Subscriber access provided by Universiteit Utrecht
Review
Food Forensics: Using Mass Spectrometry to Detect Foodborne Protein Contaminants, as Exemplified by Shiga Toxin Variants and Prion Strains. Christopher J. Silva J. Agric. Food Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jafc.8b01517 • Publication Date (Web): 02 Jun 2018 Downloaded from http://pubs.acs.org on June 2, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 70
Journal of Agricultural and Food Chemistry
1
Food Forensics: Using Mass Spectrometry to Detect Foodborne Protein Contaminants, as
2
Exemplified by Shiga Toxin Variants and Prion Strains.
3 4
Christopher J. Silva*
5 6
Produce Safety & Microbiology Research Unit, Western Regional Research Center, United
7
States Department of Agriculture, Agricultural Research Service, Albany, California 94710,
8
United States of America.
9 10
* Corresponding Author (Tel: 510.559.6135; Fax: 510.559.6429; E-mail:
11
[email protected])
12 13 14 15 16 17 18 19 20 21 22 23
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
24
Page 2 of 70
Abstract
25 26
Food forensicists need a variety of tools to detect the many possible food contaminants. Due to
27
its analytical flexibility, mass spectrometry is one of those tools. Use of the multiple reaction
28
monitoring (MRM) method expands its use to quantitation as well as detection of infectious
29
proteins (prions) and protein toxins such as Shiga toxins.
30
inactivate prions and Shiga toxins; the proteins are digested with proteases to yield peptides
31
suitable for MRM-based analysis. Prions are detected by their distinct physicochemical
32
properties and by differential covalent modification. Shiga toxin analysis is based on detecting
33
peptides derived from the five identical binding B subunits comprising the toxin.
34
internal standards are prepared from cloned proteins. These examples illustrate the power of
35
MRM, in that the same instrument can be used to safely detect and quantitate protein toxins,
36
prions and small molecules that might contaminate our food.
37 38 39 40 41 42 43 44 45 46
ACS Paragon Plus Environment
The sample processing steps
15
N-labeled
Page 3 of 70
47
Journal of Agricultural and Food Chemistry
Introduction
48 49
As food comes from more diverse sources, in an ever-astonishing variety, and is
50
processed to widely varying extents, the possibility for deliberate or inadvertent food
51
manipulation or contamination also increases. As a result, food forensicists are faced with
52
substantial and evolving challenges. A common form of manipulation is economic substitution
53
of more expensive food with less expensive, but similar food. Seafood harvested from around
54
the world and often sold to consumers in locations far removed from its harvest is sometimes
55
mislabeled accidentally or deliberately.1
56
species, such as salmon, is both farmed and wild-harvested. In Europe, marginally cheaper horse
57
meat has been mislabeled as modestly more expensive beef or pork.2 Olive oil is sometimes
58
adulterated with other cheaper vegetable oils.3-4 It is not surprising that food forensicists would
59
need an equally diverse set of tools to test for food substitution.5
Accurate labeling is further complicated when a
60
Adulteration for unintentional, natural or economic reasons can cause more serious
61
problems. In China, melamine (Figure 1) was used to confound the Kjeldahl-based protein
62
analysis to make milk appear more concentrated than it actually was.6-8 Turmeric produced in
63
India or Bangladesh and sold in the United States was discovered to be adulterated with lead (II)
64
chromate to enhance the color of and increase the weight of the product.9 Ciguatoxin 1B (Figure
65
1) is an example of a large and structurally complicated marine toxin that may naturally
66
contaminate a variety of seafoods.10 In addition to specific chemicals, food may be contaminated
67
with other small molecule toxins, infectious proteins (prions), protein toxins or bacteria that
68
produce toxins.11-13 The variety of possible contaminants means that no universal analytical
69
method can be used by a forensicist to detect all possible contaminants.5
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 4 of 70
70
The cost of food adulteration or contamination is substantial, both in terms of human
71
suffering and economic losses. The consequences of melamine adulteration to the Chinese dairy
72
industry and the Chinese economy in general have been substantial.14 The human consequences
73
are greater; approximately 294,000 children were sickened, 50,000 required hospitalization and
74
at least 6 died.14 In the United States approximately 175,000 cases of Shiga toxin-producing E.
