| |

Inside Scientific Webinar: Comfort Food: Effects of Stress and High-Fat Diets on Neuronal Activity and Mitochondrial Remodeling in Mice

Julio Ayala, PhD and Matthew Robinson, PhD discuss their research focusing on high-fat feeding behavior in mice and the effects of stress and exercise on metabolism and obesity.

To watch the full archived webinar, click here.

VIDEO NOT LOADING? Viewing videos requires acceptance of all cookies. To change consent settings, open the MANAGE CONSENT tab at the bottom right of this window and click ACCEPT.

Sarah: Good morning, good afternoon, and good evening, everyone, and welcome to our webinar titled “Comfort Food, Effects of Stress and High Fat Diets on Neuronal Activity and Mitochondrial Remodeling in Mice.” This webinar has been sponsored by Sable Systems International, so a big thank you to them for helping to make this event possible.

Joining us today, we’re fortunate to have Dr. Julio Ayala, an associate professor at Vanderbilt University, and Dr. Matt Robinson, an assistant professor at Oregon State University. Their presentations will discuss their research focusing on high-fat feeding behaviors in mice and the effects of stress and exercise on metabolism and obesity. I’m Sarah McFarland from the events team here at Inside Scientific, and I’m very pleased to be your host for today’s event. And with that, I am pleased to welcome our first presenter, Dr. Julio Ayala. Julio, thank you so much for being here with us today, and the floor is yours whenever you’re ready.

Dr. Ayala: Excellent, well, thank you, Sarah, and good morning, good afternoon, good evening to all of you attending. Thank you for coming to this webinar. So, as Sarah mentioned, my title implies I will be discussing work from our lab, looking at the effects of obesity on the response to stress with a particular emphasis on how obesity affects stress-associated phenotypes that can ultimately influence body weight and our primary focus is really looking at feeding behavior. I will then introduce some preliminary data exploring a mechanism that we hypothesize is behind some of the effects that stress has on feeding behavior and how this mechanism may actually be affected in the setting of obesity to influence the effect of stress on feeding.

So, why the interest in obesity and stress? Well, I think it should be no surprise to anyone listening to this presentation that we have been in the midst of an obesity epidemic in the United States and worldwide for quite some time. And at least in the U.S. this is exemplified by this map that we have seen many, many times showing the dramatic rise in the percentage of the population in each state of the U.S. that can be considered obese as defined by a body mass index or BMI of over 30. What is also not likely to be surprising to anyone here is that stress is also on the rise in the US and worldwide. And this has been particularly true over the past year, not only because of the effects of the COVID pandemic, but also because of political and social unrest, both domestically and abroad.

Now, while there are plenty of rigorous studies supporting this assertion, I just want to illustrate the point using this Gallup poll that has been taken routinely over the past few years that shows how self-reported feelings of stress in the U.S. have been consistently rising over the past 15 years or so. And based on these non-scientific findings, about eight in 10 Americans report feeling either moderate stress or severe stress on a daily basis. Now, of course, just because two phenomena that is, stress and obesity are rising in parallel, does not prove a direct relationship. However, it’s well documented that stress can affect behaviors that can either lead to weight gain over time or exacerbate weight gain in already obese individuals, as I will highlight here.

We know that stress can evoke a variety of responses with different degrees of magnitude and directionality in different individuals. That being said, one relatively common effect that stress can have on individuals is an alteration in feeding behavior. Now what is particularly interesting about the effect of stress on feeding behavior, it’s that it tends to be fairly unidirectional in that the vast majority of individuals, typically between 60 and 75% of individuals, depending on the study, report that they increase their caloric intake in response to stress. And this increase in caloric intake tends to occur in the form of increased snacking, as shown highlighted in this box. And this is fairly consistent in both men and women. Unfortunately, and probably not surprisingly, these snacks are not of the healthy variety, such as fruits and vegetables, but they tend to be in the form of foods and beverages that are high in sugar and fat, as shown here. And this has given rise to the terms comfort food or comfort feeding, which refers to the increased consumption of palatable, highly caloric substances based on the rewarding and soothing properties mitigating the negative feelings associated with stress.

Now, interestingly, we can actually mimic this comfort feeding behavior in pre-clinical models, such as rats and mice. Now in general, although not exclusively, the typical response in rodents to a stressor is a decrease in food intake or hypophagia. This is a term I will be using multiple times throughout this presentation. And this hypophagia is probably a defense mechanism for rodents to shift away from eating when they encounter a stressor, which in the wild would typically be something like a predator. However, what many studies show is that when rodents are presented with a palatable substance such as sucrose or a high-fat dye, the ability of stress to reduce food intake is either attenuated or completely lost. And that is shown in these studies. The study on the left showing that when rats are presented with sugar, the effect of restraint stress to reduce body weight is attenuated. And the studies on the right in mice show that the effect of this chronic social defeat stress to reduce food intake is also attenuated when they are presented with a high fat diet. Now not only does the introduction of stress actually attenuate the typical, I’m sorry, the introduction of palatable substance, not only does that attenuate the typical stress induced hypophagia, but it can actually enhance the drive and preference towards the consumption of palatable substances, even beyond the normal tendency to consume these substances in the unstressed state.

