Plenary Session 3

Immunogenomics

Chairs: Stuart Carter, Imre Kacskovics

Room 242

Wednesday, September 6th

9.00-10.30

45
 Towards Xenotransplantation: Somatic cloning of pigs expressing various
PS3-01

immunmodulatory transgenes
Niemann H.
Head of Dept. Biotechnology, Institut für Tierzucht (FAL), Mariensee, 31535 Neustadt, Germany
The hyperacute rejection response (HAR) which was the first hurdle in porcine–to-human xeno-
transplantation, can now be overcome in a clinically relevant manner by expression of human
complement regulatory proteins (DAF, hCD59, CD46) in transgenic pigs. However, despite seve-
re immunosuppressive treatment the acute vascular rejection (AVR) with the disseminated intra-
vascular coagulation (DIC) as the preeminent feature and the cellular rejection remain major obs-
tacles for long-term survival of a porcine xenograft. DIC is frequently observed in a pig-to-pri-
mate xenotransplant model and is caused by activation of the endothelial cells mainly attributed
to incompatibilities between human and porcine coagulation factors. The goal of our research is
the generation and characterization of improved lines of multi-transgenic pigs targeting the AVR
and specifically this coagulation disorder. We produced transgenic pigs expressing constructs for
the human thrombomodulin (hTM) against the genetic background of already existing, well cha-
racterized hCD59/DAF transgenic pigs which allow the control of HAR. These triple transgenic
pigs were produced by somatic nuclear transfer and expressed hTM at the mRNA and protein
level in a variety of organs. Porcine kidneys from these multi-transgenic pigs are used in in vitro
perfusion assays with human blood. We will also employ pigs with a knock-out for the 1,3-α
galactosyltransferase, the gene that codes for the epitope that is primarily responsible for induc-
tion of HAR. The generation of multi-transgenic pigs with high expression of the various mole-
cules will significantly improve long-term survival of porcine xenografts and will be a major step
forward in the clinical application of xenotransplantation.

Genomics of Salmonella enterica serovar Typhimurium invasion in pig

PS3-02

jejunum
Niewold T.A.1,3*, Veldhuizen E.J.A.2, van der Meulen J.1, Haagsman H.P.2, Smits M.A.1,
Hulst M.M.1
1
Animal Sciences Group of Wageningen University and Research, Lelystad, The Netherlands
2
Utrecht University, Utrecht, The Netherlands
3
Present address: Katholieke Universiteit Leuven, Heverlee, Belgium
Salmonella enterica serovar Typhimurium (S. typhimurium) species are an important cause of
invasive gastroenteritis in pigs and humans. Whereas there are in vitro data on gene expression
upon Salmonella interaction with cells and cell lines, very little is known about in vivo interac-
tions during actual invasion. We investigated the early interaction of Salmonella with intact pig
small intestinal mucosa using the pig small intestinal segment perfusion (SISP) model. Whole
mucosal gene expression of infected vs uninfected segments was analyzed by cDNA array at 0-
8h post-infection. Invasion of Salmonella caused upregulation of only eight jejunal mucosal
genes. The genes involved are part of the innate immune system, and regulate the inflammatory
response. We conclude that the most striking is the limited number of differentially expressed
genes in intact mucosa compared to in vitro results. This suggests that S. typhimurium is adapted
to evade in vivo host responses by down regulating the local inflammatory response.

Modification of the porcine immature dendritic cell transcriptome upon

PS3-03

interaction with Pseudorabies virus

Flori L.1, Mariani V.1,2*, Rogel-Gaillard C.1, Cochet M.3, Bidanel J.P.2, Chardon P.1, Lefèvre F.3
1
INRA-CEA, Laboratoire de Radiobiologie et d’Étude du Génome
2
INRA, Station de Génétique Quantitative et Appliquée

46
3
INRA, Laboratoire de Virologie et Immunologie Moléculaires, Jouy-en-Josas, France
Development of innate and adaptive immune response during the course of a viral infection is
dependant, at the mucosal entry site, upon early interactions between incoming virions and two
cell types: the cells of the mucosal epithelium and immature dendritic cells (iDCs) which are the
first immune cells interacting with the virus. The Pseudorabies virus (PrV) is a well-studied patho-
gen and a good model to study these mechanisms in the pig species. In order to identify key cel-
lular genes involved in these processes, we used a transcriptomic approach in which modifications
of cellular transcriptome and expression of all PrV genes were simultaneously analysed at several
times post-infection. We used in vitro systems for each cell type: the porcine PK-15 epithelial cell
line, in which PrV replicates very efficiently, and primary iDC differentiated in vitro from pig
monocytes using GM-CSF and IL-4. Cell types were mock-infected or PrV-infected and RNA
extracted at different times post-infection. Four replicates of PK15 cells and iDCs from 4 animals
were analysed. Two complementary DNA chips were hybridized. One array corresponded to a
laboratory-designed cDNA chip comprising 1856 probes including 530 probes targeting the swine
leukocyte antigen (SLA) complex region that contains many immune function genes, 80 immune
non-SLA genes, 80 PrV genes and 1056 randomly chosen probes for data normalization. The
second array referred as Qiagen-NRSP8 is an oligo chip comprising 13297 probes. The analysis
of the transcriptomic profiles will be presented.

BAC transgenic mouse model to analyze the role of the bovine FcRn in

PS3-04
IgG and albumin metabolism
Bender B.1, Bodrogi L.1, Mayer B.2, Schneider Z.2, Baranyi M.1, Zhao Y.3, Hammarstrom
L.3, Bosze Z.1, Kacskovics I.2*
1
Department of Animal Biology, Agricultural Biotechnology Center, Godollo, Hungary
2
Faculty of Veterinary Science, Szent Istvan University, Budapest, Hungary
3
Division of Clinical Immunology, Karolinska Institute, Huddinge, Sweden
The MHC class I related Fc receptor for IgG (FcRn), which is composed of the FcRn heavy chain
and the beta2-microglobulin (b2m), protects IgG from intracellular catabolic degradation, plays
important roles in IgG transcytosis in several mucosal layers and is involved in the maternal
immune transport. Most recently, it has been shown that FcRn binds albumin and protects it from
degradation just as it does IgG. Both IgG and albumin bind FcRn at low pH at distinct sites.
Recently, we detected that bovine FcRn is expressed in endothelial cells too and showed that IgG
with enhanced binding to FcRn has increased serum persistence in cattle, and thus FcRn can effec-
tively protect IgG from degradation. In order to study the regulation of the bovine FcRn heavy
chain gene and analyze its role in IgG and albumin metabolism we generated BAC transgenic
mice overexpressing the bovine FcRn heavy chain. Pharmacokinetic studies showed that the IgG
half-life was about two and eight times longer in transgenic compared to WT and KO mice, res-
pectively. Preliminary data show that the endogenous albumin level was slightly higher in TG
mice than in WT mice whereas KO animals show about half the level of it compared to WT ani-
mals. These data indicate that the bovine FcRn heavy chain is indeed expressed in the mouse
endothelial cells and form a functional receptor that protects both IgG and albumin.
Supported by grants of OTKA T049015, GAK-CALVES05, COST B20 Action and the Swedish
Research Council.

Analysis of the CEA gene family in the dog reveales a species-specific

PS3-05

evolution of CEACAM involved in the regulation of immune responses

Kammerer R.1,2*, Popp T.1, Singer B.3, Sievers B.1,2, Zimmermann W.1
1
LIFE-Center, LMU, Munich, Germany

47
 2
GSF, Munich, Germany
3
Charité, Berlin, Germany
The carcinoembryonic antigen-related cell adhesion molecule (CEACAM) gene family is loca-
ted in the expanded leukocyte receptor complex. The primordial CEACAM gene family was
composed of five genes (CEACAM1, CEACAM16, CEACAM18 - CEACAM20), two of them
(CEACAM1, CEACAM19) are expressed by immune cells. CEACAM1 regulates NK, T and B
cell functions, and induces dendritic cell maturation. Human CEACAM1 contains two immuno-
receptor tyrosine-based inhibition motifs (ITIM); in mice, one ITIM is replaced by an ITSM. The
CEACAM1 gene gave rise to a recent expansion of the CEACAM family by gene multiplication,
resulting in significant differences of the gene families in humans and rodents. We analyzed the
CEACAM family in further mammals and in one marsupial and identified orthologous genes for
all primordial gene family members in cattle, dog, elephant and opossum. These analysis indi-
cate that the primordial CEACAM1 contained a ITIM-ITSM motif as found in mice. In cattle and
dogs three and seven different CEACAM1-related genes were identified, respectively.
Surprisingly, this group of genes codes for proteins, which differ dramatically in the anchorage
to the cell membrane in humans, mice and dogs. While these molecules are predominantly secre-
ted in mice and GPI anchored in men, in the dog, most of them represent transmembrane mole-
cules with cytoplasmic domains harbouring ITAM-like motifs. Based on these data and the
notion that a number of diverse murine and human pathogens use CEACAM1 as a cellular recep-
tor, we have established a model that explains the species-specific evolution of the CEACAM1-
related genes.

Transcriptional profiling of peripheral blood mononuclear cells (PBMC)

PS3-06

from tuberculosis-infected and healthy control cattle: Identification of a

gene expression signature of infection
Meade K.G.1*, Gormley E.1, Farrelly C.O.2, Costello E.3, Keane J.4, Zhao Y.5, MacHugh
D.E.1
1
Animal Genomics Laboratory and Bovine Tuberculosis Research Unit, UCD, Ireland
2
Education and Research Centre, St. Vincent’s University Hospital, UCD, Ireland
3
Central Veterinary Research Laboratory, Ireland
4
Dublin Molecular Medicine Centre, St. James’s Hospital, Ireland
5
Computational and Systems Biology Group, National Cancer Institute, USA
Microarray analysis of messenger RNA abundance was used to investigate differential gene
expression of PBMC from six cattle infected with Mycobacterium bovis and six uninfected
controls. A targeted immunospecific cDNA microarray platform representing 1 391 genes was
used with a reference experimental design. Comparisons revealed significant expression differen-
ce for genes with important roles in the immune response to infection. In total, 378 gene features
on the BOTL-5 microarray were differentially expressed between tuberculosis-infected and
control animals at the P < 0.05 level and 151 at P < 0.01. The direction of the gene expression dif-
ferences was markedly asymmetrical, with 134 gene features upregulated and 244 downregulated
in infected animals (P < 0.05). Unsupervised hierarchical cluster analysis revealed a panel of 15
genes with an expression profile across 12 animals (6 infected and 6 controls) that predicted infec-
tion status. These results suggest that gene expression profiling may form the basis of novel
immunological tests to enhance diagnosis of M. bovis infection in cattle. Real-time quantitative
reverse transcription PCR in eight animals per treatment group validated the microarray results.
In broad terms, the combined results were suggestive of immunosuppression in the tuberculosis-
infected cattle relative to uninfected control animals.

48
DNA typing of swine class II major histocompatibility complex

PS3-07
polymorphisms
Rochelle E.S., Schook L.B.
Animal Science, University of Illinois, 374 Edward R Madigan Lab, 1201 West Gregory Drive,
Urbana IL, 61801, USA
Tissue transplant experiments have revealed a cluster of genes responsible for graft rejection, the
major histocompatibility complex (MHC). The MHC is a region dense in immune genes with a
role in antigen presentation. Porcine MHC genes are highly polymorphic, varying amongst breeds.
Unlike other species, such as mouse, cattle, and human, serological approaches to derive MHC
haplotypes have been unsuccessful. To identify porcine haplotypes it is essential to utilize DNA
typing approaches. In humans and swine (who have 85% homology in MHC class II), exon 2 of
MHC class II genes encodes the first domain of the beta-chain and this domain is in direct contact
with the processed peptide and/or the T-cell receptor during antigen presentation. Specific poly-
morphisms in these class II genes leads to an increase in susceptibility to diseases such as pneu-
moconiosis, autoimmune diseases, and allo/xeno-graft rejection. In an effort to understand poly-
morphisms and tissue expression of MHC class II genes (DQB and DRB) DNA of divergent
breeds were analyzed. Primers targeting exon 2 of DQB and DRB were shown to be locus speci-
fic, and amplified all tested haplotype alleles. Expression of exon 2 PCR amplicons was verified
through sequence analysis revealing over 35 different alleles among the 12 breeds with an avera-
ge of 7 animals per breed represented. Preliminary data reveals breed specific polymorphisms
within both genes. A subsequent analysis was done to analyze expression in various tissue
samples. These tissue samples were analyzed through RT-PCR to quantitate RNA expression in
each.

Toward global analysis of porcine alveolar macrophage proteins through

PS3-08
two-dimensional electrophoresis and mass spectrometry
Pérez E.*, Ramirez-Boo M., Garrido J.J., Moreno A.
Departamento de Genética, Universidad de Córdoba, Campus Rabanales, Córdoba, Spain
Alveolar macrophages (AM) belong to a phenotype of macrophages with distinct biological func-
tions and they are the primary phagocytes of the innate immune system. However, many cellular
mechanisms that control these processes are largely unknown. Because proteins determine all bio-
logical processes of the cell, proteomics is a useful tool to detect infection-associated proteins in
order to determine target proteins related with immune response. High-throughput protein analy-
sis would be greatly enhanced by a more detailed knowledge of porcine AM biology and the dyna-
mics of pig-pathogen cell interactions. In this study, we obtained a reference 2-DE map of the por-
cine AM proteins by carrying out a systematic proteomic analysis. The proteins were separated by
2-DE using a 5-8 range pH gradient in isoelectric focusing and approximately 800 spots were
detected. A representative set of the major proteins was subjected to MALDI-TOF analysis and
106 proteins were identified on the basis to their similarity with the human homologous proteins.
A large number of cytoskeletal and metabolic proteins were found as well as some proteins rela-
ted to cell mobility and immunological functions. Finally, other proteins implicated in the cell
signalling process, transport or apoptosis were also identified giving a wide overview of the por-
cine AM protein map. The results presented here provide a useful tool for future studies on AM
biology, as well as for identifying the proteins implicated in the immune response or physiologi-
cal changes that could be correlated with porcine diseases.

