Sample holder and sample cooling system for cryo-electron microscopy

Abstract

A sample holder design and sample cooling device for single-particle cryo-electron microscopy simplifies sample preparation and handling, significantly reduces errors, improves result reproducibility, and significantly reduces the total cost. The system consists of a grid-based sample holder system, a grid handling tool, a grid blotting tool, an immersion cooling system, and a jet cooling system.

Description

[0001] Citation of Related Applications

[0002] This application claims priority to U.S. Provisional Application No. 62 / 910,511, filed October 4, 2019, the entire content of which is incorporated herein by reference.

[0003] Acknowledgment of Government Support

[0004] This work was supported by the U.S. National Institutes of Health, General Medical Sciences, award number R43 GM137720-01. Technical Field

[0005] The present invention belongs to the field of biotechnology. More specifically, the design of sample supports and sample cooling systems for cryo-electron microscopy. Background Art

[0006] Single-particle cryo-electron microscopy (cryo-EM) is a powerful method for obtaining near-atomic resolution structures of large biomolecular complexes, membrane proteins, and other major scientific, pharmaceutical, and biotechnological targets. The development of high-efficiency, high-frame-rate direct electron detectors, algorithms for correcting the motion induced by the acquired electron beam in "movies", and computational tools for classifying and averaging 10 5 -10 6 molecular images has significantly improved the achievable resolution and throughput. The substantial investment in new cryo-EM facilities and the development of user-friendly software have greatly expanded access, especially for non-experts. Unlike X-ray crystallography, cryo-EM requires only a small amount of biomolecular sample dispersed in solution. It allows for structural studies of systems that are difficult to crystallize and is becoming the method of choice for initial attempts at structure determination.

[0007] As with X-ray cryo-crystallography, the key challenges in single-particle cryo-EM are related to sample preparation and handling. The basic principles and methods currently in use were developed in the 1980s, and recent developments in sample preparation techniques are firmly rooted in the ideas and methods developed at that time. A biological molecule sample must be expressed, isolated, and purified. A cryoprotectant-free buffer containing approximately 0.3 mg / mL of the biological molecule of interest is dispensed onto a glow-discharge cleaned and charged carbon or metal (usually gold) foil that is 10 - 50 nm thick and supported by a metal (usually copper or gold) grid (grip) with 200 - 400 mesh, 10 - 25 microns thick, and 3 mm in diameter. Excess sample is removed by blotting and evaporation to a target thickness that is several times the diameter of the biological molecule or approximately 10 - 50 nm to maximize image quality while limiting the proportion of biological molecules that are preferentially oriented by interaction with the interface. To vitrify the buffer for optimal imaging, the foil + grid containing the sample is plunged into liquid ethane at T ~ 90 K (generated by cooling gas in a liquid nitrogen-cooled cup) at a speed of 1 - 2 m / s. The sample is transferred from ethane to liquid nitrogen (LN2), loaded into a grid box, transferred to other containers, and then transferred to a storage dewar. The sample is removed from the storage dewar and the grid box, loaded onto a cold microscope stage or "clamped" and loaded into a cold sample holder; then the stage or holder is loaded into the microscope.

[0008] These complex procedures are fraught with difficulties. In each of the many manual handling steps, the grids, especially the foils, are often bent, torn, or otherwise damaged. Sample dispensing, blotting, and evaporation are imprecise. The film thickness of the final sample is poorly controlled. Biological molecules accumulate at interfaces where they may have a preferential orientation or undergo denaturation. Plunge-cooling the sample often results in the formation of distinct crystalline ice and is contaminated by ice formed on ethane, nitrogen, and other cold surfaces exposed to moisture. Instruments widely used for sample blotting and plunge-cooling, especially FEI's Vitrobot™, Gatan's Cryoplunge™, and Leica's EM GP™, do not adequately address these challenges. A new generation of instruments, such as TTP LabTech's Chameleon™ and VitroJet™, automate the sample preparation process, combining sample dispensing, blotting / wicking, plunge-cooling, and transfer to a grid box. However, these instruments are both complex and expensive - approximately $500,000 - and require long-term service contracts that are beyond the reach of most research groups. More critically, it is not clear whether they address the key sample preparation challenges in a robust and flexible manner. Summary of the Invention

[0009] The present invention relates to the design, function, and use of a sample holder and a sample cooling device for cryo-electron microscopy.

[0010] This application describes a number of innovations in sample holder design and sample cooling devices. These innovations will simplify sample preparation and handling, significantly reduce errors, improve result reproducibility, and significantly reduce overall costs.

[0011] The sample holder for cryo-EM consists of a metal grid covered on its top surface with a much thinner sample holder film / foil of carbon or metal. The grid has a mesh pattern of through-holes and a solid, non-porous outer edge region. The foil has a much smaller through-hole pattern. These are all handled with metal tweezers with pointed tips and are often damaged during handling.

[0012] This application first describes a series of innovations to the grid, foil, grid + foil assembly, and tools for handling the grid, which together form a cryo-EM sample holding and handling system that will improve functionality and useful throughput.

[0013] According to one embodiment, the grid under the sample holder film has a substantial area on its solid or near-solid side, preferably at least 10% and less than 50% of the grid area, to provide an area where the grid can be safely clamped and handled without damaging the grid or foil. According to one embodiment, the grid has one or more indentations at its outer edge, which can be used to precisely orient the grid relative to a matching clamping tool.

[0014] According to one embodiment, the grid has distinct solid regions or other structures or markings that are located at a smaller radius from the grid center than the solid outer edge region of the grid and at a smaller radius than the inner radius of any grid "clip" used to simplify automated grid handling, and within the grid region that can be imaged in the electron microscope, which allows determination of its orientation about the central axis during plunge cooling and subsequent measurements in the electron microscope. Such markings are preferably easily visible to the naked eye.

[0015] According to one embodiment, the grid has a surface marking or an array of through-holes in the grid region away from its edge and in any region covered by a "clip" that can be imaged in the electron microscope, which forms a pattern or code that can be used to uniquely identify each individual grid optically or using the electron microscope.

[0016] According to one embodiment, the grid bars below the sample support film / foil have a reduced width in the grid plane within a selected area of the grid, the selected area including less than 25% or less than 10% of the total grid area, and individually including less than 5% and preferably less than 2% of the grid area. The width of the grid bars is preferably reduced from a standard 25 μm or more on a 300 mesh grid to 1 - 10 μm. The grid area with reduced width preferably elongates in the direction of sample movement during the loading and cooling process, as indicated by the clamping area or other features on the grid, which remain visible when the grid is clamped.

[0017] According to one embodiment, the grid bars below the sample support film / foil have a reduced thickness perpendicular to the grid plane in the selected area, preferably including less than 25% or less than 10% of the total grid area. The thickness of the grid bars is preferably reduced from a standard 10 μm to 1 to 5 μm, or from a standard 25 μm to 1 to 10 μm. The grid area with reduced thickness preferably elongates in the direction of sample movement during the loading and cooling period.

[0018] According to one embodiment, the width and / or thickness of the grid bars can only be reduced in a small area equivalent to one grid square or cell to form a weak link where grid deformation due to stress generated during the cooling process is concentrated, and the deformation allows significant movement between the grid bars to release the stress in the sample support film.

[0019] According to one embodiment, the grid has a pattern of apertures and grid bars, and the central region of the grid (preferably less than 25% of the total grid area) has grid bars that have a smaller width, a smaller thickness, and / or a larger mesh size and a smaller solid area fraction compared to the outer part of the grid.

[0020] According to one embodiment, the grid can have square and hexagonal mesh regions and can have regions with different mesh sizes and open area fractions.

[0021] According to one embodiment, the grid is made of two independent, substantially circular parts that are joined together after formation. According to one embodiment, one part is thicker and can have one or more holes / apertures, each of which encloses an area much larger than the area of a single grid square (or hexagon). According to one embodiment, the thinner part has a grid pattern or mesh that covers the larger holes / apertures in the thicker part.

[0022] According to one embodiment, the grid is made of a conductive material such as molybdenum, titanium, tungsten, or tantalum and / or their alloys, whose average coefficient of thermal expansion is between 77K and 300K, and the sample support foil is made of materials such as gold, copper, or nickel and / or their alloys, which can undergo significantly greater thermal contraction.

