As rates of autism spectrum disorder (ASD) diagnosis continue to rise, the investigation into the potential causes of autism has intensified within the scientific community. As a result of these investigations, autism is currently understood to result from complex interactions between genetic and environmental variables. Now, research indicating that ASD may be caused by propionic acidemia (PA), an uncommon metabolic disorder, may give us even greater etiological insight.
Propionic acidemia (PA) is a rare inborn error of metabolism that prevents patients from properly converting amino acids into sugars during digestion. This results in a toxic byproduct, propionic acid. As propionic acid builds up in the bloodstream, patients experience vomiting, seizures, anorexia, and behavioral symptoms similar to ASD. To prevent this, PA patients are typically treated with specialty diets that avoid the use of the defective enzyme by minimizing protein intake. When these diets are unsuccessful and propionic acid buildup has caused extensive liver damage, liver transplantation may necessary as a last resort.
Due to its neurological impact, patients with PA often present with symptoms of autism that may not be prevalent enough for an ASD diagnosis. However, in some newly reported cases, patients experience both PA and a plethora of ASD symptoms. This comorbidity suggests that there may be a causal relationship between the two, with researchers noting that PA may increase vulnerability to prenatal development of comorbid ASD. However, given the weak understanding of PA’s prevalence in ASD, researchers have only recently taken the first steps to describing this relationship.
While scientists are still learning more about the causes of autism and how PA may play a part, there is reason to believe that they may be onto a useful hypothesis; the gut-brain axis may be implicated in the onset of autism via certain genetic errors of metabolism which cause disruptions of the brain and also the intestinal tract. If scientists can discover how to ease diseases which originate in metabolism, they thus may thus be able to more fully address ASD symptoms as well.
Initial Observations of the Link Between Propionic Acidemia and Autism
The first known foray into PA’s comorbidity in an ASD patient was described in a 2012 paper published in the JIMD Reports journal by Drs M. Al-Owain and N. Kaya of the King Faisal Specialist Hospital and Research Center in Riyadh. This paper examined a young patient from Saudi Arabia who presented with ASD-like symptoms in addition to prenatally diagnosed PA. The potential connection between PA and ASD was very tenuous, but not surprising; a link between ASD and GI tract issues fits cleanly under the mainstream theories regarding autism’s pathogenesis.
The potential relationship between PA and ASD was compelling enough for other researchers to undertake more in-depth investigations. Research produced in 2016 by Drs Peter Witters and Eric Debold corroborates the link between the two conditions in a longitudinal study of 12 patients investigating blood metabolite balance in patients with PA and behavioral disruptions. This investigation was one of the first to document more than a single isolated case of PA and ASD.
Of the 12 patients involved in the study, 8 had features of ASD and these 8 patients were the focus of the analysis published by the authors after the conclusion of the study. Importantly, only 5 of the patients had enough features of ASD to meet the diagnostic criteria for ASD despite all 8 patients exhibiting abnormal serum levels of common physiological molecules associated with ASD. This result upends the conventional understanding of the molecules implicated in causing autism; the 3 non-ASD patients had symptoms of ASD, but not in enough severity or number to fulfill the diagnostic criteria for ASD, which means that their symptoms were caused by another pathology.
Of the common physiological molecules, Witters and Debold examined the patients’ blood concentrations of amino acids, ammonia, and lactate, and the patient’s urine concentrations of propionic acid to guide their analysis. An abundance of soluble amino acids and ammonia would indicate that the patients’ metabolism was insufficiently breaking down proteins and would serve as an experimental proof of PA or another metabolic disorder. Once the patients in the study cohort were established to have PA, the researchers could then draw a quantitative link between the concentration of certain molecules and the incidence or severity of ASD.
Witters and Debold’s study was groundbreaking for its connection of a metabolic disorder to the pathogenesis of ASD and opens up an entirely new subfield of clinical research that has the opportunity to immediately help patients find relief from symptoms. More specifically, if caretakers know that PA and ASD are linked, they may be able to implement smarter diet-based treatment plans which control the symptoms of both at the same time.
Clarifying the Links Between Propionic Acidosis and Autism
One of the fundamental projects of Witters and Debold’s investigation was quantifying and differentiating between ASD and PA symptoms. To facilitate this process, they characterized the symptoms of their patient cohort before proceeding to establish a baseline level of impairment in their study group. Among the patients in the study, all had significant delays in motor development, and all but two had marked intellectual disabilities in keeping with the traditional characterization of PA in isolation. All patients also had delayed or absent speech, and half had central nervous system disorders like poor muscle tone or difficulty maintaining a normal gait, similar to ASD.
