Human Blood and Plasma Partition Coefficients for C4-C8 n-alkanes, Isoalkanes, and 1-alkenes International Journal of Toxicology 31(3) 267-275 ª The Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1091581812442689 http://ijt.sagepub.com Paweł Mochalski1,2, Julian King1,3, Alexander Kupferthaler1, Karl Unterkofler1, Hartmann Hinterhuber4, and Anton Amann1,5 Abstract Human blood:air and plasma:air partition coefficients for C4-C8 n-alkanes, isoalkanes, and 1-alkenes were determined using multiple headspace extraction coupled with solid phase microextraction and gas chromatography. Mean blood:air partition coefficients expressed in the form of dimensionless blood-to-air concentration ratio (g/mLb/g/mLa) were 0.183, 0.416, 1.08, 2.71, and 5.77 for C4-C8 n-alkanes; 0.079, 0.184, 0.473, 1.3, and 3.18 for C4-C8 isoalkanes; and 0.304, 0.589, 1.32, 3.5, and 7.01 for C4-C8 1-alkenes, respectively (n ¼ 8). The reported partition coefficient values increased exponentially with boiling points, molecular weights, and the carbon atoms in the particle. The solubility of 1-alkenes in blood was higher than in plasma, whereas the blood:air and plasma:air partition coefficients of n-alkanes and isoalkanes did not differ significantly. Consequently, additional interactions of 1-alkenes with whole blood seem to occur. The presented findings are expected to be particularly useful for assessing the uptake, distribution, and elimination of hydrocarbons in human organism. Keywords hydrocarbons, partition coefficient, blood, plasma, multiple headspace extraction. Introduction The blood:air partition coefficient (PC, lb:a) is a fundamental physicochemical parameter governing the behavior and fate of volatile chemicals in living organisms. In particular, lb:a is a key determinant of pulmonary gas exchange patterns, which together with ventilatory flow and cardiac output governs both the inhalational uptake of exogenous vapors (eg, of volatile anesthetic agents or environmental contaminants) and the elimination of endogenous compounds via exhalation.1–2 This is especially crucial in the context of breath gas analysis, where a major research goal is to exploit the breath levels of volatile organic compounds (VOCs) for diagnostic purposes.3 Indeed, according to classical pulmonary gas exchange theory,4 breath concentrations of VOCs with low affinity for blood will react very sensitively with respect to changes in ventilation and perfusion, whereas VOCs with higher affinity for blood changes in blood solubility can induce changes in the associated breath VOC output that might in turn be misinterpreted as fluctuations in the endogenous level. Hence, in order to quantitatively assess the variability of breath VOC levels, a precise and reliable determination of the corresponding blood:air PCs is mandatory. Aliphatic hydrocarbons (HCs) with isoprene as their main representative have received special attention in the field of exhaled breath analysis.5–7 This is due to the fact that they have been proposed as the noninvasive markers of numerous diseases or metabolic disorders in the human organism. For instance, n-alkanes and methylated alkanes proved to be useful in distinguishing patients with lung cancer from healthy controls,3,8–12 for recognizing heart rejection after transplantation, 13 breast cancer, 14 or for the detection of oxidative stress.15–18 Within this framework, the knowledge of fundamental physicochemical parameters like blood:gas PCs governing the exhalation behavior of HCs would be highly desirable. 1 Breath Research Institute, Austrian Academy of Sciences, Rathausplatz 4, A-6850 Dornbirn, Austria 2 Institute of Nuclear Physics PAN, Radzikowskiego 152, PL-31342 Kraków, Poland 3 Faculty of Mathematics, University of Vienna, Nordbergstr.15, A-1090 Wien, Austria, 4 University Clinic for Psychiatry, Innsbruck Medical University, Anichstr. 35, A-6020 Innsbruck, Austria 5 University Clinic for Anesthesia, Innsbruck Medical University, Anichstr. 35, A-6020 Innsbruck, Austria Corresponding Author: Paweł Mochalski, Austrian Academy of Sciences, Rathausplatz 4, Dornbirn, A-6850, Austria Email: pawel.mochalski@ifj.edu.pl; anton.amann@i-med.ac.at Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012 268 International Journal of Toxicology 31(3) Currently, the methods for determining blood:air PCs can be classified as experimental and predictive approaches. The experimental methods (often referred as headspace techniques) employ direct measurements of the gas and blood concentrations of an analyte in closed containers under equilibrium conditions.1,19–22 Consequently, all phenomena affecting the solubility of analytes under study (eg, interactions with blood/plasma proteins) are taken into consideration. Despite their direct applicability, such