Analysis of porcine MHC expression profile

ARTICLESChinese Science Bulletin 2005 Vol. 50 No. 9 880 - -890 .Genes in SLA class I region can be divided into foursubclasses: Ia, Ib, Ic and non-MHC. la genes are alsoAnalysis of porcine MHCknown as“classical class I”genes. There are three func-expression profiletional genes in this subclass: SLA-I, -2, -3. They are ex-pressed in all of nucleated cells with different abundances.As is believed, their function is to bind endogenous anti-JIANG Fanbo1,, CHEN Chen', DENG Yajun',gens and present them to CD8 + T cell. Subclass Ib alsoYU Jun' & HU Songnian'contains three genes: SLA-6, -7, -8. They are expressed1. Bijing Genomics Instiute, Chinese Academy of Sciences, Beiingpreferentially in some tissues and their function is still not101318, China;very clear. But their homologous human genes mainly2. Graduate School of Chinese Academy of Sciences, Beijing 100038,conduct immuno-tolerance (such as maternal-fetal toler-Correspondence should be addressed to Hu Songnian (email: husn@Chinaance). Subclass Ic has two genes: MICI and MIC2. Theformer may be a pseudogene. Tissue expression modelgenomics.org.cn)and function of this subclass are to be studied. In human,Abstract The porcine major histocompatibility complexthis subclass encodes the ligand of NKG2D receptor in(MHC, also named swine leukocyte antigen, SLA) is associ-aaT cell and NK cell; the connection of them causes theated not only with immune responsibility and disease suscep-cells expressing MIC on its membrane to die. This may betibility, but also with some reproductive and productive traitsa kind of immuno-supervise mechanism. Also, there aresuch as growth rate and carcass composition. As yet system-atical research on SLA expression profle is not reported.In more than 30 non-MHC genes in SLA class 1 region.order to ilustrate SLA expression comprehensively andTheir function is not related to histocompatibility. Some ofdeepen our understanding of its function, we outlined thethem are highly conservative and regarded as“frameworkexpression profile of SLA in 51 tissues of Landrace by ana- genes". Regions among the framework genes are“permis-lyzing a large amount of ESTs produced by“Sino-Danish sive places", where the SLA subclasses Ia, Ib, and IcPorcine Genome Project". In addition, we also compared thegenes duplicated, mutated and converted”.expression profile of SLA in several tissues from differentGenes in SLA class II region can be divided into threedevelopment stages and from another breed (Erhualian). Thesubclasses. (i) Classical class II genes. They are SLA-result shows: (i) classical SLA genes are highly expressed inDQA, -DQB, -DRA, -DRB and -DPA. The proteins theyimmune tissues and middle part of intestine; (i) althoughSLA-3 is an SLA la gene, its expression abundance and pat-encode are also transmembrane and expressed on the sur-tern are quite different from those of the other two SLA laface of antigen presenting cells (APCs). They bind ex-genes. The same phenomenon is seen in HLA-C expression, ogenous antigens and present them to CD4+ T cells. Butsuggesting that the two genes may function similarly and no mRNA of SLA-DPA is found as yet8), so it might be aundergo convergent evolution; (ii) except in jejunum, the pseudogene. (ii) Class- I-like genes, including SLA-DMantigen presenting genes are more highly expressed in breedand -DO. The sequence of this subclass is similar to thatErhualian than in Landrace. The difference might associateof classical class II genes; however, the proteins they en-with the higher resistance to bad conditions (includingcode are not located on the surface of APCs but inside thepathogens) of Erhualian and higher growth rates of Land-cells and function as molecular companion to help therace.classical class II proteins to be folded correctly. (ili)Key words: swine leukocyte antigen (SLA), major histocompatibilityNon-MHC genes. Some genes of this subclass are impor-complex (MHC), expression sequence tag (EST), expression profile,tant in endogenous antigens processing and transferringpi(e.g. TAPI, TAP2, TAPASIN, LMP, etc.), and the othersDOI: 10.1360/982004-525function individually and differently.Genes in class II mainly encode complements andMajor histocompatibility complex (MHC) exists in ge-cytokines, which are important for innate immune system.nomes of all vertebrate, and becomes more and moreBecause SLA contains members of both innate and ac-complicated during evolution. In mammalian, MHC canquired immune system, it is correlated with disease resis-be divided into three parts, namely class I, II and II.tance/susceptibility directly. People have found that dis-MHC class I is the most evolutional variable part and theease susceptibility is dependent on SLA haplotypeother two are relatively conservative. Swine MHC is alsocently, it is found that SLA haplotype is also correlatedcalled Swine Leukocyte Antigen (SLA). It is located inwith pig productive traits such as reproductivity 0 andchromosome 7， a region containing the centromerecarcass comp1上reeding scientists(7p11- -7q11)， and about 2 Mb in size". Recently, the are laying n中国煤化IsLA. Itis wellsequence of SLA class I has been published5(0.9 Mb),known that MMHCNMHGinisogeneicandand structures of classes II and II also have been clarified heterogous transplantno*.. Despite some difficulties,by genome hybridization with some known swine and/or pig has been thought to be the best heterogous organ do-human/mouse genes as probes5.6l.nor for human. Therefore, detailed understanding of SLA88分方数据Chinese Science Bulletin Vol. 50 No. 9 May 2005ARTICLESmay help people in resolving the problem.taken from breed Landrace, 14 adult tissues from breedThough many reports on SLA are published, systematic Erhualian. All cDNA libraries were not subtracted andresearch on expression of SLA has not been reported. Toabout 10000 clones were randomly selected from eachcomprehend the expression of SLA systematically andlibrary and sequenced. Therefore, the frequency of ESTsdeepen our understanding of its function, by analyzingreflected the abundance of a gene proportionally. Afterlarge amount of expressed sequence tag (EST) producedremoval of vector sequence, ESTs that were longer thanby“Sino-Danish Porcine Genome Project”， we con-100 bp, with ambiguous bases less than 5%，were re-structed an expression profile of SLA genes in 51 tissuesgarded as “high- quality ESTs". Only high-quality ESTsof breed Landrace, and compared the expression level intissues from several developmental stages and two breeds.were put into analysis. The high-quality ESTs of all thelibraries numbered 2208 (prostate)一 10276 (jejunum),11 Material and methodaveraged at 7000. Tissue origin and high-quality EST.1 EST datanumbers were listed in Table 1. Part of these EST se-ESTs used in this paper were all produced by“Sino- quences that could be definitely annotated as anti-Danish Porcine Genome Project”. There were 51 adultgen-presenting genes (Table 2) can be accessed by numbertissues and 23 tissues from several developmental stagesDN593918- -DN596190.Table 1 Tissue origin and high-quality EST numbers for each library)"SystemTissueLF50LN115AEAImmunethymus gland57717247spleen75256297tracheobronchial lymph node7107Musclesemitendinosus805152517997semimembranosus6808triceps brachia6595supraspinatus7010infraspinatus6141vastus intermedius6993biceps femoris6312longissimus dorsi10421masseter4836left ventricle7229cerebra8863624563247853cerebellum719874446357brain stem86695896hypothalamus7201hippocampus6019spinal cord7695retina8386ye6535Respiratorylung754952416762trachea9031Digestivesalivary gland5279gullet5685stomach51536870duodenum5835jejunum10210中国煤化工765834ileumcolonCNMHG)1MHrectumliver6678500869676637(To be continued on the next page)Chj所數费ience Blletin Vol.50 No.9 May 2005881ARTICLES(Continued)SystemTissueLF50 _LF100LN115LAEAReproductiveprostate2208testis51406137ovary757477313642mamma, 14 days after delivery5785mamma, 7 days after weaning6045uterus76265737Urinarykidney86675967bladderEndocrinethyroid7589adrenal gland8287pituitary9068pancreas6808Circulationheart70555470Bloodblood899310071Othersplacenta7522mnion2617umbilical cord8331skin81958451694069707336mucosa8229rhinal mucosa4896tongue mucosa5749cartilage7373joint capsule8809mediastinum8818a) LF50, Landrace fetus 50 days; LF100, Landrace fetus 100 days; LN115, Landrace newborn 115 days; LA. Landrace adult; EA, Erhualian adult.Table 2 Accession number for 10 antigen-presenting genes in GenBankGeneAccession numberSLA-IAY135592.1AY135587.1AY247767.1AY459298.1AF464036.1AY135591.1AF464045.1AY135600.1AF464043.1AYI35594.1AY459297.1AY459299.1AY135595.1AF464016.1AY459306.1AF464015.1AF464013.1SLA-2AY247774.1AY247775.1AY247773.1AF464023.1AF464049.2AF464039.1AY135599.1AF464059.2AF464058.1AY135598.1AF464004.1AF464005.1AF464003.1SLA-3AY247769.1AY247770.1AF464042.1AF464041.1AF464040.1AF464012.1AY247768.1AF464011.1AY135602.1AF464010.1AF464009.1SLA-6 .AF464019.1AY459304.1AF464020.2AY247772.1AY247771.1AF464008.1AF464007.1 .AF464006.1SLA-7AY463541.1SLA-8AY463542.1SLA-DQBAY243101.1AY243105.1AY135574.1AF464024.1AF464028.1AY135569.1AY191778.1AF464025.1AF464026.1AY102476.1AY 102477.1AF464047.1AB009659.1AF464061.1AY135571.1AY135572.1AF464038.1AF464027.1AY135568.1AF464029.1AY 102478.1AF464030.1SLA-DRBAY243103.1AY243107.1AY135581.1AY135579.1AY135582.1AY 135583.1AY191776.1AY135578.1AY102479.1AY135575.1AF464051.1AY126721.1AF464060.1中国煤化工A14033AY102481.1AF464034.1YHCNMHGSLA-DRAAY243102.1SLA-DQAAY243100.1AY243104.1AY191777.1AY102473.1AY102474.1AY 102475.188务方数据.Chinese Science Bulletin Vol. 50 No. 9 May 2005ARTICL .ES1.4 Statistic and expression cluster of genes1.2 SLA genome and mRNA sequencesResearches on SLA genomel2-6.8 have clarified theplied in comparison between two tissues of differentgenes in each class and subclass. Only 10 of the 84 genesbreeds or development stages. When more than two tis-in SLA region have porcine mRNA sequences in Gen-sues were to be compared, general + was used5,t0. Bor-Bank. They are SLA-I, -2, -3 (subclass b), SLA-6, -7, -8toluzzi et al.!17] provided very convenient statistic software(subclass Ib) and SLA-DQA, -DQB, -DRA, and -DRBIDEG6 (Identification Differentially Expressed Genes 6(classical class II). Except MIC2， all the anti-test statistic) on their web (http://telethon.bio.unipd.it/gen-presenting genes are included. It has been proved that bioinfo/IDGE6_ form/index.html).SLA is the only region containing these genes in swineThe abundance of 10 and 74 genes in 51 tissues madegenome. For the rest 74 genes, human mRNAs take place. up virtual arrays, which was similar to microarray data.Therefore, mRNAs for the 10 and 74 genes make up the Usually, in microarray data analysis, different environ-dataset for EST to align with. .ment situations/stimuli are x-Coordinate, and genes studiedIt is well known that there are a lot of alleles for anti-are y-coordinate. Genes can be clustered according to thegen-presenting genes. The alleles distinguish from eachvariation of expression level; that is, the more concor-other only at some special positions, where the nucleic dantly genes are up/down regulated, the closer they areacid mutated during the so-called“co-evolution”of MHCclustered. In this study, taking 51 tissues as xcoordinateand pathogens. Because the allele type is not clear, forand 10/74 genes in MHC region as ycoordinate, we goteach gene, all the aleles clleted in Genbank were put expression arays by free software“Cluter' provided byinto the mRNA dataset (Table 2). For the 74 human genes,Michael Eisen's lab at the Lawrence Berkeley Nationalalleles and variant splicings were processed in the sameLab (LBNL) and the University of California at Berkeleyray. ClustalW analysis for the 100 alleles of 10 geneshttp://rana.lbl.gov/downloads/Cluster.zip). Then, 10 anti-showed that alleles for the same gene were clustered to-gen-presenting genes were clustered based on their ex-gether clearly. The sequence similarity among 10 anti-pression levels in 51 tissues by the software. The moregen-presenting genes from SLA and HLA was analyzedconcordant their expression level was, the closer theirby ClustalW, which demonstrated that genes of the sameup-stream mechanism should be, and therefore, the closertheir function should be.species were clustered firstly, therefore it is difficult to2 Resultdecide the homologous genes just according to their se-quence similarity (data not shown).2.1 Expression of 10 antigen-presenting genes1.3 Aligning parametersThe expression profile of 10 antigen-presenting genesBecause the aligningwas demonstrated (Fig. 1). It shows that: (i) There is greattwo kinds of aligning methods were used. For 10 anti-difference among 51 tissues. In a few tissues, such as tra-gen-presenting genes, the program BLATI3I (similar tocheobronchial lymph node, joint capsule, duodenum, il-BLAST but faster) was used to align the nucleic acid se- eum, jejunum and colon, the genes are relatively morequences drectly. The threshold is 100 bp, which meanshighly expressed than the others. (i) Expression level isthat if an EST scceeds in aligning to an mRNA with at dffent among sub- casses. The highest one is subclassleast 100 bp, the EST are annotated as a transcript of the Ia, then classical class II, and subclass Ib is the lowest one.gene delegated by the mRNA. If an EST aligns to more (ii) Expression patterns of these 10 genes are similar inthan one mRNAs, the one with the longest accurate align-principle, i.e. the highly expressed tissues are mainly im-ing is selected to annotate the EST. This is called parame- mune tissue and digestive tract, but there is tiny differenceter“max". Then, ESTs aligning with all the alleles for the even among genes in the same subclass. Specifically genesame gene were counted. This sum represents transcriptsSLA-3, which belongs to subclass Ia, but its expressionquantity for the gene. The abundance of a gene in a library level and high-express tssues are different from the otheris calculated bytwo subclass Ia genes.Abundance (%小= (Sum of ESTs for a gene x 1000)/For the three subclass h genes, SLA-1 and SLA-2 wereHigh-quality EST number in the library.expressed almost always as much in 51 tissues. They wereFor the rest 74 genes, the“translated” proteins were lowly expressed in respiratory, blood vessel, nerve, muscle,aligned because pig ESTs had to align with human mRNA urinary, repr中国煤化Isues, with abun-and there is more difference in nucleic acid than in protein. dance less tha. exceeds in pitui-The program TBLASTX was used in six open reading tary). In tracl|YHC N M H Gimmune sytem),frames with threshold 100 bp and parameter max. Allelesduodenum and colon (digestive tract), mediastinum andand variant transcripts were processed as the 10 genes.joint capsule, they were rather highly expressed withChj所數费ience Blletin Vol.50 No.9 May 2005883ARTICLESg己志恿SLA-8SLA-DRBSLA-DOBSLA-DRASLA-DQAFig. 