Combination of Capillary Electrophoresis and Ion mobility coupled to Mass Spectrometry and Theoretical Calculations for cysteine connectivity identification in peptides bearing two intra-molecular disulfide bonds

(1)

Combination of Capillary Electrophoresis and Ion mobility

coupled to Mass Spectrometry and Theoretical

Calculations for cysteine connectivity identification in

peptides bearing two intramolecular disulfide bonds

Cédric Delvaux

(1)

_{and Philippe Massonnet}

(1)

_{, Christopher Kune}

(1)

_{, Gregory Upert}

(2)

_{, Gilles}

Mourier

(2)

_{, Jean R.N. Haler}

(1)

_{, Nicolas Gilles}

(2)

_{, Loïc Quinton}

(1)

_{, Johann Far}

(1)

_{and Edwin de}

Pauw

(1)

(1) Laboratory of Mass Spectrometry, University of Liege, Allée de la Chimie 3, B-4000 Liege, Belgium

(2) Commissariat à l’Energie Atomique, DSV/iBiTec – S/SIMOPRO, F91191 Gif-sur-Yvette, France

(2)

Context of the study : disulfide connectivity assignment

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

• Major post translational modification playing crucial roles in peptide

stabilization and protein structures

• In some cases, the native disulfide pattern is essential for biological

activities

(1)

_{or to preserve the biological activity}

(2)

• Misfolded variants can lead to reduced biological activity

(2)

_{and are}

generally degraded or recycled by enzymes to the native form

(3)

The presence of multiple disulfide bonds leads to various disulfide isomers/variants :

(1) Matsumura, M.; Signor, G.; Matthews, B. W. Nature 1989, 342 (6247), 291–293 (2) Wu, Y.; Wu, X.; Yu, J.; Zhu, X.; Zhangsun, D.; Luo, S. Molecules 2014, 19 (1), 966–979

SH

Same amino acid

sequence but different

disulfide connectivities

Potential variable

biological activities

(3)

Characterization methods for S-S bonds connectivities

in peptides and proteins: State-of-the-art

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

Method

Main Advantage

Main Drawback

X-Ray Crystallography

High structural resolution

Need for a crystal/fold

Nuclear Magnetic Resonance

High structural resolution

Large amount of sample needed

Bioinformatics

Only sequence is required

Not experimentally confirmed

Mass Spectrometry

Large number of approaches available

(MALDI-ISD, ETD, CID, IM-MS,

LC-MS/MS,…)

Complex spectral information

(4)

The use of Ion Mobility Spectrometry (IMS) for

disulfide connectivity identification

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

Ion Mobility :

Separation

in the gas phase

according to both

charge (q) and collision cross section (Ω)

K: mobility in gas phase (m2_.V-1_.s-1₎

q: charge of the ion (C) N: density number of buffer gas T: temperature (K) m: mass of buffer gas (Da) M: mass of ion (Da)

k: Boltzmann’s constant (1.38065.10-23_J.K-1₎

E

N2 N2 N2 N₂ _N₂ N2 Mobility Cell N2 N2

Courtesy of Bowers group

Projection approximation F_electric F_friction N₂ N₂ N₂ N2 C N C P K E_T A L C R R A C Q Q H A N2

Mason-Schamp

Model peptide :

27-residue synthetic peptide containing 4 cysteines

with 3 possible intramolecular disulfide pairings

(conceptual rendering) :

1 SH SH C E G W F R F T K T G L E SH SH G W R C C Y C T P G L L K L 15 20 27 C E G W C R C G W F R F TK TG L E Y C T P G L L K L

1) ModGlo

(Cys1-Cys20 / Cys15-Cys27)

C

1

-C

3

/ C

2

-C

4 C_E GWR _C C G WF R F T K T G L E Y C T P G L L K_L

2) ModBea

(Cys1-Cys15 / Cys20-Cys27)

C

1

-C

2

/ C

3

-C

4 G W C R C F R F TK TG L E Y C TP G L L K L

3) ModRib

(Cys1-Cys27 / Cys15-Cys20)

C

1

-C

4

/ C

2

-C

3

K =

3q

16N

.

2π

kT

1 2

.

m + M

mM

1 2

.

1 Ω

= constant.

m + M

mM

1 2

.

