Cutting Edge: Subunit booster vaccination confers sterilizing immunity against liver-stage malaria in mice initially primed with a weight-normalized dose of radiation attenuated sporozoites

Abstract

Radiation attenuated sporozoite (RAS) vaccination offers hope for global malaria control through induction of protective liver-stage-specific memory CD8 T cells. Effective RAS vaccination regimens exist; however, widespread implementation remains unfeasible. A key difficulty resides in the need to administer three or more doses intravenously to achieve sufficient immunity. Strategies to reduce the number of RAS doses are therefore desirable. Here we used mice to model human immune responses to a single, suboptimal weight-normalized RAS dose administered intravenously followed by subunit vaccination to amplify liver-stage-specific memory CD8 T cells. RAS-subunit prime-boost regimens increased the numbers of liver-stage-specific memory CD8 T cells to a level greater than present after one RAS vaccination. Both intravenous and intramuscular subunit vaccine delivery induced immunity in mice, and many vaccinated mice completely cleared liver infection. These findings are particularly relevant to human vaccine development because RAS-prime-subunit-boost vaccination would reduce the logistical challenges of multiple RAS-only immunizations.

INTRODUCTION

Plasmodium, the causative agent of malaria, infects millions of people annually and causes hundreds of thousands of deaths (1). Anti-malarial interventions such as insecticides and medications are problematic with regards to access and compliance (2, 3). Additionally troublesome is the tendency of mosquitos and parasites to develop resistance to chemical agents (4). Thus, efficacious malaria vaccines are of critical importance.

Successful vaccine development relies on understanding the parasite life cycle in the human. After being delivered to the host skin by infected mosquito bite, Plasmodium parasites enter the bloodstream and eventually infect liver hepatocytes. Parasites replicate and differentiate into merozoites over several days (~two in mice and ~seven in humans) to transition to potentially lethal blood-stage infection (5, 6). Liver-stage can be cleared by memory CD8 T cell-mediated immunity (5).

The “gold standard” for induction of liver-stage immunity in mouse models or humans is vaccination with radiation attenuated sporozoites (RAS) (7, 8). Intravenously (iv) injected attenuated sporozoites undergo an abortive liver infection that primes clonally diverse liver-stage-specific cytotoxic memory CD8 T cells (5, 9). These memory CD8 T cells, when present in sufficient numbers, can clear infected hepatocytes and prevent progression to blood-stage infection, termed “sterilizing immunity” (5, 10, 11). RAS vaccination is complicated, however, by logistical hurdles including the need for sporozoite harvest from infected mosquitos by hand, a cold chain, and iv delivery of three or more large doses of parasites (8, 12–14).

RAS immunization is well studied in mice, but often with doses that vastly exceed those used in human vaccine trials (~1–5×10⁴ RAS/mouse versus 1–10×10⁵ RAS in humans with a liver of 10³-fold greater size) (15–17). Our studies used a weight-normalized RAS dose of 500 RAS/mouse that is analogous to doses used in humans.

We used a murine immunization model to explore strategies for limiting the logistical burdens of RAS vaccination by boosting with subunit vaccines. Subunit vaccines generate CD8 T cell populations specific for one or more epitopes (18), and, unlike RAS, can be produced, transported, and administered rapidly (19, 20) by either iv or the more efficient intramuscular (im) route. We find that liver-stage malaria-specific memory CD8 T cells primed by weight-adjusted RAS vaccination can be numerically amplified with iv- or im-delivered subunit vaccines to achieve sterilizing liver-stage immunity.

MATERIALS AND METHODS

Mice, Immunizations, and Infections

6–10-week-old female CB6F1 (F1 cross of C57Bl/6 (H-2^b) and BALB/c (H-2^d)) mice (NCI) were injected iv (tail vein) with 10⁴ Plasmodium berghei ANKA (Pb) RAS for initial vaccination. For booster vaccination, mice received iv 10⁷ ActA^−/−InlB^−/− Listeria monocytogenes (Lm) expressing the indicated Pb peptide, 10⁶ pfu of MVA-PbCSP (21) or MVA-PR8-NP (21) iv or im (hamstring muscle) at 28 days after RAS vaccination. Mice were injected with iv 10³ freshly isolated Pb sporozoites for evaluation of sterilizing immunity and liver parasite burden.

Sporozoites

Pb sporozoites were purified from female mosquito salivary glands (21). RAS were created by irradiating virulent sporozoites with 200Gy using the University of Iowa Free Radical and Radiation Biology Program core facility. Infected mosquitos harboring luciferase expressing Pb sporozoites were sourced from the SporoCore, University of Georgia, Athens, GA.

