Identification of the bovine α 1-acid glycoprotein in colostrum and milk

735Vet. Res. 36 (2005) 735–746© INRA, EDP Sciences, 2005DOI: 10.1051/vetres:2005029

Original article

Identification of the bovine α1-acid glycoprotein in colostrum and milk

Fabrizio CECILIANI*, Vanessa POCACQUA, Elena PROVASI, Claudio COMUNIAN, Alessandra BERTOLINI, Valerio BRONZO,

Paolo MORONI, Paola SARTORELLI

Dipartimento di Patologia Animale, Igiene e Sanità Pubblica Veterinaria, University of Milan, via Celoria 10, 20133 Milano, Italy

(Received 15 October 2004; accepted 21 February 2005)

Abstract – α1-acid glycoprotein (AGP) is an immunomodulatory protein expressed by hepatocytesin response to the systemic reaction that follows tissue damage caused by inflammation, infectionor trauma. This paper presents the detection of bovine AGP (boAGP) in mammary secretions(colostrum and milk) and mammary gland tissue. Bovine AGP was detected by Western blotting inall the samples analysed, and could be quantified in colostrum at 162 (± 63.7) µg/mL and 114.5(± 67.8) µg/mL during the first 12 h and 24 h respectively. In mature milk, the boAGP concentrationclearly decreased and was no longer detectable using the Radial Immunodiffusion (RID) technique.The concentration of mature milk boAGP was therefore semi-quantified using an anion-exchangechromatographic procedure that allowed the concentration of the protein to be determined. Thepresence of AGP in bovine milk was confirmed by the internal sequence analysis performedfollowing purification to homogeneity of the protein from milk. The concentration of AGP inbovine milk with low SCC (< 250 000) was very similar to that from bovine milk with high SCC(> 250 000). In order to investigate the origin of AGP in bovine milk, a search for mRNA wascarried out in somatic cells and mammary gland tissue: mRNA expression of the boAGP gene wasdetected in mammary gland tissue, but not in somatic cells. Finally, the cDNA sequence of theboAGP was determined, and is hereby presented.

α1-acid glycoprotein / mastitis / colostrum / milk

1. INTRODUCTION

Several bovine milk proteins, includingcasein components, whey proteins (α-lac-talbumin for example) and other peptides,have an immunomodulatory activity [9]. Itis likely that the biological function of thesepeptides is not only to regulate the immunefunction of the newborn, but also to modu-late, i.e. enhance or inhibit, the inflamma-

tory reactions during infections. In fact themechanisms that usually protect the udderfrom infection can also cause extensive tis-sue injury and these collateral damages thataccompany mastitis are often responsiblefor severe lesions to the mammary epithe-lium [1].

The aim of this paper was to investigatethe expression of α1-acid glycoprotein (AGP)in bovine colostrum and mature milk.

* Corresponding author: [email protected]

Article published by EDP Sciences and available at http://www.edpsciences.org/vetres or http://dx.doi.org/10.1051/vetres:2005029

http://www.edpsciences.org/vetres

http://dx.doi.org/10.1051/vetres:2005029

736 F. Ceciliani et al.

AGP belongs to the “lipocalins” family,a group of proteins that are deputed to thebinding and transport of small hydrophobicmolecules [6]: AGP has been further clas-sified in a subset of lipocalins, the so called“immunocalins”, a subfamily of proteinsthat may also immunomodulate the inflam-matory reaction [19].

Therefore, AGP features at least two dif-ferent biological activities, apparently verydifferent from each other: AGP may immu-nomodulate the inflammatory response, and,meanwhile, act as a plasma transport pro-tein [8, 13].

Bovine AGP (boAGP) is a glycoproteinof 42 kDa, as determined by SDS-PAGE,the carbohydrate moiety accounting for26.6% of its molecular weight [30]. Morerecently the MW of boAGP was fixed to33.8 kDa by MALDI-TOF mass spectrom-etry analysis [23].

From a clinical perspective, AGP belongsto the Acute Phase Proteins (APP), a struc-turally un-related group of mainly liver-derived plasma proteins that are associatedwith an acute phase reaction [10, 26]. AGPis considered as a minor acute phase proteinin cattle since its concentration increasesonly two- to four-fold during inflammation[30]. As the other acute phase proteins,AGP is produced mainly by hepatic cells,but local expression has been reported inhuman breast epithelial cells [11], stimu-lated alveolar macrophages [7] and humanendothelial cells [29]. The detection of twoof the major bovine APP, haptoglobin andserum amyloid A (SAA) in milk duringmastitis [5] may suggest a role for APP alsoin the immunomodulation of the local inflam-matory reaction in the udder. Among theseveral immunomodulatory activities thathave been proposed for AGP [13], some ofthem including the control of neutrophilactivity [34] and the expression of severalcytokines during inflammation [31] mightplay an important role in the context ofintramammary defences against microor-ganisms [28]. Bovine AGP may thereforerepresent a conceivable candidate to be one

of several host factors that regulate therecruitment from blood of immune cells,and their rate of activation.

