Laboratory Exercises

Using HPLC-Mass Spectrometry to Teach Proteomics Concepts withProblem-based Techniques

Received for publication, February 24, 2009, and in revised form, November 16, 2009

Michael Short¶, Anne Short§, Rachel Vankempen‡*, Michael Seymour‡, and Maria Burnatowska-Hledin‡§

From the ‡Department of Chemistry, Hope College, Holland, Michigan 49423, §Department of Biology,Hope College, Holland, Michigan 49423, *Montana State University, Bozeman, Montana

Practical instruction of proteomics concepts was provided using high-performance liquid chromatogra-phy coupled with a mass selective detection system (HPLC-MS) for the analysis of simulated proteindigests. The samples were prepared from selected dipeptides in order to facilitate the mass spectralidentification. As part of the prelaboratory preparation, students calculated the parent ion patterns of thedipeptides using peptide calculator websites. Following instruction on the use of the HPLC-MS instru-ment, students analyzed mixtures of the dipeptides and identified the individual dipeptides in theunknowns. In addition, purchased chicken egg white lysozyme alkylated with iodoacetamide anddigested with trypsin was analyzed using the same approach. Key tryptic peptides were identified fromthe HPLC-MS chromatogram with information generated with the FindPept tool. This experimentdemonstrates that complex concepts can be taught in the undergraduate biochemistry laboratory usinga problem-based approach.

Keywords: Proteomics, protein sequence, mass spectrometry, HPLC-MS, lysozyme.

‘‘The great difficulty in education is to get experi-ence out of ideas’’

George Santayana

Santayana, the Spanish-born philosopher and socialcommentator, recognized that communication of ideas inthe classroom did not always result in understanding,which he called ‘‘experience.’’ The experiment describedhere is used in our biochemistry course to communicatecomplex ‘‘ideas’’ (i.e. concepts) in a practical, problem-based laboratory exercise. We wanted students to under-stand how the researcher can use high-performance liq-uid chromatography-mass spectrometry (HPLC-MS) toidentify peptides that are present in a mixture.

Our upper-level undergraduate biochemistry laboratoryis a five-hour course that meets weekly for one-half ofthe semester. Its purpose is to introduce students toconcepts and techniques used in research labs. Thecourse exercises include kinetic analysis of mushroomtyrosinase [1], the purification and analytical characteriza-tion of lysozyme [2], a bioinformatic exercise for lyso-zyme [3], and the HPLC-MS analysis of a simulatedproteolytic digest using purchased dipeptides and tryp-sin-digested lysozyme. Hands-on use of the HPLC-MSlets students become familiar with a potent, bioanalytical

technique, which can then be applied to samples thatwill be isolated and prepared in the research lab.

Students choose to purify lysozyme from one of sev-eral avian egg whites sources. Our laboratory courseuses five species of avian eggs in the exercise: Chicken(Gallus gallus), turkey (Meleagris gallopavo), duck (Anasplatyrhynchos), bobwhite quail (Colinus virginianus), andostrich (Struthio camelus). Students learn how to purifyproteins by ion-exchange chromatography. They analyzetheir protein containing fractions for lysozyme enzymeactivity against suspensions of Micrococcus lysodeikticuscell walls. Students also determine the molecular weightof their isolated lysozyme using SDS-Polyacrylamide gelelectrophoresis against proteins of known molecularweights. Finally, they review the differences in the lyso-zyme proteins from different species with a bioinfor-matics exercise.

The tertiary structure of a protein, and hence, the pro-tein’s function is determined by the amino acid sequence[4]. Sequencing protein is generally more difficult thansequencing DNA. The emergence of mass spectral tech-niques, however, has made the determination of proteinsequences and molecular size much easier. Finehout andLee have provided an excellent review of MS techniquesin biological research [5]. Matrix-assisted laser desorp-tion/ionization (MALDI) time-of-flight (TOF) has becomethe instrument of choice for protein sequence/proteomicsresearch [6, 7]. In fact, Albright et al. have recently dem-onstrated that in-gel digestion of proteins followed byMALDI-TOF MS analysis can be used in the classroom

¶To whom correspondence should be addressed. AP SchaapScience Center, 35 E. 12th St., Holland, MI 49423, USA. Tel.:616-395-7281. E-mail: short@hope.edu.

