FcRn-enhancing mutations lead to increased and prolonged levels of the HIV CCR5-blocking monoclonal antibody leronlimab in the fetuses and newborns of pregnant rhesus macaques

ABSTRACT

Prenatal administration of monoclonal antibodies (mAbs) is a strategy that could be exploited to prevent viral infections during pregnancy and early life. To reach protective levels in fetuses, mAbs must be transported across the placenta, a selective barrier that actively and specifically promotes the transfer of antibodies (Abs) into the fetus through the neonatal Fc receptor (FcRn). Because FcRn also regulates Ab half-life, Fc mutations like the M428L/N434S, commonly known as LS mutations, and others have been developed to enhance binding affinity to FcRn and improve drug pharmacokinetics. We hypothesized that these FcRn-enhancing mutations could similarly affect the delivery of therapeutic Abs to the fetus. To test this hypothesis, we measured the transplacental transfer of leronlimab, an anti-CCR5 mAb, in clinical development for preventing HIV infections, using pregnant rhesus macaques to model in utero mAb transfer. We also generated a stabilized and FcRn-enhanced form of leronlimab, termed leronlimab-PLS. Leronlimab-PLS maintained higher levels within the maternal compartment while also reaching higher mAb levels in the fetus and newborn circulation. Further, a single dose of leronlimab-PLS led to complete CCR5 receptor occupancy in mothers and newborns for almost a month after birth. These findings support the optimization of FcRn interactions in mAb therapies designed for administration during pregnancy.

Introduction

Mother-to-child transmission (MTCT) of chronic viruses like HIV necessitates lifelong management. For people living with HIV, including pregnant and breastfeeding women, antiretroviral therapy (ART) is the current standard of care. Despite progress, approximately 120,000 children worldwide were newly infected with HIV in 2023.¹ MTCT represents over 95% of HIV-infected pediatric cases.² Without timely treatment, one-third of infants infected at birth succumb to the virus by the age of one, while half of infants facing the same fate will not reach the age of two.³ The development of alternative, effective, and durable prophylactic approaches can help improve the outcomes of at-risk pregnancies.

Over the past years, considerable efforts have been invested in optimizing monoclonal antibodies (mAbs) as preventative therapies against infections through the development of high-concentration formulations⁴ and engineering techniques that extend half-life and reduce the need for frequent dosing,^5–7 These same approaches can be used to optimize mAbs to prevent MTCT. Conceptually, particularly in the context of HIV, where vertical transmission is presumed to occur predominantly during and after labor,⁸ a single dose of mAb administered once by a healthcare provider late in pregnancy could be optimized to engender protection in infants for months after birth. However, the effectiveness of this strategy relies on ensuring that sufficient mAb reaches the fetus.

In primates, the majority of immunoglobulin (Ig) transfer occurs before birth, particularly during the final trimester of pregnancy, when the intake of IgG Abs into the fetus increases exponentially through active transplacental transport.⁹ This transport mechanism is primarily regulated by the neonatal Fc receptor (FcRn),¹⁰ so it is possible that optimizing FcRn-IgG interactions could also enhance the transplacental mAb delivery to the fetus. Notably, numerous FcRn-enhancing mutations have been extensively studied regarding their impact on IgG recycling and consequent prolongation of half-life in primates, a process also governed by FcRn.^5,10 Thus, we hypothesized that FcRn-IgG enhancement would lead to higher levels of therapeutic mAb reaching the fetal compartment.

To test the effect of FcRn enhancement on transplacental mAb transport, we modified the CCR5-blocking mAb leronlimab, termed leronlimab-WT, by incorporating FcRn-enhancing and stabilizing mutations to create leronlimab-PLS.^5,11 Leronlimab is a humanized anti-CCR5-specific IgG4 mAb that has demonstrated promise as a once-weekly subcutaneous HIV therapy in multiple clinical trials as well as preexposure prophylaxis in simian HIV-infected rhesus macaques.^12,13 We administered leronlimab-PLS or leronlimab-WT, along with a reference control mAb, in two groups of three pregnant rhesus macaques each and compared the mAb levels transferred into the fetal circulation. Using this experimental design, we found elevated mAb levels in the fetal compartment in the leronlimab-PLS treatment group compared to the leronlimab-WT group. Consistent with the higher leronlimab-PLS levels in infant circulation, infants in the group that received leronlimab-PLS maintained higher and prolonged CCR5 receptor occupancy (RO). Importantly, the differences in mAb transfer were specific to leronlimab, as we observed no variations in the transfer of the control mAb or the native rhesus Abs between the groups. Our results suggest that engineering mAbs for FcRn-enhancement can be used to optimize fetus-targeted Ab therapies.

Results

Generation of leronlimab-PLS with enhanced FcRn affinity and stability

Our modified version of leronlimab, termed leronlimab-PLS incorporates two sets of mutations: M428L/N434S (LS)¹⁴ for FcRn-IgG enhancement and S228P^11,15 for core-hinge stabilization. The LS mutations were selected because these have consistently increased the half-life of mAbs in monkeys and humans^14,16–18 without modifying mAb thermostability profiles.¹⁹ The S228P mutation (Eu numbering) was incorporated to stabilize the IgG4 heavy chains, which are prone to forming half-mAbs during production and can undergo Fab-arm Exchange (FAE) in vivo.^11,15,20 Furthermore, to address conflicting findings related to the impact of glycosylation on antibody sieving,^21–23 we analyzed the glycan profiles of the two leronlimab variants. Our analysis revealed similar compositions of glycan species for both leronlimab-PLS and leronlimab-WT mAbs (Tables S1 and S2). We next evaluated the thermal stability (Table S4) and the binding affinity of the leronlimab-PLS variant to rhesus FcRn (Fig S1, Table S3). The results confirmed the affinity enhancement effects of LS on the binding interaction with rhesus FcRn. Further, the thermal stability profiles ensured that the incorporated mutations on leronlimab-PLS did not render the molecule unstable. Thus, with the aim of improving mAb delivery to the fetus, we proceeded to test the leronlimab-PLS variant in animals.

