RETROVIRAL VECTORS

Abstract

There is provided a retroviral RNA vector comprising a 5 cap, a transgene, a 3 long terminal repeat (LTR) and an RNA packaging sequence, wherein translation of the transgene is initiated at the 5 end of the transgene in a cap-dependent manner, and wherein the 3 LTR and the RNA packaging sequence are located 3 of the transgene. Also provided is a nucleotide sequence encoding a vector genome. In addition, there is provided a host cell, a virion and a pharmaceutical composition comprising the vector or nucleotide sequence, and the use of the vector in delivering a transgene to a cell or subject.

Claims

1. A retroviral RNA vector comprising a 5 cap, a transgene, a 3 long terminal repeat (LTR) and an RNA packaging sequence, wherein translation of the transgene is initiated at the 5 end of the transgene in a cap-dependent manner, wherein the 3 LTR and the RNA packaging sequence are located 3 of the transgene.

2. A retroviral RNA vector according to claim 1, wherein the vector does not comprise a 5 LTR.

3. A retroviral RNA vector according to claim 1, wherein the RNA packaging sequence comprises the RNA packaging signal () and a portion of the gag gene.

4. A retroviral RNA vector according to claim 3, wherein the RNA packaging sequence also comprises the Rev Response Element (RRE).

5. A retroviral RNA vector according to claim 1, wherein the RNA packaging sequence is located 3 of the 3 LTR.

6. A retroviral RNA vector according to claim 1, wherein the vector does not comprise a primer binding site (PBS).

7. A retroviral RNA vector according to claim 1, wherein the vector further comprises an intron, optionally wherein the intron is located between the 5 cap and the transgene.

8. A retroviral RNA vector according to claim 1, wherein the transgene further comprises a Kozak sequence.

9. A retroviral RNA vector according to claim 1, wherein the vector further comprises a post-transcriptional regulatory element (PRE), such as a woodchuck hepatitis virus post-transcriptional regulatory element (WPRE).

10. A retroviral RNA vector according to claim 1, wherein the vector further comprises a nuclear export signal, such as the constitutive transport element (CTE).

11. A retroviral RNA vector according to claim 1, wherein the 3 LTR is a self-inactivating (SIN) LTR.

12. A retroviral RNA vector according to claim 1, wherein the vector is a lentiviral vector.

13. A retroviral RNA vector according to claim 12, wherein the vector is an HIV-1 vector.

14. A retroviral RNA vector according to claim 1, wherein the vector further comprises a poly(A) tail at the 3 end of the vector.

15. A nucleotide sequence encoding a vector genome, the nucleotide sequence comprising a promoter operably linked to a vector genome, wherein the vector genome comprises a transgene, a 3 LTR and an RNA packaging sequence, wherein the transgene is located at the 5 end of the vector genome such that, following transcription and capping of the vector genome, translation is initiated at the 5 end of the transgene in a cap-dependent manner, wherein the 3 LTR and the RNA packaging sequence are located 3 of the transgene.

16. A nucleotide sequence according to claim 15, the nucleotide sequence further comprising a poly(A) signal.

17. A host cell containing: (a) retroviral RNA vector comprising a 5 cap, a transgene, a 3 long terminal repeat (LTR) and an RNA packaging sequence, wherein translation of the transgene is initiated at the 5 end of the transgene in a cap-dependent manner, wherein the 3 LTR and the RNA packaging sequence are located 3 of the transgene or (b) a nucleotide sequence encoding a vector genome, the nucleotide sequence comprising a promoter operably linked to a vector genome, wherein the vector genome comprises a transgene, a 3 LTR and an RNA packaging sequence, wherein the transgene is located at the 5 end of the vector genome such that, following transcription and capping of the vector genome, translation is initiated at the 5 end of the transgene in a cap-dependent manner, wherein the 3 LTR and the RNA packaging sequence are located 3 of the transgene.

18. A virion containing the retroviral vector according to claim 1.

19. A pharmaceutical composition comprising the retroviral vector according to claim 1 or a virion encoding the retroviral vector.

