Medical Hypotheses 80 (2013) 264–270

Contents lists available at SciVerse ScienceDirect

Medical Hypotheses
journal homepage: www.elsevier.com/locate/mehy

Hypothesis for a systems connectivity model of autism spectrum disorder

pathogenesis: Links to gut bacteria, oxidative stress, and intestinal permeability
Colin A. Heberling a, Prasad S. Dhurjati a,⇑, Myron Sasser b
a
University of Delaware, Department of Chemical and Biomolecular Engineering, 150 Academy Street Newark, DE 19716, USA
b
MIDI, Inc., 125 Sandy Drive Newark, DE 19713, USA

a r t i c l e i n f o a b s t r a c t

Article history: Autism Spectrum Disorders are neurodevelopmental disorders with symptoms that include cognitive
Received 10 September 2012 impairments, stereotyped behaviors, and impairments in social skills. The dramatic increase in incidence
Accepted 29 November 2012 of autism in recent years has created an increased need to find effective treatments. This paper proposes a
hypothesis for a systems model of the connections between Autism Spectrum Disorder pathogenesis
routes observed in recent studies. A combination treatment option is proposed to combat multiple path-
ogenesis mechanisms at once. Autism has been cited as being linked to gastrointestinal symptoms and is
thought to be caused by a combination of genetic predisposition and environmental factors. Neuroin-
flammation as a result of increased gastrointestinal permeability has been noted as being a likely cause
of Autism Spectrum Disorders, with possible primary causes stemming from abnormal intestinal bacteria
and/or sulfur metabolic deficiencies. Our pathogenesis model proposes a circular relationship: oxidative
stress and sulfur metabolic deficiencies could cause changes in colonic bacterial composition; and envi-
ronmental bacterial contaminants could lead to elevated oxidative stress in individuals. It would thus be
a self-perpetuating process where treatment options with single targets would have short-lived effects. It
is believed that bacterial toxins, oxidative stress and dietary allergens such as gluten could all lead to
increased epithelial permeability. Therefore, we propose a combination treatment to combat intestinal
permeability, abnormal bacteria and/or bacterial overgrowth, and sulfur metabolic deficiencies. It is
our hope that the proposed model will inspire new studies in finding effective treatments for individuals
with Autism Spectrum Disorders. We suggest possible future studies that may lend more credibility to
the proposed model.
Ó 2012 Elsevier Ltd. All rights reserved.

Introduction
Autism Spectrum Disorders (ASDs) are neurodevelopmental
disorders that include Autism, Asperger’s Syndrome, and Pervasive
Developmental Disorder Not Otherwise Specified (PDD-NOS). It
has implications in cognitive impairments, sociability impair-
ments, impairments in language and communication skills, re-
stricted interests, and stereotyped behaviors [1]. Currently,
observations of these behaviors are the best and only way to diag-
nose ASDs. A diagnosis is usually made before a child reaches age 3,
and diagnoses have dramatically increased in the past several
years. Current estimates state that 1 in 88 (11.4 in 1000) children
in the US in 2008 had some form of autism, up from 9 in 1000 in
2006 and 6.4 in 1000 in 2002. Incidence of autism could possibly
⇑ Corresponding author. Address: Department of Chemical Engineering, 150
Academy Street, University of Delaware, Newark, DE 19716, USA. Fax: +1 302 831
1048.
E-mail address: dhurjati@udel.edu (P.S. Dhurjati).
be more now. ASDs are about four times more prevalent in males
than females, for reasons that are still unclear.
Gastrointestinal problems have been implicated in people with
ASDs by some, but there is much discrepancy on this topic: some
studies seem to confirm this correlation as significant when com-
pared to healthy, age-matched controls [2–4], and yet others state
that the differences in gastrointestinal problems in ASD patients
are not statistically significant [5–7]. Despite these discrepancies,
when ASD patients are put on gluten and casein exclusion diets
there is usually a partial alleviation of symptoms soon afterwards
[2]. Gluten allergies are pervasive in Celiac Disease and type 1 dia-
betes [8–10]. It is an autoimmune condition, thought to be aggra-
vated by a weakening of the small or large intestine’s epithelial
barrier function, intended to limit trafficking of cells and molecules
between the gut and blood circulation. This suggests that there is a
link between the gut and autism pathogenesis. Gastrointestinal
problems (such as diarrhea, constipation, ulcerative colitis) might
not be present in all people with ASDs, but there does seem to
be a link between gastrointestinal symptom severity and autism
severity, which gives more evidence for the gut-to-brain link [2].

0306-9877/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.mehy.2012.11.044
 C.A. Heberling et al. / Medical Hypotheses 80 (2013) 264–270 265

Proposed pathogenesis
Fig. 1, depicted below, represents a summary of our proposed
systems connectivity model for autism spectrum disorder patho-
genesis. A review of the recent literature is necessary to under-
stand the full implications and reasoning behind the model and
will be addressed after the introduction of the model. The model
is quite complex, involving both linear and cyclic relationships.
The model can be broken down into three main sections or patho-
genesis routes: oxidative stress and subsequent sulfur metabolic
deficiencies; abnormal intestinal bacteria and/or bacterial over-
growth; and increased intestinal permeability leading to increased
blood circulation of molecules and cells normally contained within
the gut and subsequent neuroinflammation after breach of the
blood–brain barrier. The relationships after an increase in intesti-
nal permeability are simply linear. The chief new theory that we
propose involves the inter-relationship between the first two
pathogenesis routes: sulfur metabolic deficiencies and intestinal
bacterial composition imbalance. We hypothesize a connection
between various intestinal bacteria involved in ASD. Overgrowth
of Clostridia and/or Desulfovibrio and a decrease in Bifidobacteria
could be involved in ASD pathogenesis, and sulfur metabolic
deficiencies found in individuals with ASDs may be causing over-
growth of Desulfovibrio. The sulfur metabolic deficiencies may be

Fig. 1. Proposed pathogenesis model. Note: ‘‘Met’’ and ‘‘Cys’’ are abbreviations for amino acids Methionine and Cysteine respectively. ‘‘SAH’’ is an abbreviation for S-
adenosylhomocysteine, a metabolite in the transsulfuration cycle.
266 C.A. Heberling et al. / Medical Hypotheses 80 (2013) 264–270
self-perpetuated by decreased levels of total and reduced
glutathione, and a decrease in Bifidobacteria.
The inter-relationships between these three major pathogenesis
routes suggests that a combination treatment may be the best op-
tion for those with Autism. Assuming that the proposed cyclic rela-
tionship is true, then treatment of only one of these problems may
only have short-lived effects. Treating all three problems at once
may have better efficacy. The pathogenesis mechanism would then
be ‘‘knocked out’’ at three different places, which would minimize
the mechanisms’ inter-related effects.
The following descriptions involve more explicit details of the
systems connectivity model, which will become clearer in the liter-
ature review sections of this paper. Impairment in the gut epithe-
lial barrier function is proposed to be critical in development
of autism so that toxins, bacterial products, lymphocytes, pro-
inflammatory cytokines, and neurotransmitters can reach blood
circulation and breach the blood–brain barrier, such as through
endotoxin (otherwise known as lipopolysaccharide, or LPS) [11].
In the systems connectivity model, three items lead to increased
epithelial permeability: gut inflammation, increased zonulin levels
such as from an allergic reaction to gluten, and from cell morpho-
logical changes from clostridial toxins. Furthermore, six factors
are proposed as leading to gut inflammation in autism pathogene-
sis: clostridial toxins, hydrogen sulfide from Desulfovibrio, dietary
allergens such as gluten, reduced removal of toxic heavy metals,
reduced sulfation detoxification of xenobiotic compounds, and
an inflammatory cytokine imbalance skewed towards pro-
inflammatory cytokines.
It should be noted that particular mechanisms for environmen-
tal exposure to Clostridia and Desulfovibrio are not addressed in the
connectivity model below because that is not the model’s main fo-
cus. Use of antibiotics that Clostridia and/or Desulfovibrio are resis-
tant to as well as poor diet choices may be factors that lead to
environmental contamination with pathogenic Clostridia and Des-
ulfovibrio. More information on this topic is given in the following
sections, but for a more comprehensive coverage (including rela-
tional diagrams), see the 2011 and 2012 papers written by Fine-
gold et al. [12,13].
Literature review
Abnormal gut bacteria
Given the gut-to-brain link in ASDs, there have been numerous
studies on intestinal bacteria composition in relation to ASDs.
Anaerobic bacteria naturally reside in the human large intestine
(and similar mammals) acting as another collective ‘‘organ’’ of
the human body [14–16,8]. They ferment polysaccharides from
the diet that are indigestible by human enzymes and produce short
chain fatty acids (SCFAs), mainly acetate, propionate, and butyrate.
The main bacterial genera responsible for SCFA production are Bac-
teroides, Lactobacillus, and Bifidobacterium. SCFA production is ben-
eficial for the human host, providing energy for gut epithelial cells
and influencing the immune system. Human infants are born ster-
ile, and quickly gain bacteria that are essential for natural develop-
ment of the immune system. A proper balance of gut bacteria
composition is important as some induce the production of anti-
inflammatory cytokines such as Bifidobacterium and Lactobacillus,
and others are known to induce production of pro-inflammatory
cytokines, such as members of Clostridium and Ruminococcus.
Studies where fecal samples from ASD patients were analyzed
for microbial composition have had differing results. Many showed
presence of abnormal Clostridia taxa and increased growth of Clos-
tridia in general along with decreased numbers of Bifidobacteria
[17–20,4]. Siblings of individuals with ASDs were shown to have
intermediate changes in gut bacteria compositions, though the dif-
ferences were not statistically significant from individuals with
ASDs or controls. Therefore, Finegold hypothesized that clostridial
spores as an environmental contaminant were largely to blame for
the dramatic increase in regressive autism [21]. Clostridium difficile
is the cause of hospital-contracted ulcerative colitis and is a known
toxin producer [22]. Other members of Clostridia are known toxin
producers as well, such as C. perfringens. Finegold et al. cite evi-
dence for clostridial pathogenesis in that infants who have ear
infections are typically prescribed the antibiotics sulfamethoxazole
and trimethoprim [17,18,21]. Clostridia are Gram-positive bacteria
that are usually resistant to these antibiotics, whereas Bacteroides
and Bifidobacteria are Gram-negative bacteria that are susceptible
to such antibiotics. Therefore, it is reasonable to assume that the
result would be a shift in gut bacterial composition in these indi-
viduals. They would also be much more susceptible to contracting
more harmful bacterial strains from their environments, and their
immune systems would be compromised.
Finegold et al. cite additional evidence in that individuals with
ASDs respond well to the antibiotics vancomycin and metronida-
zole, which are able to target Gram-positive bacteria. Vancomycin
is also minimally absorbed by the gastrointestinal tract, providing
maximum efficacy to antibiotic activity in the colon. However,
after termination of antibiotic treatment, patients have recurring
symptoms a few weeks afterwards [23]. Clostridia produce spores:
dormant organisms with copies of their DNA that vegetate, or acti-
vate, when the environment is favorable for growth again [21,24].
These spores are thought to cause the recurring symptoms. Spores
are highly resistant to a number of factors that would ordinarily
kill vegetative cells, such as antibiotics, heat exposure, changes in
pH, substrate starvation, or physical pressure. They can thus exist
outside of the host and stay dormant on the surfaces of our exter-
nal environments until they are able to contaminate another host.
This is thought to be an important factor in the increased incidence
of ASDs.
In the most recent study by Finegold et al. [19], more advanced
pyrosequencing technology was used to analyze fecal samples of
ASD patients. The results of this study have conflicted with the past
studies of Finegold et al. and of many others. Amounts of Clostridia
were not statistically higher than controls although relative
amounts of Bacteroides and Desulfovibrio were higher than controls.
Desulfovibrio was of special note, because even though it repre-
sented a very small proportion of the total bacterial population
(less than 0.3%), its proportion was 10 times higher in individuals
with ASDs versus healthy controls. Finegold wrote a paper [2]
hypothesizing that Desulfovibrio was in fact responsible for the
autism pathogenesis originally thought to be caused by Clostridia.
Desulfovibrio produces hydrogen sulfide as its main metabolic
product, which is toxic to humans, and it is highly resistant to
many of the same antibiotics that Clostridia are, originating at least
in part from beta-lactamase inhibitors produced by the organism.
Common antibiotics such as penicillins rely on the activity of
beta-lactamases. However, it is hard to neglect the number of
studies pointing to Clostridial species as being a part of ASD patho-
genesis, so perhaps both Clostridia and Desulfovibrio contribute.
Sulfur metabolism deficiencies
In the quest to find reliable biomarkers for ASDs, samples taken
from blood sera and urine have revealed much about the possible
pathogenesis of autism. Among the results are evidence that many
individuals with ASDs have deficiencies in the transmethylation
and transsulfuration metabolic pathways, associated with the
metabolism of sulfur-containing amino acids methionine and cys-
teine [25,26]. Methionine is an ‘‘essential’’ amino acid, meaning
that it must be supplied in the diet, and it is then required to
 C.A. Heberling et al. / Medical Hypotheses 80 (2013) 264–270 267

synthesize cysteine. Fig. 2 taken from James et al. [25] depicts the
transmethylation and transsulfuration pathways. The metabolites
adenosine and SAH (S-adenosyl homocysteine) were found in ele-
vated amounts compared to controls, and methionine, SAM (S-
adenosyl methionine), homocysteine, cystathionine, and cysteine
were found in depleted amounts compared to controls. The results
suggest that there is a metabolic bottleneck at SAH and adenosine.
Oxidative stress is thought to be the cause, as oxidative stress is
known to inhibit methionine synthase activity, and adenosine
takes up active sites on SAH to protect against oxidative stress,
inhibiting the conversion from SAH to homocysteine. Homocyste-
ine is required for re-methylation back to methionine and for
transsulfuration conversion to cysteine. A metabolic bottleneck at
SAH would account for the depletions seen in the other
metabolites.
This metabolic deficiency has several harmful implications. Cys-
teine is the rate-limiting step in glutathione production, the body’s
natural anti-oxidant and a major player in excretion of toxic heavy
metals. James et al. also found decreased amounts of glutathione in
ASDs and a lower ratio of reduced glutathione to oxidized glutathi-
one, indicating oxidative stress. As glutathione is the body’s major
anti-oxidant, the metabolic inhibition would then self-perpetuate
and cause further deficiency. With the dual responsibility of gluta-
thione to help in removal of toxic heavy metals, individuals with
ASDs and the metabolic deficiencies would be more susceptible
to the toxic effects of heavy metals. Additional evidence for im-
paired heavy metal excretion in individuals with ASDs is that urine
samples taken from those with ASDs studied had significantly less
amounts of mercury than controls [27].
Methionine is required for methylation of key body molecules,
such as DNA, RNA, proteins, and lipids. For example, methylation
of DNA results in ‘‘silencing’’ of specific genes, inhibiting expres-
sion of those genes [28]. A deficiency in methionine would there-
fore disrupt gene expression and could possibly cause problems
associated with autism or other ailments. Studies on intestinal bac-
teria and genetics have revealed that genetics can dictate bacterial
compositions [29,30]. Even identical twins separated for several
years still have remarkably similar gut bacteria compositions. With
this evidence it would then seem possible that methionine defi-
ciencies could cause shifts in intestinal bacteria compositions in
individuals with ASDs, which may lead to other problems.
Individuals with ASDs have also been shown to have impaired
capacity to detoxify certain xenobiotic compounds, especially phe-
nolic compounds [31,26]. One such example is paracetamol (more
commonly known as acetaminophen). Detoxification of these com-
pounds requires transfer of a sulfate moiety to the xenobiotic com-
pound. Cysteine is required for synthesis of sulfate, so oxidative
stress may be causing this impairment as well. Individuals with
ASDs are thus often more susceptible to the toxic effects of oxida-
tive stress, heavy metals, and phenolic compounds as result of ini-
tial oxidative stress.
Testosterone has been shown to impair the transsulfuration
pathway as well, which may be a reason why ASDs are so much
more prevalent in males than in females [26]. Those with ASDs
have also been shown to have increased levels of testosterone
compared to controls.
‘‘Leaky gut’’ syndrome
‘‘Leaky Gut’’ Syndrome is a term given to the condition where
the epithelial barrier function of the small or large intestine is im-
paired, causing less discrimination in the numbers and types of
molecules and cells that are able to pass from the gut to the circu-
latory system and vice versa [32–34]. Tight junctions are responsi-
ble for the epithelial barrier function, and consist of a system of
several proteins in the paracellular space between each cell in
the gut lining. Increased gut permeability has been implicated in
ASDs, and is thought to be the link between the gut and brain in
autism pathogenesis. For example, propionic acid, which is benefi-
cial in the colon when fermented by intestinal bacteria [35], has
been shown to cause neuroinflammation when injected into the
brains of rats and cause symptoms similar to ASDs [36]. The study
was conducted because food preservatives containing propionic
acid have anecdotally resulted in increased symptoms in individu-

Fig. 2. (From James et al., 2004, with permission): Methylation and transsulfuration pathway. The methionine cycle involves the remethylation of homocysteine to
methionine by either the folate-vitamin B-12-dependent methionine synthase (MS) reaction or the folate-vitamin B-12-independent betaine homocysteine methyltrans-
ferase (BHMT) reaction. Methionine is then activated by methionine adenosyltransferase (MAT) to S-adenosyl-methionine (SAM), the major methyl donor for cellular
methyltransferase (MTase) reactions. After methyl group transfer, SAM is converted to S-adenosylhomocysteine (SAH), which is further metabolized in a reversible reaction
to homocysteine and adenosine. Adenosine may be phosphorylated to adenosine nucleotides by adenosine kinase (AK) or catabolized to inosine by adenosine deaminase
(ADA). Homocysteine may be permanently removed from the methionine cycle by irreversible conversion to cystathionine by vitamin B-6-dependent cystathione b-synthase
(CBS). Cystathionine is converted to cysteine, which is the rate-limiting amino acid for the synthesis of the tripeptide glutathione (Glu-Cys-Gly). THF, tetrahydrofolate; 5-CH3
THF, 5-methyltetrahydrofolate; SAHH, SAH hydrolase.
268 C.A. Heberling et al. / Medical Hypotheses 80 (2013) 264–270
als with ASDs. With increased gut permeability, it would then
seem logical that propionic acid produced by gut bacteria would
then be able to leak into the bloodstream and through some mech-
anism breach the blood–brain barrier
There are several mechanisms that can cause increased gut per-
meability, such as intestinal inflammation [33]. Intestinal inflam-
mation can result from a response to toxins, such as those from
Clostridia, Desulfovibrio, lipopolysaccharide (LPS, or endotoxin)
from the cell walls of Gram-negative bacteria like Bacteroides, or
quite possibly heavy metals and phenolic compounds that haven’t
been sulfated. With the increased prevalence of Clostridia and de-
creased prevalence of Bifidobacteria found in some individuals with
ASDs, pro-inflammatory cytokines would be more abundant. This
imbalance in inflammatory cytokines results in an impaired capac-
ity to resolve inflammation after a threat [8]. This would allow in-
creased gut permeability for a longer amount of time, and possibly
cause tissue damage, another cause of increased gut permeability.
Food allergens like gluten can also cause increased gut perme-
ability. In fact, it was found that a molecule termed ‘‘zonulin’’
was pinpointed as being a part of the cell signaling mechanism in-
volved in the loosening of tight junctions [10]. Zonulin was found
to be in elevated amounts in sera of individuals with autoimmune
conditions such as Celiac Disease, Crohn’s Disease, Irritable Bowel
Syndrome, and Type 1 Diabetes. First-degree relatives had moder-
ately elevated levels of zonulin. Watts et al. showed that a compet-
itive inhibitor of this mechanism, larazotide acetate, was shown to
nullify the effects of gluten on rats predisposed to development of
type 1 diabetes. After exposure to gluten, the control group devel-
oped type 1 diabetes and the group given the zonulin inhibitor did
not. This suggests that increased gut permeability is a factor that
determines and precedes the development of type 1 diabetes. As
individuals with ASDs have also been shown to have an impaired
epithelial barrier function and allergies to gluten it is not unrea-
sonable to assume that zonulin may be important in autism path-
ogenesis as well. It would then be especially useful to test
individuals with ASDs for abnormal zonulin levels and study the
effects of its inhibitor.
Clostridial toxins such as those from Clostridium difficile have
been shown to be able to cause gut permeability by inducing
changes in morphology of gut epithelial cells [37]. Epithelial cells
are rectangular, with very little paracellular space between them.
Clostridium difficile toxins cause epithelial cells to ‘‘round-up’’,
through inhibition of the effects on the host’s Rho-GTPases (i.e.
guanosine triphosphatases), creating a substantial increase in para-
cellular space. Other less-studied clostridial species found in indi-
viduals with ASDs may have similar virulence factors that impair
the epithelial barrier function.
Endotoxin from cell walls of Gram-negative bacteria has been
shown to be able to breach the blood–brain barrier [11]. Thus if
there is a means for endotoxin to leak out of the large intestine,
then it could allow molecules and cells to enter the cerebrospinal
fluid, causing neuroinflammation. These molecules and cells could
include: any of the toxins that induced an immune response in the
first place; pro-inflammatory cytokines; or host immuno-regula-
tory cells themselves (i.e. lymphocytes).
Background on Desulfovibrio
With Finegold et al.’s finding of increased incidence and abun-
dance of Desulfovibrio in those with ASDs it becomes imperative
to learn more about this organism’s physiology, background, and
ecological effects on other intestinal bacteria. Desulfovibrio is a nat-
urally occurring organism in the human gut [38]. Fermentation of
polysaccharides produce large amounts of hydrogen gas as a side
product. Through similar biochemical reactions, sulfate-reducing
bacteria such as Desulfovibrio and methanogenic archaea consume
hydrogen gas. Individuals usually have one or the other, but if sul-
fate-reducing bacteria are present, they will usually out-compete
methanogens as their reaction for removal of hydrogen is more
thermodynamically favorable. The end product of methanogens is
methane, which is non-toxic, whereas the end product of sulfate-
reducing bacteria is hydrogen sulfide, which is toxic to humans.
Unlike most intestinal bacteria that have limited ability to
catabolize sulfur-containing compounds [39], sulfur metabolism
is quite extensive in Desulfovibrio, similar to that of humans [40].
Sulfate is its terminal electron acceptor (in place of oxygen as in
aerobic organisms), and given access to a metabolite in the methi-
onine and/or cysteine pathways it can synthesize cysteine and then
through a few more conversions synthesize sulfate. With the sulfur
metabolic deficiencies found in individuals with ASDs, it may be
that those with ASDs have an increased need for Desulfovibrio. This
coupled with the elevated amounts of SAH (with limited use be-
cause of the metabolic bottleneck) found in individuals with ASDs
(which Desulfovibrio can use to foster its own growth) may be a
possible cause for the increased amounts of Desulfovibrio found
in Finegold et al.’s study.
In a study on Desulfovibrio desulfuricans conducted in two
anaerobic bioreactors (chemostats) with 14 common intestinal
bacterial species, it was found that Desulfovibrio could cause signif-
icant changes in the bacterial ecosystem [41]. Introduction of Des-
ulfovibrio and sulfates in the growth media to one of the
chemostats resulted in a ten-fold reduction in Bifidobacterium lon-
gum and B. pseudolongum, thought to be two of the most important
and most beneficial Bifidobacteria species (B. longum and B. adole-
scentis predominate in adults) [41]. No changes were seen in Bacte-
roides, and mixed results were seen in Clostridia; one clostridial
species increased ten-fold, while another one decreased ten-fold.
The results seen in changes on Clostridia seem inconclusive, given
the diversity of Clostridia seen in humans and individuals with
ASDs especially. Further studies including more diverse popula-
tions of Clostridia would be useful.
The inclusion of Desulfovibrio in one of the chemostats also re-
sulted in the formation of a biofilm on the walls of the chemostat,
dominated by Desulfovibrio and Bacteroides. This suggests that fecal
analysis of gut bacteria compositions may not be sufficient in indi-
viduals with ASDs. Intestinal biopsies are usually avoided if possi-
ble because they are much more invasive than fecal analyses, and
therefore patient compliance is decreased. In at least one recent
study on ASDs, intestinal biopsies were elected for measurement
of bacterial composition instead of fecal analyses [20]. The authors
cited that fecal matter only has about 50% viable cells, and may be
a better representation of what cells are excreted from the body
rather than the true bacterial composition within the colon. Wil-
liams et al. cited this as to why they found lower levels of SCFAs
in individuals with ASDs (seemingly contradicting MacFabe et al.
on effects of propionic acid in the brains of rats) and why they
found lower levels of Bacteroides in the biopsy and Finegold et al.
found higher amounts of Bacteroides in their pyrosequencing fecal
analysis. The technique used in the study (Williams et al.) may not
have been able to detect individual species and less common gen-
era (such as Desulfovibrio) as well as the pyrosequencing technique
used by Finegold et al. [19], but the general results thus may be
more representative of individuals with ASDs. The biopsy revealed
increased amounts and diversity of Clostridia over controls and de-
creased amounts of Bifidobacterium and Bacteroides. There very
well could have been increased abundance of Desulfovibrio that
went undetected in this study. Intestinal biopsies should be
encouraged to study this further and see if there are significant dif-
ferences in bacterial composition observed between the two meth-
ods in ASD patients.
From the chemostat study [41], it seems likely that elevated
growth of Desulfovibrio could cause compositional changes in at
 C.A. Heberling et al. / Medical Hypotheses 80 (2013) 264–270
least Bifidobacterium that could lead to autism pathogenesis. In an-
other study, strains of Bifidobacterium and Lactobacillus were
shown to have anti-oxidative effects, with the effects of Bifidobac-
terium being greater than those of Lactobacillus [42]. It would seem
that a decrease in amounts of Bifidobacterium would result in an in-
creased susceptibility to oxidative stress, making the sulfur meta-
bolic deficiencies and possibly enhanced growth of Desulfovibrio
(and hence even more inhibited growth of Bifidobacterium) even
greater. Perhaps this compositional shift would allow invasion by
harmful clostridial species and strains as well.
Treatment options & future studies
ASD pathogenesis is complex, and may involve both environ-
mental and genetic factors, and differences in data between studies
may suggest that there may be more than one possible cause. The
best treatment options may need to be personalized according to
individual differences. In past studies some effective treatments
have been found. James et al. treated patients with dietary supple-
ments folinic acid, betaine, and methyl vitamin B12, the combina-
tion of which restored transmethylation and transsulfuration
metabolites to similar levels to that of controls, resulting in
improvements of symptoms [25]. Finegold proposed treating clos-
tridial overgrowth with antibiotics vancomycin or metronidazole
[21], and Desulfovibrio overgrowth with aztreonam and a beta-lac-
tamase inhibitor [12]. Watts et al. [10] have successfully prevented
onset of type 1 diabetes in diabetes-prone rats using a competitive
antagonist to zonulin, larazotide acetate, to prevent loosening of
tight junctions involved in maintaining the gut epithelial barrier
function. All of these interventions could have great potential in
treating autism symptoms. Perhaps a combination of treatments
would result in a greater improvement of symptoms. The idea
would be to stop autism pathogenesis at multiple different steps.
In this way the sulfur metabolic deficiencies, abnormal intestinal
bacteria and bacterial overgrowth, and increased gut permeability
would all be targeted in the treatment. In particular, if the pro-
posed model were true that the metabolic deficiencies and prob-
lems associated with intestinal bacteria could both lead to one
another, then treatment of only one would have short-lived
improvements in symptoms.
Future studies should therefore incorporate studying the effects
when the suggested treatments are combined, and compare them
to the effects observed when only one treatment method is used.
The use of designed probiotics or prebiotics might also be useful.
Thus far though, probiotics have had limited effect on individuals
with autism [2]. Perhaps combining probiotic/prebiotic therapy
with the other therapies would improve their efficacy.
There are still several things that are uncertain in the proposed
systems connectivity model, and further studies could remedy
that. Further chemostat studies would be helpful in reproducing
the results of Newton et al. [41] and add to it by studying the eco-
logical effects that Desulfovibrio has on a more diverse consortium
of Clostridial species. Observing Desulfovibrio levels before and after
treatment with the dietary supplements described in James et al.
[25] might be useful in establishing a link between sulfur metabo-
lite depletion and Desulfovibrio overgrowth. In general, what is
needed is a more inter-disciplinary approach to the study of autism
pathogenesis mechanisms. And of course, further testing of thera-
pies should be beneficial in finding effective treatments for individ-
uals with autism spectrum disorders.
Conflict of interest
There are to the best knowledge of the authors no conflicts of
interests associated with this manuscript.
269
Acknowledgement
None.
References
[1] Baio J. Prevalence of Autism Spectrum Disorders – Autism and Developmental
Disabilities Monitoring Network, 14 Sites, United States, 2008. Centers for
Disease Control and Prevention, Surveillance Summaries (2012) 61(SS03): 1–
19.
[2] Adams JB, Johansen LJ, Powell LD, Quig D, Rubin RA. Gastrointestinal flora and
gastrointestinal status in children with autism – comparisons to typical
children and correlation with autism severity. BMC Gastroenterol 2011;11:22.
[3] Angelidou A, Alysandratos K, Asadi S, et al. Brief Report: ‘‘Allergic Symptoms’’
in children with autism spectrum disorders more than meets the eye? J Autism
Dev Disord 2011;41:1579–85.
[4] Parracho HMRT, Bingham MO, Gibson GR, McCartney AL. Differences between
the gut microflora of children with autistic spectrum disorders and that of
healthy children. J Med Microbiol 2005;54:987–91.
[5] Fernell E, Fagerburg UL, Hellström PM. No evidence for a clear link between
active intestinal inflammation and autism based on analyses faecal
calprotectin and rectal nitric oxide. Acta Paediatr 2007;96:1076–9.
[6] Ibrahim SH, Voigt RG, Katusic AK, Weaver AL, Barbaresi WJ. Incidence of
gastrointestinal symptoms in children with autism in Olmsted county,
Minnesota, 1976–1997: a population-based study. Pediatrics
2009;124(2):680–6.
[7] Sandhu B, Steer C, Golding J, Emond A. The early stool patterns of young
children with autistic spectrum disorder. Arch Dis Child 2009;94:497–500.
[8] Kranich J, Maslowski KM, Mackay CR. Commensal flora and the regulation of
inflammatory and autoimmune responses. Semin Immunol 2011;23:139–45.
[9] Tommasini A, Not T, Ventura A. Ages of celiac disease: from changing
environment to improved diagnostics. World J Gastroenterol
2011;17(32):3665–71.
[10] Watts T, Berti I, Sapone A, et al. Role of the intestinal tight junction modulator
zonulin in the pathogenesis of type I diabetes in BB diabetic-prone rats. Proc
Natl Acad Sci USA 2005;102:2916–21.
[11] Theije CG, Wu J, Silva SL, et al. Pathways underlying the gut-to-brain
connection in autism spectrum disorders as future targets for disease
management. Eur J Pharmacol 2011;668:S70–80.
[12] Finegold SM. Desulfovibrio species are potentially important in regressive
autism. Med Hypotheses 2011;77:270–4.
[13] Finegold SM, Downes J, Summanen PH. Microbiology of regressive autism.
Anaerobe 2012;18:260–2.
[14] Bäckhed F, Ley RE, Sonnenburg JL, Peterson DA, Gordon JI. Host-bacterial
mutualism in the human intestine. Science 2005;307:1915–20.
[15] O’Hara AM, Shanahan F. The gut flora as a forgotten organ. EMBO Rep
2006;7(7):688–93.
[16] Sears C. A dynamic partnership: celebrating our gut flora. Anaerobe
2005;11:247–51.
[17] Finegold SM, Molitoris D, Song Y, et al. Gastrointestinal microflora studies in
late-onset autism. Clin Infect Dis 2002;35(Suppl 1):S6–S16.
[18] Song Y, Liu C, Finegold SM. Real-time PCR quantitation of clostridia in feces of
autistic children. Appl Environ Microbiol 2004;70(11):6459.
[19] Finegold SM, Dowd SE, Gontcharova V, et al. Pyrosequencing study of fecal
microflora of autistic and control children. Anaerobe 2010;16:444–53.
[20] Williams BL, Hornig M, Buie T, et al. Impaired carbohydrate digestion and
transport and mucosal dysbiosis in the intestines of children with autism and
gastrointestinal disturbances. PLoS ONE 2011;6(9):e24585.
[21] Finegold SM. Therapy and epidemiology of autism-clostridial spores as key
elements. Med Hypotheses 2008;70:508–11.
[22] Bartlett JG, Gerding DN. Clinical recognition and diagnosis of clostridium
difficile infection. Clin Infect Dis 2008;46:S12–8.
[23] Sandler RH, Finegold SM, Bolte ER, et al. Short-term benefit from oral
vancomycin treatment of regressive-onset autism. J Child Neurol
2000;15:429–35.
[24] Hogan M. Bacteria. Encyclopedia of Earth. Eds. Sidney Draggan and C.J.
Cleveland, National Council for Science and the Environment (2012),
Washington DC.
[25] James SJ, Cutler P, Melnyk S, et al. Metabolic biomarkers of increased oxidative
stress and impaired methylation capacity in children with autism. Am J Clin
Nutr 2004;80:1611–7.
[26] Geier DA, Kern JK, Garver CR, Adams JB, Audhya T, Geier MR. A prospective
study of transsulfuration biomarkers in autistic disorders. Neurochem Res
2009;34:386–93.
[27] Wang L, Angley MT, Gerber JP, Sorich MJ. A review of candidate urinary
biomarkers for autism spectrum disorders. Biomarkers 2011;16(7):537–52.
[28] Phillips, T. The role of methylation in gene expression. Nature, Education
(2008) 1(1).
[29] Dicksved J, Halfvarson J, Rosenquist M, et al. Molecular analysis of the gut
microbiota of identical twins with Crohn’s disease. The ISME J 2008;2:716–27.
[30] Katchatryan ZA, Ktsoyan ZA, Manukayan GP, Kelly D, Ghazaryan KA, Aminov
RI. Predominant role of host genetics in controlling the composition of Gut
Microbiota. PLoS ONE 2008;3(8):e3064.
270 C.A. Heberling et al. / Medical Hypotheses 80 (2013) 264–270
[31] Alberti A, Pirrone P, Elia M, Waring RH, Romano C. Sulphation deficit in
‘‘Low-Functioning’’ autistic children: a pilot study. Biol Psychiat 1999;46:
420–4.
[32] Fasano A. Leaky gut and autoimmune diseases. Clinic Rev Alleg Immunol
2012;42:72–8.
[33] Teshima CW, Meddings JB. The measurement and clinical significance of
intestinal permeability. Curr Gastroenterol Rep 2008;10:443–9.
[34] Liu Z, Li N, Neu J. Tight junctions, leaky intestines, and pediatric diseases. Acta
Paediatr 2007;94:386–93.
[35] Al-Lahham SH, Peppelenbosch MP, Roelofsen H, Vonk RJ, Venema K. Biological
effects of propionic acid in humans; metabolism, potential applications and
underlying mechanisms. Biochim Biophys Acta 2010;1801:1175–83.
[36] MacFabe DF, Cain NE, Boon F, Ossenkopp KP, Cain DP. Effects of the enteric
bacterial metabolic product propionic acid on object-directed behavior, social
behavior, cognition, and neuroinflammation in adolescent rats: relevance to
autism spectrum disorder. Behav Brain Res 2011;217:47–54.
[37] Aktories K, Just I. Clostridial Rho-Inhibiting protein Toxins. CTMI
2005;291:113–45.
[38] Gibson GR, Macfarlane GT, Cummings JH. Sulphate reducing bacteria and
hydrogen metabolism in the human large intestine. Gut 1993;34:437–9.
[39] Smith EA, Macfarlane GT. Dissimilatory amino acid metabolism in human
colonic bacteria. Anaerobe 1997;3:327–37.
[40] Kyoto Encyclopedia of Genes and Genomes (KEGG). Desulfovibrio: cysteine
and methionine metabolisms. <genome.jp/keg-bin/show_pathway?dde00270>
[accessed May, 2012].
[41] Newton DF, Cummings JH, Macfarlane S, Macfarlane GT. Growth of a human
intestinal Desulfovibrio desulfuricans in continuous cultures containing defined
populations of saccharolytic and amino acid fermenting bacteria. J Appl
Microbiol 1998;85:372–80.
[42] Lin MY, Chang FJ. Antioxidative effect of intestinal bacteria Bifidobacterium
longum ATCC 15708 and Lactobacillus acidophilus ATCC 4356. Dig Dis Sci
2000;45(8):1617–22.