Analysis of sla class ii polymorphism to study disease resistance in pigs

Development of simultaneous high resolution typing of three SLA class II genes, SLA-DQA, DQB1 and DRB1 using genomic DNA multiplex PCR and direct sequencing ..... Here, we report the suc

Dissertation for Degree of Doctor

Supervisor: Prof Chankyu Park

Analysis of SLA class II polymorphism

to study disease resistance in pigs

Submitted by

LE MINH THONG

February, 2015

Department of Animal Biotechnology

Graduate School of Konkuk University

Analysis of SLA class II polymorphism

to study disease resistance in pigs

A Dissertation submitted to the Department of Animal Biotechnology

and the Graduate School of Konkuk University

in partial fulfillment of the requirements for the degree of

Doctor of Philosophy in Animal Biotechnology

Submitted by

LE MINH THONG

October, 2014

This certifies that the Dissertation of

Le Minh Thong is approved.

Approved by Examination Committee:

i

TABLE OF CONTENTS

List of Tables iv

List of Figures v

List of Abbreviations vi

Abstract ix

Chapter 1 Literature Review 1

1.1 Pig and infectious disease 1

1.2 Biology of swine leukocyte antigens 5

1.2.1 Structure and function 5

1.2.2 Disease resistance and susceptibility 7

1.2.3 Association with productive traits 10

1.2.4 Genetic diversity and fitness 10

1.3 Genetics of SLA 11

1.3.1 Mapping 11

1.3.2 Polymorphism and linkage disequilibrium 13

1.4 Nomenclature of SLA genes 14

1.5 Comparison of typing methods for swine leukocyte antigen 15

1.5.1 Serology 16

1.5.2 Cell-based typing 16

1.5.3 Sequence-based Typing 16

Chapter 2 Development of simultaneous high resolution typing of three SLA class II genes, SLA-DQA, DQB1 and DRB1 using genomic DNA multiplex PCR and direct sequencing 19

ii

2.1 Introduction 19

2.2 Materials and methods 21

2.2.1 Animals 21

2.2.2 Multiplex PCR for the simultaneous amplification of SLA-DQA, DQB1 and DRB1 21

2.2.3 Direct sequencing and allelic determination of SLA-DQA, DQB1 and DRB1 22

2.2.4 Determination of SLA class II haplotypes 22

2.3 Results and discussion 25

Chapter 3 Association between the polymorphisms of SLA-DQB1and DRB1 and piglet survival 29

3.1 Introduction 30

3.2 Material and method 32

3.2.1 Animals 32

3.2.2 High resolution typing of SLA-DQB1 and SLA-DRB1 32

3.2.3 Statistic association test 33

3.2.3.1 Association of SLA-DQB1, DRB1 alleles and haplotypes 33

3.2.3.2 Association of exon 2 polymorphisms 33

3.2.3.3 Impact of substituted amino acids 34

3.2.4 Prediction of peptide and SLA binding 34

3.3 Results 35

3.3.1 Typing and association of DQB1, DRB1 35

3.3.2 Identify associated polymorphism 41

3.3.3 Effect of amino acids 43

iii

3.3.4 In silico binding analysis 50

3.4 Discussion 53

Chapter 4 Establishing immortalized cell panel with diverse SLA genotypes 60

4.1 Introduction 60

4.2 Materials and methods 62

4.2.1 Tissue and primary cells 62

4.2.2 SLA typing 62

4.2.3 Cell culture, transfection and selection 63

4.2.4 Cell growth characterization 63

4.2.5 Immune fluorescent analysis 64

4.3 Results 64

4.3.1 SLA genotypes 64

4.3.2 Characterization of immortalization 67

4.3.3 Correlation of morphology and growth rate 70

4.3.4 Expression of SLA class I on cell surface 71

4.4 Discussion 72

References 75

Appendix 99

Abstract (in Korean) 154

iv

List of Tables

Table 1.1 Recent emerging pathogens in pigs 3 Table 2.1 Primers used for multiplex PCR and direct sequencing of SLA-DQA, SLA-DQB1, and SLA-DRB1 24 Table 3.1 Distribution and statistic association of 16 SLA-DQB1 alleles to piglet survivability 36

Table 3.2 Distribution and statistic association of 18 SLA-DRB1 alleles to piglet

survivability 38

Table 3.3 Distribution and statistic association of 20 major SLA-DQB1:DRB1

haplotypes to piglet survivability 40 Table 3.4 Interaction between phenotype associated amino acids to piglet survivability 48 Table 4.1 Typing results of panel comprised 26 cell lines 66 Table 4.2 Summary of typing results of cell panel 67

v

List of Figures

Figure 1.1 Molecular structure of MHC class I and II 6

Figure 1.2 Schematic molecular organization of the SLA genes 7

Figure 1.3 Comparative genomic organization of the human and swine major histocompatibility complex (MHC) class I region 12

Figure 1.4 Comparative genomic organization of the human and swine major histocompatibility complex (MHC) class II region 12

Figure 1.5 Polymorphism of MHC genes in various species (adapted from the Immuno Polymorphism Database-MHC (IPD-MHC) 13

Figure 2.1 Successful amplification of SLA-DQA, SLA-DQB1, and SLA-DRB1 by using multiplex PCR 27

Figure 3.1 The association across the exon 2 of SLA-DQB1and DRB1 42

Figure 3.2 Plot of the association to the phenotypes of healthy and illness/early died piglet of amino acid substitutions across DQB1 exon 2 44

Figure 3.3 Plot of the association to putative disease of amino acid substitutions across DRB1 exon 2 47

Figure 3.4 Number and proportion of unique pathogenic peptides bind to each SLA-DRB1 alleles with predictably strong affinity 51

Figure 3.5 Boxplot of difference of predicted affinity to peptides with strong binding (pIC50 < 50nM) among 4 SLA-DRB1 alleles, 0101, 0201, 0603Q and 04kn05 53

Figure 4.1 Comparison of the cell morphology between primary cultures and permanent cell line by the transfection of hTERT and and SV40LT 68