𝐪

𝛀

(5)

Published IM-MS method

(1)

_{on a synthetic model peptide}

(1) Massonnet, P.; Haler, J. R. N.; Upert, G.; Degueldre, M.; Morsa, D.; Smargiasso, N.; Mourier, G.; Gilles, N.; Quinton, L.; De Pauw, E. 2016, 27 (10), 1637–1646

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

ModGlo

(

Cys1-Cys20 / Cys15-Cys27

)

C

₁

-C

₃

/ C

₂

-C

₄

ModRib

(

Cys1-Cys27 / Cys15-Cys20

)

C

₁

-C

₄

/ C

₂

-C

₃

ModBea

(

Cys1-Cys15 / Cys20-Cys27

)

C

₁

-C

₂

/ C

₃

-C

₄

Mix of 3 model

peptides isomers at

equal concentrations

(6)

Ion Mobility Spectrometry and Capillary Electrophoresis :

mobility-based separation techniques

Ion Mobility :

Separation

in the gas phase

according to both

charge (q) and collision cross section (Ω)

Capillary Electrophoresis :

Separation

in solution

according to both charge (q)

(pH dependent) and hydrodynamic radius (R

_h

)

μe=

_6πηR

q

h

=

1 6πη

.

q

𝐑

_𝐡

E

N2 N2 N2 N₂ _N₂ N2 Mobility Cell N2 N2

Courtesy of Bowers group

F_electric F_friction N₂ N₂ N₂ N2 C N C P K E_T A L C R R A C Q Q H A N2

+

HV supply

-1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

_F

electric

= F

friction

→ v

stat

=

mobility constant

. E

K =

3q

16N

.

2π

kT

1 2

.

m + M

mM

1 2

.

1 Ω

= constant.

m + M

mM

1 2

.

𝐪

𝛀

(7)

BGE = 80mM formic acid in 20% isopropanol

30µm x 150µm x 90cm BFS @+30kV

CZE method development on a synthetic model peptide

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

RMT = MT peptide / MT GluFib

15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Time (min)

0 25 50 75 100

R

elat

iv

e

Abu

nd

an

ce

17.98 18.98 18.45 26.01 RMT = 0,72972 ModRib

Glufib = reference migration time

RMT = 0,70934 ModBea RMT = 0,69127 ModGlo

ModGlo : C

₁

-C

₃

/C

₂

-C

₄

ModBea : C

₁

-C

₂

/C

₃

-C

₄

ModRib : C

₁

-C

₄

/C

₂

-C

₃

Determination of RMT (separate disulfide isomers)

σ

RMT = relative migration time = standard deviation

n = number of replicates Peptide RMT (n=6) σ (n=6) %σ (n=6) ModGlo 0,69741 0,00158 0,23% ModBea 0,70650 0,00088 0,12% ModRib 0,72882 0,00110 0,15% 0.69 0.70 0.71 0.72 0.73 0.74

0.5

ModGlo

1.5

ModBea

2.5

ModRib

3.5

RMT to

G

lufib

Determination of RMT (mix of the disulfide isomers)

Peptide RMT (n=6) σ (n=6) %σ (n=6)

ModGlo 0,69388 0,00217 0,31%

ModBea 0,71225 0,00228 0,32%

ModRib 0,73169 0,00185 0,25%

RMT = relative migration time σ = standard deviation n = number of replicates 0.69 0.70 0.71 0.72 0.73 0.74

0.5

ModGlo

1.5

ModBea

2.5

ModRib

3.5

RMT to

G

(8)

Expanding the method to biologically relevant peptides :

apamins

1 st

_{biologically relevant peptide : Apamin}

Naturally occurring 18-residue peptide contained in the venom of bees

With 3 possible intramolecular disulfide pairings (conceptual rendering) :

(Cys1 – Cys11 / Cys3 – Cys15)

C

1

-C

3

/C

2

-C

4

Naturally occuring Apamin

(Cys1 – Cys15 / Cys3 – Cys11)

C

1

-C

4

/C

2

-C

3

Purely synthetic

(Cys1 – Cys3 / Cys11 – Cys15)

C

1

-C

2

/C

3

-C

4

Purely synthetic

C N C P K E T A L C R R A C Q Q H A C N C K P E T A L C A R R C Q Q H A CN C_K _P E _{T A L C}A R R C Q Q H A SH SH SH SH

ApaRib

Apamin

ApaBea

-CO-NH₂

N

_ter 1 3 11 15 C N C K A P E T A L C A R R C Q Q H

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

(9)

