ABSTRACTImmunodominant and highly conserved flavivirus envelope proteins can trigger cross-reactive IgG antibodies against related flaviviruses, which shapes subsequent protection or disease severity. This study examined how prior dengue serotype 3 (DENV-3) infection affects subsequent Zika virus (ZIKV) plasmablast responses in rhesus macaques (n = 4). We found that prior DENV-3 infection was not associated with diminished ZIKV-neutralizing antibodies or magnitude of plasmablast activation. Rather, characterization of 363 plasmablasts and their derivative 177 monoclonal antibody supernatants from acute ZIKV infection revealed that prior DENV-3 infection was associated with a differential isotype distribution toward IgG, lower somatic hypermutation, and lesser B cell receptor variable gene diversity as compared with repeat ZIKV challenge. We did not find long-lasting DENV-3 cross-reactive IgG after a ZIKV infection but did find persistent ZIKV-binding cross-reactive IgG after a DENV-3 infection, suggesting non-reciprocal cross-reactive immunity. Infection with ZIKV after DENV-3 boosted pre-existing DENV-3-neutralizing antibodies by two- to threefold, demonstrating immune imprinting. These findings suggest that the order of DENV and ZIKV infections has impact on the quality of early B cell immunity which has implications for optimal immunization strategies.IMPORTANCEThe Zika virus epidemic of 2015–2016 in the Americas revealed that this mosquito-transmitted virus could be congenitally transmitted during pregnancy and cause birth defects in newborns. Currently, there are no interventions to mitigate this disease and Zika virus is likely to re-emerge. Understanding how protective antibody responses are generated against Zika virus can help in the development of a safe and effective vaccine. One main challenge is that Zika virus co-circulates with related viruses like dengue, such that prior exposure to one can generate cross-reactive antibodies against the other which may enhance infection and disease from the second virus. In this study, we sought to understand how prior dengue virus infection impacts subsequent immunity to Zika virus by single-cell sequencing of antibody producing cells in a second Zika virus infection. Identifying specific qualities of Zika virus immunity that are modulated by prior dengue virus immunity will enable optimal immunization strategies.

Human Cytomegalovirus (HCMV) is the leading infectious congenital infection globally and the most common viral infection in transplant recipients, therefore identifying a vaccine for HCMV is a top priority. Humoral immunity is a correlate of protection for HCMV infection. The most effective vaccine tested to date, which achieved 50% reduction in acquisition of HCMV, was comprised of the glycoprotein B protein given with an oil-in-water emulsion adjuvant MF59. We characterize gB-specific monoclonal antibodies isolated from individuals vaccinated with a disabled infectious single cycle (DISC) CMV vaccine, V160, and compare these to the gB-specific monoclonal antibody repertoire isolated from naturally-infected individuals. We find that vaccination with V160 resulted in gB-specific antibodies that bound homogenously to gB expressed on the surface of a cell in contrast to antibodies isolated from natural infection which variably bound to cell-associated gB. Vaccination resulted in a similar breadth of gB-specific antibodies, with binding profile to gB genotypes 1-5 comparable to that of natural infection. Few gB-specific neutralizing antibodies were isolated from V160 vaccinees and fewer antibodies had identifiable gB antigenic domain specificity compared to that of naturally-infected individuals. We also show that glycosylation of gB residue N73 may shield binding of gB-specific antibodies.

Abstract Human Cytomegalovirus (HCMV) is the leading infectious congenital infection globally and the most common viral infection in transplant recipients, therefore identifying a vaccine for HCMV is a top priority. Humoral immunity is a correlate of protection for HCMV infection. The most effective vaccine tested to date, which achieved 50% reduction in acquisition of HCMV, was comprised of the glycoprotein B protein given with an oil-in-water emulsion adjuvant MF59. We characterize gB-specific monoclonal antibodies isolated from individuals vaccinated with a disabled infectious single cycle (DISC) CMV vaccine, V160, and compare these to the gB-specific monoclonal antibody repertoire isolated from naturally-infected individuals. We find that vaccination with V160 resulted in gB-specific antibodies that bound homogenously to gB expressed on the surface of a cell in contrast to antibodies isolated from natural infection which variably bound to cell-associated gB. Vaccination resulted in a similar breadth of gB-specific antibodies, with binding profile to gB genotypes 1-5 comparable to that of natural infection. Few gB-specific neutralizing antibodies were isolated from V160 vaccinees and fewer antibodies had identifiable gB antigenic domain specificity compared to that of naturally-infected individuals. We also show that glycosylation of gB residue N73 may shield binding of gB-specific antibodies.