Tetramers

Tetramers were conjugated from the following MHC class I monomers: K^d-CS_252–260(SYIPSAEKI), D^b-GAP50_41–48(SQLLNAKYL), D^b-TRAP_130–138(SALLNVDNL), K^b-RPL6_120–127(NVFDFNNL), and D^b-NP₃₆₆(ASNENMETM) (11, 22). 1:100 Fc block was used to prevent non-specific tetramer binding.

Cell isolation, purification, and flow cytometry

Livers were chopped and digested in digest medium (ThermoFisher, 17703034) for 30 minutes at 37°C, shaking at 100 RPM. Tissue was mashed through 70μM cell strainers. Cell suspensions were spun at 5 min 1500 RPM in a 35% Percoll (Cytiva #17089101)/65% HBSS (Gibco #24020–117) gradient. Blood was collected retro-orbitally. Erythrocytes were lysed with 1x Vitalyse (Cmdg, #WBL0100). Cell viability was assayed using the Fix Viability Dye 780.

Liver-stage immunity

Mouse liver parasite burden was assayed by bioluminescence using the Xenogen IVIS 200 instrument. Mouse abdominal fur was shaved prior to imaging. VivoGlo^™ Luciferin (Ref #P1043) was injected intraperitoneally 10 minutes before imaging. Fluorescent image capture occurred for 2 minutes with 1.5 cm assumed for subject height.

RESULTS AND DISCUSSION

Weight-normalized dose RAS-prime followed by Lm-CS₂₅₂-boost vaccination induces protective immunity to liver-stage malaria in mice

CB6F1 mice, which respond to H-2^d (BALB/c) and H-2^b (C57Bl/6)-restricted CD8 T cell epitopes (22), were vaccinated with iv Pb RAS to generate clonally diverse liver-stage-protective liver Trm cells and potentially protective circulating memory CD8 T cells (17, 23). Vaccinations contained either 10⁴ RAS, a typical mouse dose (17, 22), or a weight-normalized dose of 500 RAS (derived from a human RAS dose of ~5×10⁵) (15, 16) (Supplemental Fig. 1A). Mice were challenged with virulent Pb sporozoites four weeks later. All mice vaccinated with 10⁴ RAS – equivalent to 10⁷ RAS in a human – exhibited sterilizing immunity, while one of eight mice receiving 500 RAS had sterilizing immunity (Supplemental Fig. 1B). Since humans do not develop sterilizing immunity after one RAS vaccination (15, 16), this result supported using 500 RAS vaccination followed by subunit vaccination of mice to study amplification of protective memory CD8 T cell responses and liver-stage immunity.

We determined if a “proof-of-concept” subunit vaccine could boost RAS-induced memory CD8 T cell responses. Recombinant Lm vaccination was used because it stimulates robust expansion of antigen-specific memory CD8 T cells (23). One month after 500 RAS vaccination, mice received iv attenuated Lm expressing secreted CS₂₅₂, a liver-stage protective CD8 T cell H-2K^d-restricted epitope, or GAP50₄₁, a non-protective H-2D^b restricted epitope control (22). Frequencies of circulating CS₂₅₂-specific memory CD8 T cells in RAS-prime, Lm-CS₂₅₂-boost (RAS+Lm-CS₂₅₂)-vaccinated mice exceeded those of RAS+Lm-GAP50₄₁-vaccinated mice 28 days post-Lm (Fig. 1A, Supplemental Fig. 1C). Mouse livers contained CS₂₅₂-specific memory CD8 T cells of both liver-stage protective resident memory (Trm, CD69⁺CD62L⁻) and circulating effector memory (Tem, CD69⁻CD62L⁻) phenotypes, and both populations were numerically enriched in RAS+ Lm-CS₂₅₂-vaccinated mice versus RAS+Lm-GAP50₄₁-vaccinated controls (Fig. 1B & Supplemental Fig. 1D).

Fig. 1: Weight-normalized dose RAS-prime followed by Lm-CS₂₅₂-boost vaccination induces protective immunity to liver-stage malaria in mice.

CB6F1 mice received iv 500 Pb RAS followed by iv Lm-CS₂₅₂ or Lm-GAP50₄₁ 28 days later and were challenged with iv 10³ virulent Pb sporozoites 28 days after Lm vaccination. Blood was collected from all mice 2 days before challenge, while livers were collected from unchallenged mice for T cell analysis.

