Bispecific antibody for t cell activation for the killing of HIV infected cells

Abstract

The present invention relates to bispecific antibodies against CD3 and Human Immunodeficiency Virus (HIV) Env, to treat HIV infection.

Description

[0001] Bispecific antibody for T cell activation for the killing of HIV infected cells.

[0002] Field of the invention

[0003] The present invention relates to bispecific antibodies against CD3 and Human Immunodeficiency Virus (HIV) Env, to treat HIV infection.

[0004] Background of the invention

[0005] Since the beginning of the epidemic caused by HIV, it has led to 88.4 million infections worldwide and 42.3 million deaths. In 2023, 1.3 million people acquired HIV, 0.63 million died, and 39.9 million continue living with HIV. Despite the tremendous success of antiretroviral therapy (ART) in enhancing the health of people with HIV and turning the infection into a chronic condition, ART remains inaccessible to 30% of people with HIV.

[0006] Moreover, a functional cure for HIV remains elusive for people with HIV due to the rapid establishment of a latent viral reservoir early in infection. This reservoir, composed of long-lived, HIV-infected cells that persist despite ART, serves as a significant barrier to viral eradication. Consequently, strategies aimed at identifying and eliminating this reservoir are critical for achieving a functional cure.

[0007] Among the multiple strategies under study to control or eliminate the HIV reservoir, bispecific antibodies (bsAbs) have emerged as a promising class of therapeutic agents capable of redirecting cytotoxic immune cells to recognize and eliminate the HIV reservoir. One of these bsAbs simultaneously engages T cells via the T cell receptor (TCR or CD3) and targets the HIV-1 Envelope (Env) glycoprotein expressed on the surface of infected cells. Thus, bsAbs facilitate targeted immune-mediated clearance of the viral reservoir in people with HIV. WO13163427 A1 discloses multispecific antibodies that specifically bind a human immunodeficiency virus (HIV)-I envelope protein and CD3, and their use to treat an HIV-1 infection. Specifically, WO13163427 A1 discloses in example 1 a nucleic acid encoding for a bispecific antibody, comprising an anti-Env ScFv region, a linker and an anti-CD3 ScFv region, as well as the purified antibody. Other examples disclose the effect of the antibody in killing HIV infected cells and HIV latently infected cells.

[0008] New and more efficient bsAbs are needed to improve the specificity and efficacy in the killing of HIV infected cells. The present invention relates to a next-generation, small-sized bispecific antibody encoded by an mRNA vector — mRNA Env Light T cell engager or mRNA Env-LiTE. This novel therapeutic approach was designed to engage T cells through CD3 while specifically recognizing HIV-1 Env on infected cells (Figure 1A-B). The three critical functional aspects of mRNA Env-LiTE which show improvement over the prior art are: (1) its ability to bind specifically to each of its target antigens (CD3 and Env), (2) its capacity to activate T cells upon engagement, and (3) its effectiveness in mediating the selective killing of latently infected cells expressing HIV Env.

[0009] These results demonstrate the therapeutic potential of the bsAbs of the present invention for HIV Immunotherapeutics and strategies that control and eliminate the HIV reservoir.

[0010] Summary of the invention

[0011] A first aspect of the present invention relates to a nucleic acid comprising a sequence encoding for a bispecific antibody comprising an anti-CD3, a linker and an anti-Env-gp120 or anti-Env-gp140, wherein the anti-CD3 is a ScFv (single chain variable fragment) and / or wherein the anti-Env-gp120 or anti-Env-gp140 is a VHH (Llama variable heavy-chain antibody fragment).

[0012] A second aspect of the present invention relates to the bispecific antibody encoded by the nucleic acid of the first aspect.

[0013] A third aspect of the present invention relates to the nucleic acid of the first aspect or the bispecific antibody of the second aspect, for use in immunotherapy, particularly in immunotherapy against HIV, for killing HIV infected cells, even HIV latently infected cells.

[0014] Brief description of the figures

[0015] Figure 1. A. Schematic representation of Env-LiTE design. B. Schematic representation of the mechanism of action for Env-LiTE, redirecting T cells, particularly CD8+ cell cytotoxicity, to CD4+ latently infected cells expressing Env (HIV+) and promoting the specific elimination of the viral reservoir.

[0016] Figure 2. Representation of the Env-LiTE design and production in the form of DNA and mRNA molecules.

[0017] Figure 3. Binding of Env-LiTE. A. Bar graph representing the binding of Env-LiTE to CD3 and Env(gp120) by ELISA. The supernatant of the DNA, mRNA transfections and purified Env-LiTE, were analysed via ELISA by immobilizing CD3 and Env(gp120) onto the plate. B. Schematic of the binding and detection of Env-LiTE to Jurkat and K562-Env cell lines. C. Gating strategy for the detection of Env-LiTE in Jurkat and K562-Env cells. D. Bar graph representing the specific binding of Env-LiTE to CD3 and Env(gp120) in the membrane of Jurkat and K562-Env, respectively, as detected by anti-His-tag staining. Each bar represents the median value for a technical replicate (n=3). Figure 4. Env-LiTE activation in Jurkat and off-target effect. A. Schematic of the experimental design B. Gating strategy for the detection of CD69 in Jurkat cells. C. Dose response curve representing the activation capacity of a gradient of OKT3, purified Env-LiTE, in the presence / absence of Env(gp120), at concentrations ranging from 0.1-10,000 ng / ml measured by flow cytometry of the CD69 activation marker. Points in graph represent median ±SD (n=6). D. Dose response curve representing the activation capacity of a gradient of OKT3, purified Env-LiTE, DNA supernatant and mRNA supernatant transfection in the presence / absence of Env(gp120), at concentrations 0, 10, 25 and 50 ng / ml measured by flow cytometry of the CD69 activation marker. Points in graph represent median ±SD (n=6). E. Bar graph representing the activation under expossure to 25 ng / ml of OKT3, purified Env-LiTE, DNA supernatant and mRNA supernatant transfection in the presence / absence of Env(gp120) measured by flow cytometry of the CD69 activation marker, significance calculated by One Way-ANOVA. Bar represents median ±SD and each point of a technical replicate.

[0018] Figure 5. Env-LiTE activation in PBMCs and off-target effect. A. Schematic of the experimental design. B. Gating strategy for the detection of CD25 and CD69 activation markers in CD8+ PBMCs. C. Curves representing the activation capacity of a gradient of OKT3, purified Env-LiTE, DNA supernatant and mRNA supernatant transfection in the presence / absence of Env(gp120) measured by flow cytometry of CD69 alone (top) and CD69, CD25 co-expression (bottom). Points in graph represent median ±SD (n=3). D. Bar graph representing the activation capacity of a 25 ng / ml of OKT3, purified Env-LiTE, DNA supernatant and mRNA supernatant transfection in the presence / absence of Env(gp120) measured by flow cytometry CD69 alone (top) and CD69, CD25 co-expression (bottom) significance calculated by One Way-ANOVA (n=3). Bar represents median ±SD; each point is a replicate from 3 independent experiments and donors.

[0019] Figure 6. mRNA Env-LiTE activation in PBMCs and evaluation of off-target effect with the BSDB-258 as comparator. A. Dose-response curves of 0, 10, 25, 50, and 100 ng / ml of OKT3, mRNA Env-LiTE and BSDB-H258 in the presence or absence of Env(gp120), response was measured by flow cytometry by CD69 (top) and CD69 / CD25 co-expression. Dots indicate median values with ranges (n=3). B. Activation capacity at 25 ng / ml of OKT3, mRNA supernatant transfection, and BSDB-H258, with or without Env(gp120), measured by flow cytometry for CD69 and CD69 / CD25 co-expression. Significance was assessed using One-Way ANOVA. Bars represent median ± SD; each point represents a replicate (n=3 donors, tested in duplicates).

[0020] Figure 7. mRNA Env-LiTE redirected cytotoxicity in co-cultures. A. Schematic of the experimental design. B. Gating strategy for the differentiation of the PBMCs and the target cell lines, the detection of activation of CD8+ PBMCs and the determination of cell line viability. C. Box and whiskers graph representing the activation capacity in CD8+ primary cells (top) and the cytotoxicity in the target cell lines (bottom) of 25 ng / ml of OKT3, purified Env-LiTE and mRNA supernatant transfection, in co-culture with ACH-2 (HIV+) and A3.01(HIV-) cells at 1 :1 and 1 :5 effector to target ratio measured by flow cytometry as the expression of CD69 (n=3 donors).

