INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

AGING IS ASSOCIATED WITH A progressive decline in physiological function that is manifested at the molecular, cellular, and organ system levels. This decline in function is usually accompanied by a diminished ability to cope with environmental stress (2-4, 8, 17, 20, 25). However, the molecular mechanisms underlying age-related changes in stress responses are not well defined. Previous studies investigating alterations in gene expression have generally been confined to small sets of genes and have been focused on discrete cellular pathways. Specific areas of focus have included modifications in transcriptional regulation of stress-inducible genes (25), blunted synthesis of acute-phase proteins such as heat-shock proteins (5, 20), accumulative damage to proteins and DNA and resultant alterations in gene expression patterns (3, 4, 6, 25), and increased oxidative stress and subsequent biomolecular damage resulting from an increased rate of reactive oxygen species (ROS) generation (2).

The aim of this study was to examine the transcriptional response pattern of a wide array of stress-response genes to aging and environmental heating. We postulated that a comparison of gene expression profiles between young and aged animals responding to a physiologically relevant stress such as hyperthermia would provide insight into molecular mechanisms underlying age-related alterations in stress responses. To accomplish this aim, a cDNA expression array system was utilized. This recently available technique is advantageous because the expression profile of a large number of genes can be assessed in a defined set of physiological conditions (26). In addition, this strategy has the potential for identifying specific genes, expression patterns, or cellular pathways associated with both stress and aging that have not been previously identified.

In the present study, oligonucleotide-based arrays containing 207 stress-related genes were utilized to simultaneously compare gene expression profiles in the livers of young and senescent rats at a selected time point after hyperthermic challenge. Alterations in the expression of selected genes on the array were also confirmed with RT-PCR techniques. We focused on the liver because it shows age-dependent evidence of increased cellular ROS production (33) and is a prime target of tissue injury in physiological challenges such as heat stress (11, 18) and ischemia-reperfusion (10). Experimental animals were heat stressed on two consecutive days, separated by 24 h. This design was based on our observations that older rats are markedly less thermotolerant and have a much greater degree of cellular injury than their younger counterparts after repeated heat challenge (11, 12, 21). Thus this experimental design provides an excellent in vivo model for the high morbidity and mortality rates observed in older humans with a stress such as hyperthermia (23, 27).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Experimental Procedures

RNA isolation. Total liver RNA was isolated by the guanidinium acid-phenol-chloroform extraction method (7). First, liver tissues from three animals were ground in liquid nitrogen and then pooled. One gram of ground liver tissue was homogenized in a denaturing buffer containing guanidinium acid. Liver protein and DNA were separated from total RNA by three rounds of phenol-chloroform extraction. Total liver RNA was then precipitated by isopropanol. The concentration and quality of the RNA samples were determined with both a spectrophotometric method (a ratio of 260 nm-to-280 nm) and agarose gel techniques.

cDNA array hybridization. Rat stress-related expression cDNA arrays from Clontech Laboratory (Palo Alto, CA) were used for gene expression profiling. In these arrays, 207 stress-related genes were immobilized in duplicate on a nylon membrane. A total of four membranes were used, and all procedures were followed as recommended by the manufacturer. Liver mRNA was first purified from DNase-treated total RNA with oligo(dT)-linked Oligotex resin (Qiagen). One microgram of mRNA was used to generate [³²P]dATP-labeled double-strand cDNA fragments by RT. The labeled fragments were then hybridized onto the DNA array membranes overnight at 68°C. Membranes were subjected to stringent washing three times at 68°C and then exposed at room temperature overnight to a phosphorimaging filter. Images were generated using a phosphorimaging system (Molecular Dynamics Typhoon 8600 variable mode imager, Amersham Pharmacia Biotech).

Microarray phosphorimage assessments. To ensure accuracy in comparisons among arrays and to avoid saturation of the microarray images, the saturation level of the phosphorimager was first tuned to a low setting so that the genes most abundantly expressed were compared. The majority of the genes were then compared with a normal saturation level, with a few very abundant genes showing oversaturation levels. Very weakly expressed genes were also compared at a setting at which ~50% of genes were oversaturated. However, this condition did not provide additional information compared with values obtained at the second setting. Instead, the third setting served as a confirmation of the results obtained from the second setting for the genes that were weakly expressed.