Selection of the control rhesus anti-desipramine mAb

To account for the intrinsic differences between individual animals both within and between the groups, we co-administered the rhesus anti-desipramine (DSP) mAb, denoted ‘control mAb’, at the same time as the therapeutic leronlimab mAbs. The control mAb was engineered as a rhesus IgG1 mAb and has been safely administered to dozens of rhesus macaques in different studies.^24–26 Further, the control mAb features a native rhesus Fc domain and complementarity-determining regions (CDRs) based on a mAb targeting the small-molecule drug desipramine. Therefore, desipramine is not naturally present within a host unless specifically administered.^27,28 The rationale for the selection of a non-binder, species-matched IgG1 control was to simplify mAb pharmacokinetics by eliminating target-mediated drug disposition and minimizing the likelihood of anti-drug antibody formation. Lastly, the IgG1 Fc domain is associated with higher transfer rates based on the IgG subclass transfer hierarchy,²⁹ so we anticipated that these characteristics would render the anti-DSP mAb a quantifiable control under the designed dosing and measurement protocols. The reference control mAb was co-administered to all animals to facilitate the measurement and comparison of each animal’s baseline Ab transfer capacity. Thus, we administered both groups a matching intravenous dose of 25 mg/kg of the control mAb.

Experimental design to study transplacental mAb transfer rates

Leronlimab is an IgG4 mAb designed to occlude the CCR5 receptor, which serves as the primary co-receptor during HIV transmission.³⁰ By binding to the CCR5 receptor, leronlimab outcompetes virus-receptor binding and hinders the fusion of HIV with host cell membranes.³⁰ We treated pregnant rhesus macaques to examine the effect of the FcRn-enhancing LS mutations on the placental transfer of leronlimab (Figure 1, Table S5). The rhesus macaque FcRn molecule shares conserved IgG interacting amino acid residues with humans (NM_001257520).^31,32 This structural conservation enables studies of FcRn-mediated placental delivery of human Abs therapeutics.³³ Importantly, rhesus macaques also express a conserved CCR5 receptor on CD4+ T cells, albeit at higher per-cell levels.¹² This enables the evaluation of CCR5 RO levels on blood CD4+ T cells as a measure of the mAb’s binding efficacy to the CCR5 receptor, leronlimab’s mechanism of action.^12,30

Figure 1.

Experimental design to evaluate the transplacental transfer of leronlimab in paired dam-infant rhesus macaques. Six uninfected pregnant macaques received an intravenous (IV) dose of two mAbs at 25 mg/kg each. Each group (N = 3) received the mAb of interest, leronlimab-PLS (group 1) or leronlimab-wt (group 2), and the anti-desipramine (anti-DSP) termed control mAb. The mAbs were infused 14 days prior to the C-section (which was performed at gestational days 164–166). Dam, fetal, and newborn blood samples were collected at the specified time points shown on the timeline. These samples were analyzed to measure recombinant mAb, native Ab, and anti-drug antibody (ADA) levels by ELISA and CCR5 receptor occupancy (RO) by flow cytometric staining. Black circles indicate the type of measurements performed for each time point.

As placental IgG transfer increases exponentially during the final trimester, we aimed to control the timing of mAb administration among dams. Further, we opted to time pregnancy deliveries accordingly, thereby performing Cesarean section (C-section) 2 weeks after mAb administration. Our study included six uninfected pregnant macaques divided into two groups (N = 3), with each group receiving a single 25 mg/kg intravenous (IV) dose of mAb 2 weeks prior to giving birth via C-section. In Group 1, three animals were administered the FcRn-enhanced leronlimab-PLS, while in Group 2, three animals received leronlimab-WT.

Higher levels of leronlimab-PLS mAb reached the fetal compartment

The primary focus of our study was to enhance leronlimab transfer from the dam to the fetus. We used cord blood serum as a surrogate to estimate fetal Ab levels at the time of C-section by enzyme-linked immunosorbent assay (ELISA). Our results reveal greater levels of leronlimab-PLS (approximately 65.7–95.3 μg/mL) in fetal circulation compared to the group receiving leronlimab-WT (approximately 5.6–31.7 μg/mL) at the time of C-section (Figure 2a). Additionally, all dams in Group 1 (leronlimab-PLS) exhibited higher concentrations of leronlimab-PLS (approximately 277.8–375.0 μg/mL) compared to the WT group (approximately 28.1–152.9 μg/mL) at the time of C-section (Figure 2b). For the control mAb, we saw less variability in the mAb concentrations among all dams at the time of C-section (Figure 2d). We consistently observed lower levels of the control mAb and leronlimab (PLS or WT) in all fetuses compared to their respective dams at C-section (Figure 2a–d). However, leronlimab-PLS consistently showed higher concentrations in the fetus compared to both leronlimab-WT and the control mAb.

Figure 2.

Increased concentration of leronlimab-PLS in fetal circulation at C-section. Pregnant macaques were administered a single dose of leronlimab (leronlimab-pls or leronlimab-wt) at 25 mg/kg, and a matching dose of the control mAb two weeks prior to C-section. (a) Fetal leronlimab levels at C-section. (b) Maternal leronlimab levels at C-section. (c) Fetal control mAb levels at C-section. (d) Maternal control mAb levels at C-section. Dam-Newborn pairs are shown with corresponding numbers.

