Jian Cao, John E. Burke and Edward A. Dennis*

Supplemental Material

Using Hydrogen-Deuterium Exchange Mass Spectrometry to Define the Specific Interactions of the Phospholipase A2 Superfamily with Lipid Substrates, Inhibitors and Membranes

Jian Cao, John E. Burke and Edward A. Dennis*

Department of Chemistry and Biochemistry and Department of Pharmacology, University of California, San Diego, La Jolla, California 92093-0601

*Corresponding author. EAD: Phone, 858-534-3055; FAX, 858-534-7390;

E-mail,

Experimental Approach for DXMS of Phospholipase A2s - A critical step in the DXMS experiment is optimizing the protease digestion conditions, which provide the best protein peptide coverage map. This map determines the resolution and coverage of the H/D exchange experiments. The process can be optimized by varying the denaturant concentration, HPLC flow rate and pepsin incubation times. For disulfide bonded proteins like the sPLA2s, reducing agents such as tris-(2-carboxyethyl) phosphine (TCEP) are added to the quench solution to break the disulfide bonds for further digestion (1). The detergent concentration used in protein buffers should be kept as low as possible, since most detergents, especially ionic ones, can suppress mass spectrometry signals and interfere with peptide recognition (1). Lipid vesicles have been effectively used in PLA2/lipid membrane interaction studies to mimic lipid bilayers (1-3). A higher lipid to protein molar ratio (60:1 or higher) is usually employed to ensure that all enzyme molecules are bound. Due to the large amount of lipids present, a trap column is placed before the HPLC C8 column to remove some of these phospholipids. Recently in a nanodisc DXMS study, ZrO2 beads were also shown to be effective at removing phospholipids (4). Selection of the proper acyl chain phospholipids can avoid hydrolysis of a large amount of phospholipids by PLA2. A null substrate is usually the best choice, but for enzymes without any substrate specificity, phospholipids with less than a 10% hydrolysis rate under the experimental conditions employed were found to show similar protection in the membrane binding region in the presence of the inhibitor MAFP (2,5). For a typical PLA2/membrane interaction experiment, two sets of parallel DXMS experiments are carried out; one is on the pure PLA2 enzyme alone; the other is on the PLA2 enzyme in the presence of phospholipid membranes (Figure 1B). The use of soluble phospholipid substrate mimics can also be used to determine conformational changes induced by substrate binding of a single phospholipid versus bulk lipid substrate (2). The PLA2 regions that interact with membranes or have an allosteric shift upon membrane binding will show a difference in deuterium exchange rates compared with the PLA2 enzyme alone. It is important to note that H/D changes induced upon membrane binding can be caused by a variety of factors (oligomerization, allosteric conformational changes distant from the membrane binding site, as well as direct membrane interactions). These H/D exchange differences must be interpreted in the context of previous biophysical and structural experiments.

Figure S1 shows typical results in a DXMS experiments. The peptic digestion map of GIA PLA2 is shown in Figure S1A which shows the peptides digested by pepsin using an online column and identified by mass spectrometry. As we discussed above, different proteases, denaturation reagents, flow rate etc. may help to improve the peptide coverage map. The on-exchange data can be presented as a deuterium heat map (Figure S1B) for a whole protein and a deuterium incorporation time course for specific peptides (Figure S1C). One can quickly get an idea about the protein deuterium exchange rate at different sites from the heat map and identify the regions demonstrating different H/D exchange between different protein states (or protein with different ligands). Deuterium incorporation time courses are useful for examining the differences in individual peptides throughout the H/D exchange timecourse. .

Mechanism of Amide Hydrogen Deuterium Exchange - The rate of exchange of any given amide hydrogen in a folded protein is given by the following equation

kex = kf +ku (1)

where kex represents the determined experimental rate , and kf and ku are the individual contributions to the rate from the folded and unfolded forms of the protein, respectively. The rate of exchange of an amide hydrogen in a fully folded protein that does not undergo an unfolding reaction is given by the equation

kf = βki (2)

where ki represents the intrinsic rate constant in an unfolded peptide and β represents the protection factor, which ranges in value from 0 to 1 depending on the hydrogen’s involvement in secondary structures (intramolecular hydrogen bonding) and on solvent accessibility. However, the experimental rate constant (kex) also contains a potential contribution from exchange due to the unfolding of the protein given by ku. The equations governing this reaction are shown below

(3)

where k1 and k-1 are the rate constants for unfolding and folding, respectively. FH represents the folded protein at that amide hydrogen and UH represents the unfolded state around that hydrogen. When ki > k-1, the equation for ku becomes

ku = k1 (4)

This situation is known as EX1 kinetics and very few proteins undergo global EX1 kinetics under physiological conditions. EX1 kinetics can be thought of as a cooperative unfolding process where amides in the unfolded region exchange before refolding (k-1). In contrast to EX1 kinetics, EX2 kinetics is predominant under physiological conditions where k-1 > ki and ku becomes

ku = (k1/k-1)*ki = Kunf *ki (5)

EX2 kinetics can be thought of as a case where there are few large cooperative unfolding reactions and exchange with solvent occurs through micro-fluctuations in secondary structures and solvent accessibility. Proteins can be forced into EX1 kinetics by use of denaturants (6), but certain regions of proteins may undergo both EX1 and EX2 kinetics depending on lid or hinge movements common in some proteins. EX1 or EX2 kinetics can be easily examined with mass spectrometry based isotopic profiles (7,8).

