Embed Size (px)
Citation preview
1
Original research article resubmitted to the Journal of Proteomics
Identification of protein markers for extracellular vesicle (EV) subsets
in cow’s milk
Abderrahim Benmoussa1, Clarisse Gotti2, Sylvie Bourassa2, Caroline Gilbert1 and Patrick Provost1
1 CHU de Québec Research Center/CHUL Pavilion, 2705 Blvd Laurier, Quebec City, QC, G1V
4G2 and Department of Microbiology, Infectious Diseases and Immunology, Faculty of Medicine,
Université Laval, Quebec City, QC, G1V 0A6, Canada; 2 Proteomics platform, Genomics Center,
CHU de Québec Research Center/CHUL Pavilion, 2705 Blvd Laurier, Quebec City, QC, G1V 4G2.
Running title: Protein markers for EV subsets in cow’s milk
* Corresponding author: Dr. Patrick Provost
CHU de Québec Research Center/CHUL Pavilion
2705 Blvd Laurier, Room T1-65
Quebec City, QC G1V 4G2 Canada
Phone: +1 418 525 4444 (ext. 48842)
Fax: +1 418 654 2765
E-mail: [email protected]
Web page: http://www.crchuq.ca/en/research/researchers/4691
3
ABSTRACT 1
Extracellular vesicles (EVs), like exosomes, are small membrane vesicles involved in cell-to-cell 2
communications that modulate numerous biological processes. We previously discovered a new EV 3
subset in milk (sedimenting at 35,000 g; 35K) that protected its cargo (RNAs and proteins) during 4
simulated digestion and was more enriched in microRNAs than exosomes (sedimenting at 100K). 5
Here, we used LC-MS/MS to push further the comparison between these two pellets. Commonly 6
used EV markers were not differentially enriched between the pellets, questioning their use with 7
cow’s milk EVs. Similarly, the majority of the quantified proteins were equally enriched between 8
the two pellets. Nevertheless, 20 proteins were specific to 35K, while 41 were specifically enriched 9
in 100K (p<0.05), suggesting their potential use as specific markers. Loaded with these proteins, the 10
EVs in these pellets might regulate translation, proliferation and cell survival for 35K, and 11
metabolism, extracellular matrix turnover and immunity for 100K. This approach also brought new 12
insights into milk EV-associated integrins and their possible role in specifically targeting recipient 13
cell types. These findings may help better discriminate between milk EVs, improve our 14
understanding of milk EV-associated protein function and their possible use as therapeutic tools for 15
the management of immunity- and metabolism-associated disorders. 16
17
18
19
20
21
22
23
24
Keywords • Extracellular vesicles / protein markers / tetraspanins / mass-spectrometry / 25
exosomes / cow milk 26
4
GRAPHICAL ABSTRACT 27
28
29
SIGNIFICANCE 30
This manuscript is in line with the view of the International Society for Extracellular Vesicles 31
(ISEV) on the importance of characterizing the different EV subsets present in a given biological 32
fluid [1]. Characterization of milk EVs by mass spectrometry unveiled new specific markers for 33
different EV subsets, which may help (i) ensure reproducibility in EV research, (ii) promote their 34
use for milk EV selection, (iii) predict their possible effects and functions in recipient cells, and (iv) 35
envision their use as specific (e.g. drug) delivery vehicle to target specific cell types/organs. 36
5
1. INTRODUCTION 37
Extracellular vesicles (EVs) are small lipid membrane vesicles released by numerous cell types [1]. 38
These EVs are among the factors exchanged by cells to ensure communication and homeostasis [2]. 39
EV populations are heterogeneous, but two major subsets capture most of the interest. The first and 40
most studied one comprise the exosomes, small vesicles of ~100 nm in diameter, generated within 41
the multivesicular bodies (MVB) by the escort complex (ESCRT) and associated proteins, and 42
released in the extracellular milieu upon the fusion of MVB with the cell membrane [3]. The second 43
one is composed of several larger EVs, sometimes termed ectosomes or membrane vesicles (MVs), 44
that are directly generated from budding of the cell membrane [3]. Other EV subsets have been 45
described, including apoptotic bodies, high density lipoproteins [4], and milk fat globules [5]. 46
EVs’ functions range from regulation of cell proliferation [6], modulation of inflammation, 47
receptor mediated signaling [7] to regulation of metabolism-associated pathways [8]. 48
In milk, EVs are relatively heterogeneous [9], with functional implications upon internalization 49
or ingestion [10-12], including regulation of T-lymphocyte maturation [13], modulation of 50
macrophage activity [12] and disease management in mouse models of rheumatoid arthritis [11]. 51
In our previous work, we found that processed dairy cow’s milk EVs and associated 52
microRNAs resisted degradation in an in vitro system that mimics human digestion [14]. We also 53
described the existence of a new subset of milk EVs that pellet at 12,000 g (12K) and 35,000 g 54
(35K), which contained the bulk of microRNAs present in milk [9]. 55
Overall, the EVs present in the 12K and 35K pellets were comparable and likely a population 56
of diverse small EVs, whereas the 100,000 g (100K) pellet is trusted to contain mostly milk 57
exosomes [15]. Previous non-quantitative liquid chromatography-tandem mass spectrometry (LC-58
MS/MS) and associated Western Blots suggested that some proteins are specifically enriched, either 59
in the 12K/35K pellets (e.g. Xanthine dehydrogenase, XDH) or in the 100K (e.g. Tumor 60
susceptibility gene 101, TSG101) pellet [9]. 61
6
EV subtype classification is a complicated matter, and there is no consensus about the 62
terminology for naming EVs [16]. The most frequently used term that refers to milk EVs is 63
“exosomes”, which derives specifically from MVB [17]. This denomination is based on (i) the 64
current practices in milk EV studies, which are mostly modeled on other biological fluids [18], (ii) 65
isolation methodology [19], and/or (iii) the presence of some protein markers usually enriched in 66
exosomes [20], although not being specific to this EV subset [21]. In the context of functional 67
studies, this has led to the exclusion of many other EV subpopulations found in bovine milk [10, 11, 68
22-25]; some of those are highly enriched in bioactive compounds, like microRNAs [9], and resist 69
digestion in vitro [14]. Therefore, there are very few resources available providing markers to 70
compare EV subsets in milk and to ensure purity of milk exosomes in isolation procedures or to 71
define other EV subsets present in milk. In this study, we aimed to discriminate milk EV subtypes 72
based on their content, rather than on their possible cellular origin. This discrimination is of high 73
importance when studying the biological activity of EVs, because it is only by ensuring 74
reproducible isolation procedures and relative purity of EVs that one can associate a function to a 75
specific EV subset [1, 15, 19, 26]. It is also of importance to determine which integrins are present 76
on the surface of these EVs, since these might lead to the accumulation of specific EV subsets in 77
specific cell types/organs; milk EVs may thus be used as a vehicle to develop strategies based on 78
their potential ability to deliver therapeutics to specific diseased cell types/organs [27-29]. 79
Here, we used quantitative LC-MS/MS to push our investigations further and provide a full 80
comparison of the EV-associated proteins found in the 35K and 100K pellets of dairy milk, with the 81
aim to define their content and comparative enrichment in specific proteins, including EVs markers 82
and integrins. This work may help (i) identify protein markers specific to each EVs subsets, (ii) 83
define their nature, cellular origin and possible cell type/organ targets, (iii) provide insights into the 84
function of the proteins they contain and the pathways they might impact upon cellular 85
internalization, and (iv) guide researchers interested in functional studies of milk EVs and their 86
potential for disease management. 87
7
2. MATERIALS AND METHODS 88
2.1 Dairy milk samples 89
For all experiments, we used commercially available, filtered, skimmed dairy milk (Lactantia 90
PurFiltre brand; product: http://www.lactantia.ca/food_product/lactantia-purfiltre-skim-milk/) 91
bought at a local grocery store in Quebec City, QC in biological triplicates (three milk tetra packs 92
with different expiry dates). 93
94
2.2 Sedimentation of dairy milk extracellular vesicles (EVs) by differential ultracentrifugation 95
Milk EV pellets were obtained by following a previously described protocol [9, 30], with slight 96
modifications. One hundred (100) mL of dairy milk samples were mixed with 1 volume of 2% 97
sodium citrate (in MilliQ water) that had been filtered with 0.22 µm membrane microfilters 98
(Corning). The samples were subjected to successive differential ultracentrifugation steps at 35,000 99
g (35K) for 2 h, then 70,000 g and 100,000 g (100K) for 1 h each at 4°C in a Sorvall® WX TL-100 100
ultracentrifuge, equipped with a T-1250 rotor (ThermoFisher). After each step, the pellets were 101
suspended in 1 mL of 0.22 µm filtered sterile phosphate buffered saline (PBS) containing 100 nM 102
ethylenediaminetetraacetic acid (EDTA) pH 7.4 and processed for subsequent LC-MS/MS analysis 103
of the 35K and 100K pellets. 104
105
2.3 Trizol LS protein isolation and resuspension 106
Proteins from the 35K and 100K pellets were precipitated with Trizol-LS by following the 107
manufacturer’s protocol, with slight modifications. Briefly, interphase and organic phase obtained 108
after adding chloroform to the pellets homogenized with Trizol-LS were mixed with 2 volumes of 109
