Environ. Sci. Technol. 2010, 44, 2535–2541
Mitochondrial Toxicity of Microcystin-LR on Cultured Cells: Application to the Analysis of Contaminated Water Samples GRZEGORZ JASIONEK,† ALEXANDER ZHDANOV,† ´ HA,§ AND ˇ K BLA JOHN DAVENPORT,‡ LUDE D I M I T R I B . P A P K O V S K Y * ,† Biochemistry Department, University College Cork, College Road, Cork, Ireland, Department of Zoology, Ecology and Plant Science, Environmental Research Institute, University College Cork, College Road, Cork, Ireland, and Research Centre for Environmental Chemistry and Ecotoxicology RECETOX, Faculty of Science, Masaryk University, Kamenice 3, CZ62500 Brno, Czech Republic
Received October 21, 2009. Revised manuscript received February 9, 2010. Accepted February 17, 2010.
Microcystins (MC) are potent hepatic toxins delivered into the cells by organic anion transporting peptides (OATP) where they target protein phosphatases and mitochondria. We analyzed the effects of MC-LR on primary hepatocytes, HepG2, and Jurkat T cells, and isolated rat liver mitochondria by measuring changes in O2 consumption by optical oxygen sensing technique. Respiration of fresh primary hepatocytes was inhibited by MC-LR with EC50 ) 2.74 ( 0.65 nM, whereas an uncoupling effect on mitochondrial state 2 and state 3 respiration was observed with glutamate/malate as a substrate. HepG2 and Jurkat T cells lacking OATP showed no sensitivity to MCLR; however, facilitated delivery of MC-LR resulted in a marked enhancement of HepG2 O2 consumption and inhibition of Jurkat O2 consumption at g0.1 nM. The respiratory response did not coincide with changes in viability, total cellular ATP, extracellular acidification, ROS formation, or protein phosphorylation, which were detectable at higher MC-LR doses. Such prominent effect on cellular respiration was therefore used for the detection of MC-LR in environmental samples. A simple and sensitive screening assay for MC-LR toxicity was developed, which uses Jurkat cells, facilitated delivery of the toxin(s) and measurement on a fluorescent reader. The assay was applied to a panel of environmental samples suspected to contain MC and benchmarked against the ELISA test. It allowed identification of toxic samples and quantification of both nonspecific and MC-LR type of toxicity.
Introduction Microcystins (MC) are a group of cyclic heptapeptides produced by cyanobacterial species such as Microcystis aeuruginosa associated with poisoning of animals and * Corresponding author phone: +353 21 4901698; fax: +353 21 4274034; e-mail:
[email protected]. Corresponding author address: Biochemistry Department, University College Cork, Cavanagh Pharmacy Building, College Road, Cork, Ireland. † Biochemistry Department, University College Cork. ‡ Environmental Research Institute, University College Cork. § Masaryk University. 10.1021/es903157h
2010 American Chemical Society
Published on Web 03/01/2010
humans during cyanobacterial and algal blooms (1). Due to their widespread distribution, high toxicity, and threat to public health, MC levels have become an important parameter in water quality control, environmental monitoring, and toxicology. Although toxicity of individual MC varies considerably, they share a common structural feature: the unusual β-amino acid ADDA essential for their activity (2). The action of MCLR (one of the most toxic MC) results in disruption of the cytoskeleton causing apoptosis (3). MC-LR molecules are cell-impermeable; however, they are taken up by cells via organic anion transporting polypeptides (OATP) present mainly in liver and brain (4). Inside the cell MC-LR inhibits protein phosphatases 1 and 2A (PP1, PP2A) perturbing the phosphorylation/dephosphorylation balance (5) and activating a number of signaling pathways which lead to cell damage and apoptosis. MC-LR also induces DNA damage of susceptible cell lines and has tumor promoting potential upon chronic exposure to low doses in drinking water (6). Toxic effects of MC on mitochondrial function through elevation of reactive oxygen species (ROS) (7), depletion of glutathione (8), decrease in mitochondrial membrane potential, and triggering of mitochondrial permeability transition (9) have also been observed. Therefore, oxygen consumption may also be affected by MC-LR; however, this has not been studied in detail. A deeper