Introduction
Septic shock, the most severe stage of the septic process, is a deadly disease defined as severe sepsis with hypotension that persists after resuscitation with intravenous fluid [
1,
2]. Critically ill patients may display a state of functional immune system suppression, which predisposes them to infections with opportunistic bacteria found in the ICU environment [
3] and results in excess morbidity and mortality.
To identify genes and/or cellular pathways involved in the pathogenesis of ICU-acquired sepsis, we initially undertook leukocyte genomic expression analysis of polytrauma patients who developed severe sepsis during their ICU stay as compared with trauma patients who suffered from sepsis not escalating to a higher septic stage [
2,
4]. Following robust statistical selection of differentially expressed genes, results were compared with publicly available microarray datasets, thus creating a unique list of likely disease modifiers. Among these genes, aquaporin-1 (Aqp1), a water channel protein [
5,
6], was one of the three genes found deregulated in all datasets.
Aqp1, a 28 kDa integral membrane protein [
7], exists as a tetramer with intracellular N-termini and C-termini – an organization similar to several ion channel proteins [
8]. The functional significance of Aqp1 in migratory cells lies presumably in cell volume regulation and cytokinesis [
9,
10]. Aquaporin-9, an isoform of Aqp1, is constitutively expressed and regulates water influx at the leading edge of migrating neutrophils [
11]. However, the role of Aqp1 in the molecular events taking place after leukocyte activation by pathogens remains unclear.
Based on the findings of the genomic expression analysis of our initial polytrauma patient group, we hypothesized that Aqp1 expression is induced in leukocytes of critically ill patients afflicted with nosocomial sepsis in a manner dependent on sepsis severity. To address this hypothesis and to study potential underlying mechanisms, we measured levels of leukocyte Aqp1 expression in critically ill patients at ICU admission, upon diagnosis of sepsis and at occurrence of septic shock; in additional in vitro experiments we probed the capacity of lipopolysaccharide (LPS) to regulate Aqp1 expression in polymorphonuclear granulocytes (PMNs) and investigated potential signaling pathways and functions.
Materials and methods
The study was approved by the Evangelismos Hospital Research Ethics Committee and all procedures carried out on patients were in compliance with the Helsinki Declaration. Informed written consent was obtained from all patients’ next-of–kin prior to any study procedure.
Gene expression profiling study population
We initially screened consecutive admissions of polytrauma patients to the Evangelismos Hospital ICU, Athens, Greece. Exclusion criteria were: no consent to participate, sepsis upon ICU admission, body mass index >35 kg/m
2, age <18 years, pregnancy, brain death, end-stage cancer, total ICU stay <3 days, re-admission or transfer from another ICU, contagious diseases (human immunodeficiency virus, hepatitis), and oral intake of corticosteroids at an equivalent dosage ≥1 mg/kg prednisone/day for a period >1 month. The first five initially nonseptic patients (one female), who subsequently became septic were included in the study. Two blood samples were obtained per patient: the first upon admission to the ICU, and the second within 48 hours of sepsis. Total RNA was isolated and cRNAs were hybridized to Affymetrix Hu133A 2.0 GeneChips Affymetrix, Santa Clara, CA, USA according to the manufacturer’s instructions. This dataset has been deposited in ArrayExpress [ArrayExpress:E-MEXP-3621]. All methods employed in the aforementioned gene expression profiling study are described in detail in Additional file
1.
Aqp1 gene expression in septic ICU patients: study population
Following the initial microarray study stated above, we sought to investigate Aqp1 leukocyte expression in critically ill adult, initially nonseptic, patients who were admitted to the multi-disciplinary ICU of Evangelismos Hospital and who subsequently developed sepsis and deteriorated to septic shock; an additional group of patients who did not develop sepsis during their stay in the ICU served as the control.
