Adeno-Associated Virus Serotype-Specific Inverted Terminal Repeat Sequence Role in Vector Transgene Expression

Abstract

Adeno-associated viral vectors have been successfully used in laboratory and clinical settings for efficient gene delivery. In these vectors, 96% of the adeno-associated virus (AAV) genome is replaced with a gene cassette of interest, leaving only the 145 bp inverted terminal repeat (ITR) sequences. These cis-elements, primarily from AAV serotype 2, are required for genome rescue, replication, packaging, and vector persistence. Previous work from our lab and others have demonstrated that the AAV ITR2 sequence has inherent transcriptional activity, which may confound intended transgene expression in therapeutic applications. Currently, AAV capsids are extensively study for vector contribution; however, a comprehensive analysis of ITR promoter activity of various AAV serotypes has not been described to date. Here, the transcriptional activity of AAV ITRs from different serotypes (1-4, 6, and 7) was compared in numerous cell lines and a mouse model. Under the conditions used here, all ITRs tested were capable of promoting transgene expression both in vitro and in vivo. However, we observed three classes of AAV ITR expression in vitro. Class I ITRs (AAV2 and 3) generated the highest level, whereas class II (AAV 4) had intermediate levels, and class III (AAV1 and 6) had the lowest levels. These expression levels were consistent across multiple cell lines. Only ITR7 demonstrated cell-type dependent transcriptional activity. In vivo, all classes had promoter activity. Next-generation sequencing revealed multiple transcriptional start sites that originated from the ITR sequence, with most arising from within the Rep binding element. The collective results demonstrate that the serotype ITR sequence may have multiple levels of influence on transgene expression cassettes independent of promoter selection.

Keywords: AAV; ITR; promoter.

Figure 1.

Sequence and structures of AAV ITRs from serotypes 1–4, 6, and 7. (A) ITR2 with the RBE and RBE’ in bold. The terminal resolution nicking site TT dinucleotide is in red. (B) Consensus ITR sequence. Locations of nucleotide differences between ITR sequences 1–4, 6–7 are highlighted in red. The red nucleotides are in IUPAC code, where Y is C or T, R is A or G, S is G or C, W is A or T, K is G or T, M is A or C, B is G or T or C, V is G or C or A, and N is any nucleotide. Colored outlines denote the A (black), B (blue), C (gray), and D (green) regions in the ITR. (C) The sequence and structures of ITRs 1–4, 6–7. Bold letters denote non-conserved nucleotides between the ITR sequences. AAV, adeno-associated virus; RBE, Rep binding element; trs, terminal resolution site.

Figure 2.

Luciferase activity from cell lines infected with AAV(1–4, 6–7)/2—ITR-luciferase vectors. (A) HEK293 cells were infected with AAV2/2-ITR-luciferase or AAV2/2-CBA-luciferase at 1E5 vg/cell. Two days post-infection, the cells were lysed and luciferase activity was measured by using luciferin substrate. Values shown are raw RLUs. (B–D) Indicated cell lines were infected with AAV1/2-ITR-luciferase, AAV2/2-ITR-luciferase, or AAV7/2-ITR-luciferase at 2E5 vg/cell. Two days post-infection, the cells were lysed and luciferase activity was measured by using luciferin substrate. RLU values were normalized to total cellular protein added to the luciferase assay, as measured by BCA; then, all values were normalized to ITR2. Each AAVN/2-ITR-luciferase was made in triplicate batches and titered together; then, cells were infected in triplicate. (E–G) Indicated cell lines were infected with AAV1/2-ITR-luciferase, AAV2/2-ITR-luciferase, AAV3/2-ITR-luciferase, AAV4/2-ITR-luciferase, or AAV6/2-ITR-luciferase at 2E5 vg/cell. Two days post-infection, the cells were lysed and luciferase activity was measured by using luciferin substrate. RLU values were normalized to total cellular protein added to the luciferase assay, as measured by BCA; then, all values were normalized to ITR2. Each AAVN/2-ITR-luciferase was made in triplicate batches and titered together; then, cells were infected in triplicate. p-Values are indicated as *<0.0001, ^#<0.001, and ^<0.01. ns, nonsignificant; RLU, relative light units.

Figure 3.

Luciferase activity from HEK293 cells infected with AAV1/1 and AAV2/1—ITR-luciferase vectors. HEK293 cells were infected in triplicate with three biological replicates of AAV1/1, AAV2/1, AAV1/2, or AAV2/2-ITR-luciferase vectors at 2E5 vg/cell. Two days post-infection, the cells were lysed and luciferase activity was measured by using luciferin substrate. RLU values were normalized to total cellular protein added to the luciferase assay, as measured by BCA; then, all values were normalized to ITR2 values within each capsid group (i.e., AAV1/1 and 2/1 were normalized to AAV2/1. AAV1/2 and 2/2 were normalized to 2/2). There is no statistical difference between AAV1/1 and AAV1/2.

Figure 4.

TSS from ITR1–4, 6, or 7 promoted luciferase mRNA. HEK293 cells were infected with AAV(1–4, 6, or 7)/2-ITR-luciferase at 2E5 vg/cell, and total RNA was harvested 3 days post-infection. RNA was reverse transcribed by using luciferase specific primers. Final cDNA products were analyzed by next-generation sequencing. Black arrows indicate a TSS with 1% or higher of read sequences. The RBE’ and RBE are indicated in bold. The trs is denoted in red. TSS, transcription start sites.

Figure 5.

Luciferase activity from mice injected with AAV(1–4, 6)/9-ITR-cre recombinase. Four- to 6 week-old male FVB.129S6(B6)-Gt(ROSA)26Sor^tm1(Luc)Kael/J mice were injected with 100 μL of 1E9 vg of AAV(1–4, 6)/9-ITR-cre recombinase. At 3 weeks (A) and at 9 weeks (B) post-AAV injection, mice were given 100 μL of luciferase substrate i.p. and photons were recorded.