Application of 2D-Fluorescence Spectroscopy for On-Line Monitoring

2D-Fluorescence spectroscopy has been shown to be effective for the on-line monitoring of spectroscopic detectable substrates ...
1 downloads 0 Views 423KB Size
Download PDF
Anal. Chem. 2005, 77, 6184-6189

Application of 2D-Fluorescence Spectroscopy for On-Line Monitoring of Pseudoenantiomeric Transformations in Supercritical Carbon Dioxide Systems Torsten Knu 1 ttel,§ Hartmut Meyer,† and Thomas Scheper*,‡

Institut fu¨r Organische Chemie der Universita¨t Hannover, Schneiderberg 1b, 30167 Hannover, Germany, and Institut fu¨r Technische Chemie der Universita¨t Hannover, Callinstrasse 3, 30167 Hannover, Germany

2D-Fluorescence spectroscopy has been shown to be effective for the on-line monitoring of spectroscopic detectable substrates L-phenylalanine-7-amido-4-methylcoumarine (L-PheAMC) and D-phenylalanine-7-amido-4trifluoromethylcoumarine (D-PheAFC) in supercritical carbon dioxide. Earlier investigations with the coumarine substrates in watery and organic phases showed their potential for on-line enantiomeric evaluations of enzymatic reactions in different reaction media. The solubility of the different substrates and their fluorescence maximums were investigated in SCCO2. The sole hydrolyzations of L-PheAMC and D-PheAFC with r-chymotrypsin and the esterase from porcine liver were tracked on-line in the supercritical medium; however, different solubility characteristics of the methyl- and trifluoromethyl-substituted coumarins influence the simultaneous detection of the Land D-substrate within the applied high-pressure reactor system. The usage of enzymes as biocatalysts in supercritical carbon dioxide (SCCO2) has been exploited extensively for the last 20 years. Various reports have described the influence of several parameters toward the catalytic activity and stability of enzymes in SCCO2, such as the influence of temperature, pressure, and humidity and the number of pressurization/depressurization steps.1-3 Fluorescence spectroscopy studies with trypsin showed that enzymes become modified during processing, which leads to changes in the protein conformation.4 Such changes in the unique three-dimensional structure of the biocatalysts may alter their enantioselectivity toward the deployed substrates in SCCO2, resulting in modified product formations. We recently described * Corresponding author. Tel.: +49-511-762-2509; fax: +49-511-762-3004. E-mail: [email protected]. † Institut fu ¨ r Organische Chemie der Universita¨t Hannover. ‡ Institut fu ¨ r Technische Chemie der Universita¨t Hannover. § Present address: GE Healthcare, Bio-Sciences Medical Diagnostics, P.O. Box 4220 Nydalen, 0401 Oslo, Norway. (1) Knez, Z.; Habulin, M.; Primozic, M. Bioprocess Biosyst. Eng. 2003, 25 (5), 279-284. (2) Turner, C.; Whitehand, L; C.; Nguyen, T. J. Agric. Food Chem. 2004, 52 (1), 26-32. (3) Kasche, V.; Schlothauer, R.; Brunner, G. Biotechnol. Lett. 1988, 10, 569574 (4) Zagrobelny, J.; Bright, F. V. Biotechnol. Prog. 1992, 8, 421-423.

6184 Analytical Chemistry, Vol. 77, No. 19, October 1, 2005

the development of a pseudoenantiomeric reaction that can be followed on-line via 2D-fluorescence spectroscopy. With the application of coumarine substrates and products, all of which possess different fluorescent behaviors, it was possible to evaluate on-line the enantioselectivity of proteases and esterases in enzymatic reactions in water and organic solvents.5,6 Our future work has the aim to transfer the pseudoenantiomeric reaction into supercritical liquids to monitor potential changes in the enzyme enantioselectivity due to high pressure, water content of the biocatalyst, pH, and water content and composition of the supercritical mixture on-line. However, before a detailed evaluation of the enzymatic reaction in the supercritical solvent may occur, several aspects, such as the solubility of the coumarine substrates and products in SCCO2 and the position of their fluorescence maximums in the 2D spectra in supercritical CO2, have to be investigated. The present work describes the first-time application of the pseudoenantiomeric substrates in combination with 2Dfluorescence spectroscopy for the evaluation of the enzymatic processes in SCCO2. EXPERIMENTAL SECTION Experimental Setup. In Figure 1, the experimental setup is shown. For our studies, the “source-trap” reactor is used; a more detailed description of the reactor is given in the literature.7 The stainless steel vessel has a wall thickness of 1 cm and an inner volume of 108 mL. The top of the vessel is fitted with six hexagonal screws and sealed with O-rings of Viton 500 (Otto Gehrkens GmbH, Pinneberg, Germany). In addition, the top holds an integral quartz glass window (8.0 mm, Spindler and Hoyer, Germany), which is permeable for the different wavelengths of the 2D spectrophotometer. An adopted fiber-reinforced Teflon packing reduces the risk of breakage of the glass. A liquid-light conductor (2 m, Lumatec, Denzlingen) is fixed 0.1 mm above the glass, centrally held by a plastic adapter (TCI, University of Hanover). The reactor is segmented into two vessels, which allows a combined process of substrate solution in SCCO2 in the source (5) Knu ¨ ttel, T.; Meyer, H. H.; Scheper, T. Enzyme Microb. Technol. 2001, 29, 150-159. (6) Knu ¨ ttel, T.; Meyer, H. H.; Scheper, T. Enzyme Microb. Technol. 2004, in press. (7) Hartmann, T.; Meyer, H. H.; Scheper, T. Enzyme Microb. Technol. 2001, 28, 653-660. 10.1021/ac050747d CCC: $30.25

