Durable Immunity to Ricin Toxin Elicited by Intranasally Administered Monoclonal Antibody–Based Immune Complexes

Ricin toxin (RT) is a heterodimeric glycoprotein derived from the castor bean (Ricinus communis). The toxin’s extreme lethality following injection or inhalation, coupled with the fact that there are no preventive or postexposure therapeutics for ricin, has earned it classification as a category B biothreat by the Centers for Disease Control and Prevention (1, 2). At the cellular level, RT’s galactose/galactosamine-specific binding subunit, RTB, promotes toxin attachment to surface-exposed glycoproteins and glycolipids. RT is then internalized via endocytosis and trafficked retrograde to the trans-Golgi network and endoplasmic reticulum (ER). Within the ER, RT’s enzymatic subunit, RTA, is liberated from RTB and translocated into the cell cytoplasm where it functions as an extraordinarily efficient ribosome-inactivating protein (3, 4). Ribosome damage activates multiple stress-activated protein kinase pathways involved in inflammation and programmed cell death.

In a mouse model, RT exposure by injection results in rapid weight loss, hypoglycemia, organ failure, and death within 24–72 h depending on dose (5, 6). Inhalation of RT results in a severe acute inflammatory response marked by ablation of alveolar macrophages (AMs) and an overwhelming influx of inflammatory cells (namely polymorphonuclear cells) within the lungs (7). AMs are notoriously sensitive to the effects of RT and are likely primary drivers of toxin-induced pathophysiology in the lung by virtue of their ability to secrete an array of proinflammatory cytokines and chemokines that not only promote polymorphonuclear neutrophil recruitment, but also may sensitize epithelial cells to RT-mediated killing (8, 9). Indeed, AMs and other cells contribute to the “cytokine storm” that propagates local inflammation and tissue destruction (10–12) that ultimately culminates in acute respiratory distress syndrome (13, 14).

Despite RT’s extreme pulmonary toxicity, we have demonstrated in a mouse model that intranasal (i.n.) coadministration of RT with a mAb mixture abrogates RT-induced morbidity (e.g., weight loss, hypoglycemia), pulmonary inflammation (e.g., IL-1, IL-6 in bronchoalveolar lavage fluids), AM cell death, and gross tissue pathology (12). The mAb mixture consists of two murine IgGs, PB10, directed against RTA, and SylH3, directed against RTB (15, 16). The combination of PB10 and SylH3 was more potent at RT neutralization in the mouse model then either of the mAbs alone (11, 12). The same PB10/SylH3 mixture also conferred immunity against lethal dose RT challenge by injection, and protected liver sinusoidal epithelial cells (LSECs) and Kupffer cells from RT-induced damage ex vivo (17).

Considering the ability of the PB10/SylH3 mixture to neutralize RT in in vitro, ex vivo, and in vivo, we assumed that RT–mAb immune complexes (RICs) are inert and simply cleared from the lungs by mechanical forces and/or degraded by AMs without consequence. However, the fate of immune complexes in the lungs in general has not been studied in detail, and there is evidence to suggest that such complexes may in fact be immunostimulatory (18–21). For that reason, we chose to investigate the fate of RICs in mice following both i.n. instillation and i.p. injection.

RT (R. communis agglutinin II; RCA₆₀) was purchased from Vector Laboratories (Burlingame, CA) and dialyzed against PBS at 4°C in 10,000 Da molecular mass cutoff Slide-A-Lyzer dialysis cassettes (Thermo Fisher Scientific, Pittsburgh, PA) prior to use in mice. Unless noted otherwise, all reagents were purchased from Sigma-Aldrich (St. Louis, MO).

Murine mAbs listed in Table I (16, 22–25) were purified by protein G chromatography by the Dana Farber Cancer Institute’s mAb Core facility (16, 22). Mouse and human chimeric IgG1 and LALA (leucine to alanine) derivatives of PB10 and SylH3 were provided by Mapp Biopharmaceutical (San Diego, CA).