[0023] According to one embodiment, the size and shape of the sample support film or foil are such that it substantially does not overlap with the solid clamping portions of embodiments of the grid having solid clamping portions, such that the grid can be clamped on the solid regions without contacting or damaging the foil.

[0024] According to one embodiment, the foil covering the grid preferably has regions of at least two different thicknesses, one or more of these regions having an array of through-holes.

[0025] According to one embodiment, the sample support foil is made of a metal alloy with low thermal conductivity but high electrical conductivity, preferably having a thickness of 10 to 100 nm, preferably 50 nm, having holes with a size of 0.1 to 5 μm, preferably 1 μm, and is placed on a cryo-electron microscopy grid, preferably made of gold, copper, titanium, nickel, tungsten, or molybdenum.

[0026] According to one embodiment, the low thermal conductivity, high electrical conductivity alloy is an alloy of chromium and gold, having a chromium content between 0.1 wt% and 10 wt%, preferably 1 wt%.

[0027] According to one embodiment, the metal or carbon sample support foil is continuous and has no holes in the regions forming a pattern matching the grid bar pattern of the support, and the support foil has an array of holes in each open region away from the grid bars. According to one embodiment, the center of the holes is separated from the grid bar positions by at least 1 / 8 of the opening width between the grid bars.

[0028] According to one embodiment, the hole-free region of the sample support foil to be registered with the grid bars can be limited to a selected region of the foil, so that the grid bars can be seen elsewhere under the foil, thus facilitating the alignment of the foil strips and the grid bars.

[0029] According to one embodiment, the metal grid and the metal foil are fabricated together in a single manufacturing process, causing them to be automatically aligned, rather than in two separate processes that require an alignment step. According to one embodiment, the process includes depositing a release layer on a substrate; depositing a foil layer; depositing a photoresist; exposing the hole pattern of the foil in the photoresist; etching the hole pattern in the foil; removing the photoresist; depositing a second photoresist; exposing and developing the grid pattern in the photoresist; electroforming the grid onto the foil through the openings in the photoresist; removing the photoresist; and releasing the completed grid + foil from the substrate.

[0030] The present invention also includes the design of a tool / tweezers for holding a cryo-EM grid.

[0031] According to one embodiment, the tool / tweezers has a sample / grid clamping end that has a substantially flat region, the width of which is less than but equivalent to the 3.05 mm width of the grid.

[0032] According to one embodiment, the gripping end of the tool / tweezers is shaped to contact only the flat gripping area of the grid according to one embodiment, and is preferably configured to prevent the tweezers from contacting the foil-covered portion of the grid.

[0033] According to one embodiment, the gripping end of the tool / tweezers has a profile or protrusion that matches the outer edge of the grid, including any notches in the outer edge, such that the gripping end slides a fixed distance along the edge of the clip before the grid etching contacts the profile or protrusion, enabling precise positioning of the grid in the tweezers. According to one embodiment, the grid may have larger through-holes in the gripping area that are aligned with the posts in the gripping tool.

[0034] According to one embodiment, the gripping end of the tool / tweezers is made of a polymer. According to one embodiment, the tool / tweezers body is made of metal or polymer and has a spring action to keep them open or closed until squeezed.

[0035] The invention also includes a device for cooling samples for cryo-electron microscopy that does not use ethane or any other flammable liquid refrigerant, but only uses liquid nitrogen for cooling and storage.

[0036] According to one embodiment, the device includes

[0037] A vertical linear sample translation stage that can insert the sample into liquid nitrogen at a speed of 1 - 10 m / s;

[0038] A gripping mechanism attached to the stage that can grip a cryo-EM grid and keep its plane precisely perpendicular to the liquid nitrogen surface;

[0039] A device for removing all cold air above the liquid nitrogen surface and ensuring a sudden (on the order of 100 μm or less) transition between gas at T > 273 K and liquid nitrogen at 63 K < T < 77 K; such a device may include suction / vacuum to remove the cold air and a dry ambient temperature air stream (N2);

[0040] A Dewar flask or insulated container filled with liquid nitrogen;

[0041] A container present in the liquid nitrogen in which the cryo-cooled cryo-EM sample can be deposited.

[0042] According to one embodiment, the gripping mechanism automatically releases the cryo-EM grid into the storage container after the grid has been inserted for cooling.

[0043] According to one embodiment, a device is provided for keeping the liquid nitrogen level in the Dewar flask almost constant.

[0044] According to one embodiment, the sample resides before being introduced into a humidification chamber where the humidity can be controlled up to 100% to prevent or control dehydration of the sample.

[0045] According to one embodiment, a device for automatically or manually blotting excess liquid from a grid is provided.

[0046] According to one embodiment, a Dewar flask or insulated container containing liquid nitrogen is replaced by a first container containing liquid nitrogen into which the sample is introduced. The first container is in good thermal contact with a second container containing liquid nitrogen, the temperature of which has been lowered below its boiling point temperature but is close to and not lower than its freezing point temperature, so that the temperature of the liquid nitrogen in the first container is lowered below its boiling point temperature.

[0047] According to one embodiment, the first container is mostly placed inside the second container to maximize the thermal contact between the liquid nitrogen in the first and second containers.

[0048] According to one embodiment, the temperature of the liquid nitrogen in the second container is lowered below its boiling point by evaporative cooling.

[0049] According to one embodiment, the second container can be largely sealed, except for being connected to a port of a vacuum pump, to reduce the gas pressure inside the container and evaporate the coolant liquid nitrogen.

[0050] According to one embodiment, a mechanical stage in the main liquid nitrogen chamber receives standard cryo-EM sample holder storage boxes / cartridges and automatically positions them to be aligned with the sample introduction path defined by a vertical translation stage, such that each cold sample is placed in a separate compartment of each sample holder through a combination of only the vertical movement of the vertical translation stage and only the horizontal movement of the mechanical stage on which the sample holder storage box is placed.

[0051] The present invention also includes a device for removing excess sample solution from the surfaces of the grid and foil before introduction of cooling.

[0052] According to one embodiment, an absorbent material such as filter paper is cut to substantially match the size and area of the cryo-EM grid.

[0053] According to one embodiment, the absorbent disk is directly pressed into contact with the surface of the grid.

[0054] According to one embodiment, the absorbent disk is embossed or patterned to create raised areas such that when the absorbent disk is pressed into contact with the grid, only the raised areas contact the grid.

[0055] According to one embodiment, the raised areas on the absorbent disk cover only a small portion of the total area of the grid, preferably less than 25% or 10%, such that most of the grid area is not contacted by the absorbent disk.

[0056] According to one embodiment, the sample support foil on the grid has a pattern of regions with and without holes that matches the embossed pattern on the absorbent material such that the foil regions contacted by the absorbent material do not have holes.

[0057] The invention also includes a device for cooling cryo-EM samples using cryogenic liquids such as liquid ethane or liquid nitrogen, where a jet or jets of cryogenic liquid are directed at the sample, and a device is provided to prevent the sample from being pre-cooled by cold gas ahead of the cryogenic liquid from its injection tube or nozzle.

[0058] According to one embodiment, such a device includes

[0059] a sample holder that holds at least one cryo-EM grid;

[0060] a source of liquid refrigerant in an adiabatic container;

[0061] one or more tubes or conduits from the adiabatic container through which the liquid refrigerant flows and from which the liquid refrigerant emerges and flows as a jet towards the sample;

[0062] a device for pushing the liquid refrigerant out of the container and through the tube towards the sample; and

[0063] a device for removing the cold gas in front of the cryogenic liquid jet so that sample cooling only begins when the liquid jet impinges on the sample.

[0064] According to one embodiment, the device for pushing the liquid refrigerant out of the container can be a plunger, a pump, or a pressurized gas present in the container.

[0065] According to one embodiment, a valve prevents the liquid refrigerant from flowing out of the tube or conduit into the sample.

[0066] According to one embodiment, the sample can be stationary relative to the axis or axes of the conduit portion that produces the cryogenic liquid jet when the liquid refrigerant flows towards the sample.

[0067] According to one embodiment, the sample can translate relative to the axis of the conduit portion that produces the cryogenic liquid jet when the liquid refrigerant flows towards the sample.

