Real-Time Quantitative PCR for the Design of Lentiviral Vector Analytical Assays

Gene Therapy (2005) 12, S36–S50 & 2005 Nature Publishing Group All rights reserved 0969-7128/05 $30.00 www.nature.com/gt CONFERENCE PAPER Real-time quantitative PCR for the design of lentiviral vector analytical assays

C Delenda1 and C Gaillard2 1Genethon, CNRS UMR 8115, 1bis rue de l’Internationale, Evry Cedex, France; and 2GenoSafe, 1 rue Pierre Fontaine, Evry Cedex, France

From the recent and emerging concerns for approving context, have been included in the effort to dress an lentiviral vector-mediated gene transfer in human clinical exhaustive list. Also, great variations have been observed applications, several analytical methods have been applied from interlaboratory results, we have tempted to compare in preclinical models to address the lentiviral vector load between them the different analytical methods that have in batches, cells or tissues. This review points out the oldest been used to consider (i) the titration of lentiviral vector generation methods (blots, RT activity, standard PCR) as batches, (ii) the absence of the susceptible emerging well as a full description of the newest real-time quantitative replicative lentiviruses or (iii) the lentiviral vector biodistribu- PCR (qPCR) applications. Combinations of primer and probe tion in the organism. sequences, which have worked in the lentiviral ampliﬁcation Gene Therapy (2005) 12, S36–S50. doi:10.1038/sj.gt.3302614

Keywords: real-time PCR; lentiviral vector; standardization

Introduction This review points out the major progress undertaken for the standardization of some of these technical Lentiviral-derived transfer vectors have gained increased expertises. Associated with conventional detection attention because their karyophilic properties allow their methods, techniques derived from qPCR have been use for the transduction of quiescent cells.1,2 The applied in the lentiviral vector context for the evaluation acceptance of their use in clinical settings will require of titration, RCL and biodistribution in animal models. profound and vigorous assays in order to evaluate the quality as well as the quantity of lentiviral particles in vector batches. Description of pre-existing lentiviral vector Validation and safety tests that will be required for the use of lentiviral vectors in human gene transfer should titration methods include (i) the search for susceptible emerging replicating The main options for titering lentiviral vectors include forms (assays for replication-competent lentiviruses assessment of the reverse transcriptase (RT) activity and (RCL)), (ii) several sterility tests in order to detect the viral RNA content in supernatants, as well as the remnants from mycoplasma, fungi or bacteria as well number of proviral DNA and the efficacy of transgene as contaminants originating from adventitious agents, expression in transduced target cells. Specifically dedi- (iii) the determination of the viral vector titers as well cated to lentiviral vectors derived from the human as (iv) in vivo preclinical studies for demonstrating immunodeficiency virus (HIV), an enzyme-linked the proof of complementation concept. In addition, the immunosorbent assay (ELISA) directed against the p24 organ biodistribution of lentiviral vectors should also capsid protein has also been routinely employed for be addressed. vector titer determination. In this concern, GenoSafes www.GenoSafe.com, a service company, has been recently founded by Gene- Lentiviral vector titration from transgene protein thon to provide gene transfer safety and quality tests for expression preclinical and clinical studies. Overall, this company is dedicated to vector characterization, containing titration For lentiviral vectors containing a reporter transgene and replication-competent virus experiments as part expression cassette, ‘infectious’ particles (ip) are usually of quality control for preclinical products, and safety, determined by evaluating the percentage of target cells preclinical in vivo studies, as well as gene therapy- that express the transgene protein. specific patient follow-up. GenoSafes operates under Depending on the type of reporter used, lentiviral validated GLP conditions. ‘infectious’ units are defined in several ways. Colony- forming units are generally used for drug selection genes.3 Relative luciferase units are defined for renilla as well as firefly luciferase gene activities.4–6 The term of Correspondence: Dr C Delenda, Genethon, CNRS UMR 8115, 1bis rue de transducing unit used in the context of lentiviral vector l’Internationale, 91002 Evry Cedex, France gene transfer has been routinely employed, especially for qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S37 the expression of living color and lacZ genes.7,8 Protocols Whereas the sensitivity of first generation RT assays for the determination of functional viral titers by drug- ranges from 104 to 106 viral particles,27 it decreases to resistance colony assay, expression of green fluorescent 1–10 virions when PERT assays are used.18,20,21 Such protein (GFP) and b-galactosidase staining can be found assays, which are based on the endogenous RT activity of in Tonini et al.9 retrovirions, have been commonly employed for the Additionally, infectious titers of vectors containing verification of the absence of retroviral-based adventi- nonreporter gene cassettes have been determined by tious viruses in different biological materials,28,29 but FACS analysis using primary antibodies directed against they have also served for isolating lentiviral packaging the transgene protein. However, this application is cell clones able to secrete functional empty capsids, limiting as it requires antibody for all type of transgene where the term ‘functional’ argued for RT encapsida- and because it is restricted to some specific cellular tion.30 compartments, predominantly at the plasma membrane. By comparing different analytical methods for the Nevertheless, mild permeabilization of the plasma titration of lentiviral vectors derived from the equine membrane has already been described to allow cellular infectious anemia virus (EIAV), Martin-Rendon et al31 antibody uptake and vector recognition within the have shown that there was no correlation between PERT cytoplasmic compartment.10,11 assays and eGFP titers, by contrast to the results obtained Several optimizations undertaken in some of these with a RT-PCR titration method (cf Quantification of reporter genes have been addressed in the context of the viral RNA genomes). lentiviral gene transfer. To improve the LacZ gene expression in mammalian cells, Anson and Limberis12 have codon optimized the Escherichia coli LacZ gene. HIV-1-based vector titration from capsid content From this mutagenesis, they have observed a 15-fold measurement increase in b-galactosidase expression as well as a five- The assay commonly used for titering HIV-1-derived fold enhancement in HIV-1 vector titers, apparently due vectors that do not contain a reporter gene cassette is to the new opportunity to detect transduced cells that based on an ELISA test directed toward the capsid did not highly express the transgene. Concerning the protein (p24). This approach is generally used to GFP reporter protein, a lower cytotoxic mutant (from determine the number of ‘physical’ particles (pp), but the sea pansy Renilla reniformis) has been compared to the is limited by the variable amount of free, nonparticle- original enhanced GFP (eGFP) from the jellyfish Aequorae associated p24 proteins that are produced by standard victoria. However, in vivo lentiviral gene transfer studies plasmid cotransfection methods of lentiviral production. did not show a clear advantage in term of cytotoxicity of Based on the estimation that a HIV core particle is this newly used GFP.13,14 Whereas one report has shown composed of around two thousands capsid proteins,32 evidence of rejection in the retinal epithelium of rats 1 fg of p24 therefore represents around 12 pp. From this when using the eGFP reporter transgene,15 Duisit et al13 calculation, it is therefore able to depict the ip/pp ratio did not observe neither sign of rejection nor difference in for any HIV-1-based lentiviral vector production, as from the long-term expression. In addition, although cytotoxi- the moment when ip titer measurements are possible. city has been observed after intracorneal injection of cats, The ip/pp ratio estimation is routinely employed for it was only noticed in conditions where the eGFP protein quality control validation of the different batches being has been overexpressed, and was solved by lentiviral produced. In fact, it allows the rejection of vector batches vector dose reduction.14 having a too great quantity of non-ip, which may result from gag-pol polyprotein overexpression and/or matura- Lentiviral vector titration from RT activity tion problems. Although concrete specifications are not The first generation of RT assays measures the incorpora- formally described in the litterature, it is generally tion of radiolabelled deoxyribonucleoside triphosphates accepted to consider that a relevant lentiviral vector into DNA using viral genomic RNA as template.16 The batch is defined when the ip/pp ratio is contained in a quantity of incorporated radiolabelled nucleotides is window of 1/100–1/1000. then measured by a scintillation counter or by auto- radiography. This RT protocol dedicated for retroviral and lentiviral vector titration can be found in Tonini Lentiviral vector titration from viral nucleic acid et al.17 An additional radiolabelled technique has also detection been developed where the detection system relied on In the absence of marker genes, lentiviral vector titers hybridization of newly synthesized cDNAs by Southern expressing therapeutic genes have to be determined by blotting.18 alternative methods. Semiquantitative Northern blotting To circumvent the use of radiolabelled elements, experiments have been used to estimate the particle nonradioisotopic assays have been conducted. They are number in lentiviral supernatants.33,34 In a concern to based on the incorporation of digoxigenin-labelled dUTP generate and isolate stable producer cell clones, some during the RT step followed by conventional biotin/ laboratories have also conducted screening experiments peroxidase secondary reactions.19 To increase the signal on retroviral supernatants by the use of dot blot detection, therefore allowing the specificity improvement assays.35,36 However, that kind of blotting assays cannot of RT-based assays, product-enhanced RT (PERT) assays be employed to exactly quantitate the viral RNA content. have been designed, where neosynthesized cDNAs were Moreover, they generally acknowledged to greatly over- further amplified by standard PCR.20–23 Newest PERT estimate functional viral titers. generations have also been performed in the design of Semiquantitative Southern blotting has been described qPCR by the addition of Taqman-based fluorescent for the detection of proviral DNA genomes.37 In addition, oligonucleotide probes.24–26 competitive RT-PCR and standard PCR have been