understanding of toxic action of MC on cells and higher organisms and development of techniques for their detection in environmental samples are important for ecotoxicology. Analytical techniques such as LC-MS and ELISA provide sensitive and selective detection of MC (10) but are expensive and have limited value in assessing the biological hazard of samples, where higher animal, primary cells and animal tissue isolates, protein phosphatases inhibition assays are mainly used (11, 12). These techniques are complex, expensive, have ethical issues, limited throughput, often lack specificity, and cannot satisfy current demands. Many other cells lack OATP transporters and are not sensitive to MC-LR at environmentally relevant doses (13). Primary hepatocytes were suggested for alternative testing; however, during isolation and culturing they lose OATP and therefore their responsiveness to MC-LR (14). Engineered cells overexpressing the OATP (15, 16) are also complex and not widely applicable. O2 is a key metabolite of aerobic cells and O2 consumption rate is a sensitive marker of mitochondrial and cellular (dys)function (17). Optical O2 microrespirometry facilitates simple, rapid, high throughput analysis of O2 consumption by isolated mitochondria (18), various cell and animal models (19) using conventional microtiter plates, and fluorescence readers. In this study we present an alternative approach to the study of MC toxicity through the use of facilitated transport of the toxicant, which eliminates the need of cells expressing OATP transporters. By analyzing changes in O2 consumption and other bioenergetic parameters, we demonstrate toxic effects of low, environmentally relevant concentrations of MC-LR on the two common lines - HepG2 and Jurkat cells which are normally immune to MC-LR. We also describe a new method for screening for MC-LR type of toxicity in environmental samples using facilitated delivery of the toxicants to Jurkat cells, optical oxygen respirometry, and validate it with a panel of environmental samples.
Materials and Methods Materials. The MitoXpress and pH-Xtra probes and mineral oil were from Luxcel Biosciences. Endo-Porter (EP), EscortIII VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2535
(ESC), Lipofectamine 2000 (LP), and FuGene (FG) transfection reagents were from GeneTools, Sigma, Invitrogen, and Roche, respectively. Mouse antibodies against phospho-threonine, polyclonal rabbit antimouse-HRP, antimouse anti-Fc-IgG, and monoclonal anti-MC-LR antibodies were from Qiagen, Dako, Sigma, and ALEXIS, respectively. Protease inhibitors cocktail was from Roche, and PROTRAN nitrocelulose membrane was from Whatman. ROS indicator 3′-(p-aminophenyl) fluorescein (APF) was from Invitrogen. Microcystin-LR, Aflatoxin B1, Aroclor 1254, monoclonal anti-Rtubulin antibody, tetramethylbenzidine, Dulbecco’s Modified Eagle’s Medium (DMEM), low glucose DMEM, RPMI-1640, and fetal bovine serum (FBS) were from Sigma-Aldrich. Environmental Samples. Samples from drinking water reservoirs, lakes, and fish ponds (more than 300 samples from over 100 localities) were collected during the 2007 summer season within the National monitoring program on toxic cyanobacteria, Czech Republic (20). The samples were analyzed by competitive ELISA using a monoclonal antibody against MC-LR and a HRP-MC-LR tracer in triplicates, using 0.125-2 nM calibration with MC-LR standards. Samples with MC concentrations above the calibration range were diluted 20-fold and reanalyzed. Based on the results, 17 samples from 11 localities with variable microcystin content were selected for this study. Cell Culture and Treatment. HepG2 and Jurkat cells obtained from ATCC were cultured in 75 cm2 polystyrene flasks in DMEM or RPMI media supplemented with 10% FBS, penicillin/streptomycin, and also 2 mM glutamine in case of DMEM. Primary hepatocytes were isolated from Sprague-Dawley male rats by the two-step collagenase perfusion method (21), seeded on collagen-coated (6 µg/ mL) 96-well plates at 30,000 cells/well in low glucose (1 g/L) DMEM and left to adhere for 3 and 24 h. Jurkat cells were seeded on 24WP (Sarstedt) at 2 × 106 cells/ml in 1 mL of medium. HepG2 were seeded on collagen-coated 96WP at 50,000 cells/well in 0.2 mL of medium and returned to culture overnight. Toxicant stock (MC-LR in