Stress, rhythm, choice and the munchies – tribute to Mary F. Dallman

Abstract

In December 2021, we lost a pioneer in the field of stress research who inspired generations of scientists. Mary Dallman was an expert on the hypothalamic-pituitary-adrenal (HPA) axis, its interactions with a wide variety of other physiological systems and the impact of chronic changes of HPA function on energy metabolism and adiposity. She was not only an excellent scientist, she was a great role model and mentor for young scientists, especially women. She encouraged and supported many of her trainees even long after they left the lab. Her outside-the-box thinking, the fun and crazy discussions we had in the lab proved to be a beautiful basis for my own future research.

Keywords:

• Comfort food
• restrained stress
• food choice
• diet-induced obesity
• corticosterone
• snacking

The rhythm in B

My first interaction with Mary Dallman was in 1998 at a conference in San Antonio, Texas. I presented my first poster as a graduate student on the diurnal rhythm in plasma glucose concentrations with rising levels prior to the activity period. I showed that this rhythm was controlled by the suprachiasmatic nucleus (SCN) and independent of the feeding rhythm (la Fleur et al., 1999). For these experiments, we used a 6-meals-a-day feeding schedule (with a feeding opportunity for the rats every 4 hours) to abolish the daily rhythm in food intake and ensure that the animals were eating similar amounts during the day and night. Different feeding schedules (such as having access to food for only 2 hours a day or only in the light period) are known to shift/alter hormone rhythms (Hiroshige, 1984; Wilkinson et al., 1979) thus could indirectly affect glucose metabolism. As a proof that this schedule only abolished the feeding rhythm, we showed that the diurnal rhythm in plasma corticosterone (B) concentrations was not altered (la Fleur et al., 1999). Several people passed by while I presented my poster, and then someone mumbled while passing “You just wait till Mary Dallman sees this.” Thereafter, I nervously awaited the arrival of the expert on (rhythms in) glucocorticoids (GCs), expecting a heated critical discussion as it seemed that I should be scared of her. She came, she read the poster and then looked at me, and started to ask me many, many questions and blurting out ideas and experiments I should do next. That day, my scientific life took a turn as I got to know this brutally honest and critical, and above all, the most generous and caring person in the world. The effect of a 6-meals-a-day schedule on the rhythm in B was later on also repeated in mice with similar results (Sen et al., 2017). This is distinct from the described meal anticipatory behavior which occurs when animals are restricted to a short time window to eat, which is accompanied by a rise in B prior to the feeding bout. It has been shown that B is important for this meal anticipation to occur (Namvar et al., 2016).

Alphabet soup for adrenal steroids

Mary used the letter B to refer to corticosterone in her talks and publications. She nicely reviews the origin of this abbreviation in her essay in 2005 on ‘adrenocortical function, feedback and alphabet soup’ (Dallman, 2005). In the 30s, adrenal steroids were isolated by Edward Kendall and Tadeus Reichstein from adrenal glands by extraction of organic solvents and using chromatography. The separated fractions of the steroid mixture were labeled by letters, Kendall isolated 6 whereas Reichstein isolated 7. Kendall’s Compound B was corticosterone, the main glucocorticoid in the rat, and his Compound F was cortisol (or Reichstein’s Substance M), the major adrenal glucocorticoid secreted by humans. As the chemical names of these steroids were long, it became common to use the letters instead.

Mary, as she states in her essay (Dallman, 2005), preferred using letters B and F to the other common abbreviation ‘cort,’ because it did not distinguish between the two active adrenal glucocorticoids secreted by rats and humans.

