Available online at Minireview

Available online at www.sciencedirect.com Virology 375 (2008) 315 320 www.elsevier.com/locate/yviro Minireview Recent advances in the characterization of HIV-1 neutralization assays for standardized evaluation of the antibody response to infection and vaccination Victoria R. Polonis a,, Bruce K. Brown b, Andrew Rosa Borges b, Susan Zolla-Pazner c, Dimiter S. Dimitrov d, Mei-Yun Zhang d, Susan W. Barnett e, Ruth M. Ruprecht f, Gabriella Scarlatti g, Eva-Maria Fenyö h, David C. Montefiori i, Francine E. McCutchan b, Nelson L. Michael a a Walter Reed Army Institute of Research, Washington, DC, USA b The Henry M. Jackson Foundation, Rockville, MD, USA c The VA Hospital and New York University School of Medicine, New York, NY, USA d National Cancer Institute (NCI), US NIH, Frederick, MD, USA e Novartis Vaccines and Diagnostics Inc., Emeryville, CA, USA f Dana Farber Cancer Institute and Harvard Medical School, Boston, MA, USA g San Raffaele Scientific Institute, Milan, IT, Italy h Lund University, Lund, Sweden i Duke University, Durham, NC, USA Received 4 December 2007; returned to author for revision 30 January 2008; accepted 11 February 2008 Available online 25 March 2008 Abstract In AIDS vaccine development the pendulum has swung towards a renewed emphasis on the potential role for neutralizing antibodies in a successful global vaccine. It is recognized that vaccine-induced antibody performance, as assessed in the available neutralization assays, may well serve as a gatekeeper for HIV-1 subunit vaccine prioritization and advancement. As a result, development of a standardized platform for reproducible measurement of neutralizing antibodies has received considerable attention. Here we review current advancements in our knowledge of the performance of different types of antibodies in a traditional primary cell neutralization assay and the newer, more standardized TZM-bl reporter cell line assay. In light of recently revealed differences (see accompanying article) in the results obtained in these two neutralization formats, parallel evaluation with both platforms should be contemplated as an interim solution until a better understanding of immune correlates of protection is achieved. Published by Elsevier Inc. Keywords: Human immunodeficiency virus type 1 (HIV-1); Antibodies; Neutralization assays; Pseudovirus; Vaccine immune assessment; Standardization Historically, vaccines that induce antibodies have been the most effective strategy to combat viral diseases such as polio, hepatitis, measles, and influenza. While antibodies are known to play an important role in protection in these diseases, the importance of antibodies in human immunodeficiency virus type 1 (HIV-1) protection and pathogenesis remains to be further defined (Huber and Trkola, 2007; McMichael, 2006; Srivastava Corresponding author. WRAIR, 13 Taft Court, Rockville, MD 20850, USA. Fax: +301 762 4177. E-mail address: vpolonis@hivresearch.org (V.R. Polonis). et al., 2005; Zolla-Pazner, 2004). Despite this and because of knowledge gained from other successful vaccines, the design of a vaccine that will elicit antibody responses capable of neutralizing primary isolates of multiple HIV clades continues to be an important goal in laboratories that are developing and testing vaccines. It has been proposed that, although the virus envelope (Env) proteins have evolved an extraordinary ability to evade neutralizing antibodies, a vaccine that can elicit protective antibodies remains the best hope for developing an HIV vaccine that confers sterilizing immunity. Vaccine strategies that exclusively stimulate T-cell immunity may at best generate 0042-6822/$ - see front matter. Published by Elsevier Inc. doi:10.1016/j.virol.2008.02.007

316 Minireview persistent and broadly reactive T-cell responses that can suppress virus and limit damage caused by the virus, without preventing infection (McMichael, 2006). In light of recent developments in HIV vaccine research (Cohen, 2007), humoral responses may be a critical component of an efficacious product. The challenge to develop an immunogen that elicits antibodies effective against the genetically diverse strains of the HIV-1 pandemic is defined not only by the need to elucidate the appropriate Env structure(s) and identify epitopes that induce protective antibodies, but also by the requirement for standardized in vitro assays that will allow for a meaningful comparison of the quality and potency of neutralizing antibodies in sera or other fluids from HIV-positive