This is a scientific skepticism task where I need to critically evaluate four hypotheses about selective 4R-tau silencing via Cas13 RNA editing. Let me go through each hypothesis systematically.
## General Context
- MAPT (Microtubule-Associated Protein Tau) gene has two major isoforms: 3R-tau and 4R-tau
- 3R-tau excludes exon 10; 4R-tau includes exon 10
- The ratio of 3R:4R tau is important for neuronal function
- Dysregulation of this ratio is implicated in various tauopathies (CBD, PSP, etc.)
## Critical Evaluation Framework
For each hypothesis I need to:
1. Identify specific weaknesses and challenges
2. Cite potential counter-evidence
3. Propose falsification experiments
4. Assign revised confidence scores
Let me work through each:
---
### Hypothesis 1: Splice-Site-Adjacent mRNA Targeting
**Core claim:** By targeting the 5' splice site of exon 10, Cas13 can discriminate 4R vs 3R mRNA because 3R mRNA skips exon 10 entirely.
**Weaknesses:**
1. **Sequence overlap issue:** The 5' splice site (GUAGGG) is part of exon 10. While 3R-tau excludes exon 10, the pre-mRNA still contains this sequence in the nascent transcript. Cas13 acts on nascent RNA in the nucleus, so it could potentially bind to the pre-mRNA before splicing occurs.
2. **Off-target risk:** The GUAGGG motif is not unique to MAPT. This is a very short 6-nucleotide target, which increases off-target potential significantly.
3. **Kinetic competition:** Splicing occurs co-transcriptionally very rapidly.
Cas13 cleavage might not outpace the splicing machinery before exon 10 gets excised in 3R transcripts, undermining the intended specificity. Additionally, the targeted region sits within the coding sequence of 4R-tau, so even successful cleavage could produce truncated proteins with unpredictable consequences. There's also the fundamental challenge that 3R and 4R share the majority of their sequence, making any targeting strategy that relies on differential expression inherently difficult to execute with precision.
Alternative splicing of MAPT is tightly regulated with multiple splicing factors (SR proteins, hnRNPs) binding at the exon 10 splice site, which means an RNA-targeting approach might fail to outcompete these natural mechanisms. The specific 5' splice site sequence, while somewhat distinctive, may not be unique enough to avoid targeting other transcripts with similar motifs.
Testing this hypothesis would require designing gRNAs against the exon 10 5' splice site and measuring changes in 3R:4R ratio after Cas13 expression—ideally in a cell line that naturally produces both isoforms. If 3R levels remain unchanged, that would suggest the approach lacks the necessary specificity to achieve the desired discrimination.
Beyond that, I should examine whether any transcripts unrelated to MAPT share the same 5' splice site sequence and how the efficiency of Cas13 targeting stacks up against the rate at which splicing occurs. The current evidence seems to support this concept in other diseases like SMA, though the mechanistic basis for that precedent remains unclear.
The confidence score probably needs adjustment downward given the off-target concerns with a short 6-nucleotide target and the technical difficulty of outpacing co-transcriptional splicing. Something closer to 0.45-0.55 seems more realistic than the initial estimate of 0.65.
Moving to the second hypothesis about disrupting the branch point and polypyrimidine tract in intron 10—targeting these elements could theoretically shift the 3R:4R ratio toward more 3R expression. The branch point adenosine and polypyrimidine tract are essential for splicing at the 3' splice site of exon 10, so if these sequences are disrupted or blocked, exon 10 splicing could be compromised. This would preferentially reduce 4R-tau production. The key advantage here is that these sequences are only present in 4R-tau pre-mRNA since 3R-tau excludes this region entirely, making this approach potentially more specific than targeting the splice site itself.
However, there are significant technical obstacles. The branch point is located within intron 10, which requires Cas13 to access the nascent pre-mRNA in the nucleus with high efficiency—something that's proven difficult with current delivery methods. Additionally, the branch point sequence is relatively short (around 30-40 nucleotides), limiting the gRNA design options. Even with successful targeting, disrupting this region might simply reduce 4R-tau expression rather than eliminate it, and there's a risk of interfering with normal splicing of other genes that share similar branch point consensus sequences.
The specificity concern is substantial: the branch point consensus (YNYURAY in humans) appears frequently in transcripts genome-wide, and while intron 10 is large (approximately 10kb), off-target effects remain plausible. Moreover, the splicing machinery might adapt or compensate for partial disruption of the branch point region, diminishing the effect over time. There's also the possibility that 4R-tau uses alternative branch points within intron 10, similar to what occurs in other disease contexts. The mechanistic prediction itself is problematic—disrupting the branch point could trigger exon 10 skipping across both isoforms rather than selectively reducing 4R-tau, since this branch point is used by both splicing pathways.
Splicing occurs on a tight timeline (approximately 60-90 nucleotides from the branch point to the spliceosome assembly), so Cas13 may not have sufficient time to interact with the target before splicing is complete. Current evidence for branch point targeting is largely theoretical; functional studies are limited. Testing this hypothesis would require mapping the branch point using available datasets like branch-point-seq, then designing gRNAs to match it and assessing whether splicing is disrupted in the expected manner.
One key control would be validating that off-target genes with similar branch points show no splicing changes, and testing whether the reduction is actually due to the branch point disruption or simply to RNA cleavage elsewhere in the intron. If 4R-tau expression doesn't change despite successful Cas13 delivery, or if the effect persists even when the branch point is restored, that would suggest the mechanism isn't working as proposed. This hypothesis is more speculative than the first, making it harder to execute but potentially offering higher specificity if successful. I'm estimating around 0.35-0.45 confidence given the technical demands and mechanistic uncertainty.
For the alternative polyadenylation hypothesis, there's documented support for variable polyA sites in MAPT, and this would be the most selective approach if true since polyA signals are inherently transcript-specific. The mechanism is also distinct from splicing, avoiding some complications. However, alternative polyA sites are less well-characterized in this system and the efficiency could be inconsistent. Targeting the 150-200 nucleotide region might miss the actual polyA site, and even if the site is confirmed, simply targeting nearby doesn't guarantee the signal will be disrupted. There's also the risk that 3R-tau transcripts might use overlapping polyA signals, and Cas13-mediated cleavage could trigger broader RNA degradation rather than precise silencing.
The mechanistic concern is significant: if the alternative polyA site itself is part of intron 10, Cas13 cleavage might not prevent 4R-tau export if cleavage occurs downstream of the splice point, and the cleaved mRNA fragment could still be spliced normally. More critically, polyA site choice depends heavily on