Berberine improves central memory formation of CD8⁺ T cells: Implications for design of natural product-based vaccines

Abstract

Berberine (BBR) as one of the most effective natural products has been increasingly used to treat various chronic diseases due to its immunosuppressive/tolerogenic activities. However, it is unknown if BBR can be applied without abrogating the efforts of vaccination. Here we show that priming of CD8⁺ T cells in the presence of BBR lead to improved central memory formation (Tcm) with substantially reduced effector proliferation, primarily orchestrated through activation of AMPK and Stat5. Tcm derived from vaccinated mice fed with BBR were able to adoptively transfer protective immunity to naïve recipients. Vaccination of BBR-fed mice conferred better memory protection against infection without losing immediate effector efficacy, suggesting appreciable benefits from using BBR in vaccination. Thus, our study may help to lay the groundwork for mechanistic understanding of the immunomodulatory effects of natural products and their potential use as adjuvant that allows the design of novel vaccines with more desirable properties.

Graphical abstract

Berberine potentially modulates the adaptive immune response by targeting different signaling pathways in the priming phase of CD8⁺ T cells to enhance the generation of central memory and reduce the effector pool.

1. Introduction

Berberine (BBR), a quaternary isoquinoline alkaloid found in many kinds of plants including Hydrastis canadensis and Coptis chinensis, is becoming increasingly widely used in humans due to its therapeutic potentials in treating microbial infections, cardiovascular diseases, dyslipidemia, obesity, and diabetes1, 2, 3, 4, 5. BBR also exerts immunosuppressive effects by inhibiting proinflammatory responses of immune and related cells, such as epithelial cells, macrophages, or adipocytes, and the release of proinflammatory cytokines, like IFN-γ, TNF-α, and IL-2, from these cells, in colitis, arthritis, as well as diabetes and insulin resistance6, 7, 8, 9, 10, 11, 12, 13. More importantly, BBR mediates reduction in number of CD8⁺ T cells and application of BBR can lead to tolerogenic T cell responses systemically and locally^14,15. CD8⁺ T cells, play a central role in cell-mediated immunity, are pivotal in early protection after vaccination and continue to be essential in boost immunity16, 17, 18, 19. Even though all the immunoregulatory responses appear beneficial in many diseases or medical conditions, one cannot overlook the concern over the immunological consequence of BBR use in vaccinees who need an adequate CD8⁺ T cell-mediated response against forthcoming infectious threats.

Antigen stimulation is the key driver for proliferation and differentiation of naïve CD8⁺ T cell into effector cells which represent an expendable population and memory cells which provide a strong means of ensuring responsiveness in future encounters. While both cell pools could be generated simultaneously and CD62L-expressing memory precursors may emerge even prior to the peak response, a small fraction of early effector cells could possibly differentiate into memory cells with upregulated CD62L as well²⁰. Recent studies have demonstrated that most memory cells are likely to experience early and later effector phases²¹, and the increased memory potential is due to asymmetry cell division²². Effector and memory differentiation is transcriptionally controlled by competing sets of transcription factors, such as T-bet, Blimp1, ID2, Stat4 versus Eomes, BCL6, TCF1, and Stat3, that produce effector phenotypes and facilitate acquisition of memory properties, respectively²³. The resulting populations are thus heterogenous and classified into two major subsets, central memory (Tcm, CD62L⁺CCR7⁺CD44⁺) and effector memory/effector (Tem/Teff, CD62L⁻CCR7⁻CD44⁺) cells based on their lymphoid homing properties and activation status. Cytolytic Tem/Teff recirculate between peripheral blood and peripheral tissues, while less-cytotoxic Tcm reside primarily in secondary lymphoid organs with potential to enter the periphery. Because of their higher proliferative potential, Tcm have a greater capacity than Tem in the memory response.