Mouse studies were conducted under strict compliance with the Wadsworth Center’s Institutional Animal Care and Use Committee (IACUC). Female BALB/c and ΔFcγR (C.129P2(B6)-Fcer1g^tm1RavN12) mice 7–8 wk of age were purchased from Taconic Biosciences (Rensselaer, NY). For RIC immunizations, RT (1 μg) was mixed with 40 μg total of mAb (20 μg each for mAb combinations) and administered to mice in a final volume of 40 μl i.n. or 200 μl i.p. For challenges, RT (5 × LD₅₀; 1 μg; unless otherwise noted) was administered to mice in a volume of 40 or 200 μl for i.n. or i.p. delivery, respectively. Following RIC or RT challenge, mice were monitored daily for at least 7 d for symptoms of RT intoxication, including weight loss, low blood glucose levels, and clinical signs of morbidity. Morbidity was measured using an IACUC-approved grading sheet stratified on a scale from 0 to 3, with 0 indicating normal activity and appearance, 1 indicating ruffled fur, 2 indicating hunching/solitary nesting/inactivity when cage opened/mild dyspnea, and 3 indicating inactivity when handled, rapid heartbeat, and/or labored breathing. Mice were euthanized by CO₂ asphyxiation followed by cervical dislocation when they exceeded predetermined thresholds for weight loss, blood glucose levels, or physical signs of morbidity. Blood was procured from the submandibular vein (5).

Ninety-six–well microtiter plates (Thermo Fisher Scientific) were coated overnight with RT (0.1 μg/well in PBS) at 4°C. The plates were then washed and blocked as previously described (26) before the addition of mouse sera or mAbs. For detection of mAbs in serum, the linear PB10 peptide E12 was coated on plates at a concentration of 25 μg/ml. Sera and mAbs were diluted in PBS. HRP-labeled goat anti-mouse IgG- and IgM-specific polyclonal Abs (SouthernBiotech, Birmingham, AL) were used as secondary reagents. Plates were developed using 3,30,5,50-tetramethylbenzidine (Kirkegaard & Perry Laboratories, Gaithersburg, MD) and analyzed with a SpectraMax iD3 spectrophotometer equipped with SoftMax Pro 7.1 software (Molecular Devices, Sunnyvale, CA).

MagPlex-C microspheres (Luminex, Austin, TX) were coupled to RT as follows: microspheres were washed with activation buffer (0.1 M monosodium phosphate, pH 6.2) and activated by adding 50 mg/ml N-hyroxysulfosuccinimide and EDC (1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride; Thermo Scientific Pierce). RT (5 μg) was mixed with 1 × 10⁶ beads in coupling buffer (0.5 M MES, pH 5.0). Coupled beads were diluted in storage buffer (PBS with 1% BSA, 0.02% Tween 20, 0.05% azide, pH 7.4) to 2.5 × 10⁶ beads/ml. Serum samples were diluted 1:100 (v/v) in PBN buffer (PBS, 1% BSA, 0.05% sodium azide, pH 7.4) on nonbinding 96-well plates (Greiner Bio-One, Monroe, NC) along with 25 μl of beads (2500 beads/well). Samples and beads were incubated for 1 h in the dark at room temperature with shaking (300 rpm). Samples were then washed three times using wash buffer (PBS, 2% BSA, 0.02% Tween 20, 0.05% azide, pH 7.5) and incubated with 50 μl of PE-tagged goat-anti mouse IgG1, IgG2a, IgG2b, or IgG3 (Invitrogen, Thermo Fisher Scientific) as above. Plates were washed as described above. Beads were resuspended in 90 μl of xMAP sheath fluid (Luminex, Austin, TX) and incubated for 1 min. Samples were analyzed using a FlexMap three-dimensional instrument (Luminex). Values are reported as median fluorescence intensity.

