Bacillus anthracis' PA63 Delivers the Tumor Metastasis Suppressor Protein NDPK-A/NME1 into Breast Cancer Cells.

Some highly metastatic types of breast cancer show decreased intracellular levels of the tumor suppressor protein NME1, also known as nm23-H1 or nucleoside diphosphate kinase A (NDPK-A), which decreases cancer cell motility and metastasis. Since its activity is directly correlated with the overall outcome in patients, increasing the cytosolic levels of NDPK-A/NME1 in such cancer cells should represent an attractive starting point for novel therapeutic approaches to reduce tumor cell motility and decrease metastasis. Here, we established the Bacillus anthracis protein toxins’ transport component PA63 as transporter for the delivery of His-tagged human NDPK-A into the cytosol of cultured cells including human MDA-MB-231 breast cancer cells. The specifically delivered His6-tagged NDPK-A was detected in MDA-MB-231 cells via Western blotting and immunofluorescence microscopy. The PA63-mediated delivery of His6-NDPK-A resulted in reduced migration of MDA-MB-231 cells, as determined by a wound-healing assay. In conclusion, PA63 serves for the transport of the tumor metastasis suppressor NDPK-A/NME1 into the cytosol of human breast cancer cells In Vitro, which reduced the migratory activity of these cells. This approach might lead to development of novel therapeutic options.


Introduction
Metastasis is the leading cause of mortality in cancer patients, even if the primary tumor has been removed. Due to the complexity of the processes in metastasis major breakthroughs in inhibiting this often deadly spreading of cancer cells have not been achieved so far [1,2]. Metastasis suppressors are proteins which exert no inhibitory effect on primary tumor growth, but significantly reduce metastasis. Thus, specifically increasing the amount of such a protein present in highly metastatic tumor cells might offer an opportunity to interfere with these processes [3,4].
The human nm23-H1 gene, now more generally named nme1, was the first metastasis suppressor identified [5]. By comparison with its D. melanogaster orthologue, awd, it became evident that it encoded the A isoform of the enzyme nucleoside diphosphate kinase (NDPK-A, also named NME1) [6]. In mammals, the nme family consists of 10 genes, although the gene products NME1 and NME2 members have been studied with regard to metastasis in more detail. Overexpression of, for example, cancer cells and reduce their motility. The human breast carcinoma cell line MDA-MB-231 was used as model in this study because these cells are highly metastatic and show relatively low basal expression levels of endogenous NDPK-A/NME1. In conclusion, the results demonstrate that His 6 -NDPK-A is delivered into MDA-MB-231 cells by PA 63 -mediated transport, which reduced migration of these breast cancer cells.

Purification and Characterization of Recombinantly Expressed His 6 -NDPK-A
His 6 -NDPK-A was recombinantly expressed in E. coli BL21 and purified via affinity chromatography. The identity of the purified His 6 -NDPK-A protein was confirmed by Western blotting with a specific antibody directed against the amino acid residues 134-152 of human NDPK-A ( Figure 1B). To verify the enzyme activity of the purified His 6 -NDPK-A, its intrinsic nucleoside diphosphate kinase activity was analyzed by Western blotting. After performing an autophosphorylation assay, the 1-phosphohistidine-specific antibody confirmed the presence of the enzyme intermediate ( Figure 1C). Furthermore, the His 6 -NDPK-A-catalyzed conversion of ADP to ATP was quantified by In Vitro kinase assay. The result shows a concentration-dependent increase of the luminescence signal and the control reactions revealed the substrate dependency ( Figure 1D). Taken together, the recombinant His 6 -NDPK-A was enzymatically active and could be used in further experiments to analyze its transport into cells via the PA 63 transporter.
Int. J. Mol. Sci. 2020, 21,3295 3 of 20 MDA-MB-231 was used as model in this study because these cells are highly metastatic and show relatively low basal expression levels of endogenous NDPK-A/NME1. In conclusion, the results demonstrate that His6-NDPK-A is delivered into MDA-MB-231 cells by PA63-mediated transport, which reduced migration of these breast cancer cells.