ethanol, Aroclor1254, and Aflatoxin B1 in DMSO) was diluted with appropriate growth medium containing 5% FBS to the desired final concentrations (MC-LR 0.01-50 nM, Aroclor 1254 0.1-20 mg/L, Aflatoxin B1 0.1-100 µM). For cell treatment, medium was replaced with fresh media (100 µL for 96WP and 1 mL for 24WP), and an equal amount of MC-LR diluted in medium was dispensed in assay wells. In the case of environmental samples, these were passed through a 0.22 µm sterile filter (to prevent microbial contamination), diluted 3:1 in 4x RPMI (1.33 dilution) and added to Jurkat. Facilitated transport of MC-LR was achieved through the addition either 6 µL/ml of EndoPorter (EP), 6 µL/ml EscortIII (ESC), 4 µL/ml FuGene (FG), or 4 µL/ml Lipofectamine 2000 (LP). OptiMem I medium was used for FG and LP. Positive control samples contained cells without toxicant and/or transport reagent. Negative controls contained no cells. After incubation for 6, 12, and 24 h at 37 °C, samples were measured. Isolation of Mitochondria. Rat liver mitochondria were isolated as previously described (22). Briefly, 3 g of liver was minced and washed in solution I (210 mM mannitol, 70 mM sucrose, 5 mM Hepes, 1 mM EGTA, and 0.5% fatty acid free BSA, pH 7.4) until the homogenate became blood free. Five volumes of solution I were added, and the tissue was homogenized on a smooth glass grinder with Teflon pestle. The homogenate was then adjusted to eight volumes with solution I and centrifuged at 700 g for 10 min at 4 °C. The supernatant was filtered through cheesecloth and centrifuged at 10,000 g for 10 min at 4 °C. Precipitated mitochondria were washed in 20 mL of solution I and spun down at 10,000 g for 10 min at 4 °C. The washing step was repeated in solution II (210 mM mannitol, 70 mM sucrose, 10 mM MgCl2, 5 mM K2HPO4, 10 mM 3-(N-morpholino)propanesulfonic acid 2536
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
(MOPS), and 1 mM EGTA, pH 7.4). Finally, mitochondria were resuspended in 0.7 mL of solution II, and protein concentration was determined using the BCA kit (Thermo Scientific). Respirometric Measurements. To measure mitochondrial O2 consumption, MitoXpress probe was reconstituted in 10 mL of respiration buffer (250 mM sucrose, 15 mM KCl, 1 mM EGTA, 5 mM MgCl2, and 30 mM K2HPO4, pH 7.4), and 100 µL of this solution was added to the wells of 96WP. Then 50 µL of mitochondria (diluted to desired concentration with respiration buffer) and 50 µL of substrate giving a final concentration of 25 mM succinate or 12.5/12.5 mM glutamate/malate for state 2 respiration measurements and 1.65 mM ADP for state 3 measurements were added. Then 100 µL of heavy mineral oil was added to each well to seal the samples from ambient oxygen, and the plate was read kinetically for 45 min at 30 °C on a fluorescence plate reader Genios Pro (Tecan), with readings in each well taken every 60 s. Measurement settings were as follows: excitation/ emission - 380/650 nm, gain - 90, delay time - 30 µs, gate time - 100 µs. For the analysis of O2 consumption by primary hepatocytes and HepG2 cells, a vial of MitoXpress probe was reconstituted in 1 mL of assay medium. Following exposure to MC-LR, 10 µL of this stock was added to test wells (150 µL final volume). The wells were then sealed with oil and measured at 37 °C for 90 min, as described above for isolated mitochondria. Respiration of Jurkat cells was measured on the LightCycler system (Roche) (23). The cells were spun down and resuspended in RPMI containing 5 µM of probe to give final concentration of 3 × 106 cells/ml, and 20 µL of cell suspension was added to LightCycler glass capillary cuvettes. These capillaries were then centrifuged for 5 s at 5000 rpm to bring the sample to the bottom and read on the LightCycler reader at 37 °C for 1 h using a 650 nm emission filter. From the measured profiles of probe fluorescence intensity signal, the rate of signal increase was determined for each sample (∆I/∆t) and corrected for negative control (probe and media only). The resulting slopes of treated samples were expressed as % of untreated samples to determine relative changes in respiration rate. Cell Function Assays. Cell viability/membrane integrity was assessed on the Guava 96-PCA flow cytometer using Guava Via Count kit (Guava Technologies) at ∼5 × 104 cells/ mL as per