Effects of coupled dose and rhythm manipulation of plasma cortisol levels on leukocyte transcriptional response to endotoxin challenge in humans

Abstract

Severe traumas are associated with hypercortisolemia due to both disruption of cortisol secretion rhythm and increase in its total concentration. Understanding the effects of altered cortisol levels and rhythms on immune function is of great clinical interest, to prevent conditions such as sepsis from complicating the recovery. This in vivo study assesses the responses of circulating leukocytes to coupled dose and rhythm manipulation of cortisol, preceding an immune challenge induced by endotoxin administration. Through continuous infusion, plasma cortisol concentration was increased to and kept constant at a level associated with major physiologic stress. In response, transcriptional programming of leukocytes was altered to display a priming response before endotoxin exposure. Enhanced expression of a number of receptors and signaling proteins, as well as lowered protein translation and mitochondrial function indicated a sensitization against potential infectious threats. Despite these changes, response to endotoxin followed very similar patterns in both cortisol and saline pre-treated groups except one cluster including probe sets associated with major players regulating inflammatory response. In sum, altered dose and rhythm of plasma cortisol levels engendered priming of circulating leukocytes when preceded an immune challenge. This transcriptional program change associated with stimulated surveillance function and suppressed energy-intensive processes, emphasized permissive actions of cortisol on immune function.

Keywords

Cortisol endotoxin gene expression leukocytes systems biology

Introduction

Healthy function of the immune system is intrinsically tied to the biological rhythms running in harmony. There is a bi-directional information flow between immune and neuroendocrine systems such that while a wide range of immune parameters goes through daily fluctuations under the control of autonomic and endocrine rhythms, immune effectors, such as cytokines, can, in turn, influence the phase of these rhythms.¹ In the case of acute stress, such as trauma or infection, an inflammatory response is mounted by the coordinated action of immune and neuroendocrine systems. The hypothalamus–pituitary–adrenal axis is a crucial mediator of this action because of its control on circadian and ultradian glucocorticoid (GC) secretion patterns.² These fluctuating patterns are hypothesized to provide the basis for a ‘continuous dynamic equilibration’ that is fundamental to the establishment of homeostasis and allostasis.³ Cortisol is the major circulating human GC which, through feedforward and feedback regulatory loops, maintains this dynamic state that is essential for proper response to stressors. Through mathematical modeling of these regulatory mechanisms, our group previously showed that the ultradian oscillations of GC concentration are critical for maintaining the homeostatic balance on downstream responses. Mean levels of GC responsive mRNA synthesis was shown to be greater when the pulsatility in plasma cortisol levels was diminished despite a constant total amount of GC exposure.⁴ Besides cortisol secretion frequency, parameters such as concentration and the timing with respect to a stressor, also affect the characteristics of stress response, contributing greatly to the emergence of inter-individual variations in reaction to similar stressors.⁵

GCs are mostly involved through their immunosuppressive and anti-inflammatory actions, which have been clinically utilized for the treatment of autoimmune and inflammatory disorders for decades; however, the very first recognized view on GC physiology in the 1940s by Selye suggested that release of endogenous GCs enhances and mediates the response to external stressors, rather than limiting it. This early statement was later reshaped and extended by emerging anti-inflammatory actions of GCs⁶ and a view that synthesizes both permissive and suppressive actions of GCs on immune function was put forward. This view proposes that GCs can have disparate effects on the individual’s response to a stressor depending on the timing, duration or concentration of the exposure. These can include permissive, suppressive, stimulatory or preparatory effects. With regard to the concentration of GCs, only permissive effects on immune function are typically associated with basal levels of GCs, whereas the other three effects (suppressive, stimulatory or preparatory) are commonly observed when GC concentration is at high or stress-induced levels. However, permissive actions can also be associated with higher cortisol levels if these instances precede an acute stressor.⁷ Concentration-dependent specific actions of GCs are exemplified by an in vivo study in which low-dose corticosterone administered to rats stimulated an inflammatory phenotype in macrophages, while high-dose corticosterone acted in a purely immunosuppressive fashion.⁸ Related to the temporal dependence of a GC surge and the incidence of an acute stressor; research has shown contrasting effects even in the same concentration range. When GC administration preceded an immune challenge (i.e. LPS exposure) in rats, both peripheral and central pro-inflammatory responses were potentiated as reflected by increased cytokine secretion. However, when the same concentration of GCs was administered after the inflammatory stimulus, attenuation of the inflammatory reaction was observed.⁹

