Design: Problem Identification & Strategy
ALS is a neurodegenerative disease characterized by TDP-43 mislocalization and abnormal stress granule dynamics. Despite extensive research, there are currently no therapeutics that directly prevent TDP-43 aggregation. We identified two interconnected molecular problems and designed corresponding RNA-based solutions.
Strategic Framework
Problem 1
Stress granule proteins (DAZAP1, FAM98A, SND1) drive abnormal aggregation when dysregulated
Solution 1: ASOs
Antisense oligonucleotides to knock down expression of key stress granule proteins
Problem 2
TDP-43 mislocalization and aggregation is a hallmark of ALS pathology
Solution 2: RNA Aptamers
RNA aptamers to bind TDP-43 C-terminal domain and regulate demixing behavior
ASO Gapmer Design Optimization
Through iterative testing, we optimized a 5-10-5 gapmer design with chemical modifications to improve stability and binding affinity while minimizing off-target effects.
Click on each segment below to learn more about its modifications and purpose.
5' Flank (Positions 1-5)
Modification: 2'-O-methoxyethyl ribose (2MOEr)
Purpose: Improves nuclear resistance and binding affinity
DNA Gap (Positions 6-14)
Modification: 5-methylcytosine (iMe-dC) substitutions
Purpose: Central DNA region optimized for stability
3' Flank (Positions 15-19)
Modification: 2'-O-methoxyethyl ribose (2MOEr)
Purpose: Improves nuclear resistance and binding affinity
This systematic design approach balances chemical stability, binding strength, and biological relevance, producing ASOs suitable for in vitro testing.
Build: Computational Pipeline & Chemical Optimization
Our build phase involved both computational design and chemical optimization to create the most promising candidates for experimental testing.
Parallel Development Tracks
ASO Development
Computational Screening
Scored candidates based on GC content, secondary structure, conservation, and specificity
Chemical Modification
Optimized 5-10-5 gapmer with 2MOEr flanks and iMe-dC DNA gap
Primer Design & Validation
Created primers for DAZAP1, FAM98A, and GAPDH with efficiency testing
Aptamer Development
AptaTrans Generation
Deep learning framework predicted sequences targeting TDP-43 CTD
HADDOCK Docking
In silico simulations assessed binding stability and surface area
Monte Carlo Optimization
Tested multiple iterations and random seeds to identify top candidates
Both tracks employed iterative computational refinement to maximize the likelihood of experimental success while minimizing time and resource investment in wet-lab testing.
Test: Experimental Validation
We conducted two main experimental cycles to test our designs: ASO transfection with RT-qPCR analysis, and aptamer binding visualization with immunocytochemistry.
Preliminary qPCR Workflow Practice
Before conducting our main ASO transfections, we performed a preliminary qPCR experiment to familiarize ourselves with the workflow. We prepared a qPCR plate containing technical triplicates for GAPDH, TDP-43, METTL3, YTHDF2, CE, and Full constructs across multiple biological replicates. Different cDNA dilutions were tested to ensure amplification was within a detectable and consistent range.
Purpose
Practice plate setup, pipetting, and qPCR cycling workflow
Outcome
Gained confidence and troubleshooting experience before ASO knockdown experiments
Status
Not intended to generate final data - preparatory learning step
This preliminary experiment was crucial for ensuring technical proficiency before moving to our critical ASO validation experiments.
Primer Efficiency Validation
Standard curves were generated to validate primer performance. Target efficiency range: 90-110%.
DAZAP1 Primer Efficiency
Slope: -3.98
Efficiency: 78.4%
Status: Below target - optimize for future experiments
FAM98A Primer Efficiency
Slope: -3.46
Efficiency: 94.4%
Status: Optimal range ✓
GAPDH Primer Efficiency
Slope: -4.17
Efficiency: 73.7%
Status: Below target - optimize for future experiments
ASO Knockdown Results
RT-qPCR quantification revealed successful gene expression reduction for stress granule proteins.
DAZAP1
ASO-A & ASO-B
Both achieved 23% reduction in expression
FAM98A
ASO-C & ASO-D
Reduced expression by 38-49%
SND1
Primer Efficiency: 377%
Excluded from analysis; redesigning primers for future experiments
Detailed ASO Knockdown Analysis
RT-qPCR data showing relative gene expression (normalized to GAPDH) comparing ASO-treated samples to controls.
DAZAP1 Knockdown Results
Relative expression calculated using primer efficiency-corrected Pfaffl method. Both ASOs achieved ~23% knockdown compared to lipofectamine-only control.
FAM98A Knockdown Results
Relative expression calculated using ΔΔCt method. ASO-D achieved 49% knockdown, demonstrating strong efficacy for FAM98A targeting.
RNA Aptamer Cellular Localization
We performed two imaging trials to visualize aptamer localization and assess potential interactions with TDP-43 and stress granules.
Trial 1: Initial Merged Channel Imaging
RNA aptamer signal was visualized alongside DAPI (nuclear stain) and G3BP1 (stress granule marker). The overlap of aptamer and DAPI confirms successful cellular uptake and nuclear localization. The aptamer does not consistently co-localize with G3BP1-positive structures in these images, indicating the aptamer binds a structure whose molecular identity is currently unknown.
Trial 2: Refined Imaging with Z-Stack and Maximum Intensity Projection
To build on the first trial and gain a more detailed view of aptamer localization, we performed higher-resolution imaging using both single slices and a maximum intensity projection of a Z-stack. Sodium arsenite treatment was applied to ensure stress granule formation, and this 3D imaging approach allowed us to better assess potential colocalization between the aptamer, TDP-43, and G3BP1 across the entire cell volume, rather than relying solely on a single focal plane.
Improvement: Z-stack imaging provides superior spatial information for assessing colocalization across 3D cellular structures.
Although direct binding to TDP-43 was not fully confirmed, these images provided preliminary evidence that the aptamer reaches relevant subcellular locations for potential functional activity.
These findings confirmed that ASOs can feasibly reduce stress granule-related transcripts in neuronal-like cells. Aptamer experiments demonstrated successful cellular uptake and nuclear localization, with iterative imaging improvements providing better spatial resolution for future binding validation studies.
Learn: Engineering Cycle Analysis
The Design-Build-Test-Learn cycle is not complete until insights from testing inform the next iteration. Our experiments revealed both successes and areas for optimization.
Design-Build-Test-Learn Cycle
Click each phase to see what we accomplished and what we learned for the next iteration.
Design Phase
What We Designed:
- Computational pipeline for ASO sequence generation targeting DAZAP1, FAM98A, SND1
- 5-10-5 gapmer architecture with 2MOEr and iMe-dC modifications
- AptaTrans-generated RNA aptamers targeting TDP-43 C-terminal domain
- HADDOCK docking simulations with Monte Carlo optimization
Design Rationale:
Dual strategy addresses both upstream (stress granule proteins) and downstream (TDP-43 aggregation) pathological events in ALS.
Build Phase
What We Built:
- Computationally screened hundreds of ASO sequences, narrowed to top candidates
- Chemically optimized gapmers with iterative modification testing
- Validated primer sets for RT-qPCR quantification
- Generated and computationally ranked aptamer candidates
Build Achievements:
Created experimentally-ready ASOs and aptamers with predicted high efficacy and minimal off-target effects or immune activation.
Test Phase
What We Tested:
- ASO Experiment: Transfected SH-SY5Y cells, measured knockdown by RT-qPCR
- Results: DAZAP1 reduced 23%, FAM98A reduced 38-49%
- Aptamer Experiment: Immunocytochemistry with fluorescence microscopy
- Results: Nuclear localization confirmed, partial stress granule colocalization
Testing Challenges:
SND1 primer efficiency too high (377%); aptamer binding to TDP-43 not definitively confirmed.
Learn Phase
Key Learnings - ASOs:
- Current gapmer design is effective for knockdown in neuronal-like cells
- Primer efficiency varies; DAZAP1 and GAPDH need optimization
- SND1 requires complete primer redesign for future cycles
Key Learnings - Aptamers:
- Aptamers successfully enter cells and localize to nucleus
- Binding specificity needs validation with additional assays
- Z-stack imaging provides superior spatial information
Next Iteration:
ASOs: Test in stressed cells, optimize concentration, add more sequence variants. Aptamers: Refine sequences, develop binding assays, test functional effects on stress granule dynamics.
Future Directions: Cycle 2 Planning
Based on our learnings, we have identified specific improvements for the next engineering cycle.
ASO Optimization
Iteration 2Aptamer Refinement
Iteration 2Engineering success is not defined by perfect results, but by systematic iteration and learning. Our dual-strategy approach has demonstrated feasibility at the proof-of-concept level, and the clear path forward ensures continued progress toward an RNA-based therapeutic for ALS.