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S38 applied for a more exact quantitation of viral RNAs and tical titration methods. Owing to the absence of proviral DNAs, respectively.38,39 comparison in some of these reports, it was therefore Interestingly, nucleic acid testing (NAT) methods have sometimes difficult to dress an exhaustive assessment. been certificated by suppliers for HIV diagnosis such Moreover, concerning the determination of eGFP as nucleic acid-based sequence amplification (NASBA, titers, it should be important to take into consideration Organon Nuclisens RNA assay),40 branched DNA that result variations between laboratories are commonly (bDNA, Chiron Quantiplex and Bayer Versant RNA observed since eGFP titers are dependent on the nature assays),41,42 or transcription-mediated amplification of the promoter used to express the transgene as well as (TMA, Gen-Probe RNA assay).43 The main reason why on the cell type being chosen. A recent article has shown these assays have not been used for the titration of that the lentiviral vector titration varies with changes of nonreplicating HIV vectors reside upon the fact that their conditions in the transduction process, including the target sequences, which are mostly derived from viral inoculum volume, the type but also the number of target genes, are no longer present in the recombinant transfer cells and the length of period of the vector adsorption vector genomes. to target cells.52 They have shown that the variation of these parameters has resulted in a greater than 50-fold differences in the vector titer from the same Description of the different qPCR-based vector stock. methods applied to lentiviral vector titration In the same extent, transfer vectors that contain the central polypurine tract (cPPT) will give higher titers Real-time qPCR is an important scientific advancement in transduced cells (eGFP and proviral DNA titers) as that greatly extends the usefulness of PCR technology. It compared to vectors that do not have the cis-active facilitates the monitoring of the reaction as it progresses. nuclear translocation facilitator, and since not all labora- One can start with minimal amounts of nucleic acid tories have conducted their assays with such a sequence and quantify the end product accurately. Moreover, inserted in their lentiviral vector backbone, it was there is no need for the post-PCR processing which therefore difficult to depict solid conclusions. However, saves the resources and the time. These advantages of we have tried in this chapter to highlight the real benefit the fluorescence-based qPCR technique have completely of qPCR-based assays for the standardization of lenti- revolutionized the approach to PCR-based quantification viral vector titration over classical existing techniques. of DNA and RNA. Real-time assays are now easy to perform, have high sensitivity, more specificity, and Concern for plasmid DNA carryover provide scope for automation. Susceptible DNA contamination due to plasmid carry- The fluorescent reporter molecules include dyes that over from the lentiviral transfection process is generally bind to the double-stranded DNA (ie SybrGreen I) or eliminated by a DNA nuclease treatment. Depending on sequence-specific probes (ie TaqMan, molecular beacons the source of the nucleic acid being amplified, this prior or Scorpion probes). Taqman- and SybrGreen I-based nucleasic incubation has been addressed directly on qPCR strategies have already been used in the design of lentiviral vector preparations and on isolated viral RNAs lentiviral vector titrations, where different viral nucleic to avoid plasmid DNA contamination in genomic DNA acid species have been successfully quantified, such 31,44–47 and viral RNA extracts, respectively. as viral RNAs, and ‘strong stop’ minus strand 45,53 44,48 Sastry et al have followed, after DNase I and cDNAs, in vector batches, as well as proviral benzonase treatments, the disappearance of plasmid DNAs,31,44,45,47,49,50 and transgenic mRNAs in transduced 47,51 DNA species, by amplifying by qPCR a plasmidic target cells. sequence, that is, an ampicillin (ampr) DNA fragment. This chapter tempts to highlight and compare between Although a large amount of the ampr gene was detected several laboratories and different analytical assays the in vector supernatant, only a low percentage (o0.01%) principal titration results they have obtained. All qPCR- of the plasmid DNA was transferred to cells upon vector based assays that have been developed to date for transduction. The lack of plasmid contamination after lentiviral vector titrations are summarized in Table 1A. DNase I treatment has also been confirmed in different Primers/probe location and sequences related to lentivir- lentiviral vector preparations by conducting standard al-derived transfer vector genomes are designed either on PCR also directed against the ampr gene.47 endogenous (LTR, primer-binding site), encapsidation From the next mode of lentiviral vector production, signal) or exogenous (eGFP transgene, cytomegalovirus which is based on the design of stable producer cell lines, (CMV) promoter, WPRE) DNA elements. of course, plasmidic DNA remnants in vector batches In addition, Table 1A also shows several primers/ will not be anymore an impediment. probe sets that have been used for different control testing experiments. These include (i) the amplification of human cellular endogenous genes (b-actin and Measurement of viral nucleic acids within vector albumin DNA sequences) for exactly measuring, after batches transduction, the number of proviral genomes per target Assessing titer in lentiviral vector supernatants is cell, (ii) the verification of the absence of plasmid DNA technically the easiest method and provides the most remnants in transduced cells (amplification of the rapid turnaround. However, viral genome titration ampicillin gene) or (iii) the exact measurement of the methods may not accurately reflect the functional titer concentration of genomic DNA extracts (amplification of due to the presence of defective interfering particles, the human mitochondrial DNA). inhibitors of transduction, and eventually carryover In an attempt to depict solid conclusions, we have DNA from vector production when prior nuclease decided to compare between them the different analy- treatment was not a concern.

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S39 Table 1A List of qPCR-based assays applied to the titration of lentiviral vectors

References Vector and region Primers and probes dedicated for the Primers and probes dedicated for the amplification of amplified amplification of lentiviral or transgenic internal control DNAs (cellular genome or plasmid) DNA sequences

Scherr et al44 HIV-1 (R-U5) Forward primer (78–97)(1) Normalization with the ampliﬁcation of a human AGCTTGCCTTGAGTGCTTCA albumin DNA fragment (intron 12) Reverse primer (161–141) Forward primer TGACTAAAAGGGTCTGAGGGA GCTGTCATCTCTTGTGGGCTGT Probe (107–129) Reverse primer F-TGCCCGTCTGTTGTGTGACTCTG-T GCTGTCATCTCTTGTGGGCTGT Probe V-CCTGTCATGCCCACACAAATCTCTC-T

Butler et al49 HIV-1 (U5-Psi) Forward primer (103–122) TGTGTGCCCGTCTGTTGTGT Reverse primer (245–226) GAGTCCTGCGTCGAGAGAGC Probe (179–198) Quantiﬁcation of genomic DNAs by prior ampli- F-CAGTGGCGCCCGAACAGGGA-T ﬁcation of a human mitochondrial DNA fragment Forward primer HIV-1 (30LTR-Alu) Forward primer (9131–9153) ACCCACTCCCTCTTAGCCAATATT AACTAGGGAACCCACTGCTTAAG Reverse primer Reverse primer (Alu) GTAGGGCTAGGCCCACCG TGCTGGGATTACAGGCGTGAG Probe Probe (9644–9251) Te-CTAGTCTTTGCCGCCTGCGAAGCA-T F-ACACTACTTGAAGCACTCAA GGCAAGCTTT-T

HIV-1 (U5-U3, Forward primer (9131–9153) 2LTR) Reverse primer (245–226) TCCACAGATCAAGGATATCTTGTC Probe (9644–9251)

Ikeda et al46 HIV-1 and EIAV Forward primer (438–461)(2) Absence of normalization (only quantiﬁcation of (eGFP) CAACAGCCACAACGTCTATATCAT viral RNA genomes) Reverse primer (515–495) ATGTTGTGGCGGATCTTGAAG Probe (464–488) F-CCGACAAGCAGAAGAACGGCATCAA-T

Sastry et al45 HIV-1 (PBS-Psi) Forward primer (198–218) Veriﬁcation of plasmid DNA remnants in ACCTGAAAGCGAAAGGGAAAC transduced target cells by the ampliﬁcation of an Reverse primer (342–320) ampicillin DNA fragment CACCCATCTCTCTCCTTCTAGCC Forward primer (472–490) Probe (225–248) ACTCGCCTTGATCGTTGGG F-AGCTCTCTCGACGCAGGACTCGGC-T Reverse primer (563–543) GTTGCCATTGCTACAGGCATC Probe Te-CTAGTCTTTGCCGCCTGCGAAGCA-T

Martin-Rendon EIAV (CMV Normalization with the ampliﬁcation of et al31 promoter) Forward primer a human b-actin DNA fragment CATATATGGAGTTCCGCGTTACAT Forward primer (584–601) Reverse primer AGCGCGGCTACAGCTTCA GTATGTTCCCATAGTAACGCCAATAG Reverse primer (660–641) Probe GGCGACGTAGCACAGCTTCT F-TGGCTGACCGCCCAACGACC-T Probe (632–612) V-TCACGCACGATTTCCCGCTCG-T

