Microbial exposure during early human development primes fetal immune cells

Abstract

The human fetal immune system begins to develop early during gestation; however, factors responsible for fetal immune-priming remain elusive. We explored potential exposure to microbial agents in utero and their contribution toward activation of memory T cells in fetal tissues. We profiled microbes across fetal organs using 16S rRNA gene sequencing and detected low but consistent microbial signal in fetal gut, skin, placenta, and lungs in the 2^nd trimester of gestation. We identified several live bacterial strains including Staphylococcus and Lactobacillus in fetal tissues, which induced in vitro activation of memory T cells in fetal mesenteric lymph node, supporting the role of microbial exposure in fetal immune-priming. Finally, using SEM and RNA-ISH, we visualized discrete localization of bacteria-like structures and eubacterial-RNA within 14^th weeks fetal gut lumen. These findings indicate selective presence of live microbes in fetal organs during the 2^nd trimester of gestation and have broader implications toward the establishment of immune competency and priming before birth.

Keywords: Tem; Treg; bacteria; fetal Development; fetal immunity; immune memory; immune priming; microbes; microbiome.

Graphical abstract

Figure S1

Study design and T cell diversity during the 2^nd trimester of human fetal development, related to Figure 1 (A) Schematic for introducing the study design and techniques used. Precisely, fetal tissues were harvested from the human fetuses under sterile conditions and were subjected to mass cytometry (CyTOF) for T cell identification, 16S rRNA gene sequencing for bacterial detection, anaerobic culture based bacterial propagation and in-vitro T cell expansion assay. (B) Flow cytometry plots and gating strategy for fetal and adult Tregs and PD1⁺ Ki67⁺ Tregs. CD25⁺ FOXP3⁺ Tregs were further checked for CD152 expression and more than 80% of Tregs were found to be CD152⁺ as well. (C) Representative flow cytometry plots for fetal and adult Tregs (D) Percentage of PD1⁺ Ki67⁺ Tregs in fetal versus adult tissues and in blood samples (fetal, adult and cord blood) as determined by mass cytometry analysis (input gating strategy and reference plots in Figure S1B). (E) Representative flow cytometry plots for cytotoxic T lymphocytes (CTL: CD8⁺ Tbet⁺ Granzyme B⁺ T cells) in fetal and adult samples. CTL’s are gated on total CD8⁺ T cells. (F) Percentage of cytotoxic T cells (CD8⁺ CTL) in fetal versus adult tissues and in blood samples (fetal, adult and cord blood) as determined by mass cytometry analysis. (G) Input gating strategy and representative flow-cytometry plots for identification of TNF⁺ and IFNγ⁺ cells in CD4⁺ and CD8⁺ T cells (H) Percentage of IFNγ⁺ cells in CD4⁺ (left) and CD8⁺ (right) T cells in fetal and adult tissues and in blood samples (fetal, adult and cord blood) as determined by mass cytometry analysis. (I) Input gating strategy and representative flow-cytometry plots for identification of CD4⁺/CD8⁺ CD45RO⁺ memory and effector memory T cells (Tem). For effector memory T cell phenotype identification, CD45RO⁺ memory T cells were gated and assessed analyzed for TNF and IFNγ expression as shown. (J) Percentage of IFNγ⁺ effector memory T cells (Tem) in CD4⁺ (left) and CD8⁺ (right) memory T cells in fetal and adult tissues and in blood samples (fetal, adult and cord blood) as determined by mass cytometry analysis.

