Simultaneous Assessment of Protein Heterogeneity and Affinity by

Page 1 of 29

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

1

Simultaneous assessment of protein heterogeneity and affinity by capillary

2

electrophoresis-mass spectrometry

3

E. Domínguez-Vega1*, R. Haselberg1, G.W. Somsen1, G.J. de Jong2

4 5

6 7

8 9

10

1

Division of BioAnalytical Chemistry, VU University, de Boelelaan 1083, 1081 HV

Amsterdam, The Netherlands 2

Biomolecular Analysis, Utrecht University, P.O. Box 80082, 3508 TB Utrecht, The

Netherlands *Corresponding author: [email protected]

11

12

13

14

15

16

17

18

19

20 1

ACS Paragon Plus Environment


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

Page 2 of 29

21

Abstract

22

Most conventional analytical tools for the assessment of protein-protein interactions yield

23

information on the bulk sample. Employing the efficient separation of intact proteins, affinity

24

capillary electrophoresis (ACE) can measure the interaction of components of heterogeneous

25

proteins with a target protein. In this work, the hyphenation of ACE with mass spectrometry

26

(MS) is presented as a novel highly-selective tool for the assessment of protein-protein

27

interactions. The binding of the protease inhibitor aprotinin to trypsinogen was used as

28

protein-protein affinity model. A trypsinogen sample comprising several modifications was

29

analyzed using a background electrolyte of 25 mM ammonium acetate (pH 8.0) containing

30

increasing concentrations of aprotinin (0-300 µM). A capillary coating of Polybrene-dextran

31

sulfate-Polybrene (PB-DS-PB) was employed to prevent adsorption of the proteins to the

32

capillary wall. The trypsinogen variants were separated and could be assigned based on

33

detected molecular masses and relative migration. In presence of aprotinin, both free and

34

aprotinin-bound trypsinogen were detected revealing a 1:1 binding stoichiometry. For most

35

trypsinogen variants, shifts in electrophoretic mobility were observed upon raising the

36

aprotinin concentration, allowing determination of their dissociation constants (Kds). The

37

interacting trypsinogen variants showed similar affinity towards aprotinin (Kds of 3-9 µM),

38

which were not significantly different from the values obtained with ACE-UV and were in

39

agreement with an earlier reported value. The use of the ratio of obtained MS signal

40

intensities of free and protein-protein complex for the determination of Kds was also

41

explored. Derived Kd values (20-104 µM) for the binding variants were similar to those

42

obtained with direct-infusion MS, but higher and less precise as compared with values based

43

on mobility shifts. The suitability of the ACE-MS methodology for the affinity profiling of

44

heterogeneous protein samples was evaluated and components with high, medium or low -

45

affinity towards aprotinin could be successfully discriminated. 2


Page 3 of 29

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


46

1. Introduction

47

The bioactivity of proteins is based on their capacity to bind to their targets, which frequently

48

are protein receptors or enzymes. The availability of efficient analytical tools for the study of

49

protein-protein interactions is essential in protein science

50

methods can be used for the determination of protein-protein affinities. Among these are

51

surface plasmon resonance, calorimetry, fluorescence, circular dichroism and nuclear

52

magnetic resonance techniques 3,4. Most of these methodologies, however, are not suitable for

53

the simultaneous determination of the equilibrium dissociation constants (Kds) of multiple

54

sample components, such as protein isoforms and/or impurities. Nevertheless, proteins

55

seldomly are pure, unimolecular compounds and may comprise different post-translational

56

modifications (PTMs) or degradation products. For instance, during production, isolation

57

and/or storage, proteins may undergo modifications that may dramatically influence

58

bioactivity, efficacy and toxicity, thereby compromising their biological properties 5. In other

59

cases, PTMs (e.g. glycosylation) may be essential to maintain proper protein activity.

60

Consequently, the possibility to assess the affinity of all protein variants present in a sample

61

is essential.

62

Affinity capillary electrophoresis (ACE) represents a powerful technique for the assessment

63

of interactions of protein sample components towards a target protein

64

proteins is well established, providing efficient separations under aqueous conditions and in

65

absence of a stationary phase

66

proper capillary coatings. For instance, positively- and negatively-charged capillary coatings

67

have shown very successful in achieving high resolution separation of diverse proteins with

68

various pIs

69

ACE offers the possibility to simultaneously study the interaction of multiple proteins under

70

homogeneous and near-physiological conditions 11. Affinity measurement by CE is based on

1-3

. A number of in-solution

6-11

. CE of intact

12-14

. Protein adsorption can be prevented effectively by use of

15,16

. Employing the efficient separation of intact proteins as provided by CE,

3



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

Page 4 of 29

71

the fact that the electrophoretic mobility of the probed protein differs from the electrophoretic

72

mobility of the protein-ligand or protein-receptor complex. When association/dissociation

73

kinetics are slower or similar to the CE separation time, pre-equilibrium or kinetic ACE can

74

be applied, respectively

75

significant time and, subsequently, ligand, receptor and complex are separated by CE

76

assuming that the complex does not dissociate in the course of the analysis. In kinetic ACE,

77

the observed change in electrophoretic mobility and shape (but not area) of the ligand peak

78

when analyzed in presence of protein receptor, may be used for the determination of

79

dissociation constants

80

applying specific experimental settings and data analysis approaches for the determination of

81

dissociation constants under non-equilibrium

82

equilibrium kinetics are fast, a change in effective electrophoretic mobility will be observed

83

for the injected protein(s) when the receptor or ligand is added to the BGE. From the

84

observed mobility shifts at increasing concentrations of receptor or ligand in the BGE, Kds

85

can be calculated. Mobility-shift ACE has the main advantage that Kds of multiple sample

86

components can be determined simultaneously. Using ACE with UV absorbance detection,

87

reliable Kd determination requires components to be fully separated. However, no

88

information on the identity of the sample components is obtained and the stoichiometry of the

89

formed complex cannot be determined.

