Immune-driven gene expression loss following intramuscular AAV delivery to non-human primates is only transient

Abstract

Recombinant adeno-associated virus (rAAV) vectors stand out as highly promising for in vivo gene transfer, particularly in targeting the skeletal muscle for treating muscular genetic diseases or secreting therapeutic factors. Despite the simplicity and efficacy of the established intramuscular (IM) route, it has been often associated with an immune-induced rapid loss of transgene expression, in particular in large animal models, and generally considered irreversible as a consequence of a cytotoxic elimination of transduced cells. Here, we report in a non-human primate model that transgene expression loss after IM delivery of an rAAV1 expressing an immunogenic protein is only transient, with the re-expression of the transgene lasting up to 5 years post-injection. We show that the recovery of transgene expression is due to persisting viral genomes in the injected muscles despite the detection of peripheral anti-transgene cellular immunity. Persisting genomes were observed in the presence of infiltrated mononuclear CD8 and CD4 T lymphocytes, among which we were able to detect FoxP3⁺ regulatory cells. This is to our knowledge the first report of a transient immune-mediated loss of gene expression in a large animal model after rAAV delivery that should shed new light on the issue of rAAV vector immunogenicity.

Keywords: AAV; Tregs; gene transfer; immune response; immunomodulation; intramuscular; non-human primate; transitory loss.

Graphical abstract

Figure 1

Persistent rAAV viral genomes in injected muscles despite a loss of transgene expression Detection of viral genomes in injected muscles. Three primates (Mac1, Mac2, and Mac3) were injected with ssAAV2/1 rtTA/EPO in the left tibialis anterior at 1e11 vg/kg divided in 3–4 IM injection sites. A biopsy was performed in one muscle injection site and in the contralateral tibialis muscle at 26 months post-injection. After DNA extraction, viral genomes were measured by qPCR and expressed as viral genomes per diploid genome (vg/dg). The gray area indicates the limit of quantification of the PCR assay: 4.82e−4 vg/dg.

Figure 2

Re-expression of the transgene despite its initial loss Follow-up of erythropoietin (EPO) levels after doxycycline induction cycles (dotted lines) in Mac1, Mac2, and Mac3 after IM rAAV1 delivery. The red line indicates the physiological basal level of EPO (7 mU/mL) established as the mean ± 2 SD of 182 measures of EPO in serum samples from 32 cynomolgus primates. The blue arrow indicates the time point at which the muscle biopsies were performed.

Figure 3

Transgene re-expression results from persistent viral genomes in injected muscles and not from untargeted tissues Detection of viral genomes and transgene transcripts in the injected muscles. (A) Viral genomes were measured by qPCR in injected tibialis, non-injected muscles from the treated limb, non-injected muscles from the contralateral limb, and distant muscles (upper limb and diaphragm). Results are expressed as vg/dg. The gray area indicates the limit of quantification of the PCR assay: 4.82e−4 vg/dg. (B) Transgene transcripts were detected by quantitative reverse-transcription PCR (RT-qPCR) in the same tissues and expressed as RQ (relative quantity) to an endogenous gene. The gray area indicates the limit of quantification of the RT-qPCR assay: 8.68e−2.

Figure 4

Detection of mononuclear cellular infiltrates in IM sites of vector delivery Injected and contralateral tibialis muscles were collected at 5 years post-injection. Hematoxylin-eosin-saffron staining was performed on formalin-fixed paraffin-embedded muscle sections. Mononuclear infiltrated cells were observed in all IM injection sites of the vector. Scale bars: 50 μm.

Figure 5

Absence of PD1⁺ and CTLA4⁺ exhausted T lymphocytes in cellular muscle infiltrates Characterization of infiltrated cells in injected muscles on PFA-fixed paraffin-embedded muscles sections subjected to DAPI (blue), CD3 (green), CTLA4 (red) (A) and PD-1 (purple) (B) co-stainings. Positive control consists in PFA-fixed embedded-paraffin lymph node and spleen section subjected to the same co-stainings. (White scale bar: 50 μm).

Figure 6

Detection of MHC class I⁺ muscular fibers in injected tissues Staining of MHC class I antigen on PFA-fixed paraffin-embedded muscle sections using anti-MHC class I (red) and DAPI (blue). The orange arrows correspond to immune infiltrates next to MHC class I⁺ muscular fibers. Scale bars: 50 μm.

Figure 7

Detection of FoxP3⁺CD4⁺ T lymphocytes in cellular muscle infiltrates Characterization of infiltrated cells in injected muscles on PFA-fixed paraffin-embedded muscle sections subjected to DAPI (blue), CD4 (red), and FoxP3 (green) co-staining. White scale bars: 50 μm. The red boxes correspond to an enlargement of the merge without visualization of DAPI, to show more precisely the CD4 membrane staining surrounding the Foxp3 intracellular staining. Yellow scale bars: 20 μm.