Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.


How a Mutation Turned Almonds from Toxin to Treat

Eating just 50 bitter almonds can release enough hydrogen cyanide to kill an adult in under 3 minutes. Fortunately, the sweet variety that we scoop from bins at grocery stores is safe to eat, thanks to a mutation.

While the single-gene glitch behind almond palatability has been recognized for a century, it took genome sequencing to reveal the complex control of the trait. Raquel Sánchez-Pérez, a biochemist at CEBAS-CSIC, an agricultural research center in Spain, and colleagues at the University of Copenhagen and elsewhere in Europe, published their findings in Science in June.

A Beloved Nut Through History

Almonds lead pistachios, brazil nuts, walnuts, pecans, cashews, and pine nuts in tree nut popularity, with peanuts the most popular groundnut.

(Federico Dicenta)

In 2016 Michelle Obama joked that her husband consumes exactly 7 almonds every night. “That’s it!” she exclaimed. When The New York Times reported the observation seriously, eating 7 almonds became a thing.

Headlines shout the attributes. “9 Evidence-Based Health Benefits of Almonds” lists vitamins, minerals, and anti-oxidants, crediting the nuts with lowering blood sugar, cholesterol, and hunger. So many articles tout the ability of almonds to “melt away body fat” that, well, it must be true.

Evidence of almond farming dates to the Fertile Crescent, beginning about 11,650 years ago. Pliny the Elder wrote in his encyclopedia Naturalis Historia in the first century AD that the Romans knew how to remove the bitterness and poison from almonds. The nuts were discovered in King Tut’s tomb from 1324 BC. And a Christian text from the fourth century describes piercing and plugging a tree’s trunk to cause “the … bitter almonds” to “lose the acidity of their juice, and become delicious fruits.”


The hard cores of the bitter nuts of the almond tree Prunus amygdalus share with other members of the Rosaceae family the ability to release hydrogen cyanide, which is a gas consisting of a carbon held by three bonds to a nitrogen. Seeds of apples, flax, and manioc (aka cassava), and the hard innards of peaches, plums, and apricots also release cyanide.

Hydrogen cyanide

The aroma of hydrogen cyanide gas used as a poison is described as “bitter almond” in detective novels, the most famous probably Agatha Christie’s Sparkling Cyanide, published as Remembered Death in 1945. It was part of Zyklon B,  the key component of what the Nazi’s sent into the gas chambers. The cyanide pills that captured spies purportedly pop are salts, potassium cyanide or sodium cyanide. They kill in about 5 minutes.

The cyanide in almonds comes from reaction of a biochemical called amygdalin, which forms from a molecule called prunasin, which forms from the amino acid phenylalanine. As in any biochemical pathway, an enzyme catalyzes (speeds) each step.

In a phenomenon similar to snapping a glow stick into action at a Phish concert, the plant harbors the chemical precursors of cyanide production in different parts that come into contact when an insect or other herbivore crunches a nut. In the “wild type” (non-mutant) plants with bitter nuts, prunasin that has accumulated beneath the seed coat is propelled towards the developing leaflets, where enzymes convert it to amygdalin.


Crunching jaws release the cyanide and benzaldehyde, the source of the bitter taste. While the animal quickly succumbs, the tree gets a brief blast of carbon and nitrogen nutrients.

Audrey the carnivorous plant of “Little Shop of Horrors” had a different sort of defense.

But in mutant almond plants, prunasin is destroyed before it can activate the defense, which Dr. Sánchez-Pérez described in 2008. No toxin, no bitterness, and a sweet taste. The mutation that we enjoy harms the tree by rendering it vulnerable to herbivores.

Cyanide essentially suffocates cells, replacing oxygen in the mitochondria, blocking conversion of the energy held in the bonds of nutrient molecules into ATP, the biological energy currency. That’s why the end comes so swiftly to animals unfortunate enough to bite into a death nut.

Amygdalin enjoyed fame under another name in the 1970s: laetrile, The claimed cancer cure-all came from apricot pits, eventually earning the compound the distinction of being one of the best examples of medical quackery. Laetrile has also been promoted as “vitamin B17,” although it isn’t a vitamin.

More interesting was amygdalin’s role in “death by peach,” practiced in Egypt in the second millennium when the pits were used to poison priests declared traitors. Chemists first isolated the compound, from almonds, in the early 1800s. It belongs to a class of natural toxins called cyanogenic glycosides that more than 3,000 plant species produce.

Finding Sweeter Nuts

Exactly a century ago, experiments reminiscent of Gregor Mendel himself revealed that a single gene controls the bitterness or sweetness of almonds.

