This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

Reduction of Age-Associated Pathology in Old Mice by Overexpression of Catalase in Mitochondria

Departments of ¹ Comparative Medicine, ² Pathology, and ⁴ Biostatistics, University of Washington, Seattle.
³ Fred Hutchinson Cancer Research Center, Seattle, Washington.

Address correspondence to Piper M. Treuting, DVM, T140 Health Sciences Center, Box 357190, Seattle, WA, 98195-7190. E-mail: treuting{at}u.washington.edu

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
We describe the effects of mitochondrially targeted catalase (MCAT) expression on end-of-life pathology in mice using detailed semiquantitative histopathological evaluation. We previously reported that the median and maximum life spans of MCAT mice were extended relative to those of wild-type littermates. We now report that MCAT expression is associated with reduced malignant nonhematopoietic tumor burden, reduced cardiac lesions, and a trend toward reduced systemic inflammation, with no effect on hematopoietic neoplasia or glomerulonephropathy. Combined disease burden and comorbidity are also reduced, and MCAT expression is not associated with any detrimental clinical effects. The results suggest that oxidative damage is involved in aging of C57BL/6J mice via modulation of a subset of age-associated lesions. Antioxidant interventions targeting mitochondria may therefore be a viable strategy for prevention or postponement of some age-associated diseases. The variability of the MCAT effect across tissues, however, illustrates the importance of developing semiquantitative histopathology for assessment of comorbidity in life-span studies.

Key Words: Pathology • Free radicals • Oxidative stress

To test the free radical theory of aging in mice, and particularly the notion that mitochondria are important sources and targets of ROS damage, transgenic mice were generated that express mitochondrially targeted catalase (MCAT). Median and maximum life spans were increased 20% in MCAT animals (6). Additional observations previously noted on cross-sectional analysis of young and old MCAT mice included delayed onset of cardiac lesions and cataract development and reduction in oxidative damage, H₂O₂ production, H₂O₂-induced aconitase inactivation, and mitochondrial DNA deletions.

These results provided strong in vivo support for the free radical theory of aging in mammals. However, in the initial report, end-of-life (EOL) pathology was not addressed. In the present report, we characterize the pathology at EOL using a quantitative multilevel approach that can routinely and systematically be applied to any aging studies of mouse models.

	MATERIALS AND METHODS

Top Abstract Materials and Methods Results Discussion References

 
Animals
Two independent founder lines, B6.C3H-Tg(mCAT)4033Wcl and B6.C3H-Tg(mCAT)4403Wcl, were generated (6) and back-crossed as hemizygotes onto the C57BL/6J background (Jackson Laboratories, Bar Harbor, ME) for more that 10 generations. Control animals in this study were wild type (WT) littermates of MCAT transgenic animals. The line purity was confirmed by analysis of genetic markers (Charles River Laboratories, Wilmington, MA). Animals were multiply housed (4–5 per cage, separate sexes) in ventilated cages containing Bed-O-Cob (Andersons, Maumee, OH) in a specific pathogen-free facility at the University of Washington. Mice were fed irradiated Picolab Rodent Diet 20 #5053 (PMI Nutrition International, Brentwood, MO) and provided reverse osmosis water. All supplies entering animal rooms were autoclaved, and rooms were maintained at 70–74°F, 45–55% humidity, with 28 air changes per hour and a 12-hour light/dark cycle. Sentinel mice were tested quarterly and were negative for endo- and ectoparasites, mouse hepatitis virus, mouse parvovirus, and rotavirus, and they were tested annually for Mycoplasma pulmonis, pneumonia virus of mice, reovirus-3, Sendai virus, and Theiler's murine encephalomyelitis virus. Experimental mice were not tested for Helicobacter species or the newly recognized mouse Norovirus. All animal procedures were approved by the University of Washington Animal Care and Use Committee.

Cohort Description and EOL Determination
There were 1567 mice born into the current study cohort over 4 years. Collection of pathology specimens was closed 2 years later, although the overall cohort remains under observation for other studies. Mice that had died by the end of the pathology collection period were stratified into three categories: (i) EOL and used for the histopathological analysis reported here (n = 139); (ii) EOL but found dead in cage and thus not used for pathology (n = 269); and (iii) not EOL but killed for humane reasons as required by institutional guidelines (n = 184). In addition, animals were withdrawn from the cohort independent of genotype or illness for use in experiments (n = 593), and 382 mice were still alive and under observation at the end of the EOL pathology collection period.

Mice were considered to be at EOL and euthanized by CO₂ asphyxiation when they were moribund and demonstrated one or more clinical signs suggestive of imminent death within 24 hours. These signs included (i) nonresponsiveness to being touched; (ii) cold body temperature to the touch; (iii) slow or labored respiration; (iv) hunched body position with matted fur; (v) failure to eat and drink (as determined by food hopper weights and degree of dehydration); or (vi) prominent appearing ribs and spine, and sunken hips or >20% loss of body weight. All mice were observed daily by trained and experienced veterinary scientists 7 days a week. Mice that were nearing EOL, as determined by less severe clinical signs and up to 20% weight loss, were closely examined twice a day, 7 days a week.