So, this is a fairly reasonable modeling of what we call comfort feeding in humans. But interestingly, the vast majority of studies exploring this behavior and the mechanisms behind it in rodents have actually been conducted in lean rodents, and these palatable substances are typically presented either shortly before the introduction of a stressor or at the same time as the stressor is initiated. And while this has led to important discoveries on how stress may promote obesity, much less is known about stress-related behaviors and mechanisms that are relevant once obesity has already set in and from my perspective, the importance of knowing the effects of stress in the already obese state is highlighted in this slide.

So, these are two separate studies assessing emotional eating, quote unquote, emotional eating behavior in response to stress as it relates to BMI. And what you can see in the study on the left is that emotional eating here defined by the term coping on the x-axis, is not only proportional to BMI on the y-axis, but the slope is significantly more positive in overweight and obese individuals shown in the black circles. Now the study on the right is more topical to current events, since it assessed emotional eating in response to the stress of social isolation that has occurred during the COVID pandemic. And similar to the study on the left, this study showed a strong effect of higher BMI being associated with maladaptive coping and increased emotional eating. So, the take-home message of these studies and others is that obese individuals are more susceptible to stress and more prone to increase their caloric intake in response to stress.

So, our goal for the studies that I’ll be sharing today is to assess the effect of stress on energy balance parameters that can ultimately affect body weight in lean versus obese mice, and then to explore potential mechanisms that could explain these difference or possible differences in the metabolic response to stress in lean and obese mice. So, that’s how I’ve broken up my talk for today into these two parts. What are the effects of obesity on the response to stress focusing on energy balance parameters and what mechanisms may regulate feeding in response to stress and how are these potentially affected by obesity.

So, we first asked a very simple question, what effects does obesity have, if any, on the response to stress with a specific emphasis on parameters that control body weight such as food intake and energy expenditure. For that, we took a fairly straightforward approach. We placed C57 black six mice on a 60% very high fat diet for several weeks to make them obese and compare them to lean chow fed mice. To induce stress, we use a well-established stress modality in rodents by placing them in a restrainer for one hour. And these stress sessions were conducted during the hour before dark onset, which is when mice tend to consume most of their food. And then to measure energy balance and other parameters, we conducted these experiments in a Promethion instrument, which measures food and water intake, energy expenditure, substrate oxidation, and activity amongst other parameters. Now all the data that I will show today were obtained in male mice. We’re still conducting studies and parsing through the data in female mice to see if we not only see sex-specific differences, but also differences according to the stage in the estrus cycle, because there are reports in the literature suggesting that these factors can actually affect the response to stress.

So, now going on into the data. When we look at the effect of one-hour restraint stress in lean chow fed mice and these food intake measurements were begun right at the onset of the dark cycle, so right at the end of the stress session, you can see the expected hypophagia that is induced by restraint stress. Now interestingly there is a delay of about four hours before you see the effect of stress on food intake, the hypophagia, which is actually in line with previous observations in the literature. Now when we subject obese mice to the same stress protocol, you can see that obesity is characterized by a resistance to the hypophagic effects of stress. So, obese mice do not exhibit the expected hypophagia in response to restraint stress. What about other parameters? So, here we’re looking at energy expenditure in RQ, and this really highlights the advantage of performing these experiments in the Promethion in that we’re able to determine what happens to these parameters, such as energy expenditure and RQ, which is a marker of substrate oxidation, not only after the mice are released from the restrainer, but actually during the restraint session itself. And as you can see, if my cursor would work, here we go. As you can see here in the shaded box, this actually represents the one-hour restraint period. What you see is that both lean and obese mice display a relatively equivalent increase in energy expenditure pretty much throughout the entire restraint session. But they recover fairly quickly once they are released from the restraint. This rise in energy expenditure during the stress session is accompanied by a sharp drop in the respiratory quotient, or RQ, which is seen in the graphs below, which is the ratio between carbon dioxide expired and oxygen consumed by the mice. And this provides us information about substrate oxidation, and RQ of 0.7 is indicative of fat oxidation, whereas the RQ of 1 indicates carbohydrate oxidation. And therefore, the drop in RQ towards 0.7 in both lean and obese mice during stress suggests an increased reliance on fat oxidation. Now, interestingly, as you can see in the graphs on the bottom, this lower RQ persists for quite some time after the stressor is over and doesn’t really catch up in lean mice until the beginning of the next light cycle at around 6 a.m. So, since lower RQ can be indicative of decreased food intake and a reliance on fat stores when the mice aren’t feeding for fuel, and since stress causes a decrease in food intake in lean mice, we lined up our food intake and RQ measurements to see if this lower RQ in lean mice corresponded to the stress-induced hypophagia. But what you can clearly see here is that, on the graph on the left, is that the lower RQ in stressed mice was observed even though their food intake had not yet decreased. So, this reliance on fat oxidation appears to be a direct effect of stress itself and not secondary to an effect on food intake. This is also supported by the data in obese mice on the right which also display a decreased RQ even though stress had absolutely no effect on food intake.

So, we looked at other parameters that are available through the Promethion measurements, such as locomotor activity and water intake. And what you see here is that locomotor activity was not significantly affected by restraint stress in either a lean or obese mice, and neither was water intake, which is a little surprising given the fact that mice tend to match their intake of water to their intake of food.