isopropanol and centrifuged at 13,000 g. Pellets were rinsed thrice with absolute ethanol before 110
being solubilized in 25 µl of 50 mM ammonium bicarbonate containing 1% sodium deoxycholate, 111
and heated at 95°C for 5 min. Protein concentration of each sample was determined by colorimetric 112
Bradford assay (Thermo-Fisher Scientific). 113
8
114
2.4 Sample preparation for liquid chromatography-tandem mass spectrometry (LC-MS/MS) 115
Sample preparation and LC-MS/MS analyses were performed by the Proteomics Platform 116
(Genomics Center, CHU de Québec Research Center/CHUL Pavilion). Protein samples (10 µg) in 117
50mM ABC / 1% DOC were reduced with DTT (0.2 mM) at 37°C for 30 min, and cysteins were 118
alkylated by iodoacetamide (0.8 mM) at 37°C for 30 min. Trypsin (0.2 µg) was added and the 119
samples were incubated overnight at 37°C. Following the digestion, an acidification of the samples 120
was performed with 2% ACN / 1% TFA / 0.5% acetic acid in order to stop the trypsin reaction and 121
precipitated the deoxycholate. After centrifugation at 16000g for 5min, peptides contained in the 122
surpernatants were purified on stage tip C18 reversed phase (3M Empore C18 extraction disks), and 123
vacuum dried. Prior to MS injection, the peptides were solubilized at 0.2µg/uL in 2% acetonitrile / 124
0.05% TFA. 125
126
2.5 LC-MS/MS 127
Samples were analysed by nanoLC-MS/MS. For each injection, 1µg of peptide digest were 128
separated by online reversed-phase (RP) nanoscale capillary liquid chromatography (nanoLC) and 129
analysed by electrospray mass spectrometry (ESI MS/MS). The experiments were performed with a 130
Dionex UltiMate 3000 nanoRSLC chromatography system (Thermo Fisher Scientific / Dionex 131
Softron GmbH, Germering, Germany) connected to an Orbitrap Fusion mass spectrometer (Thermo 132
Fisher Scientific, San Jose, CA, USA) driving with Orbitrap Fusion Tune Application 2.0 and 133
equipped with a nanoelectrospray ion source. Peptides were trapped at 20 µL/min in loading solvent 134
(2% acetonitrile, 0.05% TFA) on a 5mm x 300 µm C18 PepMap cartridge pre-column (Thermo 135
Fisher Scientific / Dionex Softron GmbH, Germering, Germany) during 5 minutes. Then, the pre-136
column was switch online with a 50cm length, 75 µm ID Acclaim PepMap 100 C18 analytical 137
column (Thermo Fisher Scientific / Dionex Softron GmbH, Germering, Germany) and the peptides 138
were eluted with a linear gradient from 5-40% solvent B (A: 0,1% formic acid, B: 80% acetonitrile, 139
9
0.1% formic acid) in 90 minutes, at 300 nL/min. Mass spectra were acquired using a data dependent 140
acquisition mode using Thermo XCalibur software version 3.0.63. Full scan mass spectra (350 to 141
1800m/z) were acquired in the orbitrap using an AGC target of 4e5, a maximum injection time of 142
50 ms and a resolution of 120 000. Internal calibration using lock mass on the m/z 445.12003 143
siloxane ion was used. Each MS scan was followed by acquisition of fragmentation MSMS spectra 144
of the most intense ions for a total cycle time of 3 seconds (top speed mode). The selected ions were 145
isolated using the quadrupole analyser in a window of 1.6 m/z and fragmented by Higher energy 146
Collision-induced Dissociation (HCD) with 35% of collision energy. The resulting fragments were 147
detected by the linear ion trap in rapid scan rate with an AGC target of 1e4 and a maximum 148
injection time of 50 ms. Dynamic exclusion of previously fragmented peptides was set for a period 149
of 20 sec and a tolerance of 10 ppm. 150
151
2.6 Database searching and Label Free Quantification 152
Spectra were searched against a Bos taurus proteins database (Uniprot BosTaurus TaxID 9913) 153
using the Andromeda module of MaxQuant [31] software v. 1.5.5.1. Trypsin/P enzyme parameter 154
was selected with two possible missed cleavages. Carbamidomethylation of cysteins was set as 155
fixed modifications, and methionine oxidation and acetylation of protein N terminus as variable 156
modifications. Mass search tolerances were set to 7 ppm and 0.6 Da for MS and MS/MS 157
respectively. For protein validation, a maximum False Discovery Rate (FDR) of 1% at peptide and 158
protein level was used based on a target/decoy search. MaxQuant was also used for Label Free 159
Quantification (LFQ). The 'match between runs' option was used with a 20 min value as an 160
alignment time window and 5 min as a match time window. Only unique and 'razor' peptides were 161
used for quantification. The LFQ intensity values (non-normalized values from the ProteinGroups 162
file) extracted by MaxQuant for each protein in each sample replicate were normalized with the 163
median value and were used to calculate the ratio between the two samples. When LFQ intensity 164
values were missing, they were replaced by a noise value corresponding to the first percentile of 165
10
LFQ values of all proteins of the sample replicate. A protein was considered to be quantifiable only 166
if at least two of the three replicate values were present in one of the compared samples. 167
168
2.7 Statistical analysis 169
All statistical analysis were performed using R. Experiments were conducted in biological 170
triplicates (n=3) for the 35K and 100K pellets. A protein was considered to be variant if its z-score 171
is higher than 1.96 and had a Benjamini-Hochberg-corrected Welch p-value below 0.05. 172
173
2.8 Cellular component analysis 174
For cellular component analysis, we used the Gene-Ontology (GO) PANTHER 175
Overrepresentation Test (Released 20171205, PANTHER version 13.1 Released 2018-02-03) [32] 176
and Bos taurus (all genes in database) as a reference list. More specifically, we used Panther GO-177
slim cellular component analysis with Fisher's Exact multiple test correction with FDR, and 178
selecting only the results with FDR < 0.05. 179
180
2.9 Pathway analysis 181
Biological pathways potentially involving milk pellets’ proteins were determined using 182
Reactome database (release 63) with Pathways Analysis Tool (v3.5) [33], converting all non-human 183
identifiers to their human equivalents in order to get a hint at the possible impact of milk pellets’ 184
proteins on recipient human cells. 185
186
2.10 Figures and illustrations 187
Figures displayed in this manuscript were generated using R (Free Software Fondation) along 188
with the ggplot2(volcano) and gplot (heatmap) packages, Inkscape software (Free Software 189
Fondation) and Prism 7 (GraphPad Software, Inc.). 190
191
11
3. RESULTS 192
Milk 35K and 100K pellets have similar protein content, but specific protein markers 193
We subjected our cow milk samples to differential ultracentrifugation (Fig. 1A), except that we 194
skipped the 12,000 g centrifugation, since the 12K and 35K pellets contain closely related EV 195
subsets [9]. The 70K pellet was excluded from this analysis, as it contains a mixture of the two EV 196
subsets previously reported [9]. 197
We thus isolated all the proteins from the 35K and 100K pellets (n=3), and used quantitative 198
LC-MS/MS to determine their protein content and enrichment profiles (Fig. 1A). We identified and 199
quantitated a total of 1,974 different proteins with a FDR below 1% at the peptide and protein level 200
across all samples (Supplementary file S1). The complete list of the common and specific proteins, 201
with their related p-values, Z-scores and fold enrichment, are available in Supplementary File S1. 202
Some proteins were found in all samples, while others were identified only in some of the 203
replicates or only in one of the pellets (Fig. 1B). There were 1,974 different identified proteins in 204
these samples and 1899 quantified proteins Supplementary File S1. 205
There were 1,838 proteins common to both pellets, 20 proteins specific to the 35K pellet and 41 206
proteins specific to the 100K pellet (Fig. 1B). The 35K and 100K pellets contained several proteins 207
with a z-score above threshold (1.96) and an adjusted p-value below 0.05, making these proteins 208
possible specific markers (Fig. 2A). 209
Cellular localization analysis, through the Gene Ontology Panther tools, suggested that the 210
proteins the two pellets share in common were part of the Golgi apparatus, lysosomes, vesicular 211
coating, vacuoles, plasmic membrane, cytoplasm, endoplasmic reticulum or mitochondria, thereby 212
supporting the enrichment of extracellular vesicles in these pellets (Fig. 1B). Specific cellular origin 213
could be defined for 35K proteins, with the most significant ones involving actin cytoskeleton, 214
peroxisome and plastid, which support the presence of membrane-derived extracellular vesicles in 215
this pellet (Fig. 1B). Most of the proteins specific to the 100K pellet originate from endosomal 216
12
compartment, vacuole and extracellular region, which is in accordance with exosome enrichment in 217
this fraction (Fig. 1B). 218
The three most abundant proteins across all samples were Beta-lactoglobulin (LGB), bovine 219
progestagen-associated endometrial protein analog (PAEP) and Alpha-S1-casein (CSN1S1). These 220
proteins were equally enriched in all the samples (Fig. 1C). Most of the top 15 proteins include 221
common milk contaminant proteins (e.g. caseins, lactoglobulins) and half of them were more 222
enriched in the 100K compared to the 35K pellet (Fig. 1C). 223
Altogether, these results suggest that both 35K and 100K pellets have a very similar protein 224
content, with the most enriched proteins found more enriched in 100K pellet. Cellular origin 225
analysis suggests the presence of multiple EV subsets in these pellets, with exosomes dominating 226
the 100K pellet and cytoplasmic membrane-derived vesicles the 35K pellet. 227
