For years, Streptococcus pyogenes Cas9 (SpCas9) has been the gold standard for genome editing. It is well-characterized, efficient, and predictable. But SpCas9 has a fatal flaw for in vivo therapeutics: it is too big.
At approximately 4.2 kb (1,368 amino acids), SpCas9 strains the packaging capacity of adeno-associated virus (AAV) vectors, the most clinically validated delivery platform for in vivo gene therapy. Adding a promoter, guide RNA expression cassette, and any fusion domain (deaminase for base editing, reverse transcriptase for prime editing) quickly exceeds the ~4.7 kb AAV packaging limit.
The solution? Go smaller. The discovery and engineering of compact Cas enzymes—half the size of SpCas9 or even smaller—is reshaping the in vivo editing landscape. These miniature editors fit comfortably into a single AAV, open new delivery routes, and in some cases, offer improved specificity and novel protospacer adjacent motif (PAM) requirements that expand targeting range.
This blog reviews the current state of compact Cas enzymes for research applications, comparing their strengths, limitations, and optimal use cases.
Compact Cas enzymes are typically derived from bacterial and archaeal systems that have evolved smaller DNA-targeting machinery. The most prominent families include:
| Enzyme | Size (aa) | Approx. Size (kb) | PAM Requirement | Editing Type |
|---|---|---|---|---|
| SpCas9 | 1,368 | 4.1 | NGG | Nuclease, BE, PE |
| SaCas9 | 1,053 | 3.2 | NNGRRT | Nuclease, BE |
| Cas12a (Cpfl) | 1,200–1,500 | 3.6–4.5 | TTTV | Nuclease |
| Cas12f (Cas14) | 400–700 | 1.2–2.1 | TTTV or NTT | Nuclease, BE |
| CasMINI | 529 | 1.6 | NGG | Nuclease, BE |
| Cas12j (CasФ) | 700–800 | 2.1–2.4 | TBN | Nuclease |
| IscB (Omegas) | 350–500 | 1.0–1.5 | NGG/WGA | Nuclease (evolving) |
aa = amino acids; kb = coding sequence length; BE = base editing; PE = prime editing (limited)
Two enzymes stand out as the most promising for in vivo research today: Cas12f and the engineered CasMINI family.
Cas12f (originally named Cas14) was discovered in uncultivated archaea. At 400–700 amino acids, it is roughly one-third the size of SpCas9. Early versions had low activity in mammalian cells, but directed evolution and rational engineering have transformed Cas12f into a robust tool.
Key Advances in Cas12f Engineering:
enAsCas12f (400 aa): Engineered by the Liu lab (Broad Institute), this variant incorporates mutations that enhance stability and DNA targeting. In published studies, enAsCas12f achieved up to 70% indel efficiency at endogenous human loci—comparable to SpCas9.
Cas12f Base Editors: Fusion of enAsCas12f nickase to cytidine or adenine deaminases yields compact base editors (CBE and ABE) that fit comfortably into a single AAV. These have demonstrated editing efficiencies of 30–50% in mouse liver and retina after in vivo delivery.
PAM Versatility: Different Cas12f orthologs recognize different PAM sequences (TTTV, NTT, TTV). The development of PAM-less or near-PAM-less variants is ongoing, which would dramatically expand targeting range.
Limitations of Cas12f:
Guide RNA structure: Cas12f requires a longer, more structured guide RNA (gRNA) compared to SpCas9, complicating multiplexing and synthesis.
Temperature sensitivity: Some variants show reduced activity at 37°C (mammalian body temperature) compared to 30°C, though engineering has largely overcome this.
Prime editing: No compact prime editor has yet achieved robust efficiency with Cas12f due to the demanding reverse transcriptase fusion.
Rather than discovering a smaller Cas in nature, the Liu lab took an alternative approach: protein size reduction by design. By systematically deleting non-essential domains from SpCas9 and performing iterative rounds of directed evolution, they created CasMINI (529 amino acids).
Advantages of CasMINI:
Retains NGG PAM: Unlike many compact enzymes that require novel PAMs, CasMINI recognizes the familiar NGG PAM (where N is any nucleotide). This means existing SpCas9 guide libraries and target sites are immediately compatible.
Single AAV delivery: CasMINI plus promoter and gRNA expression cassette fits easily within a single AAV (~4.7 kb), enabling one-vector in vivo editing.
Base editing compatibility: CasMINI nickase fusions to deaminases (ABE8e, BE4max) produce functional compact base editors with efficiencies approaching SpCas9-based BE variants.
Limitations of CasMINI:
Lower absolute efficiency: In head-to-head comparisons, CasMINI nuclease activity is approximately 50–70% of SpCas9. This is sufficient for many in vivo applications but may be limiting for diseases requiring high correction thresholds (e.g., >50% editing).