[0068] According to one embodiment, a mechanical shutter or blade initially blocks the liquid refrigerant flow and then quickly moves away such that the liquid refrigerant flow impinges on the sample when the liquid refrigerant flow begins to flow from the conduit towards the sample.

[0069] According to one embodiment, the shutter is made of heat-insulating material.

[0070] According to one embodiment, a high-speed air stream or gas “blade” having a temperature between 0° C. and the ambient temperature is directed through the outlet of a tube that generates a liquid refrigerant stream such that the cold air emerging from the tube in front of the liquid deflects the sample.

[0071] According to one embodiment, a device for transferring a cooled sample into a liquid nitrogen container is provided.

[0072] According to one embodiment, the sample is transferred into a grid box or other grid holder instead of being stored in liquid nitrogen or cold nitrogen gas within a thermally insulated container. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] FIG. 1(A) is a top view of a prior art cryo-EM grid covered with a sample support film or foil and a sample.

[0074] FIG. 1(B) is a top view of a prior art cryo-EM grid covered with a sample support film or foil and a sample.

[0075] FIG. 1(C) is a top view of a prior art cryo-EM grid covered with a sample support film or foil and a sample.

[0076] FIG. 1(D) is a side view of a prior art cryo-EM grid covered with a sample support film or foil and a sample.

[0077] FIG. 1(E) is a side view of a prior art cryo-EM grid covered with a sample support film or foil and a sample.

[0078] FIG. 1(F) is a side view of a prior art cryo-EM grid covered with a sample support film or foil and a sample.

[0079] FIG. 1(G) is a perspective view of a prior art cryo-EM grid that has been “clipped” for automated processing.

[0080] FIG. 2(A) is a perspective view of a unit of a 300 mesh cryo-EM grid covered with a thin sample support foil. The arrow indicates the flow of liquid refrigerant.

[0081] FIG. 2(B) is a perspective view of a liquid sample containing biomolecules flowing through holes in the foil and wetting the grid bars.

[0082] FIG. 3(A) is a perspective view of a grid design having solid areas and notches for clamping and orientation marking, and a bar code or marking for grid identification and tracking.

[0083] FIG. 3(B) is a perspective view of a grid design having solid areas and notches for clamping and orientation marking, and a bar code or marking for grid identification and tracking.

[0084] Figure 4(A) is a perspective view of a Cryo-EM grid according to an embodiment, the grid having hexagonal mesh holes, a set of solid non-porous regions, and asymmetric features or markings.

[0085] Figure 4(B) is a perspective view of the grid, which includes an orientation notch and a clamping region.

[0086] Figure 5(A) is a perspective view of a prior art cryo-EM grid and a close-up of its grid bar pattern.

[0087] Figure 5(B) is a perspective view of a cryo-EM grid according to an embodiment, having variable grid bar widths and geometries in selected regions. The arrow indicates the direction of grid movement relative to the liquid cryogen during plunge cooling.

[0088] Figure 5(C) is an enlarged perspective view of the cryo-EM grid in Figure 5(C) according to an embodiment, showing a selected region with reduced width and removed grid bars.

[0089] Figure 6 is a perspective view of a cryo-EM grid according to an embodiment, wherein in a selected region of the grid area, the grid bars have a reduced thickness in a direction perpendicular to the grid plane.

[0090] Figure 7 is a perspective view of a cryo-EM grid according to an embodiment, wherein in one or more selected regions of the grid, the grid pattern has a larger mesh size and / or a larger opening area fraction than the surrounding regions.

[0091] Figure 8(A) is a perspective view of the top of a cryo-EM grid made of two parts aligned and joined together according to an embodiment.

[0092] Figure 8(B) is a perspective view of the bottom of a cryo-EM grid made of two parts aligned and joined together according to an embodiment.

[0093] Figure 9 is a perspective view of a thin metal or carbon sample support foil on the grid. According to an embodiment, the foil has no holes where it overlaps with the grid bars, but has an array of holes within the regions surrounded by the grid bars.

[0094] Figure 10(A) is a perspective view of a pair of tweezers for clamping and handling a cryo-EM grid without damaging it according to an embodiment.

[0095] Figure 10(B) is a perspective view of the bottom grid clamping element of the tweezers according to an embodiment.

[0096] FIG. 10(C) is a perspective view of the bottom grid clamping element of the tweezers according to an embodiment.

[0097] Figure 11 is a perspective view of a conceptual design of a cryo-EM sample cooler with only liquid nitrogen input.

[0098] Figure 12 is a perspective view of an alternative cryo-EM sample cooler design based on the prior art. The sample grid is fixed on a translation stage, and a jet of liquid cryogen is aligned with the sample to cool the sample.

[0099] FIG. 13(A) is similar to Figure 12 a perspective view of an embodiment of a cryo-EM sample cooler according to an embodiment of the present invention, which incorporates a baffle that blocks cold air and liquid cryogen.

[0100] FIG. 13(B) is similar to Figure 12 a perspective view of an embodiment of a cryo-EM sample cooler shown in

[0101] FIG. 14(A) is a perspective view of a cryo-EM grid according to an embodiment.

[0102] FIG. 14(B) is a top view and a cross-sectional view of a blotting disc cut from absorbent material, which is embossed to produce a surface relief that matches the grid design in (A). DETAILED DESCRIPTION

[0103] The goal of this work is to develop tools, including sample holders, sample handling tools, and cooling instruments, that allow for reliable vitrification of cryo-EM samples, reduce grid and foil damage rates, and facilitate automated sample tracking.

[0104] SCIENTIFIC MOTIVATION

[0105] The current generation of cryo-EM sample holders and cooling devices is largely based on research and principles developed in the 1980s. Since then, significant progress has been made in understanding the physics of cooling small samples with liquid cryogens, particularly in cryocrystallography; understanding the physics of ice formation; and our ability to design and manufacture complex structures at reasonable cost. These advancements enable the design and implementation of improved sample preparation techniques to address key bottlenecks in biomolecular structure determination using cryo-EM. We outline below the key concepts and challenges that motivated our invention.

[0106] The cooling rate required for vitrification of cryo-EM samples is approximately 220,000 K / s. The critical cooling rate (CCR) - the minimum cooling rate required for vitrification of a sample - depends on the maximum tolerable or detectable ice fraction in other vitrified samples. By extrapolating the CCRs of different solutes with measured solute concentrations to zero concentration, the CCR of pure water has been determined to be 250,000 K / s (for a crystalline ice fraction determined by X-ray methods of less than approximately 1%). The CCR decreases exponentially with solute concentration, but solutes reduce electron density and EM contrast. For a typical salt concentration of approximately 0.5% w / v in cryo-EM buffers, the CCR is approximately 220,000 K / s.

[0107] The cooling rates achieved in current cryo-EM practice are far below the theoretical limit. Despite using liquid ethane (one of the most effective liquid cryogens), and despite the moderate cooling rate required for vitrifying water, samples in single-particle cryo-EM typically show substantial regions of crystalline ice. For a sample with a diameter of 3 mm, consisting of 50 nm of water and 50 nm of gold or 12 nm of carbon, and plunged edge-on into liquid ethane at a speed of 2 m / s at approximately 90 K, our approximate analysis of heat transfer based on boundary layer approximation predicts a cooling rate of approximately 10 7 K / s. The cooling rate calculated using LN2 in the nucleate boiling regime (starting from a sample temperature below approximately 140 K) is comparable to the cooling rate in liquid ethane; even plunging through N2 gas at 100 K should produce a cooling rate of approximately 200,000 K / s, sufficient to vitrify cryo-EM sample films. For a 30-μm-diameter water sample plunged at 2 m / s, the predicted cooling rate in liquid ethane is approximately 300,000 K / s, comparable to the cooling rate measured using a 30-μm beaded thermocouple. Since crystalline ice is commonly observed in cryo-EM, the sample cooling rate must typically be far below 200,000 K / s.

