Therapeutic potential of highly functional codon-optimized microutrophin for muscle-specific expression

Intramuscular administration of rAAV9-μUtrn leads to robust expression of µUtrns and muscle function improvement in mdx mice

Previous studies provided compelling evidence that murine and human µUtrns ΔR4-R21/ΔCT (Fig. 1A) can functionally compensate for the lack of dystrophin in mouse models of DMD^17,19,20,21. Building on their findings, we designed a human version of µUtrn ΔR4-R21/ΔCT and customized it through codon optimization in order to enhance expression in striated muscles and heart, as described in Materials and Methods (Fig. 1B). The human µUtrn sequence had 37% optimal codons, and this index reached 100% after optimization (Supplementary Fig. S1). For direct comparison, the murine, human, and codon-optimized human µUtrn cDNAs (designated in this study as M-µUtrn, H-µUtrn, and Hco-µUtrn, respectively) were cloned into the AAV vector expression cassette under a cytomegalovirus (CMV) promoter. As a control, a similar AAV construct with the cDNA of microdystrophin ΔR4-R23/ΔCT¹⁶ was created. To produce rAAV viruses encoding µUtrns (rAAV-CMV-µUtrn), we used AAV serotype 9, which is currently being tested in clinical trials for microdystrophin delivery (NCT003362502, NCT03368742).

To characterize different µUtrn sequences in dystrophic mice, rAAV9-CMV-µUtrn viruses were administered to 7-week-old mdx mice via intramuscular injections of 2 × 10¹² vg into the tibialis anterior (TA) muscle. rAAV9-CMV-µDys viruses were used as a control. The expression, potency, and immunogenicity of µUtrns were assessed after 2 weeks. As expected, the CMV promoter drove robust mRNA expression in the muscle tissue for all three µUtrn-coding vectors (Fig. 2c). H-µUtrn transcripts were the least abundant among M-µUtrn and Hco-µUtrn. Recombinant utrophin production was confirmed via western blotting analysis with antibodies against the N-terminus of the utrophin protein (Fig. 2A). Quantification of the protein bands revealed lower expression levels for H-μUtrn (Fig. 2B), in agreement with mRNA expression. Importantly, codon usage bias in Hco-µUtrn did not compromise protein translation within murine muscles. Immunofluorescence analysis with antibodies against utrophin indicated that µUtrn proteins successfully localized to the sarcolemma of muscle cells, where they can perform their function, substituting dystrophin protein in the DAGC of mdx mice (Fig. 2D). Immunostaining of treated muscles with antibodies against DAGC proteins α-sarcoglycan and α1-syntrophin indicated that, in comparison with non-treated mdx muscles, the levels of both increased considerably (Fig. 2D,E).

Striking improvements of TA contractile properties were observed in rAAV9-µUtrn-treated mdx mice, with specific force increasing roughly twofold relative to that in untreated mdx mice (Fig. 2G). Similarly to µDys, H-µUtrn exhibited a pronounced effect. All miniaturized proteins (H-µUtrn, M-µUtrn, Hco-µUtrn, µDys) ensured muscle susceptibility to repeated eccentric contraction at levels comparable to those in control B10 mice (Fig. 2F). Force drop dynamics were indistinguishable between Hco-μUtrn-treated, wild-type control, and µDys group muscles, indicative of the high functionality of the recombinant protein.

In order to verify the lower immunogenicity of utrophin, we assessed the presence of CD8 + cytotoxic T-lymphocytes (CTLs) in muscle sections. Lymphocytes were detected around muscle fibers in all analyzed groups (Fig. 2H), which is typical for the dystrophic environment and has been previously investigated by other groups²⁵. We observed a decrease in CTL count in M-µUtrn- and Hco-µUtrn-treated muscles compared to untreated mdx muscles reflecting lowering of inflammation, while infiltration rate remained unchanged in the µDys-treated group (Fig. 2I). At the same time µDys-treated muscles showed similar functional recovery as µUtrn-treated, indicating successful expression and functionality of all transgenes (Fig. 2F,G). One can speculate that µDys may be targeted by CTLs to a greater extent than µUtrn, counterbalancing reduced dystrophy-associated infiltration as a consequence of treatment. Moreover, sections of Hco-µUtrn-treated muscle exhibited more CTLs than observed for the M-µUtrn group. A possible reason could be that the mice received human codon-optimized µUtrn, which differs from endogenous murine utrophin. The cellular immune response may be triggered against the AAV capsid as well as the transgene product^3,26. Since we used the rAAV9 vector for microgene delivery in all groups, we assume that the difference in the number of infiltrating T-cells was associated with the different transgenes. Found observations need further investigation at longer time points and in the context of systemic administration, as intramuscular AAV delivery is known to enhance immune responses²⁷.