This paper presents the detection of AGPin bovine colostrum and mature milk. Inorder to investigate the origin of boAGP inmilk, i.e. whether the protein is produced bymilk somatic cells, the mammary gland orderived from plasma extravasation, a searchfor mRNA has been carried out in bothsomatic and mammary gland tissue. As apreliminary step towards elucidating thecorrelation between the structure and func-tion of boAGP, the cDNA sequence of theboAGP gene was also determined.

2. MATERIALS AND METHODS

2.1. Animals

Colostrum samples were collected after12 h, 24 h, 3 days and 5 days post-partumfrom 10 clinically healthy dairy cows (Hol-stein Friesian).

Mature milk samples were collectedfrom 18 clinically healthy Holstein Friesiandairy cows, divided on the basis of thesomatic cell counts (SCC) into two groups:

Group A: 11 animals with a SCC <250 000 cells/mL.

Group B: 7 animals with a SCC >251 000 cells/mL.

Group B animals were considered as hav-ing mastitis. Three mammary gland sampleswere used for detection of boAGP mRNAfrom mammary tissue, and were collected atthe slaughterhouse from healthy animals,with a normal somatic cell count (< 250 000)and no evidence of clinical mastitis or otherdiseases. Mammary gland samples werestored in RNAlater (Qiagen S.p.A., Milano,Italy). Three liver samples, which wereused for the determination of the cDNAsequence of AGP and as a positive controlthroughout PCR experiments, were col-lected from healthy animals, immediatelyfrozen using liquid nitrogen and succes-sively stored at –80 °C.

AGP in bovine milk 737

2.2. Bacteriological procedures

Milk samples (10 µL) were spread onblood agar plates (5% defibrinated bovineblood). The plates were incubated aerobi-cally at 37 °C and examined at 24 h postseeding. The colonies were provisionallyspeciated based on morphology, haemoly-sis pattern and Gram stain. The numbers ofeach distinct colony type were recorded.The representative colonies were subcul-tured on blood agar plate and incubated aer-obically at 37 °C to obtain pure colonies.Gram positive cocci were tested for catalaseand coagulase production. Mastitis status ofmilk samples was determined by diagnosticprocedures recommended by the NationalMastitis Council.

2.3. Determination of somatic cell counts (SCC)

The SCC were determined for each milksample by an automated fluorescent micro-scopic somatic cell counter (Bentley Soma-count 150, Bentley Instruments, USA). Ethid-ium bromide dye was used for specificbinding to the DNA in the cell nuclei.

2.4. Detection of boAGP in milk and colostrum

The concentration of boAGP was deter-mined using a radial immunodiffusion assaycommercial kit (Bovine α1 AG Plate, TrideltaDevelopment Ltd., Maynooth, Ireland). Twoexperiments were performed:

(a) Determination of boAGP from whey:milk and colostrum whey were prepared asfollows: 10 mL of mammary secretions werecentrifuged at 1 200 × g for 15 min at 4 °Cin order to remove fat and cells. Four mil-lilitres of whey were further centrifuged at13 000 × g at RT, and 5 µL of the superna-tant was used following kit instruction.

(b) Determination of boAGP from wheyafter Q-Sepharose chromatography: wheywas prepared as previously described. Fourmillilitres of whey after 13 000 × g centrif-ugation were dialyzed overnight against the

starting buffer used for the chromatography(10 mM citrate-phosphate buffer, pH 4.0)at 4 °C. The dialyzed solution was then cen-trifuged at 13 000 × g for 5 min in order toremove insoluble proteins, and the superna-tant was finally applied onto HPLC (Jasco,Great Dunmow, UK) equipped with anHiTrap Q Sepharose XL (7/25 mm) stronganionic exchange column (Amersham Bio-sciences, Nerviano, Italy), that exploits thelow pI of AGP (3.2 to 3.7), and equilibratedwith the starting buffer. The column waswashed with the same buffer and the proteinwas eluted with 100 mM citrate-phosphatebuffer, pH 4.0. The elution was monitoredat 280 nm. The fraction containing boAGP(approximately 1 mL) was dried with Cen-tricon 30 (Millipore, Vimercate, Italy), resus-pended in 50 µL and 5 µL of this concen-trated protein solution was used followingthe bovine AGP detection kit instructions.

