
The spatial organization of biomolecules is critical for regulating key cellular processes, including cell signaling, transport, and communication (1-4). Fluorescence microscopy is a powerful tool for visualizing these molecular distributions within cells, providing valuable insights into their structural arrangement. Traditionally, chemical fixation has been employed to preserve biological samples for fluorescence imaging. However, this approach is often accompanied by significant drawbacks, such as induction of structural changes and the introduction of artifacts, which can lead to misinterpretation of molecular localization and cellular architecture (5-7). Addressing these limitations is crucial for advancing our understanding of biomolecular interactions in their native environments.
Cryo-fixation methods, such as cryo-electron microscopy (Cryo-EM) or cryo-fluorescence microscopy (Cryo-FM), offer alternative approach to overcome the limitations of chemical fixation. Fast freezing techniques offer an alternative by inducing a phase transition of the solution inside the biological sample into an amorphous ice state—a non-crystalline form of ice. This process helps preserve the biological sample in a state close to its native condition (8-13). To maintain the amorphous ice state, the biological sample must be kept below the devitrification point (∼136 K) (9). Currently, the most widely used fast freezing techniques are plunge freezing and high-pressure freezing. While plunge freezing is suitable for relatively thin samples, such as proteins and certain mammalian cells, high-pressure freezing can vitrify samples up to 200 μm thick (10-13). Consequently, fast freezing techniques are applicable for preserving a wide range of biological structures, from individual proteins to complex tissues. Apart from fast freezing techniques, studies have shown that hydrophilic polymers can be also used to preserve the structures of some proteins or protein complexes at low temperatures (14-17).
One of the key advantages of cryo-fixation is that it allows fluorescence imaging without the interference of light scattering caused by water crystallization, a common issue in conventional cryo-preservation techniques. The ability to maintain the native state of biological samples under such conditions has made cryo-fixation invaluable for high-resolution fluorescence microscopy, particularly in super-resolution techniques and correlative imaging approaches (18).
These advancements in cryo-fixation, combined with improved imaging technologies, continue to offer new possibilities for studying biomolecular dynamics and interactions with greater accuracy and resolution, furthering our understanding of cellular processes in their true physiological state.
The fluorescence lifetime of a fluorophore increases at low temperatures (19), offering several advantages in single-molecule imaging, such as an improved signal-to-noise ratio, reduced photobleaching, better temporal resolution, minimized autofluorescence interference, and extended observation times. These benefits make single-molecule imaging more accurate and informative, particularly in complex biological systems.
The primary reason for the fluorescence lifetime increases at low temperatures is the decrease in non-radiative decay pathways (20). Normally excited fluorophores can lose energy through non-radiative decay processes, such as molecular vibrations or rotations, where the energy is dissipated as heat rather than emitted as fluorescence. However, as the temperature decreases, these non-radiative processes are less efficient because molecular vibrations and motions are reduced, leading to fewer opportunities for energy loss through non-radiative pathways. Consequently, more energy is directed towards the fluorescent decay pathway, effectively increasing the quantum yield. A higher quantum yield means that the molecule spends more time in the excited state before emitting a photon, which translates into a longer fluorescence lifetime.
Additionally, collisional quenching, where the fluorophore loses energy by interacting with surrounding molecules, is also diminished at low temperatures. With less molecular movement, these quenching interactions occur less frequently, further contributing to the longer lifetime of the fluorophore. Another important factor is the diffusion of molecular oxygen, which plays a crucial role in the photobleaching of fluorophores (21). In frozen samples, oxygen diffusion is significantly reduced, leading to a marked decrease in photobleaching (21). As a result, fluorescence imaging can be performed for longer durations at low temperatures compared to room temperature conditions (Fig. 1A) (22, 23).