[0108] Current cryo-EM grids are not optimized to provide the fastest possible cooling rate. Cryo-EM sample holders consist of a film or foil (porous carbon, amorphous carbon, gold, graphene, or other materials) on a metal (copper, gold, molybdenum, nickel, etc.) grid. The sample holder film / foil typically has holes of size 1 μm through which the sample is imaged. The grid diameter is 3.05 mm and the thickness ranges between 10 μm and 25 μm. The grid is stiff enough to be handled manually with pliers or tweezers and does not bend or break during handling or when plunged into liquid ethane for cooling. For a 300 mesh grid, which is popular in single particle cryo-EM, the grid bars are typically 25 μm wide and spaced 58 μm apart, covering approximately 50% of the total grid area. For a gold or copper disk with a diameter of 3 mm and a thickness of 5 μm (average thickness of a 300 mesh grid) plunged into liquid ethane at a speed of 2 m / s, the calculated cooling rate is approximately 10 5 K / s. Crystalline ice is expected to form, and crystalline ice regions 3-5 μm wide (usually wider) adjacent to the grid bars are typically observed.

[0109] Cryo-EM sample buffer (usually water plus about 0.5% salt) and carbon (porous, amorphous) have very low thermal conductivities. For an amorphous carbon film approximately 50-100 nm thick, the calculated heat transfer rate during plunge cooling from the Au / Cu grid bars to the center of each grid opening in the film is approximately 1% of the calculated heat transfer rate from the film to the ethane. Thus, the film at the center of each grid opening is thermally isolated from the grid bars and - if the liquid cryogen flow and boundary layer are uniform - can be cooled at a faster rate.

[0110] However, several factors complicate this situation. First, gold foils approximately 50 nm thick (on gold grids) are becoming increasingly popular because they reduce beam-induced sample motion. The thermal conductivity of gold is approximately 10 3 times that of amorphous carbon, and the calculated heat transfer from the grid bars and through the foil is equivalent to the heat transfer from the foil to the ethane, thus affecting the sample cooling rate.

[0111] Second, during dispensing and blotting, the sample flows through the holes in the support foil and accumulates on the back of the foil, making blotting more difficult due to the protruding grid bars. Where the gap between the foil and the grid bars is small, a large amount of sample can be pinned. When cooling from approximately 300 K to approximately 90 K, water absorbs approximately the same amount of heat per unit volume as gold or copper and has a much lower thermal conductivity. The additional thermal mass and thermal insulation provided by the adhered sample significantly reduce the cooling rate of the sample, foil, and grid bars.

[0112] Third, the large thickness and small spacing of the grid bars create complex fluid flows and may cause gas trapping / entrainment near the grid bars on the back side of the grid bars, reducing the cooling rate.

[0113] Finally, in conventional handling using tweezers, the grid often bends (e.g., like a taco) and the foil bends and is damaged, complicating the ethane flow and potentially causing non-uniform sample cooling.

[0114] The large temperature difference between the grid bars and the sample + foil during cooling may contribute to sample movement during irradiation. Sample movement during irradiation is a major factor limiting single-particle cryo-EM resolution. Movement occurs even during low-dose irradiation and is most rapid at the start of irradiation (per dose), when radiation damage is moderate and the highest-resolution structural information can be obtained. If the signal-to-noise ratio is sufficient, sample "movies" can be analyzed to correct for movement and improve the final resolution. Due to stress / strain caused by the difference in shrinkage between the holder and the grid material during cooling, the bare holder undergoes significant beam-induced movement. This movement is minimized by using the same material (e.g., Au) for both. For samples spanning foil holes, the main mode of movement corresponds to "doming" of the sample film (like a drumhead); for 1.2-μm-diameter holes, the radiation-induced displacement of the sample perpendicular to the foil plane is approximately 150 Å, with a radius of curvature of 25 μm.

[0115] These observations suggest that sample movement results from sample stress-driven radiation-induced creep: as creep proceeds, the driving stress is released and the creep rate decreases. During cooling, the difference in shrinkage between the sample and the foil holder creates sample stress. Between room temperature and the glass transition temperature of water, T g ~136 K, the sample volume expands by approximately 8%, but the sample is liquid, so this expansion is independent of the shrinkage of the foil. Below T g , vitrified water has a positive coefficient of thermal expansion (comparable to that of hexagonal ice), but now the sample shrinkage combines with the shrinkage of the holder foil. Then, by matching the expansion coefficients of the glassy ice and the foil between T g and the final storage temperature of 77 K, the sample stress caused by cooling can be reduced. In this temperature range, the shrinkage of amorphous carbon, copper, and gold foils is less than that of glassy ice, and the shrinkage of gold provides the best match. All of these create tensile stress in the sample. However, in the presence of tensile sample stress, radiation-induced creep does not result in the observed doming.

[0116] The stress that drives radiation-induced sample motion may be partly due to the transient temperature difference between the grid bars and the sample + foil during cooling. Since the foil is much thinner and has much less mechanical stiffness than the grid bars, its shrinkage depends largely on that of the grid bars. Because the grid bars cool and shrink more slowly than the foil between them, the foil develops a transient tension. The sample vitrifies and adheres firmly to this tensile stress foil. When the grid bars cool to the final temperature of the foil + sample, their spacing decreases and the tension in the foil is released, placing the sample under compressive stress. In the foil holes, where the sample is unconstrained by the foil in its relaxed state, radiation-induced creep will produce a "bulge" to relieve this stress. The "bulge" with a height of about 150 observed in foil holes with a diameter of 1.2 μm may be due to a temperature difference as small as 20 K between the grid bars and the foil when the sample vitrifies.

[0117] Current cryo-EM cooling instruments are complex to use and not optimized to provide the fastest possible cooling rate. When the sample is plunged into the liquid cryogen, the cold gas above the liquid cryogen pre-cools the sample. For a plunge speed of about 1 m / s, only a cold gas layer about 2 cm thick dominates the cooling of protein crystals smaller than about 500 μm. Most cryo-EM cooling instruments plunge the sample into a small cup containing liquid ethane, which is surrounded by a larger liquid nitrogen cup. The liquid ethane level is at least a few millimeters below the cup top, ensuring that the cold gas layer is at least this thick. In cold N2 gas at about 200,000 K / s, for a sample with a diameter of 3 mm and a thickness of 50 nm, the calculated cooling rate for a sample plunged at 2 m / s will pass through only 1.6 mm of this gas before cooling below Tg of about 136 K - a distance less than the cold gas thickness. Thus, cooling of the sample and foil between the grid bars may occur mainly in the cold gas before the sample reaches the ethane. The variability in the gas layer thickness and the cooling fraction it provides may lead to variability in the observations.

[0118] Cryo-EM samples can also be cooled by jetting a stationary sample with a liquid ethane jet, as in the VitroJetÔ. The cold gas generated in the ethane jet tube is pushed out before the liquid and similarly pre-cools the sample.

[0119] Description of the Invention

[0120] A sample holder for single-particle cryo-EM that uses liquid ethane or liquid nitrogen to facilitate efficient vitrification, minimize beam-induced sample motion, simplify sample handling and tracking, and facilitate diagnosis of sample preparation issues. Figure 1 (A-F) (from Russo and Passmore, Current Opinion in Structural Biology 37, 81-89 (2016)) illustrates the design of current sample holders used in cryo-EM. The rigid base of the sample holder is a metal "grid" 10 that is 3.05 mm in diameter and 10-25 μm thick. The grid is typically made of copper or gold, but can also be made of molybdenum, nickel, titanium, or other materials. The grid has a solid boundary around the mesh openings (usually sized 40 to 200 μm, depending on the mesh size and open area fraction), and 200, 300, and 400 mesh grids are most popular. The grid is covered with a film or "foil" 20 of porous carbon, amorphous carbon, gold, or other conductive material in the range of 10-50 nm thick, or covered with a graphene layer. The foil has an array of holes 30 typically sized 0.5-2 μm. An aqueous buffer containing the biomolecule to be imaged is placed on the foil, and excess liquid is removed to produce a sample film 40 that is approximately 10-100 nm thick. The sample is imaged in the holes of the foil. Cryo-EM grids are "clamped" (Figure 1 (G)) to make them easier to handle automatically. The grid 10 is placed in an annular holder 50, and a resilient metal clip 60 is inserted to hold the grid in firm contact with the holder 50. The outer diameter of the holder 50 is 3.51 mm, the inner diameter is 2.49 mm, and the thickness is 0.40 mm

[0121] Figure 2 (A) shows how the cryogenic fluid 70 flows around the grid 10 when the grid is plunged into the liquid cryogen. For a typical 300 mesh cryo-EM grid, the grid bar thickness (25 μm) is several times smaller than the spacing between the grid bars (58 μm), so the grid bars interfere with the flow of the liquid cryogen across the back of the foil. Bubbles 80 entrained when the grid crosses the gas-liquid cryogen interface or generated via boiling of the liquid cryogen at the grid surface can be trapped between the holder foil and the grid bars and reduce the cooling rate there. Figure 2 (B) shows how, during the blotting process, the sample solution wets the grid bars and accumulates beside the grid bars through the holes in the foil, thus reducing the cooling rate there.

