Renal medullary genes in salt-sensitive hypertension: a chromosomal substitution and cDNA microarray study
Mingyu Liang,
Baozhi Yuan,
Elizabeth Rute,
Andrew S. Greene,
Ai-Ping Zou,
Paulo Soares,
Gregory D. MCQuestion,
Glenn R. Slocum,
Howard J. Jacob and
Allen W. Cowley, Jr.
Department of Physiology, Medical College of Wisconsin, Milwaukee, Wisconsin 53226
 |
ABSTRACT
|
---|
Substitution of chromosome 13 from Brown Norway BN/SsNHsd/Mcw (BN/Mcw) rats into the Dahl salt-sensitive SS/JrHsd/Mcw (SS/Mcw) rats resulted in substantial reduction of blood pressure salt sensitivity in this consomic rat strain designated SSBN13. In the present study, we attempted to identify genes associated with salt-sensitive hypertension by utilizing a custom, known-gene cDNA microarray to compare the mRNA expression profiles in the renal medulla (a tissue playing a pivotal role in long-term blood pressure regulation) of SS/Mcw and SSBN13 rats on either low-salt (0.4% NaCl) or high-salt (4% NaCl, 2 wk) diets. To increase the reliability of microarray data, we designed a four-way comparison experiment incorporating several levels of replication and developed a conservative yet robust data analysis method. Using this approach, from the 1,751 genes examined (representing more than 80% of all currently known rat genes), we identified 80 as being differentially expressed in at least 1 of the 4 comparisons. Substantial agreements were found between the microarray results and the results predicted on the basis of the four-way comparison as well as the results of Northern blots of 20 randomly selected genes. Analysis of the four-way comparison further indicated that
75% of the 80 differentially expressed genes were likely related to salt-sensitive hypertension. Many of these genes had not previously been recognized to be important in hypertension, whereas several genes/pathways known to be involved in hypertension were confirmed. These results should provide an informative source for designing future functional studies in salt-sensitive hypertension.
renal medulla; Dahl S rat; blood pressure; kidney; consomic rats
 |
INTRODUCTION
|
---|
ENHANCED SENSITIVITY OF BLOOD pressure to dietary salt intake is observed in a significant portion of hypertensive patients, especially in the African-American population (20). To elucidate the mechanism for the development and progression of salt-sensitive hypertension, the present study was directed toward determining the pattern of gene expression that underlies the pathophysiology of this disease. Despite the significant progress that has been made in the last few decades, the effort to identify those genes in the past was inevitably compromised by the high degree of genetic heterogeneity and complexity even in the animal models. Moreover, one could only examine the expression of one or a few genes at a time so that determining patterns and differences in gene expression in complex diseases such as salt-sensitive hypertension was cumbersome and inefficient.
To explore the relationship between genes and environment in hypertension, we have studied a colony of inbred Dahl salt-sensitive SS/JrHsd/Mcw (SS/Mcw) and salt-resistant Brown Norway BN/SsNHsd/Mcw (BN/Mcw) rats maintained at the Medical College of Wisconsin (5). Utilizing these inbred strains, we have developed panels of consomic inbred rat strains in which chromosomes from BN/Mcw rats are introgressed, one at a time, into the genetic background of SS/Mcw rats. One of these consomic strains in which chromosome 13 of BN/Mcw was substituted into the background of SS/Mcw (SSBN13) exhibited substantial reduction of blood pressure salt sensitivity. The mean arterial pressure (MAP) of SSBN13 on a high-salt (4% NaCl) diet for 4 wk averaged 119 ± 2 mmHg, compared with 170 ± 3 mmHg in SS/Mcw and 103 ± 1 mmHg in BN/Mcw (4). Comparison of the SS/Mcw and SSBN13 rat strains, therefore, provides a potentially robust model system for identification of genes associated with salt-sensitive hypertension.
The present study was designed to identify genes associated with salt-sensitive hypertension by using a custom, known-gene cDNA microarray to compare the mRNA expression profiles in the renal medulla, a tissue known to play a pivotal role in long-term regulation of blood pressure (2, 3), of SS/Mcw and SSBN13 rats on either low- or high-salt diets. We used a custom microarray containing 1,751 genes representing more than 80% of all currently known rat genes. In addition, a four-way comparison experimental design and a conservative data analysis method were developed, which allowed us to better utilize the power of cDNA microarray while minimizing several of its potential pitfalls. Using this approach, we identified a number of genes/pathways that had not previously been recognized to be important in salt-sensitive hypertension, while several genes/pathways known to be involved in salt-sensitive hypertension were confirmed.
 |
MATERIALS AND METHODS
|
---|
Chromosomal substitution.