Side-by-side comparison of µUtrn variants following intramuscular injections in mdx mice revealed the potential of rAAV9-Hco-μUtrn, which induced the robust expression of functional protein with lower immunogenicity. Taking into account the potential of microutrophins as a gene therapy for DMD, human codon-optimized µUtrn was chosen for the subsequent experiments.

Systemic delivery of rAAV9-Hco-μUtrn for long-term studies in mdx mice

In order to explore the systemic effects of transgene delivery, 6-week-old mdx mice were injected with rAAV9-Hco-µUtrn at a dose 6 × 10¹⁴ vg kg⁻¹. The twenty-week study included functional testing throughout the experiment (the hanging wire test) and multiple terminal examinations such as transgene expression, muscle histopathology, CK levels, force deficit, and humoral immunity assessment. Protein and RNA expression analysis confirmed the successful AAV9-mediated delivery of transgenes to all target organs, including striated muscle, heart, and diaphragm (Fig. 3A,B). Transcript levels in TA muscle were comparable to those detected after direct intramuscular injection. Human codon-optimized µUtrn was efficiently delivered to the sarcolemma, thus restoring the DAGC (Fig. 3C).

Morphological analysis of skeletal muscle, heart, and diaphragm showed myofiber damage and many centronucleated muscle cells, in agreement with the histopathological aspects of the disease. There were no prominent differences between treated and untreated mdx mice. Features observed in all mice from both groups included areas of necrosis and regeneration as well as sites of inflammation and fibrosis, which are typical for DMD (Fig. 3D). There were no changes in other organs, except for the thyroid (Supplementary Fig. S2). One mouse exhibited a reduction in follicle number and hypertrophy of follicle epithelium 20 weeks after the injection of human codon-optimized µUtrn. We cannot definitively conclude whether these changes were related to rAAV9-Hco-µUtrn treatment or were individual age-related changes. It was previously shown that the systemic delivery of rAAV-based therapeutics can result in hypertrophy of the thyroid epithelium due to a reduction in thyroid hormone levels²⁸.

Long-term functional assessment determined via hanging wire tests showed significant improvements in the group of mdx mice treated with rAAV9-Hco-µUtrn at several time points (Fig. 3E, weeks 6 and 17). Unexpectedly, on other weeks, functional improvements in the Hco-µUtrn group were not significant, if any. In addition, at week 8, wild-type animals did not exhibit any difference from model mice due to unknown reasons (Fig. 3E). We suppose that the small number of animals in the experimental groups (n = 4) did not allow to exclude the influence of individual characteristics of animals and that greatly distorts the average value in the presence of outliers. Serum CK levels were markedly reduced to 6000–7000 in treated mdx mice, as compared to approximately 9000 U/L in control mdx mice (Fig. 3F). The specific force of isolated skeletal muscles from rAAV9-Hco-µUtrn–injected mice was higher than that of untreated mdx. Similarly, moderate force drop improvements after repeated eccentric contractions were detected in treated animals, although not significant (Fig. 3G).

To assess the humoral immune response to long-term transgene expression, we detected Hco-μUtrn-specific antibodies in the serum of treated animals. In parallel, we analyzed the sera of animals treated with the same dose of AAV9-µDys. We detected µDys-specific antibodies in sera from three of four AAV9-CMV-µDys-injected mice (Fig. 3H, №1, 2, 4 of the µDys group). Antibody levels varied between samples, as indicated by the intensity of bands. The sera of wild-type (B10) and mdx mice treated with DPBS did not contain transgene-specific antibodies. Hco-µUtrn-specific antibodies were identified in sera from two out of four treated mice (Fig. 3H, №2, 3 of the µUtrn group). However, band intensity was notably lower than for µDys-treated animals.

Taken together, the long-term monitoring of mdx mice after intravenous injection of rAAV9-Hco-µUtrn demonstrated persistent transgene expression in the skeletal muscle, heart, and diaphragm, with appropriate localization to the sarcolemma. Expression of Hco-µUtrn allowed for functional recovery and protected muscle against contraction-induced damage. Further, the histopathological analysis did not reveal any toxicity effect associated with the expression of rAAV9-Hco-µUtrn and vector administration at high doses.