90

Coupling CE with mass spectrometry (MS) yields a powerful tool for protein analysis, as it

91

combines efficient separation of intact proteins with the mass-selective detection and analyte

92

characterization possibilities provided by MS 13,14,23,24. In addition, resolution of co-migrating

93

compounds can be achieved by MS by monitoring component-specific m/z values. Recently,

94

our group has shown the potential of the parallel use of ACE-UV and CE-MS for the study of

95

the affinity of a nanobody product with the epidermal growth factor receptor 6. The nanobody

10

. In pre-equilibrium ACE, ligand and protein are incubated for a

17,18

. Krylov et al. have developed a series of kinetic ACE methods

19

and equilibrium

20-22

conditions. When

4


Page 5 of 29

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


96

comprised of several variants and ACE-UV was used to establish their individual Kd values.

97

CE-MS was employed for characterization of these variants. Instead of using two parallel

98

techniques, the direct coupling of ACE with MS would provide a potentially strong platform

99

for the simultaneous assessment of protein heterogeneity and affinity. Moreover, MS can

100

reveal the stoichiometry of the protein-complex and protein signal intensities potentially can

101

also be used to gain information about protein affinity. In fact, stand alone MS has been used

102

for the determination of the affinity of proteins 10,25-28. Hence, in principle MS data could be

103

used to gain additional affinity information, but some considerations as different ionization

104

efficiencies of unbound and bound protein and/or different stabilities of the complexes in the

105

gas phase should be taken into account 26-28.

106

So far, few studies have been reported employing the concept of ACE-MS for affinity

107

determination by monitoring mobility shifts. The majority of these works focused on the

108

screening of the affinity of di- tri- and tetrapeptides towards antibiotics 29-33. Other mobility-

109

shift based ACE-MS studies deal with the interaction of oligosaccharides with the stromal

110

cell-derived factor-1 receptor 34, and the affinity of small drugs for β-cyclodextrin 25. To our

111

knowledge, protein-protein interactions have never been studied by ACE-MS before. In the

112

present work, the on-line combination of ACE and MS was evaluated as a new selective tool

113

for the simultaneous assessment of protein heterogeneity and protein-protein affinity. The

114

protease trypsinogen (24 kDa) and its inhibitor aprotinin (6.5 kDa) were selected as a model

115

system. These proteins can be efficiently analyzed by CE-MS and have relatively fast

116

equilibration kinetics (koff > 4·10-2 s-1) with a Kd in the low-µM range 35, in principle allowing

117

affinity analysis monitoring changes in electrophoretic mobility. We also studied the

118

usefulness of the developed ACE-MS methodology for the screening of aprotinin-affinity

119

components in heterogeneous protein samples.

120 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

Page 6 of 29

121

2. Materials and methods

122

2.1.

123

All reagents employed were of analytical grade. Ammonium hydroxide (25% solution), was

124

obtained from Fluka (Steinheim, Germany). Isopropanol (IPA) and acetic acid was provided

125

by Merck (Darmstadt, Germany). Formic acid was supplied by Riedel-De Haen (Seelze,

126

Germany). Trypsinogen, trypsin and α-chymotrypsin from bovine pancreas, aprotinin

127

(bovine lung trypsin inhibitor), cortisone, Polybrene (hexadimethrine bromide, PB) and

128

dextran sulfate (DS) were obtained from Sigma-Aldrich (Steinheim, Germany). Deionized

129

water was obtained from a Milli-Q purification system (Millipore, Bedford, USA).

130

2.2.

131

CE analyses were performed using a Beckman PA 800 plus instrument (Beckman Coulter,

132

Brea, USA). Fused silica capillaries were from Polymicro Technologies (Phoenix, AZ, USA)

133

having an internal diameter of 50 µm and a total length of 60 cm (UV detection) and 80 cm

134

(MS detection). Hydrodynamic injections were performed at 1 psi for 8 s (60-cm capillaries)

135

or 1 psi for 10 s (80-cm capillaries). The separation voltage was 30 kV and the capillary

136

temperature was 20 °C. UV detection was performed at 214 nm after 50 cm on the capillary.

137

The background electrolyte (BGE) was 25 mM ammonium acetate (pH 8.0) (prepared by

138

diluting the appropriate amount of ammonium hydroxide in water and setting the pH with

139

acetic acid) containing 0-300 µM aprotinin. New fused-silica capillaries were rinsed with 1

140

M NaOH for 30 min at 20 psi, and with water for 15 min at 20 psi. After this treatment,

141

capillaries were coated with a PB-DS-PB coating. For this, solutions of 10% (w/v) PB and

142

0.5% (w/v) DS in deionized water were prepared. The solutions were filtered over a 0.45 µm

143

filter type HA (Millipore, Molsheim, France) prior to use. Capillaries were coated by

144

subsequently rinsing 30 min with 10% (w/v) PB solution at 5 psi, 10 min with deionized

Chemicals

Capillary electrophoresis

6


Page 7 of 29

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


145

water at 10 psi, 30 min with 0.5% (w/v) DS solution at 5 psi, 10 min with deionized water at

146

10 psi, 30 min with 10% (w/v) PB solution at 5 psi, and 10 min with deionized water at 10

147

psi. After the final coating step, the capillary was rinsed for 10 min with 25 mM ammonium

148

acetate (pH 8.0) at 20 psi. Before each run, coated capillaries were flushed with water for 2

149

min, 25 mM ammonium acetate (pH 8.0) for 2 min and the BGE for 2 min at 20 psi.

150

Overnight, capillaries were filled with water and tips were immersed in vials with water.

151

2.3.