To address whether and how BBR impacts CD8⁺ T cells, we performed in vitro experiments on naïve T cells by costimulating TCR and CD28 together with BBR and explored the molecular mechanisms behind effector and memory differentiation during the priming phase. We demonstrated that BBR-modulated T cells exhibited central memory bias in differentiation through activation of AMPK and Stat5, which are crucial in the downstream transcriptional activation for enhanced phenotypic plasticity with inhibited formation of effector cells. In vivo studies validated the effectiveness of BBR-improved Tcm and the multifunctional natural compound as a potential immunomodulator to improve vaccine efficacy in response to antigen invasion in the memory phase and particularly not to impair the immediate response in the effector phase.

2. Materials and methods

2.1. Animals

C57BL/6 CD45.2 mice with 6–8 weeks of age were obtained from Institute of Laboratory Animal Sciences (CAMS&PUMC, Beijing, China) or Beijing Vital River Laboratory Animal Technology (a subsidiary of Charles River Laboratories International, Beijing, China). C57BL/6 CD45.1 mice were acquired from Peking University Health Science Center (Beijing, China). OT-I mice were originally acquired from The Jackson Laboratory. Prkaa1^+/− mice were purchased from Shanghai Model Organisms Center. Animals had ad libitum access to water and food and were housed in a 12 h/12 h light/dark cycle. Research was conducted in accordance with all institutional guidelines and ethics and approved by the Laboratories Institutional Animal Care and Use Committee of the Chinese Academy of Medical Sciences and Peking Union Medical College. All experimental protocols were approved by the Institute of Materia Medica Animal Authorities (Beijing, China).

2.2. Reagents

FSL-1 and poly(I:C) were purchased from Invivogen (San Diego, CA, USA). CpG 1826 were synthesized by Chinese Peptide Company (Hangzhou, Zhejiang, China). All were free of endotoxin and protein contamination. SIINFEKL was synthesized by Chinese Peptide Company (Hangzhou, Zhejiang, China). Berberine, curcumin, and metformin were purchased from J&K and Sigma, respectively. iCRT3 and compound C (CC) were obtained from MCE (MedChem Express, USA), ICG001, BD750 and AS1842856 from Sellck (Plymouth, MI, USA). Compounds and small molecule inhibitors were dissolved in DMSO and further diluted with cell culture medium before use. Final DMSO concentration did not exceed 0.1%. The MHC tetramers (tet) loaded with SIINFEKL peptides were obtained from the NIH Tetramer Core Facility. Antibodies for flow cytometry were purchased from BioLegend, eBiosciences (San Diego, CA, USA), or Signaling Technology (Danvers, CO, USA) unless otherwise noted.

2.3. In vitro activation of lymphocytes

Splenocytes were isolated from wild-type mice either naïve or pretreated with BBR (100 mg/kg/day mixed in regular chow and fed for 1 month), and incubated with ACK lysis buffer for 5 min to remove erythrocytes. Cells cultured in RPMI-1640 supplemented with 10% fetal bovine serum, penicillin (100 U/mL), streptomycin (100 mg/mL), and IL-2 (100 U/mL) (Peprotech) at 37 °C in 5% CO₂, were stimulated with anti-CD3 (clone 145-2C11, 5 μg/mL, BioLegend) and anti-CD28 (clone 37.51, 2.5 μg/mL, BioLegend) mAbs for 3 days. BBR (13.4 μmol/L), CCM (15 μmol/L), and MTF (15 mmol/L) were added. To evaluate Tcm-mediated protection, OT-I cells were stimulated with SIINFEKL (0.01 μg/mL), poly(I:C) (30 μg/mL), and CpG 1826 (1 mg/mL) in addition to BBR. In some experiments, small molecule inhibitors were added together with TCR stimulation and incubated 3 days except CC and ICG001 added in the last 24 h of cell culture.

2.4. In vitro proliferation assay

Splenocytes isolated from naïve mice were labeled with CFSE (10 μmol/L, Sigma, MUSA) at 37 °C for 5 min. After removing excess CFSE, cells were cultured in a 96-well, round-bottom plate with 2 × 10⁵ cells in the presence of anti-CD3 and anti-CD28 mAbs in combination of BBR, CCM, or MTF for 48 h before flow cytometry analysis.

