Evaluation of the roles of the cytosolic N-terminus and His-rich loop of ZNT proteins using ZNT2 and ZNT3 chimeric mutants

The physiological roles of Zn transporter (ZNT) proteins are being increasingly recognized, and three dimensional structures of ZNT bacterial homologs have facilitated our understanding of their biochemical characteristics at the molecular level. However, the biological role of the unique structural features of vertebrate ZNTs, which are absent in their bacterial homologues, is not completely understood. These ZNT sequences include a cytosolic His-rich loop between transmembrane helices IV and V and the cytosolic N-terminus. This study investigated the contribution of these features to zinc transport by ZNT proteins. The importance of the His residues in the cytosolic His-rich loop was investigated using ZNT2 Ala substitution and deletion mutants. The presence of His residues was not essential for zinc transport, even though they possibly participate in modulation of zinc transport activity. Furthermore, we determined the role of the N-terminus by characterizing ZNT2 and ZNT3 domain-swapped and deletion mutants. Unexpectedly, the N-terminus was also not essential for zinc transport by ZNT2 and the domain-swapped ZNT2 mutant, in which the cytosolic His-rich loop was substituted with that of ZNT3. These results provide molecular insights into understanding the roles of the cytosolic parts of ZNT2, ZNT3, and probably other members of their subgroup.


Results
His residues of the ZNT2 cytosolic His-rich loop. We have previously reported that znt1 −/− mt −/− znt4 −/− cells stably expressing the known H205D ZNT2 mutant (listed in the dbSNP database) showed zinc resistance similar to that of cells expressing WT ZNT2 40 . This was unexpected as His 205 is in the cytosolic Hisrich loop between TM helices IV and V (Fig. 1A), which is considered important for zinc transport activity of ZNTs 3,4 . Hence, we examined another dbSNP ZNT2 mutant, H197R, in znt1 −/− mt −/− znt4 −/− cells; similar to the H205D mutant, no defects in zinc transportation were observed ( Fig. 2A) (Supplementary Table 1).
These results prompted us to investigate the importance of loop His residues using Ala scanning. The His-rich loop of ZNT2 contains four His residues; however, H197 may belong to the TM helix IV. When stably expressed in znt1 −/− mt −/− znt4 −/− cells, monosubstituted H197A, H201A, H203A, and H205A mutants showed zinc transport similar to that of WT ZNT2 and H197R and H205D mutants (Fig. 2B). These observations indicate that a specific His residue possibly does not control zinc transport. Similar results were obtained with double Ala mutants ZNT2 (H201AH203A) , ZNT2 (H201AH205A) , and ZNT2 (H203AH205A) (Fig. 2C), suggesting that substitution of two His residues in the loop with Ala does not affect zinc transport activity. Furthermore, cells expressing triple (ZNT2 (Loop3H-3A) with Ala substituting H201, H203, and H205) and quadruple Ala mutants (all loop His residues substituted; ZNT2 (Loop4H-4A) ) still showed (moderately decreased) zinc-resistance (Fig. 2D).
