Cross-protection between attenuated Plasmodium berghei and P. yoelii sporozoites

An attenuatedPlasmodium falciparum sporozoite (PfSPZ) vaccine is under development, in part, based on studies in mice withP. berghei. We usedP. berghei andP. yoelii to study vaccine-induced protection against challenge with a species of parasite different from the immunizing parasite in BALB/c mice. One-hundred percent of mice were protected against homologous challenge. Seventy-nine percent immunized with attenuatedP. berghei sporozoite (PbSPZ)(six experiments) were protected against challenge withP. yoelii sporozoite (PySPZ), and 63% immunized with attenuatedPySPZ(three experiments) were protected against challenge withPbSPZ. Antibodies in sera of immunized mice only recognized homologous sporozoites and could not have mediated protection against heterologous challenge. Immunization with attenuatedPySPZ orPbSPZ induced CD8+ T cell-dependent protection against heterologous challenge. Immunization with attenuatedPySPZ induced CD8+ T cell-dependent protection against homologous challenge. However, homologous protection induced by attenuatedPbSPZ was not dependent on CD8+ or CD4+ T cells, and depletion of both populations only reduced protection by 36%. Immunization of C57BL/10 mice withPbSPZ induced CD8+ T cell-dependent protection againstP. berghei, but no protection againstP. yoelii. The cross-protection data in BALB/c mice support testing a human vaccine based on attenuatedPfSPZ for its efficacy againstP. vivax.


INTRODUCTION
In 1967 it was reported that immunization of A/J mice with radiation-attenuated Plasmodium berghei sporozoites ( Pb SPZ) protected the mice against malaria (1). Several years later, it was reported that humans immunized by exposure to radiation-attenuated P. falciparum sporozoites ( Pf SPZ) (2)(3)(4) were also protected. There now is a major effort to develop an attenuated sporozoite (SPZ) vaccine to protect humans against P. falciparum , the malaria parasite responsible for greater than 98% of the 1-3 million deaths caused by malaria annually (5). It is not known if immunization with attenuated Pf SPZ will protect against exposure to P. vivax , P. malariae or P. ovale , the three other Plasmodium species that infect humans. One individual was immunized with Pf SPZ, shown to be protected against P. falciparum , and then challenged with P. vivax SPZs, and shown not to be protected (6), raising the possibility that there may not be cross-protection. Determination of cross-protection in humans must await clinical trials. However, because of the importance murine studies have played in providing the foundation for human studies with attenuated SPZs, we have used two rodent malaria parasites, P. berghei and P. yoelii , to study cross-protection in mice. The results indicate that there is significant cross-protection, and that different immune responses may be active against closely related parasite species.

Mice
Female, 6-to 8-week-old BALB/CByJ and C57B1/l0 mice from Jackson Laboratories (Bar Harbor, ME, USA) were used in this study.
Correspondence : Stephen L. Hoffman, Sanaria Inc., 9800 Medical Center Drive, Suite A209, Rockville, MD 20850, USA (e-mail: slhoffman@sanaria.com). The assertions herein are the private ones of the authors and are not to be construed as official or as reflecting the views of the US Navy or the Naval service at large. The experiments reported herein were conducted according to the principles set forth in the 'Guide for the Care and Use of Laboratory Animals,' Institute of Laboratory Animal research, National Research Council, National Academy Press (1996). Re-use of this article is permitted in accordance with the Creative Commons Deed, Attribution 2·5, which does not permit commercial exploitation.

Parasites and immunizations
Sporozoites of P. berghei (ANKA) (7) and P. yoelii (17NL, 1.l clone) (8) were produced in laboratory-bred and infected Anopheles stephensi mosquitoes. Sporozoites for immunization were separated from infected mosquitoes using a renograffin discontinuous gradient as previously described (7). Sporozoites were prepared from infected mosquitoes that had been exposed to 10 000 rad from a 137 Ce source. Harvested SPZs were counted, diluted to the required concentration in M199 containing 5% normal mouse serum, and a volume of 0·2 mL was injected i.v. through the tail vein. Mice received multiple doses at 2-week intervals consisting of 7 × 10 4 irradiated SPZs as the first dose and 3 × 10 4 irradiated SPZs as subsequent doses.
In challenge experiments, hand-dissected, infected mosquito glands were triturated in medium to obtain SPZs. Challenge doses varied from 1 to 10 × 10 3 nonirradiated SPZs and were injected i.v. Minimum infective doses and the 50% infectious dose (ID 50 ) of SPZs used to challenge the immune animals were determined during most challenges by making fivefold dilutions of the challenge inocula.

