The Effects of Shaker β-Subunits on the Human Lymphocyte K+ Channel Kv1.3*

The activation of T-lymphocytes is dependent upon, and accompanied by, an increase in voltage-gated K+ conductance. Kv1.3, a Shaker family K+ channel protein, appears to play an essential role in the activation of peripheral human T cells. Although Kv1.3-mediated K+ currents increase markedly during the activation process in mice, and to a lesser degree in humans, Kv1.3 mRNA levels in these organisms do not, indicating post-transcriptional regulation. In other tissues Shaker K+ channel proteins physically associate with cytoplasmic β-subunits (Kvβ1–3). Recently it has been shown that Kvβ1 and Kvβ2 are expressed in mouse T cells and that they are up-regulated during mitogen-stimulated activation. In this study, we show that the human Kvβ subunits substantially increase K+ current amplitudes when coexpressed with their Kv1.3 counterpart, and that unlike in mouse, protein levels of human Kvβ2 remain constant upon activation. Differences in Kvβ2 expression between mice and humans may explain the differential K+ conductance increases which accompany T-cell proliferation in these organisms.

Voltage-gated K ϩ channels play an important role in the propagation of electrical signals in the nervous system of higher organisms (1). A large number of voltage-gated K ϩ channel proteins are expressed throughout the mammalian nervous system. These proteins are encoded by a large number of genes which fall into various families or subfamilies of homology. The mammalian Shaker family of K ϩ channels contains at least seven different genes, Kv1.1-Kv1.7 (2), which form functional homo-and heterotetrameric channel complexes (3,4). Furthermore, it has been shown that in the mammalian nervous system the channelforming Shaker proteins are usually complexed with cytoplasmic Kv␤ subunits (Kv␤1-Kv␤3) (5).
Coexpression studies, utilizing mRNA injection and voltageclamp analysis of Xenopus oocytes, have shown that the major brain Kv␤ subunit (6), Kv␤2, and at least one splice form of Kv␤1 (Kv␤1a) are able to alter the inactivation properties of a number of neuronally expressed Kv1 channel proteins (7,8). In contrast, neither of these two Kv␤ subunits significantly alters the inactivation properties of Kv1.3, a K ϩ channel that is sparingly expressed within the nervous system (9). Perhaps more importantly, Kv␤ subunits are also able to increase K ϩ current amplitudes of neuronal Kv1 K ϩ channels in the oocyte plasma membrane (8). Furthermore, coexpression of several neuronal Kv1 channels with Kv␤2 in mammalian cell lines has shown that Kv␤2 increases the number of these Kv1 proteins reaching the membrane surface (10). Taken together, these studies indicate that the Kv␤ subunits are likely to increase the surface expression of functional, neuronal K ϩ channels.
In addition to their role in the nervous system, K ϩ channel proteins are expressed in other cell types, where they may help determine membrane potential and maintain osmotic equilibrium. What specific physiological roles might these K ϩ channel proteins play outside the nervous system? They appear to play an essential role in the stimulation and maintenance of cellular proliferation of T cells (11), B cells (12), macrophages (13), and brown adipocytes (14). In T-lymphocytes, the role has been extensively investigated: mitogens cause an immediate shift in K ϩ conductance (15,16); activated T cells show substantially greater K ϩ conductances than quiescent cells (11,17); activation is attenuated by membrane depolarization (18); and pharmacological agents that inhibit K ϩ channel conductances block T-cell proliferation (19). In human peripheral T-lymphocytes, the Kv1.3 K ϩ channel plays a critical role in mediating the K ϩ current increase, which accompanies proliferation (20), and is also likely to be the site of blockade, which inhibits this event (21). Interestingly, Kv1.3 mRNA does not appear to be upregulated during proliferation, indicating that the Kv1.3 gene product is post-transcriptionally regulated (22,23).
Recent studies have shown that Kv␤2, and Kv␤1 to a lesser extent, are both expressed in mouse T-lymphocytes (24). Moreover, the murine Kv␤ mRNA and protein levels are markedly increased upon interleukin-2 stimulation. If the Kv␤ subunits were able to increase the surface expression of Kv1.3 channels, then up-regulation of Kv␤ subunits would result in greater K ϩ channel surface expression and therefore greater K ϩ conductance during the proliferation of T-lymphocytes and perhaps other cell types (11)(12)(13)(14) The extent of Kv␤ subunit up-regulation in response to T-cell activation could therefore account for the extent to which Kv1.3-mediated K ϩ current is elevated in different organisms. Utilizing the Xenopus oocyte expression system, we report the effects of coexpression of the human Kv␤1a and Kv␤2 subunits on the expressed current levels of Kv1.3 channels.