A new view of comfort food

In 2001, I was ready to take my next scientific step and started looking for a postdoc. In the era of molecular biology, and trends to go smaller and smaller, I was looking for something different. I wanted to look at the bigger picture, learn about the physiological system as a whole and understand the mechanisms that connect the brain with the periphery and vice versa. I remembered Mary’s unique perspective and her enthusiasm when discussing science at the meeting. I wrote her a letter to ask about opportunities in her lab. She returned my message immediately and hired me on the spot. It was the time that she put forward her ideas about “comfort food” and the response to stress (Dallman et al., 2003). Modern times in our society come with palatable high-caloric dense food items that are available at every corner, contributing to the obesity epidemic. In addition, modern society is also associated with high levels of external and psychosocial stress which interact bi-directionally with energy balance. While some individuals may experience reduced appetite in response to stress, many will increase caloric consumption and particularly palatable food items (Dallman, 2010). With the “comfort food hypothesis,” Mary postulated that people eat palatable “comfort” food to reduce the activity of the chronic stress-response network. This chronic stress-response network is recruited by chronically high GCs. Central to this network is the amygdala, where chronically enhanced GCs increase corticotrophin releasing factor (CRF), which enables the recruitment of the stress-response network through its innervation of cortical, subcortical and brainstem nuclei. GCs also enhance the salience of pleasurable activities (like consuming palatable foods) and, indeed, both stress and high levels of B in rats increase the intake of sucrose and saturated fat (Dallman et al., 2006; la Fleur et al., 2004; Pecoraro et al., 2004). Moreover, GCs also have systemic effects, increasing adiposity especially in visceral fat depots (Strack et al., 1995). A series of experiments with sucrose feeding in adrenalectomized rats, showing that sucrose reduced CRF mRNA expression in the paraventricular nucleus (PVN) and the finding that mesenteric fat weight negatively correlated with CRF mRNA expression in the PVN irrespective of rats having adrenals or not, led to the hypothesis that “mesenteric fat stores serve a signal of energy stores that feed back to inhibit the CRF activity in the HPA axis” (Dallman et al., 2003). It remains, however, still to be determined what this signal is and whether it indeed comes from visceral fat.

In an interesting paradigm to understand the role of “comfort food” in stress responsivity, Dr. Ulrich-Lai introduced a model in which rodents are presented with a small amount of sugar at the beginning and the end of the dark period (Ulrich-Lai et al., 2011). This limited sucrose intake (LSI) paradigm reduces the stress response, but does not alter body weight. In addition, it was also shown that using saccharine instead of sucrose or giving access to another type of natural reward (partner of the opposite sex) reduced the stress response pointing to a reward related signal mediating this effect rather than a metabolic signal (Ulrich-Lai, 2016). From my own work together with Mary, it seems that the signal can come from the periphery, as in metabolically stressed diabetic rats, consuming lard reduced CRF mRNA in the chronic stress-response network, this effect was (la Fleur, Manalo, et al., 2005). Thus, it is clear that comfort food reduces the stress response, how exactly this is mediated remains to be determined.

Choice of lard, but not total lard calories, dampens the response to restraint stress

After publishing this idea of comfort food relieving stress, which became her most widely cited paper (Dallman et al, 2003), Mary pointed out that she contributed to a paper in which it was shown that dietary fat intake induced chronic stress with elevated B levels and an enhanced HPA response to restraint stress (Tannenbaum et al., 1997). This study did not support the comfort food theory and was a matter of debate in the lab. Starting as a joke, which then became a wild idea, ended up as a serious experiment. We hypothesized that the way we render rodents obese using a pelleted high fat diet induces stress, masking the effects of comfort food on the stress response. The idea was born to provide animals with the choice to consume chow and lard and compare this with a group provided with a mixture of chow and lard as being a non-choice control. We subjected both groups to restraint stress and measured ACTH and B levels after 7 days. It turned out that, indeed, animals having the choice to consume lard voluntarily had reduced ACTH and B responses to restraint stress, whereas the animals provided with a 50% lard-chow mixture did not, suggesting that an involuntary dietary fat intake is masking the effect of comfort food on stress (la Fleur, Housyar, et al., 2005). The experiment was repeated in the lab a couple of years later, using not only lard-chow choice, but also sugar-lard-chow choice and sugar-chow choice. For all these palatable diets, a reduced stress response was detected, all be it only at 240 min and not at 15 min as previously shown (Foster et al., 2009). Interestingly, when evaluating the CRF response in the chronic-stress network, they found that sucrose ingestion affected a different neural network compared to lard ingestion (Foster et al., 2009). This was in line with the results I obtained while starting my own research lines in the Netherlands.

