Wherever arsenic goes, its sinister reputation precedes it. And arsenic really gets around. While other elements may be found far more abundantly in the Earth’s crust, oceans or atmosphere, at least some measurable amount of arsenic compounds will crop up in almost any sample of soil, water, food or tissue from our bodies. Many if not most of these forms pose little or no threat to human health, but that knowledge flies in the face of a long poisonous past.

Human acquaintance with arsenic goes back more than 2,000 years, when the ancient Greek, Chinese and Egyptian civilizations independently identified two minerals that became known as orpiment and realgar. Both were arsenic sulphides, although they were not recognized as related compounds. They were, however, widely used for dealing with a range of medical conditions, from cancer and fistulas to asthma and other breathing problems. Chinese dragon boat contenders even used to fortify themselves for competition by drinking realgar wine. 
The early alchemists subsequently extracted from the sulphides a third compound, arsenic trioxide, the deadly white powder that is still most typically associated with murder mystery assassins. By the 13th century this pioneering work had isolated from all three agents the common element arsenic, which was named after the Latin word for bold and potent, referring to how well this element combines with others. 

This quality of arsenic explains why it can be found in a wide variety of forms that make their way into an equally wide variety of ecological niches; it is actually unusual to encounter a natural setting where at least trace amounts cannot be detected. For just that reason, none of us should be surprised when arsenic appears in analyses of common staples such as leafy vegetables or rice, which can readily take it in while growing in soil or water with high levels of this element. Instead, we should be asking much more specific questions about what form that arsenic takes. “There should be arsenic in everything,” says Ken Reimer, a professor emeritus of chemistry and chemical engineering at the Royal Military College in Kingston, Ont. “It’s not arsenic the element we’re concerned about. It’s the combination of elements, each with its own physiological, toxicological properties,” Reimer says. “We as chemists should be starting discussions around that theme; otherwise you’re fighting an uphill battle against chemophobia.”

Among Reimer’s favourite ways of initiating that discussion is to ply an audience with wine and shrimp before giving a talk about arsenic. He then points out to them just how much arsenic they have ingested as part of these tempting treats. Once the initial shock of that revelation wears off, Reimer points out that most forms of seafood contain the organic molecule arsenobetaine (C5H11AsO2), a benign form of arsenic that we can consume without hazard. No less compelling is his observation that this molecule is essential to the survival of these ocean dwellers, who use it as an osmolyte to obtain fresh water. 

According to Chris Le, a professor in the University of Alberta’s Department of Laboratory Medicine and Pathology, the dividing line between toxic and non-toxic arsenic is often presented as non-organic versus organoarsenicals. “That is not strictly speaking capturing the essence of the toxicity,” Le says. “It’s true that the arsenic in shrimp and lobster is non-toxic and an organic form, but it doesn’t mean all organic arsenic is non-toxic. There are organic forms of arsenic that are very toxic.”

As for the wine, it contains more toxic arsenicals, but they are rarely if ever present at concentrations that pose a risk. In fact, both the chemical form (speciation) and dose must be considered when assessing risk. California wine fans are regularly dismayed to learn that some of their tipples contain several times the amount of arsenic that the US Environmental Protection Agency (EPA) mandates as the limit for drinking water (10 parts per billion, or ppb). And to be fair, this inorganic variety of the element could be toxic except for the fact that like most mammals, we are capable of methylating it in such concentrations and eliminating it in urine. Such an ability is assumed to be an adaptation we developed as a species in order to inhabit parts of the world where the water or soil might have higher levels of arsenic. 

Researchers continue to marvel at just how sophisticated that adaptation has become in some communities. Earlier this year the journal Molecular Biology and Evolution published a study of native people who have inhabited South America’s high altitude Atacama Desert for more than 10,000 years. This intensely dry landscape is dominated by volcanic rock laced with arsenic, a natural hazard the locals can withstand to a remarkable degree. More specifically, some 70 percent of the population carries a gene that appears to help the liver methylate the arsenic so that higher proportions of this element can be flushed out of the kidneys. In this way, generations of these individuals have been able to survive by drinking water that contains up to 20 times the amount of arsenic considered to be safe. 

Without this genetic advantage, the prognosis for the rest of us attempting such a diet would be grim. The commonly accepted model of arsenic trioxide’s toxicity suggests that As2O3 binds to thiols, a key funcational group on the enzymes crucial to many of the biochemical processes that keep us alive. Meanwhile, arsenic(V) is understood to mimic phosphate in a way that inhibits the output of adenosine triphosphate (ATP), a coenzyme that transports chemical energy between cells. This disruption ultimately leads to apoptosis, a cascade responsible for cellular death on a massive scale. 

At doses as low as 0.5 milligrams a day, the most obvious symptoms might be as comparatively benign as a headache, diarrhea and drowsiness. At these levels, it could take five years or longer for someone to manifest more serious signs such as changing pigmentation of the skin. Meanwhile, internal damage, such as deterioration of nerves leading to the brain and spinal cord, will also be occurring, along with a weakening of blood vessel walls. 