Statistical analyses of weights and titers were carried out using GraphPad Prism 9.2.0, and survival analyses were performed in R 4.1.1 (https://www.R-project.org/.), as well as the R package survival 3.2-13 (https://CRAN.R-project.org/package=survminer). Survival distributions following RT challenge were compared between groups using a log-rank test. When performing pairwise survival comparisons, the resulting p values from log-rank tests were adjusted using the Benjamini–Hochberg method. When comparing weights between two groups, data were analyzed with Mann–Whitney U tests. When comparing weights between more than two groups, data were analyzed by mixed effects modeling followed by Dunnett’s multiple comparisons tests. Unpaired, two-tailed Welch’s t tests were performed to determine the significance of differences in end-point titers between groups. Finally, morbidity scores were compared between groups with Mann–Whitney U tests. In all cases, the following apply: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p < 0.0001.

To examine the immunostimulatory capacity of ricin–mAb immune complexes, groups of adult BALB/c mice were administered RICs i.n. on study day 0 and then monitored for up to 3 mo for the presence of RT-specific serum IgG. The RICs consisted of 1 μg of RT (∼5 × LD₅₀) in combination with ∼20-fold molar excess of PB10 (20 μg) and SylH3 (20 μg). The dose of 40 μg of mAb (roughly 2 mg/kg) was chosen because it was previously shown to be the minimum amount of mAb required for protection against respiratory RT exposure (27, 28). Groups of control animals received RT (1 μg) or PBS. For i.n. delivery, RICs (40 μl total) were administered as droplets to the nares of mice under isoflurane anesthesia. Mice were monitored for 7 d following RIC administration and scored for signs of distress (e.g., ruffled fur, dyspnea), weight loss, and hypoglycemia. We also established cohorts of mice that received RICs by the i.p. route.

Mice that received RICs by either the i.p. or i.n. route did not experience any measurable weight loss or display any outward signs of intoxication. In contrast, mice that received RT alone underwent a dramatic reduction in body mass and experienced hypoglycemia within hours. RT-treated mice expired (or were euthanized) within 2 d following i.p. exposure and 3 d following i.n. exposure (Fig. 1A, 1B). These results confirm the capacity of PB10 and SylH3 to attenuate RT when administered mucosally or systemically (11, 12).

In both the i.n. and i.p. RIC-treated groups of animals, RT-specific serum IgG was detectable within ∼7 d (Fig. 1E, 1F). However, in neither case was there measurable RT-specific IgM or IgA at any time point assessed (e.g., 15 or 30 d postinoculation) (Supplemental Fig. 1A, 1C, 1D). To determine whether the early IgG Abs were derived de novo (endogenous) or residual PB10/SylH3 mAbs from the RICs, we evaluated serum for Ab reactivity to PB10’s linear epitope (peptide E12) by ELISA. Measurable anti-E12 Ab titers were diminished within 1 wk (Supplemental Fig. 1B), indicating that IgG from 7 d onward is endogenous. In the i.p. immunized animals, serum IgG titers rose until day ∼60, then plateaued or declined, whereas the Ab titers in the i.n. immunized mice continued to rise until day 90, at which point the study was concluded. In the i.p. immunized mice, the maximal endpoint titer achieved was 1:6300 (on day 60), whereas the i.n. immunized mice had maximal RT-specific serum IgG titers of 1:35,100 (on day 90). Analysis of serum by Luminex revealed that in the case of both i.n. and i.p. immunization, IgG is predominantly IgG1 with lower levels of IgG2a and IgG2b and even lower but still detectable levels of IgG3 (Supplemental Fig. 2).

To determine whether RICs stimulate protective immunity to RT and how the immunization route impacts protection, groups of mice were immunized with RICs by the i.n. or i.p. route and then challenged by the same or opposite route on day 90 with 5 × LD₅₀ RT. RT-specific Ab titers in sera were monitored roughly every week beginning 1 wk prior to RIC administration and every month after the first 30 d following immunization. Routine survival and weight loss monitoring was performed for 1 wk following vaccination (Supplemental Fig. 4).

Mice administered RICs by the i.n. route survived lethal RT challenge, irrespective of the challenge route, albeit with varying degrees of morbidity (Fig. 2B). Mice challenged i.n. displayed mild signs of illness and experienced ∼10% weight loss during the first 3 d, but then rebounded beginning on day 4 (Fig. 2D). Groups of mice immunized i.n. and then challenged by the i.p. route survived RT challenge with minimal or no measurable weight loss or morbidity.

