A neuronal MCT2 knockdown in the rat somatosensory cortex reduces both the NMR lactate signal and the BOLD response during whisker stimulation.

Leslie Mazuel,Véronique Bouchaud,Cendrine Repond,Jordy Blanc,Luc Pellerin,Gérard Raffard,Nicole Déglon,Anne-Karine Bouzier-Sore,Gilles Bonvento,Alexander A Mongin

doi:10.1371/journal.pone.0174990

Abstract

Although several in vitro and ex vivo evidence support the existence of lactate exchange between astrocytes and neurons, a direct demonstration in vivo is still lacking. In the present study, a lentiviral vector carrying a short hairpin RNA (shRNA) was used to downregulate the expression of the monocarboxylate transporter type 2 (MCT2) in neurons of the rat somatosensory cortex (called S1BF) by ~ 25%. After one hour of whisker stimulation, HRMAS 1H-NMR spectroscopy analysis of S1BF perchloric acid extracts showed that while an increase in lactate content is observed in both uninjected and shRNA-control injected extracts, such an effect was abrogated in shMCT2 injected rats. A 13C-incorporation analysis following [1-13C]glucose infusion during the stimulation confirmed that the elevated lactate observed during activation originates from newly synthesized [3-13C]lactate, with blood-derived [1-13C]glucose being the precursor. Moreover, the analysis of the 13C-labeling of glutamate in position C3 and C4 indicates that upon activation, there is an increase in TCA cycle velocity for control rats while a decrease is observed for MCT2 knockdown animals. Using in vivo localized 1H-NMR spectroscopy, an increase in lactate levels is observed in the S1BF area upon whisker stimulation for shRNA-control injected rats but not for MCT2 knockdown animals. Finally, while a robust BOLD fMRI response was evidenced in control rats, it was absent in MCT2 knockdown rats. These data not only demonstrate that glucose-derived lactate is locally produced following neuronal activation but also suggest that its use by neurons via MCT2 is probably essential to maintain synaptic activity within the barrel cortex.

Highlights

As early as the end of the 19th century, a link between cerebral activity and blood flow modifications has been established [1,2,3]
No apparent difference in monocarboxylate transporter type 2 (MCT2) immunofluorescence could be observed between non-infected neurons from different animals or between non-infected neurons and neurons infected with the shUNIV vector
Neurons of the somatosensory cortex infected with the shMCT2 vector exhibited a significant reduction (-26%) of MCT2 expression compared to neurons infected with the shUNIV vector (Fig 1D)

Summary

Introduction

As early as the end of the 19th century, a link between cerebral activity and blood flow modifications has been established [1,2,3]. With the emergence of neuroimaging techniques such as PET or fMRI, an uncoupling phenomenon between glucose and oxygen consumption during brain activation was revealed, suggesting a non-oxidative glucose consumption during neuronal activity [4,5]. The ANLS proposed a metabolic cooperation between astrocytes and neurons; astrocytes providing glycolytic lactate as an energetic substrate to the more oxidative neurons during brain activation, the release of the neurotransmitter glutamate being the signal for this coupling. Astrocytes are ideally located between blood vessels, which bring oxygen and glucose to the brain parenchyma, and neurons. Brain microvessels are almost entirely covered by astrocyte endfeet in the rat hippocampus [7], suggesting it might constitute a prominent blood-borne glucose uptake site

Objectives

Methods

Results

Conclusion