Available online at www.sciencedirect.

com

Biochimie 91 (2009) 68e75

www.elsevier.com/locate/biochi

Research paper

Characterisation of single nucleotide polymorphisms identified in the

bovine lactoferrin gene sequences across a range of dairy cow breeds
F. O‘Halloran a, B. Bahar b, F. Buckley c, O. O’Sullivan a, T. Sweeney b, L. Giblin a,*
a
Teagasc, Moorepark Food Research Centre, Fermoy, Co. Cork, Ireland
b
Veterinary Science Centre, UCD, Belfield, Dublin 4, Ireland
c
Teagasc, Moorepark Dairy Production Centre, Fermoy, Co. Cork, Ireland
Received 25 March 2008; accepted 15 May 2008
Available online 23 May 2008

Abstract

The lactoferrin gene sequences of 70 unrelated dairy cows representing six different dairy breeds were investigated for single nucleotide
polymorphisms to establish a baseline of polymorphisms that exist within the Irish bovine population. Twenty-nine polymorphisms were iden-
tified within a 2.2 kb regulatory region. Nineteen novel polymorphisms were identified and some of these were found within transcription factor
binding sites, including GATA-1 and SPI transcription factor sites. Forty-seven polymorphisms were identified within exon sequences with
unique polymorphisms that were associated with amino acid substitutions. These included a T/A SNP, identified in a Holstein Friesian animal,
which resulted in a valine to aspartic acid substitution (Val89Asp) in the mature lactoferrin protein. Other SNPs of interest were associated with
amino acid substitutions in the lactoferricin B peptide sequence and an A/G SNP, identified in a Jersey animal, was associated with a tyrosine to
cysteine change (Tyr181Cys). The polymorphisms identified in the promoter region may have implications relating to lactoferrin expression
levels in cows and those identified in the coding sequence indicate the existence of protein variants in the Irish bovine population. The data
presented in this study emphasises the potential for lactoferrin to serve as a candidate gene to select for mastitis resistance with the aim of im-
proving animal health.
Ó 2008 Elsevier Masson SAS. All rights reserved.

Keywords: Lactoferrin; Single nucleotide polymorphisms; Bovine; Mastitis

1. Introduction
Lactoferrin is a bioactive glycoprotein found in the secre-
tions of a wide range of mammalian species, including bovine,
human, goat and porcine. The amino acid sequences of the lac-
toferrin proteins are highly conserved among the species of
higher mammals [1]. The level of lactoferrin in milk however,
varies enormously between species with the highest concentra-
tions recorded in human milk (1.0e5.0 mg/ml) [1]. However,
bovine milk, which is the most common source of commer-
cially produced lactoferrin contains about one-tenth of the lac-
toferrin found in other mammals with concentrations ranging
* Corresponding author: Tel.: þ353 25 42614; fax: þ353 25 42340.
E-mail address: linda.giblin@teagasc.ie (L. Giblin).
between 0.02 and 0.2 mg/ml [2]. This can however, dramati-
cally increase during the dry period and during a mastitis in-
cidence, with concentrations increasing by about 100-fold
[3]. This suggests that the protein plays important physiolog-
ical roles, particularly during a mastitis event.
The bovine lactoferrin gene is located on chromosome 22
spanning approximately 34.5 kb of genomic DNA and com-
prises 17 exons ranging from 82 (exon 1) to 225 bp (exon
17). The 50 -regulatory region of the bovine lactoferrin gene
contains the promoter sequence, a non-canonical TATA box
and a number of transcription factor binding sites necessary
for expression of the gene [4]. The coding sequence of the
functional bovine lactoferrin protein defines a single polypep-
tide chain of 708 amino acids, which is folded into two sepa-
rate lobes (N- and C-terminals) that each contains two iron-
binding and two bicarbonate-binding sites [5]. There are five

