Mouse β-Defensin 3, A Defensin Inhibitor of Both Its Endogenous and Exogenous Potassium Channels

The human defensins are recently discovered to inhibit potassium channels, which are classical targets of the animal toxins. Whether other vertebrate defensins are potassium channel inhibitors remains unknown. In this work, we reported that the mouse β-defensin 3 (mBD3) was a novel inhibitor of both endogenous and exogenous potassium channels. The structural analysis showed that mBD3 is the most identical to human Kv1.3 channel-sensitive human β-defensin 2 (hBD2). However, the pharmacological profiles indicated that the recombinant mBD3 (rmBD3) weakly inhibited the mouse and human Kv1.3 channels. Different from the pharmacological features of human β-defensins, mBD3 more selectively inhibited the mouse Kv1.6 and human KCNQ1/KCNE1 channels with IC50 values of 0.6 ± 0.4 μM and 1.2 ± 0.8 μM, respectively. The site directed mutagenesis experiments indicated that the extracellular pore region of mouse Kv1.6 channel was the interaction site of rmBD3. In addition, the minor effect on the channel conductance-voltage relationship curves implied that mBD3 might bind the extracellular transmembrane helices S1-S2 linker and/or S3-S4 linker of mouse Kv1.6 channel. Together, these findings not only revealed mBD3 as a novel inhibitor of both endogenous and exogenous potassium channels, but also provided a clue to investigate the role of mBD3-Kv1.6 channel interaction in the physiological and pathological field in the future.


Expression and Purification of rmBD3 Fusion Protein
The prokaryotic expression system was used to express the rmBD3. After being transformed into E. coli Rosetta (DE3) cells, the bacteria cells were cultured at 37 • C in Luria-Bertani (LB) medium with ampicillin (100 µg/mL). 1 mM IPTG was added to induce peptide expression at the temperature of 25 • C when the bacteria reached its logarithmic growth phase. The bacteria cells were harvested after 8 to 10 h post-induction and resuspended into chilled 20 mM imidazole buffer containing 20 mM Tris-HCl, 0.5 M NaCl, 10% glycerinum (pH = 7.9). Then they were cracked using ultrasonic bath, the supernatant from the lysate was loaded to a nickel affinity column. The purified rmBD3 fusion protein was dialyzed with Enterokinase buffer containing 25 mM Tris-HCl, 50 mM NaCl, and 2 mM CaCl 2 for 4 h and digested by Enterokinase (Sangon Biotech, China) at the temperature of 25 • C for at least 12 h but no more than 16 h. High performance liquid chromatograph (HPLC) on a C18 column (10 × 250 mm, 5 µm) (Elite-HPLC) was used to further purify and isolate the digested protein by using a linear gradient of 5% to 95% acetonitrile with 0.1% trifluoroacetic acid (TFA) in 60 min at a constant flow rate of 4 mL/min, and the absorbance was detected at 230 nm [15]. The molecular weight was confirmed by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) [16]. The confirmed protein was sub-packed by pierce bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific, Pittsburgh, PA, USA) and stored at −80 • C refrigerator.

Cell Culture and Transfection
Human embryonic kidney 293 (HEK293) cells were cultured Dulbecco's modified Eagle medium (Thermo Fisher Scientific, Pittsburgh, PA. USA) with 10% heat-inactivated fetal calf serum supplemented with penicillin (100 units/mL) and streptomycin (100 µg/mL) in a humidified 5% CO 2 incubator at 37 • C. Plasmids were transfected into HEK293 cells using the TurboFect in vitro Transfection Reagent (Thermo Fisher Scientific, Pittsburgh, PA, USA). Potassium currents were recorded after transfection for 1 to 3 days and positive cells were selected based on the presence of GFP fluorescence.