manufacturer’s protocol. Total cellular ATP levels were measured on the Victor2 (PerkinElmer) multilabel reader using CellTiter-Glo assay (GE Healthcare). The cells were lysed with CellTiterGlo reagent, shaken for 2 min, dispensed into wells of white 96WP, incubated in room temperature for 10 min, and then read under standard luminescence settings. The rate of extracellular acidification (ECA) was measured using pH-Xtra probe as described previously (24). Briefly, HepG2 cells were exposed to MC-LR as described above, but the last 3 h of exposure and ECA measurement were carried out in unbuffered DMEM (without Na-bicarbonate) in CO2free (atmospheric air) incubator at 37 °C. pH-Xtra probe was reconstituted in 1 mL of medium, and 0.01 mL of this stock was added to each well. Changes in probe signal reflecting changes in extracellular pH were then measured on a Victor2 plate reader at 30 °C, kinetically over 1-2 h. Measurement settings were as follows: 340/615 ex/em, delay times 100 and 300 µs, gate time 30 µs. Probe signal was converted to lifetime values: LT ) (t1-t2)/ln(F1/F2), where LT is fluorescence lifetime, t1 and t2 are delay times, and F1 and F2 are fluorescence intensity signals. From the resulting LT profiles ECA values were determined using known calibration function (24). ROS generation was assessed using APF indicator (25). After the exposure to MC-LR, serum and phenol red free
FIGURE 1. Relative changes in O2 consumption by the following: A) - primary rat hepatocytes 4 h [-9-] and 24 h [-0-] after seeding induced by 3 h treatment with MC-LR, HepG2 [-2-] and Jurkat [-O-] cells after 24 h exposure to 0.1-50 nM MC-LR, measured at 37 °C and B) isolated rat liver mitochondria for the different substrates and respiration states measured immediately after MC-LR addition at 30 °C. medium containing 5 µM APF was added, cells were incubated for 45 min and then measured on the Olympus FV1000 confocal laser scanning microscope under excitation at 488 nm (5.0% of laser power), and emission was collected at 500-550 nm. Both fluorescent and differential interference contrast (DIC) images were collected with a 60X oil immersion objective in eight planes with 0.5 µm steps. The resulting z-stacked images were analyzed using FV1000 Viewer software (Olympus) and Adobe Photoshop and Illustrator. Western blot analysis of protein phosphorylation was performed using standard method (see the SI). All experiments were performed several times with 5-6 repeats for each data point to ensure consistency of results. Results are presented as mean values with standard deviations where appropriate.
Results The Effects of MC-LR on Cellular and Mitochondrial Oxygen Consumption. To analyze alterations in metabolism induced by MC-LR we measured cellular O2 consumption (Figure 1). Freshly isolated rat hepatocytes were seeded and after 4 or 24 h of culturing exposed to different doses of MC-LR (0.150 nM) for 3 h and then measured. At 4 h after isolation a significant drop in O2 consumption was observed (p < 0.05 for 0.1 nM and p < 0.005 above that), producing IC50 ) 2.74 ( 0.65 nM MC-LR, whereas at 24 h after isolation a reduced susceptibility was seen (Figure 1A). This can be explained by a gradual decrease in OATP expression in primary hepatocytes after isolation (14). HepG2 and Jurkat cells lacking the OATP showed no changes in O2 consumption after the exposure to higher doses of MC-LR for 24 h. This supports the findings that uptake through the OATP is required for the cytotoxic effects, while the cells lacking OATP are normally immune to MC-LR (13). Since most of the cellular oxygen is consumed by the mitochondria, the direct effect of MC-LR on isolated rat liver mitochondria was tested by exposing them to MC-LR and analyzing in state 2 and state 3 (with ADP added) respiration in glutamate/malate (complex I substrate) and succinate (complex II substrate) buffers as described in ref 18. In glutamate/malate, a significant increase in O2 consumption was detectable at >3 nM MC-LR, reaching ∼130% of the untreated control at 12 nM. The increase in state 2 respiration was higher than in state 3 (Figure 1B). In succinate no significant changes in respiration were seen for both states. These results demonstrate that MC-LR has an uncoupling effect on the mitochondria and that its target is associated with ETC complex I functioning. Facilitated Transport of MC-LR into the Cells. To bypass the association of MC-LR toxicity with the presence of OATP,