152

MS experiments were carried out on a micrOTOF-QII mass spectrometer (Bruker Daltonics,

153

Bremen, Germany). Source and transfer parameters were optimized by direct infusion of the

154

proteins via the CE capillary. CE-MS coupling was performed using a sheath-liquid

155

electrospray interface from Agilent Technologies (Waldbronn, Germany). The sheath liquid

156

was a mixture of isopropanol-water-25 mM ammonium acetate (pH 8.0) (25:50:25, v/v/v)

157

delivered at a flow rate of 2 µL·min-1 using a syringe pump from Cole-Parmer (Vernon Hill,

158

IL USA). The mass spectrometer was operated in positive-ion mode with an electrospray

159

voltage of 4.5 kV. The nebulizer and drying gas conditions were 0.2 bar and 3 L·min-1

160

nitrogen at 80 ºC, respectively. Transfer parameters were adjusted for optimal detection of

161

non-covalent protein complexes. Quadrupole ion and collision cell energy were 5 and 7 eV,

162

respectively. Transfer and pre plus storage times were set at 200 and 20 µs, respectively. The

163

monitored mass range was 250 to 3000 m/z. Extracted ion electropherograms (EIEs) were

164

obtained with an extraction window of ± 0.5 m/z and using the smooth option of the software

165

(Gaussian at 1 point).

166

2.4.

Mass spectrometry

Affinity CE

7



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

Page 8 of 29

167

BGEs containing increasing concentrations of aprotinin (0–300 µM) were prepared and used

168

for the analysis of trypsinogen (40 and 160 µM for UV and MS detection, respectively). The

169

effective electrophoretic mobility (µeff) of trypsinogen and its components were

170

experimentally determined using Eq. (1):

171

172

(1)

ߤ݂݂݁ =

‫ ݀ܮ ݐܮ‬1 1 ቆ − ቇ [ܿ݉2 ܸ −1 ‫ ݏ‬−1 ] ܸ ‫݂݋݁ݐ ܲݐ‬

173 174

where Lt is the total capillary length, Ld is the capillary length to detector, V is the applied

175

voltage, and tp and teof are the migration times of the protein and EOF marker, respectively.

176

For EOF measurements, cortisone (50 mM) was injected (0.5 psi for 5 s) immediately after

177

trypsinogen injection. For each aprotinin concentration, the difference of the measured µeff of

178

trypsinogen and derived products with the µeff obtained when no ligand was added to the

179

BGE, was determined and plotted versus the ligand concentration. These plots were fitted

180

using Prism 5.0 (GraphPad Software Inc., La Jolla, CA, USA) applying nonlinear regression

181

assuming a 1:1 binding stoichiometry according to Eq. (2):

182

(2)

∆ߤ݂݂݁ =

‫ݎܥ ݔܽ݉ܤ‬ ‫ ݀ܭ‬+ ‫ݎܥ‬

183

where ∆µeff is the effective mobility difference of trypsinogen obtained at a specific aprotinin

184

concentration compared to when no aprotinin was added to the BGE, Bmax is the maximum

185

mobility shift, Kd is the dissociation constant, and cr is the ligand concentration added to the

186

BGE. Kd values and their standard deviations were derived by the software using the

187

Levenberg-Marquardt algorithm.

188

For the evaluation of potential viscosity changes upon addition of aprotinin to the BGE, the

189

CE instrument was employed as a viscosimeter. Applying a constant pressure of 1.5 psi, the

8


Page 9 of 29

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


190

time needed for a trypsinogen solution (40 µM) to reach the detector was measured

191

repeatedly (n=6) for a BGE containing no or 300 µM of aprotinin.

192 193 194

3. Results and discussion

195

3.1. Set up of ACE-MS system

196

ACE can be performed using near-physiological BGEs containing e.g. phosphate or Tris

197

buffer. However, coupling with ESI-MS requires volatile BGEs to avoid analyte ionization

198

suppression and source contamination. Recent studies show that replacing sodium phosphate

199

or Tris acetate for e.g. ammonium acetate or bicarbonate at medium pH does not significantly

200

alter protein-ligand interactions

201

8.0

202

selected. As trypsinogen (pI 9.3) and aprotinin (pI 10.5) are overall positively charged at pH

203

8.0, a triple layer coating (PB-DS-PB) was employed in order to avoid adsorption of the

204

proteins to the capillary wall

205

by CE-UV. Trypsinogen showed a cluster of peaks migrating in the 4.5-5.5 min range. A

206

single peak was obtained for aprotinin with a significantly longer migration time than

207

trypsinogen (6.8 min). In order to test whether the protein-protein affinity in principle can be

208

monitored by CE, trypsinogen (~40 µM) was injected employing a BGE of 25 mM

209

ammonium acetate (pH 8.0) containing 35 µM aprotinin. A significant shift of the effective

210

electrophoretic mobility of trypsinogen was observed in comparison with the analysis of

211

trypsinogen in absence of aprotinin, indicating measurable protein-protein interaction.

212

Subsequently, infusion experiments were performed in order to determine optimal MS

213

conditions for the detection of free and complexed protein. Trypsinogen (40 µM) was infused

214

via the CE capillary at a flow rate of 300 nL·min−1 into the mass spectrometer using the CE-

35

36,37

. Trypsinogen and aprotinin show optimal affinity at pH

and, therefore, for CE analysis a BGE of 25 mM ammonium acetate (pH 8.0) was

38

. Both trypsinogen and aprotinin were analyzed individually

9



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

Page 10 of 29

215

MS sprayer. Trypsinogen showed an ESI mass spectrum with a charge envelope of four

216

charge states (8 to 11 positive charges). Deconvolution of this mass spectrum yielded a

217

molecular weight of 23,981 Da, which is in good agreement with the expected mass for

218

trypsinogen. Next, trypsinogen (40 µM) was incubated with aprotinin (150 µM) for 10 min

219

(the minimum CE analysis time) and then infused into the mass spectrometer. After

220

deconvolution, the mass spectrum showed three main masses of 6,511 Da, 23,981 Da, and

221

30,493 Da which correspond to aprotinin, trypsinogen and the non-covalent trypsinogen-

222

aprotinin complex (stoichiometry of 1:1), respectively. However, the signal intensity for the

223

protein-protein complex was relatively low, even though significant binding was expected at

224

these relative concentrations. Probably, dissociation of the non-covalent complex occurred

225

during ionization and ion transfer through the ion optics. Therefore, ESI-MS parameters were

226

evaluated in order to achieve efficient detection of the aprotinin-trypsinogen complex.