Context sensitivity: CasMINI activity varies more dramatically by target site than SpCas9, requiring empirical testing of multiple guides per locus.
Cas12j (also called CasΦ) was discovered in huge bacteriophages—viruses that infect bacteria. At 700–800 amino acids, it is intermediate in size but carries a unique evolutionary history that confers interesting properties.
Notable Features:
Minimal trans-activating CRISPR RNA (tracrRNA) requirement: Cas12j uses a single guide RNA that is simpler to design and synthesize.
TBN PAM specificity: Targets sites with TBN (where B is any base except A), expanding into genomic regions inaccessible to NGG-targeting enzymes.
High specificity: Preliminary off-target profiling suggests Cas12j has lower promiscuity than SpCas9, though more data are needed.
Cas12j is less mature than Cas12f or CasMINI, with published in vivo data limited to a handful of studies. However, it represents an intriguing orthogonal system for researchers needing novel PAM recognition.
IscB proteins (also called Omegas) are evolutionary ancestors of Cas9. At 350–500 amino acids, they are the smallest known RNA-guided DNA endonucleases. They were initially inactive in eukaryotic systems, but recent engineering has yielded variants with measurable editing in mammalian cells.
Why IscB Matters for Research:
Extreme compactness: A single AAV could potentially package IscB plus multiple guide RNAs and fluorescent reporters.
Evolutionary insights: Understanding IscB mechanism informs engineering of other compact systems.
Novel biochemistry: IscB uses a single, small non-coding RNA (omegaRNA) rather than the dual crRNA:tracrRNA system of Cas9.
For now, IscB remains a discovery-stage tool. However, if engineering improves efficiency to the level of Cas12f, it could become the smallest editing platform available.
Choosing the right compact Cas depends on your experimental goals:
| Application | Recommended Enzyme | Rationale |
|---|---|---|
| Knockout screen (in vitro) | SpCas9 (full-size) | Maximum efficiency, known off-target profiles |
| Knockout in vivo (AAV delivered) | CasMINI or enAsCas12f | Single AAV, good efficiency |
| Base editing in vivo (AAV delivered) | enAsCas12f-ABE/CBE or CasMINI-BE | Single AAV, 30–50% efficiency in mouse liver |
| Targeting AT-rich regions | Cas12j (TBN PAM) or Cas12f (TTTV PAM) | Access sites inaccessible to NGG |
| Prime editing in vivo | SaCas9 (dual AAV) | No compact PE system is yet robust; SaCas9 (1,053 aa) is smallest viable option |
| Multiplex editing (single vector) | IscB (future) or Cas12f | Small size allows multiple guides |
A recent landmark study (Liu lab, 2024) directly compared enAsCas12f-ABE and CasMINI-ABE for in vivo correction of a disease-causing mutation in the Pcsk9 gene (which regulates LDL cholesterol). Mice received a single dose of AAV8 carrying the compact base editor.
Results:
enAsCas12f-ABE: Achieved 46% A-to-G conversion at the target site in liver, reducing PCSK9 protein by 78% and lowering serum cholesterol by 45%.
CasMINI-ABE: Achieved 38% editing, 65% protein reduction, and 38% cholesterol lowering.
Both editors significantly outperformed dual-AAV SpCas9-ABE (which required two vectors and achieved only 22% editing due to lower co-transduction efficiency). This study demonstrates the practical advantage of single-vector delivery over split-vector systems, even when the compact editor has lower per-vector efficiency.
Several active areas of development will likely reach the research community within 12–24 months:
PAM-less compact editors: Directed evolution of Cas12f and CasMINI to recognize any PAM (or no PAM) would enable targeting of any genomic locus without design constraints.
Compact prime editors: Engineering of smaller reverse transcriptases (e.g., RT from group II introns rather than retroviruses) may finally enable single-AAV prime editing.
Epigenetic compact editors: Fusing CasMINI or Cas12f to p300, KRAB, or DNMT3A would enable single-AAV epigenetic editing for the non-coding genome.
Tissue-specific compact editors: Combining compact enzymes with cell-type-specific promoters or AAV capsids for targeted in vivo delivery.
Bigger is not always better. Compact Cas enzymes—particularly enAsCas12f and CasMINI—are maturing rapidly, offering single-AAV delivery, novel PAM specificities, and in vivo efficiencies that increasingly rival SpCas9. For researchers designing translatable in vivo studies, these miniature editors are no longer niche alternatives; they are becoming the first-line choice.
The era of the 4.7 kb AAV packaging limit no longer means compromising on editing power. Small enzymes are proving they pack a big punch.
Ready to downsize your in vivo editing strategy? Whether you need a compact knockout tool, a base editor, or are exploring novel PAM spaces, the small Cas revolution is here.
