Wageningen Academic P u b l i s h e r s
Beneficial Microbes, 2016; 7(2): 265-273
http://www.wageningenacademic.com/doi/pdf/10.3920/BM2015.0129 - Thursday, October 05, 2017 9:12:55 AM - Göteborgs Universitet IP Address:130.241.16.16
Glucolytic fingerprinting reveals metabolic groups within the genus Bifidobacterium: an exploratory study D. Rios-Covián1, B. Sánchez1, I. Cuesta2, S. Cueto-Díaz3, A.M. Hernández-Barranco2, M. Gueimonde1 and C.G. De los Reyes-Gavilán1* 1Probiotics
and Prebiotics Group, Department of Microbiology and Biochemistry of Dairy Products, Instituto de Productos Lácteos de Asturias, Consejo Superior de Investigaciones Científicas (IPLA-CSIC), Paseo Rio Linares s/n, 33300 Villaviciosa, Asturias, Spain; 2Scientific and Technical Facilities, IPLA-CSIC, Paseo Rio Linares s/n, 33300 Villaviciosa, Asturias, Spain; 3Scientific and Technical Facilities, University of Oviedo, Fernando Bonguera s/n, 33006 Oviedo, Spain; [email protected] Received: 4 September 2015 / Accepted: 30 November 2015 © 2015 Wageningen Academic Publishers
RESEARCH ARTICLE Abstract Microorganisms of the genus Bifidobacterium are inhabitants of diverse niches including the digestive tract of humans and animals. The species Bifidobacterium adolescentis, Bifidobacterium animalis, Bifidobacterium bifidum, Bifidobacterium breve and Bifidobacterium longum have qualified presumption of safety status granted by EFSA and several strains are considered probiotic, and are being included in functional dairy fermented products. In the present work we carried out a preliminary exploration of general metabolic characteristics and organic acid production profiles of a reduced number of strains selected from these and other species of the genus Bifidobacterium. The use of resting cells allowed obtaining metabolic fingerprints without interference of metabolites accumulated during growth in culture media. Acetic acid was the most abundant organic acid formed per mol of glucose consumed (from 1.07±0.03 to 1.71±0.22 mol) followed by lactic acid (from 0.34±0.06 to 0.90±0.12 mol), with moderate differences in production among strains; pyruvic, succinic and formic acids were also produced at considerably lower proportions, with variability among strains. The acetic to lactic acid ratio showed lower values in stationary phase as regard to the exponential phase for most, but not all, the microorganisms; this was due to a decrease in acetic acid molar proportions together with increases of lactic acid proportions in stationary phase. A linear discriminant analysis allowed to cluster strains into species with 51-100% probability, evidencing different metabolic profiles, according to the relative production of organic acids from glucose by resting cells, of microorganisms collected at the exponential phase of growth. Looking for a single metabolic marker that could adequately discriminate metabolic groups, we found that groups established by the acetic to lactic acid ratio fit well with differences previously evidenced by the discriminant analysis. The proper establishment of metabolic groups within the genus Bifidobacterium could help to select the best suited probiotic strains for specific applications. Keywords: Bifidobacterium, resting cells, acetic acid, lactic acid, metabolic group
1. Introduction Microorganisms of the genus Bifidobacterium are inhabitants of diverse niches, including the gastrointestinal tract and oral cavity of humans, animal gut, insect hindgut, sewage and food products (Lugli et al., 2014). Bifidobacteria are among the first colonisers of the human gut; they are dominant in the intestine of breast-fed infants and are also part of the adult human colon microbiota (Gueimonde et
al., 2010). In the colon, these microorganisms can grow at the expense of the fermentation of undigested complex carbohydrates from diet. The species Bifidobacterium adolescentis, Bifidobacterium animalis, Bifidobacterium bifidum, Bifidobacterium breve and Bifidobacterium longum have the qualified presumption of safety (QPS) status granted by the European Food Safety Authority (BIOHAZ, 2015); particular strains from these species are considered probiotics and are being included in functional
