The observed production of hyperpolarized (HP) [¹³C]lactate from [¹³C]pyruvate substrate has been shown to reflect largely exchange of the ¹³C label with the endogenous lactate pool, as opposed to net metabolic flux, both in cells¹ and in vivo². α-ketobutyrate (αKB) is a good alternative substrate to pyruvate for LDH (especially for LDHB expressed subunits, but also active with LDHA), and ¹³C hyperpolarization of αKB has recently been demonstrated³. HP αKB yields a similar spectrum as HP pyruvate after injection, but the observed production of corresponding reduction product (α-hydroxybutyrate or αHB) by definition represents net metabolic flux as opposed to exchange, since the endogenous pool of αHB is very small (low μm range)⁴. The purpose of this study was to investigate the effect of manipulating the pool sizes of LDH-catalyzed reduction products (αHB and lactate) on experiments with HP αKB, via co-injections of unlabeled material, to clarify the nature of the metabolic conversion observed in experiments using HP αKB.We performed 3D MRSI of HP [1-¹³C]αKB with and without simultaneous co-injection of unlabeled αHB or lactate, in a group of five rats (for a total of 15 HP experiments). For hyperpolarization, the neat acid form of [1-¹³C]αKB (Sigma Isotec) was liquified by addition of a small amount of H₂O (12% by weight) and mixed with 15mM trityl radical OX063 for polarization via DNP. Polarization of approximately 19% (time of dissolution) was attained, with measured aqueous T₁= 73s at 3T. For data acquisition, MRSI data (1cm isotropic 3D EPSI) was acquired in a clinical 3T scanner at 30s after the start of a 12s injection of 2.5mL 80mM HP [1-¹³C]αKB, with or without simultaneous co-injection of unlabeled 80mM αHB or lactate. The chronological order of injections in each imaging session was alternated in an attempt to exclude effects from prior injections, and each injection was separated by approximately one hour. The mean αHB/αKB ratio averaged over the rat left kidney was measured using SIVIC⁵ and Osirix. Notably, substantial HP bicarbonate signal (e.g. Fig. 1, folding into the EPSI window) was also observed in both the kidneys and the liver (more than typically observed with HP pyruvate), presumably due to decarboxylation via PDH or similar enzyme. αKB is converted to propionyl-CoA following decarboxylation, feeding the TCA cycle via succinyl-CoA. Transamination product was not detected in the localized spectra.The appearance of HP αHB product after injection of HP αKB only (Fig. 1B) is definitively due to net metabolic flux via LDH, since the preexisting endogenous pool of αHB is very small. Moreover, the HP αHB-to-αKB ratio was not much affected (change= -4.9%, p= 0.80) by co-injection of unlabeled αHB (e.g. Fig. 1C, results summarized in Fig. 2), indicating that the conversion of αKB to αHB does not reflect label exchange even in the presence of a large reduction product pool (i.e. not pool size limited), and further suggesting that, unlike lactate, back-conversion of αHB to αKB is negligible. The αHB-to-αKB ratio on the other hand increased by 74% (p= 0.057) with co-infusion of unlabeled lactate (e.g. Fig. 1D, Fig. 2). The added lactate thus appears to exert an indirect pool size effect on the production of αKB which is coupled via the LDH system, as more of the supraphysiologic HP αKB (which is much more highly concentrated than either of the unlabeled pools of oxidized substrate, pyruvate or αKB) is converted to HP αHB to balance the rapid conversion of the infused unlabeled lactate to pyruvate as their pools equilibrate. However, according to the absence of an effect from co-injecting unlabeled αHB, no αHB appears to be converted back to αKB, and the conversion of αKB to αHB still solely reflects net metabolic flux. In this analysis, we ignore any possible differences in transport between molecular analogs since prior studies have indicated similar transport characteristics⁶. Another complication is the preferential activity of αKB with LDHB, although αKB also has significant activity with LDHA.