EXPERIMENTAL IMMUNOLOGY doi: 10.1111/sji.12161 ..................................................................................................................................................................
Influence of the Gut Microbiota on Blood Acute-Phase Proteins W. Schroedl*, B. Kleessen*, L. Jaekel†, A. A. Shehata* & M. Krueger*
Abstract *Institute of Bacteriology and Mycology, Veterinary Faculty, University of Leipzig, Leipzig, Germany; and †Dairy farm, Schwabhausen, Germany
Received 19 November 2013; Accepted in revised form 27 January 2014 Correspondence to: W. Schroedl and A. A Shehata, Veterinary Faculty, Institute of Bacteriology and Mycology, University of Leipzig, Germany. E-mails: [email protected]; shehata@ vetmed.uni.leipzig.de
Little is known about the bovine intestinal microbiota influence on systemic innate immune responses. The objective of the present study was to determine relationships between acute-phase proteins in blood serum of cows [C-reactive protein (CRP), LPS-binding protein (LBP) and haptoglobin (Hp)] and the faecal microbiota. Fifty-two healthy cows (2–8 years old) were investigated. Faecal bacteria were determent characterized by in situ hybridization with 16S/23S rRNA-targeted probes and by conventional culture methods. The population of Gram-negative faecal bacteria (Enterobacteriaceae) was correlated negatively with CRP and positively with LBP in blood plasma, independent of the method used. Similar results were observed with Clostridium perfringens. No correlation was found between the faecal population of intestinal bacteria and Hp levels in blood plasma. This datum indicates that intestinal bacteria, especially Enterobacteriaceae and C. perfringens, may influence the level of CRP and LBP in blood plasma. These findings can be very important for diagnostic evaluations of the intestinal microbiota and provide specific information about its regulation.
Introduction The commensal microbiota is known to influence the body’s defence and to improve immune responses and resistance to infection [1]. It is generally accepted that a close interaction exists between the host, the microbial environment and the intestinal microflora from the beginning of life [2]. The immune system is of essential regulatory importance in these interactions. Specific relationships between the different bacterial species and the gut-associated lymphoid tissue are necessary for the development and maintenance of an individual host’s immune responses, immune defence, immune tolerance and homoeostasis. It is well established that a variety of microbial species in the gut is critical for the optimal function of the immune system. Inflammation of the gut mucosa by pathogens such as Salmonella Typhimurium creates competition with the intestinal microbiota [3]. Besides specific cellular and humoral factors, non-specific humoral defence factors are activated [4–6] and include acute-phase proteins (APP), which may have greater importance in the interaction with the microbial flora than generally assumed: immunoregulatory and antibacterial effects are described for recognition proteins such as Hp, CRP and LBP [7–9]. In recent years, there has been an increased interest in the pathogenetic importance of CRP and LBP, which are
Ó 2014 John Wiley & Sons Ltd
associated with cardiovascular and atherosclerotic diseases [10–12]. Limited information of correlations between major APP and intestinal microbiota indicates that these may be important for functioning of the innate immune system. The purpose of the present study was to investigate this relationship. Healthy dairy cows that originated from one stock with identical genetics and different ages but without any symptoms of clinical illnesses were chosen for these investigations. It was not possible to standardize all common stress factors such as different stages of lactation, number of births and length of milk production. Our hypothesis about a direct interaction between gastrointestinal microbiota and different APP is based on information obtained from human and animal studies using probiotics to improve health [13–15]. More basic knowledge of CRP, LBP and Hp is needed to understand the complex biological system involving intestinal microbiota and related host response. Quantification of bacteria by culture methods is limited in complex microbial ecological studies, as some bacteria are not detectable by culture methods [16]. To overcome this limitation, we used agar plates to cultivate aerobic bacteria and anaerobic bacteria, Lactobacillus and Clostridium perfringens in combination with culture-independent fluorescent in situ hybridization with a set of different oligonucleotide probes targeting 16S rRNA. This molecular technique permits a more comprehensive investigation of the intestinal microbiota [17].
299
300 Gut Microbiota Influence Blood Acute-Phase Proteins W. Schroedl et al. ..................................................................................................................................................................
