Pathophysiology of fatty acid oxidation disorders
- First Online:
- Cite this article as:
- Bennett, M.J. J Inherit Metab Dis (2010) 33: 533. doi:10.1007/s10545-010-9170-y
- 702 Views
Mitochondrial fatty acid oxidation represents an important pathway for energy generation during periods of increased energy demand such as fasting, febrile illness and muscular exertion. In liver, the primary end products of the pathway are ketone bodies, which are released into the circulation and provide energy to tissues that are not able to oxidize fatty acids such as brain. Other tissues, such as cardiac and skeletal muscle are capable of direct utilization of the fatty acids as sources of energy. This article provides an overview of the pathogenesis of fatty acid oxidation disorders. It describes the different tissue involvement with the disease processes and correlates disease phenotype with the nature of the genetic defect for the known disorders of the pathway.
fatty acid oxidation
long-chain 3-hydroxyacyl-CoA dehydrogenase
medium- and short-chain 3-hydroxyacyl-CoA dehydrogenase
medium-chain acyl-CoA dehydrogenase
short-chain acyl-CoA dehydrogenase
(mitochondrial) trifunctional protein
very long-chain acyl-CoA dehydrogenase
Normal physiology of the fatty acid oxidation pathway
In order to understand the pathophysiology of fatty acid oxidation (FAO) disorders, it is first important to understand the functional physiology of this important energetic pathway. In normal individuals, FAO functions primarily as an endocrine-mediated response to fasting and also to increased energy demands due to glycogen depletion which may result from either fasting or increased muscular activity. During periods of fasting, falling circulating levels of blood glucose and insulin activate hormone-sensitive lipase in adipose tissues and increase levels of circulating non-esterified fatty acids, which are delivered to peripheral tissues bound to albumin.
In liver, these fatty acids are targeted primarily towards ketogenesis. The ketone bodies are subsequently transported to tissues that cannot adequately generate energy from FAO such as brain. In cardiac and skeletal muscle and in kidney, the acetyl-CoA end product of fatty acid oxidation is targeted for direct oxidation through the tricarboxylic acid cycle. Cardiac muscle is a tissue that can equally utilize lipids or carbohydrates and FAO may be the preferred substrate for energy in heart (Rinaldo et al. 2002). Other tissues including pancreas, placenta, the gastrointestinal tract and the eye may also have capacity for performing FAO although the metabolic significance has still to be determined for some of these tissues (Strauss et al. 2009; Shekhawat et al. 2003, 2007; Oey et al. 2006).
The rate-limiting step for mitochondrial FAO in liver is the activity of carnitine palmitoyltransferase IA (CPT1A) which is situated at the outer mitochondrial membrane (McGarry and Brown 1997). High intracellular levels of malonyl-CoA that are present in the postprandial state normally inhibit this enzyme and prevent transport of fatty acids into the inner mitochondrial compartment. Malonyl-CoA levels fall during fasting and therefore activate CPT1, driving fatty acids towards oxidation. There are tissue-specific isoforms of CPT1. CPT1A is present in liver, kidney and skin fibroblasts and CPT1B in cardiac and skeletal muscle. The kinetics of malonyl-CoA inhibition of CPT1 are such that it is possible to preferentially direct fatty acids to the liver during fasting. This allows for the generation of ketone bodies for tissues such as brain at times when demand for muscular FAO may be low but when glycogen reserves are also low (Snider et al. 2006). This preferential regulation of FAO provides the first indication that there is differential channelling of substrates into the various tissues depending on the individual tissues energetic requirement. CPT1C, a third isoform of CPT1 has been identified in brain but its function remains uncharacterized (Price et al. 2002).
Pathophysiology of FAO defects
The pathophysiology associated with FAO defects results from failure of the pathway to generate energy during periods of increased requirement (typically a combination of fasting, febrile illness, gastrointestinal illness, cold exposure or muscular exertion). Clinical signs and symptoms generally reflect the respective tissue needs at the time of the metabolic stress. Caloric restriction due to fasting, for instance, will invariably impact hepatic function but does not universally impact skeletal or cardiac muscle functions, whereas exercise is more likely to have an effect on muscle FAO and be associated with myopathic signs and symptoms, particularly in patients with the less severe mutations.
