Toward Compositional Contrast by Cryo-STEM

Electron microscopy (EM) is a versatile tool for studying matter at various scales. However, the high vacuum and charged irradiation necessitate careful specimen stabilization, particularly for sensitive biological samples. Cryogenic fixation, or vitrification, avoids dehydration and preserves the native state of samples. Cryo-EM has advanced resolution from membranes to molecules and atoms, winning Nobel prizes in 1974 and 2017. While cryo-EM excels at high-resolution imaging, quantitatively interpreting pixel intensities for compositional information, especially in cellular samples, remains a challenge. Conventional cryo-transmission electron microscopy (TEM) is limited due to its strong dependence on defocus and lack of contrast at low spatial frequencies. Soft X-ray cryo-tomography, offering quantitative contrast based on atomic absorption differences, serves as inspiration. This paper investigates transferring aspects of this quantitative contrast to cryo-electron microscopy, focusing on the potential of scanning transmission electron microscopy (STEM).

The authors review previous work demonstrating the use of cryo-STEM tomography (CSTET) for thick biological specimens, including whole bacteria (Wolf et al., Nat. Methods 2014), phosphorus detection in vitrified bacteria (Wolf et al., J. Microsc. 2015), 3D visualization of mitochondrial calcium stores (Wolf et al., eLife 2017), and detection of isolated protein-bound metal ions (Elad et al., PNAS 2017). These studies highlight the capability of cryo-STEM to provide compositional information and three-dimensional structural detail in various biological systems. The literature also reveals limitations of traditional cryo-TEM for compositional analysis and emphasizes the need for alternative techniques, such as STEM, to overcome these limitations. Existing methods like energy-dispersive X-ray spectroscopy (EDS or EDX) and electron energy-loss spectroscopy (EELS) require high exposures incompatible with cryo-EM's low-dose requirement for preserving sample integrity.

The study explores the application of different STEM imaging modes to cryo-EM. STEM utilizes a focused beam scanned across the specimen, and the transmitted flux is recorded by various detectors (bright field (BF), annular dark field (ADF), high-angle annular dark field (HAADF)). The choice of probe convergence and detector configuration determines image properties. The authors emphasize that incoherent elastic scattering in STEM provides quantitative contrast sensitive to atomic number (Z contrast). Heavier atoms scatter more strongly, providing compositional information. CSTET, which involves collecting tilt series of STEM images for 3D reconstruction, is a key methodological approach. They discuss different STEM modes such as bright field (BF), annular dark field (ADF), high angle annular dark field (HAADF), and incoherent bright field (IBF) highlighting their advantages and disadvantages for different applications. They use image simulations and experimental data to compare and contrast the performance of STEM with traditional TEM, focusing on resolution, contrast transfer, and radiation damage. Single-particle analysis of ADF-STEM is applied to study heavy-chain ferritin decorated with zinc and iron.

The authors' key findings demonstrate the advantages of cryo-STEM for cryo-EM applications: 1. **CSTET for thick samples:** CSTET provides superior contrast and resolution in the axial direction compared to cryo-TEM, particularly for thicker specimens. The unipolar contrast in STEM allows for easier interpretation of density and opacity. The study also reveals that the radiation damage threshold is higher for STEM than for TEM. The authors show that in thicker specimens the incoherent BF mode is more efficient than ADF, especially when the collection aperture matches the illumination. 2. **Quantitative estimation of ion concentrations:** The contrast in cryo-STEM images is shown to be sensitive to elemental composition. By analyzing the annular dark field scattering as a function of collection angle in bacteria containing polyphosphate bodies, the authors successfully estimated the phosphorus concentration. Similarly, in a study of mitochondrial matrix granules, they identified calcium phosphate and estimated the granule density by comparing their intensity with the intensity of water and ribosomes. 3. **Single-atom detection:** Single-particle ADF-STEM analysis of ferritin protein decorated with zinc and iron demonstrated the detection of isolated metal ions. The positions of zinc atoms were identified with precision, while iron atoms showed a probabilistic distribution consistent with transport and deposition pathways. The use of 3D deconvolution enhanced visibility of high-density granules and membranes. These findings show that cryo-STEM can provide high quality, quantitative compositional information complementary to high-resolution morphological data provided by cryo-TEM.

The findings address the limitations of conventional cryo-TEM in providing compositional information. Cryo-STEM offers a powerful complementary technique, enabling quantitative compositional analysis alongside high-resolution morphology. The unipolar contrast and efficient signal collection in STEM make it particularly suitable for thick specimens, extending the scope of cryo-EM to larger volumes and more biologically relevant samples. The ability to detect isolated ions opens new avenues for studying the precise localization and interactions of metal ions in biological systems. The superior depth penetration of STEM compared to TEM is particularly useful in studying thick, fully hydrated specimens. This addresses a limitation in cryo-TEM where near-atomic resolution is typically limited to specimens of only a few tens of nanometers in thickness. The use of 3D deconvolution methods in cryo-STEM tomography enhances the quality of reconstructions, improves the visibility of subtler details, and reduces artifacts caused by high-density inclusions. The interplay between morphology and composition is a significant focus for future work.

Cryo-STEM offers a promising approach for obtaining compositional information in cryo-EM. The ability to combine compositional contrast with high-resolution imaging opens exciting new possibilities. Future work will focus on improving resolution through phase imaging techniques, such as ptychography and differential phase contrast (DPC) STEM, enabling even more precise and detailed analysis of biological specimens.

While cryo-STEM offers advantages over conventional cryo-TEM for compositional analysis, limitations remain. The resolution achieved in some applications may not match the highest resolutions achievable with cryo-TEM. Furthermore, radiation damage remains a challenge, though the use of high dose rates, fast data collection, and flexible scan schemes can help to mitigate this effect. The interpretation of quantitative compositional data can be complex due to multiple scattering in thick specimens. Further developments in image processing and analysis are needed to refine the methods for extracting quantitative information.