A) Frequencies of CS₂₅₂-specific CD8 T cells in the blood of RAS+Lm-vaccinated mice 26 days after Lm vaccination. n=21 (RAS+Lm-GAP50₄₁), n=23 RAS+Lm-CS₂₅₂)

B) Numbers of CS₂₅₂-specific Tem (CD69⁻CD62L⁻) and Trm (CD69⁺CD62L⁻) cells in livers of RAS+Lm vaccinated mice 26 days post-Lm vaccination. n=5 (RAS+Lm-GAP50₄₁), n=5 RAS+Lm-CS₂₅₂)

C) Liver parasite burden quantification (luciferin radiance) 44 hpi with 10³ Pb sporozoites. n=10 (infected control, n=16 (RAS+Lm-GAP50₄₁), n=18 RAS+Lm-CS₂₅₂), n=3 (uninfected control)

D) Depiction of whether mice exhibited sterilizing immunity (blood-stage-) or developed parasitemia (blood-stage+). Numbers indicate #blood-stage+ mice/#total mice per group. n=10 (infected control, n=16 (RAS+Lm-GAP50₄₁), n=18 RAS+Lm-CS₂₅₂), n=3 (uninfected control).

Data are combined from two independent experiments, each with at least 3 mice per group. Each circle represents an individual mouse and mean ± SEM are depicted. (A) & (B) Unpaired 2-tailed t-test (C) 1 way ANOVA (Turkey’s post-hoc correction) comparing the mean of each column to the mean of every other column. (D) Chi squared test with 95% confidence interval between individual groups. *p≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, ns p > 0.05.

Immunized and naïve mice were then challenged with iv 10³ luciferase-expressing virulent Pb sporozoites. Liver parasite burden was quantified using IVIS imaging near the end of liver-stage infection (44 hpi) to determine the ability of the memory CD8 T cell response to clear infection. RAS+Lm-GAP50₄₁-vaccinated mice reduced parasite burden (~0.5 Log₁₀ radiance on average) as compared to non-vaccinated controls (Fig. 1D & Supplemental Fig. 1E), however, no RAS+Lm-GAP50₄₁ mouse exhibited sterilizing immunity (Fig. 1E). RAS+Lm-CS₂₅₂-vaccinated mice, in contrast, exhibited greater reduction in liver parasite burden (~1.5 Log₁₀ radiance on average), and ten out of eighteen exhibited sterilizing immunity (Fig. 1C, 1D & Supplemental Fig. 1E). Of note, humans also develop partial non-sterilizing immunity after a single RAS vaccination (15, 16). These results suggest that subunit vaccination can augment partial liver-stage malaria protection engendered by one weight-normalized RAS vaccination via inducing numerical expansion of antigen-specific memory CD8 T cells.

Subunit vaccination with distinct liver-stage-protective epitopes induces protective immunity in RAS-vaccinated mice

To generalize these results beyond the CS₂₅₂ epitope, groups of 500 RAS-vaccinated mice received booster vaccinations of Lm expressing different liver-stage protective epitopes, including H-2D^b-restricted TRAP and H-2K^b-restricted RPL6 (11, 22). RAS-vaccinated mice that received Lm-CS₂₅₂ or Lm-TRAP or Lm-RPL6 booster vaccination exhibited antigen-specific memory CD8 T cells in the blood (Fig. 2A) and (Fig. 2B), liver as well as reduced liver parasite burden and sterilizing immunity ranging from 53–90% of mice after sporozoite challenge (Fig. 2C). These results suggested that distinct protective epitopes can be incorporated into subunit booster vaccines, which would give flexibility in designing human vaccines and permit antigen combination in booster vaccinations.

Fig. 2: Lm-boost vaccination utilizing distinct liver-stage-protective epitopes of RAS-primed mice induces protective immunity.

CB6F1 mice received iv 500 Pb RAS followed by iv Lm-CS₂₅₂, Lm-TRAP, Lm-RPL-6, or Lm-GAP50₄₁ 28 days later and were challenged with iv 10³ virulent Pb sporozoites 28 days after Lm vaccination. Blood was collected from all mice 2 days before challenge for T cell analysis.