0300-9084/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.biochi.2008.05.011
 F. O‘Halloran et al. / Biochimie 91 (2009) 68e75
N-linked glycosylation sites present in bovine lactoferrin lo-
cated at asparagine (Asn) residues 233, 281, 368, 476 and
545. The significance of glycosylation for lactoferrin is not
yet completely understood [6].
Single nucleotide polymorphisms (SNPs) are the most
abundant genetic variations in mammalian genomes, occurring
in regulatory, exon and intron regions of genes [7]. To date, ge-
netic polymorphisms in the regulatory region of bovine and
human [5,7,8] lactoferrin genes have been described. How-
ever, the significance and associated functional effects of these
polymorphisms have yet to be established. Polymorphisms
have also been identified in the coding sequence of lactoferrin
genes. Mutations in the coding regions of a protein can affect
protein structure, which may potentially alter the biological
functions of that protein [9]. In addition, polymorphisms that
cause amino acid substitutions on the surface of the protein
that have no affect on protein structure could alter the surface
properties of the protein [10]. Alterations to the N-terminal se-
quence of the protein are particularly significant as this region
is largely responsible for the biological activity of lactoferrin.
To date, polymorphisms in the coding region of human and
goat lactoferrin that affect the antimicrobial properties of the
lactoferrin have been identified [11,12]. In addition Mohamed
et al. identified a unique polymorphism in human lactoferrin
genes that leads to increased susceptibility to diarrhoea [13].
Bovine mastitis is defined as inflammation of a cows’ udder
and is predominantly caused by bacterial infections. The finan-
cial loss to the dairy industry caused by mastitis is significant
[14]. A logical approach to reducing mastitis incidence in
dairy cattle would be to select for animals in breeding pro-
grammes that are genetically resistant or less susceptible to
the disease [15]. Lactoferrin plays a vital role in the natural de-
fence mechanism of the mammary gland and therefore, is a po-
tential candidate gene for imparting resistance to mastitis in
dairy cows [4,16]. This study was conducted to establish
a baseline of genetic polymorphisms present in the lactoferrin
genes of Irish dairy cattle and identify single genetic events
which could potentially affect milk lactoferrin concentration
or produce lactoferrin protein variants.
2. Materials and methods
2.1. Sample population and blood collection
Seventy cows representing six different breeds were used in
this study. The group comprised 47 Holstein Friesians, 6 Jer-
sey, 5 JerseyeFriesian crossbreeds, 7 Norwegian Reds, 4
Montbéliard and 1 Norwegian RedeFriesian crossbreed. To
provide a diverse genetic pool, the animals selected were
largely unrelated being from six different breeds and within
each breed they had different sires. Blood samples (7 ml)
were collected (in duplicate) by caudal venipuncture from
each animal into sample tubes containing the anticoagulant
potassium ethylenediaminetetraacetic acid (EDTA K3E 15%,
0.12 ml; BD Vacutainer, BD Vacutainer Systems, Plymouth,
UK) and immediately placed on ice for subsequent DNA
and RNA extractions.
69
2.2. Genomic DNA isolation
To extract genomic DNA, blood samples were initially
centrifuged at 2500  g for 20 min at 4  C and the buffy layer
was collected with the surrounding erythrocytes into a 1.5 ml
sterile Eppendorf tube. DNA was extracted using a QIAGEN
Flexigene DNA isolation kit (QIAGEN Ltd., West Sussex,
UK), according to the manufacturers’ instructions. The DNA
was quantified spectrophotometrically and the integrity as-
sessed via gel electrophoresis (1.2%). DNA was stored at
20  C for subsequent analysis.
2.3. RNA isolation and cDNA synthesis
Blood samples for RNA extraction were initially processed
as for DNA extraction above and the RNA was isolated from
the buffy layer of each sample using the QIAGEN RNeasy
midi kit (QIAGEN Ltd., West Sussex, UK). RNA was quanti-
tated spectrophotometrically and the integrity assessed by
electrophoresis in a 1.5% glyoxyl gel with 1 glyoxyl buffer
(Ambion, Applied Biosystems, Foster City, USA). All glass-
ware and gel apparatus used for RNA analysis was pre-treated
with RNAzap (Ambion, Applied Biosystems, Foster City,
USA) and DEPC-treated water. RNA was aliquoted and stored
at 80  C prior to cDNA synthesis. The latter was achieved by
converting 1 mg of RNA to cDNA using the QIAGEN Quanti-
Tect reverse transcription kit (QIAGEN Ltd., West Sussex,
UK).
2.4. PCR amplification of lactoferrin gene
To amplify a 2.2 kb promoter region of bovine lactoferrin
gene, three sets of oligonucleotide primers were designed,
based on published sequence data (GenBank Ref. NC_
007320), to amplify three overlapping PCR fragments (Table
1). To amplify the coding sequence of the genes, five sets of
primers were designed (Table 1) based on the published se-
quences of Seyfert et al. [17] (GenBank accession no.
L19985) and Mead and Tweedie [18] (GenBank accession
no. X54801). To generate PCR products from the regulatory
region, 100 ng of genomic DNA was used in the presence of
200 pM of both forward and reverse primers. Cycling condi-
tions were as follows: initial denaturation at 94  C for 1 min,
followed by 40 cycles of 94  C for 30 s, 56  C for 30 s,
72  C for 1 min, with a final extension at 72  C for 10 min.
To amplify products representing the coding sequence 3 ml
of cDNA was used in each reaction and PCR was carried
out in 50 ml volumes containing 10 pmol of each forward
and reverse primer, 5 ml 10  NH4 buffer (160 mM
(NH4)2SO4, 670 mM TriseHCl (pH 8.8), 0.1% Tween 20),
MgCl2 at a final concentration of 1.5 mM, 200 mM each
dNTP, and 2.5 U Taq polymerase (Bioline Ltd., London,
UK). The cycling conditions used were as follows: initial de-
naturation at 95  C for 5 min, followed by 35 cycles of 94  C
for 1 min, 55e66  C (see Table 1) for 30 s, 72  C for 1 min,
with a final extension at 72  C for 5 min. All PCR products
were analysed by agarose gel electrophoresis (1.5%) and
70 F. O‘Halloran et al. / Biochimie 91 (2009) 68e75

