Tuesday, May 10, 2016

Why some dogs (and humans) are born hungry

The brain is the central regulator of appetite and body fatness, and genetic variation that affects body fatness tends to act in the brain.  One important site of variation is the POMC gene, which codes for a signaling molecule that suppresses food intake.  A new study shows that Labrador retrievers often carry an inactive version of the POMC gene, causing them to be highly food motivated, obesity-prone-- and perhaps more easily trainable. 

Introduction: Brain regulation of body fatness

In mammals, body fatness is regulated by the brain, primarily in response to the hormone leptin.  Fat tissue produces leptin in proportion to its size, and this informs the brain of how much fat a person carries.  Regulation of body fatness happens principally via the (direct and indirect) actions of leptin on specific neurons in the brain, the most important of which are POMC neurons and AgRP neurons (1).

POMC and AgRP neurons are the yin and yang of body fatness.  When POMC neuron activity goes up, appetite and body fatness go down.  When AgRP activity goes up, appetite and body fatness go up.  Leptin affects these populations reciprocally: it activates POMC neurons and shuts down AgRP neurons, explaining much of its effect on food intake and body weight.  These neuron populations are both located in the hypothalamus, and more specifically in a tiny region called the arcuate nucleus (1).

POMC neurons exert their effects on appetite and body fatness in large part using a signaling protein called-- you guessed it-- POMC.  This signal instructs downstream brain regions to engage the behaviors and physiology that cause us to eat less and lose fat.  When the POMC signal goes away, appetite and body fatness increase (2).

Each of us carries a complement of genes that varies in small but important ways, and this accounts for much of why individuals are different from one another.  Not surprisingly, this genetic variance affects eating behavior and body fatness: up to 70 percent of individual difference in body fatness is explained by genetic differences (3).

Many researchers have studied the genetic variation that impacts body fatness, and they have come to a consistent conclusion: genes that affect brain structure and function are heavily overrepresented among those that affect fatness (4, 5, 6).  This is particularly true of genes in the brain pathways that transmit the leptin signal.  This gives us a high degree of confidence that the leptin-brain axis really is the primary regulator of body fatness.

One of the sites of genetic variation that has repeatedly popped up is the POMC gene.  Stephen O'Rahilly and Sadaf Farooqi have shown that mutations that inactivate the POMC gene are the most common (known) cause of single-gene obesity in humans, accounting for up to 6 percent of cases of severe childhood obesity (7).  Other researchers have demonstrated that similar mutations cause obesity in mice (8).  A new study by O'Rahilly's group shows that the same principle operates in dogs as well, and that it may explain the behavior and body shape of your favorite pooch.

The study

The researchers began by comparing a small sample of lean and obese Labrador retrievers, looking for mutations in genes known to influence body weight (9).  Among the obese dogs, 10 out of 15 carried a mutation in the POMC gene that is predicted to inactivate its protein product.  Only 2 out of 18 lean dogs carried the mutation.

They went on to look for associations between the mutation and obesity in a larger group of 310 Labs, and found that each copy of the mutation was associated with an extra 4.2 lbs (1.9 kg) body weight.  Dogs with two copies of the mutation weighed 8.4 lbs (3.8 kg) more than dogs with two functional copies of POMC.

Not only were Labs with the POMC mutation heavier, they were also more motivated by food.  This was determined by administering a questionnaire to the owners that measured the dogs' behavior around food.

Interestingly, they found that Labs with the POMC mutation were much more likely to be service dogs.  We don't know exactly why that is, but the most likely explanation is that highly food-motivated dogs are easier to train.  By selecting for trainability in dogs, we may have inadvertently also selected for mutations that predispose to excessive food motivation and obesity.


Genetic variation can impact food motivation and the susceptibility to obesity.  In mice and humans, the genes in question tend to be components of the brain pathways that regulate appetite and body fatness.  This new study shows that the same is true in dogs, further supporting the fundamental importance of these pathways across diverse species.  It also suggests that we may have inadvertently selected for an obesity gene in Labrador retrievers.

Some dogs-- and humans-- are simply born hungry, and this study adds to our understanding of why this happens.


guiseppe014 said...

Surely there is an obesity epidemic in dogs and cats. A 2014 survey suggested 18% of US dogs were obese and 35% overweight, while 28% of US cats were obese, and 30% overweight. http://www.petobesityprevention.org/pet-obesity-fact-risks/ Apparently it´s twice as high as in European countries.

