- Top of page
The whole-cell configuration of the patch-clamp technique, immunoprecipitation experiments and unilateral naris occlusions were used to investigate whether the voltage-gated potassium channel Kv1.3 was a substrate for neurotrophin-induced tyrosine phosphorylation and subsequent functional modulation of current properties in cultured rat olfactory bulb (OB) neurons. Membrane proteins of the OB included all three Trk receptor kinases, but the truncated form of the receptor, lacking an intact kinase domain, was the predominant form of the protein for TrkA and TrkC, while TrkB was predominantly found as the full-length receptor. Acute (15 min) stimulation of OB neurons with bath application of 50 ng ml−1 brain-derived neurotrophic factor (BDNF), which is a selective ligand for TrkB, caused suppression of the whole-cell outward current and no changes in the kinetics of inactivation or deactivation. Acute stimulation with either nerve growth factor or neurotrophin-3 failed to evoke any changes in Kv1.3 function in the OB neurons. Chronic exposure to BDNF (days) caused an increase in the magnitude of Kv1.3 current and speeding of the inactivation and deactivation of the channel. Acute BDNF-induced activation of TrkB receptors significantly increased tyrosine phosphorylation of Kv1.3 in the OB, as shown using a combined immunoprecipitation and Western blot analysis. With unilateral naris occlusion, the acute BDNF-induced tyrosine phosphorylation of Kv1.3 was increased in neurons lacking odour sensory experience. In summary, the duration of neurotrophin exposure and the sensory-dependent state of a neuron can influence the degree of phosphorylation of a voltage-gated ion channel and its concomitant functional modulation by neurotrophins.
Recent evidence has shown that neurotrophins, growth factors and cytoplasmic protein kinases have a neuromodulatory role in addition to their well-studied roles in growth and differentiation (Schlessinger & Ullrich, 1992; Lev et al. 1995; Levine et al. 1995; Berninger & Poo, 1996, 1999; Sherwood et al. 1997; Kafitz et al. 1999). Neurotrophins, namely nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT3) are preferred ligands for TrkA, TrkB and TrkC receptor kinases, respectively. Trk receptors, when bound by their preferred ligands, phosphorylate key tyrosine residues within the intracellular domain of the receptor that serve as a scaffold to which various signalling proteins are recruited and activated (for review see Kaplan & Miller, 2000). Phosphorylation of tyrosine residues, as a result of intercellular communication, has been shown to modulate the activity of voltage- and ligand-gated ion channels, either via direct protein-protein interaction with the tyrosine kinase or through a combination of cell signalling events involving activation of a protein tyrosine kinase (e.g. Huang et al. 1993; Wang & Salter, 1994; Lesser & Lo, 1995; Lev et al. 1995; Berninger & Poo, 1996, 1999; Holmes et al. 1996; Yu et al. 1997; Bolton et al. 2000). Interestingly, tyrosine phosphorylation can produce rapid, short-term changes in ion channel function or long-term changes in ion channel targeting or ion channel protein synthesis.
The voltage-gated potassium channel Kv1.3 is most highly expressed in T-lymphocytes, the dendate gyrus, the olfactory bulb and the olfactory cortex (Kues & Wunder, 1992). Pharmacological evidence has indicated that Kv1.3 carries 60–80 % of the outward current in olfactory bulb neurons (Fadool & Levitan, 1998) and the channel is predominantly expressed in the mitral and granule nerve cell layers in a developmentally regulated manner (Fadool et al. 2000). We also know from previous experiments that Kv1.3 is a molecular target for multiple phosphorylation by four different tyrosine kinases: epidermal growth factor kinase, insulin receptor kinase and src kinase, as well as an unknown kinase that is revealed under pervanadate treatment (Bowlby et al. 1997; Fadool et al. 1997, 2000; Fadool & Levitan, 1998). Each of these tyrosine kinases is expressed in the olfactory bulb and could use Kv1.3 as a substrate for tyrosine phosphorylation to modulate information processing in the olfactory system. The adult olfactory system is unique in its ability to regenerate neurons from a basal cell type, which must continually reestablish circuitry back to the olfactory bulb in order to process odour information leading to sensory perception (Graziadei & Monti-Graziadei, 1978). This regenerative property of the olfactory system has spurred investigations exploring the expression patterns and function of growth factors in this region of the nervous system as a means to understand growth factor-dependent neural development and differentiation and neuronal survival. In the past couple of years the regenerative power of the olfactory system has come into focus as a potential avenue to treat spinal cord injuries (Ramon-Cueto et al. 1998; Bartolomei & Greer, 2000; Franklin & Barnett, 2000; Imaizumi et al. 2000a,b; Ramon-Cueto, 2000; Fry, 2001; Lu et al. 2001; Raisman, 2001). The neurotrophin family of growth factor receptors may play central roles in neuroprotection and enhanced regenerative recovery following spinal cord or nerve injury and animal trials of the delivery of ligands that activate these receptors to the injured site have been performed (Blesch et al. 1998; Zigova et al. 1998; Liu et al. 1999; Namiki et al. 2000; Novikova et al. 2000a,b; Bamber et al. 2001).
