You are here

How I really feel about five-mispair oligos

I try to send this message with any designs involving five-mispair oligos. A brief version is farther down.


** Argument for using a different specificity control instead of the five-mispair

I don't like the five mispair experiment. In the "Guidelines for morpholino use in zebrafish" (useful regardless of your model system), Stainier et al. write: "5-base mismatch MO ... cannot serve as controls for the specificity of the experimental MO."[1] It is not surprising if a five mispair oligo causes some phenotype. Many five mispair oligos do not cause phenotypes when used at the same concentration for which the targeted oligo is just high enough concentration to cause a phenotype. However, we do occasionally hear reports of phenotypes associated with five mispair oligos, which makes interpreting the five mispair experiment very difficult.

It is common practice to use two non-overlapping oligos targeting the same RNA as a specificity control set. If the oligos (used in separate experiments) phenocopy one another, that indicates that the phenotype is most likely due to interaction with the target RNA and not an unexpected interaction with a different RNA. For an oligo-based specificity control, I prefer the two-nonoverlapping-oligo technique instead of the five mispair technique. The two non-overlapping oligo experiment directly addresses the question "is the phenotype I am observing due to knockdown of my targeted gene or due to an off-target RNA interaction?" The five mispair experiment, in contrast, addresses the question "if a Morpholino has five mispairs distributed though its sequence, will it still elicit the same phenotype as my targeted oligo when the same dose is administered?" I think the first question is far more important than the second. The second non-overlapping oligo can also be co-injected with the first targeted oligo to test dose synergy in eliciting phenotype, which if observed both:
- provides further support for specificity of the knockdown, and
- may be useful as a tactic to cause a more complete knockdown. [2]

When an mRNA rescue works, it is also a very good specificity control. Unfortunately, developmental changes due to ectopic expression from a rescue mRNA sometimes occludes successful rescue from a knockdown phenotype.

My favorite specificity control is where a Morpholino is tested in a mutant that is a null for the Morpholino target, then subsequently tested in a wild-type organism. For some thoughts on using an engineered mutant as a Morpholino specificity control, see Stainier et al.'s discussion [1].

[1] Stainier DYR, Raz E, Lawson ND, Ekker SC, Burdine RD, Eisen JS, Ingham PW, Schulte-Merker S, Yelon D, Weinstein BM, Mullins MC, Wilson SW, Ramakrishnan L, Amacher SL, Neuhauss SCF, Meng A, Mochizuki N, Panula P, Moens CB. Guidelines for morpholino use in zebrafish. PLoS Genet. 2017 Oct 19;13(10):e1007000. doi: 10.1371/journal.pgen.1007000.

[2] Bill BR, Petzold AM, Clark KJ, Schimmenti LA, Ekker SC. A primer for morpholino use in zebrafish. Zebrafish. 2009 Mar;6(1):69-77.


Let me rephrase the discussion regarding mismatched controls, presenting my opinion more directly. Mismatched "controls" are scientifically worthless, prone to have unexpected RNA interactions that make interpretation difficult, devoid of useful information about specificity even when there is no associated phenotype, and reviewers are finally realizing that. The zebrafish community paper I cited above says that mispair oligos are unacceptable as specificity controls. I strongly suggest you select a different control strategy. That paper (Stainier et al. 2017) is a very good place to start for evaluating controls for steric-blocking oligos (like Morpholinos).

Add new comment