Periodic Reporting for period 1 - Cells-in-drops (High throughput screening of single-cells using droplet microfluidics)
Reporting period: 2016-03-21 to 2018-03-20
Single-cell profiling has gained considerable attention in the past few years triggering significant efforts to develop various analytical techniques to isolate, amplify and sequence the genetic material of single-cells. Despite a great progress in this field, one major limitation remains a challenge – the throughput. Using state-of-the-art technologies such as commercial microfluidic devices or FACS sorting into microtiter plates, a few hundreds to a thousand of cells can be isolated and sequenced. Although these numbers are impressive, they are not sufficient for extensive analysis of heterogeneous cell populations. For example, a tumor is comprised of many cell types and thorough analysis requires sequencing of core biopsies that typically yield ~10.000 single-cells. Extensive single-cell analysis is also required to study the immune response, in which many different cells work in a highly coordinated fashion to fight infection and clear damaged, and diseased cells. In addition, accumulative evidence suggests that rare cell types (or sub-groups of cells) constituting only a small fraction of a population can drive the collective response of an entire population. While working as a post-doctoral researcher at Harvard University, the fellow and co-workers has developed powerful droplet-based approach for single-cell RNA-Seq [Klein*, Mazutis* et a., Cell, 2015, * -equal contribution]. This method, dubbed as inDrops (for indexing Droplets), makes use of droplet microfluidic technology to index individual cells at a rate of >10,000 cells/hour in nanolitre-scale droplets. The goal of this project was to implement the reported high-end technology at the host institution (Vilnius University, Lithuania), and apply it for high-throughput immune cell transcriptional profiling.
The fellow has successfully implemented droplet microfluidics RNA-Seq platform at Vilnius University and applied it to profile the individual cells. Results of this work have been published in Nature Groups journal (Zilions et al., Nature Protocols, 2017). The basic principle of the implemented technology is relatively easy to understand: a mixture of cells is isolated into microfluidic droplets together with barcoding oligonucleotide primers (attached to hydrogel beads), and a mix of reverse transcription (RT) and lysis reagents. The mRNA released from lysed cells remains trapped inside the same droplet and is tagged (barcoded) with oligonucleotide primers during RT reaction. After barcoding, the material from all cells is pooled by breaking the droplets, and the copy DNA (cDNA) library is processed for next-gen sequencing.
In addition, during the course of the project the fellow and his group has developed a unique approach for performing single DNA molecule amplification and condensation into micrometer size particles (Zubaite et al., Micromachines, 2017). During isothermal amplification reaction DNA becomes packed into crystalline like particles making it possible to purify them from the reaction mix, and use them as a template for improved protein synthesis in vitro. The suggested approach, in principle, could be adapted to various proteins or enzymes when their expression in vivo is inefficient or incompatible with living functions. The catalytic activity of in vitro synthesized protein in a 384-well plate and in microfluidic droplets using purified DNA particles supported high yields of protein synthesis. It is important to emphasize that DNA isolation, amplification and condensation inside droplets prevents the newly synthesized DNA from forming loose, long-range catenated structures between multiple particles, which is an important consideration when preparing clonally amplified gene libraries. The reported technique should benefit biological applications relying on completely in vitro gene expression assays such as directed evolution or drug and enzyme screening.
Taking advantage of the microfluidics platform implemented during this project we have established a collaboration with the Methodist Research Institute (USA) to study drug distribution in tumours. The results of this collaboration have been recently reported in a scientific article Kiseliovas et al., Journal of Controlled Release, 2017. Therefore, out of planned 2 scientific articles, the results of this project have been already reported in 3 publications,
Finally, in collaboration with Prof. Dana Pe'er we have developed a computational algorithm to analyze sparse single-cell RNA-Seq data. The work has been uploaded on BioRxiv and describes Markov Affinity-based Graph Imputation of Cells (MAGIC) for imputing missing values, and restoring the structure of the RNA-Seq data. After MAGIC, we find that two- and three-dimensional gene interactions are restored and that MAGIC is able to impute complex and non-linear shapes of interactions. The algorithm also retains cluster structure, enhances cluster-specific gene interactions and restores trajectories, as demonstrated in mouse retinal bipolar cells, hematopoiesis, and our newly generated epithelial-to-mesenchymal transition dataset.
Social impact.