As GCs are not only major endogenous hormones regulating the stress response but also pharmacological agents that are used clinically for the treatment of many inflammatory diseases, building a solid understanding of their effects on immune cells is of primary importance. Studies have previously investigated the effects of varying doses of GCs on the response to LPS challenge in healthy humans, establishing both permissive and stimulatory roles of GCs, as determined by the measurement of circulating cytokines.^10,11 Study of a single immunological parameter is not adequate to define the complexity of immune system. System-wide approaches, such as global transcriptional profiling, provide a wider view of the physiological alterations in response to stimuli, enabling better definition of complex biological phenomena.¹² This approach has been used to profile transcriptional responses of blood cells stimulated by GCs ex vivo.^13,14 Importantly, major physical and psychological traumas are associated with hypercortisolemia, which emerges as a result of the disruption of the cortisol secretion rhythm and the increase in total cortisol concentration.^15,16 Understanding how this alteration challenges the immune system and its function is of great clinical interest, and is particularly relevant to the management of critically ill or injured patients, in order to prevent complications. The present in vivo study seeks to assess the transcriptional responses of whole blood leukocytes [white blood cells (WBCs)] to coupled dose and rhythm manipulation of cortisol preceding an immune challenge modeled by LPS administration. A healthy human model of LPS administration has been used as a reproducible experimental platform providing mechanistic insights into how cells and organs respond to systemic inflammation. The model transiently alters many physiologic and metabolic processes in a manner that is qualitatively similar to those observed after acute injury and systemic inflammation.¹⁷ It therefore serves as a suitable platform for investigating the changes in immune response due to hormonal manipulations. The current study design involves blunting the natural circadian rhythm of cortisol in healthy volunteers by continuous cortisol infusion for 30 h at a concentration that mimics hypercortisolemia induced by a major physiologic stressor. At hour 24 of this infusion period, LPS was administered as a bolus injection and transcriptional responses of leukocytes were investigated by gene expression analysis for 24 h periods both before and after the LPS challenge. Differential gene expression and clustering analyses were done separately for these two periods to allow focusing both on the response to GC infusion alone and on the effects of this treatment to subsequent LPS exposure.

Materials and methods

Human participants

Nine healthy participants aged between 18 and 40 yr were recruited by public advertisement for participation. All participants provided written, informed consent under guidelines approved by the Institutional Review Board of Robert Wood Johnson Medical School. Inclusion criteria were being 18–40 years of age and having ‘normal general health’, as demonstrated by medical history and physical examination, as well as laboratory testing. Women of child-bearing potential were screened for pregnancy risk and were included if utilizing reliable contraception. Participants underwent an initial screening visit that included the comprehensive history, physical and laboratory testing to establish suitability for inclusion; eligible participants returned within 3 wk for admission to the clinical research center.

Volunteers were randomized to receive intravenous saline (n = 4) infusion or continuous cortisol infusion at a dose of 3 µg/kg (n = 5) for a total of 30 h starting 24 h before endotoxin administration. Participants in both groups were intravenously administered 2 ng/kg endotoxin (CC-RE, lot 2) as a bolus dose to induce moderate systemic inflammation at 0 h (09:00 clock time), as previously described.^17,18 Vital signs and blood samples were obtained and recorded at various times both pre- and post-endotoxin challenge. Throughout admission, all participants were attended by nursing staff on a 24-h basis. Subjects were precluded from any physical activity other than walking about the study unit.

Blood-derived plasma was analyzed by radioimmunoassay for plasma cortisol concentration, as previously described.¹⁹ Whole blood total leukocyte levels and differentials were measured by standard laboratory procedures at the Robert Wood Johnson University Hospital clinical pathology laboratories. Blood samples were collected and leukocytes were isolated at multiple time points throughout a 48-h period (indicated in Figure 1) starting with the cortisol/saline infusion and ending 24 h after LPS administration. Following RNA extraction and cRNA synthesis, total cellular RNA was hybridized on to a GeneChip Human Genome Focus Array (Affymetrix, Santa Clara, CA, USA), as described previously.²⁰

Figure 1.

Study design: nine healthy participants received continuous cortisol (or saline) infusion starting 24 h before the bolus intravenous endotoxin (LPS) injection and continuing until 6 h after the injection. Blood samples were collected and leukocytes were isolated at multiple time points over 24 h both before and after LPS administration.

Microarray data analysis

Data analysis included extraction of expression values, identification of treatment group members, filtering for probe sets of interest and clustering. DNA chip analyzer (dChip) software was used to generate expression values from the microarray data, using invariant-set normalization and perfect match modeling. Probe sets of interest were identified using software for the extraction and analysis of gene expression (EDGE).²¹ EDGE was used to determine probe sets differentially expressed over time for all participants.²² Principal component analysis (PCA) was then performed using these probe sets, and the first principal component (PC1) was plotted versus time for each participant. The resulting projections were clustered in MATLAB using wavelet model-based clustering to identify the members of the two treatment groups. EDGE was then used to identify probe sets differentially expressed between the two treatment groups. Finally, the filtered data sets were clustered using ‘consensus clustering’, to identify subsets of the probe sets with common coherent expression patterns.²³ Functional annotation of these probe sets was done by using the pathway enrichment function in Ingenuity Pathway Analysis (IPA) tools (Ingenuity Systems, www.ingenuity.com), as well as analyzing functions of each individual gene corresponding to the probe sets, using databases of IPA and Kyoto Encyclopedia of Genes and Genomes.

Promoter analysis

Promoters of genes were extracted from a rich database of promoter information (Genomatix). In order to identify cortisol-responsive genes, we examined the presence of GC response elements (GREs) on gene promoters. Conserved regions across orthologous promoters of the selected genes were first identified and then the GRE binding site motif (TGTTCT²⁴) was scanned against those conserved promoter regions (detailed methodology in Nguyen et al.²⁵). The routine was applied to all genes associated with the clusters obtained by analyzing the −24 h to 0 h time period.

