Tải bản đầy đủ - 0 (trang)


Tải bản đầy đủ - 0trang

Figure 2

Structure-based design cycle.


For most noncrystallographers, protein crystallography tends be a black

box full of jargon. Here, we give a brief overview of the technology in an

attempt to demystify some of the terms used.

A. Crystallization

Obtaining large single crystals that diffract to high resolution remains the

primary bottleneck of protein crystallography. The most widely used

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

crystallization method is the hanging-drop method of vapor diffusion

(Fig. 3), in which a drop (1 or 2 AL) of protein is mixed with an equal

volume of a precipitant on a glass coverslip and is sealed over a well

containing the same precipitant added to the protein. Many factors are

known to be important in protein crystallization: protein purity (preferably >95% pure) and concentration (typically 10 mg/mL), the nature of

precipitant [e.g., poly(ethylene glycol) or various salts] and its concentration, the nature, concentration, and pH of the buffer, the presence or

absence of additives (e.g., metal ions, reducing agents, protease inhibitors,

metal chelators, detergents) and effectors (e.g., ligands, cofactors, substrates, inhibitors), the rate of equilibrium between the protein and the

precipitant, the crystallization temperature, and so on. Since there are no

general rules to correlate all these factors to the eventual success in

obtaining crystals, protein crystallization remains a trial-and-error process

and a significant bottleneck in protein crystallography: failure rate is

typically 50% even with thousands of crystallization trials. Many methods

and techniques have been employed to enhance one’s ability to obtain

protein crystals. Molecular biology and biochemical methods have been

utilized to generate domains of large proteins that may be less flexible and

thus more amenable to crystallization. Biophysical tools such as dynamic

light scattering [12] and ultracentrifugation [13] have been used to study

protein aggregation in solution. Molecular biology has been employed to

generate mutants that do not aggregate or are more soluble. Crystallization trials using incomplete factorial designs [14] allow the screening of a

much wider range of conditions with a modest number of experiments, and

Figure 3 Protein crystallization: diagrammatic representation of the hangingdrop method of vapor diffusion.

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

thus less protein. Miniaturization and automation made possible by the

use of advanced crystallization robots may also have a great impact on the

future success of protein crystallization.

B. X-Ray Diffraction Data Acquisition

The next step is to measure x-ray diffraction data from a single crystal

(Fig. 4). Data are usually measured by means of an area detector

such as a phosphorus image plate or a charge-coupled device (CCD).

Through several steps of computational analysis, the position and

amplitude or intensity of the each diffraction spot can be obtained.

Because diffraction intensities are proportional to the volume of the

crystal and generally decrease at higher resolution, protein crystals must

be reasonably large to give strong enough diffraction signals at high

resolution. While a cube of 0.1 to 0.5 mm in each dimension is still

preferred by most crystallographers, the availability of powerful synchrotron radiation sources has made the analysis of much smaller

crystals feasible. Crystals also must be stable enough in the x-ray beam

to allow the measurement of a complete diffraction data set from a

single crystal. In this regard, flash-freezing of protein crystals under

proper conditions at cryogenic temperatures [15] has virtually eliminated

radiation decay problems.

Figure 4 Diagrammatic representation of single-crystal x-ray diffraction and

data collection.

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

C. Phasing

The ultimate goal of an x-ray diffraction experiment is to produce an

electron density map that is then used to build an atomic model of the

molecule being studied (Fig. 5). The use of single-crystal x-ray diffraction

techniques to determine the three-dimensional structure of molecules

requires the measurement of amplitudes and the calculation of phases

for each diffraction spot. Although amplitudes can be directly measured

from diffracting crystals, as noted earlier, phases are indirectly determined. The inability to directly measure phases is known as the ‘‘phase

problem’’ [16]. In practice, there are several ways to get around the phase

problem. If the protein of interest is small (f100 amino acids) and highresolution data (1.2 A˚ or better) are available, phases can be obtained

computationally by using the so-called direct method. This is basically the

same technique used to determine crystal structures of small organic

molecules. If the protein being studied is known to have a fold similar

to that of a protein with a known three-dimensional structure, one uses the

molecular replacement (MR) method, in which the known structure serves

as a model to generate approximate phases that are then refined against

the experimental data obtained from crystals of the protein under study.

Until recently, multiple isomorphous replacement (MIR) was the most

widely used method for ab initio phase determination. This technique

requires the introduction into the protein under study of atoms of high

atomic number (heavy atoms) such as mercury, platinum, and uranium,

without disrupting the protein’s three-dimensional structure or the

packing in the crystal. This is achieved by soaking crystals in a solution

Figure 5 Steps in the use of protein crystallography for structure determination.

Copyright 2004 by Marcel Dekker, Inc. All Rights Reserved.

Tài liệu bạn tìm kiếm đã sẵn sàng tải về


Tải bản đầy đủ ngay(0 tr)