Accepted Manuscript
Title: Solid Phase Microextraction fills the gap in tissue
sampling protocols
Author: Barbara Bojko Krzysztof Gorynski German
Gomez-Rios Jan Matthias Knaak Tiago Machuca Vinzent
Nikolaus Spetzler Erasmus Cudjoe Michael Hsin Marcelo
Cypel Markus Selzner Mingyao Liu Shaf Keshjavee Janusz
Pawliszyn
PII:
DOI:
Reference:
S0003-2670(13)01124-0
http://dx.doi.org/doi:10.1016/j.aca.2013.08.031
ACA 232789
To appear in:
Analytica Chimica Acta
Received date:
Revised date:
Accepted date:
18-6-2013
12-8-2013
17-8-2013
Please cite this article as: B. Bojko, K. Gorynski, G. Gomez-Rios, J.M. Knaak, T.
Machuca, V.N. Spetzler, E. Cudjoe, M. Hsin, M. Cypel, M. Selzner, M. Liu, S.
Keshjavee, J. Pawliszyn, Solid Phase Microextraction fills the gap in tissue sampling
protocols, Analytica Chimica Acta (2013), http://dx.doi.org/10.1016/j.aca.2013.08.031
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
- In vivo SPME sampling was used during lung and liver transplantation.
- Selection of the probe, transportation, storage conditions and analyte coverage
were discussed.
Ac
ce
pt
e
d
M
an
us
cr
ip
t
- Organ preservation procedures were investigated using principal component
analysis.
Page 1 of 21
us
cr
ip
t
Barbara Bojko 1 , Krzysztof Gorynski 1 , German Gomez-Rios 1 , Jan Matthias
Knaak 2 , Tiago Machuca 3 , Vinzent Nikolaus Spetzler 2 , Erasmus Cudjoe 1 ,
Michael Hsin 3 , Marcelo Cypel 3 , Markus Selzner 2 , Mingyao Liu 3 , Shaf
Keshjavee 3 and Janusz Pawliszyn 1
1Department
of Chemistry, University of Waterloo, Waterloo ON, Canada
of Surgery, Multi Organ Transplant Program, Toronto General
Hospital, Toronto, ON, Canada
3Latner Thoracic Surgery Research Laboratories, Toronto General Research
Institute, University Health Network and Department of Surgery, University of
Toronto, Toronto, ON, Canada
an
2Department
Keywords: in vivo SPME, metabolomics, tissue sampling, surgery
d
M
Abstract
Metabolomics and biomarkers discovery are an integral part of bioanalysis.
However, untargeted tissue analysis remains as the bottleneck of such studies due
to the invasiveness of sample collection, as well as the laborious and timeconsuming sample preparation protocols. In the current study, technology
integrating in vivo sampling, sample preparation and global extraction of
metabolites – solid phase microextraction was presented and evaluated during liver
and lung transplantation in pig model. Sampling approaches, including selection of
the probe, transportation, storage conditions and analyte coverage were discussed.
The applicability of the method for metabolomics studies was demonstrated during
lung transplantation experiments.
pt
e
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
ce
2
Solid Phase Microextraction fills the gap
in tissue sampling protocols
Introduction
Tissue analysis is one of the most challenging, labor-extensive and time-consuming
analytical tasks. It is also less frequently utilized when compared to blood, plasma,
urine, saliva or stool analysis because of the difficulty in accessing the sample.
However, the abovementioned easily accessible matrices carry only global
information about the condition of the organism. In order to obtain organ-specific or
spatially resolved data (e.g. tumor vs. healthy tissue, different organ regions), an
analysis of tissue sample is necessary.
The area of metabolomics studies has gathered increasing interest over the past
years, becoming an integrated part of the ‘-omics’ family. In particular, a rising
interest in tissue analysis for global screening of small molecules and biomarkers
discovery has been observed. While the use of nuclear magnetic resonance (NMR)
voids the need for a sample preparation step, mass spectrometry coupled to
Ac
1
Page 2 of 21
ce
pt
e
d
M
an
us
cr
ip
t
different separation techniques (liquid or gas chromatography, capillary
electrophoresis) requires effective sample clean-up and extraction of analytes.
Several articles providing tissue preparation protocols and comparisons between
the effectiveness of these extraction techniques can be found in the literature [1–4].