Purification and Characterization of Recombinantly Expressed His6-NDPK-A
His6-NDPK-A was recombinantly expressed in E. coli BL21 and purified via affinity chromatography. The identity of the purified His6-NDPK-A protein was confirmed by Western blotting with a specific antibody directed against the amino acid residues 134-152 of human NDPK-A ( Figure 1B). To verify the enzyme activity of the purified His6-NDPK-A, its intrinsic nucleoside diphosphate kinase activity was analyzed by Western blotting. After performing an autophosphorylation assay, the 1-phosphohistidine-specific antibody confirmed the presence of the enzyme intermediate ( Figure 1C). Furthermore, the His6-NDPK-A-catalyzed conversion of ADP to ATP was quantified by in vitro kinase assay. The result shows a concentration-dependent increase of the luminescence signal and the control reactions revealed the substrate dependency ( Figure 1D). Taken together, the recombinant His6-NDPK-A was enzymatically active and could be used in further experiments to analyze its transport into cells via the PA63 transporter.  coli and purification via affinity chromatography, protein identity was confirmed by Western blotting with a specific anti-NDPK-A antibody (B). The kinase activity of His 6 -NDPK-A was analyzed by measuring the proteins autophosphorylation activity (C). His 6 -NDPK-A was incubated with ATP (1 mM) and MgCl 2 (2 mM) for 5 min at room temperature. The reaction was stopped by adding EDTA (5 mM) and the phosphorylated His 6 -NDPK-A intermediate detected by Western blotting with an anti-N1-phosphohistidine (1-pHis) antibody. (D) The kinase activity of His 6 -NDPK-A was determined by measuring the substrate conversion status after incubation of increasing amounts His 6 -NDPK-A with substrate for 30 min at room temperature. Controls include His 6 -NDPK-A only, substrate only and buffer only. ATP conversion was quantified with the Kinase-Glo ® Luminescent Kinase Assay.

MDA-MB-231 Cells are Susceptible to PA 63 -Mediated Delivery of His-Tagged Protein Cargo
To investigate whether His 6 -NDPK-A can be delivered into the cytosol of MDA-MB-231 breast cancer cells via the PA 63 transporter, we first confirmed that this cell line is susceptible for the PA 63 -mediated protein transport using a well-established His 6 -tagged cargo protein, His 6 -DTA, the recombinant enzyme domain of diphtheria toxin (DT). When His 6 -DTA reaches the cytosol, it ADP-ribosylates elongation factor EF-2 and thereby inhibits protein synthesis. This results in cell rounding, which can be easily monitored by microscopic methods and represents a sensitive and specific endpoint to monitor the transport of enzymatically active His 6 -DTA into the cytosol. His 6 -DTA was applied to the cells in presence or absence of PA 63 and the DTA-induced cell rounding analyzed (Figure 2A). The morphological changes were obvious after 4 h and even more evident after 24 h. The changes in cell morphology were only observed after application of the combination of His 6 -DTA plus PA 63 but not after treatment with His 6 -DTA alone, clearly indicating the specific PA 63 -mediated delivery of His 6 -DTA into the cytosol of MDA-MB-231 cells. Moreover, the treatment of the cells with bafilomycin A1 (BafA1), which inhibits endosomal acidification, prevented the changes of cell morphology after treatment with His 6 -DTA plus PA 63 , confirming the specific PA 63 -mediated transport. However, MDA-MB-231 cells responded slower to PA 63 /His 6 -NDPK-A than HeLa cells ( Figure 2B), where rounding was complete after 4 h. Therefore, the incubation times for the next experiments where adjusted accordingly and HeLa cells were tested in parallel to the MDA-MB-231 cells. For both cell lines, these results were confirmed by quantitative analysis of rounded cells. In conclusion, the results clearly showed that this breast cancer cell line can be used to investigate the transport of His-NDPK-A via PA 63 .

PA 63 Mediates Delivery of His 6 -NDPK-A into Cells
Prompted by the finding that PA 63 serves for delivery of His 6 -tagged DTA into the cytosol of MDA-MB-231 cells, we next tested whether His 6 -NDPK-A can be introduced into these cells by this approach. To this end, we first analyzed the binding of His 6 -NDPK-A in the absence and presence of PA 63 to MDA-MB-231 ( Figure 3A) and HeLa cells ( Figure 3B). After incubation of the cells at 4 • C, where binding but not uptake occurs, with His 6 -NDPK-A in presence or absence of PA 63 , the amounts of cell-bound His 6 -NDPK-A were analyzed by Western blotting. For both cell lines, there was some unspecific binding of His 6 -NDPK-A in the absence of the transporter. However, for both cell lines, the amount of cell-bound His-NDPK-A clearly increased in the presence of PA 63 (Figure 3), indicating a specific binding of His-NDPK-A to the cells via PA 63 . HeLa cells bound more His-NDPK-A via PA 63 , which might be due to higher expression of anthrax toxin receptor in this cell line. (1 µg/mL) was added and cells were further incubated at 37 °C for 24 h. Representative pictures are shown before application of the proteins and after 4 and 24 h of treatment. The number of rounded cells at the indicated time points were quantified using ImageJ and the data is given as mean ± SD, n = 3. Significance was tested using One-Way ANOVA followed by Dunnett's multiple comparison test. * p ≤ 0.05; *** p ≤ 0.001, **** p ≤ 0.0001 vs. untreated controls. Then, His 6 -DTA (5 µg/mL) with or without PA 63 (1 µg/mL) was added and cells were further incubated at 37 • C for 24 h. Representative pictures are shown before application of the proteins and after 4 and 24 h of treatment. The number of rounded cells at the indicated time points were quantified using ImageJ and the data is given as mean ± SD, n = 3. Significance was tested using One-Way ANOVA followed by Dunnett's multiple comparison test. * p ≤ 0.05; *** p ≤ 0.001, **** p ≤ 0.0001 vs. untreated controls.
