[HTML][HTML] Groundhog day for rodent models of acute lung injury: clear relevance or renewed debate?

BB Moore - American Journal of Respiratory Cell and Molecular …, 2017 - atsjournals.org
American Journal of Respiratory Cell and Molecular Biology, 2017atsjournals.org
Lung injury and acute respiratory distress syndrome (ARDS), caused by either direct or
indirect insults, are vexing clinical problems with poor patient outcomes (reviewed in
Reference 1). The pathophysiologies of these diseases are poorly understood, making it
difficult to identify promising therapeutic targets. To this day, no pharmacologic treatments
have proven effective, and patients are managed with supportive care and lung-protective
ventilation strategies (2). Over the years, biomedical researchers have turned to rodent …
Lung injury and acute respiratory distress syndrome (ARDS), caused by either direct or indirect insults, are vexing clinical problems with poor patient outcomes (reviewed in Reference 1). The pathophysiologies of these diseases are poorly understood, making it difficult to identify promising therapeutic targets. To this day, no pharmacologic treatments have proven effective, and patients are managed with supportive care and lung-protective ventilation strategies (2). Over the years, biomedical researchers have turned to rodent models to try to recapitulate the host response to lung injury, sepsis, and trauma in an attempt to identify pathologic mechanisms and potential therapies. With the advent of genomic technologies, investigators have shown great interest in determining whether the transcriptional responses seen in murine models of acute lung injury and sepsis recapitulate the responses noted in humans suffering from these disorders, with the hope of identifying common mechanisms to exploit for therapeutic gain. However, in many cases these studies have led to more confusion than clarity. In one particularly well-publicized study in 2013, Seok and colleagues compared the gene-expression profiles of humans and mice subjected to various insults, including burn, trauma, and endotoxemia, and concluded that “genomic responses in mouse models poorly mimic human inflammatory diseases”(3). These conclusions set off a flurry of editorials and responses in both the scientific and lay press that questioned the relevance and merits of rodent modeling of human disease. Summarizing the debate, Osuchowski and colleagues thoughtfully discussed potential confounders of human versus murine modeling (4). Adding even more complexity, in a study reevaluating the same data sets used in the Seok study, but focusing only on genes whose expression was significantly changed, Takao and Miyakawa reached the opposite conclusion and noted that “gene expression levels in the mouse models showed extraordinarily significant correlations with those of the human conditions”(5).
Into this fray enters a new study by Sweeney and colleagues (pp. 184–192) in this issue of the Journal that once again addresses the relevance of rodent models to human lung-injury pathways (6). Taking advantage of publically available gene-expression data sets, Sweeney and colleagues examined 21 data sets from animal lung tissue and three data sets from human lung-injury bronchoalveolar lavage (BAL). Of interest, they examined gene-expression changes from studies in both mice and rats across quite distinct modes of lung injury (ozone exposure [n= 3], ventilator-induced lung injury [n= 14], intratracheal lipopolysaccharide treatment [n= 3], and hemorrhagic trauma [n= 1]) in comparison with human BAL data sets representing experimental lipopolysaccharide or ozone exposure. The authors analyzed the data sets using sophisticated, integrated multi-cohort and all-models meta-analyses to show that common signals could be identified across all data sets. They found 50 genes common to rodent and human lung-injury models that were significantly differentially expressed, lending support for the notion that common pathways contribute to diverse lung-injury responses
ATS Journals