So, to summarize what I’ve shown is that the main difference we observe between lean obese mice with regards to the response to acute restraint stress is that lean mice reduce their food intake as expected in response to stress, but obese mice are completely resistant to this effect of stress. And so now what we’re interested in is exploring what are the mechanisms that drive this difference in the stress response, and that’s what I’ll be describing some of the work we’ve done recently in this area. So, my lab has long been interested in metabolic actions of this hormone, glucagon-like peptide-1 or GLP-1, which was originally identified as a hormone secreted by the gut in response to food intake that engages a GLP-1 receptor in pancreatic beta cells to stimulate insulin secretion. Well, as it turns out that there is also a population of neurons in the hindbrain that produces GLP-1, and the GLP-1 receptor is widely expressed throughout the central nervous system in many areas that regulate feeding behavior. And the main outcome of activating either brain GLP-1 production or brain GLP-1 receptors is hypophagia or a decrease in food intake, which is what we typically observe in response to stress. Now role for brain GLP-1 action in response to stress has been shown by several groups as will be detailed in the slide here. So, the GLP-1 producing neurons in the brain, called NTS PPG neurons, extend projections to various GLP-1 receptor-expressing brain regions that have been associated not only with stress responses, but also with promoting hypophagia, such as the paraventricular nucleus or PVN, bed nucleus of the stria terminalis, BNST, and the lateral septum or LS.

Now, elegant work by Marie Holt and Stefan Trapp in the UK has shown that the hypophagia that occurs in lean mice in response to acute restraint stress is blocked by inhibiting NTSPPG neurons. And that’s shown in this graph here. So, as you can see on the left panel, in response to acute restraint stress, food intake goes down. But when the NTSPPG neurons are chemo-genetically silenced, that hypophagia is no longer observed, suggesting that activation of these neurons is necessary for the typical response stress in lean mice, which is hypophagia.

Similarly, work done in James Herman’s group at Cincinnati and Diana Williams’ group at Florida State have shown that blocking the GLP-1 receptor in the PVN, the BNST, or the LS also blocks the reduction in caloric intake that is seen in response to stress in lean mice. So overall, what these studies show is that stress activates GLP-1 action in the brain and that this is required for the expected hypophagia that is observed in response to stress. Now I’ll highlight that all of these studies have been done in lean rodents. So, we’re interested in determining whether this system is impaired in obese mice and whether this in turn explains why obese mice are resistant to stress-induced hypophagia. And our work thus far has focused on the lateral septum because it is perhaps the least studied of the three brain regions that I’ve highlighted here with regards to effects of GLP-1. As you can see, using two different reporter mouse lines in our group, the GLP-1 receptor is densely expressed in the lateral septum, particularly in the dorsal lateral septum.

Now, recent work by Sarah Stern has shown that either activation of GLP-1 receptor signaling in the LS, as shown on the left graph here by targeting the GLP-1 receptor agonist exendin-4 or EX-4 to the lateral septum, or activation of GLP-1 receptor expressing neurons in the LS using chemo-genetic approaches or DREDS, as shown on the right, resulted in hypophagia or reduced food intake, as shown here. And this was really exciting to us because we had actually, during this time, have been investigating the effects of stress on the activity of GLP-1 receptor-expressing neurons in the lateral septum in lean mice and whether this activity is affected in obese mice. And to do that, we have been using genetically encoded calcium sensors, or GCaMPs, which convert the rise in intracellular calcium that occurs when neurons are stimulated into a fluorescent signal that we can then detect in real time. The advantage of this GCaMP is that it is CRE dependent, so we can express it specifically in GLP1 receptor expressing neurons in the lateral septum by targeting an adeno-associated virus expressing this GCaMP to the lateral septum in GLP-1 receptor CRE mice in our colony. Once expressed, we implant a fiber optic probe to target the LS and then we can measure neuronal activity in real time using a fiber photometry rig.

So, our protocol is once we’ve expressed this GCaMP and implanted the fiber optic probe targeting the LS is to train mice to being handled for several days before the test date. And then on the test day, they’re acclimated to the fiber optic patch cord, as shown here. Then we take 10 minutes of baseline readings, followed by readings in both mice that remain unstressed or that are put in a restrainer for 60 minutes. And then this is followed by 10 minutes of post-restraint measurements. And to show you what a typical trace looks like, it’s shown here. So, this is a trace of a mouse that was not restrained. And the big spike on the left is actually indicative of increased neural activity when the mouse is picked up by the tail. And we use this actually to timestamp the beginning of all of our tests. And as many of you know, picking a mouse up by the tail is stressful to the mouse. Now, when we compare the trace of a mouse that is not stressed, as shown here on the left, with a mouse that has been restrained for 60 minutes, you can clearly see an increase in the frequency of spikes indicative of increased neuronal activity specifically in GLP-1 receptor expressing neurons in the lateral septum. And so, we’ve now performed a variety of different, we performed these analyses in different mice and I’m just showing you some representative traces. You can clearly see here that when lean mice are undergoing restraint stress, they again show a significant number of spikes during the restraint stress period. But now when we look at the traces in obese mice, what you can clearly see is a significant reduction, not only in the frequency of the spikes, but also in the amplitude of the spikes.

So, these results suggest that obesity is associated with an impairment in the activation of GLP-1 receptor-expressing neurons in the lateral septum in response to stress. This is important, given the previously published observations that I showed, previous slides that the activation of these neurons promotes hypophagia. So, we speculate that the impaired activation of these neurons in obese mice directly contributes to their resistance to the effects of stress on caloric intake, to their attenuated hypophagia, and we’re directly testing that hypothesis right now.

Now, as we drill deeper into the mechanisms that could potentially dampen the activity of GLP-1 receptor-expressing neurons in the lateral septum, focusing on the cannabinoid receptor agonist 2-arachidonoyl glycerol or 2-AG because 2-AG primarily acts on presynaptic cannabinoid receptors to inhibit postsynaptic neuronal activity. And this may explain, and if we see changes in 2-AG that would correspond to an inhibitory drive on LS GLP-1 receptor-expressing neurons, this could explain the effect that we see in obesity. Importantly, 2-AG opposes many of the effects of GLP-1. For example, it is known to stimulate caloric intake, which is the opposite effect of what GLP-1 does in the brain. Interestingly, when we compare 2-AG levels in the LS in lean and obese mice, we see that these levels are significantly elevated in obese mice, which could explain both why neuronal activity is dampened in response to stress in obese mice and why obese mice are resistant to stress-induced hypophagia. Finally, we also show that the GLP-1 receptor agonists, like Exendin-4, reduce levels of 2-AG in the LS. So, this proposes a model whereby GLP-1 receptor activation in the LS in lean mice reduces 2-AG levels and therefore relieves the inhibitory effect of 2-AG on neuronal activity, allowing for LS neurons to activate in response to stress and promoting hypophagia. But what this also suggests is the possibility that elevated 2-AG levels in obese mice are actually a marker of GLP-1 resistance, if you will, that allows for LS neurons to remain inhibited, or at least their activity not to become so active, and therefore stimulating food intake even in the presence of stress.

So, to summarize what I’ve shown is that obesity is characterized by reduced activation of GOP1 receptor expressing Ls neurons in response to stress as well as attenuated stress induced hypophagia and we are currently attempting to link these observations directly and are also exploring molecular mechanisms such as 2-AG that may explain this link. I mentioned early on that stress tends to enhance the preference towards the consumption of palatable substances. And it turns out that activation of GLP-1 receptor signaling reduces the motivation towards the consumption of palatable substances, as shown here by Diana Williams’ group, which raises the possibility that defects in the system in obese mice may contribute to the increased preference towards fat and sugar consumption in response to stress. But we’re also interested in exploring whether these mechanisms are relevant in other brain regions, also whether they’re relevant with regards to other metabolic phenotypes, such as the effects on energy expenditure and substrate oxidation that we’ve observed, and also in response to other stressors, especially chronic stressors, since these are more relevant to the human experience.

And so one last piece of data that I will show is data that were obtained using repeated restraint stress. And what these data show is that repeated restraint stress progressively attenuates hypophagia in lean mice and actually stimulates overt hyperphagia in obese mice. Which is interesting since previous studies by Sachin Patel here at Vanderbilt, as shown on the right, show that in response to repeated restraint stress, the ability of stress to actually activate neurons in the lateral septum is mitigated. So, as you progress towards chronic stress, the activity of these LS neurons actually decreases. And this may explain why with repeated or chronic stress, you see attenuated hypophagia in lean mice and overt hyperphagia in obese mice. So now, interestingly, even in response to repeated restraint stress, other phenotypes, such as increased energy expenditure and fat oxidation during stress persist, even after a repeated session of restraint stress, suggesting that these phenotypes are potentially regulated by mechanisms other than LS-GLP1 receptor expressing neurons.

So, I will end with a plug for the Vanderbilt Mouse Metabolic Phenotyping Center, since we have recently acquired a Promethion system, the same Promethion system that was used for the studies I showed earlier in the talk. But this one allows for fiber optic cables to actually be fed into the metabolic chamber while you are taking measurements in the mouse, while the mouse is in that chamber. So, this will actually allow us in the end, you can see a picture of a run there on the right. And so we’re excited about this because this will allow us to combine fiber photometry type of experiments as well as optogenetic manipulations, or even feeding catheters into the mice to obtain blood samples while we’re doing a run in the Promethion. So again, this is a system that is now up and running at the Vanderbilt Mouse Metabolic Phenotyping Center.

So, with that, I’d just like to acknowledge the people who actually did the work. This was project spearheaded by Dr. Michelle Bales, a talented postdoc in the lab with significant assistance from Danny Winder’s group here at Vanderbilt, in particular Sam Centanni, a very talented research assistant professor who helped us with the fiber photometry studies. And of course, the MMPC, Merrygay and Louise, as well as funding from the NIH for the acquisition of the Promethion system that I just mentioned. So thank you again for your attention.

Sarah: Great, thanks so much for that fantastic presentation, Julio. I would love to welcome Matt to the floor. Matt, thank you so much for being here with us today, and the floor is yours whenever you’re ready.

Dr. Robinson: Well, thank you very much for the introduction, Sarah, and thank you for the opportunity to share some of our ongoing research from Oregon State University. I co-direct the Translational Research Metabolism Lab at OSU. I joined this about five years ago with my colleague, Dr. Sean Newsom, and together we pursue the development of obesity and how that impacts skeletal muscle physiology. What we’re interested in is what goes wrong, basically, with obesity on skeletal muscle, and how does exercise help restore some of those challenges that occur at skeletal muscle? We’ll be focusing on the mitochondria, and I’ll walk through different aspects of why we’re focusing on obesity and exercise.

Now, as Dr. Ayala discussed in very illuminating talk, so I appreciate that, we have a major obesity epidemic in the United States and worldwide. Now, obesity is a major risk factor for the development of insulin resistance and type 2 diabetes, such that at current rates about 10% of Americans have type 2 diabetes. If the sedentary factors and obesity development continue on, we’re looking at about one in three adults will have diabetes by 2050. And now at those rates, within this one-hour webinar, about 190 people will be newly diagnosed. So, this represents a major personal and public health burden. We’re trying to understand a bit more why this develops with obesity and how we can reverse it.

Now, why are we focusing on skeletal muscle? Well, there’s a lot of challenges that occur in skeletal muscle during sedentary lifestyles and the development of obesity. So, low exercise habits mean that we’re not burning off as many calories. With the additional nutrient intake, we have accumulation of fat mass, nutrient excess, and an increase in circulating insulin concentrations to compensate for high blood sugars. All of these are a major challenge on skeletal muscle, including the nutrient overload matched with low energy demands. So, the mitochondria are at the center of this. The mitochondria are our main site for substrate oxidation and energy production. So, if we have low demands on the energy system, along with these high nutrient pushes, that in essence results in what’s been described as a metabolic traffic jam. This concept has really been pushed forward by Dr. Dan Moyo at Duke with this idea of metabolic gridlock. And it’s a fantastic analogy. It’s just like our five o’clock traffic jams. We have a continuous onset of cars onto our roadway. And eventually what happens? Well, we have slowdown on the roadway, we have backups, we even have accidents, cars get broken down, and this leads to a general slowdown. Now this same concept can be applied to metabolism, that as we have these excess nutrients coming in, we can have then spillover or some aspects of metabolic dysfunction. Now at some level we can blame the road system and say, well if there’s a poor road system then that would make a person more susceptible and sure enough there are examples at which low mitochondria make a person more susceptible to the development of diabetes. Now it’s also pretty clear that a person can also develop diabetes without overt dysfunction. And so what we’re trying to understand is what is it about this excess nutrient that can push a person towards developing insulin resistance and type 2 diabetes?

So, we’re focusing on the mitochondria as our metabolic roadway. Now the mitochondria’s main function is oxidative metabolism through the electron transfer system. Now we have multiple on ramps onto this from different substrates. So, we can broadly classify those into lipid and non-lipid substrates. So, for our non-lipid substrates, these provide reducing equivalents that enter the electron transfer system through complex one, complex two, and also glycerol phosphate dehydrogenase. Now our interest for this talk is on lipid oxidation. And that’s because lipid oxidation is really tied to the development of diabetes. Lipids are oxidized through beta-oxidation, which generates reducing equivalents that enter through the electron transfer flavoprotein, this is the ETF, in concert with complex one. Now, we’re trying to understand a bit more of how the ETF regulates this lipid oxidation. So, the bulk of this talk is going to be focused on how exercise and obesity regulate oxidation through the ETF.

Now what’s a byproduct of this excess or this hard push on lipid oxidation? Well much like our cars can undergo disrepair, excess lipid oxidation seems to be associated with the development of reactive oxygen species and oxidative stress. So, nutrient excess when it’s not matched with energy demands, just like our cars can start to accumulate this damage, our mitochondria and other proteins seem to also accumulate damage and potentially contribute to their dysfunction. Our goal is to have a nice, well-oiled functioning machine. One way at which we can push this is through the replacement of damaged parts. Now, much like we could take our cars to the mechanic, we can also stimulate the removal of damaged parts and the synthesis of new parts. This is called protein turnover when we degrade an older protein and replace it with a newly synthesized protein. Our focus is on how can exercise stimulate protein turnover. We will be discussing aspects of how the regulation of protein breakdown and the re-synthesis of new proteins can help shift our broken or rusty mitochondria into these well-oiled machines. So, to summarize this talk, we are walking through how mitochondria respond to exercise and nutrition through the turnover of mitochondrial proteins and how that underlies our changes in whole body fuel oxidation.

So, first off, what is our main impact of high-fat diet and exercise on respiration. I’ll discuss two of our recent projects using mouse models, and then we’ll transition over to two of our recent human studies. So, to investigate the development of high-fat feeding and exercise, we turn to our obesity model using C57 Black 6 mice, and we perform four weeks of high-fat feeding, which induces obesity and skeletal muscle insulin resistance. We then perform exercise training for an additional eight weeks or allow the mice to remain sedentary. This was a project that was led by recently graduated Dr. Sarah Ehrlicher, who’s currently at the University of Alabama. She’s pursuing a clinical nutrition program and she’s been assisting with Dr. Jim Hill and a few of his ongoing projects. This was funded through NIDDK, a K award awarded. So, we’re focused on mitochondrial respiration. The way we did this is isolating out mitochondria, and then we can perform high-resolution respirometry, which allows us to use specific substrates in order to test the specific on-ramps onto the mitochondria. Our main output is measuring oxygen consumption, and so much like we can measure how fast a car is moving, we can also measure how fast the mitochondria are able to oxidize specific substrates. The data that I’ll be presenting here are the respiration rates specifically for lipids and we’ll orient you to the data. The vertical axis is our respiration rates and on the horizontal, we’ve separated out between the low-fat and the high-fat diet group. The exercise groups are in the black bars. So, focusing first on the low-fat diet group, we see that exercise can stimulate mitochondrial lipid oxidation and this also occurs in the high-fat feeding groups. Now, what’s quite apparent is that when we express these data on an absolute basis, the high-fat diet itself is a strong drive to stimulate mitochondrial lipid oxidation. Now, this would make sense from a physiological stress perspective because the high-fat diet includes 60% fat. So, we’re basically forcing oxidation of lipids. Now when we account for changes in the amount of mitochondria, these are the data that are on the right, this effect of high-fat feeding to drive mitochondrial respiration remains, such that there’s an increase in the total amount and also the relative amount per mitochondrial protein. We’ve done proteomic analyses and seen that the high-fat feeding remodels the mitochondria for more proteins of lipid oxidation. Now, because synthesis and degradation underlie all these changes, we next started to ask questions about, well, how does the contribution of degradation and synthesis contribute to these changes in the mitochondrial proteome and their function?

So, the next part of Sarah’s project was investigating “do those mitochondrial adaptations with exercise depend on the activation of autophagy?” Autophagy is a major degradation pathway to turn to a model that was developed by Dr. Beth Levene’s group. This is a model of impaired autophagy in response to exercise. So, exercise is a strong way to stimulate the breakdown of proteins through autophagy. And one of the ways it does that is through this Beclin-BCL2 complex, wherein exercise stimulates the phosphorylation of BCL2 and that causes the release and then an activation of autophagy through beclin. The model is through BCL2 is not able to be phosphorylated, so there’s been some substitutions such that there’s alanine residues and BCL2 is not 4 related in this model. That leads to a restriction of autophagy activation with exercise. Now this allows us to ask about exercise-induced autophagy. The rest of the time these mice are fully able to regulate their protein turned over just fine. It’s that specific exercise induction.

What Dr. Levine’s group showed is that these mice had impaired glucose adaptations to exercise. We wanted to focus in on the mitochondria. So, we performed a similar high-fat feeding and exercise paradigm, but now using these BCL2 mice, asking the question, does the exercise induction of autophagy, is that required for the mitochondrial adaptations? So, again, we’re looking at the vertical axis here is mitochondrial lipid oxidation. And what we’ve found is that there’s similar results compared to the wild type mice, in that in both the context of the low-fat and the high-fat diet, exercise-stimulated lipid oxidation in both the total amount of mitochondria and when we make it relative to any changes in mitochondrial protein abundance. So again, we see this robust ability for mitochondria to adapt to the high-fat feeding stimulus and the exercise, and that the BCL2 did not seem to be required for any of those adaptations. This was a contrary to our hypothesis, so we wanted to start pursuing other aspects of this to say is this limited to lipid oxidation? So, we looked at other substrates specifically here for complexes one and two in combination and we’re comparing between the wild type and the BCL mice and again in the context of the high fat diet we saw that exercise stimulated the respiration of these non-lipid substrates and that occurred in both wild type and the BCL mice. So, we’re seeing that this BCL2 is not required for some of these adaptations.

We wanted to ask the question then, while we’ve been focusing on the degradation, what about the synthesis side? So, we performed isotopic labeling in these mice in collaboration with Dr. Ben Miller at Oklahoma Medical Research Foundation. And what Sarah did is she isolated out the mitochondria. And then the mice were already labeled up with deuterium oxide. This allows us to measure long-term protein synthesis rates over 14 days. This integrates the feeding, the exercise, and any fasting considerations as well. So, a nice integrative measure. What we saw is that the BCL2 mice had a greater synthesis rate with exercise. So here we have a mouse that is now stimulating more mitochondrial proteins and it doesn’t seem that BCL2 activation was required for that activation.

This led us to then hypothesize, well perhaps there’s a compensatory pathway by which BCL is operating and we can have for adaptive mitochondrial responses. So, what Sarah did next is look at Parkin as an activation pathway for autophagy. So, she did the high-fat feeding paradigm in these mice for 12 weeks, and then instead of doing exercise training, she performed an acute bout of exercise and then sacrificed the mice immediately afterwards. She then isolated out the mitochondria and then tested the Parkin localization to the mitochondria. Parkin localization to mitochondria is a stimulation for the degradation of mitochondria. And what she found was that the low-fat diet exercise stimulated the localization of mitochondria or Parkin to the mitochondria. But interestingly enough, in the high-fat diet, that response was blunted in the wild type mice. This is consistent with some other findings that have shown that high-fat diet seems to be blunting the mitochondrial degradation response, potentially rendering these a little bit more susceptible to accumulation of damage. Now the BCL2 mice, interestingly enough, had this increase in Parkin localization, suggesting to us that perhaps this added Parkin localization indicates a compensatory pathway by which the exercise-induced autophagy is occurring, perhaps independent of BCL. So, we have strong remodeling occurring with high fat feeding. Exercise, again, is a strong way to stimulate the remodeling of mitochondria.

So, we wanted to translate these findings to humans. And asking some questions revolving around, how did the mitochondria undergo rapid changes to demands on mitochondrial fuel oxidation with aerobic exercise? To this, we performed two studies in collaboration with the Samaritan Athletic Medicine Center on campus of Oregon State University. This is an outstanding outpatient sports medicine clinic that we can perform insulin clamps, muscle biopsies, exercise testing, and a number of other clinical research procedures just a few minutes away from our laboratory. We really appreciate the clinical staff support and the enthusiasm.

We first asked a question revolving what are around the acute effects of exercise on mitochondrial respiration. This was a project led by Dr. Harrison Stierwalt who’s now at the University of Kansas Medical Center, pursuing a postdoc with Dr. John Thyfault. Harrison’s project was looking at the acute effects of a single session of exercise on insulin sensitivity and substrate oxidation. These specific data we’re looking at immediately post-exercise. We took a muscle biopsy sample and then performed our respiration, looking at how did that acute session of exercise alter substrate-specific respiration. Now, we performed an isolated mitochondrial preparation. What that allowed us to do is look at the intrinsic effects on mitochondria, removing any considerations such as the cell membrane or blood flow considerations, all of which are known to help regulate substrate metabolism. We focused in on, could that single session of exercise alter substrate metabolism specific within the mitochondria.

The data that I’m showing here are the respiration that are coupled to energy production. We’ve separated out based on the different on roads to the mitochondria, whether we’re coming in through our lipid oxidation that we’ve been talking about or through complexes one or complexes two. What we saw across the board was that there was very mild effects for greater respiration after exercise. And this was really across multiple different substrates. So, what we concluded out of this was that of the major metabolic changes and shifts in substrate oxidation that occur at the whole body level, we’re really seeing minor changes within the mitochondria itself. And if anything, those shifts seem to be consistent across mitochondria and not specific to a certain substrate.

So, we next turned our attention to what about a little bit longer term? Could we perform short term training and push mitochondrial adaptations? This was a project that was led by current doctoral student, Phil Batterson. This was something that we just completed a few months ago. So, within the research interruptions that occurred with the COVID, Phil has been doing just an outstanding job of leading this project.

Our question was, we know that aerobic exercise is a strong stimulus on mitochondria, and we know that high intensity interval training is a way that we can push more mitochondrial remodeling, stimulating the degradation and the synthesis of new proteins. So, for Phil’s project, what we did is performed seven sessions of interval training. This had people perform one-minute bouts of exercise at near maximum, then rest for one minute and repeat that. So, a single session included 10 repeats, then they rested for a day, and then repeated it on the next day. So, over two weeks of training, they performed seven of these exercise days. We were using this high intensity to determine could we see some of the similar activation of the degradation and the lipid pathways that we’ve seen before. So, in order to test this, we performed meal tolerance tests, looking at whole body substrate oxidation. And then we also performed our isolated mitochondrial respiration.

So, we’ll focus first on the whole body substrate oxidation. These are data similar to what we saw with Dr. Ayala looking at the respiratory exchange ratio. This allows us to test whether a person is oxidizing more lipids, so value of around 0.7, or if a person is oxidizing more carbohydrates at a higher value of about 1.0. So, after a meal, there’s a shift towards oxidizing more substrates, and that occurred similarly between the pre and the post training. But what caught our attention here was that in the resting state, post-training, participants were oxidizing more lipids. So, even after just seven sessions, we had this shift in substrate oxidation towards lipids at the whole body level.

We next considered, well, if we’re seeing this at the whole body level, what’s going on at the mitochondria? So again, we returned back to our isolated mitochondrial prep, and through work with Erin McGowan, second year PhD student who spent many hours in the lab respiring these mitochondria, she performed a lipid titration protocol whereby we injected in sub-saturating amounts of these lipid substrates to see what is the sensitivity of lipid oxidation and then also what’s the maximal capacity. And what we saw is that after training, there’s no real change in the sensitivity of the mitochondria to lipids, but there’s this trend towards greater lipid oxidation capacity. Now what’s remarkable here is that from an absolute standpoint, this is just seven sessions of exercise being able to push more lipid oxidation here. Now these are preliminary data that we’re working on accounting for changes in total protein as possibly driving these effects. So, this is something that Phil is working on actively, and we’re looking at how are the degradation pathways are also regulated here.

So, in summary, we’ve walked through the electron transfer system as our metabolic roadway and the on ramps through the electron transfer of flavoprotein in combination with other respiratory substrates. We’ve walked through data relating that acute exercise seems to have minor effects that are consistent across substrates and that high fat feeding and exercise training can really drive greater lipid oxidation capacity. We’re looking at how that changes the damage to the ETF, are there oxidative modifications that are possibly lowered and that helps improve the lipid oxidation capacity, and we’re looking at the individual protein synthesis as well. Our next studies are also considering how does substrate competition occur between all of these different nutrients, with some evidence that there’s perhaps an ability for the mitochondria to be overloaded when there’s many competing substrates.

So, the take-home goals of the talk have really been to understand how mitochondria adapt to the high-fat feeding and exercise stimuli, and that this adaptation occurs through the turnover of proteins, and that these underlie changes in respiratory functions with the possibility that there is an overload of the mitochondria themselves. These are new hypotheses that we’re pursuing and I’m looking forward to discussing these further with you in the next coming months. So, with that, we’ve had a variety of discussions on stress feeding. So, a final encouragement here is if the stress feeding in these contexts of traffic jams are leading you to comfort foods, remember that exercise could possibly be a way that we can drive our well-oiled metabolic machines. I look forward to discussing these concepts a bit further with you. And Sarah, I’ll turn it back to you for the questions and answers.

Sarah: Thanks so much, Matt. That was a fantastic presentation. And myself, I am a mitochondria nerd. So particularly interesting to us “mitochondriacs.” Great. So, I’m going to welcome Matt and Julio back online. Are you with us guys? Yes. Yes, I am. Yes. Great. Okay. So, we’re going to kick off the Q&A with our first question. This question is for you, Julio. Did the locomotor activity measures include wheel running behaviors?