Lizeé et al47 HIV-1 (WPRE) Normalization with the amplification of a human albumin DNA fragment Forward primer (1554–1578) TGAAACATACGTTCCCAAAGAGTTT Reverse primer (1634–1609) CTCTCCTTCTCAGAAAGTGTGCATAT Forward primer Probe (1581–1607) CCGTTGTCAGGCAACGTG F-TGCTGAAACATTCACCTTCCATGCAGA-T Reverse primer AGCTGACAGGTGGTGGCAAT Normalization with the amplification of Probe a human b-actin RNA fragment V-TGCTGACGCAACCCCCACTGGT-T Forward primer (347–366) GCGAGAAGATGACCCAGATC

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S40 Table 1A Continued

Reverse primer (449–431) CCAGTGGTACGGCCAGAGG Probe (388–413) F-CCAGCCATGTACGTTGCTATCCAGGC-T

Van Maele et al57 HIV-1 (U5-Psi) Forward primer (103–122) Absence of normalization Reverse primer (245–226) Probe (179–198) Cf Butler et al49

HIV-1 (30LTR-Alu) Forward primer (9635–9655) GTGCCCGTCTGTTGTGTGACT Reverse primer (Alu) Cf Butler et al49 Probe (9665–9697) F-TCCACACTGACTAAAAGGG TCTGAGGGATCTCT-T

HIV-1 (U5-U3, Forward primer (9131–9153) 2LTR) Cf Butler et al49 Reverse primer (8639–8664) CTTGTCTTCTTTGGGAGTGAATTAGC Probe (9665–9697)

Charrier et al50 HIV-1 (R-U5) Forward primer (78–98) AGCTTGCCTTGAGTGCTTCAA Reverse primer (153–129) AGGGTCTGAGGGATCTCTAGTTACC Normalization with the ampliﬁcation of a human albumin DNA fragment Scherr et al51 HIV-1 (RFP) Forward primer (157–174) forward and reverse primers GCCCTTCGCCTGGGACAT cf Scherr et al44 Reverse primer (221–200) GGTGCTTCACGTACACCTTGGA Probe (176–198) F-CTGTCCCCCCAGTTCCAGTACGG-T

Forward and reverse primers as well as probe sequences are indicated. In addition to the amplification scheme of DNA elements contained in lentiviral genomes (3rd raw), additional primer and probe sequences that are used as controls are also noticed in the last raw. Numbers in parentheses indicate nucleotide positions in the HIV-1 viral genome(1) or in human cDNAs.(2)Nucleotide positions referred to cis-acting sequences (CMV promoter, WPRE) are not given since their sequence delimitation is not identical according to different Genbank accession numbers. Relative to the Taqman-based design, fluorophores added in the 50- and 30-ends of probe sequences are also indicated; F: 6-FAM, T: TAMRA, V: VIC, Te: TET. For assays where no probe sequence is listed, the amplification signal is given by the incorporation of the SybRGreen I binder agent.

Quantification of viral RNA genomes not) in the transfer vector genome as well as on the type Viral genomic RNAs are isolated from lentiviral vector of cell being transduced, it may participate to the overall preparations and used as templates in one-step RT-qPCR variation observed. Additional intrinsic experimental reactions. To eliminate the participation of plasmid DNA factors may also contribute to the absence of correlation background, RNA concentration is generally determined between eGFP and RT-qPCR titers, such as technical by substracting plasmid copy number in reaction runs parameter deviations: primer/probe specificity, validity without RT from those with RT. of the DNA concentration measurement, quality of the Such RT-qPCR titration assays have already been RNA extraction, hand-to-hand variations. Moreover, it conducted in several laboratories.44–47 From all results has been shown that the RT-qPCR titration method gained, the number of vector molecules was in each case severely overestimates functional titers because of the higher than that observed by the eGFP titration method known presence of defective contaminating lentiviral (Table 1B). This suggests a considerable disparity particles, which can frequently outnumber functional between the number of vector sequences and the number particles by 20-fold or more.33,34 Since technical details of vector particles able of successfully transducing target for the lentiviral production design may vary between cells. groups, differences in the particle formation and matura- However, the level of amplitude is not identical tion will not be surprising. between groups, varying to 200- to 10 000-fold increase. In spite of the lack of homogeneity between groups, As already mentioned above, since the transgene it has been, however, possible in the same report to expression can be affected according to the type of compare RNA packaging and transduction efficiency enhancer sequences (promoter, cPPT) that are present (or for two different origins of lentiviral vectors, that is,

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S41 Table 1B Description of qPCR-based lentiviral vector titration methods by directly measuring in batches different viral nucleic acid species

Mode of titration Comparison with Additional conclusions References Lentiviral vector standard eGFP type titration (target cells, days p.t.)

Quantification 200-fold increase o1% of the vector particles are Scherr et al44 HIV-1 of viral RNAs (293 T, 3) infectious (RT-qPCR) 10 000-fold increase As compared to HeLa and Mus Sastry et al45 (293, 4) dunni cells, the 293 cell line has the highest transduction efficiency 500-fold increase Same RNA packaging activity Ikeda et al46 HIV-1 and EIAV (HIV-1) 10 000-fold between EIAV and HIV-1 but increase (EIAV) EIAV vectors are 20-fold less (HeLa, 3) efficient for transduction in human cells

1200-fold increase Vectors expressing cytotoxic Lize´e et al47 HIV-1 (HeLa, 4) proteins have consistantly lower titers than an eGFP-expressing reference vector Positive correlation No correlation with a PERT assay Martin-Rendon et al31 EIAV

ND: values in Ct (D17, 7–10)

Quantiﬁcation of 100-fold increase Sensitivity o20 copies Scherr et al44 HIV-1 strong stop cDNAs (293 T, 3) (R-U5 qPCR)

Two different assays have been applied for lentiviral vector titration: one directed against the viral RNA genome (RT-qPCR), the other based on the direct ampliﬁcation of encapsidated strong-stop cDNAs (qPCR).

ND ¼ not determined. Ct ¼ threshold cycle.

HIV-1- and EIAV-related genomes.46 In fact, the direct of viral RNA genomes), Scherr et al44 have shown that the comparison between both vectors has permitted to show number of viral particles measured was similar, with that although the amount of RNA packaged was similar, only twice as many particles detected when the RT-qPCR EIAV-based vectors were 20-fold less efficient for assay was used (Table 1B). This difference may be due to transduction in human cells. In the context of EIAV- the fact that not all lentiviral cores contain the strong- derived gene transfer, another report has also confirmed stop cDNA.54 In fact, some experimental conditions the positive correlation between this RNA-based titration during the production scheme, for example high approach and eGFP titers.31 confluency, cause differences in the pool of free nucleo- Concerning the RNA packaging efficiency in HIV-1 sides available in producer cells, therefore abrogating or vector virions, Lizeé et al47 have shown that it is allowing partial endogenous RT process in some viral dependent on the nature of the transgene being cores. From this latter event, uncompletely reverse- expressed in producer cells. Lentiviral transfer vectors transcribed cDNAs, that have been termed ‘weak-stop’ that encode cytotoxic transgenic proteins, that is, tumor- cDNAs, have been observed in wild-type HIV-1 associated antigens and neo-poly(A) polymerase, have virions.55 consistently lower genomic RNA titers as compared to an eGFP-expressing reference vector. Measurement of proviral DNA genomes within transduced target cells Quantification of strong stop cDNAs In the most majority of the assays already been Before retroviral entry into the cell, RT of the viral RNA conducted, lentiviral vector integration events have been genome is initiated in virion particles. In this endogen- calculated by normalizing the number of proviral ous step, the RT reaction, primed by the primer-binding molecules to the number of target cells, by a parallel or site anchored tRNA, starts by minus (À) strand cDNA duplex quantification of an endogenous cellular gene (cf synthesis, through R-U5, and leads to the formation Table 1A), such as DNA fragments derived from human of (À) sense cDNA species. Therefore, the viral RNA albumin,44,50 or b-actin genes.31 load present in any vector batch can be indirectly estimated by measuring the amount of these encapsi- Quantification of total proviral copies per cell dated cDNAs. Amplification of these so-called ‘strong- Quantitation of proviral DNA in transduced target cells stop’ cDNA species has been addressed in lentiviral is likely to provide a more accurate measurement of viral vector particles,44 by reproducing what have been titer than assessing viral particles released by producer previously performed for an oncoretroviral vector cells, because only transduction-competent recombinant titration design.48 vectors will be detected by this method. By comparing this DNA-based method to the pre- The majority of laboratories, who have already viously reported RT-qPCR approach (cf Quantification developed a RNA-based analytical method, have also

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S42 Table 1C Description of different qPCR-based lentiviral vector ampliﬁcations by measuring in transduced target cells distinct proviral DNA species

Mode of titration Comparison with standard Additional conclusions References Lentiviral eGFP titration (target cells, vector type days p.t.)

