When confronted with a new pathogen, the humoral immune system works to clear the pathogen and prevent a future infection. How pathogen clearance and memory formation are balanced when Plasmodium yoelii infects mice, a model of human malaria, is unclear. In this issue of Nature Immunology, Vijay et al. suggest that plasmablasts may act as a competitive sink for l-glutamine, thereby constraining the germinal center (GC) response1. Early treatment with l-glutamine or plasmablast reduction increased the GC response and improved parasite control (Fig. 1). The authors propose an intriguing model whereby competition for metabolites might regulate the transition from the early plasmablast response to the late GC response.

Fig. 1: Model of the influence of l-glutamine and plasmablasts on the GC response to Plasmodium infection.
figure 1

Gretchen Harms Pritchard

a. Early in the response to blood-stage Plasmodium infection in mice, pathogen-specific B cells in the spleen differentiate into plasmablasts to rapidly secrete antibody and, as a result, consume l-glutamine. Plasmodium-specific B cells differentiate into GC B cells more slowly, and this process also requires l-glutamine. These GC B cells can become plasma cells that secrete higher affinity antibodies and help control the infection. b. Supplementation of drinking water with l-glutamine or genetic depletion of plasmablasts increases the number of GC B cells, the number of T follicular helper cells (TFH), the amount of Plasmodium-specific antibody and improves parasite control.

Full size image

Plasmodium parasites first infect the liver, where they multiply until their numbers reach the tens of thousands. Next, they burst into the blood and begin infecting red blood cells2. Studies, including the current study by Vijay et al., model this by infecting mice intravenously with 1 × 106 Plasmodium-infected red blood cells1,3. In response to infection, Plasmodium-specific B cells rapidly differentiate into plasmablasts, cells that pump out high amounts of Plasmodium-specific antibody3. Plasmodium-specific antibodies bind parasites and support macrophages and T cells as they control the infection4,5. However, plasmablasts are short lived, and the population wanes after approximately 12 days.

After a week of infection, GCs begin to form, which constitutes the second wave of the B cell response. GCs are clusters of pathogen-specific B cells and follicular helper T (TFH) cells that work to diversify the B cell receptor repertoire and select B cells that can secrete antibodies with enhanced affinity for the pathogen. These B cells then become long-lived plasma and memory B cells. Plasma cells sustain Plasmodium-specific antibody production, while memory B cells can rapidly differentiate into plasmablasts to boost antibody production when they detect antigen. Both function to protect the host from a future infection6,7. Studies of blood-stage Plasmodium infection in mice demonstrated that parasite clearance during the primary infection required both the formation of GCs and the production of antibodies by GC-derived plasma cells or activated memory B cells8. Thus, it is important to understand what regulates the formation of GCs and protective antibodies during Plasmodium infection.

Both plasmablasts and GC B cells proliferate rapidly. They have a high demand for glutamine, which supports their proliferation and other functions via ATP production, the Krebs cycle and mTORC1 signaling9. However, how plasmablasts and GC B cells might compete for this resource has not been previously interrogated. Vijay et al. investigate the possibility of metabolic competition during mouse infection with blood-stage P. yoelii. The authors show that supplementation with l-glutamine reduced peak parasitemia and increased the formation of GCs. By contrast, treatment with l-alanine or l-valine did not decrease peak parasitemia. l-glutamine treatment increased the numbers of GC B and TFH cells early in the response and increased the titer of P. yoelii–specific antibodies. Additionally, oxygen consumption rates and the basal respiration of activated B cells increased, supporting the idea that l-glutamine can enhance GC B cell proliferation. The authors found l-glutamine was reduced at day 5 of infection in untreated mice, and the effects of l-glutamine addition were only seen when l-glutamine was supplied early after infection. This work suggests that increasing the level of l-glutamine early in the response enables GCs to form earlier, grow bigger, or both.

Interested in whether proliferating plasmablasts may be consuming the l-glutamine that the GC B cells need, Vijay et al. depleted plamablasts in several ways during the response to P. yoelii infection. Most convincingly, they treated CD138-DTR (diphtheria toxin receptor) bone-marrow-chimeric mice with diphtheria toxin early during the response. This treatment depleted CD138-expressing plasmablasts while still enabling the formation of GC-derived plasma cells. Similar to l-glutamine supplementation, plasmablast depletion reduced peak parasitemia and increased the quantity of GC B cells, GC TFH cells and P. yoelii–specific antibodies.

Vijay et al. suggest that plasmablasts, which express higher levels of glutamine transporters and expand earlier in the response, deplete the available l-glutamine, thereby limiting early GC formation. In support of this model, combining depletion of plasmablasts and l-glutamine supplementation had only a small additive effect. Further study is needed to establish the mechanism of this redundancy and to determine whether the supplemental l-glutamine is directly affecting lymphocytes in the spleen. Identifying which cell type is consuming the l-glutamine will require cell-specific deletion of amino acid transport channels such as Slc7a5, and Slc7a5f/f mice have been bred10. For example, P. yoelii–specific T cells rapidly expand early in response to Plasmodium infection and express Slc7a5 transcript. Interferon-γ-secreting T helper 1 (TH1) cells and GC-supporting TFH cells can both help control parasitemia. These cells may also be competing for l-glutamine and may therefore benefit from supplementation. In addition, it would be interesting to understand whether the limited GC response is due to constrained mTORC1 signaling, proliferation or another cellular process.

While these are important first observations, future studies are needed to understand the mechanism of improved parasite control in this system. The authors detected increased anti-Plasmodium antibodies after l-glutamine supplementation at day 12 post-infection, and they imply that this antibody is aiding parasite control and depends on the presence of GCs. While GC-derived plasma cells produce, on average, higher affinity antibodies than plasmablasts, the authors only detected increases in plasma cell numbers after day 15. Previous work with protein immunization showed memory B cells are produced from GCs earlier than plasma cells11. Since memory B cells can turn into plasmablasts and produce antibodies, perhaps the antibodies derived from memory B cells are contributing to parasite control. Additional studies are needed to determine whether antibodies are directly required for the improved parasite control and what distinguishes these protective antibodies from early plasmablast-derived antibodies.

It will be fascinating to learn whether the plasmablast response constrains the GC response in the context of other infections or whether this is specific to infection with Plasmodium. Additional studies using immunization would allow for the investigation of how antigen load, plasmablast expansion and the parasites themselves might contribute to this metabolic balance. Work in mouse models demonstrated that the formation of GCs against the liver-stage of Plasmodium infection can be disrupted when the subsequent blood-stage Plasmodium parasites emerge12. As this coincides with the rapid proliferation of plasmablasts and reduced l-glutamine, it would be interesting to determine whether l-glutamine treatment can rescue the production of liver-stage-specific GCs and antibodies. These antibodies can prevent a second liver-stage infection13. These findings may also have important implications for vaccination strategies. While there is no licensed vaccine for malaria, there are several vaccines that rely on repeated immunization currently in trials. As each boost induces a rapid plasmablast response, it will be important to determine whether this disrupts the previous GC and impairs the formation of more protective and long-lived antibody responses to the vaccine.

Humans form antibodies to the blood-stage of P. falciparum after one infection and are mostly protected from severe disease14. Vijay et al. take the first step in investigating their findings by describing a correlation between the rapid plasmablast response and parasitemia during blood-stage Plasmodium infection in humans. This confirms previous work demonstrating rapid plasmablast formation in naturally infected children15. Future studies correlating plasmablasts, parasitemia and GC kinetics in humans are needed but will require fine needle aspirate sampling of human lymph nodes.

Finally, we are currently experiencing a SARS-CoV-2 pandemic. Massive plasmablast responses and high-antibody titers have been described in patients with severe COVID-19 disease. Determining how plasmablasts, GC-dependent antibody and memory formation interact will be of the utmost importance to developing treatments and vaccines.