Elevated circulating levels and prolonged CCR5 occupancy in leronlimab-pls treated animals

The LS mutations have been reported to prolong the half-life of numerous mAbs.¹⁴ Consequently, we anticipated observing higher circulating levels of leronlimab-PLS. As expected, dam-infant pairs that received leronlimab-PLS maintained prolonged elevated mAb levels postpartum. Newborns in the leronlimab-PLS group maintained higher mAb levels for up to 30 days postpartum or 44 days post-infusion (~29.09–50.4 μg/mL) compared to newborns in the WT group (~3.8 μg/mL) (Figure 3a). Similarly, all dams in leronlimab-PLS treated group sustained higher levels of mAb (~18.4–116.9 μg/mL) compared to the WT group (~0.5–4.9 μg/mL) at ~44 days post-infusion (study completion) (Figure 3b). Overall, at the time of C-section, 2/3 dams treated with leronlimab-WT, had detectable anti-drug antibody (ADA) levels, while eventually ADA against leronlimab (WT or PLS) was developed for all dams in both treatment groups (Figure 3c,d). A rapid decline of leronlimab-WT was observed in one animal, Dam 6, coinciding with the high and escalating ADA response in that animal (Figure 3b–d). Its offspring, Newborn 6, was the only newborn with detectable ADA levels at the time of C-section, indicating potential ADA transfer from the dam to the fetus (Figure 3c,d). The elevated ADA levels in Newborn 6 aligned with the faster washout of leronlimab-WT observed in that animal (Figure 3a–c).

Figure 3.

Higher circulating leronlimab-PLS and CCR5-receptor occupancy levels in dam-newborn pairs. Leronlimab concentrations in (a) newborn plasma and (b) dam serum. Anti-drug antibody (ADA) responses against leronlimab in (c) newborns and (d) dams. (e) Newborn and (f) dam CCR5 receptor occupancy (RO). (g) ADA responses against the control mAb in newborns and dams. ADA titers were defined as the highest dilution factor of each sample that yielded a positive signal. A positive signal was defined as twice the background. The background was defined as the signal yielded from animal samples not treated with leronlimab or the control mAb. (h) Area under the concentration‐time curve (AUC) analysis of the post-C-section leronlimab-PLS and leronlimab-wt drug exposure. AUC post C-section (day 0-end of treatment) was derived from the leronlimab concentration (µg/mL) versus time (days) graphs for each dam and newborn.

We next examined the CCR5 RO throughout the treatment period (Figure 3e,f). Consistent with the increased levels of leronlimab-PLS, two of three dams in the PLS group maintained prolonged CCR5 receptor occupancy compared to the WT group, achieving near full occupancy approximately 30 days after the C-section (Figure 3f). Likewise, all newborns receiving leronlimab-PLS showed near full receptor occupancy 2-week postpartum (96–98% RO), whereas two of three newborns (newborns 5 and 6) in the WT group had dropped to 75% and 16% RO, respectively (Figure 3e). The decline of both leronlimab (Figure 3a,b) and RO levels (Figure 3e,f) coincided with the appearance of high levels of ADA (Figure 3c,d). To ensure the ADA response was mAb-specific and not due to a cross-reactive response, we also evaluated the ontogeny of ADA against the co-administered control mAb (Figure 3g). It is noteworthy that we observed no development of ADA against the control mAb in any of the newborns and all but one dam (Figure 3g).

Next, we evaluated leronlimab exposure post C-section between the groups (Figure 3h). Overall, dams and newborns in the group receiving leronlimab-PLS treatment showed greater exposure. Ultimately, the levels of administered leronlimab should be sufficient to maintain a high RO in neonates. To understand the differences in the levels of circulating leronlimab needed to afford high RO in dams versus newborns, we compared the ratio of CCR5 occupancy to leronlimab concentrations in the two populations. Our findings suggest that achieving near-full RO in newborns might require lower leronlimab concentrations than dams (Fig. S2a, S2b).

Intrinsic transplacental Ab transfer capacity was similar among groups

We next determined if the higher levels of Abs reaching the fetal compartment were specific to the engineered leronlimab-PLS form and not due to intrinsic differences in the transfer potentials of the animals (Figure 4). The transplacental Ab transfer capacity was defined as the ratio between fetal and maternal Ab concentrations at the time of C-section, termed the transfer ratio. This ratio can vary greatly due to gestation time, individual biology, and even between different pregnancies in the same animals. Native rhesus IgG served to evaluate the animals’ ability to transport natural Abs throughout pregnancy, while the control mAb assessed their capacity to transfer recombinant mAb infused 14 days before C-section.

Figure 4.

Relationship between native rhesus IgG transfer ratio, fetal leronlimab levels, and control mAb ratio. Control mAb and native IgG levels at C-section were quantified to evaluate the capacity for placental transfer of recombinant and native Abs (shown as shaded areas). Percentages indicate the ratio between fetal and maternal concentrations at C-section (transfer ratio). (a) Fetal leronlimab levels at C-section vs. native rhesus IgG transfer ratio. (b) Control mAb ratio vs. Native IgG ratio.

