CHICAGO—The "Triple A-S" meeting is like no other scientific gathering in that it is not really for scientists but for journalists who follow science. Scientists do come to present talks or to serve on AAAS governing sections, but to understand the meeting's central purpose, think of AAAS as the world's largest annual science press conference.
In recent years as the Internet and the Great Recession smashed through the traditional news media, there has been some question of who would turn up to cover science. Yet AAAS 2014 showed that reports of the death of the science "press" have been greatly exaggerated. Although only a few seemed to have any connection with a printing press, science communication practitioners by the hundreds—bloggers, tweeters, content generators, streamers, site curators, long formers, casters, and posters—flocked to AAAS in Chicago. Five days of workshops, talks, and symposia ranged from dark matter to fine art conservation at the nanoscale to nuclear non-proliferation to "Santa's Revenge: The Impacts of Arctic Warming on the Mid-Latitudes" to Alan Alda (as plenary speaker, not as research topic).
Still there wasn't enough cell and molecular biology. Given the breadth of the AAAS program, the shortage might be understandable but "Biology" was stuck in with "Neuroscience" as a symposium track, and only one program stood out as the real cellular deal—a panel promising an update on the field of non-coding RNAs. There the weather gods played a cruel trick on the non-coding RNAs, leaving Chicago cold but clear while burying the Atlantic Coast in snow and ice. Airport havoc caused three of the five "non-coding" panelists to stay home, leaving Joshua Mendell of the University of Texas Southwestern Medical Center and Arun Chinnaiyan of the University of Michigan, Ann Arbor, to carry the non-coding microRNA torch.
The AAAS Annual Meeting treads a fine line between providing "news" and scientific context for the journalists. Whole fields—"cultural heritage science" (a.k.a. high-tech art conservation) or microRNAs—could be startling news to many science writers at AAAS. But reporters/science writers are conditioned to seek "newsy" news—breakthroughs, emerging threats, and, of course, cures. AAAS panels try to balance "breaking" news against the reality that major scientific discoveries don't pop out of some wizard's forehead once a day.
Non-coding microRNAs are not really news in that sense. They first popped up in the literature in the early 1990s, Mendell said. MicroRNAs (miRNAs) are short, single-stranded RNA molecules that we now know silence genes by targeting messenger RNA (mRNA) molecules and causing the mRNAs to be degraded before they can be translated to a protein. Mendell explained that miRNAs were discovered through classic genetic mutant screening in Caenorhabditis elegans, first by Victor Ambros and then by Gary Ruvkun, who were postdocs together in Bob Horvitz's famous C. elegans lab at MIT. With Ambros at the University of Massachusetts Medical School and Ruvkun at Harvard Medical School, the two were parallel collaborators, exploring different ends of the emerging miRNA story. It was their "back to back" papers in the same 1993 issue of Cell that put non-coding miRNAs on the map. In 1993, it was a very small map—a worm-sized one. These "small RNAs with antisense complementarity" were in the parlance of the day "junk" DNA because they did not code for proteins but rather for short nucleotide transcripts. Ambros and Ruvkun demonstrated that instead of being "junk," these short strings of nucleotides were, in fact, systematically regulating the expression of other genes at the mRNA level. Many were found to be active in regulating larval development in C. elegans.
Mendell brought the story rapidly up to date. miRNAs turned out to be highly conserved in nearly all eukaryotic species. Since the completion of the Human Genome Project in 2000, hundreds of miRNAs have been identified in humans by matching analogous sequences to those first discovered in worms. Mendell estimated that in people, there are 300-500 miRNAs significantly involved in regulation of everything from X-chromosome inactivation to Alzheimer's disease. The first miRNAs discovered in C. elegans were critical in larval differentiation, but the consensus now is that development is a rarer sphere of operations for miRNAs. The vast majority of miRNAs, Mendell said, are responses to stress in cells, physiological or genomic. This points to a major role for miRNAs in disease. Today miRNAs have been studied in virtually all human diseases, but the link to cancer is the most compelling, according to Mendell. Studies of miRNA levels in a variety of tumors have shown abnormal levels of activity, and those abnormalities are correlated to response to treatments and clinical outcomes.
Whether it is causative or symptomatic, cancer cells co-opt specific miRNAs, manipulating them to drive tumor progression by turning off suppressor genes or releasing oncogenes. Mendell's lab has studied the co-option of miRNAs in several major cancer pathways including three of the most notorious—the oncogenes Myc and KRAS and the tumor suppressor p53. The question is now whether miRNAs could be therapeutic agents in their own right. Mendell described his lab's recent work in a mouse model of liver cancer where using a virus vector to deliver miR-122, they rescued function and stopped tumor progression. This is exciting work, but for journalists it tends to fall into the classic category of things that haven't cured cancer yet.
Something that didn't promise to cure cancer immediately but could have a direct impact on men's health was Arun Chinnaiyan's presentation on using long non-coding RNAs (lncRNAs) as a biological marker for sorting out "indolent" from aggressive prostate cancer. Indeed Chinnaiyan already has an agreement with a Canadian diagnostics firm, GenomeDx Bioscience (which he most properly disclosed at the outset of his talk), to explore his discovery that a long (meaning over 200 nucleotides in length) non-coding RNA called SChLAP1 (second chromosome locus associated with prostate-1) could be the basis for such a clinical test. This comes in the wake of one of the great examples of unintended consequences in modern diagnostics, the rise and fall of the PSA (prostate specific antigen) test. The problem is that most prostate cancer is asymptomatic for life, that is, most older men have cancerous cells in their prostate but the tumor never progresses. Only about 20% of all prostate cancers metastasize, but this aggressive form is often lethal. The National Cancer Institute says that 239,000 American men were diagnosed with some form of prostate cancer in 2013, while 30,000 died of it.
Back in the late 1980s, the PSA was supposed to be the game changer in early diagnosis of prostate cancer. After identification by this mass screening, "positive" individuals would be sent for clinical follow-up. First approved by the FDA in 1986, the PSA test did catch many metastatic cancers but it also channeled thousands of men with the indolent form into unnecessary and harmful surgeries, radiation, and drug treatments. In 2012, the U.S. Preventative Services Task Force issued a recommendation that the PSA should no longer be used for mass screening.
That reopened or redoubled the search for a more reliable prostate cancer test. Chinnaiyan's novel use of miRNAs for such a test mechanism has to be seen against a background of other approaches to the same end, seeking to identify key genes, antigen patterns, and other bio telltales.
But the promising aspect of the miRNA approach, as presented by Chinnaiyan, is its specificity for the metastatic form of prostate cancer. Chinnaiyan explained that SChLAP-1, a lncRNA active only in the cell nucleus, was identified in a search for PCATs (prostate cancer associated transcripts) by screening libraries of prostate cancer tissues. PCAT-114 was found at much higher levels in these tissues and, looking more closely, the researchers zeroed in on a segment containing the gene for SchLAP1. High levels of SChLAP-1 were associated with only 25% of all prostate cancers and yet those were more frequently from tumors identified by pathologists as metastatic. Working in vitro, Chinnaiyan's lab was able to demonstrate that up-regulating SChLAP-1 promotes cell invasion by prostate tumors. Knocking it down in a mouse model blocks metastasis. Moreover, Chinnaiyan believes that he has found the mechanism. He says that SChLAP-1 acts on a particular subunit of the SWI/SNF (SWITch/Sucrose Non Fermentable) nucleosome remodeling complex that wraps and unwraps DNA from its chromatin packaging in the nucleus. SWI/SNF is a known tumor suppressor, and if SChLAP-1 antagonizes its action it would remove a stop mechanism against tumor progression and metastasis. Levels of SChLAP-1 can be analyzed from urine samples, which would make screening that much easier.
So what should science writers make of this? All basic biomedical research news is a hard sell for a general audience, either because the nature of the discovery is so complex or because a news story about diagnosing metastatic prostate cancer should always stress the unpredictability of the bench-to-bedside pathway. (In this case, it would be from the bench to the urine specimen.) Yet beyond the possible impact of miRNAs on diagnostics, there is another problem here for writers. Success in basic research is largely visible only in retrospect. The 2013 Nobel Prize in Medicine or Physiology for Randy Schekman, James Rothman, and Thomas Südhof illustrates how unpredictable basic science can be. The 2013 Nobel committee cited Schekman's first paper on vesicle trafficking published in 1979. Rothman's first citation appeared in 1980, Südhof's in 1990. That's 34, 29, and 23 years of hindsight for one Nobel Prize.
In contrast, the big "biology" news from Chicago this month about long non-coding miRNAs is a mere 20 years old, going back to Ambros and Ruvkun. The good news is that miRNAs are just gathering steam.