sgRNA Promoters Compared: U6 vs T7 vs H1 Cassettes

An sgRNA promoter is the sequence that drives transcription of a CRISPR guide RNA, either inside cells (commonly human or mouse U6 or H1) or in vitro (commonly bacteriophage T7). Choosing the wrong class is a design error, not a minor plasmid preference: U6 does not replace T7 polymerase in a tube, and T7 does not transcribe in a typical mammalian nucleus.

This comparison is for teams building or reading CRISPR plasmids and deciding among U6, H1, and T7 before oligos go on order.

Three Promoter Classes Used With sgRNA

Mammalian guide plasmids almost always use a type III RNA polymerase III promoter. Human U6 is the default for a single sgRNA cassette. H1 is a second Pol III option used when a plasmid already contains U6 or when a slightly different expression profile is wanted. T7 is a bacteriophage promoter used to make sgRNA in vitro for ribonucleoprotein (RNP) delivery, or in specialized cells that already express T7 RNA polymerase.

U6: The Default for Mammalian sgRNA Plasmids

U6 promoters transcribe small nuclear RNAs and are well suited to short, uncapped RNAs with defined 3' ends. In a CRISPR cassette, U6 usually starts transcription at a defined G, drives the spacer plus scaffold, and stops at a poly(T) tract. That architecture is why so many lab plasmids look the same on a map: U6 arrow, 20-nt spacer, sgRNA scaffold, TTTTTT.

Design constraints follow from Pol III. A stretch of four or more T's inside the spacer can cause premature termination. The +1 nucleotide of U6 is often constrained, so some labs prepend a G if the spacer does not start with G. The promoter is RNA Pol III, not Pol II, so you should not expect a Kozak sequence, intron, or poly(A) signal to "help" guide expression.

On a plasmid map, confirm the U6 feature actually points into the spacer, not into Cas9. Mixed-up annotation is a common reason a "CRISPR plasmid" produces Cas9 protein and no guide.

H1: A Second Pol III Cassette, Not a Drop-in Upgrade

H1 is another mammalian Pol III promoter. It is smaller than U6 and appears in multiplex constructs so two guides are not both driven by identical U6 copies, which can increase recombination risk in bacteria and in lentiviral genomes.

H1 is not automatically "weaker" or "stronger" in every cell type; that is an empirical question for your line. Treat H1 as a second cassette with the same T-stretch termination rules. If you clone a spacer into an H1 position using oligos designed for a U6 BbsI cassette, check the overhangs and the +1 base. Cassette architecture is not interchangeable just because both promoters are Pol III.

For multiplex plasmids, name each cassette on the map (U6-sgRNA-A, H1-sgRNA-B) and store both spacers in the design record. Off-target review should include every spacer, not only the one discussed in the last meeting.

T7: In Vitro Transcription, Not a Mammalian Plasmid Default

T7 promoters are recognized by T7 RNA polymerase. In cloning practice that usually means a linear DNA template is transcribed in a tube to produce sgRNA for RNP electroporation or microinjection. Some bacterial CRISPR systems also use T7 when the strain expresses the polymerase.

Putting a T7-sgRNA cassette on a mammalian expression plasmid does not create guide RNA in HEK293 cells unless T7 polymerase is supplied. Conversely, a U6 plasmid is the wrong template style if your protocol is "transcribe 100 µg of sgRNA overnight." Match the promoter to the delivery mode: plasmid or virus in cells (U6/H1) versus RNA or RNP (T7).

Template design for T7 often needs a 5' G in the transcript and a clean run-off or terminator. If you PCR-amplify a T7-sgRNA fragment, confirm the T7 promoter is complete in the oligo and that the scaffold is not truncated. In vitro RNA should be handled as a reagent lot: concentration, integrity, and the spacer sequence belong in the experiment record.

How to Choose a Promoter for the Actual Experiment

If you will transfect or transduce a mammalian cell with a DNA cassette, start with U6. Add H1 (or a second U6 with a stuffer strategy you understand) only when you need a second guide on the same plasmid. If you will deliver RNP or microinject RNA, design a T7 template and skip Pol III promoters entirely.

