In the last two decades, there is growing evidence that NO deficiency develops in patients with mitochondrial diseases and contributes to the complex pathophysiology of these disorders. In patients with MELAS, Koga et al. (2005) showed that NO metabolite (nitrate and nitrites) concentrations were lower compared to controls during the acute phase. Stable isotope infusion techniques used to evaluate NO synthesis in adult and pediatric subjects with MELAS (El-Hattab et al., 2016, 2012) showed lower NO production in affected subjects compared to controls, providing additional evidence of NO deficiency in mitochondrial disorders.

The mechanism of NO deficiency in patients with mitochondrial disorders is complex and results from several factors that act at different levels of NO synthesis. As mentioned earlier, both arginine and citrulline are substrates for NO synthesis, therefore their deficiencies can negatively impact NO synthesis. Koga et al. (2005) showed that plasma concentrations of arginine and citrulline were lower in patients with MELAS compared to control subjects and this observation was seen in both acute and interictal phases (p = 0.01). Moreover, arginine levels in the acute phase were significantly lower than in the interictal phase (Koga et al., 2005). Arginine clearance was found to be higher in subjects with MELAS syndrome and this might contribute to the observed low levels (El-Hattab et al., 2012). As arginine is synthesized de novo from citrulline, low citrulline levels can also explain the low arginine levels. In fact, de novo arginine synthesis was shown to be lower in subjects with MELAS syndrome (El-Hattab et al., 2012). Hypocitrullinemia was documented in patients with MELAS and other mitochondrial disorders (Parfait et al., 1999; Naini et al., 2005). Citrulline is synthesized de novo by mitochondrial enzymes and therefore mitochondrial dysfunction was speculated to be the underlying reason for hypocitrullinemia (Parfait et al., 1999; Naini et al., 2005). In a more recent study, Al Jasmi et al. (2020) showed that patients with mitochondrial disorders other than MELAS had low plasma arginine and citrulline levels compared to controls.

Besides deficiency of NO precursors, reduced activity of NOS could also contribute to low NO levels in patients with mitochondrial disorders. In hybrid cell-lines, Desquiret-Dumas et al. (2012) showed that the m.3243A > G mutation in MT-TL1 gene, the most common mutation in MELAS, resulted in a significant decrease in the nitrite-nitrate concentration indicating reduced NOS activity. Through quantification of NADPH diaphorase histochemistry, Rodrigues et al. (2016) performed a quantitative analysis of NOS activity in muscle fibers from patients with variable mitochondrial diseases. They showed that muscle fibers with reduced cytochrome-c-oxidase (COX) activity have low sarcoplasmic NOS activity (Rodrigues et al., 2016). In addition, a high NADH/NAD⁺ ratio that results from defective mitochondrial respiration also inhibits NOS (Koga et al., 2012).

Mitochondrial proliferation is a compensatory mechanism that develops in mitochondrial disorders and results in increased COX activity. As NO has a strong affinity for COX (Vos et al., 2001), this increase in COX activity results in more NO binding, NO sequestration, and decreased local availability (Naini et al., 2005). Interestingly, while the ragged-red fibers (RRFs) in mtDNA disorders that are associated with mitochondrial proliferation are COX negative; in MELAS, the RRFs are typically COX positive and this is called “MELAS paradox” (Naini et al., 2005). The elevated amount of COX in MELAS blood vessels in turn results in NO sequestration.

In mitochondrial disorders, oxidative stress results in elevated levels of asymmetric dimethylarginine (ADMA) as well as overproduction of reactive oxygen species (ROS); both of which impair NOS activity (El-Hattab et al., 2012). Finally, reduced NOS activity can result from down regulation of NOS to compensate for impaired cellular respiration in mitochondrial disorders (Tengan et al., 2007).

MELAS is one of the most common mitochondrial disorders caused by mutations in mtDNA (Sproule and Kaufmann, 2008). Most data on NO deficiency in mitochondrial disorders are derived from studies on this particular disease. One of the hallmarks of MELAS is the development of stroke-like episodes, in which the ischemic regions do not correspond to the typical vascular territories seen in classic thrombotic or embolic strokes (Koenig et al., 2016). Stroke-like episodes are an important cause of morbidity and mortality in MELAS subjects. In MELAS, NO deficiency results in vasoconstriction leading to ischemia and hypoxemia (Koga et al., 2012). Flow-mediated vasodilation (FMD) was found to be lower in subjects with MELAS compared to controls (Koga et al., 2006). Neuropathological findings in patients with MELAS include ischemic-like lesions characterized by neuronal loss, capillary proliferation, and peri-lesional astrogliosis. These findings suggest that vascular dysfunction is an important factor in the complex pathophysiology of MELAS (Betts et al., 2006). Other disease manifestations seen in MELAS and other mitochondrial disorders are fatigue and exercise intolerance. It is estimated that 20% of patients with mitochondrial disorders experience such symptoms (Mancuso et al., 2012). NO enhances exercise performance and skeletal muscle function through various signaling pathways (Dyakova et al., 2015). In fact, NO triggers mitochondrial biogenesis through the activation of different signaling pathways, including the peroxisome proliferator-activated receptor-γ coactivator 1α (PGC-1α) axis (Lira et al., 2010; Bishop et al., 2018). In mice, ablation of eNOS resulted in impaired exercise performance and increased plasma lactate levels (Lee-Young et al., 2010). Therefore tackling NO deficiency could help in ameliorating fatigue and exercise intolerance.