2.5. Electrophoresis and Western blotting

Fractions containing boAGP (15 µg totalproteins) were subjected to SDS-PAGE (12%polyacrylamide gels) in a discontinuous pHsystem [17] using a Miniprotean II appara-tus (Bio-Rad, Segrate, Italy). The proteinconcentrations were determined using aBradford Assay (Bio-Rad, Segrate, Italy).Western blotting was performed with amini trans blot electrophoresis cell (Bio-Rad, Segrate, Italy) onto nitrocellulose. BovineAGP was detected with an anti-feline AGPpolyclonal antibody raised in sheep whichwas generously provided by Dr Addie(Department of Veterinary Pathology, Uni-versity of Glasgow, United Kingdom) andan alkaline phosphatase-conjugated goatanti-sheep secondary antibody. The cross-reactivity of anti-feline AGP antibody withboAGP was tested using a commercialboAGP (Sigma, Milano, Italy) as a positivecontrol. The blots were developed using theAmplified AP Immun-Blot Kit (Bio-Rad,Segrate, Italy).


2.6. Purification and aminoacid sequence analysis of the purified boAGP from milk

The purification of AGP from bovinemilk was carried out starting from 100 mLof whey. Q-Sepharose chromatography wasperformed as previously described, and theeluted fractions (50 mL) were concentratedwith Centricon 30 (Millipore, Vimercate,Italy) to 5 mL and loaded onto a PorosHeparin Affinity (20 µm 4.6/100 mm) chro-matographic column (Applied Biosystems,Monza, Italy) at a flow rate of 5 mL/min.The column was previously equilibrated inTrisHCl 10 mM, pH 7.6. In these conditionsboAGP is not retained by the column andis therefore eluted in a void volume. Thefraction containing AGP (35 mL) was directlyloaded on a Sephasil Protein C4 (5 µm ST4.6/100 mm) chromatographic column(Amersham Biosciences, Nerviano, Italy)equilibrated with 0.065% TFA in 2% ace-tonitrile. Protein separation was accomplishedusing a 0–65% gradient of acetonitrile +0.05% TFA over 33 min, at a flow rate of1 mL/min. The fraction containing boAGP,which was eluted after 20 min, was finallyconcentrated to 1 mg/mL and submitted tosequence analysis. Sequence analysis wascarried out on an Applied Biosystems Pro-cise Sequencer. Purified boAGP (1 nmol)was subjected to automated Edman degra-dation. In order to obtain internal amino acidsequences, SDS-PAGE samples (2 nmol)were electrophoresed on 12% SDS- poly-acrylamide gel and blotted onto a polyvi-

nylidene difluoride (PVDF) membrane asdescribed [20]. The blots were stained withCoomassie Blue and excised bands weresubmitted to in situ CNBr fragmentation asdescribed in [2] directly on a TFA-treatedglass fiber disk used for sample loading.The sequence obtained was overlapped tothe cDNA sequence of boAGP.

2.7. Determination of boAGP gene sequence

Total RNA was extracted from the bovineliver using the RNeasy Mini Kit (Qiagen,Milano, Italy). The reverse transcription (RT)reaction was carried out using Ready To GoYour-Prime First-Strand Beads (Amer-sham Biosciences, Nerviano, Italy) and theprimers Random Hexamers and Oligo dT.The thermal profile was as follows: 50 minat 42 °C and 15 min at 70 °C. The cDNAwas used as the template for the PCR(Eppendorf Mastercycler). PCR reactionswere performed in 10 µL final volumesunder the following condition: 1X bufferEppendorf, 1.5 mM MgCl2, 0.2 mM of eachdNTP, 1 µM of each primer and 0.5 unit ofTaq Polymerase (Eppendorf, Milano, Italy).The primers used to amplify the codingsequence of bovine AGP are listed in Table Iand were obtained following the alignmentof the known AGP cDNA sequence avail-able in Genbank, i.e. man (M13692), mouse(M17376), rat (J00696), pig (M35990). Thethermal profile (35 cycles) was as follows:pre-PCR of 2 min at 94 °C; denaturation for45 s at 94 °C; annealing for 50 s at 56 °C

Table I. The degenerate primers used for cDNA sequencing of bovine AGP.

Primer Sequence Size (bp)

First pair of primers AGP1 F CCYATCACCAAYGMSACC 18AGP1 R GCRTAGAVRGASAGYCCC 18

Second pair of primers AGP1 F1 ACCAGTGYRTCTATAAC 17AGP1 R1 TTTATTGATGCAASTGAGGGA 21

Third pair of primers AGP F3 ATGGCACAAAGAACGTGG 18AGP R3b TTTATTGATGCAASTGAGGGA 21

Fourth pair of primers AGP1 F2 ATGGCGCTGYMCWSGG 16AGP1 R2 GTGGCATTGGTGATGGG 17


for the first, 45 °C for the second and 53 °Cfor the third pairs of primers; extension for80 s at 72 °C; final 10 min extension at72 °C. The thermal protocol applied for thefourth pair of primers was a Touch-downPCR: pre-PCR of 60 s at 94 °C; 8 cycles ofdenaturation for 30 s at 94 °C; annealing for45 s at 58 °C and extension for 80 s at 60 °Cfollowed by 32 cycles of denaturation for30 s at 94 °C; annealing for 45 s at 55 °Cand extension for 80 s at 60 °C; then a finalextension at 72 °C for 5 min.

PCR products were applied on a 2% aga-rose gel electrophoresis and the segments ofpredicted molecular weight obtained weregel-purified using the QIAquick gel extrac-tion kit (Qiagen, Milano, Italy) and then werequantified. The fragments were sequenceddirectly with ABI technology using an auto-mated DNA sequencer (ABI PRISM 310Genetic Analyzer). The predicted amino acidsequence was obtained using the ExPASyproteomic server.

2.8. Detection of AGP mRNA from cells and mammary tissue

Aliquots of the total RNA samples wereprepared from bovine liver (used through-out the experiment as positive controls forthe expression of boAGP), the somatic cells(all 18 milk samples) and three mammarygland samples using the RNeasy Mini Kit(Qiagen, Milano, Italy).

Somatic cells were purified as describedby [18]. cDNA synthesis was performed byusing Ready To Go Your-Prime First-StrandBeads (Amersham Biosciences, Nerviano,Italy) and oligo dT primers. Since mam-mary tissue stored in RNAlater solutionrevealed a partial degradation of RNA, twodifferent set of primers (Tab. I) were used.Primers AGPF and AGPR were used forsomatic and liver cells, and primers AGPF3and AGPR3b were used for mammary tis-sue and liver cells. One microlitre of cDNAwas amplified by PCR. The amplificationsof somatic cells and liver cDNA were per-formed using 35 cycles of 94 °C for 45 s

(denaturation), 56 °C for 40 s (annealing)and 72 °C for 80 s (extension). The condi-tions for the amplification of mammary tis-sue were the same, with the exception of theannealing step (53 °C for 40 s). The ampli-fied products were analysed on 2% agarosegels.

3. RESULTS

3.1. Quantification of boAGP in milk and colostrum

Bovine AGP was directly detected in allwhey samples from colostrum withdrawnafter 12 h (162 µg/mL ± 63.7 µg/mL) and24 h (114.5 µg/mL ± 67.8 µg/mL) (Fig. 1).Since the AGP in colostrum from day 3 andday 5, and from 18 whey samples frommature milk, was not directly detectable,the one-step procedure of purification andconcentration was set up. The boAGP-enriched fractions were finally semi-quan-tified using the Radial Immunodiffusiontest. The results are shown in Figure 1. Theconcentration of boAGP in colostrum

Figure 1. AGP concentration in bovine colos-trum and milk. Concentration of AGP in bovinecolostrum (12 h, 24 h, 3 days, 5 days) from10 cows (colostrum) and 12 cows (maturemilk). The data of mean concentration (µg/mL)are shown as means ± St.Dev. The asterisks (*)indicate that samples have been previously con-centrated (approximately 80 fold) by anionexchange chromatography and ultrafiltration, asdescribed in the Materials and methods section.


clearly decrease to 1.5 µg/mL ± 0.6 µg/mL(day 3) and 1.2 µg/mL ± 0.52 µg/mL(day 5).

The concentration in mature milk (SCC< 250 000) was also similar to colostrum(1.2 µg/mL, ± 0.69 µg/mL). The concentra-tion of AGP in bovine milk with low SCC(< 250 000) was very similar to that frombovine with high SCC (> 250 000), whichwas determined as 1.3 µg/mL ± 0.4 µg/mL.It should be pointed out that these valueswere probably underestimated since theexact amount of boAGP in milk could notbe determined, and therefore it was not pos-sible to evaluate the yield of boAGP afterthe chromatographic step. We reduced thevariability among the samples as much aspossible by using the same staring amount(4 mL) of whey.

The bovine AGP expression in colos-trum and milk was also analysed by West-ern blotting, using an anti-feline AGP pol-yclonal antibody that cross-reacts withboAGP (Fig. 2, lane AGP). All sampleswere analysed (10 colostrum and 18 milksamples). Figure 2 presents the Westernblot of the proteins eluted from anion-exchange fractions of some representative

milk (Fig. 2a) and colostrum samples (Fig. 2b).In order to determine the specificity of thepolyclonal antibody, 1.5 µg of commercialboAGP were loaded. The WB analysisrevealed that in the whole eluted fraction, aband of 45 kDa with the same MW as thecommercial boAGP was present. The datapresented indicate that (a) boAGP is presentin all colostrum and milk samples and (b)its MW is identical to that obtained fromplasma (commercial boAGP).

3.2. Purification and identification of bovine AGP in milk

Bovine AGP was further purified asdescribed in the Materials and methods sec-tion following three successive chromato-graphic steps, that allowed to obtainboAGP in the homogeneous form neces-sary for N-terminal amino acid sequencingand the results are reported in Figure 3. Theboxed insert of Figure 3 presents the SDS-PAGE of the three purification steps. Thepurification efficiency and recovery from100 mL of milk whey is reported in Table II.The purification table lacks the informationabout the recovery of boAGP from milkwhey after the first step, because the

Figure 2. Western blotting of AGP from anion exchange chromatography and ultrafiltration. LaneAGP shows the Western blotting of 1.5 µg of commercial boAGP. (a) Mature milk SDS-PAGE andWestern blotting from fractions obtained after Q-Sepharose anionic-exchange chromatography fromcows 32, 258, 189, 128, 140 and 269. 15 µg were loaded on each lane. (b) Colostrum Western blottingfrom fractions obtained after Q-Sepharose anionic-exchange chromatography from cows 505 and553 (1 = after 12 h, 2 = after 24 h, 3 = after 3 days, 4 = 5 days).


boAGP content of whey is under the sensi-tivity limit of the detection kit. Thereforethe yields were calculated starting withstep 2 (Q-Sepharose chromatography).

Aliquots of boAGP from fraction 3 weresubjected to (a) direct amino-terminalsequencing and (b) in situ fragmentationwith CNBr-formic acid followed bysequencing of the fragments.

Direct sequence analysis of boAGP gaveno results, indicating that the amino-termi-nal residue of the protein was blocked. Insitu fragmentation followed by the amino acidsequence resulted in a single the amino acid

sequence of 11 residues, LAASXVGTKNV.A comparison of this sequence with thetranslated cDNA sequence of boAGP (pre-sented below) revealed that the sequencedresidues are identical to the sequence ofboAGP starting from Leu 112. It can there-fore be concluded that the purified proteinfrom milk is bovine AGP.

3.3. Determination of boAGP cDNA sequence

The cDNA sequence of boAGP wasdetermined with the walking primer method.

Figure 3. Purification of AGP from bovine milk. (1) Q-Sepharose Chromatography of bovine whey.(3) Heparin Affinity Chromatography: the void volume chromatogram is presented. (2) SephasilProtein C4. Boxed insert: SDS-PAGE of boAGP-containing protein fractions of each purificationstep Stained with Coomassie Blue. Molecular mass standards are shown in the first lane. Lane 1:fraction 1 from Q-Sepharose Chromatography. Lane 2: fraction 2 from Heparin Chromatography.Lane 3: fraction 3 from C4-Sephasil Chromatography.


The nucleotide sequence obtained with thefirst pair of primers was the starting pointfor the design of the second pair of primers.In order to determine the sequence of thefirst 48 bases of boAGP mature protein cod-ing sequence, the third pair of primers wasused. The strategy of the sequencing ofboAGP, and the position of the degeneratedprimers, are summarised in Figure 4a, withthe exception of primer F2, which was posi-tioned from nucleotide 9 of the human AGPsequence, and is not shown in the Figure.

The nucleotide sequence of the fragmentobtained from each PCR was subjected toa BLAST search in order to match AGP’ssequence to that of other species. The foursegments of boAGP were overlapped togetherin order to form a 658-bp cDNA sequence,shown in Figure 4a, that codes for an aminoacid sequence of 219 residues. As deducedby sequence comparison with other pro-teins belonging to the AGP family, themature protein starts at the first amino acid(Q) and ends with the first Stop codon andaccounts for a MW of 20410 kDa. ThecDNA sequence of bovine AGP has beendeposited in the EMBL database under theaccession number AJ844606.

Figure 4b shows an alignment of boAGPwith other known sequences. Eighty-fourresidues out of 185 are conserved, or con-servatively substituted in all the proteins(45% similarity).

Prosite analyses are shown in Figure 4b,and revealed five potential N-glycosylationsites at residues 17, 40, 77, 87 and 119.Interestingly, bovine α1-acid glycoproteinshows seven potential phosphorylationsites: Ser24, Ser42 and Ser95 are Proteinkinase C phosphorylation sites while Thr69,

Ser91 and Ser95 are casein kinase II phos-phorylation sites. Finally, Lys179 is acAMP- and cGMP-dependent protein kinasephosphorylation site. Surprisingly, the socalled “Lipocalin signature”, i.e. the motifthat characterises the protein that transportssmall hydrophobic molecules, is absent inthe boAGP. However, Figure 4b shows thatin other AGP sequences, in the mouse andrat for example, this lipocalin signature isalso absent.

3.4. Expression of AGP mRNA from Somatic cells and mammary tissue

In order to investigate the origin of AGPin bovine milk, the expression of boAGPmRNA from mammary gland tissue andsomatic cells was examined by RT-PCR(Fig. 5). RNAlater stored samples (mam-mary glands) revealed partial degradationof RNA, in particular in the 5’ portion of thesequence. Two pairs of primers were there-fore used: the first pair (AGP1F andAGP1R) for somatic cells, and the third pairof primers (AGPF3 and AGPR3b) formammary gland samples (Tab. I). cDNAfrom the liver was used as a positive controlin both PCR. The transcripts of the expectedsize (according to the boAGP cDNAsequence) were detected in tissues from allthree mammary glands examined, but not insomatic cells. The very low expression ofAGP in sample number 15 can be attributedto the partial degradation of mRNA fromthe mammary gland tissue. This degrada-tion was particularly evident in samplenumber 15. The 350-bp PCR product (somaticcells and positive control) and 305-bp (mam-mary gland and positive control) were

Table II. Purification efficiency and recovery of AGP from 100 mL of bovine whey.

Purification step Total volume (mL)

Total protein (mg)

Protein (mg/mL)

BoAGP(µg/mL)

Total boAGP(µg)

Specific yield (µg/mL)

Q-Sepharose 50 mL 24.75 0.495 1.25 62.5 2.5He-Affinity 35 mL 1.35 0.04 1.21 42.2 31.4Sephasil C4 15 mL 36.66 2.44 1.88 28.2 769


Figure 4. The primary structure of bovine AGP. (a) cDNA and amino acid sequence of mature bovineAGP from the liver. (b) Homology comparison of the primary structure of boAGP with that of otherproteins belonging to the same family. The primary accession numbers (Swiss-Prot) are the follow-ing: pig: Q29014, human: P02763, rabbit: P25227, mouse: Q60590, rat: P02764. indicates the-oretical N-glycosylation sites. In bold are indicate the potential phosphorylated positions as predictedby Prosite (Expasy). Identical residues are indicated with *. Conservative substitutions are indicatedwith :. The box indicates the lipocalin signature.

N


sequenced and compared with the boAGPsequence (Fig. 4a) in order to confirm thecorrect identities of the PCR products.

4. DISCUSSION

The “somatic cells” found in milk arecomposed mainly of cells of the immunesystem, milk macrophages (66–80% in themilk of healthy cows, or of chronic mastitisaffected cows) and neutrophils (> 90% inacute mastitis) [24]. These cells form one ofthe most important defence mechanisms inthe udder. The ability of somatic cells tosequester and/or kill the pathogens is a crit-ical step in the development of mastitis, andis strictly regulated by several host factors,including cytokines, soluble receptors andreceptor antagonists that can increase, orinhibit, the immune defence in the mam-mary gland [1, 25]. In the present study animmunomodulatory plasma protein, theα1-acid glycoprotein, was identified incolostrum and milk. Purification of the pro-tein to homogeneity and internal sequenceanalysis further confirmed the identity ofboAGP. The MW of 20.4 kDa, determinedby the amino acid sequence derived fromthe cDNA sequencing of the boAGP gene,

corresponds to the 20 kDa of the deglyco-sylated polypeptide chain reported in theliterature [16].

Bovine AGP was detectable by RID andWB in all tested samples. The concentrationwas very high during the first two daysof lactation (colostrum) but significantlydecreased in mature milk.

In order to investigate the origin of AGPin bovine milk, AGP gene expression insomatic cells and mammary tissues wasexamined. Our results revealed that AGPmRNA levels remained undetectable insomatic cells. This was not surprising, sincethe level of boAGP semi-quantified after achromatographic concentration proceduredid not increase in samples with high SCC.On the contrary, AGP gene expression wasdetected in all three mammary tissues usedin this experiment. Nonetheless, these expres-sion studies have to be repeated using bovineepithelial mammary gland culture, such asMAC-T for example, that have been reportedto produce other APP (Serum Amyloid Aprotein) [21]. Moreover, a Real Time PCRapproach, instead of the simple RT-PCR,may be useful in quantifying the amount ofboAGP expressed by mammalian cells indifferent pathological conditions.

Figure 5. Expression of bovine AGP mRNA in mammary tissue and somatic cells. Two differentDNA concentrations were used in mammary gland tissue, 1:1 and 1:10 dilution. Where not expresslyindicated, a 1:1 dilution was used.


What can the biological significance ofthe presence of AGP in milk be? We canonly speculate about that. As an immu-nomodulatory molecule, boAGP could playseveral roles during inflammatory chal-lenges in the mammary gland. For example,AGP has been shown to inhibit severalactivities of neutrophils [4, 34] that cancause severe cell and tissue damages in themammary gland [1, 27]. Moreover humanAGP significantly inhibits the proliferationof peripheral blood T-cells [3], induces theexpression of several pro- and anti-inflam-matory cytokines by monocytes [31] andcan directly antagonise the capillary leak-age induced by some inflammatory media-tors, such as histamine, platelet activatingfactor and bradykinin [22].

AGP might exert these activities not nec-essarily by increasing its expression rate,but also by modifying its glycan moiety.Acute phase reaction may lead to otheralterations than concentration changes inheavily glycosylated proteins such as AGP,for example changes in glycan compositionas shown for AGP in other species [32, 33].To the best of our knowledge, only onepaper is available reporting the differencesin glycosylation patterns of boAGP frombovines of different ages [14] and thisaspect should be investigated further. Thispaper also showed that boAGP can be foundat high concentrations in colostrum in non-pathological conditions. Several compo-nents within colostrum can influence theimmunological development of the off-spring. The presence of several pro-inflam-matory (IL-1β, TNF-α) and anti-inflamma-tory (IL-1Ra, TNF-α receptor I and II, andIL-10) molecules has been widely demon-strated in bovine colostrum [12]. These fac-tors usually decrease in mature milk, andare likely to be involved in the developmentof the immature immune system of neonates.Interestingly, a very high level of serumAGP was found in calves (four times theadult value) [14] and piglets (40 times theadult value) [15]. Therefore, due to its immu-nomodulatory properties, it is conceivablethat boAGP may also contribute to the com-

plex framework of the immunoregulatorymolecules expressed in bovine colostrum.

ACKNOWLEDGMENTS

We thank Prof. Saverio Paltrinieri and DrLaura Kramer for the critical reading of thepaper. We acknowledge the very precious gift ofthe anti-feline AGP from Dr Addie, Departmentof Veterinary Pathology, University of Glasgow,United Kingdom. This work was financed byGrant FIRST/2003 (F. Ceciliani), FIRB/2001(P. Sartorelli) and COFIN 2003 (P. Sartorelli).

REFERENCES

[1] Burvenich C., Van Merris V., Mehrzad J.,Diez-Fraile A., Duchateau L., Severity ofE.coli mastitis is mainly determined by cowfactors, Vet. Res. 34 (2003) 521–564.

[2] Ceciliani F., Faotto L., Negri A., Colombo I.,Berra B., Bartorelli A., Ronchi S., The pri-mary structure of UK114 tumor antigen,FEBS Lett. 393 (1996) 147–150.

[3] Chiu K.M., Mortesen R.F., Osmand A.P.,Gewurz H., Interactions of α1-acid glycopro-tein with the immune system. I. Purificationand effects upon lymphocyte responsiveness,Immunology 32 (1977) 997–1005.

[4] Costello M.J., Gewurz H., Siegel J.N., Inhibi-tion of neutrophil activation by alpha1-acidglycoprotein, Clin. Exp. Immunol. 55 (1984)465–472.

[5] Eckersall P.D., Young F.J., McComb C.,Hogarth C.J., Safi S., Weber A., McDonald T.,Nolan A.M., Fitzpatrick J.L., Acute phaseproteins in serum and milk from dairy cowswith clinical mastitis, Vet. Rec. 148 (2001)35–41.

[6] Flower D.R., North A.C., Sansom C.E., Thelipocalin protein family: structural andsequence overview, Biochim. Biophys. Acta1482 (2000) 9–24.

[7] Fournier T., Bouach N., Delafosse C.,Crestani B., Aubier M., Inducible expressionof the α1-Acid glycoprotein gene by alveolarmacrophages: prostaglandin E2 and cyclinAMP act as new positive stimuli, J. Immunol.163 (1999) 1883–2890.

[8] Fournier T., Medjoubi-N N., Porquet D.,Alpha-1-acid glycoprotein, Biochim. Bio-phys. Acta 1482 (2000) 157–171.

[9] Gill H.S., Doull F., Rutherford K.J., CrossM.L., Immunoregulatory peptides in bovinemilk, Br. J. Nutr. 84 Suppl.1 (2000) 111–117.


[10] Gabay C., Kushner I., Acute-phase proteinsand other systemic responses to inflammation,N. Engl. J. Med. 340 (1999) 448–454.

[11] Gendler S.J., Dermer G.B., Silverman L.M.,Tokes Z.A., Synthesis of alpha 1-antichymo-trypsin and alpha 1-acid glycoprotein byhuman breast epithelial cells, Cancer Res. 42(1982) 4567–1573.

[12] Hagiwara K., Kataoka S., Yamanaka H.,Kirisawa R., Iwai H., Detection of cytokinesin bovine colostrum, Vet. Immunol. Immun-opathol. 76 (2000) 183–190.

[13] Hochepied T., Berger F.G., Baumann H.,Libert C., α1-Acid glycoprotein: an acutephase protein with inflammatory and immu-nomodulating properties, Cytokine GrowthFactor Rev. 14 (2003) 25–34.

[14] Itoh H., Tamura K., Izumi M., Motoi Y.,Funayama Y., Characterization of serumalpha 1-acid glycoprotein in fetal and new-born calves during development, Am. J. Vet.Res. 54 (1993) 591–595.

[15] Itoh H., Tamura K., Izumi M., Motoi Y.,Kidoguchi K., Funayama Y., The influence ofage and health status on the serum alpha 1-acidglycoprotein level of conventional and spe-cific pathogen-free pigs, Can. J. Vet. Res. 57(1993) 74–78.

[16] Iwata H., Murase K., Inoue T., Monoclonalantibody against bovine alpha1-acid glyco-protein, J. Vet. Med. Sci. 62 (2000) 1099–1100.

[17] Laemmli U.K., Cleavage of structural pro-teins during assembly of the head of bacteri-ophage T4, Nature 227 (1970) 680–685.

[18] Leutenegger C.M., Alluwaimi A.M., SmithW.L., Perani L., Cullor J.S., Quantitation ofbovine cytokine mRNA in milk cells ofhealthy cattle by real-time TaqMan polymer-ase chain reaction, Vet. Immunol. Immuno-pathol. 77 (2000) 275–287.

[19] Logdberg L., Wester L., Immunocalins: alipocalin subfamily that modulates immuneand inflammatory responses, Biochim. Bio-phys. Acta 1482 (2000) 284–297.

[20] Matsudaira P., Sequence from picomolequantities of proteins electroblotted onto pol-yvinylidene difluoride membranes, J. Biol.Chem. 262 (1987) 10035–10038.

[21] McDonald T.L., Larson M.A., Mack D.R.,Weber A., Elevated extrahepatic expressionand secretion of mammary-associated serumamyloid A2 (M-SAA3) into colostrum, Vet.Immunol. Immunopathol. 82 (2001) 203–211.

[22] Muchitsch E.M., Teschner W., Linnau Y.,Pichler L., In vivo effect of alpha 1-acid glyc-oprotein on experimentally enhanced capil-lary permeability in guinea-pig skin, Arch.Int. Pharmacodyn. Ther. 331 (1996) 313–321.

[23] Nakano M., Kakehi K., Tsai M. Lee Y.C.,Detailed structural feautures of glycan chainsderived from α1-acid glycoproteins of severaldifferent animals: the presence of hypersia-lylated, O-acetylated sialic acids but not disi-alyl residues, Glycobiol. 14 (2004) 431–441.

[24] Östensson K., Hageltorn M., Astrom G., Dif-ferential cell counting in fraction-collectedmilk from dairy cows, Acta Vet. Scand. 29(1988) 493–500.

[25] Paape M.J., Bannerman D.D., Zhao X., LeeJ.W., The bovine neutrophil: structure andfunction in blood and milk, Vet. Res. 34(2003) 597–627.

[26] Petersen H.H., Nielsen J.P., Heegaard P.M.,Application of acute phase protein measure-ments in veterinary clinical chemistry, Vet.Res. 35 (2004) 163–187.

[27] Shuster D.E., Kehrli M.E., Rainard P., PaapeM., Complement fragment C5a and inflam-matory cytokines in neutrophil recruitmentduring intramammary infection with Escheri-chia coli, Infect. Immun. 65 (1997) 3286–3292.

[28] Sordillo L.M., Shafer-Weaver K., DeRosa D.,Immunobiology of the mammary gland, J.Dairy Sci. 80 (1997) 1851–1865.

[29] Sörensson J., Matejka G.L., Ohlson M.,Haraldsson B., Human endothelial cells pro-duce orosomucoid, an important componentof the capillary barrier, Am. J. Physiol. 276(1999) 530–534.

[30] Tamura K., Yatsu T., Itoh H., Motoi Y., Iso-lation, characterization, and quantitativemeasurement of serum α1-acid glycoproteinin cattle, Nippon Juigaku Zasshi. 51 (1989)987–994.

[31] Tilg H., Vannier E., Vachino G., Dinarello C.,Mier J.J.W., Antiinflammatory properties ofhepatic acute phase proteins: preferentialinduction of interleukin 1 (IL-1) receptorantagonist over IL-1 beta synthesis by humanperipheral blood mononuclear cells, J. Exp.Med. 178 (1993) 1629–2636.

[32] Van Dijk W., Havenaar E.C., Brinkman-vander Linden E.C., Alpha 1-acid glycoprotein(orosomucoid): pathophysiological changesin glycosylation in relation to its function,Glycoconj. J. 12 (1995) 227–233.

[33] Van Dijk W., Brinkman-Van der LindenE.C.M., Havenaar E.C., Glycosylation ofalpha(1)-acid glycoprotein (orosomucoid) inhealth and disease: Occurrence, regulationand possible functional implications, TrendsGlycosci. Glycotechnol. 10 (1998) 235–245.

[34] Vasson M.P., Roch-Arveiller M., Couderc R.,Baguet J.C., Raichvarg D., Effects of α1-acidglycoprotein on human polymorphonuclearneutrophils: influence of glycan microhetero-geneity, Clin. Chim. Acta 224 (1994) 65–71.

Identification of the bovine α 1-acid glycoprotein in colostrum and milk

Documents