Species matters. Human U6 is standard in human cells; mouse U6 variants exist for murine work. Do not assume a zebrafish or plant CRISPR plasmid uses human U6. Read the feature table, then confirm against the source paper or vendor map.

Software should make the promoter, spacer, scaffold, and terminator visible as separate features. Zettalab's molecular biology tools are relevant when you are assembling that cassette in silico and need to see whether a T-stretch in the spacer collides with U6 termination before you order oligos. A CRISPR vector library search is useful only after you know which promoter class the experiment requires.

Plasmid Map Checks Before You Order Oligos

Verify six things on the annotated map. The promoter class matches delivery. The transcription start sits at the intended first guide base. The spacer orientation is sense relative to the scaffold, not reverse-complemented twice. The scaffold is complete. The terminator is present for Pol III cassettes. Cas9 or Cas12a, if on the same plasmid, is on a Pol II promoter and is a separate transcription unit.

Then translate none of the guide. An ORF arrow drawn across an sgRNA cassette is a mapping mistake that causes people to "optimize codons" in a noncoding RNA. Keep the cassette labeled as RNA.

Record the promoter choice in the cloning experiment, including which genome the spacer was designed against. Promoter and spacer are different decisions; swapping U6 for T7 later means a new template, not a silent plasmid edit.

FAQ

Which promoter should I use for sgRNA expression in mammalian cells?

Use a Pol III promoter, almost always human U6, for DNA-encoded sgRNA in mammalian cells. H1 is a reasonable second cassette on multiplex plasmids. T7 is the wrong default unless you also provide T7 RNA polymerase. If you deliver RNP, you may not need any cellular promoter on a plasmid; you need a T7 template for in vitro transcription. Confirm the cassette on the map rather than trusting the plasmid nickname. Many "all-in-one" CRISPR vectors use U6-sgRNA plus a Pol II Cas9, and mixing those labels is a common reading error.

Can I use a T7 promoter on a lentiviral CRISPR plasmid?

Not as the guide driver in ordinary mammalian packaging and infection. Lentiviral guide plasmids use Pol III promoters so the integrated cassette can be transcribed by the host. A T7 promoter will not be recognized in standard human or mouse lines. You can still have T7 on a separate bacterial expression plasmid or on a PCR template used to make RNA. If a map shows T7 next to an sgRNA on a lenti backbone, treat that as a red flag and re-read the feature table before packaging virus.

Why do U6-transcribed spacers sometimes start with an extra G?

U6 transcription often prefers a G as the first nucleotide. Design tools and cloning protocols therefore add a 5' G when the genomic spacer does not start with G. That extra base is part of the RNA, so off-target search should use the RNA that will actually be made, not only the 20-nt genomic string. If your nuclease is sensitive to 5' mismatches, record whether the extra G is present. Do not "correct" it away during codon-style cleanup; it is not a coding sequence.

When is H1 better than a second U6?

H1 is useful when you want two different Pol III promoters on one plasmid to reduce long identical repeats, which can recombine in E. coli and in viral genomes. It is not a guarantee of equal guide dose. If multiplex cutting is the goal, validate both spacers empirically. Some compact vectors use H1 because it is shorter. Check cloning overhangs; an H1 cassette may use different Type IIS sites than the lab's usual U6 oligo protocol.

How should promoter choice be documented in an ELN?

Write the promoter class, species source (human U6 versus mouse U6), cassette position, spacer sequence including any leading G, scaffold variant, and terminator. Attach the map version used to order oligos. If the experiment later switches from plasmid transfection to RNP, start a new design record for the T7 template rather than editing the old U6 entry in place. Connected workspaces such as Zettalab help when the map and the experiment note stay in the same project, but the scientific requirement is the same on any ELN: a reviewer must see which RNA was supposed to be made.

Conclusion

U6 is the workhorse promoter for mammalian sgRNA plasmids, H1 is the usual second Pol III cassette, and T7 belongs to in vitro transcription and specialized polymerase systems. Match the promoter to delivery, then confirm start site, spacer orientation, scaffold, and terminator on the map. Connected CRISPR and plasmid tools, including those in Zettalab, are useful when those features stay visible during cloning. If you are choosing a backbone rather than a promoter class, search annotated CRISPR vectors after the expression mode is fixed, and keep the cassette design in the experiment record.