*2.3. DNA Binding of 1MEY# ZFP Is Influenced by Toxic Metal Ions* 2.3.1. Cd(II)

Previously, we have shown that Zn(II)-saturated 1MEY# could bind DNA by EMSA experiments. A clear difference in the affinity of the protein towards the 34 bp DNA probes with and without the specific 50 -GAGGCAGAA-30 sequence was observed [52]. Here, we studied the interaction of DNA with the Cd(II)-loaded 1MEY# ZFP obtained from the apo-protein. EMSA experiments revealed a single, well defined shifted DNA band with the nonspecific DNA probe. By fitting the quantified band intensities, a stability constant could be determined as log*β* <sup>0</sup> = 6.04 ± 0.02 (Figure 5a,c). This corresponds to ~0.2 log units weaker binding compared to the interaction of the Zn(II)-loaded protein with nonspecific DNA. A more significant ~0.6 log unit decrease was found in the affinity of the Cd(II)-loaded 1MEY#, with log*β* <sup>0</sup> = 7.62 ± 0.04 (Figure 5b,d), compared to that of the Zn(II)-loaded ZFP (log*β* 0 = 8.20). Nevertheless, the above listed stability constants are still substantially high values. It has to be mentioned that the titration curves for nonspecific DNA binding show a slight sigmoidal pattern instead of the saturation curve expected from the simple binding scheme. This, however, might be attributed to the ambiguity of the gel staining when the amount of bound DNA is too small. It was also visible that the DNA binding of the Cd(II)-loaded 1MEY# resulted in similar changes in the CD spectra of the system around 190 nm, as in case of the Zn(II)-loaded protein. However, the rate of the increase in ellipticity for the Cd(II)-1MEY# was approximately half of that of Zn(II)-1MEY#, which may indicate that the interaction of the Cd(II)-loaded ZFP with DNA induces smaller structural rearrangements, i.e., the process is weaker (Figure 5g).

duces smaller structural rearrangements, i.e., the process is weaker (Figure 5g).

*2.3. DNA Binding of 1MEY# ZFP Is Influenced by Toxic Metal Ions*

Previously, we have shown that Zn(II)-saturated 1MEY# could bind DNA by EMSA experiments. A clear difference in the affinity of the protein towards the 34 bp DNA probes with and without the specific 5′-GAGGCAGAA-3′ sequence was observed [52]. Here, we studied the interaction of DNA with the Cd(II)-loaded 1MEY# ZFP obtained from the apo-protein. EMSA experiments revealed a single, well defined shifted DNA band with the nonspecific DNA probe. By fitting the quantified band intensities, a stability constant could be determined as log*β*' = 6.04 ± 0.02 (Figure 5a,c). This corresponds to ~0.2 log units weaker binding compared to the interaction of the Zn(II)-loaded protein with nonspecific DNA. A more significant ~0.6 log unit decrease was found in the affinity of the Cd(II)-loaded 1MEY#, with log*β*' = 7.62 ± 0.04 (Figure 5b,d), compared to that of the Zn(II)-loaded ZFP (log*β*' = 8.20). Nevertheless, the above listed stability constants are still substantially high values. It has to be mentioned that the titration curves for nonspecific DNA binding show a slight sigmoidal pattern instead of the saturation curve expected from the simple binding scheme. This, however, might be attributed to the ambiguity of the gel staining when the amount of bound DNA is too small. It was also visible that the DNA binding of the Cd(II)-loaded 1MEY# resulted in similar changes in the CD spectra of the system around 190 nm, as in case of the Zn(II)-loaded protein. However, the rate of the increase in ellipticity for the Cd(II)-1MEY# was approximately half of that of Zn(II)- 1MEY#, which may indicate that the interaction of the Cd(II)-loaded ZFP with DNA in-