Necropsy and Tissue Preparation
Routine necropsies were performed, and samples of all major organ systems and any grossly abnormal tissues were preserved by immersion in 10% neutral phosphate-buffered formalin. Fixed tissues were trimmed using a standard protocol developed by the University of Washington Veterinary Diagnostic Laboratory so that the same section of organ was examined from each animal. The tissues were routinely processed, and the paraffin-embedded samples were sectioned at 4–5 µm and stained with hematoxylin and eosin. Cross-sectional cohorts were used to collect fresh organ weight measurements. Detailed description of tissue collection, preparation, and staining are available in the online Supplemental Methods.

Grading of Histopathological Lesions
All neoplastic processes were graded according to Ikeno and colleagues (7), with modifications. Neoplastic processes were morphologically diagnosed and categorized as hematopoietic (HP; lymphomas, histiocytic sarcomas) or non-HP (other sarcomas, adenomas, and carcinomas), and the severity of each primary tumor was assigned a grade based on the extent of tumor size and distribution (metastasis) within the tissues examined. Grade 1 included small (<3 mm), focal tumors contained within the primary site and causing no observable changes to the surrounding parenchyma. Grade 2 tumors were larger with multiple foci within the primary site or had observable effect on surrounding parenchyma (compression or necrosis) or metastatic foci to one other organ. Grade 3 tumors had metastatic foci to 2–3 organs. Grade 4 tumors had metastases to >4 organs. The neoplastic lesion burden was calculated as the sum of severity scores (grades) for each primary tumor within a mouse. The population prevalence, severity, and neoplastic burden were then calculated.

Non-neoplastic lesions were graded on a severity scale from 0 to 4 with 0 as normal and 4 as a marked or severe change. To focus our analysis, we categorized non-neoplastic lesions as either incidental lesions (defined as those minor lesions that would not have contributed to the animal's moribund state) or contributing lesions (defined as those lesions that would affect the moribund state). The non-neoplastic contributory lesions included glomerulonephropathy (Supplemental Figure 1D and E) cardiac lesions (Supplemental Figure 1A–C), systemic inflammation (Supplemental Figure 1H and I), and, less commonly, hydronephrosis, renal infarction, hepatocellular necrosis, or inflammation associated with neoplasia, arterial medial hyperplasia, necrotizing and proliferative arteritis (Supplemental Figure 1F), and acidophilic macrophage pneumonia/histiocytosis (Supplemental Figure 1G). Common contributing lesions (those with >50% prevalence in both genotypes) were analyzed individually and included glomerulonephropathy, cardiac lesions, and systemic inflammation. Detailed histopathological descriptions of these non-neoplastic lesions, and material and methods for catalase expression and clinical chemistry assays, are provided in the Supplemental Methods section on page 821.

Determination of Disease Burden
The overall disease burden was calculated as the average sum of neoplastic and non-neoplastic severity scores for all mice within each genotype cohort. We only considered non-neoplastic lesions with Grade 3 or 4 (moderate to severe), with the rationale that Grade 1 and 2 lesions are too mild to contribute any significant burden. Neoplastic severity scores for all neoplastic lesions were used, because neoplasia of any severity would be burdensome. The number of non-neoplastic moderate to severe (Grade 3+) lesions was calculated as the number of all independent non-neoplastic processes (including incidental lesions) per mouse with Grade 3 or higher severity.

Determination of Cause of Death
Technically, EOL mice used for pathology assessment in this study did not die a natural death, but were killed when in a moribund state. Therefore, we used moribundity as a surrogate for death, because in our experience aged moribund mice will die within 24 hours. Our system is a modification of the cause-of-death assignment criteria according to Kodell and colleagues (8), where disease processes were determined to be probable, contributory, equivocal, or unknown to cause of death, as this allows for the potential of more than one process contributing to an aged animal's demise. A Contributing Causes of Moribundity (CCOM) score of 1 thus indicated a single cause of moribundity (death) whereas a CCOM of 0 indicated that a reasonable determination of moribundity could not be determined. The degree of comorbidity in a moribund mouse was then the sum of independent CCOM scores per mouse. For each category of potentially contributing lesions, neoplasias with severities of 2 or higher and non-neoplastic lesions of severities 3 or higher were considered potential CCOMs. The CCOM categories analyzed independently included HP neoplasia, non-HP neoplasia, and the common non-neoplastic lesions: glomerulonephropathy, cardiac lesions, and systemic inflammation.

Statistical Assessment
Fisher's exact test was used to compare proportions of mice between groups having specific kinds of pathology (e.g., proportion of MCAT vs WT mice with non-HP tumors). Welch's t test was used to compare the average severity score between groups for specific lesions (e.g., average neoplastic severity score). The Mann–Whitney U test was used to compare the mean degree of comorbidity between groups. To test for genotype effects on age-at-death for the various pathologic outcomes, regression models were fit for each pathology outcome with Genotype, Sex, Age, and Age x Genotype interaction terms for each model. Age was modeled continuously. Similarly, to test for the possibility of differential genotype effects by sex, regression models were fit for each pathology outcome with Genotype, Sex, and Sex x Genotype interaction terms in each model.

