Circadian Mutant Mice with Obesity and Metabolic Syndrome are Resilient to Cardiovascular Disease

21 Obesity and metabolic syndrome commonly underly cardiovascular disease. ClockΔ19/Δ19 mice 22 fed a normal diet develop obesity and metabolic syndrome; however, it is not known whether they 23 develop or are resilient to cardiovascular disease. We found that ClockΔ19/Δ19 mice do not develop 24 cardiac dysfunction, despite their underlying conditions. Moreover, in contrast to wildtype controls 25 fed a high-fat diet (HFD), ClockΔ19/Δ19 HFD mice still do not develop cardiovascular disease. 26 Indeed, ClockΔ19/Δ19 HFD mice have preserved heart weight despite their obesity, no 27 cardiomyocyte hypertrophy, and preserved heart structure and function, even after 24-weeks of 28 a HFD. To determine why ClockΔ19/Δ19 mice are resilient to cardiac dysfunction despite their 29 underlying obesity and metabolic conditions, we examined global cardiac gene expression 30 profiles by microarray and bioinformatics analyses, revealing that oxidative stress pathways were 31 involved. We examined the pathways in further detail and found 1) SIRT-dependant oxidative 32 stress pathways were not directly involved in resilience. 2) Increased 4-hydroxynonenal (4-HNE) 33 in wildtype HFD but not ClockΔ19/Δ19 mice, suggesting less reactive oxygen species in ClockΔ19/Δ19 34 mice. 3) Increased cardiac catalase (CAT) and glutathione peroxidase (GPx) suggesting strong 35 antioxidant defences in ClockΔ19/Δ19 hearts. 4) Upregulation of Pparγ in ClockΔ19/Δ19 hearts; this 36 circadian-regulated gene drives transcription of CAT and GPx, providing a molecular basis for 37 resilience in the ClockΔ19/Δ19 mice. These findings shed new light on the circadian regulation of 38 oxidative stress, and demonstrate an important role for the circadian mechanism in resilience to 39 cardiovascular disease. 40 41


INTRODUCTION
Interestingly, previous studies have suggested that the circadian mechanism gates diet related 80 cardiometabolic outcomes. For example, time-of-day restricted feeding or intermittent fasting (e.g. 81 circadian strategies) can improve cardiac metabolic homeostasis (28,55). Conversely, food 82 intake at the wrong time of day can exacerbate cardiometabolic dysfunction (9, 59). The latter 83 findings are especially a caveat for shift workers, who frequently eat meals during the night shift 84 (11), and for whom there is an increased risk of obesity, metabolic disorders, and cardiovascular 85 disease (25,56,64). Intriguingly, CLOCK mutant mice (Clock Δ19/Δ19 ) develop obesity and 86 metabolic syndrome (61), which are underlying risk factors for cardiovascular disease. However, 87 it's not known whether or not the Clock Δ19/Δ19 mice actually develop or are resilient to obesity 88 induced cardiovascular disease. 89 To investigate, we used circadian mutant Clock Δ19/Δ19 mice, which have profoundly disrupted 90 metabolic energy balance and significant body weight gain (61). We found that despite this 91 underlying phenotype, the Clock Δ19/Δ19 mice did not develop cardiac dysfunction. We next used a 92 high-fat diet (HFD) to see if this in combination would precipitate the development of heart 93 disease. As expected in wild type (WT) controls, the HFD led to obesity, metabolic syndrome, and 94 cardiac dysfunction with cardiac hypertrophy and compensatory adverse structural and functional 95 remodeling. However, in contrast, the Clock Δ19/Δ19 mice fed a HFD remained resilient to 96 cardiovascular disease. At a molecular level, we found that the Clock Δ19/Δ19 mice had reduced 97 oxidative stress and increased antioxidant responses under circadian regulatory control, which 98 help explain their better cardiac outcomes. Thus, these findings reveal that Clock Δ19/Δ19 mice do 99 not develop cardiac remodeling and contractile dysfunction despite the underlying obesity and 100 metabolic disorder, and they continue to be resilient even when fed a HFD. These studies shed 101 new light on CLOCK, identifying it as a target that mediates resilience to obesity or HFD induced 102 cardiovascular disease. 103 After 24 weeks of HFD or SC, mice were fasted for 6 h, and blood was collected at zeitgeber 130 time (ZT) 0. For glucose, cholesterol, and triglyceride measurements, nonterminal blood collection 131 was performed from manually restrained, non-anesthetized mice via the saphenous vein. Fasted 132 blood glucose levels were measured from a drop of blood from the saphenous vein using a hand-133 held glucometer (Freestyle Lite, Abbott). Cholesterol and triglyceride levels were assessed from 134 ~200 µl of blood collected from the saphenous vein into a microvette capillary tube with clotting 135 activator (Sarstedt), clotted on ice for 1 h, centrifuged at 10,000xg for 5 min at room temperature, 136 and serum was aliquoted and stored at -80°C until use. From these serum samples, triglyceride 137 levels were determined using the IDEXX Rodent Lipid Panel (IDEXX BioAnalytics). Serum 138 cholesterol levels were measured using a commercially available kit, according to the 139 manufacturer's instructions (Cholesterol E kit, Wako Diagnostic). 27 mice were used, n=6-7 140 mice/group. For non-fasted insulin measurements, another set of Clock Δ19/Δ19 and WT mice on a 141 standard ad libitum diet were used. The mice were anesthetized with 4% isoflurane and 142 euthanized at 4 h intervals over one 24 h period (ZT03,07,11,15,19,23). Approximately 1 mL 143 of blood was collected at each timepoint, via cardiac puncture into EDTA-treated microcentrifuge 144 tubes (Sarstedt), and centrifuged at 1,500xg for 10 min at 4°C, and plasma was pooled and stored 145 at -80°C until use. Insulin levels were determined by ELISA (Crystal Chem), according to the 146 manufacturer's instructions, as described (61). 36 mice were used, n=18 mice per genotype, n=3 147 mice/timepoint. 148 149