Figure 4.2 A colony of cells which escaped from the crisis stage 69

Figure 4.3 Growth curves of selected cell lines 70

Figure 4.4 Expression of SLA class I on the cell surface of selected lines 71

vi

List of Abbreviations

ACP Annealing control primer

ASFV African swine fever virus

ATCC American Type Culture Collection

BLAST Basic Local Alignment Search Tool

BSA Bovine serum albumin

ELISPOT Enzyme-Linked ImmunoSpot assay

FAO Food and Agriculture Organization of the United Nations FBS Fetal bovine serum

FMD Foot-and-mouth disease

GGP Great grandparent

GSBT Genomic sequence-based typing

GWAS Genome-wide association study

HIV Human immunodeficiency virus

HLA Human leukocyte antigen

hTERT Human telomerase

IEBD immune Epitope Database

vii

IMGT ImMunoGeneTics information system

IPD Immuno Polymorphism Database

ISAG International Society for Animal Genetics KIR Killer cell immunoglobulin-like receptors KNP Korean native pig

MHC Major histocompatibility complex

mRNA Messenger ribonucleic acid

NIH National Institutes of Health

PBS Phosphate buffered saline

PCR Polymerase chain reaction

PCR-SSP PCR using sequence-specific primers

PCV2 Porcine Circovirus Type 2

PCVAD Porcine circovirus type 2-associated diseases PEF Porcine embryo fibroblast

PMWS Post weaning Multisystemic Wasting Syndrome PRRS Porcine Reproductive & Respiratory Syndrome RFLP Restriction fragment length polymorphism RT-PCR Reverse transcriptase-PCR

SLA Swine leukocyte antigen

SNP Single-nucleotide polymorphism

SNU Seoul National University

SV40 LT Simian Vacuolating Virus 40 large T antigen

viii TAE Tris-acetate-EDTA

TGEV Transmissible gastroenteritis virus

ix

Abstract

Analysis of SLA class II polymorphism to study

disease resistance in pig

Le Minh Thong Department of Animal Biotechnology Graduate School of Konkuk University

As one of the main components of the immune system, major histocompatibility complex (MHC) type shapes how the host reacts to endogenous

or exogenous risk factors in term of immune struggles Pig MHC or so called swine leukocyte antigen (SLA) genes also were reported with associations to many aspects

in pig production industry as well as in biomedical research as potential large animal model In parallel with MHC from other mammalian species, SLA genes also expose extremely polymorphism and complicated over entire pig genome For decades, many typing methods had being developed to examine the variation and biological meaning of MHC system in the species with the resolution and accuracy more improved Nevertheless, the methods revealed certain inherent technical limitations Here, we report the successful high-resolution typing of the three most polymorphic genes of SLA class II, DQB1, DQA and DRB1 using locus specific amplification from genomic DNA and direct sequencing We enhanced the robustness by simultaneous amplification of the three SLA class II genes and subsequent analysis

of individual loci using direct sequencing The new method has significant benefits over the individual locus typing, including lower typing cost, use of less biomaterial, and fewer errors in handling large samples for multiple loci In order to dissert and verify in detail the influence of SLA to piglet survivability which is assumptively attributed by infectious disease that wasting disease as potential risk, we carried out

x

a case-control study with 2 major SLA class II genes DQB1 and DRB1 as a first attempt to assess the effect of genetic differences to the occurrence of wasting disease like phenotypes after weaning We have found the variances contributing to the phenotype of disease-like resistance and susceptibility at the level of haplotype, assigned allele, SNP and amino acid Further in silico analysis of peptide binding affinity coincided and reinforced the above assessments For aiming research of interaction between SLA and specific disease antigen in vitro, we conducted transfection both transform factors SV40 LT and hTERT to the pig primary cells, make up the cell panel comprise 20 immortalized cell lines with diverse SLA genotype from major classical genes of SLA class I (SLA1, 2, 3) and class II (DQA, DQB1 and DRB1) Along with 8 commercial cell line (ATCC) that had been typed for SLA1, 2, 3, DQB1 and DRB1, we carried out DQA typing to raise the total number of cells in the panel to 28 cell lines Examination on the cell lines that most has overcome passage 40 shows the high uniform of phenotype, relatively high speed replication and continuous expression of SLA class I on the cell surface We hope this cell panel will be important contributions in research of interactions between antigens and SLA, the biological diversity of the SLA as well as improve the immune genetic studies in pigs

keyword: SLA, typing, disease, association, immortalization, cell

1

Chapter 1 Literature Review

1.1 Pig and infectious disease

Along with the process of human civilization, pig (Sus scrofa domesticus) is

one of the domesticated animals from an early era and has become very popular livestock, with important implications in the structure of global agriculture Except for some areas with cultural and religious issues, pork is preferred as most consumed red meat Pig been widely adopted and raised throughout the nations of the world with number of herbs is rapidly increasing According to the forecasts of FAO (Food and Agriculture Organization of the United Nations), in 2015, the total number of pigs will reach one billion heads (FAO 2014)

One of the main issues of the present pig production industry is preventing, reducing and remedy the damage of the affects by disease in pigs Especially are infectious diseases that often disrupts the production process, causing a lot of difficulties and the loss of millions dollars per year In 1998, the estimated cost of damages caused by the pig diseases in US exceeded 1.5 billion USD In 2005, only PRRS (Porcine Reproductive & Respiratory Syndrome) has led to the livestock

industry in the USA consume up to 561 million dollars (Neumann et al 2005)

In modern agriculture, remarkable resources have been invested to improve pig production in both nutritional value and profit Large part in which is to improve the

pig health as well as against the ravage of infectious diseases (Uddin Khan et al

2013) Remedial therapy, probiotics, and vaccines in particular have a huge effect in this work Comparing with non-transmissible diseases, infectious diseases are more complicated due to evolutionary variation in pathogens, environmental affect, and reciprocal interaction with the host Therefore, the effect of the above treatments is heterogeneous and sometimes very low when faced with the dangerous diseases, especially caused by viral pathogens One of those determined long ago that caused severe damage around the world is foot-and-mouth disease (FMD) Caused by FMD

