The metabolic hormone leptin promotes the function of TFH cells and supports vaccine responses

Abstract

Follicular helper T (T_FH) cells control antibody responses by supporting antibody affinity maturation and memory formation. Inadequate T_FH function has been found in individuals with ineffective responses to vaccines, but the mechanism underlying T_FH regulation in vaccination is not understood. Here, we report that lower serum levels of the metabolic hormone leptin associate with reduced vaccine responses to influenza or hepatitis B virus vaccines in healthy populations. Leptin promotes mouse and human T_FH differentiation and IL-21 production via STAT3 and mTOR pathways. Leptin receptor deficiency impairs T_FH generation and antibody responses in immunisation and infection. Similarly, leptin deficiency induced by fasting reduces influenza vaccination-mediated protection for the subsequent infection challenge, which is mostly rescued by leptin replacement. Our results identify leptin as a regulator of T_FH cell differentiation and function and indicate low levels of leptin as a risk factor for vaccine failure.

Fig. 1. Low leptin levels are associated with lower antibody responses after vaccination in the general population.

a Comparison of serum leptin levels between responders (n = 71, grey) and non-responders (n = 5, red) in healthy adults (18–60 years, n = 76) immunized with the influenza vaccine. Dotted line indicates healthy individuals with low leptin levels (lower 25%, <4.8 ng/mL) and sufficient leptin levels (higher 75%, ≥4.8 ng/mL). b Comparison of response rates between low leptin (n = 19) and sufficient leptin (n = 57) groups. c Comparison of serum leptin levels between responders (n = 68, grey) and non-responders (n = 15, red) in healthy elderly adults (>65 years, n = 83) immunized with the influenza vaccine. Dotted line indicates healthy individuals with low leptin levels (lower 25%, <8.8 ng/mL) and sufficient leptin levels (higher 75%, ≥8.8 ng/mL). d Comparison of response rates between low leptin (n = 21) and sufficient leptin (n = 62) groups. e Comparison of serum leptin levels between responders (n = 14, grey) and non-responders (n = 8, red) in healthy young adults (18–25 years, n = 22) immunized with the HBV vaccine. Dotted line indicates healthy individuals with low leptin levels (lower 25%, <3.2 ng/mL) and sufficient leptin levels (higher 75%, ≥3.2 ng/mL). f, Comparison of response rates between low leptin (n = 8) and sufficient leptin (n = 14) groups. Data from individuals (dots) and the median values (bars) are shown for a, c and e; Mann–Whitney test was used for analysis (a, c, e). Percentages and numbers are shown for b, d and f; Chi-Square tests were used for analysis (b, d, f). Detailed demographics of each cohort are shown in Supplementary Tables 1-3. For correlation of leptin levels and antibody responses see Supplementary Fig. 1.

Fig. 2. Leptin potentiates the differentiation and function of human T_FH cells.