Chromosome 13 from the salt-resistant BN/Mcw rat was substituted for chromosome 13 of the Dahl salt-sensitive SS/Mcw rat through selective breeding as described previously (4). The resulting consomic SSBN13 rat, therefore, contained homozygous SS/Mcw alleles on every chromosome, as confirmed by a total genome scan, except chromosome 13, which was homozygous BN/Mcw at all alleles as confirmed by a scan at higher density. The consomic line was maintained by brother-sister mating.
Construction of rat known-gene cDNA microarrays.
We purchased 1,687 clones of rat genes from Research Genetics (Huntsville, AL), and together with 64 rat genes cloned in our department, we amplified these by PCR using appropriate primer sets. The vast majority of these clones represent currently known rat genes that have been assigned defined names and in most cases have some known functions. Each PCR product was analyzed by agarose gel electrophoresis. Approximately 90% of the PCR products appeared as single bands. The PCR products were diluted with 50% DMSO and spotted in duplicate using a 4-pin arrayer (Affymetrix, Santa Clara, CA) on micro-glass slides (Corning Glass Works, Corning, NY) that had been coated with poly-L-lysine (6). Negative controls including DMSO, PCR buffer, PCR buffer with primers, vectors, and Arabidopsis genes were printed in several areas throughout the slide. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and ß-actin were also printed in several areas and used as positive controls. The design of this microarray is shown in Fig. 1. Printed arrays were blocked with succinic anhydride (6) and stored in the dark at room temperature.

View larger version (52K):
[in this window]
[in a new window]
|
Fig. 1. Design of the custom, rat known-gene cDNA microarray. Top left: an overview of an entire microarray after hybridization. Center: enlarged view of one of the four quadrants of the microarray. The four quadrants had similar spot layouts. Top right: a detailed view of one of the spots on the microarray showing the definitions of signal and local background areas that were used in data analysis. The microarray contained 4,608 spots that included 1,751 genes, representing more than 80% of all currently known rat genes, printed in duplicate, and also included hundreds of spots of housekeeping genes and negative controls printed in several areas throughout the microarray.
|
|
Dietary salt intake protocol and tissue harvesting.
Six SS/Mcw rats and six SSBN13 rats were maintained on a low-salt (0.4% NaCl) diet until 1112 wk of age. Three rats of each strain were then switched to a high-salt (4% NaCl) diet for 2 wk, while the other three rats of each strain were maintained on the low-salt diet. On the day of tissue harvest, rats were anesthetized with Inactin (25 mg/kg ip) and ketamine (20 mg/k im). Kidneys were removed quickly and the renal medulla (both papilla and outer medulla) was selectively dissected using a sterile surgical blade and scissors, snap frozen in liquid nitrogen, and stored at -80°C. Tissue was collected between 10:00 AM and noon.
cDNA labeling and microarray hybridization.
Total RNA was isolated from the renal medulla using the TRIzol reagent (Invitrogen/Life Technologies, Carlsbad, CA). Total RNA of 50 µg was reverse-transcribed to cDNA in a reaction primed by 2 µg of oligo-(dT)1218. The reverse transcription reaction contained 1x RT buffer (Invitrogen/Life Technologies), 2.5 mM MgCl2, 10 mM dithiothreitol, 500 µM each of dATP, dGTP, and dCTP, 40 µM dTTP, 40 µM of either Cy3-dUTP or Cy5-dUTP (Amersham Pharmacia, Piscataway, NJ), and 7.5 U/µl SuperScript II RT enzyme (Invitrogen/Life Technologies) in a total volume of 20 µl. The reaction was performed at 39°C for 2 h with an additional 75 U of SuperScript II RT enzyme added at the middle of the reaction. After RNase H digestion, the cDNA products that had been labeled with either the Cy3 or Cy5 fluorescent dye were purified using Centri-Sep spin columns (Princeton Separations, Adelphia, NJ) and QIAquick purification columns (Qiagen, Valencia, CA). Labeled cDNAs were then concentrated using Microcon YM-30 filters (Millipore, Bedford, MA) and hybridized to a microarray at 65°C for 16 h in a hybridization solution containing mouse Cot-1 DNA, 1 µg/µl poly-dA, 1.14 µg/µl yeast tRNA, 3.4x SSC, and 0.3% SDS. Slides were washed and scanned using the ScanArray 5000 [Packard Bioscience (formerly GSI Lumonix), Meriden, CT] to quantify the fluorescent intensity of Cy3 and Cy5 in each spot. Renal medullary gene expression profiles were examined in four comparisons: SSBN13 rats on the low-salt diet compared with SS/Mcw rats on the low-salt diet; SS/Mcw rats on the high-salt diet compared with SS/Mcw rats on the low-salt diet; SSBN13 rats on the high-salt diet compared with SSBN13 rats on the low-salt diet; and SSBN13 rats on the high-salt diet compared with SS/Mcw rats on the high-salt diet. For any two samples (rats) involved in a comparison, one was labeled with Cy3 and the other was labeled with Cy5. The two samples were pooled after labeling and hybridized to a microarray. To control for dye variations, these two samples were labeled again with Cy3 and Cy5 switched between the two samples and hybridized to a second microarray. Thus 4 groups of rats with 3 rats in each group were studied, and 2 hybridizations were performed for each pair of rats, resulting in a total of 24 microarray hybridizations. The schematic of this experimental design is shown in Fig. 2.

