Orthopaedic Research Laboratory, Carolinas Medical Center, Charlotte, North Carolina 28232-2861
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Hyp; Phex; microarray; Affymetrix
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Two experimental protocols are available to study this system. Low-phosphate diets will stimulate the homeostatic mechanisms for phosphate conservation (16, 26). In addition, loss-of-function mutations of the Phex gene will block this adaptation and suppress phosphate conservation (26). Mutations of this gene are known to occur in human patients with X-linked hypophosphatemia (33) and in the Hyp and Gy mutations of mice (8, 13, 37).
In this project, mRNA gene expression in the mouse kidney was studied by DNA microarray technology. We hypothesized that genes participating in the renal response to changing dietary phosphate intake would respond to stimulation by a low-phosphate diet and/or inhibition by the Hyp mutation.1
![]() |
Materials And Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The animals were then anesthetized, one capillary tube (70 µl) of blood was collected from the orbital sinus, and both kidneys were harvested. The kidneys were frozen in liquid nitrogen and stored at 75°C. The kidneys were then weighed and homogenized, and total RNA was extracted with TRIzol (GIBCO-BRL; Invitrogen, Gaithersburg, MD) (15). Plasma inorganic phosphate was measured by the method of Chen et al. (5).
Experimental design.
Equal amounts of RNA from three mice, matched for genotype, sex, diet, and time on the diet were pooled to create each sample for microarray analysis. Four treatment groups were done: 1) Normal mice fed the control diet, 2) normal mice fed the low-phosphate diet, 3) Hyp mice fed the control diet, or 4) Hyp mice fed the low-phosphate diet. Five replicates were done with each replicate containing one sample from each of the four treatment groups for a total of 20 independent samples (60 mice total). Each replicate was matched for littermates, sex (3 replicates of male mice and 2 of female mice), time on the diet (3 replicates at 5 days and 2 at 3 days), and parallel processing.
Microarray analysis.
Samples were processed as described in the Affymetrix GeneChip Expression Analysis Technical Manual (Affymetrix, Santa Clara, CA; Rev. 1, part no. 701021, http://www.affymetrix.com). The sample preparation is described here in brief. Samples with 30 µg RNA were purified on RNeasy columns from Qiagen (Valencia, CA; product no. 74104) and then converted to double-stranded cDNA with a SuperScript double-stranded cDNA synthesis kit (product no. 11917-010; Invitrogen, Carlsbad, CA). The cDNA was then expressed as biotin-labeled cRNA by in vitro transcription (IVT) with the Enzo RNA transcript labeling kit (Affymetrix, product no. 900182). Each sample was spiked with bioB, bioC, bioD, and cre (Affymetrix, product no. 900299). The biotin-labeled cRNA was fragmented nonenzymatically. The fragmented cRNA from the 20 independent samples was hybridized to 20 mouse U74Av2 microarrays (Affymetrix, product no. 900343) in Affymetrix hybridization buffer for 16 h at 45°C. The hybridized arrays were washed and stained in the Affymetrix Fluidics Station 400 to attach fluorescent labels to the biotin, followed by biotin-labeled antibody, and then a second staining with fluorescent labeling of the biotin. Each array was scanned twice by the Agilent GeneArray scanner G2500A (Agilent Technologies, Palo Alto, CA).
Data analysis.
The data were analyzed with Affymetrix Microarray Suite 5.0 and Affymetrix Data Mining Tool 3.0 software. Microarray Suite was used to scale the mRNA expression (signal value) of all genes to an average of 500 for each array. Data for all genes for the 20 samples were transferred to an Excel spreadsheet (Microsoft, Redmond, WA). The data were paired by replicates. Paired t-tests were then computed to test for an effect of the Hyp mutation by comparing normal mice to Hyp mice fed the control diet (Table 1). The effect of the low-phosphate diet was tested by comparing mice fed the control diet to mice fed the low-phosphate diet for mice of the same genotype (Tables 2 and 3). The data were expressed as means ± SE for the five replicates.
|
|
|
|
|
|
|
|
GenMAPP.
The "Hs-Mm-Rn-Affy_database" (http://www.GenMAPP.org; Conklin Lab, J. David Gladstone Institute, University of California at San Francisco, San Francisco, CA) was used to transform the spreadsheet described above into a GenMAPP compatible format. Results from the paired t-tests and the analysis of variance were submitted to the GenMAPP software (Gene Microarray Pathway Profiler, version 1.0, http://www.GenMAPP.org). MAPPFinder (version 1.0, http://www.GenMAPP.org) was then used to identify metabolic pathways most affected by treatment.