A) Frequencies of antigen-specific CD8 T cells in the blood of RAS+Lm-vaccinated mice 26 days after the Lm vaccination. Each group depicts the frequency of CD8 T cells that were specific for the antigen expressed by the respective Lm vaccination. n=10 (RAS+Lm-GAP50₄₁), n=18 RAS+LmTRAP, n=15 RAS+ Lm-RPL6, n=10 RAS+Lm-CS₂₅₂)

B) Liver parasite burden quantification (luciferin radiance) 44 hpi with 10³ Pb sporozoites. n=9 (infected control), n=10 (RAS+Lm-GAP50₄₁), n=18 RAS+Lm-TRAP, n=15 RAS+ Lm-RPL6, n=10 RAS+Lm-CS₂₅₂), n=6 (uninfected control).

C) Depiction of whether mice exhibited sterilizing immunity (blood-stage-) or developed parasitemia (blood-stage+). Numbers indicate #blood-stage+ mice/#total mice per group. n=9 (infected control), n=10 (RAS+Lm-GAP50₄₁), n=18 RAS+Lm-TRAP, n=15 RAS+ Lm-RPL6, n=10 RAS+Lm-CS₂₅₂).

Data are combined from 2 independent experiments, each with a minimum of 4 mice per group. Each circle represents an individual mouse and mean ± SEM are depicted. (C) 1 way ANOVA (Turkey’s post-hoc correction) comparing the mean of each column to the mean of every other column. (D) Chi squared test with 95% confidence interval between individual groups. *p≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, ns p > 0.05.

Intravenous MVA-CS₂₅₂ booster vaccination induces liver-stage protection in RAS-vaccinated mice

Although attenuated Lm has been evaluated as a vaccine delivery platform in clinical trials (24), it has not found wide usage in the clinic. Attenuated viral vectors, such as modified vaccinia virus Ankara (MVA) have in contrast been widely used to deliver malaria antigens in mice (19, 20) and to humans in prime-boost clinical trials (25, 26).

To address the utility of MVA-vectored boosting, RAS-vaccinated mice were injected with iv MVA-CS₂₅₂ or MVA-NP₃₆₆, the latter an irrelevant CD8 T cell influenza virus epitope. RAS+MVA-CS₂₅₂-vaccinated mice exhibited higher levels of circulating and liver-localized Tem and Trm CS₂₅₂-specific memory CD8 T cells than RAS+MVA-NP₃₆₆-vaccinated counterparts (Fig. 3A & 3B). RAS+MVA-CS₂₅₂-vaccinated mice demonstrated significantly reduced liver parasite burden (~1 Log₁₀ radiance on average) as compared to RAS+MVA-NP₃₆₆-vaccinated controls, and 83% of RAS+MVA-CS₂₅₂-vaccinated mice exhibited sterilizing immunity (Fig. 3C & 3D). Of note, single vaccination with either Lm-CS₂₅₂ or MVA-CS₂₅₂ did not discernibly reduce parasite burden or induce sterilizing immunity in mice despite inducing circulating CS₂₅₂-specific memory CD8 T cells (Supplemental Fig. 2A-D). These data suggested that human MVA vaccines could be used to enhance immunity engendered by one RAS vaccination.

Fig. 3: IV MVA-CS₂₅₂ vaccination induces liver-stage protection in RAS-vaccinated mice.

CB6F1 mice received iv 500 Pb RAS followed by iv MVA-CS₂₅₂ or MVA-NP₃₆₆ 28 days later and were challenged with iv 10³ virulent Pb sporozoites 28 days after MVA vaccination. Blood was collected from all mice 2 days before challenge, while livers were collected from unchallenged mice for T cell analysis.

A) Frequencies of CS₂₅₂-specific CD8 T cells in the blood of RAS+MVA-vaccinated mice 26 days after the MVA vaccination. n=19 (RAS+MVA-CS252), n=19 (RAS+ MVA-NP)

B) Numbers of CS₂₅₂-specific Tem (CD69⁻CD62L⁻) and Trm (CD69⁺CD62L⁻) cells recovered from the livers of RAS+Lm vaccinated mice 26 days post-Lm vaccination. n=7 (RAS+MVA-CS252), n=7 (RAS+ MVA-NP)

C) Liver parasite burden quantification (luciferin radiance) 44 hpi with 10³ Pb sporozoites. n=8 (infected control), n=12 (RAS+MVA-CS252), n=12 (RAS+ MVA-NP), n=8 (uninfected control).

D) Depiction of whether mice exhibited sterilizing immunity (blood-stage-) or developed parasitemia (blood-stage+). Numbers indicate #blood-stage+ mice/#total mice per group. n=8 (infected control), n=12 (RAS+MVA-CS252), n=12 (RAS+ MVA-NP).