2

virus with 7 serotypes in pig including A, O, C, SAT 1, SAT 2, SAT 3 and Asia 1 as well as many strains, the disease is highly infectious, spreads rapidly throughout animal populations and over long distances by the aerosol particles and hence it is costly difficult to control There is no treatment and infected animals should be destroyed (De Clercq & Goris 2004) Or PRRS virus causes Porcine Reproductive & Respiratory Syndrome (PRRS), also known as blue ear disease Once penetrated to the herb, the virus will actively and indefinitely proliferate in lung alveolar macrophages Up to 40% of the macrophages are destroyed which removes a major part of the body defense barrier and allows bacteria and other viruses to proliferate and do damage Virus appears to spread rapidly in a herd it may be from 4 to 5 months, at least 90% of the sows have become serum-positive of the virus There is

still no specific treatment method for PRRS (Done et al 1996) Newly discovered

and recently has been emerging and become the significant and considerable concern in many countries, particularly in countries with developed livestock industry such as Canada, US and Europe, is Post weaning Multi-systemic Wasting Syndrome (PMWS) As the name implies, the disease requires co-infection several viral pathogens but PCV2 (Porcine Circovirus Type 2) is an essential factor The disease almost spread on worldwide scale and can cause mortality up to 40% of the

herd There is still no effectively vaccine and specific treatment (Madec et al 2000)

Table 1.1 describes some emerging pathogens of concern recently

to maintain such traits in set of prime animals (Mellencamp et al 2008) Recently,

with the complete sequencing of the pig genome, along with the deeper knowledge

of animal biology promising new advances in this field will be more explosive Novel technologies such as pig gene microarrays, single nucleotide polymorphism (SNP) panels and bioinformatics are being used intensively to identify new health candidate genes for these economically important diseases

There are about 200 diseases on large domestic animals which are caused by

sequence polymorphism in single genes (Ibeagha-Awemu et al 2008) However, in

pigs, for infectious diseases cause severe damage for production, for example on pig, such as porcine reproductive and respiratory syndrome (PRRS), porcine circovirus

4

type 2-associated diseases (PCVAD), Haemophilus parasuis, and swine influenza

virus, resistance is complex and polygenic trait Therefore, the identification of markers tightly linked to resistance against such diseases is a challenge in terms of cost and level of technology

Modern immunology has revealed that a group of genes, called the major histocompatibility complex (MHC) genes, had been identified as having intimate association with both immune responsiveness and disease resistance (Charles A

Janeway et al 2001) From early time after discovery, several experiments have

showed the association of SLA complex with immune responses following vaccination Evidence of SLA association with parasite infection has been reported (Lunney & Murrell 1988) A more complete list of disease associations of the SLA

complex has been reviewed by (Warner et al 1987) and more updated and (Vaiman

et al 1998) On the other hand, the variations of MHC genes also associate to

important traits of well-being as well as reproduction process thereby affecting the survival and evolution of vertebrate population under influence from environmental

selection (Reusch et al 2001; Penn 2002; Wegner et al 2003) Given such variety of

functional importance, MHC genes have been concerned as one of the best candidates for the disease resistance and improve the productive performance traits and population genetic study (Potts & Wakeland 1993; Hedrick 1994; Bernatchez & Landry 2003)

Next to mouse and non-human primate and compare to other livestock species,

pig may be the best-studied domestic species (Meurens et al 2012) Not only be

regarded as a source of nutrition, recently, pig has emerged as a potential objects for biomedical research because traditional animal model revealed many drawbacks Rodent models do not always prove relevant in studying human diseases because of the vast genetic differences, when non-human primates show the inadequacies of bioethical issues, as well as in maintaining price Similar to other mammals, pigs have a full set of innate and adaptive immune effectors A number of miniature pig strains have been developed with distinct advantages for the use in laboratory settings Besides, an additional advantage that porcine pathogen closely related or identical to the human pathogen which most popular case is influenza virus, Ebola

5

virus Therefore, research against infectious diseases in the pig also has implications for our understanding of human infectious diseases and knowledge from human can also apply vice versa, contributes to the acquisition of new knowledge to improve both animal and human health

1.2 Biology of swine leukocyte antigens

1.2.1 Structure and function

Was first identified in mice in efforts to investigate the genetic factors that govern tissue rejection in transplantation, discovery of major histocompatibility complex (MHC) locus has brought the 1980 Nobel Prize in Physiology or Medicine

to George Snell together with Baruj Benacerraf and Jean Dausset (Snell & Higgins 1951) Comparative study shows great similarities in terms of molecular structure

and genetics of the MHC systems among vertebrates (Figure 1.1) (Yuhki et al 2007)

The MHC encodes cell-surface glycoproteins that bind antigens derived from pathogens or parasites and present them to T-lymphocytes which trigger the appropriate immune response There are two classes of MHC molecules

The class I antigen is a heterodimeric protein composed of a membrane-bound

α chain which is bonded to the monomorphic β2-microglobulin (Figure 1.1) The α chain has three domains, of which the α1 and α2 domains resemble to form the peptide-binding groove; whereas the α3 domain is a binding site for the CD8 co-receptor present on T lymphocytes as well as β2-microglobulin The class I molecules are found virtually on all nucleated cells and function in presenting peptides of 8-10 amino acids derived from intracellular compartments to CD8+ cytotoxic T lymphocytes, including pathogenic epitopes derived from viral protein and cancer cells They also interact with 9 killer cell immunoglobulin-like receptors (KIR) expressed on natural killer (NK) cells to prevent NK-mediated cytotoxicity The class II antigens are also heterodimeric proteins consisting of α chain non-covalently bound to β chain (Figure 1.1) Those two subunits comprising 2 extracellular domains for each (α1 and α2; β1 and β2, respectively) The α1 and β1

6

domains resemble and form the peptide-binding groove, whereas the β2 domain is a binding site for the CD4 co-receptor present on T lymphocytes The class II molecules are found primarily on professional antigen presenting cells such as macrophages, B lymphocytes and dendritic cells They function mainly in presenting exogenous peptides of 13-25 amino acids which are mainly derived from environment parasites (e.g bacteria, nematodes…) to CD4+ helper T lymphocytes However, recent reports have demonstrated class II molecules may be involved in

presenting of viral antigens by autophagy (Dengjel et al 2005; Paludan et al 2005;