The transfer ratios of native IgG were distributed similarly between the two groups (Figure 4; ranging from 40.7% to 68.1%). This result suggests that neither group had a superior inherent IgG transfer capacity, so the differences in the levels of transferred mAb are likely attributable to the mAb constructs themselves (Figure 4a). Leronlimab transfer ratios to fetuses were much lower than the native Ab, ranging from 8% to 26% of maternal levels without clear differences among the groups (leronlimab-PLS: Pair 1, 26%; Pair 2, 17.5%; Pair 3. 26.4%; leronlimab-WT: Pair 4, 20.7%; Pair 5, 26.4%; Pair 6, 8.3%). It is possible that the 14-day exposure window from mAb administration to C-section was insufficient to allow the equilibration of transferred mAb rates to those of native Abs. To assess that, we measured the transfer of the control mAb, a rhesus IgG1 construct co-infused into all animals at the same dose as leronlimab. The control mAb transferred similarly among all animals and at equivalent rates as the native Abs (Figure 4b), indicating that the 14-day timeframe was sufficient for the administered mAbs to reach transfer rates comparable to the endogenous Abs.

Discussion

Prenatal care offers a critical opportunity to provide clinical interventions that can be life-transforming. During this period, active immunization is widely used as a strategy to protect the mother, fetus, and future infant.³⁴ However, viruses that are not yet preventable through vaccination, such as Zika, HIV, and cytomegalovirus, can be transmitted vertically and pose a lifelong risk to pregnant, lactating, and neonate persons,^35–37 New antiviral mAbs are being quickly developed and deployed in response to emerging viral infections and have the potential to modify the outcomes of infections during pregnancy and early life.^38–42 However, it is unclear if mAbs will require optimized designs to be delivered at protective levels to both mothers and infants.

Our previous study confirmed that therapeutic mAbs administered to dams antenatally can be transported to fetuses.⁴³ However, the administration of a highly potent antiviral human mAb cocktail to Zika virus-infected pregnant macaques, at the time of peak viremia, was insufficient to block viral replication in the fetal compartment.⁴³ Treatment failure could be attributed in part to the administered mAbs not reaching sufficient prophylactic or therapeutic levels in the fetus on time or not being effective in the fetus. These findings suggest that mAb therapies designed for the treatment of pregnant women might need specific optimization to achieve a therapeutic effect. Therefore, we aimed to explore optimization strategies for administering protective mAbs to pregnant, lactating, and infant populations through a single injection administered late in pregnancy.

Receptor-blocking therapeutic mAbs such as leronlimab are often based on the IgG4 subclass due to their ability to attenuate immune activation.²⁰ However, human IgG4s can uniquely form bispecific monovalent molecules in vivo through FAE.^44,45 This naturally occurring change in Ab pairing can functionally modify the IgG4s by decreasing their potential for antigen crosslinking and Fc receptor-mediated crosslinking, among other effects.^44,46 Because this FAE is uncontrolled and unpredictable in both the extent and nature of bridging different specificities,¹¹ this is widely considered a potential liability for pharmacological development because it can complicate drug manufacturing and drug behavior in patients.^20,47 In fact, most new IgG4 therapies control FAE potential with a core-hinge stabilizing S228P mutation.^11,47,48 Importantly, FcRn-mediated effects on IgG rely on the avid binding of one FcRn to each IgG heavy chain.⁴⁹ Thus, we opted for a core-hinge stabilized, LS-mutated, and symmetric IgG4 mAb as a rational and well-controlled engineering approach to enhance the transfer levels of the CCR5-blocking mAb, leronlimab, from mother to fetus.

Leronlimab-WT, atypically, lacks stabilization mutations that prevent FAE, raising the possibility that FAE might influence or even contribute to its safe clinical profile.⁵⁰ For example, if FAE decreases leronlimab-WT’s potential to crosslink CCR5 receptors or super-crosslink CCR5⁺ cells with Fcγ Receptor (FcγR)-expressing effector lymphocytes, it will also diminish the theoretical side effects of receptor agonism and target cell killing. Assessing FAE effects in rhesus monkeys of Indian origin is challenging because the key amino acids at positions 228 and 409 in IgG4s, which are crucial for FAE, are either absent or exhibit allotypic variations.⁵¹ The infused recombinant human IgG4 mAbs would, therefore, be unable to undergo FAE with native IgG4s and remain mostly divalent due to more favorable thermodynamics.^44,52 The absence of random FAE in nonhuman primates, nonetheless, reduces the complexity of the infused mAbs (Figure 1) and simplifies the pharmacokinetics and the interpretation of the mAb transfer outcomes.

We chose to study pregnant rhesus macaques because the pattern of IgG transfer over the course of pregnancy is related to that of humans.⁵³ Further, rhesus macaques also express a conserved CCR5 molecule on CD4+ T cells, which allows us to measure RO, a measure directly related to leronlimab’s proposed mechanism of action. Our results indicated higher levels of leronlimab-PLS reaching the fetal compartment compared to the group receiving leronlimab-WT treatment. Furthermore, these higher levels of leronlimab-PLS persisted in infant circulation for up to 30 days post-C-section. More importantly, the increased mAb levels were reflected in the CCR5-RO levels, as infants in the group receiving leronlimab-PLS sustained prolonged and near full RO for up to 30 days post-C-section, indicating that developing FcRn-enhanced mAbs is an effective strategy to deliver a more robust and sustained mAb presence in the fetus.

Cohorts of pregnant rhesus macaques are valuable and limited in availability, which restricted the group sizes for the experiments. Since numerous factors can influence the ability of Ab transplacental transfer throughout the course of pregnancy,^54,55 to increase the level of confidence in our findings, despite the small group sizes, we rigorously examined other biological factors that could also affect mAb delivery to the fetus and confound interpretation. We focused on controlling the timing of the administration of mAbs in relation to C-section, measuring variations in animal-specific transfer capacities, detecting the presence of ADA responses, and comparing the glycan profiles of the administered mAbs. Measuring animal-specific Ab transfer capacity was critical to control this experiment. If animals with higher transfer capacities were unevenly distributed between the groups, it could easily confound the assessment of each group’s performance.