1. Expression profile of 10 SLA genes in 51 tissues of adult Landrace. Different brightness represents the EST abundance. Morelight means higher abundance, while pure black means no EST. The figure was drawn by free software Cluster witten by LawranceBerkeley National Lab and the University of California at Berkeley.the abundance of about 2 %o- - 3%o In ileum and jejunumuterus and colostrums-producing mamma, their abundance(middle part of the digestive tract), they were very highly was also rather high. In respiratory system, blood vessels,expressed: the abundance of SLA-I was 7.90%o and muscle, nerve, urinary and endocrine system, their abun-9.73%oSLA-2 was 8.78%and 10.51%orespectively.dance was rather low. The highest one was SLA-DRB inAlthough as SLA-I and SLA-2, SLA-3 also belongs to tracheobronchial lymph node (3.10%), but this abundancesubclass Ia, its abundance was di stinctively lower thanwas only 1/3 of the maximum of subclass Ia.that of the former two. Except in duodenum (1.54%), il-To ilustrate the tissue abundance distribution of classi-eum (1.58%b and liver (1.44%), its abundance was lower cal SLA genes, we set an arbitrary“average abundance",than 1%oOn the other side, in some tssues where SLA-I which equals the total EST number in 51 tssuesxs 1000/and -2 were too lowly expressed to be detected, such ashigh-quality EST number in 51 tissues. This“averagelung, trachea, bladder, kidney, adrenal, prostate, skin and abundance" can be interpreted as an average abundancesome muscles, SLA-3 could be detected. Also, in liver, throughout the whole body. Average and max abundanceSLA-3 was more highly expressed than SLA-I and SLA-2.of seven classical SLA genes are listed in Table 3.The abundance of subclass Ib genes was lower than0.25%in all the 51 tissues and this subclass appeared onlyTable 3 Average and max EST abundance of classical SLA genes__in several tissues. SLA-6 could be detected in mediastinum,Gene Total EST Average abundance (%b Max abundance (%bjoint capsule, spleen, jejunum, gullet and left ventricle.SLA-DQA 860.241.23SLA-7 was only detected in spleen and mediastinum.740.201.36SLA-8 appeared in nine tissues: tracheobronchial lymphSLA-DQB710.191.69node, colon, joint capsule, heart, placenta, salivary gland,)40.263.10ovary, uterus and kidney. In some of tissues which sub-SLA-I2710.749.73class lb genes were expressed, classical SLA genes (i.e.SLA-20.7710.51subclass Ia and classical class I) were also highly ex-SLA-3)80.271.58pressed, e.g., mediastinum, joint capsule, spleen, jejunum,tracheobronchial lymph node and colon. Some tissuesObviously, the average abundances of classical class IIsuch as placenta, commonly, it is thought that classicalgenes were similar, which were much lower than those ofMHC genes are not expressed (in fact, there was no ESTSLA-I and SLA-2. The average abundance of SLA-3 wasaligned to the genes), but SLA-8 was expressed. This may far closer to classical class II genes than to subclass la.hint that SLA-8 is the counterpart of HLA-G, which is ex- Using two folds of averaoe ahundance 2s a borderline, 51pressed in human placenta chorionic epithelium and func- tissues can be中国煤化ILowly expressed.tions in maternal-fetal tolerance by inhibiting the activityThe abundandcNMHGrthantwotimesof NK cell.of average abururcuct, ntiiuing Cuiuuons that nSimilar to subclass la, classical class II genes were also was detected. (i) Highly expressed. The abundance washighly expressed in immune system and digestive tract. Inhigher than two times of average abundance. For tissue88分方数据Chinese Science Bulletin Vol. 50 No. 9 May 2005ARTICL .ESdistribution of the seven genes, see Table 4.; SLA-8Table 4 Tissue numbers of classical SLA gene classified by EST abun-. SLA-7danceClass. SLA-6Gene< 2xaverage> 2xaverage, SLA-2SLA-DQA447SLA-DRA4:8L SLA-ISLA-DQB4SLA-DRB45; SLA-3SLA-ISLA-2SLA-3 .42The highly expressed tissue numbers of the seven genesare quite similar. These highly expressed tissues are listed' SLA-DQAin Table 5. Among the 14 tissues, three are immune tissues,six are digestive in which except liver, and the other fiveare intestines (from duodenum to colon). The rest fivetissues are mediastinum, joint capsule, mamma, fat andFig. 2. SLA la, Ib and casial SLA II genes clustering according totheir expression.prostate. In mediastinum, there are a lot of lymph nodes.In joint capsule, there is a kind of M cells in the synovial same abundance in 51 tissues. It is not strange that theymembrane, which look like microphages under electronic were clustered firstly. The two genes (SLA-DRA, -DQA)microscope. They contain lysosome and may function as aencoding a chain of the MHC molecular were similar inkind of antigen processing cells. Therefore, in conclusion, average and max abundance, lowly/highly expressed tis-the high expression of SLA genes in these two tissues sue numbers, but there is slight difference in highly ex-could be attributed to immune cells. In colostrums-pro- pressed tissues, so they were clustered secondly, and thenducing mamma, fat and prostate, only classical class II,the two genes encoding the a chain (SLA-DRB, -DQB),but not subclass a genes were highly expressed. The ex- and then SLA-3. SLA-3 is a special subclass Ia gene. Ipression might come from professional antigen processing abundance was close to classical class II, but tissue dis-cells (APCs) and some non-professional APCs, such as tribution was similar to subclass a. SLA-6, -7 were clus-blood vessel endothelium cells.tered together because of their low abundance. The lastBy the software“Cluster”, these 10 genes were clus-one to be clustered in was SLA-8, which has a specifictered according to their abundance in the 51 tissues (Fig.tissue distribution.The“Sino-Danish Porcine Genome Project" acquiredAs analyzed before, SLA-I and SLA-2 have almost the four development stages for several tissues (Table 1).Table 5 Tissues of classical SLA genes "high-expressed" and the ESTs abundance of each gene (%bTissueSLA-DQA SLA-DRASLA-1SLA-2 .SLA-.Tracheobronchial Lymph node0.981.273.102.953.380.84leum1.232.117.908.781.581.030.511.371.541.71Jejunum0.390.881.09.7310.51 .0.68Colon0.962.882.310.58Joint capsule0.450.61.932.040.79Mediatinum1.02 .0.912.49Spleen1.060.66Thymus gland0.41Mamma, colostrums producing1.080.4中国煤化工MYHCNMH GProstate1.36Live1.44Chj所數费ience Blletin Vol.50 No.9 May 2005885ARTICLES .Comparisons of the expression abundance among different breeds, no difference was found except for SLA-6 in thedevelopment stages may deepen our understanding of dy- following four tissues (numbers in the brackets are Er-namics and functions of the genes. The developmenthualian versus Landrace, p<0.05): kidney(3 : 0), jejunumstages include fetus 50 days, fetus 100 days, newborn 115(3 : 1), lymph node (6 : 0), and spleen(5 : 1). Subclass Iadays and adult. For mamma, three physiological phasesand classical class II were lowly expressed in skin, cere-(colostrums-producing, 14 days after delivery, 7 days afterbra, testis, uterus and kidney for both breeds (no moreweaning) were also compared.than three ESTs in all the tissues), and no difference wasSubclass Ib genes were lowly expressed in all the tis-detected. In the other nine tissues, some of these sevensues and stages, so no difference was found. Subclass agenes were differently expressed between the two breeds.and classical class II genes were also lowly expressed inIn a word, except in liver (only SLA-3) and jejunum, anti-brain stem, cerebellum, cerebra, semitendinosus and skin,gen-presenting genes in the other tissues were more highlyand no developmental difference was detected for theseexpressed in breed Erhualian than in I andrace (Table 7).tissues. In the left six tissues, there was tiny or distinct2.2 Expression of the rest 74 genesdifference. In conclusion, these seven genes were mosthighly expressed in fetus 100 days for thymus gland, inThe expression profile for the rest 74 genes is demon-newborn 115 days for lung, in adult for colon, liver and strated in: Fig. 3, which shows: (i) Most of the 74 genesjejunum. As to mamma, colostrums-producing was higher were lowly expressed. (ii) Unlike the 10 anti-than the other two stages (Table 6).gen-presenting genes, which were mainly expressed inBreed Landrace and Erhualian are different in many immune system and digestive tract, there was no obviousveterinary aspects. Are these 10 genes also expressed dif-tissue preference for these non-MHC genes. According toferently between the two breeds? We tried to answer this their total EST numbers in 51 tissues, the 74 genes can bequestion in the following analysis.classified into three groups: (i) Without EST, there are 20There were 14 tissues from Breed Erhualian (Table 1).genes (including DPAI and SEEKI, no transcript has beenBecause subclass lb genes were lowly expressed in both discovered for the former, the latter has been proved to beTable 6 Tissues with classical SLA genes dfferenly expressed during development stages"TissueColonLiverJejunumThymus glandLungMammaDevelopment LNIS LA F50 F100 L/F50 LN1I5 LA F100AF50 LN115 LAcolostrums lactation weaningstagcTotal EST6242 5201 6678 5008 6967 10210 9503 10276 5771 7247 7549 5241 676274065785 6045SLA-DQA8*2SLA-DQBSLA-DRASLA-DRBSLA-1 .SLA-2C108*SLA-310a) Numbers are ESTs for the gene. * p < 0.05;** p < 0.01.Table 7 Tissues with classical SLA genes diferentially expressed between Erhualian and Landrace"AdrenalLymph nodeOvarySpleenStomachHeartglandBreedLELE834 10276 8397 7107 7731 7574 6297 7: 5 25 6637 6967 3642 5785 6680 8287 6870 5153 7055 488336**0 15** 1134**53**SLA-I100**中国煤化工108**:MYHC NMHG 013**10** 2E, Erhualian; L, Landrace. Numbers in the cells are ESTS for the gene. * p < 0.05;** p< 0.01.88.Chinese Science Bulletin Vol. 50 No. 9 May 2005ARTICL .ES4188POU5FIB-BPigT1T-SW-CL56:PSMB8G2A42MRPSI8B18DOBCATS13158中国煤化工HCNMH GFig. 3. EST abundance of the other 74 genes in SL A region in 51 tissues from Landrace.Different brightness represent the ES I abundance. More lightmeans higher abundance, while pure black means no EST. The figure was drawn by free software Cluster written by Lawrance Berkeley National Laband the University of California at Berkeley.Chj所數费ience Blletin Vol.50 No.9 May 2005887ARTICLES .a psudogene). (ii) Less than 40 ESTs, there are 42 genes. Metzger et al.t9) discovered that in addition to immuneAll the framework genes are in these two groups. (ii) tissue and some blood vessel epithelium, they are alsoAbove 40 ESTs. Only 10 genes are in this group; they are highly expressed in many epitheliums, including intestine(number in the brackets is the EST number for the gene):epithelium.OK-SW-CL.56 (273), HSPAIB (194), PRDMI (194)In some highly expressed tissues, the changes in ESTHSPAIA (182), COL1IA2 (111), BATI (98), TNX (54)，abundance in various development stages can be detected.PSMB8 (56), CLICI (48), and PSMB9 (41).For liver, ileum and colon, the abundance of antigen- pre-It is mentionable that in human genome, only PSMB9，senting genes peaks in adult tissue, which indicates thatPSMB8, TNX and PRDM1 are single locus genes. For the these genes are expressed in parallel with function activityother genes, there are at least two homologues in the ge- of the organ. For lung, the expression peaks 115 days afternome. OK-SW-CL.56 can even align with 17 contigs from its birth. In this period, the lung grows fastest. Thymus13 chromosomes. So ESTs for these genes can only begland is the organ where T cells develop and differentiate.interpreted as transcripts of the homologue gene family.It atrophies gradually after adolescence. We found that, inAnd because of the high similarity of these homologue this tissue, the abundance of SLA genes decreased in adultgenes, EST cannot be annotated accurately. Therefore，compared to 100 days fetus, which agrees with theanalysis for these genes has to be stopped here.physiological change. As to mamma, it is well known thatin colostrums, besides high concentration of immunoglo-3 Discussionbin, fat, vitamin and other nutritions are also of differentBy analyzing EST data, we constructed expression pro-abundance from that of common lactation. In this phase,file of SLA genes in 51 tissues for breed Landrace, andmetabolism of mamma should be different from the othercompared expression level in some tissues of severaltwo phases. It is to be elucidated why SLA genes are moredevelopment stages and two breeds. In total, most of SLAhighly expressed and what the physiological significancegenes were lowly expressed. Highly expressed genes can this phenomenon has.be classified into two parts: (i) Antigen- presenting genes,In conclusion, the result we got is consistent with ex-including subclass la and classical class II genes. They aresting knowledge and theory, so the data and analyzingexpressed preferentially in some tssues. (i) Non-MHC method we used should be reliable.genes, which were non- specifically expressed in the 51When analyzing SLA expression profile in 51 tissues oftissues.adult Landrace, we found that the expression abundanceClassical SLA genes were highly expressed in immune and tissue distribution patterm were different among 10and digestive system, especially in the middle part of di- antigen-presenting genes. The expression-based genegestive tract. Spleen and tracheobranchial lymph node are cluster was mostly consistent with function-based class-immune tssues that contain a large amount of immune ification, except for SLA-3. The gene belongs to subclasscells. High expression of subclass la and classical class IIla but its abundance was more like classical class II genes.genes reflects the antigen-presenting function of theseSimilar phenomena also exist in HLA. In the three HLAcells. For digestive tract, in the lamina propria, there area subclass la genes (HLA-A, HLA-B, and HLA-C), HLA-C islot of lymph cells and lymph node (isolated in duodenumserologically negative in 10%- -50% individualsl20o. Theand jejunum, aggregated in ileum). In lamina propria ofprotein on the membrane is 1/10- -1/2 of that of HLA-B21.colon, there are also interspersed lymph nodes. As the McCutcheon et al.201 studied this phenomenon with somemost important organ for food absorption, intestine is alsochimera genes made from HLA-B7 and HLA-Cw3. Hethe front of contacting with heterogeneous proteins andsuggested that some segment between exon 3 and 3 -UTRpathogens. A lot of antigen-presenting cells participatingof HLA-C cause the mRNA to decay more easily. SLA-3 isin extraneous antigen processing and presenting might besimilar to HLA-C not only in expression level, but also inthe reason of high expression of classical class II genes. distribution of polymorphism sites. In contrast to HLA-AUsing monoclone antibodies, Oliver et al.18] found thatand -B, the antigen-binding region (ABR, mainly encodedthere are a lot of DR+, DP+, DQ+ cells in the intestineby exon 2) of HLA-C is less polymorphic than transmen-lamina propria and lymph node, furthermore, these cells brane encoding regionl211, as the case of SLA-3mainly exist in the middle of digestive tract. Subclass h Phylogenic tree of the six genes (HLA-A, -B, -C, SLA-I,-2,genes were also highly expressed in intestine. Besides-3) shows that the distance between the twoimmune cells, epithelium cells should also contribute to it.species is greater than that of the genes. But the similarityFirstly, in ileum and jejunum, the abundance of SLA-I, -2in expression中国煤化τicates that SLA-3was 2- -3 fold higher than that of immune tissue (tracheo- and HLA-C my have undergonebranchial lymph node)， and in these two tissues, the“convergentYHC N M H G particular func-abundance of classical class II genes was not concordanttions, which is different from the other two subclass Iawith that of SLA-1 and SLA-2. Secondly, in a research on genes.rat, using monoclonal antibody of these two subclasses,Making comparison between breed Landrace and Er-88号方数据.Chinese Science Bulletin Vol. 50 No. 9 May 2005ARTICL .EShualian, we found that except in jejunum, most of the an- microphage that expresses MHC class II in adrenal zonatigen- presenting genes were more highly expressed in reticularis. This kind of cell can also produce interleukin-6breed Erhualian. Erhualian belongs to Taihu strain, Jiang-(IL-6) and interleukin-1 (IL-1). These two interleukins, .hai/River and Sea type in China. Taihu strain is famous forespecially IL-1, are not only important mediators betweenits high reproductivity and Erhualian is the highest one.immune system and hypothalamus-pituitary- adrenal axis,Erhualian has better bad-condition (including cold/hot- but also promote synthesid and secretion of adrenal ster-ness/diet/high attitude) resistance, and shorter sex-mature oid (mineralocoticoid, glucocorticoid, androgen, and es-time than Landrace. It is also characterized by its excellenttrogen). These hormones are important in immuoregula-meat quality. But this breed has slower growth rate, less tion, stress response, glucose, protein and fat metabolism.lean ratio (only 38.8%- 45%)21. Landrace is a famous The higher expression of subclass h and classical class IIlean-type breed. It has faster growth rate, higher efficiency genes reflects larger quantity or higher activity of micro-of feed utilization and higher lean ratio (65%), but is not phage in adrenal of breed Erhualian, which may functionbad-condition resistant and requires better feed. There indirectly to promote the sickness resistance of the breed.might be many genetic backgrounds in charge of the dif- This might be one of the many reasons for sickness-resis-ference between the two breeds in reproductivity, bad- tance among different porcine breeds. But we cannot 28-condition resistance, growth rate and feed utilizing effi-certain whether the difference between breed Erhualianciency. SLA is undoubtedly an important onel8]. It is cor- and Landrace is caused by cell quantity or expressionrelated with sickness resistance and immuno-responsivity. level of each cell, and we should make clear whether theBesides, there are some quantitative trait loci (QTL) in difference is breed- or ale specific, and furthermore theand near SLA region. These QTLs are correlated withupstream regulation factors.productive traits such as growth rate, lean ratio, fat com-The only tissue where antigen- presenting genes areposition and meat quality-23- 261. Although the correlationmore highly expressed in breed I andrace than in Er-between sickness resistance, immuno-resposivity and pro- hualian is jejunum. Jejunum is an important organ whereductive traits is uncertain (see Rothchild MF, htp://www.food are digested and adsorbed. Antigen-presenting genes,nsif.om/Cnfere99rothil.html), this uncerta- especially subclass h genes are highly expressed in thisinty might just be caused by linkage disequilibrium be-tissue. We have reason to believe that intestine epitheliumtween immmo-genes and QTLs in this region. To cultivateshould contribute mainly to this high expression. Butbreed with both good productive traits and high sickness whether the higher expression is crrelated with higherresistance, SLA should be a noteworthy region.efficiency of feed utilization of breed Landrace needs toErhualian has better sickness-resistance than Landrace.be further researched.It is demonstrated by our EST data that antigen-presentingWe found that there was significant difference betweengenes were more highly expressed in the former thanthe two porcine breeds in expression level of antigen-pre-latter. Is there any correlation between the two phenomenasenting genes, and the difference was more significantFirst, in lymph node (classical class II genes) and spleenthan that among development stages. But whether the dif-(subclass Ia and classical class II genes), antigen-pre-ference is just individual or common to the breed? To an-senting genes were much more highly expressed in Er swer this question, more systematical research need to behualian than in Landrace (Table 7). As these genes aredone.mainly expressed in immune cells, we may infer that thereWe proved the feasibility of using EST data to con-are more active immune cells in Erhualian. Second, sub-struct expression profile (or named“electric Northern",class Ia and classical class II genes were much morefor highly expressed genes, and the difference betweenhighly expressed in adrenal of Erhualian_ than that oflibraries can be displayed digitally. But this method isLandrace (Table 7). Marx and Bornsteinl27] studied theineffective for lowly expressed genes. This is a commondistribution of HLA class II molecular in human adrenal limitation for EST analysis.sby immuno-histochemistry and proved that zona reticu-Acknowledgements We thank the Agriculture Academy of Denmarklaris of the cortex is the main area for the molecular. Thisin“Sino-Danish Porcine Genome Projct for providing all the cDNAlibraries for breed I ,andrace.zona mainly secretes androgen and glucocorticoid. Thestudy also showed that the expression level of MHC class ReferenceII molecules is parallel to the development of cells in zona1. Geffotin, C, Popescu, C. P.Cribiu, E. P. et al., Assignment ofreticularis, i.e. it increases from infancy, peaks at adult andMHC in swine to chromosome 7 by in situ hybridization and sero-decreases when geting old. Additionally, the expression中国煤化工-219.was also reversely correlated to the malignant degree of2. Shigenari,de sequencing analy-adrenal tumor28. Therefore, the molecules may partici-sis of theYHC N M H Glocated between thenon-classical ana ClassICal SLA class 1 gene clusters, Immunoge-pate in the apoptosis procedure of adrenal cell. Combin-netics, 2004. 55(10): 695- -705.ing immuno-histochemistry and in situ hybridization,3. Renard, C., Vaiman, M, Chianilkulchai, N. et al .. Sequence of theGonzalez-Hernandez et al.2930 suggested that it was thepig major histocompatibility region containing the classical class ICh月毁糖ience Bulletin Vol. 50 No. 9 May 2005889ARTICLES19. Metzger, R., Mempel, T., Joppich, I. et al., Organ-specific distribu-4. Chardon, P.. Rogel-Gillard, C, Cattolico, L. et al., Sequence of thetion of major histocompatibility antigens in rats, Pediatr. Surg. Int,swine major histocompatibility complex region containing all2000, 16(4): 285- -292.non-classical class I genes, Tissue Antigens, 2001, 57(1): 55- -65.20. McCutcheon, J. A, Gumperz, J, Smith, K. D. et al, Low HLAC5. Velten, F. W, Renard, C, Rogel-Gaillard, C. et al, Spatial arexpression at cell surfaces correlates with increased turnover ofrangement of pig MHC class I sequences, Immunogenetics, 1999,49<11-12):919- -930.6. Peelman, L. J., Chardon, P.. Vaiman, M. et al, A detailed physical21. Zemmour, J., Parham, P, Distinctive polymorphism at the HLA-Cmap of the porcine major hstocompatibility complex (MHC) classlocus: Implications for the expression of HLA-C, J. Exp. Med,III region: Comparison with human and mouse MHC class II re-1992, 176(4):937- -950.gions, Mamm. Genome, 1996, 7(5): 363- -367.22. Chinese Livestock Records Compiling Comitee, ed, Chinese7. Amadou, C, Evolution of the Mhc class I region: The frameworkPig Breeds Record, Shanghai: Publish House of Science and Tech-hypothesis, Imunogenetics, 1999, 49(4): 362- -367.nology of Shanghai, 1986.8Chardon, P., Renard, C, Gallard, C. R. et al, The porcine major3. Bidanel, J. P., Milan, D., Ianuelli, N. et al, Detection of quanti-histocompatibility complex and related paralogous regions: A e-tative trait loci for growth and fatness in pigs, Genet. Sel. Evol.,view, Genet. Sel. Evol, 2000, 32(2): 109- -128.2001, 33(3): 289- -309.9. Madden, K. B, Murell, K. D., Lunney,J. K. et al, Tihinella spi-24. Demeure, O., Renard, C., Yerle, M. et al., Rearranged gene orderralis: Major histocompatility complex associated elimination ofbetween pig and human in a QTL region on SSC 7, Mamm. Geencysted muscle larvae in swine, Exp. Parasitol, 1990, 70(4):443- -451.nome, 2003, 14(1): 71-80.10. Renard, C, Vaiman, M. Possible rlationships between sLA and 25. Genet, C. Renard, C. Cabau, C. et al, In the QTL region sur-porcine reproduction, Reprod. Nut. Dev., 1989, 29(5): 569- -576.rounding porcine MHC, gene order is conserved with human昌-11. Chen, F. X., Tang, J, Li, N. L. et al., Novel SLA class I alleles ofnome, Mamm. Genome, 2001, 12(3): 246- -249.Chinese pig strains and their significance in xenotransplantation, 26. Milan, D., Bidanel, J. P.. Ianuelli, N. et al, Detection of quanti-Cell Res. 2003, 13(4): 285- 294.tative trait loci for carcass composition traits in pigs, Genet. Sel.12. Xu, X. C, Naziruddin, B.. Sasaki, H. et al, Allele speific andEvol, 2002, 34(6): 705- -728.peptide - dependent recognition of swine leukocyte antigen class I27. Marx, C, Bornstein, S. R., Wolkersdorfer, G. W. et al, Relevanceby human cytotoxic Tcell clones, Transplantation, 1999, 68(4):of major histocompatibility complex class II expression as a hall-473- -479.mark for the cellular differentiation in the human adrenal cortex, J.13. Kent, W. J., BLAT-- -The BLAST-like alignment tool, GenomeClin. Endocrinol. Metab., 1997, 82(9): 3136- -3140.Res, 2002, 12(4): 656- -664.14. Audic, s., Claverie, J. M., The significance of digital gene expres-28. Wolkersdorfer, G. W., Marx, C, Brown, J. W. et al, Evaluation ofsion profiles, Genome Res., 1997, 7(10): 986- -995.apoptotic parameters in normal and neoplastic human adrenal, En-15. Stekel, D. J. Git, Y, Falciani, F.. The comparison of gene expres-docr. Res., 1996, 22(4):411-419.sion from multiple cDNA libraries, Genome Res.. 2000, 10(12):29. Gonzalez Hernandez, J. A, Bornstein, S. R., Ehrhart Bornstein, M.2055- -2061.et al, Interleukin-6 messenger ribonucleic acid expression in hu-6. Claverie, J. M, Computational methods for the identification ofman adrenal gland in vivo: New clue to a paracrine or autocrinedifferential and coordinated gene expression, Hum. Mol. Genet.,regulation of adrenal function, J. Clin. Endocrinol. Metab., 1994,1999, 8(10): 1821-1832.79(5): 1492- - 1497.17. Bortoluzzi, s.. d' Alessi, F, Danieli, G. A., A computational recon-30. Gonzalez Hermandez, J. A, Bomstein, S. R., Ehrhart-Bornstein, M.struction of the adult human heart transcriptional profile, J. Mol.et al, IL-1 is expressed in human adrenal gland in vivo. PossibleCell Cardiol, 2000, 32(11): 1931-1938.role in a local immune-adrenal axis, Clin. Exp. Immunol, 1995,18. Oliver, A. M., Thomson, A. W.. Sewell, H. F et al., Major histo-99(1): 137- -141compatibility complex (MHC) class II antigen (HLA-DR, DQ, andDP) expression in human fetal endocrine organs and gut, Scand. J.(Received October 26, 2004; accepted March 17, 2005)Immunol, 1988, 27(6): 731- -737.中国煤化工MHiCNMH G899方数据Chinese Science Bulletin Vol. 50 No. 9 May 2005