Münz 2009)

Figure 1.1 Molecular structure of MHC class I and II

The molecular organization of the class I genes includes a leader sequence (exon 1), three exons (exon 2, 3 and 4) encoding corresponding extracellular α1, α2 and α3 domains, a transmembrane exon (exon 5) and three cytoplasmic exons (exon

6, 7 and 8) (only two cytoplasmic exons for the class Ib genes) (Figure 1.2) For the class II antigens (SLA-DR and SLA-DQ), α chains are coded by the SLA-DRA or

7

SLA-DQA genes which consist of a leader sequence (exon 1), two exons (exon 2 and 3) encoding corresponding extracellular α1 and α2 domains, and an exon (exon 4) encoding both transmembrane and cytoplasmic domains Whereas the β chains are coded by the SLA-DRB1 and SLA-DQB1 genes, which have essentially the same structure as the α-chain genes except that they have additional one or two cytoplasmic exons (exon 5 and 6)

Figure 1.2 Schematic molecular organization of the SLA genes (Lunney et al 2009)

Note: Exons are represented by the gray ovals and introns by lines

1.2.2 Disease resistance and susceptibility

As a consequence of the role of the MHC molecule in the center of the immune system, the MHC region has been associated with more diseases than any other loci

in the mammalian genomes The involvement of MHC complex in physiological conditions and pathological syndromes is mostly attributable to the antigen-presenting properties of MHC proteins in the adaptive immune system as well as those important immune-related genes that are in strong linkage with the MHC genes (Erickson 1987) Genetic variability of the MHC genes and alleles has been repeatedly associated with difference in immune responses, disease resistance and

susceptibility against infections in various species (Potts et al 1994; Jeffery &

Bangham 2000; Sommer 2005) For example, in humans, the B52 and

HLA-8

B44 phenotypes have been associated with HIV resistance, while the HLA-B51 was

associated with HIV susceptibility (Fabio et al 1992) In cattle, a significant

association was found between the haplotype 1A and susceptibility to clinical

mastitis (Lundén et al 1990) In chickens, the B2 and B21 MHC specificities

appeared to provide natural resistance against Marek‘s disease (Simonsen 1987b)

In pig, the variations of innate immune responses relating to SLA haplotype also were reported Examining the response on three homozygous SLA-defined strains of miniature swine (SLAa, SLAc and SLAd) and one recombinant strain SLAg (ABcDd), Mallard et al determined that pigs bearing the dd, dg or gg

haplotypes had significantly higher serum IgG than other haplotype pig (Mallard et

al 1989b) Besides, comparisons between haplotypes suggested that ac, dg, and gg

pigs tended to have low pre-immunization of total serum hemolytic complement

activity (Mallard et al 1989a) Influence of SLA upon phagocytic and bactericidal

activities of peripheral blood monocytes derived from miniature pig also were

measured by the work of Lacey el al (Lacey et al 1989) The authors indicated that

haplotype significantly influenced uptake and killing of each bacterium by monocytes In that, SLA ad and aa pigs were significantly better than all others at

phagocytizing S aureus, while uptake and killing of S typhimurium was highest in

homozygous aa and cc haplotypes

So that, activities of SLA molecules resulted in eliminating the development of specific pathogens in pig For examples, Lunney et al analyzed genetic effect to

Trichinella spiralis infection in NIH miniature pigs and found that the individuals of

the SLA c/c haplotype exhibited a lower burden of T spiralis larvae in muscles

(Lunney & Murrell 1988) This lower muscle burden correlated with the earlier development of a humoral antibody response in these genetically-defined pigs However, only pigs of the a/a genotype exhibited an unusual and highly significant reduction in the numbers of encysted muscle larvae after second inoculation

(Madden et al 1990) In detail, series of T-lymphocyte cell lines was isolated and analyzed from a/a minipigs after the second challenge with T spiralis These T-

lymphocyte lines showed antigen specific, cytokine dependent proliferation in

response to T spiralis antigen stimulation higher than other non-responsive

9

haplotypes (Dillender & Lunney 1993)

Also, using 82 SLA defined miniature pigs, Lumsden el al found that "dd" and

"gg" homozygous and "dg" heterozygous individuals behaved immunologically as a group distinct from the other genotypes in the responses to immunization with an

aromatic-dependent mutant of Salmonella typhimurium (Lumsden et al 1993a)

Recently, Bao et al identified the resistance to ETEC F18 in post-weaning piglets is related to up-regulation of mRNA expression of SLA-DQA to a certain

extent (Bao et al 2012) However, the analysis also suggested that SLA-DQA may not directly resisted the Escherichia coli F18, but the resistance perhaps due to

enhancing humoral immunity and cell immunity to reduce the transmembrane signal transduction of ETEC F18 bacterial LPS Yang et al also reported about association

of sequence variations in SLA-DQA and DRA to piglet diarrhea in diverse pig

breeds (Yang et al 2013;Yang et al 2014)

Not only revealed the correlation with resistance to bacterial pathogens, the SLA also shows the association with the responses to viral pathogens Using interferon-gamma ELISPOT assays on lymphocytes from two SLA defined strains

of inbred miniature pigs (c/c and d/d haplotype) infected with foot-and-mouth disease virus (FMDV), Gerner el al identified distinct novel T-cell epitopes

restricted by each haplotype (Gerner et al 2006) This suggested variable effects of

vaccine on different haplotypes Double immune histochemical staining for alpha and leukocyte markers on mesenteric lymph nodes of piglets infected with transmissible gastroenteritis virus (TGEV) showed that IFN-alpha-producing cells

IFN-were mainly positive for SLA class II (Riffault et al 2001) Checking pig immune

response following African swine fever virus (ASFV) inoculation, Juarrero et al suggested that the recovery of SLA expression during infection of pigs with ASFV is associated with survival or replacement of macrophages in the spleen

González-leading to an effective immune response against the virus (González-Juarrero et al

1992)