we investigated the use of facilitated delivery into the cells. Several transfection reagents having different chemical nature and transport mechanism were examined for their ability to induce MC-LR toxicity in HepG2 and Jurkat cells which are normally resistant to MC-LR. For the initial assessment of different transport reagents, we measured the loading of the MitoXpress probe, which also comprises a polypeptide structure and which can be quantified with high sensitivity by time-resolved fluorometry. The cells were exposed to 1 µM of MitoXpress for 24 h with transfection reagent and without (to account for nonspecific binding), then washed, and measured on a Time Resolved Fluorescence (TR-F) reader. The resulting fluorescent signals were normalized for cell numbers and compared. Table 1 shows that EP and FG provided a higher loading of HepG2 and Jurkat cells, whereas Escort and LF were less efficient. Based on these results, the first two reagents were selected to study the effects of MC-LR on the respiration. The exposure of HepG2 and Jurkat cells to MC-LR in the presence of EP was seen to induce marked changes in O2 consumption (Figure 2). HepG2 showed a characteristic dosedependent increase in O2 consumption after the exposure which peaked at ∼0.5-1 nM of MC-LR. Significant increases (p < 0.005) were observed after 6 and 12 h exposure even at 0.1 nM MC-LR. The highest increase was after 12 h - 220% of untreated control. Higher MC-LR concentrations (>10 nM) inhibited the respiration producing a bell-shape doseresponse curve similar to ETC uncouplers (17). Jurkat cells also showed high sensitivity to MC-LR (Figure 2B) but with a strong decrease in O2 consumption. This inhibition was detectable at 0.5 nM MC-LR after 24 h exposure or g3 nM MC-LR after 6 h (p > 0.005). Shorter exposure times produced smaller responses, with a minor elevation after 3 h exposure. Exposure of HepG2 or Jurkat cells to EP or FG without MCLR had no effect on cell respiration. To examine the broader applicability of this facilitated delivery approach in toxicological studies, we analyzed the effects of several other toxicants on cell respiration under similar conditions (Table 1). Aflatoxin B1, a potent food poison and hepatocarcinogen (26), had a direct effect on Jurkat cells decreasing their O2 consumption after 24 h exposure (EC50 ) 41.55 ( 13.7 µM), in the presence of EP the respiration decreased at lower doses of Aflatoxin B1 (EC50 ) 8.18 ( 1.57 µM). Aroclor1254, a mixture of polychlorinated biphenyls known to affect the mitochondria (27), also inhibited O2 consumption in HepG2 cells (EC50 ) 5.21 ( 1.02 mg/L), and this effect was enhanced by EP (EC50 ) 3.24 ( 0.45 mg/L). Several other relevant parameters of cellular function were analyzed and related to the changes in O2 consumption by VOL. 44, NO. 7, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2537
TABLE 1. Assessment of Facilitated Delivery of the Phosphorescent Probe and Toxins into the Cells cell model (cell number) HepG2 (90,000/well)
Jurkat (100,000/well)
a
compound analyzed
parameter assessed
MitoXpress - probe
fluorescent signal
MC-LR - marine toxin
O2 consumption
Aroclor1254 - pesticide
O2 consumption
MitoXpress - probe
fluorescent signal
MC-LR - marine toxin
O2 consumption
Aflatoxin B1 - food toxin
O2 consumption
transport reagent (loading time)
observed effect
none (24 h) Endo-Porter (24 h) Fugene (24 h) Lipofectamine2000 (24 h) EscortIII (24 h) none (12/24 h) Endo-Porter (12 h) Fugene (12 h) Lipofectamine2000 (12 h) EscortIII (12 h) none (24 h) Endo-Porter (24 h)
2000 cpsa 10,000 cps 30,000 cps 7000 cps 6000 cps no toxicity peak - 220% peak - 245% no changes no changes EC50 ) 5.21 EC50 ) 3.24
none (24 h) Endo-Porter (24 h) Fugene (24 h) Lipofectamine2000 (24 h) EscortIII (24 h) none (24 h) Endo-Porter (24 h) Fugene (24 h) Lipofectamine2000 (24 h) EscortIII (24 h) none (24 h) Endo-Porter (24 h)
1000 cps 27,000 cps 25,000 cps 10,000 cps 12,000 cps no toxicity inhibition at >0.1 nM inhibition at >0.1 nM no changes no changes EC50 ) 41.55 µM ( 13.7 EC50 ) 8.18 µM ( 1.57
at 0.5 nM at 0.5 nM ( 1.02 mg/L ( 0.45 mg/L
TR-F intensity signal (counts per second, cps).
