CRISPR

Question 1

where would you target sgRNAS to create a knockout of a gene?

Answer 1

-closer to the 5’ end of the gene in the earlier exons like exon two to ensure that you are in the ORF and have some translation then the introduction of a frameshift
-need to ensure you have a PAM sequence

Question 2

if you want a homozygous null mutant, what repair strategy would you hope the cells undergo for the protein knockout?

Answer 2

non-homologous end joining (NHEJ) that would create an insertion or deletion that results in a frameshift that comes from a change in base pairs that is not a multiple of three

Question 3

CRISPR systems in the bacterial system

Answer 3

bacterial adaptive immune system, which kills foreign DNA like phages or plasmids and has memory of previous encounters with phages which is called a spacer in the CRISPR system

Question 4

spCas9 in the lab

Answer 4

-nuclear localization signal (NLS) is added to enable nuclear import in eukaryotes
-crRNA and tracrRNA are combined into a single RNA called a single guide RNA (sgRNA)

Question 5

PAM

Answer 5

-protospacer adjacent motif
-NGG
-important to have a PAM in bacterial cell since the nuclease would cut in the bacterial genome @ the CRISPR array and you don’t want to cut the memories

Question 6

cutting with the sgRNA/spCas9 complex

Answer 6

-spCas9 PAM is NGG and must be immediately downstream (3’) of the protospacer on the non-target strand
-spCas9 creates a double stranded DNA break 3 nucleotides upstream (5’) of the PAM

Question 7

what are the two types of DNA repair?

Answer 7

1. non-homologous end joining (NHEJ)
2. homology directed repair (HDR)

Question 8

NHEJ

Answer 8

-dominant repair mechanism and introduces indels (insertions or deletions)
-group of proteins that will stick the double-stranded break back together –> often done incorrectly and you get insertion or deletion @ cut site and cell has done editing

Question 9

HDR

Answer 9

-can add donor DNA to introduce specific sequences
-overwhelm cells with another type of DNA that has an edit that yyou want to make, you can cut the DNA and if there’s HDR and repair occurs with donor DNA

Question 10

knock out the human PLK4 gene

Answer 10

-target sgRNA to an early exon on the opposite strand of PAM
-looking for a point mutation that leads to insertion or deletion from frameshift –> not in multiples of three
-confirm KO with western blot to ensure no protein is produced or PCR amplify the region of interest and send it out for sequencing
-to deal with potential off-target events, BLAST guide against genome to see where off-targets may exist and if you have mismatch far from PAM you may not have issues

Question 11

how would you demonstrate a KO phenotype is not caused by off-target cleavage?

Answer 11

rescue the phenotype with an sgRNA-resistant cDNA encoding the gene (this cDNA sequence cannot be cut by the Cas9 protein)

Question 12

what are some options for Cas9/sgRNA delivery?

Answer 12

1. assemble the Cas9/sgRNA RNP complex- labor and money are extensively needed and in this method you’re just putting sgRNA and Cas9 together in cells through electroporation, lipofection, virus
2. use an sgRNA and an mRNA encoding Cas9- somewhat fast for cutting but may be transient where mRNA may go away as it gets degraded
3. clone the sgRNA (+ Cas9) into a plasmid and generate a lentivirus- good packing efficiency so you can deliver more Cas9 to target
4. clone the sgRNA (+ Cas9) into a plasmid- transiently expressed but since we are using plasmid you could use selectable marker to select for cells that have received sgRNA and Cas9
   -genomic integration with lentiviruses, which are RNA viruses that get reverse transcribed into DNA and DNA goes into the nucleus and is integrated into the genome

Question 13

how would you knock in an AID tag into the human PLK4 gene?

Answer 13

-you would place it in the C terminal @ the end of the gene but before the stop codon or you could add it in internally in protein but would require knowing a lot about protein domains
-ensure you have a PAM sequence and you want break so that when you put in repair template, you can introduce break right before stop codon and ideally cut will disrupt the target site so once edit is made Cas9 will no longer cut
-homology arms on either side of what you’re trying to produce are the hway in which the cells know to use that DNA as repair
-use exogenous DNA containing our tag and then homology arms up and down the break site
-put selectable marker on the plasmid and select for those clones –> selection where you spatially separate clones on a plate and use machine that separates individual cells into different wells and grow them up in liquids
-GOI and guide RNA overlaps with stop codon where you introduce AID tag and Cas9 knows not to cut anymore then use primers to amplify the AID tag and on your primers are different homology arms for GOI

Question 14

how do you confirm knock in of the AID?

Answer 14

you can design PCR to tell you if your clone has given you the edit you want- separate the edited cells into individual wells in a plate and you can either just dilute them to one cell per well and grow them up

Question 15

some HDR tips

Answer 15

-cut close to the insertion site (less than 30 bp_
-homology arms: ~35 bps
-recode (wobble) the sequence between the insertion site and the cut
-include mutations that prevent cutting of the edited locus by Cas9
-gene still makes the same protein but because of DNA sequence is recoded the guide RNA no longer perfectly binds

Question 16

why introduce mutations in the HDR template to prevent cutting by Cas9?

Answer 16

-prevent the introduction of unintended INDELs at the edited locus
-stabilize the HDR template

Question 17

knock in an L89G mutation into the human PLK4 gene

Answer 17

-Cas9 cuts sequence and if it gets repaired back to WT, it will just cut again but hopefully at some rate it recombines with DNA you gave it –> PAM has been disrupted and mutations have been introduced to screen for clones with the Cas9
-take each clone and do PCR and you can add in restriction enzyme since restriction site was introduced and if it cuts amplicon in two, an edit was made
-cutting close to the site where the edits are made increases the frequency of correctly edited clones

Question 18

what does cutting close to the site where the edits are made do?

