*2.1. Dicarboxylate Carriers (DICs)*

DICs are members of the MCF reported to facilitate the transport of dicarboxylates such as malate and related compounds as well as phosphate, sulfate, and thiosulfate across the IMM [44]. Phylogenetic analysis showed that distinct DIC kingdom-level subclades can be distinguished clearly, separating those of fungal, animal, or plant origins (Figure 1). All higher land plants in our analysis possess at least one copy of DIC. However, there is no DIC homolog in algae and it seems to have first evolved in bryophytes (*Selaginella moellendor*ffi*i* and *Physcomitrella patens*), which have three and five DIC homologs, respectively. This is consistent with the analysis of MCFs in *Ostreococcus lucimarinus* where a DIC homolog could not be identified [11]. DIC gene duplication seemed to be more common in plants than in animals. Consistent with our analysis, *A. thaliana* was reported to have three DIC homologs—AtDIC1 (AT2G22500), AtDIC2 (AT4G24570), and AtDIC3 (AT5G09470) [44]. *Populus trichocarpa* has eight DICs that all belong to the AtDIC2 subgroup. Cassava (*Manihot esculenta*) possesses five DICs, clustering with the AtDIC2 subgroup, and two additional DICs in the AtDIC3 subgroup.

Arabidopsis DICs have been characterized in vitro and were shown to have varying transport kinetics but similar substrate selectivity, transporting mainly malate, OAA, succinate, maleate, malonate, phosphate, sulfate, and thiosulfate [44]. In the case of malate homo exchange, the Vmax for AtDIC3 (2.21 <sup>±</sup> 0.31 mmol min−<sup>1</sup> g protein<sup>−</sup>1) was at least double of AtDIC2 (1.01 <sup>±</sup> 0.11 mmol min−<sup>1</sup> g protein<sup>−</sup>1) and at least seven times compared to that of AtDIC1 (0.29 <sup>±</sup> 0.06 mmol min−<sup>1</sup> g protein<sup>−</sup>1). While AtDIC3 seems to be the most efficient transporter for malate, its transcript could not be detected in any of the tissues tested [44]. Recent Arabidopsis mitochondrial proteomic surveys support the observation that AtDIC3 does not appear to be highly expressed [43]. By contrast, AtDIC1 (59 protein copies per mitochondria) was found to be slightly more abundant than AtDIC2 (21 protein copies per mitochondria) [43]. However, transcript levels of *AtDIC1* and *2* appear to vary across tissues [44]. The mRNA level of the former was found to be higher in roots and in flowers, while the latter was more abundant in leaves, stems, and seedlings. It should be noted that the precise roles of the AtDIC homologs remain to be investigated *in planta*. Thus, it is still unknown whether these homologs are

functionally redundant or if indeed they can transport such a wide array of substrates. Furthermore, *K*m values were reported to be in the millimolar range, e.g., 0.4 ± 0.09 mM for AtDIC1 during malate/malate exchange [44] as opposed to the ADP/ATP carrier (AtAAC1) that was shown to be in the millimolar range [45]. The orders of magnitude of the *K*<sup>m</sup> and *V*max values are similar to ones measured for DIC homologs in *Drosophila* [46] rat and *C. elegans* [47]. We hypothesize that this might be a form of metabolic control, i.e., metabolites are exported/imported only when a certain threshold concentration is achieved. For example, malate synthesized in the mitochondria is not exported as soon as it is formed by the action of DICs, rather, only when a critical concentration is reached would DICs transport the metabolites in excess. The transport process is also rapid as reflected by the *V*max values consistent with our hypothesis of *relief* from metabolite accumulation.