in vitro techniques suffer from various drawbacks related to blood sampling and analytical treatment (eg, adsorption/losses of analytes, changes in solubility due to the blood ageing). On the other hand, in predictive approaches lb:a values are calculated/modeled using other known physicochemical parameters of the compound under scrutiny, including water:air and n-octanol:air PCs, vapor pressures, blood composition, or previously obtained lb:a of homolog compounds.1,23–25 Predictive approaches can, however, lead to wrong results when unknown or additional factors affecting the analyte’s solubility occur (eg, protein interactions).24,26,27 Consequently, in order to improve their predictive power, such algorithms necessarily have to be calibrated against experimentally obtained solubility values. In this context, it becomes clear that ultimately only a carefully performed experimental approach involving real blood samples can provide precise and reliable values for the desired blood:air PCs. Unfortunately, for many classes of compounds experimentally determined values for the blood:air PC are still lacking. This is particularly true for aliphatic HCs. To the best our knowledge, the most extensive study of this parameter for aliphatic HCs was performed by Perbellini et al,21 with other studies providing single values or focusing on selected HCs only.22,26,28,29 The solubility of a few special aliphatic HCs received more attention mainly due to their toxicity (1,3-butadiene)29 or importance in the field (isoprene).28,30 Data on the plasma solubility of aliphatic HCs are even sparser, despite the fact that this parameter could give valuable information on the transport mechanisms of HCs in human blood.26 In our last article,30 we presented a method for the determination of liquid:air PCs, combining the multiple headspace (MHE) extraction technique31,32 with solid-phase microextraction (SPME). The principles and advantages underlying the MHESPME method have been thoroughly discussed in our previous article, and therefore only a very brief description will be presented here. A certain mass (m0) of the compound under study is introduced into a closed vial, containing known volumes of the respective liquid and gas phases. After a liquid:gas equilibrium has been established, the vial headspace is pressurized by injecting a certain volume (Vs) of solute-free gas with the help of a headspace syringe. Next, the same amount of the headspace vapor is withdrawn from the vial using the same syringe and transported into the second vial where the SPME extraction takes place. Finally, the extracted solutes are desorbed in the injector of gas chromatograph and analyzed using gas chromatographic systems. As soon as a new liquid:gas equilibrium is attained in the sample vial, the second pressurization/venting cycle starts and another headspace measurement is conducted, using a new vial for each SPME extraction. This procedure is iterated n times and the liquid:air PC ll:a ¼ Cl/Ca (ie, the dimensionless liquid-to-air ratio [g/ mLl/g/mLa] representing the compound’s solubility into blood or plasma from air33) is calculated from the formula ll:a ¼ ð1 þ Vs  bÞ  b ðVa þ Vs Þ ð1Þ Here, b is the slope of the linear regression performed with the sum of the gas chromatography (GC) signals (peak areas) of the first (i  1) headspace measurements as dependent variable and the signal of the ith measurement as independent variable, b ¼ Va/Vl is the phase ratio (Vl is the volume of liquid phase), Va is the volume of the gas phase in the vial, and Vs is the volume of the gas injected and subsequently withdrawn from the vial with the help of a headspace syringe. Further details on the MHE method are given elsewhere.31,32 The aforementioned method inherits major advantages of the MHE technique. First, each PC value is determined on the basis of several chromatographic analyses and a calibration of GC data (peaks areas) is not necessary. Consequently, the uncertainty of the GC measurement is improved. Moreover, the initial mass of the solute in the vial does not have to be known precisely. These features are of particular importance when the reliable reference materials are very expensive or difficult to produce. The application of SPME with MHE results in some additional benefits. The sample is preconcentrated and focused before introduction into the GC column. This results in narrower peaks and significantly improves the signal-to-noise ratio. In turn, narrow peaks improve peak separation and integration, while a better signal-to-noise ratio and improved detection limits decrease the concentration/mass limit which can be used during the PC estimation. This advantage can be very useful for studies aiming at the determination of PC values for predefined low