2.3.1. Cd(II)

**Figure 5.** Representative electrophoretic gel mobility shift assays of (**a**) nonspecific S0 DNA and (**b**) specific S1 DNA in the presence of increasing equivalents of Cd(II)-loaded 1MEY# zinc finger pro-**Figure 5.** Representative electrophoretic gel mobility shift assays of (**a**) nonspecific S0 DNA and (**b**) specific S1 DNA in the presence of increasing equivalents of Cd(II)-loaded 1MEY# zinc finger protein. *c*DNA = 0.88–1 µM. (**c**) Distribution of S0 DNA or (**d**) S1 DNA in the presence of increasing equivalents of Cd(II)-1MEY# ZFP. DNA fractions (separate points) were calculated based on the intensities of three independent electrophoretic gel mobility shift assays. Band intensity calculations were performed in ImageJ [68]. (**e**) Electrophoretic gel mobility shift assay of Zn(II)-1MEY# with specific S1 DNA in the presence of increasing equivalents of Cd(II). *c*DNA = 1 µM, *c*Zn(II)-1MEY# = 0.8 µM. (**f**) Distribution of S1-DNA among the free and protein-bound forms as calculated from the band intensities of the electrophoretic gel mobility shift assay image. (**g**) Comparison of the CD spectra of S1 DNA in the presence of 0.5 eq Zn(II)-1MEY# (black) and 0.5 eq Cd(II)-1MEY# (orange). The dashed line represents the CD spectrum calculated by summing the appropriate protein and DNA component spectra. All CD spectra were normalised to the intensity of the starting Zn(II)-loaded 1MEY# spectrum recorded at 18.8 µM protein concentration.

After it was proven that a fully Cd(II)-saturated ZFP can bind its DNA target, competition reactions were performed, where the Zn(II)-loaded 1MEY# ZFP in complex with the specific DNA probe was titrated with increasing amounts of Cd(II). During the process, no change could be seen in the DNA binding ability of the protein, which suggests that if mixed Zn2Cd11MEY# and Zn1Cd21MEY# complexes form, these can also bind DNA with a similar affinity to the Zn(II)-loaded protein (Figure 5e,f). On the other hand, previously we have shown that the interaction with specific DNA increases the stability of the Zn(II) complex of 1MEY# [52]. Therefore, it can be assumed that metal ion exchange is a minor process in this experiment.

Based on these findings, we could assume that the toxicity of cadmium in the living organism cannot be directly attributed to competition with high-stability Cys2His2 ZFPs. If the protein is in the Zn(II)-loaded form, a large excess of Cd(II) would be necessary to com-

pete with Zn(II) and it is unlikely to completely substitute Zn(II). The Cd(II)-loaded 1MEY# and the mixed metal complexes can also bind DNA, the only visible differences during our measurements were in their affinity, similar to the limited literature data [43,44,46]. Although the competition with the high stability Zn(II)-loaded ZFP is not significant, if Cd(II) meets with the apo-ZFP then it can form stable complexes due to its high affinity towards to protein. Regardless, it cannot be ruled out that such a Cd(II)-loaded high stability ZFP may still be able to perform its function, since once it is coordinated in Cys2His2 mode, the structure and function may differ only slightly compared to the Zn(II)-loaded protein. Furthermore, even if the structure of the protein differs, it does not necessarily mean that the DNA binding function is also completely vanished. In the case of the Tramtrack ZFP, where based on CD, the α-helix content of the protein was reduced significantly during Cd(II) binding, the complex could still recognise its target DNA but ~1 order of magnitude more weakly [46].

### 2.3.2. Hg(II)

Electromobility shift assay titrations revealed that the protein binding to its DNA target had no inhibitory effect on Hg(II) competition. Three equivalents of Hg(II) (one eq. per 1MEY# bs) could completely eliminate the DNA binding of the ZFP (Figure 6a,b). Thus, the effect of Hg(II) was unambiguous in the absence of other competing ligands. By using various buffer conditions (Cl− ions and DTT) this effect can be significantly reduced, yet according to most of the literature data, Hg(II) still could effectively inhibit the DNA binding of a ZFP [40,44,51]. *Inorganics* **2023**, *11*, x FOR PEER REVIEW 13 of 19

**Figure 6.** (**a**) Representative electrophoretic gel mobility shift assay of Zn(II)-1MEY# with specific S1 DNA in the presence of increasing equivalents of Hg(II). *c*DNA = 1 µM, *c*Zn(II)-1MEY# = 1 µM. (**b**) Distribution of S1-DNA based on electrophoretic gel mobility shift assay gel intensities. **Figure 6.** (**a**) Representative electrophoretic gel mobility shift assay of Zn(II)-1MEY# with specific S1 DNA in the presence of increasing equivalents of Hg(II). *c*DNA = 1 µM, *c*Zn(II)-1MEY# = 1 µM. (**b**) Distribution of S1-DNA based on electrophoretic gel mobility shift assay gel intensities.

### **3. Materials and Methods 3. Materials and Methods**

#### *3.1. Protein Expression and Purification 3.1. Protein Expression and Purification*

The protein expression and purification steps were described earlier [52]. Briefly, the 1MEY# protein (a consensus peptide-based 1MEY [69,70] derivative) was expressed in *E. coli* BL21 (DE3) cells from a pETM11 vector, with an N-terminal His-SUMO affinity tag. Ni(II) affinity purification was applied, then the N-terminal affinity and SUMO tag were cleaved specifically using the ULP1 protease [71]. Purification was followed by buffer exchange to 10 mM Cl⁻-free 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), pH = 7.40, using Amicon 3K 15 mL filters (Merck KGaA, Darmstadt, Germany) at 4000× *g* 8 × 30 min 15 °C and filtration through 0.22 µm, Ø =13 mm PES filters (Merck KGaA, Darmstadt, Germany). This process provided the Zn(II)-loaded ZFP; thus, for the Cd(II)- 1MEY# experiments, Zn(II) was removed from the protein by 0.5 mM EDTA (~25× excess over 1MEY#) and 0.2 mM TCEP (tris(2-carboxyethyl)phosphine) treatment for 10 min at 25 °C. EDTA was washed away during ultra-filtration (Amicon 3K 0.5 mL filters (Merck KGaA, Darmstadt, Germany), 14,000× *g* 5 × 5 min 15 °C) with 10 mM HEPES, 0.2 mM TCEP and 50 mM NaClO<sup>4</sup> (pH = 7.40) buffer. The TCEP reducing agent was included in The protein expression and purification steps were described earlier [52]. Briefly, the 1MEY# protein (a consensus peptide-based 1MEY [69,70] derivative) was expressed in *E. coli* BL21 (DE3) cells from a pETM11 vector, with an N-terminal His-SUMO affinity tag. Ni(II) affinity purification was applied, then the N-terminal affinity and SUMO tag were cleaved specifically using the ULP1 protease [71]. Purification was followed by buffer exchange to 10 mM Cl−-free 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), pH = 7.40, using Amicon 3K 15 mL filters (Merck KGaA, Darmstadt, Germany) at 4000× *g* 8 × 30 min 15 ◦C and filtration through 0.22 µm, Ø = 13 mm PES filters (Merck KGaA, Darmstadt, Germany). This process provided the Zn(II)-loaded ZFP; thus, for the Cd(II)- 1MEY# experiments, Zn(II) was removed from the protein by 0.5 mM EDTA (~25× excess over 1MEY#) and 0.2 mM TCEP (tris(2-carboxyethyl)phosphine) treatment for 10 min at 25 ◦C. EDTA was washed away during ultra-filtration (Amicon 3K 0.5 mL filters (Merck KGaA, Darmstadt, Germany), 14,000× *g* 5 × 5 min 15 ◦C) with 10 mM HEPES, 0.2 mM TCEP and 50 mM NaClO<sup>4</sup> (pH = 7.40) buffer. The TCEP reducing agent was included in

order to protect the free cysteines from potential oxidation. After the calculated EDTA content of the protein sample dropped to below 1 μM, 150 μM final concentration

additional buffer exchange was performed with 10 mM HEPES and 50 mM NaClO<sup>4</sup> (pH

Intact protein analysis was performed on an LTQ-Orbitrap Elite (Thermo Scientific, Ca, USA) mass spectrometer coupled with a TriVersa NanoMate (Advion, Ithaca, USCA) chip-based electrospray ion source. Measurements were carried out in positive mode at 120,000 resolution in 8.2 mM ammonium hydrogen carbonate buffer (pH ~7.8), as described previously [72]. Fitting of the ESI-MS data was performed by the Solver add in of

A J-1500 Jasco spectrophotometer was used during spectroscopic measurements under constant nitrogen flow in stepwise scanning mode over the range of 180–330 nm. Synchrotron radiation (SR) CD spectra were recorded at the CD1 beamline of the storage ring ASTRID at the Institute for Storage Ring Facilities (ISA), University of Aarhus, Denmark

Microsoft Excel based on statistical considerations [73,74].

between the 170 and 330 nm wavelength range [75,76].