227

Various dry gas flow rates (2-7 L·min−1), dry gas temperatures (70-200 ºC) and nebulizer gas

228

pressures (0.1-0.6 bar) were studied. Increase of protein-complex signals (relative to free

229

proteins signals) were obtained using relatively soft source conditions (i.e., dry gas of 80 ºC

230

at 3 L·min−1, and nebulizer gas pressure of 0.2 bar) and increased transfer and storage times

231

(200 and 20 µs, respectively). Special attention was paid to the quadrupole ion and collision

232

cell energy, since they are known to affect complex stability 26. Decrease of the collision cell

233

energy to values below 10 eV caused an increase of the complex intensity (Figure 1). A

234

similar trend was observed when lowering the quadrupole ion energy. Optimal aprotinin-

235

trypsinogen signal was achieved using a quadrupole ion energy of 5 eV and a collision cell

236

energy of 7 eV. Values below 5 eV for both collision cell and quadrupole ion energy did not

237

allow effective transfer of protein-protein complex to the detector and, no signals for the

238

protein-complex were observed (Figure 1).

10


Page 11 of 29

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


239

For the coupling of ACE with MS, the use of a sheath liquid consisting of IPA and 100 mM

240

acetic acid (75:25, v/v) was initially employed as it provided good ionization of the free and

241

aprotinin-bound proteins. However, in order to achieve a stable CE current over time,

242

addition of BGE to the sheath liquid was needed. Different IPA and BGE ratios were tested.

243

Good sensitivity and CE current stability were obtained with a sheath liquid consisting of

244

IPA-water-ammonium acetate (25 mM, pH 8.0) (25:50:25, v/v/v) and was further employed

245

for CE-MS analysis.

246

3.2. ACE-MS performance

247

3.2.1 Characterization of trypsinogen variants and their aprotinin complexes

248

The trypsinogen sample (including cortisone as EOF marker) was analyzed with the

249

optimized CE-MS method using a BGE without aprotinin. For trypsinogen five partially

250

separated peaks were detected indicating the presence of impurities and/or modified protein

251

forms (Figure 2A). Table 1 lists the masses obtained for each peak after deconvolution of the

252

ESI mass spectra. Peak 4 could be assigned to native trypsinogen. Based on the masses and

253

migration times, peaks 3, 2 and 1 were provisionally assigned to singly-, doubly- and triply-

254

deamidated trypsinogen. Adding one negative charge (at pH 8.0) and an increase of 0.984 Da

255

per modification, deamidation causes a change in the protein charge-to-size ratio, facilitating

256

CE separation. Next to triply-deamidated trypsinogen, peak 1 comprised two other

257

proteinaceous species of relatively low mass, which may be products resulting from auto-

258

proteolytic activity of trypsinogen. Peak 5 corresponded to a protein with a mass that was 18

259

Da lower than trypsinogen, but could not be assigned to an evident modification.

260

In order to study the affinity of the trypsinogen sample components, 25 µM of aprotinin was

261

added to the BGE and trypsinogen was analyzed (Figure 2B). Addition of aprotinin caused a

262

reduction of the EOF. In order to allow proper calculation of the effective electrophoretic 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

Page 12 of 29

263

mobility (µeff) of the trypsinogen variants, a neutral EOF marker (cortisone) was included in

264

the samples. Using the CE instrument as viscosimeter (see Experimental section) no

265

significant changes in viscosity were observed for these concentrations of aprotinin in the

266

BGE. Hence, no viscosity correction of the µeff was needed. The reproducibility of the µeff

267

determination was assessed in absence and presence (25 µM) of aprotinin in the BGE on

268

three different days, including installation of capillary. RSDs for µeff were below 2.3% and

269

4.7%, respectively. The extracted-ion electropherograms show trypsinogen and its variants,

270

but with shifted effective electrophoretic mobilites with respect to Figure 2A, due to

271

interaction with aprotinin in the BGE. The peaks of trypsinogen and variants remained nicely

272

symmetric upon addition of aprotinin, confirming that the equilibration kinetics were fast.

273

The mass spectra of trypsinogen and the variants showed the characteristic masses of free

274

proteins (trypsinogen variants and aprotinin), but also clear signals of protein-aprotinin

275

complexes with total masses, indicating a 1:1 binding stoichiometry. Interestingly, the two

276

unassigned proteins comigrating with the triply-deamidated trypsinogen in Figure 2A (peaks

277

1’ and 1’’) showed no shift in effective electrophoretic mobility when aprotinin was added to

278

the BGE. Hence, these proteins exhibit no measurable interaction with aprotinin. This was

279

confirmed by the fact that the mass spectra obtained for these two compounds did not show

280

signals of protein-aprotinin complexes. Overall, these results indicate that ACE-MS can be

281

employed for the simultaneous characterization and affinity assessment of impure proteins

282

samples, including comigrating species.