[0021] Figure 8. Activation of CD8+ and cytotoxicity by mRNA Env-LiTE and BSDB-H258. Box and whiskers graphs representing the activation capacity in CD8+ primary cells (top) and the cytotoxicity in the target cell lines (bottom) of 25 ng / ml of OKT3 (OKT3), mRNA Env-LiTE supernatant transfection (mRNA-LiTE 25) and BSDB-H258 (BSDB-H258 25), 1000 ng / ml mRNA Env-LiTE supernatant transfection (mRNA-LiTE 1000), BSDB-H258 (BSDB-H258 25) in co-culture with ACH- 2 (HIV+; left) and A3.01 (HIV-;right) cells at 1 :1 effector to target ratio. Cytotoxicity was estimated by comparison of cell count relative to the control condition, and activation was measured by flow cytometry as the expression of CD69. Lines represent median ± range, and each dot, a replicate (n=3 donors, tested in duplicate).

[0022] Detailed description of the invention

[0023] A first aspect of the present invention relates to a nucleic acid comprising a sequence encoding for a bispecific antibody comprising an anti-CD3, a linker and an anti-Env-gp120 or anti-Env-gp140, wherein the anti-CD3 is a ScFv (single chain variable fragment) and / or wherein the anti-Env-gp120 or anti-Env-gp140 is a VHH (Llama variable heavy-chain antibody fragment). This bispecific antibody of the invention, herein referred to as Env-LiTE, is lighter than those of the prior art and has been found to be more efficacious in killing HIV infected cells than the bispecific antibody of the prior art with anti-CD3 scFv and anti-Env-gp120 scFv, herein called BSDB-H258. The inventors have found that the bispecific antibody of the present invention achieves a much higher and significant Env- dependent activation than the prior art bispecific antibody BSDB-H258, such as the one described in WO13163427 A1 , and it is efficient in killing HIV latently infected cells, in contrast with the bispecific antibody BSDB-H258, such as the one disclosed in WO13163427 A1 , which does not show this ability.

[0024] As used herein, a bispecific antibody is a recombinant molecule composed of two different antigenbinding domains that consequently bind to two different antigenic epitopes. Bispecific antibodies include chemically or genetically linked molecules of two antigen-binding domains. The antigenbinding domains can be linked using a linker. The antigen binding domains can be monoclonal antibodies, antigen-binding fragments (e.g., Fab, scFv), or combinations thereof. A bispecific antibody can include one or more constant domains but does not necessarily include a constant domain. An example of a bispecific antibody is a bispecific single chain antibody including an scFv that specifically binds to Env(gp120) joined (via a peptide linker) to an scFv that specifically binds CD3, such as the bispecific antibody BSDB-H258 and such as the one disclosed in WO13163427 A1 , which is used herein as a comparator. Unlike said comparator, the bispecific antibody of the present invention comprises an anti-Env-gp120 VHH or an anti-Env-gp140 VHH, preferably, it comprises an anti-Env-gp120 VHH.

[0025] BSDB-H258 is a commercial antibody by Creative Biolabs described as a recombinant Human Anti- CD3 x Anti-HIV-1 Env(gp120) Bispecific Antibody (Diabody), for research use only and not intended for diagnostic use, designed to be expressed as two polypeptide chains respectively. The two chains are cross-over pairing to form a dimer.

[0026] As used herein, VHH is a Llama variable heavy-chain antibody fragment, which consists of a single immunoglobulin domain, rather than the two (heavy and light chains) found in traditional antibodies.

[0027] In a preferred embodiment of the first aspect of the present invention, the nucleic acid comprises a sequence encoding for an anti-CD3 ScFv and an anti-Env-gp120 VHH.

[0028] In a preferred embodiment of the first aspect of the present invention, the nucleic acid comprises a sequence encoding for an anti-CD3 ScFv and an anti-Env-gp140 VHH.

[0029] In a preferred embodiment of the first aspect of the present invention, said nucleic acid is RNA. Preferably, said nucleic acid is mRNA. More preferably, said mRNA comprises methylpseudouridines.

[0030] In an advantageous embodiment, the nucleic acid of the first aspect could be delivered to the patients in the form of an mRNA comprising methylpseudouridines, so that the bispecific antibody would directly be produced in the body of the HIV infected patient.

[0031] In a preferred embodiment of the first aspect, the nucleic acid encodes for the bispecific antibody comprising SEQ ID NO: 3. In a preferred embodiment of the first aspect, the nucleic acid encodes for the bispecific antibody with SEQ ID NO: 1. Sequences comprising conservative amino acid substitutions of SEQ ID NO: 3 or SEQ ID NO:1 are also comprised by the invention, as well as any nucleic acid comprising sequence encoding for said conservative amino acid substitutions.

[0032] In a preferred embodiment of the first aspect, the nucleic acid comprises SEQ ID NO: 4. In a preferred embodiment of the first aspect, the nucleic acid comprises SEQ ID NO: 2. In a preferred embodiment of the first aspect, the nucleic acid consists of SEQ ID NO: 2.

[0033] The second aspect of the present invention is the bispecific antibody encoded by the nucleic acid of the first aspect. This bispecific antibody preferably comprises an anti-CD3 ScFv and an anti-Env- gp120 VHH or an anti-CD3 ScFv and an anti-Env-gp140 VHH. The sequence of the anti-Env-gp120 VHH is preferably SEQ ID NO: 3. In a preferred embodiment, the bispecific antibody of the invention comprises or consists of SEQ ID NO: 1. Sequences comprising conservative amino acid substitutions of this SEQ ID NO:1 are also comprised by the invention.

[0034] The nucleic acid of the first aspect and / or the bispecific antibody of the second aspect can be used for immunotherapy, as they have proven to activate T cells and promote the killing of the HIV-infected cells, including those latently infected. Thus, a third aspect of the present invention relates to the nucleic acid of the first aspect and / or the bispecific antibody of the second aspect for use in immunotherapy against HIV, preferably for use in immunotherapy against HIV for killing HIV infected cells, even for use in immunotherapy against HIV for killing HIV latently infected cells.

[0035] In any of the aspects of the invention, the present invention also refers to the embodiments where the expression “comprises” or “comprising” means “consists of” or “consisting of”. The present invention includes any combination of the different embodiments and preferred embodiments disclosed herein for any of the aspects of the invention.

[0036] The present invention is illustrated by the following examples:

[0037] Methodology

[0038] 1. Prototype design and production. Env-LiTE was designed by selecting the Camelid broadly neutralizing anti-Env(gp120) antibody J3 and the OKT-3 anti-T cell receptor (TCR / CD3) antibody. The single-chain variable fragment (scFv) sequences of both antibodies were retrieved from the EMBL-EBI database. To generate the fusion protein, the scFv sequences were linked using a (GGGS)3linker. His-Tag and Myc-Tag sequences were appended to the N-terminus using SnapGene software to facilitate purification and detection. The resulting sequences were optimized for human codon usage to ensure efficient expression in mammalian cells, and as a result, SEQ ID NO: 2 was obtained.

[0039] Generated sequences were produced using the p.C.DNA 3.4 plasmid from GenScript and as mRNA containing methylpseudouridines. Both the plasmid DNA and the mRNA were transfected into EXPI-293 cells. After three days of incubation, supernatants were collected from both DNA- and mRNA-transfected cells. Env-LiTE was then purified from the collected supernatants using metal affinity chromatography (Figure 2).

[0040] 2. Binding capacity and specificity of Env-LiTE to CD3 and Env were assessed by ELISA using immobilized CD3 and Env(gp120) recombinant proteins to coat plates and indirect detection using a primary goat anti-His antibody and a biotinylated secondary antibody. Moreover, the binding capacity to proteins expressed on cell membranes was also assessed in CD3-expressing Jurkat T cells and Env-expressing K562 cells. After incubating the prototypes for 1 hour, a secondary anti-His Tag PE (phycoerythrin) was used for their indirect detection by flowcytometry.

[0041] 3. On and off-target activation of Env-LiTE were analysed in Jurkat cells and peripheral blood cells (PBMCs) (n=3). On and off-target effects were established in Jurkat and PBMCs in a dosedependent concentration of Env-LiTE using anti-CD3 IgG (OKT-3) as a positive control for activation and Creative Biolabs ref. BSDB-H258 as a commercial comparator. After 48 hours of incubation, cells were immunostained for activation markers (CD69 and CD25), and the frequency of activated cells was measured by flow cytometry.

[0042] 4. Env-LiTE specific cytotoxicity on HIV-latently infected cells. PBMCs from healthy donors (n=3) were co-cultured with HIV latently infected with the cell line (ACH-2) and the parental, noninfected cell line (A3.01). Cell lines were stained with the cell tracker CFSE (Carboxyfluorescein succinimidyl ester), for population differentiation. Co-cultures were then treated with 25 ng / ml Env-LiTE, 25 ng / ml mRNA-LiTE, 1000 ng / ml mRNA-LiTE and, 25 ng / ml BSDB-H258, 1000 ng / ml BSDB-H258 as commercial comparator and anti-CD3 IgG (OKT-3) as a positive control. After 48 hours, cells were stained for viability and activation (CD69) and analysed using flow cytometry.

[0043] Example 1. Binding specificity of mRNA Env-LiTE

[0044] We assessed the binding of Env-LiTE using both DNA and mRNA transfection supernatants, as well as purified Env-LiTE protein obtained from DNA-transfected EXPI293F cells. The His-tag on Env- LiTE was used for purification via metal affinity chromatography. We observed consistent specific binding to CD3 and Env from the prototype in mRNA, DNA and protein (Figure 3A).

[0045] To test the binding of Env-LiTE to CD3 expressed in cell membranes, the CD3-expressing cell line Jurkat was cultured in the presence of 10 pg / ml of the different Env-LiTE prototypes (in the form of purified protein, DNA or mRNA) for 1 hour, and a secondary anti-His Tag (PE) was used for the indirect detection of Env-LiTE by flow-cytometry (Figure 3B). In addition, to test the binding of Env- LiTE to Env(gp120) expressed on the surface of a cellular membrane, we generated a CD3-negative, Env(gp120)-positive cell line. We selected K562 cells, a lymphoblast cell line that does not express CD3, and transfected it with a plasmid encoding HIV-1 Env(gp120). The expression of Env(gp120) in the transfected K562-Env cells was confirmed using a primary Env(gp120) bnAb (broadly neutralizing antibody B12) and a secondary anti-mouse IgG antibody in PE. Once the expression of Env was confirmed, K562-Env cells were cultured with 10 pg / ml of the different Env-LiTE prototypes (in the form of purified protein, DNA, or mRNA) for 1 hour. After the incubation, a secondary anti-His Tag (PE) was used for the indirect detection of Env-LiTE (Figure 3C). In summary, these data confirm the specific binding of Env-LiTE to CD3 and Env(gp120) in two complementary in vitro experiments, including cellular membranes and it is consistent for the three Env-LiTE forms studied (protein, DNA and mRNA).

[0046] Example 2. On-target activation assessment in Jurkat cells

[0047] We analyzed the in vitro capacity of Env-LiTE to activate the Jurkat cell line, and determined the concentration of Env-LiTE ensuring specificity in the absence of off-target effects in the presence of both CD3 and Env. For these experiments, Fc-Env(gp120) recombinant protein was coated on 96- well plates, while BSA was used as a negative control. A commercial OKT-3 (anti-CD3) antibody was selected as a positive control for Jurkat cells activation and comparator with Env-LiTE in the context of anti-CD3 engagement (Figure 4A).

[0048] First, we determined the optimal concentration range to minimize the off-target effect of Env-LiTE while having specific T cell activation. Concentrations between 0.1 ng / ml and 10 pg / ml of Env-LiTE and OKT-3 were tested with or without coated Fc-Env(gp120) for 24h. We observed a high effect in activation, measured as CD69 expression, by OKT-3 starting from 10 ng / ml and reaching a plateau at 100 ng / ml (Figure 4C). Comparatively, we observed an Env-dependent specific activation by Env- LiTE in the range of 10 and 50 ng / ml (Figure 4C- D).

[0049] After determining the optimal concentration for Env(gp120)-specific Env-LiTE-mediated activation to avoid potential off-target effects, we tested all Env-LiTE prototypes (protein, DNA, mRNA) under the same experimental conditions in a dose-dependent manner (0, 10, 25, and 50 ng / ml). As expected, we observed consistent Env-dependent Jurkat activation with all Env-LiTE prototypes (protein, DNA, and mRNA) compared to OKT-3 (Figure 4D-E). These experiments demonstrated the capacity of Env-LiTE to activate Jurkat cells through the CD3 receptor in an Env-dependent manner at concentrations of 10-50 ng / ml. At the same time, no off-target activation effect for Jurkat cells was detected for Env-LiTE in the absence of Env(gp120) in the culture medium.

[0050] Example 3. Env-LiTE specific activation of primary CD8+ and CD4+ T cells

[0051] Next, we assessed in vitro the capacity of Env-LiTE to activate specifically, only in the presence of Env(gp120), T cells from total PBMCs. Briefly, we coated the Fc-Env(gp120) recombinant protein on 96-well plates, while BSA was coated in the negative control wells (Figure 5A). An OKT-3 antibody was selected as a positive control with Env-LiTE. CD8+ T cell activation was evaluated by measuring CD25 and CD69 expression on the cell surface after incubation (Figure 5B).

[0052] We tested Env-LiTE at 0, 10, 25, and 50 ng / ml over 48 hours. After incubation, we observed an increase in CD69 and CD69 / CD25 activation markers on the surface of CD8+ T cells for all the forms of Env-LiTE tested (protein, DNA, and mRNA) in a dose-dependent and Env(gp120)-dependent manner (Figure 5C-D). By comparison, we observed unspecific activation by OKT-3, with increased CD69 expression in CD8+ T cells independent of Env(gp120). Therefore, these data further support the ability of Env-LiTE to activate CD8+ T cells through the CD3 receptor and Env(gp120)-specific manner when compared to OKT3.

[0053] Moreover, we performed additional experiments with the mRNA Env-LiTE and a commercially available diabody BSDB-H258 (anti-CD3 x anti-Env(gp120); Creative Biolabs), as a direct comparator of activity. We evaluate Env-dependent activation of T cells by measuring CD4+ and CD8+ T cell activation by the expression of CD25 and CD69 following 48 hours of incubation of PBCMs with OKT-3, mRNA Env-LiTE, or BSDB-H258 at concentrations of 0, 10, 25, 50, and 100 ng / ml and in the presence or absence of Env(gp120). As shown in Figure 6, we found a statistically significant increase in activation for both CD69 / CD25 CD8+ and CD69 / CD25 CD4+ T cells for mRNA Env-LiTE in the absence of changes in activation for the BSDB-H258 diabody comparator at similar concentrations. These findings indicate the superiority of mRNA Env-LiTE to induce CD4+ and CD8+ T cell Env(gp120)-dependent activation to the BSDB-H258 diabody comparator.

[0054] Example 4. Env-LiTE induces specific cytotoxicity of HIV-1 latently infected cells

[0055] Last, we evaluated the capacity of Env-LiTE to induce the specific killing of HIV latently infected cells.

[0056] Briefly, PBMCs from healthy donors were collected and cocultured with two different cell lines, A3.01 and ACH-2. The A3.01 is an HIV uninfected lymphocyte immortalized parental cell line used as negative control, and the ACH-2 is a descendant immortalized cell line from the A3.01 with a competent LAI HIV integrated. We cocultured PBMCs as effector cells (E) and A3.01 or ACH-2 as target cells (T) representing uninfected (A3.01) or latently HIV-infected cells (ACH-2) (Figure 7A).

[0057] Target cells (A3.01 and ACH-2) were stained with the CFSE cell tracker to facilitate their gating by flow cytometry. Cells were co-cultured at two different EffectorTarget (E:T) ratios, 1 :1 and 5:1 , in the presence of PBS (negative control), OKT-3 (positive control), purified Env-LiTE, or mRNA Env-LiTE at 25 ng / ml. After 48 hours, the cells were analyzed using flow cytometry to determine CD8+ T cell activation, as indicated by CD69 expression and cytotoxicity, as indicated by elimination of CFSE labelled target cells (Figure 7A-B).

[0058] We corroborated the activation of CD8+ T cells in the co-culture with HIV latently infected cells (ACH- 2) by an increase of CD69 in the presence of Env-LiTE and OKT3. By contrast, when target cells were A3.01 , CD8+ T cells were only activated by OKT3 and not Env-LiTE, indicating the specificity of Env-LiTE activation only in the presence of HIV-1 Env expression in ACH-2 (Figure 7C). Furthermore, we observed a specific killing of latently infected ACH-2 cells induced by Env-LiTE. Meanwhile, the killing in A3.01 was only produced in the presence of OKT3, further supporting the specificity of Env-LiTE to induce cell death of HIV-1 latently infected cells (Figure 7D). Moreover, the commercially available bispecific diabody BSDB-H258 comparator, tested at 25 ng / ml and 1000 ng / ml, was not able to induce Env-dependent CD8+ activation measured by CD69, nor the specific killing of HIV latently infected cells in co-culture at any of the concentrations tested when compared to our mRNA Env-LiTE prototype at similar concentrations (Fig. 8).

[0059] These experiments demonstrated that Env-LiTE can induce specific CD8+ cell activation in the coculture with HIV latently infected cells but not in uninfected cells. Importantly, we showed specific cytotoxicity to HIV latently infected by Env-LiTE and mRNA Env-LiTE.

[0060] Our results confirm that mRNA Env-LiTE binds specifically to CD3 and HIV Env. This binding capability is critical for targeted therapeutic interventions.

[0061] We demonstrated mRNA Env-LiTE's capacity to activate CD3+ cell lines and CD8+ cells from PBCMS in vitro in an envelope-dependent manner without off-target effects. mRNA Env-LiTE showed the capacity to activate total PBMCs, particularly CD8+ T cells, leading to specific cytotoxicity against HIV-1 latently infected cell lines in a cellular co-culture model. mRNA Env-LiTE demonstrated superior Env(gp120)-specific activation of both CD4+ and CD8+ T cells compared to the commercially available bispecific diabody BSDB-H258 comparator. This enhanced activation was evident through increased surface expression of activation markers CD25 and CD69 in a Env(gp120)-dependent manner, highlighting the improved ability of mRNA Env-LiTE to engage and stimulate T cells. Furthermore, the superiority of mRNA Env-LiTE was maintained in the co-culture assay, where it was able to activate CD8+T cells and specifically eliminate ACH-2 HIV-latently infected cells. In contrast, the BSDB-H258 comparator neither activated the cells nor eliminated HIV-latently infected cells in this assay at similar concentrations or 40 times higher.

[0062] These findings highlight the therapeutic potential of mRNA Env-LiTE in targeting and eliminating latently HIV-infected cells in a specific manner, dependent on CD3 and Env dual engagement and with superior efficacy compared to BSDB-H258.

Claims

CLAIMS1. A nucleic acid comprising a sequence encoding for a bispecific antibody comprising an anti- CD3, a linker and an anti-Env-gp120 or anti-Env-gp140, wherein the anti-CD3 is a ScFv (single chain variable fragment) and / or wherein the anti-Env-gp120 or anti-Env-gp140 is a VHH (Llama variable heavy-chain antibody fragment).

2. The nucleic acid of claim 1, comprising a sequence encoding for an anti-CD3 ScFv and an anti-Env-gp120 VHH.

3. The nucleic acid of claim 1, comprising a sequence encoding for an anti-CD3 ScFv and an anti-Env-gp140 VHH.

4. The nucleic acid of any one of claims 1 to 3, wherein said nucleic acid is RNA.

5. The nucleic acid of claim 4, wherein said nucleic acid is mRNA.

6. The nucleic acid of claim 5, wherein said mRNA comprises methylpseudouridines.

7. The nucleic acid of any one of claims 1 to 6, wherein said nucleic acid encodes for the bispecific antibody comprising SEQ ID NO: 3 or wherein said nucleic acid encodes for the bispecific antibody with SEQ ID NO: 1.

8. The nucleic acid of claim 7, comprising SEQ ID NO: 4 or comprising SEQ ID NO: 2.

9. A bispecific antibody encoded by the nucleic acid of any one of claims 1 to 8.

10. The bispecific antibody of claim 9, comprising SEQ ID NO: 3 or SEQ ID NO: 1.

11. The bispecific antibody of claim 10, consisting of SEQ ID NO: 1.

12. The nucleic acid of any one of claims 1 to 8 or the bispecific antibody of any one of claims 9 to 11 , for use in immunotherapy.

13. The nucleic acid of any one of claims 1 to 8 or the bispecific antibody of any one of claims 9 to 11, for use in immunotherapy against HIV.

14. The nucleic acid of any one of claims 1 to 8 or the bispecific antibody of any one of claims 9 to 11, for use in immunotherapy against HIV for killing HIV infected cells.

15. The nucleic acid of any one of claims 1 to 8 or the bispecific antibody of any one of claims 9 to 11, for use in immunotherapy against HIV for killing HIV latently infected cells.

Images