Figure 1

Identification of T cell diversity during the 2^nd trimester of human fetal development (A) Suspension mass cytometry (CyTOF) was performed on PMA/ionomycin-stimulated fetal T cells. The UMAP projection shows all the fetal and adult organs and blood samples used for T cell analysis: fetal gut (7), fetal lung (9), fetal liver (5), fetal skin (6), fetal thymus (8), fetal mLN (3), adult gut (5), adult lung (2), fetal (4), cord (3), and adult blood (4). Numbers in parenthesis represent number of samples being analyzed. (B) UMAP projection of all the blood samples, along with individual projections of each sample type (fetal, cord, and adult blood): fetal blood (4), cord blood (3), adult blood (4). (C) UMAP projection of the T cells from fetal and adult organs, along with individual projections of each organ over the UMAP: fetal gut (7), fetal lung (9), fetal liver (5), fetal skin (6), fetal thymus (8), fetal mLN (3), adult gut (5), and adult lung (2). (D) UMAP projection of fetal versus adult organs (in orange and blue, respectively). CD4⁺ and CD8⁺ T cell clustering is also shown on the same UMAP projection, in both fetal and adult tissues. Tregs in CD4 compartment is also shown based on CD4⁺CD25⁺FOXP3⁺CD152⁺ markers. (E) UMAP feature projection of T cell markers (CD3, CD4, and CD8) and Treg-associated markers (CD25, FOXP3, and CD152) are shown on top and bottom panels, respectively. (F) Percentage of Tregs in fetal versus adult tissues and in blood samples (fetal, adult, and cord blood) as determined by mass cytometry analysis (input gating strategy and reference plots in Figures S1B and S1C). As shown, Tregs are enriched in fetal organs as compared to adult organs and blood. (G) TNF cytokine expression in CD4⁺ (top) and CD8⁺ (bottom) T cells as determined by mass cytometry analysis of fetal and adult tissues (input gating strategy and reference plots in Figure S1G). (H) Percentage of CD4⁺ (top) and CD8⁺ (bottom) CD45RO⁺ memory T cells in fetal and adult tissues and blood samples (input gating strategy and reference plots in Figure S1I). (I) Percentage of TNF⁺ cells in CD4⁺ (top) and CD8⁺ (bottom) CD45RO⁺ effector memory T (Tem) cells in fetal and adult tissues and blood samples. See also Table S1.

Figure S2

Contamination controls and high-throughput 16S rRNA gene sequencing of fetal tissues, related to Figure 2 (A) Schematic to show the contamination controls taken at each step ranging from fetal dissection to tissue processing, bacterial culturing, DNA extraction, and sequencing. The controls taken at each step can be divided into four major categories. Operator control includes the hand swabs of the operators handling the fetal tissues while dissection and processing and while DNA extraction. Environment control denotes control swabs taken from work surfaces like PCR hood, laminar flow, work bench etc. PBS buffer control denotes PBS used at each step of tissue processing and DNA extraction. Reagent control encompasses the molecular reagents used for DNA extraction thereby depicting the contaminations present in extraction kits and associated reagents. (B) Workflow employed for fetal tissue processing, dissection, DNA extraction and sequencing. Contamination controls taken at each step are denoted by swab-handles and color coded according to panel A (left). (C) Principal Coordinate (PC) analysis based on Weighted UniFrac distance to the PBS. PC analysis is this plot is shown in terms of early versus late estimated gestational age (ega). 2^nd trimester fetal samples were divided in two groups, early (< 16 weeks) and late (≥16 weeks) EGA, and color coded accordingly. The percentage of variance explained by the PC’s is indicated next to each axis. Weighted UniFrac distance of fetal tissues to their matched PBS controls, segregated by early (< 16 weeks) and late (≥16 weeks) EGA. Unpaired Mann-Whitney test: ^∗∗∗p < 0.0005. (D) Similar analysis as in panel (B), further broken down into tissue types. Weighted UniFrac distances between all available PBS samples is also displayed as a reference (within PBS bar). Unpaired Mann-Whitney test: ^∗p < 0.05; ^∗∗∗p < 0.0005 (E) CT values for fetal organs plotted to determine the extent of bacterial DNA present in each sample. Lower CT value for a sample represents higher content of DNA being amplified. Here the CT values for each organ plotted and compared with those of those in fetal thymus. As shown, most of the fetal tissues had significantly higher microbial DNA content than fetal thymus. (F) Shannon index boxplot of samples, including reagent, environmental and PBS controls, and fetal tissues. The gray shade represents the cutoff value which is 95th percentile of all PBS samples (~6.324) and is used to perform the Fisher’s exact test in main Figure 2D. (G) Analysis similar to the panel E, except for Chao1 index used as the α-diversity metric. (H) Analysis similar to Figure 2E, except for Bray-Curtis distance used as the β-diversity metric (instead of Weighted UniFrac distance) (I) Unnormalized Weighted UniFrac distances of fetal tissues against their matched PBS controls. Analysis was restricted to fetuses with data available for ≥ 2 tissues. Paired Mann-Whitney test: ^∗p < 0.05; ^∗∗p < 0.005, ^∗∗∗p < 0.0005. This analysis is similar to that in Figure 2F, except that ‘unnormalized’ Weighted UniFrac distances were used for β-diversity metric. (J) Bray Curtis distances of fetal tissues against their matched PBS controls. Analysis was restricted to fetuses with data available for ≥ 2 tissues. Distances were normalized within fetus, by dividing each value by the average Bray-Curtis distance within each fetus. Paired Mann-Whitney test: ^∗p < 0.05; ^∗∗p < 0.005, ^∗∗∗p < 0.0005. (K) Relative abundance plot showing the significantly enriched taxa (log2FC > 2) in fetal Gut as compared to fetal thymus (internal control) and control PBS samples (external control). Each sphere represents differential taxa and the size of the sphere represents the percentage relative abundance of the given genera.