RT-PCR. To confirm results of the cDNA array, RT-PCR was performed for four genes that exhibited distinct changes in expression with aging and after heating. The genes examined were heat shock protein 27 (HSP27), liver catalase, glutathione transferase subunit P, and c-Jun NH₂-terminal kinase 1. Two micrograms of pooled total liver RNA were first converted to cDNA by Moloney murine leukemia virus RT and oligo(dT) according to a protocol provided by Clontech. DNA fragments corresponding to the four genes of interest were synthesized with the use of primers specific to these genes (Clontech). The PCR reaction mixture contained 1 unit of Taq DNA polymerse, 1× reaction buffer with 1.5 mM MgCl, 0.4 µM each of forward and reverse primers, 0.2 mM each of dATP, dTTP, dGTP, and dCTP, 10 µl of RT reaction product, and 1 µl each of sense and anti-sense primers (10 pmol/µl). The primers for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were added simultaneously for each of the four genes being assessed. The PCR cycle was started with a denaturing period (94°C for 3 min), followed by 38 cycles of 1 min at 94°C, 1 min at 55°C, and 1.5 min at 72°C. The PCR cycle was finished with a 7-min extension time at 72°C. DNA fragments were stored at 4°C until they were separated on a 2.5% agarose gel. The separated DNA fragments were visualized under an ultraviolet light with ethidium bromide, and images were obtained using an AlphaImager digital camera (Alpha-Innotech, San Leandro, CA). RT-PCR was repeated three to four times for each gene.

Data Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

DNA Arrays

View larger version (96K): [in this window] [in a new window]   Fig. 1.   Rat cDNA arrays demonstrating differential gene expression in liver samples from young and old rats after 2 h of recovery from heat stress compared with the sham control condition. A: young control. B: young heated. C: old control. D: old heated. Representative phosphorimages from cDNA arrays containing 207 stress-related genes are presented. Each gene is represented by parallel points on the phosphorimages. Boxed areas indicate positions of 2 genes [heat shock protein 27 (HSP27) and catalase] highlighted to demonstrate differences observed in gene expression between age groups and the control and heat stress conditions.

Validation of DNA Array Data

View larger version (30K): [in this window] [in a new window]   Fig. 2.   DNA array results were confirmed by RT-PCR analysis. Representative RT-PCR images are presented for liver samples obtained from young and old rats that underwent a euthermic sham protocol or a heating protocol. Samples were obtained at 2-h recovery from heat stress. Two micrograms of pooled total liver RNA were first converted to cDNA by Moloney murine leukemia virus RT and oligo(dT). DNA fragments corresponding to the 4 genes of interest were synthesized with use of primers specific to these genes. The housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was assessed simultaneously with the other genes and used as a loading control. RT-PCR was repeated 3 or 4 times for each gene. Representative RT-PCR figures (top) and bar graphs (bottom) illustrate different comparisons for four genes: c-Jun NH2-terminal kinase 1 (JNK1; A), HSP27 (B), catalase (CAT; C), and glutathione S-transferase P subunit (GST-P; D). YC, young control; YH, young heated; OC, old control; OH, old heated. Values presented are means ± SE.

Global Changes in Gene Expression With Age and Heat Stress

Both aging and heat stress produced changes in gene expression (induction or suppression) in the liver. Table 2 shows the percentage of genes that had at least twofold changes in gene expression with aging and after heat stress. In comparisons between age groups, 9.7% of the 207 genes measured had changes in expression, with 5.3% of the genes induced and 4.3% of genes suppressed in old compared with young animals. After heat stress, changes in gene expression of 12.1 and 15.9% were obtained in the young and old groups, respectively. In both groups, a greater number of genes were induced than suppressed with heat stress.

