2.1

Materials

Poly( d,l -lactic acid) (PDLLA), ( M _n = 31,750 and M _w = 100,000) with an inherent viscosity of 1.87 dL/g was purchased from Purasorb ^® (PURAC Biochem, The Netherlands) and was used as received. The 45S5 Bioglass ^® , with the composition: 45 SiO ₂ , 24.5 CaO, 24.5 Na ₂ O and 6.0 P ₂ O ₅ in wt%, was supplied by US Biomaterials Corp. (Florida, USA). The particle size of the Bioglass ^® particles (BG), measured by laser scattering analysis (Coulter LS 100 particle size analyzer, Coulter, USA), was found to be lower than 20 μm. All the other reagents and solvents used were of reagent grade and were used without further purification.

2.2

Preparation of PDLLA and PDLLA/Bioglass ^® membranes

All PDLLA membranes were prepared based on a solvent casting technique. The PDLLA films were prepared by dissolving 0.50 g of PDLLA in 30 mL of chloroform. After total dissolution, the solution was transferred to a Petri dish with 9 cm of diameter and covered with an aluminum sheet. The Petri dish was settled in a horizontal position to facilitate the formation of a cast film with uniform thickness. The assembly was kept in a hood for 24 h, and chloroform was allowed to evaporate at a very slow rate. Then, the films were vacuum dried for 48 h at 40 °C.

The PDLLA/BG membranes were prepared in the exact same process as the pure PDLLA membranes. The PDLLA/BG dispersions were prepared by dissolving 0.40 g of PDLLA in 30 mL of chloroform. After total dissolution, 0.10 g of Bioglass ^® was dispersed in the above solution. During solvent evaporation the particles were deposited by gravity to the bottom side, creating an asymmetric of 80/20 PDLLA/BG membrane along the thickness.

2.3

Bioactivity tests

For the in vitro bioactivity tests an acellular simulated body fluid (SBF) (1.0×) with ions concentration nearly equal to human blood plasma was prepared . Sample membranes of 20 × 15 mm ² were cut from the original processed films for the bioactivity tests. Three replicates for each sample were immersed in 45 mL SBF for 2, 5, 7, 14 and 21 days at 37 °C. After being removed from SBF the membranes were gently rinsed with distilled water and dried at room temperature.

2.4

Physico-chemical characterization

2.4.1

Scanning Electron Microscopy (SEM)

Qualitative information of the morphology of PDLLA and PDLLA/BG membranes surfaces, before and after the immersion in SBF, was obtained using a Scanning Electron Microscope, SEM (Nova NanoSEM 200-FEI Company, USA), at an accelerated voltage of 5 kV. Before being observed by SEM, the membranes were gold coated.

2.4.2

Fourier Transform Infrared Spectroscopy (FTIR)

The chemical structure of each side of the composite membrane were analyzed using FTIR. Spectra were recorded in an IR Prestige 21 FTIR spectrophotometer (Shimadzu, Japan) with the attenuated total reflection accessory (128 scans, resolution 4 cm ⁻¹ ) in the spectral range 2000–400 cm ⁻¹ .

2.5

Mechanical characterization

The tensile properties were determined using an INSTRON 4505 Universal Machine (Instron Int. Ltd., USA) equipped with a 1 kN load cell, with a loading rate of 5 mm min ^−1, up to 20% of strain, at room temperature. Samples were analyzed in dry and wet conditions. For the wet condition, samples were immersed for 3 h in PBS before being tested. The values reported represent an average of at least five testing specimens. Tensile stress was taken as the maximum stress in the stress-strain curve. Tensile modulus was estimated from the initial linear section of the stress-strain curve.

2.6

Cell culture studies using hBMSC and hPDL

Two types of cells were used in this study, namely human periodontal ligament cells (hPDL) and human bone marrow stromal cells (hBMSC). Both cells types were collected from two different donors.

hPDL were obtained from human third molar according to the following procedure. After extraction, the teeth were washed three times for 10 min in PBS with 100 units/mL penicillin and streptomycin. PDL tissue was scraped from the middle third of the root with a scalpel blade, to avoid contamination by epithelial or pulpal cells. The freed portions of the periodontal ligament were minced and transferred to a small culture flask, filled with 5 mL alpha minimal essential medium (α-MEM, Gibco) with 10% (v/v) fetal calf serum (FCS, Gibco), 50 mg/mL ascorbic acid (Sigma), 10-8 M dexamethasone (Sigma), 50 mg/mL gentamycin (Gibco) and 10 mM sodium β-glycerophosphate (Sigma). Cells were cultured at 37 °C in a humidified atmosphere of 5% CO ₂ and medium was replaced every 2–3 days. Upon reaching confluence, cells were released with trypsin/EDTA (0.25%, (w/v) crude trypsin and 1 mM EDTA, pH 7.2) and sub-cultured for two passages in standard culture flasks. The cells were then frozen in liquid nitrogen until used for the experiments.