FIGURE 2. Changes in oxygen consumption for HepG2 (A) and Jurkat (B) cells induced by MC-LR in the presence of Endo-Porter at different exposure times. MC-LR treatment. Flow cytometry showed that plasma membrane integrity of Jurkat and HepG2 cells remained unaffected even after 24 h treatment 100 nM MC-LR in the presence of EP (see SI, Figure S1). Extracellular acidification (ECA), which reflects the rate of glycolytic flux, was measured using a pH-sensitive probe (24). Under the conditions which produce maximal respiratory responses to MC-LR treatment in HepG2 cells, only a small dose-dependent increase in ECA was observed at g30 nM MC-LR, i.e. when the respiration started to be inhibited (Figure 3A). Total cellular ATP was measured in conditions producing maximal respiratory effect: in primary hepatocytes after 3 h exposure without EP and in HepG2 and Jurkat cells after 12 and 24 h exposure with EP, respectively. Figure 3B shows a reduction in ATP induced by the MC-LR treatment in primary hepatocytes at 5 nM and in Jurkats at 0.1 nM (p > 0.05) but no measurable changes in HepG2 cells. These results suggest that at g30 nM MC-LR HepG2 cells elevate their glycolysis to maintain ATP level. Since elevated ROS production is associated with MC-LR induced cytotoxicity (7, 28), we measured ROS levels using APF probe and confocal fluorescent microscopy. In HepG2 2538
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 7, 2010
cells exposed to MC-LR/EP for 12 h, a dose dependent increase in ROS levels was observed, which became significant at g1 nM MC-LR (Figure 3C). Analysis of protein phosphorylation in HepG2 and Jurkat cells exposed to MC-LR/EP for 24 h was conducted by Western blotting using antibodies against phospho-threonine, as serine/threonine phosphatase is one of MC-LR target within the cell (5). At MC-LR concentrations 10 nM and higher, we observed an increase in protein phosphorylation in the 2050 kD range (see SI, Figure S2), which reflects the inhibition of phosphatase activity. Toxicological Assessment of Contaminated Water Samples. The above findings allowed us to develop a simple and sensitive screening test for the presence of MC in water samples, using standard cell lines, facilitated (nonspecific) transport of MC-LR by EP, and detection of changes in cellular O2 consumption by optical microrespirometry. HepG2 cells having a bell-shape response (Figure 2) are more prone to producing ambiguous results; therefore, Jurkat cells were chosen for this. As a proof of concept, the test was applied to a panel of 17 environmental water samples collected from
TABLE 2. Effects of Environmental Samples on O2 Consumption of Jurkat (Treated with or without EP for 24 h) and the Results of Microcystins Content Determination by O2 Respirometry and ELISAa sample no.
relative OCR, no EP (%)
relative OCR, with EP (%)
[MC], nM respirometry
[MC], nM ELISA
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
96.3 ( 4.9 98.3 ( 3.9 96.8 ( 5.3 99.5 ( 6.1 5.6 ( 5.2 93.1 ( 6.5 98.1 ( 3.6 93.4 ( 6.1 91.6 ( 3.7 65.2 ( 6.6 72 ( 5.3 68.8 ( 4.5 92.5 ( 6.3 96.5 ( 3.5 94.1 ( 5.2 83.6 ( 5.1 95.2 ( 4.8
63.4 ( 6.4 66.7 ( 4.4 72.3 ( 5.2 73.3 ( 5.6 0.1 ( 0.8 80.2 ( 5.7 75.6 ( 4.8 65.7 ( 5.3 59.4 ( 6.2 45.2 ( 4.7 40.8 ( 6.8 38.8 ( 7.1 85.6 ( 4.3 81.6 ( 5.7 88.5 ( 3.3 71.5 ( 4.5 90.1 ( 5.4
36.7 17.2 4.1 2.2 n.d. 1.0 2.1 26.6 n.d. n.d. n.d. n.d. 0.44 0.73 0.25 12.84 0.21
29.75 19.86 7.74 1.7 1.32 1.15 1.3
