The physiological function of many proteins is associated with
conformational changes in tertiary and quaternary structure. Prominent examples
are proteins which show cooperativity (binding of a substrate enhances
additional substrate binding) such as: haemoglobin (oxygen transport in
vertebrates), haemocyanin (oxygen transport in invertebrates), glutamine
synthetase (control of nitrogen metabolism) and many other proteins. A review is
given by Perutz (1989). Relevant non-cooperative proteins include myoglobin
(oxygen transport in vertebrates) and enzymes such as lysozyme.
Time-Resolved Crystallography
Provided
suitable crystals of both states can be grown it is now possible to rapidly
obtain detailed structural information concerning the beginning and end states
of structural transitions in protein molecules using protein crystallography
with synchrotron radiation. The Laue (polychromatic) method (Amoros et
al., 1975) has previously been used to investigate intermediate states of
protein transitions. For examples see Schlicting et al. (1990) and Singer
et al. (1993). However the quality of data obtained using the Laue method
is lower than that for monochromatic methods (Hajdu et al., 1991).
Cryocrystallography
Consequently,
crystallographers are seeking alternative methods for studying protein
transitions and the associated intermediate states. Cryocrystallography is used
routinely for data collections as cooling a crystal to liquid nitrogen
temperatures greatly reduces the effects of radiation damage thus increasing the
crystal lifetime in the beam, and can in some cases increase order within a
crystal lattice and therefore the resolution limit. A review is given by Garman
& Schneider (1997). Also, as the flash-freezing of protein crystals
effectively halts any chemical or enzymatic processes within the crystal, it is
feasible to conduct time-resolved cryocrystallographic experiments on protein
transitions. Frozen crystals of possible structural intermediates can be stored
in liquid nitrogen almost indefinitely (Mancia et al., 1995), therefore
provided the reaction timescale is well defined crystals can be extracted,
frozen and stored at regular intervals once the transition has been initiated.
Data can then be collected from these crystals at any time.
Insulin - A model System
The Co(2+)
insulin hexamer is an ideal model system for time-resolved studies using
cryocrystallography. Co(2+) insulin has been shown to undergo three different,
chemically induced transitions in solution (Thomas & Wollmer, 1989) and at
least one of them, the Co(2+) to Co(3+) insulin transition, can be performed in
the crystalline state with minimal loss of diffraction (McMichael, 1997).
Experimental
Data
Collections
Crystals of Co(2+) insulin were soaked in
20 mM H(2)O(2) (Hydrogen Peroxide) for 20 minutes; after which time the crystals
were observed to have changed colour from light to dark pink and the Co(2+) to
Co(3+) transition was judged to be complete (Thomas & Wollmer, 1989).
During the soaking crystals were extracted at regular intervals, flash frozen
and stored. Data sets were then collected from each crystal in turn (as outlined
in table 1) using a Kappa arc (Mancia et al., 1995) for crystal mounting.
A typical diffraction pattern obtained is displayed in figure 1.
| Crystal
| Co(2+) Insulin
| Co(2+) Insulin
| States 1, 2, 3 and 4
|
| SRS Station
| 7.2
| 9.6
| 7.2
|
| Wavelength(Å)
| 1.488
| 0.87
| 1.488
|
| Temperature(K)
| 100
| 100
| 100
|
| Rotation Range (total)(degs)
| 90
| 110
| 90
|
| Rotation Step (degs)
| 3
| 1
| 3
|
| Exposure Time (per image)(s)
| 60
| 30
| 60
|
| Resolution Limit(Å)
| 2.4
| 1.2
| 2.4 |
Table 1 Data collection parameters for Co(2+) insulin
prior to soaking in H(2)O(2), 5, 10, 15 and 20 minutes after initiation of
transition (denoted states 1, 2, 3 and 4 respectively) collected at the SRS,
Daresbury.
![]()
Figure 1
A typical diffraction pattern obtained from a crystal of Co(2+) insulin soaked
in H(2)O(2) for 20 minutes (state 4). This image was collected on station 7.2 of
the SRS at Daresbury using a MAR image plate detector
Data Processing
The
six data sets were individually processed using programs from the CCP4 suite of
programs for protein crystallography (CCP4, 1994). A schematic overview of the
data processing method used is shown in figure 2. The data processing results
are summarised in table 2. Note that the results for Co(2+) insulin are for the
two data sets merged together.
Figure 2 Flow diagram outlining the data processing
procedure applied to each diffraction data set. The programs shown are
available through the CCP4 suite.
| Crystal
| Co(2+) Insulin
| State 1
| State 2
| State 3
| State 4
|
| Resolution (Å)
| 1.2
| 2.4
| 3.0
| 2.4
| 2.4
|
| Spacegroup
| R3
| R3
| R3
| R3
| R3
|
| Unit Cell (Å) (a=b, c)
| 80.75, 33.63
| 80.78, 33.56
| 79.71, 33.74
| 81.12, 33.97
| 80.76, 33.68
|
| Reflections (Full/Partial)
| 69321/3079
| 4471/1413
| 2421/1397
| 4886/2496
| 6371/1974
|
| Merging R-Factor (%)
| 7.5
| 6.7
| 8.1
| 6.7
| 5.1
|
| Completeness (%)
| 99.1
| 92.3
| 91.6
| 98.2
| 98.5
|
| Multiplicity
| 2.8
| 1.8
| 2.6
| 2.2
| 2.6
|
| I/{sigI} (overall)
| 6.4
| 6.5
| 5.1
| 9.7
| 9.1 |
Table 2 Summary of the data processing results
obtained for each data set. Note that the state 2 data could only be processed
to a resolution of 3.0 Å; due to the broad and diffuse nature of the spots at
higher resolutions. This together with the large decrease in unit cell
dimensions for state 2 suggested that the crystal had partially dried out during
the mounting procedure. The state 2 data was thus excluded from any subsequent
analysis
Results
Electron Density
Maps