2.5. Flow cytometry

Fluorescent dye-labeled antibodies against cell-surface markers CD8 (53–6.7), CD44 (IM7), CD62L (MEL-14), CD127 (IL-7Ra, A7R34), and CD39 (Duha59) were used to stain cells for flow cytometry. For intracellular staining, cells were fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton-X100, and blocked in 5% BSA before 1-h incubation with primary antibodies, which include TCF-1 (C63D9), Foxo1 (C29H4), p-AMPK (Thr172, 40H9), β-catenin (D10A8). After washing, cells were incubated with secondary antibodies, including F(ab′)2 anti-rabbit IgG (Abcam, USA), or mouse anti-rabbit IgG (Santa Cruz Biotechnology, USA), for 30 min. p-STAT5 (Tyr694, D47E7) and Ki-67 (SolA15) were used for direct intracellular staining. Prior to all flow cytometry staining, FcγIII/II receptors were blocked by incubating cells with homemade antiCD16/32 (2.4G2). All samples were acquired and analyzed using FACSVerse cytometer (Becton Dickson, USA) and FlowJo software (TreeStar, USA).

2.6. Adoptive transfer of OT-I cells

Lymphocytes from OT-I (CD45.2) mouse spleen and lymph nodes were stimulated in vitro as described above. Cells were enriched for CD8⁺ T by magnetic separation using CD8a (Ly2) microbeads (Miltenyi Biotec, Germany) followed by sorting for CD62L⁺CD44⁺tet⁺ cells on a FACS Aria III (Becton Dickson, USA). Purified cells were resuspended in PBS and 5 × 10⁴ cells per mouse adoptively transferred i.v. into naïve CD45.1 mice.

2.7. Vaccination and challenge

The ΔactA rLM-OVA strain of Listeria monocytogenes (rLM-OVA) expressing OVA protein was incubated to mid-logarithmic phase in BHI broth (BD, Sparks, MD, USA) supplemented with streptomycin and erythromycin (Sangon Biotech, Shanghai, China) and aliquoted and stored at −80 °C before use. All mice were subjected to 5 h fasting prior to vaccination. Infected mice recovered from weight loss within 1 week. Samples were collected 2 weeks after the first inoculation. Mice either fed BBR or regular diet were primed by s.c. injection at the footpad with a mixture of 5 μg OVA, 30 μg poly(I:C), 3 μg CpG1826 ODNs in a volume of 50 μL through Days 0, 1, and 2. For boost immunization, mice received the same vaccine formulations on Days 11, 12, and 13. Mice were then challenged i.v. with 1 × 10⁶ CFU of LmOVA 21 days or 35 days after the initial immunization. To challenge recipient mice, at 1 day posttransfer, mice were infected i.v. with 1 × 10⁵ CFU LmOVA. Body weights were recorded regularly. Peripheral blood was harvested from recipient mice for donor cell characterization.

2.8. ELISA

To measure intracellular cytokines, cells were stimulated for 6 h at 37 °C with SIINFEKL peptide at 0.1 μmol/L. CD107a mAbs were added to the cells at the beginning of the incubation. Supernatants were collected after the restimulation. IFN-γ and TNF-α were quantified by enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's protocol (Elabscience, China). The optical density was read at 405 nm on a plate reader (Ensipre, PerkinElmer, Waltham, USA).

2.9. Imaging flow cytometry

Cells were labeled with CD8, CD44, and CD62L. Hoechst was used to exclude dead cells. Cells at 1 × 10⁶ cells per milliliter were run on the ImageStream Mark ΙΙ (Milli-Q, USA) and the image data were analyzed using the IDEAS Software.

2.10. Western blotting

Cellular lysates were prepared with lysis buffer (Solarbio, China) and then separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 10% polyacrylamide gels and transferred to polyvinylidene fluoride membranes. After transfer, the membranes were incubated in blocking buffer (TBST solution containing 5% skim milk) for 1 h at room temperature, and then washed three times with TBST for 5 min each. Membranes were then incubated overnight at 4 °C with the primary antibodies against p-AMPK, total AMPK, and Foxo1 purchased from Cell Signaling Technology (USA). The anti-β-actin antibody was purchased from Beyotime (Beyotime Biotechnology, China). After three washes with TBST, membranes were incubated with horseradish peroxidase-conjugated anti-rabbit IgG secondary antibody for 1 h at room temperature. After the three washing steps with TBST, protein bands were visualized using Tanon™ High-sig ECL Western Blotting Substrate (Tanon, China).

