What exactly is it you're expecting the ASOs to do, here? What chemistry do they use?

My experience of ASOs is primarily using them as steric masking agents (exon skipping/splice modulation, or interfering with microRNAs), whereas you seem to be expecting ASO-mediated target cleavage?

Are you basically doing siRNA, here?

Anyway, aside from that: a 7.5% increase in mRNA level, using qPCR, with only a single reference gene (which is also GAPDH), is very unlikely to be anything more than measurement noise (or biological noise).

To be sure this was real, you'd need a ton of biological replicates, and two or even three stable reference genes (how confident are you that GAPDH is even an appropriate choice, here?).

After all that, you'd still be looking at a scenario of "even if real, is 7.5% increase particularly meaningful?" i.e. would this tiny shift actually make a difference, biologically? Especially since gene expression is inherently noisy and sloppy anyway.

A similar argument could be applied to the 23% decrease: this simply isn't a particularly meaningful fold change.

Could you be measuring exon skipping instead? (i.e. you're forcing exclusion of exon 15, with a ~25% efficiency, such that you measure a modest reduction in exon 15 levels, but no change in exon 6 levels?)

What results do you get if you use random priming instead of oligodT?

Do you have biological and technical replicates of these data?

If you are seeing consistent results, that suggests that you are seeing splice variants. These are VERY common in some organisms (e.g. human).

• While inefficiently, RNA primes itself, RNA primes other RNA, and DNA fragments also prime RNA, beyond the intended oligodT priming. Did you compare with random primers and with target-specific priming?
• Sometimes, RNA gets cleaved but not fully digested/degraded, likely because of protective factors (like RNA binding proteins, secondary structures, etc.).
• Normalization against a single reference RNA is not recommended. Use at least two validated (stably expressed reference genes).
• Cleaved RNA can be measured by ligating an oligo 5' and 3' of the cleavage site, and doing RT-qPCR with a target specific primer and a primer complementary to the ligated oligo (see RACE strategy in

  Article Tissue-specific gene silencing monitored in circulating RNA

  , but also other similar methods).

John Hildyard

Thank you for your answer

1. ASO mechanism

I am using phosphorothioate (PS)-modified ASOs designed to induce RNase H–mediated cleavage, with the goal of suppressing protein expression and knocking down the function of oncogenic proteins.

2. Choice of GAPDH as reference gene

In my experimental conditions, GAPDH mRNA and protein levels have remained essentially constant. Before cDNA synthesis, I roughly quantify RNA concentration using a Nanodrop (Denovix) and use the same amount of RNA (in ng) for each sample. Even if HER2 mRNA levels decrease, its contribution to the total mRNA pool is minimal. In practice, qPCR consistently shows nearly identical Cq values for GAPDH regardless of ASO treatment or cleavage site, so at this stage I consider it an appropriate reference gene.

3. Biological significance of observed changes

As you pointed out, the 7.5% increase or 23% decrease observed by RT-qPCR is unlikely to be biologically meaningful. I am screening multiple ASO candidates, and ASO15 is just one of them. Since the qPCR results were difficult to interpret, I first ran Western blots, which showed that ASO15 had lower knockdown efficiency compared to other ASOs. Based on this, I plan to discard ASO15.

4. Possibility of splicing effects

My intention has been to trigger RNase H–mediated cleavage, so I had not considered splicing changes as a major factor. However, as you suggest, even PS-only ASOs could, depending on their sequence and target site, unintentionally influence splicing. That is an interesting point I will keep in mind.

5. Priming method

Your suggestion to try random priming is reasonable and aligns with common recommendations. I plan to test it.