Animals that received RICs by the i.p. route and challenged by that same route survived RT challenge without any overt signs of illness and with minimal or no weight loss. However, mice that received RICs by the i.p. route and were challenged i.n. fared less well. One animal in the group died and the remainder lost ∼15% of their initial body weight (Fig. 2A, 2C). We conclude from these studies that a single i.n. or i.p. dose of RICs to mice results in active immunization and protection against lethal dose ricin challenge.

The disparities in weight loss, clinical illness, and survival between groups of mice challenged i.n. prompted us to investigate how prechallenge anti-RT serum IgG titers compared between mice that received RICs i.n. or i.p. RT-specific serum IgG titers in mice administered RICs by the i.p. route rose until around day 30, after which they plateaued (average reciprocal endpoint titers ∼3500) and persisted up to the time of challenge. RT-specific serum IgG levels in mice treated with RICs by the i.n. route continued to rise and reached an endpoint of ∼30,000 just before RT challenge (Fig. 2E, 2F), revealing that i.n. exposure to RICs resulted in a 10-fold higher Ab response than the same dose given systemically (i.p.). One day prior to challenge, mice immunized with RICs by the i.n. route had toxin-neutralizing titers of 905 (±624), whereas those receiving RICs i.p. had neutralizing titers of 150 (±107). This ∼6-fold difference in toxin neutralization assay (TNA) between the i.n. and i.p. treated mice suggests that i.n. delivery not only accelerated the onset of total anti-ricin IgG but also influenced the maturation of toxin-neutralizing Abs.

To determine whether differences in RT-specific serum IgG titers at the time of RT challenge explained the observed discrepancies in morbidity and mortality, we performed an additional study in which mice were challenged with RT on day 30, at which time RT-specific endpoint titers were essentially equivalent between the i.n. and i.p. immunized groups of mice (Supplemental Fig. 3). Following 5 × LD₅₀ RT challenge by the i.n. or i.p. routes, mice that had been administered RICs by the i.n. route fared well: they lost only ∼10% of their starting body weight and displayed only mild signs of illness. The exception was a single mouse that had been immunized i.n. and challenged i.p. that succumbed to RT intoxication.

Mice that received RICs by the i.p. route and challenged by the same route experienced little to no weight loss or signs of illness. However, mice administered RICs by the i.p. route and then challenged by the i.n. route fared less well, with one decedent and the remaining animals experiencing ∼15% reduction in starting body weight with mild symptoms of intoxication (e.g., ruffled coat, hunching). These results suggest that i.n. RIC delivery is superior to i.p. delivery in eliciting immunity to a subsequent i.n. ricin challenge.

We speculated that the active immunization elicited by RICs might simply be the result of delivering an attenuated (low) dose of RT, rather than immune complexes per se. To investigate this question, groups of mice were i.n. administered a low dose of RT (0.2 μg; ∼1 × LD₅₀) without or with the PB10 and SylH3 mAb mixture. Importantly, the amount of RT used in these experiments was 5-fold less than that used to generate the RICs in the experiments described up to this point in the study. Mice that received RT alone lost 10–20% of their initial body weight during 3–4 d and displayed moderate signs of illness (Fig. 3B). Ultimately, 12 out of 15 mice expired within 7 d, indicating that the actual dose of RT the mice received was slightly higher than the target dose of 1 × LD₅₀ (Fig. 3A). In contrast, mice that received the same dose of RT complexed with PB10 and SylH3 did not experience weight loss or display signs of morbidity (Fig. 3B, 3C). In the RIC-treated mice, RT-specific serum IgG titers were detectable by day 7 and persisted until the end of the study. This contrasts with RT-treated mice whose toxin-specific serum IgG titers were not detectable until study day 30 (Fig. 3D). We conclude that the immunostimulatory activity of RICs is not simply a result of low-dose toxin exposure.