Materials and methods Animals and samples. Blood and faecal samples were collected from 52 (2- to 8-year-old) healthy dairy cows (German Black and White breed 9 Holstein-Friesian breed, sbHF) at one farm. Blood samples were taken from the jugular vein and transferred to heparinized blood tubes. Plasma was separated by centrifugation at 3000 g for 15 min at room temperature. The faecal samples were taken with pressure on the ampulla recti. Cultural enumeration of selected faecal bacteria. Following homogenization, a series of 10-fold dilutions (102–108) of the faecal specimens was made in Dulbecco’s phosphatebuffered saline solution (PBS, pH 7.35), and duplicate 10 ll samples of each dilution were dropped onto nonselective nutrient agar for enumeration of total aerobes (SIFIN, Berlin, Germany), to selective MacConkey agar (SIFIN) for determining aerobic Gram-negative bacteria (BCGN) and to MRS agar (DeMan; Rogosa and Sharpe; Oxoid, Wesel, Germany) for counting Lactobacillus. C. perfringens was enumerated by plating on nutrient agar plates supplemented with 2% glucose, 200 lg/ml neomycin, 100 lg/ml polymyxin B and 5% sheep blood. Plates were incubated aerobically or in anaerobic jars (MERCK Diagnostics, Darmstadt, Germany) at 37 °C for 1–2 days as appropriate. Following incubation, the colonies were counted, and representative bacteria were purified further for identification using standard methods. The bacterial counts are expressed as log10 of colony-forming units (CFU)/gram of wet faeces. Fluorescent in situ hybridization and oligonucleotide probes. For whole-cell hybridization, a 1-ml aliquot of each faecal sample was fixed in ice-cold ethanol (1:1, v/v), methanol (1:1, v/v) and fresh paraformaldehyde (1:3, v/v) as described previously [18,19] and hybridized on silanized microscope slides [20] with fluorophore (indocarbocyanine Cy3)labelled 16S/23S rRNA-targeted oligonucleotide probes consisting of (1) an equimolar mixture of five Bacteriadirected probes EUB 338, EUB 785, EUB 927, EUB 1055 and EUB 1088 [19, 21], referred to as EUB mix, to detect all bacteria, (2) ARCH 915 for the Domain Archea [22], (3) HGC69a to detect Gram-positive bacteria with high GC content of DNA [19], (4) a set of two probes for species of the Bacteroides fragilis group (Bfra602) and the species Bacteroides distasonis (Bdis656) [23], (5) Erec 482 specific for most of the clostridia and eubacteria belonging to the Clostridium coccoides–Eubacterium rectale group [23], (6) the bifidobacterial probe Bif 164 [24], (7) Lab 158 for nearly all species of the genera Lactobacillus and Enterococcus in the gut [25], (8) Chis150 to detect bacteria of the Clostridium histolyticum group, for example Clostridium perfringens, C. butyricum and C. paraputrificum, (9) Clit135 to detect clostridia belonging to the Clostridium lituseburense group [23], (10) GAM 42a specific for the gamma subdivision of Proteobacteria [26] and (11) Ec 1531 specific for a number of
