Having just defended his thesis at the end of May, MIT’s Joe Shuga has even more to celebrate: he is first author on a recent PNAS paper that might take a common in vivo drug toxicity test and make it both high-throughput and human-specific. When it comes to the promise of pharmacogenomics, those words are music to the ears.
Affiliated with three PIs — Linda Griffith in the department of biological engineering, Harvey Lodish of the Whitehead Institute, and Leona Samson, director of the Center for Environmental Health Sciences — Shuga optimized a common drug toxicity test that is usually performed on mice in vivo into one that can be conducted in vitro on cultured adult mouse bone marrow cells. The test might be not only cheaper, but could also show promise of being more effective in the preclinical toxicity testing process.
“Normally before you start clinical trials, you’re required to conduct this in vivo micronucleus test in mice,” Shuga says. “So basically what we did is create an in vitro correlate to that.”
The preclinical, FDA-approved test that he and his colleagues tweaked is the in vivo micronucleus (MN) genotoxicity assay, which looks at red blood cells in mice to detect visible signs of DNA damage during cell formation. During normal erythropoiesis, a red blood cell formed in the bone marrow will end up expelling its nucleus. However, if DNA damage occurs, the reticulocyte, or budding red cell, will retain small micronuclei, or fragments of nuclear membrane-bound DNA.
“What the test actually looks for is this separate part of the DNA that has resulted from breakage or chromosomal loss,” Shuga says, “and that’s basically what you’re testing for here: … Is the substance genotoxic?”
Utilizing expertise from all three labs, Shuga optimized an in vitro protocol that could successfully simulate erythropoiesis in adult mouse bone marrow cells. He then tested whether exposing the culture to three different genotoxic substances would have the same effect on reticulocyte formation as did dosing a mouse. In both cases, the frequency of the micronucleated reticulocytes went up with genotoxic exposure. The team found that in mouse bone marrow mutated to lack a key DNA repair protein, there was increased MN frequency as well as decreased reticulocytes when exposed to a genotoxic agent both in vivo and in vitro.
The motivation behind developing in vitro systems, something that Griffith’s lab has been working on, is to make the whole process of preclinical toxicity testing more efficient and effective. One future direction is extending this testing to many toxic agents at once. A key advantage to an in vitro test is that the bone marrow can be extracted and exposed, rather than dosing the entire mouse; so more than one test can be done using bone marrow from a single mouse. Now, for each genotoxic agent tested, about 2,000 reticulocytes are extracted from the mouse to analyze, says Shuga. With a culture system, they can move beyond one agent, and test as many as 2,000 individual cultures. “Normally, you get one data point per mouse,” he says. “But there’s a lot more reticulocytes in the mouse than you look at.”
At the end of the day, there are still 2,000 slides to analyze, though, so Shuga hopes that using high-throughput analytical techniques such as flow cytometry and laser scanning cytometry will make the process more efficient. “One of the key future directions is to mesh those two technologies,” he says. “Take the cultures that we’re conducting our tests in and have them feed into a high-throughput analytical device so that you can realize the high-throughput potential here, the thousands of conditions per mouse.”
Another direction is moving toward using human bone marrow, though finding a reliable source of healthy cells might be an initial obstacle, Shuga says. “If you extended the test to human bone marrow, then you might be able to predict human response better,” he says.