10

1.2.3 Association with productive traits

SLA haplotypes have been correlated with difference in swine reproductive performance including ovulation rate, embryonic development and survival, male genital tract development, androstenone level, litter size, as well as birth and

weaning weight of the piglets In addition, as reviewed by Vaiman et al (Vaiman et

al 1998), a number of commercial production traits were also significantly

associated with SLA specificities These included the back fat thickness, carcass composition, ham development, muscle malic enzyme activity and average daily weight gain

1.2.4 Genetic diversity and fitness

The swine industry nowadays is continuously facing new risks Pig production has rapidly changed from multiple small farms to a large corporate enterprise These corporate farms often rely on a small number of prime animals for breeding in order

to maintain desirable production traits, such as meat quality, litter size and growth rate Over time this heavy selection results in excessive inbreeding which leads to reduced overall genetic diversity within the herds As a result, inbreeding depression-associated genetic defects, such as developmental abnormalities, sterility and high postnatal mortality have become increasingly common as well as poor

reproductive performance (Signer et al 1999)

The reduced genetic diversity allows pathogens, particularly viruses, to easily adapt to the environment and develop as ―escape mutants‖, which can quickly spread among individuals with similar genetic background and often maintained in close confinement In addition, the establishment of livestock with genetically increased resistance to infectious diseases is also highly desirable, particularly with the emergence of multiple antibiotic-resistant zoonotic pathogens Characterizing the genes, gene products and their diversities that have significant influence on swine immune performance provides a molecular basis of understanding genetic resistance and susceptibility This information could lead to targets for genetic selection and

11

used as markers to assist in selective breeding programs of genetically resistant pigs, aiming to improve overall animal health and provide better disease protection across the herds With diverse functions in pig, SLA loci are particularly suited to this role (Sommer 2005; Ujvari & Belov 2011)

disease-1.3 Genetics of SLA

1.3.1 Mapping

There are about 150 loci identified in the SLA complex, of which at least 120 genes are predicted to be functional This fact also occurs in other species made MHC complex is one of the most gene-dense regions in the vertebrate genomes The SLA complex consists of three major gene clusters and has been mapped to porcine chromosome 7 spanning the centromere, with the class II region on the long arm and

the class I and class III regions on the short arm (Smith et al 1995) Sequencing and

mapping of the entire SLA complex based on a very common H01 haplotype has

been completed (Renard et al 2006) Detailed descriptions of the other haplotypes

were also reported The SLA class I, class III and class II regions are found to span approximately 1.1, 0.7 and 0.5 Mbs, respectively Further, extensive comparison and phylogenetic analysis of the SLA and HLA systems indicated that the SLA class I genes have more sequence homology to each other than to the HLA class I genes

(Smith et al 2005b) Their genomic organizations are also quite different (Figure 1.3) According to a recent report (Groenen et al 2012), there are 4 classical Ia

genes including SLA1, SLA2, SLA3, SLA12, together with 3 non-classical with unknown function, SLA6, SLA7, SLA8 and 2 pseudogenes SLA4 and SLA9 SLA5 and SLA11 are intermingled between haplotypes

12

Figure 1.3 Comparative genomic organization of the human and swine major

histocompatibility complex (MHC) class I region

(Lunney et al 2009)

In contrast, the SLA class II genes demonstrated much stronger sequence homology with their HLA counterparts and their overall genomic organizations are also very similar (Figure 1.4) (Lunney et al 2009) The functional genes in SLA class II included SLA-DRA, DRB1, DQA, DQB1, DMA, DMB, DOA, DOB1 And identified pseudogenes comprised SLA-DRB2, DRB3, DRB4, DRB5, DQB2, wDYB, DOB2

Figure 1.4 Comparative genomic organization of the human and swine major

histocompatibility complex (MHC) class II region (Lunney et al 2009)

13

In addition to the typical class I and class II genes, some of the most important immune-related genes were also mapped to the SLA complex, particularly in the class III region These included the genes that code for the cytokines TNFα, TNFβ and the complements C2, C4a and Bf; while those involved in the peptide presentation pathway such as the TAP and the proteasomes were mapped to the class

II region The close proximity of these genes through evolution is believed to allow for more effective and coordinate regulation of gene expression during immune

responses (Renard et al 2006; Groenen et al 2012)

1.3.2 Polymorphism and linkage disequilibrium

Like other vertebrates, SLA complex also revealed extremely high degree of genetic polymorphism within loci (Figure 1.5)

Figure 1.5 Polymorphism of MHC genes in various species (adapted from the

Immuno Polymorphism Database-MHC (IPD-MHC) website at http://www.ebi.ac.uk/ipd/mhc/)

Pig Cattle Canine Chimpan

zee GorillaClass I 116 114 48 131 20 DRB 82 132 52 31 13

DQB 44 81 36 11 10

0 20 40 60 80 100 120 140

14

This extreme polymorphism is believed as the result of evolutionary struggle of the vertebrate species under pressures generated by diverse environment pathogens (Potts & Slev 1995) This extreme polymorphism is essentially concentrated in the exon 2 and 3 of the class I genes and the exon 2 of the class II beta-chain genes which mainly due to these genomic regions encode for the protein domains which directly interact with countless number of peptides The class II alpha-chain genes,

on the other hand, display limited polymorphism despite the fact that they also encode for part of the peptide-binding groove of the class II molecules

Analysis of genes in the MHC region demonstrates a very strong linkage Some alleles occur more frequently together than expected by chance and they always tend

to be inherited as a haplotype (Vaiman et al 1979) This is also due to a low level of

recombination between the MHC genes

1.4 Nomenclature of SLA genes

Many SLA alleles and haplotypes have been characterized by diverse typing methods in various breeds of pig However, direct comparison of these findings was very difficult due to the discrepancy in the naming conventions adopted by different laboratories and the molecular nomenclature for the SLA system had not been standardized Therefore, an internationally recognized nomenclature system for the SLA genes and alleles is essential to resolve the confusion and further the development and communication of research in swine immunology and diseases Due to the efforts of numerous investigators around the world, DNA sequences of many genes and alleles in the swine MHC system have been determined and accumulated in several nucleotide sequence databases The Nomenclature Committee for Factors of the SLA System was formed at the 2002 International

Society for Animal Genetics (ISAG) conference in Göettingen, Germany to establish

the principles of a systematic nomenclature system for SLA class I and class II genes and alleles that have been defined by DNA sequencing The nomenclature for

factors of the SLA class I and class II systems were published and updated (Smith et

