Rational Design and Characterization of Trispecific Antibodies Targeting the HIV-1 Receptor and Envelope Glycoprotein

Multitudinous broadly neutralizing antibodies (bNAbs) against HIV-1 have been developed as novel antiviral prophylactic and therapeutic agents. Combinations of bNAbs are generally even more effective than when they are applied individually, showing excellent neutralization coverage and limiting the emergence of escape mutants. In this study, we investigated the design and characterization of three trispecific antibodies that allow a single molecule to interact with independent HIV-1 envelope determinants—(1) the host receptor CD4, (2) the host co-receptor CCR5 and (3) distinct domains in the envelope glycoprotein of HIV-1—using an ELISA, an HIV-1 pseudovirus neutralization assay and in vivo antiviral experiments in humanized mice. We found that trispecific bNAbs and monovalent ones all had satisfactory binding activities against the corresponding antigens in the ELISA, exhibited higher potency and breadth than any previously described single bnAb in the HIV-1 pseudovirus neutralization assay and showed an excellent antiviral effect in vivo. The trispecific antibodies simultaneously recognize the host receptor CD4, host co-receptor CCR5 and HIV-1 envelope glycoprotein, which could mean they have promise as prophylactic and therapeutic agents against HIV-1.


Introduction
Due to the development of multiple antiviral drugs, HIV-1 has become a treatable disease with near-to-normal life expectancy; however, better tolerated antiviral agents with superb efficacy, few side effects and the ability to limit escape viruses still need to be developed [1].Anti-HIV-1 drugs cannot only be used for post-viral infection treatment but also the pre-exposure prevention of HIV-1.Currently, the primary pre-exposure prophylactic drugs for HIV-1 that have been marketed are Tenofovir Alafenamide (TAF) [2,3] and Dapivirine [4,5].However, these drugs may have some side effects, and there are also some HIV-1-resistant strains [6,7].Developing new and practical pre-exposure prophylactic drugs against HIV-1 can help humans more effectively prevent HIV-1 infection.A variety of broadly neutralizing antibodies (bnAbs) have been generated with broad neutralization spectra and excellent antiviral activities in recent years [8][9][10], but their abilities to treat HIV-1 patients may be limited by the breadth and potency of their viral neutralization [11,12].HIV-1 bnAbs have been characterized to recognize several conserved regions on the HIV-1 envelope glycoprotein (Env); these include the CD4 binding site (e.g., N6 [13]), the N332 glycan supersite (e.g., PGT121) [14], the membrane-proximal external region of gp41 (e.g., 10E8 [15]) and the V2 apex (e.g., PGDM1400) [16].In addition, the host receptor and co-receptor of HIV-1 are also important targets for bnAbs because the receptor proteins do not mutate as a virus does-the antibodies that target the receptor possess extraordinary neutralization breadth and efficacy against diverse HIV-1 strains and inhibit the virus from entering host cells [17].
iMab, an anti-CD4 antibody, has shown a broad spectrum of antiviral effects and prevents HIV-1 infection in a non-competitive manner [18]; another antibody, PRO140, exerts effective antiviral potency by inhibiting the binding between HIV-1 and the host coreceptor CCR5 [19].The HIV-1 virus can generate escape mutants rapidly under pressure from a single bNAb; while antibody cocktails have shown improved antiviral efficacy and breadth in preclinical studies [20], multispecific "single agents" are desirable for enhancing neutralization breadth and potency [17,21,22].bNAbs, which recognize the CD4 binding site, variable region glycans and MPER, have been engineered into various multispecific antibodies and provide optimal neutralization [22,23].However, to date, no trispecific antibodies that simultaneously recognize the host CD4 receptor, CCR5 receptor and HIV-1 envelope glycoprotein have been developed; this study aims to explore the antiviral activity and breadth of this class of trispecific antibodies to provide a supplement for the development of novel antiviral drugs for HIV-1.