We demonstrated comparable transfer ratios for both native IgG and the control mAb between the groups, indicating that both groups consisted of animals with similar Ab or mAb transfer potentials. Hence, the effects observed with the leronlimab constructs were likely specific. Likewise, it can be inferred that the elevated levels observed in fetuses, specifically for the leronlimab-PLS construct, are likely not related to individual variations in the transfer capacities across the animals. From our analysis of other factors that impact mAb delivery, we identified one dam (Dam 6) that displayed a heightened ADA response at the time of C-section. The ADA response in this dam might explain the lower leronlimab-WT transfer ratios for that animal (~8%). Apart from that animal, ADA responses did not seem to affect leronlimab transfer into the fetus, as transfer ratios consistently remained around 20% regardless of treatment group and presence of ADA. Together, our data also indicate that higher mAb levels at the time of C-section in dams receiving leronlimab-PLS treatment corresponded with higher mAb levels in the fetus and overall higher mAb exposure post C-section in newborns and dams (Figures 2a–b, 3h). Notably, leronlimab-PLS consistently exhibited higher fetal levels compared to the control mAb of the IgG1 subclass, despite the latter having higher transfer ratios. Interestingly, even though leronlimab-PLS reached higher levels in the fetuses, leronlimab-PLS transfer ratios remained unchanged compared to those of leronlimab-WT. This result might suggest that more complicated mechanisms may influence Ab transfer.

Before maternal Abs reach fetal circulation, they must transverse multiple placental layers.^56,57 These layers contain several receptors speculated to participate in the transport of Ig across the placenta.^21,56,57 Notably, the fact that albumin, a FcRn-regulated protein, does not reach fetal circulation suggests that additional mechanisms govern Ab transport by either enabling Ig transport or blocking the transport of other proteins. Control of albumin transport has been speculated to be the result of unique retrograde recycling pathways or degradation mechanisms regulated by different receptors.^58–61 Although the necessity of FcRn in Ig transport is well-documented,^10,61 FcRn-mediated transport alone does not fully explain how Abs reach fetal circulation.⁵⁶ The presence of FcγRs in certain placental layers has highlighted their potential role in Ab transport.^21,56,57 The involvement of an additional receptor could offer possible explanations for more complex transfer dynamics observed in our study. We observed higher levels of leronlimab-PLS in maternal circulation (Figure 3) due to FcRn recycling, but the same was not observed for transfer ratios. This might suggest that FcRn-enhancing mutations either do not affect transfer kinetics, affect rates bidirectionally or have reached a limit. Alternatively, the Ig interaction with an additional receptor could be a rate-limiting step, possibly the FcγR as proposed by Wessel et al.⁶² Importantly, we conclude that regardless of the specific mechanism, engineering mAbs with LS mutations would be advisable for therapies aiming to achieve higher mAb levels in the fetus.

The transfer of maternal Abs increases within the last trimester of pregnancies, and Ab concentration in the fetal circulation can be even higher than that of the mothers.⁵⁴ The levels of native Abs in the fetus, presumably of maternal origin, are expected to reach close to 100% of maternal concentrations at the time of C-sections. Here, we observed that the transfer ratios of native IgG, transferred throughout the 2nd and 3rd trimesters, and the transfer ratios of the control mAb, administered only 2 weeks prior to C-section, increased similarly, indicating that our control mAb was a good surrogate for measuring intrinsic transfer capacity. However, none of the transfer ratios reached close to 100%, which might be attributed to the possibility of prematurely performing the C-section or inherent disparities between human and rhesus macaques.

The heightened mAb levels identified in dams and newborns in this study could translate to protection against HIV viral challenges for dams and neonates. In the context of HIV, vertical transmission is presumed to occur predominantly during and after labor.⁸ Therefore, maintaining elevated mAb levels during and after this period, along with prolonged RO, may be adequate for containing the infection. Supporting this hypothesis, preexisting Abs targeting the C5 region of the HIV envelope were associated with delayed HIV acquisition among infants exposed to the virus, as well as reduced circulating viral loads in infants who did acquire HIV.⁶³ While our study did not directly test protection against viral challenges, CCR5 blockade by leronlimab does not necessitate additional effector or immune components from the host. Therefore, measuring RO provides a reliable, albeit not exclusive, indirect estimate of its efficacy, especially since our measurements were limited to blood CD4+ T-cells. Additionally, the fact that leronlimab functions independently of immune effector molecules may confer a distinct advantage over other anti-HIV immunotherapies by not requiring host factors. This significant difference is especially relevant in immunocompromised and neonate populations with compromised or immature immune systems. Ultimately, however, we will evaluate the protection of newborns against viral infection afforded by the in-utero leronlimab-PLS transfer in our follow-up study to confirm treatment efficacy.

MAb-based therapies possess several attributes that could be advantageous for treating and preventing congenital diseases. MAbs are well-tolerated, highly specific, and utilize natural mechanisms to transfer immunity to the fetus. Supporting this notion, subcutaneous administration of broadly neutralizing Abs, such as VRC01, VRC01-LS, and VRC07-523LS, has been demonstrated to be safe in neonates exposed to HIV.⁶⁴ Moreover, mAbs, due to their distinct mechanism of protection compared to antiretroviral drugs, may potentially mitigate the risk of viral resistance. In areas with limited access to perinatal care, long-acting therapeutic interventions are particularly valuable. A single injection of a long-acting mAb could potentially provide coverage from late pregnancy through early infancy, reducing the need for frequent healthcare visits.