15

al 2005a; Smith et al 2005b; Ho et al 2009b)

For class I alleles, phylogenetic analyses and sequence comparisons were based

on the polymorphism in the exon 2 and 3 sequences Given that these regions code for the α1 and α2 domains of the MHC class I mature protein, which form the peptide-binding groove and are in direct contact with the T-cell receptors and therefore they are considered to be functionally more important For class II alleles, phylogenetic analyses were based on the polymorphism in the exon 2 sequence of the alpha-chain and beta-chain genes, as they respectively codes for the α1 and β1 domains which together form the peptide-binding domains of the MHC class II mature protein

Due to the extensive polymorphic nature of the SLA genes (Figure 1.5), the complexity of the SLA system and the difference in execution of the SLA characterization methods by different laboratories, quality control of the sequence data has always been very difficult Therefore, the ISAG SLA Nomenclature Committee has also established a set of guidelines and criteria for accepting and naming new alleles in order to maintain the highest quality of the submitted DNA sequences and to prevent the creation of a large number of non-existent alleles In addition, the Committee has established a publicly available SLA sequence database

at the Immuno Polymorphism Database-MHC (IPD-MHC) website (http://www.ebi.ac.uk/ipd/mhc/sla/index.html) to serve as a repository for maintaining a list of all recognized genes and their allelic sequences

1.5 Comparison of typing methods for swine leukocyte antigen

Due to complicated organization of SLA complex as well as the extensive polymorphism within loci, the assessment of SLA genes diversity in a population is extremely challenging Currently, there is no perfect solution for MHC typing This fact made biases in analysis of population genetic as well as association study In order to study the influence of SLA polymorphisms on animal health and other traits,

16

a robust typing method is needed to precisely characterize the alleles of SLA genes

A number of typing methods have been described

1.5.1 Serology

Serologic typing using alloantisera or monoclonal antibodies has been practiced

extensively (Lie et al 1988; Renard et al 1988; Ivanoska et al 1991) This method

is fast, simple and inexpensive However, there is limited availability of typing sera with well-defined specificities such as highly crossreactive, low resolution, not available of the typing reagents

1.5.2 Cell-based typing

A cellular-based typing method using the mixed lymphocyte culture has been

well-documented (DeWolf et al 1979; Thistlethwaite et al 1984) The mixed

lymphocyte reaction is a T-cell response by means of proliferation to class II antigens incompatibilities present on the stimulating cells The key advantage of this typing method is the sensitivity to the class II antigen mismatches, but class I antigen However, it is labor intensive, technically demanding and very time consuming

1.5.3 Sequence-based Typing

A variety of molecular techniques have been widely used for tissue typing The popularity of these techniques, especially DNA sequencing, is mainly because of the robustness and precision that they can offer SLA characterization by direct determination of DNA sequence is the most accurate method, yet it is relatively labor-intensive and technically demanding Locus-specific amplification of genomic DNA followed by cloning and nucleotide sequencing has been demonstrated by

various laboratories worldwide (Våge et al 1994; Brunsberg et al 1996) This

17

approach is usually intended only for the polymorphic portions of the genes (e.g exon 2 and 3 for class I genes; exon 2 for class II genes) as the entire coding sequence is difficult to obtain in a single reaction because of the massive interspersed intron sequences in genomic DNA A disadvantage of amplifying only portions of the entire coding region is that it is sometimes very difficult to distinguish the sequences of expressed loci from pseudogenes Therefore the specificity of the PCR primers is crucial to the reliability of the typing method and thus this requires close examination An alternative approach to address the issues mentioned above would be the cloning and sequencing of reverse transcriptase-PCR (RT-PCR) product which has also been practiced extensively Working with RT-PCR products eliminates the possibility of amplifying the pseudogene loci as pseudogenes do not usually transcribe into mRNAs It is also easier to obtain the

entire coding sequence as introns have been spliced out at the mRNA level (Ando et

al 2003) Also, by designing primers in different exons or in the untranslated

regions, it can also detect and eliminate the possibility of genomic DNA contamination Restriction fragment length polymorphism (RFLP) analysis has been

described to examine the SLA allelic differences (Chardon et al 1985) RFLP

analysis of genomic DNA on Southern blots hybridization and locus-specific amplification of genomic DNA or RT-PCR products followed by RFLP analysis

have been described (Shia et al 1995) Typing by RFLP analysis is generally fast,

easy and relatively inexpensive to perform However, the resolution greatly depends

on the number of restriction enzymes used and the resulting patterns of digestion fragments are sometimes difficult to interpret with the increasing number of genotypes PCR using sequence-specific primers (PCR-SSP) has also been

developed for several closed herds of pigs (Ando et al 2005) It is fast, accurate and

inexpensive to perform, yet it is limited to alleles with known DNA sequences and is not able to detect new alleles unless the mutations interfere with primer-binding or create unexpected PCR patterns

In this study, we developed a genomic sequence-based typing (GSBT) method using a combination of genomic DNA PCR and direct sequencing We applied our new GBST method on major genes of SLA class II including DQB1, DQA and

19

Chapter 2 Development of simultaneous high resolution typing of three SLA class II genes, SLA- DQA, DQB1 and DRB1 using genomic DNA multiplex

PCR and direct sequencing

We previously reported the development of high-resolution individual locus typing methods for three of the most polymorphic swine leukocyte antigens (SLA) class II loci, namely, SLA-DQA, SLA-DQB1, and SLA-DRB1 In this study, we improved our previous protocols and developed a method for the simultaneous amplification of the three SLA class II genes and subsequent analysis of individual loci using direct sequencing The unbiased and simultaneous amplification of alleles from the all three hyper-polymorphic and pseudogene containing genes such as MHC genes is extremely challenging However, using this method, we demonstrated the successful typing of SLA-DQA, SLA-DQB1, and SLA-DRB1 for 31 selected individuals comprising 26 different SLA class II haplotypes which were identified from the previous experiment involving 700 animals using the single locus typing methods The results were identical to the known genotypes from the individual locus typing The new method has significant benefits over the individual locus typing, including lower typing cost, use of less biomaterial, and fewer errors in handling large samples for multiple loci This new convenient method can be widely used for the typing of SLA class II genes and enable to enrich information on the genetic diversity of MHC molecules in pigs