[0122] Orientation markers with a clamping / handling area。Figure 3 shows a grid 100 according to an embodiment. Grids used in single-particle cryo-EM (e.g., Quantifoil®, C-flat®) have a simple symmetric mesh pattern, with the entire area covered by foil and are prone to bending or damage. In FIG. 3(A), the grid 100 has a mesh with holes 110 on which the sample support foil is placed. The grid has a notch or other serrated feature 120 on one edge, a solid region 130 on one side, and markings 140 and alphanumeric codes 150 patterned on the solid region that uniquely identify each grid. The identification pattern can be produced using, for example, laser or inkjet / dot matrix printing. In FIG. 3(B), the grid is marked by forming holes 160 through the grid. The identification markings are located within the inner edge of the "clip" in FIG. 1(G) and can thus be read optically and in the electron microscope.

[0123] The solid region on one side of the grid facilitates gripping using tweezers, pliers, or preferably tweezers, and according to an embodiment, the gripping surface matches the flat and curved regions of the grid, reducing the risk of grid deformation. If the foil does not significantly overlap this solid region, the risk of foil damage during grid handling will be reduced.

[0124] The notch 120 and the solid region 130 allow the grid to be oriented during immersion in the liquid cryogen and to determine its orientation in the electron microscope. During the immersion cooling process, the fluid flow, the thickness of the thermal boundary layer, and the heat transfer rate vary with position on the grid, especially along the direction of the immersion path. With a dedicated gripping area, the grid orientation during the immersion cooling process will be fixed, so the properties of the ice and EM inspection along and perpendicular to the immersion direction of the position can be used to characterize the heat transfer through the grid, optimize the grid and cooling instrument design, and speed up the identification of grid areas with well-vitrified samples.

[0125] Only a small number of grid squares are needed to collect a complete data set. Near the leading edge of the grid where the thermal boundary layer is thinnest, the cooling rate should be maximum. Thus, most of the grid area can be sacrificed for the gripping area. The preferred immersion direction is indicated by the arrow in FIG. 3.

[0126] Figure 4(A) shows an alternative grid embodiment 200 with hexagonal mesh holes. The grid has an asymmetric marking or feature 210 located in the grid area, which remains visible once clamped. This marking is a feature of some prior art grids that allows the use of an electron microscope and the use of a magnifying glass or optical microscope to determine the grid orientation - rotation about the central axis perpendicular to the grid plane, and which side of the grid is facing up. The grid has a set of solid, non-porous areas 220, sized one or several grid units, which can be arranged in an asymmetric pattern to assist in the visual orientation of the grid. Figure 4(B) shows a similar grid with a solid clamping area and a notch.

[0127] Barcodes for sample tracking . The flat clamping area in Figure 3 or the smaller solid area 220 in Figure 4 can be patterned with a unique identifier for each grid. These can be photographed in the cooling instrument before cooling, verified during EM imaging, and used with a spreadsheet or laboratory information management system (LIMS) for tracking individual grids. The grids can be marked individually after manufacture so that each manufactured grid has a unique marking. Alternatively, there can be a limited number of different markings encoded in the photomasks used in manufacture. Approximately 2000 individually marked grids can be manufactured from a single 6" mask. This limited set of markings, combined with grid carriers equipped with barcodes and / or cryogenic RFID tracking capabilities, may be sufficient to achieve robust sample tracking.

[0128] Grid patterns that reduce thermal gradients and increase sample thermal and mechanical isolation during cooling . The current grids used in single-particle cryo-EM have a uniform thickness of 10 μm (Au) or 18 - 25 μm (Cu, Mo). During manual handling and trimming, the overall grid thickness cannot be significantly reduced without sacrificing mechanical rigidity. However, since only a small portion of the grid area is sufficient for structure determination, the grid bar thickness and / or width can be significantly reduced in selected areas away from the "clamped" grid periphery to increase the grid bar cooling rate and reduce the grid bar stiffness to allow more deformation to occur during cooling in response to the shrinkage difference between the grid bars and the foil. Since the orientation of the grid can be fixed during the cooling process, the reduced-thickness grid bar pattern can be arranged to form "channels" for the laminar liquid cryogen flow and maximize the heat transfer rate from the sample and the foil.

[0129] Figure 5(A) shows a standard cryo-EM grid 10, where the grid bars 230 have a uniform thickness and width, and the grid openings have fixed dimensions. Figure 5(B) shows a grid 240 according to an embodiment, where, in a grid region including at least 4 grid squares, the grid bars 250 have a reduced width, and where the horizontal grid bars 260, i.e., the grid bars oriented parallel to the liquid cryogen surface during plunge-in, can be removed. The reduced width reduces the grid bar heat capacity and heat conduction in these regions. This will increase the grid bar cooling rate and reduce the temperature difference between the grid bar and the sample + foil during cooling with liquid cryogen. The regions with reduced grid bar width preferably extend along the direction of sample movement during plunge-in cooling to reduce flow perturbations caused by the grid bars and improve heat transfer. The placement, size, and structure of the regions where the grid bars have a reduced width and / or are removed should not significantly reduce the overall mechanical stiffness and robustness of the grid in a way that also increases beam-induced sample movement on the foil. The total area of the modified grid should be less than 25% of the total grid area, preferably less than 10%, and an individual modified region should be less than 5% of the grid area, preferably less than 2%.

[0130] Figure 6 Figure 4 shows a cryo-EM grid 270 according to an embodiment, where, in a selected region of a grid region including at least 4 grid squares, the grid bars 280 have a reduced thickness perpendicular to the grid plane. The reduced thickness reduces the grid bar heat capacity and heat conductance and reduces perturbations in the liquid cryogen flow and gas trapped by the grid bars during plunge-in cooling. This will increase the grid bar cooling rate and reduce the temperature difference between the grid bar and the sample + foil during cooling in liquid cryogen. The grid regions with reduced thickness preferably extend along the direction of sample movement during plunge-in cooling. The placement, size, and structure of the grid regions with reduced thickness should not significantly reduce the overall mechanical stiffness and robustness of the grid when handling at the grid edges or regions designed for grid clamping in a way that also increases beam-induced sample movement on the foil.

[0131] Figure 7 Figure 6 shows a cryo-EM grid 290 according to an embodiment, where, in one or more regions of the grid, the grid pattern has a larger mesh size and / or narrower grid bars than the surrounding regions. The grid bars in the selected regions should cool faster. The large mesh openings will increase the fraction of the sample + foil that is far from the grid bars (e.g., at least 10 μm away). These regions of the sample + foil cool faster, making vitrification more likely. The mechanically stronger parts around the grid will facilitate grid handling without damage. In Figure 7 Figure 7, the grid has a hexagonal aperture pattern and grid bars, and the central region has an area of approximately 25% of the grid area. The circular central region of the modified grid layout 300 minimizes anisotropic grid stresses that may occur during cooling.

[0132] Cryo-EM grids are typically fabricated using electroforming, which deposits a layer of metal of uniform thickness on a mandrel. A multi-thickness grid can be fabricated by performing multiple photoresist deposition, patterning, and electroforming steps, or by bonding together two separately electroformed grids. FIG. 8 shows a cryo-EM grid according to one embodiment, which includes a top grid 310 and a bottom grid 320, with a sample support foil placed on the top grid 310 and the bottom grid 320 having a different thickness and a complementary pattern. The top and bottom are fabricated separately and then bonded together. The bottom grid 320 can have an aperture 330 such that the thickness of the bonded grid within the aperture is exactly the thickness of the top grid. The resulting composite grid can have regions of very different thicknesses and cool at very different rates while maintaining mechanical rigidity. The two grids can be of the same material to ensure a thermal shrinkage match, or can be of different materials if, for example, the goal is for the top grid to be in tensile stress when cold.

