Commentary on "incidence and nature of testicular toxicity findings…".
K. Boekelheide
DOI: https://doi.org/10.1002/bdrb.20319
2011-12-01
Abstract:Testicular toxicity is an adverse endpoint of significant concern in preclinical studies conducted during drug development, as documented in the companion article (Incidence and Nature of Testiculary). This commentary gives my perspective on what to do with a preclinical testicular toxicity signal, focusing on some of the critical points of concern, with a particular emphasis on the role of toxicogenomics in the decision-making process. The testis is composed of paracrine-interacting cells of four major types: germ, Leydig, peritubular, and Sertoli. Each of these cell types may be a target for toxicity. Sertoli cells, known as ‘‘nurse cells,’’ are particularly important in orchestrating and supporting the germ cells, yet they represent only 3% of the population of cells within the adult testis. This is a significant experimental problem: How do you observe cell-type-specific changes in such a minor population? Furthermore, Sertoli cells are intricately intertwined with germ cells and physically quite close to peritubular cells within the intact testis. Germ cells are also quite complicated, because over a period of B8 weeks (in the rat), they emerge from their stem cell niche and undergo remarkable proliferation and morphological transformation. During this process of maturation and differentiation, germ cells have a continuously changing cellular composition and gene expression profile. During the last 50 years, experimentalists have recognized this problem and tried to separate and culture-enriched testicular cell populations in vitro. Many publications have resulted from these efforts (including some from my lab), even though separating the testicular cellular components and placing them in an artificial in vitro environment can markedly disrupts their underlying physiology. Although in vitro techniques may be quite useful for some testicular cells and endpoints (Leydig cell steroidogenesis, for example), the limitations are often insurmountable. Sertoli cells, usually cultured as primary cells isolated from peri-pubertal animals, do not display many of the differentiated features characteristic of in situ adult Sertoli cells. Because of this lack of differentiation, in vitro Sertoli cells often do not respond to testicular toxicants in the same way as they would if they were in situ. For germ cells, while many claims of success have been made, recreating spermatogenesis in vitro remains an illusory goal, presumably because of an inability to adequately recreate the required paracrine microenvironments. This means that if you are trying to study a Sertoli or germ cell problem (the most common testicular toxicities), the molecular characteristics of these cells that make them targets for toxicity are frequently not present in the in vitro setting. Therefore, while in vitro screening for testicular toxicity will undoubtedly ultimately be useful, for now the development of appropriate testicular in vitro models is largely an academic research effort. LESSON x1: IN VITRO TECHNIQUES ARE OF LIMITED UTILITY. This lesson conforms to the experience of pharmaceutical companies as indicated in the accompanying survey (1). As noted, only 4 of 16 respondents to the survey reported the use of in vitro techniques with none reporting the use of these methodologies for screening or mechanistic studies. The discussion correctly identifies a need for ‘‘some in vitro methods which could be used to either investigate mechanism or to screen compounds,’’ although the significant limitations of current methods, which ‘‘fall well short,’’ are noted. When testicular toxicity is observed in a preclinical animal study, an issue of primary concern is whether the lesion is related or unrelated to the known pharmacology of the drug candidate (‘‘on-target’’ vs. ‘‘off-target’’ effect). This issue is usually addressed using traditional techniques for detecting the presence of the target in testis, and for measuring the known cellular consequences of activation or inhibition of this target. If the testicular toxicity is due to target-related pharmacology in an animal model, then it is important to know whether the human testis functions with the same underlying biology, and whether the anticipated therapeutic dose is of concern. Although the studies to resolve these questions may become quite complex, such on-target toxicity can usually be managed because the underlying mechanistic basis for the effect is understood. Toxicogenomic techniques, primarily mRNA microarrays, can play a significant role in this context, allowing for the detection of the target mRNA, and for the observation, during a time course of exposure, of the characteristic on-target effects of drug candidate exposure. However, observing changes occurring in only a minor cell population can be a real problem. Using the example of a Sertoli-cellspecific toxicant, discerning mRNA signals of interest while examining whole testis mRNA may be difficult because of overlapping expression by multiple cells types, dilution, and noise. For mRNAs that are expressed only in the Sertoli cell, this may not be much of a problem; however, for pathway analysis involving canonical shared components, the results will need to be interpreted with caution. Immunohistochemical