Dr. Ayala: That’s a great question. No, unfortunately, we did not include wheel running. We have the capability to measure that. And that’s something that’s on our radar. But this was just ambulatory activity within the cage.

Sarah: Okay, great. Matt, this question is for you. Is it possible that mitochondrial capacity differs among major tissue groups? For example, mitochondria in muscle versus neurons versus hepatic tissue.

Dr. Robinson: Yeah, certainly. So, there’s definitely evidence for differences across tissues from a respiration perspective. A couple of things that underlie that question is what are the nutrition and energy demands within each of those tissues? So, for example, within skeletal muscle, the mitochondria are being driven to generate ATP versus at the liver, they’re also being driven to generate ATP but they’re flooded with a number of nutrients. So, what we’re trying to consider a little bit more in the field is how do those tissue responses vary to their specific demands? And also, how does the remodeling response vary between each of them as well? And for example, the liver has a very high turnover rate of mitochondria. So, they might be able to adapt fairly well.

Sarah: Moving on to our next question, Julio, this question is for you. How does restraint stress compare to other forms of stress, like predator-induced stressed acute injury, thermal stress, infection, et cetera?

Dr. Ayala: Yes, that’s a great question. something that we’re interested in pursuing, especially with regards to whether these not only does obesity affect the response to stress, but whether the mechanisms that we’re interested in are also a general response to different stressors or are there specific mechanisms that are relevant to specific types of stress and whether or not obesity affects those. So yeah, that’s a very good question. My hunch is that given the fact that rodents tend to respond differently to different stressors, that maybe some of the phenotypes that we are looking at and some of the mechanisms may be specific or relevant to one or maybe a couple of stressors, but not necessarily to all stressors. And that’s something that is in the long-term plans for our studies. So, great question.

Sarah: Fantastic. Matt, this next question is for you. The natural exercise regime of mice is to run full speed for several seconds and then take a break and then run full speed again. Is that similar enough to be considered analogous to high intensity interval training in humans?

Dr. Robinson: Yeah, that’s such a great question on mouse exercise behavior. So, yeah, so mice are of those natural interval runners, they’re not made for running long distances. So, for example, our exercise program on the mice tries to mimic that. So, we have the mice run at shorter, higher exercise intensity for a single bout, and then we lower the grade, lower the speed, give them a little bit to rest, and then repeat that over time. So, the mice do allow us some understanding of what is the effects of these high-intensity bouts. Now what we’re trying to understand a bit more in humans is we know those high-intensity bouts can really drive some pretty strong adaptations, so can that be useful within, in our context, for example, obesity or in other disease populations, such as cardiovascular diseases, can HIIT training be really useful for inducing rapid metabolic gains.

Sarah: Great. That’s really, really interesting. Julio, this next question is for you. Do the obese mice have lower neuronal response to initial restraint because they are, quote unquote, fat and happy?

Dr. Ayala: Yeah, great question. And that’s sort of the $64,000 question is something that we’ve been discussing a lot in the lab is why are obese mice less prone to neuronal activation in response to stress? Because yeah, it is very possible that there’s something completely independent of what we’re looking at that just makes them less susceptible to feeling stressed and therefore their neurons are not becoming activated and they’re not as susceptible to stress. So, we’re looking at that in two different, at least a few different ways. One is, I mean, just the indirect calorimetry data shows that their energy expenditure increases equivalently to that what is seen in lean mice. Now that is not a direct indicator that they are just as stressed as lean mice, but you know, could potentially be used as a proxy. So, that would suggest that they are feeling stressed, it’s just that their neuronal activity just doesn’t come on as much as lean mice. The other way to approach it is to, as in the previous question, look at different stressors that aren’t necessarily tied down to a physical restraint, something that you would assume would equivalently stress a lean and an obese mouse. And then finally, also just what happens if we expose mice to a palatable substance like a high fat diet or sucrose, not allow them to become obese, but then stress them having had that exposure, and then seeing whether in the absence of obesity, just the mere exposure to these palatable substances attenuates the activity of these neurons in response to a stressor.

Sarah: Matt, so this is our last question, and it’s for you. Have you measured ROS, or reactive oxygen species production, in the oxygraph with either substrate?

Dr. Robinson: Yes. Now this is an outstanding technical question there. So, yes, we have. And what we find is that the lipid oxidation generates more leak to reactive oxygen species. So, it’s very substrate specific. Now what we’re trying to understand a bit more is what happens when we have multiple competing substrates there. So, if we have several nutrients simultaneously, can we drive greater ROS production, and is that specific to an individual complex? So, yes, we’ve seen that. We see it mostly through lipid oxidation pressures, and that’s one of the reasons why we’re looking at some of the substrate competition and the reactive oxygen species damage to the mitochondrial proteins.

Sarah: Very cool. I used the oxygraph in my master’s research, so I’m particularly interested in these questions, and I’m excited to see the report that we get out later with all of answers from both Matt and from Julio to all of these questions. They’re really fantastic. So, thank you everyone for sending them in. And first I want to thank our presenters Julio and Matt. Thank you so much for these really awesome presentations. These talks are so important and I think your research is very relatable to everyone, especially in the current climate that the world is in. So, thank you again for your insights.

Dr. Ayala: Thank you for the opportunity.

Dr. Robinson: Yeah, we appreciate the opportunity to share.

Sarah: Okay, great. So, with that, I wanted to thank everyone who attended today. In closing, thank you again for taking a part in this webinar, and we look forward to having you with us again soon.