environmental/physiological concentrations, or using natural background levels of the investigated species. In our recent work, we successfully applied the MHE-SPME method for the measurements of isoprene solubility in water, human blood, and plasma. The present study extends these results by determining the blood:air and plasma:air PCs of C4-C8 n-alkanes, isoalkanes, and 1-alkenes. The blood:air PCs of lipophilic species such as HCs were shown to be associated with the blood lipid composition and, consequently, with the diet regime.29 For instance, for 1,3butadiene a rapid 20% increase in lb:a was reported shortly after the intake of a high-fat meal. Although blood lipid levels were not measured within this study, an effort was made to eliminate diet-related short-time fluctuations of PCs by sampling volunteers that had been fasting for 12 hours. Experimental Materials and Standards Multicompound HC gas mixtures were prepared in a 0.5-L glass bulb (Supelco, Canada) from pure substances. The majority of the latter were purchased from Fluka (Buchs, Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012 269 Mochalski et al Switzerland); n-butane (99%), n-pentane (99.8%), n-heptane (99%), n-octane (99.8%), i-pentane (99.5%), i-hexane (99.5%), i-heptane (98%), 2-methyl heptane (97%), 1-butene (99%), 1-pentene (98.5%), and 1-heptene (99.5%). Moreover, i-butane (99%), 1-hexene (97%), and 1-octene (97%) were obtained from Aldrich (St. Louis, USA), whereas, n-hexane (99%) was provided by Merck (Germany). Prior to its use, the bulb was thoroughly cleaned with methanol and dried at 70 C for at least 12 hours. Then, the bulb was evacuated using a vacuum pump and approximately 6 to 8 mL of liquid (or 1 mL of gaseous) pure HCs were injected through a rubber septum. After complete evaporation of all HCs, the bulb was balanced with nitrogen to achieve ambient pressure. Glass vials with a nominal volume of VT ¼ 10 mL were purchased from Markus Bruckner Analysentechnik (Austria). The volumes of these vials were precisely measured prior to each experiment in order to estimate the exact phase ratios. Butyl/PTFE (Polytetrafluoroethylene) septa, 1.3 mm, were obtained from Macherey-Nagel (Germany). The type of septa was carefully selected with respect to the background and recovery of the compounds of interest. Human Participants and Blood Sampling A cohort of 8 healthy normal volunteers (5 males, 3 females, age range 29-54 years, median 35.5 years) were recruited. All participants gave informed consent to participate and the blood sample collection was approved by the Ethics Commission of Innsbruck Medical University. The volunteers were asked to be in fasting condition for at least 12 hours before blood sampling. An overnight fasting was enforced in order to eliminate any diet-related short-time effects on blood solubility as reported elsewhere.29 The solubility data to be presented thus reflect normalized physiological conditions. Venous blood was sampled in the morning from the median cubital vein into two 9-mL plastic monovette vessels (Sarstedt, Germany), previously rinsed with high-purity air for 4 to 6 hours at 50 C to remove contaminants emitted by the monovettes material. Prior to the sampling procedure, a small amount of heparin (Ebewe Pharma, Austria) was added to the rinsed monovettes. Heparin addition was the only blood modification applied during these experiments. Sample Preparation First, 5 mL of blood were transferred into a sterile glass vial (measured volume 12 mL) containing a stirring bar. Such conditions resulting in a phase ratio of b ¼ 1.4 provided the minimal measurement uncertainty and were selected a priori on the basis of recommendations given by Chai et al31 and measured the blood:air PC data available in the literature.21,22 Next, the vial was crimped and approximately 0.5 mL of the HCs standard gas mixture was injected into the vial with the help of the gas-tight syringe (Hamilton, Bonaduz, Switzerland). Finally, the headspace of the vial was pierced with the help of a needle to balance overpressure. While 5 mL of sampled blood were transferred into the glass vial immediately after sampling, the remaining blood volume was centrifuged (CS-6R Centrifuge, Beckman) to produce plasma. Subsequently, 5 mL of plasma were introduced into a glass vial (as in the case of blood samples) and frozen at 20 C. An effort was made to minimize the storage time of plasma samples, which was usually not longer than 8 hours. Defrosting was performed at room temperature directly before the measurements. The MHE-SPME Procedure and Chromatographic Analysis The MHE-SPME procedure was performed automatically using a multipurpose sampler MPS2 XL (Gerstel, Germany). The sampler was equipped with 2 arms, 1 operating in an SPME mode and 