20. (canceled)

21. (canceled)

22. A method of delivering a transgene to a target cell, the method comprising administering an effective amount of the retroviral vector according to claim 1 or a virion encoding the retroviral vector to the target cell.

23. A method of delivering a transgene to a target cell in a subject, the method comprising administering an effective amount of the retroviral vector of claim 1 or a virion encoding the retroviral vector to the subject.

24. A cell produced by a method comprising administering to a subject an effective amount of: (a) a retroviral vector comprising a 5 cap, a transgene, a 3 long terminal repeat (LTR) and an RNA packaging sequence, wherein translation of the transgene is initiated at the 5 end of the transgene in a cap-dependent manner, wherein the 3 LTR and the RNA packaging sequence are located 3 of the transgene; or (b) a virion encoding the retroviral vector.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0088] FIG. 1 is a schematic showing lifecycle of LV vectors designed for direct mRNA expression. a) In the left panel, a conventional LV transduces the cell and reverse-transcribes its ssRNA into dsDNA, which then enters the nucleus to either integrate into the host genome or exist as a circular episome. The nuclear DNA can then use host machinery to drive transgene mRNA production and express the transgene protein. On the right panel, RT-deficient LV vectors contain a mutated reverse-transcriptase (RT) enzyme, which means that packaged ssRNA is free to interact with cell ribosomes to express contents as mRNA. b) Vector schematics show the structure of the DNA template used to generate vector genome RNA. White boxes indicate elements that are present in DNA alone, whilst grey boxes indicate the elements that are transcribed into the encapsidated RNA genomes. IRES-mediated mRNA expression has previously been reported, where the vector RNA genome contains an internal IRES which allows translation of the downstream coding sequence (in this case EGFP). With the inventors' novel Cap-mediated mRNA expression system, the transgene coding sequence is positioned at the extreme 5 terminus of vector ssRNA, meaning that translation initiation occurs from the 5 m7G cap. Additionally, the vector lacks a HIV-1 primer binding site, further reducing any probability of the mutated reverse-transcriptase binding its canonical target.

[0089] FIG. 2 identifies the optimal genome structure for CDLV translation. a) Six versions of CDLV (CDLV1-6) were developed, with designs aimed to enhance RNA processing for optimal packaging and translation upon cell entry. The schematics represent the structure of lentiviral vector RNA genomes that would be packaged into vector particles. LTR, HIV-1 long terminal repeat; Gag, truncated and inactive HIV-1 gag gene, including packaging sequence; RRE, HIV-1 rev response element; cPPT, HIV-1 central polypurine tract; IRES, internal ribosome entry site (IRES) of the encephalomyocarditis virus; EGFP, enhanced green fluorescent protein; WPRE, woodchuck-hepatitis virus post-transcriptional regulatory element; Intron, chimera between introns from human -globin and immunoglobulin heavy chain genes; bGHpA, bovine growth hormone polyadenylation sequence; U5, U5 domain of the HIV-1 LTR. b) Comparison of functional titres gained from each of the ssRNA genomes. Titres were determined by the percentage of EGFP+ cells in transduced cells, indicating transduction units per milliliter (TU/ml). Statistical comparison of the titres was made by Kruskal Wallis test with Dunn's posthoc analysis (*P<0.001). Averages were calculated from 4 experimental replicates in each case.

[0090] FIG. 3 shows the comparison of CDLV transduction kinetics versus IRES-based RT-deficient vectors. a) IRES-EGFP and CDLV-EGFP vectors were produced using the depicted genomes. b) HEK 293T cells were transduced with the vectors. Longitudinal analysis of EGFP expression shows CDLV vector expression profile compared versus the IRES-based system over the course of the experiment (P=0.0196 by t-test). c) Flow cytometry plots show EGFP mean fluorescence intensity (MFI) per transduced cell, comparing CDLV-EGFP expression versus IRES-EGFP. d) Vectors dosed by p24 capsid protein mass allow crude estimation of the number of vector capsids required to achieve optimal gene expression. This analysis demonstrates that 1 ng p24 per cell is necessary to achieve optimal expression intensity with CDLV (approximately 110.sup.7 LV particles).

