Extreme prematurity and sepsis strongly influence frequencies and functional characteristics of circulating γδ T and natural killer cells

Abstract

Objectives: Extremely low gestational age neonates with extremely low birthweight (ELGAN/ELBW) are highly susceptible to infection. This is linked to their relatively immature immune system which is not yet fully compatible with an extra-uterine environment. Here, we performed a longitudinal characterisation of unconventional T and natural killer (NK) cells in ELGAN/ELBW during their first months of life.

Methods: Peripheral blood mononuclear cells were collected from 97 ELGAN/ELBW at 14 and 28 days of life and at a time point corresponding to postmenstrual week 36 + 0. γδ T-cell, NKT-cell, mucosa-associated invariant T-cell and NK cell frequencies and characteristics were analysed by flow cytometry. As control, cells from 14-day-old full-term (FT) infants were included.

Results: Extreme prematurity had significant bearing on γδ T-cell and NK cell frequencies and characteristics. ELGAN/ELBW had significantly higher proportions of γδ T cells that were skewed towards effector and effector memory phenotypes, characteristics that were maintained throughout the study period. Expression of the gut homing receptor CCR9 was also more common in γδ T cells from ELGAN/ELBW. Conversely, NK cell frequencies were markedly lower and skewed towards a cytotoxic phenotype in the ELGAN/ELBW group at 14 days of age. Culture-proven sepsis with an onset during the first 14 days after birth further manifested these differences in the γδ T- and NK cell populations at 14 days of age.

Conclusion: Prematurity strongly influences the levels of γδ T and NK cells, in particular in cases where sepsis debuts during the first 2 weeks of life.

Keywords: extreme preterm; gestational age; natural killer cells; neonatal immunity; sepsis; unconventional T cells.

Figure 1

Flow chart of the number of subjects recruited, time points for blood sampling, number of PBMC samples analysed by flow cytometry and the antibody panels used for the experiments to characterise NK, NKT γδ T and MAIT cells. The final number of PBMC samples analysed for each parameter were dependent on cell numbers and viability, as indicated in each of the subsequent figures.

Figure 2

Schematic presentation of the gating strategies used for the identification and characterisations of the NK cells and unconventional T cells and subsets. NK/NKT cells and γδ T‐/MAIT cells were analysed in two different panels as indicated in the flow chart with red and black arrows, respectively. Lymphocytes (G1) were gated based on their physical properties as side scatter (SSC) and forward scatter (FSC) followed by gating on single cells (G2). From G2 gate, cells were either divided as live CD3⁺ cells (G3) (black arrow) or as live lymphocytes (G4) (red arrow), which were further gated on the basis of the cell population‐specific markers as pan γδ TCR⁺ cells (γδ T cells) and CD161⁺Vα7.2⁺ cells (MAIT cells) (black arrows) or as CD3⁺ Vα24jα18⁺ cells (NKT cells) and CD3^‐CD56⁺ cells (total NK cells)/CD56⁺CD16⁺(NK^bright/dim) (red arrows), respectively. γδ T cells were further analysed for either the subpopulations (CM, NA, EC and EM) or the expression of the molecule as CCR9.

Figure 3

Peripheral γδ T‐, NKT‐cell and NK cell frequencies are altered by extreme prematurity. Frequencies of (a) CD3⁺, (b) γδ TCR⁺, (c) NKT, (d) MAIT, and (e) NK cells among live lymphocytes in 14‐day‐old ELGAN/ELBW and FT infants are shown. The correlations between gestational age at birth and the percentages of (f) CD3⁺, (g) γδ T, (h) NKT, (i) MAIT and (j) NK cells at 14 days of life are shown for the ELGAN/ELBW. The postnatal development of these populations is shown as proportions of (k) CD3⁺, (l) γδ T, (m) NKT, (n) MAIT and (o) NK cells in longitudinal paired samples from ELGAN/ELBW at D14, D28 and W36 (k, l, n) or D14 and W36 (m, o), respectively. The Mann–Whitney U‐test for unpaired samples and the Wilcoxon signed‐rank test for paired samples were used for group comparisons. Box and whisker plots show median as the central line and error bars represent minimum to maximum values. The line with error bars shows mean and error where the Spearman correlation test was used to analyse correlation between gestational age and the proportions of the different cell types analysed. Symbols and lines were overlaid on floating bars plotted with line at mean (k, l).