	RESULTS

Top Abstract Materials and Methods Results Discussion References

 
Two founder lines were examined in this study, 4403 and 4033. These had similar life spans, as previously described (6). No differences in prevalence or severity of lesions were seen between the founder lines; results from both were therefore analyzed as one group.

MCAT Mice Have Decreased Prevalence, Severity, and Burden of Non-HP Neoplasia
Most animals in this study (74% of MCAT, 79% of WT) had neoplasia at EOL, regardless of genotype. The complete spectrum and count of tumors identified in this study are listed in Table 1. The majority of the identified neoplasias were HP, and these tumors were found in 68% of MCAT mice and 66% of WT (Figure 1A). The HP neoplasia types were lymphoma, primarily pleomorphic/follicular, and histiocytic sarcoma, which are common in aged C57BL/6J mice (9,10).

View larger version (26K): [in this window] [in a new window]   Figure 1. Prevalence and severity of neoplastic lesions. A, The prevalence of hematopoietic (HP) and non-HP tumors was calculated as the number of affected animals at death/total number in population. This is a measure of lesion accumulation over life. Any lesion that appears and then disappears would not be counted. There were no significant differences by Fisher's exact test. B, The severity of HP and non-HP tumors within affected mice was graded on a scale from 1 to 4. Mitochondrially targeted catalase (MCAT) reduced the severity of non-HP tumors (p =.002). *Significant (p <.05) by Welch's t test. C, Combining the prevalence and severity gives the total population burden for each lesion. The burden of non-HP tumors was reduced in the MCAT population (p =.014). *Significant (p <.05) by Welch's t test. D, Benign and malignant non-HP tumors were examined separately. The burden of malignant non-HP tumors was reduced in the MCAT population (p =.007). *Significant (p <.05) by Welch's t test. WT = wild type

MCAT Mice Have Decreased Severity and Burden of Cardiac Lesions
The prevalence of cardiac lesions at EOL was >90% for both MCAT and WT, suggesting that the heart is a consistent target of age-associated pathology in our cohort (Figure 2A). Cardiac lesions consisted primarily of cardiomyopathy (97% of WT and 100% of MCAT) and arteriosclerosis (90% of WT and 94% of MCAT). Cardiac amyloidosis (17% of WT and 7% of MCAT) and valvular changes (86% of WT and 73% of MCAT) showed a trend toward reduced prevalence in MCAT (Supplemental Table 1). The cardiac lesion severity was calculated as the maximum lesion severity for any of these lesions (scored on a 0–4 scale), with the consideration that any of these lesions, if severe, could cause functional impairment to the animal. The cardiac lesion severity was reduced in MCAT mice relative to WT littermates (Figure 2B, p =.035) leading to reduced cardiac lesion burden in the MCAT population (p =.035, Figure 2C). Representative histological cardiac lesions are shown in Supplemental Figure 1A–C.

View larger version (31K): [in this window] [in a new window]   Figure 2. Prevalence and severity of non-neoplastic lesions. A, The prevalence of common contributory lesions (cardiac lesions, renal lesions, and systemic inflammation) was calculated as the number of affected animals at death/total number in population. This is a measure of lesion accumulation over life. Any lesion that appears and then disappears would not be counted. There were no significant differences by Fisher's exact test. B, The severity of cardiac lesions, renal lesions, and systemic inflammation was graded on a scale from 1 to 4. Mitochondrially targeted catalase (MCAT) reduced the severity of cardiac lesions (p =.035) with a trend in systemic inflammation (p =.071). *Significant (p <.05) by Welch's t test. C, Combining the prevalence and severity gives the total population burden for each lesion. The burden of cardiac lesions (p =.035) and systemic inflammation (p =.036) was reduced in the MCAT population. *Significant (p <.05) by Welch's t test. D, Other lesions were scored including rare contributory and incidental lesions. MCAT expression reduced the burden of contributory (p =.009) but not incidental (p =.176) non-neoplastic pathology. *Significant (p <.05) by Welch's t test. WT = wild type

Renal Disease Prevalence, Severity, and Burden Are Unaffected by MCAT
The prevalence of glomerulonephropathy at EOL was >90% for both MCAT and WT, suggesting that the kidney is a consistent target of age-associated pathology in our cohort (Figure 2A). Glomerulonephropathy, considered a result of immune-mediated glomerulonephritis (11), is common in aged C57BL/6J mice (12). However, we did not find an effect of MCAT on the prevalence, severity, or population burden of this lesion based on screening of hematoxylin and eosin–stained sections (Figure 2A–C). Detailed morphological analysis or evaluation by electron microscopy was beyond the scope of this study. Supplemental Figure 1D–F shows representative histological renal lesions.