Figure 2

High-throughput 16S rRNA gene sequencing of fetal tissues and bacterial genera identification (A) Principal coordinate (PC) analysis based on weighted UniFrac distance to the PBS. The percentage of variance explained by the PC’s is indicated next to each axis. Each sample depicted by color and numbers in parenthesis indicate samples analyzed. (PERMANOVA R² = 0.17741; p = 0.0009; analyzed using adonis function of vegan (v2.5.7) package in R). (B) Comparisons of bacterial 16S rRNA gene detection using qPCR. Comparison of CT (cycle threshold) values is shown along the y axis for each sample type plotted to determine the extent of bacterial DNA present in each sample. Unpaired Mann-Whitney test: ^∗∗∗∗p < 0.0001. ns, not significant. (C) Comparison of CT values for each fetal organ against its PBS negative control. Unpaired Mann-Whitney test: ^∗p < 0.05; ^∗∗p < 0.005; ^∗∗∗∗p < 0.0001. ns, not significant. (D) Fraction of samples above and below the 95^th percentile of Shannon indices in all PBS controls. A Fisher’s exact test was performed for each sample type against PBS controls (PBS, n = 42; reagents control, n = 10; environment, n = 55; fetal tissues, n = 169). ^∗p < 0.05. ns, not significant. For reference to Shannon indices prediction see Figures S2F and S2G. (E) Weighted UniFrac distances between all PBS controls or between fetal tissues and their matched PBS controls, across two research facilities using two independent microbiome sequencing and analysis protocols: SIgN (Singapore Immunology Network) and WIS (Weizmann Institute of Science). Unpaired Mann-Whitney test: ^∗∗p < 0.005; ^∗∗∗∗p < 0.0001. ns, not significant. (F) Weighted UniFrac distances of fetal tissues against their matched PBS controls. Analysis was restricted to fetuses with data available for ≥2 tissues. Distances were normalized within fetus by dividing each value by the average weighted UniFrac distance within each fetus to determine β-diversity metric. Paired Mann-Whitney test: ^∗p < 0.05; ^∗∗p < 0.005, ^∗∗∗p < 0.0005. (G) Dot-plot showing the distribution of bacterial genera as identified by OTUs across all sample types. Bacterial genera high in PBS and low or equal in samples are depicted as “PBS-enriched taxa” (orange, potential contaminants). Bacterial genera that were absent or very low in PBS and high in fetal tissues are denoted as “fetal-enriched taxa” (green, potential signals). Mean signal intensity of each genus was calculated and plotted using matplotlib (v.3.2.1) and seaborn (v.0.9.0) Python libraries. The samples are arranged in descending order of their signal strength in fetal gut. (H) Relative abundance plot showing the significantly enriched taxa (log2FC >2) in fetal gut as compared to fetal spleen (internal control) and control PBS samples (external control). Each sphere represents differential taxa and the size of the sphere represents the percentage relative abundance of the given genera. See also Tables S1, S2, S3, and S4.