a Comparison of serum leptin levels between healthy individuals with T_FH increase (fold changes of T_FH cells, the ratio of day 7/day 0 > 1) (n = 61, grey) and healthy individuals without T_FH increase (fold changes of T_FH cells, the ratio of day 7/day 0 ≤ 1) (n = 15, red) in healthy adults (18–60 years, n = 76) immunized with the influenza vaccine (Supplementary Table 4). Dotted line indicates healthy individuals with low leptin levels (lower 25%) and sufficient leptin levels (higher 75%). b Comparison of T_FH changes and vaccine response in low leptin group (n = 19) and sufficient leptin group (n = 57). c Real-time PCR analysis of T_FH-related genes in naïve CD4⁺ T cells from PBMC of healthy controls with anti-CD3/CD28 activation for 3 days, and further treated with leptin for 12 h (CXCR5: *P = 0.0176, Bcl6: **P = 0.0046, IL21: **P = 0.0085, PDCD1: **P = 0.0050, ICOS: **P = 0.0344) (n = 7). d, e Representative FACS plots and statistics showing the differentiation of CXCR5⁺PD-1⁺ T_FH cells (d) (100 ng/mL: *P = 0.0257, 200 ng/mL: **P = 0.0012, 400 ng/mL: **P = 0.0005, 800 ng/mL: **P < 0.0001), and the secretion of IL-21 (e) (100 ng/mL: **P = 0.0004, 200 ng/mL: **P < 0.0001, 400 ng/mL: **P < 0.0001, 800 ng/mL: **P < 0.0001) from purified human naïve CD4⁺ T cells stimulated with anti-CD3/CD28 and indicated concentrations of leptin for 4 days (dots, n = 5; bars, mean ± SEM). f, g Statistics showing the percentages of CXCR5⁺PD-1⁺ T_FH cells (f) (IL-12: *P = 0.0105, *P = 0.0107; IL-23: **P = 0.0063, **P = 0.0087; TGF-β: *P = 0.0187,*P = 0.0161; IL-12+TGF-β: **P = 0.0063,**P = 0.0052; IL-23+TGF-β: *P = 0.0452,*P = 0.0211; IL-12+IL-23+TGF-β: *P = 0.0206,**P = 0.0023) and IL-21⁺ cells (g) (IL-12: **P = 0.0068, **P = 0.0071; IL-23: **P < 0.0001, **P = 0.0012; TGF-β: *P = 0.0287,**P = 0.0096; IL-12+TGF-β: **P = 0.0023,*P = 0.0129; IL-23+TGF-β: *P = 0.0145,**P = 0.0044; IL-12+IL-23+TGF-β: **P = 0.0084,**P = 0.0082) differentiated from naïve CD4⁺ T cells with anti-CD3/CD28 stimulation and indicated cytokine combinations for 4 days (dots, n = 5; bars, mean ± SEM). Data from individuals (dots) and the median values (bars) are shown for (a); Mann–Whitney test was used for analysis (a). Percentages and numbers are shown for (b); Chi-Square test was used for analysis (b). Real-time PCR and FACS statistics are shown for individuals (dots, n = 3) and mean ± SEM (bars) values, and analysed by Mann–Whitney U-test (a, c), one-way ANOVA Dunnett’s multiple comparisons test (d–g). *P < 0.05, **P < 0.01. Results are representative of three independent experiments.

Fig. 3. LepR deficiency leads to impaired antibody responses after influenza virus infection.

a Real-time PCR analysis of nucleoprotein (Np) in lung tissues of WT (grey) and db/db (red) mice 9 days post H1N1 influenza viral infection (**P = 0.0007). b, c ELISA measurement of H1N1-specific IgG1, IgG2b, IgG2c and IgG3 in sera (b) (IgG1: *P = 0.0190, IgG2b: **P < 0.0001, IgG2c: *P = 0.0299, IgG3: *P = 0.0209) and BALF (c) (IgG1: **P = 0.0080, IgG2b: **P = 0.0036, IgG2c: **P = 0.0079, IgG3: **P = 0.0007). d–f Representative FACS plots and statistics showing CXCR5⁺Bcl6⁺ T_FH cells (d) (*P = 0.0407), GL-7⁺Fas⁺ germinal centre (GC) B (e) (*P = 0.0240) and CD138^high antibody-secreting cells (ASCs) (f) (*P = 0.0411) in mediastinal lymph nodes. g ELISPOT assay showing virus-specific ASCs in spleens. Data are shown for individuals (dots, n = 5 per genotype) and mean (bars) values, and analysed by Mann–Whitney U-test (a–f). *P < 0.05, **P < 0.01. Results are representative of three independent experiments.