2.1 Introduction

The major histocompatibility complex (MHC) class II genes encode heterodimeric molecules on the cell surface, which are responsible for presenting

20

foreign antigens to CD4 lymphocytes, thereby activating the immune system against

specific exogenous antigens (Lechler et al 1991) It has been suggested that the

genetic diversity of MHC molecules was increased and maintained by driven selection (Hughes & Nei 1989; Takahata & Nei 1990) Therefore, the variation of MHC genes has been shown to affect the survival and evolution of

pathogen-vertebrate population under influence from environmental selection (Reusch et al 2001; Penn 2002; Wegner et al 2003) It has been denoted that the dynamic

evolution of the SLA, not only actively occur in natural selection but also

maintained under artificial selection (Groenen et al 2012) Consequently, MHC

genes have been exploited as one of the best candidates for population genetic studies, especially molecular mechanisms of adaptation (Potts & Wakeland 1993; Hedrick 1994; Bernatchez & Landry 2003)

MHC or Swine leukocyte antigens (SLA) genes have been reported to be

associated with the MHC-mediated immune response (Lumsden et al 1993) as well

as production traits (Gautschi & Gaillard 1990) Currently, there is increasing

evidence for pigs as ideal large animal models for biomedical research (Vodicka et

al 2005; Meurens et al 2012; Lee et al 2014) and potential xenotransplantation

donors for humans (Mezrich et al 2003) Considering these diverse developments,

rapid, accurate, and convenient assessment of SLA polymorphisms is necessary Among SLA class II genes, genetic diversity is highest for DQB1 and DRB1 which are the beta () chains of DQ and DR, and followed by their alpha (α) chain genes, DQA and DRA The current Immuno Polymorphism Database of SLA (IPD-SLA) (http://www.ebi.ac.uk/ipd/mhc/sla) contains 82 DRB1, 44 DQB1, 20 DQA, and 13 DRA alleles and 21 SLA class II haplotypes Parallel with other vertebrates, such polymorphisms create challenges for the identification and classification of the genetic variations of SLA genes

For the efficient typing of these molecules, we previously reported the development of individual locus typing methods by using genomic PCR, followed

by direct sequencing for the high-resolution typing of SLA-DQA, SLA-DQB1, and

SLA-DRB1 (Park et al 2010; Thong et al 2011; Barsh et al 2012) These methods

have important advantages compared to other available typing methods, including

21

comprehensive allelic coverage, no requirement of preexisting allelic diversity, use

of genomic DNA, and removal of cloning steps, which can be applicable to field samples without much difficulty

In multiplex PCR, more than one pairs of primers are used to specifically amplify multiple targets in the same reaction to reduce time and effort in the

laboratory (Chamberlain et al 1988) It would be ideal if we could simplify the

amplification processes of the three SLA class II genes

In this study, we attempted to introduce multiplex PCR for the high resolution typing of three SLA class II genes and achieved the successful typing results by systematically modifying the previous individual locus typing We also noticed that our multiplex PCR based typing method showed improvement in robustness of the typing by sufficient elimination of SLA-DRB1- related pseudogenes

2.2 Materials and methods

2.2.1 Animals

To maximize the genetic diversity of the three SLA class II genes, SLA-DQA, DQB1 and DRB1, in the typing population, we selected 31 individuals including 29

outbred (Landrace x Yorkshire) pigs, one Korean native pig (KNP) (Park et al 2009)

and one Seoul National University (SNU) miniature pig derived from Chicago Medical University (Setcavage & Kim 1976) The single locus typing information of SLA class II genes for animals were based on the results of previous studies obtained by genomic PCR and direct sequencing that we have developed for each

locus (Park et al 2010; Thong et al 2011; Barsh et al 2012)

2.2.2 Multiplex PCR for the simultaneous amplification of SLA-DQA, DQB1 and DRB1

Simultaneous amplification reactions of SLA-DQA, DQB1 and DRB1 were carried out in a 25-μl reaction mixture containing 50 ng of genomic DNA, 0.32 μM

at 95◦C, 1 min at 65◦C, 1 min at 72◦C and a 5-min final extension at 72◦C by using T3000 Thermocycler (Biometra, Germany) PCR products were checked by electrophoresis on 1% agarose gel in 1.5X Tris-acetate-EDTA (TAE) buffer The gel was stained with ethidium bromide and visualized under ultraviolet light

2.2.3 Direct sequencing and allelic determination of SLA-DQA, DQB1 and DRB1

Direct sequencing of multiplex PCR products was carried out in the same way

as previously described for single locus typing (Park et al 2010; Thong et al 2011; Barsh et al 2012) Briefly, for degrading remained primers and dephosphorylate

dNTPs, 5 μl of amplified products were incubated for 30 min at 37◦C together with

4 U of exonuclease I (Fermentas, Canada), and 0.8 U of shrimp alkaline phosphatase (USB Corporation, USA) in 1.5X reaction buffer, then stopped after 15-min incubation at 80◦C Sequencing reactions were separately performed for each locus using ABI PRISM BigDye™ Terminator Cycle Sequencing Kits (Applied Biosystem, USA) following manufacture‘s manual Sequencing primers for each locus is described in Table 2.1 Sequencing results with quality values >20 were then imported into the CLC Workbench (CLC Bio, Arhus N, Denmark) The entire exon 2 sequences of each locus were used for National Center for Biotechnology (NCBI) BLAST analysis to identify the genotype

2.2.4 Determination of SLA class II haplotypes

The haplotype analysis was conducted using the Pypop program (Lancaster et

23

al 2007) and manual examinations For manual haplotyping analysis, we considered

haplotypes confirmed when typing results of all three loci, DQB1, DRB1 and DQA, were homozygous Other cases were the haplotypes were shared by more than one animal, or when the separation of haplotypes was possible among heterozygotes after removing one known confirmed haplotype from more than one animal

24

Table 2.1 Primers used for multiplex PCR and direct sequencing of SLA-DQA, SLA-DQB1, and SLA-DRB1