IM-MS/MS results of the apamins

3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 % 0 100 4.85 [M+4H+_]4+

ApaBea

C

1

-C

2

/C

3

-C

4 Drift Time (ms) 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 % 0 100 4.41 [M+4H+_]4+

Apamin

C

1

-C

3

/C

2

-C

4 Drift Time (ms) 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 % 0 100 4.41 4.96 [M+4H+_]4+

ApaRib

C

1

-C

4

/C

2

-C

3 Drift Time (ms) 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 % 0 100 4.41 4.85 [M+4H+_]4+

Mix of ApaRib,

Apamin

and ApaBea

Drift Time (ms) y13 y14 y12 y11 y10 y9 y8 [y13]3+ [y9]2+ [y10]2+ [y11]2+ [y12]2+ [y13]2+ [y14 ]2+ m/z 400 450 500 550 600 650 700 750 800 850 900 950 1000 % 0 100 _706.90 755.43 642.38 503.95 591.84 556.32 790.95 [M-APETA-H2O]3+ [M-APET or PETA-H2O]3+ [M-KAPE TAL-H2O]2+ [M+4H+_]4+ [ M-PE-H2O]3+ [M-KAPET-H2O]2+ [M-KAPE-H2O]2+ m/z 400 450 500 550 600 650 700 750 800 850 900 950 1000 % 0 100 543.62 507.77 519.94 [M-PET-H2O]3+ 567.30 759.38 600.98 723.87 809.92 667.29 874.40 929.94 [M-KAPE TA-H2O]2+ m/z 400 450 500 550 600 650 700 750 800 850 900 950 1000 % 0 100 874.40 929.94 [M-APETA-H2O]3+ 519.91 [M+4H+_]4+ 507.49 [M-APET or PETA-H2O]3+ 543.59 [M-PET-H2O]3+ 567.28 [ M-PE-H2O]3+ 600.96 [M-KAPE TAL-H2O]2+ 667.29 723.84 [M-KAPE TA-H2O]2+ [M-KAPE-H2O]2+ 809.92 [M-KAPET-H2O]2+ 759.35

MS/MS spectra

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

C N CK P E T A L C A R R C Q Q H A ApaRib: C1-C4/C2-C3 C N C P K E_T A L C R R A C Q Q H A Apamin: C1-C3/C2-C4 CN CK PE T A L CA R_R C Q Q H A ApaBea: C1-C2/C3-C4

(10)

CZE-MS results of the apamins in an

acidic

buffer

R MT to G lu fib 0.45 0.47 0.49 0.51 0.53 0.55 0.45 0.47 0.49 0.51 0.53 0.55

Determination of RMT (separate disulfide isomers)

RMT = relative migration time σ = standard deviation n = number of replicates

Determination of RMT (mix of the disulfide isomers)

RMT = relative migration time σ = standard deviation n = number of replicates R MT to G lu fib Peptide RMT (n=6) σ (n=6) %σ (n=6) ApaRib 0,46764 0,00086 0,18% Apamin 0,49250 0,00075 0,15% ApaBea 0,52866 0,00312 0,59% Peptide RMT (n=6) σ (n=6) %σ (n=6) ApaRib 0,47529 0,00154 0,32% Apamin 0,48571 0,00211 0,43% ApaBea 0,52124 0,00169 0,32%

BGE =

100mM formic acid

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Time (min) 0 25 50 75 100 R elati ve Abundance 12.09 13.27 12.31 25.47 RMT = 0,47468 ApaRib RMT = 0,48331 Apamin RMT = 0,52101 ApaBea Glufib

30µm x 150µm x 90cm BFS @+30kV

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

(11)

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

Could pH optimization of the buffer improve the CZE-MS

separation of the apamins ?

ApaRib: C1-C4/C2-C3 Apamin: C1-C3/C2-C4 ApaBea: C1-C2/C3-C4 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Time (min) 0 25 50 75 100 R elati ve Abundance 12.09 13.27 12.31 25.47 RMT = 0,47468 ApaRib RMT = 0,48331 Apamin RMT = 0,52101 ApaBea Glufib

Insufficient separation

?