49
 Hardy-Weinberg expectation in dog breeds: Implications for genetic
PS3-09

studies
Short A.D.1*, Barnes A.2, Kennedy L.J.1, Fretwell N.3, Jones C.3, Thomson W.1, Ollier
W.E.R.1
1
University of Manchester, UK
2
University of Liverpool, UK
3
WALTHAM Centre for Pet Nutrition, Leicestershire, UK
Hardy-Weinberg (HW) is based on five basic assumptions - absence of natural selection, random
mating, large population size, negligible mutation rates and absence of migration and gamete
transfer. It is used as the 'gold standard' of preliminary data analysis in population based genetic
studies and as a measure of genotyping accuracy. The canine genome has increased the potential
for canine genetic studies but statistical packages used for data analysis use algorithims which
require the population to be within the specifications of HW. Canine breeds have arisen from a
desire to enhance and maintain certain phenotypic traits, opposing many HW assumptions -
mating is non-random, incestuous and subject to bottlenecks, selection is paramount for breed trait
retention and founder populations are small. We performed a candidate gene study of 33 pedigree
breeds and a set of crossbreeds (n = 8-97) using 114 single nucleotide polymorphisms (SNPs)
from 13 genes. We calculated minor allele frequencies (MAF) and HW by breed and by sub breed
(e.g. Poodle as a collective breed, then separately as Miniature, Toy and Standard) for each SNP,
to establish whether breeds would fit the HW model. MAF were highly variable and some SNPs
breed specific. For all breeds, the majority of SNPs were in agreement with the HW expectations
(p > 0.05). The percentage of SNPs in HW ranged from 68.6 for Golden Retrievers to 98.3 for
Basset Hound. Compliance with the HW model has major implications for canine case-control
studies and genetic analysis. Statistical analysis can be performed using standard, readily available
methods without the need to develop specialised packages targeted at canine genetic analysis.

DLA-DRB1, DQA1 and BQB1 alleles in Alaskan grey wolves

PS3-10

Kennedy L.J.1*, Angles J.M.2, Barnes A.3, Ollier W.E.R.1, Happ G.M.4
1
University of Manchester, UK; 2 University College Dublin, Ireland; 3 University of Liverpool, UK;
4
University of Alaska, Fairbanks, Alaska, USA
The Major Histocompatibility Complex in the dog contains highly polymorphic genes, many of
which are critical in regulating the immune response. Since domestic dog breeds were derived
from the Grey Wolf, Canis lupus, it would be anticipated that at least some of the DLA class II
alleles identified in dogs should also be present in wolves and vice versa.
Sequence based typing was used to characterise a series of 150 Alaskan Grey Wolves for their
DLA-DRB1, DQA1 and DQB1 alleles and haplotypes. DLA data from a panel of 3500 dogs from
over 85 different breeds were available for comparison.
Within these wolves, many new DLA class II alleles were identified: 10 DRB1, 2 DQA1 and 7
DQB1 alleles. All of these alleles were found in at least two different wolves, and in both homo-
zygous and heterozygous animals. We also found two DRB1, 9 DQA1 and 11 DQB1 alleles that
had previously been identified in dogs. The new alleles formed many new haplotypes (14). Only
one three locus haplotype, but several DQ haplotypes, have also been found in dogs.
The wolf population studied had relatively few DLA alleles previously found in dogs, and also
contained a range of new DLA alleles. These wolves may represent a remnant population, proba-
bly descended from Asian wolves, whereas most of the dogs tested had a European origin. A
single European wolf has also been typed and carried a haplotype found in many of the dogs tes-
ted. One DQB1 allele originally identified in wolves has since been found in Shih Tzu, a breed of
Asian origin. These data suggest that the original wolf ancestors of Asian and European dogs may
have had different gene pools, which may be reflected in the DLA alleles present in dog breeds.
50