The graft response to transplantation: a gene expression profile analysis
Kenneth Christopher1,
Thomas F. Mueller1,
Rachel DeFina1,
Yurong Liang1,
Jianhua Zhang2,
Robert Gentleman2 and
David L. Perkins1
1 Laboratory of Molecular Immunology, Department of Medicine, Brigham and Womens Hospital, Harvard Medical School
2 Biostatistics Unit, Department of Biostatistical Science, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115
 |
ABSTRACT
|
---|
Little is known regarding the graft response to transplantation injury. This study investigates the posttransplantation response of genes that are constitutively expressed in the heart. Constitutive heart and lymph node tissue-restricted gene expression was first analyzed with DNA microarrays. To demonstrate changes following transplantation in genes constitutively expressed in the heart, we performed vascularized murine heart transplants in allogeneic (BALB/c to B6), syngeneic (B6 to B6), and alymphoid (BALB/c-RAG2-/- to B6-RAG1-/-) experimental groups. Temporal induction of genes posttransplant relative to constitutive expression was evaluated with DNA microarrays. Dendrograms and self-organizing maps were generated to determine the dissimilarity between the experimental groups and to identify subsets of differentially expressed genes within the groups, respectively. Expression patterns of selected genes were confirmed by real-time PCR. Biological processes were assigned to genes induced posttransplant using the AnnBuilder package via the Gene Ontology Database. Post-transplant, a shift was noted in genes classified as defense, communication, and metabolism. Our results identify novel components of the graft response to transplantation injury and rejection.
gene regulation; rodent
 |
INTRODUCTION
|
---|
THE PROCESS OF ALLOGRAFT REJECTION involves a robust inflammatory and immune response that is mediated by effector cells infiltrating the graft. In previous studies, we have focused on the innate immune response to the graft posttransplantation (4, 11). A poorly understood corollary is the graft response to transplantation. Little is known regarding the graft response to the surgical wounding, stress, ischemia, and rejection associated with transplantation. Thus the objective of this study was to evaluate the graft response to transplantation injury.
Our assumption was that the graft response would be complex, involve multiple genes, and be detectable as changes in gene expression. A previous report analyzing global expression profiles of both mRNA and protein found that 94% of differentially regulated genes had detectable changes in the level of mRNA (13). This suggests that a substantial proportion of differential regulation involves detectable changes in transcription or mRNA stability. Our experimental design thus uses DNA microarrays to monitor the expression of large subsets of genes that demonstrate differential constitutive tissue-restricted expression. Based on the presence or absence of constitutive gene expression in untransplanted heart (graft source) or lymph node (phenotype of at least a portion of the infiltrating cells), we assigned genes to one of four of the following subsets: 1) heart restrictive, 2) lymph node restrictive, 3) expressed in neither heart nor lymph node (neither subset), and 4) expressed in both heart and lymph node (both subset). To differentiate changes in gene expression due to adaptive immunity vs. antigen-independent mechanisms, expression of each tissue-restricted subset gene was compared with our microarray database (4) generated to evaluate gene expression in alymphoid, syngeneic, and allogeneic experimental groups of heart transplantation. The alymphoid group consists of recombinase activating gene (RAG)-deficient donors and recipients that lack functional T and B cell antigen receptors and thus cannot mount an adaptive immune response. The syngeneic group eliminates the presence of alloantigen by using genetically identical donor and recipient strains. The allogeneic group uses donor and recipient strains with a complete major histocompatibility complex (MHC) class I and II mismatch. This experimental approach should suggest gene subsets expressed by the graft tissue that are differentially regulated following transplantation.
To identify small numbers of groups of genes based on similar expression patterns we turned to the self-organizing map (SOM) algorithm. The SOM algorithm is an artificial neural network that has been used in many different engineering and scientific applications such as image recognition, signal processing, and gene expression analysis (23). Because of its high classification power, the SOM is an efficient strategy for mining a wide range of genomic information from large numbers of sequences. Using a combination of hierarchical clustering algorithms and SOMs, we identified unique profiles of gene expression in each experimental group. These profiles identified modulated gene expression in both the cardiac-restricted and lymph node-restricted categories. To illustrate changes of biological functions between and within the experimental groups, each gene identified in the profiles was assigned a broad biological process via the Gene Ontology database (1). Gene Ontology is designed to produce a controlled vocabulary that applies to all eukaryotes. Biological processes are defined as broad biological goals (i.e., defense). The assignment of individual genes to biological processes allows for a broad perspective on differential gene expression.
Current immunosuppressive treatments commonly used in clinical transplantation are directed at the adaptive immune response, in particular, the T cell response. Our findings show that transplantation elicits a differential genomic profile within the graft of increasing expression of genes involved in the biological processes of defense and cell communication. Genes constitutively expressed in the graft prior to transplantation which are upregulated at day 1 following transplantation are those involved in cytoprotection. Genes upregulated at day 7 following transplantation include genes involved in innate immunity and cytoprotection. Our results provide novel insights and identify components of the graft response to the complex process of transplantation.
 |
METHODS
|
---|
Mice.
Eight to 12-wk-old male mice including BALB/cByJ (BALB/c) (H-2d), C57BL/6J (B6) (H-2b), C57BL/6J-Rag-1tm1Mom (B6-Rag) (H-2b) (Jackson Laboratories, Bar Harbor, MA), and BALB/c-AnNTac-Rag2tm1N12 (BALB/c-Rag) (H-2d) (Taconic) were used as donor and recipients in the transplant experiments. H-2 refers to the murine major histocompatibility complex. An MHC haplotype (H-2b, H-2d) is the haploid chromosomal combination of alleles of MHC loci in a given mouse strain. Mice are maintained in vented racks with constant temperature and humidity in our animal facility under virus-antibody-free conditions.
Vascularized heterotopic cardiac transplantation.
