When you’re having a bad day, consider the teratogenic effects of the Veratrum alkaloid, cyclopamine. Scientists say that an attempt to understand sodium channel openers, in and of itself, may produce similar effects.
The Cardiovascular Toxicity of Plant Sodium Channel Openers
Mad Honey, Cyclopic Ruminants, Sneezing Powder, and Herbal Death
Grayanotoxins, Veratrum Alkaloids, and Aconitum
by Steven Curry, M.D.
Banner – University Medical Center Phoenix
University of Arizona College of Medicine – Phoenix
Medical toxicology fellows commonly memorize a list of sodium channel “openers”. Examples include ciguatoxin, scorpion venom, batrachotoxins, amantadine, DDT and related organochlorine pesticides, pyrethroids, conotoxins, some spider venoms, and from plants, grayanotoxins, Veratrum alkaloids, and Aconitum alkaloids. Signs and symptoms of poisoning by these agents vary somewhat. But the last three, from plants, can produce an illness that resembles cardiac glycoside toxicity, poisoning by sodium channel blockers, and toxicity from agents that antagonize potassium channels.
How can you get a similar illness from one toxin that blocks sodium channels, another that opens sodium channels, a drug that prevents K+ efflux, and yet another that inhibits Na+-K+-ATPase? And aren’t medical toxicology fellows supposed to recall something about plant sodium channel openers in sneezing powder that produced syncope . . .
. . . and that Mithridates VI of Pontus, in an act of chemical warfare, poisoned and incapacitated an estimated 1,000 Roman soldiers with toxic honey in 97 BC . . .
. . . and that aconite, the root of wolfsbane, can produce fatal intoxication, including torsades des pointes . . .
. . . and how do those cyclopic lambs and goats fit in?
There is much to sort out in one post, but let’s try. Today I will emphasize the mechanisms of the cardiovascular toxicity of grayanotoxins and the alkaloids of Veratrum and Aconitum, while only touching on a bit of history and some non-cardiovascular effects. First, we’ll review some simplified and basic physiology we learned in college and medical school so we’re all on the same page.
Voltage-Gated Sodium Channels
I have drawn a simplified illustration of ventricular myocardial cell sitting at resting membrane potential (RMP), typically about -90 mV.
On top, the Na+-K+-ATPase pump moves 3 Na+ out of and 2 K+ into the cell at the expense of ATP hydrolysis. On the left, a Na+-Ca2+ antiporter (exchanger) pumps a calcium ion out of the cell at the expense of 3 Na+ moving in. Ionized, non-protein-bound Ca2+ is about 10,000 times more concentrated outside the cell than inside. Thus, the exchanger is powered by Na+ moving down its electrochemical gradient into the cell, driven by the concentration gradient and attracted by the relatively negative cellular interior. An L-subtype voltage-gated calcium channel (bottom right) is closed at RMP. An ATPase in the sarcoplasmic reticulum (SR) membrane has been pumping free Ca2+ out of the cytoplasm into the SR for storage and subsequent release upon cellular depolarization. And on the top right sits one of countless voltage-gated sodium channels that is in resting state and closed.
Ion channels are categorized by the mechanism by which they conduct ions. Some
Sodium channels comprise various subunits arising from different genes. Each channel comprises one alpha subunit and 2 beta subunits. The alpha subunit is responsible for sensing membrane potential and for pore formation and is the site of action for the toxins of interest to us today. We have 9 genes that code for 9 different alpha subunits (and one or two others being figured out). Different alpha subunits in combination with various beta subunits differ in their behavior and distribution. Nav1.5 is the main alpha subunit in cardiac myocytes, coded for by the gene, SCN5A. Skeletal muscle mainly contains Nav1.4 proteins. Nav 1.1, 1.2, 1.3 and 1.6 are primarily found in the CNS, while Nav 1.7, 1.8 and 1.9 are highly expressed in the peripheral nervous system. In the chart below, I’ve summarized examples of locations and associated inherited disorders (mutations) associated with various alpha subunits.
For our purposes, a sodium channel exists in one of three states (additional states have been described).
At RMP (e.g., -90 mV in cardiac myocytes, during action potential phase 4), an activation gate (top) is closed, and an inactivation gate (bottom) is open. The channel is said to be in a resting state and is closed and cannot conduct sodium.
When the cell membrane becomes depolarized (less negative) from, as examples, propagation of an action potential or depolarization by other ion channels, the activation gate opens and sodium quickly traverses the pore into the cell. Thus, the channel is activated and open. If depolarization is large enough (to about -70 mv) and there is en mass opening of activation gates among a cell’s sodium channels, threshold is said to have been reached and an action potential is generated. This inward sodium current produces a rapid upslope (phase 0) of the action potential.