Expression of Recombinant Human CD4 and CCR5 Proteins
Recombinant CD4 protein (aa 26-396, detailed sequences are provided in the Supplementary Materials) and CCR5 protein (aa 1-352, detailed sequences are provided in the Supplementary Materials) were expressed in human HEK293F cells as soluble proteins.The soluble human CD4 and CCR5 proteins were purified using a HisTrap HP column (GE Healthcare, Chicago, United States) and were further purified via size-exclusion chromatography with a Superdex 200 column (GE Healthcare).Then, the purified proteins were dialyzed in 1×PBS dialysate, and the dialysate was replaced at 4 h intervals.Finally, the dialyzed protein was concentrated to 1 mg/mL at 4 • C with an ultrafiltration tube, and samples were analyzed via SDS-PAGE and stored at −80 • C until use.

Construction, Expression, and Purification of IgG and Trispecific Antibodies
iMab, PRO140, 10E8, PGT121 and PGDM1400 were cloned as human IgG1 antibodies in cytomegalovirus (CMV)-driven expression plasmids [24].Trispecific molecules were engineered in the DVD-Ig format, and the sequences for two scFvs were cloned in frame with sequences encoding connecting G4S linkers on both the N and C termini of the full IgG1 antibody.Specifically, the iMab variable domain of the heavy chain and PRO140 variable domain of the heavy chain were fused with a GGGGSGGGGS linker, followed by a constant region (CH1-CH2-CH3); then, each ScFv from 10E8, PGDM1400 and PGT121 was connected to the C terminus of the CH3 via a GGGGSGGGGS linker in the heavy chain.For the light chain of the trispecific antibodies, the iMab variable domain and PRO140 variable domain were connected via the GGGGSGGGGS linker, followed by a constant region (CL).The plasmids encoding the heavy chain and light chain were cotransfected into HEK293F cells with a 1:1.5 molar ratio of the heavy chain and the light chain.Cells were cultured with SMM 293-TII medium (Sinobiological, Beijing, China) at 310 K and 5% CO 2 ; then, the supernatants were collected on the fifth day, filtered through a 0.22 µm membrane, purified via protein A and then further purified via size-exclusion chromatography in PBS buffer (pH 7.4).The small-scale production of transfected HEK293F cell culture of the three trispecific antibodies iMab+PRO140+10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 is 5-10 mg/liter.The associated antibody amino acid sequences are presented in the Supplementary Materials.

ELISA
In the next step, 96-well microtiter plates were coated with 100 ng/well-purified proteins or viruses overnight at 4 • C. The plates were washed thrice with PBST (1 × PBS containing 0.1% v/v Tween-20) and blocked with 200 µL blocking buffer (1 × PBS containing 2% w/v non-fat dry milk) at 37 • C for 2 h.The microtiter plates were then washed once with 1*PBST and shaken dry.Serial dilutions of purified HIV-1 neutralizing monoclonal antibodies were added to the wells at 100 µL/well and incubated for 30 min at 37 • C. Following that, 100 µL of horseradish peroxidase (HRP)-conjugated goat anti-human IgG antibody solution was added to each well after three washes, and incubated at 37 • C for 30 min.Then, 100 µL of tetramethylbenzidine (TMB) substrate was added to each well after five washes, at room temperature in the dark.The reaction was stopped with a 2 M H 2 SO 4 solution after 15 min, and the absorbance was measured at 450 nm.All samples were run in triplicate.The relative affinity of antibody binding to purified viruses or proteins was determined by measuring the concentration of antibody required to achieve the EC 50 .

K d Determination
HIV-1 proteins were amine coupled to a carboxylated dextran polymer matrix, which was coated on CM-5 sensor chips for use in the BIAcore 3000.The carboxyl groups on the chips were activated with 35 µL of a 1:1 (v/v) solution of N-hydroxysuccinimide and Nethyl-N-(3-diethylaminopropyl) carbodiimide for 7 min.HIV-1 proteins at a concentration of 50 µg/mL were prepared using 20 mM sodium acetate as a diluent and each coated onto one cell, whereas the other cell was left blank as a reference and blocked with 1 M ethanolamine HCl, pH 8.5 (GE Healthcare, Chicago, United States) as a control.The affinity measurements for antibody binding with different antigens were started by allowing HBS (10 mM HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid), 150 mMNaCl and 3 mM EDTA) to flow over the sensor surface at 30 µL/minute for 100 s of baseline acquisition, followed by injection of each serially diluted antibody at 30 µL/minute for approximately 100 s.The results were used to fine-tune the affinity fit.The ratio of k on to k off determined the K d reported here.