Results

This study aims to elucidate the effects of coupled dose and rhythm manipulation of plasma cortisol levels on the transcriptional changes that occur in response to acute LPS exposure in humans. The study design, which is schematically shown in Figure 1, involves two groups of healthy participants treated either with cortisol or saline infusion for 30 h. Both groups were administered a bolus dose of LPS 24 h after the initiation of infusion. The blood samples were collected from the participants at multiple time points and transcriptional response of leukocytes were determined via microarray analysis. In addition, blood cell counts and plasma cortisol concentrations were monitored.

Plasma cortisol levels, total number of leukocytes and leukocyte differentials

Baseline (before −24 h) plasma cortisol concentrations were 10–15 µg/dl for both the saline + LPS and cortisol + LPS groups. After cortisol infusion commenced at −24 h, plasma cortisol levels increased to ∼ 35 µg/dl in the cortisol + LPS group, but fluctuated between 2 to 15 µg/dl levels in the saline + LPS group, as expected from its natural circadian rhythm (Figure 2). After LPS administration at 0 h, there was no further elevation in the plasma cortisol concentrations for cortisol + LPS group; however, for the saline + LPS group, there was a steep increase in the cortisol concentration from a mean of 10 µg/dl up to 23 µg/dl within the first 3 h of LPS administration, after which plasma cortisol levels started to return to normal. For the cortisol + LPS group, cortisol concentration fluctuated around 30–35 µg/dl until the infusion period was ended 6 h after LPS administration and returned to normal in the following 7 h.

Figure 2.

Change in the level of plasma cortisol (A) and total number of WBCs (B) and percentage of WBC subpopulations (C–E) in response to continuous cortisol infusion and LPS administration (dashed line: saline + LPS; solid line: cortisol + LPS).

There was a significant difference between saline- and cortisol-treated groups in the total number of leukocytes at time 0. High concentration and altered rhythm of cortisol by continuous infusion resulted in a significant alteration in both the total number of leukocytes (WBCs) and the WBC differentials. The total number of cells at 0 h (before LPS administration) was at a level comparable to that observed after LPS administration for the saline-infused control group with a relative increase in the percentage of neutrophils, and a significant decrease in the percentage of lymphoctyes and monocytes (Figure 2C–E). Following LPS administration, both groups demonstrated an increase in the number of WBCs at a similar rate (Figure 2B). Interestingly, the leukocyte subpopulations as a percentage of the total leukocyte count were not significantly different by 3 h post LPS administration.

Identification of pre-treatment groups from gene expression data

We performed an initial evaluation of the microarray data in which we allocated each participant into one of the two groups according to the gene expression patterns, while initial assignments of the participants were kept anonymous. For this purpose, first, the probe sets, which were differentially expressed over time for all participants, were identified (P < 0.01, q < 0.01), resulting in 799 probe sets of interest out of 8793 total. Then, PCA was performed on the differentially expressed probe sets. The first principal component (PC1), which accounted for ∼ 50% of the total variance, was plotted over time for each of the nine participants and the resulting profiles were then clustered. Averaged PC1 profiles of two resulting clusters are shown in Figure 3. The final assignment according to these clusters, which was later verified by the physicians who undertook the study, placed four participants in one group (saline pre-treated, dashed line) and five in the other (cortisol pre-treated, solid line).

Figure 3.

Averaged first principal components (PC1) for saline- (dashed line) and cortisol- (solid line) treated groups.

Transcriptional effects of manipulated plasma cortisol levels

After the initial evaluation of the data, the focus of the analysis was narrowed down to the first 24 h, where the response to cortisol infusion was independent from LPS exposure. In the time period between −24 to 0 h, we identified 332 out of total 8793 probe sets to be differentially expressed between cortisol- and saline-treated subjects (P < 0.05 and q < 0.05). According to their temporal expression profiles, 200 out of 332 of the probe sets were classified into six clusters by consensus clustering analysis.²⁵ Resulting clusters were further grouped according to whether cortisol group shows up- or down-regulation with respect to saline group, as shown in Figure 4.

Figure 4.

Transcriptional responses of leukocytes to cortisol prior to endotoxin administration in time period between −24 h and 0 h.

In the cortisol-treated group, 120 probe sets were up-regulated and 80 probe sets were down-regulated compared to control. Table 1 lists representative genes from the up- and down-regulated groups of clusters together with the functional groups that they are associated with in the context of this study. A more detailed list of probe sets within the clusters is given in Supplementary Table S1. The temporal profiles and annotated functions of the up-regulated probe sets reveal a progressively increasing sensitization in response to continuous presence of cortisol. Numerous cytokine and pattern recognition receptors, together with receptor regulatory elements and signal transduction elements, are up-regulated, whereas protein translation machinery and mitochondrial complex elements are down-regulated. In addition to the down-regulation of protein translation, two genes (UB2B, UBE2D1) encoding for enzymes with protein degradation function were up-regulated, indicating an overall catabolic effect on total cellular protein levels. Up-regulation of cytoskeletal proteins might indicate enhanced capability to phagocytose and increased motility. In line with this, extracellular matrix degradation enzymes are also up-regulated, which points to the fact that leukocytes are becoming ready to extravasate and be recruited to the site of a potential infection.