was some unspecific binding of His6-NDPK-A in the absence of the transporter. However, for both cell lines, the amount of cell-bound His-NDPK-A clearly increased in the presence of PA63 (Figure 3), indicating a specific binding of His-NDPK-A to the cells via PA63. HeLa cells bound more His-NDPK-A via PA63, which might be due to higher expression of anthrax toxin receptor in this cell line. Having confirmed the PA63-mediated binding of His6-NDPK-A to MDA-MB-231 cells, the delivery of His6-NDPK-A via PA63 into these cells was analyzed. Here, His6-DTA was again included as a control because the PA-mediated transport into the host cell cytosol can be easily monitored in terms of cell rounding. For the analysis via immunofluorescence microscopy, MDA-MB-231 cells were incubated for 48 h at 37 °C with His6-NDPK-A in the absence or presence of PA63, and His6-NDPK-A was detected with a specific antibody against the His-tag and Alexa488-conjugated secondary antibody ( Figure 4A). In parallel, the same approach was performed with His6-DTA instead of His6-NDPK-A. The results show an increased signal for both His6-tagged cargo proteins (green) in the presence of PA63, clearly indicating the specific PA63-dependent transport into the cells. Moreover, cells treated with His6-DTA plus PA63 rounded up, indicating that His6-DTA reached the After washing, the amount of His 6 -NDPK-A bound to the cells was determined by Western blotting with an anti-His-antibody. Hsp90 was detected to confirm comparable protein loading. Densitometric analysis was performed using the ImageJ software. Signal intensity of His 6 -NDPK-A was quantified against the Hsp90 signal.
Having confirmed the PA 63 -mediated binding of His 6 -NDPK-A to MDA-MB-231 cells, the delivery of His 6 -NDPK-A via PA 63 into these cells was analyzed. Here, His textsubscript6-DTA was again included as a control because the PA-mediated transport into the host cell cytosol can be easily monitored in terms of cell rounding. For the analysis via immunofluorescence microscopy, MDA-MB-231 cells were incubated for 48 h at 37 • C with His 6 -NDPK-A in the absence or presence of PA 63 , and His 6 -NDPK-A was detected with a specific antibody against the His-tag and Alexa488-conjugated secondary antibody ( Figure 4A). In parallel, the same approach was performed with His 6 -DTA instead of His 6 -NDPK-A. The results show an increased signal for both His 6 -tagged cargo proteins (green) in the presence of PA 63 , clearly indicating the specific PA 63 -dependent transport into the cells. Moreover, cells treated with His 6 -DTA plus PA 63 rounded up, indicating that His 6 -DTA reached the cytosol. Here, the signal was stronger compared to the cells treated with His 6 -NDPK-A plus PA 63 . The same set of experiments were performed with HeLa cells to exclude a cell-line specific effect. HeLa cells also showed the PA 63 -dependent transport of His 6 -NDPK-A and His 6 -DTA or ( Figure 4B). Already after 3 h of incubation, the signal was obviously increased for His 6 -NDPK-a as well as His 6 -DTA when the transporter was present. Furthermore, co-staining of filamentous actin (red, bottom row) visualized the His 6 -DTA-induced cell-rounding.
