Abstract
Mycobacterium avium subsp. paratuberculosis (MAP) is the causative
agent of bovine paratuberculosis, a chronic granulomatous enteritis
leading to economic losses and posing a risk to human health due to its
zoonotic potential. The pathogen cannot reliably be detected by
standard methods, and immunological procedures during the infection are
not well understood. Therefore, the aim of our study was to explore
host–pathogen interactions in MAP-infected dairy cows and to improve
diagnostic tests. Serum proteomics analysis using quantitative
label-free LC-MS/MS revealed 60 differentially abundant proteins in
MAP-infected dairy cows compared to healthy controls from the same
infected herd and 90 differentially abundant proteins in comparison to
another control group from an uninfected herd. Pathway enrichment
analysis provided new insights into the immune response to MAP and
susceptibility to the infection. Furthermore, we found a higher
abundance of Cathepsin S (CTSS) in the serum of MAP-infected dairy
cows, which is involved in multiple enriched pathways associated with
the immune system. Confirmed with Western blotting, we identified CTSS
as a potential biomarker for bovine paratuberculosis. This study
enabled a better understanding of procedures in the host–pathogen
response to MAP and improved detection of paratuberculosis-diseased
cattle.
Keywords: Mycobacterium avium subsp. paratuberculosis, serum
proteomics, quantitative label-free LC-MS/MS, immune system,
host–pathogen response, Cathepsin S, bovine
1. Introduction
Bovine paratuberculosis, also known as Johne’s disease, is a bacterial
infectious disease caused by Mycobacterium avium subsp.
paratuberculosis (MAP), an intracellular pathogen [[38]1,[39]2].
Paratuberculosis manifests as a chronic granulomatous enteritis,
leading to high economic losses due to reduced milk production,
premature culling and decreased slaughter value [[40]3,[41]4,[42]5].
Transmission primarily occurs fecal–orally, whereby calves are more
susceptible to infection in the first few months of life [[43]6,[44]7].
During a long asymptomatic subclinical phase, infected cattle shed the
pathogen continuously or intermittently, posing a high risk of
spreading within the herd [[45]8]. Additionally, MAP is highly
resistant to environmental influences, with survival over several
months in fecal excretions and soil [[46]9].
The detection of MAP-infected cattle is problematic due to the
unreliability of rapid testing methods [[47]10]. Standard diagnostic
tests for identifying infected cattle can be divided into indirect and
direct pathogen detection methods [[48]10]. The common serology-based
test is the detection of MAP-specific antibodies in an enzyme-linked
immunosorbent assay (ELISA) [[49]10]. The reliable detection of
MAP-infected cattle in ELISA is restricted due to its limited
sensitivity and specificity, and the absent seroconversion in the
subclinical phase leads to a high rate of undetected diseased cattle
[[50]10,[51]11,[52]12]. The direct pathogen detection occurs through
polymerase chain reaction (PCR) or cultivation [[53]10,[54]11]. The PCR
method is also characterized by limited sensitivity and specificity
[[55]10], whereas the cultivation as a gold standard method reveals a
high specificity [[56]11]. However, the method is time-consuming due to
the slow growth of MAP, requiring cultivation for at least 8 weeks
[[57]13].
MAP has been found in various human foods, including raw and
pasteurized milk [[58]14,[59]15], infant formula [[60]16] and meat
[[61]17], raising concerns about its zoonotic potential. The
involvement of MAP in human autoimmune diseases like Crohn’s disease,
type 1 diabetes and Hashimoto’s thyroiditis is being discussed
[[62]18,[63]19,[64]20,[65]21,[66]22].
There are several different control and eradication strategies to
reduce the risk of within-herd transmission, especially to cattle of
susceptible age, and eliminate infected animals [[67]23,[68]24].
Vaccination against bovine paratuberculosis also reduces shedding of
MAP and thereby decreases prevalence [[69]25]. However, vaccination
interferes with an intradermal diagnostic test of bovine tuberculosis;
therefore, the vaccine is not available in most European countries
[[70]26,[71]27]. The strategy of eliminating infected animals through
testing and culling is also being pursued but only results in a limited
reduction in seroprevalence [[72]28]. To avoid MAP exposure to calves
and achieve a lower prevalence of MAP infections in dairy herds,
hygiene and management strategies, e.g., restricting calves’ access to
adult cattle and their feces, exist [[73]29]. Nevertheless, the success
of an eradication program based solely on hygiene and management
strategies is limited [[74]30].
Susceptibility and resistance to bovine paratuberculosis are associated
with genetic variation in interferon gamma (IFNγ) and interleukin (IL)
receptor 10, among other factors [[75]31,[76]32]. In a previous study,
we detected two different immune phenotypes in cattle in Germany
[[77]33]. These phenotypes, characterized by different immune
capacities, have previously been studied in detail with co-incubation
of MAP and peripheral blood mononuclear cells (PBMCs) from cows with
negative MAP infection status and different immunophenotypes [[78]34].
In the PBMCs of the control group, we showed a higher abundance of
IL-12-mediated signaling pathways, while we observed a higher abundance
of CCR4-NOT transcription complex, subunit 1 in immune deviant cows,
which may promote infectious diseases by repressing MHC class II
expression [[79]34,[80]35]. In another co-incubation study of MAP and
the PBMCs of cows with different MAP infection statuses, we detected a
defensive immune response of MAP-infected cows to MAP, realized by an
increased abundance of the involved proteins toll-like receptor 2 and
major histocompatibility complex (MHC) class II [[81]36]. These
findings underscore the complexity of the immune response to MAP and
highlight the need for further investigation into factors that
contribute to the persistence of MAP in cattle.
