A self-deletion lentiviral vector to reduce the risk of replication-competent virus formation

Reducing the transcriptional read-through rate of a lentiviral vector for β-thalassemia gene therapy

BACKGROUND

LentiGlobin BB305 is a self-inactivating lentiviral vector carrying a human β-globin expressing cassette for treating β-thalassemia. Initially, a 2 × 250 bp chicken Locus Control Region fragment of cHS4, functioning as an insulator, was placed at its ΔU3, which was removed after the first clinical trial led by a French team to avoid abnormal splicing, etc. This action could potentially lead to an increasing risk of the transcriptional read-through rate driven by the β-globin promoter to a significant level, posing a biosafety risk in clinical trials.

METHODS

In the present study, a read-through reducing agent (C-U+ or WPRE) was designed to be placed at the 3' UTR of the β-globin gene. The Enhancer Activities and/or Transcriptional Read-Through (EATRT) rate at the mRNA level and the protein expression level regarding lentiviral preparation titer were examined.

RESULTS

We found that the insertion of the element (C-U+ or WPRE) reduced the EATRT effectively by 53% or 41%, respectively. C-U+ has less impact on virus package efficiency. Furthermore, there was no significant difference in the protein expression level after the C-U+ or WPRE insertion.

CONCLUSIONS

The results of the present study show that inserting C-U+ or WPRE before the polyA sequence of the BB305 would reduce the EATRT rate at no cost of its expressing efficacy and viral preparation titers. Thus, we present an alternative improvement for a safer lentiviral vector for β-thalassemia clinical trials.

HIV-1 sequences in lentiviral vector genomes can be substantially reduced without compromising transduction efficiency

Many lentiviral vectors used for gene therapy are derived from HIV-1. An optimal vector genome would include only the viral sequences required for transduction efficiency and gene expression to minimize the amount of foreign sequence inserted into a patient’s genome. However, it remains unclear whether all of the HIV-1 sequence in vector genomes is essential. To determine which viral sequences are required, we performed a systematic deletion analysis, which showed that most of the gag region and over 50% of the env region could be deleted. Because the splicing profile for lentiviral vectors is poorly characterized, we used long-read sequencing to determine canonical and cryptic splice site usage. Deleting specific regions of env sequence reduced the number of splicing events per transcript and increased the proportion of unspliced genomes. Finally, combining a large deletion in gag with repositioning the Rev-response element downstream of the 3’ R to prevent its reverse transcription showed that 1201 nucleotides of HIV-1 sequence can be removed from the integrated vector genome without substantially compromising transduction efficiency. Overall, this allows the creation of lentiviral vector genomes that contain minimal HIV-1 sequence, which could improve safety and transfer less viral sequence into a patient’s DNA.

Improved biosafety of a lentiviral vector by reducing cellular gene activation

BACKGROUND

Lentiviral vectors (LVs) have enhancer activity and/or transcriptional read-through (EATRT) properties that can lead to the activation of adjacent genes. Consequently, patients may be at increased risk for adverse effects if such vectors are used clinically.

METHODS

In the present study, we assessed the abilities of different "pro-LV"-like constructs with respect to decreasing its EATRT, including the "pro-LV" vector bearing a chimeric ΔLTR of the human foamy virus R-U5 region replaced by that of an LV (HF).

RESULTS

By analyzing the EATRT of "pro-LV" constructs transfected in 293T cells, we observed that the inclusion of the first 400 bp of the chicken β-globin locus HS4 insulator core sequence oriented in the reverse direction (C-) combined with two copies of the simian virus 40 upstream-sequence element (U) at the ΔU3 of ΔLTR region of "pro-LV" tended to shield the adjacent genomic sequences, such that the EATRT rate was lower than when either of the C- or U was included in the "pro-LV". Moreover, upon transduction, the pro-HF appears to reduce the EATRT rate in the chromosomes of 293T (by 80%) and human peripheral blood mononuclear cells (PBMCs) (by 75%) compared to when pro-LV C-U was included (with a 60% and 89% reduction in 293T and PBMCs, respectively). The HF construct had a significant reduction of viral biological titer compared tiowhen the pro-LV C-U was used in 293T cells.

CONCLUSIONS

The results of the present study provide an important basis for the clinical applicability of LVs in gene and cell therapy.

Gene Therapy With Regulatory T Cells: A Beneficial Alliance

Gene therapy aims to replace a defective or a deficient protein at therapeutic or curative levels. Improved vector designs have enhanced safety, efficacy, and delivery, with potential for lasting treatment. However, innate and adaptive immune responses to the viral vector and transgene product remain obstacles to the establishment of therapeutic efficacy. It is widely accepted that endogenous regulatory T cells (Tregs) are critical for tolerance induction to the transgene product and in some cases the viral vector. There are two basic strategies to harness the suppressive ability of Tregs: in vivo induction of adaptive Tregs specific to the introduced gene product and concurrent administration of autologous, ex vivo expanded Tregs. The latter may be polyclonal or engineered to direct specificity to the therapeutic antigen. Recent clinical trials have advanced adoptive immunotherapy with Tregs for the treatment of autoimmune disease and in patients receiving cell transplants. Here, we highlight the potential benefit of combining gene therapy with Treg adoptive transfer to achieve a sustained transgene expression. Furthermore, techniques to engineer antigen-specific Treg cell populations, either through reprogramming conventional CD4⁺ T cells or transferring T cell receptors with known specificity into polyclonal Tregs, are promising in preclinical studies. Thus, based upon these observations and the successful use of chimeric (IgG-based) antigen receptors (CARs) in antigen-specific effector T cells, different types of CAR-Tregs could be added to the repertoire of inhibitory modalities to suppress immune responses to therapeutic cargos of gene therapy vectors. The diverse approaches to harness the ability of Tregs to suppress unwanted immune responses to gene therapy and their perspectives are reviewed in this article.

Eliminating HIV-1 Packaging Sequences from Lentiviral Vector Proviruses Enhances Safety and Expedites Gene Transfer for Gene Therapy

Lentiviral vector genomic RNA requires sequences that partially overlap wild-type HIV-1 gag and env genes for packaging into vector particles. These HIV-1 packaging sequences constitute 19.6% of the wild-type HIV-1 genome and contain functional cis elements that potentially compromise clinical safety. Here, we describe the development of a novel lentiviral vector (LTR1) with a unique genomic structure designed to prevent transfer of HIV-1 packaging sequences to patient cells, thus reducing the total HIV-1 content to just 4.8% of the wild-type genome. This has been achieved by reconfiguring the vector to mediate reverse-transcription with a single strand transfer, instead of the usual two, and in which HIV-1 packaging sequences are not copied. We show that LTR1 vectors offer improved safety in their resistance to remobilization in HIV-1 particles and reduced frequency of splicing into human genes. Following intravenous luciferase vector administration to neonatal mice, LTR1 sustained a higher level of liver transgene expression than an equivalent dose of a standard lentivirus. LTR1 vectors produce reverse-transcription products earlier and start to express transgenes significantly quicker than standard lentiviruses after transduction. Finally, we show that LTR1 is an effective lentiviral gene therapy vector as demonstrated by correction of a mouse hemophilia B model.