Prior to enrollment, we screened all consecutive admissions over a 12-month period for eligibility. Exclusion criteria were the same as those used for our initial gene expression profiling study population (above). Of the 118 subjects screened over the study period, finally 29 medical, surgical and trauma patients were enrolled in the study, 16 of whom developed septic shock and 13 who served as our control patients (did not develop sepsis during their stay in the ICU). Clinical data and blood samples were obtained from all patients enrolled. Patients were considered to have sepsis when they developed systemic inflammatory response syndrome as a result of documented infection, septic patients with evidence of organ dysfunction were considered to have severe sepsis, and septic patients with persisting hypotension (despite adequate fluid resuscitation) were considered to have septic shock, in accordance with international guidelines and recommendations [
2,
4].
Following study enrollment, baseline (upon ICU admission) anthropometric data and detailed organ system-oriented medical history were recorded. A venous blood sample was drawn at three different time points: at the time of ICU admission (baseline), at the onset of sepsis (within 6 to 12 hours) and upon septic shock. In patients who did not develop sepsis, blood samples were drawn at admission and at ICU discharge only. The age of the septic patients ranged from 21 to 82 years whereas the age of the patients that did not develop sepsis varied from 31 to 82 years. The characteristics of the abovementioned patients are presented in Table
1.
Table 1
Characteristics of the two ICU patient groups
Number of patients | 16 | 13 | |
Age (years) | 47 ± 18 | 57 ± 17 | 0.12 |
Gender | | | |
Male | 13 (81.3%) | 6 (46.2%) | 0.06 |
Female | 3 (18.7%) | 7 (53.8%) |
Diagnosis | | | |
Medical | 5 (31.3%) | 5 (38.5%) | 0.71 |
Surgical or trauma | 11 (68.7%) | 8 (61.5%) |
ICU mortality | 6 (37.5%) | 0 (0%) | 0.02 |
APACHE II score | 15.3 ± 6.0 | 14.6 ± 5.0 | 0.76 |
SOFA score | 8.3 ± 2.3 | 5.2 ± 1.8 | <0.001 |
Sepsis day | 4.2 ± 1.8 | N/A | |
Septic shock day | 12.7 ± 10.0 | N/A | |
Chemicals and reagents
Whole blood was collected in Tempus Blood RNA tubes (Applied Biosystems, Life Technologies Corporation, Carlsbad, CA, USA). RNA extraction was performed with the Tempus spin RNA isolation kit (Applied Biosystems, Life Technologies Corporation). Taq DNA polymerase was purchased from Fermentas (Thermo Fisher Scientific Inc., Waltham, MA, USA), reverse transcriptase from New England Biolabs (Ipswich, MA, USA), the SYBR-Green assay from Finnzymes (Thermo Fisher Scientific, Waltham, MA, USA), and the nuclear extraction kit and the NF-κB (p65) transcription factor assay kit from Cayman Chemical Company (Ann Arbor, MI, USA). LPS, EF-24, dithiothreitol (DTT) and phenylmethylsulfonyl fluoride were from Sigma (Sigma-Aldrich Corp, St Louis, MO, USA). The PCR products were analyzed by agarose gel electrophoresis. All reagents used were of analytical grade.
Blood collection
Three milliliters of venous blood were collected within the first 24 hours post ICU admission, at the onset of sepsis (within 6 to 12 hours) and upon septic shock or discharge from the ICU. Samples were collected in tubes containing RNA stabilizing solution (Applied Biosystems, Life Technologies Corporation).
Total RNA extraction and evaluation of its quality
Total RNA was isolated from total blood cells (>95% leukocytes) using the Tempus spin RNA isolation kit (Applied Biosystems, Life Technologies Corporation) and following the manufacturer's instructions. Total RNA concentration and quality were determined spectrophotometrically at 260 and 280 nm, while RNA integrity was evaluated with formaldehyde agarose gel electrophoresis. Total RNA was stored at −80°C until use.
