Mukul Sonker, Arizona State University

Abstract

Time-Resolved Structure Elucidation Enabled by 3D-Printed Droplet Microfluidics for Reduced Sample Consumption during Serial Femtosecond Crystallography

Mukul Sonker^1,2, Diandra Doppler^1,2, Ana Egatz-Gomez^1,2, Katerina Doerner³, Romain Letrun³, Joachim Schulz³, Garrett Nelson^2,4, Mohammad Towshif Rabbani^1,2, Abhik Manna^1,2, Jorvani Cruz Villarreal^1,2,, Sahba Zaare^2,4, Konstantinos Karpos^2,4, Roberto Alvarez^2,4, Sabine Botha^2,4, Gihan Ketwala^1,2, Thomas Grant⁵,  Angel L. Pey⁶, Alice Grieco⁷, Juan Luis Pacheco⁶, Miguel Angel Ruiz-Fresneda⁶, Alexandra Tolstikova⁸, Reza Nazari^2,4, Uwe Weierstall^2,4, Adrian Mancuso³, Petra Fromme^1,2, Richard Kirian^2,4, Jose Manuel Martin Garcia⁷, and Alexandra Ros^1,2

1. School of Molecular Sciences, Arizona State University, AZ, USA, 2. Center for Applied Structural Discovery, Biodesign Institute, Arizona State University, AZ, USA 3. European XFEL, Schenefeld, Germany 4. Department of Physics, Arizona State University, AZ, USA, 5. Department of Structural Biology, Jacobs School of Medicine and Biomedical Sciences, NY, USA, 6. University of Granada, Granada, Spain, 7. Institute of Physical Chemistry Blas Cabrera, CSIC, Madrid, Spain, 8. Center for Free-Electron Laser Science, DESY, Hamburg, Germany.

Determining the atomic-level catalytic mechanisms of enzymes has long been one of the primary objectives of structural biologists. Recent advancements in time-resolved serial femtosecond crystallography (TR-SFX) have offered valuable insights into unraveling these intricate mechanisms. One enzyme of particular interest is NQO1, a flavoenzyme that is overexpressed in various tumor types, including those in the thyroid, breast, and lung. Understanding the mechanism of NQO1 is crucial for the development of quinone-based chemotherapeutics. However, a significant obstacle in studying NQO1, as well as many other proteins of interest, is the challenge of obtaining sufficient quantities of purified crystals. Traditional continuous injection methods, such as gas dynamic virtual nozzles (GDVNs), require substantial amounts of protein, severely limiting the utility of these techniques for structural investigations. To address this critical issue of sample consumption, we have developed capillary-coupled droplet injector devices (CCDIDs).¹ These innovative devices can create and manipulate protein crystal-laden droplets, which are isolated by an immiscible oil phase before being introduced into the path of an X-ray Free-Electron Laser (XFEL) using a 3D-printed GDVN. With the integration of a unique electrical triggering feedback system, these devices enabled precise control over droplet generation frequency and the efficient delivery of protein crystals to the Linear Coherent Light Source (LCLS) XFEL. This breakthrough allowed us to determine the first room-temperature structure of NQO1 at a resolution of 2.7 Å using a modular version of the droplet Injector (MDI) while reducing sample consumption by 75% compared to traditional injection methods.²

Building on the success of these novel CCDIDs, we enhanced the droplet generator by increasing the inner channel dimensions from 100 to 150 µm, facilitating droplet production at a rate of 10 Hz, compatible with the pulse structure of the European XFEL (EuXFEL). We designed two variants, CCDID-A and CCDID-B, with CCDID-B featuring a 100 µm dispersed aqueous inner channel, resulting in droplets nearly half the size of those generated by CCDID-A, further improving sample conservation at 10 Hz. By optimizing flow rate stability at rates below 5 µL/min using a syringe pump, we achieved consistent 10 Hz droplet generation for several hours within the vacuum chamber of the SPB/SFX instrument at EuXFEL. This enhanced flow stability eliminated the need for a feedback mechanism, requiring only a continuous electrical trigger to synchronize droplet alignment with the XFEL. Furthermore, when CCDID-B is combined with an integrated modular mixer (MDID-B), it enabled 10 Hz droplets generation while conducting time-resolved mixing experiments with NQO1 microcrystals and the substrate NADH. During our experiments at EuXFEL in 2022 (Beamtime P3083), utilizing MDID-A and M-DID-B over three sixteen-hour shifts, we consumed less than 0.5 mL of NQO1. Additionally, we amassed a large dataset, including 37,000 diffraction patterns from the apo NQO1 structure and over 40,000 indexable patterns from a 305 ms time point upon reaction initiation. With the latter, we successfully determined the structure of NQO1 in complex with NADH at a resolution of 2.5 Å. This groundbreaking achievement provides the first structural evidence of the negative cooperativity between the two catalytic sites of the homodimer of the NQO1 enzyme catalyzed by NADH (manuscript in preparation).

References:

1.            Sonker, M.; Doppler, D.; Egatz-Gomez, A.; Zaare, S.; Rabbani, M. T.; Manna, A.; Cruz Villarreal, J.; Nelson, G.; Ketawala, G. K.; Karpos, K.; Alvarez, R. C.; Nazari, R.; Thifault, D.; Jernigan, R.; Oberthür, D.; Han, H.; Sierra, R.; Hunter, M. S.; Batyuk, A.; Kupitz, C. J.; Sublett, R. E.; Poitevin, F.; Lisova, S.; Mariani, V.; Tolstikova, A.; Boutet, S.; Messerschmidt, M.; Meza-Aguilar, J. D.; Fromme, R.; Martin-Garcia, J. M.; Botha, S.; Fromme, P.; Grant, T. D.; Kirian, R. A.; Ros, A., Electrically stimulated droplet injector for reduced sample consumption in serial crystallography. Biophysical Reports 2022, 2 (4), 100081.

2.            Doppler, D.; Sonker, M.; Egatz-Gomez, A.; Grieco, A.; Zaare, S.; Jernigan, R.; Meza-Aguilar, J. D.; Rabbani, M. T.; Manna, A.; Alvarez, R. C.; Karpos, K.; Cruz Villarreal, J.; Nelson, G.; Yang, J.-H.; Carrion, J.; Morin, K.; Ketawala, G. K.; Pey, A. L.; Ruiz-Fresneda, M. A.; Pacheco-Garcia, J. L.; Hermoso, J. A.; Nazari, R.; Sierra, R.; Hunter, M. S.; Batyuk, A.; Kupitz, C. J.; Sublett, R. E.; Lisova, S.; Mariani, V.; Boutet, S.; Fromme, R.; Grant, T. D.; Botha, S.; Fromme, P.; Kirian, R. A.; Martin-Garcia, J. M.; Ros, A., Modular droplet injector for sample conservation providing new structural insight for the conformational heterogeneity in the disease-associated NQO1 enzyme. Lab on a Chip 2023, 23 (13), 3016-3033.