These genes were further grouped into four categories according to the proposed functions of the proteins they encode. These categories included 1) proteins in cell growth, death, and signaling pathways; 2) stress-response proteins; 3) antioxidant and drug metabolism proteins; and 4) DNA, RNA, and protein synthesis-related enzymes. An overview of the changes in gene expression with aging and heat stress is shown in Table 3. Aging per se induced marked alterations in gene expression. These alterations included a greater basal expression of genes for cell growth, death, and signaling, as well as genes involved in stress responses (e.g., heat shock proteins). However, a striking suppression of genes for antioxidant and drug metabolism enzymes was noted in senescent livers. In addition, both young and old animals responded to heat stress with changes in their gene expression profiles. The most striking distinction between young and old animals was the difference in expression of stress-response genes associated with heat stress. After being subjected to heating, stress-response genes were strongly induced in young animals, whereas transcription levels for these stress-response genes were suppressed in the old animals.

Changes in Gene Expression With Age

Changes in Gene Expression With Heat Stress

When these genes were grouped into functional categories similar to Table 3, old animals had a larger number of changes in gene expression for transcripts in the cell growth, death, and signaling category than young animals after heating (10 vs. 6 genes; Table 5). Within this category, 4 of 6 genes in young animals and 6 of 10 genes in old animals were induced. For stress-response genes, aged animals exhibited a general downregulation in gene expression (5 of 7), whereas young animals had a strong pattern of induction (4 of 5). In the category of genes encoding antioxidant and drug metabolism enzymes, both young and aged animals showed a broad induction in gene expression (6 of 7 for young and 10 of 10 for old). Many genes in the DNA, RNA, and protein synthesis category were upregulated in both groups (4 of 6 for young; 5 of 6 for old) in response to heat stress.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The objective of this study was to examine the transcriptional response pattern of an array of stress-response genes to environmental heating and evaluate the influence of biological aging on these responses. The cDNA array technique provides a sensitive and rapid method for comparing the expression of a large number of genes among a variety of physiological conditions. Information obtained with the use of this strategy also provides insight into the selection of specific genes to evaluate in greater detail and allows for the development of a mechanistic framework regarding the molecular changes associated with the decline in stress tolerance in aged organisms. In our study, use of cDNA expression arrays specific for stress-related genes produced a broad profile of changes in gene expression in the liver after hyperthermic challenge in young and old rats. We focused on the liver because previous studies from our laboratory have demonstrated that this organ plays an important role in an organism's ability to cope with an environmental stress such as hyperthermia (11, 12, 19).

In one of the first studies to use oligonucleotide-based arrays to address changes in gene expression with aging, Lee et al. (22) reported that 113 of the 6,347 genes surveyed (2%) had greater than a twofold difference in gene expression in skeletal muscle samples from young compared with old mice (22). In the present study, a higher percentage (~10%) of genes surveyed showed differential expression with aging. These results appear quite reasonable because all 207 genes contained on the cDNA array utilized are thought to be indicators of stress responses. In addition, the higher rate of differences in expression of stress genes between young and old animals noted in the present study is consistent with the general concept that biological aging is characterized by reduced stress tolerance (2, 20, 25). In support of this concept, our results demonstrate that the older animals had less detectable gene expression signals (53 genes detected) than young animals (66 genes detected) in control conditions but a larger induction of genes after heating (15.9 vs. 12.1%).

Aging is associated with elevated morbidity and mortality rates in response to a variety of challenges (2, 13, 20, 25, 27), thus suggesting a decreased tolerance to stress with advancing age. In studies that compared young and old rats, we previously observed that aged animals are less thermotolerant, have a blunted stress protein response, and manifest extensive cellular injury in tissues such as the liver when heat challenged (11, 20, 21). Therefore, we hypothesized that the pattern of gene expression in old animals would be significantly different from that in young animals. Our results indicate that there is a blunted gene expression profile in old compared with young rats in the nonstressed control condition but a greater response to heat stress with advanced age.