Table 1
Primer details for PCR to amplify the regulatory (R) and coding (C) sequence of the lactoferrin genes
Primer name Sequence 50 e30 Optimum annealing temperature ( C) Product size (bp)
LFC1-F TGAGGTCACCGAGCACTGGATAAA 66 555
LFC1-R CAGCTCAAGTACGGGCGAAGGATT
LFC2-F GTCTCCCCAAACCCACTATTATGC 62 570
LFC2-R TCCTGCGCCTTGCTGAGAG
LFC3-F GCCAGTGGATGCGTTCA 58 586
LFC3-R CCCTTCCGTTGGTCTCAG
LFC4-F TGAAAGGGGAAGCAGATG 56 553
LFC4-R AAGTCCTCACGATTCAAGTT
LFC5-F CTGGGAGAACACGAATGG 58 547
LFC5-R ATGAGCTTGGCAAAAATCCGACTT
LFR1-F GACAGCCTTTGGGCACTTAG 56 1068 (652 to þ416)
LFR1-R GGGTAGGACAGAAGCGACAG
LFR2-F CTTGGGCTGGCTCTTCTTC 56 947 (1391 to 443)
LFR2-R TAAGACAGACGGAGACCCTTG
LFR3-F TTATGCTGCCTTGCCTCTTC 56 949 (2193 to 1243)
LFR3-R AAAGCTGGCCTTGATTCTTTG

ethidium bromide staining and visualised using the AlphaIm- 3. Results

ager (Alpha Innotech, San Leandro, CA, USA). PCR products
were purified prior to sequencing using the QIAGEN PCR pu-
rification kit (QIAGEN Ltd., West Sussex, UK).
2.5. DNA sequence analysis
PCR products were sequenced (Cogenics and MWG-Bio-
tech) and the data analysed using the Lazergene 6 suite of pro-
grams (DNAStar Inc., Madison WI, USA) and Molecular
Evolutionary Genetic Analysis 4.0 (MEGA). Some of the
PCR products were consistently difficult to sequence and
therefore these products were cloned in order to sequence
each allele separately. For cloning, the TOPO Cloning kit (In-
vitrogen Ltd., Paisley, UK) was used according to the manu-
facturers’ instructions and a selection of clones were used
for DNA sequence analysis. SNP data was analyzed using
population genetic data analysis software, Arlequinn 3.11.
The allele and genotype frequencies obtained were expressed
as percentages. Evaluation of the association of individual
SNPs with breed was performed through estimating Pearson’s
correlation co-efficient and P < 0.05 was considered statisti-
cally significant. AliBaba 2.1 software was used to identify
SNPs that affected putative transcription factor binding sites.
2.6. Predictive 3D protein structural images
To predict if the polymorphisms identified in the lactoferrin
coding regions could potentially lead to structural alterations
of the lactoferrin protein, translated DNA sequences were an-
alysed using homology modelling software (Swiss-PDB
Viewer) [19]. Lactoferrin protein variants were modelled using
the 3D-bovine lactoferrin structure taken from the protein da-
tabank [20] and models were visualised and verified using py-
mol [21]. Structural superposition was performed between the
protein variants and wild-type protein using the Match-Align
subroutine of Chimera [22].
In total, 2.2 kb of the promoter region of the bovine lacto-
ferrin gene was sequenced in both forward and reverse direc-
tions to confirm the presence of polymorphisms. The
consensus sequence identified in this study deviated from the
published bovine lactoferrin sequence (GenBank Ref. NC_
007320) by 29 nucleotides, giving rise to an average of one
polymorphism per 76 bases. The SNP loci, percentage
homo- and heterozygosity of the SNPs and the major allele
frequency determined in this study are presented in Table 2.
There were 14 G/A, 9 T/C, 3 G/T, and 1 each of C/A, G/C
and G/e identified across the study population (n ¼ 70).
The only SNP loci that correlated with the breed type was
the SNP at 1560 (Pearson correlation co-efficient
r ¼ 0.60, P < 0.001). For this SNP, 100% Friesians
(n ¼ 47) and 100% Norwegian Reds (n ¼ 7) were associated
with the CC genotype. Nineteen unique polymorphisms were
identified in this study (Table 2, highlighted in bold type).
The A/C SNP at position 28 occurs immediately adjacent
to the TATA box, the transcription factor recognition site. At
this locus, the major allele was A with an allele frequency of
72% and C was the minor allele (frequency 28%), with
56.9%, 12.1% and 31% of animals identified as A/A, C/C
and C/A genotypes, respectively. Also of interest is the G/A
SNP located at position 1155, which was found to be lo-
cated on two putative transcription factor binding sites in-
cluding (GATA-1 and SP1), based on predictions using
AliBaba2 software. The majority of animals (98.6%) were
homozygous for the G allele (genotype G/G), with a single
animal identified as heterozygous for this SNP (genotype
G/A). Haplotype analysis was performed, using Haploview
version 4.0, to identify the major haplotypes in the study
population. Eight different haplotypes were identified (data
not shown) with two major haplotypes predominating. The
major haplotype occurred with a frequency of 50% of the
study population and the second major type occurred with
a 12% population frequency.
 F. O‘Halloran et al. / Biochimie 91 (2009) 68e75 71