Figure 3

Culture and isolation of live bacteria from freshly harvested fetal samples (A) Schematic representation for anaerobic culturing of fetal bacteria, followed by colony selection, PCR, and Sanger sequencing of the isolates. (B) Plate images of serially diluted cultures of fetus-matched tissue inoculates (D4 donor). Negative controls include PBS samples used at hospitals to clean surgical equipment during harvesting and tissue dissection (PBS-Hospital) and PBS samples used in the laboratory to clean tools used during tissue biopsy (PBS-Lab). (C) Number of unique bacteria species identified from each fetal tissue inoculate. Data is combined from plates incubated under aerobic and anaerobic conditions. The proportion of aerobes, facultative anaerobes, and obligate anaerobes was determined from available literature. (D) Dot-plots representing the bacterial genera isolated and identified from fetal samples and their prevalence across fetuses and tissues. Colored dots next to the names of each genus indicate the oxygen requirement of known species with that genus (color coding same as in C). Multiple dots reflect variability within the genus with respect to their oxygen requirements. The orange star at the right side of the plot represents the bacterial genera that were identified by both pre-culture and post-culture sequencing for the same tissues (for reference see Figure S3D). See also Tables S1, S2, S3, S5, and S6.

Figure S3

Plating specificity, efficiency, and reproducibility of fetus-associated bacteria, related to Figure 3 (A) Plate images of serially diluted cultures of fetus-matched tissue inoculates from a separate donor compared to main Figure 3B (D8), plated under either aerobic (left) or anaerobic conditions (right). All negative controls (plated as in Figure 3) were clean under both plating conditions, while the colony morphology, color and size across different plating conditions had notable differences even within the same fetal tissue inoculate. (B) Plating efficiency (expressed as CFUs/mL) of fetal tissue cultures under aerobic and anaerobic plating conditions across donors and tissues. The area of each pie chart is proportional to the sum of the CFUs/mL values obtained in the two growth conditions. (C) Reproducibility of CFU/mL measurements of selected fetus-matched tissue cultures under aerobic and anaerobic plating conditions, between two independent culture experiments. (D) Dot-plot showing the distribution of fetal genera present across 7 out of 8 donors which were simultaneously sequenced before (and after) culturing by high-throughput 16S rRNA sequencing. This analysis is complementary to that in main Figure 3D where the same fetal samples were sequenced after culturing for 48 hours. Mean signal intensity of each genus was calculated and plotted using matplotlib (v.3.2.1) and seaborn (v.0.9.0) Python libraries. For main Figure 3D, matched PBS were not sequenced post-culturing since there were no detectable colonies in PBS samples. While for pre-culturing, as shown we also sequenced PBS samples, for comparisons.

Figure 4

Visualization of bacterial structures in fetal gut lumen using SEM and RNA-ISH (A) Representative SEM images of 14-weeks EGA fetal mid-gut from three individual fetal samples (n = 3) showing the mucosal area with bacteria (left) and area devoid of bacteria (right) in low and high magnification images. One 10-weeks EGA fetal mid-gut lumen is also shown below with no detectable bacterial structures. Red arrowheads indicate bacteria-like structures within the intestinal lumen. Cocci-like bacteria were seen with an average size of 1 μm in all the three 14-weeks EGA fetal mid-gut lumens. Scale bars, 10 μm for low magnified views and 1 μm for high magnified views. (B) SEM images of cultured bacteria (isolated from 18-weeks EGA fetal gut) attached on the nitrocellulose membrane. Cocci-like bacterial structures were prominently seen with an average size of 1 μm. Scale bar, 1 μm. (C) Representative RNA-in situ hybridization (RNA-ISH) images of fetal gut lumen demonstrating the pan-bacterial probe against bacterial 16S rRNA in green and eukaryotic epithelial cell marker Epcam-specific RNA probe in white. Image on the right shows merged file and the image in inset represents zoomed-in view to visualize the bacterial RNA along-side epithelial boundaries within the lumen of fetal intestine (mid-gut). Scale bar, 40 μm. (D) Bar plot representation of EuBac 16S rRNA and Epcam signal quantification with respect to DAPI signal at 100% saturation (n = 3). The fetal liver and mLN samples were taken as controls. (E) Bar plot representation of EuBac 16S rRNA signal quantification in fetal mid-gut lumen beside fetal liver and mLN samples. Significant detection of bacterial 16S rRNA found in fetal gut lumen (n = 3). Statistical significance between experimental groups was determined by two-tailed, unpaired Student’s t test (^∗∗∗∗p < 0.0001, ^∗∗∗p < 0.001, ^∗∗p < 0.01). See also Figure S4.