Virus Neutralization Assay
HIV-1 viral neutralization by the neutralizing antibodies was assessed using a singlecycle assay with huCD4 + CCR5 + TZM-BL target cells and HIV-1 pseudoviruses, as described previously [21].The parental antibodies or the trispecific antibodies were serially diluted and incubated with the HIV-1 pseudoviruses to determine the IC 50 values of each antibody, which reflect the neutralization of HIV-1.The IC 50 values were determined using the GraphPad Prism 6.0 software by fitting the neutralization dose-response curves using nonlinear regression.

Humanized Mouse Model
Animal procedures that might cause more than slight pain or distress were performed with appropriate anesthesia.Immunodeficient NOG mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd (Beijing, China).The 4-to 6-week-old NOG mice were humanized as previously described [25].

HIV-1 RNA Load Measurement
HIV-1 RNA was extracted using the QIAamp viral RNA mini kit and reverse transcribed to cDNA with the RT-PCR Prime Script Kit from Takara.Then, 2 µL cDNA was used in a 20 µL qRT-PCR reaction using the Takara PCR Master Mix with the specific HIV-1 P17 gene primers (5 ′ -TCTCGA CGCAGG ACTCG-3 ′ and 5 ′ -TACTGA CGCTCT CGCACC-3 ′ ) and a TaqMan probe (5 ′ -FAM-CTCTCT CCTTCT AGCCTC-MGB-3 ′ ), in the following conditions: 1 cycle of 50 • C for 2 min, 1 cycle of 95 • C for 10 min and 40 cycles of 95 • C for 15 s and 60 • C for 1 min.The HIV-1 viral titer was determined via comparison with a standard curve generated using RNA extracted from a serially diluted reference HIV-1 viral stock.

In Vivo Assay in Humanized Mice
All animal experimental procedures were approved by the Committee on the Use of Live Animals in Shenzhen Third People's Hospital.The t 1/2 of the three trispecific antibodies was computed as ln (2)/k, where k is a rate constant expressed reciprocally of the x-axis time units by the 1-phase decay equation in the GraphPad software (version 6.0).One day before the HIV-1 challenge, 200 µg (10 mg/kg) or 20 µg (1 mg/kg) of each antibody was injected intraperitoneally (i.p.) into NOG mice; after 24 h of antibody administration, the mice were challenged through the i.p. route with 500 TCID 50 live HIV-Shenzhen-BC .These mice were subjected to daily blood sampling to monitor viral load for 14 days.

Antibody-Dependent Cell-Mediated Cytotoxicity (ADCC) Assay
ADCC was determined using a previously described assay utilizing two fluorescent dyes to distinguish between viable and dead cells [24].TZM-BL cells were labeled with a general cell membrane labeling dye PKH67 (Sigma-Aldrich, Darmstadt, Germany).The cells were then washed twice with PBS, added to 96-well plates in 5 × 10 4 cells per well in triplicate and incubated with serial dilutions of single antibodies (0.5-20 mg/µL) for 30 min at 37 • C. NK cells (natural killer cells) isolated from human PBMCs were added to the plates (1 × 10 6 per well), incubated at 37 • C for 3 h and treated with 1 µL of 7-AAD (eBioscience, California, United States) to stain dead cells.Cell death was determined using a FACSAria III flow cytometer with BD FACS Diva software (version 6.1.2).