Fig. 4. Defective T_FH cells responses in Cd4-Cre:LepR^fl/fl mice after influenza virus infection.

a–e Cd4-Cre:LepR^+/+ mice (grey, n = 11) and Cd4-Cre:LepR^fl/fl mice (red, n = 8) were infected with influenza virus A/X-31 (H3N2). Mediastinal lymph nodes were analysed at day 9 post-infection. Representative FACS plots and statistics showing the CD4⁺Foxp3^-CD44^high effector cells, CD4⁺Foxp3⁺ T_REG cells (a), CD44⁺CXCR5⁺Bcl6⁺ T_FH (b), (*P = 0.0499, *P = 0.0347), IL-21 production (c), (*P = 0.0313, *P = 0.0142), B220⁺GL-7⁺Fas⁺ GC B cells (d) (*P = 0.0151, *P = 0.0249) and B220^-CD138⁺TACI⁺ ASCs (e) (*P = 0.0250, *P = 0.0409) 9 days post influenza viral infection. f Cd4-Cre:LepR^+/+ and Cd4-Cre:LepR^fl/fl mice were intranasally challenged with H1N1 influenza virus and virus-specific IgG1, IgG2b and IgG2c in sera were measured by ELISA 9 days post-infection (n = 7 per genotype) (IgG1: *P = 0.0183, IgG2b: *P < 0.0181, IgG2c: *P = 0.0168). Data are shown for individuals (dots) and mean (bars) values, and analysed by Mann–Whitney U-test (a–f). *P < 0.05, **P < 0.01. Results are representative of two independent experiments.

Fig. 5. Leptin promotes IL-21 production in a STAT3-dependent manner.

a B220^-CD4⁺CD25^-CD44⁺CD62L^-CXCR5⁺PD-1⁺ T_FH cells (1 × 10⁵) and B220⁺Fas⁺GL-7⁺ GC B cells (1 × 10⁵) from 4-Hydroxy-3-nitrophenylacetyl hapten conjugated to ovalbumin (NP-OVA) immunized WT and db/db mice (9 dpi) were co-cultured with 5 µg/mL NP-OVA for 9 days. In the db/db T_FH:WT B-cell culture, IL-21 (10 ng/mL) was added. Anti-NP IgG1 titers in the supernatant were measured by ELISA (IL-21_10 ng/mL: *P = 0.0403, WT: **P = 0.0033). b ELISA measurement of IL-21 in cultured naive CD4⁺ T cells from WT mice with anti-CD3/CD28 and leptin (0-200 ng/mL) stimulation for 3 days (100 ng/mL: **P = 0.0100, 200 ng/mL: **P = 0.0441). c Western blot showing Stat3 phosphorylation (p-Stat3) in WT naive CD4⁺ T cells stimulated with 200 ng/mL leptin for 2 h. Values showing the fold changes relative to the non-treated control. d Binding of Stat3 to the Il21 promoter. Stat3-ChIP assays were performed on WT naive CD4⁺ T cells treated with 200 ng/mL leptin for 3 h. Results showing PCR products (left) and values for the fold changes in ChIP enrichment relative to non-treated control (right) (100 ng/mL: **P = 0.0096, 200 ng/mL: **P = 0.0043). e Schematic of Il21 promoter construction and luciferase assay. f The transcriptional activity of the Il21 promoter. Dual-Luciferase reporter assay for the Il21 promoter in 293 T cells with 200 ng/mL leptin treatment for 12 h (100 ng/mL: **P = 0.0082, 200 ng/mL: **P = 0.0016). g Cd4-Cre:Stat3^+/+ or Cd4-Cre:Stat3^fl/fl naive CD4⁺ T cells were stimulated under IL-21-inducing conditions with 200 ng/mL leptin for 3 days. Results for IL-21 secretion showing representative FACS plots (left) and statistics (right) (nil: *P = 0.0111, Leptin: *P = 0.0144). Data are shown for individuals (dots) and mean (bars) values, and analysed by two-way ANOVA (a) one-way ANOVA (b, d, f) or Mann–Whitney U-test (g). *P < 0.05, **P < 0.01. Results are representative of three independent experiments.