Murine hearts were transplanted as previously described (5). Briefly, hearts were harvested from freshly killed donors and immediately transplanted into 8- to 12-wk-old recipients that were anesthetized intraperitoneally with 60 µg/kg of pentobarbital sodium. The donor aorta was attached to the recipient abdominal aorta by end-to-side anastamosis, and the donor pulmonary artery was attached to the recipient vena cava by end-to-side anastamosis. All surgical procedures were completed in less than 60 min from the time that the donor heart was harvested. Donor hearts that did not beat immediately after reperfusion or stopped within 1 day following transplantation were excluded (>95% of all grafts functioned at 1 day following transplantation). The native heart of the recipient was not surgically manipulated and remained functional. Donor grafts were harvested at 24 h or 7 days following transplantation and prepared for microarray analysis or real-time PCR.
DNA microarrays.
Experimental procedures for the GeneChips were performed according to the Affymetrix GeneChip Expression Analysis Technical Manual. Briefly, double-stranded cDNA was synthesized by means of the SuperScript Choice system (Invitrogen, Carlsbad, CA) and a T7(dT)24 primer (Genset Oligos, La Jolla, CA). The cDNA was purified using phenol/chloroform extraction with Phase Lock Gel (Eppendorf) and concentrated by ethanol precipitation. In vitro transcription was performed to produce biotin-labeled cRNA using a BioArray HighYield RNA Transcript Labeling Kit (Enzo Diagnostics, Farmingdale, NY) according to the manufacturers instructions. cRNA was linearly amplified
40-fold with T7 polymerase using double-stranded cDNA that was synthesized. The biotinylated RNA was cleaned with an RNeasy Mini kit (Qiagen, Valencia, CA), fragmented to 50 to 200 nt, and then hybridized to an Affymetrix murine array (Mu11kB), which contains probe sets for 6,500 genes and expressed sequence tags (ESTs). After being washed, the array was stained with streptavidin-phycoerythrin (Molecular Probes, Eugene, OR), amplified by biotinylated anti-streptavidin (Vector Laboratories, Burlingame, CA), and then scanned on an HP GeneArray scanner. The intensity for each feature of the array was captured with Affymetrix GeneChip Software, according to standard Affymetrix procedures.
Statistics and data analysis.
Array data was analyzed with Microarray Suite 4.0. A single expression level for each gene was derived from the 20 probe pairs representing each gene, 20 perfectly matched (PM) and mismatched (MM) control probes. The MM probes act as specificity control that allow the direct subtraction of background and cross-hybridization signals. Each array was normalized to a standard of 2,500 units per probe set. To determine the quantitative RNA level, the average of the differences (avg diff) representing PM - MM for each gene-specific probe set was calculated. The expression of each probe set was categorized as present (P), marginal (M), or absent (A). Calculations of means and variances were performed with JMP statistical software (SAS Institute, Cary, NC).
Cluster analysis.
Cluster analysis was performed using Cluster and TreeView (8) and GeneCluster software (23) software. Average difference values were analyzed by Cluster without normalization using the hierarchical clustering algorithm with average linkage clustering. Briefly, dissimilarity is determined by calculation of the Pearson correlation coefficient between each series of values from each experimental group (8). After processing, the dendrogram was visualized by TreeView. SOMs were generated by GeneCluster on genes showing a minimum of a twofold change among the experimental groups using a 2 x 3 geometry of six seed maps (23). Six maps were selected empirically to eliminate clusters with few genes or large standard deviations. The centroids and standard deviations of the groupings were analyzed using 100 epochs. Based on multiple heuristic observations, increased numbers of nodes produced clusters with low numbers of genes, whereas decreased numbers of nodes produced larger standard deviation. Increasing the number of epochs (=500) did not produce detectable changes in the clusters or standard deviations. SOM output is then graphically presented as a set of maps (Fig. 3 and 6) showing a relatively small group of genes that exhibit similar expression characteristics across all control and experimental conditions. Tables 2 and 3 list groups of genes with similar expression patterns as determined by the SOM algorithm output.

View larger version (26K):
[in this window]
[in a new window]
|
Fig. 6. SOMs of expression following transplantation of genes constitutively expressed in neither heart nor lymph node. Clusters of genes with similar patterns of expression in the experimental groups were identified using SOMs. Experimental groups distributed on x-axis include untransplanted BALB control heart (), untransplanted B6 control heart ( ), alymphoid graft heart (day 1) ( ), allogeneic graft heart (day 1) ( ), alymphoid graft heart (day 7) ( ), allogeneic graft heart (day 7) ( ), and control lymph node ( ). The y-axis is relative gene expression, which is auto-scaled to enhance visualization and thus variable among the profiles. The mean expression is depicted as the symbols with error bars indicating 2 standard deviations. The number of genes in each cluster is 10 (cluster 0), 6 (cluster 1), 2 (cluster 2), and 9 (cluster 3). The algorithm was initiated with 2 x 2 geometry using 100 epochs. Genes corresponding to each map are listed in Table 3.
|
|
Gene Ontology Process assignment.
The mapping of the GenBank accession numbers associated with a microarray oligo sequence to Gene Ontology (GO) terms is performed using the AnnBuilder package of the Bioconductor project (http://www.bioconductor.org). The AnnBuilder package is an R-based system (http://www.r-project.org) that maps a given set of GenBank accession numbers to LocusLink identifiers and then to any data in a public database that can be linked to LocusLink identifiers. As both UniGene (Mm.data.gz) and LocusLink (LL_tmpl.gz) provide mappings between GenBank accession numbers and LocusLink identifiers with intersections, the GenBank accession numbers were first mapped to LocusLink identifiers using both sources. A unified mapping was obtained by including the mappings that were agreed by both sources, and the mapping by UniGene was included when they disagree. In a trial based on the genes on Affymetrix U95AV2 GeneChip, we only found 27 disagreements for a total of 12,625 genes.
The unified mapping was then used to process the LL_tmpl file from LocusLink to link GenBank accession numbers to GO identifiers and then to GO terms from the Gene Ontology Consortium (go.xml) using GO identifiers as the linkage for data. The end product of the mapping is an extended markup language (XML) file containing entries of GenBank accession numbers and the GO terms associated with each accession number. The tree structure of the GO terms was reflected by associating a GO term to the GenBank accession numbers mapped to the term as well as all the accession numbers mapped to the offspring of that GO term. The one-to-many relationship between GenBank accession numbers and GO terms was also maintained by associating all GO terms to all the related GenBank accession numbers. When a mapping was provided by only one source, that mapping was included in the unified mapping.
Based on the XML file generated by the AnnBuilder package, we then assigned biological processes as defined by the Gene Ontology database to the microarray data. The biological process is defined as the biological objective to which the gene contributes. Most of the genes analyzed could be assigned to the following five broad processes: cell communication, cell growth/ maintenance (including cell organization and biogenesis), metabolism, defense response, and the stress response. The AnnBuilder package cannot assign a biological processes to genes that do not have a GO identifier. Thus genes that do not have GO identifiers are not included in the Figs. 4 and 7. In Tables 2 and 3, genes without GO identifiers were assigned broad biological processes based on the function reported in the literature. A detailed description of each gene including the identifier, standard name, characterization, and relevant references, as well as related biological processes according to the GO classification is provided on our web site (http://perkinslab.bwh.harvard.edu).

View larger version (42K):
[in this window]
[in a new window]
|
Fig. 4. AnnBuilder package output of the Gene Ontology defined biological processes based on the entire mu11kB GeneChip (A) and genes constitutively expressed in heart upregulated posttransplantation at day 1 (B) or day 7 (C). Pie charts of biological processes based on genes constitutively expressed in heart subsequently downregulated at day 1 (D) or day 7 (E) posttransplantation. Up- or downregulation is determined by normalized gene expression values at day 1 or day 7 less normalized heart control values.
|
|

View larger version (18K):
[in this window]
[in a new window]
|
Fig. 7. Independent verification of microarray quantification. A: relative mRNA levels of C1qb subunit, calpactin I, collagen- 1 (Col3A1), apolipoprotein E (ApoE), metallothionein-1 (MT-1), smooth muscle -actin (Actg2) genes. B: relative mRNA levels of decorin, cytochrome C oxidase ViaH (Cytcox), troponin-I (Trop), creatine kinase (CK), lactate dehydrogenase 2B (LDH2B), myosin light chain 2 (Mylpc), and ventricular alkali myosin light chain (Mylc) genes. Relative mRNA levels were measured via the increase in fluorescence caused by the binding of SYBR Green to double-stranded DNA produced during PCR amplification. Real-time PCR data is shown in fold change (2- CT) relative to control heart.
|
|
Real-time quantitative PCR.
Total murine heart RNA was individually isolated from three untransplanted control hearts, three day 1 allogeneic graft hearts, and three day 7 allogeneic graft hearts. Ten micrograms of RNA was separately reverse transcribed using SuperScript II RNase Reverse Transcriptase (GIBCO, Carlsbad, CA). Specific apolipoprotein E, C1qb subunit, calpactin I, collagen-
1, creatine kinase, cytochrome C oxidase ViaH, decorin, GAPDH, lactate dehydrogenase 2B, metallothionein-1, myosin light chain 2, smooth muscle
-actin, troponin-I, and ventricular alkali myosin light chain primer pairs were designed using the Primer Express software (Applied Biosystems, Foster City, CA). Primer sequences used follow: apolipoprotein E, forward 5'-GAGCCGGAGGTGACAGATCA-3', reverse 5'-CTCCCAGGGTTGGTTGCTT-3'; C1qb subunit, forward 5'-AGGACCATCAACAGCCCCTT-3', reverse 5'-CTTTTCGAAGCGAATGACCTG-3'; calpactin I, forward 5'-CCAAGTGCCTACGGGTCAGT-3', reverse 5'-CCTCTCAGCATCGAAGTTGGT-3'; collagen-
1, forward 5'-TGAAGATGTCGTTGATGTGCAG-3', reverse 5'-CGGCTGGAAAGAAGTCTGAGG-3'; creatine kinase, forward 5'-AAGGTGCTGACCCCTGACCT-3', reverse 5'-TGGTGTCTCCTTATCGCGAAG-3'; cytochrome C oxidase ViaH, forward 5'-TGACCTTTGTGCTGGCTCTTC-3', reverse 5'-AGCAGTTAAGGGAGCAGAGGG-3'; decorin, forward 5'-AGCTTCAACAGCATCACCGTT-3', reverse 5'-AGGAACATTGGCCAGACTGC-3'; GAPDH, forward 5'-TTGTGGAAGGGCTCATGACC-3', reverse 5'-TCTTCTGGGTGGCAGTGATG-3'; lactate dehydrogenase 2B, forward 5'-CTGTAGTGGGCGTTGGACAA-3', reverse 5'-CCAGAATGCTGATGGCACAT-3'; metallothionein-1, forward 5'-TGCTCCACCGGCGG-3', reverse 5'-TTTGCAGACACAGCCCTGG-3'; myosin light chain 2, forward 5'-CACTTCTCCTTGGTCCACTATGC-3', reverse 5'-GCCCATAATGTTGTAGTCCACG-3'; smooth muscle
-actin Actg2, forward 5'-TGTGTGACAATGGCTCTGGC-3', reverse 5'-CATCATCTCCTGCAAAGCCTG-3'; troponin-I, forward 5'-CCCACCTCAAGCAGGTGAA-3', reverse 5'-CCTCCCGGTTTTCCTTCTCA-3'; and ventricular alkali myosin light chain, forward 5'-AGAGAAACTGATGGCTGGTCAAG-3', reverse 5'-TCATAGTTGATGCAGCCGTTG-3'. Direct detection of the PCR product is monitored by measuring an increase in fluorescence due to the binding of SYBR Green to double-stranded DNA. Reactions are performed in a MicroAmp Optical 96-well reaction plate (Applied Biosystems) using for each separate well 5 µl of cDNA mix, 5 µl of primer, and 10 µl of SYBR Green Master Mix (Applied Biosystems). Each well contains the primer pair for amplification of one of the parameters of interest. The gene-specific PCR products are continuously measured by means of the GeneAmp 5700 Sequence Detection System (Applied Biosystems) during 40 cycles. The threshold cycles, i.e., the PCR cycle at which an increase in reporter fluorescence above a baseline signal can first be detected, of each target product is determined and set in relation to the amplification plot of housekeeping gene GAPDH. All experiments are run in duplicate, and the same thermal cycling parameters are used. Nontemplate controls and dissociation curves are used to detect primer-dimer conformation and nonspecific amplification. Fold change is calculated relative to control heart cycle threshold (CT). The CT value is defined as the number of PCR cycles required for the fluorescence signal to exceed the detection threshold value. With the PCR efficiency of 100%, the CT values of two separate genes can be compared (
CT), and the fold difference = 2-(CT - CTcontrol) = 2-
CT.
 |
RESULTS
|
---|
To determine baseline constitutive tissue-restricted gene expression, we analyzed RNA from two untransplanted control hearts and control lymph node with DNA microarrays. Results were analyzed with Microarray Suite software to calculate absolute difference values and to determine absent or present calls for each gene. This analysis indicated that in the untransplanted control tissues, 343 genes were expressed only in heart, 301 only in lymph node, 479 in both heart and lymph node (both subset), and 1,492 in neither heart nor lymph node (neither subset) (Fig. 1). Based on this analysis, those genes were then assigned to one of the four subsets (heart-restricted, lymph node-restricted, both, or neither). Changes in gene expression following transplantation of the four subsets were then determined by comparison to our murine array databank (one array per experimental condition)(4). Although the murine array (Mu11kB) contains probe sets for 6,500 genes and ESTs, only 5% of the probes are related to the immune response (Fig. 4A).

View larger version (32K):
[in this window]
[in a new window]
|
Fig. 1. Order-three Venn diagram of categories of constitutive gene expression. Categories include heart only (343 genes), lymph node only (301 genes), both heart and lymph node (479 genes), and neither heart nor lymph node (1,492 genes). Genes with upregulation of less than a twofold change in expression relative to mean expression were excluded.
|
|
To analyze changes in gene expression following transplantation, we performed vascularized murine heart transplants in allogeneic (BALB/c > B6), syngeneic (B6 > B6), and alymphoid (BALB/c-RAG2-/- > B6-RAG1-/-) experimental groups (Table 1). The alymphoid group, which lacks a functional adaptive immune response, used both donor and recipient mice deficient in the RAG gene, which is necessary to produce functional T and B cell antigen receptors. The donor and recipient mice in the syngeneic group were genetically identical. Thus both the alymphoid and syngeneic groups lack an alloantigen-driven adaptive immune response, and all changes in gene expression are attributed to antigen-independent mechanisms. In contrast, the allogeneic group has a complete MHC class I and II mismatch, plus minor antigen differences, and rejects grafts with a potent adaptive immune response in
8 days.
We analyzed changes in the expression of genes in transplanted tissue according to one of the four subsets (heart-restricted, lymph node-restricted, both, or neither). We began by assessing changes of expression posttransplantation in the heart-restricted subset. To determine dissimilarity of genes expression between the experimental groups, we generated dendrograms with hierarchical clustering algorithms. The dissimilarity between groups in the dendrogram is calculated as the sum of the Pearson correlation coefficients between expression values of genes from each group. Figure 2A demonstrates the changes in gene expression at day 1 following transplantation. As expected, lymph node is markedly dissimilar from the control and graft heart groups. The syngeneic and alymphoid groups show only mild dissimilarity and modestly differ from the allogeneic group. Although the Fig. 2A groups allogeneic day 1 separate from syngenic and alymphoid day 1, the dissimilarity distances are quite small relative to that of the lymph node. Because of the small dissimilarity distances, a firm conclusion relative to the relationship of differential gene expression found in allogeneic, syngenic, and alymphoid day 1 grafts is not possible. However, in the kinetic analysis at day 7, the allogeneic group is most similar to lymph node and markedly dissimilar from the control and day 1 group (Fig. 2B). Thus the dendrograms in Fig. 2 serve to illustrate that the gene expression pattern in the allogeneic graft at day 1 changes at day 7 to an expression pattern more similar to that of a lymph node.

View larger version (29K):
[in this window]
[in a new window]
|
Fig. 2. Dendrogram of expression following transplantation of genes constitutively expressed in the heart. Agglomerative hierarchical clustering algorithm was used to generate dendrogram of experimental groups including allogeneic graft heart (day 1), alymphoid graft heart (day 1), syngeneic graft heart (day 1), untransplanted control heart (BALB), untransplanted control heart (B6), and control lymph node (A); and allogeneic graft heart (day 1), allogeneic graft heart (day 7), untransplanted control heart (BALB), untransplanted control heart (B6), and control lymph node (B). The x-axis distance is proportional to the dissimilarity between groups.
|
|
To identify subsets of genes within the heart-restricted subset that were differentially expressed in the experimental groups, we generated SOMs analyzing expression in control heart (BALB), control heart (B6), alymphoid graft heart day 1, allogeneic graft heart day 1, allogeneic graft heart day 7, and untransplanted control lymph node (Fig. 3 and Table 2). Table 2 lists the complete output of maps generated for the heart-restricted subset. In Fig. 3, map 1 generated a list of genes upregulated in all groups (alymphoid and allogeneic days 1 and 7) following transplantation. Genes in map 1 (Fig. 3) have previously reported associations with stress and injury responses and include
B2-crystallin, metallothionein-1, and collagen-
1. Maps 2 and 3 (Fig. 3), which are similar, generated lists of genes (Table 2) that are downregulated following transplantation and include cytochrome c, troponin-I, and myoglobin. Interestingly, we did not detect a subset of constitutive heart genes highly upregulated in the allogeneic group at day 7. This observation suggests that the similarity observed between the control lymph node and allogeneic day 7 groups in the dendrogram is due to the loss of constitutively expressed cardiac genes, rather than the gain of lymph node related genes in graft heart sample.
