Patent Application
20250123223
The invention relates to a method for investigating nanoscale biological specimens in an electron beam instrument. Such a method is for example known from U.S. Pat. No. 7,413,872.
Electron beams are frequently used to examine the structure of nanoscale biological objects such as proteins or virus particles via direct microscopy or via diffraction methods. A typical instrument to investigate nanoscale biological objects is a transmission electron microscope (TEM); note that nanoscale biological objects subject to an investigation are also called nanoscale biological specimens.
To achieve a spatial resolution in the sub-nanometer range, significant doses of electrons of an electron beam are required. There are specimens that can stand even high doses of electrons, such as metallic nanoparticles. However, a broad range of specimens is rapidly damaged in the electron beam, such as polymers, certain battery materials involving lithium, and most biological specimens such as proteins and virus particles. Typically, the electron dose limit for proteins and virus particles is 10 electrons per Å2, for which sufficient signal-to-noise-ratio (SNR) is obtained for a spatial resolution of approximately 5 nm under standard conditions. A better resolution would require a higher electron dose, whereby the dose scales with at least the square of the spatial resolution (see textbook: L. Reimer, H. Kohl, Transmission Electron Microscopy: Physics of Image Formation, Springer, 2008, page 479, equation 11.17), but a stronger scaling up to the fourth power under certain conditions has also been reported (N. de Jonge, Ultramicroscopy 187 (2019) 113-125). To resolve the structure of a wide range of biological nanoscale objects, a spatial resolution of at least 0.5 nm would be needed, and it is thus highly desired to mitigate radiation damage.
The key problem in electron microscopy of biological samples is the occurrence of radiation damage via inelastic electron scattering (R. F. Egerton, Micron 119 (2019) 72-87). Damaging processes include the creation of excited states in molecules, and the creation of charge pairs (ionization). If charge pairs are created at the surface of the sample, emission of secondary electrons may take place resulting in electrical charging of the sample. Structural damage to the sample mainly occurs when atomic bonds are broken following inelastic scattering processes. Heating up of the sample is not relevant for typical conditions involving biological samples. Another process, so-called knock-on damage, is typically of much smaller magnitude than damage via charging effects, and it can be overcome by lowering the beam energy below the knock-on threshold, although the energy level to prevent damage on hydrogen is so low that a limited spatial resolution can be achieved.
A first partial remedy to radiation damage is to lower the influx of electrons (also simply called electron flux) such that the number of created excited states and charge pairs per unit time is reduced (E. E. Fill, F. Krausz and M. G. Raizen, New J. Phys. 10 (2008) 93015-1-7). At the typical experimental conditions in electron microscopy, multiple adjacent charge pairs are created such that the net Coulomb force is in an outward direction, promoting beam damage. Below a threshold influx, only one adjacent charge pair is present at one time point, such that the net Coulomb force is towards recombination. Mitigation of beam damage by operating the electron beam at a low electron flux has been reported in the open literature, for example, S. Keskin and N. de Jonge, Nano Lett. 18 (2018) 7435-7440. Examination of a biological specimen with transmission electron microscopy was conducted with a low flux of the electron beam, below 5 e·Å−2s−1, such that electron beam damage by electrical charging was reduced, while radiation damage occurred at higher flux for the same sample.
For charge separation to be equilibrated in organic molecules, an electron transfer process needs to take place. The rate constant in electron transfer between two fixed redox states is given by (R. A. Marcus, N. Sutin, Biochim. Biophys. Acta 811 (1985) 265-322):
with driving force −ΔG0, temperature T, Boltzmann's constant kb, reorganization energy λ, and electronic coupling Hab. In one model, Hab is given by (C. C. Moser et al., Nature 355 (1992) 796-802):
with distance between donor and acceptor R, tunneling factor β, and maximal electronic coupling V0. Depending on how far an electron created in an ionization event travelled, the recombination time thus varies, but the time also depends on the exact electron transfer pathways in the molecule (D. N. Beratan et al., Science 252 (1991) 1285-1288). Electron transfer in a protein typically occurs around room temperature, and at distances less than a few nanometers.
Important is furthermore that the object is placed on an electrically conductive support, which is connected to the electrical neutral (in general “earth” potential) of the electron beam instrument. The latter is typically achieved by placing the sample on an electrically conductive foil, for example, a carbon foil that is placed in the instrument by means of a specimen holder.
A second partial remedy against radiation damage is by freezing the sample of specimens that are normally in a liquid state, such as most biological specimens. Primary damage occurs when an electron directly hits the object under observation, while secondary damage occurs after the medium surrounding the object has been hit, and chemical radicals and free electrons diffuse towards the object, creating damage. In addition, molecular structures in which bonds are broken, remain fixed in a solid ice matrix thus reducing structural degradation. Freezing the medium protects the sample against secondary radiation damage because of the reduced diffusion and the increased structural stability compared to the situation in liquid (L. F. Kourkoutis et al., Annu. Rev. Mater. Res. 42 (2012) 33-58).
A practical problem is that freezing a biological may lead to ice crystals damaging the morphology of the investigated sample. The state-of-the-art method used in structural biology is to freeze a sample sufficiently fast and down to liquid nitrogen temperatures such as to obtain amorphous ice. The sample typically consists of a saline layer containing protein placed on a carbon film with holes. After removing excess liquid and rapid freezing, the resulting free standing amorphous ice layer in the holes in the carbon film has a thickness of about 50-70 nm. During the examination of proteins in this free standing ice layer, the temperature is kept below the so-called glass transition temperature of ice (typically about 100° K). Ice crystal formation is thus prevented.