Serum proteomics is currently often used for the identification of
potential disease biomarkers and the detection of immunological
processes during specific diseases [[82]37,[83]38]. Among others, this
method already revealed IL-8 and Pentaxin as potential biomarkers for
bovine tuberculosis in the serum of cattle [[84]39]. In our proteomics
study, we analyzed the serum of naturally MAP-infected dairy cows and
healthy controls from the same infected herd and additional controls
from another uninfected herd using label-free LC-MS/MS. The primary aim
of our study was to discover proteomic changes in the serum of dairy
cows with different MAP infection statuses to better characterize the
host–pathogen response of cattle, particularly to MAP. In addition to
this, this study aimed to identify a serum biomarker for the reliable
detection of MAP-infected cattle for the control of bovine
paratuberculosis.
2. Materials and Methods
2.1. Animals and Detection of MAP Infection Status
In this study, samples from 16 dairy cows from a MAP-infected farm were
used. Serum, milk and fecal samples were collected. Different standard
methods for detecting MAP were applied to categorize these cows into
two groups based on their MAP infection status. Healthy controls (n =
7) showed no antibodies against MAP in ELISA with serum and milk
samples (cattletype MAP Ab, Qiagen, Hilden, Germany; Indirect, IDVet,
Grabels, France), and after cultivation of fecal samples on commercial
Herrold’s Egg Yolk Agars (HEYM agar, Becton Dickinson, Heidelberg,
Germany) for 12 weeks, no bacterial growth could be observed. Cows with
positive results (n = 9) were grouped as MAP-infected cows. For mass
spectrometry analysis, the serum of seven MAP-infected cows and seven
healthy controls were used. All animals were from the same farm and
were kept under the same environmental conditions. For additional mass
spectrometry analysis with a further control group, the serum of 21
dairy cows from an uninfected farm was examined. All cows from this
farm showed negative results in ELISA with serum and milk samples.
Additionally, no positive samples could ever be detected in the
regularly tested fecal samples and sock swab samples from this farm. To
verify differences in CTSS expression in a Western blot analysis, five
dairy cows from another infected farm were examined. The MAP status of
these cows was determined by cultivating fecal samples on HEYM agar for
12 weeks and conducting ELISA with serum samples. In detail, two cows
were categorized as healthy controls, and three cows were categorized
as MAP-infected. A comprehensive overview of the MAP infection status
of the cows in each group and a schematic of the experimental design
are provided in [85]Supplementary Table S1. Sampling of whole blood by
venipuncture and experimental protocols were approved by the local
authority, the Government of Upper Bavaria, permit no.
ROB-55.2-2532.Vet_03-17-106. No experimental animals were used in this
study. All animals were kept for the purpose of milk production.
Permission from the dairy farms to use the blood samples from their
animals for study purposes was obtained.
2.2. Sample Digestion for Differential Proteome Analysis
The serum was proteolyzed with trypsin using the PreOmics iST Kit
(PreOmics GmbH, Martinsried, Germany) according to the manufacturer’s
specifications. The resulting peptides were dried using SpeedVac and
stored at −80 °C until mass spectrometric analysis.
2.3. Mass Spectrometric Analysis and Label-Free Quantification
500 ng of peptides per sample was analyzed in data-independent (DIA)
mode on a Q Exactive HF mass spectrometer (Thermo Fisher Scientific
Inc., Waltham, MA, USA) coupled online to a RSLC (Ultimate 3000, Thermo
Fisher Scientific Inc.) HPLC system. Samples were automatically
injected and loaded onto a nano-trap column (300-µm inner diameter (ID)
× 5 mm, packed with Acclaim PepMap100 C18, 5 µm, 100 Å; LC Packings,
Sunnyvale, CA, USA) before separation using reversed-phase
chromatography (Acquity UPLC M-Class HSS T3 Column 75 µm ID × 250 mm,
1.8 µm; Waters, Eschborn, Germany) at 40 °C. Peptides were eluted from
the column at a flow rate of 250 nl/min using increasing concentrations
of ACN in 0.1% formic acid from 3 to 40% over a 45 min gradient. The
DIA method consisted of a survey scan from 300 to 1500 m/z at 120,000
resolution and an automatic gain control (AGC) target of 3e6 or
120-millisecond maximum injection time. Fragmentation was carried out
with higher energy collisional dissociation with a target of 3e6 ions,
determined with predictive AGC. With an AGC target of 3e6 and automatic
injection time, precursor peptides were isolated using 37 variable
windows ranging from 300 to 1650 m/z at 30,000 resolution. Spectra were
recorded in profile type with a normalized collision energy of 28.
DIA files were processed with Spectronaut (Version 18, Biognosys,
Schlieren, Switzerland) as direct DIA analysis against the Ensembl cow
database (Release 111, June 2023, 23,842 sequences) using BSG factory
settings for Pulsar search. For DIA analysis, default settings were
applied. For quantification, precursor filtering was set on Qvalue, the
proteotypicity filter was on, and there was run wise imputing, protein
LFQ method Quan2 algorithm, quantity MS level MS2, quantity type area,
cross run normalization and mean top3 quantity. A ratio-based unpaired
t-test was performed on the peptide level. The results were considered
significant as Benjamini–Hochberg-corrected p-values at q ≤ 0.05.
