High‐resolution mass spectrometric analysis of myo‐inositol hexakisphosphate using electrospray ionisation Orbitrap

Rationale The phosphorus storage compound in grains, phytic acid, or myo‐inositol hexakisphosphate (IP6), is important for nutrition and human health, and is reportedly the most abundant organic phosphorus compound in soils. Methods for its determination have traditionally relied on complexation with iron and precipitation, acid digestion and measurement of phosphate concentration, or 31P NMR spectroscopy. Direct determination of phytic acid (and its homologues) using mass spectrometry has, as yet, found limited application to environmental or other complex matrices. The behaviour of phytic acid in electrospray ionisation high‐resolution mass spectrometry (ESI‐HRMS) and its fragmentation, both in‐source and via collision‐induced dissociation, have not been studied so far. Methods The negative ion mass spectrometry and tandem mass spectrometry (MS/MS) of IP6, and the lower inositol pentakisphosphate (IP5), using an ESI‐Orbitrap mass spectrometer is described. The purity of the compounds was investigated using anion‐exchange chromatography. Results IP6 is highly anionic, forming multiply charged ions and sodium adduct ions, which readily undergo dissociation in the ESI source. MS/MS analysis of the phytic acid [M−2H]2− ion and fragment ions and comparison with the full MS of the IP5 reference standard, and the MS/MS spectrum of the pentakisphosphate [M−2H]2− ion, confirm the fragmentation pattern of inositol phosphates in ESI. Further evidence for dissociation in the ion source is shown by the effect of increasing the source voltage on the mass spectrum of phytic acid. Conclusions The ESI‐HRMS of inositol phosphates is unusual and highly characteristic. The study of the full mass spectrum of IP6 in ESI‐HRMS mode indicates the detection of the compound in environmental matrices using this technique is preferable to the use of multiple reaction monitoring (MRM).

the hydrolysis of IP6 using concentrated H 2 SO 4 and HNO 3 and the quantification of the released phosphate using the molybdenum blue test. 10 The determination of IP6 along with the inositol phosphate stereoisomers from soils by Cosgrove 11 in the 1960s was achieved by the hydrolysis of the inositol phosphates followed by paper chromatography of the inositol core. An alternative method for the determination of IP6 using phytase enzymatic digestion has also been used widely; 12 the concentration of phosphate released from the digested IP6 is measured using molybdenum colorimetry. Phytases may, however, not be IP6-specific, and may digest other phosphatecontaining compounds co-occurring in complex environmental matrices.
In recent decades, a range of more instrumental analytical methods for determining IP6 has been developed. Liquid and anion-exchange chromatography have been used to separate, identify and quantify inositol phosphates in food and biological samples on the basis of retention times and peaks areas. 13,14 The methods have contended with the presence of the homologous compounds, the lower inositol phosphates, e.g. pentakisphosphate, tetrakisphosphate, etc., and the stereoisomers of the inositol phosphates in the chiro, scillo, neo, etc., forms making chromatographic separation of the compounds difficult.
These lower myo-inositol phosphates are intermediates in the biosynthesis of IP6, and so are commonly found associated with IP6 in plant extracts. In ion-exchange chromatography systems, IP6 detection uses electrochemical conductivity detection, 15 or post-column derivatisation with Fe(NO 3 ) 3 for spectrophotometric detection. 16 Liquid chromatographic systems have also used refractive index detection of IP6, 17 or more recently inductively coupled plasma mass spectrometry. 18 Surprisingly few studies (see below) have employed direct determination of IP6 using mass spectrometry, perhaps because ion-exchange chromatography liquid chromatography (LC) systems are generally incompatible with mass spectrometers due to metal components in the interface pumping systems and the high ionic strength of mobile phases.
Currently, 31 P NMR spectroscopy is the principal method of characterisation of P in matrices such as soils and manures. 19 To date, there has been little work on the mass spectrometric analysis of IP6. One of the features of electrospray ionisation (ESI) is the formation of salt adducts with ions present in solution. These salt adducts can result in multiple analyte-adduct ions, complicating the recorded spectra and reducing ion yields. This is particularly relevant in the case of IP6 where there is potential for the compound to form adducts with up to twelve cations. The complexity this adds to the identification of IP6 using ESI-MS is seen in the report of Heighton et al 24 where cations were added to the IP6 solution in order to use the formation of adducts to identify acid dissociation constants. Up to 16 ions are identified as IP6 per cluster in the spectrum with Fe 3+ , Na + or Cu 2+ adducts, or a mixture of these metals. The addition of different metals complicated, rather than aided, the interpretation of the mass spectra. Rougemont et al 25 developed a method where ion-pairing chromatography was used to separate IP6 from a whole blood matrix.
The addition of modifiers to the LC eluent resulted in fewer adducts, and therefore simplified the mass spectra and improved the identification of IP6. Accurate mass analysis was, however, not employed in this study, nor was the behaviour of IP6 under ESI conditions studied.
Two studies 26,27 have aimed to determine inositol phosphates in sediments using multiple reaction monitoring (MRM) mass spectrometry. Identification of, not only IP6, but also the lower inositol phosphates (IP5, IP4, IP3, etc.), was on the basis of fragmentation reactions. A third study 28     Negative ion mass spectra obtained by direct infusion on an ESI-Orbtirap: A, IP6 reference standard, B, IP5 reference standard, and C, isolated IP6 in fraction 1 (F1, Figure 4b). Ions a to a' are detailed in Table 1   IC fraction solutions were directly infused at 10 μL.min −1 . The source voltage was set to −3.4 kV, the sheath gas flow rate to 30 arb, the auxiliary gas flow rate to 15 arb and the sweep gas flow rate to 9 arb.

| RESULTS AND DISCUSSION
The HRMS negative ion mass spectrum of IP6 is presented in Figure 2A. The major ions in the mass spectrum are given in Table 1 The HRMS negative ion mass spectrum of the IP5 reference standard is given in Figure 2B. The major ions (Table 1)   We therefore sought to confirm or refute the in-source fragmentation hypothesis by purifying the reference standard. This was achieved by collecting fractions from the ion chromatograph. The mass spectrum of the leading front edge of the IP6 peak (the purest IP6 fraction F1, Figure 4B) is given in Figure 2C, and is very similar to that of the IP6 reference standard, confirming that the ions isobaric with IP5 and IP4 ions observed in the reference spectra arise from in-source fragmentation of IP6 and not contamination by these homologues.
Further analysis of the mass spectra corresponding to the 30-s fractions collected from the chromatograph, suggests that the IP6 peak tails into the later eluting peaks. IP6 appears in each fraction collected although its abundance reduces from fraction 3 to fraction 7. The IP5 [M−2H] 2− ion is the major ion in the mass spectrum for fractions 3, 4,  Table 2 chromatogram are isomers of IP5, and that the IP6 peak elutes in fractions 1 and 2 and then tails through the chromatogram to fraction 7. The larger peak in fraction 4 was determined to be IP5, as confirmed by co-injection of the IP6 and IP5 reference standards.  Letters correspond to annotated ions in the mass spectra shown in Figure 5. The results of this investigation demonstrate the potential for using full scan ESI-HRMS to study inositol phosphates, with clear gains to be made in incorporating the technique into protocols for the exploration of organic phosphorous cycling in the environment at the molecular level.