As deletion of the loop segment containing the His residues (His-rich cluster) leads to loss of zinc transport activity of cation diffusion facilitator (CDF) transporters, including ZNTs 31,32 , we designed a deletion mutant of ZNT2, in which the loop segment between H201 and H205 residues was deleted (ZNT2 (Δ201-205) ), and evaluated zinc resistance of cells expressing this mutant. Compared to the resistance of cells expressing ZNT2 (Loop3H-3A) , the stable expression of ZNT2 (Δ201-205) in znt1 −/− mt −/− znt4 −/− cells significantly decreased zinc resistance (Fig. 2E). Next, we assessed the differences in stability between ZNT2 (Δ201-205) and ZNT2 (Loop4H-4A) for investigating the importance of these His residues in detail. Cells expressing each mutant were treated with cycloheximide to block further protein synthesis, and the protein expression levels were monitored periodically over 8 h by  . His residues of the cytosolic ZNT2 His-rich loop are not essential for zinc transport. (A) Zinc transport activity of znt1 −/− mt −/− znt4 −/− cells expressing H197R or H205D ZNT2 SNP mutants (ZNT2 (H197R) , ZNT2 (H205D) ) was comparable to that of WT ZNT2-expressing cells. (B) Expressing ZNT2 mutants with single His to Ala substitution in the His-rich loop (ZNT2 (H197A) , ZNT2 (H201A) , ZNT2 (H203A) , and ZNT2 (H205A) ) in znt1 cells resulted in zinc transport activity comparable to that of WT ZNT2. Similar observations were made with the double substitution mutants (ZNT2 (H201AH203A) , ZNT2 (H201AH205A) , and ZNT2 (H203AH205A) ) (C), whereas expression of ZNT2 mutants with three or four Ala-substituted His loop residues (ZNT2 (Loop3H-3A) and ZNT2 (Loop4H-4A) ), moderately decreased zinc resistance (D). (E) Expression of ZNT2 mutant with the deleted cytosolic His-rich loop (ZNT2 (Δ201-205) ) in znt1 −/− mt −/− znt4 −/− cells did not confer zinc resistance. In experiments depicted in panels (A-E), cells were grown in the presence of indicated concentrations of ZnSO 4 and the number of surviving cells was estimated using the alamarBlue assay in triplicate (representative results shown). Expression of WT or mutant ZNT2 proteins in znt1 −/− mt −/− znt4 −/− cells was confirmed by immunoblotting (lower sub-panels). Tubulin was used as the loading control. (F) Comparison of the stability of ZNT2 (Loop4H-4A) and ZNT2 (Δ201-205) mutants with those of WT ZNT2 or ZNT2 (G280R) mutant. The expression levels of each protein at each time point are shown, with representative results of immunoblotting depicted in the lower panel. The results of ZNT2 (G280R) is shown as a control for the destabilized ZNT2 mutant 42 . Data show mean ± SEM of triplicate experiments (lower sub-panels). Tubulin was used as the loading control. immunoblotting (Fig. 2F). Compared to WT ZNT2, the stability of both proteins decreased slightly, but they were more stable than the ZNT2 G280R mutant, which is known to be a TNZD-causing unstable mutant with impaired zinc transport activity 42 . These results suggest that His residues of the His-rich ZNT2 loop are not required for zinc transport. However, these residues possibly participate in zinc coordination during zinc transport, as zinc resistance was moderately reduced in ZNT2 (Loop3H-3A) and ZNT2 (Loop4H-4A) mutants.
In addition, we constructed 15 ZNT2 SNP mutants (Supplementary Table 1), among which E279K displayed reduced zinc transport and protein stability ( Supplementary Fig. 1A,B). This result is interesting, as Glu279 is hypothesized to form intermolecular salt-bridges in ZNT2, analogous to the salt bridges observed in YiiP structures 17,18 , and may provide useful information for diagnosing TNZD (see Discussion).
Interestingly, expression of the ZNT2 mutant ZNT2 (ZNT3Nter,ZNT3Loop) , in which both the N-terminus and the cytosolic His-rich loop of ZNT2 are substituted with the corresponding segments of ZNT3, in znt1 −/− mt −/− znt4 −/− cells resulted in zinc resistance comparable to that of WT ZNT2 (Fig. 3E). The stability of ZNT2 (ZNT3Nter,ZNT3Loop) was higher than that of WT ZNT2 (Fig. 3F). This observation and the above results strongly suggest that the ZNT3 N-terminus impairs zinc transport by ZNT2 and reduces its stability by perturbing the cytosolic His-rich loop of ZNT2; association with the ZNT3 N-terminus and ZNT3 His-rich loop reverses these defects.

Analysis of the importance of the ZNT N-terminus using deletion mutants.