Determination of Antigen-Binding Activity of the Antibodies Generated in This Study
The monoclonal antibodies and their recognition targets involved in this study are shown in Table S1.Through HEK293F expression and Protein A purification, we obtained highly purified monoclonal antibodies iMab, PRO140, 10E8, PGT121, PGDM1400 and VRC01; trispecific antibodies iMab + PRO140 + 10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 were expressed and purified (Figure S1).First, the binding activity of the expressed antibodies against the corresponding antigens was evaluated through the ELISA experiment.CD4 receptor protein, CCR5 co-receptor protein, iMab and PRO140 expressed by eukaryotic cells showed good activity.The EC 50 value of iMab for the CD4 receptor protein was 17 ng/mL (Figure 1A), and the EC 50 value of PRO140 for the CCR5 co-receptor was 92 ng/mL (Figure 1B).The EC 50 values of the 10E8 antibody against gp41 protein of HIV-1 subtype A, B, C and AC strains were 1145, 381, 306 and 434 ng/mL, respectively (Figure 1C).PGDM1400 and PGT121 showed better binding activity to HIV-1 antigen than that of 10E8.The EC 50 values of PGDM for HIV-1 A, B, C and AC subtypes were 33, 64, 23 and 294 ng/mL, respectively (Figure 1D), while the corresponding values of PGT121 were 57, 69, 115 and 51 ng/mL, respectively (Figure 1E).After characterization of the activity of a single antibody, the antibodies were engineered into three trispecific antibodies in the form shown in Figure 1F: iMab + PRO140 + 10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121, respectively.The three trispecific antibodies maintained favorable binding activity to CD4 receptor and CCR5 co-receptor and showed good binding activity to different subtypes of HIV-1 antigens (Figure 1G-I).In addition, the binding of trispecific antibody iMab + PRO140 + 10E8 to HIV-1 gp41 protein, as well as of trispecific antibodies iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 to HIV-1 gp120 protein, was measured via surface plasmon resonance, with the HIV-1 proteins being firmly bound by the trispecific antibodies: K d (dissociation constant) = 6.12 nM for iMab + PRO140 + 10E8 versus HIV-1 gp41 protein; K d = 0.46 nM for iMab + PRO140 + PGMD1400 versus HIV-1 gp120; K d = 0.59 nM for iMab + PRO140 + PGT121 versus gp120 (Figure S2).

Determination of In Vitro Neutralization Activity against HIV-1 Pseudoviruses and In Viv Antiviral Activity of the Antibodies Generated in This Study
In further experiments, the neutralizing activities of the monoclonal antibodies, an body cocktails and three trispecific antibodies against different subtypes of HIV-1 pse dovirus were evaluated.A total of 23 pseudoviruses of HIV-1 were selected in the ne tralization assay in this study, including HIV-1 subtype A, subtype B, subtype C, subty G, subtype AE, subtype AG, subtype BC and subtype AC.All five single monoclonal a tibodies showed broad-spectrum and potent activity in neutralizing the HIV-1 viru However, several sporadic virus strains can escape neutralization by a single antibod iMab and 10E8 could not neutralize the X1632 strain from HIV-1 subtype G, wh KER2008.12 from subtype A and T26-60 from subtype AG escaped the neutralization fro 10E8.PRO140 was unable to neutralize the BJOX2000 and XJN0181ENV04 strains fro subtype BC, PGDM1400 failed to effectively neutralize one subtype B and one subtype HIV-1 strain, and PGT121 and the control antibody VRC01 did not show neutralizing a tivity against four of the HIV-1 strains (Figure 2).In contrast, the three-antibody cockta showed a more robust and broader spectrum of virus-neutralizing activity than a sing