ISSN 1876-2883 print, ISSN 1876-2891 online, DOI 10.3920/BM2015.0129265
D. Rios-Covián et al.
http://www.wageningenacademic.com/doi/pdf/10.3920/BM2015.0129 - Thursday, October 05, 2017 9:12:55 AM - Göteborgs Universitet IP Address:130.241.16.16
dairy fermented products (Huys et al., 2013) where simple sugars constitute the main carbon source. Probiotics can play an important role in host defence mechanisms and maturation of the immune system (Floch et al., 2011; Verhasselt, 2010). Bifidobacteria belong to the Actinobacteria phylum. The genus currently encompasses 48 taxa, representing 39 species and 9 subspecies (Ventura et al., 2007). The saccharolytic ability of bifidobacteria varies depending on the species; they can degrade complex carbohydrates or oligosaccharides, with the concomitant production of acetate and lactate through the so-called bifid shunt (De los Reyes-Gavilán et al., 2005). These organic acids are the basis for cross-feeding interaction of these bacteria with other members of the colonic microbiota, such as the butyrate-producers (De Vuyst et al., 2013) which could contribute to the beneficial properties attributed to bifidobacteria (Belenguer et al., 2006; Flint et al., 2015). Therefore, it could be assumed that differential production patterns of organic acids by different bifidobacteria strains, may condition their interaction with other microbial inhabitants of the gut. Moreover, differences in metabolites produced by different Bifidobacterium strains grown on simple sugars during the manufacture of probiotic products may be of practical importance for the standardisation of organoleptic properties of such products, finally determining the beneficial action of the ingested
microorganisms. Nonetheless, the possibility of shaping different metabolic profiles in the genus Bifidobacterium remains scarcely studied. In the present work we have conducted a preliminary exploration of the general metabolic features and different profiles of organic acids produced by resting cells from strains belonging to different species of the genus Bifidobacterium, focusing specially on those microorganisms having QPS status and that are being used as probiotics.
2. Material and methods Bifidobacterium strains and culture conditions Bifidobacterial strains used in this study are listed in Table 1. Cells from frozen stocks were reactivated in MRS broth (BioKar Diagnostics, Beauvais, France) supplemented with 0.25% (w/v) L-cysteine (Sigma Chemical Co., St. Louis, MO) (MRSc). Bacteria were incubated in an anaerobic chamber (Mac 1000, Don Whitley Scientific, Shipley, UK) under a 10% H2, 10% CO2, and 80% N2 atmosphere. Growth curves were performed in MRSc inoculated from overnight cultures of the appropriate strains at a final concentration of 1% (v/v). Cultures were monitored spectrophotometrically at 600 nm (OD600). Growth rates (µ) were estimated from the growth curve by fitting the data to the equation Nt = N0 × e µ × t, in which Nt and N0 are the cell densities at time t and time zero (Pirt, 1975). Bacterial cell counts were
Table 1. Bifidobacterium strains used in this study and main parameters of growth in MRSc medium. Species
Strain
Origin1
Reference
OD600 2
Growth rate (h-1)
B. pseudocatenolatum
IPLA20026 IPLA20014 IPLA20004 IPLA20005 IPLA20006 IPLA20030 IPLA20031 Bb12 NB667 IPLA20001 NCIMB8809 IPLA 20015 IPLA 20017 IPLA 20016 LMG11039 DSM20099 CECT5781 LMG10735
infant faeces infant faeces breast milk breast milk breast milk dairy product bile-adapted dairy product infant faeces (1) breast milk infant faeces (2) infant faeces infant faeces infant faeces human faeces (3) pig faeces (4) adult intestine (5) intestine of Apis mellifera (3)
Solis et al., 2010 Solis et al., 2010 Solis et al., 2010 Ruas-Madiedo et al., 2010 Ruas-Madiedo et al., 2010 Solis et al., 2010 López et al., 2010 López et al., 2011 López et al., 2011 -
5.25±1.65 5.56±0.52 5.64±0.23 5.90±0.51 5.10±0.61 5.14±0.18 5.55±0.62 5.91±0.88 4.10±0.31 5,81±0,62 6.19±0.61 2.31±1.33 5.01±0.99 4.58±0.50 5.99±0.45 3.85±0.51 6.70±2.48 2.59±1.01
0.22±0.06 0.20±0.03 0.49±0.09 0.48±0.08 0.48±0.06 0.24±0.04 0.28±0.03 0.41±0.09 0.32±0.06 0.34±0.07 0.32±0.02 0.26±0.06 0.34±0.09 0.43±0.06 0.44±0.07 0.75±0.34 0.33±0.02 0.38±0.16
B. breve
B. animalis
B. longum
B. bifidum
B. angulatum B. pseudolongum B. adolescentis B. asteroides 1
Culture collections: (1) NIZO food research collection; (2) NCIMB culture collection; (3) LMG culture collection; (4) DSMZ culture collection; (5) CECT culture collection. 2 At the time of collection in stationary phase.
