Application of a high-throughput microcrystal delivery system to serial femtosecond crystallography

A new one-dimensional fixed-target sample delivery method combined with a real-time visual servo scan system has been developed for serial femtosecond crystallography. The developed high-throughput microcrystal delivery system shows significant improvements for microcrystal delivery efficiency at an optimal crystal stability and significantly reduced sample consumption to reveal protein structures utilizing X-ray free-electron laser pulses.


S1. HT-MCD chamber configuration
The chamber body is manufactured from aluminium plates and is 480 × 320 × 510 mm 3 . The chamber contains an L-shaped major plate that supports the high-speed delivery module and a 45° tilted mirror. The mirror position is adjusted by 2D motorized linear stages attached to the major plate (KGB40A-B1-S-50S, travel range: 50 mm; repeatability: 20 μm; ERAEtech, Republic of Korea). An aluminium plate with a 1-mm pinhole aperture is installed at the entrance of the chamber to reduce parasitic noise from upstream around the XFEL beam, and the pinhole plate position is adjusted by the same type of 2D motorized linear stage used for mirror manipulation. To prevent damage to the detector by ultra-strong XFEL pulses, a beam stopper is installed at the back wall of the chamber.
Two commercial LED light sources (LEDMM2, Misumi, Japan) are mounted at the front wall of the chamber for the vision acquisition module, and another LED (YL-6011, SMG, China) is mounted to an arm connected to a rotational manipulator (AX-12a, ROBOTIS, Republic of Korea) at the back wall of the chamber for sample coordinate setup and alignment.

S2. Sample delivery module controller structure
To track the specific region of the MCC tube, the controller is designed as a double-loop structure: the inner loop is for position regulation (velocity-level PID controller), and the outer loop is for tube tracking (position-level P controller) (Supporting Fig. S5). The inner-loop controller is a low-level controller that controls the velocity of the piezo actuators to maintain a reference velocity ( ref ); it calculates the control input of the piezo actuator with the position feedback from the linear encoder.
The outer-loop controller calculates the reference velocity of each actuator with visual feedback.
Using the calculated tube position information from the processed image, the reference velocity is defined as where ℎ is the lateral distance between the XFEL beam position and the centroid of the detected tube ( in Fig. 2b), and p is the control gain. The lateral offset Δℎ is used to align the XFEL beam with the lower part of the MCC tube so that the XFEL pulses can interact with the crystals settled in the lower part of the tube. In our experiment, p is set to 100, and Δℎ is set to −30 μm. The desired velocity ( des ) of the MCC chip is 1.5 mm/s for a 50-μm interval at a 30 Hz repetition rate in the longitudinal direction.

S3. Settling effect of the crystals in the tube during data collection
We have calculated the path lengths in the polyimide tube with XFEL beam (~ 8 μm) as shown in the following figure. When the XFEL beam passes through just the bottom of the inner space, the total path lengths are between 40 and 76 μm in the figure. Based on the CXRO X-ray database, the expected X-ray transmission values according to those polyimide thicknesses, are very similar with that at the center region within about 2 %. The background signal from polyimide tube will not give a significant effect on the diffraction signals. The detailed description is shown as following figure.

Table S1
Overall statistics of four datasets using the MCC chip as a function of resolution range.
Four tables are presented here to show the overall data quality for each data collection.

Figure S6
Schematic drawing of ideal crystal densities. The crystal density ranges from 2.5 × 10 6 crystals/ml (minimum density) to 8.4 × 10 6 crystals/ml (maximum density) in the tube with minimized crystal consumption. When the crystals are located at positions that match the scanning step for XFEL pulse illumination, data collection can be performed with high efficiency. We have prepared a sample with a slightly higher crystal density of 5 × 10 7 crystals/ml to achieve more efficient data collection, considering the occurrence of crystal stacking inside the tube.

Figure S7
Two X-ray diffraction images showing diffraction resolution and background scattering.
(a) X-ray diffraction quality of the proteinase K microcrystals from T. album. Resolution rings at 7.61, 3.92, 2.74, 2.17, and 1.85 Å are annotated. (b) X-ray diffraction image acquired without a sample.

Figure S8
Enzymatic assay result of Proteinase K. The enzyme assay was performed using the protease activity assay kit (AB112153, Abcam). We obtained the activity profile using 24 nM of proteinase K at Ex/Em = 540/590 nm. The enzymatic reaction was performed at 22˚C and monitored by Tecan Infinite fluorimeter F200.   using a Pilatus 6M CCD detector.