A tumor identical to a natural one
Scientists have learned how to extract "ideal" 3D cell cultures for cancer research
HSE blog, Naked Science
A group of researchers from Hungary, Russia and Finland has developed a SpheroidPicker system that creates organoids of a certain shape and size from cancer cells. This first-of-its-kind device, powered by artificial intelligence, will standardize the work with tumor samples.
The scheme of operation of SpherodPicker: the system includes a stereomicroscope, a syringe, a slide table and a manipulator controller. The automatic screening function allows you to get images of spheroids. After selecting the spheroids, the spheroid collector automatically transfers the spheroids to the target plate / ©HSE Press Service.
The results of the study are published in the journal Scientific Reports (Grexa et al., SpheroidPicker for automated 3D cell culture manipulation using deep learning). A junior researcher of the Scientific and Educational Laboratory of Artificial Intelligence for Computational Biology took part in the work on the project HSE Nikita Moshkov.
To study cancer treatment methods, the ex vivo method is used, which literally means "what happens outside the body" in Latin. A sample of tumor cells is taken from the patient, and under special conditions they are exposed, which makes it possible to select means and methods of individual treatment. This approach makes it possible to investigate rare forms of cancer when the use of new treatments is difficult due to the small number of patients.
In ex vivo studies, two-dimensional samples consisting of a single layer of cancer cells are widely used. However, the properties of real tumors differ from two-dimensional samples, and three-dimensional cellular structures give a more realistic response to the effects of drugs, since they allow us to study the penetration of drugs and the development of tumors.
Among the variety of possible three-dimensional shapes, the spheroid is the closest to the real properties. Due to their structure, tumor spheroids have a microenvironment that resembles the characteristics of a real tumor and better mimic cancer lesions than single-layer cultures.
The use of three-dimensional cellular structures is still fraught with various difficulties. Firstly, there is no single protocol for creating spheroids, samples of different shapes are used in analyses, and so far, in most cases, cell selection is done manually. Secondly, modern devices do not allow convenient transfer of selected spheroids to a separate place for further study.
The authors of the paper proposed a solution to both problems. They were able to combine two main stages of creating three-dimensional cancer cell cultures into one mechanism: the selection of spheroids of the correct shape and their subsequent transfer to the necessary environment. Scientists have developed a fast and accurate method of searching for spheroids based on deep learning technologies. To train models, a unique database of images of cancer cells of various shapes has been created, and a model capable of detecting and segmenting the necessary objects has been trained.
The installation uses a microscope with a large field of view, which allows for efficient and rapid examination of samples. The micromanipulator moves a glass capillary rod with selected spheroids.
"The peculiarity of our development is that it can quickly indicate the morphological properties that are required for the selected objects, for example, the size range. Spheroid transfer does not affect the morphology and viability of cells, so in the future SpheroidPicker may become an indispensable tool for the study of cancer drugs and new treatment protocols," says Nikita Moshkov, one of the authors, a junior employee of the Research and Training Laboratory of Artificial Intelligence for Computational Biology at the HSE.
Portal "Eternal youth" http://vechnayamolodost.ru