© 2005 American Chemical Society Published on Web 08/23/2005

Figure 1. Schematic illustration of the experimental setup.

reactor (10 mL volume) and enzymatic reaction in SCCO2 in the trap reactor (65 mL volume). ′′ Steel capillaries (1/16-in.) can be connected to the side of the reaction vessel with conventional HPLC-type ferrules of steel or PEEK (Knauer, Berlin, Germany). CO2 is compressed by a high-pressure pump (1) (Milton Roy, Pont St. Pierre, France) up to the required pressure and pumped into the stainless steel reactor. The pressure is controlled by a manometer (2) (Hensinger & Salmon, Germany), and needlevalves (3) (ERC, Altegolfsheim, Germany) are used to open and close connections. A thermostatic oven (4) (Memmert, Schwabach, Germany) contains the whole apparatus, maintaining a constant temperature of 45 °C in the system. The temperature is controlled by a digital thermometer. From the 2D-fluorescence spectrometer (5), the liquid-light conductor (6) is connected to the reaction vessel via an opening in the oven wall, and the fluorescence measurements take place at the top through the quartz-glass window (7). All spectra were collected on-line with a Hitachi F-4500 2Dfluorescence spectrometer; the measuring setup was already described in our earlier work.5 Due to the high-pressure experimental setup, special care was taken for any activities directly concerning the reaction setup (burst plates were used to avoid pressures higher than 140 bar) or the area surrounding the SCCO2 experiments. Chemicals and Reagents. Carbon dioxide (purity >99.94%) was purchased from Linde AG, Germany. The fluorescence substrate L-phenylalanine-7-amido-4-methylcoumarine (L-PheAMC) was purchased from Sigma (St. Louis, MO), D-phenylalanine-7amido-4-trifluoromethylcoumarine (D-PheAFC) was synthesized by the Department of Organic Chemistry, University of Hanover.5 R-Chymotrypsin (E.C. 3.4.21.1) and the esterase from porcine liver (E.C. 3.1.1.1) were purchased from Sigma (St. Louis, MO). Methods. The solubility studies of the cleavage products 7-amino-4-methylcoumarin (AMC) and 7-trifluoromethylcoumarin (AFC) were monitored on-line in the source-trap reactor. Three milligrams of each coumarin product was placed into the source vessel, and thereafter, the whole reactor was tempered for 2 h at