Table 1.

Functional annotation and representative probe sets associated with up- and down-regulated clusters in response to continuous cortisol infusion in the −24 h to 0 h time period.

Classification	Molecules
Up (a)	Cytokine and pattern recognition receptors	CSF2RA, IL-10RB, IL-13RA1, IL-1R2, IL-4R, TNFRSF10C, TNFRSF1A, TNFRSF9, TLR1, TLR8
	Receptor regulatory elements	IL-18RAP, GRB10, IRS2
	Signal transduction proteins	NFKBIA, CAMK1D, LIMK2, MAP2K4, MAP2K6, PPP2R5A
	Transcription factors	ATF6, FOS, STAT5B
	Negative regulators of complement system	CD55, CD59, SERPINB1
	Cytoskeletal proteins	IQGAP1, GIT2, TUBA1A
	Protein degradation enzymes	UB2B, UBE2D1
	Extracellular matrix degradation enzymes and their inhibitor	MMP25, MMP9, TIMP2
	Amino acid degradation	ARG1
Down (b)	Protein translation regulatory elements and translation machinery	EEF2, EIF3H, CCNC, RPSs (5-probe sets), RPLs (8-probe sets)
	Mitochondrial proteins	NDUFs (3-probe sets), UQCRFS1
	Free radical scavenging	SOD1
	Transcription factors	LEF1, STAT4
	Antigen presentation	HLA–DPB1

Within the genes associated with these clusters in the −24 h to 0 h period, we performed a promoter analysis by scanning for the GRE binding site (TGTTCT motif) on conserved regions across sets of orthologous promoters. Among 200 genes, there were 82 genes that contained a GRE-binding site, 50 genes with no GRE-binding site and 68 genes that had to be excluded owing to limitation of (orthologous) promoter information. Genes that are identified to have GRE binding sites within their promoter regions did not show any particular directionality in regulation. Distribution of the genes to up- and down-regulated clusters had almost equal weights. The genes included in the promoter analysis are listed in Supplementary Tables S2 S3, together with their distribution to up- and down-regulated clusters and basic classifications of corresponding proteins.

Overall response to LPS

After concentrating on the effects of cortisol on leukocyte transcription dynamics, the analysis was extended to cover the entire duration of the study in order to comprehensively evaluate the response to LPS and how it is affected by preceding cortisol infusion. One hundred and ninety-nine out of 8793 probe sets were differentially expressed between time points −24 h and 24 h for the two treatment groups (P < 0.05 and q < 0.05). One hundred and fifty-seven out of 199 probe sets were distributed among five clusters according to their expression patterns, which were further grouped according to the direction of the response in the first 24 h of the study as LPS-dominant, cortisol-enhanced and cortisol-suppressed (Figure 5A–C).

Figure 5.

Overall differential response to LPS between cortisol + LPS and saline + LPS groups. Five clusters obtained by consensus clustering were grouped into three based on response to cortisol between −24 h and 0 h as (A) LPS-dominated (17 probe sets), (B) cortisol-enhanced (74 probe sets) and (C) cortisol-suppressed (66 probe sets).

Heatmaps of the probe sets included in these groups, shown in Figure 5B 5C, as well as their functional annotations, listed in Table 2, corresponded well with the previous analysis of the time period of −24 to 0 h, displaying up- and down-regulation in response to cortisol infusion, respectively. Response patterns after LPS administration (between 0 h and 24 h) were very similar in both groups across all the clusters. However, the magnitude of the up-regulation response observed after LPS administration was higher in the cortisol-pretreated group compared with control. The cluster shown in Figure 5A displays a unique pattern where the overall response seems to be driven predominantly by LPS, which was reflected as strong up-regulation within the first 2 h of LPS administration, and had a higher intensity for the cortisol treated group. The detailed list of the probe sets included in these clusters is given in Supplementary Table S4.

Table 2.

Functional annotation and representative probe sets associated with up- and down-regulated clusters in response to continuous cortisol infusion within the full time period of the study (–24 h to 24 h).

Classification	Molecules
LPS-dominated (a)	Cell–cell and cell–extracellular matrix interaction	ICAM1, THBS1
	Cytokines, chemokine, receptor antagonist	IL-1B, CCL20, CCL3, IL-1RN
	Signal transduction elements	NFKBIA, TNFAIP3
	Transcription regulator	EGR1
Cortisol-enhanced (b)	Cytokine, complement, scavenger receptors	CXCR1, IFNGR2, IL-13RA1, IL-18R1, IL-1R2, CD163
	Receptor regulatory proteins	GAB2, IL-1RAP, IRAK3, IRS2
	Signal transduction elements	DUSP1, GRB10
	Transcription factors	ATF6, BCL6, CEBPD, FOS
	Nucleotide metabolism	ENTPD1, ENTPD7, MAK, UPP1
	Chaperone and stress response proteins	FKBP5, GADD45A
	Amino acid degradation	ARG1
	Anti-apoptotic proteins	BCL2A1, NAIP
	Pro-apoptotic protein	CASP5
	Acute phase protein	HP
	Complement regulatory protein	CD59
Cortisol-suppressed (c)	Transcription and protein translation	POLR1C, RPSs (5-probe sets), RPLs (9-probe sets), EEF2, EIF3H
	Protein degradation	PSMC5
	Mitochondrial proteins	UQCRH, COX4I1, NDUFs (3-probe sets)
	Antigen presentation	HLA–DA1, HLA–DRA
	Antigen recognition	LAT
	Free radical scavenging	SOD1
	Receptors	CD48, FPR3
	Nucleotide metabolism	ADSL, NME1
	Citric acid cycle enzyme	IDH3B

Discussion

The design of the current study involved blunting the natural circadian rhythm of cortisol in healthy volunteers by continuous cortisol infusion for 30 h. This increased and maintained plasma cortisol concentrations at a level associated with major physiologic stress.²⁶ At hour 24 of this continuous infusion period, LPS was administered as a bolus injection. Transcriptional response of leukocytes was monitored by microarrays, and data were analyzed specifically for cortisol infusion period in conjunction with the whole time period. Both GCs^27–29 and LPS^17,30–32 are known to alter the total and differential peripheral blood leukocyte counts. As gene expression analysis in this study was conducted on peripheral blood leukocytes without any process for separating specific cell subtypes, observed transcriptional changes represent an omnibus response that reflects both cellular responses and changes in subpopulations of cells in peripheral blood.

Plasma cortisol levels showed the expected circadian pattern²⁶ in control participants and rapidly increased following LPS administration. In cortisol-infused participants, cortisol levels were steady throughout the 30 h infusion period around twice the normal daily peak level of control participants (∼ 35 µg/dl), and did not show any further increase upon LPS administration. Cortisol concentration returned to normal levels by 12 h in both groups. This pattern shows that the followed dosing strategy successfully blunted the natural rhythm of cortisol and clamped the concentration at a level associated with stress-induced physiologic conditions. The number of leukocytes was significantly increased in response to continuous cortisol treatment before LPS administration, and there was an additional increase immediately following LPS administration at a similar extent observed in the control participants. This contrasts with previous stress-related studies on rats showing an inverse relationship between plasma corticosterone levels and total number of leukocytes.^27,28 One reason for this, apart from obvious difference in species, might be related to the way in which GC levels are manipulated. Animal studies are based on increasing endogenous corticosterone levels by exposing the animals to restrain stress for a short amount of time (1–2 h) to study the acute changes in leukocyte distributions. However, in our study design, cortisol concentrations were fixed at a level associated with stress by exogenous infusion of cortisol. As it has been already shown that not only the concentration, but also pulsatility, duration and timing of cortisol can affect physiological processes in virtually all the organs,⁷ this major difference in study designs can be the reason for observing contradictory results.

Transcriptional responses within the first 24 h of the study reflect a number of cortisol-induced changes, the most important of which are likely to be the changes in functional gene regulation and in the leukocyte subpopulations. As a consequence of the continuous cortisol infusion, we observed enhanced expression of genes encoding for a number of pattern recognition and cytokine receptors, receptor regulatory elements and signal transduction elements, as well as reduced expression of the genes encoding for elements of protein translation process, and mitochondrial proteins. Increased receptor and signal transduction protein expression were indicative of a priming effect of the cortisol on the immune function, where cells were sensitized to recognize potential infectious threats or endogenous danger signals. These observations agree well with the previous studies investigating the effects of stress or direct cortisol delivery on the immune function and response to LPS challenge which demonstrated amplified pro-inflammatory receptor expression and function on immune cells^33–35 and subsequently enhanced inflammatory response to LPS.^{8,9,13,36–38}

Up-regulation of cytoskeletal proteins and extracellular matrix degradation enzymes also imply an increased mobility and enhanced ability for extravasation to penetrate the tissues or engulf pathogens. These are key events for the inflammatory function of leukocytes and the enhancing effect of cortisol on these functions supports its permissive role on inflammatory response as supported by earlier studies.^13,39

Tuning down the protein translation process can also be regarded as part of the priming effect as it points toward re-prioritization of cellular resources and energy, which would presumably allow better energy management when the anticipated threat is encountered. Decreased protein synthesis together with the increased expression of protein degradation enzymes indicates a catabolic state that is characteristic of cortisol action and in line with its role as a stress hormone.^7,40,41 Overall, cortisol infusion seems to induce an anticipatory state in the immune cells and have preparatory effects to increase the competency in combatting with the anticipated threat. Although this might seem counter-intuitive considering the fame of GCs for their immunosuppressive and anti-inflammatory actions utilized in clinic, depending on the timing, duration and concentration of the exposure, cortisol is shown to also have permissive and enhancing effects on immune function.^{7,8,13,35,42,43}

Promoter analysis within these genes associated with altered regulation due to constant cortisol concentration indicated that, mechanistically, only less than half of differential expression patterns might have been as a direct consequence of GC receptor (GR)-DNA binding. However, a quarter of the genes do not contain any GRE-binding site in their promoter regions, pointing to the previously recognized regulatory function of GR on other transcription factors via mechanisms other than direct DNA binding, such as tethering.^44–46 Furthermore, distribution of these genes with or without GRE-binding site in their promoters within up- or down-regulated clusters indicated that there is possibly no preferential directionality in different modes of transcriptional regulation through GR. This underscores the complexity of mechanisms that govern the transcriptional programming through GR,^47,48 which is unsurprising considering the elicited vastly different functional consequences of GCs in the cellular milieu.^7,49,50

Despite the significant transcriptional program changes in the leukocytes prior to 0 h, the overall response to LPS did not significantly differ between cortisol- and saline-treated groups, as reflected in the clusters obtained from differentially expressed genes throughout the full study period. Both cortisol- and saline-treated groups displayed similar up- and down-regulation responses as reflected in the 0–24 h periods of the heatmap shown in Figure 5(B, C). These groups of clusters correspond well to the first part of the study, and are therefore named according to the observed response to cortisol as ‘cortisol enhanced’ (Figure 5B) and ‘cortisol suppressed’ (Figure 5C). However, the expression pattern in the cluster shown in Figure 5A substantially differs from the rest. Intensity change within the first 2 h of LPS administration is so steep in this cluster that, even if cortisol effect was present prior to LPS administration, it is overridden by this steep change. The important pro-inflammatory NFKBIA gene, which was in the up-regulated cluster in Figure 4 (and listed in Table 1) for the previous clustering analysis, appears as neutral in this analysis in the same time period. The reason for this is that NFKBIA is a gene predominantly controlled by pro-inflammatory stimuli and although its expression is moderately up-regulated as a result of continuous cortisol infusion, the intense up-regulation after LPS is predominant. Furthermore, the fact that LPS response was much higher for this cluster in the cortisol group when compared with control suggests that while cortisol might not have directly affected transcription of the genes in this cluster, it may have had an indirect effect, perhaps through the modulation of up-stream signaling cascades. The net result is that the response to the subsequent inflammatory challenge (i.e. LPS) was intensified.

As noted above, at the 0 time point, there were significant differences in both the total leukocyte count and the composition of the leukocyte subpopulations. In particular, the subpopulation of neutrophils was larger in the cortisol-treated group (90% of the total leukocyte count) when compared with the control group (55%). This is an important consideration because studies from our own group^20,31 and from others³⁶ indicate that the reported change in whole blood inflammatory gene transcription/translation at early time points (2–4 h) after LPS administration can be attributed mainly to circulating neutrophils. In the present study, however, the leukocyte subpopulations (as percent of the total leukocyte count) were not significantly different by 3 h post-LPS when cortisol and control groups are compared, making it unlikely that the robust LPS response observed in this gene cluster (Figure 5A) can be ascribed simply to the altered cell differential numbers. Rather, the changes may reflect alterations in the maturity of circulating neutrophils (e.g. CD16dim⁵¹) or the bone marrow recruitment of neutrophils following LPS challenge in the presence of cortisol. One of the challenges of this study is that the subpopulations of cells were not further investigated or isolated to examine their unique transcription signatures.

The probe sets associated with the LPS-dominant cluster include the major players known to regulate the inflammatory response, such as cytokines and chemokines (Il1B, CCl20, CCL3 and IL1RN), signal transduction elements that negatively regulate pro-inflammatory signaling (NFKBIA and TNFAIP3), and receptors mediating recruitment of leukocytes to sites of inflammation (ICAM, THBS1). Together, these results suggest that continuous cortisol infusion primes circulating and bone marrow leukocytes, increases the efficiency of their surveillance function and dampens their energy-intense cellular processes so that they can more rapidly and efficiently respond to a subsequent inflammatory challenge. Although these changes in the transcriptome coherently point out to the permissive roles of cortisol on circulating leukocytes, this study is limited to describe the resulting phenotype and how inflammatory processes are actually controlled and implemented. It is worth noting that tuning down the protein translation processes by continuous cortisol treatment also inherently implies that produced transcripts might not exactly mirror the protein end products, such as inflammatory cytokines. Therefore, to fully describe the effect of cortisol on immune response, further studies focusing on proteomic and phosphoproteomic profiles of leukocytes, as well as on plasma metabolomics are essential.

Conclusions

Natural daily fluctuation of plasma cortisol concentration in healthy volunteers was blunted by continuous cortisol infusion prior to an immune challenge with a bolus dose of LPS. Transcriptional responses of leukocytes, as well as the change in total leukocyte count in blood and plasma cortisol levels, were monitored over 24 h both before and after LPS administration. By continuous infusion, plasma cortisol concentration was increased to, and kept constant at, around 35 µg/dl, which is a level associated with major physical stress and almost twice as much as the normal daily peak. This treatment induced a significant increase in the total number of leukocytes and altered their transcriptional programming. A priming response of leukocytes at the end of the 24-h infusion period associated with enhanced expression of a number of pattern recognition and cytokine receptors, receptor regulatory and signal transduction elements, as well as tuned-down protein translation and mitochondrial function was observed. Promoter analysis within cortisol-responsive genes indicated that only less than half of this response might have been as a direct consequence of GR–DNA binding; therefore, transcriptional regulation mechanisms other than DNA binding, such as tethering, plays a significant role. Furthermore, absence of any correlations between the direction of the response and different modes of transcriptional regulation underscored the complexity of mechanisms that govern the transcriptional programming through GR. Despite these significant changes prior to LPS administration, response to LPS followed very similar patterns in both cortisol and saline pre-treated groups except one significantly different cluster including probe sets associated with major players known to regulate the inflammatory response such as cytokines and chemokines, anti-inflammatory signal transduction elements and receptors mediating recruitment of leukocytes to the sites of inflammation. The much more intense response observed in the cortisol group in this cluster indicated that although continuous cortisol infusion did not directly affect the transcription of major inflammatory players, it indirectly affected their regulation, possibly through modulating their up-stream signaling cascades. Overall, coupled dose and rhythm manipulation of plasma cortisol levels promoted the priming of circulating leukocytes associated with sensitization of their surveillance function and tuning down of energy-intensive processes which may be linked to increasing their efficiency in responding to subsequent infectious threats.