283

3.2.2 Quantitative assessment of protein-protein interaction

284

In order to determine the Kds of aprotinin towards the trypsinogen variants, the trypsinogen

285

sample was analyzed in BGEs containing increasing concentrations of aprotinin. As

286

trypsinogen and aprotinin have a Kd in the low µM range, the aprotinin concentration in the 12


Page 13 of 29

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


287

BGE was varied between 0 and 300 µM. For most trypsinogen variants, increased µeff values

288

were observed when the concentration of aprotinin in the BGE was raised. The µeff was

289

determined for each trypsinogen variant and binding curves were constructed by plotting the

290

change in µeff against the aprotinin concentration (Figure 3A). Subsequent curve fitting using

291

non-linear regression yielded a Kd value of 8 ± 5 µM for trypsinogen, which is in reasonable

292

agreement with the only value reported so far in literature (2 µM) 35. Based on the changes in

293

effective electrophoretic mobility, Kd values of 4 ± 2, 4 ± 2, 9 ± 4 and 3 ± 2 µM were

294

determined for compounds 1, 2, 3 and 5 of the trypsinogen sample respectively (Table 1).

295

These values are not significantly different from the Kd of the native trypsinogen, indicating

296

that the chemical modifications did not affect the aprotinin binding. For the two variants

297

which co-migrated with the triply deamidated trypsinogen in absence of aprotinin, no shifts in

298

the mobility were observed in presence of aprotinin, confirming their lack of affinity.

299

In order to evaluate the accuracy of the ACE-MS method, Kds were also determined for

300

trypsinogen and its variants using ACE-UV employing the same BGEs. As several

301

components co-migrated in peak 1, it was not possible to determine their individual µeffs with

302

ACE-UV. The Kd values obtained for the rest of variants and trypsinogen were between 2 ± 1

303

and 6 ± 3 µM (Table 1). Seemingly slightly lower, these values do not differ significantly

304

from the values obtained by ACE-MS. Potential differences could be attributed to the fact

305

that in CE-MS, in contrast to CE-UV, a large part of the capillary is not thermostated. The

306

resulting somewhat higher temperature may affect protein-protein binding and, thus, the

307

observed Kd value 39,40.

308

ACE-MS in principle also allows assessment of protein-protein binding by using the obtained

309

MS data 10. Notably, the absolute abundances of the protein signals decrease due to ionization

310

suppression when the aprotinin concentration in the BGE is increased. However, as can be

311

seen in Figure 4, with increasing aprotinin concentration in the BGE, the abundance of the 13



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

Page 14 of 29

312

aprotinin-trypsinogen complex ions relatively to the signal of the free trypsinogen ions

313

increases. These relative abundances can be used to determine Kd values, circumventing

314

signal suppression effects by aprotinin

315

the trypsinogen-aprotinin complex and the trypsinogen ions as function of the aprotinin

316

concentration in the BGE indeed provides a binding curve (Figure 3B), yielding a Kd value of

317

92 ± 36 µM. For the variants of trypsinogen, the MS-derived Kd values were between 20 ± 17

318

and 104 ± 71 µM (Table 1). These values are higher and show larger spread than the Kds

319

obtained using electrophoretic mobility shifts.

320

In order to appreciate the MS-based values, the protein-protein affinity was also assessed

321

using stand-alone ESI-MS, i.e., by infusion of pre-incubated mixtures of trypsinogen (40 µM)

322

and aprotinin (0-500 µM) into the mass spectrometer. From the signal intensities measured

323

for free and complexed trypsinogen, a Kd of 96 ± 34 µM was calculated for the aprotinin-

324

trypsinogen interaction. This value represents the overall Kd of trypsinogen, i.e., including its

325

deamidated forms, as it was not possible to reliably discern these variants from the native

326

protein by MS (mass differences of 0.984 Da only) without pre-separation. For the 23,963-Da

327

trypsinogen variant, a Kd of 59 ± 20 µM was determined. These Kds were similar to the MS-

328

based Kds obtained with ACE-MS, indicating that the higher values and spread is caused by

329

MS detection and not by the ACE analysis. However, the MS-based values obtained were

330

higher compared with the electrophoretic-mobility-shift values and the reference value 35. For

331

the MS-based Kd determination it is assumed that the ratio of the measured MS signals of the

332

free protein and the protein complexes reflect the concentration ratio of these species in

333

solution. However, the ionization efficiencies for the free protein and the protein-protein

334

complex most probably are not the same

335

protein complex may vary between gas and liquid phase (e.g. hydrophobic interactions are

336

weaker than noncovalent hydrogen bonds in gas phase) 28. Kd values based on changes in µeff

26

. Plotting the ratio of the cumulative abundance of

26

. Moreover, the stability of the non-covalent

14


Page 15 of 29

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


337

are measured in solution and seem more suitable for the determination of protein-protein

338

affinities using ACE-MS.

339

It can be concluded that the proposed ACE-MS method represents a promising means for the

340

simultaneous assessment of heterogeneity and the affinity of individual protein sample

341

constituents. The method, allows determination of Kd values for closely related sample

342

components of similar mass (e.g. deamidated variants) and for co-migrating components.

343

3.3. ACE-MS for affinity screening of heterogeneous proteins

344

The developed ACE-MS system can also be used for the relatively fast affinity screening of

345

heterogeneous proteins, offering qualitative information on the affinity of mixture

346

components/protein variants towards a target protein without the need for repeated

347

experiments using different target concentrations and extensive dissociation constant

348

calculations. In order to demonstrate the usefulness of ACE-MS for this purpose, the proteins

349

trypsin and ɑ-chymotrypsin were analyzed in absence and presence of aprotinin. Aprotinin is

350

known to bind strongly to both trypsin and α-chymotrypsin with Kd values in the pM and nM

351

range, respectively

352

using 0 and 25 µM aprotinin in the BGE. The trypsin sample showed five peaks (Figure 5A1)

353

which on the basis of their mass spectra were assigned to didehydro, acetylated and

354

methylated trypsin (peaks 2-4) and trypsin (peak 5); peak 1 could not be assigned to a