Koga et al. (2002) reported three patients with MELAS (age 15–18 years) who received either arginine (0.5 g/kg) or placebo within 1 h of the onset of stroke-like symptoms. Using ^99mTc-ECD single-photon emission computed tomography (SPECT), the authors showed that arginine administration resulted in improved microcirculation. The symptoms associated with stroke-like episodes also improved with arginine treatment (Koga et al., 2002). Following this study, a larger trial was conducted on 24 patients with MELAS with a total of 34 stroke-like episodes (Koga et al., 2005). As compared to placebo, arginine administration was associated with significant clinical and biochemical improvement following stroke-like episodes. Acute administration of arginine during stroke-like episodes resulted in the resolution of the transient changes observed in brain magnetic resonance imaging (MRI) within 1 week (Kitamura et al., 2016). Koga et al. (2018) also evaluated the efficacy of intravenous arginine administered within 6 h of stroke-like episodes in 10 patients with MELAS. The primary endpoint was the improvement in rates of headache and nausea/vomiting at 2 h after completion of the initial intravenous administration. Although there was improvement in symptoms, this outcome measure did not reach the study target of 30% improvement in 2 h. The authors speculated that extending the window for intravenous arginine administration from 3 h in their previous study to 6 h in this study may partly explain these findings (Koga et al., 2018; Ikawa et al., 2020; Table 1).

In a consensus statement from the mitochondrial medicine society (MMS) regarding patient care standards for primary mitochondrial disease that was based on Delphi-consensus, it was stated that “IV arginine hydrochloride should be administered urgently in the acute setting of a stroke-like episode associated with the MELAS m.3243 A > G mutation in the MT-TL1 gene and considered in a stroke-like episode associated with other primary mitochondrial cytopathies as other etiologies are being excluded. Patients should be reassessed after 3 days of continuous IV therapy” (Parikh et al., 2017). In a review regarding the recommendations for the use of arginine in the management of acute stroke-like episodes in patients with MELAS, it was recommended that a bolus of 0.5 g/kg continues infusion over 24 h of arginine should be given within 3 h of symptoms onset and continued for 3–5 days (Koenig et al., 2016). A recent, Delphi method based consensus statements for the management of mitochondrial stroke-like episodes from Europe stated that the use arginine is controversial as there is little evidence to support it (Ng et al., 2019). Based on the published studies we reviewed, the MMS consensus statement (Parikh et al., 2017), opinions of experts in this field (Koenig et al., 2016), and the authors personal experience, there are clear benefits for the use of arginine in MELAS syndrome. Therefore, it is recommended to treat patients with MELAS with intravenous arginine during stroke-like episodes.

Six patients with MELAS received oral arginine (0.15–0.3 g/kg/day) for 18 months. This resulted in significant improvement in the frequency and severity of stroke-like episodes after treatment (Koga et al., 2005). In a multicenter, prospective, clinical trial, oral administration of arginine at a dose of 0.3–0.5 g/kg/day in three divided doses for the duration of 2 years in 13 patients with MELAS was not associated with a statistically significant difference in the study outcomes including MELAS stroke scale, mitochondrial disease severity scores, or migraine severity scores (Koga et al., 2018). Nevertheless, there was a tendency for arginine to improve symptoms as there was a trend toward improvement (p = 0.055) on the MELAS stroke scale. In addition, the interictal phases were extended after arginine treatment. The maximum plasma arginine concentration was 167 μmol/L when seizures developed and the authors suggested that maintaining plasma arginine levels ≥ 168 μmol/L may therefore prevent seizures (Koga et al., 2018). Arginine administration was also associated with an improvement in the endothelial function in patients with MELAS as measured by FMD (Koga et al., 2006).