Morphometry and Histology 150
Clock Δ19/Δ19 and WT mice fed a HFD or SC for 24 weeks were euthanized with 4% isoflurane 151 and cervical dislocation at ZT07. Upon sacrifice, body weight (BW), heart weight (HW), epididymal 152 white adipose tissue weight (eWAT), and tibia length (TL), were collected from each mouse. A 153 total of 27 mice were used, with n=6-7 mice/group. Hearts were collected for histopathology, as 154 previously described (1,3,49). Briefly, hearts were removed, perfused with 1 M KCl to arrest in 155 diastole, and fixed in 10% neutral buffered formalin for 24 h. Formalin fixed hearts were 156 processed, embedded, and 5 μm sections were collected at the mid-papillary level. Sections were 157 stained with Masson's trichrome for quantification of myocyte cross sectional area (MCSA) from 158 at least 100 cardiomyocytes/heart, over at least 3 sections, with n=3 hearts/group. Images were 159 taken using Q-Capture (QImaging) and analyzed in Image J 1.46 (NIH). 160 161 Echocardiography 162 At baseline (8 weeks of age, SC) and after 4, 8, 12, 16, and 24 weeks of HFD or SC, cardiac 163 structure and function were assessed under light anesthesia (1.5% isoflurane) using a GE Vivid 164 7 Dimension ultrasound machine (GE Medical Systems) with a i13L 14MHz linear-array 165 transducer, as described previously (3,17,49). All echocardiography assessments were 166 performed between ZT07 -ZT09. Measurements were taken at the mid-papillary level from at 167 least 5 images per mouse. End diastolic (EDV) and systolic (ESV) volumes were calculated using 168 the cube formula, stroke volume (SV) was calculated as EDV-ESV, and cardiac output (CO) was 169 calculated as SV x heart rate (HR). A total of 24 mice were used, n=6 mice/group. 170 171

In vivo hemodynamics 172
At the 24 week endpoint, in vivo hemodynamics measurements were collected in animals 173 anesthetized with 4% isoflurane, intubated, and ventilated (Harvard Apparatus model 687), using 174 our described methods (1-3, 5, 49). The right carotid artery was isolated and a 1.2Fr pressure 175 catheter (Transonic) was advanced through the ascending aorta into the left ventricle (LV). In vivo 176 LV and aortic pressure measurements were recorded with ADInstruments PowerLab and 177 analyzed using Lab Chart 7 (Colorado Creeks). Following hemodynamic recordings, mice were 178 euthanized with 4% isoflurane and cervical dislocation. 24 mice were used, n=6 mice/group. 179

RNA isolation, microarray, and bioinformatics analyses 181
Clock Δ19/Δ19 and WT mice fed a HFD or SC for 24 weeks were euthanized with 4% isoflurane 182 and cervical dislocation at ZT07. Hearts were collected, snap frozen in liquid nitrogen, and stored 183 at -80°C until use. RNA isolation, microarray, and bioinformatics analyses were performed as 184 described previously (6,33,36,60). Briefly, total RNA was isolated from hearts using TRIZOL 185 (Invitrogen). RNA quantity and quality were assessed by Nanodrop (Thermo Scientific) and 186 Agilent 2100 Bioanalyzer (Agilent Technologies Inc.

Statistics 227
All values are mean ± SEM. Statistical comparisons were made using an unpaired Student's 228 t-test or two-way analysis of variance (ANOVA) followed by Tukey post-hoc test for multiple 229 comparisons, as applicable. All analyses were performed using GraphPad Prism 8 (GraphPad First, we confirmed the adverse metabolic effects of circadian disruption in Clock Δ19/Δ19 mice, 236 so we could next investigate the effects on cardiac structure and function. To characterize the 237 phenotype of the Clock Δ19/Δ19 mice, we first examined 24 h locomotor running wheel activity under 238 standard L:D conditions (Fig. 1A). The Clock Δ19/Δ19 mice actigraphy was as expected, with less 239 activity in the dark phase as compared to WT controls (Fig. 1B). We also analyzed whole-body 240 metabolism by indirect calorimetry using CLAMS metabolic cages. Both genotypes exhibited the 241 anticipated diurnal rhythm in locomotor activity, however, the Clock Δ19/Δ19 mice showed 242 significantly blunted (P<0.001) wake time activity, along with increased (P<0.001) activity in the 243 light (sleep time) phase, as compared to WT controls (Fig. 1C). Moreover, Clock Δ19/Δ19 mice 244 showed a loss of diurnal feeding rhythms, with an average of 47% of food intake occurring in the 245 light phase, as compared to only 29% for WT mice during this time (P<0.001) (Fig. 1D). Similarly, 246 the Clock Δ19/Δ19 mice exhibited a significantly attenuated (P<0.05) daily rhythm in oxygen 247 consumption as compared to WT controls (Fig. 1E). The Clock Δ19/Δ19 mice also showed a loss of 248 rhythmic substrate utilization, as measured by respiratory exchange ratio (RER) (Fig. 1F) Next, we characterized the development of obesity in Clock Δ19/Δ19 and WT mice, in groups of 253 animals fed SC, and also in groups fed a HFD for 24 weeks. We found greater body weight in the 254 WT mice on the HFD ( Fig. 2A), increasing significantly (P<0.0001) over the 24-week period, as 255 compared to WT SC controls (Fig. 2B, C, Table 2). Increased body weight occurred more rapidly 256 in the Clock Δ19/Δ19 mice on the HFD (Fig. 2D), and persisted (P<0.0001) over 24 weeks, as 257 compared to Clock Δ19/Δ19 SC controls (Fig. 2E, F, Table 2). Moreover, the HFD fed mice had 258 increased (P<0.001) epididymal white adipose tissue (eWAT) weight, suggesting that the body 259 weight gain was likely due in part to an increase in visceral fat (Fig. 2G, Table 3). Interestingly, 260 overall daily caloric intake was similar for both genotypes (Fig. 2H), despite greater weight gain 261 in the Clock Δ19/Δ19 mice. We also found that HFD fed mice had a similar rise in serum cholesterol 262 levels regardless of genotype (Fig. 2i), however, only the WT mice showed elevated fasting 263 glucose levels under the HFD conditions (Fig. 2j). Together these findings confirm that at baseline 264 the Clock Δ19/Δ19 mice have obesity and metabolic dysfunction, but not WT mice, as anticipated. 265 Moreover, they show that both genotypes respond to a HFD with obesity and metabolic 266 dysfunction. 267 268 Clock Δ19/Δ19 mice with obesity and metabolic dysfunction do not develop cardiac 269

hypertrophy. 270
We next looked at whether the obesity and metabolic dysfunction in the Clock Δ19/Δ19 mice was 271 associated with the development of heart disease. We found that even though the Clock Δ19/Δ19 272 mice had greater heart weight (HW) than the WT mice at baseline (Fig. 3A), their HW:BW (Fig.  273 3B) and HW:TL ( Fig. 3C) ratios were proportionate, suggesting that any increase in HW was a 274 normal physiological response to increased BW. Furthermore, as shown in Figure 3A-C and 275 Table 3, even when fed a HFD, the Clock Δ19/Δ19 mice showed no significant increase in HW as 276 compared to Clock Δ19/Δ19 SC controls, despite the HFD induced BW gain. In contrast, the WT HFD 277 mice had a significant increase in HW disproportionate to their BW gain, suggestive of 278 pathological remodeling in this group. The adverse cardiac remodeling was also evident on 279 histological analyses, as the WT HFD mice exhibited cardiomyocyte hypertrophy (Fig. 3D, left) 280 and increased myocyte cross-sectional area (Fig. 3D,  Given that Clock Δ19/Δ19 mice showed resilience to cardiomyocyte hypertrophy, we next 288 examined whether this correlated with cardiac structure and function, by echocardiography. 289 Representative M-mode echocardiography images after 24 weeks of diet are shown in Figure  290 4A. The time-series data are illustrated in Figure 4B. We found that the Clock Δ19/Δ19 mice 291 maintained normal cardiac structure and function, consistent with the lack of pathological findings 292 in the heart (Fig. 4B, Table 2). Moreover, the Clock Δ19/Δ19 mice showed normal physiological 293 cardiac adaptations to a HFD, with increased end diastolic volume (EDV), end systolic volume 294 (ESV), stroke volume (SV), and cardiac output (CO), while maintaining normal cardiac function, 295 by 24 weeks (Fig. 4B, Table 2). In contrast, the WT HFD mice developed significant 296 pathophysiologic changes in structure and function by echocardiography, consistent with the 297 earlier findings of cardiac hypertrophy in these animals ( Fig. 4B, Table 2). The WT HFD hearts 298 also exhibited significantly impaired contractility by in vivo hemodynamics (Fig. 5A). Moreover, 299 although systolic function was preserved in the WT HFD mice (Fig. 5B), there was significant 300 (P<0.005) diastolic dysfunction indicated by an impaired relaxation rate (Fig. 5A), increased 301 LVEDP (Fig. 5C), and and an increase (P<0.005) in the LV diastolic time constant tau (Fig. 5D). 302 Together these data show that WT mice exhibit a number of cardiometabolic risk responses to a 303 HFD, which are associated with left ventricular remodeling and diastolic dysfunction. The 304 Clock Δ19/Δ19 mice also develop cardiometabolic risk profiles, on either SC or HFD, however 305 surprisingly, they are resilient to cardiac dysfunction. 306 307

Cardiac transcriptional analyses and oxidative stress pathways. 308
To investigate the underlying gene expression patterns driving adverse cardiac remodeling in 309 WT mice and, in parallel, the resilience observed in Clock Δ19/Δ19 mice, we performed genome-wide 310 microarray analysis. First, we examined transcriptional changes driven by HFD in WT hearts. A 311 total of 174 transcripts (≥1.35-fold change) showed differential expression in WT HFD versus WT 312 SC hearts (Fig. 6A Further Gene Ontology (GO) analysis revealed that the differentially regulated genes in WT HFD 314 hearts mapped to functional biological categories of stress, growth/remodeling, transcription, and 315 metabolism (Fig. 6B). In contrast, differential expression of these same cardiac remodeling genes 316 were not found in the hearts of the Clock Δ19/Δ19 mice, consistent with their resilience to 317 cardiovascular disease ( Table 4). Ontological mapping further revealed a role for the circadian 318 mechanism in driving these transcriptional responses (Fig. 6C anticipated blunted expression of core circadian mechanism genes (Fig. 6D), and KEGG analysis 321 revealed a link with the oxidative stress and antioxidant pathways (Supplemental Table 2 In terms of mechanism, we interrogated the oxidative stress gene data in more detail. Our 326 microarray data showed significantly increased nicotinamide phosphoribosyltransferase (Nampt) 327 expression only in WT HFD hearts, but not in WT SC, nor Clock Δ19/Δ19 SC nor Clock Δ19/Δ19 HFD 328 mice (Fig. 7A). Since NAMPT enhances susceptibility to oxidative stress via the sirtuin (SIRT) 329 dependent oxidative stress pathway, we next investigated this pathway. However, we found no 330 differences for any groups in the expression of downstream Sirt1, Sirt2, Sirt3, Sirt4, and Sirt6 331 genes (Fig. 7B), nor for downstream Foxo1 and Foxo3 genes (Fig. 7C), nor did we observe 332 differences in the abundance of the downstream antioxidant protein manganese-dependent 333 superoxide dismutase (MnSOD) which mitigates oxidative stress (Fig. 7D). Thus, while WT HFD 334 are susceptible to cardiac dysfunction, and Clock Δ19/Δ19 mice are resilient, that protection does not 335 appear to be mediated by changes in the overall gene and protein expression of the SIRT-336 dependent oxidative stress pathways. 337

H 2 O 2 -dependent antioxidant signaling. 339
Next, we investigated the H 2 O 2 -dependent antioxidant signaling pathways. First, we found that 340 4-hydroxynonenal (4-HNE) was increased only in the WT HFD mice (Fig. 8A), but not in the 341 Clock Δ19/Δ19 mice (Fig. 8B), suggesting increased activation of oxidative stress driving 342 cardiovascular disease in the WT mice. Second, we found that catalase (CAT) protein levels were 343 increased in both the WT HFD heart (Fig. 8C) and Clock Δ19/Δ19 HFD heart (Fig. 8D), consistent 344 with there being activation of antioxidant pathways in response to HFD in both genotypes. Third, 345 we found that glutathione peroxidase (GPx) protein levels were increased only in the Clock Δ19/Δ19 346 HFD mice (Fig. 8E), suggesting better antioxidant protection in the hearts of these mice. Finally, 347 in order to better understand the molecular drivers, we next evaluated Pparγ mRNA levels, a 348 transcription factor that underlies GPx production. We found that that peroxisome proliferator-349 activated receptor gamma (Pparγ) was significantly (P<0.005) upregulated in Clock Δ19/Δ19 hearts 350 ( Fig. 8F), consistent with the finding that GPx is upregulated and protective. Taken together, these 351 data demonstrate that Clock Δ19/Δ19 mice are resilient to cardiovascular disease, even though they 352 have underlying obesity and metabolic syndrome, concurrent with reduced oxidative stress and 353 increased antioxidant protection in the heart (Fig. 9). 354

DISCUSSION 355
In this study, we demonstrate that Clock Δ19/Δ19 mice have obesity and metabolic syndrome -356 well known risk factors for cardiovascular disease -yet they do not develop cardiac dysfunction. 357 Moreover, even when fed a HFD that normally precipitates obesity, metabolic syndrome and 358 cardiovascular disease (as in WT mice) the Clock Δ19/Δ19 mice continue to be resilient to heart 359 disease. That is, WT mice fed a HFD develop cardiac hypertrophy with compensatory cardiac 360 remodeling and diastolic dysfunction, but in contrast the Clock Δ19/Δ19 HFD mice have cardiac 361 physiology similar to healthy controls. We used microarrays and bioinformatics analyses to 362 investigate underlying mechanisms for resilience. We found that SIRT-dependant oxidative stress 363 pathways did not appear to be directly involved in resilience. However, the H 2 O 2 -dependant 364 pathways involving 4-HNE, CAT, and GPx revealed that the Clock Δ19/Δ19 hearts had reduced 365 oxidative stress and better antioxidant responses. Both genetics and lifestyle contribute to serious 366 chronic health conditions such as obesity, metabolic syndrome and the subsequent development 367 of cardiovascular disease. Notably, these findings shed new light on how the circadian 368 mechanism is an important player in mediating resilience to the cardiovascular disease outcomes. 369 One of the important foundations of our study is that Clock Δ19/Δ19 mice have underlying obesity 370 and metabolic syndrome. This is consistent with earlier studies that showed that the circadian 371 clock is an important regulator of mammalian energy balance, and that disruptions to the circadian 372 mechanism can impair metabolic homeostasis (61). However, even though the Clock Δ19/Δ19 mice 373 have obesity and metabolic syndrome, which are risk factors for cardiovascular disease, they do 374 not develop heart disease. Moreover, even when challenged with a HFD the Clock Δ19/Δ19 mice 375 exhibit normal physiologic cardiac adaptations associated with obesity, yet are resilient to 376 pathological cardiac remodeling and contractile dysfunction, in contrast to their WT littermates. 377 Thus, these studies shed new light on a role for the circadian mechanism, that is, as a factor that 378 mediates resilience to cardiovascular disease. 379 An intriguing outcome of this study, however, is that by revealing a cardioprotective role in 380 Clock mutant mice, our findings appear counter-intuitive to earlier reports that an intact circadian 381 mechanism is needed to benefit the heart (e.g. (1, 3, 8, 17-19, 36, 39, 68)). Collectively, the 382 message has long been that maintaining circadian rhythms promotes heart health, and disruption 383 causes or exacerbates disease. Why then are the circadian mutant mice resilient to heart 384 disease? In this study, the protective outcomes relate to the underlying condition. That is, diet, 385 obesity, and metabolic dysfunction trigger adverse oxidative stress pathways that are modifiable 386 by products of the circadian mechanism. Previous studies also strongly support this notion, and 387 our findings, that cardiac remodeling can be improved on even in the absence of an intact 388 circadian mechanism; the circadian mechanism remains fundamentally important because the 389 genes and proteins that drive the outcome are under circadian mechanism transcriptional control. circadian mechanism transcriptional outputs can change with aging, and sex hormones, and thus 396 resilience may change to susceptibility in old Clock Δ19/Δ19 mice (1, 2). 4) Disruption of CLOCK is 397 cardioprotective in the myocardial infarction ischemia reperfusion (mI/R) model, with reduced 398 infarct size in CCM mice versus WT controls (18). 5) Also, pharmacologically targeting the 399 circadian mechanism with the repressive REV-ERB agonist SR9009, temporarily holding back 400 the clock, improves outcomes post-mI/R in mice (49). Importantly suspending the circadian 401 mechanism in a manner in which it can mitigate outcomes is driven by the changes in output 402 genes and proteins under its regulatory control, including outputs involved in cardiac remodeling. 403 That is, perhaps what we think of as desynchrony should not be so much about the circadian 404 clock being "broken" -but rather we should consider how it works differently, and how those 405 changes in controlled output genes can improve outcome. Thus, the message remains that the 406 circadian mechanism is important for cardiovascular health, but one must also consider the 407 circadian mechanism regulated outputs that are specific to the disease process, and that those 408 can directly influence outcome. 409 Mechanistically, in this study, resilience of the Clock Δ19/Δ19 mice was mediated through 410 oxidative stress pathways. We found that Clock Δ19/Δ19 mice have less 4-HNE response to HFD, 411 as compared to the WT mice, and consistent with their better outcomes. 4-HNE is a product of 412 lipid peroxidation with well-known adverse oxidative stress responses in vitro (13, 62), and 413 elevated 4-HNE corresponds with adverse cardiac remodeling in human heart failure (42). We 414 also found greater antioxidant (CAT, GPx) responses in the Clock Δ19/Δ19 HFD hearts, as compared 415 to WTs, and consistent with their resilience to HFD induced cardiovascular disease. These 416 antioxidant enzymes are involved in detoxification of H 2 O 2 , with cardioprotective benefits as 417 shown in experimental heart failure models in vivo (24, 47). Importantly, the increased antioxidant 418 protection may be driven by its transcription regulator PPARγ (12, 14, 21), which is a circadian 419 mechanism regulated product (67). The Clock Δ19/Δ19 hearts have increased Pparγ expression, thus 420 providing a molecular explanation for the improved outcomes in these animals. 421 It has long been known that oxidative stress pathways underlie cardiac dysfunction, and that 422 strategies aimed at reducing damage could be beneficial (43). However it is only recently that 423 these oxidative stress pathways have been linked to the circadian mechanism in the 424 cardiovascular system (16,23,27,44,52). Together, our findings demonstrate that the circadian 425 mechanism drives oxidative stress and antioxidant pathways and in doing so it acts as a driver 426 modulating resilience to cardiovascular disease. 427 428

Conclusions 429
In this study we show that Clock Δ19/Δ19 mice develop obesity and metabolic syndrome, yet 430 remarkably they are protected from cardiac dysfunction, even when fed a HFD. Resilience 431 appears to be mediated by down regulating adverse oxidative stress pathways and up regulating 432 beneficial antioxidant pathways that are under circadian regulatory control. It is worth noting that 433 these studies were done in mice, which raises some caveats with regards to translation to 434 humans. Interestingly though, our findings reflect the common observation that although many 435 individuals in contemporary society have obesity and metabolic syndrome, only some will develop 436 cardiovascular disease. Additional studies in humans also support the notion that the circadian 437 mechanism mediates resilience to cardiovascular disease. For example, minor allele carriers for  Darley-Usmar V, Zhang J, Chatham JC, and Young ME. Genetic disruption of the 590 cardiomyocyte circadian clock differentially influences insulin-mediated processes in the heart. J Clock Δ19/Δ19 hearts (fold change ≥ 1.35; ZT07; see Table 4 and Supplemental Table S1,   increased abundance in WT HFD heart and liver by 24 weeks, but B) not in Clock Δ19/Δ19 mice, 742 whereas C) CAT is increased in both WT and D) Clock Δ19/Δ19 hearts in response to HFD, but E) 743