ASGCT 2026 Poster: Preclinical Data Supporting Design of TSHA-102 for Rett syndrome

Published: 2026-05-12 08:48:27 ET TSHA

Published on 05/12/2026 at 08:48 am EDT

‌Superior expression of self-complementary AAV and comparable functionality of mini and full-length MeCP2 support the design of TSHA-102 gene therapy for Rett syndrome

Emdadul Haque, Andrew D. Wallace, Ryan R. Chaparian, Rayvon Moore, Ian Jones, Annika N. Alicardi, Carlos D. Robles, Frederick Porter

Taysha Gene Therapies, 3000 Pegasus Park Dr., Dallas, Texas 75274, United States

Aims

Methods

Results

TSHA-102 (scAAV9-mini-MECP2) was evaluated against an analogous ssAAV9

Figure 3: Transduction with scAAV9 vs ssAAV9 vectors demonstrated greater transgene expression and ~30-fold greater protein expression across MeCP2 and

Figure 5: Protein functional analysis confirms mini-MeCP2 is functionally equivalent to FL-MeCP2 demonstrating a comparable half-life, mDNA binding, and DNMT3 inhibition

mini-MeCP2 and FL-MeCP2 proteins showed similar half-lives in

Background

Rett syndrome (RTT) is a rare, neurodevelopmental disorder caused by pathologic loss-of-function mutations

construct in which the mini-MECP2 and mutant inverted terminal repeat (ITR)

were replaced with FL-MECP2 and WT ITR, while maintaining identical backbone and regulatory elements

N2a cells expressing either WT MECP2 or the clinically relevant R294X allele (N2a R294X, RTT model cells) were transduced with either TSHA-102 or the ssAAV9-FL-MECP2 construct

Transgene and protein expression were quantified using reverse transcription droplet digital polymerase chain reaction (RT-ddPCR) and enzyme-linked immunosorbent assay (ELISA), respectively. The measured concentration of

GFP in RTT model cells

MeCP2

scAAV9 (TSHA-102, mini-MECP2) ssAAV9 (FL-MECP2)

1200

Relative transgene expression

1000

800

600

400

1.0

scAAV9 (TSHA-102, mini-MeCP2) ssAAV9 (FL-MeCP2)

0.5

Half-life*

mini-MeCP2 FL-MeCP2

% of protein (pmol/µg) remaining

100

mDNA binding*

mini-MeCP2 FL-MeCP2

Binding to mDNA (pmol MeCP2/mg cell lysate)

0.08

0.06

0.04

DNMT activity inhibition†

mini-MeCP2 FL-MeCP2

% DNMT enzyme activity relative to control

150

100

the RTT model cells (12 h vs 9 h, respectively; Figure 5A), indicating comparable steady-state abundance and protein synthesis rates

mini-MeCP2 and FL-MeCP2 proteins both demonstrated binding to mDNA with a similar magnitude (Figure 5B)

Both proteins inhibited DNMT3

in the X-linked MECP2 gene, encoding the MeCP2 protein1-3

For gene regulation, MeCP2 binds to both methylated (m) and unmethylated DNA and functions with partner proteins, including DNA methyltransferases (DNMTs), to modulate chromatin structure, DNA methylation and transcriptional activity4-7

A truncated MeCP2 protein retaining the essential methyl-CpG-binding domain (MBD) and the transcriptional repressor NCoR/SMRT interaction domain (NID) restored a normal phenotype in Mecp2-/Y mice and showed comparable functionality to full-length MeCP28

MeCP2 protein was within 25% of the coefficient of variation

N2a and N2a R294X cells were also transduced with ssAAV9-green fluorescent protein (GFP) or scAAV9-GFP vector. Transgene and protein expression were measured by RT-ddPCR and GFP fluorescence from live cells, respectively

N2a R294X cells were transiently transfected with equimolar plasmids using FuGENE 6 (Promega) to express a similar amount of MeCP2 protein (Figure 2)

200

GFP

scAAV9-GFP ssAAV9-GFP

Relative transgene expression

2.0x105

1.5x105

2x106

4x106 6x106 8x106 1x107

MOI (vg/cell)

0 1x106

scAAV9-GFP ssAAV9-GFP

800

600

2x106 3x106 MOI

4x106

5x106

Protein expression (pmol MeCP2/mg cell lysate)