Quantiﬁcation of total Similar titration results No bias observed between proviral Scherr et al44 HIV-1 proviral DNAs per (293 T, 3) DNA-based and eGFPtiters target cell (qPCR) Similar titration results No bias observed between proviral Scherr et al53 (K562, 3) DNA-based and RFP titers 10-fold increase (CMV o0.01% of the plasmid DNA present in Sastry et al45 promoter) vector supernatants is transferred to cells 60-fold increase (PGK upon vector transduction promoter) (293, 4 or 14) six-fold increase (HeLa, 4) Due to lower RNA titers, vectors encoding Lize´e et al47 cytotoxic proteins generate lesser amounts of integrated proviruses as compared to an eGFp reference vector ND: nonreporter gene Also included primers and Taqman probe Charrier et al50 (HT1080, 4) directed against the gene responsible for the human Wiskott–Aldrich syndrome for mRNA transgene follow-up (RT-qPCR) Correlation depending on Although EIAV-based vectors generate low Martin- EIAV the cell type (HeLa& D17, eGFP titers in HeLa cells, the number of Rendon et al31 7–10) proviruses is equivalent as the one obtained in a more permissive cell line (D17)

Quantiﬁcation of ND (293 T and SupT1, 30) Most of the late RT proviruses are not Butler et al49 HIV-1 integrated proviral converted into integrated DNAs per target cell provirusesCorrelation with Southern blotting (Alu-LTR qPCR) experiments Quantiﬁcation of the At 72 h p.t., integrated copies account for the amount of 2LTR-circle total proviral DNA titration molecules over total 2-LTR circles decline in abundance at 24 h proviral DNAs (U5-U3 p.t., primarily due to dilution by cell division qPCR)

Three different assays have been designed in order to detect several types of proviruses (total, integrated or 2-LTR circles).

addressed, by qPCR, the possibility to determine the effect in a low extent since the number of such DNA lentiviral vector titration by measuring the number of species is in minority at the time where DNA extracts proviral molecules per target cells (Table 1C). were isolated (cf Evaluation of the disappearance over- With the exception of one laboratory which obtained time of circle proviral genomes). Therefore, the major roughly similar titer results to standard reporter gene reason for overestimating the number of transducing- (eGFP and RFP) titration,44,51 the other data in the competent particles will preferably be explained by the literature suggest, in any case, that DNA proviral-based proviral integration in heterochromatin regions, where qPCR assays overestimate eGFP titers, varying from six- highly condensed structures do not allow transgene to 60-fold.31,45,47,49,50 The major differences were observed transcription. in Sastry et al,45 where a 60-fold variation was noticed according to the type of promoter used. The difference observed between these two analytical Quantification of integrated proviral copies per cell titration methods could be explained by the fact that not Detection of integrated proviruses has been accom- all proviral DNA molecules correspond to transcriptio- plished, in the context of HIV-1 gene transfer, by nally active forms. These qPCR assays are amenable to amplification with primers complementary to LTR and detect in transduced cells the totality of proviruses that chromosomal Alu repeats (Table 1C).49 Integrated viral corresponds to unintegrated linear and circular mole- DNA was detectable by 24 h post-transduction (p.t.) and cules as well as integrants. The common idea that many reached a plateau by 48 h p.t. By comparing the number retrovirologists thought from the past, that consists to of integrated proviruses over the one obtained after the consider that circular proviral genomes are transcrip- amplification of total proviral DNAs, the authors have tionally incompetent, has been recently reconsidered. shown that not all proviruses were converted to In the context of vectors derived from HIV-1 or feline integrated copies. The rate of fall of the amount of total immunodeficiency virus, integrase-defective mutants proviruses was greater than could be accounted for by have been shown to be able to deliver the transgene dilution during cell growth, indicating that the viral signal,56 suggesting the participation of unintegrated DNA was probably degraded by host cells. The robust- DNA proviral forms in the transcription process. ness of this assay has been confirmed by Southern Although these unintegrated proviruses could interfere blotting which seemingly reflected the same number of to the exact titer measurement, they probably exert their integrated proviruses.

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S43 The kinetic results have been confirmed by another a 30.0 laboratory which had used a similar Alu-LTR amplification design.57 In addition to this confirmation, the 25.0 authors have shown that integration of HIV-1 DNA 20.0 was detected as early at 12 h p.t. and that the integrated proviral DNAs accounted for about 5% of the initial viral 15.0 DNA synthesis. 10.0 As a result of the low frequency of Alu repeats in the 1E+02 1E+03 1E+04 1E+05 1E+06 1E+07 1E+08 human genome – approximately one Alu repeat every Vector DNACopy Number 5000 bp – one may argue that not all integrated proviral 30.0 genomes could be detected. Moreover, it is well known b that the real-time qPCR method works better with very 25.0 low DNA distances. In addition, because of the intrinsic 20.0 unidirectional design of the Alu primer, only half of Alu repeats could be detected. 15.0

10.0 Evaluation of the disappearance overtime of circle 1E+02 1E+03 1E+04 1E+05 1E+06 1E+07 1E+08 proviral genomes hALB DNACopy Number By taking advantage of the intrinsic molecular nature of Figure 1 Sensitivity of a SybrGreen I-based qPCR assay. A third the 2-LTR circles that allows the tandem orientation of generation HIV-1 transfer vector encoding the eGFP reporter protein97 LTRs, Butler et al49 have designed a qPCR assay for (a) and human albumin (hALB) (b) DNA curves were realized using serial 3 7 evaluating overtime the number of such unintegrated dilutions of plasmidic molecules (10 –10 ). The threshold cycle (Ct) corresponds to the cycle number at which the amplification has reached half proviruses (Table 1C). They have shown that the 2-LTR the threshold value. circles peaked at 24 h p.t. and declined thereafter. The production of 2-LTR circles has been slightly increased by prior incubation of transduced cells with integrase originate from prokaryotic (plasmid) or eukaryotic inhibitors, probably because blocking integration pro- 57 (mammalian cell) species. From our earlier described vides more substrate for ligation to form 2-LTR circles. SybrGreen I-based qPCR model,50 we have compared the Although it is interesting to get some information from standard plasmid DNA curve to genomic DNAs derived 2-LTR proviruses, however, general conclusions concern- from transduced clonal cells. Figure 1 first shows the ing total unintegrated circles cannot be dressed since the variation of the threshold cycle (C ) as a function of ratio of 1-LTR to 2-LTR circles has been estimated of t 49,57 57 plasmid DNA molecules that contain either the HIV-1 around 9:1. Nevertheless, Van Maele et al have transfer vector backbone (A) or a human endogenous shown, from their qPCR assay, that the 2-LTR circles albumin sequence (B). A five-log linear relationship was represented only 0.7% of the total viral DNA present and noted over the range tested (103–107 plasmid molecules), that the impact on nuclear import was evidenced by a with correlation coefficients higher than 0.995. Figure 2a three-fold increase after transduction with cPPT vectors. measures the copy number integrated in different clonal cells by using normalized values from plasmid DNA Choice of DNA standard curves curves. Values calculated from this qPCR assay agree The amount of proviral molecules within a cell and the with the copy number estimated by Southern blot amount of cell genome in the sample are determined (Figure 2b), suggesting that there is no difference in according to plasmid DNA curves. It may be argued that, amplification regarding the quality and origin of DNA, in some specific context, amplifications with genomic for example genomic versus plasmidic DNAs. DNA extracts cannot be extrapolated to those established with a plasmid DNA curve. Analysis of the transgene mRNA expression As a first example, if the work presented in chapter For future applications, it will be essential not only to Quantification of integrated proviral copies per cell, that accurately determine the transducing titers of lentiviral concerns proviral integration measurement, has been vector preparations but also to assess the transgene possible, it only resides on the fact that the authors have expression in transduced target cells. used, for the Alu–LTR curve, DNA extracts from a In such an attempt, Lizeé et al47 have taken advantage transduced bulk cell population.49,57 In fact, since retro- of their qPCR-based assay – that is, primers/probe set viral integration can occur at many location sites in the derived from the cis-active WPRE sequence that is still human genome, each provirus has a unique distance to present in the 30 mRNA extremity – to apply it for the the nearest Alu sequence and therefore generates quantitation of transgene transcripts. In a RT-qPCR amplification products of variable lengths. To be able to format, the copy number of a cellular endogenous quantitate the number of integrated proviruses per target mRNA, that is, from the human b-actin gene (cf Table cell, the DNA standard curve must represent the full 1A), was measured simultaneously as a normalization distribution of distances between LTRs and flanking Alu control. By analyzing lentiviral transgene expression at elements. the mRNA level, they have observed functional differ- In a second concern, one may argue that the quality of ences between three distinct lentiviral vectors that DNA extraction between genomic and plasmidic ele- express different kind of transgene. Messenger RNA ments is not comparable. Since some remnants from expression from neo-poly(A) polymerase cDNA has been these extracts can participate as susceptible PCR inhibi- shown to be 56-fold lower than that of eGFP transcrip- tors, they could have a different impact level as they tion, despite only a three-fold difference in proviral DNA

