AIDS Res Hum Retroviruses. 2016 Feb;32(2):178-86. doi: 10.1089/AID.2015.0020. Epub 2015 Jun 22.
Full Length Single Chain Fc Protein (FLSC IgG1) as a Potent Antiviral Therapy Candidate: Implications for In Vivo Studies
Latinovic OS1,2, Medina-Moreno S1, Schneider K1, Gohain N1,3, Zapata JC1, Hippler LM1, Tagaya Y1,4, Pazgier M1,3, Reitz M1,4, Bryant J1,5, Redfield RR1,4
1 Institute of Human Virology, University of Maryland School of Medicine, Baltimore, Maryland.
2 Department of Microbiology and Immunology, University of Maryland School of Medicine, Baltimore, Maryland.
3 Department of Biochemistry and Molecular Biology, University of Maryland School of Medicine, Baltimore, Maryland.
4 School of Medicine, University of Maryland School of Medicine, Baltimore, Maryland.
5 Department of Pathology, University of Maryland School of Medicine, Baltimore, Maryland.
We have previously shown that FLSC, a chimeric protein containing HIV-1 BAL gp120 and the D1 and D2 domains of human CD4, blocks the binding and entry of HIV-1 into target cells by occluding CCR5, the major HIV-1 coreceptor. In an effort to improve the antiviral potential of FLSC, we fused it with the hinge-CH2-CH3 region of human IgG1. The IgG moiety should increase both the affinity and stability in vivo of FLSC, due to the resultant bivalency and an extended serum half-life, thereby increasing its antiviral potency. We previously showed that (FLSC) IgG1 indeed had greater antiviral activity against T cell infections. Here, we extend these results to macrophages, for which (FLSC) IgG1 has a more potent antiviral activity than FLSC alone, due in part to its higher binding affinity for CCR5. We also test both compounds in a relevant humanized mouse model and show that, as anticipated, the IgG1 moiety confers a greatly extended half-life. These data, taken together with previous results, suggest potential clinical utility for (FLSC) IgG1 and support further developmental work toward eventual clinical trials.
Confocal microscopy and flow cytometry: Binding competition of FLSC IgG1 and anti- HIV-1 monoclonal antibody (mAb) 17b for CCR5. We previously visualized binding of fusion protein FLSC IgG1 to CCR5 molecules on the target cell surface [Latinovic et al, 2014]  using super resolution microscopy. Here, we extend our visualization study on FLSC IgG1 and CCR5 binding with additional control experiments and with the use of anti- HIV-1 mAb 17b. mAb 17b recognizes the gp120 domain of HIV-1 envelope (Env) and binds within the coreceptor binding site (CoRBS), thus blocking the Env interaction with CCR5 . This extended study can give more insights into the binding activity of FLSC IgG1 to the target cell surface and its ability to specifically block HIV-1 Env-CCR5 interactions.
The HeLa derived, JC53 clonal cell line, utilized in these experiments expresses high CCR5 levels due to stable transfection of the parental HI-J clone with an expression vector for CCR5 (a kind gift from Dr. D. Kabat, OHSU, OR) . A fluorescent FLSC IgG1 complex was created by tagging the FLSC IgG1 molecule with fluorescent Alexa 488 using a Zenon Alexa 488 human IgG kit (Invitrogen). Cells (5X105) were incubated with FLSC IgG1 (20 µg/ml) at room temp for 30 mins, washed twice with 1% FBS-PBS (FCM buffer), then incubated with 5 µl of the Zenon Alexa 488 reagent for green fluorescence. Stained cells were washed twice with FCM buffer, and further incubated with 20 µg/ml PE mouse anti-human CCR5 antibody (clone 3A9, BD Biosciences) in the dark for 30 mins at room temperature. Cells were analyzed by both techniques (confocal microscopy and flow cytometry) on the same day. To analyze the inhibitory effect of 17b mAb (a kind gift from Dr. G. Lewis – IHV, UMB, MD) on FLSC IgG1, JC53 cells were simultaneously incubated with 17b mAb at 100 µg/ml, and with or without FLSC IgG1 (10 µg/ml) complexed with Zenon Alexa 488 per manufacturer’s instructions. Cell-associated fluorescence was visualized using a Zeiss META LSM 510 confocal microscope and ZEN 2012 lite imaging software (Zeiss), and quantified using a FACSAria II multi-color cell analyzer/sorter or a FACSCalibur flow cytometer (BD Biosciences). Flow cytometry data were further analyzed using the FlowJo software (Tree Star, Inc).