Electrophysiological Recordings and Data Analysis
Electrophysiological experiments were carried out at room temperature using the whole-cell recording mode by EPC10 patch clamp amplifier (Heka Elektronik, Lambrecht, Germany), which was controlled by Patchmaster software version 2 × 65 (Heka Elektronik, Holliston, MA, USA) [5,6,13]. For all the channels, HEK293 cells were bathed with mammalian Ringer's solution contained 5 mM KCl, 140 mM NaCl, 10 mM HEPES, 2 mM CaCl 2 , 1 mM MgCl 2 and 10 mM D-glucose (pH adjusted to 7.4 with NaOH). When rmBD3 was applied, 0.01% BSA was added to the Ringer's solution. A multichannel microperfusion system MPS-2 (INBIO Inc., Wuhan, China) was used to exchange the external recording bath solution. The internal pipette solution contains 140 mM KCl, 1 mM MgCl 2 , 1 mM EGTA, 1 mM Na 2 ATP and 5 mM HEPES (pH adjusted to 7.4 with NaOH). All channel currents were elicited by depolarizing voltage steps of 200 ms from the holding potential −80 mV to +50 mV. Currents were typically digitized at 20 kHz and filtered at 2.9 kHz (Bessel). Electrode resistances were 3.0-5.0 MΩ, and the series resistance was usually compensated by 85%. The currents through Kv channels [6] and KCNQ1/KCNE1 channel [17] were recorded according to previously published references.
The IC 50 values were obtained by fitting a modified Hill equation to the data with the formula: I rmBD3 /I control = 1/(1 + [rmBD3]/IC 50 ), where [rmBD3] is the concentration of rmBD3, I rmBD3 and I control are the peak currents in the presence and absence of rmBD3 at four different concentrations. The conductance of mouse Kv1.6 channel was calculated by the formula: G = I peak /(V − E Kv1.6 ), where I peak is the peak current, E Kv1.6 is the reversal potential of the Kv1.6 channel. The conductance-voltage (G-V) curves were generated by the measured peak currents and fitted with a V 50 -V equation using the formula: where G is the conductance of the channels, G max is the maximal channel conductance, V is the membrane voltage, V 50 is the voltage of half-maximal activation, and k is the slope factor. The data were presented as the means ± SE of at least three times repeated experiments. Sigmaplot 12.5 and GraphPad Prism 5 software were used for analysis.

Structural Analysis of mBD3 as Potential Potassium Channel Inhibitor
So far, only two vertebrate β-defensins hBD1 and hBD2 were reported to work as Kv1.3 inhibitors with IC 50 values of 11.8 µM and 22.0 pM, respectively [12,13]. Here, the Kv1.3 channel-sensitive hBD2 was used as the molecular template to search potential mouse β-defensin from the biological databases. As shown in Figure 1A, the mouse β-defensin mBD3 was obtained with 42.9% identity to hBD2. Besides the sequence similarity, the distribution of six cysteine amino acid residues was also identical. In order to better understand the structural feature of mBD3, its structure was modeled by the template of hBD2 (PDB code: 1FD4) ( Figure 1B). Structural comparison of mBD3 and hBD2 showed the similar distribution of the positively charged amino acid residues in the spatial structures, that is, most of them distributed around the C-terminal of mBD3 and hBD2. Since the positively charged amino acid residues often play a critical role in mediating animal toxins to bind potassium channel pore region with many negatively charged amino acids [6,18,19], these similar sequence and residue distribution between mBD3 and hBD2 suggested that mBD3 was a novel inhibitor of potassium channels.
Molecules 2018, 23, x FOR PEER REVIEW 4 of 10 sequence and residue distribution between mBD3 and hBD2 suggested that mBD3 was a novel inhibitor of potassium channels.

Expression, Purification and Identification of Recombinant mBD3
The prokaryotic expression method was used to express the recombinant mBD3 in this work. The target fragment was cloned into the expression vector pET-32a(+) to obtain a fusion protein containing an N-terminal His tag sequence. As shown in Figure 2A, the first nickel affinity purification showed that the fusion protein band was between 20 to 27 kDa, which was in line with the calculated molecular weight of HIS6-mBD3 with molecular weight of 21.7 kDa (Figure 2A, columns 1-3). Through the Enterokinase digestion for HIS6-mBD3 ( Figure 2A, columns 4 and 5), the rmBD3 was obtained through the high performance liquid chromatograph by manual collection at 26-28 min, which is adjacent to the elution peak of another protein ( Figure 2B). These two proteins were well validated by Tricine-SDS-PAGE analysis (Figure 2A, columns 6 and 7). Also, the MALDI-TOF-MS was further used to analyze rmBD3. It was found that the determined molecular weight was 4615.4 Da, which was corresponded to its calculated value of 4614.4 Da ( Figure 2C). All these data showed that the recombinant mBD3 was successfully expressed.