DOI 10.1002/bmb.20380 This paper is available on line at http://www.bambed.org242

Q 2010 by The International Union of Biochemistry and Molecular Biology BIOCHEMISTRY AND MOLECULAR BIOLOGY EDUCATION

Vol. 38, No. 4, pp. 242–246, 2010

or laboratory [8]. Unfortunately, MALDI-TOF and MALDI-TOF/TOF instruments are very expensive, sometimescosting in excess of $0.5 million. HPLC systemsequipped with mass selective detectors (MSD) can bepurchased for about a fifth of the cost of MALDI-TOFinstrumentation. Although this is still expensive, theacquisition of an MSD is within the means of manyundergraduate institutions. Our experiment describedhere was developed using an HPLC-MS instrument,which had a MSD.

EXPERIMENTAL PROCEDURES

Equipment

An Agilent 1100 High Performance Liquid Chromatographequipped with a binary gradient pumping system and polysty-rene-divinylbenzene HPLC column, PRP-1 (Hamilton) was usedfor the peptide separation. The HPLC system was equippedwith an Agilent 1000 Diode Array Detector (DAD) in series withan Agilent 1100 Series LC/MSD mass spectrometer detectionsystem equipped operated in the electrospray ionization mode.The HPLC-MS system was controlled with integrated HPCOREChemStation software (Creation Date 2004).

Materials

Dipeptides were purchased from Sigma-Aldrich (St. Louis,MO) and are shown in the Table I.

Chicken egg white lysozyme, dithiothreitol, and iodoacet-amide were obtained from Sigma-Aldrich (St. Louis, MO). Tryp-sin was purchased from Invitrogen (Carlsbad, CA). The HPLCcolumn, a poly(styrene-divinylbenzene) PRP-1 (4.1 mm 3 150mm, 5 l), was purchased from Hamilton (Reno, NV). DuraporeHPLC syringe filters were obtained from Millipore (Billerica, MA).HPLC-grade acetonitrile and formic acid used to prepare mobilephases for HPLC-MS were obtained from Sigma-Aldrich (St.Louis, MO). Distilled, deionized water was used to prepare buf-fers and HPLC mobile phases. Chemicals used to prepare buf-fers and solutions that are not specified were of reagent gradequality or better.

Procedure

Solutions of the individual dipeptides were prepared at con-centrations of 5 mg mL21 in 0.3% aqueous formic acid. Binarymixtures of the dipeptides were then prepared by mixing equalparts of the individual dipeptide solutions. The concentration ofeach dipeptide in the binary mixture was 2.5 mg mL21. Thishigh concentration of the dipeptide mixtures was chosen sothat students would not struggle with the detection limits of theinstrument.

Chicken egg white lysozyme (100 mg) was weighed and dis-solved in 20 mL of 0.025 M ammonium bicarbonate, pH 8.5. A10-fold molar excess of dithiothreitol solution (0.070 mL at 1 Min 0.025 M ammonium bicarbonate buffer, pH 8.5) was added

to the 5 mg mL21 solution of lysozyme. The solution was mixedwell and incubated at 378C for 30 min to reduce the disulfidebonds in lysozyme. A fivefold molar excess of 1 M iodoacet-amide (0.035 mL, made up in 0.025 M ammonium bicarbonate,pH 8.50) was added to the lysozyme-DTT solution and incu-bated at 378C for 1 hour. The alkylated lysozyme was dialyzedat 2–88C against 50 mL of 0.025 M ammonium bicarbonate, pH8.5. Buffer was changed (33) every 45 min. The dialyzed lyso-zyme solution (assumed to be 20 mL) was then treated with2.5 mL of a 1.0 mg mL21 trypsin solution prepared in 0.025 Mammonium bicarbonate, pH 8.5. The mixture was incubated at378C for 4 hours. The reaction mixture was quenched by adjust-ment to a pH of about 3 with 6 M acetic acid [9].

HPLC mobile phase was programmed as a binary gradient: i)0.3% (v/v) aqueous formic acid (polar mobile phase) and ii) ace-tonitrile containing 0.3% (v/v) formic acid (nonpolar mobilephase). The flow rate for the separation was 0.4 mL min21. Sol-utions were filtered through 0.45 lm Durapore (PVDF) filtersbefore injection on the HPLC-MSD system, with an injectionvolume of 5 lL. The column temperature for the separationswas maintained at 258C.

The mobile phase program used for the separation of dipep-tides is shown in Table II.

The mobile phase program used for the separation of thetrypsin digest of acetamidylated lysozyme is shown in Table III.