Given the high level of TrkB receptors expressed on the dendrites of the mitral cells in the olfactory bulb and the reported release of BDNF ligand from neighbouring granule cells (Masana et al. 1993; Katoh-Semba et al. 1998; Mackay-Sim & Chuah, 2000; Carter & Roskams, 2002), we have developed a cultured olfactory bulb (OB) neuron preparation to explore the degree of modulation of electrical properties in these neurons by BDNF activation of TrkB. In our patch-clamp study we performed acute (min) and long-term (days) treatment of OBNs with neurotrophins and demonstrate that the two protocols have opposite effects on the same voltage-gated ion channel, Kv1.3. Secondarily, BDNF phosphorylation of Kv1.3 is affected by sensory deprivation to the olfactory bulb, suggesting that the sensory-dependent state of a neuron can influence the extent of phosphorylation of an ion channel and concomitant functional modulation by BDNF.
- Top of page
The OB is the first processing station in the olfactory pathway. Members of the Kv family of potassium channels in the OB partly determine the cell's resting potential, regulate the level of neuronal excitability by influencing the duration of the action potential, determine the frequency of repetitive firing and time the interspike interval (Jan & Jan, 1994). Here we show that BDNF activation of TrkB receptor kinase in these neurons uses Kv1.3 as a substrate for tyrosine phosphorylation and alters the functional properties of this channel based upon the duration of trophic factor stimulation and prior odour sensory experience.
Our patch-clamp study in the olfactory bulb suggests that the neurotrophin BDNF may serve multiple functions. The persistent expression of TrkB receptors in both the developing and adult OB implies that BDNF could influence cell phenotype, synaptic plasticity and neuronal regeneration in the olfactory system (Ming et al. 1999) and simultaneously serve a parallel role to modulate electrical activity at the level of the ion channel Kv1.3 by tyrosine phosphorylation. In these TrkB receptor-expressing OB neurons, the duration of neurotrophin exposure affected the type and presumably mechanism of neuromodulation. Acute stimulation (min) with BDNF elicited a suppression of Kv1.3 current magnitude while chronic exposure to BDNF (days) caused an increase in the current magnitude and speeding of the inactivation and deactivation kinetics of the channel. Neurotrophin signalling is known to be activity dependent; for example, release of BDNF can be upregulated by patterned electrical activity (Berninger & Poo, 1999; Balkowiec & Katz, 2000; Poo, 2001); membrane depolarization can increase the expression of TrkB receptor, which can subsequently increase targeting of voltage-gated ion channels to the plasma membrane (Tongiorgi et al. 1997; Urbano & Buno, 2000); and certain disease states, such as Alzheimer's disease, show a decrease in BDNF synthesis (Yan et al. 1997). We show that, while the expression of TrkB and Kv1.3 is not altered by sensory deprivation to the olfactory bulb induced by unilateral naris occlusion, the BDNF-induced tyrosine phosphorylation of Kv1.3 is increased in neurons lacking odour sensory experience. Hence the experience or sensory-dependent state of a neuron can influence the extent of phosphorylation of an ion channel and concomitant functional modulation by BDNF.