1 operating in a headspace mode. During equilibration, the sample vials (containing blood or plasma) were incubated at 37 C in a temperature-controlled tray and stirred intensively (1200 rpm). After a liquid:air equilibrium state had been reached in the sample vial, 2.5 mL of HC-free air were injected into the headspace of the vial using a Hamilton syringe installed on the headspace arm of the autosampler. The syringe was maintained at 40 C during the whole extraction. After 10 seconds, the same volume of 2.5 mL of headspace gas was transferred into another glass vial (10 mL) prefilled with high-purity nitrogen using the same Hamilton syringe and the headspace arm of the MPS2 sampler. This procedure guaranteed a constant pressure in the sample vial despite removing a portion of the mass of the investigated compounds. The SPME was performed automatically by means of the second arm of the autosampler by inserting a selected SPME fiber (see section on ‘‘Selection of the Optimal SPME Fiber’’ for the selection procedure) into this vial and exposing it to its gas content for 20 minutes. Subsequently, the fiber was immediately introduced into the inlet of the gas chromatograph where the extracted HCs were thermally desorbed at 250 C. The extraction procedure was repeated 7 times for both blood and plasma samples, always using new vial for each SPME extraction. The time between consecutive MHE extractions (equilibration time) was equal to the time of the gas chromatography–mass spectrometry (GC-MS) analysis (45 minutes). This time period was found to be sufficient for attaining a proper equilibrium between the tested liquid phases and air. The GC-MS analyses were performed using an Agilent 7890/5975C GCMS system (Agilent). During the fiber desorption, the split/splitless inlet operated in the splitless mode (1 min), followed by the split mode at a ratio of 1:20. The injector temperature was maintained at 250 C during the whole analysis. The analytes under study were separated using a RtQ-BOND column (30m  0.32mm, film thickness 10 m, Restek, Bellefonte, USA) operated at a constant helium flow of 1 mL/min. The column temperature program was as follows: 40 C for 1 minute, increased to 260 C at a rate of 7 C/min and 260 C for 7 minutes. The mass spectrometer worked in selective ion monitoring (SIM) mode. The selected m/z ratios (see Table 1) allowed for a complete baseline separation of the compounds under study. The dwell time was 50 ms in all cases. Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012 270 International Journal of Toxicology 31(3) Table 1. Retention Times of the Compounds Under Study, Together With Their Respective SIM Parameters (m/z ratios, dwell times) of the MS Measurements and the Relative Standard Deviations of the lb/p:a Estimation Compound isobutane isopentane isohexane isoheptane 2-methyl heptane n-butane n-pentane n-hexane n-heptane n-octane 1-butene 1-pentene 1-hexene 1-heptene 1-octene CAS Number Retention Time (min) m/z (SIM) RSD (%) 115-11-7 78-78-4 107-83-5 591-76-4 592-27-8 106-97-8 109-66-0 110-54-3 142-82-5 111-65-9 106-98-9 109-67-1 592-41-6 592-76-7 111-66-0 11.18 16.27 20.35 23.97 27.20 12.23 16.99 21.09 24.67 27.84 11.68 16.48 20.70 24.34 27.56 43 57 43 43 43 43 43 57 43 43 56 70 56 56 56 15 12 9.5 9.3 9.1 12.5 9.5 9.2 9.9 12.9 13.0 7.5 8.5 6.9 13.2 Abbreviations: MS, mass spectrometry; SIM, selective ion monitoring; RSD, relative standard deviation. The quadrupole, ion source, and transfer line were kept at 150 C, 230 C, and 280 C, respectively. The retention times of the investigated compounds for the applied chromatographic parameters are presented in Table 1. Selection of the Optimal SPME Fiber In our recent article, a carboxen-polydimethylsiloxane (CARPDMS) SPME fiber has been applied for the MHE-SPME measurements of isoprene solubility in water, human blood, and plasma.30 The CAR-PDMS fiber provides excellent extraction efficiency, however, it also has some limitations. Most notably, these include the saturation of its porous solid coating by an excess of volatiles in the sample and the competition between species during adsorption.34 Such phenomena would inevitably distort the MHE-SPME measurements and the ll:a calculations. The aforementioned effects turned out to be negligible when the blood solubility of a single compound is investigated.30 However, in the present study the blood:air PCs were measured simultaneously for all investigated species and hence such problems need to be considered. Consequently, 2 SPME fibers (100 mm PDMS fiber and 75 mm CAR-PDMS fiber, both supplied by Supelco, Canada) were tested by comparing the regression parameters b (see Equation (1)) of n-butane and nheptane obtained for blood samples spiked with all HCs under study (as described in the experimental