Human BMSC were isolated from bone blocks of human iliac crest biopsies of donors. The biopsies were discarded tissues during standard surgical procedures at Radboud University Nijmegen Medical Center (Nijmegen, The Netherlands). The bone blocks were cut into small pieces and subsequently placed in a 50 mL tube to which 20 mL alpha-minimal essential medium (α-MEM) was added. After that the tube was shaken vigorously and the medium with cells was collected. This procedure was repeated several times. The collected medium with cells was plated in culture flasks (T175, Greiner Bio-one) and expanded in proliferation medium. Cells were characterized and showed stem cells phenotype. Additionally, a multipotential differentiation test was applied, demonstrating their stem cells capacity. Cells were cultured at 37 °C in a humid atmosphere with 5% CO ₂ and its passage was performed at 80% confluence using trypsin EDTA (Gibco).

Specific culture medium (O ⁻ ) was used for each cell type. The hPDL O ⁻ was composed of α-MEM (Gibco) with 10% fetal bovine serum (FBS, Greiner Bio-one) and 100 units mL ⁻¹ penicillin/streptomycin (Gibco). The hBMSC O ⁻ was composed of α-MEM (Gibco) with 15% FBS (Greiner Bio-one), 1% l -glutamine, 1% ascorbic acid (Sigma), 100 units mL ⁻¹ penicillin/streptomycin (Gibco) and 1%, by volume added to each cell culture flask, basic fibroblast growth factor (bFGF). After the first generation, cells were plated at a density of 5000 cells/cm ² in culture flasks (T175, Greiner Bio-one). The culture medium was changed twice a week. Cells from passage 3 (hBMSC) and 5 (hPDL) were used in the biological experiments.

All Gibco products are from Life Technologies BV, Breda, the Netherlands, all Greiner Bio-one products from Greiner Bio-one BV, Alphen aan de Rijn, the Netherlands, and all Sigma products from Sigma–Aldrich Chemie B.V., Zwijndrecht, the Netherlands.

2.6.1

Cell seeding

Metal rings (15 mm diameter × 3 mm thickness) were glued to the membrane with RTV silicone adhesive (Nusil Technology, Carpinteria, CA), to keep them in the bottom of the culture well, preventing fluctuation. Prior to cell seeding, the samples were sterilized with 70% (v/v) ethanol for 60 min and then washed three times immersed in PBS. The samples were placed in 25-well plates and soaked in cell culture medium overnight. After removing the culture medium, 50 μL of a cell suspension with a 2.0 × 10 ⁴ /sample cell density, was seeded onto the surface of each sample. After incubation for 4 h at 37 °C in a 5% CO ₂ atmosphere incubator, osteogenic medium (referred as O ⁺ ), specific for each cell type, was added to the seeded samples, according to the type of assay performed. The hPDL O ⁺ was composed of α-MEM (Gibco) with 10% FBS (Greiner Bio-one), 1% ascorbic acid (Sigma), 1% b-glycerophosphate (Sigma), 1% dexamethasone (Sigma) and 100 units mL ⁻¹ penicillin/streptomycin (Gibco). The hBMSC O ⁺ was composed of a-MEM (Gibco) with 15% FBS (Greiner Bio-one), 1% l -glutamine, 1% ascorbic acid (Sigma), 1% b-glycerophosphate (Sigma), 1% dexamethasone (Sigma) and 100 units mL ⁻¹ penicillin/streptomycin (Gibco). On the control groups, cells were seeded directly on the well-plates and osteogenic medium was added immediately.