Reverse transcription and PCR
One microgram of total RNA from each sample was reverse transcribed into single-stranded cDNA in a 20 μl reaction mixture, using the M-MuLV Reverse Transcriptase (New England Biolabs, Ipswich, MA, USA) and Oligo(dT) primers, following the manufacturer's instructions. The success of the synthesis of the single-stranded cDNA was tested by its PCR amplification. PCR was performed using 1.0 μl cDNA, 1.5 mM MgCl2, 400 μM dNTPs, 500 nM primers, 1.5 U Dream-Taq DNA polymerase and 1× reaction buffer (Fermentas, Thermo Fisher Scientific Inc.), on a PTC-200 thermocycler (MJ Research Inc., Waltham, MA, USA). Equal amounts (10 μl) of the 108 bp glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and the 177 bp Aqp1 products were electrophoresed on a 2.5% agarose gel, visualized following ethidium bromide staining and photographed under ultraviolet light with a Kodak DC120 digital camera (Rochester, NY, USA).
Quantitative real-time PCR
A highly sensitive quantitative real-time PCR method has been developed for the quantification of both GAPDH and Aqp1 mRNAs, with the use of SYBR® Green Dye detection systems. For the amplification of GAPDH (endogenous reference gene) as well as Aqp1 (target gene) mRNA sequences, gene-specific sets of primers were designed according to the information on the NCBI Sequence database and the advanced software of the Primer Express program. To avoid genomic DNA amplification, the primers, mentioned above, were chosen to span at least two exons. The sequences of the primers used were: for GAPDH, the forward 5′-ATGGGGAAGGTGAAGGTCG-3′ primer, consisting of 19 nucleotides, and the reverse 23-nucleotide 5′-TACATGAGGGCACGGAAGATG-3′ primer, which gave rise to a 108 bp amplicon; for the Aqp1 gene, the forward 5′-GGTGGGGAACAACCAGACG-3′ primer and the reverse 5′-TACATGAGGGCACGGAAGATG-3′ primer, each with a length of 23 nucleotides, produced an amplicon of 177 bp.
Quantitative real-time PCR analysis was performed in 96-well plates on a PTC-200 thermocycler (MJ Research Inc.). The 25 μl reaction mixture contained 5 ng cDNA, 300 nM primers and 1× DyNAmo SYBR® Green PCR Master Mix (Finnzymes, Thermo Fisher Scientific), in which modified Tbr DNA polymerase is included. The thermal protocol conditions consisted of 2 minutes at 50°C and 10 minutes at 95°C polymerase activation step, 40 cycles of denaturation at 95°C for 15 seconds, primer annealing at 54°C for 30 seconds, extension at 72°C for 10 seconds and a final extension step at 72°C for 10 minutes. All samples were amplified in triplicate and the average CT values were calculated for their subsequent expression analysis. Following amplification, a dissociation curve was generated to distinguish the PCR products of interest from the nonspecific ones or any primer dimers, through their particular melting temperatures (Tm), recorded in the software. Using the comparative 2
−ΔΔCT method [
12] and the baseline samples as a calibrator, the relative quantification of the expression analysis of all blood samples was carried out. GAPDH expression was used for the normalization of Aqp1 expression levels between the different samples.
Polymorphonuclear granulocyte isolation
PMNs were isolated from peripheral blood of consecutive healthy volunteers – blood donors supplied by the Blood Bank of Evangelismos Hospital. Blood was collected in 17% citric acid and allowed to sediment in 3% Dextran/0.9% NaCl for 1 hour. Leukocyte-rich supernatant was separated by centrifugation on Histopaque (Sigma, Sigma-Aldrich Corp.). After lysis of residual erythrocytes by hypotonic shock in ice-cold distilled water, the cells were resuspended in Hank's balanced salt solution (Gibco, Life Technologies Corporation, Carlsbad, CA, USA) and stored on ice until use. The cells were counted and diluted to the desired concentration.