266
Beneficial Microbes 7(2)
Metabolic groups in the genus Bifidobacterium
obtained by spread plating decimal serial dilutions of liquid cultures on MRSc plates supplemented with 2% (w/v) agar. Microbial cultures were incubated at 37 °C.
http://www.wageningenacademic.com/doi/pdf/10.3920/BM2015.0129 - Thursday, October 05, 2017 9:12:55 AM - Göteborgs Universitet IP Address:130.241.16.16
Preparation and incubation of resting cell suspensions Studies were performed with buffered cell suspensions to analyse glucose consumption and organic acids formation from glucose, which greatly facilitate product analysis avoiding the interference of components from the culture medium as well as the accumulation of different microbial metabolites during growth. Ten ml of each strain culture were collected by centrifugation (10,000×g for 15 min) at an OD600 of 0.6-0.8, which corresponds with early exponential phase. Ten ml of cultures were also harvested at early stationary phase (OD600 of cultures at the time of collection are indicated in Table 1). Cultures were washed twice with 10 ml of 33 mM potassium phosphate buffer pH 7.0 and resuspended in 10 ml of 33 mM potassium phosphate buffer pH 5.6 containing 25 mM glucose. The suspensions were incubated anaerobically with constant mild stirring during 4 h at 37 °C. Cells were removed by centrifugation and cell-free supernatants were analysed by high performance liquid chromatography (HPLC) to quantify glucose and organic acid levels. Bacterial counts were obtained before and after stirring to confirm bacterial survival in resting cell preparations. Resting cell experiments for each strain and growth phase were carried out in triplicate.
Glucose consumption and organic acids production For the determination of glucose consumption, acetic, lactic, pyruvic, and formic acid production by resting cells, cell-free supernatants were filtered through 0.2 µm pore-size filters. HPLC analysis was carried out as described previously (Salazar et al., 2009). Briefly, an Alliance 2695 module (Waters, Milford, MA, USA) with an ICSep ICE-ION-300 column (Transgenomic, New Haven, CT, USA) was used to separate metabolites. Organic acids were quantified with a photodiode array detector (Waters 996) at 210 nm whereas glucose was quantified with a refractive index detector (Waters 2414) at 410 nm. A 1290 Infinity LC coupled with a 6460 triple quadrupole mass spectrometer (Agilent, Santa Clara, CA, USA), was used for the quantitation of succinic acid. Separation was performed using a Zorbax Eclipse Plus C18 column (2.1×50 mm, 1.8 µm) (Agilent), under isocratic conditions (98:2, 0.1% formic acid:acetonitrile) at a flow rate of 0.3 ml/min. The column temperature was 30 °C and the injection volume 2 µl. Detection was carried out using electrospray ionisation in the negative mode and a selected reaction monitoring method (transition 117.1→73.1) was optimised for maximum selectivity and sensitivity. A calibration curve in the range 20-2000 ng/ml (r2=0.998) was built for the quantification of succinic acid in the samples. Each sample was run in duplicate.
Beneficial Microbes 7(2)
Carbon recovery and molar proportion calculations Carbon recovery (CR) in buffered cell suspensions was calculated using the following formula: CR ≥ 100 × (4×[succinic acid] + 3×[lactic acid] + 3×[pyruvic acid] + 2×[acetic acid] + 1×[formic acid] / 6×[glucose consumed]) Results were expressed in percentages. Molar proportions were calculated as the percentage of each organic acid with respect to the total concentration of organic acids produced (succinic, lactic, pyruvic, acetic, and formic) by resting cells.
