I can remember where I was and what I was doing when it happened. I was sitting in my usual spot at Sound Café in the Marigny neighborhood of New Orleans sipping coffee on a Sunday morning. It was mid-August, 2006. It wasn’t a tragic event that locked this place and time in my memory, like the hurricane that blew through this neighborhood the year before and changed the lives of so many, but an article I was reading in the New York Times Magazine – a publication that guaranteed millions of people would see and hopefully read.
The article – a cover story – was titled Fat Factors and made the case to the public, what scientists had suspected for years and now had the technological tools to demonstrate: that bacteria in our gut play a major role in energy balance. That is, bacteria can make you fat. The gut microbiome had just gone mainstream.
For the first time, large swaths of the general public would learn that humans are really superorganisms, a mash up of microbial and human genes – with the former outnumbering us 10 to 1 on a good day. We would also learn that mice raised in sterile conditions are not as efficient at extracting calories from a meal as their germy litter mates. That is, depending on the make up of the meal, your gut bacteria can increase the caloric yield of your food above that which is printed on the ‘calories per serving’ panel on the side of the package. All of this – and much more discussed in that seminal article – meant that microbes, as much as the five boxes of nutrient poor, highly-processed macaroni and cheese for $1, might play a role in the obesity epidemic crippling the U.S. and much of the developed world.
The realization and public announcement that microbes can make you fat was exciting – at least in my mind – as it might get the public to focus on the imbalance between our modern diet and lifestyle and the microbial world we live in. The possibility – which is quickly turning into reality with the pass of time and new research – that a great many diseases that plague the modern world are in fact an imbalance with the microbial and natural world from which we co-evolved came more into focus for me that balmy Sunday morning.
But the article also gave the impression that it was somehow bad that microbes were extracting additional calories from our food and that by identifying these microbes and reducing, altering, or wiping them out would be a good thing. So I fired off a short Letter to the Editor that essentially suggested in so many words that from an evolutionary perspective the bacteria were doing exactly what they were supposed to do – extract energy from otherwise useless dietary ingredients for their host. And maybe we shouldn’t tinker with the microbes too much as they are simply doing what they are supposed to do. After all, our track record as a species when it comes to creating unintended consequences from tweaking ecological systems – in which the human gut is one – is not good.
My little Letter to the Editor was accepted and appeared two weeks later in the magazine alongside another letter that declared more or less that the “obvious” reasons for the obesity epidemic were all around us and that while “peering through a microscope” was interesting, we should stay focused on portion sizes, fast food, and processed food as the real culprits and “Any theory that diverts our attention from them may do more harm than good.”
Despite the dire warnings in that second letter, I would argue – 6 years on – that the publication of that article in the New York Times Magazine along with the years of research that formed its content, encouraged tens of millions in additional research dollars in human-microbe interactions that has resulted in the most significant advances in our understanding of human health and disease of any previous decade in human history. And that includes our understanding of obesity.
Building on that earlier research and utilizing many of the genomic and analytical tools that were either completely created in the last decade or greatly improved upon, researchers in labs all over the world are unraveling the role of the gut microbes in the pathogenesis of obesity. But a 26-year-old man that wondered into a microbial lab at the Shanghai Jiao Tong University in Shanghai, China in 2008 that may have provided a critical border piece to the puzzle that scientists are methodically filling in. The researchers would soon discover, at five-foot eight and 385 pounds, this morbidly obese man from the Shanxi Province was carrying a pathogen that was likely contributing to his obesity.
Following his initial visit to the university, the 26-year-old was asked to join a study the researchers were conducting. During that first visit, the researchers learned the man had no family history of obesity or disease, didn’t smoke, drank very little alcohol, rarely took medications, and hadn’t taken any antibiotics in the previous 3 months. Based on the initial evaluation the man agreed to a new diet composed of whole grains, traditional Chinese medicine and prebiotics. This cuisine was provided in four, carefully made and weighed cans that were consumed daily. In all, the daily gruel was 70% carbohydrate, 17% protein, and 13% fat. In all, only 1,364 calories a day!
