JANUARY 15, 2015 COVER STORY TOOLS & TECHNIQUES 1 THE RIGHT CULTURE Stakeholders agree that mandating new methods to authenticate cell lines can improve reproducibility in THE RIGHT CULTURE research, but they don’t agree on who should lead the charge. By Michael J. Haas, Associate Editor Using the right cell line might sound like a topic for Lab Work 101, but PRODUCT R&D it is an issue at the heart of the reproducibility problem in preclinical research. A panel that met to discuss strategies for tackling the problem 6 EPIZYME’S MANTLE PIECE agreed that some emerging technologies for authenticating cell lines Epizyme adds PRMT5 to the list of epigenetic targets could make a difference, but found little common ground on which and believes it has a first-in-class inhibitor of the entities — journals, funding agencies or the private sector — should enzyme. be responsible for mandating their use. The panel was convened by the Global Biological Standards Institute TOOLS & TECHNIQUES in November as part of its inaugural BioPolicy Summit and focused on 9 BOOSTING IgG WITH IgA cell line authentication — a key factor identified in the institute’s 2013 A UT Austin team has engineered chimeric mAbs white paper on the need for standards in many areas of preclinical that induce more potent leukocyte-mediated killing of research. targeted cells than conventional IgGs. GBSI President Len Freedman told BioCentury that although several issues were outlined in that analysis, the institute chose to start with FINANCE this topic “because it is a clearly identifiable problem with known effects on results in research.” 12 CONSORTIA CROSS-TALK To include a cross-section of stakeholders in the discussion, GBSI drew Consortia are wising up to the need to avoid the four panelists from Roche’s Genentech Inc. unit, NIH’s National duplication of efforts and starting to coordinate their activities, but still need to bring more academics into Institute of General Medical Sciences (NIGMS), the Prostate Cancer the fold. Foundation (PCF) and Nature Publishing Group (NPG). Multiple studies have reported that up to one third of all cell lines used TRANSLATION IN BRIEF in preclinical research are misidentified — a problem that stems from lack of best handling practices and results in cross-contamination of 18 TRUTH IN A CELL cell lines. Cell line misidentification actually involves multiple issues, A single cell partnership for Fluidigm, Wellcome and including whether a cell line was correctly identified before use, EBI; allergy center in Stanford; Takeda joins Monash whether that line becomes contaminated or otherwise altered during for GI discovery experimentation, and whether multiple daughter cell lines generated from one cell line retain identity with one another and with their DISTILLERY parent line. 21 THE WEEK IN THERAPEUTICS However, the panelists said despite growing awareness of the issue, many researchers still don’t validate their cell lines. BNIP3 inhibitors for doxorubicin-induced cardiotoxicity; malaria vaccine from an engineered “Cell lines that were tagged as misidentified back in the 1960s are still parasite; NRTIs to treat dry AMD; and more... used” under their wrong identity by some researchers, said panelist Véronique Kiermer, executive editor and director of author and 27 THE WEEK IN TECHNIQUES reviewer services at NPG. “The consequences of using misidentified Cell-free detection of CTCs in blood; rare mutations cells can be very different depending on the study being done.” for predicting MI risk; 3D printing of meniscus tissue; and more... PRODUCT R&D TOOLS & TECHNIQUES FINANCE TRANSLATION IN BRIEF DISTILLERY “We need to set new standards to increase the integrity of the science we do from now on.” Richard Neve, Genentech Inc. Panelist Richard Neve, director of discovery oncology at Neve told BioCentury that to use the technique “at the level you Genentech, noted that although the implications for science should” was too time-consuming and costly even for Genentech. already done would never be fully known, the emphasis now “This is why we went with SNP-based profiling over STR, which should be forward looking. “We need to set new standards to could cost us tens of thousands of dollars annually.” increase the integrity of the science we do from now on,” he told In 2008, Genentech launched an in-house facility that banks the panel. all cell lines coming into the company and authenticates them Jon Lorsch, director of NIGMS, said his institute recognizes the before and after use by company researchers with a method — problem as a serious one and is implementing its own strategy also developed in-house — that utilizes 48 SNPs. “The initial for addressing it at the source of funding. “We’re formulating cost of reagents is about $6 per sample for the SNP-based new guidelines and procedures and asking grant applicants to method and $15-$30 per sample for STR, depending on which have a statement about authenticating cell lines,” he said. kit you use,” Neve told Biocentury. “The SNP profiling also cuts Howard