Biological process assignment following transplantation of the genes upregulated in the heart-restricted subset at day 1 (Fig. 4B) shows cell maintenance (33%) and cell communication (42%) are prominent. At day 7 (Fig. 4C), the stress and immunologic consequence of transplantation is reflected with increases in the expression of defense (22%) and metabolism genes (17%). The majority of heart-restricted subset genes downregulated at day 1 following transplantation (Fig. 4D) assigned by Gene Ontology are cell maintenance (56%) and metabolism (27%). The proportion of heart-restricted subset genes downregulated genes at day 7 (Fig. 4E) following transplantation was similar to day 1. This observation suggests the transcriptional program is altered in favor of defense genes at the expense of genes involved in cell growth and maintenance. Genes assigned to metabolism were both up- and downregulated in response to transplantation, which is illustrative of the complexity of the graft response to transplantation.
To assess changes of expression posttransplant in the "neither" genes subset, we generated hierarchical dendrograms. The neither subset should contain genes not constitutively expressed in heart or lymph node, but might be inducible following transplantation in either graft tissue or infiltrating cells. Analysis of changes in expression at day 1 following transplantation of the neither subset showed only low levels of dissimilarity, even between the lymph node and heart tissues (Fig. 5A). This observation is consistent with the fact that the genes are not constitutively expressed but are upregulated to low levels of expression following transplantation. Similarly, the kinetic dendrogram shows only low levels of dissimilarity even in the day 7 allogeneic and lymph node groups (Fig. 5B). Detection of specific subsets of genes with similar expression profiles using SOM identified small numbers of genes with these characteristics relative to the heart-restricted and lymph node-restricted categories (Fig. 6 and Table 3). Map 0 (Fig. 6) contains genes that were upregulated in all groups following transplantation but not expressed constitutively. Consistent with this profile, genes included in this subset are IFN-ß and 24p3. Map 1 (Fig. 6) contains 11 genes upregulated at day 1, but not day 7, following transplantation and includes Eta-1/osteopontin, fibronectin, ngal, and gly96. Taken together, genes from the neither subset include a small number of genes induced following transplantation.

View larger version (29K):
[in this window]
[in a new window]
|
Fig. 5. Dendrogram of posttransplantation expression of genes constitutively expressed in neither heart nor lymph node. Agglomerative hierarchical clustering algorithm was used to generate dendrogram of experimental groups including allogeneic graft heart (day 1), alymphoid graft heart (day 1), syngeneic graft heart (day 1), untransplanted control heart (BALB), untransplanted control heart (B6), and control lymph node (A); and allogeneic graft heart (day 1), allogeneic graft heart (day 7), untransplanted control heart (BALB), untransplanted control heart (B6), and control lymph node (B). The x-axis distance is proportional to the dissimilarity between groups.
|
|
To identify the temporal induction of genes posttransplant we compared gene expression of allogeneic day 1 and allogeneic day 7 to control heart. The 343 heart-restricted subset genes (Fig. 1) were identified in the posttransplant allogeneic day 1 and day 7 microarray data. The expression values were normalized and expressed in standard deviation (z) units. The normalized values of each individual control B6 and BALB/c gene were averaged to obtain a control z value of that particular gene. The difference in gene expression between normalized control and allogeneic data sets was calculated (Tables 4 and 5). To limit changes based on background noise, normalized gene expression difference values
z greater than 0.1 and less than -0.1 were considered minimal for differential gene expression from control tissue. Heart-restricted subset genes with the largest increase in normalized expression values (
z > 1.2) in the allogeneic day 1 relative to control include
B2-crystallin, metallothionein-1, and collagen-
1 (Table 4). Heart-restricted subset genes with the largest increase in normalized expression values (
z >2.5) in the allogeneic day 7 relative to control include C1qc, apolipoprotein E, C1qa, and C1qb (Table 5). This observation is consistent with the presumption that between days 1 and 7, differential expression of graft genes likely reflects the response of the graft to the graft ischemia and or surgical stress of transplantation at day 1 followed by an antigen-independent immune response at day 7. Differential gene expression in the microarray data was confirmed in repeat murine cardiac transplants by fluorogenic quantitative PCR assay. In the allogeneic transplants, calpactin I, collagen-
1 (Col3A1), and metallothionein-1 fold change mRNA expression increased at day 1 relative to control heart (Fig. 7A). At day 7 posttransplant, C1qb expression increased 10-fold relative to control. In addition, calpactin I, Col3A1, and apolipoprotein E are upregulated greater than twofold relative to control at day 1. The real-time PCR data in Fig. 7B shows greater than twofold downregulation in response to transplantation at day 1 of myoglobin, LDH2B, myosin light chain 2 (Mylpc), ventricular alkali myosin light chain (Mylc), and troponin-I. Also consistent with the microarray data, at day 7 posttransplant, real-time PCR expression of creatine kinase, cytochrome C oxidase ViaH, decorin, lactate dehydrogenase 2B, myoglobin, myosin light chain 2, smooth muscle
-actin, troponin-I, and ventricular alkali myosin light chain are all downregulated (Fig. 7). These results all correlate with the temporal increase in
z relative to control observed in the microarray data. An exception, calpactin I shows a relatively low
z and a robust fold change at day 1. This is likely from a low amount of calpactin I mRNA present in the control heart, whereby any change in mRNA expression at day 1 will result in a high fold change relative to control heart.
 |
DISCUSSION
|
---|
It is well established that the infiltration and activation of inflammatory cells is an important component of allograft rejection. Our hypothesis is that graft responses to transplant injury are important in initiating pro-inflammatory stimuli that amplify subsequent immunity. Our results identify subsets of genes and ESTs that are candidates to mediate this graft response.
Changes in expression of genes constitutively expressed in heart, but not lymph node tissue, should be enriched for graft genes responding to transplant injury. However, we cannot exclude that some genes were constitutively expressed only in heart tissue and subsequently upregulated in inflammatory cells following transplantation. Importantly, based on the degree of dissimilarity generated in the dendrogram between the control lymph node and the three graft heart groups, it is likely that most genes remain differentially expressed following transplantation. Also, the selection of heart-restricted subset genes with modulated expression in both the allogeneic group and alymphoid group should identify genes regulated independently of adaptive immunity. Map 1 of the SOM of heart only subset genes (see Fig. 3 and Table 2) contains six such genes. Among the known genes, metallothionein-1 is induced by stress including exposure to heavy metals. Interestingly, metallothionein has also been shown to provide protection against electrophiles and oxidant stress and to be induced during acute liver injury and regeneration (24). Similarly,
B2-crystallin, a member of the heat shock protein family, is a negative regulator of apoptosis that acts distally in the conserved cell death apparatus by interfering with the maturation of caspase-3 (14).
B2-crystallin is upregulated by ischemia in congestive heart failure (27) and upregulated in an ischemia/reperfusion model in swine (7). Collagen type 3 is produced by myofibroblasts following ischemia in response to angiotensin II, fibroblast growth factor, platelet-derived growth factor, atrial natriuretic factor, prostaglandin E2, and nitric oxide (22). Thus previously reported functions of genes in map 1 genes are consistent with induction due to transplantation injury. Maps 2 and 3, which show overlapping profiles of expression, include genes downregulated following transplantation. Members of these maps such as creatine kinase, lactate dehydrogenase, and troponin-I are released following myocardial ischemia. Decorin, a small proteoglycan produced by vascular smooth muscle cells, has been shown to inhibit TGF-ß and diminish fibrosis (10).
Analysis of genes constitutively expressed in neither heart nor lymph node, but detected following transplantation in both the alymphoid and allogeneic groups, identified a subset of inducible genes (see Fig. 6 and Table 3). Map 0, and to a lesser extent map 1, in the SOM identified several pro-inflammatory genes. Members of these maps include interferon-ß, a cytokine shown to be important in innate immune responses and in interactions with cells of the adaptive immune system and 24p3, a lipocalin acute phase response protein that has been suggested to function as a regulator of inflammation (15). Eta-1/osteopontin, a cytokine produced by multiple cell types including chondrocytes, has been shown to promote TH1 T cell responses (2). Gly96 which is highly expressed in fibroblasts, is an immediate early gene with a short half-life inducible by serum growth factors (3), and fibronectin has been shown to modulate TGF-ß induction of IL-8 (26). Because all of these genes are upregulated in both the alymphoid and syngeneic groups, the mechanism appears to be independent of adaptive immunity. Interestingly, maps 2 and 3 contain genes detected during the late phase of rejection (day 7) and include the interferon inducible IP-30, the antigen processing gene Ma, and serum amyloid A3 which is a major contributor to the acute phase response. Based on previously reported functions, some of these genes may be induced by the severe injury to the graft including myocyte necrosis that is caused by the rejection process.
The differential expression of graft genes following transplantation illustrates the involvement of the innate immune response. Identification of the genes with the greatest difference in gene expression between normalized control heart and normalized experimental data sets further demonstrates the involvement of stress and innate immunity. Comparison of the genes with the largest positive change relative to control in days 1 and 7 posttransplantation shows both groups include
B2-crystallin, a negative regulator of apoptosis (14), the cytoprotectant metallothionein-1, collagen-
1, the smooth muscle cell mitogen HBEGF, the lysosomal protease cathepsin D, SPARC, and calpactin I, a binding site related to macrophage migration. Consistent with ischemic injury and eventual graft rejection, genes with the largest negative change at days 1 and 7 include myoglobin, troponin-I, creatine kinase, and
-cardiac myosin heavy chain. High
B2-crystallin expression at day 1 is likely in response to the ischemia and reperfusion of transplantation.
B2-crystallin is upregulated in an ischemia reperfusion model in swine (7) and is noted to be a negative regulator of apoptosis that acts distally in the conserved cell death apparatus by interfering with the maturation of caspase-3 (14). At day 7, the four graft genes with the largest change in normalized expression values were C1qc, apolipoprotein E, C1qa, and C1qb. C1q is a member of the innate immune system that is involved in pattern recognition and opsonization of foreign particles. C1q initiates the classic complement pathway via binding to IgG or IgM that has bound to antigen. In skin grafting experiments, C4 knockout mice deficient in classic pathway activation have a notably impaired antibody response against donor tissue antigens (16). In xenografts, experimental inhibition of the interaction between C1q and immunoglobulin prolongs graft survival (9). Experimental evidence support the role of apolipoprotein E in the modulation of the immune response. Apolipoprotein E antagonizes growth-factor-induced lymphocyte proliferation (17). Compared with normal mice, apolipoprotein E-deficient mice fail to suppress early proliferation of Listeria monocytogenes (21) and are more susceptible to Klebsiella pneumoniae endotoxemia (6).
Our experimental design used DNA microarrays to investigate gene expression in response to transplantation injury. The advantage of this approach is the feasibility of readily analyzing large numbers of genes, whereas the limitation is the lack of information on protein expression or function of individual genes. Caution must be exercised when evaluating microarray data. With the large numbers of genes analyzed, false positives are likely to occur. Evaluating genes on fold changes can magnify the importance of genes expressed at very low levels with changes that are relatively small compared with genes expressed at higher levels. Also, calculating the change between normalized gene expression data sets may not include genes with low baseline expression and small but significant changes that may be important in the process of rejection. In this analysis, we separately included aspects of both fold change (dendrograms and SOM) and change between normalized data sets (
z). Our control constitutive expression data were obtained by hybridizing samples onto two GeneChips. Data obtained posttransplantation was hybridized to one GeneChip per experimental condition. Microarray patterns of gene expression in highly differentiated genes were independently verified in repeated transplants in triplicate by real-time PCR. The genes identified in each of our experimental groups using clustering algorithms and SOM are highly consistent with previously reported functions. Thus we believe that these observations do indeed provide a framework to evaluate specific genes and functions in future studies.
A possible limitation to our approach is the inability to distinguish genes expressed by graft tissue and infiltrating cells. We have elected to analyze the graft in total but limited our analysis to genes not expressed in control lymph node. By focusing on genes expressed constitutively in untransplanted heart and following those particular genes posttransplant, we gain insight into the gene modulation of the graft tissue. Genes expressed in the neither subset are a mixture of upregulated graft genes and infiltrating cell genes. Our experimental design does not differentiate inducible gene expression of infiltrating cells from graft cells. Studies in progress will focus on purifying cell populations infiltrating the graft and evaluating differential gene expression of infiltrating cells vs. graft cells during rejection.
Further limiting our gene expression profile study is the inability to measure gene expression in individual graft cell populations such as myocytes, endothelial cells, and fibroblasts during rejection. Cardiac fibroblasts are known to proliferate and synthesize collagen following acute injury and hypoxia-reoxygenation (20). Fibroblasts isolated from rejecting cardiac allografts show increased proliferation in cell culture (12). Although fibroblast activation is more commonly associated with chronic rejection, expression profiles of the fibroblast-enriched myocardial interstitium may be helpful in elucidating peri-transplant cellular responses that precede formation of graft-destructive fibrosis, coronary vasculopathy, and cardiac allograft hypertrophy. In cardiac allograft rejection, myocytes undergo necrosis and apoptosis resulting from cytotoxic T lymphocyte-myocyte interaction. Apoptosis of interstitial and endothelial cells is present in heart allografts and may contribute to the vascular injury associated with allograft rejection. Future studies to elucidate the gene expression profiles from individual cell populations should provide insight into the mechanisms of cardiac myocyte death and vascular injury.
The data presented in our study on the graft response lead us to consider the following hypothesis: Local production of the innate immune collectin C1q subspecies (C1qa, C1qb, C1qc) are important for acute rejection via apoptotic cell clearance. C1q, a member of the classic complement cascade, has the ability to bind C-reactive protein and serum amyloid P component and activate the classic complement pathway (18). Additionally, C1q has been shown to bind to and stimulate ingestion of apoptotic cells in vitro (19) and to enhance apoptotic cell uptake in vivo (25). Further understanding of the role of C1q and apoptosis and transplant rejection may illuminate the mechanisms involved in cardiac myocyte, endothelial, and interstitial cell death and may give clues into therapeutic targets aimed at suppressing allograft rejection.
Current therapeutic strategies to prevent allograft rejection are designed to inhibit the adaptive immune response, in particular, T cell responses. Our analysis of changes in gene expression attributed to the graft tissue in response to transplantation injury provides insights into the graft response to transplantation prior to and during T cell activation.
 |
DISCLOSURES
|
---|
This work was supported by an American Heart Association Established Investigator Award, an Arthritis Foundation Research Award, and National Institutes of Health Grant RO1-AI-44085 (all to D. L. Perkins).
 |
FOOTNOTES
|
---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: D. L. Perkins, Laboratory of Molecular Immunology, Brigham and Womens Hospital, PBB 170 75 Francis St., Boston, MA 02115 (E-mail: dperkins{at}rics.bwh.harvard.edu).
10.1152/physiolgenomics.00139.2002.