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S44 a 100.0 All assays already been conducted consist in the amplification of RCL by culturing vector-containing 50 ng supernatant on a permissive cell line and require the use 25 ng 10 ng of an internal replicative control. In order to increase the stringency of RCL assays, some laboratories have asso- 10.0 ciated to the standard detection of gag-pol-env replicative particles the alternative possibility to discriminate gag-pol recombinants by amplification in permissive cells that express an envelope glycoprotein, such as the one derived from the vesicular stomatitis virus (VSV/G).32 1.0 Specifically dedicated to HIV-1-derived lentiviral vectors, several methods have been designed for the follow-up of susceptible emerging tat recombinants (tat 58–60 Number of proviral copies / cell transactivation test), as well as gag recombinants (p24 enzyme-linked immunosorbent ELISA assay).61–64 These assays only detect gene products and not actually RCL; 0.1 the second assay is thought to be relatively insensitive, b HT1 HT2 HT3 and none have been tested or validated. On the model of what has previously been carried out for oncoretroviral vectors, additional methods related to any lentiviral genus have been proposed to detect RCL such as standard PCR,3,65 transgene signal follow-up,66 or marker rescue assay.64,67–69 The latter method is based on the mobilization of a proviral genome by susceptible gag-pol-env recombinants and therefore should represent the most appropriate assay by its ability to mimics all trans-active packaging functions. Moreover, this assay is very sensitive and discriminates replication-defective virus from RCL, generating a compelling need for its standardization and validation. Figure 2 Genomic characterization of individual HT1080 cell clones Although the mobilization/marker rescue assay has containing in their genome different integrated proviral copies. HT1080 been extensively employed for the agreement of retro- cells transduced at a MOI of 5 with the eGFP-containing HIV-1 vector viral vector batch release, the parallel development of were sorted according to their eGFP expression and three clones (HT1– nucleic acid-based methods such as standard PCR or HT3) were isolated by the limiting dilution technique. The number of qPCR may probably be a relevant alternative to confirm vector proviral genome sequences present per cell clone was determined by the absence of any type of recombination. In parallel to two different strategies. (a) The number of proviral copies per cell was verified by the SybrGreen I-based qPCR approach, by using standard DNA the interest to follow the decrease of the capsid p24 plasmid curves for molecule number extrapolation. (b) Southern blotting protein after successive passages of HIV-1-based vectors was realized after cutting the human genome with a restriction enzyme in permissive cells, some laboratories have also designed (BamHI) that only cuts once in the proviral genome; proviral-host DNA qPCR assays to certify the absence in DNA genomic fragments were detected by using a radioactive probe directed against the extracts of the VSV/G envelope gene (Table 2 for eGFP gene. primers/probe designs).65,70 Since the appropriate positive control cannot be available with such an ubiquitous envelope gene titer. Presumably, toxic effects resulting from overexpres- inserted in the proviral genome because of safety sion of a normally tightly regulated gene product could concern, it was therefore difficult to clearly evaluate the select for overgrowth of transductants expressing rela- robustness of these assays. However, the sensitivity of tively low levels of transgene. VSV/G detection has been estimated to be around five Very recently, the analysis of small-interfering and to 10 copies. Surprisingly, RCL assay cultures contained small-hairpin RNAs delivered from lentiviral proviruses 51 an initial level of VSV/G sequence, depending on the has also been monitored by a RT-qPCR assay. By using original vector concentration,65 as it was noticed in different primers/probe sets specific of the bcr-abl chapter Concern for plasmid DNA carryover, this was oncogene, the authors have measured the level of most likely due to the presence of residual plasmid DNA expression of these RNA polymerase III-dependent originating from the transient transfection procedure of RNAs in a cell depletion model. vector production. This initial level decreased gradually to reach the limit of detection within the 35 days of the assay. No increase in the VSV/G concentration has been Description of qPCR-based methods for observed in any of HIV-1 vector batches, indicating that others lentiviral-derived analytical assays no VSV/G propagation caused by the replication of a recombinant virus had occurred in the cultures tested. Development of RCL assays From their respective discussion, the authors claimed Lentiviral vectors have demonstrated exceptional pro- that the VSV/G qPCR end-point analysis could not mise as tools for gene therapy applications, but have also replace existing RCL assays but suggested that it should raised safety concerns because of potential creation of remain a complementary informative assay for the RCL by uncontrolled recombination. discrimination of gag-pol-VSV/G recombinants.

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S45 Table 2 List of qPCR-based assays applied to the detection of susceptible gag-pol-VSV/G RCL contaminants

References Vector and region amplified Primers and probes dedicated for the amplification of envelop DNA fragments in amplified lentiviral vector preparations

Escarpe et al70 HIV-1 (VSV/G) Forward primer (759–780)(3) CGAGATGGCTGATAAGGATCTC Reverse primer (1008–986) ATTGATTATGGTGAAAGCAGGAC Probe (783–806) F-TGCTGCAGCCAGATTCCCTGAATG-T Sastry et al65 Forward primer (322–342) TGCAAGGAAAGCATTGAACAA Reverse primer (438–420) GAGGAGTCACCTGGACAATCACT Probe (351–376) F-AGGAACTTGGCTGAATCCAGGCTTCC-T

The susceptible detection of VSV/G recombinant DNA is performed in DNA genomic extracts from transduced permissive cells at late passages. Numbers in parentheses indicate nucleotide positions in the VSV/G cDNA.(3)

Biodistribution studies lung (0–0.30%), heart (0–0.021), brain (0–0.16%), kidney The tissue biodistribution in animal models is an (0–0.003%) and gastrointestinal tract (0–0.004%). important aspect for characterizing new transfer vectors In other reports that have also based their biodistribu- and has been defined as an essential tool for analysis by tion studies upon the real-time PCR technique, intra- regulatory agencies, such as the Food and Drug Admin- venous (i.v.) injection of VSV/G-pseudotyped lentiviral istration (FDA) or the Recombinant Advisory Committee particles, whether they are derived from HIV-1,92,93 or from the National Institute of Health’s, in order to EIAV,91 have led to a better transduction signal in the approve downstream human gene therapy trials.71–73 liver and spleen. High sensitivity and reproducibility of such assays are of Real-time PCR analyses have also been performed major concern for successful biodistribution studies, after intraperitoneal (i.p.) injection of a VSV/G-pseudo- especially when very low gene transfer is anticipated. typed HIV-1 vector preparation in immunocompromised To satisfy such requirements, PCR-related DNA analysis mice (SCID),94 showing that the vector sequence was also has been specified by the FDA as an adequate method for detected predominantly in the liver (0.1–1%) and the the transduction survey in animals.72,74 spleen (0.8–2%). Although gene transfer and transgene expression of A recent model for in vivo cell targeting of HIV-1 lentiviral vectors have been demonstrated by several vectors pseudotyped with a modified chimeric Sindbis laboratories in various organs,58,75–78 the biodistribution virus envelope glycoprotein (named m168) has also been and systemic effects have not yet been fully addressed. assessed with a qPCR monitoring.90 By using this type of Nevertheless, different groups have already initiated pseudotyping, the authors have shown, after i.v. vector that kind of expertise, where the common measurement delivery in a mouse melanoma cancer model, a substan- analyses for determining which organ is targeted are tially lower infectivity in the liver and spleen, with based either on reporter protein expression profiles,79–81 successfully targeted metastatic melanoma cells growing or on the detection of proviral DNA elements by using in the lungs. standard PCR,80 fluorescence in situ hybridization,82 or Fleury et al85 have evaluated the ability of VSV/G- qPCR amplification method.83–93 pseudotyped HIV-1 vectors to transduce the myocar- As equivalent to the in vitro strategy (cf Measurement dium of rats. At day 3 after intramyocardial injection, of proviral DNA genomes within transduced target the qPCR analysis performed on genomic DNAs showed cells), most of these qPCR-based assays measured the around 10% of cardiac cells being transduced, with half number of proviral copies per target cell by normalizing of them expressing the eGFP reporter protein. Proviral it to the number of endogenous gene amplicons. In Table DNAs declined thereafter, upon day 14 and week 10 3 are summarized primers and probes dedicated for analyses, by three- to five-fold. In parallel, systemic parallel or duplex amplifications of lentiviral DNA vector dissemination was also evaluated; small amounts proviruses and fragment DNAs derived from animal of HIV-1 proviruses have only been detected in the liver genomes. (0.0003%) and spleen (0.007%), probably attributable, Tail-vein injection of VSV/G-pseudotyped HIV-1 according to the authors, to inadvertent vector injection vectors in BALB/c mice has resulted, at 40 days post- into the cardiac circulation or to susceptible migration of injection (p.i.), to a very high gene transfer in the bone cardiac leukocytes to remote organs. marrow, ranging from 0.21 to 22.7% of transgene Biodistribution studies have also been attempted in frequency determined by qPCR amplification.83 This BALB/c fetuses by intramuscular and intrahepatic (i.h.) result was consistent with the observation that high injections of HIV-1 vectors pseudotyped by different eGFP protein levels (410%) were detected in peripheral glycoproteins.84 From the results gained by their non- blood leukocytes from these animals. Additional organs normalized qPCR assay, the authors have shown that the have also been transduced but to a lesser extent: liver efficiency and distribution of transduction was highly (0.26–1.3%), spleen (0.045–0.38%), bladder (0–0.85%), dependent upon the route of administration and the

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S46 Table 3 List of qPCR-based assays applied to lentiviral vector biodistribution studies in animal models