To examine whether other RT-specific mAbs besides PB10 and SylH3 stimulate the onset of RT-specific Ab responses, RT was mixed with various combinations of RTA- and RTB-specific IgG mAbs (20 μg each/40 μg of total mAb) that differ in their relative binding affinities and neutralizing capacities (Tables I, II). For example, we combined PB10 and SyH7, two toxin-neutralizing mAbs that recognize spatially distinct epitopes on RTA (22). We also combined PB10 with two different anti-RTB mAbs; 8B3, which has moderate binding affinity and moderate TNA, and LC5, which has weak binding affinity and no TNA (16). Finally, RT was complexed with PB10 alone. The five different RICs were administered to groups of mice by the i.n. route. Mice were monitored daily for weight loss and survival for a period of 13 d. Serum was collected from animals on days 7, 15, 35, 50, 80, and 113 and assessed for RT-specific IgG and TNA (day 113).

Mice that received RICs consisting of any two combinations of different mAbs, regardless of their epitope specificities or TNA, survived RICs treatment. In contrast, mice that received RT alone expired within 72 h, while a single mouse in the PB10-only RIC group died on day 10 (Fig. 4). There was a clear stratification in weight loss based on RIC composition. Mice that received the PB10/SylH3 RIC experienced no weight loss, whereas mice that received PB10 RIC experienced severe weight loss that reached a nadir on days 7–8 (Fig. 4). The other RIC-treated groups of animals experienced significant degrees of weight loss, as compared with PB10/SylH3. In terms of immunostimulatory activity, all four of the two-component RICs induced the onset of RT-specific serum IgG with similar kinetics and to nearly identical degrees on days 50 and 80 (Fig. 4, Table II). In comparison, mice that received PB10-only RIC had significantly lower RT-specific serum IgG titers. In terms of TNA, all RIC-treated animals regardless of whether they were treated with PB10 or an mAb mixture, had similar levels of neutralizing activity in sera collected on day 113 (Table II). These results reveal that the immunostimulatory activity of RICs is not simply proportional to the acute inflammatory responses elicited by the RT–Ab complexes, but rather is a function of RT opsonization.

The generation of protective humoral immunity elicited by IC delivery has been previously reported to occur via an Fc-dependent process (18, 25, 29, 30). To address the role of Fc interactions in the case of RICs, we employed mice deficient in the Fc γ-chain subunit (31). The so-called ΔFcγR mice are devoid of FcγRI and FcγRIII, which might be expected to play a role in immune complex uptake (19, 32). These mice also lack additional Fc receptors, including FcεR1 (33).

RICs were administered to age- and sex-matched ΔFcγR and wild-type BALB/c mice by the i.n. or i.p. routes. All mice survived RIC treatment with no signs of morbidity (Supplemental Fig. 4). Serum samples were collected on study days 0, 7, 14, 30, 60, and 90 after RIC treatment. RT-specific titers arose with nearly identical kinetics and to the same magnitude in wild-type and ΔFcγR mice (Fig. 5A, 5B) irrespective of whether RICs were administered by i.n. or i.p. routes. By day 90, the RT-specific serum IgG endpoint titers in the wild-type and ΔFcγR groups of mice differed by <4-fold in the i.p. treated groups and <3-fold in the i.n. treated groups. These results demonstrate that the immune stimulatory activity of RICs occurs in the absence of Fc–FcγR interactions.

To confirm these results and to eliminate a role for FcγRII, we took advantage of having access to LALA derivatives of PB10 and SylH3. The LALA mutations in IgG1 Fc region render the Fc elements of the Abs “silent” in terms of Fc–FcγR interactions (34, 35). The LALA versions of PB10 and SylH3 consisted of mouse V_H and V_L elements grafted onto the human IgG1 Fc or human κ L chain segments. The resulting LALA RICs were delivered i.n. to mice. For comparison with LALA RICs, which contain human Fc regions, chimeric versions of the unaltered PB10 and SylH3 mAbs with human Fc regions were used to generate conventional RICs. Neither the LALA nor the conventional RIC formulations induced any morbidity or mortality following administration (Supplemental Fig. 4). Serum IgG titer production was unaffected by LALA RICs, as the mean group titers were similar to those of mice that received conventional RICs (Fig. 5C). These results are consistent with RIC stimulatory activity being independent of Fc–FcγR interactions.

