Dataset DOI: 10.5061/dryad.2jm63xt21

Description of the data and file structure

Associated Figure Legends

Some data is organized into folders by figure where multiple datasets from related experiments were presented together. In order to understand and analyze the data, here we provide brief descriptions of the figures from the related manuscript.

Figure 1. Bacterial tissue burdens are driven by extensive replication of small numbers of initial founders. (A) Bacterial burdens in Peyer’s patch (PP), mesenteric lymph node (MLN), spleen, and liver tissues at day 5 post-infection. (B) Bacterial burdens in small intestinal pyogranuloma (PG+) and adjacent non-granulomatous (PG-) tissue isolated day 5 post-infection. Each circle represents the total CFU of 5-10 pooled punch biopsies divided by number of pooled punch biopsies from one mouse. (C) Founding population (Ns) in indicated tissues at day 5 post-infection. (D) Founding population (Ns) in small intestinal PG+ and PG- tissue isolated day 5 post-infection. Each circle represents the total Ns of 5-10 pooled punch biopsies divided by number of pooled punch biopsies from one mouse. (E) Y. ptb bacterial burden in flushed small intestinal contents (IC) at day 5 post-infection. (F) Founding population (Ns) in flushed IC at day 5 post-infection. Samples labeled liver are showing the median lobe as representative of all liver lobes. For all data, circles represent one mouse, bars are mean ± standard error of the mean (SEM), and data are pooled from three experiments. Statistical significance was only calculated on B,D using Mann-Whitney tests and **** p<0.0001.

Figure 2. Peyer’s patches and intestinal pyogranulomas contain distinct bacterial populations. (A) Bacterial burdens and (B) founding population (Ns) in small intestinal PP, PG+, and PG- tissue isolated day 5 post-infection. Each circle represents one biopsy.  Bars represent (A) mean ± SEM or (B) median ± 95% CI. Pooled data from four mice from one experiment. (C) Frequency of barcodes per biopsy for mouse 1. Each bar represents one biopsy, and each color represents one barcode. (D) Similarity between Y. ptb populations in each biopsy as assessed by chord distance (CD) for mouse 1, depicted as (TOP) samples ordered by location along small intestine or (BOTTOM) samples ordered by biopsy type. Red boxes highlight two examples of sharing dynamics where PP share with adjacent PG-, but not with adjacent PG+. (n=4, one experiment, equivalent data of mice 2-4 is depicted in Fig S2E-J). (E) Histograms showing the frequency of binned CD values for each comparison: (E) PP vs PP (blue), PP vs PG- (cyan), and PP vs PG+(magenta); (F) PG+ vs PG+ (red) and PG+ vs PG- (yellow); and (G) PG- vs PG-(green). For E-G, color-coding of comparisons is diagrammed by triple Venn diagram with specific comparisons highlighted in the figure key, bin width = 0.1, and the center of bins are shown on x-axis. Data is cumulative from 4 mice, one experiment.

Figure 3. Bacterial populations in the spleen and liver are distinct from those in the intestine. (A) Similarity between Y. ptb populations in each sample as assessed by chord distance (CD) for one representative mouse. (B) CD between spleen and indicated samples. (C-D) Prevalence of barcodes from (C) the spleen in other organs (FRDorgan-spleen) or (D) from various organs in the spleen (FRDspleen-organ). (E) Fraction of barcodes from the spleen that are not found in other sites outside of the spleen or liver lobes. All data collected at day 5 post-infection, each circle represents one mouse, n=10, pooled data from three independent experiments, where bars are mean ± SEM and violin plots indicate median ± interquartile range. Statistical significance was calculated using one way ANOVA with post-hoc Dunn’s Multiple Comparison Test, where ns = not significant and * p<0.05.

Figure 4. Blood is a conduit for systemic Y. ptb populations. (A) CD between the MLN and indicated samples. (B) Bacterial burdens and (C) Founding population (Ns) in blood; blood samples with no CFU are represented as 0.1 CFU. (D) Log-log regression with Pearson’s correlation of fit between Ns and CFU measured in blood. (C-D) Each circle represents one mouse with detectible burden of *Y. ptb *in blood. (E) CD between the blood and indicated samples. Samples labeled liver are showing CD comparisons to the median lobe as representative of all liver lobes. All data were collected at day 5 post-infection, from three-four independent experiments (except B, which has pooled data from seven experiments), where each circle represent one mouse unless indicated otherwise and bars are mean ± SEM.

Fig. 5. TNFR1 contributes to Y. ptb infection bottleneck and MLN colonization. (A) Bacterial burdens and (B) founding population (Ns) in small intestinal PP, PG+, and PG- biopsies at day 5 post-infection, where each symbol represents the total of 5-10 pooled punch biopsies divided by the number of biopsies from one mouse. (C) Bacterial burdens and (D) founding population (Ns) in whole blood. (E) Bacterial burdens and (F) founding populations (Ns) in systemic tissues: mesenteric lymph nodes (MLN), spleen, and liver (median lobe). (G-I) Chord Distance (CD) between (G) PG+, (H) spleen, or (I) MLN and the indicated systemic (MLN, spleen, median liver lobe, blood) and intestinal (PP and PG-) tissues. All data collected on day 5 post-infection and are pooled from three independent experiments. (J-K) H&E-stained paraffin-embedded mesenteric lymph node sections from naïve and Y.ptb-infected WT and Tnfr1−/− mice at day 5 after infection were used for (J) histopathological scoring and (K) imaging. (J) Each mouse was scored between 0 and 4 (minimal to extensive) for the metrics shown in Fig. S4E-G. Scores for each mouse were added together to obtain the aggregate score shown. (K) Naïve mesenteric lymph nodes from WT mice (LEFT), but not Tnfr1−/− mice (RIGHT), have lymphoid follicles in the cortex (marked by arrowheads). White scale bars = 100 μm. Infected mesenteric lymph nodes (BOTTOM) show pyogranulomatous lesions in WT mice (marked by asterisks) and necrosis in Tnfr1−/− mice (marked by arrows). Black scale bars = 200 μm. Histological data are from two independent experiments with representative images shown. Each symbol represents one mouse unless otherwise indicated, bars are mean ± SEM. Statistical significance was calculated using multiple Mann-Whitney tests and ns = not significant, * p<0.05, ** p<0.01, *** p<0.001. 

Figure S1. Validation of the Y. ptb barcode library and impact of antibiotic pre-treatment on barcoded Y.ptb dissemination. (A) The ability to determine the size of the founding population from STAMP libraries was validated in culture by comparing the number of plated colonies (CFU; known number of founders) to the size of the founding population following STAMPR analysis of those plated populations (Ns) for 3 cultures across 9 doses. (B) Bacterial burdens and (C) founding population (Ns) in stool isolated 1-, 6-, or 12-hours post-infection. Pooled data from four mice from one experiment. (D) Graphical representation of the experimental design for antibiotic (ABX) pretreatment: C57BL/6J mice were given either a cocktail of ABX or control sterile lab drinking water ad libitum for 7 days and returned to normal drinking water one day prior to infection. At day 0, mice were inoculated with 2x108 CFU Y. ptb library via oral gavage. Mice were euthanized 5 days following the inoculation and Y. ptb populations were assessed. (E) Bacterial burden and (F) founding population (Ns) in flushed small intestinal contents. (G) Bacterial burdens and (H) founding population (Ns) in small intestinal PP, PG+, and PG- biopsies. Each circle represents the total CFU of 5-10 pooled punch biopsies divided by number of pooled punch biopsies from one mouse. (I) Bacterial burden and (J) founding population (Ns) in MLN, spleen, and liver. Unless otherwise indicated, each circle represents one mouse, and all graphs are pooled data from three independent experiments.  For bar graphs, bars represent mean ± SEM, and statistical significance was determined using Mann-Whitney tests: ns = not significant, * p<0.05, ** p<0.01, **** p<0.0001.

Figure S2. Intestinal biopsy maps, analysis of barcodes in mouse intestine, and pooled comparisons of chord distance. (A-D) Map showing a representative layout of individually harvested and processed tissue biopsies from the small intestine, cecum, and colon of mice 1-4 on day 5 post-infection. (E-G) Frequency of barcodes per biopsy for mice 2-4 (mouse 1 is depicted in Fig. 1C), where each bar represents one biopsy, and each color represents one barcode. (H-J) Similarity between Y. ptb populations in each biopsy as assessed by chord distance (CD) for mice 2-4 (mouse 1 is depicted in Fig. 1D), where (LEFT) data is organized by biopsy type and (RIGHT) data is ordered by location along gastrointestinal tract. (K) CD between biopsies within the indicated small intestinal region. Each circle represents one comparison between biopsies within the indicated region for that mouse. (L) Proportion of PG+ biopsies that share or do not share with PP or PG- biopsies in the same mouse. Sharing was determined by presence of single PG+ founder in other biopsies or, for PG+ containing more than one founder, the most abundant barcode was used to determine sharing. (M-N) CD between the indicated biopsy types. Each circle represents one comparison between the (M) same or (N) different biopsy types as indicated, where biopsies are matched within mouse. Unless otherwise indicated, for all data bars represent mean ± SEM, (pooled for n=4 mice, one experiment collected at 5 days post-infection). Statistical significance was determined using one-way ANOVA with post-hoc Dunn’s Multiple Comparison Test, where ns = not significant and **** p<0.0001.

Figure S3. Populations of Y. ptb in spleen and liver are similar to each other with one highly abundant clone. (A) CD between the spleen and gastrointestinal tract organ tissues. (B) CD between the spleen and individual liver lobes. (C) Frequency of barcodes per tissue for one representative mouse. Each bar represents one sample, and each color represents one barcode. (D) Frequency of most abundant barcode per sample. All samples collected at 5 days post-infection. Data are pooled from three to five experiments, each circle represents one mouse, bars are mean ± SEM unless otherwise indicated.

Figure S4. TNFR1 signaling affects bacterial colony containment in mesenteric lymph nodes but not clone expansion or lymphadenitis. (A) Bacterial burdens and (B) founding populations (Ns) in flushed small intestinal contents at day 5 post-infection. (C-F) **Bacterial burden per founder, as determined by CFU/Ns, in (C) flushed small intestinal contents, (D) small intestinal PP, PG+, and PG- biopsies, (E) whole blood, and (F) indicated systemic tissues at day 5 post-infection. Liver samples show data from the median lobe as representative of all liver lobes. Pooled data from three independent experiments. (G-I) H&E-stained paraffin-embedded mesenteric lymph node sections from naïve and Y.ptb-infected WT and Tnfr1−/− mice at day 5 after infection were scored between 0 and 4 (minimal to extensive) for the metrics shown. Histological data are from two independent experiments. Each symbol represents one mouse, and bars are mean ± SEM. Statistical significance was determined using multiple Mann-Whitney tests and ns = not significant, * p<0.05, ** p<0.01.

Files and variables

File: Frequencyplotanalysis.zip

Description: This file contains .xlsx files with processed data from multiple FrequenciesWithoutNoise.csv files. We have left this data as .xlsx to keep related data together in multiple tabs and maintain formatting that makes the data more readable. However, this formatting is not required for accessibility or reanalysis. This processed data was used to describe and depict the frequency of barcodes in samples displayed in Fig. 2, Fig. 3, Fig. S2, and Fig. S3.

File: RawCFU.zip

Description: This folder contains raw counts of bacterial colonies used to determine the CFU and CFU/g for harvested mouse tissues. Files are organized in subfolders by figure. We have left this data as .xlsx to maintain text formatting that makes the data more readable and to provide access to formulas used in calculating CFU and CFU/g. This formatting is not required for accessibility or reanalysis. Each file lists samples in rows containing (left to right): experiment name, sample (key provided in file), group (genotype or treatment group), colony counts from triplicate spot plating, the dilution that was counted, the tube weights pre-and post-tissue collection, calculated CFU and CFU/g, the number of colonies on the plate collected for sequencing (if countable). Blood samples also have the volume of blood collected listed. Additionally, the number of intestinal biopsies of Peyer’s patches, intestinal pyogranulomas, and adjacent nongranulomatous tissue is listed for each mouse. These files were used to generate CFU/g or CFU/biopsy plots and CFU tables for STAMPR input.

File: Frequencies.zip

Description: This file contains .csv files with the barcode frequencies as determined by analysis of raw sequencing reads and removal of barcode hopping. These files were used as input into the STAMPR analysis code provided and as previously described (Holmes et al., 2025; Hullahalli et al., 2021). Each sequencing batch of 64 samples was labeled in order with “spXX”. This batch ID identifies all downstream files generated through STAMPR analysis.

File: STAMPRcode.zip

Description: This file contains R scripts used for STAMPR analysis after index hopping correction. getFP.R determines founding population estimates using the frequency table and the CFU table as inputs. MajorityDistance.R calculates genetic distance metrics referred to as “Chord distance” in our work.

File: STAMPRinputs.zip

Description: This folder contains .csv files input into the provided STAMPR analysis pipeline. Each analysis has a CFU table that was fed into the getFP.R script to determine founding population estimates. Subsequently, GD_metadata files were used to calculate genetic distance metrics using the MajorityDistance R script. Samples can be identified by the BioSample ID that lists “experiment#_Sample” (e.g., “SP051_S1”). Files are organized into subfolders based on the relevant figure.

File: MappedBiopsies.zip

Description: This file contains original images of hand-drawn maps made when harvesting individual intestinal biopsies in Fig. 2 and Fig. S2.

File: STAMPRoutputs.zip

Description: This file contains .csv files generated by the STAMPR analysis pipeline. Each file is identified by the sequencing batch ID “spXX”. Occasionally, samples were sequenced in different batches. Thus, some files have been modified to include samples from multiple sequencing batches. In these cases, the experiment number “SP0XX” is used to identify the samples in the file. Analysis of the files in Frequencies.zip generated each NsNb.csv, FrequenciesWithoutNoise.csv, GD.csv, RD.csv, and FRD.csv file.

File: Histology.zip

Description: This file contains raw data relevant to MLN histology shown in Fig 5J,K and Fig S4G-I. The two .xlsx files are from two independent experiments where blinded analysis of histological sections was done by the comparative pathology core. We have left this data as .xlsx to keep the full analysis together in multiple tabs to aid in comprehension and any future reanalysis. The tab labeled “Heatmap of Lesions” has scores plotted in Fig 5J and Fig S4G-I. Formatting, as provided by the comparative pathology core, was left on for clarity and visual assessment. However, this formatting is not required for accessibility or reanalysis. Unmodified .tif images of H&E-stained histological sections are present for each of the representative sections shown in Fig 5K.

Code/software

No Code or Software other than Microsoft Excel is needed to view this data. To reanalyze, R scripts are provided for STAMPR analysis.

Access information

All raw data were generated from new experiments described in the manuscript associated with this dataset.

Please disregard any other highlighting or colors, as they were used for personal reference.

 Abbreviations: WT = wild-type, ABX = antibiotics-treated, S = spleen, L1 = left liver lobe, L2 = median liver lobe,  L3 = right liver lobe, L4 = caudate liver lobe, M = mesenteric lymph node, B = blood, P = Peyer’s patches, + = intestinal pyogranuloma biopsies, - = intestinal biopsies that do not contain pyogranulomas, IC = intestinal contents, CFU = colony forming units, STAMPR = Sequence tag-based analysis of microbial populations in R