References Vector and Primers and probes dedicated for the amplification Primers and probes dedicated for the amplification region amplified of lentiviral or transgenic DNA sequences of internal control genes (rodent cellular genome)

Pan et al83 HIV-1 (eGFP) Forward primer (434–460) Normalization with the ampliﬁcation of ACTACAACAGCCACAACGTCTATATCA a mouse apolipoprotein B DNA fragment Reverse primer (508–489) Forward primer GGCGGATCTTGAAGTTCACC CGTGGGCTCCAGCATTCTA Probe (464–487) Reverse primer F-CCGACAAGCAGAAGAACGGCATCA-T TCACCAGTCATTTCTGCCTTTG Probe V-CCTTGAGCAGTGCCCGACCATTC-T

Indraccolo et al94 SIV and MLV Forward primer (457–480) Normalization with the ampliﬁcation of (eGFP) ATCATGGCCGACAAGCAGAAGAAC a mouse 18S ribosomal DNA fragment Reverse primer (714–691) Forward primer GTACAGCTCGTCCATGCCGAGAGT CCATCGAACGTCTGCCCTA Probe (657–642) Reverse primer V-CAGGACCATGTGATCGCGCTTCTCGT-T TCACCCGTCGTCACCATG Probe F-CGATGGTGGTCGCCGTGCCTA-T

MacKenzie et al84 HIV-1 (U3) Forward primer (201–220) Absence of normalization TGAAAGCGAAAGGGAAACCA Reverse primer (266–252) CCGTGCGCGCTTCAG Probe (225–248) F-AGCTCTCTCGACGCAGGACTCGGC-T

Fleury et al85 Forward primer (11–33) Normalization with the ampliﬁcation of a rat GTTAGACCAGATCTGAGCCTGGG b-actin DNA fragment Reverse primer (96–119) Forward primer CAACAGACGGGCACACACTACTTG GATACTCCCAGCACACTTAACTTAGC Reverse primer CAAAGTCCTAAAGCCACAAGAAACAC

Yamada et al86 HIV-1 (R-U5) Cf Scherr et al44 Normalization with the ampliﬁcation of the neomycin-resistance gene inserted during the knock-out mouse model Forward primer (394–411) CCATTCGACCACCAAGCG Reverse primer (452–438) CCGGCTCTCCATCCGA Probe (413–434) Te-AACATCGCATCGAGCGAGCACG-T

Kurre et al87 HIV-1 (eGFP) Forward primer (144–158) Normalization with the ampliﬁcation of a mouse CTGCACCACCGGCAA IL3 DNA fragment Reverse primer (225–208) Forward primer (23–41)(4) GTAGCGGCTGAAGCACTG CCAGCATCCACACCATGCT Probe (188–207) Reverse primer (96–78) F-CCACCCTGACCTACGGCGTG-T CCGGCCACTGATTGAAGCT Probe (49–75) F-CTCCTGATGCTCTTCCACCTGGGACTC-T

Hanawa et al88 SIV (R-U5) Forward primer (363–385) Rhesus macaque ribosomal RNA control TACGGCTGAGTGAAGGCAGTAAG ampliﬁcation for normalization (unspeciﬁed Reverse primer (454–436) sequences) CTCCTCACGCCGTCTGGTA Probe (392–416) F-AGGAACCAACCACGACGGAGTGCTC-T

Jimenez et al89 HIV-1 (eGFP) Forward primer (457–480) Normalization with the ampliﬁcation of a rhesus ATCATGGCCGACAAGCAGAAGAAC macaque e-globin DNA fragment Reverse primer (714–691) Forward primer GTACAGCTCGTCCATGCCGAGAGT TGGCAAGGAGTTCACCCCT Probe (663–638) Reverse primer F-ATCATGGCCGACAAGCAGAAGAAC-T AATGGCGACAGCAGACACC Probe F-TGCAGGCTGCCTGGCAGAAGC-T

Morizono et al90 HIV-1 (firefly Forward primer (93–112) Normalization with the amplification of a mouse luciferase) GAGATACGCCCTGGTTCCTG b-actin DNA fragment

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S47 Table 3 Continued

Reverse primer (237–219) Forward primer (837–858) GCATACGACGATTCTGTGATTTG CAACTCCATCATGAAGTGTGAC Reverse primer (1020–1000) CCACACGGAGTACTTGCGCTC

Lamikanra et al91 EIAV (Psi) Forward primer (478–499) Normalization with the ampliﬁcation of a mouse ATTGGGAGACCCTTTGACATTG glyceraldehyde-3-phosphate dehydrogenase Reverse primer (562–534) (GAPDH) DNA fragment ACCAGTAGTTAATTTCTGAGACCCTTGTA Forward primer (21–38) Probe (529–500) CGGATTTGGCCGTATTGG F-CACCTTCTCTAACTTCTTGAGCGCCTTGCT-T Reverse primer (91–71) TGGCAACAATCTCCACTTTGC Probe (42–63) F-CCTGGTCACCAGGGCTGCCATT-T

Forward primer (655–674) Normalization with the ampliﬁcation of a mouse TGAAAGCGAAAGGGAAACCA b-actin DNA fragment Reverse primer (720–706) Forward primer (614–634) Di Domenico et al92 HIV-1CCGTGCGCGCTTCAG AGAGGGAAATCGTGCGTGAC Di Natale et al93 (Psi)Probe (679–702) Reverse primer (751–733) V-AGCTCTCTCGACGCAGGACTCGGC-T CAATAGTGACCTGGCCGT Probe (684–706) V-CACTGCCGCATCCTCTTCCTCCC-T

In the last row are listed primer and probe sequences derived from animal genomes that allow the percentage estimation of transduced cells in different types of organs. Numbers in parentheses indicate nucleotide positions in different animal cDNAs.(4) pseudotype of vector used. A widespread distribution of detectable levels.83,84 Nevertheless, Jimenez et al89 have vector sequences has been observed in multiple tissues observed from i.v. vector delivery some significant gene but mainly in the liver, heart and muscle. Overall, scoring in male (0.0034%) and female (1.5%) gonads of hepatocytes were better transduced with the VSV/G macaque fetuses. pseudotype, whereas myocytes were preferably targeted Ex vivo applications have also been monitored by the with Mokola or Ebola glycoproteins. real-time PCR assay in order to know which differen- In another fetal model (rhesus macaque), however, no tiated cells are scored after engraftment and in how significant differences have been found in the level of many copies per colony. Autologous CD34+ cells eGFP gene marking when comparing i.p. and i.h. routes transduced with SIV-derived lentiviral vector prepara- of HIV-1-derived lentiviral vector administration.89 At tions have been engrafted in three rhesus macaques.88 7 months after vector delivery, the percentage of gene Here again, the authors have observed a clear correlation transduction was higher in abdominal organs (omentum between gene marking (qPCR) and gene expression (1.7%), diaphragm (1.4%) and peritoneum (0.6%)) as (FACS) from different reconstituted cell lineages. In a compared to nonabdominal organs (thorax (0.8%), mouse model for Fanconi anemia disease, Yamada et al86 trachea, esophagus, lung and pericardium (0.05%), have shown that one proviral copy per cell was detected cerebrum, heart, thymus and bone marrow (o0.005%)). in all lineage-committed cells in the peripheral blood of Interestingly, the authors have also designed a RT-qPCR both primary and secondary recipients, allowing the assay, as already described above for in vitro transcript phenotypic correction by ex vivo lentiviral-mediated analysis (cf Analysis of the transgene mRNA expression), transgenesis. In the concern of reducing the number of and have shown a positive correlation between gene integrated proviral copies in repopulated cells, Kurre (qPCR) and mRNA (RT-qPCR) biodistribution. Since et al87 have identified conditions for lentiviral gene differences may exist in the level of gene transfer and transfer involving minimal ex vivo target cell manipula- expression from two different specimens from the same tion. From the results gained by qPCR monitoring, they tissue, the authors have chosen to determine the amount have shown that the initial multiplicity of infection was of eGFP transgene and mRNA transcripts present in the a critical determinant of proviral copy number in same specimen. transduced murine long-term repopulating cells. One of the most important concerns for in vivo gene therapy in humans is the possibility of germline integration, which might result in the introduction of Conclusion heritable genetic changes into the offspring of treated patients. Insertional mutations that lead to devastating Many laboratories, which are implicated in the lentiviral consequences have already been reported in the germ vectorology, have settled several qPCR-based analytical line of transgenic mice produced by pronuclear micro- methods in order to measure and/or standardize the injection.95,96 From the distribution reports that have vector titration and its biodistribution in the organism. been discussed above, vector sequences in gonad organs In addition, the qPCR assay has also been dedicated for were either undetectable,85 or rarely scored as barely the follow-up of susceptible emerging recombinants;