A characteristic of sodium channels is that they quickly undergo auto-inactivation (deactivation). That is, within a few milliseconds of activation, they change
As the myocyte cell membrane repolarizes (K+ efflux, action potential phases 1-3), the probability of a sodium channel reverting back to the resting state increases as RMP becomes more and more negative and the alpha subunit changes
If an action potential comes down the line before enough sodium channels have reverted to resting, that is, well before RMP has returned to about -90 mV, the cell cannot depolarize to
Scientists perform voltage-clamp experiments and measure the maximal upslope of phase 0 (reflecting how many sodium channels were at rest at the onset of the action potential) at different membrane potentials. Consider this figure:
The X-axis is the RMP when an action potential is generated. The more negative, the more sodium channels are at rest and the greater the inward sodium current and upslope of phase 0 (slope = volts/sec) during the next action potential. The Y-axis represents the maximal slope at any point during phase 0. We’ll come back to this when we consider similarities between sodium channel openers and blockers.
A good way to understand the actions of sodium channel openers is first to be reminded of the actions of cardiac glycosides. We will find similarities with sodium channel openers with regard to cellular events, and this will help us understand why cardiac glycoside toxicity can appear similar, as well. I’ll consider actions of cardiac glycosides in three tissues – myocytes, baroreceptor cells, and skeletal muscle, and will ignore their vascular smooth muscle actions today. We begin with a myocyte at rest, and this time I have added a K+ channel that opens during repolarization.
Cardiac glycosides bind to and inhibit the alpha subunit of Na+-K+-ATPase. Intracellular Na+ concentrations subsequently rise. This decreases the electrochemical gradient driving Na+ into the cell via the Na+-Ca2+ exchanger, secondarily raising intracellular Ca2+ concentrations from decreased export. Because of the Na+-Ca2+ exchanger, concentrations of cytoplasmic Ca2+ follow Na+. That is, rises in cytoplasmic Na+ concentrations raise cytoplasmic Ca2+ concentrations.
But cells do not like large rises in intracellular Ca2+ and use various methods to remove Ca2+ from the cytoplasm. For example, Ca2+ is pumped into organelles such as endoplasmic reticulum and mitochondria; it is exported from the cell via a membrane ATPase (not shown); it is exported from the cell via the Na+-Ca2+ exchanger; and it is pumped into SR where it is concentrated by binding to calsequestrin, a Ca2+-binding protein. When extra Ca2+ is stored in SR, the SR is said to be “loaded”. Positive inotropy from cardiac glycosides is explained by a loaded SR. The next time an action potential depolarizes the cell, Ca2+ influx during phase 2 of the action potential will be accompanied by an increased outflow of Ca2+ from the SR into the cytoplasm, activating contraction. There is also evidence supporting cardiac glycoside-induced changes in sodium channels so they allow Ca2+ influx that may contribute to rises in intracellular Ca2+ concentrations – aka “slip-mode conductance”.
And as you know, skeletal muscle is the main reservoir of the body’s potassium. At high enough cardiac glycoside concentrations (e.g., acute digoxin overdose), the inhibition of Na+-K+-ATPase in skeletal muscle is mainly responsible for hyperkalemia from decreased cellular uptake of K+.
Let’s create a figure summarizing the actions of cardiac glycosides. Thus far we have:
Unfortunately, an overloaded SR in a cardiac myocyte can’t hold all the Ca2+ completely, and after repolarization Ca2+ begins leaking from the SR, raising cytoplasmic Ca2+ concentrations and depolarizing the cell.
In fact, near RMP (phase 4) and when cytoplasmic Ca2+ levels have risen high enough, the Na+-Ca2+ exchanger will export 2 positive charges (a Ca2+ atom) while bringing in 3 positive charges (3 Na+), adding to depolarization, despite some elevation in cytoplasmic Na+ concentration. The cumulative sum of cytoplasmic cations produces what are termed delayed (late) afterdepolarizations, since they occur after repolarization, during phase 4 of the action potential. As seen in the cartoon, below, if an afterdepolarization becomes large enough, an early action potential will be initiated. This explains the myocardial automaticity induced by cardiac glycosides. Clinically, we see arrhythmias such as PVCs, PACs, atrial tachycardia, junctional tachycardia, ventricular tachycardia (sometimes with bidirectional V-tach), and ventricular fibrillation.
Adding to our figure we began, we now have the following:
Cardiac glycosides are best known for positive inotropy, increased automaticity, and increased vagal tone. We’ve addressed the first two, but why would they increase vagal tone, producing heart block, bradycardia, and vomiting? We now must jump to baroreceptors.
