Stimulation of the human mitochondrial transporter ABCB10 by zinc-mesoporphrin

Heme biosynthesis occurs through a series of reactions that take place within the cytoplasm and mitochondria, so intermediates need to move across these cellular compartments. However, the specific membrane transport mechanisms involved in the process are not yet identified. The ATP-binding cassette protein ABCB10 is essential for normal heme production, as knocking down this transporter in mice is embryonically lethal and accompanied by severe anemia plus oxidative damage. The role of ABCB10 is unknown, but given its location in the inner mitochondrial membrane, it has been proposed as a candidate to export either an early heme precursor or heme. Alternatively, ABCB10 might transport a molecule important for protection against oxidative damage. To help discern between these possibilities, we decided to study the effect of heme analogs, precursors, and antioxidant peptides on purified human ABCB10. Since substrate binding increases the ATP hydrolysis rate of ABC transporters, we have determined the ability of these molecules to activate purified ABCB10 reconstituted in lipid nanodiscs using ATPase measurements. Under our experimental conditions, we found that the only heme analog increasing ABCB10 ATPase activity was Zinc-mesoporphyrin. This activation of almost seventy percent was specific for ABCB10, as the ATPase activity of a negative control bacterial ABC transporter was not affected. The activation was also observed in cysteine-less ABCB10, suggesting that Zinc-mesoporphyrin’s effect did not require binding to typical heme regulatory motifs. Furthermore, our data indicate that ABCB10 was not directly activated by neither the early heme precursor delta-aminolevulinic acid nor glutathione, downsizing their relevance as putative substrates for this transporter. Although additional studies are needed to determine the physiological substrate of ABCB10, our findings reveal Zinc-mesoporphyrin as the first tool compound to directly modulate ABCB10 activity and raise the possibility that some actions of Zinc-mesoporphyrin in cellular and animal studies could be mediated by ABCB10.

Response: we would love to determine the effect of ZnMP in cells. However, performing such experiments in cultured mammalian cells is beyond the current capabilities in our laboratory since we do not have the necessary equipment nor expertise. We are a small new laboratory (Maria E. Zoghbi is an Assistant Professor) in a relatively new campus (UC Merced was established only 15 years ago) that still lacks many resources commonly available in more established research institutions. UC Merced is in the California Central Valley, with no nearby facilities from other institutions. In addition, we do not have a collaborator currently available to help us with such experiments. Our goal with this article is to report ZnMP as a new tool for the study of ABCB10 and we hope this information will open the doors for future research, including studies at the cellular level. We do intend to expand our research capabilities once we can secure the space and financial resources needed to establish our mammalian cell culture room.
Since we are only using the transporter reconstituted in nanodiscs as our experimental system to study the effect of drugs on the protein, we have now included in the introduction (page 5) some references to studies performed using a similar approach for the well-known multidrug ABC transporter P-glycoprotein: "Monitoring changes in basal ATPase activity is an effective biochemical approach to help identify substrates of ABC transporters and reconstitution of the detergent purified protein into lipid nanodiscs has been successfully used to study the effect of substrate on the activity of other well-known ABC transporters such as P-glycoprotein [35,[37][38][39].