Soule, CSO of PCF, indicated the not-for-profit sector the run-time by a factor of 4 or 5 over STR analysis.” is taking action as well. “We take this issue very seriously as a “We’ve not looked at whether we see more reproducibility, but responsibility to our donors,” he said. His organization is also we do catch mistakes earlier,” Neve told the panel. “We can have tying it to funding, and about a year ago began requiring its more confidence moving forward that the data being produced researchers to show cell line authentication results within one will be reproducible.” year of receiving grant money. Lorsch called Genentech’s approach “a step in the right direction,” but was not convinced it would provide a comprehensive STR-AIGHT TALK solution because not every researcher has the resources or All four panelists pointed to short tandem repeat (STR) motivation to set up an SNP-based authentication facility. analysis as the best available — if not the optimum — tool for authenticating cell lines. The approach involves profiling “We want to make it more feasible for researchers to do cell line hypervariable regions in DNA that are unique to an individual authentications in terms of the time, money and number of cell genome, which provides a distinct genetic fingerprint for any lines tested” — making it as easy as using a pH meter in a lab — human cell line. and NIH is considering how to encourage development of new technologies in this area, he told the panel. Many CROs and service companies now offer STR analysis to customers for about $150 per cell line with a turnaround time Lorsch also told BioCentury that new technologies are needed of about a week. But according to Kiermer, data collected by because STR lacks the genomic resolution needed to authenticate NPG from manuscript submissions show that only about 10% gene-edited cell lines. “For instance, it can’t tell you whether of researchers use STR analysis. She thinks that as its costs your HeLa cell line is the same as mine. So it would be useful to continue to decrease, the technique will become routine. have a rapid, cost-effective technique that could do this.” In the meantime, Lorsch told BioCentury, the cost of STR Keith Yamamoto, keynote speaker at the GBSI summit, agreed analysis presents researchers with a significant barrier to that STR analysis is limited in its ability to identify cell lines, compliance. “For example, if you have eight or nine people on especially in light of gene-editing techniques that are becoming your team working with a dozen cell lines and you’re doing STR more common. “In my view it would be more useful to have a on all of them every few months, this adds up,” he said. broader discussion on whether STR is really useful and how to develop new assays to identify cell lines,” he told BioCentury. 2 JanuarY 15, 2015 TOC PRODUCT R&D TOOLS & TECHNIQUES FINANCE TRANSLATION IN BRIEF DISTILLERY TOOLS & TECHNIQUES “I think cell lines will eventually have to be bar-coded.” Yamamoto is vice chancellor for research at JOURNALS TAKE ON REPRODUCIBILITY the University of California, San Francisco, and executive vice dean at the UCSF School While any one journal cannot single-handedly mandate best practices to the of Medicine. entire research community, together they can promote consensus standards to improve the reproducibility of published studies. Late last year, NIH released a POLICE QUESTION set of proposed principles and guidelines for reporting preclinical data that it put There was less harmony among panelists together after discussions with leaders of more than 100 life sciences journals. on who should shoulder the responsibility The guidelines were the result of about two years of work, according to Véronique for mandating standards or policing Kiermer, executive editor and director of author and reviewer services at Nature compliance. Publishing Group. As concern over irreproducibility in preclinical research rose in 2012, several NIH institutes — including the National Cancer Institute and the Kiermer’s position was that journals can National Institute of Neurological Disorders and Stroke — convened meetings impose standards for publication that might for scientists and journal editors to discuss the problem in the second half of that provide an incentive for better behaviors, but year, she told BioCentury. journals cannot be responsible for ensuring that researchers act as they should. That Following those meetings, some journals began instituting measures to encourage researchers to adhere to best practices while others augmented their requires other stakeholders, she said. existing measures or hung back waiting for further developments, said Kiermer. “It’s a community issue about good lab NPG, for example, created an 18-point checklist for submitters that requests management and best practices — like details about the study, including methods of statistical analysis, the source of asking a cook to use the freshest ingredients,” cell lines used, and whether the cell lines and reagents used had been validated.