Table 2 polymorphism that was associated with an Arg17Lys substitu-

Single nucleotide polymorphisms (SNPs) present in the promoter region of the tion. The consensus sequence (AGG) encodes the arginine res-
bovine lactoferrin gene (loci positions relative to the transcription start site),
the proportion of homozygous heterozygous genotypes and the major allele
idue and the variant sequence (AAG) encodes a lysine residue.
frequency in Bos taurus (n ¼ 70) cows This SNP is potentially significant as it occurs within the en-
Locus SNPs Homozygosity Heterozygosity Major allele
coding sequence of the antimicrobial peptide, Lactoferricin B
(major/minor (%) (%) frequency (LfcinB). Other SNPs identified that affected the LfcinB se-
allele) quence included the SNPs associated with Lys-10Arg, Cy-
2151 G/A 61.9 38.1 0.67 s11Arg and Lys19Arg substitutions. Another interesting SNP
2140 G/A 69.1 30.9 0.76 was the A/G SNP associated with a tyrosine to cysteine
2122 C/T 60.0 40.0 0.69 (Tyr181Cys) substitution. This SNP was identified in a single
2101 T/C 60.0 40.0 0.69 heterozygous Jersey animal (Fig. 2b) and is of interest as
2077 A/G 61.4 38.6 0.69
2047 A/G 60.0 40.0 0.69
a new Cys residue could potentially incorporate new disul-
2009 C/T 98.6 1.4 0.99 phide bridges into the lactoferrin protein structure.
1953 G/A 60.0 40.0 0.69 3D structural models of two lactoferrin protein variants,
1807 C/T 60.0 40.0 0.69 VT50349 and VT4514 were generated (Fig. 2). To achieve
1762 T/C 82.9 17.1 0.91 this, the variant protein sequences were superimposed onto
1753 G/T 80.0 20.0 0.90
1702 A/G 60.0 40.0 0.69
the consensus bovine lactoferrin protein sequence (shown in
1560 C/T 89.9 10.1 0.95 blue). Highlighted in red are the locations of the amino acid
1448 C/G 72.5 27.5 0.80 substitutions that were found in these variant proteins.
1335 G/A 81.4 18.6 0.91 Fig. 2a shows the predicted structure of the variant protein
1155 G/A 98.6 1.4 0.99 VT50349 with the location of two amino acid substitutions,
1099 A/G 70.0 30.0 0.79
929 G/A 95.1 4.9 0.96
Leu51Pro and Gly60Ser, highlighted in red. Both of these res-
918 T/G 95.1 4.9 0.96 idues are located within b-sheet secondary structures. Fig. 2b
602 G/T 73.8 26.2 0.82 shows the predicted structure of variant protein VT4514 with
600 G/A 96.8 3.2 0.97 its Lys10Arg substitution, which occurs within the a-helix
586 C/T 65.6 34.4 0.80 structure of the LfcinB peptide and its Tyr181Cys substitution.
479 G/e 100.0 0.0 0.88
287 G/A 98.3 1.7 0.99
The latter substitution occurs within a loop region that is not
270 T/C 71.7 28.3 0.83 usually functionally active itself but serves to link functionally
190 G/A 76.7 23.3 0.87 active domains. However, as the substitution results in the ad-
156 G/A 73.3 26.7 0.80 dition of a new cysteine residue this may introduce new disul-
131 T/C 69.4 30.6 0.78 phide bridges into the protein structure.
28 A/C 69.0 31.0 0.72