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S48 although additional techniques such as the marker 14 Loewen N et al. Long-term, targeted genetic modification of the rescue assay are better suited for RCL detection, the aqueous humor outflow tract coupled with noninvasive imaging qPCR-related approach could be used in parallel to of gene expression in vivo. Invest Ophthalmol Vis Sci 2004; 45: confirm functional results and also to evaluate the 3091–3098. genetic structure of recombinants. 15 Doi K et al. Lentiviral transduction of green fluorescent protein All assays already being conducted to date in the in retinal epithelium: evidence of rejection. Vision Res 2002; 42: lentiviral vector design have shown that the qPCR 551–558. strategy is a robust and solid application with at least 16 Roy-Burman P et al. Assay for type C virus in mouse sera based five-log over-range of magnitude. Most of the qPCR- on particulate reverse transcriptase activity. J Virol 1976; 19: based titration expertises have suggested a positive 1107–1110. correlation with titers determined by the eGFP expres- 17 Tonini T, Claudio PP, Giordano A, Romano G. Retroviral and lentiviral vector titration by the analysis of the activity of viral sion. However, titers derived from the measurement of reverse transcriptase. Methods Mol Biol 2004; 285: 155–157. the viral nucleic acid species (RNA genome and strong- 18 Heneine W et al. Detection of reverse transcriptase by a stop cDNA) in vector particles could only determine the highly sensitive assay in sera from persons infected with number of pp contained in each lentiviral vector batch. human immunodeficiency virus type 1. J Infect Dis 1995; 171: By contrast, DNA proviral-based titration assays seem 1210–1216. to be more consistent for the exact quantitation of the 19 Eberle J, Seibl R. A new method for measuring reverse number of transducing-competent particles (ip). transcriptase activity by ELISA. J Virol Methods 1992; 40: 347–356. The overall disparity observed between the different 20 Silver J, Maudru T, Fujita K, Repaske R. An RT-PCR assay for the groups and assays suggest that laboratories should take enzyme activity of reverse transcriptase capable of detecting into account the necessity to harmonize and standardize single virions. Nucl Acids Res 1993; 21: 3593–3594. the lentiviral-based control testing in order to set up a 21 Pyra H, Boni J, Schupbach J. Ultrasensitive retrovirus detection liable and robust expertise from the next clinical trials by a reverse transcriptase assay based on product enhancement. that will be conducted in a near future. From discussions Proc Natl Acad Sci USA 1994; 91: 1544–1548. and exchanges, scientific experts in this field should 22 Busso M, Resnick L. Development of an assay that detects gather their effort to decide which technique will be the transcriptionally competent human immunodeficiency virus gold standard. type one particles. J Virol Methods 1994; 47: 129–139. 23 Sano K et al. Comparable sensitivities for detection of HIV-1 reverse transcriptase (RT) and other polymerases by RT assays requiring no radioisotopic materials. J Virol Methods 1995; 53: References 235–244. 24 Arnold BA, Hepler RW, Keller PM. One-step fluorescent probe 1 Zufferey R et al. Multiply attenuated lentiviral vector achieves product-enhanced reverse transcriptase assay. Biotechniques efficient gene delivery in vivo. Nat Biotechnol 1997; 15: 871–875. 1998; 25: 98–106. 2 Miyoshi H et al. Transduction of human CD34+ cells that 25 Lovatt A et al. High throughput detection of retrovirus- mediate long-term engraftment of NOD/SCID mice by HIV associated reverse transcriptase using an improved fluorescent vectors. Science 1999; 283: 682–686. product enhanced reverse transcriptase assay and its compar- 3 Metharom P et al. Novel bovine lentiviral vectors based on ison to conventional detection methods. J Virol Methods 1999; 82: Jembrana disease virus. J Gene Med 2000; 2: 176–185. 185–200. 4AnDSet al. Efficient lentiviral vectors for short hairpin RNA 26 Sears JF, Khan AS. Single-tube fluorescent product-enhanced delivery into human cells. Hum Gene Ther 2003; 14: 1207–1212. reverse transcriptase assay with Ampliwax (STF-PERT) for 5 De A, Lewis XZ, Gambhir SS. Noninvasive imaging of lentiviral- retrovirus quantitation. J Virol Methods 2003; 108: 139–142. mediated reporter gene expression in living mice. Mol Ther 2003; 27 Sears JF, Repaske R, Khan AS. Improved Mg2+-based reverse 7: 681–691. transcriptase assay for detection of primate retroviruses. J Clin 6 Ray P et al. Imaging tri-fusion multimodality reporter gene Microbiol 1999; 37: 1704–1708. expression in living subjects. Cancer Res 2004; 64: 1323–1330. 28 Andre M, Morgeaux S, Fuchs F. Quantitative detection of RT 7 D’Costa J et al. HIV-2 derived lentiviral vectors: gene transfer in activity by PERT assay: feasibility and limits to a standardized Parkinson’s and Fabry disease models in vitro. J Med Virol 2003; screening assay for human vaccines. Biologicals 2000; 28: 67–80. 71: 173–182. 29 Schupbach J. Induction/activation and detection of occult viral 8 Sirin O, Park F. Regulating gene expression using self-inactivat- agents present in mammalian cells. Dev Biol (Basel) 2001; 106: ing lentiviral vectors containing the mifepristone-inducible 425–437; discussion 465–475. system. Gene 2003; 323: 67–77. 30 Kuate S, Wagner R, Uberla K. Development and characterization 9 Tonini T, Claudio PP, Giordano A, Romano G. Determination of of a minimal inducible packaging cell line for simian immuno- functional viral titer by drug-resistance colony assay, expression deficiency virus-based lentiviral vectors. J Gene Med 2002; 4: of green fluorescent protein, and beta-galactoside staining. 347–355. Methods Mol Biol 2004; 285: 149–153. 31 Martin-Rendon E et al. New methods to titrate EIAV-based 10 Grutzkau A et al. Detection of intracellular interleukin-8 in lentiviral vectors. Mol Ther 2002; 5: 566–570. human mast cells: flow cytometry as a guide for immunoelec- 32 Farson D et al. A new-generation stable inducible packaging cell tron microscopy. J Histochem Cytochem 1997; 45: 935–945. line for lentiviral vectors. Hum Gene Ther 2001; 12: 981–997. 11 Ghosh C, Iversen PL. Intracellular delivery strategies for 33 Kirkwood TB, Bangham CR. Cycles, chaos, and evolution in antisense phosphorodiamidate morpholino oligomers. Antisense virus cultures: a model of defective interfering particles. Proc Nucleic Acid Drug Dev 2000; 10: 263–274. Natl Acad Sci USA 1994; 91: 8685–8689. 12 Anson DS, Limberis M. An improved beta-galactosidase 34 Higashikawa F, Chang L. Kinetic analyses of stability of simple reporter gene. J Biotechnol 2004; 108: 17–30. and complex retroviral vectors. Virology 2001; 280: 124–131. 13 Duisit G et al. Five recombinant simian immunodeficiency virus 35 Murdoch B et al. A rapid screening procedure for the pseudotypes lead to exclusive transduction of retinal pigmented identification of high-titer retrovirus packaging clones. Gene epithelium in rat. Mol Ther 2002; 6: 446–454. Therapy 1997; 4: 744–749.