Figure 1. FLSC IgG1 binds to surface CCR5 on JC53 cells. Panel A shows co-localization of FLSC IgG1 Zenon Alexa 488 nm complex (green) and PE CCR5 (clone 3A9) antibody (565 nm, red) as visualized by confocal microscopy. Split images show JC 53 cells stained with FLSC IgG1 Zenon Alexa 488 nm complex (green- Inset A-a) and PE CCR5 (clone 3A9) antibody (565 nm, red- Inset A-b); Panel B shows a lack of staining by IgG control – Zenon Alexa 488 and PE isotype control antibody (as expected); Panel C shows staining by IgG control – Zenon Alexa 488 with PE CCR5 (clone 3A9) antibody (565 nm, red); Panel D shows JC53 cells simultaneously incubated with 17b mAb and FLSC IgG1 Zenon Alexa 488 complex and presents a control for the image in inset A-a. Size bar is 10 µm as indicated in Panel A. A-1, B-1, C-1, A-a-1, and D-1 show flow cytometric analyses of the corresponding samples using a FACSAria II flow analyzer/sorter. Log fluorescence is shown as 10N on all flow cytometry panels.
FLSC IgG1 binding to CCR5. To confirm our previous observations showing FLSC IgG1 binding to CCR5, we used multi-color confocal microscopy. FLSC IgG1 Zenon Alexa 488 complex (green) and the PE conjugated CCR5 (clone 3A9 ) antibody (red, 565 nm) used in this study bind to different sites on CCR5, allowing FLSC IgG1 to bind to CCR5 in the presence of 3A9. Confocal microscopy of JC53 cells stained with FLSC IgG1 Zenon Alexa 488 (green) and PE CCR5 (clone 3A9) antibody (red, 565 nm) showed a yellow ring image [Figure 1A], consistent with both stained proteins co-localizing and recognizing CCR5 on the cell surface. Flow cytometric analyses show that both proteins are indeed present on JC53 cells [Figure 1, Panel A-1]. Split images are shown in the inset of Panel A: Cells stained with FLSC IgG1 Zenon Alexa 488 complex are presented in image A-a (green); the same cells stained with PE CCR5 (clone 3A9) antibody (red) are presented in image A-b. Treatment with Zenon Alexa 488 by itself did not result in green fluorescence as expected [Figure 1B] or in yellow fluorescence in cells previously stained with PE CCR5 (clone 3A9) mAb [Figure 1C]. The corresponding flow cytometry data validated the results [Figure 1, Panels B-1 and C-1]. The image and corresponding histogram in Figures 1A-a and 1A-a-1 demonstrate confocal visualization and flow cytometry readings of staining with FLSC IgG1 Zenon Alexa 488 complex (green), and its binding to CCR5.
The image and histogram in Figures 1D and 1D-1 demonstrate an inhibitory effect of anti-HIV-1 CoRBS-specific mAb 17b  to the staining of cells by FLSC IgG1 Zenon Alexa 488. mAb 17b binds to FLSC IgG1 within the site which is required for its interaction with CCR5 and thus blocks its binding to target cells. Using ZEN 2012 lite analysis software (Carl Zeiss) we estimated the inhibition of staining (fluorescence) with FLSC IgG1 Zenon Alexa 488 by 17b mAb (against gp120) to be 80% on average. Collectively, these experiments confirmed that FLSC IgG1 recognizes cell-surface CCR5 and specifically binds within the sites involved in interaction of CCR5 with the CoRBS. An additional control experiment was conducted using HI-J HeLa cells that do not express CCR5 . As expected, we did not observe any fluorescence after staining the parental CCR5-negative HI-J HeLa clone with FLSC IgG1 Zenon Alexa 488 (green) (Figure 2A) and PE CCR5 (clone 3A9) antibody (red) (Figure 2B) using the same laser power as with CCR5-expressing JC53 cells. As a further control, JC53 cells were treated with 17b mAb and Zenon Alexa 488, without FLSC IgG1 (Figure 2C). The signal was low, comparable to that of cells treated with 17b mAb plus FLSC IgG1 Zenon Alexa 488 [Figure 1D] and that of untreated cells (autofluorescence/background control samples- data not shown). We performed the same set of imaging experiments in JC 57 cells, which express lower levels of CCR5. A similar trend was observed to that of JC 53 cells (data not shown).