However, ice at the above-mentioned low temperature is an insulator. Pure ice exhibits a poor static electrical conductivity of σ=10−7 Ω−1m−1 at freezing point (see G. W. Gross et al., J. Glaciol. 21 (1978) 143) and decreases to values below 10−12 Ω−1m−1 for temperatures below 200 K. Ice at the later temperature remains an insulator even when ions are mixed with the ice (see Stillman and Grimm, Lunar and Planetary Science XXXIX (2008) 2277). Charge compensation from the electrical neutral of the instrument to the sample upon the emission of secondary electrons is thus impossible. Above a certain electron dose, the sample charges up so much that the electric field in the sample exceeds the dielectric strength of ice of 8×107 Vm−1 (T. Kohno et al., IEEE Transactions on Electrical Insulation, EI-15 (1980) 27-32), and sample damage occurs, for example, the formation of hydrogen gas.
Also, charge transfer rates within the object become highly limited. To exemplify the influence of the above-mentioned temperature on electron transfer, one may consider eq. 1 for typical values of ΔG0=−1.5 eV, and λ=0.75 eV in a protein, and a temperature change from 293 K to 77 K. This temperature change would result in a reduction of the electron transfer rate by a factor of 6×108, which would basically block electron transfer in the protein. Yet, electron transfer is possible within the protein if −ΔG0=λ (H. Wang et al., J. Phys. Chem. B 113 (2009) 818-824), which is only rarely the case in practice. Therefore, at low temperature, radiation damage mitigation is impossible via operating at a low electron flux such that charge pairs can recombine. Cryogenic sample damage thus occurs above a certain threshold electron dose, typically 10 e·Å−2.
Consequently, the above mentioned first partial remedy against radiation damage cannot be combined with the above mentioned second partial remedy against radiation damage, limiting the effect of the available radiation damage mitigation strategies.
The current method to overcome the radiation damage limit of frozen biological specimens is via so-called single-particle tomography (“Single-particle reconstruction of biological molecules—story in a sample”, Nobel Lecture, Dec. 8, 2017 by Joachim Frank). In this method, images of many similar objects, for example, a single type of purified protein or virus particle, are recorded via transmission electron microscopy (TEM). The objects are embedded in a thin layer of ice and are deposited at random orientations. The electron dose is limited to a value below the damage threshold of 10 electrons per Å2, and as a result, each image exhibits a poor signal-to-noise-ratio (SNR). A three-dimensional structure is then computed by averaging many tens of thousands of images, such to increase the SNR. But in this case, information of the unique structure of single objects, for example, a protein is not available, and information about differences between structural sub-classes is not available when the number of objects in a class is smaller than the number of images of objects that needs to be averaged to achieve the required SNR, which severely limits the study of protein function.
In U.S. Pat. No. 7,659,510 it is proposed to dope ice cooled at the temperature of liquid nitrogen, for example, with ions, for example Na+Cl−, with the purpose of making the ice conductive at liquid nitrogen temperature. The doping changes the bandgap of the insulating ice, and then an electrical potential is placed over the sample such as to charge the sample. Similar approaches are described in US 2012/0220046 A, US 2010/0243482 A and Chih-Yu Chao, Chinese Journal of Physics Vol. 45, No. 6-I, December 2007, 557-578.
However, typically, ion mobility approaches zero in ice at liquid nitrogen temperature. Moreover, amorphous ice does not contain crystalline domains as known in a typical semiconductor, for example, a silicon crystal. Therefore, there is no continuous conductive path as needed for the sample to become electrically conductive.
U.S. Pat. No. 7,413,872 describes a device for preparing specimens for a cryo-electron microscope. In an environmental chamber, there is arranged a holder for a sample or a carrier, and at least one blotting element to which a medium for absorbing liquid is attached. A grid is immersed in a suspension of particles to be examined. As a result of blotting, a film thickness in the openings of the grid is reduced for example from 3-4 μm to 100 nm. The sample is cooled down very quickly so that the fluid becomes vitreous. The specimen is examined by cryo-microscopy for example at −170° C.
Further, U.S. Pat. No. 7,420,184 describes a sample holder for electron microscopy with a temperature switch, so that a sample can switched between a temperature of liquid nitrogen and room temperature.
The present invention provides a method for investigating nanoscale biological specimens, wherein a single nanoscale biological specimen can be investigated at a high spatial resolution in a simple manner. This is achieved, in accordance with the invention, by a method for investigating at least one nanoscale biological specimen, the method comprising
The inventive method suggests preparing the nanoscale biological specimens to be investigated in a thin film on a substrate in accordance with step b), with the thin film having a small average thickness AT, with AT≤30 nm. The nanoscale biological specimens have a maximum diameter MD, with MD≤30 nm.
The small film thickness of the thin film has the advantage that the thin film can be tempered (in general cooled) in step c) to a moderately low temperature, chosen between −1° C. and −100° C., preferably between −1° and −50° C., without the formation of ice crystals which might damage the nanoscale biological specimen.
The thin film tempered to said temperature is subjected to measurement in an electron beam instrument, typically a TEM (transmission electron microscope), in step d); this is why said temperature is also called the measurement temperature MT.
By tempering the thin film in step c) to the moderately low measurement temperature MT, the thin film or its embedding liquid becomes immobilized or stabilized (or at least less mobile and more stable, as compared to the situation at a preparation temperature of step b), which is typically at room temperature). In general, the thin film applied from the embedding liquid is not in liquid state at the measurement temperature MT. First, the thin film can longer resist a vacuum present during step d) in the electron beam instrument; the thin film will exhibit a low and often negligible evaporation (over the measurement time) in the vacuum. Second, diffusion of chemical radicals and free electrons is reduced, which reduces radiation damage (due to “secondary damage”) during the measurement of at least one nanoscale biological specimen under the electron beam irradiation. Third, structural stability is provided to the sample.