[0133] The grids used in single-particle cryo-EM are typically made of copper or gold and are covered with a foil made of amorphous carbon or gold. According to one embodiment, the grid is made of a conductive material having an average thermal expansion coefficient between 77 K and 300 K that is less than that of the foil material. The thermal expansion coefficients of tungsten, molybdenum, tantalum, and titanium grids are all very small (4.3, 5, 6.5, and 9 ppm / °C, respectively), while those of gold, copper, and nickel are much larger (14, 16, and 13 ppm / °C, respectively). With this combination, the faster cooling of the sample + foil relative to the grid bars does not result in a net compressive stress in the vitrified sample at the final sample temperature (between 77 K and 120 K), which could drive beam-induced sample motion. During cooling, the sample remains liquid at a temperature close to its glass transition temperature Tg of approximately 136 K. Once vitrified, further cooling of the sample + foil causes both to want to shrink, but initially they are restricted in their shrinkage because the grid bars are hotter and shrink less, and the grid bars are much harder than the sample + foil. When the grid bars cool to the final temperature, the grid bars shrink, reducing the tensile stress on the foil, which could compress the sample on the foil. By minimizing the thermal shrinkage of the grid material and making its total shrinkage from room temperature to cryogenic temperatures much less than that of the foil, the sample + foil will remain in tension during cooling. Preferably, the thermal shrinkage of the grid material between room temperature and 77 K is equivalent to or less than the thermal shrinkage of the foil material between 136 K and 77 K, and is equivalent to or less than the thermal shrinkage of amorphous and hexagonal ice between 136 K and 77 K. This is approximately the case for Au foil and Ti, W, or Mo grids. Doped silicon grids shrink very little, much less than any metal, and also meet this criterion, but are too fragile at the required thickness.

[0134] Foil materials and patterns for optimal sample cooling, optimal imaging, and minimal sample movement. Using gold foil on a gold grid reduces the overall differential shrinkage of the foil and the grid when cooling from 300 K to 77 K. The thermal shrinkage of gold between the glass transition temperature T g = 136 K and 77 K of water roughly matches that of vitreous ice, but its relatively high thermal conductivity may conduct too much heat from the grid bars and impede the vitrification of the sample on the foil between the grid bars. To thermally isolate the sample + foil from the grid bars while maintaining the attractive thermal expansion properties of gold, the foil can be made of an alloy of gold and chromium. For example, the thermal conductivity of an Au-Cr alloy containing 1% Cr is about 1 / 10 of that of Au, but the electrical conductivity at 77 K is similar. Other alloys (e.g., Au-Pt with a concentration of 1 - 5% Pt) reduce the thermal conductivity while maintaining good electrical conductivity. Other low-thermal-conductivity, high-electrical-conductivity alloys may be suitable.

[0135] During sample deposition and / or blotting, liquid samples typically wet the grid bars through the holes in the foil. This increases the thermal mass near the grid bars and reduces the cooling rate there. Once the liquid solidifies, the flow and wetting of the fluid will firmly attach the foil to the grid and strongly couple their shrinkage during cooling, which may generate stress in the sample and cause sample movement induced by the electron beam. By eliminating the fluid in contact with the foil and the grid, the foil can slide freely on the grid bars during cooling.

[0136] One way to reduce the flow and wetting of liquid samples on the grid bars is to use a non-porous foil directly above and near the grid bars. In Figure 9 , the grid 10 has a foil 340 that has no holes where it overlaps the grid bars and has an array of holes 360 in the area surrounded by the grid bars. The gap between the edge of the grid bar and the nearest hole in the foil should be at least 5% of the spacing between the grid bars or the diameter of several holes to minimize the chance that liquid may bulge through the holes and expand to wet the grid bars.

[0137] The foil may have fiducial points or other markings to facilitate the alignment of the foil and the grid, so that the solid regions of the foil are aligned with the grid bars. The non-porous area registered with the grid bars can be limited to only selected regions of the foil, such that the grid bars can be seen elsewhere under the foil, thus facilitating the alignment of the foil and the grid bars.

[0138] The foil covering the grid may have multiple different hole sizes, between 0.2 and 5 μm, distributed over its area. This diversity of hole sizes can create a diversity of ice thicknesses and cooling rates, which helps to obtain optimal imaging conditions.

[0139] The foil covering the grid can be made of regions having two or more different thicknesses, where one or more of these regions have an array of through-holes. When the sample is properly blotted, the thickness of the holes determines the thickness of the ice. Having two thicknesses on the same foil increases the chance of obtaining the optimal sample thickness for imaging. The periphery of the foil can be made much thicker than the rest to simplify handling of the foil and its placement on the grid. During the continuous deposition of metal or carbon, a foil with increasing thickness as the radius increases can be made by using a disk-shaped shadow mask with an increasing radius. The innermost foil region will be the thinnest, while the outermost peripheral region will be the thickest. The shadow mask can be made of, for example, 1-mil-thick copper foil or an exposable polymer such as SU-8 or polyimide using standard photolithography and etching processes. The disks of the foil array must be connected by thin lines. The masking by these lines can be reduced by using different line positions on successive masks, or by placing the masks in close contact and depositing the metal or carbon at an angle.

[0140] The metal grid and the metal foil can be fabricated together in a single manufacturing process, making them automatically aligned, rather than in two separate processes that require alignment steps. The process can include depositing a release layer on a substrate; depositing a foil layer; depositing a photoresist; exposing the hole pattern of the foil in the photoresist; etching the hole pattern in the foil; removing the photoresist; depositing a second photoresist; exposing and developing the grid pattern in the photoresist; electroforming the grid onto the foil through the openings in the photoresist; removing the photoresist; and releasing the completed grid-plus-foil from the substrate.

[0141] Tool for handling grids。The present invention also includes the design of tools / tweezers for handling cryo-EM grids without damaging them. Standard tip stainless steel tweezers can damage the foil if the tip hits the foil, if the grid bends by contacting a surface, for example, such that the foil hits the tip, or if the grip of the tip slips. Since the tip is much narrower than the grid, hitting the side of the grid into a slot in, for example, a grid storage box will bend it into a "taco" shape. In the embodiment shown in FIG. 10, the sample end of the spring-actuated tweezers 370 (which is normally open or normally closed when not squeezed, depending on whether the two arms of the tweezers are substantially straight or crossed such that the "bottom" tip is attached to the arm at the top of the squeeze end) has gripping elements 380 and 390 attached to each arm near its tip. The inner surfaces of these gripping elements in the portion gripping the grid are flat, and their width is preferably comparable to the diameter of the grid (3.05 mm), i.e., within ±20% of this diameter. This width will minimize the risk of the grid being bent by the force applied to its sides when held. The gripping elements can be configured to "mate" with the grid such that the elements only contact specific portions of the grid, such as the solid area in the lower part of the grid in FIG. 3. For example, the lower gripping element 380 can have a thin or recessed area 400 (possibly with an aligned "nose") that matches the profile / shape of the grid. This recessed area should ideally be slightly shallower than the grid thickness, typically 10 or 25 μm. The lower grid element can also have a protrusion 410 that mates with a hole 420 in the upper gripping element 390 to help position the grid relative to the gripping element and prevent the grid from slipping past the recessed area 400 while allowing the tweezers to close and firmly grip the grid.

[0142] The tweezer body / arms can be made of metal, such as stainless steel commonly used in tweezer bodies, or of a polymer. The gripping elements can be metal with a smooth (possibly polished or electroplated) inner surface to minimize the risk of wear and damage to the foil, polymer, or bendable glass or ceramic. The tweezer gripping elements only need to extend - a short distance - a few millimeters - beyond the ends of the tweezer arms. Thus, the gripping elements can be thin, slightly flexible, and optically transparent polymers such as polyimide, PDMS, or COC, allowing the position of the grid within the gripping element to be directly seen. The more rigid but opaque polymer SU-8 may also be suitable. These polymers can be machined into the desired shape - and with the desired micron-level thickness tolerances - using microfabrication methods based on standard lithography techniques. Other polymers suitable for injection molding can also be used. The grid gripping elements can be attached to the jaws of standard spring-loaded metal or polymer tweezers using cryo-compatible binders, ultrasound, using screws, or by other standard means. Note that these tweezers should preferably be usable both at room temperature and when the grid is immersed in liquid ethane and liquid nitrogen.