The data in Table 3 represents 47 polymorphisms in the lac-
toferrin coding sequences that deviated from the consensus se-
quence. The amino acid positions affected by the SNP,
together with a description of the wild-type and variant-type
amino acids (aa) are given. The frequency of each allele and
the percentage of homozygotes and heterozygotes are also pre-
sented. As can be seen 18 of these SNPs were nonsense muta-
tions, causing no change to the amino acid sequence, while 27
SNPs were associated with amino acid changes. The majority
of the SNPs identified were infrequent, with 80% of the vari-
ant-type alleles occurring with <2% frequency. High fre-
quency SNPs were located at aa positions 6, 118, 312, 344,
369, 412, 422 and 594, with only two of these SNPs (aa118
and aa412) associated with aa changes. As can be seen from
the data, the animals investigated in this study were largely
heterozygous for the polymorphisms identified with only 11
of the 47 SNPs (23%) identified as homozygote genotypes.
Study animals were heterozygous for the remaining SNPs.
Some PCR products were consistently difficult to sequence
and thus the products were cloned and representative clones
were sequenced. Fig. 1a shows the chromatograms generated
for two cloned PCR products. Sequence analysis confirmed
the presence of two different alleles that identified a G/A
4. Discussion
Twenty-nine polymorphisms were identified within the
2.2 kb regulatory region of the bovine lactoferrin gene. Among
these, 10 were previously reported [5,8] while others (n ¼ 19)
were unique polymorphisms, some of which were found to be
associated with putative transcription factor binding sites. For
example, the SNP at position 1155 is located on the GATA-1
and SPI transcription factor binding sites. SP1 DNA-binding
sites are known enhancer elements in eukaryotic genomes
that are considered to be constitutive activators of gene expres-
sion [23] and thus alterations to the DNA sequence of this site
could potentially affect lactoferrin expression levels. GATA-1
is a zinc finger transcription factor involved in the expression
of the lactoferrin gene during hematopoesis in humans [24].
Mutations in GATA-1 binding sites could potentially affect
the binding of the transcription factor to the lactoferrin pro-
moter, thus polymorphisms at this site may have important
physiological implications.
Haplotype analysis identified eight different haplotypes in
the study population with two haplotypes predominating at
high frequency (50% and 12%). Further studies will be re-
quired to evaluate the transcription efficiencies of these differ-
ent promoter haplotypes.
72 F. O‘Halloran et al. / Biochimie 91 (2009) 68e75

Table 3
Polymorphisms identified in the coding regions of bovine lactoferrin genes
aa SNP WT aa VT aa WT allele VT allele Homozygosity Homozygosity Heterozygotes
position frequency (%) frequency (%) WT (%) VT (%) (%)
1 T/C Cys Cys 99 1 98.6 0.0 1.4
2 C/T Thr Thr 98.6 1.4 98.6 1.4 0.0
4 C/T Ser Ser 99 1 98.6 0.0 1.4
6 T/C Pro Pro 10.7 89.3 84.3 5.7 10
10 A/G Lys Arg 99 1 98.6 0.0 1.4
11 T/C Cys Arg 99 1 98.6 0.0 1.4
13 A/G Arg Arg 99 1 98.6 0.0 1.4
17 G/A Arg Lys 99 1 98.6 0.0 1.4
19 A/G Lys Arg 99 1 98.6 0.0 1.4
25 T/A Ser Ser 99 1 98.6 0.0 1.4
27 A/G Thr Ala 99 1 98.6 0.0 1.4
29 T/C Val Ala 99 1 98.6 0.0 1.4
33 T/C Phe Phe 99 1 98.6 0.0 1.4
47 T/C Asp Asp 99 1 98.6 0.0 1.4
51 T/C Leu Pro 99 1 98.6 0.0 1.4
53 G/A Gly Asp 99 1 98.6 0.0 1.4
56 T/C Val Ala 99 1 98.6 0.0 1.4
59 C/T Ala Val 99 1 98.6 0.0 1.4
60 G/A Gly Ser 99 1 98.6 0.0 1.4
61 C/T Arg Trp 99 1 98.6 0.0 1.4
62 A/G Asp Gly 99 1 98.6 0.0 1.4
64 T/C Tyr His 99 1 98.6 0.0 1.4
69 T/C Val Ala 99 1 98.6 0.0 1.4
70 A/G Ala Ala 99 1 98.6 0.0 1.4
85 T/A Tyr Asn 99 1 98.6 0.0 1.4
87 G/A Val Met 99 1 98.6 0.0 1.4
89 T/A Val Asp 98.6 1.4 98.6 1.4 0.0
94 A/G Ser Gly 99 1 98.6 0.0 1.4
102 A/G Gln Gln 99 1 98.6 0.0 1.4
104 C/T Arg Trp 99 1 98.6 0.0 1.4
116 G/A Gly Arg 99 1 98.6 0.0 1.4
118 A/G Ile Val 84.3 15.7 77.1 8.6 14.3
120 C/A Pro Thr 99 1 98.6 0.0 1.4
150 T/C Val Ala 99 1 98.6 0.0 1.4
181 A/G Tyr Cys 99 1 98.6 0.0 1.4
209 C/T Ala Val 98 2 95.7 0.0 4.3
301 A/G Arg Gly 99 1 98.6 0.0 1.4
312 G/T Leu Leu 84.3 15.7 75.8 7.1 17.1
344 T/C Pro Pro 85.7 14.3 78.6 7.1 14.3
357 C/T Ser Ser 99 1 98.6 0.0 1.4
362 T/C Thr Thr 98.6 1.4 98.6 1.4 0.0
369 T/C Thr Thr 83 17 75.7 10.0 14.3
371 C/G Asp Glu 98.6 1.4 98.6 1.4 0.0
377 C/T Leu Leu 99 1 98.6 0.0 1.4
412 C/T His Tyr 81.4 18.6 70.0 7.1 22.9
422 G/A Thr Thr 82 18 70.0 5.7 24.3
594 T/C Asp Asp 94 6 87.2 0.0 12.8
Single nucleotide polymorphism (SNP) with amino acid (aa) changes shown in bold type; aa position is taken from the mature protein sequence. WT, wild-type;
VT, variant type.

