Hypothesis: 2 Major Environmental and Pharmaceutical Factors—Acetaminophen Exposure and Gastrointestinal Overgrowth of Clostridia Bacteria Induced By Ingestion of Glyphosate-Contaminated Foods—Dysregulate the Developmental Protein Sonic Hedgehog and Are Major Causes of Autism

Abstract

Epidemiological studies have found 2 significant factors associated with the increased incidence of autism spectrum disorder (ASD): the increased use of acetaminophen in the 1970s when this drug largely replaced the use of aspirin for many patients because of a fear of Reye syndrome, and the agricultural use in the 1990s of the herbicide glyphosate on crops that were genetically modified (GM) to tolerate glyphosate. The incidence of autism in the United States, where acetaminophen is widely available, is more than 1000 times greater than in Cuba, where acetaminophen is available only by prescription. Metabolites of both glyphosate and acetaminophen likely alter the function of the developmental protein sonic hedgehog (SHH). Glyphosate likely affects SHH indirectly by decreasing the beneficial flora of the gastrointestinal tract and increasing pathogenic Clostridia bacteria, which are resistant to glyphosate. The marked increase of certain Clostridia species caused by glyphosate results in Clostridia production of large amounts of 3-(3-hydroxyphenyl)-3-hydroxypropionate (HPHPA) and 4-cresol (p-cresol). The 4-cresol metabolite 4-methyl-o-hydroquinone and the acetaminophen metabolite N-acetyl-p-benzoquinone imine (NAPQI) likely react with the sulfhydryl group of the N-terminal cysteine of SHH, blocking the function of this critical amino acid required for the activation of SHH. HPHPA and 4-cresol also inhibit dopamine β-hydroxylase, resulting in overproduction of dopamine and its toxic metabolites, such as aminochrome, that cause biochemical damage to mitochondria and structural proteins in brain cells. Elevated amounts of these Clostridia products in body fluids in people with autism and in animals with autistic signs have been documented in laboratories throughout the world. The synthesis of the HPHPA molecule in extremely large quantities depletes the body of free coenzyme A, which is needed for the palmitoylation of SHH. SHH covalently coupled to palmitic acid is 30 times more active than SHH without palmitic acid. These possible modifications of SHH help to explain the significantly altered quantities of SHH in the blood serum of patients with autism. The severity of autism is related to the degree of SHH abnormality. The spread of pathogenic Clostridia worldwide from soil to food animals to humans, which may be promoted by glyphosate use, is a great public health concern, not only for autism but perhaps for all the neuropsychiatric diseases that appear to be related to gastrointestinal Clostridia overgrowth These diseases include seizures, tremors, tic disorders, Parkinson disease, chronic fatigue syndrome, obsessive compulsive disorder, schizophrenia, bipolar and unipolar depression, ADHD, and anorexia nervosa.

Introduction

Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder characterized by repetitive and characteristic patterns of behavior and difficulties with social communication. Perhaps the greatest controversy in autism research is how to explain the increased incidence of autism in the last 3 decades. Some researchers think that improved diagnosis is the reason, but others dispute this claim.

For example, a website (last reviewed 2023) intended for the public states the following :

“There is no blood test or brain scan that is used to diagnose autism. Instead, autism is diagnosed based on observations of behavior. Statistics from health organizations suggest that autism spectrum disorder diagnoses are on the rise. This doesn’t necessarily mean that autism itself is becoming more common. Although this might be the case, it could also be that more people are receiving autism diagnoses.”¹

In addition, a recent review of autism in the Journal of the American Medical Association states the following:

“The estimated prevalence of ASD has been increasing in the US, from 1.1% in 2008 to 2.3% in 2018, which is likely associated with changes in diagnostic criteria, improved performance of screening and diagnostic tools, and increased public awareness. No biomarkers specific to the diagnosis of ASD have been identified.”²

On the other side of the debate are environmental scientists and epidemiologists who state the following:

“Diagnosed autism prevalence has risen dramatically in the U.S. over the last several decades and continued to trend upward as of birth year 2005. The increase is mainly real and has occurred mostly since the late 1980s.”³

Similarly, a study from 2009 states the following:

“In summary, the incidence of autism rose 7- to 8-fold in California from the early 1990s through the present. Quantitative analysis of the changes in diagnostic criteria, the inclusion of milder cases, and an earlier age at diagnosis during this period suggests that these factors probably contribute 2.2-, 1.56-, and 1.24-fold increases in autism, respectively, and hence cannot fully explain the magnitude of the rise in autism.”⁴

Curiously, the first two of these quotes^1,2 that promote improved diagnosis as the cause of the increased autism incidence also state there is no laboratory test for autism diagnosis.

Contrast the statements above concerning no markers for autism with an article from a Chinese research group that indicates that, in 62 children with autism, elevated concentrations in the urine of 3 compounds from certain Clostridia bacteria in the intestinal tract, namely 3-(3-hydroxyphenyl)-3-hydroxypropionate (HPHPA), 3-hydroxyphenylacetic acid, and 3-hydroxyhippuric acid, were 98.4% specific for autism.⁵ In addition,an elevated concentration of HPHPA has also been identified as one of the most important markers of autism in studies in Italy,^6,7 Turkey,⁸ Latvia,⁹ and the United States.^10-14 It is extremely important that the public and academic institutions stay up to date on important biochemical studies.

The rest of the article will focus on abnormalities in the amount of sonic hedgehog (SHH) protein in people with autism and its relationship to 2 major factors that I propose as the main causes of autism¹⁵: (1), an alteration in the function of SHH protein due to increased prenatal and postnatal exposure to acetaminophen; and (2), increased exposure to glyphosate-contaminated food leading to overgrowth of gastrointestinal bacteria producing neurotoxic chemicals affecting brain function by depletion of free coenzyme A needed for palmitic acid activation of SHH. The epidemiological evidence for the involvement of acetaminophen and glyphosate with autism will also be reviewed.

SHH protein

The SHH protein is a key signaling molecule that plays a crucial role in the development of various tissues and organs in vertebrates, including humans.^16-20 SHH protein functions as a morphogen, meaning it provides positional information to cells during embryonic development, guiding their differentiation and growth. SHH plays a particularly important role in the development of the nervous system, limbs, and other structures.^16-20 The SHH protein is initially produced as a precursor preprotein molecule, which is then processed and changed to produce the active form (Figure 1).

Figure 1.

Formation of Activated SHH From SHH Preprotein and the Covalent Attachment of Cholesterol and Palmitate to Processed SHH

The SHH preprotein is synthesized as a precursor molecule of 424 amino acids and undergoes various modifications before SHH is released from the cell.¹⁹ First, the N-terminal peptide consisting of amino acids 1 to 23 is removed from the preprotein. Then, a peptide corresponding to amino acids 24 to 197 of the preprotein is split off and becomes the active part of the protein that affects embryonic development after the insertion of cholesterol at the C-terminal end and palmitic acid at the N-terminal end of the newly formed peptide. Cholesterol is covalently bonded to amino acid 197 as part of the splitting off from the signaling domain of the preprotein. One of the critical modifications is the addition of palmitic acid to the N-terminal amino group of the N-terminal cysteine residue of the protein. Palmitoylation is thought to initially occur via thioester linkage to the cysteine residue.¹⁹ The thioester intermediate then rearranges to an amide linkage via an intramolecular S-to-N shift, producing an amide or N-linked palmitate. This modification is catalyzed by the enzyme hedgehog acyltransferase. As shown in Figure 1, the addition of cholesterol is essential for SHH protein stability and secretion, and proper signaling activity from palmitoylation is essential for the correct trafficking and secretion of SHH. The lipid-tethered forms of SHH are approximately 30-fold higher in potency than unmodified soluble SHH, as measured in a cell-based assay, suggesting that the lipid tether plays an important role in SHH function.¹⁹

Palmitic acid, a saturated fatty acid, is one of the major components of cell membranes. It is also used for the post-translational modification of proteins, such as protein palmitoylation. Palmitoylation involves the attachment of palmitic acid to specific cysteine residues in proteins, which can affect their localization, stability, and activity.^19,20 Abnormalities in the availability or metabolism of palmitic acid can impact the palmitoylation of SHH. This impact, in turn, can affect the proper localization and secretion of the protein. Without proper palmitoylation, SHH may fail to reach its target cells or may have reduced signaling activity, leading to developmental defects or diseases associated with disrupted SHH signaling. As shown in Figure 1, the palmitate donor for SHH is palmitoyl coenzyme A, which requires an adequate supply of palmitic acid and free (nonesterified) coenzyme A. I recently published evidence for a deficiency of palmitoyl coenzyme A in people with autism¹⁴; this evidence will be discussed later in this guest editorial. Figure 1 also shows a form of SHH in which the important acetaminophen metabolite N-acetyl-p-benzoquinone imine (NAPQI) has replaced palmitate at the N-terminal cysteine of SHH (amino acid sequence 24-197). The reaction of NAPQI with virtually all proteins, peptides, and cysteine is very common and will also be discussed later in this guest editorial.^21,22 Although NAPQI reacts predominantly at the sulfhydryl functional group of cysteine residues of proteins, palmitate likely first binds to this sulfhydryl group before transfer to the amino group of the N-terminal cysteine, so the attachment of NAPQI likely prevents palmitate attachment to and activation of SHH.²⁰

SHH acts by binding to its receptor, called patched, on target cells, which activates another receptor, called smoothened.¹⁷ This activation triggers a series of intracellular signaling events, culminating in the activation of specific target genes that regulate cell fate, proliferation, and differentiation.

