = most fl exibility. The MMP12 catalytic
domain ’ s secondary structure is that of a typical matrix metalloproteinase:
threeα - helices, a twisted fi ve - strand β - sheet, and eight connector loops (see
Figure 3.18 ). Loops L5 and L8 are quite long. The resulting topology reads as
follows: L1 - β 1 - L2 - α 1 - L3 - β 2 - L4 - β 3 - L5 - β 4 - L6 - β 5 - L7 - α 2 - L8 - α 3. Three calcium
ions and their binding sites lie approximately in the plane of theβ 3, β 5, and
β 4 strands of the β - sheet and are bound by residues in the β - turns of the sheet.
Two zinc(II) ions, Zn1 and Zn2, are found in the domain. Zn1, the catalytic
zinc ion, is fi ve - coordinate with ligands: N ε 2 atoms of his218, his222, and his228
and two oxygen ligands of the NNGH inhibitor (see Figure 3.18 ). Zn2 is tet-
rahedrally coordinated with ligands: N ε 2 atoms of his168, his183, and his196
and the O δ 1 atom of asp170.