The majority of polymorphisms detected in the lactoferrin
coding sequence occurred within the N-lobe of the lactoferrin
protein. From a structural point of view these polymorphisms
are significant if they are associated with amino acid substitu-
tions that can potentially affect the protein structure and/or
function. To date, several authors have described polymor-
phisms in the lactoferrin genes of different species. Teng
and Gladwell identified 21 SNPs in human lactoferrin se-
quences, with nine of these associated with amino acid substi-
tutions [7]. Kang et al. bioinformatically analysed 60 different
lactoferrin genes representing 11 different species [25]. In
their study they identified 50 novel polymorphisms across
five different species that were associated with amino acid
substitutions in lactoferrin proteins. Ten of these polymor-
phisms were found within the lactoferrin genes of Bos taurus
species. Li et al. [5] identified five SNPs within their bovine
study population with only one of these being associated
with an aa change. This polymorphism (A/G) was located in
exon 4 and resulted in an isoleucine (Ile) to valine (Val) sub-
stitution at position 118 in the mature protein. Velliyagounder
 F. O‘Halloran et al. / Biochimie 91 (2009) 68e75 73

Fig. 1. Sequence chromatographs of heterozygote SNPs identified in the coding lactoferrin gene. (a) DNA sequence of two different clones (A and B) which iden-
tified a heterozygote G/A SNP that leads to an arginine-lysine substitution at amino acid position 17 in the mature protein. (b) The A/G heterozygote SNP identified
in exon 5 that caused a tyrosine to cysteine substitution at position 181 in the mature protein. Shown are the forward and reverse sequence data that confirmed the
SNP.