The World Health Organization and the Joint United Nations Program have identified the development of durable HIV interventions as a primary public health objective.⁶⁴ Recent advancements in mAb therapies have introduced numerous engineering techniques to prolong half-life and reduce the necessity for frequent dosing. Our study findings suggest that engineering mAbs with FcRn-enhancing mutations, such as the LS mutations, may enhance the effectiveness of fetus-targeted mAb therapies by achieving higher mAb levels in the fetal compartment and maintaining prolonged therapeutic levels. This approach, aimed at improving mAb access to the fetal compartment, could facilitate the management of various emerging and established congenital diseases. Thus, optimization of mAb-based therapies holds promise for improving health outcomes for both mothers and newborns.

Materials and methods

Animal care and procedures

All animals were rhesus macaques (Macaca mulatta) born and raised in the breeding colony of the California National Primate Research Center (CNPRC), which is negative for type D retrovirus, simian immunodeficiency virus (SIV), and simian T-lymphotropic virus type 1. The CNPRC is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC). Animal care was performed in compliance with the 2011 Guide for the Care and Use of Laboratory Animals provided by the Institute for Laboratory Animal Research. The study was approved by the Institutional Animal Care and Use Committee of the University of California, Davis.

Six dams, pregnant via timed breedings, were randomly assigned to the two study groups (N = 3 per group). Fourteen days prior to scheduled C-sections, animals were infused intravenously with two mAbs at 25 mg/kg each, either leronlimab-PLS (Group 1) or leronlimab-WT (Group 2); both groups received the control anti-DSP mAb. To reduce the risk of anaphylactic reactions, animals were pre-treated with diphenhydramine (4 mg/kg, intramuscular (IM)) approximately 30 min prior to the mAb infusions. C-sections were performed at estimated gestational days 164–166 to deliver the newborn and collect cord blood. After C-section, infants were reared by their respective dams, except for two infants that were rejected by their dams: infant 5 was accepted by dam 4 (who thus raised two infants simultaneously), and infant 1 was nursery reared.

When necessary, animals were immobilized with ketamine HCl (10 mg/kg intramuscularly), or for better immobilization (such as for mAb infusions) with dexmedetomidine (0.0075 to 0.0150 mg/kg intramuscularly) with atipamezole reversal. Blood was collected at defined time points via venipuncture. EDTA-anticoagulated blood was collected for complete blood counts with differential counts (performed by Clinical Laboratories) and for isolation of plasma and peripheral blood mononuclear cells (PBMC) according to standard methods. Blood without anticoagulant was used to get the serum. All serum, plasma, and PBMC samples were cryopreserved and shipped on dry ice to the laboratories for further analysis.

Leronlimab CCR5 RO

We measured the percentage of CCR5 RO on the surface of CD4+ T cells as previously described.⁶⁵ Samples were washed twice with phosphate-buffered saline (PBS) and divided evenly into three tubes. Unconjugated leronlimab (5 µg/mL of parental human leronlimab) was added to tube 3 and incubated for 30 min at room temperature in the dark. Cells were washed once with PBS, and then anti-human IgG4 FITC (Sigma Aldrich, clone HP-6025) was added to tubes 2 and 3 and incubated for 30 min at room temperature in the dark. Cells were then washed once with PBS + 10% fetal bovine serum (FBS) followed by three additional washes with PBS. Cells in all three tubes were then stained for surface markers and viability and incubated for 30 min at room temperature. Cells were finally washed twice with PBS + 10% FBS and then 100 μL of 2% paraformaldehyde (PFA) was added to fix the cells at 4°C for at least 10 min before running on a FACSymphony^TM A5 flow cytometer (BD Biosciences). Flow analysis was done using FlowJo 10.4. Staining tube 1 served as the fluorescence minus one (FMO) control to assist with gating on desired cell populations and for background subtraction, tube 2 stained for the frequency of leronlimab-occupied CCR5 receptors on CD4+ T cells, and tube 3 was saturated with leronlimab ex vivo to measure for the total CCR5 receptors on CD4+ T cells. The equation used to calculate CCR5 RO is as follows:

$$\mathit{\%RO}=\frac{\mathit{\%IgG}4(\mathit{tube}2-\mathit{tube}1)}{\mathit{\%IgG}4(\mathit{tube}3-\mathit{tube}1)}\times 100\mathit{\%}$$

Antibodies against the following antigens were used: CD3 AF700 (BD Biosciences, clone SP34–2, RRID:AB_396938), CD4 BUV395 (BD Biosciences, clone L200, RRID:AB_2738596), CD8 BUV737 (BD Biosciences, clone SK1, RRID:AB_2870086), CD45 PE-Cy7 (BD Biosciences, clone D058–1283, RRID:AB_10612014), and CCR5 APC (BioLegend, clone J418F1, RRID:AB_2564073), Live/dead Fixable near-IR (Invitrogen).

Leronlimab-WT, leronlimab-PLS and anti-DSP mAb

“Leronlimab-WT” was provided by CytoDyn Inc. “Leronlimab-PLS” and “anti-DSP” were produced recombinantly by the Nonhuman Primate Reagent Resource (NHPRR, RRID:SCR_012986). Chinese hamster ovary (CHO) stable cell lines were generated to express the control rhesus anti-DSP mAb and the PLS-modified leronlimab using the methionine sulfoximine selection system for cell-line generation. Cells expressing each mAb were harvested on different days according to their viability. Supernatants were collected, clarified, and sterile-filtered using the Millistak POD filters, 2.2 m² D0HC & 1.1 m² X0HC (Millipore). The mAbs were purified using Protein A Amsphere A3 protein A resin (JSR Life Sciences). The pH of the eluted mAb was immediately lowered to 3.50 using 4 M acetic acid and incubated for 60 min for viral inactivation. After an hour, the mAb was neutralized to pH 6.0 using 2 M Tris and was stored at 2–8°C. After low-pH treatment, an anion exchange chromatography membrane was used as a polishing step. The mAbs were concentrated and the buffer was exchanged into the desired formulation using an ultrafiltration/diafiltration step with a 30 kDa Biomax A tangential flow filtration membrane. The control mAb was formulated in 20 mM Citrate, 150 mM Sodium Chloride, 0.02% Polysorbate 80, pH 6.0. Leronlimab-PLS was formulated in 20 mM histidine, 150 mM sodium chloride, pH 6.0. The concentration was determined by spectrophotometer at A280 nm using the appropriate extinction coefficient. The percentage of IgG monomers and higher molecular weight species was measured by size exclusion-high-performance liquid chromatography. Purity with respect to endotoxin content was measured using a kinetic chromogenic Limulus amebocyte lysate assay. A bioburden test was also performed for quality control.