patients and vaccine recipients. To support the evaluation of phase I, II or I/II HIV vaccine clinical trials that are being conducted to test new immunogens, it will be important to utilize high throughput, validated neutralization assays. In the context of analyzing new candidate vaccines or new monoclonal antibodies, the relevant in vitro measurement of neutralizing antibodies becomes a complex task, largely due to a long list of confounding variables attributable to the virus, the antibodies and the cells (Zolla-Pazner, 1996). Importantly, the neutralization assay outcome may be determined by numerous assay parameters. These variable parameters include: target cell used and cell density, host cell used for viral stock propagation, virus dose and antibody dilution/concentration (virus particle: antibody ratio), the inclusion of complement, volumes of components added, duration of pre-incubation of virus and antibody, duration of infection with or without antibody, cell washing steps to remove unbound antibody and virus, length of culture time, the endpoint measured and other variables. To address specific aspects of these variables, several laboratories have produced panels of virologic and serologic reagents from HIV-1-positive patients, drawn from multiple geographic regions, to model some of the extensive genetic variation of HIV-1. Neutralization platforms employing these reagents have been evolving over two decades. In the majority of neutralization assays, virus and antibody are incubated together and then added to CD4+ target cells. The early neutralization assays relied on the ability of T-cell line-adapted (TCLA) viruses to infect cell lines; viral proteins, multinucleated giant cells (syncytia), or cell survival could then be quantified to measure reduction of infection. In general, neutralization of TCLA viruses proved poorly predictive of primary isolate neutralization and was largely excluded as a gatekeeper assay. The critical concern regarding the use of any cell line-based model system is its physiologic relevance and value as a surrogate for in vivo outcomes. Subsequently, assays were developed so that patient viruses (primary or clinical isolates) could be used to infect HIVseronegative peripheral blood mononuclear cells (PBMC) from healthy human donors, an approach thought to be more physiologic. These primary cell target assays require the use of donor PBMC from different individuals; the freshly isolated or frozen PBMC are then stimulated with phytohemagglutinin (PHA) and cultured with interleukin-2 (IL-2) before infection in the presence or absence of test antibodies. It has long been known that PBMC from different donors display differential susceptibility to HIV-1 infection. These differences in HIV replication may be due, in part, to host genetics and CD8+ cell factors, the number of CD4+ cells or expression levels of CD4 molecules on host cells, effects of host cell-derived molecules on the viral surface, and host genetic polymorphisms in chemokines or chemokine receptors, such as CCR5, which function as HIV-1 coreceptors. Because of these variables, the inter-lab, and even the inter-experiment variation within a single lab, is often quite problematic in PBMC-based neutralization assays. Prior to the development of PBMC assays, early efforts to standardize and compare neutralization assays were initiated by D'Souza et al. at the US NIH Division of AIDS, together with investigators from the World Health Organization (WHO). A large study (as part of the Antibody Serologic Project) was conducted in 1990 91 by over 40 investigators from 25 organizations in 7 countries to compare assays for the evaluation of HIV-1 monoclonal antibodies (mabs). Later, a second mab study was performed comparing PBMC assays (D'Souza et al., 1991; D'Souza et al., 1997). While considerable variations in the protocols and the data were observed, these comparative studies were very informative for the characterization of the potential clinical utility of the mabs tested. These studies were also pioneering efforts directed towards the generation of standardized approaches to neutralization assessment. Recognizing that the issue of standardization of immunomonitoring assays is critically important, several investigators from a number of organizations around the world have subsequently convened meetings and initiated studies to help identify and reduce the variability across the assays currently in use. In 2003, a meeting organized by WHO UNAIDS was held in Milan, Italy to