The DAD was set to collect and store data at two wave-lengths, 222 nm and 274 nm. The MSD was used in the electro-spray ionization mode with positive polarity, that is, ESIþ. TheMSD was set to collect masses for ions from 100 to 350 massunits and 100 to 1000 mass units for dipeptides and trypsinizedlysozyme, respectively. The gas temperature for the MSDwas 3508C and the drying gas was delivered at a rate of 13.0 Lmin21.

Parent ion masses for the dipeptides were determined usingthe following web-based calculator [10]:

http://db.systemsbiology.net:8080/proteomicsToolkit/FragIonServlet.html

The interpretation of the tryptic map of acetamidylatedchicken egg white lysozyme was facilitated using the ExPASyFindPept tool [11] provided at the following URL: http://www.expasy.org/tools/findpept.html

TABLE IDipeptides used in this study

Dipeptide3-Letter code(1-letter code) Property

Glycine-D,L-aspartic acid Gly-Asp (GD) Polar, acidicTyrosine-alanine Tyr-Ala (YA) NonpolarLysine-valine Lys-Val (KV) Polar, basicAspartic acid-phenylalanine

-O-methyl esterAsp-Phe-O-Me

(DF-O-Me)Nonpolar

TABLE IIThe mobile phase program used for the separation of dipeptides

Time (min)% Nonpolar mobile phasevs. the total composition

0.00 0.05.00 4.05.10 35.0

12.00 85.013.00 85.015.00 0.0

TABLE IIIThe mobile phase program used for the separation of the trypsin

digest of acetamidylated lysozyme

Time (min)% Nonpolar mobile phasevs. the total composition

0.00 0.05.00 4.05.10 35.07.00 35.07.10 75.0

15.00 75.020.00 95.020.10 0.0

243

The protein sequence of the chicken egg white lysozymewas obtained from the National Library of Medicine public data-base [12] located at the following URL address: http://www.ncbi.nlm.nih.gov/sites/entrez?db¼protein

Analysis of the peptide fragments of the chicken egg whitelysozyme was performed with the FindPept tool assuming uni-form acetamidylation of cysteines following reduction of cystinedisulfides.

RESULTS AND DISCUSSION

The concept to use unknown mixtures of dipeptidesfor this experiment derives from the model of elementalion identification used in qualitative analysis lab experi-ments [13]. There is no reason that tri-, tetra-, penta- andother oligo-peptides could not be incorporated into thestudent’s unknown mixtures. As four dipeptides werepurchased (i.e. Gly-Asp, Lys-Val, Tyr-Ala, and Asp-Phe-O-Me), there were six possible unknown mixtures con-taining two of the dipeptides.

The dipeptides were selected because of their differentpolarities. Gly-Asp is acidic, and hence, is a polar dipep-tide. Lys-Val is basic, and hence, is also a polar mole-cule. The PRP-1 column binds molecules on the basis oftheir nonpolarity, releasing them when the equilibriumsolubility of the dipeptide in the mobile phase exceeds

the equilibrium affinity of the dipeptide for the solid sup-port. The HPLC-MS chromatogram of the mixture isshown in Fig. 1a. Lys-Val and Gly-Asp coelute with eachother on the PRP-1 (Fig. 1b). Tyr-Ala is less polar than ei-ther of the previous two dipeptides and elutes second inretention order on the column (Fig. 1c). The presence ofthe methyl ester in Asp-Phe-O-Me causes this dipeptideto be retained the longest and the acetonitrile concentra-tion of the mobile phase had to exceed 50% in order forthis nonpolar dipeptide to leave the solid phase HPLCsupport and travel with the nonpolar mobile phase(Fig. 1d).

The theory of reverse phase HPLC separations, whichis based on nonpolar affinities of the dipeptides for theHPLC solid support, became real for the students asthey observed the retention order of different dipeptides.Also, there was a simple elegance in this mixture of fourdipeptides because the mass spectrometric detectorsoftware permits extraction of an ion chromatogram fromthe total ion chromatogram (TIC). The Gly-Asp parent ionhad an m/z of 191.0, that is, (MþH)1þ. The Lys-Val parention was found at m/z of 246.1, that is, (MþH)1þ, and the(Mþ2H)2þ ion was observed at m/z of 123.6. The TICcontains the sum of the ion intensities observed in theprespecified mass range and showed the presence of

FIG. 1. HPLC-MS Chromatogram of a Mixture of Four Dipeptides with Associated Mass Spectra of Chromatographicpeaks. (a) The total ion HPLC-MS chromatogram (TIC) of the mixture of four dipeptides. The mass spectra of the componentswithin the separated peaks are the following: (b) Mass spectrum of peak containing Gly-Asp with parent ion (MþH)1þ of191.0 amu, and Lys-Val with parent ion (MþH)1þ of 246.1 amu, (c) mass spectrum of peak containing Tyr-Ala with parent ion(MþH)1þ of 253.1 amu, and (d) mass spectrum of peak containing Asp-Phe-O-Me with parent ion (MþH)1þ of 295.1 amu.

