*2.4. Functional Consequences of Changing All 11 Amino Acids of the Modeled rAbcc6 Substrate-Binding Site into Those That Comprise the bAbcc1 LTC<sup>4</sup> Binding Site*

As most mutants with single amino acid changes in the modeled rAbcc6 substrate binding cavity retained at least 25% activity, we wondered what the consequences would be if the entire modeled substrate binding cavity was altered such that it more closely mimicked that of ABCC1. Of the fourteen amino acids corresponding to those in the bAbcc1 cryoEM structure forming the LTC4-binding cavity, three are identical in rAbcc6 (rAbcc6 R1168, W1217 and R1220) (Table 1 and Figure 2). Thus, the remaining eleven amino acids of the modeled substrate-binding cavity were changed into residues found in bAbcc1 at the same positions (L316K, S319H, E365L, M369F, H424Y, F537W, L534T, K578F, T1064M, T1214Y, and Q1216N) to generate a rAbcc6 mutant protein we have termed rAbcc6-11aa.

As shown in Figure 7A, the rAbcc6-11aa was expressed at about 6.5-fold lower levels than the wild-type rAbcc6 protein when overexpressed in HEK293 cells and appeared to have a faster electrophoretic mobility (Figure 7A). The reason for the altered mobility of rAbcc6-11a is not known but may be due to changes in glycosylation or other posttranslational modification. If underglycosylated, the mutant rAbcc6-11aa protein may have a faster turn-over time thus explaining why it is expressed at lower levels than the wildtype protein. However, even if the mutations caused some misfolding of the transporter during its biosynthesis that impaired its glycosylation, the altered protein structure has remained stable enough to traffic to the plasma membrane where it can still carry out active transport. Thus, rAbcc6-11aa still mediated ATP release, as illustrated by both PPi accumulation in the medium and by real-time ATP efflux assays (Figure 7C,D, respectively). After adjustment for lower protein levels, rAbcc6-11aa seemed to even display higher activity than the wild-type protein. These data indicate that changing these 11 amino acids may have affected the stability of rAbcc6 but had minimal effect on the intrinsic activity of the protein. Consistent with the significant activity of the rAbcc6-11aa mutant protein, its abundance in the plasma membrane was also relatively high (Figure 7B).

We next set out to test if the rAbcc6-11aa mutant with the hABCC1/bAbcc1 LTC4 binding cavity residues transports LTC4. Of note, an initial characterization of hABCC6 indicated it could transport LTC<sup>4</sup> [6], although this was not the case for its ortholog rAbcc6 [35]. Nevertheless, we reasoned that introducing the amino acids in bAbcc1 thought to form the bipartite LTC4-binding cavity at the corresponding positions in rAbcc6 might establish LTC<sup>4</sup> transport in the latter ABC transporter. The cellular ATP efflux capacity of the rAbcc6-11aa indicated the protein retained activity. However, in vitro vesicular transport experiments (widely held to be the gold-standard to confirm that a molecule is a substrate of a specific ABC transporter) failed to demonstrate LTC<sup>4</sup> transport by the rAbcc6- 11aa protein (data not shown). Low levels of LTC<sup>4</sup> transport might have been missed, however, because of the low expression levels of rAbcc6-11aa in our system (HEK293 cells).

‐ ‐ ‐ ‐ μ ‐ ‐ ‐ ‐ **Figure 7.** Expression, activity, and subcellular localization of the rAbcc6-11aa mutant. (**A**): Detection of rAbcc6-11aa in HEK293 cells by immunoblot analysis. (**B**): Subcellular localization of rAbcc6-11aa in HEK293 cells. Red: Na+/K<sup>+</sup> -ATPase, a marker for the plasma membrane; Green: rAbcc6; Blue: DAPI. All scale bars represent 30 µm. (**C**): PPi accumulation in the medium of the indicated HEK293 cell lines (**D**): ATP efflux from the indicated HEK293 cell lines. Data are presented as means ± SD for (**C**), while means of representative experiments with at least 4 replicates are shown for (**D**). wild-type: wild-type rAbcc6, control: parental HEK293 cells. The dashed line in (**C**) indicates the average PPi level in medium of HEK293 cells overproducing wild-type rAbcc6, which was set at 100%. The slight differences in electrophoretic mobility of some of the mutants may be attributed to altered glycosylation or other post-translational modification. In panels C and D values have been adjusted to take differences in protein expression of the mutants relative to wild type rAbcc6 into account.