The same set of experiments were performed with HeLa cells to exclude a cell-line specific effect. HeLa cells also showed the PA63-dependent transport of His6-NDPK-A and His6-DTA or ( Figure 4B). Already after 3 h of incubation, the signal was obviously increased for His6-NDPK-a as well as His6-DTA when the transporter was present. Furthermore, co-staining of filamentous actin (red, bottom row) visualized the His6-DTA-induced cell-rounding. To analyze the PA63-mediated transport of His6-NDPK-A in an alternative approach, cells were analyzed by Western blotting. As shown in Figure 5A, the difference in the molecular mass of the endogenous NDPK-A (17 kDa) and the recombinant His6-NDPK-A (18 kDa) taken up was clearly visible. There was no signal for His6-NDPK-A in untreated cells further confirming the identity of the introduced His6-NDPK-A protein. After incubation for 48 h with His6-NDPK-A in the presence To analyze the PA 63 -mediated transport of His 6 -NDPK-A in an alternative approach, cells were analyzed by Western blotting. As shown in Figure 5A, the difference in the molecular mass of the endogenous NDPK-A (17 kDa) and the recombinant His 6 -NDPK-A (18 kDa) taken up was clearly visible. There was no signal for His 6 -NDPK-A in untreated cells further confirming the identity of the introduced His 6 -NDPK-A protein. After incubation for 48 h with His 6 -NDPK-A in the presence of PA 63 , MDA-MB-231 cells showed a concentration-dependent increase of the amount of recombinant protein present. Interestingly, lower amounts of His 6 -NDPK-A were apparently taken up into the cells in the absence of PA 63 . Noteworthy, the introduction of His 6 -NDPK-A by PA 63 into MDA-MB-231 cells did not impair cell viability at least within 48 h, as confirmed by the MTS test, monitoring mitochondrial activity ( Figure 5B). As expected, the positive control DMSO (20% by vol.) reduced cell viability by about 50%. of PA63, MDA-MB-231 cells showed a concentration-dependent increase of the amount of recombinant protein present. Interestingly, lower amounts of His6-NDPK-A were apparently taken up into the cells in the absence of PA63. Noteworthy, the introduction of His6-NDPK-A by PA63 into MDA-MB-231 cells did not impair cell viability at least within 48 h, as confirmed by the MTS test, monitoring mitochondrial activity ( Figure 5B). As expected, the positive control DMSO (20% by vol.) reduced cell viability by about 50%.  -taged NDPK-A only which was detected at 18 kDa. Endogenous NDPK-A is detected in the cell lysates at 17 kDa. Hsp90 was detected as loading control and densitometric analysis was performed using the ImageJ software. The signal intensity of His 6 -NDPK-A was quantified against Hsp90. (B) Cell viability is not affected by the treatment of cells with His 6 -NDPK-A and PA 63 . After incubation of the cells with His 6 -NDPK-A and PA 63 or His 6 -NDPK-A-only for 48 h at 37 • C, cell viability was determined by the MTS assay. Data were normalized to control (untreated cells). The known reduction by 20% (by vol.) DMSO was used as control. Data are given as mean ± SD, n = 3. (C) Uptake of His 6 -DTA is mediated by PA 63 and can be blocked by BafA1. Cells were incubated with BafA1 (30 nM) for 30 min at 37 • C followed by 12 h incubation with His 6 -DTA (5 µg/mL) with or without PA 63 (1 µg/mL). Representative pictures depict no morphological changes after incubation and before sample preparation. His 6 -DTA was detected by an anti-His 6 -antibody. Detection of Hsp90 served as loading control. (D) PA 63 mediates cytosolic localization His 6 -DTA in MDA-MB-231 cells. After incubation with BafA1 (30 nM) for 30 min, His 6 -DTA (5 µg/mL) with or without PA 63 (1 µg/mL) was added for 12 h at 37 • C. After cell lysis, In Vitro ADP-ribosylation was performed using biotinylated NAD + (10 mM) as substrate. Modified proteins were precipitated by strep-POD and biotinylated EF-2 was detected by immunoblotting. GAPDH expression was used as loading control. (E) Cell viability is decreased after treatment of cells with His 6 -DTA and PA 63 . Cell viability was determined via the MTS assay after 24 h of incubation at 37 • C with His 6 -DTA with or without PA 63 . Staurosporine (1 µM) was used as positive control. Data are normalized to untreated controls and given as mean ± SD, n = 3. Significance was tested using One-Way ANOVA followed by Dunnett's multiple comparison test. ** p ≤ 0.01 vs. untreated controls.