Generally, all these methods are based on removal of a piece of tissue (few
milligrams to few grams) or the entire organ (depending on application; human or
animal, in vivo or post mortem study etc.), followed by tissue disruption,
centrifugation and solvent extraction. The complexity of the subsequent steps
depends on the procedure used, such as various modifications of bead beating,
grinding, homogenization, as well as various approaches to solvent addition (e.g.
stepwise, two-step, all-in-one). The use of different solvents and different
configurations of solvents results in different analyte coverage. While this precludes
comparison of data originating from samples subjected to different analytical
protocols, on the other hand, it also complements the information, allowing
flexibility in the selection of the appropriate tools for a given application.
Considering the ultimate goal of metabolomics, analysis of the entire metabolome,
the chosen method should yield the best possible coverage. However, other factors
also need to be taken into consideration; in the recent protocol developed for global
metabolite profiling of animal and human tissues via UPLC-MS [4], a variety of
issues were brought up, such as losing labile metabolites during the period between
sample collection and preservation, shortage of material collected for multiple
diagnostic purposes, and the heterogeneity of the sample, among others.
An alternative to the invasive sample collection-based methods is microdialysis
(MD); however, MD is usually used for targeted analysis only, due to limitations in
the extraction of hydrophobic species as well as severe matrix effects, prohibiting
non-targeted determinations. Recent metabolomics analyses conducted with the use
of MD confirmed the biased nature of this methodology [5, 6]. Another setback with
MD analysis is timing: it is usually required to give patients time to recover after a
probe implantation. In addition, the MD probe must be constantly connected to
tubing and supplied with power, making it inconvenient for some applications.
Solid phase microextraction (SPME) is an analytical technique already validated in
many different areas of bioanalysis, including ex vivo as well as in vivo sampling for
targeted and untargeted studies [7–10]. Particularly for in vivo sampling, the
technology offers some unique features, such as a lack of sample collection and
extraction of unstable species [11–13]. It should be emphasized that by using one of
the many calibration approaches available, fully quantitative data is ensured [14],
while the wide variety of extraction phase chemistries permits selection of the most
suitable phase for a specific application. The applicability of in vivo SPME for blood
sampling was reported for pharmacokinetics [15–18] and metabolomics study [19],
while direct tissue extraction so far has mainly focused on the determination of
pharmaceutical concentrations in fish muscle, and neurotransmitter and drug levels
in rat brain [20]. The availability of biocompatible SPME fibers opens the door for
low-invasive, sample-free tissue analysis, also in the view of global metabolomics. In
this study, we would like to present for the first time the use of SPME for untargeted
direct extraction from lung and liver performed in a pig model.
Ac
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
Page 3 of 21
us
cr
ip
t
Experimental
Materials and methods
Acetonitrile, methanol and water (all LC/MS Optima grade) were purchased from
Fisher Scientific (Ottawa, Canada). Prototypes of biocompatible SPME mix-mode
probes (C18 with benzenesulfonic acid, 45 µm thickness, 4, 7 and 15 mm length of
coating) were provided by Supelco (Bellefonte, PA, USA). The standards used for
instrumental QC (test mixture) preparation were purchased from Sigma-Aldrich (Dphenylalanine, L-tryptophan, cholic acid, deoxycholic acid) and CDN Isotopes (Lphenyl-d5-alanine).
Normothermic Ex Vivo Liver Perfusion (NEVLP) and Normothermic Ex Vivo Lung
Perfusion (EVLP)
an
Information about Normothermic Ex Vivo Liver Perfusion (NEVLP) and Ex Vivo
Lung Perfusion (EVLP) procedures can be found in [21] and [22], respectively.
pt
e
d
M
In vivo extraction from liver and lung
Prior to use, all fibers underwent preconditioning for a minimum of 60 min
exposure to methanol:water 1:1 v/v. Additionally, for the lung sampling,
sterilization of the fibers was conducted by 30 min steam autoclaving at 121 oC. The
preconditioning mixture (methanol:water 1:1 v/v) was also sterilized by filtering,
using 0.45 µm pore size sterile syringe filters. The extraction time was 20 min for
the lung, 30 min for the liver. Immediately after sampling, fibers were rinsed with
deionized water, dried with Kimwipe and placed on dry ice in the empty vials. After
being transported to the laboratory, all fibers were placed at -80 oC until analysis.