section) and for blood samples spiked with only n-butane or n-heptane. In the case of the CAR-PDMS fiber, significant differences (20%) between the resulting b parameters were observed. Moreover, the quality of regression in Equation (1) for samples containing the mixture of all HCs was less pronounced, with coefficients of variations R2 <.98. This effect was particularly prominent for lighter, more volatile species (e.g. C4) and could be caused by their displacement from the adsorption sites on the fiber by heavier HCs. Analogous experiments involving the PDMS fiber showed no such effect. The calculated parameters b were very similar in both cases, with differences smaller than the relative standard deviation (RSD) of the method (<10%). This superiority of the PDMS fiber is not surprising, since in liquid PDMS the coating extraction is based on absorption, whereas the solid porous coating of the CAR-PDMS fiber extracts primarily via adsorption—a competitive process. Consequently, the 100 mm PDMS fiber was selected for the MHE-SPME measurements of ll:a. However, due to the lower extraction efficiency of this coating, about 10 times higher masses of the investigated species had to be added in the MHE procedure in order to obtain similar signal-to-noise ratios as with the CARPDMS fiber. The aforementioned finding evidences also the fact that the presence of other HCs in blood or plasma during extraction with PDMS fiber did not modify significantly the PCs of individual species under study. Consequently, PCs of all 15 HCs could be determined simultaneously. Results and Discussion Method Optimization and Verification Stability of plasma samples during storage. The stability of the plasma samples stored at 20 C was assessed by performing several measurements of lp:a in a set of plasma samples produced from blood obtained from 1 single healthy volunteer. The first sample was measured immediately after the plasma had been produced; 2 more after 6 and 24 hours of storage at 20 C. Additionally, 1 plasma sample was stored at room temperature (24 C) and analyzed after 10 hours. The frozen plasma was found to be stable over the investigated time period. The solubility differences between fresh and frozen plasma were smaller than 10% even after 1 day. In the case of plasma stored at room temperature, the calculated PCs were 30% to 130% higher than those computed for fresh plasma, thus confirming a considerable plasma ageing effect. Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012 271 Mochalski et al Table 2. Human Blood:Air (lb:a) and Plasma:Air (lp:a) Partition Coefficients (37C) Obtained Within This Study and Reported in the Literaturea Compound CAS No isobutane isopentane isohexane isoheptane 2-methyl heptane n-butane n-pentane n-hexane n-heptane n-octane 1-butene 1-pentene 1-hexene 1-heptene 1-octene 115-11-7 78-78-4 107-83-5 591-76-4 592-27-8 106-97-8 109-66-0 110-54-3 142-82-5 111-65-9 106-98-9 109-67-1 592-41-6 592-76-7 111-66-0 a Reported lb:a [g/mLb/g/mLa] 0.4121 0.24826 0.3821 0.8021 1.921 2.8522 lb:a [g/mLb/g/mLa] lp:a [g/mLp /g/mLa] Mean Range Mean Range Wilcoxon signed-rank test 0.079 0.184 0.473 1.30 3.18 0.183 0.416 1.08 2.71 5.77 0.304 0.589 1.32 3.50 7.01 0.025-0.131 0.106-0.278 0.307-0.584 1.08-1.55 2.9-3.82 0.096-0.264 0.221-0.534 0.799-1.30 2.36-3.21 5.05-6.46 0.183-0.532 0.351-0.74 1.02-1.51 2.88-4.13 5.51-7.66 0.102 0.242 0.653 1.67 3.88 0.173 0.393 1.06 2.63 6.1 0.242 0.451 1.16 2.76 6.82 0.042-0.201 0.167-0.369 0.402-0.881 1.08-2.16 2.36-4.97 0.064-0.272 0.214-0.572 0.688-1.36 1.69-3.36 3.52-8.62 0.13-0.436 0.283-0.636 0.769-1.47 1.79-3.63 5.78-7.74 n.s. n.s. P < .05 n.s. n.s. n.s. n.s. n.s. n.s. n.s. P < .05 P < .05 P < .05 P < .05 n.s. The right column shows the outcome of a Wilcoxon signed rank test for different median values of blood and plasma solubility. Precision. The RSDs were calculated on the basis of 3 lp:a measurements from consecutive plasma samples. The RSD values presented in Table 1 range from 6.9% for 1-heptene to 15% for isobutane (median 9.5%). The slightly higher RSDs for C4 and C8 HCs result from the fact that the MHE method was optimized for PCs spreading around 1 and, consequently, deviations from this value will affect the RSD31. The coefficients of variation (R2) associated with the MHE regression procedure (ie, leading to the estimation of the parameter b) were very high, taking values over .995 in all cases. In particular, the RSD of b was always better than 5%. Blood:Air and Plasma:Air PCs of C4-C8 n-alkanes, Isoalkanes, and 1-alkenes Table 2 summarizes the lb:a and lp:a of C4-C8 HCs at 37 C. The blood:air PCs have also been presented graphically in Figure 1. The mean PC values ranged from 0.079 for isobutane to 7.01 for 1-octene. The accuracy of the lb:a estimation