The Antibody Targeting the E314 Peptide of Human Kv1.3 Pore Region Serves as a Novel, Potent and Specific Channel Blocker

Selective blockade of Kv1.3 channels in effector memory T (TEM) cells was validated to ameliorate autoimmune or autoimmune-associated diseases. We generated the antibody directed against one peptide of human Kv1.3 (hKv1.3) extracellular loop as a novel and possible Kv1.3 blocker. One peptide of hKv1.3 extracellular loop E3 containing 14 amino acids (E314) was chosen as an antigenic determinant to generate the E314 antibody. The E314 antibody specifically recognized 63.8KD protein stably expressed in hKv1.3-HEK 293 cell lines, whereas it did not recognize or cross-react to human Kv1.1(hKv1.1), Kv1.2(hKv1.2), Kv1.4(hKv1.4), Kv1.5(hKv1.5), KCa3.1(hKCa3.1), HERG, hKCNQ1/hKCNE1, Nav1.5 and Cav1.2 proteins stably expressed in HEK 293 cell lines or in human atrial or ventricular myocytes by Western blotting analysis and immunostaining detection. By the technique of whole-cell patch clamp, the E314 antibody was shown to have a directly inhibitory effect on hKv1.3 currents expressed in HEK 293 or Jurkat T cells and the inhibition showed a concentration-dependence. However, it exerted no significant difference on hKv1.1, hKv1.2, hKv1.4, hKv1.5, hKCa3.1, HERG, hKCNQ1/hKCNE1, L-type Ca2+ or voltage-gated Na+ currents. The present study demonstrates that the antibody targeting the E314 peptide of hKv1.3 pore region could be a novel, potent and specific hKv1.3 blocker without affecting a variety of closely related Kv1 channels, KCa3.1 channels and functional cardiac ion channels underlying central nervous systerm (CNS) disorders or drug-acquired arrhythmias, which is required as a safe clinic-promising channel blocker.

Antibodies have the characteristics of high affinity and specificity. We herein generated the antibody directed against one peptide of human Kv1.3 extracellular loop as a novel and specific Kv1.3 blocker.

The E314 antibody generation
The E314 peptide containing 14 amino acids is located at the external end of hKv1.3 pore region. The amino acid sequence is shown as follows: Glu-Ala-Asp-Asp-Pro-Thr-Ser-Gly-Phe-Ser-Ser-Ile-Pro-Asp (China patent application number:201110044416.X Fig. 1A). By immunizing rabbits with the hapten, we generated the polyclonal antibody against the hKv1.3 E314 peptide with a high titre. After three immunizations, the antibody titre in serum was markedly boosted and reached a high and stable level at the termination of the immunization (Fig. 1B).

The E314 antibody inhibits human Kv1.3 currents stably expressed in HEK 293 cells
Using the whole-cell patch clamp technique, we tested the ability of the E314 antibody inhibiting hKv1.3 currents stably expressed in HEK 293 cells. The hKv1.3 currents were gradually inhibited after addition of the 300 nM E314 antibody to the external solution and the inhibition reached about 55% steadystate level in 10-15 minutes whcih was not reversible by washout (e.g., Figure S1 for the first supporting information figure). To obtain potent binding between the antigen and the antibody, cells stably expressing hKv1.3 channels were preincubated with the peptide-specific polyclonal antibody for two hours at 36uC and then superfused to wash off unbound antibodies, as undergone in immunoexperiments. I Kv1.3 was elicited by the 300-ms voltage steps from 280 mV to between 260 and +60 mV (as shown in the inset) in HEK 293 cells stably expressing hKv1.3 channel and I Kv1.3 was absent in raw HEK 293 cells (Fig. 3A). The E314 antibody with a concentration of 300, 150,75, or 37.5 nM significantly decreased hKv1.3 current densities at test potentials from 230 to +60 mV and the inhibition was stronger at more positive potentials. The inhibition showed a concentration-dependence ( Fig. 3B). At the depolarizing pulse +50 mV, the E314 antibody with concentrations ranging from 37.5 nM to 300 nM inhibited human Kv1.3 current densities respectively by 66%, 84%, 88% or 94% (0.1095160.0165 nA/pF, 0.0528560.01825 nA/pF, 0.0384860.01049 nA/pF, 0.019146 0.0043 nA/pF, vs 0.3209460.06573 nA/pF,P,0.001 vs control) (Fig. 3C).
To verify that the E314 antibody does bind to the external end of hKv1.3 pore region where the E314 peptide was generated and that the inhibiting effect is attributed to the binding of the E314 antibody to the hKv1.3 channel, we recorded hKv1.3 currents in the presence of the 300 nM E314 antibody that was preincubated with an excess of the E314 peptide. Supposed that supression was due to binding of the antibody to the peptide in external pore region, the inhibition should be prevented by preincubation with the peptide. As shown in Figure 3B and 3C, the inhibiting effect of the E314 antibody on I Kv1.3 was abolished after preincubation with the peptide, which indicated that the inhibition was due to specific binding of the E314 antibody to the E314 peptide around hKv1.3 pore region.
Voltage dependence of hKv1.3 channel activation (I/Imax) was determined by normalizing I Kv1.3 in the absence and presence of the 300 nM E314 antibody. Data were fitted to a Boltzmann distribution to obtain the half-activation voltage (V 0.5 ) and the Figure 1. The E314 peptide selection and the E314 antibody generation. (A) Six-membrane spanning (S1-S6) of hKv1.3 channel a subunit and pore region between S5 and S6 was depicted by hydrophilicity analysis of its constituent amino acid aligment. The E314 peptide located at pore region was selected according to amino acid antigenic index. (B) The E314 antibody titre was assayed by enzymelinked immunosorbent assay (ELISA). The E314 antibody wth a high titre was generated after 5 immunizations. doi:10.1371/journal.pone.0036379.g001 slope factor (S). The V 0.5 of I Kv1.3 activation conductance was positively shifted by 10.2 mV (from 8.560.8 mV of control to 18.761.0 mV of the E314 antibody, n = 6, P,0.01) by the 300 nM E314 antibody, and the slope factor was slightly decreased (17.860.2 mV for control, 15.860.9 mV for the E314 antibody, P.0.05) (Fig. 3D). The E314 antibody affects the activation gating of hKv1.3 channels.