discuss these issues. As a result of this meeting, a study entitled NeutNet was initiated and funded through the European Commission for HIV Vaccine Research. The NeutNet study was conducted from 2005 to 2006 and involved international laboratories from several countries. Employing a variety of assay formats, study participants assessed four test reagents against a panel of eleven HIV-1 isolates or a sub-panel of pseudoviruses. While there was some degree of consistency in certain data elements, there was also considerable variation in the results observed (Scarlatti et al., manuscript in preparation). In addition, in 2005, a standardized approach to the measurement of neutralizing antibodies for vaccine trials was proposed as a result of meetings convened by the Laboratory Standardization Subcommittee for the Global HIV/AIDS Vaccine Enterprise (GHAVE) (Mascola et al., 2005). This standardization of approaches to monitor the function(s) of vaccine-induced antibodies is reflected in the methodologies currently being employed and investigated through the efforts of the co-investigators of the Comprehensive Antibody Vaccine Immuno-Monitoring Consortium (CA-VIMC), centered at Duke University Medical Center and funded by The Bill and Melinda Gates Foundation in July of 2006. A global workshop, entitled Standardisation of HIV Neutralisation Assays for Use in Vaccine Research and Clinical Trials, was convened on March 17 18, 2007 in Varese, Italy, by members of the WHO and the European Commission for HIV Vaccine Research (Meeting report in preparation). A recent summary meeting on Humoral Immune Responses to HIVand Approaches to Design

Minireview 317 Antigens that Induce Neutralizing and Other Potentially Protective Antibodies was also sponsored by the GHAVE in Reston, VA, USA on May 14 15, 2007 (Montefiori et al., 2007). Thus, significant levels of effort and resources have now been committed to the evaluation and implementation of platforms for assessing vaccine-induced antibodies. The newer technologic approaches exploit the use of pseudoviruses that incorporate molecularly cloned HIV-1 Envs into defective virus particles capable of only a single round of infection. The pseudoviruses are generated in the 293T cell line by cotransfection of an env-mutated viral backbone, along with the env clone of choice, and then used to infect a transformed cell line expressing the appropriate receptors. One such format is an assay that employs an epithelial HeLa-derived cell line (TZM-bl) that carries the luciferase reporter gene sensitive to the presence of the HIV Tat protein. This TZM-bl assay has been advanced as a readily transferable method for assessing vaccine-elicited neutralizing antibodies in the good clinical laboratory practices (GCLP) environment. Table 1 shows a comparison of several of the variable parameters (as described above) that distinguish the PBMC assay from the pseudovirus approach. The cell target, viral inocula, and culture conditions are very different in the two assay systems. The pseudoviral system has several advantages as a platform for application to front-line assessment of antibodies, to include: the ability to rapidly test for neutralization against primary patient Envs from multiple clades, a high degree of inter-experiment reproducibility and throughput, ease and safety of reagent distribution for the assay, and facilitation of assay validation and global transfer (note Table 1). However, specific discrepancies in the data obtained when reporter cell line-based pseudovirus assays are compared with PBMC-based assays, have recently been reported. Binley et al. have demonstrated that the 4E10 gp41 mab is broadly neutralizing in a pseudovirus system, but shows only moderate activity in the PBMC assay. In contrast, in the same report it was shown that the X5 gp120 mab neutralizes considerably better in the PBMC assay (Binley et al., 2004). A study by Choudhry et al., using the m46 gp41 mab derived from an HIV+ patient with broadly cross-neutralizing antibodies, clearly demonstrated cross-clade neutralization in a PBMC assay, but no neutralization in the pseudovirus platform (Choudhry et al., 2007). Neutralization data are presented in Table 2 showing results using antibodies, screened at 30 μg/ml (mabs) or at a 1:40 Table 1 Comparison of HIV neutralization assay parameters Assay PBMC TZM-bl (JC53-bl) Cells PBMC Epithelial HeLa Virus Uncloned primary Pseudovirus Assay length 4 6 days 2 3 days Common endpoint Extra- or intra-cellular p24 Luciferase activity Rounds of