Again, His 6 -DTA was used as a control cargo to monitor the cytosolic localization of His 6 -tagged proteins after delivery into MDA-MB-231 via PA 63 . Western blot analysis revealed His 6 -DTA uptake for cells treated with His 6 -DTA plus PA 63 , but not in cells treated with His 6 -DTA alone ( Figure 5C). Accordingly, treatment of cells with BafA1 prevented the PA 63 -mediated transport of His 6 -DTA into the cells ( Figure 5C). The presence of enzymatically active His 6 -DTA in the host cells cytosol was confirmed by sequential-ADP-ribosylation for the ADP-ribosylation status of the elongation factor EF-2 ( Figure 5D). Lysates were incubated with fresh DTA and biotin-NAD and the amount of biotinylated, i.e., ADP-ribosylated EF-2, was determined by Western blotting. In contrast to all other conditions investigated, biotinylated EF-2 was detected in lysates from cells which were treated before with His 6 -DTA plus PA 63 , indicating that the EF-2 of these cells was already ADP-ribosylated by His 6 -DTA in the cytosol during the incubation of the still living cells. The presence of biotinylated EF-2 in all other samples indicates the absence of His 6 -DTA in the cytosol of the living cells prior to their lysis and the In Vitro assay ( Figure 5D). Of note, also treatment of living cells with His 6 -DTA alone did not show any reduction, indicating that His 6 -DTA was not taken up into the cytosol of cells without PA 63 . Accordingly, treatment with BafA1 abolished the effect of His 6 -DTA plus PA 63 , indicating the specific uptake of His 6 -DTA via the PA 63 -dependent transport mechanism. In line with the EF-2 ADP-ribosylation data, the cell viability was significantly reduced after the 24-h incubation of cells with His 6 -DTA plus PA 63 ( Figure 5E). Again, this effect was prevented by BafA1. Treatment with staurosporine, a known inducer of apoptosis, reduced cell viability by about 50%.
Comparable results were obtained for HeLa cells (Figure 6), indicating that the findings are not restricted to a single cell line but have a broader implication. Again, the uptake of His 6 -NDPK-A into cells was clearly increased in the presence of PA 63 , as analyzed by Western blotting (Figure 6A), but did not exhibit adverse effects on the cell viability within at least 48 h of incubation ( Figure 6B). The PA 63 dependency was also observed for the uptake of His 6 -DTA and pre-incubation of cells with BafA1 blocked this uptake as analyzed in terms of His 6 -DTA-induced cell-rounding and Western blot detection of the internalized His 6 -DTA ( Figure 6C). Moreover, only in the presence of PA 63 , His 6 -DTA reached the cytosol of cells as analyzed by sequential ADP-ribosylation of EF-2 ( Figure 6D). In line with these results, cell viability was decreased when cells were treated with His 6 -DTA plus PA 63 due to the His 6 -DTA activity in the cytosol ( Figure 6E). However, even after 24 h, BafA1 prevented this effect, indicating the specificity of the His 6 -DTA transport into cells.
In conclusion, these results confirm a PA 63 -dependency for the delivery of His 6 -NDPK-A and His 6 -DTA into MDA-MB-231 and HeLa cells. For His 6 -DTA, the cytosolic localization as an active enzyme could be demonstrated due to its specific modification of the cytosolic substrate EF-2. Moreover, the PA 63 -mediated delivery of His 6 -NDPK-A is not cell-line specific.

Figure 6. PA63 mediates delivery of His6-NDPK-A and His6-DTA into HeLa cells. (A) PA63 delivers
His6-NDPK-A into HeLa cells in a PA63-dependent manner. After incubation of the cells with His6-NDPK-A (10 µg/mL) and PA63 (1 µg/mL) or His6-NDPK-A only for 6 h at 37 °C, the cells were analyzed by immunoblotting. The uptake of His6-NDPK-A was detected with an anti-His-antibody. Hsp90 was detected as loading control and densitometric analysis performed using the ImageJ software. Signal intensity of His6-NDPK-A was quantified against Hsp90. (B) Cell viability is not reduced by treatment of cells with His6-NDPK-A plus PA63. After incubation of the cells with His6-NDPK-A plus PA63 or His6-NDPK-A alone for 48 h at 37 °C, the cell viability was determined by the MTS assay. Data were normalized to untreated cells (control). DMSO served as positive control. Data are given as mean ± SD, n = 3. (C) Uptake of His6-DTA is mediated by PA63 and is sensitive to BafA1. Cells were incubated with BafA1 (30 nM) for 30 min at 37 °C followed by 24 h incubation with His6-DTA (5 µg/mL) with or without PA63 (1 µg/mL). Representative pictures are shown. Thereafter, cell lysates were analyzed by immunoblotting for the presence of His6-DTA with an anti-His-antibody. Detection of Hsp90 was used as loading control. (D) PA63 mediates delivery of His6-DTA into the cytosol of HeLa cells. Substrate modification was determined via strep-POD after sequential ADP-ribosylation of EF-2 as described. To confirm comparable protein loading, GAPDH was again used as loading control. (E) Cell viability is decreased after treatment of cells with His6-DTA plus PA63. Cells were incubated with His6-DTA alone or in combination with PA63 for 24 h at 37 °C. Cell viability was determined by the MTS assay. Staurosporine (1 µM) served as positive control. Data are given as mean ± SD, n = 3. Significance was tested using One-Way ANOVA followed by Dunnett's multiple comparison test. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001 vs. untreated controls.  63 (1 µg/mL) or His 6 -NDPK-A only for 6 h at 37 • C, the cells were analyzed by immunoblotting. The uptake of His 6 -NDPK-A was detected with an anti-His-antibody. Hsp90 was detected as loading control and densitometric analysis performed using the ImageJ software. Signal intensity of His 6 -NDPK-A was quantified against Hsp90. (B) Cell viability is not reduced by treatment of cells with His 6 -NDPK-A plus PA 63 . After incubation of the cells with His 6 -NDPK-A plus PA 63 or His 6 -NDPK-A alone for 48 h at 37 • C, the cell viability was determined by the MTS assay. Data were normalized to untreated cells (control). DMSO served as positive control. Data are given as mean ± SD, n = 3. (C) Uptake of His 6 -DTA is mediated by PA 63 and is sensitive to BafA1. Cells were incubated with BafA1 (30 nM) for 30 min at 37 • C followed by 24 h incubation with His 6 -DTA (5 µg/mL) with or without PA 63 (1 µg/mL). Representative pictures are shown. Thereafter, cell lysates were analyzed by immunoblotting for the presence of His 6 -DTA with an anti-His-antibody. Detection of Hsp90 was used as loading control. (D) PA 63 mediates delivery of His 6 -DTA into the cytosol of HeLa cells. Substrate modification was determined via strep-POD after sequential ADP-ribosylation of EF-2 as described. To confirm comparable protein loading, GAPDH was again used as loading control. (E) Cell viability is decreased after treatment of cells with His 6 -DTA plus PA 63 . Cells were incubated with His 6 -DTA alone or in combination with PA 63 for 24 h at 37 • C. Cell viability was determined by the MTS assay. Staurosporine (1 µM) served as positive control. Data are given as mean ± SD, n = 3. Significance was tested using One-Way ANOVA followed by Dunnett's multiple comparison test. * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001 vs. untreated controls.

Treatment with His 6 -NDPK-A Plus PA 63 Decreased Migration of MDA-MB-231 Cells
Finally, it was investigated whether the PA 63 -mediated transport of the tumor metastasis suppressor protein His 6 -NDPK-A into the cytosol of MDA-MB-231 cells has an effect on the migration of these breast cancer cells. To this end, an In Vitro wound-healing assay was performed, where the cells were seeded into a dual chamber insert which was placed into a 24-well plate. When the cells were attached, the insert was removed thereby introducing the scratch. After removal of the detached cells by washing, His 6 -NDPK-A was applied in the presence or absence of PA 63 . In parallel, the same experiment was performed with His 6 -DTA instead of His 6 -NDPK-A. The migration of the MDA-MB-231-cells was monitored by microscopic analysis and photo-documentation. As shown in Figure 7A, there were less cells in the wound (scratch) area after 12 and 24 h when these cells were incubated with His 6 -NDPK-A or His 6 -DTA in the presence of PA 63 , suggesting a reduced migration of these cells. The representative pictures show a reduced closure of the scratch as soon as PA 63 was present in addition to the His-tagged proteins. After 12 h, the scratch remained more open when cells were treated with either His 6 -NDPK-A or His 6 -DTA. The automated determination of the scratch area after 24 h in correlation to the scratch area at the start of the experiment confirmed this result ( Figure 7B). There was a sustained significant reduction in migration of MDA-MB-231 cells after treatment with His 6 -NDPK-A plus PA 63 after 12 h for up to 24 h. For this particular experiment, treatment with His 6 -DTA plus PA 63 served as positive control. As expected, the migration of the treated cells was nearly abolished after 8 h and after 24 h cell-rounding was obvious. In line, a reduced cell viability occurred only in cells treated His 6 -DTA and PA 63 . In Figure 7D

Discussion
With the results presented here, we provided proof-of-concept that the transport component of anthrax toxin PA 63 serves for delivery of the tumor metastasis suppressor protein NDPK-A into human cells including breast cancer cells In Vitro. We used the established approach to fuse a common His 6 -tag to NDPK-A and purified and characterized this recombinant protein, as depicted in Figure 8. Delivery of His6-NDPK-A by PA63 was confirmed by a set of experiments using different approaches. However, there was also some non-specific binding of His6-NDPK-A to the cells in the absence of PA63. In the presence of the transporter, this binding was obviously increased. Moreover, it could be confirmed that the transport of His6-NDPK-A occurred specifically via the PA63 uptake mechanism as it was prevented when cells were pretreated with BafA1 to inhibit endosomal acidification.
