Pneumonia is the most common infectious cause of hospital admission in the United States and the most common cause of sepsis and septic shock (1). As pneumonia-related sepsis mortality has decreased over time (2), more patients are surviving with deficits in physical ability, cognitive function (3, 4), increased risk of cardiovascular events (5), and early mortality (6, 7). There are no targeted therapies to prevent long-term morbidity and mortality in survivors of sepsis due to pneumonia. Our understanding of mechanisms leading to persistent morbidity and risk of mortality in patients surviving pneumonia is limited.
Animal modeling of pneumosepsis survival is therefore necessary to understand the pathophysiology of long-term sepsis complications. Surprisingly, animal models of antibiotic-treated pneumonia are lacking (8, 9). Models of fecal peritonitis with and without antibiotic therapy, such as cecal ligation and puncture (CLP), are more commonly utilized to study sepsis beyond the acute phase of illness. However, in CLP extrinsic factors such as technical aspects of surgery can lead to limitations in the reproducibility and generalizability of these techniques (10). Surgical interventions, independent of sepsis, may also cause inflammation and brain injury leading to long-term cognitive dysfunction (11). Intrinsic factors such as microbiological composition of the cecum also impact the severity of sepsis and can vary highly with diet, vendor supplying the mice, and differences in housing (12). Thus, experimental results can differ dramatically between labs despite within lab reproducibility. Furthermore, the septic response to infection in the peritoneal cavity, which is sterile at baseline, may be inherently different than infection at other more common sites of infection (lung, urinary tract) that are regularly exposed to microbiota and microbial products. Pneumonia modeling is an advantageous alternative as primary infection can be delivered in a standardized inoculum with a specific pathogen. Furthermore, intranasal administration of a small inoculum mimics the natural microaspiration of pathogenic bacteria from the oropharynx leading to clinical bacterial pneumonia in humans. Allowing time between inoculum administration and antibiotic exposure mimics the clinical development of localized pneumonia with progression to more advanced infection which is often when patients present for evaluation. Clinical relevance is enhanced through the administration of antibiotics to treat acute infection and promote survival.
Pathophysiologic mechanisms of long-term organ dysfunction after sepsis are not well understood. Acute inflammation during sepsis occurs with the simultaneous expression of numerous pro- and anti-inflammatory mediators that promote short-term organ dysfunction and failure (13). Persistence of this inflammatory cascade has been demonstrated in patients hospitalized for pneumonia with elevated circulating interleukin (IL)-6 and IL-10 at hospital discharge and is associated with increased risk of long-term mortality (6). Damage associated molecular patterns (DAMPs) such as S100A8/A9, an endogenous ligand of Toll-like receptor (TLR) 4 and the receptor for advanced glycation end-products (RAGE), have also been demonstrated in circulation both in early and late sepsis indicating a potential role in the pathology of sepsis recovery (13–15). Acute immune tolerance with decreased response to secondary stimuli has been well characterized in experimental and clinical sepsis and is associated with poor short-term outcomes (16). However, the immune response to secondary endotoxin exposure has also been shown to be primed/enhanced in long-term survival of experimental peritonitis (17–19). Tolerant and primed responses to secondary stimulation in sepsis survival may further influence long-term outcome by predisposing to secondary infection or subsequent organ injury.
We hypothesized that survival of bacterial pneumonia would be associated with dynamic, tissue-specific expression of inflammatory mediators. Based on our prior work (17), we also hypothesized that persistent and primed neuroinflammation would be present in association with brain dysfunction in sepsis survival. We developed a highly reproducible model of intranasal Klebsiella pneumoniae pneumosepsis treated with antibiotics to promote long-term survival. We find acute heterogeneous tissue-specific expression of pro-inflammatory cytokines and DAMPs in pneumosepsis mice. We further demonstrate persistent neuroinflammation, myeloid cell infiltration, and primed immune responses in the brain that parallel previously identified phenotypes in CLP. These findings support the use of this non-surgical pneumosepsis model with antibiotic therapy to study long-term complications of pneumonia and sepsis.
MATERIALS AND METHODS
Male C57BL/6J mice 8 to 12 weeks of age were obtained from Jackson Labs. All applicable international, national, and institutional guidelines for the care and use of animals were followed. All procedures performed were in accordance with the ethical standards of the University of Michigan and the National Institutes of Health. The study protocol was approved by the University Committee on the Use and Care of Animals of the University of Michigan (protocol #PRO00007115). Mice were housed in SPF, temperature and humidity controlled environments with a 12-hour day and night cycle. Food and water were available ad libitum.
Mice were randomized by cage (five animals) to treatment group. Due to technical aspects of treatment, no blinding was performed during treatment administration. For tissue culture, gene expression, flow cytometry assays investigators were blinded to treatment groups.
Intranasal Klebsiella pneumoniae
K pneumoniae, strain ATCC 43816, serotype 2, at 1 × 104 colony forming units (CFU) was used for all experiments and prepared as previously described (20). Mice were lightly anaesthetized with ketamine and xylazine. Mice were given 30 μL (15 μL per nostril) of K pneumoniae (infection) or saline (sham) intranasally via pipette tip. Both infection and sham mice were given intraperitoneal ceftriaxone 75 mg/kg once daily starting on day 3 (72 h after infection or saline) for 5 days (Fig. 1A). Mice were weighed on day 3, 5, 7, and 14 to track the development of and recovery from sepsis. In the rare event that a sham-inoculated mouse lost weight, the animal was presumed to have sustained an injury due to multiple intraperitoneal injections and excluded from the study (two mice total). Animals were euthanized if moribund, immobile, and unable to access food or water. Euthanasia was performed with CO2 inhalation in all experiments.
Measurement of tissue colony-forming units (CFU)
Mice were transcardially perfused with ice-cold phosphate-buffered saline (PBS). Whole lungs, spleen and brain were isolated, homogenized, and plated for CFU as previously described (20). CFU were expressed as log10 (CFU/mL+1) to allow for logarithmic transformation of 0 values.
All mice were acclimated to the testing environment for 30 min prior to initiation of behavioral testing. Behavior was analyzed using Ethovision 11.5 software (Noldus Information Technology, Inc, Wagenigen, Netherlands). Based on prior behavioral testing of CLP survivor mice, we expected a coefficient of variation of most behavioral measures of 30%, and so 12 to 15 mice were tested per group to detect a 30% change in outcome with a power of >80%.
Mice underwent one baseline 5-min exploration of a dimly lit (25–30 lux) 36 × 36 cm open field. Fourteen days after treatment (KPA or saline) mice underwent one additional 5-min exploration of the open field. Cumulative distance, cumulative duration, average velocity, and percent change in locomotion from baseline were calculated. Post pneumosepsis changes in each measure were calculated as a fractional change from baseline: (M14 days – Mbaseline)/Mbaseline.
Novel object recognition
Mice underwent a 5-min training period in a dimly lit open field with two identical ethanol cleansed objects. Repeat testing was performed 90 min later with the introduction of a novel object. Objects were cleansed with ethanol between animals. Cumulative distance, cumulative duration of exploration of an object (either center point and/or nose point is within the object zone), frequency of exploration of an object, latency to exploring an object were calculated. The novel object ratio was calculated as: tnovel object/(tnovel object+ tfamiliar object). Animals that did not explore the objects in either the training or the testing trials were excluded from analysis. In each group, one mouse was excluded from analysis due to no exploration of either object.
Open field behavioral testing
Mice were placed in the middle of a brightly lit (200–250 lux) 72 × 72 cm open field. Movement in the arena was recorded for 5 min. Cumulative distance, average velocity, latency to first entrance into the center, frequency of entrances into the center, cumulative distance in the center, and cumulative duration in the center were recorded.
Flow cytometry and cell sorting
Brain single-cell suspensions were prepared as previously described (17). Perfused spleens were dissected, minced, and gently macerated through 40 μm cell strainer to form a single-cell suspension. Strainers were washed with 30 mL of PBS with 1% bovine serum albumin, 2 mM EDTA, 25 mM HEPES. Cells were washed, Fc receptors blocked, and stained with fluorophore-conjugated antibodies prior to analysis on a FACSAria II flow cytometer and cell sorter (BD). Antibodies included anti-CD11b (clone M1/70, BD), anti-CD45 (clone 30-F11, BD), anti-Ly6G (clone 1A8, Biolegend), anti-Ly6C (clone HK1.4, Biolegend). Gating strategy for myeloid cells in the brain and spleen is available in Supplemental data (Figure S1, Supplemental Digital Content 1, http://links.lww.com/SHK/A928).
In vivo inflammatory challenge assay
Two weeks after control or experimental treatment, mice were injected intraperitoneally with 5 mg/kg lipopolysaccharide (E coli O111:B4, Sigma). Three hours later mice were euthanized and perfused with ice-cold PBS.
RNA isolation and gene expression analysis
RNA was isolated from tissues using TRIzol (Ambion), as previously described (17). RT-PCR was performed on a StepOnePlus thermocycler (Applied Biosystems, Foster City, Calif) as previously described (21). IL-17 RNA transcript was rarely detectable by RT-PCR in non-sepsis and unstimulated controls. In cases where this occurred, cycle count (CT) was imputed at 38 so baseline values could be calculated. RNA expression was analyzed using the comparative CT method (22). Results were normalized through Log2 transformation and plotted on a heat map by animal.
Analyses included either ANOVA or unpaired t test testing as indicated in the text, followed by post-hoc testing when appropriate for multiple comparisons (Tukey's multiple comparisons for CFU data, Sidak's multiple comparisons for weight loss data). All figures show mean and standard error unless otherwise specified. Statistical analyses were carried out in Prism (Graphpad).
Pneumosepsis due to K pneumoniae is associated with acute sickness behavior, invasive infection, and death
Intranasal administration of 1 × 104 CFU of K pneumoniae resulted in notable acute sickness behavior, including decreased locomotion and grooming behaviors (data not shown) within 48 to 72 h. At 24 h postinfection, small sub-segmental pulmonary infiltrates are noted on necropsy with progression toward lobar pneumonia and patchy multifocal infiltrates by 48 to 72 h (Fig. 1B). First death occurs within 72 h, with 100% mortality by 11 days postinfection (Fig. 1C). Acutely infected mice demonstrated significant weight loss as compared with sham controls by 72 h after infection (Fig. 1D). K pneumoniae was detected in the lung, spleen, and brain at 24 and 72 h after infection (Fig. 1E). Detection of bacteria in multiple organs distant from the lungs suggests the presence of bacteremia and disseminated infection. Blood was not used as a primary source of culture as it is relatively low biomass and was not a useful indicator of disseminated infection in our prior work with CLP (23).
Generation of a low mortality survival model with antibiotic administration
Antibiotic treatment with ceftriaxone IP reduced 14-day mortality in the infection group to approximately 30% across multiple experiments (Fig. 1C). Sham treatment groups had no significant mortality. Mice surviving infection had marked weight loss with statistically different weight at day 3 through 14 as compared with saline controls (Fig. 1D). Postinfection mice regained weight through day 14 with no difference in weight loss when compared with baseline as measured by 2-way ANOVA with Tukey's correction for multiple comparisons (Fig. 1D). Given that this method may underestimate a statistically significant difference between groups, a paired t test was performed on absolute weights in postinfection mice at day 0 (24.92 ± 0.55 g, n = 28) to day 14 (24.15 ± 0.43 g, n = 28). This showed that mice surviving infection had significantly lower body weight at day 14 (P = 0.011, data not shown). The absolute difference in means was 0.77 g (3% of the baseline body weight) between time points. Antibiotic therapy reduced detectable K pneumoniae CFU in the lungs, spleen, and brain at 14 days (Fig. 1E). While a limited number of mice remained culture positive at a reduced titer in the lungs and brain, most mice completely cleared K pneumoniae infection. Despite significantly lower day 14 body weights in postinfection mice by direct paired comparison, given the overall trajectory of weight regain, small absolute difference in weight between day 0 and day 14, and resolution of detectable CFU in antibiotic-treated mice, these results suggest that antibiotic treatment leads to significant physiologic improvement following acute infection.
Survival of pneumosepsis is associated with persistently altered behavior
A 2-week (14-day) endpoint was selected as a “long-term” as this duration of observation exceeds prior infectious pneumonia models to a significant degree. This time point has also been observed to be critical for persistent neuroinflammation and behavioral change in our prior work (17, 24). Other models of murine polymicrobial sepsis using CLP have demonstrated persistent neurocognitive dysfunction and behavior abnormalities for up to 2 weeks after sepsis (25–27). Behavioral testing revealed that murine pneumosepsis survivors exhibit impaired exploratory locomotor behavior 14 days after infection compared to saline controls (Fig. 2A). Habituation to open field testing led to a decline in total distance traveled in both groups when comparing pre- and postintervention testing. However, the decline in total distance traveled from baseline (prior to infection/saline) to 14 days was significantly greater in pneumonia survivors compared to the effect of repeated testing in controls. In a brightly lit open field, pneumonia survivors move significantly less than do uninfected controls (Fig. 2B). Locomotor activity 14 days after infection is decreased when measured by total time spent moving, indicating that total distance traveled is not solely decreased due to reduced motor velocity (Fig. 2C). When presented with novel and familiar objects, both pneumonia survivors and saline inoculated mice preferentially explore the novel object, indicating that they do not have gross deficits in hippocampal-dependent memory (Fig. 2D).
Acute pneumosepsis and sepsis recovery are associated with enhanced expression of pro-inflammatory mediators and DAMP signaling in the brain and spleen
The inflammatory response to sepsis is highly time and organ specific (28). Given evidence of persistent behavioral change in survivors, we sought to characterize both the systemic (splenic) and brain inflammatory response in this model. In the brain (Fig. 3A, left panel) and spleen (Fig. 3A, right panel), IL-1β, the neutrophil chemoattractant KC/CXCL1, the lymphocyte chemoattractant IP-10/CXCL10 and DAMP S100A8 were elevated at 72 h postinfection but prior to initiation of antibiotics. Cytokine responses were heterogeneous across body compartments. Splenic gene expression revealed significant elevations in IL-6, MCP-1/CCL2, and IL-17 which were not observed in the brain. There was a small but statistically significant increase in TNFα expression observed in the brain, but not the spleen.
Persistent neuroinflammation has been proposed to contribute to chronic brain dysfunction in sepsis survivors. We have previously demonstrated persistent expression of inflammatory cytokines and chemokines in the murine brain 14 days after CLP (17, 29). Basal gene expression 14 days after K pneumoniae infection demonstrates a small but statistically significant increase in TNFα and KC in the brains of pneumosepsis mice as compared with sham control (Fig. 3B, left panel). In the spleen, S100A8 expression was increased nearly 20-fold, while no changes in other cytokines or chemokines were observed (Fig. 3B, right panel).
Brain cytokine expression is primed in response to secondary LPS stimulation in pneumosepsis survivors
Post-sepsis immune reprogramming has been described in both animal models and patients with acute sepsis (30). We sought to characterize the long-term response to secondary immune challenge with intraperitoneal LPS administration. We found a significant and broadly primed response to LPS with enhanced expression of IL-1β, TNFα, S100A8, MCP-1, KC, IL-17, and IP-10 in the brains of pneumosepsis survivors as compared with LPS stimulated sham controls (Fig. 4A). There was also modest but statistically significant inflammatory priming in spleen IL-1β, IP-10 and MCP-1 (Fig. 4B). S100A8 expression remained significantly elevated following LPS challenge in pneumosepsis survivor mice compared to sham controls.
Long-term infiltration of myeloid cells in the brain occurs independent of changes in systemic myeloid populations
Murine CLP and streptococcal pneumonia are associated with infiltration of neutrophils and monocytes into the brain that persist weeks after sepsis has resolved (17, 31). Given enhanced chemokine expression in the brain and spleen during pneumosepsis, we measured myeloid cell recruitment to end organs at 14 days postinfection. We found a significant increase in the proportion of infiltrating neutrophils and monocytes in whole brain single-cell suspensions at 14 days in pneumosepsis mice as compared with sham controls (Fig. 5, A and B). This occurred in the absence of a significant change in the proportion of myeloid subsets in the spleen (Figure S2, Supplemental Digital Content 2, http://links.lww.com/SHK/A929).
Pneumonia remains a significant burden to patients and the healthcare system. As more patients survive sepsis due to pneumonia, the potential burden of this illness grows. To address this issue, we must first understand how long-term complications develop in organs distant to the initial site of infection. Tissue-specific consequences of acute infection have been examined in experimental models, but often focus on inflammatory responses at the initial site of infection and during the acute phase of illness (32, 33). Here, we utilize a non-surgical, mouse model of antibiotic-treated pneumosepsis survival to examine late neurocognitive dysfunction and tissue-specific inflammation.
Using a well-studied isolate of K pneumoniae(34, 35), we have produced a model with highly morbid acute sepsis. Disseminated infection and systemic inflammation are subsequently rescued with antibiotic therapy allowing the study of surviving mice. Despite weight regain and resolution of infection 2 weeks after sepsis, mice display abnormal exploratory behavior in association with persistently elevated TNFα expression in the brain, as well as primed production of numerous cytokines/chemokines in association with myeloid cell recruitment.
Elevations in whole brain cytokines and chemokines have been noted in several small studies of patients who die of sepsis and are at high risk of acute brain dysfunction (23, 36). In rodents, administration of IL-1β and TNFα systemically or directly into the brain leads to development of depressed behavior responses including decreased motor activity (37). We have previously demonstrated that persistent expression of TNFα and other cytokines/chemokines weeks after murine polymicrobial sepsis is associated with increased fear behavior (29). Persistent expression of inflammatory mediators can therefore reasonably be hypothesized to contribute to our observed deficits in behavior. Expression of the neutrophil chemotactic genes KC/CXCL1 and S100A8 were also enhanced in early (72 h) sepsis and in recovery (14 days). This was associated with persistent neutrophil recruitment to the brain. Recruitment of inflammatory cells to bystander organs distant from the initial site of infection may be a mechanism through which tissue-specific inflammation is perpetuated. In a recent evaluation of non-antibiotic-treated Streptococcal pneumoniae lung infection, it was demonstrated that neutralization of early monocyte, but not neutrophil, recruitment to the brain reduced measures of neurocognitive dysfunction using a Morris water maze at 2 weeks (31). However, a direct link between recruited inflammatory cells and persistent cytokine expression with persistent brain dysfunction in our model requires further investigation.
Primed cytokine production has been observed in splenic (18) and bone marrow monocytes (19) of CLP survivors. Our lab has previously demonstrated enhanced TNFα production in brain microglia 2 weeks after CLP (17). In this study brain expression of the cytokines IL-1β, TNFα, and IL-17 was primed after IP LPS injection. This primed program of immunity is in contrast to the heavily investigated endotoxin tolerant program observed in acute sepsis in mice and humans (33, 38). This discrepancy is likely due to the fact that immune tolerance is likely a consequence of the acute phase of cytokine storm and is prominent in early time points of highly morbid models. In addition, the use of high mortality models of sepsis selects for surviving mice to be those with limited systemic illness. Priming may therefore represent a unique immune program that is a result of recent, but not active, severe inflammation. Priming of the brain immune response may subsequently lead to locally enhanced cytokine production. Neuroinflammatory priming is known to be associated with progression of chronic neurodegenerative disease in murine models (39–42) and represents a mechanism through which the brain may remain vulnerable to persistent dysfunction in sepsis survival. More broadly, priming of the systemic immune response may increase susceptibility of other tissues (e.g., liver, kidney, bone marrow) to secondary inflammatory insults.
Mechanisms driving primed immune responses and persistent inflammation remain unclear. Certainly, an increase in the number of inflammatory myeloid cells in the brain in sepsis survival could result in enhanced expression of inflammatory cytokines at baseline and upon secondary stimulation. Persistent and primed tissue chemokine expression may drive further myeloid cell infiltration despite resolution of infection. We did however observe priming of inflammatory mediators in the spleen without changes in proportions of inflammatory myeloid cells. Resident immune cells can also be poised toward a pro-inflammatory state during recovery from sepsis, as we have previously demonstrated primed TNFα production mediated by the DAMP S100A8/A9 in microglia 2 weeks after CLP (17). Chronic systemic elevations in other DAMPs, such as high mobility group box 1 protein (HMGB1), have also been implicated in priming of splenic inflammatory monocytes after CLP (18). We speculate that cell-specific primed immune responses occur as a result of trained immunity, a type of innate immune reprogramming characterized by changes in chromatin leading to persistently enhanced cytokine production in response to secondary stimuli (43).
Our study is limited in that we have focused primarily on gene expression profiles and not protein expression. However, gene expression profiles of peripheral blood mononuclear cells have been used extensively to characterize human sepsis and provide relevant information about morbidity and mortality (13). We conclude that primed immune phenotypes may be due to cell-specific enhanced cytokine production; however, our use of whole tissue does not allow for cell-specific analysis. While infiltrating myeloid cells may be a primary source of systemic cytokines, in the brain multiple residents cells are known to produce IL-1β and TNFα including microglia, neurons, astrocytes, and even endothelial cells in mice and humans (40, 44–46). Furthermore, we currently restricted in our ability to analyze infiltrating cells as they represent a very small fraction of the entire brain population that is recovered from our assays. Future studies will need to focus on cell-specific gene expression and protein production to further evaluate mechanisms of persistent inflammation and priming in sepsis survival. Neutralization of leukocyte trafficking to the brain during early and late sepsis may be useful to evaluate the role of infiltrating myeloid cells in driving changes in persistent inflammation and behavior. We have previously identified persistent neuroinflammation for up to 50 days after murine CLP (24). Our current study only focused on 14-days after infection. Observations at later time points (e.g., 8 or 12 weeks) will be important in future studies to determine if persistent neuroinflammation resolves and whether resolution is associated with improvement in behavioral outcomes.
Finally, we observed a deficit in locomotor behavior which may represent a persistence of sickness-type behavior, though mice in this study appear to recover from systemic illness. Others have shown that CLP survivor mice do not have persistent intrinsic motor deficits, suggesting that our observed deficits are not due to simply to a primary mobility issue (47). Given that the open field test for anxiety-like behavior relies on locomotion, we conservatively interpret these behaviors as a deficit in exploratory behavior rather than a defect in affective behavior. We did observe significant changes in weight loss and regain through 14 days in the postinfection group as compared to controls in association with behavioral abnormalities. As we did control for food intake per cage, we are unable to comment on any contribution of differences in feeding to the observed behavioral deficits. As feeding may impact performance on behavioral tasks, food intake is an important variable to measure and control in future studies. It should be noted that despite resolution of infection in the majority of mice, however there were detectable CFU in small portion of the brains (2/8) and lungs (4/13) of K pneumoniae mice at 2 weeks. While this could represent persistent infection and a mechanism of sickness-like behavior, the majority of these mice still had a multilog reduction in CFU at 14 days compared to 72 h. Overall, we conclude these findings as indicators of resolving infection and in the presence of weight regain are unlikely to represent failed treatment.
In this novel model of antibiotic-treated K pneumoniae pneumosepsis, we found that survival of sepsis is associated with persistent neuroinflammation, brain infiltration of myeloid cells, neuroinflammatory priming, and decreased exploratory locomotor behavior. Findings of this model parallel observations in models of polymicrobial sepsis survival and support a common pathway to chronic brain dysfunction in sepsis survival. This reproducible, nonsurgical model of pneumosepsis and recovery is a valuable resource as an adjunct to commonly used models of peritoneal sepsis and will improve the rigor and reproducibility of murine sepsis studies.
The authors acknowledge the Michigan Multi-Disciplinary Intensive Care Research Workgroup (MICReW) for their support. Mouse image in Figure 1 was used through the creative commons license (Iconic, thenounproject.com)
1. McDermott KW, Elixhauser A, Sun R. Trends in Hospital Inpatient Stays in the United States, 2005–2014. 1–18, 2017.
2. Stevenson EK, Rubenstein AR, Radin GT, Wiener RS, Walkey AJ. Two decades of mortality trends among patients with severe sepsis
: a comparative meta-analysis. Crit Care Med
3. Davydow DS, Hough CL, Levine DA, Langa KM, Iwashyna TJ. Functional disability, cognitive impairment, and depression after hospitalization for pneumonia
. Am J Med
4. Shah FA, Pike F, Alvarez K, Angus D, Newman AB, Lopez O, Tate J, Kapur V, Wilsdon A, Krishnan JA, et al. Bidirectional relationship between cognitive function and pneumonia
. Am J Respir Crit Care Med
5. Corrales-Medina VF, Alvarez KN, Weissfeld LA, Angus DC, Chirinos JA, Chang C-CH, Newman A, Loehr L, Folsom AR, Elkind MS, et al. Association between hospitalization for pneumonia
and subsequent risk of cardiovascular disease. JAMA
6. Yende S, D’Angelo G, Kellum JA, Weissfeld L, Fine J, Welch RD, Kong L, Carter M, Angus DC. Inflammatory markers at hospital discharge predict subsequent mortality after pneumonia
. Am J Respir Crit Care Med
7. Myint PK, Hawkins KR, Clark AB, Luben RN, Wareham NJ, Khaw K-T, Wilson AM. Long-term mortality of hospitalized pneumonia
in the EPIC-Norfolk cohort. Epidemiol Infect
8. Mehrad B, Standiford TJ. Use of animal models in the study of inflammatory mediators of pneumonia
. ILAR J
9. Mizgerd JP, Skerrett SJ. Animal models of human pneumonia
. Am J Physiol Lung Cell Mol Physiol
10. Lewis AJ, Seymour CW, Rosengart MR. Current murine models of sepsis
. Surg Infect (Larchmt)
11. Terrando N, Eriksson LI, Ryu JK, Yang T, Monaco C, Feldmann M, Jonsson Fagerlund M, Charo IF, Akassoglou K, Maze M. Resolving postoperative neuroinflammation and cognitive decline. Ann Neurol
12. Laukens D, Brinkman BM, Raes J, De Vos M, Vandenabeele P. Heterogeneity
of the gut microbiome in mice: guidelines for optimizing experimental design. FEMS Microbiol Rev
13. Tang BM, Huang SJ, McLean AS. Genome-wide transcription profiling of human sepsis
: a systematic review. Crit Care
14. Gao S, Yang Y, Fu Y, Guo W, Liu G. Diagnostic and prognostic value of myeloid-related protein complex 8/14 for sepsis
. Am J Emerg Med
15. van Zoelen MAD, Vogl T, Foell D, Van Veen SQ, van Till JWO, Florquin S, Tanck MW, Wittebole X, Laterre P-F, Boermeester MA, et al. Expression and role of myeloid-related protein-14 in clinical and experimental sepsis
. Am J Respir Crit Care Med
16. Boomer JS, Green JM, Hotchkiss RS. The changing immune system in sepsis
: is individualized immuno-modulatory therapy the answer? Virulence
17. Denstaedt SJ, Spencer-Segal JL, Newstead MW, Laborc K, Zhao AP, Hjelmaas A, Zeng X, Akil H, Standiford TJ, Singer BH. S100A8/A9 drives neuroinflammatory priming and protects against anxiety-like behavior after sepsis
. J Immunol
18. Valdés-Ferrer SI, Rosas-Ballina M, Olofsson PS, Lu B, Dancho ME, Ochani M, Li JH, Scheinerman JA, Katz DA, Levine YA, et al. HMGB1 mediates splenomegaly and expansion of splenic CD11b+ Ly-6C(high) inflammatory monocytes in murine sepsis
survivors. J Intern Med
19. Bomans K, Schenz J, Sztwiertnia I, Schaack D, Weigand MA, Uhle F. Sepsis
induces a long-lasting state of trained immunity in bone marrow monocytes. Front Immunol
20. Deng JC, Cheng G, Newstead MW, Zeng X, Kobayashi K, Flavell RA, Standiford TJ. Sepsis
-induced suppression of lung innate immunity is mediated by IRAK-M. J Clin Invest
21. Ballinger MN, Newstead MW, Zeng X, Bhan U, Horowitz JC, Moore BB, Pinsky DJ, Flavell RA, Standiford TJ, Theodore J. TLR signaling prevents hyperoxia-induced lung injury by protecting the alveolar epithelium from oxidant-mediated death. J Immunol
22. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative CT method. Nat Protoc
23. Singer BH, Dickson RP, Denstaedt SJ, Newstead MW, Kim K, Falkowski NR, Erb-Downward JR, Schmidt TM, Huffnagl GB, Standiford TJ. Bacterial dissemination to the brain in sepsis
. Am J Respir Crit Care Med
24. Singer BH, Newstead MW, Zeng X, Cooke CL, Thompson RC, Singer K, Ghantasala R, Parent JM, Murphy GG, Iwashyna TJ, et al. Cecal ligation and puncture results in long- term central nervous system myeloid inflammation. PLoS One
25. Barichello T, Martins MR, Reinke A, Constantino LS, Machado RA, Valvassori SS, Moreira JCF, Quevedo J, Dal-Pizzol F. Behavioral deficits in sepsis
-surviving rats induced by cecal ligation and perforation. Braz J Med Biol Res
26. Gao R, Li G, Yang R, Yuan H, Zhang S, Yang R, Yuan H, Yuan H, Zhang S, Zhang S. Hippocampal β2-microglobulin mediates sepsis
-induced cognitive impairment. Mol Med Rep
27. Barichello T, Martins MR, Reinke A, Feier G, Ritter C, Quevedo J, Dal-Pizzol F. Cognitive impairment in sepsis
survivors from cecal ligation and perforation. Crit Care Med
28. Denstaedt SJ, Singer BH, Standiford TJ. Sepsis
and nosocomial infection: patient characteristics, mechanisms, and modulation. Front Immunol
29. Singer BH, Newstead MW, Zeng X, Cooke CL, Thompson RC, Singer K, Ghantasala R, Parent JM, Murphy GG, Iwashyna TJ, et al. Cecal ligation and puncture results in long-term central nervous system myeloid Inflammation. PLoS One
30. Hotchkiss RS, Karl IE. The pathophysiology and treatment of sepsis
. N Engl J Med
31. Andonegui G, Zelinski EL, Schubert CL, Knight D, Craig LA, Winston BW, Spanswick SC, Petri B, Jenne CN, Sutherland JC, et al. Targeting inflammatory monocytes in sepsis
-associated encephalopathy and long-term cognitive impairment. JCI insight
3 (9):pii: 99364, 2018.
32. Roquilly A, McWilliam HEG, Jacqueline C, Tian Z, Cinotti R, Rimbert M, Wakim L, Caminschi I, Lahoud MH, Belz GT, et al. Local modulation of antigen-presenting cell development after resolution of pneumonia
induces long-term susceptibility to secondary infections. Immunity
33. Muenzer JT, Davis CG, Chang K, Schmidt RE, Dunne WM, Coopersmith CM, Hotchkiss RS. Characterization and modulation of the immunosuppressive phase of sepsis
. Infect Immun
34. Greenberger MJ, Strieter RM, Kunkel SL, Danforth JM, Goodman RE, Standiford TJ. Neutralization of IL-10 increases survival in a murine model of Klebsiella pneumonia
. J Immunol
35. Greenberger MJ, Strieter RM, Kunkel SL, Danforth JM, Laichalk LL, McGillicuddy DC, Standiford TJ. Neutralization of macrophage inflammatory protein-2 attenuates neutrophil recruitment and bacterial clearance in murine Klebsiella pneumonia
. J Infect Dis
36. Warford J, Lamport A-C, Kennedy B, Easton AS. Human brain chemokine and cytokine expression in sepsis
: a report of three cases. Can J Neurol Sci
37. Dantzer R, O’Connor JC, Freund GG, Johnson RW, Kelley KW. From inflammation to sickness and depression: when the immune system subjugates the brain. Nat Rev Neurosci
38. Boomer JS, To K, Chang KC, Ii SDJ, Kreisel D. Immunsuppression in patients who die of sepsis
and multiple organ failure. JAMA
39. Fonken LK, Frank MG, Kitt MM, D’Angelo HM, Norden DM, Weber MD, Barrientos RM, Godbout JP, Watkins LR, Maier SF. The alarmin HMGB1 mediates age-induced neuroinflammatory priming. J Neurosci
40. Hennessy E, Griffin ÉW, Cunningham C. Astrocytes are primed by chronic neurodegeneration to produce exaggerated chemokine and cell infiltration responses to acute stimulation with the cytokines IL-1β and TNF-α. J Neurosci
41. Niraula A, Sheridan JF, Godbout JP. Microglia priming with aging and stress. Neuropsychopharmacology
42. Wendeln A-C, Degenhardt K, Kaurani L, Gertig M, Ulas T, Jain G, Wagner J, Häsler LM, Wild K, Skodras A, et al. Innate immune memory in the brain shapes neurological disease hallmarks. Nature
43. Dominguez-Andres J, Netea MG. Long-term reprogramming of the innate immune system. J Leukoc Biol
105 (2):329–338, 2019.
44. Fu HQ, Yang T, Xiao W, Fan L, Wu Y, Terrando N, Wang TL. Prolonged neuroinflammation after lipopolysaccharide exposure in aged rats. PLoS One
45. Gahring LC, Carlson NG, Kulmar RA, Rogers SW. Neuronal expression of tumor necrosis factor alpha in the murine brain. Neuroimmunomodulation
46. Vitkovic L, Bockaert J, Jacque C. ‘Inflammatory’ cytokines: neuromodulators in normal brain? J Neurochem
47. Chavan SS, Huerta PT, Robbiati S, Valdes-Ferrer SI, Ochani M, Dancho M, Frankfurt M, Volpe BT, Tracey KJ, Diamond B. HMGB1 mediates cognitive impairment in sepsis
survivors. Mol Med