Interview by Marion Gruner
How long has your career in microbiology been?
Well, my first paper was published in 1951. It was based on work I did as a medical student in the late ‘40s and we finally had enough data to publish it in ’51. And at that point, my wife and I and another medical student all agreed to take some antibiotics to see what the impact would be on the bowel flora, and so we documented changes, even with the crude techniques available at that time. Now, I’ve been at it ever since then. Our first anaerobe entered our stock collection in 1957, but we were working with them for several years before that.
When did you start making a connection between gut bacteria and autism?
In 1998, I had a phone call from Dr. Richard Sandler in Chicago. He’s a pediatric gastroenterologist, and he had been seeing the autistic son of a lady named Ellen Bolte. She had done a remarkable job reviewing the medical literature on autism and related diseases, and concluded this might well be a bacterial infection. And she thought the best example of this was infant botulism, wherein the organism grows in the gut of the baby, produces toxin, and it’s absorbed. It goes to the central nervous system and produces the disease. That’s in contrast to the usual type of botulism where we eat a food and it’s contaminated with the toxin from the organism and there’s no involvement of the gut at all. The disease occurs directly. So, she had recommended to Dr. Sandler that he treat her child with oral Vancomycin, reasoning that this drug would be active against the type of organism she thought might be causing the disease, and that would be a Clostridia group. They’re Gram-positive and the Vancomycin is particularly active against Gram-positive organisms. So, he agreed to do this and the child had a dramatic improvement, starting within a few days and persisting for six weeks while he was still on the drug. This involved improvement in language skills. He actually had no language beforehand, but he picked up a few words and even began to string together tiny sentences towards the end, like, “No, Mummy, I’d like this.” He was much more malleable. He would listen to people and respond. He would look at them, which was quite different from his usual behaviour. He didn’t have any fits of anger and generally was a much more, nearly normal child.
What impact would repeated antibiotic use have on gut bacteria?
Antibiotics have an impact on the bowel flora. Virtually all of them do. Some, when given systemically, do not enter the bowel, so they would not, but most antibiotics, even given systemically, are secreted into the bowel to some extent. So, virtually all the time that we use antibiotics, one of the lesser-known problems with antibiotic use is that they do impact the bowel flora. The bowel flora, we’re finding out more and more in recent years, does a number of amazing things for the body. Our innate immunity is developed by virtue to exposure from bacteria living in the gut. And we know now that at least some forms of obesity are related to changes in the bacterial flora from the norm. We don’t know that antibiotics are involved in that particular problem, but they certainly seem to be in autism and clearly are in Clostridium difficile-associated colitis. So, you have a normal complement of organisms in your gut. It varies according to your diet and if you have an immune system problem. If your immune system is not fully effective either because of genetic problems or, more commonly, toxins in the environment, then you are not able to control bacteria that get out of the gut into other parts of the body, as you would if you had a normal immune system. So, we have that kind of background, and then on top of that, the key thing is that these children are susceptible to ear infections; probably to other types of infection. And so, they receive antibiotics, and the antibiotics that are favoured for that type of infection, ear infections, do have a significant impact on the bowel flora and tend to select out certain organisms while they suppress much of the rest of the flora. And the organisms that tend to persist are Clostridia, and this accounts, in Clostridium difficile colitis, for that problem. And we think it accounts for the problem in autism as well. Maybe not due to Clostridia specifically. Recent work suggests that there’s another organism that’s much more important.
What is this new bacterium you’re looking at and how is it related to autism?
It’s quite a different bug from the Clostridia. It’s Gram-negative instead of being Gram-positive. It does not produce spores, but it’s a very virulent organism. We have seen it over the years in serious infections like blood poisoning. So, we know about the organism, but it’s difficult to grow and it’s not been cultured many times when it has been involved in infection. So, laboratories such as ours and many others that have worked with anaerobes for years can grow this and it turns out to be important in autism. We found about this when some of the more high-throughput, detailed molecular approaches to studying the bowel flora became available. So, this allows you to find at least 1,000 different organisms per gram of fecal content, whereas with the old techniques, if you stayed for some time you could eventually pick up 300 different species. And now, there’s talk that there may be as many as 10,000. So, this is a very powerful tool and using it, too, we found out that the flora of autistic children is basically quite different from that of normal control children. The organism of interest is in a phyla called Proteobacteria. Our studies suggested that Desulfovibrio, one of the Proteobacteria, might be important in autism because it was seen with some frequency; about 50% of patients with autism but not at all in any of the controls. Now, Desulfovibrio, as the name suggests, changes a sulfate compound such as a hydrogen sulfate. It desulfates them and you end up with hydrogen sulfide rather than sulfate as the principal end product of that metabolism. Hydrogen sulfide is a major toxic compound, known for centuries, actually, but not known to be important in relation to bacteria. So, that may be one mechanism of its action. Also, the cell wall of this organism, Desulfovibrio, contains a potent endotoxin, and endotoxins are known to be very damaging to people, and animal models as well. And there are probably other factors that account for its virulence.
Where is Desulfovibrio found and how could it get into the body?
It’s found in the environment. It likes the environment around oil wells, for example. And it’s such a powerful toxin producer that the oil people are concerned about it because it corrodes metal in their apparatus and is a real pest to them. We ingest it with certain foods; meats in particular and cooked meats especially. And so, it is found in the normal flora in low numbers of probably half the people. But again, it depends on the diet, so that in the United Kingdom where probably they eat more meat than we do, they have higher counts in normal individuals of Desulfovibrio. If you take antibiotics for prophylaxis or treatment of infections, and if you take certain ones that are active against common elements of the normal flora of the bowel but not active against Desulfovibrio—and that’s true for a number of compounds, antibiotic compounds—then you tend to select out Desulfovibrio. get much higher counts and so more toxins produced and then, you can get disease in that manner.
What type of autism have you been studying?
Regressive autism is the type that we have worked with exclusively. As the name suggests, these children develop normally up to about eighteen months of age, and then they begin to go backwards. They lose their social graces. They’re not warm and friendly with their parents. They don’t maintain eye contact. They have difficulty with other children of the same age. They have gastrointestinal problems that often are quite striking and may be the main feature of their illness. These include abdominal distension, abdominal pain. Constipation is a principal problem but not always obvious overtly. They may have so-called compensatory diarrhea, attempting to get rid of the obstruction from the constipation. And they, in the extreme, may bang their head against the wall; be very damaging to themselves, to their playmates and siblings, and even their parents.
Why would Clostridia and Desulfovibrio be important in regressive autism?
Well, first of all, these children are only susceptible to autism up to the age of four. Anybody that has suspect autism after the age of four, unless they’ve had it starting before the age four, it’s not autism; it’s something else. So, you have to have a central nervous system that’s being developed at a certain critical stage. Then, you have to have impairment of the immune system. This can be on a genetic basis but we think much more commonly it’s due to environmental toxins. So, that sets the stage, and then the bacteria get into it when antibiotics are given, appropriately or inappropriately. And the typical scenario for that to happen is when the child develops an ear infection. If a beta-lactam antibiotic, such as a penicillin or a cephalosporin, is given, that eliminates certain parts of the bowel flora, and that creates a niche for organisms such as Desulfovibrio and Clostridia to grow out to larger numbers where they produce enough toxin to cause the disease.
What are the reasons to wonder if this condition might be somehow “infectious”?
Yeah, this leads to consideration of how these diseases spread from one to another and why the incidence of autism should be increasing so much, and why C. difficile colitis is a big problem in hospitals. So, let’s start with C. difficile colitis as the model. Patients acquire Clostridium difficile. Again, it may be present as a normal flora. Seems to be the case in about 3% of adults. But we think it’s been pretty well established now with C. difficile colitis that when the organism contaminates the environment of a hospital, it’s a big problem and accounts for spread from patient to patient. Clostridium difficile, when it’s exposed to antibiotics, tries to protect itself. It does this by converting the ordinary vegetative bacterial cells into the spore form where they can withstand anything short of autoclaving. So, we’ve demonstrated spores from Clostridium difficile on the floor of a hospital room that was cleaned after the patient left and left without patients in it for over a month, and we were still able to culture C. difficile. They knew the study was being performed, so the housekeepers tried to do an extra good job on cleaning, and we could still find them. I think that same sort of thing is going on in the case of autism, but it remains to be documented. With Desulfovibrio. now, that organism is not a spore-former, but it has other mechanisms, various enzymes that protect it from exposure to oxygen and to other deleterious influences. So, Desulfovibrio can live in the environment for months on end in its vegetative state until conditions are improved and it has the opportunity to survive and multiply.
How does Dr. MacFabe’s work relate to yours?
We don’t have the full connection that we need to be able to say that Dr. MacFabe’s work fits in key and lock with what we’re doing, but it seems like a likely possibility. Dr. MacFabe takes certain compounds that are by-products of the activity of bacteria in the gut—we call them short-chain fatty acids—and he injects them into the brain of rats and that leads to a set of symptoms and findings that are characteristic of what we see in autism in humans. So, it may be a very good animal model for us, and so that would be very helpful. Research would be sped up considerably if we knew that was a reliable animal model and we could do a lot of manipulation of the bacteria in the gut of the rat and save time in terms of coming up with answers that apply to humans. Now, Clostridia do produce some of the compounds that MacFabe uses in his model and Desulfovibrio does not, but it does it in a roundabout way. In other words, it doesn’t produce those compounds itself, but it leads to their production in a roundabout way. So, both of those things would fit with his hypothesis and with his animal model.
Can you comment on the connection of these bacteria to the foods we eat?
Well, I think we’re beginning to appreciate that diet is extremely important. It has a definite impact on the bowel flora and the impact can be good or bad depending on the diet. I think we need to study it further, but there’s been one study done where subjects, volunteers who were not ill, agreed to go on several different diets with intervals between the diets of whatever they chose to eat normally. And that showed clear concordance with a high meat content diet, cooked meat, and overgrowth of Desulfovibrio. So, I think that’s an important lead. Also, others have found over the years that certain diets benefit autistic children. So, a gluten-free, casein-free diet, which is difficult to make, time-consuming, and expensive, but many autistic children improve to a degree—sometimes a significant degree—on that diet. There’s also a specific carbohydrate diet, which is even more difficult to make and more expensive, which is helpful to other autistic children. Ellen Bolte’s son, for example, did not respond to the gluten, casein-free diet, but did respond well to the specific carbohydrate diet.
What do you think about the future of gut bacteria research and its connection with autism?
We think that bacteria are very important in autism, as I mentioned, and knowing what bacteria do in other circumstances, it’s easy to visualize spread of these bacteria to siblings of autistic children. And we do see now more multiple cases in families than we did before and an increased frequency of autism, probably because of spread of these organisms by way of the environment and even direct transfer from one person to another. The very encouraging thing about all of this is that we know a lot about bacteria and if we know for sure which bacteria are involved in autism, we have ways of combating them. Knowing what we know about bacteria, we can visualize making a vaccine that could be given by mouth. It would be very well-tolerated in all respects and would build up antibodies and other parts of the immune system that fight Clostridia and Desulfovibrio, and could not only control it once it’s established but could prevent it entirely. So, the day may come when we have a vaccine that could be given just as any other vaccine could be given—polio vaccine, for example—to totally prevent the disease and wipe it out. But besides waiting for the vaccine, we can be working on developing these good bacteria that we know about. And we don’t know enough about them, so we have to do more research there. But we have things called probiotics. And you’ll hear and see a lot of advertisements for this yoghurt or that yoghurt, which supposedly is good in its own right and is supplemented by some of the good guys. You can hardly buy ice cream these days, especially yoghourt-based ice creams, without addition of these things. That’s because it’s a good advertising gimmick. But the scientific background for it is not solid. There’s no research to prove that these foods can do something good for autism, for example. But there are ways of putting good bacteria back into people to fight the bad ones, and as soon as we have enough research done to know which particular types of bacteria, and sub-types, are the absolute good ones which would not backfire on us, then we could put them into probiotics.
Not long ago people thought autism was a permanent condition but, now, are there suggestions that it may be at least party reversible?
The reversibility of autism is a very exciting proposition. As I indicated, children who have had it for even a year or so, within two to three weeks of antibiotic therapy are much better, and it’s within the realm of possibility that we can maintain the improvement and take it all the way back to a totally normal child if we treat them early enough. There’s been a lot in the lay press recently about the importance of diagnosing autism early and suggestions have been made to both physicians and parents as to how one can suspect it and verify the diagnosis. I’m afraid that as the child gets older, when they get into their teens and, particularly, into adulthood, it may be much less reversible.