75
coli (STEC) occurred yearly from 2000-2008.15 The economic cost of a single outbreak in 2006
76
was estimated to be $200 million (2006 $US).16 The STEC outbreak that occurred in 1992 cost
77
the company $160 million in lost sales (1992 $US) and a 30% drop in its stock price.17 More
78
than 8300 school children were affected by the 1996 STEC outbreak in Sakai, Japan.18 In 2011,
79
an STEC outbreak in Germany sickened more than 3800 people.19
80
encephalopathy (BSE), a prion disease of domestic cattle, devastated British agriculture.20 Its
81
human form, variant Creutzfeldt-Jakob disease (vCJD), has killed at least 230 people.21 It is
82
clear that in addition to being challenging, food forensics is crucial to human health.
83
Bovine spongiform
In practice, mass spectrometry is a multiplex analytical tool that is used to detect an
84
astonishingly diverse range of contaminants.11
85
contaminants, such as melamine and a suite of pesticide residues.22
86
aflatoxin B1, microcystin-LR, tetrodotoxin, ciguatoxin 1B, and other shellfish toxins (Figure 1)
87
are also readily detectable by mass spectrometry.13, 23-26 The development of newer ionization
88
methods has permitted the detection of whole proteins by mass spectrometry. Sector instruments
89
allow for the “filtering” of molecules based on their molecular weights. The multiple reaction
90
monitoring (MRM) method exploits these properties to allow the detection and quantitation of
91
small amounts of specific small molecules in complex mixtures.27 The MRM method is not
92
suited for the direct analysis of proteins, but it can be used to analyze peptides derived from
It has been used to detect small molecule
ACS Paragon Plus Environment
Mycotoxins, such as
Page 5 of 70
Journal of Agricultural and Food Chemistry
93
those proteins after they have been digested with proteases, such as trypsin or chymotrypsin.
94
This also provides greater confidence that a protein has been correctly identified when more than
95
one peptide derived from it can be detected. Mass spectrometry can potentially provide more
96
information about potential food contaminants than any other method.
97
Prions represent a novel detection challenge, since they are infectious proteins.28 The
98
research standard for prion detection is the bioassay.29 It can be used to establish that the sample
99
material is infectious and that it causes the lesions characteristic of a prion disease.30 Incubation
100
times for prion diseases vary from two months to several years, which precludes using bioassay
101
as a routine detection method.29 Cell-based systems have been used to amplify prions, but not all
102
prions amplify in cell culture.31
103
hematoxylin and eosin stained fixed tissue slice, or the presence of prions stained by
104
immunohistochemistry, are also diagnostics used by regulatory agencies to demonstrate the
105
presence of prions and the damage that they cause in sheep and domestic cattle.
106
Unfortunately, pathological changes are only observable in the later stages of the disease.34-35
107
Western blotting is also used by regulatory agencies to rapidly diagnose prion diseases.32-33
108
Prions have been detected using prion amplification methods related to protein misfolding cyclic
109
amplification (PMCA).36 Not all prions amplify using these approaches and prion amplification
110
techniques also spontaneously produced infectious prions from bacterially derived recombinant
111
PrP (rPrP).37 Mass spectrometry remains the most sensitive means of directly detecting prions,
112
without the use of bioassay.38 In principle, mass spectrometry can also be used to detect the
113
prions amplified by techniques such as PMCA sooner, thereby permitting more rapid detection
114
of prions.
The presence of appropriate vacuolization (spongiform) in a
ACS Paragon Plus Environment
32-33
Journal of Agricultural and Food Chemistry
Page 6 of 70
115
Shiga toxins are large multi-component toxins produced by phage-infected bacterial food
116
contaminants. There are a variety of means to detect Shiga toxins to determine the source of
117
contamination.39 Regulatory agencies focus on the E. coli serotypes (vide infra) associated with
118
toxin production.40-41 Shiga toxins have been detected using animal bioassay and cell assay.42-45
119
While these methods are sensitive, they cannot distinguish among the possible Shiga toxins,
120
unless the molar amount of Shiga toxin present in the sample is already known. Shiga toxin
121
production is also inferred by the presence of specific virulence genes.46
122
methods have been widely used to detect and distinguish among the Shiga toxins.47-56 To detect
123
any new Shiga toxins using antibodies, new antibodies need to be developed. Top down mass
124
spectrometry is also used to detect Shiga toxins. 57-62 An MRM-based approach has advantages
125
over these other methods in that it can be used to both distinguish among types and variants of
126
Shiga toxins and quantitate them (vide infra).63-64
Antibody based
127
MRM can also be used to analyze proteins following digestion with proteases, such as
128
trypsin or chymotrypsin, to yield a characteristic set of small molecule peptides.27 By using this
129
approach, forensicists can apply mass spectrometry to protein toxins such as Shiga toxins, which
130
are enzymatic toxins.
131
pathogenic proteins whose pathology is enciphered in their conformation. Identifying Shiga
132
toxins represents a challenge to detect and distinguish among closely related known proteins and
133
to adapt that technique to the detection of novel Shiga toxins as they are discovered. In the case
134
of prions, the challenge is to detect and distinguish among protein conformations. How MRM
135
has been used to accomplish both goals is described.
In addition, they can use MRM to detect the presence of prions,
136 137
Multiple reaction monitoring (MRM) method
ACS Paragon Plus Environment
Page 7 of 70
Journal of Agricultural and Food Chemistry
138 139
The multiple reaction monitoring method is a powerful means of analyzing the peptides
140
derived from proteins.27 MRM requires the use of sector instruments, which are expensive and
141
not well suited for high throughput sample processing.
142
identification of suitable peptides, which necessitates an initial qualitative proteomic analysis to
143
identify peptides suitable for MRM-based analysis. In addition, synthetic peptides need to be
144
prepared and stable-isotope internal standards acquired (vide infra).
145
methods have the distinct advantage of greater sensitivity and flexibility. The MRM method
146
directly detects molecules, provides more information than is available by other means, and can
147
be used to simultaneously detect a variety of diverse molecules (multiplex).
148
power of triple quadrupole instruments to apply mass filters which permit selected ions to pass
149
on to subsequent sectors and thereby detect only ions that have the user defined properties, while
150
discarding the others (Figure 2). Modern sector instruments have cycling times (milliseconds)
151
that are much faster than chromatographic elution times (seconds). This means that the rapid
152
sequential analysis of multiple ions appears to be simultaneous as samples elute from the
153
column. In this way a stable isotope-labeled internal standard can be used to demonstrate that an
154
unknown molecule shares the same physicochemical properties, e.g., chromatographic retention
155
time and characteristic fragmentation, as the known stable isotope-labeled molecule. A cartoon
156
of the triple quadrupole mass spectrometer and the multiple reaction monitoring method is in
157
Figure 3. The one drawback to this method is that it detects small molecules and not proteins.
The MRM method requires the
However, MRM-based
It exploits the
158
Proteins, however, can be analyzed by MRM, provided that they are cleaved into shorter
159
peptides prior to analysis. This is usually accomplished by digestion with a protease, such as
160
trypsin. Such procedures are now automated. They involve the reduction of any disulfide bonds
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 8 of 70
161
with a reducing agent, such as dithiothreitol (DTT). The now free thiol groups (S-H) are reacted
162
with an alkylating agent, such as iodoacetamide, to yield the corresponding carbamidomethyl
163
group (S-CH2-CO-NH2). The covalent carbamidomethyl modification prevents the reformation
164
of disulfide bonds. The reduced and alkylated protein is then digested with a protease such as
165
trypsin to yield a set of tryptic peptides.
166
phosphines, have been used instead of DTT.66-67 Alkylating agents other than iodoacetamide can
167
be used covalently modify free thiols.68
168
variety of other proteases, such as chymotrypsin or thermolysin.69 Proteins can also be cleaved
169
by reagents such as cyanogen bromide. This variety of tools can be exploited to analyze specific
170
proteins.70
Other reducing agents, such as water soluble
Reduced and alkylated proteins can be cleaved by a
171
The resulting set of peptides is analyzed by a qualitative proteomic analysis to identify
172
peptides suitable for MRM analysis. Such peptides usually have molecular weights between 500
173
and 2,000. Those possessing the most intense signals are candidates for an MRM based analysis.
174
Physicochemical properties of the peptides need to be considered. For example, methionines can
175
be oxidized to their corresponding sulfoxides and N-terminal glutamines (and to a lesser extent
176
glutamate) can spontaneously form N-terminal pyroglutamates in solution.
177
transformations would result in the signal of a single peptide to be spread among two or more
178
different chemical analogs of that peptide.
Such chemical
179
Once a set of suitable peptides has been identified, the instrument parameters for the
180
triple quadrupole can be optimized for each peptide. The precursor ion is determined from the
181
qualitative analysis.
182
increasing the collision energy of the carrier gas and monitoring the resulting fragmentation of
183
the peptide. The values m/z of the most intense product ions become the mass filters for the third
The optimal fragmentation conditions are determined empirically by
ACS Paragon Plus Environment
Page 9 of 70
Journal of Agricultural and Food Chemistry
184
quadrupole.
This process is repeated for the remaining peptides.
This results in the
185
identification of an analyte peptide, which is the one with the most intense signal. The other
186
peptides can be used to confirm the presence of the progenitor protein. The protease digestion
187
yields a characteristic set of peptides, which can be used to quantitate and confirm the presence
188
of the progenitor protein, often in the attomole (10-18 mole) range.
189
Standard molecular biology techniques can be used to conveniently generate stable
190
isotope-labeled internal standards for the peptides optimized for MRM analysis.63, 71 A desired
191
gene is cloned into an appropriate vector and overexpressed in a suitable E. coli host. This
192
permits the convenient production of proteins and permits the generation of internal standards. If
193
the E. coli host is grown in minimal medium supplemented with an essential nutrient that is
194
highly enriched (99.7%) with 15N (e.g., such as 15NH4Cl), then all the nitrogen-bearing molecules
195
synthesized by that E. coli will also be highly enriched with
196
(14N) nitrogen. Digestion of purified proteins produced in this way would yield
197
peptides enriched with the 15N-label. The extent of isotopic incorporation is dependent upon the
198
isotopic purity of the 15NH4Cl and the number of nitrogens present in the peptide. Peptides used
199
in the work cited in this review contain between 10 and 21 nitrogens. Between 96 and 93%
200
(respectively) of such peptides would be expected to contain exclusively 15N, with the remainder
201
comprising molecules containing mostly 15N, save for one 14N. Thus, 15N-enriched peptides can
202
be used as internal standards to identify and quantitate the peptides present in a sample.
15
N instead of natural abundance 15
N-labeled
203
The mass filter settings will exclude the portion of molecules containing the various
204
natural abundance stable isotopes. The proportion of internal standard peptides or peptide
205
fragments containing one 14N will be excluded by these mass filters. Analyte peptides derived
206
from samples containing naturally occurring
13
C (~1% natural abundance) are similarly
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 10 of 70
207
excluded. The calibration curves establish an empirical relationship between synthetic analyte
208
peptides containing only
209
digestion of the isotopically (15N) enriched proteins, containing only
210
relationships are linear over a 100-fold range and have excellent correlation coefficients.
12
C and the internal standard peptides, derived from the protease 15
N. In practice, such
211 212
Prions
213 214
Prions are pathological proteins that cause transmissible spongiform encephalopathy
215
(TSE), a fatal neurological disease.28,
72
216
incubation, and a comparatively short symptomatic period, followed by death. The human TSEs
217
are Creutzfeldt-Jakob disease (CJD),73-74 kuru,75 Gerstmann–Sträussler–Scheinker (GSS)
218
disease,76 and fatal familial insomnia (FFI).77-78
219
spongiform encephalopathy (BSE) when they consume BSE-contaminated feed.79 BSE is the
220
only known zoonotic prion disease. The human form of BSE is called variant Creutzfeldt-Jakob
221
(vCJD) disease.80 Sheep scrapie is transmitted among domestic sheep and is the oldest known
222
TSE, first described in the middle of the 18th century.81-82 Transmissible mink encephalopathy
223
(TME) is a prion disease transmitted to farmed mink by contaminated feed; the last outbreak of
224
TME was in 1985.83 Chronic wasting disease (CWD) is a prion disease of cervids (white-tailed
225
deer, mule deer, elk, moose, European moose, red deer, and reindeer) and the only one naturally
226
transmitted among wild animals.84-86 More recently a camel prion disease (CPD) has been
227
discovered in Algeria that appears to be transmitted among domesticated camels in a manner
228
analogous to that of CWD and not through feed.87
TSEs are characterized by a long asymptomatic
Domestic cattle are infected with bovine
ACS Paragon Plus Environment
Page 11 of 70
229
Journal of Agricultural and Food Chemistry
Prions (PrPSc) cause disease by inducing a normal cellular prion protein (PrPC) to adopt
230
the prion conformation.71,
89-90
Both PrPC and PrPSc possess identical covalent structures and
231
only differ in their conformation.89 PrPC is a highly conserved monomeric protein of uncertain
232
function (Figure 4).91-92 It is soluble in non-denaturing detergents and its secondary structure is
233
comprised of α-helices, unstructured regions, and a small amount of β-sheet (Figure 5).93 In
234
contrast, the pathogenic (PrPSc) conformation is multimeric, insoluble in non-denaturing
235
detergents, and has no known biological function (Figure 5).28 A large body of spectroscopic
236
evidence supports the hypothesis that the secondary structure of PrPSc is composed of β-sheet
237
and unstructured regions, with no α-helices.94 Additional spectroscopic evidence suggests that
238
the β-sheet secondary structure is contained in a tertiary β-helical structure known as a four rung
239
β-solenoid.95-98
240
paired and intertwined protofilments.97 PrPC has no resistance to proteinase K (PK) while PrPSc
241
has some resistance.28 The difference in resistance to PK is presumed to be due to the β-sheet
242
secondary structure. Detecting and distinguishing among protein conformations is challenging.99
243
Although resistance to PK digestion is associated with PrPSc there is considerable
244
variation among the different strains and even among the different sizes of multimers in a given
245
prion strain.103-104 Prions can be differentially sedimented by ultracentrifugation into populations
246
of smaller multimers that are PK sensitive and larger multimers that are PK resistant.105 Even
247
though one fraction is PK resistant and the other fraction is PK sensitive, both are infectious and
248
possess the same primary, secondary, and tertiary structures.106
249
monomers present in the PrPSc multimer, resistance to PK is thought to be related to the presence
250
of β-sheet secondary structure.95, 98
251
adapted scrapie, are much more sensitive to PK than are other hamster-adapted strains, even
These β-solenoids are stacked and interlocked to form a quaternary structure of
In addition to the number of
Some prion strains, such as the drowsy strain of hamster-
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 12 of 70
252
though each strain is composed of the same PrPC monomers, albeit in different conformations.107-
253
108
254
strains.71 The same is true of the prions that cause chronic wasting disease in elk.71 This means
255
that PK resistance is not always a reliable means of distinguishing between PrPSc and PrPC.
The prions associated with scrapie are more sensitive to PK digestion than are other prion
256
The normal cellular prion protein (PrPC) is a glycolipoprotein. It possesses a single
257
disulfide bond. It has two sites that are variably glycosylated with large glycans, so the protein is
258
a mixture of diglycosylated, monoglycosylated, and unglycosylated proteins.109-111 There is
259
considerable variation in the composition of the sugars comprising the glycans.111 The protein
260
has a glycosylphosphatidylinositol (GPI) anchor.112 The sugar composition of the GPI anchor
261
also varies.113-114 Although the protein sequence remains constant, the variations in the sugar
262
composition of the glycans and the saccharide portion of the GPI anchor means that
263
comparatively few of the PrPC molecules have the same molecular weight, which complicates a
264
mass spectrometry-based analysis of the whole prion protein.
265
Prion nomenclature can be confusing. In this review, PrP refers to the prion protein, PrPSc refers to the infectious
266
regardless of conformation, glycosylation or GPI anchor.
267
conformation or isoform. PrPC refers to the natively expressed normal cellular prion protein
268
conformation, which is not infectious. Recombinant PrP (rPrP) is expressed in E. coli, so it has
269
no glycosylation or GPI anchor, but does possess a disulfide bond.
270
conformation as PrPC. Regardless of name, all prion proteins (PrPC, PrPSc, rPrP, and PrP) have
271
the same protein sequence and possess a single disulfide bond.
rPrP has the same
272
Unlike the natively expressed monomeric PrPC, prions are multimeric.103 This means that
273
the PrPSc multimeric complex needs to be denatured before it can be analyzed by mass
274
spectrometry. Denaturing in concentrated guanidine hydrochloride (GuHCl) is an effective
ACS Paragon Plus Environment
Page 13 of 70
Journal of Agricultural and Food Chemistry
275
method of inactivating prions.115 Methanol precipitation of these solutions removes the GuHCl
276
which would interfere with subsequent analysis and concentrates the protein contained in the
277
sample.38 In this way prions can be analyzed without contaminating instruments.
278
Prion proteins are irregularly glycosylated with glycans of varying composition, so
279
simple analysis of the whole protein is intractable.109-111 Instead the inactivated PrPSc can be
280
reduced, alkylated, and then digested with trypsin to yield a set of characteristic tryptic or
281
chymotryptic peptides suitable for MRM-based analysis.116-117 The required internal standards
282
can be conveniently generated from recombinant
283
once in the prion protein, so peptides can be used to demonstrate the presence of PrPSc and
284
quantitate the amount of PrPSc present in the sample.38,
285
permits researchers to quantitate covalent modifications of PrPSc, such as methionine oxidation
286
(Figure 6).71,
287
polymorphisms present in PrPSc from heterozygous sheep naturally infected with classical or
288
atypical scrapie.65, 119 MRM permits the straightforward detection and quantitation of a protein
289
that otherwise would be extremely difficult.
118
15
N-labeled PrP.71 Each peptide occurs only
116-117
Such peptide-based analysis
This approach was used to quantitate the relative proportion of PrP
290
Prions are known to infect and thereby contaminate the tissues of variety of food animals.
291
The prototypical prion disease is scrapie in sheep.81 The classical form of the disease was
292
described in the 18th century and perhaps even earlier.82 It is readily transmitted among sheep
293
and goats. A more recent form of scrapie, atypical scrapie, was first described in the late 20th
294
century.120 Its epidemiology is consistent with a sporadic origin.121 Based on analysis of human
295
consumption of sheep, there is no evidence that scrapie is transmissible to humans (zoonotic).122-
296
123
297
moose).124 It was first identified as a prion disease in 1978 and there is no epidemiological
Chronic wasting disease (CWD) is a prion disease of wild and farmed cervids (deer, elk
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
298
evidence that it is zoonotic.125 BSE transmitted to domestic cattle by the consumption of BSE
299
contaminated feed is referred to as classical BSE.126 The atypical forms of BSE (H-type or L-
300
type) are consistent with a very rare sporadic origin. The human manifestation of BSE is variant
301
Creutzfeldt-Jakob disease (vCJD), which is distinguishable from CJD.80
302
Different prions or strains, having characteristic phenotypes, are formed from the same
303
refolded PrPC.127 Furthermore, these prion strains replicate their properties in a given host.128
304
For example, if a sheep is infected with a BSE prion, the resulting prions are potentially zoonotic
305
even though they propagate in a sheep, whereas scrapie prions propagated in a sheep are not
306
zoonotic, despite the sheep expressing the same PrPC sequence as the one infected with BSE.128
307
Classical BSE (cBSE) is transmitted to cattle by consumption of BSE-contaminated feed, while
308
atypical BSE is thought to be a spontaneous disease.129-130 Sheep can be infected with classical
309
scrapie, which is acquired by transmission from a scrapie-infected sheep or by contact with a
310
scrapie-contaminated environment.81 The atypical form of scrapie, like BSE, is thought be a rare
311
spontaneous disease.121 There are varieties or strains of prions, and each can replicate its own
312
phenotype, which means that prions have the capacity to respond to selection pressures, i.e., they
313
can evolve.
314
among the prion strains.
The challenge of prion detection is to not only detect prions, but to distinguish
315 316
Detecting prions using mass spectrometry
317 318
Prions have been extensively analyzed by mass spectrometry.131 As noted previously the
319
work has been performed on trypsin or chymotrypsin digests of PrP. 65, 99, 117, 119, 132 The earliest
320
work demonstrated that the post-translational modifications present in PrPSc were also present in
ACS Paragon Plus Environment
Page 14 of 70
Page 15 of 70
Journal of Agricultural and Food Chemistry
321
PrPC.71, 89-90 Later, mass spectrometry was used to verify that the composition of the asparagine
322
bound glycans and the GPI anchor were similarly varied in both PrPC and PrPSc. The presence
323
of significant amounts of β-sheet is presumed to impede PK digestion.95, 98 Mass spectrometry
324
has been used to establish the boundaries of the β-sheet structure by determining the identity of
325
the truncated peptides remaining after a partial PK digestion. It has been used to determine the
326
relative susceptibility of the two polymorphisms present in a heterozygous sheep naturally
327
infected with scrapie or the sporadic atypical scrapie. Mass spectrometry has been used to detect
328
prions and diagnose prion diseases without the need to digest the sample with PK.
329
The conformational changes that occur when PrPC adopts the PrPSc conformation can
330
dramatically change the chemical environment of an amino acid contained within the protein.95,
331
98
332
on the PrP conformation. In principle such a difference in reactivity would result in a covalent
333
difference that would remain after PrPSc had been denatured to PrP prior to analysis (Figure 7).
334
This conformation dependent difference in reactivity is observed in a lysine that is part of the
335
epitope of the 3F4 or the 6D11 monoclonal antibodies (mAb).88 In the PrPC conformation, the
336
lysine in the epitope is exposed. In the PrPSc conformation it is a cryptotope or hidden epitope.
337
It has been used as part of a conformation dependent immunoassay (CDI) that measures the
338
difference in signal intensity from denatured and non-denatured samples.104 The ε-amino group
339
of this lysine also reacts readily with N-hydroxysuccinimide esters of carboxylic acids to form
340
the corresponding ε-amide. When the ε-amino group of this lysine is amidated, it is no longer
341
recognized by the 3F4 or 6D11 mAbs. The manipulations necessary for these reactions adds 30
342
minutes to the analysis time.
343
additional structural information it generates. In this way a mAb can be used to measure the
This means that the same amino acid can react differently with the same reagent, depending
This modest addition is more than compensated for by the
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
Page 16 of 70
344
amount PrPSc in a sample without using PK and in the presence of PrPC.88 Mass spectrometry
345
was used to observe the reaction of acetic anhydride or tetranitromethane with PrPSc.133
346
Mass spectrometry has been used to distinguish among prion strains.99 Often prion
347
strains differ in terms of their PK cleavage sites. Mass spectrometry has been used to identify
348
these sites and thereby distinguish between scrapie strains in sheep. Another approach exploits
349
the reactivity of the ε-amino group of lysine with the N-hydroxysuccinimide ester of acetic acid
350
(Ac-NHS). In this example, four of the five prion strains have identical PK cleavage products
351
(Figure 8).99 The strains were reacted with fixed concentrations of Ac-NHS for a fixed time.
352
The samples were not digested with PK and MRM was used to monitor the extent of acetylation
353
for the various lysines in PrP (Figure 9). This was necessary, since the appropriate mAbs to
354
detect all lysines in PrP do not exist. The extent of acetylation was different for each of the five
355
prions examined, which is consistent with each strain having a characteristic conformation.99
356
Mass spectrometry can be used to both detect prions and to distinguish among prion strains.
357
Fortunately, the one zoonotic prion disease, BSE, is rapidly disappearing. In 2016 there
358
were no cases of classical BSE identified in the United Kingdom, the place where BSE was first
359
identified and where the largest number of cases was diagnosed. 134 It is likely that classical BSE
360
soon will cease to exist.
361
Although CWD and scrapie have not been shown to be zoonotic,
122-123
there is some
362
experimental evidence to suggest they have that potential. The zoonotic potential of CPD
363
(camels) is not currently known, but will be defined by future work.87 Transgenic mice were
364
used to show that some scrapie strains have zoonotic potential.135 Scrapie was successfully
365
transmitted to cynomolgus macaques, a relevant human model, suggesting the potential
366
zoonosis.136 Squirrel monkeys succumbed to a prion disease after infection with CWD.
ACS Paragon Plus Environment
137-138
Page 17 of 70
Journal of Agricultural and Food Chemistry
367
Analogous attempts to infect cynomolgus macaques with CWD failed, suggesting that the tested
368
strains of CWD are not zoonotic.
369
CWD strains may be transmissible to cynomolgus macaques.140
370
appears that the need to monitor prions and prion strains will remain as long as prions exist.
371
Since mass spectrometry can directly detect prions and distinguish among prion strains without
372
using PK,65, 99, 117, 119 it will remain an invaluable analytical tool for the food forensicist.
373
137-139
Preliminary work in Canada suggests that Canadian Even though BSE is waning, it
There are other proteins that form amyloids and are found in food.141-142 Foie gras
374
contains amyloid.
Researchers have demonstrated that this amyloid can be transmitted to
375
vulnerable mice by oral transmission or by intravenous injection. It is not clear if the mouse
376
model applies to humans, but it does suggest the possibility that other protein refolding diseases
377
may also be transmissible by food.
378
provide a direct means of identifying protein conformations.
Fortunately for food forensicists, mass spectrometry may
379 380
Shiga toxins143
381 382
Shiga toxin was originally described, in the 19th century, as a toxin produced by the
383
bacterium Shigella dysenteriae.144 Much later, a similar toxin was isolated from cultures of E.
384
coli and referred to as type 1 Shiga toxin.44,
385
structurally and functionally related toxins, from other cultures of E. coli, that are referred to as
386
type 2 Shiga toxins (Figure 10).
387
verotoxins and Shiga-like toxins. Now they are all referred to as Shiga toxins.146 They are
388
classified into two types of Shiga toxins, type 1 and type 2, based on their amino acid sequence.
389
Both toxins are AB5 hexamers, in which the B subunits bind to a target eukaryotic cell and the A
145
Other researchers identified different, but
At various times these toxins have been referred to as
ACS Paragon Plus Environment
Journal of Agricultural and Food Chemistry
390
subunit contains a catalytic domain that causes the death of the eukaryotic host cell.147 Both
391
types of toxins can cause serious human disease, but type 2 Shiga toxins are most often
392
associated with the more serious symptoms of Shiga intoxication.
393
The Shiga toxins produced by Shiga toxin-producing E. coli (STEC) are the major
394
virulence factor associated with these food-borne illnesses.149 Shiga toxins are hexameric (AB5)
395
proteins (~70kDa) composed of an A subunit (~32 kDa), which contains the catalytic domain,
396
and is non-covalently bound to five identical B subunits (~7-8 kDa), which bind to specific
397
glycolipids on the surface of target eukaryotic cells.150 The B subunits bind to the saccharide
398
portion of the glycolipids globotriaosylceramide (Gb3) and/or globotetraosylceramide (Gb4).151-
399
155
400
binds to the Gb3 and/or Gb4 glycolipids on a detergent resistant portion of a membrane, then the
401
toxin can enter the cell. This binding of the toxin to a cell surface determines which cells are
402
affected by the toxin.
Mutations in the B subunits can alter the binding specificity of the B subunits.156 If the toxin
403
The binding of the toxin to a eukaryotic cell surface induces the endocytosis of the Shiga
404
toxin into the cell.157 The endocytosed toxins are sorted for retrograde trafficking instead of
405
lysosomal digestion. Retrograde trafficking moves the toxin through the Golgi apparatus to the
406
endoplasmic reticulum. During this process the A subunit is cleaved by a host enzyme (furin)
407
and a disulfide bond is cleaved to free the catalytic domain of the toxin (A1). The A1 domain
408
contains an N-terminal signal sequence which facilitates its retrotranslocation from the
409
endoplasmic reticulum to the cytosol. Once in the cytosol, the A1 domain is free to cleave a
410
specific adenine (A4324 in rat) of the 28S rRNA component of the ribosome, thereby halting
411
protein synthesis.158-159 A single functioning A1 subunit is sufficient to kill a eukaryotic cell
412
(Figure 11).
ACS Paragon Plus Environment
Page 18 of 70
Page 19 of 70
Journal of Agricultural and Food Chemistry
413 414
A sixteen-year survey (1998-2014) determined that while STEC was associated with only
415
2.5% of foodborne illness, it accounted for 13% of hospitalizations, and nearly 11% of the deaths
416
associated with foodborne illness.160 Every year approximately 175,000 Americans are infected
417
with STEC.15 Most patients (83%) have comparatively mild symptoms, such as severe stomach
418
cramps and sometimes diarrhea, but they are not serious enough to require hospitalization.161
419
There is often no fever or only a mild one (