244 BAMBED, Vol. 38, No. 4, pp. 242–246, 2010

only three peaks from the mixture of the four dipeptides(Fig. 1a). However, using the extracted ion chromato-grams (EIC) software tool for specified ions, peaks origi-nating from Gly-Asp and Lys-Val, were extracted fromthe first peak using unique ion masses of the two dipep-tides shown in the mass spectrum (Fig. 2a). A slight shiftin the peak retention times (2.928 and 3.001 min) and adifferent peak shape was observed for the two dipepti-des (Figs. 2b and 2c, respectively). The unique capabil-ities of the MSD were used to show the students howamino acid sequence data could be derived, as in thecase of Lys-Val, which had an (Mþ2H)2þ ion of 123.6mass units in the ESIþ detection mode (Fig. 2b). Thesecond and third peaks from a mixture of all four dipepti-des were identified as Tyr-Ala and Asp-Phe-O-Me,respectively, using their expected parent ion m/z ratios,(MþH)1þ (Figs. 1c and 1d).

Sequencing of proteins using HPLC-MS involves chro-matographic separation of a proteolytically digested pro-tein followed by identification of the resultant peaks inthe protein digest. A TIC of the tryptic digest of pur-chased chicken egg white lysozyme (Fig. 3) is shownsuperimposed with unique peptides (3i, 3ii, and 3iii) thatwere identified in the protein digest using the EIC soft-ware tool. No attempts were made to curb autoproteoly-sis of the lysozyme in the presence of the trypsin. Someof the peaks contain peptides that were autoproteolyti-

cally degraded beyond the specific cleavages by trypsinat the C-terminus of lysine and arginine residues. Forexample, the peak at about 4.5 min (3i) shows evidencefor the peptide C(Cam)NDGR with a (MþH)1þ of 621(Note: (Cam) indicates a cysteine, which has been car-boxamidomethylated with iodoacetamide) using the Find-Pept tool [11]. This smaller peptide derives from the peakat about 12.2 (3iii) min (MþH)1þ of 993 corresponding to(R79)/WWC(Cam)NDGR/(T87) by specific tryptic cleavageat the arginines, 79 and 86. Autoproteolysis causesremoval of the N-terminal tryptophan residues andresults in the smaller pentapeptide.

At the final stage, students were asked to use bioinfor-matic tools to identify the peak with (MþH)1þ of 993 fromlysozyme digested with trypsin. Students successfullyidentified the peptide with the skills they had acquiredfrom the previous 5-hour bioinformatics lab. In a post-labexercise, the chromatogram of the trypsinized chickenegg white lysozyme was further probed using theextracted ion chromatograph tool for (MþH)1þ and(Mþ2H)2þ ions that would be produced by cleavage withchymotrypsin. Only two small peaks were observed inthe mass range from 500 to 1000 m/z. These were notsignificant because the signal-to-noise, S/N, ratios forthe peaks were less than 10. By comparison, peptidepeaks derived by trypsin cleavage had S/N ratiosthat ranged from 20 to 60. This showed that the 4-hour

FIG. 2. Chromatograms of a Binary Co-Eluting Mixture of Gly-Asp and Lys-Val. (a) Mass spectrum of peak at 3.082 min fromwhich extracted ion chromatograms in (b) and (c) were obtained. (b) Extracted ion chromatogram (EIC) for (MþH)1þ of 191.0 forGly-Asp dipeptide from TIC peak at retention time 2.961 min observed in Fig. 2. Note retention time of EIC peak is recordedat 2.928 min. (c) Extracted ion chromatogram (EIC) for (Mþ2H)2þ of 123.6 for Lys-Val dipeptide from TIC peak at retention time2.961 min shown in Fig. 1. Note retention time of EIC peak is recorded at 3.001 min.

245

trypsinization of lysozyme was successful and that stu-dents could use the resulting chromatograms of thedigest for doing preliminary protein sequence analysis.

In summary, we have shown that HPLC-MS instrumen-tation can be incorporated into the upper-level under-graduate or graduate level biochemistry laboratory toillustrate how protein sequence information is derived.Unique chromatographic peaks of unknown mixtures canbe separated and identified by the students with this toolusing peptides with different polarities. By analyzingunknowns in the laboratory, students learn how to obtainthe needed data from the instrument to make identifica-tion of their unknown mixtures. Finally, students areexposed to proteomics tools that are often saved for theresearch laboratory at the graduate level. In fact, as onestudent wrote: ‘‘This is a very fun and interesting labbecause you get to use instruments that others have noidea on how to run.’’

Acknowledgments—The authors thank the biochemistry labo-ratory students from Winters 2008 and 2009.

REFERENCES

[1] R. Boyer (2006) Biochemistry Laboratory: Modern Theory and Tech-niques, Pearson Benjamin Cummings, San Francisco.

[2] M. O. Hurst, M. V. Keenan, C. C. Soon (1992) Purification of lyso-zyme by linear salt gradient and SDS gel electrophoresis, J. Chem.Educ. 69, 850–851.

[3] A. L. Feig, E. Jabri (2002). Incorporation of Bioinformatics Exercisesinto the Undergraduate Biochemistry Curriculum, Biochem. Mol.Biol. Educ. 30, 224–231.

[4] D. Givol, F. De Lorenzo, R. F. Goldberger, C. B. Anfinsen (1965) Di-sulfide interchange and the three-dimensional structure of proteins,Proc. Natl. Acad. Sci. USA 53, 676–684.

[5] E. J. Finehout, K. L. Lee (2004) An introduction to mass spectrometryapplications in biological research, Biochem. Mol. Biol. Educ. 32, 83–100.

[6] F. Brouillard, N. Bensalem, A. Hinzpeter, D. Tondelier, S. Trudel, A.D. Gruber, M. Ollero, A. Edelman (2005) Blue native/sds-page anal-ysis reveals reduced expression of the mcica3 protein in cystic fi-brosis knock-out mice, Mol. Cell. Proteomics 4, 1762–1771.

[7] B. Meyer, I. Wittig, E. Trifilieff, M. Karas, H. Shagger (2007) Identifi-cation of two proteins associated with ATP synthase, Mol. Cell. Pro-teomics 6, 1690–1699.

[8] J. C. Albright, D. J. Dassenko, E. A. Mohamed, D. J. Beussman(2009) Identifying gel-separated proteins using in-gel digestion,mass-spectrometry, and database searching, Biochem. Mol. Biol.Educ. 37, 49–55.

[9] W. S. Richardson, III, L. Burns (1988) HPLC of the polypeptides in ahydrolyzate of egg-white lysozyme, J. Chem. Educ. 65, 162–163.

[10] Fragment Ion Calculator. Available at: http://db.systemsbiology.net:8080/proteomicsToolkit/FragIonServlet.html. Accessed on Decem-ber 2007 through May 2008.

[11] ExPASy FindPept Tool. Available at: http://www.expasy.org/tools/findpept.html. December 23, 2008.

[12] Entrez Protein. Available at: http://www.ncbi.nlm.nih.gov/sites/entrez?db¼protein. December 23, 2008.

[13] E. J. Slowinski, W. L. Masterton (1997) Qualitative Analysis and theProperties of Ions in Aqueous Solutions, 2nd ed., Thompson Learn-ing, Stamford.

FIG. 3. Superimposed HPLC-MS Chromatograms (TIC and EICs) of Chicken Egg White Lysozyme Digested with Trypsin.Arrows identify the location of the trypsin digest chromatographic trace (TIC) in the regions where unique peptide peaks are identified.The chromatograms of unique peptides (EICs) are superimposed atop the peptide digest. Individual peptide peaks were obtainedfrom the tryptic digest chromatogram using the extracted ion chromatogram software tool. The EICs of the peptides are labeled tothe right of each peak as follows: i) peak containing C(Cam)NDGR with parent ion (MþH)1þ of 621 amu, ii) peak containing tryptic pep-tide TPGSR with parent ion (MþH)1þ of 517 amu, and iii) peak containing WWC(Cam)NDGR with parent ion (MþH)1þ of 993 amu.

246 BAMBED, Vol. 38, No. 4, pp. 242–246, 2010