Assessment of acetone as an alternative to ... - Wiley Online Library

RAPID COMMUNICATIONS IN MASS SPECTROMETRY Rapid Commun. Mass Spectrom. 2009; 23: 2139–2145 Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rcm.4122

Assessment of acetone as an alternative to acetonitrile in peptide analysis by liquid chromatography/mass spectrometry Ria Fritz, Wolfgang Ruth and Udo Kragl* Institute of Chemistry, University of Rostock, Albert-Einstein-Str. 3A, 18059 Rostock, Germany Received 14 April 2009; Revised 10 May 2009; Accepted 10 May 2009

Acetonitrile as a solvent used in liquid chromatography/mass spectrometry (LC/MS) of peptides and proteins is a relatively toxic solvent (LD50 oral; rat; 2,460 mg/kg) compared to alternatives like methanol (LD50 oral; rat; 5,628 mg/kg) and acetone (LD50 oral; rat; 5,800 mg/kg). Strategies to minimize its consumption in LC are either to reduce the inner diameter of the column or replace acetonitrile with a suitable alternative. Methanol is often recommended to replace acetonitrile in peptide analysis. In this study however, the main focus lies on another alternative solvent for LC/MS of peptides; acetone. A number of model proteins were tryptically digested and the peptide solutions were analyzed on a linear trap quadrupole (LTQ) mass spectrometer. The performances of acetonitrile, methanol and acetone were compared according to the quality of the chromatograms obtained and identification of the peptides using the BioWorksTM software developed by Thermo Scientific. In accordance to the elutropic series, acetone was found to significantly reduce the retention times of peptides separated by C18 column material with regard to acetonitrile while methanol led to increased retention times. Acetone was the superior solvent to methanol for most of the tested model proteins reaching similar sequence coverage and numbers of identified peptides as acetonitrile. We therefore propose acetone as an alternative to acetonitrile in LC/MS of peptides. Copyright # 2009 John Wiley & Sons, Ltd.

Acetonitrile as a solvent used in liquid chromatography/ mass spectrometry (LC/MS) of peptides and proteins is relatively toxic compared to alternatives like methanol and acetone. Toxicity data as well as properties of the solvent are summarized in Table 1.1,5 Acetonitrile is most commonly used in the production of pharmaceuticals, especially in the synthesis of peptides. Analytical applications of the popular solvent most often include chromatographic separations and mass spectrometric detection.3,4 Particularly, the analysis of proteins and peptides relies on the unique properties of acetonitrile as the organic component in mobile phases and gradient elution.5 The combination of liquid chromatography and mass spectrometry is applied in the determination of protein primary sequences from enzymatic digests (peptide mapping), localization of post-translational modifications of peptides and proteins and identification of protein degradation products.3,4 Since the onset of the world financial crisis in October 2008 the scientific community is confronted with a shortage of acetonitrile. The solvent is not produced in a specific process, only as a by-product in the production of acrylonitrile and *Correspondence to: U. Kragl, Institute of Chemistry, University of Rostock, Albert-Einstein-Str. 3A, 18059 Rostock, Germany. E-mail: [email protected] Contract/grant sponsor: BMBF within the project Bio-OK and the DFG graduate school 1213 ‘Sustainability in catalysis and technique’.

polyacrylonitrile (PAN). Given the high demand for the polymer a separate production of acetonitrile was not necessary. Due to the economic recession and the resulting crisis in the automobile industry the demands for the polymer are declining and industrial manufacturers were forced to reduce the production or shut down the production plants. A direct result is the current shortage that is likely to last until the end of 2009.2 One strategy to minimize the consumption of acetonitrile in the laboratory is exchanging a column with a relatively large inner diameter (4–5 mm) for a narrow- or nanobore column (2.1–4 mm and 25–100 mm). A reduced i.d. leads to smaller flow rates and smaller quantities of acetonitrile needed for an analysis.6 Another possibility is to replace acetonitrile as the organic compound in the mobile phase with a suitable but more available solvent. In conventional liquid chromatography peptides are detected by monitoring the absorption below 220 nm thereby requiring the mobile phase to ideally have an UV cutoff below 210 nm.7 With a cutoff of 330 nm acetone is not a suitable organic solvent for conventional LC peptide separations (Table 1). Detection of peptides by mass spectrometry, however, does not necessitate UV detection of the peptides thus eliminating the drawback to using acetone instead of acetonitrile or methanol. The aim of this study was therefore to assess alternatives to acetonitrile in LC/MS. Methanol as an often referred to alternative in Copyright # 2009 John Wiley & Sons, Ltd.

2140 R. Fritz, W. Ruth and U. Kragl

Table 1. Selected physical and chemical properties of acetonitrile, methanol and acetone1,5 Property Density [g/mL] Viscosity [cP] Boiling point [8C] Vapor pressure [hPa at 208C] Elutropic strength on C18 [e8] UV cutoff [nm] LD50 (oral; rat) [mg/kg] LC50 (inhalation; rat) LD50 (dermal) [mg/kg]

Acetonitrile

Methanol

Acetone

0.7822 0.38 81.6 118.39 3.1 190 2,460 7,551 ppm (8 h) 2,000 (rabbit)

0.7913 0.55 64.7 33.33 1 205 5,628 64,000 ppm (4 h) 15,800 (rabbit)

0.79 0.36 56.29 245.98 8.8 330 5,800 50,100 mg/m3 (8 h) 7,426 (guinea pig)

peptide analysis2,5 as well as acetone, a solvent that has to date never been mentioned in context with LC/MS and peptide analysis are compared to acetonitrile.

EXPERIMENTAL Materials The proteins holo-transferrin (bovine, approx. 98%), albumin from bovine serum (BSA, 96%,), myoglobin (from horse heart, minimum 90%, and lysozyme (from chicken egg white) were purchased from Sigma Aldrich (Steinheim, Germany). Hemoglobin (from bovine blood >90% was purchased from Fluka BioChemika (Steinheim, Germany). Trypsin (proteomics grade, Sigma, Steinheim, Germany) was used to digest the chosen model proteins in solution. The solvents applied in the LC/MS measurements were acetonitrile with 0.1% formic acid (v/v), methanol with 0.1% formic acid and water with 0.1% formic acid (v/v; all LC/MS Chromatosolv1) purchased from Riedel-de Hae¨n (Seelze, Germany) and acetone (Baker ultra resi-analysed) from J.T. Baker (Deventer, The Netherlands). Ultra-pure water was generated by an Ultra Clear system from SG Wasseraufbereitung und Regenierstation GmbH (Hamburg, Germany). Tryptic digestions were performed in LoBind Protein tubes obtained from Eppendorf (Hamburg, Germany).

Tryptic digestion of model proteins For the assessment of a suitable alternative to acetonitrile as co-solvent for peptide and protein analysis by LC/MS five model proteins were tryptically digested. First 1 mg of each protein was dissolved in 1 mL 100 mM ammonium bicarbonate (pH 8.3) solution in ultra-pure water. The protein solutions were further diluted with the ammonium bicarbonate solution to the final concentration of 100 mg/ mL protein. A total volume of 100 mL was placed in special non-binding 1.5 mL reaction tubes. Then 1 mL of a 1 mg/mL trypsin solution in 1 mM hydrochloric acid and 400 mL ammonium bicarbonate were added. The resulting enzymeto-substrate ratio was 1:100. The reaction tubes were placed in a thermic cycler at 378C overnight for approximately 19 h.

Liquid chromatography/mass spectrometry The digested solutions were analyzed on a Finnigan linear trap quadrupole (LTQ) with a Finnigan Surveyor highpressure liquid chromatography (HPLC) system (Thermo Scientific, Dreieich, Germany). Chromatographic separation of 5 mL injection volume per sample (3 injections per sample) Copyright # 2009 John Wiley & Sons, Ltd.

occurred on a C18 reversed-phase column (BioBasic C18, ˚; 2.1 mm 150 mm, 5 mm particle size, pore diameter 300 A Thermo Scientific, Dreieich, Germany). Gradient elution was applied for the three tested solvents. The acetonitrile gradient program started at 5% acetonitrile in water (0.1% formic acid (FCA)) for 10 min, was then increased in a linear gradient to 80% acetonitrile within 40 min, held at that level for 5 min and was decreased over the next 5 min back to 5% acetonitrile in water. Finally the column was re-equilibrated at that level for 10 min. The same gradient program with a total running time of 70 min at a flow of 150 mL/min was applied to the assessed alternative solvents acetone and methanol (with 0.1% FCA). The column temperature was controlled at 258C. To ensure the stability of the column material for both of the tested solvents at high levels of organic phase the backpressure of the LC system was noted at different elution compositions using the XCalibur Tune program (Thermo Scientific, Dreieich, Germany). Mass spectrometric detection was accomplished with an electrospray ionization (ESI) source in positive ion mode at 5.4 kV spray voltage and 2758C capillary temperature. Data acquisition was carried out by the XCalibur software in a data-dependent triple play. In a first scan event a full scan of all the ions in the trap over the m/z range of 150–2000 was conducted. The four highest mass-to-charge ratios over a threshold of 1000 counts are automatically selected for ZoomScans (scan event 2) and collision-induced dissociation (scan event 3) in the ion trap. The collected data was qualitatively evaluated utilizing the XCalibur software, detected peptides were identified using BioWorksTM 3.3.1 SP1 (Thermo Scientific, Dreieich, Germany).

RESULTS AND DISCUSSION The choice of solvent is a major factor in the separation of peptides in liquid chromatography. Depending on the physical and chemical properties of the organic solvent retention and separation of peptides occurs on the column (Table 1). The relative elution strengths can be estimated using the e8 value. In the elutropic series, each solvent is assigned a e8 value based on relative retention times on a specific support material. The higher the e8 value of a solvent, the lower is the retention of an analyte.5 According to the e8 value for C18 column material the elution strength increases in the order methanol < acetonitrile < acetone. Methanol as the Rapid Commun. Mass Spectrom. 2009; 23: 2139–2145 DOI: 10.1002/rcm

Acetone as alternative to acetonitrile in peptide analysis

2141

Figure 1. Chromatograms of a tryptic digest of BSA using (A) acetonitrile, (B) methanol, and (C) acetone as organic solvent in the mobile phase.

Copyright # 2009 John Wiley & Sons, Ltd.

Rapid Commun. Mass Spectrom. 2009; 23: 2139–2145 DOI: 10.1002/rcm


weakest solvent results, while well resolving the peptide peaks, in higher retention times for the detected peptides of a tryptic BSA digest (Fig. 1(B)). The use of acetonitrile shortens the retention time and leads to sharpened peaks in the chromatogram (Fig. 1(A)). Acetone as the strongest solvent shows further reduction in overall retention time of tryptic peptides (Fig. 1(C)). Similar trends were observed for the other analyzed tryptic digests (data not shown). The results indicate that the application of acetone to the separation of peptides in LC/MS could reduce the overall analysis time in comparison to acetonitrile and specifically methanol. Consumption of organic solvent, total runtime of the chromatographic separation and thereby the costs per analysis could be minimized using acetone instead of acetonitrile or methanol. In HPLC the viscosity of the mobile phase is an important issue, since it is directly related to the backpressure generated. Methanol with the highest viscosity of the three solvents creates greater backpressures than acetonitrile or acetone which possess similar viscosities to each other. Using gradient elution for peptide analysis by LC/MS the influence of water/organic solvent mixtures has to be considered. Viscosities of mixtures are higher than that of the pure solvents alone causing the backpressure to pass through a maximum depending of mobile phase composition.5 Based on the lower viscosities of acetonitrile and acetone the strain on the solvent pumps will due to lower backpressures be less than for methanol, probably extending column lifetime. To assess the potentials of the solvents in peptide analysis the generally preferred solvent acetonitrile is further compared to methanol and the proposed alternative acetone in a database search using the SEQUEST algorithm embedded in the BioWorksTM software.8,9 The identified proteins and their coverage by amino acids are shown in Table 2. It should be noted that the configuration of the mass spectrometer was not optimized for any of the proteins specifically. The numbers of detected peptides and the resulting coverage of amino acid sequences differ strongly depending on the protein digested. Despite the partially low coverage a comparison of the three solvents was undertaken. For BSA and hemoglobin subunit beta the percentages of covered amino acid sequence were similar for all the solvents

Table 2. Amino acid coverage [%] of tryptic protein digests in acetonitrile, methanol and acetone Protein BSA gij1351907 Transferrin gij113911795 hemoglobin, subunit alpha gij122272 hemoglobin, subunit beta gij122571 Myoglobin gij2554649 Lysozyme IPI00600859.1

Acetonitrile

Methanol

Acetone

29.2

31.68

32.46

12.9

n.d.

10.89

77.3

78.25

54.61

71.03

72.41

77.47

66.01

23.75

69.03

25.17

16.33

29.93


tested. Taking into consideration that acetonitrile is routinely used as organic solvent in LC/MS of peptides the obtained results for methanol and acetone were evaluated in comparison to acetonitrile. For three of the six protein digests methanol was the most disadvantageous choice of solvent. Transferrin was not detected in the database search; myoglobin coverage was reduced by approx. 64%, lysozyme coverage by approx. 35%. Sequence coverage was equal to acetonitrile for BSA and hemoglobin increasing only by approximately 1% sequence coverage. Acetone was equal to or better for the analysis of BSA (increased by 11.2%), myoglobin (increased by 9.1%), lysozyme (increased by 18.9%) and hemoglobin subunit beta (increased by 4.6%). Amino acid coverage of transferrin and hemoglobin subunit alpha was significantly decreased, by 15.6% and 29%, respectively. To further evaluate the potentials of methanol and acetone as alternatives to acetonitrile the peptides detected in each solvent were directly compared (Table 3). Based on the assumption that a complete tryptic digest has occurred the identified peptides were labelled according to their position within the amino acid sequence of the proteins.10 For BSA the highest number of peptides was identified by BioWorksTM. Of 80 theoretical peptides in total, 18 were identified in acetonitrile, 23 in methanol and 25 in acetone. The detected peptides however varied with the organic solvent used. Of 38 detected peptides in total, 10 were identified in all 3 solvents, 4 peptides were found only in the acetonitrile runs, 7 only in acetone and 9 only in methanol. Respectively, 4 peptides were detected in both acetonitrile/acetone and methanol/acetone runs. For hemoglobin 10 peptides were identified for subunit alpha (theoretical 14) and 12 for subunit beta (theoretical 17). The numbers of detected peptides differ for the hemoglobin subunit and the organic solvent. Using methanol 9 peptides were identified for subunit alpha, 8 using acetonitrile and 7 using acetone. Subunit beta was detected with 10 peptides using acetonitrile, 11 and 12 peptides using methanol and acetone. The percentage of peptides found in all three solvents is higher than for the other proteins, 6 for subunit alpha and 9 for subunit beta, indicating the usefulness of all three solvents for the analysis of a tryptic digest of hemoglobin. Acetone showed a minor drawback in the coverage of the alpha subunit, missing 3 peptides found in methanol and acetonitrile, while being the most suitable solvent for the beta subunit. According to the gathered data, both coverage of amino acid sequence and identified peptides, the digested peptides of transferrin were least successfully analyzable with the proposed methods. Of theoretically 78 possible peptides only 10 were detected in the different solvents. The most striking result of the comparison is the total lack of detected peptides when methanol is used as co-solvent for the LC/MS analysis. While 4 peptides were identified in both acetonitrile and acetone, 4 could be detected only in acetone and 3 only in acetonitrile. At this point there can only be speculation about the cause of this phenomenon, yet it is possible that a combination of the elution strength of the solvents and the specific properties of transferrin and its peptides could Rapid Commun. Mass Spectrom. 2009; 23: 2139–2145 DOI: 10.1002/rcm


2143

Table 3. Peptides detected by BioWorksTM in acetonitrile, methanol and acetone Protein

Detected peptides

Acetonitrile

Methanol

Acetone

SEIAHR DLGEEHFK LVNELTEFAK SLHTLFGDELCK VASLR DDSPDLPK LKPDPNTLCDEFK YLYEIAR RHPYFYAPELLYYANK YNGVFQECCQAEDK IETMR CASIQK FGER AWSVAR AEFVEVTK LVTDLTK ADLAK YICDNQDTISSK SHCIAEVEK DAIPENLPPLTADFAEDK DAIPENLPPLTADFAEDKDVCK NYQEAK DAFLGSFLYEYSR HPEYAVSVLLR DDPHACYSTVFDK HLVDEPQNLIK LGEYGFQNALIVR VPQVSTPTLVEVSR PESER LCVLHEK TPVSEK RPCFSALTPDETYVPK AFDEK LFTFHAICTLPDTEK KQTALVELLK ATEEQLK TVMENFVAFVDK LVVSTQTALA

VLSAADK VGGHAAEYGAEALER MFLSFPTTK GHGAK VAAALTK AVEHLDDLPGALSELSDLHAHK LRVDPVNFK VDPVNFK LLSHSLLVTLASHLPSDFTPAVHASLDK FLANVSTVLTSK MLTAEEK AAVTAFWGK VKVDEVGGEALGR VDEVGGEALGR LLVVYPWTQR FFESFGDLSTADAVMNNPK VLDSFDNGMK HLDDLK LHVDPENFK LLGNVLVVVLAR EFTPVLQADFQK VVAGVANALAHR

BSA B5 B7 B9 B11 B12 B16 B17-18 B21 B22 B23 B25 B30 B31 B33 B36 B37 B40 B41 B45 B46 B46-47 B48 B49 B50 B53 B55 B57 B59 B63 B65 B66 B69 B70 B71 B73 B76 B77 B80 Hemoglobin subunit alpha

subunit beta

Ha1 Ha4 Ha5 Ha7 Ha8 Ha9 Ha10-11 Ha11 Ha12 Ha13 Hb1 Hb2 Hb3-4 Hb4 Hb5 Hb6 Hb9 Hb10 Hb12 Hb13 Hb15 Hb16

(Continues)




Table 3. (Continued) Protein

Detected peptides

Acetonitrile

Methanol

Acetone

Transferrin T18 T20 T26 T31-32 T34 T36 T49 T50 T59 T61 T76

SAGWNIPMGK ELPDPQESIQR HSTVFDNLPNPEDR DKPDNFQLFQSPHGK DSADGFLK MDFELYLGYEYVTALQNLR GYLAVAVVK TSDANINWNNLK YYGYTGAFR GDVAFVK TYDSYLGDDYVR

M1 M2 M3 M7 M8 M9 M10-11 M14 M15 M17 M18-19

GLSDGEWQQVLNVWGK VEADIAGHGQEVLIR LFTGHPETLEK TEAEMK ASEDLK HGTVVLTALGGILK GHHEAELKPLAQSHATK YNEFISDAIIHVLHSK HPGDFGADAQGAMTK NDIAAK YKELGFQG

Myoglobin

Lysozyme L4 L5 L7 L8 L15

be responsible. Methanol as the weakest solvent may not be able to elute the separated peptides from the column under the given parameters, especially the relatively low column temperature of 258C. The tryptic digestion of the glycoprotein transferrin also leads to glycosylated peptides with high molecular masses that result in low responses in LC/ESI-MS. The identification of such peptides through MS/MS fragmentation is very complex due to the labile nature of the glycosylic bonds that break simultaneously to the peptide bonds.11 A very similar result is shown for myoglobin. Of 19 possible peptides under ideal digest conditions, 11 peptides were identified, but using methanol as organic component of the mobile phase only three peptides were detected. In acetonitrile and acetone two of these peptides were also identified. Furthermore, four peptides were detected in both acetonitrile and acetone, two in acetone alone, two only in acetonitrile and one in acetonitrile and methanol. The lowest number of peptides was detected for lysozyme. An ideal tryptic digest with no missed cleavage sites would lead to 16 peptides. In total 5 peptides were identified, 3 with acetonitrile as organic solvent, 2 with methanol and 4 using acetone. One reason for the low recovery may be the highly basic nature of lysozyme (pI 11) and the resulting peptides. Out of the 11 peptides that were not detected, 8 have a pI value higher than 8. At the low pH of the mobile phase basic peptides receive a higher positive charge than neutral or acidic peptides in the ESI process and therefore a lower mass-to-charge ratio. Depending on the set mass range of the mass spectrometer the charged peptides could not be detected. Copyright # 2009 John Wiley & Sons, Ltd.

CELAAAMK HGLDNYR FESNFNTQATNR NTDGSTDYGILQINSR GTDVQAWIR

CONCLUSIONS Alternative solvents to acetonitrile in peptide liquid chromatography/mass spectrometry (LC/MS) were assessed. Methanol as well as acetone influenced peptide retention times according to the elutropic series. Acetone was found to significantly lower peptide retention times in LC compared to acteonitrile and methanol. In addition to possible reduction of overall analysis time acetone may lead to a more efficient desolvation of the peptides during the ESI process.12 The high volatility represented by the vapour pressure (Table 1) causes a quicker formation of positively charged ions and improved spray stability due to the decreased surface tension of the mobile phase. The main criterion in the assessment was the identification of standard proteins from tryptic digests. While acetone as co-solvent led to similar results in amino acid coverage and number of identified peptides as acetonitrile; the use of methanol mostly decreased both considerably. Acetone is therefore a suitable alternative to acetonitrile in peptide analysis by LC/MS to consider along with the more often recommended methanol. Although the long-term influence of up to 80% acetone on the C18 column has not yet been determined, acetone showed remarkably good qualities as organic solvent in the analysis of tryptic digests in solution. The next step will be to apply the results to peptides gained from in-gel digestions of standard proteins after one-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis (1D-SDS-PAGE) and the investigation of the influence of protein and peptide properties on the detection and identification by LC/MS Rapid Commun. Mass Spectrom. 2009; 23: 2139–2145 DOI: 10.1002/rcm


using acetone as organic component in gradient elution. Based on these findings methods will be optimized according to specific properties like pI, hydrophobicity or post-translational modifications.

Acknowledgements Part of this work is financially supported by the BMBF within the project Bio-OK and the DFG graduate school 1213 ‘Sustainability in catalysis and technique’.

REFERENCES 1. Material Safety Data Sheets for acetonitrile (version 3.1), methanol (version 3.3) and acetone (version 3.1); revision


2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

2145

date 12/11/2008. Available: http://www.sigmaaldrich.com/ safety-center/msds-search.html Tullo A. Chem. Eng. News 2008; 86: 27. Nguyen DN, Becker GW, Riggin RM. J.Chromatogr. 1995; 705: 21. Yates JR. Annu. Rev. Biophys. Biomol. Struct. 2004; 33: 297. Sadek PC. The HPLC Solvent Guide. John Wiley: New York, 1996. Rozing G. LC GC Eur. 2003; 16: 14. Reversed Phase Chromatography: Principles and Methods. Amersham Pharmacia Biotech: Va¨stra, Sweden, 1997. Yates JR, Speicher S, Griffin PR, Hunkapiller T. Anal. Biochem. 1993; 214: 397. Fenyo¨ D. Curr. Opin. Biotechnol. 2000; 11: 391. Olsen JV, Ong SE, Mann M. Mol. Cell. Proteomics 2004; 3: 608. Yu YQ, Fournier J, Gilar M, Gebler JC. Anal. Chem. 2007; 79: 1731. Cech NB, Enke CG. Mass Spectrom. Rev. 2001; 20: 362.