Baroreceptors are neurons that increase rates of firing in response to rises in blood pressure, or, more specifically, in response to physical deformation. Baroreceptors of interest to us are those that reside in the carotid sinus where they sense arterial pressure, and in the ventricle, where they sense the force of contraction. These specialized neurons contain not only voltage-gated sodium channels (mainly Nav1.7, Nav1.8, and Nav1.9), but also contain a type of sodium channel that allows inward sodium currents, not in response to binding by ligands or by changes in membrane potential, but by actual physical deformation. These channels are named epithelial sodium channels (ENaC). Scientists have proposed two different models to explain their action, but both rely on physical deformation of the channel to cause channel opening and conduction of Na+. Similar ENaC proteins are responsible for our sense of touch, but ENaC proteins are found throughout our body, including kidneys, brain, bladder, skeletal muscle and other locations. With each bolus of blood that distends the carotid sinus during systole, or with each force of ventricular contraction, ENaC proteins open and depolarize baroreceptor cells to
If we would have plotted individual pulsatile changes in arterial pressure, we would see the rate of neuronal firing go up and down with each rise and fall in pressure. But here we plot mean arterial pressure to average things out and demonstrate that the higher the pressure, the more frequently the baroreceptors fire.
As we learned long ago, the afferents from baroreceptors travel to our brainstem where rapid-firing results in increased vagal tone (increased ACh release) and decreased sympathetic tone, with subsequent declines in heart rate and BP, sometimes with varying degrees of AV block and even asystole. Below is a baroreceptor cell.
Here we find endoplasmic reticulum that takes up Ca2+ with a pump very similar to that found in SR, the Na+-Ca2+ exchanger, Na+-K+-ATPase, voltage-gated sodium channels, etc. The ENaC protein is shown in blue. And similar to
With regard to digoxin, the carotid sinus baroreceptors appear most important in fooling the brainstem. The evidence for this is that digoxin-induced bradycardia and heart blocks are completely prevented not only by cutting the vagus nerve, but also by cutting the glossopharyngeal nerve. Let’s complete our figure before moving on to plant sodium channel openers.
Plant Sodium Channel Openers
We are now ready to discuss plant sodium channel openers in more detail, again focusing on their cardiovascular actions. First, let’s look at a cartoon drawing of a generic sodium channel alpha subunit that has been stretched out in 2 dimensions and discuss its organization and sites of binding for drugs and neurotoxins in more detail.
Alpha subunits contain about 2,000 amino acids and are divided into four major domains (I – IV). Each domain spans the plasma membrane six times (S1 – S6). The S1–4 spans of the four domains collectively form the voltage sensing portion of the subunit, while S5-6 spans of each domain combine to form the pore. Usually, two of at least five types of regulatory beta subunits (not shown) bind covalently or noncovalently to complete the sodium channel.
Binding sites for various neurotoxins and drugs that affect sodium channel function have been identified through competitive binding studies and through experiments in which binding is affected by specific amino acid substitutions. Some of the agents in the above figure are sodium channel blockers, while others are openers. Note that all of the plant sodium channel openers we are discussing in this column bind at neurotoxin site 2 along with batrachotoxins. The exact residues at site 2 where individual neurotoxins bind may overlap some, and there is evidence that batrachotoxins interact with S6 spans in all four domains.
Numerous actions have been described for plant sodium channel openers, depending on the species, the tissue and the subtype of sodium channel. Site 2 neurotoxins display greatest affinity for activated (open) sodium channels and, upon binding, cause the channels to open more easily at a given membrane potential and/or to remain open longer (prevent or slow inactivation). More specifically:
- The voltage-dependence of activation is shifted negatively, causing channels to open near or at RMP, and/or . . .
- Inactivation is slowed or inhibited, resulting in sustained inward sodium currents after the peak of the action potential. The red bar, below, indicates a block of inactivation.
Before we address specific sodium channel openers, general statements can be made about their actions. Whether a rise in cytoplasmic Na+ concentration results from inhibition of Na+-K+-ATPase (cardiac glycoside) or from a site 2 neurotoxin, the results are similar:
Initialincreased inotropy in myocytes
- Increased automaticity, and
- Increased vagal tone
Baroreceptor activation both in the carotid sinus and ventricular wall contribute to increased vagal tone in animals receiving Veratrum alkaloids, grayanotoxins, or Ac
If a toxin mainly increases automaticity in baroreceptor cells, then bradycardia and AV blocks will prevail, without much in the way of premature contractions or ectopic tachyarrhythmias. On the other hand, if myocytes or conducting cells are significantly affected, then tachyarrhythmias, including bidirectional ventricular tachycardia and ventricular fibrillation can result.
Slowing of sodium channel inactivation and a persistent inward sodium current at or near RMP prevents complete repolarization of the cell membrane between action potentials. There will be fewer channels at rest and available for activation when the next action potential comes down the line from above. Here is Vmax versus RMP again.
If a ventricular myocyte or Purkinje cell is only partially repolarized, yet enough so to be within the relative refractory period, depolarization can generate an action potential, but the upslope of phase 0 (Vmax) will decrease, slowing intraventricular conduction. Furthermore, the third property of site 2 toxins is that they directly lessen the peak rate of Na+ influx to various degrees. Although total Na+ influx and intracellular sodium levels are increased by channel openers, the maximal rate of Na+ influx during the short phase 0 is somewhat decreased by this third action of sodium channel openers. Some models suggest a binding site for openers within the pore (affecting inactivation and rate of Na+ influx), but this is not established. Regardless, the decreased number of channels at rest at the initiation of an action potential along with the ability of channel openers to decrease the rate of Na+ influx directly can decrease upslope of phase 0 and produce a lengthened QRS complex, just as sodium channel blockers do.
The consequences and action of sodium channel openers
Early afterdepolarizations are those that occur during phase 2 and 3 of the action potential. When induced by medications, they usually result from blockade of HERG potassium channels with subsequent prolongation of repolarization, lengthening of the action potential, and prolongation of QT intervals on the ECG. During the prolonged depolarization from sodium channel openers, excessive Ca2+ influx through L-subtype Ca2+ channels depolarizes the cell, and such depolarizations can oscillate and reach threshold to produce ectopic beats, ventricular tachycardia, torsades des pointes, and ventricular fibrillation. About 20 years ago, Lewis Nelson visited us in Phoenix and delivered an excellent lecture in which he pointed out that in vitro voltage-clamp experiments had demonstrated that during prolonged depolarization, voltage-gated Ca2+ channels repeatedly inactivate and reactivate (close and open), even without repolarization of the plasma membrane to RMP, allowing continued Ca2+ influx during the prolonged phase 2.
Most Na+ entry into an excitable cell results from influx through sodium channels and via the Na+-Ca2+ exchanger. I previously wrote that normally the action potential peak occurred when sodium channels underwent rapid auto-inactivation. This is not really the entire story. It turns out that roughly 0.1 to 0.5% of sodium channels normally do not undergo rapid inactivation, but continue to conduct inward sodium currents into later phases of the action potential. This late sodium current contributes to the shape of action potential phase 2. (Indeed, at least four inheritable long QT syndromes are from sodium channel mutations.) While the rate of Na+ influx (current) certainly is greatest during the brief upslope of phase 0, up to roughly 30% of total sodium influx actually occurs as a consequence of the late sodium current, since the total duration of action potential phases 1, 2 and part of 3 are much longer than phase zero. Thus, a toxin, disease or illness that only mildly prevents channel inactivation and increases the late sodium current may easily double, or more, Na+ influx into the cell, greatly contributing to afterdepolarizations.
The Na+-Ca2+ exchanger, like most antiporters, is reversible. Even in a normal cell near the peak of the action potential, the electrical potential and incremental Na+ concentration along the plasma membrane make it unfavorable for Na+ influx via the Na+-Ca2+ exchanger. The exchanger normally reverses direction when the cell is briefly depolarized, with 3 Na+ efflux being coupled to Ca2+ influx. (Animal data indicate this Ca2+ influx may contribute to inotropy.)
Thus, rises in cytoplasmic Ca2+ concentrations resulting from an increased late sodium current (or from blockade of K+ efflux produced by a drug) result both from the prolonged opening of voltage-gated L-subtype Ca2+ channels as well as from the prolonged reversal of the Na+-Ca2+ exchanger. In other words, early afterdepolarizations from sodium channel openers are understood to result from a combination of Ca2+ and Na+, but with a greater Ca2+ contribution from reversal of the exchanger and from prolonged Ca2+ influx via Ca2+ channels, rather than from Ca2+ leakage from overloaded SR (or endoplasmic reticulum), as with cardiac glycosides. However, exaggerated late sodium currents certainly can cause delayed afterdepolarizations, as well. The above figure could also be a baroreceptor cell, with endoplasmic reticulum and ENaC proteins. I have attempted to simplify the major mechanisms of increased automaticity in myocardium and baroreceptors in the following figure:
If a sodium channel opener both delays inactivation to produce an increased late sodium current and also increases inward sodium currents at RMP, then a combination of all of the aforementioned may occur. While we attempt to categorize and simplify molecular actions of toxins, nature may combine all of our nice categories into one complicated illness. In fact, a fourth action of some sodium channel openers is to change ion selectivity of sodium channels so they allow Ca2+ influx.
Now we understand the general mechanisms by which plant sodium channel openers produce bradyarrhythmias, heart blocks, and asystole (increased vagal tone), increase myocardial automaticity with premature beats and tachyarrhythmias, widen QRS complexes, and lengthen QT intervals. And it is apparent why poisonings by plant sodium channel openers can resemble those from cardiac glycosides (but without hyperkalemia), potassium channel blockers (long QT, torsades des pointes), and even sodium channel blockers (bradycardia, blocks, prolonged QRS). With this background, let’s briefly examine each of the three groups of toxins and see what clinical effects are reported with acute poisoning. There are countless review articles on the clinical aspects of these specific plant sodium channel openers, and some are included in the references for those wanting more information.
Plants of the genera Veratrum and Toxicoscordion (formerly Zigadenus) in the Liliaceae family contain more than 200 alkaloids divided into various groups, based on their structure.
Many of these alkaloids, such as veratridine, act as sodium channel openers.
The medicinal use of Veratrum alkaloids began long ago, including by Native Americans. Pliny the Elder referred to the use of Veratrum album as a diuretic in his pharmacopeia published in A.D. 77. In the 1950s, extracts of Veratrum were used to treat hypertension in the U.S. and other countries, and use was accompanied by bradycardia. Humans receiving IV Veratrum alkaloids demonstrated bradycardia, ventricular arrhythmias (PVCs, bigeminy), and AV junctional rhythms.
Reports of syncope, vomiting, bradycardia, and hypotension following exposure to sneezing powder manufactured in Germany appeared in Europe in the early 1980s. Sodium channel openers are effective at triggering the sneezing reflex through generation of action potentials in nasal sensory nerve endings. Investigations led to the recognition that this product comprised 50% Panama wood and 50% Veratrum album. It was rapidly removed from the market.
In Arizona, most Veratrum alkaloid poisonings we encounter result from mistaking Toxicoscordion species (known as various types of “death camas”) for the edible onion-like common camas.
Patients who eat the bulbs present with sinus bradycardia, AV blocks, hypotension, vomiting, syncope, and occasional diarrhea. It is uncommon to see cardiac extrasystolic beats or ectopic pacemakers. QRS and QT durations remain normal at doses our patients consume. Similar findings have been described when Veratrum album has been mistaken for wild garlic or other foods. Paresthesias with burning and numbness, impaired vision, and sedation reflect actions on CNS and peripheral neuronal sodium channels. When more than a single dose of atropine is required to correct bradyarrhythmias, AV blocks, and vomiting, we have used glycopyrrolate so as not to produce central anticholinergic effects. Authors have alluded to large ingestions of plants containing Veratrum alkaloids producing long QT intervals and ventricular arrhythmias, but these are exceptional.
In 1958 livestock owners associated cyclopia in lambs with grazing on Veratrum californicum. Subsequent studies demonstrated that feeding of V. californicum on the 14th day of gestation was responsible for cyclopia and associated congenital defects, and in 1968 the responsible alkaloid, cyclopamine, was isolated from the plant.
Teratogenicity of cyclopamine and since-identified related alkaloids results from interference with the hedgehog signaling pathway, not from sodium channel opening. This interference has led to studies examining the effectiveness of structurally similar substances for the treatment of some human malignancies. As humans do not feed on Veratrum species in large amounts, these plants do not represent a teratogenic risk to man.
Many genera of the Ericaceae family, including Rhododendron, Pieris, and Kalmia, contain terpenoid toxins known as grayanotoxins (formerly andromedotoxins).
Ingestions of these plants or, far more commonly, honey produced by bees that have fed on these plants, produce an illness characterized by bradycardia, hypotension, syncope, AV blocks, paresthesias, vomiting and, at times, diarrhea. In some cases, altered mentation, including “intoxication” and hallucinations ensue. In severe cases, asystole, atrial fibrillation, and convulsions have appeared. The toxic honey is known as “mad honey” or “sneezing honey” and is also intentionally collected and consumed for purposes of recreation or as an alleged aphrodisiac. Most reports of grayanotoxin poisoning today result from consumption of honey from the eastern Black Sea region in Turkey. Nevertheless, reports of mad honey from North America and other parts of the world have been documented.
As is the case with Veratrum alkaloids, grayanotoxin poisoning is not associated with long QT intervals or ventricular arrhythmias in most cases, but these would be possible with large doses.
Historical events describing intoxication by mad honey begin in 401 B.C. with Xenophon, an army commander and historian. King Mithridates, Eupator of Pontus, in 97 B.C., intentionally and successfully placed mad honeycombs in the path of Roman soldiers, which they consumed, in order to incapacitate them. A similar event is found in Russian history.
It is interesting to compare the report of Xenophon in 401B.C. with that of Kebler in 1896. Xenophon wrote that his troops who ate honey near the Black Sea were “off their heads”, experienced diarrhea and vomiting, and appeared drunk and crazy. He stated they lay on the ground as though defeated, but the next day began to return to their senses and began to arise on the 3rd and 4th days from their states of intoxication, without a single death. In 1896, Kebler described a husband and wife in New Jersey who ingested small quantities of honey and noted a burning taste immediately. This was followed by vomiting, abdominal pain, loss of consciousness, cold extremities, and a “feebly acting heart”. The husband had suddenly lost vision before collapsing to the floor. Vomiting and abdominal cramps with cold skin and pallor continued for hours, and no pulse could be detected in the man’s wrist for two hours. His heart tones were irregular, as were his wife’s. The wife regained consciousness in about 3 hours, and the husband did not feel himself for 19 hours. Overall “restoration of strength was very slow”. As sick as Xenephon’s soldiers were compared to the New Jersey couple, I’d say the soldiers took in more grayanotoxins. It would be fascinating to know what their blood pressures and ECG tracings had been.
While the clinical picture of poisoning by Veratrum alkaloids or grayanotoxins most commonly comprises increased vagal tone (e.g., bradycardia, hypotension, blocks, syncope) from baroreceptor firing, Aconitum alkaloids are an entirely different story; we have come to the big guns of plant sodium channel openers. The genus Aconitum (e.g., monkshood) has been used as an herb for centuries to treat various unrelated disorders.
Aconite, an herbal preparation from the tuber, contains numerous terpenoid alkaloids, including aconitine, mesaconitine, yunaconitine and others (all parts of the plant are toxic). Tincture of aconite may be prepared from either whole plant, tubers, and/or leaves. A paste is also used for medicinal purposes.
Aconite toxicity from medical herbs or herbal soup has been best-described in China, but acute and fatal toxicity has been reported from around the world. In traditional Chinese medicine, the roots/tubers first undergo processing which reduces alkaloid content dramatically. Ingestions of wild plants, which are not first processed, typically result in more severe toxicity.
Patients commonly present with facial and extremity paresthesias. Cardiovascular toxicity includes bradyarrhythmias, asystole, hypotension, junctional rhythms, PVCs, ventricular tachycardia, ventricular fibrillation, intraventricular conduction delays (wide QRS), long QT interval, torsades des pointes, and bidirectional ventricular tachycardia. Many reports describe difficulties in treating shock and refractory ventricular arrhythmias. Shock and, presumably, neurotoxic CNS effects explain reports of coma and seizures, though deaths result from cardiovascular toxicity. GI findings include vomiting, abdominal cramping, and diarrhea.
It is not the main purpose of this column to discuss treatment in any detail. As a general statement, atropine and, at times, vasopressors, are all that are needed to correct cardiovascular toxicity from Veratrum alkaloids or grayanotoxins.
When it comes to ventricular arrhythmias from Aconitum species, there are no randomized controlled trials to guide us. Anecdotal observations support common findings of ventricular arrhythmias refractory to treatment, whether by drugs or cardioversion. Amiodarone has been used successfully, but this may reflect more common use rather than greater effectiveness than other antiarrhythmics. Magnesium has been given for long QT and torsades des pointes, of course. Most other antiarrhythmics have been tried with varying results. Some patients have required ECMO and repeated cardioversion with good outcomes.
One might presume that if the illness is produced by a sodium channel opener, then a sodium channel blocker would be quite effective. But carefully think this through. Both openers, which mainly bind to activated sodium channels to slow or prevent inactivation, and blockers, which mainly bind to inactivated sodium channels to slow recovery, will decrease the number of sodium channels at rest. This is especially true during tachycardia, when more channels are in the active and inactive states per unit time, making them available for binding to openers and blockers. This is summarized in the following figure:
Thus, one possibility is that the combination of an opener and a blocker may dramatically worsen conduction and contractility. On the other hand, if a sodium channel blocker competed for binding with an opener, or bound elsewhere, but allosterically decreased the action of the opener on the same channel, perhaps a beneficial effect might result. For example, tetrodotoxin, a sodium channel blocker, has been reported in vitro to lessen increased sodium influx due to plant sodium channel openers on many types of sodium channels. I have heard physicians recite this fact to support the conclusion that sodium channel blockers would be effective antidotes for plant sodium channel openers. But there are a couple of problems with this. First, tetrodotoxin displays little activity on Nav1.5 (cardiac) channels. Second, tetrodotoxin binds to site 1, while our antiarrhythmics (and TCAs) mainly bind to the LA (local anesthetic) site (F
In 1998 I was discussing a new paper that had recently been published on slip-mode conductance of sodium channels with Paul Pentel, a very smart medical toxicologist from the University of Minnesota and Hennepin County Medical Center. Paul had developed and published extensively on a rat model of tricyclic antidepressant (TCA)-induced sodium channel blockade, and we had adopted this same model to examine chloroquine toxicity in our lab. Paul happened to mention an experiment that I’ve not forgotten. Paul had wondered if a sodium channel opener might reverse imipramine-induced sodium channel blockade in the rat. Imipramine-toxic animals were given batrachotoxin, a sodium channel opener that also binds to neurotoxin site 2, like plant toxins we have been discussing. The moment batrachotoxin was given, QRS complexes dramatically widened and blood pressures plummeted. Bad things were happening, and it was clear the combination of an opener and closer was dramatically decreasing sodium channels at rest at the initiation of action potentials. Some will say this was a model of sodium blocker toxicity treated with an opener, while we are concerned about sodium channel opener toxicity treated with a blocker. But Paul’s observations support our concern for worsening cardiac conduction with a combination of a blocker and opener.
The end result is that physicians confronted with life-threatening ventricular arrhythmias try various drugs in attempts to see what works, recognizing that some of them might make things worse. When the patient has a high probability of death, potential benefits certainly outweigh risks, such as when recurrent ventricular fibrillation is controlled. But when amiodarone, a drug with multiple pharmacologic actions, appears effective, it may have little to do with its action blocking sodium channels. Amiodarone’s effectiveness may result from blocking Ca2+ influx during prolonged depolarization from enhanced late sodium currents. The ability to prevent Ca2+ influx, in fact, explains why amiodarone, a drug that also blocks K+ efflux and lengthens QT intervals, uncommonly produces torsades des pointes.
For some incredible scenery and mad honey collecting, interested readers may want to view a short documentary on honey gatherers in Nepal. Go full-screen on a large desk-top computer for full visual impact, but be cautioned that there are some obscenities.
For those who enjoy toxicologic history, pick up Adrienne Mayor’s superb book, The Poison King: The Life and Legend of Mithridates, Rome’s Deadliest Enemy (ISBN-13: 978-0691150260). Mithridates’ use of mad honey to poison Roman soldiers was but a small part of his experience with poisons and antidotes throughout his life. Simply a great read.
Fellows, please, please leave some feedback. It really does help us decide on topics and on how deep to go into the molecular and cellular toxicology. Really. Please.
Antzelevitch C, Nesterenko V, Shryock JC, Rajamani S, Song Y, Belardinelli L. The role of late I Na in development of cardiac arrhythmias. Handb Exp Pharmacol. 2014;221:137-68. 24737235
Drummond HA, Jernigan NL, Grifoni SC: Sensing tension: epithelial sodium channel/acid-sensing ion channel proteins in cardiovascular homeostasis. Hypertension. 2008 May;51(5):1265-71. 18378856
Larbig R, Torres N, Bridge JH, Goldhaber JI, Philipson KD: Activation of reverse Na+-Ca2+ exchange by the Na+ current augments the cardiac Ca2+ transient: evidence from NCX knockout mice. J Physiol. 2010 Sep 1;588(Pt 17):3267-76. 20643777
Lytton J, Westlin M, Burk SE, Shull GE, MacLennan DH. Functional comparisons between isoforms of the sarcoplasmic or endoplasmic reticulum family of calcium pumps. J Biol Chem. 1992 Jul 15;267(20):14483-9. 1385815
Mangold KE, Brumback BD, Angsutararux P, Voelker TL, Zhu W, Kang PW, Moreno JD, Silva JR: Mechanisms and models of cardiac sodium channel inactivation. Channels (Austin). 2017 Nov 2;11(6):517-533. 28837385
Mullins M. Mon bête noir (my pet peeve). J Med Toxicol. 2011;7(2):181. 21461788
Santo-Domingo J, Demaurex N: Calcium uptake mechanisms of mitochondria. Biochim Biophys Acta 2010;1797(6-7):907-12. 20079335
Schild JH, Kunze DL: Differential distribution of voltage-gated channels in myelinated and unmyelinated baroreceptor afferents. Auton Neurosci. 2012 Dec 24;172(1-2):4-12. 23146622
Shryock JC, Song Y, Rajamani S, Antzelevitch C, Belardinelli L: The arrhythmogenic consequences of increasing late INa in the cardiomyocyte. Cardiovasc Res. 2013 Sep 1;99(4):600-11. 23752976
Stevens M, Peigneur S, Tytgat J: Neurotoxins and their binding areas on voltage-gated sodium channels. Front Pharmacol. 2011 Nov 9;2:71. doi: 10.3389/fphar.2011.00071. eCollection 2011. 22084632
Bronsky D, Berstein M, Chesrow EJ: Cardiac arrhythmia from Veratrum viride. Geriatrics. 1957 Jun;12(6):389-93. 13428009
Carlier P, Efthymiou ML, Garnier R, Hoffelt J, Fournier E: Poisoning with Veratrum-containing sneezing powders. Hum Toxicol. 1983 Apr;2(2):321-5. 6862477
Chandler CM, McDougal OM: Medicinal history of North American Veratrum. Phytochem Rev. 2014 Sep;13(3):671-694. 25379034