45 °C. One minute before the filling of the reactor with CO2, the on-line measurement with the 2D-fluorescence spectrometer was started. For these experiments, the time scale courses of AMC and AFC were recorded at λEX ) 340 nm/λEM ) 430 nm and λEX ) 375 nm/λEM ) 440 nm, respectively. A pressure of 100 bar and an oven temperature of 45 °C were applied; the solution of AFC into SCCO2 was monitored for 1 h; and for AMC, the solution was observed for 80 min. The quantitative solubility of AMC and AFC in supercritical CO2 have been previously described.8 To determine the solubility of L-PheAMC and D-PheAFC, the substrates were placed into a stainless steel 66-mL reactor vessel with equal connections as the source-trap reactor but without the integral boric-quartz glass window in the top. Since a low solubility of the substrates in SCCO2 was expected,9,10 no on-line monitoring was applied. Ten milligrams of L-PheAMC and 8 mg of L-PheAFC, placed on a glass vessel, were transferred into the reactor and incubated for 2 h at 45 °C. Thereafter, the reaction vessel was filled with CO2 until a pressure of 100 bar was reached. The substrates were extracted into SCCO2 for 2 h. Thereafter, the pressure was carefully released through the depressurization valve, and the solubility of the coumarine substrates was determined through difference weighing. The stand-alone enzymatic hydrolyses of L-PheAMC and D-PheAFC with R-chymotrypsin or the esterase of PLE was monitored on-line via time scale in the source-trap reactor. An amount of 9 × 10-6 mol of L-PheAMC was placed into the source vessel, the trap vessel was filled with 7000 units R-chymotrypsin (solved in 4 mL of water11), and the reactor was incubated at 45 °C for 2 h. One minute after the time scale measurement was (8) Zimmermann, C. Diploma Thesis. Fluoreszenzspektroskopische Untersuchungen an konventionellen und nicht-konventionellen Lo¨semittelsystemen. University of Hanover, 1997. (9) Calvey, E. M.; Page, S. W.; Taylor, L. T. J. Supercrit. Fluids 1990, 3, 115120. (10) Choi, Y. H.; Kim, J.; Noh, M. J.; Choi, E. S.; Yoo, KP. Chromatographia 1998, 47 (1/2), 93-97. (11) Mishima, K.; Matsuyama, K.; Baba, M.; Chidori, M. Biotechnol. Prog. 2003, 19, 281-284.

Analytical Chemistry, Vol. 77, No. 19, October 1, 2005

6185

Figure 2. Course of solubility of AFC and AMC in SCCO2; time scale measurements, T ) 45 °C, 100 bar.

started, the reactor was filled with CO2 until 100 bar was reached. The reaction was monitored on-line at λEX ) 340 nm/λEM ) 430 nm for 225 min at 45 °C and 100 bar through the supercritical phase. With D-PheAFC, 9 × 10-6 mol of the coumarine substrate was placed into the source vessel, and 1500 units PLE (solved in 300 µL water12) was filled into the trap container. Thereafter, the reaction vessel was incubated at 45 °C for 2 h. The time scale measurement was started, and after one minute, the reactor was filled with CO2 until 100 bar was reached. The reaction was monitored on-line at λEX ) 375 nm/λEM ) 440 nm for 160 min at 45 °C and 100 bar through the supercritical phase. For the simultaneous hydrolysis of L-PheAMC/D-PheAFC in SCCO2 with R-chymotrypsin and the esterase from porcine liver, 9 × 10-6 mol of each substrate was deployed into the source vessel, and 7000 units of R-chymotrypsin (solved in 4 mL water) or 1500 units of PLE (solved in 300 µL water) was placed into the trap container. The reactor was incubated for 2 h at 45 °C and then filled with CO2 until a pressure of 100 bar was reached. 2DFluorescence spectra were recorded for 17 h, and the area for the fluorescence spectra was adjusted to 280-450 nm for excitation and 300-600 nm for emission. The scan rate for each 2Dfluorescence spectrum was 12 000 nm/sec, the increment for the excitation wavelength was 5 nm, the increment for the emission wavelength was 10 nm, and the excitation slit was adjusted to 10 nm and the emission slit to 20 nm. The voltage of the photomultiplier was set at 700 V. RESULTS AND DISCUSSION Table 1 lists the fluorescence maximums of the different coumarin substrates/products in SCCO2. For the coumarine substrates L-phenylalanine-7-amido-4-methylcoumarine and D-phenylalanine-7-amido-4-trifluoromethylcoumarine, the fluorescence maximums are to close to the Rayleigh band for detection. The cleavage products AMC and AFC show their fluorescence maximums at λEM ) 345 nm and λEX ) 430 nm and λEM ) 375 nm and λEX ) 440 nm, respectively. In comparison to the fluorescence maximums in water and organic solutions,3 all maximums of the coumarin substrates and products were shifted toward lower excitation and emission wavelength in SCCO2. (12) Bauer, C.; Steinberger, D. J.; Schlauer, G.; Gamse, T.; Marr, R. J. Supercrit. Fluids 2000, 19, 79-86.

6186 Analytical Chemistry, Vol. 77, No. 19, October 1, 2005

Table 1. Fluorescence Maximums of the Different Coumarin Substrates/Products in SCCO2 coumarin substrate/product

ex/em maximum [nm]

L-PheAMC D-PheAFC

AMC AFC

345/430 375/440

Table 2. Solubility of the Coumarin Substrates/ Products in SCCO2 at 45 °C, 100 Bar, 2-h Extraction coumarin substrate/product

solubility [mol]

