Subscriber access provided by the University of Exeter
Technical Note
Characterization of fluorescently labeled protein with ESI-MS and fluorescence spectroscopy: how random is random labeling? Qiaoqiao Ruan, Cheng Zhao, Carol S Ramsay, and Sergey Y. Tetin Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b01748 • Publication Date (Web): 20 Jul 2018 Downloaded from http://pubs.acs.org on August 2, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
Characterization of Fluorescently Labeled Protein with ESI-MS and Fluorescence Spectroscopy: How Random is Random Labeling? Qiaoqiao Ruan, Cheng Zhao, Carol S. Ramsay, and Sergey Y. Tetin* Applied Research and Technology, Abbott Diagnostics Division, Abbott Laboratories, Abbott Park, IL 60064 *To whom correspondence should be addressed
[email protected] ABSTRACT: Solvent exposed lysine residues are abundantly present in many proteins. Their highly reactive ε-amino groups serve as universal targets for coupling with active esters of various extrinsic labels including a vast arsenal of fluorescent probes. Here, we describe fluorescent properties and preferential localization of two frequently used fluorescent labels, AlexaFluor488 (AF488) and Cy3, on the surface of a small highly soluble serum protein Neutrophil Gelatinase-Associated Lipocalin (NGAL), which serves as a diagnostic marker for acute kidney failure. Using a standard protocol for labeling with either AF488-SDP or AF488-NHS, we achieved >95% labeling efficiency of the protein as determined by UV-VIS absorption and ESI-MS. However, fluorescence intensity of the labeled protein was less than 10% of the expected value. To understand the unusually high quenching of the probe, we identified the sites of AF488 attachments by means of LC-MS/MS combined with trypsin digestion. Surprisingly, we found that the AF488 label is not randomly distributed among accessible lysines, but predominantly bound to the residues K125, K126 or K135, which are located in the NGAL calyx and are likely quenched by neighboring tryptophans and tyrosines. In contrast, when NGAL was labeled with Cy3, the probe’s fluorescence was almost fully retained. The LC-MS/MS data indicated that Cy3 was predominately bound to another lysine, K31, on the protein surface on the opposite side of the calyx. Our findings suggest that a combination of the inherent properties of the label and the specifics of the protein microenvironment may selectively lead probes to specific lysine residues and thus challenge the common view that protein labeling is a random process.
Keywords: fluorescence, probes, protein labeling, lysine, mass-spectrometry
INTRODUCTION Using fluorescent labels in the form of their Nhydroxysuccinimide (NHS) or sulfodicholorphenol (SDP) active esters is common for performing covalent protein labeling by targeting accessible ε-amino groups of lysine residues 1,2 . It is well-known that the reactivity of lysine side-chains varies depending on their accessibility and pKa in the local microenvironment 3-5. These differences in reactivity can be conveniently used for covalent protein modification when performing structural analysis 6. It is also well-known that lysine modifications affect protein charge and result in heterogeneous protein populations 7-9. However, from the probe perspective, it is often generalized that, due to the abundant number of exposed lysine residues, labeling results in a random distribution of the probe across all lysines, despite reported contradictions 10. Such an assumption neglects that, at low labeling ratios, even weak protein-probe interactions and/or subtle differences in accessibility and chemical reactivity of the side chains may lead to the preferential labeling of specific lysines. The latter is critical to recognize when performing ligand binding, structure-function or other biophysical characterizations. Low incorporation ratios are frequently implemented to prevent self-quenching of fluorophores and protein aggregation
11