Enterobacteriaceae (e. g. Escherichia coli and Klebsiella pneumonia) [27]. If more stringent conditions were needed, formamide was added to the hybridization buffer from 5% to 40% (v/v). Cells were counted visually using a Zeiss Axioscope 50 fluorscense microscope (Carl Zeiss, Jena, Germany). Blood analyses. The parameters Hp, CRP and LBP were determined in blood plasma using ELISA. Common ELISA steps: All ELISAs were carried out with ELISA plates (microlon, 96-well, high binding; Greiner, Frickenhausen, Germany) filled with 100 ll per well and incubated for 1 h at room temperature (RT) on a microtiter plate shaker at 400 rpm. The coating buffer was 0.1 M NaHCO3, and the wash buffer (PBST) was PBS with 0.1% (v/v) Tween-20 (Sigma-Aldrich, Taufkirchen, Germany). The plates were then washed three times with wash buffer using a 12 channel-microtiter-plate-washer (NUNC, Wiesbaden, Germany). Horseradish peroxidase activity was determined on the solid phase with H2O2 and 3, 30 , 5, 50 tetramethylbenzidine substrate. The substrate reaction was stopped with 50 ll/well 1 M H2SO4 before reading the optical density with a microplate ELISA reader at 450 nm. ELISA for haptoglobin: The Hp was determined as described previously [15]. The coating antibody was IgG fraction (rabbit) anti-human Hpt (DAKO, Hamburg, Germany) diluted 1:3000. The bovine standard serum ranged from 3 to 200 ng/ml bovine Hpt. The samples were diluted 1:1000 and 1:100,000. The detection antibody was the same polyclonal IgG fraction used for coating, but conjugated with horseradish peroxidase (ELISA grade; Roche, Mannheim, Germany) as described by [28]. The detection antibody was diluted 1:10,000 in assay buffer. ELISA for CRP: C-reactive protein was determined as described previously [15]. The coating antibody was IgG fraction (rabbit) anti-human CRP (DAKO) diluted 1:2000. The bovine standard serum ranged from 1 to 100 ng/ml bovine CRP. The samples were diluted 1:1000 and higher. The detection antibody was IgG fraction (rabbit) antihuman CRP conjugated with horseradish peroxidase (DAKO) diluted 1:2000. ELISA for LBP: The LBP-coating antibody was affinitypurified monoclonal IgG2a (mouse) anti-LBP-human (mAbAbi-202) at 1.2 lg/ml. The standard LBP range in the ELISA was 0.3–20 ng/ml human LBP (LBP standard serum). The samples were diluted 1:1000 and higher. The assay buffer for dilution of the standard and plasma samples was 50 mM Tris/HCl (pH 8.0), 0.15 M NaCl, 10 mM EDTA and 0.1% Tween-20 (v/v). The detection antibody was affinity-purified monoclonal IgG1 (mouse) anti-LBP-human conjugated with horseradish peroxidase (mAb-Abi-204) diluted 1:6000 in assay buffer. The two mAbs and the standard serum were provided by Prof. Ch. Schuett, Institute of Immunology, University of Greifswald, Germany [30]. Data analysis. Bacteriological counts were transformed to log10 numbers before statistical analysis. Protein
Scandinavian Journal of Immunology, 2014, 79, 299–304
W. Schroedl et al. Gut Microbiota Influence Blood Acute-Phase Proteins 301 ..................................................................................................................................................................
concentrations were calculated with Table Curve software (Systat Science Software GmbH, Erkrath, Germany), and SigmaStat (Systat Science Software GmbH) was used to perform the analyses. The figures were made with SigmaPlot software (Systat Science Software GmbH), showing box plots with the median (solid line within the box), 25th and 75th percentiles (top and bottom solid lines of the box) and 10th and 90th percentiles (small lines outside the top and bottom of the box). The level of significance was calculated with the Mann–Whitney rank sum test, and the correlation coefficient was determined with the Spearman rank order correlation.