= 7.40) buffer by ultra-filtration as described earlier.

*3.2. Mass Spectrometric Analysis of the Protein*

*3.3. CD Spectroscopy*

order to protect the free cysteines from potential oxidation. After the calculated EDTA content of the protein sample dropped to below 1 µM, 150 µM final concentration Cd(ClO4)<sup>2</sup> was added to a portion of the sample and incubated for 5 min at 25 ◦C. Then, additional buffer exchange was performed with 10 mM HEPES and 50 mM NaClO<sup>4</sup> (pH = 7.40) buffer by ultra-filtration as described earlier.

### *3.2. Mass Spectrometric Analysis of the Protein*

Intact protein analysis was performed on an LTQ-Orbitrap Elite (Thermo Scientific, Thousand Oaks, CA, USA) mass spectrometer coupled with a TriVersa NanoMate (Advion, Ithaca, NY, USA) chip-based electrospray ion source. Measurements were carried out in positive mode at 120,000 resolution in 8.2 mM ammonium hydrogen carbonate buffer (pH ~7.8), as described previously [72]. Fitting of the ESI-MS data was performed by the Solver add in of Microsoft Excel based on statistical considerations [73,74].

### *3.3. CD Spectroscopy*

A J-1500 Jasco spectrophotometer was used during spectroscopic measurements under constant nitrogen flow in stepwise scanning mode over the range of 180–330 nm. Synchrotron radiation (SR) CD spectra were recorded at the CD1 beamline of the storage ring ASTRID at the Institute for Storage Ring Facilities (ISA), University of Aarhus, Denmark between the 170 and 330 nm wavelength range [75,76].

All spectra were recorded with 1 nm steps and a dwell time of 2 s per step, using *l* = 0.2 mm cylindrical quartz cells (SUPRA-SIL, Hellma GmbH, Müllheim, Germany). Each sample containing 8–20 µM protein was prepared separately in 6.6–7.5 mM HEPES buffer (pH = 7.40) and was kept at room temperature for at least 5 min prior to measurement. Samples for CD measurements involving DNA contained 33–45 mM NaClO<sup>4</sup> as well. If CD data were fitted, the 215–260 nm wavelength range was selected and the PSEQUAD program was used for calculations [58].

### *3.4. Electrophoretic Mobility Shift Assay (EMSA)*

Electrophoretic mobility shift assays were performed using 34 bp DNA probes containing zero (later referred to as S0 DNA) or one (later referred to as S1 DNA) specific 1MEY# target sequence: 50–GAGGCAGAA–30 . The S0 DNA probe was obtained by the hybridisation of the Forward-S0: 50–CTAGTTTGCTGAACTGGGGTCACATAGATTAATA– 3 0 and Reverse-S0: –50 -TATTAATCTATGTGACCCCAGTTCAGCAAACTAG-30 oligonucleotides. The S1 DNA probe was obtained by the hybridisation of the Forward-S1: 50– GAATTCCTGCTGAGAGGCAGAAACATAGGGGTCG–30 and Reverse-S1: 50–CGACCCC-TATGTTTCTGCCTCTCAGCAGGAATTC–30 oligonucleotides (target sequence of 1MEY# is underlined). Oligonucleotides were obtained by solid phase synthesis (Invitrogen—Thermo Scientific, CA, USA). EMSA experiments were performed as described earlier in 10 mM HEPES, 150 mM NaClO<sup>4</sup> and 10 m/V% glycerol buffer (pH = 7.40) [77]. The FastRuler Ultra Low Range DNA Ladder (Thermo Scientific, CA, USA) served as a reference (marker) and the ImageJ program was used for the quantification of gel intensities [68]. Calculations were performed by the PSEQUAD program [58].