0 20 40 60

Hours

*Following plasmid transfection in RTT model cells; † In a DNMT3A/3L assay

0.02

0.00

0 100 200 300 400 500

Plasmid concentration (fmol)

0 50 100 150 500

MeCP2 protein concentration (nM)

activity in a dose-responsive manner, with similar levels of inhibition observed (Figure 5C), suggesting that both proteins directly interacted with the DNMT enzyme

These findings established RTT is reversible in this model and the mechanistic basis for miniaturized

MECP2 gene therapy to restore expression in MeCP2-deficient cells

TSHA-102 is a recombinant scAAV9 gene therapy that encodes a functional, miniaturized MECP2 transgene that retains the essential MBD and NID,9,10 currently under investigation in the REVEAL pivotal trials11

In contrast to ssAAV vectors that require conversion of the single-stranded DNA to double-stranded DNA prior to expression, TSHA-102 contains complementary strands of the MECP2 transgene, delivering a

transcription-ready transgene to MeCP2-deficient cells for enhanced stability and efficiency9,12

Figure 2: Plasmid constructs

MeP426-mini-MECP2-miRARE

MeP426

mini-MECP2

miRARE

MeP426-FL-MECP2-miRARE

MeP426

FL-MECP2

miRARE

TSHA-102 vector map was used for mini- and FL-MeCP2

1.0x105

5.0x104

GFP protein expression (RFU)

5x105

1x106

400

200

5x105

1x106

Figure 6: RTT model cells demonstrate mini-MeCP2 normalizes the RTT multiomic signature resulting in improvement of dysregulated biomolecules to an equivalent magnitude to FL-MeCP2

Median = 0.943

Degree of improvement is similar between mini-MeCP2 and

FL-MeCP2 on the global scale

90th Percentile

Improvement ratio

Improvement

mini-MeCP2

ImprovementFL-MeCP2

10th Percentile

1 = same molecular effect Interpretation >> or << 1 = distinct

molecular effect

Conclusions

scAAV9 enables markedly higher (30x) MeCP2 protein expression than an analogous ssAAV9 for MeCP2 delivery, consistent with the literature from

A miRNA-Responsive Auto-Regulatory Element (miRARE) further refines expression by silencing transgene expression in MeCP2-healthy cells while permitting expression in MeCP2-deficient cells9,10

MOI

other sc- and ss-model vector comparisons12-14

mini-MeCP2 is functionally comparable to FL-MeCP2 across biochemical

scAAV vectors have been shown to drive robust transgene expression across disease models and genetic payloads,12-14 yet a direct comparison with ssAAV for MECP2 delivery in RTT models is lacking. Furthermore, although miniaturized (mini-)MeCP2 is sufficient to rescue the disease phenotype in MeCP2-deficient mice,9,10 a comprehensive comparison of the biochemical and molecular properties of mini-MeCP2 and full-length

(FL-MeCP2) versions of MeCP2 have not been fully characterized

The current study addresses both of these gaps through a multi-level panel of functional assays.

Complementary assays compared scAAV vs ssAAV and mini-MeCP2 vs FL-MeCP2 at three levels: vector delivery, transgene expression, and biochemical/molecular function (Figure 1)

Figure 1: Overview of experimental design for scAAV9 vs ssAAV9 and mini- vs

Half-life: Transfected cells were treated with 15 µM cycloheximide, and cellular extracts were prepared at various time intervals (cycloheximide chase experiment15). MeCP2 levels were quantified by ELISA

When looking at MeCP2 expression in RTT model cells, the scAAV9 vector (TSHA-102) produced up to

Improvement ratio

30-fold more MeCP2 protein than the ssAAV9 vector (ssAAV9-FL-MECP2), which expressed an 2

insignificant amount of MeCP2 protein (Figure 3A, right panel)

Using a GFP model transgene in RTT model cells, scAAV9 expressed ~30-fold more GFP transgene and GFP protein compared with ssAAV9, for which expression was negligible (Figure 3B)

Figure 4: Following plasmid transfection, mini-MeCP2 protein expression was comparable to FL-MeCP2 and upregulated 3.5- to 4-fold in RTT model vs WT cells via

miRARE-driven regulation

Proteins

functions, with both proteins exhibiting similar behavior in RTT model cells:

Both mini-MeCP2 and FL-MeCP2 are stably expressed in neuronal cells. miRARE controls protein expression in a genotype-dependent manner, independent of whether the expressed protein is

mini- or-FL-MeCP2

The half-life of the mini-MeCP2 protein is similar to the FL-MeCP2 protein, suggesting comparable protein stability and supporting

the likelihood of similar biological functions

Both mini-MeCP2 and FL-MeCP2 similarly bind to mDNA, which is a

FL-MeCP2 comparisons

Binding to mDNA: Cell lysates were prepared from transfected cells and DNA-binding (D-)ELISA performed16,17 using microplates coated with custom biotin-labeled 45 base pair (bp) oligonucleotides containing three methylated CpG sites

DNMT inhibition: DNMT3A/3L complexes (Active Motif) were incubated

with different molar amounts of either mini-MeCP2 or FL-MeCP2 recombinant

WT N2a cells

Protein expression (pmol MeCP2/mg cell lysate)

0.3

0.2

N2a R294X cells

Protein expression (pmol MeCP2/mg cell lysate)

0.3

0.2

RTT model cells displayed highly dysregulated biomolecule profiles, with 5023 proteins, lipids and metabolites dysregulated compared with WT cells. Treatment with either expression plasmid (TSHA-102 mini-MECP2 or FL-MECP2) resulted in improvement of dysregulated biomolecules to an equivalent magnitude (Figure 6), suggesting that the activity of mini-MeCP2 protein is

fundamental process for establishing and maintaining normal gene expression patterns, essential for brain health18

mini-MeCP2 inhibits DNMT enzymatic activity, further showing that

mini-MeCP2 is functionally analogous to FL-MeCP2

ssAAV9-FL-MECP2 ssAAV9-GFP

scAAV9-mini-MECP2 scAAV9-GFP

Arm 1: AAV9 transduction

AAV-based assays

Evaluation of transgene and protein expression

GFP fluorescence: Transduction efficiency

RT-ddPCR: Transgene mRNA quantification

ELISA (protein quantification): mini-MeCP2/FL-MeCP2 protein expression

Tested at equivalent multiplicity of infection (MOI)

Arm 2: Plasmid transfection

Plasmid constructs tested

mini-MECP2 plasmid

FL-MECP2 plasmid

Plasmid-based assays

Evaluation of protein function

ELISA (protein half-life): Stability of mini- vs FL-MeCP2 protein

D-ELISA: mDNA binding to assess MBD function comparison

Multiomics: System-level functional comparison

Tested at equivalent plasmid and protein level

proteins. The Epiquick DNMT Activity/Inhibition Assay Ultra kit (Epigentek) was used to quantify DNMT activity

Multiomics analysis was performed on cell lysates of transfected cells using liquid chromatography/mass spectrometry (LC/MS) to define biomolecule profiles

Statistical associations between biochemicals and transfection groups were tested using linear regression models. Biomolecules were considered significant

0.1

0.0

ns, non-significant

mini-MeCP2

FL-MeCP2

0.1

0.0

mini-MeCP2

FL-MeCP2

comparable to MeCP2 protein across multiple classes of biomolecules on a cellular level

Key takeaway

- Almost no effect on biomolecule dysregulation was observed in WT background (dysregulated 1 and 8 for mini-MeCP2 and FL-MeCP2, respectively), suggesting no off-target activities in WT background as expected due to miRARE

- mini- and FL-MeCP2 proteins produce similar improvements across

dysregulated multiomics signatures, indicating preservation of core MeCP2 function, resulting in broad cellular correction with mini-MeCP2 equivalent to that of FL-MeCP2

Cell lines: Neuro2a (N2a; wildtype [WT]), N2a-MeCP2 R294X (N2a R294X, RTT model cells)

hits if the adjusted p-value was ≤0.05

The fold change (FC) in protein abundance was compared from baseline to post-transfection; biomolecule regulation was considered improved if their FC shifted closer to 1 (indicating no dysregulation) by at least 20%