To further evaluate the expression pattern of young and old animals, genes that had at least a twofold change in hybridization were grouped into four functional categories (Table 3). A distinct response pattern was observed in two of the four categories. For example, the old group exhibited an upregulation of several stress-response genes in control conditions. However, these older animals appeared unable to respond to heat stress properly, as demonstrated by the number of stress-response genes that were suppressed after hyperthermic challenge. In contrast, there was a strong induction of these stress-response genes in young animals. These results are in agreement with previous observations from our laboratory, indicating that aging is associated with decreased stress tolerance (11, 12, 19, 20). Our gene profile analysis also gives insight into specific genes that may be responsible for the age-related decrease in stress tolerance. Further studies on the regulation and expression of these genes, as well as the functional characteristics of their proteins, will advance our understanding of the mechanisms underlying age-related alterations in stress tolerance.

One proposed mechanism for aging is the oxidative stress hypothesis, which states that the loss of functional capacity associated with senescence is due to the progressive and irreversible accumulation of oxidative damage (2, 29, 32). It is postulated that the increase in oxidative stress and subsequent damage associated with aging are due to an increased rate of ROS generation, decreased removal of ROS, and/or a greater susceptibility of tissues to oxidative damage. Besides their damaging effects related to aging, ROS can also modulate gene expression, which certainly expands their role in the aging process (15). For example, it has been suggested that the increased oxidative injury that accompanies aging can result from a decrease in antioxidant enzyme defenses (9). The observation in the present study that old animals had a markedly reduced gene expression pattern in the antioxidant enzyme and drug metabolism category supports the basic tenets of the oxidative stress theory of aging.

Transcriptional levels of several genes encoding antioxidant enzymes were altered in response to heat stress. These observations, along with previous studies from our laboratory (11), indicate that ROS generation may play an important role in the liver pathology that accompanies heat stress. A potential source of superoxide radical generation could be from the leakage of mitochondrial electron transport chain reactions (14). Increased production of superoxide radicals by xanthine and/or xanthine oxidase after heat shock has also been reported (28). These superoxide radicals can be further converted to hydrogen peroxide by superoxide dismutases. Excess amounts of hydrogen peroxide, superoxide, and their by-product, hydroxyl radicals, which are produced through Haber-Weiss reactions, can either cause direct damage to cellular macromolecules or act as secondary messengers to produce intracellular oxidative stress. Thus our data implicate the involvement of an oxidative stress mechanism in the decline in stress tolerance that accompanies aging.

When evaluating the current set of data, there are potential limitations impacting the interpretation of the results that warrant discussion. The first involves the use of a single time point after heating (2 h) to evaluate differences in gene expression between age groups. It is possible that the time required to express genes could be altered with age. For example, the observation that there was no age-related difference in the expression of a particular gene at the 2-h time point after heat stress does not preclude the possibility that a difference in gene responses could be manifested at other time points poststress. However, preliminary experiments associated with the present study suggest that the general pattern of alterations in stress gene expression observed between age groups was similar at 2, 12, and 24 h after heat challenge. Additional studies addressing the temporal pattern of gene expression in response to a stress are certainly warranted. A second consideration involves the analysis of pooled liver samples from three animals for each microarray experiment. An experimental design in which a sample from a single animal was analyzed on a microarray would have allowed some insight into the consistency or repeatability of the gene expression patterns observed with aging and heat stress. However, several laboratories have used pooled samples in their gene expression studies (16, 24, 30), in part because this approach can reduce interanimal variability and increase the reliability of results.

Taken together, the DNA array results suggest that older animals do not respond to thermal challenges with an appropriate pattern of gene induction. Thus the ability of older animals to maintain cellular integrity and stability may be compromised. Our data also indicate that a physiological challenge such as heat stress generates a gene expression profile in the liver indicative of decreased stress protein transcription and increased expression of oxidative stress-related genes. These findings support the postulate that changes in transcriptional responses to physiological challenges such as hyperthermia contribute to the reduction in stress tolerance in older organisms. The use of the cDNA expression array approach will likely assist in delineating specific genes and molecular pathways associated with aging and the integrated cellular responses to physiologically relevant stressors.