et al. and Lee et al. demonstrated that individual SNPs in hu-
man and goat lactoferrin sequences (respectively) affected the
antimicrobial properties of the protein [11,12].
In the present study several SNPs (n ¼ 8) were found to be
associated with conservative amino acid substitutions, in that
they could be predicted to have no significant effect on protein
structure. These include valine to alanine, isoleucine to valine
and threonine to alanine substitutions, which are all aliphatic
aa with similar chemical properties. Whether such substitu-
tions affect the substrate recognition properties of the protein
however, remains to be determined.
Three lysine/arginine substitutions were located within the
LfcinB bioactive peptide. Two SNPs were identified within the
Lfcin B encoding sequence that were associated with lysine to
arginine substitutions (Lys10Arg, Lys19Arg) and another SNP
resulted in an arginine to lysine change (Arg17Lys). These
substitutions could be considered conservative changes as
both are positively charged, polar amino acids. However, Vel-
liyagounder et al. demonstrated that this type of substitution
can affect the antimicrobial property of the Lfcin B peptide
[11]. In their study they confirmed that a Lys29Arg polymor-
phism altered the bactericidal activity of the peptide. The Arg-
containing peptide was less active against the gram positive
species Streptococcus mutans and Streptococcus mitis. Some
of the major pathogens that cause bovine mastitis are gram
positive species, including streptococcal and staphylococcal
species [26]. The Cys11Arg substitution is also noteworthy
as the replacement of the negatively charged cysteine by the
positively charged arginine affects the overall cationic charge
of the LFcin B peptide. In vivo Lfcin B peptide is released via
pepsin digestion of the mature protein and plays an important
physiological role by helping to stimulate both innate and
adaptive immune response mechanisms. The peptide binds
to susceptible bacteria and destabilizes their membranes via
membrane perturbation and it can also help regulate an innate
immune response by alerting host cells to the potential for in-
fection [27]. Mutations in the Lfcin B peptide could poten-
tially alter either of these activities by modifying the
susceptibility of microbes and/or affect the capacity of the
peptide to bind and stimulate host cell immune response.
Polymorphism that results in the removal or addition of
cysteine residues are structurally significant as they can influ-
ence the formation of disulphide bridges, which serve to stabi-
lise protein structures. In this study an A/G SNP was identified
that resulted in a tyrosine (Tyr) to cysteine (Cys) change at po-
sition 181 in the mature protein. To examine the effect of the
Tyr181Cys substitution on protein structure a 3D model of the
protein, variant VT4514, was generated (Fig. 3). To allow
74 F. O‘Halloran et al. / Biochimie 91 (2009) 68e75
Fig. 2. Predicted 3D structure of two bovine lactoferrin protein variants (VT):
(a) lactoferrin protein variant VT50349 with two aa substitutions (Gly60Ser
and Leu51Pro); (b) protein variant VT4514 with the Lys10Arg substitution,
occurring within Lfcin B sequence, and the Tyr181Cys substitution.
a new disulphide bridge to form the distance between the al-
pha carbons on two Cys residues must be less than 8 Ang-
stroms (Å). Therefore three new bonds could potentially
form between Cys181 residue and (Cys149) 6.10 Å,
(Cys165) 6.14 Å and (Cys249) 7.37 Å.
Fig. 3. Cysteine bridges in VT4514. The Cys181 residue is shown with the
new disulfide bridges that can potentially form.
Other SNPs of interest included a T/A SNP, identified in
a Holstein Friesian animal, which resulted in a Val to Asp sub-
stitution (Val89Asp). This could be a potentially significant
change as it introduces a negatively charged amino acid into
the structure.
Another exciting amino acid change identified is the
Leu15Pro (Fig. 2a), which was identified in a single Holstein
Friesian animal. The leucine side chain is very non-reactive
and thus rarely directly involved in protein function, although
it can play a role in substrate recognition. Proline however, of-
ten plays important roles in molecular recognition, particularly
in intracellular signalling. It can also function to introduce
kinks into alpha helices, since it is unable to adopt a normal
helical conformation.
Four polymorphisms were identified within the lactoferrin
signal peptide sequence Lys2Arg, Ser10Pro, Leu18Pro and
Ala23Ala. Whether any or all of these polymorphisms affect
lactoferrin secretion in vivo will be determined by analysing
the lactoferrin content in mammary secretions.
5. Conclusions
Bovine mastitis is a costly disease to the dairy industry. The
documented association between lactoferrin concentrations in
milk and bovine mastitis incidence makes this gene an obvious
candidate to investigate whether variants impart differences in
disease resistance. This study has identified a baseline of poly-
morphisms in lactoferrin gene sequences that are present
within the Irish bovine population. Several potentially interest-
ing novel polymorphisms were identified within the coding re-
gion, including polymorphisms occurring within the sequence
of the antimicrobial peptide Lactoferricin B, as well as SNPs
that affect the formation of new cysteine bridges in the lacto-
ferrin structure. Novel SNPs were also found within the pro-
moter sequence which could influence the expression of
lactoferrin in vivo. To determine if the polymorphisms identi-
