Republished from the official company website. Author and content source: Crystal NAX.
mRNA, RNP, plasmid, and viral vectors each make different biological and regulatory tradeoffs for Cas9 delivery.
Originally published by Crystal NAX. Republished with company attribution. Original technical perspectives, references and source links are retained. The original website also credits A C.
Four formats are in active use: plasmid DNA, AAV delivered single stranded DNA (ssDNA), ribonucleoprotein complexes (RNP), and messenger RNA (mRNA). This article maps which format fits which program context, what each one costs in biological and operational terms, and where mRNA becomes the most defensible choice for programs moving toward clinical translation.
For the broader landscape of viral and non-viral delivery platforms beyond gene editing, the companion article DNA and RNA Delivery Methods for Gene Therapy Compared covers mechanisms, indication fit, and CMC requirements.
Plasmid, AAV, RNP, and mRNA each deliver Cas9 through a different mechanism, and the best format depends on cell type, integration tolerance, and delivery route.
Plasmid DNA delivers a circular dsDNA construct encoding Cas9 and, in most designs, the guide RNA. It must enter the nucleus to be transcribed, which in non-dividing cells depends on active nuclear import and is therefore inefficient. Expression onset is slow, typically 24 to 48 hours, and the construct can persist for days to weeks depending on cell division rate. It is usually delivered by lipid nanoparticles, abbreviated LNP, or transfection reagents.
AAV delivers Cas9 as a linear, single stranded DNA fragment of approximately 4.7 kb. Once inside the nucleus, this ssDNA is converted to dsDNA and transcribed into Cas9 mRNA, which is then translated in the cytoplasm. AAV persists primarily as an episome in post-mitotic cells and does not integrate into the host genome at meaningful frequency under normal conditions. The gRNA is encoded in the same vector as a U6 driven expression cassette when space allows, or in a second separate AAV vector when the Cas9 cassette consumes most of the payload budget.
RNP delivers pre-assembled Cas9 protein already complexed with guide RNA directly into the cytoplasm. No transcription or translation is required. Editing begins within hours of delivery and the protein is cleared within one to two days. RNP is usually delivered via electroporation.
mRNA delivers the Cas9 coding sequence as a translatable transcript. It acts in the cytoplasm, bypasses the nucleus entirely, is translated into functional nuclease within two to four hours, and is degraded within 24 to 48 hours. Guide RNA is co-delivered as a separate synthetic molecule. mRNA is usually delivered via Lipid Nanoparticles (LNP).
For programs where transient expression, no integration risk, and in vivo LNP delivery are all requirements, mRNA is the only format that satisfies all three simultaneously.
Plasmid DNA works best for stable cell line generation and screening in dividing, immortalized cell lines, where integration risk and slow nuclear entry are not limiting factors.
Plasmid DNA is inexpensive, stable under standard storage conditions, and technically accessible for most molecular biology laboratories. For stable cell line generation, where integration is the objective rather than a liability, plasmid remains a standard tool. For early screening workflows in immortalized, rapidly dividing cell lines where nuclear entry is efficient and off-target accumulation is not a primary concern, plasmid is often adequate.
The constraints become binding in three situations:
For research workflows in dividing cell lines where these constraints do not apply, plasmid remains a practical and cost-effective choice. For primary cell editing, clinical translation, or any context where integration risk requires active management, its limitations become program-level liabilities.
Yes, AAV delivers Cas9 as a linear ssDNA genome and works well for post-mitotic tissues such as retina, CNS, or muscle, though its 4.7 kb packaging limit restricts which editing tools fit in a single vector.
AAV packages Cas9 as a linear ssDNA genome that persists episomally in post-mitotic tissues without meaningful integration. That persistence makes it therapeutically relevant for tissues like muscle, retina, and CNS, where long-term expression from a single administration is the goal, and AAV carries strong clinical precedent across multiple approved gene therapies.