[0143] A benchtop plunge-cooling system for reliable vitrification of cryo-EM samples using only liquid nitrogen. Since the work of Dubochet in the 1980s, all cryo-EM sample cooling devices have used liquid ethane (or ethane / propane) maintained above its melting point as the cooling medium. Flammable ethane gas must be stored in the laboratory, liquefied by flowing into a cup cooled by liquid nitrogen (LN2), and then cooled to and maintained at the desired temperature; the plunge-cooled sample must be transferred through air into liquid nitrogen for storage and transport; the remaining liquid ethane is evaporated and safely vented. In current-generation plunge-cooling instruments, ethane and nitrogen are exposed to room air, so moisture condenses into frost and contaminates the grids plunged into and stored in it.

[0144] The current cryo-EM sample cooling rate is limited by grid design and pre-cooling of the cold gas above liquid ethane. In addition, the choice of ethane was based in part on incorrect interpretations and extrapolated measurements using large thermocouples, suggesting a cooling rate 20 times higher than LN2. In fact, for small samples (protein crystals and cryo-EM grids), the cooling rate advantage of liquid ethane is at most about 3-fold, which can be easily compensated for by appropriate cooling instruments and grid design. MiTeGen developed NanuqÔ, an instrument for high-throughput cryo-cooling and automated storage of protein crystallography samples using LN2. It provides cooling rates of over 50,000 K / s and 140,000 K / s for 30-μm thermocouples in film boiling and nucleate boiling regimes, respectively. In commercial applications, it has greatly simplified the cryo-preservation and storage of samples. This performance is achieved by combining a high-speed sample translation stage, a gas management manifold, and a precision liquid nitrogen level control system, where the gas management manifold removes and replaces the cold gas above LN2 with vacuum and thermal supplement gases immediately before sample plunge. These components reduce the cold gas layer (gas temperature below 273 K) to < 50 μm, reduce the pre-cooling time of the sample in this cold gas to < 25 μs, and provide the highest cooling rate ever reported using liquid nitrogen. Principle experiments using NanuqÔ demonstrated that thin buffered solution samples on cryo-EM grids can be fully vitrified using LN2 as the cooling medium.

[0145] Figure 11Shows the design of an LN2-based benchtop cryo-EM plunge cooler according to an embodiment, which eliminates manual grid handling between plunge and microscope loading. Compared with protein crystal sample holders, the small mass and impact cross-section of cryo-EM grids allow the vertical translation stage 500 to transfer speeds up to 7 m / s (and possibly 10 m / s), which should double the cooling rate compared to NanuqÔ. The grid holding mechanism 510 holds the grid absolutely perpendicular to the LN2 surface to maximize the cooling rate and minimize foil damage, and preferably automatically releases the grid into a storage container after plunge cooling. The grid holding mechanism preferably has a minimum hydrodynamic cross-section over a length of at least a few centimeters along the plunge direction, as larger plunge speeds require a significant stopping distance (which makes them impractical when using liquid ethane).)

[0146] As in the prior art NanuqÔ of MiTeGen, the plunge cooler incorporates a gas management manifold 520 above the main plunge chamber, which removes the cold gas present in the plunge hole by suction and replaces it with ambient temperature nitrogen, thus creating a transition between ambient temperature gas and liquid nitrogen within a distance of less than 100 μm and eliminating grid pre-cooling in the cold gas.

[0147] To reduce the waves that reduce cold gas layer removal associated with LN2 surface boiling and to improve cooling performance by allowing LN2 to absorb some heat before it evaporates, the LN2 in the main sample plunge chamber 530 is cooled below its boiling point of 77 K by thermal contact with a second (insulated) chamber 540, where the LN2 evaporatively cools to its solidification temperature of 63 K. This second chamber is sealed from the atmosphere and is connected via a port to a vacuum pump 550 or other vacuum source to reduce the gas pressure above the LN2 in the second chamber.

[0148] The initial sample position is within the humidification chamber 560 or within the humidification air stream. The humidity control system 565 preferably produces at least a few percent humidity up to 100% saturation to allow control of the amount of sample evaporating from the grid.

[0149] After plunge cooling, the grid is preferably automatically deposited into a commercial electron microscope cassette or other holder 570 within the main plunge chamber. In one embodiment, a mechanical stage 580 under automatic control is provided within the main chamber, which receives a commercial cryo-EM microscope cassette or a custom storage box and automatically positions them to be aligned with the sample plunge path defined by the vertical translation stage, such that each cold sample can be deposited in a separate compartment within each holder by a combination of only the vertical movement of the vertical translation stage and only the horizontal movement of the mechanical stage that places the sample holder storage box.

[0150] Using the liquid level control system in the prior art NanuqÔ, the LN2 liquid levels in the main input chamber and the outer chamber are precisely maintained.

[0151] Improvements in jet-cooling cryo-EM samples using cryogenic liquids 。 Figure 12 An alternative cryo-EM sample cooler design that has been implemented in the prior art is shown. In this design, the cryo-EM grid sample 600 is translated from an initial position (usually within a humidification chamber (not shown)) to an intermediate position by an automated translation stage 610. Then, it is cooled by a jet 620 of one or more liquid refrigerants (usually liquid ethane), which exits a pipe or conduit 630 (which may have a nozzle at the exit end). Once cooled, the sample is rapidly transferred to a thermally insulated container 630 containing liquid nitrogen and then to a storage case or cassette 640. The advantage of this method is that the grid can be "clamped" before cooling because the liquid refrigerant is not blocked by the clip on its way to the sample holder foil as it would be if the clamped grid were dropped into the liquid refrigerant. However, when the jet of the refrigerant flow first starts towards the sample, the cold gas 650 present in the tube / nozzle or generated by contact with the liquid refrigerant will leave the conduit before the liquid refrigerant. When the distance between the conduit outlet and the sample is small, this cold gas will reach the sample first and pre-cool the sample at a much lower cooling rate than when the cryogenic liquid finally lands on the sample. At the cooling rate provided by the liquid cryogen, the sample cools from room temperature to near 100 K in only about 0.4 milliseconds. At a typical jet velocity of 5 m / s (limited by the impact force on the grid and foil), the liquid only moves 2 mm during this time, which is approximately half the grid diameter. According to the analysis by Kriminski et al., the rate at which the cold gas present in the tube cools the sample may be 20 times smaller than that of the cryogenic liquid. Thus, 4 cm of cold gas is sufficient to cool the sample to near 100 K, and this slower cooling rate may not produce vitreous ice.

[0152] Figure 13 shows two embodiments of the present invention, which solve the problem by preventing cold gas from reaching the sample. Figure 12The problem of pre-cooling the sample with cold gas in a cryogenic jet cooler of the type shown. In the first embodiment, in (A), by using a very fast baffle or vane 660, the cold gas 650 emerging from the jet tube 630 before the liquid cryogen 640 is prevented from reaching the sample. The baffle initially blocks the gas and cryogen flow, and then suddenly moves away once the cold gas has been cleared and a steady cryogen flow towards the baffle has been established. The baffle can have a linear or rotational motion, driven by, for example, a stepper or synchronous motor, solenoid, compressed gas, or other standard devices. The time for the baffle to move a distance equal to the jet diameter (a few millimeters) should be equivalent to or less than the sample cooling time <1 ms, corresponding to a speed of at least several meters per second so that the entire grid is hit by the jet. The baffle is preferably made of an insulating material such as a polymer, glass (e.g., quartz), or ceramic, with a small mass and sufficient thickness such that the sample is not pre-cooled by conduction, convection, or radiative cooling caused by the cold baffle during the short time between the start of the liquid cryogen flow and the opening of the baffle.