The technological platform developed by the fellow and co-workers has unleashed a huge interest among many researchers worldwide, allowing to strengthen the existing and establish new collaborations, attract external funding and provide unprecedented career opportunities for younger generation of Lithuanian researchers. For example, Rapolas Zilionis, a graduate student from Dr. Mazutis laboratory, is currently conducting a PhD work at HMS, Harvard Medical School (Dr. Allon Klein laboratory). Another student, Greta Stonyte has completed her internship at Prof. George Church laboratory (HMS), a world leader in human genomics. Columbia University at NYC has hosted two PhD students from Lithuania for 1 year internships (Juozas Nainys and Vaidotas Kiseliovas). In all these cases the financial support has been provided by the US, which is a truly remarkable opportunity for younger generation to improve their professional qualities and to establish their future careers. During the course of this project the fellow have consulted, assisted and collaborated with various institutions such as Helsinki University, Glasgow, NUI Galway, CRG Barcelona, Columbia, Caltech, Oxford to name a few. Number of scientists from abroad have visited the host institution to master droplet microfluidics techniques.
In addition, during the course of the project the fellow and his group has developed a unique approach for performing single DNA molecule amplification and condensation into micrometer size particles (Zubaite et al., Micromachines, 2017). During isothermal amplification reaction DNA becomes packed into crystalline like particles making it possible to purify them from the reaction mix, and use them as a template for improved protein synthesis in vitro. The suggested approach, in principle, could be adapted to various proteins or enzymes when their expression in vivo is inefficient or incompatible with living functions. The catalytic activity of in vitro synthesized protein in a 384-well plate and in microfluidic droplets using purified DNA particles supported high yields of protein synthesis. It is important to emphasize that DNA isolation, amplification and condensation inside droplets prevents the newly synthesized DNA from forming loose, long-range catenated structures between multiple particles, which is an important consideration when preparing clonally amplified gene libraries. The reported technique should benefit biological applications relying on completely in vitro gene expression assays such as directed evolution or drug and enzyme screening.
Taking advantage of the microfluidics platform implemented during this project we have established a collaboration with the Methodist Research Institute (USA) to study drug distribution in tumours. The results of this collaboration have been recently reported in a scientific article Kiseliovas et al., Journal of Controlled Release, 2017. Therefore, out of planned 2 scientific articles, the results of this project have been already reported in 3 publications,
Finally, in collaboration with Prof. Dana Pe'er we have developed a computational algorithm to analyze sparse single-cell RNA-Seq data. The work has been uploaded on BioRxiv and describes Markov Affinity-based Graph Imputation of Cells (MAGIC) for imputing missing values, and restoring the structure of the RNA-Seq data. After MAGIC, we find that two- and three-dimensional gene interactions are restored and that MAGIC is able to impute complex and non-linear shapes of interactions. The algorithm also retains cluster structure, enhances cluster-specific gene interactions and restores trajectories, as demonstrated in mouse retinal bipolar cells, hematopoiesis, and our newly generated epithelial-to-mesenchymal transition dataset.
Social impact.
The technological platform developed by the fellow and co-workers has unleashed a huge interest among many researchers worldwide, allowing to strengthen the existing and establish new collaborations, attract external funding and provide unprecedented career opportunities for younger generation of Lithuanian researchers. For example, Rapolas Zilionis, a graduate student from Dr. Mazutis laboratory, is currently conducting a PhD work at HMS, Harvard Medical School (Dr. Allon Klein laboratory). Another student, Greta Stonyte has completed her internship at Prof. George Church laboratory (HMS), a world leader in human genomics. Columbia University at NYC has hosted two PhD students from Lithuania for 1 year internships (Juozas Nainys and Vaidotas Kiseliovas). In all these cases the financial support has been provided by the US, which is a truly remarkable opportunity for younger generation to improve their professional qualities and to establish their future careers. During the course of this project the fellow have consulted, assisted and collaborated with various institutions such as Helsinki University, Glasgow, NUI Galway, CRG Barcelona, Columbia, Caltech, Oxford to name a few. Number of scientists from abroad have visited the host institution to master droplet microfluidics techniques.
Based on results obtained during this project the fellow has established a start-up company "Droplet Genomics" (http://dropletgenomics.com) which aims to commercialize single-cell technologies. The company has created 7 new working places and represents very clearly the outstanding impact of the project. In addition, the patent application (PCT/IB2017/050124) describing single-DNA molecule amplification using droplet microfluidics, and condensation into DNA particles, have been submitted to international patent office, which is another important achievement beyond the state of the art.