Equally important with respect to radiation damage, the measurement temperature MT between −100° C. and −1° C., and preferably between −50° C. and −1° C., is still relatively high, as compared to state-of-the-art investigation at about liquid nitrogen temperature (−196° C.=77K). This relatively high measurement temperature improves electron transfer within a respective biological specimen, and improves electron transfer between a respective biological specimen and the (electrically conductive) substrate and its sample support. Also, the improved electron transfer, and in particular the improved electric conductivity within the immobilized embedding liquid, improves recombination of charge pairs, and accumulation of charges only occurs at relatively high electron fluxes then.
Therefore, with the sample preparation and measurement of the inventive method, both higher electron doses (in particular 10 electrons per Å2 or more, and often of 50 electrons per Å2 or more) and higher electron fluxes (often at 5 electrons per Å2 and per s or more) can be tolerated during measurement of the at least one nanoscale biological specimen, without suffering significant radiation damage. Accordingly, a relatively high spatial resolution in an electron beam measurement (in particular TEM measurement) may be achieved, within a relatively short measurement time or with a relatively small number of objects that need to be averaged to obtain sufficient SNR.
However, note that the electron flux should still be low enough in order to allow charge recombination without charge accumulation. Typically a flux of electrons at or below 50 electrons per Å2 per s and preferentially around 10 electrons per Å2 per s for a typical electron beam energy between 30 and 300 keV used in high-resolution transmission electron microscopy is adequate in the inventive method; note here that lower electron energies may also be used in the range of 50 eV-30 keV as used in other electron beam instruments.
The nanoscale biological specimen to be investigated is typically a protein, but may also be a peptide, a virus particle, a DNA sequence or something else that can be found in a human or animal body or a plant or a fungus.
In order to facilitate preparing the thin film, the substrate is wettable for the embedding liquid at least in the application zone (in particular with a contact angle alpha <90°, preferably <45°). In this way, the embedding liquid spreads itself over the application zone. In order to achieve the low thickness, liquid may be deliberately removed, for example by blotting or evaporation, and/or only a few femtoliters of embedding liquid are initially placed on the substrate (see also below).
Note that measuring at least one nanoscale biological specimen with the electron beam in step d) is typically done in order to examine the structure of the nanoscale biological specimen. The measurement may in particular include phase contrast (including bright field imaging and dark field imaging), diffraction, holography or ptychography, typically in a transmission electron microscope.
An exemplary variant of the inventive method is characterized in that the measurement temperature MT is less than a freezing temperature of the embedding liquid in bulk. The thin film of embedding liquid is reliably immobilized or stabilized by choosing a measurement temperature below the freezing point of the embedding liquid in bulk. Note that due to the low film thickness, the thin film can be prevented from forming ice crystals, even below the freezing point of the embedding liquid in bulk. The measurement temperature may be kept above a freezing temperature of the thin film of the embedding liquid (as applicable), the latter temperature may be lower than the freezing temperature of the embedding liquid in bulk due to boundary layer effects. Note that the thin film of embedding liquid may, in particular, assume a gel-like state or an amorphous state or a cured state or some different state at the measurement temperature. The freezing temperature of the embedding liquid in bulk may easily be determined by experiment, for example, with a macroscopic sample of the embedding liquid in a test tube and slowly lowering the test tube temperature; the freezing of the embedding liquid is well visible with the naked eye or can be detected with a sensor easily. If desired, the film may also be cooled below a freezing temperature of the thin film of the embedding liquid, and here, AT should be smaller than MD to avoid ice crystal formation. Whether ice crystals are formed can easily be checked via a transmission electron microscopy image, and AT decreased as needed.
That ice crystal formation can be prevented for a very thin layer of liquid follows from the ice nucleation rate. For example, for pure water, the nucleation rate for ice nucleation is 1014 m−3s−1, but the latter value drops eight orders of magnitude down to 106 m−3s−1 for a nanodroplet of 2.4 nm in radius, thus obtaining a nanoscale droplet of supercooled water. Ice nucleation is altogether avoided for nanodroplets placed on a flat surface. Thus, ice crystals would not form in a water layer of a thickness of about 3 nm at a surface of a graphene film as sample support. In practice, the liquid layer can be somewhat thicker (for example, 5 nm) when using not pure water but saline, for example, water mixed with, for example, NaCl, which reduces the nucleation for crystallization compared to ice, so that a thicker layer still does not crystallize.
By tempering the thin film of embedding liquid and the nanoscale biological specimens in the embedding liquid on the substrate (or simply tempering the sample) to the measurement temperature below the freezing temperature of the embedding liquid in bulk, diffusion of chemical radicals and free electrons is reduced, and the thin film of embedding liquid is prevented from evaporating in the vacuum of the electron microscope.
In another variant, there is chosen AT≤MD. This helps to avoid the formation of ice crystals in the embedding liquid of the thin film when tempering the thin film to the measurement temperature.
A further variant provides that in step d), an electron beam energy of the electron beam is between 50 eV and 300 keV,
In yet another variant, step a) and step b) are done at a preparation temperature PT, with PT>0° C. This simplifies performing steps a) and b). Note that the preparation temperature is typically in a range of 0° C.<PT≤40° C., and often with PT at room temperature (20° C.).
Also provided is a variant including during step b), placing an initial large amount of embedding liquid on the substrate, with the large amount being larger than needed for the thin film, and then reducing the amount of embedding liquid on the substrate until only the thin film remains. This process is particularly simple to put into practice, and application zones can be loaded with embedding liquid fast and reliably. Further, if desired, it is easy to load plenty of application zones (in particular 5 or more, or even 20 or more) at a time with embedding liquid.
A further development of this variant is characterized in that reducing the amount of embedding liquid includes placing a liquid absorbing medium in contact with the embedding liquid,
Advantageous is further a variant wherein during step b), for preparing the thin film of embedding liquid in the application zone having an application zone area AZA, an initial small volume SV of embedding liquid is placed on the substrate in the application zone, with SV≤AT*AZA. When the initial small volume of the embedding liquid, after having been applied to the application zone, spreads over the application zone of the substrate (which is wettable for the embedding liquid there), the desired average thickness of the thin film can be achieved “by itself” in this case. Note that in this variant, reducing the amount of applied embedding liquid in step b) is generally not necessary; however sometimes in this variant some evaporation of embedding liquid is performed, too.
Further preferred is a variant wherein in step b) an initial amount of embedding liquid is placed on the substrate in the application zone by first dipping an application tip, in particular an atomic force microscopy (AFM) tip, into a supply pool of the embedding liquid containing the nanoscale biological specimens, and then touching the application zone with the application tip. Touching the application zone with the tip includes, in the sense of the present invention, both that the application tip immediately touches the application zone, or a drop of embedding liquid adhering to the tip touches the application zone. This is a practical procedure to place a small amount of embedding liquid on the substrate or an application zone of it, in particular a small volume SV of embedding liquid as mentioned above. The material of the application tip should be wettable for the embedding liquid (with a contact angle beta<90°, preferably beta<45°), but preferably not as good as the substrate in the application zone (that is preferably alpha<beta, with alpha being the contact angle of the embedding liquid on the substrate in the application zone, and beta being the contact angle of the embedding liquid on the tip material). The amount of liquid stuck to the application tip can be adjusted via the dipping depth of the application tip into the surface of the embedding liquid in the supply pool.
Also provided is a variant wherein in step b) an initial amount of embedding liquid is placed on the substrate in the application zone with an application tip having a microchannel through which the initial amount of the embedding liquid is discharged. This is also a practical way to place a small amount of embedding liquid on the substrate or its application zone, in particular a small volume SV as mentioned above.
Further preferred is a variant wherein during step b), an initial volume of embedding liquid placed on the substrate is between 1 fL and 2 μL. Note that a typical initial volume of embedding liquid is 0.2 μL-2 μL in case of subsequently reducing the amount of embedding liquid, for example, by blotting or evaporation. Typical amounts of embedding liquid corresponding to a small volume SV as mentioned above, according to the invention, are in general on the order of 1 fL to 100 fL (for example, for a typical 3 mm TEM grid with a mesh of 200, corresponding to a typical TEM grid window of length 60 μm, has AZA=3600 μm2, and AT=10 nm, then results SV=36 fL (3.6*10−17 m3).
In an advantageous variant, experimental parameters applied during step b) in order to prepare the thin film with the average thickness AT have been determined in advance in calibration experiments, in particular wherein the experimental parameters include at least one of: temperature of an atmosphere surrounding the substrate, pressure of an atmosphere surrounding the substrate, humidity of an atmosphere surrounding the substrate, composition of an atmosphere surrounding the substrate, temperature of the substrate, type of a liquid absorbing medium, contact time with a liquid absorbing medium, contact pressure to a liquid absorbing medium, evaporation time, initial large amount of embedding liquid placed on the substrate, initial small volume SV of embedding liquid placed on the substrate in an application zone, initial amount of embedding liquid placed on the substrate, initial volume of embedding liquid placed on the substrate, touchdown pressure of an application tip, discharging flow of embedding liquid through a microchannel. By means of calibration experiments, samples may be prepared in a reproducible way from the nanoscale biological specimens, and the samples can be investigated in a standardized and optimized procedure, wherein the results of different samples can be compared with high information content. Note that the calibration experiments typically each include determining an obtained average thickness of the obtained thin film, which is typically measured by electron microscopy in a frozen (or immobilized) state. Experimental parameters in the calibration experiments may be adapted until a desired average film thickness is obtained.
Particularly advantageous is a variant wherein the embedding liquid is chosen as water with added salt, for example wherein the added salt is NaCl, particularly wherein the added salt is NaCl with a concentration of 10 mM. By adding salt, a high electrical conductivity of the thin film of (immobilized) embedding liquid may be obtained.
In yet another variant, the substrate is a graphene foil or carbon foil. Graphene foils or carbon foils provide a good electric conductivity for providing a good electron transfer from the thin film (or the specimen within the thin film) via the substrate to a substrate holder/sample holder and finally to earth, and therefore help to minimize radiation damage. Further, the graphene foil or carbon foil is well suited for TEM investigation, and in particular for transmission by an electron beam without distorting the TEM image of the imaged specimen too much. Note that the graphene foil or carbon foil is typically very thin, such as with a thickness of 5 nm or less, in particular wherein the graphene foil may be a monoatomic film. Typically, such a very thin graphene or carbon foil is supported by a stronger film having holes (for example a holey film with a thickness of 50 nm or more), such as a carbon film with holes or a silicon nitride membrane with holes.
Further advantageous is a variant wherein before step b) the substrate is treated with a plasma at least in the area of one or more application zones, preferably wherein the plasma is formed from oxygen and/or argon, most preferably wherein the plasma is formed from oxygen and/or argon at a pressure of 10−3 mbar. The plasma can make the substrate hydrophilic, at least in the application zone/zones of the substrate where the embedding liquid will be applied. Further, the plasma treatment cleans the substrate surface.
In an advantageous variant, after step b) and before step c), the thin film of the embedding liquid is covered with a covering substrate, in particular wherein the covering substrate is a graphene foil or a carbon foil. The covering substrate can help to obtain small liquid pockets and/or minimize evaporation of (a fraction of) the embedding liquid of the thin film in the vacuum of the electron beam instrument in step d). Further, in case of an electrically conductive cover substrate, the covering substrate can improve electron transfer from the thin film to the substrate holder/sample holder and finally to earth, and thus helps to minimize radiation damage.
In another advantageous variant, the embedding liquid contains electrically conductive polymers and/or electrically conductive proteins. The electrically conductive polymers and/or electrically conductive proteins improve the electron transfer from the nanoscale biological specimen investigated to the substrate and the substrate holder/sample holder and finally to earth, and thus help to minimize radiation damage in step d). Note that an electrically conductive polymer (or protein) can at the same time function as a linker molecule.
Preferred is a variant wherein the embedding liquid contains negative stain, in particular wherein the negative stain is chosen as α-D-Glucose 1-phosphate dipotassium salt hydrate, applied preferably with a concentration of 2 mM. The chemical stain may preserve the structure of the nanoscale biological specimen, in particular by maintaining a molecular layer of water around the specimen. The stain may further provide electrical conductivity to the substrate or the substrate holder/sample holder and finally to earth, thus minimizing radiation damage. By applying a negative stain, a thin film stable in the vacuum of the electron beam instrument can be obtained without a cover substrate. Typically, the thin film is gel-like at measurement temperature MT. In this variant, typically the amount of the embedding liquid is reduced applying evaporation in step b), in particular at least at the end of the reducing procedure.
Also provided is a variant wherein in step c), the thin film on the substrate is first cooled down to a low temperature LT, with LT≤−196° C., and then warmed up again to the measuring temperature MT. Typically the first cooling down is done at a high cooling rate, such as within 240 s or less, or even 120 s or less (assuming a preparation temperature PT between 0° C. and 40° C.). By the first cooling down (at a cooling rate high enough), the embedding liquid of the thin film can be transformed into amorphous ice, without forming ice crystals. The later warming up to the measurement temperature MT, even with MT above the glass transition temperature of the embedding liquid, will not lead to crystallization of the embedding liquid of the thin film due to its low film thickness. Note that the low temperature LT is at or below the boiling temperature of liquid nitrogen (LN2) at 1 bar.
In an alternative variant, during steps b) through d) the temperature of the thin film is always kept above or at the measurement temperature MT. This simplifies and accelerates the sample preparation procedure. In particular, only moderately low temperatures need to be provided for, which in general can be provided without handling of cryogen liquids such as LN2 or LHe and with inexpensive equipment.
Further preferred is a variant wherein the substrate is placed on a TEM grid, with the TEM grid having a plurality of crossing grid bars defining grid windows between the crossing grid bars. Each grid window may be used as an application zone. The TEM grid simplifies sample handling in a TEM. Further the TEM grid, which is typically made of copper, provides good electrical conduction paths.
A preferred further development of the above variant provides that during step b), embedding liquid is initially applied as a drop spreading over a plurality of grid windows, in particular all grid windows of the TEM grid. Typically, some embedding liquid is removed then, for example by blotting, preferably wherein the TEM grid is pressed onto a filter paper. This is particularly fast and reliably wets all grid windows or respective application zones with the embedding liquid.
A particular variant provides that the substrate is placed on a TEM grid, with the TEM grid having a plurality of crossing grid bars defining grid windows between the crossing grid bars, and that during step b), embedding liquid is applied between grid bars in at least one grid window, in particular wherein embedding liquid is applied deliberately between the grid bars in the at least one grid window. The grid bars can form a (maximum) limitation of the application zones. The grid windows or the respective application zones are easy to spot in the electron beam instrument, and accordingly the nanoscale biological specimens there are easy to investigate. The TEM grid, which is typically made of copper, further provides good electrical conduction paths along the grid bars. In case of application of small volumes SV of embedding liquid, a respective small volume SV of embedding liquid can easily be placed onto a respective grid window, in particular centrally.
In another variant, the substrate comprises at least one local coating defining at least one application zone, wherein the local coating provides that the substrate is wettable for the embedding liquid in the application zone. By means of the local coating, the application zone can be well defined, in particular its area or extension. Further, originally unsuitable substrate materials for being wetted by the embedding liquid can be made suitable by the local coating. This enlarges the scope of substrate materials for the invention. The local coating can in particular be made from linker molecules (see below).
An advantageous variant provides
In a further development of the above variant, the linker molecules contain an electrically conductive chain, in particular wherein the electrically conductive chain is based on carbon atoms with unsaturated carbon bonds or a carotenoid or a chain of carbon atoms with oxygen atoms. In this way, the electron transfer from the thin film or the nanoscale biological specimens to the substrate may be improved, and thus radiation damage may be minimized.
Also provided is a further development of the above variant wherein the linker molecules comprise molecules for specially binding to the nanoscale biological specimens to be investigated, in particular wherein the nanoscale biological specimens comprise a protein, an antibody, a binding peptide, a DNA sequence, a ligand for a receptor, a chelated metal, or a histidine. Said molecules have a high affinity for binding to the nanoscale biological specimens to be investigated are also called ligand molecules here. Various compounds (containing a ligand molecule and a nanoscale biological specimen) exist depending on the biological substance (that is nanoscale biological specimen) to be investigated. For example, linker molecules may contain biotin that is a ligand for the protein streptavidin, and the compound with attached linker molecule is Amine-PEG2-Biotin.
Also within the scope of the present invention is a method for determining a flux threshold of electrons for measuring of a nanoscale biological specimen with an electron beam, comprising the following steps:
In the course of the method, it is determined what electron flux the chosen nanoscale biological specimens can just stand in the electron beam instrument at the chosen electron dose, before the molecular structure becomes distorted as visible from, for example, changes of the original TEM image or diffraction pattern. Note that several series of measurements according to steps α) through γ) can be performed for corresponding several different electron doses. Investigations of the nanoscale biological specimen just below the determined flux threshold will result in an optimum spatial resolution of sample investigation in the electron beam instrument. The flux at which the structure becomes distorted may be identified, for example, by a disappearing of diffraction peaks at higher spatial frequencies, or the TEM image becoming fainter, indicating mass loss of the (diffracting) protein. Aiming for a specific spatial resolution, for example, 4 Å, structural damage is identified when diffraction signals above a certain spatial frequency disappear, for example, ¼ Å aiming for a spatial resolution of 4 Å. Structural damage may be concluded when the relative measure of mass reduces by, for example, 20% or more compared to the first image in an image series. A relative measure of mass is the integrated signal of image pixels at the location of the protein from which is subtracted the integrated signal of image pixels surrounding the protein. A good flux of electrons to start with is 0.5 electrons per Å2 and per s. Flux may be doubled from step to step, for example.
In a variant of the above method for determining the flux threshold, the preselected electron dose is at least 50 electrons per Å2, preferably at least 200 electrons per Å2. These electron doses can be tolerated well in accordance with the invention without significant radiation damage, and lead in general to a good resolution of the nanoscale biological specimens investigated.
Also provided is an inventive method for investigating at least one nanoscale biological specimen as described above, wherein in step d) the nanoscale biological specimen is exposed to an electron dose and a flux of electrons, wherein the flux of electrons is below a flux threshold determined for that electron dose as preselected electron dose in an inventive method for determining a flux threshold as described above. Then the investigation of the nanoscale biological specimens leads to excellent spatial resolution, in particular TEM image resolution, often at 0.5 nm or better (lower), and accurate structure analysis is available within short time, without significant radiation damage to the sample. A typical flux threshold for reduced radiation damage is 10 electrons per Å2 and per s.
Also within the scope of the present invention is a sample for investigating at least one nanoscale biological specimen in an electron beam instrument, with the sample comprising:
Further within the scope of the present invention is the use of a sample preparation apparatus in an inventive method for investigating at least one nanoscale biological specimen as described above, wherein the sample preparation apparatus is adapted to automatically perform at least step b), and preferably also step c). With the sample preparation apparatus used according to the invention, simple sample preparation as well as fast and accurate sample investigation may be achieved.
Further advantages can be extracted from the description and the enclosed drawing. The features mentioned above and below can be used in accordance with the invention either individually or collectively in any combination. The embodiments mentioned are not to be understood as exhaustive enumeration but rather have exemplary character for the description of the invention.
The invention is shown in the drawings.
The present invention provides, as can be seen from the schematic diagram of
In one aspect of this invention it is proposed to mount a nanoscale organic object, for example, a protein or a virus particle, on a sample support in such a way that electron transfer between the object and the sample support is possible, wherein first it is suggested to apply the object within a liquid (also called embedding liquid), to then reduce the liquid thickness to a very thin liquid layer (also called thin film) at the sample support, and to then freeze (cool) the sample to a temperature below freezing point of the embedding liquid in bulk, whereby the thickness of the liquid layer is below the minimum thickness required for ice crystal formation, and the temperate of said layer is sufficiently high such that electron transfer still takes place. The sample is then examined in the vacuum environment of an electron beam instrument at an electron flux below the threshold for radiation damage by ionization, whereby the total dose can be increased above 10 e Å2.
The sample support is mounted in a sample holder with means for cooling and measuring the temperature, which is commercially available as a so-called cryo holder for TEM specimens. The sample is prepared from an ultra-thin layer of liquid embedding the object immobilized at a surface of a thin membrane, for example, of graphene or carbon. The object is, for example, a protein.
The preparation protocol, the wetting behavior of the thin membrane, and the chemical content and thickness of the liquid layer are chosen such that 1) the liquid forms a thin layer over the object upon exposure of the sample to vacuum such to protect the structure of the object, 2) the liquid provides electrical conductance between object and sample support, and 3) ice crystallization does not take place upon freezing even though the temperature is above the glass transition temperature.
In one possible embodiment of the method, the sample is prepared by placing a droplet of 0.2 μl volume of saline with dissolved protein, for example, hemocyanin on a supporting membrane consisting of a single layer of graphene on a holey carbon film mounted in a standard 3 mm gold grid for TEM. After a waiting time of a few minutes, a major fraction of excess liquid is removed by blotting with a filter paper until the surface appears dry but with a shiny (light reflecting) appearance, indicating the presence of a thin liquid film. The sample is then covered with a thin membrane consisting of a single layer of graphene.
The sample, after having been covered with the second graphene layer, is inserted into the vacuum chamber of the electron beam instrument by means of a TEM specimen holder. Upon exposure to vacuum, some excess liquid evaporates but a layer of liquid remains enclosed between both graphene membranes. Note that after the final thickness of the thin film has been reached, the sample is cooled to the measuring temperature, so the embedding liquid becomes immobilized.
In particular, blotting and/or evaporation is done under calibrated conditions, preferably such that the (average) liquid layer thickness becomes thinner than the size of the protein and forms a surface layer at the supporting thin membrane, whereby the layer becomes so thin that ice crystals do not form upon freezing. Calibrated conditions involve humidity of surrounding air, time of blotting, pressure of blotting, type of liquid adsorbing medium. Calibration is done by executing the blotting for a set of conditions and examining the sample via electron microscopy to measure the thickness of the sample and inspect for the presence of ice crystals. Considering pure water, the nucleation rate for ice nucleation is 1014 m−3s−1. A nanodroplet of 2.4 nm in radius would exhibit a reduced ice nucleation rate of 106 m−3s−1 (T. Li et al, Nat. Commun. 4 (2013) 1887-1-6). Ice nucleation is altogether avoided for nanodroplets placed on a flat surface (Xiang-Xiong Zhang et al., J. Chem. Phys. 141 (2014) 124709). Thus, ice crystals would not form in a water layer of a thickness of about 3 nm at a surface of a graphene film as sample support. In practice, the liquid layer can be somewhat thicker (e.g. 5 nm) if the embedding liquid applies not pure water but saline, for example, water mixed with Na+Cl−, which reduces the nucleation for crystallization compared to ice, so that a thicker layer still does not crystallize. In addition, the ice nucleation rate is reduced since the liquid is enclosed between two sheets of graphene providing a mechanical barrier against ice crystal formation.
The sample is then slowly brought to a temperature (measurement temperature) just below the freezing point of the sample, that is below the freezing point of the embedding liquid in bulk. Note that this state is sometimes referred here to as “frozen”, even though no ice crystals form due to the small thickness of the thin film formed by the embedding liquid on the substrate.
The temperature (measurement temperature) of the tempered/cooled sample needs to be chosen according to three criteria. 1) The liquid surrounding the object has to be immobilized, typically by choosing the measurement temperature below the freezing point of the embedding liquid in bulk. 2) The temperature needs to be sufficiently high to allow electron transfer to take place in the object of interest, for example, a protein. 3) The temperature needs to be sufficiently high to allow electron transfer between object and sample support.
To fulfill criteria 1, a temperature below 273° K (=0° C.) would be needed to ensure freezing of the thin aqueous liquid layer at ambient pressure. The temperature depends on the presence and concentration of saline. For a liquid containing a physiological concentration of NaCl of 100 mol/l (6% by weight), the freezing temperature would be about 269° K (Wikipedia page saline water, https://en.wikipedia.org/wiki/Saline_water, downloaded on 22 Sep. 2023). To allow for some tolerance in the exact freezing point, a temperature (measurement temperature MT) of 260° K may be chosen. The latter temperature would still fulfill criteria 2 for typical values of driving force ΔG0=−1.5 eV, and λ=0.75 eV since it would allow electron transfer in the protein, for which the electron transfer rate would reduce by a factor of 2 only compared to room temperature.
For charge transfer between the nanoscale biological specimen and sample support, one method is to mount the specimen as close as possible to the sample support, and the temperature is chosen such that electron transfer between the object and the sample support takes place. As an example, one may consider a spherical protein of a diameter of 50 Å receiving an electron flux of 10 e·Å−2s−1 from the electron beam, and thus a current of 10 e·Å−2s−1×π(50 Å/2)2=2×104 e·s−1, corresponding to about 3×10−15 A. As an example, one may assume an ideal case of a free electron created in the protein's surface at a distance of 10 Å from the supporting layer providing electrical neutral. The electron transfer rate between a donor and acceptor spaced at this distance in a protein is of the order of 109 s−1 (C. C. Moser et al., Nature 355 (1992) 796), sufficient for charge compensation by 5 orders of magnitude, and so criterion 3 would be fulfilled.
To further enhance electrical conductivity of the sample, a second method for charge transfer between object and sample support is by including ions in the liquid surrounding the object. The measurement temperature is then chosen such that the thin film with ions is still electrically conductive, fulfilling criterion 3. Doping ice with NaCl of 7 μM increases the conductivity towards σ=2×10−5 Ω−1m−1 at freezing point compared to pure ice (G. W. Gross et al., J. Glaciol. 21 (1978) 143, or see also Stillman and Grimm, Lunar and Planetary Science XXXIX (2008) 2277).
The effect of doping on charge transfer can be estimated as follows. Suppose a protein is surrounded by a layer of mixed ice of thickness t=20 Å and positioned on a conductive support at electrical neutral. Suppose the current from the protein to the support would take the shortest pathway and pass through an area A=π(50 Å/2)2. Thus, a resistive path is obtained with a resistance R=t/(Aσ). For ice mixed with NaCl, R=5×1012 Q. The emission of secondary electrons from the sample upon electron beam irradiation can be considered as electrical source that is connected via the resistance presented by the mixed ice layer between the protein and the support connected to electrical neutral. With I=3×10−15 A at electron flux of 10 e·Å−2s−1 received by the protein, the potential drop over the mixed ice layer amounts to U=IR=1.6×10−2 V. The corresponding electric field strength would then reach F=8×106 Vm−1, which is below the dielectric strength of ice of 8×107 Vm−1 (T. Kohno et al., IEEE Transactions on Electrical Insulation, EI-15 (1980) 27) and so the electrical circuit would function. If on the other hand, no doping was used, the system would reach F=109 Vm−1, which would be above the dielectric strength, and so sample damage would occur, typically by the generation of bubbles of hydrogen gas.
The thus obtained sample conditions provide the object embedded in a thin frozen immobilizing layer while being electrically conductive such to optimize radiation damage mitigation.
For obtaining structural information of the sample, electron beam exposure in transmission mode is preferred, such that an electron dose efficient contrast mechanisms can be applied, for example, phase contrast, diffraction, holography, or ptychography.
