Systemic overexpression of C-C motif chemokine ligand 2 promotes metabolic dysregulation and premature death in mice with accelerated aging

Ccl2 overexpression reduced the lifespan of mice with accelerated aging

The increase in body weight significantly differed among strains, and Ccl2 overexpression was associated with different body weight changes, significant impairment in fertility, earlier-than-normal dermal thinning and considerable loss of muscle and fat. Mice with a single mutation in lamin A (LMNAG609G/+) gradually lost weight with respect to the weight of the wild-type mice from 25 to 32 weeks and the mice with a mutation in lamin A and overexpressing Ccl2 (LMNAG609G/+;CGCCL2+/-) from 12 to 14 weeks (Figure 1A). The progeroid strains also progressively diverged in terms of the development of features of accelerated aging, including growth rate, lipodystrophy and lordokyphosis (Figure 1B). The median lifespan was significantly reduced (33%) in the mice with overexpressed Ccl2 (45 weeks for the LMNAG609G/+ mice and 30 weeks for the LMNAG609G/+;CGCCL2+/- mice) (Figure 1C). Degenerative cardiovascular changes, cachexia and sarcopenia also developed earlier in the LMNAG609G/+;CGCCL2+/- mice than in LMNAG609G/+ mice. The progeroid strains showed major histological alterations in the aorta and femoral arteries with respect to the controls. The extent of arterial fibrosis was greater in the LMNAG609G/+;CGCCL2+/- and LMNAG609G/+ mice than in the controls. Although the expression of actin and the number of stained macrophages were significantly higher in the mice with overexpressing Ccl2, the differences in fibrosis between the progeroid strains did not reach statistical significance (Supplementary Figure 1). The histological structure of the cardiac muscles in the progeroid mice was also significantly altered, and the changes were immediately apparent with routine staining procedures. Increased fibrosis of cardiac muscles was the most prominent feature caused by Ccl2 overexpression (Supplementary Figure 2). Similarly, the generalized loss of fat depots and the extent of the abnormal structure of brown and epididymal white adipose tissue were significantly greater in the mice overexpressing Ccl2. In the tissues from both progeroid mouse strains, the adipocyte area was lower with higher expression of uncoupling protein one and an increased number of stained F4/80 cells compared to these levels in the controls. All these features and fibrosis were significantly affected by Ccl2 overexpression (Supplementary Figure 3). The sarcopenia and cachexia causing the skeletal muscle loss in the genetically modified mice were significantly altered by Ccl2 overexpression. The histological changes in both models of accelerated aging showed that fibrosis was significantly enhanced, but the response against oxidation was decreased, and the number of anti-inflammatory macrophages in muscle was significantly lower compared to those of the controls (Figure 2). The data indicate that Ccl2 overexpression promoted death in mice with accelerated aging and confirmed the role of Ccl2 as a fibrosis-inducing agent with a negative impact on biological outcomes associated with aging and/or longevity. A caveat of this set of experiments suggests that the ages at which the animals were studied differed for each experimental group (38-40 weeks for WT, 32 weeks for LMNAG609G/+, 18-22 weeks for LMNAG609G/+;CGCCL2+/-), since we decided to sacrifice the animals when they presented a similar degree of senescence, according to the criteria described in Materials and Methods. The age difference could explain, at least partially, the difference in the results that we observed in the arteries.

Ccl2 overexpression drove specific signatures of metabolic dysregulation in the skeletal muscles of the progeroid mice

To better understand the changes induced by lamin A mutation, we first performed a metabolomic analysis in muscle from fasted LMNAG609G/+ and WT mice. We selected this tissue due to its important implications in aging development. Metabolites from energy balancing and 1-C metabolism were quantitatively assessed, and normalized concentrations from all strains are depicted in Supplementary Tables 1, 2. The sets of metabolites studied demonstrated a clear distinction between LMNAG609G/+ and the controls. The partial least square discriminant analysis and the heatmap showed a clear distinction between animals from both groups (Supplementary Figure 4). The examination of individual metabolites showed that the LMNAG609G/+ animals had increased concentrations of compounds involved in energy metabolism. Notably, the increase in the concentrations of metabolites related to glycolysis, the analyzed amino acids, and the anaplerotic reactions in the CAC was important. 1-C metabolism was also altered in these mice, and the methionine level was significantly increased with a decrease in 5-methyl-tetrahydrofolate (5-methyl-THF), S-adenosylmethionine (SAM) and S-adenosylhomocysteine (SAH) levels, although we did not observe any significant alterations in DNA methylation (Figure 3B). These data likely indicate a combined defective function in the folate and methionine cycles.

The relative impact on the concentrations of these metabolites was different in the LMNAG609G/+;CGCCL2+/- mice than it was in the controls (Supplementary Figure 5). The main alterations observed were an increase in branched-chain amino acids (valine, isoleucine and leucine), β-hydroxybutyrate, fructose 1,6-bisphosphate, glucose 6-phosphate and ribose-5-phosphate concentrations, together with a depletion in 3-phosphoglycerate and citrate-isocitrate. Regarding 1-C metabolism, we observed an increase in riboflavin and methionine levels and a decrease in 5-methyl-THF, S-adenosylmethionine (SAM) and SAH. These alterations in 1-C metabolism induced an increase in the percentage of 5-methylcytosine (5-mC), which is a marker of DNA methylation. Interestingly, with respect to the controls, in both progeroid strains, the accumulation of the amino acids valine, leucine and isoleucine was significant.

We finally compared the metabolic alterations of LMNAG609G/+-;CGCCL2+/- with those of LMNAG609G/+, and we found a significant impact on the amount of metabolites in skeletal muscle that distinguished the strains, according to the results of the principal component analysis (Figure 3). Regarding energy metabolism, the LMNAG609G/+;CGCCL2+/-^- mice showed alterations in the glucose and citric acid cycles without significant alterations in amino acid concentrations with respect to the LMNAG609G/+ mice. In relation to 1-C metabolism, the LMNAG609G/+;CGCCL2+/- mice showed the downregulation of methionine, choline-dimethylglycine, 5 methyl-THF and 5 formyl-THF. These alterations may have been a consequence of the increase in SAM, SAH and AMP levels that we found in these animals. One of the most important consequences of SAM to SAH alterations was the increase in 5-mC. These findings might be relevant because they strongly suggest alterations in muscle DNA methylation and demethylation. DNA methyltransferases use SAM as a substrate to produce SAH; DNA demethylation depends on α-ketoglutarate and other metabolites from the CAC [24]. Ccl2 overexpression significantly increased the methylation of cytosine bases in muscle DNA and significantly decreased 5-hmC levels with respect to both the LMNAG609G/+ mice and controls (Figure 3C). Taken together, our data indicate that Ccl2 overexpression in the experimental model of progeria alters energy and 1-C metabolism and drives dynamic changes in DNA methylation. The role of these epigenetic marks remains to be ascertained.

Ccl2 overexpression disrupted mitochondrial dynamics and altered AMPK/m-TOR-driven pathways in the progeroid mice

Our data might suggest key roles of mitochondrial metabolites in pathological aging. Although the mechanisms are uncertain, Ccl2 overexpression interferes with mitochondrial function, particularly in the oxidative phosphorylation system (OXPHOS) and the production of ATP and reactive oxygen species (ROS) [25]. Indeed, our group previously reported that transgenic mice with a WT background that overexpress Ccl2 show alterations in mitochondrial morphology; therefore, the matrix is less electron-dense. We also found altered fusion dynamics. In transgenic mice fed a chow diet, a tendency of the unbalanced dynamics shifted towards mitochondrial fusion, but dietary manipulation elicited a significant shift towards fission. These putative mechanisms of the altered metabolism seem to be related to the autophagic response [20].

We therefore explored the expression of the mitochondrial respiratory complexes in skeletal muscle. Complexes I and II are major players in mitochondrial signaling and ROS generation, and their expression was significantly decreased in the progeroid mice compared with that of the controls. The expression levels, especially of complex I, were lower in the LMNAG609G/+;CGCCL2+/- mice than in the LMNAG609G/+ mice. Additionally, complexes III and IV appeared normal in the muscles of the LMNAG609G/+ mice with respect to the muscles of the controls, but Ccl2 overexpression in the LMNAG609G/+;CGCCL2+/- mice significantly decreased complex expression. Finally, the expression of complex V (ATP synthase) in the muscles of the LMNAG609G/+ mice was significantly decreased, and this effect was significantly more prominent in the LMNAG609G/+;CGCCL2+/- mice. In addition, the protein concentrations of translocase of outer membrane 20 (TOM20) and mitofusin 2 (MFN2) in the muscles of the progeroid mice were decreased, indicating alterations in the proper formation of the mitochondrial network (Figure 4A).

Ccl2 overexpression had a negative impact on the cellular bioenergetics in progeroid skeletal muscle, and this negative effect was associated with reduced efficiency of mTOR-related homeostatic mechanisms. In the muscles of the LMNAG609G/+ mice, the expression and phosphorylation of p53 was increased with respect to that of the controls, and the level of p53 activation in the LMNAG609G/+;CGCCL2+/- mice was significantly increased with respect to that in the LMNAG609G/+ mice, which likely indicates the triggering of aging through a p53-mediated transcriptional program. As p53 is a major factor in the cell cycle and proliferation and because senescent cells may be more or less metabolically active, the differential response may be the result of the differences in the expression and activation of related molecules, most likely those associated with the mechanistic target of rapamycin (mTOR). We explored major upstream and downstream pathway players and found that the phosphoinositide 3-kinase (PI3K)/AKT (also known as protein kinase B) pathway, with critical roles in regulating diverse cellular functions, was abnormal. PI3K and AKT expression was not affected in the cells of either the progeroid or the Ccl2-overexpressing strains. However, AKT was not phosphorylated at S473 in the muscle cells of either progeroid mouse strain, and S6 phosphorylation was significantly decreased in the LMNAG609G/+ mice and undetectable in the LMNAG609G/+;CGCCL2+/- mice, indicating that mTOR signaling is significantly inhibited in these models of accelerated aging. AMP-activated protein kinase (AMPK) expression was significantly lower in the cells of the genetically modified mice, but its relative phosphorylation was significantly higher. The effect was more pronounced in the LMNAG609G/+;CGCCL2+/- mice than in LMNAG609G/+ mice. The cells overexpressing Ccl2 in the LMNAG609G/+;CGCCL2+/- mice apparently favored autophagy with respect to the cells in the LMNAG609G/+ mice, as indicated by a higher LC3II/I ratio. However, the expression levels of the lysosomal membrane proteins LAMP2A and p62 were decreased in the progeroid mice, suggesting the involvement of chaperone-mediated autophagy in aging and the ability of Ccl2 to interfere in this process (Figure 4B). However, we cannot verify that CCL2 decreases ATP levels AMPK or, in contrast, AMPK is activated because of other changes, such as the upregulation of Ca²⁺ levels.

Animals and experimental design

Genetically modified mice that overexpress Ccl2 (CgCcl2) were generated in our laboratory as described [20]. The mice with the Lmna ^G609G mutation (LA), which constitutes a model of accelerated aging and progeria, were generated at the University of Oviedo [23]. Crossbreeding resulted in the first progeny (F1) of heterozygous animals with the LMNAG609G/+ mutation and Ccl2 overexpression in tissues (LMNAG609G/+;CGCCL2+/-), which was confirmed by routine genotyping and Ccl2 measurements (n = 34). In the LMNAG609G/+ mice, the Ccl2 concentrations in tissues were similar to those observed in the controls, except in adipose tissues, but the Ccl2 levels in the LMNAG609G/+;CGCCL2+/- mice were significantly higher in all tissues. The second progeny (F2) of the LMNAG609G/+;CGCCL2+/- mice were not fertile, and consequently, homozygotes for the two characteristics were not available. Wild-type (WT) (n = 20) and LMNAG609G/+ mice (n = 28) were used as controls (Supplementary Figure 6). The mice were fed a normal standard diet from Scientific Animal Food & Engineering (SAFE, Augy, France) and water ad libitum for the duration of the experiments and maintained under controlled temperature (22° C), humidity (50%) and a light/dark cycle (12 h/12 h). The animal procedures were approved by the Ethics Review Committee for Animal Experimentation of Universitat Rovira i Virgili (protocols 10281 and FUE-2018-00849209) and designed according to European guidelines (Directive 2010/63/EU).

The mice included in the study were male with a C57BL/6J genetic background, used at 4 weeks of age and assessed daily for lifespan indicators with manipulation maintained at a minimum. To avoid unnecessary suffering and to ensure the lack of bias, we established strict criteria for the decision to sacrifice that was based on a previously reported grading score of senescence (Supplementary Table 3) [47].

Tissue collection, the preparation of extracts and Ccl2 measurements

All tissues were collected upon mouse sacrifice, and selected pieces were frozen and stored at -80° C until batch analyses. Portions of each 30 mg sample of liver, pancreas, skeletal muscle (from quadriceps muscles), heart, major arteries (aorta and femoral) and brown and epididymal white adipose tissue were homogenized using a sonicator (Branson Sonifer 150, Thistle Scientific, Glasgow, UK) in 50 mM Tris, 1 mM EDTA, 1% IGEPAL CA-630, 150 mM NaCl, 0.10% Triton, 50 mM NaF, 100 mM phenylmethanesulfonyl fluoride (PMS) and 1 mM Na₃VO₄. Protein concentration was determined with the Pierce BCA protein assay kit. To measure Ccl2 by ELISA (PeproTech, London, UK), following the manufacturer’s instructions, 200 μg of protein from each extract was used.

Histological analysis and immunohistochemistry

Separate portions of the same tissues were collected in formalin (3.7-4% formaldehyde buffered to pH 7 and stabilized with 1-1.5% methanol) and embedded in paraffin. Hematoxylin and eosin and Sirius red staining were used for each tissue. At least 10 sections per mouse were assessed, and the tissue was blindly evaluated to assess defects in structure and detect fibrosis. For immunohistochemistry, antigens were retrieved in Tris 10 mM/EDTA 1 mM buffer at pH 9 in a microwave oven that reached 90° C. Bovine serum albumin (2%) and hydrogen peroxide (1%) were sequentially used to block nonspecific binding sites and endogenous peroxidase, and the sections were rinsed with phosphate-buffered saline. Then, the sections were incubated with primary and secondary antibodies (Supplementary Table 4). All sections were counterstained with Mayer’s hematoxylin and quantified (at least 10 fields for each sample) using ImageJ software.

Immunoblot analysis

Frozen tissues from each mouse were homogenized in lysis buffer containing 50 mM Tris at pH 7.4, 1 mM EDTA, 1% IGEPAL CA-630, 150 mM NaCl, 0.10% Triton, 50 mM NaF, 100 mM phenylmethanesulfonyl fluoride and 1 mM Na₃VO₄. Protein concentration was determined, and the samples were denatured in Laemmli buffer with β-mercaptoethanol at 100° C for 5 min, and 50 μg was loaded per lane onto 8%-14% sodium dodecyl sulfate-polyacrylamide gels. The gels were then transferred to polyvinylidene difluoride or nitrocellulose membranes, blocked with 5% nonfat dry milk or bovine serum albumin in 20 mM Tris at pH 7.4, 150 mM NaCl and 1% Tween-20, and incubated with the respective antibodies (Supplementary Table 4). Immunoreactive bands were developed with a SuperSignal West Femto chemiluminescent substrate (Pierce, Rockford, IL, USA) in a ChemiDoc system (Bio-Rad Laboratories, Madrid, Spain). The bands were quantified using Image Lab 2.0 software (Bio-Rad Laboratories, Hercules, CA, USA).

Quantitative metabolomics platform and chromatographic procedures

Targeted metabolomics analysis was performed as reported with some modifications [48]. To limit variability, batch analysis was required, and the number of samples measured per day was restricted to n = 24 (n = 8 for each strain). To measure the metabolites of energy metabolism, the tissues were homogenized using a Precellys 24 homogenizer (Izasa, Barcelona, Spain) in 1 mL of methanol/water (8:2) and D4-succinic acid and stored at -20° C for 2 h to precipitate the proteins. After centrifugation at 15,000 rpm for 10 min at 4° C, the supernatants were dried under N₂ and derivatized with methoxyamine hydrochloride dissolved in pyridine (40 mg/mL) and N-methyl-N-trimethylsilyl trifluoroacetamide. We used a 7890A gas chromatograph coupled with an electron impact source to a 7200 quadrupole time-of-flight mass spectrometer (Agilent Technologies, Santa Clara, USA). For measuring metabolites of 1-C metabolism, tissues were added to 1 mL of an extraction solution containing methanol:water (8:2 v/v), 1% ascorbic acid (m/v) and 0.5% β-mercaptoethanol (v/v) and homogenized with a Precellys 24 homogenizer (Izasa). After protein precipitation, the samples were centrifuged, and the supernatants were dried under N₂ and then resuspended in ultrapure water containing 50 mM ammonium acetate and 0.2% formic acid. The analysis was performed with an ultra-high pressure liquid chromatography-quadrupole time-of-flight mass spectrometer (Agilent Technologies). Liquid chromatography was also used in separate analyses to quantitate guanine, 5-methylcytosine (5-mC) and 5-hydroxymethylcytosine (5-hmC) as described [49].

Statistical analyses

The statistical comparisons were performed using nonparametric Mann-Whitney U tests with IBM SPSS Statistics for Windows, Version 21 (Armonk, NY, USA, IBM Corp.). Metabolites were quantitated in all the samples using MassHunter Quantitative Analysis B.07.00 software (Agilent Technologies) according to the calibration curve of the corresponding standard. Partial least squares-discriminant analysis (PLS-DA), hierarchical clustering and the variable importance in the projection score were calculated with MetaboAnalyst 4.0 (http://www.metaboanalyst.ca) software [50]. GraphPad Prism 6.01 software (GraphPad Software, San Diego, CA) was also used for relevant statistical confirmation and graphics. Differences were considered statistically significant when the p value was ≤ 0.05. Unless otherwise indicated, the results are expressed as the means ± standard error of the mean.