Previous studies report that mice lacking BACE1 have seizures, but disagree about why. Now, German scientists led by Christian Alzheimer and Tobias Huth of Friedrich-Alexander-Universität Erlangen-Nürnberg report that the answer lies in dysfunctional potassium channels. According to their study, published February 25 in the Journal of Neuroscience, BACE1 normally binds the KCNQ, or Kv7, subtype of potassium channel to help it release a steady stream of K+ from the neuron and dampen neuronal firing. Without BACE1, the channel malfunctions and brain cells fire action potentials more frequently. The results suggest that BACE1—normally known for cleaving substrates, including the amyloid precursor protein—interacts with the channel in a non-proteolytic way. While the researchers are unsure what this means for Alzheimer's disease, they suggest it is unlikely that inhibitors of BACE1 protease activity will alter potassium currents. Several BACE inhibitors are in clinical trials for AD.

Riqiang Yan's group at the Cleveland Clinic had previously reported that BACE1 knockouts experience seizures, which the researchers attributed to too many sodium channels (see Jul 2010 news). Subsequently, Robert Vassar and colleagues at Northwestern University in Evanston, Illinois, reported no relationship between sodium channels and neuronal excitability in BACE1 knockouts, calling into question whether the channels caused the seizures (see Hitt et al., 2010). Sodium channels funnel positively charged sodium ions into neurons, depolarizing them sometimes to the point where an action potential will fire.

In the current study, first author Sabine Hessler and colleagues re-examined hyperexcitability in hippocampal slices from BACE1 knockout mice. They found that the threshold needed to spark action potentials, and their amplitudes, matched those in neurons from wild-type mice. If sodium currents were stronger in these mice, Hessler and colleagues would have expected significant changes in the action-potential patterns. Instead, they saw only that the membrane depolarized slightly faster in knockouts. They interpreted this to mean that sodium channels account for just a small part of the hyperactivity.

However, the scientists noticed that while the time between sequential action potentials grew in wild-type neurons, it remained unchanged in neurons from knockout mice, which kept firing rapidly. This pattern of steady firing resembled that in neurons with reduced "M" current. The M-current is a continuous efflux of positive potassium ions through KCNQ channels that helps keep the cells polarized and reduces excitability. Without that current, a cell is more likely to fire an action potential when positive ions enter. The nervous system expresses four such channels (KCNQ2-5), while the cardiovascular system expresses only KCNQ1. Mutations in KCNQ2 and KCNQ3 cause a rare seizure disorder in newborns (see Miceli et al., 2011). Hessler found that a KCNQ blocker called XE991 reduced M-current in hippocampal slices from wild-type mice, but not in those from BACE1 knockouts, suggesting that their KCNQ channels were malfunctioning. The researchers wondered how this related to a lack of BACE1.

Voltage-clamp experiments revealed less M-current in hippocampal pyramidal neurons from knockouts than from wild-type mice. To determine whether BACE1 and KCNQ channels interact, the researchers artificially expressed them in HEK293T cells, which lack voltage-gated ion channels. Cells expressing only KCNQ channels had a moderate M-current that was stronger when accompanied by BACE1. Interestingly, current was equally strong with an enzymatically inactive version of BACE1, suggesting that its interaction with KCNQ channels relied on something besides proteolysis. Immunoprecipitation and proximity ligation assays indicate that BACE1 and KCNQ channels physically interact.

The scientists propose that BACE1 binds KCNQ channels and makes the channel more likely to remain open. They contend that this physical interaction parallels that between KCNQ1 channels on cardiac cells and their current-boosting KCNE1 β-subunits. KCNE1, like BACE1, is a type I transmembrane protein, though it lacks protease activity.

While these results do not exclude a role for excess sodium-channel activity in hyperexcitable neurons from BACE1 knockouts, they suggest that the main effect comes from a drop in M-current. This adds to the understanding of how BACE1 functions normally in the neuron, said Alzheimer (a relative of Alois). It hints that BACE1 inhibitors, which interfere with enzymatic function, may not affect all the physiological functions of BACE1, he added. The authors also speculate that the boost in BACE1 seen in early AD could be a neuron’s attempt to compensate for hyperexcitability by ramping up M-current (see Sep 2002 newsCheng et al., 2014). However, they have not tested whether higher-than-normal BACE1 raises the M-current or reduces hyperexcitability, Alzheimer said.

Yan agreed that this study could highlight an additional contributor to seizures in BACE1 knockouts. He said that changes in sodium versus potassium channels could vary by hippocampal region and that he examined mossy fibers of the hippocampus, while this study focused on neurons from the CA1 region. Yan thinks that a non-proteolytic role for BACE1 is plausible, though he would have been more convinced if the researchers had seen a drop in M-current in BACE1 heterozygotes. Halving BACE1 may have little impact on catalytic cleavage of substrates, but a more dramatic effect on a binding mechanism that depends on the level of expression, he explained. BACE1 heterozygotes have no obvious epileptic seizures, Yan said. Alzheimer disagreed. "It is not obvious to me why halving BACE1 would preferentially affect its non-proteolytic role," he wrote to Alzforum. He noted that a partial pharmacological block of BACE has been recommended as sufficient to reduce Aβ production. Yan also wondered if the BACE1 and KCNQ channels interact normally in neurons, since Hessler and colleagues only demonstrated binding when the proteins were overexpressed in HK293T cells.—Gwyneth Dickey Zakaib


  1. Although such a function may seem extremely unusual, it fits with a previous study from the same group and with increasing reports about non-proteolytic functions of other proteases, such as a transcription factor function of MMP-12 in antiviral immunity (see Marchant et al., 2014) or the γ-secretase-independent functions of presenilin in calcium homeostasis (see McBrayer and Nixon, 2013). I am sure that the group will follow up on these observations and demonstrate the direct (functional) interaction of BACE1 with KCNQ proteins in brain. Additionally, it would be great to generate a catalytically inactive BACE1 knock-in line to observe which functions require proteolytically active or inactive BACE1. In any case, the good news for drug development is that the non-proteolytic functions should not be affected by BACE inhibitors.


    . A new transcriptional role for matrix metalloproteinase-12 in antiviral immunity. Nat Med. 2014 May;20(5):493-502. Epub 2014 Apr 28 PubMed.

    . Lysosome and calcium dysregulation in Alzheimer's disease: partners in crime. Biochem Soc Trans. 2013 Dec;41(6):1495-502. PubMed.

    View all comments by Stefan Lichtenthaler

Make a Comment

To make a comment you must login or register.


News Citations

  1. Off BACE: Lack of Secretase Causes Seizures in Knockout Mice
  2. BACE Above Base in Alzheimer’s Patients

Paper Citations

  1. . BACE1-/- mice exhibit seizure activity that does not correlate with sodium channel level or axonal localization. Mol Neurodegener. 2010;5:31. PubMed.
  2. . The Voltage-Sensing Domain of K(v)7.2 Channels as a Molecular Target for Epilepsy-Causing Mutations and Anticonvulsants. Front Pharmacol. 2011;2:2. Epub 2011 Feb 1 PubMed.
  3. . High activities of BACE1 in brains with mild cognitive impairment. Am J Pathol. 2014 Jan;184(1):141-7. PubMed.

Further Reading


  1. . Control of M-current. Annu Rev Physiol. 1997;59:483-504. PubMed.
  2. . Dorsoventral differences in Kv7/M-current and its impact on resonance, temporal summation and excitability in rat hippocampal pyramidal cells. J Physiol. 2014 Dec 12; PubMed.
  3. . Cortical KCNQ2/3 channels; insights from knockout mice. Channels (Austin). 2014;8(5):389-90. PubMed.
  4. . Early-Onset Epileptic Encephalopathy Caused by Gain-of-Function Mutations in the Voltage Sensor of Kv7.2 and Kv7.3 Potassium Channel Subunits. J Neurosci. 2015 Mar 4;35(9):3782-93. PubMed.

Primary Papers

  1. . β-Secretase BACE1 regulates hippocampal and reconstituted M-currents in a β-subunit-like fashion. J Neurosci. 2015 Feb 25;35(8):3298-311. PubMed.