View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2. Experimental design. Six microarray hybridizations were performed for each of the four comparisons. As an example, the comparison of SS/Mcw and SSBN13 on the high-salt diet is described in detail on the right.
|
|
Microarray data analysis.
Raw values of fluorescent intensity were extracted from microarray images using the software ImaGene 4.01 (BioDiscovery, Los Angeles, CA). This yielded values for the mean fluorescent intensity and the standard deviation of pixels within the signal and the local background areas (see Fig. 1 for definitions) in the Cy3 and Cy5 channels for each spot. Negative controls in each microarray, a total of 372 spots, were pooled, and the mean of their signal intensity and the standard deviation were calculated for Cy3 and Cy5. A spot was defined as "nondetectable" if its mean signal intensity was less than the mean of the negative control signal intensities plus two times the negative control standard deviation for the same slide. A spot was defined as "low quality" if it was detectable but its mean signal intensity was less than its local background mean intensity plus two times the local background standard deviation. The resulting data set, designated as the "defined raw data," was then analyzed following the scheme shown in Fig. 3 to identify differentially expressed genes. Based on this approach, spots that had low fluorescent intensities in both Cy3 and Cy5 channels as well as spots that had high local background in either or both channels were discarded. These spots would likely have generated misleading ratios and had a disproportionate impact on the overall ratio distribution that served as the basis for several subsequent steps of analysis. Spots that had low intensity in only one channel and had good quality in the detectable channel were kept. Since, however, a low intensity value in a single channel might result in a falsely large ratio, this potential variation was reduced by adding a threshold value to both channels of these spots before the ratio calculation. This may have caused underestimation of the fold changes of some genes, but the process reduced the chance of obtaining unreliably large ratios. Data from one hybridization were normalized by adjusting the mean Ln(ratio) to be zero, based on the assumption that the expression levels of the vast majority of genes did not differ between the two samples being compared. Another important step in our analysis was the averaging of log-transformed, normalized ratios from all hybridized microarrays for each gene in any given comparison and using only these averages to identify differentially expressed genes. Rats were not paired for treatment (dietary salt intake or chromosomal substitution) in the present study, but they were randomly paired for hybridization. This pairing and the subsequent two-color hybridization allowed cohybridization of the two samples being compared on the same microarray and thus minimized the interference generated by array-to-array variations. The ratios obtained from any single pair of rats can be influenced, however, by the manner of pairing. Different sets of ratios would be obtained if the rats were paired differently. The averaged ratios, however, in combination with the previous log transformation, are independent of the way that the rats are paired. It is, therefore, necessary to use the averaged ratios, instead of ratios from individual hybridizations, to identify differentially expressed genes. At the end, a gene was considered differentially expressed between the two groups in a comparison if averaged, log-transformed, and normalized ratio of the gene was beyond mean ± 2 SD of the entire set of ratios from that comparison and if the raw data for that gene passed the selection process and yielded ratios in at least five of the six hybridizations in that comparison (Fig. 3).

View larger version (28K):
[in this window]
[in a new window]
|
Fig. 3. Data analysis process. See METHODS AND MATERIALS for definitions of "nondetectable" and "low quality". Ch, channel (i.e., Cy3 or Cy5); bkg, background; neg, negative control; s.d., standard deviation. *The larger (neg mean + 2 x neg s.d.) value of the two channels was used to adjust both channels.
|
|
Microarray result validation.
For genes identified as differentially expressed in at least one comparison, the original clones obtained from Research Genetics were resequenced to verify their identity. Homologies to the partial sequences obtained were searched using BLAST. The expression of 20 genes randomly selected from the list of genes identified as differentially expressed in at least one comparison in the microarray experiments were further examined by Northern blotting (24). Due to the limited quantity of RNA samples available, the membranes were stripped and rehybridized up to five times.
 |
RESULTS
|
---|
Table 1 summarizes the number of spots in each of the categories defined by our data analysis method described in Fig. 3 and indicates the action applied to them. From the 1,751 genes stamped on each microarray,
1,000 of these genes passed the data selection process and yielded ratios that were used for identification of differential expressions.