To compare mini-MeCP2 and FL-MeCP2, the improvement ratio was calculated by dividing the ΔFC of mini-MeCP2 by the ΔFC of MeCP2, with a ratio of 1 indicating the same molecular activity

mini-MeCP2 and FL-MeCP2 both achieved comparable, stable expression in neuronal cells when transfected with equimolar amounts of plasmid (Figure 4)

mini-MeCP2 and FL-MeCP2 exhibited the same expression pattern in WT and RTT model cells

Protein levels were approximately 3.5- to 4-fold higher in RTT model cells compared with WT cells, consistent with miRARE-driven genotypic regulation of expression

Presenting author: Ryan Chaparian ([email protected])

Acknowledgments: We thank Steven J Gray (University of Texas Southwestern Medical Center) for providing GFP vectors, Jeffrey L Neul (Vanderbilt University) for providing N2a-R294X cells, and Dalton Bioanalytics Inc. for conducting the multiomics LC-MS experiments. This study was funded by Taysha Gene Therapies. Medical writing support for this poster was provided by Leigh O'Connor-Jones, PhD, of Avalere Health US Inc., funded by Taysha Gene Therapies Disclosures: All authors are employees of Taysha Gene Therapies and hold equity in the company

Abbreviations: bp, base pair; DNMT, DNA methyltransferase; (D-)ELISA, (DNA-binding) enzyme-linked immunosorbent assay; FC, fold change; FL, full-length; GFP, green fluorescent protein; h, hour; ITR, inverted terminal repeat; LC, liquid chromatography; mDNA, methylated DNA; MeCP2, methyl-CpG binding protein 2; miRARE, miR-Responsive Auto-Regulatory Element; MOI, multiplicity of infection; MS, mass spectrometry; N2a, Neuro2a cells; NCoR, nuclear receptor corepressor; NID, NCoR/SMRT interaction domain; ns, non-significant; RFU, relative fluorescence units; RT-ddPCR, reverse transcription droplet digital polymerase chain reaction; RTT, Rett syndrome; scAAV(9), self-complementary adeno-associated virus (serotype 9); SMRT, silencing mediator of retinoid and thyroid hormone receptors; ssAAV(9), single-stranded adeno-associated virus (serotype 9); vg, vector genome; WT, wild type

References: 1. Samanta D. Front Neurol. 2026;17:1766679; 2. Amir RE, et al. Nat Genet. 1999;23:185-188; 3. Amir RE, et al. Ann Neurol. 2000;47:670-679; 4. Nan X, at al. Cell. 1997;88(4):471-481;

5. Yasui DH, et al. Proc Natl Acad Sci USA. 2007;104(49):19416-19421; 6. Pejhan S, Rastegar M. Biomolecules. 2021;11(1):75; 7. Rajavelu A, et al. Nucleic Acids Res. 2018;46(17):9044-9056; 8.Tillotson R, et al. Nature. 2017;550(7676):398-401; 9. Sinnett SE, et al. Brain. 2021;144(10):3005-3019; 10. Sadhu, et al. Genes. 2024;15(1):31; 11. Rossignol E, et al. REVEAL Adolescent/Adult and Pediatric Clinical Trial Update: Safety and Efficacy Data on TSHA-102 AAV9 Investigational Gene Therapy in Clinical Evaluation for Rett Syndrome. Presented at IRSF Rett Syndrome Scientific Meeting 2025: June 9-11, Boston MA, USA;

12. McCarty DM. Mol Ther. 2008;16 (10) :1648-1656; 13. Bailey RM, et al. Mol Ther Methods Clin Dev. 2018;9:160-171; 14. Gray SJ, et al. Mol Ther. 2011;19(6):1058-1069; 15. Wallace AD, Cidlowski JA.

J Biol Chem. 2001;276(46):42714-42721 16. Underwood KF, et al. J Vis Exp. 2013;(78):50512; 17. Renard P, et al. Nucleic Acids Res. 2001;29(4):E21; 18. Tillotson R, Bird A. J Mol Biol. 2020;432(6):1602-1623

ASGCT Annual Meeting: May 11-15, 2026, Boston, MA, USA

Disclaimer

Taysha Gene Therapies Inc. published this content on May 12, 2026, and is solely responsible for the information contained herein. Distributed via Public Technologies (PUBT), unedited and unaltered, on May 12, 2026 at 12:47 UTC.