Skip to main content
NCBI home page
Molecular Therapy logoLink to Molecular Therapy
. 2012 May 22;20(9):1699–1712. doi: 10.1038/mt.2012.96

Lentiviral Vectors Displaying Modified Measles Virus gp Overcome Pre-existing Immunity in In Vivo-like Transduction of Human T and B Cells

Camille Lévy 1,2,3, Fouzia Amirache 1,2,3, Caroline Costa 1,2,3, Cecilia Frecha 1,2,3, Claude P Muller 4, Hasan Kweder 5, Robin Buckland 2,3,5, François-Loïc Cosset 1,2,3,*, Els Verhoeyen 1,2,3,*
PMCID: PMC3437580  PMID: 22617109

Abstract

Gene transfer into quiescent T and B cells is important for gene therapy and immunotherapy approaches. Previously, we generated lentiviral vectors (LVs) pseudotyped with Edmonston (Ed) measles virus (MV) hemagglutinin (H) and fusion (F) glycoproteins (H/F-LVs), which allowed efficient transduction of quiescent human T and B cells. However, a major obstacle in the use of H/F-LVs in vivo is that most of the human population is vaccinated against measles. As the MV humoral immune response is exclusively directed against the H protein of MV, we mutated the two dominant epitopes in H, Noose, and NE. LVs pseudotyped with these mutant H-glycoproteins escaped inactivation by monoclonal antibodies (mAbs) but were still neutralized by human serum. Consequently, we took advantage of newly emerged MV-D genotypes that were less sensitive to MV vaccination due to a different glycosylation pattern. The mutation responsible was introduced into the H/F-LVs, already mutated for Noose and NE epitopes. We found that these mutant H/F-LVs could efficiently transduce quiescent lymphocytes in the presence of high concentrations of MV antibody-positive human serum. Finally, upon incubation with total blood, mimicking the in vivo situation, the mutant H/F-LVs escaped MV antibody neutralization, where the original H/F-LVs failed. Thus, these novel H/F-LVs offer perspectives for in vivo lymphocyte-based gene therapy and immunotherapy.

Introduction

Efficient gene transfer into quiescent T and B lymphocytes for gene therapy or immunotherapy purposes may not only allow the treatment of several genetic dysfunctions of the hematopoietic system, such as immunodeficiencies, but also the development of novel therapeutic strategies for cancers and acquired diseases.1 Until now, most clinical trials based on genetic modification of T cells have used VSV-G-LVs, a lentiviral (LV) pseudotype, which demands extended ex vivo culture and T cell receptor activation or stimulation with T-cell survival cytokines to allow their efficient transduction.2,3,4,5,6 For B cells a complex coculture with stroma cells in the presence of a cytokine cocktail is required to allow efficient VSV-G-LV transduction.7,8 For both B and T cells this kind of manipulation may change the phenotype of the cells.1 Moreover, VSV-G-LVs are not applicable in vivo since they are inactivated by the human complement9 and the majority of T cells in the body are resting cells which are not efficiently transduced by classical VSV-G-LVs, unless they enter the G1b phase of the cell cycle.3,4,6 We previously engineered LVs carrying Edmonston (Ed) hemagglutinin (H) and fusion (F) gp at their surface (H/F-LVs), which conserved the original MV-Ed tropism through CD46 and SLAM receptors.10 Most importantly, they represent the first tool to allow efficient transduction of quiescent human T cells and both healthy and cancerous B cells without inducing entry into the cell cycle or changes in phenotype.11,12 Of importance, we found that efficient quiescent lymphocyte transduction only occurs when CD46 and SLAM are correctly engaged by these H/F-LVs which triggers an entry mechanism that strongly resembles macropinocytosis.13

Thus, H/F-LVs represent for the first time a potential tool for efficient in vivo transduction of T and B lymphocytes since the majority of these target cell are quiescent in vivo. However, a major obstacle in the use of H/F-LVs in vivo for transduction of these cells is that most of the human population is vaccinated against measles virus (MV). Current live attenuated vaccines induce a vigorous and long-lasting immune response that protects against MV reinfection.14 Neutralizing activity of antibodies is highlighted by the fact that newborns and infants are protected by maternal antibodies against MV infection.15 Indeed, H/F-LVs systemic delivery directly exposes the therapeutic vector to these pre-existing neutralizing antibodies, which will probably degrade the vector before it can transduce the target T or B cells. Surprisingly, the human humoral immune response appears to be almost exclusively directed against the H protein of MV with anti-MV-F antibodies having little effect.16 Although most of the surface of a protein is antigenic, the antibody response against MV-H is biased toward a limited number of immunodominant epitopes.17 The major B cell epitope on the MV-H protein localizes to the region between amino acids 379 and 410 on the globular head. This region, conserved between the Morbillivirus attachment proteins, has been called the “noose” (HNE) epitope.18 The HNE domain contains three cysteine residues of which two form a surface-exposed loop.19 In addition, a secondary epitope (NE) has been identified on the MV-H globular head at residues 245–250.20 Structural analysis of the MV H gp, revealed that both Noose and NE epitopes are well exposed and not adjacent to SLAM and CD46 receptor binding sites in H. Human antibodies appear to react more weakly with the H regions that bind CD46 receptor and bind more strongly around the putative SLAM-binding site.17

It was reported that MVs with mutations in the noose region could be selected by anti-MV-H antibodies during MV infection and as a result persist – as in cases of subacute sclerosing panencephalitis (SSPE; ref. 21). Inspection of nine MV-H sequences from UK SSPE cases20 revealed that five possessed mutations in Noose epitope residues and another had a mutation in the NE epitope. Selection of a mutant MV that escapes neutralization could possibly explain the elevated levels of (ineffective) anti-MV antibodies in SSPE cases. Importantly, although routine measles vaccine coverage is high (in some countries >95%) several outbreaks of measles have been reported worldwide. Some of these emerging MV strains carry mutations in the H epitopes. Indeed, two H1 strains circulating endemically in Asia showed an exchange of proline 397 to leucine 397 which leads to a loss of the neutralizing Noose epitope in H.22,23 Moreover, several outbreaks of a new MV-D genotype (D7) were reported in Germany which were characterized by the mutation D416N in H-D7 gp, resulting in an additional N-linked glycosylation.24 The same glycosylation site has also been found in recently emerged D4 and D11 strains.25,26,27

In summary, pre-existing antibodies are likely to neutralize MV H and F gp displaying LVs in vivo. Thus, for in vivo use, we need to arm these vectors against the effect of MV Hgp-specific antibodies. Based on SSPE cases and on the appearance of new MV genotypes, we identified residues in the principal B cell epitopes of the MV-H gp whose mutation allowed escape from the humoral immune response without compromising HF-LV-lymphocyte transduction. Here we have engineered lentiviral vectors (LVs) expressing mutant H gps that can escape neutralization from blood of individuals immunized against MV while still allowing an efficient transduction of quiescent T and B cells.

Results

MV gp pseudotyped LVs (H/F-LVs) are neutralized by pre-existing antibodies in human serum

Since H/F-LVs confer efficient transduction of quiescent T and B cells, their in vivo administration could allow therapeutic gene transfer of these targets, which are for the majority in a quiescent state. However, the high prevalence of pre-existing antibodies due to MV vaccination14 could largely reduce gene transfer efficacy upon systemic administration.

Therefore, we first evaluated the neutralizing effect on H/F-LVs of seven different sera from donors who had been vaccinated against measles. Human sera were heat-treated to inactivate human complement which allowed to assess solely the effect of anti-MV pre-existing antibodies on H/F-LV transduction performance. The H/F-LVs were incubated with three dilutions of each serum (Figure 1a) and subsequently incubated with target 293T cells to evaluate their transduction efficiency. As a control H/F-LVs were incubated with the target cells in the absence of human serum. Neutralization was calculated as the transduction efficiency in the presence of serum relative to the transduction efficiency in the absence of serum.

Figure 1.

Figure 1

Open in a new tab

Measles virus (MV)-derived lentiviral vectors (LVs) are neutralized by MV-specific antibodies present in human and macaque serum but resistant to macaque complement. (a) 293T (CD46+) cells were transduced with H/F-LVs at a multiplicity of infection (MOI) of 0.2 after preincubation of the lentivectors with serial dilutions of human serum from seven different donors. (b) 293T cells were transduced with H/F-LVs or VSV-G-LVs at a MOI of 0.2 upon preincubation with heat-treated complement-inactivated human sera or purified IgGs from the sera 3 and 7; sera were applied at 100-fold dilutions. (c) 293T cells were transduced with the indicated LVs upon incubation with heat-treated complement-inactivated sera from four preimmune (0 day) or vaccinated macaques (14 days and 230 days) at a dilution of 1/40 (means ± SD; n = 3); In (d) the indicated vectors were incubated with fresh and complement-inactivated sera from two naive macaques applied at a twofold dilution and subsequently their transduction efficiency was evaluated on 293T cells Neutralization was calculated as the transduction efficiency in presence of serum relative to the transduction efficiency in the absence of serum (means ± SD; n = 3).

