Eating a potato that has turned green from exposure to sunlight can be toxic and
even fatal. In 1924, Science magazine reported on one such fatal
James Matheney of Vandalia, Illinois, had gathered about one and a
half bushels of tubers, which had turned green. Two days after eating the
potatoes, most of his family – wife, two daughters and four sons – showed
symptoms of poisoning; the only exceptions were James himself, who didn’t eat
the potatoes, and a breast-fed baby boy. His wife, aged 45, died a week later,
followed by their 16-year-old daughter.
The other five members of the
Although such fatalities are rare among human beings,
farm animals often get sick or die after eating green potatoes. Symptoms include
damage to the digestive system as well as loss of sensation, hallucinations and
other neurological disturbances.
Death can be caused by a disruption of
the heartbeat. The culprits are the toxic substances solanine and chaconine;
their concentration rises sharply with exposure to light or during sprouting,
and they protect the tubers from insects and disease.
chaconine belong to the large family of glycoalkaloids, which includes thousands
of toxins found in small amounts in other edible plants, including tomatoes and
eggplant. These substances have been known for over 200 years, but only recently
has Prof. Asaph Aharoni of the Weizmann Institute of Science’s plant sciences
department begun to unravel how they are produced in plants.
He and his
team have mapped out the biochemical pathway responsible for manufacturing
glycoalkaloids from cholesterol.
Their findings will facilitate the
breeding of toxin-free crops and the development of new crop varieties from wild
strains that contain such large amounts of glycoalkaloids they are currently
considered inedible. On the other hand, causing plants to produce glycoalkaloids
if they don’t do so naturally or increasing their glycoalkaloid content can help
protect them against disease.
Two years ago, in research reported in The
Plant Cell, the scientists identified the first gene in the chain of reactions
that leads to the production of glycoalkaloids.
In a new study published
recently in Science, they have now managed to identify nine other genes in the
chain by using the original gene as a marker and comparing gene expression
patterns in different parts of tomatoes and potatoes. Disrupting the activity of
one of these genes, they found, prevented the accumulation of glycoalkaloids in
potato tubers and tomatoes. The team then revealed the function of each of the
genes and outlined the entire pathway, consisting of 10 stages, in which
cholesterol molecules turn into glycoalkaloids.
An analysis of the
findings produced an intriguing insight – most of the genes involved are grouped
on chromosome 7 of the potato and tomato genome. Such grouping apparently
prevents the plants from passing on to their offspring an incomplete
glycoalkaloid pathway, which can result in the manufacture of chemicals harmful
to the plants.
How do we decide which purchases,
news stories and political candidates are worth our time, money and support? In
the Age of Internet, we increasingly rely on the opinions of others as expressed
through aggregated online ratings. But does the so-called wisdom of crowds
produce unbiased information that helps us make better decisions? Or does subtle
online peer pressure create a group-think mentality with negative consequences
for markets, politics and health? To quantify how social influence affects
online decision-making, Israeli and American researchers designed a large-scale
experiment on a popular news aggregation website similar to Digg.com and
Their study, titled “Social Influence Bias: A Randomized
Experiment,” appeared recently in the journal Science. It was conducted by Dr.
Lev Muchnik of Hebrew University’s School of Business Administration, Prof.
Sinan Aral at the Massachusetts Institute of Technology’s Sloan School of
Management and Sean Taylor at New York University’s Stern School of
Over five months, the researchers randomly assigned over
100,000 comments submitted on the site to one of three groups – a positively
manipulated group (up-treated), a negatively manipulated group (down-treated)
and a control group.
To reflect the natural proportions of votes on the
site, they artificially up-voted (+1 rating) 4,049 comments upon their creation
and down-voted (-1 rating) 1,942 others. This created a small random signal of
positive or negative judgment by prior raters, while holding all other factors
The experimental comments were viewed over 10 million times and
rated over 300,000 times by subsequent users.
The researchers found that
positively manipulating a single vote created significant bias in the rating
behavior of subsequent users. This positive herding was topic dependent: in the
seven most active topic categories on the website, there were significant
positive herding effects for comment ratings in politics, culture and business,
but no detectable herding behavior for comments in economics, IT, entertainment
and general news. There was no significant negative herding in any category. The
positive herding was also affected by whether individuals were viewing the
opinions of friends or enemies.
Interestingly, while positive
manipulation had a significant effect, negative manipulation resulted in final
ratings that were not statistically different from the control group. The
researchers explained that although artificial negative ratings generated
negative bias in subsequent user ratings, they also stimulated a countervailing
phenomenon of users fixing the inadequate negative score, thus neutralizing the
effect of social influence.
Muchnik concluded:“As new communication and
information processing technologies assume a more dominant role in our
decisionmaking, this research has implications for electoral polling, stock
market prediction, product recommendation and many other areas. To interpret
collective judgment more accurately and make better use of collective
intelligence, we need to adapt online rating and review technologies to account
for social influence bias.”