Neurotoxicity of cassava cyanogens in rodents and non-human primates

Abstract

Background: Cassava is staple food to over half a billion people globally. Consumption of insufficiently processed cassava and dietary sulfur amino acid (SAA) deficiency has been implicated in the pathogenesis of Konzo, a paralytic condition prevalent in SubSaharan countries. Cyanogenic cassava have also been associated with pervasive cognitive deficits in humans, an overlooked phenomena raising global health concerns. Probable candidate for such neurodegeneration and/or disability include cyanide, or cyanate, sulfur deficiency, or their respective combinations. The susceptibility factors and mechanisms underlying the toxicity of cyanogenic cassava have remained poorly understood partly due the lack of an experimental model. Objectives: To investigate the neurotoxicity of cassava cyanogens in rodents and non human primates. Methods: Young 6-8 weeks old male rats were treated intraperitoneally with either 2.5 mg/kg body weight (bw) NaCN, or 50 mg/kg bw NaOCN, or 1 µl/g bw saline, daily for up to 6 weeks and assessed for cognitive performance. Short and long-term memories were evaluated using a radial arm maze (RAM) testing paradigm. Additionally, young adult male rats (Crl: NIH-Fox1 rnu/Fox 1+, 6-8 weeks old) were fed either a diet rich in all amino acids (AAA) or 75%-deficient in SAA and treated intraperitoneally with either 2.5 mg/kg/body weight (bw) NaCN, or 50 mg/kg/bw NaOCN, or 1µl/g/bw saline, for up to 6 weeks and studied for cyanide detoxification capabilities (CDC) and protein carbamoylation, respectively. Further, male Macaca fascicularis monkeys (~12 year-old) (N=12) were exclusively fed on cassava for 5 weeks. CDC was assessed in plasma, spinal cord, and brain of rodents and in plasma of monkeys. Carbamoylation of albumin and spinal cord proteins was analyzed by liquid chromatography mass spectrometry (LC-MS/MS). Results: Behaviourally, NaOCN impaired short-term working memory with fewer correct arm entries (CAE) (F 2, 19 = 4.57 p <0.05), higher working memory errors (WME) (F 2, 19 = 5.09, p <0.05) and longer RAM navigation time (NT) (F2, 19 = 3.91, p <0.05) for NaOCN compared to NaCN and saline treatments. Long-term working mem induced acute seizures whereas NaOCN induced limb paralysis under SAA-restricted diet. No deficits were observed in non-human primates. Under normal diet, the CDC were up to ~ 80X faster in the nervous system (14 milliseconds to produce one µmol of thiocyanate from the detoxification of cyanide) relative to plasma. The spinal cord CDC was impaired by NaCN, NaOCN, or SAA-deficient diet. In non-human primates, the plasma CDC changed proportionally to total proteins (r=0.43; p<0.001). The plasma CDC was ~ twice faster compared to the rodent capabilities. Metabolically, there was a time-dependent decrease in BUN/creatinine ratio under the cassava diet (p<0.001). Additionally, high levels of carbamoylation relative to NaCN and vehicle (P<0.001). At Day 14, we found a diet-treatment interaction effect on albumin carbamoylation (p=0.07) at day 14 while there was no effect attributed to diet (p=0.71) at day 28. The mean number of NaCN-associated carbamoylated sites on albumin became 47.4% significantly higher relative to vehicle (95% CI:16.7-86.4%) at day 28. Spinal cord proteins were only carbamoylated by NaOCN prominently under the SAA-restricted diet. Differentially carbamoylated proteins included myelin basic protein, myelin proteolipid protein, neurofilament light polypeptide, glial fibrillary acidic protein, and 2', 3' cyclicnucleotide 3'-phosphodiesterase. Conclusion: The findings provide an experimental evidence for the biological plausibility that cassava cyanogens may induce cognition deficits. The deficits may result from dual toxicity effect of cyanide and cyanate reflecting their differential toxicity machanisms. The nervous system susceptibility to food (cassava) cyanogenesis and neurotoxic insults seen in konzo subjects may result from a “multiple hit” process including cyanide, cyanate toxicity, deficiency in sulfane sulfur, and cyanate-induced carbamoylation. The multiple hit processes may combine direct mitochondrial insults, protein carbamoylation and a thiol-redox derangement. This level of pathogenetic complexity should be considered in biomarker studies and efforts to prevent neurotoxicity effects of cassava.