2.7

Cell adhesion, proliferation and metabolic activity

2.7.1

DNA content

After the different experimental time points, medium was removed from the wells and the samples were washed twice with PBS. The analysis was performed on the supernatant of the substrates after day 1, 3, 7, 14 and 28 of culture. Cells were lysed using milliQ with subsequent sonification for 10 min between two cycles of freeze/thaw from −80 °C. The supernatant was stored at −20 °C until further analysis. A PicoGreen dsDNA Quantification Kit (Molecular Probes, Eugene, USA) was used according to manufacturer’s instructions. To each 100 μL sample, 100 μL PicoGreen working solution was added. The samples must incubate for 2–5 min at room temperature, in the dark. After incubation, the fluorescence was measured on a fluorescence cuvette reader (microplate fluorescence reader, Bio-Tek, Winooski, USA) with a 485 nm excitation filter and a 530 nm emission filter.

2.7.2

AlamarBlue® staining

Cell metabolic activity was measured using AlamarBlue ^® staining (Invitrogen) according to the instructions of the manufacturer. A solution was made with AlamarBlue and culture medium in a proportion 1:9 (v/v) and was placed at 37 °C for 5 min. The medium was removed from wells and replaced with the solution. Plates were incubated (37 °C and 5% CO ₂ ) for 4 h. After incubation, 200 μL of each sample solution was transferred to 96-well plates (Greiner Bio-one). Fluorescence was measured using a microplate reader (FL 600; Bio-Tek) at 570 nm. The assay was performed on day 1, 3, 7, 14 and 28 of culture.

2.7.3

Scanning Electron Microscopy (SEM) observation

Adhesion of both cell types (hBMSC and hPDL) on membranes was analyzed by SEM ( n = 2). After day 3 and day 28 time points, cells were fixed in 2% (v/v) glutaraldehyde in 0.1 M sodium-cacodylate buffered solution, for 5 min. Cells were rinsed in cacodylate buffered solution, dehydrated in a series of ethanol dilutions in water (70%, 80%, 90%, 96% and 100% (v/v)), 1 h in each, and dried in tetramethylsilane (TMS, Merck) to air. Finally, specimens were sputtercoated with a thin layer of gold, and examined in a JEOL 6310 scanning electron microscope.

2.8

Cell differentiation and mineralization

2.8.1

Alkaline Phosphatase activity measurements (ALP)

The same supernatants as used for PicoGreen assay were also used to measure alkaline phosphatase (ALP) activity (Sigma). To each 80 μL of sample, 20 μL of 0.5 M Alkaline Buffer (Sigma) was added. Thereafter 100 μL substrate solution 5 mM paranitrophenylphosphate (PNP, Sigma) was added to each well. After 60 min of incubation at 37 °C, 100 μL stop solution (0.3 M NaOH) was added to each well. Finally, ALP activity was measured at 405 nm using an ELISA microplate reader (Bio-Tek Instruments Inc, USA).

2.8.2

Von Kossa staining

Cells were fixed with 2% glutaraldehyde, stained with fresh 5% silver nitrate (AgNO ₃ ), washed with distilled water, developed with 5% sodium carbonate (Na ₂ CO ₃ ) in 25% formalin, and fixed with 5% sodium thiosulphate (Na ₂ S ₂ O ₃ ). Stained samples were observed under a Leica MZ12 stereomicroscope and images were captured.

2.8.3

Calcium content

Calcium content was assessed after 21 and 28 days of culture to obtain information about mineralized matrix formation. The samples were rinsed twice with milliQ. 1 mL of acetic acid was added to each sample. The samples were incubated overnight under vigorous constant shaking and the acetic acid with the diluted calcium was frozen and kept at −20 °C, until further investigation. After thawing, the calcium content was determined using the OCPC method. Optic density was read with an ELISA reader (Bio-Tek Instruments Inc, USA) at a wavelength of 570 nm. Bare membranes were also assessed in order to further exactly quantify and distinguish cellular from acellular mineralization on the membranes.

2.9

Statistical analysis

All samples were measured in triplicate. Biological tests were performed twice, excepting Von Kossa. All results are presented as mean ± standard deviation. Statistical analysis of experimental data was performed using an unpaired ordinary ANOVA with standard parametric methods. Calculations were performed in InStat (v. 3.0 GraphPad Software Inc, San Diego, CA). Statistical significance was set to p -value ≤ 0.1 (*), to p -value ≤ 0.01 (**) and to p -value ≤ 0.001 (***).