Rhesus anti-drug antibody assay

ELISA was used to detect ADA responses against leronlimab and the control mAb in the animals. Briefly, high-binding ELISA plates (Corning, Cat No. 9018) were coated with 0.1 μg/well of leronlimab-PLS, leronlimab-WT, or control mAb overnight. Plates were washed three times using 1×PBS, 0.05% Tween 20 (PBST) and blocked with 5% milk in PBST for 1 h at 37°C. Serum samples, run in duplicate, were serially diluted in 5% milk in PBST and tested against the plates coated with leronlimab (PLS or WT) or coated with the control mAb. Plates were incubated for 1 h at 37°C. Plates were then washed three times before adding the anti-lambda light-chain detection Ab (SouthernBiotech, Cat No. 2070–05, RRID:AB_2795753), to detect the presence of ADA containing lambda light chains. Lambda detection Ab was added at a 1:5,000 dilution in all wells. An anti-lambda detection Ab was selected to avoid unwanted cross-reactivity with leronlimab (PLS or WT) or the control mAb which are engineered with kappa light chains. Plates were incubated for 1 h at 37°C and washed three times before developing with 100 µl/well of One Step TMB substrate (ThermoFisher, Cat No. 34029). The reaction was stopped after 3 min using 100 µl/well of Stop Solution (Life Technologies, Cat No. SS04). Plates were read on E-max Precision Microplate Reader (Molecular Devices) at 450 nm. The presence of ADA was determined relative to ADA-positive and ADA-negative controls from previous animal studies. As a positive control for the lambda detection, we re-engineered the variable regions of a mouse anti-rhesus kappa chain (anti-rhesus kappa chain [kappitan], Nonhuman Primate Reagent Resource (NHPRR), Cat No. PR-2710, RRID:AB_2819300), to express as a rhesus IgG1 chimeric mAb with rhesus lambda light chains (HQ214073.1). This newly developed reagent, therefore, expresses a rhesus lambda chain and binds specifically to the kappa chains of the coating reagent. The positive control for lambda detection was designated rhesus lambda anti-IGK [adala] (NHPRR, Cat No. PR-0122, RRID:AB_3086856). ADA titers were defined as the highest dilution factor of each sample that yielded a positive signal. A positive signal was defined as twice the background, where the background was determined as the signal obtained from samples of animals that did not receive either leronlimab or the control mAb.

Quantification of leronlimab-WT and leronlimab-PLS

The same ELISA method was used to quantify both versions of leronlimab (leronlimab-WT and leronlimab-PLS) in maternal serum and newborn plasma post-C-section. MAb concentrations were determined using human-specific reagents. Briefly, high-binding ELISA plates (Corning, Cat No. 9018) were coated with 0.5 μg/well Goat anti-Human IgG antibody (SouthernBiotech, Cat No. 2049–01, RRID:AB_2795694) overnight. Plates were washed three times using PBST and blocked with 5% milk in PBST for 1 h at 37°C. A leronlimab standard was diluted in 5% milk in PBST ranging from 0.3 μg/mL − 1.46 × 10⁻⁴ μg/mL. Serum samples were diluted in 5% milk in PBST. Each animal sample was run in triplicate. Serum dilutions were adjusted depending on estimated leronlimab concentrations in the serum. Overall, the animal receiving the leronlimab-PLS required higher serum dilutions for quantification, indicating higher serum leronlimab-PLS concentrations. Rhesus serum inoculated with rhesus control mAb was used as a blank matrix sample. After washing, 1:50,000 dilution of mouse anti-Human IgG4 pFc-HRP (SouthernBiotech, Cat No. 9190–05, RRID:AB_2796684) was added to all wells and incubated for 1 h at 37°C. Plates were then washed three times and were developed using 100 µl/well of One Step TMB substrate (ThermoFisher, Cat No. 34029). The reaction was stopped after 5 min using 100 µl/well of Stop Solution (Life Technologies, Cat No. SS04). Plates were read on E-max Precision Microplate Reader (Molecular Devices) at 450 nm.

Quantification of control rhesus anti-DSP

ELISA was used to quantify the control mAb (NHPRR, control antibody rhesus IgG1 (anti-DSP) [DSPR1], Cat No. PR-1117, RRID:AB_2716330) concentration in maternal serum and cord blood at the time of C-section. Briefly, high-binding ELISA plates (Corning, Cat No. 9018) were coated with 1 μg/well anti-DSPR1 idiotype [idiotypeDSPR1] (NHPRR, Cat No. PR-4500, RRID:AB_2819286) overnight. Plates were washed three times using PBST and blocked with 5% milk in PBST for 1 h at 37°C. The control mAb standard was diluted in 5% milk in PBST ranging from 4 μg/mL–1.95 × 10⁻³ μg/mL. Serum samples were diluted in 5% milk in PBST. Each animal sample was run in triplicate. Serum dilutions were subjected to changes depending on estimated control mAb concentrations in serum. Rhesus serum inoculated with human leronlimab at different concentrations was used as the blank matrix sample. After washing, 1:10,000 dilution of Goat anti-Human Kappa (SouthernBiotech, Cat No. 2060–05, RRID:AB_2795720) was added to all wells and incubated for 1 h at 37°C. Plates were then washed three times and were developed using 100 μl/well of One Step TMB substrate (ThermoFisher, Cat No. 34029). The reaction was stopped after 5 min using 100 μl/well of Stop Solution (Life Technologies, Cat No. SS04). Plates were read on an E-max Precision Microplate Reader (Molecular Devices) at 450 nm.

