molecular

Transcriptomics is the total complement of RNA transcripts. It consists of coding mRNA (95%), and noncoding (rRNA, snRNA, siRNA, miRNA, lncRNA). Can teach cell and tissue specific gene expression features, quantify transcripts, post transcriptional modifications, evaluate splicing/isoforms, quantitate genotype influence on gene expression, and predict gene function. 

RNA sequencing removes the RNA from the cell. Next Generation Sequencing only reads DNA, so we have to turn the RNA into cDNA. Break the cDNA into smaller fragments to run. See how many times a specific gene is read and this shows levels of gene expression. DNA, not RNA, is run using NGS. mRNA is around 1% of the cell’s contents. In order to sequence it we need to get rid of the other 99% of rRNA and tRNA.

 There are two mechanisms that isolate mRNA: Poly(A)-selection leaves the mRNA sequences, the mRNA molecules bind beads (since mRNA has polyA that will bind TTTT but other RNAs do not). Pour all the RNA on the beads and only mRNA stays. rRNA-depletion  gets rid of the rRNA. It has sequences that bind rRNA. Put in all of the RNA, and the other RNAs pass through. In this method, only rRNA is filtered out, so our product is not pure mRNA. The isolated mix from both methods are turned to cDNA and sequenced with NGS. 

The RNA can be analyzed by cluster analysis, which identifies genes that share expression patterns to that gene. It identifies sets of genes that are regulated in similar ways, since genes that are turned off or on together are likely to be involved in similar pathways. Red shows increased expression and blue shows decreased expression. Single-Cell RNA-Seq captures transcription profiles of a single cell (fruit bowl instead of smoothie). Cells act very differently from each other due to differences in microenvironments. It allows transcriptomes of single cells to be separated from the whole population and analyzed individually. An example of this is with mouse retinal cells, they sequenced them to find there are 39 distinct subtypes of retinal cells (many more than they thought). We first separate the cells from each other (use enzymes that break the extracellular junctions). Cells are isolated in a number of ways: (a) technical- count the amount of cells in mixture, ensure each cell is planted in a different cell, and then putRNA into every cell, turn it to cDNA, and have created a RNA library. This is subject to technical error if two or zero cells are in place. Now, (c) FACS is used to isolate cells (using charge separating cells in small drops) very quickly and much more specifically. The next step also requires the generation of cDNA for sequencing. Microfluidics (e) uses a machine. It has cells on one side, which are separated into singular cells. The machine flows the cells and oil that makes fatty droplets, and then beads. The cells and beads are flowed at a rate that wraps every cell in one bead, so that each cell has its own microenvironment with one bead. This allows creation of a cDNA library from each cell since they each have unique microenvironments inside a fatty envelope. Cells are separated, and three types of beads (barcodes) are added (red, green, and blue). This gives a cell a fatty coat and blue bead. Inside the fatty bubbles are enzymes that lyse the cell, which bursts and spills its contents outside. The beads have a barcode and many PolyT sequences that bind the polyA sequences of each mRNA. This results in beads with millions of different mRNA molecules. We put the barcode sequence into the amplification so that we can see where the sequences came from. Every cell has its own unique barcode. When all the cells are sequenced, we will know which cell the RNA came from because of its barcode. When the cell bursts we are left with a mix of all the cDNA, but we still know which cell it came from because of the barcode. This can now be done for hundreds of thousands of sequences (exponential increase from what we could do before this technology). Once we get the results, first do cluster analysis to see how many types of cells are present. For example, first isolate the amount of neurons found in brain tissue using FACS, and then put them into microfluidics to see different expression levels, and analyze using cluster analysis to see which genes are expressed at what points. The amount of RNA combinations shows the amount of cells (since each cell will have its own compilation); if one combination is expressed more than that cell is more common. We can analyze how similar cells are to each other by comparing expression levels of RNA. scRNA can identify rare subpopulations and understand how genes are expressed at high resolutions. 

Multiplexed SIngle-Cell CRISPR Screening combines (barcode mediated) CRISPR with scRNA. For example, if we want to find out how KO or overexpression of a specific gene affects gene expression in the cell. CRISPRi/CRISPR-KO can create a single, known mutation, and then scRNA can show exact RNA expression for that cell. Checked this with ER stress and UPR. can also do this with cancerous tissue and check how different mutations led to different RNA formations in those cells. scRNA works with CRISPR since RNA can be modified to cDNA. This does not work for genomes, although we can do single cell genomes to sequence genomes for one cell (and also single cell proteomics, which sequences the proteasome for a single cell). If we are working on neurological diseases, and a neuron has a specific mutation that causes Parkinson’s. If we sequence all of the neurons we will lose the resolution of this singular mutation, but sequencing of lone cells can show mutation. The same is true for cancer, to show the gene presence that caused the rest of the cancer mutations to happen. 

RNA shows gene expression levels- if there is no RNA, there is no gene expression. If there is RNA, we know the potential for gene expressions but not which proteins will actually form (can have high RNA and low protein or vice versa). There are many regulations on RNA- for example, it can be highly stable or degraded; regulation is by lncRNA and siRNA. There are also many regulations on translation of RNA. Once a protein is translated it can also be broken apart more or less easily. An experiment checked the correlation between mRNA and protein expression and found it to be very low. Need to examine the proteasome to get information about proteins. 

Ribosome profiling is a mechanism in between the transcriptome and proteasome. It uses the strength of RNAseq while also establishing which protein was formed by checking expression of each RNA in the cell (ribosome seq). We don’t check the amount of proteins in the cell; rather check translation by checking which mRNA is on ribosomes and the amount of RNA found on ribosomes (this is the part that uses RNAseq). RNAs are like chains with beads of ribosomes sitting on them. The goal is to extract the mRNA with ribosomes on them. In order to know which parts of the mRNA have ribosomes on them, we add RNase to fragment all of the RNA except for the parts found in the ribosomes (meaning, protects the ribosomes in the process of translation). We then separate the large and small subunits of the ribosomes and sequence the fragment. The more ribosomes found on a fragment, the more it is being translated, and the higher gene expression is. Most ribosomes will sit on the reading frame- the UTR will not usually be sequenced since no ribosomes translate there. So, ribosome profiling can show where the reading frame of a gene is, as well as how much of the protein is translated. Then, we need to compare this amount with the total amounts of mRNA in the cell. RNAseq of cells to show how much RNA there is in cells and then riboseq to show how much of it is being translated, and the ratio between them shows the amount of translation per RNA. Riboseq of RNA being translated to show amounts of RNA actively being expressed.