Pathophysiology of genetic defects of long-chain FAO
Carnitine palmitoyltransferase IA (CPT1A) deficiency
At this time the only human defect that has been identified for CPT1 is that of the hepatic (and renal) CPT1A deficiency. Reflecting the tissue localization of this isoform of CPT1, the deficiency state is characterized by fasting-induced hepatic failure (Reye-like hepatic encephalopathy) with hypoglycaemia and without ketone body formation (non-ketotic hypoglycaemia). This is a potentially fatal presentation if unrecognized. The hypoglycaemia results from a combination of increased glucose utilization as a result of failure to make ketone bodies and reduced capacity for gluconeogenesis due to lack of the acetyl-CoA product of FAO, which is an essential activator of gluconeogenesis. Acetyl-CoA is also required for the synthesis of N-acetylglutamate, the activator of carbamylphosphate synthetase, the rate-limiting enzyme of hepatic ureagenesis. Therefore, there is also impaired conversion of ammonium to urea and subsequent hyperammonaemia during catabolic periods. CPT1A deficiency has also been associated with renal tubular acidosis during catabolic periods, which is another reflection of the tissue distribution of this enzyme. All of these pathophysiological correlations have been derived from patients with profound deficiency of CPT1A, which is characterized usually by undetectable to less than 5% residual enzyme activity.
The P479L DNA change which was initially described in an individual of Inuit origin (Brown et al. 2001) has been identified in a large number of Alaskan Inuit and Canadian First Nations individuals (Greenberg et al. 2009). These groups have higher than normal rates of infant death. A significant number of newborns identified in the Alaskan newborn screening programme have this DNA variant (Gillingham et al. 2006). To date, it has not been determined whether this is a pathological mutation or an adaptive polymorphism. A direct pathological correlation of this DNA change with infant mortality or morbidity has not been established. It has been hypothesized that this DNA change is adaptive to the ancient Inuit lifestyle as the amino acid substitution removes any inhibition of CPT1A by malonyl-CoA (Brown et al. 2001). Although the enzyme has reduced maximum activity at 20% of normal enzyme, it also has unregulated capacity for FAO. Theoretically, this would allow an individual to slowly and continuously generate energy via ketone body production from fatty acids even when not fasted. One could conceive this as being advantageous during long winter months when food is scarce and provided that there are ample fat stores generated during the summer season (Greenberg et al. 2009). Fasting in the absence of adequate fat stores may possibly be harmful in this scenario, particularly in infants, and this may account for the high infant mortality rate. Cross-sectional and longitudinal studies of individuals with this DNA variant are required in order to answer questions regarding the significance of P479L.
Other long-chain FAO defects
Other membrane-associated enzymes involved with the carnitine shuttle and long-chain FAO are systemically localized and all fatty acid-metabolizing tissues can be affected. Typically neonatal presentation of carnitine-acylcarnitine translocase (CACT), carnitine palmitoyltransferase II (CPT2), very-long-chain acyl-CoA dehydrogenase (VLCAD) and the mitochondrial trifunctional protein (TFP) defects present with early-onset multisystem failure with hepatic, cardiac and skeletal muscle signs and symptoms. Renal tubular disease may also be apparent in some patients. Neurological signs are a secondary result of the failure to deliver substrate (glucose or ketones) to the brain. Neonatal presentations are generally well-correlated with the severity of the mutation and typically have undetectable enzyme activity or protein expression (Strauss et al. 2009).