One main advantage of this approach is the intrinsic T-domain of PA63 for cytosolic delivery of the cargo proteins, which is a bottle neck of most available cellular transport systems including cell-penetrating peptides. Such peptides and their fused cargo proteins are efficiently internalized into endosomal vesicles but have limited capacity to translocate their cargo into the cytosol where it finds its cellular target substrate, i.e., drug target, for modification. However, it is not possible to monitor the delivery of His6-NDPK-A into the cytosol by a direct Delivery of His 6 -NDPK-A by PA 63 was confirmed by a set of experiments using different approaches. However, there was also some non-specific binding of His 6 -NDPK-A to the cells in the absence of PA 63 . In the presence of the transporter, this binding was obviously increased. Moreover, it could be confirmed that the transport of His 6 -NDPK-A occurred specifically via the PA 63 uptake mechanism as it was prevented when cells were pretreated with BafA1 to inhibit endosomal acidification.
One main advantage of this approach is the intrinsic T-domain of PA 63 for cytosolic delivery of the cargo proteins, which is a bottle neck of most available cellular transport systems including cell-penetrating peptides. Such peptides and their fused cargo proteins are efficiently internalized into endosomal vesicles but have limited capacity to translocate their cargo into the cytosol where it finds its cellular target substrate, i.e., drug target, for modification. However, it is not possible to monitor the delivery of His 6 -NDPK-A into the cytosol by a direct approach. Therefore, we used the established His 6 -DTA as a model cargo for PA 63 -mediated transport into the cytosol in the same experiments in parallel to His 6 -NDPK-A. As the substrate for DTA, EF-2 is exclusively localized in the cytosol of cells and the DTA-catalyzed inactivation of EF-2 results in cell rounding and the simple microscopic analysis allows direct correlation of the cell rounding with the presence of enzymatically active His 6 -DTA in the cytosol. We took advantage of this approach to confirm that the breast cancer cell line MDA-MB-231 is sensitive for PA 63 -mediated protein delivery. Although these cells bound and internalized PA 63 -associated cargo proteins, they were less sensitive towards PA 63 then HeLa cells. One reason might be that HeLa cells express more of the anthrax toxin receptors (CMG2/TEM8) on their surface. The uptake of His 6 -DTA into the cytosol was clearly confirmed. However, although His 6 -NDPK-A is a different protein with a different structure and unfolding/refolding abilities, its transport as His-tagged protein via the same PA 63 -dependent mechanism from endosomes into the cytosol might be an overall comparable process as observed for His 6 -DTA.
To prove whether NDPK-A reached the cytosol of MDA-MB-231 cells as an active enzyme, similar functional analysis was performed as after transfection-based overexpression of NME1/NDPK-A described before [11].
To this end, the migration of these cells was monitored over time in a wound healing/scratch assay. When the amounts of cells in the scratch area were monitored at different time points after generating the scratch, there were less cells present after treatment with PA 63 plus His 6 -NDPK-A compared to cells treated with His 6 -NDPK-A alone or untreated cells. In contrast to His 6 -DTA uptake, increased delivery of His 6 -NDPK-A did not diminish the number of living cells, a cytotoxic effect of PA 63 plus His 6 -NDPKA in terms of cell viability or division was excluded. Therefore, the reduced number of cells in the scratch likely results from the well described and mechanistically at least partially understood reduction in motility of MDA-MB-231 in which NME1/NDPK-A expression has been restored [8][9][10][11].
This suggests that indeed functional active NDPK-A reaches the cytosol of those cells and mediated this effect. In conclusion, as the anti-migratory effect of PA 63 plus His 6 -NDPKA on MDA-MB-231 cells was apparently as effective as NME1/NDPK-A overexpression by conventional methods, our data provide proof-of-feasibility that the specific PA 63 -mediated introduction of a His 6 -tagged version of the tumor metastasis suppressor protein NME1/NDPK-A into the cytosol of cancer cells can be an attractive novel approach to reduce the migration and metastasis, e.g., of certain breast cancer types.
