Dynamic metabolic modeling of ¹³C MRS data is a powerful tool to measure compartmentalized metabolic fluxes in the brain. The goal of the present study was to determine whether the use of ¹³C labeled glucose or acetate substrates or combination of substrates leads to improved precision on fitted metabolic rates in the model using Monte-Carlo simulations. The metabolic network used was based on a published two compartment neuronal-glial model [1].

Results

The table shows the relative standard deviation on the determination of the neuronal and glial TCA cycle rates (Vtca(n), Vtca(g)) and glutamate-glutamine cycle (Vnt) using two approaches: (i) infusion of [1,6-¹³C₂]glucose and analysis with a positional model (current approach) or (ii) double infusion of [1,6-¹³C₂]glucose and [1,2-¹³C₂]acetate and analysis with an isotopomer model (new approach). The new approach led to great improved precision on all free parameters in the model. The relative SD with the new approach was comprised between 4% and 12%, compared to 7 to 77% with the classic approach with glucose alone.

Conclusion

This study suggests that ¹³C metabolic modeling using co-infusion of [1,2-¹³C₂]acetate and [1,6-¹³C₂]glucose should provide superior results for the precise determination of metabolic rates using two compartment metabolic modeling in the brain. This double infusion protocol can be expected to advantageously replace more commonly used approaches that utilize ¹³C-glucose alone.

References

[1] Gruetter et al, AJP, 281, E100, 2001

This work was supported by NIH grants R01NS38672, P41RR0807 and P30NS057091

Gln dilution included in the model?	Metabolic flux	Relative SD with  [1,6-13C2]glucose, positional model	Relative SD with [1,6-13C2]glucose + [1,2-13C2]acetate, isotopomer model
NO	Vtca(n)	7%	4%
	Vtca(g)	52%	9%
	Vnt	77%	8%
YES	Vtca(n)	7%	5%
	Vtca(g)	61%	12%
	Vnt	22%	9%