Figure S4

Bacterial structure visualization in fetal gut lumen using scanning electron microscopy, related to Figure 4 (A) Representative SEM images of 14 weeks old EGA fetal mid-gut from three individual fetal samples (n = 3) from multiple fields. Red arrowheads indicate bacteria-like morphologies within the intestinal lumen. Cocci like bacteria were embedded in the mesh-like mucus covering the intestinal cells. Red arrows indicate bacteria-like morphologies. (Scale bars = 1 μm) (B) EM images of 14 weeks old fetal foregut (top) and hindgut (bottom) showing the mucosal area with bacteria and mucosal area devoid of bacteria from one fetus (14 weeks EGA). Cocci bacteria (average size of 1 μm) were observed and there were also presence of cocci like bacterial structures found in 14 weeks’ old fetal foregut and hindgut. Red arrows indicate bacteria-like morphologies. (Scale bars; 10 μm for low magnified views and 1 μm for high magnified views) (C) SEM images of cultured bacteria (isolated from 18 weeks old fetal Lung) attached on the nitrocellulose membrane. Cocci-like bacterial structures were prominently seen with an average size of 1 μm. (Scale bar; 1 μm). (D) Representative image of Hematoxylin & eosin stain (H&E staining) of fetal mid-gut cross-section (isolated from 18 weeks old fetal gut), from serial sections used for RNA-ISH imaging. The tissue were embedded in FFPE blocks prior to slide preparation. 200 μm magnification. (E) Representative RNA-ISH imaging (RNAscope) images of fetal liver demonstrating the pan-bacterial probe against bacterial 16S rRNA in green and eukaryotic epithelial cell marker Epcam specific RNA probe in white. Image on the right shows the merged file. There were no significant spots detected for bacterial 16S rRNA in fetal liver. Scale bar, 40 μm. (F) Representative RNA-ISH imaging (RNAscope) images of fetal mesenteric Lymph Nodes (mLN) demonstrating the pan-bacterial probe against bacterial 16S rRNA in green and eukaryotic epithelial cell marker Epcam specific RNA probe in white. Image on the right shows the merged file. There were no significant spots detected for bacterial 16S rRNA in fetal mesenteric lymph nodes. Scale bar, 40 μm.