MGED.
This report conforms to the MIAME standards of MGED (http://www.mged.org). A copy of the full microarray data set has been deposited in the NCBI Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo) as series GSE868.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Plasma inorganic phosphate at the time of euthanasia was decreased in the low-phosphate diet groups for both the normal mice [3.08 ± 0.11 (15) vs. 1.01 ± 0.05 (15) mM, control diet vs. low-phosphate diet, P < 0.001] and for the Hyp mice [1.51 ± 0.05 (15) vs. 0.24 ± 0.03 (15) mM, P < 0.001]. Hyp mice remained hypophosphatemic relative to the normal mice when fed the control diet [3.08 ± 0.11 (15) vs. 1.51 ± 0.05 (15) mM, P < 0.001] or the low-phosphate diet [1.01 ± 0.05 (15) vs. 0.24 ± 0.03 (15) mM, P < 0.001].
Microarray.
The Affymetrix U74Av2 array has probe sets for 12,473 genes. Of this total, an average of 5,719 genes were scored as present in the kidney samples. Results of the microarray analysis were examined for changes known to occur in the kidney. mRNA for the renal tubular sodium phosphate cotransporter, Npt2 (SLC34a1), was significantly depressed in the Hyp samples (Fig. 1, top left) as was the mRNA level for Npt1 (SLC17a1; Fig. 1, top right). There was a nonsignificant increase in the transcript level for Npt2 in normal mice fed the low-phosphate diet (Fig. 1, top left). mRNA for the vitamin D 1-hydroxylase was significantly increased in normal mice fed the low-phosphate diet and significantly decreased in Hyp mice fed the low-phosphate diet (Fig. 1, middle left). The vitamin D 24-hydroxylase was significantly upregulated in the Hyp samples (Fig. 1, middle right). mRNA for the two vitamin D-dependent proteins, calbindin D9K and calbindin D28K, are shown in Fig. 1 (bottom). Both were depressed in Hyp mice on the low-phosphate diet. mRNA for calbindin D9K was not significantly increased in normal mice fed the low-phosphate diet.
Normal vs. Hyp.
For mice fed the control diet, 73 genes were increased in mRNA expression in the Hyp mice compared with normal mice and 64 decreased at P < 0.01 for a total of 137 genes in the kidney affected by the Hyp mutation. These 137 significantly exceeded the randomly expected 57 genes (P < 0.001). The genes with the strongest response to the Hyp mutation (P < 0.001) are shown in Table 1, and all genes significant at P < 0.01 are shown in Supplemental Table 5. Four examples with prominent changes are shown in Fig. 2. Most of these genes did not respond to low-phosphate diet. Of the 137 genes responding to the Hyp mutation, only 13 responded to low-phosphate diet in the normal mice.
Response to low-phosphate diet.
Low-phosphate diet caused an altered mRNA expression in 267 genes in the normal mice (127 increased and 140 decreased at P < 0.01). This was significantly greater than the 57 genes expected to change at random (P < 0.001). The strongest responders to low-phosphate diet (those significant at P < 0.001) are shown in Table 2, and all significant responders at P < 0.01 are shown in Supplemental Table 6. Typical responses were for low-phosphate diet to have a significant effect in normal mice with no significant change in the Hyp mice. Four examples of this are shown in Fig. 3.
|
The Hyp mutation decreases phosphate conservation in the kidney, whereas low-phosphate diet stimulates phosphate conservation. Thus we hypothesized that some genes would be increased by low-phosphate diet in normal mice but decreased in Hyp mice on the control diet. (Or the converse may occur: decreased by low-phosphate diets in normal mice but increased by the Hyp mutation). This only happened for four genes (3 decreased and 1 increased in normal mice on low-phosphate diet), which was at the level of a random event. Transthyretin (Fig. 2) and deoxyribonuclease I (Fig. 4) are examples of this.
Genotype-by-diet interaction.
To further search for genes differentially affected by low-phosphate diet in normal and Hyp mice, the genotype-by-diet interaction statistic was calculated by analysis of variance. This analysis used all four treatment group means. It tested whether the response to low-phosphate diet in the Hyp mice differed significantly from the response of normal mice. The numerical result was an F test for genotype-by-diet interaction. The strongest interactions (those at P < 0.001) are shown in Table 4, and all significant responders at P < 0.01 are shown in Supplemental Table 8.