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S49 36 Onodera M et al. A simple and reliable method for screening deficiency virus type 1 vector transduction. J Virol 2003; 77: retroviral producer clones without selectable markers. Hum Gene 4685–4694. Ther 1997; 8: 1189–1194. 58 Naldini L et al. In vivo gene delivery and stable transduction of 37 Forghani B, Hurst JW, Shell GR. Detection of the human nondividing cells by a lentiviral vector (see comments). Science immunodeficiency virus genome with a biotinylated DNA 1996; 272: 263–267. probe generated by polymerase chain reaction. Mol Cell Probes 59 Gallichan WS et al. Lentivirus-mediated transduction of 1991; 5: 221–228. islet grafts with interleukin 4 results in sustained gene 38 Woo JC et al. Investigation of recombinant human insulin-like expression and protection from insulitis. Hum Gene Ther 1998; growth factor type I in thymus regeneration in the acute stage of 9: 2717–2726. experimental FIV infection in juvenile cats. AIDS Res Hum 60 Kafri T et al. A packaging cell line for lentivirus vectors. J Virol Retroviruses 1999; 15: 1377–1388. 1999; 73: 576–584. 39 Watson A et al. Plasma viremia in macaques infected with simian 61 Dull T et al. A third-generation lentivirus vector with a immunodeficiency virus: plasma viral load early in infection conditional packaging system. J Virol 1998; 72: 8463–8471. predicts survival. J Virol 1997; 71: 284–290. 62 Gasmi M et al. Requirements for efficient production and 40 Dyer JR et al. Comparison of NucliSens and Roche transduction of human immunodeficiency virus type 1-based Monitor assays for quantitation of levels of human immuno- vectors. J Virol 1999; 73: 1828–1834. deficiency virus type 1 RNA in plasma. J Clin Microbiol 1999; 37: 63 Mochizuki H et al. High-titer human immunodeficiency virus 447–449. type 1-based vector systems for gene delivery into nondividing 41 Cao Y et al. Clinical evaluation of branched DNA signal cells. J Virol 1998; 72: 8873–8883. amplification for quantifying HIV type 1 in human plasma. 64 Segall HI, Yoo E, Sutton RE. Characterization and detection of AIDS Res Hum Retroviruses 1995; 11: 353–361. artificial replication-competent lentivirus of altered host range. 42 Elbeik T et al. Comparative analysis of HIV-1 viral load assays on Mol Ther 2003; 8: 118–129. subtype quantification: Bayer Versant HIV-1 RNA 3.0 versus 65 Sastry L et al. Certification assays for HIV-1-based vectors: Roche Amplicor HIV-1 Monitor version 1.5. J Acquir Immun Defic frequent passage of gag sequences without evidence of replica- Syndr 2002; 29: 330–339. tion-competent viruses. Mol Ther 2003; 8: 830–839. 43 Linnen JM et al. Sensitive detection of genetic variants of HIV-1 66 Waddington SN et al. Long-term transgene expression and HCV with an HIV-1/HCV assay based on transcription- by administration of a lentivirus-based vector to the fetal mediated amplification. J Virol Methods 2002; 102: 139–155. circulation of immuno-competent mice. Gene Therapy 2003; 10: 44 Scherr M et al. Quantitative determination of lentiviral vector 1234–1240. particle numbers by real-time PCR. Biotechniques 2001; 31: 520, 67 Srinivasakumar N, Schuening FG. A lentivirus packaging 522, 524, passim. system based on alternative RNA transport mechanisms to 45 Sastry L et al. Titering lentiviral vectors: comparison of DNA, express helper and gene transfer vector RNAs and its use to RNA and marker expression methods. Gene Therapy 2002; 9: study the requirement of accessory proteins for particle 1155–1162. formation and gene delivery. J Virol 1999; 73: 9589–9598. 46 Ikeda Y et al. Gene transduction efficiency in cells of different 68 Wu X et al. Development of a novel trans-lentiviral vector that species by HIV and EIAV vectors. Gene Therapy 2002; 9: 932–938. affords predictable safety. Mol Ther 2000; 2: 47–55. 47 LizeéGet al. Real-time quantitative reverse transcriptase- 69 Mautino MR, Ramsey WJ, Reiser J, Morgan RA. Modified polymerase chain reaction as a method for determining human immunodeficiency virus-based lentiviral vectors display lentiviral vector titers and measuring transgene expression. decreased sensitivity to trans-dominant Rev. Hum Gene Ther Hum Gene Ther 2003; 14: 497–507. 2000; 11: 895–908. 48 Towers GJ et al. One step screening of retroviral producer clones 70 Escarpe P et al. Development of a sensitive assay for detection by real time quantitative PCR. J Gene Med 1999; 1: 352–359. of replication-competent recombinant lentivirus in large-scale 49 Butler SL, Hansen MS, Bushman FD. A quantitative assay for HIV-based vector preparations. Mol Ther 2003; 8: 332–341. HIV DNA integration in vivo. Nat Med 2001; 7: 631–634. 71 Administration FaD. Guidance for industry: guidance for 50 Charrier S et al. A lentiviral vector encoding the human Wiskott– human somatic cell therapy and gene therapy. http://www.fda. Aldrich syndrome protein corrects immune and cytoskeletal gov/cber/gdlns/somgene.pdf. 1991. defects in WASP knockout mice. Gene Therapy 2005; 12: 597–606. 72 Pilaro AM, Serabian MA. Preclinical development strategies for 51 Scherr M et al. Stable RNA interference (RNAi) as an option for novel gene therapeutic products. Toxicol Pathol 1999; 27: 4–7. anti-bcr-abl therapy. Gene Therapy 2005; 12: 12–21. 73 Administration FaD. Points to consider in human somatic cell 52 Zhang B et al. The significance of controlled conditions in therapy and gene therapy. http://www.fda.gov/cber/gdlns/ lentiviral vector titration and in the use of multiplicity of ptcsomat.pdf 1991. infection (MOI) for predicting gene transfer events. Genet 74 Gonin P, Gaillard C. Gene transfer vector biodistribution: pivotal Vaccines Ther 2004; 2:6. safety studies in clinical gene therapy development. Gene 53 Sastry L et al. Evaluation of plasmid DNA removal from Therapy 2004; 11 (Suppl 1): S98–S108. lentiviral vectors by benzonase treatment. Hum Gene Ther 2004; 75 Kafri T et al. Sustained expression of genes delivered directly 15: 221–226. into liver and muscle by lentiviral vectors. Nat Genet 1997; 17: 54 Trono D. Partial reverse transcripts in virions from human 314–317. immunodeficiency and murine leukemia viruses. J Virol 1992; 66: 76 Johnson LG, Olsen JC, Naldini L, Boucher RC. Pseudotyped 4893–4900. human lentiviral vector-mediated gene transfer to airway 55 Klarmann GJ, Schauber CA, Preston BD. Template-directed epithelia in vivo. Gene Therapy 2000; 7: 568–574. pausing of DNA synthesis by HIV-1 reverse transcriptase during 77 Miyoshi H et al. Development of a self-inactivating lentivirus polymerization of HIV-1 sequences in vitro. J Biol Chem 1993; 268: vector. J Virol 1998; 72: 8150–8157. 9793–9802. 78 Woods NB et al. Lentiviral gene transfer into primary and 56 Saenz DT et al. Unintegrated lentivirus DNA persistence and secondary NOD/SCID repopulating cells (in process citation). accessibility to expression in nondividing cells: analysis with Blood 2000; 96: 3725–3733. class I integrase mutants. J Virol 2004; 78: 2906–2920. 79 Ohashi K, Park F, Kay MA. Role of hepatocyte direct hyperplasia 57 Van Maele B, De Rijck J, De Clercq E, Debyser Z. Impact of the in lentivirus-mediated liver transduction in vivo. Hum Gene Ther central polypurine tract on the kinetics of human immuno- 2002; 13: 653–663.

Gene Therapy qPCR applied to lentiviral analytical assays C Delenda and C Gaillard S50 80 Peng KW et al. Organ distribution of gene expression after virus-based lentiviral vector system. Blood 2004; 103: intravenous infusion of targeted and untargeted lentiviral 4062–4069. Epub 2004 Feb 4019. vectors. Gene Therapy 2001; 8: 1456–1463. 89 Jimenez DF et al. HIV-1-derived lentiviral vectors and fetal route 81 VandenDriessche T et al. Lentiviral vectors containing the human of administration on transgene biodistribution and expression in immunodeficiency virus type-1 central polypurine tract can rhesus monkeys. Gene Therapy 2005; 7:7. efficiently transduce nondividing hepatocytes and antigen- 90 Morizono K et al. Lentiviral vector retargeting to P-glycoprotein presenting cells in vivo. Blood 2002; 100: 813–822. on metastatic melanoma through intravenous injection. Nat Med 82 Vargas Jr J et al. Novel integrase-defective lentiviral episomal 2005; 11: 346–352. Epub 2005 Feb 2013. vectors for gene transfer. Hum Gene Ther 2004; 15: 361–372. 91 Lamikanra A et al. In vivo evaluation of an EIAV vector for the 83 Pan D et al. Biodistribution and toxicity studies of VSVG- systemic genetic delivery of therapeutic antibodies. Gene Therapy pseudotyped lentiviral vector after intravenous administration 2005; 17: 17. in mice with the observation of in vivo transduction of bone 92 Di Domenico C et al. Gene therapy for a mucopolysaccharidosis marrow. Mol Ther 2002; 6: 19–29. type I murine model with lentiviral-IDUA vector. Hum Gene Ther 84 MacKenzie TC et al. Efficient transduction of liver and muscle 2005; 16: 81–90. after in utero injection of lentiviral vectors with different 93 Di Natale P et al. Treatment of the mouse model of mucopoly- pseudotypes. Mol Ther 2002; 6: 349–358. saccharideosis type IIIB with lentiviral-naglu vector. Biochem J 85 Fleury S et al. Multiply attenuated, self-inactivating lentiviral 2005; 13: 13. vectors efficiently deliver and express genes for extended 94 Indraccolo S et al. Gene transfer in ovarian cancer cells: a periods of time in adult rat cardiomyocytes in vivo. Circulation comparison between retroviral and lentiviral vectors. Cancer Res 2003; 107: 2375–2382. Epub 2003 Apr 2314. 2002; 62: 6099–6107. 86 Yamada K et al. Phenotype correction of Fanconi anemia group 95 Woychik RP et al. An inherited limb deformity created by A hematopoietic stem cells using lentiviral vector. Mol Ther 2003; insertional mutagenesis in a transgenic mouse. Nature 1985; 318: 8: 600–610. 36–40. 87 Kurre P et al. Efficient marking of murine long-term repopulat- 96 McNeish JD, Scott Jr WJ, Potter SS. Legless, a novel mutation ing stem cells targeting unseparated marrow cells at low found in PHT1-1 transgenic mice. Science 1988; 241: 837–839. lentiviral vector particle concentration. Mol Ther 2004; 9: 914–922. 97 Follenzi A et al. Gene transfer by lentiviral vectors is limited 88 Hanawa H et al. Efficient gene transfer into rhesus repopulating by nuclear translocation and rescued by HIV-1 pol sequences. hematopoietic stem cells using a simian immunodeficiency Nat Genet 2000; 25: 217–222.

Gene Therapy