Figure 2. Additional confocal experiments with HI-J (A and B) and JC53 (C) clones. Panel A shows a lack of fluorescence after staining HI-J cells with FLSC IgG1 Zenon Alexa 488 (green), with minor autofluorescence detected as green spots; Panel B shows lack of fluorescence after staining HI-J cells with PE CCR5 (clone 3A9) antibody (red). The HI-J clones do not express CCR5. Panel C shows JC53 cells stained with 17b mAb with Zenon Alexa 488 (no FLSC IgG1). The fluorescence signal was comparable to the sample with untreated cells as an autofluorescence/background control (data not shown). Size of the bar is 10 µm.
Fusion inhibitory activity of FLSC IgG1 and two CCR5 blockers
Figure 3. Kinetics of R5 HIV-1 entry after addition of FLSC IgG1, CCR5 mAbs and fusion inhibitor T-20.
We previously identified FLSC IgG1 as a more potent CCR5 blocker than some commonly used mAbs against CCR5 (N-terminus, ECL2, or multiepitope) [Latinovic et al, 2014; 2016] [1,5]. Figure 3 shows results from β lactamase (BlaM) virus-cell fusion assays , which uses HIV-1 pseudovirions bearing an R5 Env and a fusion marker protein, BlaM-Vpr . The mAbs used in this experiment bind to different domains of CCR5 (see color coded legend in Figure 3). Fusion inhibitor T-20 was the positive control. We believe that the slower kinetics of R5 HIV-1 fusion following addition of FLSC IgG1 compared to the CCR5 mAbs further supports the idea that FLSC IgG1 has greater access to CCR5 conformations permissive for HIV-1 binding. The data are normalized to the maximum fusion signal observed after 90 min and presented as a percentage for each compound. The results are given as average of triplicates with ±SD.
Figure 4. Competition between FLSC IgG1 and MAbs for CCR5 sites.
C2f Th/CCR5 cells, lacking CD4, were stained with FLSC IgG1 alone or in competition with CCR5 mAbs. Panel A shows cells treated with FLSC IgG1 Zenon Alexa 488, Panel B shows cells treated with ECL2 mAb and FLSC IgG1 Zenon Alexa 488, Panel C shows cells treated with the N-terminus antibody and FLSC IgG1 Zenon Alexa 488, and Panel D shows cells after combined and simultaneous treatment with both mAbs and FLSC IgG1 Zenon Alexa 488. Panel E shows the normalized relative staining intensity of CCR5 in samples pretreated or not with antibodies and then with FLSC IgG1 Zenon Alexa 488. Autofuorescence (AF) samples were an unstained autofluorescence control. Bar size is 10 µm.
Comparison of binding activity between FLSC IgG1 and two CCR5 blockers in CD4 lacking cells
We previously reported FLSC IgG1’s specificity for two sites of CCR5 [Latinovic et al, 2014, 2015] [1,7]. It appears to bind to at least two distinct CCR5 domains, as judged by its competition with mAbs against both the N-terminal and ECL2 domains [1,7]. Based on lack of complete fluorescence knock down upon treatment with both NT and ECL2 Abs, we speculated  that FLSC IgG1 may be specific for an alternate CCR5 conformation or CCR5 site which is not recognized by either N terminus and ECL2 mAb, or perhaps caused by its contact with cellular CD4 (e.g. displacement of the CD4 portion of FSLC IgG1 caused by cellular CD4). To evaluate the possibility of FLSC IgG1’s interactions with cellular CD4, we used CD4 receptor lacking cells C2fTh/CCR5  and tested binding of FLSC IgG1 alone and in presence of the same anti – CCR5 mAbs used in the previous experiments. C2f Th/CCR5 cells were preincubated (with antibodies against the CCR5 N-terminus (NT) (purified mouse anti-human CCR5, clone T21/8) and the ECL2 (mouse monoclonal IgG2b, clone 45531), alone or in combination, then incubated the cells with FLSC IgG1 conjugated with Zenon Alexa 488. Panel A in Figure 4 shows a sample treated with FLSC IgG1 Zenon Alexa 488 alone. There is clear inhibition of FLSC IgG1 binding to cells pretreated with ECL2 [Panel B] or N-terminus [Panel C] antibodies or a combination of both [Panel D], in contrast to that of cells not pre-treated [Panel A]. Panel E presents the quantitation of fluorescence in the same samples. The N-terminus antibody decreased FLSC IgG1 bound to CCR5 fluorescence by more than 40%; ECL2 antibody decreased the CCR5 fluorescence for more than 70% and their combination decreased the CCR5 fluorescence by 90+%. Cell autofluorescence was less than 10% [Panel E]. The results are given as average of ten images ± SD.
