Circulating CXCR5⁺CD4⁺ T Follicular-Like Helper Cell and Memory B Cell Responses to Human Papillomavirus Vaccines

Abstract

Through the interaction of T follicular helper (Tfh) cells and B cells, efficacious vaccines can generate high-affinity, pathogen-neutralizing antibodies, and memory B cells. Using CXCR5, CXCR3, CCR6, CCR7, PD1, and ICOS as markers, Tfh-like cells can be identified in the circulation and be classified into three functionally distinct subsets that are PD1+ICOS+, PD1+ ICOS-, or PD1-ICOS-. We used these markers to identify different subsets of CXCR5+CD4+ Tfh-like cells in response to highly immunogenic and efficacious vaccines for human papillomaviruses (HPV): Cervarix and Gardasil. In this small study, we used PBMC samples from 11 Gardasil recipients, and 8 Cervarix recipients from the Vaccine Research Center 902 Study to examine the induction of circulating Tfh-like cells and IgD-CD38HiCD27+ memory B cells by flow cytometry. PD1+ICOS+ CXCR3+CCR6-CXCR5+CD4+ (Tfh1-like) cells were induced and peaked on Day (D) 7 post-first vaccination, but not as much on D7 post-third vaccination. We also observed a trend toward increase in PD1+ICOS+ CXCR3-CCR6-CXCR5+CD4+ (Tfh2-like) cells for both vaccines, and PD1+ICOS+ CXCR3-CCR6+CXCR5+CD4+ (Tfh17-like) subset was induced by Cervarix post-first vaccination. There were also minimal changes in the other cellular subsets. In addition, Cervarix recipients had more memory B cells post-first vaccination than did Gardasil recipients at D14 and D30. We found frequencies of memory B cells at D30 correlated with anti-HPV16 and 18 antibody titers from D30, and the induction levels of memory B cells at D30 and PD1+ICOS+Tfh1-like cells at D7 post-first vaccination correlated for Cervarix. Our study showed that induction of circulating CXCR5+CD4+ Tfh-like subsets can be detected following immunization with HPV vaccines, and potentially be useful as a marker of immunogenicity of vaccines. However, further investigations should be extended to different cohorts with larger sample size to better understand the functions of these T cells, as well as their relationship with B cells and antibodies.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Fig 1. Gating strategy to identify CXCR5+ CM cells.

(A) Gating scheme used to select live CD3+ CD4+ cells. (B) Live CD3+ CD4+ cells were used to identify naive, central memory, and effector memory cells, and CXCR5 expression on these cellular populations was examined.

Fig 2. Longitudinal analysis of circulating Tfh1-, Tfh2-, Tfh17-like subsets and their PD1 and ICOS expression.

One of the HPV vaccine samples (Gardasil) is shown. The contour plots on the left most column show the expression of CXCR3 and CCR6 on CXCR5+ CM cells in samples collected from the indicated time points that are shown on the right. These two markers were used to identify Tfh1-like (blue), Tfh2-like (red), and Tfh17-like (purple) subsets. The percentages of each of these subsets are shown inside the plots. The second, third, and fourth columns of plots show the expression of PD1 and ICOS in each of the Tfh-like subset. In the upper right quadrants, the percentages of PD1+ ICOS+ Tfh-like cells are shown. The percentages of PD1+ ICOS- and double negative cells are indicated in the lower right and lower left quadrants, respectively.

Fig 3. Frequency distributions of different subsets of circulating Tfh-like cells.

(A) Percentages of circulating Tfh1-, Tfh2-, and Tfh17-like subsets from the Gardasil and Cervarix groups are shown. (B) Frequencies of PD1/ICOS double positive cells, (C) frequencies of PD1+ ICOS- cells, and (D) frequencies of double negative cells within each subset were plotted over time. N = 10–11 (pre-vac to D30) and N = 6–8 (M6 to M7) for Gardasil. N = 8 (pre-vac to D30) and N = 5–6 (M6 to M7) for Cervarix. Paired, one-tailed Wilcoxon rank sum analyses were performed between the pre-vac time point and each of the post-vaccination time point for the first vaccination. For the third vaccination (M6 to M7), M6 was used as the baseline for the analyses. Black horizontal bars indicate the medians. For the comparison of the two vaccine groups described in the “HPV vaccines induced circulating PD1+ICOS+ Tfh1-like cells” section, paired, two-tailed Wilcoxon rank sum analyses were performed.

Fig 4. Generation of IgD-CD38HiCD27+ memory B cells after vaccination.

(A) Gating strategy used to identify IgD-CD38HiCD27+ memory B cells is shown. (B) Percentages of memory B cells in Gardasil or Cervarix immunized recipients were plotted over time. Statistical analyses were performed as described in Fig 3. Black bars indicate the medians. For the comparison of the two vaccine groups described in the “A higher frequency of IgD-CD38HiCD27+ memory B cells was present in Cervarix recipients at D30 post-first vaccination” section, paired, two-tailed Wilcoxon rank sum analyses were performed. N = 10–11 (pre-vac to D30) and N = 6–8 (M6 to M7) for Gardasil. N = 8 (pre-vac to D30) and N = 5–6 (M6 to M7) for Cervarix. (C and D) ELISA were performed to determine the titers of anti-HPV16 (C) and-HPV18 (D) IgG. Geometric mean antibody titers (EU/ mL) ± 95% confidence intervals in log10 scale were plotted as a function of time. The plots on the left show the titers post-first vaccination, and the plots on the right show the titers after the third vaccination. Paired, two-tailed Wilcoxon rank sum analyses were performed. N = 12–15 Gardasil, and N = 8–12 for Cervarix.

Fig 5. Memory B cell frequencies at D30 correlate with the titers of anti-HPV16 and-HPV18 antibodies.

Spearman rank correlational analyses were performed on the percentages of IgD-CD38HiCD27+ B cells at D30 post-first vaccination with the antibody titers of anti-HPV16 and-HPV18 antibodies from D30. The data were plotted in log10 scale for both axes.