To further investigate the involvement of the N-terminus of ZNT subgroup II in zinc transportation, we constructed the ZNT2 N-terminus (Glu2 to His62) deletion mutant (ZNT2 (Nter-del) ) based on the sequences of the short N-termini of homologous E. coli and S. oneidensis YiiP proteins (Fig. 1B) 16,18 . Unexpectedly, znt1 −/− mt −/− znt4 −/− cells expressing ZNT2 (Nter-del) showed zinc resistance similar to that of cells expressing WT ZNT2 (Fig. 5A). Minor contribution of the ZNT2 N-terminus to zinc transport was confirmed using the ZNT2 (Nter-del,Loop3H-3A) mutant, in which His 201, 203, and 205 residues of the cytosolic His-rich loop (see Fig. 2D) were replaced by Ala, and the N-terminus was deleted (Fig. 5B). Cells expressing ZNT2 (Nter-del,Loop3H-3A) showed zinc resistance similar to that of cells expressing ZNT2 (Nter-del) . This observation was additionally verified by constructing ZNT2 (Nter-del,ZNT3Loop) , in which the cytosolic His-rich loop of ZNT2 (Nter-del) was substituted by the corresponding ZNT3 sequence. Expression of ZNT2 (Nter-del,ZNT3Loop) in znt1 −/− mt −/− znt4 −/− cells resulted in zinc resistance similar to that of cells expressing ZNT2 (Nter-del) (Fig. 5A). These results indicate that the N-terminus itself is not required for ZNT2 zinc transport, even in the ZNT2 mutant in which the His-rich loop is substituted with that of ZNT3. The stability of these N-terminal deletion mutants was investigated in detail. The ZNT2 (Nter-del,ZNT3Loop) mutant was more stable, whereas the ZNT2 (Nter-del) and ZNT2 (Nter-del,Loop3H-3A) mutants were less stable than WT ZNT2 (Fig. 5C). As the stability of ZNT2 (Nter-del,Loop3H-3A) was comparable to that of ZNT2 (Nter-del) , the replacement of three His residues with Ala did not affect the stability of the ZNT2 (Nter-del) mutant.
Taken together, our results suggest that the ZNT3 N-terminus is incompatible with ZNT2 structure and perturbs its zinc transportation ability and stability; however, zinc transporting ability can be restored by combining the N-terminus and cytosolic His-rich loop of ZNT3, even when grafted on the ZNT2 backbone (TM helices).

Discussion
In this study, we obtained two unexpected but important results regarding the molecular characteristics of ZNTs, which are summarized in Table 1. First, the His residues of the cytosolic His-rich loop are not essential for zinc transport by ZNTs, although these residues possibly participate in modulating zinc transport activity (Fig. 6A). Second, the ZNT N-terminus alone is not essential for zinc transport (Fig. 6B). Although these results were obtained from studies on the members of ZNT subgroup II, to our knowledge, this is the first report presenting evidence for these conclusions, which contribute to our understanding of ZNTs at the molecular level. This is the first study to evaluate the importance of His residues of the cytosolic His-rich loop of ZNTs. His residues of the loop are important for zinc transportation, as this function is disrupted in mutant ZNTs and homologues lacking the His-rich cluster in the loop 31,32 . However, complete substitution analysis of this region, similar to that performed on ZIP4 proteins 43 , is required. Our observations that zinc resistance was moderately reduced in cells expressing ZNT2 (Loop3H-3A) and ZNT2 (Loop4H-4A) mutants indicate that His residues are not required for ZNT2 zinc transport, although they do participate in zinc transport. Previous studies suggest that His residues possibly coordinate zinc ions during transport or modulate substrate metal specificity 34,35 . Based on the observations that the stabilities of ZNT2 (Loop4H-4A) and ZNT2 (Δ201-205) mutants were not significantly altered, and that the ZNT2 (Δ201-205) mutant possesses negligible zinc resistance, we suggest that the spatial configuration of the loop is possibly more important than the presence of His residues for movement of the TM helices during zinc transport. Further studies on the role of loop sequence length are required.