[0153] In the second embodiment, shown in Fig. 13 (B), a jet or "blade" of moisture-free room-temperature gas is directed through the gas flow from the jet tube, with a flow rate large enough to sweep away the cold gas but small enough that the path of the liquid cryogen, which has a much greater momentum, from the jet tube to the sample is not overly disturbed. This hot gas flow can be turned on before the liquid cryogen jet is activated and turned off once the sample has been cooled. The warm gas can be directed downward as shown, or upward, away from the insulated container storing the cold sample. The gas can be dry air, N2, He, or any other non-flammable, non-reactive gas with a boiling point higher than the freezing temperature of the liquid cryogen. The gas can come from a compressed gas cylinder or other compressed gas source.

[0154] Improvements in removing excess liquid from grids by blotting。In single-particle cryo-EM, biomolecules in an aqueous buffer are deposited on the sample support foil of the grid. The initially deposited sample volume is typically 0.1 - 1 mL, and the resulting droplet extends a fraction of a millimeter above the foil surface. To obtain sufficient image contrast in the electron microscope, it must be reduced to a film approximately 50 nm thick that covers the holes in the foil. A combination of evaporation in a controlled humidity environment (which results in concentration of the biomolecules in the remaining liquid) and blotting using an absorbent material, typically filter paper (e.g., manufactured by Whatman), is used to remove the excess sample. Neither is precise and difficult to control. Manual blotting can be done by gently touching the grid with a thin strip of filter paper. Commercial cryo-EM cooling instruments, such as the Vitrobot® from Thermo Fisher, use blotting paper circles a few centimeters in diameter on pads that are pushed at an angle against one or both sides of the grid with a controlled force and for a controlled duration. The angled attack of the blotting paper allows the liquid to be drawn out from one side of the grid, which helps reduce the force of the liquid passing through the holes in the foil to the back of the foil, where the grid bars make it more difficult to remove.

[0155] According to one embodiment of the present invention, blotting of the cryo-EM grid is performed using small discs of absorbent material having a diameter equal to or slightly larger than the grid, and the absorbent material preferably has a surface relief pattern. These discs are preferably attached to a solid disc-shaped pad attached to a rod, and the rod is moved such that the discs are always parallel to the grid plane. As the discs are moved towards the foil and the grid, when they first contact the liquid meniscus, the liquid is first drawn away by the raised areas of the discs. When the discs are pressed into direct contact with the foil, the liquid continues to be drawn laterally towards the raised areas of the discs for contact. At the same time, the disc material is soft, so as the pressure increases, the relief on its surface decreases, further helping to remove the excess liquid. As a result, if the approach of the discs to the grid is not too fast, the liquid can be effectively removed without being pushed through the holes in the foil to the back of the foil. In addition, because the absorbent discs are flattened onto the grid, they effectively seal the grid from the surrounding air, preventing liquid evaporation. As a result, the removal of the liquid can be more finely controlled.

[0156] In a preferred embodiment, the absorbent material is cut to substantially match the size and area of the cryo-EM grid. If the grid is blotted while being held in an plunge or spray cooler using, for example, tweezers, the absorbent disc can have a cutout such that it does not contact the tweezers and such that it can be flattened onto the part of the grid not covered by the tweezers.

[0157] In a preferred embodiment, the absorbent material is filter paper, and various grades of filter paper are available from suppliers such as Whatman.

[0158] Surface relief of the filter paper can be produced by embossing, which involves pressing the filter paper against a master die with sufficient force to transfer the master die pattern to the filter paper. The thickness of a typical Whatman filter paper is 180 μm, which roughly limits the spatial scale of the surface relief pattern. To obtain a finer pattern, the filter paper can be wet before embossing. It can also be chopped with water in, for example, a blender, pressed into a master die, and then dried.

[0159] In another embodiment, the sample support foil on the grid has a hole pattern complementary to the raised area of the absorption disk, so the foil is continuous where the filter paper contacts the foil and has holes elsewhere. This will further reduce the chance of liquid being pushed through the holes in the foil to its back side.

[0160] Figure 14 (A) shows a cryo-EM grid (covered with a foil not shown) according to one aspect of the present invention, which has a solid clamping area and a central area with a larger opening area fraction and a smaller grid bar width. Figure 14 (B) shows a top view and a side cross-sectional view of the absorption disk 700, which is cut away to avoid the clamping area of the grid. The grid is embossed to produce a raised area 710 around the outside of the mesh area.

Claims

1. A sample holder device, comprising: a grid having a first surface, a second surface, a grid thickness, an inner portion, and an outer portion, the inner portion including a plurality of grid bars defining a plurality of holes passing through the first and second surfaces between the grid bars, the plurality of holes having a first diameter or width; and a sample holder film in contact with the first surface of the grid, the sample holder film having a film thickness less than the grid thickness, the sample holder film further defining a plurality of orifices extending through the sample holder film and including a second diameter or width less than the first diameter or width of the holes in the grid, wherein the grid comprises a first metallic material having a first electrical conductivity and a first thermal conductivity, and the sample holder film comprises a second metallic material having a second electrical conductivity and a second thermal conductivity less than the first thermal conductivity, such that during cooling of the sample holder device with a cryogenic fluid, the second thermal conductivity of the sample holder film inhibits heat transfer from the grid bars to the sample support film while allowing the sample support film to cool faster than the grid bars near the holes defined by the grid bars.

2. The sample holder device according to claim 1, wherein the second metallic material comprises a metallic alloy of chromium and gold having a chromium content between 0.1% and 10%.

3. The sample holder device according to claim 1, wherein the second thermal conductivity of the second metallic material is 1 / 10 of the first thermal conductivity of the first metallic material.

4. The sample holder device according to claim 1, wherein the grid thickness is between 10 microns and 25 microns.

5. The sample holder device according to claim 1, wherein the film thickness is between 10 nanometers and 100 nanometers.

6. The sample holder device according to claim 1, wherein the grid bars form a mesh having 200 to 400 cells per linear inch.

7. The sample holder device according to claim 1, wherein the second diameter or width of the orifices in the sample holder film is between 0.5 microns and 2.0 microns.

8. The sample holder device according to claim 1, wherein the sample holder film has a solid overlapping region without through holes, the solid overlapping region overlapping with the grid bars and located within a region adjacent to the grid bars.

9. The sample holder device according to claim 8, wherein the solid overlapping region of the sample holder film includes a region adjacent to the grid bars, the width of the region being 5% of the width of the grid bars or at least twice the first diameter or width of the holes in the grid.

10. The sample holder device according to claim 1, wherein the first metallic material has a first average coefficient of thermal expansion at a temperature between 77 Kelvin (K) and 300 K, and the second metallic material has a second average coefficient of thermal expansion greater than the first average coefficient of thermal expansion at a temperature between 77 K and 300 K, such that cooling the sample support device to cryogenic temperatures tensions the sample support film.

11. The sample holder device according to claim 10, wherein the first metallic material comprises tungsten, titanium, molybdenum, tantalum, and / or alloys thereof, and the second metallic material comprises gold, copper, nickel, and / or alloys thereof.

12. The sample holder device according to claim 10, wherein a first thermal shrinkage of the first metal material at a temperature between 293K and 77K is equal to or less than a second thermal shrinkage of the second metal material at a temperature between 136K and 77K.

13. The sample holder device according to claim 10, wherein a first thermal shrinkage of the first metal material at a temperature between 293K and 77K is equal to or less than an ice thermal shrinkage of amorphous ice and hexagonal ice at a temperature between 136K and 77K.

14. The sample holder device according to claim 10, wherein a first linear thermal shrinkage of the first metal material at a temperature between 293K and 77K is less than half of a second linear thermal shrinkage of the second metal material at a temperature between 293K and 77K.

15. The sample holder device according to claim 10, wherein a first linear thermal shrinkage of the first metal material at a temperature between 293K and 77K is between 0.26 and 0.70 times a second linear thermal shrinkage of the second metal material at a temperature of 293K and 77K.

16. The sample holder device according to claim 10, wherein the sample support film further comprises a first film region having a first film thickness and a second film region having a second film thickness different from the first film thickness.

17. The sample holder device according to claim 16, wherein both the first and second film regions of the sample holder film include the orifice.

18. The sample holder device according to claim 17, wherein a periphery of the sample holder film is thicker than an inner region of the sample holder film.

19. The sample holder device according to claim 18, wherein the periphery of the sample holder film has no orifice.

20. The sample holder device according to claim 16, wherein the first thickness of the first film region is between 10 nanometers and 100 nanometers.

Images