Antibody responses
An indirect fluorescent antibody test (IFAT) was used to detect antibodies to P. yoelii and P. berghei SPZs in sera of mice by methods previously described (8). Briefly in the IFAT, diluted sera were reacted with air-dried SPZs and end point titres were detected using fluorescein isothiocyanate (FITC)-labelled rabbit anti-mouse immunoglobulin.

Protection against challenge
Two weeks after the last immunization, mice were challenged by i.v. inoculation in the tail vein with viable P. yoelii or P. berghei SPZs. Giemsa stained blood films were examined with a light microscope under an oil immersion objective (1000 × ) for the presence of asexual erythrocytic stage parasites. Mice showing no parasites in their blood for 15 days following challenge with infective SPZs were designated as negative.

Depletion of CD8 + + + + T lymphocytes
To determine if CD8 + T cells were required for protective immunity, we depleted mice of CD8 + T lymphocytes just prior to challenge. Mice immunized with five doses of irradiated SPZs were given three 0·5-mg i.p. injections of the mAb 19/178 at 24-h intervals to deplete them of CD8 + T lymphocytes as previously described (9). Peripheral lymphocyte samples were prepared and analysed for depletion of the population carrying the CD8 + T cell marker in a fluorescent-activated cell sorter (FACS). The mice were challenged with infective SPZs 4 days later. Treatment with mAB 19/178 was continued every third day after challenge until patency occurred. Control immunized mice were injected with the same quantities of rat immunoglobulin.

Calculation of the ID 50 and challenge units
Because the infectivity of individual preparations of SPZs may vary, the infectivity of the SPZs used for challenge was monitored during challenge in some of the experiments (Tables 1-3  dose at which the probability of infection is 0·5. To be acceptable for calculation of the ID 50 , at least two consecutive dilutions had to have less than 100% and greater than 0 infectivity. This is because the log of the odds for these results had to be finite. Confidence intervals for this point estimate were calculated using the asymptotic variance obtained by the delta method; log ID 50 values were exponentiated back to numbers of SPZs. After determining the ID 50 for the individual SPZ preparation, the number of ID 50 units contained in the challenge dose was calculated by dividing the challenge dose by the calculated ID 50 . Table 4 shows that immunization with irradiated PySPZ or PbSPZ resulted in the production of high antibody titres to the parasite used for the immunization, but no cross-reacting antibodies to the other species.