### *3.5. Fluorimetry*

The competition reactions of the Zn(II)-loaded 1MEY# were monitored using the FluoZin-3 fluorescent probe in a CLARIOstar Plus plate reader (BMG Labtech, Ortenberg, Germany). FluoZin-3 has an absorption maximum at 494 nm and exhibits strong fluorescence at 516 nm when binding Zn(II), with a conditional stability constant of log*β* 0 = 8.04 (pH = 7.40) [78]. In a typical measurement, 480–490 nm extinction and 510–520 nm emission filters were used during the titrations of 200 µL protein-FluoZin-3 sample in 96-well, polystyrene, non-binding, flat-bottom, black microplates (Greiner Bio-One, Kremsmünster, Austria) in 10 mM HEPES and 150 mM NaClO<sup>4</sup> buffer (pH = 7.40). An amount of 3 µL titrant (or buffer) was injected to the samples at each titration point by the two built-in

injectors of the CLARIOstar Plus plate reader. Each injection was followed by 30 s 150 rpm double orbital shaking of the plate and 5 min incubation at 25 ◦C to reach equilibrium. The titration process was automated using custom built scripts (Table S1). The FluoZin-3 concentration was determined spectrophotometrically (*λ*max = 491 nm, *ε*max = 71,143 M−<sup>1</sup> cm−<sup>1</sup> (pH = 7.40)). During each measurement, additional control samples were prepared containing only FluoZin-3 or FluoZin-3 and the equivalent amount of Zn(II) which can be released from the 1MEY# ZFP during competition reactions, in order to calculate correct relative fluorescence values (Figure S5). Fitting of the fluorometric data was performed by PSEQUAD [58] and by the Solver add in of Microsoft Excel.

### **4. Conclusions**

The interaction of Zn(II)-loaded Cys2His2 ZFPs with other metal ions can decrease, alter or destroy their DNA binding function. Therefore, a better understanding of these systems is essential. Several research works have aimed at studying the interactions between toxic metal ions and ZFPs, but the available data with model peptides and natural ZFPs cannot be directly compared. In a previous publication, we quantitatively characterised the Zn(II) and DNA binding of a ZFP consisting of three CP1-like model peptide subunits [52]. Based on CD measurements, we could prove that Cd(II) binding of 1MEY# resulted in an almost identical secondary structure to the Zn(II) complex. The protein could bind Cd(II) with the highest affinity determined so far for the Cys2Hi2 binding sites. Yet, the Zn(II) binding affinity of 1MEY# is even higher by 1–2 orders of magnitude. Similar tendencies were observed between the Zn(II) and Cd(II) binding of other investigated ZFs [33,38]. Thus, a large excess of Cd(II) would be necessary to compete with Zn(II), while the Cd(II) substituted ZFP can be reverted into Zn(II)-complex by adding two equivalents of Zn(II) per binding site, based on CD measurements. The metal exchange reaction occurred stepwise, with each ZF subunit behaving independently. Formation of Cd(II)-Cys3 or Cd(II)-Cs4 complexes could not be observed by ESI-MS and CD measurements. The protein bound three Cd(II) ions and no further spectral changes were visible, even when applying 120 equivalents of Cd(II). We could demonstrate that the Cd(II) complex was also capable of specific DNA binding by EMSA and CD measurements, although the affinity decreased by log*β* 0 = 0.6 units compared to the Zn(II) complex. The DNA binding ability of Zn(II)-loaded 1MEY# was not influenced even by ~100 eqs of Cd(II), suggesting that either no exchange occurred or the mixed complexes could recognise the DNA target as well.

Hg(II) behaved in a completely different manner. By using perchlorate salt, we could observe the affinity of Hg(II) towards the 1MEY# ZFP to be log*β* 0 Hg(II)-1MEY# bs ≥ 16.7. During the quantitative exchange of Zn(II) with Hg(II), the well-defined ββα secondary structure of 1MEY# collapsed. Excess of Hg(II) could also bind to 1MEY# with nanomolar affinity. Even a 13 Hg(II)-bound form was visible in ESI-MS, which could also be followed by CD spectroscopy through the charge transfer bands of Hg(II). In contrast, the Hg(II) affinity of the Ros87 ZFP investigated by Sivo et al. using HgCl<sup>2</sup> salt was three orders of magnitude weaker than the Zn(II) affinity of the same protein [51]. Hg(II) not only disrupted the structure of 1MEY#, but also destroyed the DNA binding of the ZFP. An investigation into the multi-metal binding sites in the Hg(II)-bound 1MEY# would be of interest.