Inhalation of trace amounts of RT can trigger severe inflammatory responses in the lung that ultimately culminate in acute respiratory distress syndrome (26, 36, 37). However, when delivered to mice as an immune complex consisting of equal parts PB10 and SylH3, RT is effectively neutralized, as recipient animals do not experience any weight loss or significant lung inflammation in the hours and days following exposure (12). Based on that and other observations, we had assumed that RICs were biologically inert and cleared from the airways through physical forces and/or mucus entrapment without consequence (38). Contrary to these expectations, we found that RICs are immunostimulatory. Mice that received a single dose of RICs mounted an anti-RT serum IgG response that persisted for months. Those same mice had unusually high levels of toxin-neutralizing Abs and were immune to lethal dose RT challenge 90 d later. The fact that RIC-induced Ab responses were essentially identical between ΔFcγR and wild-type mice was also unexpected, as it suggests that RICs are sampled by APCs by an Fc-independent pathway. Finally, RIC-induced Ab responses were not simply a result of residual toxicity, as the most inflammatory RIC preparations in vivo (as measured by previous postexposure morbidity) were not necessarily the most immunogenic. Collectively, these findings raise fundamental questions about the fate of ICs within the upper and lower airways and the degree to which properties intrinsic to the Ag itself contribute to how ICs are processed in the context of the lung.

There are only a handful of examples in the literature in which the fate of ICs following i.n. delivery has been examined (20, 39). In one well-characterized model, i.n. delivery of “contrived” ICs consisting of the Streptococcus pneumoniae Ag, PspA, fused to a single-chain variable fragment directed against human FcγRI-stimulated anti-PspA IgG and IgA in serum and bronchoalveolar lavage fluids was associated with protective immunity (21). In another model, ICs consisting of inactivated whole-cell Francisella tularensis (iFT) coated with an anti-LPS IgG mAb delivered i.n. to mice potentiated the onset of bactericidal Abs, as compared with iFT alone (40–43). In the PspA model, the total number, as well as the activation status, of dendritic cells and macrophages within the nasal-associated lymphoid tissues increased within hours after immune complex administration. Finally, the immunomodulatory activity of both the PspA and iFT immune complexes was FcγR-dependent. Taken together, a model emerges in which i.n. delivered immune complexes are sampled by mucosal APCs via one or more FcR-mediated pathways. RICs differ from these examples in two respects: only a single dose (rather than a prime/boost) of RICs is required to induce the onset of long-term immunity, and RIC-induced Ab responses were independent of FcγR interactions.

Indeed, we postulate that RT is likely an active participant in its own uptake into mucosal APCs, even when bound to neutralizing mAbs such as PB10 and SylH3. It is well established that RT gains entry into mammalian cells by two distinct pathways: a lactose-dependent pathway mediated by RTB’s two galactose/N-acetylgalactosamine (Gal/GalNAc)–specific carbohydrate recognition domains and a mannose-dependent pathway mediated by RT’s three different high mannose side chains (two on RTB, one on RTA) that are recognized by mannose-specific C-type lectins, including the mannose receptor (MR) (44–46). SylH3 is effective at inhibiting RT binding and uptake in certain cell types (e.g., epithelial cells), but not others (e.g., LSECs, primary macrophages) (17, 24, 47, 48), whereas PB10, which is directed against ricin’s enzymatic subunit, actually enhances RT uptake into Vero and HeLa cells (49). In fact, when evaluated using mouse lung-derived macrophages, the PB10/SylH3 mixture only reduced RT surface binding by ∼20% (12). In effect, RICs may actively promote their own uptake into APCs (and other cell types) in the nasal-associated lymphoid tissues and/or respiratory mucosa.