Fig. 6. Stat3 and mTOR are dispensable for leptin-induced T_FH differentiation and Bcl6 expression.

a Real-time PCR analysis of mouse T_FH-related genes in naïve CD4⁺ T cells from WT mice stimulated with anti-CD3/CD28 for 2 days, further treated with leptin for 12 h (Cxcr5: **P = 0.0029, Bcl6: **P = 0.0004, Il21: **P = 0.0007, Icos: **P = 0.0028, Cd40l: **P = 0.0023) (n = 5). b Representative FACS plots and statistics of CXCR5⁺PD-1⁺ T_FH cells in cultured naïve CD4⁺ T cells from WT mice with anti-CD3/CD28 stimulation for 48 h, followed with or without leptin treatment for 12 h (Ratio: *P = 0.0200, GMFI: *P = 0.0243). c–e Western blot assay of Bcl6 with or without anti-CD3/CD28 activation (c), the phosphorylation of Akt, mTOR, p70S6K and S6 (d) and Bcl6 expressions in WT naïve CD4⁺ T cells with Stat3 inhibition (S3I-201, 100 μM), mTOR inhibition (Rapamycin, 200 nM) plus leptin (200 ng/mL) treatment for 12 h (e). Values showing the fold changes relative to the non-treated control. f Representative FACS plots and statistics of PD-1⁺Bcl6⁺ T_FH cells in cultured naïve CD4⁺ T cells from WT mice with anti-CD3/CD28 activation for 2 days, followed with Stat3 inhibition (S3I-201, 100 μM), mTOR inhibition (Rapamycin, 200 nM) plus leptin (200 ng/mL) treatment for 12 h (0 vs nil: **P = 0.0004, S3I vs 0: **P = 0.0023, Rapa vs 0: **P = 0.0026) (n = 5). g, h Representative FACS plots of PD-1⁺Bcl6⁺ T_FH cells in cultured naïve CD4⁺ T cells from Cd4-Cre:Stat3^fl/fl mice (g) and Cd4-Cre:Rictor^fl/fl mice (h) with anti-CD3/CD28 activation for 2 days, followed with or without leptin (200 ng/mL) treatment for 12 h (n = 5). Data are shown for individuals (dots) and mean (bars) values, and analysed by Mann–Whitney U-test (a, b), or two-way ANOVA (f). *P < 0.05, **P < 0.01. Results are representative of three independent experiments.

Fig. 7. Fasting impairs vaccination-mediated protection, which can be compensated by leptin replacement.

a WT female mice were immunized (s.c.) with human influenza vaccine, following various feeding regimens: normal diet (Fed) (n = 18, black line), fasting with PBS (Fast + PBS) (n = 18, red line), fasting with leptin treatment (Fast + leptin) (n = 19, cyan line); mice were intranasally challenged with H1N1 influenza virus and analysed 6 days post-infection. b Accumulative survival curve of mice after viral challenge (Fast + PBS vs Fed: *P = 0.0369). c ELISA measurement of H1N1-specific IgG1, IgG2b and IgG2c in sera (IgG1: **P = 0.0073, **P = 0.0076; IgG2b: *P = 0.0133, *P = 0.0258; IgG2c: *P = 0.0326, *P = 0.0252) and bronchoalveolar lavage fluid (BALF) (IgG1: *P = 0.0488, *P = 0.0240; IgG2c: *P = 0.0271, *P = 0.0481). d–f Representative FACS plots and statistics showing frequencies of GC B cells in B cells (d) (*P = 0.0446, *P = 0.0197) and ASCs in lymphocytes (e) (*P = 0.0261, **P = 0.0081), and T_FH cells in CD44⁺Foxp3⁺ CD4⁺ T cells (f) (**P = 0.0022, *P = 0.0197) in the mediastinal lymph nodes 6 days post-infection. Data are shown for individual (dots, n = 7) and mean (bars) values, and analysed by Log-rank tests (b), two-way ANOVA (c–f). *P < 0.05, **P < 0.01. Results are representative of two independent experiments.