Cytomegalovirus (CMV) is a leading cause of infant hearing loss and neurodevelopmental delay, but there are no clinically licensed vaccines to prevent infection, in part due to challenges eliciting neutralizing antibodies. One of the most well-studied targets for CMV vaccines is the viral fusogen glycoprotein B (gB), which is required for viral entry into host cells. Within gB, antigenic domain 2 site 1 (AD-2S1) is a target of potently neutralizing antibodies, but gB-based candidate vaccines have yet to elicit robust responses against this region. We mapped the genealogy of B cells encoding potently neutralizing anti-gB AD-2S1 antibodies from their inferred unmutated common ancestor (UCA) and characterized the binding and function of early lineage ancestors. Surprisingly, we found that a single amino acid heavy chain mutation A33N, which was an improbable mutation rarely generated by somatic hypermutation machinery, conferred broad CMV neutralization to the non-neutralizing UCA antibody. Structural studies revealed that this mutation mediated key contacts with the gB AD-2S1 epitope. Collectively, these results provide insight into potently neutralizing gB-directed antibody evolution in a single donor and lay a foundation for using this B cell-lineage directed approach for the design of next-generation CMV vaccines.

Cytomegalovirus (CMV) is a leading cause of infant hearing loss and neurodevelopmental delay, but vaccine candidates have faced challenges eliciting neutralizing antibodies. One of the most well-studied targets for CMV vaccines is the viral fusogen glycoprotein B (gB), which is required for viral entry into host cells. Within gB, antigenic domain 2 site 1 (AD-2S1) is a target of potently neutralizing antibodies, but gB-based candidate vaccines have yet to elicit robust responses against this region. We mapped the genealogy of B cells encoding potently neutralizing anti-gB AD-2S1 antibodies from their inferred unmutated common ancestor (UCA) and characterized the binding and function of early lineage ancestors. Remarkably, we found that the single amino acid heavy chain mutation A33N, an improbable mutation rarely generated by somatic hypermutation machinery, conferred broad CMV neutralization to the UCA antibody. Structural studies revealed that this mutation mediated key contacts with the gB AD-2S1 epitope. Collectively, these results provide insight into potently neutralizing gB-directed antibody evolution and a foundation for designing next-generation CMV vaccines. One-sentence summary This manuscript identifies an early B cell lineage mutation that confers neutralizing function to antibodies targeting CMV fusogen gB.

Despite the worldwide availability of antiretroviral therapy (ART), approximately 150,000 pediatric HIV infections continue to occur annually. ART can dramatically reduce HIV mother-to-child transmission (MTCT), but inconsistent drug access and adherence, as well as primary maternal HIV infection during pregnancy and lactation are major barriers to eliminating vertical HIV transmission. Thus, immunologic strategies to prevent MTCT, such as an HIV vaccine, will be required to attain an HIV-free generation. A primary goal of HIV vaccine research has been to elicit broadly neutralizing antibodies (bnAbs) given the ability of passive bnAb immunization to protect against sensitive strains, yet we previously observed that HIV-transmitting mothers have more plasma neutralization breadth than non-transmitting mothers. Additionally, we have identified infant transmitted/founder (T/F) viruses that escape maternal bnAb responses. In this study, we examine a cohort of postpartum HIV-transmitting women with neutralization breadth to determine if certain maternal bnAb specificities drive the selection of infant T/F viruses. Using HIV pseudoviruses that are resistant to neutralizing antibodies targeting common bnAb epitopes, we mapped the plasma bnAb specificities of this cohort. Significantly more transmitting women with plasma bnAb activity had a mappable plasma bnAb specificity (six of seven, or 85.7%), compared to that of non-transmitting women with plasma bnAb activity (seven of twenty-one, or 33.3%, p=0.029 by 2-sided Fisher exact test). Our study suggests that having multispecific broad activity and/or uncommon epitope-specific bnAbs in plasma may be associated with protection against the vertical HIV transmission in the setting of maternal bnAb responses. Importance As mother to child transmission (MTCT) of HIV plays a major part in the persistence of the HIV/AIDS epidemic and bnAb-based passive and active vaccines are a primary strategy for HIV prevention, research in this field is of great importance. While previous MTCT research has investigated the neutralizing antibody activity of HIV-infected women, this is to our knowledge the largest study identifying differences in bnAb specificity of maternal plasma between transmitting and non-transmitting women. Here, we show that among HIV-infected women with broad and potent neutralization activity, more postpartum-transmitting women had a mappable plasma broadly neutralizing antibody (bnAb) specificity, compared to that of non-transmitting women, suggesting that the non-transmitting women more often have multi-specific bnAb responses or bnAb responses that target uncommon epitopes. Such responses may be required for protection against vertical HIV transmission in the setting of maternal bnAb responses.

