Storytelling and science communication (part 1)

A few thoughts and resources for teachers

There are as many ways to tell a science story as there are writers, and as many ways to give a talk or make a poster as there are scientists. Still, there’s a difference between effective communication and efforts that miss the mark. After years of writing about science and trying to teach others to do so, I’ve gained some experience that may be useful to other teachers, or may at least get the ball rolling on a larger discussion.

A note on the context: most of my teaching has taken place in Germany, one of those places that has been notorious for failing to develop a notion of functional communication and teaching it across the curriculum. The results have been predictable; to quote William Zinsser, “Literature professors shouldn’t be left alone in teaching a skill that is inherent to every field.” (His book, On Writing Well, is a must for teachers, editors, and anyone who cares about clear, effective communication.)

In Germany (and too many other places), academics still ride the dead horse of prose which is purposefully obscure, which has to be dissected and reassembled before it can be understood, presumably to show how smart an author is. Usually, though, it simply reveals a sloppiness of thought, a failure to crystallize ideas into a pure and simple form, and no concern for readers who are short on time. It requires way too much effort in decoding texts that often have little to say in the first place. The idea that complex ideas can only be expressed in complex sentences is a myth; just check out the banquet speeches given by Nobel prize-winners. Or Albert Einstein’s maxim, “Everything should be as simple as it can be, but no simpler.”

I often begin my classes with a citation from the book And the Band Played On, an account of the early days of AIDS research written by Randy Shilts. The text comes from a press conference held in June 1982, at a time before the discovery of the virus, as the Centers for Disease Control released its first findings on the epidemiology of the disease. A reporter asked CDC Director James Curran whether AIDS was a sexually transmittable disease. He gave the following answer:

“The existence of a cluster study provides evidence for an hypothesis that people in the study are not randomly associated with each other and the study is a sexual cluster. On the other hand, we don’t have enough scientific evidence to say for certain that one person gives it to another person. We have to focus much more research into this area so that we don’t prematurely release information that’s not validated. On the other hand, we’re not holding back any information that might provide important health benefits. Thank you.”

My students immediately recognize that this statement is scientifically correct; on the other hand, it is so obscure that reporters had to interpret it themselves (or turn to scientist friends for help in doing so). Interestingly, students immediately assume that its opaqueness was politically motivated – which, in fact, it probably was. If scientists feel this way, is it any wonder that the public often reacts the same way toward other scientific pronouncements they can’t understand (“If we don’t get it, they must be hiding something”)?

In the course we talk about the value of shaping and controlling a message – there are easy ways to explain cluster studies that show how strong the “sexual transmission” hypothesis is, vis-à-vis other possible interpretations of the data. Curran could have tailored the information to his audience and given an answer that would have sent a stronger message to the public at a time when people desperately needed information about a dangerous disease.

Rewriting Curran’s statement is a useful exercise, but we usually leave it to discuss what constitutes an effective communicative strategy overall. We usually arrive a single basic principle that, so far, has always gotten the point across. I ask the students to imagine attending a scientific talk, and as they leave, they meet someone outside the door who says, “Oh, I wanted to attend but just missed it – what did the speaker say?” Everyone who leaves the room should be able to give a short, sensible account of the story. Their versions should agree with each other, and they should also agree with what the speaker would say if asked the same question. If that happens, and if nobody passed out or died or spent the whole time answering e-mails because the speaker failed to hold his attention, then the talk must have fulfilled its function.

With this single criterion in hand, I tell students, “So imagine you’re giving the talk – why don’t you just hand-deliver the answer? Just put in a statement like, ‘Now, when you leave and someone asks you what this talk was about, here’s what you should say…'” Call it a take-home message, a conclusion, or whatever you like; this forces the student to reduce the content to a clear, sensible story that should be told in a way that can be remembered and repeated, no matter what type of audience is on hand, and it leaves the speaker in control of the message.

It’s a fine idea, but translating this simple principle into all the steps of preparing and presenting a talk, a poster, or a text usually requires intensive practice. I’ll discuss some of the strategies we use in the next entry.


Zinsser, William. On Writing Well, 30th Anniversary Edition: The Classic Guide to Writing Nonfiction. New York: Harper Paperbacks, 2006.

Shilts, Randy. And the Band Played On: Politics, People, and the AIDS epidemic. New York: St Martin’s Griffin, 2007

Manipulating a matrix of pain

Molecules that bind skin cells together influence the transmission of pain

“Where does it hurt?” a doctor says, and it sounds like a simple question; a patient is supposed to point to the offending region of the body. But a look at the details of how sensations arise reveals that the question isn’t so straightforward: pain begins with a response from nerves near the site of an injury. For the brain to perceive it, those neurons must generate electrical impulses that travel to other nerves in the spinal column and on to the brain. The impulse – and the sensations it causes – can be blocked at many places along the way. Now Gary Lewin’s group at the MDC has discovered that some signals are interrupted right at the source. A matrix of proteins that bind cells together in the skin can interfere with the contact between neurons and dampen touch sensation. Without this mechanism, people suffer from severe pain in a condition called epidermolysis bullosa, sometimes termed butterfly disease. The lab’s findings, published in the July 3 issue of Nature Neuroscience, reveal that a matrix protein called laminin-332 helps tune down sensations of pain and touch.

The skin needs to be tightly sealed to protect the body from water and dangerous substances in the environment. The seal consists of a dense matrix of glue-like proteins that stick cells to each other, including laminin-332. Sensory nerves end in the extracellular space in the skin and are surrounded by the same glue-like matrix of proteins. Brushing or probing the skin leads to an electrical excitation of these endings in their matrix-glue.

Normally, proteins in the matrix are required for an efficient transfer of the stress and strain produced by skin movement to the sensory endings in the matrix. The Lewin group has found that one component of the matrix normally acts to dampen down or inhibit the mechanical activation of the sensory endings in the skin. Laminin-332 was identified by two scientists in the Lewin lab, Li-Yang Chiang and Kate Poole, as a protein which naturally brakes the initiation of sensory signals that lead to pain. In human patients who lack this protein, the same sensory endings can be activated by normal stroking and probing of the skin, thereby amplifying pain.

As nerves grow, their tips branch many times to establish contact with other cells. If the endings of sensory branch excessively, the result will be amplified response to touch and pressure. The scientists also discovered that the presence of laminin-332 helps prevent too much branching in the skin.

The skin is made up of layers consisting of several types of cells, each of which is bound to its neighbors in different ways. This creates environments that develop different types of nerve architecture. One goal of the recent work was to show how the differences affect the transmission of sensory information.

“Pain-detecting nociceptors extend to the surface of the skin, into the layer called the epidermis,” Gary says. “Mechanosensory cells that detect touch reside exclusively in a lower layer, the dermis. Li-Yang and Kate found that laminin-332 blocks or lowers transmission from the nociceptor cells as they extend into the epidermis.”

In other words, if the keratinocyte cells that make up the dermis lack a working version of laminin-332, more nerves will be stimulated and will transmit stronger signals.

“People who suffer from epidermolysis bullosa don’t have laminin-332 – or they have a version of the molecule that doesn’t function well,” Gary says. “This work offers a mechanistic explanation for the intense pain that they experience.”

– Russ Hodge

A roadblock for metastases

A drug used to treat patients with tapeworms may help fight deadly colon cancer

Caught at an early stage, many cases of colon cancer come with a fairly good prognosis: 90 percent of patients survive at least five years. But if the diagnosis comes too late, after metastases have formed, the numbers flip. Only about ten percent of such patients reach the five-year mark, making this one of the most frequent causes of cancer death worldwide. Researchers are actively searching for ways of detecting the disease before it becomes deadly and new forms of treatment. A few years ago, Ulrike Stein’s group at the MDC and Charité discovered that high levels of a protein called S1004A was a good indicator of whether tumor cells were likely to migrate through the body and form metastases. Now the scientists have found that treating mice with a substance commonly used to rid the body of tapeworms reduces the likelihood that this will happen and improves the prognosis for the animals. They hope that the discovery can now lead to an effective treatment in human patients. Their work was reported in the July 6 edition of the Journal of the National Cancer Institute.

“In aggressive colon cancer and many other types of tumors, cells produce up to 60 times the normal amount of S100A4 protein,” Ulrike says. “The reason lies with the disruption of a biochemical signaling pathway within tumor cells. This route, called the beta-catenin signaling pathway, frequently becomes far too active during cancer. One result to strongly trigger the production of the S100A4 protein.”

While S100A4 itself does not cause cancer, it has deadly effects on animals already susceptible to tumors. When mice that produce too much of the protein are crossed with another strain that has a high rate of cancer, tumors spread like wildfire. On the other hand, if mice that don’t produce the protein are injected with highly metastatic breast cancer cells, the tumors don’t metastasize.

Over the past few years, researchers have discovered some of the reasons: S100A4 helps cells loosen ties to their neighbors and promotes their migration through the body and the formation of new colonies of cells. These processes are vital in the formation of tissues and organs during the development of embryos, but it becomes deadly during cancer.

Until now, Ulrike says, scientists haven’t found a substance that can block the expression and thereby the metastatic potential of S100A4. The current study changes that situation. During a research stay at the National Cancer Institute-Frederick in Maryland, USA, Ulrike and Wolfgang Walther carried out a high-throughput screen of 1280 small molecules with Robert Shoemaker’s group, to find something that could inhibit the protein’s production. The most powerful effect came from a drug called niclosamide, which is already used to treat tapeworm infections in human patients.

With this information, postdoc Ulrike Sack and her colleagues in Ulrike Stein’s group first investigated the effects of niclosamide on the behavior of cells in the test tube, including human cancer cells. They discovered that the treatment had a strong impact on cell migration and their ability to form colonies – two steps which are essential to metastases.

Next the scientists turned to mice, using a well-established method of studying metastases. They injected tumor cells into the spleens of control animals and another group that was treated with niclosamide. All the animals developed tumors in their spleens. In the control mice, metastases spread to the liver within just eight days.

But the livers of the treated animals remained almost metastases-free. The size and the number of the metastases that developed in their livers were significantly reduced, and the animals survived about twice as long as their control counterparts. This was true whether they were given the drug for just a few days after injection of the tumor cells or over the entire course of the disease.

The scientists discovered that niclosamide specifically blocks production of high amounts of S100A4 and its metastatic effects. “We’re lucky that the drug has already been thoroughly tested in human patients,” Ulrike Stein says. “That is normally a huge bottleneck in the development of drugs. So we’re hopeful that we can quickly move to human trials, to study the effects of niclosamide on deadly forms of colon cancer.”

– Russ Hodge