Results Microbiological and immunological results from dairy faecal and blood samples are presented in Table 1. Total counts of BCGN obtained by culture methods were similar to those obtained by fluorescent in situ hybridization (FISH) with oligonucleotide probes Gam 42a and Ec 1531 (correlation coefficient higher than 0.75). Cells of Clostridium histolyticum and/or C. lituseburence groups were detected by FISH in 25 of 52 faecal samples, while C. perfringens was found in 11 of the 52 samples identified by cultural methods. This study identified correlations between faecal bacteria and immunological parameters in blood plasma (Table 2). The levels of CRP Table 1 Faecal bacterial counts and concentration of non-specific immune defence markers in blood plasma. 90th percentile
Parameter
Median
10th percentile
Hpt (mg/ml) CRP (lg/ml) LBP (lg/ml) TBCa BCGNa BCLBa ARCH 915b Bacteriab HGC 69ab LGC 254b BIF 164b LAB 158b EREC 482b ATO 291b BACb Fprau 645b Gam 42ab Ec 1531b Clostridium perfringensa Chis 150/Clit 125b
1.182 73.310 13.226 7.540 6.810 6.600 9.040 9.410 8.590 6.895 8.560 6.860 8.835 7.750 8.855 8.975 7.830 7.535 n: 41/52 (
Influence of the Gut Microbiota on Blood Acute-Phase Proteins W. Schroedl*, B. Kleessen*, L. Jaekel†, A. A. Shehata* & M. Krueger*
Abstract *Institute of Bacteriology and Mycology, Veterinary Faculty, University of Leipzig, Leipzig, Germany; and †Dairy farm, Schwabhausen, Germany
Received 19 November 2013; Accepted in revised form 27 January 2014 Correspondence to: W. Schroedl and A. A Shehata, Veterinary Faculty, Institute of Bacteriology and Mycology, University of Leipzig, Germany. E-mails: [email protected]; shehata@ vetmed.uni.leipzig.de
Little is known about the bovine intestinal microbiota influence on systemic innate immune responses. The objective of the present study was to determine relationships between acute-phase proteins in blood serum of cows [C-reactive protein (CRP), LPS-binding protein (LBP) and haptoglobin (Hp)] and the faecal microbiota. Fifty-two healthy cows (2–8 years old) were investigated. Faecal bacteria were determent characterized by in situ hybridization with 16S/23S rRNA-targeted probes and by conventional culture methods. The population of Gram-negative faecal bacteria (Enterobacteriaceae) was correlated negatively with CRP and positively with LBP in blood plasma, independent of the method used. Similar results were observed with Clostridium perfringens. No correlation was found between the faecal population of intestinal bacteria and Hp levels in blood plasma. This datum indicates that intestinal bacteria, especially Enterobacteriaceae and C. perfringens, may influence the level of CRP and LBP in blood plasma. These findings can be very important for diagnostic evaluations of the intestinal microbiota and provide specific information about its regulation.
Introduction The commensal microbiota is known to influence the body’s defence and to improve immune responses and resistance to infection [1]. It is generally accepted that a close interaction exists between the host, the microbial environment and the intestinal microflora from the beginning of life [2]. The immune system is of essential regulatory importance in these interactions. Specific relationships between the different bacterial species and the gut-associated lymphoid tissue are necessary for the development and maintenance of an individual host’s immune responses, immune defence, immune tolerance and homoeostasis. It is well established that a variety of microbial species in the gut is critical for the optimal function of the immune system. Inflammation of the gut mucosa by pathogens such as Salmonella Typhimurium creates competition with the intestinal microbiota [3]. Besides specific cellular and humoral factors, non-specific humoral defence factors are activated [4–6] and include acute-phase proteins (APP), which may have greater importance in the interaction with the microbial flora than generally assumed: immunoregulatory and antibacterial effects are described for recognition proteins such as Hp, CRP and LBP [7–9]. In recent years, there has been an increased interest in the pathogenetic importance of CRP and LBP, which are
Ó 2014 John Wiley & Sons Ltd
associated with cardiovascular and atherosclerotic diseases [10–12]. Limited information of correlations between major APP and intestinal microbiota indicates that these may be important for functioning of the innate immune system. The purpose of the present study was to investigate this relationship. Healthy dairy cows that originated from one stock with identical genetics and different ages but without any symptoms of clinical illnesses were chosen for these investigations. It was not possible to standardize all common stress factors such as different stages of lactation, number of births and length of milk production. Our hypothesis about a direct interaction between gastrointestinal microbiota and different APP is based on information obtained from human and animal studies using probiotics to improve health [13–15]. More basic knowledge of CRP, LBP and Hp is needed to understand the complex biological system involving intestinal microbiota and related host response. Quantification of bacteria by culture methods is limited in complex microbial ecological studies, as some bacteria are not detectable by culture methods [16]. To overcome this limitation, we used agar plates to cultivate aerobic bacteria and anaerobic bacteria, Lactobacillus and Clostridium perfringens in combination with culture-independent fluorescent in situ hybridization with a set of different oligonucleotide probes targeting 16S rRNA. This molecular technique permits a more comprehensive investigation of the intestinal microbiota [17].
299
300 Gut Microbiota Influence Blood Acute-Phase Proteins W. Schroedl et al. ..................................................................................................................................................................