2.1

Materials

Poly( d,l -lactic acid) (PDLLA), ( M _n = 31,750 and M _w = 100,000) with an inherent viscosity of 1.87 dL/g was purchased from Purasorb ^® (PURAC Biochem, The Netherlands) and was used as received. The 45S5 Bioglass ^® , with the composition: 45 SiO ₂ , 24.5 CaO, 24.5 Na ₂ O and 6.0 P ₂ O ₅ in wt%, was supplied by US Biomaterials Corp. (Florida, USA). The particle size of the Bioglass ^® particles (BG), measured by laser scattering analysis (Coulter LS 100 particle size analyzer, Coulter, USA), was found to be lower than 20 μm. All the other reagents and solvents used were of reagent grade and were used without further purification.

2.2

Preparation of PDLLA and PDLLA/Bioglass ^® membranes

All PDLLA membranes were prepared based on a solvent casting technique. The PDLLA films were prepared by dissolving 0.50 g of PDLLA in 30 mL of chloroform. After total dissolution, the solution was transferred to a Petri dish with 9 cm of diameter and covered with an aluminum sheet. The Petri dish was settled in a horizontal position to facilitate the formation of a cast film with uniform thickness. The assembly was kept in a hood for 24 h, and chloroform was allowed to evaporate at a very slow rate. Then, the films were vacuum dried for 48 h at 40 °C.

The PDLLA/BG membranes were prepared in the exact same process as the pure PDLLA membranes. The PDLLA/BG dispersions were prepared by dissolving 0.40 g of PDLLA in 30 mL of chloroform. After total dissolution, 0.10 g of Bioglass ^® was dispersed in the above solution. During solvent evaporation the particles were deposited by gravity to the bottom side, creating an asymmetric of 80/20 PDLLA/BG membrane along the thickness.

2.3

Bioactivity tests

For the in vitro bioactivity tests an acellular simulated body fluid (SBF) (1.0×) with ions concentration nearly equal to human blood plasma was prepared . Sample membranes of 20 × 15 mm ² were cut from the original processed films for the bioactivity tests. Three replicates for each sample were immersed in 45 mL SBF for 2, 5, 7, 14 and 21 days at 37 °C. After being removed from SBF the membranes were gently rinsed with distilled water and dried at room temperature.

2.4

Physico-chemical characterization

2.4.1

Scanning Electron Microscopy (SEM)

Qualitative information of the morphology of PDLLA and PDLLA/BG membranes surfaces, before and after the immersion in SBF, was obtained using a Scanning Electron Microscope, SEM (Nova NanoSEM 200-FEI Company, USA), at an accelerated voltage of 5 kV. Before being observed by SEM, the membranes were gold coated.

2.4.2

Fourier Transform Infrared Spectroscopy (FTIR)

The chemical structure of each side of the composite membrane were analyzed using FTIR. Spectra were recorded in an IR Prestige 21 FTIR spectrophotometer (Shimadzu, Japan) with the attenuated total reflection accessory (128 scans, resolution 4 cm ⁻¹ ) in the spectral range 2000–400 cm ⁻¹ .

2.5

Mechanical characterization

The tensile properties were determined using an INSTRON 4505 Universal Machine (Instron Int. Ltd., USA) equipped with a 1 kN load cell, with a loading rate of 5 mm min ^−1, up to 20% of strain, at room temperature. Samples were analyzed in dry and wet conditions. For the wet condition, samples were immersed for 3 h in PBS before being tested. The values reported represent an average of at least five testing specimens. Tensile stress was taken as the maximum stress in the stress-strain curve. Tensile modulus was estimated from the initial linear section of the stress-strain curve.

2.6

Cell culture studies using hBMSC and hPDL

Two types of cells were used in this study, namely human periodontal ligament cells (hPDL) and human bone marrow stromal cells (hBMSC). Both cells types were collected from two different donors.

hPDL were obtained from human third molar according to the following procedure. After extraction, the teeth were washed three times for 10 min in PBS with 100 units/mL penicillin and streptomycin. PDL tissue was scraped from the middle third of the root with a scalpel blade, to avoid contamination by epithelial or pulpal cells. The freed portions of the periodontal ligament were minced and transferred to a small culture flask, filled with 5 mL alpha minimal essential medium (α-MEM, Gibco) with 10% (v/v) fetal calf serum (FCS, Gibco), 50 mg/mL ascorbic acid (Sigma), 10-8 M dexamethasone (Sigma), 50 mg/mL gentamycin (Gibco) and 10 mM sodium β-glycerophosphate (Sigma). Cells were cultured at 37 °C in a humidified atmosphere of 5% CO ₂ and medium was replaced every 2–3 days. Upon reaching confluence, cells were released with trypsin/EDTA (0.25%, (w/v) crude trypsin and 1 mM EDTA, pH 7.2) and sub-cultured for two passages in standard culture flasks. The cells were then frozen in liquid nitrogen until used for the experiments.