Theoretical Approach and Considerations of DXMS - A generalized scheme for a deuterium exchange mass spectrometry experiment is outlined in Figure 1A. The protein of interest is incubated in deuterated water for a certain period of time, generally ranging from seconds to hours, a process known as continuous labeling. Disordered and unstructured regions exchange hydrogen atoms very rapidly while rigid, solvent-protected regions may not exchange for days following deuterium exposure. The structural information gained from these experiments is limited to the spatial localization of this exchange rate. There are currently two different methods for spatially locating the amide exchange rates. The first approach is the classical bottoms-up approach where the protein of interest is labeled and then placed in a quench solution (as shown in Figure 1A) at pH = 2.5 and 0o C along with a denaturant guanidine hydrochloride (GuHCl)), which minimizes amide hydrogen exchange, and partially denatures the protein to allow for enzymatic digestion. The protein is then treated with an acid functional protease that cleaves the protein into discrete peptide fragments. Proteases used for deuterium exchange experiments must be active under low pH conditions with minimum disturbances by denaturants and reducing agents. The most popular protease used is pepsin. Other proteases or combinations of two or more proteases have been reported and they sometimes have better coverage than pepsin for specific proteins (9-12). The digested peptides are then rapidly separated using lipid chromatography (LC) and injected into the mass spectrometer. By creating enough overlapping peptide fragments, exchange rates can theoretically be isolated for single amide hydrogens.

Techniques have been developed to prevent back exchange during digestion and separation. Significant work has been done to improve separation, including fast ultra- performance liquid chromatography (UPLC) gradients (13), ion mobility separation in the mass spectrometer (14), and online HPLC protease digestion (15). The amount of back exchange can also be corrected for if both an on-exchange control and back-exchange control are prepared. The on-exchange control is prepared by adding D2O to a sample after the addition of the quench buffer to control for deuterium addition during the processing steps. The back-exchange control is prepared by generating a fully deuterated sample, with all amides containing deuterium, which receives the same treatment as the sample studied. This controls for deuterium that back-exchange during the processing steps. Equation 1 governs quantification of deuterons while controlling for both the back-exchange and on-exchange during the HPLC separation and digestion

(6)

where mi is the centroid mass of the peptide exposed to s = sample conditions, 0% = on-exchange quench control, 100% = fully deuterated, and N = maximum deuterium level in the peptide (16).

New High Resolution Approach to DXMS - In addition to the classic proteolytic digestion/LC-MS DXMS approach, resolution of the deuterium label can also be obtained by a “top-down” approach in which the fragmentation of the intact protein taking place in the gas phase either bypasses or complements the proteolytic process. In this “top down” approach, different peptide ions are fragmented depending on the dissociation method used in the mass spectrometer. Initially reported experiments used collision-induced dissociation (CID) to produce b and y ion pairs (Figure S1). However, the collisional heating can cause hydrogen and deuterium scrambling, which would invalidate this technique as a method of locating the deuterium (17,18). Therefore in our experimental protocols, CID was only used to identify the peptides, but not for deuterium quantification. Electron transfer dissociation (ETD) and electron capture dissociation (ECD) methods transfer much less energy to the gas-phase protein, generating c and z ion pairs (Figure S1), which avoid the scrambling seen in the CID method (19-26). The major advantage of this approach is that the removal of the digestion and peptide separation steps allow for much lower back-exchange rates. ETD and ECD fragmentation methods were also shown to reach single amide resolution by using traditional “bottom-up” workflows (27). The most exciting prospect for spatial resolution is a method that combines both the “bottom up” and “top down” approaches, allowing for quick digestion separation in the LC, followed by further fragmentation within the mass spectrometer itself. This will allow for the study of much larger proteins at a much higher level of detail than currently possible.

Several user friendly automated data processing programs are available that are able to accelerate the data processing time. This type of software includes DXMS Explore and HDExaminer (Sierra Analytics Inc.), DynamX Data AnalysisHDX software (Waters Inc.), HX-Express (28), Hydra (29), HDXFinder (30) and HD Desktop (31).

Figure S1. A typical data presentation for a DXMS experiment. A. Protein digestion map. B. Deuterium exchange heat map. C. Deuterium exchange time course data. Figures are adapted from (1).

A

B

C

Figure S2. Basis of hydrogen-deuterium exchange. Possible fragmentation patterns for a peptide segment composed of 5 amino acids. The peptide can fragment into a-x pairs (shown in blue), b-y pairs (shown in red), and c-z pairs (shown in green).

References:

1. Burke, J. E., Karbarz, M. J., Deems, R. A., Li, S., Woods, V. L., Jr., and Dennis, E. A. (2008) Biochemistry 47, 6451-6459

2. Burke, J. E., Hsu, Y. H., Deems, R. A., Li, S., Woods, V. L., Jr., and Dennis, E. A. (2008) J Biol Chem 283, 31227-31236

3. Cao, J., Hsu, Y. H., Li, S., Woods, V. L., and Dennis, E. A. (2011) Biochemistry 50, 5314-5321

4. Hebling, C. M., Morgan, C. R., Stafford, D. W., Jorgenson, J. W., Rand, K. D., and Engen, J. R. (2010) Analytical chemistry 82, 5415-5419

5. Hsu, Y. H., Burke, J. E., Li, S., Woods, V. L., Jr., and Dennis, E. A. (2009) J Biol Chem 284, 23652-23661

6. Deng, Y., and Smith, D. L. (1998) Biochemistry 37, 6256-6262

7. Miranker, A., Robinson, C. V., Radford, S. E., Aplin, R. T., and Dobson, C. M. (1993) Science 262, 896-900

8. Weis, D. D., Wales, T. E., Engen, J. R., Hotchko, M., and Ten Eyck, L. F. (2006) J Am Soc Mass Spectrom 17, 1498-1509

9. Rey, M., Man, P., Brandolin, G., Forest, E., and Pelosi, L. (2009) Rapid Commun Mass Spectrom 23, 3431-3438

10. Marcoux, J., Thierry, E., Vives, C., Signor, L., Fieschi, F., and Forest, E. (2009) J Am Soc Mass Spectrom

11. Cravello, L., Lascoux, D., and Forest, E. (2003) Rapid Commun Mass Spectrom 17, 2387-2393

12. Zhang, H. M., Kazazic, S., Schaub, T. M., Tipton, J. D., Emmett, M. R., and Marshall, A. G. (2008) Anal Chem 80, 9034-9041

13. Wales, T. E., Fadgen, K. E., Gerhardt, G. C., and Engen, J. R. (2008) Anal Chem 80, 6815-6820

14. Iacob, R. E., Murphy, J. P., 3rd, and Engen, J. R. (2008) Rapid Commun Mass Spectrom 22, 2898-2904

15. Brudler, R., Gessner, C. R., Li, S., Tyndall, S., Getzoff, E. D., and Woods, V. L., Jr. (2006) J.Mol.Biol. 363, 148-160

16. Zhang, Z., and Smith, D. L. (1993) Protein Sci 2, 522-531

17. Demmers, J. A., Rijkers, D. T., Haverkamp, J., Killian, J. A., and Heck, A. J. (2002) J Am Chem Soc 124, 11191-11198

18. Jorgensen, T. J., Gardsvoll, H., Ploug, M., and Roepstorff, P. (2005) J Am Chem Soc 127, 2785-2793

19. Hoerner, J. K., Xiao, H., Dobo, A., and Kaltashov, I. A. (2004) J Am Chem Soc 126, 7709-7717

20. Kaltashov, I. A., and Eyles, S. J. (2002) Mass Spectrom Rev 21, 37-71

21. Kaltashov, I. A., and Eyles, S. J. (2002) J Mass Spectrom 37, 557-565

22. Zehl, M., Rand, K. D., Jensen, O. N., and Jorgensen, T. J. (2008) J Am Chem Soc 130, 17453-17459

23. Abzalimov, R. R., Kaplan, D. A., Easterling, M. L., and Kaltashov, I. A. (2009) J Am Soc Mass Spectrom 20, 1514-1517

24. Pan, J., Han, J., Borchers, C. H., and Konermann, L. (2009) J Am Chem Soc 131, 12801-12808

25. Rand, K. D., Zehl, M., Jensen, O. N., and Jorgensen, T. J. (2009) Anal Chem 81, 5577-5584

26. Rand, K. D., Pringle, S. D., Morris, M., Engen, J. R., and Brown, J. M. (2011) Journal of the American Society for Mass Spectrometry 22, 1784-1793

27. Landgraf, R. R., Chalmers, M. J., and Griffin, P. R. (2012) Journal of the American Society for Mass Spectrometry 23, 301-309

28. Weis, D. D., Engen, J. R., and Kass, I. J. (2006) Journal of the American Society for Mass Spectrometry 17, 1700-1703

29. Slysz, G. W., Baker, C. A., Bozsa, B. M., Dang, A., Percy, A. J., Bennett, M., and Schriemer, D. C. (2009) BMC bioinformatics 10, 162

30. Miller, D. E., Prasannan, C. B., Villar, M. T., Fenton, A. W., and Artigues, A. (2012) Journal of the American Society for Mass Spectrometry 23, 425-429

31. Pascal, B. D., Chalmers, M. J., Busby, S. A., and Griffin, P. R. (2009) Journal of the American Society for Mass Spectrometry 20, 601-610