Diagram of theoretical average

charge distribution (in-solution) of

apamins according to

Henderson-Hasselbalch

1 3 11 15

C N C K A P E T A L C A R R C Q Q H

Nter

(12)

CZE-MS results of the apamins in a

neutral

buffer

0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 R MT to MR FA

RMT = relative migration time σ = standard deviation n = number of replicates RMT to MRFA Pe ptide RMT (n=6) σ (n=6) %σ (n=6) Apamin 0,86142 0,00199 0,23% ApaRib 0,88990 0,00376 0,42% ApaBe a 0,96019 0,00644 0,67% Pe ptide RMT (n=6) σ (n=6) %σ (n=6) Apamin 0,85550 0,00230 0,27% ApaRib 0,88404 0,00292 0,33% ApaBe a 0,91875 0,00133 0,15%

BGE =

NH

₄

Ac 50mM pH 7

Time (min) 0 25 50 75 100 R el ativ e Abundanc e

(1) : RMT = 0,85460 → Apamin (C

1

-C

3

/C

2

-C

4

)

(2) : RMT = 0,88172 → ApaRib (C

1

-C

4

/C

2

-C

3

)

(1) : RMT = 0,91774 → ApaBea (C

1

-C

2

/C

3

-C

4

)

15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 19.22 20.64 19.83 MRFA (1) 22.49 (2) (3)

MRFA = reference migration time

(Glufib is anionic at pH=7)

30µm x 150µm x 90cm BFS @+20kV

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

(13)

Expanding the method to biologically relevant peptides :

conotoxins

2 nd

_{biologically relevant peptide : α and χ conotoxins}

Naturally occurring 11-residue peptide contained in the venom of marine cone snails

With 2 possible intramolecular disulfide pairings (conceptual rendering) :

SH

C C

H

S

S W C K H L C

1 2 7 11 C C S H S W C K H L C

Conotoxin χ

(Cys1 – Cys11 / Cys2 – Cys7) C₁-C₄/ C₂-C₃ Naturally occuring C C S H S W C K H L C

Conotoxin α

(Cys1 – Cys7 / Cys2 – Cys11) C₁-C₃/ C₂-C₄ Purely synthetic

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

(14)

C C H S S W C K H L C

b7 b8 b9 b10 W 159.10 H [M + 2H+_]2+ m/z 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 % 0 100 110.07 651.76 223.16HL-28 [y9 2SH]2+ 549.71 [M-CO + 2H+_]2+ 637.70 [a7 1SH]+ 777.29 [b7 1SH]+ 805.29 [b8 1SH]+ 933.35 [b9 1SH - H2O]+ 1054.36 [b10 1SH - H2O]+ 1165.44 [M + 2H+_]2+ m/z 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 % 0 100 H 110.08 651.77 [b9 1SH - H2O]+ 1054.36 W 159.10 [y9 2SH]2+ 549.71 [y31SH]+ 372.18 HL-28 223.17 [b10 1SH - H2O]+ 1165.46 [a9 1SH]2+ 521.72 [M-CO + 2H+_]2+ 637.70 [b7 1SH]+ 805.29 [b8 1SH]+ 933.35 [a7 1SH]+ 777.29 % 0 100 6.65 6.00 6.50 7.00 7.50 8.00 Drift Time (ms)

[M+2H

+

_]

2+

α conotoxin

C

₁

-C

₃

/C

₂

-C

₄ % 0 100 7.06 6.00 6.50 7.00 7.50 8.00 Drift Time (ms)

[M+2H

+

_]

2+

χ conotoxin

C

₁

-C

₄

/C

₂

-C

₃ % 0 100 7.06 6.65 6.00 6.50 7.00 7.50 8.00

[M+2H

+

_]

2+ Drift Time (ms)

Mix conotoxins

(B)

(C)

(A)

IM-MS/MS results of α and χ conotoxins

MS/MS spectra

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

C C H S S W C K H L C

b7 b8 b9 b10

(15)

CZE-MS results of the conotoxins in an

acidic

buffer

-2,00 -1,00 0,00 1,00 2,00 3,00 4,00 5,00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 N et C ha rg e pH

Poor separation

?