All sera completely neutralized H/F-LV transduction at high concentration (dilution 1/10; Figure 1a). Nevertheless, at lower serum concentrations, a difference in the neutralizing capacity between the human sera could be appreciated. The sera from donors 3 and 5 less neutralized the transduction activity of H/F-LVs compared to the other sera whereas sera from donors 4 and 7 contained the highest content of MV-specific neutralizing antibodies (Figure 1a). To discard the possibility that other factors in human serum induced the degradation of infectious H/F-LVs, total immunoglobulin G (IgG) antibodies were isolated from the sera from donors 3 and 7. We confirmed the specific neutralizing effect of the IgGs on H/F-LVs in both cases whereas the control VSV-G-LVs were only slightly affected (Figure 1b).

To confirm these data we collected the sera of macaques preimmunization and at 14 and 230 days post-MV vaccination. It is well known that macaques can be easily vaccinated against MV.28

H/F-LVs and VSV-G-LVs were not inactivated upon incubation with four different macaque sera collected preimmunization (Figure 1c; 0 days). In agreement with the results obtained with human sera (Figure 1a), H/F-LVs were neutralized completely by sera collected shortly after MV vaccination of the four macaques (14 days) or at a much later timepoint (230 days) whereas VSV-G-LVs were not (Figure 1c). Furthermore, incubation with sera containing functional complement or complement-inactivated sera from two different unvaccinated macaques, showed that the complement system played no major role in the inhibition of the H/F-LVs, whereas VSV-G-LVs were highly sensitive to macaque complement (Figure 1d and ref. 9).

In summary, the major hurdle for using H/F-LVs in vivo is the neutralization by MV-specific antibodies present in the blood.

Ablation of the major epitopes in Ed Hgp in H/F-LVs allows them to escape neutralization by mAbs

The majority of the MV neutralizing antibodies are directed against the Hgp.16 Indeed, two major epitopes were identified on H gp, the Noose epitope (aa379-410)18,19 and the NE epitope (aa245-250).17,20,29

In a first strategy, therefore, we chose to generate H/F-LVs mutated in these major epitopes of the H. We introduced four point mutations (K387G-G388A-Q391E-E395A) into the Ed Hgp (H) in order to inactivate the Noose epitope (HNo; Figure 2a). A second mutant, HNoNE, carried in addition to these Noose mutations, a mutation into the second major epitope NE (L246S-S247P; Figure 2a).

Figure 2.

Figure 2

Open in a new tab

Measles virus gp-pseudotyped lentiviral vectors (LVs) mutated in the major epitopes of the hemagglutinin escape from measles virus (MV)-specific monoclonal antibodies. (a) Schematic representation of the different MV hemagglutinin constructs. In Edm-Hgp (H) the binding sites to SLAM and CD46 are presented. In addition, the Noose and NE major epitopes that bind anti-MV antibodies in H are indicated; point mutations were introduced in Edm-H to knock out the Noose epitope55 and both the Noose and NE epitopes (HNoNE). (b) Surface staining of H gp on 293T cells transfected with hemagglutin encoding constructs (H, HNo, HNoNE) was performed with mouse monoclonal antibodies (mAbs) specific to the SLAM binding site (cl55), the Noose (BH216) and the NE (BH129) epitopes in H gp. The data are representative of three experiments. (c) 293T (CD46+), CHO-SLAM (SLAM+) and Raji (CD46+SLAM+) cells were transduced with H/F-, HNo/F-, HNoNE/F-, or VSV-G-LVs at a multiplicity of infection (MOI) of 0.2 after preincubation of these LVs with mAbs specific for the Noose (BH216, BH6) or NE (BH47, BH59, BH129) epitopes (means ± SD; n = 3). The specificity of the antibodies is shown in (a). Neutralization was calculated as the transduction efficiency in the presence of neutralizing antibodies relative to the transduction efficiency in the absence of antibodies.

H, HNo, and HNoNE gp expression was confirmed at the surface of transfected 293T cells by a monoclonal antibody (mAb) recognizing the SLAM-binding site in Hgp (cl55; Figure 2a,b). We confirmed that both HNo and HNoNE mutant gps lost affinity for a Noose-specific mAb (BH216; Figure 2a,b). As expected the HNoNE gp was, in addition, unable to bind an NE epitope-specific mAb (BH129; Figure 2a,b).

Next, we evaluated whether the H/F-LVs, HNo/F-LVs or HNoNE/F-LVs escaped neutralization by anti-Noose or anti-NE-specific mAbs. Titers of the two latter LVs were similar to H/F-LV titers on CHO-SLAM cells. However, titers determined on 293T (CD46+) cells were lower than those obtained for the original H/F-LV (Table 1). 293T, CHO-SLAM and Raji (CD46+SLAM+) cells were transduced with H/F-, HNo/F-, HNoNE/F-LVs, or VSV-G-LVs after preincubation with mAbs specific for the Noose (BH216 and BH6) or NE (BH47, BH59, and BH129; Figure 2a) epitopes. As a control, transductions were also performed in the absence of antibody. Since the SLAM epitope was intact in all three H gps, the specific antibody against the SLAM-binding site (cl55) was able to specifically neutralize H/F-, HNo/F-, HNoNE/F-LVs but not the VSV-G-LVs. Interesting, the anti-SLAM antibody also neutralized the different LVs for transduction of CD46+ SLAM– 293T cells (Figure 2c), which can be explained by the fact that SLAM– and CD46-binding sites are partially overlapping and that this antibody also masks the CD46-binding site in Hgp;30 As expected, only H/F-LVs were neutralized after preincubation with anti-Noose specific antibodies (BH216 and BH6), whereas the HNo/F-, HNoNE/F-LVs escaped their neutralizing effect, thus confirming that the Noose epitope was nonfunctional in the HNo gp as well as the HNoNE gp. The NE-specific antibodies (BH47, BH59, and BH129) inactivated H/F- and HNo/F-LVs when evaluated for transduction on CHO-SLAM and Raji cells since these carried no mutation in the NE epitope (Figure 1c). In contrast, HNoNE/F-LVs mutated in the NE epitope, clearly escaped neutralization by anti-NE antibodies in the same cell lines (Figure 2c). Interestingly, the same anti-NE antibodies did not neutralize the H/F-LVs for their transduction of CD46+-cells (293T), although they were able to bind the wild-type (wt) H gp (Figure 2b).

Table 1. Titers of measles-virus gp pseudotyped LVs.

graphic file with name mt201296t1.jpg

Open in a new tab

Since HNoNE/F-LVs were able to escape neutralization by NE- and Noose-specific antibodies, we next evaluated their capacity to escape inactivation by human sera. However, HNoNE/F-LVs were as sensitive to neutralization by different MV antibody-positive sera as the original H/F-LVs (Supplementary Figure S1).

In conclusion, solely disabling the Noose and NE neutralizing epitopes for anti-MV antibody binding was not sufficient to escape from the polyclonal MV-specific antibodies present in human sera.

Introduction of an additional glycosylation site in HNoNE/F-LVs allows escape from neutralization by human serum

In a second approach, we chose to take advantage of a glycosylation polymorphism present in several D genotypes. Indeed, the emergent MV D-clade strains (D4, D7, D11) carry an additional glycosylation site at amino acid (aa) position 416 (D416N) as compared to Ed MV in their hemaglutttinin (Supplementary Figure S2). It has been suggested that the glycosylation at aa position 416 might shield some H gp epitopes so that they become inaccessible for binding by MV-neutralizing antibodies in these D-clade strains.24 The D416N mutation was inserted in the HNoNE mutant (HNoNE-D416N; Figure 3a) to test this hypothesis. Presence of the glycosylation site was confirmed by western blot analyses, demonstrating that the molecular weight (MW) of HNoNE-D416N was higher than that of the original H gp but was similar to that of the HD4 gp, which also harbors the extra D416N glycosylation site (Figure 3b). Importantly, the insertion of the glycosylation site in HNoNE did not significantly affect titers on CHO-SLAM and 293T cells (Table 2).

Figure 3.

Figure 3

Open in a new tab

Introduction of an extra glycosylation site in H-Edm allows the H/F-LVs to escape neutralization by anti-measles virus (MV) antibodies in human sera. (a) Schematic representation of the different MV hemagglutinin glycoproteins. In Edm-Hgp (H) the binding sites for SLAM and CD46 are presented. In addition, the Noose and NE major epitopes that bind specific anti-MV antibodies in H are indicated. The MV clinical strain H gp (HD4) contains only SLAM binding residues and the Noose and NE epitopes. As indicated, the HD4 contains an extra glycosylation site (D416N) compared to H; the D416N glycosylation site was introduced in HNoNE mutant (HNoNE-D416N). (b) Immunoblots of lentiviral particles displaying H, HD4, or HNoNE-D416N and F gps at their surface. Lentiviral vectors (LVs) were purified over a sucrose cushion by ultracentrifugation. The upper part of the membrane was stained with a monoclonal antibody (mAb) against the ectodomain of H, the middle part with a monoclonal antibody against the ectodomain of F and the lower part of the membrane was stained with anti-HIV p24 antibody directed against the HIV capsid. (c) Raji (CD46+SLAM+) cells were transduced with H/F-, HNoNE/F-, HNoNE-D416N/F, or VSV-G-LVs at a multiplicity of infection (MOI) of 0.2 after preincubation of these LVs with mAbs specific for the Noose (BH216, BH6) or NE (BH47, BH59, BH129) epitopes or an antibody (BH99) sensitive to the D416N glycosylation (means ± SD; n = 3); (d) Raji cells were transduced with H/F-LVs or HNoNE-D416N/F-LVs at a MOI of 0.2 after preincubation with serial dilutions of human serum from two different donors. Neutralization was calculated as the transduction efficiency in presence of neutralizing antibodies or serum relative to the transduction efficiency in absence of antibodies or serum. Both results obtained for each serum of a different donor are expressed on the same graph; the line represents the mean of both results while the points represent the actual data.