infection Either multiple or single Single only DEAE-dextran used? No Yes Measure inhibition of: Attachment/entry Yes Yes Cell cell transmission Potentially No Coreceptors used CCR5, CXCR4, others (CCR5 physiologic) CCR5, CXCR4 (N2 logs more CCR5 than PBMC) dilution (USHIV+ pool), against 30 viruses representing six major clades (Brown et al., 2005). The antibodies were tested in a PBMC assay against primary isolates, or in the TZM-bl assay against a pseudovirus prepared from the matched primary isolate. Red boxes indicate the positive results and the negative values are highlighted in blue, using 50% neutralization as a cutoff (Brown et al., 2005). It can be seen that the m9 anti-gp120 single chain mab and the m47 anti-gp41 mab (M. Zhang et al., in preparation) both appear to show better neutralization in the PBMC system as compared to the TZM-bl assay. However, as previously shown by Binley et al. (2004), the 4E10 mab displays much broader neutralizing activity in the pseudovirus assay, neutralizing 29/30 pseudoviruses as compared to 13/30 primary isolates in the PBMC assay. As might be expected, the 2F5 mab does not neutralize clade C viruses, in which the target epitope is absent, in either assay (Table 2). A recent study of an anti-phosphatidyl inositol phosphate (PIP) mab has suggested that targeting lipids in close proximity to the viral and/or host cell membrane during viral fusion and entry, may be sufficient to neutralize HIV-1. The anti-pip mab appears to neutralize the virus only in assays where PBMC are used as target cells, and not at all in the TZM-bl cell line-based pseudovirus assay (Brown et al., 2007). Additional information regarding the activity of anti-lipid mabs derived from humans with anti-phospholipid antibody syndrome and systemic lupus erythematosus has also recently been presented. A subset of these antibodies, which recognize a variety of lipids (Alam et al., 2007), have been shown to neutralize primary isolates in a PBMC assay, but not the matched pseudoviruses in the TZM-bl platform (B Haynes, personal communication). This finding is similar to the observation using the PIP anti-lipid mab. It has been recognized that some investigators in the field of HIV vaccine research and development have perceived these two neutralization assay formats to be interchangeable in the data that they provide. Taking into account that there are several differences between the individual parameters of the two neutralization platforms (Table 1), we sought to compare a primary PBMC format to the TZM-bl pseudovirus assay, in totality and as they are currently performed, using a large panel of sequenced viruses from patients in the chronic stages of infection and polyclonal plasma pools. The pools were comprised of 6 10 pure clade plasmas from HIV-1-seropositive individuals from 6 different countries, each with a different predominant circulating HIV-1 clade. These plasma pools were tested against a panel of 60 primary isolates (10 each from 6 major clades A D, CRF01_AE and CRF02_AG), in a PBMC assay using reduction of supernatant p24 as an endpoint. Pseudoviruses prepared from a single env clone representing 56 of the 60 virus isolates, were then tested using the same 6 plasma pools in the TZM-bl cell assay. The virus antibody pairs showing 50% neutralization at a 1:40 plasma dilution were further evaluated to obtain 50% and 80% endpoint titers, and the reciprocal titer values from the two neutralization assays were compared by linear regression. Fig. 1 shows the comparison of the 336 plasma/virus pairs in the two assays. No direct correlation was observed (R 2 =0.002, Fig. 1A). Overall, there was 60% concordance in qualitative neutralizing activity measured by both assays. Within the

318 Minireview Table 2 Comparison of neutralization data from PBMC versus TZM-bl assays using a single antibody concentration of 30 μg/ml (mabs) or a 1:40 dilution (USHIV+ pool) Five different antibody sources were tested for neutralization ( 50%, indicated by red boxes) of 30 primary isolates in a PBMC assay or 30 pseudoviruses (derived from the corresponding strains) in the TZM-bl assay. The m9 is a single chain anti-gp120 mab and m47, 2F5, and 4E10 are gp41 mabs, while USHIV+ is a pool of sera from N. American HIV-positive patients. concordant results, 47% were negative in both assays and only 13% were concordant positive (Fig. 1B), in agreement with commonly observed neutralization resistance of primary isolate Envs. The remaining 40% of the paired titers were positive in one assay, and