[0091] FIG. 4 shows the comparison of CDLV transduction kinetics to conventional IDLV and IPLV vectors. a) Integration deficient (IDLV-SFFV-EGFP) and integration-proficient (IPLV-SFFV-EGFP) lentiviral vectors containing an SFFV-EGFP expression cassette were compared to a CDLV-EGFP vector. b) HEK 293T cells transduced and the percentage of EGFP+ cells was tracked for 2 weeks, to monitor expression longevity. c) The intensity of EGFP expression gained from each vector is shown for the first 72 hours after transduction. d) The percentage of EGFP-positive cells is compared for CDLV-EGFP at 24-hours post-transduction and IDLV-SFFV-EGFP at 72-hours post-transduction (the relative timing of peak expression for each vector platform) (P=0.05 by t-test). e) The same data set shown in d) was analyzed for EGFP fluorescence (MFI), to compare the strength of EGFP expression per transduced cell (P=0.005 by t-test).

[0092] FIG. 5 tracks the in vivo expression kinetics of CDLV compared to DNA-based LV vectors. a) IDLV or IPLV vectors were produced containing a bicistronic luciferase-EGFP expression cassette, driven by the SFFV promoter (IPLV/IDLV-SFFV-Luc-EGFP). A CDLV vector was additionally produced containing the same Luc-EGFP coding sequence (CDLV-Luc-EGFP). b) In vivo bioluminescence was imaged at various time-points for 10 days (240 hours) after vector injection, to track the onset and persistence of transgene expression. Data are expressed as photons per second per cm 2 per steradian (n=4 mice per group). c) The associated heat maps for each mouse indicate bioluminescence signal intensity over the initial 96-hours of the experiment.

DETAILED DESCRIPTION OF THE INVENTION

Example 1

[0093] Results

[0094] The aim of this investigation was to develop an LV-based vector that can efficiently deliver its genome as mRNA to target cells in vivo and ex vivo. It has previously been reported that the HIV-1 Reverse-Transcriptase component of LV vectors can be mutated to remove its ability to convert RNA into DNA. This reverse transcription deficient (RT-deficient) LV vector platform then achieves transgene expression in target cells without forming a DNA intermediate (FIG. 1a).

[0095] Engineering LV Vector Genomes to Maximize Translation of Packaged ssRNA

[0096] In previous studies, RT-deficient vectors have achieved gene expression in a subset of target cells, but the expression level was insufficient to mediate targeted disruption of the human CCR5 gene in HSCs. The inventors hypothesized that this lack of efficacy was due to reliance on an IRES element to initiate translation, due to the presence of 1.5 kb wild-type HIV-1 DNA in the 5 region of the vector RNA (FIG. 1b).

[0097] In order to eliminate this problem, the inventors restructured the vector genome by moving the 1.5 kb HIV-1 DNA downstream of the therapeutic transgene (FIG. 1b). This configuration was designed to leave a Kozak consensus sequence at the extreme 5 terminus of the vector RNA, which would be 5 m.sup.7G cap capped during vector production. Additionally, the HIV-1 primer-binding site (PBS) was completely removed to eliminate canonical reverse-transcriptase priming. The inventors' expectation was that the 5 m.sup.7G cap-mediated LV vector would produce a greater level of gene expression than the IRES-mediated configuration.

[0098] The inventors engineered six iterations of the cap-mediated CDLV vector, with the aim of identifying a configuration that could exceed IRES-based LV expression (FIG. 2a). In each case, the depicted vector RNA is derived from a plasmid using the Cytomegalovirus (CMV) promoter to transcribe vector genome RNA. Each variant contained variable combinations of non-coding domains of the transcribed region. Version 1 (CDLV1) was designed to include a chimeric intron (fusion of introns from human (3-globin and immunoglobulin heavy chain genes), based in the expectation that the presence of splice sites at the 5 end would enhance nuclear export of the RNA. CDLV2 was identical to version CDLV1, but for omission of the chimeric intron. CDLV3 included a polyA motif upstream of the 3LTR to enhance LV ssRNA expression, whilst CDLV4 additionally contained an intron downstream of the polyA, to minimise premature termination during vector production. Finally, CDLV5 and CDLV6 aimed to promote read-through of the 3LTR by relocating WPRE to the extreme 3 terminus and deleting portions of the U3-R domains that contain HIV-1 polyA motifs.