Answer 18

increases the frequency of correctly edited clones

Question 19

make a large genomic deletion

Answer 19

make repair template where fusing homology regions of DNA that has a restriction site within it

Question 20

base editing with CRISPR-Cas9-Nickase Cas9 fused to deaminase

Answer 20

-Cas9 introduces a double strand break that allows you to use HDR and gives you homology repair but this type of break can be stressful on the cells
-target Cas9 to some DNA target but instead of cutting and repairing, using base editor –> fuse Cas9 to a cytidine deaminase and once Cas9 binds to a target and creates R loop with a free single stranded DNA
-Cas9 hanging @ target and other protein domain will deaminate cytidines and convert them to uridines –> if you have U in DNA, it base pairs with A instead of G
-when this DNA is replicated, U will be matched with A and created edit without double strand break
-you can also use a nick to create single stranded break to tell cell to repair top strand with bottom strand and bias the replication machinery to fix it

Question 21

genetic screens

Answer 21

-selection- growth condition that allows for the propagation of genetically altered cells
-screen- condition where both mutant and WT can grow and be distinguished phenotypically
-reverse genetics- ID the phenotype caused by gene mutations

Question 22

arrayed vs pooled approaches

Answer 22

-arrayed- different mutants in each of the wells and examining behavior of cells in the wells
–typically can see the cell morphology
-pooled- typically selection- flask/plate full of cells with a different mutant and all co-mingled in the same vessel

Question 23

generate genome-scale human CRISPR/Cas9 KO library

Answer 23

-target towards the 5’ end of the first or secodn exon of protein-coding genes
-ideally want 4-6 sgRNAs per gene
-order oligo library of 80,000 guides (20,000*4) then do massive cloning into the vector then deliver into the cells
-use a lentiviral vector so that it integrates
-use the pooled approach to ID the mutated gene and look for depleted guide RNAs after pooling all of the cells
-for the hit list, deep sequencing of the library before and after selection

Question 24

positive selection- ID a proteinaceous receptor for norovirus

Answer 24

-given CRISPR KO library, murine norovirus that kills cells it infects
-use lentiviral library to generate your library (does not kill cells) and norovirus and trying to find gene that encodes receptors for this virus
-perform selection using puromycin and use guide RNA libraries and select for those that uptake the guide through blasticidin and those that survive after norovirus would be screened further
-to rank and ID hits: look @ sgRNA enrichment or depletion and compare that to the sgRNAs that you sequenced @ the start and after you treat with norovirus to look @ surviving cells
-the controls would be before and after infection
-to ID a real hit: ID multiple independent sgRNAs that are targeting the same gene
-hits are per gene and for that gene you zoom in and look @ the performance of the guides and hope they behave the same way

Question 25

lentivirus production

Answer 25

start with an sgRNA library –> transfect into a cell line that will be specialized to produce lentiviruses –> your DNA gets integrated into target cell and target cell can’t make any more viruses

Question 26

transduction

Answer 26

-once you have your viruses, you get CRISPR-Cas9 expressing cells and do some selection for cells making puromycin resistance

Question 27

cell populations and conditions

Answer 27

-untreated control- time zero library- initial cell population with the introduction of sgRNAs that were selected for in the puromycin and measure their abundance per cell
-let these cells grow for the same amount of time as the treated cells without the selective pressure, so this cell population becomes the untreated control
-need untreated control to normalize what you see in the treated condition
-you also want a no-transduction control plate

Question 28

amplify the sgRNAs

Answer 28

-once you have the library, take your time zero and treated library and untreated controls
-in the genome, the construct has been integrated and you have different guide RNAs in each cell but flanking this single guide RNA you have the promoter (driver expression of that guide and then you have tracer RNA that serves as a scaffold)
-take complex library and do PCR with these two primers you will amplify that different guide in every cell
-look for where sgRNAs are enriched or depleted

Question 29

timeline of example CRISPR KO experiment

Answer 29

-day 0: design sgRNA targeting 5’ early exon
-day 1: transduce cells with Cas9, sgRNA-expressing lentivirus
-day 2: apply selection to kill non-transduced cells and include no-transduction control plate
-day 5-10: after selection, evaluate cells for KO by western blot, IF, or sanger sequencing

Question 30

how do you increase confidence in your hits?

Answer 30

repeat the screen using an independently transduced population of cells

Question 31

types of CRISPR screens

Answer 31

-gene KO provides growth advantage –> sgRNAs will enrich
-gene KO provides growth disadvantage/lethality –> sgRNAs will deplete

Question 32

positive selection

Answer 32

strong selective pressure is introduced so that only survival-enhancing sgRNAs will enrich (ID perturbations that cause resistance to drugs, pathogens, or toxins/stress)

Question 33

negative selection

Answer 33

ID genes that affect proliferation or survival of cells, which cause the perturbed cells to deplete during selection (time/cell divisions)

Question 34

improving robustness of hits list

Answer 34

-multiple sgRNAs per gene (4-6): if all sgRNAs give a similar phenotype, it is unlikely to be the result of off-target cutting
-a secondary screen with a new set of separately transduced cells and more sgRNAs per gene
-deeper sequencing depth (especially if you are looking for depletion of sgRNAs in the synthetic lethality screen)

Question 35

pooled CRISPR screen workflow

Answer 35

1. generate Cas9-expressing cells
2. transduce with library of sgRNAs and select
3. treat cells with drug/virus or negative control for 42 days
4. compare gene essentiality between samples by counting sgRNA representation