turned out to be sufficient for differentiating between the investigated HC families. Effectively, a comparison of lb:a values for compounds with the same number of carbon atoms shows that 1-alkenes exhibit the highest solubility in blood, which is on average 20% higher than the solubility of the corresponding n-alkanes. This difference could be explained by the higher polarity of 1-alkenes induced by the presence of an unsaturated bond and, consequently, their tendency to interact with highly polar components of the blood. On the other hand, the blood:air PCs of isoalkanes are approximately two times smaller than the PCs of their nonbranched isomers. However, this difference diminishes with the increase of the number of carbon atoms in the particle. It is worth noting that the PC values of CN isoalkanes are almost identical with the ones of CN-1 n-alkanes, suggesting that the addition of methyl group (as a side chain) to the linear carbon chain does not influence this parameter significantly. The obtained data agree reasonably well with the available literature on human lb:a values (see Table 2). For each family of HCs, an exponential increase of blood:air PCs with boiling point, molecular weights and the number of carbons in the particle was found (see Figure 2). For each individual volunteer, the lb:a dependence with respect to these parameters was modeled using the equation: Log -In lb:a ¼ A þ B  X ð2Þ where X represents boiling point temperature T (K), or number of carbon atoms N (-), or molecular weight m (g). The ranges and means of parameters A and B obtained by curve fitting are presented in Table 3. The corresponding slopes B for n-alkanes and 1-alkenes were found to be very similar for all dependencies, indicating that the PC differences between these 2 families remain roughly similar for higher-order n-alkanes and 1-alkenes. Conversely, the associated slopes for isoalkanes were slightly higher. This suggests a gradual diminution in PC divergences between n- and isoalkanes with an increase of the number of carbon atoms in the particle. As such, it is likely that at a certain point both families will follow the same solubility dependence. It is also interesting to note that the PC data for C4 n-alkanes and 1-alkenes as well as for C4-C5 isoalkanes slightly deviate from the general exponential trend seen in Figure 2. This ‘‘hockey stick’’ effect closely resembles a similar phenomenon reported by Sojak et al35 for the relation between Kovats retention indices and boiling points for HCs. Indeed, this resemblance is not surprising as in both cases partitioning between 2 phases is involved (blood:air and stationary:mobile phases). Particularly, the above-mentioned effect seems to be strongest for isoalkanes and weaker for n-alkanes and 1-alkenes. However, that the first members of the n-alkanes (C1-C3) and 1-alkenes (C2-C3), for which the highest degree of Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012 272 International Journal of Toxicology 31(3) Figure 1. Blood:air partition coefficients for C4-C8 n-alkanes, isoalkanes, and 1-alkenes. deviation can be anticipated, were not measured within this study. The computed parameters A and B were subsequently used to predict blood:air PCs for higher (C9-C11) n-alkanes, isoalkanes, and 1-alkenes (see Table 4). The solubility of HCs in human plasma has received little attention so far. This parameter, however, could give valuable insights into the transport mechanisms of HCs in human blood. For example, some HCs (methane, propane, butane, propene, and butene) were reported to exhibit a higher solubility in whole blood than in plasma, which might be attributable to interactions of these species with blood cells and hemoglobin.26 An inverse effect was observed for isoprene, showing approximately 16% higher plasma:air PCs than blood:air PCs.30 The human plasma:air PC data obtained within this study are presented in Table 2. A Wilcoxon signed rank test was used to compare the median values of blood:air and plasma:air PCs, and a P value of <.05 was considered as significant (Table 2). Apart from 1-octene, the solubility of all investigated 1-alkenes in blood was significantly higher than in plasma. This finding may suggest that some binding mechanisms might exist between the members of this HC family and whole blood proteins (probably due to the increased polarity related to the presence of an unsaturated bond). No such effect could be observed for the remaining two HC families. Within the group of alkanes, only isohexane exhibited a difference between blood and plasma; however, its lb:a was found to be smaller than lp:a. Conclusions The MHE-SPME method coupled with GC-MS was used for determining the blood:air and plasma:air PCs of C 4 -C 8 n-alkanes, isoalkanes, and 1-alkenes. This method has proven to be sufficiently accurate for detecting solubility changes resulting from molecular structure differences among the investigated HC families. In general, 1-alkenes were found to be 20% more soluble in blood than the corresponding n-alkanes. On the other hand, isoalkanes exhibited approximately 2-fold lower values of blood:air PCs than the associated n-alkanes. An exponential increase of HCs PCs with boiling points, molecular weights, and the number of carbons in the particle was found. This dependency was used to predict lb:a values for higher members (C9-C11) of the investigated HC families. 