Photoblinking, or fluorescence intermittency, can also exhibit unique characteristics at low temperatures. While the overall fluorescence signal may become more stable due to suppressed non-radiative decay pathways at low temperatures, the duration of the ”off” periods in the photoblinking cycle can increase (24). This occurs because the fluorophore becomes trapped in these dark states for longer periods due to the lack of thermal energy required for transitioning back to the emissive state. As a result, low temperatures can reduce the frequency of blinking, but they may also extend the duration of the blinking events, making photoblinking more pronounced in some cases (Fig. 1B). These effects are particularly observed in organic dyes (e.g., Alexa, Cy, ATTO) and some fluorescent proteins (14-17, 25-28), where temperature plays a critical role in modulating the photophysical properties of the fluorophore. Although there have been numerous studies on the stochastic photoblinking of fluorophores, many aspects remain unexplored (29-31).
The ability to perform fluorescence imaging for longer periods than at room temperature has spurred numerous studies exploring the application of super-resolution methods at low temperatures (1, 3, 14-17, 25-28, 32-39). Specifically, single-molecule localization microscopy (SMLM) can be achieved by localizing single fluorophores within diffraction-limited spots, utilizing the properties of photoactivation or stochastic photoblinking at low temperatures (14-17, 25-28, 32-37).
Despite the significant advantages of cryo-fluorescence imaging at the single-molecule level, such as enhanced fluorescence lifetime and reduced photobleaching, there are notable technical and practical challenges that must be addressed. Single-molecule fluorescence imaging relies on a high numerical aperture (NA) optical system to capture the small number of photons emitted by a single fluorophore. However, only air objective (NA < 1) can be used at low temperatures because immersion fluids are not available at cryogenic temperatures, such as in liquid nitrogen (LN2, ∼77 K) or liquid helium (LHe, ∼4 K) (35, 40). This limitation in NA reduces the photon collection efficiency, which is critical for accurate localization of single fluorophores.
Additionally, performing single-molecule fluorescence imaging at low temperatures requires specialized cryostats, in which mechanical and thermal instabilities often lead to drift in the cryo-stage, where the frozen sample is positioned (41). This drift issues significantly reduce the photon detection efficiency of the microscope, making it challenging to accurately localize a single fluorophore and thereby limiting the achievable spatial resolution. Furthermore, to prevent sample devitrification, laser power must be kept low, resulting in weak signal strength (33). Collectively, these factors make single-molecule fluorescence imaging at low temperatures a difficult endeavor.
In recent years, significant advancements have been made to address the technical challenges of single-molecule fluorescence imaging at low temperatures. One of the major limitations in low-temperature single-molecule imaging is the restriction on using high numerical aperture (NA) objectives. To address this, solid-immersion lenses (SILs) have been introduced to increase the effective NA. When paired with an air objective, the effective NA can theoretically reach the refractive index of the material from which the SIL is made (Fig. 2A) (42-45). For instance, a cryo-fluorescence microscope with an effective NA of 2.17 was developed using a cubic zirconia SIL (refractive index = 2.17) in combination with a 0.55 NA objective lens (Fig. 2B) (27, 46, 47). Additionally, total internal reflection fluorescence (TIRF) microscopy has been achieved using a SIL with an air objective lens at room temperature (45). This result suggests that TIRF imaging is available at low temperatures, which is widely used for single-molecule imaging.
Mechanical and thermal stability in cryo-imaging has also seen significant improvements. A cryostat equipped with a vacuum chamber was developed to improve stability of the cryo-stage at low temperatures (Fig. 2C) (48). Liquid helium (LHe) was chosen as the cooling medium based on findings that demonstrating increased photostability of single fluorophores at 4 K compared to 77 K (48). This cryostat design minimized stage drift and improved the SNR compared to cryostats that use liquid nitrogen (LN2) (28, 48). By using this cryo-fluorescence microscope, a resolution of 4–8 angstrom was achieved when imaging the structure of a single protein labeled with ATTO647N at 4 K (15, 48). However, relying solely on the photoblinking kinetics of ATTO647N resulted in low yields (15, 48). To address this limitation, a novel method was developed that combines polarization differences and photoblinking kinetics of ATTO647N, using the fixed dipole orientations of fluorophores at low temperatures to obtain relatively higher yields (16). Recently, the trimer protein and the hexamer protein complex structures were successfully imaged with 4-8 angstrom resolution using this method (Fig. 2D) (17). Despite its remarkable resolution, the LHe cryostat remains expensive and complex to operate, limiting its accessibility.