Quantification of native rhesus IgG

ELISA was used to quantify native rhesus IgG concentration in maternal serum and newborn plasma post-C-section as described above with the following modifications. Briefly, high-binding ELISA plates (corning, Cat No. 9018) were coated with 1 μg/well of anti-Rhesus IgG1/3 [1B3] (NHPRR, Cat No. PR-1230, RRID:AB_2819287) overnight. Plates were washed three times using PBST and blocked with 5% milk in PBST for 1 h at 37°C. Purified rhesus polyclonal IgG standard was diluted in 5% milk in PBST ranging from 10 μg/mL–5.64 × 10⁻⁵ μg/mL. Serum samples were diluted in 5% milk in PBST. Each animal sample was run in duplicates. Different concentrations of the human leronlimab and the rhesus control mAb were used to test the reactivity of rhLambda-IgG Assay against the administered mAbs. To minimize cross-reactivity with the control rhesus anti-DSP mAb that is engineered with a kappa light chain, an anti-human lambda detection reagent was used. After washing, 1:5,000 dilution of Goat Anti-Human Lambda-HRP (SouthernBiotech, Cat No. 2070–05) was added to all wells and incubated for 1 h at 37°C. Plates were then washed three times and were developed using 100 µl/well of One Step TMB substrate (ThermoFisher, Cat No. 34029). The reaction was stopped after 5 min using 100 µl/well of Stop Solution (Life Technologies, Cat No. SS04). Plates were read on an E-max Precision Microplate Reader (Molecular Devices) at 450 nm.

Data analysis

Data analysis was conducted using GraphPad Prism software. Median absorbance readings for each sample were analyzed using the four-parameter logistic regression model (4PL). Interpolated values within the quantifiable range of the standard curve were multiplied by the equivalent dilution factor to quantify the mAb concentration in study samples.

For the quantification of leronlimab and control mAb, the limit of quantitation was defined as the lowest concentration of mAb that yielded absorbance readings twice that of the background. In the leronlimab assay, the background was determined as the absorbance readings at 450 nm of rhesus serum from animals not administered leronlimab or control mAb, diluted in 5% milk in PBST and inoculated with the control mAb. Conversely, in the control mAb quantification assay, the background was determined using the same method, but rhesus serum was inoculated with leronlimab instead. In the native IgG quantification assay, the background was determined as the absorbance readings at 450 nm yielded by leronlimab and the control mAb.

AUC post C-section (Day 0-end of treatment) was calculated using the trapezoidal analysis method in GraphPad. The AUC was derived from the leronlimab concentration (µg/mL) versus time (days) graphs for each dam and newborn.

Funding Statement

This study was funded by the United States Department of Defense (DOD) under award number W81XWH-18-1-0094. Research reported in this study was also supported by The National Institute of Allergy and Infectious Diseases (NIAID) and The Office of Research Infrastructure Programs (ORIP) of the National Institutes of Health under award numbers U24AI126683, P40OD028116, P51OD011107. This work was supported by the National Institutes of Health grants AI112433, AI170547, AI1666969, and AI129703 awarded to J.B.S., and P51 OD011092 awarded to the Oregon National Primate Research Center (ONPRC). “Leronlimab - WT” was provided by CytoDyn Inc. Reagents used in these studies were provided by the NIH Nonhuman Primate Reagent Resource (P40OD028116, U24AI126683)

Abbreviations

Anti-drug antibody

ADA

Anion exchange chromatography membrane

AEX

Antibodies

Abs

Antiretroviral therapy

ART

Area under the concentration-time curve

AUC

Cesarean section

C-section

Chinese hamster ovary

CHO

Complementarity-determining regions

CDRs

Desipramine

DSP

Differential scanning fluorimetry

DSF

Enzyme-linked immunosorbent assay

ELISA

Fab-arm exchange

FAE

Fetal bovine serum

FBS

Fluorescence minus one

FMO

Immunoglobulin

Ig

Intravenous

IV

M428L/N434S

LS

Monoclonal antibody

mAb

Mother-to-child transmission

MTCT

Neonatal Fc receptor

FcRn

Paraformaldehyde

PFA

Peripheral blood mononuclear cells

PBMC

Receptor occupancy

RO

Simian immunodeficiency virus

SIV

Surface plasmon resonance

SPR

The Nonhuman Primate Reagent Resource

NHPRR

Wild-type

WT

Disclosure statement

J.Z., J.B.S., and D.M. are authors of intellectual property documents related to this manuscript. J.B.S. has a significant financial interest in and serves on the scientific advisory board of CytoDyn, Inc., a company that may have a financial interest in the results of this research and technology. This potential individual conflict of interest has been reviewed and managed by Oregon Health and Science University.

Grants

DOD W81XWH-18-1-0094, P40 OD028116, U24 AI126683 to D.M.M.

P51 OD011107 to CNPRC

AI112433, AI170547, AI1666969, and AI129703 to J.B.S., and P51OD011092 to ONPRC

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19420862.2024.2406788