Table 2. Titers of measles-virus gp pseudotyped LVs.

graphic file with name mt201296t2.jpg

Open in a new tab

Next, we evaluated whether LVs pseudotyped with HNoNE-D416N escaped neutralization by anti-MV mAbs. Raji cells were transduced with H/F-, HNoNE/F-, HNoNE-D416N/F-, or VSV-G-LVs upon preincubation with MV-specific mAbs. The anti-H gp antibody, BH99, has been mapped to the top loop connecting sheets β4 and β517 and binds to H when aa position 416 is not glycosylated. Antibodies specific for Noose and NE were also included in the neutralization experiment (Figure 3a). As expected, HNoNE-D416N/F-LVs still escaped neutralization by anti-Noose and anti-NE antibodies (Figure 3c). Additionally, HNoNE-D416N/F-LVs escaped neutralization by the B99 mAb. In constrast, the H/F-LVs and HNoNE/F-LVs, which lack the 416 glycosylation, were neutralized after preincubation with this antibody (Figure 3c). These results indicate that glycosylation at aa 416 in H indeed masks one or several epitopes present at the top loop connecting the β4 and β5 sheets. We then evaluated HNoNE-D416N-LV's ability to escape inactivation by MV antibody-positive human sera. Upon incubation with increasing dilutions of sera from two donors, HNoNE-D416N-LVs were 50% less sensitive to MV-specific neutralization than H/F-LVs (Figure 3d).

This last strategy confirmed our hypothesis that by modifying the glycosylation pattern of Ed H, H/F-LVs are rendered less sensitive to neutralizing antibodies present in human serum.

HD4Escape-Mut/F-LVs efficiently escape polyclonal MV antibodies in human sera

The H of Ed and the clinical D4 isolates (HD4 gp) have, apart from the D416A mutation and the CD46 receptor-binding sites, several other aa mismatches that might also play a role in MV antibody escape in vivo (Supplementary Figure S2 and Figure 4a). To test this hypothesis, we preincubated HD4/F-LVs with MV antibody-positive sera before incubation with Raji cells, which revealed a lower degree of neutralization compared to H/F-LVs (Figure 4b). In a second step, we introduced the Noose and NE major epitope mutations in HD4 gp resulting in HD4NoNE (Figure 4a). HD4NoNE/F-LVs escaped neutralization by MV antibody-positive sera very efficiently as compared to H/F-LVs (Figure 4b). Moreover, the former escaped neutralization better than the HD4/F-LVs suggesting a major impact of the D4 mutations in synergism with the epitope mutations in Noose and NE (Figure 4b).

Figure 4.

Figure 4

Open in a new tab

The HD4-NoNE gp that gained CD46 receptor-binding resists neutralization by human serum. (a) Schematic representation of the different measles virus (MV) hemagglutinin gps. The indicated residues, 481Y and 492G (marked in white), confer CD46 binding in Edm-Hgp while the clinical strain based gps HD4 and HD4NoNE contain the 481N and 492E residues (marked in black). The N481Y and E492G mutations were introduced in HD4NoNE resulting in HD4-NoNE-YG. In parallel, a chimeric mutant HChim5 was engineered by swapping the domain corresponding domain in H D4 for the one encoding for the mutated Noose region and the CD46-binding sites in HNoNEgp (see Figure 2a). (b,d) Raji cells were transduced with H/F-LVs, H-D4/F-LVs and HD4-NoNE/F-LVs (b), HD4-NoNE-YG/F-LVs or HChim5-LVs (d) at a multiplicity of infection (MOI) of 0.2 upon preincubation of these vectors with serial dilutions of human sera from two different donors. Both results obtained for each serum of a different donor are expressed on the same graph; the line represents the mean of both results while the points represent the actual data. (c) Immunoblots of lentiviral particles displaying H, HD4, HD4-NoNE-YG, or HChim5 at their surface. Lentiviral vectors (LVs) were purified over a sucrose cushion by ultracentrifugation. The upper part of the membrane was stained with a monoclonal antibody (mAb) against the ectodomain of H, the central part with a mAb against the ectodomain of F and the lower part of the membrane with an anti-HIV p24 antibody directed against the capsid.

Previously, we have amply documented that H/F-LV transduction of quiescent T and B lymphocytes requires binding of H gp to both the SLAM and CD46 receptors.10,11,13 Thus, HD4/F-LVs and HD4NoNE/F-LVs, which only recognize the SLAM receptor are not able to transduce quiescent lymphocytes as we previously demonstrated (ref. 13 and data not shown). Therefore, we chose to introduce hCD46 receptor-binding sites into HD4NoNE gp. Several studies have demonstrated that a single asparagine to tyrosine mutation at position 481 (N481Y) enables wt MV strains to bind the CD46 receptor.31 Moreover, additional introduction of the E492G mutation, adjacent to the N481Y aa substitution in a wt MV strain resulted in a still stronger binding to the CD46 receptor.31 By introduction of both of these H mutations into D4-HNoNE (Figure 4a), the corresponding HD4NoNE-YG-LVs indeed gained efficient entry through CD46 receptor as confirmed by their gain of titer on 293T cells (Table 2). In parallel, another mutant of HD4NoNE (Hchim5) was engineered by switching its corresponding domain for that containing the major CD46-binding sites and the Noose epitope of HNo gp (Figure 2a and Figure 4a). As expected, these mutant Hchim5/F-LVs also gained titer on 293T cells (Table 2), whereas HD4NoNE/F-LVs were unable to infect these target cells. In accordance with the production of high levels of infectious particles, we detected for both HD4NoNE-YG-LVs and Hchim5/F-LVs (collectively renamed HD4Escape-Mut/F-LVs) an efficient incorporation of H as well as F gps (Figure 4c). It should be noted that, the lower signal obtained for HD4NoNE-YG and Hchim5 compared to the HD4 and H gps is not due to lower incorporation on the LVs but to the lower binding affinity of the anti-H antibody (BH195) for its target epitope once the Noose epitope is mutated (Figure 4c). We next preincubated both HD4Escape-Mut/F-LVs and the original H/F-LVs with sera from donor 3 and 6 and then infected Raji cells. A very clear neutralization escape profile was detected for both the HD4Escape-Mut/F-LVs and the HD4NoNE-YG-LVs clearly outperformed Hchim5/F-LVs (Figure 4d).

These data underline that the two HD4Escape-Mut/F-LVs displaying Hgp from emerging wt D4 MV strains, which are also mutated in Noose and NE and have gained CD46 binding, are able to escape from the polyclonal MV-specific antibodies present in human sera.

The escape mutant H/F-LVs retain the capacity to efficiently transduce memory and naive T and B cells

Although we were able to restore CD46-tropism into a SLAM-tropic clinical strain HD4, we needed to verify that the two HD4Escape-Mut/F-LVs (HD4NoNE-YG/F- and Hchim5/F-LVs) also acquired the capacity to transduce primary resting lymphocytes, of importance for in vivo application.13

Thus, we isolated T and B lymphocytes from peripheral blood by negative selection to avoid their activation and we immediately transduced them with H/F-LVs, either of the HD4Escape-Mut/F-LVs, HNoNE-D416N/F-LVs, and the VSV-G-LVs. Three days post-transduction, green fluorescent protein (GFP) fluorescence was detected by fluorescence-activated cell sorting (FACS) analysis of the memory and naive cell subsets, which were distinguished by cell surface staining. The HD4Escape-Mut/F-LVs and HNoNE-D416N/F-LVs, achieved similar transduction levels as H/F-LVs in primary B cells. Detailed analysis showed that H/F-LV transduction reached 41% for the memory cell subset (CD27+) whereas the HD4Escape-Mut/F-LV transduction levels ranged from 40 to 56%. H/F-LV transduction levels of the naive cell subset (CD27–) was 27% whereas HD4Escape-Mut/F-LVs resulted in 23–40% transduction (Figure 5a). Similar observations were made for CD45RA+-naive T cells (transduction levels: 17% for H/F-LVs, 11–22% for mutant H/F-LVs) and for CD45RO+-memory T cells (transduction levels: 25% for H/F-LVs, 17–26% for the mutant H/F-LVs; Figure 5b). Moreover, bona fide CD45+CD62L+ T cells were transduced by all the H/F-LV pseudotypes (Supplementary Figure S3).

Figure 5.