Other critics
1. The flow of the paper is unusual with the authors referring to fig 3A in the introduction. If the point of 3A is to introduce the structures of Abcb10, perhaps it should be either included in figure 1 or referenced in the results section. Either way, both figure 1 and figure 3A are already published information and do not advance the field, thus, they do not seem unnecessary for the reader.
Response: We completely agree with the observation of unusual flow and we have therefore removed the reference to fig 3A from the introduction. However, we do think that having a figure showing the structure of many of the molecules we are testing (Fig 1) to the reader facilitates the understanding of the article. In a similar way, Fig 3A (Fig 4A in the new version) is based on an available crystal structure, but it helps the reader to visually locate the cysteine residues that are being mutated in the transporter.
2. It would be nice to see if a Cysteine-less Abcb10 can complement the loss of Abcb10 in cell culture.
Response: We absolutely agree. Unfortunately, as mentioned above, we do not have the capability to perform experiments in mammalian cell cultures.
3. The authors use bacterial Abc transporter MsbA as a negative control. Is there a positive control Abc transporter that could be used to show that it can be activated by a known substrate but that the substrate does not affect Abcb10? Possible Positive control would be Abcb6, which is proposed to be a porphyrin transporter.
Response: We used MsbA as a negative control since we have priorly studied this transporter and we can express and purify this protein in our laboratory using E. coli. Likewise, as we stated in the methods, we have expressed and purified ABCB10 using E. coli by a procedure we have developed, and since that article has been recently accepted (Saxberg et al, https://doi.org/10.1016/j.pep.2020.105778), we are now citing that work in the revised version of this manuscript (reference #41, in page 6-7 of the method section). In that article we show that ABCB10 produced in E. coli is properly folded and functional, similarly to the protein produced in insect cells. However, the available protocols for expression and purification of ABCB6 use mammalian cells and/or Pichia pastoris. Once again, unfortunately, we do not currently have the capability of using those expression systems in our laboratory. Therefore, purified ABCB6 was not available to be included as a positive control for some of the porphyrins in our experiments. Nevertheless, we have taken precautions to use freshly prepared stocks of all porphyrins, all of them were purchased from the same manufacturer, and the stocks were tested simultaneously during the same experiment. Under those conditions, ABCB10 was activated by ZnMP but not by the other porphyrins.

What is the kinetics of MsbA activity compared to Abcb10? Is it a very effective ATPase?
Response: We appreciate this very relevant question from the reviewer. The ATPase activity of MsbA is about ten times higher than that of ABCB10 and we have now added information in the methods to address this issue (page 8-9). The low ATPase activity of ABCB10 is typical for human ABC transporters and has also been reported by Shintre et al. (DOI: 10.1073/pnas.1217042110). This comparison of ATPase rates is also directly related to the reviewer's comment about Sup fig 1. The rate of the Abs340 decay could be faster if we add more protein per well. However, we intentionally design the experiments to record the activity for a long period of time as a way to confirm the protein's stability during the assay. We want to be sure we are measuring the response of a stable protein, capable of withstanding several hours of activity at 37 o C. Since the ATPase activity of MsbA is higher, we use less MsbA for the assay, so the time scales can be comparable. If we use too much transporter, the Abs340 will decrease to zero in a few minutes and we will miss important information.
Sup fig 1 shows a time course of absorbance change. Seems rather slow…..on the order of several hours to get 50% reduction. Is this a normal rate for an Abc transporter? Some comment is needed Response: Thank you for the comment. Please refer to point #4 above.
Sup fig 2 atpase activity inhibited by increasing hemin -# of replicates? The graph looks like there may be only 2 replicates at some time points. If this is the case, the scientific rigor seems to be lacking for this supplemental figure.
Supplemental figures seem import -not sure why they are supplemental and the reader would recommend moving into the results section as figures.
Response: We have followed the reviewer's recommendations about the supplemental figures. We moved former S1 Fig (part B) to the main text (found now as Fig 2D,  Text has been modified accordingly to take into consideration the new distribution of figures (all modifications can be followed as red track changes). We have also included the following statement in the methods (page 9): "Data are presented as mean ± standard deviation of at least three independent experiments performed with independent protein purifications and reconstitutions." Previous studies with Glutathione were done in submitochondrial particles which would be more reflective of what happens in vivo and should be mentioned as a caveat to the nanodisc experiments, which do not have potential activation of partner proteins.
Response: we would like to mention that in the original version we did acknowledge the prior experiments done in submitochondrial particles and the fact that we cannot have glutathionylation of the purified protein. The statement, now in page 13-14 of the revised version, says: "Under our experimental conditions, similar concentrations of glutathione did not affect the ATPase activity of the transporter ( Fig  2D 3D), suggesting that neither GSSG nor GSH have a direct effect on ABCB10.