Culture of polymorphonuclear granulocytes
PMNs (2×106) were cultured in Hank’s balanced salt solution using six-well plates. LPS from Pseudomonas aeruginosa 10 (Sigma, Sigma-Aldrich Corp.) was added at a final concentration of 100 ng/ml in the absence or presence of 1 μM NF-κB inhibitor EF-24 (Sigma, Sigma-Aldrich Corp.). The cells were incubated for 3 hours at 37°C. Following incubation, the cells were harvested for RNA and/or protein isolation. Each experiment was independently performed five times.
RNA isolation from polymorphonuclear cells
RNA extraction was performed using the Trizol® reagent and the PureLink RNA Mini kit (Invitrogen, Life Technologies Corporation) following the manufacturer’s instructions. RNA was stored at −80°C until use. Reverse transcription and quantitative real-time PCR were performed as described above.
Polymorphonuclear cell homogenate
Approximately 2×106 cells were homogenized in PBS buffer with the addition of 5 μM DTT, 0.1 mM phenylmethylsulfonyl fluoride in isopropanol, and protease inhibitors, using a sonicator (Sonic Dismembrator 550; Fisher Scientific, Loughborough, UK) and 3×10 second bursts. The homogenate was centrifuged in a 5804 R Eppendorf centrifuge (rotor A-4-44; Eppendorf AG, Hamburg, Germany) for 5 minutes at 4°C and 1,000×g. The supernatant, referred to as total protein, was collected and stored at −80°C until use.
Protein determination
Total protein concentration was determined according to the method of bicinchonic acid [
13], using bovine serum albumin as standard.
SDS-polyacrylamide gel electrophoresis
SDS-PAGE was performed on a Biorad Mini Protean II electrophoresis apparatus (Bio-Rad, Hercules, CA, USA), as described [
14], using 12% polyacrylamide slab gels. Electrophoresis was carried out at 150 V for 1 hour at room temperature.
Immunoblotting of polymorphonuclear homogenates with Aqp1 antibody
Following electrophoresis, samples were transferred onto an Immobilon-P polyvinylidene difluoride membrane (Millipore, 0.45 μl pore size; Millipore Corporation, Billerica, MA, USA). Western transfer was performed on a wet transfer apparatus (Bio-Rad). Immunological detection was performed according to the method of Batteiger and colleagues [
15]. The polyclonal antibody against Aqp1 purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz Biotechnology, Inc., Dallas, TX, USA) was used for the detection.
Cell volume measurements
Osmotic swelling was monitored as in [
16] with slight modifications. Briefly, swelling was monitored at 22°C with a Zeiss Axiovert 25 phase contrast microscope (Carl Zeiss AG, Oberkochen, Germany) with two 10× eyepieces and a 40× objective on a trinocular base, 0.5-numerical aperture long working distance Canon objective equipped with a Canon digital camera (Canon USA Inc., Melville, NY, USA). PMN were transferred from 300 mosM (in) to 150 mosM (out) Hank’s balanced salt solution diluted with water. PMN images were taken at 30-second intervals for a total of 600 seconds or until the time of PMN rupture. The relative volume V/Vo, that is the ratio of measured volume at various time-points during the osmotic challenge to the baseline volume, was determined. Changes in volume were measured by ImageJ software (NIH, Bethesda, MD, USA).
Cytokine measurement
IL-8 levels were measured at the supernatants of the PMN cells by ELISA, according to the manufacturer’s instructions. The human IL-8 immunoassay of R&D Systems Inc. (Minneapolis, MN, USA) was used. This assay uses two different polyclonal antibodies against the cytokines as the catching antibody and the tagging antibody. Samples were assayed in triplicate.
Polymorphonuclear cell nuclear extraction
Nuclear extraction was performed with the kit by Cayman Chemical Company, according to the manufacturer’s instructions. In brief, approximately 2×106 cells were homogenized in pre-extraction buffer containing DTT and protease inhibitors. The cytoplasmic extract was removed from the nuclear pellet, which was treated further with extraction buffer containing DTT and protease inhibitors. The extract was incubated on ice for 15 minutes with vortex (5 seconds) every 3 minutes. The suspension was centrifuged in an Espresso centrifuge (Thermo Fisher Scientific Inc.) for 10 minutes at 9,800×g at 4°C and transferred into a new microcentrifuge vial. The supernatant, referred to as nuclear extract, was used for the measurement of NF-κB activation. Total protein concentration was determined by the bicinchonic acid method, as mentioned in the Protein determination section above.