228
Can usual EV markers be used to classify commercial dairy cow milk EVs? 229
We assessed whether widely accepted EV protein markers [21, 34] could be used to discriminate 230
the nature and subsets of the EVs isolated from commercial cow milk. Two previous studies, 231
performed on milk or another biological fluid (e.g. cell culture supernatant), suggested several 232
protein markers for different EV subsets, based either on isolation procedures and physical 233
properties of the EVs (light, dense, large or multiple EVs) [21], or on cellular origin of the proteins 234
associated with those EVs – proteins from the endosomal sorting complexes required for transport 235
(ESCRT) and proteins from the Rab family strongly associated to MVB-derived EVs (i.e. 236
exosomes) [34]. The concomitant presence of the three tetraspanins CD9, CD63 and CD81 was also 237
suggested to be primordial to define an EV subset as exosomes [21]. Samuel et al. [34] proposed 238
that the presence of integrins is an important marker for EV internalization and biological activity, 239
and those could be used to predict possible cell targets. We thus looked for the relative enrichment 240
of each of these markers and proteins of interest between the two pellets (Fig. 2 and 241
Supplementary File S3). 242
13
The tetraspanins CD9, CD63 and CD81 tended to be more enriched in the 100K pellet, 243
compared to the 35K pellet (Fig. 2). 244
For ESCRT-associated proteins, there was no discriminatory pattern between the two EV 245
populations: some components were equally enriched between the pellets (e.g., VPS4A, VTA1 and 246
CHMP6), some tended to be more enriched in the 100K pellet (e.g., VPS37C, MVB12A and 247
VPS28), while others (e.g., STAM, CHAMP4A and PTPN6) tended to be in higher abundance in 248
the 35K pellet (Fig. 2). 249
When we looked at the Rab family, the majority was found in comparable or close enrichment 250
between the 35K and 100K pellets (e.g., Rab22A, RAB13 and RAB10). However, the 35K pellet 251
tended to be more enriched in some of the Rab family proteins (e.g. RABGAP1, RAB3IP and 252
RAB9A). Only one Rab family protein (RAB33B) was slightly more enriched in the 100K pellet 253
versus the 35K pellet (Fig. 2). 254
Altogether, these results support the enrichment of CD9high, CD81high and CD63high exosomes 255
with some ESCRT proteins in the 100K pellet, but not without questioning the use of Rab proteins 256
as exosome markers. On the contrary, the enrichment of these proteins in the 35K pellet suggests 257
their importance in the biogenesis of EVs sedimenting at 35,000 g. 258
However, following our stringent methodology, none of these tendencies fell within the 259
significance threshold defined in the Materials and Methods section. The lack of statistically 260
different enrichment of these usual EV markers precludes their use to discriminate the nature and 261
subsets of cow milk EVs sedimenting at 35,000 and 100,000 g. 262
Further analyses of markers described for different EV subsets (light, dense, large or multiple 263
EVs) revealed that 3 of the 4 proteins associated to light EVs (including TSG101) were more 264
enriched in the 100K pellet, compared to the 35K pellet. EH Domain Containing 4 (EHD4), 265
however, was slightly more enriched in the 35K pellet (Fig. 2). Together, these findings suggest 266
that the 100K EVs may be termed “light EVs”. 267
14
On the other hand, 3 of the 4 proteins associated to dense EVs (e.g., complement proteins C2, 268
C6 and C7) were also more enriched in the 100K pellet, compared to the 35K pellet, with 269
Fibronectin 1(FN1) having comparable levels between the two pellets (Fig. 2). These observations 270
indicate the presence of multiple EV subsets in the 100K pellet. 271
Proteins usually found in large EVs and proteins associated to multiple EVs were slightly more 272
represented within the 35K pellet than in the100K pellets, with a tendency for enrichment of Actin 273
1 (ACTN1), Heat Shock Protein 90 Alpha Family Class B Member 1 (HSP90AB1) and Major 274
Vault protein (MVP) in the 35K pellet (Fig. 2). 275
Finally, most integrin proteins of importance in EVs internalization (ITGA1, ITFG1, ITGA6, 276
ITGAM and ITGB2) were found in equal levels between the two pellets. Only Integrin alpha-V 277
(ITGAV) was slightly, but not significantly, more enriched in the 100K pellet (Fig. 2), suggesting 278
that both pellets contain multiple EV subsets that can be internalized, albeit with a differential 279
capacity to bind extracellular matrix. 280
We observed a slight enrichment of some protein markers usable to discriminate at least 281
between the 35K and 100K pellets and their associated EVs, with STAM, CHMP4A, PTPN6, 282
RAB3IP and other Rab proteins being possible markers for 35K EVs, and TSG101, SDCBP, CD9, 283
CD63, CD81, Complement proteins C2, C6 and C7 and ITGAV being the ones most markedly 284
enriched in the 100K EV subsets. However, although the p-values of those markers were 285
significantly lower than 0.05, none of these canonical EV markers passed the z-score threshold, 286
suggesting that they may not be used alone to discriminate between commercial milk EVs. Their 287
possible use in combination, as previously suggested [21], could circumvent such limitations. 288
289
Unraveling specific protein markers for the 35K and 100K pellets, and their associated EVs 290
Unable to use canonical EV markers, we turned to the 20 proteins that were significantly more 291
enriched in the 35K pellet and on the 41 proteins specific to the 100K pellet to find specific markers 292
15
capable of discriminating between milk EV subsets (Fig. 1B and 3). All the proteins present in both 293
pellets are listed in Supplementary File S1. 294
The five proteins most specific to the 35K pellet included Epidermal growth factor receptor 295
substrate 15 (EPS15), Phosphoglycerate Dehydrogenase (PHGDH), dynactin subunit 2 (DCTN2), 296
Protein Kinase CAMP-Dependent Type II Regulatory Subunit Beta (PRKAR2B) and Glutaredoxin-297
3 (GLRX3), in decreasing order of significance (Fig. 3A and 3B). 298
Concerning the 100K pellet, the most specific proteins were Complement C8 Beta Chain 299
(C8B), C1GALT1-specific chaperone 1 (C1GALT1C1), Cartilage-associated protein (CRTAP), 300
Alpha-mannosidase 2x (MAN2A2) and Procollagen-lysine 2-oxoglutarate 5-dioxygenase 3 301
(PLOD3) (Fig. 3A and 3B). 302
These proteins may thus be used, alone or in combination, as markers to discriminate between 303
the EV subsets found in the 35K and 100K milk pellets. 304
305
Molecular function analysis unveils functional role of 100K milk EVs in translation 306
regulation, galactose and protein metabolism 307
Use of the GO Panther analysis tool revealed that proteins found at comparable levels in both 308
milk pellets (common) may be involved mainly in the regulation of translation initiation, protein 309
modification, peroxidase activity and metabolism-associated functions (Fig. 4A). 310
No defined function could be found for the 20 proteins specific to the 35K milk pellet. 311
For the proteins specifically enriched in the 100K pellet, three functions reached statistical 312
significance: galactosidase activity and, to a lesser extent, glycosyl transferase and peptidase 313
activity (Fig. 4B). 314
Together, these results suggest that the proteins found in the milk pellets are mainly involved in 315
translation regulation, protein maturation and cell structural maintenance, with 100K EVs and 316
associated proteins concealing specific metabolic activity. 317
318
16
Milk EVs and associated proteins may impact metabolism, translation and immunity 319
modulating pathways upon internalization 320
Milk EVs can be internalized by multiple human cells [22, 23, 25, 35, 36] and have been 321
suggested to enter blood stream along with their RNA cargo [37, 38] or to release their protein load 322
into recipient cells, thereby impacting their function [22, 23, 25, 35, 36]. Several studies previously 323
underlined the potential impact of milk EVs on immunity, although most of them used 100K EVs 324
[22, 23, 36, 39, 40] or a crude mixture of milk EVs [13]. 325
Here, we used the Reactome Pathways Analysis Tool to analyze the pathways potentially 326
impacted by a surge of milk EVs within human cells. We found that both 35K and 100K pellets 327
contained proteins capable of impacting translation regulation (e.g. 60S and 40S ribosomal subunits 328
modulation, translation initiation and elongation, etc.), immunity (e.g. neutrophil degranulation, 329
innate immune system), vesicular trafficking (e.g. vesicle-mediated transport, membrane 330
trafficking) and cell migration (e.g. axon guidance, signaling by ROBO receptors) (Table 1). 331
The 35K pellet proteins are likely more involved in cell proliferation and survival (e.g. 332
apoptosis-induced DNA fragmentation, G2/M transition, apoptotic execution phase), intercellular 333
structure (e.g. gap junction trafficking, formation of annular gap junctions, gap junction 334
degradation) and vesicle formation and vesicular communication (cell-extracellular matrix 335
interactions, clathrin-mediated endocytosis, membrane trafficking) (Table 2). As for the 100K 336
pellet proteins, they may have a wider range of impact on metabolism and extracellular matrix 337
turnover (e.g. glycosphingolipid metabolism, HS-GAG metabolism, extracellular matrix 338
organization), hemostasis and immunity modulation (e.g. platelet degranulation, innate immune 339
system, terminal pathway of complement) (Table 3). 340
Together, these results support a role for the EVs and associated proteins, found in both pellets, 341