Starting with Day 1, and continuing every few days for 23 weeks, the researchers collected the standard data like height, weight, and blood pressure in addition to blood, urine and stool samples. The first round of samples confirmed what the researchers likely expected – the morbidly obese man suffered from high blood pressure, diabetes, and in the researchers words “serious metabolic deteriorations.” However, they didn’t anticipate what the DNA analysis of the Day 1 stool sample revealed about the bacteria living in his gut. After tallying up all the snippets of DNA from the sample, the researchers discovered that 35% of all the bacteria in the sample belonged to Enterobacter, a genus that belongs to a major group (Phylum) of bacteria known as Proteobacteria. For most us, our guts are home to hundreds (sometimes a lot more) of bacterial species that belong mostly (>80% +) to the major phyla of Firmicutes and Bacteroidetes. It was the ratio of Firmicutes to Bacteroidetes that first tipped researchers off that bacteria play some role in obesity that was reported on in that New York Times Magazine article. While members of Proteobacteria are present in all of us – and include some of the Who’s Who of bad guys like certain strains of E. coli and Salmonella – a whopping 35% of all the bacteria in a sample belonging to one genus that was not part of one of the major phyla known to dominate the human gut was interesting. So the researchers kept an eye on this particular genus over the course of the 23 weeks to see what happened.
After 23 weeks of four cans a day of the researcher’s gruel, the 26-year-old man stepped on the scale to find he had gone from 385 to 272 pounds. He had shed an astounding 113 pounds. At this point it’s difficult to account for the weight loss. At only 1,364 calories a day, this was definitely a calorie-restriction diet. And given the history of the difficulty individuals have had in following such diets – and the fact the man was not locked up in a medical ward, but living freely at home from what I can tell – it is highly unlikely he didn’t cheat, at least a little. It doesn’t appear the man was put on a Biggest Loser-style exercise program either, or any program for that matter.
Accompanying the weight loss was a normalizing of his triglycerides, total cholesterol, fasting plasma glucose and insulin, and so on – all of this despite the fact that he still weighed 272 pounds and had a BMI of 41. But one of the most intriguing changes at the end of the 23 weeks was the abundance of Enterobacter. This genus went from 35% of all bacteria in his stool sample on Day 1, to 1.77% at week 9, to 0% at week 23. While the abundance of other bacteria went either up or down over the 23 weeks, the changes were minor. The researchers wondered if this shift had something to do with his initial weight gain, and subsequent weight loss. That is, were the Enterobacter contributing or causing the obesity, or were they a result of it? Classic causation versus correlation.
The genus Enterobacter – and many other members of the Enterobacteriaceae family for that matter – are significant producers of lipopolysaccharide (LPS) endotoxins. LPS is the primary structural component of the outer membrane of Gram-negative bacteria like Enterobacter. Numerous studies have shown that if endotoxins translocate (leak) from the gut into serum (blood), they can cause low-grade inflammation that can contribute to obesity and insulin-resistance (think type 2 diabetes).
When the researchers looked at the serum-endotoxin load from Day 1 to Week 23, they noticed a sharp drop. They also noted a substantial improvement in inflammation, as measured by several biomarkers in the blood. So, as the abundance of Enterobacter dropped over the 23 weeks, so did the LPS load and inflammation. In addition, the researchers also conducted a more in-depth shotgun metagenomic analysis of his stool samples, which reveals the genes present with his microbial community (Note: the initial DNA work provided a list of the taxa present, while the shotgun metagenomics provides insights into the capability/function [genes] of the taxa). This allowed them to measure the LPS biosynthetic pathway – that is, the metabolic endotoxin-producing capacity of his entire gut microbial system, across all the bacteria. As expected, and in keeping with the reduction of the big endotoxin producing Enterobacter throughout the 23 weeks, the overall metabolic output of LPS from the system (i.e., all the gut microbes), was greatly diminished over the 23 weeks. So whether it was something in the new diet, or something removed by discontinuing the old diet, his microbiome shifted in a significant way and system-wide inflammation had been greatly reduced in the process as the endotoxin load went down.
To explore the possible role of Enterobacter in this one man’s weight loss (and initial weight gain), the researchers isolated the specific strain in his stool and compared it to a databank of known bacteria (Binder full of Microbes – sorry, couldn’t resist). A match came back for Enterobacter cloacae subsp. For ease, B29 became its new name. Once they isolated B29, they grew bunch of them in the lab and prepared to inject them in germ free mice. Previous research has shown that germ free mice fed a high fat diet (HFD) resisted weight gain, but when inoculated with bacteria (in their gut), they would gain weight. In other words, the bacteria played some role in the weight gain.
