1 Scientific RepoRts | 7: 13107 | DOI:10.1038/s41598-017-13454-5 www.nature.com/scientificreports Mitochondrial glycerol 3-phosphate facilitates bumblebee pre-fight thermogenesis stewart W. C. Masson 1 , Christopher P. Hedges 1,2 , Jules B. L. Devaux 1 , Crystal S. James 1 & Anthony J. R. Hickey 1 Bumblebees (Bombus terrestris) fy at low ambient temperatures where other insects cannot, and to do so they must pre-warm their fight muscles. While some have proposed mechanisms, none fully explain how pre-fight thermogenesis occurs. Here, we present a novel hypothesis based on the less studied mitochondrial glycerol 3-phosphate dehydrogenase pathway (mGPDH). Using calorimetry, and high resolution respirometry coupled with fuorimetry, we report substrate oxidation by mGPDH in permeabilised fight muscles operates, in vitro, at a high fux, even in the absence of ADP. This may be facilitated by an endogenous, mGPDH-mediated uncoupling of mitochondria. This uncoupling increases ETS activity, which results in increased heat release. Furthermore, passive regulation of this mechanism is achieved via dampened temperature sensitivity of mGPDH relative to other respiratory pathways, and subsequent consumption of its substrate, glycerol 3-phosphate (G3P), at low temperatures. Mitochondrial GPDH may therefore facilitate pre-fight thermogenesis through poor mitochondrial coupling. We calculate this can occur at a sufcient rate to warm fight muscles until shivering commences, and until fight muscle function is adequate for bumblebees to fy in the cold. Bumblebees, such as the burrowing bumblebee (Bombus terrestris), undergo pre-fight thermogenesis to facilitate fight at low temperatures 1–4 . While this is well established, the mechanisms mediating pre-fight thermogenesis are not fully understood. Tere are two main hypotheses, futile cycling and shivering. In futile cycling, phosphof- ructokinase and fructose 1,6-bisphosphatase simultaneously convert fructose 6-phopshate to fructose 1,6- bis- phosphate and vice versa. Tis stimulates heat production via increased ATP hydrolysis, and the resulting ADP increasing ETS activity 5,6 . While this has been an example in biochemistry texts, estimates based on maximal enzymatic capacities suggest that this process likely generates less than 7% of the pre-warming heat required to enable fight 7 . Shivering also releases signifcant heat given that the actin-myosin interaction is thermodynami- cally inefcient 8 , and also increasing ADP concentrations will stimulate heat production from the ETS. However, fight muscle contraction is arrested below 15 °C, and insufcient for shivering below 25 °C 9 . Indeed, shivering is the current explanation for hymenopteran fight muscle heating, and given that there is minimal lactate dehy- drogenase (LDH) expression 10 or phosphogen transfer system 11,12 , all fight muscle ATP is generated through oxidative phosphorylation (OXPHOS). It remains unknown whether bumblebee fight muscle mitochondria can adequately support ATP-dependent heating processes such as shivering, or can respire to release sufcient heat to warm muscles independent of OXPHOS. In bumblebees, mitochondria occupy approximately 40% of fight muscle volume, and account for some of the highest mass specifc respiration rates measured 13 . Moreover, fight muscle mitochondria have high activities of mGPDH, which is also highly expressed in BAT and has been suggested to have a thermogenic function in mammals 14–17 . To sustain glycolytic fux, the glycerol 3-phosphate (G3P) pathway efectively oxidises cytosolic NADH. G3P is produced by cytosolic GPDH through the reduction of dihydroxyacetone phosphate using NADH formed during glycolysis 16 . However, mGPDH then converts NADH (ΔG°′ - 220 kJ mole -1 ) to FADH 2 (ΔG°′ - 181.6 kJ mol -1 ), imparting a 17.5% loss in reducing power. Tis energy loss may be resolved as heat. Moreover, the coupling of FADH 2 to OXPHOS results in ~40% less ATP per oxygen reduced, and equates to approximately 288 kJ mol -1 loss of energy relative to NADH, assuming ΔG′ of ATP = -72 kJ mol -1 18,19 . 1 School of Biological Sciences, University of Auckland, 3a Symonds St, Auckland, 1010, New Zealand. 2 institute of Sport, Exercise and Active Living, Victoria University, Melbourne, VIC, Australia. Correspondence and requests for materials should be addressed to A.J.R.H. (email: a.hickey@auckland.ac.nz) Received: 19 July 2017 Accepted: 22 September 2017 Published: xx xx xxxx OPEN