The defining constraint is the 4.7 kb packaging limit. SpCas9 and a U6-gRNA cassette together consume approximately 4.2 kb, leaving minimal space for regulatory elements in a single vector. When space is insufficient, Cas9 and gRNA are split across two separate AAV vectors, which requires co-infection of the same cell and adds manufacturing and titration complexity. The size constraint is more severe for next-generation editing tools. Prime editors run approximately 6.3 kb and base editors exceed the limit once a deaminase domain is added to SpCas9, making dual-vector delivery the only AAV-compatible option for prime editing and base editors. One way to recover packaging space is to use smaller Cas9 orthologs such as SaCas9, at approximately 3.2 kb, which fits alongside a gRNA cassette in a single vector, though with a different PAM requirement and a narrower characterized editing profile than SpCas9.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
RNP is the preferred choice for ex vivo editing of primary T cells and hematopoietic stem cells by electroporation, where its fast onset and short residence time limit off-target exposure.
RNP delivers the fastest editing onset of any format. Because Cas9 protein arrives pre-assembled with guide RNA, no transcription or translation is required, and the complex is active within hours of cell entry. Clearance occurs within one to two days through normal protein degradation, producing the narrowest editing window and the lowest cumulative off-target exposure of any format. In human CD34+ hematopoietic stem and progenitor cells, RNP delivered by electroporation has been reported to produce robust editing efficiency, and for ex vivo workflows in HSCs and T cells where cold chain management is feasible, it remains a strong option.
The constraints are decisive beyond the ex vivo setting:
The format choice therefore follows directly from the program context: ex vivo electroporation or local injection favors RNP, and systemic in vivo delivery requires mRNA.
mRNA is preferred for programs that need transient expression, no integration risk, and in vivo delivery through lipid nanoparticles all at once, a combination no other format supports.
Cas9 mRNA satisfies the constraints that eliminate the other formats in the most clinically relevant editing contexts. It bypasses nuclear entry, produces no integration risk, supports low immunogenicity delivery via Lipid Nanoparticles (LNP), enables rapid guide RNA iteration without re-cloning, and degrades predictably within 24 to 72 hours to produce a defined editing window. Manufacturing and regulatory complexity are substantially lower than plasmid DNA, which carries integration assessment obligations, and RNP, which requires protein production and cold chain management.
For in vivo editing programs targeting liver, lung, or muscle, RNA-LNP formats are preferred for low immunogenicity and repeatable dosing. For IND-enabling programs where integration assessment and immunogenicity characterization must be minimized, mRNA offers the most tractable regulatory path.
The advantage is conditional on one thing: mRNA quality. A preparation with high dsRNA content activates innate immune sensors and suppresses translation. Low capping efficiency reduces functional nuclease yield to the point where editing efficiency becomes batch dependent. Truncated transcripts cannot produce full-length Cas9 protein. The performance difference between Cas9 mRNA and other formats in practice comes down to the quality of the preparation. Read here how to Choose the Best Cas9 mRNA for Gene Editing.
At NAX, Cas9 mRNA is manufactured with m1Ψ modification, validated dsRNA control below 0.1 percent by J2 ELISA, capping efficiency above 95 percent Cap 1 confirmed by LC-MS, and functional potency testing in a cell based editing assay.
The format follows directly from program type: plasmid for dividing cell line work, AAV for long-term expression in post-mitotic tissue, RNP for ex vivo primary cell editing, and mRNA for in vivo systemic delivery or IND-enabling programs.
For research cell line engineering in immortalized dividing cells where integration is the objective or off-target accumulation is not a primary concern, plasmid is adequate and cost-effective.
For in vivo editing of post-mitotic tissues such as retina, CNS, or muscle where long-term expression is acceptable, AAV-ssDNA has the strongest clinical precedent. Standard SpCas9 approaches must account for the packaging constraint through smaller orthologs like SaCas9 or a dual-vector split. Base editing and prime editing require dual-vector AAV or a switch to mRNA delivery.
For ex vivo editing of primary T cells or HSCs by electroporation, RNP delivers the highest editing efficiency and shortest nuclease residence time. mRNA is the appropriate alternative when RNP is unavailable or when the workflow must scale toward manufacturing.
For CAR-T programs, the dominant delivery format has historically been viral transduction using lentiviral or gammaretroviral vectors, which produce stable, permanent CAR expression. mRNA-LNP is emerging as a non-integrating alternative for both ex vivo and in vivo approaches. In ex vivo manufacturing, LNP delivery preserves T cell fitness and reduces exhaustion relative to electroporation, which matters when T cell quality at the point of infusion is a clinical variable. For in vivo CAR-T, mRNA removes integration risk and supports repeatable dosing, properties that viral formats cannot offer. For autoimmune indications where prolonged CAR persistence is unnecessary, the transient expression window of mRNA is a regulatory and safety advantage.