Statistical analyses One-way analyses of variance (ANOVA), Pearson’s correlations, and linear regressions were performed using the IBM SPSS software, version 22.0 (IBM, Armonk, NY, USA). Linear discriminant analysis (LDA) was carried out with the XLSTAT 2015 software (Addinsoft, SaintHubert, Canada). ANOVA were run using acetic to lactic acid ratio as factor and either strain or species as categories. Post hoc comparison was achieved by a least significant difference test (LSD; P
Beneficial Microbes, 2016; 7(2): 265-273
http://www.wageningenacademic.com/doi/pdf/10.3920/BM2015.0129 - Thursday, October 05, 2017 9:12:55 AM - Göteborgs Universitet IP Address:130.241.16.16
Glucolytic fingerprinting reveals metabolic groups within the genus Bifidobacterium: an exploratory study D. Rios-Covián1, B. Sánchez1, I. Cuesta2, S. Cueto-Díaz3, A.M. Hernández-Barranco2, M. Gueimonde1 and C.G. De los Reyes-Gavilán1* 1Probiotics
and Prebiotics Group, Department of Microbiology and Biochemistry of Dairy Products, Instituto de Productos Lácteos de Asturias, Consejo Superior de Investigaciones Científicas (IPLA-CSIC), Paseo Rio Linares s/n, 33300 Villaviciosa, Asturias, Spain; 2Scientific and Technical Facilities, IPLA-CSIC, Paseo Rio Linares s/n, 33300 Villaviciosa, Asturias, Spain; 3Scientific and Technical Facilities, University of Oviedo, Fernando Bonguera s/n, 33006 Oviedo, Spain; [email protected] Received: 4 September 2015 / Accepted: 30 November 2015 © 2015 Wageningen Academic Publishers
RESEARCH ARTICLE Abstract Microorganisms of the genus Bifidobacterium are inhabitants of diverse niches including the digestive tract of humans and animals. The species Bifidobacterium adolescentis, Bifidobacterium animalis, Bifidobacterium bifidum, Bifidobacterium breve and Bifidobacterium longum have qualified presumption of safety status granted by EFSA and several strains are considered probiotic, and are being included in functional dairy fermented products. In the present work we carried out a preliminary exploration of general metabolic characteristics and organic acid production profiles of a reduced number of strains selected from these and other species of the genus Bifidobacterium. The use of resting cells allowed obtaining metabolic fingerprints without interference of metabolites accumulated during growth in culture media. Acetic acid was the most abundant organic acid formed per mol of glucose consumed (from 1.07±0.03 to 1.71±0.22 mol) followed by lactic acid (from 0.34±0.06 to 0.90±0.12 mol), with moderate differences in production among strains; pyruvic, succinic and formic acids were also produced at considerably lower proportions, with variability among strains. The acetic to lactic acid ratio showed lower values in stationary phase as regard to the exponential phase for most, but not all, the microorganisms; this was due to a decrease in acetic acid molar proportions together with increases of lactic acid proportions in stationary phase. A linear discriminant analysis allowed to cluster strains into species with 51-100% probability, evidencing different metabolic profiles, according to the relative production of organic acids from glucose by resting cells, of microorganisms collected at the exponential phase of growth. Looking for a single metabolic marker that could adequately discriminate metabolic groups, we found that groups established by the acetic to lactic acid ratio fit well with differences previously evidenced by the discriminant analysis. The proper establishment of metabolic groups within the genus Bifidobacterium could help to select the best suited probiotic strains for specific applications. Keywords: Bifidobacterium, resting cells, acetic acid, lactic acid, metabolic group
1. Introduction Microorganisms of the genus Bifidobacterium are inhabitants of diverse niches, including the gastrointestinal tract and oral cavity of humans, animal gut, insect hindgut, sewage and food products (Lugli et al., 2014). Bifidobacteria are among the first colonisers of the human gut; they are dominant in the intestine of breast-fed infants and are also part of the adult human colon microbiota (Gueimonde et
al., 2010). In the colon, these microorganisms can grow at the expense of the fermentation of undigested complex carbohydrates from diet. The species Bifidobacterium adolescentis, Bifidobacterium animalis, Bifidobacterium bifidum, Bifidobacterium breve and Bifidobacterium longum have the qualified presumption of safety (QPS) status granted by the European Food Safety Authority (BIOHAZ, 2015); particular strains from these species are considered probiotics and are being included in functional