 |
References
|
---|
- Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, and Sherlock G. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25: 2529, 2000.[ISI][Medline]
- Ashkar S, Weber GF, Panoutsakopoulou V, Sanchirico ME, Jansson M, Zawaideh S, Rittling SR, Denhardt DT, Glimcher MJ, and Cantor H. Eta-1 (osteopontin): an early component of type-1 (cell-mediated) immunity. Science 287: 860864, 2000.[Abstract/Free Full Text]
- Charles CH, Abler AS, and Lau LF. cDNA sequence of a growth factor-inducible immediate early gene and characterization of its encoded protein. Oncogene 7: 187190, 1992.[ISI][Medline]
- Christopher K, Mueller TF, Ma C, Liang Y, and Perkins DL. Analysis of the innate and adaptive phases of allograft rejection by cluster analysis of transcriptional profiles. J Immunol 169: 522530, 2002.[Abstract/Free Full Text]
- Corry RJ, Winn HJ, and Russell PS. Heart transplantation in congenic strains of mice. Transplant Proc 5: 733735, 1973.[ISI][Medline]
- de Bont N, Netea MG, Demacker PN, Verschueren I, Kullberg BJ, van Dijk KW, van der Meer JW, and Stalenhoef AF. Apolipoprotein E knock-out mice are highly susceptible to endotoxemia and Klebsiella pneumoniae infection. J Lipid Res 40: 680685, 1999.[Abstract/Free Full Text]
- Depre C, Tomlinson JE, Kudej RK, Gaussin V, Thompson E, Kim SJ, Vatner DE, Topper JN, and Vatner SF. Gene program for cardiac cell survival induced by transient ischemia in conscious pigs. Proc Natl Acad Sci USA 98: 93369341, 2001.[Abstract/Free Full Text]
- Eisen MB, Spellman PT, Brown PO, and Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95: 1486314868, 1998.[Abstract/Free Full Text]
- Fryer JP, Leventhal JR, Pao W, Stadler C, Jones M, Walsh T, Zhong R, Zhang Z, Wang H, Goodman DJ, Kurek M, dApice AJ, Blondin B, Ivancic D, Buckingham F, Kaufman D, Abecassis M, Stuart F, and Anderson BE. Synthetic peptides which inhibit the interaction between C1q and immunoglobulin and prolong xenograft survival. Transplantation 70: 828836, 2000.[ISI][Medline]
- Fukui N, Fukuda A, Kojima K, Nakajima K, Oda H, and Nakamura K. Suppression of fibrous adhesion by proteoglycan decorin. J Orthop Res 19: 456462, 2001.[ISI][Medline]
- He H, Stone JR, and Perkins DL. Analysis of robust innate immune response after transplantation in the absence of adaptive immunity. Transplantation 73: 853861, 2002.[ISI][Medline]
- Hellkvist J, Tufveson G, Gerdin B, and Johnsson C. Characterization of fibroblasts from rejecting tissue: the hyaluronan production is increased. Transplantation 74: 16721677, 2002.[ISI][Medline]
- Ideker T, Thorsson V, Ranish JA, Christmas R, Buhler J, Eng JK, Bumgarner R, Goodlett DR, Aebersold R, and Hood L. Integrated genomic and proteomic analyses of a systematically perturbed metabolic network. Science 292: 929934, 2001.[Abstract/Free Full Text]
- Kamradt MC, Chen F, and Cryns VL. The small heat shock protein
B-crystallin negatively regulates cytochrome c- and caspase-8-dependent activation of caspase-3 by inhibiting its autoproteolytic maturation. J Biol Chem 276: 1605916063, 2001.[Abstract/Free Full Text]
- Liu Q and Nilsen-Hamilton M. Identification of a new acute phase protein. J Biol Chem 270: 2256522570, 1995.[Abstract/Free Full Text]
- Marsh JE, Farmer CK, Jurcevic S, Wang Y, Carroll MC, and Sacks SH. The allogeneic T and B cell response is strongly dependent on complement components C3 and C4. Transplantation 72: 13101318, 2001.[ISI][Medline]
- Mistry MJ, Clay MA, Kelly ME, Steiner MA, and Harmony JA. Apolipoprotein E restricts interleukin-dependent T lymphocyte proliferation at the G1A/G1B boundary. Cell Immunol 160: 1423, 1995.[ISI][Medline]
- Nauta AJ, Bottazzi B, Mantovani A, Salvatori G, Kishore U, Schwaeble WJ, Gingras AR, Tzima S, Vivanco F, Egido J, Tijsma O, Hack EC, Daha MR, and Roos A. Biochemical and functional characterization of the interaction between pentraxin 3 and C1q. Eur J Immunol 33: 465473, 2003.[ISI][Medline]
- Ogden CA, deCathelineau A, Hoffmann PR, Bratton D, Ghebrehiwet B, Fadok VA, and Henson PM. C1q and mannose binding lectin engagement of cell surface calreticulin and CD91 initiates macropinocytosis and uptake of apoptotic cells. J Exp Med 194: 781795, 2001.[Abstract/Free Full Text]
- Piacentini L and Karliner JS. Altered gene expression during hypoxia and reoxygenation of the heart. Pharmacol Ther 83: 2137, 1999.[ISI][Medline]
- Roselaar SE and Daugherty A. Apolipoprotein E-deficient mice have impaired innate immune responses to Listeria monocytogenes in vivo. J Lipid Res 39: 17401743, 1998.[Abstract/Free Full Text]
- Sutton MG and Sharpe N. Left ventricular remodeling after myocardial infarction: pathophysiology and therapy. Circulation 101: 29812988, 2000.[Free Full Text]
- Tamayo P, Slonim D, Mesirov J, Zhu Q, Kitareewan S, Dmitrovsky E, Lander ES, and Golub TR. Interpreting patterns of gene expression with self-organizing maps: methods and application to hematopoietic differentiation. Proc Natl Acad Sci USA 96: 29072912, 1999.[Abstract/Free Full Text]
- Theocharis SE, Kanelli H, Margeli AP, Spiliopoulou CA, and Koutselinis AS. Metallothionein and heat shock protein expression during acute liver injury and regeneration in rats. Clin Chem Lab Med 38: 11371140, 2000.[ISI][Medline]
- Vandivier RW, Ogden CA, Fadok VA, Hoffmann PR, Brown KK, Botto M, Walport MJ, Fisher JH, Henson PM, and Greene KE. Role of surfactant proteins A, D, and C1q in the clearance of apoptotic cells in vivo and in vitro: calreticulin and CD91 as a common collectin receptor complex. J Immunol 169: 39783986, 2002.[Abstract/Free Full Text]
- Wang AC and Fu L. Fibronectin regulates the activation of THP-1 cells by TGF-ß1. Inflamm Res 50: 142148, 2001.[ISI][Medline]
- Yang J, Moravec CS, Sussman MA, DiPaola NR, Fu D, Hawthorn L, Mitchell CA, Young JB, Francis GS, McCarthy PM, and Bond M. Decreased SLIM1 expression and increased gelsolin expression in failing human hearts measured by high-density oligonucleotide arrays. Circulation 102: 30463052, 2000.[Abstract/Free Full Text]
Copyright © 2003 by the American Physiological Society.