I´m not sure what role the commercial feed play in this, and how much it is related to dogs and cats being given leftover human foods. In the past it was more common to give meat and bones, now it´s corn, soy and whole grains etc. What was the rate of obesity in cats and dogs 50 or 100 years ago?

The website of one of the major dog food manufacturers list «some of the good stuff we use in our recipes»: Ground whole corn helps build lean muscles and is a good source of energy. Corn gluten meal helps build, maintain & repair tissue, improves the immune system and gives a healthy skin & coat. Beet pulp and brewers rice help digestion and give extra energy.

WilliamS said...

A key question is whether those mutant labs also gain weight and beg for food on their (our) ancestral diets, or just on industrial crap in a bag. I bet that just as the common human genomic variations only cause obesity on industrial diets (obesity was really rare before they existed), those mutant labs would weigh and eat just like all the others on a decent diet. Of course that could be easily tested.

Alex aka Primalkid said...

Alternatively, these hungry genes worked independent of dietary composition. Rather, the modern food environment allows for over-consumption whereas our ancestors did not have food readily available despite hunger. Therefore, they stayed lean.

Stella B. said...

I have three ppodles; two skinny ones and one bottomless pit. The skinny ones won't overeat. They all get the same, high quality food. It's not food, it's not macronutrient ratio, it's not willpower, it's not nurture, it's not breed. It's an inborn inability to balance energy input and output.

karl said...

This reads as a statement of fact:

"The brain is the central regulator of appetite and body fatness, and genetic variation that affects body fatness tends to act in the brain."

Or just this bit - "The brain is the central regulator of appetite..." emphasis mine.

BUT! Do we really know that? If humans and dogs were engineered there might be a single 'central regulator' - but we are not - we evolved. Why would nature select for a system where a single point mutation would be lethal? Shouldn't we expect to see overlapping control loops?

I'm not saying that the brain is not important - I just don't buy that appetite, unlike most every other system, lacks redundancy. Being sexual creatures matters - the random mixing of genes produces problems - and those sets of genes that have redundant control loops have a chance to survive to pass on their genes - those that don't die out.

No one wants to see this - it makes things much harder to understand. But to find the truth means to see biology as it really is - not simplified to fit some humanly understandable meme. To really understand appetite means understanding all of the evolutionary history - and that is why the feedback loops are nested, confounded, non linear, integrating, differentiating etc - an insanely confusing mess.

I think we have high level loops listening in to archaic systems on the cell level (for example see Adipocyte-selective reduction of the leptin receptors induced by antisense RNA leads to increased adiposity, dyslipidemia, and insulin resistance. where they note: "Importantly, this suggests the possibility that leptin resistance at the adipocyte level might be a molecular link between obesity and type 2 diabetes." - in other words adipose internal signaling with leptin. When did the nervous system evolve to listen in?

Why do humans survive a Vagotomy?

Had the same thought - dogs in nature don't eat rolled corn full of carbs. CIAB.

From the paper:
"We recruited a cohort of 310 pet and assistance dog Labrador retrievers."

Farming started recently - and I suspect it will take time - as it has for lactose tolerance in humans - to evolve compatibility in dogs to a different diet.

This does not mean that this is not an important control loop. Dogs evolved as scavengers and pack hunters - they really have to compete to get their pound of flesh. They need the motivation.

I suspect that the fairly common cataracts that we see in old dogs is likely an effect of an unnatural diet.

Stephan Guyenet said...

Hi Karl,

The brain is the only organ that controls the muscle contractions that put food into the mouth. The brain is the only "computer" in the body capable of integrating and processing all the information that is relevant to eating vs. not eating. There is a lot of redundancy in the systems that regulate food intake, but it's all in the brain. Note that animals with POMC mutations aren't totally dysfunctional, they're just fatter than animals without the mutation. They still are able to homeostatically control food intake and adiposity, just with a higher setpoint.

That's because of a) redundancy in the functions of POMC neurons and AgRP neurons (recall that they are reciprocally regulated by leptin), b) redundancy between the hypothalamus and brainstem, and c) redundancy with other neuron populations that I left out of the discussion for simplicity's sake.