In these previous studies the mice were often injected with a cocktail of bacteria, represented numerous species. In Shanghai, the researchers wanted to know how B29 alone acted when challenged with either a high fat diet (HFD) or a normal chow diet (NCD). For this study, HFD was defined as ~35% of calories and normal mouse chow was ~5% fat. (The researchers do not go into much detail on what the rest of the diet was – but let’s focus on the fat content for the moment.)
The researchers took a group (initially 14 in total, but a few died along the way) of genetically identical mice and split them into two equal groups. One group was fed NCD diet and was inoculated with one trillion B29 cells (our Enterobacter strain), and the other group was fed a HFD and inoculated with the same number of B29 cells.
After the first week, the HFD + B29 group showed steady weight gain (and insulin resistance) and eventually reached an obese state by the end of the 16-week trial. The NCD + B29 group of mice remained lean throughout.
Unlike previous studies that used a cocktail of microbes and a HFD, this research demonstrates that a single endotoxin producing bacteria like B29 can function in the same capacity as a whole microbiota (multiple species/strains in a cocktail) to induce obesity and insulin-resistance. The researchers also tested a group of mice using another single-strain strategy, this time with Bifidobacterium animalis. Using the same diets, neither group (HFD or NCD) inoculated with the single strain became obese demonstrating that obesity cannot be induced by just any bacteria under a HFD. In other words, it’s not the HFD alone in the presence of bacteria in the gut that causes weight gain. The specific species or strain of bacteria is important.
The researchers also noted that LPS-binding proteins were higher in the blood of the HFD + B29 mice than the NCD + B29 littermates. This was striking, as the amount of LPS endotoxins was higher in the gut of the NCD + B29 mice than in the HFD mice. In other words, higher levels of LPS endotoxins in your gut does not mean you will have elevated levels in you blood, which increases inflammation that is associated with insulin resistance and obesity. And since the B29 was the only microbe in the mice, then the inflammation-causing LPS endotoxins could have only come from the gut. But how did the endotoxins get from the gut into the blood, and why was this most pronounced during a high fat diet, even though a normal chow diet produced greater numbers of endotoxins in the gut – and presumably a larger source pool of endotoxins that could end up in the blood?
Previous research has shown that a HFD can lead to elevated levels of LPS endotoxins, which then lead to inflammation and metabolic issues likes diabetes and obesity (click here for background on this). In these studies, it was suggested that an impaired gut barrier or leaky gut was to blame. That is, some break down in the mechanisms and signals that control the movement of nutrients across the wall separating your gut contents from your blood went haywire and allowed LPS endotoxins to pass through. This permeability hiccup has focused on the tight junction(s) between the cells lining your intestinal tract. Think of the tight junctions like the mortar between the exposed bricks on a house. If a crack or impairment of any kind happens, stuff can leak in (of course it’s a bit a more complicated than that, but you get the idea).
In the case of our Shanghai mice, the researchers didn’t note any differences in the gene expression levels of the tight junction proteins “occluding and ZO-1” – which would have indicated a crack, if you will, in the tight junctions – between the high fat or normal chow eaters inoculated with B29. This has led them to the conclusion – albeit tentative – that the tiny LPS endotoxins are actually hitching a ride on lipoproteins called chylomicrons. These chylomicrons occur in the human gut and help transport dietary lipids (fats) from the intestines to other locations in the body. Unlike ‘stuff’ that tries to get through cracks in the tight junctions between cells (our mortar between the bricks), the chylomicrons go straight through the brick itself (enterocytes), where they then enter the body. Recent research with the gut microbiota of Zebra fish support this hitchhiker theory for the movement of endotoxins from the gut to the blood.
So what are we to make of our 26-year-old man, who came in with never-before-seen levels of endotoxin producing Enterobacter and walked out 23 weeks later with undetectable levels and over 100 pounds lighter. We know his new low-calorie diet – which was high in carbohydrates by the way – probably played some role, but it’s hard to say for sure if it gets all the credit. We do know that a strain of Enterobacter isolated from his stool (B29) was able to induce inflammation and weight gain in mice eating a high fat diet, but their genetic cousins stayed lean on a low fat diet, even though the low fat diet resulted in higher levels of inflammation-causing endotoxins in the gut. The higher levels of endotoxins in the gut (think of it as a reservoir) was not enough to force themselves into the blood. The fat in the diet seems to be the key – but only in the presence of our Enterobacter B29. If you will recall the presence of Bifidobacterium animalis did not result in obesity even under a high fat diet.