Figure 5

Syngeneic T cell expansion and memory activation assay for fetal T cells primed with fetal bacteria (A) Schematic representation of the syngeneic T cell expansion assay. DCs isolated from fetal mLN were primed with heat-killed fetal bacteria (for 8 h) and presented to CTV-labeled T cells from the same organ (mLN). At day 6, T cell expansion is assayed by dye dilution using flow cytometry. (B) Sort profile of fetal DCs and T cells for T cell expansion assay. HLA-DR high and lineage negative cells were selected for DC population, which were further enriched for CD14-negative cells. T cells were sorted by negative selection by gating on lineage-negative and HLA-DR-negative cells. Further, to assess the purity of T cells, a small portion was assayed by post-sort for T cell profile and found to be more than 75% pure. (C) Representative flow cytometry plots for T cell expansion (16 weeks EGA fetal mLN). CTV-labeled T cells were incubated with bacteria (fetal Staphylococci or Lactobacilli)-primed DCs for 6 days. For controls, T cells were incubated with unprimed DCs (no bacteria) or CD3/28 Dynabeads as a negative and positive control, respectively. T cell expansion is assayed by measuring dye dilution using flow cytometry. The conditions shown are T cells + DCs alone (gray), T cells + DCs primed with fetal Staphylococci (blue), T cells + DCs primed with fetal Lactobacilli (green), and T cells + CD3/28 Dynabeads (magenta). For comparisons, T cell profile at day 0 is shown at the left. (D) Total T cell number (absolute count) as counted by CountBright beads and manually, post 6 day incubation, plotted for all the conditions (n = 4). Statistical significance between experimental groups was determined by paired t test (^∗∗∗∗p < 0.0001, ^∗∗∗p < 0.001, ^∗∗p < 0.01). (E) CD45RO⁺ memory T cell expansion post 6 day incubation measured by CD45RO expression and dye dilution (CTV^neg) for all the conditions in comparison to T cell + DCs alone (gray). For comparisons, T cell at day zero is shown at the left. (F) Frequency of CTV^neg expanded memory T cells (CD45RO⁺, CTV^neg) plotted for all conditions. (G) CD69⁺ activated memory T cell expansion post 6 day incubation measured by CD69 expression and dye dilution (CTV^neg) for all the conditions, in comparison to T cell + DCs alone (gray). For comparisons, T cell at day 0 is shown at the left. (H) Frequency of expanded CD69⁺ active memory T cells (CTV^neg) plotted for all conditions. Statistical significance between experimental groups was determined by two-tailed, unpaired Student’s t test (^∗∗∗∗p < 0.0001, ^∗∗∗p < 0.001, ^∗∗p < 0.01). See also Figure S5.

Figure S5

Gating strategy for in vitro syngeneic T cell expansion assay, related to Figure 5 (A) Reference input gating strategy for the Figure 5C, showing the live CD45⁺ CD3⁺ T cells gate obtained from fetal mesenteric lymph nodes T cells, post 6 days of incubation with the T cells + DC alone, T cells + DC primed with fetal Staphylococcus, T cells + DC primed with fetal Lactobacillus and T cells + CD3/28 dynabeads respectively (color coded in the figure). (B) As shown, the CD4/CD8 population frequencies post culturing, and CTV staining, gated on total CD3⁺ cells from the above panel a. (C) Another representative flow-cytometry plots for T cell expansion (15 weeks EGA fetal mLN). CellTrace violet (CTV) dye labeled T cells were incubated with bacteria (fetal Staphylococci and Lactobacilli) primed DC for 6 days. For controls, T cells were incubated with unprimed DC (no bacteria) or CD3/28 dynabeads as a negative and positive control, respectively. T cell expansion is assayed by measuring dye dilution using flow cytometry. The conditions shown are T cells + DC alone (gray), T cells + DC primed with fetal staphylococci (blue), T cells + DC primed with fetal lactobacilli (green) and T cells + CD3/28 dynabeads (magenta). For comparisons, T cell profile at day zero is shown at the left. (D) CD45RO⁺ memory T cell expansion post 6 day incubation measured by CD45RO expression and dye dilution (CTV^neg) for all the conditions in comparison to T cell + DC alone (gray). For comparisons, T cell at day zero is shown at the left. (E) CD69⁺ activated memory T cell expansion post 6 day incubation measured by CD69 expression and dye dilution (CTV^neg) for all the conditions, in comparison to T cell + DC alone (gray). For comparisons, T cell at day zero is shown at the left. (F) Levels of TNF released in the culture supernatant of T cells expanded post 6 day exposure to fetal bacteria as determined by drop array-based Luminex assays (n = 4) (G) Levels of IFNγ released in the culture supernatant of T cells expanded post 6 day exposure to fetal bacteria as determined by drop array-based Luminex assays (n = 4)

Figure 6

Fetal T cell expansion and memory activation with PFA-fixed DCs (A) Schematic representation of the syngeneic T cell expansion assay. DCs isolated from fetal mLN were primed with heat-killed fetal bacteria (for 8 h) followed by 1% PFA-induced fixation. Fixed DCs were then presented to CTV-labeled T cells from the same organ (mLN). At day 6, T cell expansion is assayed by dye dilution using flow cytometry. (B) Representative flow cytometry histograms for T cell expansion (17-weeks EGA fetal mLN). CTV-labeled T cells were incubated with bacteria (fetal Staphylococci) primed DCs for 6 days. For controls, T cells were incubated with unprimed DCs (no bacteria) or CD3/28 Dynabeads as a negative and positive control, respectively. T cell expansion is assayed by measuring dye dilution using flow cytometry. The conditions shown are T cells + DCs alone (gray), T cells + fetal Staphylococci primed DCs (blue), T cells + fetal Staphylococci primed PFA-fixed DCs (orange), and T cells + CD3/28 Dynabeads (magenta). (C) Total T cell number (absolute count) as counted by CountBright beads, post 6 day incubation, plotted for all the conditions (n = 4). Statistical significance between experimental groups was determined by paired t test (^∗∗∗∗p < 0.0001, ^∗∗∗p < 0.001, ^∗∗p < 0.01). (D) Frequency of CTV^neg-expanded memory T cells (CD45RO⁺, CTV^neg) plotted for all conditions. (E) Frequency of expanded CD69⁺ active memory T cells (CTV^neg) plotted for all conditions. Statistical significance between experimental groups was determined by two-tailed, unpaired Student’s t test (^∗∗∗∗p < 0.0001, ^∗∗∗p < 0.001, ^∗∗p < 0.01). See also Figure S6.

Figure S6

Flow-cytometry plots for in vitro syngeneic T cell expansion assay (with PFA fixed DC), related to Figure 6 Representative flow cytometry plots of in vitro syngeneic T cell expansion of fetal mLN T cells (treated with bacteria primed DC fixed with 1% PFA) showing the: (A) frequency of CD4 and CD8 T cells ratio, (B) CD45RO⁺ memory T cell plotted against CTV dye to show the dye dilution post 6 days, and (C) CD69⁺ activated memory T cell plotted against CTV dye to show the dye dilution post 6 days. The conditions shown are T cells + fetal staphylococci primed DC (blue), T cells + fetal staphylococci primed PFA fixed DC (orange), T cells + unprimed DC alone (gray), and T cells + CD3/28 dynabeads (magenta). Fetal T cell expansion and memory activation assay (with PTI treated DC) (D) Schematic representation of the syngeneic T cell expansion assay. DC isolated from fetal mLN were primed with heat-killed fetal bacteria (for 8 hours) followed by Protein Transport Inhibitor (PTI) treatment (a combination of Monensin and Brefeldin A). PTI treated DC were then presented to CTV labeled T cells from the same organ (mLN). At day 6, T cell expansion is assayed by dye dilution using flow cytometry. (E) Representative flow-cytometry histograms for T cell expansion (17 weeks EGA fetal mLN). CellTrace Violet (CTV) dye labeled T cells were incubated with bacteria (fetal staphylococci) primed DC for 6 days. For controls, T cells were incubated with unprimed DC (no bacteria) or CD3/28 dynabeads as a negative and positive control, respectively. T cell expansion is assayed by measuring dye dilution using flow cytometry. The conditions shown are T cells + DC alone (gray), T cells + fetal staphylococci primed DC (blue), T cells + fetal staphylococci primed PTI treated DC (orange) and T cells + CD3/28 dynabeads (magenta). (F) Total T cell number (absolute count) as counted by CountBright beads, post 6 day incubation, plotted for all the conditions (n = 4). Statistical significance between experimental groups was determined by paired t test (^∗∗∗∗p < 0.0001, ^∗∗∗p < 0.001, ^∗∗p < 0.01) (G) Frequency of CTV^neg expanded memory T cells (CD45RO⁺, CTV^neg) plotted for all conditions. (H) Frequency of expanded CD69⁺ active memory T cells (CTV^neg) plotted for all conditions. Statistical significance between experimental groups was determined by two-tailed, unpaired Student’s t test (^∗∗∗∗p < 0.0001, ^∗∗∗p < 0.001, ^∗∗p < 0.01)