**4. Mastitis and inflammatory-related biomarkers in goats**

Mastitis is inflammation of the mammary gland that manifests in a wide range of physical indications, chemical changes in the milk, and pathological changes in the udder [56]. Clinical symptoms of mastitis include swelling and pain in the udder, increased rectal temperature, reduced feed intake and milk production, and the watery appearance or presence of clots in the milk. Mastitis research has drawn immense interest over the years because of its profound economic impact on the dairy industry. The incidence of clinical mastitis can be reduced by the application of management strategies, including a greater awareness of efficient milking and hygienic measures. However, despite the development of vaccines and other preventative methods, mastitis caused by Gram negative environmental pathogens and coliform bacteria remain problematic to treat and to manage [57]. Coliform bacteria are normal inhabitants of soil, manure, bedding, and water; thus, coliform mastitis can occur due to the contact of teats with the infected environment. Although innate immune responses in the mammary gland are effective to some extent, the mammary gland defense mechanisms can be compromised by environmental and physiological conditions [58]. Despite extensive knowledge of the bovine host response to mastitis pathogens and the effects of mastitis infection on the bovine milk proteome, only a limited understanding of the goat innate immune response to mastitis pathogens or the subsequent changes in goat milk protein expression over the course of a clinical infection exists. Nonetheless, like other ruminants that are managed for milk production, goats are also susceptible to and affected by mastitis.

In this study, they selected α-casein and two synthesized mono- and di-phosphopeptides as a

with the improvements of IMAC for phosphoproteomic experiments, instrument enhancements including improved acquisition speed allowed the identification of many more phosphopeptides per analysis. As mentioned before, in a comprehensive analysis of goat milk

Using nLC-MS/MS and high resolution mass spectrometer, they characterized the phosphorylation of several key mammary gland proteins in goat MFGM. This group, leveraging the strengths of high resolution and faster acquisition time, reported the detection of 271 sites of

The identification of PTMs is especially useful for the detection and characterization of acute phase proteins (APPs) during disease because APPs are subjected to modification. The N-glycan profiles of goat milk lactoferrin were compared with human and bovine milk using advanced mass spectrometry techniques [47]. The characterization of glycan composition established high mannose, hybrid, and complex N-glycans. Among the N-glycan compositions, 37% were sialylated and 34% were fucosylated. This group highlighted the existence of similar glycan composition between human and goat milk and discovered a novel glycan in goat milk that was not detected in human milk. A recent 2016 study investigated N-glycoproteome analysis of MFGMs from a number of mammals' milk [46]. They observed different glycosylation patterns in certain proteins that were previously reported with varying molecular weights based on the analysis by SDS-PAGE. They concluded that these discrepancies were the result of the differences in carbohydrate content

Mastitis is inflammation of the mammary gland that manifests in a wide range of physical indications, chemical changes in the milk, and pathological changes in the udder [56]. Clinical symptoms of mastitis include swelling and pain in the udder, increased rectal temperature, reduced feed intake and milk production, and the watery appearance or presence of clots in the milk. Mastitis research has drawn immense interest over the years because of its profound economic impact on the dairy industry. The incidence of clinical mastitis can be reduced by the application of management strategies, including a greater awareness of efficient milking and hygienic measures. However, despite the development of vaccines and other preventative methods, mastitis caused by Gram negative environmental pathogens and coliform bacteria remain problematic to treat and to manage [57]. Coliform bacteria are normal inhabitants of soil, manure, bedding, and water; thus, coliform mastitis can occur due to the contact of teats with the infected environment. Although innate immune responses in the mammary gland are effective to some extent, the mammary gland defense mechanisms can be compromised by environmental and physiological conditions [58]. Despite extensive knowledge of the bovine host response to mastitis pathogens and the effects of mastitis infection on the

was highly selective for multi-phosphopeptides

for enrichment of the MFGM samples.

had a higher affinity for mono-phosphopeptides. Along

O4

model system to demonstrate that NiZnFe<sup>2</sup>

, and ZnFe<sup>2</sup>

phosphorylation on 124 unique goat MFGM proteins [48].

MFGM phosphoproteome, Henry et al. used TiO2

O4

**4. Mastitis and inflammatory-related biomarkers in goats**

whereas Fe3

152 Goat Science

of these proteins.

O4 , NiFe2 O4

> Soluble mediators of inflammation in bovine milk and plasma during clinical mastitis have been studied extensively using antibody-based strategies [59]. Although antibody-based methodologies are both quantitative and accurate, they have limited detection capabilities. Conversely, mass spectrometric-based proteomic technologies allow for the simultaneous analysis of a larger number of proteins without the reliance on antibodies. Using MS-based proteomics, a number of biomarkers including APPs were identified in bovine serum and milk, which were correlated with pain and disease status [19, 60]. The concentration of most APPs typically increased during infection or inflammation, and the increased levels were relatively stable and persisted for a number of days, or even weeks, after the original insult or stimulus [19]. Despite the fact that our knowledge of the modulation of the bovine milk proteome during mastitis continues to expand, very little comparative data exists on lactating dairy goats.

> In regards to the study of the goat milk proteome, our group detected increases in haptoglobin (Hp), serum amyloid A (SAA), and lactoferrin in the milk of goats following an intramammary infusion of lipopolysaccharide (LPS) to induce coliform mastitis [28]. Other studies also documented significantly increased blood levels of Hp and SAA in an experimentally induced subacute ruminal acidosis in goats [61]. The majority of APPs are known to be glycosylated. Due to the high extent of its carbohydrate moiety, the APP alpha-1-acid glycoprotein (AGP) has been established as a biomarker of inflammation in goats [62]. AGP was also reported to potentially inhibit neutrophil migration to the site of infection, leading to inadequate bacterial clearance and resulting in increased risk of mortality [63]. Further, Heller et al. determined the species-specific reference intervals for four APPs including Hp, SAA, AGP, and lipopolysaccharide-binding protein (LBP), which is a soluble polypeptide that binds to bacterial LPS and increases its proinflammatory activity up to 1000 fold, in goat milk [64, 65].

#### **4.1. Effects of experimentally-induced mastitis on the goat proteome**

Modulations in the expression of goat milk proteins have been examined following an experimental induction of endotoxin mastitis by intra-mammary infusion with LPS. For details of challenge study and sample preparation, see materials and methods section in Ref. [28]. Crude milk samples were separated by 2DE prior to nLC-MS/MS analysis. The unique proteins identified following the 2DE analysis of skim milk from healthy goats and skim milk collected from the same goats 18 h post infusion with LPS are summarized in **Table 2**. In the absence of goat specific database, we used the Swiss-Prot other mammalia taxonomy, which includes only a limited number of goat sequences. Though some goat specific proteins were identified, the majority of the protein identifications were assigned to other species. As shown in this table, caseins constitute the most abundant proteins in milk; thus, a marked number of casein variants, specifically β- and αS2-caseins, which were detected in 13 and 6 separate


spots at varying isoelectric points on the gel, respectively, were observed in the milk of the goats prior to LPS infusion. The presence of full lengths β- and αS2-casein and in several spots of varying mass and isoelectric charges was most likely due to the presence of multiple fragments of the two dominant caseins in the skim milk as a result of proteolysis. Conversely, the αS1- and κ-caseins were detected in only two spots on the gel. Similar to the protein expression profiles generated from milk samples collected prior to infection, the β- and αS2-caseins dominated the profiles of the skim milk samples collected 18 h following LPS infusion. The number of β-casein fragments at lower-MW detected on the gel was reduced, but the number of αS2-casein spots, both full length protein and corresponding fragments increased in the 18 h samples. In addition to the caseins and the whey proteins β-lactoglobulin and α-lactalbumin, the lower abundance proteins serum albumin, gelsolin, retinol binding protein, fatty acid binding protein, and β-2-microglobulin were likewise detected in goat skim milk samples

**Table 2.** Proteins detected in milk of healthy goats and experimentally induced with endotoxin mastitis (LPS).

**Protein ID Protein name Species1 Peptides2 Sample3** P19661 Cathelicidin-3 Bovine 2 LPS

Proteomic Analysis of Goat Milk

155

http://dx.doi.org/10.5772/intechopen.70082

P22226 Cathelicidin-1 Bovine 3 LPS

P42819 Serum amyloid A Ovine 4 LPS P02584 Profilin-1 Bovine 2 LPS

Species of highest scoring assignment from Swiss-Prot (http://www.uniprot.org/).

Samples were either obtained from healthy goats or induced by LPS.

In sharp contrast to prior reports of bovine milk protein profiles during coliform mastitis, vascular-derived proteins such as complement factors and serotransferrin (known to leak into bovine milk following cytokine induction and the subsequent breakdown of the blood-milk barrier) were not detected in our analysis. Similarly, although serum albumin was detected, no significant increase in the abundance of the vascular-derived protein was apparent. However, along with the increases in SCC, several proteins with antimicrobial properties that are known to be found in the granules of neutrophils, the primary component of SCC during mammary infections, were detected in the goat skim milk samples at 18 h after induction of endotoxin mastitis including cathelicidin-1 and cathelicidin-3 and lactoferrin. The induction of the acute phase response during endotoxin mastitis in goats was also apparent as the APP haptoglobin (Hp) and SAA were likewise detected in the mastitic goat skim milk samples col-

The intensity of the spot corresponding to serum albumin remained the same in the gels of the goat milk samples 18 h post challenge, which could indicate that the breakdown of the blood-milk barrier during endotoxin mastitis might not be as profound in goats as has been observed in dairy cattle. The inflammatory response was however, supported by elevated

collected just prior to challenge with LPS.

1

2

3

Number of peptide assignments.

lected 18 h after intra-mammary infusion with LPS.


1 Species of highest scoring assignment from Swiss-Prot (http://www.uniprot.org/).

2 Number of peptide assignments.

**Protein ID Protein name Species1 Peptides2 Sample3** Q28372 Gelsolin Equine 2 Healthy

Q3SX14 Gelsolin Bovine 6 LPS

P14639 Serum albumin Ovine 9 LPS

P18626 αS1-Casein Caprine 15 LPS

P11839 β-Casein Ovine 4 LPS

P33049 αS2-Casein Caprine 5 LPS

P02670 κ-Casein Caprine 6 LPS

P02756 β-Lactoglobulin Caprine 10 LPS

P02694 Retinol-binding

154 Goat Science

P02694 Retinol-binding

Q4TZH2 Fatty acid-binding

protein-1

protein-1

protein

P85295 Serum albumin Caprine 6 Healthy

P18626 αS1-Casein Caprine 12 Healthy

P11839 β-Casein Ovine 3 Healthy

P04654 αS2-Casein Ovine 4 Healthy

P02670 κ-Casein Caprine 6 Healthy

P02756 β-Lactoglobulin Caprine 22 Healthy

P00712 α-Lactalbumin Caprine 3 Healthy

Q6QAT4 β-2-microglobulin Ovine 3 healthy

Q6QAT4 β-2-microglobulin Bovine 4 LPS

Q29477 Lactotransferrin Caprine 19 LPS

Q32PJ2 Apolipoprotein A-IV Bovine 9 LPS

B6E141 Haptoglobin Ibex 3 LPS

P00711 α-Lactalbumin Bovine 2 LPS

Bovine 4 Healthy

Bovine 5 Healthy

Bovine 2 LPS

3 Samples were either obtained from healthy goats or induced by LPS.

**Table 2.** Proteins detected in milk of healthy goats and experimentally induced with endotoxin mastitis (LPS).

spots at varying isoelectric points on the gel, respectively, were observed in the milk of the goats prior to LPS infusion. The presence of full lengths β- and αS2-casein and in several spots of varying mass and isoelectric charges was most likely due to the presence of multiple fragments of the two dominant caseins in the skim milk as a result of proteolysis. Conversely, the αS1- and κ-caseins were detected in only two spots on the gel. Similar to the protein expression profiles generated from milk samples collected prior to infection, the β- and αS2-caseins dominated the profiles of the skim milk samples collected 18 h following LPS infusion. The number of β-casein fragments at lower-MW detected on the gel was reduced, but the number of αS2-casein spots, both full length protein and corresponding fragments increased in the 18 h samples. In addition to the caseins and the whey proteins β-lactoglobulin and α-lactalbumin, the lower abundance proteins serum albumin, gelsolin, retinol binding protein, fatty acid binding protein, and β-2-microglobulin were likewise detected in goat skim milk samples collected just prior to challenge with LPS.

In sharp contrast to prior reports of bovine milk protein profiles during coliform mastitis, vascular-derived proteins such as complement factors and serotransferrin (known to leak into bovine milk following cytokine induction and the subsequent breakdown of the blood-milk barrier) were not detected in our analysis. Similarly, although serum albumin was detected, no significant increase in the abundance of the vascular-derived protein was apparent. However, along with the increases in SCC, several proteins with antimicrobial properties that are known to be found in the granules of neutrophils, the primary component of SCC during mammary infections, were detected in the goat skim milk samples at 18 h after induction of endotoxin mastitis including cathelicidin-1 and cathelicidin-3 and lactoferrin. The induction of the acute phase response during endotoxin mastitis in goats was also apparent as the APP haptoglobin (Hp) and SAA were likewise detected in the mastitic goat skim milk samples collected 18 h after intra-mammary infusion with LPS.

The intensity of the spot corresponding to serum albumin remained the same in the gels of the goat milk samples 18 h post challenge, which could indicate that the breakdown of the blood-milk barrier during endotoxin mastitis might not be as profound in goats as has been observed in dairy cattle. The inflammatory response was however, supported by elevated SCC in the goat milk following inoculation with endotoxin, as well as by the presence of both antimicrobial and APPs. The results likewise provided preliminary information regarding protein modulations of goat milk during disease as well as added knowledge of the host response during endotoxin mastitis in goats [28].

#### **4.2. Analysis of multiphosphorylation sites in caseins**

Caseins exist in four different variants including αS1-, αS2, β-, and κ-casein. Despite little homology, there is a rare conserved sequence (SSSEE) present in αS1-, αS2, and β-casein that serves as a multiphosphorylation site [66]. This conserved sequence motif does not exist in κ-casein, although it does possess two phosphorylation sites embedded in the C-terminus of the protein. The existence of the conserved sequence domain in casein multi-phosphopeptides makes their detection more challenging as the two glutamic acid (E) residues further increase their hydrophilic nature. In the presence of other peptides, they often go undetected by mass spectrometry. Thus, a vast majority of researchers use IMAC to enrich multi-phosphopeptides prior to detection. Without using any enrichment strategies, we reported changes in the levels of caprine casein phosphorylation [29].

In our experiments, we used milk samples obtained from healthy goats before and after experimental induction of endotoxin mastitis with LPS. We isolated casein bands using 1D-SDS/PAGE prior to in-gel digestion and analysis by nLC-MS/MS. Despite their large size, the majority of these tryptic phosphopeptides eluted early during chromatographic separation. As well, many were not detected following database searching or were assigned a very low peptide score. Consequently, manual inspections of the MS/MS spectra were necessary to validate the identifications.

In addition to mono- and di-phosphorylation sites, a number of multiphosphorylation sites exist in αS1-, αS2-, and β-casein (**Figure 1**). As shown, the conserved SSSEE domain forms a hexa-phosphopeptide in αS1-casein, a tetra-phosphopeptide in β-casein, and two different multi-phosphopeptides in αS2-casein. We characterized 18 different phosphorylation sites from a series of mono- and multi-phosphopeptides.

Examples of multi-phosphopeptides detected in our analysis are presented in **Figure 2**. The top MS/MS spectrum corresponds to a di-phosphopeptide (D58−K73) in αS1-casein that was detected as a doubly charged ion (MH2+= 951.32) corresponding to (MH+ = 1901.64, **Figure 2A**). In the αS2-casein, a triply charged ion (MH3+ = 1039.89) with the monoisotopic mass (MH<sup>+</sup> = 3117.08) was isolated and fragmented in a linear ion trap to produce the MS/MS spectrum as shown in **Figure 2B**. The bottom MS/MS spectrum corresponds to the tetra-phosphopeptide (E17−K43) that was detected in β-casein as a triply charged ion (MH3+ = 1103.10) corresponding to (MH+ = 3307.29, **Figure 2C**). The phosphorylation sites were accurately assigned in each multi-phosphopeptide despite the presence of multiple other serine and threonine resides. In αS1-casein, we did not detect the hexa-phosphopeptide, which also contains the conserved domain. Likely, the proximity of numerous phosphorylated sites made it even more fragile. Nevertheless, we clearly characterized one mono- and one di-phosphopeptide in αS1-casein [29].

Despite the lower apparent abundance, the multi-phosphopeptides shown in **Figure 2** were also detected in milk samples obtained from animals with experimentally-induced endotoxin mastitis. In αS2-casein, we also detected a tetra-phosphopeptide (N62−K86) with the amino acid sequence NANEEEYSIRSSSEESAEVAPEEIK [29]. The phosphorylation sites were found at S72, S73, S74, and S77. However, this multi-phosphopeptide was never observed in the mastitic goat milk samples. Instead, this ion was isolated and fragmented in the linear ion trap in which we readily detected two unmodified peptides. The first tryptic

**Figure 1.** Identification of potential serine and threonine phosphorylation sites within β-casein (A), αS1-casein (B), and αS2-casein (C). The conserved sequence domains (SSSEE) that serve as the multiphosphorylation sites are present in all

Proteomic Analysis of Goat Milk

157

http://dx.doi.org/10.5772/intechopen.70082

three caseins.

SCC in the goat milk following inoculation with endotoxin, as well as by the presence of both antimicrobial and APPs. The results likewise provided preliminary information regarding protein modulations of goat milk during disease as well as added knowledge of the host

Caseins exist in four different variants including αS1-, αS2, β-, and κ-casein. Despite little homology, there is a rare conserved sequence (SSSEE) present in αS1-, αS2, and β-casein that serves as a multiphosphorylation site [66]. This conserved sequence motif does not exist in κ-casein, although it does possess two phosphorylation sites embedded in the C-terminus of the protein. The existence of the conserved sequence domain in casein multi-phosphopeptides makes their detection more challenging as the two glutamic acid (E) residues further increase their hydrophilic nature. In the presence of other peptides, they often go undetected by mass spectrometry. Thus, a vast majority of researchers use IMAC to enrich multi-phosphopeptides prior to detection. Without using any enrichment strategies, we reported changes in the levels of caprine

In our experiments, we used milk samples obtained from healthy goats before and after experimental induction of endotoxin mastitis with LPS. We isolated casein bands using 1D-SDS/PAGE prior to in-gel digestion and analysis by nLC-MS/MS. Despite their large size, the majority of these tryptic phosphopeptides eluted early during chromatographic separation. As well, many were not detected following database searching or were assigned a very low peptide score. Consequently, manual inspections of the MS/MS spectra were necessary to

In addition to mono- and di-phosphorylation sites, a number of multiphosphorylation sites exist in αS1-, αS2-, and β-casein (**Figure 1**). As shown, the conserved SSSEE domain forms a hexa-phosphopeptide in αS1-casein, a tetra-phosphopeptide in β-casein, and two different multi-phosphopeptides in αS2-casein. We characterized 18 different phosphorylation sites

Examples of multi-phosphopeptides detected in our analysis are presented in **Figure 2**. The top MS/MS spectrum corresponds to a di-phosphopeptide (D58−K73) in αS1-casein that was

In the αS2-casein, a triply charged ion (MH3+ = 1039.89) with the monoisotopic mass (MH<sup>+</sup>

3117.08) was isolated and fragmented in a linear ion trap to produce the MS/MS spectrum as shown in **Figure 2B**. The bottom MS/MS spectrum corresponds to the tetra-phosphopeptide (E17−K43) that was detected in β-casein as a triply charged ion (MH3+ = 1103.10) correspond-

each multi-phosphopeptide despite the presence of multiple other serine and threonine resides. In αS1-casein, we did not detect the hexa-phosphopeptide, which also contains the conserved domain. Likely, the proximity of numerous phosphorylated sites made it even more fragile. Nevertheless, we clearly characterized one mono- and one di-phosphopeptide

= 3307.29, **Figure 2C**). The phosphorylation sites were accurately assigned in

= 1901.64, **Figure 2A**).

=

response during endotoxin mastitis in goats [28].

casein phosphorylation [29].

156 Goat Science

validate the identifications.

ing to (MH+

in αS1-casein [29].

from a series of mono- and multi-phosphopeptides.

detected as a doubly charged ion (MH2+= 951.32) corresponding to (MH+

**4.2. Analysis of multiphosphorylation sites in caseins**

**Figure 1.** Identification of potential serine and threonine phosphorylation sites within β-casein (A), αS1-casein (B), and αS2-casein (C). The conserved sequence domains (SSSEE) that serve as the multiphosphorylation sites are present in all three caseins.

Despite the lower apparent abundance, the multi-phosphopeptides shown in **Figure 2** were also detected in milk samples obtained from animals with experimentally-induced endotoxin mastitis. In αS2-casein, we also detected a tetra-phosphopeptide (N62−K86) with the amino acid sequence NANEEEYSIRSSSEESAEVAPEEIK [29]. The phosphorylation sites were found at S72, S73, S74, and S77. However, this multi-phosphopeptide was never observed in the mastitic goat milk samples. Instead, this ion was isolated and fragmented in the linear ion trap in which we readily detected two unmodified peptides. The first tryptic

peptide in the amino acid sequence (NANEEEYSIR) was detected as an intense doubly

phorylated segment (SSSEESAEVAPEEIK) was detected as a doubly charged ion (MH2+ =

the peptide remained dephosphorylated in mastitic goat milk. The attachment of multiple phosphate moieties might have sterically hindered the trypsin cleavage site and as a result, we observed the tetra-phosphopeptide as a missed cleavage. To this end, it has been reported that the proximity of the cleavage sites to the phosphorylated amino acids could impair

Advances in separation and mass spectrometry capabilities, enable our abilities to identify and characterize proteins and their PTMs. This chapter provides an overview of proteomics investigations in goat milk, from identifying major milk proteins to comprehensive analysis in different fractions and PTMs. Many challenges still exist, but technological advances have led to an increased in research contributing to a better understanding of the proteomic analysis of goat milk. Low-abundance proteins and disease-specific proteins have been identified as potential biomarkers. Using proteomics strategies, the efforts of our group and others have shed some light on the role of APPs during coliform mastitis and other diseases in goats. The host response during infection and related changes in the goat milk proteome remains comparatively limited. Nonetheless, our comparative proteomic analysis suggests that the caprine host response to endotoxin could differ from other ruminant species. Finally, our precise characterization of casein phosphorylation in goat milk before and after challenge with LPS

The views expressed in this article are those of the author do not necessarily reflect the official policy of the Department of Health Human Services, the U.S. Food Drug Administration, or

1 US Food and Drug Administration Center for Veterinary Medicine, Laurel, MD, USA

offers new insights into protein modulations in goat milk during mastitis.

\* and Jamie L. Boehmer2

\*Address all correspondence to: zohra.olumee-shabon@fda.hhs.gov

2 The University of Arizona, Tucson, AZ, USA

= 1227.36 Da). The second dephos-

http://dx.doi.org/10.5772/intechopen.70082

Proteomic Analysis of Goat Milk

159

= 1592.36 Da). These observations clearly indicated that

charged ion (MH2+ = 614.18 Da) corresponding to (MH+

796.68 Da) corresponding to (MH+

tryptic digestion [67].

**5. Conclusions**

**Disclaimer**

the U.S. Government.

Zohra Olumee-Shabon<sup>1</sup>

**Author details**

**Figure 2.** Tandem mass spectra of multi-phosphopeptides: (A) di-phosphopeptide in αS1-casein, (B) tetra-phosphopeptide αS2-casein, and (C) tetra-phosphopeptide β-casein. The *y-* and *b-*ions are marked in each spectrum with losses of phosphoric acid (neutral loss) probably owing to an in-source fragmentation.

peptide in the amino acid sequence (NANEEEYSIR) was detected as an intense doubly charged ion (MH2+ = 614.18 Da) corresponding to (MH+ = 1227.36 Da). The second dephosphorylated segment (SSSEESAEVAPEEIK) was detected as a doubly charged ion (MH2+ = 796.68 Da) corresponding to (MH+ = 1592.36 Da). These observations clearly indicated that the peptide remained dephosphorylated in mastitic goat milk. The attachment of multiple phosphate moieties might have sterically hindered the trypsin cleavage site and as a result, we observed the tetra-phosphopeptide as a missed cleavage. To this end, it has been reported that the proximity of the cleavage sites to the phosphorylated amino acids could impair tryptic digestion [67].