ce
Evaluation of transportation and storage condition
In order to evaluate storage and transportation conditions, nine 7 mm fibers were
divided into three groups, and each group was subjected to a different protocol. Six
of the fibers were used for in vivo extraction from the liver; immediately after the 30
min exposure to the tissue, fibers were removed from the organ, quickly rinsed with
deionized water and gently dried with Kimwipe. Three fibers were placed in the
empty 0.3 mL polypropylene vials and three others in identical vials containing 0.3
mL desorption solvent (acetonitrile:water, 1:1 v/v). All vials were placed on dry ice
and transported to the laboratory, where they were stored at -80 oC for two weeks.
For comparative analysis, a piece of liver was collected at the beginning of the in
vivo extraction and immediately placed in dry ice together with the fibers used for
the in vivo study. Similarly to the fibers, the liver tissue was kept frozen at -80 oC for
two weeks, and then thawed on ice and used for 30 min SPME extraction using the
last three fibers. Desorption was performed simultaneously for all nine fibers; the
desorption time was 90 min with 1000 rpm vortex agitation (model DVX-2500,
VWR International, Mississauga, ON, Canada). Extracts were further injected to the
LC-MS system for analysis.
Ac
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
Page 4 of 21
Liquid chromatography-mass spectrometry analysis (LC/MS)
The samples were analyzed by reverse phase liquid chromatography (Accela,
Thermo Scientific) coupled to bench top orbitrap mass spectrometer (Exactive,
Thermo Scientific). The details and conditions of the LC/MS method can be found
elsewhere [19, 23]
M
an
us
cr
ip
t
Data processing and statistical analysis
The raw data files .raw obtained from Thermo Xcalibur were converted to .mzXML
using MSConvert software. Subsequently, MZmine software version 2.10 was used
for data processing. For filtering, the Savitzky-Golay algorithm was used, and mass
detection was conducted with exact mass method using 5e 3 noise level. For the
chromatogram builder m/z tolerance of 0.001 or 5 ppm was used. The
chromatogram deconvolution was performed using the Savitzky-Golay method and
for the alignment, the ‘join aligner’ option was selected. After data processing, the
first minute (0-1 min) corresponding to column void volume and the last five
minutes of column re-equilibration (35-40 min) were excluded from the analysis.
Given chromatograms were subjected to manual peak picking and all signals with
unacceptable shape, as well as those present in blank were removed from the
analysis. The final dataset containing signal intensities for all molecular features
defined by accurate mass and retention time were exported to .cvs file and used for
statistical analysis using SIMCA-P software version 12.0.1. The Pareto scaling was
applied in principal component analysis and accurate masses were used as variables
ce
pt
e
d
Results and discussion
Tissue sampling with SPME fibers
The sampling approach involved direct insertion of biocompatible mix mode fibers
to liver and lung. The fibers were assembled in a 24 ga hypodermic needle, which
served as guide cannula when placing the fiber inside the sampling area. Figure 1
presents the sampling of liver and lung performed in pig model during organ
transplantation.
Ac
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
Page 5 of 21
t
us
cr
ip
Figure 1. SPME sampling of pig liver (A) and lung (B) with biocompatible mix-mode
fibers
Ac
166
167
168
169
170
171
172
173
174
175
176
177
178
179
ce
pt
e
d
M
an
164
165
Effects of fiber length on sampling efficacy
The length of the fiber coating can be adjusted to the requirements of the particular
application. To achieve the best sensitivity of the method, the longest possible
coating should be used; however, the length of the coating should ensure good
spatial resolution, and coating should be immersed in the desired sampling area
only. For instance, the fiber length cannot extend beyond the studied region for
organs with structures that are not homogenous, or anatomically divided into
sections, lobs or surfaces such as lungs, liver or kidneys, or if sampling needs to be
limited to altered tissue (e.g. malignant, infected, inflammatory). Also, it should be
Page 6 of 21
t
us
cr
ip
an
M
d
pt
e
ce
207
209
211
213
215
217
219
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
remembered that in order to achieve good reproducibility, the entire length of the
coating must be covered by the tissue matrix. Taking into consideration the need for
different coating lengths for different applications, mix mode fibers with three
different coating lengths were tested during liver sampling. The extraction time
used for all three coating types was 30 min. As expected, the 15 mm coating gave the
highest sensitivity and coverage and the 4 mm coating the lowest (Fig. 2).
Conversely, fibers with 4 and 7 mm coating lengths provided better reproducibility
when compared with the 15 mm coating, which could be related to the structure of
the liver. As a compromise between sensitivity and reproducibility, probes with 7
mm extraction phase were selected for further studies.
Figure 2. The principal component analysis showing clustering of three types of
fibers used for liver sampling: 4, 7 and 15 mm coating length.
Ac
180
181
182
183
184
185
186
187
188
189
190
191
192
193
195
197
199
201
203
Sample preservation and transportation
When using standard methods of tissue analysis, how samples are handled after
tissue collection is a major concern. In any case, the collected tissue should be
placed on ice, preventing loss of unstable analytes by slowing down tissue
metabolism. If the sample cannot be immediately processed, it has to be snap frozen
at a minimum of – 20 oC, and preferably at -80 oC. To further minimize the progress
of metabolic reactions, and to accelerate freezing of the samples, it is recommended
that obtained tissue is cut into small pieces, and stored this way. This procedure also
helps avoid multiple freezing and thawing of the same sample, which can also
influence results. For in vivo SPME, direct sampling of the metabolites from the
Page 7 of 21
d
M
an
us
cr
ip
t
living system voids the need for tissue collection. Because of the biocompatible
nature of the coating, which restricts access for large molecules such as proteins
(enzymes) to the coating, the extracted analyte is protected from enzymatic
degradation. This feature allows analysis of short-lived species, and has been
already discussed elsewhere for blood sampling [19]. This sample-free sampling
gives us different options for transportation and storage of the analytes, either
leaving analytes on the fiber until analysis, or transporting and storing as an extract.
In both cases, the fiber needs to be quickly rinsed with deionized water immediately
after sampling. In the first scenario, after the fiber is rinsed, it should be placed on
ice securely hidden in the needle assembly. It can be transported and stored in such
way at a low temperature (-30 or -80 oC). The second option involves desorption on
site, with transportation and storage of the extract conducted in sealed vials.
However, a principal component analysis plot (Fig. 3) shows that data points
corresponding to three different extractions, transportation and storage approaches
do not cluster, showing that there is an influence of the used protocol on the final
results. The orientation of the clusters indicates that both principal components,
PC1 and PC2, influence the separation of the samples, suggesting that different
factors are responsible for the changes in the metabolic profile. The differences
observed between the two in vivo sets of samples (followed by transport of the
analytes on fiber and transport of the analytes in the extract after on-site
desorption) can be explained by the partial evaporation of the desorption solvents
and the preconcentration of the analytes in the extract with respect to in-lab
desorption.
277
280
281
282
283
284
285
286
Ac
ce
pt
e
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
262
Figure 3. The influence of storage and transportation condition on PCA clustering;
blue diamonds – ex vivo extraction from the lung tissue stored at -80 oC for two
weeks; green triangles – in vivo extraction and on-fiber transportation and storage
for two weeks; red dots – in vivo extraction, desorption on-site, transport and
storage of extract.
Page 8 of 21
an
us
cr
ip
t
The differences between ex vivo and in vivo sampling are observed in the loss of fast
turnover compounds in the case of the former approach, as reported previously for
blood sampling [19]. Moreover, degradation products can be produced in the tissue
after sample collection and before metabolism quenching, thus not being present in
the in vivo extracts. Tentative identification of the compounds differentiating in vivo
sampling, followed by on-fiber transport from ex vivo sampling of stored tissue,
suggested the presence of several intermediate metabolites, such as itaconic acid,
hypoxanthine, lactaldehyde or hydroxyacetone. The chromatogram and
corresponding spectra with masses assigned as different adducts of lactaldehyde or
hydroxyacetone and hypoxantine are presented in Figure 4. On the other hand, in
the extracts obtained from ex vivo lung sampling, the dominant compounds, when
compared with in vivo samples, were diacylglycerol. Their presence can be
explained by the remaining enzymatic activity of phospholipase, and the
degradation of the cell membrane. However, due to the fact that in global
metabolomics, the analysis of unknown compounds does not permit prior
evaluation of the stability of all metabolites of interest, the transportation/storageon-fiber protocol seems to be the most convenient method.
ce
Figure 4. Chromatograms and corresponding mass spectra indicating masses
assigned as adducts of lactaldehyde or hydroxyacetone (A) and hypoxantine (B)
Effects of positive vs. negative ionization mode of molecular analysis
The analyses of molecular features detected in positive and negative ionization
mode after 20 min extraction from liver and 30 min extraction from lung, using 7
mm mix-mode coating and separation on pentafluorophenyl column, are shown in
Fig. 5A and C and 5B and D, respectively. Only signals that were present in at least
two out of three replicates of the sample were taken into account,. For the liver
sampling, the number of molecular features detected in negative mode was 311, and
ranged between 100.07596 and 971.44040 m/z, while in positive, the number was
1580, and ranged between 80.96398 and 947.44186 m/z. Corresponding results
obtained for the lung study showed the number of molecular features detected in
negative and positive mode were 239 and 1041, respectively. The masses ranges for
negative and positive modes were 86.948939 - 781.25821 and 110.5100 -
Ac
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
pt
e
d
M
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
Page 9 of 21
ce
pt
e
d
M
an
us
cr
ip
t
997.77161, respectively. The range of detected metabolites support the observation
that SPME provides a good and balanced coverage, as shown in the previous
metabolomics study [19]. However, the currently used coating does not give as wide
a yield of metabolites as solvent methods. This can be addressed by implementing
alternative extraction phases based on SPE particles i.e. hydrophilic interaction
liquid chromatography (hilic) , hydrophilic-lipophilic balance (HLB), or polystyrene
divinylbenzene (PS-DVB), which can go with polarities as low as logP -3 and extract
sugars or peptides. In terms of hydrophobic compounds, the limit is primarily
dependent on the concentration of free molecules in the studied matrix. It should be
noted that an SPME fiber interacts only with free molecules; its coating does not
pick up the bound fraction of the analyte. Despite the fact that the percentage of
bound hydrophobic metabolites is usually high, because of poor water solubility, the
sorbents used for SPME extraction have a high affinity towards these compounds,
therefore providing good extraction recovery. However, the determination of free
concentrations has an important advantage over the total concentration extraction
given by solvent-based extraction methods, since the free portion exhibits biological
activity, and the information taken from SPME directly correlates to the actual
response of the organ. It is worth mentioning that SPME can go beyond untargeted
analysis of small molecules, and offer very selective extraction if a molecularly
imprinted polymer or aptamer based coating is used. This should be considered as a
useful tool for monitoring biomarkers, which is the ultimate goal of metabolomics
study. By providing additional information not accessible from standard techniques,
in vivo SPME may serve as a complementary or alternative method, particularly
when sample collection is restricted, or when spatial or temporal resolution is
required. The preconcentration of the analytes on the fiber and their desorption in a
small volume ensures good sensitivity of the method (e.g. intensity for the presented
liver study was within the range of 5e3 to 7.5e7 for positive mode and 5.3e3 to 3.7e6
in negative mode). The total time “from sampling to extract” would take 2-3 hours
with the current experimental setup, versus over 15-20 hours of standard solvent
based approaches, including 3-hour and overnight dryness of aqueous and organic
extracts, respectively [3].
353
Ac
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
Page 10 of 21
t
us
cr
ip
an
M
pt
e
d
Figure 5. Ion map (m/z versus retention time) for lung (A, C) and liver (B, D)
extracted with 7 mm mix-mode fiber during in vivo exposure followed by LC-MS
analysis in positive (ESI +) (A and B) and negative (ESI -) (C and D) ionization mode.
ce
Detection of metabolites after different lung preservation condition
The in vivo SPME-LC/MS study, followed by multivariate principal component
analysis employed to the liver and lung transplantation experiment, allowed us to
distinguish the clinically relevant separation of the investigated individuals, and
indicate metabolites resolving the clustered data. Figure 6 presents the example of
PCA plot for the lungs subjected to two different organ preservation procedures.
The cluster on the left side of the plot represents a standard preservation method
(Cold Standard Preservation after flush with perfadex), and the cluster on the right
represents the samples collected from the lung after ex vivo perfusion (EVLP). It can
be noticed that the first principal component distinguishes differences between the
two preservation methods, while the second characterizes changes between
different samples obtained during the certain experiment.
Ac
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
Page 11 of 21
an
Figure 6. PCA showing clustering of the samples collected under different clinical
conditions of the lung (ESI -)
ce
pt
e
d
M
Conclusions
The results of the present study indicate that in vivo SPME can be successfully used
as an alternative approach to tissue sampling. The method offers enrichment, low
invasiveness, balanced yield and the possibility of capturing labile compounds.
Further, development of the extraction phases will not only extend the range of
extracted analytes and make the technology more attractive for global screening,
but also increase the chance of extraction biomarkers directly from the studied
tissue by designing highly selective coatings (i.e. molecularly imprinted polymers,
aptamers). The mentioned feature of low invasiveness and good spatial resolution
provided by SPME fibers permits repeated sampling from the tissue and
consequently makes the method suitable for monitoring of drug and biomarkers
distribution and concentration over the period of time. This application of in vivo
SPME can be also found as a useful tool in the research area, especially in animal
studies, where minimum invasiveness during repeated sampling is a major concern.
Although time resolution associated with relatively long exposure of the probe to
the organ for optimum sensitivity for metabolomics remains the main drawback of
the approach, it may be improved for targeted analysis, where extraction time can
be optimized prior to the experiment. With its given simplicity and solventless
nature, this technique can be an attractive solution for on-site analysis. For
diagnostic purposes, ongoing studies are being carried out in our research group for
improvement of previously reported direct couplings of SPME with mass
spectrometer [24–29] for rapid bedside or intrasurgical analysis.
Ac
386
388
390
392
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
us
cr
ip
t
374
Reference
Page 12 of 21
423
424
[1] P. Masson, A.C. Alves, T.M.D. Ebbels, J.K. Nicholson, E.J. Want, Anal. Chem. 82
(2010) 7779–7786
426
427
[3] E.J. Want, P. Masson, F. Michopoulos, I.D. Wilson, G. Theodoridis, R.S. Plumb, J.
Shockcor, N. Loftus, E. Holmes, J.K. Nicholson, Nat. Protoc. 8 (2013) 17–32.
t
428
[2] F.M. Geier, E.J. Want, A.M. Leroi, J.G. Bundy, Anal. Chem. 83 (2011) 3730–3736.
[4] H. Wu, A.D. Southam, A. Hines, M.R. Viant, Anal. Biochem. 372 (2008) 204–212.
us
cr
ip
425
429
430
[5] C. Wibom, I. Surowiec, L. Mörén, P. Bergstro, M. Johansson, H. Antti, A.T.
Bergenheim, J. Proteome. Res. 9 (2010) 2909–2919.
433
[7] H. Kataoka, K. Saito, J. Pharm. Biomed. Anal. 54 (2011) 926–950.
440
441
442
443
444
an
M
[10] B. Bojko, E. Cudjoe, M. Wasowicz, J. Pawliszyn, TrAC Trends Anal. Chem. 30
(2011) 1505–1512.
d
438
439
[9] S. Risticevic, V.H. Niri, D. Vuckovic, Anal. Bioanal. Chem. 393 (2009) 781–795.
pt
e
437
[8] B. Bojko, E. Cudjoe, G. Gómez-Ríos, K. Gorynski, R. Jiang, N. Reyes-Garcés, S.
Risticevic, É. S. Silva, O. Togunde, D. Vuckovic, J. Pawliszyn, Anal. Chim. Acta. 750
(2012) 132–151.
[11] D. Vuckovic, S. Risticevic, J. Pawliszyn, Angew. Chem. Int. Ed. 50 (2011) 5618–
5628.
[12] D. Vuckovic, Bioanalysis, 3 (2011) 1305–1308.
ce
434
435
436
[6] O. Hrydziuszko, M.A Silva, M.T. Perera, D.A Richards, N. Murphy, D. Mirza, M.R.
Viant, OMICS, 14 (2010) 143–150.
[13] G. Ouyang, D. Vuckovic, J. Pawliszyn, Chem. Rev. 111 (2011) 2784–2814.
[14] G. Ouyang, J. Pawliszyn, Anal. Chim. Acta. 627 (2008) 184–197.
Ac
431
432
445
446
[15] H.L. Lord, R.P. Grant, M. Walles, B. Incledon, B. Fahie, J.B. Pawliszyn, Anal. Chem.
75 (2003) 5103–5115.
448
449
[17] F.M. Musteata, I. de Lannoy, B. Gien, J. Pawliszyn, J. Pharm. Biomed. Anal. 47
(2008) 907–912.
447
450
451
[16] F.M. Musteata, M.L. Musteata, J. Pawliszyn, Clin. Chem. 52 (2006) 708–715.
[18] D. Vuckovic, I. De Lannoy, B. Gien, Y. Yang, F.M. Musteata, R. Shirey, L. Sidisky, J.
Pawliszyn, J. Chromatogr. A, 1218 (2011) 3367–3375.
Page 13 of 21
452
453
[19] D. Vuckovic, I. de Lannoy, B. Gien, R.E. Shirey, L.M. Sidisky, S. Dutta, J. Pawliszyn,
Angew. Chem. Int. Ed. 50 (2011) 5344–5348.
455
456
457
[21] M.U. Boehnert, J.C. Yeung, F. Bazerbachi, J.M. Knaak, N. Selzner, I.D. McGilvray,
O.D. Rotstein, O.A. Adeyi, S.M. Kandel, P. Rogalla, P.M. Yip, G.A. Levy, S. Keshavjee,
D.R. Grant, M. Selzner, Am. J. Transplant. 13(2013) 1441-1449.
t
[20] E. Cudjoe, B. Bojko, P. Togunde, J. Pawliszyn, Bioanalysis, 4 (2012) 2605–2619.
us
cr
ip
454
458
459
460
461
[22] M. Cypel, J.C. Yeung, M. Liu, M. Anraku, F. Chen, W. Karolak, M. Sato, J. Laratta, S.
Azad, M. Madonik, C.-W. Chow, C. Chaparro, M. Hutcheon, L.G. Singer, A.S. Slutsky, K.
Yasufuku, M. de Perrot, A.F. Pierre, T.K. Waddell, S. Keshavjee, New Engl. J. Med. 364
(2011) 1431–4140.
463
464
[24] Y. Wang, M. Walles, B. Thomson, S. Nacson, J. Pawliszyn, Rapid Commun. Mass
Spectrom. 18 (2004) 157–162.
466
467
[26] S. Perera, A. Berthod, E. Dodbiba, D.W. Armstrong, Rapid Commun. Mass
Spectrom. 26 (2012) 853–862.
469
470
[28] M. Walles, Y. Gu, C. Dartiguenave, F.M. Musteata, K. Waldron, D. Lubda, J.
Pawliszyn, J. Chromatogr. A, 1067 (2005) 197–205.
471
472
473
474
475
476
477
478
479
an
M
d
[27] J.K. Lokhnauth, N.H. Snow, J. Sep. Sci. 28 (2005) 612–618.
pt
e
468
[25] F.S. Mirnaghi, J. Pawliszyn, Anal. Chem. 84 (2012) 8301–8309.
[29] H. Tong, N. Sze, B. Thomson, S. Nacson, J. Pawliszyn, The Analyst, 127 (2002)
1207–1210.
ce
465
[23] V. Bessonneau, B. Bojko, J. Pawliszyn, Bioanalysis, 5 (2013) 783–792.
Acknowledgment
The authors want to acknowledge the Natural Sciences and Engineering Research
Council (NSERC) of Canada and Supelco (Sigma-Aldrich) for financial support. The
authors express their sincere gratitude to Dr. Marcin Wasowicz for his help in
establishing the collaboration between Toronto General Hospital and University of
Waterloo. We also thank Ms. Nathaly Reyes-Garcés for her help in SPME sampling.
Ac
462
Page 14 of 21
Ac
ce
pt
ed
M
an
us
c
Page 15 of 21
Ac
pt
a
M
ed
ce
Page 16 of 21
Ac
ce
pt
ed
M
an
us
c
Page 17 of 21
Ac
ce
pt
ed
M
Page 18 of 21
an
us
c
Ac
ce
pt
ed
M
an
us
cr
ip
t
Page 19 of 21
Ac
ce
pt
ed
M
an
us
cr
ip
t
Page 20 of 21
Ac
ce
pt
ed
M
an
us
cr
ip
t
Page 21 of 21