Identification of a biliverdin geometric isomer by means of HPLC/ESI–MS and NMR spectroscopy. Differentiation of the isomers by using fragmentation “in-source”
- 480 Downloads
A commercially available biliverdin sample was analyzed by means of HPLC/ESI–MS and NMR spectroscopy. It was been found that beside the main IXα 5Z,10Z,15Z isomer, the sample contains also the geometric isomer IXα 5Z,10Z,15E. It was also found the isomers behave differentially upon “in-source” fragmentation in negative ion mode (in contrast to the their behavior upon “in-source” fragmentation in positive ion mode and to their behavior upon MS/MS fragmentation in both modes): the relative abundances of deprotonated molecules and fragment ions are significantly different for both isomers, which can be used as an analytical tool to differentiate between the isomers.
KeywordsBiliverdin Geometric isomer Electrospray ionization Mass spectrometry NMR spectroscopy
Biliverdin (BV) is a tetrapyrrolic pigment, a product of heme catabolism. Usually the term “biliverdin” refers to the main isomer of biliverdin, namely to IXα isomer. However, one should note that other positional isomers of biliverdin have been identified (e.g. XIIIα or IXδ) [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], and also besides the “natural” (5Z,10Z,15Z) isomer the geometric isomers (e.g. (5Z,10Z,15E), at the exocyclic double bonds are possible [12, 13]. Furthermore, the commercially available biliverdin may contain other positional isomers as well . Therefore, we decided to check using HPLC/ESI–MS if biliverdin obtained from a commercial source as hydrochloride contains other isomers.
Results and discussion
As clearly shown in Fig. 1, for both ions two chromatographic peaks are obtained, thus the analyzed biliverdin sample contains two isomers. Taking into account the height of the peaks in the positive ion mode, the ratio of the main isomer to the minor isomer is about 5/1. A similar ratio was obtained by HPLC–UV/Vis analysis as shown in the Supplementary Material. As described further a similar ratio was also obtained by NMR spectroscopy. In the negative ion mode the peak ratio is different (Fig. 1), it indicates that the isomers may behave differentially upon ESI(−) conditions and this problem is discussed further in the text.
As clearly seen in Fig. 4, the spectra of both isomers obtained in the positive ion mode are quite similar. There is no difference in relative abundances of fragment ions. There are only minor differences in relative abundances of protonated dimers and sodium adducts. However, the spectra of both isomers obtained in the negative ion mode are different. Beside the differences in relative abundances of deprotonated dimers and sodium adducts, there is a significant difference in relative abundances of fragment ions (Fig. 4). It should be noted that in a few papers ESI/MS has been successfully applied for biliverdin analysis [16, 17, 18, 19, 20, 21]. However, to the best of our knowledge, our finding is the first one which demonstrates the different ESI/MS behavior of two biliverdin isomers.
We also performed HPLC/ESI–MS/MS analysis of commercial biliverdin sample. However, the MS/MS spectra were very similar in both positive and negative ion mode. The results of HPLC/ESI–MS/MS analysis are presented in the Supplementary Material. In other words, there are differences in MS behavior of the isomers upon “in-source” fragmentation in negative ion mode, however there are no differences in MS behavior of the isomers upon MS/MS fragmentation (in collision chamber). The fragmentation of ions upon MS/MS experiments occurs later than that “in-source”. Therefore, it is reasonable to conclude that before the isomer ions reach the collision chamber, they isomerize to the identical structure. The fragmentation “in source” occurs almost immediately after the transfer of the ions from solution to the gas phase, thus the fragmentation reflects the structural differences of the biliverdin isomers present in solution.
In negative ion mode, deprotonation of biliverdin molecule occurs at a carboxyl group. It is clear that the isomerization of deprotonated biliverdin molecules is not easy as the isomerization of protonated biliverdin molecules.
As shown in Fig. 5 the gas phase stability of the [2BV−H]− ion (deprotonated dimer) of main isomer is definitely higher than the gas phase stability of the [2BV−H]− ion of the minor isomer (it is difficult to rationalize why at a cone voltage of 70 V we deal with an increase in [2BV−H]− ion abundances for both isomers). The gas phase stability of the [BV−H]− ion of the main isomer is also definitely higher than the gas phase stability of [BV−H]− ion of the minor isomer. For both isomers decomposition of ions [BV−H]− begins from the cone voltage 40 V (fragment ion at m/z = 285 is formed, Fig. 5). However, for the main isomer decomposition of [BV−H]− ion is amply compensated by the sensitivity increase (at higher cone voltage more ions reach the high vacuum region). It is also worth adding that the decomposition of ions [BV−H]− is compensated by the decomposition of ions [2BV−H]−. Taking into account the behavior of ions [2BV−H]− (Fig. 5) it is clear that compensation of ion [BV−H]− for the main isomer is more effective.
We have performed the breakdown plots of respective positive ions against cone voltage. As shown in the Supplementary Material, the breakdown plots for both isomers are similar.
Using HPLC-ESI/MS and NMR spectroscopy, the minor isomer IXα 5Z,10Z,15E was found in a commercially available biliverdin sample (beside the main IXα 5Z,10Z,15Z isomer). The isomers behave differentially upon “in-source” fragmentation in the negative ion mode (in contrast to their behavior upon “in-source” fragmentation in the positive ion mode and to their behavior upon MS/MS fragmentation in both modes). It is difficult to rationalize why this very geometric isomer is present in the analyzed biliverdin sample. The geometric isomers are often formed as a result of exposure to light. However, our sample had been stored in the dark and frozen. It should be emphasized that our finding does not exclude such a commercial biliverdin sample from its use for scientific (e.g. analytical) purposes. Quite the opposite, this sample may be useful for analysis of both isomers, at least for semi-quantitative analysis.
Biliverdin (as hydrochloride) was obtained from Sigma-Aldrich (Poznań, Poland) and used without purification.
1H NMR spectra were recorded on Agilent DD2 800 spectrometer, operating at frequency 799.83 MHz. All spectra were measured at 298 K. The signal assignment has been made on the basis of 2D spectra (gCOSY, gHSQCAD, gHMBCAD) and 1D selective NOE measurements (mixing time 500 ms). Samples were prepared by dissolution of 5 mg of biliverdin hydrochloride in 0.7 cm3 of [2H]4-methanol, containing 10% of [2H]2-water and 0.05% of NaOD or in 0.7 cm3 of [2H]3-acetonitrile, containing 10% of [2H]2-water.
1H NMR (CD3CN/D2O): δ = 7.97 (bs, 1H, H-10), 6.69 (bm, 1H, H-21), 6.51 (bs, 1H, H-15?), 6.49 (bs, 1H, H-5?), 6.17 (bm, 1H, H-32), 5.78 (bd, 1H, J = 11.4 Hz, H-22), 5.64 (bd, 1H, J = 18.0 Hz, H-22), 5.41 (bs, 1H, H-33), 4.98 (bs, 1H, H-33), 3.23 (bs, 4H, H-24, H-27), 2.69 (bs, 4H, H-25, H-28), 2.28 (s, 3H, H-31?), 2.27 (s, 3H, H-23?), 2.10 (s, 3H, H-30), 1.48 (s, 3H, H-20) ppm.
1H NMR data assigned to the 5Z,10Z,15Z-isomer (main) (CD3OD/D2O/NaOD): δ = 7.02 (s, 1H, H-10), 6.73 (dd, 1H, J = 11.7, 17.9 Hz, H-21), 6.52 (dd, 1H, J = 11.6, 17.6 Hz, H-32), 6.13 (s, 1H, H-15), 6.08 (s, 1H, H-5), 5.97 (dd, 1H, J = 2.2, 17.6 Hz, H-33), 5.62 (dd, 1H, J = 1.6, 17.9 Hz, H-22), 5.57 (dd, 1H, J = 1.6, 11.7, H-22), 5.36 (dd, 1H, J = 2.2, 11.6, H-33), 2.92 (m, 4H, H-24, 27), 2.35 (m, 4H, H25, 28), 2.18 (s, 3H, H-31), 2.13 (s, 3H, H-23), 2.11 (s, 3H, H-30), 1.84 (s, 3H, H-20) ppm.
1H NMR data assigned to the 5Z,10Z,15E -isomer (minor) (CD3OD/D2O/NaOD): δ = 6.93 (s, 1H, H-10), 6.70 (dd, 1H, J = 11.7, 17.9 Hz, H-21), 6.50 (dd, 1H, J = 11.6, 17.8 Hz, H-32), 6.13 (s, 1H, H-15), 6.03 (s, 1H, H-5), 5.87 (dd, 1H, J = 2.0, 17.8 Hz, H-33), 5.62 (dd, 1H, J = 1.6, 17.9 Hz, H-22), 5.56 (dd, 1H, J = 1.6, 11.7, H-22), 5.37 (dd, 1H, J = 2.0, 11.6, H-33), 2.92 (m, 2H, H-24), 2.89 (m, 2H, H-27), 2.35 (m, 4H, H25, 28), 2.19 (s, 3H, H-31), 2.14 (s, 3H, H-23), 2.08 (s, 3H, H-30), 1.86 (s, 3H, H-20) ppm.
The HPLC-ESI/MS analyses were performed using a Waters model 2690 HPLC pump (Milford, MA, USA), a Waters/Micromass ZQ2000 mass spectrometer (single quadrupole type instrument equipped with electrospray ion source, Z-spray, Manchester, UK). The software used was MassLynx V3.5 (Manchester, UK). Using an autosampler, the sample solutions were injected onto the XBridge C18 column (3.5 µm, 100 × 2.1 mm i.d., Waters). The injection volume was 10 mm3 of biliverdin-containing solution at concentration 0.05 mg/cm3. The solutions were analyzed using linear gradient of CH3CN–H2O with a flow rate of 0.3 cm3/min. The gradient started from 0% CH3CN—95% H2O with 5% of a 10% solution of formic acid in water, reaching 95% CH3CN after 10 min, and the latter concentration was maintained for 10 min.
The mass spectra were recorded in the m/z range 200–1200, in positive and negative modes simultaneously (during the HPLC/ESI–MS analyses the mass spectrometer was switched in the fast mode between the positive and negative ion modes). The electrospray source potentials were: capillary 3 kV, lens 0.5 kV, extractor 4 V, and cone voltage 20–70 V (indicated in each mass spectrum shown). The source temperature was 120 °C and the desolvation temperature 300 °C. Nitrogen was used as the nebulizing and desolvation gas at the flow rates of 100 and 300 dm3/h, respectively.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.