Expression, Purification and Identification of Recombinant mBD3
The prokaryotic expression method was used to express the recombinant mBD3 in this work. The target fragment was cloned into the expression vector pET-32a(+) to obtain a fusion protein containing an N-terminal His tag sequence. As shown in Figure 2A, the first nickel affinity purification showed that the fusion protein band was between 20 to 27 kDa, which was in line with the calculated molecular weight of HIS 6 -mBD3 with molecular weight of 21.7 kDa (Figure 2A, columns 1-3). Through the Enterokinase digestion for HIS 6 -mBD3 ( Figure 2A, columns 4 and 5), the rmBD3 was obtained through the high performance liquid chromatograph by manual collection at 26-28 min, which is adjacent to the elution peak of another protein ( Figure 2B). These two proteins were well validated by Tricine-SDS-PAGE analysis ( Figure 2A, columns 6 and 7). Also, the MALDI-TOF-MS was further used to analyze rmBD3. It was found that the determined molecular weight was 4615.4 Da, which was corresponded to its calculated value of 4614.4 Da ( Figure 2C). All these data showed that the recombinant mBD3 was successfully expressed.
Molecules 2018, 23, x FOR PEER REVIEW 4 of 10 sequence and residue distribution between mBD3 and hBD2 suggested that mBD3 was a novel inhibitor of potassium channels.

Expression, Purification and Identification of Recombinant mBD3
The prokaryotic expression method was used to express the recombinant mBD3 in this work. The target fragment was cloned into the expression vector pET-32a(+) to obtain a fusion protein containing an N-terminal His tag sequence. As shown in Figure 2A, the first nickel affinity purification showed that the fusion protein band was between 20 to 27 kDa, which was in line with the calculated molecular weight of HIS6-mBD3 with molecular weight of 21.7 kDa (Figure 2A, columns 1-3). Through the Enterokinase digestion for HIS6-mBD3 (Figure 2A, columns 4 and 5), the rmBD3 was obtained through the high performance liquid chromatograph by manual collection at 26-28 min, which is adjacent to the elution peak of another protein ( Figure 2B). These two proteins were well validated by Tricine-SDS-PAGE analysis (Figure 2A, columns 6 and 7). Also, the MALDI-TOF-MS was further used to analyze rmBD3. It was found that the determined molecular weight was 4615.4 Da, which was corresponded to its calculated value of 4614.4 Da ( Figure 2C). All these data showed that the recombinant mBD3 was successfully expressed.

Inhibition Effects of Potassium Channels by mBD3
In order to verify whether the mBD3 could be an inhibitor of potassium channels, we examined it on different potassium channels subtypes through the patch clamp technique according to our previously procedure [5,6,13]. All the potassium channels were expressed in HEK293 cells, and the effects of rmBD3 on channels currents were measured. When the potential interactions between rmBD3 and mouse potassium channels were investigated, it was found that 1 µM rmBD3 could inhibit 19.0 ± 2.2%, 4.9 ± 0.6%, 14.0 ± 3.7% of mouse potassium currents mediated by Kv1.1, Kv1.2, Kv1.3, respectively ( Figure 3A-C). These weak activities were different from previous human β-defensins [12,13], which prompted us to investigate whether other mouse potassium channels are possible targets of mBD3. It is interesting that we found 1 µM rmBD3 could inhibit 56.1 ± 4.2% of mouse Kv1.6 channel ( Figure 3D), which indicated rmBD3 was a novel inhibitor of mouse potassium channels.
Molecules 2018, 23, x FOR PEER REVIEW 6 of 10 channels, but strong activities on Kv1.6 and KCNQ1/KCNE1 channels, whose 31.2 ± 3.6% and 46.3 ± 4.2% currents were inhibited by 1 μM rmBD3, respectively ( Figure 3H-J,L). It was noticed that 1 μM rmBD3 could inhibit 26.1 ± 4.8% of KCNQ1 channel currents, which was much weaker than inhibition potency of rmBD3 on KCNQ1/KCNE1 channel ( Figure 3J,K). Such functional differences suggested that auxiliary protein KCNE1 subunit potentially affected the rmBD3 binding. Together, all these pharmacological data indicated that mBD3 was a novel both endogenous and exogenous inhibitor of potassium channels.