Numbers of differentially expressed renal medullary genes in the four comparisons are summarized in Table 2. Names of these genes, together with their clone IDs, locations on the original Research Genetics stock plates, actual ratios, and related pathways/functions are shown in Tables 36, with each table containing genes identified in one of the four comparisons. These genes are grouped under several categories. Overall, 80 genes were differentially expressed in at least one comparison. Note that the categorization and pathways/functions included in Tables 36 were intended strictly as reference only, since a complete description of these genes is beyond the intended scope of this report. Moreover, many of these genes are either not well-characterized in terms of functions or have multiple functions, which is not reflected in the categorization presented in the Tables 36.
View this table:
[in this window]
[in a new window]
|
Table 2. Numbers of differentially expressed genes in renal medulla in response to the substitution of chromosome 13 and/or changes in dietary salt intake
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Renal medullary genes differentially expressed between SS/Mcw and SSBN13 rats on a low-salt (0.4% NaCl) diet
|
|
View this table:
[in this window]
[in a new window]
|
Table 4. Renal medullary genes differentially expressed between SS/Mcw rats on a low-salt (0.4% NaCl) or a high-salt (4% NaCl, 2 wk) diet
|
|
View this table:
[in this window]
[in a new window]
|
Table 5. Renal medullary genes differentially expressed between SSBN13 rats on a low-salt (0.4% NaCl) or a high-salt (4% NaCl, 2 wk) diet
|
|
View this table:
[in this window]
[in a new window]
|
Table 6. Renal medullary genes differentially expressed between SS/Mcw and SSBN13 rats on a high-salt (4% NaCl, 2 wk) diet
|
|
The identity of the Research Genetics clones showing differential expression in at least one comparison was examined by resequencing the original clones. The partial sequences obtained for 56 of the 74 clones resequenced exhibited high degrees of homology with the clone identities provided by Research Genetics. Eight clones did not match the Research Genetics identities, whereas the sequences of another 10 clones were not obtained due to either multiple amplification products or the absence of products. These 18 clones were indicated in the Tables 37 with a question mark in parentheses following their gene names and clone IDs. More extensive sequencing is needed to assign exact identities to these 18 clones.
Because of the involvement of genetic and dietary factors in the present study, the differential expression of these genes might be related to both the substitution of chromosome 13 and changes in dietary salt intake or related to either one of them alone. Because of the substantial reduction in blood pressure salt sensitivity caused by the substitution of chromosome 13, genes related to both the substitution of chromosome 13 and changes in dietary salt intake are more likely associated with salt-sensitive hypertension. These three categories of genes can be roughly separated by comparing the results from the four comparisons. When the genes in Table 3 (SSBN13 vs. SS/Mcw on a low-salt diet) were compared with those in Table 6 (SSBN13 vs. SS/Mcw on a high-salt diet), we found 12 genes that were in both Tables 3 and 6 and were altered toward the same direction in both Tables 3 and 6. These 12 genes were, therefore, differentially expressed between the two strains of rats independent of changes in dietary salt intake. Thus these 12 genes are likely related to the substitution of chromosome 13 alone. Similar inspection of Table 4 (SS/Mcw on a high-salt diet vs. SS/Mcw on a low-salt diet) and Table 5 (SSBN13 on a high-salt diet vs. SSBN13 on a low-salt diet) indicated that seven genes appeared in both Tables 4 and 5 and changed in the same direction in Tables 4 and 5. In other words, these seven genes were altered by changes in dietary salt intake independent of the differences in chromosome 13 between the two strains and were, therefore, likely related to the changes in dietary salt intake alone. The rest of the genes in Tables 36,
60 genes, were related to both the substitution of chromosome 13 and changes in dietary salt intake. These genes were, therefore, more likely associated with salt-sensitive hypertension. Note that the separation of genes described above was a rough estimate. Closer inspection revealed that there were a few exceptions in each category. For example, methylacyl-CoA racemase-
was downregulated by the high-salt diet in both SS/Mcw (Table 4) and SSBN13 (Table 5) and was, therefore, classified as a gene related to changes in dietary salt intake alone. However, this gene appeared to be downregulated to a larger extent in SSBN13 because it was expressed at a lower level when SSBN13 on a high-salt diet was directly compared with SS/Mcw on a high-salt diet (Table 6). Therefore this gene might actually be related to both the substitution of chromosome 13 and changes in dietary salt intake.
