Recognition of the importance of community diseases caused by human coronaviruses has increased in recent years; however, detailed information on pathogenesis, immunity, and viral characteristics remains limited. As such, significant efforts have been made to develop more sensitive diagnostic tools and molecular detection methods32. In this manuscript, we described the application of a label-free method for direct visualization and relative quantitative analysis of two human coronaviruses, i.e., HCoV-OC43 and HCoV-229E in a cellular environment by combining the enhanced darkfield microscopy and hyperspectral mapping. First, using hyperspectral image analysis we mapped the commercially acquired and prepared viral samples and compared their spectral profiles to verify whether the PEG precipitated virions display the same spectral characteristic as the commercially acquired samples. Here, we found that their spectral signatures were identical. Next, we compared the mean spectral profiles of HCoV-OC43 with HCoV-229E virions in the solution and noted a distinct spectral signal for each viral strain. Thus, these spectral profiles can be used as a distinguishable parameter for the differentiation of these two human coronavirus strains. Further, the spectral profiles of uninfected and infected cells were compared to understand to what extent the spectral signature of virions would be detectable in the cellular environment. Here, we noted that the spectral signatures of the infected cells exhibited a considerable peak shift compared to the uninfected cells. As the spectral profiles for HCoV-OC43 and HCoV-229E in the infected cells are distinct from each other, these unique spectral responses from the virions can be further mapped in the infected cells and utilized for the generation of the reference libraries for future analysis. Thus, the application of EDHM analysis allows differentiation between two HCoV strains when present in the solution or cellular milieu.
EDHM offers a significant advantage over conventional imaging techniques, as it involves minimum and non-destructive sample preparation, fast image acquisition, and rapid analysis Additionally, it has the capability to determine the spatial distribution and characterization of samples in complex biological environments33. Moreover, EDHM can be operated by a relatively inexperienced individual with minimal training compared to other microscopic techniques, such as transmission electron microscopy (TEM) or scanning electron microscopy (SEM). EDHM system is also significantly less expensive than alternative options with estimated costs for the system averaging at approximately $155,000. In comparison, TEM system costs on average $4.0 million, while SEM system expenses are near $1.0 million24.
EDHM has been applied to deliver real-time images of biomarker information and to examine cell pathophysiology depending on the spectral resonance characteristics in relevant tissues. For example, EDHM has been approved for in vitro screening of chemical entities for amyloidogenesis modulatory activity and for detection of Aβ aggregate in Alzheimer's mouse brain and retina by analyzing unique signatures for Aβ plaques34. EDHM is also a convenient, noninvasive tool for assessing signs of hemorrhagic shock (HEM), for example, the quantification of changes in the surface tissue saturation of oxygen (SHSIO2), wherein the brightness of the oxygen saturation images generated through EDHM is proportional to the intensity of SHSIO235. EDHM in combination with an artificial neural network (ANN) has been used for the diagnosis of urolithiasis recidivism36, which affects 10-20 % of the population in developed countries37. In this case, EDHM provides rapid characterization and classification of renal calculi within the urinary tract compared to conventional techniques like stereoscopic microscopy38 and infrared analysis39.
Previously an electronic biosensor based on single-walled carbon nanotubes (SWCNTs) network was used for recognition of the dengue virus in infected cells40,41. However, numerous studies have indicated that SWCNTs have the potential to cause pulmonary injury42–45, and increase the susceptibility of small airway epithelial cells (SAEC) to influenza A virus (IAV)46. EDHM has been used to evaluate the unique spectral profiles and cellular localization of SWCNTs and viral particles in fixed SAEC. This analysis suggested that the co-exposure of the viral particles and SWCNTs to SAEC increased the intracellular localization of IAV46.
Recently EDHM has been applied to track gold nanoparticles capped with antisense oligonucleotides (Au-ASOmix) against SARS CoV-2. In the presence of SARS-CoV2 infection, a notable number of agglomerated gold nanoparticles were detected indicating the specific binding of nanoparticles to SARS CoV2 RNA. Further, a significant hyperspectral shift and broadening of hyperspectral signatures were also observed for the Au-ASOmix nanoparticles in the presence of viral RNA47.
Besides, a modified version of the EDHM system including the outlier removal auxiliary classifier generative adversarial nets (OR-AC-GAN), has been applied to detect early symptoms of the disease caused by Tomato Spotted Wilt Virus (TSWV). OR-AC-GAN is a popular neural network architecture in the deep learning domain48–50, that in combination with EDHM, has been utilized for image segmentation, feature extraction, and spectrum classification. This modified EDHM system can distinguish the pixels of healthy plants and plants infected with TSWV at early stages before the symptoms are visible on the plants, thus facilitating the management and spread of disease51. Thus, there is an upward trend in the utilization of EDHM for medical diagnostics, as well as for image-guided surgeries.
Although the EDHM confers various advantages over conventional microscopy techniques, in terms of cost and time reduction, the amount of data generated to form the hypercubes (hyperspectral objects resulting from the hyperspectral measurements of wavelength information for all the bands in a hyperspectral image)41 requires substantial processing, large data storage, and accurate analysis to extract appropriate conclusive information26,52. Occasionally, along with the reflected/scattered light from the sample, the out-of-focus light also reaches the objective, reducing the spatial resolution of the image. Therefore, appropriate instrumental adjustments are required to provide the users with vertical scanning capabilities53.
In summary, the EDHM offers a novel diagnostic tool that can be utilized to detect and differentiate infectious agents, e.g., HCoVs in the solution and within infected cells. This technique can also be modified to obtain the real-time images of biomarkers related to virus-induced pathogenesis, e.g., acquired immunodeficiency syndrome (AIDS) which results in dark skin lesions, Eczema herpeticum, and for analyzing the cell pathophysiology based on the spectral characteristics of relevant tissue. This will aid in the early detection of the symptoms and efficient treatment of diseases. Additionally, combining EDHM with other techniques like Raman spectroscopy will improve the diagnostic applications of this technique.