Gilotta I, Brvar M: Accidental poisoning with Veratrum album mistaken for wild garlic (Allium ursinum). Clin Toxicol (Phila). 2010 Nov;48(9):949-52. 21171854
Jaffe AM, Gephardt D, Courtemanche L: Poisoning due to ingestion of Veratrum viride (false hellebore). J Emerg Med. 1990 Mar-Apr;8(2):161-7. 2362117
Kauntzer R, Trounce J: The hypotensive action of veriloid (Veratrum viride). A clinical investigation. Lancet. 1951 Mar 10;1(6654):549-55. 14805125
Lee ST, Welch KD, Panter KE, Gardner DR, Garrossian M, Chang CW: Cyclopamine: from cyclops lambs to cancer treatment. J Agric Food Chem. 2014 Jul 30;62(30):7355-62. 24754790
Quatrehomme G, Bertrand F, Chauvet C, Ollier A: Intoxication from Veratrum album. Hum Exp Toxicol. 1993 Mar;12(2):111-5. 8096707
Schep LJ, Schmierer DM, Fountain JS: Veratrum poisoning. Toxicol Rev. 2006;25(2):73-8. 16958554
Ulbricht W. Effects of veratridine on sodium currents and fluxes. Rev Physiol Biochem Pharmacol. 1998;133:1-54. 9600010
Anonymous. Poisonous honey. Lancet 1896; 148 (3815):1027-28. DOI: https://doi.org/10.1016/S0140-6736(01)76117-X
Brown BS, Akera T, Brody TM: Mechanism of grayanotoxin III-induced afterpotentials in feline cardiac Purkinje fibers. Eur J Pharmacol. 1981 Nov 5;75(4):271-81. 7318913
Erenler AK: Cardiac Effects of Mad Honey Poisoning and Its Management in Emergency Department: A Review from Turkey. Cardiovasc Toxicol. 2016 Jan;16(1):1-4. 25613735
Gunduz A, Turedi S, Russell RM, Ayaz FA: Clinical review of grayanotoxin/mad honey poisoning past and present. Clin Toxicol (Phila). 2008 Jun;46(5):437-42. 18568799
Gunduz A, Turedi S, Oksuz H: The honey, the poison, the weapon. Wilderness Environ Med. 2011 Jun;22(2):182-4. 21664563
Jansen SA, Kleerekooper I, Hofman ZL, Kappen IF, Stary-Weinzinger A, van der Heyden MA. Grayanotoxin poisoning: ‘mad honey disease' and beyond. Cardiovasc Toxicol. 2012 Sep;12(3):208-15. 22528814
Kalkan A, Gökçe M, Memetoğlu ME: An unusual clinical state: atrial fibrillation due to mad-honey intoxication. Anadolu Kardiyol Derg. 2012 Jun;12(4):365-6. 22498862
Ku DD, Akera T, Frank M, Brody TM, Iwasa J: The effects of grayanotoxin I and alpha-dihydrograyanotoxin II on guinea-pig myocardium. J Pharmacol Exp Ther. 1977 Feb;200(2):363-72. 138734
Lee SW, Choi SH, Hong YS, Lim SI: Grayanotoxin poisoning from
O’Neil W: Poisonous honey. Lancet 1896; 148(3818):1262. DOI: https://doi.org/10.1016/S0140-6736(01)76281-2
Silici S, Atayoglu AT. Mad honey intoxication: A systematic review on the 1199 cases. Food Chem Toxicol. 2015 Dec;86:282-90. 26547022
Chan TY: Aconite poisoning. Clin Toxicol (Phila). 2009 Apr;47(4):279-85. 19514874
Chan TY: Incidence and Causes of Aconitum Alkaloid Poisoning in Hong Kong from 1989 to 2010. Phytother Res. 2015 Aug;29(8):1107-11. 25974837
Chan TY: Aconitum alkaloid poisoning related to the culinary uses of aconite roots. Toxins (Basel). 2014 Sep 2;6(9):2605-11. 25184557
Coulson JM, Caparrotta TM, Thompson JP: The management of ventricular dysrhythmia in aconite poisoning. Clin Toxicol (Phila). 2017 Jun;55(5):313-321. 28421842
El-Shazly M, Tai CJ, Wu TY, Csupor D, Hohmann J, Chang FR, Wu YC: Use, history, and liquid chromatography/mass spectrometry chemical analysis of Aconitum. J Food Drug Anal. 2016 Jan;24(1):29-45. 28911407
Fu M, Wu M, Qiao Y, Wang Z. Toxicological mechanisms of Aconitum alkaloids. Pharmazie. 2006 Sep;61(9):735-41. 17020146
Li H, Liu L, Zhu S, Liu Q. Case reports of aconite poisoning in mainland China from 2004 to 2015: A retrospective analysis. J Forensic Leg Med. 2016 Aug;42:68-73. 27266651
Martinez A, Dobos N, Rotella JA, Greene SL. Life-threatening cardiovascular toxicity following ingestion of Chinese herbal medicine. Emerg Med Australas. 2014 Oct;26(5):512-3. 25159087
Sheth S, Tan EC, Tan HH, Tay L. Herb-induced cardiotoxicity from accidental aconitine overdose. Singapore Med J. 2015 Jul;56(7):e116-9. 26243980