. Under these conditions, a uniform distribution of the label cannot be assumed a priori. We learned this when labeling human neutrophil gelatinase-associated lipocalin (NGAL) with AlexaFluor488 (AF488) and Cy3 dyes using their NHS and SDP active esters at pH 7.2. NGAL is a 22 kDa globular protein that serves as a biomarker for acute kidney injury. The fluorescence intensity of the AF488-labeled NGAL dropped 90% in comparison with an equimolar solution of AF488. This effect was observed regardless of which active form of the dye, AF488-SDP or AF488-NHS, was used for labeling. In contrast, when NGAL was labeled with Cy3-NHS, the fluorescence of the probe was almost unaffected. To understand this effect, we employed mass spectrometry to localize the labeling sites 12. In both labeled NGAL preparations, the number of fluorophores were characterized by intact molecular weight profiling and the specific locations were identified using LC-MS/MS after trypsin digestion13,14. Apparently, AF488 was predominantly attached to the residues K125, K126 or K135 in the NGAL calyx, while Cy3 was found on the opposite side of the calyx on the residue K31. Therefore, we suggest that a combination of the side chain orientation/reactivity and inherent properties of the label direct AF488 and Cy3 to different lysines, thus resulting in biased labeling. We suggest that this
ACS Paragon Plus Environment
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
effect may be common for proteins and should not be overlooked. EXPERIMENTAL SECTION Protein expression and purification: NGAL was produced as previously described 15. Briefly, the sequence of human NGAL (GenBank AAD14168) was fused with a N-terminal methionine and a C-terminal 6-histidine tag and cloned into pKRR826 (a pLbased expression vector). To prevent dimer formation, NGAL Cys88 was mutated to serine. The recombinant human NGAL protein was expressed in E. coli BL21 cells and purified by nickel affinity chromatography, and finally dialyzed in PBS. Protein labeling: Two forms of dyes, sulfodicholorphenol (SDP) and N-hydroxysuccinimide (NHS) active esters, were used for protein labeling. AF488-SDP, AF488-NHS (ThermoFisher Scientific) and Cy3-NHS (GE Healthcare) were dissolved in DMSO to 10 mM. NGAL was diluted to 1.5mg/ml in PBS pH 7.2. AF488SDP, AF488-NHS or Cy3-NHS was added to NGAL at molar excess varying from 2- to 20-fold. The reaction conditions were coded as following (AF488-NGAL 2x (SDP), pH 7.2: represents NGAL reacted with 2-fold excess of AF488-SDP at pH 7.2; AF488-NGAL 2x(NHS), 7.2: represents NGAL reacted with 2fold excess of AF488-NHS at pH 7.2). All reactions were carried out overnight at 2-8ºC. Excess labels were removed by passing the samples twice through pre-packed NAP-5 columns (GE Healthcare) equilibrated with PBS.
Page 2 of 5
sent the label distribution of AF488-NGAL. The mass increment between each peak is 516Da, which matches the molecular weight of AF488, confirming that the structure of AF488 was not modified during labeling. In the AF488-NGAL 2X(SDP), pH 7.2 sample, the number of AF488 attached to NGAL ranges from 0 to 3. In the AF488-NGAL 15X(SDP), pH 10 sample, the number of the conjugated AF488 ranges from 4 to 9. Figure 1C illustrates the distribution of AF488 in NGAL prepared under different labeling conditions using AF488-SDP. The deconvoluted ESI-MS spectra of Cy3-NGAL and AF488-NGAL(NHS) are included in the supplementary section (S2). The average I.R. of the labeled NGAL was calculated from the formula: ∑(number of fluorophore x peak intensity) ⁄ ∑(peak intensity). Alternatively, the I.R. of labeled NGAL can be determined by absorption spectrum. Calculations and results of AF488-NGAL (SDP) are shown in the supplementary section (see S3). The direct correlation (r2=0.976) of the average I.R. obtained from absorption and ESI-MS for each AF488-NGAL(SDP) sample validated both approaches.
Confirmation of NGAL labeling: The labeled NGAL samples were profiled by QSTAR® ESI-MS (Hybrid Quadrupole–TOF LC-MS/MS mass spectrometer, AB Sciex, Framingham, MA) and then the multiple charged states distribution of the protein samples was deconvoluted using the Bayesian Protein Reconstruct function available in Analyst QS software from AB Sciex. Deconvoluted spectra reflect the molecular weight profile of the labeled protein. Localization of labeling sites: 5 µg of labeled NGAL samples were dried with a Vacufuge (Eppendorf, Hauppage, NY) and reconstituted in 15 µL 100mM ammonium bicarbonate and then incubated with 1/50 (w/w) Trypsin at 37oC for 18 hrs. The samples were then desalted using ZipTipC18 (Millipore Corporation) and analyzed by Q-TOF instrument coupling with a C18 column (5 µm particle size, 150 × 0.3 mm i.d., Michrom) on ESI- MS. Fluorescence measurements: The AF488 dye (control) and AF488-NGAL samples, as well as the Cy3 dye (control) and Cy3NGAL samples, were diluted to the equal AF488 and Cy3 concentrations, (A496=0.025 OD) and (A550=0.025 OD) accordingly. The fluorescence emission spectra were recorded on a Fluorolog spectrofluorimeter (Horiba Jobin Yvon, Edison, NJ.)
