Prenatal Exposure to Bacterial Extracellular Vesicles Influences Fetal Gut Immunity and Immune Programming

84 85

The fetal immune system undergoes pivotal development during gestation, 86 preparing for postnatal antigenic challenges. Bacterial extracellular vesicles (bEVs), bioactive 87 particles shed by bacteria, are emerging as modulators of host immunity. However, their role in 88 shaping fetal intestinal immune development remains largely unexplored. 89

This study aimed to investigate the effects of bEV exposure on lymphoid and myeloid 90 populations in the fetal murine gut, focusing on their role in priming intestinal immunity, promoting 91 differentiation, and modulating immune cell phenotypes in both normal and germ -free (GF) 92 environments. 93

We used a murine model to evaluate the immune-modulating effects of 94 bEVs during fetal development. bEVs were isolated from bacterial cultures and introduced into 95 the amniotic sac of embryonic day 15.5 (E15.5) fetuses through intra-amniotic injection. Fetal and 96 neonatal mice were either raised under conventional conditions (normal environment, NE) or in 97 germ-free (GF) environments to assess microbiota -dependent effects. Immune profiling of fetal 98 (E17) and postnatal (4 weeks) gut tissues was performed using high-dimensional mass cytometry 99 (CyTOF) in both conventionally housed and germ -free (GF) mice. Clustering and differential 100 expression analyses identified lymphoid and myeloid subpopulations, including progenitors, 101 antigen-presenting cells, and intestinal stem cells (ISCs). secondary immune challenge (LPS or 102 TSST-1) was conducted in postnatal bEV-primed mice to assess immune memory responses. 103

bEV exposure significantly increased the prevalence of CD45 - CD24+ CD44+ ISCs, 104 promoting intestinal renewal and defense via differentiation into Paneth and tuft cells. These ISCs 105 exhibited potential antigen -presenting capabilities through MHC expression. CD45+ lymphoid 106 progenitors were upregulated, highlighting their role in early differentiation pathways. Myeloid 107 progenitors, particularly monocyte-dendritic progenitor subsets, showed a bias toward antigen -108 presenting phenotypes.Germ -free models revealed heightened sensitivity to bEVs, with 109 pronounced activation of progenitors and a reduction in exhaustion markers. Interestingly, 110 macrophage and neutrophil populations displayed dose -dependent modulation, with low bEV 111 concentrations promoting their expansion and higher doses leading to reduced incidence. Our 112 findings suggest that bEVs act as immune priming agents in the fetal gut, promoting progenitor 113 expansion and differentiation while preparing the intestine for postnatal challenges. Differences 114 in responses between NE and GF models emphasize the importance of environmental influences, 115 including microbiota, on bEV-mediated immune modulation. 116

bEVs play a pivotal role in shaping fetal intestinal immunity by priming lymphoid and 117 myeloid progenitors and enhancing ISC function. These results open potential avenues for 118 leveraging bEVs in immunomodulation and vaccine strategies. Future studies should explore the 119 functional responses of bEV-primed cells and their translational relevance in humans. 120 121

bacterial extracellular vesicles, fetal immunity, prenatal priming, CyTOF, immune 122 development, dendritic cells, lymphoid progenitors 123 124 125 126 127 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 128 1. Introduction 129 The regulatory factors and mechanisms that control immune system development and function 130 are highly complex. Human immune system development begins in utero. Programming of the 131 fetal immune system occurs at discrete stages of gestation and is known to be influenced by both 132 maternal and fetal signaling. However, this is an understudied area primarily due to the difficulty 133 of carefully accessing the human fetus during pregnancy. Current research on reproductive 134 immunology has focused on the tolerogenic mechanisms at the feto-placental-maternal interface 135 that allow pregnancy maintenance1-12. Protective immunity is the arm of reproductive immunology 136 concerned with how progressive education of the fetal immune system programs its maturation 137 in utero13-16. Protective mechanisms, like early training and education of the fetal immune system, 138 are of interest as this system must produce existential responses to rapid microbial and other 139 antigenic challenges ex utero17-19. Consequently, in utero education of the immune system and 140 its imprinting in response to exposures during pregnancy (e.g., maternal vaccination) is 141 fundamental to priming of the emerging human immune system20-25. Among exposures, maternal 142 inflammation during pregnancy has been shown to program long -term immune output from 143 hematopoietic progenitors and subsequently impact offspring immune function in mouse 144 models26; however, the precise mechanisms by which this programming occurs are unknown. 145 Extracellular vesicles (EVs; exosomes , 30- 200nm size particles) are one of the most 146 dynamic products that cells produce throughout their lifespan27. EVs were once considered 147 carriers of cellular metabolic waste; however, emerging data demonstrate that their roles in 148 various cellular biological functions are vast and mostly unknown 28. During pregnancy, feto -149 maternal paracrine communication by EVs is one of the fundamental bases of maintaining 150 homeostasis. EVs also signal parturition at term and preterm pregnancies 29,30. Recently, we 151 reported the discovery of bacterial extracellular vesicles (bEVs) released from commensal 152 microbes of the host (various maternal body sites) and their presence in the placenta. bEVs are 153 also reported in the amniotic fluid of pregnant women during normal pregnancy 31. As reported, 154 bEVs can elicit an immune response in host macrophages at supraphysiologic doses. While bEVs 155 in the placenta are not proinflammatory, they do generate an immune response in the fetus. 156 It is postulated that exosomes can interact with various immune cells in the body. In innate 157 immunity, nucleic acids can be packaged inside exosomes derived from T-cells, virus-infected 158 cells, and cancer cells to be subsequently targeted by dendritic cells for presentation and 159 interferon release32-35. Macrophages are also able to be either polarized or inactivated by tumor-160 associated exosomes via the delivery of non -coding RNAs and proteins 36-39. Neutrophil 161 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint extracellular trap release and interleukin production from neutrophils have also been observed to 162 be abrogated via exposure to exosomes40. In adaptive immunity, plasma exosomes have been 163 shown to activate T-cells, possibly through packaged MHCs inside exosomes of dendritic 164 cells41,42, while also transform ing the T -cells into a suppressive phenotype via interactions with 165 multiple tumor-derived exosomes43. 166 During pregnancy, EVs coming from various body sources can be deposited in the 167 placenta and traverse the placenta to allow exposure to the growing fetus and affect the 168 physiology of fetal tissues29,30,44. Little is known about the propagation and effect of bEVs on fetal 169 tissues. In a cellular environment where the majority of immune cells are undergoing training and 170 development, bEV exposure might alter the differentiation of specific precursor cell populations 171 toward certain phenotypes. Here, we reasoned that training may not necessarily be induced via 172 inflammation, as overwhelming inflammation is likely to be deleterious for the pregnancy; instead, 173 a fine-tuned balance of stimulation would be a more favorable environment for appropriate fetal 174 immune development. We specifically tested the hypothesis that bEV from maternal body sites 175 systemically deposited in the placenta can reach the fetus , causing in utero immune education 176 and rendering the neonatal immune system programmed to recognize these microbes as self. 177 In this study, we used intraamniotic injections of bEVs in pregnant mice and high -178 dimensional mass cytometry to investigate how bEVs affect the development of lymphoid and 179 myeloid immune cell populations in fetal and postnatal gut tissues. The results reveal that bEV 180 exposure induces a dose -dependent upregulation of diverse immune —including progenitors, 181 activated lymphocytes, and dendritic -like cells ,with distinct patterns observed between mice 182 raised in a conventional (normal-) environment and those raised in a reduced-germ load 183 environment as well as between embryonic and young mice. Overall, our findings suggest that 184 bEVs can prime and educate the developing immune system, offering promising insights into 185 potential novel early-life immunization strategies based on extracellular vesicles. 186 187 2. Materials and Methods: 188 189 2.1 Institutional ethics approval 190 191 All Animal procedures were followed in accordance with the Institutional Animal Care and Use 192 Committee (IACUC) at the University of Texas Medical Branch, Galveston, under approved 193 protocol number 041107F. 194 195 196 197 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 2.2 Husbandry and mating 198 Normal-environment (NE) CD-1 mice were maintained at 4–5 mice/cage in standard cages in the 199 UTMB Animal Research Center housing facility under a strict 12-h light/dark cycle. Normal chow 200 (Lab Diets, #5001) and osmotically-sterilized drinking water were provided ad libitum. 201 Germ-free (GF) C57BL/6 mice were kindly provided by Dr. Richard Pyles’ laboratory and 202 maintained in a separate building under a reduced-germ load environment. Autoclaved chow and 203 water were provided to these mice ad libitum. Nulligravid mice were mated with male mice (n=3 204 for breeding) and, upon confirmation of a vaginal plug (designated as embryonic day 1, E1), the 205 bred dams were placed in individual cages until E15. 206 207 2.3 Protocol for the isolation of bEVs from the E. coli and human placenta 208 bEVs were isolated following the established protocol from Menon et al.31 Escherichia (E.) coli 209 ATCC 12014 strain O55:K59(B5):H, obtained from Remel Laboratory of Thermo Fisher (Thermo 210 Fisher Scientific, Remel Products, Lenexa, KS, United States, Lot # 496291) , was cultured in 211 nutrient broth (BD Biosciences) with stocks stored in 20% glycerol at −80°C . Logarithmic phase 212 cultures were processed using the ExoBacteria ™ OMV Isolation Kit (Cat# EXOBAC100A -1, 213 System Biosciences, Palo Alto, CA, United States) to isolate bEVs. Placental specimens for 214 human placenta bEV isolation were deidentified and considered as discarded human specimens 215 that do not require institutional review board (IRB) approval. Placental specimens were collected 216 from John Sealy Hospital at the University of Texas Medical Branch at Galveston, Texas, USA, 217 in accordance with the relevant guidelines and regulations of approved protocols for various 218 studies (UTMB 11-251; University of Texas Medical Branch at Galveston). Specimens were first 219 processed according to the established protocol from Vidal et al.45 Briefly, tissues were minced 220 into uniform 2 mm³ pieces, digested in endotoxin-free PBS containing collagenase D (2 mg/mL) 221 and DNAse I (40 U/mL) at 37°C with constant rotation, and then filtered and centrifuged to obtain 222 a crude extract, which was concentrated using 10 -kDa Amicon tubes. This concentrated extract 223 was layered into an iodixanol density gradient (comprising 50%, 40%, 20%, and 10% iodixanol 224 solutions topped with PBS) and subjected to ultracentrifugation at 100,000×g for 18 hours at 4°C, 225 after which the 8th and 9th fractions—containing the placental bEVs—were carefully collected on 226 ice. All bEVs were stored at -80°C prior to use. 227 228 2.4 Induction of bEV exposure in pregnant mice 229 At E15, the pregnant mice were brought to our Animal Surgery Roo m. Each mouse was 230 anesthetized with 1 -2% isoflurane in oxygen delivered through a nosecone using a controlled -231 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint delivery anesthetic machine. Once unconsciousness was determined, individual mice were 232 placed supine in a sterile surgical field. Limb restraints were put into place. A midline abdominal 233 incision was performed, and the uterine horns were exposed. A single intraamniotic injection of 234 each treatment [10 µL of one of the following: E. coli-derived bEVs (9 x 106 particles) (EC-bEV), 235 “low-dose” human placental bEVs (9 x 106 particles), “high-dose” human placental bEVs (9 x 108 236 particles), or endotoxin-free PBS] was administered per amniotic sac. After injections had been 237 completed, the uterine horns were repositioned to their original location, the abdominal walls were 238 positioned, and the defect was closed using sterile surgical staples. A single subcutaneous 239 injection of buprenorphine (0.05 mg/kg/dose) was given to each mouse. The pregnant mice were 240 transferred back to clean individual cages. Heat lamps were employed to warm the mice during 241 recovery. Monitoring was performed every 2 hours, and any signs of weakness were accounted 242 for. Mice were transferred to the Animal Satellite Room to recover. The experimental design is 243 shown in Figure 1. 244 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 245 Figure 1. Experimental design for the study. (A) Flow for the first part of the study, 246 exploring cell populations during fetal life. (B) Flow for the second part of the study, 247 exploring early responses to bacterial extracellular vesicle (bEV) -primed mice. PBS, 248 phosphate-buffered saline. hpbEV, human placenta bEV. LPS, lipopolysaccharide. 249 TSST-1, toxic syndrome shock toxin-1. 250 251 252 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 2.4.1 Tissue harvesting 253 On E17, the treated pregnant mice were transferred again to our Animal Surgery Room. The mice 254 were euthanized via carbon dioxide asphyxiation. Once death was confirmed, mice were placed 255 supine in a sterile surgical field. A midline abdominal incision was performed, and the uterine 256 horns were exposed. Incisions were made per amniotic sac, and fetuses were harvested. Each 257 pup’s intestines were individually obtained and stored in RPMI media with 10% FBS and 1% 258 penicillin-streptomycin for cell isolation. 259 260 2.5 bEV priming and postnatal immune challenge 261 NE CD-1 mice were assigned to three groups. Each of the groups were given different “priming” 262 exposures – the first batch was given an intraamniotic injection of endotoxin-free PBS, the second 263 batch was given EC -bEV (9 x 10 5 particles), and the third batch was given LD -bEV (9 x 10 5 264 particles) (Figure 1). The doses for this priming challenge were predetermined by exposing GF 265 mice to various doses of EC-bEV, and the lowest dose, which did not cause mortality, was chosen. 266 The administration method was similar to the aforementioned surgical technique. The animals 267 were allowed to deliver spontaneously, and the pups were kept with the mother until four weeks 268 of age. At this point, three young mice from each group were then separated accordingly and 269 given separate treatments via oral gavage – one group was given endotoxin-free PBS, one group 270 was given lipopolysaccharide (50 mg/kg), and one group was given toxin shock syndrome toxin-271 1 (TSST-1) (0.01 ug/kg). The mice were then allowed to recover, and, four hours later, they were 272 individually euthanized. Embryonic pup intestines were harvested according to the protocol as 273 mentioned earlier. All isolated tissues were kept at -80°C until cell isolation was performed. 274 275 2.5.1 Isolation of cells from tissues 276 Upon collection of all necessary tissues, 500 µL of accutase was added , and the tissues were 277 homogenized. The homogenates were then incubated at 37°C with shaking for 1 hr. Another 500 278 µL of RPMI was added to the homogenates, which were strained using a 70 µm cell strainer. The 279 strained suspension was washed with repeated additions of RPMI and centrifugation at 1500 rpm 280 for 5 min at 20 °C. After the washing, the supernatant was removed, and the pellet was 281 resuspended in 1 mL of RBC lysis buffer. After incubation for 10 mins, the suspension was 282 neutralized with 9 mL of DMEM-F12/10% FBS and spun down at 1500 rpm for 5 min at 20°C. The 283 supernatant was removed again, and 500 µL of cell freezing media was added to allow for storage 284 at -80°C until staining. 285 286 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 2.6 High-dimensional single -cell profiling of feto -maternal tissues by mass cytometry 287 (CyTOF) 288 2.6.1 Antibodies 289 The CyTOF panel was designed based on available literature for known lymphoid and myeloid 290 populations. A summary of antibodies used for each panel can be found in Table 1. The antibodies 291 were sourced from the MD Anderson Cancer Research Center Flow core facilities. (MDACC, 292 Texas, Houston), or custom conjugated using the Maxpar antibody conjugation kit (Fluidigm, 293 Markham, ON, Canada) following the manufacturer’s protocol. After being labeled with their 294 corresponding metal conjugate, the percentage yield was determined by measuring their 295 absorbance at 280 nm using a Nanodrop 2000 spectrophotometer (Thermo Scientific, 296 Wilmington, DE). Antibodies were diluted using Candor phosphate-buffered saline (PBS) antibody 297 stabilization solution (Candor Bioscience GmbH, Wangen, Germany) to 0.3 mg/mL and then 298 stored at 4°C. A summary of the marker antibodies is shown in Table 1. 299 300 Table 1. List of conjugated antibodies used for both lymphoid and myeloid panels, 301 respectively. 302 LYMPHOID MYELOID Target Label Target Label CD69 156Gd Target label CD27 148Nd CD11c 209Bi CD183, CXCR3 143Nd I-A/I-E, MHC-II 115ln CD62L 164Dy Ly-6G/C, Gr-1 141Pr CD44 153Eu CD11b 143Nd CD4(Ms) 115In CD14 156Gd CD184, CXCR4 159Tb FceR1a 144Nd CD197, CCR7 155Gd CD117, c-kit 166Er CD24 169Tm XCR1, GPR5, CCXCR1 168Er CD279, PD1 165Ho CD206, MMR 169Tm CD25 167Er F4/80 174Yb NK1.1, CD161b/c, Ly- 55 170Er CD86 172Yb Granzyme B 173Yb CD8a 146Nd .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint CD38 175Lu Siglec-F 172Yb CD8a 168Er Ly-6C 150Nd CD45(Ms) 89Y Arg-1, ARG1 164Dy TCRgd 160Gd CD103 147Sm CD19 149Sm Foxp3 158Gd GATA3 145Nd RORg(t) 163Dy T-bet 161Dy 303 2.6.2 Antibody staining 304 Single-cell suspension samples were resuspended in Maxpar staining buffer for 10 min at room 305 temperature on a shaker to block F c receptors. Cells were mixed with a cocktail of metal -306 conjugated surface marker antibodies (Table 1), yielding 500 -μL final reaction volumes, and 307 stained at room temperature for 30 min on a shaker. Following staining, cells were washed twice 308 with PBS with 0.5% BSA and 0.02% NaN3. Next, cells were permeabilized with 4°C methanol for 309 10 min at 4 °C. Cells were then washed twice in PBS with 0.5% BSA and 0.02% NaN3 to remove 310 the remaining methanol. They were stained with intracellular antibodies in 500 μL buffer for 30 min 311 at room temperature on a shaker. Samples were then washed twice in PBS with 0.5% BSA and 312 0.02% NaN 3. Cells were incubated overnight at 4°C with 1 mL of 1:4,000 191/193Ir DNA 313 intercalator (Standard BioTools, Inc., Markham, ON) diluted in Maxpar fix/perm overnight. The 314 following day, cells were washed once with PBS with 0.5% BSA and 0.02% NaN 3 and then two 315 times with double-deionized water. 316 317 2.6.3 Mass cytometry 318 Prior to analysis, the stained and intercalated cell pellet was resuspended in ddH 2O containing 319 polystyrene normalization beads containing lanthanum -139, praseodymium -141, terbium -159, 320 thulium-169, and lutetium-175 as described previously46. Stained cells were analyzed on a CyTOF 321 2 (Standard BioTools Inc ., Markham, ON) outfitted with a Super Sampler sample introduction 322 system (Victorian Airship & Scientific Apparatus, Alamo, CA) at an event rate of 200-to-300 cells 323 per second. All mass cytometry files were normalized using the mass cytometry data 324 normalization algorithm freely available for download from https://github.com/nolanlab/bead-325 normalization. 326 327 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 2.6.4 Data Analysis 328 CyTOF data sets were first manually gated using Standard BioTools/Fluidigm clean -up 329 procedure, including Gaussian discrimination (Markham, ON) in FlowJo V10 (FlowJo LLC). Then, 330 each sample was given a unique Sample ID, then all samples were concatenated into a single 331 .fcs file. This concatenated file was further analyzed by t-distributed stochastic neighbor 332 embedding (t-SNE) in FlowJo V10, using equal numbers of cells from individual treatment groups. 333 The analyses were done for both lymphoid and myeloid panels. tSNE was performed using the 334 following settings: Iterations, 3000; Perplexity, 50; Eta (learning rate), 28000. Heatmaps of marker 335 expression were generated using the Color Map Axis function. Visualizing the resulting t-SNE plot 336 as a heatmap of marker expression of immune cells. To explore the phenotypic diversity of 337 immune cell populations in the different groups of mice FM tissues, we applied a K -nearest-338 neighbor density -based clustering algorithm called Phenograph. This algorithm allows the 339 unsupervised clustering analysis of data from single cells. The output was organized using the 340 Cluster Explorer tool to visualize the phenotypic continuum of cell populations . This tool creates 341 an interactive cluster Profile graph and heatmap and displays the cluster populations on a tSNE 342 plot. For each cluster, marker positivity was set at ≥10x the relative expression level of the marker 343 with the lowest expression level. Two-way ANOVA was employed to determine statistical 344 differences between the groups; a p-level of <0.05 was considered significant. 345 346 3. RESULTS 347 3.1 bEV introduction drives distinct immune cell differentiation programs 348 First, to determine whether bEV introduction will stimulate mucosal fetal immunity, we introduced 349 bEV particles mid-gestation intraamniotically and analyzed populations within the immediate term 350 period. Using mass cytometry via time-of-flight (CyTOF) in isolated cells from murine gut tissues, 351 we observed variably expressed clusters in the NE CD-1 mice in both lymphoid (35 clusters) and 352 myeloid (21 clusters) populations. The delineated clusters with their corresponding markers are 353 outlined in Tables 2 and 3. tSNE maps of lymphoid and myeloid populations in this setup are 354 shown in Figure 2. Interestingly, we see an increase in the intensity of the populations identified, 355 implying a general upregulation of cell populations upon bEV exposure. 356 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 357 Figure 2. tSNE maps for lymphoid (first panel) and myeloid (second panel) 358 populations in the normal environment setups (A, B) and germ -free (C, D) mice 359 setups. Each tSNE panel features populations that vary in number of absolute events 360 across different treatment setups, as visually apparent in the myeloid panel. Absent or 361 present populations also vary across different treatment setups, as visually apparent in 362 the lymphoid panel. PBS, phosphate buffered saline. bEV, bacterial extracellular vesicles. 363 hpbEV, human placental bacterial extracellular vesicles. 364 365 Table 2. Cell clusters observed in the CyTOF analysis using a lymphoid and myeloid panel 366 in normal-environment mice. 367 CLUSTER LYMPHOID MARKERS CLUSTER MYELOID MARKERS 1 CD44, CXCR3 1 CD14, CD86, CD11c, Ly6G, XCR1 2 CD44, T -bet, CXCR3, RORyt, CD24, CD8 2 CD86, Ly6G .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 3 CD45, CD44, CD24 3 CD11b, CD14, CD86, Ly6C, Ly6G, CD8, CD193, XCR1, FcER1a 4 CD44, CD24 4 CD86, FcER1a 5 CD24 5 CD86, Ly6G, XCR1 6 CD69, CCR7 6 CD86 7 CD38 7 XCR1 8 CD69, PD -1, CD44, FoxP3, CD38 8 CD14, CD86, Ly6G, c-kit 9 CD24 9 CD86 10 CD44 10 CD86, XCR1 11 CCR7 11 Ly6G 12 CXCR3 12 CD11b, CD14, CD86, Ly6C, Ly6G, CD8, CD103, CD193, XCR1, FcER1a 13 GATA3 13 CD14, CD11c 14 CD45, NK1.1, PD -1, CD44, CCR7, CXCR3, CD19, CD24, CD38, 14 --- 15 TCRyd 15 CD86, c-kit 16 CD62L 16 CD11b, F4 -80, CD86, MMR, CD11c, Ly6G, Siglec-F, XCR1 17 CD44, CXCR4 17 CD11b, CD14, CD86, Ly6C, Ly6G, XCR1, FcER1a, c-kit 18 CD24 18 19 FoxP3 19 CD14, CD86, CD11c, Ly6G, XCR1 20 --- 20 21 PD-1 21 CD11b, CD14, CD86, CD11c, MHC-II, Ly6C, Ly6G, CD8, CD103, Siglec -F, CD193, XCR1, FcER1a, c-kit .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 22 CD45, NK1.1, CD25, CD69, PD- 1, CD44, CD62L, CCR7, CD3, CXCR3, GATA3, CXCR4, FoxP3, CD24, CD27, TCRyd, CD8 23 T-bet 24 CD19, CD24 25 CD3 26 CD38 27 CD27 28 RORyt, CD24 29 RORyt 30 CD25 31 --- 32 NK1.1 33 CD45, CD44, CD3, CD4, CD27, TCRyd, CD8 34 NK1.1, CD44, CD3, CXCR3, CD8 35 GranB 368 Table 3. Cell clusters observed in the CyTOF analysis using a lymphoid and myeloid panel 369 in germ-free mice. 370 CLUSTER LYMPHOID MARKERS CLUSTER MYELOID MARKERS 1 CCR7, CD24 1 CD14, Ly6G 2 CD69, CD44, CCR7, CXCR3 2 Ly6G 3 CD45, CD44, CD24 3 CD14, Ly6G, XCR1, FceR1a 4 CD44 4 XCR1, FcER1a 5 CD69, CCR7, CXCR3 5 FcER1a 6 CD69, PD -1, CD44, CD3, FOXP3, CD38 6 CD14 7 CD69, CD44, CCR7, CD24 7 CD14, CD11c, Ly6C, Ly6G, CD8, CD103, XCR1, FcER1a, .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 8 CCR7 8 CD14 9 CD8, CD44 9 c-kit 10 CD44 10 XCR1 11 CXCR3 11 CD14, XCR1 12 CD24 12 Arg-1 13 TCRyd 13 MMR, XCR1 14 PD-1, GATA3 14 CD11b 15 T-bet 15 CD193, XCR1 16 CD62L, T-bet, RORyt 16 Ly6G, XCR1 17 CD19, CD24 17 CD86 18 --- 18 XCR1 19 CD45, CD69, CD44, CCR7, CXCR3, CD38 19 Ly6G 20 CD44 20 CD14, CD11c 21 FOXP3 21 CD103 22 CD62L 22 CD11b, F4 -80, Ly6G, Siglec - F, FCeR1a 23 CXCR4 23 Ly6C 24 --- 24 CD8 25 CD86 25 Ly6G 26 CD3 26 XCR1 27 --- 27 28 CD45, CD8, CD25, CD69, CCR7, CD4 28 FCeR1a 29 CD45, NK1.1, CD8, CD25, CD69, PD -1, CD44, CD62L, CCR7, CD3, CD4, CXCR3, GATA-3, CXCR4, RORyt, FOXP3, CD19, CD24, CD27 29 371 In Figure s 3 and 4 , various diverse populations with different markers display general 372 upregulation upon bEV exposure. In the lymphoid subset, the majority of the cells are CD45- but 373 express either of the memory markers CD4447, CCR748-50, and CD62L51-53. The remaining CD45- 374 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint clusters may represent populations of intestinal stem cells that can differentiate into epithelial 375 cells, providing an initial barrier to pathogens and continually providing a pool of ready ISCs54,55. 376 Additionally, we identified five CD45+ populations in the fetal gut of NG mice. One of the clusters 377 (Cluster 29: CD45+ CD3+ CD8+ NK1.1+ CD25+ CD69+ PD-1+ CD44+ CD62L+ CCR7+ CXCR3+ 378 GATA3+ CXCR4+ FoxP3+ CD24+ CD27+ TCRyd+) may indicate a unique progenitor in fetal life. 379 Another population, CD45+ CD24+ CD44+ cluster (Cluster 17), may point towards a unique set 380 of lymphocytes, while the CD45+ CD4+ CD8+ population may point towards a set of double -381 positive T cells. Interestingly, two of the lymphoid clusters in NG mice are exhausted phenotypes 382 (PD-1+) of NK cells (NK1.1+ CCR7+ CXCR3+ CD38+) (Clusters 32, 34) 56-58and regulatory T cell 383 markers (FoxP3+ CD69+) (Cluster 8) 59,60. Therefore, it may be presumed that in the normal 384 environment, the majority of the lymphoid cells are poised for memory phenotypes, although early 385 activation of certain populations is observable – an increase in progenitors, and a reduction in 386 exhausted phenotype. 387 388 Figure 3. Heatmaps for the myeloid (A) and lymphoid (B) populations across 389 different treatments in the normal environment mice setups. Each population can be 390 broken down into different clusters, as shown in the heatmap (blue palette). Each cluster 391 represents a specific cell type, whose marker positivities vary across each marker. 392 Overall, 20 clusters were identified for the myeloid population and 35 clusters were 393 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint identified for the lymphoid population. The relative abundances of the different clusters, 394 or cell types, vary across different treatments, as shown in the heatmap (red palette). 395 PBS, phosphate buffered saline. bEV, bacterial extracellular vesicles. hpbEV, human 396 placental bacterial extracellular vesicles. 397 398 Figure 4. Heatmaps for the myeloid (A) and lymphoid (B) populations across 399 different treatments in the germ -free mice setups. Each population can be broken 400 down into different clusters, as shown in the heatmap (blue palette). Each cluster 401 represents a specific cell type, whose marker positivities vary across each marker. 402 Overall, 29 clusters were identified for the myeloid population and 29 clusters were 403 identified for the lymphoid population. The relative abundances of the different clusters, 404 or cell types, vary across different treatments, as shown in the heatmap (red palette). 405 PBS, phosphate buffered saline. hpbEV, human placental bacterial extracellular vesicles. 406 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint In the myeloid subset, we observed clusters of CD11b+ cells, which are markers for 407 monocyte/dendritic-like cells. We presume that four clusters (Cluster 3, 12, 17, 21) are putative 408 monocyte-dendritic progenitors, since there were multiple monocyte/macrophage and dendritic 409 cell markers that are present in PBS and high -dose bEV setups only. Another CD11b+ positive 410 cluster (Cluster 16) was observed to increase only in low-dose bEV setups, which may represent 411 intestinal macrophages (CD11b+ F4 -80+ CD86+ CD11c+ XCR1+) . Six clusters of CD11b - Ly-412 6g+ clusters are present (Clusters 1, 2, 5, 8, 11, 19), which in general increase with increasing 413 bEV concentration. The third group of clusters is CD11b- Ly6G- CD86+ cells, which may be 414 transient cells brought about by the gradual differentiation of progenitors into cells of monocyte 415 and dendritic phenotype . Neutrophil-like Ly-6g+ clusters (Cluster 2, 5, 8, 11) increase 416 proportionally to the bEV dose. Early response to bEVs, then, seems to be reliant on an increase 417 in neutrophil-like cells , with potential differentiation of monocyte -dendritic progenitors into 418 macrophage/dendritic-like cells. 419 420 3.2 Germ-free mice reveal activation of granulocytic -like cells and lymphoid 421 progenitors in response to bEV stimulation 422 Second, to shed light on which immune cells will be activated in the mucosa of a bacterially-naïve 423 system, we characterized the baseline state of gut immune cell populations in the embryonic gut 424 of GF versus conventionally colonized mice with or without bEV treat ment. Notably, due to a 425 limited number of samples, we compared only a low-dose exposure of hpbEVs against a negative 426 control in our GF setups. Overall, we observe differentially expressed lymphoid (29 clusters) and 427 myeloid (29 clusters) populations in the pups of controls versus bEV -treated mice (side panels, 428 Figure 4). 429 In the myeloid group (Figure 4 A), there is an XCR1 + population (Cluster 18), Ly6G+ 430 population (Cluster 25) and a FCeR1a+ population (Cluster 28) that are only present in the low -431 dose hpbEV, while there is an XCR1 + population (Cluster 26) that is only present in the PBS 432 group. It appears that without the influence of native environment , neutrophil-like and eosinophil-433 like cells are the primary innate immune responses to bEV exposure, with varying populations of 434 dendritic-like cells being activated concomitantly. 435 Interestingly, in the lymphoid group ( Figure 4B), the majority of cell populations are at a 436 higher incidence in the PBS group compared to the low -dose bEV group. Both groups express 437 two multi-marked populations, although a more progenitor-like state is more prevalent in the PBS 438 group (Cluster 29: CD45+ NK1.1+ CD8+ CD25+ CD69+ PD -1+ CD44+ CD62L+ CCR7+ CD3+ 439 CD4+ CXCR3+ GATA-3+ CXCR4+ RORyt+ FOXP3+ CD19+ CD24+ CD27+ ) compared to the 440 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint low-dose bEV group (Cluster 28: CD45+ CD8+ CD25+ CD69+ CCR7+ CD4+) . A CD44+ 441 population (Cluster 20) is also present in the PBS group but not in the low-dose bEV group. Four 442 populations seem to be more activated upon bEV exposure – memory T-cell-like populations 443 (Cluster 2: CD69+ CD44+ CCR7+ CXCR3+ and Cluster 19: CD45+ CD69+ CD44+ CCR7+ 444 CXCR3+ CD38+) and B-cell-like populations (Cluster 17: CD24+ CD44+ cells). Taken together, 445 it appears that without the influence of the native environment, the lymphoid response upon bEV 446 exposure shifts into an early memory response with concomitant humoral regulation. 447 448 3.3 Embryonic progenitors, mucosal macrophage-like populations, and dendritic-cell-449 like populations comprise the majority of the myeloid cell populations activated upon 450 second encounter 451 Given that there is observable stimulation of fetal mucosal immunity upon bEV exposure, w e 452 theorized that (1) bEVs may contain bacterial antigens that can activate the immune system, (2) 453 the initial encounter with bEVs may stimulate “priming”, that is, enhancement of immune response 454 upon second encounter, and (3) myeloid cells respond earlier to acute insults prior to utilization 455 of lymphoid cells. Therefore, for the third part, we challenged bEV-primed young mice with E. coli 456 antigens (LPS, TSST-1) and checked which myeloid populations are upregulated in response to 457 this “second” encounter. tSNE maps are shown in Figure 5, in which we included the subset of 458 normal environment mouse populations, both fetal and young “primed” mice. A summary of the 459 population incidence changes across different setups is shown in Table 4 and 5. 460 461 Figure 5. tSNE graphs for the myeloid population in the germ -free mice and the 462 young mice setups. The overall tSNE map is shown (A), which is comprised of all cell 463 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint populations identified from the germ-free mice setups (B) and the young mice setups (C). 464 In the germ-free mice setups (B), we observe drastically different tSNEs across different 465 treatment groups, although hpbEVs appear to activate similar populations. In the young 466 mice setups (C), each priming agent, activates different populations entirely. Across 467 different challenge treatments (LPS, TSST-1), cell populations remain apparently similar 468 to controls (PBS), with the presence and presence of some populations. PBS, phosphate 469 buffered saline. bEV, bacterial extracellular vesicles. hpbEV, human placental bacterial 470 extracellular vesicles. LPS, lipopolysaccharide. TSST-1, toxic shock syndrome toxin-1. 471 472 Table 4. Cell clusters observed in the CyTOF analysis of myeloid cells in young, primed 473 mice. 474 CLUSTER MYELOID MARKERS 1 CD11c c-kit XCR1 CD14 CD11b FceR1a CD8a Gr-1 CD86 2 CD11c Arg1 c-kit XCR1 CD14 CD11b FceR1a CD193 CD8a Ly -6g Gr-1 CD103 CD86 CD206 F4-80 3 c-kit XCR1 CD86 CD206 4 CD11c Arg1 c-kit XCR1 FceR1a Gr-1 CD206 5 Arg1 c-kit XCR1 CD14 CD11b FceR1a CD193 CD8a Ly -6g Gr-1 CD103 CD86 CD206 SiglecF F4-80 6 CD11c XCR1 CD11b CD86 CD206 7 CD11c Arg1 XCR1 CD86 CD206 8 CD11c 9 CD11c XCR1 FceR1a CD86 CD206 10 XCR1 CD86 CD206 F4-80 11 XCR1 12 CD11c XCR1 13 CD11c c-kit XCR1 CD86 CD206 Siglec-F 14 XCR1 CD86 15 CD11c c-kit XCR1 CD103 CD86 CD206 16 CD11c Arg1 c-kit XCR1 CD11b FCeR1a CD193 GR-1 CD86 CD206 17 CD11c Arg1 CD11b CD86 CD206 18 CD11c XCR1 CD11b CD86 F4-80 19 CD11c XCER1 CD14 CD11b FCeR1a CD193 CD8a Ly -6g Gr-1 CD103 CD86 CD206 Siglec-F F4-80 20 CD11c Gr-1 CD86 CD206 21 XCR1 MHC-II .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 22 -- 23 CD86 24 CD11c Arg1 XCR1 CD11b FceR1a CD193 CD8a Ly-6g Gr-1 CD86 25 CD11b 26 Arg1 CD11b FceR1a CD206 27 CD11b Ly-6g Gr-1 CD86 28 CD11c 29 CD11c Arg1 c-kit XCR1 CD14 CD11b FceR1a Gr-1 30 CD11c Arg1 c -kit XCR1 CD14 CD11b FceR1a CD193 CD8a Ly -6g Gr-1 CD86 CD206 Siglec-F 475 476 Table 5. Comparisons across different setups of myeloid populations, and the absolute 477 differences between the two populations shown using two -way ANOVA. A significant 478 difference is denoted by p < 0.05. 479 Setups Comparator Difference Significance Cluster 3 (c-kit+ XCR1+ CD86+ CD206+) Fetal (PBS) Young hpBEV-primed, PBS exposed -10.91 0.0003 Young hpBEV-primed, LPS exposed -12.19 <0.0001 Young hpBEV -primed, TSST -1 exposed -8.835 0.0103 Fetal (EC bEV) Young hpBEV-primed, PBS exposed -10.29 0.0008 Young hpBEV-primed, LPS exposed -11.57 <0.0001 Young hpBEV -primed, TSST -1 exposed -8.22 0.026 Fetal (LD hpbEV) Young hpBEV-primed, PBS exposed -10.23 <0.0001 Young hpBEV-primed, LPS exposed -11.5 <0.0001 Young hpBEV -primed, TSST -1 exposed -8.153 0.0066 Fetal (HD hpbEV) Young hpBEV-primed, PBS exposed -10.99 <0.0001 Young hpBEV-primed, LPS exposed -12.27 <0.0001 Young hpBEV -primed, TSST -1 exposed -8.918 0.0015 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint Young PBS primed, PBS exposed Young hpBEV-primed, PBS exposed -8.117 0.007 Young hpBEV-primed, LPS exposed -9.393 0.0006 Young PBS primed, LPS exposed Young hpBEV-primed, PBS exposed -7.93 0.0098 Young hpBEV-primed, LPS exposed -9.207 0.0008 Young PBS primed, TSST-1 exposed Young hpBEV-primed, PBS exposed -7.883 0.0106 Young hpBEV-primed, LPS exposed -9.16 0.0009 Young EC-bEV primed, PBS exposed Young hpBEV-primed, PBS exposed -7.357 0.0258 Young hpBEV-primed, LPS exposed -8.633 0.0026 Young EC-bEV primed, LPS exposed Young hpBEV-primed, PBS exposed -7.587 0.0177 Young hpBEV-primed, LPS exposed -8.863 0.0017 Cluster 4 (CD44+) Fetal (PBS) Young hpBEV-primed, PBS exposed -11.33 0.0001 Fetal (EC bEV) Young hpBEV-primed, PBS exposed -11.32 0.0001 Fetal (LD hpbEV) Young EC bEV-primed, LPS exposed -7.161 0.0351 Young EC bEV -primed, TSST -1 exposed -7.548 0.0189 Young hpbEV-primed, PBS exposed -11.29 <0.0001 Fetal (HD hpbEV) Young EC bEV-primed, LPS exposed -7.238 0.0312 Young EC bEV -primed, TSST -1 exposed -7.625 0.0166 Young hpbEV-primed, PBS exposed -11.37 <0.0001 Cluster 6 (CD69+ PD-1+ CD44+ CD3+ FOXP3+ CD38+) Fetal (PBS) Young EC bEV-primed, PBS exposed -7.792 0.047 Fetal (LD hpbEV) -7.487 0.0209 Fetal (HD hpbEV) -7.737 0.0137 Cluster 7 (CD69+ CD44+ CCR7+ CD24+) Fetal (LD hpbEV) Young EC bEV-primed, LPS exposed -7.197 0.0332 Fetal (HD hpbEV) -7.412 0.0236 Cluster 8 (CCR7+) Fetal (HD hpbEV) Young hp bEV-primed, PBS exposed 7.433 0.0228 Young hp bEV-primed, LPS exposed 7.653 0.0158 Young hp bEV -primed, TSST -1 exposed 6.933 0.0496 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 480 Heat maps for the myeloid population are shown in Figure 6. Overall, two-way ANOVA 481 suggests that most of the clusters that significantly change in mean counts across various 482 treatments occur in Clusters 1, 2, and 3 (p < 0.0001). Cluster 1 (c -kit+ XCR1+ CD14+ CD11c+ 483 CD11b+ FceR1a+ CD8a+ Gr-1+ CD86+) and Cluster 2 (c-kit+ XCR1+ CD14+ CD11c+ CD11b+ 484 FceR1a+ CD8a+ Gr-1+ CD86+ Arg1+ CD193+ Ly-6g+ CD103+ CD206+ F4-80+) both represent 485 widely-marked myeloid progenitor cell populations. In both clusters, significant differences in the 486 mean populations are observed against all young mouse myeloid cells across all treatments, 487 strongly indicating that these cell types are only found in fetal environments and in embryonic 488 tissues. Comparing across fetal populations, it is interesting to note that both populations are not 489 significantly different in (1) control conditions vs. a high dose of hpbEVs (p > 0.9999) and (2) EC 490 bEV treated mice vs. a low dose of hpbEVs (p > 0.9999). 491 492 Figure 6. Heat maps for the myeloid populations in the germ-free mice setups and 493 the young mice setups. (A) Each population can be broken down into different clusters, 494 as shown in the heatmap (blue palette). Overall, due to the large number of populations 495 (200), the study was limited only to the top 30 most abundant clusters for succeeding 496 analyses. The relative abundances of the different clusters, or cell types, vary across 497 different treatments, as shown in the heatmap (dark blue palette). Relatively, clusters 1 498 and 2 in the germ -free mice setups are significantly higher (two -way ANOVA, p <0.05) 499 compared to most of the populations within their setups and across different treatments. 500 Clusters 3-8 significantly differ in abundances (p 0.05). (B) Each cluster represents a specific cell 503 type, whose marker positivities vary across each marker. A summary of all markers is 504 presented in Table 5 in the text. PBS, phosphate buffered saline. EC bEV, E. coli bacterial 505 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint extracellular vesicles. hpbEV, human placental bacterial extracellular vesicles. LPS, 506 lipopolysaccharide. TSST-1, toxic shock syndrome toxin-1. 507 508 In Cluster 3 (c-kit+ XCR1+ CD86+ CD206+), markers for monocytes/macrophages in the 509 intestinal epithelia, there are significant differences between fetal and young mice . These cells 510 are activated more in hpbEV -primed young mice compared to fetal mice, and only certain 511 treatment groups in young mice have more of these cellpopulous. Interestingly, these cells do not 512 significantly differ between fetal mouse setups. 513 Various dendritic cell groups, characterized by CD11c+ CD206+ and varying other 514 markers, have varying population differences across different setups. Among treatment groups 515 with significant differences, Clusters 4,6, and 7 generally showed lower counts in fetal setups 516 versus young bEV-primed mice. Interestingly, Cluster 8, which is the only single -marked cluster 517 (CD11c+ only), showed higher population counts in fetal setups compared to young hpbEV -518 primed mice. 519 In summary, our data suggest that upon second encounter, embryonic progenitors, 520 macrophage/dendritic-cell-like cells , and dendritic cell -like populations dominate the myeloid 521 response in GF mice. In contrast, in NE mice, a more targeted macrophage-like and dendritic cell-522 like response dominates. 523 524 4. Discussion 525 We established that bEVs influence fetal mucosal immunity and foster immune 526 programming for a more targeted response to subsequent challenges. Introducing bEVs into the 527 fetal gut drove distinct immune cell differentiation programs, as shown by CyTOF analysis in 528 murine models. In conventionally colonized mice, bEV exposure led to increased memory and 529 regulatory markers in lymphoid populations, with a concomitant dose -dependent expansion of 530 monocyte/dendritic and neutrophil-like lineages in myeloid clusters. In GF setups, this exposure 531 primarily activated granulocytic -like cells and shifted lymphocytes toward early memory 532 phenotypes. Upon a second encounter with E. coli antigens, progenitor cells, 533 macrophage/dendritic-like populations, and dendritic cell subsets dominated the myeloid 534 response. 535 In the intestine, multiple lymphoid microenvironments exist, including the mucosa -536 associated lymphoid tissue (MALT), the lamina propria, and the epithelial cell tracts lined with 537 intraepithelial lymphocytes (IEL) 61. Via in utero fetal swallowing, any bEVs introduced in the 538 amniotic sac can be transferred into the fetal gut and activate naïve immunity in the tissues. Our 539 observations in lymphoid and myeloid cells led us to speculate that bEVs stimulate the immune 540 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint environment in the murine fetal gut. Our findings have the potential to advance our understanding 541 of how maternal inflammation during pregnancy programs long-term immune output from 542 hematopoietic progenitors and subsequently impacts offspring immune function. 543 In the lymphoid population, we identified five CD45+ populations in the fetal gut of normal-544 environment mice. One interesting population is likely lymphoid progenitors (Cluster 22) 545 expressing multiple markers of various canonical lymphocytes ( CD45+ CD3+ CD8+ NK1.1+ 546 CD25+ CD69+ PD -1+ CD44+ CD62L+ CCR7+ CXCR3+ GATA3+ CXCR4+ FoxP3+ CD24+ 547 CD27+ TCRyd+ ). The conventional outline for lymphocyte differentiation starts from the 548 hematopoietic stem cells, differentiating into lymphoid-primed multipotent progenitors (LMPP), to 549 common lymphoid progenitors (CLP), to committed progenitors, and finally to mature 550 lymphocytes62. The slew of markers in this cluster may point towards a new lymphoid progenitor 551 phenotype in the intestines that can readily differentiate into specific lineages, as needed. In our 552 GF mice, we also observe a similar lymphoid cluster bearing multiple markers (Cluster 29). The 553 relatively low incidence of these cells indicates they may represent transient populations since 554 extrathymic development in the intestines is naturally repressed in the presence of a normal 555 thymus63. 556 PD-1 is a marker typically associated with exhaustion , but may alternatively represent a 557 marker of activation and chronic stimulation 64,65. Both clusters decrease in incidence with 558 increasing bEV exposure, possibly indicating a decrease in exhausted phenotypes with 559 concomitant release from PD-1 inactivation that would allow for enhanced action in response to 560 bEV exposure. We also observe a small population of double -positive T cells (CD45+ CD4+ 561 CD8+) in both NE and GF mice (Cluster 33 and Cluster 28, respectively) that increases with 562 exposure to increasing doses of bEVs. Albeit relatively low in incidence, we postulate that double-563 positive T cells may play important roles in fetal immunity due to their memory nature as well as 564 their potential to produce higher levels of cytokines and chemokines compared to CD4+ or CD8+ 565 only cells66,67. 566 In the intestine, CD24+ CD44+ cells represent fetal intestinal stem cells (ISCs), which are 567 important in potentiating the self -renewing capacity of the tissue to respond to antigenic 568 challenges and injury. Although Lgr5 is a more putative marker for ISCs, CD24 provides an 569 acceptable marker for identifying stem cells in mice, with similar counterparts in human intestinal 570 cells 68-71. ISCs provide crosstalk with other lymphocytes for further differentiation into Paneth 571 cells or tuft cells. Paneth cells are secretory cells that may aid intestinal defense by secreting 572 granules containing a host of defensive proteins, including antimicrobial peptides and enzymes72-573 74. ISCs can also express MHCs, allowing them to serve as adjunct antigen -presenting cells for 574 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint T-cell activation and proliferation 75,76. We observed a general increase in the incidence of the 575 clusters upon bEV exposure compared to negative controls, suggesting that bEV presence 576 preemptively induces preparation of the intestines for possible damage. However, the presence 577 of CD45 antigen in the CD24+ cluster strongly suggests that this is a lymphocyte population . A 578 population of TCR ɣ𝛿 cells has been demonstrated to produce a similar phenotype, but the 579 absence of a TCR ɣ𝛿 marker in this population suggests otherwise 77. It might be that the 580 populations are an early phenotype of naïve T cells that do not express CD4+ or CD8+ yet, but 581 are early activated via CD44 upregulation during encounter with antigens78. 582 As for the CD45 - cells, the majority of the populations express CD44 and CCR7, in 583 combination with other markers; interestingly, the populations would either have only these 584 markers present or have them in combination with a few markers that would be incomplete for us 585 to classify them into one of the canonical lymphocyte types. In line with the above hypothesis, we 586 suspect that the majority of the clusters are transition states into other lymphocytes with early 587 expression of CD44, CCR7, and/or CD62L as early memory or activation markers. 588 In general, we observe the predominant upregulation of markers as bEV exposure 589 increases, displaying the capacity of bEVs to induce memory and/or activation phenotypes in 590 these cell clusters. Therefore, two populations of naïve memory populations are fine-tuned by 591 CD45 expression : (1) CD45+ populations are cells trending towards maturity via increasing 592 survival signal receptivity, while (2) CD45 - populations are cells that are geared toward an 593 augmented inflammatory response to the presence of bEVs79-82. However, in our GF mice, we do 594 not see any significant differences between negative controls and bEV -exposed pups in the 595 CD45+ CD24+ CD44+ cluster, instead we see a similar trend in the increase with the CD45 - 596 CD24+ CD44+ cluster. Therefore, if we reduce any influences from the environment, the acute 597 course of fetal murine gut immunity is to expand inflammatory responses against the presence of 598 bEVs, while maintaining all populations in the course of maturation. 599 In the myeloid population, we observe six (6) clusters of CD11b+ cells in the NE mice. The 600 choice in utilizing CD11b as a primary gate is secondary to the abundance of CD11b-expressing 601 monocytes, macrophages, and dendritic cells that lie within the murine epithelium at the start of 602 myeloid-derived cell residency coming from the fetal liver and blood 83-85. Three clusters (Cluster 603 3, 12, 21) display multiple markers reminiscent of the lymphoid progenitor population we have 604 postulated previously. The observed markers correspond to dendritic cells ( CD11b+, CD103+) 605 and macrophages/monocytes (F4/80+, CD14+ , CD86+ , XCR1+, Ly6C+, Gr -1+), with a few 606 uncommon markers as well (CD8+, FcεRIα+, CD193+, Siglec-F+)83-89. We purport that these cells 607 may be monocyte-dendritic progenitors, which can produce monocyte-derived dendritic cells90,91. 608 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint Interestingly, we do not see a strong expression of MHCII in any of the cell clusters, potentially 609 indicating that the levels of bEV are not high enough to induce antigen presentation, and that 610 development of exposed immune cells in the gut may be instead driven by direct interactions with 611 bEVs; this remain to be proven via functional studies. Another alternative explanation is that 612 MHCII is in the myeloid panel where antibodies against epithelial cells (CD24, CD44) were not 613 accounted for; indeed, Paneth cells and crypt villi in the intestine natively express MHCII and may 614 play important roles in antigen presentation92,93; future research may also attempt to elucidate the 615 role of these cells in inflammation and immune education upon bEV exposure. Nonetheless, we 616 see a similar progenitor cluster in GF mice (Cluster 7) that also observes the same trend, 617 indicating that this population may be an important player in myeloid immunity. 618 There are three clusters of cells (Myeloid Cluster 1, 16, 19) that are upregulated in low 619 concentrations of bEVs but downregulated upon exposure to high concentrations in NE mice. 620 Although closely resembling some markers present in the previously mentioned monocyte -621 dendritic progenitors, the fewer markers present in the clusters, as well as their differential 622 expression, may point to an entirely different population . We suspect that the clusters may 623 represent intestinal macrophages (CD11b+/- F4/80+ CD86+ CD11c+ XCR1+) . It is yet unclear 624 why there is a concentration-dependent decrease in the incidence of these cells . Still, it can be 625 speculated that there might be a switch towards a different phenotype of macrophages that were 626 not captured by our panel. Possible alternative populations may include (1) CD169+ 627 macrophages, which are mainly regulatory in nature94,95, and (2) TCR+ macrophages, which are 628 TNF-tuned, CCL2 -releasing, and highly phagocytic subpopulations96,97. We observe a similar 629 macrophage cluster in RG mice (Cluster 14) that also exhibits the same trend. 630 Interestingly, we see an increase in neutrophils (Ly6G+) in both groups of mice (Cluster 631 11 for NE mice, Clusters 16, 19, 25 in GF mice). Upon acute inflammation, resident monocytes in 632 the intestine recruit neutrophils to aid in the response against pathogens 98,99. Neutrophils can 633 release chemokines that can recruit a second wave of macrophages 100; however, as mentioned 634 previously, neutrophil-assisted recruitment of macrophages may only contribute to an increase of 635 macrophage counts in lower concentrations of bEVs. 636 For the GF mice, an interesting cluster is the B -cell-like population, characterized by 637 CD19⁺ CD24⁺. This mimics a tolerogenic phenotype, similar to an IL-10 suppressed Treg cell; in 638 an experimental model (collagen -induced arthritis), the adoptive transfer of CD19 ⁺CD24⁺ 639 transitional B cells reduced disease severity via secretion of IL -10 in an inflammatory 640 environment101. It is attractive to think that in an environment without native influences, Bregs 641 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint contribute to maintenance of homeostatic neutrality that prevents over -exertion of the immune 642 system to a transient exposure and may serve to also train the immune system 102,103. 643 Overall, we observe differentially expressed populations in the pups of controls versus 644 bEV treated mice, with general upregulation of a majority populations as bEV concentration 645 increases. Lymphoid progenitors, intestinal stem cells, and monocyte -dendritic progenitors are 646 increased in both NE and GF mice, indicating that bEV exposure has the propensity to produce 647 a pool of cells readily available for targeted differentiation and maturation. Exhausted phenotypes 648 and memory/activated populations also decrease upon bEV exposure. Macrophages are 649 produced in a concentration -dependent manner, with lower concentrations of bEVs providing 650 higher incidences compared to higher concentrations. Neutrophil production is also stimulated 651 upon bEV exposure, but it remains to be seen whether cells are mature enough to contribute to 652 fetal intestinal immunity. 653 Our experiments with young mice provide insights as to how prenatal bEV stimulation may 654 modulate second encounter responses. Similar to what we have seen in myeloid populations in 655 embryonic conditions, a majority of the populations upregulated in young mice are progenitors 656 characterized by multiple cellular markers (Cluster 1 and Cluster 2) . There are two important 657 points with these progenitors: (1) they are highly present in fetal life at baseline, as indicated by 658 the absence of significant differences among embryonic controls, and (2) they do not persist in 659 later life, as seen with the significant decreases from fetal setups to young mice setups. We can 660 potentially think of these cells as native progenitors resting within the murine intestine that slowly 661 differentiate into other cells over time . Whether the progenitors function differently compared to 662 the aforementioned CD45+ intestinal stem cells is yet unknown. In both clusters, there are two 663 interesting scenarios where we see no difference in progenitor counts. In high -dose hpbEVs 664 versus controls, we theorize that the high dose leads to an immune response that decreases the 665 antigen burden and decreases the recruitment of progenitors. In mice treated with EC bEV versus 666 a low dose of hpbEV, we theorize that both exposures are biological equivalents of each other 667 leading to no-difference in progenitor counts upon encounter. 668 Cluster 3 may represent dendritic cells in the intestine that are activated upon introduction 669 of an unrecognized antigen. CD86 and CD206 are canonical macrophage markers for M1 and 670 M2 macrophages, respectively 104 . However, c-kit positivity is an essential feature of dendritic 671 cells 105 The presence of both CD86 and CD206 markers points more toward a monocyte-derived 672 dendritic cell. Noticeably, only hpbEV-primed setups show a significant increase in population 673 counts regardless of exposure status in young life. Therefore, we postulate that in response to 674 hpbEV exposure, dendritic cells become activated to serve as an initial defense mechanism for 675 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint the mucosal environment. However, the lack of MHC II positivity in these cells may imply the 676 beginning of clonal expansion upon initial contact. 677 Other dendritic cell-like phenotypes also increase upon bEV exposure. In CD44+ clusters, 678 we see that hpbEV primed cells have baseline higher CD44+ cells compared to fetal setups. 679 ECbEV-primed setups have notably higher CD44+ counts compared to hpbEV -exposed fetal 680 mice. We can speculate that the CD44+ cells are higher in primed mice due to their relative ly 681 longer exposure time with the hpbEVs compared to embryonic ones. Interestingly, EC bEV-682 primed mice also showed higher CD44+ counts versus hpbEV exposed fetal mice, showing that 683 subsequent exposure to LPS/TSST-1 in EC bEV-primed mice may trigger activation of a primed 684 state that enables a higher response to a secondary exposure. 685 Cluster 6 may also represent progenitor cells, as the cluster expresses neutrophil cell 686 (CD38+) and dendritic cell markers (CD69+ CD44+) . CD3+ expression is a marker of 687 macrophage cells106 while regulation of FOXP3 in myeloid cells is necessary for differentiation 688 and maturation. Deletion of PD-1 in myeloid cells stimulates an increase in T effector memory cell 689 proliferation, increasing memory responses among T cells with improved functionality107 . In tumor 690 infiltrating myeloid cells, PD -1 positively regulates the differentiation of myeloid -derived 691 suppressive cells, subsequently leading to a suppression of T effector cells population and 692 function. The expression of this cluster increases in EC-bEV primed, PBS exposed mice, implying 693 that EC-bEV priming, on baseline, stimulates an increase in suppressive clusters. Non-difference 694 upon priming with hpbEV and LPS/TSST post-exposure in EC bEV young mice may imply that 695 (1) hpbEV exposure does not stimulate production of suppress ive clusters and (2) LPS/TSST 696 post-exposure results in dampening of suppressors to promote an increased response upon 697 recognition of these antigens. 698 CCR7+ is an inflammatory marker expressed by myeloid dendritic cells and lymphoid 699 cells, resulting in chemotaxis of recruited cells 108,109 . Thus, Cluster 8 cells may represent myeloid 700 cells that are increased secondary to a controlled inflammation from hpbEV priming. It is attractive 701 to think that the decrease in hpbEV -primed setups in Cluster 1 and 2 is due to differentiation to 702 these monocyte-dendritic cell hybrids; however, further trajectory analysis would be required to 703 confirm this hypothesis. Interestingly, the observed increases are mild and may thus represent an 704 early inflammatory response to priming and/or antigen exposure. 705 Notably different from the primed setups versus embryonic fetal setups is the absence of 706 purely macrophage or neutrophil populations. In fetal life , fetal macrophages , neutrophils, and 707 monocytes can respond to antigens as the major pro -inflammatory cells in the developing 708 fetus89,110. We identified several clusters that may be macrophage and neutrophil candidates, 709 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint although the clusters do not vary significantly between different setups. We speculate that these 710 cells may be more apparent in later time points, since it is expected that dendritic cells would be 711 the first responders to a foreign antigen, subsequently activating other myeloid cells to initiate an 712 immune response111,112. It may also be advantageous that the differentiated cells are kept to a 713 lower state of activation of proliferation to prevent unwarranted, widespread immune response 714 within the gut and promote a constant low exposure of antigens that can be exploited for intestinal 715 immune education112. 716 Our study has certain limitations. First, it is notable that most of our populations do not 717 readily fit into the classical descriptions of conventional immune cells in terms of surface markers. 718 Although we provide the phenotypic landscape of the immune repository upon exposure to 719 bacterial extracellular vesicles, functional studies are still necessary to provide context as to how 720 these cells may act locally in response to a relatively mild immunogen. Second, we also utilized 721 mice as our subjects due to their convenience and manipulability. Although this model may 722 explain how humans would also react toward bEV exposure, studies utilizing human-derived cells 723 or a humanized mouse model may be more representative. Third, interaction with the environment 724 will also play a crucial role in modifying the early immunophenotype of our models. Therefore, 725 studies that employ purely GF mice under GF environments will provide a more conclusive look 726 into how naïve immunity is affected by environmental changes and the evolving microbiome. 727 Fourth, o ur approach also utilizes a simple k -means clustering network – more powerful 728 approaches toward dissecting the overall population may be warranted in future studies, and 729 trajectory analysis will determine how one phenotype of cells evolve s into other phenotypes. 730 Finally, additional research, especially functional studies, are needed to further elucidate the exact 731 mechanisms of fetal immune development upon bEV exposure. 732 As the establishment of the presence of placental bEVs has only come into recent light, 733 this study addresses the lack of studies into the importance of bEVs on the developing immune 734 system. Our study provides an initial look at how bEVs can influence protective responses in early 735 life. In summary, we see that (1) progenitors are highly prominent during fetal life, (2) there is a 736 decrease of progenitor populaitnos upon birth going to young life, and (3) there is a propensity for 737 myeloid cells to differentiate into antigen-presenting cells in a primed response – all of which are 738 expected trends in immune system development upon exposure to an antigen, such as bEVs. 739 This study also raises several important research avenues to address in the future, including (1) 740 Which specific populations do the identified cell populations differentiate into upon longer 741 exposure? (2) What are the corresponding lymphoid responses in the primed setup? (3) Would 742 responses between animal and human immunity be the same ? (4) What are the functional 743 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint responses of an EV -primed system against other antigens ? and (5) What is the value of bEVs 744 and any other modifications in effecting a memory response similar to a vaccine ? The last point 745 is a relevant point, since EVs can provide a more sustainable and convenient approach to 746 vaccination113,114Finally, an exciting question is whether EVs coming from various sources (plants, 747 viruses, bacteria) would provoke similar or discrete responses in early-life immunity and whether 748 the responses can be harnessed to develop more targeted immune tolerance/immunity against 749 specific pathogens. 750 751 5. Conclusion 752 This study elucidates the role of bEVs in shaping fetal immune development. We observed key 753 immune trends, including the prominence of progenitor cells during fetal life, their decline 754 postnatally, and the primed differentiation of myeloid cells into antigen -presenting cells. Our 755 findings align with expected immune maturation patterns following antigenic exposure. Future 756 studies should explore the long-term differentiation of these cells, their interaction with lymphoid 757 responses, and the potential for bEVs to induce memory -like immunity. Given their stability and 758 bioavailability, bEVs may serve as a novel platform for immune modulation, with applications in 759 neonatal immunity, vaccine development, and immune tolerance strategies. 760 Acknowledgment 761 We sincerely thank Ms. Talar Kechichian for her invaluable assistance in securing the animal 762 amendment in a timely manner. Ms. Phyllis Gamble for her dedication to maintaining the animal 763 colony. Additionally, we are grateful to Dr. Enkhtuya Radnaa , Ph.D., for her expertise and 764 support in performing surgical procedures. The contributions of the aforelisted individuals were 765 instrumental in the successful completion of this study. 766 767 Conflict of interest 768 Authors declare there is no conflict of interest. 769 770 Data Availability Statement 771 Mass Cytometry Data are available in https://doi.org/10.5281/zenodo.15658656 These datasets 772 provide comprehensive insights into the cellular composition and protein expression profiles 773 analyzed through mass cytometry techniques. 774 775 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint Consent of publication 776 Not applicable 777 778 Funding : The Robert and Janice McNair Foundation, R01 HD109095, and R01 HD109780 to 779 S.A.B. R01HD100729-05 to RM 780 781 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint

782 783 1. Stout MJ, Conlon B, Landeau M, et al. Identification of intracellular bacteria in the basal 784 plate of the human placenta in term and preterm gestations. Am J Obstet Gynecol. 785 2013;208(3):226 e221-227. 786 2. PrabhuDas M, Bonney E, Caron K, et al. Immune mechanisms at the maternal -fetal 787 interface: perspectives and challenges. Nat Immunol. 2015;16(4):328-334. 788 3. Liu S, Diao L, Huang C, Li Y, Zeng Y, Kwak-Kim JYH. The role of decidual immune cells 789 on human pregnancy. J Reprod Immunol. 2017;124:44-53. 790 4. Mor G, Kwon JY. Trophoblast -microbiome interaction: a new paradigm on immune 791 regulation. Am J Obstet Gynecol. 2015;213(4 Suppl):S131-137. 792 5. Lash GE. Molecular Cross-Talk at the Feto-Maternal Interface. Cold Spring Harb Perspect 793 Med. 2015;5(12). 794 6. Gomez-Lopez N, StLouis D, Lehr MA, Sanchez -Rodriguez EN, Arenas -Hernandez M. 795 Immune cells in term and preterm labor. Cell Mol Immunol. 2014;11(6):571-581. 796 7. Alijotas-Reig J, Llurba E, Gris JM. Potentiating maternal immune tolerance in pregnancy: 797 a new challenging role for regulatory T cells. Placenta. 2014;35(4):241-248. 798 8. Nancy P, Tagliani E, Tay CS, Asp P, Levy DE, Erlebacher A. Chemokine gene silencing 799 in decidual stromal cells limits T cell access to the maternal -fetal interface. Science. 800 2012;336(6086):1317-1321. 801 9. Williams PJ, Searle RF, Robson SC, Innes BA, Bulmer JN. Decidual leucocyte populations 802 in early to late gestation normal human pregnancy. J Reprod Immunol. 2009;82(1):24-31. 803 10. Harris LK, Benagiano M, D'Elios MM, Brosens I, Benagiano G. Placental bed research: II. 804 Functional and immunological investigations of the placental bed. Am J Obstet Gynecol. 805 2019;221(5):457-469. 806 11. Rouas-Freiss N, Moreau P, LeMaoult J, Papp B, Tronik -Le Roux D, Carosella ED. Role 807 of the HLA-G immune checkpoint molecule in pregnancy. Hum Immunol. 2021;82(5):353-808 361. 809 12. Jacobs SO, Sheller -Miller S, Richardson LS, Urrabaz -Garza R, Radnaa E, Menon R. 810 Characterizing the immune cell population in the human fetal membrane. Am J Reprod 811 Immunol. 2021;85(5):e13368. 812 13. True H, Blanton M, Sureshchandra S, Messaoudi I. Monocytes and macrophages in 813 pregnancy: The good, the bad, and the ugly. Immunol Rev. 2022;308(1):77-92. 814 14. Motomura K, Hara M, Ito I, Morita H, Matsumoto K. Roles of human trophoblasts' pattern 815 recognition receptors in host defense and pregnancy complications. J Reprod Immunol. 816 2023;156:103811. 817 15. Muralidhara P, Sood V, Vinayak Ashok V, Bansal K. Pregnancy and Tumour: The Parallels 818 and Differences in Regulatory T Cells. Front Immunol. 2022;13:866937. 819 16. Xie M, Li Y, Meng YZ, et al. Uterine Natural Killer Cells: A Rising Star in Human Pregnancy 820 Regulation. Front Immunol. 2022;13:918550. 821 17. Levy M, Blacher E, Elinav E. Microbiome, metabolites and host immunity. Curr Opin 822 Microbiol. 2017;35:8-15. 823 18. Netea MG, Joosten LA, Latz E, et al. Trained immunity: A program of innate immune 824 memory in health and disease. Science. 2016;352(6284):aaf1098. 825 19. Mandal M, Donnelly R, Elkabes S, et al. Maternal immune stimulation during pregnancy 826 shapes the immunological phenotype of offspring. Brain Behav Immun. 2013;33:33-45. 827 20. Mishra A, Lai GC, Yao LJ, et al. Microbial exposure during early human development 828 primes fetal immune cells. Cell. 2021;184(13):3394-3409 e3320. 829 21. Zengeler KE, Lukens JR. Innate immunity at the crossroads of healthy brain maturation 830 and neurodevelopmental disorders. Nat Rev Immunol. 2021;21(7):454-468. 831 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 22. Jash S, Sharma S. In utero immune programming of autism spectrum disorder (ASD). 832 Hum Immunol. 2021;82(5):379-384. 833 23. Scott RL, Vu HTH, Jain A, Iqbal K, Tuteja G, Soares MJ. Conservation at the uterine -834 placental interface. Proc Natl Acad Sci U S A. 2022;119(41):e2210633119. 835 24. Hodyl NA, Stark MJ, Osei -Kumah A, Clifton VL. Prenatal programming of the innate 836 immune response following in utero exposure to inflammation: a sexually dimorphic 837 process? Expert Rev Clin Immunol. 2011;7(5):579-592. 838 25. Langley-Evans SC, Carrington LJ. Diet and the developing immune system. Lupus. 839 2006;15(11):746-752. 840 26. Lopez DA, Apostol AC, Lebish EJ, et al. Prenatal inflammation perturbs murine fetal 841 hematopoietic development and causes persistent changes to postnatal immunity. Cell 842 Rep. 2022;41(8):111677. 843 27. Zhang Y, Bi J, Huang J, Tang Y, Du S, Li P. Exosome: A Review of Its Classification, 844 Isolation Techniques, Storage, Diagnostic and Targeted Therapy Applications. Int J 845 Nanomedicine. 2020;15:6917-6934. 846 28. Chen YF, Luh F, Ho YS, Yen Y. Exosomes: a review of biologic function, diagnostic and 847 targeted therapy applications, and clinical trials. J Biomed Sci. 2024;31(1):67. 848 29. Shepherd MC, Radnaa E, Tantengco OA, et al. Extracellular vesicles from maternal 849 uterine cells exposed to risk factors cause fetal inflammatory response. Cell Commun 850 Signal. 2021;19(1):100. 851 30. Sheller-Miller S, Trivedi J, Yellon SM, Menon R. Exosomes Cause Preterm Birth in Mice: 852 Evidence for Paracrine Signaling in Pregnancy. Sci Rep. 2019;9(1):608. 853 31. Menon R, Khanipov K, Radnaa E, et al. Amplification of microbial DNA from bacterial 854 extracellular vesicles from human placenta. Front Microbiol. 2023;14:1213234. 855 32. Baglio SR, van Eijndhoven MA, Koppers -Lalic D, et al. Sensing of latent EBV infection 856 through exosomal transfer of 5'pppRNA. Proc Natl Acad Sci U S A. 2016;113(5):E587-857 596. 858 33. Nabet BY, Qiu Y, Shabason JE, et al. Exosome RNA Unshielding Couples Stromal 859 Activation to Pattern Recognition Receptor Signaling in Cancer. Cell. 2017;170(2):352-860 366 e313. 861 34. Torralba D, Baixauli F, Villarroya -Beltri C, et al. Priming of dendritic cells by DNA -862 containing extracellular vesicles from activated T cells through antigen -driven contacts. 863 Nat Commun. 2018;9(1):2658. 864 35. Diamond JM, Vanpouille-Box C, Spada S, et al. Exosomes Shuttle TREX1-Sensitive IFN-865 Stimulatory dsDNA from Irradiated Cancer Cells to DCs. Cancer Immunol Res. 866 2018;6(8):910-920. 867 36. Fabbri M, Paone A, Calore F, et al. MicroRNAs bind to Toll -like receptors to induce 868 prometastatic inflammatory response. Proc Natl Acad Sci U S A. 2012;109(31):E2110-869 2116. 870 37. Li X, Lei Y, Wu M, Li N. Regulation of Macrophage Activation and Polarization by HCC -871 Derived Exosomal lncRNA TUC339. Int J Mol Sci. 2018;19(10). 872 38. Liu J, Fan L, Yu H, et al. Endoplasmic Reticulum Stress Causes Liver Cancer Cells to 873 Release Exosomal miR-23a-3p and Up-regulate Programmed Death Ligand 1 Expression 874 in Macrophages. Hepatology. 2019;70(1):241-258. 875 39. Zhao S, Mi Y, Guan B, et al. Tumor-derived exosomal miR-934 induces macrophage M2 876 polarization to promote liver metastasis of colorectal cancer. J Hematol Oncol. 877 2020;13(1):156. 878 40. Loh JT, Zhang B, Teo JKH, et al. Mechanism for the attenuation of neutrophil and 879 complement hyperactivity by MSC exosomes. Cytotherapy. 2022;24(7):711-719. 880 41. Thery C, Duban L, Segura E, Veron P, Lantz O, Amigorena S. Indirect activation of naive 881 CD4+ T cells by dendritic cell-derived exosomes. Nat Immunol. 2002;3(12):1156-1162. 882 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 42. Hong X, Schouest B, Xu H. Effects of exosome on the activation of CD4+ T cells in rhesus 883 macaques: a potential application for HIV latency reactivation. Sci Rep. 2017;7(1):15611. 884 43. Maybruck BT, Pfannenstiel LW, Diaz-Montero M, Gastman BR. Tumor-derived exosomes 885 induce CD8(+) T cell suppressors. J Immunother Cancer. 2017;5(1):65. 886 44. Sheller-Miller S, Choi K, Choi C, Menon R. Cyclic-recombinase-reporter mouse model to 887 determine exosome communication and function during pregnancy. Am J Obstet Gynecol. 888 2019;221(5):502 e501-502 e512. 889 45. Vidal MS, Jr., Radnaa E, Vora N, et al. Establishment and comparison of human term 890 placenta-derived trophoblast cellsdagger. Biol Reprod. 2024;110(5):950-970. 891 46. Timmons BC, Reese J, Socrate S, et al. Prostaglandins are essential for cervical ripening 892 in LPS -mediated preterm birth but not term or antiprogestin -driven preterm ripening. 893 Endocrinology. 2014;155(1):287-298. 894 47. Baaten BJ, Tinoco R, Chen AT, Bradley LM. Regulation of Antigen -Experienced T Cells: 895 Lessons from the Quintessential Memory Marker CD44. Front Immunol. 2012;3:23. 896 48. Campbell JJ, Murphy KE, Kunkel EJ, et al. CCR7 expression and memory T cell diversity 897 in humans. J Immunol. 2001;166(2):877-884. 898 49. Kueberuwa G, Gornall H, Alcantar-Orozco EM, et al. CCR7(+) selected gene-modified T 899 cells maintain a central memory phenotype and display enhanced persistence in 900 peripheral blood in vivo. J Immunother Cancer. 2017;5:14. 901 50. Mullen KM, Gocke AR, Allie R, et al. Expression of CCR7 and CD45RA in CD4+ and CD8+ 902 subsets in cerebrospinal fluid of 134 patients with inflammatory and non -inflammatory 903 neurological diseases. J Neuroimmunol. 2012;249(1-2):86-92. 904 51. Unsoeld H, Pircher H. Complex memory T -cell phenotypes revealed by coexpression of 905 CD62L and CCR7. J Virol. 2005;79(7):4510-4513. 906 52. Kohn LA, Hao QL, Sasidharan R, et al. Lymphoid priming in human bone marrow begins 907 before expression of CD10 with upregulation of L-selectin. Nat Immunol. 2012;13(10):963-908 971. 909 53. Ivetic A, Hoskins Green HL, Hart SJ. L-selectin: A Major Regulator of Leukocyte Adhesion, 910 Migration and Signaling. Front Immunol. 2019;10:1068. 911 54. Guiu J, Hannezo E, Yui S, et al. Tracing the origin of adult intestinal stem cells. Nature. 912 2019;570(7759):107-111. 913 55. Hou Q, Huang J, Ayansola H, Masatoshi H, Zhang B. Intestinal Stem Cells and Immune 914 Cell Relationships: Potential Therapeutic Targets for Inflammatory Bowel Diseases. Front 915 Immunol. 2020;11:623691. 916 56. Blasius AL, Barchet W, Cella M, Colonna M. Development and function of murine 917 B220+CD11c+NK1.1+ cells identify them as a subset of NK cells. J Exp Med. 918 2007;204(11):2561-2568. 919 57. Wendel M, Galani IE, Suri-Payer E, Cerwenka A. Natural killer cell accumulation in tumors 920 is dependent on IFN-gamma and CXCR3 ligands. Cancer Res. 2008;68(20):8437-8445. 921 58. Marquardt N, Wilk E, Pokoyski C, Schmidt RE, Jacobs R. Murine CXCR3+CD27bright NK 922 cells resemble the human CD56bright NK -cell population. Eur J Immunol. 923 2010;40(5):1428-1439. 924 59. Yu L, Yang F, Zhang F, et al. CD69 enhances immunosuppressive function of regulatory 925 T-cells and attenuates colitis by prompting IL -10 production. Cell Death Dis. 926 2018;9(9):905. 927 60. Cortes JR, Sanchez -Diaz R, Bovolenta ER, et al. Maintenance of immune tolerance by 928 Foxp3+ regulatory T cells requires CD69 expression. J Autoimmun. 2014;55:51-62. 929 61. Gibbons DL, Spencer J. Mouse and human intestinal immunity: same ballpark, different 930 players; different rules, same score. Mucosal Immunol. 2011;4(2):148-157. 931 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 62. Ghaedi M, Steer CA, Martinez -Gonzalez I, Halim TYF, Abraham N, Takei F. Common -932 Lymphoid-Progenitor-Independent Pathways of Innate and T Lymphocyte Development. 933 Cell Rep. 2016;15(3):471-480. 934 63. Guy-Grand D, Azogui O, Celli S, et al. Extrathymic T cell lymphopoiesis: ontogeny and 935 contribution to gut intraepithelial lymphocytes in athymic and euthymic mice. J Exp Med. 936 2003;197(3):333-341. 937 64. Sharpe AH, Pauken KE. The diverse functions of the PD1 inhibitory pathway. Nat Rev 938 Immunol. 2018;18(3):153-167. 939 65. Jubel JM, Barbati ZR, Burger C, Wirtz DC, Schildberg FA. The Role of PD-1 in Acute and 940 Chronic Infection. Front Immunol. 2020;11:487. 941 66. Pahar B, Lackner AA, Veazey RS. Intestinal double-positive CD4+CD8+ T cells are highly 942 activated memory cells with an increased capacity to produce cytokines. Eur J Immunol. 943 2006;36(3):583-592. 944 67. Overgaard NH, Jung JW, Steptoe RJ, Wells JW. CD4+/CD8+ double -positive T cells: 945 more than just a developmental stage? J Leukoc Biol. 2015;97(1):31-38. 946 68. Nefzger CM, Jarde T, Rossello FJ, et al. A Versatile Strategy for Isolating a Highly 947 Enriched Population of Intestinal Stem Cells. Stem Cell Reports. 2016;6(3):321-329. 948 69. Gracz AD, Fuller MK, Wang F, et al. Brief report: CD24 and CD44 mark human intestinal 949 epithelial cell populations with characteristics of active and facultative stem cells. Stem 950 Cells. 2013;31(9):2024-2030. 951 70. Wang F, Scoville D, He XC, et al. Isolation and characterization of intestinal stem cells 952 based on surface marker combinations and colony -formation assay. Gastroenterology. 953 2013;145(2):383-395 e381-321. 954 71. King JB, von Furstenberg RJ, Smith BJ, McNaughton KK, Galanko JA, Henning SJ. CD24 955 can be used to isolate Lgr5+ putative colonic epithelial stem cells in mice. Am J Physiol 956 Gastrointest Liver Physiol. 2012;303(4):G443-452. 957 72. Lueschow SR, McElroy SJ. The Paneth Cell: The Curator and Defender of the Immature 958 Small Intestine. Front Immunol. 2020;11:587. 959 73. Ouellette AJ. Paneth cells and innate mucosal immunity. Curr Opin Gastroenterol. 960 2010;26(6):547-553. 961 74. Satoh Y. Effect of live and heat-killed bacteria on the secretory activity of Paneth cells in 962 germ-free mice. Cell Tissue Res. 1988;251(1):87-93. 963 75. Visan I. Stem cell-immune cell cross-talk. Nat Immunol. 2019;20(1):1. 964 76. Biton M, Haber AL, Rogel N, et al. T Helper Cell Cytokines Modulate Intestinal Stem Cell 965 Renewal and Differentiation. Cell. 2018;175(5):1307-1320 e1322. 966 77. Sumaria N, Grandjean CL, Silva -Santos B, Pennington DJ. Strong TCRgammadelta 967 Signaling Prohibits Thymic Development of IL -17A-Secreting gammadelta T Cells. Cell 968 Rep. 2017;19(12):2469-2476. 969 78. Goldrath AW, Bogatzki LY, Bevan MJ. Naive T cells transiently acquire a memory -like 970 phenotype during homeostasis-driven proliferation. J Exp Med. 2000;192(4):557-564. 971 79. Cho JH, Kim HO, Ju YJ, et al. CD45-mediated control of TCR tuning in naive and memory 972 CD8(+) T cells. Nat Commun. 2016;7:13373. 973 80. Zikherman J, Doan K, Parameswaran R, Raschke W, Weiss A. Quantitative differences 974 in CD45 expression unmask functions for CD45 in B -cell development, tolerance, and 975 survival. Proc Natl Acad Sci U S A. 2012;109(1):E3-12. 976 81. McNeill L, Salmond RJ, Cooper JC, et al. The differential regulation of Lck kinase 977 phosphorylation sites by CD45 is critical for T cell receptor signaling responses. Immunity. 978 2007;27(3):425-437. 979 82. Trowbridge IS, Thomas ML. CD45: an emerging role as a protein tyrosine phosphatase 980 required for lymphocyte activation and development. Annu Rev Immunol. 1994;12:85-116. 981 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 83. Gross M, Salame TM, Jung S. Guardians of the Gut - Murine Intestinal Macrophages and 982 Dendritic Cells. Front Immunol. 2015;6:254. 983 84. Bain CC, Bravo -Blas A, Scott CL, et al. Constant replenishment from circulating 984 monocytes maintains the macrophage pool in the intestine of adult mice. Nat Immunol. 985 2014;15(10):929-937. 986 85. Farache J, Zigmond E, Shakhar G, Jung S. Contributions of dendritic cells and 987 macrophages to intestinal homeostasis and immune defense. Immunol Cell Biol. 988 2013;91(3):232-239. 989 86. Bain CC, Scott CL, Uronen -Hansson H, et al. Resident and pro -inflammatory 990 macrophages in the colon represent alternative context -dependent fates of the same 991 Ly6Chi monocyte precursors. Mucosal Immunol. 2013;6(3):498-510. 992 87. Dunay IR, Damatta RA, Fux B, et al. Gr1(+) inflammatory monocytes are required for 993 mucosal resistance to the pathogen Toxoplasma gondii. Immunity. 2008;29(2):306-317. 994 88. Shin JS, Greer AM. The role of FcepsilonRI expressed in dendritic cells and monocytes. 995 Cell Mol Life Sci. 2015;72(12):2349-2360. 996 89. Lakhdari O, Yamamura A, Hernandez GE, et al. Differential Immune Activation in Fetal 997 Macrophage Populations. Sci Rep. 2019;9(1):7677. 998 90. Yanez A, Coetzee SG, Olsson A, et al. Granulocyte-Monocyte Progenitors and Monocyte-999 Dendritic Cell Progenitors Independently Produce Functionally Distinct Monocytes. 1000 Immunity. 2017;47(5):890-902 e894. 1001 91. Merad M, Sathe P, Helft J, Miller J, Mortha A. The dendritic cell lineage: ontogeny and 1002 function of dendritic cells and their subsets in the steady state and the inflamed setting. 1003 Annu Rev Immunol. 2013;31:563-604. 1004 92. Wosen JE, Mukhopadhyay D, Macaubas C, Mellins ED. Epithelial MHC Class II 1005 Expression and Its Role in Antigen Presentation in the Gastrointestinal and Respiratory 1006 Tracts. Front Immunol. 2018;9:2144. 1007 93. Jamwal DR, Laubitz D, Harrison CA, et al. Intestinal Epithelial Expression of MHCII 1008 Determines Severity of Chemical, T -Cell-Induced, and Infectious Colitis in Mice. 1009 Gastroenterology. 2020;159(4):1342-1356 e1346. 1010 94. Oetke C, Vinson MC, Jones C, Crocker PR. Sialoadhesin-deficient mice exhibit subtle 1011 changes in B- and T-cell populations and reduced immunoglobulin M levels. Mol Cell Biol. 1012 2006;26(4):1549-1557. 1013 95. Hiemstra IH, Beijer MR, Veninga H, et al. The identification and developmental 1014 requirements of colonic CD169(+) macrophages. Immunology. 2014;142(2):269-278. 1015 96. Fuchs T, Puellmann K, Hahn M, et al. A second combinatorial immune receptor in 1016 monocytes/macrophages is based on the TCRgammadelta. Immunobiology. 1017 2013;218(7):960-968. 1018 97. Beham AW, Puellmann K, Laird R, et al. A TNF -regulated recombinatorial macrophage 1019 immune receptor implicated in granuloma formation in tuberculosis. PLoS Pathog. 1020 2011;7(11):e1002375. 1021 98. Ajuebor MN, Das AM, Virag L, Flower RJ, Szabo C, Perretti M. Role of resident peritoneal 1022 macrophages and mast cells in chemokine production and neutrophil migration in acute 1023 inflammation: evidence for an inhibitory loop involving endogenous IL -10. J Immunol. 1024 1999;162(3):1685-1691. 1025 99. Soehnlein O, Lindbom L. Phagocyte partnership during the onset and resolution of 1026 inflammation. Nat Rev Immunol. 2010;10(6):427-439. 1027 100. Rosales C. Neutrophil: A Cell with Many Roles in Inflammation or Several Cell Types? 1028 Front Physiol. 2018;9:113. 1029 101. Evans JG, Chavez-Rueda KA, Eddaoudi A, et al. Novel suppressive function of transitional 1030 2 B cells in experimental arthritis. J Immunol. 2007;178(12):7868-7878. 1031 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint 102. van de Veen W, Stanic B, Wirz OF, Jansen K, Globinska A, Akdis M. Role of regulatory B 1032 cells in immune tolerance to allergens and beyond. J Allergy Clin Immunol. 1033 2016;138(3):654-665. 1034 103. Blair PA, Norena LY, Flores -Borja F, et al. CD19(+)CD24(hi)CD38(hi) B cells exhibit 1035 regulatory capacity in healthy individuals but are functionally impaired in systemic Lupus 1036 Erythematosus patients. Immunity. 2010;32(1):129-140. 1037 104. Smith TD, Tse MJ, Read EL, Liu WF. Regulation of macrophage polarization and plasticity 1038 by complex activation signals. Integr Biol (Camb). 2016;8(9):946-955. 1039 105. Krishnamoorthy N, Oriss TB, Paglia M, et al. Activation of c-Kit in dendritic cells regulates 1040 T helper cell differentiation and allergic asthma. Nat Med. 2008;14(5):565-573. 1041 106. Rodriguez-Cruz A, Vesin D, Ramon -Luing L, et al. CD3(+) Macrophages Deliver 1042 Proinflammatory Cytokines by a CD3- and Transmembrane TNF-Dependent Pathway and 1043 Are Increased at the BCG-Infection Site. Front Immunol. 2019;10:2550. 1044 107. Strauss L, Mahmoud MAA, Weaver JD, et al. Targeted deletion of PD -1 in myeloid cells 1045 induces antitumor immunity. Sci Immunol. 2020;5(43). 1046 108. Brandum EP, Jorgensen AS, Rosenkilde MM, Hjorto GM. Dendritic Cells and CCR7 1047 Expression: An Important Factor for Autoimmune Diseases, Chronic Inflammation, and 1048 Cancer. Int J Mol Sci. 2021;22(15). 1049 109. Forster R, Davalos -Misslitz AC, Rot A. CCR7 and its ligands: balancing immunity and 1050 tolerance. Nat Rev Immunol. 2008;8(5):362-371. 1051 110. Strunk T, Temming P, Gembruch U, Reiss I, Bucsky P, Schultz C. Differential maturation 1052 of the innate immune response in human fetuses. Pediatr Res. 2004;56(2):219-226. 1053 111. Filardy AA, Ferreira JRM, Rezende RM, Kelsall BL, Oliveira RP. The intestinal 1054 microenvironment shapes macrophage and dendritic cell identity and function. Immunol 1055 Lett. 2023;253:41-53. 1056 112. Faria AM, Weiner HL. Oral tolerance. Immunol Rev. 2005;206:232-259. 1057 113. Bhatta R, Han J, Liu Y, et al. Metabolic tagging of extracellular vesicles and development 1058 of enhanced extracellular vesicle based cancer vaccines. Nat Commun. 2023;14(1):8047. 1059 114. Sabanovic B, Piva F, Cecati M, Giulietti M. Promising Extracellular Vesicle -Based 1060 Vaccines against Viruses, Including SARS-CoV-2. Biology (Basel). 2021;10(2). 1061 .CC-BY-NC 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 15, 2026. ; https://doi.org/10.64898/2026.01.15.699706doi: bioRxiv preprint