Content uploaded by Peter Docker
Author content
All content in this area was uploaded by Peter Docker on Jul 29, 2015
Content may be subject to copyright.
The Development of an Automated Nano Sampling Handling System for Nanometre
Protein Crystallography Experiments
P. Docker+, D. Axford+, M. Prince*, B. Cordovez~, J. Kay+, D. Stuart+, G. Evans+
+ Diamond light Source Harwell OX110DE
*Aston University Birmingham B4 7ET
~ Optek Systems Philadelphia, PA, 19104
ABSTRACT
As the world’s synchrotrons and X-FELs endeavour to
meet the need to analyse ever-smaller protein crystals, there
grows a requirement for a new technique to present nano-
dimensional samples to the beam for X-ray diffraction
experiments.The work presented here details developmental
work to reconfigure the nano tweezer technology developed
by Optofluidics (PA, USA) for the trapping of nano
dimensional protein crystals for X-ray crystallography
experiments. The system in its standard configuration is
used to trap nano particles for optical microscopy. It uses
silicon nitride laser waveguides that bridge a micro fluidic
channel. These waveguides contain 180 nm apertures of
enabling the system to use biologically compatible 1.6
micron wavelength laser light to trap nano dimensional
biological samples. Using conventional laser tweezers, the
wavelength required to trap such nano dimensional samples
would destroy them. The system in its optical configuration
has trapped protein molecules as small as 10 nanometres.
Keywords: microfluidics, protein crystallography, X rays,
optical nano traps.
1 INTRODUCTION
Typically, workers in the synchrotron community look
to employ a macro, top-down approach using robots to
automate human protocols [1]. This is a methodology often
seen in the micro and nano engineering community at large.
Currently, as the handling of nano crystals is beyond top
down capabilities, X-FEL facilities are currently
employing an injector approach. This requires micro litre
quantities of reagant containing nanocrystals to be injected
in front of the X-ray beam and the diffraction pattern is
imaged by the detector. Such systems have a reported ‘hit’
rate of less that 10 % [2].
This paper details optical nano tweezers which are very
much a bottom-up approach that facilitates self-assembly,
alignment and subsequently automation. This also offers
the potential to inform the user that a crystal has been
trapped prior to interrogation. This techneque will also
present a static array of crystals to be interrogated by the X-
ray beam. This technique is not to be confused with
traditional optical tweezers which can be used to
manipulate a single sample. Although a very useful
technique it can only be used with >1 µm sized samples as
the wavelength of laser light used to manipulate the sample
must be of the same order of magnitude. To achieve the
trapping of nano crystals using this technique would require
a wavelength of laser light with too much energy thereby
destroying any sample it traps.
These nanotweezers have been developed by
Optofluidics (PA, USA). By using optical wave guides
1064 nm biologically friendly laser light can be used to trap
samples sizes down to 10 nm at predetermined sites micro
machined into the waveguides. These can be prealigned to
the X-ray source and no further alignment would be
required.
2 THE TECHNOLOGY
This potential solution is offered by Optofluidics, based
in the Philadelphia, USA. They have developed technology
which employs a microfluidic cell that uses silicon nitride
waveguides to overcome the free space limitations of
traditional optical traps to capture sub-micron particles with
laser light using a wavelength that is compatible with
biological samples (1064 nm) whilst maintaining protein
sample integrity. Briefly, light is tightly confined at the
near-surface of the silicon nitride waveguides, and the
technology uses the evanescent wave (light outside of the
waveguides) to generate strong optical gradients which are
necessary to capture such small particles. If traditional
optical trapping technology were to be implemented, such
technologies would either be unable to optically capture the
nano crystals (because of the light diffraction limit
limitations of free space traps) or they would introduce a
significant heating effect tothe surrounding volume, likely
destroying the analyte. The flow cell allows accurate
fluidic delivery (bringing the particles to the trapping
locations via pressure driven flow), and once the particles
are in the vicinity of the silicon nitride traps, the light is
guided through the traps which results in the particles being
captured (figure 1).
Figure 1 Trapping of nano particles
The technology has further advantages as the
microfluidic flow-cell can have several waveguides each
with an array of trapping sites allowing the chip to form a
sample grid. This would allow for a chip to be aligned to
the beam and present a ‘grid’ of trapped protein crystals
that would also remain static during interogation. In
operation, a slurry of crystals are flowed through the flow
cell while the 1064 nm laser light is transmitted through the
waveguides. The crystals flowing over the waveguides are
then captured at the trapping sites. Given that the traps are
very strong, the waveguide can continuously trap crystals
while the flow is maintained, but both the flow and the
trapping light can be modulated on command. If the laser
light is switched off the examined crystals are released and
the process can then be repeated.
In addition the system will allow stationary crystals to
be exposed to different types of reagents during the
measurement process.
2 THE CHALLENGE
This work sets out to develop the system to meet the
new challenges of X-ray protein crystallography
experiments to be carried out on both Synchrotrons and the
next generation of X- ray light sources, X-FEL. Often the
nano crystals are constituted of up to 98 % water and
suspended in an aqueous solution further complicating the
selection and trapping of samples for analysis. An
additional challenge is a micro engineering/fabrication one.
The standard chip architecture is designed for optical
microscopy where illumination and analysis
measurements are carried out from one side and the system
in its standard form takes place in the horizontal plane. The
former requires re-engineering to the macro engineering of
the chip holder and the latter the more complex issue of
redesign the microfabrication route to develop chips that
will cause minimal attenuation of the X-ray beam.
3 MEETING THE CHALLENGE
3.1 Trapping Experiments
The experiments required to determine whether this
technique is suitable to trapping protein cryals was carried
out using the standard Optofluidic chip. Figure 2 shows the
microfluidic chip in its holder and micrographs of the chip
architecture
Figure 2 Standard chip architecture
Trials were conducted with 400 nm beads to clarify
whether submicron specimens could be consistently
trapped, and then further studies were undertaken with one
µm protein crystals to clarify whether such structures could
be trapped and would not perish due to the laser light being
used to trap them. Figure 3 shows three 400 nm nanobeads
trapped on chip. The image is taken from live footage taken
using a standard upright microscope stage and a stemmer
IDE UI LE Camera.
Figure 3 Trapped 400 nm beads
3.2 On Chip Modifications
Figure 4 shows how the system was configured to trap
1 µm Protein crystals. This size crystal was chosen for
ease of obtaining confirmation they had been trapped
without attaching a flourescent tag. Three trapped crystals
can be seen in Figure 4.
Figure 4 1 µm protein crystals trapped on the wave guide.
For clarity figure 5 shows a micrograph of the
waveguide when it was being processed. The dark circles
are the ‘holes’ in the wave guide where the crystals are
trapped.
Figure 5 SEM micrograph of the waveguide. (Please note it
is in the vertical orientation)
So from both a biological and size perspective the system
is more than capable of meeting the demands of its new
challenge. A standard chip has 320 µm of silicon below the
waveguide which from standard calculations [3] would
attenuate 80 % of the beam only allowing 20% to be used
to obtain diffraction patterns from the sample. Predictive
software was used to ensure that back-etching of this
additional silicon would reduce this attenuation to an
acceptable level. The new proposed chip architecture can be
seen in figure 6.
Figure 6 Proposed new chip architecture
A first run of processing the modified chips utilized wet
etching as an additional stage to remove the bulk silicon
below the waveguide. This approach however yielded only
a small number of functional chips. The chips did however
facilitate X ray tests to see if the new design had adequately
reduced attenuation. The modified chip can be seen in the
figure 7.
Figure 7 Modified chip
The chip was tested using 12.8 keV X-rays and proved the
chip facilitated transmission of X-rays >99.9 %. It also
caused negligible scatter comparable to that the air presents
between the sample and the detector.
Figure 8 Scatter comparable with air between chip and
detector
3.3 Hardware Modifications
The majority of the hardware is already fully compatible for
X-ray diffraction experiments, the only component
requiring modification is the chip holder. This component is
responsible for making the fluid tight seal between the chip
and the sample see figure 9.
Figure 9 Standard chip holder
This chip holder requires a hole to be machined through
the aluminuim between the two port seals as the X ray
diffraction experiments are carried out in transmission. The
beam will pass through the the hole, difract through the
trapped crystals and onto the diffraction detector.
In service, laser light will pass throught the wave guides
and a slurry of crystal will pass over the waveguides and
will be attracted to the array of traps. The chip can then be
indexed and patterns may be determined from each crystal
in turn. The laser will then be turned off and the spent
crystals sent to waste and the fresh analyte containing the
crystals will re-enter the chip and the process repeated.
4 FUTURE CHALLENGES
The key driving challenge is to develop a completely
dry etch process route in to improve yield and ensure it
integrates with the modified chip holder. The last process
run used wet etching to hollow the chip below the
waveguides,moving forwards the next batch will employ
deep reactive ion etching (DRIE) to remove the excess
silicon. This approach is more expensive but more
controlled. There is also no requirement to freeze dry the
devices to negate issues from surface tensions caused by
using wet etching techniques
A more ‘pure science’ challenge will investigate using
manipulation of the laser light to orientate the crystals that
are trapped in different or known orientations. To get the
full structure of a protein the sample will need to be imaged
in different orientations to obtain its full structure. By
polarizing the light at the laser traps it is hypothesized that
the crystals could be trapped and rotated with careful
control of the laser light.
5 SUMMARY AND CONCLUSIONS
This paper details experiments that have been carried
out which prove that ‘typical’ protein crystals can be
trapped, with potential for trapping sub-micron crystals. A
modified chip architecture has been developed specifically
for use with a synchrotron X-ray beam. The chip is
manufactured using standard semi-conductor technology
which has been revised to produce a chip with a
transmission of 99.98 % of the beam at energies between
6.4 - 20.0 KeV whilst maintaining the structural integrity of
the flow cell part of the device. Standard chips were only
capable of transmitting 20%.
This technology when fully developed for X-ray protein
Crystalography experiments will truly be a game changing
methodology for the nano protein crystallography
community. Allowing these crystals to be held in position
whilst they are imaged and allow for automation by default.
When Synchrotron and X-FEL beam time can take users
two years to obtain an 8 hour slot, the ability to obtain as
much high-quality data as possible is of paramount
importance.
REFERENCES
[1] Cipriani, F. et al. CrystalDirect: a new method for
automated crystal harvesting based on laser-
induced photoablation of thin films. Acta
crystallographica. Section D, Biological
crystallography 68, 1393-1399,
doi:10.1107/s0907444912031459 (2012).
[2] Spence, J. C. H., Weierstall, U. & Chapman, H. N
Xray lasers for structural and dynamic biology.
Reports on Progress in Physics 75,
doi:10.1088/0034-4885/75/10/102601 (2012)
Dr Peter Docker The Diamond light Source
Peter.Docker@Diamond.ac.uk