Cross-protection in BALB/c mice
Two weeks after the last immunization with radiationattenuated SPZs, mice were challenged by i.v. injection of infective PySPZ or PbSPZ. In two challenge experiments, the number of PySPZ and PbSPZ used for challenge was 5 × 10 3 and 10 × 10 3 , respectively. In all challenges, 100% of naive mice became infected. In these two challenges, protection against challenge with PySPZ or PbSPZ was 100% in groups Groups of BALB/c mice were immunized with five doses of radiation-attenuated PySPZ. Two weeks after the last immunization, mice were injected with anti-CD8 mAbs or a control antibody prior to challenge with infective SPZs. The challenge dose for P. berghei was 4 × 10 3 SPZs, and for P. yoelii was 10 3 SPZs. a The ID 50 for P. yoelii was 13·6 SPZs (95% CI 5·4 -34·0), and the number of P. yoelii ID 50 units contained in the 1 × 10 3 SPZs used for the challenge was 73·7 (95% CI 29·4−185·1). The ID 50 units for the P. berghei challenge could not be determined (see Materials and Methods).
Groups of BALB/c mice were immunized with five doses of radiation-attenuated PbSPZ. Two weeks after the last immunization, mice were injected with anti-CD8 mAbs or control antibody prior to challenge with infective SPZs. The challenge dose for P. berghei was 4 × 10 3 infective SPZs, and for P. yoelii was 1 × 10 3 SPZs. a The ID 50 for P. yoelii used in Experiment 1 was 17·9 SPZs (95% CI 7·6−42·2), and the number of P. yoelii ID 50 units contained in the 1 × 10 3 SPZs used for this challenge was 55·9 (95% CI 23·7−132·0). ID 50 units for the P. yoelii used in Experiment 2 and the P. berghei challenge could not be determined (see Materials and Methods). ND, not done. that received the same parasite species for immunization and challenge (homologous immunization and challenge) (13 out of 13 mice protected for both parasites) (Table 1). However, BALB/c mice immunized with radiation-attenuated SPZs of either parasite strain showed different levels of cross-protection against the heterologous parasite species challenge. In two experiments, the level of cross-protection obtained when mice immunized with attenuated PbSPZs were challenged with PySPZs was greater (12 out of 13 mice protected, 92·3%) than when mice immunized with attenuated PySPZs were challenged with PbSPZ (7 out of 13 mice protected, 54%); P = 0·073, Fisher's exact test, two-sided. This difference occurred despite significantly higher infectivity of the PySPZ (mean ID 50 = 10·6) vs. PbSPZ (mean ID 50 = 571·1); P = 0·0047, independent samples t-test.

Mechanisms of cross-protection in BALB/c mice
In BALB/c mice, CD8 + T cells have been shown to be responsible for the immunity against PySPZ after immunization with radiation-attenuated PySPZ (8-10), while protection induced in BALB/c mice after immunization with irradiated PbSPZ is independent of CD8 + T cells (11).
To determine whether the same immune mechanisms were responsible for the cross-protection between P. yoelii and P. berghei, we injected immunized mice with an anti-CD8 mAb and then challenged the mice with infective SPZs. The FACS analysis confirmed that injection of the mAb resulted in depletion of more than 99% of the CD8 + T cells (data not shown). Depletion of CD8 + T cells in BALB/c mice immunized with attenuated PySPZ eliminated protection against challenge with homologous (P. yoelii) and heterologous (P. berghei) SPZs (Table 2). In mice immunized with attenuated PbSPZs, treatment with anti-CD8 antibodies had no effect on challenge with homologous (P. berghei) SPZs, but it eliminated protection against heterologous (P. yoelii) SPZs (Table 3).
Since CD8 + T cell depletion did not affect protective efficacy in BALB/c mice immunized with radiation-attenuated PbSPZ and challenged with PbSPZs (Table 3), we assessed the effect of CD4 + T cell depletion in these mice ( Table 5). Depletion of CD4 + T cells had no effect on protection ( Table 5), but depletion of both CD4 + and CD8 + T cells had a modest effect on protective efficacy (Table 5). CD4 + T cell depletion reduced, but did not eliminate protection against challenge with PySPZ.

Cross-protection between the two parasite species in C57BL/10 mice
In BALB/c mice immunized with radiation-attenuated PbSPZ, CD8 + T cells are not required for protection against PbSPZ challenge (11,12). However, in A/J (H-2a) mice immunized with radiation-attenuated PbSPZ, CD8 + T cells are required for protection (12,13). Since the mechanism of protection against P. berghei varied among different mouse species, we wanted to determine if cross-protection extended to another mouse strain. We immunized C57BL/10 (H-2b) mice with irradiated P. berghei to determine the mechanism of protection induced in this mouse strain against homologous challenge and also to determine if mice immunized with irradiated P. berghei showed cross-protection against PySPZ. In this mouse strain, C57BL/10 (H-2b), protection against the homologous P. berghei challenge was dependent on CD8 + T cells (Table 6). Furthermore, C57BL/10 (H-2b) mice immunized with irradiated P. berghei did not show cross-protection against PySPZ (Table 6).