MCAT Mice Have Reduced Burden of Rare Contributory Lesions
Other infrequent non-neoplastic lesions were consistent with those described previously in C57BL/6J mice (12–14). MCAT mice had significant reduction in the prevalence of incidental lesions (90% of WT vs 70% of MCAT, data not shown, p =.005) but no change in severity within affected individuals. MCAT mice showed reduced burden of rare contributory (p =.009) but not incidental (p =.176) non-neoplastic lesions (Figure 2D). Incidental lesions included extramedullary hematopoiesis in the spleen or liver, cystic endometrial hyperplasia, seminal vesicle ectasia, mild exocrine pancreatic atrophy, hyaline droplet nephropathy, cystic degeneration of the liver, mild liver atrophy, biliary cysts, testicular or accessory sex gland degeneration, and nonglandular stomach hyperkeratosis.

MCAT Mice Have Decreased Combined Disease Burden
Because a calculated estimate of the total lesion burden of a mouse at death is often desired for evaluation of the overall health status at EOL, we summed the scores of all moderate-to-severe (grade 3+) non-neoplastic lesions, including incidental lesions, together with the neoplastic burden to calculate combined disease burden. MCAT expression significantly reduced the number of total (p =.012) and contributory (p =.003) grade 3+ non-neoplastic lesions per mouse (Figure 3A). In addition, there was a significantly lower combined disease burden in MCAT mice relative to WT littermate controls (p =.005, Figure 3B) when both neoplastic and non-neoplastic lesions were combined. The non-neoplastic component of lesion burden was significantly lower in the MCAT mice (p =.002, Figure 3B). The neoplastic burden was not different between genotypes (p =.692, Figure 3B), due to the preponderant influence of HP neoplasms, which were not affected by genotype.

View larger version (26K): [in this window] [in a new window]   Figure 3. Combined disease burdens. A, The average number of non-neoplastic lesions per mouse was calculated. Mitochondrially targeted catalase (MCAT) reduced the total number of non-neoplastic lesions (p =.012) and the number of contributory lesions (p =.003). Prevalence of these lesions is shown by % on each bar. *Significant (p <.05) by Welch's t test. B, Combined disease burden was calculated as the sum of neoplastic (all) and moderate to severe (grade 3 and 4) non-neoplastic scores and averaged across each population to determine the total lesion burden. MCAT mice have lower combined disease burden (p =.005) and lower non-neoplastic lesion burden (p =.002). Prevalence of these lesions is shown by % on each bar. *Significant (p <.05) by Welch's t test. WT = wild type

View larger version (27K): [in this window] [in a new window]   Figure 4. Lesions present at end of life (EOL) that contributed to the moribundity of the animal. Lesions were considered to be contributing to moribundity with severity score 3 or 4 for non-neoplastic lesions and stage 2–4 for neoplastic lesions. A, The degree of comorbidity is defined as the number of separate severe lesions in each category. Fewer mitochondrially targeted catalase (MCAT) mice had high degrees of comorbidity (p =.012 by Mann–Whitney U test). B, Contributing causes of moribundity: the percentage of mice in each group with the contributory lesion category. MCAT expression reduced the number of mice with nonhematopoietic (non-HP) tumors (p =.009) with a trend toward reduction of the non-neoplastic lesions. *Significant (p <.05) by Fisher's exact test. WT = wild type

No Adverse Physiology or Pathology Was Observed To Be Associated With Mitochondrial Expression of Catalase
There were no significant differences in body weight, organ/body weight ratios, or clinical chemistry markers between MCAT and WT littermate controls (Supplemental Tables 2 and 3). The lack of difference in food intake in this cohort between MCAT and WT littermates was reported previously (6). There were no differences detected between genotypes in Disease x Age at death patterns in any of the analyses except in the case of the non-HP neoplasia score, where there was a significant Genotype x Age interaction. MCAT mice that died younger had significantly lower non-HP scores than MCAT mice that died older, whereas no such pattern was present for WT mice (p =.04 for the Genotype x Age interaction).

Catalase Expression and Activity Are Not Affected by Age
Unlike many short-term experiments, loss of transgene expression becomes a potential confounding factor in longevity experiments. To address this issue, catalase expression and activity were measured in heart samples from a cross-sectional cohort. This study included young (9–10 months), middle aged (18–20 months), old (26–30 months), and EOL mice (variable ages) (Supplemental Materials and Methods). Heart was chosen as a surrogate for whole-animal expression because expression levels across tissues are highly correlated within a mouse (data not shown, r² > 0.7 for heart, brain, and muscle). There were no changes in expression of the catalase transgene with age (p =.846), although large inter-animal variability was noted (Supplemental Figure 2A). Because RNA expression does not necessarily relate to catalase activity, transgene-specific expression was compared to total catalase activity; these were found to be closely correlated (r² = 0.96, p <.0001) (Figure 4B). To determine whether MCAT expression level was associated with life span, we examined the correlation between cardiac transgene expression and age at EOL and found no significant correlation or trend (r² = 0.007, p =.70) (Supplemental Figure 2C). There was also no significant trend between MCAT expression level at EOL and the presence of HP neoplasia (p =.50), severe (grade 3 or 4) renal lesions (p =.22), or severe (grade 3 or 4) cardiac lesions (p =.16) (Supplemental Figure 2D). The development of severe (grade 3 or 4) lesions was considered the threshold for significant burden to the animal. Non-HP neoplasia was not sufficiently common in the MCAT population to permit an analysis of trends with MCAT expression.

