The present protocol describes transmission electron microscope (TEM) modifications with a light illumination system, the fabrication of liquid cells, and in situ TEM observations of light-induced interactions between bacterial cells and a photosensitizer. The sample preparation methods, electron beam damage, and imaging are also discussed.
The current protocol describes the modifications of the transmission electron microscope (TEM) setup for in situ light-induced observations. A glass optical fiber inserted into the electron column above the objective lens polepiece, and a laser, an adjustable light source, was used to fabricate the device. After the illuminator has been calibrated using an external measuring system, it allows one to adjust the intensity of the lighting to the needs of the observed process. This lighting system was utilized to image antimicrobial photodynamic therapy phenomena, which are currently the subject of intense research. The sample was prepared by spotting a suspension of bacteria on a carbon, graphene, or silicon nitride substrate, blotting the excess solution, spotting the photosensitizer solution, blotting the excess liquid again, and then assembling the liquid cell with a second substrate or graphene film. The process of the imaging experiment itself includes choosing the right place for observation with the use of low magnification and a minimum dose of electrons, and then cyclical activation of the light source to capture subsequent images at specified intervals with the minimum amount of electrons necessary. The electron dose of each exposure and the time and intensity of lighting used need to be carefully recorded due to the complexity of the observed phenomena as, at the same time, the process is both light- and electron-driven. After the actual experiment is performed, additional control observations must be made, in which the same doses of electrons are used but without additional light influence and smaller doses of electrons are used for higher doses of light. This makes it possible to distinguish light-induced microstructural effects from those caused by electrons in both the fields of life and materials science.
Light-induced phenomena in high resolution are interesting in many fields such as nanoengineering1,2,3, catalysis4,5, and biophotonics6. Some original designs that allow such experiments can be found in the literature,including modifications of the sample holders1,4,7,8,9 and the optical fiber attached to the microscope10,11.
The combination of light illumination, a liquid environment, and transmission electron microscopy (TEM) gives a great opportunity for detailed, dynamical studies of photo-induced processes. However, the high-vacuum condition inside the microscope is rather unfavorable for many liquids, especially water solutions. Liquid encapsulation, which protects it from the environment, can be achieved using a few techniques based mainly on graphene12, silicon nitride13, or carbon14 substrates. In addition to research in materials science2, so-called liquid cells offer possibilities for conducting unconventional microscopic observations on biological specimens near their native conditions15. Such observations are extremely demanding, especially for living microorganisms such as bacterial cells. The electron beam as ionizing radiation causes irreversible damage to the hydrated specimens, so the electron dose must be specified16. This is necessary to minimize unfavorable effects, control the damage, and avoid confusing artifacts. The optimal maximum electron dose that allows observations of living cells is still a questionable topic16, but the dose of 30 e−/nm2 appears to be the threshold value, at least for bacteria17.
Some of the subjects of interest for such microscopic studies are processes during antimicrobial photodynamic therapy (APDT)18. In short, the therapy proceeds as follows. The bacterial cells are surrounded by the photosensitive liquid called photosensitizer. When light illumination is given at a specific wavelength, the cytotoxic reactive oxygen species (ROS) are generated from energy or charge transfer from the excited photosensitizer molecules to the oxygen naturally present in the solution. Pathogens exposed to ROS are quickly inactivated with very high efficiency, with no side effects19. The response to the therapy varies for distinct microbes – for example, the impact of the same photosensitizer may be quite different for Gram-positive and Gram-negative bacteria20. In general, it has been established that the main target of ROS is the outer structures of cells, where damage results in functional disorders of the cell membrane and, consequently, lead to the death of bacteria21,22. However, damage to nucleic acids and proteinscan also be considered a cause of inactivation18, so it is still unknown which cell structures are the main targets during this process19. A deeper understanding of the damaging processes could help to improve this definitive therapy. Compared to the light microscopy methods used in APDT research23, TEM techniques give more possibilities to look at the APDT mechanism with higher resolution and magnification24. TEM has already been successfully used for cell observation during ongoing therapy, which allowed us to study Gram-positive bacteriadamage and describe the changesoccurring within the cell wall in detail6,25.
The current protocol presents a suitable experimental setup for high-resolution imaging of light-induced bacteria inactivation using TEM, which requires a proper light illumination system, the encapsulation of cells with a liquid, and strict electron dose control. The bacteria used for the observation was Staphylococcus aureus, and a methylene blue solution was used as a photosensitizer. The special light illumination setup comprises a tunable semiconductor laser connected directly to the microscope column using the light fiber. This design provides uniform irradiation throughout the sample because of the almost-parallel placement of the optical fiber to the microscope axis. Monochromatic light of high intensity generated by the laser can then be used to study various photochemical effects. The light used in the experiment had a wavelength equal to 660 nm because, in the visible region, methylene blue has absorption peaks at 613 nm and 664 nm26. The protocol for liquid encapsulation is based on carbon substrates, which makes the procedure quick and uncomplicated. Finally, a method for low-dose in situ TEM observation of cells in liquid is presented. The difficulties regarding sample preparation, the electron dose effects on the sensitive specimen, and reasonable image interpretation are discussed.
1. Transmission electron microscope modification
2. Encapsulation of bacteria with the photosensitizer
3. In situ TEM observations of bacterial cells in liquid-electron dose effect
4. In situ TEM observations of light-induced processes between bacterial cells and photosensitizer
The experimental setup was designed to observe the light-induced processes occurring in a liquid with high resolution. The competitive analysis of the images allowed us to distinguish the damage caused by the electron beam from changes related to the photochemical reaction. The effect of the electron beam on the sensitive specimen (in this case, the cells encapsulated with liquid) was visible all over the irradiated area as the electrons penetrated through it uniformly. Of course, the effect depends on the electron dose, so more significant changes in the sample morphology can be observed faster for a more focused beam. Compared to this severe damage, the actual destructive effects of photodynamic inactivation occurred in specific parts of cells only in areas surrounded by the illuminated photosensitizer.
In the example of bacteria imaging, and according to the literature, the dose of 30 e−/nm2 seems to be the lethal threshold value for bacteria17, which was highly exceeded in the authors' previous studies6,25. However, the visible changes in the cell were not noticed before reaching the electron dose of 6000 e−/nm2. The results showed a notable degradation of the outer layer of the cell caused by the reactive oxygen species generated by light irradiation of the photosensitizer after 1 min (Figure 1). Further light illumination made the changes more visible (Figure 1D,G).
The encapsulation of the cells between the carbon substrates was sufficient to perform constant observations for at least 1 h. Proper observation spots with enough liquid can be easily found at low magnification, as these areas appear darker and blurred (indicated by the frames in Figure 2A). The liquid-covered bacteria are less visible than the dry ones but can still be distinguished (Figure 2B). It is useful to look at the liquid boundary during the imaging, and its movement can be easily detected, which implies that the chosen area might be unsuitable and unstable (Figure 3). Another issue during observations is the evaporation of the water and precipitation of the solution, which can occur in random areas of the sample, including the areas surrounding the bacteria (Figure 3B,C).
Figure 1: Exemplary results of the photodynamic treatment of bacteria. (A) The scheme of the liquid cell containing bacterial cells and methylene blue. (B) The cell before light illumination. (C) The cell after illumination for 1 min. (D) The cell after illumination for 10 min. (E–G) The magnified areas from (B–D). The electron dose was equal to 1600 e−/nm2, 2500 e−/nm2, and 6000 e−/nm2, respectively. The figure is reproduced from Żak et al.25. Please click here to view a larger version of this figure.
Figure 2: TEM images of S. aureus encapsulated with methylene blue between two carbon films. (A) Low magnification image presents both areas with a high amount of liquid (indicated by the yellow frames) and dry spots. A quick examination of the sample at this magnification gives information on whether the encapsulation was successful or not. (B) The liquid boundary is clearly visible in the magnified area, and the bacteria covered with the photosensitizer can be distinguished. Please click here to view a larger version of this figure.
Figure 3: TEM images of unfavorable effects that may occur during imaging. (A) In the beginning, the cell was uniformly surrounded by the liquid, which covered most of the observation area. (B) Based on the constant electron beam irradiation, the movement and evaporation of the liquid and the start of the precipitation from the solution can be noticed. (C) The continuation of the processes leads to irreversible movement, foaming, and disintegration of the photosensitizer, making further observations impossible. Please click here to view a larger version of this figure.