The inventive method largely reduces the problem of electron beam damage for a broad class of radiation sensitive samples such as proteins, virus particles, (bio)minerals, polymers, soft-matter nanoparticles, electrolytes, etc. With an increased tolerance to radiation damage, it is possible to achieve a better spatial resolution for these samples compared to state-of-the-art.
Various other embodiments of this method are possible of which examples are described in the following.
The blotted sample may be frozen (cooled) slowly to a temperature of 260° K in an atmosphere with low water vapor content provided by dry nitrogen gas such that freezing does not lead to condensation of ice on the sample. The sample is then inserted into vacuum.
The blotted sample may, in an alternative, be frozen (cooled) sufficiently fast and to a sufficient low temperature (also referred to as low temperature LT, typically with LT≤−196° C.) to obtain amorphous ice, and the sample is then inserted into vacuum and carefully heated to a temperature (measurement temperature MT) of, for example, 260° K. The blotting step prior to freezing should be executed in such way as described above, such that a liquid layer of only a few nanometers thickness remains, and thus to prevent ice crystal formation once the sample is heated above the glass transition temperature.
The sample support may comprise of carbon, graphene oxide, silicon nitride, or any other material that provides a mechanically strong support, and provides charge transfer for the conductive substrate. In case transmission of the electron beam is desired, the substrate should be thin enough to allow electron transmission. Thin enough means that the thickness of the substrate material is at least a factor of ten smaller than the mean free path length of elastic electron scattering at the set electron energy, which amounts to approximately 0.12 μm for carbon at 200 keV beam energy (K. Iakoubovskii and K. Mitsuishi, J. Phys. Condens. Matter 21 (2009) 155402). The substrate does not necessarily need to be a flat sheet but may also be curved, for example, a carbon nanotube, or tip shaped.
The method can also be applied correspondingly for thick substrates, in which case backscattered signals or amplified secondary emission signals are used for detection, for example, in scanning electron microscopy.
In yet another embodiment, the substrate is first placed in a controlled environment such as a chamber with pure argon, or a vacuum chamber, and the sample is placed on the substrate in this chamber such to reduce contamination as much as possible.
In another embodiment, instead of saline, the object is surrounded by a chemical stain preserving the structure of the biological specimen, for example, a protein or virus particle, and electrical conductance is provided via the presence of ions. An example is a stain of low atomic number materials such as provided by applying water with dissolved glucose-1-phosphate potassium salt. Further, the sample is not covered by a graphene sheet. Upon exposure to vacuum, some excess liquid evaporates but the liquid layer transitions to a gel protecting the structure of the object from being damaged in the vacuum, so that a molecular layer of water remains stable at the protein (see
The embedding liquid used may contain conductive polymers or conductive proteins providing electron transmission pathways.
The embedding liquid used may contain heavy water for additional radiation protection or an ionic liquid with very low vapor pressure for protection of the specimen against evaporation. Another type of liquid than water or saline can be used, for example, an oil or an electrolyte.
The sample preparation steps are preferably carried out in an automated way by means of a sample preparation apparatus 25, as shown in
A subset of the automated sample preparation for the present invention can be accomplished using commercially available equipment, for example using an apparatus described in US 2004/157284 (U.S. Pat. No. 7,413,872).
Further automation may be implemented including automated loading of the 3 mm grid in the tip of a specimen holder for electron microscopy. The specimen holder is first mounted in an adapter in the sample preparation apparatus. A mechanical manipulator positions the 3 mm grid in the tip.
The sample preparation apparatus may also contain one or more additional micro pipettes or micro-channeled cantilevers for adding fluids, such as a chemical stain or a liquid droplet containing conductive polymer.
A further addition may be the enclosure of the sample in a controlled environment of, for example, pure argon gas, or the equipment may also be mounted in a vacuum chamber.
In another embodiment, the apparatus contains a freezing device (cooling device), which can be of a type of a rapid plunge freezing.
The sample is placed in an electron beam instrument for investigation of the nanoscale structure of the specimen. Various types of instruments that can be used in accordance with the invention are commercially available, such as a transmission electron microscope, a scanning transmission electron microscope, a scanning electron microscope, or an electron diffraction instrument. The substrate holding the specimen that is mounted, for example, on a 3 mm grid, is placed in the vacuum chamber of the electron beam instrument, for example, by means of a cooling specimen holder, for example, a cryo transfer holder as described in US 2019/131106 or in US 2012/024086. The temperature may be controlled also using a modified electron beam instrument as described, for example, in US 2007/252090. In case a scanning electron microscope is used, the specimen can be mounted on a cooling Peltier stage.
Analogously to the TEM images of dose series described above, for a particular preselected electron dose, TEM images of a flux series can be recorded in accordance with the invention. In each TEM image, a different electron flux is applied, wherein the preselected electron dose is distributed over different irradiation/measurement times. For each flux, a different position on the sample (and therefore a different nanoscale biological specimen/protein) is chosen. A typical flux series starts with a low electron flux, such as 1 e/(Å2*s), and flux is doubled from image to image in the series, such as to 2 e/(Å2*s), 4 e/(Å2*s), 8 e/(Å2*s) and so on. The image of the sample measured with the highest flux that maintained its protein structure (that is the structure is not yet distorted) is determined. Note that the series can be ended once the first structure distortion has been identified.
Structure distortion can be determined for example using the 0.8 normalized intensity criterion described above (see
For future measurements for samples of this type and the preselected electron dose, said highest flux having maintained the protein structure can be chosen, to efficiently obtain good quality TEM images.
Note that local coatings with linker molecules can be used to define application zones 22 for thin films 2 to be prepared, in accordance with the invention, as shown in
| Number | Date | Country | Kind |
|---|---|---|---|
| EP23203613.7 | Oct 2023 | EP | regional |