Future work must focus on the improvement of this system regarding efficiency and cell type-selectivity. Efficiency might be improved by fusing the N-terminal domain of LF (LF N ) [23][24][25]33] to NDPK-A instead of the His 6 -tag or to use alternative amino acid tags instead of histidine residues [27][28][29]. It will also be important to treat other breast cancer cell lines as well as primary human breast cancer cells, which might express larger amounts of anthrax toxin receptors than the MDA-MB-231 cells, with PA 63 plus His 6 -NDPKA or LF N -NDPK-A to investigate whether these cells show enhanced sensitivity towards this system.
Cell-type selectivity can be achieved by deleting the natural receptor binding capacity of PA 63 by single amino acid exchange in its B-domain (mPA 63 ) and fusing peptide ligands for receptors on the surface of certain cancer cells such as EGF or ZHER2, the ligand for HER2 receptor. This elegant approach was developed by the Collier group earlier and successfully applied in various studies In Vitro, ex vivo and In Vivo, without obvious immunological reactions caused by the PA-based delivery system [34][35][36][37][38]. Finally, the feasibility and efficacy of such approaches have to be validated in animal models for metastasis.

Cell Culture
Cells were cultured in medium containing 10% FCS, 1.5 g/L sodium bicarbonate, 2 mM L-glutamine, 1 mM sodium-pyruvate, 0.1 mM non-essential amino acids and 1% penicillin-streptomycin at 37 • C and 5% CO 2 . MEM was used for HeLa cells from DSMZ (CCL-2, Braunschweig, Germany) and DMEM for MDA-MB-231 cells from Cell Biolabs (HTB-26, San Diego, CA, USA). MDA-MB-231 cells were additionally supplemented with 0.25 µg/mL Fungizone ® antimycotic. If not mentioned otherwise, all experiments were performed in fully supplemented medium and cells were seeded one day prior to the experiment.
The expression vector encoding for His 6 -DTA (pET-15b_His 6 -DTA) was kindly provided by R.J. Collier (Boston, MA, USA) and transformed into E.coli BL21. Protein expression was induced at an optical density of 0.6-0.8 with 0.5 mM IPTG and performed for 4-5 h, 30 • C. Cells were harvested at 6.000× g for 30 min at room temperature, the pellet was resuspended in 10 mL lysis buffer (NaH 2 PO 4 (50 mM), NaCl (300 mM), imidazole (10 mM), pH 8) per liter main culture. After a freeze-thaw cycle lysis was performed via sonication, alternating 30 s bursts, 50% intensity with 30 s intervals on ice. The lysate was cleared with centrifugation at 15.000× g for 30 min at 4 • C, filtered through 0.45-and 0.2 µM syringe filters and purification was performed using the ÄKTA™ system. After Ni 2+ -affinity chromatography with elution of the protein using an increasing gradient of elution buffer (NaH 2 PO 4 (50 mM), NaCl (300 mM), imidazole (500 mM), pH 8), His 6 -DTA was reduced by DTT (10 mM) for 40 min at room temperature. Buffer was exchanged to PBS and concentrating was done with Vivaspin ® 20, MWCO 5 kDa. Total protein amount was quantified with SDS-PAGE at 6.8 mg/mL with a total yield of 6 mg His 6 -DTA per L main culture. The protein was stored at −80 • C.

Migration Assay (Wound-Healing Assay, Scratch Assay)
The 2-well culture-Insert system for self-insertion from Ibidi was used (#80209, Ibidi, Gräfelfing, Germany). Inserts were placed into wells of a 24-well plate and MDA-MB-231 cells were seeded (25,000 cells per insert-well). The surrounding well area was filled with culture medium. To induce the wound, the insert was removed with forceps and the cells were washed twice carefully with warm culture medium. The protein mixtures were added (250 µL/well) and the cells incubated at 37 • C. Three technical replicates were performed. Migration was monitored via image acquisition; bright field pictures were taken every 4 h. Each well was analyzed with three pictures within the middle of the well and wound/scratch area, respectively. The entire scratch area could be traced. Migration was monitored for up to 48 h with subsequent determination of the cell viability using the MTS assay. For quantitative analysis of the total wound area, the MRI tool was applied using the ImageJ software. Relative wound area was determined in relation to the initial scratch.

Experimental Reproducibility and Statistics
All experiments were independently performed at least three times. The figures show representative results. Immunoblots show sections of larger membrane parts for presentation but the protein bands were always detected on the same membrane, originally. Statistical analysis was performed using the GraphPad Prism software (8.3.1 (549), San Diego, CA, USA).