It is of course true that the brain regulates (in part) by listening to what's happening in the body. For example, signals ascending from adipose tissue, the digestive tract, and the pancreas. Some of these travel via the vagus nerve, others do not (explaining why vagotomy is not excessively destructive to physiology). And yes, changes in these signals (e.g. mutations in the leptin gene) can affect food intake and body fatness. But the point is that the integration and processing of these signals happens in the brain, and the relevant outputs originate from the brain-- in other words, the brain is what regulates the system.

Nate said...

A thermostat regulates the temperature in a room. If you feed false data to any regulating device, thermostat or brain, you will get poor regulation of the system. Put a candle under a thermostat and you will get a very cold room. Develop leptin resistance and you get hungry and fat.

Maybe, I should change my motto from 'It's the Insulin, Stupid' to 'It's the Data, Stupid'. Or maybe, "Garbage in, garbage out."

Stella B. said...

Karl -- I doubt your grain hypothesis. My dogs eat a grain free food, because it's the only food the youngest is willing to eat and yet the fat one is still ravenous and fat. The fat one does get additional carbohydrates in the form of paper, socks, and whatever else she deems edible. You haven't explained why three, closely-related, identically fed dogs have such different outcomes.

Stephan Guyenet said...

Hi Nate,

I agree that the behavior of the system depends on the input it receives. However, there is no evidence that elevated insulin levels lead to obesity-promoting inputs to the brain. I think high insulin is probably just a marker of brain leptin and insulin resistance.

karl said...

The brain is not isolated from effects of the body - it obviously controlled by the body as well. Modulate leptin in the body and you effect appetite.

Take the ENS 'Enteric Nervous System' sometimes called our 'second brain'. Could it effect appetite? Fatness? How about the flora in the intestines? Can they effect what we eat? ( could it be the answer to the existential question is that we exist to feed our Escherichia coli? do they control us?) This ENS of the intestines can be severed from the brain and continues to work fairly well full of serotonin receptors and something like a BBB surrounding ganglia.

The idea that the brain is some sort of control tower for the appetite system misses the fact that it is only one node of the loop. The brain listens to the body - is controlled by what it hears. Plug a feeding tube in and the body stops telling the brain to engage in feeding activity. Really.

I think you disagreed before, but if something causes the body to over-store fat - at some point the same signals that controls the brains appetite/satiety kick in and one will eat to make up the deficit of available energy - at least long term. Leptin is not the only thing the brain listens to.

One can't get away from evolution - earlier systems worked, or we wouldn't be here. And it seems fairly obvious to me that the early control loops to maintain homeostasis within a cell are still running. Yes, the brain is listening in and responding, but what is happening on the cellular level is part of the loop.

Going back to control theory - I don't like analogies (they often mislead) but try this:

If your thermostat is broken - does the brain in your furnace still keep the temperature correct?

Thus, if adipocytes are not functioning correctly, they can account for weight gain or loss. The obvious example is that women - due to estrogen have more stored fat for child bearing - did the brain decide that? Or did estrogens effect on the body?

See Estrogen Controls Lipolysis by Up-Regulating α2A-Adrenergic Receptors Directly in Human Adipose Tissue through the Estrogen Receptor α. Implications for the Female Fat Distribution

Why do women have the appetite to supply the calories to provide for this?

But the question should not be if the brain controls the body or the body controls the brain - it should be 'how does the system work?'. IMO it is a complex of overlapping feedback loops - arranged in a way that a single point mutation is not likely to be lethal.

Stephan Guyenet said...

Hi Karl,

I think it's important to distinguish between the system that regulates a variable, and the inputs and outputs of that system. Take a thermostat. It responds to temperature inputs, and creates heat and A/C outputs. The behavior of the system is obviously heavily dependent on the inputs it receives. But the thing that is REGULATING the system-- collecting inputs and computing what the outputs will be-- is still the thermostat. That's what I'm saying here. The brain responds to signals from the body (as well as many external signals)-- the inputs-- and creates the outputs. The outputs are eating behavior, metabolic rate, etc. Yes, the brain takes signals from the body into account, but that doesn't change that fact that the brain is the regulator.

You seem to think that my argument is that the brain is operating totally independently of signals from the body and elsewhere. That's the opposite of what I'm saying. On many occasions, I've written about the numerous signals the brain responds to, not just leptin but peptide hormones from the gut and pancreas, and signals ascending from the enteric nervous system. The brain integrates an enormous amount of information into its decision to eat or not eat. Some of that comes from body signals of energy status (including leptin, insulin, and a variety of gut/pancreas signals), and some of it comes from things in your environment like the smell/sight of French fries.

karl said...