For systemic in vivo editing via LNP targeting liver, lung, or other tissues, mRNA is the only viable format. RNP cannot be LNP-encapsulated and is limited to local injection. Plasmid requires nuclear entry that LNP delivery does not facilitate efficiently. Viral vectors introduce immunogenicity and manufacturing complexity that does not fit a transient editing objective.
For high-throughput guide screening, a single validated Cas9 mRNA batch can be co-delivered with multiple synthetic guide RNAs in parallel without re-cloning, compressing the design-to-edit timeline relative to plasmid-based workflows.
| Attribute | Plasmid DNA | AAV ssDNA | RNP | mRNA |
|---|---|---|---|---|
| Onset of editing activity | 24 to 48 hours | Days, sustained expression | 1 to 4 hours | 2 to 6 hours |
| Genomic integration risk | Present, requires assessment | Low, mainly episomal | None | None |
| Primary cell compatibility | Poor, depends on nuclear entry | Good in post-mitotic tissue | Strong for HSCs and T cells | Strong across cell types |
| Compatible with in vivo LNP delivery | Not efficient | Not applicable, uses viral capsid | No, limited to local injection | Yes |
| Packaging or size constraint | None | 4.7 kb limit | None | None |
| Manufacturing and cold chain burden | Low | High, viral vector production | High, protein cold chain | Moderate, lyophilization possible |
| Typical administration route | Ex vivo transfection | In vivo, direct injection | Ex vivo electroporation, or local in vivo injection | Ex vivo or systemic in vivo via LNP |
What is the best format to deliver Cas9 in vivo?
For systemic in vivo delivery, mRNA encapsulated in LNPs is currently the only format that is both biologically feasible and operationally tractable. RNP cannot be LNP-encapsulated and is limited to local injection. Plasmid DNA requires nuclear entry that LNP delivery does not facilitate efficiently in non-dividing cells. AAV-ssDNA is an alternative for post-mitotic tissues like retina, CNS, and muscle where long-term expression from a single administration is acceptable, but carries immunogenicity and redosing constraints that mRNA avoids.
Can AAV deliver base editors or prime editors?
Not in a single vector. Base editors and prime editors exceed the 4.7 kb AAV packaging limit and must be split across two vectors using intein-mediated reconstitution, requiring co-infection of the same cell and adding manufacturing complexity. mRNA delivery imposes no size ceiling and can accommodate any editing tool without dual-vector engineering.
RNP vs mRNA for primary T cell editing: which should I use?
For ex vivo editing by electroporation where cold chain is manageable, RNP delivers the highest editing efficiency and shortest nuclease residence time. mRNA is the appropriate alternative when the workflow must scale toward IND manufacturing or when LNP delivery is preferred over electroporation.
Does Cas9 mRNA integrate into the genome?
No. mRNA is translated in the cytoplasm and degraded within 24 to 72 hours through normal RNA decay pathways. It does not enter the nucleus and cannot integrate. This removes the integration assessment burden required for plasmid DNA and viral formats and simplifies the regulatory path for IND-enabling programs.
Why does plasmid DNA perform poorly in primary cells?
In non-dividing or slowly dividing primary cells, plasmid nuclear entry depends on active import mechanisms rather than mitotic breakdown of the nuclear envelope, which is inefficient for large circular DNA constructs. mRNA and RNP both act in the cytoplasm and bypass this constraint entirely.
What Cas9 format should I use for an IND-enabling gene editing program?
mRNA combined with LNP or electroporation offers the most tractable path. Plasmid and viral formats carry integration assessment obligations. RNP requires protein production, cold chain management, and does not scale easily under defined CMC conditions. mRNA manufacturing and analytical characterization are substantially simpler and better supported by existing regulatory precedent.
Can I use the same Cas9 mRNA batch for multiple guide RNAs?
Yes. Cas9 mRNA quality attributes are independent of guide RNA sequence. A single validated batch can be co-delivered with multiple synthetic guide RNAs across parallel experiments, compressing guide optimization timelines relative to plasmid workflows where each guide variant requires a new cloning and sequence verification step.
What delivery format is used for CAR-T cell engineering?
Approved CAR-T products rely on lentiviral or gammaretroviral vectors for stable CAR transgene integration. mRNA-LNP is emerging as a non-integrating alternative for both ex vivo and in vivo CAR-T programs, with advantages in T cell fitness preservation, repeatable dosing, and removal of integration risk. For autoimmune indications where prolonged CAR persistence is not required, the transient expression window of mRNA is a regulatory and safety advantage over viral formats.
What is the difference between AAV and LNP for in vivo gene editing?
AAV and LNP function as delivery vehicles rather than Cas9 formats. AAV delivers Cas9 as a linear ssDNA expression cassette and is suited to post-mitotic tissues where long-term episomal expression is the goal. LNP delivers Cas9 as mRNA and is suited to systemic delivery targeting liver, lung, or other tissues where transient expression, no integration risk, and repeatable dosing are the design criteria. The two vehicles address different parts of the delivery problem and are not direct substitutes.
What is the fastest Cas9 delivery format for editing onset?
RNP delivers editing activity within one to four hours of delivery because no transcription or translation is required. Cas9 mRNA follows at two to six hours post-delivery. Plasmid and viral formats require 24 hours or more depending on nuclear entry efficiency and transcription rate.
Can Cas9 mRNA be delivered in vivo?
Yes. Cas9 mRNA is compatible with LNP delivery for systemic in vivo editing and has demonstrated preclinical efficacy in liver, lung, and muscle targets. Co-formulation of Cas9 mRNA with synthetic guide RNA in a single LNP particle is the standard approach for in vivo delivery. RNP and plasmid DNA are not feasible in this context for the reasons described above.
Is RNP or mRNA better for HSC editing?
For HSC editing by electroporation, RNP has demonstrated superior efficiency in some primary cell comparisons, including in human CD34+ cells where it has been reported as the only format producing robust editing under certain electroporation conditions. For HSC programs requiring scalable manufacturing toward IND or LNP-based delivery, mRNA is the more operationally tractable format.
What is the regulatory difference between research-grade and IND-aligned Cas9 mRNA?
Research-grade Cas9 mRNA is manufactured under quality practices appropriate for laboratory use, with specifications focused on functional performance. IND-enabling studies require materials manufactured under a quality system supporting regulatory submission, including raw material traceability, controlled process parameters, formal batch records, specification-based release testing, and stability data. Analytical thresholds for dsRNA, endotoxin, and integrity are typically the same or tighter, but process controls and documentation requirements are substantially more extensive. For a full breakdown of quality attributes and specification standards, see How to Choose the Best Cas9 mRNA for Gene Editing.
What distinguishes AAV from lentivirus for gene editing delivery?
AAV persists primarily as an episome in post-mitotic cells and is used for in vivo delivery to non-dividing tissues. Lentivirus integrates into the host genome and is used for stable ex vivo modification of dividing cells. For transient CRISPR editing where integration is not the objective, both formats introduce complexity that mRNA avoids. For a full mechanistic comparison, see DNA and RNA Delivery Methods for Gene Therapy Compared.
Is RNP delivery less toxic to cells than plasmid DNA?
RNP is generally reported to produce lower cytotoxicity than plasmid delivery. The protein complex does not introduce exogenous DNA into the cell and clears within one to two days, which reduces the sustained cellular burden associated with prolonged plasmid expression.
Which Cas9 delivery format carries the lowest off-target editing risk?
RNP carries the lowest cumulative off-target editing risk because Cas9 protein clears within one to two days, limiting the window during which off-target cutting can occur. Plasmid DNA carries the highest off-target risk among the four formats due to prolonged nuclear persistence and extended Cas9 expression.
Does electroporation work for all Cas9 delivery formats?
Yes. Electroporation is compatible with plasmid DNA, Cas9 mRNA, and RNP delivery, since it works by temporarily increasing membrane permeability rather than depending on a specific cargo type. Its main limitations are that it can induce measurable cell death and that plasmid integration through this route remains very low, at roughly 0.01 percent of target cells.
References