EXPERIMENTAL PROCEDURES
In Vitro Transcription and Oocyte Injection-The Kv␤1 and Kv␤2 cDNAs were subcloned into a vector containing 40 adenosine residues downstream of the 3Ј cloning site and linearized with NotI. The ␤1 cDNA had 10 nucleotides 3Ј of the stop codon. The human Kv1.3 gene was transcribed from linearized template. For more efficient expression in oocytes (8), the 5Ј untranslated sequences immediately upstream of the ATG start codon of Kv␤2 and Kv1.3 were changed to CGCCGCCAAG. Oocytes were injected with 50 nl of cRNA at varying dilutions. All electrophysiology experiments were carried out at approximately 20°C.
Microinjection and Electrophysiology-Ovaries of specimens of Xenopus laevis were surgically removed, and individual oocytes were dissected away in a saline solution (ND96) containing 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl 2 , 2 mM MgCl 2 , and 5 mM HEPES at pH 7.4 with NaOH. Stage V and VI oocytes were treated for 2 h with collagenase (1 mg/ml) in Ca 2ϩ -free ND96 to discard follicle cells. cRNA solutions of 50 nl were injected into oocytes using the Nanoject injector apparatus (Drummond Scientific, Broomall, PA). Measurement of ionic currents in Xenopus oocytes was performed using standard two-electrode voltage clamp techniques. Recordings were obtained with the GeneClamp500 voltage clamp and the PCLAMP software package (Axon Instruments, Foster City, CA). Electrodes of 0.4 -1.0 megohms were filled with 1 M KCl. Recordings were conducted in ND96 solution. Data were leak subtracted using hyperpolarizing P/4 subtraction pulses from a holding potential of Ϫ80 mV. The interpulse interval was 25 s. K ϩ conductances were calculated using a reversal potential of Ϫ80 mV.
Cell Culture-Peripheral blood mononuclear cells were isolated by Ficoll Gradient and then added to media containing 10% fetal calf serum, 100 units/ml penicillin, 100 mg/ml streptomycin, and 100 units/ml interleukin-2. One-half of the cells were frozen at Ϫ80°C immediately after separation and thawed into media. The other cells were activated with phytohemagglutinin at 5 g/ml at 37°C. Cells were activated for 72 h.
Western Blot Analysis-Human T-lymphocytes were treated with or without phytohemagglutinin as described above. These cells were then homogenized in RIPA buffer (25 mM Tris, pH 7.4, 150 mM KCl, 5 mM EDTA, 1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS). Proteins were size-fractionated on a 9% SDS-polyacrylamide gel and then transferred to Nitrocellulose membrane. The membranes were blocked in 100 mM phosphate-buffered saline containing 10% nonfat dried milk and 0.05% Tween-20. The anti-Kv␤2 polyclonal antibody (Quality Controlled Biochemicals, Inc., Hopkinton, MA) raised against the C-terminal 18 amino acids of the Kv␤2 protein was recognized by an anti-rabbit IgG secondary antibody. Membranes were washed in 100 mM phosphate-buffered saline containing 3% dried milk and 0.05% Tween-20. Immunoreactivity was visualized by a horseradish peroxidase-catalyzed color reaction (Pierce, Rockford, IL).
For all of the experiments reported below, Kv1.3 RNA was injected into Xenopus oocytes at concentrations, approximately 0.2 ng/oocyte, which produced relatively small currents. At concentrations as much as 10-fold higher, the relationship between the amount of RNA injected and current amplitudes is linear, suggesting that cellular translation and transport functions are not saturated in this concentration range (Fig. 1 D).
Functional Expression of ␣1.3␤2,K ϩ Channel Complexes-To investigate the potential effects of the ␤-subunits on Kv1.3 current, we first co-injected Kv1.3 with varying amounts of Kv␤2 cRNA. We observed that the enhancement of Kv1.3 expression increased markedly until the ␤2 cRNA level approached 10-fold that of the Kv1.3 cRNA (Fig. 2A). Beyond that ratio, there was little increase in the enhancement of current, suggesting that saturating amounts of the Kv␤2 subunit had been reached.