Materials and methods Animals and samples. Blood and faecal samples were collected from 52 (2- to 8-year-old) healthy dairy cows (German Black and White breed 9 Holstein-Friesian breed, sbHF) at one farm. Blood samples were taken from the jugular vein and transferred to heparinized blood tubes. Plasma was separated by centrifugation at 3000 g for 15 min at room temperature. The faecal samples were taken with pressure on the ampulla recti. Cultural enumeration of selected faecal bacteria. Following homogenization, a series of 10-fold dilutions (102–108) of the faecal specimens was made in Dulbecco’s phosphatebuffered saline solution (PBS, pH 7.35), and duplicate 10 ll samples of each dilution were dropped onto nonselective nutrient agar for enumeration of total aerobes (SIFIN, Berlin, Germany), to selective MacConkey agar (SIFIN) for determining aerobic Gram-negative bacteria (BCGN) and to MRS agar (DeMan; Rogosa and Sharpe; Oxoid, Wesel, Germany) for counting Lactobacillus. C. perfringens was enumerated by plating on nutrient agar plates supplemented with 2% glucose, 200 lg/ml neomycin, 100 lg/ml polymyxin B and 5% sheep blood. Plates were incubated aerobically or in anaerobic jars (MERCK Diagnostics, Darmstadt, Germany) at 37 °C for 1–2 days as appropriate. Following incubation, the colonies were counted, and representative bacteria were purified further for identification using standard methods. The bacterial counts are expressed as log10 of colony-forming units (CFU)/gram of wet faeces. Fluorescent in situ hybridization and oligonucleotide probes. For whole-cell hybridization, a 1-ml aliquot of each faecal sample was fixed in ice-cold ethanol (1:1, v/v), methanol (1:1, v/v) and fresh paraformaldehyde (1:3, v/v) as described previously [18,19] and hybridized on silanized microscope slides [20] with fluorophore (indocarbocyanine Cy3)labelled 16S/23S rRNA-targeted oligonucleotide probes consisting of (1) an equimolar mixture of five Bacteriadirected probes EUB 338, EUB 785, EUB 927, EUB 1055 and EUB 1088 [19, 21], referred to as EUB mix, to detect all bacteria, (2) ARCH 915 for the Domain Archea [22], (3) HGC69a to detect Gram-positive bacteria with high GC content of DNA [19], (4) a set of two probes for species of the Bacteroides fragilis group (Bfra602) and the species Bacteroides distasonis (Bdis656) [23], (5) Erec 482 specific for most of the clostridia and eubacteria belonging to the Clostridium coccoides–Eubacterium rectale group [23], (6) the bifidobacterial probe Bif 164 [24], (7) Lab 158 for nearly all species of the genera Lactobacillus and Enterococcus in the gut [25], (8) Chis150 to detect bacteria of the Clostridium histolyticum group, for example Clostridium perfringens, C. butyricum and C. paraputrificum, (9) Clit135 to detect clostridia belonging to the Clostridium lituseburense group [23], (10) GAM 42a specific for the gamma subdivision of Proteobacteria [26] and (11) Ec 1531 specific for a number of
Enterobacteriaceae (e. g. Escherichia coli and Klebsiella pneumonia) [27]. If more stringent conditions were needed, formamide was added to the hybridization buffer from 5% to 40% (v/v). Cells were counted visually using a Zeiss Axioscope 50 fluorscense microscope (Carl Zeiss, Jena, Germany). Blood analyses. The parameters Hp, CRP and LBP were determined in blood plasma using ELISA. Common ELISA steps: All ELISAs were carried out with ELISA plates (microlon, 96-well, high binding; Greiner, Frickenhausen, Germany) filled with 100 ll per well and incubated for 1 h at room temperature (RT) on a microtiter plate shaker at 400 rpm. The coating buffer was 0.1 M NaHCO3, and the wash buffer (PBST) was PBS with 0.1% (v/v) Tween-20 (Sigma-Aldrich, Taufkirchen, Germany). The plates were then washed three times with wash buffer using a 12 channel-microtiter-plate-washer (NUNC, Wiesbaden, Germany). Horseradish peroxidase activity was determined on the solid phase with H2O2 and 3, 30 , 5, 50 tetramethylbenzidine