355

common modification of trypsin. The lower mass of the latter component as compared with

356

the intact protein (approx. 6 kDa less) suggests a fragment of trypsin, possibly as a

357

consequence of auto-proteolytic activity. For the ACE-MS analysis of trypsin with 25 µM of

358

aprotinin in the BGE, only the aprotinin-complexed species were observed in the mass

359

spectra of peaks 2-5 (Figure 5A2, red trace), which showed a change in migration time as

360

well. This indicates that next to trypsin itself, also the didehydro, acetylated and methylated

35

. Figure 5 shows the ACE-MS analysis of trypsin and ɑ-chymotrypsin

15



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

Page 16 of 29

361

forms bind strongly to aprotinin. Interestingly, the variant with lower molecular weight (peak

362

1) did not exhibit a significant shift of the migration time (Figure 5A2, blue trace) and its

363

mass spectrum only showed ions corresponding to the free protein (blue trace). This implies

364

that this trypsin fragment has no affinity towards aprotinin.

365

CE-MS analysis of the α-chymotrypsin sample revealed four peaks (Figure 5B1). Peak 3

366

could be assigned to the unmodified protein, whereas peak 1 most probably is a deamidated

367

form. The mass spectra of peaks 2 and 4 showed higher masses compared to the unmodified

368

chymotrypsin (+216 Da for peak 2, and +196 and +246 Da for peak 4). When the α-

369

chymotrypsin was analyzed with aprotinin added to the BGE (Figure 5B2), clear relative

370

differences in affinity were observed. For the +246 Da protein species (peak 4’) only the free

371

protein was observed in the mass spectrum, indicating that this component has no affinity

372

with aprotinin. The rest of the α-chymotrypsin components showed migration time shifts in

373

presence of aprotinin, indicating binding. However, the affinities of these components

374

appeared to be different. For native and deamidated α-chymotrypsin, and the +196 Da

375

species (peak 4’’), only the aprotinin complexes were observed in their mass spectra. This

376

means that these components all have a high affinity to aprotinin. For the +216 Da species

377

(peak 2) both free and aprotinin-complexed protein was observed. This implies that this

378

variant has interaction with aprotinin, but its affinity is lower as compared with α-

379

chymotrypsin itself. Overall, these results demonstrate that the new ACE-MS method can be

380

highly useful for the screening of protein-protein affinities of components of heterogeneous

381

protein samples.

382 383

4. Concluding remarks

16


Page 17 of 29

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


384

The feasibility of ACE-MS for the simultaneous characterization of intact proteins and

385

determination of protein-protein affinity was studied using trypsinogen and the protease

386

inhibitor aprotinin as model compounds. Efficient protein separation and determination of

387

affinity-induced changes of protein electrophoretic mobilities was achieved using a PB-DS-

388

DS coated capillary. The use of soft ionization and ion transfer conditions was crucial to

389

provide adequate MS sensitivity while maintaining the formed protein-protein complexes.

390

Both CE and MS data acquired with ACE-MS, i.e. effective electrophoretic mobilities and

391

ratio of signal intensities of free and complexed protein, can be used to determine

392

dissociation constants. By measuring effective electrophoretic mobility shift of trypsinogen

393

as function of aprotinin concentration, similar Kds were found for all the interacting

394

trypsinogen variants, indicating comparable affinity. MS-based Kd assessment suffered from

395

a relative large spread in ESI-MS signal intensities and deviation in ESI efficiencies among

396

free and complexed protein. The ACE-MS approach also showed useful for screening of

397

high, medium and low-affinity components in heterogeneous protein samples. Overall, the

398

present work demonstrates that ACE-MS is a powerful tool for the study of protein-protein

399

interactions exhibiting fast equilibrium kinetics. In a single run, ACE-MS provides (a)

400

assignment of the molecular weight of all protein sample components, including modified

401

forms, other variants and impurities, (b) protein-variant-selective determination of multiple

402

dissociation constants, and (c) establishment of protein-complex stoichiometries. Application

403

of the proposed ACE-MS methodology to larger proteins and biopharmaceuticals is currently

404

being studied in our laboratory.

405

Still there are some limitations of the present method. Fast equilibration is not always the

406

case in protein-protein interactions. Alternative approaches as pre-equilibrium or kinetic ACE

407

may be employed in these cases 10. Use of BGEs at (near)-physiological conditions may not

408

provide optimal CE separation and/or ESI-MS sensitivity. Moreover, determination of Kd’s in 17



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

Page 18 of 29

409

the high µM and mM range may be an issue, as the higher amounts of ligand required in the

410

BGE may result in strong suppression of the MS signal. Sensitivity can be increased and

411

ionization suppression can be reduced by the employment of sheathless interfacing for CE-

412

MS using, e.g.,

413

thermostatting almost the entire CE capillary. We plan to study the feasibility of sheathless

414

ACE-MS in the near future.

a porous tip capillary

41,42

. Sheathless interfacing also provides

415

416

18


Page 19 of 29

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


417

Acknowledgment

418

This research was supported by the Dutch Technology Foundation STW, which is part of the

419

Netherlands Organisation for Scientific Research (NWO) and partly funded by the Ministry

420

of Economic Affairs (project number 11056).

421

E. Domínguez-Vega and R. Haselberg contributed equally to this work.

422

19



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

Page 20 of 29

423

Figure captions

424

Figure 1. Effect of the quadrupole ion and collision cell energy on the intensity of the

425

trypsinogen-aprotinin complex ions. Infusion of a pre-incubated mixture of trypsinogen (40

426

µM) and aprotinin (150 µM) at a flow rate of 300 nL·min−1. MS conditions: ESI voltage, -4.5

427

kV; nebulizer, 0.2 bar; drying gas, 3 L/min at 80 ºC. For other conditions, see Experimental

428

section.