Xenopus oocytes were injected with Kv1.3 cRNA and saturating amounts of Kv␤2 cRNA. Recordings were taken at 18, 24, 48, and 72 h after injection. At all time points investigated, there was a striking increase in the current amplitude of Kv1.3 when co-expressed with Kv␤2 (Fig. 2B). The increase in current of Kv1.3 in the presence of saturating amounts of Kv␤2 was 4-fold (Ϯ0.38, n Ͼ 10) at 18 h, 4.2-fold (Ϯ0.45, n Ͼ 10) at 24 h, 3.2-fold (Ϯ0.38, n Ͼ 10) at 48 h, and 2.6-fold (Ϯ0.44, n Ͼ 10) at 72 h (Fig. 2B). Marked increases in current were obtained when at least equal amounts of Kv1.3 and Kv␤2 cRNA were used. Unlike Kv1.4, the inactivation of Kv1.3 was not accelerated when coexpressed with Kv␤2 (Fig. 3, A and C). Scaled currents of Kv1.3 alone (Fig. 4A) and Kv1.3 with Kv␤2 (Fig. 4B) show that kinetics of Kv1.3 are not significantly altered by the presence of Kv␤2. Plots of the normalized voltage-conductance relationships (Fig. 5) rule out the possibility that the Kv␤-mediated increases in current amplitude are because of alterations in the voltage-gating properties of Kv1.3 channels; the probability of opening is saturated at approximately ϩ20 mV both in the absence and presence of the Kv␤ subunits.
Functional Expression of ␣1.3␤1,K ϩ Channel Complexes-We coinjected Xenopus oocytes with Kv1.3 cRNA and saturating amounts of Kv␤1 cRNA. Unlike most other Kv1 channels, Kv1.3 did not exhibit rapid inactivation when coexpressed with Kv␤1 (Fig. 3B), though controls performed with Kv1.4 and Kv␤1 confirmed that the ␤1 cRNA was being translated properly (data not shown). We observed an increase in current amplitude when Kv1.3 was co-expressed with Kv␤1, peaking at slightly more than 2-fold at 24 h, but by 72 h after injection the Kv1.3 current amplitude in the presence of Kv␤1 decreased to 67% of the current of Kv1.3 injected alone (Fig.  2B). The voltage-dependence of Kv1.3 was not affected by the presence of the Kv␤1 subunit (Fig. 5). Taken together, the data indicate that although the Kv␤ subunits do not alter gating or inactivation, they strongly regulate the expression of functional channels at the cell surface membrane.
Kv␤2 Subunit Protein Expression-A polyclonal antibody against Kv␤2 was used to detect the expression of ␤-subunits in T-lymphocytes. In both quiescent and activated T-lymphocytes, a 39-kDa protein is detected at levels which remain relatively constant (Fig. 6). This 39-kDa protein is identical in size to the Kv␤2 channel subunit purified from bovine brain (6) and the Kv␤2 subunit detected in activated murine T-lymphocytes (24). A faint band just below the 39-kDa band in Fig. 6 is likely to correspond to a previously reported splice variant of Kv␤2 (GenBank accession number 2827466), in which 14 amino acids near the N terminus of the protein are absent. In contrast to murine T-lymphocytes, the expression levels of Kv␤2 do not change in activated human T-cells. DISCUSSION T-lymphocytes are critical for eliciting cellular immune responses. It is well established that K ϩ channels play important roles in the activation of T-lymphocytes: important physiologic changes which T-lymphocytes undergo as a result of the activation process are inhibited by blockers of K ϩ channels, including protein synthesis, cell volume increase, and cell-cycle progression (17,19).
It has been demonstrated that channels containing the Kv1.3 subunit are the major K ϩ channel (type n channel) in T-lymphocytes (25). The K ϩ conductance increases 20-fold in murine T-cells treated with mitogen, whereas in human T-cells, the K ϩ conductance increases roughly 2-fold (11,17). However, treatment of these cells by a mitogen results in constant or decreased, rather than increased, Kv1.3 mRNA levels in mice and humans, respectively (22,23). On the other hand, it has been shown that the expression of the murine Kv␤2 subunit increases markedly in response to stimulation by interleukin-2 (24).
Our results show that Kv1.3 can form functional channels alone but that the presence of Kv␤ subunits, primarily Kv␤2, the more abundant ␤-subunit in human T-lymphocytes (24) sion can up-regulate the expression of Kv1.3 channels in activated murine lymphocytes. The much greater increase in Kv␤2 protein levels of mitogen-treated T-cells in mouse (24), compared with human (Fig. 6), is consistent with the differential increase in K ϩ current observed in mitogen-treated T-cells in these two organisms. In fact, although K ϩ current levels in human mitogen-treated T-cells increase roughly 2-fold, this effect is immediate (19), unlike in mouse (20), ruling out transcriptional or translational regulation as the means by which the K ϩ current level increases. The differences in ␤-subunit regulation in mice and humans may account for the vast difference in the extent to which Kv1.3 K ϩ conductance is upregulated in activated T-cells in these two organisms.