Taken together, the endotoxin-induced inflammation in the mice triggered by the B29 isolated from our big fella, demonstrate a clear line of causation of this microbe in the gut and the pathogenesis of obesity and other metabolic disorders. Following a slight modification of the Koch postulates – a set of four criteria designed to establish a causal relationship between a microbe and a disease and thus its moniker as a pathogen – it can be said with certainty that B29 is a pathogen. And therefore, ipso facto, elevated levels of Enterobacter cloacae subsp. (B29) in your gut in the presence of a high fat diet may cause an increase in your circulating levels of LPS endotoxins which will in turn increase your levels of inflammation which will definitely contribute to a cascade of metabolic disorders including, but not likely limited to, obesity and type 2 diabetes.
The most interesting question that remains is what was it in the man’s previous diet that led to such high levels of endotoxin-producing Enterobacter? A series of emails back and forth with the lead researcher on the project didn’t bring any further clarity. We can’t say that with certainty high fat diet will elevate levels Enterobacter (and others within the same family of bacteria), as the researchers demonstrated that low fat chow (NCD) produced higher levels of endotoxins in the gut than did the high fat regime (HFD). Remember, high levels of endotoxins in the gut do not equal high levels of endotoxins in your blood.
In our work in Africa, we note that among Bushmen groups, Enterobacter levels run 0-7% among various individuals – which is more on the low end of the our Shanghai friend following his 23 weeks of diet in a can. In general, the Bushmen diet is low in animal protein and fat, and low in processed food (in general). It’s also interesting to note that Enterobacter levels among a small cohort of Puerto Rican folks – who are eating more or less a western diet – is in the 12% + range (unpublished data).
In chatting with the researchers from Shanghai, it’s clear that our B29 microbe is not the only heavy-duty endotoxin-producing bacteria that is contributing to obesity and associated metabolic disorders in mice and humans. Several members within the larger Enterobactereriaceae family (within the phylum Proteobacteria) will emerge as opportunistic pathogens as well. Opportunistic in that they only cause significant damage to the host under elevated conditions which are brought on by a western diet and lifestyle (apparently). However, their mere presence – or tolerance if you will of our immune system – within our larger microbial ecosystem suggests they have co-evolved with humans in a symbiotic way. In other words, they are doing what they are supposed to do – maybe.
I’m going to go out on a limb with lots of cracks in it here and suggest that whatever mechanism that allows endotoxins to translocate from the gut to the blood, be it hitchhiking a ride on chylomicrons or via an impaired gut barrier, is exactly what is supposed to happen, and it’s an evolutionary conserved strategy to maximize lipid conservation, storage, and ultimately utilization worked out deep in our ancestral past. Though much is made of our meat eating past (set aside fat-bearing nuts for the moment), the millions of years before we possessed advanced weaponry the consistent taking of meat was highly variable – not to mention highly seasonal. This seasonality and variability of meat consumption – and thus animal fat intake – is the norm among hunter-gatherer groups today in those latitudes in Africa where our genus and Mio-Pliocene ancestors evolved.
Any biological deal worked out between our microbiome and our immune system (inflammation which triggers insulin resistance and thus promotes fat storage) that conserved and optimized seasonal or opportunistic intake of such a valuable nutrient as fat, would be positively selected throughout hominin evolution, but maladaptive in an ecosystem such as the middle class world of today – with its steady and easy access to fat resources. The issue here, I believe, is not so much the fat per se – although its more easily acquired today – but those aspects of the diet and lifestyle in the modern world that promote the over growth of opportunistic pathogens that are very efficient at producing endotoxins.
This is not a promotion of the Thrifty Gene Hypothesis put forward by James Neel, but possibly a rethinking that puts emphasis on the microbial contribution as well. Maybe the Thrifty Metagenome Hypothesis would be more fitting – metagenome being the sum of our human genome and our microbial bits and pieces from head to toe.
For obvious reasons there is an intense research interest in sorting out the underlying causes of obesity and the sooner we unhitch ourselves from the calorie is a calorie dogma, the faster we will progress. Just as great pathogens of our past had done before B29, the impact of our modern obesity epidemic will likely change the course of history for much of a planet struggling under the economic and political pressures that only multiply with each tick we move up the BMI chart.
*If you are wondering whether or not you carry Enterobacter and other ‘potentially opportunistic pathogens’ there is only one way to find out – join our effort over at American Gut.