Two-Dimensional Crystallization of Biological Macromolecules - Molecules: Nucleation, Aggregation and Crystallization - page 100

Table 1. Factors that may Influence 2D Crystallization

The protein itself — Size of hydrophobic/hydrophilic parts, purity, concentration, glyco-
sylation
Type of detergent — Ionic, non-ionic, zwitterionic, length of hydrophobic tail, head
group size, CMC
Type of phospholipid — Fatty acid chain length, degree of saturation, phase transition
temperature
Lipid-to-protein ratio — Range 1-500 (molar ratio)
Temperature — 40-4°C and slow annealing
Salt concentration — 0-0.5 M
pH

lipid and protein in method descriptions called LPR and given as either
w/v or molar ratios may have a pronounced impact on crystal formation,
size and order. The molar ratio may vary from below ten to several hun-
dred. Large proteins with many transmembrane domains will require a
high LPR. The temperature is a variable that will influence crystallization
time. In order to speed up low CMC detergent removal, it may be advan-
tageous to dialyze at room temperature if the protein tolerates that. In fact,
many membrane proteins seem to be more stable once they are inserted
into a lipid environment than in detergent solutions. In some cases a low
temperature may be important to slow down the process and decrease the
number of nucleation sites for the 2D crystal growth.
As mentioned in the previous section, definite evaluation of crystal-
lization trials can only be made using a TEM. However, initial hints can
be obtained simply by looking at the crystallization solution. Often an
increased turbidity is a sign of formation of proteolipid complexes and
membranes with inserted protein. Indeed, many successful 2D crystal-
lization conditions result not only in useful areas but also in precipitate.
Often the crystalline areas grow as extensions from heavy aggregates.
In some cases a membrane protein may be so tightly packed in its native
membrane that it packs into a 2D crystal. A well-known example is bac-
teriorhodopsin from the purple membrane of Halobacterium halobium .
Using such arrays and electron crystallography, Unwin and Henderson
(1975) were able to obtain the first 3D structural information of a membrane