Figure 4

Extreme prematurity alters NK cell and γδ T‐cell phenotypical characteristics. The relative frequencies of (a) CD56^dim‐ and (b) CD56^bright NK cells and (c) ratio between dim and bright NK cells at 14 days of age in ELGAN/ELBW and FT infants. (d–g) show the relative proportions of naïve (NA), effector (EC), effector memory (EM) and central memory (CM) γδ T cells in FT neonates and ELGAN/ELBW at the different sampling time points. The ratio of EC:NA γδ T cells is shown in (h), and the frequencies of CCR9⁺ γδ T cells at 14 days of age are shown in (i). The Mann–Whitney U‐test and Kruskal–Wallis test with Dunn's multiple comparison were used for group comparisons. Box and whisker plots show median as the central line and error bars represent minimum to maximum values for group comparison. PCA comparing the overall phenotypic characterisation in a longitudinal manner of MAIT and γδ T cells (j), or the maturation over time of frequencies of all four investigated cell types (k). FT neonates (blue), ELGAN/ELBW D14 (red), D28 (orange) and W36 (green).

Figure 5

Sepsis accentuates the effect of extreme prematurity on the NK cell compartment. (a–c) show the proportions of total NK, CD56^dim‐cell and CD56^bright NK‐cell frequencies, respectively, in 14‐day‐old ELGAN/ELBW with sepsis onset within 14 days of life, after day 14 of age, or no sepsis. (d–f) show the proportions of total NK, CD56^dim‐cell and CD56^bright NK cell frequencies, respectively, in 28‐day‐old ELGAN/ELBW with sepsis onset within 28 days after birth or no sepsis. The proportions of total NK, CD56^dim and CD56^bright NK cells in paired samples from day 14 and week 36 are shown for cases with (g–i) or without (j–l) sepsis. Box and whisker plots show median as the central line and error bars represent minimum to maximum. The Mann–Whitney U‐test was used for group comparisons, and the Wilcoxon signed‐rank test was used to assess differences between pairs.

Figure 6

Increased γδ T‐cell frequencies, a rise in effector populations and an altered homing capacity are parameters associated with sepsis onset within 14 days of life. (a–c) The frequencies of γδ T cells from ELGAN/ELBW at day 14, day 28 and week 36, respectively, in relation to time of disease onset as indicated in the figure. The proportions of γδ T cells in paired samples from day 14 and week 36 are shown for cases with (d) or without (e) sepsis. (f–i) show the proportions of γδ T‐cell naïve, effector and memory subpopulations at 14‐day‐old ELGAN/ELBW, (j) shows the ratio between EC:NA γδ T cells and (k) displays the frequencies of CCR9⁺ γδ T cells in relation to sepsis onset. The proportions of CCR9⁺ γδ T cells in paired samples from day 14 and week 36 are shown for cases with (l) or without (m) sepsis. Box and whisker plots show median as the central line and error bars represent minimum to maximum. The Mann–Whitney U‐test and Kruskal–Wallis test with Dunn's multiple comparison were used for group comparisons. The Wilcoxon signed‐rank test was used to assess differences between paired samples.

Figure 7

Lactobacillus reuteri supplementation has no effect on unconventional T cells. (a–c) show PCA plots at day 14, day 28 and week 36 of γδ T‐ and MAIT‐cell frequencies and phenotypic characteristics from ELGAN/ELBW receiving L. reuteri supplementation (green) or placebo (red).