The later-onset forms of these disorders tend to have a primary cardiac or skeletal muscle presentation. The high residual enzyme activity in these conditions appears to be sufficient to spare the liver and ketogenesis is sufficient to spare the brain from significant involvement. Many of the patients with later-onset disease present with exercise-induced rhabdomyolysis but do not have pronounced fasting intolerance. Cardiac symptoms are rare after childhood. The common high residual activity CPT2 mutation (S113L) is thermolabile so that during vigorous exercise, when muscle temperature rises, the enzyme activity is further reduced by degradation of the protein, exacerbating the muscle damage (Olpin et al. 2003). The release of myoglobin as a result of the rhabdomyolysis may cause glomerular damage and resulting acute renal failure in all long-chain FAO defects that present with rhabdomyolysis. It is the author’s experience that a number of young adults who are suddenly exposed to high-impact exercise such as military training have presented for the first time with muscle signs as a result of this exposure. Despite a lack of systemic signs to suggest a multisystem disease, CPT2 deficiency is a common cause of late-onset muscle disease that needs to be considered in this population.
Intermediate forms of these long-chain FAO defects also exist and many patients are compound heterozygotes for a severe and a mild mutation. These individuals are at risk for metabolic decompensation and hepatic failure when fasting, and may also progressively develop potentially fatal hypertrophic cardiomyopathy. Some patients also develop a skeletal myopathy and are at risk for acute muscle signs and symptoms including rhabdomyolysis under stressful conditions (den Boer et al. 2002; Strauss et al. 2009; Thuillier et al. 2003).
Patients with trifunctional protein defects (LCHAD deficiency and defects that affect all three enzymes including long-chain enoyl-CoA hydratase and long-chain 3-ketoacyl-CoA thiolase) are also at risk for developing additional signs and symptoms including retinal pigmentary degeneration and peripheral neuropathy (Strauss et al. 2009; Tyni et al. 1998).
Heterozygous pregnant female carriers of trifunctional protein mutations who are carrying a homozygously effected fetus are at risk for developing acute fatty liver of pregnancy (AFLP) and haemolysis, elevated liver enzymes, and low platelets (HELLP) syndrome. The pathophysiology of all of these additional complications in trifunctional protein defects has not been characterized but may be a result of toxic accumulation of unusual 3-hydroxy fatty acid intermediates or to some unrecognized requirements for the FAO pathway in retinal, placental and nervous tissues (Ibdah et al. 1999; Wilcken et al. 1993). In a single published case report, LCHAD deficiency was also found to be associated with the rare and serious complication of placental floor infarct with massive perivillous fibrin deposition, thus providing additional indication for a requirement for FAO in placenta (Matern et al. 2001). We have subsequently studied an additional three cases of this severe obstetrical complication and have identified mutations in the LCHAD coding region (Carruth et al. 2009).
Defects of mitochondrial matrix medium- and short-chain length specific enzymes
Medium-chain acyl-CoA dehydrogenase (MCAD) deficiency
Defects of the mitochondrial matrix-associated medium- and short-chain FAO enzymes, although systemically distributed, appear to each have their unique pathophysiological properties. MCAD deficiency is clinically indistinguishable from CPT1A deficiency in that the presentation is of primary hepatic failure with encephalopathy, and a Reye-like illness with relatively hypoketotic hypoglycaemia (Rinaldo et al. 2002). However, one has to be cautioned that during catabolic events, and unlike CPT1A deficiency, there may be positive ketones in the urine from patients with MCAD deficiency when tested with dipsticks. This is presumed to arise as a result of acetyl-CoA being derived from partial oxidation of long-chain acyl-CoAs to those of the medium-chain species. Urine organic acid analysis on these occasions will typically reveal a very high excretion of saturated and unsaturated medium-chain dicarboxylic acids. The successes that we have achieved with expanded newborn screening for MCAD deficiency and the greatly improved clinical outcomes from better management should ensure that future generations of metabolic physicians and laboratorians may only very rarely encounter acute-onset MCAD deficiency (Wilcken et al. 2007). To date, cardiac, renal and skeletal muscle signs and symptoms have not been shown to contribute to the MCAD phenotype, although there have been a few anecdotal reports of both muscle and cardiac signs (Iafolla et al. 1994). It may be that for the vast majority of MCAD deficient patients, partial oxidation of long-chain fatty acids provides sufficient acetyl-CoA for metabolic purposes to support the needs of cardiac and skeletal tissue.