In the OB, acute stimulation of Trk receptors elicited changes in ion channel function that were selective to activation of TrkB and were nonmodulatory upon activation of TrkA or TrkC. Chronic stimulation of Trk receptors also elicited changes in Kv1.3 function that were predominantly selective to TrkB, with the exception of NT3 modulation of Kv1.3 inactivation following 3 days treatment with the growth factor. Although our data cannot definitely demonstrate that there was no cross-activity of NGF or NT3 at TrkB receptor proteins, application of these ligands to the OBNs did not elicit changes in Kv1.3 current magnitude as was observed for BDNF. Our finding that the majority of TrkA and TrkC receptor protein exists as the truncated version that lacks kinase activity could explain why ligands for these receptors failed to modulate or inconsistently modulated Kv1.3 current in the OBNs. The role of truncated, noncatalytic forms of Trk receptor protein in the nervous system (Allen et al. 1994) has not been widely studied, but selective loss of full-length TrkB receptor protein over that of its truncated form has been reported in patients with Alzheimer's disease (Allen et al. 1999) and truncation of neurotrophic receptors in general may be developmentally or traumatically induced (Offenhauser et al. 1995; Brodeur et al. 1997; King et al. 2000).
Our previous work demonstrating acute modulation of Kv1.3 by another receptor tyrosine kinase, insulin receptor kinase, suggested that the BDNF-induced current suppression of Kv1.3 was probably attributable to direct tyrosine phosphorylation of the channel. Kv1.3 has 17 tyrosine residues, six of which lie within good recognition motifs for tyrosine-specific phosphorylation (Songyang et al. 1993; Hunter, 1995). Using a combined immunoprecipitation and Western blot strategy, we demonstrated that Kv1.3 phosphorylation is increased following acute stimulation with BDNF by immunoprecipitating Kv1.3 protein and probing this SDS-PAGE-separated protein with anti-phosphotyrosine-specific antibodies. Although we cannot exclude an indirect mechanism of phosphorylation by cross-talk with another signalling cascade, such as neurotrophin activation of src kinases, we can deduce that modulation of Kv1.3 by BDNF can occur independent of synapse formation, because Kv1.3 plus TrkB cotransfected HEK293 cells can be similarly modulated by including BDNF in the patch pipette during cell-attached recordings (data not shown). A physiological relationship between neurotrophin action and K+ channel function has been previously reported for the G protein-gated inward rectifier Kir3 (Rogalski et al. 2000). Interestingly, BDNF regulation of Kir3 required specific tyrosine residues in the carboxyl terminus of both Kir3.1 and Kir3.4 and did not affect the homotetrameric inwardly rectifying potassium channel. The fact that, under our recording conditions, Kv1.3 comprises as much as 80 % of the outward currents in the OB neurons (Fadool & Levitan, 1998), combined with the fact that homotetrameric Kv1.3 channels transfected with TrkB receptors in HEK293 cells are both suppressed and tyrosine phosphorylated by BDNF (data not shown), implies that BDNF can modulate homotetrameric voltage-gated channels by tyrosine phosphorylation of Kv1.3. It is also important to draw parallels between phosphorylation and expression of Kv1.3 channels (Figs 7-10) and BDNF effects on Kv1.3 currents (Figs 1-6), since potassium currents in OB neurons include other members of the voltage-gated potassium channel family, particularly Kv1.4. Although Kv1.3 phosphorylation cannot be analysed in single-cell culture (only whole OB) because of insufficient protein, lack of BDNF effects in margatoxin-treated neurons supports BDNF modulation of Kv1.3 explicitly over that of other family members. Additionally, one must consider that acute stimulation of central neurons by neurotrophins has been reported that is completely phosphorylation independent. Kafitz et al. (1999) have demonstrated acute, rapid modulation of Na+ currents that occurred in the order of 9 ms, much faster than the most rapid detectable tyrosine phosphorylation, which occurs in approximately 30 s.
