Deepak Srivastava

About three years ago, members of Deepak Srivastava's lab were pursuing separate projects on two microRNAs, which are short molecules that regulate genes. Some researchers were trying to figure out the role of microRNA-145, which had turned up in screens of early cardiac progenitor cells. Others were looking at microRNA-143, which is highly conserved and more abundant than any other microRNA in mouse embryonic stem cells that are differentiating into cardiac progenitors.

Though they didn't know it at the time, these researchers at the Gladstone Institute of Cardiovascular Disease in San Francisco were on their way to uncovering a single microRNA (miRNA) that efficiently shifts multipotent cells into another lineage — in this case vascular smooth muscle cells (VSMCs). Both miRNAs regulate and are regulated by proteins that cause the differentiation and proliferation of smooth muscle cells. Not only could this have applications for treating heart disease or developing technologies for reprogramming cells, but also the study provides an unusually complete picture of how miRNAs can regulate cell fate.

The data from the separate miRNA projects kept showing odd, striking similarities, and both miRNAs seemed to be in the same place at the same time. The only thing that made sense was that they were being transcribed together. But try as they might, the researchers could not identify any transcripts containing sequences for both miRNA-143 and miRNA-145. Kimberly Cordes, then a graduate student in the lab, had a hunch that a single primary transcript did in fact exist but was rapidly processed into mature, separate RNAs. She tried hunting for the primary transcript using another strategy: embryonic stem cells that lacked a gene essential for miRNA production. She found the transcript.

Now the researchers knew that both miRNAs were controlled by the same regulatory elements, but they didn't know what those were. Thus began what Srivastava calls a “frustrating, brute-force search”. They deleted large swaths of the 4,200-nucleotide enhancer region until they found the 900 bases that really mattered. A quick scan of the sequence showed likely binding sites for cardiac transcription factors. One of these, serum transcription factor (SRF), works alongside a protein called myocardin as part of a molecular switch for turning multipotent cells into VSMCs. Myocardin can also reprogram fibroblasts into VSMCs.

To see if either miRNA-143 or miRNA-145 were part of that reprogramming process, the researchers knocked them down in fibroblasts with antisense oligonucleotides. When miRNA-143 was inhibited, the researchers saw plenty of VSMCs. When they tried the same with miRNA-145, they saw only fibroblasts.

“About a year ago, when we made the discovery that miRNA-145 was required for myocardin to reprogram a cell, that's the moment that stands out the most,” says Srivastava. “We didn't understand the mechanisms, but [when you see something like that] you know you have a major revelation on hand.”

The clinching evidence for this revelation did not come quickly. Cordes knew she was getting close to a key result, and she wanted to have it in hand before the lab presented at one of Gladstone's scientific advisory meetings. “She worked day and night to make that happen.”

The next steps involved figuring out what genes miRNA-143 and miRNA-145 regulate, another laborious process. It turns out miRNA-143 is no wallflower, even if it has little effect on turning cells into smooth muscle. It represses Elk-1, a protein that competes with myocardin to interact with SRF, switching the protein from one that promotes differentiation to one that promotes proliferation. When miRNA-143 was inhibited, cultures of VSMCs doubled at twice the normal rate.

Interestingly, miRNA-145 inhibits expression of several proteins important for smooth-muscle proliferation. This includes Klf4, one of the four proteins that Shinya Yamanaka used to transform fibroblasts into induced pluripotent stem (iPS) cells.

Although miRNA-145 was recently reported to cause iPS and embryonic stem cells to differentiate in nonspecific, heterogeneous ways, Srivastava thought that, in the right cell population, the miRNA alone might be enough to shift cells to smooth muscle. Attempts with fibroblasts and early heart populations failed. “No one in the lab really thought it would work,” Srivastava recalls. Then Emily Berry, a rotation student in the lab, and Neil Sheehy, another graduate student, decided to try the experiment on a population of neural crest cells found within the developing heart. “Within 24 hours [the miRNA] had converted 75% of the cells into smooth muscle.”

Srivastava thinks this might be the most remarkable effect of any miRNA published to date. “How often does a rotation student get to contribute that way?” he asks. Berry will be joining the lab full time this summer.