**Supplementary Materials:** The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/inorganics11020064/s1, Section S1. Statistical considerations during ESI-MS measurements. Scheme S1. Schematic representation of the Zn(II)/Cd(II) exchange in the 1MEY# ZFP. In the transition states, Cd(II) can replace Zn(II) of any zinc finger subunit with equal probability. Scheme S2. Schematic representation of the Zn(II)/Hg(II) exchange in the 1MEY# ZFP. In the transition states, Hg(II) can replace Zn(II) of any zinc finger subunit with equal probability. Figure S1: Cartoon representation of the crystal structure of (a) the 1st ZF subunit of the 1MEY ZFP and (b) the whole 1MEY ZFP. ZFP: blue, Zn(II): grey sphere, cysteine thiols: yellow (PyMOL representation of 1MEY PDB [4]). (c) Alignment of the amino acid sequence of 1MEY# ZFP (constructed from 1MEY ZFP [4]) with the 26 amino acid-long consensus Cys2His2 model peptide CP1 and CP1 K/S mutant,

established and investigated by Berg et al. [5]. The identical amino acids of 1MEY# compared to CP1 are marked with green, while the ones that differ both compared to the CP1 and CP1 K/S mutant are marked with red. The amino acids differing only compared to CP1 are marked with light red. Figure S2: UV–Vis absorption spectra of 1MEY# in either Zn(II) (blue) or Cd(II) (orange) saturated form. *c*1MEY# = 13.5 µM in 10 mM HEPES and 50 mM NaClO<sup>4</sup> (pH 7.4); *l* = 1 cm. Figure S3: Measured (separate symbols) and calculated (full lines) relative fluorescence values of Zn(II)–FluoZin-3 systems in the presence of an increasing amount of (**a**) Cd(ClO<sup>4</sup> )2 ; and (**b**) Hg(ClO<sup>4</sup> )2 . Samples (200 µL) were loaded into the plate wells and titrated with 3 µL aliquots of the titrant at 25 ◦C. *c*FluoZin-3 = 3.98 µM, 10 mM HEPES and 150 mM NaClO<sup>4</sup> (pH 7.40). The calculations were performed by the PSEQUAD program [6]. Figure S4: Circular dichroism spectra of Zn(II)-loaded 1MEY# (full black line), Hg(II) loaded (red) and metal-free form using 5 eqs of EDTA per 1MEY# (1.7 eqs per binding site) (dashed black) are also presented. *c*1MEY# = 16.4 µM in 7.5 mM HEPES (pH = 7.4) buffer. CD1 beamline of the storage ring ASTRID, Aarhus *l* = 0.2 mm. Figure S5: Fluorometric titration procedure. Baseline fluorescence was determined by applying a 10-fold excess of EDTA over FluoZin-3. The maximal achievable fluorescence value was determined by applying 0.5 eq Zn(II) to FluoZin-3 ('Max'). A two-fold excess of FluoZin-3 was necessary to make sure 100% of Zn(II) is complexed. A sample containing an identical amount of Zn(II) to the 'Max' reference well was titrated with the titrant. The dilution effect during titration was determined by the injection of buffer (instead of the titrant) to the reference wells and to an additional sample well. Table S1: Automated titration script for CLARIOstar Plus plate reader.

**Author Contributions:** Conceptualisation, B.H. and B.G.; investigation, B.H. and É.H.-G.; data curation, B.H., É.H.-G. and B.G.; writing—original draft preparation, B.H. and B.G.; writing—review and editing, B.H., É.H.-G. and B.G.; funding acquisition, B.H., É.H.-G. and B.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** Supported by the ÚNKP-22-4-SZTE-491 New National Excellence Program of the Ministry for Culture and Innovation from the source of the National Research, Development and Innovation Fund and by the Hungarian National Research, Development and Innovation Office (GINOP-2.3.2- 15-2016-00038, GINOP-2.3.2-15-2016-00001, GINOP-2.3.2-15-2016-00020, 2019-2.1111-TÉT-2019-00089, and K\_16/120130) by the EU Horizon 2020 grant no. 739593. This research was partially funded by the CM\_SMP\_471080\_2021 Campus Mundi Student Mobility Traineeship from the Tempus Public Foundation. The support of SRCD measurements from the CALIPSOplus (EU Framework Programme for Research and Innovation HORIZON 2020, grant no. 730872) is also greatly acknowledged.

**Institutional Review Board Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors thank Peter Baker for the development and maintenance of the ELKH Cloud (https://science-cloud.hu/ (accessed on 27 December 2022)) for hosting the ProteinProspector search engine. The authors would like to thank Milan Kožíšek for providing the pETM11-SUMO3 plasmid.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

## *Article*