Data are combined from 2 independent experiments, each with at least 4 mice per group. Each circle represents an individual mouse and mean ± SEM are depicted. (A) & (B) Unpaired 2-tailed t-test. (C) 1 way ANOVA (Turkey’s post-hoc correction) comparing the mean of each column to the mean of every other column. (D) Chi squared test with 95% confidence interval between individual groups. *p≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, ns p > 0.05.

Intramuscular subunit vaccine delivery induces liver-stage protection in RAS-vaccinated mice

Intramuscular delivery – widely used in human vaccines (27) – is preferred over the iv route due to simplicity and logistics. Im RAS delivery, however, is ineffective at inducing protective immune responses in humans as compared to iv (12). Because RAS vaccination induced CS₂₅₂-specific Tem and Trm in previous experiments, iv RAS-vaccinated mice were injected with im MVA-CS₂₅₂ to determine if one of two iv vaccinations could be replaced while still generating protective memory CD8 T cell-mediated immunity.

RAS-vaccinated mice were given im MVA-CS₂₅₂ (RAS+ im MVA-CS₂₅₂) or MVA-NP₃₆₆. RAS+ im MVA-CS₂₅₂-vaccinated mice exhibited higher levels of circulating CS₂₅₂-specific memory CD8 T cells than RAS+ im MVA-NP₃₆₆-vaccinated counterparts (Fig. 4A). RAS+ im MVA-CS₂₅₂ booster vaccination reduced liver parasite burden as compared to RAS+ im MVA-NP₃₆₆ controls (~1 Log₁₀ radiance on average), and 77% of RAS+ im MVA-CS₂₅₂-vaccinated mice exhibited sterilizing immunity (Fig 4B. & 4C). These data suggest that im subunit vaccine delivery with a translationally relevant vector expands memory CD8 T cells induced by RAS prime vaccination, and that these responses confer increased protection as compared to RAS vaccination alone.

Fig. 4: im MVA-CS₂₅₂ vaccination induces liver-stage protection in RAS-vaccinated mice.

CB6F1 mice received iv 500 Pb RAS followed by im MVA-CS₂₅₂ or MVA-NP₃₆₆ 28 days later. Mice were challenged with iv 10³ virulent Pb sporozoites 28 days after MVA vaccination. Blood was collected from all mice before challenge for T cell analysis.

A) Frequencies of CS₂₅₂-specific CD8 T cells in the blood of RAS+MVA-vaccinated mice 26 days after the MVA vaccination. n=13 (RAS+MVA-CS252), n=12 (RAS+ MVA-NP)

B) Liver parasite burden quantification (luciferin radiance) 44 hpi with 10³ Pb sporozoites. n=9 (infected control), n=13 (RAS+MVA-CS252), n=12 (RAS+ MVA-NP), n=6 (uninfected control).

C) Depiction of whether mice exhibited sterilizing immunity (blood-stage-) or developed parasitemia (blood-stage+). Numbers indicate #blood-stage+ mice/#total mice per group. n=9 (infected control), n=13 (RAS+MVA-CS252), n=12 (RAS+ MVA-NP),.

Data are combined from 2 independent experiments, each with at least 2 mice per group. Each circle represents an individual mouse and mean ± SEM are depicted. (A) Unpaired 2-tailed t-test (B) 1 way ANOVA (Turkey’s post-hoc correction) comparing the mean of each column to the mean of every other column. (C) Chi squared test with 95% confidence interval between individual groups. *p≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001, ns p > 0.05.

Our findings demonstrate that a weight-normalized dose RAS vaccination – which elicits a full range of malaria-specific Trm and circulating CD8 T cell memory cells – albeit at low non-protective numbers, primed hosts for a substantial secondary response of both CD8 populations following subunit boosting and conferred enhanced pathogen control. Limiting RAS usage to one vaccination would help overcome the logistical challenges of sporozoite immunization while still taking advantage of RAS-generated liver Trm and providing the opportunity to boost clonally distinct CD8 T cell responses. RAS-subunit prime-boost vaccination could therefore be a strategy to reduce the number of sporozoite-based vaccines necessary to induce sterilizing immunity to liver-stage malaria in humans.

KEY POINTS.

• Weight-normalized RAS vaccination dosing in mice models human immunity

• RAS prime-subunit boost vaccination induces immunity to liver-stage malaria in mice

Abbreviations

RAS

radiation attenuated sporozoites

Pb

Plasmodium berghei ANKA

Lm

Listeria monocytogenes

MVA

Modified vaccinia virus Ankara

Trm

Tissue resident memory CD8 T cells

iv

intravenous

im

intramuscular