Chronic stimulation of OBN cultures with BDNF evoked opposite changes in OB current, namely an increase in outward potassium current as opposed to current suppression. Studies have demonstrated that phosphorylation of TrkB by BDNF is achievable in a 5 min time course and will progressively decline over the course of an hour (Yuen & Mobley, 1999). Unfortunately, the protocol of stimulating whole OBs with BDNF did not permit long chronic exposures to the neurotrophin that could then be assessed biochemically to measure the phosphorylation state of either TrkB or Kv1.3. This protocol of maintaining whole bulbs over days in situ would induce inevitable death of neurons, whereas harvesting cultured OBNs chronically supplemented with neurotrophins did not yield enough protein for immunoprecipitation and SDS-PAGE analysis. So it remains to be determined whether there is a downregulation of either TrkB receptor expression or phosphorylation state that would be consistent with a reduced modulation (suppression) of Kv1.3 current in the OB under chronic BDNF stimulation. The left-shifted conductance plots, however, indicate that chronic BDNF stimulation may alter the voltage dependence of Kv1.3. The lower V1/2 values reported for BDNF-treated cells could account for greater whole-cell currents because the neuron could activate sooner, at lower potentials, than time-matched controls. Another likely interpretation of the mechanism of chronic neurotrophin upregulation can be derived from the fact that the neurotrophin NGF is known to regulate the number and distribution of delayed rectifying K+ channels in PC12 cells and SK-SH neuroblastoma cells (Lesser et al. 1997; Yan et al. 1997). Increased targeting of potassium ion channels to the membrane of OB neurons could account for the increased current magnitude or density observed in BDNF-treated cultures. Equally plausible is the fact that expression or targeting of another class of delayed rectifier could be upregulated under chronic BDNF treatment that would mask a continued Kv1.3 current suppression. Independent of mechanism, our data demonstrate that long-term treatment with neurotrophins, such as in treatment for nerve injury or neurogenerative diseases, upregulates potassium currents in neurons, an effect that is the inverse of acute or rapid exposure. BDNF activation of TrkB receptors can thus have different functional effects depending upon duration of stimulation, a finding also reported by Sherwood & Lo (1999), who compared acute versus chronic effects of BDNF on synaptic transmission in CA1 hippocampal cultures.
Finally, our data in sensory-deprived animals with unilateral naris occlusion demonstrate that odour sensory experience is necessary to maintain kinase activity, as observed in control animals for acute BDNF-induced tyrosine phosphorylation of Kv1.3. Odour sensory deprivation failed to alter either TrkB kinase protein expression (Fig. 10) or Kv1.3 protein expression (Fadool et al. 2000), yet phosphorylation of the channel after acute BDNF stimulation was significantly increased. Most recent data have shown that BDNF levels fall in the granule cell layer following sensory deprivation via naris occlusion (McClean et al. 2001), which implies, for our data, that there must be an increased efficacy of the TrkB kinase to yield increased Kv1.3 tyrosine phosphorylation following occlusion. Alternatively, a different signalling cascade could be activated after occlusion to increase Kv1.3 tyrosine phosphorylation in the presence of BDNF. Independent of mechanism, all increases in tyrosine phosphorylation of Kv1.3 by a variety of receptor and cellular tyrosine kinases have been correlated with a decreased total current and would imply that an increase in basal channel phosphorylation induced by sensory deprivation would decrease Kv1.3 current magnitude. If neurotrophic factors are being researched as potential treatments to increase nerve cell generation and growth following injury or deprivation (Sinson et al. 1996; Zochodne, 1996), it is important also to consider altered receptor kinase or channel functions induced by such deprivation, because potential treatments would be administered from an altered basal state. This patch-clamp study of modulation of voltage-gated potassium current (Kv1.3) in olfactory bulb neurons and the parallel biochemical analysis of the phosphorylated state of Kv1.3 ion channel protein demonstrates that Kv1.3 is tyrosine phosphorylated by acute stimulation with the neurotrophin BDNF. BDNF-evoked Kv1.3 phosphorylation induces current suppression and no changes in inactivation or deactivation kinetics or voltage dependence of the potassium current in these OB neurons. Conversely, chronic stimulation with BDNF over days induces an enhancement of the potassium current, probably by a mechanism that changes the voltage dependence of Kv1.3, and a speeding of both the inactivation and deactivation kinetics. Acute stimulation of OB neurons by BDNF is dependent upon odour sensory experience, where unilateral naris occlusion shortly after birth upregulates the increase in tyrosine phosphorylation of Kv1.3 induced by BDNF. Studies of genetically modified mice lacking TrkB or BDNF indicate that lack of expression of the kinase or ligand does not alter the topographical positioning or number of OB glomeruli (Nef et al. 2001). Unfortunately, functional studies have not been undertaken with such a model, so this is an important future directive in understanding how neurotrophins affect ion channel biophysics and electrical patterning for coding of olfactory information.