[0099] In each case, vectors were designed to deliver an enhanced green fluorescent protein (EGFP) RNA payload. Vector titres were determined by transduction of HEK 293T cells and quantifying the number of EGFP-positive cells by flow cytometry. This comparison revealed that CDLV version 2 yielded titres one order of magnitude greater than the IRES-based vector (P<0.001) (FIG. 2b). This structure was therefore taken forwards for further characterization of the technology.

[0100] Cap-Dependent Translation Provides Greater Gene Expression than IRES-Dependent Translation

[0101] The transduction kinetics of CDLV version 2 were compared in greater detail versus the RES-mediated system. HEK 293T cells were transduced with an equal dose of either CDLV-EGFP or IRES-EGFP (FIG. 3a). Longitudinal analysis of EGFP expression showed that the CDLV vector produced a greater number of EGFP-positive cells than IRES-based LV throughout the course of the experiment (P=0.0196), with both vectors achieving peak expression by 24 hours and dropping to minimal levels by day 6 (FIG. 3b). EGFP fluorescence intensity was also quantified at each timepoint, with CDLV expression exceeding that of IRES-based vectors up to 72-hours post-transduction (P<(FIG. 3c).

[0102] Additionally, in order to clarify that these expression values were being derived from enhanced expression, rather than excess physical particle mass, dose response profiles were also compared in investigations in which vectors were dosed by capsid mass. These investigations showed that the CDLV vector gave the greatest level of EGFP expression at 4 hours and 24 hours post-transduction, when administered at a dose of 0.64 ng p24/cell. The IRES-based vector was unable to match this level, even when administered to cells at higher doses of 0.7 ng p24/cell and 3.5 ng p24/cell (FIG. 3d). The intensity of EGFP expression from IRES vectors was not enhanced at any of the tested doses, indicating that much higher doses would be needed to accumulate equivalent EGFP levels as those observed for CDLV.

[0103] CDLV Vectors Provide a Transient Burst of Gene Expression In Vitro

[0104] After demonstrating the potential advantages of CDLV as an mRNA delivery platform, the inventors set out to investigate how its longitudinal expression profile compared to DNA-based gene delivery systems. In this experiment, EGFP was again employed as a transgene, driven by the spleen focus-forming virus (SFFV) promoter in the DNA-based IPLV and IDLV vectors (FIG. 4a).

[0105] CDLV-EGFP was delivered to HEK 293T cells at a multiplicity of infection (MOI) of 41 EGFP-forming units per cell (EFU/cell), whilst IPLV-SFFV-EGFP and IDLV-SFFV-EGFP were delivered at doses of 10 EFU/cell. As expected, CDLV produced a transient expression profile, with peak expression occurring around 24 hours post-transduction, matching the longitudinal profile seen with previous retrovirus-based mRNA delivery platforms. whereas IPLV and IDLV vectors peaked at around 48 hours post-transduction, with the IDLV profile falling to 2.2% EGFP-positive cells by day 14 (FIG. 4b). Analysis of mean fluorescence intensity again showed that CDLV expression appeared sooner than IDLV and IPLV, with CDLV expression level remaining detectable throughout the initial 72 hours post-transduction (FIG. 4c). Quantification of vector DNA in the transduced cells did not reveal presence of proviruses in any of the CDLV-transduced cells (data not shown), confirming that the RT mutation has eliminated RT functionality.