429

Figure 2. CE-MS analysis of trypsinogen sample using a PB-DS-PB coated capillary and a

430

BGE of 25 mM ammonium acetate (pH 8.0) containing no (A) or 25 µM (B) aprotinin.

431

Extracted-ion elecropherograms of trypsinogen (2398 m/z), trypsinogen variants (2397-2399,

432

1814.1 and 1798.5 m/z) and cortisone (EOF marker) (361.2 m/z) are depicted. For other

433

conditions, see Experimental section.

434

Figure 3. ACE-MS binding curves for trypsinogen-aprotinin obtained by plotting (A)

435

effective electrophoretic mobility shifts and (B) the ratio between the relative intensity of

436

trypsinogen-aprotinin complex/free trypsinogen against the concentration of aprotinin in the

437

BGE. For experimental conditions, see Experimental section.

438

Figure 4. Mass spectra obtained for peak 4 during ACE-MS of trypsinogen using a BGE of

439

25 mM ammonium acetate (pH 8.0) containing no (A), 2 (B), 15 (C), 50 (D) or 150 µM (E)

440

aprotinin. Arrows indicate trypsinogen-aprotinin complex ions. For other conditions, see

441

Experimental section.

442

Figure 5. CE-MS analysis of trypsin (A) and α-chymotrypsin (B) using a PB-DS-PB coated

443

capillary and a BGE of 25 mM ammonium acetate (pH 8.0) containing no (A1 and B1) or 25

444

µM (A2 and B2) aprotinin. Cumulative extracted-ion electropherograms of free proteins

20


Page 21 of 29

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


445

(blue traces) and protein complexes (red traces) are depicted. For other conditions, see

446

Experimental section.

447

21



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

Page 22 of 29

448

References

449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504

(1) Johnsson, N. Biochem Bioph Res Co 2014, 445, 739-745. (2) Heck, A. J. R.; van den Heuvel, R. H. H. Mass Spectrom Rev 2004, 23, 368-389. (3) Berggard, T.; Linse, S.; James, P. Proteomics 2007, 7, 2833-2842. (4) Syafrizayanti; Betzen, C.; Hoheisel, J. D.; Kastelic, D. Expert Rev Proteomic 2014, 11, 107120. (5) Kalman-Szekeres, Z.; Olajos, M.; Ganzler, K. J Pharmaceut Biomed 2012, 69, 185-195. (6) Haselberg, R.; Oliveira, S.; van der Meel, R.; Somsen, G. W.; de Jong, G. J. Anal Chim Acta 2014, 818, 1-6. (7) He, X. Y.; Ding, Y. S.; Li, D. Z.; Lin, B. C. Electrophoresis 2004, 25, 697-711. (8) Albishri, H. M.; El Deeb, S.; AlGarabli, N.; AlAstal, R.; Alhazmi, H. A.; Nachbar, M.; El-Hady, D. A.; Watzig, H. Bioanalysis 2014, 6, 3369-3392. (9) El Deeb, S.; Watzig, H.; Abd El-Hady, D. Trac-Trend Anal Chem 2013, 48, 112-131. (10) Chen, Z.; Weber, S. G. Trends in analytical chemistry : TRAC 2008, 27, 738-748. (11) Galievsky, V. A.; Stasheuski, A. S.; Krylov, S. N. Anal Chem 2015, 87, 157-171. (12) Zhao, S. S.; Chen, D. D. Y. Electrophoresis 2014, 35, 96-108. (13) Haselberg, R.; de Jong, G. J.; Somsen, G. W. Electrophoresis 2011, 32, 66-82. (14) Haselberg, R.; de Jong, G. J.; Somsen, G. W. Electrophoresis 2013, 34, 99-112. (15) Haselberg, R.; Brinks, V.; Hawe, A.; de Jong, G. J.; Somsen, G. W. Anal Bioanal Chem 2011, 400, 295-303. (16) Haselberg, R.; de Jong, G. J.; Somsen, G. W. J Sep Sci 2009, 32, 2408-2415. (17) Avila, L. Z.; Chu, Y. H.; Blossey, E. C.; Whitesides, G. M. J Med Chem 1993, 36, 126-133. (18) Heegaard, N. H. H. J Chromatogr A 1994, 680, 405-412. (19) Krylov, S. N. Electrophoresis 2007, 28, 69-88. (20) Kanoatov, M.; Galievsky, V. A.; Krylova, S. M.; Cherney, L. T.; Jankowski, H. K.; Krylov, S. N. Anal Chem 2015, 87, 3099-3106. (21) Cherney, L. T.; Krylov, S. N. Anal Chem 2011, 83, 1381-1387. (22) Cherney, L. T.; Krylov, S. N. Analyst 2012, 137, 1649-1655. (23) Haselberg, R.; de Jong, G. J.; Somsen, G. W. J Chromatogr A 2007, 1159, 81-109. (24) Monton, M. R.; Terabe, S. Anal Sci 2005, 21, 5-13. (25) Mironov, G. G.; Logie, J.; Okhonin, V.; Renaud, J. B.; Mayer, P. M.; Berezovski, M. V. J Am Soc Mass Spectr 2012, 23, 1232-1240. (26) Liu, J.; Konermann, L. J Am Soc Mass Spectrom 2011, 22, 408-417. (27) Daniel, J. M.; Friess, S. D.; Rajagopalan, S.; Wendt, S.; Zenobi, R. Int J Mass Spectrom 2002, 216, 1-27. (28) Peschke, M.; Verkerk, U. H.; Kebarle, P. J Am Soc Mass Spectr 2004, 15, 1424-1434. (29) Chu, Y. H.; Dunayevskiy, Y. M.; Kirby, D. P.; Vouros, P.; Karger, B. L. J Am Chem Soc 1996, 118, 7827-7835. (30) Chu, Y. H.; Kirby, D. P.; Karger, B. L. J Am Chem Soc 1995, 117, 5419-5420. (31) Lynen, F.; Zhao, Y.; Becu, C.; Borremans, F.; Sandra, P. Electrophoresis 1999, 20, 24622474. (32) Machour, N.; Place, J.; Tron, F.; Charlionet, R.; Mouchard, L.; Morin, C.; Desbene, A.; Desbene, P. L. Electrophoresis 2005, 26, 1466-1475. (33) Dunayevskiy, Y. M.; Lyubarskaya, Y. V.; Chu, Y. H.; Vouros, P.; Karger, B. L. J Med Chem 1998, 41, 1201-1204. (34) Fermas, S.; Gonnet, F.; Sutton, A.; Charnaux, N.; Mulloy, B.; Du, Y. G.; Baleux, F.; Daniel, R. Glycobiology 2008, 18, 1054-1064. (35) Vincent, J. P.; Lazdunski, M. Febs Lett 1976, 63, 240-244. (36) Bao, J.; Krylov, S. N. Anal Chem 2012, 84, 6944-6947. (37) Vuignier, K.; Veuthey, J. L.; Carrupt, P. A.; Schappler, J. Electrophoresis 2012, 33, 33063315. (38) Haselberg, R.; de Jong, G. J.; Somsen, G. W. Anal Chim Acta 2010, 678, 128-134. (39) Musheev, M. U.; Filiptsev, Y.; Krylov, S. N. Anal Chem 2010, 82, 8692-8695. (40) Musheev, M. U.; Filiptsev, Y.; Krylov, S. N. Anal Chem 2010, 82, 8637-8641. (41) Haselberg, R.; de Jong, G. J.; Somsen, G. W. Anal Chem 2013, 85, 2289-2296. (42) Medina-Casanellas, S.; Dominguez-Vega, E.; Benavente, F.; Sanz-Nebot, V.; Somsen, G. W.; de Jong, G. J. J Chromatogr A 2014, 1328, 1-6.