ISSN 1876-2883 print, ISSN 1876-2891 online, DOI 10.3920/BM2015.0129265
D. Rios-Covián et al.
http://www.wageningenacademic.com/doi/pdf/10.3920/BM2015.0129 - Thursday, October 05, 2017 9:12:55 AM - Göteborgs Universitet IP Address:130.241.16.16
dairy fermented products (Huys et al., 2013) where simple sugars constitute the main carbon source. Probiotics can play an important role in host defence mechanisms and maturation of the immune system (Floch et al., 2011; Verhasselt, 2010). Bifidobacteria belong to the Actinobacteria phylum. The genus currently encompasses 48 taxa, representing 39 species and 9 subspecies (Ventura et al., 2007). The saccharolytic ability of bifidobacteria varies depending on the species; they can degrade complex carbohydrates or oligosaccharides, with the concomitant production of acetate and lactate through the so-called bifid shunt (De los Reyes-Gavilán et al., 2005). These organic acids are the basis for cross-feeding interaction of these bacteria with other members of the colonic microbiota, such as the butyrate-producers (De Vuyst et al., 2013) which could contribute to the beneficial properties attributed to bifidobacteria (Belenguer et al., 2006; Flint et al., 2015). Therefore, it could be assumed that differential production patterns of organic acids by different bifidobacteria strains, may condition their interaction with other microbial inhabitants of the gut. Moreover, differences in metabolites produced by different Bifidobacterium strains grown on simple sugars during the manufacture of probiotic products may be of practical importance for the standardisation of organoleptic properties of such products, finally determining the beneficial action of the ingested
microorganisms. Nonetheless, the possibility of shaping different metabolic profiles in the genus Bifidobacterium remains scarcely studied. In the present work we have conducted a preliminary exploration of the general metabolic features and different profiles of organic acids produced by resting cells from strains belonging to different species of the genus Bifidobacterium, focusing specially on those microorganisms having QPS status and that are being used as probiotics.
2. Material and methods Bifidobacterium strains and culture conditions Bifidobacterial strains used in this study are listed in Table 1. Cells from frozen stocks were reactivated in MRS broth (BioKar Diagnostics, Beauvais, France) supplemented with 0.25% (w/v) L-cysteine (Sigma Chemical Co., St. Louis, MO) (MRSc). Bacteria were incubated in an anaerobic chamber (Mac 1000, Don Whitley Scientific, Shipley, UK) under a 10% H2, 10% CO2, and 80% N2 atmosphere. Growth curves were performed in MRSc inoculated from overnight cultures of the appropriate strains at a final concentration of 1% (v/v). Cultures were monitored spectrophotometrically at 600 nm (OD600). Growth rates (µ) were estimated from the growth curve by fitting the data to the equation Nt = N0 × e µ × t, in which Nt and N0 are the cell densities at time t and time zero (Pirt, 1975). Bacterial cell counts were
Table 1. Bifidobacterium strains used in this study and main parameters of growth in MRSc medium. Species
Strain
Origin1
Reference
OD600 2
Growth rate (h-1)
B. pseudocatenolatum
IPLA20026 IPLA20014 IPLA20004 IPLA20005 IPLA20006 IPLA20030 IPLA20031 Bb12 NB667 IPLA20001 NCIMB8809 IPLA 20015 IPLA 20017 IPLA 20016 LMG11039 DSM20099 CECT5781 LMG10735
infant faeces infant faeces breast milk breast milk breast milk dairy product bile-adapted dairy product infant faeces (1) breast milk infant faeces (2) infant faeces infant faeces infant faeces human faeces (3) pig faeces (4) adult intestine (5) intestine of Apis mellifera (3)
Solis et al., 2010 Solis et al., 2010 Solis et al., 2010 Ruas-Madiedo et al., 2010 Ruas-Madiedo et al., 2010 Solis et al., 2010 López et al., 2010 López et al., 2011 López et al., 2011 -
5.25±1.65 5.56±0.52 5.64±0.23 5.90±0.51 5.10±0.61 5.14±0.18 5.55±0.62 5.91±0.88 4.10±0.31 5,81±0,62 6.19±0.61 2.31±1.33 5.01±0.99 4.58±0.50 5.99±0.45 3.85±0.51 6.70±2.48 2.59±1.01
0.22±0.06 0.20±0.03 0.49±0.09 0.48±0.08 0.48±0.06 0.24±0.04 0.28±0.03 0.41±0.09 0.32±0.06 0.34±0.07 0.32±0.02 0.26±0.06 0.34±0.09 0.43±0.06 0.44±0.07 0.75±0.34 0.33±0.02 0.38±0.16
B. breve
B. animalis
B. longum
B. bifidum
B. angulatum B. pseudolongum B. adolescentis B. asteroides 1
Culture collections: (1) NIZO food research collection; (2) NCIMB culture collection; (3) LMG culture collection; (4) DSMZ culture collection; (5) CECT culture collection. 2 At the time of collection in stationary phase.