The unique four-way comparison design of the present study (Fig. 2) also allowed us to use the results from any three of the four comparisons to predict the result of the fourth comparison. The prediction could be made either qualitatively or quantitatively. Qualitative prediction took advantage of the significant upregulation or downregulation of a particular gene in three comparisons and derived the tendency of change for this gene in the fourth comparison. For example, if gene X was not differentially expressed between SS/Mcw and SSBN13 on the low-salt diet, was significantly upregulated by the high-salt diet in SS/Mcw, and was not changed by the high-salt diet in SSBN13, we could then predict that gene X would be expressed higher in SS/Mcw than in SSBN13 when both were on the high-salt diet. Quantitative prediction used ratios for a gene obtained from three comparisons and calculated the predicted ratio for that gene in the fourth comparison. For example, if gene Y had a ratio of 1.0 for the comparison of SSBN13 over SS/Mcw when both were on the low-salt diet, had a ratio of 1.4 for SS/Mcw on the high-salt diet over SS/Mcw on the low-salt diet, and had a ratio of 0.7 for SSBN13 on the high-salt diet over SSBN13 on the low-salt diet, we could then predict that the ratio of gene Y for SSBN13 on the high-salt diet over SS/Mcw on the high-salt diet would be 0.5. When the predicted results were compared with the actual experimental results, we found that, of the 118 differential expressions of 80 genes, 74 (63%) were confirmed by at least one way of prediction, and 31 (26%) were confirmed by both ways of prediction. The substantial agreement between actual experimental results and the derived results indicated a high level of consistency in our microarray data.
As another way to test the reliability of our microarray method, we used a random number generator to randomly select 20 genes from the genes identified as differentially expressed by microarray and performed conventional Northern blot analysis to examine their expression levels in the renal medulla of the four groups of rats used for microarray experiments. The actual ratios for these genes in each comparison obtained from both microarray and Northern blot analyses are shown in Table 7. According to a Spearman rank-order analysis, a highly significant correlation coefficient of 0.447 (P < 0.001) was found between the Ln(ratio)s obtained from the two methods. Of the 40 ratios that exceeded the threshold for the microarray analysis, all but 5 were directionally confirmed by Northern blot analysis. The large majority of genes whose ratios were not significantly changed based on the threshold criteria for the microarray analysis, also had ratios that were close to one according to Northern blot analysis.
We also attempted to locate the differentially expressed genes on rat chromosomes. Of the 80 differentially expressed genes, 61 have their identities confirmed, including 56 Research Genetics clones, sequence-verified by the authors, and 5 in-house clones. We were able to retrieve chromosome locations for 30 of the 61 genes from the Rat Genome Database maintained at the Medical College of Wisconsin (http://rgd.mcw.edu/). Chromosome locations for another 16 genes were derived from human and/or mouse genomes by using the virtual comparative mapping tool also located on the RGD web site. Of the 46 mapped genes, one, kynurenine 3-hydroxylase, was found to be located on chromosome 13. The other 45 genes are widely distributed on many chromosomes.
 |
DISCUSSION
|
---|
Assessment of microarray results.
cDNA microarray is a powerful technique that has been primarily used for high-throughput gene expression profiling (7, 17, 18). However, the application of this technique has caused substantial concerns for two main reasons: 1) the lack of necessary replicates (11) and 2) the lack of extensive assessment of raw data reliability (22). These flaws often lead to unreliable or uninterpretable results, especially a large percentage of false positives.
To address these problems, we first developed an experimental design that incorporated several levels of replication. In addition to printing each gene in duplicate on the microarray (Fig. 1), we included three individual rats in each group and the hybridization was repeated for each pair of rats with dye switching (see METHODS AND MATERIALS, and Fig. 2). At the end, we had six hybridized microarrays representing three pairs of rats for each of the four comparisons.
We then developed a systematic, conservative approach to perform data analysis (Fig. 3). To be included in the ratio calculation, a spot had to have a signal fluorescent intensity that was significantly above the negative controls and the local background, as defined in METHODS AND MATERIALS. Only about 60% of the spots passed this selection process. The adjustment applied to the 1% of spots (Fig. 3 and Table 1) also reduced the chance of obtaining falsely large ratios, while allowing us to keep these potentially important spots because they might represent induction of a gene not initially expressed or complete quenching of the expression of a gene. We used two times the standard deviation as the threshold for differential expression. This was based on the assumption that the expression levels of the vast majority of genes did not differ between the two groups being compared. This threshold is data set dependent and less arbitrary than a fixed fold change threshold. Furthermore, a gene that had a ratio beyond the threshold had to meet another criterion to be considered differentially expressed, i.e., the raw data for that gene should pass the selection process in at least five of the six hybridizations in that comparison.
In addition to the confidence gained by incorporating replicates and employing a conservative data analysis method, the four-way comparison design allowed use of the results from any three comparisons to predict the result of the fourth comparison. The substantial overlapping between the predicted results and the actual experimental results indicated a high level of consistency in our microarray data and further increased our level of confidence on these data.
Finally, the validity of our microarray method was supported by Northern blot analysis. Northern blot has been used by many studies to validate results from microarrays (14, 19). Since the genes used for Northern blot analysis in the present study were randomly selected from the list of genes identified by the microarray, the high degree of agreement between the results of Northern blot and microarray not only provided a level of validation for the changes of those specific genes but also strongly supported the reliability of all the significant changes identified by microarrays and indicated a low false-positive rate in our microarray results.
Identification of differentially expressed renal medullary genes.
