J. Appl. Cryst. (2001). 34, 365±370 F. J. Lo  pez-Jaramillo et al.  Crystallization and cryocrystallography 365 cryocrystallography papers Journal of Applied Crystallography ISSN 0021-8898 Received 25 August 2000 Accepted 26 February 2001 # 2001 International Union of Crystallography Printed in Great Britain ± all rights reserved Crystallization and cryocrystallography inside X-ray capillaries F. J. Lo  pez-Jaramillo,* J. M. Garcõ Âa-Ruiz, J. A. Gavira and F. Ota Âlora Laboratorio de Estudios Cristalogra Âficos, Instituto Andaluz de Ciencias de la Tierra, CSIC-UGRA, Facultad de Ciencias, Campus Fuentenueva, 18002 Granada, Spain. Correspondence e-mail: fjljara@ugr.es This paper presents a modi®cation of the gel acupuncture method to grow isolated crystals inside X-ray capillaries. Protein crystals are grown from 2±12 ml of gelled agarose±protein solution, cryoprotected and immobilized by the gel matrix. The same X-ray capillary that acts as a crystallization reactor is used to transport the crystals to the X-ray source and to collect data at both room temperature and 100 K, without any post-crystallization manipulation. To enhance the ¯ash-cooling stage, two additional elements are proposed for inclusion in the cryosystems currently in use: a laser pointer to illuminate the crystal to be ¯ash-cooled and a trap to divert the N 2 ¯ow and switch from room temperature to 100 K without misalignment of the crystal. With the proposed implementation, data can be collected at different temperatures from the same crystal in exactly the same orientation. This permits the study, at lattice level, of changes in unit-cell parameters, mosaic spread and crystal quality induced by cryogenic temperatures and annealing techniques. 1. Introduction Nowadays, the Human Genome project is making available thousands of sequences of potential biotechnological interest. But in most cases, the function of these genes remains unknown. One clue, however, may be provided by the three- dimensional structure of the proteins that they encode. So a new effort, known as structural genomics, is ongoing to crys- tallize and determine the three-dimensional structure of thousands of proteins, much as the genome project deter- mined gene sequences en masse . This structural information will lead to a better understanding of the molecular basis of life and diseases, and to drugs that speci®cally bind to the target protein. That outlook obviously depends on improve- ments in crystallization techniques and minimization of the post-crystallization manipulation of crystals. Hanging-drop, sitting-drop and dialysis-button techniques are among the classical approaches currently used to crystal- lize proteins (McPherson, 1999; Ducruix & Giege Â, 1999). All these traditional methods share a common feature: once protein crystals are obtained they must be either placed in a capillary that is transparent to X-rays or transferred to a cryoloop prior to collecting diffraction data. In both cases, this manipulation is problematic because it is time consuming, affects the integrity of the crystal (i.e. crystal quality and, hence, resolution limit) and provokes inevitable losses of some crystals. Our group has developed a new crystallization method based on the counter-diffusion of the precipitant agent and the protein solution under conditions starting far from equili- brium. A practical implementation is the gel acupuncture method (GAME), in which the protein solution is con®ned in a capillary punched in a gel layer and the precipitating agent is poured on the gel layer (Garcõ Âa-Ruiz, 1991; GarcõÂa-Ruiz & Moreno, 1994; Garcõ Âa-Ruiz & Moreno, 1997). The capillary acts as a long protein chamber and removes convection from the system. When the precipitating system contacts the protein solution, a wave of supersaturation is triggered (Garcõ Âa-Ruiz, Ota  lora et al., 2001). This wave travels along the protein chamber and screens different crystallization conditions of progressively lower supersaturation values as it moves further from the lower end of the capillary. Unlike traditional methods, one of the main advantages of GAME is its ability to yield crystals obtained under different nucleation and growth conditions in every experiment. This implies that a single GAME experiment is equivalent to many experiments with traditional methods. As well as crystals of higher quality, GAME is the only technique that produces crystals that ®ll the capillary diameter and do not slip, allowing data collection at room temperature from the same capillary in which the crystals were grown (Ota  lora et al., 1996; Gavira, 2000). However, when the crystals do not ®ll the capillary or do not attach to the capillary walls, they sediment and have to be mounted in a new capillary in order to collect data.