negative in the other. Of these pairs, 23% were positive only in the TZM-bl and 17% only in the PBMC assay. Thus, assay discordance was bi-directional and not attributable to assay sensitivity. The 80% reciprocal neutralization titers also showed no direct correlation (R 2 =0.004, data not shown). Using a panel of pseudoviruses prepared with clade B and C envelopes from acute and early infection, we have seen a similar discrepancy among 31 plasma/virus pairs tested in both assays (R 2 =0.05). In this smaller data set with acute/early samples, a 61% concordance was found in the qualitative neutralization between the two assays (data not shown). Identification of the mechanisms underlying the significant differences observed between these two assay formats will be important and may require examination of the biologic parameters at play in both assays. The assays in which primary cells are used have been proposed to more closely resemble the in vivo host cells for HIV, although mitogen-stimulated lymphoblasts are not typical of tissue or circulating T cells. Assays employing primary cells typically capture all stages of the virus life cycle during neutralization assays incorporating multiple rounds of infection. The TZM-bl assay is a single round assay and primarily assesses inhibition of virus binding and entry. It is possible that some antibody subpopulations may not be detected through the use of a single assay focusing only on virus entry. Another fundamental difference between T cells and the HeLa-derived TZM-bl cells may lie in the location of HIV/receptor interaction(s). The majority of HIV-1 entry into T cells occurs at the plasma membrane via CD4 and chemokine receptor-mediated engagement and fusion. In HeLa cells, endocytosis has been reported to account for approximately 85% of virus entry (Marechal et al., 1998). Moreover, TZM-bl assays often include the use of DEAEdextran during the infection phase. If endocytosis is playing a significant role in viral entry in TZM-bl cells, this compound most likely enhances productive infection by buffering endosomes, thus allowing the pseudovirus to avoid lysosomal degradation and to enter the cell through the endosomal vesicle. Data from our laboratory and others also suggest that certain cell line models, ie. TZM-bl cells, express vastly different levels of the cell surface coreceptor CCR5, compared to primary PBMC (Table 1). The CD4 receptor expression on the two cell types is relatively similar, but there appears to be about 100-fold more CCR5 on TZM-bl cells than on either activated or nonactivated PBMC. The CD4:CCR5 ratio may actually be more important to the kinetics of virus entry than simply the increased level of CCR5 on the TZM-bl and other cell lines. While the CD4:CCR5 ratio on TZM-bl cells is approximately 2, this same ratio is about 20, on average, for PHA-stimulated PBMC (Rosa Borges et al., submitted for publication). These cell surface differences may significantly alter measurements of virus antibody host cell interactions and kinetics, as has recently been shown using the X5 gp120 mab (Choudhry et al., 2006). The difference in the efficacy of neutralization by the X5 mab Fig. 1. (A) A comparison of 50% endpoint titers from the PBMC and TZM-bl assays using 56 primary viruses or 56 corresponding pseudoviruses neutralized using 6 clade-specific plasma pools. To allow the use of a log scale, all weak titers (b40) were arbitrarily assigned the value of 1.0. Linear regression (R 2 =0.002) shows no direct correlation in the values observed. (B) The proportion (%) of titers for virus/plasma pairs which are positive only in the PBMC or TZM-bl assays, or concordant positive or negative in both assays.

Minireview 319 in cells with reduced CCR5 expression underscores the role of the cell substrate in HIV neutralization assays. The mechanism by which neutralizing antibodies function also may influence how well the neutralization results obtained in the two platforms correlate. In the study by Binley et al., the results using the potently neutralizing mab IgG1 b12 (b12) correlated well between the two assays (84%), while the mab 4E10 showed only 52% concordance between the two assays (Binley et al., 2004). It is believed that CD4 engagement is the first step of HIV-1 infection, allowing the virus to recognize its target cell. The b12 mab is thought to neutralize by binding to the CD4 binding site in gp120 and stopping the engagement of gp120 and CD4 on the cell surface. Blockage of CD4 engagement has been shown to halt both