Discussion
The defensins are a large molecular resource, which are naturally produced by fungus, insects, invertebrates and vertebrate animals. So far, only several defensins were found to be the inhibitors of potassium channels in recent years, however, the functional roles of most defensins in binding potassium channels remains unknown, which further limits our understanding of defensins in physiological and pathological functions. For example, the binding of human defensins to human Kv1.3 channels in T cells could affect the cytokine secretion [11,13], which provides new evidence for defensin roles in the adaptive immunity. Therefore, the search of potassium channel-interacting defensins and its functional characterization undoubtedly remains an emerging area.
In this work, we reported the potassium channel-binding defensins from the second vertebrate animals, that is mouse β-defensin mBD3 since the discoveries of human defensins [11][12][13][14]. Although the mBD3 shared about 42.9% sequence identity with the Kv1.3 channel-selective human β-defensin hBD2 (Figure 1), the recombinant mBD3 showed very weak inhibition activity on both mouse and human Kv1.3 channels ( Figure 3C,G). Among the mouse potassium channels investigated, the mBD3 could selectively and dose-dependently inhibit the currents of mouse Kv1.6 channels ( Figure 3A-D and Figure 4A,B), and its IC 50 value was 0.6 ± 0.4 µM. Such findings first demonstrated that mouse defensin was an endogenous inhibitor of own potassium channels. Besides the inhibition of mBDs towards endogenous Kv1.6 potassium channels, rmBD3 displayed weaker inhibition on human Kv1.6 channel, and some strong inhibition on human KCNQ1/KCNE1 channel with an IC 50 value of 1.2 ± 0.8 µM ( Figure 3E-L and Figure 4C,D). Such findings also indicated that mBD3 was an exogenous inhibitor of human potassium channels. These distinct pharmacological profiles of mouse mBD3 from human β-defensins undoubtedly expanded our knowledge of vertebrate defensing-potassium channel interactions.
The human β-defensins hBD1 and hBD2 presented differential binding mechanisms when they interacted with human Kv1.3 channels [12][13][14]. Here, the mouse mBD3 might adapt hBD2-like mechanism to recognize the pore region of mouse Kv1.6 channel, which was demonstrated by the site directed mutagenesis ( Figure 5A-C). The gating effect of mBD3 was further supported by its minor effect on Kv1.6 channel activation ( Figure 5D), which was similar to hBD2 modifier, different from hBD1 blocker [12,13]. During the scorpion toxin-potassium channel interactions, the electrostatic interactions usually play a critical role in their recognition process since the scorpion toxins as the basic peptides block the channel pore region with many acidic residues. As shown in Figure 1, the mBD3 is also a basic peptide with about 10 basic residues, which likely plays an important role in blocking the negatively-charged pore region of Kv1.6 channel through the dominant electrostatic interactions. However, the differential pharmacological profiles among hBD1, hBD2 and mBD3 were likely caused by the differential distribution of their potential functional residues (Figure 1), which deserves in-depth investigation in the future.
To our knowledge, the mouse Kv1.6 channel is mainly expressed in the cerebellum and brain [21,22], and mBD3 is widely distributed in various tissues and organs, including lung/trachea and bowel [23]. As a novel endogenous inhibitor of Kv1.6 channel, mBD3 would play a potential role in Kv1.6 channel-related physiological and pathological activities, which remains to be answered in the future.
In addition, the human KCNQ1/KCNE1 channel, closely related to heart disease [24], was shortage of potent peptide blocker until now. Previously, we found that the scorpion venom could not effectively inhibit the KCNQ1/KCNE1 channel, and engineered MT2-2 peptide, an analog derived from a scorpion toxin BmKTX, was a weak blocker with IC 50 value of 1.5 µM [25]. Here, we found that mBD3 could inhibit the human KCNQ1/KCNE1 channel with IC 50 value of 1.2 ± 0.8 µM (Figure 4C,D). Two such peptides with these distinct structural features will help in discovering a potent blocker of the human KCNQ1/KCNE1 channel, as a potential therapeutic agent. Also, our work would stimulate the discoveries of more potassium channel-acting defensins from mouse and other vertebrate species, and enrich our understanding of defensin functions in the physiological and pathological activities.