NF-κB transcription factor activation measurement
The activation of NF-κB transcription factor was measured for the nuclear extracts of PMN cells by the Transcription Factor Assay kit (Cayman Chemical Company), according to the manufacturer’s instructions. A specific double-stranded DNA sequence containing the NF-κB response element is immobilized onto the bottom of wells of a 96-well plate. NF-κB contained in a nuclear extract binds specifically to the NF-κB response element. NF-κB (p65) is detected by addition of specific primary antibody directed against NF-κB (p65). A secondary antibody conjugated to horseradish peroxidase is added to provide a sensitive colorimetric readout at 450 nm. Samples were assayed in triplicate.
Statistical analysis
Data are presented as individual values, mean ± standard deviation for normally distributed variables and median with interquartile range for variables with skewed distribution. For continuous variables, the t test was used for two-group comparisons. Associations between qualitative variables were examined by the chi-square test. The Friedman test was used to examine differences within more than two groups, and the Wilcoxon signed-rank test for differences within two groups, one-way and two-way analyses of variance for repeated measures and Kruskal–Wallis analysis of variance followed by Newman–Keuls or Bonferroni’s post-hoc test were performed as appropriate. The Spearman’s rank correlation coefficient was calculated to measure the strength of relationship between levels of Aqp1 expression and time. All aforementioned analyses were performed using the Graphpad prism 5 statistical program (GraphPad Software, Inc., La Jolla, CA, USA). All P values are two-sided; P <0.05 was considered significant.
Discussion
Hospital-acquired sepsis is a major cause of mortality in critically ill patients. It is becoming increasingly evident that critical illness is associated with a state of functional immunosuppression [
3,
18,
19], possibly following an acute phase of disproportionate immune system activation. This paradigm has triggered interest in mechanistic understanding of immune system (dys)function in sepsis. In this context, we initially undertook gene expression profiling in a small sample of polytrauma patients, who subsequently acquired sepsis, which returned a large set of genes with differential expression between patients who developed sepsis versus severe sepsis. Owing to the limited size of this pilot study, and the variability inherent to this methodology, we compared our dataset with published work in animal models of acute insults relevant to the ICU population, such as LPS-induced lung injury, a common organ failure in severe sepsis, using the ortholog gene approach. We were thus able to identify Aqp1 as a gene differentially expressed in experimental animals and humans on a consistent basis, indicating a potential involvement in the acute phase response. Importantly, Aqp1 upregulation seems to take place not only in lung tissue [
20] but, as shown in our study, also in leukocytes, suggesting a potentially fundamental role in cellular adaptation to acute stress. Although Aqp1 can be found in various tissues and cell types, most notably red blood cells, renal tubules and endothelial cells [
7,
21‐
24], details on Aqp1 expression patterns and functional roles in leukocytes have not yet been described.
Aquaporins regulate cell membrane permeability in most cell types. Most isoforms are selective to water, although some, including aquaporin-7 and aquaporin-9, are permeable to other micromolecules, including urea and glycerol [
25]. Leukocytes are known to express aquaporin-9, which is crucial to the formation of filopodia, finger-like membrane protrusions involved in cytokinesis [
26,
27]. We here report the inducible expression of Aqp1 in leukocytes in clinical sepsis and in response to LPS
in vitro. This finding reveals a hitherto unknown and potentially fundamental function of this molecule in innate immunity.
To validate the results of the initial pilot gene expression study and meta-analysis, we prospectively followed Aqp1 expression in leukocytes of critically ill patients who were not septic upon ICU admission. We observed an increase in Aqp1 mRNA expression with the appearance of sepsis and septic shock. This effect was not dependent on the length of ICU stay, but was related to the septic process severity and strongly implies that Aqp1 is inducible in humans in the context of the immune response to infection.
Induction in Aqp1 was recently reported by Madonna and coworkers [
28] in mouse hearts following septic endotoxemia. In contrast, Aqp1 was found downregulated in human pleural mesothelial cell lines (MET-5A) and primary rat lung microvessel endothelial cells following LPS administration [
29,
30], as well as in rat and murine models of LPS-induced lung injury [
30,
31]. In spite of these disparities, which may merely reflect tissue-specific effects of the molecule and differences in experimental design, the data support a novel role for this protein in the stress response to sepsis.
Since ICU-acquired sepsis is most commonly caused by LPS found in Gram-negative bacteria, we tested the role of LPS as an Aqp1 inducer in human PMNs. We found a statistically and biologically significant upregulation of Aqp1 by LPS. This was abrogated by pharmacological disruption of NF-κB signaling pathway, known to transduce signals downstream of the Toll-like-receptor 4, implying that LPS stimulates Aqp1 expression via NF-κB activation. LPS measurably stimulated PMNs, as indicated by a significant increase of IL-8 production and activation of NF-κB.
Placing naive cells in a hypoosmotic environment results in a time-dependent and HgCl2-sensitive increase in cell volume. This transient effect is followed by a gradual decline to the baseline values, possibly as a result of a compensatory escape of water and osmotically active molecules from the cell or derecruitment of Aqp1 from the cell membrane. In the presence of LPS, however, membrane permeability remained increased throughout the observation period (600 seconds), consistent with excess Aqp1.
The implications of LPS-induced cell membrane permeability on PMN function in sepsis are unknown. Besides permeability, however, Aqp1 could regulate the PMN stress response by other putative mechanisms. In this context, Aqp1 contributes to endothelial cell migration [
32] by forming a critical scaffold for a plasma membrane-associated multiprotein complex important for cytoskeleton remodeling, adhesion and motility. Additionally, the formation of filopodia depends on assembly of an actin cytoskeletal scaffold to support the cell membrane protrusion. An early event in protrusion formation is a local increase in water permeability facilitated by aquaporins, which are thus critical to cell motility [
33]. The theoretical significance of the herein demonstrated induction of Aqp1 may be to enhance the neutrophil’s ability to migrate to sites of infection. However, additional involvement in neutrophil activation (for example, phagocytosis and oxidative burst) and resistance to apoptosis [
34] cannot be excluded.
Our study has the following limitations. The population of our initial study on gene expression profiling (that is, five polytrauma patients) was small due to limitation of funds. However, analysis identified genes, among them Aqp1, with statistically significant changes in their expression. Expression of Aqp1 in patients was probed in total leukocytes isolated from blood, whereas in vitro studies were performed on the PMN fraction, which contains mostly neutrophils. Nevertheless, it should be noted that leukocytes from septic patients are mainly neutrophils. Aside from this being one of the first reports of Aqp1 expression in human leukocytes and PMNs, our results document that sepsis and LPS are associated with Aqp1 induction in leukocytes and neutrophils and this could be an important event in neutrophil activation in the setting of sepsis, as well as a biomarker of the latter. Further studies on the regulation of the Aqp1 signal transduction pathway and functional significance may enhance our understanding of the host’s immune response, which appears crucial to the outcome of this devastating syndrome.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
AGV designed and conducted experiments and drafted the manuscript. NAM designed the study and drafted the manuscript. SEO designed and supervised the study and edited the manuscript. ZM collected and processed specimens and recorded patient data. EJ collected and processed specimens and recorded patient data. TP conducted the microarray experiments. AA edited the manuscript and provided valuable input to study design. CR edited the manuscript and provided valuable input to study design. VA designed the microarray study and edited the manuscript. AK designed the study, collected specimens, supervised patient enrollment, data collection and recording, and edited the manuscript. All authors read and approved the final manuscript.