plasma membrane fusion as well as entry into the cell through endocytosis in epithelial cells (Schaeffer et al., 2004); concordance in the two assays may reflect inhibition of an early event of HIV entry into primary cells and cell lines alike. While the mechanism of neutralization by 4E10 is unknown, it has been shown that this mab loses the ability to bind to the membrane proximal external region of gp41 (MPER) as the level of fusion increases. The inhibition of infection by 4E10 follows a similar time course to that of C34, a peptide that blocks virion Env fusion with the cell membrane (Dimitrov et al., 2007). If endocytosis is the major route of virus entry into TZM-bl cells, it is possible that the process of viral entry through the endosome results in extended (or otherwise altered) exposure of the MPER, thus allowing 4E10 to more effectively neutralize the virus in TZM-bl assays. Despite the differences in the data described herein (Table 2 and Fig. 1), without a measurable humoral immune correlate of protection from HIV-1 infection or viremic control as established in a clinical setting, it is impossible to predict which assay may better quantify protective neutralization. Thus, to better understand the differences between the assays, research should focus on understanding the results in which the neutralizing antibody virus pairs show discordant results between the two assays. Additionally, it may be valuable to study how the TZM-bl assay could be altered such that the parameters would more closely represent the physiologic state. This could include using TZM-bl clones or different reporter cell clones that express CCR5 at levels comparable to those levels measured on primary in vivo target cells (ie. lymphocytes and monocyte macrophages). The 293T cell clones used to produce the pseudoviruses could also be modified to express more common host factors that HIV particles incorporate while budding from primary host cells. These host cell molecules include such major immune system proteins as ICAMs, MHC class I and MHC class II (Cantin et al., 2005). Another approach to achieve parity between the two platforms involves using hybrid assays wherein pseudoviruses are used as viral inocula for PBMC target cells and primary isolates are used to infect TZM-bl cells. Studies such as these are currently underway in several laboratories and will help provide insight into the possible underlying mechanisms that influence the differences observed between these two assays. Functional assays using PBMC from different HIV-negative donors may also be impacted significantly by variations in the viral stocks. Viruses isolated from the plasma or cells of HIV-1- positive individuals on PHA-activated, donor PBMC represent a quasispecies, comprised of related but non-identical sequences. With serial passage on different PBMC, perturbations in the composition of the quasispecies can occur. Thus, virus stocks isolated on primary PMBC are impermanent, non-standardizable, and difficult to precisely define in terms of nucleotide sequence(s). Nevertheless, viruses isolated and propagated on primary PBMC are considered to most closely represent those circulating in vivo. In contrast, viruses adapted to growth in T-cell lines, and those produced in cell lines by transfection of infectious molecular clones (IMC), have altered neutralization susceptibility compared to PBMC-derived viruses (Louder et al., 2005). In addition, the pseudoviruses prepared using an env clone that represents only one of the possible quasispecies present in clinical isolates may or may not reflect the neutralization sensitivity of the quasispecies population from which the env clone was derived. The preparation of IMC (with the env clone or clones of choice incorporated) that could be passaged once in PBMC may prove useful for assessing the role of env diversity and host cell molecules in neutralization. Systematic comparisons of viral quasispecies, viral stocks of defined sequence produced in cell lines or primary PBMC, and pseudoviruses with corresponding Envs may also shed light on the reasons for the current discrepancies that have been reported between assays. Significant challenges remain for the development of a global vaccine for HIV. Foremost amongst these challenges is the identification, in a clinical setting, of assay platform(s) that will best provide predictive value with regard to protection of humans against HIV infection and spread. The performance of passive antibody transfer and protection studies in animal