fied in this study are suitable as genetic markers for mastitis
resistance, further investigations will be necessary, not least
to determine if associations exist between the prevalence of
specific polymorphisms and superior udder health, i.e. reduced
incidence of clinical mastitis and/or low somatic cell counts.
In the event of evidence of association, screening of animals,
including AI sires and potential young test sires, could take
place, thus helping to expedite genetic improvement in udder
health in a background where other production and functional
traits are monitored. The overall objective would be to provide
the dairy industry with increased disease-resistant robust
animals.
References
[1] C.T. Teng, Lactoferrin gene expression and regulation: an overview, Bio-
chem. Cell Biol. 80 (2002) 7e16.
[2] F.L. Schanbacher, R.E. Goodman, R.S. Talhouk, Bovine mammary lacto-
ferrin: implications from messenger ribonucleic acid (mRNA) sequence
and regulation contrary to other milk proteins, J. Dairy Sci. 76 (1993)
3812e3831.
 F. O‘Halloran et al. / Biochimie 91 (2009) 68e75
[3] T. Kutila, S. Pyorala, H. Saloniemi, L. Kaartinen, Antibacterial effect of
bovine lactoferrin against udder pathogens, Acta Vet. Scand. 44 (2003)
35e42.
[4] J. Zheng, J.L. Ather, T.S. Sonstegard, D.E. Kerr, Characterisation of the
infection-responsive bovine lactoferrin promoter, Gene 353 (2005) 107e
117.
[5] G.-H. Li, Y. Zhang, D.-X. Sun, N. Li, Study on the polymorphism of bo-
vine lactoferrin gene and its relationship with mastitis, Anim. Biotechnol.
15 (2004) 67e76.
[6] H.A. van Veen, M.E.J. Geerts, P.H.C. van Berkel, J.H. Nuijens, The role
of N-linked glycosylation in the protection of human and bovine lactofer-
rin against tryptic proteolysis, Eur. J. Biochem. 271 (2004) 678e684.
[7] C.T. Teng, W. Gladwell, Single nucleotide polymorphisms (SNPs) in hu-
man lactoferrin gene, Biochem. Cell Biol. 84 (2006) 381e384.
[8] M. Daly, P. Ross, L. Giblin, Polymorphisms within the lactoferrin gene
promoter in various cattle breeds, Anim. Biotechnol. 17 (2006) 33e42.
[9] L.-H. Liu, W. Gladwell, C.T. Teng, Detection of exon polymorphisms in
the human lactoferrin gene, Biochem. Cell Biol. 80 (2002) 17e22.
[10] E.N. Baker, H.M. Baker, R.D. Kidd, Lactoferrin and transferrin: func-
tional variations on a common structural framework, Biochem. Cell
Biol. 80 (2002) 27e34.
[11] K. Velliyagounder, J.B. Kaplan, D. Furgang, D. Legarda, G. Diamond,
R.E. Parkin, D.H. Fine, One of two human lactoferrin variants exhibits
increased antibacterial and transcriptional activation activities and is as-
sociated with localized juvenile periodontitis, Infect. Immun. 71 (2003)
6141e6147.
[12] T.H. Lee, K. Shimazaki, S.L. Yu, M.S. Nam, S.J. Kim, K.K. Lee,
D.Y. Yu, Polymorphic sequence of Korean native goat lactoferrin exhib-
iting greater antibacterial activity, Anim. Genet. 28 (1997) 367e369.
[13] J.A. Mohamed, H.L. DuPont, Z.D. Jiang, J. Belkind-Gerson, J.F. Figueroa,
L.Y. Armitige, A. Tsai, P. Nair, F.J. Martinez-Sandoval, D.C. Guo,
P. Hayes, P.C. Okhuysen, A novel single-nucleotide polymorphism in
the lactoferrin gene is associated with susceptibility to diarrhea in North
American travellers to Mexico, Clin. Infect. Dis. 44 (2007) 945e952.
[14] K. Wojdak-Maksymiec, M. Kmiec, J. Ziemak, Association between bo-
vine lactoferrin gene polymorphism and somatic cell count, Vet. Med. 51
(2006) 14e20.
75
[15] S.M. Youngerman, A.M. Saxton, S.P. Oliver, G.M. Pighetti, Association
of CXCR2 polymorphisms with sub clinical and clinical mastitis in dairy
cattle, J. Dairy Sci. 87 (2004) 2442e2448.
[16] K. Kawai, S. Hagiwara, A. Anri, H. Nagahata, Lactoferrin concentration
in milk of bovine clinical mastitis, Vet. Res. Commun. 23 (1999) 391e
398.
[17] H.-M. Seyfert, A. Tuckoricz, H. Interthal, D. Koczan, G. Hobom, Struc-
ture of the bovine lactoferrin-encoding gene and its promoter, Gene 143
(1994) 265e269.
[18] P.E. Mead, J.W. Tweedie, cDNA and protein sequence of bovine lactofer-
rin, Nucleic Acids Res. 18 (1990) 7167.
[19] N. Guex, M.C. Peitsch, SWISS-MODEL and the Swiss-PdbViewer: an
environment for comparative protein modelling, Electrophoresis 18
(1997) 2714e2723.
[20] H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat,
H. Weissig, I.N. Shindyalov, P.E. Bourne, The Protein Data Bank, Nu-
cleic Acids Res. 28 (2000) 235e242.
[21] W.L. DeLano, The PyMOL Molecular Graphics System on World Wide
Web.http://www.pymol.org (2002).
[22] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt,
E.C. Meng, T.E. Ferrin, UCSF Chimera e a visualization system for ex-
ploratory research and analysis, J. Comput. Chem. 25 (2004) 1605e
1612.
[23] A. Khana-Gupta, T. Zibello, C. Simkevich, A.G. Rosmarin, N. Berliner,
Sp1 and C/EBP are necessary to activate the lactoferrin gene promoter
during myeloid differentiation, Blood 95 (2000) 3734e3741.
[24] J.J. Johnston, P. Rintels, J. Chung, J. Sather, E.J. Benz, N. Berliner, Lac-
toferrin gene promoter: structural integrity and nonexpression in HL60
cells, Blood 79 (1992) 2998e3006.
[25] J.-F. Kang, X.-L. Li, R.-Y. Zhou, L.-H. Li, F.-J. Feng, X.-L. Guo, Bioin-
formatic analysis of lactoferrin gene for several species, Biochem. Genet.
46 (2008) 312e322.
[26] N.-Y. Lee, K. Kawai, I. Nakamura, T. Tanaka, H. Kumura, K. Shimazaki,
Susceptibilities against bovine lactoferrin with micro-organisms isolated
from mastitic milk, J. Vet. Med. Sci. 66 (2004) 1267e1269.
[27] H. Wakabayashi, M. Takase, M. Tomita, Lactoferricin derived from milk
protein lactoferrin, Curr. Pharm. Design 9 (2003) 1277e1287.