Determination of In Vitro Neutralization Activity against HIV-1 Pseudoviruses and In Vivo Antiviral Activity of the Antibodies Generated in This Study
In further experiments, the neutralizing activities of the monoclonal antibodies, antibody cocktails and three trispecific antibodies against different subtypes of HIV-1 pseudovirus were evaluated.A total of 23 pseudoviruses of HIV-1 were selected in the neutralization assay in this study, including HIV-1 subtype A, subtype B, subtype C, subtype G, subtype AE, subtype AG, subtype BC and subtype AC.All five single monoclonal antibodies showed broad-spectrum and potent activity in neutralizing the HIV-1 virus.However, several sporadic virus strains can escape neutralization by a single antibody.iMab and 10E8 could not neutralize the X1632 strain from HIV-1 subtype G, while KER2008.12 from subtype A and T26-60 from subtype AG escaped the neutralization from 10E8.PRO140 was unable to neutralize the BJOX2000 and XJN0181ENV04 strains from subtype BC, PGDM1400 failed to effectively neutralize one subtype B and one subtype C HIV-1 strain, and PGT121 and the control antibody VRC01 did not show neutralizing activity against four of the HIV-1 strains (Figure 2).In contrast, the three-antibody cocktails showed a more robust and broader spectrum of virus-neutralizing activity than a single antibody.The three cocktails neutralized all viruses, and the trispecific form of antibodies showed the best antiviral activity and reactive spectrum (Figure 2).cines 2024, 12, x FOR PEER REVIEW 6 of antibody.The three cocktails neutralized all viruses, and the trispecific form of antibodie showed the best antiviral activity and reactive spectrum (Figure 2).The median IC50 neutralizing titers of iMab, PRO140, 10E8, PGMD1400 and PGT12 against all HIV-1 viruses were 1.38, 3.34, 3.18, 2.06 and 3.78 μg/mL, respectively; the ant viral mean IC50 titers of the iMab+PRO140+10E8, iMab+PRO140+PGMD1400 an iMab+PRO140+PGT121 cocktails against all HIV-1 viruses were 0.24, 0.15 and 0.18 μg/m while the values of the trispecific antibodies iMab+PRO140+10E iMab+PRO140+PGMD1400 and iMab+PRO140+PGT121 were 0.12, 0.05 and 0.08 μg/m respectively (Figure 3).The IC50 values of trispecific antibodies iMab+PRO140+10E iMab+PRO140+PGMD1400 and iMab+PRO140+PGT121 against the HIV-1 strains showe statistically significant differences compared to the IC50 values of monoclonal antibodi iMab, PRO140, 10E8, PGMD1400 and PGT121; in contrast, the IC50 values of the three coc tail antibodies showed significant differences compared to the values of iMab, PRO14 10E8 and PGMD1400, while they did not show significant increases when compared that of PGT121; the IC50 values of the trispecific antibodies iMab+PRO140+PGDM1400 an iMab+PRO140+PGT121 were significantly higher than those values of their correspondin parental antibody cocktail combinations (Table S2).The median IC 50 neutralizing titers of iMab, PRO140, 10E8, PGMD1400 and PGT121 against all HIV-1 viruses were 1.38, 3.34, 3.18, 2.06 and 3.78 µg/mL, respectively; the antiviral mean IC 50 titers of the iMab + PRO140 + 10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 cocktails against all HIV-1 viruses were 0.24, 0.15 and 0.18 µg/mL, while the values of the trispecific antibodies iMab + PRO140 + 10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 were 0.12, 0.05 and 0.08 µg/mL, respectively (Figure 3).The IC 50 values of trispecific antibodies iMab + PRO140 + 10E8, iMab+PRO140+PGMD1400 and iMab + PRO140 + PGT121 against the HIV-1 strains showed statistically significant differences compared to the IC 50 values of monoclonal antibodies iMab, PRO140, 10E8, PGMD1400 and PGT121; in contrast, the IC 50 values of the three cocktail antibodies showed significant differences compared to the values of iMab, PRO140, 10E8 and PGMD1400, while they did not show significant increases when compared to that of PGT121; the IC 50 values of the trispecific antibodies iMab + PRO140 + PGDM1400 and iMab + PRO140 + PGT121 were significantly higher than those values of their corresponding parental antibody cocktail combinations (Table S2).
The dual specificity against CD4 and CCR5 could lead to CD 4+ T cell depletion in vivo due to several mechanisms such as ADCC; the trispecific antibodies could bind CD4 + CCR5 + T cells at low concentrations and then recruit effector cells such as NK cells that express the Fc receptor to kill the CD4+CCR5+ T cells.To address this concern, we performed in vitro assays to test the ADCC effect of the trispecific antibodies on TZM-BL cells as target cells and human NK cells as effector cells in the presence of trispecific antibodies.The results showed that the trispecific antibodies iMab + PRO140 + 10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 showed no ADCC killing activity on TZM-BL cells (Figure S3).We then measured the pharmacokinetics and bioavailabilities of iMab + PRO140 + 10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 in five healthy NSG-HuPBL mice (2 male (M) and 2 female (F)) and found that the calculated peripheral half-lives (t 1/2 ) of iMab + PRO140 + 10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 were 5.279, 4.355 and 4.378 days, respectively (Figure 4A).These results showed that trispecific iMab + PRO140 + 10E8, iMab + PRO140 + PGMD1400 and iMab + PRO140 + PGT121 were stable in vivo.Finally, the in vivo protective activity of three trispecific antibodies was tested in NOG humanized mice, which were given the antibodies one day before HIV-1 infection.The results showed that HIV titers in the plasma of mice in the antibody-negative control group gradually increased during the two weeks of HIV-1 infection; in contrast, when the antibody was administered at a dose of 10 mg/kg, no HIV-1 RNA was detected in the plasma of mice treated with the three trispecific antibodies (Figure 4B).When the antibody was administered at 1 mg/kg, the HIV-1 virus could be detected on the third day after the infection of the mice in the iMab + PRO140 + 10E8-treated group, and the virus titer gradually increased during the observation period, but the virus titer of the mice treated by the iMAB + PRO140 + 10E8 group was significantly reduced compared with the negative antibody control group.Encouragingly, no HIV-1 virus was detected in mice treated with 1 mg/kg of iMab + PRO140 + PGMD1400 or iMab + PRO140 + PGT121 (Figure 4C).The dual specificity against CD4 and CCR5 could lead to CD 4+ T cell depletion in vivo due to several mechanisms such as ADCC; the trispecific antibodies could bind CD4 + CCR5 + T cells at low concentrations and then recruit effector cells such as NK cells that express the Fc receptor to kill the CD4+CCR5+ T cells.To address this concern, we per- Encouragingly, no HIV-1 virus was detected in mice treated with 1 mg/kg of iMab+PRO140+PGMD1400 or iMab+PRO140+PGT121 (Figure 4C).