The E314 antibody inhibits human Kv1.3 currents in Jurkat T cells
To further study the ability of the E314 antibody inhibiting hKv1.3 currents, we also tested the effect of the anibody on human leukemia T cell line, Jurkat E6-1 cells. I Kv1.3 expressed in Jurkat T cells preincubated with the E314 antibody for two hours at 36uC was recorded with the voltage protocol as described previously in the absence and presence of the 300 nM E314 antibody ( Fig. 4A and B). The 300 nM E314 antibody significantly decreased hKv1.3 current densities at test potentials from 230 to +60 mV and the inhibition was stronger at more positive potentials (Fig. 4C). At the depolarizing pulse +50 mV, the 300 nM E314 antibody inhibited hKv1.3 current densities by 90% (3.4955260.89790 nA/pF vs 34.5790862.21566 pA/ pF,P,0.001) (Fig. 4D).  The E314 antibody has no significant effect on I Kv1.1 , I Kv1.2 , I Kv1.4 , I Kv1.5 , I KCa3.1 , I HERG , I hKCNQ1/hKCNE1, I CaL or I Na To further verify the specificity of the E314 antibody, we examined the effect of E314 antibody on other closely related K v 1family channels, functional cardiac ion channels and KCa3.1 channel which is another important potassium channel expressed in T lymphocytes. Human atrial myocytes were used to observe the effect of E314 on I Kv1. 5 (Fig. 5, 6  and 7).