L-PheAMC

3.47 × 10-7 1.38 × 10-6 4.00 × 10-7 3.92 × 10-6

D-PheAFC

AMC8 AFC8

Figure 2 shows the solution course of the relative fluorescence intensity (FI) of AFC and AMC in SCCO2. For AFC, the pressure was raised from 100 to 105 bar after 33 min. The pressure increase was connected with a strong increase of the FI from ∼200 to ∼320 and clarifies the influence of the pressure toward the solubility of AFC in SCCO2. Due to the pressure change, the fluorescence maximums of the AFC emission wavelength were lowered. This indicates the influence of the fluid density toward the fluorescence of the coumarine substrates. The cleavage product AMC was less soluble in supercritical CO2, after 1 h. The FI just increased from ∼20 to ∼50. After 60 min, the pressure was raised from 100 to 105 bar, which increased the FI to ∼60. The higher solubility of the trifluoro-substituted coumarin derivate toward the methylsubstituted coumarin derivate in SCCO2 is also shown in Table 2, where quantitative data about the extraction of all the different coumarin substrates and products are listed. Here, the product AFC shows a 10-fold higher solubility in supercritical CO2 than AMC. The coumarin substrates L-PheAMC and D-PheAFC are less soluble in SCCO2 than their hydrolysis products; nevertheless, D-PheAFC possesses a 4-fold higher solubility in the supercritical medium than L-PheAMC. When highly compressed, CO2 has weak van der Waals forces, similar to, for example, fluorocarbon or

Figure 3. Hydrolysis of L-PheAMC and D-PheAFC in SCCO2 with R-chymotrypsin and PLE, respectively. T ) 45 °C, 100 bar.

Figure 4. Simultaneous hydrolysis of L-PheAMC/D-PheAFC in SCCO2 with R-chymotrypsin. T ) 45 °C, 100 bar. (a) 20 min after start in SCCO2, (b) 60 min after start in SCCO2, (c) 17 h after start in SCCO2.

fluoroether,13,14 which favor the solution of the fluorinated coumarine substrate/product. Additionally, coumarin derivates made up by the substitution of two functional groups show a lower solubility in SCCO2 than do those derivates with one substituted functional group.15 (13) Johnston, K. P.; Harrison, K. L.; Clarke, M. J.; Howdle, S. M.; Heitz, M. P.; Bright, F. V.; Carlier, C.; Randolph, T. W. Science 1996, 271, 624-626. (14) Sarkari, M.; Darrat, I.; Knutson, B. Biotechnol. Prog. 2003, 19, 448-454.

The enzymatic hydrolysis of L-PheAMC and D-PheAFC with R-chymotrypsin and esterase PLE, respectively, are shown in Figure 3. For L-PheAMC with R-chymotrypsin, the course of the appearing fluorescence maximums of AMC was clearly visible. After 220 min, an FI of ∼450 was reached, and the hydrolysis was nearly finished. With D-PheAFC and the esterase from porcine (15) Yoo, KP.; Shin, H. Y.; Noh, M. J.; You, S. S. Korean J. Chem. Eng. 1997, 14 (5), 341-346.

Analytical Chemistry, Vol. 77, No. 19, October 1, 2005

6187

Figure 5. Simultaneous hydrolysis of L-PheAMC/D-PheAFC in SCCO2 with esterase PLE. T ) 45 °C, 100 bar. (a) 20 min after start in SCCO2, (b) 60 min after start in SCCO2, (c) 17 h after start in SCCO2.

liver, the end point of the reaction was already reached after ∼120 min with an FI of ∼90, evoked by the higher solution ratio of fluorinated substrate/product in supercritical CO2. The tarnishing of the reaction solution due to the crude enzyme usage caused an overall low FI. This was not experienced when highly purified R-chymotrypsin was used. Proteins are generally insoluble in the nonpolar solvent SCCO2;8 the on-line monitoring of the enzymatic reaction was likely to take place at the water/SCCO2 border. Additional investigations showed a loss of 75% of the FI of the coumarine sustrates/products (data not shown), shown by fluorescence measurements through supercritical phases with the source-trap reactor, which leads to an overall loss of the FI shown by measurements in SCCO2, as compared to the FI of the coumarins experienced in water or organic solvents. However, on-line monitoring of the separated hydrolyzations in supercritical carbon dioxide was possible. The simultaneous hydrolysis of L-PheAFC/D-PheAFC in SCCO2 with R-chymotrypsin is shown in Figure 4. Twenty minutes after the beginning of the reaction in supercritical CO2, just the fluorescence area of the deployed enzyme was visible in the 2Dfluorescence spectrum at low wavelengths. After 60 min in SCCO2, peak maximums at λEM ) 345 nm and λEX ) 430 nm were appearing with an FI of ∼66, which can be attributed to the cleavage product AMC. After 17 h, a fluorescence maximum at λEM ) 375 nm and λEX ) 440 nm with an FI of 290, which belongs to the coumarin product AFC was visible. Earlier investigations 6188 Analytical Chemistry, Vol. 77, No. 19, October 1, 2005