Disruptions in the SHH signaling pathway can lead to developmental abnormalities and various human diseases. Mutations or dysregulation of the SHH gene have been implicated in conditions such as holoprosencephaly (a disorder affecting brain development), certain types of cancer (including basal cell carcinoma and medulloblastoma), and limb malformations.¹⁶

In summary, both cholesterol and palmitate are critical for the proper processing, secretion, and signaling activity of SHH. Without cholesterol and palmitate, SHH may not be able to effectively regulate development and tissue patterning during embryogenesis.

These key roles of SHH are especially important, since SHH serum concentrations are higher in children with autism than in children without autism, with the most severe autism signs associated with the highest SHH serum concentrations (Figure 2).¹⁵ My hypothesis is that the aberrant forms of SHH fail to provide adequate negative feedback to cells producing SHH, leading to overproduction of aberrant nonactive forms of SHH, as shown in Figure 1. The data shown in Figure 2 come from an immunoassay for the SHH protein, which likely measures all forms of SHH to approximately the same degree, since the amino acid sequences of aberrant SHH forms are the same as the activated forms of the protein.

Figure 2.

Increased Serum Concentrations of Sonic Hedgehog Protein in Children with Autism

Anatomical abnormalities in autism, possibly related to SHH malfunction

Since SHH is involved in fetal and post-fetal development, it was suspected that children with autism might have significant fetal abnormalities if their SHH function was impaired. A recent study from Israel is very important in that it is the first to comprehensively examine prenatal organ development in children with ASD via an examination of the fetal anatomy survey.²³ The authors showed that fetuses that developed into children later diagnosed with ASD had significantly higher rates of ultrasonography fetal anomalies compared with both their typically developing siblings and with matched typically developing children from the general population.

The Purkinje neurons secrete SHH to sustain the division of granule neuron precursors in the external granule layer in cerebral development. According to Ritvo et al,²⁴ the number of Purkinje cells in the vermis of the cerebellum was approximately 15 standard deviations below the mean, and approximately 8 standard deviations below the mean bilaterally in the cerebellar hemispheres, in subjects with autism compared with normal controls. Another study found significantly smaller Purkinje cells in the cerebella of children with autism.²⁵

A neuropathological study by Xiao et al²⁶ showed lower Purkinje cell numbers, missed or ectopic neurons of the deeper cerebellar nuclei, cortical-thickness alterations, foliation dysplasia, and migration impairments in the cerebellar cortex of mice with autistic signs. In addition, Stewart Klar found that all children of 49 patients with autism with a lung illness who were referred for persistent cough had a lung defect she called “doublets.”²⁷ This abnormality was not found in 410 children without autism.

The SHH protein is a key signaling molecule involved in embryonic development and tissue patterning in animals. It plays a crucial role in cell differentiation, proliferation, and organogenesis. In addition, brain abnormalities found by neuropathological studies are also consistent with aberrant SHH function in autism. Thus, all of the fetal, lung, and Purkinje cell abnormalities reported in autism are consistent with the abnormal serum SHH concentrations found in autism.

Common factors that may alter SHH function and cause autism

The following is a summary of evidence linking 2 factors as causes of autism: (1) acetaminophen use, and (2) gastrointestinal overgrowth of Clostridia bacteria.

Acetaminophen use

Acetaminophen is also known as paracetamol and N-acetyl-p-aminophenol (AAP or APAP). More than 70% of the population in Western countries has taken acetaminophen at least once, and a relevant percentage takes the drug chronically as a mild pain reliever and antipyretic.²⁸ Acetaminophen is used to treat pain and fever, and it has become one of the most popular over-the-counter (OTC) non-narcotic analgesic agents. For instance, this compound has been taken at least once by more than 85% of children under the age of 91 months in the UK.²⁸ In the US, approximately 79% of the general population regularly takes acetaminophen, including 35% to 60% of pregnant women.²⁸ Acetaminophen use has grown in popularity in large part because of its reputation for safety. However, evidence for the claim that paracetamol is safe was evaluated using a systematic literature search.²⁸ Publications on PubMed between 1974 and 2017 that contained the keywords infant and either paracetamol or acetaminophen were considered in the search. Of those initial 3096 papers, 218 made claims that paracetamol was safe for use on infants or children. Of these 218, 103 papers were identified as sources of authority for the safety claim. A total of 52 papers contained actual experiments designed to test safety; these experiments had a median follow-up time of 48 hours. None monitored neurodevelopment. Furthermore, no trial considered total exposure to the drug since birth, eliminating the possibility that the effects of drug exposure on long-term neurodevelopment could be accurately assessed. Both of these deficiencies should be corrected by future studies.

A total of 73 881 mother-child pairs were included in a meta-analysis of autism and attention-deficit/hyperactivity disorder (ADHD) in 6 European population-based cohorts.²⁹ Results indicated that children prenatally exposed to acetaminophen were 19% more likely to subsequently have borderline ASD or ASD and were 21% more likely to have ADHD compared with nonexposed children.

Another study showed that offspring of mothers who used acetaminophen before pregnancy had a higher risk of low birthweight and being small for gestational age.³⁰ Acetaminophen use less than once per week was associated with a 46% increased risk of being small for gestational age. Acetaminophen use greater than once per week was associated with significantly increased risk of being both small for gestational age and having low birthweight.

A study conducted at Johns Hopkins University of 996 mother-baby pairs in which acetaminophen was used during pregnancy revealed that the cord blood of all babies tested had detectable acetaminophen.³¹ Sensitivity and subgroup analyses found consistent associations between acetaminophen burden and ADHD and between acetaminophen burden and ASD across strata of potential confounders, including maternal indication, substance use, preterm birth, and child age and sex, for which point estimates for the odds ratios varied from 2.3 to 3.5 for ADHD and from 1.6 to 4.1 for ASD. Furthermore, cord biomarkers of fetal exposure to acetaminophen were associated with significantly increased risk of childhood ADHD and ASD in a dose-response fashion.

Another research group used data from the First Baby Study, a prospective cohort study conducted in Pennsylvania, USA, with 2423 mother-child pairs.³² They found 1011 women (41.7%) reported using acetaminophen during pregnancy. Children who were exposed to acetaminophen during pregnancy scored significantly higher on 3 of the 7 Child Behavior Checklist scales: withdrawn, sleep problems, and attention problems.

Using a data set obtained from an internet-based survey among parents on 1515 children from the US, an adjusted odds ratio and gender-specific adjusted odds ratios for number of doses of postnatal acetaminophen provided before age 2 were calculated against the outcome of ASD.³³ Using this data set, the population-attributable fraction associated with postnatal acetaminophen was estimated to be approximately 40% of the risk of ASD among male children in the US. The authors of this study suggest that postnatal acetaminophen use may be a significant contributor to the risk of ASD among males in the US.

A sibling-control study revealed that children exposed to acetaminophen prenatally for more than 28 days had higher activity levels and poorer gross motor development, communication, and externalizing and internalizing behaviors.³⁴ Children exposed prenatally to short-term use of paracetamol (1-27 days) also had poorer gross motor outcomes, but the effects were smaller than with long-term use. Ibuprofen exposure was not associated with neurodevelopmental changes. Schultz et al³⁵ were the first to specifically link increased acetaminophen use to increased autism incidence. Specifically, the authors temporally related increased autism incidence in California with increased acetaminophen use in the United States, and decreased acetaminophen use with a decreased rate of autism in California (Figure 3). The first significant increase in the incidence of autism occurred after a concern in 1980 that aspirin may have been a cause of Reye syndrome; this concern greatly increased acetaminophen use in children. The authors also noted that the rates of autism incidence stopped increasing in the months following 2 attempted extortion events in 1982 and 1986 in which Tylenol, the most popular brand of acetaminophen, was deliberately poisoned with cyanide. The change in the incidence of autism at exactly the time Tylenol use slowed down after news of the contamination and then stopped completely because it was withdrawn by the manufacturer for a significant period is remarkable. This study also found that, compared with controls, children aged 1 to 5 years with autism were 6 times more likely to have taken acetaminophen. There was no increased incidence of autism associated with ibuprofen use.

Figure 3.

Number of People With a Diagnosis of Autism in California by Year of Birth Annotated With Events in Acetaminophen Use. The rates of autism incidence stopped increasing in the months following 2 attempted extortion events in 1982 and 1986 in which Tylenol, the most popular brand of acetaminophen, was deliberately poisoned with cyanide. The first major increase in autism incidence began after acetaminophen use increased after the use of aspirin for children was increasingly stopped. Reprinted with permission.³⁵

One of the difficulties with chemical studies of autism associations is that many chemicals are used worldwide, making it difficult to find a “clean” environment where autism might be less prevalent. Thalidomide is an example of a chemical whose use differed between world regions. One of the clues that led to the discovery of thalidomide as the causative agent of deformed limbs in infants was that it was much more commonly used in Europe than in the United States.³⁶ Babies with deformed limbs occurred in 10 000 babies in Europe, where the drug was widely available, and occurred in only 40 babies in the United States, where women with babies with deformed limbs had mostly purchased thalidomide in Europe. Countries with the greatest use of thalidomide by pregnant women during pregnancy were those with the highest incidence of babies with deformed limbs.

If there were a geographic region in the world in which the incidence of autism was much lower than in the United States, a comparison of medical or dietary differences might provide a significant clue to the major cause of autism. Such a country is Cuba. The highest estimate of the total incidence of autism in 2016 in Cuba is 241 cases out of a total population of 11 000 000 (0.002 19% of the population)³⁷ compared with an estimate of 1 in 36 (2.78%) people in the United States.³⁸ Based on this data, the percentage of the population with autism in the United States is thus 1269 times higher than in Cuba. The per capita income of Cuba is approximately 8 times lower than that of the United States. Despite the economic challenges presented to the government of Cuba, basic health care is readily available, and there are a large number of physicians, trained in 14 medical schools. A possible explanation for the dramatic differences in autism between the United States and Cuba is the dramatic difference in the use of acetaminophen. Acetaminophen is not approved as an OTC product in Cuba but has been available as an OTC product in the United States since 1959.³⁶ Furthermore, in Cuba, prophylactic use of antipyretic drugs is not the standard medical treatment for vaccine-related fever (phone conversation with Spanish-speaking staff with Olympio Rodriquez Santos, MD, MSc, allergist, Camaguey, Cuba). In Cuba, if high fever continues for more than 2 days after vaccination of children, parents are advised to visit the physician’s office, where the drug metamizole is most commonly prescribed. Prescription of acetaminophen in such cases is rare; OTC acetaminophen is only available in special tourist pharmacies. If we can confirm the dramatic difference in autism rates between the United States and Cuba and its relation to the availability of acetaminophen, this finding would be close to an absolute proof of acetaminophen causation of autism. The US government should immediately undertake such studies to confirm the autism incidence in Cuba.

Metabolism of acetaminophen, NAPQI toxicity, and SHH

The metabolism of acetaminophen is shown in Figure 4. There are 4 possible pathways for its detoxification.³⁹ Acetaminophen can be converted to acetaminophen sulfate by phenol sulfotransferase. It can also be converted to a glucuronide or deacetylated to an aminophenol. In addition, it can be converted to its extremely toxic metabolite NAPQI by the P450 2E1 enzyme. In addition, acetaminophen exposure itself induces increased P450 2E1 activity, thus increasing the amount of NAPQI formed with each exposure to the drug.^40,41 The common prophylactic use of acetaminophen for days before vaccination as well as for multiple vaccinations would likely greatly increase the conversion of acetaminophen to its extremely toxic NAPQI metabolite. Glucuronidation is commonly at low capacity in the fetus and in newborns and young infants, such that exposure to acetaminophen at these times leads to greater metabolism by other pathways, such as the one producing NAPQI.⁴² Acetaminophen can also be deacetylated to form p -aminophenol, which can also be sulfated or converted to an active cannabinoid substance by being conjugated with arachidonic acid to form a cannabinoid conjugate termed AM404.⁴³ Para-aminophenol may also be detoxified by phenol sulfotransferase.

Figure 4.

Metabolism of Acetaminophen and Its Relation to the Metabolism of 4-Cresol

4-Cresol, a metabolite made by the types of gastrointestinal bacteria commonly found in the intestines of people with autism, is also metabolized by phenol sulfotransferase. Reduced phenol sulfotransferase activity caused by excessive competitive inhibition by 4-cresol would likely shift more acetaminophen to the production of NAPQI as well as convert 4-cresol to 4-methyl-o-hydroquinone, which is highly reactive and depletes the glutathione that assists in the detoxification of NAPQI, thus increasing NAPQI toxicity.⁴⁴ As shown in Figure 1, NAPQI can form a covalent bond with the terminal cysteine of SHH, potentially preventing the activation of SHH by blocking the addition of palmitic acid. NAPQI may be formed even at therapeutic doses of acetaminophen and may be formed at even higher amounts in the developing fetus and in newborns and young infants who lack glucuronidation capacity.⁴⁰ As in the case of the teratogen thalidomide, the time of exposure to NAPQI during pregnancy may be extremely critical as to the damage that might be done by SHH alteration.

Clostridia bacteria and autism

I reviewed metabolomic studies of autism¹⁴ and summarized the evidence that the presence of HPHPA and a related bacterial metabolite, 4-cresol, from 8 species of Clostridia bacteria (difficile, botulinum, caloritolerans, mangenotii, ghonii, bifermentans, sordellii, and sporogenes) is a major biochemical abnormality found in samples from people with autism in laboratories throughout the world. I proposed these abnormalities to be at least one of the major causes of autism. My reasons are as follows:

1. HPHPA and 4-cresol, widely found in urine samples of children with autism throughout the world, inhibit dopamine β-hydroxylase (DBH) in both the brain and the sympathetic nervous system, leading to excessive dopamine accumulation, which can cause brain damage, and which explains the common use of neuroleptics like the dopamine inhibitor risperidone to treat severe autism. The metabolism of dopamine is outlined in Figure 5. Excess dopamine, caused by the HPHPA and 4-cresol inhibitors of DBH, leaks from the synaptic vesicles at pH 5.2 into the cytosol at pH 7.4. Dopamine at pH 7.4 is very unstable, because the protons of the hydroxyl groups are dissociated and much more rapidly converted to unstable quinones, such as cyclized dopamine o-quinone, dopamine o-semiquinone, and aminochrome, that generate severe oxidative stress.^45,46 Each molecule of unstable cyclized dopamine o-semiquinone generates an excess of oxygen superoxide radicals, causing severe oxidative damage to the neurons. In addition, cysteine and glutathione adducts of these unstable quinones are converted to metabolites that cause apoptosis of neurons. The optimum enzyme activity of DBH corresponds to the low pH of the synaptic vesicles. Thus, the leakage of dopamine into the cytosol further diminishes DBH activity. The dopamine metabolite aminochrome damages neuronal mitochondria by covalently bonding to the mitochondrial complexes I and III of the electron transport chain and to the Krebs cycle enzyme isocitrate dehydrogenase. Aminochrome can also bind to and damage neuronal structural proteins, including α-synuclein, actin, parkin, α-tubulin, and β-tubulin.⁴³ Based on the ability of aminochrome to react to form covalent bonds with proteins, it seems quite likely that aminochrome, a benzoquinone like NAPQI and like the benzoquinone metabolite of 4-cresol, 4-methyl-o-benzoquinone, may also react with the critical N-terminal cysteine group of SHH.

2. Phenylpropionic acid (the precursor of HPHPA) depletes free (nonesterified) coenzyme A required for its metabolism,¹⁴ leading to multiple impairments in intermediary metabolism, depletion of cholesterol, and decreased activation of palmitic acid by coenzyme A (Figure 6). As shown by Knoop⁴⁷ in the early 1900s, phenyl-modified fatty acids such as phenylpropionic acid are broken down by the beta-oxidation pathway. A large steady-state production of phenylpropionic acid from overgrowth of Clostridia bacteria in the gastrointestinal tract may cause depletion of free coenzyme A to such an extent that palmitoyl coenzyme A cannot be maintained in adequate amounts to activate SHH to its active palmitoyl form, resulting in abnormal development. Since free coenzyme A is also required for cholesterol synthesis, the lack of free coenzyme A may also result in inadequate cholesterol for multiple functions such as neuron myelination, steroid hormone production, and SHH activation. Of note, cholesterol deficiency is common in autism.⁴⁸

Figure 5.

Abnormal Dopamine Metabolism in the Presence of Bacterial Inhibitors of DBH

Figure 6.

Depletion of Free CoA by Its Sequestration With PPA and Biochemical Consequences

If Clostridia bacterial products are a major cause of autism, what environmental factor could have radically increased over the past decades to increase the incidence of autism?

Some researchers have proposed that glyphosate may be a cause of autism, based on epidemiological data from 1990 to 2011 that correlate increased use of glyphosate on corn and soy crops with an increased autism rate (Figure 7).^49-52 The graph shows a remarkable increase in autism over this time. Not shown in Figure 7, however, is that glyphosate increasingly was used on non–genetically modified (non-GM) crops as a drying agent at harvest for important grains such as wheat and oats. Sampling by the Canadian Food Inspection Agency found glyphosate in 90% of pizza, 88% of wheat flour, 84% of crackers, 84% of fresh pasta, 83% of cooked pasta, 80% of dried pasta, 75% of oats, 70% of chickpea flour, and 67% of lentils.53 With such a wide range of foods containing glyphosate it is not surprising that a very high number of people have glyphosate in their urine. The Centers for Disease Control and Prevention reported that more than 80% of children and adults had detectable glyphosate in the urine, and a 2017 study from the University of California, San Diego reported that glyphosate use increased by a factor of approximately 15 and that the prevalence of glyphosate in human urine increased approximately 500% since 1994.⁵⁴ In addition, in a review of the effects of multiple chemicals, glyphosate was shown to be the major chemical associated with ASD in children and with ASD signs in rodents^.55

Figure 7.

Correlation Between Number of Children With Autism and Glyphosate Use

Some countries use glyphosate in greater amounts than all other agricultural chemicals combined. Monsanto brought glyphosate to market in 1974 under the trade name Roundup, and Monsanto’s last commercially relevant United States patent expired in 2000.⁵⁶ Many companies now produce the herbicide. By 2016 there was a 100-fold increase from the late 1970s in the frequency of application and volume of glyphosate-based herbicides, with further increases expected in the future, partly in response to the global emergence and spread of glyphosate-resistant weeds.⁵⁶ The shikimic acid pathway in weeds and common non-GM food plants produces aromatic amino acids needed for plant growth⁵⁷; glyphosate inhibits a key enzyme in this pathway, leading to the death of the weed or non-GM plant.

Although glyphosate was designed to kill weeds, it also kills susceptible bacteria that have biochemical pathways like the shikimic acid pathway in plants and weeds. Shehata et al⁵⁸ found that glyphosate exposure in poultry markedly increased pathogenic bacteria, such as Salmonella and Clostridia species, that were resistant to glyphosate in the stool samples of poultry and significantly decreased beneficial flora, such as Enterococcus faecalis, Enterococcus faecium, Bacillus badius, Bifidobacterium adolescentis, and Lactobacillus species, which are susceptible to glyphosate.⁵⁷ In addition, ingestion of the herbicide could be a significant predisposing factor that has been associated with an increase in diseases mediated by Clostridium botulinum in cattle^59,60 and perhaps even in the farmers exposed to infected cattle.⁶¹ Researchers have detected evolution over the past 2 decades of a new disease in cattle and in human caretakers of diseased animals that may be a chronic gastrointestinal form of C botulinum infection.⁶¹

Another research group added different amounts of glyphosate to ventilated incubation vessels filled with rumen fluid of a cow to simulate conditions in the rumen and found that both C botulinum bacteria and botulinum neurotoxin increased in the presence of high glyphosate.⁶² Based on these results, they recommended that the global regulations restrictions for the use of glyphosate should be re-evaluated. C botulinum is one of 3 species of Clostridia bacteria that produces large quantities of the precursors of HPHPA.⁶³ I also found high levels of HPHPA in the urine of a child with autism (William Shaw, PhD, unpublished data, June 2014) who died shortly thereafter with a diagnosis of botulism caused by C botulinum, one of the Clostridia species predicted to be an HPHPA producer based on its production of the precursor phenylpropionic acid.

Puigbò et al⁶⁴ found that 54% of most common gut bacterial species are intrinsically sensitive to glyphosate; that is, these species present amino acid biomarkers that determine the susceptibility to glyphosate. Of these species, 29% are potentially resistant, 7% vary intraspecifically, and 10% are unclassified. Bacteria with sensitive copies of the 5-enolpyruvylshikimate-3-phosphate synthase enzyme that is the target for glyphosate include Faecalibacterium, Bifidobacterium, and Citrobacter, whereas Clostridium, Dorea, and Ruminococcus mostly have resistant sequences. These glyphosate-resistant genera are associated with irritable bowel syndrome.

Del Castilo et al⁶⁵ exposed mice to low amounts of glyphosate in drinking water from pregnancy to adulthood. Exposed mice developed impaired social behavior and increased repetitive behavior. Exposed mice also showed an activation of phagocytic cells (positive for ionized calcium binding adaptor molecule 1) in the cortical brain tissue. The herbicide also led to changes in the gut microbiome composition, which is also crucial for the establishment of the intestinal barrier. Altogether, their findings suggest that long-term glyphosate exposure leads to morphological and functional changes in the gut, which correlate with behavioral changes that are like those observed in patients with neurodevelopmental disorders. Some of the abnormalities were specific for male mice, reflecting the increased incidence of ASD in male humans.

Barnett et al⁶⁶ have stated that environmental exposure to glyphosate and glyphosate-based herbicides has the potential to negatively influence neurodevelopment and behavior across generations indirectly through the gut-brain-microbiome axis. Potential mechanisms by which glyphosate may elicit these effects are through the disruption of the normally symbiotic relationship of the host and the gut microbiome. Glyphosate can kill commensal members of the microbiome like Lactobacillus, Ruminococaeae, and Butyricoccus species, reducing key microbial metabolites that act through the gut-brain-microbiome axis, including indoles, L-glutamate, and short-chain fatty acids.

Another example of environmental exposure to glyphosate that may negatively influence neurodevelopment is the case study of triplets exposed to high amounts of glyphosate due to frequent ingestion of genetically modified corn tortillas¹³; corn and soy are the 2 food crops that are most exposed to glyphosate.⁵³ Two of the children (male) had autism, while a third (female) had a suspected seizure disorder. The 2 male triplets with autism had abnormalities on at least 1 organic acids test, including elevated phenolic compounds such as 4-cresol, 3-(3-hydroxyphenyl)-3-hydroxypropionic acid, and 4-hydroxyphenylacetic acid, which have been previously associated with Clostridia bacteria and autism. All the children had high biochemical markers indicating mitochondrial damage. All 3 children had markedly elevated urinary glyphosate with their mean baseline value—34.4 μg/g creatinine—being 25.5 times higher than the median value of 1.35 µg/g creatinine and 24.1 times higher than the mean value of 1.43 µg/g creatinine of the study’s internal reference range. The glyphosate value in urine in 1 of the retested triplets decreased 94% after switching to an organic food diet. Signs of autism declined considerably after organic diet implementation. Other unpublished cases of autism with high Clostridia markers and high glyphosate in urine samples were also found in my laboratory. In addition, I found extremely high amounts of glyphosate and HPHPA in the urine of an adult with severe psychiatric signs (William Shaw, PhD,unpublished data,July 2017). Although this person ate organic food exclusively, he used chewing tobacco on a regular basis. The likely source of his exposure was glyphosate used as a drying agent on the tobacco plants.

Summary

It appears that the marked increase in ASD may be largely due to 2 factors: (1) the switch from aspirin to acetaminophen use, largely due to a proposed association with Reye syndrome, and (2), a marked increase in the use of genetically modified foods that tolerate the herbicide glyphosate together with the use of glyphosate as a drying agent at harvest on foods that are not genetically modified to be resistant to glyphosate. Both factors may be responsible for altering the function of SHH, which is present at abnormal concentrations in the serum of people with both mild and severe autism. Children exposed prenatally and/or postnatally to both factors may be especially vulnerable to alteration of SHH function. At a minimum, warning labels should be required for acetaminophen and for foods contaminated with glyphosate. These changes will not be easy and will require significant input from all medical and agricultural authorities worldwide. Given the evidence that glyphosate may contaminate agricultural fields with pathogenic Clostridia, the appropriate government agencies should immediately begin an assessment of the microbiomes in glyphosate-treated fields.

The human safety of glyphosate has been seriously questioned by a broad group of 14 health experts who are predominantly professors in institutions of environmental health or medical schools or are other environmental professionals.⁶⁶ A similar consensus statement reflecting the concerns about the safety of acetaminophen is currently supported by 91 scientists, clinicians, and public health professionals from across the globe.⁶⁷ My focus on these 2 factors does not mean that other factors may not also be important as contributing causes of autism, such as heavy metals, nutritional deficiencies, genetic factors, nonmetal toxic chemicals, and other microorganisms such as molds, yeasts, and viruses. However, the 2 factors emphasized in this guest editorial help to explain the marked increase in the rate of autism since the 1980 switch from the use of children’s aspirin to acetaminophen combined with the marked increase in the use of glyphosate from 1992. Furthermore, the common metabolic detoxification pathways for acetaminophen and the bacterial metabolite 4-cresol link the 2 factors, so that the presence of both factors may combine to have an even greater toxicity. Both substances form benzoquinone metabolites that are detoxified by glutathione. In addition, NAPQI reacts covalently with sulfhydryl groups of proteins, as does the benzoquinone metabolite of 4-cresol, making the critical N-terminal cysteine of SHH a likely target of both compounds.

Additional experiments and empirical evidence will be required to determine the correctness of the hypothesis that aberrant SHH forms fail to provide adequate negative feedback, leading to overproduction of relatively inactive SHH forms. Such research would investigate the interactions between aberrant SHH forms, cellular feedback mechanisms, and the resulting effects on SHH production and signaling. Additional research will be required to determine if SHH dysfunction is a major reason for the anatomical abnormalities common in people with autism. Confirmation of the autism incidence in Cuba would also help to confirm the possible relation between autism and acetaminophen exposure.

Clostridia relevance to other neuropsychiatric diseases and danger that glyphosate is spreading pathogenic Clostridia worldwide

Marked increases in the concentration of both HPHPA and 4-cresol with concomitant increases in the dopamine metabolite homovanillic acid and a significant increase in the homovanillic acid: vanillylmandelic acid ratio have been noted in my laboratory for many neuropsychiatric diseases, including seizures, tremors, tic disorders, Parkinson disease, chronic fatigue syndrome, obsessive compulsive disorder, schizophrenia, bipolar and unipolar depression, ADHD, and anorexia nervosa.⁶⁵ Consultations by me with the attending physicians in such cases have indicated substantial reduction in signs by treating the gastrointestinal Clostridia species. In one of the most remarkable cases, the patienthad a complete recovery from first-onset schizophrenia with severe auditory hallucinations with only oral antibiotics to kill Clostridia and no use of standard psychiatric drugs.¹⁰ Cases of neuropsychiatric diseases associated with Clostridia are not rare and so deserve the attention of the entire medical community.

Finally, this article relied heavily on my review of multiple metabolomic studies of autism14 as well as my own metabolomic studies of approximately 500 000 urine samples from people with autism throughout the world over a period of 25 years and shows what a useful tool metabolomics may be for studying other complex diseases. The spread of pathogenic Clostridia worldwide from soil to food animals to humans, as proposed by Rodloff and Krüger,⁶⁰ which may be promoted by glyphosate use, is a great public health concern, not only for autism but perhaps for all the neuropsychiatric diseases that appear to be related to gastrointestinal Clostridia overgrowth.