[0106] Data presented in FIG. 4b and FIG. 4c appear to show that CDLV expression peaks at 24 hours post-transduction, whereas IDLV expression peaks at 72 hours. To compare their relative gene expression efficacy at these time-points, CDLV and IPLV vectors were titered by RNA genome copy number and administered to HEK 293T cells at known vector genome doses, with EGFP expression measured at their respective peak time-points. This investigation revealed that CDLV produced a similar number of EGFP-positive cells to IDLV, even when administered at a lower RNA copy number (P=0.05) (FIG. 4d). However, when examining the intensity of EGFP expression in the transduced cells, it was seen that IDLV treatment produced a greater level of expression per administered RNA genome (FIG. 4e) (P=0.005), likely due to the use of a strong SFFV promoter in the IDLV format, leading to mRNA abundance.

[0107] LV Vectors can Express ssRNA Payloads in Mouse Liver In Vivo

[0108] LV vectors are commonly pseudotyped with VSVg, a glycoprotein that confers broad tissue tropism by targeting the low-density lipoprotein receptor (LDLR) for cell entry (Finkelshtein, D et al. (2013), Proc. Natl. Acad. Sci. U.S.A 110: 7306-11). VSVg-pseudotyped LV vectors are particularly effective for in vivo liver transduction (Pan, D et al. (2002), Mol. Ther. 6: 19-29). The inventors investigated the effectiveness of their engineered CDLV vector for gene transfer to neonatal mouse liver in vivo, comparing its longitudinal expression profile to a conventional IDLV vector.

[0109] The inventors used a bicistronic transgene expressing luciferase and EGFP, separated by a 2A cleavage peptide derived from Thosea asigna virus (Szymczak, A L et al. (2004), Nat. Biotechnol. 22: 589-594). This Luc-EGFP reporter was packaged into IPLV and IDLV vectors driven by the SFFV promoter (IPLV-SFFV-Luc-EGFP and IDLV-SFFV-Luc-EGFP, respectively). The Luc-EGFP transgene was additionally packaged into a CDLV vector (CDLV-Luc-EGFP) (FIG. 5a). Neonatal outbred mice received intravenous injections of each vector on the day of birth and bioluminescent imaging began 2 hours after vector administration and continued for days (FIGS. 5b and 5c). Vector doses were calculated based on ssRNA titres, with IPLV being delivered at 110.sup.13 viral genomes per milliliter (vg/ml), IDLV delivered at 510.sup.12 vg/ml and CDLV delivered at 410.sup.11 vg/ml. As expected, IPLV vector expression was detectable at early stages post-transduction and expression intensity continued to increase throughout the course of the investigation. IDLV vectors showed reduction of bioluminescent signal after 96 hours post-transduction, indicating reduced persistence in transduced liver. CDLV vector expression was detectable by 2 hours post-injection, with its expression appearing to peak at 24 hours before falling below the limit of detection thereafter.

[0110] Discussion

[0111] Lentiviral vectors are effective gene transfer agents, with an ability to transduce a variety of cell types in vitro and in vivo. This has led to their application in a number of gene and cell therapies, particularly in circumstances where transgene capacity precludes use of AAV vectors, or cell targeting is suboptimal with non-viral vector technologies. Additionally, their ability to permanently integrate their DNA into dividing and non-dividing cells has made them a valuable tool in stem cell therapies, as modified cells will retain the therapeutic payload throughout cell division.

[0112] Here, the inventors show that HIV-1-based LV vectors can be used as transient mRNA delivery vehicles by engineering the reverse-transcriptase and RNA genome to promote translation of the transgene from the 5 m.sup.7G cap, which they show delivers immediate, but transient mRNA expression both in vitro and in vivo.

[0113] The findings of their work are of relevance to retrovirology and the mechanism of HIV-1 uncoating. The precise mechanism of lentivirus uncoating and the timing of genomic RNA release is not well defined. A number of mechanisms have been proposed, one of which is that capsid disassembly occurs during the early stages of reverse-transcription. Given that reverse-transcription is dysfunctional in the inventors' system, but EGFP is clearly detectable soon after cell entry, this suggests that uncoating may not be absolutely dependent on reverse-transcription and a significant amount of vector RNA is released immediately after cell entry, irrespective of reverse transcription.

[0114] During the development of the CDLV system, the inventors engineered a variety of gene expression cassettes designed to modify the packaging and expression of vector ssRNA. Version 2 was taken forwards to further investigations, given that it produced the most efficient EGFP expression. The inventors demonstrated that this structure clearly maximizes in vitro expression compared to an IRES-based version, in a conventional 3.sup.rd generation LV backbone. Further to this, the inventors also demonstrated that CDLV can transduce cells in vitro with comparable frequency to non-integrating and integrating 3.sup.rd generation LV vectors, but with complete transiency. Additionally, direct comparison of CDLV to an integration-deficient LV (IDLV) vector showed that CDLV could achieve greater transduction frequency than IDLV, albeit with weaker expression levels per transduced cell.

[0115] Finally, given efficient liver targeting of VSVg-pseudotyped LV vectors the inventors investigated the ability of CDLV to deliver transient gene expression to hepatocytes in vivo, employing a luciferase reporter to track live vector expression kinetics in vivo. This study showed that CDLV could indeed provide short-term transgene expression in vivo, demonstrating a similar profile to that obtained in vitro. Luciferase expression from the CDLV platform was comparable to IPLV and IDLV 3.sup.rd generation vectors during the early stages after injection, despite the 3.sup.rd generation vectors being delivered at higher doses. Additionally, it is important to note that in all of experiments 3.sup.rd generation IPLV and IDLV vectors were driven by a strong SFFV promoter, which will produce high levels of mRNA in hepatocytes and HEK 293T cells. In gene therapy, promoters weaker than SFFV are usually preferred, particularly in a lentiviral context, due to potential safety concerns. Therefore, the gene expression profiles that the inventors have detected from CDLV technology is likely to be comparable to clinically relevant lentiviral vector cassettes. However, validation of the in vivo scalability of their platform will require further studies beyond neonatal mice, given that larger animal models may not be easily transduced in vivo with lentiviral vectors.

[0116] The ability to deliver LV genomes as mRNA in vitro and in vivo brings some potential benefits in gene and cell therapy. LV vectors have a relatively large packaging capacity, able to package the mRNA of the majority of human genes. Therefore, CDLV technology has the potential to express large therapeutic transcripts transiently with high efficiency. This presents an advantage over non-viral gene transfer technologies, as transfection efficiency is known to reduce in correlation with increasing nucleic acid length. Indeed, LV vector gene transfer efficacy is also known to reduce in correlation with payload size, but it is noteworthy that this effect is thought to be limited primarily by inefficient reverse-transcription of large payloads, rather than ssRNA packaging, which suggests that CDLV payload tolerance could be even higher than that of conventional LV vectors.

[0117] Perhaps one of the most interesting avenues for exploiting CDLV technology would be delivery of genome editing nucleases. Cas9 nuclease mRNA has been used for in vitro and in vivo applications, although novel mRNA delivery strategies are continually being explored to enhance gene transfer efficiency. Lentiviral delivery of mRNA holds significant advantages here, as it has been shown over the past 20 years that LV vectors can be pseudotyped with a range of glycoproteins for targeted transduction of a wide range of cell types. Additionally, the packaging capacity of LV vectors means that Cas9 expression transcripts can be easily modified to include any of the recently developed modules (e.g. base editors, transcriptional activators and repressors), without compromising packaging ability.

[0118] An additional platform that could benefit from CDLV technology is in vaccinology, where researchers are developing methods to express antigens in antigen presenting cells (APCs) in situ. Lentiviral vectors have been investigated extensively for this purpose and it has been shown that a lack of pre-existing immunity allows repeated administrations. However, a potential limitation of LV use for APC transduction is the expression of SAMHD1 (SAM and HD domain-containing deoxynucleoside triphosphate triphosphohydrolase 1) in these immune cells, which lowers intracellular dNTP pools and reduces transduction efficiency by restricting the activity of reverse-transcriptase. Therefore, given that CDLV technology is not dependent on reverse-transcription for mediating expression, the inventors' platform technology could provide an advantage in this area of gene therapy.