Another critical challenge in cryogenic imaging is balancing signal strength with prevention of sample devitrification. To address this issue, mammalian cells were cultured on sapphire disks (3 mm diameter and 50 μm thick) that could minimize sample heating and ice recrystallization even at intense laser illumination (∼1-10 kW/cm2) due to the high thermal conductivity of sapphire (28). Subsequently, high-pressure freezing, capable of vitrifying samples up to 200 μm thick, was applied to mammalian cells (28). After vitrification, the frozen samples were transferred to a vacuum-sealed cryostat cooled with liquid helium (LHe, ∼8 K) for fluorescence imaging (28). When imaging the photoblinking process of six different fluorophores (e.g., eGFP, mEmerald, mTagYFP, JF525, JF549, mCherry) bound to the outer mitochondrial membrane protein TOMM20 in mammalian cells, all six fluorophores exhibited enhanced single-molecule contrast at 8 K compared to 77 K (28). This enhanced contrast enabled us to localize the position of endoplasmic reticulum proteins and outer mitochondrial membrane proteins in U2OS cells with a lateral precision of 2-5 nm at 8 K (28).
The field of cryo-fluorescence imaging has made remarkable advancements, enabling unprecedented resolution and insights into biomolecular structure at near-atomic levels (49). However, as we look to the future, several areas offer promising opportunities for further innovation and expansion.
While cryo-fluorescence imaging has already benefited from incorporating techniques such as photoactivated localization microscopy (34), there is significant potential to enhance super-resolution methods for cryogenic environments. Improvements in fluorophore photostability at low temperatures have opened the door to extended observation times, but challenges remain in achieving sub-nanometer precision consistently. Future developments could focus on combining super-resolution microscopy with advanced cryogenic setups to map larger and more complex cellular structures with higher accuracy, such as entire organelles or multi-protein complexes.
Combining the high spatial resolution of cryo-electron microscopy (cryo-EM) with the molecular specificity of fluorescence imaging will be a critical step in advancing correlative microscopy techniques; cryo-correlative light and electron microscopy (Cryo-CLEM) (21, 41, 50). By overlaying fluorescence data with cryo-EM structural information, we will be able to achieve both molecular localization and ultra-structural resolution in a single experiment. This hybrid approach could enable the simultaneous study of protein interactions, structural dynamics, and functional roles within a biological context, revealing more comprehensive molecular insights. However, there is still a significant resolution gap between Cryo-FM and Cryo-EM, making it challenging to implement ideal Cryo-CLEM (21, 41, 50). To reduce this resolution gap, it is necessary to localize a larger number of single fluorophores within diffraction-limited spots. Understanding the photophysics of single fluorophores at low temperatures is crucial for achieving Cryo-CLEM (21). Currently, many studies are being conducted to address this challenge (26, 28, 32-34, 36, 37). Although significant work remains, Cryo-CLEM is anticipated to provide unprecedented insights into structural biology research.
As imaging datasets grow in complexity, the demand for automated analysis tools and machine learning approaches will increase. AI-driven techniques can significantly accelerate image processing, photoblinking analysis, and protein structure reconstructions, helping to overcome human-imposed limitations in manual data interpretation (51). Automated systems capable of recognizing patterns and correcting for photobleaching or drift will streamline experiments and improve accuracy. Moreover, the use of AI to predict molecular behavior or interactions based on cryo-imaging data will open new avenues for understanding dynamic biological systems.
In conclusion, the future of cryo-single-molecule fluorescence imaging lies in the convergence of technological advancements, interdisciplinary collaborations, and innovative fluorophore design. By overcoming existing limitations and integrating cutting-edge techniques, this field is set to expand its impact across structural biology, cell biology, and therapeutic research. As the technology becomes more accessible, cryo-single-molecule fluorescence imaging will likely play an increasingly pivotal role in elucidating molecular mechanisms with unparalleled detail.
This research was supported by the National Research Foundation of Korea funded by the Ministry of Science and ICT (2022R1A2C2091815 to C.U.K., and RS-2023-00280169 and RS-2023-00218927 to J.-B.L).
The authors have no conflicting interests.
![]() |
![]() |