To elucidate the biological significance of these changes, the significant interactions (P < 0.05) were submitted to the GenMAPP software. MAPPFinder was used to identify specific metabolic pathways affected by low-phosphate diet and the Hyp mutation. MAPPFinder identified the map "Mm_microtubule-based process." Most - and ß-tubulins and most kinesin genes were affected by genotype, diet, and/or their interaction (Fig. 5). These genes responded to low-phosphate diet in normal mice but not in Hyp mice.
The GenMAPP "Mm_antioxidant" pointed out changes in mRNA levels for genes related to antioxidant activity. Expression levels for glutathione reductase 1 (AI851983, Supplemental Table 8), peroxiredoxin 2 (U20611, Supplemental Table 8), and sequestosome 1 (Fig. 3) were increased by low-phosphate diet in normal mice, but not in Hyp mice.
In addition, two of the four genes presented on the GenMAPP "Mm_ligase, forming carbon-sulfur bonds," had significant genotype-by-diet interaction terms: U15977 fatty acid coenzyme A ligase, long chain 2 (P < 0.005, Supplemental Table 8) and AK003867 succinate-coenzyme A ligase, ADP-forming, ß-subunit (P < 0.05). This suggested a change in fatty acid metabolism. There were also significant effects of sex on the mRNA gene expression in the kidney that will be reported elsewhere.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Data were found with the microarray that agreed with previously published studies of individual genes. mRNA expression was found for both Npt1 and Npt2. Npt2 transcript levels increased nonsignificantly in normal mice fed the low-phosphate diet; significant increases have been noted before (27). mRNA levels for both phosphate transporters were decreased in the Hyp mice. It is the low levels of the phosphate transporters that cause the low rate of phosphate reabsorption in the kidney (42). Transcript levels for vitamin D 1-hydroxylase were upregulated by low-phosphate diet in normal mice, but downregulated in Hyp mice (16). The transcript levels for vitamin D 24-hydroxylase are known to be increased in Hyp mice (39). mRNA for two vitamin D-dependent calcium-binding proteins were found in the kidney. We have shown previously that the calbindin D9K is stimulated by treatment with 1,25-dihydroxyvitamin D3 in the kidneys of both normal and Hyp mice, whereas the renal calbindin D28K does not respond to exogenous 1,25-dihydroxyvitamin D3 in either genotype (3). In this project, transcript levels for both calbindins were downregulated by low-phosphate diet in the Hyp mice in concordance with the reduced vitamin D 1
-hydroxylase message. Search of the other genes on the array failed to find any other gene that showed this biphasic response to low-phosphate diet. Thus there were only two vitamin D-dependent proteins in the kidney among the genes studied in this project.
The Hyp mutation abolished the renal adaptation to the low-phosphate diet for many genes. Of the 267 genes that had a significant response in normal mice, only 7 of these genes were affected by low-phosphate diet in the Hyp mice at P < 0.01. It has been known for some time that the Hyp and Gy mutations of the Phex gene block both the stimulation of renal tubular reabsorption of phosphate (13, 26, 43) and the increased secretion of 1,25-dihydroxyvitamin D3 (16, 41). Although there continues to be increased Npt2 protein on the renal brush-border membrane and increased brush-border transport of phosphate in the mutant mice (40, 41), this is not evident in the whole kidney as significantly increased phosphate reabsorption (26, 43) in mutant mice fed a low-phosphate diet. This adaptation of Npt2 must be a postgenomic change since no significant change in mRNA for this transporter was found in either the normal or the Hyp mice fed the low-phosphate diet. Such postgenomic changes are known to occur with phosphate transport (27). The possibility of time effects have not been explored, and it is possible that the lack of significant adaptation of Npt2 transcript levels could be explained by an earlier response than that measured in this study.
The Hyp mice have a mutation in the gene Phex (37). This gene belongs to a family of membrane-bound neutral endopeptidases (45). The normal physiological role for this gene is not clear, nor is the mechanism by which it inhibits this renal adaptation. It is presumed that Phex acts to activate or inactivate a humoral factor affecting the renal reabsorption of phosphate (32). There are elevated levels of FGF23 in Hyp mice, in human patients with X-linked hypophosphatemia, and in other phosphaturic syndromes (32, 46). Injections of FGF23 will acutely affect renal phosphate conservation (32). Whether this is the only phosphaturic factor present in this disease is unclear. Unfortunately, there were no probe sets for FGF23 on the microarray used for this project.
