Biophysical studies of TIMP-1.
This study had two aspects. The first was the production and purification of
TIMP-l. The second was a series of biophysical studies of TIMP-l and a TIMP-l
A monoclonal antibody affinity column was developed and used to purifY large
quantities of human TIMP-l for further experiments. Two E.coli expression systems
were studied to determine whether they would be suitable for large scale production of
recombinant protein. In the first system TIMP-I was to be secreted as a fusion protein
which could be cleaved, leaving a free N-terminus. It was discovered that it was not
possible to cleave off the fusion protein. In the second system, the protein was secreted,
without additions to the periplasm. Although a~tive protein, with the correct N-tenninus,
was obtained, the yields were too low to be of use for large scale expression.
Secondary structure analysis by CD and FTIR showed TIMP-l to be a mostly f3-
sheet protein (approaching 50%) with around 20% a-helix. A temperature study using
these techniques found that little change occurs until temperatures of over 60°C where
the protein aggregates. The small changes appear to be a general loosening of the
In analyses of the surface of TIMP-l, additional carbohydrate was identified
(other than the two N-linked chains) using Con-A probing of Western blots. TIMP-l
purified from WI-38 foetal lung fibroblast cells can be separated into two pools by
Concanavalin A-Sepharose chromatography. These two pools were found to have a
different set of pIs and a different monosaccharide composition.
The use of NMR paramagnetic probes identified a hydrophobic region exposed
on the surface of TIMP-I. This region probably includes a tyrosine residue and either a
tryptophan or phenylalanine. The presence of an exposed hydrophobic region was also
shown in binding studies using the fluorescent probe ANS. These studies identified a
single, low affinity binding site. An additional study with the N-terminal fragment of
type-I collagenase found no binding sites on the enzyme, but a change in fluorescence
occurred when TIMP-I was present.
A pt~ptide was designed based on the N-terrninal sequence of TIMP-I. High
homology, susceptibility to mutation and an interesting resemblance to the Bowman-Birk
family of inhibitors suggested that this peptide might be inhibitory. It was found to have
only a weak inhibitory activity against gelatinase. NMR studies of this peptide in water
showed a large number of conformers as a result of stabilisation of the cis isomer of its
proline residues. This preference for the cis form was retained for one proline in the
Preliminary NMR studies were also carried out which concluded that TIMP-I
should be suitable for further structural studies using isotopic labelling.