RESULTS AND DISCUSSION The first three-dimensional structure of NGAL was determined by NMR and published in 1999 (pdb code: 1NGL)16. In this paper, we are using the same protein construct and expression system, and therefore will follow the same residue numbering sequence. According to the structural data16 17, the NGAL molecule contains 16 lysine residues, of which 11 are exposed to solution according to its structure determined by NMR spectroscopy (See S1). AF488-SDP, AF 488-NHS and Cy3-NHS effectively reacted with NGAL. AF488-NGAL and Cy3-NGAL with various incorporation ratio (I.R.) were achieved under different reaction conditions. As expected, labeling NGAL at higher pH and higher molar excess of the probe lead to higher I.R. values. Figure 1A and 1B shows two examples of the deconvoluted ESI-MS spectra of AF488-NGAL (SDP). The corresponding peak intensities repre-
ACS Paragon Plus Environment
2
Page 3 of 5
1x106
Fluorescence Intensity (a.u.)
Cy3-NGAL 5X, pH 7.2 Cy3
B
9x105 8x105 7x105 6x105 5x105 4x105 3x105 2x105 1x105 0 550
600
650
700
750
800
Wavelength (nm)
Figure 1: A, B) Deconvoluted ESI-MS of AF488-NGAL 2X(SDP), pH 7.2 and AF488-NGAL 15X(SDP), pH 10 samples obtained using Bayesian Protein Reconstruct function available in Analyst QS software from AB Sciex. The mass of each peak is labeled on the graph followed with the number (in parentheses) of AF488 attached to the protein based on the inset table. The inset table listed the theoretical molecular weight of AF488-NGAL and corresponding number of AF488 attached to one NGAL molecule. C) Distribution of the numbers of AF488 attached to NGAL under various labeling conditions determined by ESI-MS. The emission spectra of AF488-NGAL(SDP) and free AF488 dye measured under the same conditions and equal fluorophore concentrations indicated strong quenching of AF488-NGAL (SDP), (shown in Figure 2). The total fluorescence of AF488-NGAL (SDP) is quenched ~90% in comparison with free AF488. Switching AF488-SDP to AF488-NHS in the labeling procedure did not improve fluorescence efficiency of the conjugate. The AF488NGAL (NHS) is also quenched ~90% comparing to free AF488 (data not shown). In contrast, the fluorescence intensity of Cy3NGAL is quenched only by ~15% compared to free Cy3. Labeling of NGAL with Alexa488 SDP was performed more than five times, each time resulting in a highly-quenched product (>90%). Similarly, multiple labeling experiments (n>3) with Cy3-NHS all showed strong unquenched fluorescence. Figure 2B shows the emission spectra of Cy3-NGAL and free Cy3 dye measured at equal concentrations.
1.0x106
Fluorescence Intensity (a.u.)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
A AF488-NGAL 2X pH7.2 AF488-NGAL 8X pH 7.2 AF488
8.0x105
6.0x105
4.0x105
2.0x105
0.0 500
550
600
Wavelength (nm)
650
700
Figure 2: A) Emission spectra of AF488-NGAL(SDP) and free AF488 dye measured at equal AF488 concentration under same conditions. B) Examples of the emission spectra of Cy3NGAL(NHS) and free Cy3 dye measured at equal Cy3 concentration under same conditions. The loss of AF488 fluorescence after conjugation to the protein can result from unanticipated chemical modification of the probe or from its quenching by the neighboring amino acids. The ESIMS results confirmed the structural integrity of both probes, AF488 and Cy3, in all labeled NGAL samples. Thus, strong AF488 quenching is likely caused by the local protein environment. To identify individual labeling sites, AF488-NGAL and Cy3-NGAL samples were digested by trypsin and analyzed using the LC-MS/MS method on an ESI-MS instrument equipped with a C18 column. Trypsin digestion of AF488-NGAL 2X(SDP), pH 7.2, and AF488-NGAL 2X(NHS), pH 7.2, samples was performed independently and resulted in identification of 32 and 52 peptides, respectively. These peptides are listed in the supplemental Table S4. The experimental molecular weights of all identified peptides match the theoretical values with the mass accuracy within 20 ppm. The overall peptide coverage is ~ 98% in both cases. Peptides containing AF488 labels are listed in Tables 1A and 1B. In all peptides, the difference between the experimental and theoretical M.W. is 516 Da, which is the molecular weight of AF488. The relative signal intensities were calculated based on the MS spectra as follows: The relative signal intensity (%) =signal intensity/ ∑signal intensity *100%. For comparison, using the same method, we also identified the probe attachment site in NGAL labeled with Cy3 (Table 1C). In all NGAL samples labeled at a low incorporation ratio we found that the probes are preferentially attached only to a few specific residues, which was rather unexpected. Approximately 68% of AF488 labels in the AF488-NGAL (SDP) sample were distributed between the residues K126 and K135, identified in peptides 126-131 and 132-141 (mass-spec data are shown in supplementary Figure S5). Similarly, in the AF488-NGAL (NHS) samples ~84% of AF488 labels were attached to the residues K125 or K126, identified in peptides 126-131 and 100-125. In contrast, in the Cy3-NGAL samples more than 60% of the Cy3 labels were attached to the residue K31. When targeting a high label incorporation ratio, the probe attachment sites became more randomly distributed, as would be initially expected. For instance, we labeled NGAL with 15-times molar excess of AF488 at pH10 (AF488-NGAL 15X, pH 10) and observed practically even distribution of AF488 among all surface exposed lysine residues (Fig 1C). However, such conjugates demonstrated weak fluorescence and decreased solubility which indicates that aiming for higher incorporation is not always practical.
ACS Paragon Plus Environment
3
Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Table 1 A) Lists of identified NGAL peptides with AF488 attachment from sample AF488-NGAL (SDP) 2X, pH7.2.
B) Lists of identified NGAL peptides with AF488 attachment from sample AF488-NGAL (NHS) 2X, pH7.2.
Page 4 of 5
ther the emission intensity/anisotropy nor the diffusion rate of the probes changed, which does not support potential complex formation. Perhaps, a higher steric accessibility and/or reactivity of certain side chain amine(s) acts as the dominant cause of the preferential labeling of the NGAL described in this paper. Considering the identified sites, low fluorescence efficiencies of the AF488-NGAL reagents can be explained as follows: at low I.R, the probe was mainly attached to three lysine residues (K125, K126 or K135) located in the calyx in proximity of two tryptophan (Trp32 and Trp80) and six tyrosine residues (Tyr53, Tyr57, Tyr65, Tyr101, Tyr107, Tyr133). It is well-known that tryptophans and tyrosines are effective fluorescence quenchers21. When labeling with Cy3, the probe was predominantly conjugated to the surface-exposed K31 which is located far from the aromatic side chains on the opposite side from the calyx. Therefore, the Cy3labeled NGAL preparations remain highly fluorescent. Given the relatively small size of an NGAL molecule (22 kDa), higher I.R. labeling implies high probe density, thus increasing the likelihood that the fluorophores are self-quenched. Our findings suggest that when aiming for a low probe incorporation ratio during labeling, a combination of the inherent properties of the label and the uniqueness of the protein microenvironment may direct the label to certain lysine residues. This observation opposes the common assumption that the labeling of protein lysines is generally random. A bigger problem could arise in twocolor studies, where changing the labeling site may affect proteinligand and protein-protein interactions, creating a difficult to recognize bias. The possibility of non-random labeling may be a help or hindrance for a given protein-label combination and thus should not be overlooked.
C) Lists of identified NGAL peptides with Cy3 attachment from sample Cy3-NGAL 2X, pH7.2.
Supporting Information: Additional data of deconvoluted ESI-MS spectra of labeled NGAL; calculation of incorporation ratio; list of identified peptides by ESI-MS.
AUTHOR INFORMATION NGAL is a member of the lipocalin protein family. It has an eight-strand beta-barrel, forming a cup-shaped, lipophilic pocket often referred to as the calyx16. NGAL binds small, bacteriaproduced formyl peptides or siderophores18. Goetz et al.17 found that the natural NGAL ligand is centered between the guanidinium group of Arg81 and the Nε atoms of Lys126 and Lys135. As follows from the mass spectrometry data, AF488 preferentially reacts with the lysine side chains K125, K126 or K135 located in close proximity to the calyx. In contrast, the Cy3 ester mainly reacts with the residue K31, located on the protein surface on opposite side of the calyx. In the three presented examples, preferential labeling may be caused by weak inherent interactions of NGAL with AF488 or Cy3 (structure of AF488 and Cy3 are shown in S6). According to the concept proposed by Wofsy et al. in 196219, formation of the initial non-covalent, reversible complex increases local concentration of the label, thus enhancing the rate of covalent bond formation between the protein and the probe. It was later shown by Watt & Voss that an anti-fluorescein antibody can be specifically labeled in the binding site using a highly reactive fluorescein isothiocyanate20. Labeling reactions are usually performed at micromolar reagent concentrations, at which, given at least a millimolar affinity, a detectable amount of the non-covalent complex should be formed. However, we were not able to detect such binding directly by mixing 1nM AF488 or Cy3 and 5uM NGAL. Nei-
Corresponding Author *Address: AP-20, Abbott Laboratories, 100 Abbott Park Road, Abbott Park, IL 60064-6016. Tel: 224-668-4661; Fax: 224-6686498; E-mail:
[email protected].
Author Contributions Study conceived and designed by Q. Ruan and S.Y. Tetin. Q. Ruan performed labeling and fluorescence experiments. C. Zhao and C.S. Ramsay performed mass-spec experiments. Manuscript was written by Q. Ruan and S.Y. Tetin. All authors reviewed the manuscript and approved the final version.
Notes The authors declare no competing financial interests.
ACKNOWLEDGMENT The work is funded by Abbott Laboratories. We thank our colleagues Larry Birkenmeyer for preparing NGAL protein and Patrick Macdonald and Richard Haack for critical reading the manuscript and useful suggestions.
ACS Paragon Plus Environment
4
Page 5 of 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Analytical Chemistry
For Table of Contents Only
REFERENCES (1) Lundblad, R. L.; Noyes, C. M. The modification of lysines; CRC: Boston, MA, 1984; Vol. 1, pp 127-170. (2) Wong, S. S. Reactive groups of proteins and their modifying agents; CRC: Boston, MA, 1991. (3) Fitch, C. A.; Karp, D. A.; Lee, K. K.; Stites, W. E.; Lattman, E. E.; Garcia-Moreno, E. B. Biophys. J. 2002, 82, 3289-3304. (4) Isom, D. G.; Castaneda, C. A.; Cannon, B. R.; Garcia-Moreno, B. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 5260-5265. (5) Haugland, R. P. Fluorescent labels; Humana Press: New York, 1991, pp 85-109. (6) Limpikirati, P.; Liu, T.; Vachet, R. W. Methods 2018, DOI 10.1016/j.ymeth.2018.04.002. (7) Gao, J.; Gomez, F. A.; Harter, R.; Whitesides, G. M. Proc. Natl. Acad. Sci. U. S. A. 1994, 91, 12027-12030. (8) Richards, D. P.; Stathakis, C.; Polakowski, R.; Ahmadzadeh, H.; Dovichi, N. J. Chromatogr. A. 1999, 853, 21-25. (9) Stoyanov, A. V.; Ahmadzadeh, H.; Krylov, S. N. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 2002, 780, 283-287. (10) Carilli, C. T.; Farley, R. A.; Perlman, D. M.; Cantley, L. C. J. Biol. Chem. 1982, 257, 5601-5606. (11) Vira, S.; Mekhedov, E.; Humphrey, G.; Blank, P. S. Anal. Biochem. 2010, 402, 146-150. (12) Schnaible, V.; Przybylski, M. Bioconjug. Chem. 1999, 10, 861866. (13) Aebersold, R.; Goodlett, D. R. Chem. Rev. 2001, 101, 269-295. (14) Pierce, W. M.; Cai, J. Contrib. Nephrol. 2004, 141, 40-58. (15) Olejniczak, E. T.; Ruan, Q.; Ziemann, R. N.; Birkenmeyer, L. G.; Saldana, S. C.; Tetin, S. Y. Biopolymers 2010, 93, 657-667. (16) Coles, M.; Diercks, T.; Muehlenweg, B.; Bartsch, S.; Zolzer, V.; Tschesche, H.; Kessler, H. J. Mol. Biol. 1999, 289, 139-157. (17) Goetz, D. H.; Holmes, M. A.; Borregaard, N.; Bluhm, M. E.; Raymond, K. N.; Strong, R. K. Mol. Cell 2002, 10, 1033-1043. (18) Flower, D. R. Biochem. J. 1996, 318 (Pt 1), 1-14. (19) Wofsy, L.; Metzger, H.; Singer, S. J. Biochemistry 1962, 1, 1031-1039. (20) Watt, R. M.; Voss, E. W., Jr. Immunochemistry 1978, 15, 875882. (21) Chen, H.; Ahsan, S. S.; Santiago-Berrios, M. B.; Abruna, H. D.; Webb, W. W. J. Am. Chem. Soc. 2010, 132, 7244-7245.
ACS Paragon Plus Environment
5