Oral arginine administration to 10 adults with MELAS at a dose of 10 g/m² body surface area divided every 4 h for 48 h resulted in an increased NO synthesis rate as measured by arginine-to-citrulline flux which represents NO synthesis rate (El-Hattab et al., 2012). The increase in NO synthesis was associated with a concomitant increase in de novo arginine synthesis and plasma arginine concentration. Citrulline supplementation at the same dose resulted in higher de novo arginine synthesis and greater increase in NO synthesis, suggesting that citrulline is a better precursor for NO as compared to arginine (El-Hattab et al., 2012). Similar results were also observed when the same study was performed on five children with MELAS (El-Hattab et al., 2016). Based on these findings, an open-label dose-finding and safety clinical trial will soon start to establish the maximum tolerated dose of citrulline in individuals with MELAS syndrome by measuring the incidence of dose-limiting toxicities. The study will also evaluate changes in cerebral blood flow and cerebrovascular reactivity by using arterial spin-labeling (ASL) magnetic resonance imaging (MRI) as a secondary outcome measure.¹

Three siblings with MELAS were studied with ASL MRI to evaluate regional cerebral blood flow (CBF) and arterial cerebrovascular reactivity (CVR). CVR was measured using changes in blood oxygen level dependent (BOLD) signal in combination with changes in end tidal PCO2. CVR reflects the capacity of blood vessels to dilate in response to vasodilatory stimulus and it is therefore a marker for brain vascular reserve. Compared to controls, subjects with MELAS were found to have decreased CVR (p ≤ 0.002) and increased CBF (p < 0.0026) and this difference correlated with disease severity and percentage of mutant mtDNA (Rodan et al., 2015). CVR was inversely proportional to the increase in cerebral blood flow. The authors speculated that increased CBF may result from adaptive response to compensate for energy deficiency or it may represent a passive response to tissue acidosis. Endothelial dysfunction that results from abnormal mitochondria may also cause functional impairment of vasodilation in response to stimuli, thereby reducing CVR and increasing blood flow (Rodan et al., 2015). Following this observation, the authors designed a pilot study to assess the response of CBF and CVR to arginine supplementation in three siblings with MELAS (aged 17, 21 and 22 years) compared to four healthy controls. Visual cortex response in MELAS and controls were evaluated using functional MRI with a stimulus consisting of an alternating black and white checkerboard (Rodan et al., 2020). Subjects with MELAS received a single oral dose of arginine at a dose of 100 mg/kg and a CVR study was performed 1 h later. Two weeks later, subjects with MELAS were started on a 6-week course of oral arginine at 100 mg/kg divided three times daily and CVR study was repeated after the course. CVR in MELAS subjects after a single dose and a 6-week arginine course was not significantly different than baseline, but there was a trend toward increasing CVR in the frontal cortex and a decrease in the occipital cortex. Following 6 weeks of arginine, one patient had a 29% reduction in baseline CBF in the frontal cortex and a 37% reduction in CBF in the occipital cortex. In this patient, pre-treatment functional MRI activation in response to visual cortex stimulus was significantly decreased compared to controls and showed marked increase toward control values following a single dose and a 6-week arginine course (Rodan et al., 2020).

In a consensus statement from the MMS regarding patient care standards for primary mitochondrial disease, it was stated that “The use of daily oral arginine supplementation to prevent strokes should be considered in MELAS syndrome” (Parikh et al., 2017). Following the first stroke in patients with MELAS, it is recommended to start oral arginine at a daily dose of 0.15–0.30 g/kg in three divided doses (Koenig et al., 2016).

While initial studies focused on the use of arginine and citrulline in MELAS, mitochondrial disorders share common pathophysiological mechanisms and long term complications. In fact, stroke-like episodes could develop in mitochondrial disorders other than MELAS (Barca et al., 2020). In a retrospective study that evaluated the use of intravenous arginine therapy for acute metabolic strokes in pediatric patients with mitochondrial diseases other than MELAS, nine out of 71 subjects received acute arginine treatment for one or more stroke-like episodes (total 17 episodes for the nine subjects) during the study duration of 8 years (Ganetzky and Falk, 2018). These nine patients include four with mtDNA pathogenic point mutations, one with mtDNA deletion, and the remaining four had nuclear gene related disorders (FBXL4, POLG, NDUFS8, and SURF1). A positive clinical response to intravenous arginine occurred in 47% of stroke-like episodes. Additional 6% of episodes showed clinical benefit from multiple simultaneous treatments that included arginine. The presence of unilateral symptoms strongly predicted arginine response. There was also a trend toward increased arginine responsiveness in patients with mtDNA mutations compared to those with nuclear genes mutations (P = 0.1) and in older pediatric subjects compared to younger subject (P = 0.24).

In a recent study, nine individuals with variable mitochondrial diseases, less than 18 years of age, were randomized to receive either arginine or citrulline (500 mg/kg/day if weight is < 20 kg and 10 g/m²/day if weight is ≥ 20 kg; divided in three doses) for 2 weeks. The subjects then crossed over to the other intervention following a 2 week washout period (Al Jasmi et al., 2020). Peripheral arterial tonometry was used in this study to measure the reactive hyperemic index (RHI) as a marker for endothelial dysfunction. RHI averages increased after arginine or citrulline supplementation. Average RHI increased 15 and 19% with arginine and citrulline supplementation, respectively (Al Jasmi et al., 2020).