in the modulation of immunity and translation, with additional implications of the proteins specific 342
to each of the milk pellets; 35K pellet proteins are potentially more involved in translation 343
regulation, cell migration, proliferation and survival, whereas those in the 100K pellet are possibly 344
17
oriented towards an enhanced metabolic and extracellular matrix turnover and immunity 345
modulation fate. The multiple EV subsets and associated proteins present in milk may thus 346
constitute a complex system that ensures the delivery of bioactive regulatory molecules naturally 347
selected to modulate digestive tract integrity and inflammatory processes. 348
349
4. DISCUSSION 350
There is an ever increasing interest in EVs among laboratory scientists and clinicians [41, 42], 351
with tremendous opportunities for EVs to be used as markers for disease [42] and product quality 352
assessment [43, 44], and as carriers for drug delivery and therapeutics [27-29]. With an isolation 353
method that is scalable virtually to the industrial level [45-48], milk EVs are among the most 354
interesting and promising EV population, as they may be used for inflammatory disease 355
management [11] and, even more, as drug carrier for cancer treatment [49]. 356
In this study, we characterized further two distinct EV populations present in cow milk, 357
including the well-studied exosomes and a new EV population that we recently reported [9], by LC-358
MS/MS. These analyses unveiled that this latter new EV population bears a wider variety of 359
proteins than canonical milk exosomes. However, both EV subsets share as many as 1,838 different 360
proteins, including XDH, BTN1A1 and MFGE8, strongly suggesting that they have a common 361
mammary gland cell origin [50, 51] – both EV subsets may come from the same cell type [20]. 362
Some of these proteins, like XDH, are highly enriched in milk fat globules (MFG). Comparison of 363
the milk proteins associated with EVs versus MFGs, using a similar approach, could help define 364
whether milk EVs derive from MFGs upon processing or if they are indeed secreted during 365
lactation. Several of these proteins are well-known for their immunity modulating properties [52] 366
and their implication in health and disease management, such as ulcerative colitis for MFGE8 [53]. 367
These properties of dietary (milk) EVs may thus provide an additional mechanism by which 368
nutrition may help prevent inflammatory diseases and improve health. 369
18
The complete list of the new markers of interest that we have found for each of the milk 35K 370
and 100K pellets (Supplementary Files S1 and S3) may help discriminate the EV subsets under 371
study and ensure reproducibility of milk EV isolation and functional analyses [1, 15, 19, 26]. 372
Because of its low enrichment in microRNAs [9], its high volumes and its relatively high 373
enrichment in caseins and whey proteins, the residual sample obtained after ultracentrifugation 374
(supernatant, SN) was not analyzed in this study. Nevertheless, we systematically kept the SN 375
fraction and used it as a control for the milk EVs tested in our experimental settings the same way 376
as proposed by the ISEV position paper for EV RNA functional studies [54]. 377
Based on these results, the 100K milk pellet likely contains the EV subset closest to 378
“exosomes”, as they are usually referred to in the literature [55]. However, considering the presence 379
of ESCRT and Rab family proteins in both pellets, it would be more prudent to stick to the generic 380
terminology for milk EVs, fully describe the isolation methodology and provide a full description of 381
those EVs based on their content in specific markers (e.g. CD63high or +, CD81high, CD9high, bta-miR-382
223low EVs), as it is the common practice for immune cell types. In fact, the relatively high 383
enrichment of CD9, CD63 and CD81 proteins in the 100K pellet support their combined use as 384
“exosomes” markers – their concomitant presence is likely necessary to define EVs as exosomes, as 385
previously suggested [21]. 386
Common EV type markers, such as ESCRT complex proteins, Rab family proteins and 387
tetraspanins, could not be used alone to distinguish these EV subsets. The presence of ESCRT 388
machinery-associated proteins in both EV subsets does not go without reminding of a non-canonical 389
type of EVs that are generated by a cellular mechanism very close to the one ensuring exosome 390
biogenesis, but occurring at the cell membrane instead of the MVB membrane [56]. 391
It may be interesting to note that membrane proteins bared at the surface of EVs may become 392
surface receptors upon EV fusion with the recipient cells’ membrane and have functional 393
implications [57]. In the case of MFGE8, such a transfer would confer the ability to recipient cells 394
to link integrin αvβ3-5 and modulate immune and inflammatory processes [58-67]. Similarly, the 395
19
relatively high enrichment of integrin ITGAV in 100K milk EVs suggest the possible enhanced 396
internalization of milk exosomes harboring it [34] by gastrointestinal tissues and cells that are 397
enriched in the ITGAV ligand vitronectin (GeneAtlas, U133A, gcrma) [68]. 35K EVs, which are 398
impoverished in this specific integrin, might not bind to the gut cells as efficiently and end up in the 399
colon, in contact with gut bacteria, and impact the host-bacteria crosstalk through extracellular 400
vesicles [69]. Therefore, the use of LC-MS/MS to determine the integrin enrichment between 401
different EV subset could be applied to specifically identify and select an EV subset that bears a 402
specific integrin profile targeting a specific tissue, which may allow its use as a targeted drug 403
delivery vehicle [27-29]. 404
We have identified more than 20 and 40 new possible protein markers specific to the EVs 405
present in the 35K and 100K pellets, respectively, which will help researchers identify their milk 406
EVs of interest and facilitate data dissemination and comparison. Admittedly, the population of 407
milk EVs is likely more complex than the two subsets that we have isolated and characterized, and 408
our data may help optimize the selection of specific EV subsets and their study. 409
Intriguingly, the 35K EVs, which are most often readily discarded when isolating exosomes, 410
seem to be more abundant and more enriched in microRNAs [9] than the canonical exosomes found 411
in the 100K milk pellet. We have shown previously that the bulk of microRNAs that they carry 412
resist digestion [14], supporting a protective effect conferred upon EV association. Containing 413
comparable amounts of immunity modulating proteins and harboring surface integrins of 414
importance for cellular internalization, milk EV subsets, which are likely heterogeneous in nature, 415
may cooperate towards an enhanced regulation of immunity and health status. Whereas 100K EVs 416
may exert more pronounced effects on the regulation of metabolism and inflammation, 35K EVs 417
may be involved more closely in digestive tract maintenance and integrity, as previously reported 418
for milk exosomes [34]. 419
Providing anti-inflammatory proteins, growth factors and gene regulatory microRNAs, the 420
relatively abundant EVs present in maternal milk [10] likely play an essential role in infant gut 421
20
development. Supplementation of infant formulations of milk, with specific EV subsets and protein 422
content, as previously done with milk fat globule membranes [5], should be considered in order to 423
better simulate maternal milk composition and properties, and improve the nutritional value and 424
health benefits of the formulation. 425
The relative scarcity of reliable antibodies recognizing the most interesting milk proteins of 426
bovine origin may hamper validation of our results and further projects based on them. 427
Nevertheless, now that these proteins have been identified, approaches like Targeted Proteomics by 428
Selected Reaction Monitoring (SRM) or Parallel-Reaction Monitoring (PRM) might be used to 429
confirm the relevance of these markers in different biological samples [70]. Therefore, the present 430
report does provide the milk EV research community and manufacturers with a list of bovine 431
proteins of interest for monitoring milk EVs and possibly producing and developing antibodies 432
aimed to detect specific milk EV subsets. 433
In this study, we chose to use the z-sore approach instead of setting an arbitrary 2 or 3 fold 434
change, like previously described [21, 34]. The z-score approach is more stringent and takes into 435
account the differences between the biological replicates, the biological background and the 436
variability of the samples. These differences might lead to discrepancies between our work and 437
previous ones on milk EVs, but not without convergence when looking at the most expressed 438
proteins [21, 34]. 439
Finally, the experiments reported in this study used commercial skimmed, filtered and 440
pasteurized cow’s milk of the same brand and type as our previous studies [9, 14]. We chose to 441
keep the same cow’s milk type and associated experimental protocols so to allow comparison of our 442
data and reproducibility in our experiments and between studies. It is possible that other cow’s milk 443
brands or types (raw or processed [homogenized], ultra-high temperature [UHT] sterilized or 444
pasteurized, filtered or not filtered, skimmed, half and half or whole and any combination of the 445
theses possibilities) exhibit different protein profiles the same way as milk processing impacted 446
milk EV and microRNA profiles [9, 71]. 447
21
448
5. CONCLUSION 449
Milk EVs are emerging as a novel research arena of interest that focuses increasingly on their 450
characterization, biological role, function and use as therapeutic tools and vehicle [41, 42]. The EV 451
isolation protocols, protein markers available and the current trends in the field have led researchers 452
to focus almost exclusively on milk exosomes, while ignoring numerous types of EV subsets whose 453
characteristics, content and function may also be of interest [21]. By documenting the resistance of 454
milk EVs to digestion [14], their enrichment in immunity-modulating microRNAs and immunity-455
associated proteins [9], our previous work unveiled the potential importance of milk EVs 456
sedimenting at ultracentrifugations speeds lower than 100K. The current study unveiled major 457
protein markers that are differentially enriched between the EVs sedimenting at 35K and those 458
sedimenting at 100K, and provides specific markers that may be used to ensure reproducibility in 459
milk EV research. The suitability and reliability of these markers remain to be determined and is 460
limited by the availability of antibodies recognizing these bovine proteins. Although our results 461
may be transposed to other biological fluids, one has to recognize and appropriately consider the 462
several features that may be unique to milk EVs. 463
464
ACKNOWLEDGMENTS AND FUNDINGS 465
This work was supported by the Canadian Institutes of Health Research (CIHR) [Grants No. 466
319618 and 327522] through the Institute of Genetics (to P.P.). 467
468
DISCLOSURE OF CONFLICT OF INTERESTS 469
The authors state that they have no conflict of interests. 470
22
REFERENCES
[1] J. Lotvall, A.F. Hill, F. Hochberg, E.I. Buzas, D. Di Vizio, C. Gardiner, Y.S. Gho, I.V. Kurochkin, S. Mathivanan, P. Quesenberry, S. Sahoo, H. Tahara, M.H. Wauben, K.W. Witwer, C. Thery, Minimal experimental requirements for definition of extracellular vesicles and their functions: a position statement from the International Society for Extracellular Vesicles, Journal of extracellular vesicles 3 (2014) 26913. [2] M. Yanez-Mo, P.R. Siljander, Z. Andreu, A.B. Zavec, F.E. Borras, E.I. Buzas, K. Buzas, E. Casal, F. Cappello, J. Carvalho, E. Colas, A. Cordeiro-da Silva, S. Fais, J.M. Falcon-Perez, I.M. Ghobrial, B. Giebel, M. Gimona, M. Graner, I. Gursel, M. Gursel, N.H. Heegaard, A. Hendrix, P. Kierulf, K. Kokubun, M. Kosanovic, V. Kralj-Iglic, E.M. Kramer-Albers, S. Laitinen, C. Lasser, T. Lener, E. Ligeti, A. Line, G. Lipps, A. Llorente, J. Lotvall, M. Mancek-Keber, A. Marcilla, M. Mittelbrunn, I. Nazarenko, E.N. Nolte-'t Hoen, T.A. Nyman, L. O'Driscoll, M. Olivan, C. Oliveira, E. Pallinger, H.A. Del Portillo, J. Reventos, M. Rigau, E. Rohde, M. Sammar, F. Sanchez-Madrid, N. Santarem, K. Schallmoser, M.S. Ostenfeld, W. Stoorvogel, R. Stukelj, S.G. Van der Grein, M.H. Vasconcelos, M.H. Wauben, O. De Wever, Biological properties of extracellular vesicles and their physiological functions, Journal of extracellular vesicles 4 (2015) 27066. [3] E.R. Abels, X.O. Breakefield, Introduction to Extracellular Vesicles: Biogenesis, RNA Cargo Selection, Content, Release, and Uptake, Cell Mol Neurobiol 36(3) (2016) 301-12. [4] R. Crescitelli, C. Lasser, T.G. Szabo, A. Kittel, M. Eldh, I. Dianzani, E.I. Buzas, J. Lotvall, Distinct RNA profiles in subpopulations of extracellular vesicles: apoptotic bodies, microvesicles and exosomes, Journal of extracellular vesicles 2 (2013). [5] G. Bhinder, J.M. Allaire, C. Garcia, J.T. Lau, J.M. Chan, N.R. Ryz, E.S. Bosman, F.A. Graef, S.M. Crowley, L.S. Celiberto, J.C. Berkmann, R.A. Dyer, K. Jacobson, M.G. Surette, S.M. Innis, B.A. Vallance, Milk Fat Globule Membrane Supplementation in Formula Modulates the Neonatal Gut Microbiome and Normalizes Intestinal Development, Scientific reports 7 (2017) 45274. [6] M. Takasugi, R. Okada, A. Takahashi, D. Virya Chen, S. Watanabe, E. Hara, Small extracellular vesicles secreted from senescent cells promote cancer cell proliferation through EphA2, Nature communications 8 (2017) 15729. [7] P.D. Robbins, A.E. Morelli, Regulation of immune responses by extracellular vesicles, Nat Rev Immunol 14(3) (2014) 195-208. [8] N. Iraci, E. Gaude, T. Leonardi, A.S.H. Costa, C. Cossetti, L. Peruzzotti-Jametti, J.D. Bernstock, H.K. Saini, M. Gelati, A.L. Vescovi, C. Bastos, N. Faria, L.G. Occhipinti, A.J. Enright, C. Frezza, S. Pluchino, Extracellular vesicles are independent metabolic units with asparaginase activity, Nature chemical biology 13(9) (2017) 951-955. [9] A. Benmoussa, S. Ly, S.T. Shan, J. Laugier, E. Boilard, C. Gilbert, P. Provost, A subset of extracellular vesicles carries the bulk of microRNAs in commercial dairy cow’s milk, Journal of extracellular vesicles 6(1) (2017) 1401897. [10] J. Zempleni, A. Aguilar-Lozano, M. Sadri, S. Sukreet, S. Manca, D. Wu, F. Zhou, E. Mutai, Biological Activities of Extracellular Vesicles and Their Cargos from Bovine and Human Milk in Humans and Implications for Infants, The Journal of nutrition (2016). [11] O.J. Arntz, B.C. Pieters, M.C. Oliveira, M.G. Broeren, M.B. Bennink, M. de Vries, P.L. van Lent, M.I. Koenders, W.B. van den Berg, P.M. van der Kraan, F.A. van de Loo, Oral administration of bovine milk derived extracellular vesicles attenuates arthritis in two mouse models, Molecular nutrition & food research (2015). [12] B.C. Pieters, O.J. Arntz, M.B. Bennink, M.G. Broeren, A.P. van Caam, M.I. Koenders, P.L. van Lent, W.B. van den Berg, M. de Vries, P.M. van der Kraan, F.A. van de Loo, Commercial cow milk contains physically stable extracellular vesicles expressing immunoregulatory TGF-beta, PLoS One 10(3) (2015) e0121123. [13] C. Admyre, S.M. Johansson, K.R. Qazi, J.J. Filen, R. Lahesmaa, M. Norman, E.P.A. Neve, A. Scheynius, S. Gabrielsson, Exosomes with Immune Modulatory Features Are Present in Human Breast Milk, The Journal of Immunology 179(3) (2007) 1969-1978.
23
[14] A. Benmoussa, C.H. Lee, B. Laffont, P. Savard, J. Laugier, E. Boilard, C. Gilbert, I. Fliss, P. Provost, Commercial Dairy Cow Milk microRNAs Resist Digestion under Simulated Gastrointestinal Tract Conditions, The Journal of nutrition (2016). [15] D.W. Greening, R. Xu, H. Ji, B.J. Tauro, R.J. Simpson, A protocol for exosome isolation and characterization: evaluation of ultracentrifugation, density-gradient separation, and immunoaffinity capture methods, Methods Mol Biol 1295 (2015) 179-209. [16] G. Raposo, W. Stoorvogel, Extracellular vesicles: Exosomes, microvesicles, and friends, 2013. [17] N.P. Hessvik, A. Llorente, Current knowledge on exosome biogenesis and release, Cellular and molecular life sciences : CMLS 75(2) (2018) 193-208. [18] E. Zeringer, T. Barta, M. Li, A.V. Vlassov, Strategies for isolation of exosomes, Cold Spring Harb Protoc 2015(4) (2015) 319-23. [19] C. Gardiner, D. Di Vizio, S. Sahoo, C. Thery, K.W. Witwer, M. Wauben, A.F. Hill, Techniques used for the isolation and characterization of extracellular vesicles: results of a worldwide survey, Journal of extracellular vesicles 5 (2016) 32945. [20] A. Bobrie, M. Colombo, S. Krumeich, G. Raposo, C. Thery, Diverse subpopulations of vesicles secreted by different intracellular mechanisms are present in exosome preparations obtained by differential ultracentrifugation, Journal of extracellular vesicles 1 (2012). [21] J. Kowal, G. Arras, M. Colombo, M. Jouve, J.P. Morath, B. Primdal-Bengtson, F. Dingli, D. Loew, M. Tkach, C. Thery, Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes, Proceedings of the National Academy of Sciences of the United States of America (2016). [22] P. Rani, M. Vashisht, N. Golla, S. Shandilya, S.K. Onteru, D. Singh, Milk miRNAs encapsulated in exosomes are stable to human digestion and permeable to intestinal barrier in vitro, Journal of Functional Foods 34 (2017) 431-439. [23] Y. Liao, X. Du, J. Li, B. Lonnerdal, Human milk exosomes and their microRNAs survive digestion in vitro and are taken up by human intestinal cells, Molecular nutrition & food research 61(11) (2017). [24] T. Chen, M.Y. Xie, J.J. Sun, R.S. Ye, X. Cheng, R.P. Sun, L.M. Wei, M. Li, D.L. Lin, Q.Y. Jiang, Q.Y. Xi, Y.L. Zhang, Porcine milk-derived exosomes promote proliferation of intestinal epithelial cells, Scientific reports 6 (2016) 33862. [25] R.J. Kusuma, S. Manca, T. Friemel, S. Sukreet, C. Nguyen, J. Zempleni, Human vascular endothelial cells transport foreign exosomes from cow's milk by endocytosis, American journal of physiology. Cell physiology 310(10) (2016) C800-7. [26] K.W. Witwer, E.I. Buzas, L.T. Bemis, A. Bora, C. Lasser, J. Lotvall, E.N. Nolte-'t Hoen, M.G. Piper, S. Sivaraman, J. Skog, C. Thery, M.H. Wauben, F. Hochberg, Standardization of sample collection, isolation and analysis methods in extracellular vesicle research, Journal of extracellular vesicles 2 (2013). [27] P. Gangadaran, C.M. Hong, B.C. Ahn, An Update on in Vivo Imaging of Extracellular Vesicles as Drug Delivery Vehicles, Frontiers in pharmacology 9 (2018) 169. [28] B.B. Aggarwal, W. Yuan, S. Li, S.C. Gupta, Curcumin-free turmeric exhibits anti-inflammatory and anticancer activities: Identification of novel components of turmeric, Molecular nutrition & food research 57(9) (2013) 1529-42. [29] K.I. Mentkowski, J.D. Snitzer, S. Rusnak, J.K. Lang, Therapeutic Potential of Engineered Extracellular Vesicles, The AAPS journal 20(3) (2018) 50. [30] T. Hata, K. Murakami, H. Nakatani, Y. Yamamoto, T. Matsuda, N. Aoki, Isolation of bovine milk-derived microvesicles carrying mRNAs and microRNAs, Biochem Biophys Res Commun 396(2) (2010) 528-33. [31] J. Cox, M. Mann, MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification, Nat Biotechnol 26(12) (2008) 1367-72.
24
[32] H. Mi, X. Huang, A. Muruganujan, H. Tang, C. Mills, D. Kang, P.D. Thomas, PANTHER version 11: expanded annotation data from Gene Ontology and Reactome pathways, and data analysis tool enhancements, Nucleic acids research 45(D1) (2017) D183-D189. [33] A. Fabregat, S. Jupe, L. Matthews, K. Sidiropoulos, M. Gillespie, P. Garapati, R. Haw, B. Jassal, F. Korninger, B. May, M. Milacic, C.D. Roca, K. Rothfels, C. Sevilla, V. Shamovsky, S. Shorser, T. Varusai, G. Viteri, J. Weiser, G. Wu, L. Stein, H. Hermjakob, P. D'Eustachio, The Reactome Pathway Knowledgebase, Nucleic acids research 46(D1) (2018) D649-D655. [34] M. Samuel, D. Chisanga, M. Liem, S. Keerthikumar, S. Anand, C.S. Ang, C.G. Adda, E. Versteegen, M. Jois, S. Mathivanan, Bovine milk-derived exosomes from colostrum are enriched with proteins implicated in immune response and growth, Scientific reports 7(1) (2017) 5933. [35] H. Izumi, M. Tsuda, Y. Sato, N. Kosaka, T. Ochiya, H. Iwamoto, K. Namba, Y. Takeda, Bovine milk exosomes contain microRNA and mRNA and are taken up by human macrophages, Journal of dairy science 98(5) (2015) 2920-33. [36] T. Wolf, S.R. Baier, J. Zempleni, The Intestinal Transport of Bovine Milk Exosomes Is Mediated by Endocytosis in Human Colon Carcinoma Caco-2 Cells and Rat Small Intestinal IEC-6 Cells, The Journal of nutrition 145(10) (2015) 2201-6. [37] L. Wang, M. Sadri, D. Giraud, J. Zempleni, RNase H2-Dependent Polymerase Chain Reaction and Elimination of Confounders in Sample Collection, Storage, and Analysis Strengthen Evidence That microRNAs in Bovine Milk Are Bioavailable in Humans, The Journal of nutrition 148(1) (2018) 153-159. [38] S.R. Baier, C. Nguyen, F. Xie, J.R. Wood, J. Zempleni, MicroRNAs are absorbed in biologically meaningful amounts from nutritionally relevant doses of cow milk and affect gene expression in peripheral blood mononuclear cells, HEK-293 kidney cell cultures, and mouse livers, The Journal of nutrition 144(10) (2014) 1495-500. [39] T. Chen, Q.Y. Xi, R.S. Ye, X. Cheng, Q.E. Qi, S.B. Wang, G. Shu, L.N. Wang, X.T. Zhu, Q.Y. Jiang, Y.L. Zhang, Exploration of microRNAs in porcine milk exosomes, BMC Genomics 15 (2014) 100. [40] Q. Zhou, M. Li, X. Wang, Q. Li, T. Wang, Q. Zhu, X. Zhou, X. Wang, X. Gao, X. Li, Immune-related microRNAs are abundant in breast milk exosomes, Int J Biol Sci 8(1) (2012) 118-23. [41] D.K. Kim, J. Lee, S.R. Kim, D.S. Choi, Y.J. Yoon, J.H. Kim, G. Go, D. Nhung, K. Hong, S.C. Jang, S.H. Kim, K.S. Park, O.Y. Kim, H.T. Park, J.H. Seo, E. Aikawa, M. Baj-Krzyworzeka, B.W. van Balkom, M. Belting, L. Blanc, V. Bond, A. Bongiovanni, F.E. Borras, L. Buee, E.I. Buzas, L. Cheng, A. Clayton, E. Cocucci, C.S. Dela Cruz, D.M. Desiderio, D. Di Vizio, K. Ekstrom, J.M. Falcon-Perez, C. Gardiner, B. Giebel, D.W. Greening, J.C. Gross, D. Gupta, A. Hendrix, A.F. Hill, M.M. Hill, E. Nolte-'t Hoen, D.W. Hwang, J. Inal, M.V. Jagannadham, M. Jayachandran, Y.K. Jee, M. Jorgensen, K.P. Kim, Y.K. Kim, T. Kislinger, C. Lasser, D.S. Lee, H. Lee, J. van Leeuwen, T. Lener, M.L. Liu, J. Lotvall, A. Marcilla, S. Mathivanan, A. Moller, J. Morhayim, F. Mullier, I. Nazarenko, R. Nieuwland, D.N. Nunes, K. Pang, J. Park, T. Patel, G. Pocsfalvi, H. Del Portillo, U. Putz, M.I. Ramirez, M.L. Rodrigues, T.Y. Roh, F. Royo, S. Sahoo, R. Schiffelers, S. Sharma, P. Siljander, R.J. Simpson, C. Soekmadji, P. Stahl, A. Stensballe, E. Stepien, H. Tahara, A. Trummer, H. Valadi, L.J. Vella, S.N. Wai, K. Witwer, M. Yanez-Mo, H. Youn, R. Zeidler, Y.S. Gho, EVpedia: a community web portal for extracellular vesicles research, Bioinformatics 31(6) (2015) 933-9. [42] S. Roy, F.H. Hochberg, P.S. Jones, Extracellular vesicles: the growth as diagnostics and therapeutics; a survey, Journal of extracellular vesicles 7(1) (2018) 1438720. [43] S. Raynel, M.P. Padula, D.C. Marks, L. Johnson, Cryopreservation alters the membrane and cytoskeletal protein profile of platelet microparticles, Transfusion 55(10) (2015) 2422-32. [44] J. Sun, K. Aswath, S.G. Schroeder, J.D. Lippolis, T.A. Reinhardt, T.S. Sonstegard, MicroRNA expression profiles of bovine milk exosomes in response to Staphylococcus aureus infection, BMC Genomics 16 (2015) 806.
25
[45] K. Vaswani, Y.Q. Koh, F.B. Almughlliq, H.N. Peiris, M.D. Mitchell, A method for the isolation and enrichment of purified bovine milk exosomes, Reproductive biology 17(4) (2017) 341-348. [46] K. Blans, M.S. Hansen, L.V. Sorensen, M.L. Hvam, K.A. Howard, A. Moller, L. Wiking, L.B. Larsen, J.T. Rasmussen, Pellet-free isolation of human and bovine milk extracellular vesicles by size-exclusion chromatography, Journal of extracellular vesicles 6(1) (2017) 1294340. [47] M.S.H. Kristine Blans, Laila Sørensen, Michael L. Hvam, Lotte B. Larsen, Lars Wiking and Jan T. Rasmussen, Effective and gentle isolation of extracellular vesicles in human and bovine milk without ultracentrifugation , ISEV International Meeting 2016, 2016. [48] M. Somiya, Y. Yoshioka, T. Ochiya, Biocompatibility of highly purified bovine milk-derived extracellular vesicles, Journal of extracellular vesicles 7(1) (2018) 1440132. [49] F. Aqil, R. Munagala, J. Jeyabalan, A.K. Agrawal, R. Gupta, Exosomes for the Enhanced Tissue Bioavailability and Efficacy of Curcumin, The AAPS journal 19(6) (2017) 1691-1702. [50] X. Zheng, C. Ning, Y. Dong, P. Zhao, J. Li, Z. Fan, J. Li, Y. Yu, R. Mrode, J.F. Liu, Quantitative proteome analysis of bovine mammary gland reveals protein dynamic changes involved in peak and late lactation stages, Biochemical and biophysical research communications 494(1-2) (2017) 292-297. [51] J. Jeong, A.U. Rao, J. Xu, S.L. Ogg, Y. Hathout, C. Fenselau, I.H. Mather, The PRY/SPRY/B30.2 domain of butyrophilin 1A1 (BTN1A1) binds to xanthine oxidoreductase: implications for the function of BTN1A1 in the mammary gland and other tissues, The Journal of biological chemistry 284(33) (2009) 22444-56. [52] C.H. Wang, C. Zhang, X.H. Xing, Xanthine dehydrogenase: An old enzyme with new knowledge and prospects, Bioengineered (2016) 1-11. [53] A. Otani, S. Ishihara, M.M. Aziz, N. Oshima, Y. Mishima, I. Moriyama, T. Yuki, Y. Amano, M.M. Ansary, Y. Kinoshita, Intrarectal administration of milk fat globule epidermal growth factor-8 protein ameliorates murine experimental colitis, International journal of molecular medicine 29(3) (2012) 349-56. [54] B. Mateescu, E.J. Kowal, B.W. van Balkom, S. Bartel, S.N. Bhattacharyya, E.I. Buzas, A.H. Buck, P. de Candia, F.W. Chow, S. Das, T.A. Driedonks, L. Fernandez-Messina, F. Haderk, A.F. Hill, J.C. Jones, K.R. Van Keuren-Jensen, C.P. Lai, C. Lasser, I.D. Liegro, T.R. Lunavat, M.J. Lorenowicz, S.L. Maas, I. Mager, M. Mittelbrunn, S. Momma, K. Mukherjee, M. Nawaz, D.M. Pegtel, M.W. Pfaffl, R.M. Schiffelers, H. Tahara, C. Thery, J.P. Tosar, M.H. Wauben, K.W. Witwer, E.N. Nolte-'t Hoen, Obstacles and opportunities in the functional analysis of extracellular vesicle RNA - an ISEV position paper, Journal of extracellular vesicles 6(1) (2017) 1286095. [55] J. Zempleni, Milk exosomes: beyond dietary microRNAs, Genes & nutrition 12(1) (2017). [56] G. Odorizzi, Membrane manipulations by the ESCRT machinery, F1000Research 4(F1000 Faculty Rev) (2015) 516. [57] K. Al-Nedawi, B. Meehan, J. Micallef, V. Lhotak, L. May, A. Guha, J. Rak, Intercellular transfer of the oncogenic receptor EGFRvIII by microvesicles derived from tumour cells, Nature cell biology 10(5) (2008) 619-24. [58] Y. Hirano, W.L. Yang, M. Aziz, F. Zhang, B. Sherry, P. Wang, MFG-E8-derived peptide attenuates adhesion and migration of immune cells to endothelial cells, Journal of leukocyte biology 101(5) (2017) 1201-1209. [59] W. Huang, J. Wu, H. Yang, Y. Xiong, R. Jiang, T. Cui, D. Ye, Milk fat globule-EGF factor 8 suppresses the aberrant immune response of systemic lupus erythematosus-derived neutrophils and associated tissue damage, Cell death and differentiation 24(2) (2017) 263-275. [60] A. Das, S. Ghatak, M. Sinha, S. Chaffee, N.S. Ahmed, N.L. Parinandi, E.S. Wohleb, J.F. Sheridan, C.K. Sen, S. Roy, Correction of MFG-E8 Resolves Inflammation and Promotes Cutaneous Wound Healing in Diabetes, Journal of immunology (Baltimore, Md. : 1950) 196(12) (2016) 5089-100.
26
[61] Y.S. Yi, Functional Role of Milk Fat Globule-Epidermal Growth Factor VIII in Macrophage-Mediated Inflammatory Responses and Inflammatory/Autoimmune Diseases, Mediators of inflammation 2016 (2016) 5628486. [62] E. Albus, K. Sinningen, M. Winzer, S. Thiele, U. Baschant, A. Hannemann, J. Fantana, A.K. Tausche, H. Wallaschofski, M. Nauck, H. Volzke, S. Grossklaus, T. Chavakis, M.C. Udey, L.C. Hofbauer, M. Rauner, Milk Fat Globule-Epidermal Growth Factor 8 (MFG-E8) Is a Novel Anti-inflammatory Factor in Rheumatoid Arthritis in Mice and Humans, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 31(3) (2016) 596-605. [63] Y. Zhang, M. Brenner, W.L. Yang, P. Wang, Recombinant human MFG-E8 ameliorates colon damage in DSS- and TNBS-induced colitis in mice, Lab Invest 95(5) (2015) 480-90. [64] M. Aziz, W.L. Yang, L.M. Corbo, W.W. Chaung, S. Matsuo, P. Wang, MFG-E8 inhibits neutrophil migration through alphavbeta(3)-integrin-dependent MAP kinase activation, International journal of molecular medicine 36(1) (2015) 18-28. [65] S. Akhtar, X. Wang, H.-F. Bu, X.-D. Tan, Role of MFG-E8 in Protection of Intestinal Epithelial Barrier Function and Attenuation of Intestinal Inflammation, in: P. Wang (Ed.), MFG-E8 and Inflammation, Springer Netherlands, Dordrecht, 2014, pp. 55-63. [66] S. Ishihara, R. Kusunoki, Y. Kinoshita, Anti-Inflammatory Role of MFG-E8 in the Intestinal Tract, in: P. Wang (Ed.), MFG-E8 and Inflammation, Springer Netherlands, Dordrecht, 2014, pp. 137-148. [67] S.R. Greisen, Y. Yan, A.S. Hansen, M.T. Venø, J.R. Nyengaard, S.K. Moestrup, M. Hvid, G.J. Freeman, J. Kjems, B. Deleuran, Extracellular Vesicles Transfer the Receptor Programmed Death-1 in Rheumatoid Arthritis, Frontiers in immunology 8 (2017) 851. [68] M. Uhlen, L. Fagerberg, B.M. Hallstrom, C. Lindskog, P. Oksvold, A. Mardinoglu, A. Sivertsson, C. Kampf, E. Sjostedt, A. Asplund, I. Olsson, K. Edlund, E. Lundberg, S. Navani, C.A. Szigyarto, J. Odeberg, D. Djureinovic, J.O. Takanen, S. Hober, T. Alm, P.H. Edqvist, H. Berling, H. Tegel, J. Mulder, J. Rockberg, P. Nilsson, J.M. Schwenk, M. Hamsten, K. von Feilitzen, M. Forsberg, L. Persson, F. Johansson, M. Zwahlen, G. von Heijne, J. Nielsen, F. Ponten, Proteomics. Tissue-based map of the human proteome, Science (New York, N.Y.) 347(6220) (2015) 1260419. [69] S. Liu, A.P. da Cunha, R.M. Rezende, R. Cialic, Z. Wei, L. Bry, L.E. Comstock, R. Gandhi, H.L. Weiner, The Host Shapes the Gut Microbiota via Fecal MicroRNA, Cell host & microbe 19(1) (2016) 32-43. [70] C. Carapito, P. Duek, C. Macron, M. Seffals, K. Rondel, F. Delalande, C. Lindskog, T. Freour, Y. Vandenbrouck, L. Lane, C. Pineau, Validating Missing Proteins in Human Sperm Cells by Targeted Mass-Spectrometry- and Antibody-based Methods, Journal of proteome research 16(12) (2017) 4340-4351. [71] K.M. Howard, R. Jati Kusuma, S.R. Baier, T. Friemel, L. Markham, J. Vanamala, J. Zempleni, Loss of miRNAs during processing and storage of cow's (Bos taurus) milk, J Agric Food Chem 63(2) (2015) 588-92.
27
FIGURE LEGENDS
Figure 1. Quantitative proteomic identification of milk proteins that can be sedimented by
differential ultracentrifugation. (A) Milk was subjected to differential ultracentrifugation, and the
35K and 100K pellets were lysed with Trizol LS to recover all proteins. After performing mass-
spectrometry analysis, the quantitative analysis of protein enrichment was performed using the
MaxQuant Andromeda algorithm. Panel adapted from previous work [9]. (B) Venn diagram
comparing the 35K and 100K pellets, in terms of protein enrichment. Proteins with a z-score higher
than 1.96 and a Benjamini-Hochberg adjusted Welch p-value below 0.05 were considered specific
to each subset. Panther GeneOntology (GO)-slim cellular component analysis was used to predict
the cellular origin of the proteins in each pellet, as indicated in the circles. The complete list of the
specific proteins and their respective enrichment in each pellets are provided in Supplementary
File S1. (C) Top 15 of the most enriched proteins found in both 35K and 100K pellets based on the
Exponentially Modified Protein Abundance Index (emPAI) normalized for each sample. The levels
for the emPAI in each pellets are provided in Supplementary File S2.
Figure 2. Comparative enrichment of canonical EV protein markers in the 35K and 100K
pellets. Comparison of the 35K and 100K pellets for their enrichment (ratio intensities pellet/total)
in different EV subset protein markers (light EVs, dense EVs, large EVs or multiple EVs markers),
for multivesicular body (MVB) markers/exosome-enriched proteins (tetraspanins, ESCRT complex
proteins, Rab family) or proteins necessary for docking and internalization of EVs (Integrins), as
defined previously [21, 34]. From inside to outside: 100K n1; 100K n2; 100K n3; 35K n1, 35K n2,
35K n3. The entire dataset and associated statistical analysis are provided as Supplementary File
S3.
Figure 3. Identification of the proteins differentially enriched between 35K and 100K pellets.
(A) Volcano plot representing the differential enrichment in milk proteins between the two pellets
28
as well as the most significantly enriched proteins for each of them (z-score > 1.96 and adjusted p-
value < 0.05). x axis = log2(35K/100K). y axis = −log10 (adjusted p-value). The horizontal line
indicates p-value = 0.05, vertical blue lines indicate log2(35K/100K) threshold limits as defined
through the z-score. Green and red dots correspond to the proteins above Z-score and below p-value
thresholds for each pellet. The blue dots identify the proteins common to both pellets. The name of
the most promising specific markers are displayed. (B) Heatmap and clustering of milk pellets
based on the ratio [log2(35K/100K)] of the most specific proteins for each pellets. The complete list
of the specific proteins and their respective enrichment in each pellets are provided in
Supplementary File S1.
Figure 4. The predicted molecular function of the proteins associated to the milk pellets.
Panther GeneOntology (GO)-slim molecular fonction analysis was used to predict the molecular
function of the proteins common to the two pellets (A) or specific to the 100K pellet (B). Functions
associated to the milk 100K pellet, for which fold change versus database > 2 and p-value < 0.05,
are displayed. No specific molecular function could be defined for 35K pellet-specific proteins.
29
FIGURES
Figure 1
30
Figure 2
31
Figure 3
32
Figure 4
33
TABLES
Table 1. Pathways involving the proteins common to both 35K and 100K pellets. The proteins
found in both pellets were submitted to the Reactome pathways analysis tool, and the top 20 of the
pathways most likely to involve 35K proteins are displayed in this table.
Table 2. Pathways involving 35K proteins. The proteins found in the 35K pellets were analyzed
with Reactome pathways analysis tool, and the top 20 of the pathways involving specific 35K
proteins are displayed in this table.
Table 3. Pathways involving 100K proteins. The proteins found in the 100K pellets were
analyzed with Reactome pathways analysis tool, and the top 20 of the pathways involving specific
100K proteins are displayed in this table.
34
FDR, False Discovery Rate. Table 1.
Pathway name Entities
found
Entities
Total
Entities
ratio
Entities
p-value
Entities
FDR
Reactions
found
Reactions
total
Reactions
ratio
GTP hydrolysis and joining of the 60S ribosomal subunit 73 122 0.009 1.11E-16 1.44E-14 3 3 0
Formation of a pool of free 40S subunits 63 106 0.008 1.11E-16 1.44E-14 2 2 0
Cap-dependent translation initiation 74 131 0.01 1.11E-16 1.44E-14 18 18 0.002
Eukaryotic translation elongation 57 104 0.008 1.11E-16 1.44E-14 9 9 0.001
Neutrophil degranulation 213 480 0.035 1.11E-16 1.44E-14 10 10 0.001
Eukaryotic translation initiation 74 131 0.01 1.11E-16 1.44E-14 20 21 0.002
Membrane trafficking 234 668 0.049 1.11E-16 1.44E-14 181 218 0.019
Vesicle-mediated transport 246 827 0.061 1.11E-16 1.44E-14 199 251 0.022 L13a-mediated translational silencing of ceruloplasmin expression 72 118 0.009 1.11E-16 1.44E-14 2 3 0
Axon guidance 188 583 0.043 1.11E-16 1.44E-14 166 297 0.026
Innate immune system 335 1,291 0.095 1.11E-16 1.44E-14 358 649 0.058 Regulation of expression of SLITs and ROBOs 92 183 0.014 1.11E-16 1.44E-14 7 20 0.002
Signaling by ROBO receptors 106 235 0.017 1.11E-16 1.44E-14 20 60 0.005 Nonsense mediated decay (NMD) independent of the exon junction complex (EJC)
55 101 0.007 4.44E-16 5.37E-14 1 1 0
Peptide chain elongation 54 99 0.007 7.77E-16 8.78E-14 5 5 0 AUF1 (hnRNP D0) binds and destabilizes mRNA 40 56 0.004 9.99E-16 1.06E-13 3 4 0
Immune system 513 2,537 0.187 1.11E-15 1.11E-13 919 1,458 0.13 Translation initiation complex formation 42 63 0.005 1.89E-15 1.77E-13 2 2 0
Ribosomal scanning and start codon recognition 42 66 0.005 8.66E-15 7.71E-13 2 2 0
Activation of the mRNA upon binding of the cap-binding complex and eIFs, and subsequent binding to 43S
42 67 0.005 1.41E-14 1.2E-12 6 6 0.001
35
FDR, False Discovery Rate. Table 2.
Pathway name Entities found
Entities Total
Entities ratio
Entities p-value
Entities FDR
Reactions found
Reactions total
Reactions ratio
Formation of annular gap junctions 2 11 0.001 1.65E-4 1.88E-2 2 2 0
Gap junction degradation 2 12 0.001 1.96E-4 1.88E-2 4 4 0
Apoptosis induced DNA fragmentation 2 13 0.001 2.3E-4 1.88E-2 9 11 0.001
Activation of DNA fragmentation factor 2 13 0.001 2.3E-4 1.88E-2 9 11 0.001
Cell-extracellular matrix interactions 2 19 0.001 4.88E-4 3.17E-2 3 10 0.001
Clathrin-mediated endocytosis 3 155 0.011 2.23E-3 9.5E-2 21 34 0.003
Phenylalanine and tyrosine catabolism 2 42 0.003 2.33E-3 9.5E-2 2 12 0.001
Apoptosis 3 177 0.013 3.24E-3 9.5E-2 10 122 0.011
Gap junction trafficking 2 50 0.004 3.27E-3 9.5E-2 4 20 0.002
Programmed cell death 3 185 0.014 3.67E-3 9.5E-2 10 135 0.012
Recycling pathway of L1 2 54 0.004 3.8E-3 9.5E-2 4 14 0.001
Apoptotic execution phase 2 54 0.004 3.8E-3 9.5E-2 9 54 0.005
Gap junction trafficking and regulation 2 54 0.004 3.8E-3 9.5E-2 4 24 0.002
Membrane trafficking 5 668 0.049 4.64E-3 9.78E-2 36 218 0.019
Cilium assembly 3 207 0.015 5.02E-3 9.78E-2 10 50 0.004
G2/M Transition 3 210 0.015 5.22E-3 9.78E-2 15 78 0.007
Mitotic G2-G2/M phases 3 212 0.016 5.36E-3 9.78E-2 15 80 0.007 Loss of proteins required for interphase microtubule organization from the centrosome 2 71 0.005 6.45E-3 9.78E-2 3 3 0
Loss of Nlp from mitotic centrosomes 2 71 0.005 6.45E-3 9.78E-2 2 2 0 Phenylketonuria 1 4 0 6.77E-3 9.78E-2 1 1 0
36
FDR, false discovery rate. Table 3.
Pathway name Entities found
Entities Total
Entities ratio
Entities p-value
Entities FDR
Reactions found
Reactions total
Reactions ratio
Glycosphingolipid metabolism 8 86 0.006 4.59E-10 5.79E-8 8 31 0.003
Sphingolipid metabolism 8 157 0.012 4.77E-8 3E-6 8 61 0.005
HS-GAG degradation 5 44 0.003 3.95E-7 1.66E-5 5 14 0.001
Collagen formation 6 104 0.008 1.29E-6 4E-5 13 77 0.007 Collagen biosynthesis and modifying enzymes 5 76 0.006 5.62E-6 1.4E-4 12 51 0.005
Heparan sulfate/heparin (HS-GAG) metabolism 5 88 0.006 1.14E-5 2.38E-4 5 32 0.003
Neutrophil degranulation 9 480 0.035 2.44E-5 4.39E-4 4 10 0.001
Glycosaminoglycan metabolism 6 182 0.013 3.05E-5 4.57E-4 11 82 0.007
Keratan sulfate degradation 3 22 0.002 5.77E-5 8.07E-4 2 6 0.001
Extracellular matrix organization 7 329 0.024 1E-4 1.2E-3 21 318 0.028
Innate immune system 13 1,291 0.095 2.06E-4 2.25E-3 20 649 0.058
Mucopolysaccharidoses 3 35 0.003 2.25E-4 2.25E-3 3 11 0.001
Terminal pathway of complement 2 8 0.001 3.4E-4 3.05E-3 4 5 0
Metabolism of carbohydrates 7 413 0.03 4E-4 3.6E-3 12 214 0.019
Keratan sulfate/keratin metabolism 3 52 0.004 7.1E-4 5.68E-3 2 15 0.001
Sialic acid metabolism 3 60 0.004 1.07E-3 7.5E-3 1 20 0.002
Platelet degranulation 4 137 0.01 1.12E-3 7.82E-3 3 11 0.001 Response to elevated platelet cytosolic Ca2+ 4 144 0.011 1.34E-3 9.39E-3 3 14 0.001
Hyaluronan uptake and degradation 2 18 0.001 1.68E-3 1.01E-2 3 9 0.001 Diseases of carbohydrate metabolism 3 80 0.006 2.42E-3 1.45E-2 3 34 0.003
37
SUPPLEMENTARY DATA
Supplementary File S1. Identification and specific enrichment analysis of extracellular vesicle
(EV)-associated proteins in milk pellets by LC-MS/MS.
Supplementary File S2. Quantification of the proteins found in each pellet based on the
Exponentially Modified Protein Abundance Index (emPAI) normalized for each sample.
Supplementary File S3. Relative enrichment of extracellular vesicle (EV)-associated proteins
in milk pellets.