1-alkenes were found to be more soluble in blood than Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012 273 Mochalski et al Figure 2. Dependence of the measured blood:air partition coefficients on the following parameters: number of carbon atoms in particle (panel A), molecular weight (panel B), and boiling point (panel C). in plasma, suggesting additional interactions between these compounds and whole blood proteins. A similar comparison done for the members of the other two families did not show any significant difference (with the exception of isohexane). Since all blood samples were taken following an overnight fast the obtained solubility data are likely to reflect normalized physiological conditions and are expected to be particularly useful for quantitatively modeling the uptake, distribution, and elimination of light HCs in the human organism. Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. Funding The research leading to these results has received funding from the European Community’s Seventh Framework Program (FP7/2007-13) under grant agreement no. 217967 (‘‘SGL for USaR’’ project, Second Generation Locator for Urban Search and Rescue Operations, www.sgl-eu.org). We appreciate funding from the Austrian Federal Ministry for Transport, Innovation and Technology (BMVIT/BMWA, project 818803, KIRAS). We greatly appreciate the generous support of the government of Vorarlberg, Austria. References 1. Fiserova-Bergerova V, Diaz ML. Determination and prediction of tissue-gas partition coefficients. Int Arch Occup Environ Health. 1986;58(1):75-87. 2. Fiserova-Bergerova VE. Modeling of Inhalation Exposure to Vapors: Uptake, Distribution, and Elimination. Boca Raton: CRC Press; 1983. 3. Amann A, Corradi M, Mazzone P, Mutti A. Lung cancer biomarkers in exhaled breath. Expert Rev Mol Diagn. 2011;11(2):207-217. 4. Farhi LE. Elimination of inert gas by the lung. Respir Physiol. 1967;3(1):1-11. 5. King J, Kupferthaler A, Unterkofler K, et al. Isoprene and acetone concentration profiles during exercise on an ergometer. J Breath Res. 2009;3(2):027006. 6. King J, Mochalski P, Kupferthaler A, et al. Dynamic profiles of volatile organic compounds in exhaled breath as determined by a Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012 274 International Journal of Toxicology 31(3) Table 3. Mean Values and Ranges of Parameters of Curve Fittinga for In lb:a ¼ A þ B  X. Number of carbon atoms dependency B (-) R2 A (-) HC family Mean Range Mean Range Mean Range n-Alkanes Isoalkanes 1-Alkenes 0.8494 0.9577 0.8192 0.7904-0.9934 0.8160-1.1948 0.6971-0.9175 5.306 6.493 4.584 6.129 to4.630 8.284 to 5.351 5.391 to 3.656 .9945 .9953 .9905 .9841-.9996 .9883-.9997 .9766-.9981 Boiling point dependency B’ (1/Kelvin) R2 A’ (-) HC Family Mean Range Mean Range Mean Range n-Alkanes Isoalkanes 1-Alkenes 0.0284 0.0343 0.0255 0.0252-0.0318 0.0311-0.0404 0.0218-0.0288 9.566 12.185 8.169 10.943 to 8.430 14.621 to 10.932 9.410 to 6.754 .9927 .9991 .9862 .9736-.9996 .9974-1.0000 .9631-.9979 Molecular weight dependency B’’ (1/g) R2 A’’ (-) HC family Mean Range Mean Range Mean Range n-Alkanes Isoalkanes 1-Alkenes 0.0614 0.0683 0.0584 0.0563-0.0708 0.0582-0.0852 0.0497-0.0654 5.433 6.629 4.583 6.270 to 4.745 8.454 to 5.467 5.390 to 3.655 .9945 .9953 .9905 .9841-.9996 .9883-.9997 .9767-.9981 a X represents either boiling point, molecular weight, or the number of carbon atoms in the particle. Table 4. Average Predicted Blood:Air Partition Coefficients for C9-C11 HCs According to the Exponential Dependences Discussed in Table 3 lb:a (g/mLb/g/mLa) n-Alkanes Mean Range n-Nonane n-Decane n-Undecane 13.9 32.1 78.9 11.1-16.3 23-38 50.2-105 7. 8. 9. 10. 11. lb:a (g/mLb/g/mLa) Isoalkanes Mean Range 2 Methyloctane 2 Methylnonane 2 Methyldecane 8.4 21.5 59.4 6.7-11.9 14.9-33.7 35-107 coupled PTR-MS/GC-MS study. Physiol Meas. 2010;31(9): 1169-1184. Salerno-Kennedy R, Cashman KD. Potential applications of breath isoprene as a biomarker in modern medicine: a concise overview. Wien Klin Wochenschr. 2005;117(5-6):180-186. Phillips M, Altorki N, Austin JH, et al. Prediction of lung cancer using volatile biomarkers in breath. Cancer Biomark. 