Human BMSC were isolated from bone blocks of human iliac crest biopsies of donors. The biopsies were discarded tissues during standard surgical procedures at Radboud University Nijmegen Medical Center (Nijmegen, The Netherlands). The bone blocks were cut into small pieces and subsequently placed in a 50 mL tube to which 20 mL alpha-minimal essential medium (α-MEM) was added. After that the tube was shaken vigorously and the medium with cells was collected. This procedure was repeated several times. The collected medium with cells was plated in culture flasks (T175, Greiner Bio-one) and expanded in proliferation medium. Cells were characterized and showed stem cells phenotype. Additionally, a multipotential differentiation test was applied, demonstrating their stem cells capacity. Cells were cultured at 37 °C in a humid atmosphere with 5% CO ₂ and its passage was performed at 80% confluence using trypsin EDTA (Gibco).

Specific culture medium (O ⁻ ) was used for each cell type. The hPDL O ⁻ was composed of α-MEM (Gibco) with 10% fetal bovine serum (FBS, Greiner Bio-one) and 100 units mL ⁻¹ penicillin/streptomycin (Gibco). The hBMSC O ⁻ was composed of α-MEM (Gibco) with 15% FBS (Greiner Bio-one), 1% l -glutamine, 1% ascorbic acid (Sigma), 100 units mL ⁻¹ penicillin/streptomycin (Gibco) and 1%, by volume added to each cell culture flask, basic fibroblast growth factor (bFGF). After the first generation, cells were plated at a density of 5000 cells/cm ² in culture flasks (T175, Greiner Bio-one). The culture medium was changed twice a week. Cells from passage 3 (hBMSC) and 5 (hPDL) were used in the biological experiments.

All Gibco products are from Life Technologies BV, Breda, the Netherlands, all Greiner Bio-one products from Greiner Bio-one BV, Alphen aan de Rijn, the Netherlands, and all Sigma products from Sigma–Aldrich Chemie B.V., Zwijndrecht, the Netherlands.

2.6.1

Cell seeding

Metal rings (15 mm diameter × 3 mm thickness) were glued to the membrane with RTV silicone adhesive (Nusil Technology, Carpinteria, CA), to keep them in the bottom of the culture well, preventing fluctuation. Prior to cell seeding, the samples were sterilized with 70% (v/v) ethanol for 60 min and then washed three times immersed in PBS. The samples were placed in 25-well plates and soaked in cell culture medium overnight. After removing the culture medium, 50 μL of a cell suspension with a 2.0 × 10 ⁴ /sample cell density, was seeded onto the surface of each sample. After incubation for 4 h at 37 °C in a 5% CO ₂ atmosphere incubator, osteogenic medium (referred as O ⁺ ), specific for each cell type, was added to the seeded samples, according to the type of assay performed. The hPDL O ⁺ was composed of α-MEM (Gibco) with 10% FBS (Greiner Bio-one), 1% ascorbic acid (Sigma), 1% b-glycerophosphate (Sigma), 1% dexamethasone (Sigma) and 100 units mL ⁻¹ penicillin/streptomycin (Gibco). The hBMSC O ⁺ was composed of a-MEM (Gibco) with 15% FBS (Greiner Bio-one), 1% l -glutamine, 1% ascorbic acid (Sigma), 1% b-glycerophosphate (Sigma), 1% dexamethasone (Sigma) and 100 units mL ⁻¹ penicillin/streptomycin (Gibco). On the control groups, cells were seeded directly on the well-plates and osteogenic medium was added immediately.