GenMAPP software was useful in locating specific metabolic pathways affected by adaptation to changing dietary phosphate levels and the Hyp mutation. The statistical interaction between diet and genotype from the analysis of variance table was a sensitive marker for genes that responded to the low-phosphate diet in normal mice and whose response was affected by the Hyp mutation. The most prominent pathway identified was that of microtubules with most - and ß-tubulins and most kinesins affected by treatment. Microtubules are known to be involved with the trafficking of transport molecules onto and off the cell membrane (12), including renal phosphate transporters (12), but not in all cases (1).
Several specific genes and gene families are worthy of comment.
1) Mutations of klotho create a model of accelerated aging (21). These mutations elevate renal vitamin D 1-hydroxylase (47). Interestingly, low-phosphate diets block many of the effects of the mutation of the klotho gene (24). The mRNA levels for klotho were altered in the Hyp mice (Fig. 2).
2) Carbonic anhydrase XIV (Fig. 2) is involved in renal tubular acidification (23). Levels of the mRNA for this enzyme were decreased in the Hyp mice.
3) The deoxyribonuclease I shown in Fig. 4 is related to microfilament function. This enzyme binds G-actin and blocks actin polymerization (4, 7). The low level of mRNA for this enzyme in the Hyp mice may affect microfilament formation and alter the functioning of the renal tubules.
4) One member of family 1 of the serine protease inhibitors is shown in Fig. 2. All six members of this family, all separate genes, have elevated transcript levels in the kidneys of Hyp mice. In general, these family members act to inhibit elastase (22). Neutrophil elastase elevates epithelial permeability which is blocked by serine protease inhibitor 11 (22, 31). Serine protease inhibitors are also known as 1-antitrypsin family 1, and serine protease inhibitor 14 is also known as kallistatin (35).
5) Transthyretin mRNA is greatly elevated in the Hyp mice (Fig. 2), and it has a significant genotype x diet interaction (Table 4). It is the major thyroxin-binding protein in the plasma of rodents (30). It is found in the glomerular peripolar (9) and proximal tubule (44) cells. Transthyretin can act to deliver both retinol and thyroxin to the tissues (34).
6) Transcript levels for collagen types I, III, V, VI, XIV, and XV were decreased by low-phosphate diet in the kidney of normal mice. This suggests a decrease in extracellular matrix synthesis as part of the adaptation to reduced extracellular phosphate levels. Type III collagen (Fig. 3) is found in the blood vessels of the kidney (11, 25).
7) Antioxidant transcript levels were found to be increased with low-phosphate diet for glutathione reductase 1, peroxiredoxin 2, and sequestosome 1 in normal mice. Glutathione reductase 1 is part of the intracellular defense systems that reduce the damage to the cells from reactive oxygen species (ROS) (20). Peroxiredoxin 2 is also an antioxidant protein that is found in cells such as renal tubular cells that have high levels of oxygen metabolism (10). Sequestosome 1, also known as the protein p62, is known to respond to oxidative stress (38). Klotho may also be involved in preventing damage from ROS since the potent antioxidant -tocopherol will prevent the cognition impairment and lipid peroxide accumulation, which occur in klotho-deficient mice (28).
In this project, renal adaptation to low-phosphate diet was studied by measuring mRNA levels of many genes expressed in the kidney. Changes in mRNA levels may not correlate perfectly to changes in protein level for that gene. Not all transcripts undergo translation to a peptide sequence, and posttranslational modifications alter levels of specific proteins independent of the transcript level.
In summary, low-phosphate diet caused changes in the mRNA expression level of specific genes in the kidneys of normal mice. These changes were abolished by the Hyp mutation. Understanding the physiological roles of these genes may increase our understanding of the control over phosphate homeostasis.
![]() |
GRANTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
FOOTNOTES |
---|
Address for reprint requests and other correspondence: R. A. Meyer, Jr., Cannon Research Center, Rm. 304, Carolinas Medical Center, P.O. Box 32861, Charlotte, NC 28232-2861 (E-mail: rmeyer{at}carolinas.org).
1 The American Physiological Society sponsored a meeting at the Riverfront Augusta Hotel in Augusta, Georgia, October 14, 2003, titled "Understanding Renal and Cardiovascular Function through Physiological Genomics" and organized by David Pollock of the Medical College of Georgia (See the meeting report by Moreno C and Pollock DM. Physiol Genomics 16: 178179, 2004; 10.1152/physiolgenomics.00195.2003. http://physiolgenomics.physiology.org/cgi/content/full/16/2/178).
2 The Supplementary Material for this article (Supplemental Tables 58) is available online at http://physiolgenomics.physiology.org/cgi/content/full/00210.2003/DC1.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|