We speculate that the handling of RICs by APCs following endocytosis may be at the heart of the immunostimulatory activity observed in vivo. Classical retrograde trafficking of RT through the lactose pathway is known to be disrupted by PB10 (49), so RICs internalized by that route are likely directed to lysosomes. However, we favor a model wherein binding of RT by PB10 and SylH3 shunts RT toward a lectin-dependent pathway, by which RT is processed and presented. In addition to its roles in innate immune activation and cytokine production, the MR pathway has also been described as a robust pathway for Ag presentation by both MHC class I and II (50). This putative MR-mediated uptake pathway provides a reasonable basis for the rapid induction of anti-RT serum IgG as well as for the FcγR independence of RICs. Of course, several C-type lectin receptors, including those of the dendritic cell–specific ICAM–3–grabbing nonintegrin (DC-SIGN) family, may be facilitating this process, but none have been as well documented as the MR pathway in the context of RT. Studies evaluating the role of C-type lectin receptors in recognizing and presenting RIC Ag are currently ongoing.

RICs were also immunostimulatory when administered to mice by the i.p. route, similar to the effects of an i.v. administration route reported by Lemley et al. (51). RT given by these routes is known to traffic to the liver, where it is internalized by Kupffer cells and LSECs that then undergoes apoptosis (52). Although the exact fate of RICs given by these systemic routes is currently unknown, the neutralizing activity of PB10 and SylH3 would protect Kupffer cells and LSECs from toxin-induced apoptosis, potentially allowing for Ag degradation and presentation (53).

An ∼20-fold molar excess of mAb to RT was used in the current study, as previous work determined that lower concentrations of mAbs yield less protection against lethal respiratory RT challenge (27, 28). However, both studies provided mAbs prior to RT exposure as opposed to in combination with RT. Additional studies evaluating various ratios of mAbs and RT within RICs and the consequent impact on immunity have not been performed.

We established in the present study that RIC-conferred protection is impacted by both initial administration and challenge routes, with data suggesting that mucosal RIC delivery affords improved protection against respiratory RT exposure compared with systemic protection. Similar trends have been reported in other mucosal immunization studies (54) including with SARS-CoV-2 and respiratory syncytial virus (55–57). A recent study in a mouse model of influenza infection supports existence of tissue-resident memory B cells that differentiate into local plasma cells at the site of reinfection (58). Others have reported roles for BALT tissues in mediating local Ag presentation and lymphocyte activation within the lower respiratory tract (59–61). It is tempting to speculate that similar forces are at play following i.n. RIC exposure and RT challenge. Investigation into the formation of BALT and tissue-resident lymphocytes following i.n. RIC delivery is currently ongoing.

Regardless of the mechanisms at play, the durable protection afforded by RICs is impressive and opens the door for future exploration. Immune complexes have shown experimental success as vaccine candidates against infectious disease in livestock (23, 62) but have not yet made their way into humans. However, interest in developing immune complex vaccines is high, as they may allow for new long-term prophylactic applications of therapeutic Abs, including those already in existence, which are currently limited to treatment of developed disease and limited to relatively short-term benefits. There is a growing body of work supporting the use of immune complexes as vaccinations for a variety of pathogens, including HIV, F. tularensis, tick-borne encephalitis virus, hepatitis B virus, Zika virus, and influenza A virus (20, 29, 30, 63). The success of RICs in inducing a vaccinal effect by the i.n. route also creates opportunity for further exploration of mucosal immunization, another key frontier in vaccine development. The development of concomitant systemic and local immunity is perhaps the most enticing advantage of all, and it is perfectly exemplified by the dual systemic and mucosal protection conferred by RIC immunization.

We thank members of the Wadsworth Center’s Veterinary Sciences team for caring for animals during COVID-19 facility closings. We extend our thanks to Drs. Michael Pauly, Miles Brennan, and Larry Zeitlin (Mapp Biopharmaceutical, San Diego, CA) for providing chimeric LALA derivatives of PB10 and SylH3.

The authors have no financial conflicts of interest.