The development of a maternal HIV vaccine to synergize with current antiretroviral drug prophylaxis can overcome implementation challenges and further reduce mother-to-child transmission (MTCT) of HIV. Both the epitope-specificity and autologous neutralization capacity of maternal HIV envelope (Env)-specific antibodies have been implicated in decreased risk of MTCT of HIV. Our goal was to determine if heterologous HIV Env immunization of SHIV.C.CH505-infected, ART-suppressed female rhesus macaques (RMs) could boost autologous Env-specific antibodies. SHIV.C.CH505-infected female RMs ( n = 12), began a daily ART regimen at 12 weeks post-infection (wpi), which was continued for 12 weeks. Starting 2 weeks after ART initiation, RMs received 3 monthly immunizations with HIV b.63521/1086.C gp120 or placebo ( n = 6/group) vaccine with adjuvant STR8S-C. Compared to the placebo-immunized animals, Env-vaccinated, SHIV-infected RMs exhibited enhanced IgG binding, avidity, and ADCC responses against the vaccine immunogens and the autologous SHIV.C.CH505 Env. Notably, the Env-specific memory B cells elicited by heterologous vaccination were dominated by cells that recognized the SHIV.C.CH505 Env, the antigen of primary exposure. Thus, vaccination of SHIV-infected, ART-suppressed RMs with heterologous HIV Envs can augment multiple components of the antibody response against the Env antigen of primary exposure, suggesting antigenic seniority. Our results suggest that a universal maternal HIV vaccination regimen can be developed to leverage antigenic seniority in targeting the maternal autologous virus pool.

Despite considerable reduction of mother-to-child transmission (MTCT) of HIV through use of maternal and infant antiretroviral therapy (ART), over 150,000 infants continue to become infected with HIV annually, falling far short of the World Health Organization goal of reaching <20,000 annual pediatric HIV cases worldwide by 2020. Prior to the widespread use of ART in the setting of pregnancy, over half of infants born to HIV-infected mothers were protected against HIV acquisition. Yet, the role of maternal immune factors in this protection against vertical transmission is still unclear, hampering the development of synergistic strategies to further reduce MTCT. It has been established that infant transmitted/founder (T/F) viruses are often resistant to maternal plasma, yet it is unknown if the neutralization resistance profile of circulating viruses predicts the maternal risk of transmission to her infant. In this study, we amplified HIV-1 envelope genes (env) by single genome amplification and produced representative Env variants from plasma of 19 non-transmitting mothers from the U.S. Women Infant Transmission Study (WITS), enrolled in the pre-ART era. Maternal HIV Env variants from non-transmitting mothers had similar sensitivity to autologous plasma as observed for non-transmitting variants from transmitting mothers. In contrast, infant variants were on average 30% less sensitive to paired plasma neutralization compared to non-transmitted maternal variants from both transmitting and non-transmitting mothers (p = 0.015). Importantly, a signature sequence analysis revealed that motifs enriched in env sequences from transmitting mothers were associated with broadly neutralizing antibody (bnAb) resistance. Altogether, our findings suggest that circulating maternal virus resistance to bnAb-mediated neutralization, but not autologous plasma neutralization, near the time of delivery, predicts increased MTCT risk. These results caution that enhancement of maternal plasma neutralization through passive or active vaccination during pregnancy may potentially drive the evolution of variants fit for vertical transmission.

Development of a human cytomegalovirus (HCMV) vaccine is a Tier 1 priority by the National Institutes of Medicine, as HCMV is the most common congenital infection globally and most frequent infectious complication in transplant patients. Relevant preclinical non-human primate models used for testing HCMV vaccine immunogenicity are rhesus and cynomolgous monkeys. However, a complication in using these models is that species-specific CMV variants are endemic in non-human primate breeding colonies. We hypothesize that natural immunity to species-specific CMV in rhesus and cynomolgous monkeys impacts HCMV vaccine immunogenicity and may interfere with our ability to fully interpret vaccine immunogenicity. A modified mRNA vaccine encoding HCMV glycoprotein (gB) and the pentameric complex (PC) packaged in lipid nanoparticles (LNP) was delivered intramuscularly to groups of cynomolgous (n = 16, CyCMV-seropositive) and rhesus macaques (n = 24, RhCMV-seropositive). High pre-vaccination IgG binding responses to HCMV gB were present in both species, but pre-vaccination binding responses to PC were mostly present in rhesus macaques. Yet, at least a log increase in both PC and gB-specific plasma IgG levels was detected post-second HCMV mRNA vaccination in both species. Both species responded with high epithelial cell neutralizing antibody responses at 4 weeks post second HCMV mRNA vaccination, but limited fibroblast neutralizing antibodies. HCMV gB + PC mRNA/LNP vaccine also elicited IgG binding responses to cell-associated gB, an identified immune correlate of protection, in both species after the second vaccination, and there was a moderately strong direct correlation between this pre- and post-vaccination response in rhesus macaques. Based on the correlation between pre-existing and post-vaccine gB-specific binding responses in rhesus macaques, we conclude that species-specific CMV variant-specific antibody responses contribute to antibody responses to HCMV vaccination in primate models, indicating that pre-existing immunity must be taken into account in non-human primate preclinical models and will impact immunogenicity of HCMV vaccines seropositive human vaccinees.