2007;3(2): 95-109. Phillips M, Altorki N, Austin JH, et al. Detection of lung cancer using weighted digital analysis of breath biomarkers. Clin Chim Acta. 2008;393(2):76-84. Poli D, Carbognani P, Corradi M, et al. Exhaled volatile organic compounds in patients with non-small cell lung cancer: cross sectional and nested short-term follow-up study. Respir Res. 2005;6(1):71. Gordon SM, Szidon JP, Krotoszynski BK, Gibbons RD, O’Neill HJ. Volatile organic compounds in exhaled air from patients with lung cancer. Clin Chem. 1985;31(8):1278-1282. lb:a (g/mLb/g/mLa) 1-Alkenes 1-Nonene 1-Decene 1-Undecene Mean Range 15.2 33.1 73 10.9-19.4 20.8-45 40-108 12. Bajtarevic A, Ager C, Pienz M, et al. Noninvasive detection of lung cancer by analysis of exhaled breath. BMC Cancer. 2009;9:348. 13. Phillips M, Boehmer JP, Cataneo RN, et al. Heart allograft rejection: detection with breath alkanes in low levels (the HARDBALL study). J Heart Lung Transplant. 2004;23(6): 701-708. 14. Phillips M, Cataneo RN, Saunders C, Hope P, Schmitt P, Wai J. Volatile biomarkers in the breath of women with breast cancer. J Breath Res. 2010;4(2):026003. 15. Kanoh S, Kobayashi H, Motoyoshi K. Exhaled ethane: an in vivo biomarker of lipid peroxidation in interstitial lung diseases. Chest. 2005;128(4):2387-2392. 16. Phillips M, Cataneo RN, Greenberg J, Grodman R, Salazar M. Breath markers of oxidative stress in patients with unstable angina. Heart Dis. 2003;5(2):95-99. 17. Scholpp J, Schubert JK, Miekisch W, Geiger K. Breath markers and soluble lipid peroxidation markers in critically ill patients. Clin Chem Lab Med. 2002;40(6):587-594. Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012 275 Mochalski et al 18. Mendis S, Sobotka PA, Leja FL, Euler DE. Breath pentane and plasma lipid peroxides in ischemic heart disease. Free Radic Biol Med. 1995;19(5):679-684. 19. Sato A, Nakajima T. Partition coefficients of some aromatic hydrocarbons and ketones in water, blood and oil. Br J Ind Med. 1979;36(3):231-234. 20. Smith AQ, Campbell JL, Keys DA, Fisher JW. Rat tissue and blood partition coefficients for n-alkanes (C8 to C12). Int J Toxicol. 2005;24(1):35-41. 21. Perbellini L, Brugnone F, Caretta D, Maranelli G. Partition coefficients of some industrial aliphatic hydrocarbons (C5-C7) in blood and human tissues. Br J Ind Med. 1985;42(3):162-167. 22. Gargas ML, Burgess RJ, Voisard DE, Cason GH, Andersen ME. Partition coefficients of low-molecular-weight volatile chemicals in various liquids and tissues. Toxicol Appl Pharmacol. 1989; 98(1):87-99. 23. Beliveau M, Krishnan K. Estimation of rat blood:air partition coefficients of volatile organic chemicals using reconstituted mixtures of blood components. Toxicol Lett. 2000;116(3): 183-188. 24. Paterson S, Mackay D. Correlation of tissue, blood, and air partition coefficients of volatile organic chemicals. Br J Ind Med. 1989;46(5):321-328. 25. Meulenberg CJ, Vijverberg HP. Empirical relations predicting human and rat tissue:air partition coefficients of volatile organic compounds. Toxicol Appl Pharmacol. 2000;165(3):206-216. 26. Poyart C, Bursaux E, Freminet A, Bertin M. Interactions of short chain aliphatic hydrocarbons with human blood and haemoglobin a solutions. Biomedicine. 1976;25(6):224-227. 27. Poyart C, Bursaux E, Guesnon P, Bohn B. Interaction of butene with human hemoglobin A. Eur J Biochem. 1979;99(1):85-88. 28. Filser JG, Csanady GA, Denk B, et al. Toxicokinetics of isoprene in rodents and humans. Toxicology. 1996;113(1-3): 278-287. 29. Lin YS, Smith TJ, Wypij D, Kelsey KT, Sacks FM. Association of the blood/air partition coefficient of 1,3-butadiene with blood lipids and albumin. Environ Health Perspect. 2002;110(2): 165-168. 30. Mochalski P, King J, Kupferthaler A, Unterkofler K, Hinterhuber H, Amann A. Measurement of isoprene solubility in water, human blood and plasma by multiple headspace extraction gas chromatography coupled with solid phase microextraction. J Breath Res. 2011;5(4):046010. 31. Chai XS, Zhu JY. Simultaneous measurements of solute concentration and Henry’s constant using multiple headspace extraction gas chromatography. Anal Chem. 1998;70(16): 3481-3487. 32. Kolb B, Ettre LS. Static Headspace-Gas Chromatography: Theory and Practice. New York, NY: Wiley-VCH; 1997. 33. Staudinger J, Roberts PV. A critical compilation of Henry’s law constant temperature dependence relations for organic compounds in dilute aqueous solutions. Chemosphere. 2001;44(4): 561-576. 34. Pragst F. Application of solid-phase microextraction in analytical toxicology. Anal Bioanal Chem. 2007;388(7):1393-1414. 35. Sojak L, Krupcik J, Rijks J. Correlations between boiling points and relative retention data for hydrocarbons. Chromatographia. 1974;7(1):26-29. Downloaded from ijt.sagepub.com at ETH Zurich on June 10, 2012