190
J. Y. CHANG
(Creighton, 1992; Weissmann & Kim, 1992a), the folding pathway of BPTI has
been established as one of the central dogma of protein folding.
For the past 5 years, our lab analyzed the disulfide folding pathways of four
single domain proteins that have the size and disulfide numbers similar to that of
BPTI. These four proteins are hirudin core domain (Hir, 49 amino acids) (Chatrenet & Chang, 1993), potato carboxypeptidase inhibitor (PCI, 39 amino acids)
(Chang et aI., 1994), human epidermal growth factor (EGF, 58 amino acids)
(Chang et aI., 1995) and tick anticoagulant peptide (TAP, 60 amino acids) (Chang,
1996). We found that these four proteins refold through a similar mechanism. But
the mechanism is very different from what has been described in the case of
BPTI. There are two major differences; (1) One is that the folding intermediates
are much more heterogeneous than in the case of BPTI (Creighton, 1990;
Weissmann & Kim, 1991). (2) Another important difference is that scrambled
3-disulfide species were found as folding intermediates in all these four proteins.
Some of our data will be presented here to illustrate these major differences. It is
important to mention that TAP is structurally homologous to BPTI in terms of
3-D conformation and disulfide pattern (Autuch et aI., 1994; Lim-Wilby et aI.,
1995). Yet, the folding mechanism of TAP is simply incompatible with those
described in the case of BPTI.
3
Experimental Procedures
Materials
All four proteins, hirudin core domain (Hir. residues 1-49), tick anticoagulant
pep tides (TAP), potato carboxypeptidase inhibitor (PCI), and human epidermal
growth factor (EGF), are recombinant proteins. Their purity was greater than
95 % as judged by HPLC, mass analysis and N-terminal sequence analysis.
Reduced glutathione (GSH), oxidized glutathione (GSSG), cysteine (Cys), cystine
(Cys-Cys), j3-mercaptoethanol were obtained from Sigma.
Folding experiments performed in the absence of redox agent
The native proteins (1.5 mg) were first reduced and denatured in 0.5 ml of TrisHCI buffer (0.5 M, pH 8.5) containing 5 M of GdmCI and 30 mM of dithiothreitol.
Reduction and denaturation was carried out at 220C for 90 mins. To initiate the
folding, the sample was passed through a PD-lO column (Pharmacia) equilibrated in 0.1 M Tris-HCI buffer (pH 8.5). Desalted and unfolded protein was
recovered in a volume of 1.1 ml, which was immediately diluted with the same
Tris-HCI buffer to a final protein concentration of 1 mg/ml, both in the absence
(control -) and presence (control +) of 0.25 mM 2-mercaptoethanol. Folding
intermediates were trapped in a time-course manner by mixing aliquots of the
sample with equal volume of 4 % trifluoroacetic acid in water or with 0.4 M
iodoacetic acid in the Tris-HCI buffer (0.5 M, pH 8.5). In the case of iodoacetate
trapping, carboxymethylation was performed at 220C for 30 mins, followed by
desalting using the PD-lO column. Trapped folding intermediates were analyzed
by HPLC.
J. Y. CHANG
(Creighton, 1992; Weissmann & Kim, 1992a), the folding pathway of BPTI has
been established as one of the central dogma of protein folding.
For the past 5 years, our lab analyzed the disulfide folding pathways of four
single domain proteins that have the size and disulfide numbers similar to that of
BPTI. These four proteins are hirudin core domain (Hir, 49 amino acids) (Chatrenet & Chang, 1993), potato carboxypeptidase inhibitor (PCI, 39 amino acids)
(Chang et aI., 1994), human epidermal growth factor (EGF, 58 amino acids)
(Chang et aI., 1995) and tick anticoagulant peptide (TAP, 60 amino acids) (Chang,
1996). We found that these four proteins refold through a similar mechanism. But
the mechanism is very different from what has been described in the case of
BPTI. There are two major differences; (1) One is that the folding intermediates
are much more heterogeneous than in the case of BPTI (Creighton, 1990;
Weissmann & Kim, 1991). (2) Another important difference is that scrambled
3-disulfide species were found as folding intermediates in all these four proteins.
Some of our data will be presented here to illustrate these major differences. It is
important to mention that TAP is structurally homologous to BPTI in terms of
3-D conformation and disulfide pattern (Autuch et aI., 1994; Lim-Wilby et aI.,
1995). Yet, the folding mechanism of TAP is simply incompatible with those
described in the case of BPTI.
3
Experimental Procedures
Materials
All four proteins, hirudin core domain (Hir. residues 1-49), tick anticoagulant
pep tides (TAP), potato carboxypeptidase inhibitor (PCI), and human epidermal
growth factor (EGF), are recombinant proteins. Their purity was greater than
95 % as judged by HPLC, mass analysis and N-terminal sequence analysis.
Reduced glutathione (GSH), oxidized glutathione (GSSG), cysteine (Cys), cystine
(Cys-Cys), j3-mercaptoethanol were obtained from Sigma.
Folding experiments performed in the absence of redox agent
The native proteins (1.5 mg) were first reduced and denatured in 0.5 ml of TrisHCI buffer (0.5 M, pH 8.5) containing 5 M of GdmCI and 30 mM of dithiothreitol.
Reduction and denaturation was carried out at 220C for 90 mins. To initiate the
folding, the sample was passed through a PD-lO column (Pharmacia) equilibrated in 0.1 M Tris-HCI buffer (pH 8.5). Desalted and unfolded protein was
recovered in a volume of 1.1 ml, which was immediately diluted with the same
Tris-HCI buffer to a final protein concentration of 1 mg/ml, both in the absence
(control -) and presence (control +) of 0.25 mM 2-mercaptoethanol. Folding
intermediates were trapped in a time-course manner by mixing aliquots of the
sample with equal volume of 4 % trifluoroacetic acid in water or with 0.4 M
iodoacetic acid in the Tris-HCI buffer (0.5 M, pH 8.5). In the case of iodoacetate
trapping, carboxymethylation was performed at 220C for 30 mins, followed by
desalting using the PD-lO column. Trapped folding intermediates were analyzed
by HPLC.