BGE =

100mM formic acid

Diagram of theoretical average

charge distribution (in-solution) of

conotoxins according to

Henderson-Hasselbalch

30µm x 150µm x 90cm BFS @+30kV

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

Electropherogram of the disulfide isomers mix

R MT to Gl uf ib 0.690 0.695 0.700 0.705 0.710 0.715 0.720 0.00 1.00 2.00 Peptide RMT (n=6) σ (n=6) %σ (n=6) χ conotoxin 0,70396 0,00094 0,13% α conotoxin 0,70281 0,00016 0,02%

(A)

χ conotoxin α conotoxin

(B)

0 5 10 15 20 Time (min) 0 25 50 75 100 R el ativ e abundanc e 11.30 8.49 Glufib

Unresolved mix

of conotoxins

(16)

CZE-MS results of the conotoxins in a

neutral

buffer

(C)

(A)

(B)

Peptide RMT (n=6) σ (n=6) %σ (n=6) χ conotoxin 1,02815 0,00305 0,30% α conotoxin 1,07469 0,00181 0,17% Peptide RMT (n=6) σ (n=6) %σ (n=6) χ conotoxin 1,01371 0,00128 0,13% α conotoxin 1,06252 0,00166 0,16% MRFA R el ativ e Abundanc e 0 25 50 75 100 RMT = 1,01185 χ conotoxin RMT = 1,06122 α conotoxin 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0 16.12 15.37 15.19 1.00 1.02 1.04 1.06 1.08 0 χ conotoxin 1 α conotoxin 2 1.00 1.02 1.04 1.06 1.08 0 1 2 R M T to M R F A R M T to M R F A χ conotoxin α conotoxin χ conotoxin : C1-C4/C2-C3 α conotoxin : C1-C3/C2-C4

BGE =

NH

₄

Ac 25mM pH 7

Base-peak resolution achieved

30µm x 150µm x 90cm BFS @+30kV

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

(17)

IM-MS and CZE-MS results highlight

differential migration behaviors

between α and χ conotoxins

1) α conotoxin

_CC S H S W C K H L C

2) χ conotoxin

_CC S H SW C K H L C

2) α conotoxin

_CC S H S W C K H L C

1) χ conotoxin

_CC S H SW C K H L C % 0 100 7.06 6.65 6.00 6.50 7.00 7.50 8.00

[M+2H

+

_]

2+ Drift Time (ms) C C S H SW C K H L C C C S H S W C K H L C α conotoxin χ conotoxin Time (min) R el ativ e Abundanc e 0 25 50 75 100 χ conotoxin α conotoxin 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0 16.12 15.37 C C S H S W C K H L C C C S H SW C K H L C R el ativ e Abundanc e

Comparison between the migrations in the gas phase (IMS)

and the solution (CZE)

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

CZE-MS

IM-MS

(18)

Theoretical calculations for structure elucidation

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

IMS: CCS estimation from structure optimization (in vacuo)

CZE: µ

_e

estimation from structure and charge state optimization (in water)

Non optimized folding

Structure optimization

for

studied charge states

Molecular Mechanics

AMBER in vacuo

Optimized geometry

CCS calculation

Averaged surface

seen by IMS gas

IMOS N

2

PA/EHSS

n+

Non optimized

folding

Structure optimization

for the

each charge state

Molecular Mechanics

AMBER CPCM water

Pre-optimized

geometry

n+

Protonation state

energies calculation

H++ computation

(1)

Computed pK

_a

shift

induced by the folding

Stable pK

a

shift ?

Structure reoptimization for the

most probable charged residues

Molecular Mechanics

AMBER CPCM water

Final optimized

geometry

n+

Averaged

hydrodynamic

radius determination

µ

_e

estimation for

each pH value

(MS buffer compatible)

Form factor

evaluation

R

g

/R

h

conversion

n+

(n+1)+

(n-1)+

Relevant optimized

geometries for a given pH

(19)

Theoretical calculations for structure elucidation

1) Context

2) Model

peptide

3) Apamin

4) Conotoxin

5) Modeling

6) Conclusion

IMS: CCS estimation from structure optimization (in vacuo)

Disulfide

isomer

Calculated CCS

(EHSS)

ApaRib

377Å

Apamin

405Å

ApaBea

474Å

3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 % 0 100 4.85 [M+4H+_]4+ Drift Time (ms) 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 % 0 100 4.41 [M+4H+_]4+ Drift Time (ms) 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 % 0 100 4.41 4.96 [M+4H+_]4+ Drift Time (ms) 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 % 0 100 4.41 4.85 [M+4H+_]4+

Mix of ApaRib,

Apamin

and ApaBea

Drift Time (ms) C N CK P E T A L C A R R C Q Q H A ApaRib: C1-C4/C2-C3 C N C P K E_T A L C R R A C Q Q H A Apamin: C1-C3/C2-C4 CN CK PE T A L CA R_R C Q Q H A ApaBea: C1-C2/C3-C4

7% CCS difference

<

IMS resolution

(20)