models, followed by assessment of the sera from these studies in current neutralization assay platforms, may help to determine which assay has greater correlative value. In the current absence of a gold standard for comparative vaccine studies, it is important to consider exactly what processes or parameters are being measured in each type of neutralization assay. Effective inhibition of HIV-1 entry may require different neutralizing antibody species, depending on the target cell. Although there seem to be identifiable disparities between neutralization profiles obtained using TZM-bl vs. PBMC assays, it is presently unclear what the relative contribution of either assay will be toward elucidating the correlates of neutralization susceptibility, and toward selecting the best candidate vaccine. Evaluation of antibodies in different assay platforms, reflecting the biologic variation in the interaction of HIV-1 with its target cells in vivo, should be contemplated as an interim solution to this problem. The TZM-bl assay, through a significant level of multinational effort and commitment of resources, has undergone substantial development and should be maintained as a widely applied, validated standard platform for HIV-1 vaccine development. The field should also acknowledge that further standardization of the PBMC assay has lagged and take steps to readdress this deficiency. The recent international conference convened by the WHO and NeutNet Working Group (March 17 18, 2007, Varese, Italy) highlighted the significant amount of ground to be covered before the PBMC assay achieves a level of consistency, reproducibility, and portability. Systematic

320 Minireview studies directed at reducing the variability within PBMC assays will be important in making progress towards standardization. Manipulation of the pseudovirus assay to more closely reflect physiologic parameters is a second approach that is currently being pursued in several laboratories. The results from studies within NeutNet and the GHAVE CA-VIMC will be a key in advancing our ability to identify differences between neutralization assays, either PBMC or reporter cell line-based. Indeed, the NeutNet study, while using smaller panels of reagents, will be able to address differences between multiple PBMC assays, as well as reporter cell line assays. The continuation of these studies, combined with the efforts and collaborations of multiple organizations, will provide insight into the inter-lab variability that occurs within and between assay systems. Finally, more detailed studies should and will be performed to compare platforms, as these studies may further reveal assay parameters or antibody properties that result in higher or lower degrees of concordance between different formats. These combined approaches will facilitate optimal progress towards the critical goal of achieving vaccine-elicited, broadly protective neutralizing antibodies. Acknowledgments This work was supported in part by Cooperative Agreement no. DAMD17-93-V-3004, between the U.S. Army Medical Research and Material Command and the Henry M. Jackson Foundation for the Advancement of Military Medicine, working with the Division of AIDS, National Institute for Allergy and Infectious Diseases National Institutes of Health. Funding also came in part from the National Cancer Institute, National Institutes of Health, under contract N01-CO-12400 and the Intramural Research Program of the NIH, NCI. We would like to thank Dr. John Mascola for the generous gift of 10 pseudovirus clones. We would also like to thank Tammy Oblander, Lindsay Wieczorek, Erik Odom and Kara Lombardi for their help with the neutralization assays. We also thank Eric Sanders-Buell, Sodsai Tovanabutra and Julie Horak for sequencing and cloning. The views and content expressed are those of the authors and should not be construed to represent the positions of the US Army, the Dept of Defense or the Dept of Health and Human Services. References Alam, S.M., McAdams, M., Boren, D., Rak, M., Scearce, R.M., Gao, F., Camacho, Z.T., Gewirth, D., Kelsoe, G., Chen, P., Haynes, B.F., 2007. The role of antibody polyspecificity and lipid reactivity in binding of broadly neutralizing anti-hiv-1 envelope human monoclonal antibodies 2F5 and 4E10 to glycoprotein 41 membrane proximal envelope epitopes. J. Immunol. 178 (7), 4424 4435. Binley, J.M., Wrin, T., Korber, B., Zwick, M.B., Wang, M., Chappey, C., Stiegler, G., Kunert, R., Zolla-Pazner, S., Katinger, H., Petropoulos, C.J., Burton, D.R., 2004. Comprehensive cross-clade neutralization analysis of a panel of anti-human immunodeficiency virus type 1 monoclonal antibodies. J. Virol. 