We observed in these experiments with cells lacking CD4 that the intensity of CCR5 staining was lower than in corresponding experiments with cell lines expressing CD4 and CCC5 (JC 53 and JC 57) . We also observed decreased binding of FLSC IgG1 to cells lacking CD4, C2fTh/CCr5, compared to CD4 containing cells studied previously . These data indicate that in addition to binding in a way that may require cellular CD4 (where displacement of FLSC IgG1’s CD4 part could happen), most likely FLSC IgG1 recognizes some alternate CCR5 confirmations that are present on the target cell surface, perhaps induced by contact with or proximity of CD4. This ongoing study requires more functional assay support and other evidence that will be presented in future reports on FLSC IgG1 binding activity.
The importance of this study. The new, additional experiments reported in this short study give clearer insight into the binding activity of the fusion protein, FLSC IgG1, and its fusion inhibitory activity in comparison with other CCR5 blockers. The study has direct relevance to our ongoing in vivo studies using FLSC IgG1 in HIV-1 infected humanized NSG mice.
- Latinovic O, Schneider K, Szimansky H, Lakowicz J, Heredia A, et al. (2014): Binding of fusion protein FLEC IgG1 to CCR5 is enhanced by CCR5 antagonist Maraviroc. Antiviral Research; 112: 80-90.
- Kwong PD, Wyatt R, Robinson J, Sweet RW, Sodorski J, Hendrickson WA (1998): Structure of an HIVgp120 envelope glycoprotein in complex with the CD4 receptor and a neutralizing human antibody. Nature; 393: 648-659.
- Platt EJ, Wehrly K, Kuhmann SE, Chesebro B, Kabat D (1998): Effects of CCR5 and CD4 cell surface concentrations on infections by macrophagetropic isolates of human immunodeficiency virus type 1. J Virol ; 72:2855-2864.
- Konigs C, Pustowka A, Irving J, Kessel C, Klich K, Wegner V, Rowley MJ, Mackey IR, Kreuz W, Griesinger C, Dietrich U (2007): Peptide mimotopes selected with HIV-1 blocking monoclonal antibodies against CCR5 represent motifs specific for HIV-1 entry, Immunol Cell Biol 85 (7): 511-7.
- Latinovic OS, Zhang J, Tagaya Y, DeVico AL, Fouts TR, Schneider K, Lakowicz JR, Heredia A and Redfield RR (2016): Synergistic inhibition of R5 HIV-1 by Maraviroc and FLSC IgG in primary cells: Implications for prevention and treatment, Current HIV Research- Volume 14, Number 1.
- Cavrois M, De Noronha C, Green WC (2002): A sensitive and specific enzyme-based assay detecting HIV-1 virion fusion in primary T lymphocytes. Nat Biotechnology; 20: 1151-1154.
- Latinovic OS, Medina-Moreno S, Schneider K, Gohain N, Zapata JC, Pazgier M, Reitz M, Bryant J, Redfield RR (2015): Full length single chain Fc protein (FLSC IgG1) as a potent anti-viral therapy candidate: Implications for in vivo studies, AIDS Research and Human Retroviruses- Volume 31, Number 00.
- Y. Guan, Sajadi MM, Kamin-Lewis R, Fouts TR, Dimitrov A, Zhang Z, Redfield RR, DeVico AL, Gallo RC, Lewis GK (2009): Discordant memory B cell and circulating anti-Env antibody reposnes in HIV-1 infection. PNAS; Vol 106. No. 10, p. 3952-3957.
Olga S. Latinovic, PhD, MSc
Institute of Human Virology
725 W. Lombard St.
Baltimore, MD 21209
410 706 2769 / email: firstname.lastname@example.org