266
Beneficial Microbes 7(2)
Metabolic groups in the genus Bifidobacterium
obtained by spread plating decimal serial dilutions of liquid cultures on MRSc plates supplemented with 2% (w/v) agar. Microbial cultures were incubated at 37 °C.
http://www.wageningenacademic.com/doi/pdf/10.3920/BM2015.0129 - Thursday, October 05, 2017 9:12:55 AM - Göteborgs Universitet IP Address:130.241.16.16
Preparation and incubation of resting cell suspensions Studies were performed with buffered cell suspensions to analyse glucose consumption and organic acids formation from glucose, which greatly facilitate product analysis avoiding the interference of components from the culture medium as well as the accumulation of different microbial metabolites during growth. Ten ml of each strain culture were collected by centrifugation (10,000×g for 15 min) at an OD600 of 0.6-0.8, which corresponds with early exponential phase. Ten ml of cultures were also harvested at early stationary phase (OD600 of cultures at the time of collection are indicated in Table 1). Cultures were washed twice with 10 ml of 33 mM potassium phosphate buffer pH 7.0 and resuspended in 10 ml of 33 mM potassium phosphate buffer pH 5.6 containing 25 mM glucose. The suspensions were incubated anaerobically with constant mild stirring during 4 h at 37 °C. Cells were removed by centrifugation and cell-free supernatants were analysed by high performance liquid chromatography (HPLC) to quantify glucose and organic acid levels. Bacterial counts were obtained before and after stirring to confirm bacterial survival in resting cell preparations. Resting cell experiments for each strain and growth phase were carried out in triplicate.
Glucose consumption and organic acids production For the determination of glucose consumption, acetic, lactic, pyruvic, and formic acid production by resting cells, cell-free supernatants were filtered through 0.2 µm pore-size filters. HPLC analysis was carried out as described previously (Salazar et al., 2009). Briefly, an Alliance 2695 module (Waters, Milford, MA, USA) with an ICSep ICE-ION-300 column (Transgenomic, New Haven, CT, USA) was used to separate metabolites. Organic acids were quantified with a photodiode array detector (Waters 996) at 210 nm whereas glucose was quantified with a refractive index detector (Waters 2414) at 410 nm. A 1290 Infinity LC coupled with a 6460 triple quadrupole mass spectrometer (Agilent, Santa Clara, CA, USA), was used for the quantitation of succinic acid. Separation was performed using a Zorbax Eclipse Plus C18 column (2.1×50 mm, 1.8 µm) (Agilent), under isocratic conditions (98:2, 0.1% formic acid:acetonitrile) at a flow rate of 0.3 ml/min. The column temperature was 30 °C and the injection volume 2 µl. Detection was carried out using electrospray ionisation in the negative mode and a selected reaction monitoring method (transition 117.1→73.1) was optimised for maximum selectivity and sensitivity. A calibration curve in the range 20-2000 ng/ml (r2=0.998) was built for the quantification of succinic acid in the samples. Each sample was run in duplicate.
Beneficial Microbes 7(2)
Carbon recovery and molar proportion calculations Carbon recovery (CR) in buffered cell suspensions was calculated using the following formula: CR ≥ 100 × (4×[succinic acid] + 3×[lactic acid] + 3×[pyruvic acid] + 2×[acetic acid] + 1×[formic acid] / 6×[glucose consumed]) Results were expressed in percentages. Molar proportions were calculated as the percentage of each organic acid with respect to the total concentration of organic acids produced (succinic, lactic, pyruvic, acetic, and formic) by resting cells.
Statistical analyses One-way analyses of variance (ANOVA), Pearson’s correlations, and linear regressions were performed using the IBM SPSS software, version 22.0 (IBM, Armonk, NY, USA). Linear discriminant analysis (LDA) was carried out with the XLSTAT 2015 software (Addinsoft, SaintHubert, Canada). ANOVA were run using acetic to lactic acid ratio as factor and either strain or species as categories. Post hoc comparison was achieved by a least significant difference test (LSD; P
