One of the more exciting demonstrations of an RNA sequencing application at AGBT was by Patrik Stahl from Joakim Lundeberg’s lab at the Karolinska Institute. His talk, entitled “Spatially Resolved Analysis of Differential Expression within Single Tissue Sections” described a cool idea, laying a tissue section onto a gridded array with approximately 10 x 10 micron features that contain about 10e6 oligo-dT probes/feature with location-specific barcodes. RNA diffuses directly from the tissue section onto the microarray to hybridize to the oligo-dT followed by cDNA synthesis. The cDNA are then sequenced on the HiSeq after library QC on the MiSeq. The barcodes now define the original 2-dimenisonal location of the RNA at single cell resolution, and the resulting array contained about 135,000 features. The unique assay did not suffer from loss of spatial resolution by diffusion, shown in examples from the olfactory bulb and prostate cancer. In one of the examples, 1.6 x 10e9 reads were used for the RNA library created from one 135k array. The method offers an incredible combination of single cell resolution with the complete transcriptome while conserving the histological information. Very cool!
On the other extreme, Christine Malboeuf from the Broad Institute, demonstrated the ability to sequence viral RNA at ultra-low copy number in the presence of host RNA contamination at 5 pg (10e6 RNA copies). She used the NuGen Ovation RNA-Seq kit to detect 5-50 fg, or 1000-10,000 copies. The Ovation RNA-seq method employs viral specific RT-PCR followed by double stranded cDNA synthesis and RNA amplification of the cDNA template. The final cDNA product is sequenced on Illumina sequencers. She showed data from the sequencing of West Nile virus, HIV, and RSV with > 97% of the genome covered with about 5 million reads. The method was so reproducible and with such high success rates that it has been made the new standard for sequencing viruses at the Broad Institute. In January 2013, about 400 viral genomes were processed and they expect this method to be useful for surveillance of endemic or emerging viral pathogens, detecting co-infection with multiple viruses, and for viral pathogen discovery.
Leonid Moroz, from the University of Florida had another very interesting presentation with great science and a tour de force demonstration of single cell and sub-cellular RNA sequencing. His focus was on memory persistence or ‘how do we remember our first kiss for the rest of our life’ as he put it in one of his slides. Long-term memory leads to altered gene expression and the growth of new synaptic connections. These connections perisist, but the strength of the connections is altered by experience. His group compared the transcriptome and methylome of different types of neurons in snails. He showed a strong preference for snail brains over humans as human brains contain about 100,000,000,000 neurons each sized a few microns whereas snails have only 10,000 neurons of about 1 mm diameter (yes, 1,000 µm!). To further simplify his model system, his group reconstructed memory circuits in cell culture, performed RNA sequencing on these, and identified about 3,000 genes. When these genes were compared with rat brain genes, about 692 overlapped. Makes me wonder what happens when rats dream of snails? He also showed that facilitatory neurotransmitters induce a massive, rapid genome-wide demethylation. Hydroxymethyl DNA seemed to be an important intermediary in the demethylation process. What was surprising that demethylation often happened only on one strand. No hint of how this happens. There were more slides on RNA cargo transport and active demethylation, but by then the limited number of neurons in my brain were overloaded. Leonid had a slide with different organisms and another intriguing statement, that the complex brain has evolved at least 9 times. And that memory may have evolved as a memory to injury to aid survival. Now if only I could remember my first kiss.
Josh Levin, also from the Broad Institute had a very intriguing presentation on a different approach on single cell transcriptomics in stimulated immune cells. But no blogging of this talk!! Sorry.