505 22


Page 23 of 29

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


Table 1. Peak assignment and Kd values obtained by ACE-MS, ACE-UV and DIMS for trypsinogen sample components.

Peak

Mass

Tentative assignment

1’

14379.0 Da

not assigned

1’’

14505.0 Da

not assigned

1

23984.0 Da

2

Kd based on electrophoretic mobility shift ACE-MS ± s.d. ACE-UV ± (µM) s.d. (µM) no affinity n.d.

Kd based on relative MS signal intensity ACE-MS ± s.d. DIMS ± s.d. (µM) (µM) no affinity no affinity

no affinity

n.d.

no affinity

triply deamidated trypsinogen

4±2

n.d.

25 ± 37

23983.0 Da

doubly deamidated trypsinogen

4±2

3±1

3

23982.0 Da

singly deamidated trypsinogen

9±4

4±2

43 ± 22

4

23981.0 Da

trypsinogen

8±5

6±3

92 ± 36

5

23963.0 Da

not assigned

3±2

2±1

104 ± 71

no affinity

20 ± 17 96 ± 34*

59 ± 20

s.d., standard deviation n.d., not determined *overall value for components 1-4.

23



Figure 1

Relative intensity aprotinin-trypsinogen complex (%)

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

Page 24 of 29

100 90 80 70 60

Quadrupole ion energy (eV) 3 5 10 15

50 40 30 20 10 0 5 7 10 15 Collision cell energy (eV)

20

24


Page 25 of 29

Figure 2

Intensity (x104 cnts)

4

A

8

3

6 2 4 EIE 2398 ± 2 m/z

1’

EIE 1814.1 ± 0.5 m/z

2

5

1

1’’

EIE 1798.5 ± 0.5 m/z EIE 361.2 ± 0.5 m/z (EOF marker) 0

2

4

6

8

10

12

14

16

18

time (min)

4


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


B 4

3 3 2

2

1

0

EIE 1814.1 ± 0.5 m/z

1’

EIE 1798.5 ± 0.5 m/z EIE 361.2 ± 0.5 m/z (EOF marker)

1’’

2

4

5

1

EIE 2398 ± 2 m/z

6

8

10

12

14

16

18

time (min)

25



Figure 3

A

B

X10-5 1.5

Icomplexed protein /Ifree protein

Effective electrophoretic mobility shift (cm2V-1s-1)

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

Page 26 of 29

1.0

0.5

0.0 0

100

200

300

Concentration aprotinin in BGE (µ µM)

Peak 1 Peak 2 Peak 3 Peak 4 Peak 5

1.5

1.0

0.5

0.0 0

100

200

300

Concentration aprotinin in BGE (µ µM)

26


Page 27 of 29

Figure 4

+10 2399.0

100

A

75

+9 2665.5

50 25

B

100 75

Relative abundance (%)

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


50

+12 2543.4

+11 2774.5

25

↓

↓

C

100 75 50

↓

↓

25

D

100

↓

75

↓

50 25 ↓

100

E ↓

75 50 25 2200

2300 2400 2500 2600 2700 2800 m/z

27



Figure 5

A1

2.5 2.0 4

1.5

3

1.0

2 1

0.5 0

B1

3.0 5



3.0

2.5

3

2.0 1.5 1.0 1 2

0.5

10

20

30

0

40

10

20

40

30

B2

1.5 3 1.0


A2

2.0

4

time (min)

time (min)


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

Page 28 of 29

5

2 4

0.5

1

2.0 1.5

3

1.0

12

4’’

4’

0.5

2

0

10

20

30

40

0

time (min)

10

20

30

40

time (min)

28


Page 29 of 29

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


85x47mm (300 x 300 DPI)


Simultaneous Assessment of Protein Heterogeneity and Affinity by

Recommend Documents