DISCUSSION
There were several major new findings in these studies. There was significant cross-protection between P. yoelii and P. berghei in BALB/c mice (Tables 1-3 and 5), antibodies against SPZs did not play a role in the cross-protection and Groups of BALB/c mice were immunized with five doses of radiation-attenuated PbSPZ and depleted of either T cells expressing CD4 + or T cells expressing CD4 + and CD8 + markers, and challenged with 7 × 10 3 PbSPZ or 1 × 10 3 PySPZ.
No claim to original US government works Volume 29, Number 11, November 2007 Cross-protection between Plasmodium species the immune mechanisms required for protection against homologous challenge, but not heterologous challenge, in BALB/c mice were apparently different for P. yoelii (Table 2) and P. berghei (Tables 3 and 5). The potentially most important finding was the crossprotection. One-hundred percent of mice were protected against homologous challenge. Thirty-one of 39 (79%) BALB/c mice immunized with attenuated PbSPZ were protected against challenge with PySPZ, and 12 of 19 (63%) BALB/c mice immunized with attenuated PySPZ were protected against challenge with PbSPZ (Tables 1-3 and 5).
Mice immunized with PySPZ did not produce antibodies that recognized PbSPZ, and mice immunized with PbSPZ did not produce antibodies that recognized PySPZ (Table 4). Thus, antibodies against SPZs could not have been responsible for the protection against heterologous challenge.
The protection against heterologous challenge in mice immunized with either radiation-attenuated PySPZ or PbSPZ was dependent on CD8 + T cells (Tables 2 and 3). When mice were immunized with radiation-attenuated PySPZ and challenged with nonirradiated PbSPZ, or immunized with radiation-attenuated PbSPZ and challenged with nonirradiated PySPZ, the protective immunity was eliminated by treatment of the mice before challenge with an antibody to CD8 + T cells. These data demonstrated that CD8 + T cells were required for the cross-protection. The requirement for T cells to provide protection against heterologous parasites was consistent with the finding that these mice made no antibodies against the heterologous SPZs (Table 4).
In BALB/c mice immunized with radiation-attenuated PySPZ, protection against challenge with P. yoelii was dependent on CD8 + T cells (Table 2). However, in BALB/c mice immunized with radiation-attenuated PbSPZ, protection against P. berghei was not eliminated by depletion of either CD8 + (Table 3) or CD4 + T cells (Table 5) and only modestly reduced by depletion of both (Table 5). The finding of lack of dependence on CD8 + T cells in BALB/c mice immunized by i.v. injection of irradiated PbSPZ model was first reported more than 16 years ago (11) Interestingly this did not occur in mice immunized by the bite of irradiated, PbSPZ-infected mosquitoes (14). Regardless, the other results have not been previously reported. There are several possible explanations for why T cell depletion did not reduce protection against P. berghei in mice immunized with irradiated PbSPZ. One possibility is that non-T cell immunity is more effective against PbSPZ infection than it is against P. yoelii. This could be because the infectivity (ID 50 ) of PbSPZ is lower than that of PySPZ (Table 1 and see  below), and the parasites are easier to eliminate by non-T cell mechanisms because fewer of them are infective. This differential infectivity could explain why, in contrast to PbSPZ, the protection against PySPZ in these same mice was eliminated by depletion of CD8 + T cells. Another explanation for why T cell depletion of mice immunized with irradiated PbSPZ did not eliminate protection is that the effect of in vivo T cell depletion on effector T cells could have been less in mice immunized with irradiated PbSPZ as compared to after immunization with irradiated PySPZ. Although the efficiency of in vivo depletion of total splenic CD8 + and CD4 + T cells was 99% in these experiments, we did not measure antigen-specific T cell populations responding to P. berghei or P. yoelii antigens after depletion. Thus, there remains a possibility that sufficient effector cells remained in the residual 1% of undepleted total T cells to protect mice against PbSPZ infection, but not P. yoelii infection. There is a region, amino acids 281-289 on the P. yoelii circumsporozoite protein (PyCSP) (SYVPSAEQI) and P. berghei circumsporozoite protein (PbCSP) (SYIPSAEKI), which contains a protective cytotoxic T lymphocytes (CTL) epitope (15). Preliminary studies by one of us (W. Weiss, unpublished data) indicate that the precursor frequency of CD8 + CTLs responding to the specific epitopes were 10-fold higher in BALB/c mice after immunization with irradiated PbSPZ than with PySPZ. Thus, while in vivo depletion removed 99% of all CD8 + T cells, it is possible that it removed 10-fold less CD8 + T effector cells against this protective CTL epitope in mice immunized with irradiated PbSPZ as compared to irradiated PySPZ. Thus, it is possible that T cells were responsible for protection of BALB/c mice immunized with irradiated PbSPZ, but that our methods of in vivo depletion were inadequate to demonstrate this. This line of reasoning does not explain why in these same mice T cell depletion eliminated protection against challenge with PySPZ. In mice immunized with irradiated PbSPZ and challenged with PySPZ, CD4 + T cell depletion modestly reduced protection (Table 5), indicating that CD4 + T cells play a role in this heterologous protection. Interestingly, this effect was not seen with homologous challenge, but depletion of both CD8 + and CD4 + T cells reduced protection against homologous (P. berghei) challenge, again showing that CD4 + T cells do play some role in protection. This finding also raises the possibility of the complementary effects of CD8 + and CD4 + T cells working together. Nevertheless, our findings and more than 20 years of research on mice immunized with irradiated SPZ have demonstrated that CD8 + T cells are the major effector arm of the immune system, but that CD4 + T cells also play a role, albeit minor role, and that in some cases, antibodies can be the major effector immune response (16).
Despite the excellent cross-protection in BALB/c mice, there was no cross-protection in C57BL/10 mice. It is possible that the immune response induced with the irradiated PbSPZ was not strong enough to overcome the highly infectious PySPZ challenge in the C57BL/10 mice. The high susceptibility of C57BL/10 mice to PySPZ may also explain why in some experiments not all mice immunized with irradiated PySPZ become protected after challenge with PySPZ (12,17).
It is generally thought that optimally radiation-attenuated SPZs pass through the bloodstream to the liver, invade hepatocytes and partially develop expressing proteins not expressed in hepatocytes. Many malariologists believe the CD8 + T cell-dependent protective immunity elicited by immunization with radiation-attenuated SPZs is primarily directed against these 'new' proteins first expressed in hepatocytes (10,16,17) This view has been based on a report that 'over-irradiated' SPZs do not provide protection even though they invade hepatoctyes (18), and killed SPZs do not provide high level of protection (19). Thus, despite the fact that immunization with PyCSP and PbCSP (11,20 -24) based vaccines elicit CD8 + T cell-dependent protection against homologous challenge, and transfer of an anti-PyCSP T cell clone protects against P. yoelii and P. berghei challenge (15,25), it is thought that these SPZ-derived proteins are not adequate for the high level of protective immunity seen after immunization with radiation-attenuated SPZs. These conclusions are supported by a recent study with mice transgenic for the PyCSP, and therefore tolerant to the PyCSP (26,27). When these mice received two doses of irradiated PySPZ, there was only minimal protection, indicating that PyCSP was important in protective immunity. However, when these mice received a full immunizing regimen of three doses of irradiated PySPZ, they were fully protected against challenge with infectious PySPZ, indicating that in this model system, immune responses against the PyCSP were not required for protection. While the data reported herein do not indicate what antigens are responsible for the protection, the findings that a PyCSP T cell clone was protective against challenge with the heterologous PbSPZ (15) possibly through the mediation of antigenspecific induction of IFNγ is significant. The availability of the genomic sequences of P. yoelii (28) and P. berghei (29), and gene expression and proteomic analyses should facilitate the discovery of other antigens involved.
The finding of cross-protection in mice between P. yoelii and P. berghei after immunization with radiation-attenuated SPZs is provocative, but cannot be used to predict what will occur in humans immunized with radiation-attenuated SPZs. However, when the radiation-attenuated PfSPZ vaccine is developed and tested (33), it will be important to determine if immunization with this vaccine protects against P. vivax. The data presented herein suggest that crossprotection could occur. Comparative analyses of the genomes of P. falciparum, P. vivax, P. yoelii and P. berghei may provide more insight into the likelihood of cross-protection. However, there will be no substitute for carefully executed clinical trials to determine if there is cross-protection.