Discussion
Autoimmune diseases or autoimmune-associated diseases afflict millions of people in the world. Overreaction of immune activities plays a pivotal role in the pathogenesis of these diseases [11]. Thus immunosuppression therapy is of broad use in the management of autoimmune diseases [20,38]. However pleiotropic actions of immunosuppressants clinically available such as methylprednisolone limit therapeutic values [39,40]. For patients, more specific immunosuppression would help to ameliorate disease with less adverse or side effects [24,41]. In recent years, selective blockade of Kv1.3 in T EM cells has exhibited the potential of specific inhibition of T lymphocyte subsets and leaves protective immunity unharmed, which encourages increasing efforts to explore selective or specific Kv1.3 blockers [36,[41][42][43]. In this study, we presented a novel, potent and specific Kv1.3 blocker.
Unlike the heteromultimeric Kv1.3 channel expressed in neurons, Kv1.3-containing homotetramers in lymphocytes comprise 4 identical subunits [44,45]. We expressed the identical subunits in a stable HEK 293 cell lines, which represents Kv1.3 channels expressed in lymphocytes. The antibody designed targeting one subunit would be possible to exert its effect on the Kv1.3 channel. By computing the antigenic index and hydrophilicity of hKv1.3 constituent amino acid aligment, we selected an antigenic peptide that is located at the external end of the pore region which has been successfully used to generate polyclonal antibodies against several ion channels [46,47]. The peptide location was indicated by immunostaining of the E314 antibody binding. The pore region is the way mediating K + efflux and determines ion permeability [48,49]. The E314 antibody blocks the pore region from the external end with a large interacting surface, which utilizes a ''cork in the bottle'' strategy exemplified by peptide toxins, covering and plugging the external end [24,42]. The external block by the E314 antibody yields the K + impermeability. In addition, the external block is stable, in contrast with the internal block by small molecules which is greatly affected by gating-associated conformational changes and statedependent [37,50]. The external block gives the E314 antibody more potency. In this study, the E314 antibody in nanomolar concentrations showed a strong inhibition, comparable to some selective kv1.3 blockers [38,51,52].
In clinics, a variety of drugs including immunosuppressants can increase the risk of drug-acquired arrhythmias due to the impact on functional cardiac ion channels [53][54][55][56]. It is documented that drug binding sites among Kv1.3, Kv1.5, HERG and hKCNQ1/ hKCNE1 channels are conservative [21,36,43,57]. The conservation hinders some Kv1.3-blocking candidates from being developed into good drugs. Thus, to prevent the occurrence of drug-acquired arrhythmias, the ICH E14 and S7B guidance issued by FDA calls for the assessment of the potential of any drug to delay cardiac repolarisation [58][59][60]. Besides, loss function of closely related Kv1 channels, such as Kv1.1 or Kv1.2 is able to result in CNS disorders [61,62]. By blast, we found that there is a maximal homology between the hKv1.3-E314 peptide and the corresponding peptides of hKv1.1, hKv1.2, hKv1.4 or hKv1.5. In addition KCa3.1 channels in T lymphocytes might be of importance for immune-mediated side-effects [63]. Recently the tandem of pore domains in a weak inwardly rectifying potassium channel-related acid-sensitive potassium channels (TASK1-3) have been found to be expressed in the nervous system, T lymphocytes and heart and play important role in cardiac repolarization, autoimmune inflammation, cancer development and CNS disorders [64][65][66]. Therefore, this study was necessary to be focused on the effect of the E314 antibody as a potential candidate for specific Kv1.3 blockers on functional cardiac ion channels, closely related Kv1 channels, KCa3.1 channels and TASK1-3 channels.
Immunoexperiment results indicate that the E314 antibody can specifically recognize hKv1.3 protein, whereas it is not able to recognize or cross-react to hKv1.1, hKv1.2, hKv1.4, hKv1.5, hKCa3.1, HERG, hKCNQ1/hKCNE1 or human Na v 1.5, Ca v 1.2 proteins. By the patch clamp technique, we demonstrated that the E314 antibody with a high concentration exhibits no significant effect on these closely related K v 1 channels, KCa3.1 channels or functional cardiac ion channels. By blast analysis, we found that there is no homology between Kv1.3 and TASK1-3 channels. All the results indicate that the E314 antibody is able to function as a novel specific hKv1.3 blocker without worries about its potential proarrhythmias, immune-mediated side-effects or CNS disorders, which is required as a safe clinic-promising channel blocker.
Based on the E314 antibody generation regimen, one monoclonal antibody with more affinity and specificity or a vaccine targeting hKv1.3 E314 peptide, which can serve as a novel hKv1.3 blocker to inhibit autoreactive T lymphocyte activities, would probably be developed as a novel drug for the treatment of autoimmune diseases or autoimmune-associated diseases. The attractiveness arises from the excellent track record of several monoantibodies and a vaccine, such as infliximab [67], adalimumab [68], natalizumab [69], Anti-IL-17A vaccine [70], which is typical of targeting biological therapy, gradually superior to traditional immunosuppressants.

Ethics statement
In this study, the procedure of obtaining human atrial and ventricle specimens from patients receiving cardiac surgery conforms to the principles outlined in the Declaration of Helsinki. The study was approved by the Ethics Committee of Tongji Medical College of Huazhong University of Science and Technology and patients provided written informed consent.
In this study, the animal use and care protocol was approved by the Ethical Committee on Animal Experimentation of Tongji Medical College, Huazhong University of Science and Technology (Approval ID: 00009678). All surgery was performed under sodium pentobarbital anesthesia, and every effort was made to minimize suffering.