Footnotes

Funding

The authors acknowledge support from NIH GM082974 and NIH GM34695.

References

Mavroudis

Scheff

Calvano

Androulakis

. Systems biology of circadian-immune interactions. J Innate Immun 2013; 5: 153–162.

de Kloet

Sarabdjitsingh

. Everything has rhythm: focus on glucocorticoid pulsatility. Endocrinology 2008; 149: 3241–3243.

Lightman

Conway-Campbell

. The crucial role of pulsatile activity of the HPA axis for continuous dynamic equilibration. Nat Rev Neuroscience 2010; 11: 710–718.

Scheff

Calvano

Lowry

Androulakis

. Transcriptional implications of ultradian glucocorticoid secretion in homeostasis and in the acute stress response. Physiol Genomics 2012; 44: 121–129.

Young

Abelson

Lightman

. Cortisol pulsatility and its role in stress regulation and health. Front Neuroendocrinol 2004; 25: 69–76.

Munck

Guyre

. Glucocorticoid physiology, pharmacology and stress. Adv Exp Med Biol 1986; 196: 81–81.

Sapolsky

Romero

Munck

. How do glucocorticoids influence stress responses? Integrating permissive, suppressive, stimulatory, and preparative actions. Endocr Rev 2000; 21: 55–89.

Lim

Müller

Herold

. Glucocorticoids exert opposing effects on macrophage function dependent on their concentration. Immunology 2007; 122: 47–53.

Frank

Miguel

Watkins

Maier

. Prior exposure to glucocorticoids sensitizes the neuroinflammatory and peripheral inflammatory responses to E. coli lipopolysaccharide. Brain Behav Immun 2010; 24: 19–30.

10.

Barber

Coyle

Marano

. Glucocorticoid therapy alters hormonal and cytokine responses to endotoxin in man. J Immunol 1993; 150: 1999–2006.

11.

Yeager

Rassias

Pioli

. Pretreatment with stress cortisol enhances the human systemic inflammatory response to bacterial endotoxin. Crit Care Med 2009; 37: 2727–2732.

12.

Ricciardi-Castagnoli

Granucci

. Interpretation of the complexity of innate immune responses by functional genomics. Nat Rev Immunol 2002; 2: 881–887.

13.

Galon

Franchimont

Hiroi

. Gene profiling reveals unknown enhancing and suppressive actions of glucocorticoids on immune cells. FASEB J 2002; 16: 61–71.

14.

Ehrchen

Steinmüller

Barczyk

. Glucocorticoids induce differentiation of a specifically activated, anti-inflammatory subtype of human monocytes. Blood 2007; 109: 1265–1274.

15.

Deuschle

Schweiger

Weber

. Diurnal activity and pulsatility of the hypothalamus-pituitary-adrenal system in male depressed patients and healthy controls. J Clin Endocrinol Metab 1997; 82: 234–238.

16.

Vaughan

Becker

Allen

. Cortisol and corticotrophin in burned patients. J Trauma Acute Care Surg 1982; 22: 263–273.

17.

Lowry

. Human endotoxemia: a model for mechanistic insight and therapeutic targeting. Shock 2005; 24(Suppl. 1): 94–100.

18.

Haimovich

Calvano

Haimovich

. In vivo endotoxin synchronizes and suppresses clock gene expression in human peripheral blood leukocytes. Crit Care Med 2010; 38: 751–758.

19.

Hawes

Fischer

Marano

. Comparison of peripheral blood leukocyte kinetics after live Escherichia coli, endotoxin, or interleukin-1 alpha administration. Studies using a novel interleukin-1 receptor antagonist. Ann Surg 1993; 218: 79–90.

20.

Calvano

Xiao

Richards

. A network-based analysis of systemic inflammation in humans. Nature 2005; 437: 1032–1039.

21.

Leek

Monsen

Dabney

Storey

. EDGE: extraction and analysis of differential gene expression. Bioinformatics 2006; 22: 507–508.

22.

McDunn

Husain

Polpitiya

. Plasticity of the systemic inflammatory response to acute infection during critical illness: development of the riboleukogram. PLoS ONE 2008; 3: e1564–e1564.

23.

Nguyen

Nowakowski

Androulakis

. Unsupervised selection of highly coexpressed and noncoexpressed genes using a consensus clustering approach. OMICS 2009; 13: 219–237.

24.

Almon

Lai

DuBois

Jusko

. Corticosteroid-regulated genes in rat kidney: mining time series array data. Am J Physiol Endocrinol Metab 2005; 289: E870–E882.

25.