The genes identified in the present study included those related to several pathways that are known to be associated with salt-sensitive hypertension. Examples of these genes included the cytochrome P-450 enzyme family (16) and collagen-related genes. Several collagen-related genes were substantially upregulated by the high-salt diet in SS/Mcw but not in SSBN13 (Tables 4 and 5). This is consistent with the well-known accumulation of collagen in tissues of hypertensive subjects, including the kidney tissues of Dahl salt-sensitive rats on a high-salt diet (13, 21). It is interesting to note that some genes that are often found to be involved in hypertension, such as renin and nitric oxide synthases, were not identified as differentially expressed in the present study. Closer inspection of the data indicated that the expression levels of many of these genes in the renal medulla, which was the tissue examined in this study, were below the conservative detection threshold and, therefore, were discarded before ratio calculation.
Importantly, many genes that had not been recognized before to be associated with salt-sensitive hypertension were identified in the present study. Deoxyribonuclease I (DNase I) is an interesting example. DNase I is a major enzyme involved in DNA hydrolysis. As shown in the present study (Table 7), the renal medullary expression of this gene was not significantly different between SS/Mcw and SSBN13 on the low-salt diet. However, the exposure to the high-salt diet significantly downregulated the expression of this gene in SS/Mcw but not in SSBN13. Recent studies in both mice (12) and humans (23) have indicated an important role of the deficiency of DNase I in the pathogenesis of systemic lupus erythematosus presumably due to the inefficient clearance of nuclear DNA-protein complexes after cell death. It would be interesting to determine whether a similar pathophysiological process also occurs in salt-sensitive hypertension and, if so, whether it plays any role in the progression of this disease. Another example is kidney-specific protein. Kidney-specific protein is encoded by a rat gene specifically expressed in the kidney (8). The function of this protein is unknown. The present study (Table 7) showed that this gene was expressed at a lower level in the renal medulla of SSBN13 compared with SS/Mcw on the low-salt diet. When the rats were exposed to the high-salt diet, the expression of this gene was further decreased in SSBN13 but did not change significantly in SS/Mcw. Interestingly, the predicted peptide sequence of this protein had
70% similarity to the product of SA gene (8) whose renal expression was substantially higher in spontaneously hypertensive rats compared with Wistar-Kyoto rats (10), although the involvement of SA gene in hypertension was not supported by congenic rat studies (9). The distinct regulation of the kidney-specific protein in SSBN13 and SS/Mcw rats strongly suggests the need for further studies on the function of this protein and its role in hypertension.
Taken together, the mRNA expression profiles established in the present study indicate several possible phenotypic features in the renal medulla of SS/Mcw and SSBN13 rats that may be related to salt-sensitive hypertension. For example, a number of genes related to extracellular matrix/fibrosis, cellular stress response, or apoptosis were altered by the transfer of the chromosome 13 and the changes of dietary salt intake (Tables 36). The changes of those genes generally indicated higher fibrotic activity, higher stress levels, and higher apoptotic activity in the renal medulla of SS/Mcw compared with SSBN13 when exposed to the high-salt diet. In addition, a large number of genes related to pathways such as arachidonic acid metabolism, energy metabolism and/or fatty acid oxidation, and so on, were altered. The pattern of changes of these genes, however, was more complicated, and the net effect on tissue functions and phenotype was less straightforward. Clearly, much more work needs to be done to interpret and integrate these results.
It is widely accepted that hypertension is a complex disease often determined by both genetic and nongenetic factors and that multiple genes are involved in the development and progression of this disease (1, 15). Indeed, quantitative trait loci for blood pressure have been located on many rat chromosomes (15). The multigene nature of hypertension, in particular, of salt-sensitive hypertension, is supported by the present study. As described in the RESULTS, of the 1,751 genes examined,
60 exhibited changes of expression in association with both the substitution of chromosome 13 and changes in dietary salt intake. These results indicate that there are likely multiple known rat genes in the renal medulla that are associated with salt-sensitive hypertension, although at the current stage it is difficult to clearly distinguish genes causing salt-sensitive hypertension from those that exhibited changes in expression in response to hypertension.
Equally evident from these results is the extensive interchromosomal gene-gene interactions involved in salt-sensitive hypertension. The genetic backgrounds of SS/Mcw and SSBN13 were identical except for the chromosome 13. As discussed above, a few dozen genes were differentially expressed in one or more of the comparisons because of the difference in the chromosome 13. As shown in the RESULTS, these genes are widely distributed on many rat chromosomes. Of the 46 genes that we were able to map, only one, kynurenine 3-hydroxylase, was mapped to chromosome 13. Kynurenine 3-hydroxylase is a mitochondrial membrane enzyme involved in the synthesis of NAD/NADP from tryptophan. It is noteworthy that relatively few known genes have been mapped to chromosome 13. Of the 1,751 genes included in our microarray, only 13 have been mapped to the chromosome 13. Therefore, it is likely that some of the differentially expressed genes that we were not able to map, mostly due to uncertain identities, might be actually located on chromosome 13. Moreover, the known rat genes, which were the majority of genes examined in this study, constitute only a small portion of the entire set of rat genes. Furthermore, it is possible that the regulation of some genes on the chromosome 13 did not occur at the level of steady-state mRNA content and, therefore, could not be identified by the approach used in the present study. It is also possible that differential expression of genes on the chromosome 13 in tissues outside the renal medulla affected the expression of genes on other chromosomes in the renal medulla. Taken together, it is likely that the difference of chromosome 13 between SS/Mcw and SSBN13 can, through extensive interchromosomal gene-gene interactions, result in differential expression of genes on many chromosomes. Collectively, this may contribute to the changes in phenotypes caused by the substitution of chromosome 13, such as the reduction of blood pressure salt sensitivity.
Perspectives
It is self-evident that the regulation of steady-state mRNA levels, which was the parameter assessed in the present study, is by no means the only mechanism that the body invokes to cope with genetic or environmental alterations. It is not the intention of the present study to obtain a complete picture of molecular events occurring in the renal medulla in salt-sensitive hypertension. Yet the information generated by the present study is undoubtedly vast in quantity and highly thought-provoking in nature. Although the consistency with the current understanding of hypertension provides another level of confirmation of the validity of the novel approaches taken by the present study, the identification of dozens of genes by a single study that had not previously been recognized to be associated with hypertension constitutes the most fascinating part of the present work. Further studies on the functional roles of these genes in hypertension hold the promise of discovering novel pathways critical to this disease and hopefully the design of new therapeutic approaches. Equally exciting and important is the integration of this large collection of differentially expressed genes in the context of organ function and disease process, a formidable task that can only be attempted after we have a better understanding of the functional roles of all these genes. In addition to the lack of a full understanding of the functional roles of these genes, the interpretation of the results of the present study at this point of time is also limited by several other factors such as the relatively small number of currently known rat genes, the small number of genes that have been mapped to chromosome 13, the lack of comparable studies in other strains of rats, and so on. A more complete understanding and appreciation of these results should be possible a few years from now when these limitations will likely have been at least partially overcome.
In summary, we have utilized a combination of chromosomal substitution and custom known-gene cDNA microarray techniques to identify rat renal medullary genes associated with salt-sensitive hypertension. Using a unique four-way comparison experimental design and a conservative data analysis process, a total of 80 genes, including many genes that were not previously known to be involved in hypertension, were identified to be differentially expressed in at least one of the comparisons. Our data supported the multigene nature of salt-sensitive hypertension, indicated extensive interchromosomal gene-gene interactions in this disease, and should provide an informative source for designing future functional studies.
 |
ACKNOWLEDGMENTS
|
---|
We gratefully acknowledge Dr. Michael Olivier and Regina Cole for assistance with clone resequencing, Randy Berdan and SPS Productions for assistance with computer program development for compiling gene identification files, and Dr. Richard J. Roman, Dr. David P. Basile, and Meredith Skelton for critical review of the manuscript.
This study was supported by National Institutes of Health, National Heart, Lung, and Blood Institute Grants HL-66579, HL-54998, and HL-29587.
 |
FOOTNOTES
|
---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: M. Liang, Dept. of Physiology, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226 (E-mail: mliang{at}mcw.edu).
10.1152/physiolgenomics.00083.2001.
 |
References
|
---|
-
Cowley AW Jr. Genetic and nongenetic determinants of salt sensitivity and blood pressure. Am J Clin Nutr 65, Suppl 2: 587S593S, 1997.
-
Cowley AW Jr. Role of the renal medulla in volume and arterial pressure regulation. Am J Physiol Regulatory Integrative Comp Physiol 273: R1R15, 1997.[Abstract/Free Full Text]
-
Cowley AW Jr and Roman RJ. The role of the kidney in hypertension. JAMA 275: 15811589, 1996.[ISI][Medline]
-
Cowley AW Jr, Roman RJ, Kaldunski ML, Dumas P, Dickhout JG, Greene AS, and Jacob HJ. Brown Norway chromosome 13 confers protection from high salt to consomic Dahl S rat. Hypertension 37: 456461, 2001.[Abstract/Free Full Text]
-
Cowley AW Jr, Stoll M, Greene AS, Kaldunski ML, Roman RJ, Tonellato PJ, Schork NJ, Dumas P, and Jacob HJ. Genetically defined risk of salt sensitivity in an intercross of Brown Norway and Dahl S rats. Physiol Genomics 2: 107115, 2000.[Abstract/Free Full Text]
-
Kita Y, Shiozawa M, Jin W, Majewski RR, Besharse JC, Greene AS, and Jacob HJ. Implications of circadian gene expression in kidney, liver, and the effects of fasting on pharmacogenomic studies. Pharmacogenetics 12: 5565, 2002.[ISI][Medline]
-
Kurella M, Hsiao LL, Yoshida T, Randall JD, Chow G, Sarang SS, Jensen RV, and Gullans SR. DNA microarray analysis of complex biologic processes. J Am Soc Nephrol 12: 10721078, 2001.[Abstract/Free Full Text]
-
Hilgers KF, Nagaraj SK, Karginova EA, Kazakova IG, Chevalier RL, Carey RM, Pentz ES, and Gomez RA. Molecular cloning of KS, a novel rat gene expressed exclusively in the kidney. Kidney Int 54: 14441454, 1998.[ISI][Medline]
-
Hubner N, Lee YA, Lindpaintner K, Ganten D, and Kreutz R. Congenic substitution mapping excludes Sa as a candidate gene locus for a blood pressure quantitative trait locus on rat chromosome 1. Hypertension 34: 643648, 1999.[Abstract/Free Full Text]
-
Iwai N and Inagami T. Isolation of preferentially expressed genes in the kidneys of hypertensive rats. Hypertension 17: 161169, 1991.[Abstract]
-
Lee ML, Kuo FC, Whitmore GA, and Sklar J. Importance of replication in microarray gene expression studies: statistical methods and evidence from repetitive cDNA hybridizations. Proc Natl Acad Sci USA 97: 98349839, 2000.[Abstract/Free Full Text]
-
Napirei M, Karsunky H, Zevnik B, Stephan H, Mannherz HG, and Moroy T. Features of systemic lupus erythematosus in Dnase1-deficient mice. Nat Genet 25: 135136, 2000.[ISI][Medline]
-
Otsuka F, Yamauchi T, Kataoka H, Mimura Y, Ogura T, and Makino H. Effects of chronic inhibition of ACE and AT1 receptors on glomerular injury in Dahl salt-sensitive rats. Am J Physiol Regulatory Integrative Comp Physiol 274: R1797R1806, 1998.[Abstract/Free Full Text]
-
Pendurthi UR, Allen KE, Ezban M, and Rao LV. Factor VIIa and thrombin induce the expression of Cyr61 and connective tissue growth factor, extracellular matrix signaling proteins that could act as possible downstream mediators in factor VIIa x tissue factor-induced signal transduction. J Biol Chem 275: 1463214641, 2000.[Abstract/Free Full Text]
-
Rapp JP. Genetic analysis of inherited hypertension in the rat. Physiol Rev 80: 135172, 2000.[Abstract/Free Full Text]
-
Roman RJ. P-450 metabolites of arachidonic acid in the control of cardiovascular function. Physiol Rev 82: 131185, 2002; 10.1152/physrev.00021.2001.[Abstract/Free Full Text]
-
Schena M, Shalon D, Davis RW, and Brown PO. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270: 467470, 1995.[Abstract]
-
Schena M, Shalon D, Heller R, Chai A, Brown PO, and Davis RW. Parallel human genome analysis: microarray-based expression monitoring of 1000 genes. Proc Natl Acad Sci USA 93: 1061410619, 1996.[Abstract/Free Full Text]
-
Sudarsanam P, Iyer VR, Brown PO, and Winston F. Whole-genome expression analysis of snf/swi mutants of Saccharomyces cerevisiae. Proc Natl Acad Sci USA 97: 33643369, 2000.[Abstract/Free Full Text]
-
Sullivan JM, Prewitt RL, and Ratts TE. Sodium sensitivity in normotensive and borderline hypertensive humans. Am J Med Sci 295: 370377, 1988.[ISI][Medline]
-
Tamaki K, Okuda S, Nakayama M, Yanagida T, and Fujishima M. Transforming growth factor-ß1 in hypertensive renal injury in Dahl salt-sensitive rats. J Am Soc Nephrol 7: 25782589, 1996.[Abstract]
-
Yang MCK, Ruan QG, Yang JJ, Eckenrode S, Wu S, McIndoe RA, and She JX. A statistical method for flagging weak spots improves normalization and ratio estimates in microarray. Physiol Genomics 7: 4553, 2001. First published August 8, 2001; 10.1152/physiolgenomics.00020.2001.[Abstract/Free Full Text]
-
Yasutomo K, Horiuchi T, Kagami S, Tsukamoto H, Hashimura C, Urushihara M, Kuroda Y. Mutation of DNASE1 in people with systemic lupus erythematosus. Nat Genet 28: 313314, 2001.[ISI][Medline]
-
Yuan B and Cowley W Jr. Evidence that reduced renal medullary nitric oxide synthase activity of Dahl S rats enables small elevations of arginine vasopressin to produce sustained hypertension. Hypertension 37: 524528, 2001.[Abstract/Free Full Text]