Figure 4: The unfavorable effects leading to wrong conclusions. (A) The crystallized liquid. (B) The contamination settled on the cell. (C) Foaming caused by the electron beam. Please click here to view a larger version of this figure.
The installation and commissioning of the illuminator require basic service knowledge and can damage the microscope. The simplest way to introduce the light to the microscope is to connect the optical fiber from the top of the objective lens, where there is usually room for the TEM deflection coils and additional detectors. More free space can also be expected from older top-entry devices, where the same location houses the sample vacuum lock and the sample installation mechanism. This configuration is widely described in a previous report25, which contains an exemplary illuminator design. If the interior of the microscope does not allow the installation of the optical fiber above the objective lens, the illuminator may be installed inside the objective lens polepiece from the side32. Other solutions assume using a dedicated sample holder4 or the reverse use of the holder for cathodoluminescent observations1.
In general, the preparation of liquid samples for TEM is complicated and requires experience and manual skills. The critical step in liquid cell preparation is step 2.6. of the protocol. It is important to place the second grid quickly to avoid the evaporation of the liquid. The method presented here is simple and does not require uncommon materials; however, it might not suit all liquids. In some cases, it is beneficial to clean the substrate with plasma to increase its hydrophilicity. For liquid samples, especially water solutions, which do not contain any particles and/or cells that could serve as a spacer between the carbon films, the encapsulation might be unsuccessful as all the liquid may flow out from between the flat surfaces. Using a graphene liquid cell prepared using a holey substrate, which creates small “pockets” for liquid, might be more suitable33.
TEM observations performed in liquid can be disturbed by a few factors related to the sample preparation and the unfavorable effects caused by the electron beam. The first includes the detachment of substrates and the movement of the liquid, often followed by the breaking of the carbon film, which occurs for high liquid volumes in the area and/or upon electron irradiation. This can be avoided by minimizing the liquid volume in step 2.5. of the protocol.
The irradiation with highly energetic electrons leads to energy transfer, resulting in heating or radiolysis. After interaction with radiation, water decomposes into radicals (H· and OH·), H2, and solvated electrons. These species participate in further reactions, thus generating different products and causing damage to the sample34. Therefore, it is very important to keep the electron dose low during the observations (step 3.2.), especially for sensitive specimens like cells. Other effects caused by the electron beam are precipitation from the solution (Figure 4A) and foaming of the liquid (Figure 4C). Lowering the electron dose helps to avoid these two negative effects, but sometimes it is not possible to eliminate them. There may also be some radiation-sensitive contamination in the solution, which settles on the cell (Figure 4B) and changes upon electron beam influence. This is unfavorable as the impurity might be taken as part of the cell, which is damaged upon the light irradiation, leading to wrong conclusions. The best way to obtain reliable images is to find an uncontaminated area and keep the electron dose as low as possible.
Reliable in situ TEM studies of photoreactions require a good comparison between the processes induced by light with those caused by the electron beam. Except for lowering the electron dose described in step 3.2., it is helpful to turn off the electron source for the time of light illumination so that the undisturbed results can be observed, as emphasized in step 4.3. of the protocol.
The present protocol can be used to observe light-induced reactions (especially in liquid), for example, light on metal nanoclusters2, light-induced catalysis processes1, and the influence of light on biological specimens (e.g., cells with a photosensitizer, chloroplasts).
The authors have nothing to disclose.
The research was supported by the Miniatura grant (2019/03/X/NZ3/02100, National Science Center, Poland).
Carbon film on 200 mesh copper grid | Agar Scientific | AGS160 | The standard TEM grids for observations and liquid cell preparation |
Crossover Tweezers | Dumont | N5 | The tweezers are neecesarry for liquid cell preparation |
Photodiode Power Sensor | ThorLabs | S130C | The sensor used for light intensity measurement |
Polyimide-Coated Multimode Fiber | Thorlabs | FG400UEP | Must be built into the microscope using the on-site built adapter, according to the 10.1016/j.ultramic.2021.113388 |
Transmission Electron Microscope | Hitachi | H-800 | Can be replaced with any side-entry microscope, available for modification |
Tuneable Diode Laser | CNI | MRL-III-660D | The light wavelength must be chosen basing on photosensitizer's absorption spectrum |