This order of events can be accelerated considerably by a higher dosage. Although some individuals have taken in as much as 10 grams of arsenic(III) and lived to talk about it, 100 to 200 milligrams is generally regarded as sufficient to kill an adult human within 24 hours. Within hours the poisoned subject will begin suffering severe abdominal pain, vomiting, bloody diarrhea and uncontrollable drooling. 

Such an assured outcome has made arsenic among the most popular non-violent tools of murder for literary and historical characters. That status was ensured by the rise of industrial-scale chemical manufacturing in the early 19th century, which featured arsenic as a widely marketed retail solution to a common household problem — rats — which thrived in Europe’s growing urban centres. Arsenic as a rodent pesticide was routinely sold over-the-counter and therefore became easily obtained for more nefarious uses. Perhaps not surprisingly, when Mary Ann Cotton was tried and hanged in 1873 as Britain’s first convicted serial killer, it was for murdering 21 people with arsenic she had obtained from a local store.  

Today we can fight back against arsenic poisoning thanks to the efforts of chemical weapons manufacturers of the early 20th century. Following the advent of toxic gases such as phosgene and chlorodiphenylarsine during First World War, researchers looking for even more effective agents began studying other arsenicals. Among the most daunting of these was beta-chlorovinyldichloroarsinine, which emerged in 1918 from a research group at the Catholic University of America in Washington, DC led by Capt. Winford Lee Lewis. This new chemical weapon was given the name Lewisite and, with the outbreak of the Second World War, a new generation of researchers began looking for an antidote to possible gas attacks. The result was dimercaprol — dubbed British anti-Lewisite (BAL) — a chelating agent that binds strongly with arsenic species. If it is administered before too much arsenic has attached to the body’s essential enzymes, BAL can capture this element as a chelated species that can be eliminated in urine. 

While Lewisite never became the battlefield hazard that had originally been feared, variations of BAL continue to be used wherever arsenic threatens human health. This treatment was a central part of the response to a crisis in Bangladesh that began in the late 1990s, when high levels of arsenic were discovered in tube wells that were being drilled all over this densely populated country. By 2002 an intensive testing regime had concluded that no fewer than 52 million people were drinking water contaminated with arsenic at more than the internationally accepted safety level of 10 parts per billion, with 32 million of those drinking water contaminated to 50 parts per billion. 

At least a portion of this contamination is natural, which has led researchers to begin looking for genetic markers that might indicate whether this is another population that has adapted to these conditions. If that turns out to be the case, it will add to the many facets of this element’s intricate portrait. William Cullen, a professor emeritus in the University of British Columbia’s Department of Chemistry, attempted to draft just such a portrait in 2008 with the publication of a weighty and fascinating tome bearing the provocative title Is Arsenic an Aphrodisiac? Here Cullen explored the complex past of this element, as well as the unanswered questions that still surround its ability to both harm and help us. The current determination of the toxicity of arsenic species depends on the mantra that there is no threshold dose below which no harm is done. A linear extrapolation is currently used by the EPA beyond the ranges where useful epidemiological or animal information is available. There is considerable dispute over this “linear default” model. Some investigators believe that there is a threshold concentration below which arsenic will be harmless and suggest 100 parts per billion would be a good place to start. Others think that extremely small concentrations can be harmful, warranting reduction of that limit to as little as three parts per billion. “That number is wrong when the rice is not nearly as toxic as you think it is,” Cullen says, adding that the fundamental mechanisms that differentiate benign and bioactive arsenic are still being explored. This includes such basic aspects of poisoning such as the effect on the body’s production of ATP, that energy-transporting molecule needed for metabolism. “Although it’s postulated that this takes place, there’s very little evidence that it is a harmful process.”

Cullen, along with Ken Reimer, is currently revising his book for a new edition that will explore the latest insights into how many humans appear to have achieved a biological accommodation with arsenic. This also includes examples of traditions, such as Chinese medicine, that have successfully incorporated this element, leading to the now universal use of arsenic trioxide as a cure for acute promyelocytic leukemia.

For his part, Chris Le is taking an even closer look at exactly how arsenic disrupts the activity of cells in a way that can leave the body vulnerable to longer-term threats such as cancer. “New analytical techniques allow us to determine how arsenic binds to proteins and where on the protein it binds and what sort of changes that causes,” he says. “Certain forms of arsenic bind to certain DNA repair proteins, enzymes that are responsible for repairing DNA.” 

Since those repairs are necessary to counter the impact of environmental carcinogens that we encounter on a daily basis, if this work begins to falter or stops altogether, these agents can gain the upper hand. In this way, some forms of arsenic could be acting through mechanisms that we have yet to uncover, leaving us uncertain of just how much arsenic is bad for us, and how little might actually be good. “We still don’t know how arsenic causes so many adverse outcomes and we don’t know the acceptable level below which there is no substantial increase of health risk,” concludes Le. “Both of these questions are quite fundamental.”