2.11. Statistics

Comparisons between groups were analyzed by two-tailed Student's t test. Analyses were performed with SPSS for Windows. P values less than 0.05 are considered statistically significant.

3. Results and discussion

In view of the fact that the clinical benefits of BBR are associated with at least a few weeks of use, mice were fed with BBR-containing diet (BBRpre mice) in the initial experiments. The number of total lymphocytes and CD8⁺ T cells from these mice following in vitro stimulation with α3/28 only was not affected (Fig. 1A), suggesting unimpaired proliferative capacity. There is an interesting fact that lymphocytes from BBRpre mice displayed a visible but insignificantly higher proliferative response to TCR stimulation than those from naïve (nonBBRpre) mice. It could be a post-suppression rebound of the lymphocytes. However, it is worth noting that cell numbers were significantly decreased by BBR with α3/28 stimulation, regardless of prior exposure to BBR or not (Fig. 1A and Supporting Information Fig. S1B, the gating strategy shown in Fig. S1A), indicating that BBR was inhibitory to proliferation of stimulated T cells. CD44 expression in α3/28-stimulated T cells was actually increased, rather than decreased, by the antiproliferative BBR (Fig. 1B), and a similar upregulation was seen for CD127 and CD69 (Fig. 1B and Fig. S1C). BBR actually shares many functional similarities with the natural compound curcumin (CCM) and the guanidine derivative metformin (MTF)¹⁵. It was interested to see that CCM and MTF were also inhibitory to primed T cells (Fig. 1C), confirmed in a CFSE-based proliferation assay (Fig. 1D).

Figure 1.

BBR inhibits Tem/Teff proliferation but improves Tcm differentiation. Equal numbers (2 × 10⁵) of splenocytes were isolated from mice either fed with 150 mg/kg/day BBR (BBRpre) for 4 weeks or untreated (nonBBRpre naïve), and stimulated with anti-CD3 and anti-CD28 (α3/28) in combination of BBR for 3 days. In some experiments, CCM or MTF was substituted for BBR. For proliferation assay, cells were labeled with CFSE prior to stimulation. (A) Assessment of absolute numbers of total lymphocytes and CD8⁺ T cells after α3/28 + BBR stimulation. (B) Mean fluorescence intensity (MFI) of CD44 and CD127 in CD8⁺ T cells. (C) Number of total lymphocytes and CD8⁺ T cells after α3/28 stimulation in the presence of BBR, CCM, or MTF. (D) CFSE assay used to analyze CD8⁺ T cell proliferation following BBR, CCM, or MTF treatment. Results are representative of two to four independent experiments (n = 3–4 per group). Data shown in bar graphs represent mean ± SD; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

The initial observation prompted us to investigate the role of BBR in memory and effector differentiation of CD8⁺ T cells by gating on CD62L and CD44. Stimulation with α3/28 resulted in Tcm (CD44⁺CD62L⁺) and Tem/Teff (CD44⁺CD62L⁻) cells in a ratio of approximately 1:3; however, inclusion of BBR increased the ratio above 1:1 (Fig. 2A). It is worth mentioning that there was a nearly negligible fraction of CD44⁻CD62L⁺ naïve or CD44⁻CD62L⁻ double-negative cells, apparently as a result of the high-level increase of CD44 expression in Tcm and Tem/Teff, respectively. The heightened Tcm-to-Tem/Teff ratio correlated with marked reduction in expansion capacity of Tem/Teff compared to that of Tcm (Fig. 2B). Indeed, regardless of BBR, Tcm increased in number in the first 24 h and remained relative stable by 72 h (Fig. 2C). Conversely, α3/28-stimulated Tem/Teff started increasing in 48 h but α3/28+BBR-stimulated cells failed to do so (Fig. 2C). Of note, both CD62L and CD44 levels were markedly elevated in Tcm (Fig. 2C), suggesting enhanced recirculation ability and activation status. The increased expression of CD127 in both Tcm and Tem (Fig. 2C) indicated more Tem than Teff induced by BBR. Such insufficient effector differentiation was verified by the low level of effector function CD39 (Fig. 2D). The above data reveal that BBR sustained Tcm with enhanced expression of CD62L and CD44 on the one hand and highly restrain Tem/Teff proliferation on the other. Obviously on this account, the overall expansion of CD8⁺ T cells shown above was dramatically suppressed.

Figure 2.

BBR improves Tcm formation while inhibiting Tem/Teff differentiation. Splenic lymphocytes were isolated from naïve mice and stimulated with α3/28 and BBR for 3 days and analyzed by flow cytometry. (A) Flow cytometry analysis of CD62L⁺CD44⁺ Tcm and CD62L⁻CD44⁺ Tem/Teff CD8⁺ T cells. CD62L⁺CD44⁻ naïve and CD62L⁻CD44⁻ double negative populations were also included. (B) Histogram showing CFSE labeling for analysis of Tcm and Tem/Teff. (C) Mean fluorescence intensity (MFI) of CD62L, CD44, and CD127 in Tcm on Day 3 and a time course study of Tcm and Tem numbers. (D) CD39 MFI and the number of CD39⁺ CD8⁺ T cells 3 days after stimulation. (E) Formation of Tcm and Tem/Teff in BBR, CCM, or MTF-modulated CD8⁺ T cells. Results are representative of two to six independent experiments (n = 3–4 per group). Data shown in bar graphs represent mean ± SD; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. ns, not significant.

Similar to BBR, CCM induced Tcm and Tem/Teff at a ratio greater than 1:1 (Fig. 2E), but it differed from BBR due to its inability to elevate CD44, implying that CCM was probably not as effective as BBR in driving the subset differentiation. In contrast to BBR, MTF stimulated T cells to have an almost complete differentiation bias toward effector with a Tcm:Tem/Teff ratio as low as 1:8 (Fig. 2E). Thus, these phytochemicals not only share similarities but also some important differences in differentiation of T cells. MTF potently suppressed Tcm differentiation in sharp contrast to BBR and CCM. It is possible that MTF drives the trafficking of activated T cells away from the lymphoid organs toward peripheral tissues, whereas CCM facilitates cell migration into both lymphoid and nonlymphoid tissues, in which BBR may be even more efficacious.

Due to the favorable performance compared to other functionally similar phytochemicals studied, it is important to understand the mechanisms by which BBR modulates T cell differentiation. Tcm regulation is known to be highly associated with Foxo1 and TCF1^24,25. Possibly, downregulation of Foxo1 expression with TCF1 inhibition might involve additional mechanisms. An increased expression of both Foxo1 and TCF1 was observed in α3/28+BBR-treated CD8⁺ T cells (Fig. S1D), and almost completely abolished by inhibition of the β-catenin activity with either iCRT3 or ICG001 (Fig. 3A). Both inhibitors downregulated CD62L, but iCRT3 was the major cause of inhibited CD44 expression (Fig. 3B). However, neither inhibitor was able to restore the suppressed number of Tem/Teff (Fig. 3B). We then showed that direct inhibition of Foxo1 activity by AS1842856 (AS18) led to a full reduction in the level of CD62L and substantially decreased expression of CD44, but still did not reverse the reduced Tem/Teff proliferation (Fig. 3C). The Foxo1-dependent expression of TCF1 was likely due to the highly conserved forkhead-binding motif on TCF1 for Foxo1's mediator activity since deficiency in Foxo1 could impair TCF1 expression, causing deficits in memory formation but limited effects on clonal expansion^24,26. Therefore, BBR-upregulated CD62L expression might depend on Foxo1 while CD44 upregulation was likely regulated in part by Foxo1 and TCF1 probably, probably due to the presence of additional transcriptional regulators involved, such as TCF4 and NF-κB.

Figure 3.

BBR-improved CD8⁺ Tcm differentiation depends on upregulation of Foxo1 and TCF1. (A–C) Expression of indicated molecules in CD8⁺ T cells and subsets. Splenic lymphocytes were isolated from naïve mice and stimulated with α3/28 + BBR and indicated inhibitors for 3 days (except ICG001 added in the last 24 h) before flow cytometry analysis. (A) MFI of Foxo1 and TCF1 in CD8⁺ T cells treated with iCRT3 or ICG001. (B) MFI of CD62L and CD44 in CD8⁺ T cells and number of CD62L⁺CD44⁺ Tcm and CD62⁻CD44⁺ Tem in the CD8⁺ T cells treated with iCRT3 or ICG001. (C) CD62L and CD44 levels in CD8⁺ T cells and the number of CD8⁺ Tcm and Tem cells in the culture containing AS1842856 (AS18). Results are representative of two to three independent experiments (n = 4–5 per group). Data are shown as mean ± SD; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

We next investigated whether BBR functions as an AMPK agonist to improve Tcm through upregulation of Foxo1 or TCF1. Significant upregulation of p-AMPKα (Thr172) was detected in BBR-treated cells (Fig. S1E) and the AMPK inhibitor compound C (CC) completely reversed the increased level of Foxo1 and TCF1 as well as CD62L and CD44 (Fig. 4A). This suggested that increased expression of CD44 was fully regulated by AMPK. The same applied to CD62L, highlighting the important role for AMPK in BBR-mediated Tcm. AMPK phosphorylation at Thr172 has been shown to be associated with fatty acid oxidation involved in metabolic transition from an effector to CD8⁺ memory T cells and their survival^27,28. It can directly phosphorylate Foxo1 to enhance nuclear localization. The resulting transcriptional activation could induce expression of memory-associated genes including TCF1^29,30. Thus, BBR-upregulated TCF1 in this study was likely attributed to AMPK-stimulated Foxo1 activity and may explain the observation that inhibition of the transcriptional activity of Foxo1 completely blocked the increased CD62L expression and hence hampered Tcm formation. It was then noted that the number of Tem/Teff slightly rebounded in AMPK-treated Tcm/Teff cells (Fig. 4B), giving the impression that activation of AMPK made a small contribution to the suppression of Tem/Teff. We however found that AMPK heterozygous (Prkaa1^+/−) CD8⁺ T cells yielded reduced numbers of Tcm compared to their wildtype counterparts (Fig. S1F). This is in line with previous work showing an inhibitory effect of AMPK on T-cell expansion based on the fact of modestly increased proliferation in AMPK-deficient CD8⁺ effector T cells³¹.

Figure 4.

BBR improves Tcm formation through orchestrated AMPK and Stat5 signaling. Expression of indicated molecules in CD8⁺ T cells and subsets. Splenic lymphocytes were isolated from naïve mice and stimulated with α3/28 + BBR and indicated inhibitors for 3 days (except CC added in the last 24 h) before flow cytometry analysis. (A) Expression of Foxo1, TCF1, CD62L, and CD44 in CD8⁺ T cells and the number of Tcm and Tem after α3/28 + BBR stimulation in the presence of CC (B) and BD750 (C). (D) Activation of AMPK and Stat5 in CD8⁺ T cells cultured with BD750 or CC. Results are representative of two to three independent experiments (n = 3–4 per group). Data are shown as mean ± SD; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

It was intriguing that the level of Stat5 was greatly elevated also by BBR (Fig. S1G). Foxo1 and CD62L decreased their expression with the Jak3/Stat5 inhibitor BD750 as they did with CC; however, TCF1 and CD44 were not or only partially affected (Fig. 4C). This implied that Stat5 might be required for AMPK regulation of Foxo1 and CD62L, but not for TCF1 and CD44. Further, BD750 not only impeded Tcm differentiation but also substantially abrogated the suppressive effect of BBR on Tem/Teff proliferation (Fig. 4C). Previously, some work reported that BBR disrupts the interaction between Jak2 and Stat5 to decrease Stat5 signaling, whereas other investigations revealed a greater inhibitory activity of BBR against Jak3 than other members of the Jak families through direct binding to the Jak3 kinase domain^32,33. In this study, the unexpected but major role for Stat5 in inhibiting, rather than stimulating, T cells could be attributed to differences in its transactivation levels³⁴ or may reflect of selective effects the multipotent BBR exerts depending on the context and state of the cell.

To figure out how AMPK and Stat5 were both involved in the differentiation processes, we investigated further to determine whether they were influenced by each other. Interestingly, p-AMPK and p-Stat5 each could be fully downregulated by the inhibitor of the other (Fig. 4D), suggestive of a possible crosstalk between them. It was likely that BBR stimulated AMPK and Stat5 to cross-activate each other; further, AMPK might impact the increase of CD62L and CD44, while the Jak3/Stat5 pathway took the primary role in inhibiting the proliferation of Tem/Teff.

To verify that BBR-improved CD8⁺ Tcm cells could respond specifically to the targeted antigen in an infection, we stimulated CD45.2⁺ OT-I splenocytes with OVA_257–264 (SIINFEKL) peptides, poly(I:C), CpG ODN, and BBR (OPC + BBR) for 3 days, and then MACS/FACS-sorted SIINFEKL-specific (tetramer⁺) CD62L⁺CD44⁺ CD8⁺ T cells for adoptive transfer to CD45.1⁺ naïve congenic mice 1 day before infection with LmOVA to establish infection (Fig. 5A). Using imaging flow cytometry for single-cell analysis of Tcm, we noted that CD44 expression on OPC + BBR-primed Tcm was polarized compared to OPC-primed Tcm (Supporting Information Fig. S2A). Polarized expression was also observed for CD62L but on OPC-primed Tcm; CD62L expression on OPC + BBR-primed Tcm was more concentrated than OPC-primed Tcm (Fig. S2A). These data suggested a distinct CD44 and CD62L expression pattern in the two different kinds of Tcm cells. Upon antigen stimulation, the purified OPC + BBR-induced Tcm were able to produce IFN-γ several fold relative to OPC-induced Tcm, suggesting an increased cytotoxic potential for BBR-improved Tcm (Fig. S2B). In the OPC + BBR Tcm recipient which were highly resistant to infection compared to SPC Tcm recipient mice and nontransferred controls (Fig. 5A), the majority of proliferation (Fig. S2C and S2D). At d35 after immunization, there was a slight increase of BBR/vaccination-induced Tcm in percentage compared to vaccination only, but the percentage of Tem was significantly less in BBR/vaccination mice than in vaccination mice although the decrease was not as dramatic as that seen in vitro, presumably because of the longer period of time after vaccination (Fig. S2E). Therefore, BBR-modulated Tcm were antigen-responsive and likely to provide immediate protection through Tem differentiation. Given that BBR-improved Tcm showed protective effects, it was of interest to determine whether immunization of BBR-fed mice would induce protection at the memory phase. BBR-fed mice were immunized s.c. by prime-boost with a mixture of OVA protein, poly(I:C), and CpG1826 ODNs, and inoculated i.v. with LmOVA at 35 day. More effective memory protection was detected in mice fed BBR-containing diet than fed regular chow (Fig. 5B). BBR conferred better memory protection presumably owing to higher expression of CD62L and CD44 that may improve the quality of Tcm with enhanced repopulating ability to control systemic infection. To test whether immunized BBR-treated mice have an immediate effector protection, LmOVA i.v. challenge was conducted at 7 days after boost. Mice fed BBR-containing diet were protected just as equally well as those fed control diet at 7 days after boost (Fig. 5C). Therefore, vaccination of BBR-modulated mice can provide sufficient effector protection and, importantly, better memory protection. Nonetheless, strategic application of BBR and longer-term protective immunity merit further investigation.

Figure 5.

Vaccination with BBR confers better memory protection with unimpaired immediate effector function. Freshly isolated splenocytes from CD45.2⁺ OT-I mice were stimulated with SIINFEKL, poly(I:C), CpG1826 ODNs, and BBR for 3 days, followed by MACS enrichment of CD8⁺ T cells and FACS sorting for tet⁺CD62L⁺CD44⁺ Tcm cells. Purified cells (5 × 10⁴) were transferred into naïve (nonBBRpre) or BBRpre CD45.1 mice 1 day before i.v. challenge with 1 × 10⁵ CFU LmOVA. A scheme is given in each panel to show the sorting strategy for tet + Tcm for adoptive transfer or the immunization protocol. (A) Survival and bodyweight changes after challenge of recipient mice with LmOVA. (B, C) Mice were primed with OVA, poly(I:C), and CpG1826 ODNs consecutively for 3 days, followed by boost from Day 11 to Day 13. Immunized mice were then challenged with 1 × 10⁶ LmOVA by i.v. injection on Day 35 (B) or Day 21 (C) after primary vaccination. Results are representative of two independent experiments (n = 4–5 per group). Data are shown as mean ± SD; ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. Comparisons between groups were analyzed by 2-tailed Student's t test. Analyses were performed with SPSS for Windows (SPSS).

As shown in this study, BBR did not harm vaccine-induced immunity and rather surprisingly helped elicit a good desirable one. A potentially useful feature of future applying BBR in a vaccine as modulatory adjuvant would be the provision of better durable immunity and meanwhile minimization of proliferative/cytolytic Tem/Teff without losing immediate effector efficacy but making excessive cytotoxicity or immune-related adverse reactions preventable^35,36, thereby offering much safer and more satisfying experience for vaccinees. In the current pandemic or in preparing to face a new one, increasing mass vaccination rates is a critical step toward ending global pandemic infectious disease especially for countries with large populations of the elderly, many of which are on medications including a variety of natural products, some of which are known to weaken or exacerbate the response to vaccines. A systems approach to dissecting their respective effects would be very helpful to identify and discriminate among herbal medicines and natural products and eventually find the right benefits to not only older and diseased individuals but also young and healthy people when mass vaccination begins.

The T cell-mediated response induced by BBR may vary depending mostly likely on how these cells are activated. It is generally accepted that BBR-induced responses are tolerogenic. For example, a marked reduction in CD69 expression is shown in T cells in BBR-treated mice with a heart allograft³⁷. We found, however, that CD69 expression on CD8⁺ T cells was significantly upregulated after TCR stimulation in the presence of BBR. Indeed, T cells are modulated, manifested when activated by α3/28 or antigen in adjuvants with increased production of Tcm and decreased production of Tem. It appears that a more thorough investigation into the characteristics of BBR modulated T-cell differentiation is needed in future studies.

4. Conclusions

In conclusion, BBR acts as an immunomodulator to shape CD8⁺ T cell differentiation with improved central memory formation and reduced effector properties orchestrated by AMPK and Jak3/Stat5. Vaccination combined with BBR provides a better memory response at no cost of losing immediate effector efficacy. This work sheds light on the cellular and molecular basis of BBR in immunomodulation of T cell-mediated immunity and may have important implications for the use of BBR and potentially other naturally occurring or derived compounds as herbal adjuvants in future vaccine and therapeutic development.

Acknowledgments

This work was supported by National Natural Science Foundation of China (81871784 and 82171822), CAMS Major Collaborative Innovation Project (2016-I2M-1-011, China), Beijing Key Laboratory of New Drug Mechanisms and Pharmacological Evaluation Study Project (BZ0150, China), Graduate Innovation Fund (2018-1007-05, China), and Yunnan Science and Technology Talent and Platform Program (202105AG070012, China).

Author contributions

Mingyan Li and Qing Zhu designed the model and planned the experiments. Mingyan Li, Yaling Wang, Lingzhi Zhang, Changxing Gao, and Jing J. Li were involved in the experimental design, implementation of methods, and data analysis and interpretation. Mingyan Li wrote the draft and Qing Zhu revised the manuscript. Jiandong Jiang and Qing Zhu conceived the original idea and contributed to the overall direction and the interpretation of findings.

Conflicts of interest

The authors declare no conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article: