Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 20.
Published in final edited form as: Exp Neurol. 2024 Dec 26;385:115130. doi: 10.1016/j.expneurol.2024.115130

Unique Infrared Thermographic Profiles and Altered Hypothalamic Neurochemistry Associated with Mortality in Endotoxic Shock

Sean N Curtis a, Catherine A Mayer a, Tracey L Bonfield b,c, Thomas M Raffay a, Juliann M DiFiore a, Richard J Martin a, Adriana C Hoffman a, Michael A Folz d, Ryan W Bavis e, Mathias Dutschmann f, Peter M MacFarlane a,*
PMCID: PMC12918751  NIHMSID: NIHMS2143143  PMID: 39732274

Abstract

Neonatal sepsis results in significant morbidity and mortality, but early detection is clinically challenging. In a neonatal rat model of endotoxic shock, we identified unique infrared thermographic (IRT) profiles in skin temperature that could identify risk of later mortality. Ten-day old rats were placed in a thermally stable isolette and IRT images of cranial (TCR), scapula (TSC) and rump (TRU) skin temperature were obtained continuously for 8h following an intraperitoneal injection of LPS (or saline). LPS resulted in ~74% mortality (designated as non-survivors, LPSNS) between 4.5–7.5 h post-injection. LPSNS and LPSS rats displayed hypothermic tendencies with TCR, TSC and TRU decreasing at ~80–100 min (T80–100) post-injection. Compared to LPSS rats, the hypothermia of LPSNS rats was slightly earlier (T80), more severe, and failed to recover. The TCR, TSC and TRU of LPSS rats fully recovered by 4 h (T240) post-injection. In separate rats, hypothalamic microglia and ECM expression at T240 post-injection were increased in putative identified LPSNS rats and negatively correlated with IR temperatures. IRT could be a useful early identifier of infants at risk of death from endotoxic shock, which may be related to early failure of CNS thermogenic mechanisms mediated by unique hypothalamic changes in inflammatory (microglia) and ECM neurochemistry.

Keywords: Endotoxic shock, neonates, hypothalamus, microglia, thermal imaging

Introduction:

Sepsis is a dysregulated host response to infection leading to life-threatening organ dysfunction (Singer et al. 2016). With an incidence of almost 3,000 cases per 100,000 live births, sepsis is the number one cause of neonatal morbidity and mortality worldwide (Lawn et al. 2005; Fleischmann et al. 2021). The early identification of sepsis can significantly attenuate morbidity and mortality, but early detection of sepsis onset remains a significant clinical challenge (Rivers et al. 2001) largely because of variability in the physical presentation of neonatal sepsis, wide inter-observer interpretation of these physical signs, and imprecise diagnostic testing (Iroh Tam and Bendel 2017). These are compounded in preterm neonates who have a 50% mortality rate from sepsis (Stoll et al. 2020). For this reason, efforts to identify a reliable an accurate early identification marker of sepsis in the neonate could significantly improve morbidity and mortality rates.

The interplay between sepsis and thermoregulation has been known for some time (Inghammar and Sunden-Cullberg 2020; Thomas-Rüddel et al. 2021), but scarcely studied in the context of assisting with the early identification of sepsis. In both animal and human clinical studies, temperature instability and hypothermia have been demonstrated as physiologic disturbances in the neonate highly specific for sepsis and sepsis-induced mortality (Hofer et al. 2012; Lagerspetz and Väätäinen 1987). Currently employed methods of detecting temperature disturbances in neonates include point-of-care axillary probes and contact thermometry in which a probe measures the temperature of a single point on the body of the patient (MacRae et al. 2018). Both techniques have significant clinical limitations. Infrared thermography (IRT) is a technique in which skin surface thermal patterns can be measured continuously and non- invasively over the entire surface of a patient in real time and can be used to obtain thermographic profiles (Jones 1998). As a novel temperature monitoring methodology, IRT can detect subtle changes in surface temperature. It has been used to study cancer, peripheral vascular disease, trauma, wound management (Kateb et al. 2009; Katz et al. 2008; Mital and Scott 2007), and has shown some promise as a means of identifying necrotizing enterocolitis (NEC) in premature neonates (Knobel et al. 2011). Prior studies in adults (Amson et al. 2020; Kazune et al. 2023) have demonstrated the utility of IRT for identification of risk of later sepsis-induced mortality, but studies on neonates are lacking. Early detection of the most subtle changes in surface temperatures using IRT may be a major step forward in identification of infants with sepsis, which may be a novel approach to promote early intervention and prevention of later morbidity and mortality.

Sepsis induces the endogenous release of pyrogens and cryogens which serve as mediators of fever and hypothermia respectively (Kluger 1991). The preoptic region of the hypothalamus and associated nuclei coordinate central and peripheral thermal inputs in mammalian temperature regulation. These regions such as the medial preoptic area (MPA), median preoptic nucleus (MnPO), and anteroventral preoptic nucleus (AVPO) are critical in modulating thermoregulatory control under normal physiological conditions. Their role in modulating temperature responses during sepsis are also likely but have not been explored. Several studies have shown augmented microglial responses in vasopressor regions was associated with increased microglia activity (da Costa et al. 2021; Santos-Junior et al. 2018), suggesting a role in modulation of neuroinflammatory responses during sepsis as has been demonstrated in other brain regions in both humans (Lemstra et al. 2007; Westhoff et al. 2019) and murine models (Bozadas et al. 2019; Michels et al. 2020; Yang et al. 2022). Multiple components of the extracellular matrix (ECM) have also been implicated in modulation of microglia activity in the injured CNS through effects on local immune activation (Johnson et al. 2018; Sorokin 2010). We showed previously that increased microglia expression was associated with increased deposition of several ECM components in brainstem respiratory and autonomic control regions (MacFarlane et al. 2016; Stryker et al. 2018). ECM deposition also plays a pivotal role in repair of the CNS by modulating microglia activity (Rolls et al. 2008). Conversely, microglia can also modulate ECM signaling (Liu et al. 2021). Given the complex interplay between microglia and ECM in the CNS response during inflammation, we used a lipopolysaccharide (LPS) model of neonatal endotoxic shock to determine whether mortality was associated with early changes in expression of microglia and major ECM components (specifically, hyaluronan, HA and aggrecan, Agg) in thermoregulatory regions of the hypothalamus and also whether there were correlations with IRT profiles. We hypothesized that: 1) infrared thermography can detect early thermoregulatory changes associated with neonatal endotoxemia, 2) early expression of unique IRT profiles distinguishes between rats in which LPS would be lethal (or not), 3) the lethality of LPS exposure is associated early changes in microglia and ECM deposition in the MPA, MnPO, and AVPO, and 4) whether hypothalamic microglia and ECM are negatively correlated with IR thermographic profiles.

Materials and Methods

All procedures were carried out in accordance with the National Institute of Health (NIH) guidelines for care and use of laboratory animals and were approved by the Animal Care and Use Committee at Case Western Reserve University (#2014–0051). Time-pregnant Sprague-Dawley rats were purchased from a commercial vendor (Charles River) and were later observed to give birth in the animal facility of the institution; experiments were performed on 10-day old male rat pups.

LPS injections

Rat pups were removed from the dam, weighed, and randomly divided into treatment groups that received an IP injection of saline or LPS (E. coli; 0.1mg/kg). LPS dose was based on our prior study (Rourke et al. 2016). Each rat from a given litter was placed in the prone position on a thermal pad in a temperature controlled isolette for ambient temperature stability (28.5°C), but were separated from one another by dividers creating (~2.5” × 2.5”) individual containment areas (Fig. 1). Rats were allowed a 20 min adjustment period before receiving an IP injection of LPS, or equal volume of saline.

Fig. 1.

Fig. 1.

Open in a new tab

Schematic of the experimental timeline for Experiment 1 (Exp. 1; full IR profiling of LPSS and LPSNS) and Experiment 2 (Exp. 2; correlation and regression analysis of hypothalamic microglia and ECM neurochemistry). Rats were removed from the nest and nursing dam, placed within dividers on a heat pad inside a temperature-controlled neonatal isolette for 20 minutes to equilibrate before injection of LPS or saline (A). The IR camera was manually held above each rat at an equidistant height (the under-side of the isolette lid) for each experiment for IR images. Images were acquired 10 minutes prior (T−10) to injection, then at 10 minute intervals for 2 h, then 30 minute intervals for up to 8 h post injection (Exp. 1). Rats in Exp. 2 underwent the same imaging, but were euthanized at T240 (specifically at the time when IRT profiles of LPSS and LPSNS were statistically different from one another in Exp. 1; see methods) for hypothalamic analysis of microglia and ECM.

Experiment 1: Determining Unique Infrared Thermographic (IRT) Profiles Following LPS

All IRT images of cranial (TCR), scapula skin (TSC), and rump (TRU) were obtained in saline (N=9) and LPS (N=31) treated rats using a pre-calibrated FLIR E4 handheld thermal camera. These three regions were chosen to encompass anterior, thoracic, and posterior regions. At each region, three consecutive thermal images were obtained (and averaged) at baseline and then at 10 min intervals for 2 h following injection, at which point images were obtained every 30 min until 8 h post-injection or death, whichever came first. Our prior study determined that LPS resulted in ~50% mortality with death occurring between 4–7.5 h post-injection of LPS (rats designated as LPS-non-survivors, LPSNS) (Rourke et al. 2016). Any rat surviving by 8 h was placed back with the dam and monitored for another 72 h. Given the sensitive nature of these experiments in which death in a subgroup of rats was an anticipated experimental end-point, all efforts were made to implement interventional euthanasia when imminent death was visibly obvious with guidance from a veterinarian.

Experiment 2: Early Changes in Hypothalamic Neurochemistry in LPSS and LPSNS

We next aimed to determine whether it was possible to detect changes in hypothalamic neurochemistry (microglia and ECM) in LPS rats prior to subsequent mortality specifically at the time when IRT profiles of LPSS and LPSNS rats could be distinguished from one another (i.e. at T240). Therefore, the experiment was repeated in an additional 28 rats to obtain IRT profiles up to T240 minutes post-LPS injection at which time all rats were euthanized for analysis of the hypothalamus. Since it is unknown which LPS rats would have survived and which would not have, the IRT profiles at T240 were used for comparison with rats obtained from Experiment 1. From these IRT profiles, individual rats could then be putatively assigned to LPSS or LPSNS based on whether their IRT temperatures at 4 h (T240) post-injection was within 1SD of the means for LPSS and LPSNS obtained in Experiment 1, above. The T240 time-point was chosen based on the unique IRT profiles of respective LPSS and LPSNS rats described below (see results). Since the IRT profiles of TCR, TSC and TRU in response to LPS followed similar time-courses (See Fig. 2), we present subsequent data based on the profile of TCR also because it likely has greater clinical practicality than measurements made from the TSC and TRU. Using this criterion, we putatively identified n=18 LPSNS and n=9 LPSS rats. An additional n=8 rats were treated with saline.

Fig. 2.

Fig. 2.

Open in a new tab

A) Cranial (TCR), B) scapula (TSC), and C) rump (TRU) skin surface temperatures of P10 rats obtained over time via IRT following injection of either saline (n=9) or LPSS (n=8) and LPSNS (n=23). TCR decreased (vs saline rats) at T80 and T100 for LPSS and LPSNS rats, respectively (A). LPSNS rats never recovered TCR, whereas LPSS recovered TCR (not significantly different from saline rats) by T240 (shaded area). T240 was chosen for Experiment 2 since it represented the time-point when TCR of LPSS was not significantly different from saline and significantly different from LPSNS rats. TSC and TRU followed similar time-domains of IRT thermographic profiles as TCR for respective treatment groups. Solid line indicates LPSS sig. diff from saline group; short dashed line indicates saline sig. diff. from LPSNS group; long dashed line indicates LPSNS sig. diff. from LPSS group at corresponding time-point; p<0.05. Values are mean ± 1SD. Kaplan-Meier curve of P10 rats depicting drop-off in LPSNS rats is also shown (D). Non-survivors (74.2%, pie chart) began to expire at 4.5 h (T270) post-injection, and all had expired by 7.5 h (T450) post-injection.

At the T240 time point, rats were euthanized (urethane overdose), perfused with cold PBS then 4% paraformaldehyde (PFA). Brains were removed and postfixed for 4 h in PFA and then submerged in 30% sucrose (72 h). Midbrains were removed and sectioned (20μm, Leica CM 1850 cryostat). Immunohistochemistry for Iba-1 (microglia), aggrecan, hyaluronan binding protein (HA) and HAS1 (hyaluronan synthase-1) was then performed in MPA, MnPO, AVPO for semi-quantification via fluorescence densitometry within a region of interest (ROI, 100 pixels in diameter) in each brain region using ImageJ. Microglia cell bodies (Iba-1 (MacFarlane et al. 2016)) were counted manually by 2 experienced and blinded investigators; the average pixel intensity within the circle was also measured for densitometric analysis. Sections were blocked with 5% goat serum, 0.1% BSA and 1x PBS blocking buffer for 1–2 h and incubated in primary antibodies overnight at 4°C (HAS1: 1:100, BIOSSUSA; HA: Biotinylated hyaluronan binding protein from bovine nasal cartilage (1:100, Millipore); Agg: sections were pre-digested with chondroitinase ABC (Sigma Aldrich) in chondroitinase buffer (0.1M Tris-HCl, 0.03M sodium acetate, pH 8) for 1 h at 37°C). Sections were then incubated in respective secondary antibodies for 2 h at room temperature (Iba-1, Agg and HAS1: Goat-anti-rabbit alexa fluor plus 488 IgG Green (Invitrogen, 1:200); HA: Streptavidin, Alexa Fluor conjucate 488 Green (1:500, Invitrogen). Three sections were imaged (Leica DMLB microscope; QCLASX software, 20x magnification) for each animal and averaged to obtain a single value/animal.

Statistical analysis

Statistical analysis was conducted with SigmaPlot12 software. Temperature data between survivors, non-survivors and controls were compared using a general linear model with the Tukey test for post-hoc comparisons. Baseline temperatures (Exp. 1) between the three different regions (TCR, TSC, TRU) were analyzed using One-Way ANOVA. Densitometry and microglia count between survivors, non-survivors and controls were compared using a One-way ANOVA. Linear regression analysis was performed for assessing correlations in the MPA between microglia/ECM components and TCR. Values are expressed as means ± 1 SD (significant considered at P <0.05).

Results

Experiment 1: Infrared thermography and LPS-induced mortality:

Baseline temperatures for TCR, TSC and TRU within LPSS, LPSNS and saline controls were not significantly different (Table 1). Generally, baseline TSC was slightly warmer than TCR and TRU, but this was only statistically significant for the latter. The time-domains of TCR, TSC, and TRU IRT profiles are provided for saline, LPSNS and LPSS rats (Fig. 2) based on whether LPS resulted in later mortality (LPS, non-survivors, LPSNS) or survival (LPS, survivors, LPSS). Based on IRT imaging of TCR, TSC and TRU of LPSNS treated rats, temperatures began to decrease below saline values at 80 min post-injection (T80), and at ~T90–100 for LPSS rats (Fig. 2A–C). Temperatures of three regions continued to decline in LPSNS rats until death occurred. In contrast, TCR, TSC and TRU of LPSS rats fully recovered by T240 (defined as no longer significantly below temperatures of saline rats; Fig. 2A–C). This time point (T240), therefore, was chosen for further characterization with a focus on TCR since, for a given treatment group, all three regions showed very similar IRT profiles. At T240, TCR of saline (35.1 ± 0.8°C, mean ± 1SD) rats was similar to LPSS (34.5 ± 1.2°C, mean ± 1SD), whereas TCR of LPSNS rats (32.1 ± 0.7°C, mean ± 1SD) was significantly below both saline and LPSS rats. These values were used as reference points for each respective group for Experiment 2 to assess changes in hypothalamic neurochemistry. LPS caused 74.2% mortality (4.5–7.5 h post-injection) as shown by the Kaplan-Meier Curve (Fig. 2D). LPSNS rats generally appeared lethargic, exhibited labored breathing, and were often cyanotic when death appeared imminent as described previously (Rourke et al. 2016); these animals underwent interventional euthanasia. In contrast, imminent death was not always visibly noticeable, and, in these cases, rats were asymptomatic, and death was relatively sudden. LPSS rats did not visibly display any adverse physical signs at any stage of the experiment and these rats were returned to the dam for a further 72 h for daily monitoring. There was no further documented mortality beyond 7.5 h timepoint.

Table 1.

Baseline surface temperatures (Experiment 1).

Region Treatment
S
(N=9)		 LPSS
(N=8)		 LPSNS
(N=23)
T CR 34.9 ± 0.2 34.9 ± 0.2 34.8 ± 0.1
T SC 35.4 ± 0.1 35.5 ± 0.2 35.4 ± 0.1*
T RU 34.6 ± 0.2$ 34.5 ± 0.2$ 34.5 ± 0.1$
Open in a new tab

Values are means ± 1SEM

*

p<0.05 TCR vs TSC

$

p<0.05 TSC vs TRU

TCR vs TRU (N.S)

S vs LPSS vs LPSNS (NS)

Experiment 2: Early changes in TCR and associated hypothalamic neurochemistry of LPSS and LPSNS

Upon repeating Experiment 1, 18 of the 28 rats injected with LPS exhibited TCR within ±1SD of 32.1°C (i.e. mean TCR of the LPSNS group from Experiment 1) at T240 and thus, were putatively categorized as LPSNS (Fig. 3A). TCR of the remaining 9 LPS rats exhibited TSC within 1SD of 34.5°C (i.e. mean TCR of the LPSS group from Experiment 1). One rat was excluded from analysis because TCR did not fit within the SD of either group. Baseline (i.e. pre-injection) TCR was similar between saline, LPSS and LPSNS groups (Fig. 3A). Analysis of TCR at T240 estimated 66.6% anticipated mortality (c.f. 74.8% of confirmed mortality in Experiment 1; Fig. 3B). Visual comparison and corresponding IRT images of LPSS and LPSNS rats at T240 are also shown (Fig. 3B).

Fig. 3.

Fig. 3.

Open in a new tab

(A) Baseline and T240 TCR of saline treated rats and confirmed LPSS and LPSNS rats (bar graphs, means ± 1SD) as determined from Experiment 1. Superimposed on each bar are TCR values from individual rats obtained from Experiment 2 (circles), which were categorized as presumed LPSS (open circles) or LPSNS (grey circles) based on whether their TCR at T240 was within ±1SD of the mean values obtained from Experiment 1 (saline (S) 35.1 ± 0.8°C; LPSS 34.5 ± 1.2°C; LPSNS 32.1 ± 0.7°C; mean ± 1SD). (B) Representative digital (top) and infrared (bottom) images of a surviving (left, LPSS) and non-surviving (right, LPSNS) P10 day old rat taken at T240 post-injection. Note the color scale in which warmer regions display as white, red and orange and cooler regions display as yellow, green and blue. To the naked eye, these animals are barely discernible, but their infrared images are in stark contrast with one-another. The projected mortality following LPS was 66.7% (pie chart).

Microglia cell counts (shown in Fig. 4Ai) and densitometry (Fig. 4Bi) were significantly increased in the MPA only in LPSNS rats compared to LPSS and saline rats. There was a significant negative correlation between TCR and both microglia counts (Fig. 4Aii; R2=0.41, p<0.0001) and densitometry (Fig. 4Bii; R2=0.39, p<0.0001) in the MPA. Similar trends were seen in the MnPO for microglia cell counts (Fig. 5Ai) and densitometry (Fig. 5Bi), although such changes were observed in both LPSS and LPSNS rats (compared to saline rats). There was no difference in microglia expression (cell counts or densitometry) between groups in the AVPO (Fig. 5Aii and 5Bii).

Fig. 4.

Fig. 4.

Open in a new tab

Iba-1 immunofluorescent staining for microglia cell counts (A) and fluorescence (B) in the medial preoptic area (MPA) in saline, LPSS, and LPSNS rats at T240 post-injection. (C) Representative images for each group are taken at 20x magnification. Note, the increased microglia expression in the MPA (Ai and Bi) exclusively in LPSNS rats (vs saline and LPSS). Linear regression analysis also showed significant negative correlations between microglia expression and TCR (Aii and Bii). Rats were sacrificed at T240 but were putatively categorized as LPSS or LPSNS based on the cranial IRT profile of these respective groups determined from Experiment 1. Representative Iba-1 images of the MPA for each group are also provided (C). Values are mean ± 1SD; p-values and R2 values are also provided.

Figure 5.

Figure 5.

Open in a new tab

Iba-1 immunofluorescent staining for microglia cell counts (A) and fluorescence (B) in median preoptic nucleus (MnPO; Ai and Bi), and anteroventral preoptic nucleus (AVPO; Aii and Bii). Note microglia cell counts and densitometry were elevated in the MnPO in both LPSS and LPSNS compared to saline (i.e. not exclusive to LPSNS) and expression was unchanged in the AVPO. Therefore, subsequent ECM analysis (Fig. 6) was restricted to the MPA. Data are means ± 1SD and significant p-values are indicated.

Since only the MPA showed significant changes in microglia expression in LPSNS rats, this region was the focus for associated changes in ECM expression (Fig. 6A–C). There was a significant negative correlation between TCR and HAS1 (R2=0.15, p<0.021), an enzyme responsible for synthesis of high molecular weight HA (Fig. 6Ai). HAS1 expression was co-expressed with neurons (shown in Fig. 6Aii, inset). Further, Expression of HAS1was increased only in LPSNS rats (vs saline and LPSS rats; Fig. 6Av). Similarly, TCR was negatively correlated with both HA (Fig. 6B; R2=0.42, p<0.0001) and Agg (Fig. 6C; R2=0.11, p<0.038) and significantly increased only in LPSNS rats (vs saline and LPSS).

Fig. 6.

Fig. 6.

Open in a new tab

Extracellular matrix immunofluorescence including hyaluronan synthase 1 (HAS1; A), Hyaluronan (HA; B) and Aggrecan (C) immunofluorescence in medial preoptic area (MPA) in saline, LPSS and LPSNS rats at T240 post-injection. Rats were sacrificed at T240, but were categorized as LPSS or LPSNS based on the cranial IRT profile of these respective groups determined from Experiment 1. Note, the significant negative correlation (linear regression analysis) for HAS1, HA, and Aggrecan densitometry with TCR (Ai, Bi, and Ci, respectively) in the MPA. HAS1, HA, and Aggrecan expression was also exclusively elevated in the LPSNS groups compared to LPSS and saline rats (v). Representative images of the MPA for each group are also provided (ii, iii, iv). Corresponding neuronal (NeuN, red) double-labeling with HAS1 (green) for the saline rat (Aii, inset) showing localized expression within neurons. Images are taken at 20x magnification. Values are mean ± 1SD, p-values and R2 values are also provided.

Discussion:

Using a rat model of neonatal endotoxemia, this study demonstrated the utility of using IRT to detect subtle, yet unique profiles in skin surface temperature for early identification of rats who were lethally vulnerable following exposure to LPS endotoxin. Another notable observation was that the unique IRT profiles of these rats consisted of early, severe, and non-recoverable hypothermic tendencies, which were also associated with specific increases in hypothalamic microglia and ECM deposition. Specifically, in the MPA there were significant negative correlations in microglia and components of the ECM with TCR. These data suggest that failure to recover from progressive hypothermia during endotoxic shock may be linked to thermogenic collapse mediated in part by aberrant increases in microglia and several components of the ECM within specific hypothalamic thermoregulatory regions.

Infrared Thermographic Profiles Associated with LPS-induced Mortality:

All rats who received an LPS injection developed a decrease in their IRT profiles; for TCR the decrease was slightly delayed (by 20 min vs non-survivors) in rats who tolerated (survived) LPS. Progressive hypothermia is consistent with the clinical setting in preterm infants who predominantly experience hypothermia as opposed to hyperthermia when septic (Kumar et al. 2009). Sepsis-induced hypothermia is also associated with significant infant morbidity and mortality (Bang et al. 2005), which was reflected in our 66–74% mortality rate for LPS-exposed rats. However, the neonatal mortality rate of the present study was higher than our prior study (Rourke et al. 2016). The slight discrepancy in mortality rate between the two studies may be related to the circumstance surrounding the handling of the pups after injection. The study by Rourke et al., (Rourke et al. 2016) returned the pups to the dam for maternal care, nursing, and nesting after LPS injection rather than the thermally stable isolette (maintained at ~28.5°C) of the present study. The potentially warmer nesting conditions (~34°C; (Cramer and Blass 1982; Stern and Lonstein 1996)) compared to the isolette could have differentially influenced endogenous thermogenic abilities and affected whether some individuals could have thermally recovered from LPS. An influence of ambient temperature on mortality has been observed in adults hospitalized with sepsis in which hypothermia and mortality was associated with lower outside ambient temperature (Thomas-Rüddel et al. 2021). Another unique feature revealed by the IRT imaging for LPSNS rats was the more severe decrease in surface temperatures and failure to spontaneously recover compared to LPSS rats. Whether a different ambient temperature of the isolette impacts the IRT profiles or changes the proportion of LPSS:LPSNS rats may have significant clinical implications.

A major challenge with neonatal sepsis is its early detection. Using IRT, we were able to show that changes in the profile of skin (specifically cranial) heat maps ~80 minutes post-LPS could identify individual rats who would later be lethally vulnerable to LPS, separating these rats from those who fully recovered TCR. While the difference here is robust enough to allow for statistical differentiation between non-survivors and survivors at this early period, the eventual survivors of LPS recovered TCR by 4 h post-injection whereas non-survivors remained markedly hypothermic and expired between 4.5–7.5 h post-injection. The cause for the mortality is unclear although likely linked to unique proinflammatory responses. Prior work suggests that the severity of hypothermia and LPS-induced sepsis may be related to TNFα signaling through TNF-R1 (Long et al. 1992; Mul Fedele et al. 2020), which may function as a cryogenic cytokine capable of reducing body temperature (Klir et al. 1995; Moncada et al. 1991). Rourke et al (Rourke et al. 2016) demonstrated a robust brainstem TNFα mRNA response in P10 day old rats exposed to LPS. Similarly, iNOS is also increased following LPS which may be significant since NO is known to reduce blood pressure, block response to pressors, influence the distribution of circulation and suppress cardiac myocytes (MacMicking et al. 1995; Roberts et al. 1992), each of which can contribute to alterations in skin-surface temperature patterns that would be detected by infrared thermography.

Histopathologic changes in the thermoregulatory regions of the hypothalamus are associated with mortality and negatively correlated with TCR

The etiology of hypothermia during sepsis is poorly understood. Endothelial dysfunction, peripheral vasodilation and loss of endothelial integrity during sepsis may result in heat loss by hampering the body’s ability to regulate core temperature (Mallet 2002; Romanovsky et al. 1996). The results of the present study demonstrated a robust increase in hypothalamic microglia in non-surviving animals compared with surviving animals, particularly in the MPA of the hypothalamus. In fact, there was a significant negative correlation between MPA microglia and TCR indicating higher microglia counts are associated with lower TCR. As a main thermoregulatory region of the brain, it also regulates temperature homeostasis in the presence of different stimuli or immune challenges (Paladino et al. 2010; Tan and Knight 2018; Zhao et al. 2017). Multiple studies have reported increased neuronal activation in rats exposed to LPS (Uchida et al. 2014) as well as specifically within the preoptic area (POA) in mice which developed hypothermia (Werneburg et al. 2017). This work expands upon the discoveries of Mul Fedele et al in which greater neuronal activation of the POA of the hypothalamus, thought to be immune mediated, correlated with increased severity of hypothermia and increased mortality in a mouse model of sepsis. The elevated microglial burden seen in the MPA and to some extent in the MnPO in non-surviving animals and its inverse correlation with TCR in this study suggests it may not be an advantageous response to LPS possibly by mediating a more robust and prolonged hypothermic response and prevention of thermoregulatory mechanisms involved in recovery of thermogenesis.

In a different neonatal model, we showed previously that postnatal chronic hypoxia (a pro-inflammatory stimulus), which also resulted in significant mortality, was associated with increased microglia (prior to death) in the brainstem, leading to disruptions in respiratory and autonomic function (MacFarlane et al. 2016). A follow-up study also showed increased expression of various components of the ECM (Stryker et al. 2018), suggesting a potential interplay between glial and ECM expression. The results of the present study demonstrated a marked increase in HA and one of the enzymes involved in its synthesis (HAS1) as well as aggrecan in the MPA of non-surviving animals. Similarly, as was the case with microglia, these ECM components were also negatively correlated with TCR, indicating higher ECM expression is associated with lower TCR, which is likely not advantageous given the later mortality. HA is a sulfated glycosaminoglycan that functions as a main component of the ECM and is actively produced during tissue injury, tissue repair and wound healing (Jiang et al. 2005). HA regulates activation of inflammatory cells and immunological responses (Feusi et al. 1999; Teriete et al. 2004). Dysregulated HAS1 expression has been demonstrated in many diseases in both animal and human studies (Lennon and Singleton 2011). These proteins have been shown to be upregulated in the setting of LPS-induced inflammation (Morawski et al. 2004). To our knowledge, no previous work has investigated the relationship between aberrant ECM expression in the hypothalamus in the context of sepsis. Zhu et al demonstrated that induced microglial cytokine expression in the CNS was likely mediated by HA, and that inhibition of hyaluronidase, a family of enzymes that catalyzes HA degradation, reduces IL-1-induced microglial cytokine expression in vitro and in vivo. Additional studies investigating the clinical effects of the ECM would shed light on the relationship between hypothermia and mortality in neonatal sepsis as demonstrated here.

Conclusions:

We showed that IRT imaging can identify distinct thermal profiles of skin-surface temperatures, and that these early patterns can distinguish between rats in which LPS would ultimately be lethal. The hypothermic tendencies of these rats are associated with early changes in neurochemistry in key thermoregulatory regions of the hypothalamus. The microglia response and associated aberrant changes in the ECM, which occurs exclusively in the rats that later expired, may play a lethal role involving disruption of mechanisms necessary for thermogenic recovery. These data could have implications for preterm neonate hypothermic responses during sepsis or endotoxic shock and suggest further avenues for investigation into the role of the pre-optic area of the hypothalamus in mediating the clinical effects of sepsis. Further, while we did not assess sex differences in the response to LPS, these are important future studies given the sexual dimorphic characteristics of microglia, which could hold clues into potential differences in vulnerabilities to endotoxin. Overall, the possibility that IRT, especially given the convenience of assessing cranial temperatures, could be used for early identification of neonates exhibiting increased vulnerability to sepsis should be investigated. It remains unclear whether the neurochemical changes in the hypothalamus are causative or consequential to the hypothermic tendencies following LPS. However, it is clear that the neurochemical changes precede the failure of thermogenic recovery suggesting there may be a window of opportunity for intervention. Further studies are needed to address how the ambient thermal environment affects the neonatal response, how the neurochemical changes seen here affect neural function in the hypothalamus, and whether such profound effects of LPS can be prevented through targeting microglia and the ECM.

Highlights:

  • Infrared imaging of skin (e.g. cranial) surface temperature reveals unique thermographic profiles following neonatal endotoxin exposure.

  • Infrared thermography can be used for early identification of neonates at risk of later mortality from endotoxic shock in neonates.

  • Lethality from endotoxic shock is associated with irreversible hypothermic tendencies.

  • The hypothalamic microglia and extracellular matrix expression are negatively correlated with thermographic profiles.

Funding:

Funded by NIH HD-111415, the Rainbow Babies and Children’s Foundation and the William and Lois Briggs research endowment to PMM.

Footnotes

Competing Interests: Authors have no potential conflicts of interest to report.

Disclaimer: The views expressed in this article are those of the authors and do not necessarily represent those of the National Institutes of Health or the U.S. Department of Health and Human Services.

Data Availability:

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

References:

  1. Amson H, Vacheron CH, Thiolliere F, Piriou V, Magnin M, Allaouchiche B (2020) Core-to-skin temperature gradient measured by thermography predicts day-8 mortality in septic shock: A prospective observational study. J Crit Care 60:294–299. doi: 10.1016/j.jcrc.2020.08.022 [DOI] [PubMed] [Google Scholar]
  2. Bang AT, Reddy HM, Deshmukh MD, Baitule SB, Bang RA (2005) Neonatal and infant mortality in the ten years (1993 to 2003) of the Gadchiroli field trial: effect of home-based neonatal care. J Perinatol 25 Suppl 1:S92–107. doi: 10.1038/sj.jp.7211277 [DOI] [PubMed] [Google Scholar]
  3. Bozadas P, Nadella R, Sánchez-García A, Rembao Bojórquez D, Rovirosa Hernández M, Hernandez Baltazar D (2019) Lipopolysaccharide-based endotoxemia produce toxicity in peripheral organs and microglia migration in a dose-dependent manner in rat substantia nigra. Folia Neuropathol 57 (3):258–266. doi: 10.5114/fn.2019.88454 [DOI] [PubMed] [Google Scholar]
  4. Cramer CP, Blass EM (1982) The contribution of ambient temperature to suckling behavior in rats 3–20 days of age. Dev Psychobiol 15 (4):339–348. doi: 10.1002/dev.420150406 [DOI] [PubMed] [Google Scholar]
  5. da Costa LHA, Santos-Junior NN, Catalão CHR, Rocha MJA (2021) Microglial Activation Modulates Neuroendocrine Secretion During Experimental Sepsis. Mol Neurobiol 58 (5):2133–2144. doi: 10.1007/s12035-020-02241-5 [DOI] [PubMed] [Google Scholar]
  6. Feusi E, Sun L, Sibalic A, Beck-Schimmer B, Oertli B, Wüthrich RP (1999) Enhanced hyaluronan synthesis in the MRL-Fas(lpr) kidney: role of cytokines. Nephron 83 (1):66–73. doi: 10.1159/000045475 [DOI] [PubMed] [Google Scholar]
  7. Fleischmann C, Reichert F, Cassini A, Horner R, Harder T, Markwart R, Tröndle M, Savova Y, Kissoon N, Schlattmann P, Reinhart K, Allegranzi B, Eckmanns T (2021) Global incidence and mortality of neonatal sepsis: a systematic review and meta-analysis. Arch Dis Child 106 (8):745–752. doi: 10.1136/archdischild-2020-320217 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Hofer N, Müller W, Resch B (2012) Neonates presenting with temperature symptoms: role in the diagnosis of early onset sepsis. Pediatr Int 54 (4):486–490. doi: 10.1111/j.1442-200X.2012.03570.x [DOI] [PubMed] [Google Scholar]
  9. Inghammar M, Sunden-Cullberg J (2020) Prognostic significance of body temperature in the emergency department vs the ICU in Patients with severe sepsis or septic shock: A nationwide cohort study. PLoS One 15 (12):e0243990. doi: 10.1371/journal.pone.0243990 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Iroh Tam PY, Bendel CM (2017) Diagnostics for neonatal sepsis: current approaches and future directions. Pediatr Res 82 (4):574–583. doi: 10.1038/pr.2017.134 [DOI] [PubMed] [Google Scholar]
  11. Jiang D, Liang J, Fan J, Yu S, Chen S, Luo Y, Prestwich GD, Mascarenhas MM, Garg HG, Quinn DA, Homer RJ, Goldstein DR, Bucala R, Lee PJ, Medzhitov R, Noble PW (2005) Regulation of lung injury and repair by Toll-like receptors and hyaluronan. Nat Med 11 (11):1173–1179. doi: 10.1038/nm1315 [DOI] [PubMed] [Google Scholar]
  12. Johnson P, Arif AA, Lee-Sayer SSM, Dong Y (2018) Hyaluronan and Its Interactions With Immune Cells in the Healthy and Inflamed Lung. Front Immunol 9:2787. doi: 10.3389/fimmu.2018.02787 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Jones BF (1998) A reappraisal of the use of infrared thermal image analysis in medicine. IEEE Trans Med Imaging 17 (6):1019–1027. doi: 10.1109/42.746635 [DOI] [PubMed] [Google Scholar]
  14. Kateb B, Yamamoto V, Yu C, Grundfest W, Gruen JP (2009) Infrared thermal imaging: a review of the literature and case report. Neuroimage 47 Suppl 2:T154–162. doi: 10.1016/j.neuroimage.2009.03.043 [DOI] [PubMed] [Google Scholar]
  15. Katz LM, Nauriyal V, Nagaraj S, Finch A, Pearlstein K, Szymanowski A, Sproule C, Rich PB, Guenther BD, Pearlstein RD (2008) Infrared imaging of trauma patients for detection of acute compartment syndrome of the leg. Crit Care Med 36 (6):1756–1761. doi: 10.1097/CCM.0b013e318174d800 [DOI] [PubMed] [Google Scholar]
  16. Kazune S, Vasiljevs E, Caica-Rinca A, Marcinkevics Z, Grabovskis A (2023) Infrared Thermography Imaging for Assessment of Peripheral Perfusion in Patients with Septic Shock. Bioengineering (Basel) 10 (6). doi: 10.3390/bioengineering10060729 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Klir JJ, McClellan JL, Kozak W, Szelényi Z, Wong GH, Kluger MJ (1995) Systemic but not central administration of tumor necrosis factor-alpha attenuates LPS-induced fever in rats. Am J Physiol 268 (2 Pt 2):R480–486. doi: 10.1152/ajpregu.1995.268.2.R480 [DOI] [PubMed] [Google Scholar]
  18. Kluger MJ (1991) Fever: role of pyrogens and cryogens. Physiol Rev 71 (1):93–127. doi: 10.1152/physrev.1991.71.1.93 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Knobel RB, Guenther BD, Rice HE (2011) Thermoregulation and thermography in neonatal physiology and disease. Biol Res Nurs 13 (3):274–282. doi: 10.1177/1099800411403467 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kumar V, Shearer JC, Kumar A, Darmstadt GL (2009) Neonatal hypothermia in low resource settings: a review. J Perinatol 29 (6):401–412. doi: 10.1038/jp.2008.233 [DOI] [PubMed] [Google Scholar]
  21. Lagerspetz KY, Väätäinen T (1987) Bacterial endotoxin and infection cause behavioural hypothermia in infant mice. Comp Biochem Physiol A Comp Physiol 88 (3):519–521. doi: 10.1016/0300-9629(87)90074-0 [DOI] [PubMed] [Google Scholar]
  22. Lawn JE, Cousens S, Zupan J (2005) 4 million neonatal deaths: when? Where? Why? Lancet 365 (9462):891–900. doi: 10.1016/s0140-6736(05)71048-5 [DOI] [PubMed] [Google Scholar]
  23. Lemstra AW, Groen in’t Woud JC, Hoozemans JJ, van Haastert ES, Rozemuller AJ, Eikelenboom P, van Gool WA (2007) Microglia activation in sepsis: a case-control study. J Neuroinflammation 4:4. doi: 10.1186/1742-2094-4-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Lennon FE, Singleton PA (2011) Role of hyaluronan and hyaluronan-binding proteins in lung pathobiology. Am J Physiol Lung Cell Mol Physiol 301 (2):L137–147. doi: 10.1152/ajplung.00071.2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liu Y, Li Z, Khan S, Zhang R, Wei R, Zhang Y, Xue M, Yong VW (2021) Neuroprotection of minocycline by inhibition of extracellular matrix metalloproteinase inducer expression following intracerebral hemorrhage in mice. Neurosci Lett 764:136297. doi: 10.1016/j.neulet.2021.136297 [DOI] [PubMed] [Google Scholar]
  26. Long NC, Morimoto A, Nakamori T, Murakami N (1992) Systemic injection of TNF-alpha attenuates fever due to IL-1 beta and LPS in rats. Am J Physiol 263 (5 Pt 2):R987–991. doi: 10.1152/ajpregu.1992.263.5.R987 [DOI] [PubMed] [Google Scholar]
  27. MacFarlane PM, Mayer CA, Litvin DG (2016) Microglia modulate brainstem serotonergic expression following neonatal sustained hypoxia exposure: implications for sudden infant death syndrome. J Physiol 594 (11):3079–3094. doi: 10.1113/jp271845 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. MacMicking JD, Nathan C, Hom G, Chartrain N, Fletcher DS, Trumbauer M, Stevens K, Xie QW, Sokol K, Hutchinson N, et al. (1995) Altered responses to bacterial infection and endotoxic shock in mice lacking inducible nitric oxide synthase. Cell 81 (4):641–650. doi: 10.1016/0092-8674(95)90085-3 [DOI] [PubMed] [Google Scholar]
  29. MacRae BA, Annaheim S, Spengler CM, Rossi RM (2018) Skin Temperature Measurement Using Contact Thermometry: A Systematic Review of Setup Variables and Their Effects on Measured Values. Front Physiol 9:29. doi: 10.3389/fphys.2018.00029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mallet ML (2002) Pathophysiology of accidental hypothermia. Qjm 95 (12):775–785. doi: 10.1093/qjmed/95.12.775 [DOI] [PubMed] [Google Scholar]
  31. Michels M, Abatti M, Vieira A, Ávila P, Goulart AI, Borges H, Córneo E, Dominguini D, Barichello T, Dal-Pizzol F (2020) Modulation of microglial phenotypes improves sepsis-induced hippocampus-dependent cognitive impairments and decreases brain inflammation in an animal model of sepsis. Clin Sci (Lond) 134 (7):765–776. doi: 10.1042/cs20191322 [DOI] [PubMed] [Google Scholar]
  32. Mital M, Scott EP (2007) Thermal detection of embedded tumors using infrared imaging. J Biomech Eng 129 (1):33–39. doi: 10.1115/1.2401181 [DOI] [PubMed] [Google Scholar]
  33. Moncada S, Palmer RM, Higgs EA (1991) Nitric oxide: physiology, pathophysiology, and pharmacology. Pharmacol Rev 43 (2):109–142 [PubMed] [Google Scholar]
  34. Morawski M, Brückner MK, Riederer P, Brückner G, Arendt T (2004) Perineuronal nets potentially protect against oxidative stress. Exp Neurol 188 (2):309–315. doi: 10.1016/j.expneurol.2004.04.017 [DOI] [PubMed] [Google Scholar]
  35. Mul Fedele ML, Aiello I, Caldart CS, Golombek DA, Marpegan L, Paladino N (2020) Differential Thermoregulatory and Inflammatory Patterns in the Circadian Response to LPS-Induced Septic Shock. Front Cell Infect Microbiol 10:100. doi: 10.3389/fcimb.2020.00100 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Paladino N, Leone MJ, Plano SA, Golombek DA (2010) Paying the circadian toll: the circadian response to LPS injection is dependent on the Toll-like receptor 4. J Neuroimmunol 225 (1–2):62–67. doi: 10.1016/j.jneuroim.2010.04.015 [DOI] [PubMed] [Google Scholar]
  37. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M (2001) Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 345 (19):1368–1377. doi: 10.1056/NEJMoa010307 [DOI] [PubMed] [Google Scholar]
  38. Roberts AB, Vodovotz Y, Roche NS, Sporn MB, Nathan CF (1992) Role of nitric oxide in antagonistic effects of transforming growth factor-beta and interleukin-1 beta on the beating rate of cultured cardiac myocytes. Mol Endocrinol 6 (11):1921–1930. doi: 10.1210/mend.6.11.1282674 [DOI] [PubMed] [Google Scholar]
  39. Rolls A, Shechter R, London A, Segev Y, Jacob-Hirsch J, Amariglio N, Rechavi G, Schwartz M (2008) Two faces of chondroitin sulfate proteoglycan in spinal cord repair: a role in microglia/macrophage activation. PLoS Med 5 (8):e171. doi: 10.1371/journal.pmed.0050171 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Romanovsky AA, Shido O, Sakurada S, Sugimoto N, Nagasaka T (1996) Endotoxin shock: thermoregulatory mechanisms. Am J Physiol 270 (4 Pt 2):R693–703. doi: 10.1152/ajpregu.1996.270.4.R693 [DOI] [PubMed] [Google Scholar]
  41. Rourke KS, Mayer CA, MacFarlane PM (2016) A critical postnatal period of heightened vulnerability to lipopolysaccharide. Respir Physiol Neurobiol 232:26–34. doi: 10.1016/j.resp.2016.06.003 [DOI] [PubMed] [Google Scholar]
  42. Santos-Junior NN, Catalão CHR, Costa LHA, Souza AO, Mota CMD, Alberici LC, Branco LGS, Rocha MJA (2018) Experimental sepsis induces sustained inflammation and acetylcholinesterase activity impairment in the hypothalamus. J Neuroimmunol 324:143–148. doi: 10.1016/j.jneuroim.2018.08.013 [DOI] [PubMed] [Google Scholar]
  43. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche JD, Coopersmith CM, Hotchkiss RS, Levy MM, Marshall JC, Martin GS, Opal SM, Rubenfeld GD, van der Poll T, Vincent JL, Angus DC (2016) The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). Jama 315 (8):801–810. doi: 10.1001/jama.2016.0287 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Sorokin L (2010) The impact of the extracellular matrix on inflammation. Nat Rev Immunol 10 (10):712–723. doi: 10.1038/nri2852 [DOI] [PubMed] [Google Scholar]
  45. Stern JM, Lonstein JS (1996) Nursing behavior in rats is impaired in a small nestbox and with hyperthermic pups. Dev Psychobiol 29 (2):101–122. doi: 10.1002/(sici)1098-2302(199603)29:2<101::Aid-dev2>3.0.Co;2-w [DOI] [PubMed] [Google Scholar]
  46. Stoll BJ, Puopolo KM, Hansen NI, Sánchez PJ, Bell EF, Carlo WA, Cotten CM, D’Angio CT, Kazzi SNJ, Poindexter BB, Van Meurs KP, Hale EC, Collins MV, Das A, Baker CJ, Wyckoff MH, Yoder BA, Watterberg KL, Walsh MC, Devaskar U, Laptook AR, Sokol GM, Schrag SJ, Higgins RD (2020) Early-Onset Neonatal Sepsis 2015 to 2017, the Rise of Escherichia coli, and the Need for Novel Prevention Strategies. JAMA Pediatr 174 (7):e200593. doi: 10.1001/jamapediatrics.2020.0593 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Stryker C, Camperchioli DW, Mayer CA, Alilain WJ, Martin RJ, MacFarlane PM (2018) Respiratory dysfunction following neonatal sustained hypoxia exposure during a critical window of brain stem extracellular matrix formation. Am J Physiol Regul Integr Comp Physiol 314 (2):R216–r227. doi: 10.1152/ajpregu.00199.2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Tan CL, Knight ZA (2018) Regulation of Body Temperature by the Nervous System. Neuron 98 (1):31–48. doi: 10.1016/j.neuron.2018.02.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Teriete P, Banerji S, Noble M, Blundell CD, Wright AJ, Pickford AR, Lowe E, Mahoney DJ, Tammi MI, Kahmann JD, Campbell ID, Day AJ, Jackson DG (2004) Structure of the regulatory hyaluronan binding domain in the inflammatory leukocyte homing receptor CD44. Mol Cell 13 (4):483–496. doi: 10.1016/s1097-2765(04)00080-2 [DOI] [PubMed] [Google Scholar]
  50. Thomas-Rüddel DO, Hoffmann P, Schwarzkopf D, Scheer C, Bach F, Komann M, Gerlach H, Weiss M, Lindner M, Rüddel H, Simon P, Kuhn SO, Wetzker R, Bauer M, Reinhart K, Bloos F (2021) Fever and hypothermia represent two populations of sepsis patients and are associated with outside temperature. Crit Care 25 (1):368. doi: 10.1186/s13054-021-03776-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Uchida Y, Tokizawa K, Nagashima K (2014) Characteristics of activated neurons in the suprachiasmatic nucleus when mice become hypothermic during fasting and cold exposure. Neurosci Lett 579:177–182. doi: 10.1016/j.neulet.2014.07.039 [DOI] [PubMed] [Google Scholar]
  52. Werneburg S, Feinberg PA, Johnson KM, Schafer DP (2017) A microglia-cytokine axis to modulate synaptic connectivity and function. Curr Opin Neurobiol 47:138–145. doi: 10.1016/j.conb.2017.10.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Westhoff D, Engelen-Lee JY, Hoogland ICM, Aronica EMA, van Westerloo DJ, van de Beek D, van Gool WA (2019) Systemic infection and microglia activation: a prospective postmortem study in sepsis patients. Immun Ageing 16:18. doi: 10.1186/s12979-019-0158-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Yang G, Cao Y, Wang P, Mei L, Chen J, Lu W (2022) Minocycline Pretreatment Prevents Blood-Brain Barrier Disruption in Septic Rats. J Surg Res 273:247–254. doi: 10.1016/j.jss.2022.01.021 [DOI] [PubMed] [Google Scholar]
  55. Zhao ZD, Yang WZ, Gao C, Fu X, Zhang W, Zhou Q, Chen W, Ni X, Lin JK, Yang J, Xu XH, Shen WL (2017) A hypothalamic circuit that controls body temperature. Proc Natl Acad Sci U S A 114 (8):2042–2047. doi: 10.1073/pnas.1616255114 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

RESOURCES