A novel obesogenic rodent model based on free choice of palatable items

When starting my own independent research line in Utrecht in 2005, I changed gears somewhat, focusing on obesity and diabetes and the role of the brain in their development. This was complex enough without a focus on the interaction with the stress network. Although independent, I never stopped “running ideas and data” by Mary. In my search for a more human-like obesity model, I focused on the idea of choice and characterized the free-choice (fc) high fat high sugar (HFHS)-diet (Slomp et al., 2019). It turned out to be a perfect model to study the differences between sugar and fat-based diets, and to determine whether there were additive or synergistic effects of the combination of ingesting fat and sugar in obesity and diabetes development (Slomp et al., 2019). We showed that the animals on the fcHFHS diet with lard and sugar water were persistent hyperphagic, which was characterized by a snacking-like phenotype (la Fleur et al., 2014) and enhanced food-motivated behavior (la Fleur et al., 2007; Slomp et al., 2019). Thus, animals on a fcHFHS diet increased meal frequency and kept meal sizes similar to chow fed animals, whereas animals fed a non-choice HFHS diet increased meal size with fewer meals a day. This type of compensation was also shown for animals on a fcHF diet, also exhibiting larger meal sizes and reduced meal frequency (la Fleur et al., 2014). Animals on a free-choice high sugar diet showed smaller meals but increased meal frequency. Taken together, the combination of fat and sugar in a choice paradigm clearly provided us with a hyperphagic-snacking phenotype. Given the association between snacking sugar-based beverages and the obesity epidemic (Malik et al., 2013), this is a useful model to study mechanisms underlying obesity development.

As was shown for the stress-response network in the brain (Foster et al., 2009), feeding-related neural circuitry was also differentially affected by fat and/or sugar consumption. The fcHFHS diet affected Neuropeptide Y (NPY) gene expression and NPY responsiveness (Gumbs et al., 2020; van den Heuvel, Eggels, Fliers, et al., 2014; van den Heuvel, Eggels, van Rozen, et al. 2014; van den Heuvel et al., 2015) and also altered the dopamine and opioid system in reward-related areas (van de Giessen et al., 2012, 2013; van den Heuvel, Eggels, Fliers, et al., 2014). We observed many fat – sugar interactions when comparing the different diets, and more recently showed that sugar enhances fat intake, most likely via an opioid-like mechanism (Koekkoek et al., 2022).

Interestingly, different studies have highlighted the interaction between stress and high fat diet in feeding behavior and body weight gain (Peckett et al., 2011; Shpilberg et al., 2012). Restraint stress in rats enhances palatable intake (Pecoraro et al., 2004), and high GCs promote lard intake (la Fleur et al., 2004). Combining HF diet with stress paradigms accelerates obesity development in mice, with a role for NPY signaling in the amygdala (Ip et al., 2019). In 1993, Mary put NPY in the middle of her proposed peripheral feedback loop linking GCs and insulin to feeding behavior (Dallman et al., 1993). She never lost interest when it came to NPY biology and her work still inspires our current studies to unravel how NPY plays a role in fat-sugar interactions.

The best mentor

Mary was a true physiologist curious to understand the human system as whole. She could explain processes like nobody else. This tribute only covers a tiny part of her research, that I had the honor to witness and participate in. Long before my time, she identified and established many important aspects of hypothalamic-pituitary adrenal axis function that are in textbooks these days. But above all, she was the most generous person I ever met, sharing ideas freely and providing support to those in and outside the lab. She was the best mentor I could have wished for. She paved the way for women in science, breaking boundaries in unconventional ways and inspiring others to follow their dreams. I will always remember her outside-the-box thinking, the fun and crazy discussions and her never ending curiosity to hear about the next new scientific finding.

Disclosure statement

The author reports there is no competing interest to declare.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.

Notes on contributors

Susanne E. la Fleur

Susanne la Fleur is professor in Neurobiology of Energy Metabolism at the University of Amsterdam and University Medical Centers Amsterdam. Her research group focuses on the interaction between energy metabolism and the brain in the development of obesity and type 2 diabetes, with a special emphasis on the effects of highly palatable nutrients on the brain. Her research bridges basic science and the clinic to translate findings from the rodent obesity model she developed to the human setting, showing how the brain responds to an overload of highly palatable nutrients.