Figure 5

Open in a new tab

HD4Escape-Mut/F-LVs efficiently transduce memory and naive T and B cells. T and B cells were purified from peripheral blood by negative selection. These unstimulated lymphocytes were immediately transduced with H/F-LVs, HNoNE-D416N/F-LVs, HD4-NoNE-YG/F-LVs, Hchim5/F-LVs (multiplicity of infection (MOI) of 1) and VSV-G-LVs (MOI of 50). Two days post-transduction surface expression of naive (CD27–) and memory B-cell (CD27+) subsets was detected by fluorescence-activated cell sorting (FACS) analysis. These data are representative of four experiments. (a) The GFP+ cells in these B-cell subsets are indicated in the upper right quadrant. Three days post-transduction surface staining for naive (CD45RA+) and memory (CD45RO+) subsets of T cells was performed by anti-CD45RA/anti-CD45RO double staining. (b) The GFP+ cells in these T-cell subsets are indicated in the upper right quadrant. These data are representative of four experiments. (c,d) represent long-term cultures of the LV transductions of unstimulated B and T cells. Transduced B cells were washed twice after transduction and cultured for 6 more days on MS5 feeder cells and were then analyzed for GFP-expressing naive (CD27–) and memory (CD27+) cells by (c) FACS, while transduced T cells were washed, and then cultured in RPMI medium supplemented with rIL-7 (10 ng/ml) for 6 days before FACS analysis of GFP-expressing naive (CD45RA+) and memory (CD45RO+) cells (d; means ± SD; n = 4). LVs, lentiviral vectors.

In order to verify that the observed transduction was stable, T cells were maintained in culture in the presence of rhIL7 for 6 more days and B cells were cultured on MS5 feeder cells for 6 more days. Here again, the level of transduction in the B cells was similar for all the pseudotypes (Figure 5c), and the same was detected for T cells (Figure 5d). As previously shown, B cells are more permissive to H/F-LV transduction than T cells and this was also true for the escape mutant H/F-LVs. A preferential gene transfer into memory T cells as compared to naive T cells confirmed our previous results (Figure 5d and ref. 10) while naive and memory B cells were transduced to the same extent by all H/F-LV pseudotypes (Figure 5c and previous results).11

In conclusion, HD4Escape-Mut/F-LVs in which the CD46-binding site was inserted next to the naturally present SLAM binding site gained the capacity to transduce primary human T and B cells.

H/F-LV escape-mutants efficiently transduce quiescent lymphocytes in the presence of MV antibody-positive serum

Finally, we set out to evaluate the performance of the three escape mutants of H/F-LVs in an “in vivo-like” experimental setting. Therefore, quiescent peripheral blood lymphocytes (PBLs), were transduced with H/F-LVs, HNoNE-D416N/F-LVs and both HD4Escape-Mut/F-LVs at low vector doses (multiplicity of infection (MOI of 1) either without serum or in the presence of increasing human serum concentrations (1%, 10%, or 30%) from four different donors (Figure 6).

Figure 6.

Figure 6

Open in a new tab

HD4Escape-Mut/F-LVs transduce quiescent peripheral blood lymphocytes (PBLs) more efficiently than the original H/F-LVs in the presence of human serum. Freshly isolated PBLs were transduced with H/F-LVs, HNoNE-D416N/F-LVs, HD4NoNE-YG/F-LVs, Hchim5/F-LVs (MOI of 1) in the absence of or with increasing human serum concentrations from four different donors. Three days post-transduction surface staining for (a) T cells and (b) B cells was performed with anti-CD3 and anti-CD19 antibodies, respectively and transduction efficiencies in these subpopulations were analyzed by fluorescence-activated cell sorting (FACS). The transduction efficiencies in the presence of serum were expressed as the percentage of GFP+-cells relative to the percentage of transduced T or B cells in the absence of human serum was which was set to 100% (means ± SD; n = 3). LVs, lentiviral vectors.

Three days post-transduction, PBLs were washed and kept for 4 more days in the presence of the survival cytokine, rIL-7. Double surface-staining was then performed with anti-CD3 (T cells) and anti-CD19 (B cells) antibodies and transduction efficiencies in these subpopulations were analyzed by FACS. The transduction efficiencies in the presence of serum were expressed as the percentage of GFP+-cells relative to the percentage of transduced T or B cells in the absence of human serum, which was set to 100% (Figure 6). In line with the results obtained for Raji cells (Figures 3 and 4), all three newly engineered mutant H/F-LVs were able to escape MV antibody neutralization by all four different sera. Escape was most pronounced for sera 3 and 5. We also observed that HD4-NoNE-YG/F-LVs and Hchim5/F-LVs escaped neutralization slightly better than the HNoNE-D416N/F-LVs and this was true for T and B cell transduction.

To test the performance of H/F-LVs in an in vivo model, we need to point out that these vectors are human tropic and do not transduce mice cells. Thus, a normal mouse model is inappropriate for in vivo evaluation. Therefore, we have humanized these mice by injection of hCD34+-cells into newborns. At 10 weeks of human cell reconstitution we injected these mice the original H/F-LV vector (2–5 × 105 IU) intravenously (Supplementary Figure S4a). We confirmed SLAM receptor expression on human thymocytes, splenocytes, and blood lymphocytes that reconstituted the mice since this receptor is needed for H/F-LV transduction (Supplementary Figure S4b). We found transduced peripheral lymphocytes, thymocytes, and splenocytes in the humanized animals (n = 4), at 6 weeks upon vector injection (Supplementary Figure S4c). To test in vivo escape from MV neutralizing antibodies, we tempted to vaccinate these humanized mice with virus-like particles carrying the MV glycoproteins at their surface, a method show by us to work very efficiently in immunocompetent mice.32 However, in 8 out of 8 mice we were unable to induce MV neutralizing antibodies in this HIS mouse model; thus this was not a valid in vivo model to test MV antibody neutralization.

To approach better an in vivo setting, we incubated whole blood of MV vaccinated donors with the different vectors. This seemed to us the closest we can approach an in vivo setting for several reasons: (i) We added highly concentrated vectors (300-fold concentrated) to the blood sample in order to keep dilution of the blood to an absolute minimum. The volume of the vector added to the whole blood never exceeded 10%, meaning that the vectors were exposed to levels of MV antibodies similar to in vivo conditions; (ii) The vectors are exposed to the human complement system since complement is active in the whole blood sample and to other unknown factors that might interfere with efficient transduction; (iii) Incubation of the vector with the blood was performed for 16 hours, which should allow the H/F-LVs to migrate to the hematopoietic tissues.

These are very stringent conditions since we have shown that an incubation with a 1/10 dilution of human serum for only 1 hour is enough to inactivate the unmodified H/F-LVs (Figure 6a,b). In conclusion, we used very stringent conditions to approach the in vivo settings (Figure 7).

Figure 7.

Figure 7

Open in a new tab

Novel engineered H/F-LVs allow efficient gene transfer into T cells within a blood sample. (a) Schematic representation of the transduction protocol. Four different H/F-LVs (H/F-LVs, HNoNE-D416N/F-LVs, HD4-NoNE-YG/F-LVs, and Hchim5/F-LVs) were used for transduction (multiplicity of infection (MOI) = 1) of total blood and control PBMC transduction. (b) Transduction efficiency of CD3+ T cells determined by fluorescence-activated cell sorting (FACS) analysis. The percentage of transduction obtained for the PBMCs present in the whole blood was expressed relative to the transduction efficiency of PBMCs in the absence of anti-measles virus (MV) antibodies which was set to 100%. The data for three different donors and three different sets of vector preparations are shown for day 5 as well as day 10 of culture in the presence of rIL-7. (c) Determination of transduction efficiency of CD3+ T cells by FACS analysis upon incubation of isolated PBMCs or whole blood with the different H/F-LV pseudotypes. In the dot blots the expression levels (MFI) of the GFP+ cells are indicated. LVs, lentiviral vectors.

As a reference an equal amount of isolated peripheral blood mononuclear cells (PBMCs) were transduced in the absence of MV-specific antibodies overnight with the same vector dose as indicated in Figure 7a. Following transduction, the PBMCs were isolated from the transduced blood, while the directly transduced PBMCs were washed and both were continued in IL-7 containing medium to reveal stable transduction. The HD4NoNE-YG/F-LVs and to a lower extent the HNoNE-D416N/F-LVs were able to escape neutralization by MV-positive antibodies in the blood which was revealed by efficient quiescent T cell transduction (see donor A and C in Figure 7b). In contrast, the original H/F-LVs were inactivated completely upon overnight incubation in the blood sample. Stability of the transduction is clearly demonstrated in the whole blood sample for the mutant HD4NoNE-YG/F-LVs since even 10 days upon isolation and culture in rIL-7, we obtained equivalent transduction levels as at day 5 (Figure 7b). Moreover, the best performing mutant (HD4NoNE-YG/F-LV) confers similar expression levels in the T cell transduced in the control PBMCs (MFI = 11,000) in absence of MV antibodies and in the whole blood sample, in which MV neutralizing antibodies can interfere with H/F-LV transduction (Figure 7c). Phenotyping again revealed a superior transduction of memory T cells (1.5–2-fold) versus naive T cells (data not shown).

In summary, HD4NoNE-YG/F-LVs outperformed the other mutant H/F-LVs in escape from MV-antibody neutralization in adult blood and allowed efficient lymphocyte transduction.

Discussion

The objective of this study was to make H/F-LVs fit for in vivo T and B cell gene transfer. To overcome inactivation of these vectors by MV H-specific antibodies present in the blood of MV vaccinated donors, we firstly mutated the identified immunodominant epitopes Noose and NE in Hgp. The resulting H/F-LVs escaped neutralization by MV monoclonal antibodies (mAbs) but were still neutralized by human polyclonal serum from vaccinated donors. Secondly, we took advantage of newly emerged MV-D genotypes that are less sensitive to MV vaccination due to an additional glycosylation site at position 416 of H gp. This glycosylation site, was introduced in Ed Hgp of the H/F-LVs, in addition to mutations of the major H gp epitopes, Noose and NE. In parallel, the latter epitopes were inactivated in the Hgp from the MV-D4 genotype, which was additionally engineered to gain CD46 receptor binding, a requirement for resting lymphocyte transduction. Both strategies resulted in >50% escape from MV antibody neutralization and efficient transduction of quiescent lymphocytes in total blood from MV vaccinated donors.

These new H/F-LVs that escape degradation by MV-antibodies and are resistant to human complement might thus represent the first gene therapy tools for T and B cell genetic modification in vivo. Indeed, H/F-LV-mediated T cell gene transfer may allow the treatment of several hematopoietic disorders like immunodeficiencies,33 cancer,34 and acquired diseases such as AIDS.35 In vivo transduction of naive T cells, which respond to novel antigens and persist for years can be envisaged for disease correction even covering the life-span of patients. Aiuti et al. obtained immune reconstitution in Adenosine deaminase immunodeficiency patients after T cell gene therapy.36 An important anticancer strategy based on the transfer of tumor-specific T cell receptor genes into patient T cells has been proven successful in the clinic.34 More recently, a radically different strategy consisting of introducing a coding sequence for a chimeric antigen receptor, allowed to confer the desired specificity for a cancer antigen to T cells. Recent case reports from ongoing clinical trials have described durable rejection of previously refractory B-cell malignancy in patients after CD19-directed chimeric antigen receptor therapy.37,38 Transgene expression in B cells is of particular interest as B cells have the potential to induce specific immune activation and tolerance, which could improve genetic vaccination against cancer or autoimmune diseases.39,40 Indeed, autologous tumor cells manipulated to increase the anti-tumor immune response could offer a new perspective of cure for patients with B cell cancers. B cells can also function as tolerogenic antigen-presenting cells upon introduction of an antigen-IgG fusion protein.41 Moreover, the reprogramming of autologous B cells to make a specific protective antibody would provide a continuous secretion of this antibody in vivo that could potentially neutralize viruses such as HIV and HCV.42 One possibility to augment efficacy in vivo would be to administer the vectors by local injection into lymph nodes or direct injection into the spleen to augment the local vector concentration and gene transfer into the target cells. We are convinced that for certain applications even low level transduction might be sufficient in vivo. For example Kalos et al.38 very recently reported that that use of chimeric antigen receptor T cells was successful in the treatment of three CLL patients and this thanks to a 1,000-fold in vivo expansion of the gene-modified T cells. In the context of immunotherapy memory B cells can be modified by H/F-LVs expressing a membrane bound antibody that can be converted to a secreted immunoglobulin form during B cell differentiation. B cells carrying the membrane bound antibody will expand when these cells encounter the specific antigen in vivo. Upon further differentiation into plasma B cells in vivo they then will secrete the specific antibody. All of the above mentioned T and B cell gene therapy applications might benefit from in vivo gene transfer since there is no need to isolate the cells, to culture them ex vivo and to reinfuse them. These three parameters may indeed affect the phenotype of the lymphocytes, lower their long-term lifespan in vivo and are accompanied by high costs. Moreover, the majority of T cells (90%) reside in the lymph nodes or thymus and not in the periphery. The same is true for B cells: only a fraction can be found in the periphery. Thus, in vivo administration of H/F-LVs shielded against MV antibody neutralization would allow transduction of these lymphocytes in their niche, where the majority of these cells reside. One safety issue should be kept in mind: the H/F-LVs result in proviral host genome integration and this can potentially result in insertional mutagenesis. However, this risk seems to be rather low in modified T cells since it was shown that retroviral vector integration deregulates gene expression in T cells to some extent but this had no consequences on the function and biology of the transplanted T cells.43 In B cells, this risk still needs to be assessed.

As the human anti-MV humoral response appears to be almost entirely directed against the H glycoprotein,16 we introduced mutations in this protein that would potentially allow escape from neutralization. Our strategy was to introduce natural mutations that are present in existing MV strains as these should not have deleterious effects on protein conformation. Moreover, a potential mechanism for the pathogenesis of SSPE—persistence of MV in the human brain—could involve escape by mutant viruses from the anti-MV humoral response. Therefore, we searched for mutations in the major epitopes of sequenced H proteins in a study of 11 SSPE cases21 and found that a majority had point mutations in either the HNE (noose) or NE epitopes (H. Kweder and R. Buckland, unpublished results). In addition to introducing such SSPE point mutations into our LV's H protein we also introduced the mutation D416N, which results in an additional N-linked glycosylation site in the MV Hgp. MV-H contains 5 potential N-glycosylation sites (Asn 168, 187, 200, 215, and 238). All of them are used except N238.44 The extra natural occurring D416N mutation, is present in the newly emerged MV genotype D7 (Supplementary Figure S2). In contrast to the common glycosylations, which cover the stem of Hgp, this extra one is present at the globular head of the protein. The same glycosylation, also present in D4 or D11 genotypes (Supplementary Figure S2), is involved in immunogenicity: D7 viruses escape 2 different mAbs, demonstrating the loss of two neutralizing B-cell epitopes. We showed that HD4 escapes a third anti-Hgp mAb (BH99). It has been speculated that this additional glycan masks other H protein epitopes and thereby allows D7 viruses to escape from neutralization by certain MV H-specific mAbs.24 It is generally accepted that one of the main barriers to antibody neutralization of viruses is the array of protective structural carbohydrates that cover antigens on the virus surface. Much of this carbohydrate diversity is a product of antigenic selection. Glycan patterns of viruses like HIV or influenza are continuously evolving in order to escape immune pressure from hosts and hence to facilitate survival in different host environments.45 For example, in the case of HIV, the acquisition (or loss) of a glycosylation site can have a dramatic effect on the immunogenicity of the surrounding protein surface.45 The presence of a N-linked glycosylation site at aa positions 413–415 of gp120 has been recently found to be associated with escape from potent and broad neutralizing antibodies.46 The same phenomenon is observed for HCV, in which the presence of glycans at the surface of envelope glycoproteins potentially explain how HCV evades the humoral immune response and why most HCV infections lead to chronicity. In fact, several data evidence that HCV envelope protein N-glycans mask conserved neutralizing epitopes at the surface of HCV particles.47

Other strategies to escape MV antibody neutralization in the context of oncolytic MV have been explored. Since MV has only one serotype, Miest et al.48 replaced the MV envelope glycoprotein H and F for those of the related canine distemper virus (CDV), generating chimeric MV virus that was resistant to neutralization by sera from MV-immune humans. We have also attempted the same strategy by pseudotyping LVs with the Hgp of CDV carrying a deletion in its cytoplasmic tail (H-CDV/F-LVs). These vectors were indeed resistant to MV antibody-positive serum (unpublished results, EV, CL, and FLC). But the H-CDV/F-LVs did not allow productive gene transfer of resting T and B cells. This was probably because the H-CDV gp differs from the H-Ed gp in 63% of its aa sequence. However, another possible explanation for the lack of lymphocyte transduction might be that CD46 receptors are species specific and thus the H-CDVgp binding to hCD46 is low or non-existent. In addition it should not be excluded that binding of H-CDV to hSLAM is inefficient since both CD46 and SLAM receptors need to be engaged by the Hgp for efficient lymphocyte transduction.13

Finally, we observed that the escape mutant HD4-NoNE-D416/F-LV was superior to the Hchim5/F-LVs in terms of anti-MV antibody escape in human serum when evaluated on cell lines as well as on primary human lymphocytes for transduction efficiency. The former escape H mutant is closer to the HD4 strain, which contains several additional point mutations as compared to HEd gp. In contrast, The Hchim5 gp contains a large domain of the HEd gp. Therefore, we hypothesize that HD4-NoNE-D416/F-LV might better escape polyclonal MV antibodies that were induced in the first place by the MV Ed vaccine.

Our results though are of high importance for many other applications of MV-based systems that are highly selective in vivo or are already in use in clinical trials for in vivo cancer therapy.49 The MV gps have been engineered to allow specific targeted cell entry to any cell population of interest. H and F gps are especially well suited for this purpose since the MV H gps were mutated to prevent binding to the MV receptors and can be engineered to incorporate any single chain antibody that recognizes cell surface antigens. Using this strategy the authors generated LVs specific for CD105+ endothelial cells, CD19+ B cells, CD133+ hematopoietic progenitors, GluA-expressing neurons and MHC II± B cells and DCs. Even highly selective in vivo targeting was demonstrated.50,51 However, since these re-targeted vector are based on MV Ed H en F MV gps they face the same challenges as our vectors: neutralization by MV-specific antibodies present in the majority of the population. Thus, this highly selective MV-based targeted LVs will also benefit from our H gp mutants that allow to escape in vivo neutralization.

Our findings are essential for two important areas of clinical application. Firstly, Ed MV proved to be an excellent vaccine platform for protection against measles but also shows potential for other viruses such HIV, HCV, and dengue virus.52,53,54 However, because the majority of the population is vaccinated against MV these MV based vaccines can only be used in pediatric patients since at later age these vectors would be neutralized immediately. Our escape mutant H gps are compatible with this platform and could allow to use the MV based vaccines in adults against viral infections. Secondly, MV based on Ed vaccine strains have potent oncolytic activity on a wide range of many different cancers and are now tested in the clinic for the treatment of multiple myeloma, glioblastoma pancreatic, and ovarian cancer.49 Indeed, Ed MV derivatives are tumor selective and can in vivo preferentially infect and destroy cancer cells while sparing the surrounding tissue. Additionally, they provoke the activation of specific anti-tumor immune responses. These MV-based oncolytic viruses, however, are strongly neutralized by MV-specific antibodies which reduces their efficacy upon in vivo application. Again here the mutations found in Hgp allowing MV antibody escape could mean an improvement for this in vivo anticancer therapy.

In summary, we have generated upgraded H/F-LV vectors for in vivo use in humans that bypass MV vaccine-induced antibody neutralization. This will allow interesting perspectives for in vivo T and B cell gene therapy applications of these vectors.

Material and Methods

Plasmids. Edm-Hgp coming from the Ed vaccine strain (SLAM− and CD46-tropic) and H-D4gp coming from clinical strain D4 (SLAM-tropic; ref. 25) and were used for pseudotyping of LVs.

HNo is a mutant Edm-H in which the Noose epitope has been mutated by the introduction of 4 aa exchanges (K387G-G388A-Q391E-E395A). HNoNE and D4-NoNE carry, in addition, the L246S-S247P mutations in the NE epitope of Edm-Hgp and D4, respectively. The HNoNE-D416N glycosylation mutant was generated by introduction of the D416N mutation in HNoNE. The H-D4NoNE-YGgp was generated from H-D4NoNE by PCR using primers containing the N481Y and E492G mutations to allow binding to the CD46 receptor. All Hgps and Fgp are inserted into pCG plasmids under the control of the cytomegalovirus early promoter. Cytoplasmic tails of all Hgps and Fgps were deleted by truncation of 24aa or 30aa, respectively to allow efficient incorporation on LVs. The pCMV-G plasmid has been described previously.5

Cell lines. CHO-SLAM (Chinese hamster ovary cells expressing SLAM) cells and Raji (B cell line-expressing CD46 and SLAM) were grown in RPMI medium (Gibco, Invitrogen, Auckland, New Zealand) supplemented with 10% fetal calf serum and 50 µg/ml of penicillin/Streptomycin; 293T (human kidney epithelial cells expressing CD46) were grown in DMEM (Gibco, Invitrogen) medium supplemented as for RPMI medium.

Vector production and titration. Self-inactivating HIV-1–derived vectors were generated by transient transfection of 293T cells in DMEM medium (Gibco, Invitrogen) as previously described.10 Briefly, for codisplay of the different Hgps and Fgps, 3 µg of each envelope plasmid was transfected together with a gagpol packaging plasmid and a plasmid encoding a LV-expressing GFP (SIN-HIVSFFVGFP). After 18 hours of transfection the medium was replaced by Optimem supplemented with Hepes (Gibco, Invitrogen). Viral supernatants were harvested 48 hours after transfection and filtered. Low-speed concentration of the vectors was performed by overnight centrifugation of the viral supernatants at 3,000g at 4 °C. Infectious titers (TU/ml) were determined by FACS in target cells by adding serial dilutions of the supernatants to the appropriate target cell line (CHO-SLAM or 293T for SLAM- or CD46-tropic vectors, respectively).

Antibodies. The following anti-H mAbs were used: BH216, BH6, BH4720 directed against the Noose epitope; BH59 and BH129 directed against the NE epitope20 and BH99 directed against the top loop connecting sheet β4 and β5;17 BH195 recognizing a linear epitope in the H ectodomain (western blot analysis); all thes mAbs were a gift from Claude Müller (Laboratoire National de Santé, Luxembourg, Luxembourg). The mouse anti-H cl55 mAb recognizing the SLAM-binding site and was a gift from Denis Gerlier (Inserm U758, ENS de Lyon, France) and anti-F mAb used for western blot analysis was a gift from Christian Buchholz (Paul-Ehrlich-Institute, Langen, Germany).

Serum isolation. Blood from adult human or macaque donors was collected in the absence of anticoagulant and was clotted for 1 hour at 37 °C and then centrifuged at 10,000g for 30 minutes. Serum (upper phase) was separated and when needed heat-inactivated for 1 hour at 56 °C. Serum was aliquoted and stored at –80 °C.

IgG purification from human serum. Total IgG was purified from human serum using Protein G Sepharose 4 Fast Flow (GE Healthcare, Uppsala, Sweden) according to the manufacturer's protocol.

Western blot analysis. HIV-GFP vectors pseudotyped with different H and F gps were purified by ultracentrifugation over a sucrose cushion. Subsequently, viral H gp was detected by western blot with an antibody recognizing the H ectodomain and viral F gp with an antibody recognizing the F ectodomain; incubation with an anti-HIV p24 antibody was used to reveal HIV capsid (CA).

FACS analysis. Cell lines and T and B cells were analyzed for transduction and surface markers: anti-hCD3-APC, anti-hCD19-PECy7, anti-hCD27-APC, anti-hCD45RO-PE and anti-hCD45RA-APC, anti-hCD150-PE and anti-hCD46-PE (BD Biosciences, Le pont de Claix, France). Detection by a Canto-II flow cytometer (BD Biosciences). Surface staining of Hgp on 293T cells transfected with Hgp-encoding constructs was performed by incubation with mouse anti-Hgp mAbs (clones: BH216, BH129, Cl55), followed by incubation with a secondary APC-coupled mAb (Clinisciences, Montrouge, France).

Cell line neutralization assay. Serum samples from patients vaccinated against measles were decomplemented at 56 °C for 1 hour and the day before transduction target cells were seeded in a 96-well plate (5,000 cells/well). Vectors (MOI = 0.2) were preincubated with different dilutions of the serum samples or anti-H specific mAb before incubation with target cells. The transduction efficiency was determined as the percentage of GFP+ cells and analyzed by high-throughput fluorescence-activated cell sorter analysis 72 hours after transduction. The percentage of cells transduced by vectors preincubated with serum or antibodies was divided by the percentage of cells transduced by vectors preincubated with medium.

Primary lymphocyte isolation. Adult peripheral blood samples, obtained from healthy donors after informed consent, were collected in acid citrate dextrose. Human peripheral blood T and B lymphocytes were isolated by negative selection using the Rosette Sep T an B cells isolation kits (Stem Cell Technologies, Grenoble, France) to avoid cell activation. Freshly isolated unstimulated lymphocytes were immediately seeded for transduction in RPMI 1640 medium (Gibco, Invitrogen) supplemented with 10% fetal calf serum (Lonza, Verviers, Belgium) and penicillin/streptomycin (Gibco, Invitrogen).

Transduction and neutralization assay of lymphocytes. For transduction, 5 × 104—1 × 105 cells were seeded in 48-well plates in RPMI/10% fetal calf serum medium and concentrated vector supernatants were added at the indicated MOIs. Transduction efficiency of vectors (%GFP+ cells) was assessed by FACS 72 hours after transduction. For some experiments cells were washed three times in PBS at day 3 post-transduction and kept in culture for up to 12 days, either by the addition of rhIL-7 (10 ng/ml; Preprotech, Neuilly sur Seine, France) for T cells or by coculture of B cells with MS5 feeder cells as previously described.10,11

For the neutralization assays 1 × 105 PBMC were seeded in 48-well plates in the presence of different quantities of decomplemented serum samples. Subsequently, concentrated vectors were added at a MOI of 1. After 3 days, surface staining for T- and B-cells was performed by double staining with anti-CD19-PeCy7/anti-CD3-APC (BD Biosciences) mAbs and transduction was analyzed by FACS.

The “in vivo-like” neutralization assay was performed as followed: (i) total human adult blood from a donor vaccinated against MV (containing 1 × 105 PBMCs) was incubated with wt Hgp or mutant Hgp pseudotyped vectors at an MOI of 10. (ii) In parallel, 1 × 105 PBMCs (isolated by Ficol gradient from the same blood donor) were transduced at the same vector doses. Following overnight incubation, the PBMCs were isolated from the blood (i) and the transduced PBMCs (ii) were washed twice. Cells were resuspended in RPMI/10% fetal calf serum/rIL7 (10 ng /ml) and kept in culture for 5 more days before anti-CD3-APC staining and FACS analysis was performed.

SUPPLEMENTARY MATERIAL Figure S1. H/F-LVs mutated in the major epitopes of hemagglutinin are degraded in MV antibody-positive serum. Figure S2. Sequence alignment of Edmonston H and H gps from MV D4, D7 and D11 strains. Figure S3. Escape mutant H/F-LVs transduce efficiently CD45RA+ CD62L+ naive T cells. Figure S4. H/F-LVs transduce T and B cells in vivo in humanized mice.