2.7

Cell adhesion, proliferation and metabolic activity

2.7.1

DNA content

After the different experimental time points, medium was removed from the wells and the samples were washed twice with PBS. The analysis was performed on the supernatant of the substrates after day 1, 3, 7, 14 and 28 of culture. Cells were lysed using milliQ with subsequent sonification for 10 min between two cycles of freeze/thaw from −80 °C. The supernatant was stored at −20 °C until further analysis. A PicoGreen dsDNA Quantification Kit (Molecular Probes, Eugene, USA) was used according to manufacturer’s instructions. To each 100 μL sample, 100 μL PicoGreen working solution was added. The samples must incubate for 2–5 min at room temperature, in the dark. After incubation, the fluorescence was measured on a fluorescence cuvette reader (microplate fluorescence reader, Bio-Tek, Winooski, USA) with a 485 nm excitation filter and a 530 nm emission filter.

2.7.2

AlamarBlue® staining

Cell metabolic activity was measured using AlamarBlue ^® staining (Invitrogen) according to the instructions of the manufacturer. A solution was made with AlamarBlue and culture medium in a proportion 1:9 (v/v) and was placed at 37 °C for 5 min. The medium was removed from wells and replaced with the solution. Plates were incubated (37 °C and 5% CO ₂ ) for 4 h. After incubation, 200 μL of each sample solution was transferred to 96-well plates (Greiner Bio-one). Fluorescence was measured using a microplate reader (FL 600; Bio-Tek) at 570 nm. The assay was performed on day 1, 3, 7, 14 and 28 of culture.

2.7.3

Scanning Electron Microscopy (SEM) observation

Adhesion of both cell types (hBMSC and hPDL) on membranes was analyzed by SEM ( n = 2). After day 3 and day 28 time points, cells were fixed in 2% (v/v) glutaraldehyde in 0.1 M sodium-cacodylate buffered solution, for 5 min. Cells were rinsed in cacodylate buffered solution, dehydrated in a series of ethanol dilutions in water (70%, 80%, 90%, 96% and 100% (v/v)), 1 h in each, and dried in tetramethylsilane (TMS, Merck) to air. Finally, specimens were sputtercoated with a thin layer of gold, and examined in a JEOL 6310 scanning electron microscope.

2.8

Cell differentiation and mineralization

2.8.1

Alkaline Phosphatase activity measurements (ALP)

The same supernatants as used for PicoGreen assay were also used to measure alkaline phosphatase (ALP) activity (Sigma). To each 80 μL of sample, 20 μL of 0.5 M Alkaline Buffer (Sigma) was added. Thereafter 100 μL substrate solution 5 mM paranitrophenylphosphate (PNP, Sigma) was added to each well. After 60 min of incubation at 37 °C, 100 μL stop solution (0.3 M NaOH) was added to each well. Finally, ALP activity was measured at 405 nm using an ELISA microplate reader (Bio-Tek Instruments Inc, USA).

2.8.2

Von Kossa staining

Cells were fixed with 2% glutaraldehyde, stained with fresh 5% silver nitrate (AgNO ₃ ), washed with distilled water, developed with 5% sodium carbonate (Na ₂ CO ₃ ) in 25% formalin, and fixed with 5% sodium thiosulphate (Na ₂ S ₂ O ₃ ). Stained samples were observed under a Leica MZ12 stereomicroscope and images were captured.

2.8.3

Calcium content

Calcium content was assessed after 21 and 28 days of culture to obtain information about mineralized matrix formation. The samples were rinsed twice with milliQ. 1 mL of acetic acid was added to each sample. The samples were incubated overnight under vigorous constant shaking and the acetic acid with the diluted calcium was frozen and kept at −20 °C, until further investigation. After thawing, the calcium content was determined using the OCPC method. Optic density was read with an ELISA reader (Bio-Tek Instruments Inc, USA) at a wavelength of 570 nm. Bare membranes were also assessed in order to further exactly quantify and distinguish cellular from acellular mineralization on the membranes.

2.9

Statistical analysis

All samples were measured in triplicate. Biological tests were performed twice, excepting Von Kossa. All results are presented as mean ± standard deviation. Statistical analysis of experimental data was performed using an unpaired ordinary ANOVA with standard parametric methods. Calculations were performed in InStat (v. 3.0 GraphPad Software Inc, San Diego, CA). Statistical significance was set to p -value ≤ 0.1 (*), to p -value ≤ 0.01 (**) and to p -value ≤ 0.001 (***).