Intramuscular injection of rAAV9-Hco-µUtrn-FLAG virus led to expression on the sarcolemma of TA muscles and functional improvements in mdx mice

To distinguish between native utrophin and recombinant µUtrn during immunofluorescence, the FLAG epitope tag was used as a transgene marker. The addition of an N-terminal FLAG tag to the µUtrn sequence has been reported previously^20,21. In the present study, the Hco-µUtrn protein C-terminus was modified with the DYKDDDDK peptide sequence, allowing for detection of the transgene using antibodies against the utrophin N-terminus and the FLAG epitope. To determine whether the FLAG-tag had a negative impact on Hco-µUtrn expression, localization, and functionality, we injected rAAV9-Hco-µUtrn and rAAV9-Hco-µUtrn-FLAG into the TA muscles of mdx mice (2 × 10¹² vg per TA muscle). Two weeks post-injection, transcript levels and function were analyzed. Hco-µUtrn-FLAG as well as Hco-µUtrn were successfully expressed on the sarcolemma of TA muscle cells (Fig. 4A, UTRN). Antibodies against the FLAG epitope recognized the Hco-µUtrn-FLAG protein (Fig. 4A, FLAG). The expression level was similar for both transgenes (Fig. 4B). Further, both rAAV9-Hco-µUtrn- and rAAV9-Hco-µUtrn-FLAG-treated mdx TA muscles exhibited a statistically significant improvement in contractile function when compared to untreated mdx mouse muscles (Fig. 4C). Thus, FLAG-tagged Hco-µUtrn can be used for further experiments in mdx mice or other animals with endogenous utrophin expression.

High-dose rAAV9-Hco-µUtrn-FLAG administration does not cause toxicity in rats

In order to assess the toxicity of human codon-optimized µUtrn delivered via the rAAV9 vector, we administered rAAV9-Hco-µUtrn-FLAG to male rats through intravenous injection. Low (2 × 10¹⁴ vg kg⁻¹) and high (6 × 10¹⁴ vg kg⁻¹) doses were tested. Weaning rats were injected at 3 weeks of age, followed by extensive daily monitoring. Acute and subacute toxicity evaluation was performed on days 3 and 14, respectively (Fig. 5A).

All animals survived the duration of the study, with no significant changes in clinical signs, organ weights, histopathological findings, hematological and coagulation parameters, as well as urine biochemistry. Biochemical blood tests indicated that the levels of ALT, AST, ALP, and GGT did not differ between treated and control rats (Table 1). There was a significant increase in creatinine levels on day 3 in the group receiving a low dose (Table 1), but there was no difference on day 14. The groups did not differ in mass and total activity, as determined via the open field test (Fig. 5D). Daily monitoring of animals confirmed no changes in the condition of the skin, hair, and visible mucous membranes. Ophthalmoscopic examination, ECG recording, urine biochemistry and Irwin tests also confirmed the lack of toxicity associated with rAAV9-Hco-µUtrn-FLAG (Supplementary Fig. S5–S7).

At the two-week time point, several organs from the treated animals were subjected to transgene expression analysis. The Hco-µUtrn transcript was detected in the diaphragm, heart, TA, gastrocnemius (GAS), as well as the liver, thus confirming successful rAAV9-mediated delivery and efficient expression (Fig. 5B). Dose-dependent expression was evident in all analyzed organs, except for the diaphragm. The high level of transgene transcripts in the liver can be explained by the presence of the constitutive CMV promoter in the transgene construct.

Taken together, our toxicity study confirmed that neither the rAAV9-Hco-µUtrn-FLAG vector itself nor Hco-µUtrn expression exhibited toxicity in rats.

Muscle-specific SPc5-12 and MHCK7 promoters drive µUtrn expression and ensure force improvement after intramuscular injection in mdx mice

To further reduce immunogenicity, we decided to test an Hco-µUtrn construct under muscle-specific promoters. To this end, we replaced the constitutive CMV promoter with MHCK7 and SPc5-12 promoters^29,30. In contrast to previous reports²⁰, including a SPc5-12 promoter in the µUtrn-expressing construct did not lead to the reduction of AAV vector yield (Supplementary Table S8). The modified constructs were administered to 7-week-old mdx mice via intramuscular injection at a dose of 2 × 10¹² vg per TA muscle.

We confirmed the induction of µUtrn expression by all constructs via western blot analysis and immunostaining at day 14 following injection (Fig. 6A,D). Protein levels from constructs under CMV, MHCK7, and SPc5-12 promoters did not differ significantly between each other (Fig. 6B). However, the SPc5-12 promoter induced three-fold lower mRNA levels relative to CMV and MHCK7, as determined via RT-qPCR analysis (Fig. 6C). Restoration of DAGC on the sarcolemma was achieved in all groups, although highest α1-syntrophin expression was observed in the rAAV9-CMV-Hco-µUtrn-treated group (Fig. 6D).

We observed the improvement of tetanic force in all treated TA muscles when compared to the muscles of untreated mdx mice. All rAAV9-Promoter-µUtrn variants led to a similar force deficit after eccentric contractions (Fig. 6E). The SPc5-12 promoter group exhibited the best specific force relative to those in the CMV and MHCK7 groups, which was also comparable to the force of wild-type controls (Fig. 6F). Thus, muscle-specific promoters MHCK7 and SPc5-12 were as effective as the constitutive CMV promoter. Further, all tested promoters induced Hco-µUtrn expression and functional improvement at comparable levels.