Acknowledgments

We thank C. Buchholz and D. Gerlier for sharing reagents. This work was supported by grants from the “Agence Nationale pour la Recherche contre le SIDA et les Hépatites Virales” (ANRS), the “Agence Nationale de la Recherche” (ANR), the European Research Council (ERC-2008-AdG-233130 “HEPCENT”) and the European Community (FP7-HEALTH-2007-B/222878 “PERSIST” and FP7-E-Rare “GENTHALTHER”). We acknowledge the contribution of the AniRA platform (flow cytometry) of the SFR BioSciences Gerland—Lyon Sud (UMS3444/US8). The authors declared no conflict of interest.

Supplementary Material

Figure S1.

H/F-LVs mutated in the major epitopes of hemagglutinin are degraded in MV antibody-positive serum.

Click here for additional data file. (45.5KB, pdf)
Figure S2.

Sequence alignment of Edmonston H and H gps from MV D4, D7 and D11 strains.

Click here for additional data file. (1.9MB, pdf)
Figure S3.

Escape mutant H/F-LVs transduce efficiently CD45RA+ CD62L+ naive T cells.

Click here for additional data file. (110.2KB, pdf)
Figure S4.

H/F-LVs transduce T and B cells in vivo in humanized mice.

Click here for additional data file. (195KB, pdf)