	DISCUSSION

Top Abstract Materials and Methods Results Discussion References

 
Using a mouse model of mitochondrial antioxidant overexpression and applying detailed EOL pathological assessment with a semiquantitative analysis, we show that MCAT expression is associated with a reduced disease burden, principally due to reduction in non-HP neoplasias and certain non-neoplastic lesions. These data suggest that the somatic overexpression of MCAT is beneficial in improving healthy aging, by reducing or delaying the appearance non-neoplastic disease, perhaps via a reduction in the initiation and/or promotion of non-HP neoplasms. The data also demonstrate that the effects of MCAT expression are selective, with impacts only on certain, but not all, age-associated lesions. Moreover, we find no evidence of detrimental effects of MCAT overexpression in the C57BL/6J mouse. These findings underscore the utility of detailed pathology assessment at EOL when animal models are used for studies of aging, as they provide additional power to reveal mechanisms contributing to the multifactorial processes of aging and longevity (14,15). These findings encourage further study of mitochondrially targeted antioxidant therapies.

ROS have been implicated in carcinogenesis via oxidative damage to nucleic acids and proteins (16,17). We found that non-HP malignancies such as fibrosarcoma, hemangiosarcoma, and carcinomas are significantly reduced in MCAT mice. Because the tumors formed in a variety of tissues, many of which have low-to-undetectable MCAT expression (data not shown), we speculate that circulating factors or immune competence may play a role in the reduced non-HP tumor phenotype. In contrast, there is no effect of MCAT expression on the development of HP neoplasia; which particularly affects C57BL/6J mice. Although it is possible that insufficient antioxidant protection was achieved in these compartments (due to either insufficient transgene expression or mosaic expression patterns), it is also likely that these particular disease processes do not depend on mitochondrial ROS-induced damage. The cause of HP neoplasia in the C57BL/6J background is unknown but may result from an endogenous retrovirus (18,19), which could shift the balance between tumor suppressors and oncogenes and promote carcinogenesis independent of oxidative damage or altered ROS signaling. Alternatively, HP neoplasia in mice, as in humans, may be facilitated by DNA recombination processes specific to generation of immune diversity, and these may be independent of ROS or oxidative damage.

There was a significant reduction in cardiac lesion burden and prevalence of severe cardiac lesions at EOL in MCAT mice relative to WT littermates. This finding is consistent with previous results of a cross-sectional study of 20- to 25-month-old mice (6) and with findings that MCAT mice have 40%–70% improved echocardiographic indices during aging (Dai D-F, Vermulst M, Tomazela DM, et al., unpublished observations, 2008). This in vivo study suggests that the pathological assessment may be underestimating the true physiological difference between genotypes. We measured four cardiac pathology parameters in the present study: cardiomyopathy, arteriosclerosis, amyloidosis, and valvular changes; whereas the burden of any one lesion was not significant, severe cardiac lesions and overall cardiac disease burden were significantly reduced in MCAT animals.

There are a number of challenges to the implementation of comprehensive studies of EOL pathology in mouse models, partly explaining why relatively few such studies are reported. Despite systematic efforts to identify moribund animals near death, 45% (n = 269) of the EOL cohort were found dead in their cages, with tissues too autolyzed to permit satisfactory histopathology. Furthermore, it is common practice to kill animals deemed to be in severe discomfort, including the effects of dermatitis, to which the C57BL/6J line is highly susceptible (12,20). In the present study, the resulting censoring of unhealthy mice of both genotypes (31%, n = 184) greatly complicates interpretation of the effects of genotype on life span. Median age at EOL for the 139 mice in this pathology study was 26 months (range 16–36 months) for the MCAT mice (n = 69) and 26 months (range 18–36 months) for the WT mice (n = 70), and median time to humane killing was not different for wild type and MCAT mice (p =.75). However, these simple statistics should not be taken as median life-span estimates for the entire cohort, as they do not take into account effects of differential censoring, nor do they include the entire study period (mice are included only up to the closure of the collection period for the pathology study). This fuller analysis will be the subject of a follow-up report. We conclude from these observations, however, that improved guidelines and practices are required to reduce the number of both animals found dead as well as animals censored by killing. Furthermore, as hybrid animals (e.g., F1 and four-way crosses) (21,22) are generally less sensitive to dermatitis and other strain-related diseases, they should be considered for future studies of health span and life span.

In the field of aging research, use of animal models is common practice for the study of interventions that modulate life span. However, although all organisms show signs of functional declines with age, it is difficult to determine whether successful life-span extension in animal models merely corrects a particular weakness of the studied organism or whether the underlying mechanisms of aging have been altered. Typically, pathology is used in survivorship studies to determine a single probable cause of death in response to a toxic insult. The hallmark of aging, however, is the presence of multiple comorbidities and their complex contributions to the modulation of life span. It will therefore be important to go beyond currently prevailing methodologies to include semiquantitative or quantitative assessments of comorbidities.

Combined disease burden, defined as the average sum of total neoplastic burden and significant non-neoplastic lesions in a population, has been used to assess animal health status at EOL (7,23). We modified this protocol to include measures of comorbidity. A summarization statistic, based on counts of the number of co-occurring contributing lesions per mouse, revealed subsets of high-burden and low-burden animals. Furthermore, we characterized the cause of death using a method we are calling CCOM (to reflect multiple contributing causes of moribundity). We found that MCAT mice have reduced combined disease burden, primarily due to the reduction of non-HP malignancies and non-neoplastic lesions and a reduced degree of comorbidity. Our results suggest that MCAT mice accumulate less disease over their lives and, in combination with the previous life-span extension data (6), suggest that reduction of oxidative damage via overexpression of MCAT extends healthy aging (i.e., health span) as well as longevity in C57BL/6J mice.

Because loss of transgene expression during aging could affect late-life disease, we investigated the possibility that MCAT expression was lost over time. Despite large inter-animal variability, there was no trend toward a change in MCAT expression with age (Supplemental Figure 2A). Moreover, expression was highly correlated with catalase activity (Supplemental Figure 2B). The cause of interanimal variability in expression is unknown but may arise from variations in epigenetic silencing. Furthermore, MCAT expression, which correlated strongly with catalase activity in heart, did not correlate with age at EOL (Supplemental Figure 2C). Additionally, there was a similar mean and variance in MCAT expression in mice with non-HP neoplasias or renal lesions (Supplemental Figure 2D), indicating that MCAT expression level is not a strong predictor of susceptibility to death or to the development of these age-associated lesions. This result suggests that even low or mosaic expression of MCAT is sufficient to confer a protected phenotype, as was concluded previously from comparison of higher and lower MCAT expressing founder lines, both of which had similar life-span extension (6).

Summary
Using detailed semiquantitative histopathological analyses, we were able to define those organ systems and lesions in which there was a positive effect from the MCAT model of increased antioxidant expression. Importantly, no negative effects of life-long MCAT expression were detected. These data suggest that, with more information on specific sites of action, antioxidant intervention (particularly mitochondrial antioxidants) may be a viable strategy for the prevention of certain human neoplastic processes and age-associated diseases. The approach for analysis of EOL pathology that we have described, accounting for combined disease burden and comorbidity, can provide a framework for determining the effects of other interventions on the multiple spontaneous comorbidities present in aging populations of mice.

Top Abstract Materials and Methods Results Discussion References

 
SUPPLEMENTAL METHODS
Necropsy and Tissue Preparation.-- After euthanasia, the mouse was weighed and blood collected via cardiac puncture. The hearts were weighed after the remaining blood was gently expelled and placed in a heart-sectioning guide mold and cut into five approximately equal transverse sections for histopathology and biochemical analysis. Lungs were insufflated with 10% neutral phosphate-buffered formalin prior to immersion fixation. Brains were removed prior to fixation and weighed. Remaining major organs were dissected and fixed. Necropsy findings were recorded by the prosector using standard necropsy forms. In addition to the standard tissue trimming protocol, the heart was dissected using a sectioning guide mold. For each animal, 3–5 slides with 3–5 tissues per slide were sectioned. One hematoxylin and eosin-stained section was evaluated per paraffin block for most tissues. Additionally, selected sections were stained with Masson's Trichrome, Congo red, and Von Kossa to highlight fibrosis, amyloid, or calcification, respectively.

Non-Neoplastic Lesions.-- Glomerulonephropathy (Supplemental Figure 1G and H), either membranous, proliferative, or membranoproliferative, was defined as increased mesangial matrix or cellularity with variable tubular protein casts, interstitial, or tubular changes.

Cardiomyopathy (Supplemental Figure 1A) was defined as combinations of interstitial fibrosis resulting in compartmentalization of the cardiomyocytes with hyaline cytoplasmic change, vacuolization of cytoplasm, variable myocyte fiber size, hypercellularity, and collapse of sarcomeres with variable interstitial lymphohistiocytic inflammation. Arteriosclerosis (Supplemental Figure 1B) was defined as combinations of medial hypertrophy, hyalinization, and media mineralization. Valvular changes (Supplemental Figure 1C) in the atrioventricular valves of the heart included combinations of cartilaginous metaplasia, myxomatous degeneration, mineralization, and pigmentation. Amyloidosis (Supplemental Figure 1H) was defined as the presence of acellular interstitial material that, when stained with Congo red, is green under polarized light. The overall cardiac score was defined to be the maximum score for any of the graded lesions. This cardiac score was created in this manner under the reasoning that any of these lesions, if severe, could cause functional impairment to the animal.

Histological markers of systemic inflammation included the associated processes of amyloidosis and accumulations of lymphoid nodules in various organs. Amyloid A (AA amyloid) is an extracellular accumulation of serum amyloid A protein that is a sequelae of chronic inflammation or stress induced by husbandry factors (8), and occurs with aging in C57BL/6 mice most frequently in the intestines, heart, and glomeruli (8,9). Lymphoid nodules are thought to result from chronic antigenic stimulation (8). We combined these biologically related lesions into a category denoted as systemic inflammation. The systemic inflammation score was used as a combined measure and calculated as an integer scale of 0–4 according to increasing degrees of distribution and severity with 0 representing normal tissue and 4 representing six or more organs affected mildly or three or more organs affected severely. The brain was not evaluated as part of this study.

Clinical Chemistries.-- Serum was extracted from whole blood using a Microtainer tube (Becton Dickinson, Franklin Lakes, NJ) and rapidly frozen then stored at –70°C. Clinical chemistry assessments were performed according to standard procedures at Phoenix Central Laboratory (Everett, WA).

Catalase Expression and Activity.-- Tissue samples were dissected, flash frozen in liquid nitrogen, and stored at –80°C. MCAT mRNA expression from cardiac ventricle was assessed relative to GAPDH as described previously (6). Catalase activity was measured according to Yasmineh and colleagues (9). Briefly, crude lysates were prepared in pH 7.0 phosphate buffer by homogenization (Fisher PowerGen) and protein concentration was determined with the BCA kit (Pierce, Rockford, IL). Each sample was run in triplicate in a 200 µl reaction with 1 mg crude protein lysate from cardiac ventricular region. NADPH absorbance was measured at 340 nm using a Spectracount plate spectrophotometer (Packard Instruments, Meriden, CT) once per minute for 15 minutes. Purified catalase (Sigma, St. Louis, MO) was used for standard curve measurements to confirm linearity and values were normalized to a reference sample.

View larger version (161K): [in this window] [in a new window]   Supplemental Figure 1. Representative examples of non-neoplastic lesions considered to be severe in this study. Sections were stained routinely with hematoxylin and eosin except where noted. Original magnification for all images is 20x except where noted. A, Heart from a 26-month-old female wild-type (WT) mouse. Notice the collapse of the epicardial surface (arrow) due to interstitial fibrosis, loss of cardiomyocytes, and chronic inflammation. B, Arteriosclerotic coronary artery from a 31-month-old female WT mouse. C, Thickened heart valves from a 26-month-old female WT mouse with mineralization (positive staining with Von Kossa; data not shown) and hemosiderosis (black granular material). Original magnification 4x; inset 20x. D, Renal cortex from a 30-month-old female mitochondrially targeted catalase (MCAT) mouse with membranoproliferative glomerulonephritis and chronic interstitial nephritis. Note the obsolescent glomerulus (arrow). E, Periodic acid-Schiff (PAS)-stained section of kidney from the same mouse as D. Note the intralumenal PAS-positive casts (arrow) and increased PAS-positive material within the glomerular mesangia. F, Necrotizing and proliferative renal arteritis from a 26-month-old female mouse. G, Acidophilic macrophage pneumonia from a 26-month-old female WT mouse. Note the multinucleated cell with abundant eosinophilic cytoplasm (arrow). H, Small intestinal amyloidosis in a 32-month-old male WT mouse. Note the homogenous light pink material (asterisk) expanding the interstitium. I, Lymphohistiocytic aggregate within the kidney of a 27-month-old female WT mouse. Notice the mixed population of lymphocytes, plasma cells, and Mott cells (arrow inset) within a lymphoid aggregate from the same animal with the omental fat (original magnification 40x)

View larger version (23K): [in this window] [in a new window]   Supplemental Figure 2. Catalase expression is maintained with aging. A, Levels of messenger RNA (mRNA) for the mitochondrially targeted catalase (MCAT) were evaluated by quantitative reverse transcription–polymerase chain reaction in heart from a cross-sectional cohort and expressed relative to a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) control. There was no difference in MCAT mRNA level with age (p =.846; analysis of variance). B, Catalase activity was compared to MCAT expression (r2 = 0.96, p <.0001). C, Cardiac MCAT expression was compared to age at death for a group of end-of-life (EOL) animals. There was no significant correlation between age at death and level of MCAT expression. (r2 = 0.007, p =.70). D, Cardiac MCAT expression at EOL was analyzed in subsets of animals with each contributing cause of moribundity. There was no significant difference between MCAT expression levels in animals with or without life-limiting hematopoietic (HP) neoplasia (p =.50), renal lesions (p =.22), or cardiac lesions (p =.16) by Welch's t test

	Acknowledgments

Top Abstract Materials and Methods Results Discussion References

We thank H. Denny Liggitt, Charles Frevert, and Rosana Risques for critical review of the manuscript; Ruby Mangalindan and Ashot Safarli for expert care and observation of the mice; Serina Tsang for collection of tissues; and Heather Hopkins for colony data management.

P. M. T. and N. J. L. contributed equally to this work.

Nancy Linford is currently at the Huffington Center on Aging, Baylor College of Medicine, Houston, TX.

	Footnotes

Top Abstract Materials and Methods Results Discussion References

 
Decision Editor: Huber R. Warner, PhD

Received March 11, 2008

Accepted May 10, 2008

	References

Top Abstract Materials and Methods Results Discussion References

 

1. Harman D. Free radical theory of aging: an update: increasing the functional life span. Ann N Y Acad Sci. 2006;1067:10-21.[Medline]
2. Bokov A, Chaudhuri A, Richardson A. The role of oxidative damage and stress in aging. Mech Ageing Dev. 2004;125:811-826.[Medline]
3. Sun J, Folk D, Bradley TJ, Tower J. Induced overexpression of mitochondrial Mn-superoxide dismutase extends the life span of adult drosophila melanogaster. Genetics. 2002;161:661-672.[Abstract/Free Full Text]
4. Orr WC, Radyuk SN, Prabhudesai L, et al. Overexpression of glutamate-cysteine ligase extends life span in drosophila melanogaster. J Biol Chem. 2005;280:37331-37338.[Abstract/Free Full Text]
5. Linford NJ, Schriner SE, Rabinovitch PS. Oxidative damage and aging: spotlight on mitochondria. Cancer Res. 2006;66:2497-2499.[Abstract/Free Full Text]
6. Schriner SE, Linford NJ, Martin GM, et al. Extension of murine life span by overexpression of catalase targeted to mitochondria. Science. 2005;308:1909-1911.[Abstract/Free Full Text]
7. Ikeno Y, Bronson RT, Hubbard GB, Lee S, Bartke A. Delayed occurrence of fatal neoplastic diseases in ames dwarf mice: correlation to extended longevity. J Gerontol A Biol Sci Med Sci. 2003;58:291-296.[Medline]
8. Kodell RL, Blackwell BN, Bucci TJ, Greenman DL. Cause-of-death assignment at the National Center for Toxicological Research. Toxicol Pathol. 1995;23:241-247.[Abstract/Free Full Text]
9. Blackwell BN, Bucci TJ, Hart RW, Turturro A. Longevity, body weight, and neoplasia in ad libitum-fed and diet-restricted C57BL6 mice fed NIH-31 open formula diet. Toxicol Pathol. 1995;23:570-582.[Abstract/Free Full Text]
10. Ward JM. Lymphomas and leukemias in mice. Exp Toxicol Pathol. 2006;57:377-381.[Medline]
11. Porter DD, Porter HG, Cox NA. Immune complex glomerulonephritis in one-year-old C57BL-6 mice induced by endogenous murine leukemia virus and erythrocyte antigens. J Immunol. 1973;111:1626-1633.[Abstract/Free Full Text]
12. Ward JM, Anver MR, Mahler JF, Devor-Henneman DE. Pathology of mice commonly used in genetic engineering (C57Bl/6; 129; B6, 129; and FVB/N). In: Ward JM, Mahler JF, Maronpot RR, Sundberg JP, Frederickson RM. Pathology of Genetically Engineered Mice. First edition. Ames: Iowa State Press; 2000:161–179.
13. Lipman RD, Gaillard ET, Harrison DE, Bronson RT. Husbandry factors and the prevalence of age-related amyloidosis in mice. Lab Anim Sci. 1993;43:439-444.[Medline]
14. Lipman RD, Bronson RT, Wu D, et al. Disease incidence and longevity are unaltered by dietary antioxidant supplementation initiated during middle age in C57BL/6 mice. Mech Ageing Dev. 1998;103:269-284.[Medline]
15. Lipman RD, Dallal GE, Bronson RT. Lesion biomarkers of aging in B6C3F1 hybrid mice. J Gerontol Biol Sci. 1999;54A:B466-B477.[Abstract]
16. Finkel T. Oxidant signals and oxidative stress. Curr Opin Cell Biol. 2003;15:247-254.[Medline]
17. Finkel T, Holbrook NJ. Oxidants, oxidative stress and the biology of ageing. Nature. 2000;408:239-247.[Medline]
18. Gayama S, Vaupel BA, Kanagawa O. Sequence heterogeneity of murine acquired immunodeficiency syndrome virus: the role of endogenous virus. Int Immunol. 1995;7:861-868.[Abstract/Free Full Text]
19. Kubo Y, Nakagawa Y, Kakimi K, et al. Molecular cloning and characterization of a murine AIDS virus-related endogenous transcript expressed in C57BL/6 mice. J Gen Virol. 1994;75:881-888.[Abstract/Free Full Text]
20. Kastenmayer RJ, Fain MA, Perdue KA. A retrospective study of idiopathic ulcerative dermatitis in mice with a C57BL/6 background. J Am Assoc Lab Anim Sci. 2006;45:8-12.[Medline]
21. Miller RA, Nadon NL. Principles of animal use for gerontological research. J Gerontol Biol Sci. 2000;55A:B117-B123.[Abstract/Free Full Text]
22. Miller RA, Burke D, Nadon N. Announcement: four-way cross mouse stocks: a new, genetically heterogeneous resource for aging research. J Gerontol Biol Sci. 1999;54A:B358-B360.[Medline]
23. Bronson RT, Lipman RD. Reduction in rate of occurrence of age related lesions in dietary restricted laboratory mice. Growth Dev Aging. 1991;55:169-184.[Medline]

This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation