Long-term regulated expression of growth hormone in mice after intramuscular gene transfer

Abstract

Effective delivery of secreted proteins by gene therapy will require a vector that directs stable delivery of a transgene and a regulatory system that permits pharmacologic control over the level and kinetics of therapeutic protein expression. We previously described a regulatory system that enables transcription of a target gene to be controlled by rapamycin, an orally bioavailable drug. Here we demonstrate in vivo regulation of gene expression after intramuscular injection of two separate adenovirus or adeno-associated virus (AAV) vectors, one encoding an inducible human growth hormone (hGH) target gene, and the other a bipartite rapamycin-regulated transcription factor. Upon delivery of either vector system into immunodeficient mice, basal plasma hGH expression was undetectable and was induced to high levels after administration of rapamycin. The precise level and duration of hGH expression could be controlled by the rapamycin dosing regimen. Equivalent profiles of induction were observed after repeated administration of single doses of rapamycin over many months. AAV conferred stable expression of regulated hGH in both immunocompetent and immunodeficient mice, whereas adenovirus-directed hGH expression quickly extinguished in immunocompetent animals. These studies demonstrate that the rapamycin-based regulatory system, delivered intramuscularly by AAV, fulfills many of the conditions necessary for the safe and effective delivery of therapeutic proteins by gene therapy.

Figure 1

Schematic diagram of transcription factor fusions and the reporter gene constructs. TF1 vector. The enhancer/promoter from CMV drives expression of a bicistronic transcript in which the first cistron encodes a fusion of the rapamycin binding domain of FRAP (FRB) with the activation domain of the p65 subunit of NF-κB. The second cistron, translated from an internal ribosome entry site (IRES), encodes 3 copies of FKBP fused to the chimeric DNA-binding domain ZFHD1. The 3′ untranslated region consists of an intron and polyadenylation sequence (PAS) from the rabbit β-globin gene (RBG). Both fusion proteins contain an amino-terminal epitope tag (E) and nuclear localization signal (N). ZG1 vector. This target gene vector contains 12 repeats of the recognition site for ZFHD1 upstream of a minimal IL-2 promoter driving expression of hGH, followed by the polyadenylation signal of the simian virus 40 late gene. ZG2 vector. This vector contains two transcription units in series. The ZG1 cassette is followed by all the elements of the TF1 cassette except the IRES and DNA-binding domain fusion.

Figure 2

Regulated hGH secretion in mice receiving i.m. injection of adenovirus. (A) Male BALB/c nude (open and filled circles) or BALB/c immunocompetent mice (open triangles) were injected i.m. with 0.5 × 10¹¹ particles each of Ad.TF1 and Ad.ZG1 or 1 × 10¹¹ particles of Ad.CMVhGH. The filled triangles indicate administration of rapamycin (5 mg/kg, i.p.). Values of plasma hGH are mean ± SEM (n = 4). The lower limit of detection of the assay is 50 pg/ml. (B) Vector dose–response. Male BALB/c nude mice were injected i.m. with a 1:1 mixture of Ad.TF1 and Ad.ZG1 at the indicated total dose (10e9 = 10⁹, etc.). Rapamycin (5 mg/kg, i.p.) was administered 10 days after vector infusion, and plasma samples were collected 24 h later. Values of plasma hGH are mean ± SEM. (C) Rapamycin dose–response. Male BALB/c nude mice were injected i.m. with a 1:1 mixture of Ad.TF1 and Ad.ZG1 at a total dose of 1 × 10¹¹ particles per mouse. The mice were challenged with rapamycin at the indicated doses 10 days after vector infusion, and plasma samples were collected 24 h later. Values of plasma hGH are mean ± SEM (n = 5).

Figure 3

Control over the level and duration of hGH expression with rapamycin. Male BALB/c nude mice were injected i.m. with 1 × 10¹¹ total particles of Ad.TF1 and Ad.ZG1 in a 1:1 ratio (filled and open circles) or with 5 × 10¹⁰ particles of Ad.CMVhGH (open triangles). Mice receiving Ad.TF1 and Ad.ZG1 were divided into two groups. Group 1 (open circles) received six consecutive doses of rapamycin (0.05 mg/kg, i.p.) every 2 days. Plasma samples were taken before each rapamycin dose and every other day after the last dose of rapamycin. The second and third sets of rapamycin challenges (0.5 and 0.05 mg/kg, respectively) were initiated when plasma hGH returned to basal levels from the previous challenges. Group 2 (filled circles) received similar injection and bleeding schedules but a 10-fold higher dose of rapamycin initially (0.5 mg/kg), with subsequent dosing schedules of 0.05 and 0.5 mg/kg. Values of plasma hGH are mean ± SEM of 5 mice for the group receiving Ad-CMVhGH and 10 mice for groups receiving Ad.TF1 and Ad.ZG1. Arrows indicate times of rapamycin dosing. The lower limit of detection of the hGH assay is 50 pg/ml.

Figure 4

Long-term rapamycin-induced hGH secretion in mice receiving i.m. injection of AAV. Immunodeficient BALB/c mice were injected i.m. with a total of 2 × 10¹¹ genome copies of AAV.TF1 and AAV.ZG2 in a 1:1 ratio. Rapamycin (5 mg/kg filled triangles, 1 mg/kg open triangles) or a rapamycin analog with 10-fold decreased activity (5 mg/kg open circle) was administered i.p. in single or multiple doses as indicated. Values of plasma hGH are mean ± SEM (n = 5). The lower limit of detection of the hGH assay is 50 pg/ml.

Figure 5

Dose-responsive hGH secretion in immunocompetent mice receiving i.m. injection of AAV. Immunocompetent BALB/c mice were injected i.m. with 2 × 10¹¹ total genome copies of AAV.TF1 and AAV.ZG2 at a 1:1 ratio. The mice were then divided into three groups and challenged with different single doses of rapamycin as indicated. Values of plasma hGH are mean ± SEM (n = 5).