REFERENCES

  1. Frecha C, Lévy C, Cosset FL., and, Verhoeyen E. Advances in the field of lentivector-based transduction of T and B lymphocytes for gene therapy. Mol Ther. 2010;18:1748–1757. doi: 10.1038/mt.2010.178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bonini C, Ferrari G, Verzeletti S, Servida P, Zappone E, Ruggieri L.et al. (1997HSV-TK gene transfer into donor lymphocytes for control of allogeneic graft-versus-leukemia Science 2761719–1724. [DOI] [PubMed] [Google Scholar]
  3. Dardalhon V, Jaleco S, Kinet S, Herpers B, Steinberg M, Ferrand C.et al. (2001IL-7 differentially regulates cell cycle progression and HIV-1-based vector infection in neonatal and adult CD4+ T cells Proc Natl Acad Sci USA 989277–9282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Korin YD., and, Zack JA. Progression to the G1b phase of the cell cycle is required for completion of human immunodeficiency virus type 1 reverse transcription in T cells. J Virol. 1998;72:3161–3168. doi: 10.1128/jvi.72.4.3161-3168.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Maurice M, Verhoeyen E, Salmon P, Trono D, Russell SJ., and, Cosset FL. Efficient gene transfer into human primary blood lymphocytes by surface-engineered lentiviral vectors that display a T cell-activating polypeptide. Blood. 2002;99:2342–2350. doi: 10.1182/blood.v99.7.2342. [DOI] [PubMed] [Google Scholar]
  6. Verhoeyen E, Dardalhon V, Ducrey-Rundquist O, Trono D, Taylor N., and, Cosset FL. IL-7 surface-engineered lentiviral vectors promote survival and efficient gene transfer in resting primary T lymphocytes. Blood. 2003;101:2167–2174. doi: 10.1182/blood-2002-07-2224. [DOI] [PubMed] [Google Scholar]
  7. Bovia F, Salmon P, Matthes T, Kvell K, Nguyen TH, Werner-Favre C.et al. (2003Efficient transduction of primary human B lymphocytes and nondividing myeloma B cells with HIV-1-derived lentiviral vectors Blood 1011727–1733. [DOI] [PubMed] [Google Scholar]
  8. Serafini M, Naldini L., and, Introna M. Molecular evidence of inefficient transduction of proliferating human B lymphocytes by VSV-pseudotyped HIV-1-derived lentivectors. Virology. 2004;325:413–424. doi: 10.1016/j.virol.2004.04.038. [DOI] [PubMed] [Google Scholar]
  9. Sandrin V, Boson B, Salmon P, Gay W, Nègre D, Le Grand R.et al. (2002Lentiviral vectors pseudotyped with a modified RD114 envelope glycoprotein show increased stability in sera and augmented transduction of primary lymphocytes and CD34+ cells derived from human and nonhuman primates Blood 100823–832. [DOI] [PubMed] [Google Scholar]
  10. Frecha C, Costa C, Nègre D, Gauthier E, Russell SJ, Cosset FL.et al. (2008Stable transduction of quiescent T cells without induction of cycle progression by a novel lentiviral vector pseudotyped with measles virus glycoproteins Blood 1124843–4852. [DOI] [PubMed] [Google Scholar]
  11. Frecha C, Costa C, Lévy C, Nègre D, Russell SJ, Maisner A.et al. (2009Efficient and stable transduction of resting B lymphocytes and primary chronic lymphocyte leukemia cells using measles virus gp displaying lentiviral vectors Blood 1143173–3180. [DOI] [PubMed] [Google Scholar]
  12. Lévy C, Frecha C, Costa C, Rachinel N, Salles G, Cosset FL.et al. (2010Lentiviral vectors and transduction of human cancer B cells Blood 116498–500; author reply 500. [DOI] [PubMed] [Google Scholar]
  13. Frecha C, Lévy C, Costa C, Nègre D, Amirache F, Buckland R.et al. (2011Measles virus glycoprotein-pseudotyped lentiviral vector-mediated gene transfer into quiescent lymphocytes requires binding to both SLAM and CD46 entry receptors J Virol 855975–5985. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Viana PO, Ono E, Miyamoto M, Salomao R, Costa-Carvalho BT, Weckx LY.et al. (2010Humoral and cellular immune responses to measles and tetanus: the importance of elapsed time since last exposure and the nature of the antigen J Clin Immunol 30574–582. [DOI] [PubMed] [Google Scholar]
  15. Premenko-Lanier M, Hodge G, Rota P, Tamin A, Bellini W., and, McChesney M. Maternal antibody inhibits both cellular and humoral immunity in response to measles vaccination at birth. Virology. 2006;350:429–432. doi: 10.1016/j.virol.2006.02.029. [DOI] [PubMed] [Google Scholar]
  16. de Swart RL, Yüksel S., and, Osterhaus AD. Relative contributions of measles virus hemagglutinin- and fusion protein-specific serum antibodies to virus neutralization. J Virol. 2005;79:11547–11551. doi: 10.1128/JVI.79.17.11547-11551.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ertl OT, Wenz DC, Bouche FB, Berbers GA., and, Muller CP. Immunodominant domains of the Measles virus hemagglutinin protein eliciting a neutralizing human B cell response. Arch Virol. 2003;148:2195–2206. doi: 10.1007/s00705-003-0159-9. [DOI] [PubMed] [Google Scholar]
  18. Ziegler D, Fournier P, Berbers GA, Steuer H, Wiesmüller KH, Fleckenstein B.et al. (1996Protection against measles virus encephalitis by monoclonal antibodies binding to a cystine loop domain of the H protein mimicked by peptides which are not recognized by maternal antibodies J Gen Virol 77 (Pt 10)2479–2489. [DOI] [PubMed] [Google Scholar]
  19. Pütz MM, Hoebeke J, Ammerlaan W, Schneider S., and, Muller CP. Functional fine-mapping and molecular modeling of a conserved loop epitope of the measles virus hemagglutinin protein. Eur J Biochem. 2003;270:1515–1527. doi: 10.1046/j.1432-1033.2003.03517.x. [DOI] [PubMed] [Google Scholar]
  20. Fournier P, Brons NH, Berbers GA, Wiesmüller KH, Fleckenstein BT, Schneider F.et al. (1997Antibodies to a new linear site at the topographical or functional interface between the haemagglutinin and fusion proteins protect against measles encephalitis J Gen Virol 78 (Pt 6)1295–1302. [DOI] [PubMed] [Google Scholar]
  21. Jin L, Beard S, Hunjan R, Brown DW., and, Miller E. Characterization of measles virus strains causing SSPE: a study of 11 cases. J Neurovirol. 2002;8:335–344. doi: 10.1080/13550280290100752. [DOI] [PubMed] [Google Scholar]
  22. Finsterbusch T, Wolbert A, Deitemeier I, Meyer K, Mosquera MM, Mankertz A.et al. (2009Measles viruses of genotype H1 evade recognition by vaccine-induced neutralizing antibodies targeting the linear haemagglutinin noose epitope J Gen Virol 90Pt 112739–2745. [DOI] [PubMed] [Google Scholar]
  23. Shi J, Zheng J, Huang H, Hu Y, Bian J, Xu D.et al. (2011Measles incidence rate and a phylogenetic study of contemporary genotype H1 measles strains in China: is an improved measles vaccine needed Virus Genes 43319–326. [DOI] [PubMed] [Google Scholar]
  24. Santibanez S, Niewiesk S, Heider A, Schneider-Schaulies J, Berbers GA, Zimmermann A.et al. (2005Probing neutralizing-antibody responses against emerging measles viruses (MVs): immune selection of MV by H protein-specific antibodies J Gen Virol 86Pt 2365–374. [DOI] [PubMed] [Google Scholar]
  25. Domínguez A, Torner N, Barrabeig I, Rovira A, Rius C, Cayla J, Working Group for the Study of the Measles Outbreak in Catalonia et al. Large outbreak of measles in a community with high vaccination coverage: implications for the vaccination schedule. Clin Infect Dis. 2008;47:1143–1149. doi: 10.1086/592258. [DOI] [PubMed] [Google Scholar]
  26. Kremer JR, Brown KE, Jin L, Santibanez S, Shulga SV, Aboudy Y.et al. (2008High genetic diversity of measles virus, World Health Organization European Region, 2005-2006 Emerging Infect Dis 14107–114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Zhang Y, Ding Z, Wang H, Li L, Pang Y, Brown KE.et al. (2010New measles virus genotype associated with outbreak, China Emerging Infect Dis 16943–947. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Leonard VH, Hodge G, Reyes-Del Valle J, McChesney MB., and, Cattaneo R. Measles virus selectively blind to signaling lymphocytic activation molecule (SLAM; CD150) is attenuated and induces strong adaptive immune responses in rhesus monkeys. J Virol. 2010;84:3413–3420. doi: 10.1128/JVI.02304-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Bouche FB, Ertl OT., and, Muller CP. Neutralizing B cell response in measles. Viral Immunol. 2002;15:451–471. doi: 10.1089/088282402760312331. [DOI] [PubMed] [Google Scholar]
  30. Massé N, Ainouze M, Néel B, Wild TF, Buckland R., and, Langedijk JP. Measles virus (MV) hemagglutinin: evidence that attachment sites for MV receptors SLAM and CD46 overlap on the globular head. J Virol. 2004;78:9051–9063. doi: 10.1128/JVI.78.17.9051-9063.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Tahara M, Takeda M, Seki F, Hashiguchi T., and, Yanagi Y. Multiple amino acid substitutions in hemagglutinin are necessary for wild-type measles virus to acquire the ability to use receptor CD46 efficiently. J Virol. 2007;81:2564–2572. doi: 10.1128/JVI.02449-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Szécsi J, Gabriel G, Edfeldt G, Michelet M, Klenk HD., and, Cosset FL. DNA vaccination with a single-plasmid construct coding for viruslike particles protects mice against infection with a highly pathogenic avian influenza A virus. J Infect Dis. 2009;200:181–190. doi: 10.1086/599840. [DOI] [PubMed] [Google Scholar]
  33. Bordignon C, Notarangelo LD, Nobili N, Ferrari G, Casorati G, Panina P.et al. (1995Gene therapy in peripheral blood lymphocytes and bone marrow for ADA- immunodeficient patients Science 270470–475. [DOI] [PubMed] [Google Scholar]
  34. Morgan RA, Dudley ME, Wunderlich JR, Hughes MS, Yang JC, Sherry RM.et al. (2006Cancer regression in patients after transfer of genetically engineered lymphocytes Science 314126–129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Buchschacher GL., Jr, and, Wong-Staal F. Approaches to gene therapy for human immunodeficiency virus infection. Hum Gene Ther. 2001;12:1013–1019. doi: 10.1089/104303401750214249. [DOI] [PubMed] [Google Scholar]
  36. Aiuti A, Vai S, Mortellaro A, Casorati G, Ficara F, Andolfi G.et al. (2002Immune reconstitution in ADA-SCID after PBL gene therapy and discontinuation of enzyme replacement Nat Med 8423–425. [DOI] [PubMed] [Google Scholar]
  37. Kochenderfer JN, Wilson WH, Janik JE, Dudley ME, Stetler-Stevenson M, Feldman SA.et al. (2010Eradication of B-lineage cells and regression of lymphoma in a patient treated with autologous T cells genetically engineered to recognize CD19 Blood 1164099–4102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Kalos M, Levine BL, Porter DL, Katz S, Grupp SA, Bagg A.et al. (2011T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia Sci Transl Med 395ra73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Lei TC., and, Scott DW. Induction of tolerance to factor VIII inhibitors by gene therapy with immunodominant A2 and C2 domains presented by B cells as Ig fusion proteins. Blood. 2005;105:4865–4870. doi: 10.1182/blood-2004-11-4274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Stripecke R, Cardoso AA, Pepper KA, Skelton DC, Yu XJ, Mascarenhas L.et al. (2000Lentiviral vectors for efficient delivery of CD80 and granulocyte-macrophage- colony-stimulating factor in human acute lymphoblastic leukemia and acute myeloid leukemia cells to induce antileukemic immune responses Blood 961317–1326. [PubMed] [Google Scholar]
  41. Melo ME, Qian J, El-Amine M, Agarwal RK, Soukhareva N, Kang Y.et al. (2002Gene transfer of Ig-fusion proteins into B cells prevents and treats autoimmune diseases J Immunol 1684788–4795. [DOI] [PubMed] [Google Scholar]
  42. Luo XM, Maarschalk E, O'Connell RM, Wang P, Yang L., and, Baltimore D. Engineering human hematopoietic stem/progenitor cells to produce a broadly neutralizing anti-HIV antibody after in vitro maturation to human B lymphocytes. Blood. 2009;113:1422–1431. doi: 10.1182/blood-2008-09-177139. [DOI] [PubMed] [Google Scholar]
  43. Recchia A, Bonini C, Magnani Z, Urbinati F, Sartori D, Muraro S.et al. (2006Retroviral vector integration deregulates gene expression but has no consequence on the biology and function of transplanted T cells Proc Natl Acad Sci USA 1031457–1462. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Hu A, Cattaneo R, Schwartz S., and, Norrby E. Role of N-linked oligosaccharide chains in the processing and antigenicity of measles virus haemagglutinin protein. J Gen Virol. 1994;75 (Pt 5):1043–1052. doi: 10.1099/0022-1317-75-5-1043. [DOI] [PubMed] [Google Scholar]
  45. Scanlan CN, Offer J, Zitzmann N., and, Dwek RA. Exploiting the defensive sugars of HIV-1 for drug and vaccine design. Nature. 2007;446:1038–1045. doi: 10.1038/nature05818. [DOI] [PubMed] [Google Scholar]
  46. Gnanakaran S, Bhattacharya T, Daniels M, Keele BF, Hraber PT, Lapedes AS.et al. (2011Recurrent signature patterns in HIV-1 B clade envelope glycoproteins associated with either early or chronic infections PLoS Pathog 7e1002209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Helle F, Duverlie G., and, Dubuisson J. The hepatitis C virus glycan shield and evasion of the humoral immune response. Viruses. 2011;3:1909–1932. doi: 10.3390/v3101909. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Miest TS, Yaiw KC, Frenzke M, Lampe J, Hudacek AW, Springfeld C.et al. (2011Envelope-chimeric entry-targeted measles virus escapes neutralization and achieves oncolysis Mol Ther 191813–1820. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Msaouel P, Dispenzieri A., and, Galanis E. Clinical testing of engineered oncolytic measles virus strains in the treatment of cancer: an overview. Curr Opin Mol Ther. 2009;11:43–53. [PMC free article] [PubMed] [Google Scholar]
  50. Anliker B, Abel T, Kneissl S, Hlavaty J, Caputi A, Brynza J.et al. (2010Specific gene transfer to neurons, endothelial cells and hematopoietic progenitors with lentiviral vectors Nat Methods 7929–935. [DOI] [PubMed] [Google Scholar]
  51. Ageichik A, Buchholz CJ., and, Collins MK. Lentiviral vectors targeted to MHC II are effective in immunization. Hum Gene Ther. 2011;22:1249–1254. doi: 10.1089/hum.2010.184. [DOI] [PubMed] [Google Scholar]
  52. Brandler S, Ruffie C, Najburg V, Frenkiel MP, Bedouelle H, Desprès P.et al. (2010Pediatric measles vaccine expressing a dengue tetravalent antigen elicits neutralizing antibodies against all four dengue viruses Vaccine 286730–6739. [DOI] [PubMed] [Google Scholar]
  53. Garrone P, Fluckiger AC, Mangeot PE, Gauthier E, Dupeyrot-Lacas P, Mancip J.et al. (2011A prime-boost strategy using virus-like particles pseudotyped for HCV proteins triggers broadly neutralizing antibodies in macaques Sci Transl Med 394ra71. [DOI] [PubMed] [Google Scholar]
  54. Guerbois M, Moris A, Combredet C, Najburg V, Ruffié C, Février M.et al. (2009Live attenuated measles vaccine expressing HIV-1 Gag virus like particles covered with gp160DeltaV1V2 is strongly immunogenic Virology 388191–203. [DOI] [PubMed] [Google Scholar]
  55. Yanagi Y, Takeda M, Ohno S., and, Seki F. Measles virus receptors and tropism. Jpn J Infect Dis. 2006;59:1–5. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1.

H/F-LVs mutated in the major epitopes of hemagglutinin are degraded in MV antibody-positive serum.

Click here for additional data file. (45.5KB, pdf)
Figure S2.

Sequence alignment of Edmonston H and H gps from MV D4, D7 and D11 strains.

Click here for additional data file. (1.9MB, pdf)
Figure S3.

Escape mutant H/F-LVs transduce efficiently CD45RA+ CD62L+ naive T cells.

Click here for additional data file. (110.2KB, pdf)
Figure S4.

H/F-LVs transduce T and B cells in vivo in humanized mice.

Click here for additional data file. (195KB, pdf)

Articles from Molecular Therapy are provided here courtesy of The American Society of Gene & Cell Therapy

RESOURCES