Nguyen

Almon

DuBois

. Comparative analysis of acute and chronic corticosteroid pharmacogenomic effects in rat liver: transcriptional dynamics and regulatory structures. BMC Bioinformatics 2010; 11: 515–515.

26.

Yeager

Pioli

Guyre

. Cortisol exerts bi-phasic regulation of inflammation in humans. Dose Response 2011; 9: 332–347.

27.

Dhabhar

Miller

Stein

. Diurnal and acute stress-induced changes in distribution of peripheral blood leukocyte subpopulations. Brain Behav Immun 1994; 8: 66–79.

28.

Dhabhar

Miller

McEwen

Spencer

. Effects of stress on immune cell distribution. Dynamics and hormonal mechanisms. J Immunol 1995; 154: 5511–5527.

29.

Calvano

Barber

Hawes

. Effect of combined cortisol-endotoxin administration on peripheral blood leukocyte counts and phenotype in normal humans. Arch Surg 1992; 127: 181–186.

30.

Talwar

Munson

Barb

. Gene expression profiles of peripheral blood leukocytes after endotoxin challenge in humans. Physiol Genomics 2006; 25: 203–215.

31.

Richardson

Rhyne

Fong

. Peripheral blood leukocyte kinetics following in vivo lipopolysaccharide (LPS) administration to normal human subjects. Influence of elicited hormones and cytokines. Ann Surg 1989; 210: 239–245.

32.

de Kleijn

Kox

Sama

. Transcriptome kinetics of circulating neutrophils during human experimental endotoxemia. PLoS ONE 2012; 7: e38255–e38255.

33.

Shieh

Peterson

Moore

. Cytokines and dexamethasone modulation of IL-1 receptors on human neutrophils in vitro. J Immunol 1993; 150: 3515–3524.

34.

Hawrylowicz

Guida

Paleolog

. Dexamethasone up-regulates granulocyte-macrophage colony-stimulating factor receptor expression on human monocytes. Immunology 1994; 83: 274–274.

35.

Zhang

Daynes

. Glucocorticoid conditioning of myeloid progenitors enhances TLR4 signaling via negative regulation of the phosphatidylinositol 3-kinase-Akt pathway. J Immunol 2007; 178: 2517–2526.

36.

Yeager

Rassias

Pioli

. Pretreatment with stress cortisol enhances the human systemic inflammatory response to bacterial endotoxin. Crit Care Med 2009; 37: 2727–2727.

37.

Johnson

O'Connor

Deak

. Prior stressor exposure sensitizes LPS-induced cytokine production. Brain Behav Immun 2002; 16: 461–476.

38.

Gundersen

Opstad

Reistad

. Seven days’ around the clock exhaustive physical exertion combined with energy depletion and sleep deprivation primes circulating leukocytes. Eur J Applied Physiol 2006; 97: 151–157.

39.

van der Goes

Hoekstra

van den Berg

Dijkstra

. Dexamethasone promotes phagocytosis and bacterial killing by human monocytes/macrophages in vitro. J Leukoc Biol 2000; 67: 801–807.

40.

Shah

Kimball

Jefferson

. Acute attenuation of translation initiation and protein synthesis by glucocorticoids in skeletal muscle. Am J Physiol Endocrinol Metab 2000; 278: E76–E82.

41.

Shah

Iniguez-Lluhi

Romanelli

. The activated glucocorticoid receptor modulates presumptive autoregulation of ribosomal protein S6 protein kinase, p70 S6K. J Biol Chem 2002; 277: 2525–2533.

42.

Munck

Naray-Fejes-Toth

. Glucocorticoids and stress: permissive and suppressive actions. Ann N Y Acad Sci 1994; 746: 115–130.

43.

Guyre

Yeager

Munck

. Glucocorticoid effects on immune responses. Neuroimmune Biol 2007; 7: 147–167.

44.

Buckingham

. Glucocorticoids: exemplars of multi-tasking. Br J Pharmacol 2006; 147(Suppl. 1): S258–S268.

45.

Miranda

Morris

Hager

. Complex genomic interactions in the dynamic regulation of transcription by the glucocorticoid receptor. Mol Cell Endocrinol. Epub ahed of print 14 March 2013.

46.

Rao

McCalman

Moulos

. Coactivation of GR and NFKB alters the repertoire of their binding sites and target genes. Genome Res 2011; 21: 1404–1416.

47.

Surjit

Ganti

Mukherji

. Widespread negative response elements mediate direct repression by agonist-liganded glucocorticoid receptor. Cell 2011; 145: 224–241.

48.

Uhlenhaut

Barish

. Insights into negative regulation by the glucocorticoid receptor from genome-wide profiling of inflammatory cistromes. Mol Cell 2013; 49: 158–171.

49.

Goulding

. The molecular complexity of glucocorticoid actions in inflammation – a four-ring circus. Curr Opin Pharmacol 2004; 4: 629–636.

50.

Busillo

Cidlowski

. The five Rs of glucocorticoid action during inflammation: ready, reinforce, repress, resolve, and restore. Trends Endocrinol Metab 2013; 24: 109–119.

51.

Pillay

Ramakers

Kamp

. Functional heterogeneity and differential priming of circulating neutrophils in human experimental endotoxemia. J Leukoc Biol 2010; 88: 211–220.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.17 MB