I'm quite aware of what you have written - (and think your blogs on LA are important), but exactly where is the body weight set-point controlled? Some place in the brain?

The most likely source of a set-point ( and I realize there is much we don't yet know ) - is adipocytes. Evolution selects for what works - and creating an accurate set point an integrating feedback signal is what improves accuracy ( see PID loops for an introduction - you can see that integrating terms reduce the long term errors - provide loop accuracy ).

As fat accumulates in adipocytes, they produce more leptin - which has local effects but also signals the brain which SHOULD reduce appetite ( I don't think the term appetite is well defined - I think I experience more than one sensation - probably there is a serotonin effect, leptin, BG? ). The fact that it is a signal that represents the integral of fat (the amount stored) makes it a likely that it has a strong influence on set-point - IF we were eating the foods we evolved with(bugs and all). (I sort of doubt that adipocytes via leptin are the only integrating feedback - evolution favors redundancy.)

At some point this could get into 'free will' - sort of funny to think that so many have failed to lose weight by will power - which makes my point that it would be doubtful that set-points come from the brain. That is why I take issue with your thesis statement for this blog - which I repeat for clarity: "The brain is the central regulator of appetite and body fatness, and genetic variation that affects body fatness tends to act in the brain."

It sure sounds like you are saying the brain is where the set point signal is derived - I don't think that is true.

T33CH said...

Stephen, you mentioned before that obese individuals have the same amount of fatty acids in their blood stream as thinner individuals. Why is this the case? If they have more fat, shouldn't they be releasing more fatty acids? What is preventing them from doing this? Is it the brain, or have the cells been damaged in a way that prevents them from doing this?

Freeguitar coach said...

Hi @karl Where do I find more of your stuff? Do you write somewhere? What's you full name?

Galina L. said...

I am healed from being always hungry now, as well as my mom. Actually, it is what I value the most about my lowcarbing. Yes, I started it as a migraines prevention regiment(it worked), lost 30 lbs and my health got better, but the loss of my life-long food obsession is the most valuable feature of that eating regiment for me. It changed me - my brain if free to think about something else when my mind is not rotating around food.
Me and my mom (and a grandma) are from a human breed very suitable for a hard work - sturdy build, stronger than average individuals. We don't feel right when we are sedentary.

David Orton said...

Fascinating. But don't forget the "second brain". The gut has its own nervous system which is influenced by by also operates independently of the CNS and autonomic nervous systems. About 75 % of the fibers in the vagus are afferent, from the gut to the brain. From this I take that appetite and food behaviour can be influenced by what is happening in the gut.
Also don't forget epigenagtics, particularly that gene expression is modulated by social and emotional factor; closeness, acceptance, belonging etc, which are intimately concerned with food behaviours, for labradors and their human companions!
BTW I loved Galina's comments. My favourite experimantal animal is the human, and my clinical practice is more informed by what works for people than by theoretical concepts.

Libby Naylor said...

Great... so where's the therapy?

I guess CRISPR gene therapy could help, but that's years down the line...

Is there anyway to supplement your dog's food with beta-lipotropin or POMC?

Our dog is very healthy weight thanks to meal and mealtime restriction and lots of exercise. But his insatiable appetite for food is obsessive. To think that he will never feel full in his short life is sad... I've been researching potential supplements to give him at meal time to at least give him some feeling of fullness.

Does anybody here know of anything?


Galina L. said...

As an always hungry and obsessed with a food human in my past, I can tell that eating more fat has helped me to stop being a family joke. Could you experiment with giving your dog some fat before meals and giving it only meats/chicken with a lard? May be your dog needs some adaptations to eating a lot of fat. I needed an adaptation, especially because my gallbladder was removed. After decades of eating low fat (I was diagnosed with a Cholecystitis at age 10 y.o.) fatty meals were causing me a diarrhea at the beginning of a high fat diet change. Sled dogs eat 80% fat meals two times a day and are fine. I am curious - what is your dog breed?

@David Orton,
Thank you, David! My GP also likes to observe his patients experiences, and he approves my approach to keeping myself healthy. He told me he basically gave up on weight loss recommendations because normally his patients couldn't prevent a weight regain within 6 months.He initially graduated from a university with a major in chemistry and a minor in mathematics, became a chemical engineer, got bored when started to work, than went to a medical school in order to get a more people-oriented profession. It is nice to meet creative and observant people among medics.