Welcome to my talk. My name is Wendy Walker. I am an assistant professor at Texas Tech University Health Sciences Center, El Paso. My talk is entitled exploring changes in peritoneal macrophage composition and function during sepsis. I want to thank the Society of Critical Care Medicine for giving me this opportunity to present our data, as well as for funding in the form of an SCCM wheel award. We have no conflicts of interest to declare. Tissue resident macrophages exist in most tissues throughout the body. They play an important role during health to clean up dead cells, as well as during the response to infection or tissue damage. They express pattern recognition receptors such as the toll-like receptors, which sends pathogen and damage associated molecular patterns that are produced by viruses, microbes, and damaged host cells. Within the peritoneal cavity, two populations of tissue resident macrophages exist. This was shown first in mice and subsequently in people. These are the large peritoneal macrophages, the more predominant population, which express high levels of F480, but low levels of MHC2. They derive from a yolk sac progenitor and renew throughout the lifespan. There's a second population, the small peritoneal macrophages or SPM, that represent about 10 percent of the macrophages here. They express intermediate or low levels of F480, but high levels of MHC2. These derive from monocytes that infiltrate slowly in the steady state to replenish the SPM. In the context of inflammation, these cells dramatically change. The LPM seem to disappear from the peritoneal wash. This is known as the macrophage disappearance reaction. Studies by Paul Kube suggest that they may not be dying, but instead traffic to sites of sterile injury to nibble away at apoptotic cells. At the same time, there's a rapid influx of large numbers of monocytes. Within a few days, these have the capacity to differentiate to greatly increase the numbers of small peritoneal macrophages. This was studied in the context of inflammatory insults, such as thioglycolate and zymosan. A knowledge gap remains, however, because we still don't know how myeloid cells change in the peritoneal cavities, specifically during abdominal sepsis and sterile abdominal surgery. It would be important to identify patterns associated with a functional response that lead to sepsis survival versus a dysfunctional response that without intervention would lead to death of the patient, because these could represent targets for therapeutic manipulation. To study this, we use the mouse cecal ligation and puncture model of sepsis. Mixed gender C57 black six mice underwent a surgery with a midline laparotomy. The cecum was exteriorized. One centimeter was ligated with a suture and then punctured through and through with a 21-gauge needle. The cecum was returned to the abdominal cavity and it was closed in two layers. The animals received buprenorphine SR for analgesia, as well as lactated ringer solution and antibiotics after surgery. As a control, the animals underwent a sterile sham surgery where the cecum was neither ligated nor punctured. I want to mention that this model complies with the guidelines presented in the minimum quality threshold in preclinical sepsis studies. At various time points post-surgery, animals were euthanized, the peritoneal lavage was harvested, and the cells in this liquid fraction were stained with antibodies and identified through the use of flow cytometry. This slide shows the analysis pipeline used to identify different populations of macrophages and monocytes. First, cells were gated, and then cingulates, and then CD45 positive leukocytes. Some cells of non-interest were then excluded, and then the CD11B positive cells were gated. Within this population, we identified the F480 high, MHC2 low, LPM, as well as the F480 intermediate or low, and MHC2 high, SPM. Down here at the bottom, you can see the wash obtained from an animal 18 hours post-CLP, and you can see we also identified within the CD11B positive population, cells that were live 6C high, these represented monocytes. Within the peritoneal cavity of a naive mouse, shown in the gray triangles, there were about one to two million LPM. However, these rapidly disappeared at 18 hours post-CLP, shown in the black circles, or sham surgery shown in the white squares. They were still depleted at 66 hours post-surgery. At the late time point of 14 days post-sham surgery, most animals had seen a return of LPM, but many of the CLP mice still lack these cells. At the same time, while monocytes were absent from the peritoneal cavity of naive mice, they rapidly infiltrated the peritoneum at 18 hours post-CLP or sham surgery, remained abundant at 66 hours post-surgery, and by 14 days post-surgery, most of the sham mice had seen the monocytes greatly depleted while they were still present in most of the CLP mice, though at lower numbers. In regard to the SPM, these were less abundant than the LPM. Other studies had shown that monocytes rapidly differentiate to give rise to SPM within a few days after an inflammatory insult. But in contrast, with these models, we saw only a minor increase in the number of SPM at 18 and 66 hours post-surgery. So only in a subset of the CLP mice did we see a more substantial increase in SPM at the late time point of 14 days post-surgery. These data suggest to us that monocytes may be playing an important function during the acute phase of sepsis, that is, during the first few days, without differentiating to give rise to SPM. We next wanted to distinguish patterns that differed between animals that were destined to survive sepsis versus animals destined to succumb. Daniel Remick and others have shown that serum interleukin 6 is an excellent marker to predict the outcome of CLP sepsis. And we confirmed this in our own lab. We found that animals that died within three days post-CLP all exhibited IL-6 values of greater than 5 ng per ml at 18 hours post-CLP, while animals that survived sepsis had levels lower than this, as well as a few mice that exhibited a very late death. So we used this marker to stratify our cohort. We found that animals that were predicted to survive sepsis had significantly higher numbers of monocytes in their peritoneal cavity in comparison to animals predicted to die. Based on these data, we came up with a hypothesis that greater infiltration of monocytes into the peritoneal cavity aids sepsis survival. We used two methods to test this hypothesis. The first was through the use of CCR2 knockout mice. These animals lack the chemokine receptor that plays a major role in trafficking of mature classical monocytes in response to inflammation. You can see that at 18 hours post-CLP surgery, these CCR2 knockout mice had significantly lower numbers of monocytes in their peritoneal cavity in comparison to wild-type animals that possess the chemokine receptor. So we predicted that these lower numbers of monocytes would be associated with worse sepsis survival. However, contrary to our hypothesis, there was no significant difference in the survival of the two groups, even with large numbers of mice. And in fact, we observed a trend towards slightly improved survival in the CCR2 knockout mice. As a second method to test our hypothesis, we adoptively transferred three million monocytes into the peritoneal cavity. However, these animals shown in the gray line showed no better survival than their counterparts that had undergone CLP without monocyte adoptive transfer. So these data indicate that simply altering the number of monocytes in the peritoneal cavity does not affect sepsis survival. So we came up with a new hypothesis that perhaps sepsis survival is associated with the infiltration of different types of monocytes rather than merely increased numbers. This slide shows how hematopoiesis proceeds during health. Within the bone marrow, hematopoietic stem cells divide and give rise to a balanced mix of lymphocytes as well as myeloid cells. In the context of sepsis or trauma, though, the body goes into emergency myelopoiesis. Division of the hematopoietic stem cells is increased and the balance is shifted to generate more myeloid cells at the expense of generating fewer lymphocytes. Additionally, large numbers of immature myeloid cells are generated and leave the bone marrow in this form, and these have the potential to act as myeloid-derived suppressor cells. We use CD31 as a marker of immature monocytes. We found that animals predicted to survive sepsis had a significantly higher number of these immature CD31-positive monocytes in their peritoneal cavity in comparison to animals predicted to die, while in contrast, when we looked at the CD31-negative mature monocytes, the number of these was similar in animals predicted to survive or predicted to die. We wanted to further investigate the character of these cells. We found that the CD31-positive monocytes expressed higher levels of the CXCR4 chemokine receptor in comparison to their CD31-negative counterparts, and this combination of CD31 and high levels of CXCR4 is consistent with a cell called the transitional pre-monocyte that is the immediate precursor of the mature monocyte. We also found that the CD31-positive cells expressed high levels of intracellular arginase 1, and this would be consistent with an MDSC phenotype. Finally, we wanted to determine the functional importance of these immature monocytes in sepsis survival, and to test this, we used the AMD3100 inhibitor. This blocks the CXCR4-CXCL12 axis, and you can see here that 18 hours post-CLP, animals that were treated with AMD3100 showed a lower number of these immature CD31-positive monocytes within their peritoneal cavity in comparison to animals that were treated with a PBS vehicle, and additionally, treatment with this AMD3100 resulted in a worse sepsis survival in comparison to animals that were treated with PBS, so these data are consistent with the notion that trafficking of these immature CD31-positive cells to the peritoneal cavity aids sepsis survival. In summary, we found the myeloid cells shift dramatically after CLP sepsis as well as after sham abdominal surgery, resulting in the loss of LPM from the peritoneal cavity. In the future, we want to find out what is the destination or fate of these cells as well as the infiltration of monocytes, both mature CD31-negative monocytes mediated by CCL2-CCR2 as well as these immature CD31-positive cells mediated by CXCL12-CXCR4, but the differentiation into SPM seems to be delayed in comparison to other reports that use sterile models of inflammation. We found that animals that are destined to survive sepsis have increased numbers of the CD31-positive immature monocytes trafficking to their peritoneum, and these CD31-positive monocytes do seem to aid sepsis survival. In the future, we want to find out more about the phenotype of these cells and determine whether they are acting as myeloid-derived suppressor cells and whether there might be any way to manipulate their trafficking to help patients with sepsis. I want to thank the people that contributed to this study. A large portion of the study was done by master's students in the graduate program at our minority-serving institution. I also want to thank the Society of Critical Care Medicine for this SCCM Wheel Award, as well as my institution for a seed grant and startup funds and thank you all for listening.

peritoneal macrophage composition

sepsis

tissue resident macrophages

inflammation

monocytes