Antibody generation
One extracellular peptide located at E3 loop between hKv1.3 spanning S5 and S6 was selected as an antigenic determinant according to its constituent amino acid antigenic index calculated by the software DNAStar and the corresponding amino acid alignment was synthesized by PSSM8 Peptide Synthesizer (Shimadzu, Japan). Adult male New Zealand white rabbits from the Center of Experimental Animals (Tongji Medical College, Huazhong University of Science and Technology, China) were fortnightly immunized by subcutaneous injection of a mixture of the synthetic peptide conjugated with bovine serum albumin (BSA) and complete or incomplete Freund's adjuvant. Rabbits receiving sham immunization with a mixture of physiological saline conjugated with complete or incomplete Freund's adjuvant were as a control. Serum was collected before each immunization and the rabbits were sacrificed on 63 d of the experiment. Harvested serum were screened by enzymelinked immunosorbent assay (ELISA) and purified on a protein A column (UNOsphere SUPrA Affinity Cartridge, Bio-Rad, US). The E314 antibody was additionally purified on an peptide-affinity column (GL Biochem Ltd, Shanghai, China).

Cell culture and establishment of cell lines
HEK 293 cells were purchased from the American Type Culture Collection (ATCC Manassas, VA, USA) and grown in Dulbecco's modified Eagle medium (DMEM, Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad, CA) in 5% CO 2 and 95% air at 37uC.

Human atrial myocyte isolation
Atrial myocytes were isolated from specimens of human right atrial appendage obtained from patients receiving cardiac surgery aging from 25 to 70 years old (male 6, female 5). After excision, the samples were quickly immersed in oxygenated, Ca 2+ -free cardioplegic solution for transport to the laboratory. One single atrial myocyte was enzymatically dissociated under the sterile circumstance. Briefly, the atrial tissue was minced and the pieces were The isolated atrial myocytes were kept in a High-K + storage solution at room temperature for at least 1 hour until use.

Immunofluorescence
HEK 293 cells or human atrial myocytes were stained by immunofluorescence for the detection of the binding of the E314 antibody to plasma membranes. After adherent cells were fixed and blocked with 10% BSA and 1% normal donkey serum, these cells were incubated with the E314 antibody diluted at 1:200 at 4uC overnight and then with FITC-labelled donkey anti-rabbit secondary antibodies (Chemicon International) at room temperature for 2 h away from light. The cells were observed in a Olympus FluoView TM FV1000 (Olympus, Japan) laser-scanning confocal microscope. In control experiments, the primary antibody were preincubated with an excess of E314 antigenic peptide as a negative control.

Western blotting
To confirm the E314 antibody specific recognition, Westernblotting analysis was performed. Human atrial or ventricular specimens and HEK 293 cells were lysed and proteins were extracted. Proteins were analyzed by electrophoresis in 7.5% (w/v) polyacrylamide gel containing 0.1% (w/v) sodium dodecylsulfate (SDS) followed by blotting to a nitrocellulose membrane. After blocking, the membrane was incubated with the E314 antibody diluted at 1:1000 overnight at 4uC and then incubated with horseradish peroxidase-conjugated anti-rabbit IgG antibody 2 h at room temperature. The blot was visualized by a chemiluminescence's method (ECL Western blotting detection system). In control experiments, the primary antibody were preincubated with an excess of E314 antigenic peptide as a negative control.

Patch-clamp recording
HEK 293 cell lines, Jurkat T cells or human atrial myocytes were incubated at least 2 h with the E314 antibody at 36uC for total binding of the E314 antibody. Borosilicate glass electrodes (1.2-mm OD) were constructed using a Brown-Flamming puller (model P-97, Sutter Instrument Co, Novato, CA, USA) and had tip resistances of 2-3 MV when filled with pipette solution for I Kv1.3 , I Kv1.1 , I Kv1.2 , I Kv1.4 , I Kv1.5 , I KCa3.1 , I HERG , I hKCNQ1/hKCNE1 or I CaL recording and 0.5-1 MV for I Na recording. Membrane currents were recorded using an Axopatch200B amplifier and Clampex software (Molecular Devices, USA). For I Kv1.3, I Kv1.1 , I Kv1.2 , I Kv1.4 , I Kv1.5 , I KCa3.1 , I HERG or I Na recording, whole -cell patch clamp was used and perforated patch configuration was performed for I hKCNQ1/hKCNE1 or I CaL recording with additional 0.042 mg/ml Escin (Sigma-Aldrich) in the pipette solution. Electrical signals were low-pass filtered at 5 kHz and stored on the hard disk of computer for off-line analysis using clamfit or sigmaplot.

Statistical Analysis
Data were expressed as mean6SEM. Statistical analysis was performed using SPSS 12.0 software. Differences between groups were analyzed by t test or ANOVA. P value less than 0.05 was considered statistically significant.