(Federico Dicenta)

In a paper in Genetics, “The Factor for Bitterness in the Sweet Almond,” Meyer J. Heppner described experiments he did from 1916 through 1919 at the University Farm in Davis, California. He was trying to delay blooming by a few days to enable the trees to survive late frosts.

The numbers from the breeding crosses that Heppner set up flashed me right back to Mendel’s peas: 32 crosses produced 602 trees, of which 243 bloomed during the study period, yielding 208 with sweet almonds and 59 with bitter almonds.

To a geneticist, that’s a classic 3:1 Mendelian ratio, indicating that one version of the trait is dominant, and one recessive. The fact that “bitter” is recessive explaines why it shows up unexpectedly, just like Mendel’s wrinkled green peas.

Heppner spelled it out: “There is nearly a perfect 3:1 ratio, 3.028:0.972. This close approximation to the theoretical Mendelian monohybrid ratio indicates that all the almond varieties represented in the above crosses are heterozygous for sweetness of the kernel. They must have the genetic constitution Bb, where b represents the factor for bitterness as the recessive character and B the factor for sweetness, as the dominant character … it is possible that a mutation occurred in the bitter almond tree with the sweet almond as the result.”

So Heppner was trying to select for later bloomers and instead found sweeter nuts. Researchers would eventually cleverly name the gene that imparts sweetness when mutant “sweet kernel.” Bitterness can seemingly arise anew when bees carry “b”-bearing pollen to “b”-bearing ovules in flowers.

It was the reappearance of recessive traits that inspired Mendel’s work. But the theoretical is also practical. Each season new trees must be tested (tasted) and those with bitter almonds discarded. It was wasteful. Hold that thought.

The Genomic Approach

Finding the source of the sweetness was a bit circuitous. When the researchers discovered that the two genes whose protein products are necessary to synthesize amygdalin didn’t differ in bitter versus sweet trees, they looked to genes that control the expression of those genes. It’s a little like two people wearing the same piece of clothing, like a scarf, in entirely different ways.

Proteins called transcription factors control whether other genes are turned on or off. Would assembling the genome sequence of the almond tree reveal transcription factor genes that could account for the distinction between bitter and sweet nuts?

A basic helix-loop-helix protein motif

Indeed, within the almond genome, five transcription factor genes cluster on one chromosome. They’re called bHLH, which stands for “basic helix-loop-helix,” for the three-dimensional form that enables these proteins to ping on and off specific sets of genes, controlling their activities.

Of the five bHLH genes, only one, bHLH2, is turned on or off differently in trees yielding sweet versus bitter almonds. So it’s the long-sought “sweet kernel” gene.

A tiny genetic glitch, change of a cytosine (C) to a thymine (T), alters one amino acid in the transcription factor in a way that blocks it from assuming the normal shape, stacked pairs. As a result, the two genes for amygdalin aren’t activated, severing the biochemical pathway to bitterness. The almonds are sweet.

In addition to solving the mystery of the origin of tasty almonds, the new work identified variations in the DNA sequences surrounding the bHLH2 gene. These can be incorporated into a diagnostic test to select seedlings that will yield plants with sweet nuts, rather than having to grow trees three to five years to see whether the nuts are toxic or sweet. In this way researchers can select or modify other genes without concern over appearance of the recessive bitterness trait, saving on costs as well as land use.

(Federico Dicenta)

Implications reverberate beyond the popular nuts. The researchers suggest a genomic approach to detect toxins early in other plants, including gossypol in cotton (a male contraceptive), the anti-oxidant anthocyanins  in strawberries, linamarin and lotaustralin that produce cyanide in cassava, and saponins that make quinoa bitter.

My favorite part of the almond story? The demonstration of the genius of scientists like Gregor Mendel, whose observations and clever experiments, unaided by technology, nevertheless revealed the very laws of nature.

  1. It seems proof reading of this articles was not performed in 1919 before publication. In the original 1923 Genetics article the statement says “243 bloomed during the study period, yielding 208 with sweet almonds and 59 with bitter almonds”. However, the table of this article shows 184 sweet almonds instead of 208. The statement in later article of 1926 Genetics is correct though.
    Still.. Great experiments performed by Meyer J. Heppner giving sweet almonds now.

    1. Who doesn’t? Unfortunately, IMHO, it’s as much an indictment of modern publications as of brilliant work in the past… Just my 2 cents.

Leave a Reply

Your email address will not be published. Required fields are marked *

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Back to top