Micro-Dystrophin Codon Optimization for DMD AAV Gene Therapy

Rare Disease

Micro-Dystrophin Codon Optimization for DMD AAV Gene Therapy

Gene therapy biotech needed a codon-optimized micro-dystrophin construct that maximizes expression in skeletal muscle while fitting AAV packaging constraints and avoiding immunogenic epitopes.

Company

Gene therapy biotech (Series A, 22-person team)

Timeline

July to August 2025

Engagement

Codon Optimization & Expression Prediction Pipeline

AAV Codon Optimization

14 days: Computational delivery
3.2×: Expression vs. prior construct
100%: AAV packaging compliance (top 5)
4/5: Constructs passed NHP expression threshold

The Challenge

A gene therapy biotech was developing an AAV9-delivered micro-dystrophin therapy for Duchenne muscular dystrophy. Their first-generation construct expressed poorly in NHP skeletal muscle (12% of native dystrophin levels) despite acceptable manufacturing titers. The 3.8 kb coding sequence left minimal room for regulatory elements, and repeat redesign cycles with their CRO had consumed 4 months without meeting the 30% dystrophin restoration threshold required for their preclinical go/no-go decision.

Business Constraints

Budget: $340K (construct optimization line item)
Timeline: Ranked codon-optimized sequences in 14 business days
Must fit AAV packaging limit (<4.7 kb total construct including ITRs and promoter)

GeneForge Approach

Days 1 to 5: Construct Analysis and Codon Variant Generation

Input

First-generation micro-dystrophin sequence (3,812 bp coding region)
AAV9 packaging constraints (4.7 kb total including CAG promoter and polyA)
Human skeletal muscle codon usage tables (GTEx v8 muscle tissue)
Known DMD patient T-cell epitopes (IEDB, prior clinical trial data)

Methods

2.1M codon variant generation preserving amino acid sequence
Codon Adaptation Index (CAI) and tRNA adaptation scoring for human muscle
mRNA secondary structure prediction (RNAfold) for translation efficiency
Cryptic splice site and premature polyA signal scanning

Output

Top 4,200 variants ranked by predicted muscle expression score
Immunogenic epitope map across all variants

Days 6 to 11: AAV Compatibility and Expression Simulation

GC content and repeat motif filtering eliminated 680 variants with manufacturing risk (homopolymer runs, high GC peaks). CpG dinucleotide optimization reduced innate immune activation risk. In silico expression simulation (Optimus 5-Prime + proprietary muscle-specific model) predicted protein output for top 300 variants. AAV packaging size verification and ITR proximity checks filtered to 85 construct-ready candidates. Immunogenic T-cell epitope reduction scoring removed 12 variants with known DMD patient reactive epitopes.

Days 12 to 14: Ranking and Preclinical Validation Protocol

Output

Top 10 codon-optimized micro-dystrophin sequences with expression score, CAI, and manufacturability tier
Recommended NHP study design (dose, timepoints, dystrophin quantification by IF and Western)
CMC-ready sequence files with restriction map for plasmid assembly
Comparative expression prediction report vs. first-generation construct

Final Ranked Codon-Optimized Constructs

Top 5 shown; full list of 10 delivered with CMC-ready sequence files.

Top micro-dystrophin constructs by predicted muscle expression and AAV compatibility
Rank	Construct ID	Expression Score	CAI	Packaging (kb)	Status
1	GF-c001	0.93	0.91	4.62	Priority A
2	GF-c002	0.90	0.89	4.58	Priority A
3	GF-c003	0.88	0.88	4.65	Priority A
4	GF-c004	0.86	0.87	4.61	Priority A
5	GF-c005	0.84	0.86	4.59	Priority B
6–10	GF-c006–c010	0.78–0.83	0.82–0.85	4.55–4.66	Backup

Results and impact

Speed, validation, and business outcomes

Speed vs. Prior CRO Redesign Cycles

Metric	Prior CRO Approach	GeneForge	Improvement
Timeline	4 months (2 cycles)	14 business days	6× faster
Cost	$420K (2 failed cycles)	$340K	19% savings + success
Variants Evaluated	~20 manual redesigns	2.1M codon variants screened	105,000× larger search space
NHP Expression	12% native dystrophin	38% native dystrophin (GF-c001)	3.2× expression improvement

NHP Validation Outcomes (10 weeks post-delivery)

AAV9 intramuscular injection in cynomolgus macaques; dystrophin quantification at 8 weeks post-dose.

Construct	Predicted Expression	Dystrophin (% native)	Titer Achieved	Notes
GF-c001	0.93	38%	4.2×10¹³ vg/mL	Selected for GLP tox study
GF-c002	0.90	34%	3.8×10¹³ vg/mL	Backup construct
GF-c003	0.88	31%	4.0×10¹³ vg/mL	Met 30% threshold
GF-c004	0.86	29%	3.6×10¹³ vg/mL	Borderline; deprioritized
GF-c005	0.84	26%	3.9×10¹³ vg/mL	Below threshold

3.2×: Expression improvement vs. Gen 1 construct
4/5: Constructs met or approached 30% threshold
100%: AAV manufacturing success (top 5)
10 wks: Timeline to NHP expression data

Immediate Wins

Preclinical go/no-go passed: GF-c001 achieved 38% dystrophin restoration, exceeding 30% threshold
GLP tox study initiated: 8 weeks ahead of schedule after failed CRO redesign cycles
Series A extension: NHP data supported $45M extension close with validated construct

Strategic Advantages

Muscle-specific codon optimization outperformed generic human codon tables by 40% in predicted expression
Immunogenic epitope removal prevented T-cell reactivity risk flagged in prior DMD trials
CMC-ready files with restriction maps accelerated plasmid assembly by 3 weeks

Follow-on engagement
Q1 2026: promoter and UTR optimization for systemic AAV delivery. Estimated cost: $275K. Target: improved transduction efficiency across all muscle groups.

Model validation

Lessons and recommendations

What Worked

mRNA secondary structure optimization contributed more to expression gains than CAI alone (ablation study: −22% expression without structure scoring)
CpG reduction improved manufacturing consistency (titer variance dropped from ±40% to ±12%)
2.1M variant search found a codon combination missed by 4 months of manual CRO redesign

Challenges and Mitigations

GF-c004 showed strong in silico expression but fell just below the 30% NHP threshold. Root cause: cryptic splice site activated in NHP but not predicted in human muscle model.

Mitigation: Added cross-species splice site prediction (Human + Macaca fascicularis); integrated into standard DMD construct pipeline.

One backup construct (GF-c007) produced lower AAV titer due to a 42 bp direct repeat introduced during codon shuffling.

Mitigation: Added repeat motif detection filter (≥15 bp direct repeats) before final ranking.

When to use GeneForge for AAV codon optimization

AAV gene therapy constructs with tight packaging size constraints
Prior codon optimization producing sub-threshold expression in preclinical models
Programs requiring immunogenic epitope reduction for repeat-dosing strategies
Timeline pressure before GLP tox or IND-enabling studies

ROI: approximately 6:1 (avoided failed CRO cycles + accelerated GLP tox + financing milestone value).

Next steps: pair codon optimization with promoter engineering and tissue-specific regulatory element selection for systemic delivery.

About This Engagement

Client profile: Series A gene therapy biotech, 22 employees, prior AAV manufacturing experience
Project duration: 14 business days (computational delivery) + 10 weeks (NHP validation)
Total cost: $340K
Date: July to August 2025

This case study is anonymized at client request. Construct sequences and institutional affiliations have been redacted. Full protocols available under NDA.

Run a similar program with GeneForge.

Tell us about your target. We will scope a 2 to 4 week pilot.

Schedule a Consultation [email protected]