Signs and symptoms associated with SCAD deficiency have been extremely varied and the pathophysiology of the condition is poorly defined (Jethva et al. 2008). The initial description included neurological, myopathic and hepatic signs and symptoms. Neurological signs were attributed to the fact that short-chain fatty acid intermediates are volatile and have greater potential to cross the blood–brain barrier, but this has not been demonstrated despite the availability of an excellent mouse model (Wood et al. 1989). The vast majority of individuals with SCAD deficiency are defined at the molecular level to have one or more common polymorphisms that are present in the general population with a much higher frequency than clinical diagnosis could identify. Furthermore, current experience with diagnosis based upon newborn screening is starting to show that most identified newborns with SCAD deficiency do not develop a clinical phenotype (Jethva and Ficicioglu 2008). Expert opinions vary from describing this enzyme defect as a non-disease to suggesting the SCAD gene is a susceptibility gene for disease in certain, but not all individuals. It has been conclusively demonstrated that the majority of mutant SCAD protein is improperly folded within the mitochondrion and it remains possible that protein–protein interactions are impacted (Pedersen et al. 2008). It is possible that as yet unidentified proteins may interact with SCAD and that these proteins function abnormally to result in the variable phenotype. Conversely, other experts now regard SCAD deficiency as a non disease and find it inappropriate to include in newborn screening programmes. There is concern that patients with SCAD deficiency are being labelled with a metabolic disease unnecessarily and that the establishment of this as a diagnostic end point may result in insufficient investigation of the true causes of any signs and symptoms (Strauss et al. 2009)
Medium- and short-chain 3-hydroxyacyl-CoA dehydrogenase (M/SCHAD) deficiency
Finally, the most recently described FAO defect is M/SCHAD deficiency. The literature mostly names this defect SCHAD deficiency, but the chain-length specificity of the enzyme includes medium-chain-length species in addition to the traditional substrate used for diagnostic purposes which is C4 (acetoacetyl-CoA). Very few cases of M/SCHAD deficiency are known to date, but the majority of confirmed patients have been shown to have a phenotype of symptomatic hypoglycaemia associated with hyperinsulinism (Bennett et al. 2006; Clayton et al. 2001; Hussain et al. 2005; Kapoor et al. 2009; Molven et al. 2004). This is in contrast to all of the other FAO disorders in which the hypoglycaemia is associated with appropriate hypoinsulinism. The patients described to date with hyperinsulin-related hypoglycaemia mostly have undetectable to very low residual enzyme activity and when investigated have little or no detectable M/SCHAD protein by western blot. In a single patient with high residual activity and normal protein by western blot, there was no clear evidence for hyperinsulinism. This patient presented with fasting-induced hepatic disease in a fashion reminiscent of that seen in most patients with classical FAO disorders (Bennett et al. 2006). The pathophysiology of hyperinsulinism in M/SCHAD deficiency is not fully established, but it is becoming increasingly evident that it results from an interaction of the SCHAD protein with glutamate dehydrogenase (GDH) (Filling et al. 2008), which suggests that the hypoglycaemia is protein-sensitive and not necessarily a result of impaired fatty acid oxidation (Kapoor et al. 2009). GDH is known to be an important regulator of insulin secretion in the pancreas and dominantly inherited gain-of-function mutations of GDH result in the hyperinsulinaemia–hyperammonaemia syndrome (Stanley et al. 1998). The current hypothesis is that the M/SCHAD protein interacts with GDH to downregulate GDH activity and reduce the level of insulin secretion. In M/SCHAD deficiency in which there is no protein, this regulation is removed and there is a resultant increase in insulin secretion. In patients in whom there is an enzyme deficiency but protein is expressed, there appears to be no irregularity with regards to insulin secretion. Therefore, this appears to be a non-enzymatic function (a moonlighting role) for the M/SCHAD protein and a novel function for a FAO enzyme. The observation of a protein–protein interaction in this instance is the first indication of multiple roles for FAO proteins and provides a template for future studies of potential non-enzymatic interactions of FAO proteins. This approach may be particularly important for a better understanding of the pathophysiology of SCAD deficiency, for example.