The Alterations of Mouse Plasma Proteins during Septic Development Yan Ren,†,‡,§ Jiandong Wang,§,| Ji Xia,†,‡ Chaoguang Jiang,| Kang Zhao,†,‡ Rong Li,| Ningzhi Xu,†,‡ Yingxin Xu,| and Siqi Liu*,†,‡,⊥ Beijing Genomics Institute, Chinese Academy of Science, Beijing, China, Beijing Proteomics Institute, Beijing, China, The General Hospital of Military, Beijing, China, Departement of Medicine, University of Louisville, Louisville, Kentucky Received January 26, 2007
A fundamental issue for sepsis therapy is to control the development of inflammation at an early stage. With cecal ligation and puncture (CLP) surgery, the mouse model has clearly shown the septic signs triggered by chronic insult. To monitor the plasma proteomic responses to sepsis, the mouse blood was collected at intervals after sham and CLP surgery followed by the sample treatment to remove high abundance serum albumin. The treated mouse plasma proteins were well resolved by twodimensional electrophoresis (2-DE). The image analysis revealed that these 2-DE spots observed from the sham and the CLP samples 4 h after surgery were comparable, whereas more than 30 different spots appeared on the 2-DE gels between the sham and CLP mouse plasma 24 h after surgery, indicating that some plasma proteins responded to the inflammatory development. These differential spots were verified by MALDI-TOF/TOF MS, resulting in 13 unique sepsis-responsive proteins. More importantly, most of them exhibited multiple spots as difference on the 2-DE gels. Furthermore, these isospots were incubated with PNGase F to eliminate N-linked oligosaccharides on proteins and then evaluated by Western blot as well as mass spectrometry. The results of PNGase F digestion suggested that most sepsis-associated proteins remained in N-glycosylation status but changed their N-glycans during septic development. Keywords: sepsis • cecal ligation and puncture (CLP) • plasma • glycoprotein • matrix-assisted laser desorption/ ionization-time-of-flight (MALDI-TOF) mass spectrometry • two-dimensional electrophoresis (2-DE)
Introduction Trauma, a life threatening disease, is especially prevalent among the younger population.1 Along with the rapid growth of global industrialization, the number of disabling injuries and trauma-related mortalities has also grown steadily every year. In most cases, multiple organ dysfunction syndrome (MODS) caused by trauma are directly precipitated by sepsis. Although it is well-known that sepsis is induced by many microbes that can invade human organisms, the underlying mechanism of this syndrome remains a mystery that is yet to be uncovered. Currently, there are only a few limited medical treatments available for septic shock. For instance, the recombinant of high-nobility group B1 protein (HMGB1), which can stimulate human monocytes to generate cytokines, has been adopted as a late mediator in LPS-induced toxicity, and another recombinant of human activated protein C (HAPC, Drotrecogin alfa), * To whom correspondence should be addressed. Siqi Liu, Ph.D., Beijing Genomics Institute, Chinese Academy of Sciences, Beijing Airport Industrial Zone B-6, Beijing 101300, China; E-mail,
[email protected]. † Chinese Academy of Science. ‡ Beijing Proteomics Institute. § These authors made equal contributions to this study. | The General Hospital of Military. ⊥ University of Louisville.
2812
Journal of Proteome Research 2007, 6, 2812-2821
Published on Web 05/12/2007
which provides anti-inflammatory and cytoprotective activities, has been proven to lower the risk of death from severe sepsis in adult patients at a high risk of death.2,3 The judgment to the septic stages is critical to improve the therapeutic effects of sepsis. The sequential organ failure assessment (SOFA) score is usually taken in assessing the incidence and severity of organ dysfunction in critically ill patients. Moreover, the SOFA score is established not only from physiological features but also from biochemical characteristics.4 Two fundamental questions, thus, have to be clearly addressed for developing the biomarkers of sepsis. Which biochemical events induce sepsis? What biochemical criteria can guide the classification of sepsis patients and monitor the septic development?5 Sepsis could also be referred to as a process of malignant intravascular inflammation.6 When the inflammation occurs, toxins from infectious microorganisms activate the cellular and humoral immune defense systems in the human body such as phagocytes (neutrophils and macrophages) and natural-killer (NK) lymphocytes. In the meantime, antibodies, cytokines, and inflammatory factors are present to effectively regulate these defense responses.7 Several signaling events are immediately evident during the initial responses to sepsis. This is especially true for cytokines. For instance, the pro-inflammatory regula10.1021/pr070047k CCC: $37.00
2007 American Chemical Society
Mouse Plasma Proteins during Septic Development
tors, such as TNF alpha, IL-1 beta, IL-6, and IL-8, are frequently elevated in human sepsis; in addition, anti-inflammatory regulators, such as IL-1 receptor, IL-10, IL-11, and IL-13, are also produced in large quantities in patients with sepsis.8 A complex interaction of cytokines and cytokine-neutralizing molecules probably determines the clinical presentation and course of sepsis. The concentrations of individual cytokines in body fluid, therefore, may not directly reflect the septic status. The use of an animal model is a well-established method for studying the molecular mechanisms of sepsis. As compared with other animal models of sepsis, such as lipopolysaccharide (LPS) and colon ascendens stent peritonitis (CASP), CLP is reasoned to more accurately reflect the status of septic humans, even though it is short of information relevant to the incidence of sepsis caused by gram-positive bacteria and fungi.9 Through CLP surgery, the mouse model clearly shows evidence of septic shock caused by chronic insult.10 It is worthy of noting that the cytokine responses to septic development were not consistent in the CLP models. Heuer et al. observed that in the CLP mouse plasma, IL-10 and IL-11 were down-regulated 12 h after surgery but up-regulated 22 h after surgery.11 Enoh et al. reported that all wild-type mice exhibited high plasma cytokine concentrations after CLP surgery and died by 42 h after CLP.12 The question here is which biomarker could serve as a distinctive and sensitive indicator to sepsis? According to the generally accepted criterion, the potential markers employed in clinical trials should at least possess three distinct characteristics: (1) close and steady correspondence with septic development; (2) wide and consistent existence in the circulation system; and (3) quantifiable through common biochemical approaches. To date, the global responses of the plasma proteins remain to be uncovered in the septic mice. This has prompted us to hypothesize that the plasma proteomic changes induced by CLP surgery could lead to the discovery of the potential biomarkers for sepsis. The proteomic strategy has been proven successful in profiling the differential protein expressions and it is widely employed in the area of plasma proteomics.13 In our current study, we evaluated our hypothesis by comparing differential plasma proteomes of the CLP and the sham mice. Because the amount of IgG in mouse plasma was not as high as that of human plasma, we only depleted most of the serum albumin from collected plasmas using affinity chromatography. 2-DE was used to separate the rest of the plasma proteins. The 2-DE spots corresponding with septic development, which show significant disparities, were further identified by mass spectrometry and Western blot. On the basis of the proteomic data, we demonstrated that during the early stage of sepsis (4 h after surgery), there were no significantly different plasma proteomes observed between the CLP and the sham group, but proteomic changes were definitely confirmed 24 h after CLP surgery. Impressively, most sepsis-responsive proteins in the CLP mice were up-regulated with multiple modified forms. The experiments showed that digesting the plasma proteins with PNGase F further attributed these modifications to various glycans on the plasma proteins. Ultimately, the proteomic information in the CLP plasma will enable us to get a better grasp of septic development at the molecular level.
Materials and Methods Chemical Reagents. DEAE-CIBACRON BLUE 3GA, PNGase F and all chemicals of analytical grade were purchased from Sigma (St. Louis, MO). Mouse Albumin ELISA Quantitation Kit
research articles was obtained from Bethyl Laboratories, Inc. (Montgomery, TX). IPG strips and all chemicals employed for electrophoresis were from Amersham Biosciences (Uppsala, Sweden). Modified trypsin (sequence grade) was obtained from Promega (Madison, WI). The primary and secondary antibodies against serum amyloid protein P (SAP), C reactive protein (CRP), and kininogen (KNG) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). The primary and secondary antibodies of R1-acid glycoprotein (R-AGP) were endowed as gifts from Genway Biotechnology Inc. (San Diego, CA). PVDF films were acquired from Millipore Corporation (Billerica, MA). CLP Model. Male Kunming mice (3-4 weeks of age, 1921 g/body weight) were subjected to the CLP procedure.14 The protocols of animal operation were approved by Committee of Animal Experimentation in The General Hospital of Military in Beijing. Briefly, under aseptic conditions, a 1 cm midline abdominal incision was made and the cecum was exposed. The cecum was ligated at approximately two-thirds of the distance from the distal tip, so intestinal continuity was maintained. The cecum was punctured twice with a 22-gauge needle, and a small amount of cecal contents were expressed through the punctures. A sham operation (laparotomy and cecal exposure without any more manipulation) was performed as control. The incision was closed and 1 mL of normal saline was administered subcutaneously. After the surgery, the animal had unrestricted access to food and water. The blood was drawn from the inferior vena cava and the mice were sacrificed at 4 or 24 h after the procedure. The blood samples were anticoagulated and centrifuged at 3000g for 5 m to get the plasma. The protease inhibitor cocktail Sigma (St. Louis, MO) was added into the plasma to prevent the degradation of plasma proteins. Removal of Albumin and Determination of Removal Efficiency. The blue dye agarose was balanced with 10 mM TrisHCl, pH 7.4. The mouse plasmas were diluted three times with 10 mM Tris-HCl, pH 7.4, and mixed with the balanced resins (3.5 mg protein:40 mg resin). The slurries were shaken for 40 min at room temperature followed by centrifuge at 2000g for 0.5 min through the spin column, and the collected filtrations were used for later proteomic analysis. The protein concentrations in the samples were quantitatively measured by the Bradford method. The concentrations of albumin in the plasmas treated with or without by the resin were evaluated by Mouse Albumin ELISA Quantitation Kit (Bethyl Laboratories, Inc. Montgomery, TX), in which the standard curve was established from the mouse reference serum with the stock concentration of albumin 45 mg/mL. The range of dilutions of reference albumin was from 7.8 to 500 ng/mL. Two-Dimensional Electrophoresis. Approximately 35 µg of protein was loaded to each strip (18 cm) with the rehydration buffer containing 8 M urea, 2% (w/v) CHAPS, 65 mM DTT, and 0.5% IPGphor buffer. Each sample was duplicated for 2-DE. The strips were rehydrated without voltage for 4 h and with 50 V for 8 h at 18 °C. The isoelectric focusing was programmed at a gradient model for 1 h focusing at each stage of voltage changes, 500, 1000, and 8000 V, respectively, and was continued at 8000 V for a total of 72 000 V h. The focused strips were treated by reduction of 1% (w/v) DTT in the equilibration buffer of 6 M Urea, 50 mM Tris-HCl, 30% Glycerol, 2% SDS, and trace Bromophenol blue and were subsequently alkylated by 2.5% (w/v) iodoacetamide in the same equilibration buffer. The treated strips were transferred onto 12% uniform SDS-polyacrylamide gels running in 2.5 W each gel for 30 min and 15 W each gel until the bromophenol blue dye reaching the gel Journal of Proteome Research • Vol. 6, No. 7, 2007 2813
research articles bottoms. The gels were stained by silver staining according to the protocol from Amersham only without glutardialdehyde. Image Acquisition and Data Analysis. The 2-DE gels were scanned by Imagescaner (Amersham Biosciences, Uppsala, Sweden) in a transmission mode, and the image analysis was conducted with ImageMaster 2D Platinum (Amersham Biosciences, Uppsala, Sweden). To get the comparable data for quantitative analysis, several key parameters in the image analysis were fixed as the constants. The relative volume of each spot was used as an index to eliminate the density differences caused by the individual experimental errors. The threshold defined as the significant change in spot volume was 3-fold at least upon the comparison of the average gels between the CLP and the sham samples. In-Gel Tryptic Digestion. The 2-DE spots verified as the significant changes in relative spot volume were separately excised by spot picker and transferred into the Eppendorf tubes. The particles were dehydrated with acetonitrile and then treated by reduction of 10 mM DTT and alkylation of 55 mM iodoacetamide followed by a thorough wash. Finally, the treated gel particles were incubated with 10 µL 25 mM NH4HCO3 containing 0.01 µg of trypsin at 37 °C overnight. After centrifugation, the resulted supernatants were mixed with 1 µL of 5% TFA and delivered to mass spectrometry. Protein Identification with Mass Spectrometry. The digestions were applied onto the AnchorChip target from Bruker (Bruker Daltonics, Bremen, Germany) followed by adding matrix solution consisted of R-cyano-4-hydroxycinnamic acid (4 mg/mL) in 70% acetonitrile with 0.1% TFA. The dried target was subjected into a Bruker UltraFlex MALDI-TOF/TOF MS. The mass spectrometer was operated under 19 kV accelerating voltage in the reflectron mode and the m/z range from 600 to 4000. Typically 100 shots were accumulated per spectrum in MS mode and 400 shots in MS/MS mode. The energy of laser was controlled about at 20% and the baseline of spectra less than 100 in intensity was accepted. All peptide mass fingerprintings (PMFs) were externally calibrated using standard peptide mixtures and internally calibrated using the masses of trypsin autolysis products (m/z ) 842.509 and m/z ) 2211.105) to reach a typical mass measurement accuracy of 100 ppm. The monoisotopic peptide masses obtained from MALDI-TOF MS were analyzed by Flexanalysis 2.4 and Biotools 2.4 software. Peptide mass searches were performed with MASCOT algorithm. The latest versions of the NCBI and Swiss-Prot protein databases were used. The following parameters were used for database searches: Monoisotopic mass accuracy
