High-sensitivity hypothesis-free subcellular proteomics is challenging due to the limited sensitivity of mass spectrometry and the lack of amplification tools for proteins. Without such technology, it is not possible to discover proteins at specific subcellular locations of interest in cells or tissue samples. Here, we introduce a total-sync ultra-content photo-biotinylation termed "optoproteomics" that integrates microscopy, optics, mechatronics, photochemistry, and deep learning or traditional image processing to enable high-content in situ photolabeling. The Microscoop system photolabels proteins at user defined regions of interests (ROIs) under a microscope utilizing directed photochemistry one field of view (FOV) at a time for tens of thousands of FOVs with similar morphological features. Photolabeled proteins are then extracted and sent to mass spectrometry to reveal ROI-specific known and novel protein players. Thus, beyond spatial proteomics mapping, subcellular spatial proteomics discovery can be achieved.
Similar to dental curing where photochemistry is used to trigger a molecular change, Syncell's Microscoop system utilizes microscopy-focused photochemistry to achieve localized photo-induced protein labeling. One approach is opto-biotinylation. With a photosensitizer catalyzing covalent amino acid labeling with a biotin-containing molecule at the focal region of interest (ROI), one can achieve microscopy-guided subcellular protein biotinylation for later streptavidin bead pulldown, effectively creating a subcellular "pickable" microscope.

That is, under a microscope, if one finds a morphological feature of interest of a cell or tissue sample, such as beta amyloid, pTDP43 aggregates, primary cilia, focal adhesion, immune synapses, or others, the image-guided illumination performed by Microscoop with the wavelength suitable for the photochemical reaction results in protein labeling and scooping at the ROI with high spatial precision. Syncell's proprietary photolabeling probes are designed with a fast photochemical reaction rate (<ms) and low non-specific binding.
To perform photolabeling automatically with high spatial precision, the system needs a real-time pattern generation software. For non-complicated cases, traditional image processing tools such as adaptive thresholding, filtering, or other methods can be implemented to achieve pattern generation for a specific ROI.

For more complex cases, Microscoop allows the use (inference) of a trained neural network from AI deep learning to automatically recognize the ROI of an image. AI-based image segmentation facilitates precise pattern generation for images with high noise, complicated background, unclear interface, or other complexity, such as dendritic spines, T cell-cancer cell interface, etc. Syncell services can help pattern generation using traditional image processing or AI.
The integrated electronic control system of Microscoop is specifically designed to enable accurate and fast control of key steps of optoproteomics including imaging, pattern generation, patterned illumination, and stage movement. All steps are automated and require high speed to complete ultra-content photolabeling within a reasonable duration of time. Mechatronics is implemented to allow tight integration and synchronization of software, firmware, and hardware. Specifically, the sub-ms movements of illumination spots through the coordinates calculated by image processing or AI require dedicated mechatronic design to achieve.
Mass spectrometry is the method of choice for de novo proteomic discovery. Although the sensitivity of mass spectrometry improves significantly in recent years, the required copy number remains high. So far, there is no protein PCR technology to amplify proteins. Microscoop overcomes the protein amplification problem by brute-force protein accumulation, i.e. by photolabeling proteins one field of view (FOV) at a time across thousands of FOVs fully automatically by mechatronic control. The ultra-content cycle of imaging, pattern generation, and illumination results in protein labeling of millions of illumination spots at the ROIs, such that the following scooping and mass spectrometry analysis yield proteomic identification in high sensitivity, essential for discovering the novel protein composition at an ROI.
Images in, proteome out.
An example of a stress granule (SG) study showing the capability of protein biomarker discovery.

(i) Venn diagram of three biological replicates of the SG proteome discovered with optoproteomics.
(ii) Volcano plot of relative protein levels between photolabeled samples and control samples (PL/CTL ratio) in log2 scale. Enriched proteins are shown in green.
(iii-iv) List of top-ranked proteins localized at the SGs isolated by Microscoop. Some are known SG proteins, whereas others are not known to be associated with SGs. Immunostaining (see below) shows that the majority of these previously non-associated proteins are SG biomarkers (orange), demonstrating the high specificity of the optoproteomics technology leading to high-precision subcellular spatial proteomic discovery.
(v) 124 enriched proteins are subject to gene ontology analysis to reveal their SG related biological processes.
Confocal imaging showing that 12 proteins that are previously not associated with SGs are highly colocalized with G3BP1, a known SG marker, in U-2OS cells when cells are treated with arsenite stress. Green: newly identified SG proteins; red: G3BP1; blue: DAPI.
Get in Touch