substrate. The substrate reaction was stopped with 50 ll/well 1 M H2SO4 before reading the optical density with a microplate ELISA reader at 450 nm. ELISA for haptoglobin: The Hp was determined as described previously [15]. The coating antibody was IgG fraction (rabbit) anti-human Hpt (DAKO, Hamburg, Germany) diluted 1:3000. The bovine standard serum ranged from 3 to 200 ng/ml bovine Hpt. The samples were diluted 1:1000 and 1:100,000. The detection antibody was the same polyclonal IgG fraction used for coating, but conjugated with horseradish peroxidase (ELISA grade; Roche, Mannheim, Germany) as described by [28]. The detection antibody was diluted 1:10,000 in assay buffer. ELISA for CRP: C-reactive protein was determined as described previously [15]. The coating antibody was IgG fraction (rabbit) anti-human CRP (DAKO) diluted 1:2000. The bovine standard serum ranged from 1 to 100 ng/ml bovine CRP. The samples were diluted 1:1000 and higher. The detection antibody was IgG fraction (rabbit) antihuman CRP conjugated with horseradish peroxidase (DAKO) diluted 1:2000. ELISA for LBP: The LBP-coating antibody was affinitypurified monoclonal IgG2a (mouse) anti-LBP-human (mAbAbi-202) at 1.2 lg/ml. The standard LBP range in the ELISA was 0.3–20 ng/ml human LBP (LBP standard serum). The samples were diluted 1:1000 and higher. The assay buffer for dilution of the standard and plasma samples was 50 mM Tris/HCl (pH 8.0), 0.15 M NaCl, 10 mM EDTA and 0.1% Tween-20 (v/v). The detection antibody was affinity-purified monoclonal IgG1 (mouse) anti-LBP-human conjugated with horseradish peroxidase (mAb-Abi-204) diluted 1:6000 in assay buffer. The two mAbs and the standard serum were provided by Prof. Ch. Schuett, Institute of Immunology, University of Greifswald, Germany [30]. Data analysis. Bacteriological counts were transformed to log10 numbers before statistical analysis. Protein
Scandinavian Journal of Immunology, 2014, 79, 299–304
W. Schroedl et al. Gut Microbiota Influence Blood Acute-Phase Proteins 301 ..................................................................................................................................................................
concentrations were calculated with Table Curve software (Systat Science Software GmbH, Erkrath, Germany), and SigmaStat (Systat Science Software GmbH) was used to perform the analyses. The figures were made with SigmaPlot software (Systat Science Software GmbH), showing box plots with the median (solid line within the box), 25th and 75th percentiles (top and bottom solid lines of the box) and 10th and 90th percentiles (small lines outside the top and bottom of the box). The level of significance was calculated with the Mann–Whitney rank sum test, and the correlation coefficient was determined with the Spearman rank order correlation.
Results Microbiological and immunological results from dairy faecal and blood samples are presented in Table 1. Total counts of BCGN obtained by culture methods were similar to those obtained by fluorescent in situ hybridization (FISH) with oligonucleotide probes Gam 42a and Ec 1531 (correlation coefficient higher than 0.75). Cells of Clostridium histolyticum and/or C. lituseburence groups were detected by FISH in 25 of 52 faecal samples, while C. perfringens was found in 11 of the 52 samples identified by cultural methods. This study identified correlations between faecal bacteria and immunological parameters in blood plasma (Table 2). The levels of CRP Table 1 Faecal bacterial counts and concentration of non-specific immune defence markers in blood plasma. 90th percentile
Parameter
Median
10th percentile
Hpt (mg/ml) CRP (lg/ml) LBP (lg/ml) TBCa BCGNa BCLBa ARCH 915b Bacteriab HGC 69ab LGC 254b BIF 164b LAB 158b EREC 482b ATO 291b BACb Fprau 645b Gam 42ab Ec 1531b Clostridium perfringensa Chis 150/Clit 125b
1.182 73.310 13.226 7.540 6.810 6.600 9.040 9.410 8.590 6.895 8.560 6.860 8.835 7.750 8.855 8.975 7.830 7.535 n: 41/52 (