The molecular functions of the ZNT N-terminus have not been as extensively investigated as those of the cytosolic C-terminus 44,45 . The homologous E. coli and S. oneidensis YiiP proteins possess short N-termini with low sequence similarity 16,18 , and diverse N-termini sequences have been observed among ZNTs. ZNT1 and ZNT10 (ZNT subgroup I) feature short N-termini, similar to YiiP 30 , while the N-terminus of the nine ZNT5 TM helices is uniquely long 46 , which renders speculation regarding N-terminus function based on sequence conservation difficult. The N-terminus was previously considered crucial for zinc transport, based on studies on complete and partial deletion ZNT2 mutants and plant homologous proteins with no zinc transport activity 36,37 . Unlike these deletion mutants, the N-terminus deletion mutants used in this study were designed based on YiiP proteins, which harbour N-termini of lengths similar to the N-termini of ZNT1 and ZNT10 (ZNT subgroup I). We show that these mutants retain zinc transport functions. These results suggest that a long N-terminal sequence is not required for forming a zinc transport-competent ZNT conformation, and that a short N-terminus is sufficient to form a compact four-helix bundle consisting of TM helices I, II, IV, and V, as observed in YiiP structures [16][17][18][19][20] . A recent report shows that a short ZNT8 isoform with moderate-length N-terminus retains zinc transport activity 47 , which may support our observations.
What is the role of the N-terminus in ZNT subgroup II members? Important features and sequence motifs have been previously identified in the N-termini of the members of this subfamily and other homologues, offering insights into the potential roles of this structural feature. The ZNT2 N-terminus features a His cluster motif involved in mitochondrial sorting 36 , the ZNT3 N-terminus contains a zinc binding (sensor) motif 33 , and the ZNT4 N-terminus possesses four consecutive L-(x)6 repeats resembling the Leu zipper motif for potential protein-protein interactions 48 , whereas the AtMTP1 N-terminus contains two Cys residues, which possibly   In (A,B), the alamarBlue assay was performed as shown in Fig. 2. Stable expression of WT and mutant ZNT2 in znt1 −/− mt −/− znt4 −/− cells was confirmed by immunoblotting (lower sub-panels). Tubulin was used as the loading control. (C) The stabilities of ZNT2 (Nter-del) , ZNT2 ZNT3Loop) , and ZNT2 (Nter-del,Loop3H-3A) were evaluated as described in Fig. 2F. Data show mean ± SEM of triplicate experiments (lower sub-panels). Tubulin was used as the loading control. Asterisk (*) denotes a significant difference between WT ZNT2, and ZNT2 (Nter-del) mutant protein levels (P < 0.05 by Dunnett's test).
Scientific REPORTS | (2018) 8:14084 | DOI:10.1038/s41598-018-32372-8 participate in zinc transport 37 . In this study, we showed that substitution of the ZNT2 N-terminus with that of ZNT3 (ZNT2 (ZNT3Nter) ) results in loss of zinc transport function of WT ZNT2, but not of the ZNT2 mutant (ZNT2 (ZNT3Nter,ZNT3Loop) ), the His-rich loop of which was also substituted with the corresponding ZNT3 loop. On the contrary, the ZNT2 N-terminus probably has the ability to stabilize ZNT2, as ZNT2 (Nter-del) was slightly less stable than WT ZNT2 (See Fig. 5C). This hypothesis was supported by the fact that ZNT2 (Nter-del,ZNT3Loop) was less stable than ZNT2 (ZNT3Nter,ZNT3Loop) (See Figs 3F and 5C). These observations suggest that the N-termini of ZNT subgroup II members interact with the cytosolic His-rich loop and modulate zinc transport, which may be potentiated by proper combinations of each protein. In ZNT3, this function may be attributed to the sequence spanning Met1 to Phe52, as substituting the sequence between Met1 to Asn50 of ZNT2 with the sequence encompassing Met1 to Phe52 of ZNT3 (ZNT2 (ZNT3-2Nter) ) moderately decreased zinc transportation by ZNT2 (Table 1). Further studies are required to extend this conclusion to other ZNTs, such as members of ZNT subgroup III and IV.  Resistance of ZNT3-expressing znt1 −/− mt −/− znt4 −/− cells to high zinc concentrations is lower than that of cells expressing ZNT2. Similar observations have been previously reported; Unlike ZNT2, which shows potent zinc transport activity in many cell lines [49][50][51][52] , ZNT3 shows weak zinc transport activity in non-neuronal cells 49 . ZNT3 is a functional zinc transporter in synaptic vesicles, where it uses specific ions as the driving force for zinc transportation 53,54 . Our results indicate that the three ZNT3 cytosolic portions (the N-terminus, His-rich loop, and C-terminus) are functional when fused to the ZNT2 backbone, although the N-terminus has to be fused simultaneously with the His-rich loop, suggesting that the TM helices, and not the cytosolic parts of ZNT3, determine ZNT3-specific zinc transport in synaptic vesicles.