However, while there are significant overlaps between PA and ASD symptoms, PA also causes metabolic crises in which toxic byproducts reach dangerously high concentrations in the patient’s body, causing liver damage and seizures. Metabolic crises are extremely dangerous for patients with PA and may be life-threatening when combined with malnutrition. This is because in those with PA, metabolizing proteins require metabolic processes which result in higher concentrations of propionic acid and worsen metabolic crisis. Critically, this metabolic dysfunction and resulting concentrations of propionic acid may have a causal relationship with ASD symptoms.
Propionic Acid Can Induce ASD Symptoms
While propionic acid’s impact on human ASD is just starting to be uncovered, the effects of propionic acid have been investigated much more extensively in animal models of ASD. These experiments in animal models have shown that propionic acidosis can produce ASD symptoms that correlate with high concentrations of the acid. In particular, propionic acid has been found to produce symptoms such as social withdrawal and stereotyped behavior.
In an experiment by the widely cited Drs SR Shultz and DF MacFabe, adult rats were injected with propionic acid in their intracerebroventricular space. After injection, the rats rapidly exhibited numerous symptoms of ASD. Further replications of this experiment by other researchers revealed that the brain lipid concentrations of the rats changed in response to the propionic acid infusion and these lipid concentrations were directly correlated with ASD symptoms. Although the relationship between individual lipids and individual ASD symptoms remains unclear, the results suggest that diets which control PA could also help control ASD—a potentially massive breakthrough.
Potential Links Between PA, ASD, And The Microbiome
Although propionic acid concentrations in the brain are a potentially important mechanism for ASD development, they are not the only way PA may induce ASD symptoms. Disruptions to the microbiome are also a potential cause of ASD symptomatology resulting from PA, as described in the original study reporting ASD in PA by Drs Al-Owain and Kaya. The study by Al-Owain’s group notes that ASD patients typically suffer from dysbiosis of their intestinal mucosa as a result of impaired carbohydrate metabolism. Formally, dysbiosis is used to refer to situations in which a person’s microbiome has pathological effects on the host, similar to parasitism. Dysbiosis itself is caused by increased production of short chain fatty acids (SCFAs) in the gut, which allows for maladaptive bacteria to grow while suppressing normal and healthy gut microbiota. This unhealthy microbiome may affect the gut-brain axis and drive the development of ASD or it may be caused by ASD. Significantly, propionic acid is one of the SCFAs resulting from impaired metabolism.
However, even in the absence of ASD, PA patients suffer from dysbiosis. While intestinal propionic acid concentrations weren’t measured by Al-Owain’s or Witters’ group, the two potential mechanisms for dysbiosis—impaired carbohydrate or protein metabolism—both alter those concentrations markedly. It’s feasible that these dual dysbiosis mechanisms could team up to cause even more severe microbiome disruption and GI tract symptoms. Al-Owain’s group suggests that PA may thus act as one of the drivers of developing ASD under the two-hit model of ASD pathology, which posits that two separate factors must be present before ASD is developed. Under this model, one problem—such as PA in isolation—is insufficient to cause ASD, but when paired with another problem—such as another metabolic disorder affecting the gut-brain axis, like biotinidase deficiency—the combined detrimental effects of the two pathologies can cause ASD. Once developed, ASD likely combines with inborn metabolic disorders to exacerbate negative impacts on the microbiome. This hypothesis is put to the test by Witters’ group’s analysis of the genetic basis for comorbid ASD and PA.
Upending The Genetic Understanding Of PA
While a fertile area of interest, the genetic basis for ASD is not yet fully understood. The genetic basis for PA, on the other hand, was believed to be clear at the time of Debold and Witters’ study. However, Debold and Witters’ research calls into question the established genetic understanding of PA in significant ways, which has profound implications for our understanding of both the etiology of PA and the relationship between PA and ASD.
Debold and Witter study took DNA samples from each of their 12 research subjects and examined their PCCA and PCCB genes. PCCA and PCCB code for two variations of the propionyl-CoA carboxylase enzyme, which has profound implications for gut health. All people have one copy of PCCA, and one copy of PCCB, meaning that they produce two slightly different variations of an enzyme which both serve the same purpose. However, this enzyme is not correctly produced in patients with PA.
Significantly, all of the patients in the study’s cohort who were diagnosed with comorbid ASD were found to have some kind of loss-of-function mutation in their PCCB gene, resulting in PA—the patients still produced the enzyme, but the enzyme was incapable of performing its job, leading to propionic acid buildup. There are many possible loss-of-function mutations, however, and examining which particular loss-of-function mutation of the PCCB gene individual patients possessed raised more questions than it answered. Critically, two of the ASD patients had only mild PA because, the researchers speculate, they shared a less impairing loss-of-function mutation in their PCCB gene, which shouldn’t be possible; all loss of function mutations should be equally and totally impairing to the enzyme.
This result is inconsistent with what is known about the autosomal recessive inheritance of PA without respect to ASD, which predicts that individuals need two identical copies of the PCCB gene which must both contain the same loss-of-function mutation before PA develops. Under this likely incorrect understanding, if two parents are asymptomatic carriers of PA, they have a 75% chance of producing healthy offspring despite 50% of those offspring carrying the loss-of-function mutation. The other 25% of their offspring will carry two copies of the loss-of-function mutation, causing them to develop PA. The existence of mild PA means that some loss-of-function mutations are less incapacitating of the enzyme, and thus it might be possible for certain individuals to have both loss-of-function-mutations—indicating that genetically, they have PA—but only experience mild or subclinical symptoms.
Significantly, Debold and Witter found that the standard inheritance model does not describe the pathological impact of the genetic data that was harvested; when looking at data from siblings of the research subjects, it was discovered that certain people did not develop PA despite having both loss-of-function mutations. In other words, the current model for explaining the genetics of PA is wrong. Additionally, not all known loss-of-function PCCB mutations were correlated with ASD symptoms, meaning that PCCB mutations are not universally predictive of ASD. Of particular importance was the finding that the the brother of one of the patients in the study did not develop PA, ASD, or any ASD symptoms whatsoever despite the presence of loss-of-function mutations; he defied the genetic model. This means that though PA and ASD are genetically correlated, their exact relationship still eludes a comprehensive understanding. However, clinically detectable PA symptoms are still likely to behave as one contributor to developing ASD, supporting the two-hit model of ASD pathogenesis.
The Potential of Future Research
In the concluding sentences of their analysis, Witters and Debold make their strongest argument for the connection between propionic acid and ASD by observing that, given the rates of PA and of ASD occurrence in the general population, the probability of there being no link between the two pathologies is 4.34 in 10 trillion. Put differently, comorbid incidences of PA and ASD are far more common than should be expected between two unrelated diseases, meaning that there is most likely a link between the two. As such, Witters and Debold judge their study as supporting a correlation between ASD and PA, though they state that further research is required to elucidate the relationship in detail.
If subsequent research proves a connection between PA and ASD, ASD pathogenesis will be clearer than ever before. Finding PA to be a causative factor in ASD would be broadly beneficial to ASD research in many other subfields; researchers would be one step closer to having a complete inventory of environment and genetic factors that contribute to the development of ASD, potentially bringing prevention efforts to a higher level. Likewise, if research into the ASD microbiome can solidify the link between certain microbiota and propionic acid concentrations in the gut, patients will be able to calibrate their diets more effectively and potentially exploit the gut-brain axis for symptom relief. Furthermore, ASD patients without PA may be able to reap the benefits of using a PA-control diet if a definitive link is drawn between PA-negative propionic acid levels and ASD symptoms in humans.
Adams, J. B., Johansen, L. J., Powell, L. D., Quig, D., & Rubin, R. A. 2011. Gastrointestinal flora and gastrointestinal status in children with autism — comparisons to typical children and correlation with autism severity. BMC Gastroenterology, 11(1). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3072352/
Al-Owain, M., Kaya, N., Al-Shamrani, H., Al-Bakheet, A., Qari, A., et al. 2012. Autism spectrum disorder in a child with propionic acidemia. JIMD Reports, 63-66. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3573175/pdf/978-3-642-32442-0_Chapter_143.pdf
Macfabe, D., Cain, D., Rodriguezcapote, K., Franklin, A., Hoffman, J., et al. 2007. Neurobiological effects of intraventricular propionic acid in rats: Possible role of short chain fatty acids on the pathogenesis and characteristics of autism spectrum disorders. Behavioural Brain Research, 176(1):149-169. https://www.sciencedirect.com/science/article/pii/S0166432806004165?via%3Dihub
Sutton, V., Chapman, K., Gropman, A., Macleod, E., Stagni, K., et al. 2012. Chronic management and health supervision of individuals with propionic acidemia. Molecular Genetics and Metabolism, 105(1):26-33. https://www.ncbi.nlm.nih.gov/pubmed/21963082
Williams, B. L., Hornig, M., Buie, T., Bauman, M. L., Paik, M. C., et al. 2011. Impaired carbohydrate digestion and transport and mucosal dysbiosis in the intestines of children with autism and gastrointestinal disturbances. PLoS ONE, 6(9). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3174969/
Witters, P., Debbold, E., Crivelly, K., Kerckhove, K. V., Corthouts, K., et al. 2016. Autism in patients with propionic acidemia. Molecular Genetics and Metabolism, 119(4), 317-321. http://www.mgmjournal.com/article/S1096-7192(16)30177-9/fulltext