Figure 2 .
Figure 2. The neutralization breadth of the parental and trispecific antibodies was tested against a HIV-1 pseudovirus panel consisting of Envs of 23 viral strains.Heat maps of IC50 titers were gene ated in GraphPad Prism.In the heatmaps, each row represents a virus strain while columns repr sent antibodies.Red colors indicate more potent neutralization, and green indicates lower potenc an antibody virus neutralization titer more significant than 20 μg/mL is below the detection thres old.The median IC50 value for each antibody is calculated and placed in the last row.

Figure 2 .
Figure 2. The neutralization breadth of the parental and trispecific antibodies was tested against an HIV-1 pseudovirus panel consisting of Envs of 23 viral strains.Heat maps of IC 50 titers were generated in GraphPad Prism.In the heatmaps, each row represents a virus strain while columns represent antibodies.Red colors indicate more potent neutralization, and green indicates lower potency; an antibody virus neutralization titer more significant than 20 µg/mL is below the detection threshold.The median IC 50 value for each antibody is calculated and placed in the last row.

Figure 3 .
Figure 3. Analysis of IC50 values for each antibody from three independent neutralization experiments.IC50 values for each antibody against each virus are plotted as symbols.The statistical differences among the groups are shown in Table S1.IC50 values less than 20 are represented by the experimentally obtained IC50 value; IC50 values greater than 20 are uniformly represented by 20.

Figure 3 .
Figure 3. Analysis of IC 50 values for each antibody from three independent neutralization experiments.IC 50 values for each antibody against each virus are plotted as symbols.The statistical differences among the groups are shown in Table S1.IC 50 values less than 20 are represented by the experimentally obtained IC50 value; IC 50 values greater than 20 are uniformly represented by 20.