Another important aspect of this study was the characterization of ZNT2 SNP mutants, as even heterozygous ZNT2 mutations may reduce zinc levels in the breast milk of affected mothers, leading to TNZD in exclusively breastfed infants 24,25,42,52,[55][56][57][58] . TNZD-associated zinc deficiency is effectively counteracted by zinc supplements 12,13 and accumulating genetic information associated with TNZD pathogenesis assists in supporting normal growth and development of breastfed infants 13,58 . Previously, we identified four ZNT2 SNPs associated with high-TNZD risk 42 , one of which was recently confirmed to be clinically relevant 59 . In this study, we report E279K as a new ZNT2 SNP associated with high-risk of TNZD, based on similarities with TNZD-causing G280R SNP ( Supplementary Fig. 1A,B) 42 . Sequence-wise, the ZNT2 Glu279 residue corresponds to Asp207 and 209 residues of E. coli 16 and S. oneidensis YiiP 18 proteins, respectively, which form intermolecular salt-bridges with Lys77 and Lys79 YiiP residues (in E. coli and S. oneidensis, respectively), ensuring proper orientation of TM helices III and VI at the dimer interface 17 . These Lys residues correspond to the Arg138 residue in ZNT2 (see Supplementary  Fig. 1C), suggesting that Glu279 possibly participates in the formation of similar salt-bridges in ZNT2.
In conclusion, our observations revealed that sequences unique to vertebrate ZNTs, including the cytosolic His-rich loop between TM helices IV and V and the cytosolic N-terminus, both of which have been previously associated with zinc transport, are not essential for zinc transport, at least by ZNT subgroup II members. Recent evidence regarding the diverse physiological functions of ZNTs and the role of impaired ZNT function in human diseases indicates that ZNTs can be developed as novel therapeutic targets. Biochemical characterization of ZNTs is important considering the growing interest in the pathophysiological functions of these proteins.
Evaluation of WT and mutant ZNT2 and ZNT3 zinc transport function based on viability of znt1 −/− mt −/− znt4 −/− cells in the presence of high zinc concentrations. The cells were seeded at a density of 10 5 cells/ml in 96-well plates and treated with 50-90 µM ZnSO 4 for 2 days as described previously 38 . The alamarBlue reagent (AbD Serotec, Ltd., Oxford, UK) was added to the culture media and the cells were incubated for 4 h. Absorbance was determined at 570 and 600 nm using PowerScan 4 (DS Pharma Biomedical, Osaka, Japan).
Evaluation of the stability of ZNT2 mutants. The stability of mutant ZNT2 proteins was evaluated as described previously 24,42 . Cells expressing WT or mutant ZNT2 were treated with cycloheximide to block further protein synthesis and were collected periodically over 8 h. Total cell lysates were prepared and subjected to immunoblotting to monitor ZNT2 levels. The band intensity of each protein is the average of data from three independent experiments and is shown as the percentage of the intensity at 0 h (T 0 ) after normalization. Densitometric quantification was performed using ImageQuant (GE Healthcare). Experiments were performed in triplicate.

Statistical analyses.
Data are represented as means ± standard deviation (SD) or standard error of mean (SEM). Statistical significance was determined using Dunnett's test. P-values < 0.05 were considered statistically significant.

Data Availability
All data generated or analysed during this study are included in this published article (and its supplementary information files) or are available upon reasonable request.