[0119] In summary, the inventors report design and development of a novel LV gene structure that enhances translation of packaged genomic mRNA immediately upon cell entry, with limited duration. The inventors have shown that this novel CDLV vector can be used for gene expression both in vitro and in vivo, for potential applications in gene therapy.

[0120] Materials and Methods

[0121] Generation of Plasmid Constructs

[0122] All plasmid constructs were made using standard molecular cloning procedures and PCR-mediated deletion of plasmid sequences (Hansson, M D et al. (2008), Anal. Biochem. 375: 373-5). In cases where novel sequences or HIV-1 sequence deletions were incorporated, synthetic DNA fragments were designed and ordered as gBlocks (Integrated DNA Technologies).

[0123] Production of Lentiviral Vectors

[0124] Lentiviral vectors were produced as described previously (Vink, C A et al. (2017), Mol. Ther. 9: 10-20). Briefly, 1.810 7 HEK293T cells were plated per 15 cm sterile culture dish and transfected with the following components: 40 g of the relevant transfer plasmid, 20 g of pMDLg.RRE, 10 g of pRS V-Rev and 10 g of pMDG.2 (all plasmids produced by PlasmidFactory). Additionally, 10 g of pCMV-Tat (kindly provided by Professor Axel Schambach from Hannover Medical School (Huelsmann, P M et al. (2011), BMC Biotechnol. 11: 4) was supplemented for enhanced vector titres. The plasmid mixtures were added to 5 mL Opti-MEM and filtered through 220 nm sterile filter units. Filtered DNA was combined with 5 mL Opti-MEM (Life Tech/GE) containing 2 M polyethylenimine (PEI, Sigma). The resulting 10 mL mixture was incubated at room temperature for 10 minutes before addition to HEK 293T cells. After 4 hours, the transfection mixture was replaced with fresh culture medium. Virus supernatant was collected at 48 hr and 72 hr post-transfection. After each harvest, the collected medium was filtered through a cellulose acetate membrane (0.45 mm pore). Lentivirus harvests were combined before concentration by ultracentrifugation. Briefly, viruses were placed in polyallomer centrifuge tubes (Beckman Coulter) and centrifuged for 2 hr at 90,000g at 4 C. in a Sorvall Discovery 90SE Centrifuge. Following centrifugation, the supernatant was removed, and pellet recovered in 200 L Opti-MEM.

[0125] Titration of Lentiviral Vectors

[0126] Vector titration by flow cytometry: 1105 HEK293T cells were plated into each well of a 6-well plate and transduced with a dose-escalation of concentrated lentivirus. For IDLV and IPLV vectors, EGFP measurements were made at 72 hours post-transduction, whereas CDLV analysis was performed at 16 hours post-transduction. Titres were calculated based on cell populations in the range of 5-30% EGFP+, as described previously (Vink, C A et al. (2017), Mol. Ther. 9: 10-20). EGFP positive cells were identified as described below in Detection of eGFP expression in transduced cells.

[0127] Vector titration by p24 capsid antigen: A p24 ELISA kit (Clontech product 632200) was used to determine the LV vector capsid number, according to the kit manufacturer's calculations, where 1 ng p24 is equivalent to 1.25107 lentiviral particles.

[0128] Vector titration by ssRNA genome copies: ssRNA genome copies were quantified using a qRT-PCR titration kit (Clontech product 631235). In brief, vector RNA was initially extracted from viral particles using spin columns and quantified by nanodrop. The vector RNA copy number was then calculated using an RT-qPCR assay targeting the HIV-1 RNA packaging sequence and extrapolating the absolute value from a standard curve of known vector genome copy numbers.

[0129] Detection of eGFP Expression in Transduced Cells

[0130] Unless stated otherwise, 100,000 cells were analyzed for EGFP expression in a BD FACSArray Bioanalyzer. During analysis, live cells were determined by gating forward-light-scatter versus side-scatter and isolating the relevant population. EGFP-positive cells were determined by plotting EGFP fluorescence (detected using a 530/30 nm bandpass filter) versus emission from the yellow channel (detected using a 575/26 band-pass filter) to compensate for auto-fluorescence. Non-transduced controls were used to gate background expression in each channel. All flow cytometry data were analyzed by FlowJo software version 9.3.1 (Tree Star).