showed that R-chymotrypsin is L-specific toward the coumarin substrates;5 however, proteins such as trypsin and R-chymotrypsin may partly denature in SCCO2, and their unique three-dimensional structure may be affected drastically.16 This may ultimately influence their enantioselectivity; however, the appearing AFC fluorescence peak is more likely to emerge through the instability of the substrate in aqueous solutions6 than through enzymatic cleavage. Further investigations showed that the pH of aqueous solutions in SCCO2 under our experimental conditions decreases to 5.5-5.0,17 which may additionally support the autohydrolysis of the D-substrate in the high-pressure aqueous phase. The simultaneous hydrolysis of L-PheAMC/D-PheAFC with the esterase from porcine liver is shown in Figure 5. After 20 min of extraction in SCCO2, no fluorescence maximum was visible in the 2D spectra. The detection of increased FI in the low-wavelength area near the band of scattered light occurs as a result of the protein fluorescence of R-chymotrypsin. Sixty minutes after the supercritical set, a fluorescence maximum of the product AMC at λEM ) 345 nm and λEX ) 430 nm was clearly visible in the spectrum. The quantitative data showed a higher solubility of the trifluoro-substituted D-substrate in SCCO2; however, the unspecific esterase obviously hydrolyzed primarily the L-substrate in the (16) Zagrobelny, J. A.; Bright, F. Biotechnol. Prog. 1992, 8, 421-423. (17) Tservistats, M. Ph.D. Dissertation. Untersuchungen zum Einsatz von u ¨ berkritischem Kohlendioxid als Medium fu ¨ r biokatalysierte Reaktionen. University of Hanover, 1997.

beginning. After 17 h of extraction/reaction in supercritical CO2, the fluorescence peak was shifted toward a higher wavelength, which indicates the additional hydrolysis of the D-PheAFC substrate at a later time point. A clear fluorescence maximum of the AFC product, as seen in Figure 4c, was not observed. The remaining AMC peak may overlap with the AFC peak, thus complicating the on-line detection. CONCLUSIONS The experiments show that an on-line monitoring of the combined process of extraction/enzymatic reaction with the pseudoenantiomeric coumarins in supercritical carbon dioxide is possible. Factors such as the dropping pH and the solvatochromic nature of the coumarine substrates restricted the fluorescence detection in SCCO2, as compared to investigations in water and organic solutions. Additionally, caution has to be taken in the interpretation of the on-line-monitored data. Since the two coumarin substrates and products exhibit quite different solvent characteristics in the supercritical medium, the former existence of a racemic solution, as present in aqueous/organic media,5,6 is not given. Therefore, a clear on-line assignment of the enantioselectivity of the applied enzymes is restricted through the obtained 2D-fluorescence spectra. Further investigations should provide the complete solution of both coumarin substrates before an enzymatic (18) Hartmann, T. Ph.D. Dissertation. U ¨ berktitisches Kohlendioxid als Reaktionsmedium fu ¨ r die Naturstoff-Synthese - Die enantioselektive Hydrolyze von 3-Hydroxysa¨ureestern. University of Hanover, 2000.

reaction takes place. This may be possible through a stand-alone reactor which contains the biocatalyst externally. Inside the source-trap reactor, the solution of the coumarins can be monitored on-line. When the extraction of the substrates into SCCO2 is finished, the biocatalyst may be added, and a clear starting point of the enzymatic reaction is available.18 The loss of FI by measurements in supercritical CO2 may be compensated by the usage of a shorter liquid light conductor and by reduction of the reactor dimensions. To avoid deactivation/denaturation of the biocatalysts in the supercritical media, surfactant complexes and other stabilization techniques10 contain promising possibilities for utilizations of enzymes in supercritical CO2. Further investigations, such as the determination of kinetic parameters and the application of additional coumarin substrates6 in SCCO2, are necessary for a more complete description of the pseudoenantiomeric enzymatic reaction in the supercritical fluid. ACKNOWLEDGMENT This work was financially supported by the Deutsche Forschungsgesellschaft (Graduiertenkolleg: “Chemische und Technische Grundlagen der Naturstofftransformation”). The authors gratefully thank and acknowledge Thorsten Hartmann for technical support and assistance with the high-pressure reactor system. Received for review May 2, 2005. Accepted July 24, 2005. AC050747D

Analytical Chemistry, Vol. 77, No. 19, October 1, 2005

6189