The Use of Cell Lines to “Model” Ocular Tissues: Cautionary Tales

In this issue of IOVS, a Perspective by Krishnamoorthy and colleagues ¹ provides an informative “forensic path” to help us understand one of the many things that can go wrong when using transformed or immortalized cell lines to investigate the physiology, pathology, and response to therapy of ocular cells. They show that a cell line used for years to study the biology of retinal ganglion cells, RGC-5, is from a different retinal cell type. ¹  

A warning about these cells was published in IOVS several years ago from the laboratory of Jonathan Crowston, noting that the RGC-5 cells in their laboratory did not have the expected response to glutamate and were of mouse, not rat, origin. ² The authors concluded their abstract with the admonition, “Investigators using cells designated as RGC-5 should confirm the species to be of rat origin and retinal-specific marker expression before considering their use as retinal ganglion-like cells.” Because Krishnamoorthy and colleagues ¹ had access to very early passages of “RGC-5 cells,” we now know that investigators are unlikely to find any rat RGC-5 cells that meet the standard set by Crowston and colleagues. The original RGC-5 line has been replaced with a line of mouse photoreceptor–derived cells. It appears that “authentic” RGC-5 cells no longer exist and have not existed for more than a decade. Therefore, the results of most of the more than 220 articles published using this cell line are unlikely to reflect the biology of retinal ganglion cells. The “lessons learned” from this sad tale are detailed at the end of the article, with a guide to assist those receiving cell lines from other laboratories to confirm their authenticity. 

In addition to cell identity, there is an additional critical issue that should be considered when using immortalized or transformed cell lines, like ARPE-19, HLE-B3, HTM4, or transformed corneal epithelial cells lines. Do the cells accurately reflect the behavior of the tissue from which they were derived? For cells to exist in vitro, they must have at least one characteristic that is distinct from their in vivo relatives; they must proliferate… forever. Proliferation is not part of the adult phenotype of RPE, retina, lens, or trabecular meshwork cells and the differentiation of corneal epithelial cells is associated with cessation of proliferation. Worse yet, cells with faster proliferation will be favored by the unnatural selective pressure of in vitro culture. Cells adapted to the in vitro environment are likely to lose characteristics that were critical to their function in vivo. Just looking at some of these cell lines under a phase-contrast microscope demonstrates that diagnostic features, such as cell shape, cell–cell junctions, and pigmentation, are different from the cells they are meant to represent. Therefore, when possible, results obtained using cell lines should be confirmed using cultured primary tissues or in vivo. Studies that “go the extra distance” to confirm the results of in vitro studies will generally receive a more favorable assessment at IOVS. 

I recognize that cell lines often permit research to be done more easily and at lower cost than experiments that use primary tissues or are done in vivo. However, studies that come to an incorrect conclusion and mislead future investigators are much more costly in time and money.