Understanding viral rebound in pediatric HIV-1 infection may inform the development of alternatives to lifelong antiretroviral therapy (ART) to achieve viral remission. We thus investigated viral rebound after analytical treatment interruption (ATI) in 10 infant macaques orally infected with SHIV.C.CH505 and treated with long-term ART. Rebound viremia was detected within 7 to 35 days of ATI in 9 of 10 animals, with posttreatment control of viremia seen in 5 of 5 Mamu-A*01+ macaques. Single-genome sequencing revealed that initial rebound virus was similar to viral DNA present in CD4+ T cells from blood, rectum, and lymph nodes before ATI. We assessed the earliest sites of viral reactivation immediately following ATI using ImmunoPET imaging. The largest increase in signal that preceded detectable viral RNA in plasma was found in the gastrointestinal (GI) tract, a site with relatively high SHIV RNA/DNA ratios in CD4+ T cells before ATI. Thus, the GI tract may be an initial source of rebound virus, but as ATI progresses, viral reactivation in other tissues likely contributes to the composition of plasma virus. Our study provides potentially novel insight into the features of viral rebound in pediatric infection and highlights the application of a noninvasive technique to monitor areas of HIV-1 expression in children.

Study of evolution and selection pressure on HIV-1 in fetuses will lead to a better understanding of the role of immune responses in shaping virus evolution and vertical transmission. Detailed genetic analyses of HIV-1 env gene from 12 in utero transmission pairs show that most infections (67%) occur within 2 months of childbirth. In addition, the env sequences from long-term-infected fetuses are highly divergent and form separate phylogenetic lineages from their cognate maternal viruses. Host-selection sites unique to neonate viruses are identified in regions frequently targeted by neutralizing antibodies and T cell immune responses. Identification of unique selection sites in the env gene of fetal viruses indicates that the immune system in fetuses is capable of exerting selection pressure on viral evolution. Studying selection and evolution of HIV-1 or other viruses in fetuses can be an alternative approach to investigate adaptive immunity in fetuses.

Optimal bNAb immunotherapeutics will need to mediate multiple antiviral functions against a broad range of HIV strains. Our systematic assessment of triple bNAb combinations against SHIVs will identify bNAbs with synergistic, polyfunctional antiviral activity that will inform the selection of candidate bNAbs for optimal combination designs. The identified combinations can be validated in vivo in future passive immunization studies using the SHIV challenge model.

Maternal antiretroviral therapy (ART) has effectively reduced but not eliminated the burden of mother-to-child transmission of HIV across the globe, as an estimated 160,000 children were newly infected with HIV in 2018. Thus, additional preventive strategies beyond ART will be required to close the remaining gap and end the pediatric HIV epidemic. A maternal active immunization strategy that synergizes with maternal ART could further reduce infant HIV infections. In this study, we found that two historic HIV Env vaccines did not enhance the ability of HIV-infected pregnant women to neutralize autologous viruses. Therefore, next-generation maternal HIV vaccine candidates must employ alternate approaches to achieve potent neutralizing antibody and perhaps nonneutralizing antibody responses to effectively impede vertical virus transmission. Moreover, these approaches must reflect the broad diversity of HIV strains and widespread availability of ART worldwide.

Each year, >180,000 infants become infected via mother-to-child transmission (MTCT) of HIV despite the availability of effective maternal antiretroviral treatments, underlining the need for a maternal HIV vaccine. We characterized 224 maternal HIV envelope (Env)-specific IgG monoclonal antibodies (MAbs) from seven nontransmitting and transmitting HIV-infected U.S. and Malawian mothers and examined their neutralization activities against nontransmitted autologous circulating viruses and infant-transmitted founder (infant-T/F) viruses. Only a small subset of maternal viruses, 3 of 72 (4%), were weakly neutralized by maternal linear V3 epitope-specific IgG MAbs, whereas 6 out of 6 (100%) infant-T/F viruses were neutralization resistant to these V3-specific IgG MAbs. We also show that maternal-plasma broadly neutralizing antibody (bNAb) responses targeting the V3 glycan supersite in a transmitting woman may have selected for an N332 V3 glycan neutralization-resistant infant-T/F virus. These data have important implications for bNAb-eliciting vaccines and passively administered bNAbs in the setting of MTCT.IMPORTANCE Efforts to eliminate MTCT of HIV with antiretroviral therapy (ART) have met little success, with >180,000 infant infections each year worldwide. It is therefore likely that additional immunologic strategies that can synergize with ART will be required to eliminate MTCT of HIV. To this end, understanding the role of maternal HIV Env-specific IgG antibodies in the setting of MTCT is crucial. In this study, we found that maternal-plasma broadly neutralizing antibody (bNAb) responses can select for T/F viruses that initiate infection in infants. We propose that clinical trials testing the efficacy of single bNAb specificities should not include HIV-infected pregnant women, as a single bNAb might select for neutralization-resistant infant-T/F viruses.

Number Name Category email Title 1 Berendam Stella Basic stella.berendam@duke.edu Kinetics of Antibody Development, Functionality, and Breadth in SHIV.C.CH505 375H dCT-infected Infant Rhesus Macaques with Delayed ART Treatment 2 Borges* Monica SBS monicaeb@live.unc.edu Developing a predictive model for HIV clinical care disengagement 3 Bu* Fan Basic fb75@duke.edu Identifying Immune Correlates of Time to Viral Rebound for Infant Rhesus Macaques 4 Deng* Machao Basic machao.deng@duke.edu Using Machine Learning in Pharmacokinetic and Pharmacodynamic Modeling of Anti-HIV Microbicide Performance 5 Dennis Maria Basic maria.dennis@duke.edu Induction of neutralizing antibody (nAb) responses by immunization with soluble native-like HIV envelope trimer vaccines (SOSIP) in infant rhesus macaques 6 Gobeil Sophie Basic sophie.gobeil@duke.edu Structural characterization of HIV-1 Env V3-glycan DH1030lineage bnAbs induced by macaque BG505 SHIV infection. 7 Ho* Dena Clinical dzion@math.duke.edu Validation of biomarkers for estimating HIV-1 incidence 8 Holt Lauren SBS lauren.holt@duke.edu Stigma Among Women Living with HIV in Rwanda Abstract Number Name Category email TitleNumber Name Category email Title 9 Irwan Ishak Basic id32@duke.edu Reversal of Epigenetic Silencing Allows Robust HIV-1 Replication in the Absence of Integrase Function 10 Mainou* Ellie Basic ejm6007@psu.edu Mathematical modeling of autologous antibodies impact on the time to viral rebound following ART interruption in postnatally SHIV-infected infant rhesus macaques 11 Manne Kartik Basic kartik.manne@duke.edu Structural characterization of autologous Tier 2 neutralizing antibody DH1025.2 elicited by vaccination of rhesus macaques 12 MontanoCampos* Felipe SBS ose.montano.campos@duke.ed Examining Structural Barriers to Optimal HIV Care Engagement: Is Neighborhood of Residence a Risk Factor for HIV Patient Attrition? 13 Muhirwa Amnazo SBS amnazo.muhirwa@duke.edu Patient perspectives on the Helpfulness of a Community Health Worker program for HIV care engagement in Kilimanjaro, Tanzania. 14 Nemeth* Hayley Clinical hayley.nemeth@duke.edu Persistent Elevation of Liver Transaminases Following HCV Virologic Cure Among HCV Mono-infected and HIV/HCV Coinfected Adults 15 Niyonzima* Yvonne Clinical yniyonz@ncsu.edu Connectome: HIV-Associated Neurocognitie Disorders (HAND) Abstract Number Name Category email TitleNumber Name Category email Title 16 Rochat Eric Basic eric.rochat@duke.edu Cytomegalovirus glycoprotein B-specific antibody repertoire isolated from naturally-infected and V160 vaccinated individuals 17 Sass Julian Basic jasass@ncsu.edu Mathematical Modeling of SHIV Dynamics Before Treatment in Infant and Adult Rhesus Macaques 18 Steppe Justin Basic justin.steppe@duke.edu Development of synthetic DNA-encoded dimeric AgA antibodies that potently neutralize rotavirus 19 Tu Joshua Basic joshua.tu@duke.edu The role of maternal broadly neutralizing antibodies in mother to child transmission 20 Webster Helen Basic helen.webster@duke.edu Pre-existing immunity to cytomegalovirus in cynomolgous and rhesus monkeys influences human CMV vaccine responses in preclinical models 21 Yu* Yahe Basic yyu29@ncsu.edu Impact of Maternal HIV Infection on the Respiratory Microbiome of Infants *Summer Intern project 2020 CFAR FALL SCIENTIFIC RETREAT Research Category: Basic/Translational Full Name: Stella Berendam Email address: stella.berendam@duke.edu Job Title: Postdoctoral fellow Departmental Affiliation: Duke Human Vaccine Inst Mentor (if applicable): Genevieve Fouda Background: A successful pediatric HIV remission/cure strategy, especially in postnatally infected children, will likely require combination of ART with immune-based interventions that target HIV replication, limit the size of virus reservoirs, maintain virus suppression, and delay time to virus rebound. A previous small study reported that the development of HIVspecific antibodies in infants who started ART early delayed virus rebound following treatment interruption, suggesting that elicitation of HIV-specific antibodies in ART-treated infants could impact time to virus rebound. Most infants infected postnatally during the breast-feeding period do not start ART early due to delayed in a diagnosis. The kinetics of HIVspecific antibody response and their potential impact on virus rebound have never been studied in the setting of postnatal HIV transmission. To address this question, we used previously developed infant rhesus macaque (RM) model of breastmilk transmission with late ART therapy. Objective: The primary goal is to define the kinetics, specificity, breadth, and antiviral functions of HIV-specific antibody responses in an infant RM model of postnatal HIV infection with late ART initiation. Methods: Ten (n=10) infant RMs were orally challenged with SHIV.C.CH505 375H dCT and daily triple ART was initiated at 8-week post-infection (wpi). ART was interrupted after 52 weeks (ATI) and virus rebound was monitored by viral RNA detection in infant plasma. Envelope (Env)-specific antibody levels in infant plasma were measured against the autologous virus CH505 T/F gp120 and MN gp41. Meanwhile, infant plasma antibody specificity and breadth against cross clade HIV peptides were assessed using luminex-based assay. Infant plasma antibody neutralizing and nonneutralizing functions (ADCC and ADCP) were evaluated at time points pre-ART, during ART, and ATI. Results: All infant RMs established plasma viremia with median viral loads at peak (2wpi) and immediately prior to ART were ~500,000 and 100,000 copies/ml, respectively. HIV Env-specific antibodies were first detected at 4wpi and following ART initiation (8 wpi) there was a significant decline in gp120and gp41-specific antibodies, 89.7% and 88.5%, respectively. However, the antibody levels remained stable during ART until virus rebound/ATI. Infant plasma antibodies showed broad specificity and breadth against heterologous HIV Env peptides even while on ART. Upon ATI, virus rebound was observed in 9 of 10 infants (range, 7-35 days after ATI) with 5 of 10 infants developed autologous neutralization including 1 infant that did not rebound. Further assessment using purified IgG from infant plasma demonstrated limited neutralization and ADCC activities at pre-ART, during ART, rebound, and post-ATI. Infant plasma antibody from time point pre-ART, during Art, and ATI also mediated robust ADCP activity against CH505-gp120-coated beads. Conclusions: Our results demonstrate that SHIV-infected infant RMs with late ART initiation are able to generate virusspecific antibody responses with broad specificity and breadth. Additionally, SHIV-infected ART-treated infant RMs continue to develop and maintain neutralizing and non-neutralizing antibody functions even while on ART. In future work, we will determine the Env-specific antibody half-life in infant plasma and define viral evolution that confer immune escape to autologous neutralization in postnatally SHIV-infected infants following long-term ART discontinuation. Overall, this study will inform the development of strategies to augment HIV-specific antibody responses in ART-treated infants to reduce the size of the viral reservoir and delay viral rebound following ART interruption. 1 2020 CFAR FALL SCIENTIFIC RETREAT Research Category: Social/Behavioral Monica Borges monicaeb@live.unc.edu M.S. Candidate, Biostatistics (UNC Chapel Hill) Duke CFAR Quantitative Intern (Summer 2020) Dr. Lance Okeke Developing a predictive model for HIV clinical care disengagement Background Disengaging from care is associated with all-cause mortality in people living with HIV (PLWH); therefore tools that can assist with the early identification of persons at-risk for falling out of care are needed. The electronic health record (EHR) serves as an invaluable platform to create such tools. Here, we present an EHR-based predictive model that incorporates demographics, clinical history and health utilization behavior to identify persons at risk for disengagement from HIV care. Methods The outcome of interest, disengaging from care, is defined as not visiting the HIV clinic for 12+ months after previously being in care. Data came from the Duke HIV clinic spanning 2014-2018, and one year of data was used for each patient. LASSO regression was used in selecting features for a logistic model. The model performance criterion used for LASSO was area under the receiver operating characteristic curve (AUC). A non-regularized logistic model was then fit using the LASSO-selected predictors to verify performance on the testing set. AUC was calculated to assess model performance, and estimated coefficients were transformed to adjusted odds ratios. Results 2301 subjects (mean age 46 years; 72% male; 58% Black) were included in analysis. 39 candidate variables included demographics, mental health and STI diagnoses, substance use, and visits to the ER, hospital, and HIV clinic. Predictor variables positively associated with disengaging from care include number of positive gonorrhea or chlamydia tests; number of recent syphilis tests; ever being diagnosed with syphilis; a positive amphetamines test; and schizophrenia diagnosis. Predictors negatively associated with disengagement from care are higher age; number of gonorrhea/chlamydia tests; 2+ emergency visits; 1+ hospital admissions; depression diagnosis; and visiting the HIV clinic in each half of the year. A logistic regression model using these predictors for the testing set results in an AUC of 0.808. Conclusions A combination of demographic, STI/mental health diagnosis, and health behavior variables are useful for identifying PLWH in care who are at risk of disengaging. Future iterations of the model will incorporate data on structural barriers to health access. With HIV clinic visits in the model, emergency department and hospital visits still offer insight about the outcome. 2

Studies in animal models are essential prerequisites for clinical trials of candidate HIV vaccines. Small animals, such as rabbits, are used to evaluate promising strategies prior to further immunogenicity and efficacy testing in nonhuman primates. Our goal was to determine how HIV-specific vaccine-elicited antibody responses, epitope specificity, and Fc-mediated functions in the rabbit model can predict those in the rhesus macaque (RM) model. Detailed comparisons of the HIV-1-specific IgG response were performed on serum from rabbits and RM given identical modified vaccinia virus Ankara-prime/gp120-boost immunization regimens. We found that vaccine-induced neutralizing antibody, gp120-binding antibody levels and immunodominant specificities, antibody-dependent cellular phagocytosis of HIV-1 virions, and antibody-dependent cellular cytotoxicity (ADCC) responses against gp120-coated target cells were similar in rabbits and RM. However, we also identified characteristics of humoral immunity that differed across species. ADCC against HIV-infected target cells was elicited in rabbits but not in RM, and we observed differences among subdominantly targeted epitopes. Human Fc receptor binding assays and analysis of antibody-cell interactions indicated that rabbit vaccine-induced antibodies effectively recruited and activated human natural killer cells, while vaccine-elicited RM antibodies were unable to activate either human or RM NK cells. Thus, our data demonstrate that both Fc-independent and Fc-dependent functions of rabbit antibodies can be measured with commonly used in vitro assays; however, the ability of immunogenicity studies performed in rabbits to predict responses in RM will vary depending on the particular immune parameter of interest. IMPORTANCE Nonneutralizing antibody functions have been associated with reduced infection risk, or control of virus replication, for HIV-1 and related viruses. It is therefore critical to evaluate development of these responses throughout all stages of preclinical testing. Rabbits are conventionally used to evaluate the ability of vaccine candidates to safely elicit antibodies that bind and neutralize HIV-1. However, it remained unexplored how effectively rabbits model the development of nonneutralizing antibody responses in primates. We administered identical HIV-1 vaccine regimens to rabbits and rhesus macaques and performed detailed comparisons of vaccine-induced antibody responses. We demonstrated that nonneutralizing HIV-specific antibody responses can be studied in the rabbit model and have identified aspects of these responses that are common, and those that are unique, to rabbits and rhesus macaques. Our findings will help determine how to best utilize preclinical rabbit and rhesus macaque models to accelerate HIV vaccine candidate testing in human trials.

To achieve long-term viral remission in human immunodeficiency virus (HIV)-infected children, novel strategies beyond early antiretroviral therapy (ART) will be necessary. Identifying clinical predictors of the time to viral rebound upon ART interruption will streamline the development of novel therapeutic strategies and accelerate their evaluation in clinical trials. However, identification of these biomarkers is logistically challenging in infants, due to sampling limitations and the potential risks of treatment interruption. To facilitate the identification of biomarkers predicting viral rebound, we have developed an infant rhesus macaque (RM) model of oral simian-human immunodeficiency virus (SHIV) SHIV.CH505.375H.dCT challenge and analytical treatment interruption (ATI) after short-term ART. We used this model to characterize SHIV replication kinetics and virus-specific immune responses during short-term ART or after ATI and demonstrated plasma viral rebound in 5 out of 6 (83%) infants. We observed a decline in humoral immune responses and partial dampening of systemic immune activation upon initiation of ART in these infants. Furthermore, we monitored SHIV replication and rebound kinetics in infant and adult RMs and found that both infants and adults demonstrated equally potent virus-specific humoral immune responses. Finally, we validated our models by confirming a well-established correlate of the time to viral rebound, namely, the preART plasma viral load, as well as identified additional potential humoral immune correlates. Thus, this model of infant ART and viral rebound can be used and further optimized to define biomarkers of viral rebound following long-term ART as well as to preclinically assess novel therapies to achieve a pediatric HIV functional cure. IMPORTANCE Novel interventions that do not rely on daily adherence to ART are needed to achieve sustained viral remission for perinatally infected children, who currently rely on lifelong ART. Considering the risks and expense associated with ART interruption trials, the identification of biomarkers of viral rebound will prioritize promising therapeutic intervention strategies, including anti-HIV Env protein therapeutics. However, comprehensive studies to identify those biomarkers are logistically challenging in human infants, demanding the need for relevant nonhuman primate models of HIV rebound. In this study, we developed an infant RM model of oral infection with simian-human immunodeficiency virus expressing clade C HIV Env and short-term ART followed by ATI, longitudinally characterizing the immune responses to viral infection during ART and after ATI. Additionally, we compared this infant RM model to an analogous adult RM rebound model and identified virologic and immunologic correlates of the time to viral rebound after ATI.

Preventive strategies beyond ART will be required to end the pediatric HIV epidemic. A maternal vaccine capable of boosting neutralizing antibody (nAb) responses against circulating viruses in HIV-infected pregnant women could effectively decrease mother-to-child transmission of HIV. However, it is not known if an HIV envelope (Env) vaccine administered to infected pregnant women can enhance autologous virus neutralization. Here, we assessed autologous virus nAb responses in maternal plasma samples obtained from AIDS Vaccine Evaluation Group (AVEG) Protocols 104 and 102, historical Phase I safety and immunogenicity trials of recombinant HIV Env subunit vaccines in HIV-infected pregnant women ( NCT00001041 ). AVEG 104 participants were randomized to receive 300 µg Env subunit MN recombinant gp120 with alum adjuvant or alum alone. AVEG 102 participants were randomized to receive 640 µg Env subunit recombinant gp160 or placebo. HIV Env-specific maternal plasma binding and neutralizing responses were characterized before and after vaccination in 15 AVEG 104 (n=10 vaccinee, n=5 placebo) and 2 AVEG 102 (n=1 vaccinee, n=1 placebo) participants. Single genome amplification (SGA) was used to obtain HIV env gene sequences from autologous viruses for pseudovirus production in pre- and post-vaccination plasma of HIV-infected pregnant vaccinees (n=6 gp120, n=1 gp160) and placebo recipients (n=3). We detected an increase in MN gp120-specific IgG binding in the vaccinee group between the first immunization visit and the last visit at delivery (p=0.027, 2-sided Wilcoxon test). However, no difference was observed in the neutralization potency of maternal plasma collected at delivery against autologous viruses isolated from early or late pregnancy. Thus, maternal vaccination with gp120/160 did not boost maternal autologous virus nAb responses. Immunization strategies capable of more potent B cell stimulation will likely be required to effectively boost autologous virus nAb responses in pregnant women and synergize with ART to further reduce infant HIV infections. Highlights Prior maternal HIV Env vaccine trial did not assess autologous virus neutralization Circulating viruses isolated from mothers were tested against autologous plasma Maternal vaccination with HIV Env gp120/160 increased MN gp120-specific IgG binding Maternal HIV Env vaccine regimen did not boost autologous virus neutralization More potent B cell stimulation will be required to elicit autologous nAb responses

To achieve long-term viral remission in HIV-infected children, novel strategies beyond early anti-retroviral therapy (ART) will be necessary. Identifying clinical predictors of time to viral rebound upon ART interruption will streamline the development of novel therapeutic strategies and accelerate their evaluation in clinical trials. However, identification of these biomarkers is logistically challenging in infants, due to sampling limitations and potential risks of treatment interruption. To facilitate identification of biomarkers predicting viral rebound, we have developed an infant rhesus macaque (RM) model of oral SHIV.CH505.375H.dCT challenge and analytical treatment interruption (ATI) after short-term ART. We used this model to characterize SHIV replication kinetics and virus-specific immune responses during short-term ART or post-ATI and demonstrated plasma viral rebound in 5 out of 6 (83%) infants. We observed a decline in humoral immune responses and partial dampening of systemic immune activation upon initiation of ART in these infants. Furthermore, we documented that infant and adult macaques have similar SHIV replication and rebound kinetics and equally potent virus-specific humoral immune responses. Finally, we validated our models by confirming a well-established correlate of time to viral rebound, namely pre-ART plasma viral load, as well as identified additional potential humoral immune correlates. Thus, this model of infant ART and viral rebound can be used and further optimized to define biomarkers of viral rebound following long-term ART as well as to pre-clinically assess novel therapies to achieve a pediatric HIV functional cure. IMPORTANCE Novel interventions that do not rely on daily adherence to ART are needed to achieve sustained viral remission for perinatally infected children who currently rely on lifelong ART. Considering the risks and expense associated with ART-interruption trials, identification of biomarkers of viral rebound will prioritize promising therapeutic intervention strategies, including anti-HIV Env protein therapeutics. However, comprehensive studies to identify those biomarkers are logistically challenging in human infants, demanding the need for relevant non-human primate models of HIV rebound. In this study, we developed an infant RM model of oral Simian/Human Immunodeficiency virus infection expressing clade C HIV Env, and short-term ART followed by ATI, longitudinally characterizing immune responses to viral infection during ART and post-ATI. Additionally, we compared this infant RM model to an analogous adult RM rebound model and identified virologic and immunologic correlates of time to viral rebound post-ATI.