Effect of Metals in Biomimetic Dimetal Complexes ... - ACS Publications

Subscriber access provided by NEW YORK UNIV

Article

Effect of Metals in Biomimetic Dimetal Complexes on Affinity and Gas-phase Protection of Phosphate Esters Simon Svane, Thomas J.D. Jørgensen, Christine Joy McKenzie, and Frank Kjeldsen Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b00257 • Publication Date (Web): 12 Jun 2015 Downloaded from http://pubs.acs.org on June 14, 2015

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 free 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 accessible to all readers and 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.

Analytical Chemistry 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 11

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

Effect of Metals in Biomimetic Dimetal Complexes on Affinity and Gas-phase Protection of Phosphate Esters Simon Svane[a],[b], Thomas J. D. Jørgensen[a], Christine J. McKenzie[b]* and Frank Kjeldsen[a]* [a] Department of Biochemistry and Molecular Biology, University of Southern Denmark, 5230, Odense M, Denmark [b] Department of Physics, Chemistry and Pharmacy, University of Southern Denmark, 5230, Odense M, Denmark

ABSTRACT: While the biomimetic dimetal complex [LGa2(OH)2(H2O)2]3+ (L=2,6-bis((N,N’-bis(2-picolyl)amino)methyl)-4tertbutylphenolate) provide efficient protection against phosphate loss in phosphopeptides upon collision-induced dissociation tandem mass spectrometry (CID MS/MS) the underlying mechanism remains unknown. Here, we explored the mechanism in detail and investigated the selective binding to phosphate groups in solution. Dimetal complexes containing combinations of Ga3+, In3+, Fe3+, Co3+, Zn2+, Cu2+, and V2+ were reacted with HPO42-, phosphoserine, and a phosphopeptide (FQpSEEQQQTEDELQDK, abbreviated “βcas”) and studied with isothermal titration calorimetry (ITC), CID MS/MS, and density functional theory (DFT). Ka for HPO42- binding scaled with the metal charge and was 35-fold larger for [LGa2(OH)2(H2O)2]3+ (3.08±0.31x106 M-1) than for [LZn2(HCOO)2]+. CID MS/MS of [LGa2(βcas)]n+ revealed protection against phosphate detachment (3% relative signal intensity) are listed below in non-bold type. Ligands OH-, O2-, HCOO- and OAc- are either present in the crystals of the dimetal complexes or derived from solvent. RO- = L-, Ra = C3H6NO2 (pSer) and Rb = C2H6NO (Ser). 3+

2+

Heterometallic, M , M complexes: The acetate-bridged complexes [LFeZn(CH3CO2)2]2+, [LGaCu(CH3CO2)2]2+ and [LGaZn(CH3CO2)2]2+ demonstrated large differences in pSer selectivity. At a 1:1 ratio of pSer:Ser the selectivity decreased in the order of [LFeZn(CH3CO2)2]2+ ≈ [LGaCu(CH3CO2)2]2+ > [LGaZn(CH3CO2)2]2+ with [LFeZn(CH3CO2)2]2+ binding ~60% more pSer than [LGaZn(CH3CO2)2]2+. If the pSer:Ser ratio was changed to 1:50, [LFeZn(CH3CO2)2]2+ was still reasonably selective toward pSer, binding close to 70%. [LGaCu(CH3CO2)2]2+ and [LGaZn(CH3CO2)2]2+ bound 67% and 25% pSer, respectively. The stoichiometries of the ions formed by the reaction of pSer/Ser with [LFeZn(CH3CO2)2]2+ and [LGaCu(CH3CO2)2]2+ are listed in Figure 4. Homometallic, M2+, M2+ complexes: [LZn2(HCOO)2]+, [LCu2(OCH3)]2+, and [L(VO)2(H2O)2]3+ demonstrated large differences in pSer affinity. All product ions containing either pSer or Ser were of the type [LM2(pSer-H)]2+ or [LM2(Ser-H)]2+. Both [LZn2(HCOO)2]+ and [LCu2(OCH3)]2+ appeared to form very weak complexes with pSer/Ser. This

resulted in an overall low selectivity consistent with the observation of ions with different auxiliary ligands in the mass spectra (Figure 2). The vanadyl complex [L(VO)2(H2O)2]3+ selectively bound pSer at 1:1 and 1:10 ratios of pSer:Ser (95% of ions contain pSer). At 1:50 pSer:Ser the selectivity was, however, lost as 33% of the formed ions contained Ser. Association constants, Ka. Association constants (Ka) for the binding of the phosphate moiety to different dimetal complexes were determined using ITC. Since the dilution heat produced when titrating pSer into 10 mM HEPES buffer (pH 7.1) was greater than the heat of binding to the metal complexes, Na2HPO4 was chosen as substrate instead. Incidentally this made it possible to compare the Ka of our dimetal complexes to two previously published related dizinc complexes based on the ligands (2,6-bis((N,N’-bis(2picolyl)amino)methyl)-4-methylphenolate) (bpmp-)37 and (1,8-bis((N,N’-bis(2-picolyl)amino)methyl)-anthracene) (here denoted bpanth)44 The Ka for these complexes, as well as those obtained in this study, are given in Table 1.

4 ACS Paragon Plus Environment

Page 5 of 11

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. Ka, for reaction between dimetal complexes and Na2HPO4 by ITC at 30°C in 10 mM HEPES buffer, pH 7.1. Ka[106 M-1]

Complex

Technique

Reference

[LGa2(OH)2(H2O)2](ClO4)3

3.08±0.31

ITC

28

[LIn2(CH3CO2)2](ClO4)3

1.22±0.27

ITC

28

[LFeZn(CH3CO2)2](ClO4)2

0.14±0.04

ITC

This work

[L(VO)2(H2O)2](ClO4)2[a]

-

ITC

This work

[LGaCu(CH3CO2)2](ClO4)2

0.14±0.02

ITC

This work

[LGaZn(CH3CO2)2](ClO4)2[a]

-

ITC

This work

[LZn2(HCOO)2](ClO4)

0.09±0.01

ITC

28

[LCu2(OCH3)](ClO4)

-

ITC

This work

[(bpmp)Zn2]3+,10 mM HEPES[b]

0.11±0.01

ITC

37

[(bpanth)Zn2](NO3)4 , 10 mM HEPES[c]

0.42

Fluorescence

44

[a]

[a] Solubility of the compound was too low to yield measureable binding heats. [b] Hbpmp = (2,6-bis((N,N’-bis(2-picolyl)amino)methyl)-4methylphenol); no counteranion reported. [c] bpanth = (1,8-bis((N,N’-bis(2-picolyl)amino)methyl)-anthracene).

As expected, the Ka were found to be largest for the complexes with two 3+ metal ions (digallium and diindium, ~106 M-1) and generally decreased with decreasing charge states of the metals. Of the complexes with two M2+ metal ions, only [LZn2(HCOO)2]+ produced measurable heat of binding when titrated with Na2HPO4. Ka for [LZn2(HCOO)2]+ was found to be 8.60x104 M-1 or approximately 35 times lower than that of [LGa2(OH)2(H2O)2]3+. Ka for two previously published bis(2picolyl)amine-based dizinc complexes, which have been tested as receptors for phosphate under similar conditions, are listed at the end of table 1. [(bpmp)Zn2]3+ complexes were found to have a slightly higher Ka than we report here for [LZn2(HCOO)2]+. Such a relatively small difference is likely caused by small variations in pH or concentrations. The second dizinc complex [L1Zn2]4+ was measured by fluorescence spectroscopy to have a Ka of 4.2x105 M-1. The heterometallic M3+, M2+ complexes [LFeZn(CH3CO2)2]2+ and [LGaCu(CH3CO2)2]2+ showed very similar Ka at ~105 M1 (Table 1), while [LGaZn(CH3CO2)2]2+ yielded almost no signal due to low solubility in water. These results showed that the solution Ka correlates well with the selectivity experiments (Figure 3) in the sense that the complexes with the largest Ka were found also by MS to be the most selective with respect to phosphate binding. Binding isotherms are presented in the supporting information, Figure S2-S3. Phosphate group stabilization. The dimetallic precursor complexes were mixed with the βcas phosphopeptide (FQpSEEQQQTEDELQDK, MW 2060 Da, abbreviated βcas) in 1.2:1 ratio in H2O:MeCN 1:1 and left to stand for 3 h, 25°C. The most abundant mass spectrometric product signals were assigned to [LMaMb(βcas)] n+-type ions (n = 3 or 4, the protonation state of the peptide varies). To study the effect of phosphate ester protection provided by DIMPES, [LMaMb(βcas)] n+ ions were fragmented with CID. The amount of phosphate lost from each [LMaMb(βcas)] n+ was determined as the ratio of the summed abundance of all phosphate loss signals (neutral loss from the precursor ions, from fragment ions, and the metal complex bound phosphate), divided by the summed ion abundance of the most intense fragment ions (counted up to 200). In the case of βcas phosphopeptide, the predominant

loss was neutral H3PO4, whereas most of the metal complex bound peptides predominantly lost phosphate as [LMaMb(HnPO4)]2+ or [LMaMb(HnPO4)]+ (n = 0, 1 or 2) (Table 2).

Table 2. Percentage loss of H3PO4/[LMaMb(HnPO4)]+/2+ in CID MS/MS. Precursor complex/peptide

Selected ion

Phosphate loss [%]

βcas[a]

[βcas+2H]2+

40

[a]

[βcas+3H]3+

22

[LGa2(OH)2(OH2)2](ClO4)3

[LGa2(βcas-2H)]3+

1.0

[LGa2(OH)2(OH2)2](ClO4)3

[LGa2(βcas-H)]4+

2.9

βcas

dimetal

3+

71

[LGaZn(CH3CO2)2](ClO4)2

[LGaZn(βcas-H)]

[LGaZn(CH3CO2)2](ClO4)2

[LGaZn(βcas)]4+

42


[LFeZn(βcas-H)]3+

66


[LFeZn(βcas)]4+

65

2+

[LZn2(HCOO)2](ClO4)

[LZn2(βcas-H)]

51

[LZn2(HCOO)2](ClO4)

[LZn2(βcas)]3+

51

[LCu2(OCH3)](ClO4)2

[LCu2(βcas-H)]2+

65

[LCu2(OCH3)](ClO4)2

[LCu2(βcas)]3+

60

[L(VO)2(H2O)2](ClO4)3

[L(VO)2(βcas-H)]2+

61

[L(VO)2(H2O)2](ClO4)3

[L(VO)2(βcas)]3+

69


[LGaCu(βcas-H)]3+

66


[LGaCu(βcas)]4+

54

3+


[LIn2(βcas-2H)]


[LIn2(βcas-H)]4+

43 53

[LInCu(CH3CO2)2](ClO4)3

[LInCu(βcas-H)]3+

69

[LInCu(CH3CO2)2](ClO4)3

[LInCu(βcas)]4+

57

[a] βcas = FQpSEEQQQTEDELQDK

For βcas the loss of H3PO4 from precursor and fragment ions in CID corresponded to approximately 40% and 22% of the ion intensity for 2+ and 3+ precursor ions, respectively.



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

Interestingly, CID of βcas bound to {LGa2}5+ resulted in phosphate losses of only 1% and 3% from 3+ and 4+ precursor ions, respectively. Phosphate was lost from [LGa2(βcas-2H)]3+ and [LGa2(βcas-H)]4+ exclusively as [LGa2(PO4)]2+, m/z 402.0631 confirming the digallium metal complex to be coordinated to the phosphate group (Supporting information, Figure S1). Notably, metal complexes containing any other combination of metal ions than two Ga3+, resulted in an increased phosphate loss (from 42% for [LGaZn(βcas)]4+ up to 71% for [LGaZn(βcasH)]3+). The fact that the diindium complex resulted in increased phosphate loss was surprising since it has the same net charge as the digallium complex.

Page 6 of 11

Fragmentation channels for phosphate losses in DIMPES. As illustrated in Figure 5, the distribution of various forms of phosphate losses obtained by CID activation of [LMaMb(βcas)]3+ can be divided into two fragmentation channels (A and B). Channel A resulted in loss of mass equivalent to both phosphate and metal complex. It is assumed that the metal complex was bound to the phosphate group and upon CID activation the phosphate ester bond was broken. This interpretation is supported by two observations; 1) the appearance of ions of the type [LMaMb(HnPO4)]2+, n = 0, 1, or 2 depending on the charge of the metal complex, and 2) the absence of ions in the CID spectra belonging to metal complexes without phosphate. Channel A accounted for more than 90% of the loss of phosphate from most of the [LMaMb(βcas)]3+ irrespective of the species of Ma and Mb.

Figure 5. Branching ratios for loss of phosphate containing species from different [LMaMb(βcas)]3+ ions in CID MS/MS. In A the dimetal complex remains coordinated to the phosphate on detachment from the peptide. In B a rearrangement is proposed to take place and the metal complex remains coordinated to the peptide while H3PO4 is lost. The ratios of the phosphate loss between channels A and B are given for different combinations of Ma,Mb. For complexes that form phosphate loss ions with several protonation states the dominant ions have been highlighted in = βcas. green. The percent intensity of the different protonation states is given in parenthesis.


Page 7 of 11

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


In fragmentation channel B, a 98 Da loss corresponding to H3PO4 was observed. Both accurate mass and isotopic distribution of the remaining ions suggests that the metal complex was retained on the peptide while H3PO4 was lost. For [LFeZn(βcas-H)]3+, [LGaZn(βcas-H)]3+, 3+ 3+ [LGaCu(βcas-H)] , [L(VO)2(βcas)] , [LCu2(βcas)]3+, [LZn2(βcas)]3+ and [LGa2(βcas-2H)]3+ this pathway accounted for less than 10% of the total phosphate loss based on intensity and may be explained by unspecific binding. Phosphopeptide ions with {LIn2}5+ were dominated by fragmentation through channel B and produced 100% H3PO4 loss. A possible explanation is binding of this metal complex to other competing functional groups in the phosphopeptide such as carboxylate groups. To investigate this, βcas was dephosphorylated prior to the reaction with the diindium complex. Under the same conditions as reaction with native βcas approximately 31% of the mass spectral abundance could be assigned to [LIn2(dephosβcas)] n+. It was also observed that at higher amounts of complex (~5 eq.) almost all the dephosphorylated peptide had reacted. However, the peptide was never found to bind more than one diindium complex even though the peptide has seven carboxylic acids. In another experiment, carboxylic acids of the βcas phosphopeptide were methylated prior to addition of [LIn2(CH3CO2)2]3+ (Figure 6).

Figure 6. CID MS/MS spectra of a) [LIn2(βcas-2H)]3+ m/z 953.31 (unlabeled peaks belong to unidentified fragment ions) and b) [LIn2(methβcas-H)]4+ m/z 740.27. Red peaks denote phosphate loss.

The mass spectrum of the methylated βcas confirmed phosphopeptide with [LIn2(CH3CO2)2]3+ reaction between the metal complex and the phosphopeptide by the appearance of an ion which could be assigned to [LIn2(methβcas-H)]4+ (m/z 740.2651,

Effect of Metals in Biomimetic Dimetal Complexes ... - ACS Publications

Recommend Documents