(bp) SLA-DQA

SLA-DRB1

exon 2

DRB1R+284 tctaccaggcattcgcttcatiiiiiCYSCSGGCVGCSCA sequencing Sq-mul-DRB1 TAGCTGAATTCGAATGCTGCGACTA Note: i, Inosine Primer tails with random nucleotides are indicated as lowercase letters

25

2.3 Results and discussion

To enable the multiplex genomic PCR-based SLA class II typing, it is necessary to overcome the difficulty of co-amplifying these extremely polymorphic loci, while eliminating several pseudogenes with high sequence homologies and without nonspecific amplification For example, DRB1 neighbors 4 homologous

pseudogenes (Ho et al 2009b) Because the minimum criteria for SLA class II typing is the availability of the complete sequence information of exon 2 (Smith et

al 2005a), we focused on the simultaneous amplification of complete exon 2

regions from SLA-DQA, SLA-DQB1, and SLA-DRB1

We first started with some samples for optimizing the primers of multiplex PCR For DQA, no changes were made from the previously reported DQA primers

for single locus PCR (Barsh et al 2012) For DQB1 and DRB1, changes have been

made because the amplification results were not acceptable for these two loci in multiplex PCR conditions and the primers yielded non-specific amplifications, poor sequencing chromatograms from the subsequent direct sequencing, and allelic drop out

Several DQB1 forward and reverse primers were redesigned until successful amplifications of SLA-DQB1 were obtained in the multiplex PCR conditions Finally, the forward primer, DQB1F-119, which was designed to contain the region with higher sequence conservation than the previous single reaction primer, and the reverse primer, DQB1R+295, which is 3 base pairs longer in the 3′ direction with nucleotide degenerations W (A or T) and R (G or A) at two positions over the previous single reaction primer, showed satisfactory amplification of SLA-DQB1 in multiplex PCR conditions (Appendix B)

The main issue for the optimization of SLA-DRB1 primers for multiplex PCR was to prevent the amplification of related pseudogenes especially for DRB2 which possesses only a few nucleotide differences to the DRB1 sequence and thus was

difficult to sufficiently remove it previously (Thong et al 2011) For that, the

forward primer, DRB1F-22, was designed to contain the site near the 5′ end of

26

DRB1 exon 2 where the discrimination between SLA-DRB1 and DRB2 could be possible, with a tail with 18 base non-specific nucleotides at the 5′ end of the primer

to minimize information loss at the beginning regions of the sequence The reverse

primer, DRB1R+284, the annealing control primer (ACP) (Hwang et al 2003),

successfully yielded SLA-DRB1 specific amplification in the multiplex condition (Appendix C) The ACP strategy helps to obtain successful PCR results on high GC content regions and thus eliminates the amplification of DRB pseudogenes (DRB2

to DRB5) Because the DRB1 forward primer contains the first two bases of DRB1 exon 2, our DRB1 typing results failed to determine the two nucleotides at this position Nevertheless, we were able to distinguish all currently reported DRB1 alleles without the two bases If new alleles are identified and the identity of the missing two bases is necessary, the information can be obtained by SLA-DRB1

single locus typing (Thong et al 2011)

A difficulty to develop multiplex PCR arises from difference in the amplification efficiency among multiple primer pairs, leading preferential amplification of one target sequence over another (Mutter & Boynton 1995) Among the three pairs of primers, the amplification efficiency for DQA and DQB1 was higher than DRB1 due to the complexity of DRB1 primers, both forward and reverse Therefore, an increased concentration of DRB1 primers over DQA and DQB1 was used in the multiplex condition A series of PCR tests were conducted to find a quantitative balance among the primer pairs In addition, the concentrations of other PCR components such as dNTPs, MgCl2, PCR buffer, and polymerase were adjusted to improve the results of multiplex PCR We found that an increase in the concentration of the PCR reaction buffer to 1.2X gave the most desirable results, compared to other changes (data not shown) Finally, we were able to obtain successful multiplex PCR results with three specific bands of 898, 478, and 364 bp for SLA-DQA, SLA-DQB1, and SLA-DRB1, respectively, from a single PCR reaction (Figure 2.1) The chromatograms from the multiplex PCR were of similar quality to those of individual locus PCR for all three loci

27

Figure 2.1 Successful amplification of SLA-DQA, SLA-DQB1, and SLA-DRB1 by

using multiplex PCR Note: Specific bands of 898, 478, and 364 bp for SLA-DQA, SLA-DQB1, and SLA-DRB1, respectively, were consistently amplified from samples with different SLA class II haplotypes Locus names are indicated on the top Lanes 2 to 4, amplified SLA class II genes using individual locus PCR Lanes 5 to 10, products from the simultaneous amplification of the three SLA class II genes by using multiplex PCR from samples corresponding to the haplotypes Hp-0.2/Hp-0.3, Hp-0.1/Hp-0.13, Hp-0.30/new_Hp-11, Hp-0.4/new_Hp-17, Hp-0.14/new_Hp-13, and Hp-0.11/new_Hp-

19, respectively L, 100-bp DNA ladder; N, negative control for PCR

To confirm the accuracy and efficiency of the new method, we selected 31 individuals consisting of 26 different haplotypes from our single locus typing results conducted using 700 animals (including our unpublished data) and compared the results of the two methods The number of alleles included in the selected DNA panel was 13, 17, and 20 for DQA, DQB1, and DRB1, respectively Although we typed only 31 selected animals in this study, the number of selected alleles is larger than in previous single locus typing reports by approximately 480 animals in which

there were 11 DQA, 11 DQB1, and 18 DRB1 alleles (Park et al 2010; Thong et al 2011; Barsh et al 2012) The results of multiplex PCR based typing showed not

only the efficiency regarding to handling multiple loci in a single reaction but also improvement in overcoming obstacles from previous typing designs including pseudogene noise in DRB1 and the preferential amplification in SLA-DQB1

All typing results were identical to those of individual locus typing (Appendix A) The separation of alleles from heterozygotes was carried out in the same way as our previous single locus typing (Park et al., 2010; Thong et al., 2011; Le et al.,