Delivering editing reagents to cells

A cell can take up Cas9 and remain genetically unchanged at the locus you named. Delivery answers an exposure question: did this form of the reagent reach the nucleus, for how long, and at what cost in viability. It does not answer the genotype. This page compares plasmid, messenger RNA and ribonucleoprotein formats, and electroporation, chemical carriers and viral genomes as ways in. The editing chemistry those reagents are meant to perform is in how CRISPR-Cas9 editing works in research.

Reagent classes live in the molecular biology catalogue. The format you need should be written into the quote request.

Three physical forms of the editor

A plasmid is DNA that must be transcribed and translated before any nuclease exists. Expression from a strong promoter can last for days in a dividing cell, and the plasmid can integrate. That long exposure is useful when you need ongoing guide expression, and it is a liability when every extra hour of nuclease is another chance for an off-target cut. A fluorescent protein or a resistance gene on the same plasmid reports transfection or selection. It is not a scar at the target.

Messenger RNA encoding Cas9, supplied with a guide RNA, skips transcription of the editor. Translation starts soon after the RNA is in the cytoplasm. The pulse of protein is limited by RNA and protein turnover, often a window of hours to a couple of days rather than the life of a plasmid. There is no DNA copy of the editor to integrate, which removes one class of unwanted insertion. The RNA can still trigger innate responses in some cell types. Chemical modification of the RNA is a reagent class used to improve stability and to blunt that sensing. Follow the sheet for the RNA you bought. Do not assume an unmodified transcript and a modified one are interchangeable.

A ribonucleoprotein is Cas9 protein already bound to the guide. It can cut as soon as it is inside the nucleus. There is no delay for expression, and the protein decays instead of being replenished by a plasmid. In published comparisons this short exposure is often the more specific way to cut, because off-target sites that need more time or more enzyme see less of both. Often is not always. A crude protein preparation, a guide with a crowded mismatch profile, or a huge excess of complex can erase the advantage. Specificity claims still belong to an assay, as in off-target checks and their limits.

The guide has to match the format. A plasmid with a U6 cassette produces the guide in the cell. An RNP needs the RNA before the pulse, annealed or complexed in a buffer the protein tolerates. Mixing a DNA guide into a protein complex because both are nucleic acid is a failed delivery dressed as a mystery at the locus.

Electroporation, chemistry and viruses

Electroporation opens the membrane with a pulse. It is the practical route into many primary cells, blood cells and other targets that ignore lipid reagents. The pulse settings belong to the instrument, the cuvette or cartridge, and the manufacturer instructions for that cell class. Copying a voltage from a paper that used a different machine is how cultures die. This page does not list those settings. Viability and uptake move in opposite directions as the pulse gets harsher. The survivors of an overly harsh pulse are a selected population.

Chemical carriers, lipid and polymer reagents among them, are the convenient route for many immortalised adherent lines. Serum, confluence and the charge of the cargo all matter. Plasmid lipofection is routine where RNP lipofection may be poor or may need a different formulation. A reagent that works in HEK293 derivatives can fail in a lymphocyte. Treat a successful chemical transfection in one line as a property of that line. Culture conditions that make the comparison fair are part of mammalian cell culture for research labs.

Viral delivery changes the experiment's lifetime and its institutional category. Lentiviral genomes integrate. They suit pooled screens and stable expression of dead-Cas9 repressors or activators. They are a blunt instrument for a nuclease you wanted gone by the next passage. Integrants keep expressing, and the integration site is an extra mutation you did not design. Adeno-associated virus is largely episomal, with a packaging limit that makes SpCas9 plus guides and donors a tight fit. It brings a different biosafety conversation, not a free pass to call the edit clean. Neither virus proves on-target repair. Both prove that you must genotype.

Work with viral stocks belongs in the containment the institutional review assigned. A biosafety cabinet practice is described in working inside a biosafety cabinet. The cabinet does not correct a wrong plasmid.

When viability and editing disagree

If cells die and the locus is unedited, ease the physical insult or the dose before you redesign the guide. Death without cutting is delivery toxicity. If cells look well and the locus is unedited, you may have failed to deliver, delivered a guide that cannot bind, or cut a population so small that a bulk assay cannot see it. A marker-positive sort, followed by sequencing, separates "nothing entered" from "entered but did not cut". If the marker-positive cells are also unedited, the guide or the protein is the suspect, not the pulse.

If a subset of cells is highly edited and the well as a whole looks weak, you are looking at a mixture. Decisions about pooling versus cloning are in editing pools versus single-cell clones. Do not average them into a single percentage and then treat a later clone as if it had that percentage.

Essential genes punish successful editing with disappearance of the edited cells. The healthy plate can be the plate that was not edited. A viability assay without genotyping will congratulate the failed delivery.

False comfort from a marker or a cassette

Co-transfected GFP is the mildest version of the mistake. Cells glow because GFP was translated. The editor plasmid may have been left out of that tube, or may have failed to enter the same cells. A fused Cas9-fluorescent protein is tighter evidence that the nuclease polypeptide was made, and it still does not show a cut. Resistance cassettes select for survival under drug, which can come from transient expression, integration, or a well that was not actually under selection because the antibiotic had expired or was never added.

Viral titre is not an allele frequency. A high multiplicity of infection can put several integrants in a cell and increase toxicity and off-target exposure together. Report multiplicity only if you measured it. Do not infer it from a bright well.

Sequencing remains the edit assay. Enrichment steps are optional prefaces. Checking whether a genome edit worked is the order of those assays.

Biosafety changes when the vehicle changes

Naked protein and RNA in a research line are not the same review as a lentiviral preparation. The WHO laboratory biosafety manual is a general anchor for laboratory practice. Your institutional committee decides the containment for the vector, the cells and the gene you are altering. This selection guide does not approve a delivery method for therapy, and it does not describe embryo or germline editing. Electroporation can aerosolise a small amount of the cuvette contents. Treat the pulse as a procedure with a closed container and the protective equipment your local rules already require. Do not invent a lower category because the cargo is "just a protein".

Edited cells keep the hazard of the parent. Delivery does not attenuate a pathogen-derived line.

Ice, heat and an instrument that rebooted

Ribonucleoprotein and RNA belong on ice for the short time you are assembling them, including in a laboratory whose room temperature is far above the assumption in a temperate protocol. A complex left on a warm block while a biosafety cabinet is prepared is a denatured reagent. Plasmids are more forgiving and are not an excuse to handle RNA the same way. After a power cut, an electroporator that rebooted should be treated as an instrument that owes you a check with a control cell and a control cargo before a rare primary sample is loaded. A pulse delivered mid-failure is not a biological replicate.

Cold-chain handoff matters when protein or mRNA arrives. A shipment that spent the afternoon warm is a different lot from the one the paperwork describes. Quarantine it until a known guide cuts in a reference line, or replace it. Do not bury that uncertainty inside a story about chromatin.

What the sourcing note must decide

Name the format: plasmid, mRNA, protein, or a viral genome, and the guide form that matches it. Name the cell type, whether the cells are primary, and whether chemical transfection has already failed. State that you need a transient nuclease or a stable expression system, because those sentences buy different reagents. Sequencing after delivery can be discussed through the CRISPR validation sequencing enquiry reference. That page is an enquiry reference. It does not mean a delivery or a validation service is operated on your behalf. Use the molecular biology pathway for the surrounding workflow. Ask whether a quotation is possible. A request for "CRISPR kit" without a format will be answered with an assumption you did not make.