[0131] Animal Procedures

[0132] For in vivo investigations, outbred CD1 mice (Charles River), were time mated to produce neonatal animals. At postnatal day 1, non-randomised neonates were subjected to brief hypothermic anaesthesia and intravenously injected with LV vectors via the superficial temporal vein. Experimental groups were blinded during the course of in vivo investigations. Experiments were carried out under United Kingdom Home Office regulations and approved by the ethical review committee of University College London.

[0133] Longitudinal Tracking of Vector Expression In Vivo

[0134] To monitor LTR1 bioluminescence in vivo, 40 l of the relevant luciferase expression vector was administered intravenously to 1-day old neonatal CD-1 mice. Vector doses were based on ssRNA titration results. Doses were calculated as 110.sup.13 vg/ml for IPLV, 510.sup.12 vg/ml for IDLV and 410.sup.11 vg/ml for CDLV. Images and bioluminescence data were gathered continually for 10 days, as described previously (Buckley, S M K et al. (2015), Sci. Rep. 5: 11842), by intraperitoneal injection with firefly D-luciferin (150 mg/kg) and imaged after 5 minutes with a cooled charge-coupled device (CCD) camera (IVIS Lumina II, PerkinElmer). Detection of bioluminescence in the liver was performed using the auto region of interest (ROI) quantification function in Living Image 4.4 (PerkinElmer). Signal intensities were expressed as photons per second per centimeter 2 per steradian.

[0135] Statistical Analysis

[0136] All statistical analyses were carried out using Matlab 2015a. A Kruskal-Wallis test with Dunn's posthoc analysis was used to compare vector titres. Student's t-test was used to compare mean fluorescence intensities and EGFP values. In vitro experiments in cultured cells were performed in 4 experimental replicates. Mouse sample sizes were limited to 4 animals per experimental group for in vivo investigations.

[0137] Sequences

[0138] The sequences of the constructs depicted in FIG. 2a are provided in the associated sequence listing and the position of the various features of the constructs in the sequences are as follows:

[0139] CDLV1SEQ ID NO: 1

[0140] Promoter: nt 1-605; Intron: 753-885; EGFP transgene: 942-1661; WPRE: 1677-2265; 3 LTR (U3-R-U5): 2352-2585; Gag: 2723-3061; RRE: 3232-3465; pA: 3939-4073.

[0141] CDLV2SEQ ID NO: 2

[0142] Promoter: nt 1-605; EGFP transgene: 619-1338; WPRE: 1364-1932; 3 LTR (U3-R-U5): 2029-2262; Gag: 2400-2738; RRE: 2909-3142; pA: 3616-3750.

[0143] CDLV3SEQ ID NO: 3

[0144] Promoter: nt 1-605; EGFP transgene: 619-1338; WPRE: 1364-1932; bGHpA: 1952-2176; 3 LTR (U3-R-U5): 2248-2481; Gag: 2619-2957; RRE: 3128-3361; pA: 3835-3969.

[0145] CDLV4SEQ ID NO: 4

[0146] Promoter: nt 1-605; EGFP transgene: 619-1338; WPRE: 1354-1942; bGHpA: 1952-2176; Intron: 2206-2338; 3 LTR (U3-R-U5): 2453-2686; Gag: 2824-3162; RRE: 3333-2566; pA: 4040-4174.

[0147] CDLV5SEQ ID NO: 5

[0148] Promoter: nt 1-605; EGFP transgene: 619-1338; 3 LTR (U3-R-U5): 1440-1673; Gag: 1811-2149; RRE: 2320-2553; WPRE: 3031-3619; pA: 3620-3754.

[0149] CDLV6SEQ ID NO: 6

[0150] Promoter: nt 1-605; EGFP transgene: 619-1338; U5: 1355-1449; Gag: 1576-1914; RRE: 2085-2318; WPRE: 2796-3384; pA: 3385-3519.