Replication-deficient human adenovirus type 35 vectors for gene transfer and vaccination: efficient human cell infection and bypass of preexisting adenovirus immunity

Abstract

Replication-deficient human adenovirus type 5 (Ad5) can be produced to high titers in complementing cell lines, such as PER.C6, and is widely used as a vaccine and gene therapy vector. However, preexisting immunity against Ad5 hampers consistency of gene transfer, immunological responses, and vector-mediated toxicities. We report the identification of human Ad35 as a virus with low global prevalence and the generation of an Ad35 vector plasmid system for easy insertion of heterologous genes. In addition, we have identified the minimal sequence of the Ad35-E1B region (molecular weight, 55,000 [55K]), pivotal for complementation of fully E1-lacking Ad35 vector on PER.C6 cells. After stable insertion of the 55K sequence into PER.C6 cells a cell line was obtained (PER.C6/55K) that efficiently transcomplements both Ad5 and Ad35 vectors. We further demonstrate that transduction with Ad35 is not hampered by preexisting Ad5 immunity and that Ad35 efficiently infects dendritic cells, smooth muscle cells, and synoviocytes, in contrast to Ad5.

FIG. 1.

Low seroprevalence toward Ad35 in healthy individuals. (A) One hundred serum samples from the Belgium population (average age, 30.8 years; range, 18 to 62 years; 50% female) were tested for NA against human adenovirus types. Shown on the y axis is the percentage of sera that inhibited virus replication >90% at the lowest dilution (end concentration in the well: 1/4). Types are ranked according to their classification in subgroups. (B) Approximately 560 serum samples from locations in Europe (black bar), the United States (gray bar), and Asia (white bar) were tested for NA against adenovirus types. Shown on the y axis is the percentage of sera that inhibited virus replication >90% at the lowest dilution for Ad5, Ad35, and Ad11 wild-type virus.

FIG. 2.

Generation of recombinant Ad35 vectors. (A) Schematic of the genome organization of wtAd35. Indicated are the inverted terminal repeats (ITRs), packaging signal (Ψ), and early genes involved in replication (E1, E2A, E2B, and E4) and immune modulation (E3). Also, regions coding for the structural proteins (L1 to L5) are given. Total length is 34,794 bp. The NdeI sites are located at positions 6541 and 33166 in the wtAd35 sequence (GenBank accession no. AY271307). (B) Schematic overview of the two-plasmid vector system designed to allow generation of recombinant E1-lacking Ad35 virus (see Materials and Methods for details).

FIG. 3.

Improved infection of (primary) human cells with Ad35-based viruses. Cells were exposed for 2 h, in triplicate, to Ad5 or Ad35 vector carrying green fluorescent protein (GFP) as a marker gene, after which the medium was replaced by culture medium without virus. A549 cells, SMC, and human synoviocytes were infected at a multiplicity of infection of 500 (white bar), 2,000 (grey bar), or 5,000 (black bar) vp/cell. Mo-DC (from three different donors; data were pooled) were infected with 100, 1,000, or 2,500 vp/cell. After 48 h cells were harvested and gene transfer efficacy was determined.

FIG. 4.

Biodistribution of recombinant Ad35 vectors in mice. (A) Ad5.AdApt.Luc or Ad35ΔE3.AdApt.Luc (10¹¹ vp) was injected i.v. into BALB/c mice, and luciferase expression was monitored 72 h later using charge-coupled device camera technology (56). (B) Quality of Ad5 and Ad35 vector batches. Human A549 cells were exposed for two hours to an increasing concentration of Ad5 or Ad35 carrying the luciferase reporter gene. Cells were harvested after 48 h and lysed, and luciferase activity was determined (expressed in relative light units [RLU]).

FIG. 5.

Ad35 vectors circumvent Ad5 immunity. (A) Human serum samples were tested for NA against Ad5- or Ad35-luciferase viruses. Black dots represent individual sera at the serum dilution that inhibited >90% of the luciferase activity in the absence of serum. The bars represent geometric mean titers. (B) Serum samples from mice twice immunized with Ad5.empty viruses (10¹⁰ vp) were tested for NA against Ad5- or Ad35-luciferase viruses. (C) Spleen cells from mice twice immunized with Ad5.empty viruses (10¹⁰ vp) were incubated with NIH 3T3 cells transduced with either Ad5 or Ad35 empty vectors and IFN-γ release was quantified.

FIG. 6.

(A) Human serum was administered i.v. to SCID mice, and serum was isolated at various time points. The undiluted human serum fraction was tested in vitro for the ability to inhibit Ad5-mediated luciferase gene transfer to human A549 cells (black circles). The serum fraction was subsequently diluted 10-fold in PBS (black squares) to mimic the dilution of the samples retrieved from the mice that received injections. Mouse serum was isolated at 6 h (black triangles) post-serum administration and after 48 h (black diamonds). Serum from mice that received Ad5-negative human serum served as negative controls (white circles). (B) Ad35 viruses transduce muscle in presence of human serum with NA against Ad5. Results are presented as luciferase activity in 25 μl of muscle lysates (left graph), showing that Ad5.Luc expression is largely reduced in the presence of human serum with Ad5 NA (+serum) compared to Ad5.Luc activity in mice that received human serum without Ad5 NA (−serum). This reduction is not seen in mice receiving Ad35.Luc viruses. In the graph at right the luciferase activity determined in muscle from mice that received human anti-Ad5 serum (+serum) is shown as a percentage of luciferase activity compared to that in mice that received anti-Ad5-negative serum (−serum).