June 2023 ESRFnews14
SYNCHROTRON SERIAL CRYSTALLOGRAPHY
Choosing the right sample delivery , above). It depends on what the goal is of the experiment, says de Sanctis. Users can select the most suitable one for their samples and their projects.
The methods are continually being developed, and indeed users are coming up with their own strategies. Takashi Tomizaki, a structural biologist at the Paul Scherrer Institut in Villigen, Switzerland, for instance, has co-invented a device that uses ultrasonic waves to levitate and rotate a film dispensed with tiny crystals while they are scanned at ID29 in several stages. He hopes
that it will prove to be more efficient, automated and versatile than other methods, especially for samples that are hard to purify without damage, such as membrane proteins. In a recent test, he and his colleagues collected a complete dataset of microcrystalline membrane proteins from a few microlitres of medium in less than a minute. Today it is very laborious to study crystals of this size, as we need to work on each of the crystals, but we hope this will be a thing of the past soon, he says.
Although ID29 has only been open a few months, several users have already found interesting results. In
CHOOSING THE RIGHT SAMPLE DELIVERY
Fixed target Arguably the most straightforward way
to obtain serial MX data is to sandwich
crystals, in random orientations or in a
micropattern chip, between a pair of films,
which are then rastered across the X-ray
beam. Potentially the crystals can be grown
directly on the film, thereby avoiding the
possibility of damaging them by harvesting
elsewhere. Reactions in them can be
activated by light, or the crystals can be
mixed rapidly before the data collection.
Extrusion Often nicknamed the goo method, this involves
suspending crystals in high viscosity, toothpaste-
like medium, which is extruded vertically
downwards in the path of the X-ray beam.
Although this method requires the extrusion to
be very stable, it is very efficient, especially for
membrane proteins, as the natural environment is
a greasy lipid cubic phase anyway. The method is
also well suited to light activation, because of the
ease of tuning the delay between illuminating a
sample and its descent into the X-ray beam.
Tape drive Another way to finely control the delay
between reaction activation and data
collection is the tape drive method. Here,
crystals are suspended in a thin suspension
or ink that can be written with a device
similar to a fountain pen onto a moving
tape. The crystals can be exposed to light
for activation en route, or the tape itself can
contain a compound a drug under test, say
so that the reaction is activated as soon as
the crystal-ink lands on it.
DOES RADIATION DAMAGE WARM SAMPLES?
Traditional synchrotron MX employs cryogenics to protect crystals from
the ionising effects of X-rays. Conversely, serial MX, whether at XFELs
or synchrotrons, is performed at room temperature. Does that mean it
poses a higher risk of radiation damage?
The issue is more complicated than it may at first appear. XFELs
employ the most intense X-rays, and are known to destroy samples, but
their data collection, on the order of femtoseconds, is assumed to take
place before most damage actually takes place, a paradigm known
as diffraction before destruction . By contrast, synchrotron serial MX
occupies a middle ground: the X-rays are less intense, and so crystals
do not explode, but their X-ray pulses are longer, on the order
of microseconds.
In fact, all MX, serial or otherwise, involves radiation damage to some
degree. According to de Sanctis, despite the longer pulses, no concrete
evidence of global radiation damage is observed; the question is which
specific molecular sites are most vulnerable, and whether those sites
are relevant for the interactions under study. This has been investigated
a lot at cryogenic temperatures, favouring the refinement of sample-
to-structure pipelines, he says. We need to amass the same kind of
knowledge for room-temperature studies for the beam intensity that
ID29 can deliver.
The results of radiation-damage investigations can come as a surprise.
In different studies, researchers have found that cryogenics protects
samples well from global damage, but less so from specific damage,
despite the latter being less pronounced at ambient temperature (Acta
Crystallogr. D Struct. Biol. 79pt1 78). Is this true also for the intense ID29 beam? asks de Sanctis. The only way to find out is to try. We re
in uncharted territory, and ID29 is the means to explore it.
M O
N TS
E R
R A T
C A
P E LL
A S E
S P
U N
Y