2.4. Data Analysis
Volcano plots were generated using GraphPad Prism Software (version
5.04, GraphPad Software, San Diego, CA, USA). A pathway enrichment
analysis was conducted with open-source software ShinyGO (version 0.77,
[86]http://bioinformatics.sdstate.edu/go/, accessed on 13 December
2023). The p-value cutoff (FDR) was set to 0.05, and the best-matching
species (cow) was used. The p-value for the enrichment analysis was
calculated using hypergeometric distribution followed by FDR
correction. A pathway analysis with open-source software Reactome
(version 87, [87]https://reactome.org/, accessed on 13 December 2023)
was performed on human orthologues gene names. Over-representation of
pathways was determined with hypergeometric distribution corrected for
FDR using the Benjamini–Hochberg procedure.
2.5. Western Blots
To 0.5 µL serum 1× Laemmli-buffer (50 mM tris-HCl, 1% SDS (AppliChem,
Darmstadt, Germany), 10% glycerol (SERVA, Heidelberg, Germany), 4 mM
2-mercaptoethanol (Sigma-Aldrich, Taufkirchen, Germany), bromophenol
blue (Sigma-Aldrich)) was added. Protein expression was analyzed
separately for each biological replicate of controls (n = 9) and
MAP-infected cows (n = 12), as well as for each technical replicate (n
= 2). Proteins were separated with SDS-PAGE on 15% gels and blotted
semi-dry onto PVDF membranes (GE Healthcare, Freiburg, Germany). To
prevent unspecific binding, membranes were blocked with 4% PVP-T
(Sigma-Aldrich). After washing, blots were incubated with a rabbit
anti-bovine Cathepsin S polyclonal antibody (rabbit polyclonal,
MyBioSource, San Diego, CA, USA; 1:600) at 4 °C overnight. After three
washing steps with PBS-T (NaCl 136.9 mM (Sigma-Aldrich), Na[2]HPO[4] ×
2H[2]O 8.1 mM (AppliChem), KH[2]PO[4] 1.4 mM (Sigma-Aldrich), KCl 2.6
mM (Sigma-Aldrich), 0.05% Tween 20 (AppliChem); pH 7.4), HRP-conjugated
anti-rabbit IgG (Fc) secondary antibody (Bio-Rad, Feldkirchen, Germany;
1:10,000) preadsorbed with 5% heat-denatured bovine serum was used for
incubation for one hour at room temperature. Following six washing
steps, signals were detected with enhanced chemiluminescence on an
Amersham Imager 680 (GE Healthcare). Quantification of Western blot
signals was performed using ImageQuant TL v8.2.0 software (GE
Healthcare). The Gaussian distribution was determined using the
Kolmogorov–Smirnov test. Due to the non-normal distribution, statistics
were performed using the Mann–Whitney test. The result was considered
significant at p ≤ 0.05. Data were processed, analyzed and visualized
with GraphPad Prism software (version 5.04, GraphPad Software).
3. Results
3.1. Serum Proteomics Reveals Differentially Abundant Proteins in Serum of
MAP-Infected Dairy Cows When Compared to Two Healthy Control Groups from
Farms with Divergent MAP Infection Status
Using quantitative mass spectrometric analysis, we identified a total
of 394 proteins in bovine serum. Of these, 60 proteins were
significantly (q ≤ 0.05) differently abundant between MAP-infected
dairy cows (n = 7) and healthy controls (n = 7) from the same dairy
farm. In detail, 20 proteins were more abundant in the control group,
while 40 proteins showed increased abundance in MAP-infected cows
([88]Figure 1a). Furthermore, an additional investigation was carried
out with another control group comprising 21 cows from an uninfected
herd. This revealed significant differential expression of 90 proteins
(q ≤ 0.05). Of these, 38 proteins were upregulated in the control
group, representing healthy cows from the uninfected herd, while 52
proteins showed upregulation in the MAP-infected cows ([89]Figure 1b).
Figure 1.
[90]Figure 1
[91]Figure 1
[92]Open in a new tab
Volcano plot of all 394 identified proteins. (a) Comparing MAP-infected
cows with healthy controls from the same dairy farm, 40 proteins showed
significantly (q ≤ 0.05) higher abundance (fold change > 1.5, red
dots), while 20 proteins showed significantly lower abundance (fold
change < 0.
[MATH: 6¯
:MATH]
, blue dots). (b) Fifty-two proteins were significantly (q ≤ 0.05) more
abundant in MAP-infected cows compared to healthy controls from another
dairy farm with MAP-uninfected status (fold change > 1.5, red dots),
and 38 proteins were more abundant in the control group (fold change <
0.
[MATH: 6¯
:MATH]
, blue dots). Significantly (q ≤ 0.05) more abundant proteins with
strong associations to immune system pathways are displayed with their
bovine gene names: Cathepsin S (CTSS), major histocompatibility
complex, class II, DR alpha (BOLA-DRA) and Fc fragment of IgG,
high-affinity Ia, receptor (FCGR1A).
Considering the proteins with higher abundance in MAP-infected dairy
cows, 40 proteins were more abundant compared to healthy controls from
the same infected farm, whereas 52 proteins showed higher abundance in
comparison to healthy controls from another uninfected herd. Of these,
22 proteins overlapped when compared to both healthy control groups
([93]Figure 2). The overlapping proteins included immunologically
functional proteins such as “Cathepsin S” (CTSS), “major
histocompatibility complex, class II, DR alpha” (BOLA-DRA) and “Fc
fragment of IgG, high affinity Ia, receptor” (FCGR1A) ([94]Figure
1a,b).
Figure 2.
[95]Figure 2
[96]Open in a new tab
Significantly (q ≤ 0.05) more abundant proteins in MAP-infected cows:
40 proteins were more abundant compared to healthy controls from the
same infected herd (indicated by white and light blue) and 52 compared
to healthy controls from another uninfected herd (light blue and dark
blue). Of these, 22 proteins overlapped when compared to both control
groups (light blue).
3.2. Proteins with Significantly Higher Abundance in MAP-Infected Cows
Associate with Immune System Pathways
To gain deeper insights into the functional effects of the
differentially abundant proteins, we conducted a pathway enrichment
analysis of all proteins with significantly (q ≤ 0.05) higher abundance
(ratio MAP/control > 1.5) in MAP-infected cows compared to both healthy
control groups using ShinyGO. This analysis revealed an enrichment of
proteins in pathways associated with bacterial infections, such as
“Tuberculosis” (compared to controls from the infected herd)
([97]Figure 3a), “Staphylococcus aureus infection” (compared to both
control groups) ([98]Figure 3a,b) and “Salmonella infection” (compared
to controls from the uninfected herd) ([99]Figure 3b). Additional
enrichment was observed in pathways belonging to the immune system such
as “Phagosome” (compared to both control groups) ([100]Figure 3a,b) and
“Antigen processing and presentation” (compared to controls from the
uninfected herd) ([101]Figure 3b). There also was an enrichment of the
pathways “Pentose phosphate pathway” (compared to controls from the
infected herd) ([102]Figure 3a) and “Apoptosis” (compared to both
control groups) ([103]Figure 3a,b).
Figure 3.
[104]Figure 3
[105]Figure 3
[106]Open in a new tab
Pathway enrichment analysis of significantly (q ≤ 0.05) higher abundant
proteins in MAP-infected cows compared to (a) healthy controls from the
same infected herd and (b) healthy controls from another uninfected
herd. Functional enrichment displays the ten most significant
categories of all available gene sets. Hierarchical clustering was
conducted using ShinyGO v0.77. The y-axis lists the assigned pathways
in order of the enrichment of fold change ratio (FCR). The x-axis shows
the FDR values for the enrichment of the respective pathways. The color
chart shows fold enrichment for each pathway, with the size of dots
corresponding to the number of genes associated with each pathway.
Pathway enrichment analysis was conducted with bovine gene names.
Further pathway analysis using Reactome confirmed an association with
immune system pathways, showing enrichment in all three major
subdomains: “Adaptive Immune System”, “Innate Immune System” and
“Cytokine Signaling Immune System” of Reactome superpathway “Immune
System” compared to both control groups ([107]Figure 4). Specifically,
we identified over-representation in pathways “MHC class II antigen
presentation”, “Endosomal/Vacuolar pathway”, “Neutrophil
degranulation”, “Gene and protein expression by JAK-STAT signaling
after Interleukin-12” and “Interferon Signaling” with “Interferon gamma
signaling” ([108]Figure 4a,b). Pathways in MAP-infected cows that were
enriched only in comparison to healthy controls from the same infected
herd were “Butyrophilin (BTN) family interactions”, “Antigen activates
B Cell Receptor (BCR) leading to generation of second messengers”,
“CD22 mediated BCR regulation”, “Fcgamma receptor (FCGR) dependent
phagocytosis” with “Role of phospholipids in phagocytosis”, “FCGR
activation” and “Regulation of actin dynamic for phagocytic cup
formation”, “Transfer of LPS from LBP carrier to CD14” and “Classical
antibody-mediated complement activation” ([109]Figure 4a). In contrast,
compared to healthy controls from another uninfected herd, the pathways
“PKR-mediated signaling” and “Regulation of Complement cascade” were
enriched in MAP-infected cows. ([110]Figure 4b).
Figure 4.
[111]Figure 4
[112]Figure 4
[113]Open in a new tab
Voronoi diagram illustrating Reactome analysis results of enriched
pathways in MAP-infected dairy cows compared to (a) healthy controls
from the same infected herd and (b) healthy controls from another
uninfected herd. The enlarged polygon of Reactome superpathway “Immune
System” is shown. Pathway enrichment analysis was conducted with human
orthologue gene names of proteins with significantly (q ≤ 0.05) higher
abundance in MAP-infected dairy cows (ratio MAP/control > 1.5). Color
intensity represents the p-value (p ≤ 0.05) of the statistical test for
over-representation, as illustrated by the color bar. Polygons colored
in dark grey visualize pathways without significant over-representation
(p > 0.05), while light grey areas represent pathways without related
proteins.
3.3. CTSS Involved in Pathways with Strong Association to the Immune System
To delve deeper into the detected pathways, we examined the
significantly upregulated proteins (q ≤ 0.05, ratio MAP/control > 1.5)
in MAP-infected cows and the enriched pathways in which they were
involved. We found that CTSS was participating in multiple pathways
belonging to both the adaptive immune system, such as “MHC class II
antigen presentation”, and the innate immune system, such as
“Neutrophil degranulation” ([114]Table 1). Additionally, CTSS was
involved in the disease pathway “Tuberculosis” and programmed cell
death pathway “Apoptosis”. The function of CTSS in multiple enriched
pathways was observed in comparison to both control groups. All
enriched proteins associated with the respective pathway and the total
number of genes in each pathway are listed in [115]Table 1.
Table 1.
Significantly (q ≤ 0.05) abundant proteins involved in significantly (p
≤ 0.05) enriched pathways in the serum of MAP-infected dairy cows
compared to healthy controls from the same infected herd and also to
another uninfected herd. Superordinate pathways consisting of one or
more individual pathways were excluded.
Comparison Group Enriched Pathway Pathway Genes
Total Proteins
Healthy controls
(Infected herd) Pentose phosphate pathway 28 GPI, ALDOA
Phagosome 160 ACTG1, FCGR1A, HLA-DRA, CTSS, TUBA4A, SFTPD
Apoptosis 140 ACTG1, CTSC, SEPTIN4, CTSS, TUBA4A
Staphylococcus aureus infection 98 FCGR1A, HLA-DRA, KRT25
Tuberculosis 182 FCGR1A, HLA-DRA, LBP, CTSS
Neutrophil degranulation 478 GPI, CRISP3, HP, ALDOA, PGLYRP1, PRDX6,
CTSS, CTSC
MHC class II antigen presentation 137 HLA-DRA, TUBA4A, CTSS, CTSC
Regulation of actin dynamics for phagocytic cup formation 158 IGKC,
FCGR1A, IGKV2-30, ACTG1
Interferon gamma signaling 177 HLA-DRA, FCGR1A
FCGR activation 103 IGKC, FCGR1A, IGKV2-30
Role of phospholipids in phagocytosis 129 IGKC, FCGR1A, IGKV2-30
Transfer of LPS from LBP carrier to CD14 3 LBP
CD22 mediated BCR regulation 72 IGKC, IGKV2-30
Gene and protein expression by JAK-STAT signaling after Interleukin-12
73 LCP1
Butyrophilin (BTN) family interactions 12 PPL
Classical antibody-mediated complement activation 97 IGKC, IGKV2-30
Antigen activates B Cell Receptor (BCR) leading to generation of second
messengers 103 IGKC, IGKV2-30
Endosomal/Vacuolar pathway 15 CTSS
Uninfected herd Staphylococcus aureus infection 98 KRT17, KRT14,
FCGR1A, HLA-DRA, MASP1, KRT10, KRT25, KRT24
Phagosome 160 ACTG1, TUBB, FCGR1A, HLA-DRA, CTSS, TUBB1, HLA-G, TUBA4A,
SFTPD
Estrogen signaling pathway 129 ADCY6, KRT17, KRT14, KRT10, HSPA1A,
KRT25, KRT24
Antigen processing and presentation 78 HLA-DRA, CTSS, HLA-G, HSPA1A
Gap junction 87 ADCY6, TUBB, TUBB1, TUBA4A
Viral myocarditis 69 ACTG1, HLA-DRA, HLA-G
Gastric acid secretion 72 ADCY6, ACTG1, CA2
Complement and coagulation cascades 82 PROCR, MASP1
Apoptosis 140 ACTG1, CTSC, CTSS, TUBA4A
Salmonella infection 250 TXN, ACTG1, TUBB, TUBB1, TUBA4A
Neutrophil degranulation 478 SERPINA3, SERPINB1, TUBB5, CRISP3, COTL1,
HBB, LYZ, PRDX6, CTSS, CTSC, HSPA1A
MHC class II antigen presentation 137 TUBB5, TUBB1, HLA-DRA, TUBA4A,
CTSS, CTSC
Interferon gamma signaling 177 HLA-DRA, FCGR1A, HLA-G
PKR-mediated signaling 88 TUBB5, TUBB1, TUBA4A, HSPA1A
Endosomal/Vacuolar pathway 15 HLA-G, CTSS
Regulation of Complement cascade 139 CFH, MASP1
Gene and protein expression by JAK-STAT signaling after Interleukin-12
stimulation 73 LCP1
[116]Open in a new tab
Column 1 (control group) indicates the control group to which the
comparison was made. Column 2 (enriched pathway) contains the names of
enriched pathways in ShinyGO analysis or in Reactome analysis. Column 3
(pathway genes total) displays the total number of genes in each
pathway. Column 4 (proteins) shows the proteins in human orthologues
gene names clustering to the enriched pathways.
3.4. Detection of Increased Abundance of CTSS in Serum of MAP-Infected Dairy
Cows Using LC-MS/MS and Western Blotting
In the serum of MAP-infected dairy cows, we observed a 1.8 higher
abundance of CTSS (q = 0.0001) compared to healthy controls from the
same infected herd and a 1.5-fold increase in CTSS (q = 0.0005)
compared to healthy controls from another uninfected herd using
LC-MS/MS analysis. Confirmation of the observed elevated abundance of
CTSS using Western blotting revealed a 3.4-fold enrichment of CTSS (p =
0.008) in the serum of MAP-infected dairy cows compared to healthy
controls ([117]Figure 5).
Figure 5.
[118]Figure 5
[119]Figure 5
[120]Open in a new tab
Elevated CTSS abundance was verified and quantified with Western
blotting. (a) CTSS expression was significantly (** p ≤ 0.01) increased
in serum of MAP-infected dairy cows (n = 12, blue column) compared to
healthy controls (n = 9, white column, set to factor 1). (b)
Representative Western blot showed higher abundance of CTSS in serum of
MAP-infected dairy cows compared to healthy controls.
4. Discussion
Paratuberculosis is a severe and incurable disease that affects cattle
and other ruminants, leading to significant economic losses
[[121]4,[122]40,[123]41]. Moreover, due to the spreading of MAP and its
potential entry into the food chain, it poses a risk to human health
[[124]22]. Therefore, in accordance with the One Health approach, there
is a pressing need to deepen our understanding of the immune response
of cattle to MAP and to improve diagnostic methods for reliable disease
detection. Serum is a commonly used biological material for diagnostic
procedures and offers valuable insights into immune response through
the examination of metabolites and proteins [[125]42].
In this study, we identified differently abundant proteins in the serum
of MAP-infected dairy cows and controls from herds with divergent MAP
infection status. We assumed that the control group from the infected
herd had been exposed to MAP but had remained uninfected, whereas the
control group from the uninfected herd had never had contact with the
pathogen. This allowed us to discern differences in host–pathogen
interactions and detect immune reactions essential for the successful
elimination of MAP.
The pathway enrichment analysis revealed an association of
significantly higher abundant proteins with immune system-related
pathways. Notably, the pathway “Fcgamma receptor (FCGR) dependent
phagocytosis” and all its subordinated pathways “Regulation of actin
dynamics for phagocytic cup formation”, “FCGR activation” and “Role of
phospholipids in phagocytosis” were enriched in MAP-infected cows
compared to healthy controls from the infected herd ([126]Figure 3a).
None of these pathways were enriched in comparison to the control group
from the uninfected herd ([127]Figure 3b). These findings suggest a
potential influence of the “Fcgamma receptor (FCGR) dependent
phagocytosis” pathway on susceptibility to bovine paratuberculosis.
Since MAP enters the organism through phagocytosis by macrophages, the
inhibition of this pathway might hinder MAP persistence within the
organism. This may be a crucial insight because the exact mechanisms
enabling MAP survival within host cells remain elusive. The agent is
suspected of persisting inside the phagosomes within these cells, as
has already been proven for Mycobacterium tuberculosis (Mtb)
[[128]43,[129]44]. The exact mechanisms enabling MAP survival are
unknown, but it is assumed that MAP prevents maturation and
acidification of the phagosome and blocks phagolysosome fusion, as
shown for Mtb [[130]43,[131]44,[132]45]. Consequently, the phagocytosis
of MAP is a critical factor of infection.
The functional enrichment of “FCGR dependent phagocytosis” in
MAP-infected cows compared to healthy cows from the infected herd
showed that inhibition of this pathway might contribute to overcoming
the persistence of MAP in the organism. In this enriched pathway group,
the higher abundant protein FCGR1A, also known as CD64, in MAP-infected
cows compared to both control groups ([133]Figure 1), was described in
several studies of Mycobacteriaceae-induced diseases
[[134]46,[135]47,[136]48,[137]49]. In cattle, FCGR1A is expressed in
leukocytes, especially in macrophages [[138]50]. Soluble FCGR1A was
detected in human serum [[139]51]. We are the first to describe FCGR1A
in bovine serum.
The function of FCGR1A in cattle has not been fully researched.
However, in humans, it is involved in various immune effects, such as
phagocytosis, production of reactive oxygen species, cytokine
expression and antigen presentation [[140]52]. Studies have
demonstrated an increase in FCGR1A expression in the transcriptome of
ileocaecal lymph nodes in sheep diseased with paratuberculosis
[[141]46]. So far, the significance of the increase in FCGR1A in bovine
paratuberculosis is unknown. A study with tissue samples of cows with
bovine tuberculosis caused by Mycobacterium bovis (Mb) showed the
involvement of FCGR1A in the immune response against Mb [[142]47]. In
the peripheral blood of human children with intrathoracic tuberculosis
induced by Mtb, FCGR1A levels correlated with the extent of the disease
[[143]48]. An increase in high-affinity FCGR1 expression could be
induced with IFNγ in humans [[144]53]. Based on an increased level of
IFNγ in paratuberculosis-diseased cattle [[145]54], we suspected an
IFNγ-induced expression of FCGR1A in cattle and the extracellular
shedding of soluble CD64.
Due to an improved control of Mtb in FCGR1-knockout mice [[146]49] and
a decrease in FCGR1A concentrations after anti-tuberculosis treatment
in humans [[147]48], we hypothesized that soluble FCGR1A had a negative
impact on the immune response to MAP and facilitated their
intracellular survival by interacting with soluble immune complexes and
blocking Fc-dependent immune reactions [[148]55]. Another soluble
isoform of FCGR, soluble FCGR2b, inhibited antibody production in
murine spleen cells [[149]56], which could be transferred to soluble
CD64, resulting in an additional impairment of adaptive immune
response.
The effects of soluble FCGR1A in immune response, especially against
MAP, should be explored more closely. Above all, new insights into its
impact on susceptibility to bovine paratuberculosis could arise.
However, in terms of its suitability as a specific biomarker for
Mycobacteriaceae-induced diseases, especially bovine paratuberculosis,
FCGR1A is not suitable due to its characterization as a diagnostic
biomarker of infection and sepsis in humans [[150]57,[151]58].
Enriched pathway analysis provided an over-representation of
“neutrophil degranulation” in MAP-infected cows compared to both
healthy control groups ([152]Figure 4a,b). The role of neutrophils in
bovine paratuberculosis is still unknown. Neutrophils are highly
effective cells of the innate immune system and kill microorganisms
through various functions, including phagocytosis, degranulation,
production of reactive oxygen species (ROS) and neutrophil
extracellular trap (NET) release [[153]59]. An increased migration of
neutrophils in lesions at the early stages of the infection with MAP
was shown in experimentally infected calves [[154]60]. In a gene
ontology analysis of long non-coding RNA target genes in macrophages of
cattle with paratuberculosis, an enrichment of biological processes
associated with neutrophils was found [[155]61].
After analyzing the proteins involved in the enriched pathway
“neutrophil degranulation” including peroxiredoxin 6 (PRDX6),
cysteine-rich secretory protein 3 (CRISP3) and Cathepsin C (CTSC)
([156]Table 1), it is evident that they cover various neutrophil
functions. During neutrophil activation in humans, PRDX6 translocates
to the plasma membrane and increases the production of ROS [[157]62].
CRISP3 was found in granules of human neutrophils, which could suggest
an antimicrobial role of CRISP3 in innate immune responses [[158]63].
Furthermore, human CTSC is essential for optimal NET formation
[[159]64]. In a strong innate response to MAP, caprine neutrophils use
all their effector functions—phagocytosis, chemotaxis, degranulation
and NET release [[160]65]. Oral vaccination against MAP in a rabbit
infection model stimulated the phagocytosis rate of neutrophils and
additionally increased NET release against MAP and non-related
pathogens [[161]66].
The involvement of neutrophils and their effector functions in early
and advanced stages of bovine paratuberculosis is confirmed by these
findings and the over-represented neutrophil degranulation in our
pathway enrichment analysis. We put forward the hypothesis that the
immune system of MAP-infected cattle attempts to combat the disease
through increased activation of neutrophils, but the immunological
functions of neutrophils are not effective in the host defense against
MAP. MAP can escape the NETs by degrading NETs through an extracellular
DNAse, which promotes the colonization of MAP and the formation of
granulomas in mice [[162]67]. Additionally, bovine neutrophils are
effective in killing and defending against MAP, but in the presence of
macrophages, the killing rate worsens, and levels of pro-inflammatory
cytokines IL-1β and IL-8 are lower, leading to fewer pathological
injuries [[163]68].
These studies support our hypothesis and even describe a worsening of
the disease due to neutrophil effector functions. Further research
should analyze the mechanism of the host–pathogen immune response of
neutrophils against MAP and their possible influence on the degradation
of bovine paratuberculosis.
In our proteomics study, we detected a higher abundance of CTSS in the
serum of MAP-infected dairy cows compared to both healthy control
groups ([164]Figure 1a,b). Pathway enrichment analysis revealed the
involvement of CTSS in multiple enriched pathways. The involvement of
CTSS in enriched pathways was observed in comparison to both healthy
control groups ([165]Table 1), leading us to select CTSS as a candidate
for a biomarker in paratuberculosis-infected cattle. CTSS is a
lysosomal cysteine peptidase, primarily found in immune cells,
including professional antigen-presenting cells, B-cells, dendritic
cells and macrophages. It plays a role in extracellular matrix
remodeling and regulates MHC class II antigen presentation [[166]69].
Thereby, CTSS proteolyzed the MHC class II-associated chaperone
invariant chain Ii and produced class II invariant chain peptide (CLIP)
[[167]70,[168]71]. The CLIP fragment is exchanged for antigenic peptide
fragments, and the peptide-loaded MHC class II is transported to the
plasma membrane where it will be recognized by CD4^+ T cells
[[169]70,[170]71].
In an in vitro study with human macrophages, infection with
Mycobacterium bovis bacillus Calmette–Guérin (BCG) induced the
inhibition of CTSS with IL-10 and consequently reduced MHC class II
antigen presentation [[171]72]. In contrast, the infection with a BCG
strain that secreted active human CTSS led to an increased MHC class II
mycobacterial antigen presentation on the surface of macrophages
[[172]73], which shows that BCG does not directly inhibit MHC class II
presentation, but its expression is dependent on CTSS. In contrast to
infection with lower pathogen vaccine strain BCG, we detected a higher
abundance of CTSS in MAP-infected dairy cows ([173]Figure 1a,b). We
infer that the expression of CTSS could correlate with the
pathogenicity of Mycobacteriaceae.
In addition to a higher abundance of CTSS, we could detect a higher
abundance of bovine MHC class II (BOLA-DRA), which indicates a
correlation of CTSS with expression of MHC class II antigen
presentation. IL-10 expression by bovine macrophages was also reported
in an in vitro infection assay with MAP, where neutralization of IL-10
resulted in an elevated killing rate of MAP, increased expression of
tumor necrosis factor alpha, IL-8, IL-12 and MHC class II, higher rate
of phagosome acidification and apoptotic cells [[174]74]. This led to
the hypothesis that IL-10 also inhibited the expression of CTSS in
bovine macrophages, causing an inferior antigen presentation at the
plasma membrane, phagosome acidification and apoptosis rate, where CTSS
could be assigned to the corresponding pathways ([175]Table 1).
Interestingly, the pathway “MHC class II antigen presentation”, also
associated with CTSS and BOLA-DRA, was enriched in the serum of
MAP-infected cows compared to both healthy control groups ([176]Table
1).
In vitro, MAP-infected bovine macrophages showed downregulation of MHC
class II expression, which appears to be a strategy of MAP to secure
their survival by inhibiting efficient mycobacterial antigen
presentation to T cells [[177]75]. In our previous in vitro study with
PBMCs co-incubated with MAP, we observed a higher abundance of MHC
class II complex proteins in MAP-resistant cows compared to
persistently MAP-infected cattle [[178]36], which describes an
association of the expression of MHC class II with resistance to MAP.
In our serum proteome analysis, we detected a higher abundance of
BOLA-DRA in the serum of MAP-infected cows compared to both healthy
control groups ([179]Figure 1a,b).
The macrophages of mice infected with Mtb or BCG shedded MHC class II
in plasma membrane-derived microvesicles and exosomes in an
ATP-dependent manner [[180]76]. Both organelles were able to present
peptide-MHC class II complexes to T cells [[181]76]. This study reveals
an alternative process of an effective antigen presentation in
mycobacterial infections. We hypothesize that in cattle infected with
MAP, MHC class II is expressed at a lower level on the plasma membrane
of antigen-presenting cells but is instead released in high amounts
into the extracellular space for an alternative MAP-antigen presenting
mechanism. Additional in vitro studies are necessary for characterizing
the function of exosomes and microvesicles in mycobacterial antigen
presentation in cows.
For the first time, we described a higher abundance of CTSS in serum in
association with bovine paratuberculosis. We cannot yet explain why
CTSS was secreted to a higher degree in the extracellular space in
paratuberculosis-diseased cattle, but there could be a context of MHC
class II in exosomes and microvesicles and the possible alternative
mechanism of antigen presentation. So far, the mechanism of secretion
and the function of extracellular CTSS in bovine host immune response
to MAP are still unknown and need further investigation.
In our pathway enrichment analysis, CTSS, BOLA-DRA and FCGR1A revealed
a strong association with immune system pathways, whereby we chose them
as potential candidates. In relation to clinical practice, the
comparison between MAP-infected dairy cows and healthy controls from
the same infected herd was more relevant. In this group comparison,
CTSS showed the highest q-value of all these three proteins. Therefore,
we chose CTSS, which provides the best conditions for a potential
biomarker for bovine paratuberculosis to be validated.
In our Western blot analysis, we loaded samples based on volume to
validate our findings in a context relevant to potential practical
applications. Although we performed normalization based on equal total
protein content in our proteomics assay, we still observed a
significantly higher abundance of CTSS in the serum of MAP-infected
dairy cows using Western blotting ([182]Figure 5a,b). We detected CTSS
at a lower molecular weight, as expected, in recombinant CTSS expressed
in Escherichia coli (E. coli). However, the molecular weight of
naturally occurring proteins can be altered due to post-translational
modifications [[183]77], that do not occur in E. coli-expressed
proteins [[184]78]. The mature enzymatically active form of human CTSS
has 24 kDa, lower than that of inactive CTSS [[185]79]. Therefore, we
hypothesized an enzymatic activity of CTSS secreted into the
extracellular space.
Verification of CTSS using Western blotting promotes the practical
utilization of extracellular CTSS as a potential biomarker for bovine
paratuberculosis, including the development of a simple lateral flow
assay. Nevertheless, there could be potential limitations in the use of
CTSS as a biomarker regarding interindividual variability in
MAP-infected dairy cows observed with Western blotting ([186]Figure 5a)
and the potential association of extracellular CTSS with other
diseases. Therefore, it is essential to examine the expression of the
protein in relation to the stage of the disease and its association
with other mycobacterial- and non-mycobacterial-induced diseases in
further investigations. So far, it is very rare that a biomarker cannot
also be associated with other diseases [[187]80].
Serum is an easily obtainable biological sample and reveals a high
amount of information on the pathophysiological conditions in diseased
animals [[188]38]. However, the analysis of the serum proteome is
challenged and limited due to the wide dynamic range of the proteins in
it [[189]38]. Compared to mass spectrometry, the proximity extension
assay (PEA) as a novel technology has revealed a more sensitively
targeted immunoassay [[190]81]. However, due to the restricted offering
of target products, PEA is not yet applicable for the analysis of
bovine samples. Nevertheless, we will actively monitor advancements in
technology and explore other emerging methodologies for their potential
application in the analysis of bovine samples. As new technologies
develop and become more accessible, we aim to reassess their
suitability for the identification of better disease biomarkers.
Our quantified mass spectrometry analysis identified a total of 394
proteins in bovine serum, of which 70 were significantly differently
abundant in the serum of MAP-infected dairy cows compared to healthy
controls. A further study only revealed a significantly divergent
abundance of eight proteins in the plasma of MAP-infected cattle
[[191]82]. In this study, differentially expressed proteins in a
2-dimensional fluorescence difference gel electrophoresis were analyzed
with mass spectrometry [[192]82]. This methodic procedure differed
strongly from our enhanced label-free analyses and had, therefore, led
to a limited number of identified proteins. Another serum proteomics
study already detected 669 quantified proteins in bovine serum, of
which only nine proteins showed significantly different abundances in
MAP-infected cattle compared to the control group [[193]83]. One of
these nine proteins, fetuin B, also showed a significantly different
abundance in our study. Due to the association of these already
detected biomarkers with other diseases
[[194]84,[195]85,[196]86,[197]87,[198]88], there is still a need for
suitable biomarkers of bovine paratuberculosis.
In clinical medicine, no single laboratory parameter should be used for
the diagnosis of a disease without considering clinical parameters.
However, a specific biomarker can contribute to an improvement in
diagnostic tests. To ensure this, a cohort study with all already
detected potential biomarkers for bovine paratuberculosis in serum or
plasma needs to be conducted. It would also be interesting to research
their correlation with other parameters, including disease stage and
antibody levels. In addition to CTSS, our serum proteomics analysis
revealed 69 other potential biomarkers that could also be further
investigated, since it was shown that for some diseases they can be
described through biomarker panels, not only one biomarker [[199]89].
For the development of a suitable panel, our candidates identified in
this study are available in PRIDE [[200]90].
5. Conclusions
Our differential serum proteomics study reveals valuable new insights
into the molecular mechanisms underlying host–pathogen interactions in
MAP-infected cattle. Identifying CTSS as a potential biomarker provides
a promising avenue for developing more effective diagnostic tools for
bovine paratuberculosis, thereby contributing to efforts to combat
bovine paratuberculosis.
Acknowledgments