78 (23), 13232 13252. Brown, B.K., Darden, J.M., Tovanabutra, S., Oblander, T., Frost, J., Sanders- Buell, E., de Souza, M.S., Birx, D.L., McCutchan, F.E., Polonis, V.R., 2005. Biologic and genetic characterization of a panel of 60 human immunodeficiency virus type 1 isolates, representing clades A, B, C, D, CRF01_AE, and CRF02_AG, for the development and assessment of candidate vaccines. J. Virol. 79 (10), 6089 6101. Brown, B.K., Karasavvas, N., Beck, Z., Matyas, G.R., Birx, D.L., Polonis, V.R., Alving, C.R., 2007. Monoclonal antibodies to phosphatidylinositol phosphate neutralize human immunodeficiency virus type 1: role of phosphate-binding subsites. J. Virol. 81 (4), 2087 2091. Cantin, R., Methot, S., Tremblay, M.J., 2005. Plunder and stowaways: incorporation of cellular proteins by enveloped viruses. J. Virol. 79 (11), 6577 6587. Choudhry, V., Zhang, M.Y., Harris, I., Sidorov, I.A., Vu, B., Dimitrov, A.S., Fouts, T., Dimitrov, D.S., 2006. Increased efficacy of HIV-1 neutralization by antibodies at low CCR5 surface concentration. Biochem. Biophys. Res. Commun. 348 (3), 1107 1115. Choudhry, V., Zhang, M.Y., Sidorov, I.A., Louis, J.M., Harris, I., Dimitrov, A.S., Bouma, P., Cham, F., Choudhary, A., Rybak, S.M., Fouts, T., Montefiori, D. C., Broder, C.C., Quinnan Jr., G.V., Dimitrov, D.S., 2007. Cross-reactive HIV-1 neutralizing monoclonal antibodies selected by screening of an immune human phage library against an envelope glycoprotein (gp140) isolated from a patient (R2) with broadly HIV-1 neutralizing antibodies. Virology 363 (1), 79 90. Cohen, J., 2007. AIDS research. Did Merck's failed HIV vaccine cause harm? Science 318 (5853), 1048 1049. D'Souza, M.P., Durda, P., Hanson, C.V., Milman, G., 1991. Evaluation of monoclonal antibodies to HIV-1 by neutralization and serological assays: an international collaboration. Collab. Investig. Aids 5 (9), 1061 1070. D'Souza, M.P., Livnat, D., Bradac, J.A., Bridges, S.H., 1997. Evaluation of monoclonal antibodies to human immunodeficiency virus type 1 primary isolates by neutralization assays: performance criteria for selecting candidate antibodies for clinical trials. AIDS Clinical Trials Group Antibody Selection Working Group. J. Infect. Dis. 175 (5), 1056 1062. Dimitrov, A.S., Jacobs, A., Finnegan, C.M., Stiegler, G., Katinger, H., Blumenthal, R., 2007. Exposure of the membrane-proximal external region of HIV-1 gp41 in the course of HIV-1 envelope glycoprotein-mediated fusion. Biochemistry 46 (5), 1398 1401. Huber, M., Trkola, A., 2007. Humoral immunity to HIV-1: neutralization and beyond. J. Intern. Med. 262 (1), 5 25. Louder, M.K., Sambor, A., Chertova, E., Hunte, T., Barrett, S., Ojong, F., Sanders-Buell, E., Zolla-Pazner, S., McCutchan, F.E., Roser, J.D., Gabuzda, D., Lifson, J.D., Mascola, J.R., 2005. HIV-1 envelope pseudotyped viral vectors and infectious molecular clones expressing the same envelope glycoprotein have a similar neutralization phenotype, but culture in peripheral blood mononuclear cells is associated with decreased neutralization sensitivity. Virology 339 (2), 226 238. Marechal, V., Clavel, F., Heard, J.M., Schwartz, O., 1998. Cytosolic Gag p24 as an index of productive entry of human immunodeficiency virus type 1. J. Virol. 72 (3), 2208 2212. Mascola, J.R., D'Souza, P., Gilbert, P., Hahn, B.H., Haigwood, N.L., Morris, L., Petropoulos, C.J., Polonis, V.R., Sarzotti, M., Montefiori, D.C., 2005. Recommendations for the design and use of standard virus panels to assess neutralizing antibody responses elicited by candidate human immunodeficiency virus type 1 vaccines. J. Virol. 79 (16), 10103 10107. McMichael, A.J., 2006. HIV vaccines. Annu. Rev. Immunol. 24, 227 255. Montefiori, D., Sattentau, Q., Flores, J., Esparza, J., Mascola, J., 2007. Antibodybased HIV-1 vaccines: recent developments and future directions. PLoS Med. 4 (12), e348. Schaeffer, E., Soros, V.B., Greene, W.C., 2004. Compensatory link between fusion and endocytosis of human immunodeficiency virus type 1 in human CD4 T lymphocytes. J. Virol. 78 (3), 1375 1383. Srivastava, I.K., Ulmer, J.B., Barnett, S.W., 2005. Role of neutralizing antibodies in protective immunity against HIV. Hum. Vaccin. 1 (2), 45 60. Zolla-Pazner, S., 1996. Mechanisms contributing to the neutralization of HIV-1. Immunol. Lett. 51 (1 2), 89 93. Zolla-Pazner, S., 2004. Identifying epitopes of HIV-1 that induce protective antibodies. Nat. Rev. Immunol. 4 (3), 199 210.